diff --git a/.ci/generate-buildkite-pipeline-premerge b/.ci/generate-buildkite-pipeline-premerge index 9c6f5aefd6de0..f32eb7213b940 100755 --- a/.ci/generate-buildkite-pipeline-premerge +++ b/.ci/generate-buildkite-pipeline-premerge @@ -107,7 +107,7 @@ function add-dependencies() { done ;; compiler-rt|libc|openmp) - echo clang + echo clang lld ;; flang|lldb) for p in llvm clang; do diff --git a/.github/dependabot.yml b/.github/dependabot.yml new file mode 100644 index 0000000000000..8c1dfd39b82c4 --- /dev/null +++ b/.github/dependabot.yml @@ -0,0 +1,18 @@ +version: 2 +updates: + - package-ecosystem: "github-actions" + directory: "/" + schedule: + interval: "monthly" + groups: + github-actions: + patterns: + - "*" + - package-ecosystem: "pip" + directory: "/llvm/docs" + schedule: + interval: "monthly" + groups: + llvm-docs-requirements: + patterns: + - "*" diff --git a/.github/workflows/docs.yml b/.github/workflows/docs.yml index 0c8656ce6f6cb..c4b1ab9c43f75 100644 --- a/.github/workflows/docs.yml +++ b/.github/workflows/docs.yml @@ -138,7 +138,7 @@ jobs: - name: Build libcxx docs if: steps.docs-changed-subprojects.outputs.libcxx_any_changed == 'true' run: | - cmake -B libcxx-build -GNinja -DCMAKE_BUILD_TYPE=Release -DLLVM_ENABLE_RUNTIMES="libcxxabi;libcxx" -DLLVM_ENABLE_SPHINX=ON ./runtimes + cmake -B libcxx-build -GNinja -DCMAKE_BUILD_TYPE=Release -DLLVM_ENABLE_RUNTIMES="libcxxabi;libcxx;libunwind" -DLLVM_ENABLE_SPHINX=ON ./runtimes TZ=UTC ninja -C libcxx-build docs-libcxx-html - name: Build libc docs if: steps.docs-changed-subprojects.outputs.libc_any_changed == 'true' diff --git a/.github/workflows/issue-subscriber.yml b/.github/workflows/issue-subscriber.yml index 9a9c8f9c65162..111fa6d7aa67e 100644 --- a/.github/workflows/issue-subscriber.yml +++ b/.github/workflows/issue-subscriber.yml @@ -13,14 +13,20 @@ jobs: runs-on: ubuntu-latest if: github.repository == 'llvm/llvm-project' steps: + - name: Checkout Automation Script + uses: actions/checkout@v4 + with: + sparse-checkout: llvm/utils/git/ + ref: main + - name: Setup Automation Script + working-directory: ./llvm/utils/git/ run: | - curl -O -L --fail https://raw.githubusercontent.com/"$GITHUB_REPOSITORY"/"$GITHUB_SHA"/llvm/utils/git/github-automation.py - curl -O -L --fail https://raw.githubusercontent.com/"$GITHUB_REPOSITORY"/"$GITHUB_SHA"/llvm/utils/git/requirements.txt chmod a+x github-automation.py pip install -r requirements.txt - name: Update watchers + working-directory: ./llvm/utils/git/ # https://docs.github.com/en/actions/security-guides/security-hardening-for-github-actions#using-an-intermediate-environment-variable env: LABEL_NAME: ${{ github.event.label.name }} diff --git a/.github/workflows/libclang-python-tests.yml b/.github/workflows/libclang-python-tests.yml index 73edb6cf3bad2..e12acbc0f6ce8 100644 --- a/.github/workflows/libclang-python-tests.yml +++ b/.github/workflows/libclang-python-tests.yml @@ -30,6 +30,10 @@ jobs: check-clang-python: # Build libclang and then run the libclang Python binding's unit tests. name: Build and run Python unit tests + strategy: + fail-fast: false + matrix: + python-version: ["3.7", "3.11"] uses: ./.github/workflows/llvm-project-tests.yml with: build_target: check-clang-python @@ -37,3 +41,4 @@ jobs: # There is an issue running on "windows-2019". # See https://github.com/llvm/llvm-project/issues/76601#issuecomment-1873049082. os_list: '["ubuntu-latest"]' + python_version: ${{ matrix.python-version }} diff --git a/.github/workflows/libcxx-build-and-test.yaml b/.github/workflows/libcxx-build-and-test.yaml index 985790a0ee236..0cab9b841e4ee 100644 --- a/.github/workflows/libcxx-build-and-test.yaml +++ b/.github/workflows/libcxx-build-and-test.yaml @@ -35,7 +35,6 @@ concurrency: env: - CMAKE: "/opt/bin/cmake" # LLVM POST-BRANCH bump version # LLVM POST-BRANCH add compiler test for ToT - 1, e.g. "Clang 17" # LLVM RELEASE bump remove compiler ToT - 3, e.g. "Clang 15" @@ -163,31 +162,24 @@ jobs: 'generic-no-rtti', 'generic-optimized-speed', 'generic-static', - 'generic-with_llvm_unwinder', # TODO Find a better place for the benchmark and bootstrapping builds to live. They're either very expensive # or don't provide much value since the benchmark run results are too noise on the bots. 'benchmarks', 'bootstrapping-build' ] machine: [ 'libcxx-runners-8-set' ] - std_modules: [ 'OFF' ] include: - config: 'generic-cxx26' machine: libcxx-runners-8-set - std_modules: 'ON' - config: 'generic-asan' machine: libcxx-runners-8-set - std_modules: 'OFF' - config: 'generic-tsan' machine: libcxx-runners-8-set - std_modules: 'OFF' - config: 'generic-ubsan' machine: libcxx-runners-8-set - std_modules: 'OFF' # Use a larger machine for MSAN to avoid timeout and memory allocation issues. - config: 'generic-msan' machine: libcxx-runners-8-set - std_modules: 'OFF' runs-on: ${{ matrix.machine }} steps: - uses: actions/checkout@v4 @@ -197,7 +189,6 @@ jobs: CC: clang-18 CXX: clang++-18 ENABLE_CLANG_TIDY: "OFF" - ENABLE_STD_MODULES: ${{ matrix.std_modules }} - uses: actions/upload-artifact@v3 if: always() with: diff --git a/.github/workflows/llvm-project-tests.yml b/.github/workflows/llvm-project-tests.yml index fadaea129e508..a1404e1f1efa9 100644 --- a/.github/workflows/llvm-project-tests.yml +++ b/.github/workflows/llvm-project-tests.yml @@ -15,6 +15,10 @@ on: os_list: required: false default: '["ubuntu-latest", "windows-2019", "macOS-11"]' + python_version: + required: false + type: string + default: '3.11' workflow_call: inputs: build_target: @@ -38,6 +42,11 @@ on: # https://github.com/actions/virtual-environments/issues/5900 default: '["ubuntu-latest", "windows-2019", "macOS-11"]' + python_version: + required: false + type: string + default: '3.11' + concurrency: # Skip intermediate builds: always. # Cancel intermediate builds: only if it is a pull request build. @@ -67,7 +76,7 @@ jobs: - name: Setup Python uses: actions/setup-python@v4 with: - python-version: '3.11' + python-version: ${{ inputs.python_version }} - name: Install Ninja uses: llvm/actions/install-ninja@main # actions/checkout deletes any existing files in the new git directory, diff --git a/.github/workflows/new-prs.yml b/.github/workflows/new-prs.yml index 23fab598fc77d..a52e4e9436263 100644 --- a/.github/workflows/new-prs.yml +++ b/.github/workflows/new-prs.yml @@ -34,14 +34,20 @@ jobs: (github.event.pull_request.author_association != 'MEMBER') && (github.event.pull_request.author_association != 'OWNER') steps: + - name: Checkout Automation Script + uses: actions/checkout@v4 + with: + sparse-checkout: llvm/utils/git/ + ref: main + - name: Setup Automation Script + working-directory: ./llvm/utils/git/ run: | - curl -O -L --fail https://raw.githubusercontent.com/"$GITHUB_REPOSITORY"/main/llvm/utils/git/github-automation.py - curl -O -L --fail https://raw.githubusercontent.com/"$GITHUB_REPOSITORY"/main/llvm/utils/git/requirements.txt chmod a+x github-automation.py pip install -r requirements.txt - name: Greet Author + working-directory: ./llvm/utils/git/ run: | ./github-automation.py \ --token '${{ secrets.GITHUB_TOKEN }}' \ diff --git a/.github/workflows/pr-subscriber.yml b/.github/workflows/pr-subscriber.yml index 99f9be771588d..aa36e6d502c95 100644 --- a/.github/workflows/pr-subscriber.yml +++ b/.github/workflows/pr-subscriber.yml @@ -13,14 +13,20 @@ jobs: runs-on: ubuntu-latest if: github.repository == 'llvm/llvm-project' steps: + - name: Checkout Automation Script + uses: actions/checkout@v4 + with: + sparse-checkout: llvm/utils/git/ + ref: main + - name: Setup Automation Script + working-directory: ./llvm/utils/git/ run: | - curl -O -L --fail https://raw.githubusercontent.com/"$GITHUB_REPOSITORY"/main/llvm/utils/git/github-automation.py - curl -O -L --fail https://raw.githubusercontent.com/"$GITHUB_REPOSITORY"/main/llvm/utils/git/requirements.txt chmod a+x github-automation.py pip install -r requirements.txt - name: Update watchers + working-directory: ./llvm/utils/git/ run: | ./github-automation.py \ --token '${{ secrets.ISSUE_SUBSCRIBER_TOKEN }}' \ diff --git a/.github/workflows/release-binaries.yml b/.github/workflows/release-binaries.yml index 4a4ba9dcc5ca3..ebf6fa41898dc 100644 --- a/.github/workflows/release-binaries.yml +++ b/.github/workflows/release-binaries.yml @@ -1,20 +1,29 @@ name: Release Binaries on: - push: - tags: - - 'llvmorg-*' workflow_dispatch: inputs: + release-version: + description: 'Release Version' + required: true + type: string upload: description: 'Upload binaries to the release page' required: true - default: true + default: false type: boolean - tag: - description: 'Tag to build' + + workflow_call: + inputs: + release-version: + description: 'Release Version' required: true type: string + upload: + description: 'Upload binaries to the release page' + required: true + default: false + type: boolean schedule: # * is a special character in YAML so you have to quote this string - cron: '0 8 1 * *' @@ -26,21 +35,26 @@ jobs: prepare: name: Prepare to build binaries runs-on: ubuntu-22.04 - if: github.repository == 'llvm/llvm-project' outputs: - release-version: ${{ steps.validate-tag.outputs.release-version }} - flags: ${{ steps.validate-tag.outputs.flags }} - build-dir: ${{ steps.validate-tag.outputs.build-dir }} - rc-flags: ${{ steps.validate-tag.outputs.rc-flags }} - ref: ${{ steps.validate-tag.outputs.ref }} - upload: ${{ steps.validate-tag.outputs.upload }} + release-version: ${{ steps.vars.outputs.release-version }} + flags: ${{ steps.vars.outputs.flags }} + build-dir: ${{ steps.vars.outputs.build-dir }} + rc-flags: ${{ steps.vars.outputs.rc-flags }} + ref: ${{ steps.vars.outputs.ref }} + upload: ${{ steps.vars.outputs.upload }} steps: - name: Checkout LLVM - uses: actions/checkout@v4 + uses: actions/checkout@b4ffde65f46336ab88eb53be808477a3936bae11 # v4.1.1 + + - name: Check Permissions + env: + GITHUB_TOKEN: ${{ github.token }} + run: | + ./llvm/utils/release/./github-upload-release.py --token "$GITHUB_TOKEN" --user ${{ github.actor }} check-permissions - - name: Validate and parse tag - id: validate-tag + - name: Collect Variables + id: vars # In order for the test-release.sh script to run correctly, the LLVM # source needs to be at the following location relative to the build dir: # | X.Y.Z-rcN | ./rcN/llvm-project @@ -61,9 +75,9 @@ jobs: if [ -n "${{ inputs.upload }}" ]; then upload="${{ inputs.upload }}" else - upload="true" + upload="false" fi - bash .github/workflows/set-release-binary-outputs.sh "${{ github.actor }}" "$tag" "$upload" + bash .github/workflows/set-release-binary-outputs.sh "$tag" "$upload" # Try to get around the 6 hour timeout by first running a job to fill # the build cache. @@ -77,15 +91,15 @@ jobs: - ubuntu-22.04 steps: - name: Checkout LLVM - uses: actions/checkout@v4 + uses: actions/checkout@b4ffde65f46336ab88eb53be808477a3936bae11 # v4.1.1 with: ref: ${{ needs.prepare.outputs.ref }} - name: Install Ninja - uses: llvm/actions/install-ninja@main + uses: llvm/actions/install-ninja@22e9f909d35b50bd1181709564bfe816eaeaae81 # main - name: Setup sccache - uses: hendrikmuhs/ccache-action@v1 + uses: hendrikmuhs/ccache-action@ca3acd2731eef11f1572ccb126356c2f9298d35e # v1.2.9 with: max-size: 250M key: sccache-${{ matrix.os }}-release @@ -119,13 +133,13 @@ jobs: steps: - name: Checkout LLVM - uses: actions/checkout@v4 + uses: actions/checkout@b4ffde65f46336ab88eb53be808477a3936bae11 # v4.1.1 with: ref: ${{ needs.prepare.outputs.ref }} path: ${{ needs.prepare.outputs.build-dir }}/llvm-project - name: Setup sccache - uses: hendrikmuhs/ccache-action@v1 + uses: hendrikmuhs/ccache-action@ca3acd2731eef11f1572ccb126356c2f9298d35e # v1.2.9 with: max-size: 250M key: sccache-${{ matrix.target.os }}-release diff --git a/.github/workflows/release-documentation.yml b/.github/workflows/release-documentation.yml new file mode 100644 index 0000000000000..64572906988ba --- /dev/null +++ b/.github/workflows/release-documentation.yml @@ -0,0 +1,88 @@ +name: Release Documentation + +permissions: + contents: read + +on: + workflow_dispatch: + inputs: + release-version: + description: 'Release Version' + required: true + type: string + upload: + description: 'Upload documentation' + required: false + type: boolean + + workflow_call: + inputs: + release-version: + description: 'Release Version' + required: true + type: string + upload: + description: 'Upload documentation' + required: false + type: boolean + +jobs: + release-documentation: + name: Build and Upload Release Documentation + runs-on: ubuntu-latest + env: + upload: ${{ inputs.upload && !contains(inputs.release-version, 'rc') }} + steps: + - name: Checkout LLVM + uses: actions/checkout@b4ffde65f46336ab88eb53be808477a3936bae11 # v4.1.1 + + - name: Setup Python env + uses: actions/setup-python@v4 + with: + cache: 'pip' + cache-dependency-path: './llvm/docs/requirements.txt' + + - name: Install Dependencies + run: | + sudo apt-get update + sudo apt-get install -y \ + graphviz \ + python3-github \ + ninja-build \ + texlive-font-utils + pip3 install --user -r ./llvm/docs/requirements.txt + + - name: Build Documentation + env: + GITHUB_TOKEN: ${{ github.token }} + run: | + ./llvm/utils/release/build-docs.sh -release "${{ inputs.release-version }}" -no-doxygen + + - name: Create Release Notes Artifact + uses: actions/upload-artifact@v3 + with: + name: release-notes + path: docs-build/html-export/ + + - name: Clone www-releases + if: env.upload + uses: actions/checkout@b4ffde65f46336ab88eb53be808477a3936bae11 # v4.1.1 + with: + repository: ${{ github.repository_owner }}/www-releases + ref: main + fetch-depth: 0 + path: www-releases + + - name: Upload Release Notes + if: env.upload + env: + WWW_RELEASES_TOKEN: ${{ secrets.WWW_RELEASES_TOKEN }} + run: | + mkdir -p ../www-releases/${{ inputs.release-version }} + mv ./docs-build/html-export/* ../www-releases/${{ inputs.release-version }} + cd ../www-releases + git add ${{ inputs.release-version }} + git config user.email "llvmbot@llvm.org" + git config user.name "llvmbot" + git commit -a -m "Add ${{ inputs.release-version }} documentation" + git push "https://$WWW_RELEASES_TOKEN@github.com/${{ github.repository_owner }}/www-releases" main:main diff --git a/.github/workflows/release-doxygen.yml b/.github/workflows/release-doxygen.yml new file mode 100644 index 0000000000000..5e322849a1d09 --- /dev/null +++ b/.github/workflows/release-doxygen.yml @@ -0,0 +1,67 @@ +name: Release Doxygen + +permissions: + contents: read + +on: + workflow_dispatch: + inputs: + release-version: + description: 'Release Version' + required: true + type: string + upload: + description: 'Upload documentation' + required: false + type: boolean + + workflow_call: + inputs: + release-version: + description: 'Release Version' + required: true + type: string + upload: + description: 'Upload documentation' + required: false + type: boolean + +jobs: + release-doxygen: + name: Build and Upload Release Doxygen + runs-on: ubuntu-latest + permissions: + contents: write + env: + upload: ${{ inputs.upload && !contains(inputs.release-version, 'rc') }} + steps: + - name: Checkout LLVM + uses: actions/checkout@b4ffde65f46336ab88eb53be808477a3936bae11 # v4.1.1 + + - name: Setup Python env + uses: actions/setup-python@v4 + with: + cache: 'pip' + cache-dependency-path: './llvm/docs/requirements.txt' + + - name: Install Dependencies + run: | + sudo apt-get update + sudo apt-get install -y \ + doxygen \ + graphviz \ + python3-github \ + ninja-build \ + texlive-font-utils + pip3 install --user -r ./llvm/docs/requirements.txt + + - name: Build Doxygen + env: + GITHUB_TOKEN: ${{ github.token }} + run: | + ./llvm/utils/release/build-docs.sh -release "${{ inputs.release-version }}" -no-sphinx + + - name: Upload Doxygen + if: env.upload + run: | + ./llvm/utils/release/github-upload-release.py --token "$GITHUB_TOKEN" --release "${{ inputs.release-version }}" --user "${{ github.actor }}" upload --files ./*doxygen*.tar.xz diff --git a/.github/workflows/release-lit.yml b/.github/workflows/release-lit.yml new file mode 100644 index 0000000000000..36b0b6edd518f --- /dev/null +++ b/.github/workflows/release-lit.yml @@ -0,0 +1,74 @@ +name: Release Lit + +permissions: + contents: read + +on: + workflow_dispatch: + inputs: + release-version: + description: 'Release Version' + required: true + type: string + + workflow_call: + inputs: + release-version: + description: 'Release Version' + required: true + type: string + +jobs: + release-lit: + name: Release Lit + runs-on: ubuntu-latest + steps: + - name: Checkout LLVM + uses: actions/checkout@b4ffde65f46336ab88eb53be808477a3936bae11 # v4.1.1 + with: + ref: "llvmorg-${{ inputs.release-version }}" + + - name: Install dependencies + run: | + sudo apt-get update + sudo apt-get install -y python3-setuptools python3-psutil python3-github + + - name: Check Permissions + env: + GITHUB_TOKEN: ${{ github.token }} + run: | + ./llvm/utils/release/./github-upload-release.py --token "$GITHUB_TOKEN" --user ${{ github.actor }} check-permissions + + - name: Setup Cpp + uses: aminya/setup-cpp@v1 + with: + compiler: llvm-16.0.6 + cmake: true + ninja: true + + - name: Test lit + run: | + mkdir build && cd build + export FILECHECK_OPTS='-dump-input-filter=all -vv -color' + cmake ../llvm -DCMAKE_BUILD_TYPE=Release -G Ninja + ninja -v -j $(nproc) check-lit + + - name: Package lit + run: | + cd llvm/utils/lit + # Remove 'dev' suffix from lit version. + sed -i 's/ + "dev"//g' lit/__init__.py + python3 setup.py sdist + + - name: Upload lit to test.pypi.org + uses: pypa/gh-action-pypi-publish@release/v1 + with: + password: ${{ secrets.LLVM_LIT_TEST_PYPI_API_TOKEN }} + repository-url: https://test.pypi.org/legacy/ + packages-dir: llvm/utils/lit/dist/ + + - name: Upload lit to pypi.org + uses: pypa/gh-action-pypi-publish@release/v1 + with: + password: ${{ secrets.LLVM_LIT_PYPI_API_TOKEN }} + packages-dir: llvm/utils/lit/dist/ diff --git a/.github/workflows/release-tasks.yml b/.github/workflows/release-tasks.yml index 85b720e323d1b..f2a831ad3577a 100644 --- a/.github/workflows/release-tasks.yml +++ b/.github/workflows/release-tasks.yml @@ -1,7 +1,7 @@ name: Release Task permissions: - contents: read + contents: write on: push: @@ -10,112 +10,70 @@ on: - 'llvmorg-*' jobs: - release-tasks: - permissions: - contents: write # To upload assets to release. + validate-tag: + name: Validate Tag runs-on: ubuntu-latest if: github.repository == 'llvm/llvm-project' + outputs: + release-version: ${{ steps.validate-tag.outputs.release-version }} steps: - name: Validate Tag id: validate-tag run: | - test "${{ github.actor }}" = "tstellar" || test "${{ github.actor }}" = "tru" echo "${{ github.ref_name }}" | grep -e '^llvmorg-[0-9]\+\.[0-9]\+\.[0-9]\+\(-rc[0-9]\+\)\?$' release_version=$(echo "${{ github.ref_name }}" | sed 's/llvmorg-//g') echo "release-version=$release_version" >> "$GITHUB_OUTPUT" - - name: Checkout LLVM - uses: actions/checkout@v4 - - - name: Install Dependencies - run: | - sudo apt-get update - sudo apt-get install -y \ - doxygen \ - graphviz \ - python3-github \ - ninja-build \ - texlive-font-utils - pip3 install --user -r ./llvm/docs/requirements.txt - - - name: Create Release - run: | - ./llvm/utils/release/./github-upload-release.py --token ${{ github.token }} --release ${{ steps.validate-tag.outputs.release-version }} create - - - name: Build Documentation - run: | - ./llvm/utils/release/build-docs.sh -release ${{ steps.validate-tag.outputs.release-version }} - ./llvm/utils/release/github-upload-release.py --token ${{ github.token }} --release ${{ steps.validate-tag.outputs.release-version }} upload --files ./*doxygen*.tar.xz - - - name: Create Release Notes Artifact - uses: actions/upload-artifact@v3 - with: - name: release-notes - path: docs-build/html-export/ - - - name: Clone www-releases - if: ${{ !contains(steps.validate-tag.outputs.release-version, 'rc') }} - uses: actions/checkout@v4 - with: - repository: ${{ github.repository_owner }}/www-releases - ref: main - fetch-depth: 0 - path: www-releases - - - name: Upload Release Notes - if: ${{ !contains(steps.validate-tag.outputs.release-version, 'rc') }} - run: | - mkdir -p ../www-releases/${{ steps.validate-tag.outputs.release-version }} - mv ./docs-build/html-export/* ../www-releases/${{ steps.validate-tag.outputs.release-version }} - cd ../www-releases - git add ${{ steps.validate-tag.outputs.release-version }} - git config user.email "llvmbot@llvm.org" - git config user.name "llvmbot" - git commit -a -m "Add ${{ steps.validate-tag.outputs.release-version }} documentation" - git push https://${{ secrets.WWW_RELEASES_TOKEN }}@github.com/${{ github.repository_owner }}/www-releases main:main - - release-lit: + release-create: + name: Create a New Release runs-on: ubuntu-latest - if: github.repository == 'llvm/llvm-project' + needs: validate-tag steps: - - name: Checkout LLVM - uses: actions/checkout@v4 - - - name: Setup Cpp - uses: aminya/setup-cpp@v1 - with: - compiler: llvm-16.0.6 - cmake: true - ninja: true - - - name: Install dependencies + - name: Install Dependencies run: | sudo apt-get update - sudo apt-get install -y python3-setuptools python3-psutil + sudo apt-get install python3-github - - name: Test lit - run: | - mkdir build && cd build - export FILECHECK_OPTS='-dump-input-filter=all -vv -color' - cmake ../llvm -DCMAKE_BUILD_TYPE=Release -G Ninja - ninja -v -j $(nproc) check-lit + - name: Checkout LLVM + uses: actions/checkout@b4ffde65f46336ab88eb53be808477a3936bae11 # v4.1.1 - - name: Package lit + - name: Create Release + env: + GITHUB_TOKEN: ${{ github.token }} run: | - cd llvm/utils/lit - # Remove 'dev' suffix from lit version. - sed -i 's/ + "dev"//g' lit/__init__.py - python3 setup.py sdist - - - name: Upload lit to test.pypi.org - uses: pypa/gh-action-pypi-publish@release/v1 - with: - password: ${{ secrets.LLVM_LIT_TEST_PYPI_API_TOKEN }} - repository-url: https://test.pypi.org/legacy/ - packages-dir: llvm/utils/lit/dist/ + ./llvm/utils/release/./github-upload-release.py --token "$GITHUB_TOKEN" --release ${{ needs.validate-tag.outputs.release-version }} --user ${{ github.actor }} create + release-documentation: + name: Build and Upload Release Documentation + needs: + - validate-tag + uses: ./.github/workflows/release-documentation.yml + with: + release-version: ${{ needs.validate-tag.outputs.release-version }} + upload: true + + release-doxygen: + name: Build and Upload Release Doxygen + needs: + - validate-tag + - release-create + uses: ./.github/workflows/release-doxygen.yml + with: + release-version: ${{ needs.validate-tag.outputs.release-version }} + upload: true - - name: Upload lit to pypi.org - uses: pypa/gh-action-pypi-publish@release/v1 - with: - password: ${{ secrets.LLVM_LIT_PYPI_API_TOKEN }} - packages-dir: llvm/utils/lit/dist/ + release-lit: + name: Release Lit + needs: validate-tag + uses: ./.github/workflows/release-lit.yml + with: + release-version: ${{ needs.validate-tag.outputs.release-version }} + + release-binaries: + name: Build Release Binaries + needs: + - validate-tag + - release-create + uses: ./.github/workflows/release-binaries.yml + with: + release-version: ${{ needs.validate-tag.outputs.release-version }} + upload: true diff --git a/.github/workflows/set-release-binary-outputs.sh b/.github/workflows/set-release-binary-outputs.sh index 9bc459a24e801..59470cf83ba75 100644 --- a/.github/workflows/set-release-binary-outputs.sh +++ b/.github/workflows/set-release-binary-outputs.sh @@ -8,14 +8,8 @@ if [ -z "$GITHUB_OUTPUT" ]; then echo "Writing output variables to $GITHUB_OUTPUT" fi -github_user=$1 -tag=$2 -upload=$3 - -if [[ "$github_user" != "tstellar" && "$github_user" != "tru" ]]; then - echo "ERROR: User not allowed: $github_user" - exit 1 -fi +tag=$1 +upload=$2 if echo $tag | grep -e '^[0-9a-f]\+$'; then # This is a plain commit. diff --git a/bolt/docs/BAT.md b/bolt/docs/BAT.md new file mode 100644 index 0000000000000..96a9a187e0acf --- /dev/null +++ b/bolt/docs/BAT.md @@ -0,0 +1,93 @@ +# BOLT Address Translation (BAT) +# Purpose +A regular profile collection for BOLT involves collecting samples from +unoptimized binary. BOLT Address Translation allows collecting profile +from BOLT-optimized binary and using it for optimizing the input (pre-BOLT) +binary. + +# Overview +BOLT Address Translation is an extra section (`.note.bolt_bat`) inserted by BOLT +into the output binary containing translation tables and split functions linkage +information. This information enables mapping the profile back from optimized +binary onto the original binary. + +# Usage +`--enable-bat` flag controls the generation of BAT section. Sampled profile +needs to be passed along with the optimized binary containing BAT section to +`perf2bolt` which reads BAT section and produces fdata profile for the original +binary. Note that YAML profile generation is not supported since BAT doesn't +contain the metadata for input functions. + +# Internals +## Section contents +The section is organized as follows: +- Hot functions table + - Address translation tables +- Cold functions table + +## Construction and parsing +BAT section is created from `BoltAddressTranslation` class which captures +address translation information provided by BOLT linker. It is then encoded as a +note section in the output binary. + +During profile conversion when BAT-enabled binary is passed to perf2bolt, +`BoltAddressTranslation` class is populated from BAT section. The class is then +queried by `DataAggregator` during sample processing to reconstruct addresses/ +offsets in the input binary. + +## Encoding format +The encoding is specified in +[BoltAddressTranslation.h](/bolt/include/bolt/Profile/BoltAddressTranslation.h) +and [BoltAddressTranslation.cpp](/bolt/lib/Profile/BoltAddressTranslation.cpp). + +### Layout +The general layout is as follows: +``` +Hot functions table header +|------------------| +| Function entry | +| |--------------| | +| | OutOff InOff | | +| |--------------| | +~~~~~~~~~~~~~~~~~~~~ + +Cold functions table header +|------------------| +| Function entry | +| |--------------| | +| | OutOff InOff | | +| |--------------| | +~~~~~~~~~~~~~~~~~~~~ +``` + +### Functions table +Hot and cold functions tables share the encoding except difference marked below. +Header: +| Entry | Encoding | Description | +| ------ | ----- | ----------- | +| `NumFuncs` | ULEB128 | Number of functions in the functions table | + +The header is followed by Functions table with `NumFuncs` entries. +Output binary addresses are delta encoded, meaning that only the difference with +the previous output address is stored. Addresses implicitly start at zero. +Hot indices are delta encoded, implicitly starting at zero. +| Entry | Encoding | Description | +| ------ | ------| ----------- | +| `Address` | Delta, ULEB128 | Function address in the output binary | +| `HotIndex` | Delta, ULEB128 | Cold functions only: index of corresponding hot function in hot functions table | +| `NumEntries` | ULEB128 | Number of address translation entries for a function | + +Function header is followed by `NumEntries` pairs of offsets for current +function. + +### Address translation table +Delta encoding means that only the difference with the previous corresponding +entry is encoded. Offsets implicitly start at zero. +| Entry | Encoding | Description | +| ------ | ------| ----------- | +| `OutputOffset` | Delta, ULEB128 | Function offset in output binary | +| `InputOffset` | Delta, SLEB128 | Function offset in input binary with `BRANCHENTRY` LSB bit | + +`BRANCHENTRY` bit denotes whether a given offset pair is a control flow source +(branch or call instruction). If not set, it signifies a control flow target +(basic block offset). diff --git a/bolt/include/bolt/Profile/BoltAddressTranslation.h b/bolt/include/bolt/Profile/BoltAddressTranslation.h index 07e4b283211c6..01d3be4ee59be 100644 --- a/bolt/include/bolt/Profile/BoltAddressTranslation.h +++ b/bolt/include/bolt/Profile/BoltAddressTranslation.h @@ -11,6 +11,7 @@ #include "llvm/ADT/SmallVector.h" #include "llvm/ADT/StringRef.h" +#include "llvm/Support/DataExtractor.h" #include #include #include @@ -118,6 +119,16 @@ class BoltAddressTranslation { void writeEntriesForBB(MapTy &Map, const BinaryBasicBlock &BB, uint64_t FuncAddress); + /// Write the serialized address translation table for a function. + template + void writeMaps(std::map &Maps, raw_ostream &OS); + + /// Read the serialized address translation table for a function. + /// Return a parse error if failed. + template + void parseMaps(std::vector &HotFuncs, DataExtractor &DE, + uint64_t &Offset, Error &Err); + std::map Maps; /// Links outlined cold bocks to their original function @@ -125,7 +136,7 @@ class BoltAddressTranslation { /// Identifies the address of a control-flow changing instructions in a /// translation map entry - const static uint32_t BRANCHENTRY = 0x80000000; + const static uint32_t BRANCHENTRY = 0x1; }; } // namespace bolt diff --git a/bolt/lib/Passes/CacheMetrics.cpp b/bolt/lib/Passes/CacheMetrics.cpp index ba3a2a5f685f3..f6708997a4335 100644 --- a/bolt/lib/Passes/CacheMetrics.cpp +++ b/bolt/lib/Passes/CacheMetrics.cpp @@ -14,23 +14,20 @@ #include "bolt/Passes/CacheMetrics.h" #include "bolt/Core/BinaryBasicBlock.h" #include "bolt/Core/BinaryFunction.h" -#include "llvm/Support/CommandLine.h" #include using namespace llvm; using namespace bolt; -namespace opts { - -extern cl::OptionCategory BoltOptCategory; - -extern cl::opt ITLBPageSize; -extern cl::opt ITLBEntries; - -} // namespace opts - namespace { +/// The following constants are used to estimate the number of i-TLB cache +/// misses for a given code layout. Empirically the values result in high +/// correlations between the estimations and the perf measurements. +/// The constants do not affect the code layout algorithms. +constexpr unsigned ITLBPageSize = 4096; +constexpr unsigned ITLBEntries = 16; + /// Initialize and return a position map for binary basic blocks void extractBasicBlockInfo( const std::vector &BinaryFunctions, @@ -133,9 +130,6 @@ double expectedCacheHitRatio( const std::vector &BinaryFunctions, const std::unordered_map &BBAddr, const std::unordered_map &BBSize) { - - const double PageSize = opts::ITLBPageSize; - const uint64_t CacheEntries = opts::ITLBEntries; std::unordered_map Calls = extractFunctionCalls(BinaryFunctions); // Compute 'hotness' of the functions @@ -155,7 +149,8 @@ double expectedCacheHitRatio( for (BinaryFunction *BF : BinaryFunctions) { if (BF->getLayout().block_empty()) continue; - double Page = BBAddr.at(BF->getLayout().block_front()) / PageSize; + const uint64_t Page = + BBAddr.at(BF->getLayout().block_front()) / ITLBPageSize; PageSamples[Page] += FunctionSamples.at(BF); } @@ -166,15 +161,17 @@ double expectedCacheHitRatio( if (BF->getLayout().block_empty() || FunctionSamples.at(BF) == 0.0) continue; double Samples = FunctionSamples.at(BF); - double Page = BBAddr.at(BF->getLayout().block_front()) / PageSize; + const uint64_t Page = + BBAddr.at(BF->getLayout().block_front()) / ITLBPageSize; // The probability that the page is not present in the cache - double MissProb = pow(1.0 - PageSamples[Page] / TotalSamples, CacheEntries); + const double MissProb = + pow(1.0 - PageSamples[Page] / TotalSamples, ITLBEntries); // Processing all callers of the function for (std::pair Pair : Calls[BF]) { BinaryFunction *SrcFunction = Pair.first; - double SrcPage = - BBAddr.at(SrcFunction->getLayout().block_front()) / PageSize; + const uint64_t SrcPage = + BBAddr.at(SrcFunction->getLayout().block_front()) / ITLBPageSize; // Is this a 'long' or a 'short' call? if (Page != SrcPage) { // This is a miss diff --git a/bolt/lib/Profile/BoltAddressTranslation.cpp b/bolt/lib/Profile/BoltAddressTranslation.cpp index e004309e0e213..697ff1e5dd0da 100644 --- a/bolt/lib/Profile/BoltAddressTranslation.cpp +++ b/bolt/lib/Profile/BoltAddressTranslation.cpp @@ -8,8 +8,9 @@ #include "bolt/Profile/BoltAddressTranslation.h" #include "bolt/Core/BinaryFunction.h" -#include "llvm/Support/DataExtractor.h" #include "llvm/Support/Errc.h" +#include "llvm/Support/Error.h" +#include "llvm/Support/LEB128.h" #define DEBUG_TYPE "bolt-bat" @@ -44,7 +45,7 @@ void BoltAddressTranslation::writeEntriesForBB(MapTy &Map, // and this deleted block will both share the same output address (the same // key), and we need to map back. We choose here to privilege the successor by // allowing it to overwrite the previously inserted key in the map. - Map[BBOutputOffset] = BBInputOffset; + Map[BBOutputOffset] = BBInputOffset << 1; const auto &IOAddressMap = BB.getFunction()->getBinaryContext().getIOAddressMap(); @@ -61,8 +62,8 @@ void BoltAddressTranslation::writeEntriesForBB(MapTy &Map, LLVM_DEBUG(dbgs() << " Key: " << Twine::utohexstr(OutputOffset) << " Val: " << Twine::utohexstr(InputOffset) << " (branch)\n"); - Map.insert( - std::pair(OutputOffset, InputOffset | BRANCHENTRY)); + Map.insert(std::pair(OutputOffset, + (InputOffset << 1) | BRANCHENTRY)); } } @@ -101,36 +102,51 @@ void BoltAddressTranslation::write(const BinaryContext &BC, raw_ostream &OS) { } } - const uint32_t NumFuncs = Maps.size(); - OS.write(reinterpret_cast(&NumFuncs), 4); - LLVM_DEBUG(dbgs() << "Writing " << NumFuncs << " functions for BAT.\n"); + writeMaps(Maps, OS); + writeMaps(Maps, OS); + + outs() << "BOLT-INFO: Wrote " << Maps.size() << " BAT maps\n"; +} + +template +void BoltAddressTranslation::writeMaps(std::map &Maps, + raw_ostream &OS) { + const uint32_t NumFuncs = + llvm::count_if(llvm::make_first_range(Maps), [&](const uint64_t Address) { + return Cold == ColdPartSource.count(Address); + }); + encodeULEB128(NumFuncs, OS); + LLVM_DEBUG(dbgs() << "Writing " << NumFuncs << (Cold ? " cold" : "") + << " functions for BAT.\n"); + size_t PrevIndex = 0; + // Output addresses are delta-encoded + uint64_t PrevAddress = 0; for (auto &MapEntry : Maps) { const uint64_t Address = MapEntry.first; + // Only process cold fragments in cold mode, and vice versa. + if (Cold != ColdPartSource.count(Address)) + continue; MapTy &Map = MapEntry.second; const uint32_t NumEntries = Map.size(); LLVM_DEBUG(dbgs() << "Writing " << NumEntries << " entries for 0x" << Twine::utohexstr(Address) << ".\n"); - OS.write(reinterpret_cast(&Address), 8); - OS.write(reinterpret_cast(&NumEntries), 4); + encodeULEB128(Address - PrevAddress, OS); + PrevAddress = Address; + if (Cold) { + size_t HotIndex = + std::distance(ColdPartSource.begin(), ColdPartSource.find(Address)); + encodeULEB128(HotIndex - PrevIndex, OS); + PrevIndex = HotIndex; + } + encodeULEB128(NumEntries, OS); + uint64_t InOffset = 0, OutOffset = 0; + // Output and Input addresses and delta-encoded for (std::pair &KeyVal : Map) { - OS.write(reinterpret_cast(&KeyVal.first), 4); - OS.write(reinterpret_cast(&KeyVal.second), 4); + encodeULEB128(KeyVal.first - OutOffset, OS); + encodeSLEB128(KeyVal.second - InOffset, OS); + std::tie(OutOffset, InOffset) = KeyVal; } } - const uint32_t NumColdEntries = ColdPartSource.size(); - LLVM_DEBUG(dbgs() << "Writing " << NumColdEntries - << " cold part mappings.\n"); - OS.write(reinterpret_cast(&NumColdEntries), 4); - for (std::pair &ColdEntry : ColdPartSource) { - OS.write(reinterpret_cast(&ColdEntry.first), 8); - OS.write(reinterpret_cast(&ColdEntry.second), 8); - LLVM_DEBUG(dbgs() << " " << Twine::utohexstr(ColdEntry.first) << " -> " - << Twine::utohexstr(ColdEntry.second) << "\n"); - } - - outs() << "BOLT-INFO: Wrote " << Maps.size() << " BAT maps\n"; - outs() << "BOLT-INFO: Wrote " << NumColdEntries - << " BAT cold-to-hot entries\n"; } std::error_code BoltAddressTranslation::parse(StringRef Buf) { @@ -152,53 +168,52 @@ std::error_code BoltAddressTranslation::parse(StringRef Buf) { if (Name.substr(0, 4) != "BOLT") return make_error_code(llvm::errc::io_error); - if (Buf.size() - Offset < 4) - return make_error_code(llvm::errc::io_error); + Error Err(Error::success()); + std::vector HotFuncs; + parseMaps(HotFuncs, DE, Offset, Err); + parseMaps(HotFuncs, DE, Offset, Err); + outs() << "BOLT-INFO: Parsed " << Maps.size() << " BAT entries\n"; + return errorToErrorCode(std::move(Err)); +} - const uint32_t NumFunctions = DE.getU32(&Offset); - LLVM_DEBUG(dbgs() << "Parsing " << NumFunctions << " functions\n"); +template +void BoltAddressTranslation::parseMaps(std::vector &HotFuncs, + DataExtractor &DE, uint64_t &Offset, + Error &Err) { + const uint32_t NumFunctions = DE.getULEB128(&Offset, &Err); + LLVM_DEBUG(dbgs() << "Parsing " << NumFunctions << (Cold ? " cold" : "") + << " functions\n"); + size_t HotIndex = 0; + uint64_t PrevAddress = 0; for (uint32_t I = 0; I < NumFunctions; ++I) { - if (Buf.size() - Offset < 12) - return make_error_code(llvm::errc::io_error); - - const uint64_t Address = DE.getU64(&Offset); - const uint32_t NumEntries = DE.getU32(&Offset); + const uint64_t Address = PrevAddress + DE.getULEB128(&Offset, &Err); + PrevAddress = Address; + if (Cold) { + HotIndex += DE.getULEB128(&Offset, &Err); + ColdPartSource.emplace(Address, HotFuncs[HotIndex]); + } else { + HotFuncs.push_back(Address); + } + const uint32_t NumEntries = DE.getULEB128(&Offset, &Err); MapTy Map; LLVM_DEBUG(dbgs() << "Parsing " << NumEntries << " entries for 0x" << Twine::utohexstr(Address) << "\n"); - if (Buf.size() - Offset < 8 * NumEntries) - return make_error_code(llvm::errc::io_error); + uint64_t InputOffset = 0, OutputOffset = 0; for (uint32_t J = 0; J < NumEntries; ++J) { - const uint32_t OutputAddr = DE.getU32(&Offset); - const uint32_t InputAddr = DE.getU32(&Offset); - Map.insert(std::pair(OutputAddr, InputAddr)); - LLVM_DEBUG(dbgs() << Twine::utohexstr(OutputAddr) << " -> " - << Twine::utohexstr(InputAddr) << "\n"); + const uint64_t OutputDelta = DE.getULEB128(&Offset, &Err); + const int64_t InputDelta = DE.getSLEB128(&Offset, &Err); + OutputOffset += OutputDelta; + InputOffset += InputDelta; + Map.insert(std::pair(OutputOffset, InputOffset)); + LLVM_DEBUG(dbgs() << formatv("{0:x} -> {1:x} ({2}/{3}b -> {4}/{5}b)\n", + OutputOffset, InputOffset, OutputDelta, + encodeULEB128(OutputDelta, nulls()), + InputDelta, + encodeSLEB128(InputDelta, nulls()))); } Maps.insert(std::pair(Address, Map)); } - - if (Buf.size() - Offset < 4) - return make_error_code(llvm::errc::io_error); - - const uint32_t NumColdEntries = DE.getU32(&Offset); - LLVM_DEBUG(dbgs() << "Parsing " << NumColdEntries << " cold part mappings\n"); - for (uint32_t I = 0; I < NumColdEntries; ++I) { - if (Buf.size() - Offset < 16) - return make_error_code(llvm::errc::io_error); - const uint32_t ColdAddress = DE.getU64(&Offset); - const uint32_t HotAddress = DE.getU64(&Offset); - ColdPartSource.insert( - std::pair(ColdAddress, HotAddress)); - LLVM_DEBUG(dbgs() << Twine::utohexstr(ColdAddress) << " -> " - << Twine::utohexstr(HotAddress) << "\n"); - } - outs() << "BOLT-INFO: Parsed " << Maps.size() << " BAT entries\n"; - outs() << "BOLT-INFO: Parsed " << NumColdEntries - << " BAT cold-to-hot entries\n"; - - return std::error_code(); } void BoltAddressTranslation::dump(raw_ostream &OS) { @@ -209,7 +224,7 @@ void BoltAddressTranslation::dump(raw_ostream &OS) { OS << "BB mappings:\n"; for (const auto &Entry : MapEntry.second) { const bool IsBranch = Entry.second & BRANCHENTRY; - const uint32_t Val = Entry.second & ~BRANCHENTRY; + const uint32_t Val = Entry.second >> 1; // dropping BRANCHENTRY bit OS << "0x" << Twine::utohexstr(Entry.first) << " -> " << "0x" << Twine::utohexstr(Val); if (IsBranch) @@ -244,7 +259,7 @@ uint64_t BoltAddressTranslation::translate(uint64_t FuncAddress, --KeyVal; - const uint32_t Val = KeyVal->second & ~BRANCHENTRY; + const uint32_t Val = KeyVal->second >> 1; // dropping BRANCHENTRY bit // Branch source addresses are translated to the first instruction of the // source BB to avoid accounting for modifications BOLT may have made in the // BB regarding deletion/addition of instructions. diff --git a/bolt/lib/Rewrite/CMakeLists.txt b/bolt/lib/Rewrite/CMakeLists.txt index fb21c13c654b3..6890f52e2b28b 100644 --- a/bolt/lib/Rewrite/CMakeLists.txt +++ b/bolt/lib/Rewrite/CMakeLists.txt @@ -5,8 +5,8 @@ set(LLVM_LINK_COMPONENTS MC Object Support - DWARFLinkerBase DWARFLinker + DWARFLinkerClassic AsmPrinter TargetParser ) diff --git a/bolt/lib/Rewrite/RewriteInstance.cpp b/bolt/lib/Rewrite/RewriteInstance.cpp index a95b1650753cf..f5a8a5b716874 100644 --- a/bolt/lib/Rewrite/RewriteInstance.cpp +++ b/bolt/lib/Rewrite/RewriteInstance.cpp @@ -4112,6 +4112,7 @@ void RewriteInstance::encodeBATSection() { copyByteArray(BoltInfo), BoltInfo.size(), /*Alignment=*/1, /*IsReadOnly=*/true, ELF::SHT_NOTE); + outs() << "BOLT-INFO: BAT section size (bytes): " << BoltInfo.size() << '\n'; } template diff --git a/bolt/test/X86/bolt-address-translation.test b/bolt/test/X86/bolt-address-translation.test index f68a8f7e9bcb7..430b4cb007310 100644 --- a/bolt/test/X86/bolt-address-translation.test +++ b/bolt/test/X86/bolt-address-translation.test @@ -36,7 +36,7 @@ # # CHECK: BOLT: 3 out of 7 functions were overwritten. # CHECK: BOLT-INFO: Wrote 6 BAT maps -# CHECK: BOLT-INFO: Wrote 3 BAT cold-to-hot entries +# CHECK: BOLT-INFO: BAT section size (bytes): 404 # # usqrt mappings (hot part). We match against any key (left side containing # the bolted binary offsets) because BOLT may change where it puts instructions diff --git a/bolt/tools/bat-dump/bat-dump.cpp b/bolt/tools/bat-dump/bat-dump.cpp index fa53c5b122f88..71efe008d409b 100644 --- a/bolt/tools/bat-dump/bat-dump.cpp +++ b/bolt/tools/bat-dump/bat-dump.cpp @@ -75,7 +75,7 @@ static std::string GetExecutablePath(const char *Argv0) { if (llvm::ErrorOr P = llvm::sys::findProgramByName(ExecutablePath)) ExecutablePath = *P; - return std::string(ExecutablePath.str()); + return std::string(ExecutablePath); } void dumpBATFor(llvm::object::ELFObjectFileBase *InputFile) { diff --git a/bolt/tools/driver/llvm-bolt.cpp b/bolt/tools/driver/llvm-bolt.cpp index 5a3af6a44b522..cc215a5256d2b 100644 --- a/bolt/tools/driver/llvm-bolt.cpp +++ b/bolt/tools/driver/llvm-bolt.cpp @@ -174,7 +174,7 @@ static std::string GetExecutablePath(const char *Argv0) { if (llvm::ErrorOr P = llvm::sys::findProgramByName(ExecutablePath)) ExecutablePath = *P; - return std::string(ExecutablePath.str()); + return std::string(ExecutablePath); } int main(int argc, char **argv) { diff --git a/bolt/tools/heatmap/heatmap.cpp b/bolt/tools/heatmap/heatmap.cpp index f5204fa8ec60b..9b190dd288b27 100644 --- a/bolt/tools/heatmap/heatmap.cpp +++ b/bolt/tools/heatmap/heatmap.cpp @@ -46,7 +46,7 @@ static std::string GetExecutablePath(const char *Argv0) { if (llvm::ErrorOr P = llvm::sys::findProgramByName(ExecutablePath)) ExecutablePath = *P; - return std::string(ExecutablePath.str()); + return std::string(ExecutablePath); } int main(int argc, char **argv) { diff --git a/clang-tools-extra/clang-doc/HTMLGenerator.cpp b/clang-tools-extra/clang-doc/HTMLGenerator.cpp index d2466857e5110..f6355ccf4f844 100644 --- a/clang-tools-extra/clang-doc/HTMLGenerator.cpp +++ b/clang-tools-extra/clang-doc/HTMLGenerator.cpp @@ -457,7 +457,7 @@ writeFileDefinition(const Location &L, Node->Children.emplace_back(std::make_unique(" of file ")); auto LocFileNode = std::make_unique( HTMLTag::TAG_A, llvm::sys::path::filename(FileURL)); - LocFileNode->Attributes.emplace_back("href", std::string(FileURL.str())); + LocFileNode->Attributes.emplace_back("href", std::string(FileURL)); Node->Children.emplace_back(std::move(LocFileNode)); return Node; } diff --git a/clang-tools-extra/clang-tidy/GlobList.cpp b/clang-tools-extra/clang-tidy/GlobList.cpp index 694db35106fd6..dfe3f7c505b17 100644 --- a/clang-tools-extra/clang-tidy/GlobList.cpp +++ b/clang-tools-extra/clang-tidy/GlobList.cpp @@ -16,11 +16,7 @@ namespace clang::tidy { // from the GlobList. static bool consumeNegativeIndicator(StringRef &GlobList) { GlobList = GlobList.trim(); - if (GlobList.starts_with("-")) { - GlobList = GlobList.substr(1); - return true; - } - return false; + return GlobList.consume_front("-"); } // Converts first glob from the comma-separated list of globs to Regex and diff --git a/clang-tools-extra/clang-tidy/bugprone/ExceptionEscapeCheck.cpp b/clang-tools-extra/clang-tidy/bugprone/ExceptionEscapeCheck.cpp index 90bf523ffb00b..620a57194acb8 100644 --- a/clang-tools-extra/clang-tidy/bugprone/ExceptionEscapeCheck.cpp +++ b/clang-tools-extra/clang-tidy/bugprone/ExceptionEscapeCheck.cpp @@ -28,6 +28,10 @@ AST_MATCHER(FunctionDecl, isExplicitThrow) { Node.getExceptionSpecSourceRange().isValid(); } +AST_MATCHER(FunctionDecl, hasAtLeastOneParameter) { + return Node.getNumParams() > 0; +} + } // namespace ExceptionEscapeCheck::ExceptionEscapeCheck(StringRef Name, @@ -58,14 +62,16 @@ void ExceptionEscapeCheck::storeOptions(ClangTidyOptions::OptionMap &Opts) { void ExceptionEscapeCheck::registerMatchers(MatchFinder *Finder) { Finder->addMatcher( - functionDecl(isDefinition(), - anyOf(isNoThrow(), - allOf(anyOf(cxxDestructorDecl(), - cxxConstructorDecl(isMoveConstructor()), - cxxMethodDecl(isMoveAssignmentOperator()), - isMain(), hasName("swap")), - unless(isExplicitThrow())), - isEnabled(FunctionsThatShouldNotThrow))) + functionDecl( + isDefinition(), + anyOf(isNoThrow(), + allOf(anyOf(cxxDestructorDecl(), + cxxConstructorDecl(isMoveConstructor()), + cxxMethodDecl(isMoveAssignmentOperator()), isMain(), + allOf(hasAnyName("swap", "iter_swap", "iter_move"), + hasAtLeastOneParameter())), + unless(isExplicitThrow())), + isEnabled(FunctionsThatShouldNotThrow))) .bind("thrower"), this); } diff --git a/clang-tools-extra/clang-tidy/bugprone/ImplicitWideningOfMultiplicationResultCheck.cpp b/clang-tools-extra/clang-tidy/bugprone/ImplicitWideningOfMultiplicationResultCheck.cpp index 95a7c521eabb0..6f22f02f30183 100644 --- a/clang-tools-extra/clang-tidy/bugprone/ImplicitWideningOfMultiplicationResultCheck.cpp +++ b/clang-tools-extra/clang-tidy/bugprone/ImplicitWideningOfMultiplicationResultCheck.cpp @@ -9,6 +9,7 @@ #include "ImplicitWideningOfMultiplicationResultCheck.h" #include "clang/AST/ASTContext.h" #include "clang/ASTMatchers/ASTMatchFinder.h" +#include "clang/Lex/Lexer.h" #include using namespace clang::ast_matchers; @@ -99,15 +100,16 @@ void ImplicitWideningOfMultiplicationResultCheck::handleImplicitCastExpr( "make conversion explicit to silence this warning", DiagnosticIDs::Note) << E->getSourceRange(); - + const SourceLocation EndLoc = Lexer::getLocForEndOfToken( + E->getEndLoc(), 0, *Result->SourceManager, getLangOpts()); if (ShouldUseCXXStaticCast) Diag << FixItHint::CreateInsertion( E->getBeginLoc(), "static_cast<" + Ty.getAsString() + ">(") - << FixItHint::CreateInsertion(E->getEndLoc(), ")"); + << FixItHint::CreateInsertion(EndLoc, ")"); else Diag << FixItHint::CreateInsertion(E->getBeginLoc(), "(" + Ty.getAsString() + ")(") - << FixItHint::CreateInsertion(E->getEndLoc(), ")"); + << FixItHint::CreateInsertion(EndLoc, ")"); Diag << includeStddefHeader(E->getBeginLoc()); } @@ -137,7 +139,11 @@ void ImplicitWideningOfMultiplicationResultCheck::handleImplicitCastExpr( Diag << FixItHint::CreateInsertion(LHS->getBeginLoc(), "static_cast<" + WideExprTy.getAsString() + ">(") - << FixItHint::CreateInsertion(LHS->getEndLoc(), ")"); + << FixItHint::CreateInsertion( + Lexer::getLocForEndOfToken(LHS->getEndLoc(), 0, + *Result->SourceManager, + getLangOpts()), + ")"); else Diag << FixItHint::CreateInsertion(LHS->getBeginLoc(), "(" + WideExprTy.getAsString() + ")"); @@ -206,16 +212,17 @@ void ImplicitWideningOfMultiplicationResultCheck::handlePointerOffsetting( "make conversion explicit to silence this warning", DiagnosticIDs::Note) << IndexExpr->getSourceRange(); - + const SourceLocation EndLoc = Lexer::getLocForEndOfToken( + IndexExpr->getEndLoc(), 0, *Result->SourceManager, getLangOpts()); if (ShouldUseCXXStaticCast) Diag << FixItHint::CreateInsertion( IndexExpr->getBeginLoc(), (Twine("static_cast<") + TyAsString + ">(").str()) - << FixItHint::CreateInsertion(IndexExpr->getEndLoc(), ")"); + << FixItHint::CreateInsertion(EndLoc, ")"); else Diag << FixItHint::CreateInsertion(IndexExpr->getBeginLoc(), (Twine("(") + TyAsString + ")(").str()) - << FixItHint::CreateInsertion(IndexExpr->getEndLoc(), ")"); + << FixItHint::CreateInsertion(EndLoc, ")"); Diag << includeStddefHeader(IndexExpr->getBeginLoc()); } @@ -229,7 +236,11 @@ void ImplicitWideningOfMultiplicationResultCheck::handlePointerOffsetting( Diag << FixItHint::CreateInsertion( LHS->getBeginLoc(), (Twine("static_cast<") + TyAsString + ">(").str()) - << FixItHint::CreateInsertion(LHS->getEndLoc(), ")"); + << FixItHint::CreateInsertion( + Lexer::getLocForEndOfToken(IndexExpr->getEndLoc(), 0, + *Result->SourceManager, + getLangOpts()), + ")"); else Diag << FixItHint::CreateInsertion(LHS->getBeginLoc(), (Twine("(") + TyAsString + ")").str()); diff --git a/clang-tools-extra/clang-tidy/fuchsia/MultipleInheritanceCheck.cpp b/clang-tools-extra/clang-tidy/fuchsia/MultipleInheritanceCheck.cpp index 8aecd3ba27b2e..b5ce23ae8feda 100644 --- a/clang-tools-extra/clang-tidy/fuchsia/MultipleInheritanceCheck.cpp +++ b/clang-tools-extra/clang-tidy/fuchsia/MultipleInheritanceCheck.cpp @@ -53,7 +53,7 @@ bool MultipleInheritanceCheck::isCurrentClassInterface( // Interfaces should have exclusively pure methods. return llvm::none_of(Node->methods(), [](const CXXMethodDecl *M) { - return M->isUserProvided() && !M->isPure() && !M->isStatic(); + return M->isUserProvided() && !M->isPureVirtual() && !M->isStatic(); }); } @@ -103,8 +103,8 @@ void MultipleInheritanceCheck::check(const MatchFinder::MatchResult &Result) { const auto *Base = cast(Ty->getDecl()->getDefinition()); if (!isInterface(Base)) NumConcrete++; } - - // Check virtual bases to see if there is more than one concrete + + // Check virtual bases to see if there is more than one concrete // non-virtual base. for (const auto &V : D->vbases()) { const auto *Ty = V.getType()->getAs(); diff --git a/clang-tools-extra/clang-tidy/misc/StaticAssertCheck.cpp b/clang-tools-extra/clang-tidy/misc/StaticAssertCheck.cpp index 77763a9f429f2..35536b47700a6 100644 --- a/clang-tools-extra/clang-tidy/misc/StaticAssertCheck.cpp +++ b/clang-tools-extra/clang-tidy/misc/StaticAssertCheck.cpp @@ -7,6 +7,7 @@ //===----------------------------------------------------------------------===// #include "StaticAssertCheck.h" +#include "../utils/Matchers.h" #include "clang/AST/ASTContext.h" #include "clang/AST/Expr.h" #include "clang/ASTMatchers/ASTMatchFinder.h" @@ -45,13 +46,20 @@ void StaticAssertCheck::registerMatchers(MatchFinder *Finder) { IsAlwaysFalse); auto NonConstexprFunctionCall = callExpr(hasDeclaration(functionDecl(unless(isConstexpr())))); + auto NonConstexprVariableReference = + declRefExpr(to(varDecl(unless(isConstexpr()))), + unless(hasAncestor(expr(matchers::hasUnevaluatedContext()))), + unless(hasAncestor(typeLoc()))); + + auto NonConstexprCode = + expr(anyOf(NonConstexprFunctionCall, NonConstexprVariableReference)); auto AssertCondition = expr( anyOf(expr(ignoringParenCasts(anyOf( AssertExprRoot, unaryOperator(hasUnaryOperand( ignoringParenCasts(AssertExprRoot)))))), anything()), - unless(findAll(NonConstexprFunctionCall))) + unless(NonConstexprCode), unless(hasDescendant(NonConstexprCode))) .bind("condition"); auto Condition = anyOf(ignoringParenImpCasts(callExpr( diff --git a/clang-tools-extra/clang-tidy/misc/UnusedUsingDeclsCheck.cpp b/clang-tools-extra/clang-tidy/misc/UnusedUsingDeclsCheck.cpp index 051375263e53c..59e487bab3119 100644 --- a/clang-tools-extra/clang-tidy/misc/UnusedUsingDeclsCheck.cpp +++ b/clang-tools-extra/clang-tidy/misc/UnusedUsingDeclsCheck.cpp @@ -117,8 +117,10 @@ void UnusedUsingDeclsCheck::check(const MatchFinder::MatchResult &Result) { /*SkipTrailingWhitespaceAndNewLine=*/true)); for (const auto *UsingShadow : Using->shadows()) { const auto *TargetDecl = UsingShadow->getTargetDecl()->getCanonicalDecl(); - if (shouldCheckDecl(TargetDecl)) + if (shouldCheckDecl(TargetDecl)) { Context.UsingTargetDecls.insert(TargetDecl); + UsingTargetDeclsCache.insert(TargetDecl); + } } if (!Context.UsingTargetDecls.empty()) Contexts.push_back(Context); @@ -201,13 +203,16 @@ void UnusedUsingDeclsCheck::check(const MatchFinder::MatchResult &Result) { void UnusedUsingDeclsCheck::removeFromFoundDecls(const Decl *D) { if (!D) return; + const Decl *CanonicalDecl = D->getCanonicalDecl(); + if (!UsingTargetDeclsCache.contains(CanonicalDecl)) + return; // FIXME: Currently, we don't handle the using-decls being used in different // scopes (such as different namespaces, different functions). Instead of // giving an incorrect message, we mark all of them as used. - // - // FIXME: Use a more efficient way to find a matching context. for (auto &Context : Contexts) { - if (Context.UsingTargetDecls.contains(D->getCanonicalDecl())) + if (Context.IsUsed) + continue; + if (Context.UsingTargetDecls.contains(CanonicalDecl)) Context.IsUsed = true; } } @@ -224,6 +229,7 @@ void UnusedUsingDeclsCheck::onEndOfTranslationUnit() { } } Contexts.clear(); + UsingTargetDeclsCache.clear(); } } // namespace clang::tidy::misc diff --git a/clang-tools-extra/clang-tidy/misc/UnusedUsingDeclsCheck.h b/clang-tools-extra/clang-tidy/misc/UnusedUsingDeclsCheck.h index 498b3ffd2678c..7bdaf12e8aece 100644 --- a/clang-tools-extra/clang-tidy/misc/UnusedUsingDeclsCheck.h +++ b/clang-tools-extra/clang-tidy/misc/UnusedUsingDeclsCheck.h @@ -49,6 +49,7 @@ class UnusedUsingDeclsCheck : public ClangTidyCheck { }; std::vector Contexts; + llvm::SmallPtrSet UsingTargetDeclsCache; StringRef RawStringHeaderFileExtensions; FileExtensionsSet HeaderFileExtensions; diff --git a/clang-tools-extra/clang-tidy/modernize/UseEqualsDeleteCheck.cpp b/clang-tools-extra/clang-tidy/modernize/UseEqualsDeleteCheck.cpp index 059a0af60d3ee..9561cc71183d9 100644 --- a/clang-tools-extra/clang-tidy/modernize/UseEqualsDeleteCheck.cpp +++ b/clang-tools-extra/clang-tidy/modernize/UseEqualsDeleteCheck.cpp @@ -17,11 +17,12 @@ namespace clang::tidy::modernize { namespace { AST_MATCHER(FunctionDecl, hasAnyDefinition) { - if (Node.hasBody() || Node.isPure() || Node.isDefaulted() || Node.isDeleted()) + if (Node.hasBody() || Node.isPureVirtual() || Node.isDefaulted() || + Node.isDeleted()) return true; if (const FunctionDecl *Definition = Node.getDefinition()) - if (Definition->hasBody() || Definition->isPure() || + if (Definition->hasBody() || Definition->isPureVirtual() || Definition->isDefaulted() || Definition->isDeleted()) return true; diff --git a/clang-tools-extra/clang-tidy/performance/ForRangeCopyCheck.cpp b/clang-tools-extra/clang-tidy/performance/ForRangeCopyCheck.cpp index 5bfa6fb0d02d5..655e480ffa62c 100644 --- a/clang-tools-extra/clang-tidy/performance/ForRangeCopyCheck.cpp +++ b/clang-tools-extra/clang-tidy/performance/ForRangeCopyCheck.cpp @@ -97,6 +97,15 @@ bool ForRangeCopyCheck::handleConstValueCopy(const VarDecl &LoopVar, return true; } +static bool isReferenced(const VarDecl &LoopVar, const Stmt &Stmt, + ASTContext &Context) { + const auto IsLoopVar = varDecl(equalsNode(&LoopVar)); + return !match(stmt(hasDescendant(declRefExpr(to(valueDecl(anyOf( + IsLoopVar, bindingDecl(forDecomposition(IsLoopVar)))))))), + Stmt, Context) + .empty(); +} + bool ForRangeCopyCheck::handleCopyIsOnlyConstReferenced( const VarDecl &LoopVar, const CXXForRangeStmt &ForRange, ASTContext &Context) { @@ -113,9 +122,7 @@ bool ForRangeCopyCheck::handleCopyIsOnlyConstReferenced( // compiler warning which can't be suppressed. // Since this case is very rare, it is safe to ignore it. if (!ExprMutationAnalyzer(*ForRange.getBody(), Context).isMutated(&LoopVar) && - !utils::decl_ref_expr::allDeclRefExprs(LoopVar, *ForRange.getBody(), - Context) - .empty()) { + isReferenced(LoopVar, *ForRange.getBody(), Context)) { auto Diag = diag( LoopVar.getLocation(), "loop variable is copied but only used as const reference; consider " diff --git a/clang-tools-extra/clang-tidy/performance/NoexceptSwapCheck.cpp b/clang-tools-extra/clang-tidy/performance/NoexceptSwapCheck.cpp index 25a58af74f7ee..e7cba6e54e86a 100644 --- a/clang-tools-extra/clang-tidy/performance/NoexceptSwapCheck.cpp +++ b/clang-tools-extra/clang-tidy/performance/NoexceptSwapCheck.cpp @@ -39,7 +39,8 @@ void NoexceptSwapCheck::registerMatchers(MatchFinder *Finder) { .bind("type"))))), hasParameter(1, hasType(qualType(hasCanonicalType( qualType(equalsBoundNode("type"))))))); - Finder->addMatcher(functionDecl(unless(isDeleted()), hasName("swap"), + Finder->addMatcher(functionDecl(unless(isDeleted()), + hasAnyName("swap", "iter_swap"), anyOf(MethodMatcher, FunctionMatcher)) .bind(BindFuncDeclName), this); diff --git a/clang-tools-extra/clang-tidy/readability/AvoidNestedConditionalOperatorCheck.cpp b/clang-tools-extra/clang-tidy/readability/AvoidNestedConditionalOperatorCheck.cpp new file mode 100644 index 0000000000000..1b62f54d5557d --- /dev/null +++ b/clang-tools-extra/clang-tidy/readability/AvoidNestedConditionalOperatorCheck.cpp @@ -0,0 +1,52 @@ +//===--- AvoidNestedConditionalOperatorCheck.cpp - clang-tidy -------------===// +// +// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. +// See https://llvm.org/LICENSE.txt for license information. +// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception +// +//===----------------------------------------------------------------------===// + +#include "AvoidNestedConditionalOperatorCheck.h" +#include "clang/ASTMatchers/ASTMatchFinder.h" +#include "clang/ASTMatchers/ASTMatchers.h" +#include "clang/Basic/DiagnosticIDs.h" + +using namespace clang::ast_matchers; + +namespace clang::tidy::readability { + +void AvoidNestedConditionalOperatorCheck::registerMatchers( + MatchFinder *Finder) { + Finder->addMatcher( + conditionalOperator( + anyOf( + hasCondition(ignoringParenCasts( + conditionalOperator().bind("nested-conditional-operator"))), + hasTrueExpression(ignoringParenCasts( + conditionalOperator().bind("nested-conditional-operator"))), + hasFalseExpression(ignoringParenCasts( + conditionalOperator().bind("nested-conditional-operator"))))) + .bind("conditional-operator"), + this); +} + +void AvoidNestedConditionalOperatorCheck::check( + const MatchFinder::MatchResult &Result) { + const auto *CO = + Result.Nodes.getNodeAs("conditional-operator"); + const auto *NCO = Result.Nodes.getNodeAs( + "nested-conditional-operator"); + assert(CO); + assert(NCO); + + if (CO->getBeginLoc().isMacroID() || NCO->getBeginLoc().isMacroID()) + return; + + diag(NCO->getBeginLoc(), + "conditional operator is used as sub-expression of parent conditional " + "operator, refrain from using nested conditional operators"); + diag(CO->getBeginLoc(), "parent conditional operator here", + DiagnosticIDs::Note); +} + +} // namespace clang::tidy::readability diff --git a/clang-tools-extra/clang-tidy/readability/AvoidNestedConditionalOperatorCheck.h b/clang-tools-extra/clang-tidy/readability/AvoidNestedConditionalOperatorCheck.h new file mode 100644 index 0000000000000..9010156de6ce2 --- /dev/null +++ b/clang-tools-extra/clang-tidy/readability/AvoidNestedConditionalOperatorCheck.h @@ -0,0 +1,33 @@ +//===--- AvoidNestedConditionalOperatorCheck.h - clang-tidy -----*- C++ -*-===// +// +// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. +// See https://llvm.org/LICENSE.txt for license information. +// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception +// +//===----------------------------------------------------------------------===// + +#ifndef LLVM_CLANG_TOOLS_EXTRA_CLANG_TIDY_READABILITY_AVOID_NESTED_CONDITIONAL_OPERATOR_CHECK_H +#define LLVM_CLANG_TOOLS_EXTRA_CLANG_TIDY_READABILITY_AVOID_NESTED_CONDITIONAL_OPERATOR_CHECK_H + +#include "../ClangTidyCheck.h" + +namespace clang::tidy::readability { + +/// Identifies instances of nested conditional operators in the code. +/// +/// For the user-facing documentation see: +/// http://clang.llvm.org/extra/clang-tidy/checks/readability/avoid-nested-conditional-operator.html +class AvoidNestedConditionalOperatorCheck : public ClangTidyCheck { +public: + AvoidNestedConditionalOperatorCheck(StringRef Name, ClangTidyContext *Context) + : ClangTidyCheck(Name, Context) {} + void registerMatchers(ast_matchers::MatchFinder *Finder) override; + void check(const ast_matchers::MatchFinder::MatchResult &Result) override; + std::optional getCheckTraversalKind() const override { + return TK_IgnoreUnlessSpelledInSource; + } +}; + +} // namespace clang::tidy::readability + +#endif // LLVM_CLANG_TOOLS_EXTRA_CLANG_TIDY_READABILITY_AVOID_NESTED_CONDITIONAL_OPERATOR_CHECK_H diff --git a/clang-tools-extra/clang-tidy/readability/CMakeLists.txt b/clang-tools-extra/clang-tidy/readability/CMakeLists.txt index 408c822b861c5..fa571d5dd7650 100644 --- a/clang-tools-extra/clang-tidy/readability/CMakeLists.txt +++ b/clang-tools-extra/clang-tidy/readability/CMakeLists.txt @@ -5,6 +5,7 @@ set(LLVM_LINK_COMPONENTS add_clang_library(clangTidyReadabilityModule AvoidConstParamsInDecls.cpp + AvoidNestedConditionalOperatorCheck.cpp AvoidReturnWithVoidValueCheck.cpp AvoidUnconditionalPreprocessorIfCheck.cpp BracesAroundStatementsCheck.cpp diff --git a/clang-tools-extra/clang-tidy/readability/ContainerSizeEmptyCheck.cpp b/clang-tools-extra/clang-tidy/readability/ContainerSizeEmptyCheck.cpp index f0357fb49eff3..19307b4cdcbe3 100644 --- a/clang-tools-extra/clang-tidy/readability/ContainerSizeEmptyCheck.cpp +++ b/clang-tools-extra/clang-tidy/readability/ContainerSizeEmptyCheck.cpp @@ -291,10 +291,14 @@ void ContainerSizeEmptyCheck::check(const MatchFinder::MatchResult &Result) { OpCode == BinaryOperatorKind::BO_NE)) return; - // Always true, no warnings for that. - if ((OpCode == BinaryOperatorKind::BO_GE && Value == 0 && ContainerIsLHS) || - (OpCode == BinaryOperatorKind::BO_LE && Value == 0 && !ContainerIsLHS)) - return; + // Always true/false, no warnings for that. + if (Value == 0) { + if ((OpCode == BinaryOperatorKind::BO_GT && !ContainerIsLHS) || + (OpCode == BinaryOperatorKind::BO_LT && ContainerIsLHS) || + (OpCode == BinaryOperatorKind::BO_LE && !ContainerIsLHS) || + (OpCode == BinaryOperatorKind::BO_GE && ContainerIsLHS)) + return; + } // Do not warn for size > 1, 1 < size, size <= 1, 1 >= size. if (Value == 1) { @@ -306,12 +310,32 @@ void ContainerSizeEmptyCheck::check(const MatchFinder::MatchResult &Result) { return; } + // Do not warn for size < 1, 1 > size, size <= 0, 0 >= size for non signed + // types + if ((OpCode == BinaryOperatorKind::BO_GT && Value == 1 && + !ContainerIsLHS) || + (OpCode == BinaryOperatorKind::BO_LT && Value == 1 && ContainerIsLHS) || + (OpCode == BinaryOperatorKind::BO_GE && Value == 0 && + !ContainerIsLHS) || + (OpCode == BinaryOperatorKind::BO_LE && Value == 0 && ContainerIsLHS)) { + const Expr *Container = ContainerIsLHS + ? BinaryOp->getLHS()->IgnoreImpCasts() + : BinaryOp->getRHS()->IgnoreImpCasts(); + if (Container->getType() + .getCanonicalType() + .getNonReferenceType() + ->isSignedIntegerType()) + return; + } + if (OpCode == BinaryOperatorKind::BO_NE && Value == 0) Negation = true; + if ((OpCode == BinaryOperatorKind::BO_GT || OpCode == BinaryOperatorKind::BO_GE) && ContainerIsLHS) Negation = true; + if ((OpCode == BinaryOperatorKind::BO_LT || OpCode == BinaryOperatorKind::BO_LE) && !ContainerIsLHS) diff --git a/clang-tools-extra/clang-tidy/readability/ImplicitBoolConversionCheck.cpp b/clang-tools-extra/clang-tidy/readability/ImplicitBoolConversionCheck.cpp index f0fca30de3b3c..11706ffb5b7d4 100644 --- a/clang-tools-extra/clang-tidy/readability/ImplicitBoolConversionCheck.cpp +++ b/clang-tools-extra/clang-tidy/readability/ImplicitBoolConversionCheck.cpp @@ -7,6 +7,7 @@ //===----------------------------------------------------------------------===// #include "ImplicitBoolConversionCheck.h" +#include "../utils/FixItHintUtils.h" #include "clang/AST/ASTContext.h" #include "clang/ASTMatchers/ASTMatchFinder.h" #include "clang/Lex/Lexer.h" @@ -61,31 +62,6 @@ bool isUnaryLogicalNotOperator(const Stmt *Statement) { return UnaryOperatorExpr && UnaryOperatorExpr->getOpcode() == UO_LNot; } -bool areParensNeededForOverloadedOperator(OverloadedOperatorKind OperatorKind) { - switch (OperatorKind) { - case OO_New: - case OO_Delete: // Fall-through on purpose. - case OO_Array_New: - case OO_Array_Delete: - case OO_ArrowStar: - case OO_Arrow: - case OO_Call: - case OO_Subscript: - return false; - - default: - return true; - } -} - -bool areParensNeededForStatement(const Stmt *Statement) { - if (const auto *OperatorCall = dyn_cast(Statement)) { - return areParensNeededForOverloadedOperator(OperatorCall->getOperator()); - } - - return isa(Statement) || isa(Statement); -} - void fixGenericExprCastToBool(DiagnosticBuilder &Diag, const ImplicitCastExpr *Cast, const Stmt *Parent, ASTContext &Context) { @@ -105,9 +81,10 @@ void fixGenericExprCastToBool(DiagnosticBuilder &Diag, const Expr *SubExpr = Cast->getSubExpr(); - bool NeedInnerParens = areParensNeededForStatement(SubExpr); + bool NeedInnerParens = + SubExpr != nullptr && utils::fixit::areParensNeededForStatement(*SubExpr); bool NeedOuterParens = - Parent != nullptr && areParensNeededForStatement(Parent); + Parent != nullptr && utils::fixit::areParensNeededForStatement(*Parent); std::string StartLocInsertion; @@ -274,8 +251,7 @@ void ImplicitBoolConversionCheck::registerMatchers(MatchFinder *Finder) { allOf(anyOf(hasCastKind(CK_NullToPointer), hasCastKind(CK_NullToMemberPointer)), hasSourceExpression(cxxBoolLiteral()))), - hasSourceExpression(expr(hasType(booleanType()))), - unless(ExceptionCases)); + hasSourceExpression(expr(hasType(booleanType())))); auto BoolXor = binaryOperator(hasOperatorName("^"), hasLHS(ImplicitCastFromBool), hasRHS(ImplicitCastFromBool)); @@ -315,7 +291,7 @@ void ImplicitBoolConversionCheck::registerMatchers(MatchFinder *Finder) { traverse( TK_AsIs, implicitCastExpr( - ImplicitCastFromBool, + ImplicitCastFromBool, unless(ExceptionCases), // Exclude comparisons of bools, as they are always cast to // integers in such context: // bool_expr_a == bool_expr_b @@ -365,7 +341,7 @@ void ImplicitBoolConversionCheck::handleCastToBool(const ImplicitCastExpr *Cast, return; } - auto Diag = diag(Cast->getBeginLoc(), "implicit conversion %0 -> bool") + auto Diag = diag(Cast->getBeginLoc(), "implicit conversion %0 -> 'bool'") << Cast->getSubExpr()->getType(); StringRef EquivalentLiteral = @@ -382,7 +358,7 @@ void ImplicitBoolConversionCheck::handleCastFromBool( ASTContext &Context) { QualType DestType = NextImplicitCast ? NextImplicitCast->getType() : Cast->getType(); - auto Diag = diag(Cast->getBeginLoc(), "implicit conversion bool -> %0") + auto Diag = diag(Cast->getBeginLoc(), "implicit conversion 'bool' -> %0") << DestType; if (const auto *BoolLiteral = diff --git a/clang-tools-extra/clang-tidy/readability/ReadabilityTidyModule.cpp b/clang-tools-extra/clang-tidy/readability/ReadabilityTidyModule.cpp index 0b0aad7c0dcb3..f769752c5de5f 100644 --- a/clang-tools-extra/clang-tidy/readability/ReadabilityTidyModule.cpp +++ b/clang-tools-extra/clang-tidy/readability/ReadabilityTidyModule.cpp @@ -10,6 +10,7 @@ #include "../ClangTidyModule.h" #include "../ClangTidyModuleRegistry.h" #include "AvoidConstParamsInDecls.h" +#include "AvoidNestedConditionalOperatorCheck.h" #include "AvoidReturnWithVoidValueCheck.h" #include "AvoidUnconditionalPreprocessorIfCheck.h" #include "BracesAroundStatementsCheck.h" @@ -64,6 +65,8 @@ class ReadabilityModule : public ClangTidyModule { void addCheckFactories(ClangTidyCheckFactories &CheckFactories) override { CheckFactories.registerCheck( "readability-avoid-const-params-in-decls"); + CheckFactories.registerCheck( + "readability-avoid-nested-conditional-operator"); CheckFactories.registerCheck( "readability-avoid-return-with-void-value"); CheckFactories.registerCheck( diff --git a/clang-tools-extra/clang-tidy/readability/RedundantMemberInitCheck.cpp b/clang-tools-extra/clang-tidy/readability/RedundantMemberInitCheck.cpp index b5d407773bb73..015347ee9294c 100644 --- a/clang-tools-extra/clang-tidy/readability/RedundantMemberInitCheck.cpp +++ b/clang-tools-extra/clang-tidy/readability/RedundantMemberInitCheck.cpp @@ -7,6 +7,7 @@ //===----------------------------------------------------------------------===// #include "RedundantMemberInitCheck.h" +#include "../utils/LexerUtils.h" #include "../utils/Matchers.h" #include "clang/AST/ASTContext.h" #include "clang/ASTMatchers/ASTMatchFinder.h" @@ -18,33 +19,64 @@ using namespace clang::tidy::matchers; namespace clang::tidy::readability { +static SourceRange +getFullInitRangeInclWhitespaces(SourceRange Range, const SourceManager &SM, + const LangOptions &LangOpts) { + const Token PrevToken = + utils::lexer::getPreviousToken(Range.getBegin(), SM, LangOpts, false); + if (PrevToken.is(tok::unknown)) + return Range; + + if (PrevToken.isNot(tok::equal)) + return {PrevToken.getEndLoc(), Range.getEnd()}; + + return getFullInitRangeInclWhitespaces( + {PrevToken.getLocation(), Range.getEnd()}, SM, LangOpts); +} + void RedundantMemberInitCheck::storeOptions(ClangTidyOptions::OptionMap &Opts) { Options.store(Opts, "IgnoreBaseInCopyConstructors", IgnoreBaseInCopyConstructors); } void RedundantMemberInitCheck::registerMatchers(MatchFinder *Finder) { + auto ConstructorMatcher = + cxxConstructExpr(argumentCountIs(0), + hasDeclaration(cxxConstructorDecl(ofClass(cxxRecordDecl( + unless(isTriviallyDefaultConstructible())))))) + .bind("construct"); + Finder->addMatcher( cxxConstructorDecl( unless(isDelegatingConstructor()), ofClass(unless(isUnion())), forEachConstructorInitializer( - cxxCtorInitializer( - withInitializer( - cxxConstructExpr( - hasDeclaration( - cxxConstructorDecl(ofClass(cxxRecordDecl( - unless(isTriviallyDefaultConstructible())))))) - .bind("construct")), - unless(forField(hasType(isConstQualified()))), - unless(forField(hasParent(recordDecl(isUnion()))))) + cxxCtorInitializer(withInitializer(ConstructorMatcher), + unless(forField(fieldDecl( + anyOf(hasType(isConstQualified()), + hasParent(recordDecl(isUnion()))))))) .bind("init"))) .bind("constructor"), this); + + Finder->addMatcher(fieldDecl(hasInClassInitializer(ConstructorMatcher), + unless(hasParent(recordDecl(isUnion())))) + .bind("field"), + this); } void RedundantMemberInitCheck::check(const MatchFinder::MatchResult &Result) { - const auto *Init = Result.Nodes.getNodeAs("init"); const auto *Construct = Result.Nodes.getNodeAs("construct"); + + if (const auto *Field = Result.Nodes.getNodeAs("field")) { + const Expr *Init = Field->getInClassInitializer(); + diag(Construct->getExprLoc(), "initializer for member %0 is redundant") + << Field + << FixItHint::CreateRemoval(getFullInitRangeInclWhitespaces( + Init->getSourceRange(), *Result.SourceManager, getLangOpts())); + return; + } + + const auto *Init = Result.Nodes.getNodeAs("init"); const auto *ConstructorDecl = Result.Nodes.getNodeAs("constructor"); @@ -52,18 +84,15 @@ void RedundantMemberInitCheck::check(const MatchFinder::MatchResult &Result) { Init->isBaseInitializer()) return; - if (Construct->getNumArgs() == 0 || - Construct->getArg(0)->isDefaultArgument()) { - if (Init->isAnyMemberInitializer()) { - diag(Init->getSourceLocation(), "initializer for member %0 is redundant") - << Init->getAnyMember() - << FixItHint::CreateRemoval(Init->getSourceRange()); - } else { - diag(Init->getSourceLocation(), - "initializer for base class %0 is redundant") - << Construct->getType() - << FixItHint::CreateRemoval(Init->getSourceRange()); - } + if (Init->isAnyMemberInitializer()) { + diag(Init->getSourceLocation(), "initializer for member %0 is redundant") + << Init->getAnyMember() + << FixItHint::CreateRemoval(Init->getSourceRange()); + } else { + diag(Init->getSourceLocation(), + "initializer for base class %0 is redundant") + << Construct->getType() + << FixItHint::CreateRemoval(Init->getSourceRange()); } } diff --git a/clang-tools-extra/clang-tidy/readability/SimplifyBooleanExprCheck.cpp b/clang-tools-extra/clang-tidy/readability/SimplifyBooleanExprCheck.cpp index 26d9287f07049..edb67614bd558 100644 --- a/clang-tools-extra/clang-tidy/readability/SimplifyBooleanExprCheck.cpp +++ b/clang-tools-extra/clang-tidy/readability/SimplifyBooleanExprCheck.cpp @@ -277,7 +277,13 @@ class SimplifyBooleanExprCheck::Visitor : public RecursiveASTVisitor { } bool dataTraverseStmtPre(Stmt *S) { - if (S && !shouldIgnore(S)) + if (!S) { + return true; + } + if (Check->IgnoreMacros && S->getBeginLoc().isMacroID()) { + return false; + } + if (!shouldIgnore(S)) StmtStack.push_back(S); return true; } @@ -583,6 +589,7 @@ class SimplifyBooleanExprCheck::Visitor : public RecursiveASTVisitor { SimplifyBooleanExprCheck::SimplifyBooleanExprCheck(StringRef Name, ClangTidyContext *Context) : ClangTidyCheck(Name, Context), + IgnoreMacros(Options.get("IgnoreMacros", false)), ChainedConditionalReturn(Options.get("ChainedConditionalReturn", false)), ChainedConditionalAssignment( Options.get("ChainedConditionalAssignment", false)), @@ -671,6 +678,7 @@ void SimplifyBooleanExprCheck::reportBinOp(const ASTContext &Context, } void SimplifyBooleanExprCheck::storeOptions(ClangTidyOptions::OptionMap &Opts) { + Options.store(Opts, "IgnoreMacros", IgnoreMacros); Options.store(Opts, "ChainedConditionalReturn", ChainedConditionalReturn); Options.store(Opts, "ChainedConditionalAssignment", ChainedConditionalAssignment); diff --git a/clang-tools-extra/clang-tidy/readability/SimplifyBooleanExprCheck.h b/clang-tools-extra/clang-tidy/readability/SimplifyBooleanExprCheck.h index c4dad24ec3998..ccc6f3d879fc0 100644 --- a/clang-tools-extra/clang-tidy/readability/SimplifyBooleanExprCheck.h +++ b/clang-tools-extra/clang-tidy/readability/SimplifyBooleanExprCheck.h @@ -64,6 +64,7 @@ class SimplifyBooleanExprCheck : public ClangTidyCheck { StringRef Description, SourceRange ReplacementRange, StringRef Replacement); + const bool IgnoreMacros; const bool ChainedConditionalReturn; const bool ChainedConditionalAssignment; const bool SimplifyDeMorgan; diff --git a/clang-tools-extra/clang-tidy/utils/FixItHintUtils.cpp b/clang-tools-extra/clang-tidy/utils/FixItHintUtils.cpp index 226dd60b5bf5f..bbdd4326b0bac 100644 --- a/clang-tools-extra/clang-tidy/utils/FixItHintUtils.cpp +++ b/clang-tools-extra/clang-tidy/utils/FixItHintUtils.cpp @@ -224,6 +224,40 @@ std::optional addQualifierToVarDecl(const VarDecl &Var, return std::nullopt; } +bool areParensNeededForStatement(const Stmt &Node) { + if (isa(&Node)) + return false; + + if (isa(&Node) || isa(&Node)) + return true; + + if (isa(&Node) || + isa(&Node)) + return true; + + if (const auto *Op = dyn_cast(&Node)) { + switch (Op->getOperator()) { + case OO_PlusPlus: + [[fallthrough]]; + case OO_MinusMinus: + return Op->getNumArgs() != 2; + case OO_Call: + [[fallthrough]]; + case OO_Subscript: + [[fallthrough]]; + case OO_Arrow: + return false; + default: + return true; + }; + } + + if (isa(&Node)) + return true; + + return false; +} + // Return true if expr needs to be put in parens when it is an argument of a // prefix unary operator, e.g. when it is a binary or ternary operator // syntactically. diff --git a/clang-tools-extra/clang-tidy/utils/FixItHintUtils.h b/clang-tools-extra/clang-tidy/utils/FixItHintUtils.h index 15894c4bf5cf8..d25bea2c6d297 100644 --- a/clang-tools-extra/clang-tidy/utils/FixItHintUtils.h +++ b/clang-tools-extra/clang-tidy/utils/FixItHintUtils.h @@ -47,6 +47,10 @@ addQualifierToVarDecl(const VarDecl &Var, const ASTContext &Context, // \brief Format a pointer to an expression std::string formatDereference(const Expr &ExprNode, const ASTContext &Context); + +// \brief Checks whatever a expression require extra () to be always used in +// safe way in any other expression. +bool areParensNeededForStatement(const Stmt &Node); } // namespace clang::tidy::utils::fixit #endif // LLVM_CLANG_TOOLS_EXTRA_CLANG_TIDY_UTILS_FIXITHINTUTILS_H diff --git a/clang-tools-extra/clang-tidy/utils/HeaderGuard.cpp b/clang-tools-extra/clang-tidy/utils/HeaderGuard.cpp index 80580bc9888a8..3540c496515be 100644 --- a/clang-tools-extra/clang-tidy/utils/HeaderGuard.cpp +++ b/clang-tools-extra/clang-tidy/utils/HeaderGuard.cpp @@ -281,7 +281,7 @@ void HeaderGuardCheck::registerPPCallbacks(const SourceManager &SM, std::string HeaderGuardCheck::sanitizeHeaderGuard(StringRef Guard) { // Only reserved identifiers are allowed to start with an '_'. - return Guard.drop_while([](char C) { return C == '_'; }).str(); + return Guard.ltrim('_').str(); } bool HeaderGuardCheck::shouldSuggestEndifComment(StringRef FileName) { diff --git a/clang-tools-extra/clangd/CompileCommands.cpp b/clang-tools-extra/clangd/CompileCommands.cpp index f4e8e7e74a3be..166f17e9b3f71 100644 --- a/clang-tools-extra/clangd/CompileCommands.cpp +++ b/clang-tools-extra/clangd/CompileCommands.cpp @@ -313,13 +313,15 @@ void CommandMangler::operator()(tooling::CompileCommand &Command, tooling::addTargetAndModeForProgramName(Cmd, Cmd.front()); - // Check whether the flag exists, either as -flag or -flag=* - auto Has = [&](llvm::StringRef Flag) { - for (llvm::StringRef Arg : Cmd) { - if (Arg.consume_front(Flag) && (Arg.empty() || Arg[0] == '=')) - return true; - } - return false; + // Check whether the flag exists in the command. + auto HasExact = [&](llvm::StringRef Flag) { + return llvm::any_of(Cmd, [&](llvm::StringRef Arg) { return Arg == Flag; }); + }; + + // Check whether the flag appears in the command as a prefix. + auto HasPrefix = [&](llvm::StringRef Flag) { + return llvm::any_of( + Cmd, [&](llvm::StringRef Arg) { return Arg.starts_with(Flag); }); }; llvm::erase_if(Cmd, [](llvm::StringRef Elem) { @@ -327,12 +329,13 @@ void CommandMangler::operator()(tooling::CompileCommand &Command, }); std::vector ToAppend; - if (ResourceDir && !Has("-resource-dir")) + if (ResourceDir && !HasExact("-resource-dir") && !HasPrefix("-resource-dir=")) ToAppend.push_back(("-resource-dir=" + *ResourceDir)); // Don't set `-isysroot` if it is already set or if `--sysroot` is set. // `--sysroot` is a superset of the `-isysroot` argument. - if (Sysroot && !Has("-isysroot") && !Has("--sysroot")) { + if (Sysroot && !HasPrefix("-isysroot") && !HasExact("--sysroot") && + !HasPrefix("--sysroot=")) { ToAppend.push_back("-isysroot"); ToAppend.push_back(*Sysroot); } @@ -343,7 +346,7 @@ void CommandMangler::operator()(tooling::CompileCommand &Command, } if (!Cmd.empty()) { - bool FollowSymlink = !Has("-no-canonical-prefixes"); + bool FollowSymlink = !HasExact("-no-canonical-prefixes"); Cmd.front() = (FollowSymlink ? ResolvedDrivers : ResolvedDriversNoFollow) .get(Cmd.front(), [&, this] { diff --git a/clang-tools-extra/clangd/Selection.cpp b/clang-tools-extra/clangd/Selection.cpp index 8c6d5750ecefd..277cb8769a1b1 100644 --- a/clang-tools-extra/clangd/Selection.cpp +++ b/clang-tools-extra/clangd/Selection.cpp @@ -1113,6 +1113,9 @@ const DeclContext &SelectionTree::Node::getDeclContext() const { return *DC; return *Current->getLexicalDeclContext(); } + if (const auto *LE = CurrentNode->ASTNode.get()) + if (CurrentNode != this) + return *LE->getCallOperator(); } llvm_unreachable("A tree must always be rooted at TranslationUnitDecl."); } diff --git a/clang-tools-extra/clangd/SemanticHighlighting.cpp b/clang-tools-extra/clangd/SemanticHighlighting.cpp index 37939d36425a9..ee3772e3d380c 100644 --- a/clang-tools-extra/clangd/SemanticHighlighting.cpp +++ b/clang-tools-extra/clangd/SemanticHighlighting.cpp @@ -265,7 +265,7 @@ bool isStatic(const Decl *D) { bool isAbstract(const Decl *D) { if (const auto *CMD = llvm::dyn_cast(D)) - return CMD->isPure(); + return CMD->isPureVirtual(); if (const auto *CRD = llvm::dyn_cast(D)) return CRD->hasDefinition() && CRD->isAbstract(); return false; diff --git a/clang-tools-extra/clangd/XRefs.cpp b/clang-tools-extra/clangd/XRefs.cpp index 5f41f788a6939..250d571eea191 100644 --- a/clang-tools-extra/clangd/XRefs.cpp +++ b/clang-tools-extra/clangd/XRefs.cpp @@ -427,7 +427,7 @@ locateASTReferent(SourceLocation CurLoc, const syntax::Token *TouchedIdentifier, // Special case: virtual void ^method() = 0: jump to all overrides. // FIXME: extend it to ^virtual, unfortunately, virtual location is not // saved in the AST. - if (CMD->isPure()) { + if (CMD->isPureVirtual()) { if (TouchedIdentifier && SM.getSpellingLoc(CMD->getLocation()) == TouchedIdentifier->location()) { VirtualMethods.insert(getSymbolID(CMD)); diff --git a/clang-tools-extra/clangd/unittests/CompileCommandsTests.cpp b/clang-tools-extra/clangd/unittests/CompileCommandsTests.cpp index 772177b60b5ee..b64dd4acad4c2 100644 --- a/clang-tools-extra/clangd/unittests/CompileCommandsTests.cpp +++ b/clang-tools-extra/clangd/unittests/CompileCommandsTests.cpp @@ -370,9 +370,8 @@ TEST(ArgStripperTest, OrderDependent) { } TEST(PrintArgvTest, All) { - std::vector Args = { - "one", "two", "thr ee", "f\"o\"ur", "fi\\ve", "$" - }; + std::vector Args = {"one", "two", "thr ee", + "f\"o\"ur", "fi\\ve", "$"}; const char *Expected = R"(one two "thr ee" "f\"o\"ur" "fi\\ve" $)"; EXPECT_EQ(Expected, printArgv(Args)); } @@ -450,6 +449,82 @@ TEST(CommandMangler, PathsAsPositional) { Mangler(Cmd, "a.cc"); EXPECT_THAT(Cmd.CommandLine, Contains("foo")); } + +TEST(CommandMangler, RespectsOriginalResourceDir) { + auto Mangler = CommandMangler::forTests(); + Mangler.ResourceDir = testPath("fake/resources"); + + { + tooling::CompileCommand Cmd; + Cmd.CommandLine = {"clang++", "-resource-dir", testPath("true/resources"), + "foo.cc"}; + Mangler(Cmd, "foo.cc"); + EXPECT_THAT(llvm::join(Cmd.CommandLine, " "), + HasSubstr("-resource-dir " + testPath("true/resources"))); + EXPECT_THAT(llvm::join(Cmd.CommandLine, " "), + Not(HasSubstr(testPath("fake/resources")))); + } + + { + tooling::CompileCommand Cmd; + Cmd.CommandLine = {"clang++", "-resource-dir=" + testPath("true/resources"), + "foo.cc"}; + Mangler(Cmd, "foo.cc"); + EXPECT_THAT(llvm::join(Cmd.CommandLine, " "), + HasSubstr("-resource-dir=" + testPath("true/resources"))); + EXPECT_THAT(llvm::join(Cmd.CommandLine, " "), + Not(HasSubstr(testPath("fake/resources")))); + } +} + +TEST(CommandMangler, RespectsOriginalSysroot) { + auto Mangler = CommandMangler::forTests(); + Mangler.Sysroot = testPath("fake/sysroot"); + + { + tooling::CompileCommand Cmd; + Cmd.CommandLine = {"clang++", "-isysroot", testPath("true/sysroot"), + "foo.cc"}; + Mangler(Cmd, "foo.cc"); + EXPECT_THAT(llvm::join(Cmd.CommandLine, " "), + HasSubstr("-isysroot " + testPath("true/sysroot"))); + EXPECT_THAT(llvm::join(Cmd.CommandLine, " "), + Not(HasSubstr(testPath("fake/sysroot")))); + } + + { + tooling::CompileCommand Cmd; + Cmd.CommandLine = {"clang++", "-isysroot" + testPath("true/sysroot"), + "foo.cc"}; + Mangler(Cmd, "foo.cc"); + EXPECT_THAT(llvm::join(Cmd.CommandLine, " "), + HasSubstr("-isysroot" + testPath("true/sysroot"))); + EXPECT_THAT(llvm::join(Cmd.CommandLine, " "), + Not(HasSubstr(testPath("fake/sysroot")))); + } + + { + tooling::CompileCommand Cmd; + Cmd.CommandLine = {"clang++", "--sysroot", testPath("true/sysroot"), + "foo.cc"}; + Mangler(Cmd, "foo.cc"); + EXPECT_THAT(llvm::join(Cmd.CommandLine, " "), + HasSubstr("--sysroot " + testPath("true/sysroot"))); + EXPECT_THAT(llvm::join(Cmd.CommandLine, " "), + Not(HasSubstr(testPath("fake/sysroot")))); + } + + { + tooling::CompileCommand Cmd; + Cmd.CommandLine = {"clang++", "--sysroot=" + testPath("true/sysroot"), + "foo.cc"}; + Mangler(Cmd, "foo.cc"); + EXPECT_THAT(llvm::join(Cmd.CommandLine, " "), + HasSubstr("--sysroot=" + testPath("true/sysroot"))); + EXPECT_THAT(llvm::join(Cmd.CommandLine, " "), + Not(HasSubstr(testPath("fake/sysroot")))); + } +} } // namespace } // namespace clangd } // namespace clang diff --git a/clang-tools-extra/clangd/unittests/DumpASTTests.cpp b/clang-tools-extra/clangd/unittests/DumpASTTests.cpp index d1b8f21b82c65..304682118c871 100644 --- a/clang-tools-extra/clangd/unittests/DumpASTTests.cpp +++ b/clang-tools-extra/clangd/unittests/DumpASTTests.cpp @@ -186,6 +186,17 @@ TEST(DumpASTTests, Arcana) { EXPECT_THAT(Node.children.front().arcana, testing::StartsWith("QualType ")); } +TEST(DumpASTTests, UnbalancedBraces) { + // Test that we don't crash while trying to compute a source range for the + // node whose ending brace is missing, and also that the source range is + // not empty. + Annotations Case("/*error-ok*/ $func[[int main() {]]"); + ParsedAST AST = TestTU::withCode(Case.code()).build(); + auto Node = dumpAST(DynTypedNode::create(findDecl(AST, "main")), + AST.getTokens(), AST.getASTContext()); + ASSERT_EQ(Node.range, Case.range("func")); +} + } // namespace } // namespace clangd } // namespace clang diff --git a/clang-tools-extra/clangd/unittests/SelectionTests.cpp b/clang-tools-extra/clangd/unittests/SelectionTests.cpp index 4c019a1524f3c..754e8c287c514 100644 --- a/clang-tools-extra/clangd/unittests/SelectionTests.cpp +++ b/clang-tools-extra/clangd/unittests/SelectionTests.cpp @@ -880,6 +880,19 @@ TEST(SelectionTest, DeclContextIsLexical) { } } +TEST(SelectionTest, DeclContextLambda) { + llvm::Annotations Test(R"cpp( + void foo(); + auto lambda = [] { + return $1^foo(); + }; + )cpp"); + auto AST = TestTU::withCode(Test.code()).build(); + auto ST = SelectionTree::createRight(AST.getASTContext(), AST.getTokens(), + Test.point("1"), Test.point("1")); + EXPECT_TRUE(ST.commonAncestor()->getDeclContext().isFunctionOrMethod()); +} + } // namespace } // namespace clangd } // namespace clang diff --git a/clang-tools-extra/clangd/unittests/tweaks/AddUsingTests.cpp b/clang-tools-extra/clangd/unittests/tweaks/AddUsingTests.cpp index 1fd2487378d70..c2dd8e1bb8eef 100644 --- a/clang-tools-extra/clangd/unittests/tweaks/AddUsingTests.cpp +++ b/clang-tools-extra/clangd/unittests/tweaks/AddUsingTests.cpp @@ -309,6 +309,29 @@ namespace foo { void fun(); } void foo::fun() { ff(); })cpp"}, + // Inside a lambda. + { + R"cpp( +namespace NS { +void unrelated(); +void foo(); +} + +auto L = [] { + using NS::unrelated; + NS::f^oo(); +};)cpp", + R"cpp( +namespace NS { +void unrelated(); +void foo(); +} + +auto L = [] { + using NS::foo;using NS::unrelated; + foo(); +};)cpp", + }, // If all other using are fully qualified, add :: {R"cpp( #include "test.hpp" diff --git a/clang-tools-extra/docs/ReleaseNotes.rst b/clang-tools-extra/docs/ReleaseNotes.rst index b4d87e0ed2a67..d77267588db91 100644 --- a/clang-tools-extra/docs/ReleaseNotes.rst +++ b/clang-tools-extra/docs/ReleaseNotes.rst @@ -119,7 +119,7 @@ Improvements to clang-tidy - Improved `--dump-config` to print check options in alphabetical order. -- Improved :program:`clang-tidy-diff.py` script. +- Improved :program:`clang-tidy-diff.py` script. * Return exit code `1` if any :program:`clang-tidy` subprocess exits with a non-zero code or if exporting fixes fails. @@ -224,11 +224,10 @@ New checks Recommends the smallest possible underlying type for an ``enum`` or ``enum`` class based on the range of its enumerators. -- New :doc:`readability-reference-to-constructed-temporary - ` check. +- New :doc:`readability-avoid-nested-conditional-operator + ` check. - Detects C++ code where a reference variable is used to extend the lifetime - of a temporary object that has just been constructed. + Identifies instances of nested conditional operators in the code. - New :doc:`readability-avoid-return-with-void-value ` check. @@ -236,6 +235,12 @@ New checks Finds return statements with ``void`` values used within functions with ``void`` result types. +- New :doc:`readability-reference-to-constructed-temporary + ` check. + + Detects C++ code where a reference variable is used to extend the lifetime + of a temporary object that has just been constructed. + New check aliases ^^^^^^^^^^^^^^^^^ @@ -260,6 +265,14 @@ Changes in existing checks casting during type conversions at variable initialization, now with improved compatibility for C++17 and later versions. +- Improved :doc:`bugprone-exception-escape + ` check by extending the default + check function names to include ``iter_swap`` and ``iter_move``. + +- Improved :doc:`bugprone-implicit-widening-of-multiplication-result + ` check + to correctly emit fixes. + - Improved :doc:`bugprone-lambda-function-name ` check by adding option `IgnoreMacros` to ignore warnings in macros. @@ -381,9 +394,13 @@ Changes in existing checks ` check to ignore false-positives in unevaluated context (e.g., ``decltype``). +- Improved :doc:`misc-static-assert + ` check to ignore false-positives when + referring to non-``constexpr`` variables in non-unevaluated context. + - Improved :doc:`misc-unused-using-decls ` check to avoid false positive when - using in elaborated type and only check cpp files. + using in elaborated type and only check C++ files. - Improved :doc:`modernize-avoid-bind ` check to @@ -423,38 +440,45 @@ Changes in existing checks - Improved :doc:`modernize-use-using ` check to fix function pointer and - forward declared ``typedef`` correctly. Added option `IgnoreExternC` to ignore ``typedef`` - declaration in ``extern "C"`` scope. + forward declared ``typedef`` correctly. Added option `IgnoreExternC` to ignore + ``typedef`` declaration in ``extern "C"`` scope. - Improved :doc:`performance-faster-string-find ` check to properly escape single quotes. +- Improved :doc:`performance-for-range-copy + ` check to handle cases where + the loop variable is a structured binding. + - Improved :doc:`performance-noexcept-move-constructor ` to better handle - conditional noexcept expressions, eliminating false-positives. + conditional ``noexcept`` expressions, eliminating false-positives. - Improved :doc:`performance-noexcept-swap ` check to enforce a stricter match with the swap function signature and better handling of condition - noexcept expressions, eliminating false-positives. + ``noexcept`` expressions, eliminating false-positives. ``iter_swap`` function + name is checked by default. - Improved :doc:`readability-braces-around-statements ` check to ignore false-positive for ``if constexpr`` in lambda expression. - Improved :doc:`readability-avoid-const-params-in-decls - ` diagnositics to - highlight the const location + ` diagnostics to + highlight the ``const`` location - Improved :doc:`readability-container-contains ` to correctly handle - interger literals with suffixes in fix-its. + integer literals with suffixes in fix-its. - Improved :doc:`readability-container-size-empty ` check to detect comparison between string and empty string literals and support - ``length()`` method as an alternative to ``size()``. + ``length()`` method as an alternative to ``size()``. Resolved false positives + tied to negative values from size-like methods, and one triggered by size + checks below zero. - Improved :doc:`readability-function-size ` check configuration to use @@ -472,13 +496,14 @@ Changes in existing checks ``camel_Snake_Case`` now detect more invalid identifier names. Fields in anonymous records (i.e. anonymous structs and unions) now can be checked with the naming rules associated with their enclosing scopes rather than the naming - rules of public struct/union members. + rules of public ``struct``/``union`` members. - Improved :doc:`readability-implicit-bool-conversion ` check to take do-while loops into account for the `AllowIntegerConditions` and `AllowPointerConditions` options. It also now provides more consistent - suggestions when parentheses are added to the return value. + suggestions when parentheses are added to the return value or expressions. + It also ignores false-positives for comparison containing bool bitfield. - Improved :doc:`readability-misleading-indentation ` check to ignore @@ -488,6 +513,15 @@ Changes in existing checks ` check to ignore false-positives in initializer list of record. +- Improved :doc:`readability-redundant-member-init + ` check to now also + detect redundant in-class initializers. + +- Improved :doc:`readability-simplify-boolean-expr + ` check by adding the + new option `IgnoreMacros` that allows to ignore boolean expressions originating + from expanded macros. + - Improved :doc:`readability-simplify-subscript-expr ` check by extending the default value of the `Types` option to include ``std::span``. diff --git a/clang-tools-extra/docs/clang-tidy/checks/bugprone/exception-escape.rst b/clang-tools-extra/docs/clang-tidy/checks/bugprone/exception-escape.rst index e6aa8e001492a..182fade7f47a0 100644 --- a/clang-tools-extra/docs/clang-tidy/checks/bugprone/exception-escape.rst +++ b/clang-tools-extra/docs/clang-tidy/checks/bugprone/exception-escape.rst @@ -11,6 +11,8 @@ should not. The functions which should not throw exceptions are the following: * Move assignment operators * The ``main()`` functions * ``swap()`` functions +* ``iter_swap()`` functions +* ``iter_move()`` functions * Functions marked with ``throw()`` or ``noexcept`` * Other functions given as option diff --git a/clang-tools-extra/docs/clang-tidy/checks/list.rst b/clang-tools-extra/docs/clang-tidy/checks/list.rst index 2f86121ad8729..5f21449cfc3df 100644 --- a/clang-tools-extra/docs/clang-tidy/checks/list.rst +++ b/clang-tools-extra/docs/clang-tidy/checks/list.rst @@ -337,6 +337,7 @@ Clang-Tidy Checks :doc:`portability-simd-intrinsics `, :doc:`portability-std-allocator-const `, :doc:`readability-avoid-const-params-in-decls `, "Yes" + :doc:`readability-avoid-nested-conditional-operator `, :doc:`readability-avoid-return-with-void-value `, :doc:`readability-avoid-unconditional-preprocessor-if `, :doc:`readability-braces-around-statements `, "Yes" diff --git a/clang-tools-extra/docs/clang-tidy/checks/misc/const-correctness.rst b/clang-tools-extra/docs/clang-tidy/checks/misc/const-correctness.rst index b5a59fa691da4..86fba6c7e4f7c 100644 --- a/clang-tools-extra/docs/clang-tidy/checks/misc/const-correctness.rst +++ b/clang-tools-extra/docs/clang-tidy/checks/misc/const-correctness.rst @@ -8,7 +8,7 @@ This check implements detection of local variables which could be declared as coding guidelines, such as: `ES.25 `_ from the C++ Core Guidelines and `AUTOSAR C++14 Rule A7-1-1 (6.7.1 Specifiers) -`_. +`_. Please note that this check's analysis is type-based only. Variables that are not modified but used to create a non-const handle that might escape the scope are not diagnosed diff --git a/clang-tools-extra/docs/clang-tidy/checks/misc/unconventional-assign-operator.rst b/clang-tools-extra/docs/clang-tidy/checks/misc/unconventional-assign-operator.rst index 49e3fd5b6ee42..3b4b65a5cb683 100644 --- a/clang-tools-extra/docs/clang-tidy/checks/misc/unconventional-assign-operator.rst +++ b/clang-tools-extra/docs/clang-tidy/checks/misc/unconventional-assign-operator.rst @@ -13,3 +13,6 @@ types and definitions with good return type but wrong ``return`` statements. type (e.g. ``int``). * Private and deleted operators are ignored. * The operator must always return ``*this``. + +This check implements `AUTOSAR C++14 Rule A13-2-1 +`_. diff --git a/clang-tools-extra/docs/clang-tidy/checks/performance/noexcept-swap.rst b/clang-tools-extra/docs/clang-tidy/checks/performance/noexcept-swap.rst index fa4007618ba0f..2901d721706c8 100644 --- a/clang-tools-extra/docs/clang-tidy/checks/performance/noexcept-swap.rst +++ b/clang-tools-extra/docs/clang-tidy/checks/performance/noexcept-swap.rst @@ -3,11 +3,11 @@ performance-noexcept-swap ========================= -The check flags user-defined swap functions not marked with ``noexcept`` or +The check flags user-defined swap and iter_swap functions not marked with ``noexcept`` or marked with ``noexcept(expr)`` where ``expr`` evaluates to ``false`` (but is not a ``false`` literal itself). -When a swap function is marked as ``noexcept``, it assures the compiler that +When a swap or iter_swap function is marked as ``noexcept``, it assures the compiler that no exceptions will be thrown during the swapping of two objects, which allows the compiler to perform certain optimizations such as omitting exception handling code. diff --git a/clang-tools-extra/docs/clang-tidy/checks/readability/avoid-nested-conditional-operator.rst b/clang-tools-extra/docs/clang-tidy/checks/readability/avoid-nested-conditional-operator.rst new file mode 100644 index 0000000000000..44b74283292ce --- /dev/null +++ b/clang-tools-extra/docs/clang-tidy/checks/readability/avoid-nested-conditional-operator.rst @@ -0,0 +1,21 @@ +.. title:: clang-tidy - readability-avoid-nested-conditional-operator + +readability-avoid-nested-conditional-operator +============================================= + +Identifies instances of nested conditional operators in the code. + +Nested conditional operators, also known as ternary operators, can contribute +to reduced code readability and comprehension. So they should be split as +several statements and stored the intermediate results in temporary variable. + +Examples: + +.. code-block:: c++ + + int NestInConditional = (condition1 ? true1 : false1) ? true2 : false2; + int NestInTrue = condition1 ? (condition2 ? true1 : false1) : false2; + int NestInFalse = condition1 ? true1 : condition2 ? true2 : false1; + +This check implements part of `AUTOSAR C++14 Rule A5-16-1 +`_. diff --git a/clang-tools-extra/docs/clang-tidy/checks/readability/redundant-member-init.rst b/clang-tools-extra/docs/clang-tidy/checks/readability/redundant-member-init.rst index b2f86c4429cc5..c26aaf73b16f8 100644 --- a/clang-tools-extra/docs/clang-tidy/checks/readability/redundant-member-init.rst +++ b/clang-tools-extra/docs/clang-tidy/checks/readability/redundant-member-init.rst @@ -11,13 +11,14 @@ Example .. code-block:: c++ - // Explicitly initializing the member s is unnecessary. + // Explicitly initializing the member s and v is unnecessary. class Foo { public: Foo() : s() {} private: std::string s; + std::vector v {}; }; Options diff --git a/clang-tools-extra/docs/clang-tidy/checks/readability/simplify-boolean-expr.rst b/clang-tools-extra/docs/clang-tidy/checks/readability/simplify-boolean-expr.rst index 18ab84b26a259..3d00d5b043a60 100644 --- a/clang-tools-extra/docs/clang-tidy/checks/readability/simplify-boolean-expr.rst +++ b/clang-tools-extra/docs/clang-tidy/checks/readability/simplify-boolean-expr.rst @@ -82,6 +82,11 @@ Examples: Options ------- +.. option:: IgnoreMacros + + If `true`, ignore boolean expressions originating from expanded macros. + Default is `false`. + .. option:: ChainedConditionalReturn If `true`, conditional boolean return statements at the end of an @@ -99,8 +104,8 @@ Options .. option:: SimplifyDeMorganRelaxed - If `true`, :option:`SimplifyDeMorgan` will also transform negated - conjunctions and disjunctions where there is no negation on either operand. + If `true`, :option:`SimplifyDeMorgan` will also transform negated + conjunctions and disjunctions where there is no negation on either operand. This option has no effect if :option:`SimplifyDeMorgan` is `false`. Default is `false`. diff --git a/clang-tools-extra/modularize/ModularizeUtilities.cpp b/clang-tools-extra/modularize/ModularizeUtilities.cpp index 25de9c9dd05ac..d25bb09b8cad5 100644 --- a/clang-tools-extra/modularize/ModularizeUtilities.cpp +++ b/clang-tools-extra/modularize/ModularizeUtilities.cpp @@ -195,7 +195,7 @@ std::error_code ModularizeUtilities::loadSingleHeaderListsAndDependencies( // Get canonical form. HeaderFileName = getCanonicalPath(HeaderFileName); // Save the resulting header file path and dependencies. - HeaderFileNames.push_back(std::string(HeaderFileName.str())); + HeaderFileNames.push_back(std::string(HeaderFileName)); Dependencies[HeaderFileName.str()] = Dependents; } return std::error_code(); @@ -248,7 +248,7 @@ std::error_code ModularizeUtilities::loadProblemHeaderList( // Get canonical form. HeaderFileName = getCanonicalPath(HeaderFileName); // Save the resulting header file path. - ProblemFileNames.push_back(std::string(HeaderFileName.str())); + ProblemFileNames.push_back(std::string(HeaderFileName)); } return std::error_code(); } diff --git a/clang-tools-extra/test/clang-tidy/checkers/bugprone/exception-escape.cpp b/clang-tools-extra/test/clang-tidy/checkers/bugprone/exception-escape.cpp index 4a7149e81ce7e..f5e74df1621ce 100644 --- a/clang-tools-extra/test/clang-tidy/checkers/bugprone/exception-escape.cpp +++ b/clang-tools-extra/test/clang-tidy/checkers/bugprone/exception-escape.cpp @@ -586,6 +586,16 @@ void swap(int&, int&) { throw 1; } +void iter_swap(int&, int&) { + // CHECK-MESSAGES: :[[@LINE-1]]:6: warning: an exception may be thrown in function 'iter_swap' which should not throw exceptions + throw 1; +} + +void iter_move(int&) { + // CHECK-MESSAGES: :[[@LINE-1]]:6: warning: an exception may be thrown in function 'iter_move' which should not throw exceptions + throw 1; +} + namespace std { class bad_alloc {}; } diff --git a/clang-tools-extra/test/clang-tidy/checkers/bugprone/implicit-widening-of-multiplication-result-array-subscript-expression.cpp b/clang-tools-extra/test/clang-tidy/checkers/bugprone/implicit-widening-of-multiplication-result-array-subscript-expression.cpp index 2881341c87857..5b432cb266296 100644 --- a/clang-tools-extra/test/clang-tidy/checkers/bugprone/implicit-widening-of-multiplication-result-array-subscript-expression.cpp +++ b/clang-tools-extra/test/clang-tidy/checkers/bugprone/implicit-widening-of-multiplication-result-array-subscript-expression.cpp @@ -18,7 +18,7 @@ char *t0(char *base, int a, int b) { // CHECK-NOTES-CXX: static_cast( ) // CHECK-NOTES-ALL: :[[@LINE-5]]:16: note: perform multiplication in a wider type // CHECK-NOTES-C: (ptrdiff_t) - // CHECK-NOTES-CXX: static_cast() + // CHECK-NOTES-CXX: static_cast( ) } void *t1(char *base, int a, int b) { return &((a * b)[base]); @@ -35,7 +35,7 @@ char *t2(char *base, unsigned int a, int b) { // CHECK-NOTES-CXX: static_cast( ) // CHECK-NOTES-ALL: :[[@LINE-5]]:16: note: perform multiplication in a wider type // CHECK-NOTES-C: (size_t) - // CHECK-NOTES-CXX: static_cast() + // CHECK-NOTES-CXX: static_cast( ) } char *t3(char *base, int a, unsigned int b) { diff --git a/clang-tools-extra/test/clang-tidy/checkers/bugprone/implicit-widening-of-multiplication-result-int.cpp b/clang-tools-extra/test/clang-tidy/checkers/bugprone/implicit-widening-of-multiplication-result-int.cpp index cf13a7ea9fbd8..8619a4ee86550 100644 --- a/clang-tools-extra/test/clang-tidy/checkers/bugprone/implicit-widening-of-multiplication-result-int.cpp +++ b/clang-tools-extra/test/clang-tidy/checkers/bugprone/implicit-widening-of-multiplication-result-int.cpp @@ -18,7 +18,7 @@ long t0(int a, int b) { // CHECK-NOTES-CXX: static_cast( ) // CHECK-NOTES-ALL: :[[@LINE-5]]:10: note: perform multiplication in a wider type // CHECK-NOTES-C: (long) - // CHECK-NOTES-CXX: static_cast() + // CHECK-NOTES-CXX: static_cast( ) } unsigned long t1(int a, int b) { return a * b; @@ -28,7 +28,7 @@ unsigned long t1(int a, int b) { // CHECK-NOTES-CXX: static_cast( ) // CHECK-NOTES-ALL: :[[@LINE-5]]:10: note: perform multiplication in a wider type // CHECK-NOTES-C: (long) - // CHECK-NOTES-CXX: static_cast() + // CHECK-NOTES-CXX: static_cast( ) } long t2(unsigned int a, int b) { @@ -39,7 +39,7 @@ long t2(unsigned int a, int b) { // CHECK-NOTES-CXX: static_cast( ) // CHECK-NOTES-ALL: :[[@LINE-5]]:10: note: perform multiplication in a wider type // CHECK-NOTES-C: (unsigned long) - // CHECK-NOTES-CXX: static_cast() + // CHECK-NOTES-CXX: static_cast( ) } unsigned long t3(unsigned int a, int b) { return a * b; @@ -49,7 +49,7 @@ unsigned long t3(unsigned int a, int b) { // CHECK-NOTES-CXX: static_cast( ) // CHECK-NOTES-ALL: :[[@LINE-5]]:10: note: perform multiplication in a wider type // CHECK-NOTES-C: (unsigned long) - // CHECK-NOTES-CXX: static_cast() + // CHECK-NOTES-CXX: static_cast( ) } long t4(int a, unsigned int b) { diff --git a/clang-tools-extra/test/clang-tidy/checkers/bugprone/implicit-widening-of-multiplication-result-pointer-offset.cpp b/clang-tools-extra/test/clang-tidy/checkers/bugprone/implicit-widening-of-multiplication-result-pointer-offset.cpp index c812885507e8f..5ba6a5639849d 100644 --- a/clang-tools-extra/test/clang-tidy/checkers/bugprone/implicit-widening-of-multiplication-result-pointer-offset.cpp +++ b/clang-tools-extra/test/clang-tidy/checkers/bugprone/implicit-widening-of-multiplication-result-pointer-offset.cpp @@ -18,7 +18,7 @@ char *t0(char *base, int a, int b) { // CHECK-NOTES-CXX: static_cast( ) // CHECK-NOTES-ALL: :[[@LINE-5]]:17: note: perform multiplication in a wider type // CHECK-NOTES-C: (ptrdiff_t) - // CHECK-NOTES-CXX: static_cast() + // CHECK-NOTES-CXX: static_cast( ) } char *t1(char *base, int a, int b) { return a * b + base; @@ -35,7 +35,7 @@ char *t2(char *base, unsigned int a, int b) { // CHECK-NOTES-CXX: static_cast( ) // CHECK-NOTES-ALL: :[[@LINE-5]]:17: note: perform multiplication in a wider type // CHECK-NOTES-C: (size_t) - // CHECK-NOTES-CXX: static_cast() + // CHECK-NOTES-CXX: static_cast( ) } char *t3(char *base, int a, unsigned int b) { diff --git a/clang-tools-extra/test/clang-tidy/checkers/misc/static-assert.cpp b/clang-tools-extra/test/clang-tidy/checkers/misc/static-assert.cpp index da4ab9baa3948..16c6ba60c1925 100644 --- a/clang-tools-extra/test/clang-tidy/checkers/misc/static-assert.cpp +++ b/clang-tools-extra/test/clang-tidy/checkers/misc/static-assert.cpp @@ -20,6 +20,48 @@ void print(...); #define my_macro() assert(0 == 1) // CHECK-FIXES: #define my_macro() assert(0 == 1) +namespace PR24066 { + +void referenceMember() { + struct { + int A; + int B; + } S; + assert(&S.B - &S.A == 1); +} + +const int X = 1; +void referenceVariable() { + assert(X > 0); +} + + +constexpr int Y = 1; +void referenceConstexprVariable() { + assert(Y > 0); + // CHECK-MESSAGES: :[[@LINE-1]]:3: warning: found assert() that could be replaced by static_assert() [misc-static-assert] + // CHECK-FIXES-CXX11: {{^ }}static_assert(Y > 0, ""); + // CHECK-FIXES-CXX17: {{^ }}static_assert(Y > 0); +} + +void useInSizeOf() { + char a = 0; + assert(sizeof(a) == 1U); + // CHECK-MESSAGES: :[[@LINE-1]]:3: warning: found assert() that could be replaced by static_assert() [misc-static-assert] + // CHECK-FIXES-CXX11: {{^ }}static_assert(sizeof(a) == 1U, ""); + // CHECK-FIXES-CXX17: {{^ }}static_assert(sizeof(a) == 1U); +} + +void useInDecltype() { + char a = 0; + assert(static_cast(256) == 0); + // CHECK-MESSAGES: :[[@LINE-1]]:3: warning: found assert() that could be replaced by static_assert() [misc-static-assert] + // CHECK-FIXES-CXX11: {{^ }}static_assert(static_cast(256) == 0, ""); + // CHECK-FIXES-CXX17: {{^ }}static_assert(static_cast(256) == 0); +} + +} + constexpr bool myfunc(int a, int b) { return a * b == 0; } typedef __SIZE_TYPE__ size_t; diff --git a/clang-tools-extra/test/clang-tidy/checkers/performance/for-range-copy.cpp b/clang-tools-extra/test/clang-tidy/checkers/performance/for-range-copy.cpp index 1a2eedc9e65c5..f9d06898ca03d 100644 --- a/clang-tools-extra/test/clang-tidy/checkers/performance/for-range-copy.cpp +++ b/clang-tools-extra/test/clang-tidy/checkers/performance/for-range-copy.cpp @@ -47,6 +47,11 @@ struct S { S &operator=(const S &); }; +struct Point { + ~Point() {} + int x, y; +}; + struct Convertible { operator S() const { return S(); @@ -87,6 +92,10 @@ void instantiated() { // CHECK-MESSAGES: [[@LINE-1]]:16: warning: the loop variable's type is {{.*}} // CHECK-FIXES: {{^}} for (const S& S2 : View>()) {} + for (const auto [X, Y] : View>()) {} + // CHECK-MESSAGES: [[@LINE-1]]:19: warning: the loop variable's type is + // CHECK-FIXES: {{^}} for (const auto& [X, Y] : View>()) {} + for (const T T2 : View>()) {} // CHECK-MESSAGES: [[@LINE-1]]:16: warning: the loop variable's type is {{.*}} // CHECK-FIXES: {{^}} for (const T& T2 : View>()) {} @@ -123,11 +132,6 @@ struct Mutable { ~Mutable() {} }; -struct Point { - ~Point() {} - int x, y; -}; - Mutable& operator<<(Mutable &Out, bool B) { Out.setBool(B); return Out; @@ -144,6 +148,7 @@ void useByValue(Mutable M); void useByConstValue(const Mutable M); void mutate(Mutable *M); void mutate(Mutable &M); +void mutate(int &); void onceConstOnceMutated(const Mutable &M1, Mutable &M2); void negativeVariableIsMutated() { @@ -234,6 +239,22 @@ void positiveOnlyAccessedFieldAsConst() { } } +void positiveOnlyUsedAsConstBinding() { + for (auto [X, Y] : View>()) { + // CHECK-MESSAGES: [[@LINE-1]]:13: warning: loop variable is copied but + // CHECK-FIXES: for (const auto& [X, Y] : View>()) { + use(X); + use(Y); + } +} + +void negativeMutatedBinding() { + for (auto [X, Y] : View>()) { + use(X); + mutate(Y); + } +} + void positiveOnlyUsedInCopyConstructor() { for (auto A : View>()) { // CHECK-MESSAGES: [[@LINE-1]]:13: warning: loop variable is copied but only used as const reference; consider making it a const reference [performance-for-range-copy] diff --git a/clang-tools-extra/test/clang-tidy/checkers/performance/noexcept-swap.cpp b/clang-tools-extra/test/clang-tidy/checkers/performance/noexcept-swap.cpp index dfc71a2bb9ab3..6b266707d77f8 100644 --- a/clang-tools-extra/test/clang-tidy/checkers/performance/noexcept-swap.cpp +++ b/clang-tools-extra/test/clang-tidy/checkers/performance/noexcept-swap.cpp @@ -32,6 +32,10 @@ void swap(A &, A &); // CHECK-MESSAGES: :[[@LINE-1]]:6: warning: swap functions should be marked noexcept [performance-noexcept-swap] // CHECK-FIXES: void swap(A &, A &) noexcept ; +void iter_swap(A &, A &); +// CHECK-MESSAGES: :[[@LINE-1]]:6: warning: swap functions should be marked noexcept [performance-noexcept-swap] +// CHECK-FIXES: void iter_swap(A &, A &) noexcept ; + struct B { static constexpr bool kFalse = false; void swap(B &) noexcept(kFalse); diff --git a/clang-tools-extra/test/clang-tidy/checkers/readability/avoid-nested-conditional-operator.cpp b/clang-tools-extra/test/clang-tidy/checkers/readability/avoid-nested-conditional-operator.cpp new file mode 100644 index 0000000000000..847df08199e34 --- /dev/null +++ b/clang-tools-extra/test/clang-tidy/checkers/readability/avoid-nested-conditional-operator.cpp @@ -0,0 +1,24 @@ +// RUN: %check_clang_tidy %s readability-avoid-nested-conditional-operator %t + +int NestInConditional = (true ? true : false) ? 1 : 2; +// CHECK-MESSAGES: :[[@LINE-1]]:26: warning: conditional operator is used as sub-expression of parent conditional operator, refrain from using nested conditional operators +// CHECK-MESSAGES: :[[@LINE-2]]:25: note: parent conditional operator here + +int NestInTrue = true ? (true ? 1 : 2) : 2; +// CHECK-MESSAGES: :[[@LINE-1]]:26: warning: conditional operator is used as sub-expression of parent conditional operator, refrain from using nested conditional operators +// CHECK-MESSAGES: :[[@LINE-2]]:18: note: parent conditional operator here + +int NestInFalse = true ? 1 : true ? 1 : 2; +// CHECK-MESSAGES: :[[@LINE-1]]:30: warning: conditional operator is used as sub-expression of parent conditional operator, refrain from using nested conditional operators +// CHECK-MESSAGES: :[[@LINE-2]]:19: note: parent conditional operator here +int NestInFalse2 = true ? 1 : (true ? 1 : 2); +// CHECK-MESSAGES: :[[@LINE-1]]:32: warning: conditional operator is used as sub-expression of parent conditional operator, refrain from using nested conditional operators +// CHECK-MESSAGES: :[[@LINE-2]]:20: note: parent conditional operator here + +int NestWithParensis = true ? 1 : ((((true ? 1 : 2)))); +// CHECK-MESSAGES: :[[@LINE-1]]:39: warning: conditional operator is used as sub-expression of parent conditional operator, refrain from using nested conditional operators +// CHECK-MESSAGES: :[[@LINE-2]]:24: note: parent conditional operator here + +#define CONDITIONAL_EXPR (true ? 1 : 2) +// not diag for macro since it will not reduce readability +int NestWithMacro = true ? CONDITIONAL_EXPR : 2; diff --git a/clang-tools-extra/test/clang-tidy/checkers/readability/container-size-empty.cpp b/clang-tools-extra/test/clang-tidy/checkers/readability/container-size-empty.cpp index 3b9e060841830..84bdbd58b85e9 100644 --- a/clang-tools-extra/test/clang-tidy/checkers/readability/container-size-empty.cpp +++ b/clang-tools-extra/test/clang-tidy/checkers/readability/container-size-empty.cpp @@ -33,7 +33,7 @@ class TemplatedContainer { public: bool operator==(const TemplatedContainer& other) const; bool operator!=(const TemplatedContainer& other) const; - int size() const; + unsigned long size() const; bool empty() const; }; @@ -42,7 +42,7 @@ class PrivateEmpty { public: bool operator==(const PrivateEmpty& other) const; bool operator!=(const PrivateEmpty& other) const; - int size() const; + unsigned long size() const; private: bool empty() const; }; @@ -61,7 +61,7 @@ struct EnumSize { class Container { public: bool operator==(const Container& other) const; - int size() const; + unsigned long size() const; bool empty() const; }; @@ -70,13 +70,13 @@ class Derived : public Container { class Container2 { public: - int size() const; + unsigned long size() const; bool empty() const { return size() == 0; } }; class Container3 { public: - int size() const; + unsigned long size() const; bool empty() const; }; @@ -85,7 +85,7 @@ bool Container3::empty() const { return this->size() == 0; } class Container4 { public: bool operator==(const Container4& rhs) const; - int size() const; + unsigned long size() const; bool empty() const { return *this == Container4(); } }; @@ -815,3 +815,49 @@ bool testNotEmptyStringLiterals(const std::string& s) using namespace std::string_literals; return s == "foo"s; } + +namespace PR72619 { + struct SS { + bool empty() const; + int size() const; + }; + + struct SU { + bool empty() const; + unsigned size() const; + }; + + void f(const SU& s) { + if (s.size() < 0) {} + if (0 > s.size()) {} + if (s.size() >= 0) {} + if (0 <= s.size()) {} + if (s.size() < 1) + ; + // CHECK-MESSAGES: :[[@LINE-2]]:9: warning: the 'empty' method should be used to check for emptiness instead of 'size' [readability-container-size-empty] + // CHECK-FIXES: {{^ }}if (s.empty()){{$}} + if (1 > s.size()) + ; + // CHECK-MESSAGES: :[[@LINE-2]]:13: warning: the 'empty' method should be used to check for emptiness instead of 'size' [readability-container-size-empty] + // CHECK-FIXES: {{^ }}if (s.empty()){{$}} + if (s.size() <= 0) + ; + // CHECK-MESSAGES: :[[@LINE-2]]:9: warning: the 'empty' method should be used to check for emptiness instead of 'size' [readability-container-size-empty] + // CHECK-FIXES: {{^ }}if (s.empty()){{$}} + if (0 >= s.size()) + ; + // CHECK-MESSAGES: :[[@LINE-2]]:14: warning: the 'empty' method should be used to check for emptiness instead of 'size' [readability-container-size-empty] + // CHECK-FIXES: {{^ }}if (s.empty()){{$}} + } + + void f(const SS& s) { + if (s.size() < 0) {} + if (0 > s.size()) {} + if (s.size() >= 0) {} + if (0 <= s.size()) {} + if (s.size() < 1) {} + if (1 > s.size()) {} + if (s.size() <= 0) {} + if (0 >= s.size()) {} + } +} diff --git a/clang-tools-extra/test/clang-tidy/checkers/readability/implicit-bool-conversion-allow-in-conditions.cpp b/clang-tools-extra/test/clang-tidy/checkers/readability/implicit-bool-conversion-allow-in-conditions.cpp index e393e297140cb..ef591940917cd 100644 --- a/clang-tools-extra/test/clang-tidy/checkers/readability/implicit-bool-conversion-allow-in-conditions.cpp +++ b/clang-tools-extra/test/clang-tidy/checkers/readability/implicit-bool-conversion-allow-in-conditions.cpp @@ -12,13 +12,14 @@ int* functionReturningPointer(); struct Struct { int member; unsigned bitfield : 1; + bool boolfield : 1; }; void regularImplicitConversionIntegerToBoolIsNotIgnored() { int integer = 0; functionTaking(integer); - // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'int' -> bool [readability-implicit-bool-conversion] + // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'int' -> 'bool' [readability-implicit-bool-conversion] // CHECK-FIXES: functionTaking(integer != 0); } @@ -28,6 +29,8 @@ void implicitConversionIntegerToBoolInConditionalsIsAllowed() { if (!s.member) {} if (s.bitfield) {} if (!s.bitfield) {} + if (s.boolfield == true) {} + if (s.boolfield != true) {} if (functionReturningInt()) {} if (!functionReturningInt()) {} if (functionReturningInt() && functionReturningPointer()) {} @@ -51,12 +54,12 @@ void implicitConversionIntegerToBoolInConditionalsIsAllowed() { void regularImplicitConversionPointerToBoolIsNotIgnored() { int* pointer = nullptr; functionTaking(pointer); - // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'int *' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'int *' -> 'bool' // CHECK-FIXES: functionTaking(pointer != nullptr); int Struct::* memberPointer = &Struct::member; functionTaking(memberPointer); - // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'int Struct::*' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'int Struct::*' -> 'bool' // CHECK-FIXES: functionTaking(memberPointer != nullptr); } diff --git a/clang-tools-extra/test/clang-tidy/checkers/readability/implicit-bool-conversion-cxx98.cpp b/clang-tools-extra/test/clang-tidy/checkers/readability/implicit-bool-conversion-cxx98.cpp index 861677ac6c303..e4c47b06d8f84 100644 --- a/clang-tools-extra/test/clang-tidy/checkers/readability/implicit-bool-conversion-cxx98.cpp +++ b/clang-tools-extra/test/clang-tidy/checkers/readability/implicit-bool-conversion-cxx98.cpp @@ -15,31 +15,31 @@ struct Struct { void useOldNullMacroInReplacements() { int* pointer = NULL; functionTaking(pointer); - // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'int *' -> bool [readability-implicit-bool-conversion] + // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'int *' -> 'bool' [readability-implicit-bool-conversion] // CHECK-FIXES: functionTaking(pointer != 0); int Struct::* memberPointer = NULL; functionTaking(!memberPointer); - // CHECK-MESSAGES: :[[@LINE-1]]:25: warning: implicit conversion 'int Struct::*' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:25: warning: implicit conversion 'int Struct::*' -> 'bool' // CHECK-FIXES: functionTaking(memberPointer == 0); } void fixFalseLiteralConvertingToNullPointer() { functionTaking(false); - // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion bool -> 'int *' + // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'bool' -> 'int *' // CHECK-FIXES: functionTaking(0); int* pointer = NULL; if (pointer == false) {} - // CHECK-MESSAGES: :[[@LINE-1]]:18: warning: implicit conversion bool -> 'int *' + // CHECK-MESSAGES: :[[@LINE-1]]:18: warning: implicit conversion 'bool' -> 'int *' // CHECK-FIXES: if (pointer == 0) {} functionTaking(false); - // CHECK-MESSAGES: :[[@LINE-1]]:33: warning: implicit conversion bool -> 'int Struct::*' + // CHECK-MESSAGES: :[[@LINE-1]]:33: warning: implicit conversion 'bool' -> 'int Struct::*' // CHECK-FIXES: functionTaking(0); int Struct::* memberPointer = NULL; if (memberPointer != false) {} - // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion bool -> 'int Struct::*' + // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'bool' -> 'int Struct::*' // CHECK-FIXES: if (memberPointer != 0) {} } diff --git a/clang-tools-extra/test/clang-tidy/checkers/readability/implicit-bool-conversion.cpp b/clang-tools-extra/test/clang-tidy/checkers/readability/implicit-bool-conversion.cpp index f7f5d506a9ce0..6f1f297297188 100644 --- a/clang-tools-extra/test/clang-tidy/checkers/readability/implicit-bool-conversion.cpp +++ b/clang-tools-extra/test/clang-tidy/checkers/readability/implicit-bool-conversion.cpp @@ -21,30 +21,30 @@ void implicitConversionFromBoolSimpleCases() { functionTaking(boolean); functionTaking(boolean); - // CHECK-MESSAGES: :[[@LINE-1]]:23: warning: implicit conversion bool -> 'int' [readability-implicit-bool-conversion] + // CHECK-MESSAGES: :[[@LINE-1]]:23: warning: implicit conversion 'bool' -> 'int' [readability-implicit-bool-conversion] // CHECK-FIXES: functionTaking(static_cast(boolean)); functionTaking(boolean); - // CHECK-MESSAGES: :[[@LINE-1]]:33: warning: implicit conversion bool -> 'unsigned long' + // CHECK-MESSAGES: :[[@LINE-1]]:33: warning: implicit conversion 'bool' -> 'unsigned long' // CHECK-FIXES: functionTaking(static_cast(boolean)); functionTaking(boolean); - // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion bool -> 'char' + // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'bool' -> 'char' // CHECK-FIXES: functionTaking(static_cast(boolean)); functionTaking(boolean); - // CHECK-MESSAGES: :[[@LINE-1]]:25: warning: implicit conversion bool -> 'float' + // CHECK-MESSAGES: :[[@LINE-1]]:25: warning: implicit conversion 'bool' -> 'float' // CHECK-FIXES: functionTaking(static_cast(boolean)); functionTaking(boolean); - // CHECK-MESSAGES: :[[@LINE-1]]:26: warning: implicit conversion bool -> 'double' + // CHECK-MESSAGES: :[[@LINE-1]]:26: warning: implicit conversion 'bool' -> 'double' // CHECK-FIXES: functionTaking(static_cast(boolean)); } float implicitConversionFromBoolInReturnValue() { bool boolean = false; return boolean; - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: implicit conversion bool -> 'float' + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: implicit conversion 'bool' -> 'float' // CHECK-FIXES: return static_cast(boolean); } @@ -58,15 +58,15 @@ void implicitConversionFromBoolInSingleBoolExpressions(bool b1, bool b2) { boolean = b2 != false; int integer = boolean - 3; - // CHECK-MESSAGES: :[[@LINE-1]]:17: warning: implicit conversion bool -> 'int' + // CHECK-MESSAGES: :[[@LINE-1]]:17: warning: implicit conversion 'bool' -> 'int' // CHECK-FIXES: int integer = static_cast(boolean) - 3; float floating = boolean / 0.3f; - // CHECK-MESSAGES: :[[@LINE-1]]:20: warning: implicit conversion bool -> 'float' + // CHECK-MESSAGES: :[[@LINE-1]]:20: warning: implicit conversion 'bool' -> 'float' // CHECK-FIXES: float floating = static_cast(boolean) / 0.3f; char character = boolean; - // CHECK-MESSAGES: :[[@LINE-1]]:20: warning: implicit conversion bool -> 'char' + // CHECK-MESSAGES: :[[@LINE-1]]:20: warning: implicit conversion 'bool' -> 'char' // CHECK-FIXES: char character = static_cast(boolean); } @@ -75,41 +75,41 @@ void implicitConversionFromBoollInComplexBoolExpressions() { bool anotherBoolean = false; int integer = boolean && anotherBoolean; - // CHECK-MESSAGES: :[[@LINE-1]]:17: warning: implicit conversion bool -> 'int' + // CHECK-MESSAGES: :[[@LINE-1]]:17: warning: implicit conversion 'bool' -> 'int' // CHECK-FIXES: int integer = static_cast(boolean && anotherBoolean); unsigned long unsignedLong = (! boolean) + 4ul; - // CHECK-MESSAGES: :[[@LINE-1]]:32: warning: implicit conversion bool -> 'unsigned long' + // CHECK-MESSAGES: :[[@LINE-1]]:32: warning: implicit conversion 'bool' -> 'unsigned long' // CHECK-FIXES: unsigned long unsignedLong = static_cast(! boolean) + 4ul; float floating = (boolean || anotherBoolean) * 0.3f; - // CHECK-MESSAGES: :[[@LINE-1]]:20: warning: implicit conversion bool -> 'float' + // CHECK-MESSAGES: :[[@LINE-1]]:20: warning: implicit conversion 'bool' -> 'float' // CHECK-FIXES: float floating = static_cast(boolean || anotherBoolean) * 0.3f; double doubleFloating = (boolean && (anotherBoolean || boolean)) * 0.3; - // CHECK-MESSAGES: :[[@LINE-1]]:27: warning: implicit conversion bool -> 'double' + // CHECK-MESSAGES: :[[@LINE-1]]:27: warning: implicit conversion 'bool' -> 'double' // CHECK-FIXES: double doubleFloating = static_cast(boolean && (anotherBoolean || boolean)) * 0.3; } void implicitConversionFromBoolLiterals() { functionTaking(true); - // CHECK-MESSAGES: :[[@LINE-1]]:23: warning: implicit conversion bool -> 'int' + // CHECK-MESSAGES: :[[@LINE-1]]:23: warning: implicit conversion 'bool' -> 'int' // CHECK-FIXES: functionTaking(1); functionTaking(false); - // CHECK-MESSAGES: :[[@LINE-1]]:33: warning: implicit conversion bool -> 'unsigned long' + // CHECK-MESSAGES: :[[@LINE-1]]:33: warning: implicit conversion 'bool' -> 'unsigned long' // CHECK-FIXES: functionTaking(0u); functionTaking(true); - // CHECK-MESSAGES: :[[@LINE-1]]:31: warning: implicit conversion bool -> 'signed char' + // CHECK-MESSAGES: :[[@LINE-1]]:31: warning: implicit conversion 'bool' -> 'signed char' // CHECK-FIXES: functionTaking(1); functionTaking(false); - // CHECK-MESSAGES: :[[@LINE-1]]:25: warning: implicit conversion bool -> 'float' + // CHECK-MESSAGES: :[[@LINE-1]]:25: warning: implicit conversion 'bool' -> 'float' // CHECK-FIXES: functionTaking(0.0f); functionTaking(true); - // CHECK-MESSAGES: :[[@LINE-1]]:26: warning: implicit conversion bool -> 'double' + // CHECK-MESSAGES: :[[@LINE-1]]:26: warning: implicit conversion 'bool' -> 'double' // CHECK-FIXES: functionTaking(1.0); } @@ -118,11 +118,11 @@ void implicitConversionFromBoolInComparisons() { int integer = 0; functionTaking(boolean == integer); - // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion bool -> 'int' + // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'bool' -> 'int' // CHECK-FIXES: functionTaking(static_cast(boolean) == integer); functionTaking(integer != boolean); - // CHECK-MESSAGES: :[[@LINE-1]]:35: warning: implicit conversion bool -> 'int' + // CHECK-MESSAGES: :[[@LINE-1]]:35: warning: implicit conversion 'bool' -> 'int' // CHECK-FIXES: functionTaking(integer != static_cast(boolean)); } @@ -171,59 +171,59 @@ void useOfTemplateFunction() { void implicitConversionToBoolSimpleCases() { int integer = 10; functionTaking(integer); - // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'int' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'int' -> 'bool' // CHECK-FIXES: functionTaking(integer != 0); unsigned long unsignedLong = 10; functionTaking(unsignedLong); - // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'unsigned long' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'unsigned long' -> 'bool' // CHECK-FIXES: functionTaking(unsignedLong != 0u); float floating = 0.0f; functionTaking(floating); - // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'float' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'float' -> 'bool' // CHECK-FIXES: functionTaking(floating != 0.0f); double doubleFloating = 1.0f; functionTaking(doubleFloating); - // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'double' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'double' -> 'bool' // CHECK-FIXES: functionTaking(doubleFloating != 0.0); signed char character = 'a'; functionTaking(character); - // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'signed char' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'signed char' -> 'bool' // CHECK-FIXES: functionTaking(character != 0); int* pointer = nullptr; functionTaking(pointer); - // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'int *' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'int *' -> 'bool' // CHECK-FIXES: functionTaking(pointer != nullptr); auto pointerToMember = &Struct::member; functionTaking(pointerToMember); - // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'int Struct::*' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'int Struct::*' -> 'bool' // CHECK-FIXES: functionTaking(pointerToMember != nullptr); } void implicitConversionToBoolInSingleExpressions() { int integer = 10; bool boolComingFromInt = integer; - // CHECK-MESSAGES: :[[@LINE-1]]:28: warning: implicit conversion 'int' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:28: warning: implicit conversion 'int' -> 'bool' // CHECK-FIXES: bool boolComingFromInt = integer != 0; float floating = 10.0f; bool boolComingFromFloat = floating; - // CHECK-MESSAGES: :[[@LINE-1]]:30: warning: implicit conversion 'float' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:30: warning: implicit conversion 'float' -> 'bool' // CHECK-FIXES: bool boolComingFromFloat = floating != 0.0f; signed char character = 'a'; bool boolComingFromChar = character; - // CHECK-MESSAGES: :[[@LINE-1]]:29: warning: implicit conversion 'signed char' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:29: warning: implicit conversion 'signed char' -> 'bool' // CHECK-FIXES: bool boolComingFromChar = character != 0; int* pointer = nullptr; bool boolComingFromPointer = pointer; - // CHECK-MESSAGES: :[[@LINE-1]]:32: warning: implicit conversion 'int *' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:32: warning: implicit conversion 'int *' -> 'bool' // CHECK-FIXES: bool boolComingFromPointer = pointer != nullptr; } @@ -233,131 +233,131 @@ void implicitConversionToBoolInComplexExpressions() { int integer = 10; int anotherInteger = 20; bool boolComingFromInteger = integer + anotherInteger; - // CHECK-MESSAGES: :[[@LINE-1]]:32: warning: implicit conversion 'int' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:32: warning: implicit conversion 'int' -> 'bool' // CHECK-FIXES: bool boolComingFromInteger = (integer + anotherInteger) != 0; float floating = 0.2f; bool boolComingFromFloating = floating - 0.3f || boolean; - // CHECK-MESSAGES: :[[@LINE-1]]:33: warning: implicit conversion 'float' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:33: warning: implicit conversion 'float' -> 'bool' // CHECK-FIXES: bool boolComingFromFloating = ((floating - 0.3f) != 0.0f) || boolean; double doubleFloating = 0.3; bool boolComingFromDoubleFloating = (doubleFloating - 0.4) && boolean; - // CHECK-MESSAGES: :[[@LINE-1]]:39: warning: implicit conversion 'double' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:39: warning: implicit conversion 'double' -> 'bool' // CHECK-FIXES: bool boolComingFromDoubleFloating = ((doubleFloating - 0.4) != 0.0) && boolean; } void implicitConversionInNegationExpressions() { int integer = 10; bool boolComingFromNegatedInt = !integer; - // CHECK-MESSAGES: :[[@LINE-1]]:36: warning: implicit conversion 'int' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:36: warning: implicit conversion 'int' -> 'bool' // CHECK-FIXES: bool boolComingFromNegatedInt = integer == 0; float floating = 10.0f; bool boolComingFromNegatedFloat = ! floating; - // CHECK-MESSAGES: :[[@LINE-1]]:39: warning: implicit conversion 'float' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:39: warning: implicit conversion 'float' -> 'bool' // CHECK-FIXES: bool boolComingFromNegatedFloat = floating == 0.0f; signed char character = 'a'; bool boolComingFromNegatedChar = (! character); - // CHECK-MESSAGES: :[[@LINE-1]]:39: warning: implicit conversion 'signed char' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:39: warning: implicit conversion 'signed char' -> 'bool' // CHECK-FIXES: bool boolComingFromNegatedChar = (character == 0); int* pointer = nullptr; bool boolComingFromNegatedPointer = not pointer; - // CHECK-MESSAGES: :[[@LINE-1]]:43: warning: implicit conversion 'int *' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:43: warning: implicit conversion 'int *' -> 'bool' // CHECK-FIXES: bool boolComingFromNegatedPointer = pointer == nullptr; } void implicitConversionToBoolInControlStatements() { int integer = 10; if (integer) {} - // CHECK-MESSAGES: :[[@LINE-1]]:7: warning: implicit conversion 'int' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:7: warning: implicit conversion 'int' -> 'bool' // CHECK-FIXES: if (integer != 0) {} long int longInteger = 0.2f; for (;longInteger;) {} - // CHECK-MESSAGES: :[[@LINE-1]]:9: warning: implicit conversion 'long' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:9: warning: implicit conversion 'long' -> 'bool' // CHECK-FIXES: for (;longInteger != 0;) {} float floating = 0.3f; while (floating) {} - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: implicit conversion 'float' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: implicit conversion 'float' -> 'bool' // CHECK-FIXES: while (floating != 0.0f) {} double doubleFloating = 0.4; do {} while (doubleFloating); - // CHECK-MESSAGES: :[[@LINE-1]]:16: warning: implicit conversion 'double' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:16: warning: implicit conversion 'double' -> 'bool' // CHECK-FIXES: do {} while (doubleFloating != 0.0); } bool implicitConversionToBoolInReturnValue() { float floating = 1.0f; return floating; - // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: implicit conversion 'float' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:10: warning: implicit conversion 'float' -> 'bool' // CHECK-FIXES: return floating != 0.0f; } void implicitConversionToBoolFromLiterals() { functionTaking(0); - // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'int' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'int' -> 'bool' // CHECK-FIXES: functionTaking(false); functionTaking(1); - // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'int' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'int' -> 'bool' // CHECK-FIXES: functionTaking(true); functionTaking(2ul); - // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'unsigned long' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'unsigned long' -> 'bool' // CHECK-FIXES: functionTaking(true); functionTaking(0.0f); - // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'float' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'float' -> 'bool' // CHECK-FIXES: functionTaking(false); functionTaking(1.0f); - // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'float' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'float' -> 'bool' // CHECK-FIXES: functionTaking(true); functionTaking(2.0); - // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'double' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'double' -> 'bool' // CHECK-FIXES: functionTaking(true); functionTaking('\0'); - // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'char' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'char' -> 'bool' // CHECK-FIXES: functionTaking(false); functionTaking('a'); - // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'char' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'char' -> 'bool' // CHECK-FIXES: functionTaking(true); functionTaking(""); - // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'const char *' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'const char *' -> 'bool' // CHECK-FIXES: functionTaking(true); functionTaking("abc"); - // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'const char *' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'const char *' -> 'bool' // CHECK-FIXES: functionTaking(true); functionTaking(NULL); - // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'long' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'long' -> 'bool' // CHECK-FIXES: functionTaking(false); } void implicitConversionToBoolFromUnaryMinusAndZeroLiterals() { functionTaking(-0); - // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'int' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'int' -> 'bool' // CHECK-FIXES: functionTaking((-0) != 0); functionTaking(-0.0f); - // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'float' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'float' -> 'bool' // CHECK-FIXES: functionTaking((-0.0f) != 0.0f); functionTaking(-0.0); - // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'double' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'double' -> 'bool' // CHECK-FIXES: functionTaking((-0.0) != 0.0); } @@ -370,11 +370,11 @@ void implicitConversionToBoolInWithOverloadedOperators() { UserStruct s; functionTaking(s(0)); - // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'int' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'int' -> 'bool' // CHECK-FIXES: functionTaking(s(0) != 0); functionTaking(s + 2); - // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'int' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'int' -> 'bool' // CHECK-FIXES: functionTaking((s + 2) != 0); } @@ -446,27 +446,27 @@ struct S { int c : 2; S(bool a, bool b, bool c) : a(a), b(b), c(c) {} - // CHECK-MESSAGES: :[[@LINE-1]]:33: warning: implicit conversion bool -> 'int' - // CHECK-MESSAGES: :[[@LINE-2]]:45: warning: implicit conversion bool -> 'int' + // CHECK-MESSAGES: :[[@LINE-1]]:33: warning: implicit conversion 'bool' -> 'int' + // CHECK-MESSAGES: :[[@LINE-2]]:45: warning: implicit conversion 'bool' -> 'int' // CHECK-FIXES: S(bool a, bool b, bool c) : a(static_cast(a)), b(b), c(static_cast(c)) {} }; bool f(S& s) { functionTaking(s.a); - // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'int' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'int' -> 'bool' // CHECK-FIXES: functionTaking(s.a != 0); functionTaking(s.b); // CHECK-FIXES: functionTaking(s.b); s.a = true; - // CHECK-MESSAGES: :[[@LINE-1]]:9: warning: implicit conversion bool -> 'int' + // CHECK-MESSAGES: :[[@LINE-1]]:9: warning: implicit conversion 'bool' -> 'int' // CHECK-FIXES: s.a = 1; s.b = true; // CHECK-FIXES: s.b = true; s.c = true; - // CHECK-MESSAGES: :[[@LINE-1]]:9: warning: implicit conversion bool -> 'int' + // CHECK-MESSAGES: :[[@LINE-1]]:9: warning: implicit conversion 'bool' -> 'int' // CHECK-FIXES: s.c = 1; functionTaking(s.c); - // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'int' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:24: warning: implicit conversion 'int' -> 'bool' // CHECK-FIXES: functionTaking(s.c != 0); } @@ -475,14 +475,14 @@ bool f(S& s) { int implicitConversionReturnInt() { return true; - // CHECK-MESSAGES: :[[@LINE-1]]:12: warning: implicit conversion bool -> 'int' + // CHECK-MESSAGES: :[[@LINE-1]]:12: warning: implicit conversion 'bool' -> 'int' // CHECK-FIXES: return 1 } int implicitConversionReturnIntWithParens() { return (true); - // CHECK-MESSAGES: :[[@LINE-1]]:12: warning: implicit conversion bool -> 'int' + // CHECK-MESSAGES: :[[@LINE-1]]:12: warning: implicit conversion 'bool' -> 'int' // CHECK-FIXES: return 1 } @@ -490,14 +490,14 @@ int implicitConversionReturnIntWithParens() bool implicitConversionReturnBool() { return 1; - // CHECK-MESSAGES: :[[@LINE-1]]:12: warning: implicit conversion 'int' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:12: warning: implicit conversion 'int' -> 'bool' // CHECK-FIXES: return true } bool implicitConversionReturnBoolWithParens() { return (1); - // CHECK-MESSAGES: :[[@LINE-1]]:12: warning: implicit conversion 'int' -> bool + // CHECK-MESSAGES: :[[@LINE-1]]:12: warning: implicit conversion 'int' -> 'bool' // CHECK-FIXES: return true } @@ -508,3 +508,19 @@ namespace PR47000 { using IntType = int; int to_int2(bool x) { return IntType{x}; } } + +namespace PR71867 { + bool foo(bool x) { + return x ? 1 : false; + // CHECK-MESSAGES: :[[@LINE-1]]:12: warning: implicit conversion 'int' -> 'bool' + // CHECK-MESSAGES: :[[@LINE-2]]:20: warning: implicit conversion 'bool' -> 'int' + // CHECK-FIXES: return (x ? 1 : 0) != 0; + } + + bool boo(bool x) { + return x ? true : 0; + // CHECK-MESSAGES: :[[@LINE-1]]:12: warning: implicit conversion 'int' -> 'bool' + // CHECK-MESSAGES: :[[@LINE-2]]:16: warning: implicit conversion 'bool' -> 'int' + // CHECK-FIXES: return (x ? 1 : 0) != 0; + } +} diff --git a/clang-tools-extra/test/clang-tidy/checkers/readability/redundant-member-init.cpp b/clang-tools-extra/test/clang-tidy/checkers/readability/redundant-member-init.cpp index d8ca406581a38..17b2714abca07 100644 --- a/clang-tools-extra/test/clang-tidy/checkers/readability/redundant-member-init.cpp +++ b/clang-tools-extra/test/clang-tidy/checkers/readability/redundant-member-init.cpp @@ -250,3 +250,55 @@ struct NF15 { S s2; }; }; + +// Direct in-class initialization with default constructor +struct D1 { + S f1 {}; + // CHECK-MESSAGES: :[[@LINE-1]]:8: warning: initializer for member 'f1' is redundant + // CHECK-FIXES: S f1; +}; + +// Direct in-class initialization with constructor with default argument +struct D2 { + T f2 {}; + // CHECK-MESSAGES: :[[@LINE-1]]:9: warning: initializer for member 'f2' is redundant + // CHECK-FIXES: T f2; +}; + +// Direct in-class initialization with default constructor (assign) +struct D3 { + S f3 = {}; + // CHECK-MESSAGES: :[[@LINE-1]]:8: warning: initializer for member 'f3' is redundant + // CHECK-FIXES: S f3; +}; + +// Direct in-class initialization with constructor with default argument (assign) +struct D4 { + T f4 = {}; + // CHECK-MESSAGES: :[[@LINE-1]]:8: warning: initializer for member 'f4' is redundant + // CHECK-FIXES: T f4; +}; + +// Templated class independent type +template +struct D5 { + S f5 /*comment*/ = S(); + // CHECK-MESSAGES: :[[@LINE-1]]:22: warning: initializer for member 'f5' is redundant + // CHECK-FIXES: S f5 /*comment*/; +}; +D5 d5i; +D5 d5s; + +struct D6 { + UsesCleanup uc2{}; + // CHECK-MESSAGES: :[[@LINE-1]]:18: warning: initializer for member 'uc2' is redundant + // CHECK-FIXES: UsesCleanup uc2; +}; + +template +struct D7 { + V f7; +}; + +D7 d7i; +D7 d7s; diff --git a/clang-tools-extra/test/clang-tidy/checkers/readability/simplify-boolean-expr-macros.cpp b/clang-tools-extra/test/clang-tidy/checkers/readability/simplify-boolean-expr-macros.cpp new file mode 100644 index 0000000000000..7d0cfe7e27dc2 --- /dev/null +++ b/clang-tools-extra/test/clang-tidy/checkers/readability/simplify-boolean-expr-macros.cpp @@ -0,0 +1,20 @@ +// RUN: %check_clang_tidy -check-suffixes=,MACROS %s readability-simplify-boolean-expr %t + +// Ignore expressions in macros. +// RUN: %check_clang_tidy %s readability-simplify-boolean-expr %t \ +// RUN: -- -config="{CheckOptions: {readability-simplify-boolean-expr.IgnoreMacros: true}}" \ +// RUN: -- + +#define NEGATE(expr) !(expr) + +bool without_macro(bool a, bool b) { + return !(!a && b); + // CHECK-MESSAGES: :[[@LINE-1]]:12: warning: boolean expression can be simplified by DeMorgan's theorem + // CHECK-FIXES: return a || !b; +} + +bool macro(bool a, bool b) { + return NEGATE(!a && b); + // CHECK-MESSAGES-MACROS: :[[@LINE-1]]:12: warning: boolean expression can be simplified by DeMorgan's theorem + // CHECK-FIXES: return NEGATE(!a && b); +} diff --git a/clang/bindings/python/README.txt b/clang/bindings/python/README.txt index 44c715e5de56f..3e509662144fa 100644 --- a/clang/bindings/python/README.txt +++ b/clang/bindings/python/README.txt @@ -10,7 +10,7 @@ runner. For example: -- $ env PYTHONPATH=$(echo ~/llvm/clang/bindings/python/) \ CLANG_LIBRARY_PATH=$(llvm-config --libdir) \ - python -m unittest discover -v + python3 -m unittest discover -v tests.cindex.test_index.test_create ... ok ... diff --git a/clang/bindings/python/clang/cindex.py b/clang/bindings/python/clang/cindex.py index d780ee353a133..754f03d718e88 100644 --- a/clang/bindings/python/clang/cindex.py +++ b/clang/bindings/python/clang/cindex.py @@ -66,81 +66,50 @@ import clang.enumerations +import collections.abc import os -import sys - -if sys.version_info[0] == 3: - # Python 3 strings are unicode, translate them to/from utf8 for C-interop. - class c_interop_string(c_char_p): - def __init__(self, p=None): - if p is None: - p = "" - if isinstance(p, str): - p = p.encode("utf8") - super(c_char_p, self).__init__(p) - def __str__(self): - return self.value - - @property - def value(self): - if super(c_char_p, self).value is None: - return None - return super(c_char_p, self).value.decode("utf8") - - @classmethod - def from_param(cls, param): - if isinstance(param, str): - return cls(param) - if isinstance(param, bytes): - return cls(param) - if param is None: - # Support passing null to C functions expecting char arrays - return None - raise TypeError( - "Cannot convert '{}' to '{}'".format(type(param).__name__, cls.__name__) - ) - - @staticmethod - def to_python_string(x, *args): - return x.value - def b(x): - if isinstance(x, bytes): - return x - return x.encode("utf8") +# Python 3 strings are unicode, translate them to/from utf8 for C-interop. +class c_interop_string(c_char_p): + def __init__(self, p=None): + if p is None: + p = "" + if isinstance(p, str): + p = p.encode("utf8") + super(c_char_p, self).__init__(p) -elif sys.version_info[0] == 2: - # Python 2 strings are utf8 byte strings, no translation is needed for - # C-interop. - c_interop_string = c_char_p - - def _to_python_string(x, *args): - return x - - c_interop_string.to_python_string = staticmethod(_to_python_string) + def __str__(self): + return self.value - def b(x): - return x + @property + def value(self): + if super(c_char_p, self).value is None: + return None + return super(c_char_p, self).value.decode("utf8") + @classmethod + def from_param(cls, param): + if isinstance(param, str): + return cls(param) + if isinstance(param, bytes): + return cls(param) + if param is None: + # Support passing null to C functions expecting char arrays + return None + raise TypeError( + "Cannot convert '{}' to '{}'".format(type(param).__name__, cls.__name__) + ) -# Importing ABC-s directly from collections is deprecated since Python 3.7, -# will stop working in Python 3.8. -# See: https://docs.python.org/dev/whatsnew/3.7.html#id3 -if sys.version_info[:2] >= (3, 7): - from collections import abc as collections_abc -else: - import collections as collections_abc + @staticmethod + def to_python_string(x, *args): + return x.value -# We only support PathLike objects on Python version with os.fspath present -# to be consistent with the Python standard library. On older Python versions -# we only support strings and we have dummy fspath to just pass them through. -try: - fspath = os.fspath -except AttributeError: - def fspath(x): +def b(x): + if isinstance(x, bytes): return x + return x.encode("utf8") # ctypes doesn't implicitly convert c_void_p to the appropriate wrapper @@ -202,7 +171,7 @@ def __init__(self, enumeration, message): ### Structures and Utility Classes ### -class CachedProperty(object): +class CachedProperty: """Decorator that lazy-loads the value of a property. The first time the property is accessed, the original property function is @@ -392,7 +361,7 @@ def __repr__(self): return "" % (self.start, self.end) -class Diagnostic(object): +class Diagnostic: """ A Diagnostic is a single instance of a Clang diagnostic. It includes the diagnostic severity, the message, the location the diagnostic occurred, as @@ -433,7 +402,7 @@ def spelling(self): @property def ranges(self): - class RangeIterator(object): + class RangeIterator: def __init__(self, diag): self.diag = diag @@ -449,7 +418,7 @@ def __getitem__(self, key): @property def fixits(self): - class FixItIterator(object): + class FixItIterator: def __init__(self, diag): self.diag = diag @@ -468,7 +437,7 @@ def __getitem__(self, key): @property def children(self): - class ChildDiagnosticsIterator(object): + class ChildDiagnosticsIterator: def __init__(self, diag): self.diag_set = conf.lib.clang_getChildDiagnostics(diag) @@ -532,7 +501,7 @@ def from_param(self): return self.ptr -class FixIt(object): +class FixIt: """ A FixIt represents a transformation to be applied to the source to "fix-it". The fix-it should be applied by replacing the given source range @@ -547,7 +516,7 @@ def __repr__(self): return "" % (self.range, self.value) -class TokenGroup(object): +class TokenGroup: """Helper class to facilitate token management. Tokens are allocated from libclang in chunks. They must be disposed of as a @@ -603,7 +572,7 @@ def get_tokens(tu, extent): yield token -class TokenKind(object): +class TokenKind: """Describes a specific type of a Token.""" _value_map = {} # int -> TokenKind @@ -642,7 +611,7 @@ def register(value, name): ### Cursor Kinds ### -class BaseEnumeration(object): +class BaseEnumeration: """ Common base class for named enumerations held in sync with Index.h values. @@ -2059,7 +2028,7 @@ def from_cursor_result(res, fn, args): return res -class StorageClass(object): +class StorageClass: """ Describes the storage class of a declaration """ @@ -2353,7 +2322,7 @@ def argument_types(self): container is a Type instance. """ - class ArgumentsIterator(collections_abc.Sequence): + class ArgumentsIterator(collections.abc.Sequence): def __init__(self, parent): self.parent = parent self.length = None @@ -2608,7 +2577,7 @@ def __ne__(self, other): # a void*. -class ClangObject(object): +class ClangObject: """ A helper for Clang objects. This class helps act as an intermediary for the ctypes library and the Clang CIndex library. @@ -2656,8 +2625,8 @@ class _CXUnsavedFile(Structure): } -class CompletionChunk(object): - class Kind(object): +class CompletionChunk: + class Kind: def __init__(self, name): self.name = name @@ -2747,7 +2716,7 @@ def isKindResultType(self): class CompletionString(ClangObject): - class Availability(object): + class Availability: def __init__(self, name): self.name = name @@ -2849,7 +2818,7 @@ def results(self): @property def diagnostics(self): - class DiagnosticsItr(object): + class DiagnosticsItr: def __init__(self, ccr): self.ccr = ccr @@ -3003,13 +2972,13 @@ def from_source( if hasattr(contents, "read"): contents = contents.read() contents = b(contents) - unsaved_array[i].name = b(fspath(name)) + unsaved_array[i].name = b(os.fspath(name)) unsaved_array[i].contents = contents unsaved_array[i].length = len(contents) ptr = conf.lib.clang_parseTranslationUnit( index, - fspath(filename) if filename is not None else None, + os.fspath(filename) if filename is not None else None, args_array, len(args), unsaved_array, @@ -3040,7 +3009,7 @@ def from_ast_file(cls, filename, index=None): if index is None: index = Index.create() - ptr = conf.lib.clang_createTranslationUnit(index, fspath(filename)) + ptr = conf.lib.clang_createTranslationUnit(index, os.fspath(filename)) if not ptr: raise TranslationUnitLoadError(filename) @@ -3159,7 +3128,7 @@ def diagnostics(self): Return an iterable (and indexable) object containing the diagnostics. """ - class DiagIterator(object): + class DiagIterator: def __init__(self, tu): self.tu = tu @@ -3193,7 +3162,7 @@ def reparse(self, unsaved_files=None, options=0): if hasattr(contents, "read"): contents = contents.read() contents = b(contents) - unsaved_files_array[i].name = b(fspath(name)) + unsaved_files_array[i].name = b(os.fspath(name)) unsaved_files_array[i].contents = contents unsaved_files_array[i].length = len(contents) ptr = conf.lib.clang_reparseTranslationUnit( @@ -3217,7 +3186,11 @@ def save(self, filename): """ options = conf.lib.clang_defaultSaveOptions(self) result = int( - conf.lib.clang_saveTranslationUnit(self, fspath(filename), options) + conf.lib.clang_saveTranslationUnit( + self, + os.fspath(filename), + options, + ) ) if result != 0: raise TranslationUnitSaveError(result, "Error saving TranslationUnit.") @@ -3261,12 +3234,12 @@ def codeComplete( if hasattr(contents, "read"): contents = contents.read() contents = b(contents) - unsaved_files_array[i].name = b(fspath(name)) + unsaved_files_array[i].name = b(os.fspath(name)) unsaved_files_array[i].contents = contents unsaved_files_array[i].length = len(contents) ptr = conf.lib.clang_codeCompleteAt( self, - fspath(path), + os.fspath(path), line, column, unsaved_files_array, @@ -3300,7 +3273,9 @@ class File(ClangObject): @staticmethod def from_name(translation_unit, file_name): """Retrieve a file handle within the given translation unit.""" - return File(conf.lib.clang_getFile(translation_unit, fspath(file_name))) + return File( + conf.lib.clang_getFile(translation_unit, os.fspath(file_name)), + ) @property def name(self): @@ -3328,7 +3303,7 @@ def from_result(res, fn, args): return res -class FileInclusion(object): +class FileInclusion: """ The FileInclusion class represents the inclusion of one source file by another via a '#include' directive or as the input file for the translation @@ -3377,7 +3352,7 @@ def __init__(self, enumeration, message): Exception.__init__(self, "Error %d: %s" % (enumeration, message)) -class CompileCommand(object): +class CompileCommand: """Represents the compile command used to build a file""" def __init__(self, cmd, ccmds): @@ -3409,7 +3384,7 @@ def arguments(self): yield conf.lib.clang_CompileCommand_getArg(self.cmd, i) -class CompileCommands(object): +class CompileCommands: """ CompileCommands is an iterable object containing all CompileCommand that can be used for building a specific file. @@ -3460,7 +3435,7 @@ def fromDirectory(buildDir): errorCode = c_uint() try: cdb = conf.lib.clang_CompilationDatabase_fromDirectory( - fspath(buildDir), byref(errorCode) + os.fspath(buildDir), byref(errorCode) ) except CompilationDatabaseError as e: raise CompilationDatabaseError( @@ -3474,7 +3449,7 @@ def getCompileCommands(self, filename): build filename. Returns None if filename is not found in the database. """ return conf.lib.clang_CompilationDatabase_getCompileCommands( - self, fspath(filename) + self, os.fspath(filename) ) def getAllCompileCommands(self): @@ -3865,7 +3840,7 @@ def register(item): register(f) -class Config(object): +class Config: library_path = None library_file = None compatibility_check = True @@ -3880,7 +3855,7 @@ def set_library_path(path): "any other functionalities in libclang." ) - Config.library_path = fspath(path) + Config.library_path = os.fspath(path) @staticmethod def set_library_file(filename): @@ -3891,7 +3866,7 @@ def set_library_file(filename): "any other functionalities in libclang." ) - Config.library_file = fspath(filename) + Config.library_file = os.fspath(filename) @staticmethod def set_compatibility_check(check_status): diff --git a/clang/cmake/caches/Fuchsia-stage2.cmake b/clang/cmake/caches/Fuchsia-stage2.cmake index c4673c8a54c5e..eee37c5e7901f 100644 --- a/clang/cmake/caches/Fuchsia-stage2.cmake +++ b/clang/cmake/caches/Fuchsia-stage2.cmake @@ -6,7 +6,7 @@ set(LLVM_TARGETS_TO_BUILD X86;ARM;AArch64;RISCV CACHE STRING "") set(PACKAGE_VENDOR Fuchsia CACHE STRING "") -set(_FUCHSIA_ENABLE_PROJECTS "bolt;clang;clang-tools-extra;lld;llvm;polly") +set(_FUCHSIA_ENABLE_PROJECTS "bolt;clang;clang-tools-extra;libc;lld;llvm;polly") set(LLVM_ENABLE_RUNTIMES "compiler-rt;libcxx;libcxxabi;libunwind" CACHE STRING "") set(LLVM_ENABLE_BACKTRACES OFF CACHE BOOL "") @@ -22,8 +22,11 @@ set(LLVM_ENABLE_TERMINFO OFF CACHE BOOL "") set(LLVM_ENABLE_UNWIND_TABLES OFF CACHE BOOL "") set(LLVM_ENABLE_Z3_SOLVER OFF CACHE BOOL "") set(LLVM_ENABLE_ZLIB ON CACHE BOOL "") +set(LLVM_FORCE_BUILD_RUNTIME ON CACHE BOOL "") set(LLVM_INCLUDE_DOCS OFF CACHE BOOL "") set(LLVM_INCLUDE_EXAMPLES OFF CACHE BOOL "") +set(LLVM_LIBC_FULL_BUILD ON CACHE BOOL "") +set(LIBC_HDRGEN_ONLY ON CACHE BOOL "") set(LLVM_STATIC_LINK_CXX_STDLIB ON CACHE BOOL "") set(LLVM_USE_RELATIVE_PATHS_IN_FILES ON CACHE BOOL "") set(LLDB_ENABLE_CURSES OFF CACHE BOOL "") @@ -297,6 +300,39 @@ if(FUCHSIA_SDK) set(LLVM_RUNTIME_MULTILIB_hwasan+noexcept_TARGETS "aarch64-unknown-fuchsia;riscv64-unknown-fuchsia" CACHE STRING "") endif() +foreach(target riscv32-unknown-elf) + list(APPEND BUILTIN_TARGETS "${target}") + set(BUILTINS_${target}_CMAKE_SYSTEM_NAME Generic CACHE STRING "") + set(BUILTINS_${target}_CMAKE_SYSTEM_PROCESSOR RISCV CACHE STRING "") + set(BUILTINS_${target}_CMAKE_SYSROOT "" CACHE STRING "") + set(BUILTINS_${target}_CMAKE_BUILD_TYPE RelWithDebInfo CACHE STRING "") + foreach(lang C;CXX;ASM) + set(BUILTINS_${target}_CMAKE_${lang}_FLAGS "--target=${target} -march=rv32imafc -mabi=ilp32f" CACHE STRING "") + endforeach() + foreach(type SHARED;MODULE;EXE) + set(BUILTINS_${target}_CMAKE_${type}_LINKER_FLAGS "-fuse-ld=lld" CACHE STRING "") + endforeach() + set(BUILTINS_${target}_COMPILER_RT_BAREMETAL_BUILD ON CACHE BOOL "") + + list(APPEND RUNTIME_TARGETS "${target}") + set(RUNTIMES_${target}_CMAKE_SYSTEM_NAME Generic CACHE STRING "") + set(RUNTIMES_${target}_CMAKE_SYSTEM_PROCESSOR RISCV CACHE STRING "") + set(RUNTIMES_${target}_CMAKE_SYSROOT "" CACHE STRING "") + set(RUNTIMES_${target}_CMAKE_BUILD_TYPE RelWithDebInfo CACHE STRING "") + set(RUNTIMES_${target}_CMAKE_TRY_COMPILE_TARGET_TYPE STATIC_LIBRARY CACHE STRING "") + foreach(lang C;CXX;ASM) + set(RUNTIMES_${target}_CMAKE_${lang}_FLAGS "--target=${target} -march=rv32imafc -mabi=ilp32f" CACHE STRING "") + endforeach() + foreach(type SHARED;MODULE;EXE) + set(RUNTIMES_${target}_CMAKE_${type}_LINKER_FLAGS "-fuse-ld=lld" CACHE STRING "") + endforeach() + set(RUNTIMES_${target}_LLVM_LIBC_FULL_BUILD ON CACHE BOOL "") + set(RUNTIMES_${target}_LIBC_ENABLE_USE_BY_CLANG ON CACHE BOOL "") + set(RUNTIMES_${target}_LLVM_INCLUDE_TESTS OFF CACHE BOOL "") + set(RUNTIMES_${target}_LLVM_ENABLE_ASSERTIONS OFF CACHE BOOL "") + set(RUNTIMES_${target}_LLVM_ENABLE_RUNTIMES "libc" CACHE STRING "") +endforeach() + set(LLVM_BUILTIN_TARGETS "${BUILTIN_TARGETS}" CACHE STRING "") set(LLVM_RUNTIME_TARGETS "${RUNTIME_TARGETS}" CACHE STRING "") diff --git a/clang/cmake/caches/Fuchsia.cmake b/clang/cmake/caches/Fuchsia.cmake index dad434be720da..fe925901eb3dc 100644 --- a/clang/cmake/caches/Fuchsia.cmake +++ b/clang/cmake/caches/Fuchsia.cmake @@ -6,7 +6,7 @@ set(LLVM_TARGETS_TO_BUILD X86;ARM;AArch64;RISCV CACHE STRING "") set(PACKAGE_VENDOR Fuchsia CACHE STRING "") -set(_FUCHSIA_ENABLE_PROJECTS "bolt;clang;clang-tools-extra;lld;llvm;polly") +set(_FUCHSIA_ENABLE_PROJECTS "bolt;clang;clang-tools-extra;libc;lld;llvm;polly") set(LLVM_ENABLE_DIA_SDK OFF CACHE BOOL "") set(LLVM_ENABLE_LIBEDIT OFF CACHE BOOL "") @@ -18,6 +18,7 @@ set(LLVM_ENABLE_Z3_SOLVER OFF CACHE BOOL "") set(LLVM_ENABLE_ZLIB OFF CACHE BOOL "") set(LLVM_INCLUDE_DOCS OFF CACHE BOOL "") set(LLVM_INCLUDE_EXAMPLES OFF CACHE BOOL "") +set(LIBC_HDRGEN_ONLY ON CACHE BOOL "") set(LLVM_USE_RELATIVE_PATHS_IN_FILES ON CACHE BOOL "") set(LLDB_ENABLE_CURSES OFF CACHE BOOL "") set(LLDB_ENABLE_LIBEDIT OFF CACHE BOOL "") @@ -38,9 +39,25 @@ set(_FUCHSIA_BOOTSTRAP_PASSTHROUGH CURL_ROOT OpenSSL_ROOT httplib_ROOT + + # Deprecated CursesAndPanel_ROOT + + CURSES_INCLUDE_DIRS + CURSES_LIBRARIES + PANEL_LIBRARIES + + # Deprecated Terminfo_ROOT + + Terminfo_LIBRARIES + + # Deprecated LibEdit_ROOT + + LibEdit_INCLUDE_DIRS + LibEdit_LIBRARIES + FUCHSIA_ENABLE_LLDB LLDB_ENABLE_CURSES LLDB_ENABLE_LIBEDIT diff --git a/clang/cmake/caches/Release.cmake b/clang/cmake/caches/Release.cmake index 3ea65ce26296c..a7b9a8d0e29f8 100644 --- a/clang/cmake/caches/Release.cmake +++ b/clang/cmake/caches/Release.cmake @@ -2,6 +2,9 @@ # BOOTSTRAP_* options configure the second build. # BOOTSTRAP_BOOTSTRAP_* options configure the third build. +# General Options +set(LLVM_RELEASE_ENABLE_LTO THIN CACHE STRING "") + set(CMAKE_BUILD_TYPE RELEASE CACHE STRING "") # Stage 1 Bootstrap Setup @@ -33,9 +36,17 @@ set(BOOTSTRAP_CLANG_BOOTSTRAP_TARGETS check-clang CACHE STRING "") # Stage 2 Options -set(BOOTSTRAP_LLVM_ENABLE_PROJECTS "clang" CACHE STRING "") +set(STAGE2_PROJECTS "clang") +if (LLVM_RELEASE_ENABLE_LTO) + list(APPEND STAGE2_PROJECTS "lld") +endif() +set(BOOTSTRAP_LLVM_ENABLE_PROJECTS ${STAGE2_PROJECTS} CACHE STRING "") set(BOOTSTRAP_LLVM_TARGETS_TO_BUILD Native CACHE STRING "") # Stage 3 Options set(BOOTSTRAP_BOOTSTRAP_LLVM_ENABLE_RUNTIMES "compiler-rt;libcxx;libcxxabi;libunwind" CACHE STRING "") set(BOOTSTRAP_BOOTSTRAP_LLVM_ENABLE_PROJECTS "clang;lld;lldb;clang-tools-extra;bolt;polly;mlir;flang" CACHE STRING "") +set(BOOTSTRAP_BOOTSTRAP_LLVM_ENABLE_LTO ${LLVM_RELEASE_ENABLE_LTO} CACHE STRING "") +if (LLVM_RELEASE_ENABLE_LTO) + set(BOOTSTRAP_BOOTSTRAP_LLVM_ENABLE_LLD ON CACHE BOOL "") +endif() diff --git a/clang/docs/BoundsSafety.rst b/clang/docs/BoundsSafety.rst new file mode 100644 index 0000000000000..f1837675ec9bf --- /dev/null +++ b/clang/docs/BoundsSafety.rst @@ -0,0 +1,998 @@ +================================================== +``-fbounds-safety``: Enforcing bounds safety for C +================================================== + +.. contents:: + :local: + +Overview +======== + +``-fbounds-safety`` is a C extension to enforce bounds safety to prevent +out-of-bounds (OOB) memory accesses, which remain a major source of security +vulnerabilities in C. ``-fbounds-safety`` aims to eliminate this class of bugs +by turning OOB accesses into deterministic traps. + +The ``-fbounds-safety`` extension offers bounds annotations that programmers can +use to attach bounds to pointers. For example, programmers can add the +``__counted_by(N)`` annotation to parameter ``ptr``, indicating that the pointer +has ``N`` valid elements: + +.. code-block:: c + + void foo(int *__counted_by(N) ptr, size_t N); + +Using this bounds information, the compiler inserts bounds checks on every +pointer dereference, ensuring that the program does not access memory outside +the specified bounds. The compiler requires programmers to provide enough bounds +information so that the accesses can be checked at either run time or compile +time — and it rejects code if it cannot. + +The most important contribution of ``-fbounds-safety`` is how it reduces the +programmer's annotation burden by reconciling bounds annotations at ABI +boundaries with the use of implicit wide pointers (a.k.a. "fat" pointers) that +carry bounds information on local variables without the need for annotations. We +designed this model so that it preserves ABI compatibility with C while +minimizing adoption effort. + +The ``-fbounds-safety`` extension has been adopted on millions of lines of +production C code and proven to work in a consumer operating system setting. The +extension was designed to enable incremental adoption — a key requirement in +real-world settings where modifying an entire project and its dependencies all +at once is often not possible. It also addresses multiple of other practical +challenges that have made existing approaches to safer C dialects difficult to +adopt, offering these properties that make it widely adoptable in practice: + +* It is designed to preserve the Application Binary Interface (ABI). +* It interoperates well with plain C code. +* It can be adopted partially and incrementally while still providing safety + benefits. +* It is a conforming extension to C. +* Consequently, source code that adopts the extension can continue to be + compiled by toolchains that do not support the extension (CAVEAT: this still + requires inclusion of a header file macro-defining bounds annotations to + empty). +* It has a relatively low adoption cost. + +This document discusses the key designs of ``-fbounds-safety``. The document is +subject to be actively updated with a more detailed specification. The +implementation plan can be found in :doc:`BoundsSafetyImplPlans`. + + +Programming Model +================= + +Overview +-------- + +``-fbounds-safety`` ensures that pointers are not used to access memory beyond +their bounds by performing bounds checking. If a bounds check fails, the program +will deterministically trap before out-of-bounds memory is accessed. + +In our model, every pointer has an explicit or implicit bounds attribute that +determines its bounds and ensures guaranteed bounds checking. Consider the +example below where the ``__counted_by(count)`` annotation indicates that +parameter ``p`` points to a buffer of integers containing ``count`` elements. An +off-by-one error is present in the loop condition, leading to ``p[i]`` being +out-of-bounds access during the loop's final iteration. The compiler inserts a +bounds check before ``p`` is dereferenced to ensure that the access remains +within the specified bounds. + +.. code-block:: c + + void fill_array_with_indices(int *__counted_by(count) p, unsigned count) { + // off-by-one error (i < count) + for (unsigned i = 0; i <= count; ++i) { + // bounds check inserted: + // if (i >= count) trap(); + p[i] = i; + } + } + +A bounds annotation defines an invariant for the pointer type, and the model +ensures that this invariant remains true. In the example below, pointer ``p`` +annotated with ``__counted_by(count)`` must always point to a memory buffer +containing at least ``count`` elements of the pointee type. Changing the value +of ``count``, like in the example below, may violate this invariant and permit +out-of-bounds access to the pointer. To avoid this, the compiler employs +compile-time restrictions and emits run-time checks as necessary to ensure the +new count value doesn't exceed the actual length of the buffer. Section +`Maintaining correctness of bounds annotations`_ provides more details about +this programming model. + +.. code-block:: c + + int g; + + void foo(int *__counted_by(count) p, size_t count) { + count++; // may violate the invariant of __counted_by + count--; // may violate the invariant of __counted_by if count was 0. + count = g; // may violate the invariant of __counted_by + // depending on the value of `g`. + } + +The requirement to annotate all pointers with explicit bounds information could +present a significant adoption burden. To tackle this issue, the model +incorporates the concept of a "wide pointer" (a.k.a. fat pointer) – a larger +pointer that carries bounds information alongside the pointer value. Utilizing +wide pointers can potentially reduce the adoption burden, as it contains bounds +information internally and eliminates the need for explicit bounds annotations. +However, wide pointers differ from standard C pointers in their data layout, +which may result in incompatibilities with the application binary interface +(ABI). Breaking the ABI complicates interoperability with external code that has +not adopted the same programming model. + +``-fbounds-safety`` harmonizes the wide pointer and the bounds annotation +approaches to reduce the adoption burden while maintaining the ABI. In this +model, local variables of pointer type are implicitly treated as wide pointers, +allowing them to carry bounds information without requiring explicit bounds +annotations. Please note that this approach doesn't apply to function parameters +which are considered ABI-visible. As local variables are typically hidden from +the ABI, this approach has a marginal impact on it. In addition, +``-fbounds-safety`` employs compile-time restrictions to prevent implicit wide +pointers from silently breaking the ABI (see `ABI implications of default bounds +annotations`_). Pointers associated with any other variables, including function +parameters, are treated as single object pointers (i.e., ``__single``), ensuring +that they always have the tightest bounds by default and offering a strong +bounds safety guarantee. + +By implementing default bounds annotations based on ABI visibility, a +considerable portion of C code can operate without modifications within this +programming model, reducing the adoption burden. + +The rest of the section will discuss individual bounds annotations and the +programming model in more detail. + +Bounds annotations +------------------ + +Annotation for pointers to a single object +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ + +The C language allows pointer arithmetic on arbitrary pointers and this has been +a source of many bounds safety issues. In practice, many pointers are merely +pointing to a single object and incrementing or decrementing such a pointer +immediately makes the pointer go out-of-bounds. To prevent this unsafety, +``-fbounds-safety`` provides the annotation ``__single`` that causes pointer +arithmetic on annotated pointers to be a compile time error. + +* ``__single`` : indicates that the pointer is either pointing to a single + object or null. Hence, pointers with ``__single`` do not permit pointer + arithmetic nor being subscripted with a non-zero index. Dereferencing a + ``__single`` pointer is allowed but it requires a null check. Upper and lower + bounds checks are not required because the ``__single`` pointer should point + to a valid object unless it's null. + +``__single`` is the default annotation for ABI-visible pointers. This +gives strong security guarantees in that these pointers cannot be incremented or +decremented unless they have an explicit, overriding bounds annotation that can +be used to verify the safety of the operation. The compiler issues an error when +a ``__single`` pointer is utilized for pointer arithmetic or array access, as +these operations would immediately cause the pointer to exceed its bounds. +Consequently, this prompts programmers to provide sufficient bounds information +to pointers. In the following example, the pointer on parameter p is +single-by-default, and is employed for array access. As a result, the compiler +generates an error suggesting to add ``__counted_by`` to the pointer. + +.. code-block:: c + + void fill_array_with_indices(int *p, unsigned count) { + for (unsigned i = 0; i < count; ++i) { + p[i] = i; // error + } + } + + +External bounds annotations +^^^^^^^^^^^^^^^^^^^^^^^^^^^ + +"External" bounds annotations provide a way to express a relationship between a +pointer variable and another variable (or expression) containing the bounds +information of the pointer. In the following example, ``__counted_by(count)`` +annotation expresses the bounds of parameter p using another parameter count. +This model works naturally with many C interfaces and structs because the bounds +of a pointer is often available adjacent to the pointer itself, e.g., at another +parameter of the same function prototype, or at another field of the same struct +declaration. + +.. code-block:: c + + void fill_array_with_indices(int *__counted_by(count) p, size_t count) { + // off-by-one error + for (size_t i = 0; i <= count; ++i) + p[i] = i; + } + +External bounds annotations include ``__counted_by``, ``__sized_by``, and +``__ended_by``. These annotations do not change the pointer representation, +meaning they do not have ABI implications. + +* ``__counted_by(N)`` : The pointer points to memory that contains ``N`` + elements of pointee type. ``N`` is an expression of integer type which can be + a simple reference to declaration, a constant including calls to constant + functions, or an arithmetic expression that does not have side effect. The + ``__counted_by`` annotation cannot apply to pointers to incomplete types or + types without size such as ``void *``. Instead, ``__sized_by`` can be used to + describe the byte count. +* ``__sized_by(N)`` : The pointer points to memory that contains ``N`` bytes. + Just like the argument of ``__counted_by``, ``N`` is an expression of integer + type which can be a constant, a simple reference to a declaration, or an + arithmetic expression that does not have side effects. This is mainly used for + pointers to incomplete types or types without size such as ``void *``. +* ``__ended_by(P)`` : The pointer has the upper bound of value ``P``, which is + one past the last element of the pointer. In other words, this annotation + describes a range that starts with the pointer that has this annotation and + ends with ``P`` which is the argument of the annotation. ``P`` itself may be + annotated with ``__ended_by(Q)``. In this case, the end of the range extends + to the pointer ``Q``. This is used for "iterator" support in C where you're + iterating from one pointer value to another until a final pointer value is + reached (and the final pointer value is not dereferencable). + +Accessing a pointer outside the specified bounds causes a run-time trap or a +compile-time error. Also, the model maintains correctness of bounds annotations +when the pointer and/or the related value containing the bounds information are +updated or passed as arguments. This is done by compile-time restrictions or +run-time checks (see `Maintaining correctness of bounds annotations`_ +for more detail). For instance, initializing ``buf`` with ``null`` while +assigning non-zero value to ``count``, as shown in the following example, would +violate the ``__counted_by`` annotation because a null pointer does not point to +any valid memory location. To avoid this, the compiler produces either a +compile-time error or run-time trap. + +.. code-block:: c + + void null_with_count_10(int *__counted_by(count) buf, unsigned count) { + buf = 0; + // This is not allowed as it creates a null pointer with non-zero length + count = 10; + } + +However, there are use cases where a pointer is either a null pointer or is +pointing to memory of the specified size. To support this idiom, +``-fbounds-safety`` provides ``*_or_null`` variants, +``__counted_by_or_null(N)``, ``__sized_by_or_null(N)``, and +``__ended_by_or_null(P)``. Accessing a pointer with any of these bounds +annotations will require an extra null check to avoid a null pointer +dereference. + +Internal bounds annotations +^^^^^^^^^^^^^^^^^^^^^^^^^^^ + +A wide pointer (sometimes known as a "fat" pointer) is a pointer that carries +additional bounds information internally (as part of its data). The bounds +require additional storage space making wide pointers larger than normal +pointers, hence the name "wide pointer". The memory layout of a wide pointer is +equivalent to a struct with the pointer, upper bound, and (optionally) lower +bound as its fields as shown below. + +.. code-block:: c + + struct wide_pointer_datalayout { + void* pointer; // Address used for dereferences and pointer arithmetic + void* upper_bound; // Points one past the highest address that can be + // accessed + void* lower_bound; // (Optional) Points to lowest address that can be + // accessed + }; + +Even with this representational change, wide pointers act syntactically as +normal pointers to allow standard pointer operations, such as pointer +dereference (``*p``), array subscript (``p[i]``), member access (``p->``), and +pointer arithmetic, with some restrictions on bounds-unsafe uses. + +``-fbounds-safety`` has a set of "internal" bounds annotations to turn pointers +into wide pointers. These are ``__bidi_indexable`` and ``__indexable``. When a +pointer has either of these annotations, the compiler changes the pointer to the +corresponding wide pointer. This means these annotations will break the ABI and +will not be compatible with plain C, and thus they should generally not be used +in ABI surfaces. + +* ``__bidi_indexable`` : A pointer with this annotation becomes a wide pointer + to carry the upper bound and the lower bound, the layout of which is + equivalent to ``struct { T *ptr; T *upper_bound; T *lower_bound; };``. As the + name indicates, pointers with this annotation are "bidirectionally indexable", + meaning that they can be indexed with either a negative or a positive offset + and the pointers can be incremented or decremented using pointer arithmetic. A + ``__bidi_indexable`` pointer is allowed to hold an out-of-bounds pointer + value. While creating an OOB pointer is undefined behavior in C, + ``-fbounds-safety`` makes it well-defined behavior. That is, pointer + arithmetic overflow with ``__bidi_indexable`` is defined as equivalent of + two's complement integer computation, and at the LLVM IR level this means + ``getelementptr`` won't get ``inbounds`` keyword. Accessing memory using the + OOB pointer is prevented via a run-time bounds check. + +* ``__indexable`` : A pointer with this annotation becomes a wide pointer + carrying the upper bound (but no explicit lower bound), the layout of which is + equivalent to ``struct { T *ptr; T *upper_bound; };``. Since ``__indexable`` + pointers do not have a separate lower bound, the pointer value itself acts as + the lower bound. An ``__indexable`` pointer can only be incremented or indexed + in the positive direction. Indexing it in the negative direction will trigger + a compile-time error. Otherwise, the compiler inserts a run-time + check to ensure pointer arithmetic doesn't make the pointer smaller than the + original ``__indexable`` pointer (Note that ``__indexable`` doesn't have a + lower bound so the pointer value is effectively the lower bound). As pointer + arithmetic overflow will make the pointer smaller than the original pointer, + it will cause a trap at runtime. Similar to ``__bidi_indexable``, an + ``__indexable`` pointer is allowed to have a pointer value above the upper + bound and creating such a pointer is well-defined behavior. Dereferencing such + a pointer, however, will cause a run-time trap. + +* ``__bidi_indexable`` offers the best flexibility out of all the pointer + annotations in this model, as ``__bidi_indexable`` pointers can be used for + any pointer operation. However, this comes with the largest code size and + memory cost out of the available pointer annotations in this model. In some + cases, use of the ``__bidi_indexable`` annotation may be duplicating bounds + information that exists elsewhere in the program. In such cases, using + external bounds annotations may be a better choice. + +``__bidi_indexable`` is the default annotation for non-ABI visible pointers, +such as local pointer variables — that is, if the programmer does not specify +another bounds annotation, a local pointer variable is implicitly +``__bidi_indexable``. Since ``__bidi_indexable`` pointers automatically carry +bounds information and have no restrictions on kinds of pointer operations that +can be used with these pointers, most code inside a function works as is without +modification. In the example below, ``int *buf`` doesn't require manual +annotation as it's implicitly ``int *__bidi_indexable buf``, carrying the bounds +information passed from the return value of malloc, which is necessary to insert +bounds checking for ``buf[i]``. + +.. code-block:: c + + void *__sized_by(size) malloc(size_t size); + + int *__counted_by(n) get_array_with_0_to_n_1(size_t n) { + int *buf = malloc(sizeof(int) * n); + for (size_t i = 0; i < n; ++i) + buf[i] = i; + return buf; + } + +Annotations for sentinel-delimited arrays +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ + +A C string is an array of characters. The null terminator — the first null +character ('\0') element in the array — marks the end of the string. +``-fbounds-safety`` provides ``__null_terminated`` to annotate C strings and the +generalized form ``__terminated_by(T)`` to annotate pointers and arrays with an +end marked by a sentinel value. The model prevents dereferencing a +``__terminated_by`` pointer beyond its end. Calculating the location of the end +(i.e., the address of the sentinel value), requires reading the entire array in +memory and would have some performance costs. To avoid an unintended performance +hit, the model puts some restrictions on how these pointers can be used. +``__terminated_by`` pointers cannot be indexed and can only be incremented one +element at a time. To allow these operations, the pointers must be explicitly +converted to ``__indexable`` pointers using the intrinsic function +``__unsafe_terminated_by_to_indexable(P, T)`` (or +``__unsafe_null_terminated_to_indexable(P)``) which converts the +``__terminated_by`` pointer ``P`` to an ``__indexable`` pointer. + +* ``__null_terminated`` : The pointer or array is terminated by ``NULL`` or + ``0``. Modifying the terminator or incrementing the pointer beyond it is + prevented at run time. + +* ``__terminated_by(T)`` : The pointer or array is terminated by ``T`` which is + a constant expression. Accessing or incrementing the pointer beyond the + terminator is not allowed. This is a generalization of ``__null_terminated`` + which is defined as ``__terminated_by(0)``. + +Annotation for interoperating with bounds-unsafe code +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ + +A pointer with the ``__unsafe_indexable`` annotation behaves the same as a plain +C pointer. That is, the pointer does not have any bounds information and pointer +operations are not checked. + +``__unsafe_indexable`` can be used to mark pointers from system headers or +pointers from code that has not adopted -fbounds safety. This enables +interoperation between code using ``-fbounds-safety`` and code that does not. + +Default pointer types +--------------------- + +ABI visibility and default annotations +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ + +Requiring ``-fbounds-safety`` adopters to add bounds annotations to all pointers +in the codebase would be a significant adoption burden. To avoid this and to +secure all pointers by default, ``-fbounds-safety`` applies default bounds +annotations to pointer types. +Default annotations apply to pointer types of declarations + +``-fbounds-safety`` applies default bounds annotations to pointer types used in +declarations. The default annotations are determined by the ABI visibility of +the pointer. A pointer type is ABI-visible if changing its size or +representation affects the ABI. For instance, changing the size of a type used +in a function parameter will affect the ABI and thus pointers used in function +parameters are ABI-visible pointers. On the other hand, changing the types of +local variables won't have such ABI implications. Hence, ``-fbounds-safety`` +considers the outermost pointer types of local variables as non-ABI visible. The +rest of the pointers such as nested pointer types, pointer types of global +variables, struct fields, and function prototypes are considered ABI-visible. + +All ABI-visible pointers are treated as ``__single`` by default unless annotated +otherwise. This default both preserves ABI and makes these pointers safe by +default. This behavior can be controlled with macros, i.e., +``__ptrcheck_abi_assume_*ATTR*()``, to set the default annotation for +ABI-visible pointers to be either ``__single``, ``__bidi_indexable``, +``__indexable``, or ``__unsafe_indexable``. For instance, +``__ptrcheck_abi_assume_unsafe_indexable()`` will make all ABI-visible pointers +be ``__unsafe_indexable``. Non-ABI visible pointers — the outermost pointer +types of local variables — are ``__bidi_indexable`` by default, so that these +pointers have the bounds information necessary to perform bounds checks without +the need for a manual annotation. All ``const char`` pointers or any typedefs +equivalent to ``const char`` pointers are ``__null_terminated`` by default. This +means that ``char8_t`` is ``unsigned char`` so ``const char8_t *`` won't be +``__null_terminated`` by default. Similarly, ``const wchar_t *`` won't be +``__null_terminated`` by default unless the platform defines it as ``typedef +char wchar_t``. Please note, however, that the programmers can still explicitly +use ``__null_terminated`` in any other pointers, e.g., ``char8_t +*__null_terminated``, ``wchar_t *__null_terminated``, ``int +*__null_terminated``, etc. if they should be treated as ``__null_terminated``. +The same applies to other annotations. +In system headers, the default pointer attribute for ABI-visible pointers is set +to ``__unsafe_indexable`` by default. + +The ``__ptrcheck_abi_assume_*ATTR*()`` macros are defined as pragmas in the +toolchain header (See `Portability with toolchains that do not support the +extension`_ for more details about the toolchain header): + +.. code-block:: C + +#define __ptrcheck_abi_assume_single() \ + _Pragma("clang abi_ptr_attr set(single)") + +#define __ptrcheck_abi_assume_indexable() \ + _Pragma("clang abi_ptr_attr set(indexable)") + +#define __ptrcheck_abi_assume_bidi_indexable() \ + _Pragma("clang abi_ptr_attr set(bidi_indexable)") + +#define __ptrcheck_abi_assume_unsafe_indexable() \ + _Pragma("clang abi_ptr_attr set(unsafe_indexable)") + + +ABI implications of default bounds annotations +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ + +Although simply modifying types of a local variable doesn't normally impact the +ABI, taking the address of such a modified type could create a pointer type that +has an ABI mismatch. Looking at the following example, ``int *local`` is +implicitly ``int *__bidi_indexable`` and thus the type of ``&local`` is a +pointer to ``int *__bidi_indexable``. On the other hand, in ``void foo(int +**)``, the parameter type is a pointer to ``int *__single`` (i.e., ``void +foo(int *__single *__single)``) (or a pointer to ``int *__unsafe_indexable`` if +it's from a system header). The compiler reports an error for casts between +pointers whose elements have incompatible pointer attributes. This way, +``-fbounds-safety`` prevents pointers that are implicitly ``__bidi_indexable`` +from silently escaping thereby breaking the ABI. + +.. code-block:: c + + void foo(int **); + + void bar(void) { + int *local = 0; + // error: passing 'int *__bidi_indexable*__bidi_indexable' to parameter of + // incompatible nested pointer type 'int *__single*__single' + foo(&local); + } + +A local variable may still be exposed to the ABI if ``typeof()`` takes the type +of local variable to define an interface as shown in the following example. + +.. code-block:: C + + // bar.c + void bar(int *) { ... } + + // foo.c + void foo(void) { + int *p; // implicitly `int *__bidi_indexable p` + extern void bar(typeof(p)); // creates an interface of type + // `void bar(int *__bidi_indexable)` + } + +Doing this may break the ABI if the parameter is not ``__bidi_indexable`` at the +definition of function ``bar()`` which is likely the case because parameters are +``__single`` by default without an explicit annotation. + +In order to avoid an implicitly wide pointer from silently breaking the ABI, the +compiler reports a warning when ``typeof()`` is used on an implicit wide pointer +at any ABI visible context (e.g., function prototype, struct definition, etc.). + +.. _Default pointer types in typeof: + +Default pointer types in ``typeof()`` +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ + +When ``typeof()`` takes an expression, it respects the bounds annotation on +the expression type, including the bounds annotation is implcit. For example, +the global variable ``g`` in the following code is implicitly ``__single`` so +``typeof(g)`` gets ``char *__single``. The similar is true for the parameter +``p``, so ``typeof(p)`` returns ``void *__single``. The local variable ``l`` is +implicitly ``__bidi_indexable``, so ``typeof(l)`` becomes +``int *__bidi_indexable``. + +.. code-block:: C + + char *g; // typeof(g) == char *__single + + void foo(void *p) { + // typeof(p) == void *__single + + int *l; // typeof(l) == int *__bidi_indexable + } + +When the type of expression has an "external" bounds annotation, e.g., +``__sized_by``, ``__counted_by``, etc., the compiler may report an error on +``typeof`` if the annotation creates a dependency with another declaration or +variable. For example, the compiler reports an error on ``typeof(p1)`` shown in +the following code because allowing it can potentially create another type +dependent on the parameter ``size`` in a different context (Please note that an +external bounds annotation on a parameter may only refer to another parameter of +the same function). On the other hand, ``typeof(p2)`` works resulting in ``int +*__counted_by(10)``, since it doesn't depend on any other declaration. + +.. TODO: add a section describing constraints on external bounds annotations + +.. code-block:: C + + void foo(int *__counted_by(size) p1, size_t size) { + // typeof(p1) == int *__counted_by(size) + // -> a compiler error as it tries to create another type + // dependent on `size`. + + int *__counted_by(10) p2; // typeof(p2) == int *__counted_by(10) + // -> no error + + } + +When ``typeof()`` takes a type name, the compiler doesn't apply an implicit +bounds annotation on the named pointer types. For example, ``typeof(int*)`` +returns ``int *`` without any bounds annotation. A bounds annotation may be +added after the fact depending on the context. In the following example, +``typeof(int *)`` returns ``int *`` so it's equivalent as the local variable is +declared as ``int *l``, so it eventually becomes implicitly +``__bidi_indexable``. + +.. code-block:: c + + void foo(void) { + typeof(int *) l; // `int *__bidi_indexable` (same as `int *l`) + } + +The programmers can still explicitly add a bounds annotation on the types named +inside ``typeof``, e.g., ``typeof(int *__bidi_indexable)``, which evaluates to +``int *__bidi_indexable``. + + +Default pointer types in ``sizeof()`` +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ + +When ``sizeof()`` takes a type name, the compiler doesn't apply an implicit +bounds annotation on the named pointer types. This means if a bounds annotation +is not specified, the evaluated pointer type is treated identically to a plain C +pointer type. Therefore, ``sizeof(int*)`` remains the same with or without +``-fbounds-safety``. That said, programmers can explicitly add attribute to the +types, e.g., ``sizeof(int *__bidi_indexable)``, in which case the sizeof +evaluates to the size of type ``int *__bidi_indexable`` (the value equivalent to +``3 * sizeof(int*)``). + +When ``sizeof()`` takes an expression, i.e., ``sizeof(expr``, it behaves as +``sizeof(typeof(expr))``, except that ``sizeof(expr)`` does not report an error +with ``expr`` that has a type with an external bounds annotation dependent on +another declaration, whereas ``typeof()`` on the same expression would be an +error as described in :ref:`Default pointer types in typeof`. +The following example describes this behavior. + +.. code-block:: c + + void foo(int *__counted_by(size) p, size_t size) { + // sizeof(p) == sizeof(int *__counted_by(size)) == sizeof(int *) + // typeof(p): error + }; + +Default pointer types in ``alignof()`` +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ + +``alignof()`` only takes a type name as the argument and it doesn't take an +expression. Similar to ``sizeof()`` and ``typeof``, the compiler doesn't apply +an implicit bounds annotation on the pointer types named inside ``alignof()``. +Therefore, ``alignof(T *)`` remains the same with or without +``-fbounds-safety``, evaluating into the alignment of the raw pointer ``T *``. +The programmers can explicitly add a bounds annotation to the types, e.g., +``alignof(int *__bidi_indexable)``, which returns the alignment of ``int +*__bidi_indexable``. A bounds annotation including an internal bounds annotation +(i.e., ``__indexable`` and ``__bidi_indexable``) doesn't affect the alignment of +the original pointer. Therefore, ``alignof(int *__bidi_indexable)`` is equal to +``alignof(int *)``. + + +Default pointer types used in C-style casts +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ + +A pointer type used in a C-style cast (e.g., ``(int *)src``) inherits the same +pointer attribute in the type of src. For instance, if the type of src is ``T +*__single`` (with ``T`` being an arbitrary C type), ``(int *)src`` will be ``int +*__single``. The reasoning behind this behavior is so that a C-style cast +doesn't introduce any unexpected side effects caused by an implicit cast of +bounds attribute. + +Pointer casts can have explicit bounds annotations. For instance, ``(int +*__bidi_indexable)src`` casts to ``int *__bidi_indexable`` as long as src has a +bounds annotation that can implicitly convert to ``__bidi_indexable``. If +``src`` has type ``int *__single``, it can implicitly convert to ``int +*__bidi_indexable`` which then will have the upper bound pointing to one past +the first element. However, if src has type ``int *__unsafe_indexable``, the +explicit cast ``(int *__bidi_indexable)src`` will cause an error because +``__unsafe_indexable`` cannot cast to ``__bidi_indexable`` as +``__unsafe_indexable`` doesn't have bounds information. `Cast rules`_ describes +in more detail what kinds of casts are allowed between pointers with different +bounds annotations. + +Default pointer types in typedef +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ + +Pointer types in ``typedef``\s do not have implicit default bounds annotations. +Instead, the bounds annotation is determined when the ``typedef`` is used. The +following example shows that no pointer annotation is specified in the ``typedef +pint_t`` while each instance of ``typedef``'ed pointer gets its bounds +annotation based on the context in which the type is used. + +.. code-block:: c + + typedef int * pint_t; // int * + + pint_t glob; // int *__single glob; + + void foo(void) { + pint_t local; // int *__bidi_indexable local; + } + +Pointer types in a ``typedef`` can still have explicit annotations, e.g., +``typedef int *__single``, in which case the bounds annotation ``__single`` will +apply to every use of the ``typedef``. + +Array to pointer promotion to secure arrays (including VLAs) +------------------------------------------------------------ + +Arrays on function prototypes +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ + +In C, arrays on function prototypes are promoted (or "decayed") to a pointer to +its first element (e.g., ``&arr[0]``). In ``-fbounds-safety``, arrays are also +decayed to pointers, but with the addition of an implicit bounds annotation, +which includes variable-length arrays (VLAs). As shown in the following example, +arrays on function prototypes are decalyed to corresponding ``__counted_by`` +pointers. + +.. code-block:: c + + // Function prototype: void foo(int n, int *__counted_by(n) arr); + void foo(int n, int arr[n]); + + // Function prototype: void bar(int *__counted_by(10) arr); + void bar(int arr[10]); + +This means the array parameters are treated as `__counted_by` pointers within +the function and callers of the function also see them as the corresponding +`__counted_by` pointers. + +Incomplete arrays on function prototypes will cause a compiler error unless it +has ``__counted_by`` annotation in its bracket. + +.. code-block:: c + + void f1(int n, int arr[]); // error + + void f3(int n, int arr[__counted_by(n)]); // ok + + void f2(int n, int arr[n]); // ok, decays to int *__counted_by(n) + + void f4(int n, int *__counted_by(n) arr); // ok + + void f5(int n, int *arr); // ok, but decays to int *__single, + // and cannot be used for pointer arithmetic + +Array references +^^^^^^^^^^^^^^^^ + +In C, similar to arrays on the function prototypes, a reference to array is +automatically promoted (or "decayed") to a pointer to its first element (e.g., +``&arr[0]``). + +In `-fbounds-safety`, array references are promoted to ``__bidi_indexable`` +pointers which contain the upper and lower bounds of the array, with the +equivalent of ``&arr[0]`` serving as the lower bound and ``&arr[array_size]`` +(or one past the last element) serving as the upper bound. This applies to all +types of arrays including constant-length arrays, variable-length arrays (VLAs), +and flexible array members annotated with `__counted_by`. + +In the following example, reference to ``vla`` promotes to ``int +*__bidi_indexable``, with ``&vla[n]`` as the upper bound and ``&vla[0]`` as the +lower bound. Then, it's copied to ``int *p``, which is implicitly ``int +*__bidi_indexable p``. Please note that value of ``n`` used to create the upper +bound is ``10``, not ``100``, in this case because ``10`` is the actual length +of ``vla``, the value of ``n`` at the time when the array is being allocated. + +.. code-block:: c + + void foo(void) { + int n = 10; + int vla[n]; + n = 100; + int *p = vla; // { .ptr: &vla[0], .upper: &vla[10], .lower: &vla[0] } + // it's `&vla[10]` because the value of `n` was 10 at the + // time when the array is actually allocated. + // ... + } + +By promoting array references to ``__bidi_indexable``, all array accesses are +bounds checked in ``-fbounds-safety``, just as ``__bidi_indexable`` pointers +are. + +Maintaining correctness of bounds annotations +--------------------------------------------- + +``-fbounds-safety`` maintains correctness of bounds annotations by performing +additional checks when a pointer object and/or its related value containing the +bounds information is updated. + +For example, ``__single`` expresses an invariant that the pointer must either +point to a single valid object or be a null pointer. To maintain this invariant, +the compiler inserts checks when initializing a ``__single`` pointer, as shown +in the following example: + +.. code-block:: c + + void foo(void *__sized_by(size) vp, size_t size) { + // Inserted check: + // if ((int*)upper_bound(vp) - (int*)vp < sizeof(int) && !!vp) trap(); + int *__single ip = (int *)vp; + } + +Additionally, an explicit bounds annotation such as ``int *__counted_by(count) +buf`` defines a relationship between two variables, ``buf`` and ``count``: +namely, that ``buf`` has ``count`` number of elements available. This +relationship must hold even after any of these related variables are updated. To +this end, the model requires that assignments to ``buf`` and ``count`` must be +side by side, with no side effects between them. This prevents ``buf`` and +``count`` from temporarily falling out of sync due to updates happening at a +distance. + +The example below shows a function ``alloc_buf`` that initializes a struct that +members that use the ``__counted_by`` annotation. The compiler allows these +assignments because ``sbuf->buf`` and ``sbuf->count`` are updated side by side +without any side effects in between the assignments. + +Furthermore, the compiler inserts additional run-time checks to ensure the new +``buf`` has at least as many elements as the new ``count`` indicates as shown in +the transformed pseudo code of function ``alloc_buf()`` in the example below. + +.. code-block:: c + + typedef struct { + int *__counted_by(count) buf; + size_t count; + } sized_buf_t; + + void alloc_buf(sized_buf_t *sbuf, sized_t nelems) { + sbuf->buf = (int *)malloc(sizeof(int) * nelems); + sbuf->count = nelems; + } + + // Transformed pseudo code: + void alloc_buf(sized_buf_t *sbuf, sized_t nelems) { + // Materialize RHS values: + int *tmp_ptr = (int *)malloc(sizeof(int) * nelems); + int tmp_count = nelems; + // Inserted check: + // - checks to ensure that `lower <= tmp_ptr <= upper` + // - if (upper(tmp_ptr) - tmp_ptr < tmp_count) trap(); + sbuf->buf = tmp_ptr; + sbuf->count = tmp_count; + } + +Whether the compiler can optimize such run-time checks depends on how the upper +bound of the pointer is derived. If the source pointer has ``__sized_by``, +``__counted_by``, or a variant of such, the compiler assumes that the upper +bound calculation doesn't overflow, e.g., ``ptr + size`` (where the type of +``ptr`` is ``void *__sized_by(size)``), because when the ``__sized_by`` pointer +is initialized, ``-fbounds-safety`` inserts run-time checks to ensure that ``ptr ++ size`` doesn't overflow and that ``size >= 0``. + +Assuming the upper bound calculation doesn't overflow, the compiler can simplify +the trap condition ``upper(tmp_ptr) - tmp_ptr < tmp_count`` to ``size < +tmp_count`` so if both ``size`` and ``tmp_count`` values are known at compile +time such that ``0 <= tmp_count <= size``, the optimizer can remove the check. + +``ptr + size`` may still overflow if the ``__sized_by`` pointer is created from +code that doesn't enable ``-fbounds-safety``, which is undefined behavior. + +In the previous code example with the transformed ``alloc_buf()``, the upper +bound of ``tmp_ptr`` is derived from ``void *__sized_by_or_null(size)``, which +is the return type of ``malloc()``. Hence, the pointer arithmetic doesn't +overflow or ``tmp_ptr`` is null. Therefore, if ``nelems`` was given as a +compile-time constant, the compiler could remove the checks. + +Cast rules +---------- + +``-fbounds-safety`` does not enforce overall type safety and bounds invariants +can still be violated by incorrect casts in some cases. That said, +``-fbounds-safety`` prevents type conversions that change bounds attributes in a +way to violate the bounds invariant of the destination's pointer annotation. +Type conversions that change bounds attributes may be allowed if it does not +violate the invariant of the destination or that can be verified at run time. +Here are some of the important cast rules. + +Two pointers that have different bounds annotations on their nested pointer +types are incompatible and cannot implicitly cast to each other. For example, +``T *__single *__single`` cannot be converted to ``T *__bidi_indexable +*__single``. Such a conversion between incompatible nested bounds annotations +can be allowed using an explicit cast (e.g., C-style cast). Hereafter, the rules +only apply to the top pointer types. ``__unsafe_indexable`` cannot be converted +to any other safe pointer types (``__single``, ``__bidi_indexable``, +``__counted_by``, etc) using a cast. The extension provides builtins to force +this conversion, ``__unsafe_forge_bidi_indexable(type, pointer, char_count)`` to +convert pointer to a ``__bidi_indexable`` pointer of type with ``char_count`` +bytes available and ``__unsafe_forge_single(type, pointer)`` to convert pointer +to a single pointer of type type. The following examples show the usage of these +functions. Function ``example_forge_bidi()`` gets an external buffer from an +unsafe library by calling ``get_buf()`` which returns ``void +*__unsafe_indexable.`` Under the type rules, this cannot be directly assigned to +``void *buf`` (implicitly ``void *__bidi_indexable``). Thus, +``__unsafe_forge_bidi_indexable`` is used to manually create a +``__bidi_indexable`` from the unsafe buffer. + +.. code-block:: c + + // unsafe_library.h + void *__unsafe_indexable get_buf(void); + size_t get_buf_size(void); + + // my_source1.c (enables -fbounds-safety) + #include "unsafe_library.h" + void example_forge_bidi(void) { + void *buf = + __unsafe_forge_bidi_indexable(void *, get_buf(), get_buf_size()); + // ... + } + + // my_source2.c (enables -fbounds-safety) + #include + void example_forge_single(void) { + FILE *fp = __unsafe_forge_single(FILE *, fopen("mypath", "rb")); + // ... + } + +* Function ``example_forge_single`` takes a file handle by calling fopen defined + in system header ``stdio.h``. Assuming ``stdio.h`` did not adopt + ``-fbounds-safety``, the return type of ``fopen`` would implicitly be ``FILE + *__unsafe_indexable`` and thus it cannot be directly assigned to ``FILE *fp`` + in the bounds-safe source. To allow this operation, ``__unsafe_forge_single`` + is used to create a ``__single`` from the return value of ``fopen``. + +* Similar to ``__unsafe_indexable``, any non-pointer type (including ``int``, + ``intptr_t``, ``uintptr_t``, etc.) cannot be converted to any safe pointer + type because these don't have bounds information. ``__unsafe_forge_single`` or + ``__unsafe_forge_bidi_indexable`` must be used to force the conversion. + +* Any safe pointer types can cast to ``__unsafe_indexable`` because it doesn't + have any invariant to maintain. + +* ``__single`` casts to ``__bidi_indexable`` if the pointee type has a known + size. After the conversion, the resulting ``__bidi_indexable`` has the size of + a single object of the pointee type of ``__single``. ``__single`` cannot cast + to ``__bidi_indexable`` if the pointee type is incomplete or sizeless. For + example, ``void *__single`` cannot convert to ``void *__bidi_indexable`` + because void is an incomplete type and thus the compiler cannot correctly + determine the upper bound of a single void pointer. + +* Similarly, ``__single`` can cast to ``__indexable`` if the pointee type has a + known size. The resulting ``__indexable`` has the size of a single object of + the pointee type. + +* ``__single`` casts to ``__counted_by(E)`` only if ``E`` is 0 or 1. + +* ``__single`` can cast to ``__single`` including when they have different + pointee types as long as it is allowed in the underlying C standard. + ``-fbounds-safety`` doesn't guarantee type safety. + +* ``__bidi_indexable`` and ``__indexable`` can cast to ``__single``. The + compiler may insert run-time checks to ensure the pointer has at least a + single element or is a null pointer. + +* ``__bidi_indexable`` casts to ``__indexable`` if the pointer does not have an + underflow. The compiler may insert run-time checks to ensure the pointer is + not below the lower bound. + +* ``__indexable`` casts to ``__bidi_indexable``. The resulting + ``__bidi_indexable`` gets the lower bound same as the pointer value. + +* A type conversion may involve both a bitcast and a bounds annotation cast. For + example, casting from ``int *__bidi_indexable`` to ``char *__single`` involve + a bitcast (``int *`` to ``char *``) and a bounds annotation cast + (``__bidi_indexable`` to ``__single``). In this case, the compiler performs + the bitcast and then converts the bounds annotation. This means, ``int + *__bidi_indexable`` will be converted to ``char *__bidi_indexable`` and then + to ``char *__single``. + +* ``__terminated_by(T)`` cannot cast to any safe pointer type without the same + ``__terminated_by(T)`` attribute. To perform the cast, programmers can use an + intrinsic function such as ``__unsafe_terminated_by_to_indexable(P)`` to force + the conversion. + +* ``__terminated_by(T)`` can cast to ``__unsafe_indexable``. + +* Any type without ``__terminated_by(T)`` cannot cast to ``__terminated_by(T)`` + without explicitly using an intrinsic function to allow it. + + + ``__unsafe_terminated_by_from_indexable(T, PTR [, PTR_TO_TERM])`` casts any + safe pointer PTR to a ``__terminated_by(T)`` pointer. ``PTR_TO_TERM`` is an + optional argument where the programmer can provide the exact location of the + terminator. With this argument, the function can skip reading the entire + array in order to locate the end of the pointer (or the upper bound). + Providing an incorrect ``PTR_TO_TERM`` causes a run-time trap. + + + ``__unsafe_forge_terminated_by(T, P, E)`` creates ``T __terminated_by(E)`` + pointer given any pointer ``P``. Tmust be a pointer type. + +Portability with toolchains that do not support the extension +------------------------------------------------------------- + +The language model is designed so that it doesn't alter the semantics of the +original C program, other than introducing deterministic traps where otherwise +the behavior is undefined and/or unsafe. Clang provides a toolchain header +(``ptrcheck.h``) that macro-defines the annotations as type attributes when +``-fbounds-safety`` is enabled and defines them to empty when the extension is +disabled. Thus, the code adopting ``-fbounds-safety`` can compile with +toolchains that do not support this extension, by including the header or adding +macros to define the annotations to empty. For example, the toolchain not +supporting this extension may not have a header defining ``__counted_by``, so +the code using ``__counted_by`` must define it as nothing or include a header +that has the define. + +.. code-block:: c + + #if defined(__has_feature) && __has_feature(bounds_safety) + #define __counted_by(T) __attribute__((__counted_by__(T))) + // ... other bounds annotations + #else #define __counted_by(T) // defined as nothing + // ... other bounds annotations + #endif + + // expands to `void foo(int * ptr, size_t count);` + // when extension is not enabled or not available + void foo(int *__counted_by(count) ptr, size_t count); + +Other potential applications of bounds annotations +================================================== + +The bounds annotations provided by the ``-fbounds-safety`` programming model +have potential use cases beyond the language extension itself. For example, +static and dynamic analysis tools could use the bounds information to improve +diagnostics for out-of-bounds accesses, even if ``-fbounds-safety`` is not used. +The bounds annotations could be used to improve C interoperability with +bounds-safe languages, providing a better mapping to bounds-safe types in the +safe language interface. The bounds annotations can also serve as documentation +specifying the relationship between declarations. + +Limitations +=========== + +``-fbounds-safety`` aims to bring the bounds safety guarantee to the C language, +and it does not guarantee other types of memory safety properties. Consequently, +it may not prevent some of the secondary bounds safety violations caused by +other types of safety violations such as type confusion. For instance, +``-fbounds-safety`` does not perform type-safety checks on conversions between +`__single`` pointers of different pointee types (e.g., ``char *__single`` → +``void *__single`` → ``int *__single``) beyond what the foundation languages +(C/C++) already offer. + +``-fbounds-safety`` heavily relies on run-time checks to keep the bounds safety +and the soundness of the type system. This may incur significant code size +overhead in unoptimized builds and leaving some of the adoption mistakes to be +caught only at run time. This is not a fundamental limitation, however, because +incrementally adding necessary static analysis will allow us to catch issues +early on and remove unnecessary bounds checks in unoptimized builds. \ No newline at end of file diff --git a/clang/docs/BoundsSafetyImplPlans.rst b/clang/docs/BoundsSafetyImplPlans.rst new file mode 100644 index 0000000000000..4fbf87f966350 --- /dev/null +++ b/clang/docs/BoundsSafetyImplPlans.rst @@ -0,0 +1,255 @@ +============================================ +Implementation plans for ``-fbounds-safety`` +============================================ + +.. contents:: + :local: + +External bounds annotations +=========================== + +The bounds annotations are C type attributes appertaining to pointer types. If +an attribute is added to the position of a declaration attribute, e.g., ``int +*ptr __counted_by(size)``, the attribute appertains to the outermost pointer +type of the declaration (``int *``). + +New sugar types +=============== + +An external bounds annotation creates a type sugar of the underlying pointer +types. We will introduce a new sugar type, ``DynamicBoundsPointerType`` to +represent ``__counted_by`` or ``__sized_by``. Using ``AttributedType`` would not +be sufficient because the type needs to hold the count or size expression as +well as some metadata necessary for analysis, while this type may be implemented +through inheritance from ``AttributedType``. Treating the annotations as type +sugars means two types with incompatible external bounds annotations may be +considered canonically the same types. This is sometimes necessary, for example, +to make the ``__counted_by`` and friends not participate in function +overloading. However, this design requires a separate logic to walk through the +entire type hierarchy to check type compatibility of bounds annotations. + +Late parsing for C +================== + +A bounds annotation such as ``__counted_by(count)`` can be added to type of a +struct field declaration where count is another field of the same struct +declared later. Similarly, the annotation may apply to type of a function +parameter declaration which precedes the parameter count in the same function. +This means parsing the argument of bounds annotations must be done after the +parser has the whole context of a struct or a function declaration. Clang has +late parsing logic for C++ declaration attributes that require late parsing, +while the C declaration attributes and C/C++ type attributes do not have the +same logic. This requires introducing late parsing logic for C/C++ type +attributes. + +Internal bounds annotations +=========================== + +``__indexable`` and ``__bidi_indexable`` alter pointer representations to be +equivalent to a struct with the pointer and the corresponding bounds fields. +Despite this difference in their representations, they are still pointers in +terms of types of operations that are allowed and their semantics. For instance, +a pointer dereference on a ``__bidi_indexable`` pointer will return the +dereferenced value same as plain C pointers, modulo the extra bounds checks +being performed before dereferencing the wide pointer. This means mapping the +wide pointers to struct types with equivalent layout won’t be sufficient. To +represent the wide pointers in Clang AST, we add an extra field in the +PointerType class to indicate the internal bounds of the pointer. This ensures +pointers of different representations are mapped to different canonical types +while they are still treated as pointers. + +In LLVM IR, wide pointers will be emitted as structs of equivalent +representations. Clang CodeGen will handle them as Aggregate in +``TypeEvaluationKind (TEK)``. ``AggExprEmitter`` was extended to handle pointer +operations returning wide pointers. Alternatively, a new ``TEK`` and an +expression emitter dedicated to wide pointers could be introduced. + +Default bounds annotations +========================== + +The model may implicitly add ``__bidi_indexable`` or ``__single`` depending on +the context of the declaration that has the pointer type. ``__bidi_indexable`` +implicitly adds to local variables, while ``__single`` implicitly adds to +pointer types specifying struct fields, function parameters, or global +variables. This means the parser may first create the pointer type without any +default pointer attribute and then recreate the type once the parser has the +declaration context and determined the default attribute accordingly. + +This also requires the parser to reset the type of the declaration with the +newly created type with the right default attribute. + +Promotion expression +==================== + +A new expression will be introduced to represent the conversion from a pointer +with an external bounds annotation, such as ``__counted_by``, to +``__bidi_indexable``. This type of conversion cannot be handled by normal +CastExprs because it requires an extra subexpression(s) to provide the bounds +information necessary to create a wide pointer. + +Bounds check expression +======================= + +Bounds checks are part of semantics defined in the ``-fbounds-safety`` language +model. Hence, exposing the bounds checks and other semantic actions in the AST +is desirable. A new expression for bounds checks has been added to the AST. The +bounds check expression has a ``BoundsCheckKind`` to indicate the kind of checks +and has the additional sub-expressions that are necessary to perform the check +according to the kind. + +Paired assignment check +======================= + +``-fbounds-safety`` enforces that variables or fields related with the same +external bounds annotation (e.g., ``buf`` and ``count`` related with +``__counted_by`` in the example below) must be updated side by side within the +same basic block and without side effect in between. + +.. code-block:: c + + typedef struct { + int *__counted_by(count) buf; size_t count; + } sized_buf_t; + + void alloc_buf(sized_buf_t *sbuf, sized_t nelems) { + sbuf->buf = (int *)malloc(sizeof(int) * nelems); + sbuf->count = nelems; + } + +To implement this rule, the compiler requires a linear representation of +statements to understand the ordering and the adjacency between the two or more +assignments. The Clang CFG is used to implement this analysis as Clang CFG +provides a linear view of statements within each ``CFGBlock`` (Clang +``CFGBlock`` represents a single basic block in a source-level CFG). + +Bounds check optimizations +========================== + +In ``-fbounds-safety``, the Clang frontend emits run-time checks for every +memory dereference if the type system or analyses in the frontend couldn’t +verify its bounds safety. The implementation relies on LLVM optimizations to +remove redundant run-time checks. Using this optimization strategy, if the +original source code already has bounds checks, the fewer additional checks +``-fbounds-safety`` will introduce. The LLVM ``ConstraintElimination`` pass is +design to remove provable redundant checks (please check Florian Hahn’s +presentation in 2021 LLVM Dev Meeting and the implementation to learn more). In +the following example, ``-fbounds-safety`` implicitly adds the redundant bounds +checks that the optimizer can remove: + +.. code-block:: c + + void fill_array_with_indices(int *__counted_by(count) p, size_t count) { + for (size_t i = 0; i < count; ++i) { + // implicit bounds checks: + // if (p + i < p || p + i + 1 > p + count) trap(); + p[i] = i; + } + } + +``ConstraintElimination`` collects the following facts and determines if the +bounds checks can be safely removed: + +* Inside the for-loop, ``0 <= i < count``, hence ``1 <= i + 1 <= count``. +* Pointer arithmetic ``p + count`` in the if-condition doesn’t wrap. +* ``-fbounds-safety`` treats pointer arithmetic overflow as deterministically + two’s complement computation, not an undefined behavior. Therefore, + getelementptr does not typically have inbounds keyword. However, the compiler + does emit inbounds for ``p + count`` in this case because + ``__counted_by(count)`` has the invariant that p has at least as many as + elements as count. Using this information, ``ConstraintElimination`` is able + to determine ``p + count`` doesn’t wrap. +* Accordingly, ``p + i`` and ``p + i + 1`` also don’t wrap. +* Therefore, ``p <= p + i`` and ``p + i + 1 <= p + count``. +* The if-condition simplifies to false and becomes dead code that the subsequent + optimization passes can remove. + +``OptRemarks`` can be utilized to provide insights into performance tuning. It +has the capability to report on checks that it cannot eliminate, possibly with +reasons, allowing programmers to adjust their code to unlock further +optimizations. + +Debugging +========= + +Internal bounds annotations +--------------------------- + +Internal bounds annotations change a pointer into a wide pointer. The debugger +needs to understand that wide pointers are essentially pointers with a struct +layout. To handle this, a wide pointer is described as a record type in the +debug info. The type name has a special name prefix (e.g., +``__bounds_safety$bidi_indexable``) which can be recognized by a debug info +consumer to provide support that goes beyond showing the internal structure of +the wide pointer. There are no DWARF extensions needed to support wide pointers. +In our implementation, LLDB recognizes wide pointer types by name and +reconstructs them as wide pointer Clang AST types for use in the expression +evaluator. + +External bounds annotations +--------------------------- + +Similar to internal bounds annotations, external bound annotations are described +as a typedef to their underlying pointer type in the debug info, and the bounds +are encoded as strings in the typedef’s name (e.g., +``__bounds_safety$counted_by:N``). + +Recognizing ``-fbounds-safety`` traps +------------------------------------- + +Clang emits debug info for ``-fbounds-safety`` traps as inlined functions, where +the function name encodes the error message. LLDB implements a frame recognizer +to surface a human-readable error cause to the end user. A debug info consumer +that is unaware of this sees an inlined function whose name encodes an error +message (e.g., : ``__bounds_safety$Bounds check failed``). + +Expression Parsing +------------------ + +In our implementation, LLDB’s expression evaluator does not enable the +``-fbounds-safety`` language option because it’s currently unable to fully +reconstruct the pointers with external bounds annotations, and also because the +evaluator operates in C++ mode, utilizing C++ reference types, while +``-fbounds-safety`` does not currently support C++. This means LLDB’s expression +evaluator can only evaluate a subset of the ``-fbounds-safety`` language model. +Specifically, it’s capable of evaluating the wide pointers that already exist in +the source code. All other expressions are evaluated according to C/C++ +semantics. + +C++ support +=========== + +C++ has multiple options to write code in a bounds-safe manner, such as +following the bounds-safety core guidelines and/or using hardened libc++ along +with the `C++ Safe Buffer model +`_. However, these +techniques may require ABI changes and may not be applicable to code +interoperating with C. When the ABI of an existing program needs to be preserved +and for headers shared between C and C++, ``-fbounds-safety`` offers a potential +solution. + +``-fbounds-safety`` is not currently supported in C++, but we believe the +general approach would be applicable for future efforts. + +Upstreaming plan +================ + +Gradual updates with experimental flag +-------------------------------------- + +The upstreaming will take place as a series of smaller PRs and we will guard our +implementation with an experimental flag ``-fexperimental-bounds-safety`` until +the usable model is fully upstreamed. Once the model is ready for use, we will +expose the flag ``-fbounds-safety``. + +Possible patch sets +------------------- + +* External bounds annotations and the (late) parsing logic. +* Internal bounds annotations (wide pointers) and their parsing logic. +* Clang code generation for wide pointers with debug information. +* Pointer cast semantics involving bounds annotations (this could be divided + into multiple sub-PRs). +* CFG analysis for pairs of related pointer and count assignments and the likes. +* Bounds check expressions in AST and the Clang code generation (this could also + be divided into multiple sub-PRs). + diff --git a/clang/docs/ClangFormatStyleOptions.rst b/clang/docs/ClangFormatStyleOptions.rst index ac9a0b70ed5da..8bc13e45bf2f5 100644 --- a/clang/docs/ClangFormatStyleOptions.rst +++ b/clang/docs/ClangFormatStyleOptions.rst @@ -4432,6 +4432,11 @@ the configuration (without a prefix: ``Auto``). **PenaltyBreakOpenParenthesis** (``Unsigned``) :versionbadge:`clang-format 14` :ref:`¶ ` The penalty for breaking after ``(``. +.. _PenaltyBreakScopeResolution: + +**PenaltyBreakScopeResolution** (``Unsigned``) :versionbadge:`clang-format 18` :ref:`¶ ` + The penalty for breaking after ``::``. + .. _PenaltyBreakString: **PenaltyBreakString** (``Unsigned``) :versionbadge:`clang-format 3.7` :ref:`¶ ` diff --git a/clang/docs/CommandGuide/clang.rst b/clang/docs/CommandGuide/clang.rst index e1c872cdc5539..643365215f875 100644 --- a/clang/docs/CommandGuide/clang.rst +++ b/clang/docs/CommandGuide/clang.rst @@ -302,7 +302,7 @@ Language Selection and Mode Options .. option:: -fmsc-version= Set ``_MSC_VER``. When on Windows, this defaults to either the same value as - the currently installed version of cl.exe, or ``1920``. Not set otherwise. + the currently installed version of cl.exe, or ``1933``. Not set otherwise. .. option:: -fborland-extensions diff --git a/clang/docs/HLSL/FunctionCalls.rst b/clang/docs/HLSL/FunctionCalls.rst index 996ddd6944b1c..7317de2163f89 100644 --- a/clang/docs/HLSL/FunctionCalls.rst +++ b/clang/docs/HLSL/FunctionCalls.rst @@ -144,7 +144,7 @@ undefined behavior in HLSL, and any use of the argument after the call is a use of an undefined value which may be illegal in the target (DXIL programs with used or potentially used ``undef`` or ``poison`` values fail validation). -Clang Implementation +Clang Implementation ==================== .. note:: diff --git a/clang/docs/InternalsManual.rst b/clang/docs/InternalsManual.rst index a866f621c1003..b3e2b870ae5f9 100644 --- a/clang/docs/InternalsManual.rst +++ b/clang/docs/InternalsManual.rst @@ -931,7 +931,7 @@ the option appears on the command line, the argument value is simply copied. .. code-block:: text def isysroot : JoinedOrSeparate<["-"], "isysroot">, - Visibility<[ClangOption, CC1Option]>, + Visibility<[ClangOption, CC1Option, FlangOption]>, MarshallingInfoString, [{"/"}]>; **List of Strings** diff --git a/clang/docs/LibASTMatchersReference.html b/clang/docs/LibASTMatchersReference.html index fcd3114bb5231..c40d679e383bb 100644 --- a/clang/docs/LibASTMatchersReference.html +++ b/clang/docs/LibASTMatchersReference.html @@ -1629,6 +1629,17 @@

Node Matchers

+Matcher<Stmt>cxxFoldExprMatcher<CXXFoldExpr>... +
Matches C++17 fold expressions.
+
+Example matches `(0 + ... + args)`:
+  template <typename... Args>
+  auto sum(Args... args) {
+      return (0 + ... + args);
+  }
+
+ + Matcher<Stmt>cxxForRangeStmtMatcher<CXXForRangeStmt>... 
Matches range-based for statements.
 
@@ -2965,11 +2976,18 @@ 
Narrowing Matchers

 
 
 Matcher<BinaryOperator>hasOperatorNamestd::string Name
-
Matches the operator Name of operator expressions (binary or
-unary).
+
Matches the operator Name of operator expressions and fold expressions
+(binary or unary).
 
 Example matches a || b (matcher = binaryOperator(hasOperatorName("||")))
   !(a || b)
+
+Example matches `(0 + ... + args)`
+    (matcher = cxxFoldExpr(hasOperatorName("+")))
+  template <typename... Args>
+  auto sum(Args... args) {
+      return (0 + ... + args);
+  }
 

 
 
@@ -3430,6 +3448,91 @@ 
Narrowing Matchers

 

 
 
+Matcher<CXXFoldExpr>hasOperatorNamestd::string Name
+
Matches the operator Name of operator expressions and fold expressions
+(binary or unary).
+
+Example matches a || b (matcher = binaryOperator(hasOperatorName("||")))
+  !(a || b)
+
+Example matches `(0 + ... + args)`
+    (matcher = cxxFoldExpr(hasOperatorName("+")))
+  template <typename... Args>
+  auto sum(Args... args) {
+      return (0 + ... + args);
+  }
+

+
+
+Matcher<CXXFoldExpr>isBinaryFold
+
Matches binary fold expressions, i.e. fold expressions with an initializer.
+
+Example matches `(0 + ... + args)`
+    (matcher = cxxFoldExpr(isBinaryFold()))
+  template <typename... Args>
+  auto sum(Args... args) {
+      return (0 + ... + args);
+  }
+
+  template <typename... Args>
+  auto multiply(Args... args) {
+      return (args * ...);
+  }
+

+
+
+Matcher<CXXFoldExpr>isLeftFold
+
Matches left-folding fold expressions.
+
+Example matches `(0 + ... + args)`
+    (matcher = cxxFoldExpr(isLeftFold()))
+  template <typename... Args>
+  auto sum(Args... args) {
+      return (0 + ... + args);
+  }
+
+  template <typename... Args>
+  auto multiply(Args... args) {
+      return (args * ... * 1);
+  }
+

+
+
+Matcher<CXXFoldExpr>isRightFold
+
Matches right-folding fold expressions.
+
+Example matches `(args * ... * 1)`
+    (matcher = cxxFoldExpr(isRightFold()))
+  template <typename... Args>
+  auto sum(Args... args) {
+      return (0 + ... + args);
+  }
+
+  template <typename... Args>
+  auto multiply(Args... args) {
+      return (args * ... * 1);
+  }
+

+
+
+Matcher<CXXFoldExpr>isUnaryFold
+
Matches unary fold expressions, i.e. fold expressions without an
+initializer.
+
+Example matches `(args * ...)`
+    (matcher = cxxFoldExpr(isUnaryFold()))
+  template <typename... Args>
+  auto sum(Args... args) {
+      return (0 + ... + args);
+  }
+
+  template <typename... Args>
+  auto multiply(Args... args) {
+      return (args * ...);
+  }
+

+
+
 Matcher<CXXMethodDecl>isConst
 
Matches if the given method declaration is const.
 
@@ -3599,11 +3702,18 @@ 
Narrowing Matchers

 
 
 Matcher<CXXOperatorCallExpr>hasOperatorNamestd::string Name
-
Matches the operator Name of operator expressions (binary or
-unary).
+
Matches the operator Name of operator expressions and fold expressions
+(binary or unary).
 
 Example matches a || b (matcher = binaryOperator(hasOperatorName("||")))
   !(a || b)
+
+Example matches `(0 + ... + args)`
+    (matcher = cxxFoldExpr(hasOperatorName("+")))
+  template <typename... Args>
+  auto sum(Args... args) {
+      return (0 + ... + args);
+  }
 

 
 
@@ -3757,11 +3867,18 @@ 
Narrowing Matchers

 
 
 Matcher<CXXRewrittenBinaryOperator>hasOperatorNamestd::string Name
-
Matches the operator Name of operator expressions (binary or
-unary).
+
Matches the operator Name of operator expressions and fold expressions
+(binary or unary).
 
 Example matches a || b (matcher = binaryOperator(hasOperatorName("||")))
   !(a || b)
+
+Example matches `(0 + ... + args)`
+    (matcher = cxxFoldExpr(hasOperatorName("+")))
+  template <typename... Args>
+  auto sum(Args... args) {
+      return (0 + ... + args);
+  }
 

 
 
@@ -5711,12 +5828,19 @@ 
Narrowing Matchers

 

 
 
-Matcher<UnaryOperator>hasOperatorNamestd::string Name
-
Matches the operator Name of operator expressions (binary or
-unary).
+Matcher<UnaryOperator>hasOperatorNamestd::string Name
+
Matches the operator Name of operator expressions and fold expressions
+(binary or unary).
 
 Example matches a || b (matcher = binaryOperator(hasOperatorName("||")))
   !(a || b)
+
+Example matches `(0 + ... + args)`
+    (matcher = cxxFoldExpr(hasOperatorName("+")))
+  template <typename... Args>
+  auto sum(Args... args) {
+      return (0 + ... + args);
+  }
 

 
 
@@ -6453,7 +6577,7 @@ 
AST Traversal Matchers

 
 Matcher<BinaryOperator>hasEitherOperandMatcher<Expr> InnerMatcher
 
Matches if either the left hand side or the right hand side of a
-binary operator matches.
+binary operator or fold expression matches.
 

 
 
@@ -6466,7 +6590,8 @@ 
AST Traversal Matchers

 
 
 Matcher<BinaryOperator>hasOperandsMatcher<Expr> Matcher1, Matcher<Expr> Matcher2
-
Matches if both matchers match with opposite sides of the binary operator.
+
Matches if both matchers match with opposite sides of the binary operator
+or fold expression.
 
 Example matcher = binaryOperator(hasOperands(integerLiteral(equals(1),
                                              integerLiteral(equals(2)))
@@ -6885,6 +7010,112 @@ 
AST Traversal Matchers

 

 
 
+Matcher<CXXFoldExpr>calleeMatcher<Stmt> InnerMatcher
+
Matches if the call or fold expression's callee expression matches.
+
+Given
+  class Y { void x() { this->x(); x(); Y y; y.x(); } };
+  void f() { f(); }
+callExpr(callee(expr()))
+  matches this->x(), x(), y.x(), f()
+with callee(...)
+  matching this->x, x, y.x, f respectively
+
+Given
+  template <typename... Args>
+  auto sum(Args... args) {
+      return (0 + ... + args);
+  }
+
+  template <typename... Args>
+  auto multiply(Args... args) {
+      return (args * ... * 1);
+  }
+cxxFoldExpr(callee(expr()))
+  matches (args * ... * 1)
+with callee(...)
+  matching *
+
+Note: Callee cannot take the more general internal::Matcher<Expr>
+because this introduces ambiguous overloads with calls to Callee taking a
+internal::Matcher<Decl>, as the matcher hierarchy is purely
+implemented in terms of implicit casts.
+

+
+
+Matcher<CXXFoldExpr>hasEitherOperandMatcher<Expr> InnerMatcher
+
Matches if either the left hand side or the right hand side of a
+binary operator or fold expression matches.
+

+
+
+Matcher<CXXFoldExpr>hasFoldInitast_matchers::Matcher<Expr> InnerMacher
+
Matches the operand that does not contain the parameter pack.
+
+Example matches `(0 + ... + args)` and `(args * ... * 1)`
+    (matcher = cxxFoldExpr(hasFoldInit(expr())))
+  with hasFoldInit(...)
+    matching `0` and `1` respectively
+  template <typename... Args>
+  auto sum(Args... args) {
+      return (0 + ... + args);
+  }
+
+  template <typename... Args>
+  auto multiply(Args... args) {
+      return (args * ... * 1);
+  }
+

+
+
+Matcher<CXXFoldExpr>hasLHSMatcher<Expr> InnerMatcher
+
Matches the left hand side of binary operator expressions.
+
+Example matches a (matcher = binaryOperator(hasLHS()))
+  a || b
+

+
+
+Matcher<CXXFoldExpr>hasOperandsMatcher<Expr> Matcher1, Matcher<Expr> Matcher2
+
Matches if both matchers match with opposite sides of the binary operator
+or fold expression.
+
+Example matcher = binaryOperator(hasOperands(integerLiteral(equals(1),
+                                             integerLiteral(equals(2)))
+  1 + 2 // Match
+  2 + 1 // Match
+  1 + 1 // No match
+  2 + 2 // No match
+

+
+
+Matcher<CXXFoldExpr>hasPatternast_matchers::Matcher<Expr> InnerMacher
+
Matches the operand that contains the parameter pack.
+
+Example matches `(0 + ... + args)`
+    (matcher = cxxFoldExpr(hasPattern(expr())))
+  with hasPattern(...)
+    matching `args`
+  template <typename... Args>
+  auto sum(Args... args) {
+      return (0 + ... + args);
+  }
+
+  template <typename... Args>
+  auto multiply(Args... args) {
+      return (args * ... * 1);
+  }
+

+
+
+Matcher<CXXFoldExpr>hasRHSMatcher<Expr> InnerMatcher
+
Matches the right hand side of binary operator expressions.
+
+Example matches b (matcher = binaryOperator(hasRHS()))
+  a || b
+

+
+
 Matcher<CXXForRangeStmt>hasBodyMatcher<Stmt> InnerMatcher
 
Matches a 'for', 'while', 'while' statement or a function or coroutine
 definition that has a given body. Note that in case of functions or
@@ -7179,7 +7410,7 @@ 
AST Traversal Matchers

 
 Matcher<CXXOperatorCallExpr>hasEitherOperandMatcher<Expr> InnerMatcher
 
Matches if either the left hand side or the right hand side of a
-binary operator matches.
+binary operator or fold expression matches.
 

 
 
@@ -7192,7 +7423,8 @@ 
AST Traversal Matchers

 
 
 Matcher<CXXOperatorCallExpr>hasOperandsMatcher<Expr> Matcher1, Matcher<Expr> Matcher2
-
Matches if both matchers match with opposite sides of the binary operator.
+
Matches if both matchers match with opposite sides of the binary operator
+or fold expression.
 
 Example matcher = binaryOperator(hasOperands(integerLiteral(equals(1),
                                              integerLiteral(equals(2)))
@@ -7317,9 +7549,9 @@ 
AST Traversal Matchers

 

 
 
-Matcher<CXXRewrittenBinaryOperator>hasEitherOperandMatcher<Expr> InnerMatcher
-
Matches if either the left hand side or the right hand side of a
-binary operator matches.
+Matcher<CXXRewrittenBinaryOperator>hasEitherOperandMatcher<Expr> InnerMatcher
+
Matches if either the left hand side or the right hand side of a
+binary operator or fold expression matches.
 

 
 
@@ -7331,8 +7563,9 @@ 
AST Traversal Matchers

 

 
 
-Matcher<CXXRewrittenBinaryOperator>hasOperandsMatcher<Expr> Matcher1, Matcher<Expr> Matcher2
-
Matches if both matchers match with opposite sides of the binary operator.
+Matcher<CXXRewrittenBinaryOperator>hasOperandsMatcher<Expr> Matcher1, Matcher<Expr> Matcher2
+
Matches if both matchers match with opposite sides of the binary operator
+or fold expression.
 
 Example matcher = binaryOperator(hasOperands(integerLiteral(equals(1),
                                              integerLiteral(equals(2)))
@@ -7436,8 +7669,8 @@ 
AST Traversal Matchers

 

 
 
-Matcher<CallExpr>calleeMatcher<Decl> InnerMatcher
-
Matches 1) if the call expression's callee's declaration matches the
+Matcher<CallExpr>calleeMatcher<Decl> InnerMatcher
+
Matches 1) if the call expression's callee's declaration matches the
 given matcher; or 2) if the Obj-C message expression's callee's method
 declaration matches the given matcher.
 
@@ -7458,7 +7691,7 @@ 
AST Traversal Matchers

 
 
 Matcher<CallExpr>calleeMatcher<Stmt> InnerMatcher
-
Matches if the call expression's callee expression matches.
+
Matches if the call or fold expression's callee expression matches.
 
 Given
   class Y { void x() { this->x(); x(); Y y; y.x(); } };
@@ -7468,6 +7701,21 @@ 
AST Traversal Matchers

 with callee(...)
   matching this->x, x, y.x, f respectively
 
+Given
+  template <typename... Args>
+  auto sum(Args... args) {
+      return (0 + ... + args);
+  }
+
+  template <typename... Args>
+  auto multiply(Args... args) {
+      return (args * ... * 1);
+  }
+cxxFoldExpr(callee(expr()))
+  matches (args * ... * 1)
+with callee(...)
+  matching *
+
 Note: Callee cannot take the more general internal::Matcher<Expr>
 because this introduces ambiguous overloads with calls to Callee taking a
 internal::Matcher<Decl>, as the matcher hierarchy is purely
@@ -9087,8 +9335,8 @@ 
AST Traversal Matchers

 

 
 
-Matcher<ObjCMessageExpr>calleeMatcher<Decl> InnerMatcher
-
Matches 1) if the call expression's callee's declaration matches the
+Matcher<ObjCMessageExpr>calleeMatcher<Decl> InnerMatcher
+
Matches 1) if the call expression's callee's declaration matches the
 given matcher; or 2) if the Obj-C message expression's callee's method
 declaration matches the given matcher.
 
diff --git a/clang/docs/LibASTMatchersTutorial.rst b/clang/docs/LibASTMatchersTutorial.rst
index 2a58502a8de4e..d2883688ebfac 100644
--- a/clang/docs/LibASTMatchersTutorial.rst
+++ b/clang/docs/LibASTMatchersTutorial.rst
@@ -22,7 +22,7 @@ started guide `_.
 
 .. code-block:: console
 
-      cd ~/clang-llvm
+      mkdir ~/clang-llvm && cd ~/clang-llvm
       git clone https://github.com/llvm/llvm-project.git
 
 Next you need to obtain the CMake build system and Ninja build tool.
@@ -33,11 +33,11 @@ Next you need to obtain the CMake build system and Ninja build tool.
       git clone https://github.com/martine/ninja.git
       cd ninja
       git checkout release
-      ./bootstrap.py
+      ./configure.py --bootstrap
       sudo cp ninja /usr/bin/
 
       cd ~/clang-llvm
-      git clone git://cmake.org/stage/cmake.git
+      git clone https://gitlab.kitware.com/cmake/cmake.git
       cd cmake
       git checkout next
       ./bootstrap
@@ -50,7 +50,7 @@ Okay. Now we'll build Clang!
 
       cd ~/clang-llvm
       mkdir build && cd build
-      cmake -G Ninja ../llvm -DLLVM_ENABLE_PROJECTS="clang;clang-tools-extra" -DCMAKE_BUILD_TYPE=Release -DLLVM_BUILD_TESTS=ON
+      cmake -G Ninja ../llvm-project/llvm -DLLVM_ENABLE_PROJECTS="clang;clang-tools-extra" -DCMAKE_BUILD_TYPE=Release -DLLVM_BUILD_TESTS=ON
       ninja
       ninja check       # Test LLVM only.
       ninja clang-test  # Test Clang only.
@@ -65,7 +65,7 @@ Finally, we want to set Clang as its own compiler.
 .. code-block:: console
 
       cd ~/clang-llvm/build
-      ccmake ../llvm
+      ccmake ../llvm-project/llvm
 
 The second command will bring up a GUI for configuring Clang. You need
 to set the entry for ``CMAKE_CXX_COMPILER``. Press ``'t'`` to turn on
@@ -89,7 +89,7 @@ live in the ``clang-tools-extra`` repository.
 
 .. code-block:: console
 
-      cd ~/clang-llvm
+      cd ~/clang-llvm/llvm-project
       mkdir clang-tools-extra/loop-convert
       echo 'add_subdirectory(loop-convert)' >> clang-tools-extra/CMakeLists.txt
       vim clang-tools-extra/loop-convert/CMakeLists.txt
@@ -314,7 +314,7 @@ handiwork:
 
 .. code-block:: console
 
-      cd ~/clang-llvm/llvm/llvm_build/
+      cd ~/clang-llvm/build/
       ninja loop-convert
       vim ~/test-files/simple-loops.cc
       bin/loop-convert ~/test-files/simple-loops.cc
diff --git a/clang/docs/ReleaseNotes.rst b/clang/docs/ReleaseNotes.rst
index a18d36a16b1a9..cf52c1e66b91b 100644
--- a/clang/docs/ReleaseNotes.rst
+++ b/clang/docs/ReleaseNotes.rst
@@ -37,6 +37,29 @@ These changes are ones which we think may surprise users when upgrading to
 Clang |release| because of the opportunity they pose for disruption to existing
 code bases.
 
+- Fix a bug in reversed argument for templated operators.
+  This breaks code in C++20 which was previously accepted in C++17.
+  Clang did not properly diagnose such casese in C++20 before this change. Eg:
+
+  .. code-block:: cpp
+
+    struct P {};
+    template bool operator==(const P&, const S&);
+
+    struct A : public P {};
+    struct B : public P {};
+
+    // This equality is now ambiguous in C++20.
+    bool ambiguous(A a, B b) { return a == b; }
+
+    template bool operator!=(const P&, const S&);
+    // Ok. Found a matching operator!=.
+    bool fine(A a, B b) { return a == b; }
+
+  To reduce such widespread breakages, as an extension, Clang accepts this code
+  with an existing warning ``-Wambiguous-reversed-operator`` warning.
+  Fixes `GH `_.
+
 - The CMake variable ``GCC_INSTALL_PREFIX`` (which sets the default
   ``--gcc-toolchain=``) is deprecated and will be removed. Specify
   ``--gcc-install-dir=`` or ``--gcc-triple=`` in a `configuration file
@@ -74,6 +97,10 @@ C/C++ Language Potentially Breaking Changes
   outlined in "The Equality Operator You Are Looking For" (`P2468 `_).
   Fixes (`#68901: `_).
 
+- Remove the hardcoded path to the imported modules for C++20 named modules. Now we
+  require all the dependent modules to specified from the command line.
+  See (`#62707: `_).
+
 C++ Specific Potentially Breaking Changes
 -----------------------------------------
 - The name mangling rules for function templates has been changed to take into
@@ -200,6 +227,11 @@ C++2c Feature Support
 Resolutions to C++ Defect Reports
 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
 
+- Implemented `CWG2598 `_ and `CWG2096 `_,
+  making unions (that have either no members or at least one literal member) literal types.
+  (`#77924: `_).
+
+
 C Language Changes
 ------------------
 - ``structs``, ``unions``, and ``arrays`` that are const may now be used as
@@ -249,6 +281,8 @@ Non-comprehensive list of changes in this release
 
 * Added ``#pragma clang fp reciprocal``.
 
+* The version of Unicode used by Clang (primarily to parse identifiers) has been updated to 15.1.
+
 New Compiler Flags
 ------------------
 
@@ -533,6 +567,34 @@ Improvements to Clang's diagnostics
 - Clang now diagnoses unexpanded packs within the template argument lists of function template specializations.
 - Clang now diagnoses attempts to bind a bitfield to an NTTP of a reference type as erroneous
   converted constant expression and not as a reference to subobject.
+- Clang now diagnoses the requirement that non-template friend declarations with requires clauses
+  and template friend declarations with a constraint that depends on a template parameter from an
+  enclosing template must be a definition.
+- Clang now diagnoses function/variable templates that shadow their own template parameters, e.g. ``template void T();``.
+
+- Clang now emits more descriptive diagnostics for 'unusual' expressions (e.g. incomplete index
+  expressions on matrix types or builtin functions without an argument list) as placement-args
+  to new-expressions.
+
+  Before:
+
+  .. code-block:: text
+
+    error: no matching function for call to 'operator new'
+       13 |     new (__builtin_memset) S {};
+          |     ^   ~~~~~~~~~~~~~~~~~~
+
+    note: candidate function not viable: no known conversion from '' to 'int' for 2nd argument
+        5 |     void* operator new(__SIZE_TYPE__, int);
+          |           ^
+
+  After:
+
+  .. code-block:: text
+
+    error: builtin functions must be directly called
+       13 |     new (__builtin_memset) S {};
+          |          ^
 
 
 Improvements to Clang's time-trace
@@ -653,6 +715,8 @@ Bug Fixes in This Version
   Fixes (`#67317 `_)
 - Clang now properly diagnoses use of stand-alone OpenMP directives after a
   label (including ``case`` or ``default`` labels).
+- Fix compiler memory leak for enums with underlying type larger than 64 bits.
+  Fixes (`#78311 `_)
 
   Before:
 
@@ -719,7 +783,27 @@ Bug Fixes in This Version
 - Clang now emits correct source location for code-coverage regions in `if constexpr`
   and `if consteval` branches.
   Fixes (`#54419 `_)
-
+- Fix assertion failure when declaring a template friend function with
+  a constrained parameter in a template class that declares a class method
+  or lambda at different depth.
+  Fixes (`#75426 `_)
+- Fix an issue where clang cannot find conversion function with template
+  parameter when instantiation of template class.
+  Fixes (`#77583 `_)
+- Fix an issue where CTAD fails for function-type/array-type arguments.
+  Fixes (`#51710 `_)
+- Fix crashes when using the binding decl from an invalid structured binding.
+  Fixes (`#67495 `_) and
+  (`#72198 `_)
+- Fix assertion failure when call noreturn-attribute function with musttail
+  attribute.
+  Fixes (`#76631 `_)
+  - The MS ``__noop`` builtin without an argument list is now accepted
+  in the placement-args of new-expressions, matching MSVC's behaviour.
+- Fix an issue that caused MS ``__decspec(property)`` accesses as well as
+  Objective-C++ property accesses to not be converted to a function call
+  to the getter in the placement-args of new-expressions.
+  Fixes (`#65053 `_)
 
 Bug Fixes to Compiler Builtins
 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
@@ -880,6 +964,9 @@ Bug Fixes to C++ Support
   (`#64607 `_)
   (`#64086 `_)
 
+- Fixed a crash where we lost uninstantiated constraints on placeholder NTTP packs. Fixes:
+  (`#63837 `_)
+
 - Fixed a regression where clang forgets how to substitute into constraints on template-template
   parameters. Fixes:
   (`#57410 `_) and
@@ -925,6 +1012,8 @@ Miscellaneous Clang Crashes Fixed
 - Fixed a crash when a lambda marked as ``static`` referenced a captured
   variable in an expression.
   `Issue 74608 `_
+- Fixed a crash with modules and a ``constexpr`` destructor.
+  `Issue 68702 `_
 
 
 OpenACC Specific Changes
@@ -1028,6 +1117,8 @@ RISC-V Support
 ^^^^^^^^^^^^^^
 - Unaligned memory accesses can be toggled by ``-m[no-]unaligned-access`` or the
   aliases ``-m[no-]strict-align``.
+- CodeGen of RV32E/RV64E was supported experimentally.
+- CodeGen of ilp32e/lp64e was supported experimentally.
 
 - Default ABI with F but without D was changed to ilp32f for RV32 and to lp64f
   for RV64.
@@ -1089,6 +1180,9 @@ AST Matchers
 - Add ``convertVectorExpr``.
 - Add ``dependentSizedExtVectorType``.
 - Add ``macroQualifiedType``.
+- Add ``CXXFoldExpr`` related matchers: ``cxxFoldExpr``, ``callee``,
+  ``hasInit``, ``hasPattern``, ``isRightFold``, ``isLeftFold``,
+  ``isUnaryFold``, ``isBinaryFold``, ``hasOperator``, ``hasLHS``, ``hasRHS``, ``hasEitherOperand``.
 
 clang-format
 ------------
@@ -1098,6 +1192,7 @@ clang-format
 - Add ``BreakAdjacentStringLiterals`` option.
 - Add ``ObjCPropertyAttributeOrder`` which can be used to sort ObjC property
   attributes (like ``nonatomic, strong, nullable``).
+- Add ``PenaltyBreakScopeResolution`` option.
 - Add ``.clang-format-ignore`` files.
 - Add ``AlignFunctionPointers`` sub-option for ``AlignConsecutiveDeclarations``.
 
diff --git a/clang/docs/StandardCPlusPlusModules.rst b/clang/docs/StandardCPlusPlusModules.rst
index 4970457574030..22d506f0da2b1 100644
--- a/clang/docs/StandardCPlusPlusModules.rst
+++ b/clang/docs/StandardCPlusPlusModules.rst
@@ -143,7 +143,7 @@ Then we type:
 .. code-block:: console
 
   $ clang++ -std=c++20 Hello.cppm --precompile -o Hello.pcm
-  $ clang++ -std=c++20 use.cpp -fprebuilt-module-path=. Hello.pcm -o Hello.out
+  $ clang++ -std=c++20 use.cpp -fmodule-file=Hello=Hello.pcm Hello.pcm -o Hello.out
   $ ./Hello.out
   Hello World!
 
@@ -200,15 +200,15 @@ Then we are able to compile the example by the following command:
   $ clang++ -std=c++20 interface_part.cppm --precompile -o M-interface_part.pcm
   $ clang++ -std=c++20 impl_part.cppm --precompile -fprebuilt-module-path=. -o M-impl_part.pcm
   $ clang++ -std=c++20 M.cppm --precompile -fprebuilt-module-path=. -o M.pcm
-  $ clang++ -std=c++20 Impl.cpp -fmodule-file=M=M.pcm -c -o Impl.o
+  $ clang++ -std=c++20 Impl.cpp -fprebuilt-module-path=. -c -o Impl.o
 
   # Compiling the user
   $ clang++ -std=c++20 User.cpp -fprebuilt-module-path=. -c -o User.o
 
   # Compiling the module and linking it together
-  $ clang++ -std=c++20 M-interface_part.pcm -c -o M-interface_part.o
-  $ clang++ -std=c++20 M-impl_part.pcm -c -o M-impl_part.o
-  $ clang++ -std=c++20 M.pcm -c -o M.o
+  $ clang++ -std=c++20 M-interface_part.pcm -fprebuilt-module-path=. -c -o M-interface_part.o
+  $ clang++ -std=c++20 M-impl_part.pcm -fprebuilt-module-path=. -c -o M-impl_part.o
+  $ clang++ -std=c++20 M.pcm -fprebuilt-module-path=. -c -o M.o
   $ clang++ User.o M-interface_part.o  M-impl_part.o M.o Impl.o -o a.out
 
 We explain the options in the following sections.
@@ -218,7 +218,6 @@ How to enable standard C++ modules
 
 Currently, standard C++ modules are enabled automatically
 if the language standard is ``-std=c++20`` or newer.
-The ``-fmodules-ts`` option is deprecated and is planned to be removed.
 
 How to produce a BMI
 ~~~~~~~~~~~~~~~~~~~~
diff --git a/clang/docs/UsersManual.rst b/clang/docs/UsersManual.rst
index 7c30570437e8b..881d903d91a7e 100644
--- a/clang/docs/UsersManual.rst
+++ b/clang/docs/UsersManual.rst
@@ -3963,6 +3963,60 @@ implicitly included in later levels.
 - ``-march=x86-64-v3``: (close to Haswell) AVX, AVX2, BMI1, BMI2, F16C, FMA, LZCNT, MOVBE, XSAVE
 - ``-march=x86-64-v4``: AVX512F, AVX512BW, AVX512CD, AVX512DQ, AVX512VL
 
+`Intel AVX10 ISA `_ is
+a major new vector ISA incorporating the modern vectorization aspects of
+Intel AVX-512. This ISA will be supported on all future Intel processors.
+Users are supposed to use the new options ``-mavx10.N`` and ``-mavx10.N-512``
+on these processors and should not use traditional AVX512 options anymore.
+
+The ``N`` in ``-mavx10.N`` represents a continuous integer number starting
+from ``1``. ``-mavx10.N`` is an alias of ``-mavx10.N-256``, which means to
+enable all instructions within AVX10 version N at a maximum vector length of
+256 bits. ``-mavx10.N-512`` enables all instructions at a maximum vector
+length of 512 bits, which is a superset of instructions ``-mavx10.N`` enabled.
+
+Current binaries built with AVX512 features can run on Intel AVX10/512 capable
+processors without re-compile, but cannot run on AVX10/256 capable processors.
+Users need to re-compile their code with ``-mavx10.N``, and maybe update some
+code that calling to 512-bit X86 specific intrinsics and passing or returning
+512-bit vector types in function call, if they want to run on AVX10/256 capable
+processors. Binaries built with ``-mavx10.N`` can run on both AVX10/256 and
+AVX10/512 capable processors.
+
+Users can add a ``-mno-evex512`` in the command line with AVX512 options if
+they want to run the binary on both legacy AVX512 and new AVX10/256 capable
+processors. The option has the same constraints as ``-mavx10.N``, i.e.,
+cannot call to 512-bit X86 specific intrinsics and pass or return 512-bit vector
+types in function call.
+
+Users should avoid using AVX512 features in function target attributes when
+developing code for AVX10. If they have to do so, they need to add an explicit
+``evex512`` or ``no-evex512`` together with AVX512 features for 512-bit or
+non-512-bit functions respectively to avoid unexpected code generation. Both
+command line option and target attribute of EVEX512 feature can only be used
+with AVX512. They don't affect vector size of AVX10.
+
+User should not mix the use AVX10 and AVX512 options together at any time,
+because the option combinations are conflicting sometimes. For example, a
+combination of ``-mavx512f -mavx10.1-256`` doesn't show a clear intention to
+compiler, since instructions in AVX512F and AVX10.1/256 intersect but do not
+overlap. In this case, compiler will emit warning for it, but the behavior
+is determined. It will generate the same code as option ``-mavx10.1-512``.
+A similar case is ``-mavx512f -mavx10.2-256``, which equals to
+``-mavx10.1-512 -mavx10.2-256``, because ``avx10.2-256`` implies ``avx10.1-256``
+and ``-mavx512f -mavx10.1-256`` equals to ``-mavx10.1-512``.
+
+There are some new macros introduced with AVX10 support. ``-mavx10.1-256`` will
+enable ``__AVX10_1__`` and ``__EVEX256__``, while ``-mavx10.1-512`` enables
+``__AVX10_1__``, ``__EVEX256__``, ``__EVEX512__``  and ``__AVX10_1_512__``.
+Besides, both ``-mavx10.1-256`` and ``-mavx10.1-512`` will enable all AVX512
+feature specific macros. A AVX512 feature will enable both ``__EVEX256__``,
+``__EVEX512__`` and its own macro. So ``__EVEX512__`` can be used to guard code
+that can run on both legacy AVX512 and AVX10/512 capable processors but cannot
+run on AVX10/256, while a AVX512 macro like ``__AVX512F__`` cannot tell the
+difference among the three options. Users need to check additional macros
+``__AVX10_1__`` and ``__EVEX512__`` if they want to make distinction.
+
 ARM
 ^^^
 
@@ -4382,11 +4436,11 @@ Execute ``clang-cl /?`` to see a list of supported options:
       -fmerge-all-constants   Allow merging of constants
       -fms-compatibility-version=
                               Dot-separated value representing the Microsoft compiler version
-                              number to report in _MSC_VER (0 = don't define it (default))
+                              number to report in _MSC_VER (0 = don't define it; default is same value as installed cl.exe, or 1933)
       -fms-compatibility      Enable full Microsoft Visual C++ compatibility
       -fms-extensions         Accept some non-standard constructs supported by the Microsoft compiler
       -fmsc-version=   Microsoft compiler version number to report in _MSC_VER
-                              (0 = don't define it (default))
+                              (0 = don't define it; default is same value as installed cl.exe, or 1933)
       -fno-addrsig            Don't emit an address-significance table
       -fno-builtin-    Disable implicit builtin knowledge of a specific function
       -fno-builtin            Disable implicit builtin knowledge of functions
diff --git a/clang/docs/index.rst b/clang/docs/index.rst
index 5453a19564b87..a35a867b96bd7 100644
--- a/clang/docs/index.rst
+++ b/clang/docs/index.rst
@@ -35,6 +35,8 @@ Using Clang as a Compiler
    SanitizerCoverage
    SanitizerStats
    SanitizerSpecialCaseList
+   BoundsSafety
+   BoundsSafetyImplPlans
    ControlFlowIntegrity
    LTOVisibility
    SafeStack
diff --git a/clang/include/clang/AST/APNumericStorage.h b/clang/include/clang/AST/APNumericStorage.h
new file mode 100644
index 0000000000000..95eddbcd86e83
--- /dev/null
+++ b/clang/include/clang/AST/APNumericStorage.h
@@ -0,0 +1,71 @@
+//===--- APNumericStorage.h - Store APInt/APFloat in ASTContext -*- C++ -*-===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#ifndef LLVM_CLANG_AST_APNUMERICSTORAGE_H
+#define LLVM_CLANG_AST_APNUMERICSTORAGE_H
+
+#include "llvm/ADT/APFloat.h"
+#include "llvm/ADT/APInt.h"
+
+namespace clang {
+class ASTContext;
+
+/// Used by IntegerLiteral/FloatingLiteral/EnumConstantDecl to store the
+/// numeric without leaking memory.
+///
+/// For large floats/integers, APFloat/APInt will allocate memory from the heap
+/// to represent these numbers.  Unfortunately, when we use a BumpPtrAllocator
+/// to allocate IntegerLiteral/FloatingLiteral nodes the memory associated with
+/// the APFloat/APInt values will never get freed. APNumericStorage uses
+/// ASTContext's allocator for memory allocation.
+class APNumericStorage {
+  union {
+    uint64_t VAL;   ///< Used to store the <= 64 bits integer value.
+    uint64_t *pVal; ///< Used to store the >64 bits integer value.
+  };
+  unsigned BitWidth;
+
+  bool hasAllocation() const { return llvm::APInt::getNumWords(BitWidth) > 1; }
+
+  APNumericStorage(const APNumericStorage &) = delete;
+  void operator=(const APNumericStorage &) = delete;
+
+protected:
+  APNumericStorage() : VAL(0), BitWidth(0) {}
+
+  llvm::APInt getIntValue() const {
+    unsigned NumWords = llvm::APInt::getNumWords(BitWidth);
+    if (NumWords > 1)
+      return llvm::APInt(BitWidth, NumWords, pVal);
+    else
+      return llvm::APInt(BitWidth, VAL);
+  }
+  void setIntValue(const ASTContext &C, const llvm::APInt &Val);
+};
+
+class APIntStorage : private APNumericStorage {
+public:
+  llvm::APInt getValue() const { return getIntValue(); }
+  void setValue(const ASTContext &C, const llvm::APInt &Val) {
+    setIntValue(C, Val);
+  }
+};
+
+class APFloatStorage : private APNumericStorage {
+public:
+  llvm::APFloat getValue(const llvm::fltSemantics &Semantics) const {
+    return llvm::APFloat(Semantics, getIntValue());
+  }
+  void setValue(const ASTContext &C, const llvm::APFloat &Val) {
+    setIntValue(C, Val.bitcastToAPInt());
+  }
+};
+
+} // end namespace clang
+
+#endif // LLVM_CLANG_AST_APNUMERICSTORAGE_H
diff --git a/clang/include/clang/AST/Decl.h b/clang/include/clang/AST/Decl.h
index a9b215efb266a..e4e380f057b9d 100644
--- a/clang/include/clang/AST/Decl.h
+++ b/clang/include/clang/AST/Decl.h
@@ -13,6 +13,7 @@
 #ifndef LLVM_CLANG_AST_DECL_H
 #define LLVM_CLANG_AST_DECL_H
 
+#include "clang/AST/APNumericStorage.h"
 #include "clang/AST/APValue.h"
 #include "clang/AST/ASTContextAllocate.h"
 #include "clang/AST/DeclAccessPair.h"
@@ -2294,8 +2295,8 @@ class FunctionDecl : public DeclaratorDecl,
 
   /// Whether this virtual function is pure, i.e. makes the containing class
   /// abstract.
-  bool isPure() const { return FunctionDeclBits.IsPure; }
-  void setPure(bool P = true);
+  bool isPureVirtual() const { return FunctionDeclBits.IsPureVirtual; }
+  void setIsPureVirtual(bool P = true);
 
   /// Whether this templated function will be late parsed.
   bool isLateTemplateParsed() const {
@@ -3252,15 +3253,16 @@ class FieldDecl : public DeclaratorDecl, public Mergeable {
 /// that is defined.  For example, in "enum X {a,b}", each of a/b are
 /// EnumConstantDecl's, X is an instance of EnumDecl, and the type of a/b is a
 /// TagType for the X EnumDecl.
-class EnumConstantDecl : public ValueDecl, public Mergeable {
+class EnumConstantDecl : public ValueDecl,
+                         public Mergeable,
+                         public APIntStorage {
   Stmt *Init; // an integer constant expression
-  llvm::APSInt Val; // The value.
+  bool IsUnsigned;
 
 protected:
-  EnumConstantDecl(DeclContext *DC, SourceLocation L,
+  EnumConstantDecl(const ASTContext &C, DeclContext *DC, SourceLocation L,
                    IdentifierInfo *Id, QualType T, Expr *E,
-                   const llvm::APSInt &V)
-    : ValueDecl(EnumConstant, DC, L, Id, T), Init((Stmt*)E), Val(V) {}
+                   const llvm::APSInt &V);
 
 public:
   friend class StmtIteratorBase;
@@ -3273,10 +3275,15 @@ class EnumConstantDecl : public ValueDecl, public Mergeable {
 
   const Expr *getInitExpr() const { return (const Expr*) Init; }
   Expr *getInitExpr() { return (Expr*) Init; }
-  const llvm::APSInt &getInitVal() const { return Val; }
+  llvm::APSInt getInitVal() const {
+    return llvm::APSInt(getValue(), IsUnsigned);
+  }
 
   void setInitExpr(Expr *E) { Init = (Stmt*) E; }
-  void setInitVal(const llvm::APSInt &V) { Val = V; }
+  void setInitVal(const ASTContext &C, const llvm::APSInt &V) {
+    setValue(C, V);
+    IsUnsigned = V.isUnsigned();
+  }
 
   SourceRange getSourceRange() const override LLVM_READONLY;
 
diff --git a/clang/include/clang/AST/DeclBase.h b/clang/include/clang/AST/DeclBase.h
index 5b1038582bc67..9a4736019d1b1 100644
--- a/clang/include/clang/AST/DeclBase.h
+++ b/clang/include/clang/AST/DeclBase.h
@@ -548,17 +548,18 @@ class alignas(8) Decl {
     return hasAttrs() ? getAttrs().end() : nullptr;
   }
 
-  template 
-  void dropAttr() {
+  template  void dropAttrs() {
     if (!HasAttrs) return;
 
     AttrVec &Vec = getAttrs();
-    llvm::erase_if(Vec, [](Attr *A) { return isa(A); });
+    llvm::erase_if(Vec, [](Attr *A) { return isa(A); });
 
     if (Vec.empty())
       HasAttrs = false;
   }
 
+  template  void dropAttr() { dropAttrs(); }
+
   template 
   llvm::iterator_range> specific_attrs() const {
     return llvm::make_range(specific_attr_begin(), specific_attr_end());
@@ -1707,7 +1708,7 @@ class DeclContext {
     LLVM_PREFERRED_TYPE(bool)
     uint64_t IsVirtualAsWritten : 1;
     LLVM_PREFERRED_TYPE(bool)
-    uint64_t IsPure : 1;
+    uint64_t IsPureVirtual : 1;
     LLVM_PREFERRED_TYPE(bool)
     uint64_t HasInheritedPrototype : 1;
     LLVM_PREFERRED_TYPE(bool)
diff --git a/clang/include/clang/AST/DeclCXX.h b/clang/include/clang/AST/DeclCXX.h
index 648f5f9464087..9cebaff63bb0d 100644
--- a/clang/include/clang/AST/DeclCXX.h
+++ b/clang/include/clang/AST/DeclCXX.h
@@ -266,7 +266,7 @@ class CXXRecordDecl : public RecordDecl {
   friend class LambdaExpr;
   friend class ODRDiagsEmitter;
 
-  friend void FunctionDecl::setPure(bool);
+  friend void FunctionDecl::setIsPureVirtual(bool);
   friend void TagDecl::startDefinition();
 
   /// Values used in DefinitionData fields to represent special members.
@@ -1439,31 +1439,20 @@ class CXXRecordDecl : public RecordDecl {
 
   /// Determine whether this class is a literal type.
   ///
-  /// C++11 [basic.types]p10:
+  /// C++20 [basic.types]p10:
   ///   A class type that has all the following properties:
-  ///     - it has a trivial destructor
-  ///     - every constructor call and full-expression in the
-  ///       brace-or-equal-intializers for non-static data members (if any) is
-  ///       a constant expression.
-  ///     - it is an aggregate type or has at least one constexpr constructor
-  ///       or constructor template that is not a copy or move constructor, and
-  ///     - all of its non-static data members and base classes are of literal
-  ///       types
-  ///
-  /// We resolve DR1361 by ignoring the second bullet. We resolve DR1452 by
-  /// treating types with trivial default constructors as literal types.
-  ///
-  /// Only in C++17 and beyond, are lambdas literal types.
-  bool isLiteral() const {
-    const LangOptions &LangOpts = getLangOpts();
-    return (LangOpts.CPlusPlus20 ? hasConstexprDestructor()
-                                          : hasTrivialDestructor()) &&
-           (!isLambda() || LangOpts.CPlusPlus17) &&
-           !hasNonLiteralTypeFieldsOrBases() &&
-           (isAggregate() || isLambda() ||
-            hasConstexprNonCopyMoveConstructor() ||
-            hasTrivialDefaultConstructor());
-  }
+  ///     - it has a constexpr destructor
+  ///     - all of its non-static non-variant data members and base classes
+  ///       are of non-volatile literal types, and it:
+  ///        - is a closure type
+  ///        - is an aggregate union type that has either no variant members
+  ///          or at least one variant member of non-volatile literal type
+  ///        - is a non-union aggregate type for which each of its anonymous
+  ///          union members satisfies the above requirements for an aggregate
+  ///          union type, or
+  ///        - has at least one constexpr constructor or constructor template
+  ///          that is not a copy or move constructor.
+  bool isLiteral() const;
 
   /// Determine whether this is a structural type.
   bool isStructural() const {
@@ -2121,7 +2110,7 @@ class CXXMethodDecl : public FunctionDecl {
 
     // Member function is virtual if it is marked explicitly so, or if it is
     // declared in __interface -- then it is automatically pure virtual.
-    if (CD->isVirtualAsWritten() || CD->isPure())
+    if (CD->isVirtualAsWritten() || CD->isPureVirtual())
       return true;
 
     return CD->size_overridden_methods() != 0;
diff --git a/clang/include/clang/AST/Expr.h b/clang/include/clang/AST/Expr.h
index 3cc4e414f9dbe..2cca5b35478ce 100644
--- a/clang/include/clang/AST/Expr.h
+++ b/clang/include/clang/AST/Expr.h
@@ -13,6 +13,7 @@
 #ifndef LLVM_CLANG_AST_EXPR_H
 #define LLVM_CLANG_AST_EXPR_H
 
+#include "clang/AST/APNumericStorage.h"
 #include "clang/AST/APValue.h"
 #include "clang/AST/ASTVector.h"
 #include "clang/AST/ComputeDependence.h"
@@ -1479,57 +1480,6 @@ class DeclRefExpr final
   }
 };
 
-/// Used by IntegerLiteral/FloatingLiteral to store the numeric without
-/// leaking memory.
-///
-/// For large floats/integers, APFloat/APInt will allocate memory from the heap
-/// to represent these numbers.  Unfortunately, when we use a BumpPtrAllocator
-/// to allocate IntegerLiteral/FloatingLiteral nodes the memory associated with
-/// the APFloat/APInt values will never get freed. APNumericStorage uses
-/// ASTContext's allocator for memory allocation.
-class APNumericStorage {
-  union {
-    uint64_t VAL;    ///< Used to store the <= 64 bits integer value.
-    uint64_t *pVal;  ///< Used to store the >64 bits integer value.
-  };
-  unsigned BitWidth;
-
-  bool hasAllocation() const { return llvm::APInt::getNumWords(BitWidth) > 1; }
-
-  APNumericStorage(const APNumericStorage &) = delete;
-  void operator=(const APNumericStorage &) = delete;
-
-protected:
-  APNumericStorage() : VAL(0), BitWidth(0) { }
-
-  llvm::APInt getIntValue() const {
-    unsigned NumWords = llvm::APInt::getNumWords(BitWidth);
-    if (NumWords > 1)
-      return llvm::APInt(BitWidth, NumWords, pVal);
-    else
-      return llvm::APInt(BitWidth, VAL);
-  }
-  void setIntValue(const ASTContext &C, const llvm::APInt &Val);
-};
-
-class APIntStorage : private APNumericStorage {
-public:
-  llvm::APInt getValue() const { return getIntValue(); }
-  void setValue(const ASTContext &C, const llvm::APInt &Val) {
-    setIntValue(C, Val);
-  }
-};
-
-class APFloatStorage : private APNumericStorage {
-public:
-  llvm::APFloat getValue(const llvm::fltSemantics &Semantics) const {
-    return llvm::APFloat(Semantics, getIntValue());
-  }
-  void setValue(const ASTContext &C, const llvm::APFloat &Val) {
-    setIntValue(C, Val.bitcastToAPInt());
-  }
-};
-
 class IntegerLiteral : public Expr, public APIntStorage {
   SourceLocation Loc;
 
diff --git a/clang/include/clang/AST/Type.h b/clang/include/clang/AST/Type.h
index afcc52d37016a..63ab6e259f7ee 100644
--- a/clang/include/clang/AST/Type.h
+++ b/clang/include/clang/AST/Type.h
@@ -4039,6 +4039,22 @@ class FunctionType : public Type {
   /// because TrailingObjects cannot handle repeated types.
   struct ExceptionType { QualType Type; };
 
+  /// A simple holder for various uncommon bits which do not fit in
+  /// FunctionTypeBitfields. Aligned to alignof(void *) to maintain the
+  /// alignment of subsequent objects in TrailingObjects.
+  struct alignas(void *) FunctionTypeExtraBitfields {
+    /// The number of types in the exception specification.
+    /// A whole unsigned is not needed here and according to
+    /// [implimits] 8 bits would be enough here.
+    unsigned NumExceptionType : 10;
+
+    LLVM_PREFERRED_TYPE(bool)
+    unsigned HasArmTypeAttributes : 1;
+
+    FunctionTypeExtraBitfields()
+        : NumExceptionType(0), HasArmTypeAttributes(false) {}
+  };
+
   /// The AArch64 SME ACLE (Arm C/C++ Language Extensions) define a number
   /// of function type attributes that can be set on function types, including
   /// function pointers.
@@ -4046,26 +4062,37 @@ class FunctionType : public Type {
     SME_NormalFunction = 0,
     SME_PStateSMEnabledMask = 1 << 0,
     SME_PStateSMCompatibleMask = 1 << 1,
-    SME_PStateZASharedMask = 1 << 2,
-    SME_PStateZAPreservedMask = 1 << 3,
-    SME_AttributeMask = 0b111'111 // We only support maximum 6 bits because of the
-                                  // bitmask in FunctionTypeExtraBitfields.
+
+    // Describes the value of the state using ArmStateValue.
+    SME_ZAShift = 2,
+    SME_ZAMask = 0b111 << SME_ZAShift,
+
+    SME_AttributeMask = 0b111'111 // We only support maximum 6 bits because of
+                                  // the bitmask in FunctionTypeArmAttributes
+                                  // and ExtProtoInfo.
   };
 
-  /// A simple holder for various uncommon bits which do not fit in
-  /// FunctionTypeBitfields. Aligned to alignof(void *) to maintain the
-  /// alignment of subsequent objects in TrailingObjects.
-  struct alignas(void *) FunctionTypeExtraBitfields {
-    /// The number of types in the exception specification.
-    /// A whole unsigned is not needed here and according to
-    /// [implimits] 8 bits would be enough here.
-    unsigned NumExceptionType : 10;
+  enum ArmStateValue : unsigned {
+    ARM_None = 0,
+    ARM_Preserves = 1,
+    ARM_In = 2,
+    ARM_Out = 3,
+    ARM_InOut = 4,
+  };
+
+  static ArmStateValue getArmZAState(unsigned AttrBits) {
+    return (ArmStateValue)((AttrBits & SME_ZAMask) >> SME_ZAShift);
+  }
 
+  /// A holder for Arm type attributes as described in the Arm C/C++
+  /// Language extensions which are not particularly common to all
+  /// types and therefore accounted separately from FunctionTypeBitfields.
+  struct alignas(void *) FunctionTypeArmAttributes {
     /// Any AArch64 SME ACLE type attributes that need to be propagated
     /// on declarations and function pointers.
     unsigned AArch64SMEAttributes : 6;
-    FunctionTypeExtraBitfields()
-        : NumExceptionType(0), AArch64SMEAttributes(SME_NormalFunction) {}
+
+    FunctionTypeArmAttributes() : AArch64SMEAttributes(SME_NormalFunction) {}
   };
 
 protected:
@@ -4164,7 +4191,8 @@ class FunctionProtoType final
       public llvm::FoldingSetNode,
       private llvm::TrailingObjects<
           FunctionProtoType, QualType, SourceLocation,
-          FunctionType::FunctionTypeExtraBitfields, FunctionType::ExceptionType,
+          FunctionType::FunctionTypeExtraBitfields,
+          FunctionType::FunctionTypeArmAttributes, FunctionType::ExceptionType,
           Expr *, FunctionDecl *, FunctionType::ExtParameterInfo, Qualifiers> {
   friend class ASTContext; // ASTContext creates these.
   friend TrailingObjects;
@@ -4271,7 +4299,11 @@ class FunctionProtoType final
 
     bool requiresFunctionProtoTypeExtraBitfields() const {
       return ExceptionSpec.Type == EST_Dynamic ||
-             AArch64SMEAttributes != SME_NormalFunction;
+             requiresFunctionProtoTypeArmAttributes();
+    }
+
+    bool requiresFunctionProtoTypeArmAttributes() const {
+      return AArch64SMEAttributes != SME_NormalFunction;
     }
 
     void setArmSMEAttribute(AArch64SMETypeAttributes Kind, bool Enable = true) {
@@ -4291,6 +4323,10 @@ class FunctionProtoType final
     return isVariadic();
   }
 
+  unsigned numTrailingObjects(OverloadToken) const {
+    return hasArmTypeAttributes();
+  }
+
   unsigned numTrailingObjects(OverloadToken) const {
     return hasExtraBitfields();
   }
@@ -4379,6 +4415,12 @@ class FunctionProtoType final
 
   }
 
+  bool hasArmTypeAttributes() const {
+    return FunctionTypeBits.HasExtraBitfields &&
+           getTrailingObjects()
+               ->HasArmTypeAttributes;
+  }
+
   bool hasExtQualifiers() const {
     return FunctionTypeBits.HasExtQuals;
   }
@@ -4590,9 +4632,9 @@ class FunctionProtoType final
   /// Return a bitmask describing the SME attributes on the function type, see
   /// AArch64SMETypeAttributes for their values.
   unsigned getAArch64SMEAttributes() const {
-    if (!hasExtraBitfields())
+    if (!hasArmTypeAttributes())
       return SME_NormalFunction;
-    return getTrailingObjects()
+    return getTrailingObjects()
         ->AArch64SMEAttributes;
   }
 
diff --git a/clang/include/clang/ASTMatchers/ASTMatchers.h b/clang/include/clang/ASTMatchers/ASTMatchers.h
index 82a26356c58f5..dc1f49525a004 100644
--- a/clang/include/clang/ASTMatchers/ASTMatchers.h
+++ b/clang/include/clang/ASTMatchers/ASTMatchers.h
@@ -2062,6 +2062,18 @@ extern const internal::VariadicDynCastAllOfMatcher
 extern const internal::VariadicDynCastAllOfMatcher
     cxxOperatorCallExpr;
 
+/// Matches C++17 fold expressions.
+///
+/// Example matches `(0 + ... + args)`:
+/// \code
+///   template 
+///   auto sum(Args... args) {
+///       return (0 + ... + args);
+///   }
+/// \endcode
+extern const internal::VariadicDynCastAllOfMatcher
+    cxxFoldExpr;
+
 /// Matches rewritten binary operators
 ///
 /// Example matches use of "<":
@@ -3881,7 +3893,7 @@ AST_MATCHER_P(ObjCMessageExpr, numSelectorArgs, unsigned, N) {
   return Node.getSelector().getNumArgs() == N;
 }
 
-/// Matches if the call expression's callee expression matches.
+/// Matches if the call or fold expression's callee expression matches.
 ///
 /// Given
 /// \code
@@ -3893,13 +3905,32 @@ AST_MATCHER_P(ObjCMessageExpr, numSelectorArgs, unsigned, N) {
 /// with callee(...)
 ///   matching this->x, x, y.x, f respectively
 ///
+/// Given
+/// \code
+///   template 
+///   auto sum(Args... args) {
+///       return (0 + ... + args);
+///   }
+///
+///   template 
+///   auto multiply(Args... args) {
+///       return (args * ... * 1);
+///   }
+/// \endcode
+/// cxxFoldExpr(callee(expr()))
+///   matches (args * ... * 1)
+/// with callee(...)
+///   matching *
+///
 /// Note: Callee cannot take the more general internal::Matcher
 /// because this introduces ambiguous overloads with calls to Callee taking a
 /// internal::Matcher, as the matcher hierarchy is purely
 /// implemented in terms of implicit casts.
-AST_MATCHER_P(CallExpr, callee, internal::Matcher,
-              InnerMatcher) {
-  const Expr *ExprNode = Node.getCallee();
+AST_POLYMORPHIC_MATCHER_P_OVERLOAD(callee,
+                                   AST_POLYMORPHIC_SUPPORTED_TYPES(CallExpr,
+                                                                   CXXFoldExpr),
+                                   internal::Matcher, InnerMatcher, 0) {
+  const auto *ExprNode = Node.getCallee();
   return (ExprNode != nullptr &&
           InnerMatcher.matches(*ExprNode, Finder, Builder));
 }
@@ -4532,6 +4563,121 @@ AST_POLYMORPHIC_MATCHER_P2(hasArgument,
   return InnerMatcher.matches(*Arg->IgnoreParenImpCasts(), Finder, Builder);
 }
 
+/// Matches the operand that does not contain the parameter pack.
+///
+/// Example matches `(0 + ... + args)` and `(args * ... * 1)`
+///     (matcher = cxxFoldExpr(hasFoldInit(expr())))
+///   with hasFoldInit(...)
+///     matching `0` and `1` respectively
+/// \code
+///   template 
+///   auto sum(Args... args) {
+///       return (0 + ... + args);
+///   }
+///
+///   template 
+///   auto multiply(Args... args) {
+///       return (args * ... * 1);
+///   }
+/// \endcode
+AST_MATCHER_P(CXXFoldExpr, hasFoldInit, ast_matchers::internal::Matcher,
+              InnerMacher) {
+  const auto *const Init = Node.getInit();
+  return Init && InnerMacher.matches(*Init, Finder, Builder);
+}
+
+/// Matches the operand that contains the parameter pack.
+///
+/// Example matches `(0 + ... + args)`
+///     (matcher = cxxFoldExpr(hasPattern(expr())))
+///   with hasPattern(...)
+///     matching `args`
+/// \code
+///   template 
+///   auto sum(Args... args) {
+///       return (0 + ... + args);
+///   }
+///
+///   template 
+///   auto multiply(Args... args) {
+///       return (args * ... * 1);
+///   }
+/// \endcode
+AST_MATCHER_P(CXXFoldExpr, hasPattern, ast_matchers::internal::Matcher,
+              InnerMacher) {
+  const Expr *const Pattern = Node.getPattern();
+  return Pattern && InnerMacher.matches(*Pattern, Finder, Builder);
+}
+
+/// Matches right-folding fold expressions.
+///
+/// Example matches `(args * ... * 1)`
+///     (matcher = cxxFoldExpr(isRightFold()))
+/// \code
+///   template 
+///   auto sum(Args... args) {
+///       return (0 + ... + args);
+///   }
+///
+///   template 
+///   auto multiply(Args... args) {
+///       return (args * ... * 1);
+///   }
+/// \endcode
+AST_MATCHER(CXXFoldExpr, isRightFold) { return Node.isRightFold(); }
+
+/// Matches left-folding fold expressions.
+///
+/// Example matches `(0 + ... + args)`
+///     (matcher = cxxFoldExpr(isLeftFold()))
+/// \code
+///   template 
+///   auto sum(Args... args) {
+///       return (0 + ... + args);
+///   }
+///
+///   template 
+///   auto multiply(Args... args) {
+///       return (args * ... * 1);
+///   }
+/// \endcode
+AST_MATCHER(CXXFoldExpr, isLeftFold) { return Node.isLeftFold(); }
+
+/// Matches unary fold expressions, i.e. fold expressions without an
+/// initializer.
+///
+/// Example matches `(args * ...)`
+///     (matcher = cxxFoldExpr(isUnaryFold()))
+/// \code
+///   template 
+///   auto sum(Args... args) {
+///       return (0 + ... + args);
+///   }
+///
+///   template 
+///   auto multiply(Args... args) {
+///       return (args * ...);
+///   }
+/// \endcode
+AST_MATCHER(CXXFoldExpr, isUnaryFold) { return Node.getInit() == nullptr; }
+
+/// Matches binary fold expressions, i.e. fold expressions with an initializer.
+///
+/// Example matches `(0 + ... + args)`
+///     (matcher = cxxFoldExpr(isBinaryFold()))
+/// \code
+///   template 
+///   auto sum(Args... args) {
+///       return (0 + ... + args);
+///   }
+///
+///   template 
+///   auto multiply(Args... args) {
+///       return (args * ...);
+///   }
+/// \endcode
+AST_MATCHER(CXXFoldExpr, isBinaryFold) { return Node.getInit() != nullptr; }
+
 /// Matches the n'th item of an initializer list expression.
 ///
 /// Example matches y.
@@ -5709,17 +5855,27 @@ AST_POLYMORPHIC_MATCHER_P_OVERLOAD(equals,
     .matchesNode(Node);
 }
 
-/// Matches the operator Name of operator expressions (binary or
-/// unary).
+/// Matches the operator Name of operator expressions and fold expressions
+/// (binary or unary).
 ///
 /// Example matches a || b (matcher = binaryOperator(hasOperatorName("||")))
 /// \code
 ///   !(a || b)
 /// \endcode
+///
+/// Example matches `(0 + ... + args)`
+///     (matcher = cxxFoldExpr(hasOperatorName("+")))
+/// \code
+///   template 
+///   auto sum(Args... args) {
+///       return (0 + ... + args);
+///   }
+/// \endcode
 AST_POLYMORPHIC_MATCHER_P(
     hasOperatorName,
     AST_POLYMORPHIC_SUPPORTED_TYPES(BinaryOperator, CXXOperatorCallExpr,
-                                    CXXRewrittenBinaryOperator, UnaryOperator),
+                                    CXXRewrittenBinaryOperator, CXXFoldExpr,
+                                    UnaryOperator),
     std::string, Name) {
   if (std::optional OpName = internal::getOpName(Node))
     return *OpName == Name;
@@ -5789,11 +5945,12 @@ AST_POLYMORPHIC_MATCHER(
 /// \code
 ///   a || b
 /// \endcode
-AST_POLYMORPHIC_MATCHER_P(hasLHS,
-                          AST_POLYMORPHIC_SUPPORTED_TYPES(
-                              BinaryOperator, CXXOperatorCallExpr,
-                              CXXRewrittenBinaryOperator, ArraySubscriptExpr),
-                          internal::Matcher, InnerMatcher) {
+AST_POLYMORPHIC_MATCHER_P(
+    hasLHS,
+    AST_POLYMORPHIC_SUPPORTED_TYPES(BinaryOperator, CXXOperatorCallExpr,
+                                    CXXRewrittenBinaryOperator,
+                                    ArraySubscriptExpr, CXXFoldExpr),
+    internal::Matcher, InnerMatcher) {
   const Expr *LeftHandSide = internal::getLHS(Node);
   return (LeftHandSide != nullptr &&
           InnerMatcher.matches(*LeftHandSide, Finder, Builder));
@@ -5805,29 +5962,31 @@ AST_POLYMORPHIC_MATCHER_P(hasLHS,
 /// \code
 ///   a || b
 /// \endcode
-AST_POLYMORPHIC_MATCHER_P(hasRHS,
-                          AST_POLYMORPHIC_SUPPORTED_TYPES(
-                              BinaryOperator, CXXOperatorCallExpr,
-                              CXXRewrittenBinaryOperator, ArraySubscriptExpr),
-                          internal::Matcher, InnerMatcher) {
+AST_POLYMORPHIC_MATCHER_P(
+    hasRHS,
+    AST_POLYMORPHIC_SUPPORTED_TYPES(BinaryOperator, CXXOperatorCallExpr,
+                                    CXXRewrittenBinaryOperator,
+                                    ArraySubscriptExpr, CXXFoldExpr),
+    internal::Matcher, InnerMatcher) {
   const Expr *RightHandSide = internal::getRHS(Node);
   return (RightHandSide != nullptr &&
           InnerMatcher.matches(*RightHandSide, Finder, Builder));
 }
 
 /// Matches if either the left hand side or the right hand side of a
-/// binary operator matches.
+/// binary operator or fold expression matches.
 AST_POLYMORPHIC_MATCHER_P(
     hasEitherOperand,
     AST_POLYMORPHIC_SUPPORTED_TYPES(BinaryOperator, CXXOperatorCallExpr,
-                                    CXXRewrittenBinaryOperator),
+                                    CXXFoldExpr, CXXRewrittenBinaryOperator),
     internal::Matcher, InnerMatcher) {
   return internal::VariadicDynCastAllOfMatcher()(
              anyOf(hasLHS(InnerMatcher), hasRHS(InnerMatcher)))
       .matches(Node, Finder, Builder);
 }
 
-/// Matches if both matchers match with opposite sides of the binary operator.
+/// Matches if both matchers match with opposite sides of the binary operator
+/// or fold expression.
 ///
 /// Example matcher = binaryOperator(hasOperands(integerLiteral(equals(1),
 ///                                              integerLiteral(equals(2)))
@@ -5840,7 +5999,7 @@ AST_POLYMORPHIC_MATCHER_P(
 AST_POLYMORPHIC_MATCHER_P2(
     hasOperands,
     AST_POLYMORPHIC_SUPPORTED_TYPES(BinaryOperator, CXXOperatorCallExpr,
-                                    CXXRewrittenBinaryOperator),
+                                    CXXFoldExpr, CXXRewrittenBinaryOperator),
     internal::Matcher, Matcher1, internal::Matcher, Matcher2) {
   return internal::VariadicDynCastAllOfMatcher()(
              anyOf(allOf(hasLHS(Matcher1), hasRHS(Matcher2)),
@@ -6192,9 +6351,7 @@ AST_POLYMORPHIC_MATCHER(isFinal,
 ///   };
 /// \endcode
 ///   matches A::x
-AST_MATCHER(CXXMethodDecl, isPure) {
-  return Node.isPure();
-}
+AST_MATCHER(CXXMethodDecl, isPure) { return Node.isPureVirtual(); }
 
 /// Matches if the given method declaration is const.
 ///
diff --git a/clang/include/clang/ASTMatchers/ASTMatchersInternal.h b/clang/include/clang/ASTMatchers/ASTMatchersInternal.h
index 7136d0d2fe084..47d912c73dd7e 100644
--- a/clang/include/clang/ASTMatchers/ASTMatchersInternal.h
+++ b/clang/include/clang/ASTMatchers/ASTMatchersInternal.h
@@ -2195,6 +2195,9 @@ inline std::optional getOpName(const CXXOperatorCallExpr &Node) {
   }
   return BinaryOperator::getOpcodeStr(*optBinaryOpcode);
 }
+inline StringRef getOpName(const CXXFoldExpr &Node) {
+  return BinaryOperator::getOpcodeStr(Node.getOperator());
+}
 
 /// Matches overloaded operators with a specific name.
 ///
diff --git a/clang/include/clang/Analysis/FlowSensitive/ControlFlowContext.h b/clang/include/clang/Analysis/FlowSensitive/ControlFlowContext.h
index 768387a121b92..405e93287a05d 100644
--- a/clang/include/clang/Analysis/FlowSensitive/ControlFlowContext.h
+++ b/clang/include/clang/Analysis/FlowSensitive/ControlFlowContext.h
@@ -32,7 +32,8 @@ namespace dataflow {
 class ControlFlowContext {
 public:
   /// Builds a ControlFlowContext from a `FunctionDecl`.
-  /// `Func.hasBody()` must be true, and `Func.isTemplated()` must be false.
+  /// `Func.doesThisDeclarationHaveABody()` must be true, and
+  /// `Func.isTemplated()` must be false.
   static llvm::Expected build(const FunctionDecl &Func);
 
   /// Builds a ControlFlowContext from an AST node. `D` is the function in which
diff --git a/clang/include/clang/Analysis/FlowSensitive/DataflowEnvironment.h b/clang/include/clang/Analysis/FlowSensitive/DataflowEnvironment.h
index e8c27d6c12038..1543f900e401d 100644
--- a/clang/include/clang/Analysis/FlowSensitive/DataflowEnvironment.h
+++ b/clang/include/clang/Analysis/FlowSensitive/DataflowEnvironment.h
@@ -172,7 +172,8 @@ class Environment {
   ///
   /// Requirements:
   ///
-  ///  The function must have a body.
+  ///  The function must have a body, i.e.
+  ///  `FunctionDecl::doesThisDecalarationHaveABody()` must be true.
   void initialize();
 
   /// Returns a new environment that is a copy of this one.
diff --git a/clang/include/clang/Analysis/FlowSensitive/Formula.h b/clang/include/clang/Analysis/FlowSensitive/Formula.h
index 982e400c1deff..0e6352403a832 100644
--- a/clang/include/clang/Analysis/FlowSensitive/Formula.h
+++ b/clang/include/clang/Analysis/FlowSensitive/Formula.h
@@ -75,6 +75,10 @@ class alignas(const Formula *) Formula {
     return static_cast(Value);
   }
 
+  bool isLiteral(bool b) const {
+    return kind() == Literal && static_cast(Value) == b;
+  }
+
   ArrayRef operands() const {
     return ArrayRef(reinterpret_cast(this + 1),
                     numOperands(kind()));
diff --git a/clang/include/clang/Basic/Attr.td b/clang/include/clang/Basic/Attr.td
index 7fd719c354658..17d84a5d947bc 100644
--- a/clang/include/clang/Basic/Attr.td
+++ b/clang/include/clang/Basic/Attr.td
@@ -310,6 +310,9 @@ class VariadicEnumArgument values,
   bit IsExternalType = isExternalType;
 }
 
+// Represents an attribute wrapped by another attribute.
+class WrappedAttr : Argument;
+
 // This handles one spelling of an attribute.
 class Spelling {
   string Name = name;
@@ -3273,7 +3276,7 @@ def ObjCBridgeRelated : InheritableAttr {
 def NSErrorDomain : InheritableAttr {
   let Spellings = [GNU<"ns_error_domain">];
   let Subjects = SubjectList<[Enum], ErrorDiag>;
-  let Args = [DeclArgument];
+  let Args = [IdentifierArgument<"ErrorDomain">];
   let Documentation = [NSErrorDomainDocs];
 }
 
@@ -3521,16 +3524,45 @@ def ArmStreamingCompatible : TypeAttr, TargetSpecificAttr {
   let Documentation = [ArmSmeStreamingCompatibleDocs];
 }
 
-def ArmSharedZA : TypeAttr, TargetSpecificAttr {
-  let Spellings = [RegularKeyword<"__arm_shared_za">];
+def ArmNew : InheritableAttr, TargetSpecificAttr {
+  let Spellings = [RegularKeyword<"__arm_new">];
+  let Args = [VariadicStringArgument<"NewArgs">];
+  let Subjects = SubjectList<[Function], ErrorDiag>;
+  let Documentation = [ArmNewDocs];
+
+  let AdditionalMembers = [{
+    bool isNewZA() const {
+      return llvm::is_contained(newArgs(), "za");
+    }
+  }];
+}
+
+def ArmIn : TypeAttr, TargetSpecificAttr {
+  let Spellings = [RegularKeyword<"__arm_in">];
+  let Args = [VariadicStringArgument<"InArgs">];
+  let Subjects = SubjectList<[HasFunctionProto], ErrorDiag>;
+  let Documentation = [ArmInDocs];
+}
+
+def ArmOut : TypeAttr, TargetSpecificAttr {
+  let Spellings = [RegularKeyword<"__arm_out">];
+  let Args = [VariadicStringArgument<"OutArgs">];
   let Subjects = SubjectList<[HasFunctionProto], ErrorDiag>;
-  let Documentation = [ArmSmeSharedZADocs];
+  let Documentation = [ArmOutDocs];
 }
 
-def ArmPreservesZA : TypeAttr, TargetSpecificAttr {
-  let Spellings = [RegularKeyword<"__arm_preserves_za">];
+def ArmInOut : TypeAttr, TargetSpecificAttr {
+  let Spellings = [RegularKeyword<"__arm_inout">];
+  let Args = [VariadicStringArgument<"InOutArgs">];
   let Subjects = SubjectList<[HasFunctionProto], ErrorDiag>;
-  let Documentation = [ArmSmePreservesZADocs];
+  let Documentation = [ArmInOutDocs];
+}
+
+def ArmPreserves : TypeAttr, TargetSpecificAttr {
+  let Spellings = [RegularKeyword<"__arm_preserves">];
+  let Args = [VariadicStringArgument<"PreserveArgs">];
+  let Subjects = SubjectList<[HasFunctionProto], ErrorDiag>;
+  let Documentation = [ArmPreservesDocs];
 }
 
 def ArmLocallyStreaming : InheritableAttr, TargetSpecificAttr {
@@ -3539,14 +3571,6 @@ def ArmLocallyStreaming : InheritableAttr, TargetSpecificAttr {
   let Documentation = [ArmSmeLocallyStreamingDocs];
 }
 
-def ArmNewZA : InheritableAttr, TargetSpecificAttr {
-  let Spellings = [RegularKeyword<"__arm_new_za">];
-  let Subjects = SubjectList<[Function], ErrorDiag>;
-  let Documentation = [ArmSmeNewZADocs];
-}
-def : MutualExclusions<[ArmNewZA, ArmSharedZA]>;
-def : MutualExclusions<[ArmNewZA, ArmPreservesZA]>;
-
 
 def Pure : InheritableAttr {
   let Spellings = [GCC<"pure">];
@@ -3609,6 +3633,22 @@ def SwiftError : InheritableAttr {
   let Documentation = [SwiftErrorDocs];
 }
 
+def SwiftImportAsNonGeneric : InheritableAttr {
+  // This attribute has no spellings as it is only ever created implicitly
+  // from API notes.
+  let Spellings = [];
+  let SemaHandler = 0;
+  let Documentation = [InternalOnly];
+}
+
+def SwiftImportPropertyAsAccessors : InheritableAttr {
+  // This attribute has no spellings as it is only ever created implicitly
+  // from API notes.
+  let Spellings = [];
+  let SemaHandler = 0;
+  let Documentation = [InternalOnly];
+}
+
 def SwiftName : InheritableAttr {
   let Spellings = [GNU<"swift_name">];
   let Args = [StringArgument<"Name">];
@@ -3630,6 +3670,31 @@ def SwiftPrivate : InheritableAttr {
   let SimpleHandler = 1;
 }
 
+def SwiftVersionedAddition : Attr {
+  // This attribute has no spellings as it is only ever created implicitly
+  // from API notes.
+  let Spellings = [];
+  let Args = [VersionArgument<"Version">, WrappedAttr<"AdditionalAttr">,
+              BoolArgument<"IsReplacedByActive">];
+  let SemaHandler = 0;
+  let Documentation = [InternalOnly];
+}
+
+def SwiftVersionedRemoval : Attr {
+  // This attribute has no spellings as it is only ever created implicitly
+  // from API notes.
+  let Spellings = [];
+  let Args = [VersionArgument<"Version">, UnsignedArgument<"RawKind">,
+              BoolArgument<"IsReplacedByActive">];
+  let SemaHandler = 0;
+  let Documentation = [InternalOnly];
+  let AdditionalMembers = [{
+    attr::Kind getAttrKindToRemove() const {
+      return static_cast(getRawKind());
+    }
+  }];
+}
+
 def NoDeref : TypeAttr {
   let Spellings = [Clang<"noderef">];
   let Documentation = [NoDerefDocs];
diff --git a/clang/include/clang/Basic/AttrDocs.td b/clang/include/clang/Basic/AttrDocs.td
index 294cfe8469f65..abd4b3df1dfd1 100644
--- a/clang/include/clang/Basic/AttrDocs.td
+++ b/clang/include/clang/Basic/AttrDocs.td
@@ -8565,30 +8565,73 @@ without changing modes.
   }];
 }
 
-def ArmSmeSharedZADocs : Documentation {
+def ArmInDocs : Documentation {
   let Category = DocCatArmSmeAttributes;
   let Content = [{
-The ``__arm_shared_za`` keyword applies to prototyped function types and specifies
-that the function shares SME's matrix storage (ZA) with its caller.  This
-means that:
+The ``__arm_in`` keyword applies to prototyped function types and specifies
+that the function shares a given state S with its caller.  For ``__arm_in``, the
+function takes the state S as input and returns with the state S unchanged.
 
-* the function requires that the processor implements the Scalable Matrix
-  Extension (SME).
+The attribute takes string arguments to instruct the compiler which state
+is shared.  The supported states for S are:
 
-* the function enters with ZA in an active state.
+* ``"za"`` for Matrix Storage (requires SME)
 
-* the function returns with ZA in an active state.
+The attributes ``__arm_in(S)``, ``__arm_out(S)``, ``__arm_inout(S)`` and
+``__arm_preserves(S)`` are all mutually exclusive for the same state S.
   }];
 }
 
-def ArmSmePreservesZADocs : Documentation {
+def ArmOutDocs : Documentation {
   let Category = DocCatArmSmeAttributes;
   let Content = [{
-The ``__arm_preserves_za`` keyword applies to prototyped function types and
-specifies that the function does not modify ZA state.
+The ``__arm_out`` keyword applies to prototyped function types and specifies
+that the function shares a given state S with its caller.  For ``__arm_out``,
+the function ignores the incoming state for S and returns new state for S.
+
+The attribute takes string arguments to instruct the compiler which state
+is shared.  The supported states for S are:
+
+* ``"za"`` for Matrix Storage (requires SME)
+
+The attributes ``__arm_in(S)``, ``__arm_out(S)``, ``__arm_inout(S)`` and
+``__arm_preserves(S)`` are all mutually exclusive for the same state S.
   }];
 }
 
+def ArmInOutDocs : Documentation {
+  let Category = DocCatArmSmeAttributes;
+  let Content = [{
+The ``__arm_inout`` keyword applies to prototyped function types and specifies
+that the function shares a given state S with its caller.  For ``__arm_inout``,
+the function takes the state S as input and returns new state for S.
+
+The attribute takes string arguments to instruct the compiler which state
+is shared.  The supported states for S are:
+
+* ``"za"`` for Matrix Storage (requires SME)
+
+The attributes ``__arm_in(S)``, ``__arm_out(S)``, ``__arm_inout(S)`` and
+``__arm_preserves(S)`` are all mutually exclusive for the same state S.
+  }];
+}
+
+def ArmPreservesDocs : Documentation {
+  let Category = DocCatArmSmeAttributes;
+  let Content = [{
+The ``__arm_preserves`` keyword applies to prototyped function types and
+specifies that the function does not read a given state S and returns
+with state S unchanged.
+
+The attribute takes string arguments to instruct the compiler which state
+is shared.  The supported states for S are:
+
+* ``"za"`` for Matrix Storage (requires SME)
+
+The attributes ``__arm_in(S)``, ``__arm_out(S)``, ``__arm_inout(S)`` and
+``__arm_preserves(S)`` are all mutually exclusive for the same state S.
+  }];
+}
 
 def ArmSmeLocallyStreamingDocs : Documentation {
   let Category = DocCatArmSmeAttributes;
@@ -8611,13 +8654,18 @@ at the end of the function.
   }];
 }
 
-def ArmSmeNewZADocs : Documentation {
+def ArmNewDocs : Documentation {
   let Category = DocCatArmSmeAttributes;
   let Content = [{
-The ``__arm_new_za`` keyword applies to function declarations and specifies
-that the function will be set up with a fresh ZA context.
+The ``__arm_new`` keyword applies to function declarations and specifies
+that the function will create a new scope for state S.
+
+The attribute takes string arguments to instruct the compiler for which state
+to create new scope.  The supported states for S are:
+
+* ``"za"`` for Matrix Storage (requires SME)
 
-This means that:
+For state ``"za"``, this means that:
 
 * the function requires that the target processor implements the Scalable Matrix
   Extension (SME).
@@ -8628,8 +8676,8 @@ This means that:
 
 * the function will disable PSTATE.ZA (by setting it to 0) before returning.
 
-For ``__arm_new_za`` functions Clang will set up the ZA context automatically
-on entry to the function, and disable it before returning. For example, if ZA is
+For ``__arm_new("za")`` functions Clang will set up the ZA context automatically
+on entry to the function and disable it before returning. For example, if ZA is
 in a dormant state Clang will generate the code to commit a lazy-save and set up
 a new ZA state before executing user code.
   }];
diff --git a/clang/include/clang/Basic/AttributeCommonInfo.h b/clang/include/clang/Basic/AttributeCommonInfo.h
index 018b92fdc11f5..d787e4959bfee 100644
--- a/clang/include/clang/Basic/AttributeCommonInfo.h
+++ b/clang/include/clang/Basic/AttributeCommonInfo.h
@@ -255,6 +255,19 @@ class AttributeCommonInfo {
     return SpellingIndex != SpellingNotCalculated;
   }
 };
+
+inline bool doesKeywordAttributeTakeArgs(tok::TokenKind Kind) {
+  switch (Kind) {
+  default:
+    return false;
+#define KEYWORD_ATTRIBUTE(NAME, HASARG)                                        \
+  case tok::kw_##NAME:                                                         \
+    return HASARG;
+#include "clang/Basic/RegularKeywordAttrInfo.inc"
+#undef KEYWORD_ATTRIBUTE
+  }
+}
+
 } // namespace clang
 
 #endif // LLVM_CLANG_BASIC_ATTRIBUTECOMMONINFO_H
diff --git a/clang/include/clang/Basic/BuiltinsAMDGPU.def b/clang/include/clang/Basic/BuiltinsAMDGPU.def
index f66a06f060c21..b6c37a00ef2cb 100644
--- a/clang/include/clang/Basic/BuiltinsAMDGPU.def
+++ b/clang/include/clang/Basic/BuiltinsAMDGPU.def
@@ -255,6 +255,10 @@ TARGET_BUILTIN(__builtin_amdgcn_sudot4, "iIbiIbiiIb", "nc", "dot8-insts")
 TARGET_BUILTIN(__builtin_amdgcn_sdot8, "SiSiSiSiIb", "nc", "dot1-insts")
 TARGET_BUILTIN(__builtin_amdgcn_udot8, "UiUiUiUiIb", "nc", "dot7-insts")
 TARGET_BUILTIN(__builtin_amdgcn_sudot8, "iIbiIbiiIb", "nc", "dot8-insts")
+TARGET_BUILTIN(__builtin_amdgcn_dot4_f32_fp8_bf8, "fUiUif", "nc", "gfx12-insts")
+TARGET_BUILTIN(__builtin_amdgcn_dot4_f32_bf8_fp8, "fUiUif", "nc", "gfx12-insts")
+TARGET_BUILTIN(__builtin_amdgcn_dot4_f32_fp8_fp8, "fUiUif", "nc", "gfx12-insts")
+TARGET_BUILTIN(__builtin_amdgcn_dot4_f32_bf8_bf8, "fUiUif", "nc", "gfx12-insts")
 
 //===----------------------------------------------------------------------===//
 // GFX10+ only builtins.
@@ -397,14 +401,14 @@ TARGET_BUILTIN(__builtin_amdgcn_smfmac_f32_32x32x32_bf8_fp8, "V16fV2iV4iV16fiIiI
 TARGET_BUILTIN(__builtin_amdgcn_smfmac_f32_32x32x32_fp8_bf8, "V16fV2iV4iV16fiIiIi", "nc", "fp8-insts")
 TARGET_BUILTIN(__builtin_amdgcn_smfmac_f32_32x32x32_fp8_fp8, "V16fV2iV4iV16fiIiIi", "nc", "fp8-insts")
 
-TARGET_BUILTIN(__builtin_amdgcn_cvt_f32_bf8, "fiIi", "nc", "fp8-insts")
-TARGET_BUILTIN(__builtin_amdgcn_cvt_f32_fp8, "fiIi", "nc", "fp8-insts")
-TARGET_BUILTIN(__builtin_amdgcn_cvt_pk_f32_bf8, "V2fiIb", "nc", "fp8-insts")
-TARGET_BUILTIN(__builtin_amdgcn_cvt_pk_f32_fp8, "V2fiIb", "nc", "fp8-insts")
-TARGET_BUILTIN(__builtin_amdgcn_cvt_pk_bf8_f32, "iffiIb", "nc", "fp8-insts")
-TARGET_BUILTIN(__builtin_amdgcn_cvt_pk_fp8_f32, "iffiIb", "nc", "fp8-insts")
-TARGET_BUILTIN(__builtin_amdgcn_cvt_sr_bf8_f32, "ifiiIi", "nc", "fp8-insts")
-TARGET_BUILTIN(__builtin_amdgcn_cvt_sr_fp8_f32, "ifiiIi", "nc", "fp8-insts")
+TARGET_BUILTIN(__builtin_amdgcn_cvt_f32_bf8, "fiIi", "nc", "fp8-conversion-insts")
+TARGET_BUILTIN(__builtin_amdgcn_cvt_f32_fp8, "fiIi", "nc", "fp8-conversion-insts")
+TARGET_BUILTIN(__builtin_amdgcn_cvt_pk_f32_bf8, "V2fiIb", "nc", "fp8-conversion-insts")
+TARGET_BUILTIN(__builtin_amdgcn_cvt_pk_f32_fp8, "V2fiIb", "nc", "fp8-conversion-insts")
+TARGET_BUILTIN(__builtin_amdgcn_cvt_pk_bf8_f32, "iffiIb", "nc", "fp8-conversion-insts")
+TARGET_BUILTIN(__builtin_amdgcn_cvt_pk_fp8_f32, "iffiIb", "nc", "fp8-conversion-insts")
+TARGET_BUILTIN(__builtin_amdgcn_cvt_sr_bf8_f32, "ifiiIi", "nc", "fp8-conversion-insts")
+TARGET_BUILTIN(__builtin_amdgcn_cvt_sr_fp8_f32, "ifiiIi", "nc", "fp8-conversion-insts")
 
 //===----------------------------------------------------------------------===//
 // SYCL builtin.
@@ -415,6 +419,7 @@ BUILTIN(__builtin_amdgcn_implicit_offset, "Ui*5", "nc")
 // GFX12+ only builtins.
 //===----------------------------------------------------------------------===//
 
+TARGET_BUILTIN(__builtin_amdgcn_s_sleep_var, "vUi", "n", "gfx12-insts")
 TARGET_BUILTIN(__builtin_amdgcn_permlane16_var,  "UiUiUiUiIbIb", "nc", "gfx12-insts")
 TARGET_BUILTIN(__builtin_amdgcn_permlanex16_var, "UiUiUiUiIbIb", "nc", "gfx12-insts")
 TARGET_BUILTIN(__builtin_amdgcn_s_barrier_signal, "vIi", "n", "gfx12-insts")
@@ -428,6 +433,13 @@ TARGET_BUILTIN(__builtin_amdgcn_s_wakeup_barrier, "vi", "n", "gfx12-insts")
 TARGET_BUILTIN(__builtin_amdgcn_s_barrier_leave, "b", "n", "gfx12-insts")
 TARGET_BUILTIN(__builtin_amdgcn_s_get_barrier_state, "Uii", "n", "gfx12-insts")
 
+TARGET_BUILTIN(__builtin_amdgcn_global_load_tr_v2i32, "V2iV2i*1", "nc", "gfx12-insts,wavefrontsize32")
+TARGET_BUILTIN(__builtin_amdgcn_global_load_tr_v8i16, "V8sV8s*1", "nc", "gfx12-insts,wavefrontsize32")
+TARGET_BUILTIN(__builtin_amdgcn_global_load_tr_v8f16, "V8hV8h*1", "nc", "gfx12-insts,wavefrontsize32")
+
+TARGET_BUILTIN(__builtin_amdgcn_global_load_tr_i32, "ii*1", "nc", "gfx12-insts,wavefrontsize64")
+TARGET_BUILTIN(__builtin_amdgcn_global_load_tr_v4i16, "V4sV4s*1", "nc", "gfx12-insts,wavefrontsize64")
+TARGET_BUILTIN(__builtin_amdgcn_global_load_tr_v4f16, "V4hV4h*1", "nc", "gfx12-insts,wavefrontsize64")
 
 #undef BUILTIN
 #undef TARGET_BUILTIN
diff --git a/clang/include/clang/Basic/BuiltinsPPC.def b/clang/include/clang/Basic/BuiltinsPPC.def
index a35488ed3dfa5..88ae0ce940852 100644
--- a/clang/include/clang/Basic/BuiltinsPPC.def
+++ b/clang/include/clang/Basic/BuiltinsPPC.def
@@ -110,6 +110,10 @@ BUILTIN(__builtin_ppc_fctiw, "dd", "")
 BUILTIN(__builtin_ppc_fctiwz, "dd", "")
 BUILTIN(__builtin_ppc_fctudz, "dd", "")
 BUILTIN(__builtin_ppc_fctuwz, "dd", "")
+
+// fence builtin prevents all instructions moved across it
+BUILTIN(__builtin_ppc_fence, "v", "")
+
 BUILTIN(__builtin_ppc_swdiv_nochk, "ddd", "")
 BUILTIN(__builtin_ppc_swdivs_nochk, "fff", "")
 BUILTIN(__builtin_ppc_alignx, "vIivC*", "nc")
diff --git a/clang/include/clang/Basic/CMakeLists.txt b/clang/include/clang/Basic/CMakeLists.txt
index 28baa2e45e423..be216150c71bb 100644
--- a/clang/include/clang/Basic/CMakeLists.txt
+++ b/clang/include/clang/Basic/CMakeLists.txt
@@ -45,10 +45,10 @@ clang_tablegen(AttrSubMatchRulesList.inc -gen-clang-attr-subject-match-rule-list
   SOURCE Attr.td
   TARGET ClangAttrSubjectMatchRuleList)
 
-clang_tablegen(AttrTokenKinds.inc -gen-clang-attr-token-kinds
+clang_tablegen(RegularKeywordAttrInfo.inc -gen-clang-regular-keyword-attr-info
   -I ${CMAKE_CURRENT_SOURCE_DIR}/../../
   SOURCE Attr.td
-  TARGET ClangAttrTokenKinds
+  TARGET ClangRegularKeywordAttrInfo
   )
 
 clang_tablegen(AttrHasAttributeImpl.inc -gen-clang-attr-has-attribute-impl
diff --git a/clang/include/clang/Basic/DiagnosticCommonKinds.td b/clang/include/clang/Basic/DiagnosticCommonKinds.td
index 4fec97ccc5e02..835f83088e655 100644
--- a/clang/include/clang/Basic/DiagnosticCommonKinds.td
+++ b/clang/include/clang/Basic/DiagnosticCommonKinds.td
@@ -361,9 +361,9 @@ def err_cannot_open_file : Error<"cannot open file '%0': %1">, DefaultFatal;
 def err_file_modified : Error<
   "file '%0' modified since it was first processed">, DefaultFatal;
 def err_file_too_large : Error<
-  "sorry, unsupported: file '%0' is too large for Clang to process">;
+  "file '%0' is too large for Clang to process">;
 def err_sloc_space_too_large : Error<
-  "sorry, the translation unit is too large for Clang to process: ran out of source locations">, DefaultFatal;
+  "translation unit is too large for Clang to process: ran out of source locations">, DefaultFatal;
 def err_unsupported_bom : Error<"%0 byte order mark detected in '%1', but "
   "encoding is not supported">, DefaultFatal;
 def err_unable_to_rename_temp : Error<
diff --git a/clang/include/clang/Basic/DiagnosticDriverKinds.td b/clang/include/clang/Basic/DiagnosticDriverKinds.td
index 27fdd56aea985..d40dc42d184f1 100644
--- a/clang/include/clang/Basic/DiagnosticDriverKinds.td
+++ b/clang/include/clang/Basic/DiagnosticDriverKinds.td
@@ -541,7 +541,7 @@ def warn_unsupported_branch_protection: Warning <
 def err_sls_hardening_arm_not_supported : Error<
   "-mharden-sls is only supported on armv7-a or later">;
 def warn_drv_large_data_threshold_invalid_code_model: Warning<
-  "'%0' only applies to medium code model">,
+  "'%0' only applies to medium and large code models">,
   InGroup;
 
 def note_drv_command_failed_diag_msg : Note<
diff --git a/clang/include/clang/Basic/DiagnosticParseKinds.td b/clang/include/clang/Basic/DiagnosticParseKinds.td
index 088f8b74983c8..a30ab27566ec3 100644
--- a/clang/include/clang/Basic/DiagnosticParseKinds.td
+++ b/clang/include/clang/Basic/DiagnosticParseKinds.td
@@ -1366,6 +1366,15 @@ def err_acc_invalid_open_paren
     : Error<"expected clause-list or newline in OpenACC directive">;
 def err_acc_invalid_default_clause_kind
     : Error<"invalid value for 'default' clause; expected 'present' or 'none'">;
+def err_acc_invalid_tag_kind
+    : Error<"invalid tag %0 on '%1' %select{directive|clause}2">;
+def err_acc_expected_reduction_operator
+    : Error<"missing reduction operator, expected '+', '*', 'max', 'min', '&', "
+            "'|', '^', '&&', or '||', follwed by a ':'">;
+def err_acc_invalid_reduction_operator
+    : Error<"invalid reduction operator,  expected '+', '*', 'max', 'min', "
+            "'&', '|', '^', '&&', or '||'">;
+def err_acc_incorrect_bind_arg : Error<"expected identifier or string literal">;
 
 // OpenMP support.
 def warn_pragma_omp_ignored : Warning<
@@ -1609,6 +1618,9 @@ def err_pragma_invalid_keyword : Error<
 def err_pragma_pipeline_invalid_keyword : Error<
     "invalid argument; expected 'disable'">;
 
+// API notes.
+def err_type_unparsed : Error<"unparsed tokens following type">;
+
 // Pragma unroll support.
 def warn_pragma_unroll_cuda_value_in_parens : Warning<
   "argument to '#pragma unroll' should not be in parentheses in CUDA C/C++">,
diff --git a/clang/include/clang/Basic/DiagnosticSemaKinds.td b/clang/include/clang/Basic/DiagnosticSemaKinds.td
index a849bf4431407..91ca589c072b1 100644
--- a/clang/include/clang/Basic/DiagnosticSemaKinds.td
+++ b/clang/include/clang/Basic/DiagnosticSemaKinds.td
@@ -3156,6 +3156,8 @@ def err_musttail_scope : Error<
   "cannot perform a tail call from this return statement">;
 def err_musttail_no_variadic : Error<
   "%0 attribute may not be used with variadic functions">;
+def err_musttail_no_return : Error<
+  "%0 attribute may not be used with no-return-attribute functions">;
 
 def err_nsobject_attribute : Error<
   "'NSObject' attribute is for pointer types only">;
@@ -3758,6 +3760,12 @@ def err_sme_definition_using_sm_in_non_sme_target : Error<
   "function executed in streaming-SVE mode requires 'sme'">;
 def err_sme_definition_using_za_in_non_sme_target : Error<
   "function using ZA state requires 'sme'">;
+def err_conflicting_attributes_arm_state : Error<
+  "conflicting attributes for state '%0'">;
+def err_unknown_arm_state : Error<
+  "unknown state '%0'">;
+def err_missing_arm_state : Error<
+  "missing state for %0">;
 def err_cconv_change : Error<
   "function declared '%0' here was previously declared "
   "%select{'%2'|without calling convention}1">;
@@ -5199,7 +5207,7 @@ def err_non_type_template_arg_subobject : Error<
 def err_non_type_template_arg_addr_label_diff : Error<
   "template argument / label address difference / what did you expect?">;
 def err_non_type_template_arg_unsupported : Error<
-  "sorry, non-type template argument of type %0 is not yet supported">;
+  "non-type template argument of type %0 is not yet supported">;
 def err_template_arg_not_convertible : Error<
   "non-type template argument of type %0 cannot be converted to a value "
   "of type %1">;
@@ -5252,7 +5260,7 @@ def err_template_arg_not_object_or_func : Error<
 def err_template_arg_not_pointer_to_member_form : Error<
   "non-type template argument is not a pointer to member constant">;
 def err_template_arg_member_ptr_base_derived_not_supported : Error<
-  "sorry, non-type template argument of pointer-to-member type %1 that refers "
+  "non-type template argument of pointer-to-member type %1 that refers "
   "to member %q0 of a different class is not supported yet">;
 def err_template_arg_invalid : Error<
   "non-type template argument '%0' is invalid">;
@@ -7064,6 +7072,11 @@ def err_member_decl_does_not_match : Error<
   "does not match any declaration in %1">;
 def err_friend_decl_with_def_arg_must_be_def : Error<
   "friend declaration specifying a default argument must be a definition">;
+def err_friend_decl_with_enclosing_temp_constraint_must_be_def : Error<
+  "friend declaration with a constraint that depends on an enclosing "
+  "template parameter must be a definition">;
+def err_non_temp_friend_decl_with_requires_clause_must_be_def : Error<
+  "non-template friend declaration with a requires clause must be a definition">;
 def err_friend_decl_with_def_arg_redeclared : Error<
   "friend declaration specifying a default argument must be the only declaration">;
 def err_friend_decl_does_not_match : Error<
@@ -10001,10 +10014,10 @@ def warn_new_dangling_initializer_list : Warning<
   "will be destroyed at the end of the full-expression">,
   InGroup;
 def warn_unsupported_lifetime_extension : Warning<
-  "sorry, lifetime extension of "
+  "lifetime extension of "
   "%select{temporary|backing array of initializer list}0 created "
-  "by aggregate initialization using default member initializer "
-  "is not supported; lifetime of %select{temporary|backing array}0 "
+  "by aggregate initialization using a default member initializer "
+  "is not yet supported; lifetime of %select{temporary|backing array}0 "
   "will end at the end of the full-expression">, InGroup;
 
 // For non-floating point, expressions of the form x == x or x != x
diff --git a/clang/include/clang/Basic/DiagnosticSerializationKinds.td b/clang/include/clang/Basic/DiagnosticSerializationKinds.td
index 3cb2cd32cf6d0..11c706ebf84b5 100644
--- a/clang/include/clang/Basic/DiagnosticSerializationKinds.td
+++ b/clang/include/clang/Basic/DiagnosticSerializationKinds.td
@@ -129,10 +129,8 @@ def warn_module_system_bit_conflict : Warning<
   "as a non-system module; any difference in diagnostic options will be ignored">,
   InGroup;
 
-def warn_reading_std_cxx_module_by_implicit_paths : Warning<
-  "it is deprecated to read module '%0' implicitly; it is going to be removed in clang 18; "
-  "consider to specify the dependencies explicitly">,
-  InGroup>;
+def err_failed_to_find_module_file : Error<
+  "failed to find module file for module '%0'">;
 } // let CategoryName
 
 let CategoryName = "AST Serialization Issue" in {
diff --git a/clang/include/clang/Basic/LangOptions.def b/clang/include/clang/Basic/LangOptions.def
index 107216da7eae9..deab962a31afb 100644
--- a/clang/include/clang/Basic/LangOptions.def
+++ b/clang/include/clang/Basic/LangOptions.def
@@ -480,6 +480,7 @@ ENUM_LANGOPT(SignReturnAddressKey, SignReturnAddressKeyKind, 1, SignReturnAddres
              "Key used for return address signing")
 LANGOPT(BranchTargetEnforcement, 1, 0, "Branch-target enforcement enabled")
 LANGOPT(BranchProtectionPAuthLR, 1, 0, "Use PC as a diversifier using PAuthLR NOP instructions.")
+LANGOPT(GuardedControlStack, 1, 0, "Guarded control stack enabled")
 
 LANGOPT(SpeculativeLoadHardening, 1, 0, "Speculative load hardening enabled")
 
diff --git a/clang/include/clang/Basic/OpenACCKinds.h b/clang/include/clang/Basic/OpenACCKinds.h
index b0c157e002360..a27b78d2e6da8 100644
--- a/clang/include/clang/Basic/OpenACCKinds.h
+++ b/clang/include/clang/Basic/OpenACCKinds.h
@@ -14,6 +14,9 @@
 #ifndef LLVM_CLANG_BASIC_OPENACCKINDS_H
 #define LLVM_CLANG_BASIC_OPENACCKINDS_H
 
+#include "clang/Basic/Diagnostic.h"
+#include "llvm/Support/ErrorHandling.h"
+
 namespace clang {
 // Represents the Construct/Directive kind of a pragma directive. Note the
 // OpenACC standard is inconsistent between calling these Construct vs
@@ -62,6 +65,75 @@ enum class OpenACCDirectiveKind {
   Invalid,
 };
 
+inline const StreamingDiagnostic &operator<<(const StreamingDiagnostic &Out,
+                                             OpenACCDirectiveKind K) {
+  switch (K) {
+  case OpenACCDirectiveKind::Parallel:
+    return Out << "parallel";
+
+  case OpenACCDirectiveKind::Serial:
+    return Out << "serial";
+
+  case OpenACCDirectiveKind::Kernels:
+    return Out << "kernels";
+
+  case OpenACCDirectiveKind::Data:
+    return Out << "data";
+
+  case OpenACCDirectiveKind::EnterData:
+    return Out << "enter data";
+
+  case OpenACCDirectiveKind::ExitData:
+    return Out << "exit data";
+
+  case OpenACCDirectiveKind::HostData:
+    return Out << "host_data";
+
+  case OpenACCDirectiveKind::Loop:
+    return Out << "loop";
+
+  case OpenACCDirectiveKind::Cache:
+    return Out << "cache";
+
+  case OpenACCDirectiveKind::ParallelLoop:
+    return Out << "parallel loop";
+
+  case OpenACCDirectiveKind::SerialLoop:
+    return Out << "serial loop";
+
+  case OpenACCDirectiveKind::KernelsLoop:
+    return Out << "kernels loop";
+
+  case OpenACCDirectiveKind::Atomic:
+    return Out << "atomic";
+
+  case OpenACCDirectiveKind::Declare:
+    return Out << "declare";
+
+  case OpenACCDirectiveKind::Init:
+    return Out << "init";
+
+  case OpenACCDirectiveKind::Shutdown:
+    return Out << "shutdown";
+
+  case OpenACCDirectiveKind::Set:
+    return Out << "set";
+
+  case OpenACCDirectiveKind::Update:
+    return Out << "update";
+
+  case OpenACCDirectiveKind::Wait:
+    return Out << "wait";
+
+  case OpenACCDirectiveKind::Routine:
+    return Out << "routine";
+
+  case OpenACCDirectiveKind::Invalid:
+    return Out << "";
+  }
+  llvm_unreachable("Uncovered directive kind");
+}
+
 enum class OpenACCAtomicKind {
   Read,
   Write,
@@ -98,10 +170,163 @@ enum class OpenACCClauseKind {
   If,
   /// 'self' clause, allowed on Compute and Combined Constructs, plus 'update'.
   Self,
+  /// 'copy' clause, allowed on Compute and Combined Constructs, plus 'data' and
+  /// 'declare'.
+  Copy,
+  /// 'use_device' clause, allowed on 'host_data' construct.
+  UseDevice,
+  /// 'attach' clause, allowed on Compute and Combined constructs, plus 'data'
+  /// and 'enter data'.
+  Attach,
+  /// 'delete' clause, allowed on the 'exit data' construct.
+  Delete,
+  /// 'detach' clause, allowed on the 'exit data' construct.
+  Detach,
+  /// 'device' clause, allowed on the 'update' construct.
+  Device,
+  /// 'deviceptr' clause, allowed on Compute and Combined Constructs, plus
+  /// 'data' and 'declare'.
+  DevicePtr,
+  /// 'device_resident' clause, allowed on the 'declare' construct.
+  DeviceResident,
+  /// 'firstprivate' clause, allowed on 'parallel', 'serial', 'parallel loop',
+  /// and 'serial loop' constructs.
+  FirstPrivate,
+  /// 'host' clause, allowed on 'update' construct.
+  Host,
+  /// 'link' clause, allowed on 'declare' construct.
+  Link,
+  /// 'no_create' clause, allowed on allowed on Compute and Combined constructs,
+  /// plus 'data'.
+  NoCreate,
+  /// 'present' clause, allowed on Compute and Combined constructs, plus 'data'
+  /// and 'declare'.
+  Present,
+  /// 'private' clause, allowed on 'parallel', 'serial', 'loop', 'parallel
+  /// loop', and 'serial loop' constructs.
+  Private,
+  /// 'copyout' clause, allowed on Compute and Combined constructs, plus 'data',
+  /// 'exit data', and 'declare'.
+  CopyOut,
+  /// 'copyin' clause, allowed on Compute and Combined constructs, plus 'data',
+  /// 'enter data', and 'declare'.
+  CopyIn,
+  /// 'copyin' clause, allowed on Compute and Combined constructs, plus 'data',
+  /// 'enter data', and 'declare'.
+  Create,
+  /// 'reduction' clause, allowed on Parallel, Serial, Loop, and the combined
+  /// constructs.
+  Reduction,
+  /// 'collapse' clause, allowed on 'loop' and Combined constructs.
+  Collapse,
+  /// 'bind' clause, allowed on routine constructs.
+  Bind,
+
   /// Represents an invalid clause, for the purposes of parsing.
   Invalid,
 };
 
+inline const StreamingDiagnostic &operator<<(const StreamingDiagnostic &Out,
+                                             OpenACCClauseKind K) {
+  switch (K) {
+  case OpenACCClauseKind::Finalize:
+    return Out << "finalize";
+
+  case OpenACCClauseKind::IfPresent:
+    return Out << "if_present";
+
+  case OpenACCClauseKind::Seq:
+    return Out << "seq";
+
+  case OpenACCClauseKind::Independent:
+    return Out << "independent";
+
+  case OpenACCClauseKind::Auto:
+    return Out << "auto";
+
+  case OpenACCClauseKind::Worker:
+    return Out << "worker";
+
+  case OpenACCClauseKind::Vector:
+    return Out << "vector";
+
+  case OpenACCClauseKind::NoHost:
+    return Out << "nohost";
+
+  case OpenACCClauseKind::Default:
+    return Out << "default";
+
+  case OpenACCClauseKind::If:
+    return Out << "if";
+
+  case OpenACCClauseKind::Self:
+    return Out << "self";
+
+  case OpenACCClauseKind::Copy:
+    return Out << "copy";
+
+  case OpenACCClauseKind::UseDevice:
+    return Out << "use_device";
+
+  case OpenACCClauseKind::Attach:
+    return Out << "attach";
+
+  case OpenACCClauseKind::Delete:
+    return Out << "delete";
+
+  case OpenACCClauseKind::Detach:
+    return Out << "detach";
+
+  case OpenACCClauseKind::Device:
+    return Out << "device";
+
+  case OpenACCClauseKind::DevicePtr:
+    return Out << "deviceptr";
+
+  case OpenACCClauseKind::DeviceResident:
+    return Out << "device_resident";
+
+  case OpenACCClauseKind::FirstPrivate:
+    return Out << "firstprivate";
+
+  case OpenACCClauseKind::Host:
+    return Out << "host";
+
+  case OpenACCClauseKind::Link:
+    return Out << "link";
+
+  case OpenACCClauseKind::NoCreate:
+    return Out << "no_create";
+
+  case OpenACCClauseKind::Present:
+    return Out << "present";
+
+  case OpenACCClauseKind::Private:
+    return Out << "private";
+
+  case OpenACCClauseKind::CopyOut:
+    return Out << "copyout";
+
+  case OpenACCClauseKind::CopyIn:
+    return Out << "copyin";
+
+  case OpenACCClauseKind::Create:
+    return Out << "create";
+
+  case OpenACCClauseKind::Reduction:
+    return Out << "reduction";
+
+  case OpenACCClauseKind::Collapse:
+    return Out << "collapse";
+
+  case OpenACCClauseKind::Bind:
+    return Out << "bind";
+
+  case OpenACCClauseKind::Invalid:
+    return Out << "";
+  }
+  llvm_unreachable("Uncovered clause kind");
+}
 enum class OpenACCDefaultClauseKind {
   /// 'none' option.
   None,
@@ -110,6 +335,29 @@ enum class OpenACCDefaultClauseKind {
   /// Not a valid option.
   Invalid,
 };
+
+enum class OpenACCReductionOperator {
+  /// '+'.
+  Addition,
+  /// '*'.
+  Multiplication,
+  /// 'max'.
+  Max,
+  /// 'min'.
+  Min,
+  /// '&'.
+  BitwiseAnd,
+  /// '|'.
+  BitwiseOr,
+  /// '^'.
+  BitwiseXOr,
+  /// '&&'.
+  And,
+  /// '||'.
+  Or,
+  /// Invalid Reduction Clause Kind.
+  Invalid,
+};
 } // namespace clang
 
 #endif // LLVM_CLANG_BASIC_OPENACCKINDS_H
diff --git a/clang/include/clang/Basic/TargetInfo.h b/clang/include/clang/Basic/TargetInfo.h
index ac3c324c6c29c..3eb23ebdacf0e 100644
--- a/clang/include/clang/Basic/TargetInfo.h
+++ b/clang/include/clang/Basic/TargetInfo.h
@@ -1373,6 +1373,7 @@ class TargetInfo : public TransferrableTargetInfo,
         LangOptions::SignReturnAddressKeyKind::AKey;
     bool BranchTargetEnforcement = false;
     bool BranchProtectionPAuthLR = false;
+    bool GuardedControlStack = false;
   };
 
   /// Determine if the Architecture in this TargetInfo supports branch
diff --git a/clang/include/clang/Basic/TokenKinds.def b/clang/include/clang/Basic/TokenKinds.def
index c7bce28eb8994..6e1c2eac77866 100644
--- a/clang/include/clang/Basic/TokenKinds.def
+++ b/clang/include/clang/Basic/TokenKinds.def
@@ -769,9 +769,9 @@ KEYWORD(__builtin_sycl_unique_stable_id  , KEYSYCL)
 
 // Keywords defined by Attr.td.
 #ifndef KEYWORD_ATTRIBUTE
-#define KEYWORD_ATTRIBUTE(X) KEYWORD(X, KEYALL)
+#define KEYWORD_ATTRIBUTE(X, ...) KEYWORD(X, KEYALL)
 #endif
-#include "clang/Basic/AttrTokenKinds.inc"
+#include "clang/Basic/RegularKeywordAttrInfo.inc"
 
 // Clang-specific keywords enabled only in testing.
 TESTING_KEYWORD(__unknown_anytype , KEYALL)
diff --git a/clang/include/clang/Basic/TokenKinds.h b/clang/include/clang/Basic/TokenKinds.h
index e4857405bc7f4..7529b922619ad 100644
--- a/clang/include/clang/Basic/TokenKinds.h
+++ b/clang/include/clang/Basic/TokenKinds.h
@@ -109,8 +109,8 @@ bool isPragmaAnnotation(TokenKind K);
 
 inline constexpr bool isRegularKeywordAttribute(TokenKind K) {
   return (false
-#define KEYWORD_ATTRIBUTE(X) || (K == tok::kw_##X)
-#include "clang/Basic/AttrTokenKinds.inc"
+#define KEYWORD_ATTRIBUTE(X, ...) || (K == tok::kw_##X)
+#include "clang/Basic/RegularKeywordAttrInfo.inc"
   );
 }
 
diff --git a/clang/include/clang/Basic/arm_sve.td b/clang/include/clang/Basic/arm_sve.td
index 6f35e25617ade..51cf8e29129c6 100644
--- a/clang/include/clang/Basic/arm_sve.td
+++ b/clang/include/clang/Basic/arm_sve.td
@@ -1281,9 +1281,9 @@ def SVUNDEF_2 : SInst<"svundef2_{d}", "2v", "csilUcUsUiUlhfd", MergeNone, "", [I
 def SVUNDEF_3 : SInst<"svundef3_{d}", "3v", "csilUcUsUiUlhfd", MergeNone, "", [IsUndef, IsStreamingCompatible]>;
 def SVUNDEF_4 : SInst<"svundef4_{d}", "4v", "csilUcUsUiUlhfd", MergeNone, "", [IsUndef, IsStreamingCompatible]>;
 
-def SVCREATE_2 : SInst<"svcreate2[_{d}]", "2dd",   "csilUcUsUiUlhfd", MergeNone, "", [IsTupleCreate]>;
-def SVCREATE_3 : SInst<"svcreate3[_{d}]", "3ddd",  "csilUcUsUiUlhfd", MergeNone, "", [IsTupleCreate]>;
-def SVCREATE_4 : SInst<"svcreate4[_{d}]", "4dddd", "csilUcUsUiUlhfd", MergeNone, "", [IsTupleCreate]>;
+def SVCREATE_2 : SInst<"svcreate2[_{d}]", "2dd",   "csilUcUsUiUlhfd", MergeNone, "", [IsTupleCreate, IsStreamingCompatible]>;
+def SVCREATE_3 : SInst<"svcreate3[_{d}]", "3ddd",  "csilUcUsUiUlhfd", MergeNone, "", [IsTupleCreate, IsStreamingCompatible]>;
+def SVCREATE_4 : SInst<"svcreate4[_{d}]", "4dddd", "csilUcUsUiUlhfd", MergeNone, "", [IsTupleCreate, IsStreamingCompatible]>;
 
 let TargetGuard = "sve,bf16" in {
 def SVUNDEF_1_BF16 : SInst<"svundef_{d}",  "dv", "b", MergeNone, "", [IsUndef, IsStreamingCompatible]>;
@@ -1291,42 +1291,47 @@ def SVUNDEF_2_BF16 : SInst<"svundef2_{d}", "2v", "b", MergeNone, "", [IsUndef, I
 def SVUNDEF_3_BF16 : SInst<"svundef3_{d}", "3v", "b", MergeNone, "", [IsUndef, IsStreamingCompatible]>;
 def SVUNDEF_4_BF16 : SInst<"svundef4_{d}", "4v", "b", MergeNone, "", [IsUndef, IsStreamingCompatible]>;
 
-def SVCREATE_2_BF16 : SInst<"svcreate2[_{d}]", "2dd",   "b", MergeNone, "", [IsTupleCreate]>;
-def SVCREATE_3_BF16 : SInst<"svcreate3[_{d}]", "3ddd",  "b", MergeNone, "", [IsTupleCreate]>;
-def SVCREATE_4_BF16 : SInst<"svcreate4[_{d}]", "4dddd", "b", MergeNone, "", [IsTupleCreate]>;
+def SVCREATE_2_BF16 : SInst<"svcreate2[_{d}]", "2dd",   "b", MergeNone, "", [IsTupleCreate, IsStreamingCompatible]>;
+def SVCREATE_3_BF16 : SInst<"svcreate3[_{d}]", "3ddd",  "b", MergeNone, "", [IsTupleCreate, IsStreamingCompatible]>;
+def SVCREATE_4_BF16 : SInst<"svcreate4[_{d}]", "4dddd", "b", MergeNone, "", [IsTupleCreate, IsStreamingCompatible]>;
 }
 
-let TargetGuard = "sve2p1" in {
-  def SVCREATE_2_B : SInst<"svcreate2[_{d}]", "2dd",   "Pc", MergeNone, "", [IsTupleCreate]>;
-  def SVCREATE_4_B : SInst<"svcreate4[_{d}]", "4dddd", "Pc", MergeNone, "", [IsTupleCreate]>;
+let TargetGuard = "sve2p1|sme2" in {
+  def SVCREATE_2_B : SInst<"svcreate2[_b]", "2dd",   "Pc", MergeNone, "", [IsTupleCreate, IsStreamingCompatible]>;
+  def SVCREATE_4_B : SInst<"svcreate4[_b]", "4dddd", "Pc", MergeNone, "", [IsTupleCreate, IsStreamingCompatible]>;
 }
 
 ////////////////////////////////////////////////////////////////////////////////
 // Vector insertion and extraction
-def SVGET_2 : SInst<"svget2[_{d}]", "d2i", "csilUcUsUiUlhfd", MergeNone, "", [IsTupleGet], [ImmCheck<1, ImmCheck0_1>]>;
-def SVGET_3 : SInst<"svget3[_{d}]", "d3i", "csilUcUsUiUlhfd", MergeNone, "", [IsTupleGet], [ImmCheck<1, ImmCheck0_2>]>;
-def SVGET_4 : SInst<"svget4[_{d}]", "d4i", "csilUcUsUiUlhfd", MergeNone, "", [IsTupleGet], [ImmCheck<1, ImmCheck0_3>]>;
+def SVGET_2 : SInst<"svget2[_{d}]", "d2i", "csilUcUsUiUlhfd", MergeNone, "", [IsTupleGet, IsStreamingCompatible], [ImmCheck<1, ImmCheck0_1>]>;
+def SVGET_3 : SInst<"svget3[_{d}]", "d3i", "csilUcUsUiUlhfd", MergeNone, "", [IsTupleGet, IsStreamingCompatible], [ImmCheck<1, ImmCheck0_2>]>;
+def SVGET_4 : SInst<"svget4[_{d}]", "d4i", "csilUcUsUiUlhfd", MergeNone, "", [IsTupleGet, IsStreamingCompatible], [ImmCheck<1, ImmCheck0_3>]>;
 
-def SVSET_2 : SInst<"svset2[_{d}]", "22id", "csilUcUsUiUlhfd", MergeNone, "", [IsTupleSet], [ImmCheck<1, ImmCheck0_1>]>;
-def SVSET_3 : SInst<"svset3[_{d}]", "33id", "csilUcUsUiUlhfd", MergeNone, "", [IsTupleSet], [ImmCheck<1, ImmCheck0_2>]>;
-def SVSET_4 : SInst<"svset4[_{d}]", "44id", "csilUcUsUiUlhfd", MergeNone, "", [IsTupleSet], [ImmCheck<1, ImmCheck0_3>]>;
+def SVSET_2 : SInst<"svset2[_{d}]", "22id", "csilUcUsUiUlhfd", MergeNone, "", [IsTupleSet, IsStreamingCompatible], [ImmCheck<1, ImmCheck0_1>]>;
+def SVSET_3 : SInst<"svset3[_{d}]", "33id", "csilUcUsUiUlhfd", MergeNone, "", [IsTupleSet, IsStreamingCompatible], [ImmCheck<1, ImmCheck0_2>]>;
+def SVSET_4 : SInst<"svset4[_{d}]", "44id", "csilUcUsUiUlhfd", MergeNone, "", [IsTupleSet, IsStreamingCompatible], [ImmCheck<1, ImmCheck0_3>]>;
 
 let TargetGuard = "sve,bf16" in {
-def SVGET_2_BF16 : SInst<"svget2[_{d}]", "d2i", "b", MergeNone, "", [IsTupleGet], [ImmCheck<1, ImmCheck0_1>]>;
-def SVGET_3_BF16 : SInst<"svget3[_{d}]", "d3i", "b", MergeNone, "", [IsTupleGet], [ImmCheck<1, ImmCheck0_2>]>;
-def SVGET_4_BF16 : SInst<"svget4[_{d}]", "d4i", "b", MergeNone, "", [IsTupleGet], [ImmCheck<1, ImmCheck0_3>]>;
+def SVGET_2_BF16 : SInst<"svget2[_{d}]", "d2i", "b", MergeNone, "", [IsTupleGet, IsStreamingCompatible], [ImmCheck<1, ImmCheck0_1>]>;
+def SVGET_3_BF16 : SInst<"svget3[_{d}]", "d3i", "b", MergeNone, "", [IsTupleGet, IsStreamingCompatible], [ImmCheck<1, ImmCheck0_2>]>;
+def SVGET_4_BF16 : SInst<"svget4[_{d}]", "d4i", "b", MergeNone, "", [IsTupleGet, IsStreamingCompatible], [ImmCheck<1, ImmCheck0_3>]>;
 
-def SVSET_2_BF16 : SInst<"svset2[_{d}]", "22id", "b", MergeNone, "", [IsTupleSet], [ImmCheck<1, ImmCheck0_1>]>;
-def SVSET_3_BF16 : SInst<"svset3[_{d}]", "33id", "b", MergeNone, "", [IsTupleSet], [ImmCheck<1, ImmCheck0_2>]>;
-def SVSET_4_BF16 : SInst<"svset4[_{d}]", "44id", "b", MergeNone, "", [IsTupleSet], [ImmCheck<1, ImmCheck0_3>]>;
+def SVSET_2_BF16 : SInst<"svset2[_{d}]", "22id", "b", MergeNone, "", [IsTupleSet, IsStreamingCompatible], [ImmCheck<1, ImmCheck0_1>]>;
+def SVSET_3_BF16 : SInst<"svset3[_{d}]", "33id", "b", MergeNone, "", [IsTupleSet, IsStreamingCompatible], [ImmCheck<1, ImmCheck0_2>]>;
+def SVSET_4_BF16 : SInst<"svset4[_{d}]", "44id", "b", MergeNone, "", [IsTupleSet, IsStreamingCompatible], [ImmCheck<1, ImmCheck0_3>]>;
 }
 
-let TargetGuard = "sve2p1" in {
-  def SVGET_2_B : SInst<"svget2[_{d}]", "d2i", "Pc", MergeNone, "", [IsTupleGet], [ImmCheck<1, ImmCheck0_1>]>;
-  def SVGET_4_B : SInst<"svget4[_{d}]", "d4i", "Pc", MergeNone, "", [IsTupleGet], [ImmCheck<1, ImmCheck0_3>]>;
+let TargetGuard = "sve2p1|sme2" in {
+  def SVGET_2_B : SInst<"svget2[_b]", "d2i", "Pc", MergeNone, "", [IsTupleGet, IsStreamingCompatible], [ImmCheck<1, ImmCheck0_1>]>;
+  def SVGET_4_B : SInst<"svget4[_b]", "d4i", "Pc", MergeNone, "", [IsTupleGet, IsStreamingCompatible], [ImmCheck<1, ImmCheck0_3>]>;
+
+  def SVSET_2_B : SInst<"svset2[_b]", "22id", "Pc", MergeNone, "", [IsTupleSet, IsStreamingCompatible], [ImmCheck<1, ImmCheck0_1>]>;
+  def SVSET_4_B : SInst<"svset4[_b]", "44id", "Pc", MergeNone, "", [IsTupleSet, IsStreamingCompatible], [ImmCheck<1, ImmCheck0_3>]>;
+}
 
-  def SVSET_2_B : SInst<"svset2[_{d}]", "22id", "Pc", MergeNone, "", [IsTupleSet], [ImmCheck<1, ImmCheck0_1>]>;
-  def SVSET_4_B : SInst<"svset4[_{d}]", "44id", "Pc", MergeNone, "", [IsTupleSet], [ImmCheck<1, ImmCheck0_3>]>;
+let TargetGuard = "sve2p1|sme2" in {
+  def SVUNDEF_2_B: Inst<"svundef2_b", "2", "Pc", MergeNone, "", [IsUndef, IsStreamingCompatible], []>;
+  def SVUNDEF_4_B: Inst<"svundef4_b", "4", "Pc", MergeNone, "", [IsUndef, IsStreamingCompatible], []>;
 }
 ////////////////////////////////////////////////////////////////////////////////
 // SVE2 WhileGE/GT
@@ -1948,14 +1953,10 @@ def SVBGRP_N : SInst<"svbgrp[_n_{d}]", "dda", "UcUsUiUl", MergeNone, "aarch64_sv
 }
 
 let TargetGuard = "sve2p1|sme" in {
-def SVPSEL_B : SInst<"svpsel_lane_b8",  "PPPm", "Pc", MergeNone, "", [IsStreamingCompatible], []>;
-def SVPSEL_H : SInst<"svpsel_lane_b16", "PPPm", "Ps", MergeNone, "", [IsStreamingCompatible], []>;
-def SVPSEL_S : SInst<"svpsel_lane_b32", "PPPm", "Pi", MergeNone, "", [IsStreamingCompatible], []>;
-def SVPSEL_D : SInst<"svpsel_lane_b64", "PPPm", "Pl", MergeNone, "", [IsStreamingCompatible], []>;
-def SVPSEL_COUNT_ALIAS_B : SInst<"svpsel_lane_c8",  "}}Pm", "Pc", MergeNone, "", [IsStreamingCompatible], []>;
-def SVPSEL_COUNT_ALIAS_H : SInst<"svpsel_lane_c16", "}}Pm", "Ps", MergeNone, "", [IsStreamingCompatible], []>;
-def SVPSEL_COUNT_ALIAS_S : SInst<"svpsel_lane_c32", "}}Pm", "Pi", MergeNone, "", [IsStreamingCompatible], []>;
-def SVPSEL_COUNT_ALIAS_D : SInst<"svpsel_lane_c64", "}}Pm", "Pl", MergeNone, "", [IsStreamingCompatible], []>;
+def SVPSEL_B : SInst<"svpsel_lane_b8",  "PPPm", "Pc", MergeNone, "", [IsStreamingOrSVE2p1], []>;
+def SVPSEL_H : SInst<"svpsel_lane_b16", "PPPm", "Ps", MergeNone, "", [IsStreamingOrSVE2p1], []>;
+def SVPSEL_S : SInst<"svpsel_lane_b32", "PPPm", "Pi", MergeNone, "", [IsStreamingOrSVE2p1], []>;
+def SVPSEL_D : SInst<"svpsel_lane_b64", "PPPm", "Pl", MergeNone, "", [IsStreamingOrSVE2p1], []>;
 }
 
 // Standalone sve2.1 builtins
@@ -1979,6 +1980,11 @@ let TargetGuard = "sve2p1|sme2" in {
 def SVPEXT_SINGLE : SInst<"svpext_lane_{d}", "P}i", "QcQsQiQl", MergeNone, "aarch64_sve_pext", [IsStreamingOrSVE2p1], [ImmCheck<1, ImmCheck0_3>]>;
 def SVPEXT_X2     : SInst<"svpext_lane_{d}_x2", "2.P}i", "QcQsQiQl", MergeNone, "aarch64_sve_pext_x2", [IsStreamingOrSVE2p1], [ImmCheck<1, ImmCheck0_1>]>;
 
+def SVPSEL_COUNT_ALIAS_B : SInst<"svpsel_lane_c8",  "}}Pm", "Pc", MergeNone, "", [IsStreamingOrSVE2p1], []>;
+def SVPSEL_COUNT_ALIAS_H : SInst<"svpsel_lane_c16", "}}Pm", "Ps", MergeNone, "", [IsStreamingOrSVE2p1], []>;
+def SVPSEL_COUNT_ALIAS_S : SInst<"svpsel_lane_c32", "}}Pm", "Pi", MergeNone, "", [IsStreamingOrSVE2p1], []>;
+def SVPSEL_COUNT_ALIAS_D : SInst<"svpsel_lane_c64", "}}Pm", "Pl", MergeNone, "", [IsStreamingOrSVE2p1], []>;
+
 def SVWHILEGE_COUNT  : SInst<"svwhilege_{d}[_{1}]",  "}lli", "QcQsQiQl", MergeNone, "aarch64_sve_whilege_{d}", [IsOverloadNone, IsStreamingOrSVE2p1], [ImmCheck<2, ImmCheck2_4_Mul2>]>;
 def SVWHILEGT_COUNT  : SInst<"svwhilegt_{d}[_{1}]",  "}lli", "QcQsQiQl", MergeNone, "aarch64_sve_whilegt_{d}", [IsOverloadNone, IsStreamingOrSVE2p1], [ImmCheck<2, ImmCheck2_4_Mul2>]>;
 def SVWHILELE_COUNT  : SInst<"svwhilele_{d}[_{1}]",  "}lli", "QcQsQiQl", MergeNone, "aarch64_sve_whilele_{d}", [IsOverloadNone, IsStreamingOrSVE2p1], [ImmCheck<2, ImmCheck2_4_Mul2>]>;
@@ -2039,13 +2045,13 @@ let TargetGuard = "sve2p1|sme2" in {
   defm STNT1 : MultiVecStore<"stnt1">;
 }
 
-let TargetGuard = "sve2p1" in {
-def SVDOT_X2_S : SInst<"svdot[_{d}_{2}]", "ddhh", "i",  MergeNone, "aarch64_sve_sdot_x2", [], []>;
-def SVDOT_X2_U : SInst<"svdot[_{d}_{2}]", "ddhh", "Ui", MergeNone, "aarch64_sve_udot_x2", [], []>;
-def SVDOT_X2_F : SInst<"svdot[_{d}_{2}]", "ddhh", "f",  MergeNone, "aarch64_sve_fdot_x2", [], []>;
-def SVDOT_LANE_X2_S : SInst<"svdot_lane[_{d}_{2}]", "ddhhi", "i",  MergeNone, "aarch64_sve_sdot_lane_x2", [], [ImmCheck<3, ImmCheck0_3>]>;
-def SVDOT_LANE_X2_U : SInst<"svdot_lane[_{d}_{2}]", "ddhhi", "Ui", MergeNone, "aarch64_sve_udot_lane_x2", [], [ImmCheck<3, ImmCheck0_3>]>;
-def SVDOT_LANE_X2_F : SInst<"svdot_lane[_{d}_{2}]", "ddhhi", "f",  MergeNone, "aarch64_sve_fdot_lane_x2", [], [ImmCheck<3, ImmCheck0_3>]>;
+let TargetGuard = "sve2p1|sme2" in {
+def SVDOT_X2_S : SInst<"svdot[_{d}_{2}]", "ddhh", "i",  MergeNone, "aarch64_sve_sdot_x2", [IsStreamingOrSVE2p1], []>;
+def SVDOT_X2_U : SInst<"svdot[_{d}_{2}]", "ddhh", "Ui", MergeNone, "aarch64_sve_udot_x2", [IsStreamingOrSVE2p1], []>;
+def SVDOT_X2_F : SInst<"svdot[_{d}_{2}]", "ddhh", "f",  MergeNone, "aarch64_sve_fdot_x2", [IsStreamingOrSVE2p1], []>;
+def SVDOT_LANE_X2_S : SInst<"svdot_lane[_{d}_{2}]", "ddhhi", "i",  MergeNone, "aarch64_sve_sdot_lane_x2", [IsStreamingOrSVE2p1], [ImmCheck<3, ImmCheck0_3>]>;
+def SVDOT_LANE_X2_U : SInst<"svdot_lane[_{d}_{2}]", "ddhhi", "Ui", MergeNone, "aarch64_sve_udot_lane_x2", [IsStreamingOrSVE2p1], [ImmCheck<3, ImmCheck0_3>]>;
+def SVDOT_LANE_X2_F : SInst<"svdot_lane[_{d}_{2}]", "ddhhi", "f",  MergeNone, "aarch64_sve_fdot_lane_x2", [IsStreamingOrSVE2p1], [ImmCheck<3, ImmCheck0_3>]>;
 }
 
 let TargetGuard = "sve2p1|sme2" in {
@@ -2238,15 +2244,15 @@ let TargetGuard = "sme2" in {
   def SVCVT_F16_X2  : SInst<"svcvt_f16[_f32_x2]", "e2", "f", MergeNone, "aarch64_sve_fcvt_x2", [IsStreaming],[]>;
   def SVCVT_BF16_X2 : SInst<"svcvt_bf16[_f32_x2]", "$2", "f", MergeNone, "aarch64_sve_bfcvt_x2", [IsOverloadNone, IsStreaming],[]>;
 
-  def SVCVT_F32_U32_X2 : SInst<"svcvt_{d}[_u32_x2]", "2.d2.u", "f", MergeNone, "aarch64_sve_fcvtu_x2", [IsStreaming], []>;
-  def SVCVT_U32_F32_X2 : SInst<"svcvt_u32[_{d}_x2]", "2.u2.d", "f", MergeNone, "aarch64_sve_ucvtf_x2", [IsStreaming], []>;
-  def SVCVT_F32_S32_X2 : SInst<"svcvt_{d}[_s32_x2]", "2.d2.x", "f", MergeNone, "aarch64_sve_fcvts_x2", [IsStreaming], []>;
-  def SVCVT_S32_F32_X2 : SInst<"svcvt_s32[_{d}_x2]", "2.x2.d", "f", MergeNone, "aarch64_sve_scvtf_x2", [IsStreaming], []>;
+  def SVCVT_F32_U32_X2 : SInst<"svcvt_{d}[_u32_x2]", "2.d2.u", "f", MergeNone, "aarch64_sve_ucvtf_x2", [IsStreaming], []>;
+  def SVCVT_U32_F32_X2 : SInst<"svcvt_u32[_{d}_x2]", "2.u2.d", "f", MergeNone, "aarch64_sve_fcvtu_x2", [IsStreaming], []>;
+  def SVCVT_F32_S32_X2 : SInst<"svcvt_{d}[_s32_x2]", "2.d2.x", "f", MergeNone, "aarch64_sve_scvtf_x2", [IsStreaming], []>;
+  def SVCVT_S32_F32_X2 : SInst<"svcvt_s32[_{d}_x2]", "2.x2.d", "f", MergeNone, "aarch64_sve_fcvts_x2", [IsStreaming], []>;
 
-  def SVCVT_F32_U32_X4 : SInst<"svcvt_{d}[_u32_x4]", "4.d4.u", "f", MergeNone, "aarch64_sve_fcvtu_x4", [IsStreaming], []>;
-  def SVCVT_U32_F32_X4 : SInst<"svcvt_u32[_{d}_x4]", "4.u4.d", "f", MergeNone, "aarch64_sve_ucvtf_x4", [IsStreaming], []>;
-  def SVCVT_F32_S32_X4 : SInst<"svcvt_{d}[_s32_x4]", "4.d4.x", "f", MergeNone, "aarch64_sve_fcvts_x4", [IsStreaming], []>;
-  def SVCVT_S32_F32_X4 : SInst<"svcvt_s32[_{d}_x4]", "4.x4.d", "f", MergeNone, "aarch64_sve_scvtf_x4", [IsStreaming], []>;
+  def SVCVT_F32_U32_X4 : SInst<"svcvt_{d}[_u32_x4]", "4.d4.u", "f", MergeNone, "aarch64_sve_ucvtf_x4", [IsStreaming], []>;
+  def SVCVT_U32_F32_X4 : SInst<"svcvt_u32[_{d}_x4]", "4.u4.d", "f", MergeNone, "aarch64_sve_fcvtu_x4", [IsStreaming], []>;
+  def SVCVT_F32_S32_X4 : SInst<"svcvt_{d}[_s32_x4]", "4.d4.x", "f", MergeNone, "aarch64_sve_scvtf_x4", [IsStreaming], []>;
+  def SVCVT_S32_F32_X4 : SInst<"svcvt_s32[_{d}_x4]", "4.x4.d", "f", MergeNone, "aarch64_sve_fcvts_x4", [IsStreaming], []>;
 }
 
 //
diff --git a/clang/include/clang/Basic/riscv_vector.td b/clang/include/clang/Basic/riscv_vector.td
index e7d78b03511fe..8bde081052505 100644
--- a/clang/include/clang/Basic/riscv_vector.td
+++ b/clang/include/clang/Basic/riscv_vector.td
@@ -117,7 +117,7 @@ multiclass RVVIndexedLoad {
         defvar eew = eew_list[0];
         defvar eew_type = eew_list[1];
         let Name = op # eew # "_v", IRName = op, MaskedIRName = op # "_mask",
-            RequiredFeatures = !if(!eq(type, "x"), ["ZvfhminOrZvfh"],
+            RequiredFeatures = !if(!eq(type, "x"), ["Zvfhmin"],
                                                    []) in {
           def: RVVOutOp1Builtin<"v", "vPCe" # eew_type # "Uv", type>;
             if !not(IsFloat.val) then {
@@ -128,7 +128,7 @@ multiclass RVVIndexedLoad {
       defvar eew64 = "64";
       defvar eew64_type = "(Log2EEW:6)";
       let Name = op # eew64 # "_v", IRName = op, MaskedIRName = op # "_mask",
-          RequiredFeatures = !if(!eq(type, "x"), ["ZvfhminOrZvfh", "RV64"],
+          RequiredFeatures = !if(!eq(type, "x"), ["Zvfhmin", "RV64"],
                                                  ["RV64"]) in {
           def: RVVOutOp1Builtin<"v", "vPCe" # eew64_type # "Uv", type>;
             if !not(IsFloat.val) then {
@@ -222,7 +222,7 @@ multiclass RVVIndexedStore {
           defvar eew = eew_list[0];
           defvar eew_type = eew_list[1];
           let Name = op # eew  # "_v", IRName = op, MaskedIRName = op # "_mask",
-          RequiredFeatures = !if(!eq(type, "x"), ["ZvfhminOrZvfh"],
+          RequiredFeatures = !if(!eq(type, "x"), ["Zvfhmin"],
                                                  []) in  {
             def : RVVBuiltin<"v", "0Pe" # eew_type # "Uvv", type>;
             if !not(IsFloat.val) then {
@@ -233,7 +233,7 @@ multiclass RVVIndexedStore {
         defvar eew64 = "64";
         defvar eew64_type = "(Log2EEW:6)";
         let Name = op # eew64  # "_v", IRName = op, MaskedIRName = op # "_mask",
-            RequiredFeatures = !if(!eq(type, "x"), ["ZvfhminOrZvfh", "RV64"],
+            RequiredFeatures = !if(!eq(type, "x"), ["Zvfhmin", "RV64"],
                                                    ["RV64"]) in  {
           def : RVVBuiltin<"v", "0Pe" # eew64_type # "Uvv", type>;
           if !not(IsFloat.val) then {
@@ -681,7 +681,7 @@ let HasBuiltinAlias = false,
 def vlm: RVVVLEMaskBuiltin;
 defm vle8: RVVVLEBuiltin<["c"]>;
 defm vle16: RVVVLEBuiltin<["s"]>;
-let Name = "vle16_v", RequiredFeatures = ["ZvfhminOrZvfh"] in
+let Name = "vle16_v", RequiredFeatures = ["Zvfhmin"] in
   defm vle16_h: RVVVLEBuiltin<["x"]>;
 defm vle32: RVVVLEBuiltin<["i","f"]>;
 defm vle64: RVVVLEBuiltin<["l","d"]>;
@@ -689,7 +689,7 @@ defm vle64: RVVVLEBuiltin<["l","d"]>;
 def vsm : RVVVSEMaskBuiltin;
 defm vse8 : RVVVSEBuiltin<["c"]>;
 defm vse16: RVVVSEBuiltin<["s"]>;
-let Name = "vse16_v", RequiredFeatures = ["ZvfhminOrZvfh"] in
+let Name = "vse16_v", RequiredFeatures = ["Zvfhmin"] in
   defm vse16_h: RVVVSEBuiltin<["x"]>;
 defm vse32: RVVVSEBuiltin<["i","f"]>;
 defm vse64: RVVVSEBuiltin<["l","d"]>;
@@ -697,14 +697,14 @@ defm vse64: RVVVSEBuiltin<["l","d"]>;
 // 7.5. Vector Strided Instructions
 defm vlse8: RVVVLSEBuiltin<["c"]>;
 defm vlse16: RVVVLSEBuiltin<["s"]>;
-let Name = "vlse16_v", RequiredFeatures = ["ZvfhminOrZvfh"] in
+let Name = "vlse16_v", RequiredFeatures = ["Zvfhmin"] in
   defm vlse16_h: RVVVLSEBuiltin<["x"]>;
 defm vlse32: RVVVLSEBuiltin<["i","f"]>;
 defm vlse64: RVVVLSEBuiltin<["l","d"]>;
 
 defm vsse8 : RVVVSSEBuiltin<["c"]>;
 defm vsse16: RVVVSSEBuiltin<["s"]>;
-let Name = "vsse16_v", RequiredFeatures = ["ZvfhminOrZvfh"] in
+let Name = "vsse16_v", RequiredFeatures = ["Zvfhmin"] in
   defm vsse16_h: RVVVSSEBuiltin<["x"]>;
 defm vsse32: RVVVSSEBuiltin<["i","f"]>;
 defm vsse64: RVVVSSEBuiltin<["l","d"]>;
@@ -719,7 +719,7 @@ defm : RVVIndexedStore<"vsoxei">;
 // 7.7. Unit-stride Fault-Only-First Loads
 defm vle8ff: RVVVLEFFBuiltin<["c"]>;
 defm vle16ff: RVVVLEFFBuiltin<["s"]>;
-let Name = "vle16ff_v", RequiredFeatures = ["ZvfhminOrZvfh"] in
+let Name = "vle16ff_v", RequiredFeatures = ["Zvfhmin"] in
   defm vle16ff: RVVVLEFFBuiltin<["x"]>;
 defm vle32ff: RVVVLEFFBuiltin<["i", "f"]>;
 defm vle64ff: RVVVLEFFBuiltin<["l", "d"]>;
@@ -738,7 +738,7 @@ multiclass RVVUnitStridedSegLoadTuple {
             IRName = op # nf,
             MaskedIRName = op # nf # "_mask",
             NF = nf,
-            RequiredFeatures = !if(!eq(type, "x"), ["ZvfhminOrZvfh"],
+            RequiredFeatures = !if(!eq(type, "x"), ["Zvfhmin"],
                                                    []),
             ManualCodegen = [{
     {
@@ -800,7 +800,7 @@ multiclass RVVUnitStridedSegStoreTuple {
           MaskedIRName = op # nf # "_mask",
           NF = nf,
           HasMaskedOffOperand = false,
-          RequiredFeatures = !if(!eq(type, "x"), ["ZvfhminOrZvfh"],
+          RequiredFeatures = !if(!eq(type, "x"), ["Zvfhmin"],
                                                  []),
           ManualCodegen = [{
     {
@@ -852,7 +852,7 @@ multiclass RVVUnitStridedSegLoadFFTuple {
             IRName = op # nf # "ff",
             MaskedIRName = op # nf # "ff_mask",
             NF = nf,
-            RequiredFeatures = !if(!eq(type, "x"), ["ZvfhminOrZvfh"],
+            RequiredFeatures = !if(!eq(type, "x"), ["Zvfhmin"],
                                                    []),
             ManualCodegen = [{
     {
@@ -927,7 +927,7 @@ multiclass RVVStridedSegLoadTuple {
             IRName = op # nf,
             MaskedIRName = op # nf # "_mask",
             NF = nf,
-            RequiredFeatures = !if(!eq(type, "x"), ["ZvfhminOrZvfh"],
+            RequiredFeatures = !if(!eq(type, "x"), ["Zvfhmin"],
                                                    []),
             ManualCodegen = [{
     {
@@ -991,7 +991,7 @@ multiclass RVVStridedSegStoreTuple {
             NF = nf,
             HasMaskedOffOperand = false,
             MaskedPolicyScheme = NonePolicy,
-            RequiredFeatures = !if(!eq(type, "x"), ["ZvfhminOrZvfh"],
+            RequiredFeatures = !if(!eq(type, "x"), ["Zvfhmin"],
                                                    []),
             ManualCodegen = [{
     {
@@ -1040,7 +1040,7 @@ multiclass RVVIndexedSegLoadTuple {
             IRName = op # nf,
             MaskedIRName = op # nf # "_mask",
             NF = nf,
-            RequiredFeatures = !if(!eq(type, "x"), ["ZvfhminOrZvfh"],
+            RequiredFeatures = !if(!eq(type, "x"), ["Zvfhmin"],
                                                    []),
             ManualCodegen = [{
     {
@@ -1103,7 +1103,7 @@ multiclass RVVIndexedSegStoreTuple {
             NF = nf,
             HasMaskedOffOperand = false,
             MaskedPolicyScheme = NonePolicy,
-            RequiredFeatures = !if(!eq(type, "x"), ["ZvfhminOrZvfh"],
+            RequiredFeatures = !if(!eq(type, "x"), ["Zvfhmin"],
                                                    []),
             ManualCodegen = [{
     {
@@ -1345,7 +1345,7 @@ let HasMasked = false,
                                    [["v", "Uv", "UvUv"]]>;
     defm vmv_v : RVVOutBuiltinSet<"vmv_v_v", "csilfd",
                                    [["v", "v", "vv"]]>;
-    let RequiredFeatures = ["ZvfhminOrZvfh"] in
+    let RequiredFeatures = ["Zvfhmin"] in
       defm vmv_v : RVVOutBuiltinSet<"vmv_v_v", "x",
                                     [["v", "v", "vv"]]>;
   let SupportOverloading = false in
@@ -1841,7 +1841,7 @@ let HasMasked = false,
     }] in {
   defm vmerge : RVVOutOp1BuiltinSet<"vmerge", "fd",
                                     [["vvm", "v", "vvvm"]]>;
-  let RequiredFeatures = ["ZvfhminOrZvfh"] in
+  let RequiredFeatures = ["Zvfhmin"] in
     defm vmerge : RVVOutOp1BuiltinSet<"vmerge", "x",
                                       [["vvm", "v", "vvvm"]]>;
   defm vfmerge : RVVOutOp1BuiltinSet<"vfmerge", "xfd",
@@ -1869,7 +1869,7 @@ let Log2LMUL = [-3, -2, -1, 0, 1, 2] in {
   def vfwcvt_f_xu_v : RVVConvBuiltin<"Fw", "FwUv", "csi", "vfwcvt_f">;
   def vfwcvt_f_x_v : RVVConvBuiltin<"Fw", "Fwv", "csi", "vfwcvt_f">;
   def vfwcvt_f_f_v : RVVConvBuiltin<"w", "wv", "f", "vfwcvt_f">;
-  let RequiredFeatures = ["ZvfhminOrZvfh"] in
+  let RequiredFeatures = ["Zvfhmin"] in
     def vfwcvt_f_f_v_fp16 : RVVConvBuiltin<"w", "wv", "x", "vfwcvt_f"> {
       let Name = "vfwcvt_f_f_v";
       let IRName = "vfwcvt_f_f_v";
@@ -1966,7 +1966,7 @@ let ManualCodegen = [{
       }
       let OverloadedName = "vfncvt_f" in {
         defm : RVVConvBuiltinSet<"vfncvt_f_f_w", "f", [["v", "vwu"]]>;
-        let RequiredFeatures = ["ZvfhminOrZvfh"] in
+        let RequiredFeatures = ["Zvfhmin"] in
         defm : RVVConvBuiltinSet<"vfncvt_f_f_w", "x", [["v", "vwu"]]>;
       }
     }
@@ -2011,7 +2011,7 @@ let ManualCodegen = [{
     }
     let OverloadedName = "vfncvt_f" in {
       defm : RVVConvBuiltinSet<"vfncvt_f_f_w", "f", [["v", "vw"]]>;
-      let RequiredFeatures = ["ZvfhminOrZvfh"] in
+      let RequiredFeatures = ["Zvfhmin"] in
       defm : RVVConvBuiltinSet<"vfncvt_f_f_w", "x", [["v", "vw"]]>;
     }
   }
@@ -2271,7 +2271,7 @@ let HasMasked = false, HasVL = false, IRName = "" in {
     def vreinterpret_u_f : RVVBuiltin<"FvUv", "UvFv", "il", "Uv">;
     def vreinterpret_f_i : RVVBuiltin<"vFv", "Fvv", "il", "Fv">;
     def vreinterpret_f_u : RVVBuiltin<"UvFv", "FvUv", "il", "Fv">;
-    let RequiredFeatures = ["ZvfhminOrZvfh"] in {
+    let RequiredFeatures = ["Zvfhmin"] in {
       def vreinterpret_i_h : RVVBuiltin<"Fvv", "vFv", "s", "v">;
       def vreinterpret_u_h : RVVBuiltin<"FvUv", "UvFv", "s", "Uv">;
       def vreinterpret_h_i : RVVBuiltin<"vFv", "Fvv", "s", "Fv">;
diff --git a/clang/include/clang/Driver/Options.td b/clang/include/clang/Driver/Options.td
index b6e0550162e31..33ac0ff62538a 100644
--- a/clang/include/clang/Driver/Options.td
+++ b/clang/include/clang/Driver/Options.td
@@ -4538,7 +4538,7 @@ def iquote : JoinedOrSeparate<["-"], "iquote">, Group,
   Visibility<[ClangOption, CC1Option]>,
   HelpText<"Add directory to QUOTE include search path">, MetaVarName<"">;
 def isysroot : JoinedOrSeparate<["-"], "isysroot">, Group,
-  Visibility<[ClangOption, CC1Option]>,
+  Visibility<[ClangOption, CC1Option, FlangOption]>,
   HelpText<"Set the system root directory (usually /)">, MetaVarName<"">,
   MarshallingInfoString, [{"/"}]>;
 def isystem : JoinedOrSeparate<["-"], "isystem">, Group,
@@ -4663,8 +4663,8 @@ def mcmodel_EQ : Joined<["-"], "mcmodel=">, Group,
   Visibility<[ClangOption, CC1Option]>,
   MarshallingInfoString, [{"default"}]>;
 def mlarge_data_threshold_EQ : Joined<["-"], "mlarge-data-threshold=">, Group,
-  Visibility<[ClangOption, CC1Option]>,
-  MarshallingInfoInt, "65535">;
+  Flags<[TargetSpecific]>, Visibility<[ClangOption, CC1Option]>,
+  MarshallingInfoInt, "0">;
 def mtls_size_EQ : Joined<["-"], "mtls-size=">, Group,
   Visibility<[ClangOption, CC1Option]>,
   HelpText<"Specify bit size of immediate TLS offsets (AArch64 ELF only): "
@@ -5443,7 +5443,7 @@ def module_file_info : Flag<["-"], "module-file-info">, Flags<[]>,
   HelpText<"Provide information about a particular module file">;
 def mthumb : Flag<["-"], "mthumb">, Group;
 def mtune_EQ : Joined<["-"], "mtune=">, Group,
-  HelpText<"Only supported on AArch64, PowerPC, RISC-V, SystemZ, and X86">;
+  HelpText<"Only supported on AArch64, PowerPC, RISC-V, SPARC, SystemZ, and X86">;
 def multi__module : Flag<["-"], "multi_module">;
 def multiply__defined__unused : Separate<["-"], "multiply_defined_unused">;
 def multiply__defined : Separate<["-"], "multiply_defined">;
@@ -5580,7 +5580,8 @@ def pthreads : Flag<["-"], "pthreads">;
 defm pthread : BoolOption<"", "pthread",
   LangOpts<"POSIXThreads">, DefaultFalse,
   PosFlag,
-  NegFlag, BothFlags<[], [ClangOption, CC1Option]>>;
+  NegFlag,
+  BothFlags<[], [ClangOption, CC1Option, FlangOption, FC1Option]>>;
 def pie : Flag<["-"], "pie">, Group;
 def static_pie : Flag<["-"], "static-pie">, Group;
 def read__only__relocs : Separate<["-"], "read_only_relocs">;
@@ -6612,6 +6613,11 @@ def fsycl_fp32_prec_sqrt : Flag<["-"], "fsycl-fp32-prec-sqrt">, Group,
   MarshallingInfoFlag>;
 
+// OS-specific options
+let Flags = [TargetSpecific] in {
+defm android_pad_segment : BooleanFFlag<"android-pad-segment">, Group;
+} // let Flags = [TargetSpecific]
+
 //===----------------------------------------------------------------------===//
 // FLangOption + NoXarchOption
 //===----------------------------------------------------------------------===//
@@ -7327,6 +7333,8 @@ def mbranch_target_enforce : Flag<["-"], "mbranch-target-enforce">,
   MarshallingInfoFlag>;
 def mbranch_protection_pauth_lr : Flag<["-"], "mbranch-protection-pauth-lr">,
   MarshallingInfoFlag>;
+def mguarded_control_stack : Flag<["-"], "mguarded-control-stack">,
+  MarshallingInfoFlag>;
 def fno_dllexport_inlines : Flag<["-"], "fno-dllexport-inlines">,
   MarshallingInfoNegativeFlag>;
 def cfguard_no_checks : Flag<["-"], "cfguard-no-checks">,
diff --git a/clang/include/clang/Format/Format.h b/clang/include/clang/Format/Format.h
index 59b645ecab715..6fd7947bd2179 100644
--- a/clang/include/clang/Format/Format.h
+++ b/clang/include/clang/Format/Format.h
@@ -3055,6 +3055,7 @@ struct FormatStyle {
   bool isProto() const {
     return Language == LK_Proto || Language == LK_TextProto;
   }
+  bool isTableGen() const { return Language == LK_TableGen; }
 
   /// Language, this format style is targeted at.
   /// \version 3.5
@@ -3397,6 +3398,10 @@ struct FormatStyle {
   /// \version 14
   unsigned PenaltyBreakOpenParenthesis;
 
+  /// The penalty for breaking after ``::``.
+  /// \version 18
+  unsigned PenaltyBreakScopeResolution;
+
   /// The penalty for each line break introduced inside a string literal.
   /// \version 3.7
   unsigned PenaltyBreakString;
@@ -4872,6 +4877,7 @@ struct FormatStyle {
            PenaltyBreakComment == R.PenaltyBreakComment &&
            PenaltyBreakFirstLessLess == R.PenaltyBreakFirstLessLess &&
            PenaltyBreakOpenParenthesis == R.PenaltyBreakOpenParenthesis &&
+           PenaltyBreakScopeResolution == R.PenaltyBreakScopeResolution &&
            PenaltyBreakString == R.PenaltyBreakString &&
            PenaltyBreakTemplateDeclaration ==
                R.PenaltyBreakTemplateDeclaration &&
diff --git a/clang/include/clang/Interpreter/Interpreter.h b/clang/include/clang/Interpreter/Interpreter.h
index 01858dfcc90ac..292fa566ae703 100644
--- a/clang/include/clang/Interpreter/Interpreter.h
+++ b/clang/include/clang/Interpreter/Interpreter.h
@@ -129,7 +129,7 @@ class Interpreter {
   llvm::Expected
   getSymbolAddressFromLinkerName(llvm::StringRef LinkerName) const;
 
-  enum InterfaceKind { NoAlloc, WithAlloc, CopyArray };
+  enum InterfaceKind { NoAlloc, WithAlloc, CopyArray, NewTag };
 
   const llvm::SmallVectorImpl &getValuePrintingInfo() const {
     return ValuePrintingInfo;
@@ -144,7 +144,7 @@ class Interpreter {
 
   llvm::DenseMap Dtors;
 
-  llvm::SmallVector ValuePrintingInfo;
+  llvm::SmallVector ValuePrintingInfo;
 };
 } // namespace clang
 
diff --git a/clang/include/clang/Lex/Lexer.h b/clang/include/clang/Lex/Lexer.h
index 899e665e74546..b6ecc7e5ded9e 100644
--- a/clang/include/clang/Lex/Lexer.h
+++ b/clang/include/clang/Lex/Lexer.h
@@ -198,11 +198,11 @@ class Lexer : public PreprocessorLexer {
   /// from.  Currently this is only used by _Pragma handling.
   SourceLocation getFileLoc() const { return FileLoc; }
 
-private:
   /// Lex - Return the next token in the file.  If this is the end of file, it
   /// return the tok::eof token.  This implicitly involves the preprocessor.
   bool Lex(Token &Result);
 
+private:
   /// Called when the preprocessor is in 'dependency scanning lexing mode'.
   bool LexDependencyDirectiveToken(Token &Result);
 
diff --git a/clang/include/clang/Parse/Parser.h b/clang/include/clang/Parse/Parser.h
index edc65811f6e77..dc32580f42add 100644
--- a/clang/include/clang/Parse/Parser.h
+++ b/clang/include/clang/Parse/Parser.h
@@ -13,6 +13,7 @@
 #ifndef LLVM_CLANG_PARSE_PARSER_H
 #define LLVM_CLANG_PARSE_PARSER_H
 
+#include "clang/Basic/OpenACCKinds.h"
 #include "clang/Basic/OperatorPrecedence.h"
 #include "clang/Lex/CodeCompletionHandler.h"
 #include "clang/Lex/Preprocessor.h"
@@ -3569,7 +3570,21 @@ class Parser : public CodeCompletionHandler {
   void ParseOpenACCCacheVarList();
   /// Parses a single variable in a variable list for OpenACC.
   bool ParseOpenACCVar();
+  /// Parses the variable list for the variety of clauses that take a var-list,
+  /// including the optional Special Token listed for some,based on clause type.
+  bool ParseOpenACCClauseVarList(OpenACCClauseKind Kind);
+  /// Parses any parameters for an OpenACC Clause, including required/optional
+  /// parens.
+  bool ParseOpenACCClauseParams(OpenACCDirectiveKind DirKind,
+                                OpenACCClauseKind Kind);
+  /// Parses a single clause in a clause-list for OpenACC.
+  bool ParseOpenACCClause(OpenACCDirectiveKind DirKind);
+  /// Parses the clause-list for an OpenACC directive.
+  void ParseOpenACCClauseList(OpenACCDirectiveKind DirKind);
   bool ParseOpenACCWaitArgument();
+  /// Parses the clause of the 'bind' argument, which can be a string literal or
+  /// an ID expression.
+  ExprResult ParseOpenACCBindClauseArgument();
 
 private:
   //===--------------------------------------------------------------------===//
@@ -3645,6 +3660,18 @@ class Parser : public CodeCompletionHandler {
   ParseConceptDefinition(const ParsedTemplateInfo &TemplateInfo,
                          SourceLocation &DeclEnd);
 
+  /// Parse the given string as a type.
+  ///
+  /// This is a dangerous utility function currently employed only by API notes.
+  /// It is not a general entry-point for safely parsing types from strings.
+  ///
+  /// \param TypeStr The string to be parsed as a type.
+  /// \param Context The name of the context in which this string is being
+  /// parsed, which will be used in diagnostics.
+  /// \param IncludeLoc The location at which this parse was triggered.
+  TypeResult ParseTypeFromString(StringRef TypeStr, StringRef Context,
+                                 SourceLocation IncludeLoc);
+
   //===--------------------------------------------------------------------===//
   // Modules
   DeclGroupPtrTy ParseModuleDecl(Sema::ModuleImportState &ImportState);
diff --git a/clang/include/clang/Sema/Sema.h b/clang/include/clang/Sema/Sema.h
index 01c8075a3e427..1a23b14f08c5e 100644
--- a/clang/include/clang/Sema/Sema.h
+++ b/clang/include/clang/Sema/Sema.h
@@ -1143,6 +1143,10 @@ class Sema final {
     OpaqueParser = P;
   }
 
+  /// Callback to the parser to parse a type expressed as a string.
+  std::function
+      ParseTypeFromStringCallback;
+
   class DelayedDiagnostics;
 
   class DelayedDiagnosticsState {
@@ -2378,6 +2382,10 @@ class Sema final {
 
   bool CheckFunctionReturnType(QualType T, SourceLocation Loc);
 
+  /// Check an argument list for placeholders that we won't try to
+  /// handle later.
+  bool CheckArgsForPlaceholders(MultiExprArg args);
+
   /// Build a function type.
   ///
   /// This routine checks the function type according to C++ rules and
@@ -2423,7 +2431,7 @@ class Sema final {
                          SourceLocation Loc);
   QualType BuildBitIntType(bool IsUnsigned, Expr *BitWidth, SourceLocation Loc);
 
-  TypeSourceInfo *GetTypeForDeclarator(Declarator &D, Scope *S);
+  TypeSourceInfo *GetTypeForDeclarator(Declarator &D);
   TypeSourceInfo *GetTypeForDeclaratorCast(Declarator &D, QualType FromTy);
 
   /// Package the given type and TSI into a ParsedType.
@@ -2464,7 +2472,7 @@ class Sema final {
       SourceLocation TargetLoc, const FunctionProtoType *Source,
       bool SkipSourceFirstParameter, SourceLocation SourceLoc);
 
-  TypeResult ActOnTypeName(Scope *S, Declarator &D);
+  TypeResult ActOnTypeName(Declarator &D);
 
   /// The parser has parsed the context-sensitive type 'instancetype'
   /// in an Objective-C message declaration. Return the appropriate type.
@@ -3287,6 +3295,9 @@ class Sema final {
   ParmVarDecl *BuildParmVarDeclForTypedef(DeclContext *DC,
                                           SourceLocation Loc,
                                           QualType T);
+  QualType AdjustParameterTypeForObjCAutoRefCount(QualType T,
+                                                  SourceLocation NameLoc,
+                                                  TypeSourceInfo *TSInfo);
   ParmVarDecl *CheckParameter(DeclContext *DC, SourceLocation StartLoc,
                               SourceLocation NameLoc, IdentifierInfo *Name,
                               QualType T, TypeSourceInfo *TSInfo,
@@ -5087,6 +5098,29 @@ class Sema final {
   /// Valid types should not have multiple attributes with different CCs.
   const AttributedType *getCallingConvAttributedType(QualType T) const;
 
+  /// Check whether a nullability type specifier can be added to the given
+  /// type through some means not written in source (e.g. API notes).
+  ///
+  /// \param Type The type to which the nullability specifier will be
+  /// added. On success, this type will be updated appropriately.
+  ///
+  /// \param Nullability The nullability specifier to add.
+  ///
+  /// \param DiagLoc The location to use for diagnostics.
+  ///
+  /// \param AllowArrayTypes Whether to accept nullability specifiers on an
+  /// array type (e.g., because it will decay to a pointer).
+  ///
+  /// \param OverrideExisting Whether to override an existing, locally-specified
+  /// nullability specifier rather than complaining about the conflict.
+  ///
+  /// \returns true if nullability cannot be applied, false otherwise.
+  bool CheckImplicitNullabilityTypeSpecifier(QualType &Type,
+                                             NullabilityKind Nullability,
+                                             SourceLocation DiagLoc,
+                                             bool AllowArrayTypes,
+                                             bool OverrideExisting);
+
   /// Process the attributes before creating an attributed statement. Returns
   /// the semantic attributes that have been processed.
   void ProcessStmtAttributes(Stmt *Stmt, const ParsedAttributes &InAttrs,
@@ -7389,15 +7423,7 @@ class Sema final {
                                  NestedNameSpecInfo &IdInfo,
                                  bool EnteringContext);
 
-  /// The kind of conversion to check for. Either all attributes must match exactly,
-  /// or the converted type may add/drop '__arm_preserves_za'.
-  enum class AArch64SMECallConversionKind {
-    MatchExactly,
-    MayAddPreservesZA,
-    MayDropPreservesZA,
-  };
-  bool IsInvalidSMECallConversion(QualType FromType, QualType ToType,
-                                  AArch64SMECallConversionKind C);
+  bool IsInvalidSMECallConversion(QualType FromType, QualType ToType);
 
   /// The parser has parsed a nested-name-specifier
   /// 'template[opt] template-name < template-args >::'.
@@ -11645,6 +11671,11 @@ class Sema final {
   VarDecl *buildCoroutinePromise(SourceLocation Loc);
   void CheckCompletedCoroutineBody(FunctionDecl *FD, Stmt *&Body);
 
+  // Heuristically tells if the function is `get_return_object` member of a
+  // coroutine promise_type by matching the function name.
+  static bool CanBeGetReturnObject(const FunctionDecl *FD);
+  static bool CanBeGetReturnTypeOnAllocFailure(const FunctionDecl *FD);
+
   // As a clang extension, enforces that a non-coroutine function must be marked
   // with [[clang::coro_wrapper]] if it returns a type marked with
   // [[clang::coro_return_type]].
diff --git a/clang/include/clang/Support/RISCVVIntrinsicUtils.h b/clang/include/clang/Support/RISCVVIntrinsicUtils.h
index c525d3443331e..05a5e02e13900 100644
--- a/clang/include/clang/Support/RISCVVIntrinsicUtils.h
+++ b/clang/include/clang/Support/RISCVVIntrinsicUtils.h
@@ -488,7 +488,7 @@ class RVVIntrinsic {
 enum RVVRequire : uint32_t {
   RVV_REQ_None = 0,
   RVV_REQ_RV64 = 1 << 0,
-  RVV_REQ_ZvfhminOrZvfh = 1 << 1,
+  RVV_REQ_Zvfhmin = 1 << 1,
   RVV_REQ_Xsfvcp = 1 << 2,
   RVV_REQ_Xsfvfnrclipxfqf = 1 << 3,
   RVV_REQ_Xsfvfwmaccqqq = 1 << 4,
diff --git a/clang/lib/AST/ASTContext.cpp b/clang/lib/AST/ASTContext.cpp
index c5d2b4d2cdd55..e130a1132c3de 100644
--- a/clang/lib/AST/ASTContext.cpp
+++ b/clang/lib/AST/ASTContext.cpp
@@ -4496,10 +4496,11 @@ QualType ASTContext::getFunctionTypeInternal(
       EPI.ExceptionSpec.Type, EPI.ExceptionSpec.Exceptions.size());
   size_t Size = FunctionProtoType::totalSizeToAlloc<
       QualType, SourceLocation, FunctionType::FunctionTypeExtraBitfields,
-      FunctionType::ExceptionType, Expr *, FunctionDecl *,
-      FunctionProtoType::ExtParameterInfo, Qualifiers>(
+      FunctionType::FunctionTypeArmAttributes, FunctionType::ExceptionType,
+      Expr *, FunctionDecl *, FunctionProtoType::ExtParameterInfo, Qualifiers>(
       NumArgs, EPI.Variadic, EPI.requiresFunctionProtoTypeExtraBitfields(),
-      ESH.NumExceptionType, ESH.NumExprPtr, ESH.NumFunctionDeclPtr,
+      EPI.requiresFunctionProtoTypeArmAttributes(), ESH.NumExceptionType,
+      ESH.NumExprPtr, ESH.NumFunctionDeclPtr,
       EPI.ExtParameterInfos ? NumArgs : 0,
       EPI.TypeQuals.hasNonFastQualifiers() ? 1 : 0);
 
diff --git a/clang/lib/AST/ASTImporter.cpp b/clang/lib/AST/ASTImporter.cpp
index 0d3927dcf0e08..d1dff9e28890f 100644
--- a/clang/lib/AST/ASTImporter.cpp
+++ b/clang/lib/AST/ASTImporter.cpp
@@ -3902,7 +3902,7 @@ ExpectedDecl ASTNodeImporter::VisitFunctionDecl(FunctionDecl *D) {
   ToFunction->setLexicalDeclContext(LexicalDC);
   ToFunction->setVirtualAsWritten(D->isVirtualAsWritten());
   ToFunction->setTrivial(D->isTrivial());
-  ToFunction->setPure(D->isPure());
+  ToFunction->setIsPureVirtual(D->isPureVirtual());
   ToFunction->setDefaulted(D->isDefaulted());
   ToFunction->setExplicitlyDefaulted(D->isExplicitlyDefaulted());
   ToFunction->setDeletedAsWritten(D->isDeletedAsWritten());
@@ -5935,15 +5935,22 @@ ExpectedDecl ASTNodeImporter::VisitClassTemplateDecl(ClassTemplateDecl *D) {
   if (ToD)
     return ToD;
 
-  bool IsFriendTemplate = D->getFriendObjectKind() != Decl::FOK_None;
-  bool IsDependentContext = DC != LexicalDC ? LexicalDC->isDependentContext()
-                                            : DC->isDependentContext();
-  bool DependentFriend = IsFriendTemplate && IsDependentContext;
+  // Should check if a declaration is friend in a dependent context.
+  // Such templates are not linked together in a declaration chain.
+  // The ASTImporter strategy is to map existing forward declarations to
+  // imported ones only if strictly necessary, otherwise import these as new
+  // forward declarations. In case of the "dependent friend" declarations, new
+  // declarations are created, but not linked in a declaration chain.
+  auto IsDependentFriend = [](ClassTemplateDecl *TD) {
+    return TD->getFriendObjectKind() != Decl::FOK_None &&
+           TD->getLexicalDeclContext()->isDependentContext();
+  };
+  bool DependentFriend = IsDependentFriend(D);
 
   ClassTemplateDecl *FoundByLookup = nullptr;
 
   // We may already have a template of the same name; try to find and match it.
-  if (!DependentFriend && !DC->isFunctionOrMethod()) {
+  if (!DC->isFunctionOrMethod()) {
     SmallVector ConflictingDecls;
     auto FoundDecls = Importer.findDeclsInToCtx(DC, Name);
     for (auto *FoundDecl : FoundDecls) {
@@ -5959,10 +5966,13 @@ ExpectedDecl ASTNodeImporter::VisitClassTemplateDecl(ClassTemplateDecl *D) {
 
         // FIXME: sufficient conditon for 'IgnoreTemplateParmDepth'?
         bool IgnoreTemplateParmDepth =
-            FoundTemplate->getFriendObjectKind() != Decl::FOK_None &&
-            !D->specializations().empty();
+            (FoundTemplate->getFriendObjectKind() != Decl::FOK_None) !=
+            (D->getFriendObjectKind() != Decl::FOK_None);
         if (IsStructuralMatch(D, FoundTemplate, /*Complain=*/true,
                               IgnoreTemplateParmDepth)) {
+          if (DependentFriend || IsDependentFriend(FoundTemplate))
+            continue;
+
           ClassTemplateDecl *TemplateWithDef =
               getTemplateDefinition(FoundTemplate);
           if (D->isThisDeclarationADefinition() && TemplateWithDef)
@@ -6276,17 +6286,21 @@ ExpectedDecl ASTNodeImporter::VisitVarTemplateDecl(VarTemplateDecl *D) {
                                               D->getTemplatedDecl()))
         continue;
       if (IsStructuralMatch(D, FoundTemplate)) {
-        // The Decl in the "From" context has a definition, but in the
-        // "To" context we already have a definition.
+        // FIXME Check for ODR error if the two definitions have
+        // different initializers?
         VarTemplateDecl *FoundDef = getTemplateDefinition(FoundTemplate);
-        if (D->isThisDeclarationADefinition() && FoundDef)
-          // FIXME Check for ODR error if the two definitions have
-          // different initializers?
-          return Importer.MapImported(D, FoundDef);
-        if (FoundTemplate->getDeclContext()->isRecord() &&
-            D->getDeclContext()->isRecord())
-          return Importer.MapImported(D, FoundTemplate);
-
+        if (D->getDeclContext()->isRecord()) {
+          assert(FoundTemplate->getDeclContext()->isRecord() &&
+                 "Member variable template imported as non-member, "
+                 "inconsistent imported AST?");
+          if (FoundDef)
+            return Importer.MapImported(D, FoundDef);
+          if (!D->isThisDeclarationADefinition())
+            return Importer.MapImported(D, FoundTemplate);
+        } else {
+          if (FoundDef && D->isThisDeclarationADefinition())
+            return Importer.MapImported(D, FoundDef);
+        }
         FoundByLookup = FoundTemplate;
         break;
       }
diff --git a/clang/lib/AST/ASTStructuralEquivalence.cpp b/clang/lib/AST/ASTStructuralEquivalence.cpp
index a9e0d1698a917..5103fc86a8005 100644
--- a/clang/lib/AST/ASTStructuralEquivalence.cpp
+++ b/clang/lib/AST/ASTStructuralEquivalence.cpp
@@ -98,6 +98,8 @@ static bool IsStructurallyEquivalent(StructuralEquivalenceContext &Context,
                                      QualType T1, QualType T2);
 static bool IsStructurallyEquivalent(StructuralEquivalenceContext &Context,
                                      Decl *D1, Decl *D2);
+static bool IsStructurallyEquivalent(StructuralEquivalenceContext &Context,
+                                     const Stmt *S1, const Stmt *S2);
 static bool IsStructurallyEquivalent(StructuralEquivalenceContext &Context,
                                      const TemplateArgument &Arg1,
                                      const TemplateArgument &Arg2);
@@ -437,12 +439,67 @@ class StmtComparer {
 };
 } // namespace
 
+static bool IsStructurallyEquivalent(StructuralEquivalenceContext &Context,
+                                     const UnaryOperator *E1,
+                                     const CXXOperatorCallExpr *E2) {
+  return UnaryOperator::getOverloadedOperator(E1->getOpcode()) ==
+             E2->getOperator() &&
+         IsStructurallyEquivalent(Context, E1->getSubExpr(), E2->getArg(0));
+}
+
+static bool IsStructurallyEquivalent(StructuralEquivalenceContext &Context,
+                                     const CXXOperatorCallExpr *E1,
+                                     const UnaryOperator *E2) {
+  return E1->getOperator() ==
+             UnaryOperator::getOverloadedOperator(E2->getOpcode()) &&
+         IsStructurallyEquivalent(Context, E1->getArg(0), E2->getSubExpr());
+}
+
+static bool IsStructurallyEquivalent(StructuralEquivalenceContext &Context,
+                                     const BinaryOperator *E1,
+                                     const CXXOperatorCallExpr *E2) {
+  return BinaryOperator::getOverloadedOperator(E1->getOpcode()) ==
+             E2->getOperator() &&
+         IsStructurallyEquivalent(Context, E1->getLHS(), E2->getArg(0)) &&
+         IsStructurallyEquivalent(Context, E1->getRHS(), E2->getArg(1));
+}
+
+static bool IsStructurallyEquivalent(StructuralEquivalenceContext &Context,
+                                     const CXXOperatorCallExpr *E1,
+                                     const BinaryOperator *E2) {
+  return E1->getOperator() ==
+             BinaryOperator::getOverloadedOperator(E2->getOpcode()) &&
+         IsStructurallyEquivalent(Context, E1->getArg(0), E2->getLHS()) &&
+         IsStructurallyEquivalent(Context, E1->getArg(1), E2->getRHS());
+}
+
 /// Determine structural equivalence of two statements.
 static bool IsStructurallyEquivalent(StructuralEquivalenceContext &Context,
                                      const Stmt *S1, const Stmt *S2) {
   if (!S1 || !S2)
     return S1 == S2;
 
+  // Check for statements with similar syntax but different AST.
+  // A UnaryOperator node is more lightweight than a CXXOperatorCallExpr node.
+  // The more heavyweight node is only created if the definition-time name
+  // lookup had any results. The lookup results are stored CXXOperatorCallExpr
+  // only. The lookup results can be different in a "From" and "To" AST even if
+  // the compared structure is otherwise equivalent. For this reason we must
+  // treat a similar unary/binary operator node and CXXOperatorCall node as
+  // equivalent.
+  if (const auto *E2CXXOperatorCall = dyn_cast(S2)) {
+    if (const auto *E1Unary = dyn_cast(S1))
+      return IsStructurallyEquivalent(Context, E1Unary, E2CXXOperatorCall);
+    if (const auto *E1Binary = dyn_cast(S1))
+      return IsStructurallyEquivalent(Context, E1Binary, E2CXXOperatorCall);
+  }
+  if (const auto *E1CXXOperatorCall = dyn_cast(S1)) {
+    if (const auto *E2Unary = dyn_cast(S2))
+      return IsStructurallyEquivalent(Context, E1CXXOperatorCall, E2Unary);
+    if (const auto *E2Binary = dyn_cast(S2))
+      return IsStructurallyEquivalent(Context, E1CXXOperatorCall, E2Binary);
+  }
+
   // Compare the statements itself.
   StmtComparer Comparer(Context);
   if (!Comparer.IsEquivalent(S1, S2))
@@ -1410,7 +1467,7 @@ static bool IsStructurallyEquivalent(StructuralEquivalenceContext &Context,
       Method1->isConst() == Method2->isConst() &&
       Method1->isVolatile() == Method2->isVolatile() &&
       Method1->isVirtual() == Method2->isVirtual() &&
-      Method1->isPure() == Method2->isPure() &&
+      Method1->isPureVirtual() == Method2->isPureVirtual() &&
       Method1->isDefaulted() == Method2->isDefaulted() &&
       Method1->isDeleted() == Method2->isDeleted();
   if (!PropertiesEqual)
diff --git a/clang/lib/AST/Decl.cpp b/clang/lib/AST/Decl.cpp
index 824f0767d6a64..f9aa9ecd25982 100644
--- a/clang/lib/AST/Decl.cpp
+++ b/clang/lib/AST/Decl.cpp
@@ -3040,7 +3040,7 @@ FunctionDecl::FunctionDecl(Kind DK, ASTContext &C, DeclContext *DC,
   FunctionDeclBits.IsInline = isInlineSpecified;
   FunctionDeclBits.IsInlineSpecified = isInlineSpecified;
   FunctionDeclBits.IsVirtualAsWritten = false;
-  FunctionDeclBits.IsPure = false;
+  FunctionDeclBits.IsPureVirtual = false;
   FunctionDeclBits.HasInheritedPrototype = false;
   FunctionDeclBits.HasWrittenPrototype = true;
   FunctionDeclBits.IsDeleted = false;
@@ -3207,8 +3207,8 @@ void FunctionDecl::setBody(Stmt *B) {
     EndRangeLoc = B->getEndLoc();
 }
 
-void FunctionDecl::setPure(bool P) {
-  FunctionDeclBits.IsPure = P;
+void FunctionDecl::setIsPureVirtual(bool P) {
+  FunctionDeclBits.IsPureVirtual = P;
   if (P)
     if (auto *Parent = dyn_cast(getDeclContext()))
       Parent->markedVirtualFunctionPure();
@@ -5373,16 +5373,23 @@ void CapturedDecl::setBody(Stmt *B) { BodyAndNothrow.setPointer(B); }
 bool CapturedDecl::isNothrow() const { return BodyAndNothrow.getInt(); }
 void CapturedDecl::setNothrow(bool Nothrow) { BodyAndNothrow.setInt(Nothrow); }
 
+EnumConstantDecl::EnumConstantDecl(const ASTContext &C, DeclContext *DC,
+                                   SourceLocation L, IdentifierInfo *Id,
+                                   QualType T, Expr *E, const llvm::APSInt &V)
+    : ValueDecl(EnumConstant, DC, L, Id, T), Init((Stmt *)E) {
+  setInitVal(C, V);
+}
+
 EnumConstantDecl *EnumConstantDecl::Create(ASTContext &C, EnumDecl *CD,
                                            SourceLocation L,
                                            IdentifierInfo *Id, QualType T,
                                            Expr *E, const llvm::APSInt &V) {
-  return new (C, CD) EnumConstantDecl(CD, L, Id, T, E, V);
+  return new (C, CD) EnumConstantDecl(C, CD, L, Id, T, E, V);
 }
 
 EnumConstantDecl *
 EnumConstantDecl::CreateDeserialized(ASTContext &C, unsigned ID) {
-  return new (C, ID) EnumConstantDecl(nullptr, SourceLocation(), nullptr,
+  return new (C, ID) EnumConstantDecl(C, nullptr, SourceLocation(), nullptr,
                                       QualType(), nullptr, llvm::APSInt());
 }
 
diff --git a/clang/lib/AST/DeclCXX.cpp b/clang/lib/AST/DeclCXX.cpp
index 98b0a6dc28ea2..117e802dae2d9 100644
--- a/clang/lib/AST/DeclCXX.cpp
+++ b/clang/lib/AST/DeclCXX.cpp
@@ -1383,6 +1383,31 @@ void CXXRecordDecl::addedMember(Decl *D) {
   }
 }
 
+bool CXXRecordDecl::isLiteral() const {
+  const LangOptions &LangOpts = getLangOpts();
+  if (!(LangOpts.CPlusPlus20 ? hasConstexprDestructor()
+                             : hasTrivialDestructor()))
+    return false;
+
+  if (hasNonLiteralTypeFieldsOrBases()) {
+    // CWG2598
+    // is an aggregate union type that has either no variant
+    // members or at least one variant member of non-volatile literal type,
+    if (!isUnion())
+      return false;
+    bool HasAtLeastOneLiteralMember =
+        fields().empty() || any_of(fields(), [this](const FieldDecl *D) {
+          return !D->getType().isVolatileQualified() &&
+                 D->getType()->isLiteralType(getASTContext());
+        });
+    if (!HasAtLeastOneLiteralMember)
+      return false;
+  }
+
+  return isAggregate() || (isLambda() && LangOpts.CPlusPlus17) ||
+         hasConstexprNonCopyMoveConstructor() || hasTrivialDefaultConstructor();
+}
+
 void CXXRecordDecl::addedSelectedDestructor(CXXDestructorDecl *DD) {
   DD->setIneligibleOrNotSelected(false);
   addedEligibleSpecialMemberFunction(DD, SMF_Destructor);
@@ -2054,7 +2079,7 @@ void CXXRecordDecl::completeDefinition(CXXFinalOverriderMap *FinalOverriders) {
         //   A class is abstract if it contains or inherits at least one
         //   pure virtual function for which the final overrider is pure
         //   virtual.
-        if (SO->second.front().Method->isPure()) {
+        if (SO->second.front().Method->isPureVirtual()) {
           data().Abstract = true;
           Done = true;
           break;
@@ -2273,7 +2298,7 @@ CXXMethodDecl *CXXMethodDecl::getDevirtualizedMethod(const Expr *Base,
   // If the member function is marked 'final', we know that it can't be
   // overridden and can therefore devirtualize it unless it's pure virtual.
   if (hasAttr())
-    return isPure() ? nullptr : this;
+    return isPureVirtual() ? nullptr : this;
 
   // If Base is unknown, we cannot devirtualize.
   if (!Base)
@@ -2302,7 +2327,7 @@ CXXMethodDecl *CXXMethodDecl::getDevirtualizedMethod(const Expr *Base,
   // If that method is pure virtual, we can't devirtualize. If this code is
   // reached, the result would be UB, not a direct call to the derived class
   // function, and we can't assume the derived class function is defined.
-  if (DevirtualizedMethod->isPure())
+  if (DevirtualizedMethod->isPureVirtual())
     return nullptr;
 
   // If that method is marked final, we can devirtualize it.
diff --git a/clang/lib/AST/DeclPrinter.cpp b/clang/lib/AST/DeclPrinter.cpp
index 42d796698be42..b6cdac9842392 100644
--- a/clang/lib/AST/DeclPrinter.cpp
+++ b/clang/lib/AST/DeclPrinter.cpp
@@ -877,7 +877,7 @@ void DeclPrinter::VisitFunctionDecl(FunctionDecl *D) {
 
   prettyPrintAttributes(D, Out, AttrPrintLoc::Right);
 
-  if (D->isPure())
+  if (D->isPureVirtual())
     Out << " = 0";
   else if (D->isDeletedAsWritten())
     Out << " = delete";
diff --git a/clang/lib/AST/ExprConstant.cpp b/clang/lib/AST/ExprConstant.cpp
index b5ad243ccd0b6..c128d7159c9a9 100644
--- a/clang/lib/AST/ExprConstant.cpp
+++ b/clang/lib/AST/ExprConstant.cpp
@@ -5839,7 +5839,7 @@ static const CXXMethodDecl *HandleVirtualDispatch(
   // C++2a [class.abstract]p6:
   //   the effect of making a virtual call to a pure virtual function [...] is
   //   undefined
-  if (Callee->isPure()) {
+  if (Callee->isPureVirtual()) {
     Info.FFDiag(E, diag::note_constexpr_pure_virtual_call, 1) << Callee;
     Info.Note(Callee->getLocation(), diag::note_declared_at);
     return nullptr;
@@ -15784,10 +15784,22 @@ bool Expr::EvaluateAsInitializer(APValue &Value, const ASTContext &Ctx,
     LValue LVal;
     LVal.set(VD);
 
-    if (!EvaluateInPlace(Value, Info, LVal, this,
-                         /*AllowNonLiteralTypes=*/true) ||
-        EStatus.HasSideEffects)
-      return false;
+    {
+      // C++23 [intro.execution]/p5
+      // A full-expression is ... an init-declarator ([dcl.decl]) or a
+      // mem-initializer.
+      // So we need to make sure temporary objects are destroyed after having
+      // evaluated the expression (per C++23 [class.temporary]/p4).
+      //
+      // FIXME: Otherwise this may break test/Modules/pr68702.cpp because the
+      // serialization code calls ParmVarDecl::getDefaultArg() which strips the
+      // outermost FullExpr, such as ExprWithCleanups.
+      FullExpressionRAII Scope(Info);
+      if (!EvaluateInPlace(Value, Info, LVal, this,
+                           /*AllowNonLiteralTypes=*/true) ||
+          EStatus.HasSideEffects)
+        return false;
+    }
 
     // At this point, any lifetime-extended temporaries are completely
     // initialized.
diff --git a/clang/lib/AST/Interp/ByteCodeExprGen.cpp b/clang/lib/AST/Interp/ByteCodeExprGen.cpp
index 7f8bbe7873248..138ffed392fca 100644
--- a/clang/lib/AST/Interp/ByteCodeExprGen.cpp
+++ b/clang/lib/AST/Interp/ByteCodeExprGen.cpp
@@ -287,6 +287,10 @@ bool ByteCodeExprGen::VisitCastExpr(const CastExpr *CE) {
     return true;
   }
 
+  case CK_IntegralComplexToReal:
+  case CK_FloatingComplexToReal:
+    return this->emitComplexReal(SubExpr);
+
   case CK_ToVoid:
     return discard(SubExpr);
 
@@ -2030,7 +2034,7 @@ bool ByteCodeExprGen::dereference(
   }
 
   if (LV->getType()->isAnyComplexType())
-    return visit(LV);
+    return this->delegate(LV);
 
   return false;
 }
@@ -2337,7 +2341,7 @@ bool ByteCodeExprGen::visitDecl(const VarDecl *VD) {
     auto GlobalIndex = P.getGlobal(VD);
     assert(GlobalIndex); // visitVarDecl() didn't return false.
     if (VarT) {
-      if (!this->emitGetGlobal(*VarT, *GlobalIndex, VD))
+      if (!this->emitGetGlobalUnchecked(*VarT, *GlobalIndex, VD))
         return false;
     } else {
       if (!this->emitGetPtrGlobal(*GlobalIndex, VD))
@@ -2767,23 +2771,16 @@ bool ByteCodeExprGen::VisitUnaryOperator(const UnaryOperator *E) {
     if (!this->visit(SubExpr))
       return false;
     return DiscardResult ? this->emitPop(*T, E) : this->emitComp(*T, E);
-  case UO_Real: { // __real x
-    assert(!T);
-    if (!this->visit(SubExpr))
-      return false;
-    if (!this->emitConstUint8(0, E))
-      return false;
-    if (!this->emitArrayElemPtrPopUint8(E))
-      return false;
-
-    // Since our _Complex implementation does not map to a primitive type,
-    // we sometimes have to do the lvalue-to-rvalue conversion here manually.
-    if (!SubExpr->isLValue())
-      return this->emitLoadPop(classifyPrim(E->getType()), E);
-    return true;
-  }
+  case UO_Real: // __real x
+    if (T)
+      return this->delegate(SubExpr);
+    return this->emitComplexReal(SubExpr);
   case UO_Imag: { // __imag x
-    assert(!T);
+    if (T) {
+      if (!this->discard(SubExpr))
+        return false;
+      return this->visitZeroInitializer(*T, SubExpr->getType(), SubExpr);
+    }
     if (!this->visit(SubExpr))
       return false;
     if (!this->emitConstUint8(1, E))
@@ -2948,10 +2945,32 @@ bool ByteCodeExprGen::emitPrimCast(PrimType FromT, PrimType ToT,
   return false;
 }
 
+/// Emits __real(SubExpr)
+template 
+bool ByteCodeExprGen::emitComplexReal(const Expr *SubExpr) {
+  assert(SubExpr->getType()->isAnyComplexType());
+
+  if (DiscardResult)
+    return this->discard(SubExpr);
+
+  if (!this->visit(SubExpr))
+    return false;
+  if (!this->emitConstUint8(0, SubExpr))
+    return false;
+  if (!this->emitArrayElemPtrPopUint8(SubExpr))
+    return false;
+
+  // Since our _Complex implementation does not map to a primitive type,
+  // we sometimes have to do the lvalue-to-rvalue conversion here manually.
+  if (!SubExpr->isLValue())
+    return this->emitLoadPop(*classifyComplexElementType(SubExpr->getType()),
+                             SubExpr);
+  return true;
+}
+
 /// When calling this, we have a pointer of the local-to-destroy
 /// on the stack.
 /// Emit destruction of record types (or arrays of record types).
-/// FIXME: Handle virtual destructors.
 template 
 bool ByteCodeExprGen::emitRecordDestruction(const Descriptor *Desc) {
   assert(Desc);
diff --git a/clang/lib/AST/Interp/ByteCodeExprGen.h b/clang/lib/AST/Interp/ByteCodeExprGen.h
index bbb13e97e7256..48005ce05724b 100644
--- a/clang/lib/AST/Interp/ByteCodeExprGen.h
+++ b/clang/lib/AST/Interp/ByteCodeExprGen.h
@@ -294,6 +294,8 @@ class ByteCodeExprGen : public ConstStmtVisitor, bool>,
     return this->classify(ElemType);
   }
 
+  bool emitComplexReal(const Expr *SubExpr);
+
   bool emitRecordDestruction(const Descriptor *Desc);
   unsigned collectBaseOffset(const RecordType *BaseType,
                              const RecordType *DerivedType);
diff --git a/clang/lib/AST/Interp/ByteCodeStmtGen.cpp b/clang/lib/AST/Interp/ByteCodeStmtGen.cpp
index b1ab5fcf9cb64..38067be73e254 100644
--- a/clang/lib/AST/Interp/ByteCodeStmtGen.cpp
+++ b/clang/lib/AST/Interp/ByteCodeStmtGen.cpp
@@ -142,6 +142,27 @@ bool ByteCodeStmtGen::visitFunc(const FunctionDecl *F) {
   // Classify the return type.
   ReturnType = this->classify(F->getReturnType());
 
+  auto emitFieldInitializer = [&](const Record::Field *F, unsigned FieldOffset,
+                                  const Expr *InitExpr) -> bool {
+    if (std::optional T = this->classify(InitExpr)) {
+      if (!this->visit(InitExpr))
+        return false;
+
+      if (F->isBitField())
+        return this->emitInitThisBitField(*T, F, FieldOffset, InitExpr);
+      return this->emitInitThisField(*T, FieldOffset, InitExpr);
+    }
+    // Non-primitive case. Get a pointer to the field-to-initialize
+    // on the stack and call visitInitialzer() for it.
+    if (!this->emitGetPtrThisField(FieldOffset, InitExpr))
+      return false;
+
+    if (!this->visitInitializer(InitExpr))
+      return false;
+
+    return this->emitPopPtr(InitExpr);
+  };
+
   // Emit custom code if this is a lambda static invoker.
   if (const auto *MD = dyn_cast(F);
       MD && MD->isLambdaStaticInvoker())
@@ -162,29 +183,8 @@ bool ByteCodeStmtGen::visitFunc(const FunctionDecl *F) {
       if (const FieldDecl *Member = Init->getMember()) {
         const Record::Field *F = R->getField(Member);
 
-        if (std::optional T = this->classify(InitExpr)) {
-          if (!this->visit(InitExpr))
-            return false;
-
-          if (F->isBitField()) {
-            if (!this->emitInitThisBitField(*T, F, InitExpr))
-              return false;
-          } else {
-            if (!this->emitInitThisField(*T, F->Offset, InitExpr))
-              return false;
-          }
-        } else {
-          // Non-primitive case. Get a pointer to the field-to-initialize
-          // on the stack and call visitInitialzer() for it.
-          if (!this->emitGetPtrThisField(F->Offset, InitExpr))
-            return false;
-
-          if (!this->visitInitializer(InitExpr))
-            return false;
-
-          if (!this->emitPopPtr(InitExpr))
-            return false;
-        }
+        if (!emitFieldInitializer(F, F->Offset, InitExpr))
+          return false;
       } else if (const Type *Base = Init->getBaseClass()) {
         // Base class initializer.
         // Get This Base and call initializer on it.
@@ -198,6 +198,26 @@ bool ByteCodeStmtGen::visitFunc(const FunctionDecl *F) {
           return false;
         if (!this->emitInitPtrPop(InitExpr))
           return false;
+      } else if (const IndirectFieldDecl *IFD = Init->getIndirectMember()) {
+        assert(IFD->getChainingSize() >= 2);
+
+        unsigned NestedFieldOffset = 0;
+        const Record::Field *NestedField = nullptr;
+        for (const NamedDecl *ND : IFD->chain()) {
+          const auto *FD = cast(ND);
+          const Record *FieldRecord =
+              this->P.getOrCreateRecord(FD->getParent());
+          assert(FieldRecord);
+
+          NestedField = FieldRecord->getField(FD);
+          assert(NestedField);
+
+          NestedFieldOffset += NestedField->Offset;
+        }
+        assert(NestedField);
+
+        if (!emitFieldInitializer(NestedField, NestedFieldOffset, InitExpr))
+          return false;
       } else {
         assert(Init->isDelegatingInitializer());
         if (!this->emitThis(InitExpr))
diff --git a/clang/lib/AST/Interp/IntegralAP.h b/clang/lib/AST/Interp/IntegralAP.h
index d5f46409d231d..55e29caa1cd74 100644
--- a/clang/lib/AST/Interp/IntegralAP.h
+++ b/clang/lib/AST/Interp/IntegralAP.h
@@ -204,14 +204,18 @@ template  class IntegralAP final {
   }
 
   static bool rem(IntegralAP A, IntegralAP B, unsigned OpBits, IntegralAP *R) {
-    // FIXME: Implement.
-    assert(false);
+    if constexpr (Signed)
+      *R = IntegralAP(A.V.srem(B.V));
+    else
+      *R = IntegralAP(A.V.urem(B.V));
     return false;
   }
 
   static bool div(IntegralAP A, IntegralAP B, unsigned OpBits, IntegralAP *R) {
-    // FIXME: Implement.
-    assert(false);
+    if constexpr (Signed)
+      *R = IntegralAP(A.V.sdiv(B.V));
+    else
+      *R = IntegralAP(A.V.udiv(B.V));
     return false;
   }
 
diff --git a/clang/lib/AST/Interp/Interp.cpp b/clang/lib/AST/Interp/Interp.cpp
index 21ea2503b94bf..ac03837408a4d 100644
--- a/clang/lib/AST/Interp/Interp.cpp
+++ b/clang/lib/AST/Interp/Interp.cpp
@@ -53,6 +53,25 @@ static bool Jf(InterpState &S, CodePtr &PC, int32_t Offset) {
   return true;
 }
 
+static void diagnoseNonConstVariable(InterpState &S, CodePtr OpPC,
+                                     const ValueDecl *VD) {
+  if (!S.getLangOpts().CPlusPlus)
+    return;
+
+  const SourceInfo &Loc = S.Current->getSource(OpPC);
+
+  if (VD->getType()->isIntegralOrEnumerationType())
+    S.FFDiag(Loc, diag::note_constexpr_ltor_non_const_int, 1) << VD;
+  else
+    S.FFDiag(Loc,
+             S.getLangOpts().CPlusPlus11
+                 ? diag::note_constexpr_ltor_non_constexpr
+                 : diag::note_constexpr_ltor_non_integral,
+             1)
+        << VD << VD->getType();
+  S.Note(VD->getLocation(), diag::note_declared_at);
+}
+
 static bool CheckActive(InterpState &S, CodePtr OpPC, const Pointer &Ptr,
                         AccessKinds AK) {
   if (Ptr.isActive())
@@ -134,6 +153,18 @@ void cleanupAfterFunctionCall(InterpState &S, CodePtr OpPC) {
   if (CurFunc->isUnevaluatedBuiltin())
     return;
 
+  // Some builtin functions require us to only look at the call site, since
+  // the classified parameter types do not match.
+  if (CurFunc->isBuiltin()) {
+    const auto *CE =
+        cast(S.Current->Caller->getExpr(S.Current->getRetPC()));
+    for (int32_t I = CE->getNumArgs() - 1; I >= 0; --I) {
+      const Expr *A = CE->getArg(I);
+      popArg(S, A);
+    }
+    return;
+  }
+
   if (S.Current->Caller && CurFunc->isVariadic()) {
     // CallExpr we're look for is at the return PC of the current function, i.e.
     // in the caller.
@@ -159,9 +190,7 @@ bool CheckExtern(InterpState &S, CodePtr OpPC, const Pointer &Ptr) {
 
   if (!S.checkingPotentialConstantExpression() && S.getLangOpts().CPlusPlus) {
     const auto *VD = Ptr.getDeclDesc()->asValueDecl();
-    const SourceInfo &Loc = S.Current->getSource(OpPC);
-    S.FFDiag(Loc, diag::note_constexpr_ltor_non_constexpr, 1) << VD;
-    S.Note(VD->getLocation(), diag::note_declared_at);
+    diagnoseNonConstVariable(S, OpPC, VD);
   }
   return false;
 }
@@ -204,6 +233,44 @@ bool CheckLive(InterpState &S, CodePtr OpPC, const Pointer &Ptr,
   return true;
 }
 
+bool CheckConstant(InterpState &S, CodePtr OpPC, const Descriptor *Desc) {
+  assert(Desc);
+
+  auto IsConstType = [&S](const VarDecl *VD) -> bool {
+    if (VD->isConstexpr())
+      return true;
+
+    if (S.getLangOpts().CPlusPlus && !S.getLangOpts().CPlusPlus11)
+      return false;
+
+    QualType T = VD->getType();
+    if (T.isConstQualified())
+      return true;
+
+    if (const auto *RT = T->getAs())
+      return RT->getPointeeType().isConstQualified();
+
+    if (const auto *PT = T->getAs())
+      return PT->getPointeeType().isConstQualified();
+
+    return false;
+  };
+
+  if (const auto *D = Desc->asValueDecl()) {
+    if (const auto *VD = dyn_cast(D);
+        VD && VD->hasGlobalStorage() && !IsConstType(VD)) {
+      diagnoseNonConstVariable(S, OpPC, VD);
+      return S.inConstantContext();
+    }
+  }
+
+  return true;
+}
+
+static bool CheckConstant(InterpState &S, CodePtr OpPC, const Pointer &Ptr) {
+  return CheckConstant(S, OpPC, Ptr.getDeclDesc());
+}
+
 bool CheckDummy(InterpState &S, CodePtr OpPC, const Pointer &Ptr) {
   return !Ptr.isZero() && !Ptr.isDummy();
 }
@@ -292,6 +359,9 @@ bool CheckInitialized(InterpState &S, CodePtr OpPC, const Pointer &Ptr,
 bool CheckLoad(InterpState &S, CodePtr OpPC, const Pointer &Ptr) {
   if (!CheckLive(S, OpPC, Ptr, AK_Read))
     return false;
+  if (!CheckConstant(S, OpPC, Ptr))
+    return false;
+
   if (!CheckDummy(S, OpPC, Ptr))
     return false;
   if (!CheckExtern(S, OpPC, Ptr))
@@ -420,7 +490,7 @@ bool CheckThis(InterpState &S, CodePtr OpPC, const Pointer &This) {
 }
 
 bool CheckPure(InterpState &S, CodePtr OpPC, const CXXMethodDecl *MD) {
-  if (!MD->isPure())
+  if (!MD->isPureVirtual())
     return true;
   const SourceInfo &E = S.Current->getSource(OpPC);
   S.FFDiag(E, diag::note_constexpr_pure_virtual_call, 1) << MD;
@@ -593,13 +663,7 @@ bool CheckDeclRef(InterpState &S, CodePtr OpPC, const DeclRefExpr *DR) {
     }
   } else if (const auto *VD = dyn_cast(D)) {
     if (!VD->getType().isConstQualified()) {
-      S.FFDiag(E,
-               VD->getType()->isIntegralOrEnumerationType()
-                   ? diag::note_constexpr_ltor_non_const_int
-                   : diag::note_constexpr_ltor_non_constexpr,
-               1)
-          << VD;
-      S.Note(VD->getLocation(), diag::note_declared_at) << VD->getSourceRange();
+      diagnoseNonConstVariable(S, OpPC, VD);
       return false;
     }
 
diff --git a/clang/lib/AST/Interp/Interp.h b/clang/lib/AST/Interp/Interp.h
index c05dea0cc55d3..5321f9617feff 100644
--- a/clang/lib/AST/Interp/Interp.h
+++ b/clang/lib/AST/Interp/Interp.h
@@ -77,6 +77,9 @@ bool CheckSubobject(InterpState &S, CodePtr OpPC, const Pointer &Ptr,
 /// Checks if a pointer points to const storage.
 bool CheckConst(InterpState &S, CodePtr OpPC, const Pointer &Ptr);
 
+/// Checks if the Descriptor is of a constexpr or const global variable.
+bool CheckConstant(InterpState &S, CodePtr OpPC, const Descriptor *Desc);
+
 /// Checks if a pointer points to a mutable field.
 bool CheckMutable(InterpState &S, CodePtr OpPC, const Pointer &Ptr);
 
@@ -1004,12 +1007,23 @@ bool SetThisField(InterpState &S, CodePtr OpPC, uint32_t I) {
 template ::T>
 bool GetGlobal(InterpState &S, CodePtr OpPC, uint32_t I) {
   const Block *B = S.P.getGlobal(I);
+
+  if (!CheckConstant(S, OpPC, B->getDescriptor()))
+    return false;
   if (B->isExtern())
     return false;
   S.Stk.push(B->deref());
   return true;
 }
 
+/// Same as GetGlobal, but without the checks.
+template ::T>
+bool GetGlobalUnchecked(InterpState &S, CodePtr OpPC, uint32_t I) {
+  auto *B = S.P.getGlobal(I);
+  S.Stk.push(B->deref());
+  return true;
+}
+
 template ::T>
 bool SetGlobal(InterpState &S, CodePtr OpPC, uint32_t I) {
   // TODO: emit warning.
@@ -1064,15 +1078,18 @@ bool InitThisField(InterpState &S, CodePtr OpPC, uint32_t I) {
   return true;
 }
 
+// FIXME: The Field pointer here is too much IMO and we could instead just
+// pass an Offset + BitWidth pair.
 template ::T>
-bool InitThisBitField(InterpState &S, CodePtr OpPC, const Record::Field *F) {
+bool InitThisBitField(InterpState &S, CodePtr OpPC, const Record::Field *F,
+                      uint32_t FieldOffset) {
   assert(F->isBitField());
   if (S.checkingPotentialConstantExpression())
     return false;
   const Pointer &This = S.Current->getThis();
   if (!CheckThis(S, OpPC, This))
     return false;
-  const Pointer &Field = This.atField(F->Offset);
+  const Pointer &Field = This.atField(FieldOffset);
   const auto &Value = S.Stk.pop();
   Field.deref() = Value.truncate(F->Decl->getBitWidthValue(S.getCtx()));
   Field.initialize();
diff --git a/clang/lib/AST/Interp/InterpBuiltin.cpp b/clang/lib/AST/Interp/InterpBuiltin.cpp
index b55b1569a2598..754ca96b0c645 100644
--- a/clang/lib/AST/Interp/InterpBuiltin.cpp
+++ b/clang/lib/AST/Interp/InterpBuiltin.cpp
@@ -164,6 +164,8 @@ static bool retPrimValue(InterpState &S, CodePtr OpPC, APValue &Result,
   case X:                                                                      \
     return Ret(S, OpPC, Result);
   switch (*T) {
+    RET_CASE(PT_Ptr);
+    RET_CASE(PT_FnPtr);
     RET_CASE(PT_Float);
     RET_CASE(PT_Bool);
     RET_CASE(PT_Sint8);
@@ -613,15 +615,34 @@ static bool interp__builtin_ffs(InterpState &S, CodePtr OpPC,
   return true;
 }
 
+static bool interp__builtin_addressof(InterpState &S, CodePtr OpPC,
+                                      const InterpFrame *Frame,
+                                      const Function *Func,
+                                      const CallExpr *Call) {
+  PrimType PtrT =
+      S.getContext().classify(Call->getArg(0)->getType()).value_or(PT_Ptr);
+
+  if (PtrT == PT_FnPtr) {
+    const FunctionPointer &Arg = S.Stk.peek();
+    S.Stk.push(Arg);
+  } else if (PtrT == PT_Ptr) {
+    const Pointer &Arg = S.Stk.peek();
+    S.Stk.push(Arg);
+  } else {
+    assert(false && "Unsupported pointer type passed to __builtin_addressof()");
+  }
+  return true;
+}
+
 bool InterpretBuiltin(InterpState &S, CodePtr OpPC, const Function *F,
                       const CallExpr *Call) {
   InterpFrame *Frame = S.Current;
   APValue Dummy;
 
-  QualType ReturnType = Call->getCallReturnType(S.getCtx());
-  std::optional ReturnT = S.getContext().classify(ReturnType);
+  std::optional ReturnT = S.getContext().classify(Call->getType());
+
   // If classify failed, we assume void.
-  assert(ReturnT || ReturnType->isVoidType());
+  assert(ReturnT || Call->getType()->isVoidType());
 
   switch (F->getBuiltinID()) {
   case Builtin::BI__builtin_is_constant_evaluated:
@@ -820,6 +841,12 @@ bool InterpretBuiltin(InterpState &S, CodePtr OpPC, const Function *F,
     if (!interp__builtin_ffs(S, OpPC, Frame, F, Call))
       return false;
     break;
+  case Builtin::BIaddressof:
+  case Builtin::BI__addressof:
+  case Builtin::BI__builtin_addressof:
+    if (!interp__builtin_addressof(S, OpPC, Frame, F, Call))
+      return false;
+    break;
 
   default:
     return false;
diff --git a/clang/lib/AST/Interp/Opcodes.td b/clang/lib/AST/Interp/Opcodes.td
index 69068e87d5720..6e0f1c939460d 100644
--- a/clang/lib/AST/Interp/Opcodes.td
+++ b/clang/lib/AST/Interp/Opcodes.td
@@ -379,6 +379,7 @@ def CheckGlobalCtor : Opcode {}
 
 // [] -> [Value]
 def GetGlobal : AccessOpcode;
+def GetGlobalUnchecked : AccessOpcode;
 // [Value] -> []
 def InitGlobal : AccessOpcode;
 // [Value] -> []
@@ -416,7 +417,11 @@ def InitThisField : AccessOpcode;
 // [Value] -> []
 def InitThisFieldActive : AccessOpcode;
 // [Value] -> []
-def InitThisBitField : BitFieldOpcode;
+def InitThisBitField : Opcode {
+  let Types = [AluTypeClass];
+  let Args = [ArgRecordField, ArgUint32];
+  let HasGroup = 1;
+}
 // [Pointer, Value] -> []
 def InitField : AccessOpcode;
 // [Pointer, Value] -> []
diff --git a/clang/lib/AST/JSONNodeDumper.cpp b/clang/lib/AST/JSONNodeDumper.cpp
index 88b75833fb4a4..17c0cb1f2e5b2 100644
--- a/clang/lib/AST/JSONNodeDumper.cpp
+++ b/clang/lib/AST/JSONNodeDumper.cpp
@@ -930,7 +930,7 @@ void JSONNodeDumper::VisitFunctionDecl(const FunctionDecl *FD) {
     JOS.attribute("storageClass", VarDecl::getStorageClassSpecifierString(SC));
   attributeOnlyIfTrue("inline", FD->isInlineSpecified());
   attributeOnlyIfTrue("virtual", FD->isVirtualAsWritten());
-  attributeOnlyIfTrue("pure", FD->isPure());
+  attributeOnlyIfTrue("pure", FD->isPureVirtual());
   attributeOnlyIfTrue("explicitlyDeleted", FD->isDeletedAsWritten());
   attributeOnlyIfTrue("constexpr", FD->isConstexpr());
   attributeOnlyIfTrue("variadic", FD->isVariadic());
diff --git a/clang/lib/AST/ODRDiagsEmitter.cpp b/clang/lib/AST/ODRDiagsEmitter.cpp
index 9dcd2ed04f6f7..5b1cdc16e2ea2 100644
--- a/clang/lib/AST/ODRDiagsEmitter.cpp
+++ b/clang/lib/AST/ODRDiagsEmitter.cpp
@@ -1102,8 +1102,8 @@ bool ODRDiagsEmitter::diagnoseMismatch(
 
     const bool FirstVirtual = FirstMethod->isVirtualAsWritten();
     const bool SecondVirtual = SecondMethod->isVirtualAsWritten();
-    const bool FirstPure = FirstMethod->isPure();
-    const bool SecondPure = SecondMethod->isPure();
+    const bool FirstPure = FirstMethod->isPureVirtual();
+    const bool SecondPure = SecondMethod->isPureVirtual();
     if ((FirstVirtual || SecondVirtual) &&
         (FirstVirtual != SecondVirtual || FirstPure != SecondPure)) {
       DiagMethodError(MethodVirtual) << FirstPure << FirstVirtual;
diff --git a/clang/lib/AST/ODRHash.cpp b/clang/lib/AST/ODRHash.cpp
index aea1a93ae1fa8..c5aa91a81da05 100644
--- a/clang/lib/AST/ODRHash.cpp
+++ b/clang/lib/AST/ODRHash.cpp
@@ -688,7 +688,7 @@ void ODRHash::AddFunctionDecl(const FunctionDecl *Function,
   ID.AddInteger(Function->getStorageClass());
   AddBoolean(Function->isInlineSpecified());
   AddBoolean(Function->isVirtualAsWritten());
-  AddBoolean(Function->isPure());
+  AddBoolean(Function->isPureVirtual());
   AddBoolean(Function->isDeletedAsWritten());
   AddBoolean(Function->isExplicitlyDefaulted());
 
@@ -741,8 +741,55 @@ void ODRHash::AddEnumDecl(const EnumDecl *Enum) {
   if (Enum->isScoped())
     AddBoolean(Enum->isScopedUsingClassTag());
 
-  if (Enum->getIntegerTypeSourceInfo())
-    AddQualType(Enum->getIntegerType());
+  if (Enum->getIntegerTypeSourceInfo()) {
+    // FIMXE: This allows two enums with different spellings to have the same
+    // hash.
+    //
+    //  // mod1.cppm
+    //  module;
+    //  extern "C" {
+    //      typedef unsigned __int64 size_t;
+    //  }
+    //  namespace std {
+    //      using :: size_t;
+    //  }
+    //
+    //  extern "C++" {
+    //      namespace std {
+    //          enum class align_val_t : std::size_t {};
+    //      }
+    //  }
+    //
+    //  export module mod1;
+    //  export using std::align_val_t;
+    //
+    //  // mod2.cppm
+    //  module;
+    //  extern "C" {
+    //      typedef unsigned __int64 size_t;
+    //  }
+    //
+    //  extern "C++" {
+    //      namespace std {
+    //          enum class align_val_t : size_t {};
+    //      }
+    //  }
+    //
+    //  export module mod2;
+    //  import mod1;
+    //  export using std::align_val_t;
+    //
+    // The above example should be disallowed since it violates
+    // [basic.def.odr]p14:
+    //
+    //    Each such definition shall consist of the same sequence of tokens
+    //
+    // The definitions of `std::align_val_t` in two module units have different
+    // spellings but we failed to give an error here.
+    //
+    // See https://github.com/llvm/llvm-project/issues/76638 for details.
+    AddQualType(Enum->getIntegerType().getCanonicalType());
+  }
 
   // Filter out sub-Decls which will not be processed in order to get an
   // accurate count of Decl's.
diff --git a/clang/lib/AST/RecordLayoutBuilder.cpp b/clang/lib/AST/RecordLayoutBuilder.cpp
index a9ef9079216fc..ec9d1a56a8cf1 100644
--- a/clang/lib/AST/RecordLayoutBuilder.cpp
+++ b/clang/lib/AST/RecordLayoutBuilder.cpp
@@ -2349,7 +2349,7 @@ static const CXXMethodDecl *computeKeyFunction(ASTContext &Context,
     if (!MD->isVirtual())
       continue;
 
-    if (MD->isPure())
+    if (MD->isPureVirtual())
       continue;
 
     // Ignore implicit member functions, they are always marked as inline, but
@@ -3294,7 +3294,7 @@ void MicrosoftRecordLayoutBuilder::computeVtorDispSet(
   // Seed the working set with our non-destructor, non-pure virtual methods.
   for (const CXXMethodDecl *MD : RD->methods())
     if (MicrosoftVTableContext::hasVtableSlot(MD) &&
-        !isa(MD) && !MD->isPure())
+        !isa(MD) && !MD->isPureVirtual())
       Work.insert(MD);
   while (!Work.empty()) {
     const CXXMethodDecl *MD = *Work.begin();
diff --git a/clang/lib/AST/TextNodeDumper.cpp b/clang/lib/AST/TextNodeDumper.cpp
index 302fd36cab8f5..6fde21390b63c 100644
--- a/clang/lib/AST/TextNodeDumper.cpp
+++ b/clang/lib/AST/TextNodeDumper.cpp
@@ -1878,7 +1878,7 @@ void TextNodeDumper::VisitFunctionDecl(const FunctionDecl *D) {
   if (D->isModulePrivate())
     OS << " __module_private__";
 
-  if (D->isPure())
+  if (D->isPureVirtual())
     OS << " pure";
   if (D->isDefaulted()) {
     OS << " default";
diff --git a/clang/lib/AST/Type.cpp b/clang/lib/AST/Type.cpp
index 556f9a17fee40..626133d735b0c 100644
--- a/clang/lib/AST/Type.cpp
+++ b/clang/lib/AST/Type.cpp
@@ -3465,6 +3465,14 @@ FunctionProtoType::FunctionProtoType(QualType result, ArrayRef params,
     FunctionTypeBits.HasExtraBitfields = false;
   }
 
+  if (epi.requiresFunctionProtoTypeArmAttributes()) {
+    auto &ArmTypeAttrs = *getTrailingObjects();
+    ArmTypeAttrs = FunctionTypeArmAttributes();
+
+    // Also set the bit in FunctionTypeExtraBitfields
+    auto &ExtraBits = *getTrailingObjects();
+    ExtraBits.HasArmTypeAttributes = true;
+  }
 
   // Fill in the trailing argument array.
   auto *argSlot = getTrailingObjects();
@@ -3476,10 +3484,10 @@ FunctionProtoType::FunctionProtoType(QualType result, ArrayRef params,
 
   // Propagate the SME ACLE attributes.
   if (epi.AArch64SMEAttributes != SME_NormalFunction) {
-    auto &ExtraBits = *getTrailingObjects();
+    auto &ArmTypeAttrs = *getTrailingObjects();
     assert(epi.AArch64SMEAttributes <= SME_AttributeMask &&
            "Not enough bits to encode SME attributes");
-    ExtraBits.AArch64SMEAttributes = epi.AArch64SMEAttributes;
+    ArmTypeAttrs.AArch64SMEAttributes = epi.AArch64SMEAttributes;
   }
 
   // Fill in the exception type array if present.
diff --git a/clang/lib/AST/TypePrinter.cpp b/clang/lib/AST/TypePrinter.cpp
index dd4bde709da8e..dc78b061d5646 100644
--- a/clang/lib/AST/TypePrinter.cpp
+++ b/clang/lib/AST/TypePrinter.cpp
@@ -947,15 +947,20 @@ void TypePrinter::printFunctionProtoAfter(const FunctionProtoType *T,
   OS << ')';
 
   FunctionType::ExtInfo Info = T->getExtInfo();
+  unsigned SMEBits = T->getAArch64SMEAttributes();
 
-  if ((T->getAArch64SMEAttributes() & FunctionType::SME_PStateSMCompatibleMask))
+  if (SMEBits & FunctionType::SME_PStateSMCompatibleMask)
     OS << " __arm_streaming_compatible";
-  if ((T->getAArch64SMEAttributes() & FunctionType::SME_PStateSMEnabledMask))
+  if (SMEBits & FunctionType::SME_PStateSMEnabledMask)
     OS << " __arm_streaming";
-  if ((T->getAArch64SMEAttributes() & FunctionType::SME_PStateZASharedMask))
-    OS << " __arm_shared_za";
-  if ((T->getAArch64SMEAttributes() & FunctionType::SME_PStateZAPreservedMask))
-    OS << " __arm_preserves_za";
+  if (FunctionType::getArmZAState(SMEBits) == FunctionType::ARM_Preserves)
+    OS << " __arm_preserves(\"za\")";
+  if (FunctionType::getArmZAState(SMEBits) == FunctionType::ARM_In)
+    OS << " __arm_in(\"za\")";
+  if (FunctionType::getArmZAState(SMEBits) == FunctionType::ARM_Out)
+    OS << " __arm_out(\"za\")";
+  if (FunctionType::getArmZAState(SMEBits) == FunctionType::ARM_InOut)
+    OS << " __arm_inout(\"za\")";
 
   printFunctionAfter(Info, OS);
 
@@ -1803,14 +1808,6 @@ void TypePrinter::printAttributedAfter(const AttributedType *T,
     OS << "__arm_streaming_compatible";
     return;
   }
-  if (T->getAttrKind() == attr::ArmSharedZA) {
-    OS << "__arm_shared_za";
-    return;
-  }
-  if (T->getAttrKind() == attr::ArmPreservesZA) {
-    OS << "__arm_preserves_za";
-    return;
-  }
 
   OS << " __attribute__((";
   switch (T->getAttrKind()) {
@@ -1858,8 +1855,10 @@ void TypePrinter::printAttributedAfter(const AttributedType *T,
   case attr::WebAssemblyFuncref:
   case attr::ArmStreaming:
   case attr::ArmStreamingCompatible:
-  case attr::ArmSharedZA:
-  case attr::ArmPreservesZA:
+  case attr::ArmIn:
+  case attr::ArmOut:
+  case attr::ArmInOut:
+  case attr::ArmPreserves:
     llvm_unreachable("This attribute should have been handled already");
 
   case attr::NSReturnsRetained:
diff --git a/clang/lib/AST/VTableBuilder.cpp b/clang/lib/AST/VTableBuilder.cpp
index 69861c65f6e3a..a956ca5b37acf 100644
--- a/clang/lib/AST/VTableBuilder.cpp
+++ b/clang/lib/AST/VTableBuilder.cpp
@@ -422,7 +422,7 @@ void FinalOverriders::dump(raw_ostream &Out, BaseSubobject Base,
     Out << ", " << Overrider.Offset.getQuantity() << ')';
 
     BaseOffset Offset;
-    if (!Overrider.Method->isPure())
+    if (!Overrider.Method->isPureVirtual())
       Offset = ComputeReturnAdjustmentBaseOffset(Context, Overrider.Method, MD);
 
     if (!Offset.isEmpty()) {
@@ -1261,7 +1261,7 @@ ThisAdjustment ItaniumVTableBuilder::ComputeThisAdjustment(
     const CXXMethodDecl *MD, CharUnits BaseOffsetInLayoutClass,
     FinalOverriders::OverriderInfo Overrider) {
   // Ignore adjustments for pure virtual member functions.
-  if (Overrider.Method->isPure())
+  if (Overrider.Method->isPureVirtual())
     return ThisAdjustment();
 
   BaseSubobject OverriddenBaseSubobject(MD->getParent(),
@@ -1607,7 +1607,7 @@ void ItaniumVTableBuilder::AddMethods(
     // Check if this overrider needs a return adjustment.
     // We don't want to do this for pure virtual member functions.
     BaseOffset ReturnAdjustmentOffset;
-    if (!OverriderMD->isPure()) {
+    if (!OverriderMD->isPureVirtual()) {
       ReturnAdjustmentOffset =
         ComputeReturnAdjustmentBaseOffset(Context, OverriderMD, MD);
     }
@@ -1959,7 +1959,7 @@ void ItaniumVTableBuilder::dumpLayout(raw_ostream &Out) {
       std::string Str = PredefinedExpr::ComputeName(
           PredefinedIdentKind::PrettyFunctionNoVirtual, MD);
       Out << Str;
-      if (MD->isPure())
+      if (MD->isPureVirtual())
         Out << " [pure]";
 
       if (MD->isDeleted())
@@ -2010,7 +2010,7 @@ void ItaniumVTableBuilder::dumpLayout(raw_ostream &Out) {
       else
         Out << "() [deleting]";
 
-      if (DD->isPure())
+      if (DD->isPureVirtual())
         Out << " [pure]";
 
       ThunkInfo Thunk = VTableThunks.lookup(I);
@@ -2038,7 +2038,7 @@ void ItaniumVTableBuilder::dumpLayout(raw_ostream &Out) {
       std::string Str = PredefinedExpr::ComputeName(
           PredefinedIdentKind::PrettyFunctionNoVirtual, MD);
       Out << "[unused] " << Str;
-      if (MD->isPure())
+      if (MD->isPureVirtual())
         Out << " [pure]";
     }
 
@@ -3076,7 +3076,7 @@ void VFTableBuilder::AddMethods(BaseSubobject Base, unsigned BaseDepth,
     // We don't want to do this for pure virtual member functions.
     BaseOffset ReturnAdjustmentOffset;
     ReturnAdjustment ReturnAdjustment;
-    if (!FinalOverriderMD->isPure()) {
+    if (!FinalOverriderMD->isPureVirtual()) {
       ReturnAdjustmentOffset =
           ComputeReturnAdjustmentBaseOffset(Context, FinalOverriderMD, MD);
     }
@@ -3175,7 +3175,7 @@ void VFTableBuilder::dumpLayout(raw_ostream &Out) {
       std::string Str = PredefinedExpr::ComputeName(
           PredefinedIdentKind::PrettyFunctionNoVirtual, MD);
       Out << Str;
-      if (MD->isPure())
+      if (MD->isPureVirtual())
         Out << " [pure]";
 
       if (MD->isDeleted())
@@ -3194,7 +3194,7 @@ void VFTableBuilder::dumpLayout(raw_ostream &Out) {
       DD->printQualifiedName(Out);
       Out << "() [scalar deleting]";
 
-      if (DD->isPure())
+      if (DD->isPureVirtual())
         Out << " [pure]";
 
       ThunkInfo Thunk = VTableThunks.lookup(I);
diff --git a/clang/lib/ASTMatchers/ASTMatchersInternal.cpp b/clang/lib/ASTMatchers/ASTMatchersInternal.cpp
index 8ed213ca2ce09..bf87b1aa0992a 100644
--- a/clang/lib/ASTMatchers/ASTMatchersInternal.cpp
+++ b/clang/lib/ASTMatchers/ASTMatchersInternal.cpp
@@ -893,6 +893,7 @@ const internal::VariadicDynCastAllOfMatcher
     cxxOperatorCallExpr;
 const internal::VariadicDynCastAllOfMatcher
     cxxRewrittenBinaryOperator;
+const internal::VariadicDynCastAllOfMatcher cxxFoldExpr;
 const internal::VariadicDynCastAllOfMatcher expr;
 const internal::VariadicDynCastAllOfMatcher declRefExpr;
 const internal::VariadicDynCastAllOfMatcher objcIvarRefExpr;
diff --git a/clang/lib/ASTMatchers/Dynamic/Registry.cpp b/clang/lib/ASTMatchers/Dynamic/Registry.cpp
index 2e43dec331b75..15dad022df5fe 100644
--- a/clang/lib/ASTMatchers/Dynamic/Registry.cpp
+++ b/clang/lib/ASTMatchers/Dynamic/Registry.cpp
@@ -198,6 +198,7 @@ RegistryMaps::RegistryMaps() {
   REGISTER_MATCHER(cxxDependentScopeMemberExpr);
   REGISTER_MATCHER(cxxDestructorDecl);
   REGISTER_MATCHER(cxxDynamicCastExpr);
+  REGISTER_MATCHER(cxxFoldExpr);
   REGISTER_MATCHER(cxxForRangeStmt);
   REGISTER_MATCHER(cxxFunctionalCastExpr);
   REGISTER_MATCHER(cxxMemberCallExpr);
@@ -319,6 +320,7 @@ RegistryMaps::RegistryMaps() {
   REGISTER_MATCHER(hasExplicitSpecifier);
   REGISTER_MATCHER(hasExternalFormalLinkage);
   REGISTER_MATCHER(hasFalseExpression);
+  REGISTER_MATCHER(hasFoldInit);
   REGISTER_MATCHER(hasGlobalStorage);
   REGISTER_MATCHER(hasImplicitDestinationType);
   REGISTER_MATCHER(hasInClassInitializer);
@@ -344,6 +346,7 @@ RegistryMaps::RegistryMaps() {
   REGISTER_MATCHER(hasOverloadedOperatorName);
   REGISTER_MATCHER(hasParameter);
   REGISTER_MATCHER(hasParent);
+  REGISTER_MATCHER(hasPattern);
   REGISTER_MATCHER(hasPointeeLoc);
   REGISTER_MATCHER(hasQualifier);
   REGISTER_MATCHER(hasRHS);
@@ -404,6 +407,7 @@ RegistryMaps::RegistryMaps() {
   REGISTER_MATCHER(isAssignmentOperator);
   REGISTER_MATCHER(isAtPosition);
   REGISTER_MATCHER(isBaseInitializer);
+  REGISTER_MATCHER(isBinaryFold);
   REGISTER_MATCHER(isBitField);
   REGISTER_MATCHER(isCatchAll);
   REGISTER_MATCHER(isClass);
@@ -447,6 +451,7 @@ RegistryMaps::RegistryMaps() {
   REGISTER_MATCHER(isInteger);
   REGISTER_MATCHER(isIntegral);
   REGISTER_MATCHER(isLambda);
+  REGISTER_MATCHER(isLeftFold);
   REGISTER_MATCHER(isListInitialization);
   REGISTER_MATCHER(isMain);
   REGISTER_MATCHER(isMemberInitializer);
@@ -460,6 +465,7 @@ RegistryMaps::RegistryMaps() {
   REGISTER_MATCHER(isProtected);
   REGISTER_MATCHER(isPublic);
   REGISTER_MATCHER(isPure);
+  REGISTER_MATCHER(isRightFold);
   REGISTER_MATCHER(isScoped);
   REGISTER_MATCHER(isSharedKind);
   REGISTER_MATCHER(isSignedInteger);
@@ -469,6 +475,7 @@ RegistryMaps::RegistryMaps() {
   REGISTER_MATCHER(isStruct);
   REGISTER_MATCHER(isTemplateInstantiation);
   REGISTER_MATCHER(isTypeDependent);
+  REGISTER_MATCHER(isUnaryFold);
   REGISTER_MATCHER(isUnion);
   REGISTER_MATCHER(isUnsignedInteger);
   REGISTER_MATCHER(isUserProvided);
diff --git a/clang/lib/Analysis/ExprMutationAnalyzer.cpp b/clang/lib/Analysis/ExprMutationAnalyzer.cpp
index c0de9277ff866..9af818be0415f 100644
--- a/clang/lib/Analysis/ExprMutationAnalyzer.cpp
+++ b/clang/lib/Analysis/ExprMutationAnalyzer.cpp
@@ -243,9 +243,16 @@ const Stmt *ExprMutationAnalyzer::findMutationMemoized(
 
 const Stmt *ExprMutationAnalyzer::tryEachDeclRef(const Decl *Dec,
                                                  MutationFinder Finder) {
-  const auto Refs =
-      match(findAll(declRefExpr(to(equalsNode(Dec))).bind(NodeID::value)),
-            Stm, Context);
+  const auto Refs = match(
+      findAll(
+          declRefExpr(to(
+                          // `Dec` or a binding if `Dec` is a decomposition.
+                          anyOf(equalsNode(Dec),
+                                bindingDecl(forDecomposition(equalsNode(Dec))))
+                          //
+                          ))
+              .bind(NodeID::value)),
+      Stm, Context);
   for (const auto &RefNodes : Refs) {
     const auto *E = RefNodes.getNodeAs(NodeID::value);
     if ((this->*Finder)(E))
diff --git a/clang/lib/Analysis/FlowSensitive/ControlFlowContext.cpp b/clang/lib/Analysis/FlowSensitive/ControlFlowContext.cpp
index 56246066e4aa1..c9ebffe6f3780 100644
--- a/clang/lib/Analysis/FlowSensitive/ControlFlowContext.cpp
+++ b/clang/lib/Analysis/FlowSensitive/ControlFlowContext.cpp
@@ -69,7 +69,7 @@ static llvm::BitVector findReachableBlocks(const CFG &Cfg) {
 
 llvm::Expected
 ControlFlowContext::build(const FunctionDecl &Func) {
-  if (!Func.hasBody())
+  if (!Func.doesThisDeclarationHaveABody())
     return llvm::createStringError(
         std::make_error_code(std::errc::invalid_argument),
         "Cannot analyze function without a body");
diff --git a/clang/lib/Analysis/FlowSensitive/DataflowAnalysisContext.cpp b/clang/lib/Analysis/FlowSensitive/DataflowAnalysisContext.cpp
index fa114979c8e32..f4c4af022f51f 100644
--- a/clang/lib/Analysis/FlowSensitive/DataflowAnalysisContext.cpp
+++ b/clang/lib/Analysis/FlowSensitive/DataflowAnalysisContext.cpp
@@ -174,6 +174,9 @@ Solver::Result DataflowAnalysisContext::querySolver(
 
 bool DataflowAnalysisContext::flowConditionImplies(Atom Token,
                                                    const Formula &F) {
+  if (F.isLiteral(true))
+    return true;
+
   // Returns true if and only if truth assignment of the flow condition implies
   // that `F` is also true. We prove whether or not this property holds by
   // reducing the problem to satisfiability checking. In other words, we attempt
@@ -188,6 +191,9 @@ bool DataflowAnalysisContext::flowConditionImplies(Atom Token,
 
 bool DataflowAnalysisContext::flowConditionAllows(Atom Token,
                                                   const Formula &F) {
+  if (F.isLiteral(false))
+    return false;
+
   llvm::SetVector Constraints;
   Constraints.insert(&arena().makeAtomRef(Token));
   Constraints.insert(&F);
@@ -292,7 +298,7 @@ DataflowAnalysisContext::getControlFlowContext(const FunctionDecl *F) {
   if (It != FunctionContexts.end())
     return &It->second;
 
-  if (F->hasBody()) {
+  if (F->doesThisDeclarationHaveABody()) {
     auto CFCtx = ControlFlowContext::build(*F);
     // FIXME: Handle errors.
     assert(CFCtx);
diff --git a/clang/lib/Analysis/FlowSensitive/DataflowEnvironment.cpp b/clang/lib/Analysis/FlowSensitive/DataflowEnvironment.cpp
index 96fe6df88dbb9..07dc3a9f76ac2 100644
--- a/clang/lib/Analysis/FlowSensitive/DataflowEnvironment.cpp
+++ b/clang/lib/Analysis/FlowSensitive/DataflowEnvironment.cpp
@@ -386,7 +386,7 @@ void Environment::initialize() {
     return;
 
   if (const auto *FuncDecl = dyn_cast(DeclCtx)) {
-    assert(FuncDecl->getBody() != nullptr);
+    assert(FuncDecl->doesThisDeclarationHaveABody());
 
     initFieldsGlobalsAndFuncs(FuncDecl);
 
@@ -426,7 +426,7 @@ void Environment::initialize() {
 // FIXME: Add support for resetting globals after function calls to enable
 // the implementation of sound analyses.
 void Environment::initFieldsGlobalsAndFuncs(const FunctionDecl *FuncDecl) {
-  assert(FuncDecl->getBody() != nullptr);
+  assert(FuncDecl->doesThisDeclarationHaveABody());
 
   FieldSet Fields;
   llvm::DenseSet Vars;
@@ -747,6 +747,7 @@ static bool isOriginalRecordConstructor(const Expr &RecordPRValue) {
     return !Init->isSemanticForm() || !Init->isTransparent();
   return isa(RecordPRValue) || isa(RecordPRValue) ||
          isa(RecordPRValue) ||
+         isa(RecordPRValue) ||
          // The framework currently does not propagate the objects created in
          // the two branches of a `ConditionalOperator` because there is no way
          // to reconcile their storage locations, which are different. We
@@ -803,13 +804,15 @@ void Environment::setValue(const StorageLocation &Loc, Value &Val) {
 }
 
 void Environment::setValue(const Expr &E, Value &Val) {
+  const Expr &CanonE = ignoreCFGOmittedNodes(E);
+
   if (auto *RecordVal = dyn_cast(&Val)) {
-    assert(isOriginalRecordConstructor(E) ||
-           &RecordVal->getLoc() == &getResultObjectLocation(E));
+    assert(isOriginalRecordConstructor(CanonE) ||
+           &RecordVal->getLoc() == &getResultObjectLocation(CanonE));
   }
 
-  assert(E.isPRValue());
-  ExprToVal[&E] = &Val;
+  assert(CanonE.isPRValue());
+  ExprToVal[&CanonE] = &Val;
 }
 
 Value *Environment::getValue(const StorageLocation &Loc) const {
diff --git a/clang/lib/Analysis/FlowSensitive/SimplifyConstraints.cpp b/clang/lib/Analysis/FlowSensitive/SimplifyConstraints.cpp
index 00a3104a57b6b..cc20202768b92 100644
--- a/clang/lib/Analysis/FlowSensitive/SimplifyConstraints.cpp
+++ b/clang/lib/Analysis/FlowSensitive/SimplifyConstraints.cpp
@@ -131,9 +131,9 @@ void simplifyConstraints(llvm::SetVector &Constraints,
     for (const auto *Constraint : Constraints) {
       const Formula &NewConstraint =
           substitute(*Constraint, Substitutions, arena);
-      if (&NewConstraint == &arena.makeLiteral(true))
+      if (NewConstraint.isLiteral(true))
         continue;
-      if (&NewConstraint == &arena.makeLiteral(false)) {
+      if (NewConstraint.isLiteral(false)) {
         contradiction();
         return;
       }
diff --git a/clang/lib/Analysis/FlowSensitive/Value.cpp b/clang/lib/Analysis/FlowSensitive/Value.cpp
index 349f873f1e6c4..7fad6deb0e918 100644
--- a/clang/lib/Analysis/FlowSensitive/Value.cpp
+++ b/clang/lib/Analysis/FlowSensitive/Value.cpp
@@ -27,9 +27,17 @@ static bool areEquivalentIndirectionValues(const Value &Val1,
 }
 
 bool areEquivalentValues(const Value &Val1, const Value &Val2) {
-  return &Val1 == &Val2 || (Val1.getKind() == Val2.getKind() &&
-                            (isa(&Val1) ||
-                             areEquivalentIndirectionValues(Val1, Val2)));
+  if (&Val1 == &Val2)
+    return true;
+  if (Val1.getKind() != Val2.getKind())
+    return false;
+  // If values are distinct and have properties, we don't consider them equal,
+  // leaving equality up to the user model.
+  if (!Val1.properties().empty() || !Val2.properties().empty())
+    return false;
+  if (isa(&Val1))
+    return true;
+  return areEquivalentIndirectionValues(Val1, Val2);
 }
 
 raw_ostream &operator<<(raw_ostream &OS, const Value &Val) {
diff --git a/clang/lib/Basic/Targets.cpp b/clang/lib/Basic/Targets.cpp
index fceb8627389aa..5deacaa344f96 100644
--- a/clang/lib/Basic/Targets.cpp
+++ b/clang/lib/Basic/Targets.cpp
@@ -644,6 +644,8 @@ std::unique_ptr AllocateTarget(const llvm::Triple &Triple,
       return std::make_unique>(Triple, Opts);
     case llvm::Triple::PS5:
       return std::make_unique>(Triple, Opts);
+    case llvm::Triple::Hurd:
+      return std::make_unique>(Triple, Opts);
     default:
       return std::make_unique(Triple, Opts);
     }
diff --git a/clang/lib/Basic/Targets/AArch64.cpp b/clang/lib/Basic/Targets/AArch64.cpp
index 2f8395cb8932f..d47181bfca4fc 100644
--- a/clang/lib/Basic/Targets/AArch64.cpp
+++ b/clang/lib/Basic/Targets/AArch64.cpp
@@ -226,6 +226,7 @@ bool AArch64TargetInfo::validateBranchProtection(StringRef Spec, StringRef,
 
   BPI.BranchTargetEnforcement = PBP.BranchTargetEnforcement;
   BPI.BranchProtectionPAuthLR = PBP.BranchProtectionPAuthLR;
+  BPI.GuardedControlStack = PBP.GuardedControlStack;
   return true;
 }
 
@@ -532,6 +533,9 @@ void AArch64TargetInfo::getTargetDefines(const LangOptions &Opts,
   if (Opts.BranchTargetEnforcement)
     Builder.defineMacro("__ARM_FEATURE_BTI_DEFAULT", "1");
 
+  if (Opts.GuardedControlStack)
+    Builder.defineMacro("__ARM_FEATURE_GCS_DEFAULT", "1");
+
   if (HasLS64)
     Builder.defineMacro("__ARM_FEATURE_LS64", "1");
 
@@ -544,6 +548,9 @@ void AArch64TargetInfo::getTargetDefines(const LangOptions &Opts,
   if (HasD128)
     Builder.defineMacro("__ARM_FEATURE_SYSREG128", "1");
 
+  if (HasGCS)
+    Builder.defineMacro("__ARM_FEATURE_GCS", "1");
+
   if (*ArchInfo == llvm::AArch64::ARMV8_1A)
     getTargetDefinesARMV81A(Opts, Builder);
   else if (*ArchInfo == llvm::AArch64::ARMV8_2A)
@@ -615,9 +622,8 @@ AArch64TargetInfo::getVScaleRange(const LangOptions &LangOpts) const {
 unsigned AArch64TargetInfo::multiVersionSortPriority(StringRef Name) const {
   if (Name == "default")
     return 0;
-  for (const auto &E : llvm::AArch64::Extensions)
-    if (Name == E.Name)
-      return E.FmvPriority;
+  if (auto Ext = llvm::AArch64::parseArchExtension(Name))
+    return Ext->FmvPriority;
   return 0;
 }
 
@@ -627,24 +633,19 @@ unsigned AArch64TargetInfo::multiVersionFeatureCost() const {
 }
 
 bool AArch64TargetInfo::doesFeatureAffectCodeGen(StringRef Name) const {
-  auto F = llvm::find_if(llvm::AArch64::Extensions, [&](const auto &E) {
-    return Name == E.Name && !E.DependentFeatures.empty();
-  });
-  return F != std::end(llvm::AArch64::Extensions);
+  if (auto Ext = llvm::AArch64::parseArchExtension(Name))
+    return !Ext->DependentFeatures.empty();
+  return false;
 }
 
 StringRef AArch64TargetInfo::getFeatureDependencies(StringRef Name) const {
-  auto F = llvm::find_if(llvm::AArch64::Extensions,
-                         [&](const auto &E) { return Name == E.Name; });
-  return F != std::end(llvm::AArch64::Extensions) ? F->DependentFeatures
-                                                  : StringRef();
+  if (auto Ext = llvm::AArch64::parseArchExtension(Name))
+    return Ext->DependentFeatures;
+  return StringRef();
 }
 
 bool AArch64TargetInfo::validateCpuSupports(StringRef FeatureStr) const {
-  for (const auto &E : llvm::AArch64::Extensions)
-    if (FeatureStr == E.Name)
-      return true;
-  return false;
+  return llvm::AArch64::parseArchExtension(FeatureStr).has_value();
 }
 
 bool AArch64TargetInfo::hasFeature(StringRef Feature) const {
@@ -1086,8 +1087,7 @@ ParsedTargetAttr AArch64TargetInfo::parseTargetAttr(StringRef Features) const {
       FoundArch = true;
       std::pair Split =
           Feature.split("=").second.trim().split("+");
-      const std::optional AI =
-          llvm::AArch64::parseArch(Split.first);
+      const llvm::AArch64::ArchInfo *AI = llvm::AArch64::parseArch(Split.first);
 
       // Parse the architecture version, adding the required features to
       // Ret.Features.
diff --git a/clang/lib/Basic/Targets/AMDGPU.h b/clang/lib/Basic/Targets/AMDGPU.h
index 59b3c8e56cb6d..9367f565cf5d8 100644
--- a/clang/lib/Basic/Targets/AMDGPU.h
+++ b/clang/lib/Basic/Targets/AMDGPU.h
@@ -169,10 +169,8 @@ class LLVM_LIBRARY_VISIBILITY AMDGPUTargetInfo final : public TargetInfo {
     }
 
     bool HasLeftParen = false;
-    if (S.front() == '{') {
+    if (S.consume_front("{"))
       HasLeftParen = true;
-      S = S.drop_front();
-    }
     if (S.empty())
       return false;
     if (S.front() != 'v' && S.front() != 's' && S.front() != 'a') {
@@ -199,29 +197,24 @@ class LLVM_LIBRARY_VISIBILITY AMDGPUTargetInfo final : public TargetInfo {
       return true;
     }
     bool HasLeftBracket = false;
-    if (!S.empty() && S.front() == '[') {
+    if (S.consume_front("["))
       HasLeftBracket = true;
-      S = S.drop_front();
-    }
     unsigned long long N;
     if (S.empty() || consumeUnsignedInteger(S, 10, N))
       return false;
-    if (!S.empty() && S.front() == ':') {
+    if (S.consume_front(":")) {
       if (!HasLeftBracket)
         return false;
-      S = S.drop_front();
       unsigned long long M;
       if (consumeUnsignedInteger(S, 10, M) || N >= M)
         return false;
     }
     if (HasLeftBracket) {
-      if (S.empty() || S.front() != ']')
+      if (!S.consume_front("]"))
         return false;
-      S = S.drop_front();
     }
-    if (S.empty() || S.front() != '}')
+    if (!S.consume_front("}"))
       return false;
-    S = S.drop_front();
     if (!S.empty())
       return false;
     // Found {vn}, {sn}, {an}, {v[n]}, {s[n]}, {a[n]}, {v[n:m]}, {s[n:m]}
diff --git a/clang/lib/Basic/Targets/OSTargets.h b/clang/lib/Basic/Targets/OSTargets.h
index 342af4bbc42b7..4366c1149e405 100644
--- a/clang/lib/Basic/Targets/OSTargets.h
+++ b/clang/lib/Basic/Targets/OSTargets.h
@@ -74,7 +74,8 @@ class LLVM_LIBRARY_VISIBILITY DarwinTargetInfo : public OSTargetInfo {
         this->TLSSupported = !Triple.isOSVersionLT(3);
     } else if (Triple.isDriverKit()) {
       // No TLS on DriverKit.
-    }
+    } else if (Triple.isXROS())
+      this->TLSSupported = true;
 
     this->MCountName = "\01mcount";
   }
@@ -109,6 +110,9 @@ class LLVM_LIBRARY_VISIBILITY DarwinTargetInfo : public OSTargetInfo {
     case llvm::Triple::WatchOS: // Earliest supporting version is 5.0.0.
       MinVersion = llvm::VersionTuple(5U);
       break;
+    case llvm::Triple::XROS:
+      MinVersion = llvm::VersionTuple(0);
+      break;
     default:
       // Conservatively return 8 bytes if OS is unknown.
       return 64;
diff --git a/clang/lib/Basic/Targets/PPC.cpp b/clang/lib/Basic/Targets/PPC.cpp
index 045c273f03c7a..41935abfb65d3 100644
--- a/clang/lib/Basic/Targets/PPC.cpp
+++ b/clang/lib/Basic/Targets/PPC.cpp
@@ -212,6 +212,7 @@ static void defineXLCompatMacros(MacroBuilder &Builder) {
   Builder.defineMacro("__darn_32", "__builtin_darn_32");
   Builder.defineMacro("__darn_raw", "__builtin_darn_raw");
   Builder.defineMacro("__dcbf", "__builtin_dcbf");
+  Builder.defineMacro("__fence", "__builtin_ppc_fence");
   Builder.defineMacro("__fmadd", "__builtin_fma");
   Builder.defineMacro("__fmadds", "__builtin_fmaf");
   Builder.defineMacro("__abs", "__builtin_abs");
diff --git a/clang/lib/Basic/Targets/RISCV.cpp b/clang/lib/Basic/Targets/RISCV.cpp
index daaa8639ae835..c71b2e9eeb6c1 100644
--- a/clang/lib/Basic/Targets/RISCV.cpp
+++ b/clang/lib/Basic/Targets/RISCV.cpp
@@ -154,7 +154,7 @@ void RISCVTargetInfo::getTargetDefines(const LangOptions &Opts,
   else
     Builder.defineMacro("__riscv_float_abi_soft");
 
-  if (ABIName == "ilp32e")
+  if (ABIName == "ilp32e" || ABIName == "lp64e")
     Builder.defineMacro("__riscv_abi_rve");
 
   Builder.defineMacro("__riscv_arch_test");
@@ -163,9 +163,8 @@ void RISCVTargetInfo::getTargetDefines(const LangOptions &Opts,
     auto ExtName = Extension.first;
     auto ExtInfo = Extension.second;
 
-    Builder.defineMacro(
-        Twine("__riscv_", ExtName),
-        Twine(getVersionValue(ExtInfo.MajorVersion, ExtInfo.MinorVersion)));
+    Builder.defineMacro(Twine("__riscv_", ExtName),
+                        Twine(getVersionValue(ExtInfo.Major, ExtInfo.Minor)));
   }
 
   if (ISAInfo->hasExtension("m") || ISAInfo->hasExtension("zmmul"))
@@ -215,6 +214,13 @@ void RISCVTargetInfo::getTargetDefines(const LangOptions &Opts,
     Builder.defineMacro("__riscv_misaligned_fast");
   else
     Builder.defineMacro("__riscv_misaligned_avoid");
+
+  if (ISAInfo->hasExtension("e")) {
+    if (Is64Bit)
+      Builder.defineMacro("__riscv_64e");
+    else
+      Builder.defineMacro("__riscv_32e");
+  }
 }
 
 static constexpr Builtin::Info BuiltinInfo[] = {
@@ -235,39 +241,6 @@ ArrayRef RISCVTargetInfo::getTargetBuiltins() const {
                         clang::RISCV::LastTSBuiltin - Builtin::FirstTSBuiltin);
 }
 
-static std::vector
-collectNonISAExtFeature(ArrayRef FeaturesNeedOverride, int XLen) {
-  std::vector NonISAExtFeatureVec;
-
-  auto IsNonISAExtFeature = [](const std::string &Feature) {
-    assert(Feature.size() > 1 && (Feature[0] == '+' || Feature[0] == '-'));
-    StringRef Ext = StringRef(Feature).drop_front(); // drop the +/-
-    return !llvm::RISCVISAInfo::isSupportedExtensionFeature(Ext);
-  };
-  llvm::copy_if(FeaturesNeedOverride, std::back_inserter(NonISAExtFeatureVec),
-                IsNonISAExtFeature);
-
-  return NonISAExtFeatureVec;
-}
-
-static std::vector
-resolveTargetAttrOverride(const std::vector &FeaturesVec,
-                          int XLen) {
-  auto I = llvm::find(FeaturesVec, "__RISCV_TargetAttrNeedOverride");
-  if (I == FeaturesVec.end())
-    return FeaturesVec;
-
-  ArrayRef FeaturesNeedOverride(&*FeaturesVec.begin(), &*I);
-  std::vector NonISAExtFeature =
-      collectNonISAExtFeature(FeaturesNeedOverride, XLen);
-
-  std::vector ResolvedFeature(++I, FeaturesVec.end());
-  ResolvedFeature.insert(ResolvedFeature.end(), NonISAExtFeature.begin(),
-                         NonISAExtFeature.end());
-
-  return ResolvedFeature;
-}
-
 bool RISCVTargetInfo::initFeatureMap(
     llvm::StringMap &Features, DiagnosticsEngine &Diags, StringRef CPU,
     const std::vector &FeaturesVec) const {
@@ -281,10 +254,27 @@ bool RISCVTargetInfo::initFeatureMap(
     Features["32bit"] = true;
   }
 
-  std::vector NewFeaturesVec =
-      resolveTargetAttrOverride(FeaturesVec, XLen);
+  // If a target attribute specified a full arch string, override all the ISA
+  // extension target features.
+  const auto I = llvm::find(FeaturesVec, "__RISCV_TargetAttrNeedOverride");
+  if (I != FeaturesVec.end()) {
+    std::vector OverrideFeatures(std::next(I), FeaturesVec.end());
+
+    // Add back any non ISA extension features, e.g. +relax.
+    auto IsNonISAExtFeature = [](StringRef Feature) {
+      assert(Feature.size() > 1 && (Feature[0] == '+' || Feature[0] == '-'));
+      StringRef Ext = Feature.substr(1); // drop the +/-
+      return !llvm::RISCVISAInfo::isSupportedExtensionFeature(Ext);
+    };
+    llvm::copy_if(llvm::make_range(FeaturesVec.begin(), I),
+                  std::back_inserter(OverrideFeatures), IsNonISAExtFeature);
 
-  auto ParseResult = llvm::RISCVISAInfo::parseFeatures(XLen, NewFeaturesVec);
+    return TargetInfo::initFeatureMap(Features, Diags, CPU, OverrideFeatures);
+  }
+
+  // Otherwise, parse the features and add any implied extensions.
+  std::vector AllFeatures = FeaturesVec;
+  auto ParseResult = llvm::RISCVISAInfo::parseFeatures(XLen, FeaturesVec);
   if (!ParseResult) {
     std::string Buffer;
     llvm::raw_string_ostream OutputErrMsg(Buffer);
@@ -295,21 +285,9 @@ bool RISCVTargetInfo::initFeatureMap(
     return false;
   }
 
-  // RISCVISAInfo makes implications for ISA features
-  std::vector ImpliedFeatures = (*ParseResult)->toFeatures();
-
-  // parseFeatures normalizes the feature set by dropping any explicit
-  // negatives, and non-extension features.  We need to preserve the later
-  // for correctness and want to preserve the former for consistency.
-  for (auto &Feature : NewFeaturesVec) {
-     StringRef ExtName = Feature;
-     assert(ExtName.size() > 1 && (ExtName[0] == '+' || ExtName[0] == '-'));
-     ExtName = ExtName.drop_front(1); // Drop '+' or '-'
-     if (!llvm::is_contained(ImpliedFeatures, ("+" + ExtName).str()) &&
-         !llvm::is_contained(ImpliedFeatures, ("-" + ExtName).str()))
-       ImpliedFeatures.push_back(Feature);
-  }
-  return TargetInfo::initFeatureMap(Features, Diags, CPU, ImpliedFeatures);
+  // Append all features, not just new ones, so we override any negatives.
+  llvm::append_range(AllFeatures, (*ParseResult)->toFeatures());
+  return TargetInfo::initFeatureMap(Features, Diags, CPU, AllFeatures);
 }
 
 std::optional>
@@ -379,6 +357,11 @@ bool RISCVTargetInfo::handleTargetFeatures(std::vector &Features,
   if (llvm::is_contained(Features, "+experimental"))
     HasExperimental = true;
 
+  if (ABI == "ilp32e" && ISAInfo->hasExtension("d")) {
+    Diags.Report(diag::err_invalid_feature_combination)
+        << "ILP32E cannot be used with the D ISA extension";
+    return false;
+  }
   return true;
 }
 
@@ -413,7 +396,10 @@ static void handleFullArchString(StringRef FullArchStr,
     // Forward the invalid FullArchStr.
     Features.push_back("+" + FullArchStr.str());
   } else {
-    std::vector FeatStrings = (*RII)->toFeatures();
+    // Append a full list of features, including any negative extensions so that
+    // we override the CPU's features.
+    std::vector FeatStrings =
+        (*RII)->toFeatures(/* AddAllExtensions */ true);
     Features.insert(Features.end(), FeatStrings.begin(), FeatStrings.end());
   }
 }
diff --git a/clang/lib/Basic/Targets/RISCV.h b/clang/lib/Basic/Targets/RISCV.h
index f98c88cd45f83..bfbdafb682c85 100644
--- a/clang/lib/Basic/Targets/RISCV.h
+++ b/clang/lib/Basic/Targets/RISCV.h
@@ -132,6 +132,12 @@ class LLVM_LIBRARY_VISIBILITY RISCV32TargetInfo : public RISCVTargetInfo {
   }
 
   bool setABI(const std::string &Name) override {
+    if (Name == "ilp32e") {
+      ABI = Name;
+      resetDataLayout("e-m:e-p:32:32-i64:64-n32-S32");
+      return true;
+    }
+
     if (Name == "ilp32" || Name == "ilp32f" || Name == "ilp32d") {
       ABI = Name;
       return true;
@@ -156,6 +162,12 @@ class LLVM_LIBRARY_VISIBILITY RISCV64TargetInfo : public RISCVTargetInfo {
   }
 
   bool setABI(const std::string &Name) override {
+    if (Name == "lp64e") {
+      ABI = Name;
+      resetDataLayout("e-m:e-p:64:64-i64:64-i128:128-n32:64-S64");
+      return true;
+    }
+
     if (Name == "lp64" || Name == "lp64f" || Name == "lp64d") {
       ABI = Name;
       return true;
diff --git a/clang/lib/Basic/Targets/X86.cpp b/clang/lib/Basic/Targets/X86.cpp
index 64e281b888a95..a68b662d9401a 100644
--- a/clang/lib/Basic/Targets/X86.cpp
+++ b/clang/lib/Basic/Targets/X86.cpp
@@ -1418,6 +1418,14 @@ bool X86TargetInfo::validateAsmConstraint(
   case 'O':
     Info.setRequiresImmediate(0, 127);
     return true;
+  case 'W':
+    switch (*++Name) {
+    default:
+      return false;
+    case 's':
+      Info.setAllowsRegister();
+      return true;
+    }
   // Register constraints.
   case 'Y': // 'Y' is the first character for several 2-character constraints.
     // Shift the pointer to the second character of the constraint.
@@ -1715,6 +1723,9 @@ std::string X86TargetInfo::convertConstraint(const char *&Constraint) const {
     return std::string("{st}");
   case 'u':                        // second from top of floating point stack.
     return std::string("{st(1)}"); // second from top of floating point stack.
+  case 'W':
+    assert(Constraint[1] == 's');
+    return '^' + std::string(Constraint++, 2);
   case 'Y':
     switch (Constraint[1]) {
     default:
diff --git a/clang/lib/Basic/Warnings.cpp b/clang/lib/Basic/Warnings.cpp
index 92954cab6fb04..5a5ac55563388 100644
--- a/clang/lib/Basic/Warnings.cpp
+++ b/clang/lib/Basic/Warnings.cpp
@@ -199,8 +199,7 @@ void clang::ProcessWarningOptions(DiagnosticsEngine &Diags,
 
       // Check to see if this warning starts with "no-", if so, this is a
       // negative form of the option.
-      bool IsPositive = !Opt.starts_with("no-");
-      if (!IsPositive) Opt = Opt.substr(3);
+      bool IsPositive = !Opt.consume_front("no-");
 
       auto Severity = IsPositive ? diag::Severity::Remark
                                  : diag::Severity::Ignored;
diff --git a/clang/lib/CodeGen/CGBuiltin.cpp b/clang/lib/CodeGen/CGBuiltin.cpp
index f4c8f1002e0b4..48ff2113618c0 100644
--- a/clang/lib/CodeGen/CGBuiltin.cpp
+++ b/clang/lib/CodeGen/CGBuiltin.cpp
@@ -1096,7 +1096,7 @@ CodeGenFunction::emitFlexibleArrayMemberSize(const Expr *E, unsigned Type,
   FAMSize = Builder.CreateIntCast(FAMSize, ResType, IsSigned);
   Value *Res = FAMSize;
 
-  if (const auto *DRE = dyn_cast(Base)) {
+  if (isa(Base)) {
     // The whole struct is specificed in the __bdos.
     const ASTRecordLayout &Layout = Ctx.getASTRecordLayout(OuterRD);
 
@@ -13381,7 +13381,7 @@ Value *CodeGenFunction::EmitBPFBuiltinExpr(unsigned BuiltinID,
     const auto *DR = cast(CE->getSubExpr());
     const auto *Enumerator = cast(DR->getDecl());
 
-    auto &InitVal = Enumerator->getInitVal();
+    auto InitVal = Enumerator->getInitVal();
     std::string InitValStr;
     if (InitVal.isNegative() || InitVal > uint64_t(INT64_MAX))
       InitValStr = std::to_string(InitVal.getSExtValue());
@@ -18275,6 +18275,45 @@ Value *CodeGenFunction::EmitAMDGPUBuiltinExpr(unsigned BuiltinID,
     llvm::Function *F = CGM.getIntrinsic(IID, {ArgTy});
     return Builder.CreateCall(F, {Addr, Val, ZeroI32, ZeroI32, ZeroI1});
   }
+  case AMDGPU::BI__builtin_amdgcn_global_load_tr_i32:
+  case AMDGPU::BI__builtin_amdgcn_global_load_tr_v2i32:
+  case AMDGPU::BI__builtin_amdgcn_global_load_tr_v4f16:
+  case AMDGPU::BI__builtin_amdgcn_global_load_tr_v4i16:
+  case AMDGPU::BI__builtin_amdgcn_global_load_tr_v8f16:
+  case AMDGPU::BI__builtin_amdgcn_global_load_tr_v8i16: {
+
+    llvm::Type *ArgTy;
+    switch (BuiltinID) {
+    case AMDGPU::BI__builtin_amdgcn_global_load_tr_i32:
+      ArgTy = llvm::Type::getInt32Ty(getLLVMContext());
+      break;
+    case AMDGPU::BI__builtin_amdgcn_global_load_tr_v2i32:
+      ArgTy = llvm::FixedVectorType::get(
+          llvm::Type::getInt32Ty(getLLVMContext()), 2);
+      break;
+    case AMDGPU::BI__builtin_amdgcn_global_load_tr_v4f16:
+      ArgTy = llvm::FixedVectorType::get(
+          llvm::Type::getHalfTy(getLLVMContext()), 4);
+      break;
+    case AMDGPU::BI__builtin_amdgcn_global_load_tr_v4i16:
+      ArgTy = llvm::FixedVectorType::get(
+          llvm::Type::getInt16Ty(getLLVMContext()), 4);
+      break;
+    case AMDGPU::BI__builtin_amdgcn_global_load_tr_v8f16:
+      ArgTy = llvm::FixedVectorType::get(
+          llvm::Type::getHalfTy(getLLVMContext()), 8);
+      break;
+    case AMDGPU::BI__builtin_amdgcn_global_load_tr_v8i16:
+      ArgTy = llvm::FixedVectorType::get(
+          llvm::Type::getInt16Ty(getLLVMContext()), 8);
+      break;
+    }
+
+    llvm::Value *Addr = EmitScalarExpr(E->getArg(0));
+    llvm::Function *F =
+        CGM.getIntrinsic(Intrinsic::amdgcn_global_load_tr, {ArgTy});
+    return Builder.CreateCall(F, {Addr});
+  }
   case AMDGPU::BI__builtin_amdgcn_read_exec:
     return EmitAMDGCNBallotForExec(*this, E, Int64Ty, Int64Ty, false);
   case AMDGPU::BI__builtin_amdgcn_read_exec_lo:
diff --git a/clang/lib/CodeGen/CGCall.cpp b/clang/lib/CodeGen/CGCall.cpp
index 7fce654e047ab..4507be1d26f58 100644
--- a/clang/lib/CodeGen/CGCall.cpp
+++ b/clang/lib/CodeGen/CGCall.cpp
@@ -1776,14 +1776,21 @@ static void AddAttributesFromFunctionProtoType(ASTContext &Ctx,
       FPT->isNothrow())
     FuncAttrs.addAttribute(llvm::Attribute::NoUnwind);
 
-  if (FPT->getAArch64SMEAttributes() & FunctionType::SME_PStateSMEnabledMask)
+  unsigned SMEBits = FPT->getAArch64SMEAttributes();
+  if (SMEBits & FunctionType::SME_PStateSMEnabledMask)
     FuncAttrs.addAttribute("aarch64_pstate_sm_enabled");
-  if (FPT->getAArch64SMEAttributes() & FunctionType::SME_PStateSMCompatibleMask)
+  if (SMEBits & FunctionType::SME_PStateSMCompatibleMask)
     FuncAttrs.addAttribute("aarch64_pstate_sm_compatible");
-  if (FPT->getAArch64SMEAttributes() & FunctionType::SME_PStateZASharedMask)
+
+  // ZA
+  if (FunctionType::getArmZAState(SMEBits) == FunctionType::ARM_Out ||
+      FunctionType::getArmZAState(SMEBits) == FunctionType::ARM_InOut)
+    FuncAttrs.addAttribute("aarch64_pstate_za_shared");
+  if (FunctionType::getArmZAState(SMEBits) == FunctionType::ARM_Preserves ||
+      FunctionType::getArmZAState(SMEBits) == FunctionType::ARM_In) {
     FuncAttrs.addAttribute("aarch64_pstate_za_shared");
-  if (FPT->getAArch64SMEAttributes() & FunctionType::SME_PStateZAPreservedMask)
     FuncAttrs.addAttribute("aarch64_pstate_za_preserved");
+  }
 }
 
 static void AddAttributesFromAssumes(llvm::AttrBuilder &FuncAttrs,
@@ -2509,9 +2516,6 @@ void CodeGenModule::ConstructAttributeList(StringRef Name,
 
     if (TargetDecl->hasAttr())
       FuncAttrs.addAttribute("aarch64_pstate_sm_body");
-
-    if (TargetDecl->hasAttr())
-      FuncAttrs.addAttribute("aarch64_pstate_za_new");
   }
 
   // Attach "no-builtins" attributes to:
@@ -2675,6 +2679,8 @@ void CodeGenModule::ConstructAttributeList(StringRef Name,
   if (IRFunctionArgs.hasSRetArg()) {
     llvm::AttrBuilder SRETAttrs(getLLVMContext());
     SRETAttrs.addStructRetAttr(getTypes().ConvertTypeForMem(RetTy));
+    SRETAttrs.addAttribute(llvm::Attribute::Writable);
+    SRETAttrs.addAttribute(llvm::Attribute::DeadOnUnwind);
     hasUsedSRet = true;
     if (RetAI.getInReg())
       SRETAttrs.addAttribute(llvm::Attribute::InReg);
diff --git a/clang/lib/CodeGen/CGCleanup.h b/clang/lib/CodeGen/CGCleanup.h
index 079a3e25d6dc7..fcfbf41b0eaff 100644
--- a/clang/lib/CodeGen/CGCleanup.h
+++ b/clang/lib/CodeGen/CGCleanup.h
@@ -613,6 +613,7 @@ struct EHPersonality {
   static const EHPersonality MSVC_CxxFrameHandler3;
   static const EHPersonality GNU_Wasm_CPlusPlus;
   static const EHPersonality XL_CPlusPlus;
+  static const EHPersonality ZOS_CPlusPlus;
 
   /// Does this personality use landingpads or the family of pad instructions
   /// designed to form funclets?
diff --git a/clang/lib/CodeGen/CGDebugInfo.cpp b/clang/lib/CodeGen/CGDebugInfo.cpp
index cda279a39f0bb..3be77912ab75a 100644
--- a/clang/lib/CodeGen/CGDebugInfo.cpp
+++ b/clang/lib/CodeGen/CGDebugInfo.cpp
@@ -1980,7 +1980,7 @@ llvm::DISubprogram *CGDebugInfo::CreateCXXMemberFunction(
   int ThisAdjustment = 0;
 
   if (VTableContextBase::hasVtableSlot(Method)) {
-    if (Method->isPure())
+    if (Method->isPureVirtual())
       SPFlags |= llvm::DISubprogram::SPFlagPureVirtual;
     else
       SPFlags |= llvm::DISubprogram::SPFlagVirtual;
diff --git a/clang/lib/CodeGen/CGException.cpp b/clang/lib/CodeGen/CGException.cpp
index 2150c19eace8b..f222b6e7a43c6 100644
--- a/clang/lib/CodeGen/CGException.cpp
+++ b/clang/lib/CodeGen/CGException.cpp
@@ -127,6 +127,8 @@ const EHPersonality
 EHPersonality::GNU_Wasm_CPlusPlus = { "__gxx_wasm_personality_v0", nullptr };
 const EHPersonality EHPersonality::XL_CPlusPlus = {"__xlcxx_personality_v1",
                                                    nullptr};
+const EHPersonality EHPersonality::ZOS_CPlusPlus = {"__zos_cxx_personality_v2",
+                                                    nullptr};
 
 static const EHPersonality &getCPersonality(const TargetInfo &Target,
                                             const LangOptions &L) {
@@ -187,6 +189,8 @@ static const EHPersonality &getCXXPersonality(const TargetInfo &Target,
     return EHPersonality::GNU_CPlusPlus_SEH;
   if (L.hasWasmExceptions())
     return EHPersonality::GNU_Wasm_CPlusPlus;
+  if (T.isOSzOS())
+    return EHPersonality::ZOS_CPlusPlus;
   return EHPersonality::GNU_CPlusPlus;
 }
 
diff --git a/clang/lib/CodeGen/CGExpr.cpp b/clang/lib/CodeGen/CGExpr.cpp
index 1a453429f78ce..4e36596568326 100644
--- a/clang/lib/CodeGen/CGExpr.cpp
+++ b/clang/lib/CodeGen/CGExpr.cpp
@@ -4258,6 +4258,12 @@ LValue CodeGenFunction::EmitArraySubscriptExpr(const ArraySubscriptExpr *E,
       ArrayLV = EmitLValue(Array);
     auto *Idx = EmitIdxAfterBase(/*Promote*/true);
 
+    const ValueDecl *ArrayDecl = nullptr;
+    if (const auto *DRE = dyn_cast(Array->IgnoreParenCasts()))
+      ArrayDecl = DRE->getDecl();
+    else if (const auto *ME = dyn_cast(Array->IgnoreParenCasts()))
+      ArrayDecl = ME->getMemberDecl();
+
     if (SanOpts.has(SanitizerKind::ArrayBounds)) {
       // If the array being accessed has a "counted_by" attribute, generate
       // bounds checking code. The "count" field is at the top level of the
@@ -4299,12 +4305,6 @@ LValue CodeGenFunction::EmitArraySubscriptExpr(const ArraySubscriptExpr *E,
       }
     }
 
-    const ValueDecl *ArrayDecl = nullptr;
-    if (const auto *DRE = dyn_cast(Array->IgnoreParenCasts()))
-      ArrayDecl = DRE->getDecl();
-    else if (const auto *ME = dyn_cast(Array->IgnoreParenCasts()))
-      ArrayDecl = ME->getMemberDecl();
-
     // Propagate the alignment from the array itself to the result.
     QualType arrayType = Array->getType();
     Addr = emitArraySubscriptGEP(
diff --git a/clang/lib/CodeGen/CGObjC.cpp b/clang/lib/CodeGen/CGObjC.cpp
index b2b296c41a256..ab7ea1da63828 100644
--- a/clang/lib/CodeGen/CGObjC.cpp
+++ b/clang/lib/CodeGen/CGObjC.cpp
@@ -3942,6 +3942,8 @@ static unsigned getBaseMachOPlatformID(const llvm::Triple &TT) {
     return llvm::MachO::PLATFORM_TVOS;
   case llvm::Triple::WatchOS:
     return llvm::MachO::PLATFORM_WATCHOS;
+  case llvm::Triple::XROS:
+    return llvm::MachO::PLATFORM_XROS;
   case llvm::Triple::DriverKit:
     return llvm::MachO::PLATFORM_DRIVERKIT;
   default:
@@ -4025,6 +4027,9 @@ static bool isFoundationNeededForDarwinAvailabilityCheck(
   case llvm::Triple::MacOSX:
     FoundationDroppedInVersion = VersionTuple(/*Major=*/10, /*Minor=*/15);
     break;
+  case llvm::Triple::XROS:
+    // XROS doesn't need Foundation.
+    return false;
   case llvm::Triple::DriverKit:
     // DriverKit doesn't need Foundation.
     return false;
diff --git a/clang/lib/CodeGen/CGObjCGNU.cpp b/clang/lib/CodeGen/CGObjCGNU.cpp
index cd1a0b6a130ff..9cc7f32815f7e 100644
--- a/clang/lib/CodeGen/CGObjCGNU.cpp
+++ b/clang/lib/CodeGen/CGObjCGNU.cpp
@@ -1431,12 +1431,24 @@ class CGObjCGNUstep2 : public CGObjCGNUstep {
                                 const std::string &TypeEncoding) override {
     return GetConstantSelector(Sel, TypeEncoding);
   }
+  std::string GetSymbolNameForTypeEncoding(const std::string &TypeEncoding) {
+    std::string MangledTypes = std::string(TypeEncoding);
+    // @ is used as a special character in ELF symbol names (used for symbol
+    // versioning), so mangle the name to not include it.  Replace it with a
+    // character that is not a valid type encoding character (and, being
+    // non-printable, never will be!)
+    if (CGM.getTriple().isOSBinFormatELF())
+      std::replace(MangledTypes.begin(), MangledTypes.end(), '@', '\1');
+    // = in dll exported names causes lld to fail when linking on Windows.
+    if (CGM.getTriple().isOSWindows())
+      std::replace(MangledTypes.begin(), MangledTypes.end(), '=', '\2');
+    return MangledTypes;
+  }
   llvm::Constant  *GetTypeString(llvm::StringRef TypeEncoding) {
     if (TypeEncoding.empty())
       return NULLPtr;
-    std::string MangledTypes = std::string(TypeEncoding);
-    std::replace(MangledTypes.begin(), MangledTypes.end(),
-      '@', '\1');
+    std::string MangledTypes =
+        GetSymbolNameForTypeEncoding(std::string(TypeEncoding));
     std::string TypesVarName = ".objc_sel_types_" + MangledTypes;
     auto *TypesGlobal = TheModule.getGlobalVariable(TypesVarName);
     if (!TypesGlobal) {
@@ -1453,13 +1465,7 @@ class CGObjCGNUstep2 : public CGObjCGNUstep {
   }
   llvm::Constant *GetConstantSelector(Selector Sel,
                                       const std::string &TypeEncoding) override {
-    // @ is used as a special character in symbol names (used for symbol
-    // versioning), so mangle the name to not include it.  Replace it with a
-    // character that is not a valid type encoding character (and, being
-    // non-printable, never will be!)
-    std::string MangledTypes = TypeEncoding;
-    std::replace(MangledTypes.begin(), MangledTypes.end(),
-      '@', '\1');
+    std::string MangledTypes = GetSymbolNameForTypeEncoding(TypeEncoding);
     auto SelVarName = (StringRef(".objc_selector_") + Sel.getAsString() + "_" +
       MangledTypes).str();
     if (auto *GV = TheModule.getNamedGlobal(SelVarName))
@@ -1671,9 +1677,7 @@ class CGObjCGNUstep2 : public CGObjCGNUstep {
                                         const ObjCIvarDecl *Ivar) override {
     std::string TypeEncoding;
     CGM.getContext().getObjCEncodingForType(Ivar->getType(), TypeEncoding);
-    // Prevent the @ from being interpreted as a symbol version.
-    std::replace(TypeEncoding.begin(), TypeEncoding.end(),
-      '@', '\1');
+    TypeEncoding = GetSymbolNameForTypeEncoding(TypeEncoding);
     const std::string Name = "__objc_ivar_offset_" + ID->getNameAsString()
       + '.' + Ivar->getNameAsString() + '.' + TypeEncoding;
     return Name;
diff --git a/clang/lib/CodeGen/CGStmt.cpp b/clang/lib/CodeGen/CGStmt.cpp
index b89017de0bcf1..beff0ad9da270 100644
--- a/clang/lib/CodeGen/CGStmt.cpp
+++ b/clang/lib/CodeGen/CGStmt.cpp
@@ -2399,9 +2399,9 @@ EmitAsmStores(CodeGenFunction &CGF, const AsmStmt &S,
         Tmp = Builder.CreatePtrToInt(
             Tmp, llvm::IntegerType::get(CTX, (unsigned)TmpSize));
         Tmp = Builder.CreateTrunc(Tmp, TruncTy);
-      } else if (TruncTy->isIntegerTy()) {
+      } else if (Tmp->getType()->isIntegerTy() && TruncTy->isIntegerTy()) {
         Tmp = Builder.CreateZExtOrTrunc(Tmp, TruncTy);
-      } else if (TruncTy->isVectorTy()) {
+      } else if (Tmp->getType()->isVectorTy() || TruncTy->isVectorTy()) {
         Tmp = Builder.CreateBitCast(Tmp, TruncTy);
       }
     }
diff --git a/clang/lib/CodeGen/CGVTables.cpp b/clang/lib/CodeGen/CGVTables.cpp
index 05cabb8e7d961..913c24bc7e859 100644
--- a/clang/lib/CodeGen/CGVTables.cpp
+++ b/clang/lib/CodeGen/CGVTables.cpp
@@ -792,7 +792,7 @@ void CodeGenVTables::addVTableComponent(ConstantArrayBuilder &builder,
     llvm::Constant *fnPtr;
 
     // Pure virtual member functions.
-    if (cast(GD.getDecl())->isPure()) {
+    if (cast(GD.getDecl())->isPureVirtual()) {
       if (!PureVirtualFn)
         PureVirtualFn =
             getSpecialVirtualFn(CGM.getCXXABI().GetPureVirtualCallName());
diff --git a/clang/lib/CodeGen/CodeGenModule.cpp b/clang/lib/CodeGen/CodeGenModule.cpp
index 5474f35fe1a2f..b6e38e35f1e0d 100644
--- a/clang/lib/CodeGen/CodeGenModule.cpp
+++ b/clang/lib/CodeGen/CodeGenModule.cpp
@@ -240,7 +240,8 @@ createTargetCodeGenInfo(CodeGenModule &CGM) {
       ABIFLen = 32;
     else if (ABIStr.ends_with("d"))
       ABIFLen = 64;
-    return createRISCVTargetCodeGenInfo(CGM, XLen, ABIFLen);
+    bool EABI = ABIStr.ends_with("e");
+    return createRISCVTargetCodeGenInfo(CGM, XLen, ABIFLen, EABI);
   }
 
   case llvm::Triple::systemz: {
@@ -1152,6 +1153,8 @@ void CodeGenModule::Release() {
     if (LangOpts.BranchProtectionPAuthLR)
       getModule().addModuleFlag(llvm::Module::Min, "branch-protection-pauth-lr",
                                 1);
+    if (LangOpts.GuardedControlStack)
+      getModule().addModuleFlag(llvm::Module::Min, "guarded-control-stack", 1);
     if (LangOpts.hasSignReturnAddress())
       getModule().addModuleFlag(llvm::Module::Min, "sign-return-address", 1);
     if (LangOpts.isSignReturnAddressScopeAll())
@@ -1299,7 +1302,7 @@ void CodeGenModule::Release() {
       llvm::CodeModel::Model codeModel = static_cast(CM);
       getModule().setCodeModel(codeModel);
 
-      if (CM == llvm::CodeModel::Medium &&
+      if ((CM == llvm::CodeModel::Medium || CM == llvm::CodeModel::Large) &&
           Context.getTargetInfo().getTriple().getArch() ==
               llvm::Triple::x86_64) {
         getModule().setLargeDataThreshold(getCodeGenOpts().LargeDataThreshold);
@@ -2564,8 +2567,10 @@ void CodeGenModule::SetLLVMFunctionAttributesForDefinition(const Decl *D,
   if (D->hasAttr())
     B.addAttribute("aarch64_pstate_sm_body");
 
-  if (D->hasAttr())
-    B.addAttribute("aarch64_pstate_za_new");
+  if (auto *Attr = D->getAttr()) {
+    if (Attr->isNewZA())
+      B.addAttribute("aarch64_pstate_za_new");
+  }
 
   // Track whether we need to add the optnone LLVM attribute,
   // starting with the default for this optimization level.
diff --git a/clang/lib/CodeGen/CoverageMappingGen.cpp b/clang/lib/CodeGen/CoverageMappingGen.cpp
index b245abd16c3f4..4a44d113ddec9 100644
--- a/clang/lib/CodeGen/CoverageMappingGen.cpp
+++ b/clang/lib/CodeGen/CoverageMappingGen.cpp
@@ -573,6 +573,11 @@ struct EmptyCoverageMappingBuilder : public CoverageMappingBuilder {
 /// creation.
 struct MCDCCoverageBuilder {
 
+  struct DecisionIDPair {
+    MCDCConditionID TrueID = 0;
+    MCDCConditionID FalseID = 0;
+  };
+
   /// The AST walk recursively visits nested logical-AND or logical-OR binary
   /// operator nodes and then visits their LHS and RHS children nodes.  As this
   /// happens, the algorithm will assign IDs to each operator's LHS and RHS side
@@ -616,14 +621,14 @@ struct MCDCCoverageBuilder {
   ///
   /// A node ID of '0' always means MC/DC isn't being tracked.
   ///
-  /// As the AST walk proceeds recursively, the algorithm will also use stacks
+  /// As the AST walk proceeds recursively, the algorithm will also use a stack
   /// to track the IDs of logical-AND and logical-OR operations on the RHS so
   /// that it can be determined which nodes are executed next, depending on how
   /// a LHS or RHS of a logical-AND or logical-OR is evaluated.  This
   /// information relies on the assigned IDs and are embedded within the
   /// coverage region IDs of each branch region associated with a leaf-level
   /// condition. This information helps the visualization tool reconstruct all
-  /// possible test vectors for the purposes of MC/DC analysis. if a "next" node
+  /// possible test vectors for the purposes of MC/DC analysis. If a "next" node
   /// ID is '0', it means it's the end of the test vector. The following rules
   /// are used:
   ///
@@ -663,54 +668,40 @@ struct MCDCCoverageBuilder {
 private:
   CodeGenModule &CGM;
 
-  llvm::SmallVector AndRHS;
-  llvm::SmallVector OrRHS;
-  llvm::SmallVector NestLevel;
+  llvm::SmallVector DecisionStack;
   llvm::DenseMap &CondIDs;
   llvm::DenseMap &MCDCBitmapMap;
   MCDCConditionID NextID = 1;
   bool NotMapped = false;
 
+  /// Represent a sentinel value of [0,0] for the bottom of DecisionStack.
+  static constexpr DecisionIDPair DecisionStackSentinel{0, 0};
+
   /// Is this a logical-AND operation?
   bool isLAnd(const BinaryOperator *E) const {
     return E->getOpcode() == BO_LAnd;
   }
 
-  /// Push an ID onto the corresponding RHS stack.
-  void pushRHS(const BinaryOperator *E) {
-    llvm::SmallVector &rhs = isLAnd(E) ? AndRHS : OrRHS;
-    rhs.push_back(CondIDs[CodeGenFunction::stripCond(E->getRHS())]);
-  }
-
-  /// Pop an ID from the corresponding RHS stack.
-  void popRHS(const BinaryOperator *E) {
-    llvm::SmallVector &rhs = isLAnd(E) ? AndRHS : OrRHS;
-    if (!rhs.empty())
-      rhs.pop_back();
-  }
-
-  /// If the expected ID is on top, pop it off the corresponding RHS stack.
-  void popRHSifTop(const BinaryOperator *E) {
-    if (!OrRHS.empty() && CondIDs[E] == OrRHS.back())
-      OrRHS.pop_back();
-    else if (!AndRHS.empty() && CondIDs[E] == AndRHS.back())
-      AndRHS.pop_back();
-  }
-
 public:
   MCDCCoverageBuilder(CodeGenModule &CGM,
                       llvm::DenseMap &CondIDMap,
                       llvm::DenseMap &MCDCBitmapMap)
-      : CGM(CGM), CondIDs(CondIDMap), MCDCBitmapMap(MCDCBitmapMap) {}
+      : CGM(CGM), DecisionStack(1, DecisionStackSentinel), CondIDs(CondIDMap),
+        MCDCBitmapMap(MCDCBitmapMap) {}
 
-  /// Return the ID of the RHS of the next, upper nest-level logical-OR.
-  MCDCConditionID getNextLOrCondID() const {
-    return OrRHS.empty() ? 0 : OrRHS.back();
-  }
+  /// Return whether the build of the control flow map is at the top-level
+  /// (root) of a logical operator nest in a boolean expression prior to the
+  /// assignment of condition IDs.
+  bool isIdle() const { return (NextID == 1 && !NotMapped); }
+
+  /// Return whether any IDs have been assigned in the build of the control
+  /// flow map, indicating that the map is being generated for this boolean
+  /// expression.
+  bool isBuilding() const { return (NextID > 1); }
 
-  /// Return the ID of the RHS of the next, upper nest-level logical-AND.
-  MCDCConditionID getNextLAndCondID() const {
-    return AndRHS.empty() ? 0 : AndRHS.back();
+  /// Set the given condition's ID.
+  void setCondID(const Expr *Cond, MCDCConditionID ID) {
+    CondIDs[CodeGenFunction::stripCond(Cond)] = ID;
   }
 
   /// Return the ID of a given condition.
@@ -722,6 +713,9 @@ struct MCDCCoverageBuilder {
       return I->second;
   }
 
+  /// Return the LHS Decision ([0,0] if not set).
+  const DecisionIDPair &back() const { return DecisionStack.back(); }
+
   /// Push the binary operator statement to track the nest level and assign IDs
   /// to the operator's LHS and RHS.  The RHS may be a larger subtree that is
   /// broken up on successive levels.
@@ -730,68 +724,67 @@ struct MCDCCoverageBuilder {
       return;
 
     // If binary expression is disqualified, don't do mapping.
-    if (NestLevel.empty() && MCDCBitmapMap.find(CodeGenFunction::stripCond(
-                                 E)) == MCDCBitmapMap.end())
+    if (!isBuilding() && !MCDCBitmapMap.contains(CodeGenFunction::stripCond(E)))
       NotMapped = true;
 
-    // Push Stmt on 'NestLevel' stack to keep track of nest location.
-    NestLevel.push_back(E);
-
     // Don't go any further if we don't need to map condition IDs.
     if (NotMapped)
       return;
 
+    const DecisionIDPair &ParentDecision = DecisionStack.back();
+
     // If the operator itself has an assigned ID, this means it represents a
-    // larger subtree.  In this case, pop its ID out of the RHS stack and
-    // assign that ID to its LHS node.  Its RHS will receive a new ID.
-    if (CondIDs.find(CodeGenFunction::stripCond(E)) != CondIDs.end()) {
-      // If Stmt has an ID, assign its ID to LHS
-      CondIDs[CodeGenFunction::stripCond(E->getLHS())] = CondIDs[E];
-
-      // Since the operator's LHS assumes the operator's same ID, pop the
-      // operator from the RHS stack so that if LHS short-circuits, it won't be
-      // incorrectly re-used as the node executed next.
-      popRHSifTop(E);
-    } else {
-      // Otherwise, assign ID+1 to LHS.
-      CondIDs[CodeGenFunction::stripCond(E->getLHS())] = NextID++;
-    }
+    // larger subtree.  In this case, assign that ID to its LHS node.  Its RHS
+    // will receive a new ID below. Otherwise, assign ID+1 to LHS.
+    if (CondIDs.contains(CodeGenFunction::stripCond(E)))
+      setCondID(E->getLHS(), getCondID(E));
+    else
+      setCondID(E->getLHS(), NextID++);
 
-    // Assign ID+1 to RHS.
-    CondIDs[CodeGenFunction::stripCond(E->getRHS())] = NextID++;
+    // Assign a ID+1 for the RHS.
+    MCDCConditionID RHSid = NextID++;
+    setCondID(E->getRHS(), RHSid);
 
-    // Push ID of Stmt's RHS so that LHS nodes know about it
-    pushRHS(E);
+    // Push the LHS decision IDs onto the DecisionStack.
+    if (isLAnd(E))
+      DecisionStack.push_back({RHSid, ParentDecision.FalseID});
+    else
+      DecisionStack.push_back({ParentDecision.TrueID, RHSid});
+  }
+
+  /// Pop and return the LHS Decision ([0,0] if not set).
+  DecisionIDPair pop() {
+    if (!CGM.getCodeGenOpts().MCDCCoverage || NotMapped)
+      return DecisionStack.front();
+
+    assert(DecisionStack.size() > 1);
+    DecisionIDPair D = DecisionStack.back();
+    DecisionStack.pop_back();
+    return D;
   }
 
-  /// Pop the binary operator from the next level. If the walk is at the top of
-  /// the next, assign the total number of conditions.
-  unsigned popAndReturnCondCount(const BinaryOperator *E) {
+  /// Return the total number of conditions and reset the state. The number of
+  /// conditions is zero if the expression isn't mapped.
+  unsigned getTotalConditionsAndReset(const BinaryOperator *E) {
     if (!CGM.getCodeGenOpts().MCDCCoverage)
       return 0;
 
-    unsigned TotalConds = 0;
-
-    // Pop Stmt from 'NestLevel' stack.
-    assert(NestLevel.back() == E);
-    NestLevel.pop_back();
+    assert(!isIdle());
+    assert(DecisionStack.size() == 1);
 
     // Reset state if not doing mapping.
-    if (NestLevel.empty() && NotMapped) {
+    if (NotMapped) {
       NotMapped = false;
+      assert(NextID == 1);
       return 0;
     }
 
-    // Pop RHS ID.
-    popRHS(E);
+    // Set number of conditions and reset.
+    unsigned TotalConds = NextID - 1;
 
-    // If at the parent (NestLevel=0), set conds and reset.
-    if (NestLevel.empty()) {
-      TotalConds = NextID - 1;
+    // Reset ID back to beginning.
+    NextID = 1;
 
-      // Reset ID back to beginning.
-      NextID = 1;
-    }
     return TotalConds;
   }
 };
@@ -1018,13 +1011,15 @@ struct CounterCoverageMappingBuilder
     return (Cond->EvaluateAsInt(Result, CVM.getCodeGenModule().getContext()));
   }
 
+  using MCDCDecisionIDPair = MCDCCoverageBuilder::DecisionIDPair;
+
   /// Create a Branch Region around an instrumentable condition for coverage
   /// and add it to the function's SourceRegions.  A branch region tracks a
   /// "True" counter and a "False" counter for boolean expressions that
   /// result in the generation of a branch.
-  void createBranchRegion(const Expr *C, Counter TrueCnt, Counter FalseCnt,
-                          MCDCConditionID ID = 0, MCDCConditionID TrueID = 0,
-                          MCDCConditionID FalseID = 0) {
+  void
+  createBranchRegion(const Expr *C, Counter TrueCnt, Counter FalseCnt,
+                     const MCDCDecisionIDPair &IDPair = MCDCDecisionIDPair()) {
     // Check for NULL conditions.
     if (!C)
       return;
@@ -1034,6 +1029,10 @@ struct CounterCoverageMappingBuilder
     // function's SourceRegions) because it doesn't apply to any other source
     // code other than the Condition.
     if (CodeGenFunction::isInstrumentedCondition(C)) {
+      MCDCConditionID ID = MCDCBuilder.getCondID(C);
+      MCDCConditionID TrueID = IDPair.TrueID;
+      MCDCConditionID FalseID = IDPair.FalseID;
+
       // If a condition can fold to true or false, the corresponding branch
       // will be removed.  Create a region with both counters hard-coded to
       // zero. This allows us to visualize them in a special way.
@@ -1822,20 +1821,28 @@ struct CounterCoverageMappingBuilder
   }
 
   void VisitBinLAnd(const BinaryOperator *E) {
-    // Keep track of Binary Operator and assign MCDC condition IDs
+    bool IsRootNode = MCDCBuilder.isIdle();
+
+    // Keep track of Binary Operator and assign MCDC condition IDs.
     MCDCBuilder.pushAndAssignIDs(E);
 
     extendRegion(E->getLHS());
     propagateCounts(getRegion().getCounter(), E->getLHS());
     handleFileExit(getEnd(E->getLHS()));
 
+    // Track LHS True/False Decision.
+    const auto DecisionLHS = MCDCBuilder.pop();
+
     // Counter tracks the right hand side of a logical and operator.
     extendRegion(E->getRHS());
     propagateCounts(getRegionCounter(E), E->getRHS());
 
-    // Process Binary Operator and create MCDC Decision Region if top-level
+    // Track RHS True/False Decision.
+    const auto DecisionRHS = MCDCBuilder.back();
+
+    // Create MCDC Decision Region if at top-level (root).
     unsigned NumConds = 0;
-    if ((NumConds = MCDCBuilder.popAndReturnCondCount(E)))
+    if (IsRootNode && (NumConds = MCDCBuilder.getTotalConditionsAndReset(E)))
       createDecisionRegion(E, getRegionBitmap(E), NumConds);
 
     // Extract the RHS's Execution Counter.
@@ -1847,30 +1854,13 @@ struct CounterCoverageMappingBuilder
     // Extract the Parent Region Counter.
     Counter ParentCnt = getRegion().getCounter();
 
-    // Extract the MCDC condition IDs (returns 0 if not needed).
-    MCDCConditionID NextOrID = MCDCBuilder.getNextLOrCondID();
-    MCDCConditionID NextAndID = MCDCBuilder.getNextLAndCondID();
-    MCDCConditionID LHSid = MCDCBuilder.getCondID(E->getLHS());
-    MCDCConditionID RHSid = MCDCBuilder.getCondID(E->getRHS());
-
     // Create Branch Region around LHS condition.
-    // MC/DC: For "LHS && RHS"
-    // - If LHS is TRUE, execution goes to the RHS.
-    // - If LHS is FALSE, execution goes to the LHS of the next logical-OR.
-    //   If that does not exist, execution exits (ID == 0).
     createBranchRegion(E->getLHS(), RHSExecCnt,
-                       subtractCounters(ParentCnt, RHSExecCnt), LHSid, RHSid,
-                       NextOrID);
+                       subtractCounters(ParentCnt, RHSExecCnt), DecisionLHS);
 
     // Create Branch Region around RHS condition.
-    // MC/DC: For "LHS && RHS"
-    // - If RHS is TRUE, execution goes to LHS of the next logical-AND.
-    //   If that does not exist, execution exits (ID == 0).
-    // - If RHS is FALSE, execution goes to the LHS of the next logical-OR.
-    //   If that does not exist, execution exits (ID == 0).
     createBranchRegion(E->getRHS(), RHSTrueCnt,
-                       subtractCounters(RHSExecCnt, RHSTrueCnt), RHSid,
-                       NextAndID, NextOrID);
+                       subtractCounters(RHSExecCnt, RHSTrueCnt), DecisionRHS);
   }
 
   // Determine whether the right side of OR operation need to be visited.
@@ -1884,20 +1874,28 @@ struct CounterCoverageMappingBuilder
   }
 
   void VisitBinLOr(const BinaryOperator *E) {
-    // Keep track of Binary Operator and assign MCDC condition IDs
+    bool IsRootNode = MCDCBuilder.isIdle();
+
+    // Keep track of Binary Operator and assign MCDC condition IDs.
     MCDCBuilder.pushAndAssignIDs(E);
 
     extendRegion(E->getLHS());
     Counter OutCount = propagateCounts(getRegion().getCounter(), E->getLHS());
     handleFileExit(getEnd(E->getLHS()));
 
+    // Track LHS True/False Decision.
+    const auto DecisionLHS = MCDCBuilder.pop();
+
     // Counter tracks the right hand side of a logical or operator.
     extendRegion(E->getRHS());
     propagateCounts(getRegionCounter(E), E->getRHS());
 
-    // Process Binary Operator and create MCDC Decision Region if top-level
+    // Track RHS True/False Decision.
+    const auto DecisionRHS = MCDCBuilder.back();
+
+    // Create MCDC Decision Region if at top-level (root).
     unsigned NumConds = 0;
-    if ((NumConds = MCDCBuilder.popAndReturnCondCount(E)))
+    if (IsRootNode && (NumConds = MCDCBuilder.getTotalConditionsAndReset(E)))
       createDecisionRegion(E, getRegionBitmap(E), NumConds);
 
     // Extract the RHS's Execution Counter.
@@ -1913,28 +1911,13 @@ struct CounterCoverageMappingBuilder
     // Extract the Parent Region Counter.
     Counter ParentCnt = getRegion().getCounter();
 
-    // Extract the MCDC condition IDs (returns 0 if not needed).
-    MCDCConditionID NextOrID = MCDCBuilder.getNextLOrCondID();
-    MCDCConditionID NextAndID = MCDCBuilder.getNextLAndCondID();
-    MCDCConditionID LHSid = MCDCBuilder.getCondID(E->getLHS());
-    MCDCConditionID RHSid = MCDCBuilder.getCondID(E->getRHS());
-
     // Create Branch Region around LHS condition.
-    // MC/DC: For "LHS || RHS"
-    // - If LHS is TRUE, execution goes to the LHS of the next logical-AND.
-    //   If that does not exist, execution exits (ID == 0).
-    // - If LHS is FALSE, execution goes to the RHS.
     createBranchRegion(E->getLHS(), subtractCounters(ParentCnt, RHSExecCnt),
-                       RHSExecCnt, LHSid, NextAndID, RHSid);
+                       RHSExecCnt, DecisionLHS);
 
     // Create Branch Region around RHS condition.
-    // MC/DC: For "LHS || RHS"
-    // - If RHS is TRUE, execution goes to LHS of the next logical-AND.
-    //   If that does not exist, execution exits (ID == 0).
-    // - If RHS is FALSE, execution goes to the LHS of the next logical-OR.
-    //   If that does not exist, execution exits (ID == 0).
     createBranchRegion(E->getRHS(), subtractCounters(RHSExecCnt, RHSFalseCnt),
-                       RHSFalseCnt, RHSid, NextAndID, NextOrID);
+                       RHSFalseCnt, DecisionRHS);
   }
 
   void VisitLambdaExpr(const LambdaExpr *LE) {
diff --git a/clang/lib/CodeGen/TargetInfo.h b/clang/lib/CodeGen/TargetInfo.h
index 0c0781a2d5ab9..7682f197041c7 100644
--- a/clang/lib/CodeGen/TargetInfo.h
+++ b/clang/lib/CodeGen/TargetInfo.h
@@ -496,7 +496,8 @@ createPPC64_SVR4_TargetCodeGenInfo(CodeGenModule &CGM, PPC64_SVR4_ABIKind Kind,
                                    bool SoftFloatABI);
 
 std::unique_ptr
-createRISCVTargetCodeGenInfo(CodeGenModule &CGM, unsigned XLen, unsigned FLen);
+createRISCVTargetCodeGenInfo(CodeGenModule &CGM, unsigned XLen, unsigned FLen,
+                             bool EABI);
 
 std::unique_ptr
 createCommonSPIRTargetCodeGenInfo(CodeGenModule &CGM);
diff --git a/clang/lib/CodeGen/Targets/AArch64.cpp b/clang/lib/CodeGen/Targets/AArch64.cpp
index 7102d190fe008..ee7f95084d2e0 100644
--- a/clang/lib/CodeGen/Targets/AArch64.cpp
+++ b/clang/lib/CodeGen/Targets/AArch64.cpp
@@ -138,6 +138,8 @@ class AArch64TargetCodeGenInfo : public TargetCodeGenInfo {
                   BPI.BranchTargetEnforcement ? "true" : "false");
     Fn->addFnAttr("branch-protection-pauth-lr",
                   BPI.BranchProtectionPAuthLR ? "true" : "false");
+    Fn->addFnAttr("guarded-control-stack",
+                  BPI.GuardedControlStack ? "true" : "false");
   }
 
   bool isScalarizableAsmOperand(CodeGen::CodeGenFunction &CGF,
diff --git a/clang/lib/CodeGen/Targets/RISCV.cpp b/clang/lib/CodeGen/Targets/RISCV.cpp
index 1e1d249b37ac0..0851d1993d0c0 100644
--- a/clang/lib/CodeGen/Targets/RISCV.cpp
+++ b/clang/lib/CodeGen/Targets/RISCV.cpp
@@ -25,8 +25,9 @@ class RISCVABIInfo : public DefaultABIInfo {
   // ISA might have a wider FLen than the selected ABI (e.g. an RV32IF target
   // with soft float ABI has FLen==0).
   unsigned FLen;
-  static const int NumArgGPRs = 8;
-  static const int NumArgFPRs = 8;
+  const int NumArgGPRs;
+  const int NumArgFPRs;
+  const bool EABI;
   bool detectFPCCEligibleStructHelper(QualType Ty, CharUnits CurOff,
                                       llvm::Type *&Field1Ty,
                                       CharUnits &Field1Off,
@@ -34,8 +35,10 @@ class RISCVABIInfo : public DefaultABIInfo {
                                       CharUnits &Field2Off) const;
 
 public:
-  RISCVABIInfo(CodeGen::CodeGenTypes &CGT, unsigned XLen, unsigned FLen)
-      : DefaultABIInfo(CGT), XLen(XLen), FLen(FLen) {}
+  RISCVABIInfo(CodeGen::CodeGenTypes &CGT, unsigned XLen, unsigned FLen,
+               bool EABI)
+      : DefaultABIInfo(CGT), XLen(XLen), FLen(FLen), NumArgGPRs(EABI ? 6 : 8),
+        NumArgFPRs(FLen != 0 ? 8 : 0), EABI(EABI) {}
 
   // DefaultABIInfo's classifyReturnType and classifyArgumentType are
   // non-virtual, but computeInfo is virtual, so we overload it.
@@ -86,7 +89,7 @@ void RISCVABIInfo::computeInfo(CGFunctionInfo &FI) const {
   }
 
   int ArgGPRsLeft = IsRetIndirect ? NumArgGPRs - 1 : NumArgGPRs;
-  int ArgFPRsLeft = FLen ? NumArgFPRs : 0;
+  int ArgFPRsLeft = NumArgFPRs;
   int NumFixedArgs = FI.getNumRequiredArgs();
 
   int ArgNum = 0;
@@ -396,9 +399,12 @@ ABIArgInfo RISCVABIInfo::classifyArgumentType(QualType Ty, bool IsFixed,
   // Determine the number of GPRs needed to pass the current argument
   // according to the ABI. 2*XLen-aligned varargs are passed in "aligned"
   // register pairs, so may consume 3 registers.
+  // TODO: To be compatible with GCC's behaviors, we don't align registers
+  // currently if we are using ILP32E calling convention. This behavior may be
+  // changed when RV32E/ILP32E is ratified.
   int NeededArgGPRs = 1;
   if (!IsFixed && NeededAlign == 2 * XLen)
-    NeededArgGPRs = 2 + (ArgGPRsLeft % 2);
+    NeededArgGPRs = 2 + (EABI && XLen == 32 ? 0 : (ArgGPRsLeft % 2));
   else if (Size > XLen && Size <= 2 * XLen)
     NeededArgGPRs = 2;
 
@@ -480,6 +486,13 @@ Address RISCVABIInfo::EmitVAArg(CodeGenFunction &CGF, Address VAListAddr,
 
   auto TInfo = getContext().getTypeInfoInChars(Ty);
 
+  // TODO: To be compatible with GCC's behaviors, we force arguments with
+  // 2×XLEN-bit alignment and size at most 2×XLEN bits like `long long`,
+  // `unsigned long long` and `double` to have 4-byte alignment. This
+  // behavior may be changed when RV32E/ILP32E is ratified.
+  if (EABI && XLen == 32)
+    TInfo.Align = std::min(TInfo.Align, CharUnits::fromQuantity(4));
+
   // Arguments bigger than 2*Xlen bytes are passed indirectly.
   bool IsIndirect = TInfo.Width > 2 * SlotSize;
 
@@ -499,8 +512,9 @@ namespace {
 class RISCVTargetCodeGenInfo : public TargetCodeGenInfo {
 public:
   RISCVTargetCodeGenInfo(CodeGen::CodeGenTypes &CGT, unsigned XLen,
-                         unsigned FLen)
-      : TargetCodeGenInfo(std::make_unique(CGT, XLen, FLen)) {}
+                         unsigned FLen, bool EABI)
+      : TargetCodeGenInfo(
+            std::make_unique(CGT, XLen, FLen, EABI)) {}
 
   void setTargetAttributes(const Decl *D, llvm::GlobalValue *GV,
                            CodeGen::CodeGenModule &CGM) const override {
@@ -526,6 +540,7 @@ class RISCVTargetCodeGenInfo : public TargetCodeGenInfo {
 
 std::unique_ptr
 CodeGen::createRISCVTargetCodeGenInfo(CodeGenModule &CGM, unsigned XLen,
-                                      unsigned FLen) {
-  return std::make_unique(CGM.getTypes(), XLen, FLen);
+                                      unsigned FLen, bool EABI) {
+  return std::make_unique(CGM.getTypes(), XLen, FLen,
+                                                  EABI);
 }
diff --git a/clang/lib/CodeGen/Targets/X86.cpp b/clang/lib/CodeGen/Targets/X86.cpp
index 241a8667cf7a2..b3785fadd7469 100644
--- a/clang/lib/CodeGen/Targets/X86.cpp
+++ b/clang/lib/CodeGen/Targets/X86.cpp
@@ -40,6 +40,11 @@ static llvm::Type* X86AdjustInlineAsmType(CodeGen::CodeGenFunction &CGF,
     return llvm::Type::getX86_MMXTy(CGF.getLLVMContext());
   }
 
+  if (Constraint == "k") {
+    llvm::Type *Int1Ty = llvm::Type::getInt1Ty(CGF.getLLVMContext());
+    return llvm::FixedVectorType::get(Int1Ty, Ty->getScalarSizeInBits());
+  }
+
   // No operation needed
   return Ty;
 }
diff --git a/clang/lib/Driver/Driver.cpp b/clang/lib/Driver/Driver.cpp
index 4bb814dca4c65..ec54394e2e769 100644
--- a/clang/lib/Driver/Driver.cpp
+++ b/clang/lib/Driver/Driver.cpp
@@ -9803,6 +9803,7 @@ const ToolChain &Driver::getToolChain(const ArgList &Args,
     case llvm::Triple::IOS:
     case llvm::Triple::TvOS:
     case llvm::Triple::WatchOS:
+    case llvm::Triple::XROS:
     case llvm::Triple::DriverKit:
       TC = std::make_unique(*this, Target, Args);
       break;
diff --git a/clang/lib/Driver/SanitizerArgs.cpp b/clang/lib/Driver/SanitizerArgs.cpp
index 6bf55afbb507d..0b0f8c42f50a1 100644
--- a/clang/lib/Driver/SanitizerArgs.cpp
+++ b/clang/lib/Driver/SanitizerArgs.cpp
@@ -39,8 +39,6 @@ static const SanitizerMask NotAllowedWithTrap = SanitizerKind::Vptr;
 static const SanitizerMask NotAllowedWithMinimalRuntime = SanitizerKind::Vptr;
 static const SanitizerMask NotAllowedWithExecuteOnly =
     SanitizerKind::Function | SanitizerKind::KCFI;
-static const SanitizerMask RequiresPIE =
-    SanitizerKind::DataFlow | SanitizerKind::Scudo;
 static const SanitizerMask NeedsUnwindTables =
     SanitizerKind::Address | SanitizerKind::HWAddress | SanitizerKind::Thread |
     SanitizerKind::Memory | SanitizerKind::DataFlow;
@@ -303,9 +301,7 @@ bool SanitizerArgs::needsCfiDiagRt() const {
          CfiCrossDso && !ImplicitCfiRuntime;
 }
 
-bool SanitizerArgs::requiresPIE() const {
-  return NeedPIE || (Sanitizers.Mask & RequiresPIE);
-}
+bool SanitizerArgs::requiresPIE() const { return NeedPIE; }
 
 bool SanitizerArgs::needsUnwindTables() const {
   return static_cast(Sanitizers.Mask & NeedsUnwindTables);
@@ -699,8 +695,6 @@ SanitizerArgs::SanitizerArgs(const ToolChain &TC,
     MsanParamRetval = Args.hasFlag(
         options::OPT_fsanitize_memory_param_retval,
         options::OPT_fno_sanitize_memory_param_retval, MsanParamRetval);
-    NeedPIE |= !(TC.getTriple().isOSLinux() &&
-                 TC.getTriple().getArch() == llvm::Triple::x86_64);
   } else if (AllAddedKinds & SanitizerKind::KernelMemory) {
     MsanUseAfterDtor = false;
     MsanParamRetval = Args.hasFlag(
diff --git a/clang/lib/Driver/ToolChains/Arch/AArch64.cpp b/clang/lib/Driver/ToolChains/Arch/AArch64.cpp
index 912df79417ae2..0cf96bb5c9cb0 100644
--- a/clang/lib/Driver/ToolChains/Arch/AArch64.cpp
+++ b/clang/lib/Driver/ToolChains/Arch/AArch64.cpp
@@ -53,6 +53,11 @@ std::string aarch64::getAArch64TargetCPU(const ArgList &Args,
     return "apple-m1";
   }
 
+  if (Triple.isXROS()) {
+    // The xrOS simulator runs on M1 as well, it should have been covered above.
+    assert(!Triple.isSimulatorEnvironment() && "xrossim should be mac-like");
+    return "apple-a12";
+  }
   // arm64e requires v8.3a and only runs on apple-a12 and later CPUs.
   if (Triple.isArm64e())
     return "apple-a12";
@@ -67,103 +72,45 @@ std::string aarch64::getAArch64TargetCPU(const ArgList &Args,
 
 // Decode AArch64 features from string like +[no]featureA+[no]featureB+...
 static bool DecodeAArch64Features(const Driver &D, StringRef text,
-                                  std::vector &Features,
-                                  const llvm::AArch64::ArchInfo &ArchInfo) {
+                                  llvm::AArch64::ExtensionSet &Extensions) {
   SmallVector Split;
   text.split(Split, StringRef("+"), -1, false);
 
   for (StringRef Feature : Split) {
-    StringRef FeatureName = llvm::AArch64::getArchExtFeature(Feature);
-    if (!FeatureName.empty())
-      Features.push_back(FeatureName);
-    else if (Feature == "neon" || Feature == "noneon")
+    if (Feature == "neon" || Feature == "noneon") {
       D.Diag(clang::diag::err_drv_no_neon_modifier);
-    else
-      return false;
-
-    // +sme implies +bf16.
-    // +sme-f64f64 and +sme-i16i64 both imply +sme.
-    if (Feature == "sme") {
-      Features.push_back("+bf16");
-    } else if (Feature == "nosme") {
-      Features.push_back("-sme-f64f64");
-      Features.push_back("-sme-i16i64");
-    } else if (Feature == "sme-f64f64") {
-      Features.push_back("+sme");
-      Features.push_back("+bf16");
-    } else if (Feature == "sme-i16i64") {
-      Features.push_back("+sme");
-      Features.push_back("+bf16");
-    } else if (Feature == "nobf16") {
-      Features.push_back("-sme");
-      Features.push_back("-sme-f64f64");
-      Features.push_back("-sme-i16i64");
+      continue;
     }
-
-    if (Feature == "sve2")
-      Features.push_back("+sve");
-    else if (Feature == "sve2-bitperm" || Feature == "sve2-sha3" ||
-             Feature == "sve2-aes" || Feature == "sve2-sm4") {
-      Features.push_back("+sve");
-      Features.push_back("+sve2");
-    } else if (Feature == "nosve") {
-      Features.push_back("-sve2");
-      Features.push_back("-sve2-bitperm");
-      Features.push_back("-sve2-sha3");
-      Features.push_back("-sve2-aes");
-      Features.push_back("-sve2-sm4");
-    } else if (Feature == "nosve2") {
-      Features.push_back("-sve2-bitperm");
-      Features.push_back("-sve2-sha3");
-      Features.push_back("-sve2-aes");
-      Features.push_back("-sve2-sm4");
-    }
-
-    // +sve implies +f32mm if the base architecture is >= v8.6A (except v9A)
-    // It isn't the case in general that sve implies both f64mm and f32mm
-    if ((ArchInfo == llvm::AArch64::ARMV8_6A ||
-         ArchInfo == llvm::AArch64::ARMV8_7A ||
-         ArchInfo == llvm::AArch64::ARMV8_8A ||
-         ArchInfo == llvm::AArch64::ARMV8_9A ||
-         ArchInfo == llvm::AArch64::ARMV9_1A ||
-         ArchInfo == llvm::AArch64::ARMV9_2A ||
-         ArchInfo == llvm::AArch64::ARMV9_3A ||
-         ArchInfo == llvm::AArch64::ARMV9_4A) &&
-        Feature == "sve")
-      Features.push_back("+f32mm");
+    if (!Extensions.parseModifier(Feature))
+      return false;
   }
+
   return true;
 }
 
 // Check if the CPU name and feature modifiers in -mcpu are legal. If yes,
 // decode CPU and feature.
 static bool DecodeAArch64Mcpu(const Driver &D, StringRef Mcpu, StringRef &CPU,
-                              std::vector &Features) {
+                              llvm::AArch64::ExtensionSet &Extensions) {
   std::pair Split = Mcpu.split("+");
   CPU = Split.first;
-  const llvm::AArch64::ArchInfo *ArchInfo = &llvm::AArch64::ARMV8A;
 
   if (CPU == "native")
     CPU = llvm::sys::getHostCPUName();
 
   if (CPU == "generic") {
-    Features.push_back("+neon");
+    Extensions.enable(llvm::AArch64::AEK_SIMD);
   } else {
     const std::optional CpuInfo =
         llvm::AArch64::parseCpu(CPU);
     if (!CpuInfo)
       return false;
-    ArchInfo = &CpuInfo->Arch;
-
-    Features.push_back(ArchInfo->ArchFeature);
 
-    auto Extension = CpuInfo->getImpliedExtensions();
-    if (!llvm::AArch64::getExtensionFeatures(Extension, Features))
-      return false;
+    Extensions.addCPUDefaults(*CpuInfo);
   }
 
   if (Split.second.size() &&
-      !DecodeAArch64Features(D, Split.second, Features, *ArchInfo))
+      !DecodeAArch64Features(D, Split.second, Extensions))
     return false;
 
   return true;
@@ -172,30 +119,21 @@ static bool DecodeAArch64Mcpu(const Driver &D, StringRef Mcpu, StringRef &CPU,
 static bool
 getAArch64ArchFeaturesFromMarch(const Driver &D, StringRef March,
                                 const ArgList &Args,
-                                std::vector &Features) {
+                                llvm::AArch64::ExtensionSet &Extensions) {
   std::string MarchLowerCase = March.lower();
   std::pair Split = StringRef(MarchLowerCase).split("+");
 
-  std::optional  ArchInfo =
+  const llvm::AArch64::ArchInfo *ArchInfo =
       llvm::AArch64::parseArch(Split.first);
   if (Split.first == "native")
     ArchInfo = llvm::AArch64::getArchForCpu(llvm::sys::getHostCPUName().str());
   if (!ArchInfo)
     return false;
-  Features.push_back(ArchInfo->ArchFeature);
-
-  // Enable SVE2 by default on Armv9-A.
-  // It can still be disabled if +nosve2 is present.
-  // We must do this early so that DecodeAArch64Features has the correct state
-  if ((*ArchInfo == llvm::AArch64::ARMV9A ||
-       *ArchInfo == llvm::AArch64::ARMV9_1A ||
-       *ArchInfo == llvm::AArch64::ARMV9_2A)) {
-    Features.push_back("+sve");
-    Features.push_back("+sve2");
-  }
+
+  Extensions.addArchDefaults(*ArchInfo);
 
   if ((Split.second.size() &&
-       !DecodeAArch64Features(D, Split.second, Features, *ArchInfo)))
+       !DecodeAArch64Features(D, Split.second, Extensions)))
     return false;
 
   return true;
@@ -204,10 +142,10 @@ getAArch64ArchFeaturesFromMarch(const Driver &D, StringRef March,
 static bool
 getAArch64ArchFeaturesFromMcpu(const Driver &D, StringRef Mcpu,
                                const ArgList &Args,
-                               std::vector &Features) {
+                               llvm::AArch64::ExtensionSet &Extensions) {
   StringRef CPU;
   std::string McpuLowerCase = Mcpu.lower();
-  if (!DecodeAArch64Mcpu(D, McpuLowerCase, CPU, Features))
+  if (!DecodeAArch64Mcpu(D, McpuLowerCase, CPU, Extensions))
     return false;
 
   return true;
@@ -218,10 +156,10 @@ getAArch64MicroArchFeaturesFromMtune(const Driver &D, StringRef Mtune,
                                      const ArgList &Args,
                                      std::vector &Features) {
   std::string MtuneLowerCase = Mtune.lower();
-  // Check CPU name is valid
-  std::vector MtuneFeatures;
+  // Check CPU name is valid, but ignore any extensions on it.
+  llvm::AArch64::ExtensionSet Extensions;
   StringRef Tune;
-  if (!DecodeAArch64Mcpu(D, MtuneLowerCase, Tune, MtuneFeatures))
+  if (!DecodeAArch64Mcpu(D, MtuneLowerCase, Tune, Extensions))
     return false;
 
   // Handle CPU name is 'native'.
@@ -240,7 +178,8 @@ getAArch64MicroArchFeaturesFromMcpu(const Driver &D, StringRef Mcpu,
                                     const ArgList &Args,
                                     std::vector &Features) {
   StringRef CPU;
-  std::vector DecodedFeature;
+  // Check CPU name is valid, but ignore any extensions on it.
+  llvm::AArch64::ExtensionSet DecodedFeature;
   std::string McpuLowerCase = Mcpu.lower();
   if (!DecodeAArch64Mcpu(D, McpuLowerCase, CPU, DecodedFeature))
     return false;
@@ -255,9 +194,8 @@ void aarch64::getAArch64TargetFeatures(const Driver &D,
                                        bool ForAS) {
   Arg *A;
   bool success = true;
-  // Enable NEON by default.
-  Features.push_back("+neon");
   llvm::StringRef WaMArch;
+  llvm::AArch64::ExtensionSet Extensions;
   if (ForAS)
     for (const auto *A :
          Args.filtered(options::OPT_Wa_COMMA, options::OPT_Xassembler))
@@ -268,17 +206,19 @@ void aarch64::getAArch64TargetFeatures(const Driver &D,
   // "-Xassembler -march" is detected. Otherwise it may return false
   // and causes Clang to error out.
   if (!WaMArch.empty())
-    success = getAArch64ArchFeaturesFromMarch(D, WaMArch, Args, Features);
+    success = getAArch64ArchFeaturesFromMarch(D, WaMArch, Args, Extensions);
   else if ((A = Args.getLastArg(options::OPT_march_EQ)))
-    success = getAArch64ArchFeaturesFromMarch(D, A->getValue(), Args, Features);
+    success =
+        getAArch64ArchFeaturesFromMarch(D, A->getValue(), Args, Extensions);
   else if ((A = Args.getLastArg(options::OPT_mcpu_EQ)))
-    success = getAArch64ArchFeaturesFromMcpu(D, A->getValue(), Args, Features);
+    success =
+        getAArch64ArchFeaturesFromMcpu(D, A->getValue(), Args, Extensions);
   else if (isCPUDeterminedByTriple(Triple))
     success = getAArch64ArchFeaturesFromMcpu(
-        D, getAArch64TargetCPU(Args, Triple, A), Args, Features);
+        D, getAArch64TargetCPU(Args, Triple, A), Args, Extensions);
   else
     // Default to 'A' profile if the architecture is not specified.
-    success = getAArch64ArchFeaturesFromMarch(D, "armv8-a", Args, Features);
+    success = getAArch64ArchFeaturesFromMarch(D, "armv8-a", Args, Extensions);
 
   if (success && (A = Args.getLastArg(clang::driver::options::OPT_mtune_EQ)))
     success =
@@ -300,12 +240,23 @@ void aarch64::getAArch64TargetFeatures(const Driver &D,
       Diag << A->getSpelling() << A->getValue();
   }
 
+  // -mgeneral-regs-only disables all floating-point features.
   if (Args.getLastArg(options::OPT_mgeneral_regs_only)) {
-    Features.push_back("-fp-armv8");
-    Features.push_back("-crypto");
-    Features.push_back("-neon");
+    Extensions.disable(llvm::AArch64::AEK_FP);
   }
 
+  // En/disable crc
+  if (Arg *A = Args.getLastArg(options::OPT_mcrc, options::OPT_mnocrc)) {
+    if (A->getOption().matches(options::OPT_mcrc))
+      Extensions.enable(llvm::AArch64::AEK_CRC);
+    else
+      Extensions.disable(llvm::AArch64::AEK_CRC);
+  }
+
+  // At this point all hardware features are decided, so convert the extensions
+  // set to a feature list.
+  Extensions.toLLVMFeatureList(Features);
+
   if (Arg *A = Args.getLastArg(options::OPT_mtp_mode_EQ)) {
     StringRef Mtp = A->getValue();
     if (Mtp == "el3" || Mtp == "tpidr_el3")
@@ -367,147 +318,6 @@ void aarch64::getAArch64TargetFeatures(const Driver &D,
     }
   }
 
-  // En/disable crc
-  if (Arg *A = Args.getLastArg(options::OPT_mcrc, options::OPT_mnocrc)) {
-    if (A->getOption().matches(options::OPT_mcrc))
-      Features.push_back("+crc");
-    else
-      Features.push_back("-crc");
-  }
-
-  int V8Version = -1;
-  int V9Version = -1;
-  bool HasNoSM4 = false;
-  bool HasNoSHA3 = false;
-  bool HasNoSHA2 = false;
-  bool HasNoAES = false;
-  bool HasSM4 = false;
-  bool HasSHA3 = false;
-  bool HasSHA2 = false;
-  bool HasAES = false;
-  bool HasCrypto = false;
-  bool HasNoCrypto = false;
-  int FullFP16Pos = -1;
-  int NoFullFP16Pos = -1;
-  int FP16FMLPos = -1;
-  int NoFP16FMLPos = -1;
-  int ArchFeatPos = -1;
-
-  for (auto I = Features.begin(), E = Features.end(); I != E; I++) {
-    if (*I == "+v8a")   V8Version = 0;
-    else if (*I == "+v8.1a") V8Version = 1;
-    else if (*I == "+v8.2a") V8Version = 2;
-    else if (*I == "+v8.3a") V8Version = 3;
-    else if (*I == "+v8.4a") V8Version = 4;
-    else if (*I == "+v8.5a") V8Version = 5;
-    else if (*I == "+v8.6a") V8Version = 6;
-    else if (*I == "+v8.7a") V8Version = 7;
-    else if (*I == "+v8.8a") V8Version = 8;
-    else if (*I == "+v8.9a") V8Version = 9;
-    else if (*I == "+v9a")   V9Version = 0;
-    else if (*I == "+v9.1a") V9Version = 1;
-    else if (*I == "+v9.2a") V9Version = 2;
-    else if (*I == "+v9.3a") V9Version = 3;
-    else if (*I == "+v9.4a") V9Version = 4;
-    else if (*I == "+v9.5a") V9Version = 5;
-    else if (*I == "+sm4")  HasSM4 = true;
-    else if (*I == "+sha3") HasSHA3 = true;
-    else if (*I == "+sha2") HasSHA2 = true;
-    else if (*I == "+aes")  HasAES = true;
-    else if (*I == "-sm4")  HasNoSM4 = true;
-    else if (*I == "-sha3") HasNoSHA3 = true;
-    else if (*I == "-sha2") HasNoSHA2 = true;
-    else if (*I == "-aes")  HasNoAES = true;
-    else if (*I == "+fp16fml")  FP16FMLPos = I - Features.begin();
-    else if (*I == "-fp16fml")  NoFP16FMLPos = I - Features.begin();
-    else if (*I == "-fullfp16") NoFullFP16Pos = I - Features.begin();
-    else if (*I == "+fullfp16") FullFP16Pos = I - Features.begin();
-    // Whichever option comes after (right-most option) will win
-    else if (*I == "+crypto") {
-      HasCrypto = true;
-      HasNoCrypto = false;
-    } else if (*I == "-crypto" || *I == "-neon") {
-      HasCrypto = false;
-      HasNoCrypto = true;
-      HasSM4 = HasSHA2 = HasSHA3 = HasAES = false;
-    }
-    // Register the iterator position if this is an architecture feature
-    if (ArchFeatPos == -1 && (V8Version != -1 || V9Version != -1))
-      ArchFeatPos = I - Features.begin();
-  }
-
-  // Handle (arch-dependent) fp16fml/fullfp16 relationship.
-  // FIXME: this fp16fml option handling will be reimplemented after the
-  // TargetParser rewrite.
-  if (V8Version >= 4) {
-    // "-fullfp16" "+fullfp16" && "+fp16fml" "+fullfp16" && no "+fullfp16" "-fp16fml" = "+fp16fml"
-    if (FullFP16Pos > NoFullFP16Pos && FullFP16Pos > FP16FMLPos && FullFP16Pos > NoFP16FMLPos)
-      // Only entangled feature that can be to the right of this +fullfp16 is -fp16fml.
-      // Only append the +fp16fml if there is no -fp16fml after the +fullfp16.
-      Features.push_back("+fp16fml");
-    else
-      goto fp16_fml_fallthrough;
-  } else {
-fp16_fml_fallthrough:
-    // In both of these cases, putting the 'other' feature on the end of the vector will
-    // result in the same effect as placing it immediately after the current feature.
-    // "+fp16fml"  "-fullfp16" = "-fp16fml"
-    if (NoFullFP16Pos > FP16FMLPos)
-      Features.push_back("-fp16fml");
-    // "-fullfp16" "+fp16fml" = "+fullfp16"
-    else if (NoFullFP16Pos < FP16FMLPos)
-      Features.push_back("+fullfp16");
-  }
-
-  // FIXME: this needs reimplementation too after the TargetParser rewrite
-  //
-  // Context sensitive meaning of Crypto:
-  // 1) For Arch >= ARMv8.4a:  crypto = sm4 + sha3 + sha2 + aes
-  // 2) For Arch <= ARMv8.3a:  crypto = sha2 + aes
-  if (V8Version >= 4 || V9Version >= 0) {
-    if (HasCrypto && !HasNoCrypto) {
-      // Check if we have NOT disabled an algorithm with something like:
-      //   +crypto, -algorithm
-      // And if "-algorithm" does not occur, we enable that crypto algorithm.
-      if (!HasNoSM4)
-        Features.push_back("+sm4");
-      if (!HasNoSHA3)
-        Features.push_back("+sha3");
-      if (!HasNoSHA2)
-        Features.push_back("+sha2");
-      if (!HasNoAES)
-        Features.push_back("+aes");
-    } else if (HasNoCrypto) {
-      // Check if we have NOT enabled a crypto algorithm with something like:
-      //   -crypto, +algorithm
-      // And if "+algorithm" does not occur, we disable that crypto algorithm.
-      if (!HasSM4)
-        Features.push_back("-sm4");
-      if (!HasSHA3)
-        Features.push_back("-sha3");
-      if (!HasSHA2)
-        Features.push_back("-sha2");
-      if (!HasAES)
-        Features.push_back("-aes");
-    }
-  } else {
-    if (HasCrypto && !HasNoCrypto) {
-      if (!HasNoSHA2)
-        Features.push_back("+sha2");
-      if (!HasNoAES)
-        Features.push_back("+aes");
-    } else if (HasNoCrypto) {
-      if (!HasSHA2)
-        Features.push_back("-sha2");
-      if (!HasAES)
-        Features.push_back("-aes");
-      if (V8Version == 2 || V8Version == 3) {
-        Features.push_back("-sm4");
-        Features.push_back("-sha3");
-      }
-    }
-  }
-
   if (Arg *A = Args.getLastArg(options::OPT_mno_unaligned_access,
                                options::OPT_munaligned_access)) {
     if (A->getOption().matches(options::OPT_mno_unaligned_access))
diff --git a/clang/lib/Driver/ToolChains/Arch/ARM.cpp b/clang/lib/Driver/ToolChains/Arch/ARM.cpp
index 25470db2b6ceb..e6ee2f88a84ed 100644
--- a/clang/lib/Driver/ToolChains/Arch/ARM.cpp
+++ b/clang/lib/Driver/ToolChains/Arch/ARM.cpp
@@ -367,6 +367,7 @@ arm::FloatABI arm::getDefaultFloatABI(const llvm::Triple &Triple) {
   case llvm::Triple::IOS:
   case llvm::Triple::TvOS:
   case llvm::Triple::DriverKit:
+  case llvm::Triple::XROS:
     // Darwin defaults to "softfp" for v6 and v7.
     if (Triple.isWatchABI())
       return FloatABI::Hard;
@@ -836,8 +837,8 @@ llvm::ARM::FPUKind arm::getARMTargetFeatures(const Driver &D,
     if (A->getOption().matches(options::OPT_mlong_calls))
       Features.push_back("+long-calls");
   } else if (KernelOrKext && (!Triple.isiOS() || Triple.isOSVersionLT(6)) &&
-             !Triple.isWatchOS()) {
-      Features.push_back("+long-calls");
+             !Triple.isWatchOS() && !Triple.isXROS()) {
+    Features.push_back("+long-calls");
   }
 
   // Generate execute-only output (no data access to code sections).
diff --git a/clang/lib/Driver/ToolChains/Arch/RISCV.cpp b/clang/lib/Driver/ToolChains/Arch/RISCV.cpp
index 16a8b3cc42bab..a46b44f9ad2b2 100644
--- a/clang/lib/Driver/ToolChains/Arch/RISCV.cpp
+++ b/clang/lib/Driver/ToolChains/Arch/RISCV.cpp
@@ -210,6 +210,7 @@ StringRef riscv::getRISCVABI(const ArgList &Args, const llvm::Triple &Triple) {
   // rv32e -> ilp32e
   // rv32* -> ilp32
   // rv64g | rv64*d -> lp64d
+  // rv64e -> lp64e
   // rv64* -> lp64
   StringRef Arch = getRISCVArch(Args, Triple);
 
@@ -285,6 +286,7 @@ StringRef riscv::getRISCVArch(const llvm::opt::ArgList &Args,
   // 3. Choose a default based on `-mabi=`
   //
   // ilp32e -> rv32e
+  // lp64e -> rv64e
   // ilp32 | ilp32f | ilp32d -> rv32imafdc
   // lp64 | lp64f | lp64d -> rv64imafdc
   if (const Arg *A = Args.getLastArg(options::OPT_mabi_EQ)) {
@@ -292,6 +294,8 @@ StringRef riscv::getRISCVArch(const llvm::opt::ArgList &Args,
 
     if (MABI.equals_insensitive("ilp32e"))
       return "rv32e";
+    else if (MABI.equals_insensitive("lp64e"))
+      return "rv64e";
     else if (MABI.starts_with_insensitive("ilp32"))
       return "rv32imafdc";
     else if (MABI.starts_with_insensitive("lp64")) {
diff --git a/clang/lib/Driver/ToolChains/Clang.cpp b/clang/lib/Driver/ToolChains/Clang.cpp
index 79024f1481411..ab8beb715166c 100644
--- a/clang/lib/Driver/ToolChains/Clang.cpp
+++ b/clang/lib/Driver/ToolChains/Clang.cpp
@@ -45,7 +45,9 @@
 #include "clang/Driver/XRayArgs.h"
 #include "llvm/ADT/SmallSet.h"
 #include "llvm/ADT/StringExtras.h"
+#include "llvm/BinaryFormat/Magic.h"
 #include "llvm/Config/llvm-config.h"
+#include "llvm/Object/ObjectFile.h"
 #include "llvm/Option/ArgList.h"
 #include "llvm/Support/Casting.h"
 #include "llvm/Support/CodeGen.h"
@@ -993,11 +995,21 @@ static bool ContainsAppendFooterAction(const Action *A) {
   return false;
 }
 
-static bool hasClangPchSignature(const Driver &D, StringRef Path) {
-  if (llvm::ErrorOr> MemBuf =
-          D.getVFS().getBufferForFile(Path))
-    return (*MemBuf)->getBuffer().starts_with("CPCH");
-  return false;
+static bool maybeHasClangPchSignature(const Driver &D, StringRef Path) {
+  llvm::ErrorOr> MemBuf =
+      D.getVFS().getBufferForFile(Path);
+  if (!MemBuf)
+    return false;
+  llvm::file_magic Magic = llvm::identify_magic((*MemBuf)->getBuffer());
+  if (Magic == llvm::file_magic::unknown)
+    return false;
+  // Return true for both raw Clang AST files and object files which may
+  // contain a __clangast section.
+  if (Magic == llvm::file_magic::clang_ast)
+    return true;
+  Expected> Obj =
+      llvm::object::ObjectFile::createObjectFile(**MemBuf, Magic);
+  return !Obj.takeError();
 }
 
 static bool gchProbe(const Driver &D, StringRef Path) {
@@ -1009,14 +1021,14 @@ static bool gchProbe(const Driver &D, StringRef Path) {
     std::error_code EC;
     for (llvm::vfs::directory_iterator DI = D.getVFS().dir_begin(Path, EC), DE;
          !EC && DI != DE; DI = DI.increment(EC)) {
-      if (hasClangPchSignature(D, DI->path()))
+      if (maybeHasClangPchSignature(D, DI->path()))
         return true;
     }
     D.Diag(diag::warn_drv_pch_ignoring_gch_dir) << Path;
     return false;
   }
 
-  if (hasClangPchSignature(D, Path))
+  if (maybeHasClangPchSignature(D, Path))
     return true;
   D.Diag(diag::warn_drv_pch_ignoring_gch_file) << Path;
   return false;
@@ -1644,7 +1656,7 @@ static void CollectARMPACBTIOptions(const ToolChain &TC, const ArgList &Args,
         << Triple.getArchName();
 
   StringRef Scope, Key;
-  bool IndirectBranches, BranchProtectionPAuthLR;
+  bool IndirectBranches, BranchProtectionPAuthLR, GuardedControlStack;
 
   if (A->getOption().matches(options::OPT_msign_return_address_EQ)) {
     Scope = A->getValue();
@@ -1654,6 +1666,7 @@ static void CollectARMPACBTIOptions(const ToolChain &TC, const ArgList &Args,
     Key = "a_key";
     IndirectBranches = false;
     BranchProtectionPAuthLR = false;
+    GuardedControlStack = false;
   } else {
     StringRef DiagMsg;
     llvm::ARM::ParsedBranchProtection PBP;
@@ -1667,6 +1680,7 @@ static void CollectARMPACBTIOptions(const ToolChain &TC, const ArgList &Args,
     Key = PBP.Key;
     BranchProtectionPAuthLR = PBP.BranchProtectionPAuthLR;
     IndirectBranches = PBP.BranchTargetEnforcement;
+    GuardedControlStack = PBP.GuardedControlStack;
   }
 
   CmdArgs.push_back(
@@ -1679,6 +1693,8 @@ static void CollectARMPACBTIOptions(const ToolChain &TC, const ArgList &Args,
         Args.MakeArgString(Twine("-mbranch-protection-pauth-lr")));
   if (IndirectBranches)
     CmdArgs.push_back("-mbranch-target-enforce");
+  if (GuardedControlStack)
+    CmdArgs.push_back("-mguarded-control-stack");
 }
 
 void Clang::AddARMTargetArgs(const llvm::Triple &Triple, const ArgList &Args,
@@ -6484,20 +6500,24 @@ void Clang::ConstructJob(Compilation &C, const JobAction &JA,
     }
   }
 
-  if (Arg *A = Args.getLastArg(options::OPT_mlarge_data_threshold_EQ)) {
-    if (!Triple.isX86()) {
-      D.Diag(diag::err_drv_unsupported_opt_for_target)
-          << A->getOption().getName() << TripleStr;
-    } else {
-      bool IsMediumCM = false;
-      if (Arg *A = Args.getLastArg(options::OPT_mcmodel_EQ))
-        IsMediumCM = StringRef(A->getValue()) == "medium";
-      if (!IsMediumCM) {
+  if (Triple.getArch() == llvm::Triple::x86_64) {
+    bool IsMediumCM = false;
+    bool IsLargeCM = false;
+    if (Arg *A = Args.getLastArg(options::OPT_mcmodel_EQ)) {
+      IsMediumCM = StringRef(A->getValue()) == "medium";
+      IsLargeCM = StringRef(A->getValue()) == "large";
+    }
+    if (Arg *A = Args.getLastArg(options::OPT_mlarge_data_threshold_EQ)) {
+      if (!IsMediumCM && !IsLargeCM) {
         D.Diag(diag::warn_drv_large_data_threshold_invalid_code_model)
             << A->getOption().getRenderName();
       } else {
         A->render(Args, CmdArgs);
       }
+    } else if (IsMediumCM) {
+      CmdArgs.push_back("-mlarge-data-threshold=65536");
+    } else if (IsLargeCM) {
+      CmdArgs.push_back("-mlarge-data-threshold=0");
     }
   }
 
diff --git a/clang/lib/Driver/ToolChains/CommonArgs.cpp b/clang/lib/Driver/ToolChains/CommonArgs.cpp
index 7440d50f65881..9c79b60d063aa 100644
--- a/clang/lib/Driver/ToolChains/CommonArgs.cpp
+++ b/clang/lib/Driver/ToolChains/CommonArgs.cpp
@@ -744,6 +744,8 @@ void tools::addLTOOptions(const ToolChain &ToolChain, const ArgList &Args,
   const bool IsAMDGCN = ToolChain.getTriple().isAMDGCN();
   const char *Linker = Args.MakeArgString(ToolChain.GetLinkerPath());
   const Driver &D = ToolChain.getDriver();
+  const bool IsFatLTO = Args.hasArg(options::OPT_ffat_lto_objects);
+  const bool IsUnifiedLTO = Args.hasArg(options::OPT_funified_lto);
   if (llvm::sys::path::filename(Linker) != "ld.lld" &&
       llvm::sys::path::stem(Linker) != "ld.lld" &&
       !ToolChain.getTriple().isOSOpenBSD()) {
@@ -773,7 +775,7 @@ void tools::addLTOOptions(const ToolChain &ToolChain, const ArgList &Args,
   } else {
     // Tell LLD to find and use .llvm.lto section in regular relocatable object
     // files
-    if (Args.hasArg(options::OPT_ffat_lto_objects))
+    if (IsFatLTO)
       CmdArgs.push_back("--fat-lto-objects");
   }
 
@@ -833,7 +835,8 @@ void tools::addLTOOptions(const ToolChain &ToolChain, const ArgList &Args,
   // Matrix intrinsic lowering happens at link time with ThinLTO. Enable
   // LowerMatrixIntrinsicsPass, which is transitively called by
   // buildThinLTODefaultPipeline under EnableMatrix.
-  if (IsThinLTO && Args.hasArg(options::OPT_fenable_matrix))
+  if ((IsThinLTO || IsFatLTO || IsUnifiedLTO) &&
+        Args.hasArg(options::OPT_fenable_matrix))
     CmdArgs.push_back(
         Args.MakeArgString(Twine(PluginOptPrefix) + "-enable-matrix"));
 
diff --git a/clang/lib/Driver/ToolChains/Darwin.cpp b/clang/lib/Driver/ToolChains/Darwin.cpp
index 65846cace461e..9050c545cf544 100644
--- a/clang/lib/Driver/ToolChains/Darwin.cpp
+++ b/clang/lib/Driver/ToolChains/Darwin.cpp
@@ -219,7 +219,8 @@ static bool shouldLinkerNotDedup(bool IsLinkerOnlyAction, const ArgList &Args) {
 void darwin::Linker::AddLinkArgs(Compilation &C, const ArgList &Args,
                                  ArgStringList &CmdArgs,
                                  const InputInfoList &Inputs,
-                                 VersionTuple Version, bool LinkerIsLLD) const {
+                                 VersionTuple Version, bool LinkerIsLLD,
+                                 bool UsePlatformVersion) const {
   const Driver &D = getToolChain().getDriver();
   const toolchains::MachO &MachOTC = getMachOToolChain();
 
@@ -355,7 +356,7 @@ void darwin::Linker::AddLinkArgs(Compilation &C, const ArgList &Args,
   Args.AddAllArgs(CmdArgs, options::OPT_init);
 
   // Add the deployment target.
-  if (Version >= VersionTuple(520) || LinkerIsLLD)
+  if (Version >= VersionTuple(520) || LinkerIsLLD || UsePlatformVersion)
     MachOTC.addPlatformVersionArgs(Args, CmdArgs);
   else
     MachOTC.addMinVersionArgs(Args, CmdArgs);
@@ -596,9 +597,13 @@ void darwin::Linker::ConstructJob(Compilation &C, const JobAction &JA,
   const char *Exec =
       Args.MakeArgString(getToolChain().GetLinkerPath(&LinkerIsLLD));
 
+  // xrOS always uses -platform-version.
+  bool UsePlatformVersion = getToolChain().getTriple().isXROS();
+
   // I'm not sure why this particular decomposition exists in gcc, but
   // we follow suite for ease of comparison.
-  AddLinkArgs(C, Args, CmdArgs, Inputs, Version, LinkerIsLLD);
+  AddLinkArgs(C, Args, CmdArgs, Inputs, Version, LinkerIsLLD,
+              UsePlatformVersion);
 
   if (willEmitRemarks(Args) &&
       checkRemarksOptions(getToolChain().getDriver(), Args,
@@ -954,6 +959,13 @@ ObjCRuntime Darwin::getDefaultObjCRuntime(bool isNonFragile) const {
     return ObjCRuntime(ObjCRuntime::WatchOS, TargetVersion);
   if (isTargetIOSBased())
     return ObjCRuntime(ObjCRuntime::iOS, TargetVersion);
+  if (isTargetXROS()) {
+    // XROS uses the iOS runtime.
+    auto T = llvm::Triple(Twine("arm64-apple-") +
+                          llvm::Triple::getOSTypeName(llvm::Triple::XROS) +
+                          TargetVersion.getAsString());
+    return ObjCRuntime(ObjCRuntime::iOS, T.getiOSVersion());
+  }
   if (isNonFragile)
     return ObjCRuntime(ObjCRuntime::MacOSX, TargetVersion);
   return ObjCRuntime(ObjCRuntime::FragileMacOSX, TargetVersion);
@@ -961,7 +973,7 @@ ObjCRuntime Darwin::getDefaultObjCRuntime(bool isNonFragile) const {
 
 /// Darwin provides a blocks runtime starting in MacOS X 10.6 and iOS 3.2.
 bool Darwin::hasBlocksRuntime() const {
-  if (isTargetWatchOSBased() || isTargetDriverKit())
+  if (isTargetWatchOSBased() || isTargetDriverKit() || isTargetXROS())
     return true;
   else if (isTargetIOSBased())
     return !isIPhoneOSVersionLT(3, 2);
@@ -1092,6 +1104,8 @@ std::string Darwin::ComputeEffectiveClangTriple(const ArgList &Args,
     Str += "driverkit";
   else if (isTargetIOSBased() || isTargetMacCatalyst())
     Str += "ios";
+  else if (isTargetXROS())
+    Str += llvm::Triple::getOSTypeName(llvm::Triple::XROS);
   else
     Str += "macosx";
   Str += getTripleTargetVersion().getAsString();
@@ -1174,6 +1188,8 @@ void DarwinClang::AddLinkARCArgs(const ArgList &Args,
   // ARC runtime is supported everywhere on arm64e.
   if (getTriple().isArm64e())
     return;
+  if (isTargetXROS())
+    return;
 
   ObjCRuntime runtime = getDefaultObjCRuntime(/*nonfragile*/ true);
 
@@ -1309,6 +1325,8 @@ StringRef Darwin::getPlatformFamily() const {
       return "Watch";
     case DarwinPlatformKind::DriverKit:
       return "DriverKit";
+    case DarwinPlatformKind::XROS:
+      return "XR";
   }
   llvm_unreachable("Unsupported platform");
 }
@@ -1340,6 +1358,9 @@ StringRef Darwin::getOSLibraryNameSuffix(bool IgnoreSim) const {
   case DarwinPlatformKind::WatchOS:
     return TargetEnvironment == NativeEnvironment || IgnoreSim ? "watchos"
                                                                : "watchossim";
+  case DarwinPlatformKind::XROS:
+    return TargetEnvironment == NativeEnvironment || IgnoreSim ? "xros"
+                                                               : "xrossim";
   case DarwinPlatformKind::DriverKit:
     return "driverkit";
   }
@@ -1649,6 +1670,9 @@ struct DarwinPlatform {
     case DarwinPlatformKind::WatchOS:
       Opt = options::OPT_mwatchos_version_min_EQ;
       break;
+    case DarwinPlatformKind::XROS:
+      // xrOS always explicitly provides a version in the triple.
+      return;
     case DarwinPlatformKind::DriverKit:
       // DriverKit always explicitly provides a version in the triple.
       return;
@@ -1794,6 +1818,8 @@ struct DarwinPlatform {
       return DarwinPlatformKind::TvOS;
     case llvm::Triple::WatchOS:
       return DarwinPlatformKind::WatchOS;
+    case llvm::Triple::XROS:
+      return DarwinPlatformKind::XROS;
     case llvm::Triple::DriverKit:
       return DarwinPlatformKind::DriverKit;
     default:
@@ -1975,6 +2001,10 @@ inferDeploymentTargetFromSDK(DerivedArgList &Args,
       return DarwinPlatform::createFromSDK(
           Darwin::TvOS, Version,
           /*IsSimulator=*/SDK.starts_with("AppleTVSimulator"));
+    else if (SDK.starts_with("XR"))
+      return DarwinPlatform::createFromSDK(
+          Darwin::XROS, Version,
+          /*IsSimulator=*/SDK.contains("Simulator"));
     else if (SDK.starts_with("DriverKit"))
       return DarwinPlatform::createFromSDK(Darwin::DriverKit, Version);
     return std::nullopt;
@@ -2013,6 +2043,11 @@ std::string getOSVersion(llvm::Triple::OSType OS, const llvm::Triple &Triple,
   case llvm::Triple::WatchOS:
     OsVersion = Triple.getWatchOSVersion();
     break;
+  case llvm::Triple::XROS:
+    OsVersion = Triple.getOSVersion();
+    if (!OsVersion.getMajor())
+      OsVersion = OsVersion.withMajorReplaced(1);
+    break;
   case llvm::Triple::DriverKit:
     OsVersion = Triple.getDriverKitVersion();
     break;
@@ -2104,6 +2139,7 @@ std::optional getDeploymentTargetFromMTargetOSArg(
   case llvm::Triple::IOS:
   case llvm::Triple::TvOS:
   case llvm::Triple::WatchOS:
+  case llvm::Triple::XROS:
     break;
   default:
     TheDriver.Diag(diag::err_drv_invalid_os_in_arg)
@@ -2319,6 +2355,13 @@ void Darwin::AddDeploymentTarget(DerivedArgList &Args) const {
         Micro >= 100)
       getDriver().Diag(diag::err_drv_invalid_version_number)
           << OSTarget->getAsString(Args, Opts);
+  } else if (Platform == XROS) {
+    if (!Driver::GetReleaseVersion(OSTarget->getOSVersion(), Major, Minor,
+                                   Micro, HadExtra) ||
+        HadExtra || Major < 1 || Major >= MajorVersionLimit || Minor >= 100 ||
+        Micro >= 100)
+      getDriver().Diag(diag::err_drv_invalid_version_number)
+          << OSTarget->getAsString(Args, Opts);
   } else
     llvm_unreachable("unknown kind of Darwin platform");
 
@@ -2641,6 +2684,10 @@ void DarwinClang::AddCCKextLibArgs(const ArgList &Args,
     llvm::sys::path::append(P, "libclang_rt.cc_kext_ios.a");
   } else if (isTargetDriverKit()) {
     // DriverKit doesn't want extra runtime support.
+  } else if (isTargetXROSDevice()) {
+    llvm::sys::path::append(
+        P, llvm::Twine("libclang_rt.cc_kext_") +
+               llvm::Triple::getOSTypeName(llvm::Triple::XROS) + ".a");
   } else {
     llvm::sys::path::append(P, "libclang_rt.cc_kext.a");
   }
@@ -2855,6 +2902,8 @@ bool Darwin::isAlignedAllocationUnavailable() const {
   case WatchOS: // Earlier than 4.0.
     OS = llvm::Triple::WatchOS;
     break;
+  case XROS: // Always available.
+    return false;
   case DriverKit: // Always available.
     return false;
   }
@@ -2876,6 +2925,8 @@ static bool sdkSupportsBuiltinModules(const Darwin::DarwinPlatformKind &TargetPl
     return SDKVersion >= VersionTuple(99U);
   case Darwin::WatchOS:
     return SDKVersion >= VersionTuple(99U);
+  case Darwin::XROS:
+    return SDKVersion >= VersionTuple(99U);
   default:
     return true;
   }
@@ -2917,6 +2968,10 @@ void Darwin::addClangTargetOptions(const llvm::opt::ArgList &DriverArgs,
   // to fix the same problem with C++ headers, and is generally fragile.
   if (!sdkSupportsBuiltinModules(TargetPlatform, SDKInfo))
     CC1Args.push_back("-fbuiltin-headers-in-system-modules");
+
+  if (!DriverArgs.hasArgNoClaim(options::OPT_fdefine_target_os_macros,
+                                options::OPT_fno_define_target_os_macros))
+    CC1Args.push_back("-fdefine-target-os-macros");
 }
 
 void Darwin::addClangCC1ASTargetOptions(
@@ -2994,7 +3049,7 @@ Darwin::TranslateArgs(const DerivedArgList &Args, StringRef BoundArch,
   // FIXME: It would be far better to avoid inserting those -static arguments,
   // but we can't check the deployment target in the translation code until
   // it is set here.
-  if (isTargetWatchOSBased() || isTargetDriverKit() ||
+  if (isTargetWatchOSBased() || isTargetDriverKit() || isTargetXROS() ||
       (isTargetIOSBased() && !isIPhoneOSVersionLT(6, 0))) {
     for (ArgList::iterator it = DAL->begin(), ie = DAL->end(); it != ie; ) {
       Arg *A = *it;
@@ -3088,6 +3143,8 @@ void Darwin::addMinVersionArgs(const ArgList &Args,
                                ArgStringList &CmdArgs) const {
   VersionTuple TargetVersion = getTripleTargetVersion();
 
+  assert(!isTargetXROS() && "xrOS always uses -platform-version");
+
   if (isTargetWatchOS())
     CmdArgs.push_back("-watchos_version_min");
   else if (isTargetWatchOSSimulator())
@@ -3147,6 +3204,8 @@ static const char *getPlatformName(Darwin::DarwinPlatformKind Platform,
     return "tvos";
   case Darwin::WatchOS:
     return "watchos";
+  case Darwin::XROS:
+    return "xros";
   case Darwin::DriverKit:
     return "driverkit";
   }
@@ -3343,7 +3402,7 @@ void Darwin::addStartObjectFileArgs(const ArgList &Args,
 }
 
 void Darwin::CheckObjCARC() const {
-  if (isTargetIOSBased() || isTargetWatchOSBased() ||
+  if (isTargetIOSBased() || isTargetWatchOSBased() || isTargetXROS() ||
       (isTargetMacOSBased() && !isMacosxVersionLT(10, 6)))
     return;
   getDriver().Diag(diag::err_arc_unsupported_on_toolchain);
diff --git a/clang/lib/Driver/ToolChains/Darwin.h b/clang/lib/Driver/ToolChains/Darwin.h
index 815449ae8f379..fda3b1a3639a8 100644
--- a/clang/lib/Driver/ToolChains/Darwin.h
+++ b/clang/lib/Driver/ToolChains/Darwin.h
@@ -65,7 +65,7 @@ class LLVM_LIBRARY_VISIBILITY Linker : public MachOTool {
   void AddLinkArgs(Compilation &C, const llvm::opt::ArgList &Args,
                    llvm::opt::ArgStringList &CmdArgs,
                    const InputInfoList &Inputs, VersionTuple Version,
-                   bool LinkerIsLLD) const;
+                   bool LinkerIsLLD, bool UsePlatformVersion) const;
 
 public:
   Linker(const ToolChain &TC) : MachOTool("darwin::Linker", "linker", TC) {}
@@ -298,6 +298,7 @@ class LLVM_LIBRARY_VISIBILITY Darwin : public MachO {
     TvOS,
     WatchOS,
     DriverKit,
+    XROS,
     LastDarwinPlatform = DriverKit
   };
   enum DarwinEnvironmentKind {
@@ -404,6 +405,16 @@ class LLVM_LIBRARY_VISIBILITY Darwin : public MachO {
     return isTargetIPhoneOS() || isTargetIOSSimulator();
   }
 
+  bool isTargetXROSDevice() const {
+    return TargetPlatform == XROS && TargetEnvironment == NativeEnvironment;
+  }
+
+  bool isTargetXROSSimulator() const {
+    return TargetPlatform == XROS && TargetEnvironment == Simulator;
+  }
+
+  bool isTargetXROS() const { return TargetPlatform == XROS; }
+
   bool isTargetTvOS() const {
     assert(TargetInitialized && "Target not initialized!");
     return TargetPlatform == TvOS && TargetEnvironment == NativeEnvironment;
@@ -546,7 +557,8 @@ class LLVM_LIBRARY_VISIBILITY Darwin : public MachO {
   GetDefaultStackProtectorLevel(bool KernelOrKext) const override {
     // Stack protectors default to on for user code on 10.5,
     // and for everything in 10.6 and beyond
-    if (isTargetIOSBased() || isTargetWatchOSBased() || isTargetDriverKit())
+    if (isTargetIOSBased() || isTargetWatchOSBased() || isTargetDriverKit() ||
+        isTargetXROS())
       return LangOptions::SSPOn;
     else if (isTargetMacOSBased() && !isMacosxVersionLT(10, 6))
       return LangOptions::SSPOn;
diff --git a/clang/lib/Driver/ToolChains/Flang.cpp b/clang/lib/Driver/ToolChains/Flang.cpp
index 422209e6c2615..03d68c3df7fb3 100644
--- a/clang/lib/Driver/ToolChains/Flang.cpp
+++ b/clang/lib/Driver/ToolChains/Flang.cpp
@@ -403,12 +403,10 @@ void Flang::addTargetOptions(const ArgList &Args,
         if (A->getValue() == StringRef{"Accelerate"}) {
           CmdArgs.push_back("-framework");
           CmdArgs.push_back("Accelerate");
-          A->render(Args, CmdArgs);
         }
       }
-    } else {
-      A->render(Args, CmdArgs);
     }
+    A->render(Args, CmdArgs);
   }
 
   if (Triple.isKnownWindowsMSVCEnvironment()) {
diff --git a/clang/lib/Driver/ToolChains/Gnu.cpp b/clang/lib/Driver/ToolChains/Gnu.cpp
index e2e89f9c502bb..06d67e89b43f7 100644
--- a/clang/lib/Driver/ToolChains/Gnu.cpp
+++ b/clang/lib/Driver/ToolChains/Gnu.cpp
@@ -550,6 +550,11 @@ void tools::gnutools::Linker::ConstructJob(Compilation &C, const JobAction &JA,
 
     // Add crtfastmath.o if available and fast math is enabled.
     ToolChain.addFastMathRuntimeIfAvailable(Args, CmdArgs);
+
+    if (isAndroid && Args.hasFlag(options::OPT_fandroid_pad_segment,
+                                  options::OPT_fno_android_pad_segment, false))
+      CmdArgs.push_back(
+          Args.MakeArgString(ToolChain.GetFilePath("crt_pad_segment.o")));
   }
 
   // Performing link for dependency file information, undefined symbols are OK.
@@ -2624,7 +2629,7 @@ void Generic_GCC::GCCInstallationDetector::AddDefaultGCCPrefixes(
   static const char *const X86Triples[] = {
       "i586-linux-gnu",      "i686-linux-gnu",        "i686-pc-linux-gnu",
       "i386-redhat-linux6E", "i686-redhat-linux",     "i386-redhat-linux",
-      "i586-suse-linux",     "i686-montavista-linux", "i686-gnu",
+      "i586-suse-linux",     "i686-montavista-linux",
   };
 
   static const char *const LoongArch64LibDirs[] = {"/lib64", "/lib"};
@@ -2791,6 +2796,23 @@ void Generic_GCC::GCCInstallationDetector::AddDefaultGCCPrefixes(
     return;
   }
 
+  if (TargetTriple.isOSHurd()) {
+    switch (TargetTriple.getArch()) {
+    case llvm::Triple::x86_64:
+      LibDirs.append(begin(X86_64LibDirs), end(X86_64LibDirs));
+      TripleAliases.push_back("x86_64-gnu");
+      break;
+    case llvm::Triple::x86:
+      LibDirs.append(begin(X86LibDirs), end(X86LibDirs));
+      TripleAliases.push_back("i686-gnu");
+      break;
+    default:
+      break;
+    }
+
+    return;
+  }
+
   switch (TargetTriple.getArch()) {
   case llvm::Triple::aarch64:
     LibDirs.append(begin(AArch64LibDirs), end(AArch64LibDirs));
diff --git a/clang/lib/Driver/ToolChains/Hurd.cpp b/clang/lib/Driver/ToolChains/Hurd.cpp
index 7a4c2bb7ede1d..0bc114b90ffc0 100644
--- a/clang/lib/Driver/ToolChains/Hurd.cpp
+++ b/clang/lib/Driver/ToolChains/Hurd.cpp
@@ -30,13 +30,21 @@ using tools::addPathIfExists;
 std::string Hurd::getMultiarchTriple(const Driver &D,
                                      const llvm::Triple &TargetTriple,
                                      StringRef SysRoot) const {
-  if (TargetTriple.getArch() == llvm::Triple::x86) {
+  switch (TargetTriple.getArch()) {
+  default:
+    break;
+
+  case llvm::Triple::x86:
     // We use the existence of '/lib/' as a directory to detect some
     // common hurd triples that don't quite match the Clang triple for both
     // 32-bit and 64-bit targets. Multiarch fixes its install triples to these
     // regardless of what the actual target triple is.
     if (D.getVFS().exists(SysRoot + "/lib/i386-gnu"))
       return "i386-gnu";
+    break;
+
+  case llvm::Triple::x86_64:
+    return "x86_64-gnu";
   }
 
   // For most architectures, just use whatever we have rather than trying to be
@@ -126,8 +134,14 @@ Tool *Hurd::buildAssembler() const {
 }
 
 std::string Hurd::getDynamicLinker(const ArgList &Args) const {
-  if (getArch() == llvm::Triple::x86)
+  switch (getArch()) {
+  case llvm::Triple::x86:
     return "/lib/ld.so";
+  case llvm::Triple::x86_64:
+    return "/lib/ld-x86-64.so.1";
+  default:
+    break;
+  }
 
   llvm_unreachable("unsupported architecture");
 }
diff --git a/clang/lib/Format/ContinuationIndenter.cpp b/clang/lib/Format/ContinuationIndenter.cpp
index 102504182c450..e6eaaa9ab4570 100644
--- a/clang/lib/Format/ContinuationIndenter.cpp
+++ b/clang/lib/Format/ContinuationIndenter.cpp
@@ -1591,6 +1591,9 @@ unsigned ContinuationIndenter::moveStateToNextToken(LineState &State,
     State.StartOfStringLiteral = State.Column + 1;
   if (Current.is(TT_CSharpStringLiteral) && State.StartOfStringLiteral == 0) {
     State.StartOfStringLiteral = State.Column + 1;
+  } else if (Current.is(TT_TableGenMultiLineString) &&
+             State.StartOfStringLiteral == 0) {
+    State.StartOfStringLiteral = State.Column + 1;
   } else if (Current.isStringLiteral() && State.StartOfStringLiteral == 0) {
     State.StartOfStringLiteral = State.Column;
   } else if (!Current.isOneOf(tok::comment, tok::identifier, tok::hash) &&
diff --git a/clang/lib/Format/Format.cpp b/clang/lib/Format/Format.cpp
index ff5ed6c306f38..7c2f4dcf3d230 100644
--- a/clang/lib/Format/Format.cpp
+++ b/clang/lib/Format/Format.cpp
@@ -1054,6 +1054,8 @@ template <> struct MappingTraits {
                    Style.PenaltyBreakFirstLessLess);
     IO.mapOptional("PenaltyBreakOpenParenthesis",
                    Style.PenaltyBreakOpenParenthesis);
+    IO.mapOptional("PenaltyBreakScopeResolution",
+                   Style.PenaltyBreakScopeResolution);
     IO.mapOptional("PenaltyBreakString", Style.PenaltyBreakString);
     IO.mapOptional("PenaltyBreakTemplateDeclaration",
                    Style.PenaltyBreakTemplateDeclaration);
@@ -1602,6 +1604,7 @@ FormatStyle getLLVMStyle(FormatStyle::LanguageKind Language) {
   LLVMStyle.PenaltyReturnTypeOnItsOwnLine = 60;
   LLVMStyle.PenaltyBreakBeforeFirstCallParameter = 19;
   LLVMStyle.PenaltyBreakOpenParenthesis = 0;
+  LLVMStyle.PenaltyBreakScopeResolution = 500;
   LLVMStyle.PenaltyBreakTemplateDeclaration = prec::Relational;
   LLVMStyle.PenaltyIndentedWhitespace = 0;
 
@@ -1698,8 +1701,6 @@ FormatStyle getGoogleStyle(FormatStyle::LanguageKind Language) {
           /*BasedOnStyle=*/"google",
       },
   };
-  GoogleStyle.AttributeMacros.push_back("GUARDED_BY");
-  GoogleStyle.AttributeMacros.push_back("ABSL_GUARDED_BY");
 
   GoogleStyle.SpacesBeforeTrailingComments = 2;
   GoogleStyle.Standard = FormatStyle::LS_Auto;
diff --git a/clang/lib/Format/FormatToken.h b/clang/lib/Format/FormatToken.h
index 3f9664f8f78a3..dede89f260015 100644
--- a/clang/lib/Format/FormatToken.h
+++ b/clang/lib/Format/FormatToken.h
@@ -148,6 +148,7 @@ namespace format {
   TYPE(StructLBrace)                                                           \
   TYPE(StructRBrace)                                                           \
   TYPE(StructuredBindingLSquare)                                               \
+  TYPE(TableGenMultiLineString)                                                \
   TYPE(TemplateCloser)                                                         \
   TYPE(TemplateOpener)                                                         \
   TYPE(TemplateString)                                                         \
@@ -275,14 +276,15 @@ class AnnotatedLine;
 struct FormatToken {
   FormatToken()
       : HasUnescapedNewline(false), IsMultiline(false), IsFirst(false),
-        MustBreakBefore(false), IsUnterminatedLiteral(false),
-        CanBreakBefore(false), ClosesTemplateDeclaration(false),
-        StartsBinaryExpression(false), EndsBinaryExpression(false),
-        PartOfMultiVariableDeclStmt(false), ContinuesLineCommentSection(false),
-        Finalized(false), ClosesRequiresClause(false),
-        EndsCppAttributeGroup(false), BlockKind(BK_Unknown),
-        Decision(FD_Unformatted), PackingKind(PPK_Inconclusive),
-        TypeIsFinalized(false), Type(TT_Unknown) {}
+        MustBreakBefore(false), MustBreakBeforeFinalized(false),
+        IsUnterminatedLiteral(false), CanBreakBefore(false),
+        ClosesTemplateDeclaration(false), StartsBinaryExpression(false),
+        EndsBinaryExpression(false), PartOfMultiVariableDeclStmt(false),
+        ContinuesLineCommentSection(false), Finalized(false),
+        ClosesRequiresClause(false), EndsCppAttributeGroup(false),
+        BlockKind(BK_Unknown), Decision(FD_Unformatted),
+        PackingKind(PPK_Inconclusive), TypeIsFinalized(false),
+        Type(TT_Unknown) {}
 
   /// The \c Token.
   Token Tok;
@@ -318,6 +320,10 @@ struct FormatToken {
   /// before the token.
   unsigned MustBreakBefore : 1;
 
+  /// Whether MustBreakBefore is finalized during parsing and must not
+  /// be reset between runs.
+  unsigned MustBreakBeforeFinalized : 1;
+
   /// Set to \c true if this token is an unterminated literal.
   unsigned IsUnterminatedLiteral : 1;
 
@@ -416,10 +422,14 @@ struct FormatToken {
   /// to another one please use overwriteFixedType, or even better remove the
   /// need to reassign the type.
   void setFinalizedType(TokenType T) {
+    if (MacroCtx && MacroCtx->Role == MR_UnexpandedArg)
+      return;
     Type = T;
     TypeIsFinalized = true;
   }
   void overwriteFixedType(TokenType T) {
+    if (MacroCtx && MacroCtx->Role == MR_UnexpandedArg)
+      return;
     TypeIsFinalized = false;
     setType(T);
   }
@@ -1202,6 +1212,21 @@ struct AdditionalKeywords {
     kw_verilogHashHash = &IdentTable.get("##");
     kw_apostrophe = &IdentTable.get("\'");
 
+    // TableGen keywords
+    kw_bit = &IdentTable.get("bit");
+    kw_bits = &IdentTable.get("bits");
+    kw_code = &IdentTable.get("code");
+    kw_dag = &IdentTable.get("dag");
+    kw_def = &IdentTable.get("def");
+    kw_defm = &IdentTable.get("defm");
+    kw_defset = &IdentTable.get("defset");
+    kw_defvar = &IdentTable.get("defvar");
+    kw_dump = &IdentTable.get("dump");
+    kw_include = &IdentTable.get("include");
+    kw_list = &IdentTable.get("list");
+    kw_multiclass = &IdentTable.get("multiclass");
+    kw_then = &IdentTable.get("then");
+
     // Keep this at the end of the constructor to make sure everything here
     // is
     // already initialized.
@@ -1294,6 +1319,27 @@ struct AdditionalKeywords {
          kw_wildcard,     kw_wire,
          kw_with,         kw_wor,
          kw_verilogHash,  kw_verilogHashHash});
+
+    TableGenExtraKeywords = std::unordered_set({
+        kw_assert,
+        kw_bit,
+        kw_bits,
+        kw_code,
+        kw_dag,
+        kw_def,
+        kw_defm,
+        kw_defset,
+        kw_defvar,
+        kw_dump,
+        kw_foreach,
+        kw_in,
+        kw_include,
+        kw_let,
+        kw_list,
+        kw_multiclass,
+        kw_string,
+        kw_then,
+    });
   }
 
   // Context sensitive keywords.
@@ -1539,6 +1585,21 @@ struct AdditionalKeywords {
   // Symbols in Verilog that don't exist in C++.
   IdentifierInfo *kw_apostrophe;
 
+  // TableGen keywords
+  IdentifierInfo *kw_bit;
+  IdentifierInfo *kw_bits;
+  IdentifierInfo *kw_code;
+  IdentifierInfo *kw_dag;
+  IdentifierInfo *kw_def;
+  IdentifierInfo *kw_defm;
+  IdentifierInfo *kw_defset;
+  IdentifierInfo *kw_defvar;
+  IdentifierInfo *kw_dump;
+  IdentifierInfo *kw_include;
+  IdentifierInfo *kw_list;
+  IdentifierInfo *kw_multiclass;
+  IdentifierInfo *kw_then;
+
   /// Returns \c true if \p Tok is a keyword or an identifier.
   bool isWordLike(const FormatToken &Tok) const {
     // getIdentifierinfo returns non-null for keywords as well as identifiers.
@@ -1811,6 +1872,27 @@ struct AdditionalKeywords {
     }
   }
 
+  bool isTableGenDefinition(const FormatToken &Tok) const {
+    return Tok.isOneOf(kw_def, kw_defm, kw_defset, kw_defvar, kw_multiclass,
+                       kw_let, tok::kw_class);
+  }
+
+  bool isTableGenKeyword(const FormatToken &Tok) const {
+    switch (Tok.Tok.getKind()) {
+    case tok::kw_class:
+    case tok::kw_else:
+    case tok::kw_false:
+    case tok::kw_if:
+    case tok::kw_int:
+    case tok::kw_true:
+      return true;
+    default:
+      return Tok.is(tok::identifier) &&
+             TableGenExtraKeywords.find(Tok.Tok.getIdentifierInfo()) !=
+                 TableGenExtraKeywords.end();
+    }
+  }
+
 private:
   /// The JavaScript keywords beyond the C++ keyword set.
   std::unordered_set JsExtraKeywords;
@@ -1820,6 +1902,9 @@ struct AdditionalKeywords {
 
   /// The Verilog keywords beyond the C++ keyword set.
   std::unordered_set VerilogExtraKeywords;
+
+  /// The TableGen keywords beyond the C++ keyword set.
+  std::unordered_set TableGenExtraKeywords;
 };
 
 inline bool isLineComment(const FormatToken &FormatTok) {
diff --git a/clang/lib/Format/FormatTokenLexer.cpp b/clang/lib/Format/FormatTokenLexer.cpp
index 61430282c6f88..25ac9be57c81a 100644
--- a/clang/lib/Format/FormatTokenLexer.cpp
+++ b/clang/lib/Format/FormatTokenLexer.cpp
@@ -93,6 +93,8 @@ ArrayRef FormatTokenLexer::lex() {
       // string literals are correctly identified.
       handleCSharpVerbatimAndInterpolatedStrings();
     }
+    if (Style.isTableGen())
+      handleTableGenMultilineString();
     if (Tokens.back()->NewlinesBefore > 0 || Tokens.back()->IsMultiline)
       FirstInLineIndex = Tokens.size() - 1;
   } while (Tokens.back()->isNot(tok::eof));
@@ -272,6 +274,14 @@ void FormatTokenLexer::tryMergePreviousTokens() {
       return;
     }
   }
+  // TableGen's Multi line string starts with [{
+  if (Style.isTableGen() && tryMergeTokens({tok::l_square, tok::l_brace},
+                                           TT_TableGenMultiLineString)) {
+    // Set again with finalizing. This must never be annotated as other types.
+    Tokens.back()->setFinalizedType(TT_TableGenMultiLineString);
+    Tokens.back()->Tok.setKind(tok::string_literal);
+    return;
+  }
 }
 
 bool FormatTokenLexer::tryMergeNSStringLiteral() {
@@ -763,6 +773,37 @@ void FormatTokenLexer::handleCSharpVerbatimAndInterpolatedStrings() {
   resetLexer(SourceMgr.getFileOffset(Lex->getSourceLocation(Offset + 1)));
 }
 
+void FormatTokenLexer::handleTableGenMultilineString() {
+  FormatToken *MultiLineString = Tokens.back();
+  if (MultiLineString->isNot(TT_TableGenMultiLineString))
+    return;
+
+  auto OpenOffset = Lex->getCurrentBufferOffset() - 2 /* "[{" */;
+  // "}]" is the end of multi line string.
+  auto CloseOffset = Lex->getBuffer().find("}]", OpenOffset);
+  if (CloseOffset == StringRef::npos)
+    return;
+  auto Text = Lex->getBuffer().substr(OpenOffset, CloseOffset + 2);
+  MultiLineString->TokenText = Text;
+  resetLexer(SourceMgr.getFileOffset(
+      Lex->getSourceLocation(Lex->getBufferLocation() - 2 + Text.size())));
+  auto FirstLineText = Text;
+  auto FirstBreak = Text.find('\n');
+  // Set ColumnWidth and LastLineColumnWidth when it has multiple lines.
+  if (FirstBreak != StringRef::npos) {
+    MultiLineString->IsMultiline = true;
+    FirstLineText = Text.substr(0, FirstBreak + 1);
+    // LastLineColumnWidth holds the width of the last line.
+    auto LastBreak = Text.rfind('\n');
+    MultiLineString->LastLineColumnWidth = encoding::columnWidthWithTabs(
+        Text.substr(LastBreak + 1), MultiLineString->OriginalColumn,
+        Style.TabWidth, Encoding);
+  }
+  // ColumnWidth holds only the width of the first line.
+  MultiLineString->ColumnWidth = encoding::columnWidthWithTabs(
+      FirstLineText, MultiLineString->OriginalColumn, Style.TabWidth, Encoding);
+}
+
 void FormatTokenLexer::handleTemplateStrings() {
   FormatToken *BacktickToken = Tokens.back();
 
@@ -1182,6 +1223,9 @@ FormatToken *FormatTokenLexer::getNextToken() {
                                   tok::kw_operator)) {
       FormatTok->Tok.setKind(tok::identifier);
       FormatTok->Tok.setIdentifierInfo(nullptr);
+    } else if (Style.isTableGen() && !Keywords.isTableGenKeyword(*FormatTok)) {
+      FormatTok->Tok.setKind(tok::identifier);
+      FormatTok->Tok.setIdentifierInfo(nullptr);
     }
   } else if (FormatTok->is(tok::greatergreater)) {
     FormatTok->Tok.setKind(tok::greater);
diff --git a/clang/lib/Format/FormatTokenLexer.h b/clang/lib/Format/FormatTokenLexer.h
index bb6a8ab69c1be..1dec6bbc41514 100644
--- a/clang/lib/Format/FormatTokenLexer.h
+++ b/clang/lib/Format/FormatTokenLexer.h
@@ -95,6 +95,9 @@ class FormatTokenLexer {
 
   void handleCSharpVerbatimAndInterpolatedStrings();
 
+  // Handles TableGen multiline strings. It has the form [{ ... }].
+  void handleTableGenMultilineString();
+
   void tryParsePythonComment();
 
   bool tryMerge_TMacro();
diff --git a/clang/lib/Format/TokenAnnotator.cpp b/clang/lib/Format/TokenAnnotator.cpp
index 8b43438c72dfe..25fcceb878643 100644
--- a/clang/lib/Format/TokenAnnotator.cpp
+++ b/clang/lib/Format/TokenAnnotator.cpp
@@ -2209,7 +2209,8 @@ class AnnotatingParser {
         (!NextNonComment && !Line.InMacroBody) ||
         (NextNonComment &&
          (NextNonComment->isPointerOrReference() ||
-          NextNonComment->isOneOf(tok::identifier, tok::string_literal)))) {
+          NextNonComment->is(tok::string_literal) ||
+          (Line.InPragmaDirective && NextNonComment->is(tok::identifier))))) {
       return false;
     }
 
@@ -2234,6 +2235,12 @@ class AnnotatingParser {
     if (PreviousNotConst->ClosesRequiresClause)
       return false;
 
+    if (Style.isTableGen()) {
+      // keywords such as let and def* defines names.
+      if (Keywords.isTableGenDefinition(*PreviousNotConst))
+        return true;
+    }
+
     bool IsPPKeyword = PreviousNotConst->is(tok::identifier) &&
                        PreviousNotConst->Previous &&
                        PreviousNotConst->Previous->is(tok::hash);
@@ -2369,7 +2376,7 @@ class AnnotatingParser {
       }
     }
 
-    if (Tok.Next->is(tok::question))
+    if (Tok.Next->isOneOf(tok::question, tok::ampamp))
       return false;
 
     // `foreach((A a, B b) in someList)` should not be seen as a cast.
@@ -2769,13 +2776,6 @@ class ExpressionParser {
       // Consume operators with higher precedence.
       parse(Precedence + 1);
 
-      // Do not assign fake parenthesis to tokens that are part of an
-      // unexpanded macro call. The line within the macro call contains
-      // the parenthesis and commas, and we will not find operators within
-      // that structure.
-      if (Current && Current->MacroParent)
-        break;
-
       int CurrentPrecedence = getCurrentPrecedence();
 
       if (Precedence == CurrentPrecedence && Current &&
@@ -2919,6 +2919,13 @@ class ExpressionParser {
 
   void addFakeParenthesis(FormatToken *Start, prec::Level Precedence,
                           FormatToken *End = nullptr) {
+    // Do not assign fake parenthesis to tokens that are part of an
+    // unexpanded macro call. The line within the macro call contains
+    // the parenthesis and commas, and we will not find operators within
+    // that structure.
+    if (Start->MacroParent)
+      return;
+
     Start->FakeLParens.push_back(Precedence);
     if (Precedence > prec::Unknown)
       Start->StartsBinaryExpression = true;
@@ -3759,7 +3766,7 @@ unsigned TokenAnnotator::splitPenalty(const AnnotatedLine &Line,
   }
 
   if (Left.is(tok::coloncolon))
-    return 500;
+    return Style.PenaltyBreakScopeResolution;
   if (Right.isOneOf(TT_StartOfName, TT_FunctionDeclarationName) ||
       Right.is(tok::kw_operator)) {
     if (Line.startsWith(tok::kw_for) && Right.PartOfMultiVariableDeclStmt)
@@ -4668,6 +4675,10 @@ bool TokenAnnotator::spaceRequiredBefore(const AnnotatedLine &Line,
   } else if (Style.Language == FormatStyle::LK_Java) {
     if (Left.is(tok::r_square) && Right.is(tok::l_brace))
       return true;
+    // spaces inside square brackets.
+    if (Left.is(tok::l_square) || Right.is(tok::r_square))
+      return Style.SpacesInSquareBrackets;
+
     if (Left.is(Keywords.kw_synchronized) && Right.is(tok::l_paren)) {
       return Style.SpaceBeforeParensOptions.AfterControlStatements ||
              spaceRequiredBeforeParens(Right);
diff --git a/clang/lib/Format/UnwrappedLineFormatter.cpp b/clang/lib/Format/UnwrappedLineFormatter.cpp
index 56077499c39d5..adeb072434873 100644
--- a/clang/lib/Format/UnwrappedLineFormatter.cpp
+++ b/clang/lib/Format/UnwrappedLineFormatter.cpp
@@ -95,7 +95,7 @@ class LevelIndentTracker {
   /// level to the same indent.
   /// Note that \c nextLine must have been called before this method.
   void adjustToUnmodifiedLine(const AnnotatedLine &Line) {
-    if (Line.InPPDirective)
+    if (Line.InPPDirective || Line.IsContinuation)
       return;
     assert(Line.Level < IndentForLevel.size());
     if (Line.First->is(tok::comment) && IndentForLevel[Line.Level] != -1)
@@ -954,13 +954,15 @@ static void markFinalized(FormatToken *Tok) {
       // will be modified as unexpanded arguments (as part of the macro call
       // formatting) in the next pass.
       Tok->MacroCtx->Role = MR_UnexpandedArg;
-      // Reset whether spaces are required before this token, as that is context
-      // dependent, and that context may change when formatting the macro call.
-      // For example, given M(x) -> 2 * x, and the macro call M(var),
-      // the token 'var' will have SpacesRequiredBefore = 1 after being
+      // Reset whether spaces or a line break are required before this token, as
+      // that is context dependent, and that context may change when formatting
+      // the macro call.  For example, given M(x) -> 2 * x, and the macro call
+      // M(var), the token 'var' will have SpacesRequiredBefore = 1 after being
       // formatted as part of the expanded macro, but SpacesRequiredBefore = 0
       // for its position within the macro call.
       Tok->SpacesRequiredBefore = 0;
+      if (!Tok->MustBreakBeforeFinalized)
+        Tok->MustBreakBefore = 0;
     } else {
       Tok->Finalized = true;
     }
diff --git a/clang/lib/Format/UnwrappedLineParser.cpp b/clang/lib/Format/UnwrappedLineParser.cpp
index d4aa3735c2a54..50d41c9f57a67 100644
--- a/clang/lib/Format/UnwrappedLineParser.cpp
+++ b/clang/lib/Format/UnwrappedLineParser.cpp
@@ -4675,6 +4675,7 @@ void UnwrappedLineParser::readToken(int LevelDifference) {
         conditionalCompilationEnd();
       FormatTok = Tokens->getNextToken();
       FormatTok->MustBreakBefore = true;
+      FormatTok->MustBreakBeforeFinalized = true;
     }
 
     auto IsFirstNonCommentOnLine = [](bool FirstNonCommentOnLine,
@@ -4891,6 +4892,7 @@ void UnwrappedLineParser::pushToken(FormatToken *Tok) {
   Line->Tokens.push_back(UnwrappedLineNode(Tok));
   if (MustBreakBeforeNextToken) {
     Line->Tokens.back().Tok->MustBreakBefore = true;
+    Line->Tokens.back().Tok->MustBreakBeforeFinalized = true;
     MustBreakBeforeNextToken = false;
   }
 }
diff --git a/clang/lib/Format/WhitespaceManager.cpp b/clang/lib/Format/WhitespaceManager.cpp
index f1d176f182ffa..8415c8d360d64 100644
--- a/clang/lib/Format/WhitespaceManager.cpp
+++ b/clang/lib/Format/WhitespaceManager.cpp
@@ -1451,8 +1451,10 @@ WhitespaceManager::CellDescriptions WhitespaceManager::getCells(unsigned Start,
       } else if (C.Tok->is(tok::comma)) {
         if (!Cells.empty())
           Cells.back().EndIndex = i;
-        if (C.Tok->getNextNonComment()->isNot(tok::r_brace)) // dangling comma
+        if (const auto *Next = C.Tok->getNextNonComment();
+            Next && Next->isNot(tok::r_brace)) { // dangling comma
           ++Cell;
+        }
       }
     } else if (Depth == 1) {
       if (C.Tok == MatchingParen) {
diff --git a/clang/lib/Frontend/CompilerInstance.cpp b/clang/lib/Frontend/CompilerInstance.cpp
index 578306d61391f..3383b4a0fcbfc 100644
--- a/clang/lib/Frontend/CompilerInstance.cpp
+++ b/clang/lib/Frontend/CompilerInstance.cpp
@@ -2206,7 +2206,7 @@ void CompilerInstance::createModuleFromSource(SourceLocation ImportLoc,
   // Build the module, inheriting any modules that we've built locally.
   if (compileModuleImpl(*this, ImportLoc, ModuleName, Input, StringRef(),
                         ModuleFileName, PreBuildStep, PostBuildStep)) {
-    BuiltModules[std::string(ModuleName)] = std::string(ModuleFileName.str());
+    BuiltModules[std::string(ModuleName)] = std::string(ModuleFileName);
     llvm::sys::RemoveFileOnSignal(ModuleFileName);
   }
 }
diff --git a/clang/lib/Frontend/Rewrite/FrontendActions.cpp b/clang/lib/Frontend/Rewrite/FrontendActions.cpp
index 14569013b92ce..92921bf6f3d7b 100644
--- a/clang/lib/Frontend/Rewrite/FrontendActions.cpp
+++ b/clang/lib/Frontend/Rewrite/FrontendActions.cpp
@@ -77,7 +77,7 @@ class FixItActionSuffixInserter : public FixItOptions {
     SmallString<128> Path(Filename);
     llvm::sys::path::replace_extension(Path,
       NewSuffix + llvm::sys::path::extension(Path));
-    return std::string(Path.str());
+    return std::string(Path);
   }
 };
 
diff --git a/clang/lib/Headers/arm_acle.h b/clang/lib/Headers/arm_acle.h
index 9aae2285aeb1d..9cd34948e3c53 100644
--- a/clang/lib/Headers/arm_acle.h
+++ b/clang/lib/Headers/arm_acle.h
@@ -4,6 +4,13 @@
  * See https://llvm.org/LICENSE.txt for license information.
  * SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
  *
+ * The Arm C Language Extensions specifications can be found in the following
+ * link: https://github.com/ARM-software/acle/releases
+ *
+ * The ACLE section numbers are subject to change. When consulting the
+ * specifications, it is recommended to search using section titles if
+ * the section numbers look outdated.
+ *
  *===-----------------------------------------------------------------------===
  */
 
@@ -20,8 +27,8 @@
 extern "C" {
 #endif
 
-/* 8 SYNCHRONIZATION, BARRIER AND HINT INTRINSICS */
-/* 8.3 Memory barriers */
+/* 7 SYNCHRONIZATION, BARRIER AND HINT INTRINSICS */
+/* 7.3 Memory barriers */
 #if !__has_builtin(__dmb)
 #define __dmb(i) __builtin_arm_dmb(i)
 #endif
@@ -32,7 +39,7 @@ extern "C" {
 #define __isb(i) __builtin_arm_isb(i)
 #endif
 
-/* 8.4 Hints */
+/* 7.4 Hints */
 
 #if !__has_builtin(__wfi)
 static __inline__ void __attribute__((__always_inline__, __nodebug__)) __wfi(void) {
@@ -68,7 +75,7 @@ static __inline__ void __attribute__((__always_inline__, __nodebug__)) __yield(v
 #define __dbg(t) __builtin_arm_dbg(t)
 #endif
 
-/* 8.5 Swap */
+/* 7.5 Swap */
 static __inline__ uint32_t __attribute__((__always_inline__, __nodebug__))
 __swp(uint32_t __x, volatile uint32_t *__p) {
   uint32_t v;
@@ -78,8 +85,8 @@ __swp(uint32_t __x, volatile uint32_t *__p) {
   return v;
 }
 
-/* 8.6 Memory prefetch intrinsics */
-/* 8.6.1 Data prefetch */
+/* 7.6 Memory prefetch intrinsics */
+/* 7.6.1 Data prefetch */
 #define __pld(addr) __pldx(0, 0, 0, addr)
 
 #if defined(__ARM_32BIT_STATE) && __ARM_32BIT_STATE
@@ -90,7 +97,7 @@ __swp(uint32_t __x, volatile uint32_t *__p) {
   __builtin_arm_prefetch(addr, access_kind, cache_level, retention_policy, 1)
 #endif
 
-/* 8.6.2 Instruction prefetch */
+/* 7.6.2 Instruction prefetch */
 #define __pli(addr) __plix(0, 0, addr)
 
 #if defined(__ARM_32BIT_STATE) && __ARM_32BIT_STATE
@@ -101,15 +108,15 @@ __swp(uint32_t __x, volatile uint32_t *__p) {
   __builtin_arm_prefetch(addr, 0, cache_level, retention_policy, 0)
 #endif
 
-/* 8.7 NOP */
+/* 7.7 NOP */
 #if !defined(_MSC_VER) || !defined(__aarch64__)
 static __inline__ void __attribute__((__always_inline__, __nodebug__)) __nop(void) {
   __builtin_arm_nop();
 }
 #endif
 
-/* 9 DATA-PROCESSING INTRINSICS */
-/* 9.2 Miscellaneous data-processing intrinsics */
+/* 8 DATA-PROCESSING INTRINSICS */
+/* 8.2 Miscellaneous data-processing intrinsics */
 /* ROR */
 static __inline__ uint32_t __attribute__((__always_inline__, __nodebug__))
 __ror(uint32_t __x, uint32_t __y) {
@@ -248,9 +255,7 @@ __rbitl(unsigned long __t) {
 #endif
 }
 
-/*
- * 9.3 16-bit multiplications
- */
+/* 8.3 16-bit multiplications */
 #if defined(__ARM_FEATURE_DSP) && __ARM_FEATURE_DSP
 static __inline__ int32_t __attribute__((__always_inline__,__nodebug__))
 __smulbb(int32_t __a, int32_t __b) {
@@ -279,18 +284,18 @@ __smulwt(int32_t __a, int32_t __b) {
 #endif
 
 /*
- * 9.4 Saturating intrinsics
+ * 8.4 Saturating intrinsics
  *
  * FIXME: Change guard to their corresponding __ARM_FEATURE flag when Q flag
  * intrinsics are implemented and the flag is enabled.
  */
-/* 9.4.1 Width-specified saturation intrinsics */
+/* 8.4.1 Width-specified saturation intrinsics */
 #if defined(__ARM_FEATURE_SAT) && __ARM_FEATURE_SAT
 #define __ssat(x, y) __builtin_arm_ssat(x, y)
 #define __usat(x, y) __builtin_arm_usat(x, y)
 #endif
 
-/* 9.4.2 Saturating addition and subtraction intrinsics */
+/* 8.4.2 Saturating addition and subtraction intrinsics */
 #if defined(__ARM_FEATURE_DSP) && __ARM_FEATURE_DSP
 static __inline__ int32_t __attribute__((__always_inline__, __nodebug__))
 __qadd(int32_t __t, int32_t __v) {
@@ -308,7 +313,7 @@ __qdbl(int32_t __t) {
 }
 #endif
 
-/* 9.4.3 Accumultating multiplications */
+/* 8.4.3 Accumultating multiplications */
 #if defined(__ARM_FEATURE_DSP) && __ARM_FEATURE_DSP
 static __inline__ int32_t __attribute__((__always_inline__, __nodebug__))
 __smlabb(int32_t __a, int32_t __b, int32_t __c) {
@@ -337,13 +342,13 @@ __smlawt(int32_t __a, int32_t __b, int32_t __c) {
 #endif
 
 
-/* 9.5.4 Parallel 16-bit saturation */
+/* 8.5.4 Parallel 16-bit saturation */
 #if defined(__ARM_FEATURE_SIMD32) && __ARM_FEATURE_SIMD32
 #define __ssat16(x, y) __builtin_arm_ssat16(x, y)
 #define __usat16(x, y) __builtin_arm_usat16(x, y)
 #endif
 
-/* 9.5.5 Packing and unpacking */
+/* 8.5.5 Packing and unpacking */
 #if defined(__ARM_FEATURE_SIMD32) && __ARM_FEATURE_SIMD32
 typedef int32_t int8x4_t;
 typedef int32_t int16x2_t;
@@ -368,7 +373,7 @@ __uxtb16(int8x4_t __a) {
 }
 #endif
 
-/* 9.5.6 Parallel selection */
+/* 8.5.6 Parallel selection */
 #if defined(__ARM_FEATURE_SIMD32) && __ARM_FEATURE_SIMD32
 static __inline__ uint8x4_t __attribute__((__always_inline__, __nodebug__))
 __sel(uint8x4_t __a, uint8x4_t __b) {
@@ -376,7 +381,7 @@ __sel(uint8x4_t __a, uint8x4_t __b) {
 }
 #endif
 
-/* 9.5.7 Parallel 8-bit addition and subtraction */
+/* 8.5.7 Parallel 8-bit addition and subtraction */
 #if defined(__ARM_FEATURE_SIMD32) && __ARM_FEATURE_SIMD32
 static __inline__ int8x4_t __attribute__((__always_inline__, __nodebug__))
 __qadd8(int8x4_t __a, int8x4_t __b) {
@@ -428,7 +433,7 @@ __usub8(uint8x4_t __a, uint8x4_t __b) {
 }
 #endif
 
-/* 9.5.8 Sum of 8-bit absolute differences */
+/* 8.5.8 Sum of 8-bit absolute differences */
 #if defined(__ARM_FEATURE_SIMD32) && __ARM_FEATURE_SIMD32
 static __inline__ uint32_t __attribute__((__always_inline__, __nodebug__))
 __usad8(uint8x4_t __a, uint8x4_t __b) {
@@ -440,7 +445,7 @@ __usada8(uint8x4_t __a, uint8x4_t __b, uint32_t __c) {
 }
 #endif
 
-/* 9.5.9 Parallel 16-bit addition and subtraction */
+/* 8.5.9 Parallel 16-bit addition and subtraction */
 #if defined(__ARM_FEATURE_SIMD32) && __ARM_FEATURE_SIMD32
 static __inline__ int16x2_t __attribute__((__always_inline__, __nodebug__))
 __qadd16(int16x2_t __a, int16x2_t __b) {
@@ -540,7 +545,7 @@ __usub16(uint16x2_t __a, uint16x2_t __b) {
 }
 #endif
 
-/* 9.5.10 Parallel 16-bit multiplications */
+/* 8.5.10 Parallel 16-bit multiplications */
 #if defined(__ARM_FEATURE_SIMD32) && __ARM_FEATURE_SIMD32
 static __inline__ int32_t __attribute__((__always_inline__, __nodebug__))
 __smlad(int16x2_t __a, int16x2_t __b, int32_t __c) {
@@ -607,7 +612,7 @@ __rintnf(float __a) {
 }
 #endif
 
-/* 9.7 CRC32 intrinsics */
+/* 8.8 CRC32 intrinsics */
 #if (defined(__ARM_FEATURE_CRC32) && __ARM_FEATURE_CRC32) ||                   \
     (defined(__ARM_64BIT_STATE) && __ARM_64BIT_STATE)
 static __inline__ uint32_t __attribute__((__always_inline__, __nodebug__, target("crc")))
@@ -651,6 +656,7 @@ __crc32cd(uint32_t __a, uint64_t __b) {
 }
 #endif
 
+/* 8.6 Floating-point data-processing intrinsics */
 /* Armv8.3-A Javascript conversion intrinsic */
 #if defined(__ARM_64BIT_STATE) && __ARM_64BIT_STATE
 static __inline__ int32_t __attribute__((__always_inline__, __nodebug__, target("v8.3a")))
@@ -702,7 +708,7 @@ __rint64x(double __a) {
 }
 #endif
 
-/* Armv8.7-A load/store 64-byte intrinsics */
+/* 8.9 Armv8.7-A load/store 64-byte intrinsics */
 #if defined(__ARM_64BIT_STATE) && __ARM_64BIT_STATE
 typedef struct {
     uint64_t val[8];
@@ -728,7 +734,7 @@ __arm_st64bv0(void *__addr, data512_t __value) {
 }
 #endif
 
-/* 10.1 Special register intrinsics */
+/* 11.1 Special register intrinsics */
 #define __arm_rsr(sysreg) __builtin_arm_rsr(sysreg)
 #define __arm_rsr64(sysreg) __builtin_arm_rsr64(sysreg)
 #define __arm_rsr128(sysreg) __builtin_arm_rsr128(sysreg)
@@ -742,7 +748,7 @@ __arm_st64bv0(void *__addr, data512_t __value) {
 #define __arm_wsrf(sysreg, v) __arm_wsr(sysreg, __builtin_bit_cast(uint32_t, v))
 #define __arm_wsrf64(sysreg, v) __arm_wsr64(sysreg, __builtin_bit_cast(uint64_t, v))
 
-/* Memory Tagging Extensions (MTE) Intrinsics */
+/* 10.3 Memory Tagging Extensions (MTE) Intrinsics */
 #if defined(__ARM_64BIT_STATE) && __ARM_64BIT_STATE
 #define __arm_mte_create_random_tag(__ptr, __mask)  __builtin_arm_irg(__ptr, __mask)
 #define __arm_mte_increment_tag(__ptr, __tag_offset)  __builtin_arm_addg(__ptr, __tag_offset)
@@ -751,12 +757,12 @@ __arm_st64bv0(void *__addr, data512_t __value) {
 #define __arm_mte_set_tag(__ptr) __builtin_arm_stg(__ptr)
 #define __arm_mte_ptrdiff(__ptra, __ptrb) __builtin_arm_subp(__ptra, __ptrb)
 
-/* Memory Operations Intrinsics */
+/* 18 Memory Operations Intrinsics */
 #define __arm_mops_memset_tag(__tagged_address, __value, __size)    \
   __builtin_arm_mops_memset_tag(__tagged_address, __value, __size)
 #endif
 
-/* Coprocessor Intrinsics */
+/* 11.3 Coprocessor Intrinsics */
 #if defined(__ARM_FEATURE_COPROC)
 
 #if (__ARM_FEATURE_COPROC & 0x1)
@@ -815,7 +821,7 @@ __arm_st64bv0(void *__addr, data512_t __value) {
 
 #endif // __ARM_FEATURE_COPROC
 
-/* Transactional Memory Extension (TME) Intrinsics */
+/* 17 Transactional Memory Extension (TME) Intrinsics */
 #if defined(__ARM_FEATURE_TME) && __ARM_FEATURE_TME
 
 #define _TMFAILURE_REASON  0x00007fffu
@@ -837,7 +843,7 @@ __arm_st64bv0(void *__addr, data512_t __value) {
 
 #endif /* __ARM_FEATURE_TME */
 
-/* Armv8.5-A Random number generation intrinsics */
+/* 8.7 Armv8.5-A Random number generation intrinsics */
 #if defined(__ARM_64BIT_STATE) && __ARM_64BIT_STATE
 static __inline__ int __attribute__((__always_inline__, __nodebug__, target("rand")))
 __rndr(uint64_t *__p) {
diff --git a/clang/lib/Headers/ia32intrin.h b/clang/lib/Headers/ia32intrin.h
index a8b59dfaad898..7d5fede61ce85 100644
--- a/clang/lib/Headers/ia32intrin.h
+++ b/clang/lib/Headers/ia32intrin.h
@@ -58,7 +58,7 @@ __bsrd(int __A) {
   return 31 - __builtin_clz((unsigned int)__A);
 }
 
-/// Swaps the bytes in the input. Converting little endian to big endian or
+/// Swaps the bytes in the input, converting little endian to big endian or
 ///    vice versa.
 ///
 /// \headerfile 
@@ -73,6 +73,16 @@ __bswapd(int __A) {
   return (int)__builtin_bswap32((unsigned int)__A);
 }
 
+/// Swaps the bytes in the input, converting little endian to big endian or
+///    vice versa.
+///
+/// \headerfile 
+///
+/// This intrinsic corresponds to the \c BSWAP instruction.
+///
+/// \param __A
+///    A 32-bit integer operand.
+/// \returns A 32-bit integer containing the swapped bytes.
 static __inline__ int __DEFAULT_FN_ATTRS_CONSTEXPR
 _bswap(int __A) {
   return (int)__builtin_bswap32((unsigned int)__A);
@@ -173,12 +183,29 @@ __popcntq(unsigned long long __A)
 #endif /* __x86_64__ */
 
 #ifdef __x86_64__
+/// Returns the program status and control \c RFLAGS register with the \c VM
+///    and \c RF flags cleared.
+///
+/// \headerfile 
+///
+/// This intrinsic corresponds to the \c PUSHFQ + \c POP instruction sequence.
+///
+/// \returns The 64-bit value of the RFLAGS register.
 static __inline__ unsigned long long __DEFAULT_FN_ATTRS
 __readeflags(void)
 {
   return __builtin_ia32_readeflags_u64();
 }
 
+/// Writes the specified value to the program status and control \c RFLAGS
+///    register. Reserved bits are not affected.
+///
+/// \headerfile 
+///
+/// This intrinsic corresponds to the \c PUSH + \c POPFQ instruction sequence.
+///
+/// \param __f
+///    The 64-bit value to write to \c RFLAGS.
 static __inline__ void __DEFAULT_FN_ATTRS
 __writeeflags(unsigned long long __f)
 {
@@ -186,12 +213,29 @@ __writeeflags(unsigned long long __f)
 }
 
 #else /* !__x86_64__ */
+/// Returns the program status and control \c EFLAGS register with the \c VM
+///    and \c RF flags cleared.
+///
+/// \headerfile 
+///
+/// This intrinsic corresponds to the \c PUSHFD + \c POP instruction sequence.
+///
+/// \returns The 32-bit value of the EFLAGS register.
 static __inline__ unsigned int __DEFAULT_FN_ATTRS
 __readeflags(void)
 {
   return __builtin_ia32_readeflags_u32();
 }
 
+/// Writes the specified value to the program status and control \c EFLAGS
+///    register. Reserved bits are not affected.
+///
+/// \headerfile 
+///
+/// This intrinsic corresponds to the \c PUSH + \c POPFD instruction sequence.
+///
+/// \param __f
+///    The 32-bit value to write to \c EFLAGS.
 static __inline__ void __DEFAULT_FN_ATTRS
 __writeeflags(unsigned int __f)
 {
@@ -341,12 +385,32 @@ __crc32q(unsigned long long __C, unsigned long long __D)
 }
 #endif /* __x86_64__ */
 
+/// Reads the specified performance monitoring counter. Refer to your
+///    processor's documentation to determine which performance counters are
+///    supported.
+///
+/// \headerfile 
+///
+/// This intrinsic corresponds to the \c RDPMC instruction.
+///
+/// \param __A
+///    The performance counter to read.
+/// \returns The 64-bit value read from the performance counter.
 static __inline__ unsigned long long __DEFAULT_FN_ATTRS
 __rdpmc(int __A) {
   return __builtin_ia32_rdpmc(__A);
 }
 
-/* __rdtscp */
+/// Reads the processor's time stamp counter and the \c IA32_TSC_AUX MSR
+///    \c (0xc0000103).
+///
+/// \headerfile 
+///
+/// This intrinsic corresponds to the \c RDTSCP instruction.
+///
+/// \param __A
+///    Address of where to store the 32-bit \c IA32_TSC_AUX value.
+/// \returns The 64-bit value of the time stamp counter.
 static __inline__ unsigned long long __DEFAULT_FN_ATTRS
 __rdtscp(unsigned int *__A) {
   return __builtin_ia32_rdtscp(__A);
@@ -361,42 +425,146 @@ _wbinvd(void) {
   __builtin_ia32_wbinvd();
 }
 
+/// Rotates an 8-bit value to the left by the specified number of bits.
+///    This operation is undefined if the number of bits exceeds the size of
+///    the value.
+///
+/// \headerfile 
+///
+/// This intrinsic corresponds to the \c ROL instruction.
+///
+/// \param __X
+///    The unsigned 8-bit value to be rotated.
+/// \param __C
+///    The number of bits to rotate the value.
+/// \returns The rotated value.
 static __inline__ unsigned char __DEFAULT_FN_ATTRS_CONSTEXPR
 __rolb(unsigned char __X, int __C) {
   return __builtin_rotateleft8(__X, __C);
 }
 
+/// Rotates an 8-bit value to the right by the specified number of bits.
+///    This operation is undefined if the number of bits exceeds the size of
+///    the value.
+///
+/// \headerfile 
+///
+/// This intrinsic corresponds to the \c ROR instruction.
+///
+/// \param __X
+///    The unsigned 8-bit value to be rotated.
+/// \param __C
+///    The number of bits to rotate the value.
+/// \returns The rotated value.
 static __inline__ unsigned char __DEFAULT_FN_ATTRS_CONSTEXPR
 __rorb(unsigned char __X, int __C) {
   return __builtin_rotateright8(__X, __C);
 }
 
+/// Rotates a 16-bit value to the left by the specified number of bits.
+///    This operation is undefined if the number of bits exceeds the size of
+///    the value.
+///
+/// \headerfile 
+///
+/// This intrinsic corresponds to the \c ROL instruction.
+///
+/// \param __X
+///    The unsigned 16-bit value to be rotated.
+/// \param __C
+///    The number of bits to rotate the value.
+/// \returns The rotated value.
 static __inline__ unsigned short __DEFAULT_FN_ATTRS_CONSTEXPR
 __rolw(unsigned short __X, int __C) {
   return __builtin_rotateleft16(__X, __C);
 }
 
+/// Rotates a 16-bit value to the right by the specified number of bits.
+///    This operation is undefined if the number of bits exceeds the size of
+///    the value.
+///
+/// \headerfile 
+///
+/// This intrinsic corresponds to the \c ROR instruction.
+///
+/// \param __X
+///    The unsigned 16-bit value to be rotated.
+/// \param __C
+///    The number of bits to rotate the value.
+/// \returns The rotated value.
 static __inline__ unsigned short __DEFAULT_FN_ATTRS_CONSTEXPR
 __rorw(unsigned short __X, int __C) {
   return __builtin_rotateright16(__X, __C);
 }
 
+/// Rotates a 32-bit value to the left by the specified number of bits.
+///    This operation is undefined if the number of bits exceeds the size of
+///    the value.
+///
+/// \headerfile 
+///
+/// This intrinsic corresponds to the \c ROL instruction.
+///
+/// \param __X
+///    The unsigned 32-bit value to be rotated.
+/// \param __C
+///    The number of bits to rotate the value.
+/// \returns The rotated value.
 static __inline__ unsigned int __DEFAULT_FN_ATTRS_CONSTEXPR
 __rold(unsigned int __X, int __C) {
   return __builtin_rotateleft32(__X, (unsigned int)__C);
 }
 
+/// Rotates a 32-bit value to the right by the specified number of bits.
+///    This operation is undefined if the number of bits exceeds the size of
+///    the value.
+///
+/// \headerfile 
+///
+/// This intrinsic corresponds to the \c ROR instruction.
+///
+/// \param __X
+///    The unsigned 32-bit value to be rotated.
+/// \param __C
+///    The number of bits to rotate the value.
+/// \returns The rotated value.
 static __inline__ unsigned int __DEFAULT_FN_ATTRS_CONSTEXPR
 __rord(unsigned int __X, int __C) {
   return __builtin_rotateright32(__X, (unsigned int)__C);
 }
 
 #ifdef __x86_64__
+/// Rotates a 64-bit value to the left by the specified number of bits.
+///    This operation is undefined if the number of bits exceeds the size of
+///    the value.
+///
+/// \headerfile 
+///
+/// This intrinsic corresponds to the \c ROL instruction.
+///
+/// \param __X
+///    The unsigned 64-bit value to be rotated.
+/// \param __C
+///    The number of bits to rotate the value.
+/// \returns The rotated value.
 static __inline__ unsigned long long __DEFAULT_FN_ATTRS_CONSTEXPR
 __rolq(unsigned long long __X, int __C) {
   return __builtin_rotateleft64(__X, (unsigned long long)__C);
 }
 
+/// Rotates a 64-bit value to the right by the specified number of bits.
+///    This operation is undefined if the number of bits exceeds the size of
+///    the value.
+///
+/// \headerfile 
+///
+/// This intrinsic corresponds to the \c ROR instruction.
+///
+/// \param __X
+///    The unsigned 64-bit value to be rotated.
+/// \param __C
+///    The number of bits to rotate the value.
+/// \returns The rotated value.
 static __inline__ unsigned long long __DEFAULT_FN_ATTRS_CONSTEXPR
 __rorq(unsigned long long __X, int __C) {
   return __builtin_rotateright64(__X, (unsigned long long)__C);
diff --git a/clang/lib/Headers/immintrin.h b/clang/lib/Headers/immintrin.h
index 0149a1cdea633..27800f7a8202c 100644
--- a/clang/lib/Headers/immintrin.h
+++ b/clang/lib/Headers/immintrin.h
@@ -490,6 +490,15 @@ _writegsbase_u64(unsigned long long __V)
  * field inside of it.
  */
 
+/// Load a 16-bit value from memory and swap its bytes.
+///
+/// \headerfile 
+///
+/// This intrinsic corresponds to the MOVBE instruction.
+///
+/// \param __P
+///    A pointer to the 16-bit value to load.
+/// \returns The byte-swapped value.
 static __inline__ short __attribute__((__always_inline__, __nodebug__, __target__("movbe")))
 _loadbe_i16(void const * __P) {
   struct __loadu_i16 {
@@ -498,6 +507,16 @@ _loadbe_i16(void const * __P) {
   return (short)__builtin_bswap16(((const struct __loadu_i16*)__P)->__v);
 }
 
+/// Swap the bytes of a 16-bit value and store it to memory.
+///
+/// \headerfile 
+///
+/// This intrinsic corresponds to the MOVBE instruction.
+///
+/// \param __P
+///    A pointer to the memory for storing the swapped value.
+/// \param __D
+///    The 16-bit value to be byte-swapped.
 static __inline__ void __attribute__((__always_inline__, __nodebug__, __target__("movbe")))
 _storebe_i16(void * __P, short __D) {
   struct __storeu_i16 {
@@ -506,6 +525,15 @@ _storebe_i16(void * __P, short __D) {
   ((struct __storeu_i16*)__P)->__v = __builtin_bswap16((unsigned short)__D);
 }
 
+/// Load a 32-bit value from memory and swap its bytes.
+///
+/// \headerfile 
+///
+/// This intrinsic corresponds to the MOVBE instruction.
+///
+/// \param __P
+///    A pointer to the 32-bit value to load.
+/// \returns The byte-swapped value.
 static __inline__ int __attribute__((__always_inline__, __nodebug__, __target__("movbe")))
 _loadbe_i32(void const * __P) {
   struct __loadu_i32 {
@@ -514,6 +542,16 @@ _loadbe_i32(void const * __P) {
   return (int)__builtin_bswap32(((const struct __loadu_i32*)__P)->__v);
 }
 
+/// Swap the bytes of a 32-bit value and store it to memory.
+///
+/// \headerfile 
+///
+/// This intrinsic corresponds to the MOVBE instruction.
+///
+/// \param __P
+///    A pointer to the memory for storing the swapped value.
+/// \param __D
+///    The 32-bit value to be byte-swapped.
 static __inline__ void __attribute__((__always_inline__, __nodebug__, __target__("movbe")))
 _storebe_i32(void * __P, int __D) {
   struct __storeu_i32 {
@@ -523,6 +561,15 @@ _storebe_i32(void * __P, int __D) {
 }
 
 #ifdef __x86_64__
+/// Load a 64-bit value from memory and swap its bytes.
+///
+/// \headerfile 
+///
+/// This intrinsic corresponds to the MOVBE instruction.
+///
+/// \param __P
+///    A pointer to the 64-bit value to load.
+/// \returns The byte-swapped value.
 static __inline__ long long __attribute__((__always_inline__, __nodebug__, __target__("movbe")))
 _loadbe_i64(void const * __P) {
   struct __loadu_i64 {
@@ -531,6 +578,16 @@ _loadbe_i64(void const * __P) {
   return (long long)__builtin_bswap64(((const struct __loadu_i64*)__P)->__v);
 }
 
+/// Swap the bytes of a 64-bit value and store it to memory.
+///
+/// \headerfile 
+///
+/// This intrinsic corresponds to the MOVBE instruction.
+///
+/// \param __P
+///    A pointer to the memory for storing the swapped value.
+/// \param __D
+///    The 64-bit value to be byte-swapped.
 static __inline__ void __attribute__((__always_inline__, __nodebug__, __target__("movbe")))
 _storebe_i64(void * __P, long long __D) {
   struct __storeu_i64 {
diff --git a/clang/lib/Headers/llvm_libc_wrappers/string.h b/clang/lib/Headers/llvm_libc_wrappers/string.h
index b4fbf17c7e421..0ea49cb137606 100644
--- a/clang/lib/Headers/llvm_libc_wrappers/string.h
+++ b/clang/lib/Headers/llvm_libc_wrappers/string.h
@@ -51,13 +51,13 @@ char *strcpy(char *__restrict, const char *__restrict) __LIBC_ATTRS;
 size_t strcspn(const char *, const char *) __LIBC_ATTRS;
 char *strdup(const char *) __LIBC_ATTRS;
 size_t strlen(const char *) __LIBC_ATTRS;
-char *strncat(char *, const char *, size_t) __LIBC_ATTRS;
+char *strncat(char *__restrict, const char *__restrict, size_t) __LIBC_ATTRS;
 int strncmp(const char *, const char *, size_t) __LIBC_ATTRS;
 char *strncpy(char *__restrict, const char *__restrict, size_t) __LIBC_ATTRS;
 char *strndup(const char *, size_t) __LIBC_ATTRS;
 size_t strnlen(const char *, size_t) __LIBC_ATTRS;
 size_t strspn(const char *, const char *) __LIBC_ATTRS;
-char *strtok(char *__restrict, const char *) __LIBC_ATTRS;
+char *strtok(char *__restrict, const char *__restrict) __LIBC_ATTRS;
 char *strtok_r(char *__restrict, const char *__restrict,
                char **__restrict) __LIBC_ATTRS;
 size_t strxfrm(char *__restrict, const char *__restrict, size_t) __LIBC_ATTRS;
diff --git a/clang/lib/Interpreter/Interpreter.cpp b/clang/lib/Interpreter/Interpreter.cpp
index 734fe90d0d89b..d1764d07dfd24 100644
--- a/clang/lib/Interpreter/Interpreter.cpp
+++ b/clang/lib/Interpreter/Interpreter.cpp
@@ -249,7 +249,7 @@ Interpreter::~Interpreter() {
 // can't find the precise resource directory in unittests so we have to hard
 // code them.
 const char *const Runtimes = R"(
-    void* operator new(__SIZE_TYPE__, void* __p) noexcept;
+#ifdef __cplusplus
     void *__clang_Interpreter_SetValueWithAlloc(void*, void*, void*);
     void __clang_Interpreter_SetValueNoAlloc(void*, void*, void*);
     void __clang_Interpreter_SetValueNoAlloc(void*, void*, void*, void*);
@@ -257,15 +257,18 @@ const char *const Runtimes = R"(
     void __clang_Interpreter_SetValueNoAlloc(void*, void*, void*, double);
     void __clang_Interpreter_SetValueNoAlloc(void*, void*, void*, long double);
     void __clang_Interpreter_SetValueNoAlloc(void*,void*,void*,unsigned long long);
+    struct __clang_Interpreter_NewTag{} __ci_newtag;
+    void* operator new(__SIZE_TYPE__, void* __p, __clang_Interpreter_NewTag) noexcept;
     template 
     void __clang_Interpreter_SetValueCopyArr(T* Src, void* Placement, unsigned long Size) {
       for (auto Idx = 0; Idx < Size; ++Idx)
-        new ((void*)(((T*)Placement) + Idx)) T(Src[Idx]);
+        new ((void*)(((T*)Placement) + Idx), __ci_newtag) T(Src[Idx]);
     }
     template 
     void __clang_Interpreter_SetValueCopyArr(const T (*Src)[N], void* Placement, unsigned long Size) {
       __clang_Interpreter_SetValueCopyArr(Src[0], Placement, Size);
     }
+#endif // __cplusplus
 )";
 
 llvm::Expected>
@@ -280,7 +283,7 @@ Interpreter::create(std::unique_ptr CI) {
   if (!PTU)
     return PTU.takeError();
 
-  Interp->ValuePrintingInfo.resize(3);
+  Interp->ValuePrintingInfo.resize(4);
   // FIXME: This is a ugly hack. Undo command checks its availability by looking
   // at the size of the PTU list. However we have parsed something in the
   // beginning of the REPL so we have to mark them as 'Irrevocable'.
@@ -501,7 +504,7 @@ Interpreter::CompileDtorCall(CXXRecordDecl *CXXRD) {
 static constexpr llvm::StringRef MagicRuntimeInterface[] = {
     "__clang_Interpreter_SetValueNoAlloc",
     "__clang_Interpreter_SetValueWithAlloc",
-    "__clang_Interpreter_SetValueCopyArr"};
+    "__clang_Interpreter_SetValueCopyArr", "__ci_newtag"};
 
 bool Interpreter::FindRuntimeInterface() {
   if (llvm::all_of(ValuePrintingInfo, [](Expr *E) { return E != nullptr; }))
@@ -531,6 +534,9 @@ bool Interpreter::FindRuntimeInterface() {
   if (!LookupInterface(ValuePrintingInfo[CopyArray],
                        MagicRuntimeInterface[CopyArray]))
     return false;
+  if (!LookupInterface(ValuePrintingInfo[NewTag],
+                       MagicRuntimeInterface[NewTag]))
+    return false;
   return true;
 }
 
@@ -608,7 +614,9 @@ class RuntimeInterfaceBuilder
                 .getValuePrintingInfo()[Interpreter::InterfaceKind::CopyArray],
             SourceLocation(), Args, SourceLocation());
       }
-      Expr *Args[] = {AllocCall.get()};
+      Expr *Args[] = {
+          AllocCall.get(),
+          Interp.getValuePrintingInfo()[Interpreter::InterfaceKind::NewTag]};
       ExprResult CXXNewCall = S.BuildCXXNew(
           E->getSourceRange(),
           /*UseGlobal=*/true, /*PlacementLParen=*/SourceLocation(), Args,
@@ -629,8 +637,9 @@ class RuntimeInterfaceBuilder
           Interp.getValuePrintingInfo()[Interpreter::InterfaceKind::NoAlloc],
           E->getBeginLoc(), Args, E->getEndLoc());
     }
+    default:
+      llvm_unreachable("Unhandled Interpreter::InterfaceKind");
     }
-    llvm_unreachable("Unhandled Interpreter::InterfaceKind");
   }
 
   Interpreter::InterfaceKind VisitRecordType(const RecordType *Ty) {
@@ -815,3 +824,15 @@ __clang_Interpreter_SetValueNoAlloc(void *This, void *OutVal, void *OpaqueType,
   VRef = Value(static_cast(This), OpaqueType);
   VRef.setLongDouble(Val);
 }
+
+// A trampoline to work around the fact that operator placement new cannot
+// really be forward declared due to libc++ and libstdc++ declaration mismatch.
+// FIXME: __clang_Interpreter_NewTag is ODR violation because we get the same
+// definition in the interpreter runtime. We should move it in a runtime header
+// which gets included by the interpreter and here.
+struct __clang_Interpreter_NewTag {};
+REPL_EXTERNAL_VISIBILITY void *
+operator new(size_t __sz, void *__p, __clang_Interpreter_NewTag) noexcept {
+  // Just forward to the standard operator placement new.
+  return operator new(__sz, __p);
+}
diff --git a/clang/lib/Lex/UnicodeCharSets.h b/clang/lib/Lex/UnicodeCharSets.h
index 5316d2540b76c..b63908024e5a1 100644
--- a/clang/lib/Lex/UnicodeCharSets.h
+++ b/clang/lib/Lex/UnicodeCharSets.h
@@ -10,7 +10,7 @@
 
 #include "llvm/Support/UnicodeCharRanges.h"
 
-// Unicode 15.0 XID_Start
+// Unicode 15.1 XID_Start
 static const llvm::sys::UnicodeCharRange XIDStartRanges[] = {
     {0x0041, 0x005A},   {0x0061, 0x007A},   {0x00AA, 0x00AA},
     {0x00B5, 0x00B5},   {0x00BA, 0x00BA},   {0x00C0, 0x00D6},
@@ -233,9 +233,10 @@ static const llvm::sys::UnicodeCharRange XIDStartRanges[] = {
     {0x1EE8B, 0x1EE9B}, {0x1EEA1, 0x1EEA3}, {0x1EEA5, 0x1EEA9},
     {0x1EEAB, 0x1EEBB}, {0x20000, 0x2A6DF}, {0x2A700, 0x2B739},
     {0x2B740, 0x2B81D}, {0x2B820, 0x2CEA1}, {0x2CEB0, 0x2EBE0},
-    {0x2F800, 0x2FA1D}, {0x30000, 0x3134A}, {0x31350, 0x323AF}};
+    {0x2EBF0, 0x2EE5D}, {0x2F800, 0x2FA1D}, {0x30000, 0x3134A},
+    {0x31350, 0x323AF}};
 
-// Unicode 15.0 XID_Continue, excluding XID_Start
+// Unicode 15.1 XID_Continue, excluding XID_Start
 // The Unicode Property XID_Continue is a super set of XID_Start.
 // To save Space, the table below only contains the codepoints
 // that are not also in XID_Start.
@@ -299,24 +300,25 @@ static const llvm::sys::UnicodeCharRange XIDContinueRanges[] = {
     {0x1C24, 0x1C37},   {0x1C40, 0x1C49},   {0x1C50, 0x1C59},
     {0x1CD0, 0x1CD2},   {0x1CD4, 0x1CE8},   {0x1CED, 0x1CED},
     {0x1CF4, 0x1CF4},   {0x1CF7, 0x1CF9},   {0x1DC0, 0x1DFF},
-    {0x203F, 0x2040},   {0x2054, 0x2054},   {0x20D0, 0x20DC},
-    {0x20E1, 0x20E1},   {0x20E5, 0x20F0},   {0x2CEF, 0x2CF1},
-    {0x2D7F, 0x2D7F},   {0x2DE0, 0x2DFF},   {0x302A, 0x302F},
-    {0x3099, 0x309A},   {0xA620, 0xA629},   {0xA66F, 0xA66F},
-    {0xA674, 0xA67D},   {0xA69E, 0xA69F},   {0xA6F0, 0xA6F1},
-    {0xA802, 0xA802},   {0xA806, 0xA806},   {0xA80B, 0xA80B},
-    {0xA823, 0xA827},   {0xA82C, 0xA82C},   {0xA880, 0xA881},
-    {0xA8B4, 0xA8C5},   {0xA8D0, 0xA8D9},   {0xA8E0, 0xA8F1},
-    {0xA8FF, 0xA909},   {0xA926, 0xA92D},   {0xA947, 0xA953},
-    {0xA980, 0xA983},   {0xA9B3, 0xA9C0},   {0xA9D0, 0xA9D9},
-    {0xA9E5, 0xA9E5},   {0xA9F0, 0xA9F9},   {0xAA29, 0xAA36},
-    {0xAA43, 0xAA43},   {0xAA4C, 0xAA4D},   {0xAA50, 0xAA59},
-    {0xAA7B, 0xAA7D},   {0xAAB0, 0xAAB0},   {0xAAB2, 0xAAB4},
-    {0xAAB7, 0xAAB8},   {0xAABE, 0xAABF},   {0xAAC1, 0xAAC1},
-    {0xAAEB, 0xAAEF},   {0xAAF5, 0xAAF6},   {0xABE3, 0xABEA},
-    {0xABEC, 0xABED},   {0xABF0, 0xABF9},   {0xFB1E, 0xFB1E},
-    {0xFE00, 0xFE0F},   {0xFE20, 0xFE2F},   {0xFE33, 0xFE34},
-    {0xFE4D, 0xFE4F},   {0xFF10, 0xFF19},   {0xFF3F, 0xFF3F},
+    {0x200C, 0x200D},   {0x203F, 0x2040},   {0x2054, 0x2054},
+    {0x20D0, 0x20DC},   {0x20E1, 0x20E1},   {0x20E5, 0x20F0},
+    {0x2CEF, 0x2CF1},   {0x2D7F, 0x2D7F},   {0x2DE0, 0x2DFF},
+    {0x302A, 0x302F},   {0x3099, 0x309A},   {0x30FB, 0x30FB},
+    {0xA620, 0xA629},   {0xA66F, 0xA66F},   {0xA674, 0xA67D},
+    {0xA69E, 0xA69F},   {0xA6F0, 0xA6F1},   {0xA802, 0xA802},
+    {0xA806, 0xA806},   {0xA80B, 0xA80B},   {0xA823, 0xA827},
+    {0xA82C, 0xA82C},   {0xA880, 0xA881},   {0xA8B4, 0xA8C5},
+    {0xA8D0, 0xA8D9},   {0xA8E0, 0xA8F1},   {0xA8FF, 0xA909},
+    {0xA926, 0xA92D},   {0xA947, 0xA953},   {0xA980, 0xA983},
+    {0xA9B3, 0xA9C0},   {0xA9D0, 0xA9D9},   {0xA9E5, 0xA9E5},
+    {0xA9F0, 0xA9F9},   {0xAA29, 0xAA36},   {0xAA43, 0xAA43},
+    {0xAA4C, 0xAA4D},   {0xAA50, 0xAA59},   {0xAA7B, 0xAA7D},
+    {0xAAB0, 0xAAB0},   {0xAAB2, 0xAAB4},   {0xAAB7, 0xAAB8},
+    {0xAABE, 0xAABF},   {0xAAC1, 0xAAC1},   {0xAAEB, 0xAAEF},
+    {0xAAF5, 0xAAF6},   {0xABE3, 0xABEA},   {0xABEC, 0xABED},
+    {0xABF0, 0xABF9},   {0xFB1E, 0xFB1E},   {0xFE00, 0xFE0F},
+    {0xFE20, 0xFE2F},   {0xFE33, 0xFE34},   {0xFE4D, 0xFE4F},
+    {0xFF10, 0xFF19},   {0xFF3F, 0xFF3F},   {0xFF65, 0xFF65},
     {0xFF9E, 0xFF9F},   {0x101FD, 0x101FD}, {0x102E0, 0x102E0},
     {0x10376, 0x1037A}, {0x104A0, 0x104A9}, {0x10A01, 0x10A03},
     {0x10A05, 0x10A06}, {0x10A0C, 0x10A0F}, {0x10A38, 0x10A3A},
@@ -364,7 +366,8 @@ static const llvm::sys::UnicodeCharRange XIDContinueRanges[] = {
     {0x1E026, 0x1E02A}, {0x1E08F, 0x1E08F}, {0x1E130, 0x1E136},
     {0x1E140, 0x1E149}, {0x1E2AE, 0x1E2AE}, {0x1E2EC, 0x1E2F9},
     {0x1E4EC, 0x1E4F9}, {0x1E8D0, 0x1E8D6}, {0x1E944, 0x1E94A},
-    {0x1E950, 0x1E959}, {0x1FBF0, 0x1FBF9}, {0xE0100, 0xE01EF}};
+    {0x1E950, 0x1E959}, {0x1FBF0, 0x1FBF9}, {0xE0100, 0xE01EF},
+};
 
 // Clang supports the "Mathematical notation profile" as an extension,
 // as described in https://www.unicode.org/L2/L2022/22230-math-profile.pdf
diff --git a/clang/lib/Parse/ParseDecl.cpp b/clang/lib/Parse/ParseDecl.cpp
index 3b505528bd1fc..e13b526e7cdd5 100644
--- a/clang/lib/Parse/ParseDecl.cpp
+++ b/clang/lib/Parse/ParseDecl.cpp
@@ -79,7 +79,7 @@ TypeResult Parser::ParseTypeName(SourceRange *Range, DeclaratorContext Context,
   if (DeclaratorInfo.isInvalidType())
     return true;
 
-  return Actions.ActOnTypeName(getCurScope(), DeclaratorInfo);
+  return Actions.ActOnTypeName(DeclaratorInfo);
 }
 
 /// Normalizes an attribute name by dropping prefixed and suffixed __.
@@ -5012,7 +5012,7 @@ void Parser::ParseEnumSpecifier(SourceLocation StartLoc, DeclSpec &DS,
                                   DeclSpecContext::DSC_type_specifier);
       Declarator DeclaratorInfo(DS, ParsedAttributesView::none(),
                                 DeclaratorContext::TypeName);
-      BaseType = Actions.ActOnTypeName(getCurScope(), DeclaratorInfo);
+      BaseType = Actions.ActOnTypeName(DeclaratorInfo);
 
       BaseRange = SourceRange(ColonLoc, DeclaratorInfo.getSourceRange().getEnd());
 
@@ -6792,7 +6792,13 @@ void Parser::ParseDirectDeclarator(Declarator &D) {
     } else if (Tok.isRegularKeywordAttribute()) {
       // For consistency with attribute parsing.
       Diag(Tok, diag::err_keyword_not_allowed) << Tok.getIdentifierInfo();
+      bool TakesArgs = doesKeywordAttributeTakeArgs(Tok.getKind());
       ConsumeToken();
+      if (TakesArgs) {
+        BalancedDelimiterTracker T(*this, tok::l_paren);
+        if (!T.consumeOpen())
+          T.skipToEnd();
+      }
     } else if (Tok.is(tok::kw_requires) && D.hasGroupingParens()) {
       // This declarator is declaring a function, but the requires clause is
       // in the wrong place:
@@ -8046,6 +8052,71 @@ bool Parser::TryAltiVecTokenOutOfLine(DeclSpec &DS, SourceLocation Loc,
   return false;
 }
 
+TypeResult Parser::ParseTypeFromString(StringRef TypeStr, StringRef Context,
+                                       SourceLocation IncludeLoc) {
+  // Consume (unexpanded) tokens up to the end-of-directive.
+  SmallVector Tokens;
+  {
+    // Create a new buffer from which we will parse the type.
+    auto &SourceMgr = PP.getSourceManager();
+    FileID FID = SourceMgr.createFileID(
+        llvm::MemoryBuffer::getMemBufferCopy(TypeStr, Context), SrcMgr::C_User,
+        0, 0, IncludeLoc);
+
+    // Form a new lexer that references the buffer.
+    Lexer L(FID, SourceMgr.getBufferOrFake(FID), PP);
+    L.setParsingPreprocessorDirective(true);
+
+    // Lex the tokens from that buffer.
+    Token Tok;
+    do {
+      L.Lex(Tok);
+      Tokens.push_back(Tok);
+    } while (Tok.isNot(tok::eod));
+  }
+
+  // Replace the "eod" token with an "eof" token identifying the end of
+  // the provided string.
+  Token &EndToken = Tokens.back();
+  EndToken.startToken();
+  EndToken.setKind(tok::eof);
+  EndToken.setLocation(Tok.getLocation());
+  EndToken.setEofData(TypeStr.data());
+
+  // Add the current token back.
+  Tokens.push_back(Tok);
+
+  // Enter the tokens into the token stream.
+  PP.EnterTokenStream(Tokens, /*DisableMacroExpansion=*/false,
+                      /*IsReinject=*/false);
+
+  // Consume the current token so that we'll start parsing the tokens we
+  // added to the stream.
+  ConsumeAnyToken();
+
+  // Enter a new scope.
+  ParseScope LocalScope(this, 0);
+
+  // Parse the type.
+  TypeResult Result = ParseTypeName(nullptr);
+
+  // Check if we parsed the whole thing.
+  if (Result.isUsable() &&
+      (Tok.isNot(tok::eof) || Tok.getEofData() != TypeStr.data())) {
+    Diag(Tok.getLocation(), diag::err_type_unparsed);
+  }
+
+  // There could be leftover tokens (e.g. because of an error).
+  // Skip through until we reach the 'end of directive' token.
+  while (Tok.isNot(tok::eof))
+    ConsumeAnyToken();
+
+  // Consume the end token.
+  if (Tok.is(tok::eof) && Tok.getEofData() == TypeStr.data())
+    ConsumeAnyToken();
+  return Result;
+}
+
 void Parser::DiagnoseBitIntUse(const Token &Tok) {
   // If the token is for _ExtInt, diagnose it as being deprecated. Otherwise,
   // the token is about _BitInt and gets (potentially) diagnosed as use of an
diff --git a/clang/lib/Parse/ParseDeclCXX.cpp b/clang/lib/Parse/ParseDeclCXX.cpp
index d97081da4200d..c0d771dc93dae 100644
--- a/clang/lib/Parse/ParseDeclCXX.cpp
+++ b/clang/lib/Parse/ParseDeclCXX.cpp
@@ -1290,7 +1290,7 @@ TypeResult Parser::ParseBaseTypeSpecifier(SourceLocation &BaseLoc,
 
     Declarator DeclaratorInfo(DS, ParsedAttributesView::none(),
                               DeclaratorContext::TypeName);
-    return Actions.ActOnTypeName(getCurScope(), DeclaratorInfo);
+    return Actions.ActOnTypeName(DeclaratorInfo);
   }
 
   // Check whether we have a template-id that names a type.
@@ -1385,7 +1385,7 @@ TypeResult Parser::ParseBaseTypeSpecifier(SourceLocation &BaseLoc,
 
   Declarator DeclaratorInfo(DS, ParsedAttributesView::none(),
                             DeclaratorContext::TypeName);
-  return Actions.ActOnTypeName(getCurScope(), DeclaratorInfo);
+  return Actions.ActOnTypeName(DeclaratorInfo);
 }
 
 void Parser::ParseMicrosoftInheritanceClassAttributes(ParsedAttributes &attrs) {
@@ -1890,7 +1890,13 @@ void Parser::ParseClassSpecifier(tok::TokenKind TagTokKind,
         if (!SkipUntil(tok::r_paren, StopAtSemi))
           break;
       } else if (Tok.isRegularKeywordAttribute()) {
+        bool TakesArgs = doesKeywordAttributeTakeArgs(Tok.getKind());
         ConsumeToken();
+        if (TakesArgs) {
+          BalancedDelimiterTracker T(*this, tok::l_paren);
+          if (!T.consumeOpen())
+            T.skipToEnd();
+        }
       } else {
         break;
       }
@@ -4539,8 +4545,18 @@ void Parser::ParseCXX11AttributeSpecifierInternal(ParsedAttributes &Attrs,
   if (Tok.isRegularKeywordAttribute()) {
     SourceLocation Loc = Tok.getLocation();
     IdentifierInfo *AttrName = Tok.getIdentifierInfo();
-    Attrs.addNew(AttrName, Loc, nullptr, Loc, nullptr, 0, Tok.getKind());
+    ParsedAttr::Form Form = ParsedAttr::Form(Tok.getKind());
+    bool TakesArgs = doesKeywordAttributeTakeArgs(Tok.getKind());
     ConsumeToken();
+    if (TakesArgs) {
+      if (!Tok.is(tok::l_paren))
+        Diag(Tok.getLocation(), diag::err_expected_lparen_after) << AttrName;
+      else
+        ParseAttributeArgsCommon(AttrName, Loc, Attrs, EndLoc,
+                                 /*ScopeName*/ nullptr,
+                                 /*ScopeLoc*/ Loc, Form);
+    } else
+      Attrs.addNew(AttrName, Loc, nullptr, Loc, nullptr, 0, Form);
     return;
   }
 
@@ -4706,11 +4722,13 @@ SourceLocation Parser::SkipCXX11Attributes() {
       T.consumeOpen();
       T.skipToEnd();
       EndLoc = T.getCloseLocation();
-    } else if (Tok.isRegularKeywordAttribute()) {
+    } else if (Tok.isRegularKeywordAttribute() &&
+               !doesKeywordAttributeTakeArgs(Tok.getKind())) {
       EndLoc = Tok.getLocation();
       ConsumeToken();
     } else {
-      assert(Tok.is(tok::kw_alignas) && "not an attribute specifier");
+      assert((Tok.is(tok::kw_alignas) || Tok.isRegularKeywordAttribute()) &&
+             "not an attribute specifier");
       ConsumeToken();
       BalancedDelimiterTracker T(*this, tok::l_paren);
       if (!T.consumeOpen())
diff --git a/clang/lib/Parse/ParseExpr.cpp b/clang/lib/Parse/ParseExpr.cpp
index 72259d8d31f31..fee02dd813d5c 100644
--- a/clang/lib/Parse/ParseExpr.cpp
+++ b/clang/lib/Parse/ParseExpr.cpp
@@ -1243,8 +1243,7 @@ ExprResult Parser::ParseCastExpression(CastParseKind ParseKind,
 
             Declarator DeclaratorInfo(DS, ParsedAttributesView::none(),
                                       DeclaratorContext::TypeName);
-            TypeResult Ty = Actions.ActOnTypeName(getCurScope(),
-                                                  DeclaratorInfo);
+            TypeResult Ty = Actions.ActOnTypeName(DeclaratorInfo);
             if (Ty.isInvalid())
               break;
 
@@ -1541,7 +1540,7 @@ ExprResult Parser::ParseCastExpression(CastParseKind ParseKind,
 
       Declarator DeclaratorInfo(DS, ParsedAttributesView::none(),
                                 DeclaratorContext::TypeName);
-      TypeResult Ty = Actions.ActOnTypeName(getCurScope(), DeclaratorInfo);
+      TypeResult Ty = Actions.ActOnTypeName(DeclaratorInfo);
       if (Ty.isInvalid())
         break;
 
@@ -3196,7 +3195,7 @@ Parser::ParseParenExpression(ParenParseOption &ExprType, bool stopIfCastExpr,
       TypeResult Ty;
       {
         InMessageExpressionRAIIObject InMessage(*this, false);
-        Ty = Actions.ActOnTypeName(getCurScope(), DeclaratorInfo);
+        Ty = Actions.ActOnTypeName(DeclaratorInfo);
       }
       Result = ParseObjCMessageExpressionBody(SourceLocation(),
                                               SourceLocation(),
@@ -3211,7 +3210,7 @@ Parser::ParseParenExpression(ParenParseOption &ExprType, bool stopIfCastExpr,
         TypeResult Ty;
         {
           InMessageExpressionRAIIObject InMessage(*this, false);
-          Ty = Actions.ActOnTypeName(getCurScope(), DeclaratorInfo);
+          Ty = Actions.ActOnTypeName(DeclaratorInfo);
         }
         return ParseCompoundLiteralExpression(Ty.get(), OpenLoc, RParenLoc);
       }
@@ -3223,7 +3222,7 @@ Parser::ParseParenExpression(ParenParseOption &ExprType, bool stopIfCastExpr,
           TypeResult Ty;
           {
             InMessageExpressionRAIIObject InMessage(*this, false);
-            Ty = Actions.ActOnTypeName(getCurScope(), DeclaratorInfo);
+            Ty = Actions.ActOnTypeName(DeclaratorInfo);
           }
           if(Ty.isInvalid())
           {
@@ -3270,7 +3269,7 @@ Parser::ParseParenExpression(ParenParseOption &ExprType, bool stopIfCastExpr,
           TypeResult Ty;
           {
             InMessageExpressionRAIIObject InMessage(*this, false);
-            Ty = Actions.ActOnTypeName(getCurScope(), DeclaratorInfo);
+            Ty = Actions.ActOnTypeName(DeclaratorInfo);
           }
           CastTy = Ty.get();
           return ExprResult();
diff --git a/clang/lib/Parse/ParseExprCXX.cpp b/clang/lib/Parse/ParseExprCXX.cpp
index ef9ea6575205c..d61f414406f02 100644
--- a/clang/lib/Parse/ParseExprCXX.cpp
+++ b/clang/lib/Parse/ParseExprCXX.cpp
@@ -1916,7 +1916,7 @@ ExprResult
 Parser::ParseCXXTypeConstructExpression(const DeclSpec &DS) {
   Declarator DeclaratorInfo(DS, ParsedAttributesView::none(),
                             DeclaratorContext::FunctionalCast);
-  ParsedType TypeRep = Actions.ActOnTypeName(getCurScope(), DeclaratorInfo).get();
+  ParsedType TypeRep = Actions.ActOnTypeName(DeclaratorInfo).get();
 
   assert((Tok.is(tok::l_paren) ||
           (getLangOpts().CPlusPlus11 && Tok.is(tok::l_brace)))
@@ -2815,7 +2815,7 @@ bool Parser::ParseUnqualifiedIdOperator(CXXScopeSpec &SS, bool EnteringContext,
   ParseDeclaratorInternal(D, /*DirectDeclParser=*/nullptr);
 
   // Finish up the type.
-  TypeResult Ty = Actions.ActOnTypeName(getCurScope(), D);
+  TypeResult Ty = Actions.ActOnTypeName(D);
   if (Ty.isInvalid())
     return true;
 
@@ -3252,8 +3252,7 @@ Parser::ParseCXXNewExpression(bool UseGlobal, SourceLocation Start) {
     ConstructorLParen = T.getOpenLocation();
     if (Tok.isNot(tok::r_paren)) {
       auto RunSignatureHelp = [&]() {
-        ParsedType TypeRep =
-            Actions.ActOnTypeName(getCurScope(), DeclaratorInfo).get();
+        ParsedType TypeRep = Actions.ActOnTypeName(DeclaratorInfo).get();
         QualType PreferredType;
         // ActOnTypeName might adjust DeclaratorInfo and return a null type even
         // the passing DeclaratorInfo is valid, e.g. running SignatureHelp on
@@ -4037,7 +4036,7 @@ Parser::ParseCXXAmbiguousParenExpression(ParenParseOption &ExprType,
       if (DeclaratorInfo.isInvalidType())
         return ExprError();
 
-      TypeResult Ty = Actions.ActOnTypeName(getCurScope(), DeclaratorInfo);
+      TypeResult Ty = Actions.ActOnTypeName(DeclaratorInfo);
       return ParseCompoundLiteralExpression(Ty.get(),
                                             Tracker.getOpenLocation(),
                                             Tracker.getCloseLocation());
diff --git a/clang/lib/Parse/ParseObjc.cpp b/clang/lib/Parse/ParseObjc.cpp
index c0261c462b883..849fd1ac95a44 100644
--- a/clang/lib/Parse/ParseObjc.cpp
+++ b/clang/lib/Parse/ParseObjc.cpp
@@ -1332,7 +1332,7 @@ ParsedType Parser::ParseObjCTypeName(ObjCDeclSpec &DS,
                                            DS.getNullabilityLoc(),
                                            addedToDeclSpec);
 
-      TypeResult type = Actions.ActOnTypeName(getCurScope(), declarator);
+      TypeResult type = Actions.ActOnTypeName(declarator);
       if (!type.isInvalid())
         Ty = type.get();
 
@@ -1747,7 +1747,7 @@ void Parser::parseObjCTypeArgsOrProtocolQualifiers(
       // Form a declarator to turn this into a type.
       Declarator D(DS, ParsedAttributesView::none(),
                    DeclaratorContext::TypeName);
-      TypeResult fullTypeArg = Actions.ActOnTypeName(getCurScope(), D);
+      TypeResult fullTypeArg = Actions.ActOnTypeName(D);
       if (fullTypeArg.isUsable()) {
         typeArgs.push_back(fullTypeArg.get());
         if (!foundValidTypeId) {
@@ -3024,7 +3024,7 @@ bool Parser::ParseObjCXXMessageReceiver(bool &IsExpr, void *&TypeOrExpr) {
   // remainder of the class message.
   Declarator DeclaratorInfo(DS, ParsedAttributesView::none(),
                             DeclaratorContext::TypeName);
-  TypeResult Type = Actions.ActOnTypeName(getCurScope(), DeclaratorInfo);
+  TypeResult Type = Actions.ActOnTypeName(DeclaratorInfo);
   if (Type.isInvalid())
     return true;
 
diff --git a/clang/lib/Parse/ParseOpenACC.cpp b/clang/lib/Parse/ParseOpenACC.cpp
index fc82324e235d7..1f6d111f3aa57 100644
--- a/clang/lib/Parse/ParseOpenACC.cpp
+++ b/clang/lib/Parse/ParseOpenACC.cpp
@@ -89,15 +89,35 @@ OpenACCClauseKind getOpenACCClauseKind(Token Tok) {
 
   return llvm::StringSwitch(
              Tok.getIdentifierInfo()->getName())
+      .Case("attach", OpenACCClauseKind::Attach)
       .Case("auto", OpenACCClauseKind::Auto)
+      .Case("bind", OpenACCClauseKind::Bind)
+      .Case("create", OpenACCClauseKind::Create)
+      .Case("collapse", OpenACCClauseKind::Collapse)
+      .Case("copy", OpenACCClauseKind::Copy)
+      .Case("copyin", OpenACCClauseKind::CopyIn)
+      .Case("copyout", OpenACCClauseKind::CopyOut)
       .Case("default", OpenACCClauseKind::Default)
+      .Case("delete", OpenACCClauseKind::Delete)
+      .Case("detach", OpenACCClauseKind::Detach)
+      .Case("device", OpenACCClauseKind::Device)
+      .Case("device_resident", OpenACCClauseKind::DeviceResident)
+      .Case("deviceptr", OpenACCClauseKind::DevicePtr)
       .Case("finalize", OpenACCClauseKind::Finalize)
+      .Case("firstprivate", OpenACCClauseKind::FirstPrivate)
+      .Case("host", OpenACCClauseKind::Host)
       .Case("if", OpenACCClauseKind::If)
       .Case("if_present", OpenACCClauseKind::IfPresent)
       .Case("independent", OpenACCClauseKind::Independent)
+      .Case("link", OpenACCClauseKind::Link)
+      .Case("no_create", OpenACCClauseKind::NoCreate)
       .Case("nohost", OpenACCClauseKind::NoHost)
+      .Case("present", OpenACCClauseKind::Present)
+      .Case("private", OpenACCClauseKind::Private)
+      .Case("reduction", OpenACCClauseKind::Reduction)
       .Case("self", OpenACCClauseKind::Self)
       .Case("seq", OpenACCClauseKind::Seq)
+      .Case("use_device", OpenACCClauseKind::UseDevice)
       .Case("vector", OpenACCClauseKind::Vector)
       .Case("worker", OpenACCClauseKind::Worker)
       .Default(OpenACCClauseKind::Invalid);
@@ -132,6 +152,8 @@ enum class OpenACCSpecialTokenKind {
   ReadOnly,
   DevNum,
   Queues,
+  Zero,
+  Force,
 };
 
 bool isOpenACCSpecialToken(OpenACCSpecialTokenKind Kind, Token Tok) {
@@ -145,10 +167,56 @@ bool isOpenACCSpecialToken(OpenACCSpecialTokenKind Kind, Token Tok) {
     return Tok.getIdentifierInfo()->isStr("devnum");
   case OpenACCSpecialTokenKind::Queues:
     return Tok.getIdentifierInfo()->isStr("queues");
+  case OpenACCSpecialTokenKind::Zero:
+    return Tok.getIdentifierInfo()->isStr("zero");
+  case OpenACCSpecialTokenKind::Force:
+    return Tok.getIdentifierInfo()->isStr("force");
   }
   llvm_unreachable("Unknown 'Kind' Passed");
 }
 
+/// Used for cases where we have a token we want to check against an
+/// 'identifier-like' token, but don't want to give awkward error messages in
+/// cases where it is accidentially a keyword.
+bool isTokenIdentifierOrKeyword(Parser &P, Token Tok) {
+  if (Tok.is(tok::identifier))
+    return true;
+
+  if (!Tok.isAnnotation() && Tok.getIdentifierInfo() &&
+      Tok.getIdentifierInfo()->isKeyword(P.getLangOpts()))
+    return true;
+
+  return false;
+}
+
+/// Parses and consumes an identifer followed immediately by a single colon, and
+/// diagnoses if it is not the 'special token' kind that we require. Used when
+/// the tag is the only valid value.
+/// Return 'true' if the special token was matched, false if no special token,
+/// or an invalid special token was found.
+template 
+bool tryParseAndConsumeSpecialTokenKind(Parser &P, OpenACCSpecialTokenKind Kind,
+                                        DirOrClauseTy DirOrClause) {
+  Token IdentTok = P.getCurToken();
+  // If this is an identifier-like thing followed by ':', it is one of the
+  // OpenACC 'special' name tags, so consume it.
+  if (isTokenIdentifierOrKeyword(P, IdentTok) && P.NextToken().is(tok::colon)) {
+    P.ConsumeToken();
+    P.ConsumeToken();
+
+    if (!isOpenACCSpecialToken(Kind, IdentTok)) {
+      P.Diag(IdentTok, diag::err_acc_invalid_tag_kind)
+          << IdentTok.getIdentifierInfo() << DirOrClause
+          << std::is_same_v;
+      return false;
+    }
+
+    return true;
+  }
+
+  return false;
+}
+
 bool isOpenACCDirectiveKind(OpenACCDirectiveKind Kind, Token Tok) {
   if (!Tok.is(tok::identifier))
     return false;
@@ -198,16 +266,53 @@ bool isOpenACCDirectiveKind(OpenACCDirectiveKind Kind, Token Tok) {
   llvm_unreachable("Unknown 'Kind' Passed");
 }
 
+OpenACCReductionOperator ParseReductionOperator(Parser &P) {
+  // If there is no colon, treat as if the reduction operator was missing, else
+  // we probably will not recover from it in the case where an expression starts
+  // with one of the operator tokens.
+  if (P.NextToken().isNot(tok::colon)) {
+    P.Diag(P.getCurToken(), diag::err_acc_expected_reduction_operator);
+    return OpenACCReductionOperator::Invalid;
+  }
+  Token ReductionKindTok = P.getCurToken();
+  // Consume both the kind and the colon.
+  P.ConsumeToken();
+  P.ConsumeToken();
+
+  switch (ReductionKindTok.getKind()) {
+  case tok::plus:
+    return OpenACCReductionOperator::Addition;
+  case tok::star:
+    return OpenACCReductionOperator::Multiplication;
+  case tok::amp:
+    return OpenACCReductionOperator::BitwiseAnd;
+  case tok::pipe:
+    return OpenACCReductionOperator::BitwiseOr;
+  case tok::caret:
+    return OpenACCReductionOperator::BitwiseXOr;
+  case tok::ampamp:
+    return OpenACCReductionOperator::And;
+  case tok::pipepipe:
+    return OpenACCReductionOperator::Or;
+  case tok::identifier:
+    if (ReductionKindTok.getIdentifierInfo()->isStr("max"))
+      return OpenACCReductionOperator::Max;
+    if (ReductionKindTok.getIdentifierInfo()->isStr("min"))
+      return OpenACCReductionOperator::Min;
+    LLVM_FALLTHROUGH;
+  default:
+    P.Diag(ReductionKindTok, diag::err_acc_invalid_reduction_operator);
+    return OpenACCReductionOperator::Invalid;
+  }
+  llvm_unreachable("Reduction op token kind not caught by 'default'?");
+}
+
 /// Used for cases where we expect an identifier-like token, but don't want to
 /// give awkward error messages in cases where it is accidentially a keyword.
 bool expectIdentifierOrKeyword(Parser &P) {
   Token Tok = P.getCurToken();
 
-  if (Tok.is(tok::identifier))
-    return false;
-
-  if (!Tok.isAnnotation() && Tok.getIdentifierInfo() &&
-      Tok.getIdentifierInfo()->isKeyword(P.getLangOpts()))
+  if (isTokenIdentifierOrKeyword(P, Tok))
     return false;
 
   P.Diag(P.getCurToken(), diag::err_expected) << tok::identifier;
@@ -329,12 +434,65 @@ OpenACCDirectiveKind ParseOpenACCDirectiveKind(Parser &P) {
   return DirKind;
 }
 
-bool ClauseHasOptionalParens(OpenACCClauseKind Kind) {
-  return Kind == OpenACCClauseKind::Self;
+enum ClauseParensKind {
+  None,
+  Optional,
+  Required
+};
+
+ClauseParensKind getClauseParensKind(OpenACCDirectiveKind DirKind,
+                                     OpenACCClauseKind Kind) {
+  switch (Kind) {
+  case OpenACCClauseKind::Self:
+    return DirKind == OpenACCDirectiveKind::Update ? ClauseParensKind::Required
+                                                   : ClauseParensKind::Optional;
+
+  case OpenACCClauseKind::Default:
+  case OpenACCClauseKind::If:
+  case OpenACCClauseKind::Create:
+  case OpenACCClauseKind::Copy:
+  case OpenACCClauseKind::CopyIn:
+  case OpenACCClauseKind::CopyOut:
+  case OpenACCClauseKind::UseDevice:
+  case OpenACCClauseKind::NoCreate:
+  case OpenACCClauseKind::Present:
+  case OpenACCClauseKind::DevicePtr:
+  case OpenACCClauseKind::Attach:
+  case OpenACCClauseKind::Detach:
+  case OpenACCClauseKind::Private:
+  case OpenACCClauseKind::FirstPrivate:
+  case OpenACCClauseKind::Delete:
+  case OpenACCClauseKind::DeviceResident:
+  case OpenACCClauseKind::Device:
+  case OpenACCClauseKind::Link:
+  case OpenACCClauseKind::Host:
+  case OpenACCClauseKind::Reduction:
+  case OpenACCClauseKind::Collapse:
+  case OpenACCClauseKind::Bind:
+    return ClauseParensKind::Required;
+
+  case OpenACCClauseKind::Auto:
+  case OpenACCClauseKind::Finalize:
+  case OpenACCClauseKind::IfPresent:
+  case OpenACCClauseKind::Independent:
+  case OpenACCClauseKind::Invalid:
+  case OpenACCClauseKind::NoHost:
+  case OpenACCClauseKind::Seq:
+  case OpenACCClauseKind::Worker:
+  case OpenACCClauseKind::Vector:
+    return ClauseParensKind::None;
+  }
+  llvm_unreachable("Unhandled clause kind");
+}
+
+bool ClauseHasOptionalParens(OpenACCDirectiveKind DirKind,
+                             OpenACCClauseKind Kind) {
+  return getClauseParensKind(DirKind, Kind) == ClauseParensKind::Optional;
 }
 
-bool ClauseHasRequiredParens(OpenACCClauseKind Kind) {
-  return Kind == OpenACCClauseKind::Default || Kind == OpenACCClauseKind::If;
+bool ClauseHasRequiredParens(OpenACCDirectiveKind DirKind,
+                             OpenACCClauseKind Kind) {
+  return getClauseParensKind(DirKind, Kind) == ClauseParensKind::Required;
 }
 
 ExprResult ParseOpenACCConditionalExpr(Parser &P) {
@@ -345,53 +503,190 @@ ExprResult ParseOpenACCConditionalExpr(Parser &P) {
   return P.getActions().CorrectDelayedTyposInExpr(P.ParseExpression());
 }
 
-bool ParseOpenACCClauseParams(Parser &P, OpenACCClauseKind Kind) {
-  BalancedDelimiterTracker Parens(P, tok::l_paren,
+// Skip until we see the end of pragma token, but don't consume it. This is us
+// just giving up on the rest of the pragma so we can continue executing. We
+// have to do this because 'SkipUntil' considers paren balancing, which isn't
+// what we want.
+void SkipUntilEndOfDirective(Parser &P) {
+  while (P.getCurToken().isNot(tok::annot_pragma_openacc_end))
+    P.ConsumeAnyToken();
+}
+
+} // namespace
+
+// OpenACC 3.3, section 1.7:
+// To simplify the specification and convey appropriate constraint information,
+// a pqr-list is a comma-separated list of pdr items. The one exception is a
+// clause-list, which is a list of one or more clauses optionally separated by
+// commas.
+void Parser::ParseOpenACCClauseList(OpenACCDirectiveKind DirKind) {
+  bool FirstClause = true;
+  while (getCurToken().isNot(tok::annot_pragma_openacc_end)) {
+    // Comma is optional in a clause-list.
+    if (!FirstClause && getCurToken().is(tok::comma))
+      ConsumeToken();
+    FirstClause = false;
+
+    // Recovering from a bad clause is really difficult, so we just give up on
+    // error.
+    if (ParseOpenACCClause(DirKind)) {
+      SkipUntilEndOfDirective(*this);
+      return;
+    }
+  }
+}
+
+bool Parser::ParseOpenACCClauseVarList(OpenACCClauseKind Kind) {
+  // FIXME: Future clauses will require 'special word' parsing, check for one,
+  // then parse it based on whether it is a clause that requires a 'special
+  // word'.
+  (void)Kind;
+
+  // If the var parsing fails, skip until the end of the directive as this is
+  // an expression and gets messy if we try to continue otherwise.
+  if (ParseOpenACCVar())
+    return true;
+
+  while (!getCurToken().isOneOf(tok::r_paren, tok::annot_pragma_openacc_end)) {
+    ExpectAndConsume(tok::comma);
+
+    // If the var parsing fails, skip until the end of the directive as this is
+    // an expression and gets messy if we try to continue otherwise.
+    if (ParseOpenACCVar())
+      return true;
+  }
+  return false;
+}
+// The OpenACC Clause List is a comma or space-delimited list of clauses (see
+// the comment on ParseOpenACCClauseList).  The concept of a 'clause' doesn't
+// really have its owner grammar and each individual one has its own definition.
+// However, they all are named with a single-identifier (or auto/default!)
+// token, followed in some cases by either braces or parens.
+bool Parser::ParseOpenACCClause(OpenACCDirectiveKind DirKind) {
+  // A number of clause names are actually keywords, so accept a keyword that
+  // can be converted to a name.
+  if (expectIdentifierOrKeyword(*this))
+    return true;
+
+  OpenACCClauseKind Kind = getOpenACCClauseKind(getCurToken());
+
+  if (Kind == OpenACCClauseKind::Invalid)
+    return Diag(getCurToken(), diag::err_acc_invalid_clause)
+           << getCurToken().getIdentifierInfo();
+
+  // Consume the clause name.
+  ConsumeToken();
+
+  return ParseOpenACCClauseParams(DirKind, Kind);
+}
+
+bool Parser::ParseOpenACCClauseParams(OpenACCDirectiveKind DirKind,
+                                      OpenACCClauseKind Kind) {
+  BalancedDelimiterTracker Parens(*this, tok::l_paren,
                                   tok::annot_pragma_openacc_end);
 
-  if (ClauseHasRequiredParens(Kind)) {
+  if (ClauseHasRequiredParens(DirKind, Kind)) {
     if (Parens.expectAndConsume()) {
       // We are missing a paren, so assume that the person just forgot the
       // parameter.  Return 'false' so we try to continue on and parse the next
       // clause.
-      P.SkipUntil(tok::comma, tok::r_paren, tok::annot_pragma_openacc_end,
-                  Parser::StopBeforeMatch);
+      SkipUntil(tok::comma, tok::r_paren, tok::annot_pragma_openacc_end,
+                Parser::StopBeforeMatch);
       return false;
     }
 
     switch (Kind) {
     case OpenACCClauseKind::Default: {
-      Token DefKindTok = P.getCurToken();
+      Token DefKindTok = getCurToken();
 
-      if (expectIdentifierOrKeyword(P))
+      if (expectIdentifierOrKeyword(*this))
         break;
 
-      P.ConsumeToken();
+      ConsumeToken();
 
       if (getOpenACCDefaultClauseKind(DefKindTok) ==
           OpenACCDefaultClauseKind::Invalid)
-        P.Diag(DefKindTok, diag::err_acc_invalid_default_clause_kind);
+        Diag(DefKindTok, diag::err_acc_invalid_default_clause_kind);
 
       break;
     }
     case OpenACCClauseKind::If: {
-      ExprResult CondExpr = ParseOpenACCConditionalExpr(P);
+      ExprResult CondExpr = ParseOpenACCConditionalExpr(*this);
       // An invalid expression can be just about anything, so just give up on
       // this clause list.
       if (CondExpr.isInvalid())
         return true;
       break;
     }
+    case OpenACCClauseKind::CopyIn:
+      tryParseAndConsumeSpecialTokenKind(
+          *this, OpenACCSpecialTokenKind::ReadOnly, Kind);
+      if (ParseOpenACCClauseVarList(Kind))
+        return true;
+      break;
+    case OpenACCClauseKind::Create:
+    case OpenACCClauseKind::CopyOut:
+      tryParseAndConsumeSpecialTokenKind(*this, OpenACCSpecialTokenKind::Zero,
+                                         Kind);
+      if (ParseOpenACCClauseVarList(Kind))
+        return true;
+      break;
+    case OpenACCClauseKind::Reduction:
+      // If we're missing a clause-kind (or it is invalid), see if we can parse
+      // the var-list anyway.
+      ParseReductionOperator(*this);
+      if (ParseOpenACCClauseVarList(Kind))
+        return true;
+      break;
+    case OpenACCClauseKind::Self:
+      // The 'self' clause is a var-list instead of a 'condition' in the case of
+      // the 'update' clause, so we have to handle it here.  U se an assert to
+      // make sure we get the right differentiator.
+      assert(DirKind == OpenACCDirectiveKind::Update);
+      LLVM_FALLTHROUGH;
+    case OpenACCClauseKind::Attach:
+    case OpenACCClauseKind::Copy:
+    case OpenACCClauseKind::Delete:
+    case OpenACCClauseKind::Detach:
+    case OpenACCClauseKind::Device:
+    case OpenACCClauseKind::DeviceResident:
+    case OpenACCClauseKind::DevicePtr:
+    case OpenACCClauseKind::FirstPrivate:
+    case OpenACCClauseKind::Host:
+    case OpenACCClauseKind::Link:
+    case OpenACCClauseKind::NoCreate:
+    case OpenACCClauseKind::Present:
+    case OpenACCClauseKind::Private:
+    case OpenACCClauseKind::UseDevice:
+      if (ParseOpenACCClauseVarList(Kind))
+        return true;
+      break;
+    case OpenACCClauseKind::Collapse: {
+      tryParseAndConsumeSpecialTokenKind(*this, OpenACCSpecialTokenKind::Force,
+                                         Kind);
+      ExprResult NumLoops =
+          getActions().CorrectDelayedTyposInExpr(ParseAssignmentExpression());
+      if (NumLoops.isInvalid())
+        return true;
+      break;
+    }
+    case OpenACCClauseKind::Bind: {
+      ExprResult BindArg = ParseOpenACCBindClauseArgument();
+      if (BindArg.isInvalid())
+        return true;
+      break;
+    }
     default:
       llvm_unreachable("Not a required parens type?");
     }
 
     return Parens.consumeClose();
-  } else if (ClauseHasOptionalParens(Kind)) {
+  } else if (ClauseHasOptionalParens(DirKind, Kind)) {
     if (!Parens.consumeOpen()) {
       switch (Kind) {
       case OpenACCClauseKind::Self: {
-        ExprResult CondExpr = ParseOpenACCConditionalExpr(P);
+        assert(DirKind != OpenACCDirectiveKind::Update);
+        ExprResult CondExpr = ParseOpenACCConditionalExpr(*this);
         // An invalid expression can be just about anything, so just give up on
         // this clause list.
         if (CondExpr.isInvalid())
@@ -407,62 +702,6 @@ bool ParseOpenACCClauseParams(Parser &P, OpenACCClauseKind Kind) {
   return false;
 }
 
-// The OpenACC Clause List is a comma or space-delimited list of clauses (see
-// the comment on ParseOpenACCClauseList).  The concept of a 'clause' doesn't
-// really have its owner grammar and each individual one has its own definition.
-// However, they all are named with a single-identifier (or auto/default!)
-// token, followed in some cases by either braces or parens.
-bool ParseOpenACCClause(Parser &P) {
-  // A number of clause names are actually keywords, so accept a keyword that
-  // can be converted to a name.
-  if (expectIdentifierOrKeyword(P))
-    return true;
-
-  OpenACCClauseKind Kind = getOpenACCClauseKind(P.getCurToken());
-
-  if (Kind == OpenACCClauseKind::Invalid)
-    return P.Diag(P.getCurToken(), diag::err_acc_invalid_clause)
-           << P.getCurToken().getIdentifierInfo();
-
-  // Consume the clause name.
-  P.ConsumeToken();
-
-  return ParseOpenACCClauseParams(P, Kind);
-}
-
-// Skip until we see the end of pragma token, but don't consume it. This is us
-// just giving up on the rest of the pragma so we can continue executing. We
-// have to do this because 'SkipUntil' considers paren balancing, which isn't
-// what we want.
-void SkipUntilEndOfDirective(Parser &P) {
-  while (P.getCurToken().isNot(tok::annot_pragma_openacc_end))
-    P.ConsumeAnyToken();
-}
-
-// OpenACC 3.3, section 1.7:
-// To simplify the specification and convey appropriate constraint information,
-// a pqr-list is a comma-separated list of pdr items. The one exception is a
-// clause-list, which is a list of one or more clauses optionally separated by
-// commas.
-void ParseOpenACCClauseList(Parser &P) {
-  bool FirstClause = true;
-  while (P.getCurToken().isNot(tok::annot_pragma_openacc_end)) {
-    // Comma is optional in a clause-list.
-    if (!FirstClause && P.getCurToken().is(tok::comma))
-      P.ConsumeToken();
-    FirstClause = false;
-
-    // Recovering from a bad clause is really difficult, so we just give up on
-    // error.
-    if (ParseOpenACCClause(P)) {
-      SkipUntilEndOfDirective(P);
-      return;
-    }
-  }
-}
-
-} // namespace
-
 /// OpenACC 3.3, section 2.16:
 /// In this section and throughout the specification, the term wait-argument
 /// means:
@@ -527,7 +766,7 @@ bool Parser::ParseOpenACCWaitArgument() {
 ExprResult Parser::ParseOpenACCIDExpression() {
   ExprResult Res;
   if (getLangOpts().CPlusPlus) {
-    Res = ParseCXXIdExpression(/*isAddressOfOperand=*/false);
+    Res = ParseCXXIdExpression(/*isAddressOfOperand=*/true);
   } else {
     // There isn't anything quite the same as ParseCXXIdExpression for C, so we
     // need to get the identifier, then call into Sema ourselves.
@@ -554,6 +793,25 @@ ExprResult Parser::ParseOpenACCIDExpression() {
   return getActions().CorrectDelayedTyposInExpr(Res);
 }
 
+ExprResult Parser::ParseOpenACCBindClauseArgument() {
+  // OpenACC 3.3 section 2.15:
+  // The bind clause specifies the name to use when calling the procedure on a
+  // device other than the host. If the name is specified as an identifier, it
+  // is called as if that name were specified in the language being compiled. If
+  // the name is specified as a string, the string is used for the procedure
+  // name unmodified.
+  if (getCurToken().is(tok::r_paren)) {
+    Diag(getCurToken(), diag::err_acc_incorrect_bind_arg);
+    return ExprError();
+  }
+
+  if (tok::isStringLiteral(getCurToken().getKind()))
+    return getActions().CorrectDelayedTyposInExpr(ParseStringLiteralExpression(
+        /*AllowUserDefinedLiteral=*/false, /*Unevaluated=*/true));
+
+  return ParseOpenACCIDExpression();
+}
+
 /// OpenACC 3.3, section 1.6:
 /// In this spec, a 'var' (in italics) is one of the following:
 /// - a variable name (a scalar, array, or compisite variable name)
@@ -563,7 +821,8 @@ ExprResult Parser::ParseOpenACCIDExpression() {
 /// - a common block name between slashes (fortran only)
 bool Parser::ParseOpenACCVar() {
   OpenACCArraySectionRAII ArraySections(*this);
-  ExprResult Res = ParseAssignmentExpression();
+  ExprResult Res =
+      getActions().CorrectDelayedTyposInExpr(ParseAssignmentExpression());
   return Res.isInvalid();
 }
 
@@ -578,14 +837,11 @@ void Parser::ParseOpenACCCacheVarList() {
     return;
 
   // The VarList is an optional `readonly:` followed by a list of a variable
-  // specifications.  First, see if we have `readonly:`, else we back-out and
-  // treat it like the beginning of a reference to a potentially-existing
-  // `readonly` variable.
-  if (isOpenACCSpecialToken(OpenACCSpecialTokenKind::ReadOnly, Tok) &&
-      NextToken().is(tok::colon)) {
-    // Consume both tokens.
-    ConsumeToken();
-    ConsumeToken();
+  // specifications. Consume something that looks like a 'tag', and diagnose if
+  // it isn't 'readonly'.
+  if (tryParseAndConsumeSpecialTokenKind(*this,
+                                         OpenACCSpecialTokenKind::ReadOnly,
+                                         OpenACCDirectiveKind::Cache)) {
     // FIXME: Record that this is a 'readonly' so that we can use that during
     // Sema/AST generation.
   }
@@ -664,7 +920,7 @@ void Parser::ParseOpenACCDirective() {
   }
 
   // Parses the list of clauses, if present.
-  ParseOpenACCClauseList(*this);
+  ParseOpenACCClauseList(DirKind);
 
   Diag(getCurToken(), diag::warn_pragma_acc_unimplemented);
   assert(Tok.is(tok::annot_pragma_openacc_end) &&
diff --git a/clang/lib/Parse/ParseTentative.cpp b/clang/lib/Parse/ParseTentative.cpp
index 242741c15b5ff..5bfabf55f50c4 100644
--- a/clang/lib/Parse/ParseTentative.cpp
+++ b/clang/lib/Parse/ParseTentative.cpp
@@ -894,7 +894,8 @@ bool Parser::TrySkipAttributes() {
       // Note that explicitly checking for `[[` and `]]` allows to fail as
       // expected in the case of the Objective-C message send syntax.
       ConsumeBracket();
-    } else if (Tok.isRegularKeywordAttribute()) {
+    } else if (Tok.isRegularKeywordAttribute() &&
+               !doesKeywordAttributeTakeArgs(Tok.getKind())) {
       ConsumeToken();
     } else {
       ConsumeToken();
diff --git a/clang/lib/Parse/Parser.cpp b/clang/lib/Parse/Parser.cpp
index b703c2d9b8e04..0b092181bca7b 100644
--- a/clang/lib/Parse/Parser.cpp
+++ b/clang/lib/Parse/Parser.cpp
@@ -70,6 +70,11 @@ Parser::Parser(Preprocessor &pp, Sema &actions, bool skipFunctionBodies)
   PP.addCommentHandler(CommentSemaHandler.get());
 
   PP.setCodeCompletionHandler(*this);
+
+  Actions.ParseTypeFromStringCallback =
+      [this](StringRef TypeStr, StringRef Context, SourceLocation IncludeLoc) {
+        return this->ParseTypeFromString(TypeStr, Context, IncludeLoc);
+      };
 }
 
 DiagnosticBuilder Parser::Diag(SourceLocation Loc, unsigned DiagID) {
diff --git a/clang/lib/Sema/Sema.cpp b/clang/lib/Sema/Sema.cpp
index 9ad274fc8f185..58088a66b3aa6 100644
--- a/clang/lib/Sema/Sema.cpp
+++ b/clang/lib/Sema/Sema.cpp
@@ -985,7 +985,7 @@ static bool MethodsAndNestedClassesComplete(const CXXRecordDecl *RD,
        I != E && Complete; ++I) {
     if (const CXXMethodDecl *M = dyn_cast(*I))
       Complete = M->isDefined() || M->isDefaulted() ||
-                 (M->isPure() && !isa(M));
+                 (M->isPureVirtual() && !isa(M));
     else if (const FunctionTemplateDecl *F = dyn_cast(*I))
       // If the template function is marked as late template parsed at this
       // point, it has not been instantiated and therefore we have not
diff --git a/clang/lib/Sema/SemaChecking.cpp b/clang/lib/Sema/SemaChecking.cpp
index 930dd79e63469..ee5c0fa1895ac 100644
--- a/clang/lib/Sema/SemaChecking.cpp
+++ b/clang/lib/Sema/SemaChecking.cpp
@@ -3228,11 +3228,15 @@ static void checkArmStreamingBuiltin(Sema &S, CallExpr *TheCall,
 }
 
 static bool hasSMEZAState(const FunctionDecl *FD) {
-  if (FD->hasAttr())
-    return true;
-  if (const auto *T = FD->getType()->getAs())
-    if (T->getAArch64SMEAttributes() & FunctionType::SME_PStateZASharedMask)
+  if (auto *Attr = FD->getAttr())
+    if (Attr->isNewZA())
+      return true;
+  if (const auto *T = FD->getType()->getAs()) {
+    FunctionType::ArmStateValue State =
+        FunctionType::getArmZAState(T->getAArch64SMEAttributes());
+    if (State != FunctionType::ARM_None)
       return true;
+  }
   return false;
 }
 
@@ -7729,14 +7733,19 @@ void Sema::checkCall(NamedDecl *FDecl, const FunctionProtoType *Proto,
 
     // If the callee uses AArch64 SME ZA state but the caller doesn't define
     // any, then this is an error.
-    if (ExtInfo.AArch64SMEAttributes & FunctionType::SME_PStateZASharedMask) {
+    FunctionType::ArmStateValue ArmZAState =
+        FunctionType::getArmZAState(ExtInfo.AArch64SMEAttributes);
+    if (ArmZAState != FunctionType::ARM_None) {
       bool CallerHasZAState = false;
       if (const auto *CallerFD = dyn_cast(CurContext)) {
-        if (CallerFD->hasAttr())
+        auto *Attr = CallerFD->getAttr();
+        if (Attr && Attr->isNewZA())
           CallerHasZAState = true;
-        else if (const auto *FPT = CallerFD->getType()->getAs())
-          CallerHasZAState = FPT->getExtProtoInfo().AArch64SMEAttributes &
-                             FunctionType::SME_PStateZASharedMask;
+        else if (const auto *FPT =
+                     CallerFD->getType()->getAs())
+          CallerHasZAState = FunctionType::getArmZAState(
+                                 FPT->getExtProtoInfo().AArch64SMEAttributes) !=
+                             FunctionType::ARM_None;
       }
 
       if (!CallerHasZAState)
diff --git a/clang/lib/Sema/SemaCoroutine.cpp b/clang/lib/Sema/SemaCoroutine.cpp
index bee80db8d166a..0e0f8f67dcd73 100644
--- a/clang/lib/Sema/SemaCoroutine.cpp
+++ b/clang/lib/Sema/SemaCoroutine.cpp
@@ -16,6 +16,7 @@
 #include "CoroutineStmtBuilder.h"
 #include "clang/AST/ASTLambda.h"
 #include "clang/AST/Decl.h"
+#include "clang/AST/Expr.h"
 #include "clang/AST/ExprCXX.h"
 #include "clang/AST/StmtCXX.h"
 #include "clang/Basic/Builtins.h"
diff --git a/clang/lib/Sema/SemaDecl.cpp b/clang/lib/Sema/SemaDecl.cpp
index fb181a36d3ff9..2953adb915d20 100644
--- a/clang/lib/Sema/SemaDecl.cpp
+++ b/clang/lib/Sema/SemaDecl.cpp
@@ -3877,8 +3877,7 @@ bool Sema::MergeFunctionDecl(FunctionDecl *New, NamedDecl *&OldD, Scope *S,
 
   // It is not permitted to redeclare an SME function with different SME
   // attributes.
-  if (IsInvalidSMECallConversion(Old->getType(), New->getType(),
-                                 AArch64SMECallConversionKind::MatchExactly)) {
+  if (IsInvalidSMECallConversion(Old->getType(), New->getType())) {
     Diag(New->getLocation(), diag::err_sme_attr_mismatch)
         << New->getType() << Old->getType();
     Diag(OldLocation, diag::note_previous_declaration);
@@ -4449,8 +4448,8 @@ bool Sema::MergeCompatibleFunctionDecls(FunctionDecl *New, FunctionDecl *Old,
   mergeDeclAttributes(New, Old);
 
   // Merge "pure" flag.
-  if (Old->isPure())
-    New->setPure();
+  if (Old->isPureVirtual())
+    New->setIsPureVirtual();
 
   // Merge "used" flag.
   if (Old->getMostRecentDecl()->isUsed(false))
@@ -5862,7 +5861,7 @@ Decl *Sema::BuildAnonymousStructOrUnion(Scope *S, DeclSpec &DS,
   // Mock up a declarator.
   Declarator Dc(DS, ParsedAttributesView::none(), DeclaratorContext::Member);
   StorageClass SC = StorageClassSpecToVarDeclStorageClass(DS);
-  TypeSourceInfo *TInfo = GetTypeForDeclarator(Dc, S);
+  TypeSourceInfo *TInfo = GetTypeForDeclarator(Dc);
   assert(TInfo && "couldn't build declarator info for anonymous struct/union");
 
   // Create a declaration for this anonymous struct/union.
@@ -5959,7 +5958,7 @@ Decl *Sema::BuildMicrosoftCAnonymousStruct(Scope *S, DeclSpec &DS,
 
   // Mock up a declarator.
   Declarator Dc(DS, ParsedAttributesView::none(), DeclaratorContext::TypeName);
-  TypeSourceInfo *TInfo = GetTypeForDeclarator(Dc, S);
+  TypeSourceInfo *TInfo = GetTypeForDeclarator(Dc);
   assert(TInfo && "couldn't build declarator info for anonymous struct");
 
   auto *ParentDecl = cast(CurContext);
@@ -6436,12 +6435,6 @@ NamedDecl *Sema::HandleDeclarator(Scope *S, Declarator &D,
   } else if (DiagnoseUnexpandedParameterPack(NameInfo, UPPC_DeclarationType))
     return nullptr;
 
-  // The scope passed in may not be a decl scope.  Zip up the scope tree until
-  // we find one that is.
-  while ((S->getFlags() & Scope::DeclScope) == 0 ||
-         (S->getFlags() & Scope::TemplateParamScope) != 0)
-    S = S->getParent();
-
   DeclContext *DC = CurContext;
   if (D.getCXXScopeSpec().isInvalid())
     D.setInvalidType();
@@ -6497,7 +6490,7 @@ NamedDecl *Sema::HandleDeclarator(Scope *S, Declarator &D,
     }
   }
 
-  TypeSourceInfo *TInfo = GetTypeForDeclarator(D, S);
+  TypeSourceInfo *TInfo = GetTypeForDeclarator(D);
   QualType R = TInfo->getType();
 
   if (DiagnoseUnexpandedParameterPack(D.getIdentifierLoc(), TInfo,
@@ -6594,6 +6587,12 @@ NamedDecl *Sema::HandleDeclarator(Scope *S, Declarator &D,
   if (getLangOpts().CPlusPlus)
     CheckExtraCXXDefaultArguments(D);
 
+  // The scope passed in may not be a decl scope.  Zip up the scope tree until
+  // we find one that is.
+  while ((S->getFlags() & Scope::DeclScope) == 0 ||
+         (S->getFlags() & Scope::TemplateParamScope) != 0)
+    S = S->getParent();
+
   NamedDecl *New;
 
   bool AddToScope = true;
@@ -7127,8 +7126,7 @@ static void checkAttributesAfterMerging(Sema &S, NamedDecl &ND) {
   if (WeakRefAttr *Attr = ND.getAttr()) {
     if (ND.isExternallyVisible()) {
       S.Diag(Attr->getLocation(), diag::err_attribute_weakref_not_static);
-      ND.dropAttr();
-      ND.dropAttr();
+      ND.dropAttrs();
     }
   }
 
@@ -9982,7 +9980,7 @@ Sema::ActOnFunctionDeclarator(Scope *S, Declarator &D, DeclContext *DC,
     if (const CXXRecordDecl *Parent =
           dyn_cast(NewFD->getDeclContext())) {
       if (Parent->isInterface() && cast(NewFD)->isUserProvided())
-        NewFD->setPure(true);
+        NewFD->setIsPureVirtual(true);
 
       // C++ [class.union]p2
       //   A union can have member functions, but not virtual functions.
@@ -10949,9 +10947,19 @@ Sema::ActOnFunctionDeclarator(Scope *S, Declarator &D, DeclContext *DC,
     // Precalculate whether this is a friend function template with a constraint
     // that depends on an enclosing template, per [temp.friend]p9.
     if (isFriend && FunctionTemplate &&
-        FriendConstraintsDependOnEnclosingTemplate(NewFD))
+        FriendConstraintsDependOnEnclosingTemplate(NewFD)) {
       NewFD->setFriendConstraintRefersToEnclosingTemplate(true);
 
+      // C++ [temp.friend]p9:
+      //    A friend function template with a constraint that depends on a
+      //    template parameter from an enclosing template shall be a definition.
+      if (!D.isFunctionDefinition()) {
+        Diag(NewFD->getBeginLoc(),
+             diag::err_friend_decl_with_enclosing_temp_constraint_must_be_def);
+        NewFD->setInvalidDecl();
+      }
+    }
+
     if (FunctionTemplate) {
       if (NewFD->isInvalidDecl())
         FunctionTemplate->setInvalidDecl();
@@ -12164,11 +12172,12 @@ bool Sema::CheckFunctionDeclaration(Scope *S, FunctionDecl *NewFD,
         checkThisInStaticMemberFunctionType(Method);
     }
 
-    // C++20: dcl.decl.general p4:
-    // The optional requires-clause ([temp.pre]) in an init-declarator or
-    // member-declarator shall be present only if the declarator declares a
-    // templated function ([dcl.fct]).
     if (Expr *TRC = NewFD->getTrailingRequiresClause()) {
+      // C++20: dcl.decl.general p4:
+      // The optional requires-clause ([temp.pre]) in an init-declarator or
+      // member-declarator shall be present only if the declarator declares a
+      // templated function ([dcl.fct]).
+      //
       // [temp.pre]/8:
       // An entity is templated if it is
       // - a template,
@@ -12186,15 +12195,29 @@ bool Sema::CheckFunctionDeclaration(Scope *S, FunctionDecl *NewFD,
       // templated. A templated variable is a variable template or a variable
       // that is templated.
 
-      if (!NewFD->getDescribedFunctionTemplate() && // -a template
-          // defined... in a templated entity
+      bool IsTemplate = NewFD->getDescribedFunctionTemplate();
+      bool IsFriend = NewFD->getFriendObjectKind();
+      if (!IsTemplate && // -a template
+                         // defined... in a templated entity
           !(DeclIsDefn && NewFD->isTemplated()) &&
           // a member of a templated entity
           !(isa(NewFD) && NewFD->isTemplated()) &&
           // Don't complain about instantiations, they've already had these
           // rules + others enforced.
-          !NewFD->isTemplateInstantiation()) {
+          !NewFD->isTemplateInstantiation() &&
+          // If the function violates [temp.friend]p9 because it is missing
+          // a definition, and adding a definition would make it templated,
+          // then let that error take precedence.
+          !(!DeclIsDefn && IsFriend && NewFD->isTemplated())) {
         Diag(TRC->getBeginLoc(), diag::err_constrained_non_templated_function);
+      } else if (!DeclIsDefn && !IsTemplate && IsFriend &&
+                 !NewFD->isTemplateInstantiation()) {
+        // C++ [temp.friend]p9:
+        //   A non-template friend declaration with a requires-clause shall be a
+        //   definition.
+        Diag(NewFD->getBeginLoc(),
+             diag::err_non_temp_friend_decl_with_requires_clause_must_be_def);
+        NewFD->setInvalidDecl();
       }
     }
 
@@ -12278,13 +12301,15 @@ bool Sema::CheckFunctionDeclaration(Scope *S, FunctionDecl *NewFD,
   // Check if the function definition uses any AArch64 SME features without
   // having the '+sme' feature enabled.
   if (DeclIsDefn) {
+    const auto *Attr = NewFD->getAttr();
     bool UsesSM = NewFD->hasAttr();
-    bool UsesZA = NewFD->hasAttr();
+    bool UsesZA = Attr && Attr->isNewZA();
     if (const auto *FPT = NewFD->getType()->getAs()) {
       FunctionProtoType::ExtProtoInfo EPI = FPT->getExtProtoInfo();
       UsesSM |=
           EPI.AArch64SMEAttributes & FunctionType::SME_PStateSMEnabledMask;
-      UsesZA |= EPI.AArch64SMEAttributes & FunctionType::SME_PStateZASharedMask;
+      UsesZA |= FunctionType::getArmZAState(EPI.AArch64SMEAttributes) !=
+                FunctionType::ARM_None;
     }
 
     if (UsesSM || UsesZA) {
@@ -13681,6 +13706,15 @@ void Sema::AddInitializerToDecl(Decl *RealDecl, Expr *Init, bool DirectInit) {
           CreateRecoveryExpr(Init->getBeginLoc(), Init->getEndLoc(), Args);
       if (RecoveryExpr.get())
         VDecl->setInit(RecoveryExpr.get());
+      // In general, for error recovery purposes, the initalizer doesn't play
+      // part in the valid bit of the declaration. There are a few exceptions:
+      //  1) if the var decl has a deduced auto type, and the type cannot be
+      //     deduced by an invalid initializer;
+      //  2) if the var decl is decompsition decl with a non-deduced type, and
+      //     the initialization fails (e.g. `int [a] = {1, 2};`);
+      // Case 1) was already handled elsewhere.
+      if (isa(VDecl)) // Case 2)
+        VDecl->setInvalidDecl();
       return;
     }
 
@@ -15159,7 +15193,7 @@ Decl *Sema::ActOnParamDeclarator(Scope *S, Declarator &D,
 
   CheckFunctionOrTemplateParamDeclarator(S, D);
 
-  TypeSourceInfo *TInfo = GetTypeForDeclarator(D, S);
+  TypeSourceInfo *TInfo = GetTypeForDeclarator(D);
   QualType parmDeclType = TInfo->getType();
 
   // Check for redeclaration of parameters, e.g. int foo(int x, int x);
@@ -15283,6 +15317,37 @@ void Sema::DiagnoseSizeOfParametersAndReturnValue(
   }
 }
 
+QualType Sema::AdjustParameterTypeForObjCAutoRefCount(QualType T,
+                                                      SourceLocation NameLoc,
+                                                      TypeSourceInfo *TSInfo) {
+  // In ARC, infer a lifetime qualifier for appropriate parameter types.
+  if (!getLangOpts().ObjCAutoRefCount ||
+      T.getObjCLifetime() != Qualifiers::OCL_None || !T->isObjCLifetimeType())
+    return T;
+
+  Qualifiers::ObjCLifetime Lifetime;
+
+  // Special cases for arrays:
+  //   - if it's const, use __unsafe_unretained
+  //   - otherwise, it's an error
+  if (T->isArrayType()) {
+    if (!T.isConstQualified()) {
+      if (DelayedDiagnostics.shouldDelayDiagnostics())
+        DelayedDiagnostics.add(sema::DelayedDiagnostic::makeForbiddenType(
+            NameLoc, diag::err_arc_array_param_no_ownership, T, false));
+      else
+        Diag(NameLoc, diag::err_arc_array_param_no_ownership)
+            << TSInfo->getTypeLoc().getSourceRange();
+    }
+    Lifetime = Qualifiers::OCL_ExplicitNone;
+  } else {
+    Lifetime = T->getObjCARCImplicitLifetime();
+  }
+  T = Context.getLifetimeQualifiedType(T, Lifetime);
+
+  return T;
+}
+
 ParmVarDecl *Sema::CheckParameter(DeclContext *DC, SourceLocation StartLoc,
                                   SourceLocation NameLoc, IdentifierInfo *Name,
                                   QualType T, TypeSourceInfo *TSInfo,
@@ -15960,13 +16025,26 @@ static void diagnoseImplicitlyRetainedSelf(Sema &S) {
           << FixItHint::CreateInsertion(P.first, "self->");
 }
 
+static bool methodHasName(const FunctionDecl *FD, StringRef Name) {
+  return isa(FD) && FD->param_empty() &&
+         FD->getDeclName().isIdentifier() && FD->getName().equals(Name);
+}
+
+bool Sema::CanBeGetReturnObject(const FunctionDecl *FD) {
+  return methodHasName(FD, "get_return_object");
+}
+
+bool Sema::CanBeGetReturnTypeOnAllocFailure(const FunctionDecl *FD) {
+  return FD->isStatic() &&
+         methodHasName(FD, "get_return_object_on_allocation_failure");
+}
+
 void Sema::CheckCoroutineWrapper(FunctionDecl *FD) {
   RecordDecl *RD = FD->getReturnType()->getAsRecordDecl();
   if (!RD || !RD->getUnderlyingDecl()->hasAttr())
     return;
-  // Allow `get_return_object()`.
-  if (FD->getDeclName().isIdentifier() &&
-      FD->getName().equals("get_return_object") && FD->param_empty())
+  // Allow some_promise_type::get_return_object().
+  if (CanBeGetReturnObject(FD) || CanBeGetReturnTypeOnAllocFailure(FD))
     return;
   if (!FD->hasAttr())
     Diag(FD->getLocation(), diag::err_coroutine_return_type) << RD;
@@ -16047,7 +16125,8 @@ Decl *Sema::ActOnFinishFunctionBody(Decl *dcl, Stmt *Body,
 
       // MSVC permits the use of pure specifier (=0) on function definition,
       // defined at class scope, warn about this non-standard construct.
-      if (getLangOpts().MicrosoftExt && FD->isPure() && !FD->isOutOfLine())
+      if (getLangOpts().MicrosoftExt && FD->isPureVirtual() &&
+          !FD->isOutOfLine())
         Diag(FD->getLocation(), diag::ext_pure_function_definition);
 
       if (!FD->isInvalidDecl()) {
@@ -18425,7 +18504,7 @@ FieldDecl *Sema::HandleField(Scope *S, RecordDecl *Record,
   SourceLocation Loc = DeclStart;
   if (II) Loc = D.getIdentifierLoc();
 
-  TypeSourceInfo *TInfo = GetTypeForDeclarator(D, S);
+  TypeSourceInfo *TInfo = GetTypeForDeclarator(D);
   QualType T = TInfo->getType();
   if (getLangOpts().CPlusPlus) {
     CheckExtraCXXDefaultArguments(D);
@@ -18789,7 +18868,7 @@ Decl *Sema::ActOnIvar(Scope *S, SourceLocation DeclStart, Declarator &D,
   // FIXME: Unnamed fields can be handled in various different ways, for
   // example, unnamed unions inject all members into the struct namespace!
 
-  TypeSourceInfo *TInfo = GetTypeForDeclarator(D, S);
+  TypeSourceInfo *TInfo = GetTypeForDeclarator(D);
   QualType T = TInfo->getType();
 
   if (BitWidth) {
@@ -20334,7 +20413,7 @@ void Sema::ActOnEnumBody(SourceLocation EnumLoc, SourceRange BraceRange,
     // Adjust the APSInt value.
     InitVal = InitVal.extOrTrunc(NewWidth);
     InitVal.setIsSigned(NewSign);
-    ECD->setInitVal(InitVal);
+    ECD->setInitVal(Context, InitVal);
 
     // Adjust the Expr initializer and type.
     if (ECD->getInitExpr() &&
diff --git a/clang/lib/Sema/SemaDeclAttr.cpp b/clang/lib/Sema/SemaDeclAttr.cpp
index fb4010dd0f79b..5ea8469cc9ed3 100644
--- a/clang/lib/Sema/SemaDeclAttr.cpp
+++ b/clang/lib/Sema/SemaDeclAttr.cpp
@@ -9023,29 +9023,35 @@ static void handleObjCRequiresSuperAttr(Sema &S, Decl *D,
   D->addAttr(::new (S.Context) ObjCRequiresSuperAttr(S.Context, Attrs));
 }
 
-static void handleNSErrorDomain(Sema &S, Decl *D, const ParsedAttr &AL) {
-  auto *E = AL.getArgAsExpr(0);
-  auto Loc = E ? E->getBeginLoc() : AL.getLoc();
-
-  auto *DRE = dyn_cast(AL.getArgAsExpr(0));
-  if (!DRE) {
-    S.Diag(Loc, diag::err_nserrordomain_invalid_decl) << 0;
+static void handleNSErrorDomain(Sema &S, Decl *D, const ParsedAttr &Attr) {
+  if (!isa(D)) {
+    S.Diag(D->getBeginLoc(), diag::err_nserrordomain_invalid_decl) << 0;
     return;
   }
 
-  auto *VD = dyn_cast(DRE->getDecl());
-  if (!VD) {
-    S.Diag(Loc, diag::err_nserrordomain_invalid_decl) << 1 << DRE->getDecl();
+  IdentifierLoc *IdentLoc =
+      Attr.isArgIdent(0) ? Attr.getArgAsIdent(0) : nullptr;
+  if (!IdentLoc || !IdentLoc->Ident) {
+    // Try to locate the argument directly.
+    SourceLocation Loc = Attr.getLoc();
+    if (Attr.isArgExpr(0) && Attr.getArgAsExpr(0))
+      Loc = Attr.getArgAsExpr(0)->getBeginLoc();
+
+    S.Diag(Loc, diag::err_nserrordomain_invalid_decl) << 0;
     return;
   }
 
-  if (!isNSStringType(VD->getType(), S.Context) &&
-      !isCFStringType(VD->getType(), S.Context)) {
-    S.Diag(Loc, diag::err_nserrordomain_wrong_type) << VD;
+  // Verify that the identifier is a valid decl in the C decl namespace.
+  LookupResult Result(S, DeclarationName(IdentLoc->Ident), SourceLocation(),
+                      Sema::LookupNameKind::LookupOrdinaryName);
+  if (!S.LookupName(Result, S.TUScope) || !Result.getAsSingle()) {
+    S.Diag(IdentLoc->Loc, diag::err_nserrordomain_invalid_decl)
+        << 1 << IdentLoc->Ident;
     return;
   }
 
-  D->addAttr(::new (S.Context) NSErrorDomainAttr(S.Context, AL, VD));
+  D->addAttr(::new (S.Context)
+                 NSErrorDomainAttr(S.Context, Attr, IdentLoc->Ident));
 }
 
 static void handleObjCBridgeAttr(Sema &S, Decl *D, const ParsedAttr &AL) {
@@ -11882,26 +11888,74 @@ static bool MustDelayAttributeArguments(const ParsedAttr &AL) {
   return false;
 }
 
+static bool checkArmNewAttrMutualExclusion(
+    Sema &S, const ParsedAttr &AL, const FunctionProtoType *FPT,
+    FunctionType::ArmStateValue CurrentState, StringRef StateName) {
+  auto CheckForIncompatibleAttr =
+      [&](FunctionType::ArmStateValue IncompatibleState,
+          StringRef IncompatibleStateName) {
+        if (CurrentState == IncompatibleState) {
+          S.Diag(AL.getLoc(), diag::err_attributes_are_not_compatible)
+              << (std::string("'__arm_new(\"") + StateName.str() + "\")'")
+              << (std::string("'") + IncompatibleStateName.str() + "(\"" +
+                  StateName.str() + "\")'")
+              << true;
+          AL.setInvalid();
+        }
+      };
+
+  CheckForIncompatibleAttr(FunctionType::ARM_In, "__arm_in");
+  CheckForIncompatibleAttr(FunctionType::ARM_Out, "__arm_out");
+  CheckForIncompatibleAttr(FunctionType::ARM_InOut, "__arm_inout");
+  CheckForIncompatibleAttr(FunctionType::ARM_Preserves, "__arm_preserves");
+  return AL.isInvalid();
+}
 
-static void handleArmNewZaAttr(Sema &S, Decl *D, const ParsedAttr &AL) {
-  if (auto *FPT = dyn_cast(D->getFunctionType())) {
-    if (FPT->getAArch64SMEAttributes() &
-        FunctionType::SME_PStateZASharedMask) {
-      S.Diag(AL.getLoc(), diag::err_attributes_are_not_compatible)
-          << AL << "'__arm_shared_za'" << true;
+static void handleArmNewAttr(Sema &S, Decl *D, const ParsedAttr &AL) {
+  if (!AL.getNumArgs()) {
+    S.Diag(AL.getLoc(), diag::err_missing_arm_state) << AL;
+    AL.setInvalid();
+    return;
+  }
+
+  std::vector NewState;
+  if (const auto *ExistingAttr = D->getAttr()) {
+    for (StringRef S : ExistingAttr->newArgs())
+      NewState.push_back(S);
+  }
+
+  bool HasZA = false;
+  for (unsigned I = 0, E = AL.getNumArgs(); I != E; ++I) {
+    StringRef StateName;
+    SourceLocation LiteralLoc;
+    if (!S.checkStringLiteralArgumentAttr(AL, I, StateName, &LiteralLoc))
+      return;
+
+    if (StateName == "za")
+      HasZA = true;
+    else {
+      S.Diag(LiteralLoc, diag::err_unknown_arm_state) << StateName;
       AL.setInvalid();
+      return;
     }
-    if (FPT->getAArch64SMEAttributes() &
-        FunctionType::SME_PStateZAPreservedMask) {
-      S.Diag(AL.getLoc(), diag::err_attributes_are_not_compatible)
-          << AL << "'__arm_preserves_za'" << true;
-      AL.setInvalid();
+
+    if (std::find(NewState.begin(), NewState.end(), StateName) ==
+        NewState.end()) { // Avoid adding duplicates.
+      NewState.push_back(StateName);
     }
-    if (AL.isInvalid())
+  }
+
+  if (auto *FPT = dyn_cast(D->getFunctionType())) {
+    FunctionType::ArmStateValue ZAState =
+        FunctionType::getArmZAState(FPT->getAArch64SMEAttributes());
+    if (HasZA && ZAState != FunctionType::ARM_None &&
+        checkArmNewAttrMutualExclusion(S, AL, FPT, ZAState, "za"))
       return;
   }
 
-  handleSimpleAttribute(S, D, AL);
+  D->dropAttr();
+  D->addAttr(::new (S.Context)
+                 ArmNewAttr(S.Context, AL, NewState.data(), NewState.size()));
 }
 
 /// ProcessDeclAttribute - Apply the specific attribute to the specified decl if
@@ -12803,8 +12857,8 @@ ProcessDeclAttribute(Sema &S, Scope *scope, Decl *D, const ParsedAttr &AL,
     handleSimpleAttribute(S, D, AL);
     break;
 
-  case ParsedAttr::AT_ArmNewZA:
-    handleArmNewZaAttr(S, D, AL);
+  case ParsedAttr::AT_ArmNew:
+    handleArmNewAttr(S, D, AL);
     break;
 
   case ParsedAttr::AT_AcquireHandle:
diff --git a/clang/lib/Sema/SemaDeclCXX.cpp b/clang/lib/Sema/SemaDeclCXX.cpp
index 630fe45721bea..8664d735c853c 100644
--- a/clang/lib/Sema/SemaDeclCXX.cpp
+++ b/clang/lib/Sema/SemaDeclCXX.cpp
@@ -839,7 +839,7 @@ Sema::ActOnDecompositionDeclarator(Scope *S, Declarator &D,
     Diag(DS.getVolatileSpecLoc(),
          diag::warn_deprecated_volatile_structured_binding);
 
-  TypeSourceInfo *TInfo = GetTypeForDeclarator(D, S);
+  TypeSourceInfo *TInfo = GetTypeForDeclarator(D);
   QualType R = TInfo->getType();
 
   if (DiagnoseUnexpandedParameterPack(D.getIdentifierLoc(), TInfo,
@@ -6089,7 +6089,7 @@ void Sema::DiagnoseAbstractType(const CXXRecordDecl *RD) {
       if (SO->second.size() != 1)
         continue;
 
-      if (!SO->second.front().Method->isPure())
+      if (!SO->second.front().Method->isPureVirtual())
         continue;
 
       if (!SeenPureMethods.insert(SO->second.front().Method).second)
@@ -6572,8 +6572,7 @@ void Sema::checkClassLevelDLLAttribute(CXXRecordDecl *Class) {
   if ((Context.getTargetInfo().getCXXABI().isMicrosoft() ||
        Context.getTargetInfo().getTriple().isPS()) &&
       (!Class->isExternallyVisible() && Class->hasExternalFormalLinkage())) {
-    Class->dropAttr();
-    Class->dropAttr();
+    Class->dropAttrs();
     return;
   }
 
@@ -17038,7 +17037,7 @@ VarDecl *Sema::BuildExceptionDeclaration(Scope *S,
 /// ActOnExceptionDeclarator - Parsed the exception-declarator in a C++ catch
 /// handler.
 Decl *Sema::ActOnExceptionDeclarator(Scope *S, Declarator &D) {
-  TypeSourceInfo *TInfo = GetTypeForDeclarator(D, S);
+  TypeSourceInfo *TInfo = GetTypeForDeclarator(D);
   bool Invalid = D.isInvalidType();
 
   // Check for unexpanded parameter packs.
@@ -17789,7 +17788,7 @@ Decl *Sema::ActOnFriendTypeDecl(Scope *S, const DeclSpec &DS,
   // for a TUK_Friend.
   Declarator TheDeclarator(DS, ParsedAttributesView::none(),
                            DeclaratorContext::Member);
-  TypeSourceInfo *TSI = GetTypeForDeclarator(TheDeclarator, S);
+  TypeSourceInfo *TSI = GetTypeForDeclarator(TheDeclarator);
   QualType T = TSI->getType();
   if (TheDeclarator.isInvalidType())
     return nullptr;
@@ -17854,7 +17853,7 @@ NamedDecl *Sema::ActOnFriendFunctionDecl(Scope *S, Declarator &D,
   assert(DS.getStorageClassSpec() == DeclSpec::SCS_unspecified);
 
   SourceLocation Loc = D.getIdentifierLoc();
-  TypeSourceInfo *TInfo = GetTypeForDeclarator(D, S);
+  TypeSourceInfo *TInfo = GetTypeForDeclarator(D);
 
   // C++ [class.friend]p1
   //   A friend of a class is a function or class....
@@ -18355,9 +18354,7 @@ bool Sema::CheckOverridingFunctionAttributes(const CXXMethodDecl *New,
   }
 
   // SME attributes must match when overriding a function declaration.
-  if (IsInvalidSMECallConversion(
-          Old->getType(), New->getType(),
-          AArch64SMECallConversionKind::MayAddPreservesZA)) {
+  if (IsInvalidSMECallConversion(Old->getType(), New->getType())) {
     Diag(New->getLocation(), diag::err_conflicting_overriding_attributes)
         << New << New->getType() << Old->getType();
     Diag(Old->getLocation(), diag::note_overridden_virtual_function);
@@ -18525,7 +18522,7 @@ bool Sema::CheckPureMethod(CXXMethodDecl *Method, SourceRange InitRange) {
     Method->setRangeEnd(EndLoc);
 
   if (Method->isVirtual() || Method->getParent()->isDependentContext()) {
-    Method->setPure();
+    Method->setIsPureVirtual();
     return false;
   }
 
@@ -18796,7 +18793,7 @@ bool Sema::DefineUsedVTables() {
 void Sema::MarkVirtualMemberExceptionSpecsNeeded(SourceLocation Loc,
                                                  const CXXRecordDecl *RD) {
   for (const auto *I : RD->methods())
-    if (I->isVirtual() && !I->isPure())
+    if (I->isVirtual() && !I->isPureVirtual())
       ResolveExceptionSpec(Loc, I->getType()->castAs());
 }
 
@@ -18817,7 +18814,8 @@ void Sema::MarkVirtualMembersReferenced(SourceLocation Loc,
 
       // C++ [basic.def.odr]p2:
       //   [...] A virtual member function is used if it is not pure. [...]
-      if (!Overrider->isPure() && (!ConstexprOnly || Overrider->isConstexpr()))
+      if (!Overrider->isPureVirtual() &&
+          (!ConstexprOnly || Overrider->isConstexpr()))
         MarkFunctionReferenced(Loc, Overrider);
     }
   }
@@ -19211,7 +19209,7 @@ MSPropertyDecl *Sema::HandleMSProperty(Scope *S, RecordDecl *Record,
   }
   SourceLocation Loc = D.getIdentifierLoc();
 
-  TypeSourceInfo *TInfo = GetTypeForDeclarator(D, S);
+  TypeSourceInfo *TInfo = GetTypeForDeclarator(D);
   QualType T = TInfo->getType();
   if (getLangOpts().CPlusPlus) {
     CheckExtraCXXDefaultArguments(D);
diff --git a/clang/lib/Sema/SemaDeclObjC.cpp b/clang/lib/Sema/SemaDeclObjC.cpp
index c3b95e168a605..bb0d0cd2030ba 100644
--- a/clang/lib/Sema/SemaDeclObjC.cpp
+++ b/clang/lib/Sema/SemaDeclObjC.cpp
@@ -1623,7 +1623,7 @@ void Sema::actOnObjCTypeArgsOrProtocolQualifiers(
     }
 
     // Convert this to a type.
-    return ActOnTypeName(S, D);
+    return ActOnTypeName(D);
   };
 
   // Local function that updates the declaration specifiers with
@@ -5211,7 +5211,7 @@ Decl *Sema::ActOnObjCExceptionDecl(Scope *S, Declarator &D) {
   if (getLangOpts().CPlusPlus)
     CheckExtraCXXDefaultArguments(D);
 
-  TypeSourceInfo *TInfo = GetTypeForDeclarator(D, S);
+  TypeSourceInfo *TInfo = GetTypeForDeclarator(D);
   QualType ExceptionType = TInfo->getType();
 
   VarDecl *New = BuildObjCExceptionDecl(TInfo, ExceptionType,
diff --git a/clang/lib/Sema/SemaExpr.cpp b/clang/lib/Sema/SemaExpr.cpp
index f9f7ea8656239..2452d3e1bc0ca 100644
--- a/clang/lib/Sema/SemaExpr.cpp
+++ b/clang/lib/Sema/SemaExpr.cpp
@@ -5163,8 +5163,6 @@ static QualType getDependentArraySubscriptType(Expr *LHS, Expr *RHS,
   return Result->isDependentType() ? Result : Ctx.DependentTy;
 }
 
-static bool checkArgsForPlaceholders(Sema &S, MultiExprArg args);
-
 ExprResult Sema::ActOnArraySubscriptExpr(Scope *S, Expr *base,
                                          SourceLocation lbLoc,
                                          MultiExprArg ArgExprs,
@@ -5268,7 +5266,7 @@ ExprResult Sema::ActOnArraySubscriptExpr(Scope *S, Expr *base,
       return ExprError();
     ArgExprs[0] = result.get();
   } else {
-    if (checkArgsForPlaceholders(*this, ArgExprs))
+    if (CheckArgsForPlaceholders(ArgExprs))
       return ExprError();
   }
 
@@ -7022,15 +7020,13 @@ static bool isPlaceholderToRemoveAsArg(QualType type) {
   llvm_unreachable("bad builtin type kind");
 }
 
-/// Check an argument list for placeholders that we won't try to
-/// handle later.
-static bool checkArgsForPlaceholders(Sema &S, MultiExprArg args) {
+bool Sema::CheckArgsForPlaceholders(MultiExprArg args) {
   // Apply this processing to all the arguments at once instead of
   // dying at the first failure.
   bool hasInvalid = false;
   for (size_t i = 0, e = args.size(); i != e; i++) {
     if (isPlaceholderToRemoveAsArg(args[i]->getType())) {
-      ExprResult result = S.CheckPlaceholderExpr(args[i]);
+      ExprResult result = CheckPlaceholderExpr(args[i]);
       if (result.isInvalid()) hasInvalid = true;
       else args[i] = result.get();
     }
@@ -7304,7 +7300,7 @@ ExprResult Sema::BuildCallExpr(Scope *Scope, Expr *Fn, SourceLocation LParenLoc,
   if (Result.isInvalid()) return ExprError();
   Fn = Result.get();
 
-  if (checkArgsForPlaceholders(*this, ArgExprs))
+  if (CheckArgsForPlaceholders(ArgExprs))
     return ExprError();
 
   if (getLangOpts().CPlusPlus) {
@@ -9913,8 +9909,7 @@ ExprResult Sema::ActOnConditionalOp(SourceLocation QuestionLoc,
 }
 
 // Check that the SME attributes for PSTATE.ZA and PSTATE.SM are compatible.
-bool Sema::IsInvalidSMECallConversion(QualType FromType, QualType ToType,
-                                      AArch64SMECallConversionKind C) {
+bool Sema::IsInvalidSMECallConversion(QualType FromType, QualType ToType) {
   unsigned FromAttributes = 0, ToAttributes = 0;
   if (const auto *FromFn =
           dyn_cast(Context.getCanonicalType(FromType)))
@@ -9925,25 +9920,7 @@ bool Sema::IsInvalidSMECallConversion(QualType FromType, QualType ToType,
     ToAttributes =
         ToFn->getAArch64SMEAttributes() & FunctionType::SME_AttributeMask;
 
-  if (FromAttributes == ToAttributes)
-    return false;
-
-  // If the '__arm_preserves_za' is the only difference between the types,
-  // check whether we're allowed to add or remove it.
-  if ((FromAttributes ^ ToAttributes) ==
-      FunctionType::SME_PStateZAPreservedMask) {
-    switch (C) {
-    case AArch64SMECallConversionKind::MatchExactly:
-      return true;
-    case AArch64SMECallConversionKind::MayAddPreservesZA:
-      return !(ToAttributes & FunctionType::SME_PStateZAPreservedMask);
-    case AArch64SMECallConversionKind::MayDropPreservesZA:
-      return !(FromAttributes & FunctionType::SME_PStateZAPreservedMask);
-    }
-  }
-
-  // There has been a mismatch of attributes
-  return true;
+  return FromAttributes != ToAttributes;
 }
 
 // Check if we have a conversion between incompatible cmse function pointer
@@ -10112,9 +10089,7 @@ checkPointerTypesForAssignment(Sema &S, QualType LHSType, QualType RHSType,
     return Sema::IncompatibleFunctionPointer;
   if (IsInvalidCmseNSCallConversion(S, ltrans, rtrans))
     return Sema::IncompatibleFunctionPointer;
-  if (S.IsInvalidSMECallConversion(
-          rtrans, ltrans,
-          Sema::AArch64SMECallConversionKind::MayDropPreservesZA))
+  if (S.IsInvalidSMECallConversion(rtrans, ltrans))
     return Sema::IncompatibleFunctionPointer;
   return ConvTy;
 }
@@ -17105,7 +17080,7 @@ void Sema::ActOnBlockArguments(SourceLocation CaretLoc, Declarator &ParamInfo,
   assert(ParamInfo.getContext() == DeclaratorContext::BlockLiteral);
   BlockScopeInfo *CurBlock = getCurBlock();
 
-  TypeSourceInfo *Sig = GetTypeForDeclarator(ParamInfo, CurScope);
+  TypeSourceInfo *Sig = GetTypeForDeclarator(ParamInfo);
   QualType T = Sig->getType();
 
   // FIXME: We should allow unexpanded parameter packs here, but that would,
@@ -20904,7 +20879,7 @@ void Sema::MarkMemberReferenced(MemberExpr *E) {
   bool MightBeOdrUse = true;
   if (E->performsVirtualDispatch(getLangOpts())) {
     if (CXXMethodDecl *Method = dyn_cast(E->getMemberDecl()))
-      if (Method->isPure())
+      if (Method->isPureVirtual())
         MightBeOdrUse = false;
   }
   SourceLocation Loc =
diff --git a/clang/lib/Sema/SemaExprCXX.cpp b/clang/lib/Sema/SemaExprCXX.cpp
index 48430c4d48ab2..20693d01933c2 100644
--- a/clang/lib/Sema/SemaExprCXX.cpp
+++ b/clang/lib/Sema/SemaExprCXX.cpp
@@ -1936,7 +1936,7 @@ Sema::ActOnCXXNew(SourceLocation StartLoc, bool UseGlobal,
     }
   }
 
-  TypeSourceInfo *TInfo = GetTypeForDeclarator(D, /*Scope=*/nullptr);
+  TypeSourceInfo *TInfo = GetTypeForDeclarator(D);
   QualType AllocType = TInfo->getType();
   if (D.isInvalidType())
     return ExprError();
@@ -2291,6 +2291,9 @@ ExprResult Sema::BuildCXXNew(SourceRange Range, bool UseGlobal,
   bool PassAlignment = getLangOpts().AlignedAllocation &&
                        Alignment > NewAlignment;
 
+  if (CheckArgsForPlaceholders(PlacementArgs))
+    return ExprError();
+
   AllocationFunctionScope Scope = UseGlobal ? AFS_Global : AFS_Both;
   if (!AllocType->isDependentType() &&
       !Expr::hasAnyTypeDependentArguments(PlacementArgs) &&
diff --git a/clang/lib/Sema/SemaExprMember.cpp b/clang/lib/Sema/SemaExprMember.cpp
index 2abec3d86a27d..32998ae60eafe 100644
--- a/clang/lib/Sema/SemaExprMember.cpp
+++ b/clang/lib/Sema/SemaExprMember.cpp
@@ -782,7 +782,8 @@ Sema::BuildMemberReferenceExpr(Expr *Base, QualType BaseType,
                                const Scope *S,
                                ActOnMemberAccessExtraArgs *ExtraArgs) {
   if (BaseType->isDependentType() ||
-      (SS.isSet() && isDependentScopeSpecifier(SS)))
+      (SS.isSet() && isDependentScopeSpecifier(SS)) ||
+      NameInfo.getName().isDependentName())
     return ActOnDependentMemberExpr(Base, BaseType,
                                     IsArrow, OpLoc,
                                     SS, TemplateKWLoc, FirstQualifierInScope,
diff --git a/clang/lib/Sema/SemaInit.cpp b/clang/lib/Sema/SemaInit.cpp
index 763e7f9ba3ba3..847cc9c66bea1 100644
--- a/clang/lib/Sema/SemaInit.cpp
+++ b/clang/lib/Sema/SemaInit.cpp
@@ -12,6 +12,7 @@
 
 #include "clang/AST/ASTContext.h"
 #include "clang/AST/DeclObjC.h"
+#include "clang/AST/Expr.h"
 #include "clang/AST/ExprCXX.h"
 #include "clang/AST/ExprObjC.h"
 #include "clang/AST/ExprOpenMP.h"
@@ -7584,15 +7585,27 @@ static void visitLifetimeBoundArguments(IndirectLocalPath &Path, Expr *Call,
     Path.pop_back();
   };
 
-  if (ObjectArg && implicitObjectParamIsLifetimeBound(Callee))
-    VisitLifetimeBoundArg(Callee, ObjectArg);
-
   bool CheckCoroCall = false;
   if (const auto *RD = Callee->getReturnType()->getAsRecordDecl()) {
     CheckCoroCall = RD->hasAttr() &&
                     RD->hasAttr() &&
                     !Callee->hasAttr();
   }
+
+  if (ObjectArg) {
+    bool CheckCoroObjArg = CheckCoroCall;
+    // Coroutine lambda objects with empty capture list are not lifetimebound.
+    if (auto *LE = dyn_cast(ObjectArg->IgnoreImplicit());
+        LE && LE->captures().empty())
+      CheckCoroObjArg = false;
+    // Allow `get_return_object()` as the object param (__promise) is not
+    // lifetimebound.
+    if (Sema::CanBeGetReturnObject(Callee))
+      CheckCoroObjArg = false;
+    if (implicitObjectParamIsLifetimeBound(Callee) || CheckCoroObjArg)
+      VisitLifetimeBoundArg(Callee, ObjectArg);
+  }
+
   for (unsigned I = 0,
                 N = std::min(Callee->getNumParams(), Args.size());
        I != N; ++I) {
diff --git a/clang/lib/Sema/SemaLambda.cpp b/clang/lib/Sema/SemaLambda.cpp
index 536552bb23f6b..73cd730af112f 100644
--- a/clang/lib/Sema/SemaLambda.cpp
+++ b/clang/lib/Sema/SemaLambda.cpp
@@ -916,7 +916,7 @@ static TypeSourceInfo *getLambdaType(Sema &S, LambdaIntroducer &Intro,
       }
     }
 
-    MethodTyInfo = S.GetTypeForDeclarator(ParamInfo, CurScope);
+    MethodTyInfo = S.GetTypeForDeclarator(ParamInfo);
     assert(MethodTyInfo && "no type from lambda-declarator");
 
     // Check for unexpanded parameter packs in the method type.
diff --git a/clang/lib/Sema/SemaObjCProperty.cpp b/clang/lib/Sema/SemaObjCProperty.cpp
index 22540af1cda8c..349c7fc9c91bd 100644
--- a/clang/lib/Sema/SemaObjCProperty.cpp
+++ b/clang/lib/Sema/SemaObjCProperty.cpp
@@ -180,7 +180,7 @@ Decl *Sema::ActOnProperty(Scope *S, SourceLocation AtLoc,
   unsigned Attributes = ODS.getPropertyAttributes();
   FD.D.setObjCWeakProperty((Attributes & ObjCPropertyAttribute::kind_weak) !=
                            0);
-  TypeSourceInfo *TSI = GetTypeForDeclarator(FD.D, S);
+  TypeSourceInfo *TSI = GetTypeForDeclarator(FD.D);
   QualType T = TSI->getType();
   if (!getOwnershipRule(Attributes)) {
     Attributes |= deducePropertyOwnershipFromType(*this, T);
diff --git a/clang/lib/Sema/SemaOpenMP.cpp b/clang/lib/Sema/SemaOpenMP.cpp
index 61780d0f67b34..5f2000c331b32 100644
--- a/clang/lib/Sema/SemaOpenMP.cpp
+++ b/clang/lib/Sema/SemaOpenMP.cpp
@@ -7306,7 +7306,7 @@ void Sema::ActOnStartOfFunctionDefinitionInOpenMPDeclareVariantScope(
                       LookupOrdinaryName);
   LookupParsedName(Lookup, S, &D.getCXXScopeSpec());
 
-  TypeSourceInfo *TInfo = GetTypeForDeclarator(D, S);
+  TypeSourceInfo *TInfo = GetTypeForDeclarator(D);
   QualType FType = TInfo->getType();
 
   bool IsConstexpr =
@@ -22721,7 +22721,7 @@ Sema::DeclGroupPtrTy Sema::ActOnOpenMPDeclareReductionDirectiveEnd(
 }
 
 TypeResult Sema::ActOnOpenMPDeclareMapperVarDecl(Scope *S, Declarator &D) {
-  TypeSourceInfo *TInfo = GetTypeForDeclarator(D, S);
+  TypeSourceInfo *TInfo = GetTypeForDeclarator(D);
   QualType T = TInfo->getType();
   if (D.isInvalidType())
     return true;
diff --git a/clang/lib/Sema/SemaOverload.cpp b/clang/lib/Sema/SemaOverload.cpp
index f89555321a568..5ae75cecebbc5 100644
--- a/clang/lib/Sema/SemaOverload.cpp
+++ b/clang/lib/Sema/SemaOverload.cpp
@@ -1259,6 +1259,40 @@ static bool IsOverloadOrOverrideImpl(Sema &SemaRef, FunctionDecl *New,
   if ((OldTemplate == nullptr) != (NewTemplate == nullptr))
     return true;
 
+  // Is the function New an overload of the function Old?
+  QualType OldQType = SemaRef.Context.getCanonicalType(Old->getType());
+  QualType NewQType = SemaRef.Context.getCanonicalType(New->getType());
+
+  // Compare the signatures (C++ 1.3.10) of the two functions to
+  // determine whether they are overloads. If we find any mismatch
+  // in the signature, they are overloads.
+
+  // If either of these functions is a K&R-style function (no
+  // prototype), then we consider them to have matching signatures.
+  if (isa(OldQType.getTypePtr()) ||
+      isa(NewQType.getTypePtr()))
+    return false;
+
+  const auto *OldType = cast(OldQType);
+  const auto *NewType = cast(NewQType);
+
+  // The signature of a function includes the types of its
+  // parameters (C++ 1.3.10), which includes the presence or absence
+  // of the ellipsis; see C++ DR 357).
+  if (OldQType != NewQType && OldType->isVariadic() != NewType->isVariadic())
+    return true;
+
+  // For member-like friends, the enclosing class is part of the signature.
+  if ((New->isMemberLikeConstrainedFriend() ||
+       Old->isMemberLikeConstrainedFriend()) &&
+      !New->getLexicalDeclContext()->Equals(Old->getLexicalDeclContext()))
+    return true;
+
+  // Compare the parameter lists.
+  // This can only be done once we have establish that friend functions
+  // inhabit the same context, otherwise we might tried to instantiate
+  // references to non-instantiated entities during constraint substitution.
+  // GH78101.
   if (NewTemplate) {
     // C++ [temp.over.link]p4:
     //   The signature of a function template consists of its function
@@ -1296,34 +1330,6 @@ static bool IsOverloadOrOverrideImpl(Sema &SemaRef, FunctionDecl *New,
       return true;
   }
 
-  // Is the function New an overload of the function Old?
-  QualType OldQType = SemaRef.Context.getCanonicalType(Old->getType());
-  QualType NewQType = SemaRef.Context.getCanonicalType(New->getType());
-
-  // Compare the signatures (C++ 1.3.10) of the two functions to
-  // determine whether they are overloads. If we find any mismatch
-  // in the signature, they are overloads.
-
-  // If either of these functions is a K&R-style function (no
-  // prototype), then we consider them to have matching signatures.
-  if (isa(OldQType.getTypePtr()) ||
-      isa(NewQType.getTypePtr()))
-    return false;
-
-  const FunctionProtoType *OldType = cast(OldQType);
-  const FunctionProtoType *NewType = cast(NewQType);
-
-  // The signature of a function includes the types of its
-  // parameters (C++ 1.3.10), which includes the presence or absence
-  // of the ellipsis; see C++ DR 357).
-  if (OldQType != NewQType && OldType->isVariadic() != NewType->isVariadic())
-    return true;
-
-  // For member-like friends, the enclosing class is part of the signature.
-  if ((New->isMemberLikeConstrainedFriend() ||
-       Old->isMemberLikeConstrainedFriend()) &&
-      !New->getLexicalDeclContext()->Equals(Old->getLexicalDeclContext()))
-    return true;
   const auto *OldMethod = dyn_cast(Old);
   const auto *NewMethod = dyn_cast(New);
 
@@ -1784,26 +1790,6 @@ bool Sema::IsFunctionConversion(QualType FromType, QualType ToType,
     Changed = true;
   }
 
-  // Drop the 'arm_preserves_za' if not present in the target type (we can do
-  // that because it is merely a hint).
-  if (const auto *FromFPT = dyn_cast(FromFn)) {
-    FunctionProtoType::ExtProtoInfo ExtInfo = FromFPT->getExtProtoInfo();
-    if (ExtInfo.AArch64SMEAttributes &
-        FunctionType::SME_PStateZAPreservedMask) {
-      unsigned ToFlags = 0;
-      if (const auto *ToFPT = dyn_cast(ToFn))
-        ToFlags = ToFPT->getExtProtoInfo().AArch64SMEAttributes;
-      if (!(ToFlags & FunctionType::SME_PStateZAPreservedMask)) {
-        ExtInfo.setArmSMEAttribute(FunctionType::SME_PStateZAPreservedMask,
-                                   false);
-        QualType QT = Context.getFunctionType(
-            FromFPT->getReturnType(), FromFPT->getParamTypes(), ExtInfo);
-        FromFn = QT->getAs();
-        Changed = true;
-      }
-    }
-  }
-
   // Drop 'noexcept' if not present in target type.
   if (const auto *FromFPT = dyn_cast(FromFn)) {
     const auto *ToFPT = cast(ToFn);
@@ -7740,9 +7726,19 @@ bool Sema::CheckNonDependentConversions(
        ++I) {
     QualType ParamType = ParamTypes[I + Offset];
     if (!ParamType->isDependentType()) {
-      unsigned ConvIdx = PO == OverloadCandidateParamOrder::Reversed
-                             ? 0
-                             : (ThisConversions + I);
+      unsigned ConvIdx;
+      if (PO == OverloadCandidateParamOrder::Reversed) {
+        ConvIdx = Args.size() - 1 - I;
+        assert(Args.size() + ThisConversions == 2 &&
+               "number of args (including 'this') must be exactly 2 for "
+               "reversed order");
+        // For members, there would be only one arg 'Args[0]' whose ConvIdx
+        // would also be 0. 'this' got ConvIdx = 1 previously.
+        assert(!HasThisConversion || (ConvIdx == 0 && I == 0));
+      } else {
+        // For members, 'this' got ConvIdx = 0 previously.
+        ConvIdx = ThisConversions + I;
+      }
       Conversions[ConvIdx]
         = TryCopyInitialization(*this, Args[I], ParamType,
                                 SuppressUserConversions,
@@ -10138,11 +10134,23 @@ getImplicitObjectParamType(ASTContext &Context, const FunctionDecl *F) {
   return M->getFunctionObjectParameterReferenceType();
 }
 
-static bool haveSameParameterTypes(ASTContext &Context, const FunctionDecl *F1,
-                                   const FunctionDecl *F2) {
+// As a Clang extension, allow ambiguity among F1 and F2 if they represent
+// represent the same entity.
+static bool allowAmbiguity(ASTContext &Context, const FunctionDecl *F1,
+                           const FunctionDecl *F2) {
   if (declaresSameEntity(F1, F2))
     return true;
-
+  auto PT1 = F1->getPrimaryTemplate();
+  auto PT2 = F2->getPrimaryTemplate();
+  if (PT1 && PT2) {
+    if (declaresSameEntity(PT1, PT2) ||
+        declaresSameEntity(PT1->getInstantiatedFromMemberTemplate(),
+                           PT2->getInstantiatedFromMemberTemplate()))
+      return true;
+  }
+  // TODO: It is not clear whether comparing parameters is necessary (i.e.
+  // different functions with same params). Consider removing this (as no test
+  // fail w/o it).
   auto NextParam = [&](const FunctionDecl *F, unsigned &I, bool First) {
     if (First) {
       if (std::optional T = getImplicitObjectParamType(Context, F))
@@ -10346,14 +10354,14 @@ bool clang::isBetterOverloadCandidate(
     case ImplicitConversionSequence::Worse:
       if (Cand1.Function && Cand2.Function &&
           Cand1.isReversed() != Cand2.isReversed() &&
-          haveSameParameterTypes(S.Context, Cand1.Function, Cand2.Function)) {
+          allowAmbiguity(S.Context, Cand1.Function, Cand2.Function)) {
         // Work around large-scale breakage caused by considering reversed
         // forms of operator== in C++20:
         //
-        // When comparing a function against a reversed function with the same
-        // parameter types, if we have a better conversion for one argument and
-        // a worse conversion for the other, the implicit conversion sequences
-        // are treated as being equally good.
+        // When comparing a function against a reversed function, if we have a
+        // better conversion for one argument and a worse conversion for the
+        // other, the implicit conversion sequences are treated as being equally
+        // good.
         //
         // This prevents a comparison function from being considered ambiguous
         // with a reversed form that is written in the same way.
@@ -14552,7 +14560,7 @@ ExprResult Sema::CreateOverloadedBinOp(SourceLocation OpLoc,
           llvm::SmallVector AmbiguousWith;
           for (OverloadCandidate &Cand : CandidateSet) {
             if (Cand.Viable && Cand.Function && Cand.isReversed() &&
-                haveSameParameterTypes(Context, Cand.Function, FnDecl)) {
+                allowAmbiguity(Context, Cand.Function, FnDecl)) {
               for (unsigned ArgIdx = 0; ArgIdx < 2; ++ArgIdx) {
                 if (CompareImplicitConversionSequences(
                         *this, OpLoc, Cand.Conversions[ArgIdx],
@@ -15490,7 +15498,7 @@ ExprResult Sema::BuildCallToMemberFunction(Scope *S, Expr *MemExprE,
   }
 
   if (isa(CurContext) &&
-      TheCall->getDirectCallee()->isPure()) {
+      TheCall->getDirectCallee()->isPureVirtual()) {
     const FunctionDecl *MD = TheCall->getDirectCallee();
 
     if (isa(MemExpr->getBase()->IgnoreParenCasts()) &&
diff --git a/clang/lib/Sema/SemaRISCVVectorLookup.cpp b/clang/lib/Sema/SemaRISCVVectorLookup.cpp
index 3ed3e61954418..00a5ea65f3f4e 100644
--- a/clang/lib/Sema/SemaRISCVVectorLookup.cpp
+++ b/clang/lib/Sema/SemaRISCVVectorLookup.cpp
@@ -274,9 +274,8 @@ void RISCVIntrinsicManagerImpl::ConstructRVVIntrinsics(
         continue;
 
       if (BaseType == BasicType::Float16) {
-        if ((Record.RequiredExtensions & RVV_REQ_ZvfhminOrZvfh) ==
-            RVV_REQ_ZvfhminOrZvfh) {
-          if (!TI.hasFeature("zvfh") && !TI.hasFeature("zvfhmin"))
+        if ((Record.RequiredExtensions & RVV_REQ_Zvfhmin) == RVV_REQ_Zvfhmin) {
+          if (!TI.hasFeature("zvfhmin"))
             continue;
         } else if (!TI.hasFeature("zvfh")) {
           continue;
diff --git a/clang/lib/Sema/SemaStmt.cpp b/clang/lib/Sema/SemaStmt.cpp
index 24355b1a3f78a..15958613d52b5 100644
--- a/clang/lib/Sema/SemaStmt.cpp
+++ b/clang/lib/Sema/SemaStmt.cpp
@@ -786,6 +786,12 @@ bool Sema::checkMustTailAttr(const Stmt *St, const Attr &MTA) {
     return false;
   }
 
+  const auto *CalleeDecl = CE->getCalleeDecl();
+  if (CalleeDecl && CalleeDecl->hasAttr()) {
+    Diag(St->getBeginLoc(), diag::err_musttail_no_return) << &MTA;
+    return false;
+  }
+
   // Caller and callee must match in whether they have a "this" parameter.
   if (CallerType.This.isNull() != CalleeType.This.isNull()) {
     if (const auto *ND = dyn_cast_or_null(CE->getCalleeDecl())) {
diff --git a/clang/lib/Sema/SemaTemplate.cpp b/clang/lib/Sema/SemaTemplate.cpp
index 5fcc39ec70052..0655d36335206 100644
--- a/clang/lib/Sema/SemaTemplate.cpp
+++ b/clang/lib/Sema/SemaTemplate.cpp
@@ -1498,7 +1498,7 @@ NamedDecl *Sema::ActOnNonTypeTemplateParameter(Scope *S, Declarator &D,
                                           unsigned Position,
                                           SourceLocation EqualLoc,
                                           Expr *Default) {
-  TypeSourceInfo *TInfo = GetTypeForDeclarator(D, S);
+  TypeSourceInfo *TInfo = GetTypeForDeclarator(D);
 
   // Check that we have valid decl-specifiers specified.
   auto CheckValidDeclSpecifiers = [this, &D] {
@@ -2587,15 +2587,15 @@ struct ConvertConstructorToDeductionGuideTransform {
                           : ParamTy->isRValueReferenceType() ? VK_XValue
                                                              : VK_PRValue);
     }
+    // Handle arrays and functions decay.
+    auto NewType = NewDI->getType();
+    if (NewType->isArrayType() || NewType->isFunctionType())
+      NewType = SemaRef.Context.getDecayedType(NewType);
 
-    ParmVarDecl *NewParam = ParmVarDecl::Create(SemaRef.Context, DC,
-                                                OldParam->getInnerLocStart(),
-                                                OldParam->getLocation(),
-                                                OldParam->getIdentifier(),
-                                                NewDI->getType(),
-                                                NewDI,
-                                                OldParam->getStorageClass(),
-                                                NewDefArg.get());
+    ParmVarDecl *NewParam = ParmVarDecl::Create(
+        SemaRef.Context, DC, OldParam->getInnerLocStart(),
+        OldParam->getLocation(), OldParam->getIdentifier(), NewType, NewDI,
+        OldParam->getStorageClass(), NewDefArg.get());
     NewParam->setScopeInfo(OldParam->getFunctionScopeDepth(),
                            OldParam->getFunctionScopeIndex());
     SemaRef.CurrentInstantiationScope->InstantiatedLocal(OldParam, NewParam);
@@ -9229,10 +9229,8 @@ void Sema::CheckConceptRedefinition(ConceptDecl *NewDecl,
 /// that has just been explicitly specialized.
 static void StripImplicitInstantiation(NamedDecl *D, bool MinGW) {
   if (MinGW || (isa(D) &&
-                cast(D)->isFunctionTemplateSpecialization())) {
-    D->dropAttr();
-    D->dropAttr();
-  }
+                cast(D)->isFunctionTemplateSpecialization()))
+    D->dropAttrs();
 
   if (FunctionDecl *FD = dyn_cast(D))
     FD->setInlineSpecified(false);
@@ -10521,7 +10519,7 @@ DeclResult Sema::ActOnExplicitInstantiation(Scope *S,
     S = S->getParent();
 
   // Determine the type of the declaration.
-  TypeSourceInfo *T = GetTypeForDeclarator(D, S);
+  TypeSourceInfo *T = GetTypeForDeclarator(D);
   QualType R = T->getType();
   if (R.isNull())
     return true;
diff --git a/clang/lib/Sema/SemaTemplateInstantiate.cpp b/clang/lib/Sema/SemaTemplateInstantiate.cpp
index f5065a83dfce7..476c57d51ffce 100644
--- a/clang/lib/Sema/SemaTemplateInstantiate.cpp
+++ b/clang/lib/Sema/SemaTemplateInstantiate.cpp
@@ -223,6 +223,9 @@ Response HandleFunction(const FunctionDecl *Function,
       (!Pattern || !Pattern->getLexicalDeclContext()->isFileContext())) {
     return Response::ChangeDecl(Function->getLexicalDeclContext());
   }
+
+  if (ForConstraintInstantiation && Function->getFriendObjectKind())
+    return Response::ChangeDecl(Function->getLexicalDeclContext());
   return Response::UseNextDecl(Function);
 }
 
diff --git a/clang/lib/Sema/SemaTemplateInstantiateDecl.cpp b/clang/lib/Sema/SemaTemplateInstantiateDecl.cpp
index a343280a0c6db..77f1828aec36a 100644
--- a/clang/lib/Sema/SemaTemplateInstantiateDecl.cpp
+++ b/clang/lib/Sema/SemaTemplateInstantiateDecl.cpp
@@ -3272,7 +3272,7 @@ Decl *TemplateDeclInstantiator::VisitCXXMethodDecl(
                                    IsExplicitSpecialization,
                                    Method->isThisDeclarationADefinition());
 
-  if (D->isPure())
+  if (D->isPureVirtual())
     SemaRef.CheckPureMethod(Method, SourceRange());
 
   // Propagate access.  For a non-friend declaration, the access is
@@ -3573,16 +3573,21 @@ Decl *TemplateDeclInstantiator::VisitNonTypeTemplateParmDecl(
         D->getPosition(), D->getIdentifier(), T, D->isParameterPack(), DI);
 
   if (AutoTypeLoc AutoLoc = DI->getTypeLoc().getContainedAutoTypeLoc())
-    if (AutoLoc.isConstrained())
+    if (AutoLoc.isConstrained()) {
+      SourceLocation EllipsisLoc;
+      if (IsExpandedParameterPack)
+        EllipsisLoc =
+            DI->getTypeLoc().getAs().getEllipsisLoc();
+      else if (auto *Constraint = dyn_cast_if_present(
+                   D->getPlaceholderTypeConstraint()))
+        EllipsisLoc = Constraint->getEllipsisLoc();
       // Note: We attach the uninstantiated constriant here, so that it can be
-      // instantiated relative to the top level, like all our other constraints.
-      if (SemaRef.AttachTypeConstraint(
-              AutoLoc, Param, D,
-              IsExpandedParameterPack
-                ? DI->getTypeLoc().getAs()
-                    .getEllipsisLoc()
-                : SourceLocation()))
+      // instantiated relative to the top level, like all our other
+      // constraints.
+      if (SemaRef.AttachTypeConstraint(AutoLoc, /*NewConstrainedParm=*/Param,
+                                       /*OrigConstrainedParm=*/D, EllipsisLoc))
         Invalid = true;
+    }
 
   Param->setAccess(AS_public);
   Param->setImplicit(D->isImplicit());
diff --git a/clang/lib/Sema/SemaType.cpp b/clang/lib/Sema/SemaType.cpp
index 568cfc89257b8..55f12ca9ed788 100644
--- a/clang/lib/Sema/SemaType.cpp
+++ b/clang/lib/Sema/SemaType.cpp
@@ -147,8 +147,10 @@ static void diagnoseBadTypeAttribute(Sema &S, const ParsedAttr &attr,
   case ParsedAttr::AT_CmseNSCall:                                              \
   case ParsedAttr::AT_ArmStreaming:                                            \
   case ParsedAttr::AT_ArmStreamingCompatible:                                  \
-  case ParsedAttr::AT_ArmSharedZA:                                             \
-  case ParsedAttr::AT_ArmPreservesZA:                                          \
+  case ParsedAttr::AT_ArmPreserves:                                            \
+  case ParsedAttr::AT_ArmIn:                                                   \
+  case ParsedAttr::AT_ArmOut:                                                  \
+  case ParsedAttr::AT_ArmInOut:                                                \
   case ParsedAttr::AT_AnyX86NoCallerSavedRegisters:                            \
   case ParsedAttr::AT_AnyX86NoCfCheck:                                         \
     CALLING_CONV_ATTRS_CASELIST
@@ -6094,7 +6096,7 @@ static TypeSourceInfo *GetFullTypeForDeclarator(TypeProcessingState &state,
 ///
 /// The result of this call will never be null, but the associated
 /// type may be a null type if there's an unrecoverable error.
-TypeSourceInfo *Sema::GetTypeForDeclarator(Declarator &D, Scope *S) {
+TypeSourceInfo *Sema::GetTypeForDeclarator(Declarator &D) {
   // Determine the type of the declarator. Not all forms of declarator
   // have a type.
 
@@ -6772,13 +6774,13 @@ void LocInfoType::getAsStringInternal(std::string &Str,
          " GetTypeFromParser");
 }
 
-TypeResult Sema::ActOnTypeName(Scope *S, Declarator &D) {
+TypeResult Sema::ActOnTypeName(Declarator &D) {
   // C99 6.7.6: Type names have no identifier.  This is already validated by
   // the parser.
   assert(D.getIdentifier() == nullptr &&
          "Type name should have no identifier!");
 
-  TypeSourceInfo *TInfo = GetTypeForDeclarator(D, S);
+  TypeSourceInfo *TInfo = GetTypeForDeclarator(D);
   QualType T = TInfo->getType();
   if (D.isInvalidType())
     return true;
@@ -7605,6 +7607,25 @@ static bool HandleWebAssemblyFuncrefAttr(TypeProcessingState &State,
   return false;
 }
 
+/// Rebuild an attributed type without the nullability attribute on it.
+static QualType rebuildAttributedTypeWithoutNullability(ASTContext &Ctx,
+                                                        QualType Type) {
+  auto Attributed = dyn_cast(Type.getTypePtr());
+  if (!Attributed)
+    return Type;
+
+  // Skip the nullability attribute; we're done.
+  if (Attributed->getImmediateNullability())
+    return Attributed->getModifiedType();
+
+  // Build the modified type.
+  QualType Modified = rebuildAttributedTypeWithoutNullability(
+      Ctx, Attributed->getModifiedType());
+  assert(Modified.getTypePtr() != Attributed->getModifiedType().getTypePtr());
+  return Ctx.getAttributedType(Attributed->getAttrKind(), Modified,
+                               Attributed->getEquivalentType());
+}
+
 /// Map a nullability attribute kind to a nullability kind.
 static NullabilityKind mapNullabilityAttrKind(ParsedAttr::Kind kind) {
   switch (kind) {
@@ -7625,74 +7646,65 @@ static NullabilityKind mapNullabilityAttrKind(ParsedAttr::Kind kind) {
   }
 }
 
-/// Applies a nullability type specifier to the given type, if possible.
-///
-/// \param state The type processing state.
-///
-/// \param type The type to which the nullability specifier will be
-/// added. On success, this type will be updated appropriately.
-///
-/// \param attr The attribute as written on the type.
-///
-/// \param allowOnArrayType Whether to accept nullability specifiers on an
-/// array type (e.g., because it will decay to a pointer).
-///
-/// \returns true if a problem has been diagnosed, false on success.
-static bool checkNullabilityTypeSpecifier(TypeProcessingState &state,
-                                          QualType &type,
-                                          ParsedAttr &attr,
-                                          bool allowOnArrayType) {
-  Sema &S = state.getSema();
-
-  NullabilityKind nullability = mapNullabilityAttrKind(attr.getKind());
-  SourceLocation nullabilityLoc = attr.getLoc();
-  bool isContextSensitive = attr.isContextSensitiveKeywordAttribute();
-
-  recordNullabilitySeen(S, nullabilityLoc);
+static bool CheckNullabilityTypeSpecifier(
+    Sema &S, TypeProcessingState *State, ParsedAttr *PAttr, QualType &QT,
+    NullabilityKind Nullability, SourceLocation NullabilityLoc,
+    bool IsContextSensitive, bool AllowOnArrayType, bool OverrideExisting) {
+  bool Implicit = (State == nullptr);
+  if (!Implicit)
+    recordNullabilitySeen(S, NullabilityLoc);
 
   // Check for existing nullability attributes on the type.
-  QualType desugared = type;
-  while (auto attributed = dyn_cast(desugared.getTypePtr())) {
+  QualType Desugared = QT;
+  while (auto *Attributed = dyn_cast(Desugared.getTypePtr())) {
     // Check whether there is already a null
-    if (auto existingNullability = attributed->getImmediateNullability()) {
+    if (auto ExistingNullability = Attributed->getImmediateNullability()) {
       // Duplicated nullability.
-      if (nullability == *existingNullability) {
-        S.Diag(nullabilityLoc, diag::warn_nullability_duplicate)
-          << DiagNullabilityKind(nullability, isContextSensitive)
-          << FixItHint::CreateRemoval(nullabilityLoc);
+      if (Nullability == *ExistingNullability) {
+        if (Implicit)
+          break;
+
+        S.Diag(NullabilityLoc, diag::warn_nullability_duplicate)
+            << DiagNullabilityKind(Nullability, IsContextSensitive)
+            << FixItHint::CreateRemoval(NullabilityLoc);
 
         break;
       }
 
-      // Conflicting nullability.
-      S.Diag(nullabilityLoc, diag::err_nullability_conflicting)
-        << DiagNullabilityKind(nullability, isContextSensitive)
-        << DiagNullabilityKind(*existingNullability, false);
-      return true;
+      if (!OverrideExisting) {
+        // Conflicting nullability.
+        S.Diag(NullabilityLoc, diag::err_nullability_conflicting)
+            << DiagNullabilityKind(Nullability, IsContextSensitive)
+            << DiagNullabilityKind(*ExistingNullability, false);
+        return true;
+      }
+
+      // Rebuild the attributed type, dropping the existing nullability.
+      QT = rebuildAttributedTypeWithoutNullability(S.Context, QT);
     }
 
-    desugared = attributed->getModifiedType();
+    Desugared = Attributed->getModifiedType();
   }
 
   // If there is already a different nullability specifier, complain.
   // This (unlike the code above) looks through typedefs that might
   // have nullability specifiers on them, which means we cannot
   // provide a useful Fix-It.
-  if (auto existingNullability = desugared->getNullability()) {
-    if (nullability != *existingNullability) {
-      S.Diag(nullabilityLoc, diag::err_nullability_conflicting)
-        << DiagNullabilityKind(nullability, isContextSensitive)
-        << DiagNullabilityKind(*existingNullability, false);
+  if (auto ExistingNullability = Desugared->getNullability()) {
+    if (Nullability != *ExistingNullability && !Implicit) {
+      S.Diag(NullabilityLoc, diag::err_nullability_conflicting)
+          << DiagNullabilityKind(Nullability, IsContextSensitive)
+          << DiagNullabilityKind(*ExistingNullability, false);
 
       // Try to find the typedef with the existing nullability specifier.
-      if (auto typedefType = desugared->getAs()) {
-        TypedefNameDecl *typedefDecl = typedefType->getDecl();
+      if (auto TT = Desugared->getAs()) {
+        TypedefNameDecl *typedefDecl = TT->getDecl();
         QualType underlyingType = typedefDecl->getUnderlyingType();
-        if (auto typedefNullability
-              = AttributedType::stripOuterNullability(underlyingType)) {
-          if (*typedefNullability == *existingNullability) {
+        if (auto typedefNullability =
+                AttributedType::stripOuterNullability(underlyingType)) {
+          if (*typedefNullability == *ExistingNullability) {
             S.Diag(typedefDecl->getLocation(), diag::note_nullability_here)
-              << DiagNullabilityKind(*existingNullability, false);
+                << DiagNullabilityKind(*ExistingNullability, false);
           }
         }
       }
@@ -7702,44 +7714,73 @@ static bool checkNullabilityTypeSpecifier(TypeProcessingState &state,
   }
 
   // If this definitely isn't a pointer type, reject the specifier.
-  if (!desugared->canHaveNullability() &&
-      !(allowOnArrayType && desugared->isArrayType())) {
-    S.Diag(nullabilityLoc, diag::err_nullability_nonpointer)
-      << DiagNullabilityKind(nullability, isContextSensitive) << type;
+  if (!Desugared->canHaveNullability() &&
+      !(AllowOnArrayType && Desugared->isArrayType())) {
+    if (!Implicit)
+      S.Diag(NullabilityLoc, diag::err_nullability_nonpointer)
+          << DiagNullabilityKind(Nullability, IsContextSensitive) << QT;
+
     return true;
   }
 
   // For the context-sensitive keywords/Objective-C property
   // attributes, require that the type be a single-level pointer.
-  if (isContextSensitive) {
+  if (IsContextSensitive) {
     // Make sure that the pointee isn't itself a pointer type.
     const Type *pointeeType = nullptr;
-    if (desugared->isArrayType())
-      pointeeType = desugared->getArrayElementTypeNoTypeQual();
-    else if (desugared->isAnyPointerType())
-      pointeeType = desugared->getPointeeType().getTypePtr();
+    if (Desugared->isArrayType())
+      pointeeType = Desugared->getArrayElementTypeNoTypeQual();
+    else if (Desugared->isAnyPointerType())
+      pointeeType = Desugared->getPointeeType().getTypePtr();
 
     if (pointeeType && (pointeeType->isAnyPointerType() ||
                         pointeeType->isObjCObjectPointerType() ||
                         pointeeType->isMemberPointerType())) {
-      S.Diag(nullabilityLoc, diag::err_nullability_cs_multilevel)
-        << DiagNullabilityKind(nullability, true)
-        << type;
-      S.Diag(nullabilityLoc, diag::note_nullability_type_specifier)
-        << DiagNullabilityKind(nullability, false)
-        << type
-        << FixItHint::CreateReplacement(nullabilityLoc,
-                                        getNullabilitySpelling(nullability));
+      S.Diag(NullabilityLoc, diag::err_nullability_cs_multilevel)
+          << DiagNullabilityKind(Nullability, true) << QT;
+      S.Diag(NullabilityLoc, diag::note_nullability_type_specifier)
+          << DiagNullabilityKind(Nullability, false) << QT
+          << FixItHint::CreateReplacement(NullabilityLoc,
+                                          getNullabilitySpelling(Nullability));
       return true;
     }
   }
 
   // Form the attributed type.
-  type = state.getAttributedType(
-      createNullabilityAttr(S.Context, attr, nullability), type, type);
+  if (State) {
+    assert(PAttr);
+    Attr *A = createNullabilityAttr(S.Context, *PAttr, Nullability);
+    QT = State->getAttributedType(A, QT, QT);
+  } else {
+    attr::Kind attrKind = AttributedType::getNullabilityAttrKind(Nullability);
+    QT = S.Context.getAttributedType(attrKind, QT, QT);
+  }
   return false;
 }
 
+static bool CheckNullabilityTypeSpecifier(TypeProcessingState &State,
+                                          QualType &Type, ParsedAttr &Attr,
+                                          bool AllowOnArrayType) {
+  NullabilityKind Nullability = mapNullabilityAttrKind(Attr.getKind());
+  SourceLocation NullabilityLoc = Attr.getLoc();
+  bool IsContextSensitive = Attr.isContextSensitiveKeywordAttribute();
+
+  return CheckNullabilityTypeSpecifier(State.getSema(), &State, &Attr, Type,
+                                       Nullability, NullabilityLoc,
+                                       IsContextSensitive, AllowOnArrayType,
+                                       /*overrideExisting*/ false);
+}
+
+bool Sema::CheckImplicitNullabilityTypeSpecifier(QualType &Type,
+                                                 NullabilityKind Nullability,
+                                                 SourceLocation DiagLoc,
+                                                 bool AllowArrayTypes,
+                                                 bool OverrideExisting) {
+  return CheckNullabilityTypeSpecifier(
+      *this, nullptr, nullptr, Type, Nullability, DiagLoc,
+      /*isContextSensitive*/ false, AllowArrayTypes, OverrideExisting);
+}
+
 /// Check the application of the Objective-C '__kindof' qualifier to
 /// the given type.
 static bool checkObjCKindOfType(TypeProcessingState &state, QualType &type,
@@ -7962,6 +8003,49 @@ static bool checkMutualExclusion(TypeProcessingState &state,
   return true;
 }
 
+static bool handleArmStateAttribute(Sema &S,
+                                    FunctionProtoType::ExtProtoInfo &EPI,
+                                    ParsedAttr &Attr,
+                                    FunctionType::ArmStateValue State) {
+  if (!Attr.getNumArgs()) {
+    S.Diag(Attr.getLoc(), diag::err_missing_arm_state) << Attr;
+    Attr.setInvalid();
+    return true;
+  }
+
+  for (unsigned I = 0; I < Attr.getNumArgs(); ++I) {
+    StringRef StateName;
+    SourceLocation LiteralLoc;
+    if (!S.checkStringLiteralArgumentAttr(Attr, I, StateName, &LiteralLoc))
+      return true;
+
+    unsigned Shift;
+    FunctionType::ArmStateValue ExistingState;
+    if (StateName == "za") {
+      Shift = FunctionType::SME_ZAShift;
+      ExistingState = FunctionType::getArmZAState(EPI.AArch64SMEAttributes);
+    } else {
+      S.Diag(LiteralLoc, diag::err_unknown_arm_state) << StateName;
+      Attr.setInvalid();
+      return true;
+    }
+
+    // __arm_in(S), __arm_out(S), __arm_inout(S) and __arm_preserves(S)
+    // are all mutually exclusive for the same S, so check if there are
+    // conflicting attributes.
+    if (ExistingState != FunctionType::ARM_None && ExistingState != State) {
+      S.Diag(LiteralLoc, diag::err_conflicting_attributes_arm_state)
+          << StateName;
+      Attr.setInvalid();
+      return true;
+    }
+
+    EPI.setArmSMEAttribute(
+        (FunctionType::AArch64SMETypeAttributes)((State << Shift)));
+  }
+  return false;
+}
+
 /// Process an individual function attribute.  Returns true to
 /// indicate that the attribute was handled, false if it wasn't.
 static bool handleFunctionTypeAttr(TypeProcessingState &state, ParsedAttr &attr,
@@ -8094,11 +8178,18 @@ static bool handleFunctionTypeAttr(TypeProcessingState &state, ParsedAttr &attr,
 
   if (attr.getKind() == ParsedAttr::AT_ArmStreaming ||
       attr.getKind() == ParsedAttr::AT_ArmStreamingCompatible ||
-      attr.getKind() == ParsedAttr::AT_ArmSharedZA ||
-      attr.getKind() == ParsedAttr::AT_ArmPreservesZA){
-    if (S.CheckAttrTarget(attr) || S.CheckAttrNoArgs(attr))
+      attr.getKind() == ParsedAttr::AT_ArmPreserves ||
+      attr.getKind() == ParsedAttr::AT_ArmIn ||
+      attr.getKind() == ParsedAttr::AT_ArmOut ||
+      attr.getKind() == ParsedAttr::AT_ArmInOut) {
+    if (S.CheckAttrTarget(attr))
       return true;
 
+    if (attr.getKind() == ParsedAttr::AT_ArmStreaming ||
+        attr.getKind() == ParsedAttr::AT_ArmStreamingCompatible)
+      if (S.CheckAttrNoArgs(attr))
+        return true;
+
     if (!unwrapped.isFunctionType())
       return false;
 
@@ -8125,11 +8216,21 @@ static bool handleFunctionTypeAttr(TypeProcessingState &state, ParsedAttr &attr,
         return true;
       EPI.setArmSMEAttribute(FunctionType::SME_PStateSMCompatibleMask);
       break;
-    case ParsedAttr::AT_ArmSharedZA:
-      EPI.setArmSMEAttribute(FunctionType::SME_PStateZASharedMask);
+    case ParsedAttr::AT_ArmPreserves:
+      if (handleArmStateAttribute(S, EPI, attr, FunctionType::ARM_Preserves))
+        return true;
+      break;
+    case ParsedAttr::AT_ArmIn:
+      if (handleArmStateAttribute(S, EPI, attr, FunctionType::ARM_In))
+        return true;
       break;
-    case ParsedAttr::AT_ArmPreservesZA:
-      EPI.setArmSMEAttribute(FunctionType::SME_PStateZAPreservedMask);
+    case ParsedAttr::AT_ArmOut:
+      if (handleArmStateAttribute(S, EPI, attr, FunctionType::ARM_Out))
+        return true;
+      break;
+    case ParsedAttr::AT_ArmInOut:
+      if (handleArmStateAttribute(S, EPI, attr, FunctionType::ARM_InOut))
+        return true;
       break;
     default:
       llvm_unreachable("Unsupported attribute");
@@ -8980,11 +9081,8 @@ static void processTypeAttrs(TypeProcessingState &state, QualType &type,
         bool allowOnArrayType =
             state.getDeclarator().isPrototypeContext() &&
             !hasOuterPointerLikeChunk(state.getDeclarator(), endIndex);
-        if (checkNullabilityTypeSpecifier(
-              state,
-              type,
-              attr,
-              allowOnArrayType)) {
+        if (CheckNullabilityTypeSpecifier(state, type, attr,
+                                          allowOnArrayType)) {
           attr.setInvalid();
         }
 
diff --git a/clang/lib/Serialization/ASTReader.cpp b/clang/lib/Serialization/ASTReader.cpp
index 943718b1eea00..c9ec0f3295bdb 100644
--- a/clang/lib/Serialization/ASTReader.cpp
+++ b/clang/lib/Serialization/ASTReader.cpp
@@ -3037,12 +3037,17 @@ ASTReader::ReadControlBlock(ModuleFile &F,
         // location info are setup, in ReadAST.
         SourceLocation ImportLoc =
             ReadUntranslatedSourceLocation(Record[Idx++]);
-        off_t StoredSize = (off_t)Record[Idx++];
-        time_t StoredModTime = (time_t)Record[Idx++];
-        auto FirstSignatureByte = Record.begin() + Idx;
-        ASTFileSignature StoredSignature = ASTFileSignature::create(
-            FirstSignatureByte, FirstSignatureByte + ASTFileSignature::size);
-        Idx += ASTFileSignature::size;
+        off_t StoredSize = !IsImportingStdCXXModule ? (off_t)Record[Idx++] : 0;
+        time_t StoredModTime =
+            !IsImportingStdCXXModule ? (time_t)Record[Idx++] : 0;
+
+        ASTFileSignature StoredSignature;
+        if (!IsImportingStdCXXModule) {
+          auto FirstSignatureByte = Record.begin() + Idx;
+          StoredSignature = ASTFileSignature::create(
+              FirstSignatureByte, FirstSignatureByte + ASTFileSignature::size);
+          Idx += ASTFileSignature::size;
+        }
 
         std::string ImportedName = ReadString(Record, Idx);
         std::string ImportedFile;
@@ -3057,18 +3062,19 @@ ASTReader::ReadControlBlock(ModuleFile &F,
           ImportedFile = PP.getHeaderSearchInfo().getPrebuiltModuleFileName(
               ImportedName, /*FileMapOnly*/ !IsImportingStdCXXModule);
 
-        if (ImportedFile.empty()) {
-          // It is deprecated for C++20 Named modules to use the implicitly
-          // paths.
-          if (IsImportingStdCXXModule)
-            Diag(clang::diag::warn_reading_std_cxx_module_by_implicit_paths)
-                << ImportedName;
-
-          // Use BaseDirectoryAsWritten to ensure we use the same path in the
-          // ModuleCache as when writing.
-          ImportedFile = ReadPath(BaseDirectoryAsWritten, Record, Idx);
-        } else
-          SkipPath(Record, Idx);
+        // For C++20 Modules, we won't record the path to the imported modules
+        // in the BMI
+        if (!IsImportingStdCXXModule) {
+          if (ImportedFile.empty()) {
+            // Use BaseDirectoryAsWritten to ensure we use the same path in the
+            // ModuleCache as when writing.
+            ImportedFile = ReadPath(BaseDirectoryAsWritten, Record, Idx);
+          } else
+            SkipPath(Record, Idx);
+        } else if (ImportedFile.empty()) {
+          Diag(clang::diag::err_failed_to_find_module_file) << ImportedName;
+          return Missing;
+        }
 
         // If our client can't cope with us being out of date, we can't cope with
         // our dependency being missing.
@@ -5584,8 +5590,23 @@ bool ASTReader::readASTFileControlBlock(
       while (Idx < N) {
         // Read information about the AST file.
 
-        // Kind, StandardCXXModule, ImportLoc, Size, ModTime, Signature
-        Idx += 1 + 1 + 1 + 1 + 1 + ASTFileSignature::size;
+        // Skip Kind
+        Idx++;
+        bool IsStandardCXXModule = Record[Idx++];
+
+        // Skip ImportLoc
+        Idx++;
+
+        // In C++20 Modules, we don't record the path to imported
+        // modules in the BMI files.
+        if (IsStandardCXXModule) {
+          std::string ModuleName = ReadString(Record, Idx);
+          Listener.visitImport(ModuleName, /*Filename=*/"");
+          continue;
+        }
+
+        // Skip Size, ModTime and Signature
+        Idx += 1 + 1 + ASTFileSignature::size;
         std::string ModuleName = ReadString(Record, Idx);
         std::string Filename = ReadString(Record, Idx);
         ResolveImportedPath(Filename, ModuleDir);
diff --git a/clang/lib/Serialization/ASTReaderDecl.cpp b/clang/lib/Serialization/ASTReaderDecl.cpp
index 547eb77930b4e..a149d82153037 100644
--- a/clang/lib/Serialization/ASTReaderDecl.cpp
+++ b/clang/lib/Serialization/ASTReaderDecl.cpp
@@ -910,7 +910,7 @@ void ASTDeclReader::VisitEnumConstantDecl(EnumConstantDecl *ECD) {
   VisitValueDecl(ECD);
   if (Record.readInt())
     ECD->setInitExpr(Record.readExpr());
-  ECD->setInitVal(Record.readAPSInt());
+  ECD->setInitVal(Reader.getContext(), Record.readAPSInt());
   mergeMergeable(ECD);
 }
 
@@ -1135,7 +1135,7 @@ void ASTDeclReader::VisitFunctionDecl(FunctionDecl *FD) {
 
   // Defer calling `setPure` until merging above has guaranteed we've set
   // `DefinitionData` (as this will need to access it).
-  FD->setPure(Pure);
+  FD->setIsPureVirtual(Pure);
 
   // Read in the parameters.
   unsigned NumParams = Record.readInt();
@@ -3095,6 +3095,8 @@ class AttrReader {
 
   Expr *readExpr() { return Reader.readExpr(); }
 
+  Attr *readAttr() { return Reader.readAttr(); }
+
   std::string readString() {
     return Reader.readString();
   }
diff --git a/clang/lib/Serialization/ASTWriter.cpp b/clang/lib/Serialization/ASTWriter.cpp
index 78939bfd533ff..a3fa99c3d7e49 100644
--- a/clang/lib/Serialization/ASTWriter.cpp
+++ b/clang/lib/Serialization/ASTWriter.cpp
@@ -1411,15 +1411,20 @@ void ASTWriter::WriteControlBlock(Preprocessor &PP, ASTContext &Context,
       Record.push_back(M.StandardCXXModule);
       AddSourceLocation(M.ImportLoc, Record);
 
-      // If we have calculated signature, there is no need to store
-      // the size or timestamp.
-      Record.push_back(M.Signature ? 0 : M.File.getSize());
-      Record.push_back(M.Signature ? 0 : getTimestampForOutput(M.File));
-
-      llvm::append_range(Record, M.Signature);
+      // We don't want to hard code the information about imported modules
+      // in the C++20 named modules.
+      if (!M.StandardCXXModule) {
+        // If we have calculated signature, there is no need to store
+        // the size or timestamp.
+        Record.push_back(M.Signature ? 0 : M.File.getSize());
+        Record.push_back(M.Signature ? 0 : getTimestampForOutput(M.File));
+        llvm::append_range(Record, M.Signature);
+      }
 
       AddString(M.ModuleName, Record);
-      AddPath(M.FileName, Record);
+
+      if (!M.StandardCXXModule)
+        AddPath(M.FileName, Record);
     }
     Stream.EmitRecord(IMPORTS, Record);
   }
@@ -1892,7 +1897,8 @@ namespace {
 
       unsigned char Flags = (Data.AlreadyIncluded << 6)
                           | (Data.HFI.isImport << 5)
-                          | (Data.HFI.isPragmaOnce << 4)
+                          | (Writer.isWritingStdCXXNamedModules() ? 0 :
+                             Data.HFI.isPragmaOnce << 4)
                           | (Data.HFI.DirInfo << 1)
                           | Data.HFI.IndexHeaderMapHeader;
       LE.write(Flags);
diff --git a/clang/lib/Serialization/ASTWriterDecl.cpp b/clang/lib/Serialization/ASTWriterDecl.cpp
index 9e3299f049184..bb1f51786d281 100644
--- a/clang/lib/Serialization/ASTWriterDecl.cpp
+++ b/clang/lib/Serialization/ASTWriterDecl.cpp
@@ -680,7 +680,7 @@ void ASTDeclWriter::VisitFunctionDecl(FunctionDecl *D) {
   FunctionDeclBits.addBit(D->isInlined());
   FunctionDeclBits.addBit(D->hasSkippedBody());
   FunctionDeclBits.addBit(D->isVirtualAsWritten());
-  FunctionDeclBits.addBit(D->isPure());
+  FunctionDeclBits.addBit(D->isPureVirtual());
   FunctionDeclBits.addBit(D->hasInheritedPrototype());
   FunctionDeclBits.addBit(D->hasWrittenPrototype());
   FunctionDeclBits.addBit(D->isDeletedBit());
diff --git a/clang/lib/StaticAnalyzer/Checkers/CheckSecuritySyntaxOnly.cpp b/clang/lib/StaticAnalyzer/Checkers/CheckSecuritySyntaxOnly.cpp
index ce05d2d3c9058..17af1aebd6d2a 100644
--- a/clang/lib/StaticAnalyzer/Checkers/CheckSecuritySyntaxOnly.cpp
+++ b/clang/lib/StaticAnalyzer/Checkers/CheckSecuritySyntaxOnly.cpp
@@ -973,6 +973,8 @@ void WalkAST::checkMsg_decodeValueOfObjCType(const ObjCMessageExpr *ME) {
     if (VT < VersionTuple(11, 0))
       return;
     break;
+  case llvm::Triple::XROS:
+    break;
   default:
     return;
   }
diff --git a/clang/lib/StaticAnalyzer/Checkers/LocalizationChecker.cpp b/clang/lib/StaticAnalyzer/Checkers/LocalizationChecker.cpp
index de9efe17d220b..812d787e2e37c 100644
--- a/clang/lib/StaticAnalyzer/Checkers/LocalizationChecker.cpp
+++ b/clang/lib/StaticAnalyzer/Checkers/LocalizationChecker.cpp
@@ -712,7 +712,7 @@ void NonLocalizedStringChecker::setNonLocalizedState(const SVal S,
 
 
 static bool isDebuggingName(std::string name) {
-  return StringRef(name).lower().find("debug") != StringRef::npos;
+  return StringRef(name).contains_insensitive("debug");
 }
 
 /// Returns true when, heuristically, the analyzer may be analyzing debugging
@@ -1248,8 +1248,8 @@ bool PluralMisuseChecker::MethodCrawler::isCheckingPlurality(
           BO = B;
         }
       }
-      if (VD->getName().lower().find("plural") != StringRef::npos ||
-          VD->getName().lower().find("singular") != StringRef::npos) {
+      if (VD->getName().contains_insensitive("plural") ||
+          VD->getName().contains_insensitive("singular")) {
         return true;
       }
     }
diff --git a/clang/lib/StaticAnalyzer/Checkers/StdLibraryFunctionsChecker.cpp b/clang/lib/StaticAnalyzer/Checkers/StdLibraryFunctionsChecker.cpp
index 3b36565681a7f..83f972c7072ec 100644
--- a/clang/lib/StaticAnalyzer/Checkers/StdLibraryFunctionsChecker.cpp
+++ b/clang/lib/StaticAnalyzer/Checkers/StdLibraryFunctionsChecker.cpp
@@ -2220,6 +2220,9 @@ void StdLibraryFunctionsChecker::initFunctionSummaries(
                 0, WithinRange, {{EOFv, EOFv}, {0, UCharRangeMax}}))
             .ArgConstraint(NotNull(ArgNo(1))));
 
+    std::optional Off_tTy = lookupTy("off_t");
+    std::optional Off_tMax = getMaxValue(Off_tTy);
+
     // int fseek(FILE *stream, long offset, int whence);
     // FIXME: It can be possible to get the 'SEEK_' values (like EOFv) and use
     // these for condition of arg 2.
@@ -2232,6 +2235,16 @@ void StdLibraryFunctionsChecker::initFunctionSummaries(
             .ArgConstraint(NotNull(ArgNo(0)))
             .ArgConstraint(ArgumentCondition(2, WithinRange, {{0, 2}})));
 
+    // int fseeko(FILE *stream, off_t offset, int whence);
+    addToFunctionSummaryMap(
+        "fseeko",
+        Signature(ArgTypes{FilePtrTy, Off_tTy, IntTy}, RetType{IntTy}),
+        Summary(NoEvalCall)
+            .Case(ReturnsZero, ErrnoMustNotBeChecked, GenericSuccessMsg)
+            .Case(ReturnsMinusOne, ErrnoNEZeroIrrelevant, GenericFailureMsg)
+            .ArgConstraint(NotNull(ArgNo(0)))
+            .ArgConstraint(ArgumentCondition(2, WithinRange, {{0, 2}})));
+
     // int fgetpos(FILE *restrict stream, fpos_t *restrict pos);
     // From 'The Open Group Base Specifications Issue 7, 2018 edition':
     // "The fgetpos() function shall not change the setting of errno if
@@ -2279,6 +2292,15 @@ void StdLibraryFunctionsChecker::initFunctionSummaries(
             .Case(ReturnsMinusOne, ErrnoNEZeroIrrelevant, GenericFailureMsg)
             .ArgConstraint(NotNull(ArgNo(0))));
 
+    // off_t ftello(FILE *stream);
+    addToFunctionSummaryMap(
+        "ftello", Signature(ArgTypes{FilePtrTy}, RetType{Off_tTy}),
+        Summary(NoEvalCall)
+            .Case({ReturnValueCondition(WithinRange, Range(0, Off_tMax))},
+                  ErrnoMustNotBeChecked, GenericSuccessMsg)
+            .Case(ReturnsMinusOne, ErrnoNEZeroIrrelevant, GenericFailureMsg)
+            .ArgConstraint(NotNull(ArgNo(0))));
+
     // int fileno(FILE *stream);
     addToFunctionSummaryMap(
         "fileno", Signature(ArgTypes{FilePtrTy}, RetType{IntTy}),
@@ -2410,8 +2432,6 @@ void StdLibraryFunctionsChecker::initFunctionSummaries(
             .ArgConstraint(
                 ArgumentCondition(0, WithinRange, Range(0, IntMax))));
 
-    std::optional Off_tTy = lookupTy("off_t");
-
     // int truncate(const char *path, off_t length);
     addToFunctionSummaryMap(
         "truncate",
@@ -2772,18 +2792,21 @@ void StdLibraryFunctionsChecker::initFunctionSummaries(
             .ArgConstraint(NotNull(ArgNo(2))));
 
     // DIR *opendir(const char *name);
-    // FIXME: Improve for errno modeling.
     addToFunctionSummaryMap(
         "opendir", Signature(ArgTypes{ConstCharPtrTy}, RetType{DirPtrTy}),
-        Summary(NoEvalCall).ArgConstraint(NotNull(ArgNo(0))));
+        Summary(NoEvalCall)
+            .Case({NotNull(Ret)}, ErrnoMustNotBeChecked, GenericSuccessMsg)
+            .Case({IsNull(Ret)}, ErrnoNEZeroIrrelevant, GenericFailureMsg)
+            .ArgConstraint(NotNull(ArgNo(0))));
 
     // DIR *fdopendir(int fd);
-    // FIXME: Improve for errno modeling.
-    addToFunctionSummaryMap("fdopendir",
-                            Signature(ArgTypes{IntTy}, RetType{DirPtrTy}),
-                            Summary(NoEvalCall)
-                                .ArgConstraint(ArgumentCondition(
-                                    0, WithinRange, Range(0, IntMax))));
+    addToFunctionSummaryMap(
+        "fdopendir", Signature(ArgTypes{IntTy}, RetType{DirPtrTy}),
+        Summary(NoEvalCall)
+            .Case({NotNull(Ret)}, ErrnoMustNotBeChecked, GenericSuccessMsg)
+            .Case({IsNull(Ret)}, ErrnoNEZeroIrrelevant, GenericFailureMsg)
+            .ArgConstraint(
+                ArgumentCondition(0, WithinRange, Range(0, IntMax))));
 
     // int isatty(int fildes);
     addToFunctionSummaryMap(
@@ -2854,19 +2877,6 @@ void StdLibraryFunctionsChecker::initFunctionSummaries(
         "rand_r", Signature(ArgTypes{UnsignedIntPtrTy}, RetType{IntTy}),
         Summary(NoEvalCall).ArgConstraint(NotNull(ArgNo(0))));
 
-    // int fseeko(FILE *stream, off_t offset, int whence);
-    addToFunctionSummaryMap(
-        "fseeko",
-        Signature(ArgTypes{FilePtrTy, Off_tTy, IntTy}, RetType{IntTy}),
-        Summary(NoEvalCall)
-            .Case(ReturnsZeroOrMinusOne, ErrnoIrrelevant)
-            .ArgConstraint(NotNull(ArgNo(0))));
-
-    // off_t ftello(FILE *stream);
-    addToFunctionSummaryMap(
-        "ftello", Signature(ArgTypes{FilePtrTy}, RetType{Off_tTy}),
-        Summary(NoEvalCall).ArgConstraint(NotNull(ArgNo(0))));
-
     // void *mmap(void *addr, size_t length, int prot, int flags, int fd,
     // off_t offset);
     // FIXME: Improve for errno modeling.
diff --git a/clang/lib/StaticAnalyzer/Checkers/StreamChecker.cpp b/clang/lib/StaticAnalyzer/Checkers/StreamChecker.cpp
index fbfa101257d5e..95c7503e49e0d 100644
--- a/clang/lib/StaticAnalyzer/Checkers/StreamChecker.cpp
+++ b/clang/lib/StaticAnalyzer/Checkers/StreamChecker.cpp
@@ -263,13 +263,20 @@ class StreamChecker : public Checker(Call.getOriginExpr());
+  if (!CE)
+    return;
+
+  const StreamState *OldSS = State->get(StreamSym);
+  if (!OldSS)
+    return;
+
+  assertStreamStateOpened(OldSS);
+
+  NonLoc RetVal = makeRetVal(C, CE).castAs();
+  State = State->BindExpr(CE, C.getLocationContext(), RetVal);
+  SValBuilder &SVB = C.getSValBuilder();
+  auto &ACtx = C.getASTContext();
+  auto Cond = SVB.evalBinOp(State, BO_GE, RetVal, SVB.makeZeroVal(ACtx.IntTy),
+                            SVB.getConditionType())
+                  .getAs();
+  if (!Cond)
+    return;
+  ProgramStateRef StateNotFailed, StateFailed;
+  std::tie(StateNotFailed, StateFailed) = State->assume(*Cond);
+
+  StateNotFailed =
+      StateNotFailed->set(StreamSym, StreamState::getOpened(Desc));
+  C.addTransition(StateNotFailed);
+
+  // Add transition for the failed state. The resulting value of the file
+  // position indicator for the stream is indeterminate.
+  StateFailed = StateFailed->set(
+      StreamSym, StreamState::getOpened(Desc, ErrorFError, true));
+  C.addTransition(StateFailed);
+}
+
 void StreamChecker::evalUngetc(const FnDescription *Desc, const CallEvent &Call,
                                CheckerContext &C) const {
   ProgramStateRef State = C.getState();
@@ -1113,10 +1166,10 @@ void StreamChecker::evalFtell(const FnDescription *Desc, const CallEvent &Call,
   NonLoc RetVal = makeRetVal(C, CE).castAs();
   ProgramStateRef StateNotFailed =
       State->BindExpr(CE, C.getLocationContext(), RetVal);
-  auto Cond = SVB.evalBinOp(State, BO_GE, RetVal,
-                            SVB.makeZeroVal(C.getASTContext().LongTy),
-                            SVB.getConditionType())
-                  .getAs();
+  auto Cond =
+      SVB.evalBinOp(State, BO_GE, RetVal, SVB.makeZeroVal(Call.getResultType()),
+                    SVB.getConditionType())
+          .getAs();
   if (!Cond)
     return;
   StateNotFailed = StateNotFailed->assume(*Cond, true);
@@ -1124,7 +1177,7 @@ void StreamChecker::evalFtell(const FnDescription *Desc, const CallEvent &Call,
     return;
 
   ProgramStateRef StateFailed = State->BindExpr(
-      CE, C.getLocationContext(), SVB.makeIntVal(-1, C.getASTContext().LongTy));
+      CE, C.getLocationContext(), SVB.makeIntVal(-1, Call.getResultType()));
 
   // This function does not affect the stream state.
   // Still we add success and failure state with the appropriate return value.
diff --git a/clang/lib/StaticAnalyzer/Checkers/VirtualCallChecker.cpp b/clang/lib/StaticAnalyzer/Checkers/VirtualCallChecker.cpp
index 1c589e3468c2d..33a9a07f9d32d 100644
--- a/clang/lib/StaticAnalyzer/Checkers/VirtualCallChecker.cpp
+++ b/clang/lib/StaticAnalyzer/Checkers/VirtualCallChecker.cpp
@@ -116,7 +116,7 @@ void VirtualCallChecker::checkPreCall(const CallEvent &Call,
   if (!ObState)
     return;
 
-  bool IsPure = MD->isPure();
+  bool IsPure = MD->isPureVirtual();
 
   // At this point we're sure that we're calling a virtual method
   // during construction or destruction, so we'll emit a report.
diff --git a/clang/lib/StaticAnalyzer/Core/HTMLDiagnostics.cpp b/clang/lib/StaticAnalyzer/Core/HTMLDiagnostics.cpp
index 69d25120dcd43..86947b7929e9b 100644
--- a/clang/lib/StaticAnalyzer/Core/HTMLDiagnostics.cpp
+++ b/clang/lib/StaticAnalyzer/Core/HTMLDiagnostics.cpp
@@ -68,6 +68,7 @@ class HTMLDiagnostics : public PathDiagnosticConsumer {
   bool noDir = false;
   const Preprocessor &PP;
   const bool SupportsCrossFileDiagnostics;
+  llvm::StringSet<> EmittedHashes;
 
 public:
   HTMLDiagnostics(PathDiagnosticConsumerOptions DiagOpts,
@@ -301,6 +302,17 @@ void HTMLDiagnostics::ReportDiag(const PathDiagnostic& D,
       }
   }
 
+  SmallString<32> IssueHash = getIssueHash(D, PP);
+  auto [It, IsNew] = EmittedHashes.insert(IssueHash);
+  if (!IsNew) {
+    // We've already emitted a duplicate issue. It'll get overwritten anyway.
+    return;
+  }
+
+  // FIXME: This causes each file to be re-parsed and syntax-highlighted
+  // and macro-expanded separately for each report. We could cache such rewrites
+  // across all reports and only re-do the part that's actually different:
+  // the warning/note bubbles.
   std::string report = GenerateHTML(D, R, SMgr, path, declName.c_str());
   if (report.empty()) {
     llvm::errs() << "warning: no diagnostics generated for main file.\n";
@@ -332,7 +344,7 @@ void HTMLDiagnostics::ReportDiag(const PathDiagnostic& D,
              << declName.c_str() << "-" << offsetDecl << "-";
   }
 
-  FileName << StringRef(getIssueHash(D, PP)).substr(0, 6).str() << ".html";
+  FileName << StringRef(IssueHash).substr(0, 6).str() << ".html";
 
   SmallString<128> ResultPath;
   llvm::sys::path::append(ResultPath, Directory, FileName.str());
diff --git a/clang/lib/Tooling/Syntax/Tokens.cpp b/clang/lib/Tooling/Syntax/Tokens.cpp
index 2f28b9cf286a6..8d32c45a4a70c 100644
--- a/clang/lib/Tooling/Syntax/Tokens.cpp
+++ b/clang/lib/Tooling/Syntax/Tokens.cpp
@@ -401,6 +401,12 @@ std::string TokenBuffer::Mapping::str() const {
 
 std::optional>
 TokenBuffer::spelledForExpanded(llvm::ArrayRef Expanded) const {
+  // In cases of invalid code, AST nodes can have source ranges that include
+  // the `eof` token. As there's no spelling for this token, exclude it from
+  // the range.
+  if (!Expanded.empty() && Expanded.back().kind() == tok::eof) {
+    Expanded = Expanded.drop_back();
+  }
   // Mapping an empty range is ambiguous in case of empty mappings at either end
   // of the range, bail out in that case.
   if (Expanded.empty())
diff --git a/clang/test/AST/Interp/arrays.cpp b/clang/test/AST/Interp/arrays.cpp
index 4aa10da55dd3a..e14ff34dd7371 100644
--- a/clang/test/AST/Interp/arrays.cpp
+++ b/clang/test/AST/Interp/arrays.cpp
@@ -566,3 +566,35 @@ namespace GH69115 {
   static_assert(foo2() == 0, "");
 #endif
 }
+
+namespace NonConstReads {
+#if __cplusplus >= 202002L
+  void *p = nullptr; // ref-note {{declared here}} \
+                     // expected-note {{declared here}}
+
+  int arr[!p]; // ref-error {{not allowed at file scope}} \
+               // expected-error {{not allowed at file scope}} \
+               // ref-warning {{variable length arrays}} \
+               // ref-note {{read of non-constexpr variable 'p'}} \
+               // expected-warning {{variable length arrays}} \
+               // expected-note {{read of non-constexpr variable 'p'}}
+  int z; // ref-note {{declared here}} \
+         // expected-note {{declared here}}
+  int a[z]; // ref-error {{not allowed at file scope}} \
+            // expected-error {{not allowed at file scope}} \
+            // ref-warning {{variable length arrays}} \
+            // ref-note {{read of non-const variable 'z'}} \
+            // expected-warning {{variable length arrays}} \
+            // expected-note {{read of non-const variable 'z'}}
+#else
+  void *p = nullptr;
+  int arr[!p]; // ref-error {{not allowed at file scope}} \
+               // expected-error {{not allowed at file scope}}
+  int z;
+  int a[z]; // ref-error {{not allowed at file scope}} \
+            // expected-error {{not allowed at file scope}}
+#endif
+
+  const int y = 0;
+  int yy[y];
+}
diff --git a/clang/test/AST/Interp/complex.cpp b/clang/test/AST/Interp/complex.cpp
index 66490e973988b..e63693a0cfaa1 100644
--- a/clang/test/AST/Interp/complex.cpp
+++ b/clang/test/AST/Interp/complex.cpp
@@ -1,5 +1,5 @@
-// RUN: %clang_cc1 -fexperimental-new-constant-interpreter -verify %s
-// RUN: %clang_cc1 -verify=ref %s
+// RUN: %clang_cc1 -fexperimental-new-constant-interpreter -verify -Wno-unused-value %s
+// RUN: %clang_cc1 -verify=ref -Wno-unused-value %s
 
 // expected-no-diagnostics
 // ref-no-diagnostics
@@ -37,6 +37,24 @@ constexpr _Complex int I2 = {};
 static_assert(__real(I2) == 0, "");
 static_assert(__imag(I2) == 0, "");
 
+static_assert(__real(4.0) == 4.0, "");
+static_assert(__real(12u) == 12u, "");
+static_assert(__imag(4.0) == 0.0, "");
+static_assert(__imag(13) == 0, "");
+
+constexpr int ignoredCast() {
+  I2;
+  (int)I2;
+  (float)I2;
+  D1;
+  (int)D1;
+  (double)D1;
+  return 0;
+}
+static_assert(ignoredCast() == 0, "");
+static_assert((int)I1 == 1, "");
+static_assert((float)D == 1.0f, "");
+
 
 /// Standalone complex expressions.
 static_assert(__real((_Complex float){1.0, 3.0}) == 1.0, "");
diff --git a/clang/test/AST/Interp/cxx11.cpp b/clang/test/AST/Interp/cxx11.cpp
new file mode 100644
index 0000000000000..81e293fec1750
--- /dev/null
+++ b/clang/test/AST/Interp/cxx11.cpp
@@ -0,0 +1,24 @@
+// RUN: %clang_cc1 -fexperimental-new-constant-interpreter -verify=both,expected -std=c++11 %s
+// RUN: %clang_cc1 -verify=both,ref -std=c++11 %s
+
+// expected-no-diagnostics
+
+namespace IntOrEnum {
+  const int k = 0;
+  const int &p = k;
+  template struct S {};
+  S
 s;
+}
+
+const int cval = 2;
+template  struct C{};
+template struct C;
+
+
+/// FIXME: This example does not get properly diagnosed in the new interpreter.
+extern const int recurse1;
+const int recurse2 = recurse1; // ref-note {{here}}
+const int recurse1 = 1;
+int array1[recurse1];
+int array2[recurse2]; // ref-warning 2{{variable length array}} \
+                      // ref-note {{initializer of 'recurse2' is not a constant expression}}
diff --git a/clang/test/AST/Interp/cxx23.cpp b/clang/test/AST/Interp/cxx23.cpp
index bd1cf186d519c..133ab10023df0 100644
--- a/clang/test/AST/Interp/cxx23.cpp
+++ b/clang/test/AST/Interp/cxx23.cpp
@@ -25,22 +25,32 @@ constexpr int g(int n) {        // ref20-error {{constexpr function never produc
 }
 
 constexpr int c_thread_local(int n) { // ref20-error {{constexpr function never produces a constant expression}} \
-                                      // ref23-error {{constexpr function never produces a constant expression}}
+                                      // ref23-error {{constexpr function never produces a constant expression}} \
+                                      // expected20-error {{constexpr function never produces a constant expression}} \
+                                      // expected23-error {{constexpr function never produces a constant expression}}
   static _Thread_local int m = 0;     // ref20-note {{control flows through the definition of a thread_local variable}} \
                                       // ref20-warning {{is a C++23 extension}} \
                                       // ref23-note {{control flows through the definition of a thread_local variable}} \
-                                      // expected20-warning {{is a C++23 extension}}
-  return m;
+                                      // expected20-warning {{is a C++23 extension}} \
+                                      // expected20-note {{declared here}} \
+                                      // expected23-note {{declared here}}
+  return m; // expected20-note {{read of non-const variable}} \
+            // expected23-note {{read of non-const variable}}
 }
 
 
 constexpr int gnu_thread_local(int n) { // ref20-error {{constexpr function never produces a constant expression}} \
-                                        // ref23-error {{constexpr function never produces a constant expression}}
+                                        // ref23-error {{constexpr function never produces a constant expression}} \
+                                        // expected20-error {{constexpr function never produces a constant expression}} \
+                                        // expected23-error {{constexpr function never produces a constant expression}}
   static __thread int m = 0;            // ref20-note {{control flows through the definition of a thread_local variable}} \
                                         // ref20-warning {{is a C++23 extension}} \
                                         // ref23-note {{control flows through the definition of a thread_local variable}} \
-                                        // expected20-warning {{is a C++23 extension}}
-  return m;
+                                        // expected20-warning {{is a C++23 extension}} \
+                                        // expected20-note {{declared here}} \
+                                        // expected23-note {{declared here}}
+  return m; // expected20-note {{read of non-const variable}} \
+            // expected23-note {{read of non-const variable}}
 }
 
 constexpr int h(int n) {  // ref20-error {{constexpr function never produces a constant expression}} \
diff --git a/clang/test/AST/Interp/cxx98.cpp b/clang/test/AST/Interp/cxx98.cpp
new file mode 100644
index 0000000000000..79f93c8d78f16
--- /dev/null
+++ b/clang/test/AST/Interp/cxx98.cpp
@@ -0,0 +1,32 @@
+// RUN: %clang_cc1 -fexperimental-new-constant-interpreter -verify=both,expected -std=c++98 %s
+// RUN: %clang_cc1 -verify=both,ref -std=c++98 %s
+
+
+
+namespace IntOrEnum {
+  const int k = 0;
+  const int &p = k; // both-note {{declared here}}
+  template struct S {};
+  S
 s; // both-error {{not an integral constant expression}} \
+          // both-note {{read of variable 'p' of non-integral, non-enumeration type 'const int &'}}
+}
+
+const int cval = 2;
+template  struct C{};
+template struct C;
+
+
+/// FIXME: This example does not get properly diagnosed in the new interpreter.
+extern const int recurse1;
+const int recurse2 = recurse1; // ref-note {{here}}
+const int recurse1 = 1;
+int array1[recurse1];
+int array2[recurse2]; // ref-warning 2{{variable length array}} \
+                      // ref-note {{initializer of 'recurse2' is not a constant expression}} \
+                      // expected-warning {{variable length array}} \
+                      // expected-error {{variable length array}}
+
+int NCI; // both-note {{declared here}}
+int NCIA[NCI]; // both-warning {{variable length array}} \
+               // both-error {{variable length array}} \\
+               // both-note {{read of non-const variable 'NCI'}}
diff --git a/clang/test/AST/Interp/functions.cpp b/clang/test/AST/Interp/functions.cpp
index 179a195098b13..75f3c5d192b2c 100644
--- a/clang/test/AST/Interp/functions.cpp
+++ b/clang/test/AST/Interp/functions.cpp
@@ -389,3 +389,27 @@ namespace Packs {
   static_assert(foo() == 2, "");
   static_assert(foo<>() == 0, "");
 }
+
+namespace AddressOf {
+  struct S {} s;
+  static_assert(__builtin_addressof(s) == &s, "");
+
+  struct T { constexpr T *operator&() const { return nullptr; } int n; } t;
+  constexpr T *pt = __builtin_addressof(t);
+  static_assert(&pt->n == &t.n, "");
+
+  struct U { int n : 5; } u;
+  int *pbf = __builtin_addressof(u.n); // expected-error {{address of bit-field requested}} \
+                                       // ref-error {{address of bit-field requested}}
+
+  S *ptmp = __builtin_addressof(S{}); // expected-error {{taking the address of a temporary}} \
+                                      // expected-warning {{temporary whose address is used as value of local variable 'ptmp' will be destroyed at the end of the full-expression}} \
+                                      // ref-error {{taking the address of a temporary}} \
+                                      // ref-warning {{temporary whose address is used as value of local variable 'ptmp' will be destroyed at the end of the full-expression}}
+
+  constexpr int foo() {return 1;}
+  static_assert(__builtin_addressof(foo) == foo, "");
+
+  constexpr _Complex float F = {3, 4};
+  static_assert(__builtin_addressof(F) == &F, "");
+}
diff --git a/clang/test/AST/Interp/intap.cpp b/clang/test/AST/Interp/intap.cpp
index b99422dc8f931..118dc21b67e87 100644
--- a/clang/test/AST/Interp/intap.cpp
+++ b/clang/test/AST/Interp/intap.cpp
@@ -5,6 +5,7 @@
 
 
 using MaxBitInt = _BitInt(128);
+#define INT_MIN (~__INT_MAX__)
 
 constexpr _BitInt(2) A = 0;
 constexpr _BitInt(2) B = A + 1;
@@ -44,6 +45,35 @@ static_assert(MulA * MulB == 50, ""); // ref-error {{not an integral constant ex
 static_assert(MulA * 5 == 25, "");
 static_assert(-1 * MulB == -7, "");
 
+
+constexpr _BitInt(4) DivA = 2;
+constexpr _BitInt(2) DivB = 1;
+static_assert(DivA / DivB == 2, "");
+
+constexpr _BitInt(4) DivC = DivA / 0; // ref-error {{must be initialized by a constant expression}} \
+                                      // ref-note {{division by zero}} \
+                                      // expected-error {{must be initialized by a constant expression}} \
+                                      // expected-note {{division by zero}}
+
+constexpr _BitInt(7) RemA = 47;
+constexpr _BitInt(6) RemB = 9;
+static_assert(RemA % RemB == 2, "");
+static_assert(RemA % 0 == 1, ""); // ref-error {{not an integral constant expression}} \
+                                  // ref-note {{division by zero}} \
+                                  // expected-error {{not an integral constant expression}} \
+                                  // expected-note {{division by zero}}
+
+constexpr _BitInt(32) bottom = -1;
+constexpr _BitInt(32) top = INT_MIN;
+constexpr _BitInt(32) nope = top / bottom;  // ref-error {{must be initialized by a constant expression}} \
+                                            // ref-note {{value 2147483648 is outside the range}} \
+                                            // expected-error {{must be initialized by a constant expression}} \
+                                            // expected-note {{value 2147483648 is outside the range}}
+constexpr _BitInt(32) noooo = top % bottom; // ref-error {{must be initialized by a constant expression}} \
+                                            // ref-note {{value 2147483648 is outside the range}} \
+                                            // expected-error {{must be initialized by a constant expression}} \
+                                            // expected-note {{value 2147483648 is outside the range}}
+
 namespace APCast {
   constexpr _BitInt(10) A = 1;
   constexpr _BitInt(11) B = A;
diff --git a/clang/test/AST/Interp/literals.cpp b/clang/test/AST/Interp/literals.cpp
index 61825bc11438f..77b75151a1255 100644
--- a/clang/test/AST/Interp/literals.cpp
+++ b/clang/test/AST/Interp/literals.cpp
@@ -1181,8 +1181,29 @@ namespace InvalidDeclRefs {
                           // expected-error {{not an integral constant expression}} \
                           // expected-note {{initializer of 'b02' is unknown}}
 
-  /// FIXME: This should also be diagnosed in the new interpreter.
-  int b03 = 3; // ref-note {{declared here}}
+  int b03 = 3; // ref-note {{declared here}} \
+               // expected-note {{declared here}}
   static_assert(b03, ""); // ref-error {{not an integral constant expression}} \
-                          // ref-note {{read of non-const variable}}
+                          // ref-note {{read of non-const variable}} \
+                          // expected-error {{not an integral constant expression}} \
+                          // expected-note {{read of non-const variable}}
+}
+
+namespace NonConstReads {
+  void *p = nullptr; // ref-note {{declared here}} \
+                     // expected-note {{declared here}}
+  static_assert(!p, ""); // ref-error {{not an integral constant expression}} \
+                         // ref-note {{read of non-constexpr variable 'p'}} \
+                         // expected-error {{not an integral constant expression}} \
+                         // expected-note {{read of non-constexpr variable 'p'}}
+
+  int arr[!p]; // ref-error {{variable length array}} \
+               // expected-error {{variable length array}}
+
+  int z; // ref-note {{declared here}} \
+         // expected-note {{declared here}}
+  static_assert(z == 0, ""); // ref-error {{not an integral constant expression}} \
+                             // ref-note {{read of non-const variable 'z'}} \
+                             // expected-error {{not an integral constant expression}} \
+                             // expected-note {{read of non-const variable 'z'}}
 }
diff --git a/clang/test/AST/Interp/records.cpp b/clang/test/AST/Interp/records.cpp
index 280eaf34898ce..800dc6f910be6 100644
--- a/clang/test/AST/Interp/records.cpp
+++ b/clang/test/AST/Interp/records.cpp
@@ -734,8 +734,6 @@ namespace VirtualDtors {
   }
 
   static_assert(foo());
-
-
 };
 
 namespace QualifiedCalls {
@@ -1135,3 +1133,90 @@ namespace AccessOnNullptr {
                                // ref-error {{not an integral constant expression}} \
                                // ref-note {{in call to 'a2()'}}
 }
+
+namespace IndirectFieldInit {
+#if __cplusplus >= 202002L
+  /// Primitive.
+  struct Nested1 {
+    struct {
+      int first;
+    };
+    int x;
+    constexpr Nested1(int x) : first(12), x() { x = 4; }
+    constexpr Nested1() : Nested1(42) {}
+  };
+  constexpr Nested1 N1{};
+  static_assert(N1.first == 12, "");
+
+  /// Composite.
+  struct Nested2 {
+    struct First { int x = 42; };
+    struct {
+      First first;
+    };
+    int x;
+    constexpr Nested2(int x) : first(12), x() { x = 4; }
+    constexpr Nested2() : Nested2(42) {}
+  };
+  constexpr Nested2 N2{};
+  static_assert(N2.first.x == 12, "");
+
+  /// Bitfield.
+  struct Nested3 {
+    struct {
+      unsigned first : 2;
+    };
+    int x;
+    constexpr Nested3(int x) : first(3), x() { x = 4; }
+    constexpr Nested3() : Nested3(42) {}
+  };
+
+  constexpr Nested3 N3{};
+  static_assert(N3.first == 3, "");
+
+  /// Test that we get the offset right if the
+  /// record has a base.
+  struct Nested4Base {
+    int a;
+    int b;
+    char c;
+  };
+  struct Nested4 : Nested4Base{
+    struct {
+      int first;
+    };
+    int x;
+    constexpr Nested4(int x) : first(123), x() { a = 1; b = 2; c = 3; x = 4; }
+    constexpr Nested4() : Nested4(42) {}
+  };
+  constexpr Nested4 N4{};
+  static_assert(N4.first == 123, "");
+
+  struct S {
+    struct {
+      int x, y;
+    };
+
+    constexpr S(int x_, int y_) : x(x_), y(y_) {}
+  };
+
+  constexpr S s(1, 2);
+  static_assert(s.x == 1 && s.y == 2);
+
+  struct S2 {
+    int a;
+    struct {
+      int b;
+      struct {
+        int x, y;
+      };
+    };
+
+    constexpr S2(int x_, int y_) : a(3), b(4), x(x_), y(y_) {}
+  };
+
+  constexpr S2 s2(1, 2);
+  static_assert(s2.x == 1 && s2.y == 2 && s2.a == 3 && s2.b == 4);
+
+#endif
+}
diff --git a/clang/test/AST/ast-dump-invalid-initialized.cpp b/clang/test/AST/ast-dump-invalid-initialized.cpp
index 1c374ae716a9d..7fcbc41a7be40 100644
--- a/clang/test/AST/ast-dump-invalid-initialized.cpp
+++ b/clang/test/AST/ast-dump-invalid-initialized.cpp
@@ -24,4 +24,19 @@ void test() {
   auto b4 = A(1);
   // CHECK: `-VarDecl {{.*}} invalid b5 'auto'
   auto b5 = A{1};
-}
\ No newline at end of file
+}
+
+void GH72198() {
+  // CHECK: DecompositionDecl {{.*}} invalid 'int'
+  int [_, b] = {0, 0};
+  [b]{};
+}
+
+namespace GH67495 {
+int get_point();
+void f() {
+  // CHECK: DecompositionDecl {{.*}} invalid 'int &'
+  auto& [x, y] = get_point();
+  [x, y] {};
+}
+}
diff --git a/clang/test/AST/ast-dump-sme-attributes.cpp b/clang/test/AST/ast-dump-sme-attributes.cpp
index 6581fd4baba9e..133648d90a157 100644
--- a/clang/test/AST/ast-dump-sme-attributes.cpp
+++ b/clang/test/AST/ast-dump-sme-attributes.cpp
@@ -13,16 +13,16 @@ struct Foo {
 // CHECK-NEXT: |-CXXMethodDecl {{.*}} f_streaming_compatible 'void () __arm_streaming_compatible'
 // CHECK-NEXT: |-CXXMethodDecl {{.*}} f_locally_streaming 'void ()'
 // CHECK-NEXT: | `-ArmLocallyStreamingAttr
-// CHECK-NEXT: |-CXXMethodDecl {{.*}} f_shared_za 'void () __arm_shared_za'
+// CHECK-NEXT: |-CXXMethodDecl {{.*}} f_shared_za 'void () __arm_inout("za")'
 // CHECK-NEXT: |-CXXMethodDecl {{.*}} f_new_za 'void ()'
-// CHECK-NEXT: | `-ArmNewZAAttr
-// CHECK-NEXT: |-CXXMethodDecl {{.*}} f_preserves_za 'void () __arm_preserves_za'
+// CHECK-NEXT: | `-ArmNewAttr {{.*}} za
+// CHECK-NEXT: |-CXXMethodDecl {{.*}} f_preserves_za 'void () __arm_preserves("za")'
   void f_streaming() __arm_streaming;
   void f_streaming_compatible() __arm_streaming_compatible;
   __arm_locally_streaming void f_locally_streaming();
-  void f_shared_za() __arm_shared_za;
-  __arm_new_za void f_new_za();
-  void f_preserves_za() __arm_preserves_za;
+  void f_shared_za() __arm_inout("za");
+  __arm_new("za") void f_new_za();
+  void f_preserves_za() __arm_preserves("za");
 
 
 // CHECK:      |-CXXMethodDecl {{.*}} test_lambda 'int (int)' implicit-inline
diff --git a/clang/test/Analysis/Inputs/system-header-simulator-for-simple-stream.h b/clang/test/Analysis/Inputs/system-header-simulator-for-simple-stream.h
index b65b7a6b0e7b0..098a2208fecbe 100644
--- a/clang/test/Analysis/Inputs/system-header-simulator-for-simple-stream.h
+++ b/clang/test/Analysis/Inputs/system-header-simulator-for-simple-stream.h
@@ -8,9 +8,9 @@
 typedef struct __sFILE {
   unsigned char *_p;
 } FILE;
-FILE *fopen(const char * restrict, const char * restrict) __asm("_" "fopen" );
+FILE *fopen(const char *restrict, const char *restrict) __asm("_" "fopen" );
 int fputc(int, FILE *);
-int fputs(const char * restrict, FILE * restrict) __asm("_" "fputs" );
+int fputs(const char *restrict, FILE *restrict) __asm("_" "fputs" );
 int fclose(FILE *);
 void exit(int);
 
diff --git a/clang/test/Analysis/Inputs/system-header-simulator.h b/clang/test/Analysis/Inputs/system-header-simulator.h
index caae59c38a4c8..f8e3e546a7aed 100644
--- a/clang/test/Analysis/Inputs/system-header-simulator.h
+++ b/clang/test/Analysis/Inputs/system-header-simulator.h
@@ -13,6 +13,7 @@ typedef __typeof(sizeof(int)) size_t;
 typedef long long __int64_t;
 typedef __int64_t __darwin_off_t;
 typedef __darwin_off_t fpos_t;
+typedef int off_t;
 
 typedef struct _FILE FILE;
 #define SEEK_SET 0 /* Seek from beginning of file. */
@@ -55,7 +56,9 @@ int fputc(int ch, FILE *stream);
 int fputs(const char *restrict s, FILE *restrict stream);
 int ungetc(int c, FILE *stream);
 int fseek(FILE *__stream, long int __off, int __whence);
+int fseeko(FILE *__stream, off_t __off, int __whence);
 long int ftell(FILE *__stream);
+off_t ftello(FILE *__stream);
 void rewind(FILE *__stream);
 int fgetpos(FILE *restrict stream, fpos_t *restrict pos);
 int fsetpos(FILE *stream, const fpos_t *pos);
@@ -68,9 +71,9 @@ int fflush(FILE *stream);
 size_t strlen(const char *);
 
 char *strcpy(char *restrict, const char *restrict);
-char *strncpy(char *dst, const char *src, size_t n);
-char *strsep(char **stringp, const char *delim);
-void *memcpy(void *dst, const void *src, size_t n);
+char *strncpy(char *restrict dst, const char *restrict src, size_t n);
+char *strsep(char **restrict stringp, const char *restrict delim);
+void *memcpy(void *restrict dst, const void *restrict src, size_t n);
 void *memset(void *s, int c, size_t n);
 
 typedef unsigned long __darwin_pthread_key_t;
diff --git a/clang/test/Analysis/bsd-string.c b/clang/test/Analysis/bsd-string.c
index 1c7b28198dff2..93b2214786009 100644
--- a/clang/test/Analysis/bsd-string.c
+++ b/clang/test/Analysis/bsd-string.c
@@ -7,8 +7,8 @@
 #define NULL ((void *)0)
 
 typedef __typeof(sizeof(int)) size_t;
-size_t strlcpy(char *dst, const char *src, size_t n);
-size_t strlcat(char *dst, const char *src, size_t n);
+size_t strlcpy(char *restrict dst, const char *restrict src, size_t n);
+size_t strlcat(char *restrict dst, const char *restrict src, size_t n);
 size_t strlen(const char *s);
 void clang_analyzer_eval(int);
 
diff --git a/clang/test/Analysis/std-c-library-functions-POSIX.c b/clang/test/Analysis/std-c-library-functions-POSIX.c
index daa4d904c3ac5..8a26e0b0f2728 100644
--- a/clang/test/Analysis/std-c-library-functions-POSIX.c
+++ b/clang/test/Analysis/std-c-library-functions-POSIX.c
@@ -21,6 +21,8 @@
 // CHECK: Loaded summary for: FILE *freopen(const char *restrict pathname, const char *restrict mode, FILE *restrict stream)
 // CHECK: Loaded summary for: int fclose(FILE *stream)
 // CHECK: Loaded summary for: int fseek(FILE *stream, long offset, int whence)
+// CHECK: Loaded summary for: int fseeko(FILE *stream, off_t offset, int whence)
+// CHECK: Loaded summary for: off_t ftello(FILE *stream)
 // CHECK: Loaded summary for: int fileno(FILE *stream)
 // CHECK: Loaded summary for: long a64l(const char *str64)
 // CHECK: Loaded summary for: char *l64a(long value)
@@ -80,8 +82,6 @@
 // CHECK: Loaded summary for: void rewinddir(DIR *dir)
 // CHECK: Loaded summary for: void seekdir(DIR *dirp, long loc)
 // CHECK: Loaded summary for: int rand_r(unsigned int *seedp)
-// CHECK: Loaded summary for: int fseeko(FILE *stream, off_t offset, int whence)
-// CHECK: Loaded summary for: off_t ftello(FILE *stream)
 // CHECK: Loaded summary for: void *mmap(void *addr, size_t length, int prot, int flags, int fd, off_t offset)
 // CHECK: Loaded summary for: void *mmap64(void *addr, size_t length, int prot, int flags, int fd, off64_t offset)
 // CHECK: Loaded summary for: int pipe(int fildes[2])
diff --git a/clang/test/Analysis/stream-errno.c b/clang/test/Analysis/stream-errno.c
index f44ee6070708b..13981e9094d1c 100644
--- a/clang/test/Analysis/stream-errno.c
+++ b/clang/test/Analysis/stream-errno.c
@@ -129,6 +129,7 @@ void check_fseek(void) {
   int S = fseek(F, 11, SEEK_SET);
   if (S != 0) {
     clang_analyzer_eval(errno != 0); // expected-warning{{TRUE}}
+    clang_analyzer_eval(S == -1);    // expected-warning{{TRUE}}
     if (errno) {} // no-warning
     fclose(F);
     return;
@@ -136,6 +137,21 @@ void check_fseek(void) {
   if (errno) {} // expected-warning{{An undefined value may be read from 'errno'}}
 }
 
+void check_fseeko(void) {
+  FILE *F = tmpfile();
+  if (!F)
+    return;
+  int S = fseeko(F, 11, SEEK_SET);
+  if (S == -1) {
+    clang_analyzer_eval(errno != 0); // expected-warning{{TRUE}}
+    if (errno) {}                    // no-warning
+  } else {
+    clang_analyzer_eval(S == 0);     // expected-warning{{TRUE}}
+    if (errno) {}                    // expected-warning{{An undefined value may be read from 'errno'}}
+  }
+  fclose(F);
+}
+
 void check_no_errno_change(void) {
   FILE *F = tmpfile();
   if (!F)
@@ -197,6 +213,21 @@ void check_ftell(void) {
   fclose(F);
 }
 
+void check_ftello(void) {
+  FILE *F = tmpfile();
+  if (!F)
+    return;
+  off_t Ret = ftello(F);
+  if (Ret >= 0) {
+    if (errno) {}                    // expected-warning{{An undefined value may be read from 'errno'}}
+  } else {
+    clang_analyzer_eval(Ret == -1);  // expected-warning{{TRUE}}
+    clang_analyzer_eval(errno != 0); // expected-warning{{TRUE}}
+    if (errno) {}                    // no-warning
+  }
+  fclose(F);
+}
+
 void check_rewind(void) {
   FILE *F = tmpfile();
   if (!F)
diff --git a/clang/test/Analysis/stream-error.c b/clang/test/Analysis/stream-error.c
index c038348e799d2..0f7fdddc0dd4c 100644
--- a/clang/test/Analysis/stream-error.c
+++ b/clang/test/Analysis/stream-error.c
@@ -191,6 +191,23 @@ void error_fputs(void) {
   fputs("ABC", F); // expected-warning {{Stream might be already closed}}
 }
 
+void error_fprintf(void) {
+  FILE *F = tmpfile();
+  if (!F)
+    return;
+  int Ret = fprintf(F, "aaa");
+  if (Ret >= 0) {
+    clang_analyzer_eval(feof(F) || ferror(F)); // expected-warning {{FALSE}}
+    fprintf(F, "bbb");                         // no-warning
+  } else {
+    clang_analyzer_eval(ferror(F)); // expected-warning {{TRUE}}
+    clang_analyzer_eval(feof(F));   // expected-warning {{FALSE}}
+    fprintf(F, "bbb");              // expected-warning {{might be 'indeterminate'}}
+  }
+  fclose(F);
+  fprintf(F, "ccc"); // expected-warning {{Stream might be already closed}}
+}
+
 void error_ungetc() {
   FILE *F = tmpfile();
   if (!F)
@@ -295,6 +312,25 @@ void error_fseek(void) {
   fclose(F);
 }
 
+void error_fseeko(void) {
+  FILE *F = fopen("file", "r");
+  if (!F)
+    return;
+  int rc = fseeko(F, 1, SEEK_SET);
+  if (rc) {
+    int IsFEof = feof(F), IsFError = ferror(F);
+    // Get feof or ferror or no error.
+    clang_analyzer_eval(IsFEof || IsFError);
+    // expected-warning@-1 {{FALSE}}
+    // expected-warning@-2 {{TRUE}}
+    clang_analyzer_eval(IsFEof && IsFError); // expected-warning {{FALSE}}
+  } else {
+    clang_analyzer_eval(feof(F));   // expected-warning {{FALSE}}
+    clang_analyzer_eval(ferror(F)); // expected-warning {{FALSE}}
+  }
+  fclose(F);
+}
+
 void error_fseek_0(void) {
   FILE *F = fopen("file", "r");
   if (!F)
@@ -324,6 +360,68 @@ void error_fseek_0(void) {
   fclose(F);
 }
 
+void error_fseeko_0(void) {
+  FILE *F = fopen("file", "r");
+  if (!F)
+    return;
+  int rc = fseeko(F, 0, SEEK_SET);
+  if (rc) {
+    int IsFEof = feof(F), IsFError = ferror(F);
+    // Get ferror or no error, but not feof.
+    clang_analyzer_eval(IsFError);
+    // expected-warning@-1 {{FALSE}}
+    // expected-warning@-2 {{TRUE}}
+    clang_analyzer_eval(IsFEof);
+    // expected-warning@-1 {{FALSE}}
+  } else {
+    clang_analyzer_eval(feof(F));   // expected-warning {{FALSE}}
+    clang_analyzer_eval(ferror(F)); // expected-warning {{FALSE}}
+  }
+  fclose(F);
+}
+
+void error_ftell(void) {
+  FILE *F = fopen("file", "r");
+  if (!F)
+    return;
+  long rc = ftell(F);
+  if (rc >= 0)
+    clang_analyzer_warnIfReached();          // expected-warning {{REACHABLE}}
+  else
+    clang_analyzer_eval(rc == -1);           // expected-warning {{TRUE}}
+  clang_analyzer_eval(feof(F) && ferror(F)); // expected-warning {{FALSE}}
+  StreamTesterChecker_make_feof_stream(F);
+  rc = ftell(F);
+  clang_analyzer_eval(feof(F));              // expected-warning {{TRUE}}
+  clang_analyzer_eval(ferror(F));            // expected-warning {{FALSE}}
+  StreamTesterChecker_make_ferror_stream(F);
+  rc = ftell(F);
+  clang_analyzer_eval(feof(F));              // expected-warning {{FALSE}}
+  clang_analyzer_eval(ferror(F));            // expected-warning {{TRUE}}
+  fclose(F);
+}
+
+void error_ftello(void) {
+  FILE *F = fopen("file", "r");
+  if (!F)
+    return;
+  off_t rc = ftello(F);
+  if (rc >= 0)
+    clang_analyzer_warnIfReached();          // expected-warning {{REACHABLE}}
+  else
+    clang_analyzer_eval(rc == -1);           // expected-warning {{TRUE}}
+  clang_analyzer_eval(feof(F) && ferror(F)); // expected-warning {{FALSE}}
+  StreamTesterChecker_make_feof_stream(F);
+  rc = ftello(F);
+  clang_analyzer_eval(feof(F));              // expected-warning {{TRUE}}
+  clang_analyzer_eval(ferror(F));            // expected-warning {{FALSE}}
+  StreamTesterChecker_make_ferror_stream(F);
+  rc = ftello(F);
+  clang_analyzer_eval(feof(F));              // expected-warning {{FALSE}}
+  clang_analyzer_eval(ferror(F));            // expected-warning {{TRUE}}
+  fclose(F);
+}
+
 void error_fflush_after_fclose(void) {
   FILE *F = tmpfile();
   int Ret;
diff --git a/clang/test/Analysis/stream.c b/clang/test/Analysis/stream.c
index d8026247697a5..e8f06922bdb2f 100644
--- a/clang/test/Analysis/stream.c
+++ b/clang/test/Analysis/stream.c
@@ -39,6 +39,12 @@ void check_fputs(void) {
   fclose(fp);
 }
 
+void check_fprintf(void) {
+  FILE *fp = tmpfile();
+  fprintf(fp, "ABC"); // expected-warning {{Stream pointer might be NULL}}
+  fclose(fp);
+}
+
 void check_ungetc(void) {
   FILE *fp = tmpfile();
   ungetc('A', fp); // expected-warning {{Stream pointer might be NULL}}
@@ -271,9 +277,8 @@ void check_escape4(void) {
     return;
   fwrite("1", 1, 1, F); // may fail
 
-  // no escape at (non-StreamChecker-handled) system call
-  // FIXME: all such calls should be handled by the checker
-  fprintf(F, "0");
+  // no escape at a non-StreamChecker-handled system call
+  setbuf(F, "0");
 
   fwrite("1", 1, 1, F); // expected-warning {{might be 'indeterminate'}}
   fclose(F);
diff --git a/clang/test/Analysis/string.c b/clang/test/Analysis/string.c
index d47de9db8228e..85232624160c0 100644
--- a/clang/test/Analysis/string.c
+++ b/clang/test/Analysis/string.c
@@ -71,7 +71,7 @@ void clang_analyzer_eval(int);
 int scanf(const char *restrict format, ...);
 void *malloc(size_t);
 void free(void *);
-void *memcpy(void *dest, const void *src, size_t n);
+void *memcpy(void *restrict dest, const void *restrict src, size_t n);
 
 //===----------------------------------------------------------------------===
 // strlen()
@@ -1252,7 +1252,7 @@ int strncasecmp_null_argument(char *a, size_t n) {
 // strsep()
 //===----------------------------------------------------------------------===
 
-char *strsep(char **stringp, const char *delim);
+char *strsep(char ** restrict stringp, const char * restrict delim);
 
 void strsep_null_delim(char *s) {
   strsep(&s, NULL); // expected-warning{{Null pointer passed as 2nd argument to strsep()}}
diff --git a/clang/test/CXX/dcl.decl/dcl.decl.general/p4-20.cpp b/clang/test/CXX/dcl.decl/dcl.decl.general/p4-20.cpp
index 83ec78dac9fe3..c7596218db537 100644
--- a/clang/test/CXX/dcl.decl/dcl.decl.general/p4-20.cpp
+++ b/clang/test/CXX/dcl.decl/dcl.decl.general/p4-20.cpp
@@ -27,7 +27,7 @@ auto *p = new void(*)(char)
 namespace GH61748 {
 template
 struct S {
-  // expected-error@+1 {{non-templated function cannot have a requires clause}}
+  // expected-error@+1 {{non-template friend declaration with a requires clause must be a definition}}
   friend void declared_friend() requires(sizeof(T) > 1);
   // OK, is a definition.
   friend void defined_friend() requires(sizeof(T) > 1){}
diff --git a/clang/test/CXX/drs/dr0xx.cpp b/clang/test/CXX/drs/dr0xx.cpp
index e5f675fb44898..5959f0a0c8dd6 100644
--- a/clang/test/CXX/drs/dr0xx.cpp
+++ b/clang/test/CXX/drs/dr0xx.cpp
@@ -30,7 +30,7 @@ namespace dr1 { // dr1: no
   namespace X {
     void z(int = 1);
     // expected-error@-1 {{redefinition of default argument}}
-    // expected-note@#dr1-z {{previous definition is here}}
+    //   expected-note@#dr1-z {{previous definition is here}}
   }
 
   void i(int = 1);
@@ -56,7 +56,7 @@ namespace dr3 { // dr3: yes
   template void f(int);
   template<> struct A {};
   // expected-error@-1 {{explicit specialization of 'dr3::A' after instantiation}}
-  // expected-note@#dr3-f-T {{implicit instantiation first required here}}
+  //   expected-note@#dr3-f-T {{implicit instantiation first required here}}
 }
 
 namespace dr4 { // dr4: 2.8
@@ -66,7 +66,7 @@ namespace dr4 { // dr4: 2.8
     void dr4_g(int) {} // #dr4-g-int
     void dr4_g(float) {}
     // expected-error@-1 {{conflicting types for 'dr4_g'}}
-    // expected-note@#dr4-g-int {{previous definition is here}}
+    //   expected-note@#dr4-g-int {{previous definition is here}}
   }
 }
 
@@ -99,7 +99,7 @@ namespace dr7 { // dr7: 3.4
   class X { ~X(); }; // #dr7-X
   class Y : X { ~Y() {} };
   // expected-error@-1 {{base class 'X' has private destructor}}
-  // expected-note@#dr7-X {{implicitly declared private here}}
+  //   expected-note@#dr7-X {{implicitly declared private here}}
 
   namespace PR16370 { // This regressed the first time DR7 was fixed.
     struct S1 { virtual ~S1(); };
@@ -137,8 +137,8 @@ namespace dr9 { // dr9: 2.8
   } n;
   int R1() { return n.m; }
   // expected-error@-1 {{'m' is a protected member of 'dr9::B'}}
-  // expected-note@#dr9-N {{constrained by protected inheritance here}}
-  // expected-note@#dr9-m {{member is declared here}}
+  //   expected-note@#dr9-N {{constrained by protected inheritance here}}
+  //   expected-note@#dr9-m {{member is declared here}}
   int R2() { return n.m; }
 }
 
@@ -211,15 +211,15 @@ namespace dr14 { // dr14: 3.4
   T t; // ok, same type both times
   U u;
   // expected-error@-1 {{reference to 'U' is ambiguous}}
-  // expected-note@#dr14-X-U {{candidate found by name lookup is 'dr14::X::U'}}
-  // expected-note@#dr14-Y-U {{candidate found by name lookup is 'dr14::Y::U'}}
+  //   expected-note@#dr14-X-U {{candidate found by name lookup is 'dr14::X::U'}}
+  //   expected-note@#dr14-Y-U {{candidate found by name lookup is 'dr14::Y::U'}}
 }
 
 namespace dr15 { // dr15: yes
   template void f(int); // #dr15-f-decl-first
   template void f(int = 0);
   // expected-error@-1 {{default arguments cannot be added to a function template that has already been declared}}
-  // expected-note@#dr15-f-decl-first {{previous template declaration is here}}
+  //   expected-note@#dr15-f-decl-first {{previous template declaration is here}}
 }
 
 namespace dr16 { // dr16: 2.8
@@ -232,8 +232,8 @@ namespace dr16 { // dr16: 2.8
     void g() {
       f();
       // expected-error@-1 {{'f' is a private member of 'dr16::A'}}
-      // expected-note@#dr16-B {{constrained by implicitly private inheritance here}}
-      // expected-note@#dr16-A-f-decl {{member is declared here}}
+      //   expected-note@#dr16-B {{constrained by implicitly private inheritance here}}
+      //   expected-note@#dr16-A-f-decl {{member is declared here}}
       A::f(); // #dr16-A-f-call
       // expected-error@#dr16-A-f-call {{'A' is a private member of 'dr16::A'}}
       //   expected-note@#dr16-B {{constrained by implicitly private inheritance here}}
@@ -269,8 +269,8 @@ namespace dr19 { // dr19: 3.1
   struct D : B {
     int get1() { return c.n; }
     // expected-error@-1 {{'n' is a protected member of 'dr19::A'}}
-    // expected-note@#dr19-B {{constrained by protected inheritance here}}
-    // expected-note@#dr19-n {{member is declared here}}
+    //   expected-note@#dr19-B {{constrained by protected inheritance here}}
+    //   expected-note@#dr19-n {{member is declared here}}
     int get2() { return ((A&)c).n; } // ok, A is an accessible base of B from here
   };
 }
@@ -285,7 +285,7 @@ namespace dr20 { // dr20: 2.8
   X &f();
   X x = f();
   // expected-error@-1 {{calling a private constructor of class 'dr20::X'}}
-  // expected-note@#dr20-X-ctor {{declared private here}}
+  //   expected-note@#dr20-X-ctor {{declared private here}}
 }
 
 namespace dr21 { // dr21: 3.4
@@ -310,8 +310,8 @@ namespace dr23 { // dr23: yes
   template void f(T, int); // #dr23-f-T-int
   void g() { f(0, 0); }
   // expected-error@-1 {{call to 'f' is ambiguous}}
-  // expected-note@#dr23-f-T-T {{candidate function [with T = int]}}
-  // expected-note@#dr23-f-T-int {{candidate function [with T = int]}}
+  //   expected-note@#dr23-f-T-T {{candidate function [with T = int]}}
+  //   expected-note@#dr23-f-T-int {{candidate function [with T = int]}}
 }
 
 // dr24: na
@@ -320,25 +320,25 @@ namespace dr25 { // dr25: yes
   struct A {
     void f() throw(int);
     // since-cxx17-error@-1 {{ISO C++17 does not allow dynamic exception specifications}}
-    // since-cxx17-note@-2 {{use 'noexcept(false)' instead}}
+    //   since-cxx17-note@-2 {{use 'noexcept(false)' instead}}
   };
   void (A::*f)() throw (int);
   // since-cxx17-error@-1 {{ISO C++17 does not allow dynamic exception specifications}}
-  // since-cxx17-note@-2 {{use 'noexcept(false)' instead}}
+  //   since-cxx17-note@-2 {{use 'noexcept(false)' instead}}
   void (A::*g)() throw () = f;
   // cxx98-14-error@-1 {{target exception specification is not superset of source}}
   // since-cxx17-error@-2 {{different exception specifications}}
   void (A::*g2)() throw () = 0;
   void (A::*h)() throw (int, char) = f;
   // since-cxx17-error@-1 {{ISO C++17 does not allow dynamic exception specifications}}
-  // since-cxx17-note@-2 {{use 'noexcept(false)' instead}}
+  //   since-cxx17-note@-2 {{use 'noexcept(false)' instead}}
   void (A::*i)() throw () = &A::f;
   // cxx98-14-error@-1 {{target exception specification is not superset of source}}
   // since-cxx17-error@-2 {{different exception specifications}}
   void (A::*i2)() throw () = 0;
   void (A::*j)() throw (int, char) = &A::f;
   // since-cxx17-error@-1 {{ISO C++17 does not allow dynamic exception specifications}}
-  // since-cxx17-note@-2 {{use 'noexcept(false)' instead}}
+  //   since-cxx17-note@-2 {{use 'noexcept(false)' instead}}
   void x() {
     g2 = f;
     // cxx98-14-error@-1 {{target exception specification is not superset of source}}
@@ -367,7 +367,7 @@ namespace dr26 { // dr26: yes
     static C &f();
     C(const C &, C = f());
     // expected-error@-1 {{recursive evaluation of default argument}}
-    // expected-note@-2 {{default argument used here}}
+    //   expected-note@-2 {{default argument used here}}
   };
 }
 
@@ -384,29 +384,29 @@ namespace dr29 { // dr29: 3.4
   extern "C++" void g0_cxx() { void dr29_f0(); }
   extern "C" void g0_c() { void dr29_f0(); }
   // expected-error@-1 {{declaration of 'dr29_f0' has a different language linkage}}
-  // expected-note@#dr29-f0 {{previous declaration is here}}
+  //   expected-note@#dr29-f0 {{previous declaration is here}}
 
   extern "C" void dr29_f1(); // #dr29-f1
   void g1() { void dr29_f1(); }
   extern "C" void g1_c() { void dr29_f1(); }
   extern "C++" void g1_cxx() { void dr29_f1(); }
   // expected-error@-1 {{declaration of 'dr29_f1' has a different language linkage}}
-  // expected-note@#dr29-f1 {{previous declaration is here}}
+  //   expected-note@#dr29-f1 {{previous declaration is here}}
 
   void g2() { void dr29_f2(); } // #dr29-f2
   extern "C" void dr29_f2();
   // expected-error@-1 {{declaration of 'dr29_f2' has a different language linkage}}
-  // expected-note@#dr29-f2 {{previous declaration is here}}
+  //   expected-note@#dr29-f2 {{previous declaration is here}}
 
   extern "C" void g3() { void dr29_f3(); } // #dr29-f3
   extern "C++" void dr29_f3();
   // expected-error@-1 {{declaration of 'dr29_f3' has a different language linkage}}
-  // expected-note@#dr29-f3 {{previous declaration is here}}
+  //   expected-note@#dr29-f3 {{previous declaration is here}}
 
   extern "C++" void g4() { void dr29_f4(); } // #dr29-f4
   extern "C" void dr29_f4();
   // expected-error@-1 {{declaration of 'dr29_f4' has a different language linkage}}
-  // expected-note@#dr29-f4 {{previous declaration is here}}
+  //   expected-note@#dr29-f4 {{previous declaration is here}}
 
   extern "C" void g5();
   extern "C++" void dr29_f5();
@@ -425,7 +425,7 @@ namespace dr29 { // dr29: 3.4
   extern "C" void g7() {
     void dr29_f7();
     // expected-error@-1 {{declaration of 'dr29_f7' has a different language linkage}}
-    // expected-note@#dr29-f7 {{previous declaration is here}}
+    //   expected-note@#dr29-f7 {{previous declaration is here}}
   }
 
   extern "C++" void g8();
@@ -433,7 +433,7 @@ namespace dr29 { // dr29: 3.4
   extern "C++" void g8() {
     void dr29_f8();
     // expected-error@-1 {{declaration of 'dr29_f8' has a different language linkage}}
-    // expected-note@#dr29-f8 {{previous declaration is here}}
+    //   expected-note@#dr29-f8 {{previous declaration is here}}
   }
 }
 
@@ -459,7 +459,7 @@ namespace dr31 { // dr31: 2.8
   // and even though we allocated the X using ::operator delete).
   X *p = new X;
   // expected-error@-1 {{'operator delete' is a private member of 'dr31::X'}}
-  // expected-note@#dr31-delete {{declared private here}}
+  //   expected-note@#dr31-delete {{declared private here}}
 }
 
 // dr32: na
@@ -471,8 +471,8 @@ namespace dr33 { // dr33: 9
   template Z g(Y::T);
   void h() { f(&g); }
   // expected-error@-1 {{call to 'f' is ambiguous}}
-  // expected-note@#dr33-f-S {{candidate function}}
-  // expected-note@#dr33-f-T {{candidate function}}
+  //   expected-note@#dr33-f-S {{candidate function}}
+  //   expected-note@#dr33-f-T {{candidate function}}
 
   template void t(X::S);
   template void u(X::S);
@@ -543,22 +543,22 @@ namespace example2 {
     using B::i; // #dr36-ex2-B-i-first
     using B::i;
     // expected-error@-1 {{redeclaration of using declaration}}
-    // expected-note@#dr36-ex2-B-i-first {{previous using declaration}}
+    //   expected-note@#dr36-ex2-B-i-first {{previous using declaration}}
 
     using C::i; // #dr36-ex2-C-i-first
     using C::i;
     // expected-error@-1 {{redeclaration of using declaration}}
-    // expected-note@#dr36-ex2-C-i-first {{previous using declaration}}
+    //   expected-note@#dr36-ex2-C-i-first {{previous using declaration}}
 
     using B::j; // #dr36-ex2-B-j-first
     using B::j;
     // expected-error@-1 {{redeclaration of using declaration}}
-    // expected-note@#dr36-ex2-B-j-first {{previous using declaration}}
+    //   expected-note@#dr36-ex2-B-j-first {{previous using declaration}}
 
     using C::j; // #dr36-ex2-C-j-first
     using C::j;
     // expected-error@-1 {{redeclaration of using declaration}}
-    // expected-note@#dr36-ex2-C-j-first {{previous using declaration}}
+    //   expected-note@#dr36-ex2-C-j-first {{previous using declaration}}
   };
 }
 
@@ -581,22 +581,22 @@ namespace example3 {
     using B::i; // #dr36-ex3-B-i-first
     using B::i;
     // expected-error@-1 {{redeclaration of using declaration}}
-    // expected-note@#dr36-ex3-B-i-first {{previous using declaration}}
+    //   expected-note@#dr36-ex3-B-i-first {{previous using declaration}}
 
     using C::i; // #dr36-ex3-C-i-first
     using C::i;
     // expected-error@-1 {{redeclaration of using declaration}}
-    // expected-note@#dr36-ex3-C-i-first {{previous using declaration}}
+    //   expected-note@#dr36-ex3-C-i-first {{previous using declaration}}
 
     using B::j; // #dr36-ex3-B-j-first
     using B::j;
     // expected-error@-1 {{redeclaration of using declaration}}
-    // expected-note@#dr36-ex3-B-j-first {{previous using declaration}}
+    //   expected-note@#dr36-ex3-B-j-first {{previous using declaration}}
 
     using C::j; // #dr36-ex3-C-j-first
     using C::j;
     // expected-error@-1 {{redeclaration of using declaration}}
-    // expected-note@#dr36-ex3-C-j-first {{previous using declaration}}
+    //   expected-note@#dr36-ex3-C-j-first {{previous using declaration}}
   };
 }
 namespace example4 {
@@ -610,7 +610,7 @@ namespace example4 {
     using E::k; // #dr36-E-k-first
     using E::k;
     // expected-error@-1 {{redeclaration of using declaration}}
-    // expected-note@#dr36-E-k-first {{previous using declaration}}
+    //   expected-note@#dr36-E-k-first {{previous using declaration}}
   };
 }
 }
@@ -656,14 +656,14 @@ namespace dr39 { // dr39: no
     struct dr39::example2::C -> B -> A}} */
     int &x = c.x(0);
     // expected-error@-1 {{member 'x' found in multiple base classes of different types}}
-    // expected-note@#dr39-A-x-decl {{member found by ambiguous name lookup}}
-    // expected-note@#dr39-using-A-x {{member found by ambiguous name lookup}}
+    //   expected-note@#dr39-A-x-decl {{member found by ambiguous name lookup}}
+    //   expected-note@#dr39-using-A-x {{member found by ambiguous name lookup}}
 
     // FIXME: This is valid, because we find the same static data member either way.
     int &y = c.y(0);
     // expected-error@-1 {{member 'y' found in multiple base classes of different types}}
-    // expected-note@#dr39-A-y-decl {{member found by ambiguous name lookup}}
-    // expected-note@#dr39-using-A-y {{member found by ambiguous name lookup}}
+    //   expected-note@#dr39-A-y-decl {{member found by ambiguous name lookup}}
+    //   expected-note@#dr39-using-A-y {{member found by ambiguous name lookup}}
     int &z = c.z(0);
   }
 
@@ -684,7 +684,7 @@ namespace dr39 { // dr39: no
     {{non-static member 'n' found in multiple base-class subobjects of type 'A':
     struct dr39::example4::D -> B -> A
     struct dr39::example4::D -> C -> A}} */
-    // expected-note@#dr39-ex4-A-n {{member found by ambiguous name lookup}}
+    //   expected-note@#dr39-ex4-A-n {{member found by ambiguous name lookup}}
   }
 
   namespace PR5916 {
@@ -698,7 +698,7 @@ namespace dr39 { // dr39: no
     {{non-static member 'n' found in multiple base-class subobjects of type 'A':
     struct dr39::PR5916::D -> B -> A
     struct dr39::PR5916::D -> C -> A}} */
-    // expected-note@#dr39-A-n {{member found by ambiguous name lookup}}
+    //   expected-note@#dr39-A-n {{member found by ambiguous name lookup}}
 
     // expected-error@#dr39-sizeof {{unknown type name}}
 #if __cplusplus >= 201103L
@@ -707,7 +707,7 @@ namespace dr39 { // dr39: no
     {{non-static member 'n' found in multiple base-class subobjects of type 'A':
     struct dr39::PR5916::D -> B -> A
     struct dr39::PR5916::D -> C -> A}} */
-    // expected-note@#dr39-A-n {{member found by ambiguous name lookup}}
+    //   expected-note@#dr39-A-n {{member found by ambiguous name lookup}}
 #endif
   }
 }
@@ -750,8 +750,8 @@ namespace dr47 { // dr47: sup 329
   template struct A {
     friend void f() { T t; } // #dr47-f
     // expected-error@-1 {{redefinition of 'f'}}
-    // expected-note@#dr47-b {{in instantiation of template class 'dr47::A' requested here}}
-    // expected-note@#dr47-f {{previous definition is here}}
+    //   expected-note@#dr47-b {{in instantiation of template class 'dr47::A' requested here}}
+    //   expected-note@#dr47-f {{previous definition is here}}
   };
   A a;
   A b; // #dr47-b
@@ -811,7 +811,7 @@ namespace dr50 { // dr50: yes
   X *t = reinterpret_cast(p);
   X *u = dynamic_cast(p);
   // expected-error@-1 {{'dr50::X' is an incomplete type}}
-  // expected-note@#dr50-X {{forward declaration of 'dr50::X'}}
+  //   expected-note@#dr50-X {{forward declaration of 'dr50::X'}}
 }
 
 namespace dr51 { // dr51: 2.8
@@ -848,22 +848,22 @@ namespace dr54 { // dr54: 2.8
 
   A &sab = static_cast(b);
   // expected-error@-1 {{cannot cast 'struct B' to its private base class 'A'}}
-  // expected-note@#dr54-B {{declared private here}}
+  //   expected-note@#dr54-B {{declared private here}}
   A *spab = static_cast(&b);
   // expected-error@-1 {{cannot cast 'struct B' to its private base class 'A'}}
-  // expected-note@#dr54-B {{declared private here}}
+  //   expected-note@#dr54-B {{declared private here}}
   int A::*smab = static_cast(&B::b);
   // expected-error@-1 {{cannot cast 'dr54::B' to its private base class 'dr54::A'}}
-  // expected-note@#dr54-B {{declared private here}}
+  //   expected-note@#dr54-B {{declared private here}}
   B &sba = static_cast(a);
   // expected-error@-1 {{cannot cast private base class 'dr54::A' to 'dr54::B'}}
-  // expected-note@#dr54-B {{declared private here}}
+  //   expected-note@#dr54-B {{declared private here}}
   B *spba = static_cast(&a);
   // expected-error@-1 {{cannot cast private base class 'dr54::A' to 'dr54::B'}}
-  // expected-note@#dr54-B {{declared private here}}
+  //   expected-note@#dr54-B {{declared private here}}
   int B::*smba = static_cast(&A::a);
   // expected-error@-1 {{cannot cast private base class 'dr54::A' to 'dr54::B'}}
-  // expected-note@#dr54-B {{declared private here}}
+  //   expected-note@#dr54-B {{declared private here}}
 
   V &svb = static_cast(b);
   V *spvb = static_cast(&b);
@@ -905,14 +905,14 @@ namespace dr56 { // dr56: yes
     typedef int T; // #dr56-typedef-int-T-first
     typedef int T;
     // expected-error@-1 {{redefinition of 'T'}}
-    // expected-note@#dr56-typedef-int-T-first {{previous definition is here}}
+    //   expected-note@#dr56-typedef-int-T-first {{previous definition is here}}
   };
   struct B {
     struct X;
     typedef X X; // #dr56-typedef-X-X-first
     typedef X X;
     // expected-error@-1 {{redefinition of 'X'}}
-    // expected-note@#dr56-typedef-X-X-first {{previous definition is here}}
+    //   expected-note@#dr56-typedef-X-X-first {{previous definition is here}}
   };
 }
 
@@ -939,42 +939,42 @@ namespace dr59 { // dr59: yes
   A a3 = convert_to();
   A a4 = convert_to();
   // cxx98-14-error@-1 {{no viable constructor copying variable of type 'const volatile dr59::A'}}
-  // cxx98-14-note@#dr59-A {{candidate constructor (the implicit copy constructor) not viable: 1st argument ('const volatile dr59::A') would lose volatile qualifier}}
-  // cxx11-14-note@#dr59-A {{candidate constructor (the implicit move constructor) not viable: 1st argument ('const volatile dr59::A') would lose const and volatile qualifiers}}
-  // cxx98-14-note@#dr59-A {{candidate constructor (the implicit default constructor) not viable: requires 0 arguments, but 1 was provided}}
+  //   cxx98-14-note@#dr59-A {{candidate constructor (the implicit copy constructor) not viable: 1st argument ('const volatile dr59::A') would lose volatile qualifier}}
+  //   cxx11-14-note@#dr59-A {{candidate constructor (the implicit move constructor) not viable: 1st argument ('const volatile dr59::A') would lose const and volatile qualifiers}}
+  //   cxx98-14-note@#dr59-A {{candidate constructor (the implicit default constructor) not viable: requires 0 arguments, but 1 was provided}}
   A a5 = convert_to();
   // expected-error@-1 {{no viable constructor copying variable of type 'const volatile dr59::A'}}
-  // expected-note@#dr59-A {{candidate constructor (the implicit copy constructor) not viable: 1st argument ('const volatile dr59::A') would lose volatile qualifier}}
-  // since-cxx11-note@#dr59-A {{candidate constructor (the implicit move constructor) not viable: 1st argument ('const volatile dr59::A') would lose const and volatile qualifiers}}
-  // expected-note@#dr59-A {{candidate constructor (the implicit default constructor) not viable: requires 0 arguments, but 1 was provided}}
+  //   expected-note@#dr59-A {{candidate constructor (the implicit copy constructor) not viable: 1st argument ('const volatile dr59::A') would lose volatile qualifier}}
+  //   since-cxx11-note@#dr59-A {{candidate constructor (the implicit move constructor) not viable: 1st argument ('const volatile dr59::A') would lose const and volatile qualifiers}}
+  //   expected-note@#dr59-A {{candidate constructor (the implicit default constructor) not viable: requires 0 arguments, but 1 was provided}}
 
   B b1 = convert_to();
   B b2 = convert_to();
   B b3 = convert_to();
   B b4 = convert_to();
   // cxx98-14-error@-1 {{no viable constructor copying variable of type 'const volatile dr59::B'}}
-  // cxx98-14-note@#dr59-B {{candidate constructor (the implicit copy constructor) not viable: 1st argument ('const volatile dr59::B') would lose volatile qualifier}}
-  // cxx11-14-note@#dr59-B {{candidate constructor (the implicit move constructor) not viable: 1st argument ('const volatile dr59::B') would lose const and volatile qualifiers}}
-  // cxx98-14-note@#dr59-B {{candidate constructor (the implicit default constructor) not viable: requires 0 arguments, but 1 was provided}}
+  //   cxx98-14-note@#dr59-B {{candidate constructor (the implicit copy constructor) not viable: 1st argument ('const volatile dr59::B') would lose volatile qualifier}}
+  //   cxx11-14-note@#dr59-B {{candidate constructor (the implicit move constructor) not viable: 1st argument ('const volatile dr59::B') would lose const and volatile qualifiers}}
+  //   cxx98-14-note@#dr59-B {{candidate constructor (the implicit default constructor) not viable: requires 0 arguments, but 1 was provided}}
   B b5 = convert_to();
   // expected-error@-1 {{no viable constructor copying variable of type 'const volatile dr59::B'}}
-  // expected-note@#dr59-B {{candidate constructor (the implicit copy constructor) not viable: 1st argument ('const volatile dr59::B') would lose volatile qualifier}}
-  // since-cxx11-note@#dr59-B {{candidate constructor (the implicit move constructor) not viable: 1st argument ('const volatile dr59::B') would lose const and volatile qualifiers}}
-  // expected-note@#dr59-B {{candidate constructor (the implicit default constructor) not viable: requires 0 arguments, but 1 was provided}}
+  //   expected-note@#dr59-B {{candidate constructor (the implicit copy constructor) not viable: 1st argument ('const volatile dr59::B') would lose volatile qualifier}}
+  //   since-cxx11-note@#dr59-B {{candidate constructor (the implicit move constructor) not viable: 1st argument ('const volatile dr59::B') would lose const and volatile qualifiers}}
+  //   expected-note@#dr59-B {{candidate constructor (the implicit default constructor) not viable: requires 0 arguments, but 1 was provided}}
 
   A c1 = convert_to();
   A c2 = convert_to();
   A c3 = convert_to();
   A c4 = convert_to();
   // expected-error@-1 {{no viable constructor copying variable of type 'const volatile dr59::B'}}
-  // expected-note@#dr59-A {{candidate constructor (the implicit copy constructor) not viable: no known conversion from 'const volatile dr59::B' to 'const A &' for 1st argument}}
-  // since-cxx11-note@#dr59-A {{candidate constructor (the implicit move constructor) not viable: no known conversion from 'const volatile dr59::B' to 'A &&' for 1st argument}}
-  // expected-note@#dr59-A {{candidate constructor (the implicit default constructor) not viable: requires 0 arguments, but 1 was provided}}
+  //   expected-note@#dr59-A {{candidate constructor (the implicit copy constructor) not viable: no known conversion from 'const volatile dr59::B' to 'const A &' for 1st argument}}
+  //   since-cxx11-note@#dr59-A {{candidate constructor (the implicit move constructor) not viable: no known conversion from 'const volatile dr59::B' to 'A &&' for 1st argument}}
+  //   expected-note@#dr59-A {{candidate constructor (the implicit default constructor) not viable: requires 0 arguments, but 1 was provided}}
   A c5 = convert_to();
   // expected-error@-1 {{no viable constructor copying variable of type 'const volatile dr59::B'}}
-  // expected-note@#dr59-A {{candidate constructor (the implicit copy constructor) not viable: no known conversion from 'const volatile dr59::B' to 'const A &' for 1st argument}}
-  // since-cxx11-note@#dr59-A {{candidate constructor (the implicit move constructor) not viable: no known conversion from 'const volatile dr59::B' to 'A &&' for 1st argument}}
-  // expected-note@#dr59-A {{candidate constructor (the implicit default constructor) not viable: requires 0 arguments, but 1 was provided}}
+  //   expected-note@#dr59-A {{candidate constructor (the implicit copy constructor) not viable: no known conversion from 'const volatile dr59::B' to 'const A &' for 1st argument}}
+  //   since-cxx11-note@#dr59-A {{candidate constructor (the implicit move constructor) not viable: no known conversion from 'const volatile dr59::B' to 'A &&' for 1st argument}}
+  //   expected-note@#dr59-A {{candidate constructor (the implicit default constructor) not viable: requires 0 arguments, but 1 was provided}}
 
   int n1 = convert_to();
   int n2 = convert_to();
@@ -1028,19 +1028,19 @@ namespace dr62 { // dr62: 2.9
 
   X x2;
   // cxx98-error@-1 {{template argument uses unnamed type}}
-  // cxx98-note@#dr62-unnamed {{unnamed type used in template argument was declared here}}
+  //   cxx98-note@#dr62-unnamed {{unnamed type used in template argument was declared here}}
   X x3;
   // cxx98-error@-1 {{template argument uses unnamed type}}
-  // cxx98-note@#dr62-unnamed {{unnamed type used in template argument was declared here}}
+  //   cxx98-note@#dr62-unnamed {{unnamed type used in template argument was declared here}}
   NoNameForLinkagePtr p1 = get();
   // cxx98-error@-1 {{template argument uses unnamed type}}
-  // cxx98-note@#dr62-unnamed {{unnamed type used in template argument was declared here}}
+  //   cxx98-note@#dr62-unnamed {{unnamed type used in template argument was declared here}}
   NoNameForLinkagePtr p2 = get();
   // cxx98-error@-1 {{template argument uses unnamed type}}
-  // cxx98-note@#dr62-unnamed {{unnamed type used in template argument was declared here}}
+  //   cxx98-note@#dr62-unnamed {{unnamed type used in template argument was declared here}}
   int n1 = take(noNameForLinkagePtr);
   // cxx98-error@-1 {{template argument uses unnamed type}}
-  // cxx98-note@#dr62-unnamed {{unnamed type used in template argument was declared here}}
+  //   cxx98-note@#dr62-unnamed {{unnamed type used in template argument was declared here}}
 
   X x4;
 
@@ -1089,11 +1089,11 @@ namespace dr66 { // dr66: no
   // FIXME: The first two calls here should be accepted.
   int a = f();
   // expected-error@-1 {{no matching function for call to 'f'}}
-  // expected-note@#dr66-f-first {{candidate function not viable: requires single argument 'n', but no arguments were provided}}
+  //   expected-note@#dr66-f-first {{candidate function not viable: requires single argument 'n', but no arguments were provided}}
   int b = f(1);
   int c = f(1, 2);
   // expected-error@-1 {{no matching function for call to 'f'}}
-  // expected-note@#dr66-f-first {{candidate function not viable: requires single argument 'n', but 2 arguments were provided}}
+  //   expected-note@#dr66-f-first {{candidate function not viable: requires single argument 'n', but 2 arguments were provided}}
 }
 
 // dr67: na
@@ -1127,7 +1127,7 @@ namespace dr69 { // dr69: 9
   template struct Q {};
   Q<&f > q;
   // cxx98-error@-1 {{non-type template argument referring to function 'f' with internal linkage is a C++11 extension}}
-  // cxx98-note@#dr69-f {{non-type template argument refers to function here}}
+  //   cxx98-note@#dr69-f {{non-type template argument refers to function here}}
 }
 
 namespace dr70 { // dr70: yes
@@ -1145,7 +1145,7 @@ namespace dr73 { // dr73: sup 1652
   int a, b;
   static_assert(&a + 1 != &b, "");
   // expected-error@-1 {{static assertion expression is not an integral constant expression}}
-  // expected-note@-2 {{comparison against pointer '&a + 1' that points past the end of a complete object has unspecified value}}
+  //   expected-note@-2 {{comparison against pointer '&a + 1' that points past the end of a complete object has unspecified value}}
 }
 #endif
 
@@ -1228,8 +1228,8 @@ namespace dr84 { // dr84: yes
   // here. In C++17, we initialize the B object directly using 'A::operator B()'.
   B b = a;
   // cxx98-14-error@-1 {{no viable constructor copying variable of type 'B'}}
-  // cxx98-14-note@#dr84-copy-ctor {{candidate constructor not viable: expects an lvalue for 1st argument}}
-  // cxx98-14-note@#dr84-ctor-from-C {{candidate constructor not viable: no known conversion from 'B' to 'C' for 1st argument}}
+  //   cxx98-14-note@#dr84-copy-ctor {{candidate constructor not viable: expects an lvalue for 1st argument}}
+  //   cxx98-14-note@#dr84-ctor-from-C {{candidate constructor not viable: no known conversion from 'B' to 'C' for 1st argument}}
 }
 
 namespace dr85 { // dr85: 3.4
@@ -1238,26 +1238,26 @@ namespace dr85 { // dr85: 3.4
     struct B {}; // #dr85-B-def
     struct B;
     // expected-error@-1 {{class member cannot be redeclared}}
-    // expected-note@#dr85-B-def {{previous declaration is here}}
+    //   expected-note@#dr85-B-def {{previous declaration is here}}
 
     union U;
     union U {}; // #dr85-U-def
     union U;
     // expected-error@-1 {{class member cannot be redeclared}}
-    // expected-note@#dr85-U-def {{previous declaration is here}}
+    //   expected-note@#dr85-U-def {{previous declaration is here}}
 
 #if __cplusplus >= 201103L
     enum E1 : int;
     enum E1 : int { e1 }; // #dr85-E1-def
     enum E1 : int;
     // expected-error@-1 {{class member cannot be redeclared}}
-    // expected-note@#dr85-E1-def {{previous declaration is here}}
+    //   expected-note@#dr85-E1-def {{previous declaration is here}}
 
     enum class E2;
     enum class E2 { e2 }; // #dr85-E2-def
     enum class E2;
     // expected-error@-1 {{class member cannot be redeclared}}
-    // expected-note@#dr85-E2-def {{previous declaration is here}}
+    //   expected-note@#dr85-E2-def {{previous declaration is here}}
 #endif
   };
 
@@ -1266,7 +1266,7 @@ namespace dr85 { // dr85: 3.4
     struct B {}; // #dr85-C-B-def
     struct B;
     // expected-error@-1 {{class member cannot be redeclared}}
-    // expected-note@#dr85-C-B-def {{previous declaration is here}}
+    //   expected-note@#dr85-C-B-def {{previous declaration is here}}
   };
 }
 
@@ -1288,7 +1288,7 @@ namespace dr88 { // dr88: 2.8
   };
   template<> const int S::a = 4;
   // expected-error@-1 {{static data member 'a' already has an initializer}}
-  // expected-note@#dr88-a {{previous initialization is here}}
+  //   expected-note@#dr88-a {{previous initialization is here}}
   template<> const int S::b = 4;
 }
 
@@ -1336,7 +1336,7 @@ namespace dr91 { // dr91: yes
 namespace dr92 { // dr92: 4 c++17
   void f() throw(int, float);
   // since-cxx17-error@-1 {{ISO C++17 does not allow dynamic exception specifications}}
-  // since-cxx17-note@-2 {{use 'noexcept(false)' instead}}
+  //   since-cxx17-note@-2 {{use 'noexcept(false)' instead}}
   void (*p)() throw(int) = &f; // #dr92-p
   // since-cxx17-error@#dr92-p {{ISO C++17 does not allow dynamic exception specifications}}
   //   since-cxx17-note@#dr92-p {{use 'noexcept(false)' instead}}
@@ -1344,7 +1344,7 @@ namespace dr92 { // dr92: 4 c++17
   // since-cxx17-warning@#dr92-p {{target exception specification is not superset of source}}
   void (*q)() throw(int);
   // since-cxx17-error@-1 {{ISO C++17 does not allow dynamic exception specifications}}
-  // since-cxx17-note@-2 {{use 'noexcept(false)' instead}}
+  //   since-cxx17-note@-2 {{use 'noexcept(false)' instead}}
   void (**pp)() throw() = &q;
   // cxx98-14-error@-1 {{exception specifications are not allowed beyond a single level of indirection}}
   // since-cxx17-error@-2 {{cannot initialize a variable of type 'void (**)() throw()' with an rvalue of type 'void (**)() throw(int)'}}
@@ -1372,7 +1372,7 @@ namespace dr92 { // dr92: 4 c++17
 
   template struct Y {};
   // since-cxx17-error@-1 {{ISO C++17 does not allow dynamic exception specifications}}
-  // since-cxx17-note@-2 {{use 'noexcept(false)' instead}}
+  //   since-cxx17-note@-2 {{use 'noexcept(false)' instead}}
   Y<&h> yp; // ok
 }
 
@@ -1397,7 +1397,7 @@ namespace dr95 { // dr95: 3.3
   A *q = N::p; // ok, same type
   struct B { void f() { N::C::f(); } };
   // expected-error@-1 {{'f' is a private member of 'dr95::N::C'}}
-  // expected-note@#dr95-C-f {{implicitly declared private here}}
+  //   expected-note@#dr95-C-f {{implicitly declared private here}}
 }
 
 namespace dr96 { // dr96: no
@@ -1433,24 +1433,24 @@ namespace dr98 { // dr98: yes
       try { // #dr98-try
         case 0:
         // expected-error@-1 {{cannot jump from switch statement to this case label}}
-        // expected-note@#dr98-try {{jump bypasses initialization of try block}}
+        //   expected-note@#dr98-try {{jump bypasses initialization of try block}}
         x:
           throw n;
       } catch (...) { // #dr98-catch
         case 1:
         // expected-error@-1 {{cannot jump from switch statement to this case label}}
-        // expected-note@#dr98-catch {{jump bypasses initialization of catch block}}
+        //   expected-note@#dr98-catch {{jump bypasses initialization of catch block}}
         y:
           throw n;
       }
       case 2:
         goto x;
         // expected-error@-1 {{cannot jump from this goto statement to its label}}
-        // expected-note@#dr98-try {{jump bypasses initialization of try block}}
+        //   expected-note@#dr98-try {{jump bypasses initialization of try block}}
       case 3:
         goto y;
         // expected-error@-1 {{cannot jump from this goto statement to its label}}
-        // expected-note@#dr98-catch {{jump bypasses initialization of catch block}}
+        //   expected-note@#dr98-catch {{jump bypasses initialization of catch block}}
     }
   }
 }
diff --git a/clang/test/CXX/drs/dr13xx.cpp b/clang/test/CXX/drs/dr13xx.cpp
index 359c04b3e0f3d..0ff44ebbc9559 100644
--- a/clang/test/CXX/drs/dr13xx.cpp
+++ b/clang/test/CXX/drs/dr13xx.cpp
@@ -40,6 +40,8 @@ void caller() {
 #endif // __cplusplus >= 201103L
 } // namespace dr1307
 
+// dr1308: sup 1330
+
 namespace dr1310 { // dr1310: 5
   struct S {} * sp = new S::S;
   // expected-error@-1 {{qualified reference to 'S' is a constructor name rather than a type in this context}}
@@ -379,6 +381,77 @@ namespace dr1347 { // dr1347: 3.1
 #endif
 }
 
+namespace dr1350 { // dr1350: 3.5
+#if __cplusplus >= 201103L
+struct NoexceptCtor {
+  NoexceptCtor(int) noexcept {}
+};
+
+struct ThrowingNSDMI : NoexceptCtor {
+  int a = []() noexcept(false) { return 0; }();
+  using NoexceptCtor::NoexceptCtor;
+};
+
+static_assert(!__is_nothrow_constructible(ThrowingNSDMI, int), "");
+
+struct ThrowingCtor {
+  ThrowingCtor() noexcept(false) {}
+};
+
+struct ThrowingNSDM : NoexceptCtor {
+  ThrowingCtor c;
+  using NoexceptCtor::NoexceptCtor;
+};
+
+static_assert(!__is_nothrow_constructible(ThrowingNSDM, int), "");
+
+struct ThrowingCtorTemplate {
+  template 
+  ThrowingCtorTemplate() noexcept(false) {}
+};
+
+struct ThrowingNSDM2 : NoexceptCtor {
+  ThrowingCtorTemplate c;
+  using NoexceptCtor::NoexceptCtor;
+};
+
+static_assert(!__is_nothrow_constructible(ThrowingNSDM2, int), "");
+
+struct D1 : NoexceptCtor, ThrowingCtor {
+  using NoexceptCtor::NoexceptCtor;
+};
+
+static_assert(!__is_nothrow_constructible(D1, int), "");
+
+struct D2 : NoexceptCtor, ThrowingCtorTemplate {
+  using NoexceptCtor::NoexceptCtor;
+};
+
+static_assert(!__is_nothrow_constructible(D2, int), "");
+
+struct ThrowingDefaultArg {
+  ThrowingDefaultArg(ThrowingCtor = {}) {}
+};
+
+struct D3 : NoexceptCtor, ThrowingDefaultArg {
+  using NoexceptCtor::NoexceptCtor;
+};
+
+static_assert(!__is_nothrow_constructible(D3, int), "");
+
+struct ThrowingDefaultArgTemplate {
+  template 
+  ThrowingDefaultArgTemplate(ThrowingCtor = {}) {}
+};
+
+struct D4 : NoexceptCtor, ThrowingDefaultArgTemplate {
+  using NoexceptCtor::NoexceptCtor;
+};
+
+static_assert(!__is_nothrow_constructible(D4, int), "");
+#endif
+} // namespace dr1350
+
 namespace dr1358 { // dr1358: 3.1
 #if __cplusplus >= 201103L
   struct Lit { constexpr operator int() const { return 0; } };
diff --git a/clang/test/CXX/drs/dr15xx.cpp b/clang/test/CXX/drs/dr15xx.cpp
index 007b42c74affb..3d4050a5713f9 100644
--- a/clang/test/CXX/drs/dr15xx.cpp
+++ b/clang/test/CXX/drs/dr15xx.cpp
@@ -390,6 +390,13 @@ namespace dr1573 { // dr1573: 3.9
   H h(0);
   // since-cxx11-error@-1 {{constructor inherited by 'H' from base class 'G' is implicitly deleted}}
   //   since-cxx11-note@#dr1573-H {{constructor inherited by 'H' is implicitly deleted because field 'g' has no default constructor}}
+
+  // deleted definition of constructor is inherited
+  struct I { I(int) = delete; }; // #dr1573-I
+  struct J : I { using I::I; };
+  J j(0);
+  // since-cxx11-error@-1 {{call to deleted constructor of 'J'}}
+  //   since-cxx11-note@#dr1573-I {{'I' has been explicitly marked deleted here}}
 #endif
 }
 
diff --git a/clang/test/CXX/drs/dr16xx.cpp b/clang/test/CXX/drs/dr16xx.cpp
index 3f074c4d57354..1be4d2ab3ebe8 100644
--- a/clang/test/CXX/drs/dr16xx.cpp
+++ b/clang/test/CXX/drs/dr16xx.cpp
@@ -472,7 +472,7 @@ namespace dr1696 { // dr1696: 7
     const A &a = A(); // #dr1696-D1-a
   };
   D1 d1 = {}; // #dr1696-d1
-  // since-cxx14-warning@-1 {{sorry, lifetime extension of temporary created by aggregate initialization using default member initializer is not supported; lifetime of temporary will end at the end of the full-expression}}
+  // since-cxx14-warning@-1 {{lifetime extension of temporary created by aggregate initialization using a default member initializer is not yet supported; lifetime of temporary will end at the end of the full-expression}}
   //   since-cxx14-note@#dr1696-D1-a {{initializing field 'a' with default member initializer}}
 
   struct D2 {
diff --git a/clang/test/CXX/drs/dr1807.cpp b/clang/test/CXX/drs/dr1807.cpp
new file mode 100644
index 0000000000000..0e16eedac71ef
--- /dev/null
+++ b/clang/test/CXX/drs/dr1807.cpp
@@ -0,0 +1,28 @@
+// RUN: %clang_cc1 -std=c++98 %s -triple x86_64-linux-gnu -emit-llvm -o - -fexceptions -fcxx-exceptions -pedantic-errors | llvm-cxxfilt | FileCheck %s --check-prefixes CHECK,CXX98
+// RUN: %clang_cc1 -std=c++11 %s -triple x86_64-linux-gnu -emit-llvm -o - -fexceptions -fcxx-exceptions -pedantic-errors | llvm-cxxfilt | FileCheck %s --check-prefixes CHECK,SINCE-CXX11
+// RUN: %clang_cc1 -std=c++14 %s -triple x86_64-linux-gnu -emit-llvm -o - -fexceptions -fcxx-exceptions -pedantic-errors | llvm-cxxfilt | FileCheck %s --check-prefixes CHECK,SINCE-CXX11
+// RUN: %clang_cc1 -std=c++17 %s -triple x86_64-linux-gnu -emit-llvm -o - -fexceptions -fcxx-exceptions -pedantic-errors | llvm-cxxfilt | FileCheck %s --check-prefixes CHECK,SINCE-CXX11
+// RUN: %clang_cc1 -std=c++20 %s -triple x86_64-linux-gnu -emit-llvm -o - -fexceptions -fcxx-exceptions -pedantic-errors | llvm-cxxfilt | FileCheck %s --check-prefixes CHECK,SINCE-CXX11
+// RUN: %clang_cc1 -std=c++23 %s -triple x86_64-linux-gnu -emit-llvm -o - -fexceptions -fcxx-exceptions -pedantic-errors | llvm-cxxfilt | FileCheck %s --check-prefixes CHECK,SINCE-CXX11
+// RUN: %clang_cc1 -std=c++2c %s -triple x86_64-linux-gnu -emit-llvm -o - -fexceptions -fcxx-exceptions -pedantic-errors | llvm-cxxfilt | FileCheck %s --check-prefixes CHECK,SINCE-CXX11
+
+namespace dr1807 { // dr1807: 3.0
+struct S {
+  S() {}
+  ~S() {}
+};
+
+void f() {
+  S s[3];
+}
+}
+
+// CHECK-LABEL:      define dso_local void @dr1807::f()
+// CHECK:              invoke void @dr1807::S::S(){{.+}}
+// CHECK-NEXT:         {{.+}} unwind label %lpad
+// CHECK-LABEL:      lpad:
+// CHECK:              br {{.+}}, label {{.+}}, label %arraydestroy.body
+// CHECK-LABEL:      arraydestroy.body:         
+// CHECK:              [[ARRAYDESTROY_ELEMENT:%.*]] = getelementptr {{.+}}, i64 -1
+// CXX98-NEXT:         invoke void @dr1807::S::~S()({{.*}}[[ARRAYDESTROY_ELEMENT]])
+// SINCE-CXX11-NEXT:   call void @dr1807::S::~S()({{.*}}[[ARRAYDESTROY_ELEMENT]])
diff --git a/clang/test/CXX/drs/dr18xx.cpp b/clang/test/CXX/drs/dr18xx.cpp
index 1d76804907a5c..0245f03986dd7 100644
--- a/clang/test/CXX/drs/dr18xx.cpp
+++ b/clang/test/CXX/drs/dr18xx.cpp
@@ -164,6 +164,8 @@ void A::C::f4() {
 }
 } // namespace dr1804
 
+// dr1807 is in dr1807.cpp
+
 namespace dr1812 { // dr1812: no
                    // NB: dup 1710
 #if __cplusplus >= 201103L
@@ -209,7 +211,7 @@ namespace dr1815 { // dr1815: no
   // FIXME: needs codegen test
   struct A { int &&r = 0; }; // #dr1815-A 
   A a = {};
-  // since-cxx14-warning@-1 {{sorry, lifetime extension of temporary created by aggregate initialization using default member initializer is not supported; lifetime of temporary will end at the end of the full-expression}} FIXME
+  // since-cxx14-warning@-1 {{lifetime extension of temporary created by aggregate initialization using a default member initializer is not yet supported; lifetime of temporary will end at the end of the full-expression}} FIXME
   //   since-cxx14-note@#dr1815-A {{initializing field 'r' with default member initializer}}
 
   struct B { int &&r = 0; }; // #dr1815-B
@@ -329,6 +331,34 @@ namespace dr1881 { // dr1881: 7
   static_assert(!__is_standard_layout(D), "");
 }
 
+namespace dr1890 { // dr1890: no drafting
+// FIXME: current consensus for CWG2335 is that the examples are well-formed.
+namespace ex1 {
+#if __cplusplus >= 201402L
+struct A {
+  struct B {
+    auto foo() { return 0; } // #dr1890-foo
+  };
+  decltype(B().foo()) x;
+  // since-cxx14-error@-1 {{function 'foo' with deduced return type cannot be used before it is defined}}
+  //   since-cxx14-note@#dr1890-foo {{'foo' declared here}}
+};
+#endif
+} // namespace ex1
+
+namespace ex2 {
+#if __cplusplus >= 201103L
+struct Bar {
+  struct Baz {
+    int a = 0;
+  };
+  static_assert(__is_constructible(Baz), "");
+  // since-cxx11-error@-1 {{static assertion failed due to requirement '__is_constructible(dr1890::ex2::Bar::Baz)'}}
+};
+#endif
+} // namespace ex2
+} // namespace dr1890
+
 void dr1891() { // dr1891: 4
 #if __cplusplus >= 201103L
   int n;
diff --git a/clang/test/CXX/drs/dr1xx.cpp b/clang/test/CXX/drs/dr1xx.cpp
index 064b69411f0b5..1930de2f070a7 100644
--- a/clang/test/CXX/drs/dr1xx.cpp
+++ b/clang/test/CXX/drs/dr1xx.cpp
@@ -45,8 +45,8 @@ namespace dr102 { // dr102: yes
   namespace A {
     template T f(T a, T b) { return a + b; }
     // expected-error@-1 {{call to function 'operator+' that is neither visible in the template definition nor found by argument-dependent lookup}}
-    // expected-note@#dr102-instantiation {{in instantiation of function template specialization 'dr102::A::f' requested here}}
-    // expected-note@#dr102-operator-plus {{'operator+' should be declared prior to the call site or in namespace 'dr102::B'}}
+    //   expected-note@#dr102-instantiation {{in instantiation of function template specialization 'dr102::A::f' requested here}}
+    //   expected-note@#dr102-operator-plus {{'operator+' should be declared prior to the call site or in namespace 'dr102::B'}}
   }
   namespace B {
     struct S {};
@@ -114,8 +114,8 @@ namespace dr111 { // dr111: dup 535
   const B b1;
   B b2(b1);
   // expected-error@-1 {{no matching constructor for initialization of 'B'}}
-  // expected-note@#dr111-B {{candidate constructor (the implicit copy constructor) not viable: 1st argument ('const B') would lose const qualifier}}
-  // expected-note@#dr111-B {{candidate constructor not viable: requires 0 arguments, but 1 was provided}}
+  //   expected-note@#dr111-B {{candidate constructor (the implicit copy constructor) not viable: 1st argument ('const B') would lose const qualifier}}
+  //   expected-note@#dr111-B {{candidate constructor not viable: requires 0 arguments, but 1 was provided}}
 }
 
 namespace dr112 { // dr112: yes
@@ -130,11 +130,11 @@ namespace dr112 { // dr112: yes
   // FIXME: Test this somehow in C++11 and on.
   X x1;
   // cxx98-error@-1 {{non-type template argument referring to object 'a1' with internal linkage is a C++11 extension}}
-  // cxx98-note@#dr112-a1 {{non-type template argument refers to object here}}
+  //   cxx98-note@#dr112-a1 {{non-type template argument refers to object here}}
   X x2;
   X x3;
   // cxx98-error@-1 {{non-type template argument referring to object 'a3' with internal linkage is a C++11 extension}}
-  // cxx98-note@#dr112-a3 {{non-type template argument refers to object here}}
+  //   cxx98-note@#dr112-a3 {{non-type template argument refers to object here}}
   X x4;
 }
 
@@ -158,7 +158,7 @@ namespace dr114 { // dr114: yes
     void g() { f(0); }
   } b;
   // expected-error@-1 {{variable type 'struct B' is an abstract class}}
-  // expected-note@#dr114-A-f {{unimplemented pure virtual method 'f' in 'B'}}
+  //   expected-note@#dr114-A-f {{unimplemented pure virtual method 'f' in 'B'}}
 }
 
 namespace dr115 { // dr115: 3.0
@@ -168,31 +168,31 @@ namespace dr115 { // dr115: 3.0
 
   int k1 = f(&f);
   // expected-error@-1 {{no matching function for call to 'f'}}
-  // expected-note@#dr115-f {{candidate template ignored: couldn't infer template argument 'T'}}
+  //   expected-note@#dr115-f {{candidate template ignored: couldn't infer template argument 'T'}}
   int k2 = f(&f);
   int k3 = f(&g);
   // expected-error@-1 {{no matching function for call to 'f'}}
-  // expected-note@#dr115-f {{candidate template ignored: couldn't infer template argument 'T'}}
+  //   expected-note@#dr115-f {{candidate template ignored: couldn't infer template argument 'T'}}
 
   void h() {
     (void)&f;
     // expected-error@-1 {{address of overloaded function 'f' cannot be cast to type 'void'}}
-    // expected-note@#dr115-f {{candidate function template}}
+    //   expected-note@#dr115-f {{candidate function template}}
     (void)&f;
     (void)&g;
     // expected-error@-1 {{address of overloaded function 'g' cannot be cast to type 'void'}}
-    // expected-note@#dr115-g-int {{candidate function template}}
-    // expected-note@#dr115-g {{candidate function template}}
+    //   expected-note@#dr115-g-int {{candidate function template}}
+    //   expected-note@#dr115-g {{candidate function template}}
 
     &f;
     // expected-error@-1 {{reference to overloaded function could not be resolved; did you mean to call it?}}
-    // expected-note@#dr115-f {{possible target for call}}
+    //   expected-note@#dr115-f {{possible target for call}}
     &f;
     // expected-warning@-1 {{expression result unused}}
     &g;
     // expected-error@-1 {{reference to overloaded function could not be resolved; did you mean to call it?}}
-    // expected-note@#dr115-g-int {{possible target for call}}
-    // expected-note@#dr115-g {{possible target for call}}
+    //   expected-note@#dr115-g-int {{possible target for call}}
+    //   expected-note@#dr115-g {{possible target for call}}
   }
 
   struct S {
@@ -254,16 +254,16 @@ namespace dr115 { // dr115: 3.0
   template void with_default(); // #dr115-with-default
   int k10 = f(&with_default);
   // expected-error@-1 {{no matching function for call to 'f'}}
-  // expected-note@#dr115-f {{candidate template ignored: couldn't infer template argument 'T'}}
+  //   expected-note@#dr115-f {{candidate template ignored: couldn't infer template argument 'T'}}
   int k11 = f(&with_default<>);
   void k() {
     (void)&with_default;
     // expected-error@-1 {{address of overloaded function 'with_default' cannot be cast to type 'void'}}
-    // expected-note@#dr115-with-default {{candidate function template}}
+    //   expected-note@#dr115-with-default {{candidate function template}}
     (void)&with_default<>;
     &with_default;
     // expected-error@-1 {{reference to overloaded function could not be resolved; did you mean to call it?}}
-    // expected-note@#dr115-with-default {{possible target for call}}
+    //   expected-note@#dr115-with-default {{possible target for call}}
     &with_default<>;
     // expected-warning@-1 {{expression result unused}}
   }
@@ -275,11 +275,11 @@ namespace dr116 { // dr116: yes
   template void f(A) {} // #dr116-f-N
   template void f(A) {}
   // expected-error@-1 {{redefinition of 'f'}}
-  // expected-note@#dr116-f-N {{previous definition is here}}
+  //   expected-note@#dr116-f-N {{previous definition is here}}
   template void f(A) {} // #dr116-f-T
   template void f(A) {}
   // expected-error@-1 {{redefinition of 'f'}}
-  // expected-note@#dr116-f-T {{previous definition is here}}
+  //   expected-note@#dr116-f-T {{previous definition is here}}
 }
 
 // dr117: na
@@ -383,27 +383,27 @@ namespace dr126 { // dr126: partial
     virtual void dp() throw(D*);
     virtual void ep() throw(E*);
     // cxx98-14-error@-1 {{exception specification of overriding function is more lax than base version}}
-    // cxx98-14-note@#dr126-ep {{overridden virtual function is here}}
+    //   cxx98-14-note@#dr126-ep {{overridden virtual function is here}}
     virtual void fp() throw(F*);
     // cxx98-14-error@-1 {{exception specification of overriding function is more lax than base version}}
-    // cxx98-14-note@#dr126-fp {{overridden virtual function is here}}
+    //   cxx98-14-note@#dr126-fp {{overridden virtual function is here}}
     virtual void gp() throw(G*);
     virtual void hp() throw(H*);
     // cxx98-14-error@-1 {{exception specification of overriding function is more lax than base version}}
-    // cxx98-14-note@#dr126-hp {{overridden virtual function is here}}
+    //   cxx98-14-note@#dr126-hp {{overridden virtual function is here}}
 
     virtual void cr() throw(C&);
     virtual void dr() throw(D&);
     virtual void er() throw(E&);
     // cxx98-14-error@-1 {{exception specification of overriding function is more lax than base version}}
-    // cxx98-14-note@#dr126-er {{overridden virtual function is here}}
+    //   cxx98-14-note@#dr126-er {{overridden virtual function is here}}
     virtual void fr() throw(F&);
     // cxx98-14-error@-1 {{exception specification of overriding function is more lax than base version}}
-    // cxx98-14-note@#dr126-fr {{overridden virtual function is here}}
+    //   cxx98-14-note@#dr126-fr {{overridden virtual function is here}}
     virtual void gr() throw(G&);
     virtual void hr() throw(H&);
     // cxx98-14-error@-1 {{exception specification of overriding function is more lax than base version}}
-    // cxx98-14-note@#dr126-hr {{overridden virtual function is here}}
+    //   cxx98-14-note@#dr126-hr {{overridden virtual function is here}}
 
     virtual void pv() throw(C*);
 
@@ -413,7 +413,7 @@ namespace dr126 { // dr126: partial
     virtual void npm() throw(nullptr_t&);
     virtual void nr() throw(nullptr_t);
     // cxx11-14-error@-1 {{exception specification of overriding function is more lax than base version}}
-    // cxx11-14-note@#dr126-nr {{overridden virtual function is here}}
+    //   cxx11-14-note@#dr126-nr {{overridden virtual function is here}}
     virtual void ncr() throw(nullptr_t);
 #endif // __cplusplus >= 201103L
 
@@ -424,16 +424,16 @@ namespace dr126 { // dr126: partial
     virtual void w() throw(int);
     virtual void x() throw(const int*);
     // cxx98-14-error@-1 {{exception specification of overriding function is more lax than base version}}
-    // cxx98-14-note@#dr126-x {{overridden virtual function is here}}
+    //   cxx98-14-note@#dr126-x {{overridden virtual function is here}}
     virtual void y() throw(int*); // ok
     virtual void z() throw(long);
     // cxx98-14-error@-1 {{exception specification of overriding function is more lax than base version}}
-    // cxx98-14-note@#dr126-z {{overridden virtual function is here}}
+    //   cxx98-14-note@#dr126-z {{overridden virtual function is here}}
   };
 #endif // __cplusplus <= 201402L
   void f() throw(int);
   // since-cxx17-error@-1 {{ISO C++17 does not allow dynamic exception specifications}}
-  // since-cxx17-note@-2 {{use 'noexcept(false)' instead}}
+  //   since-cxx17-note@-2 {{use 'noexcept(false)' instead}}
 }
 
 namespace dr127 { // dr127: 2.9
@@ -491,10 +491,10 @@ namespace dr136 { // dr136: 3.4
   struct A {
     friend void f(int, int = 0, int);
     // expected-error@-1 {{friend declaration specifying a default argument must be the only declaration}}
-    // expected-note@#dr136-f {{previous declaration is here}}
+    //   expected-note@#dr136-f {{previous declaration is here}}
     friend void g(int, int, int = 0);
     // expected-error@-1 {{friend declaration specifying a default argument must be the only declaration}}
-    // expected-note@#dr136-g {{previous declaration is here}}
+    //   expected-note@#dr136-g {{previous declaration is here}}
     friend void h(int, int, int = 0);
     // expected-error@-1 {{friend declaration specifying a default argument must be a definition}}
     friend void i(int, int, int = 0) {} // #dr136-A-i
@@ -503,7 +503,7 @@ namespace dr136 { // dr136: 3.4
   };
   void i(int, int, int);
   // expected-error@-1 {{friend declaration specifying a default argument must be the only declaration}}
-  // expected-note@#dr136-A-i {{previous declaration is here}}
+  //   expected-note@#dr136-A-i {{previous declaration is here}}
   void q() {
     j(A(), A()); // ok, has default argument
   }
@@ -512,7 +512,7 @@ namespace dr136 { // dr136: 3.4
   struct A {
     friend void dr136::k(int, int, int, int = 0);
     // expected-error@-1 {{friend declaration specifying a default argument must be the only declaration}}
-    // expected-note@#dr136-k {{previous declaration is here}}
+    //   expected-note@#dr136-k {{previous declaration is here}}
   };
   }
   namespace NSB {
@@ -529,7 +529,7 @@ namespace dr136 { // dr136: 3.4
   struct C {
     friend void B::f(int = 0);
     // expected-error@-1 {{friend declaration specifying a default argument must be the only declaration}}
-    // expected-note@#dr136-B-f {{previous declaration is here}}
+    //   expected-note@#dr136-B-f {{previous declaration is here}}
   };
 }
 
@@ -563,7 +563,7 @@ namespace dr139 { // dr139: yes
     struct A {
       friend void f(A &);
       // expected-error@-1 {{redefinition of 'f' as different kind of symbol}}
-      // expected-note@#dr139-typedef-f {{previous definition is here}}
+      //   expected-note@#dr139-typedef-f {{previous definition is here}}
     };
   }
 
@@ -583,7 +583,7 @@ namespace dr140 { // dr140: yes
   void f(int *const) {} // #dr140-f-first
   void f(int[3]) {}
   // expected-error@-1 {{redefinition of 'f'}}
-  // expected-note@#dr140-f-first {{previous definition is here}}
+  //   expected-note@#dr140-f-first {{previous definition is here}}
   void g(const int);
   void g(int n) { n = 2; }
 }
@@ -644,23 +644,23 @@ namespace dr142 { // dr142: 2.8
   void DD::f() {
     mi = 3;
     // expected-error@-1 {{'mi' is a private member of 'dr142::B'}}
-    // expected-note@#dr142-D {{constrained by private inheritance here}}
-    // expected-note@#dr142-B-mi {{member is declared here}}
+    //   expected-note@#dr142-D {{constrained by private inheritance here}}
+    //   expected-note@#dr142-B-mi {{member is declared here}}
     si = 3;
     // expected-error@-1 {{'si' is a private member of 'dr142::B'}}
-    // expected-note@#dr142-D {{constrained by private inheritance here}}
-    // expected-note@#dr142-B-si {{member is declared here}}
+    //   expected-note@#dr142-D {{constrained by private inheritance here}}
+    //   expected-note@#dr142-B-si {{member is declared here}}
     B b_old;
     // expected-error@-1 {{'B' is a private member of 'dr142::B'}}
-    // expected-note@#dr142-D {{constrained by private inheritance here}}
-    // expected-note@#dr142-B {{member is declared here}}
+    //   expected-note@#dr142-D {{constrained by private inheritance here}}
+    //   expected-note@#dr142-B {{member is declared here}}
     dr142::B b;
     b.mi = 3;
     b.si = 3;
     B::si = 3;
     // expected-error@-1 {{'B' is a private member of 'dr142::B'}}
-    // expected-note@#dr142-D {{constrained by private inheritance here}}
-    // expected-note@#dr142-B {{member is declared here}}
+    //   expected-note@#dr142-D {{constrained by private inheritance here}}
+    //   expected-note@#dr142-B {{member is declared here}}
     dr142::B::si = 3;
     B *bp1_old = this; // #dr142-bp1_old
     // expected-error@#dr142-bp1_old {{'B' is a private member of 'dr142::B'}}
@@ -670,7 +670,7 @@ namespace dr142 { // dr142: 2.8
     //   expected-note@#dr142-D {{declared private here}}
     dr142::B *bp1 = this;
     // expected-error@-1 {{cannot cast 'dr142::DD' to its private base class 'dr142::B'}}
-    // expected-note@#dr142-D {{declared private here}}
+    //   expected-note@#dr142-D {{declared private here}}
     B *bp2_old = (B*)this; // #dr142-bp2_old
     // expected-error@#dr142-bp2_old {{'B' is a private member of 'dr142::B'}}
     //   expected-note@#dr142-D {{constrained by private inheritance here}}
@@ -761,15 +761,15 @@ namespace dr152 { // dr152: yes
   };
   A a1 = A();
   // cxx98-14-error@-1 {{no matching constructor for initialization of 'A'}}
-  // cxx98-14-note@#dr152-A-explicit-ctor {{explicit constructor is not a candidate}}
-  // cxx98-14-note@#dr152-A-ctor {{candidate constructor not viable: requires 0 arguments, but 1 was provided}}
+  //   cxx98-14-note@#dr152-A-explicit-ctor {{explicit constructor is not a candidate}}
+  //   cxx98-14-note@#dr152-A-ctor {{candidate constructor not viable: requires 0 arguments, but 1 was provided}}
   A a2((A()));
 
   A &f();
   A a3 = f();
   // expected-error@-1 {{no matching constructor for initialization of 'A'}}
-  // expected-note@#dr152-A-explicit-ctor {{explicit constructor is not a candidate}}
-  // expected-note@#dr152-A-ctor {{candidate constructor not viable: requires 0 arguments, but 1 was provided}}
+  //   expected-note@#dr152-A-explicit-ctor {{explicit constructor is not a candidate}}
+  //   expected-note@#dr152-A-ctor {{candidate constructor not viable: requires 0 arguments, but 1 was provided}}
   A a4(f());
 }
 
@@ -817,17 +817,17 @@ namespace dr161 { // dr161: 3.1
       C::B b2; // ok, accessible as a member of A
       (void)&C::b;
       // expected-error@-1 {{'b' is a protected member of 'dr161::A'}}
-      // expected-note@#dr161-B {{declared protected here}}
+      //   expected-note@#dr161-B {{declared protected here}}
       (void)&C::bs;
       (void)c.b;
       // expected-error@-1 {{'b' is a protected member of 'dr161::A'}}
-      // expected-note@#dr161-B {{declared protected here}}
+      //   expected-note@#dr161-B {{declared protected here}}
       (void)c.bs;
       f();
       sf();
       c.f();
       // expected-error@-1 {{protected}}
-      // expected-note@#dr161-f {{declared protected here}}
+      //   expected-note@#dr161-f {{declared protected here}}
       c.sf();
       A::f();
       D::f();
@@ -887,8 +887,8 @@ namespace dr166 { // dr166: 2.9
   int g(A::X);
   template int h(T t) { return t.n; }
   // expected-error@-1 {{'n' is a private member of 'dr166::A::X'}}
-  // expected-note@#dr166-h-instantiation {{in instantiation of function template specialization 'dr166::h' requested here}}
-  // expected-note@#dr166-X-n {{implicitly declared private here}}
+  //   expected-note@#dr166-h-instantiation {{in instantiation of function template specialization 'dr166::h' requested here}}
+  //   expected-note@#dr166-X-n {{implicitly declared private here}}
   int i(A::X);
 
   namespace A {
@@ -909,7 +909,7 @@ namespace dr166 { // dr166: 2.9
   template int h(A::X); // #dr166-h-instantiation
   int i(A::X x) { return x.n; }
   // expected-error@-1 {{'n' is a private member of 'dr166::A::X'}}
-  // expected-note@#dr166-X-n {{implicitly declared private here}}
+  //   expected-note@#dr166-X-n {{implicitly declared private here}}
 }
 
 // dr167: sup 1012
@@ -951,7 +951,7 @@ namespace dr171 {
   extern "C" void dr171a();
   extern "C" void dr171b();
   // expected-error@-1 {{declaration of 'dr171b' with C language linkage conflicts with declaration in global scope}}
-  // expected-note@#dr171b-int {{declared in global scope here}}
+  //   expected-note@#dr171b-int {{declared in global scope here}}
 }
 
 namespace dr172 { // dr172: yes
@@ -1002,8 +1002,8 @@ namespace dr175 { // dr175: 2.8
   struct C : B {
     A a;
     // expected-error@-1 {{'A' is a private member of 'dr175::A'}}
-    // expected-note@#dr175-B {{constrained by private inheritance here}}
-    // expected-note@#dr175-A {{member is declared here}}
+    //   expected-note@#dr175-B {{constrained by private inheritance here}}
+    //   expected-note@#dr175-A {{member is declared here}}
     dr175::A b;
   };
 }
@@ -1015,7 +1015,7 @@ namespace dr176 { // dr176: 3.1
       typedef Y A; // #dr176-A-first
       typedef Y A;
       // expected-error@-1 {{typedef redefinition with different types ('Y' vs 'Y')}}
-      // expected-note@#dr176-A-first {{previous definition is here}}
+      //   expected-note@#dr176-A-first {{previous definition is here}}
     }
   };
 
@@ -1031,8 +1031,8 @@ namespace dr176 { // dr176: 3.1
   template struct Derived2 : Base, Base {
     typename Derived2::Base b;
     // expected-error@-1 {{member 'Base' found in multiple base classes of different types}}
-    // expected-note@#dr176-Base {{member type 'dr176::Base' found by ambiguous name lookup}}
-    // expected-note@#dr176-Base {{member type 'dr176::Base' found by ambiguous name lookup}}
+    //   expected-note@#dr176-Base {{member type 'dr176::Base' found by ambiguous name lookup}}
+    //   expected-note@#dr176-Base {{member type 'dr176::Base' found by ambiguous name lookup}}
     typename Derived2::Base d;
   };
 
@@ -1057,8 +1057,8 @@ namespace dr177 { // dr177: yes
   B b;
   A a = b;
   // cxx98-14-error@-1 {{no viable constructor copying variable of type 'A'}}
-  // cxx98-14-note@#dr177-A-copy-ctor {{candidate constructor not viable: expects an lvalue for 1st argument}}
-  // cxx98-14-note@#dr177-A-ctor-from-B {{candidate constructor not viable: no known conversion from 'A' to 'const B &' for 1st argument}}
+  //   cxx98-14-note@#dr177-A-copy-ctor {{candidate constructor not viable: expects an lvalue for 1st argument}}
+  //   cxx98-14-note@#dr177-A-ctor-from-B {{candidate constructor not viable: no known conversion from 'A' to 'const B &' for 1st argument}}
 
   struct C { C(C&); }; // #dr177-C-copy-ctor
   struct D : C {};
@@ -1066,7 +1066,7 @@ namespace dr177 { // dr177: yes
   E e;
   C c = e;
   // expected-error@-1 {{no viable constructor copying variable of type 'D'}}
-  // expected-note@#dr177-C-copy-ctor {{candidate constructor not viable: expects an lvalue for 1st argument}}
+  //   expected-note@#dr177-C-copy-ctor {{candidate constructor not viable: expects an lvalue for 1st argument}}
 }
 
 namespace dr178 { // dr178: yes
@@ -1157,7 +1157,7 @@ namespace dr184 { // dr184: yes
   template class T> void A::f() { // #dr184-T
     T<> t;
     // expected-error@-1 {{too few template arguments for template template parameter 'T'}}
-    // expected-note@#dr184-T {{template is declared here}}
+    //   expected-note@#dr184-T {{template is declared here}}
   }
 
   template class T> void A::g() {
@@ -1248,7 +1248,7 @@ namespace dr197 { // dr197: yes
     char &a = f(1);
     char &b = f(T(1));
     // expected-error@-1 {{non-const lvalue reference to type 'char' cannot bind to a value of unrelated type 'int'}}
-    // expected-note@#dr197-g-e-call {{in instantiation of function template specialization 'dr197::g' requested here}}
+    //   expected-note@#dr197-g-e-call {{in instantiation of function template specialization 'dr197::g' requested here}}
     char &c = f(t);
     // expected-error@-1 {{non-const lvalue reference to type 'char' cannot bind to a value of unrelated type 'int'}}
   }
diff --git a/clang/test/CXX/drs/dr20xx.cpp b/clang/test/CXX/drs/dr20xx.cpp
index 60ee7684440f5..f7f37379e61ad 100644
--- a/clang/test/CXX/drs/dr20xx.cpp
+++ b/clang/test/CXX/drs/dr20xx.cpp
@@ -418,3 +418,5 @@ namespace dr2094 { // dr2094: 5
   static_assert(__is_trivially_assignable(A, const A&), "");
   static_assert(__is_trivially_assignable(B, const B&), "");
 }
+
+// dr2096: dup 2598
diff --git a/clang/test/CXX/drs/dr2335.cpp b/clang/test/CXX/drs/dr2335.cpp
new file mode 100644
index 0000000000000..d143aaf7cb0ac
--- /dev/null
+++ b/clang/test/CXX/drs/dr2335.cpp
@@ -0,0 +1,48 @@
+// RUN: %clang_cc1 -std=c++98 %s -verify=expected -fexceptions -fcxx-exceptions -pedantic-errors
+// RUN: %clang_cc1 -std=c++11 %s -verify=expected -fexceptions -fcxx-exceptions -pedantic-errors
+// RUN: %clang_cc1 -std=c++14 %s -verify=expected,since-cxx14 -fexceptions -fcxx-exceptions -pedantic-errors
+// RUN: %clang_cc1 -std=c++17 %s -verify=expected,since-cxx14 -fexceptions -fcxx-exceptions -pedantic-errors
+// RUN: %clang_cc1 -std=c++20 %s -verify=expected,since-cxx14 -fexceptions -fcxx-exceptions -pedantic-errors
+// RUN: %clang_cc1 -std=c++23 %s -verify=expected,since-cxx14 -fexceptions -fcxx-exceptions -pedantic-errors
+// RUN: %clang_cc1 -std=c++2c %s -verify=expected,since-cxx14 -fexceptions -fcxx-exceptions -pedantic-errors
+
+#if __cplusplus <= 201103L
+// expected-no-diagnostics
+#endif
+
+namespace dr2335 { // dr2335: no drafting
+// FIXME: current consensus is that the examples are well-formed.
+#if __cplusplus >= 201402L
+namespace ex1 {
+template  struct partition_indices {
+  static auto compute_right() {}
+  static constexpr auto right = compute_right;
+};
+template struct partition_indices;
+} // namespace ex1
+
+namespace ex2 {
+template  struct X {};
+template  struct partition_indices {
+  static auto compute_right() { return X(); }
+  // since-cxx14-error@-1 {{no member 'I' in 'dr2335::ex2::partition_indices'; it has not yet been instantiated}}
+  //   since-cxx14-note@#dr2335-ex2-right {{in instantiation of member function 'dr2335::ex2::partition_indices::compute_right' requested here}}
+  //   since-cxx14-note@#dr2335-ex2-inst {{in instantiation of template class 'dr2335::ex2::partition_indices' requested here}}
+  //   since-cxx14-note@#dr2335-ex2-I {{not-yet-instantiated member is declared here}}
+  static constexpr auto right = compute_right; // #dr2335-ex2-right
+  static constexpr int I = sizeof(T); // #dr2335-ex2-I
+};
+template struct partition_indices; // #dr2335-ex2-inst
+} // namespace ex2
+
+namespace ex3 {
+struct partition_indices {
+  static auto compute_right() {} // #dr2335-compute_right
+  static constexpr auto right = compute_right; // #dr2335-ex3-right
+  // since-cxx14-error@-1 {{function 'compute_right' with deduced return type cannot be used before it is defined}}
+  //   since-cxx14-note@#dr2335-compute_right {{'compute_right' declared here}}
+  // since-cxx14-error@#dr2335-ex3-right {{declaration of variable 'right' with deduced type 'const auto' requires an initializer}}
+};
+} // namespace ex3
+#endif
+} // namespace dr2335
diff --git a/clang/test/CXX/drs/dr23xx.cpp b/clang/test/CXX/drs/dr23xx.cpp
index d2f4e7652ab56..03077ae9239a4 100644
--- a/clang/test/CXX/drs/dr23xx.cpp
+++ b/clang/test/CXX/drs/dr23xx.cpp
@@ -48,6 +48,7 @@ void g() {
 #endif
 
 // dr2331: na
+// dr2335 is in dr2335.cxx
 
 #if __cplusplus >= 201103L
 namespace dr2338 { // dr2338: 12
diff --git a/clang/test/CXX/drs/dr25xx.cpp b/clang/test/CXX/drs/dr25xx.cpp
index 32bbfc63d0df4..502f03271d9af 100644
--- a/clang/test/CXX/drs/dr25xx.cpp
+++ b/clang/test/CXX/drs/dr25xx.cpp
@@ -208,3 +208,67 @@ namespace dr2565 { // dr2565: 16 open
 
 #endif
 }
+
+
+namespace dr2598 { // dr2598: 18
+#if __cplusplus >= 201103L
+struct NonLiteral {
+    NonLiteral();
+};
+
+struct anonymous1 {
+    union {} a;
+};
+static_assert(__is_literal(anonymous1), "");
+
+struct anonymous2 {
+    union { char c; };
+};
+static_assert(__is_literal(anonymous2), "");
+
+struct anonymous3 {
+    union { char c; NonLiteral NL; };
+};
+static_assert(__is_literal(anonymous3), "");
+
+struct anonymous4 {
+    union { NonLiteral NL; };
+};
+static_assert(!__is_literal(anonymous4), "");
+
+union empty {};
+static_assert(__is_literal(empty), "");
+
+union union1 { char c; };
+static_assert(__is_literal(union1), "");
+
+union union2 { char c; NonLiteral NL;};
+static_assert(__is_literal(union2), "");
+
+union union3 { NonLiteral NL;};
+static_assert(!__is_literal(union3), "");
+
+union union4 { union4(); };
+static_assert(!__is_literal(union4), "");
+
+union union5 { static NonLiteral NL; };
+static_assert(__is_literal(union5), "");
+
+struct Literal { constexpr Literal() {} };
+union union6 { NonLiteral NL; Literal L; };
+static_assert(__is_literal(union6), "");
+
+#if __cplusplus >= 202003L
+struct A { A(); };
+union U {
+  A a;
+  constexpr U() {}
+  constexpr ~U() {}
+};
+static_assert(!__is_literal(U), "");
+#endif
+
+
+
+#endif
+}
diff --git a/clang/test/CXX/drs/dr26xx.cpp b/clang/test/CXX/drs/dr26xx.cpp
index 47eeddcc5bf26..f151c9eea051a 100644
--- a/clang/test/CXX/drs/dr26xx.cpp
+++ b/clang/test/CXX/drs/dr26xx.cpp
@@ -106,11 +106,11 @@ int \N{Λ} = 0;
 // expected-error@-2 {{expected unqualified-id}}
 const char* emoji = "\N{🤡}";
 // expected-error@-1 {{'🤡' is not a valid Unicode character name}}
-// expected-note@-2 {{did you mean OX ('🐂' U+1F402)?}}
-// expected-note@-3 {{did you mean ANT ('🐜' U+1F41C)?}}
-// expected-note@-4 {{did you mean ARC ('⌒' U+2312)?}}
-// expected-note@-5 {{did you mean AXE ('🪓' U+1FA93)?}}
-// expected-note@-6 {{did you mean BAT ('🦇' U+1F987)?}}
+//   expected-note@-2 {{did you mean OX ('🐂' U+1F402)?}}
+//   expected-note@-3 {{did you mean ANT ('🐜' U+1F41C)?}}
+//   expected-note@-4 {{did you mean ARC ('⌒' U+2312)?}}
+//   expected-note@-5 {{did you mean AXE ('🪓' U+1FA93)?}}
+//   expected-note@-6 {{did you mean BAT ('🦇' U+1F987)?}}
 
 #define z(x) 0
 #define dr2640_a z(
diff --git a/clang/test/CXX/drs/dr2xx.cpp b/clang/test/CXX/drs/dr2xx.cpp
index bc862e26ff667..1a3ac532f93b2 100644
--- a/clang/test/CXX/drs/dr2xx.cpp
+++ b/clang/test/CXX/drs/dr2xx.cpp
@@ -42,10 +42,10 @@ namespace dr206 { // dr206: yes
   struct S; // #dr206-S
   template struct Q { S s; };
   // expected-error@-1 {{field has incomplete type 'S'}}
-  // expected-note@#dr206-S {{forward declaration of 'dr206::S'}}
+  //   expected-note@#dr206-S {{forward declaration of 'dr206::S'}}
   template void f() { S s; }
   // expected-error@-1 {{variable has incomplete type 'S'}}
-  // expected-note@#dr206-S {{forward declaration of 'dr206::S'}}
+  //   expected-note@#dr206-S {{forward declaration of 'dr206::S'}}
 }
 
 namespace dr207 { // dr207: yes
@@ -72,7 +72,7 @@ namespace dr209 { // dr209: 3.2
   class B {
     friend void A::f();
     // expected-error@-1 {{friend function 'f' is a private member of 'dr209::A'}}
-    // expected-note@#dr209-A-f {{implicitly declared private here}}
+    //   expected-note@#dr209-A-f {{implicitly declared private here}}
   };
 }
 
@@ -95,8 +95,8 @@ namespace dr213 { // dr213: yes
       char &r1 = f(t);
       int &r2 = g(t);
       // expected-error@-1 {{explicit qualification required to use member 'g' from dependent base class}}
-      // expected-note@#dr213-instantiation {{in instantiation of member function 'dr213::A::h' requested here}}
-      // expected-note@#dr213-B-g {{member is declared here}}
+      //   expected-note@#dr213-instantiation {{in instantiation of member function 'dr213::A::h' requested here}}
+      //   expected-note@#dr213-B-g {{member is declared here}}
     }
   };
   struct B {
@@ -172,16 +172,16 @@ namespace dr218 { // dr218: yes
     // expected-error@-1 {{called object type 'int' is not a function or function pointer}}
     void test2(A::S as) { void f(); f(as); }
     // expected-error@-1 {{too many arguments to function call, expected 0, have 1}}
-    // expected-note@-2 {{'f' declared here}}
+    //   expected-note@-2 {{'f' declared here}}
     void test3(A::S as) { using A::f; f(as); } // ok
     void test4(A::S as) { using B::f; f(as); } // ok
     void test5(A::S as) { int f; f(as); }
     // expected-error@-1 {{called object type 'int' is not a function or function pointer}}
     void test6(A::S as) { struct f {}; (void) f(as); }
     // expected-error@-1 {{no matching conversion for functional-style cast from 'A::S' to 'f'}}
-    // expected-note@-2 {{candidate constructor (the implicit copy constructor) not viable: no known conversion from 'A::S' to 'const f' for 1st argument}}
-    // since-cxx11-note@-3 {{candidate constructor (the implicit move constructor) not viable: no known conversion from 'A::S' to 'f' for 1st argument}}
-    // expected-note@-4 {{candidate constructor (the implicit default constructor) not viable: requires 0 arguments, but 1 was provided}}
+    //   expected-note@-2 {{candidate constructor (the implicit copy constructor) not viable: no known conversion from 'A::S' to 'const f' for 1st argument}}
+    //   since-cxx11-note@-3 {{candidate constructor (the implicit move constructor) not viable: no known conversion from 'A::S' to 'f' for 1st argument}}
+    //   expected-note@-4 {{candidate constructor (the implicit default constructor) not viable: requires 0 arguments, but 1 was provided}}
   };
 
   namespace D {
@@ -243,15 +243,15 @@ namespace dr221 { // dr221: 3.6
     a += n;
     a = c;
     // expected-error@-1 {{no viable overloaded '='}}
-    // expected-note@#dr221-S-copy-assign {{candidate function not viable: no known conversion from 'char' to 'int &' for 1st argument}}
-    // since-cxx11-note@#dr221-S {{candidate function (the implicit move assignment operator) not viable: no known conversion from 'char' to 'A' for 1st argument}}
-    // expected-note@#dr221-S {{candidate function (the implicit copy assignment operator) not viable: no known conversion from 'char' to 'const A' for 1st argument}}
+    //   expected-note@#dr221-S-copy-assign {{candidate function not viable: no known conversion from 'char' to 'int &' for 1st argument}}
+    //   since-cxx11-note@#dr221-S {{candidate function (the implicit move assignment operator) not viable: no known conversion from 'char' to 'A' for 1st argument}}
+    //   expected-note@#dr221-S {{candidate function (the implicit copy assignment operator) not viable: no known conversion from 'char' to 'const A' for 1st argument}}
     a += c;
     a = f;
     // expected-error@-1 {{no viable overloaded '='}}
-    // expected-note@#dr221-S-copy-assign {{candidate function not viable: no known conversion from 'float' to 'int &' for 1st argument}}
-    // since-cxx11-note@#dr221-S {{candidate function (the implicit move assignment operator) not viable: no known conversion from 'float' to 'A' for 1st argument}}
-    // expected-note@#dr221-S {{candidate function (the implicit copy assignment operator) not viable: no known conversion from 'float' to 'const A' for 1st argument}}
+    //   expected-note@#dr221-S-copy-assign {{candidate function not viable: no known conversion from 'float' to 'int &' for 1st argument}}
+    //   since-cxx11-note@#dr221-S {{candidate function (the implicit move assignment operator) not viable: no known conversion from 'float' to 'A' for 1st argument}}
+    //   expected-note@#dr221-S {{candidate function (the implicit copy assignment operator) not viable: no known conversion from 'float' to 'const A' for 1st argument}}
     a += f;
   }
 }
@@ -351,8 +351,8 @@ namespace dr224 { // dr224: 16
 // dr225: yes
 template void dr225_f(T t) { dr225_g(t); }
 // expected-error@-1 {{call to function 'dr225_g' that is neither visible in the template definition nor found by argument-dependent lookup}}
-// expected-note@#dr225-f {{in instantiation of function template specialization 'dr225_f' requested here}}
-// expected-note@#dr225-g {{'dr225_g' should be declared prior to the call site}}
+//   expected-note@#dr225-f {{in instantiation of function template specialization 'dr225_f' requested here}}
+//   expected-note@#dr225-g {{'dr225_g' should be declared prior to the call site}}
 void dr225_g(int); // #dr225-g
 template void dr225_f(int); // #dr225-f
 
@@ -397,7 +397,7 @@ namespace dr226 { // dr226: no
   // cxx98-error@-1 {{default template arguments for a function template are a C++11 extension}}
   template struct Foo {};
   // expected-error@-1 {{template parameter missing a default argument}}
-  // expected-note@-2 {{previous default template argument defined here}}
+  //   expected-note@-2 {{previous default template argument defined here}}
 
   template int foo(X, Y);
   // cxx98-error@-1 {{default template arguments for a function template are a C++11 extension}}
@@ -433,7 +433,7 @@ namespace dr230 { // dr230: 3.0
   struct S {
     S() { f(); }
     // expected-warning@-1 {{call to pure virtual member function 'f' has undefined behavior; overrides of 'f' in subclasses are not available in the constructor of 'dr230::S'}}
-    // expected-note@#dr230-f {{'f' declared here}}
+    //   expected-note@#dr230-f {{'f' declared here}}
     virtual void f() = 0; // #dr230-f
   };
 }
@@ -446,7 +446,7 @@ namespace dr231 { // dr231: yes
     void f() { using namespace inner; }
     int j = i;
     // expected-error@-1 {{use of undeclared identifier 'i'; did you mean 'inner::i'?}}
-    // expected-note@#dr231-i {{'inner::i' declared here}}
+    //   expected-note@#dr231-i {{'inner::i' declared here}}
   }
 }
 
@@ -502,20 +502,20 @@ namespace dr241 { // dr241: yes
     // cxx98-17-error@-1 {{use of function template name with no prior declaration in function call with explicit template arguments is a C++20 extension}}
     A::f<3>(b);
     // expected-error@-1 {{no matching function for call to 'f'}}
-    // expected-note@#dr241-A-f {{candidate function template not viable: requires 0 arguments, but 1 was provided}}
+    //   expected-note@#dr241-A-f {{candidate function template not viable: requires 0 arguments, but 1 was provided}}
     A::g<3>(b);
     C::f<3>(b);
     // expected-error@-1 {{no matching function for call to 'f'}}
-    // expected-note@#dr241-C-f {{candidate template ignored: invalid explicitly-specified argument for template parameter 'T'}}
+    //   expected-note@#dr241-C-f {{candidate template ignored: invalid explicitly-specified argument for template parameter 'T'}}
     C::g<3>(b);
     // expected-error@-1 {{no matching function for call to 'g'}}
-    // expected-note@#dr241-C-g {{candidate template ignored: invalid explicitly-specified argument for template parameter 'T'}}
+    //   expected-note@#dr241-C-g {{candidate template ignored: invalid explicitly-specified argument for template parameter 'T'}}
     using C::f;
     using C::g;
     f<3>(b);
     // expected-error@-1 {{no matching function for call to 'f'}}
-    // expected-note@#dr241-C-f {{candidate template ignored: invalid explicitly-specified argument for template parameter 'T'}}
-    // expected-note@#dr241-A-f {{candidate function template not viable: requires 0 arguments, but 1 was provided}}
+    //   expected-note@#dr241-C-f {{candidate template ignored: invalid explicitly-specified argument for template parameter 'T'}}
+    //   expected-note@#dr241-A-f {{candidate function template not viable: requires 0 arguments, but 1 was provided}}
     g<3>(b);
   }
 }
@@ -532,8 +532,8 @@ namespace dr243 { // dr243: yes
   A a1(b);
   A a2 = b;
   // expected-error@-1 {{conversion from 'struct B' to 'A' is ambiguous}}
-  // expected-note@#dr243-A {{candidate constructor}}
-  // expected-note@#dr243-B {{candidate function has been explicitly deleted}}
+  //   expected-note@#dr243-A {{candidate constructor}}
+  //   expected-note@#dr243-B {{candidate function has been explicitly deleted}}
 }
 
 namespace dr244 { // dr244: 11
@@ -548,7 +548,7 @@ namespace dr244 { // dr244: 11
   void f() {
     D_object.~B();
     // expected-error@-1 {{destructor type 'dr244::B' in object destruction expression does not match the type 'D' of the object being destroyed}}
-    // expected-note@#dr244-B {{type 'dr244::B' found by destructor name lookup}}
+    //   expected-note@#dr244-B {{type 'dr244::B' found by destructor name lookup}}
     D_object.B::~B();
     D_object.D::~B(); // FIXME: Missing diagnostic for this.
     B_ptr->~B();
@@ -601,7 +601,7 @@ namespace dr244 { // dr244: 11
     // Rejecting this seems correct, but most compilers accept, so we do also.
     f.N::F::~G();
     // expected-error@-1 {{qualified destructor name only found in lexical scope; omit the qualifier to find this type name by unqualified lookup}}
-    // expected-note@#dr244-G {{type 'G' (aka 'E') found by destructor name lookup}}
+    //   expected-note@#dr244-G {{type 'G' (aka 'E') found by destructor name lookup}}
   }
 
   // Bizarrely, compilers perform lookup in the scope for qualified destructor
@@ -615,7 +615,7 @@ namespace dr244 { // dr244: 11
       typedef typename N::S::Inner T;
       p->::dr244::QualifiedLookupInScope::N::S::Inner::~T();
       // expected-error@-1 {{no type named 'T' in 'dr244::QualifiedLookupInScope::N::S'}}
-      // expected-note@#dr244-f {{in instantiation of function template specialization 'dr244::QualifiedLookupInScope::f' requested here}}
+      //   expected-note@#dr244-f {{in instantiation of function template specialization 'dr244::QualifiedLookupInScope::f' requested here}}
     }
     template void f(N::S::Inner *); // #dr244-f
 
@@ -635,7 +635,7 @@ namespace dr245 { // dr245: yes
     enum E {}; // #dr245-E
     class E *p;
     // expected-error@-1 {{use of 'E' with tag type that does not match previous declaration}}
-    // expected-note@#dr245-E {{previous use is here}}
+    //   expected-note@#dr245-E {{previous use is here}}
   };
 }
 
@@ -647,7 +647,7 @@ X: ;
     } catch (int) {
       goto X;
       // expected-error@-1 {{cannot jump from this goto statement to its label}}
-      // expected-note@#dr246-try {{jump bypasses initialization of try block}}
+      //   expected-note@#dr246-try {{jump bypasses initialization of try block}}
     }
   };
 }
@@ -710,8 +710,8 @@ namespace dr252 { // dr252: 3.1
   };
   C::~C() {}
   // expected-error@-1 {{member 'operator delete' found in multiple base classes of different types}}
-  // expected-note@#dr252-A {{member found by ambiguous name lookup}}
-  // expected-note@#dr252-B {{member found by ambiguous name lookup}}
+  //   expected-note@#dr252-A {{member found by ambiguous name lookup}}
+  //   expected-note@#dr252-B {{member found by ambiguous name lookup}}
 
   struct D {
     void operator delete(void*, int); // #dr252-D
@@ -719,7 +719,7 @@ namespace dr252 { // dr252: 3.1
   };
   D::~D() {}
   // expected-error@-1 {{no suitable member 'operator delete' in 'D'}}
-  // expected-note@#dr252-D {{member 'operator delete' declared here}}
+  //   expected-note@#dr252-D {{member 'operator delete' declared here}}
 
   struct E {
     void operator delete(void*, int);
@@ -729,7 +729,7 @@ namespace dr252 { // dr252: 3.1
   };
   E::~E() {}
   // expected-error@-1 {{attempt to use a deleted function}}
-  // expected-note@#dr252-E {{'operator delete' has been explicitly marked deleted here}}
+  //   expected-note@#dr252-E {{'operator delete' has been explicitly marked deleted here}}
 
   struct F {
     // If both functions are available, the first one is a placement delete.
@@ -740,7 +740,7 @@ namespace dr252 { // dr252: 3.1
   };
   F::~F() {}
   // expected-error@-1 {{attempt to use a deleted function}}
-  // expected-note@#dr252-F {{'operator delete' has been explicitly marked deleted here}}
+  //   expected-note@#dr252-F {{'operator delete' has been explicitly marked deleted here}}
 
   struct G {
     void operator delete(void*, size_t);
@@ -755,8 +755,8 @@ namespace dr254 { // dr254: 2.9
                                    // it's not an elaborated-type-specifier
     typedef struct T::type foo;
     // expected-error@-1 {{typedef 'type' cannot be referenced with a struct specifier}}
-    // expected-note@#dr254-instantiation {{in instantiation of template class 'dr254::A' requested here}}
-    // expected-note@#dr254-C {{declared here}}
+    //   expected-note@#dr254-instantiation {{in instantiation of template class 'dr254::A' requested here}}
+    //   expected-note@#dr254-C {{declared here}}
   };
   struct B { struct type {}; };
   struct C { typedef struct {} type; }; // #dr254-C
@@ -786,7 +786,7 @@ namespace dr257 { // dr257: 3.4
   struct D : B {
     D() {}
     // expected-error@-1 {{constructor for 'dr257::D' must explicitly initialize the base class 'A' which does not have a default constructor}}
-    // expected-note@#dr257-A {{'dr257::A' declared here}}
+    //   expected-note@#dr257-A {{'dr257::A' declared here}}
     void f();
   };
 }
@@ -808,7 +808,7 @@ namespace dr258 { // dr258: 2.8
   int &w = b.f(0);
   int &x = b.g(0);
   // expected-error@-1 {{no matching member function for call to 'g'}}
-  // expected-note@#dr258-B-g {{candidate template ignored: invalid explicitly-specified argument for 1st template parameter}}
+  //   expected-note@#dr258-B-g {{candidate template ignored: invalid explicitly-specified argument for 1st template parameter}}
   int &y = b.h();
   float &z = const_cast(b).h();
 
@@ -826,7 +826,7 @@ namespace dr258 { // dr258: 2.8
     void f() const {}
   } f;
   // expected-error@-1 {{variable type 'struct F' is an abstract class}}
-  // expected-note@#dr258-E-f {{unimplemented pure virtual method 'f' in 'F'}}
+  //   expected-note@#dr258-E-f {{unimplemented pure virtual method 'f' in 'F'}}
 }
 
 namespace dr259 { // dr259: 4
@@ -834,34 +834,34 @@ namespace dr259 { // dr259: 4
   template struct A; // #dr259-A-int 
   template struct A;
   // expected-error@-1 {{duplicate explicit instantiation of 'A'}}
-  // expected-note@#dr259-A-int {{previous explicit instantiation is here}}
+  //   expected-note@#dr259-A-int {{previous explicit instantiation is here}}
 
   template<> struct A; // #dr259-A-float
   template struct A;
   // expected-warning@-1 {{explicit instantiation of 'A' that occurs after an explicit specialization has no effect}}
-  // expected-note@#dr259-A-float {{previous template specialization is here}}
+  //   expected-note@#dr259-A-float {{previous template specialization is here}}
 
   template struct A; // #dr259-A-char
   template<> struct A;
   // expected-error@-1 {{explicit specialization of 'dr259::A' after instantiation}}
-  // expected-note@#dr259-A-char {{explicit instantiation first required here}}
+  //   expected-note@#dr259-A-char {{explicit instantiation first required here}}
 
   template<> struct A;
   template<> struct A;
   template<> struct A {}; // #dr259-A-double
   template<> struct A {};
   // expected-error@-1 {{redefinition of 'A'}}
-  // expected-note@#dr259-A-double {{previous definition is here}}
+  //   expected-note@#dr259-A-double {{previous definition is here}}
 
   template struct B; // #dr259-B
   template struct B;
   // expected-error@-1 {{explicit instantiation of undefined template 'dr259::B'}}
-  // expected-note@#dr259-B {{template is declared here}}
+  //   expected-note@#dr259-B {{template is declared here}}
 
   template<> struct B; // #dr259-B-float
   template struct B;
   // expected-warning@-1 {{explicit instantiation of 'B' that occurs after an explicit specialization has no effect}}
-  // expected-note@#dr259-B-float {{previous template specialization is here}}
+  //   expected-note@#dr259-B-float {{previous template specialization is here}}
 }
 
 // FIXME: When dr260 is resolved, also add tests for DR507.
@@ -981,11 +981,11 @@ namespace dr275 { // dr275: no
     // FIXME: this should be rejected in c++98 too
     template void f(long);
     // since-cxx11-error@-1 {{explicit instantiation of 'dr275::N::f' must occur in namespace 'N'}}
-    // since-cxx11-note@#dr275-N-f {{explicit instantiation refers here}}
+    //   since-cxx11-note@#dr275-N-f {{explicit instantiation refers here}}
     // FIXME: this should be rejected in c++98 too
     template void N::f(unsigned long);
     // since-cxx11-error@-1 {{explicit instantiation of 'f' not in a namespace enclosing 'N'}}
-    // since-cxx11-note@#dr275-N-f {{explicit instantiation refers here}}
+    //   since-cxx11-note@#dr275-N-f {{explicit instantiation refers here}}
     template void h(long);
     // expected-error@-1 {{explicit instantiation of 'h' does not refer to a function template, variable template, member function, member class, or static data member}}
     template <> void f(double) {}
@@ -1001,22 +1001,22 @@ namespace dr275 { // dr275: no
   // FIXME: this should be rejected in c++98 too
   template void f(short);
   // since-cxx11-error@-1 {{explicit instantiation of 'dr275::N::f' must occur in namespace 'N'}}
-  // since-cxx11-note@#dr275-N-f {{explicit instantiation refers here}}
+  //   since-cxx11-note@#dr275-N-f {{explicit instantiation refers here}}
   template void N::f(unsigned short);
 
   // FIXME: this should probably be valid. the wording from the issue
   // doesn't clarify this, but it follows from the usual rules.
   template void g(int);
   // expected-error@-1 {{partial ordering for explicit instantiation of 'g' is ambiguous}}
-  // expected-note@#dr275-g {{explicit instantiation candidate function 'dr275::g' template here [with T = int]}}
-  // expected-note@#dr275-N-g {{explicit instantiation candidate function 'dr275::N::g' template here [with T = int]}}
+  //   expected-note@#dr275-g {{explicit instantiation candidate function 'dr275::g' template here [with T = int]}}
+  //   expected-note@#dr275-N-g {{explicit instantiation candidate function 'dr275::N::g' template here [with T = int]}}
 
   // FIXME: likewise, this should also be valid.
   template void f(T) {} // #dr275-f
   template void f(short);
   // expected-error@-1 {{partial ordering for explicit instantiation of 'f' is ambiguous}}
-  // expected-note@#dr275-f {{explicit instantiation candidate function 'dr275::f' template here [with T = short]}}
-  // expected-note@#dr275-N-f {{explicit instantiation candidate function 'dr275::N::f' template here [with T = short]}}
+  //   expected-note@#dr275-f {{explicit instantiation candidate function 'dr275::f' template here [with T = short]}}
+  //   expected-note@#dr275-N-f {{explicit instantiation candidate function 'dr275::N::f' template here [with T = short]}}
 }
 
 // dr276: na
@@ -1055,28 +1055,28 @@ namespace dr280 { // dr280: 2.9
     d(); // ok, public
     d(0);
     // expected-error@-1 {{'operator void (*)(int)' is a private member of 'dr280::A'}}
-    // expected-note@#dr280-D {{constrained by private inheritance here}}
-    // expected-note@#dr280-A-f1 {{member is declared here}}
+    //   expected-note@#dr280-D {{constrained by private inheritance here}}
+    //   expected-note@#dr280-A-f1 {{member is declared here}}
     d(0, 0); // ok, suppressed by member in D
     d(0, 0, 0);
     // expected-error@-1 {{'operator void (*)(int, int, int)' is a private member of 'dr280::B'}}
-    // expected-note@#dr280-B-f3 {{declared private here}}
+    //   expected-note@#dr280-B-f3 {{declared private here}}
     e();
     // expected-error@-1 {{call to object of type 'struct E' is ambiguous}}
-    // expected-note@#dr280-B-f0 {{conversion candidate of type 'void (*)()'}}
-    // expected-note@#dr280-C-f0 {{conversion candidate of type 'void (*)()'}}
+    //   expected-note@#dr280-B-f0 {{conversion candidate of type 'void (*)()'}}
+    //   expected-note@#dr280-C-f0 {{conversion candidate of type 'void (*)()'}}
     e(0);
     // expected-error@-1 {{call to object of type 'struct E' is ambiguous}}
-    // expected-note@#dr280-A-f1 {{conversion candidate of type 'void (*)(int)'}}
-    // expected-note@#dr280-C-f1 {{conversion candidate of type 'void (*)(int)'}}
+    //   expected-note@#dr280-A-f1 {{conversion candidate of type 'void (*)(int)'}}
+    //   expected-note@#dr280-C-f1 {{conversion candidate of type 'void (*)(int)'}}
     e(0, 0);
     // expected-error@-1 {{call to object of type 'struct E' is ambiguous}}
-    // expected-note@#dr280-C-f2 {{conversion candidate of type 'void (*)(int, int)'}}
-    // expected-note@#dr280-D-f2 {{conversion candidate of type 'void (*)(int, int)'}}
+    //   expected-note@#dr280-C-f2 {{conversion candidate of type 'void (*)(int, int)'}}
+    //   expected-note@#dr280-D-f2 {{conversion candidate of type 'void (*)(int, int)'}}
     e(0, 0, 0);
     // expected-error@-1 {{call to object of type 'struct E' is ambiguous}}
-    // expected-note@#dr280-B-f3 {{conversion candidate of type 'void (*)(int, int, int)'}}
-    // expected-note@#dr280-C-f3 {{conversion candidate of type 'void (*)(int, int, int)'}}
+    //   expected-note@#dr280-B-f3 {{conversion candidate of type 'void (*)(int, int, int)'}}
+    //   expected-note@#dr280-C-f3 {{conversion candidate of type 'void (*)(int, int, int)'}}
   }
 }
 
@@ -1102,10 +1102,10 @@ namespace dr283 { // dr283: yes
   struct S {
     friend class T;
     // expected-error@-1 {{declaration of 'T' shadows template parameter}}
-    // expected-note@#dr283-template {{template parameter is declared here}}
+    //   expected-note@#dr283-template {{template parameter is declared here}}
     class T;
     // expected-error@-1 {{declaration of 'T' shadows template parameter}}
-    // expected-note@#dr283-template {{template parameter is declared here}}
+    //   expected-note@#dr283-template {{template parameter is declared here}}
   };
 }
 
@@ -1152,8 +1152,8 @@ namespace dr285 { // dr285: yes
   template void f(int, T); // #dr285-f-int-T
   template<> void f(int, int) {}
   // expected-error@-1 {{function template specialization 'f' ambiguously refers to more than one function template; explicitly specify additional template arguments to identify a particular function template}}
-  // expected-note@#dr285-f-int-T {{function template 'dr285::f' matches specialization [with T = int]}}
-  // expected-note@#dr285-f-T-int {{function template 'dr285::f' matches specialization [with T = int]}}
+  //   expected-note@#dr285-f-int-T {{function template 'dr285::f' matches specialization [with T = int]}}
+  //   expected-note@#dr285-f-T-int {{function template 'dr285::f' matches specialization [with T = int]}}
 }
 
 namespace dr286 { // dr286: 2.8
@@ -1169,7 +1169,7 @@ namespace dr286 { // dr286: 2.8
 
   A::C::B absip;
   // expected-error@-1 {{'B' is a private member of 'dr286::A::C'}}
-  // expected-note@#dr286-B {{implicitly declared private here}}
+  //   expected-note@#dr286-B {{implicitly declared private here}}
 }
 
 // dr288: na
@@ -1178,11 +1178,11 @@ namespace dr289 { // dr289: yes
   struct A; // #dr289-A
   struct B : A {};
   // expected-error@-1 {{base class has incomplete type}}
-  // expected-note@#dr289-A {{forward declaration of 'dr289::A'}}
+  //   expected-note@#dr289-A {{forward declaration of 'dr289::A'}}
 
   template struct C { typename T::error error; };
   // expected-error@-1 {{type 'int' cannot be used prior to '::' because it has no members}}
-  // expected-note@#dr289-C-int {{in instantiation of template class 'dr289::C' requested here}}
+  //   expected-note@#dr289-C-int {{in instantiation of template class 'dr289::C' requested here}}
   struct D : C {}; // #dr289-C-int
 }
 
@@ -1193,7 +1193,7 @@ namespace dr289 { // dr289: yes
 namespace dr294 { // dr294: no
   void f() throw(int);
   // since-cxx17-error@-1 {{ISO C++17 does not allow dynamic exception specifications}}
-  // since-cxx17-note@-2 {{use 'noexcept(false)' instead}}
+  //   since-cxx17-note@-2 {{use 'noexcept(false)' instead}}
   int main() {
     (void)static_cast(f); // FIXME: ill-formed in C++14 and before
     // FIXME: since-cxx17-error@-1 {{static_cast from 'void (*)() throw(int)' to 'void (*)() throw()' is not allowed}}
@@ -1207,14 +1207,14 @@ namespace dr294 { // dr294: no
     // conversion, but that is being changed by core issue).
     (void)static_cast(f); // FIXME: ill-formed in C++14 and before
     // since-cxx17-error@-1 {{ISO C++17 does not allow dynamic exception specifications}}
-    // since-cxx17-note@-2 {{use 'noexcept(false)' instead}}
+    //   since-cxx17-note@-2 {{use 'noexcept(false)' instead}}
 
     void (*p)() throw() = f;
     // cxx98-14-error@-1 {{target exception specification is not superset of source}}
     // since-cxx17-error@-2 {{cannot initialize a variable of type 'void (*)() throw()' with an lvalue of type 'void () throw(int)': different exception specifications}}
     void (*q)() throw(int) = f;
     // since-cxx17-error@-1 {{ISO C++17 does not allow dynamic exception specifications}}
-    // since-cxx17-note@-2 {{use 'noexcept(false)' instead}}
+    //   since-cxx17-note@-2 {{use 'noexcept(false)' instead}}
   }
 }
 
@@ -1260,10 +1260,10 @@ namespace dr298 { // dr298: 3.1
   struct A a;
   struct B b;
   // expected-error@-1 {{typedef 'B' cannot be referenced with a struct specifier}}
-  // expected-note@#dr298-B {{declared here}}
+  //   expected-note@#dr298-B {{declared here}}
   struct C c;
   // expected-error@-1 {{typedef 'C' cannot be referenced with a struct specifier}}
-  // expected-note@#dr298-C {{declared here}}
+  //   expected-note@#dr298-C {{declared here}}
 
   B::B() {}
   // expected-error@-1 {{a type specifier is required for all declarations}}
@@ -1274,7 +1274,7 @@ namespace dr298 { // dr298: 3.1
   typedef struct D E; // #dr298-E
   struct E {};
   // expected-error@-1 {{definition of type 'E' conflicts with typedef of the same name}}
-  // expected-note@#dr298-E {{'E' declared here}}
+  //   expected-note@#dr298-E {{'E' declared here}}
 
   struct F {
     ~F();
diff --git a/clang/test/CXX/drs/dr3xx.cpp b/clang/test/CXX/drs/dr3xx.cpp
index b1fcf709c2273..4584801f9f971 100644
--- a/clang/test/CXX/drs/dr3xx.cpp
+++ b/clang/test/CXX/drs/dr3xx.cpp
@@ -26,7 +26,7 @@ namespace dr301 { // dr301: 3.5
     //   cxx20-23-note@-3 {{to match this '<'}} 
     bool c = (void(*)(S, S))operator+ < (void(*)(S, S))operator-;
     // expected-error@-1 {{expected '>'}}
-    // expected-note@-2 {{to match this '<'}}
+    //   expected-note@-2 {{to match this '<'}}
   }
 
   template void f() {
@@ -71,8 +71,8 @@ namespace dr302 { // dr302: 3.0
 #if __cplusplus < 201103L
   struct B {
   // expected-error@-1 {{implicit default constructor for 'dr302::B' must explicitly initialize the const member 'n'}}
-  // expected-note@#dr302-b {{in implicit default constructor for 'dr302::B' first required here}}
-  // expected-note@#dr302-B-n {{declared here}}
+  //   expected-note@#dr302-b {{in implicit default constructor for 'dr302::B' first required here}}
+  //   expected-note@#dr302-B-n {{declared here}}
     const int n; // #dr302-B-n
     A a;
   } b = B(); // #dr302-b
@@ -86,13 +86,13 @@ namespace dr302 { // dr302: 3.0
     A a;
   } b = B();
   // expected-error@-1 {{call to implicitly-deleted default constructor of 'B'}}
-  // expected-note@#dr302-B-n {{default constructor of 'B' is implicitly deleted because field 'n' of const-qualified type 'const int' would not be initialized}}
+  //   expected-note@#dr302-B-n {{default constructor of 'B' is implicitly deleted because field 'n' of const-qualified type 'const int' would not be initialized}}
   // C::C() is called here, because even though it's trivial, it's deleted.
   struct C {
     const int n; // #dr302-C-n
   } c = C();
   // expected-error@-1 {{call to implicitly-deleted default constructor of 'C'}}
-  // expected-note@#dr302-C-n {{default constructor of 'C' is implicitly deleted because field 'n' of const-qualified type 'const int' would not be initialized}}
+  //   expected-note@#dr302-C-n {{default constructor of 'C' is implicitly deleted because field 'n' of const-qualified type 'const int' would not be initialized}}
   struct D {
     const int n = 0;
   } d = D();
@@ -132,7 +132,7 @@ namespace dr305 { // dr305: no
     struct B {}; // #dr305-h-B
     b->~B();
     // expected-error@-1 {{destructor type 'B' in object destruction expression does not match the type 'B' (aka 'dr305::A') of the object being destroyed}}
-    // expected-note@#dr305-h-B {{type 'B' found by destructor name lookup}}
+    //   expected-note@#dr305-h-B {{type 'B' found by destructor name lookup}}
   }
 
   template struct X {};
@@ -186,8 +186,8 @@ namespace dr306 { // dr306: dup 39
   Z::X zx;
   Z::X zcx;
   // expected-error@-1 {{member 'X' found in multiple base classes of different types}}
-  // expected-note@#dr306-X {{member type 'dr306::X' found}}
-  // expected-note@#dr306-typedef-X {{member type 'const dr306::X' found}}
+  //   expected-note@#dr306-X {{member type 'dr306::X' found}}
+  //   expected-note@#dr306-typedef-X {{member type 'const dr306::X' found}}
 }
 
 // dr307: na
@@ -214,7 +214,7 @@ namespace dr308 { // dr308: 3.7
       // unreachable
     } catch (const B&) {
       // expected-warning@-1 {{exception of type 'const B &' will be caught by earlier handler}}
-      // expected-note@#dr308-catch-A {{for type 'const A &'}}
+      //   expected-note@#dr308-catch-A {{for type 'const A &'}}
       // get here instead
     }
   }
@@ -269,7 +269,7 @@ namespace dr317 { // dr317: 3.5
   void f() {} // #dr317-f
   inline void f();
   // expected-error@-1 {{inline declaration of 'f' follows non-inline definition}}
-  // expected-note@#dr317-f {{previous definition is here}}
+  //   expected-note@#dr317-f {{previous definition is here}}
 
   int g();
   int n = g();
@@ -280,7 +280,7 @@ namespace dr317 { // dr317: 3.5
   int h() { return 0; } // #dr317-h
   inline int h();
   // expected-error@-1 {{inline declaration of 'h' follows non-inline definition}}
-  // expected-note@#dr317-h {{previous definition is here}}
+  //   expected-note@#dr317-h {{previous definition is here}}
 }
 
 namespace dr318 { // dr318: sup 1310
@@ -372,12 +372,12 @@ namespace dr324 { // dr324: 3.6
   struct S { int n : 1; } s; // #dr324-n
   int &a = s.n;
   // expected-error@-1 {{non-const reference cannot bind to bit-field 'n'}}
-  // expected-note@#dr324-n {{bit-field is declared here}}
+  //   expected-note@#dr324-n {{bit-field is declared here}}
   int *b = &s.n;
   // expected-error@-1 {{address of bit-field requested}}
   int &c = (s.n = 0);
   // expected-error@-1 {{non-const reference cannot bind to bit-field 'n'}}
-  // expected-note@#dr324-n {{bit-field is declared here}}
+  //   expected-note@#dr324-n {{bit-field is declared here}}
   int *d = &(s.n = 0);
   // expected-error@-1 {{address of bit-field requested}}
   // FIXME: why don't we emit a note here, as for the rest of this type of diagnostic in this test?
@@ -387,7 +387,7 @@ namespace dr324 { // dr324: 3.6
   // expected-error@-1 {{address of bit-field requested}}
   int &g = (void(), s.n);
   // expected-error@-1 {{non-const reference cannot bind to bit-field 'n'}}
-  // expected-note@#dr324-n {{bit-field is declared here}}
+  //   expected-note@#dr324-n {{bit-field is declared here}}
   int *h = &(void(), s.n);
   // expected-error@-1 {{address of bit-field requested}}
   int *i = &++s.n;
@@ -411,13 +411,13 @@ namespace dr328 { // dr328: yes
   struct A; // #dr328-A
   struct B { A a; };
   // expected-error@-1 {{field has incomplete type 'A'}}
-  // expected-note@#dr328-A {{forward declaration of 'dr328::A'}}
+  //   expected-note@#dr328-A {{forward declaration of 'dr328::A'}}
   template struct C { A a; };
   // expected-error@-1 {{field has incomplete type 'A'}}
-  // expected-note@#dr328-A {{forward declaration of 'dr328::A'}}
+  //   expected-note@#dr328-A {{forward declaration of 'dr328::A'}}
   A *p = new A[0];
   // expected-error@-1 {{allocation of incomplete type 'A'}}
-  // expected-note@#dr328-A {{forward declaration of 'dr328::A'}}
+  //   expected-note@#dr328-A {{forward declaration of 'dr328::A'}}
 }
 
 namespace dr329 { // dr329: 3.5
@@ -426,11 +426,11 @@ namespace dr329 { // dr329: 3.5
     friend void f(A a) { g(a); }
     friend void h(A a) { g(a); }
     // expected-error@-1 {{use of undeclared identifier 'g'}}
-    // expected-note@#dr329-h-call {{in instantiation of member function 'dr329::h' requested here}}
+    //   expected-note@#dr329-h-call {{in instantiation of member function 'dr329::h' requested here}}
     friend void i(B b) {} // #dr329-i
     // expected-error@-1 {{redefinition of 'i'}}
-    // expected-note@#dr329-b {{in instantiation of template class 'dr329::A' requested here}}
-    // expected-note@#dr329-i {{previous definition is here}}
+    //   expected-note@#dr329-b {{in instantiation of template class 'dr329::A' requested here}}
+    //   expected-note@#dr329-i {{previous definition is here}}
   };
   A a; 
   A b; // #dr329-b
@@ -540,10 +540,10 @@ namespace dr331 { // dr331: 11
   };
   const A a;
   // expected-error@-1 {{no matching constructor for initialization of 'const A'}}
-  // expected-note@#dr331-A-ctor {{candidate constructor not viable: requires 1 argument, but 0 were provided}}
+  //   expected-note@#dr331-A-ctor {{candidate constructor not viable: requires 1 argument, but 0 were provided}}
   const A b(a);
   // expected-error@-1 {{no matching constructor for initialization of 'const A'}}
-  // expected-note@#dr331-A-ctor {{candidate constructor not viable: 1st argument ('const A') would lose const qualifier}}
+  //   expected-note@#dr331-A-ctor {{candidate constructor not viable: 1st argument ('const A') would lose const qualifier}}
 }
 
 namespace dr332 { // dr332: dup 577
@@ -638,7 +638,7 @@ namespace dr339 { // dr339: 2.8
     A<2> b = f(0.0f);
     A<3> c = f("foo");
     // expected-error@-1 {{no matching function}}
-    // expected-note@#dr339-f {{candidate}}
+    //   expected-note@#dr339-f {{candidate}}
   }
 
 
@@ -673,30 +673,30 @@ namespace dr341 { // dr341: sup 1708
   namespace B {
     extern "C" int &dr341_a = dr341_a;
     // expected-error@-1 {{redefinition of 'dr341_a'}}
-    // expected-note@#dr341_a {{previous definition is here}} 
+    //   expected-note@#dr341_a {{previous definition is here}} 
   }
   extern "C" void dr341_b(); // #dr341_b 
 }
 int dr341_a;
 // expected-error@-1 {{declaration of 'dr341_a' in global scope conflicts with declaration with C language linkage}}
-// expected-note@#dr341_a {{declared with C language linkage here}}
+//   expected-note@#dr341_a {{declared with C language linkage here}}
 int dr341_b;
 // expected-error@-1 {{declaration of 'dr341_b' in global scope conflicts with declaration with C language linkage}}
-// expected-note@#dr341_b {{declared with C language linkage here}}
+//   expected-note@#dr341_b {{declared with C language linkage here}}
 int dr341_c; // #dr341_c
 int dr341_d; // #dr341_d
 namespace dr341 {
   extern "C" int dr341_c;
   // expected-error@-1 {{declaration of 'dr341_c' with C language linkage conflicts with declaration in global scope}}
-  // expected-note@#dr341_c {{declared in global scope here}}
+  //   expected-note@#dr341_c {{declared in global scope here}}
   extern "C" void dr341_d();
   // expected-error@-1 {{declaration of 'dr341_d' with C language linkage conflicts with declaration in global scope}}
-  // expected-note@#dr341_d {{declared in global scope here}}
+  //   expected-note@#dr341_d {{declared in global scope here}}
 
   namespace A { extern "C" int dr341_e; } // #dr341_e 
   namespace B { extern "C" void dr341_e(); }
   // expected-error@-1 {{redefinition of 'dr341_e' as different kind of symbol}}
-  // expected-note@#dr341_e {{previous definition is here}}
+  //   expected-note@#dr341_e {{previous definition is here}}
 }
 
 // dr342: na
@@ -729,8 +729,8 @@ namespace dr345 { // dr345: yes
   };
   template  void f(T t) { typename T::X x; }
   // expected-error@-1 {{typename specifier refers to non-type member 'X' in 'dr345::A'}}
-  // expected-note@#dr345-f-a {{in instantiation of function template specialization 'dr345::f' requested here}}
-  // expected-note@#dr345-int-X {{referenced member 'X' is declared here}}
+  //   expected-note@#dr345-f-a {{in instantiation of function template specialization 'dr345::f' requested here}}
+  //   expected-note@#dr345-int-X {{referenced member 'X' is declared here}}
   void f(A a, B b) {
     f(b);
     f(a); // #dr345-f-a
@@ -768,7 +768,7 @@ namespace dr349 { // dr349: no
       return p;
       // cxx98-20-error@-1 {{cannot initialize return object of type 'const int ***' with an lvalue of type 'int ***'}}
       // since-cxx23-error@-2 {{cannot initialize return object of type 'const int ***' with an rvalue of type 'int ***'}}
-      // expected-note@#dr349-p1 {{in instantiation of function template specialization 'dr349::A::operator const int ***' requested here}}
+      //   expected-note@#dr349-p1 {{in instantiation of function template specialization 'dr349::A::operator const int ***' requested here}}
     }
   };
 
@@ -804,12 +804,12 @@ namespace dr352 { // dr352: 2.8
     void f(A::E e) {
       foo(e, &arg);
       // expected-error@-1 {{no matching function for call to 'foo'}}
-      // expected-note@#dr352-foo {{candidate template ignored: couldn't infer template argument 'R'}}
+      //   expected-note@#dr352-foo {{candidate template ignored: couldn't infer template argument 'R'}}
 
       using A::foo;
       foo(e, &arg);
       // expected-error@-1 {{attempt to use a deleted function}}
-      // expected-note@#dr352-deleted {{'arg' has been explicitly marked deleted here}}
+      //   expected-note@#dr352-deleted {{'arg' has been explicitly marked deleted here}}
     }
 
     int arg(int);
@@ -817,7 +817,7 @@ namespace dr352 { // dr352: 2.8
     void g(A::E e) {
       foo(e, &arg);
       // expected-error@-1 {{no matching function for call to 'foo'}}
-      // expected-note@#dr352-foo {{candidate template ignored: couldn't infer template argument 'R'}} 
+      //   expected-note@#dr352-foo {{candidate template ignored: couldn't infer template argument 'R'}} 
 
       using A::foo;
       foo(e, &arg); // ok, uses non-template
@@ -834,7 +834,7 @@ namespace dr352 { // dr352: 2.8
       f1(a);
       f2(a);
       // expected-error@-1 {{no matching function for call to 'f2'}}
-      // expected-note@#dr352-f2 {{candidate template ignored: couldn't infer template argument 'I'}}
+      //   expected-note@#dr352-f2 {{candidate template ignored: couldn't infer template argument 'I'}}
       f3(a, b);
     }
 
@@ -848,7 +848,7 @@ namespace dr352 { // dr352: 2.8
       g1(a);
       g2(a);
       // expected-error@-1 {{no matching function for call to 'g2'}}
-      // expected-note@#dr352-g2 {{candidate template ignored: couldn't infer template argument 'I'}}
+      //   expected-note@#dr352-g2 {{candidate template ignored: couldn't infer template argument 'I'}}
       g3(a, b);
     }
 
@@ -857,7 +857,7 @@ namespace dr352 { // dr352: 2.8
     void h() {
       h1();
       // expected-error@-1 {{no matching function for call to 'h1'}}
-      // expected-note@#dr352-h1 {{candidate template ignored: couldn't infer template argument 'T'}}
+      //   expected-note@#dr352-h1 {{candidate template ignored: couldn't infer template argument 'T'}}
       h1(0);
       h1();
       h2(0);
@@ -874,20 +874,20 @@ namespace dr352 { // dr352: 2.8
       extern int ambig(float), ambig(int);
       i1(ambig);
       // expected-error@-1 {{no matching function for call to 'i1'}}
-      // expected-note@#dr352-i1 {{candidate template ignored: couldn't infer template argument 'R'}}
+      //   expected-note@#dr352-i1 {{candidate template ignored: couldn't infer template argument 'R'}}
       i2(0, 0, ambig);
 
       extern void no_match(float), no_match(int);
       i1(no_match);
       // expected-error@-1 {{no matching function for call to 'i1'}}
-      // expected-note@#dr352-i1 {{candidate template ignored: couldn't infer template argument 'R'}}
+      //   expected-note@#dr352-i1 {{candidate template ignored: couldn't infer template argument 'R'}}
       i2(0, 0, no_match);
       // expected-error@-1 {{no matching function for call to 'i2'}}
-      // expected-note@#dr352-i2 {{candidate function [with R = int, A = int] not viable: no overload of 'no_match' matching 'int (*)(int)' for 3rd argument}}
+      //   expected-note@#dr352-i2 {{candidate function [with R = int, A = int] not viable: no overload of 'no_match' matching 'int (*)(int)' for 3rd argument}}
 
       i1(tmpl);
       // expected-error@-1 {{no matching function for call to 'i1'}}
-      // expected-note@#dr352-i1 {{candidate template ignored: couldn't infer template argument 'R'}}
+      //   expected-note@#dr352-i1 {{candidate template ignored: couldn't infer template argument 'R'}}
       i2(0, 0, tmpl);
     }
   }
@@ -923,7 +923,7 @@ namespace dr352 { // dr352: 2.8
     void h(A<1> a1, A<2> a2) {
       g(a1);
       // expected-error@-1 {{no matching function for call to 'g'}}
-      // expected-note@#dr352-g {{candidate template ignored: couldn't infer template argument 'I'}}
+      //   expected-note@#dr352-g {{candidate template ignored: couldn't infer template argument 'I'}}
       g<0>(a1);
       f(a1, a2);
     }
@@ -964,8 +964,8 @@ namespace dr354 { // dr354: yes c++11
   int b0 = both<0>();
   int b1 = both<(int*)0>();
   // cxx98-error@-1 {{no matching function for call to 'both'}}
-  // cxx98-note@#dr354-both-int-ptr {{candidate template ignored: invalid explicitly-specified argument for 1st template parameter}}
-  // cxx98-note@#dr354-both-int {{candidate template ignored: invalid explicitly-specified argument for 1st template parameter}}
+  //   cxx98-note@#dr354-both-int-ptr {{candidate template ignored: invalid explicitly-specified argument for 1st template parameter}}
+  //   cxx98-note@#dr354-both-int {{candidate template ignored: invalid explicitly-specified argument for 1st template parameter}}
 
   template struct ptr_mem {}; // #dr354-ptr_mem
   ptr_mem<0> m0; // #dr354-m0
@@ -1003,7 +1003,7 @@ namespace dr357 { // dr357: yes
   };
   template void A::f() {}
   // expected-error@-1 {{out-of-line definition of 'f' does not match any declaration in 'A'}}
-  // expected-note@#dr357-f {{member declaration does not match because it is const qualified}}
+  //   expected-note@#dr357-f {{member declaration does not match because it is const qualified}}
 
   struct B {
     template void f();
@@ -1070,10 +1070,10 @@ struct B : A {
 int main() {
   int foo = B().foo();
   // expected-error@-1 {{'foo' is a private member of 'dr360::B'}}
-  // expected-note@#dr360-using-foo {{declared private here}}
+  //   expected-note@#dr360-using-foo {{declared private here}}
   int bar = B().bar();
   // expected-error@-1 {{'bar' is a protected member of 'dr360::B'}}
-  // expected-note@#dr360-using-bar {{declared protected here}}
+  //   expected-note@#dr360-using-bar {{declared protected here}}
   int baz = B().baz();
 }
 } // namespace dr360
@@ -1120,7 +1120,7 @@ namespace dr368 { // dr368: 3.6
   template struct S {}; // #dr368-S
   template int f(S *);
   // expected-error@-1 {{template argument for non-type template parameter is treated as function type 'T ()'}}
-  // expected-note@#dr368-S {{template parameter is declared here}}
+  //   expected-note@#dr368-S {{template parameter is declared here}}
   template int g(S *); // #dr368-g
   template int g(S *); // #dr368-g-2
   struct X {};
@@ -1151,16 +1151,16 @@ namespace dr372 { // dr372: no
       T1,
       T2::Type> {};
       // expected-error@-1 {{'Type' is a protected member of 'dr372::example1::X'}}
-      // expected-note@#dr372-z1 {{in instantiation of template class 'dr372::example1::Z1' requested here}}
-      // expected-note@#dr372-ex1-Type {{declared protected here}}
+      //   expected-note@#dr372-z1 {{in instantiation of template class 'dr372::example1::Z1' requested here}}
+      //   expected-note@#dr372-ex1-Type {{declared protected here}}
 
     template class T1,
              template class T2> struct Z2 :
       T2::Type>,
       // expected-error@-1 {{'Type' is a protected member of 'dr372::example1::X'}}
-      // expected-note@#dr372-z2 {{in instantiation of template class 'dr372::example1::Z2' requested here}}
-      // expected-note@#dr372-ex1-Type {{declared protected here}}
+      //   expected-note@#dr372-z2 {{in instantiation of template class 'dr372::example1::Z2' requested here}}
+      //   expected-note@#dr372-ex1-Type {{declared protected here}}
       T1 {};
 
     Z1 z1; // #dr372-z1
@@ -1175,8 +1175,8 @@ namespace dr372 { // dr372: no
     template struct A {
       typename T::Type t;
       // expected-error@-1 {{'Type' is a private member of 'dr372::example2::X'}}
-      // expected-note@#dr372-ax {{in instantiation of template class 'dr372::example2::A' requested here}}
-      // expected-note@#dr372-ex2-Type {{declared private here}}
+      //   expected-note@#dr372-ax {{in instantiation of template class 'dr372::example2::A' requested here}}
+      //   expected-note@#dr372-ex2-Type {{declared private here}}
     };
     A ax; // #dr372-ax
   }
@@ -1190,12 +1190,12 @@ namespace dr372 { // dr372: no
     template struct B {};
     template struct C : U, B {};
     // expected-error@-1 {{'N' is a protected member of 'dr372::example3::A'}}
-    // expected-note@#dr372-x {{in instantiation of template class 'dr372::example3::C' requested here}}
-    // expected-note@#dr372-N {{declared protected here}}
+    //   expected-note@#dr372-x {{in instantiation of template class 'dr372::example3::C' requested here}}
+    //   expected-note@#dr372-N {{declared protected here}}
     template struct D : B, U {};
     // expected-error@-1 {{'N' is a protected member of 'dr372::example3::A'}}
-    // expected-note@#dr372-y {{in instantiation of template class 'dr372::example3::D' requested here}}
-    // expected-note@#dr372-N {{declared protected here}}
+    //   expected-note@#dr372-y {{in instantiation of template class 'dr372::example3::D' requested here}}
+    //   expected-note@#dr372-N {{declared protected here}}
 
     C x; // #dr372-x
     D y; // #dr372-y
@@ -1223,7 +1223,7 @@ namespace dr372 { // dr372: no
     };
     struct D : A::B, A {};
     // expected-error@-1 {{'B' is a protected member of 'dr372::std_example::A'}}
-    // expected-note@#dr372-B-std {{declared protected here}}
+    //   expected-note@#dr372-B-std {{declared protected here}}
   }
 
   // FIXME: This is valid: deriving from A::B gives access to A::B!
@@ -1235,7 +1235,7 @@ namespace dr372 { // dr372: no
     struct A::B : A {};
     struct C : A::B {};
     // expected-error@-1 {{'B' is a protected member of 'dr372::badwolf::A'}}
-    // expected-note@#dr372-B {{declared protected here}}
+    //   expected-note@#dr372-B {{declared protected here}}
   }
 }
 
@@ -1246,7 +1246,7 @@ namespace dr373 { // dr373: 5
       using namespace dr373::X;
       int k = dr373;
       // expected-error@-1 {{'dr373' does not refer to a value}}
-      // expected-note@#dr373-struct {{declared here}}
+      //   expected-note@#dr373-struct {{declared here}}
       namespace Y = dr373::X;
       k = Y::dr373;
     }
@@ -1255,10 +1255,10 @@ namespace dr373 { // dr373: 5
   struct A { struct B {}; }; // #dr373-A 
   namespace X = A::B;
   // expected-error@-1 {{expected namespace name}}
-  // expected-note@#dr373-A {{'A' declared here}}
+  //   expected-note@#dr373-A {{'A' declared here}}
   using namespace A::B;
   // expected-error@-1 {{expected namespace name}}
-  // expected-note@#dr373-A {{'A' declared here}}
+  //   expected-note@#dr373-A {{'A' declared here}}
 }
 
 namespace dr374 { // dr374: 7
@@ -1361,8 +1361,8 @@ namespace dr385 { // dr385: 2.8
   struct F : E { friend int i(E); };
   int i(E e) { return e.n; }
   // expected-error@-1 {{'n' is a protected member of 'dr385::D'}}
-  // expected-note@#dr385-E {{constrained by protected inheritance here}}
-  // expected-note@#dr385-n {{member is declared here}}
+  //   expected-note@#dr385-E {{constrained by protected inheritance here}}
+  //   expected-note@#dr385-n {{member is declared here}}
 }
 
 namespace dr387 { // dr387: 2.8
@@ -1375,10 +1375,10 @@ namespace dr387 { // dr387: 2.8
     void g() {
       number a(3);
       // expected-error@-1 {{calling a private constructor of class 'dr387::old::number'}}
-      // expected-note@#dr387-number-ctor {{implicitly declared private here}}
+      //   expected-note@#dr387-number-ctor {{implicitly declared private here}}
       number b(4);
       // expected-error@-1 {{calling a private constructor of class 'dr387::old::number'}}
-      // expected-note@#dr387-number-ctor {{implicitly declared private here}}
+      //   expected-note@#dr387-number-ctor {{implicitly declared private here}}
       a = gcd(a, b);
       b = gcd(3, 4);
       // expected-error@-1 {{use of undeclared identifier 'gcd'}}
@@ -1451,19 +1451,19 @@ namespace dr389 { // dr389: no
 
   typedef T BadArg1;
   // expected-error@-1 {{template argument uses unnamed type}}
-  // expected-note@#dr389-no-link-1 {{unnamed type used in template argument was declared here}}
+  //   expected-note@#dr389-no-link-1 {{unnamed type used in template argument was declared here}}
   typedef T BadArg2;
   // expected-error@-1 {{template argument uses unnamed type}}
-  // expected-note@#dr389-no-link-2 {{unnamed type used in template argument was declared here}}
+  //   expected-note@#dr389-no-link-2 {{unnamed type used in template argument was declared here}}
   typedef T BadArg3;
   // expected-error@-1 {{template argument uses unnamed type}}
-  // expected-note@#dr389-C {{unnamed type used in template argument was declared here}}
+  //   expected-note@#dr389-C {{unnamed type used in template argument was declared here}}
   typedef T BadArg4;
   // expected-error@-1 {{template argument uses unnamed type}}
-  // expected-note@#dr389-D {{unnamed type used in template argument was declared here}}
+  //   expected-note@#dr389-D {{unnamed type used in template argument was declared here}}
   typedef T BadArg5;
   // expected-error@-1 {{template argument uses unnamed type}}
-  // expected-note@#dr389-C {{unnamed type used in template argument was declared here}}
+  //   expected-note@#dr389-C {{unnamed type used in template argument was declared here}}
 #endif
 
   extern WithLinkage1 withLinkage1;
@@ -1530,14 +1530,14 @@ namespace dr390 { // dr390: 3.3
   struct A {
     A() { f(); }
     // expected-warning@-1 {{call to pure virtual member function 'f' has undefined behavior; overrides of 'f' in subclasses are not available in the constructor of 'dr390::A'}}
-    // expected-note@#dr390-A-int {{in instantiation of member function 'dr390::A::A' requested here}}
-    // expected-note@#dr390-f {{'f' declared here}}
+    //   expected-note@#dr390-A-int {{in instantiation of member function 'dr390::A::A' requested here}}
+    //   expected-note@#dr390-f {{'f' declared here}}
     virtual void f() = 0; // #dr390-f
     virtual ~A() = 0;
   };
   template A::~A() { T::error; }
   // expected-error@-1 {{type 'int' cannot be used prior to '::' because it has no members}}
-  // expected-note@#dr390-A-int {{in instantiation of member function 'dr390::A::~A' requested here}}
+  //   expected-note@#dr390-A-int {{in instantiation of member function 'dr390::A::~A' requested here}}
   template void A::f() { T::error; } // ok, not odr-used
   struct B : A { // #dr390-A-int
     void f() {}
@@ -1550,14 +1550,14 @@ namespace dr391 { // dr391: 2.8 c++11
   A fa();
   const A &a = fa();
   // cxx98-error@-1 {{C++98 requires an accessible copy constructor for class 'dr391::A' when binding a reference to a temporary; was private}}
-  // cxx98-note@#dr391-A {{implicitly declared private here}}
+  //   cxx98-note@#dr391-A {{implicitly declared private here}}
 
   struct B { B(const B&) = delete; }; // #dr391-B
   // cxx98-error@-1 {{deleted function definitions are a C++11 extension}}
   B fb();
   const B &b = fb();
   // cxx98-error@-1 {{copying variable of type 'B' invokes deleted constructor}}
-  // cxx98-note@#dr391-B {{'B' has been explicitly marked deleted here}}
+  //   cxx98-note@#dr391-B {{'B' has been explicitly marked deleted here}}
 
   template
   struct C {
@@ -1600,7 +1600,7 @@ namespace dr395 { // dr395: 3.0
   } null1;
   int (S::*p)() = null1;
   // expected-error@-1 {{no viable conversion from 'struct null1_t' to 'int (dr395::S::*)()'}}
-  // expected-note@#dr395-conv-func {{candidate template ignored: couldn't infer template argument 'T'}}
+  //   expected-note@#dr395-conv-func {{candidate template ignored: couldn't infer template argument 'T'}}
 
   template  using id = T;
   // cxx98-error@-1 {{alias declarations are a C++11 extension}}
@@ -1664,16 +1664,16 @@ namespace dr398 { // dr398: yes
     void test() {
       f(0);
       // expected-error@-1 {{no matching function for call to 'f'}}
-      // expected-note@#dr398-f {{candidate template ignored: substitution failure [with T = A]: no type named 'Y' in 'dr398::example2::A'}}
+      //   expected-note@#dr398-f {{candidate template ignored: substitution failure [with T = A]: no type named 'Y' in 'dr398::example2::A'}}
       f(0);
       // expected-error@-1 {{no matching function for call to 'f'}}
-      // expected-note@#dr398-f {{candidate template ignored: substitution failure [with T = B]: typename specifier refers to non-type member 'Y' in 'dr398::example2::B'}}
+      //   expected-note@#dr398-f {{candidate template ignored: substitution failure [with T = B]: typename specifier refers to non-type member 'Y' in 'dr398::example2::B'}}
       g(0);
       // expected-error@-1 {{no matching function for call to 'g'}}
-      // expected-note@#dr398-g {{candidate template ignored: substitution failure [with T = C]: missing 'typename' prior to dependent type name 'C::N'}}
+      //   expected-note@#dr398-g {{candidate template ignored: substitution failure [with T = C]: missing 'typename' prior to dependent type name 'C::N'}}
       h(0);
       // expected-error@-1 {{no matching function for call to 'h'}}
-      // expected-note@#dr398-h {{candidate template ignored: substitution failure [with T = D]: 'TT' following the 'template' keyword does not refer to a template}}
+      //   expected-note@#dr398-h {{candidate template ignored: substitution failure [with T = D]: 'TT' following the 'template' keyword does not refer to a template}}
     }
   }
 }
@@ -1690,7 +1690,7 @@ namespace dr399 { // dr399: 11
   void f() {
     D_object.~B();
     // expected-error@-1 {{destructor type 'dr399::B' in object destruction expression does not match the type 'D' of the object being destroyed}}
-    // expected-note@#dr399-B {{type 'dr399::B' found by destructor name lookup}}
+    //   expected-note@#dr399-B {{type 'dr399::B' found by destructor name lookup}}
     D_object.B::~B();
     D_object.D::~B(); // FIXME: Missing diagnostic for this.
     B_ptr->~B();
@@ -1743,7 +1743,7 @@ namespace dr399 { // dr399: 11
     // Rejecting this seems correct, but most compilers accept, so we do also.
     f.N::F::~G();
     // expected-error@-1 {{qualified destructor name only found in lexical scope; omit the qualifier to find this type name by unqualified lookup}}
-    // expected-note@#dr399-G {{type 'G' (aka 'E') found by destructor name lookup}}
+    //   expected-note@#dr399-G {{type 'G' (aka 'E') found by destructor name lookup}}
   }
 
   // Bizarrely, compilers perform lookup in the scope for qualified destructor
@@ -1757,7 +1757,7 @@ namespace dr399 { // dr399: 11
       typedef typename N::S::Inner T;
       p->::dr399::QualifiedLookupInScope::N::S::Inner::~T();
       // expected-error@-1 {{no type named 'T' in 'dr399::QualifiedLookupInScope::N::S'}}
-      // expected-note@#dr399-f {{in instantiation of function template specialization 'dr399::QualifiedLookupInScope::f' requested here}}
+      //   expected-note@#dr399-f {{in instantiation of function template specialization 'dr399::QualifiedLookupInScope::f' requested here}}
     }
     template void f(N::S::Inner *); // #dr399-f
 
diff --git a/clang/test/CXX/drs/dr4xx.cpp b/clang/test/CXX/drs/dr4xx.cpp
index fa90764df9b03..aa7a41b95ccd8 100644
--- a/clang/test/CXX/drs/dr4xx.cpp
+++ b/clang/test/CXX/drs/dr4xx.cpp
@@ -28,8 +28,8 @@ namespace dr400 { // dr400: yes
   };
   struct E : A, B { struct a b; };
   // expected-error@-1 {{member 'a' found in multiple base classes of different types}}
-  // expected-note@#dr400-A {{member type 'dr400::A::a' found by ambiguous name lookup}}
-  // expected-note@#dr400-B {{member type 'dr400::B::a' found by ambiguous name lookup}}
+  //   expected-note@#dr400-A {{member type 'dr400::A::a' found by ambiguous name lookup}}
+  //   expected-note@#dr400-B {{member type 'dr400::B::a' found by ambiguous name lookup}}
 }
 
 namespace dr401 { // dr401: 2.8
@@ -80,7 +80,7 @@ namespace dr401 { // dr401: 2.8
   //   cxx98-note@#dr402-B-type {{declared protected here}}
   void g(B b) { f(b); } // #dr402-f-b
   // since-cxx11-error@-1 {{no matching function for call to 'f'}}
-  // since-cxx11-note@#dr402-f {{candidate template ignored: substitution failure [with T = B, U = typename B::type]: 'type' is a protected member of 'dr401::B'}}
+  //   since-cxx11-note@#dr402-f {{candidate template ignored: substitution failure [with T = B, U = typename B::type]: 'type' is a protected member of 'dr401::B'}}
 }
 
 namespace dr403 { // dr403: yes
@@ -117,16 +117,16 @@ namespace dr405 { // dr405: yes
     // expected-error@-1 {{called object type 'int' is not a function or function pointer}}
     void test2(A::S as) { void f(); f(as); }
     // expected-error@-1 {{too many arguments to function call, expected 0, have 1}}
-    // expected-note@-2 {{'f' declared here}}
+    //   expected-note@-2 {{'f' declared here}}
     void test3(A::S as) { using A::f; f(as); } // ok
     void test4(A::S as) { using B::f; f(as); } // ok
     void test5(A::S as) { int f; f(as); }
     // expected-error@-1 {{called object type 'int' is not a function or function pointer}}
     void test6(A::S as) { struct f {}; (void) f(as); }
     // expected-error@-1 {{no matching conversion for functional-style cast from 'A::S' to 'f'}}
-    // expected-note@-2 {{candidate constructor (the implicit copy constructor) not viable: no known conversion from 'A::S' to 'const f' for 1st argument}}
-    // since-cxx11-note@-3 {{candidate constructor (the implicit move constructor) not viable: no known conversion from 'A::S' to 'f' for 1st argument}}
-    // expected-note@-4 {{candidate constructor (the implicit default constructor) not viable: requires 0 arguments, but 1 was provided}}
+    //   expected-note@-2 {{candidate constructor (the implicit copy constructor) not viable: no known conversion from 'A::S' to 'const f' for 1st argument}}
+    //   since-cxx11-note@-3 {{candidate constructor (the implicit move constructor) not viable: no known conversion from 'A::S' to 'f' for 1st argument}}
+    //   expected-note@-4 {{candidate constructor (the implicit default constructor) not viable: requires 0 arguments, but 1 was provided}}
   };
 
   namespace D {
@@ -165,7 +165,7 @@ namespace dr407 { // dr407: 3.8
       typedef struct S S; // #dr407-typedef-S
       struct S *p;
       // expected-error@-1 {{typedef 'S' cannot be referenced with a struct specifier}}
-      // expected-note@#dr407-typedef-S {{declared here}}
+      //   expected-note@#dr407-typedef-S {{declared here}}
     }
   }
   struct S {};
@@ -182,8 +182,8 @@ namespace dr407 { // dr407: 3.8
       using namespace B;
       struct S s;
       // expected-error@-1 {{ambiguous}}
-      // expected-note@#dr407-A-S {{candidate found by name lookup is 'dr407::UsingDir::A::S'}}
-      // expected-note@#dr407-B-S {{candidate found by name lookup is 'dr407::UsingDir::B::S'}}
+      //   expected-note@#dr407-A-S {{candidate found by name lookup is 'dr407::UsingDir::A::S'}}
+      //   expected-note@#dr407-B-S {{candidate found by name lookup is 'dr407::UsingDir::B::S'}}
     }
     namespace D {
       using A::S;
@@ -279,7 +279,7 @@ namespace dr410 { // dr410: no
   template<> void f(int) { M::A::z(); }
   void g(int) { M::A::z(); }
   // expected-error@-1 {{'z' is a private member of 'dr410::M::A'}}
-  // expected-note@#dr410-z {{declared private here}}
+  //   expected-note@#dr410-z {{declared private here}}
 }
 
 // dr412 is in dr412.cpp
@@ -302,7 +302,7 @@ namespace dr413 { // dr413: yes
   T t1 = { 1, {}, 2 };
   T t2 = { 1, 2 };
   // expected-error@-1 {{initializer for aggregate with no elements requires explicit braces}}
-  // expected-note@#dr413-T {{'dr413::T' declared here}}
+  //   expected-note@#dr413-T {{'dr413::T' declared here}}
 }
 
 namespace dr414 { // dr414: dup 305
@@ -385,7 +385,7 @@ void g2() {
   using B::f2;
   f2();
   // expected-error@-1 {{no matching function for call to 'f2'}}
-  // expected-note@#dr418-f2 {{candidate function not viable: requires 1 argument, but 0 were provided}}
+  //   expected-note@#dr418-f2 {{candidate function not viable: requires 1 argument, but 0 were provided}}
 }
 } // namespace example2
 
@@ -473,8 +473,8 @@ namespace dr422 { // dr422: yes
     typedef T type; // #dr422-typedef-T
     typedef U type;
     // expected-error@-1 {{typedef redefinition with different types ('char' vs 'int')}}
-    // expected-note@#dr422-f-int-char {{in instantiation of function template specialization 'dr422::f' requested here}}
-    // expected-note@#dr422-typedef-T {{previous definition is here}}
+    //   expected-note@#dr422-f-int-char {{in instantiation of function template specialization 'dr422::f' requested here}}
+    //   expected-note@#dr422-typedef-T {{previous definition is here}}
   }
   template void f();
   template void f(); // #dr422-f-int-char
@@ -490,7 +490,7 @@ namespace dr424 { // dr424: yes
     typedef int N; // #dr424-N
     typedef int N;
     // expected-error@-1 {{redefinition of 'N'}}
-    // expected-note@#dr424-N {{previous definition is here}}
+    //   expected-note@#dr424-N {{previous definition is here}}
 
     struct X;
     typedef X X; // #dr424-X
@@ -502,7 +502,7 @@ namespace dr424 { // dr424: yes
 
     typedef X X;
     // expected-error@-1 {{redefinition of 'X'}}
-    // expected-note@#dr424-X {{previous definition is here}}
+    //   expected-note@#dr424-X {{previous definition is here}}
   };
   struct B {
     typedef int M;
@@ -511,7 +511,7 @@ namespace dr424 { // dr424: yes
     typedef int M; // #dr424-M
     typedef int M;
     // expected-error@-1 {{redefinition of 'M'}}
-    // expected-note@#dr424-M {{previous definition is here}}
+    //   expected-note@#dr424-M {{previous definition is here}}
   };
 }
 
@@ -519,7 +519,7 @@ namespace dr425 { // dr425: yes
   struct A { template operator T() const; } a;
   float f = 1.0f * a;
   // expected-error@-1 {{use of overloaded operator '*' is ambiguous (with operand types 'float' and 'struct A')}}
-  // expected-note@-2 +{{built-in candidate}}
+  //   expected-note@-2 +{{built-in candidate}}
 
   template struct is_float;
   template<> struct is_float { typedef void type; };
@@ -544,7 +544,7 @@ namespace dr427 { // dr427: yes
   const D &d3 = (const D&)b;
   const D &d4(b);
   // expected-error@-1 {{conversion function from 'B' to 'const D' invokes a deleted function}}
-  // expected-note@#dr427-D {{'D' has been explicitly marked deleted here}}
+  //   expected-note@#dr427-D {{'D' has been explicitly marked deleted here}}
 }
 
 namespace dr428 { // dr428: yes
@@ -557,19 +557,19 @@ namespace dr428 { // dr428: yes
     throw make();
     throw x;
     // expected-error@-1 {{cannot throw}}
-    // expected-note@#dr428-X {{forward declaration of 'dr428::X'}}
+    //   expected-note@#dr428-X {{forward declaration of 'dr428::X'}}
     throw make();
     // expected-error@-1 {{cannot throw}}
-    // expected-note@#dr428-X {{forward declaration of 'dr428::X'}}
+    //   expected-note@#dr428-X {{forward declaration of 'dr428::X'}}
     throw make();
     // expected-error@-1 {{cannot throw}}
-    // expected-note@#dr428-X {{forward declaration of 'dr428::X'}}
+    //   expected-note@#dr428-X {{forward declaration of 'dr428::X'}}
     throw make();
     // expected-error@-1 {{cannot throw}}
-    // expected-note@#dr428-X {{forward declaration of 'dr428::X'}}
+    //   expected-note@#dr428-X {{forward declaration of 'dr428::X'}}
     throw make();
     // expected-error@-1 {{cannot throw}}
-    // expected-note@#dr428-X {{forward declaration of 'dr428::X'}}
+    //   expected-note@#dr428-X {{forward declaration of 'dr428::X'}}
   }
 }
 
@@ -580,7 +580,7 @@ namespace dr429 { // dr429: 2.8 c++11
     static void operator delete(void*, size_t); // #dr429-delete
   } *a = new (0) A;
   // since-cxx11-error@-1 {{'new' expression with placement arguments refers to non-placement 'operator delete'}}
-  // since-cxx11-note@#dr429-delete {{here}}
+  //   since-cxx11-note@#dr429-delete {{here}}
   struct B {
     static void *operator new(size_t, size_t);
     static void operator delete(void*);
@@ -633,12 +633,12 @@ namespace dr432 { // dr432: 3.0
   template struct A {};
   template struct B : A {};
   // expected-error@-1 {{use of class template 'B' requires template arguments}}
-  // expected-note@-2 {{template is declared here}}
+  //   expected-note@-2 {{template is declared here}}
   template struct C : A > {};
 #if __cplusplus >= 201103L
   template struct D : decltype(A()) {};
   // since-cxx11-error@-1 {{use of class template 'D' requires template arguments}}
-  // since-cxx11-note@-2 {{template is declared here}}
+  //   since-cxx11-note@-2 {{template is declared here}}
 #endif
 }
 
@@ -676,7 +676,7 @@ namespace dr436 { // dr436: yes
   enum E { f }; // #dr436-f
   void f();
   // expected-error@-1 {{redefinition of 'f' as different kind of symbol}}
-  // expected-note@#dr436-f {{previous definition is here}}
+  //   expected-note@#dr436-f {{previous definition is here}}
 }
 
 namespace dr437 { // dr437: sup 1308
@@ -686,14 +686,14 @@ namespace dr437 { // dr437: sup 1308
   struct S {
     void f() throw(S);
     // since-cxx17-error@-1 {{ISO C++17 does not allow dynamic exception specifications}}
-    // since-cxx17-note@-2 {{use 'noexcept(false)' instead}}
+    //   since-cxx17-note@-2 {{use 'noexcept(false)' instead}}
     void g() throw(T);
     // since-cxx17-error@-1 {{ISO C++17 does not allow dynamic exception specifications}}
-    // since-cxx17-note@-2 {{use 'noexcept(false)' instead}}
+    //   since-cxx17-note@-2 {{use 'noexcept(false)' instead}}
     struct U;
     void h() throw(U);
     // since-cxx17-error@-1 {{ISO C++17 does not allow dynamic exception specifications}}
-    // since-cxx17-note@-2 {{use 'noexcept(false)' instead}}
+    //   since-cxx17-note@-2 {{use 'noexcept(false)' instead}}
     struct U {};
   };
 }
@@ -716,11 +716,11 @@ namespace dr444 { // dr444: yes
   void f() {
     d = d;
     // expected-error@-1 {{overload resolution selected deleted operator '='}}
-    // expected-note@#dr444-deleted {{candidate function has been explicitly deleted}}
-    // expected-note@#dr444-D {{candidate function (the implicit copy assignment operator)}}
-    // expected-note@#dr444-B {{candidate function (the implicit copy assignment operator)}}
-    // since-cxx11-note@#dr444-B {{candidate function (the implicit move assignment operator) not viable: expects an rvalue for 1st argument}}
-    // since-cxx11-note@#dr444-D {{candidate function (the implicit move assignment operator) not viable: expects an rvalue for 1st argument}}
+    //   expected-note@#dr444-deleted {{candidate function has been explicitly deleted}}
+    //   expected-note@#dr444-D {{candidate function (the implicit copy assignment operator)}}
+    //   expected-note@#dr444-B {{candidate function (the implicit copy assignment operator)}}
+    //   since-cxx11-note@#dr444-B {{candidate function (the implicit move assignment operator) not viable: expects an rvalue for 1st argument}}
+    //   since-cxx11-note@#dr444-D {{candidate function (the implicit move assignment operator) not viable: expects an rvalue for 1st argument}}
   }
 }
 
@@ -729,7 +729,7 @@ namespace dr445 { // dr445: 3.2
   struct B {
     friend void A::f();
     // expected-error@-1 {{friend function 'f' is a private member of 'dr445::A'}}
-    // expected-note@#dr445-f {{implicitly declared private here}}
+    //   expected-note@#dr445-f {{implicitly declared private here}}
   };
 }
 
@@ -746,24 +746,24 @@ namespace dr446 { // dr446: 2.8
     void(b ? a : a);
     b ? A() : a;
     // expected-error@-1 {{call to deleted constructor of 'A'}}
-    // expected-note@#dr446-deleted {{'A' has been explicitly marked deleted here}}
+    //   expected-note@#dr446-deleted {{'A' has been explicitly marked deleted here}}
     b ? a : A();
     // expected-error@-1 {{call to deleted constructor of 'A'}}
-    // expected-note@#dr446-deleted {{'A' has been explicitly marked deleted here}}
+    //   expected-note@#dr446-deleted {{'A' has been explicitly marked deleted here}}
     b ? A() : A();
     // cxx98-14-error@-1 {{call to deleted constructor of 'A'}}
-    // expected-note@#dr446-deleted {{'A' has been explicitly marked deleted here}}
+    //   expected-note@#dr446-deleted {{'A' has been explicitly marked deleted here}}
 
     void(b ? a : c);
     b ? a : C();
     // expected-error@-1 {{call to deleted constructor of 'A'}}
-    // cxx98-14-note@#dr446-deleted {{'A' has been explicitly marked deleted here}}
+    //   cxx98-14-note@#dr446-deleted {{'A' has been explicitly marked deleted here}}
     b ? c : A();
     // cxx98-14-error@-1 {{call to deleted constructor of 'A'}}
-    // cxx98-14-note@#dr446-deleted {{'A' has been explicitly marked deleted here}}
+    //   cxx98-14-note@#dr446-deleted {{'A' has been explicitly marked deleted here}}
     b ? A() : C();
     // cxx98-14-error@-1 {{call to deleted constructor of 'A'}}
-    // cxx98-14-note@#dr446-deleted {{'A' has been explicitly marked deleted here}}
+    //   cxx98-14-note@#dr446-deleted {{'A' has been explicitly marked deleted here}}
   }
 }
 
@@ -801,11 +801,11 @@ namespace dr448 { // dr448: 2.8
   template void g(T t) {
     f(t);
     // expected-error@-1 {{call to function 'f' that is neither visible in the template definition nor found by argument-dependent lookup}}
-    // expected-note@#dr448-g {{in instantiation of function template specialization 'dr448::g' requested here}}
-    // expected-note@#dr448-f-T {{'f' should be declared prior to the call site or in namespace 'dr448::HideFromADL'}}
+    //   expected-note@#dr448-g {{in instantiation of function template specialization 'dr448::g' requested here}}
+    //   expected-note@#dr448-f-T {{'f' should be declared prior to the call site or in namespace 'dr448::HideFromADL'}}
     dr448::f(t);
     // expected-error@-1 {{no matching function for call to 'f'}}
-    // expected-note@#dr448-f-int {{candidate function template not viable: no known conversion from 'dr448::HideFromADL::X' to 'int' for 1st argument}}
+    //   expected-note@#dr448-f-int {{candidate function template not viable: no known conversion from 'dr448::HideFromADL::X' to 'int' for 1st argument}}
   }
   template void f(T); // #dr448-f-T
   namespace HideFromADL { struct X {}; }
@@ -824,14 +824,14 @@ namespace dr450 { // dr450: yes
     f1(S().n);
     f2(S().n);
     // expected-error@-1 {{no matching function for call to 'f2'}}}
-    // expected-note@#dr450-f2 {{candidate function not viable: expects an lvalue for 1st argument}}
+    //   expected-note@#dr450-f2 {{candidate function not viable: expects an lvalue for 1st argument}}
   }
 #if __cplusplus >= 201103L
   void h() {
     f1(A{});
     f2(A{});
     // expected-error@-1 {{no matching function for call to 'f2'}}}
-    // expected-note@#dr450-f2 {{candidate function not viable: expects an lvalue for 1st argument}}
+    //   expected-note@#dr450-f2 {{candidate function not viable: expects an lvalue for 1st argument}}
   }
 #endif
 }
@@ -885,7 +885,7 @@ namespace dr457 { // dr457: yes
     ea = a,
     eb = b
     // expected-error@-1 {{expression is not an integral constant expression}}
-    // expected-note@-2 {{read of volatile-qualified type 'const volatile int' is not allowed in a constant expression}}
+    //   expected-note@-2 {{read of volatile-qualified type 'const volatile int' is not allowed in a constant expression}}
   };
 }
 
@@ -909,7 +909,7 @@ namespace dr458 { // dr458: 11
   int A::g() {
     return T;
     // expected-error@-1 {{'T' does not refer to a value}}
-    // expected-note@#dr458-g-T {{declared here}}
+    //   expected-note@#dr458-g-T {{declared here}}
   }
 
   template
@@ -926,7 +926,7 @@ namespace dr458 { // dr458: 11
   int B::h() {
     return T;
     // expected-error@-1 {{'T' does not refer to a value}}
-    // expected-note@#dr458-h-T {{declared here}}
+    //   expected-note@#dr458-h-T {{declared here}}
   }
 }
 
@@ -993,7 +993,7 @@ namespace dr467 { // dr467: yes
   int g() {
     goto later;
     // expected-error@-1 {{cannot jump from this goto statement to its label}}
-    // expected-note@#dr467-k {{jump bypasses variable initialization}}
+    //   expected-note@#dr467-k {{jump bypasses variable initialization}}
     int k = stuff(); // #dr467-k
   later:
     return k;
@@ -1017,7 +1017,7 @@ namespace dr469 { // dr469: no
   template struct X {};
   X x;
   // expected-error@-1 {{implicit instantiation of undefined template 'dr469::X'}}
-  // expected-note@#dr469-X {{template is declared here}}
+  //   expected-note@#dr469-X {{template is declared here}}
 }
 
 namespace dr470 { // dr470: yes
@@ -1030,7 +1030,7 @@ namespace dr470 { // dr470: yes
   template struct A; // #dr470-A-int
   template struct A::B;
   // expected-error@-1 {{duplicate explicit instantiation of 'B'}}
-  // expected-note@#dr470-A-int {{previous explicit instantiation is here}}
+  //   expected-note@#dr470-A-int {{previous explicit instantiation is here}}
 
   // ok, instantiating C doesn't instantiate base class members.
   template struct A;
@@ -1052,7 +1052,7 @@ namespace dr471 { // dr471: 2.8
   };
   struct H : B, G { int f() { return n; } };
   // expected-error@-1 {{'n' is a private member of 'dr471::G'}}
-  // expected-note@#dr471-G-using {{declared private here}}
+  //   expected-note@#dr471-G-using {{declared private here}}
 }
 
 namespace dr474 { // dr474: 3.4
@@ -1068,7 +1068,7 @@ namespace dr474 { // dr474: 3.4
   namespace N {
     int g();
     // expected-error@-1 {{functions that differ only in their return type cannot be overloaded}}
-    // expected-note@#dr474-g {{previous declaration is here}}
+    //   expected-note@#dr474-g {{previous declaration is here}}
   }
 }
 
@@ -1096,7 +1096,7 @@ namespace dr478 { // dr478: yes
   void f(A *a);
   void f(A a[10]);
   // expected-error@-1 {{array of abstract class type 'A'}}
-  // expected-note@#dr478-f {{unimplemented pure virtual method 'f' in 'A'}}
+  //   expected-note@#dr478-f {{unimplemented pure virtual method 'f' in 'A'}}
 }
 
 namespace dr479 { // dr479: 2.8
@@ -1120,7 +1120,7 @@ namespace dr479 { // dr479: 2.8
   void g() {
     S s;
     // expected-error@-1 {{variable of type 'S' has private destructor}}
-    // expected-note@#dr479-S-dtor {{declared private here}}
+    //   expected-note@#dr479-S-dtor {{declared private here}}
     throw s;
     // expected-error@-1 {{exception object of type 'S' has private destructor}}
     //   expected-note@#dr479-S-dtor {{declared private here}}
@@ -1183,8 +1183,8 @@ namespace dr481 { // dr481: 2.8
       typedef ::dr481::C c; // #dr481-c
       typedef C c;
       // expected-error@-1 {{typedef redefinition with different types ('int' vs '::dr481::C')}}
-      // expected-note@#dr481-E-int {{in instantiation of member function 'dr481::E::f' requested here}}
-      // expected-note@#dr481-c {{previous definition is here}}
+      //   expected-note@#dr481-E-int {{in instantiation of member function 'dr481::E::f' requested here}}
+      //   expected-note@#dr481-c {{previous definition is here}}
     }
   };
   template struct E<>; // ok
@@ -1296,7 +1296,7 @@ namespace dr484 { // dr484: yes
   typedef C CT; // #dr484-typedef-CT
   struct CT {};
   // expected-error@-1 {{definition of type 'CT' conflicts with typedef of the same name}}
-  // expected-note@#dr484-typedef-CT {{'CT' declared here}}
+  //   expected-note@#dr484-typedef-CT {{'CT' declared here}}
 
   namespace N {
     struct D;
@@ -1304,7 +1304,7 @@ namespace dr484 { // dr484: yes
   }
   struct N::DT {};
   // expected-error@-1 {{definition of type 'DT' conflicts with typedef of the same name}}
-  // expected-note@#dr484-typedef-DT {{'DT' declared here}}
+  //   expected-note@#dr484-typedef-DT {{'DT' declared here}}
 
   typedef struct {
     S();
@@ -1337,10 +1337,10 @@ namespace dr486 { // dr486: yes
     int &b = f(&n);
     f(&g);
     // expected-error@-1 {{no matching function for call to 'f'}}
-    // expected-note@#dr486-f {{candidate template ignored: substitution failure [with T = void ()]: function cannot return function type 'void ()'}}
+    //   expected-note@#dr486-f {{candidate template ignored: substitution failure [with T = void ()]: function cannot return function type 'void ()'}}
     f(&n);
     // expected-error@-1 {{no matching function for call to 'f'}}
-    // expected-note@#dr486-f {{candidate template ignored: substitution failure [with T = int[10]]: function cannot return array type 'int[10]'}}
+    //   expected-note@#dr486-f {{candidate template ignored: substitution failure [with T = int[10]]: function cannot return array type 'int[10]'}}
   }
 }
 
@@ -1399,10 +1399,10 @@ namespace dr490 { // dr490: 2.8
     // (no number yet...) which will eventually supersede this one.
     friend int X::*A::i();
     // expected-error@-1 {{return type of out-of-line definition of 'dr490::A::i' differs from that in the declaration}}
-    // expected-note@#dr490-i {{previous declaration is here}}
+    //   expected-note@#dr490-i {{previous declaration is here}}
     friend int K::*A::j();
     // expected-error@-1 {{use of undeclared identifier 'K'; did you mean 'A::K'?}}
-    // expected-note@#dr490-k {{'A::K' declared here}}
+    //   expected-note@#dr490-k {{'A::K' declared here}}
 
     // ok, lookup finds B::T, not A::T, so return type matches
     friend char A::k();
diff --git a/clang/test/CXX/drs/dr6xx.cpp b/clang/test/CXX/drs/dr6xx.cpp
index 78604d480aa1c..b35d3051ab554 100644
--- a/clang/test/CXX/drs/dr6xx.cpp
+++ b/clang/test/CXX/drs/dr6xx.cpp
@@ -779,7 +779,7 @@ namespace dr657 { // dr657: partial
 
   struct C { C(Abs) {} };
   // expected-error@-1 {{parameter type 'Abs' is an abstract class}}
-  // expected-note@#dr657-Abs {{unimplemented pure virtual method 'x' in 'Abs'}}
+  //   expected-note@#dr657-Abs {{unimplemented pure virtual method 'x' in 'Abs'}}
   struct Q { operator Abs() { __builtin_unreachable(); } } q;
   // expected-error@-1 {{return type 'Abs' is an abstract class}}
 #if __cplusplus >= 201703L
@@ -809,7 +809,7 @@ namespace dr659 { // dr659: 3.0
   struct A; // #dr659-A
   int m = alignof(A&);
   // since-cxx11-error@-1 {{invalid application of 'alignof' to an incomplete type 'A'}}
-  // since-cxx11-note@#dr659-A {{forward declaration of 'dr659::A'}}
+  //   since-cxx11-note@#dr659-A {{forward declaration of 'dr659::A'}}
 }
 #endif
 
@@ -836,7 +836,7 @@ namespace dr662 { // dr662: yes
     T &tr = t;
     T *tp = &t;
     // expected-error@-1 {{'tp' declared as a pointer to a reference of type 'int &'}}
-    // expected-note@#dr662-f-call {{in instantiation of function template specialization 'dr662::f' requested here}}
+    //   expected-note@#dr662-f-call {{in instantiation of function template specialization 'dr662::f' requested here}}
 #if __cplusplus >= 201103L
     auto *ap = &t;
 #endif
@@ -1302,12 +1302,12 @@ namespace dr692 { // dr692: 16
       e2(b2);
       f(42);
       // expected-error@-1 {{call to 'f' is ambiguous}}
-      // expected-note@#dr692-f-deleted {{candidate function [with T = int, U = int] has been explicitly deleted}}
-      // expected-note@#dr692-f {{candidate function [with U = int]}}
+      //   expected-note@#dr692-f-deleted {{candidate function [with T = int, U = int] has been explicitly deleted}}
+      //   expected-note@#dr692-f {{candidate function [with U = int]}}
       g(42);
       // expected-error@-1 {{ambiguous}}
-      // expected-note@#dr692-g {{candidate function [with T = int]}}
-      // expected-note@#dr692-g-variadic {{candidate function [with T = int, U = <>]}}
+      //   expected-note@#dr692-g {{candidate function [with T = int]}}
+      //   expected-note@#dr692-g-variadic {{candidate function [with T = int, U = <>]}}
     }
   }
 
@@ -1371,7 +1371,7 @@ namespace dr696 { // dr696: 3.1
         int arr[N]; (void)arr;
         f(&N);
         // expected-error@-1 {{reference to local variable 'N' declared in enclosing function 'dr696::g'}}
-        // expected-note@#dr696-N {{'N' declared here}}
+        //   expected-note@#dr696-N {{'N' declared here}}
       }
     };
 #if __cplusplus >= 201103L
diff --git a/clang/test/CXX/over/over.load/p2-0x.cpp b/clang/test/CXX/over/over.load/p2-0x.cpp
index 8fd9a1ce1e87a..94185963ff5c7 100644
--- a/clang/test/CXX/over/over.load/p2-0x.cpp
+++ b/clang/test/CXX/over/over.load/p2-0x.cpp
@@ -55,4 +55,31 @@ static_assert(not test&>);
 static_assert(test&&>);
 
 }
+
+namespace GH78101 {
+
+template
+concept True = true;
+
+template
+struct Template {
+    static constexpr int i = I;
+    friend constexpr auto operator+(True auto f) {
+        return i;
+    }
+};
+
+template
+struct Template {
+    static constexpr int i = I;
+    friend constexpr auto operator+(True auto f) {
+        return i;
+    }
+};
+
+Template f{};
+static_assert(+Template{} == 5);
+
+}
+
 #endif
diff --git a/clang/test/CXX/over/over.match/over.match.funcs/over.match.oper/p3-2a.cpp b/clang/test/CXX/over/over.match/over.match.funcs/over.match.oper/p3-2a.cpp
index d83a176ec07ee..d757d28860a40 100644
--- a/clang/test/CXX/over/over.match/over.match.funcs/over.match.oper/p3-2a.cpp
+++ b/clang/test/CXX/over/over.match/over.match.funcs/over.match.oper/p3-2a.cpp
@@ -324,6 +324,124 @@ bool x = X() == X(); // expected-warning {{ambiguous}}
 }
 } // namespace P2468R2
 
+namespace GH53954{
+namespace friend_template_1 {
+struct P {
+  template 
+  friend bool operator==(const P&, const T&) { return true; } // expected-note {{candidate}} \
+                              // expected-note {{ambiguous candidate function with reversed arguments}}
+};
+struct A : public P {};
+struct B : public P {};
+bool check(A a, B b) { return a == b; } // expected-warning {{use of overloaded operator '==' (with operand types 'A' and 'B') to be ambiguous}}
+}
+
+namespace friend_template_2 {
+struct P {
+  template 
+  friend bool operator==(const T&, const P&) { return true; } // expected-note {{candidate}} \
+                                              // expected-note {{ambiguous candidate function with reversed arguments}}
+};
+struct A : public P {};
+struct B : public P {};
+bool check(A a, B b) { return a == b; } // expected-warning {{use of overloaded operator '==' (with operand types 'A' and 'B') to be ambiguous}}
+}
+
+namespace friend_template_class_template {
+template
+struct P {
+  template 
+  friend bool operator==(const T&, const P&) { // expected-note 2 {{candidate}}
+    return true;
+  }
+};
+struct A : public P {};
+struct B : public P {};
+bool check(A a, B b) { return a == b; } // expected-warning {{ambiguous}}
+}
+
+namespace friend_template_fixme {
+// FIXME(GH70210): This should not rewrite operator== and definitely not a hard error.
+struct P {
+  template 
+  friend bool operator==(const T &, const P &) { return true; } // expected-note 2 {{candidate}}
+  template 
+  friend bool operator!=(const T &, const P &) { return true; } // expected-note {{candidate}}
+};
+struct A : public P {};
+struct B : public P {};
+bool check(A a, B b) { return a != b; } // expected-error{{ambiguous}}
+}
+
+namespace member_template_1 {
+struct P {
+  template
+  bool operator==(const S &) const; // expected-note {{candidate}} \
+                                    // expected-note {{ambiguous candidate function with reversed arguments}}
+};
+struct A : public P {};
+struct B : public P {};
+bool check(A a, B b) { return a == b; } // expected-warning {{use of overloaded operator '==' (with operand types 'A' and 'B') to be ambiguous}}
+} // namespace member_template
+
+namespace member_template_2{
+template 
+class Foo {
+ public:
+  template 
+  bool operator==(const Foo& other) const;
+};
+bool x = Foo{} == Foo{};
+} // namespace template_member_opeqeq
+
+namespace non_member_template_1 {
+struct P {};
+template
+bool operator==(const P&, const S &); // expected-note {{candidate}} \
+                                      // expected-note {{ambiguous candidate function with reversed arguments}}
+
+struct A : public P {};
+struct B : public P {};
+bool check(A a, B b) { return a == b; } // expected-warning {{use of overloaded operator '==' (with operand types 'A' and 'B') to be ambiguous}}
+
+template bool operator!=(const P&, const S &);
+bool fine(A a, B b) { return a == b; } // Ok. Found a matching operator!=.
+} // namespace non_member_template_1
+
+namespace non_member_template_2 {
+struct P {};
+template
+bool operator==(const S&, const P&); // expected-note {{candidate}} \
+                                     // expected-note {{ambiguous candidate function with reversed arguments}}
+
+struct A : public P {};
+struct B : public P {};
+bool check(A a, B b) { return a == b; } // expected-warning {{use of overloaded operator '==' (with operand types 'A' and 'B') to be ambiguous}}
+} // namespace non_member_template_2
+
+namespace class_and_member_template {
+template 
+struct S {
+    template 
+    bool operator==(const OtherT &rhs); // expected-note {{candidate}} \
+                                        // expected-note {{reversed arguments}}
+};
+struct A : S {};
+struct B : S {};
+bool x = A{} == B{}; // expected-warning {{ambiguous}}
+} // namespace class_and_member_template
+
+namespace ambiguous_case {
+template 
+struct Foo {};
+template  bool operator==(Foo, Foo); // expected-note{{candidate}}
+template  bool operator==(Foo, Foo); // expected-note{{candidate}}
+
+void test() {
+    Foo() == Foo(); // expected-error{{ambiguous}}
+}
+} // namespace ambiguous_case
+} // namespace
 namespace ADL_GH68901{
 namespace test1 {
 namespace A {
diff --git a/clang/test/CXX/temp/temp.res/temp.local/p6.cpp b/clang/test/CXX/temp/temp.res/temp.local/p6.cpp
index e2aa0ff344291..0702966e56854 100644
--- a/clang/test/CXX/temp/temp.res/temp.local/p6.cpp
+++ b/clang/test/CXX/temp/temp.res/temp.local/p6.cpp
@@ -127,16 +127,30 @@ template struct Z { // expected-note 16{{declared here}}
 template // expected-note {{declared here}}
 void f(int T) {} // expected-error {{declaration of 'T' shadows template parameter}}
 
-// FIXME: These are ill-formed: a template-parameter shall not have the same name as the template name.
 namespace A {
   template struct T {};  // expected-error{{declaration of 'T' shadows template parameter}}
                                      // expected-note@-1{{template parameter is declared here}}
+  template struct U {
+    template struct V {}; // expected-error{{declaration of 'V' shadows template parameter}}
+                                      // expected-note@-1{{template parameter is declared here}}
+  };
 }
 namespace B {
-  template void T() {}
+  template void T() {} // expected-error{{declaration of 'T' shadows template parameter}}
+                                   // expected-note@-1{{template parameter is declared here}}
+
+  template struct U {
+    template void V(); // expected-error{{declaration of 'V' shadows template parameter}}
+                                   // expected-note@-1{{template parameter is declared here}}
+  };
 }
 namespace C {
-  template int T;
+  template int T; // expected-error{{declaration of 'T' shadows template parameter}}
+                              // expected-note@-1{{template parameter is declared here}}
+  template struct U {
+    template static int V; // expected-error{{declaration of 'V' shadows template parameter}}
+                                       // expected-note@-1{{template parameter is declared here}}
+  };
 }
 
 namespace PR28023 {
diff --git a/clang/test/CodeGen/2006-05-19-SingleEltReturn.c b/clang/test/CodeGen/2006-05-19-SingleEltReturn.c
index 16eacf3ec1625..b542242606cc3 100644
--- a/clang/test/CodeGen/2006-05-19-SingleEltReturn.c
+++ b/clang/test/CodeGen/2006-05-19-SingleEltReturn.c
@@ -24,7 +24,7 @@ struct Y bar(void) {
 
 
 // X86_32: define{{.*}} void @foo(ptr noundef %P)
-// X86_32:   call void @bar(ptr sret(%struct.Y) align 4 %{{[^),]*}})
+// X86_32:   call void @bar(ptr dead_on_unwind writable sret(%struct.Y) align 4 %{{[^),]*}})
 
-// X86_32: define{{.*}} void @bar(ptr noalias sret(%struct.Y) align 4 %{{[^,)]*}})
+// X86_32: define{{.*}} void @bar(ptr dead_on_unwind noalias writable sret(%struct.Y) align 4 %{{[^,)]*}})
 // X86_32:   ret void
diff --git a/clang/test/CodeGen/64bit-swiftcall.c b/clang/test/CodeGen/64bit-swiftcall.c
index da6f18248c2af..b1c42e3b0a657 100644
--- a/clang/test/CodeGen/64bit-swiftcall.c
+++ b/clang/test/CodeGen/64bit-swiftcall.c
@@ -30,7 +30,7 @@ SWIFTCALL int indirect_result_2(OUT int *arg0, OUT float *arg1) {  __builtin_unr
 
 typedef struct { char array[1024]; } struct_reallybig;
 SWIFTCALL struct_reallybig indirect_result_3(OUT int *arg0, OUT float *arg1) { __builtin_unreachable(); }
-// CHECK-LABEL: define {{.*}} void @indirect_result_3(ptr noalias sret(%struct.struct_reallybig) {{.*}}, ptr noalias align 4 dereferenceable(4){{.*}}, ptr noalias align 4 dereferenceable(4){{.*}})
+// CHECK-LABEL: define {{.*}} void @indirect_result_3(ptr dead_on_unwind noalias writable sret(%struct.struct_reallybig) {{.*}}, ptr noalias align 4 dereferenceable(4){{.*}}, ptr noalias align 4 dereferenceable(4){{.*}})
 
 SWIFTCALL void context_1(CONTEXT void *self) {}
 // CHECK-LABEL: define {{.*}} void @context_1(ptr swiftself
@@ -238,7 +238,7 @@ typedef struct {
 } struct_big_1;
 TEST(struct_big_1)
 
-// CHECK-LABEL: define {{.*}} void @return_struct_big_1({{.*}} noalias sret
+// CHECK-LABEL: define {{.*}} void @return_struct_big_1(ptr dead_on_unwind noalias writable sret
 
 // Should not be byval.
 // CHECK-LABEL: define {{.*}} void @take_struct_big_1(ptr{{( %.*)?}})
@@ -522,7 +522,7 @@ typedef struct {
   double d4;
 } struct_d5;
 TEST(struct_d5)
-// CHECK: define{{.*}} swiftcc void @return_struct_d5(ptr noalias sret([[STRUCT5:.+]])
+// CHECK: define{{.*}} swiftcc void @return_struct_d5(ptr dead_on_unwind noalias writable sret([[STRUCT5:.+]])
 // CHECK: define{{.*}} swiftcc void @take_struct_d5(ptr
 
 typedef struct {
@@ -709,7 +709,7 @@ typedef struct {
   long long l4;
 } struct_l5;
 TEST(struct_l5)
-// CHECK: define{{.*}} swiftcc void @return_struct_l5(ptr noalias sret([[STRUCT5:.+]])
+// CHECK: define{{.*}} swiftcc void @return_struct_l5(ptr dead_on_unwind noalias writable sret([[STRUCT5:.+]])
 // CHECK: define{{.*}} swiftcc void @take_struct_l5(ptr
 
 typedef struct {
@@ -754,7 +754,7 @@ typedef struct {
   char16 c4;
 } struct_vc5;
 TEST(struct_vc5)
-// CHECK: define{{.*}} swiftcc void @return_struct_vc5(ptr noalias sret([[STRUCT:.+]])
+// CHECK: define{{.*}} swiftcc void @return_struct_vc5(ptr dead_on_unwind noalias writable sret([[STRUCT:.+]])
 // CHECK: define{{.*}} swiftcc void @take_struct_vc5(ptr
 
 typedef struct {
@@ -799,7 +799,7 @@ typedef struct {
   short8 c4;
 } struct_vs5;
 TEST(struct_vs5)
-// CHECK: define{{.*}} swiftcc void @return_struct_vs5(ptr noalias sret([[STRUCT:.+]])
+// CHECK: define{{.*}} swiftcc void @return_struct_vs5(ptr dead_on_unwind noalias writable sret([[STRUCT:.+]])
 // CHECK: define{{.*}} swiftcc void @take_struct_vs5(ptr
 
 typedef struct {
@@ -844,7 +844,7 @@ typedef struct {
   int4 c4;
 } struct_vi5;
 TEST(struct_vi5)
-// CHECK: define{{.*}} swiftcc void @return_struct_vi5(ptr noalias sret([[STRUCT:.+]])
+// CHECK: define{{.*}} swiftcc void @return_struct_vi5(ptr dead_on_unwind noalias writable sret([[STRUCT:.+]])
 // CHECK: define{{.*}} swiftcc void @take_struct_vi5(ptr
 
 typedef struct {
@@ -872,7 +872,7 @@ typedef struct {
   long2 c4;
 } struct_vl5;
 TEST(struct_vl5)
-// CHECK: define{{.*}} swiftcc void @return_struct_vl5(ptr noalias sret([[STRUCT:.+]])
+// CHECK: define{{.*}} swiftcc void @return_struct_vl5(ptr dead_on_unwind noalias writable sret([[STRUCT:.+]])
 // CHECK: define{{.*}} swiftcc void @take_struct_vl5(ptr
 
 typedef struct {
@@ -900,7 +900,7 @@ typedef struct {
   double2 c4;
 } struct_vd5;
 TEST(struct_vd5)
-// CHECK: define{{.*}} swiftcc void @return_struct_vd5(ptr noalias sret([[STRUCT:.+]])
+// CHECK: define{{.*}} swiftcc void @return_struct_vd5(ptr dead_on_unwind noalias writable sret([[STRUCT:.+]])
 // CHECK: define{{.*}} swiftcc void @take_struct_vd5(ptr
 
 typedef struct {
@@ -924,7 +924,7 @@ typedef struct {
   double4 c2;
 } struct_vd43;
 TEST(struct_vd43)
-// CHECK: define{{.*}} swiftcc void @return_struct_vd43(ptr noalias sret([[STRUCT:.+]])
+// CHECK: define{{.*}} swiftcc void @return_struct_vd43(ptr dead_on_unwind noalias writable sret([[STRUCT:.+]])
 // CHECK: define{{.*}} swiftcc void @take_struct_vd43(ptr
 
 typedef struct {
@@ -960,7 +960,7 @@ typedef struct {
   float4 c4;
 } struct_vf5;
 TEST(struct_vf5)
-// CHECK: define{{.*}} swiftcc void @return_struct_vf5(ptr noalias sret([[STRUCT:.+]])
+// CHECK: define{{.*}} swiftcc void @return_struct_vf5(ptr dead_on_unwind noalias writable sret([[STRUCT:.+]])
 // CHECK: define{{.*}} swiftcc void @take_struct_vf5(ptr
 
 typedef struct {
diff --git a/clang/test/CodeGen/CSKY/csky-abi.c b/clang/test/CodeGen/CSKY/csky-abi.c
index a24d4d8d64077..2e549376ba933 100644
--- a/clang/test/CodeGen/CSKY/csky-abi.c
+++ b/clang/test/CodeGen/CSKY/csky-abi.c
@@ -117,7 +117,7 @@ void f_agg_large(struct large x) {
 
 // The address where the struct should be written to will be the first
 // argument
-// CHECK-LABEL: define{{.*}} void @f_agg_large_ret(ptr noalias sret(%struct.large) align 4 %agg.result, i32 noundef %i, i8 noundef signext %j)
+// CHECK-LABEL: define{{.*}} void @f_agg_large_ret(ptr dead_on_unwind noalias writable sret(%struct.large) align 4 %agg.result, i32 noundef %i, i8 noundef signext %j)
 struct large f_agg_large_ret(int32_t i, int8_t j) {
   return (struct large){1, 2, 3, 4};
 }
@@ -144,7 +144,7 @@ int f_scalar_stack_1(struct tiny a, struct small b, struct small_aligned c,
 // the presence of large return values that consume a register due to the need
 // to pass a pointer.
 
-// CHECK-LABEL: define{{.*}} void @f_scalar_stack_2(ptr noalias sret(%struct.large) align 4 %agg.result, i32 noundef %a, i64 noundef %b, i64 noundef %c, double noundef %d, i8 noundef zeroext %e, i8 noundef signext %f, i8 noundef zeroext %g)
+// CHECK-LABEL: define{{.*}} void @f_scalar_stack_2(ptr dead_on_unwind noalias writable sret(%struct.large) align 4 %agg.result, i32 noundef %a, i64 noundef %b, i64 noundef %c, double noundef %d, i8 noundef zeroext %e, i8 noundef signext %f, i8 noundef zeroext %g)
 struct large f_scalar_stack_2(int32_t a, int64_t b, int64_t c, long double d,
                               uint8_t e, int8_t f, uint8_t g) {
   return (struct large){a, e, f, g};
diff --git a/clang/test/CodeGen/CSKY/csky-hard-abi.c b/clang/test/CodeGen/CSKY/csky-hard-abi.c
index 2171da8091e2b..0bc4a5a8808e5 100644
--- a/clang/test/CodeGen/CSKY/csky-hard-abi.c
+++ b/clang/test/CodeGen/CSKY/csky-hard-abi.c
@@ -72,7 +72,7 @@ struct double_float_s {
 // CHECK: define{{.*}} void @f_double_double_s_arg([4 x i32] %a.coerce)
 void f_double_double_s_arg(struct double_double_s a) {}
 
-// CHECK: define{{.*}} void @f_ret_double_double_s(ptr noalias sret(%struct.double_double_s) align 4 %agg.result)
+// CHECK: define{{.*}} void @f_ret_double_double_s(ptr dead_on_unwind noalias writable sret(%struct.double_double_s) align 4 %agg.result)
 struct double_double_s f_ret_double_double_s(void) {
   return (struct double_double_s){1.0, 2.0};
 }
@@ -80,7 +80,7 @@ struct double_double_s f_ret_double_double_s(void) {
 // CHECK: define{{.*}} void @f_double_float_s_arg([3 x i32] %a.coerce)
 void f_double_float_s_arg(struct double_float_s a) {}
 
-// CHECK: define{{.*}} void @f_ret_double_float_s(ptr noalias sret(%struct.double_float_s) align 4 %agg.result)
+// CHECK: define{{.*}} void @f_ret_double_float_s(ptr dead_on_unwind noalias writable sret(%struct.double_float_s) align 4 %agg.result)
 struct double_float_s f_ret_double_float_s(void) {
   return (struct double_float_s){1.0, 2.0};
 }
@@ -118,7 +118,7 @@ struct double_int8_zbf_s {
 // CHECK: define{{.*}}  @f_double_int8_s_arg([3 x i32] %a.coerce)
 void f_double_int8_s_arg(struct double_int8_s a) {}
 
-// CHECK: define{{.*}} void @f_ret_double_int8_s(ptr noalias sret(%struct.double_int8_s) align 4 %agg.result)
+// CHECK: define{{.*}} void @f_ret_double_int8_s(ptr dead_on_unwind noalias writable sret(%struct.double_int8_s) align 4 %agg.result)
 struct double_int8_s f_ret_double_int8_s(void) {
   return (struct double_int8_s){1.0, 2};
 }
@@ -126,7 +126,7 @@ struct double_int8_s f_ret_double_int8_s(void) {
 // CHECK: define{{.*}} void @f_double_uint8_s_arg([3 x i32] %a.coerce)
 void f_double_uint8_s_arg(struct double_uint8_s a) {}
 
-// CHECK: define{{.*}} void @f_ret_double_uint8_s(ptr noalias sret(%struct.double_uint8_s) align 4 %agg.result)
+// CHECK: define{{.*}} void @f_ret_double_uint8_s(ptr dead_on_unwind noalias writable sret(%struct.double_uint8_s) align 4 %agg.result)
 struct double_uint8_s f_ret_double_uint8_s(void) {
   return (struct double_uint8_s){1.0, 2};
 }
@@ -134,7 +134,7 @@ struct double_uint8_s f_ret_double_uint8_s(void) {
 // CHECK: define{{.*}} void @f_double_int32_s_arg([3 x i32] %a.coerce)
 void f_double_int32_s_arg(struct double_int32_s a) {}
 
-// CHECK: define{{.*}} void @f_ret_double_int32_s(ptr noalias sret(%struct.double_int32_s) align 4 %agg.result)
+// CHECK: define{{.*}} void @f_ret_double_int32_s(ptr dead_on_unwind noalias writable sret(%struct.double_int32_s) align 4 %agg.result)
 struct double_int32_s f_ret_double_int32_s(void) {
   return (struct double_int32_s){1.0, 2};
 }
@@ -142,7 +142,7 @@ struct double_int32_s f_ret_double_int32_s(void) {
 // CHECK: define{{.*}} void @f_double_int64_s_arg([4 x i32] %a.coerce)
 void f_double_int64_s_arg(struct double_int64_s a) {}
 
-// CHECK: define{{.*}} void @f_ret_double_int64_s(ptr noalias sret(%struct.double_int64_s) align 4 %agg.result)
+// CHECK: define{{.*}} void @f_ret_double_int64_s(ptr dead_on_unwind noalias writable sret(%struct.double_int64_s) align 4 %agg.result)
 struct double_int64_s f_ret_double_int64_s(void) {
   return (struct double_int64_s){1.0, 2};
 }
@@ -150,7 +150,7 @@ struct double_int64_s f_ret_double_int64_s(void) {
 // CHECK: define{{.*}} void @f_double_int64bf_s_arg([3 x i32] %a.coerce)
 void f_double_int64bf_s_arg(struct double_int64bf_s a) {}
 
-// CHECK: define{{.*}} void @f_ret_double_int64bf_s(ptr noalias sret(%struct.double_int64bf_s) align 4 %agg.result)
+// CHECK: define{{.*}} void @f_ret_double_int64bf_s(ptr dead_on_unwind noalias writable sret(%struct.double_int64bf_s) align 4 %agg.result)
 struct double_int64bf_s f_ret_double_int64bf_s(void) {
   return (struct double_int64bf_s){1.0, 2};
 }
@@ -158,7 +158,7 @@ struct double_int64bf_s f_ret_double_int64bf_s(void) {
 // CHECK: define{{.*}} void @f_double_int8_zbf_s([3 x i32] %a.coerce)
 void f_double_int8_zbf_s(struct double_int8_zbf_s a) {}
 
-// CHECK: define{{.*}} void @f_ret_double_int8_zbf_s(ptr noalias sret(%struct.double_int8_zbf_s) align 4 %agg.result)
+// CHECK: define{{.*}} void @f_ret_double_int8_zbf_s(ptr dead_on_unwind noalias writable sret(%struct.double_int8_zbf_s) align 4 %agg.result)
 struct double_int8_zbf_s f_ret_double_int8_zbf_s(void) {
   return (struct double_int8_zbf_s){1.0, 2};
 }
@@ -179,7 +179,7 @@ void f_struct_double_int8_insufficient_fprs(float a, double b, double c, double
 // CHECK: define{{.*}} void @f_doublecomplex(double noundef %a.coerce0, double noundef %a.coerce1)
 void f_doublecomplex(double __complex__ a) {}
 
-// CHECK: define{{.*}} void @f_ret_doublecomplex(ptr noalias sret({ double, double }) align 4 %agg.result)
+// CHECK: define{{.*}} void @f_ret_doublecomplex(ptr dead_on_unwind noalias writable sret({ double, double }) align 4 %agg.result)
 double __complex__ f_ret_doublecomplex(void) {
   return 1.0;
 }
@@ -191,7 +191,7 @@ struct doublecomplex_s {
 // CHECK: define{{.*}} void @f_doublecomplex_s_arg([4 x i32] %a.coerce)
 void f_doublecomplex_s_arg(struct doublecomplex_s a) {}
 
-// CHECK: define{{.*}} void @f_ret_doublecomplex_s(ptr noalias sret(%struct.doublecomplex_s) align 4 %agg.result)
+// CHECK: define{{.*}} void @f_ret_doublecomplex_s(ptr dead_on_unwind noalias writable sret(%struct.doublecomplex_s) align 4 %agg.result)
 struct doublecomplex_s f_ret_doublecomplex_s(void) {
   return (struct doublecomplex_s){1.0};
 }
@@ -218,7 +218,7 @@ struct doublearr2_s {
 // CHECK: define{{.*}} void @f_doublearr2_s_arg([4 x i32] %a.coerce)
 void f_doublearr2_s_arg(struct doublearr2_s a) {}
 
-// CHECK: define{{.*}} void @f_ret_doublearr2_s(ptr noalias sret(%struct.doublearr2_s) align 4 %agg.result)
+// CHECK: define{{.*}} void @f_ret_doublearr2_s(ptr dead_on_unwind noalias writable sret(%struct.doublearr2_s) align 4 %agg.result)
 struct doublearr2_s f_ret_doublearr2_s(void) {
   return (struct doublearr2_s){{1.0, 2.0}};
 }
@@ -232,7 +232,7 @@ struct doublearr2_tricky1_s {
 // CHECK: define{{.*}} void @f_doublearr2_tricky1_s_arg([4 x i32] %a.coerce)
 void f_doublearr2_tricky1_s_arg(struct doublearr2_tricky1_s a) {}
 
-// CHECK: define{{.*}} void @f_ret_doublearr2_tricky1_s(ptr noalias sret(%struct.doublearr2_tricky1_s) align 4 %agg.result)
+// CHECK: define{{.*}} void @f_ret_doublearr2_tricky1_s(ptr dead_on_unwind noalias writable sret(%struct.doublearr2_tricky1_s) align 4 %agg.result)
 struct doublearr2_tricky1_s f_ret_doublearr2_tricky1_s(void) {
   return (struct doublearr2_tricky1_s){{{{1.0}}, {{2.0}}}};
 }
@@ -247,7 +247,7 @@ struct doublearr2_tricky2_s {
 // CHECK: define{{.*}} void @f_doublearr2_tricky2_s_arg([4 x i32] %a.coerce)
 void f_doublearr2_tricky2_s_arg(struct doublearr2_tricky2_s a) {}
 
-// CHECK: define{{.*}} void @f_ret_doublearr2_tricky2_s(ptr noalias sret(%struct.doublearr2_tricky2_s) align 4 %agg.result)
+// CHECK: define{{.*}} void @f_ret_doublearr2_tricky2_s(ptr dead_on_unwind noalias writable sret(%struct.doublearr2_tricky2_s) align 4 %agg.result)
 struct doublearr2_tricky2_s f_ret_doublearr2_tricky2_s(void) {
   return (struct doublearr2_tricky2_s){{}, {{{1.0}}, {{2.0}}}};
 }
@@ -262,7 +262,7 @@ struct doublearr2_tricky3_s {
 // CHECK: define{{.*}} void @f_doublearr2_tricky3_s_arg([4 x i32] %a.coerce)
 void f_doublearr2_tricky3_s_arg(struct doublearr2_tricky3_s a) {}
 
-// CHECK: define{{.*}} void @f_ret_doublearr2_tricky3_s(ptr noalias sret(%struct.doublearr2_tricky3_s) align 4 %agg.result)
+// CHECK: define{{.*}} void @f_ret_doublearr2_tricky3_s(ptr dead_on_unwind noalias writable sret(%struct.doublearr2_tricky3_s) align 4 %agg.result)
 struct doublearr2_tricky3_s f_ret_doublearr2_tricky3_s(void) {
   return (struct doublearr2_tricky3_s){{}, {{{1.0}}, {{2.0}}}};
 }
@@ -278,7 +278,7 @@ struct doublearr2_tricky4_s {
 // CHECK: define{{.*}} void @f_doublearr2_tricky4_s_arg([4 x i32] %a.coerce)
 void f_doublearr2_tricky4_s_arg(struct doublearr2_tricky4_s a) {}
 
-// CHECK: define{{.*}} void @f_ret_doublearr2_tricky4_s(ptr noalias sret(%struct.doublearr2_tricky4_s) align 4 %agg.result)
+// CHECK: define{{.*}} void @f_ret_doublearr2_tricky4_s(ptr dead_on_unwind noalias writable sret(%struct.doublearr2_tricky4_s) align 4 %agg.result)
 struct doublearr2_tricky4_s f_ret_doublearr2_tricky4_s(void) {
   return (struct doublearr2_tricky4_s){{}, {{{}, {1.0}}, {{}, {2.0}}}};
 }
@@ -292,7 +292,7 @@ struct int_double_int_s {
 // CHECK: define{{.*}} void @f_int_double_int_s_arg([4 x i32] %a.coerce)
 void f_int_double_int_s_arg(struct int_double_int_s a) {}
 
-// CHECK: define{{.*}} void @f_ret_int_double_int_s(ptr noalias sret(%struct.int_double_int_s) align 4 %agg.result)
+// CHECK: define{{.*}} void @f_ret_int_double_int_s(ptr dead_on_unwind noalias writable sret(%struct.int_double_int_s) align 4 %agg.result)
 struct int_double_int_s f_ret_int_double_int_s(void) {
   return (struct int_double_int_s){1, 2.0, 3};
 }
@@ -305,7 +305,7 @@ struct int64_double_s {
 // CHECK: define{{.*}} void @f_int64_double_s_arg([4 x i32] %a.coerce)
 void f_int64_double_s_arg(struct int64_double_s a) {}
 
-// CHECK: define{{.*}} void @f_ret_int64_double_s(ptr noalias sret(%struct.int64_double_s) align 4 %agg.result)
+// CHECK: define{{.*}} void @f_ret_int64_double_s(ptr dead_on_unwind noalias writable sret(%struct.int64_double_s) align 4 %agg.result)
 struct int64_double_s f_ret_int64_double_s(void) {
   return (struct int64_double_s){1, 2.0};
 }
@@ -319,7 +319,7 @@ struct char_char_double_s {
 // CHECK-LABEL: define{{.*}} void @f_char_char_double_s_arg([3 x i32] %a.coerce)
 void f_char_char_double_s_arg(struct char_char_double_s a) {}
 
-// CHECK: define{{.*}} void @f_ret_char_char_double_s(ptr noalias sret(%struct.char_char_double_s) align 4 %agg.result)
+// CHECK: define{{.*}} void @f_ret_char_char_double_s(ptr dead_on_unwind noalias writable sret(%struct.char_char_double_s) align 4 %agg.result)
 struct char_char_double_s f_ret_char_char_double_s(void) {
   return (struct char_char_double_s){1, 2, 3.0};
 }
@@ -338,19 +338,19 @@ union double_u f_ret_double_u(void) {
   return (union double_u){1.0};
 }
 
-// CHECK: define{{.*}} void @f_ret_double_int32_s_double_int32_s_just_sufficient_gprs(ptr noalias sret(%struct.double_int32_s) align 4 %agg.result, i32 noundef %a, i32 noundef %b, i32 noundef %c, i32 noundef %d, i32 noundef %e, i32 noundef %f, i32 noundef %g, [3 x i32] %h.coerce)
+// CHECK: define{{.*}} void @f_ret_double_int32_s_double_int32_s_just_sufficient_gprs(ptr dead_on_unwind noalias writable sret(%struct.double_int32_s) align 4 %agg.result, i32 noundef %a, i32 noundef %b, i32 noundef %c, i32 noundef %d, i32 noundef %e, i32 noundef %f, i32 noundef %g, [3 x i32] %h.coerce)
 struct double_int32_s f_ret_double_int32_s_double_int32_s_just_sufficient_gprs(
     int a, int b, int c, int d, int e, int f, int g, struct double_int32_s h) {
   return (struct double_int32_s){1.0, 2};
 }
 
-// CHECK: define{{.*}} void @f_ret_double_double_s_double_int32_s_just_sufficient_gprs(ptr noalias sret(%struct.double_double_s) align 4 %agg.result, i32 noundef %a, i32 noundef %b, i32 noundef %c, i32 noundef %d, i32 noundef %e, i32 noundef %f, i32 noundef %g, [3 x i32] %h.coerce)
+// CHECK: define{{.*}} void @f_ret_double_double_s_double_int32_s_just_sufficient_gprs(ptr dead_on_unwind noalias writable sret(%struct.double_double_s) align 4 %agg.result, i32 noundef %a, i32 noundef %b, i32 noundef %c, i32 noundef %d, i32 noundef %e, i32 noundef %f, i32 noundef %g, [3 x i32] %h.coerce)
 struct double_double_s f_ret_double_double_s_double_int32_s_just_sufficient_gprs(
     int a, int b, int c, int d, int e, int f, int g, struct double_int32_s h) {
   return (struct double_double_s){1.0, 2.0};
 }
 
-// CHECK: define{{.*}} void @f_ret_doublecomplex_double_int32_s_just_sufficient_gprs(ptr noalias sret({ double, double }) align 4 %agg.result, i32 noundef %a, i32 noundef %b, i32 noundef %c, i32 noundef %d, i32 noundef %e, i32 noundef %f, i32 noundef %g, [3 x i32] %h.coerce)
+// CHECK: define{{.*}} void @f_ret_doublecomplex_double_int32_s_just_sufficient_gprs(ptr dead_on_unwind noalias writable sret({ double, double }) align 4 %agg.result, i32 noundef %a, i32 noundef %b, i32 noundef %c, i32 noundef %d, i32 noundef %e, i32 noundef %f, i32 noundef %g, [3 x i32] %h.coerce)
 double __complex__ f_ret_doublecomplex_double_int32_s_just_sufficient_gprs(
     int a, int b, int c, int d, int e, int f, int g, struct double_int32_s h) {
   return 1.0;
@@ -376,7 +376,7 @@ struct large {
 // the presence of large return values that consume a register due to the need
 // to pass a pointer.
 
-// CHECK-LABEL: define{{.*}} void @f_scalar_stack_2(ptr noalias sret(%struct.large) align 4 %agg.result, float noundef %a, i64 noundef %b, double noundef %c, double noundef %d, i8 noundef zeroext %e, i8 noundef signext %f, i8 noundef zeroext %g)
+// CHECK-LABEL: define{{.*}} void @f_scalar_stack_2(ptr dead_on_unwind noalias writable sret(%struct.large) align 4 %agg.result, float noundef %a, i64 noundef %b, double noundef %c, double noundef %d, i8 noundef zeroext %e, i8 noundef signext %f, i8 noundef zeroext %g)
 struct large f_scalar_stack_2(float a, int64_t b, double c, long double d,
                               uint8_t e, int8_t f, uint8_t g) {
   return (struct large){a, e, f, g};
diff --git a/clang/test/CodeGen/CSKY/csky-soft-abi.c b/clang/test/CodeGen/CSKY/csky-soft-abi.c
index 1aba2df1f20a2..04fb7a4940842 100644
--- a/clang/test/CodeGen/CSKY/csky-soft-abi.c
+++ b/clang/test/CodeGen/CSKY/csky-soft-abi.c
@@ -72,7 +72,7 @@ struct double_float_s {
 // CHECK: define{{.*}} void @f_double_double_s_arg([4 x i32] %a.coerce)
 void f_double_double_s_arg(struct double_double_s a) {}
 
-// CHECK: define{{.*}} void @f_ret_double_double_s(ptr noalias sret(%struct.double_double_s) align 4 %agg.result)
+// CHECK: define{{.*}} void @f_ret_double_double_s(ptr dead_on_unwind noalias writable sret(%struct.double_double_s) align 4 %agg.result)
 struct double_double_s f_ret_double_double_s(void) {
   return (struct double_double_s){1.0, 2.0};
 }
@@ -80,7 +80,7 @@ struct double_double_s f_ret_double_double_s(void) {
 // CHECK: define{{.*}} void @f_double_float_s_arg([3 x i32] %a.coerce)
 void f_double_float_s_arg(struct double_float_s a) {}
 
-// CHECK: define{{.*}} void @f_ret_double_float_s(ptr noalias sret(%struct.double_float_s) align 4 %agg.result)
+// CHECK: define{{.*}} void @f_ret_double_float_s(ptr dead_on_unwind noalias writable sret(%struct.double_float_s) align 4 %agg.result)
 struct double_float_s f_ret_double_float_s(void) {
   return (struct double_float_s){1.0, 2.0};
 }
@@ -118,7 +118,7 @@ struct double_int8_zbf_s {
 // CHECK: define{{.*}}  @f_double_int8_s_arg([3 x i32] %a.coerce)
 void f_double_int8_s_arg(struct double_int8_s a) {}
 
-// CHECK: define{{.*}} void @f_ret_double_int8_s(ptr noalias sret(%struct.double_int8_s) align 4 %agg.result)
+// CHECK: define{{.*}} void @f_ret_double_int8_s(ptr dead_on_unwind noalias writable sret(%struct.double_int8_s) align 4 %agg.result)
 struct double_int8_s f_ret_double_int8_s(void) {
   return (struct double_int8_s){1.0, 2};
 }
@@ -126,7 +126,7 @@ struct double_int8_s f_ret_double_int8_s(void) {
 // CHECK: define{{.*}} void @f_double_uint8_s_arg([3 x i32] %a.coerce)
 void f_double_uint8_s_arg(struct double_uint8_s a) {}
 
-// CHECK: define{{.*}} void @f_ret_double_uint8_s(ptr noalias sret(%struct.double_uint8_s) align 4 %agg.result)
+// CHECK: define{{.*}} void @f_ret_double_uint8_s(ptr dead_on_unwind noalias writable sret(%struct.double_uint8_s) align 4 %agg.result)
 struct double_uint8_s f_ret_double_uint8_s(void) {
   return (struct double_uint8_s){1.0, 2};
 }
@@ -134,7 +134,7 @@ struct double_uint8_s f_ret_double_uint8_s(void) {
 // CHECK: define{{.*}} void @f_double_int32_s_arg([3 x i32] %a.coerce)
 void f_double_int32_s_arg(struct double_int32_s a) {}
 
-// CHECK: define{{.*}} void @f_ret_double_int32_s(ptr noalias sret(%struct.double_int32_s) align 4 %agg.result)
+// CHECK: define{{.*}} void @f_ret_double_int32_s(ptr dead_on_unwind noalias writable sret(%struct.double_int32_s) align 4 %agg.result)
 struct double_int32_s f_ret_double_int32_s(void) {
   return (struct double_int32_s){1.0, 2};
 }
@@ -142,7 +142,7 @@ struct double_int32_s f_ret_double_int32_s(void) {
 // CHECK: define{{.*}} void @f_double_int64_s_arg([4 x i32] %a.coerce)
 void f_double_int64_s_arg(struct double_int64_s a) {}
 
-// CHECK: define{{.*}} void @f_ret_double_int64_s(ptr noalias sret(%struct.double_int64_s) align 4 %agg.result)
+// CHECK: define{{.*}} void @f_ret_double_int64_s(ptr dead_on_unwind noalias writable sret(%struct.double_int64_s) align 4 %agg.result)
 struct double_int64_s f_ret_double_int64_s(void) {
   return (struct double_int64_s){1.0, 2};
 }
@@ -150,7 +150,7 @@ struct double_int64_s f_ret_double_int64_s(void) {
 // CHECK: define{{.*}} void @f_double_int64bf_s_arg([3 x i32] %a.coerce)
 void f_double_int64bf_s_arg(struct double_int64bf_s a) {}
 
-// CHECK: define{{.*}} void @f_ret_double_int64bf_s(ptr noalias sret(%struct.double_int64bf_s) align 4 %agg.result)
+// CHECK: define{{.*}} void @f_ret_double_int64bf_s(ptr dead_on_unwind noalias writable sret(%struct.double_int64bf_s) align 4 %agg.result)
 struct double_int64bf_s f_ret_double_int64bf_s(void) {
   return (struct double_int64bf_s){1.0, 2};
 }
@@ -158,7 +158,7 @@ struct double_int64bf_s f_ret_double_int64bf_s(void) {
 // CHECK: define{{.*}} void @f_double_int8_zbf_s([3 x i32] %a.coerce)
 void f_double_int8_zbf_s(struct double_int8_zbf_s a) {}
 
-// CHECK: define{{.*}} void @f_ret_double_int8_zbf_s(ptr noalias sret(%struct.double_int8_zbf_s) align 4 %agg.result)
+// CHECK: define{{.*}} void @f_ret_double_int8_zbf_s(ptr dead_on_unwind noalias writable sret(%struct.double_int8_zbf_s) align 4 %agg.result)
 struct double_int8_zbf_s f_ret_double_int8_zbf_s(void) {
   return (struct double_int8_zbf_s){1.0, 2};
 }
@@ -180,7 +180,7 @@ void f_struct_double_int8_insufficient_fprs(float a, double b, double c, double
 // CHECK: define{{.*}} void @f_doublecomplex([4 x i32] noundef %a.coerce)
 void f_doublecomplex(double __complex__ a) {}
 
-// CHECK: define{{.*}} void @f_ret_doublecomplex(ptr noalias sret({ double, double }) align 4 %agg.result)
+// CHECK: define{{.*}} void @f_ret_doublecomplex(ptr dead_on_unwind noalias writable sret({ double, double }) align 4 %agg.result)
 double __complex__ f_ret_doublecomplex(void) {
   return 1.0;
 }
@@ -192,7 +192,7 @@ struct doublecomplex_s {
 // CHECK: define{{.*}} void @f_doublecomplex_s_arg([4 x i32] %a.coerce)
 void f_doublecomplex_s_arg(struct doublecomplex_s a) {}
 
-// CHECK: define{{.*}} void @f_ret_doublecomplex_s(ptr noalias sret(%struct.doublecomplex_s) align 4 %agg.result)
+// CHECK: define{{.*}} void @f_ret_doublecomplex_s(ptr dead_on_unwind noalias writable sret(%struct.doublecomplex_s) align 4 %agg.result)
 struct doublecomplex_s f_ret_doublecomplex_s(void) {
   return (struct doublecomplex_s){1.0};
 }
@@ -219,7 +219,7 @@ struct doublearr2_s {
 // CHECK: define{{.*}} void @f_doublearr2_s_arg([4 x i32] %a.coerce)
 void f_doublearr2_s_arg(struct doublearr2_s a) {}
 
-// CHECK: define{{.*}} void @f_ret_doublearr2_s(ptr noalias sret(%struct.doublearr2_s) align 4 %agg.result)
+// CHECK: define{{.*}} void @f_ret_doublearr2_s(ptr dead_on_unwind noalias writable sret(%struct.doublearr2_s) align 4 %agg.result)
 struct doublearr2_s f_ret_doublearr2_s(void) {
   return (struct doublearr2_s){{1.0, 2.0}};
 }
@@ -233,7 +233,7 @@ struct doublearr2_tricky1_s {
 // CHECK: define{{.*}} void @f_doublearr2_tricky1_s_arg([4 x i32] %a.coerce)
 void f_doublearr2_tricky1_s_arg(struct doublearr2_tricky1_s a) {}
 
-// CHECK: define{{.*}} void @f_ret_doublearr2_tricky1_s(ptr noalias sret(%struct.doublearr2_tricky1_s) align 4 %agg.result)
+// CHECK: define{{.*}} void @f_ret_doublearr2_tricky1_s(ptr dead_on_unwind noalias writable sret(%struct.doublearr2_tricky1_s) align 4 %agg.result)
 struct doublearr2_tricky1_s f_ret_doublearr2_tricky1_s(void) {
   return (struct doublearr2_tricky1_s){{{{1.0}}, {{2.0}}}};
 }
@@ -248,7 +248,7 @@ struct doublearr2_tricky2_s {
 // CHECK: define{{.*}} void @f_doublearr2_tricky2_s_arg([4 x i32] %a.coerce)
 void f_doublearr2_tricky2_s_arg(struct doublearr2_tricky2_s a) {}
 
-// CHECK: define{{.*}} void @f_ret_doublearr2_tricky2_s(ptr noalias sret(%struct.doublearr2_tricky2_s) align 4 %agg.result)
+// CHECK: define{{.*}} void @f_ret_doublearr2_tricky2_s(ptr dead_on_unwind noalias writable sret(%struct.doublearr2_tricky2_s) align 4 %agg.result)
 struct doublearr2_tricky2_s f_ret_doublearr2_tricky2_s(void) {
   return (struct doublearr2_tricky2_s){{}, {{{1.0}}, {{2.0}}}};
 }
@@ -263,7 +263,7 @@ struct doublearr2_tricky3_s {
 // CHECK: define{{.*}} void @f_doublearr2_tricky3_s_arg([4 x i32] %a.coerce)
 void f_doublearr2_tricky3_s_arg(struct doublearr2_tricky3_s a) {}
 
-// CHECK: define{{.*}} void @f_ret_doublearr2_tricky3_s(ptr noalias sret(%struct.doublearr2_tricky3_s) align 4 %agg.result)
+// CHECK: define{{.*}} void @f_ret_doublearr2_tricky3_s(ptr dead_on_unwind noalias writable sret(%struct.doublearr2_tricky3_s) align 4 %agg.result)
 struct doublearr2_tricky3_s f_ret_doublearr2_tricky3_s(void) {
   return (struct doublearr2_tricky3_s){{}, {{{1.0}}, {{2.0}}}};
 }
@@ -279,7 +279,7 @@ struct doublearr2_tricky4_s {
 // CHECK: define{{.*}} void @f_doublearr2_tricky4_s_arg([4 x i32] %a.coerce)
 void f_doublearr2_tricky4_s_arg(struct doublearr2_tricky4_s a) {}
 
-// CHECK: define{{.*}} void @f_ret_doublearr2_tricky4_s(ptr noalias sret(%struct.doublearr2_tricky4_s) align 4 %agg.result)
+// CHECK: define{{.*}} void @f_ret_doublearr2_tricky4_s(ptr dead_on_unwind noalias writable sret(%struct.doublearr2_tricky4_s) align 4 %agg.result)
 struct doublearr2_tricky4_s f_ret_doublearr2_tricky4_s(void) {
   return (struct doublearr2_tricky4_s){{}, {{{}, {1.0}}, {{}, {2.0}}}};
 }
@@ -293,7 +293,7 @@ struct int_double_int_s {
 // CHECK: define{{.*}} void @f_int_double_int_s_arg([4 x i32] %a.coerce)
 void f_int_double_int_s_arg(struct int_double_int_s a) {}
 
-// CHECK: define{{.*}} void @f_ret_int_double_int_s(ptr noalias sret(%struct.int_double_int_s) align 4 %agg.result)
+// CHECK: define{{.*}} void @f_ret_int_double_int_s(ptr dead_on_unwind noalias writable sret(%struct.int_double_int_s) align 4 %agg.result)
 struct int_double_int_s f_ret_int_double_int_s(void) {
   return (struct int_double_int_s){1, 2.0, 3};
 }
@@ -306,7 +306,7 @@ struct int64_double_s {
 // CHECK: define{{.*}} void @f_int64_double_s_arg([4 x i32] %a.coerce)
 void f_int64_double_s_arg(struct int64_double_s a) {}
 
-// CHECK: define{{.*}} void @f_ret_int64_double_s(ptr noalias sret(%struct.int64_double_s) align 4 %agg.result)
+// CHECK: define{{.*}} void @f_ret_int64_double_s(ptr dead_on_unwind noalias writable sret(%struct.int64_double_s) align 4 %agg.result)
 struct int64_double_s f_ret_int64_double_s(void) {
   return (struct int64_double_s){1, 2.0};
 }
@@ -320,7 +320,7 @@ struct char_char_double_s {
 // CHECK-LABEL: define{{.*}} void @f_char_char_double_s_arg([3 x i32] %a.coerce)
 void f_char_char_double_s_arg(struct char_char_double_s a) {}
 
-// CHECK: define{{.*}} void @f_ret_char_char_double_s(ptr noalias sret(%struct.char_char_double_s) align 4 %agg.result)
+// CHECK: define{{.*}} void @f_ret_char_char_double_s(ptr dead_on_unwind noalias writable sret(%struct.char_char_double_s) align 4 %agg.result)
 struct char_char_double_s f_ret_char_char_double_s(void) {
   return (struct char_char_double_s){1, 2, 3.0};
 }
@@ -339,19 +339,19 @@ union double_u f_ret_double_u(void) {
   return (union double_u){1.0};
 }
 
-// CHECK: define{{.*}} void @f_ret_double_int32_s_double_int32_s_just_sufficient_gprs(ptr noalias sret(%struct.double_int32_s) align 4 %agg.result, i32 noundef %a, i32 noundef %b, i32 noundef %c, i32 noundef %d, i32 noundef %e, i32 noundef %f, i32 noundef %g, [3 x i32] %h.coerce)
+// CHECK: define{{.*}} void @f_ret_double_int32_s_double_int32_s_just_sufficient_gprs(ptr dead_on_unwind noalias writable sret(%struct.double_int32_s) align 4 %agg.result, i32 noundef %a, i32 noundef %b, i32 noundef %c, i32 noundef %d, i32 noundef %e, i32 noundef %f, i32 noundef %g, [3 x i32] %h.coerce)
 struct double_int32_s f_ret_double_int32_s_double_int32_s_just_sufficient_gprs(
     int a, int b, int c, int d, int e, int f, int g, struct double_int32_s h) {
   return (struct double_int32_s){1.0, 2};
 }
 
-// CHECK: define{{.*}} void @f_ret_double_double_s_double_int32_s_just_sufficient_gprs(ptr noalias sret(%struct.double_double_s) align 4 %agg.result, i32 noundef %a, i32 noundef %b, i32 noundef %c, i32 noundef %d, i32 noundef %e, i32 noundef %f, i32 noundef %g, [3 x i32] %h.coerce)
+// CHECK: define{{.*}} void @f_ret_double_double_s_double_int32_s_just_sufficient_gprs(ptr dead_on_unwind noalias writable sret(%struct.double_double_s) align 4 %agg.result, i32 noundef %a, i32 noundef %b, i32 noundef %c, i32 noundef %d, i32 noundef %e, i32 noundef %f, i32 noundef %g, [3 x i32] %h.coerce)
 struct double_double_s f_ret_double_double_s_double_int32_s_just_sufficient_gprs(
     int a, int b, int c, int d, int e, int f, int g, struct double_int32_s h) {
   return (struct double_double_s){1.0, 2.0};
 }
 
-// CHECK: define{{.*}} void @f_ret_doublecomplex_double_int32_s_just_sufficient_gprs(ptr noalias sret({ double, double }) align 4 %agg.result, i32 noundef %a, i32 noundef %b, i32 noundef %c, i32 noundef %d, i32 noundef %e, i32 noundef %f, i32 noundef %g, [3 x i32] %h.coerce)
+// CHECK: define{{.*}} void @f_ret_doublecomplex_double_int32_s_just_sufficient_gprs(ptr dead_on_unwind noalias writable sret({ double, double }) align 4 %agg.result, i32 noundef %a, i32 noundef %b, i32 noundef %c, i32 noundef %d, i32 noundef %e, i32 noundef %f, i32 noundef %g, [3 x i32] %h.coerce)
 double __complex__ f_ret_doublecomplex_double_int32_s_just_sufficient_gprs(
     int a, int b, int c, int d, int e, int f, int g, struct double_int32_s h) {
   return 1.0;
@@ -377,7 +377,7 @@ struct large {
 // the presence of large return values that consume a register due to the need
 // to pass a pointer.
 
-// CHECK-LABEL: define{{.*}} void @f_scalar_stack_2(ptr noalias sret(%struct.large) align 4 %agg.result, float noundef %a, i64 noundef %b, double noundef %c, double noundef %d, i8 noundef zeroext %e, i8 noundef signext %f, i8 noundef zeroext %g)
+// CHECK-LABEL: define{{.*}} void @f_scalar_stack_2(ptr dead_on_unwind noalias writable sret(%struct.large) align 4 %agg.result, float noundef %a, i64 noundef %b, double noundef %c, double noundef %d, i8 noundef zeroext %e, i8 noundef signext %f, i8 noundef zeroext %g)
 struct large f_scalar_stack_2(float a, int64_t b, double c, long double d,
                               uint8_t e, int8_t f, uint8_t g) {
   return (struct large){a, e, f, g};
diff --git a/clang/test/CodeGen/PowerPC/aix-alignment.c b/clang/test/CodeGen/PowerPC/aix-alignment.c
index e25fc7ec5599b..f732e94569ef2 100644
--- a/clang/test/CodeGen/PowerPC/aix-alignment.c
+++ b/clang/test/CodeGen/PowerPC/aix-alignment.c
@@ -22,8 +22,8 @@ StructDouble d1;
 // AIX: ret double %0
 double retDouble(double x) { return x; }
 
-// AIX32: define void @bar(ptr noalias sret(%struct.StructDouble) align 4 %agg.result, ptr noundef byval(%struct.StructDouble) align 4 %x)
-// AIX64: define void @bar(ptr noalias sret(%struct.StructDouble) align 4 %agg.result, ptr noundef byval(%struct.StructDouble) align 8 %x)
+// AIX32: define void @bar(ptr dead_on_unwind noalias writable sret(%struct.StructDouble) align 4 %agg.result, ptr noundef byval(%struct.StructDouble) align 4 %x)
+// AIX64: define void @bar(ptr dead_on_unwind noalias writable sret(%struct.StructDouble) align 4 %agg.result, ptr noundef byval(%struct.StructDouble) align 8 %x)
 // AIX32:   call void @llvm.memcpy.p0.p0.i32(ptr align 4 %agg.result, ptr align 4 %x, i32 16, i1 false)
 // AIX64:   call void @llvm.memcpy.p0.p0.i64(ptr align 4 %agg.result, ptr align 8 %x, i64 16, i1 false)
 StructDouble bar(StructDouble x) { return x; }
diff --git a/clang/test/CodeGen/PowerPC/builtins-ppc-xlcompat-sync.c b/clang/test/CodeGen/PowerPC/builtins-ppc-xlcompat-sync.c
index 9187bb855dac2..a5cc97161c56a 100644
--- a/clang/test/CodeGen/PowerPC/builtins-ppc-xlcompat-sync.c
+++ b/clang/test/CodeGen/PowerPC/builtins-ppc-xlcompat-sync.c
@@ -194,6 +194,18 @@ void test_dcbz() {
   __dcbz(c);
 }
 
+// CHECK-LABEL: @test_fence(
+// CHECK:         call void @llvm.ppc.fence()
+// CHECK-NEXT:    ret void
+//
+// CHECK-32-LABEL: @test_fence(
+// CHECK-32:         call void @llvm.ppc.fence()
+// CHECK-32-NEXT:    ret void
+//
+void test_fence() {
+  __fence();
+}
+
 // CHECK-LABEL: @test_builtin_ppc_popcntb(
 // CHECK:    [[TMP0:%.*]] = load i64, ptr @a, align 8
 // CHECK-NEXT:    [[POPCNTB:%.*]] = call i64 @llvm.ppc.popcntb.i64.i64(i64 [[TMP0]])
@@ -375,3 +387,15 @@ void test_builtin_ppc_dcbtst() {
 void test_builtin_ppc_dcbz() {
   __builtin_ppc_dcbz(c);
 }
+
+// CHECK-LABEL: @test_builtin_ppc_fence(
+// CHECK:         call void @llvm.ppc.fence()
+// CHECK-NEXT:    ret void
+//
+// CHECK-32-LABEL: @test_builtin_ppc_fence(
+// CHECK-32:         call void @llvm.ppc.fence()
+// CHECK-32-NEXT:    ret void
+//
+void test_builtin_ppc_fence() {
+  __builtin_ppc_fence();
+}
diff --git a/clang/test/CodeGen/PowerPC/powerpc-c99complex.c b/clang/test/CodeGen/PowerPC/powerpc-c99complex.c
index f01c7820857e8..82b5ac0fdc06c 100644
--- a/clang/test/CodeGen/PowerPC/powerpc-c99complex.c
+++ b/clang/test/CodeGen/PowerPC/powerpc-c99complex.c
@@ -9,7 +9,7 @@ _Complex float foo1(_Complex float x) {
 // CHECK-LABEL:             define{{.*}} { float, float } @foo1(float noundef %x.{{.*}}, float noundef %x.{{.*}}) #0 {
 // CHECK:                   ret { float, float }
 
-// PPC32LNX-LABEL:          define{{.*}} void @foo1(ptr noalias sret({ float, float }) align 4 %agg.result, ptr noundef byval({ float, float }) align 4 %x) #0 {
+// PPC32LNX-LABEL:          define{{.*}} void @foo1(ptr dead_on_unwind noalias writable sret({ float, float }) align 4 %agg.result, ptr noundef byval({ float, float }) align 4 %x) #0 {
 // PPC32LNX:                [[RETREAL:%.*]] = getelementptr inbounds { float, float }, ptr %agg.result, i32 0, i32 0
 // PPC32LNX-NEXT:           [[RETIMAG:%.*]] = getelementptr inbounds { float, float }, ptr %agg.result, i32 0, i32 1
 // PPC32LNX-NEXT:           store float %{{.*}}, ptr [[RETREAL]], align 4
@@ -21,7 +21,7 @@ _Complex double foo2(_Complex double x) {
 // CHECK-LABEL:             define{{.*}} { double, double } @foo2(double noundef %x.{{.*}}, double noundef %x.{{.*}}) #0 {
 // CHECK:                   ret { double, double }
 
-// PPC32LNX-LABEL:          define{{.*}} void @foo2(ptr noalias sret({ double, double }) align 8 %agg.result, ptr noundef byval({ double, double }) align 8 %x) #0 {
+// PPC32LNX-LABEL:          define{{.*}} void @foo2(ptr dead_on_unwind noalias writable sret({ double, double }) align 8 %agg.result, ptr noundef byval({ double, double }) align 8 %x) #0 {
 // PPC32LNX:                [[RETREAL:%.*]] = getelementptr inbounds { double, double }, ptr %agg.result, i32 0, i32 0
 // PPC32LNX-NEXT:           [[RETIMAG:%.*]] = getelementptr inbounds { double, double }, ptr %agg.result, i32 0, i32 1
 // PPC32LNX-NEXT:           store double %{{.*}}, ptr [[RETREAL]], align 8
@@ -36,7 +36,7 @@ _Complex long double foo3(_Complex long double x) {
 // CHECK-LDBL128-LABEL:     define{{.*}} { ppc_fp128, ppc_fp128 } @foo3(ppc_fp128 noundef %x.{{.*}}, ppc_fp128 noundef %x.{{.*}}) #0 {
 // CHECK-LDBL128:           ret { ppc_fp128, ppc_fp128 }
 
-// PPC32LNX-LABEL:          define{{.*}} void @foo3(ptr noalias sret({ ppc_fp128, ppc_fp128 }) align 16 %agg.result, ptr noundef byval({ ppc_fp128, ppc_fp128 }) align 16 %x) #0 {
+// PPC32LNX-LABEL:          define{{.*}} void @foo3(ptr dead_on_unwind noalias writable sret({ ppc_fp128, ppc_fp128 }) align 16 %agg.result, ptr noundef byval({ ppc_fp128, ppc_fp128 }) align 16 %x) #0 {
 // PPC32LNX:                [[RETREAL:%.*]] = getelementptr inbounds { ppc_fp128, ppc_fp128 }, ptr %agg.result, i32 0, i32 0
 // PPC32LNX-NEXT:           [[RETIMAG:%.*]] = getelementptr inbounds { ppc_fp128, ppc_fp128 }, ptr %agg.result, i32 0, i32 1
 // PPC32LNX-NEXT:           store ppc_fp128 %{{.*}}, ptr [[RETREAL]], align 16
diff --git a/clang/test/CodeGen/PowerPC/ppc-aggregate-abi.cpp b/clang/test/CodeGen/PowerPC/ppc-aggregate-abi.cpp
index 0c094891fec50..a7f7f3ee558c3 100644
--- a/clang/test/CodeGen/PowerPC/ppc-aggregate-abi.cpp
+++ b/clang/test/CodeGen/PowerPC/ppc-aggregate-abi.cpp
@@ -4,57 +4,57 @@
 // RUN:   -o - %s | FileCheck %s -check-prefix=CHECK-LE
 
 class agg_float_class { float a; };
-// CHECK-BE-LABEL: define{{.*}} void @_Z20pass_agg_float_class15agg_float_class(ptr noalias sret(%class.agg_float_class) align 4 %{{.*}}, float inreg %{{.*}})
+// CHECK-BE-LABEL: define{{.*}} void @_Z20pass_agg_float_class15agg_float_class(ptr dead_on_unwind noalias writable sret(%class.agg_float_class) align 4 %{{.*}}, float inreg %{{.*}})
 // CHECK-LE-LABEL: define{{.*}} [1 x float] @_Z20pass_agg_float_class15agg_float_class(float inreg %{{.*}})
 agg_float_class pass_agg_float_class(agg_float_class arg) { return arg; }
 
 class agg_double_class { double a; };
-// CHECK-BE-LABEL: define{{.*}} void @_Z21pass_agg_double_class16agg_double_class(ptr noalias sret(%class.agg_double_class) align 8 %{{.*}}, double inreg %{{.*}})
+// CHECK-BE-LABEL: define{{.*}} void @_Z21pass_agg_double_class16agg_double_class(ptr dead_on_unwind noalias writable sret(%class.agg_double_class) align 8 %{{.*}}, double inreg %{{.*}})
 // CHECK-LE-LABEL: define{{.*}} [1 x double] @_Z21pass_agg_double_class16agg_double_class(double inreg %{{.*}})
 agg_double_class pass_agg_double_class(agg_double_class arg) { return arg; }
 
 struct agg_float_cpp { float a; int : 0; };
-// CHECK-BE-LABEL: define{{.*}} void @_Z18pass_agg_float_cpp13agg_float_cpp(ptr noalias sret(%struct.agg_float_cpp) align 4 %{{.*}}, float inreg %{{.*}})
+// CHECK-BE-LABEL: define{{.*}} void @_Z18pass_agg_float_cpp13agg_float_cpp(ptr dead_on_unwind noalias writable sret(%struct.agg_float_cpp) align 4 %{{.*}}, float inreg %{{.*}})
 // CHECK-LE-LABEL: define{{.*}} [1 x float] @_Z18pass_agg_float_cpp13agg_float_cpp(float inreg %{{.*}})
 agg_float_cpp pass_agg_float_cpp(agg_float_cpp arg) { return arg; }
 
 struct empty { };
 struct agg_nofloat_empty { float a; empty dummy; };
-// CHECK-BE-LABEL: define{{.*}} void @_Z22pass_agg_nofloat_empty17agg_nofloat_empty(ptr noalias sret(%struct.agg_nofloat_empty) align 4 %{{.*}}, i64 %{{.*}})
+// CHECK-BE-LABEL: define{{.*}} void @_Z22pass_agg_nofloat_empty17agg_nofloat_empty(ptr dead_on_unwind noalias writable sret(%struct.agg_nofloat_empty) align 4 %{{.*}}, i64 %{{.*}})
 // CHECK-LE-LABEL: define{{.*}} i64 @_Z22pass_agg_nofloat_empty17agg_nofloat_empty(i64 %{{.*}})
 agg_nofloat_empty pass_agg_nofloat_empty(agg_nofloat_empty arg) { return arg; }
 
 struct agg_float_empty { float a; [[no_unique_address]] empty dummy; };
-// CHECK-BE-LABEL: define{{.*}} void @_Z20pass_agg_float_empty15agg_float_empty(ptr noalias sret(%struct.agg_float_empty) align 4 %{{.*}}, float inreg %{{.*}})
+// CHECK-BE-LABEL: define{{.*}} void @_Z20pass_agg_float_empty15agg_float_empty(ptr dead_on_unwind noalias writable sret(%struct.agg_float_empty) align 4 %{{.*}}, float inreg %{{.*}})
 // CHECK-LE-LABEL: define{{.*}} [1 x float] @_Z20pass_agg_float_empty15agg_float_empty(float inreg %{{.*}})
 agg_float_empty pass_agg_float_empty(agg_float_empty arg) { return arg; }
 
 struct agg_nofloat_emptyarray { float a; [[no_unique_address]] empty dummy[3]; };
-// CHECK-BE-LABEL: define{{.*}} void @_Z27pass_agg_nofloat_emptyarray22agg_nofloat_emptyarray(ptr noalias sret(%struct.agg_nofloat_emptyarray) align 4 %{{.*}}, i64 %{{.*}})
+// CHECK-BE-LABEL: define{{.*}} void @_Z27pass_agg_nofloat_emptyarray22agg_nofloat_emptyarray(ptr dead_on_unwind noalias writable sret(%struct.agg_nofloat_emptyarray) align 4 %{{.*}}, i64 %{{.*}})
 // CHECK-LE-LABEL: define{{.*}} i64 @_Z27pass_agg_nofloat_emptyarray22agg_nofloat_emptyarray(i64 %{{.*}})
 agg_nofloat_emptyarray pass_agg_nofloat_emptyarray(agg_nofloat_emptyarray arg) { return arg; }
 
 struct noemptybase { empty dummy; };
 struct agg_nofloat_emptybase : noemptybase { float a; };
-// CHECK-BE-LABEL: define{{.*}} void @_Z26pass_agg_nofloat_emptybase21agg_nofloat_emptybase(ptr noalias sret(%struct.agg_nofloat_emptybase) align 4 %{{.*}}, i64 %{{.*}})
+// CHECK-BE-LABEL: define{{.*}} void @_Z26pass_agg_nofloat_emptybase21agg_nofloat_emptybase(ptr dead_on_unwind noalias writable sret(%struct.agg_nofloat_emptybase) align 4 %{{.*}}, i64 %{{.*}})
 // CHECK-LE-LABEL: define{{.*}} i64 @_Z26pass_agg_nofloat_emptybase21agg_nofloat_emptybase(i64 %{{.*}})
 agg_nofloat_emptybase pass_agg_nofloat_emptybase(agg_nofloat_emptybase arg) { return arg; }
 
 struct emptybase { [[no_unique_address]] empty dummy; };
 struct agg_float_emptybase : emptybase { float a; };
-// CHECK-BE-LABEL: define{{.*}} void @_Z24pass_agg_float_emptybase19agg_float_emptybase(ptr noalias sret(%struct.agg_float_emptybase) align 4 %{{.*}}, float inreg %{{.*}})
+// CHECK-BE-LABEL: define{{.*}} void @_Z24pass_agg_float_emptybase19agg_float_emptybase(ptr dead_on_unwind noalias writable sret(%struct.agg_float_emptybase) align 4 %{{.*}}, float inreg %{{.*}})
 // CHECK-LE-LABEL: define{{.*}} [1 x float] @_Z24pass_agg_float_emptybase19agg_float_emptybase(float inreg %{{.*}})
 agg_float_emptybase pass_agg_float_emptybase(agg_float_emptybase arg) { return arg; }
 
 struct noemptybasearray { [[no_unique_address]] empty dummy[3]; };
 struct agg_nofloat_emptybasearray : noemptybasearray { float a; };
-// CHECK-BE-LABEL: define{{.*}} void @_Z31pass_agg_nofloat_emptybasearray26agg_nofloat_emptybasearray(ptr noalias sret(%struct.agg_nofloat_emptybasearray) align 4 %{{.*}}, i64 %{{.*}})
+// CHECK-BE-LABEL: define{{.*}} void @_Z31pass_agg_nofloat_emptybasearray26agg_nofloat_emptybasearray(ptr dead_on_unwind noalias writable sret(%struct.agg_nofloat_emptybasearray) align 4 %{{.*}}, i64 %{{.*}})
 // CHECK-LE-LABEL: define{{.*}} i64 @_Z31pass_agg_nofloat_emptybasearray26agg_nofloat_emptybasearray(i64 %{{.*}})
 agg_nofloat_emptybasearray pass_agg_nofloat_emptybasearray(agg_nofloat_emptybasearray arg) { return arg; }
 
-// CHECK-BE: call void @_Z24pass_agg_float_emptybase19agg_float_emptybase(ptr sret(%struct.agg_float_emptybase) align 4 %{{.*}}, float inreg %{{.*}})
+// CHECK-BE: call void @_Z24pass_agg_float_emptybase19agg_float_emptybase(ptr dead_on_unwind writable sret(%struct.agg_float_emptybase) align 4 %{{.*}}, float inreg %{{.*}})
 // CHECK-LE: call [1 x float] @_Z24pass_agg_float_emptybase19agg_float_emptybase(float inreg %{{.*}})
 void pass_agg_float_emptybase_ptr(agg_float_emptybase* arg) { pass_agg_float_emptybase(*arg); }
-// CHECK-BE: call void @_Z26pass_agg_nofloat_emptybase21agg_nofloat_emptybase(ptr sret(%struct.agg_nofloat_emptybase) align 4 %{{.*}}, i64 %{{.*}})
+// CHECK-BE: call void @_Z26pass_agg_nofloat_emptybase21agg_nofloat_emptybase(ptr dead_on_unwind writable sret(%struct.agg_nofloat_emptybase) align 4 %{{.*}}, i64 %{{.*}})
 // CHECK-LE: call i64 @_Z26pass_agg_nofloat_emptybase21agg_nofloat_emptybase(i64 %{{.*}})
 void pass_agg_nofloat_emptybase_ptr(agg_nofloat_emptybase* arg) { pass_agg_nofloat_emptybase(*arg); }
diff --git a/clang/test/CodeGen/PowerPC/ppc32-and-aix-struct-return.c b/clang/test/CodeGen/PowerPC/ppc32-and-aix-struct-return.c
index b0bb089d664c2..cfe5c1ca3e851 100644
--- a/clang/test/CodeGen/PowerPC/ppc32-and-aix-struct-return.c
+++ b/clang/test/CodeGen/PowerPC/ppc32-and-aix-struct-return.c
@@ -59,42 +59,42 @@ typedef struct {
   char c[9];
 } Nine;
 
-// CHECK-AIX-LABEL:  define{{.*}} void @ret0(ptr noalias sret(%struct.Zero) {{[^,]*}})
+// CHECK-AIX-LABEL:  define{{.*}} void @ret0(ptr dead_on_unwind noalias writable sret(%struct.Zero) {{[^,]*}})
 // CHECK-SVR4-LABEL: define{{.*}} void @ret0()
 Zero ret0(void) { return (Zero){}; }
 
-// CHECK-AIX-LABEL:  define{{.*}} void @ret1(ptr noalias sret(%struct.One) {{[^,]*}})
+// CHECK-AIX-LABEL:  define{{.*}} void @ret1(ptr dead_on_unwind noalias writable sret(%struct.One) {{[^,]*}})
 // CHECK-SVR4-LABEL: define{{.*}} i8 @ret1()
 One ret1(void) { return (One){'a'}; }
 
-// CHECK-AIX-LABEL:  define{{.*}} void @ret2(ptr noalias sret(%struct.Two) {{[^,]*}})
+// CHECK-AIX-LABEL:  define{{.*}} void @ret2(ptr dead_on_unwind noalias writable sret(%struct.Two) {{[^,]*}})
 // CHECK-SVR4-LABEL: define{{.*}} i16 @ret2()
 Two ret2(void) { return (Two){123}; }
 
-// CHECK-AIX-LABEL:  define{{.*}} void @ret3(ptr noalias sret(%struct.Three) {{[^,]*}})
+// CHECK-AIX-LABEL:  define{{.*}} void @ret3(ptr dead_on_unwind noalias writable sret(%struct.Three) {{[^,]*}})
 // CHECK-SVR4-LABEL: define{{.*}} i24 @ret3()
 Three ret3(void) { return (Three){"abc"}; }
 
-// CHECK-AIX-LABEL:  define{{.*}} void @ret4(ptr noalias sret(%struct.Four) {{[^,]*}})
+// CHECK-AIX-LABEL:  define{{.*}} void @ret4(ptr dead_on_unwind noalias writable sret(%struct.Four) {{[^,]*}})
 // CHECK-SVR4-LABEL: define{{.*}} i32 @ret4()
 Four ret4(void) { return (Four){0.4}; }
 
-// CHECK-AIX-LABEL:  define{{.*}} void @ret5(ptr noalias sret(%struct.Five) {{[^,]*}})
+// CHECK-AIX-LABEL:  define{{.*}} void @ret5(ptr dead_on_unwind noalias writable sret(%struct.Five) {{[^,]*}})
 // CHECK-SVR4-LABEL: define{{.*}} i40 @ret5()
 Five ret5(void) { return (Five){"abcde"}; }
 
-// CHECK-AIX-LABEL:  define{{.*}} void @ret6(ptr noalias sret(%struct.Six) {{[^,]*}})
+// CHECK-AIX-LABEL:  define{{.*}} void @ret6(ptr dead_on_unwind noalias writable sret(%struct.Six) {{[^,]*}})
 // CHECK-SVR4-LABEL: define{{.*}} i48 @ret6()
 Six ret6(void) { return (Six){12, 34, 56}; }
 
-// CHECK-AIX-LABEL:  define{{.*}} void @ret7(ptr noalias sret(%struct.Seven) {{[^,]*}})
+// CHECK-AIX-LABEL:  define{{.*}} void @ret7(ptr dead_on_unwind noalias writable sret(%struct.Seven) {{[^,]*}})
 // CHECK-SVR4-LABEL: define{{.*}} i56 @ret7()
 Seven ret7(void) { return (Seven){"abcdefg"}; }
 
-// CHECK-AIX-LABEL:  define{{.*}} void @ret8(ptr noalias sret(%struct.Eight) {{[^,]*}})
+// CHECK-AIX-LABEL:  define{{.*}} void @ret8(ptr dead_on_unwind noalias writable sret(%struct.Eight) {{[^,]*}})
 // CHECK-SVR4-LABEL: define{{.*}} i64 @ret8()
 Eight ret8(void) { return (Eight){123, 'a'}; }
 
-// CHECK-AIX-LABEL:  define{{.*}} void @ret9(ptr noalias sret(%struct.Nine) {{[^,]*}})
-// CHECK-SVR4-LABEL: define{{.*}} void @ret9(ptr noalias sret(%struct.Nine) {{[^,]*}})
+// CHECK-AIX-LABEL:  define{{.*}} void @ret9(ptr dead_on_unwind noalias writable sret(%struct.Nine) {{[^,]*}})
+// CHECK-SVR4-LABEL: define{{.*}} void @ret9(ptr dead_on_unwind noalias writable sret(%struct.Nine) {{[^,]*}})
 Nine ret9(void) { return (Nine){"abcdefghi"}; }
diff --git a/clang/test/CodeGen/PowerPC/ppc64-align-struct.c b/clang/test/CodeGen/PowerPC/ppc64-align-struct.c
index 2476c7149d076..50e8330fceaec 100644
--- a/clang/test/CodeGen/PowerPC/ppc64-align-struct.c
+++ b/clang/test/CodeGen/PowerPC/ppc64-align-struct.c
@@ -60,7 +60,7 @@ void test9 (int x, struct test9 y)
 {
 }
 
-// CHECK: define{{.*}} void @test1va(ptr noalias sret(%struct.test1) align 4 %[[AGG_RESULT:.*]], i32 noundef signext %x, ...)
+// CHECK: define{{.*}} void @test1va(ptr dead_on_unwind noalias writable sret(%struct.test1) align 4 %[[AGG_RESULT:.*]], i32 noundef signext %x, ...)
 // CHECK: %[[CUR:[^ ]+]] = load ptr, ptr %ap
 // CHECK: %[[NEXT:[^ ]+]] = getelementptr inbounds i8, ptr %[[CUR]], i64 8
 // CHECK: store ptr %[[NEXT]], ptr %ap
@@ -75,7 +75,7 @@ struct test1 test1va (int x, ...)
   return y;
 }
 
-// CHECK: define{{.*}} void @test2va(ptr noalias sret(%struct.test2) align 16 %[[AGG_RESULT:.*]], i32 noundef signext %x, ...)
+// CHECK: define{{.*}} void @test2va(ptr dead_on_unwind noalias writable sret(%struct.test2) align 16 %[[AGG_RESULT:.*]], i32 noundef signext %x, ...)
 // CHECK: %[[CUR:[^ ]+]] = load ptr, ptr %ap
 // CHECK: %[[TMP0:[^ ]+]] = getelementptr inbounds i8, ptr %[[CUR]], i32 15
 // CHECK: %[[ALIGN:[^ ]+]] = call ptr @llvm.ptrmask.p0.i64(ptr %[[TMP0]], i64 -16)
@@ -92,7 +92,7 @@ struct test2 test2va (int x, ...)
   return y;
 }
 
-// CHECK: define{{.*}} void @test3va(ptr noalias sret(%struct.test3) align 32 %[[AGG_RESULT:.*]], i32 noundef signext %x, ...)
+// CHECK: define{{.*}} void @test3va(ptr dead_on_unwind noalias writable sret(%struct.test3) align 32 %[[AGG_RESULT:.*]], i32 noundef signext %x, ...)
 // CHECK: %[[CUR:[^ ]+]] = load ptr, ptr %ap
 // CHECK: %[[TMP0:[^ ]+]] = getelementptr inbounds i8, ptr %[[CUR]], i32 15
 // CHECK: %[[ALIGN:[^ ]+]] = call ptr @llvm.ptrmask.p0.i64(ptr %[[TMP0]], i64 -16)
@@ -109,7 +109,7 @@ struct test3 test3va (int x, ...)
   return y;
 }
 
-// CHECK: define{{.*}} void @test4va(ptr noalias sret(%struct.test4) align 4 %[[AGG_RESULT:.*]], i32 noundef signext %x, ...)
+// CHECK: define{{.*}} void @test4va(ptr dead_on_unwind noalias writable sret(%struct.test4) align 4 %[[AGG_RESULT:.*]], i32 noundef signext %x, ...)
 // CHECK: %[[CUR:[^ ]+]] = load ptr, ptr %ap
 // CHECK: %[[NEXT:[^ ]+]] = getelementptr inbounds i8, ptr %[[CUR]], i64 16
 // CHECK: store ptr %[[NEXT]], ptr %ap
@@ -124,7 +124,7 @@ struct test4 test4va (int x, ...)
   return y;
 }
 
-// CHECK: define{{.*}} void @test8va(ptr noalias sret(%struct.test8) align 1 %[[AGG_RESULT:.*]], i32 noundef signext %x, ...)
+// CHECK: define{{.*}} void @test8va(ptr dead_on_unwind noalias writable sret(%struct.test8) align 1 %[[AGG_RESULT:.*]], i32 noundef signext %x, ...)
 // CHECK: %[[CUR:[^ ]+]] = load ptr, ptr %ap
 // CHECK: %[[NEXT:[^ ]+]] = getelementptr inbounds i8, ptr %[[CUR]], i64 8
 // CHECK: store ptr %[[NEXT]], ptr %ap
@@ -140,7 +140,7 @@ struct test8 test8va (int x, ...)
   return y;
 }
 
-// CHECK: define{{.*}} void @test9va(ptr noalias sret(%struct.test9) align 1 %[[AGG_RESULT:.*]], i32 noundef signext %x, ...)
+// CHECK: define{{.*}} void @test9va(ptr dead_on_unwind noalias writable sret(%struct.test9) align 1 %[[AGG_RESULT:.*]], i32 noundef signext %x, ...)
 // CHECK: %[[CUR:[^ ]+]] = load ptr, ptr %ap
 // CHECK: %[[NEXT:[^ ]+]] = getelementptr inbounds i8, ptr %[[CUR]], i64 8
 // CHECK: store ptr %[[NEXT]], ptr %ap
@@ -156,7 +156,7 @@ struct test9 test9va (int x, ...)
   return y;
 }
 
-// CHECK: define{{.*}} void @testva_longdouble(ptr noalias sret(%struct.test_longdouble) align 16 %[[AGG_RESULT:.*]], i32 noundef signext %x, ...)
+// CHECK: define{{.*}} void @testva_longdouble(ptr dead_on_unwind noalias writable sret(%struct.test_longdouble) align 16 %[[AGG_RESULT:.*]], i32 noundef signext %x, ...)
 // CHECK: %[[CUR:[^ ]+]] = load ptr, ptr %ap
 // CHECK: %[[NEXT:[^ ]+]] = getelementptr inbounds i8, ptr %[[CUR]], i64 16
 // CHECK: store ptr %[[NEXT]], ptr %ap
@@ -172,7 +172,7 @@ struct test_longdouble testva_longdouble (int x, ...)
   return y;
 }
 
-// CHECK: define{{.*}} void @testva_vector(ptr noalias sret(%struct.test_vector) align 16 %[[AGG_RESULT:.*]], i32 noundef signext %x, ...)
+// CHECK: define{{.*}} void @testva_vector(ptr dead_on_unwind noalias writable sret(%struct.test_vector) align 16 %[[AGG_RESULT:.*]], i32 noundef signext %x, ...)
 // CHECK: %[[CUR:[^ ]+]] = load ptr, ptr %ap
 // CHECK: %[[TMP0:[^ ]+]] = getelementptr inbounds i8, ptr %[[CUR]], i32 15
 // CHECK: %[[ALIGN:[^ ]+]] = call ptr @llvm.ptrmask.p0.i64(ptr %[[TMP0]], i64 -16)
diff --git a/clang/test/CodeGen/PowerPC/ppc64-elf-abi.c b/clang/test/CodeGen/PowerPC/ppc64-elf-abi.c
index 64bff635af953..fa7f79455fda4 100644
--- a/clang/test/CodeGen/PowerPC/ppc64-elf-abi.c
+++ b/clang/test/CodeGen/PowerPC/ppc64-elf-abi.c
@@ -15,7 +15,7 @@
 // RUN: %clang_cc1 -triple powerpc64le-unknown-linux-gnu -emit-llvm -o - %s \
 // RUN:   -target-abi elfv2 | FileCheck %s --check-prefix=CHECK-ELFv2
 
-// CHECK-ELFv1: define{{.*}} void @func_fab(ptr noalias sret(%struct.fab) align 4 %agg.result, i64 %x.coerce)
+// CHECK-ELFv1: define{{.*}} void @func_fab(ptr dead_on_unwind noalias writable sret(%struct.fab) align 4 %agg.result, i64 %x.coerce)
 // CHECK-ELFv2: define{{.*}} [2 x float] @func_fab([2 x float] %x.coerce)
 struct fab { float a; float b; };
 struct fab func_fab(struct fab x) { return x; }
diff --git a/clang/test/CodeGen/PowerPC/ppc64-soft-float.c b/clang/test/CodeGen/PowerPC/ppc64-soft-float.c
index c2e887ef7ab6e..812d9ff5a3509 100644
--- a/clang/test/CodeGen/PowerPC/ppc64-soft-float.c
+++ b/clang/test/CodeGen/PowerPC/ppc64-soft-float.c
@@ -30,53 +30,53 @@ struct fabc { float a; float b; float c; };
 struct f2a2b { float a[2]; float b[2]; };
 
 // CHECK-LE: define{{.*}} i32 @func_f1(float inreg %x.coerce)
-// CHECK-BE: define{{.*}} void @func_f1(ptr noalias sret(%struct.f1) align 4 %agg.result, float inreg %x.coerce)
+// CHECK-BE: define{{.*}} void @func_f1(ptr dead_on_unwind noalias writable sret(%struct.f1) align 4 %agg.result, float inreg %x.coerce)
 struct f1 func_f1(struct f1 x) { return x; }
 
 // CHECK-LE: define{{.*}} i64 @func_f2(i64 %x.coerce)
-// CHECK-BE: define{{.*}} void @func_f2(ptr noalias sret(%struct.f2) align 4 %agg.result, i64 %x.coerce)
+// CHECK-BE: define{{.*}} void @func_f2(ptr dead_on_unwind noalias writable sret(%struct.f2) align 4 %agg.result, i64 %x.coerce)
 struct f2 func_f2(struct f2 x) { return x; }
 
 // CHECK-LE: define{{.*}} { i64, i64 } @func_f3([2 x i64] %x.coerce)
-// CHECK-BE: define{{.*}} void @func_f3(ptr noalias sret(%struct.f3) align 4 %agg.result, [2 x i64] %x.coerce)
+// CHECK-BE: define{{.*}} void @func_f3(ptr dead_on_unwind noalias writable sret(%struct.f3) align 4 %agg.result, [2 x i64] %x.coerce)
 struct f3 func_f3(struct f3 x) { return x; }
 
 // CHECK-LE: define{{.*}} { i64, i64 } @func_f4([2 x i64] %x.coerce)
-// CHECK-BE: define{{.*}} void @func_f4(ptr noalias sret(%struct.f4) align 4 %agg.result, [2 x i64] %x.coerce)
+// CHECK-BE: define{{.*}} void @func_f4(ptr dead_on_unwind noalias writable sret(%struct.f4) align 4 %agg.result, [2 x i64] %x.coerce)
 struct f4 func_f4(struct f4 x) { return x; }
 
-// CHECK: define{{.*}} void @func_f5(ptr noalias sret(%struct.f5) align 4 %agg.result, [3 x i64] %x.coerce)
+// CHECK: define{{.*}} void @func_f5(ptr dead_on_unwind noalias writable sret(%struct.f5) align 4 %agg.result, [3 x i64] %x.coerce)
 struct f5 func_f5(struct f5 x) { return x; }
 
-// CHECK: define{{.*}} void @func_f6(ptr noalias sret(%struct.f6) align 4 %agg.result, [3 x i64] %x.coerce)
+// CHECK: define{{.*}} void @func_f6(ptr dead_on_unwind noalias writable sret(%struct.f6) align 4 %agg.result, [3 x i64] %x.coerce)
 struct f6 func_f6(struct f6 x) { return x; }
 
-// CHECK: define{{.*}} void @func_f7(ptr noalias sret(%struct.f7) align 4 %agg.result, [4 x i64] %x.coerce)
+// CHECK: define{{.*}} void @func_f7(ptr dead_on_unwind noalias writable sret(%struct.f7) align 4 %agg.result, [4 x i64] %x.coerce)
 struct f7 func_f7(struct f7 x) { return x; }
 
-// CHECK: define{{.*}} void @func_f8(ptr noalias sret(%struct.f8) align 4 %agg.result, [4 x i64] %x.coerce)
+// CHECK: define{{.*}} void @func_f8(ptr dead_on_unwind noalias writable sret(%struct.f8) align 4 %agg.result, [4 x i64] %x.coerce)
 struct f8 func_f8(struct f8 x) { return x; }
 
-// CHECK: define{{.*}} void @func_f9(ptr noalias sret(%struct.f9) align 4 %agg.result, [5 x i64] %x.coerce)
+// CHECK: define{{.*}} void @func_f9(ptr dead_on_unwind noalias writable sret(%struct.f9) align 4 %agg.result, [5 x i64] %x.coerce)
 struct f9 func_f9(struct f9 x) { return x; }
 
 // CHECK-LE: define{{.*}} i64 @func_fab(i64 %x.coerce)
-// CHECK-BE: define{{.*}} void @func_fab(ptr noalias sret(%struct.fab) align 4 %agg.result, i64 %x.coerce)
+// CHECK-BE: define{{.*}} void @func_fab(ptr dead_on_unwind noalias writable sret(%struct.fab) align 4 %agg.result, i64 %x.coerce)
 struct fab func_fab(struct fab x) { return x; }
 
 // CHECK-LE: define{{.*}} { i64, i64 } @func_fabc([2 x i64] %x.coerce)
-// CHECK-BE: define{{.*}} void @func_fabc(ptr noalias sret(%struct.fabc) align 4 %agg.result, [2 x i64] %x.coerce)
+// CHECK-BE: define{{.*}} void @func_fabc(ptr dead_on_unwind noalias writable sret(%struct.fabc) align 4 %agg.result, [2 x i64] %x.coerce)
 struct fabc func_fabc(struct fabc x) { return x; }
 
 // CHECK-LE: define{{.*}} { i64, i64 } @func_f2a2b([2 x i64] %x.coerce)
-// CHECK-BE: define{{.*}} void @func_f2a2b(ptr noalias sret(%struct.f2a2b) align 4 %agg.result, [2 x i64] %x.coerce)
+// CHECK-BE: define{{.*}} void @func_f2a2b(ptr dead_on_unwind noalias writable sret(%struct.f2a2b) align 4 %agg.result, [2 x i64] %x.coerce)
 struct f2a2b func_f2a2b(struct f2a2b x) { return x; }
 
 // CHECK-LABEL: @call_f1
 // CHECK-BE: %[[TMP0:[^ ]+]] = alloca %struct.f1, align 4
 // CHECK: %[[TMP:[^ ]+]] = load float, ptr @global_f1, align 4
 // CHECK-LE: call i32 @func_f1(float inreg %[[TMP]])
-// CHECK-BE: call void @func_f1(ptr sret(%struct.f1) align 4 %[[TMP0]], float inreg %[[TMP]])
+// CHECK-BE: call void @func_f1(ptr dead_on_unwind writable sret(%struct.f1) align 4 %[[TMP0]], float inreg %[[TMP]])
 struct f1 global_f1;
 void call_f1(void) { global_f1 = func_f1(global_f1); }
 
@@ -84,7 +84,7 @@ void call_f1(void) { global_f1 = func_f1(global_f1); }
 // CHECK-BE: %[[TMP0:[^ ]+]] = alloca %struct.f2, align 4
 // CHECK: %[[TMP:[^ ]+]] = load i64, ptr @global_f2, align 4
 // CHECK-LE: call i64 @func_f2(i64 %[[TMP]])
-// CHECK-BE: call void @func_f2(ptr sret(%struct.f2) align 4 %[[TMP0]], i64 %[[TMP]])
+// CHECK-BE: call void @func_f2(ptr dead_on_unwind writable sret(%struct.f2) align 4 %[[TMP0]], i64 %[[TMP]])
 struct f2 global_f2;
 void call_f2(void) { global_f2 = func_f2(global_f2); }
 
@@ -94,7 +94,7 @@ void call_f2(void) { global_f2 = func_f2(global_f2); }
 // CHECK: call void @llvm.memcpy.p0.p0.i64(ptr align 8 %[[TMP1]], ptr align 4 @global_f3, i64 12, i1 false)
 // CHECK: %[[TMP3:[^ ]+]] = load [2 x i64], ptr %[[TMP1]]
 // CHECK-LE: call { i64, i64 } @func_f3([2 x i64] %[[TMP3]])
-// CHECK-BE: call void @func_f3(ptr sret(%struct.f3) align 4 %[[TMP0]], [2 x i64] %[[TMP3]])
+// CHECK-BE: call void @func_f3(ptr dead_on_unwind writable sret(%struct.f3) align 4 %[[TMP0]], [2 x i64] %[[TMP3]])
 struct f3 global_f3;
 void call_f3(void) { global_f3 = func_f3(global_f3); }
 
@@ -102,7 +102,7 @@ void call_f3(void) { global_f3 = func_f3(global_f3); }
 // CHECK-BE: %[[TMP0:[^ ]+]] = alloca %struct.f4, align 4
 // CHECK: %[[TMP:[^ ]+]] = load [2 x i64], ptr @global_f4, align 4
 // CHECK-LE: call { i64, i64 } @func_f4([2 x i64] %[[TMP]])
-// CHECK-BE: call void @func_f4(ptr sret(%struct.f4) align 4 %[[TMP0]], [2 x i64] %[[TMP]])
+// CHECK-BE: call void @func_f4(ptr dead_on_unwind writable sret(%struct.f4) align 4 %[[TMP0]], [2 x i64] %[[TMP]])
 struct f4 global_f4;
 void call_f4(void) { global_f4 = func_f4(global_f4); }
 
@@ -111,14 +111,14 @@ void call_f4(void) { global_f4 = func_f4(global_f4); }
 // CHECK: %[[TMP1:[^ ]+]] = alloca [3 x i64]
 // CHECK: call void @llvm.memcpy.p0.p0.i64(ptr align 8 %[[TMP1]], ptr align 4 @global_f5, i64 20, i1 false)
 // CHECK: %[[TMP3:[^ ]+]] = load [3 x i64], ptr %[[TMP1]]
-// CHECK: call void @func_f5(ptr sret(%struct.f5) align 4 %[[TMP0]], [3 x i64] %[[TMP3]])
+// CHECK: call void @func_f5(ptr dead_on_unwind writable sret(%struct.f5) align 4 %[[TMP0]], [3 x i64] %[[TMP3]])
 struct f5 global_f5;
 void call_f5(void) { global_f5 = func_f5(global_f5); }
 
 // CHECK-LABEL: @call_f6
 // CHECK: %[[TMP0:[^ ]+]] = alloca %struct.f6, align 4
 // CHECK: %[[TMP:[^ ]+]] = load [3 x i64], ptr @global_f6, align 4
-// CHECK: call void @func_f6(ptr sret(%struct.f6) align 4 %[[TMP0]], [3 x i64] %[[TMP]])
+// CHECK: call void @func_f6(ptr dead_on_unwind writable sret(%struct.f6) align 4 %[[TMP0]], [3 x i64] %[[TMP]])
 struct f6 global_f6;
 void call_f6(void) { global_f6 = func_f6(global_f6); }
 
@@ -127,14 +127,14 @@ void call_f6(void) { global_f6 = func_f6(global_f6); }
 // CHECK: %[[TMP1:[^ ]+]] = alloca [4 x i64], align 8
 // CHECK: call void @llvm.memcpy.p0.p0.i64(ptr align 8 %[[TMP1]], ptr align 4 @global_f7, i64 28, i1 false)
 // CHECK: %[[TMP3:[^ ]+]] = load [4 x i64], ptr %[[TMP1]], align 8
-// CHECK: call void @func_f7(ptr sret(%struct.f7) align 4 %[[TMP0]], [4 x i64] %[[TMP3]])
+// CHECK: call void @func_f7(ptr dead_on_unwind writable sret(%struct.f7) align 4 %[[TMP0]], [4 x i64] %[[TMP3]])
 struct f7 global_f7;
 void call_f7(void) { global_f7 = func_f7(global_f7); }
 
 // CHECK-LABEL: @call_f8
 // CHECK: %[[TMP0:[^ ]+]] = alloca %struct.f8, align 4
 // CHECK: %[[TMP:[^ ]+]] = load [4 x i64], ptr @global_f8, align 4
-// CHECK: call void @func_f8(ptr sret(%struct.f8) align 4 %[[TMP0]], [4 x i64] %[[TMP]])
+// CHECK: call void @func_f8(ptr dead_on_unwind writable sret(%struct.f8) align 4 %[[TMP0]], [4 x i64] %[[TMP]])
 struct f8 global_f8;
 void call_f8(void) { global_f8 = func_f8(global_f8); }
 
@@ -142,7 +142,7 @@ void call_f8(void) { global_f8 = func_f8(global_f8); }
 // CHECK: %[[TMP1:[^ ]+]] = alloca [5 x i64]
 // CHECK: call void @llvm.memcpy.p0.p0.i64(ptr align 8 %[[TMP1]], ptr align 4 @global_f9, i64 36, i1 false)
 // CHECK: %[[TMP3:[^ ]+]] = load [5 x i64], ptr %[[TMP1]]
-// CHECK: call void @func_f9(ptr sret(%struct.f9) align 4 %{{[^ ]+}}, [5 x i64] %[[TMP3]])
+// CHECK: call void @func_f9(ptr dead_on_unwind writable sret(%struct.f9) align 4 %{{[^ ]+}}, [5 x i64] %[[TMP3]])
 struct f9 global_f9;
 void call_f9(void) { global_f9 = func_f9(global_f9); }
 
@@ -150,7 +150,7 @@ void call_f9(void) { global_f9 = func_f9(global_f9); }
 // CHECK: %[[TMP0:[^ ]+]] = alloca %struct.fab, align 4
 // CHECK: %[[TMP:[^ ]+]] = load i64, ptr @global_fab, align 4
 // CHECK-LE: %call = call i64 @func_fab(i64 %[[TMP]])
-// CHECK-BE: call void @func_fab(ptr sret(%struct.fab) align 4 %[[TMP0]], i64 %[[TMP]])
+// CHECK-BE: call void @func_fab(ptr dead_on_unwind writable sret(%struct.fab) align 4 %[[TMP0]], i64 %[[TMP]])
 struct fab global_fab;
 void call_fab(void) { global_fab = func_fab(global_fab); }
 
@@ -160,7 +160,7 @@ void call_fab(void) { global_fab = func_fab(global_fab); }
 // CHECK: call void @llvm.memcpy.p0.p0.i64(ptr align 8 %[[TMP0]], ptr align 4 @global_fabc, i64 12, i1 false)
 // CHECK: %[[TMP3:[^ ]+]] = load [2 x i64], ptr %[[TMP0]], align 8
 // CHECK-LE: %call = call { i64, i64 } @func_fabc([2 x i64] %[[TMP3]])
-// CHECK-BE: call void @func_fabc(ptr sret(%struct.fabc) align 4 %[[TMPX]], [2 x i64] %[[TMP3]])
+// CHECK-BE: call void @func_fabc(ptr dead_on_unwind writable sret(%struct.fabc) align 4 %[[TMPX]], [2 x i64] %[[TMP3]])
 struct fabc global_fabc;
 void call_fabc(void) { global_fabc = func_fabc(global_fabc); }
 
diff --git a/clang/test/CodeGen/PowerPC/ppc64-vector.c b/clang/test/CodeGen/PowerPC/ppc64-vector.c
index 2e685799efa07..5d3dd86a009d5 100644
--- a/clang/test/CodeGen/PowerPC/ppc64-vector.c
+++ b/clang/test/CodeGen/PowerPC/ppc64-vector.c
@@ -39,13 +39,13 @@ v8i16 test_v8i16(v8i16 x)
   return x;
 }
 
-// CHECK: define{{.*}} void @test_v16i16(ptr noalias sret(<16 x i16>) align 32 %agg.result, ptr noundef %0)
+// CHECK: define{{.*}} void @test_v16i16(ptr dead_on_unwind noalias writable sret(<16 x i16>) align 32 %agg.result, ptr noundef %0)
 v16i16 test_v16i16(v16i16 x)
 {
   return x;
 }
 
-// CHECK: define{{.*}} void @test_struct_v16i16(ptr noalias sret(%struct.v16i16) align 32 %agg.result, [2 x i128] %x.coerce)
+// CHECK: define{{.*}} void @test_struct_v16i16(ptr dead_on_unwind noalias writable sret(%struct.v16i16) align 32 %agg.result, [2 x i128] %x.coerce)
 struct v16i16 test_struct_v16i16(struct v16i16 x)
 {
   return x;
diff --git a/clang/test/CodeGen/PowerPC/ppc64le-aggregates.c b/clang/test/CodeGen/PowerPC/ppc64le-aggregates.c
index 42b179217dd2e..3ea85a6a8ff3b 100644
--- a/clang/test/CodeGen/PowerPC/ppc64le-aggregates.c
+++ b/clang/test/CodeGen/PowerPC/ppc64le-aggregates.c
@@ -41,7 +41,7 @@ struct f7 func_f7(struct f7 x) { return x; }
 // CHECK: define{{.*}} [8 x float] @func_f8([8 x float] %x.coerce)
 struct f8 func_f8(struct f8 x) { return x; }
 
-// CHECK: define{{.*}} void @func_f9(ptr noalias sret(%struct.f9) align 4 %agg.result, [5 x i64] %x.coerce)
+// CHECK: define{{.*}} void @func_f9(ptr dead_on_unwind noalias writable sret(%struct.f9) align 4 %agg.result, [5 x i64] %x.coerce)
 struct f9 func_f9(struct f9 x) { return x; }
 
 // CHECK: define{{.*}} [2 x float] @func_fab([2 x float] %x.coerce)
@@ -105,7 +105,7 @@ void call_f8(void) { global_f8 = func_f8(global_f8); }
 // CHECK: %[[TMP1:[^ ]+]] = alloca [5 x i64]
 // CHECK: call void @llvm.memcpy.p0.p0.i64(ptr align 8 %[[TMP1]], ptr align 4 @global_f9, i64 36, i1 false)
 // CHECK: %[[TMP3:[^ ]+]] = load [5 x i64], ptr %[[TMP1]]
-// CHECK: call void @func_f9(ptr sret(%struct.f9) align 4 %{{[^ ]+}}, [5 x i64] %[[TMP3]])
+// CHECK: call void @func_f9(ptr dead_on_unwind writable sret(%struct.f9) align 4 %{{[^ ]+}}, [5 x i64] %[[TMP3]])
 struct f9 global_f9;
 void call_f9(void) { global_f9 = func_f9(global_f9); }
 
@@ -161,7 +161,7 @@ struct v7 func_v7(struct v7 x) { return x; }
 // CHECK: define{{.*}} [8 x <4 x i32>] @func_v8([8 x <4 x i32>] %x.coerce)
 struct v8 func_v8(struct v8 x) { return x; }
 
-// CHECK: define{{.*}} void @func_v9(ptr noalias sret(%struct.v9) align 16 %agg.result, ptr noundef byval(%struct.v9) align 16 %x)
+// CHECK: define{{.*}} void @func_v9(ptr dead_on_unwind noalias writable sret(%struct.v9) align 16 %agg.result, ptr noundef byval(%struct.v9) align 16 %x)
 struct v9 func_v9(struct v9 x) { return x; }
 
 // CHECK: define{{.*}} [2 x <4 x i32>] @func_vab([2 x <4 x i32>] %x.coerce)
@@ -219,7 +219,7 @@ struct v8 global_v8;
 void call_v8(void) { global_v8 = func_v8(global_v8); }
 
 // CHECK-LABEL: @call_v9
-// CHECK: call void @func_v9(ptr sret(%struct.v9) align 16 %{{[^ ]+}}, ptr noundef byval(%struct.v9) align 16 @global_v9)
+// CHECK: call void @func_v9(ptr dead_on_unwind writable sret(%struct.v9) align 16 %{{[^ ]+}}, ptr noundef byval(%struct.v9) align 16 @global_v9)
 struct v9 global_v9;
 void call_v9(void) { global_v9 = func_v9(global_v9); }
 
@@ -278,7 +278,7 @@ struct v3f7 func_v3f7(struct v3f7 x) { return x; }
 // CHECK: define{{.*}} [8 x <4 x float>] @func_v3f8([8 x <4 x float>] %x.coerce)
 struct v3f8 func_v3f8(struct v3f8 x) { return x; }
 
-// CHECK: define{{.*}} void @func_v3f9(ptr noalias sret(%struct.v3f9) align 16 %agg.result, ptr noundef byval(%struct.v3f9) align 16 %x)
+// CHECK: define{{.*}} void @func_v3f9(ptr dead_on_unwind noalias writable sret(%struct.v3f9) align 16 %agg.result, ptr noundef byval(%struct.v3f9) align 16 %x)
 struct v3f9 func_v3f9(struct v3f9 x) { return x; }
 
 // CHECK: define{{.*}} [2 x <4 x float>] @func_v3fab([2 x <4 x float>] %x.coerce)
@@ -336,7 +336,7 @@ struct v3f8 global_v3f8;
 void call_v3f8(void) { global_v3f8 = func_v3f8(global_v3f8); }
 
 // CHECK-LABEL: @call_v3f9
-// CHECK: call void @func_v3f9(ptr sret(%struct.v3f9) align 16 %{{[^ ]+}}, ptr noundef byval(%struct.v3f9) align 16 @global_v3f9)
+// CHECK: call void @func_v3f9(ptr dead_on_unwind writable sret(%struct.v3f9) align 16 %{{[^ ]+}}, ptr noundef byval(%struct.v3f9) align 16 @global_v3f9)
 struct v3f9 global_v3f9;
 void call_v3f9(void) { global_v3f9 = func_v3f9(global_v3f9); }
 
diff --git a/clang/test/CodeGen/PowerPC/ppc64le-f128Aggregates.c b/clang/test/CodeGen/PowerPC/ppc64le-f128Aggregates.c
index 27c3b14f0a04b..b2e79d6be9b87 100644
--- a/clang/test/CodeGen/PowerPC/ppc64le-f128Aggregates.c
+++ b/clang/test/CodeGen/PowerPC/ppc64le-f128Aggregates.c
@@ -42,7 +42,7 @@ struct fp7 func_f7(struct fp7 x) { return x; }
 // CHECK: define{{.*}} [8 x fp128] @func_f8([8 x fp128] %x.coerce)
 struct fp8 func_f8(struct fp8 x) { return x; }
 
-// CHECK: define{{.*}} void @func_f9(ptr noalias sret(%struct.fp9) align 16 %agg.result, ptr noundef byval(%struct.fp9) align 16 %x)
+// CHECK: define{{.*}} void @func_f9(ptr dead_on_unwind noalias writable sret(%struct.fp9) align 16 %agg.result, ptr noundef byval(%struct.fp9) align 16 %x)
 struct fp9 func_f9(struct fp9 x) { return x; }
 
 // CHECK: define{{.*}} [2 x fp128] @func_fab([2 x fp128] %x.coerce)
@@ -104,7 +104,7 @@ void call_fp8(void) { global_f8 = func_f8(global_f8); }
 
 // CHECK-LABEL: @call_fp9
 // CHECK: %[[TMP1:[^ ]+]] = alloca %struct.fp9, align 16
-// CHECK: call void @func_f9(ptr sret(%struct.fp9) align 16 %[[TMP2:[^ ]+]], ptr noundef byval(%struct.fp9) align 16 @global_f9
+// CHECK: call void @func_f9(ptr dead_on_unwind writable sret(%struct.fp9) align 16 %[[TMP2:[^ ]+]], ptr noundef byval(%struct.fp9) align 16 @global_f9
 // CHECK: call void @llvm.memcpy.p0.p0.i64(ptr align 16 @global_f9, ptr align 16 %[[TMP2]], i64 144, i1 false
 // CHECK: ret void
 struct fp9 global_f9;
diff --git a/clang/test/CodeGen/RISCV/bfloat-abi.c b/clang/test/CodeGen/RISCV/bfloat-abi.c
index bfaf1043133b8..f38646c8b12ff 100644
--- a/clang/test/CodeGen/RISCV/bfloat-abi.c
+++ b/clang/test/CodeGen/RISCV/bfloat-abi.c
@@ -467,7 +467,7 @@ struct floatbfloat3 {
 // CHECK-RV64-NEXT:    ret [2 x i64] [[TMP4]]
 //
 // CHECK-RV32-LABEL: define dso_local void @fh3
-// CHECK-RV32-SAME: (ptr noalias sret([[STRUCT_FLOATBFLOAT3:%.*]]) align 4 [[AGG_RESULT:%.*]], float noundef [[A:%.*]], bfloat noundef [[B:%.*]], bfloat noundef [[C:%.*]], bfloat noundef [[D:%.*]]) #[[ATTR0]] {
+// CHECK-RV32-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_FLOATBFLOAT3:%.*]]) align 4 [[AGG_RESULT:%.*]], float noundef [[A:%.*]], bfloat noundef [[B:%.*]], bfloat noundef [[C:%.*]], bfloat noundef [[D:%.*]]) #[[ATTR0]] {
 // CHECK-RV32-NEXT:  entry:
 // CHECK-RV32-NEXT:    [[RESULT_PTR:%.*]] = alloca ptr, align 4
 // CHECK-RV32-NEXT:    [[A_ADDR:%.*]] = alloca float, align 4
@@ -545,7 +545,7 @@ struct bfloat5 {
 // CHECK-RV64-NEXT:    ret [2 x i64] [[TMP5]]
 //
 // CHECK-RV32-LABEL: define dso_local void @h5
-// CHECK-RV32-SAME: (ptr noalias sret([[STRUCT_BFLOAT5:%.*]]) align 2 [[AGG_RESULT:%.*]], bfloat noundef [[A:%.*]], bfloat noundef [[B:%.*]], bfloat noundef [[C:%.*]], bfloat noundef [[D:%.*]], bfloat noundef [[E:%.*]]) #[[ATTR0]] {
+// CHECK-RV32-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_BFLOAT5:%.*]]) align 2 [[AGG_RESULT:%.*]], bfloat noundef [[A:%.*]], bfloat noundef [[B:%.*]], bfloat noundef [[C:%.*]], bfloat noundef [[D:%.*]], bfloat noundef [[E:%.*]]) #[[ATTR0]] {
 // CHECK-RV32-NEXT:  entry:
 // CHECK-RV32-NEXT:    [[RESULT_PTR:%.*]] = alloca ptr, align 4
 // CHECK-RV32-NEXT:    [[A_ADDR:%.*]] = alloca bfloat, align 2
diff --git a/clang/test/CodeGen/RISCV/riscv-abi.cpp b/clang/test/CodeGen/RISCV/riscv-abi.cpp
index aa18afe41f6db..fe1a2b6d8595c 100644
--- a/clang/test/CodeGen/RISCV/riscv-abi.cpp
+++ b/clang/test/CodeGen/RISCV/riscv-abi.cpp
@@ -75,7 +75,7 @@ struct child3_int64_s : parent3_float_s {
 };
 
 // ILP32-ILP32F-ILP32D-LABEL: define dso_local void @_Z30float_int64_struct_inheritance14child3_int64_s
-// ILP32-ILP32F-ILP32D-SAME: (ptr noalias sret([[STRUCT_CHILD3_INT64_S:%.*]]) align 8 [[AGG_RESULT:%.*]], ptr noundef [[A:%.*]]) #[[ATTR0]] {
+// ILP32-ILP32F-ILP32D-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_CHILD3_INT64_S:%.*]]) align 8 [[AGG_RESULT:%.*]], ptr noundef [[A:%.*]]) #[[ATTR0]] {
 // ILP32-ILP32F-ILP32D:  entry:
 //
 // LP64-LABEL: define dso_local [2 x i64] @_Z30float_int64_struct_inheritance14child3_int64_s
@@ -99,7 +99,7 @@ struct child4_double_s : parent4_double_s {
 };
 
 // ILP32-ILP32F-LABEL: define dso_local void @_Z32double_double_struct_inheritance15child4_double_s
-// ILP32-ILP32F-SAME: (ptr noalias sret([[STRUCT_CHILD4_DOUBLE_S:%.*]]) align 8 [[AGG_RESULT:%.*]], ptr noundef [[A:%.*]]) #[[ATTR0]] {
+// ILP32-ILP32F-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_CHILD4_DOUBLE_S:%.*]]) align 8 [[AGG_RESULT:%.*]], ptr noundef [[A:%.*]]) #[[ATTR0]] {
 // ILP32-ILP32F:  entry:
 //
 // ILP32D-LABEL: define dso_local { double, double } @_Z32double_double_struct_inheritance15child4_double_s
@@ -130,11 +130,11 @@ struct child5_virtual_s : virtual parent5_virtual_s {
 };
 
 // ILP32-ILP32F-ILP32D-LABEL: define dso_local void @_Z38int32_float_virtual_struct_inheritance16child5_virtual_s
-// ILP32-ILP32F-ILP32D-SAME: (ptr noalias sret([[STRUCT_CHILD5_VIRTUAL_S:%.*]]) align 4 [[AGG_RESULT:%.*]], ptr noundef [[A:%.*]]) #[[ATTR0]] {
+// ILP32-ILP32F-ILP32D-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_CHILD5_VIRTUAL_S:%.*]]) align 4 [[AGG_RESULT:%.*]], ptr noundef [[A:%.*]]) #[[ATTR0]] {
 // ILP32-ILP32F-ILP32D:  entry:
 //
 // LP64-LP64F-LP64D-LABEL: define dso_local void @_Z38int32_float_virtual_struct_inheritance16child5_virtual_s
-// LP64-LP64F-LP64D-SAME: (ptr noalias sret([[STRUCT_CHILD5_VIRTUAL_S:%.*]]) align 8 [[AGG_RESULT:%.*]], ptr noundef [[A:%.*]]) #[[ATTR0]] {
+// LP64-LP64F-LP64D-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_CHILD5_VIRTUAL_S:%.*]]) align 8 [[AGG_RESULT:%.*]], ptr noundef [[A:%.*]]) #[[ATTR0]] {
 // LP64-LP64F-LP64D:  entry:
 //
 struct child5_virtual_s int32_float_virtual_struct_inheritance(struct child5_virtual_s a) {
diff --git a/clang/test/CodeGen/RISCV/riscv-func-attr-target.c b/clang/test/CodeGen/RISCV/riscv-func-attr-target.c
index 759c33a225060..7d3362e84e758 100644
--- a/clang/test/CodeGen/RISCV/riscv-func-attr-target.c
+++ b/clang/test/CodeGen/RISCV/riscv-func-attr-target.c
@@ -40,8 +40,9 @@ __attribute__((target("cpu=sifive-u54"))) void testAttrCpuOnly() {}
 // CHECK: attributes #1 = { {{.*}}"target-cpu"="rocket-rv64" "target-features"="+64bit,+a,+d,+f,+m,+save-restore,+v,+zicsr,+zifencei,+zve32f,+zve32x,+zve64d,+zve64f,+zve64x,+zvl128b,+zvl32b,+zvl64b,-relax,-zbb,-zfa" "tune-cpu"="generic-rv64" }
 // CHECK: attributes #2 = { {{.*}}"target-features"="+64bit,+a,+m,+save-restore,+zbb,+zifencei,-relax,-zfa" }
 // CHECK: attributes #3 = { {{.*}}"target-features"="+64bit,+a,+d,+experimental-zicond,+f,+m,+save-restore,+v,+zbb,+zicsr,+zifencei,+zve32f,+zve32x,+zve64d,+zve64f,+zve64x,+zvl128b,+zvl32b,+zvl64b,-relax,-zfa" }
-// CHECK: attributes #4 = { {{.*}}"target-features"="+64bit,+a,+c,+d,+f,+m,+save-restore,+zbb,+zicsr,+zifencei,-relax" }
-// CHECK: attributes #5 = { {{.*}}"target-features"="+64bit,+m,+save-restore,-relax" }
+// Make sure we append negative features if we override the arch
+// CHECK: attributes #4 = { {{.*}}"target-features"="+64bit,+a,+c,+d,+f,+m,+save-restore,+zbb,+zicsr,+zifencei,{{(-[[:alnum:]-]+)(,-[[:alnum:]-]+)*}}" }
+// CHECK: attributes #5 = { {{.*}}"target-features"="+64bit,+m,+save-restore,{{(-[[:alnum:]-]+)(,-[[:alnum:]-]+)*}}" }
 // CHECK: attributes #6 = { {{.*}}"target-cpu"="sifive-u54" "target-features"="+64bit,+a,+m,+save-restore,+zbb,+zifencei,-relax,-zfa" }
-// CHECK: attributes #7 = { {{.*}}"target-cpu"="sifive-u54" "target-features"="+64bit,+m,+save-restore,-relax" }
-// CHECK: attributes #8 = { {{.*}}"target-cpu"="sifive-u54" "target-features"="+64bit,+a,+c,+d,+f,+m,+save-restore,+zicsr,+zifencei,-relax" }
+// CHECK: attributes #7 = { {{.*}}"target-cpu"="sifive-u54" "target-features"="+64bit,+m,+save-restore,{{(-[[:alnum:]-]+)(,-[[:alnum:]-]+)*}}" }
+// CHECK: attributes #8 = { {{.*}}"target-cpu"="sifive-u54" "target-features"="+64bit,+a,+c,+d,+f,+m,+save-restore,+zicsr,+zifencei,{{(-[[:alnum:]-]+)(,-[[:alnum:]-]+)*}}" }
diff --git a/clang/test/CodeGen/RISCV/riscv32-abi.c b/clang/test/CodeGen/RISCV/riscv32-abi.c
index 040ae500fc60e..b53f9a9169146 100644
--- a/clang/test/CodeGen/RISCV/riscv32-abi.c
+++ b/clang/test/CodeGen/RISCV/riscv32-abi.c
@@ -5,6 +5,8 @@
 // RUN:   | FileCheck -check-prefixes=ILP32-ILP32F-ILP32D,ILP32F-ILP32D,ILP32-ILP32F,ILP32F %s
 // RUN: %clang_cc1 -triple riscv32 -target-feature +f -target-feature +d -target-abi ilp32d -emit-llvm %s -o - \
 // RUN:   | FileCheck -check-prefixes=ILP32-ILP32F-ILP32D,ILP32F-ILP32D,ILP32D %s
+// RUN: %clang_cc1 -triple riscv32 -emit-llvm -target-abi ilp32e %s -o - \
+// RUN:     | FileCheck -check-prefixes=ILP32-ILP32F-ILP32D,ILP32-ILP32F,ILP32,ILP32E %s
 
 #include 
 #include 
@@ -254,7 +256,7 @@ void f_agg_large(struct large x) {
 // The address where the struct should be written to will be the first
 // argument
 // ILP32-ILP32F-ILP32D-LABEL: define dso_local void @f_agg_large_ret
-// ILP32-ILP32F-ILP32D-SAME: (ptr noalias sret([[STRUCT_LARGE:%.*]]) align 4 [[AGG_RESULT:%.*]], i32 noundef [[I:%.*]], i8 noundef signext [[J:%.*]]) #[[ATTR0]] {
+// ILP32-ILP32F-ILP32D-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_LARGE:%.*]]) align 4 [[AGG_RESULT:%.*]], i32 noundef [[I:%.*]], i8 noundef signext [[J:%.*]]) #[[ATTR0]] {
 // ILP32-ILP32F-ILP32D:  entry:
 //
 struct large f_agg_large_ret(int32_t i, int8_t j) {
@@ -272,7 +274,7 @@ void f_vec_large_v16i8(v16i8 x) {
 }
 
 // ILP32-ILP32F-ILP32D-LABEL: define dso_local void @f_vec_large_v16i8_ret
-// ILP32-ILP32F-ILP32D-SAME: (ptr noalias sret(<16 x i8>) align 16 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
+// ILP32-ILP32F-ILP32D-SAME: (ptr dead_on_unwind noalias writable sret(<16 x i8>) align 16 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
 // ILP32-ILP32F-ILP32D:  entry:
 //
 v16i8 f_vec_large_v16i8_ret(void) {
@@ -292,7 +294,7 @@ int f_scalar_stack_1(struct tiny a, struct small b, struct small_aligned c,
 }
 
 // ILP32-ILP32F-ILP32D-LABEL: define dso_local void @f_scalar_stack_2
-// ILP32-ILP32F-ILP32D-SAME: (ptr noalias sret([[STRUCT_LARGE:%.*]]) align 4 [[AGG_RESULT:%.*]], i32 noundef [[A:%.*]], i64 noundef [[B:%.*]], i64 noundef [[C:%.*]], fp128 noundef [[D:%.*]], i8 noundef zeroext [[E:%.*]], i8 noundef signext [[F:%.*]], i8 noundef zeroext [[G:%.*]]) #[[ATTR0]] {
+// ILP32-ILP32F-ILP32D-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_LARGE:%.*]]) align 4 [[AGG_RESULT:%.*]], i32 noundef [[A:%.*]], i64 noundef [[B:%.*]], i64 noundef [[C:%.*]], fp128 noundef [[D:%.*]], i8 noundef zeroext [[E:%.*]], i8 noundef signext [[F:%.*]], i8 noundef zeroext [[G:%.*]]) #[[ATTR0]] {
 // ILP32-ILP32F-ILP32D:  entry:
 //
 struct large f_scalar_stack_2(int32_t a, int64_t b, int64_t c, long double d,
@@ -329,7 +331,7 @@ int f_scalar_stack_5(int32_t a, int64_t b, float c, double d, long double e,
 }
 
 // ILP32-ILP32F-ILP32D-LABEL: define dso_local void @f_scalar_stack_6
-// ILP32-ILP32F-ILP32D-SAME: (ptr noalias sret([[STRUCT_LARGE:%.*]]) align 4 [[AGG_RESULT:%.*]], float noundef [[A:%.*]], i64 noundef [[B:%.*]], double noundef [[C:%.*]], fp128 noundef [[D:%.*]], i8 noundef zeroext [[E:%.*]], i8 noundef signext [[F:%.*]], i8 noundef zeroext [[G:%.*]]) #[[ATTR0]] {
+// ILP32-ILP32F-ILP32D-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_LARGE:%.*]]) align 4 [[AGG_RESULT:%.*]], float noundef [[A:%.*]], i64 noundef [[B:%.*]], double noundef [[C:%.*]], fp128 noundef [[D:%.*]], i8 noundef zeroext [[E:%.*]], i8 noundef signext [[F:%.*]], i8 noundef zeroext [[G:%.*]]) #[[ATTR0]] {
 // ILP32-ILP32F-ILP32D:  entry:
 //
 struct large f_scalar_stack_6(float a, int64_t b, double c, long double d,
@@ -374,7 +376,7 @@ struct int_double_s { int a; double b; };
 void f_int_double_s_arg(struct int_double_s a) {}
 
 // ILP32-ILP32F-LABEL: define dso_local void @f_ret_int_double_s
-// ILP32-ILP32F-SAME: (ptr noalias sret([[STRUCT_INT_DOUBLE_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
+// ILP32-ILP32F-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_INT_DOUBLE_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
 // ILP32-ILP32F:  entry:
 //
 // ILP32D-LABEL: define dso_local { i32, double } @f_ret_int_double_s
@@ -490,7 +492,7 @@ struct double_float_s { double f; float g; };
 void f_double_double_s_arg(struct double_double_s a) {}
 
 // ILP32-ILP32F-LABEL: define dso_local void @f_ret_double_double_s
-// ILP32-ILP32F-SAME: (ptr noalias sret([[STRUCT_DOUBLE_DOUBLE_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
+// ILP32-ILP32F-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_DOUBLE_DOUBLE_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
 // ILP32-ILP32F:  entry:
 //
 // ILP32D-LABEL: define dso_local { double, double } @f_ret_double_double_s
@@ -512,7 +514,7 @@ struct double_double_s f_ret_double_double_s(void) {
 void f_double_float_s_arg(struct double_float_s a) {}
 
 // ILP32-ILP32F-LABEL: define dso_local void @f_ret_double_float_s
-// ILP32-ILP32F-SAME: (ptr noalias sret([[STRUCT_DOUBLE_FLOAT_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
+// ILP32-ILP32F-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_DOUBLE_FLOAT_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
 // ILP32-ILP32F:  entry:
 //
 // ILP32D-LABEL: define dso_local { double, float } @f_ret_double_float_s
@@ -547,7 +549,7 @@ struct double_int8_zbf_s { double d; int8_t i; int : 0; };
 void f_double_int8_s_arg(struct double_int8_s a) {}
 
 // ILP32-ILP32F-ILP32D-LABEL: define dso_local void @f_ret_double_int8_s
-// ILP32-ILP32F-ILP32D-SAME: (ptr noalias sret([[STRUCT_DOUBLE_INT8_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
+// ILP32-ILP32F-ILP32D-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_DOUBLE_INT8_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
 // ILP32-ILP32F-ILP32D:  entry:
 //
 struct double_int8_s f_ret_double_int8_s(void) {
@@ -565,7 +567,7 @@ struct double_int8_s f_ret_double_int8_s(void) {
 void f_double_uint8_s_arg(struct double_uint8_s a) {}
 
 // ILP32-ILP32F-LABEL: define dso_local void @f_ret_double_uint8_s
-// ILP32-ILP32F-SAME: (ptr noalias sret([[STRUCT_DOUBLE_UINT8_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
+// ILP32-ILP32F-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_DOUBLE_UINT8_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
 // ILP32-ILP32F:  entry:
 //
 // ILP32D-LABEL: define dso_local { double, i8 } @f_ret_double_uint8_s
@@ -587,7 +589,7 @@ struct double_uint8_s f_ret_double_uint8_s(void) {
 void f_double_int32_s_arg(struct double_int32_s a) {}
 
 // ILP32-ILP32F-LABEL: define dso_local void @f_ret_double_int32_s
-// ILP32-ILP32F-SAME: (ptr noalias sret([[STRUCT_DOUBLE_INT32_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
+// ILP32-ILP32F-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_DOUBLE_INT32_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
 // ILP32-ILP32F:  entry:
 //
 // ILP32D-LABEL: define dso_local { double, i32 } @f_ret_double_int32_s
@@ -605,7 +607,7 @@ struct double_int32_s f_ret_double_int32_s(void) {
 void f_double_int64_s_arg(struct double_int64_s a) {}
 
 // ILP32-ILP32F-ILP32D-LABEL: define dso_local void @f_ret_double_int64_s
-// ILP32-ILP32F-ILP32D-SAME: (ptr noalias sret([[STRUCT_DOUBLE_INT64_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
+// ILP32-ILP32F-ILP32D-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_DOUBLE_INT64_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
 // ILP32-ILP32F-ILP32D:  entry:
 //
 struct double_int64_s f_ret_double_int64_s(void) {
@@ -623,7 +625,7 @@ struct double_int64_s f_ret_double_int64_s(void) {
 void f_double_int64bf_s_arg(struct double_int64bf_s a) {}
 
 // ILP32-ILP32F-LABEL: define dso_local void @f_ret_double_int64bf_s
-// ILP32-ILP32F-SAME: (ptr noalias sret([[STRUCT_DOUBLE_INT64BF_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
+// ILP32-ILP32F-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_DOUBLE_INT64BF_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
 // ILP32-ILP32F:  entry:
 //
 // ILP32D-LABEL: define dso_local { double, i32 } @f_ret_double_int64bf_s
@@ -648,7 +650,7 @@ struct double_int64bf_s f_ret_double_int64bf_s(void) {
 void f_double_int8_zbf_s(struct double_int8_zbf_s a) {}
 
 // ILP32-ILP32F-LABEL: define dso_local void @f_ret_double_int8_zbf_s
-// ILP32-ILP32F-SAME: (ptr noalias sret([[STRUCT_DOUBLE_INT8_ZBF_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
+// ILP32-ILP32F-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_DOUBLE_INT8_ZBF_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
 // ILP32-ILP32F:  entry:
 //
 // ILP32D-LABEL: define dso_local { double, i8 } @f_ret_double_int8_zbf_s
@@ -687,7 +689,7 @@ void f_struct_double_int8_insufficient_fprs(float a, double b, double c, double
 void f_doublecomplex(double __complex__ a) {}
 
 // ILP32-ILP32F-LABEL: define dso_local void @f_ret_doublecomplex
-// ILP32-ILP32F-SAME: (ptr noalias sret({ double, double }) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
+// ILP32-ILP32F-SAME: (ptr dead_on_unwind noalias writable sret({ double, double }) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
 // ILP32-ILP32F:  entry:
 //
 // ILP32D-LABEL: define dso_local { double, double } @f_ret_doublecomplex
@@ -711,7 +713,7 @@ struct doublecomplex_s { double __complex__ c; };
 void f_doublecomplex_s_arg(struct doublecomplex_s a) {}
 
 // ILP32-ILP32F-LABEL: define dso_local void @f_ret_doublecomplex_s
-// ILP32-ILP32F-SAME: (ptr noalias sret([[STRUCT_DOUBLECOMPLEX_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
+// ILP32-ILP32F-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_DOUBLECOMPLEX_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
 // ILP32-ILP32F:  entry:
 //
 // ILP32D-LABEL: define dso_local { double, double } @f_ret_doublecomplex_s
@@ -762,7 +764,7 @@ struct doublearr2_s { double a[2]; };
 void f_doublearr2_s_arg(struct doublearr2_s a) {}
 
 // ILP32-ILP32F-LABEL: define dso_local void @f_ret_doublearr2_s
-// ILP32-ILP32F-SAME: (ptr noalias sret([[STRUCT_DOUBLEARR2_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
+// ILP32-ILP32F-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_DOUBLEARR2_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
 // ILP32-ILP32F:  entry:
 //
 // ILP32D-LABEL: define dso_local { double, double } @f_ret_doublearr2_s
@@ -786,7 +788,7 @@ struct doublearr2_tricky1_s { struct { double f[1]; } g[2]; };
 void f_doublearr2_tricky1_s_arg(struct doublearr2_tricky1_s a) {}
 
 // ILP32-ILP32F-LABEL: define dso_local void @f_ret_doublearr2_tricky1_s
-// ILP32-ILP32F-SAME: (ptr noalias sret([[STRUCT_DOUBLEARR2_TRICKY1_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
+// ILP32-ILP32F-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_DOUBLEARR2_TRICKY1_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
 // ILP32-ILP32F:  entry:
 //
 // ILP32D-LABEL: define dso_local { double, double } @f_ret_doublearr2_tricky1_s
@@ -810,7 +812,7 @@ struct doublearr2_tricky2_s { struct {}; struct { double f[1]; } g[2]; };
 void f_doublearr2_tricky2_s_arg(struct doublearr2_tricky2_s a) {}
 
 // ILP32-ILP32F-LABEL: define dso_local void @f_ret_doublearr2_tricky2_s
-// ILP32-ILP32F-SAME: (ptr noalias sret([[STRUCT_DOUBLEARR2_TRICKY2_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
+// ILP32-ILP32F-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_DOUBLEARR2_TRICKY2_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
 // ILP32-ILP32F:  entry:
 //
 // ILP32D-LABEL: define dso_local { double, double } @f_ret_doublearr2_tricky2_s
@@ -834,7 +836,7 @@ struct doublearr2_tricky3_s { union {}; struct { double f[1]; } g[2]; };
 void f_doublearr2_tricky3_s_arg(struct doublearr2_tricky3_s a) {}
 
 // ILP32-ILP32F-LABEL: define dso_local void @f_ret_doublearr2_tricky3_s
-// ILP32-ILP32F-SAME: (ptr noalias sret([[STRUCT_DOUBLEARR2_TRICKY3_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
+// ILP32-ILP32F-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_DOUBLEARR2_TRICKY3_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
 // ILP32-ILP32F:  entry:
 //
 // ILP32D-LABEL: define dso_local { double, double } @f_ret_doublearr2_tricky3_s
@@ -858,7 +860,7 @@ struct doublearr2_tricky4_s { union {}; struct { struct {}; double f[1]; } g[2];
 void f_doublearr2_tricky4_s_arg(struct doublearr2_tricky4_s a) {}
 
 // ILP32-ILP32F-LABEL: define dso_local void @f_ret_doublearr2_tricky4_s
-// ILP32-ILP32F-SAME: (ptr noalias sret([[STRUCT_DOUBLEARR2_TRICKY4_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
+// ILP32-ILP32F-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_DOUBLEARR2_TRICKY4_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
 // ILP32-ILP32F:  entry:
 //
 // ILP32D-LABEL: define dso_local { double, double } @f_ret_doublearr2_tricky4_s
@@ -881,7 +883,7 @@ struct int_double_int_s { int a; double b; int c; };
 void f_int_double_int_s_arg(struct int_double_int_s a) {}
 
 // ILP32-ILP32F-ILP32D-LABEL: define dso_local void @f_ret_int_double_int_s
-// ILP32-ILP32F-ILP32D-SAME: (ptr noalias sret([[STRUCT_INT_DOUBLE_INT_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
+// ILP32-ILP32F-ILP32D-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_INT_DOUBLE_INT_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
 // ILP32-ILP32F-ILP32D:  entry:
 //
 struct int_double_int_s f_ret_int_double_int_s(void) {
@@ -897,7 +899,7 @@ struct int64_double_s { int64_t a; double b; };
 void f_int64_double_s_arg(struct int64_double_s a) {}
 
 // ILP32-ILP32F-ILP32D-LABEL: define dso_local void @f_ret_int64_double_s
-// ILP32-ILP32F-ILP32D-SAME: (ptr noalias sret([[STRUCT_INT64_DOUBLE_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
+// ILP32-ILP32F-ILP32D-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_INT64_DOUBLE_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
 // ILP32-ILP32F-ILP32D:  entry:
 //
 struct int64_double_s f_ret_int64_double_s(void) {
@@ -913,7 +915,7 @@ struct char_char_double_s { char a; char b; double c; };
 void f_char_char_double_s_arg(struct char_char_double_s a) {}
 
 // ILP32-ILP32F-ILP32D-LABEL: define dso_local void @f_ret_char_char_double_s
-// ILP32-ILP32F-ILP32D-SAME: (ptr noalias sret([[STRUCT_CHAR_CHAR_DOUBLE_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
+// ILP32-ILP32F-ILP32D-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_CHAR_CHAR_DOUBLE_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
 // ILP32-ILP32F-ILP32D:  entry:
 //
 struct char_char_double_s f_ret_char_char_double_s(void) {
@@ -946,7 +948,7 @@ union double_u f_ret_double_u(void) {
 // double+double structs by the ABI.
 
 // ILP32-ILP32F-LABEL: define dso_local void @f_ret_double_int32_s_double_int32_s_just_sufficient_gprs
-// ILP32-ILP32F-SAME: (ptr noalias sret([[STRUCT_DOUBLE_INT32_S:%.*]]) align 8 [[AGG_RESULT:%.*]], i32 noundef [[A:%.*]], i32 noundef [[B:%.*]], i32 noundef [[C:%.*]], i32 noundef [[D:%.*]], i32 noundef [[E:%.*]], i32 noundef [[F:%.*]], i32 noundef [[G:%.*]], ptr noundef [[H:%.*]]) #[[ATTR0]] {
+// ILP32-ILP32F-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_DOUBLE_INT32_S:%.*]]) align 8 [[AGG_RESULT:%.*]], i32 noundef [[A:%.*]], i32 noundef [[B:%.*]], i32 noundef [[C:%.*]], i32 noundef [[D:%.*]], i32 noundef [[E:%.*]], i32 noundef [[F:%.*]], i32 noundef [[G:%.*]], ptr noundef [[H:%.*]]) #[[ATTR0]] {
 // ILP32-ILP32F:  entry:
 //
 // ILP32D-LABEL: define dso_local { double, i32 } @f_ret_double_int32_s_double_int32_s_just_sufficient_gprs
@@ -959,7 +961,7 @@ struct double_int32_s f_ret_double_int32_s_double_int32_s_just_sufficient_gprs(
 }
 
 // ILP32-ILP32F-LABEL: define dso_local void @f_ret_double_double_s_double_int32_s_just_sufficient_gprs
-// ILP32-ILP32F-SAME: (ptr noalias sret([[STRUCT_DOUBLE_DOUBLE_S:%.*]]) align 8 [[AGG_RESULT:%.*]], i32 noundef [[A:%.*]], i32 noundef [[B:%.*]], i32 noundef [[C:%.*]], i32 noundef [[D:%.*]], i32 noundef [[E:%.*]], i32 noundef [[F:%.*]], i32 noundef [[G:%.*]], ptr noundef [[H:%.*]]) #[[ATTR0]] {
+// ILP32-ILP32F-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_DOUBLE_DOUBLE_S:%.*]]) align 8 [[AGG_RESULT:%.*]], i32 noundef [[A:%.*]], i32 noundef [[B:%.*]], i32 noundef [[C:%.*]], i32 noundef [[D:%.*]], i32 noundef [[E:%.*]], i32 noundef [[F:%.*]], i32 noundef [[G:%.*]], ptr noundef [[H:%.*]]) #[[ATTR0]] {
 // ILP32-ILP32F:  entry:
 //
 // ILP32D-LABEL: define dso_local { double, double } @f_ret_double_double_s_double_int32_s_just_sufficient_gprs
@@ -972,7 +974,7 @@ struct double_double_s f_ret_double_double_s_double_int32_s_just_sufficient_gprs
 }
 
 // ILP32-ILP32F-LABEL: define dso_local void @f_ret_doublecomplex_double_int32_s_just_sufficient_gprs
-// ILP32-ILP32F-SAME: (ptr noalias sret({ double, double }) align 8 [[AGG_RESULT:%.*]], i32 noundef [[A:%.*]], i32 noundef [[B:%.*]], i32 noundef [[C:%.*]], i32 noundef [[D:%.*]], i32 noundef [[E:%.*]], i32 noundef [[F:%.*]], i32 noundef [[G:%.*]], ptr noundef [[H:%.*]]) #[[ATTR0]] {
+// ILP32-ILP32F-SAME: (ptr dead_on_unwind noalias writable sret({ double, double }) align 8 [[AGG_RESULT:%.*]], i32 noundef [[A:%.*]], i32 noundef [[B:%.*]], i32 noundef [[C:%.*]], i32 noundef [[D:%.*]], i32 noundef [[E:%.*]], i32 noundef [[F:%.*]], i32 noundef [[G:%.*]], ptr noundef [[H:%.*]]) #[[ATTR0]] {
 // ILP32-ILP32F:  entry:
 //
 // ILP32D-LABEL: define dso_local { double, double } @f_ret_doublecomplex_double_int32_s_just_sufficient_gprs
@@ -1193,7 +1195,7 @@ struct float_int32_s f_ret_float_int32_s(void) {
 void f_float_int64_s_arg(struct float_int64_s a) {}
 
 // ILP32-ILP32F-ILP32D-LABEL: define dso_local void @f_ret_float_int64_s
-// ILP32-ILP32F-ILP32D-SAME: (ptr noalias sret([[STRUCT_FLOAT_INT64_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
+// ILP32-ILP32F-ILP32D-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_FLOAT_INT64_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
 // ILP32-ILP32F-ILP32D:  entry:
 //
 struct float_int64_s f_ret_float_int64_s(void) {
@@ -1469,7 +1471,7 @@ struct int_float_int_s { int a; float b; int c; };
 void f_int_float_int_s_arg(struct int_float_int_s a) {}
 
 // ILP32-ILP32F-ILP32D-LABEL: define dso_local void @f_ret_int_float_int_s
-// ILP32-ILP32F-ILP32D-SAME: (ptr noalias sret([[STRUCT_INT_FLOAT_INT_S:%.*]]) align 4 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
+// ILP32-ILP32F-ILP32D-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_INT_FLOAT_INT_S:%.*]]) align 4 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
 // ILP32-ILP32F-ILP32D:  entry:
 //
 struct int_float_int_s f_ret_int_float_int_s(void) {
@@ -1485,7 +1487,7 @@ struct int64_float_s { int64_t a; float b; };
 void f_int64_float_s_arg(struct int64_float_s a) {}
 
 // ILP32-ILP32F-ILP32D-LABEL: define dso_local void @f_ret_int64_float_s
-// ILP32-ILP32F-ILP32D-SAME: (ptr noalias sret([[STRUCT_INT64_FLOAT_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
+// ILP32-ILP32F-ILP32D-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_INT64_FLOAT_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
 // ILP32-ILP32F-ILP32D:  entry:
 //
 struct int64_float_s f_ret_int64_float_s(void) {
@@ -1627,7 +1629,7 @@ struct double_float16_s { double f; _Float16 g; };
 void f_double_float16_s_arg(struct double_float16_s a) {}
 
 // ILP32-ILP32F-LABEL: define dso_local void @f_ret_double_float16_s
-// ILP32-ILP32F-SAME: (ptr noalias sret([[STRUCT_DOUBLE_FLOAT16_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
+// ILP32-ILP32F-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_DOUBLE_FLOAT16_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
 // ILP32-ILP32F:  entry:
 //
 // ILP32D-LABEL: define dso_local { double, half } @f_ret_double_float16_s
@@ -1729,7 +1731,7 @@ struct float16_int32_s f_ret_float16_int32_s(void) {
 void f_float16_int64_s_arg(struct float16_int64_s a) {}
 
 // ILP32-ILP32F-ILP32D-LABEL: define dso_local void @f_ret_float16_int64_s
-// ILP32-ILP32F-ILP32D-SAME: (ptr noalias sret([[STRUCT_FLOAT16_INT64_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
+// ILP32-ILP32F-ILP32D-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_FLOAT16_INT64_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
 // ILP32-ILP32F-ILP32D:  entry:
 //
 struct float16_int64_s f_ret_float16_int64_s(void) {
@@ -2005,7 +2007,7 @@ struct int_float16_int_s { int a; _Float16 b; int c; };
 void f_int_float16_int_s_arg(struct int_float16_int_s a) {}
 
 // ILP32-ILP32F-ILP32D-LABEL: define dso_local void @f_ret_int_float16_int_s
-// ILP32-ILP32F-ILP32D-SAME: (ptr noalias sret([[STRUCT_INT_FLOAT16_INT_S:%.*]]) align 4 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
+// ILP32-ILP32F-ILP32D-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_INT_FLOAT16_INT_S:%.*]]) align 4 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
 // ILP32-ILP32F-ILP32D:  entry:
 //
 struct int_float16_int_s f_ret_int_float16_int_s(void) {
@@ -2021,7 +2023,7 @@ struct int64_float16_s { int64_t a; _Float16 b; };
 void f_int64_float16_s_arg(struct int64_float16_s a) {}
 
 // ILP32-ILP32F-ILP32D-LABEL: define dso_local void @f_ret_int64_float16_s
-// ILP32-ILP32F-ILP32D-SAME: (ptr noalias sret([[STRUCT_INT64_FLOAT16_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
+// ILP32-ILP32F-ILP32D-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_INT64_FLOAT16_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
 // ILP32-ILP32F-ILP32D:  entry:
 //
 struct int64_float16_s f_ret_int64_float16_s(void) {
@@ -2064,4 +2066,5 @@ union float16_u f_ret_float16_u(void) {
 }
 
 //// NOTE: These prefixes are unused and the list is autogenerated. Do not add tests below this line:
+// ILP32E: {{.*}}
 // ILP32F: {{.*}}
diff --git a/clang/test/CodeGen/RISCV/riscv32-ilp32e-error.c b/clang/test/CodeGen/RISCV/riscv32-ilp32e-error.c
new file mode 100644
index 0000000000000..0afe7b025efdb
--- /dev/null
+++ b/clang/test/CodeGen/RISCV/riscv32-ilp32e-error.c
@@ -0,0 +1,4 @@
+// RUN: not %clang_cc1 -triple riscv32 -target-feature +d -emit-llvm -target-abi ilp32e %s 2>&1 \
+// RUN:     | FileCheck -check-prefix=ILP32E-WITH-FD %s
+
+// ILP32E-WITH-FD: error: invalid feature combination: ILP32E cannot be used with the D ISA extension
diff --git a/clang/test/CodeGen/RISCV/riscv32-vararg.c b/clang/test/CodeGen/RISCV/riscv32-vararg.c
index 02b1ed38e2655..1c4e41f2f54c8 100644
--- a/clang/test/CodeGen/RISCV/riscv32-vararg.c
+++ b/clang/test/CodeGen/RISCV/riscv32-vararg.c
@@ -1,9 +1,11 @@
 // NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py UTC_ARGS: --version 2
 // RUN: %clang_cc1 -triple riscv32 -emit-llvm %s -o - | FileCheck %s
 // RUN: %clang_cc1 -triple riscv32 -target-feature +f -target-abi ilp32f -emit-llvm %s -o - \
-// RUN:     | FileCheck %s
+// RUN:     | FileCheck %s -check-prefixes=CHECK,CHECK-ILP32F
 // RUN: %clang_cc1 -triple riscv32 -target-feature +d -target-feature +f -target-abi ilp32d -emit-llvm %s -o - \
-// RUN:     | FileCheck %s
+// RUN:     | FileCheck %s -check-prefixes=CHECK,CHECK-ILP32D
+// RUN: %clang_cc1 -triple riscv32 -target-abi ilp32e -emit-llvm %s -o - \
+// RUN:     | FileCheck %s -check-prefixes=CHECK,CHECK-ILP32E
 
 #include 
 #include 
@@ -102,24 +104,60 @@ int f_va_1(char *fmt, ...) {
 // used to pass varargs with 2x xlen alignment and 2x xlen size. Ensure the
 // correct offsets are used.
 
-// CHECK-LABEL: define dso_local double @f_va_2
-// CHECK-SAME: (ptr noundef [[FMT:%.*]], ...) #[[ATTR0]] {
-// CHECK-NEXT:  entry:
-// CHECK-NEXT:    [[FMT_ADDR:%.*]] = alloca ptr, align 4
-// CHECK-NEXT:    [[VA:%.*]] = alloca ptr, align 4
-// CHECK-NEXT:    [[V:%.*]] = alloca double, align 8
-// CHECK-NEXT:    store ptr [[FMT]], ptr [[FMT_ADDR]], align 4
-// CHECK-NEXT:    call void @llvm.va_start(ptr [[VA]])
-// CHECK-NEXT:    [[ARGP_CUR:%.*]] = load ptr, ptr [[VA]], align 4
-// CHECK-NEXT:    [[TMP0:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR]], i32 7
-// CHECK-NEXT:    [[ARGP_CUR_ALIGNED:%.*]] = call ptr @llvm.ptrmask.p0.i32(ptr [[TMP0]], i32 -8)
-// CHECK-NEXT:    [[ARGP_NEXT:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR_ALIGNED]], i32 8
-// CHECK-NEXT:    store ptr [[ARGP_NEXT]], ptr [[VA]], align 4
-// CHECK-NEXT:    [[TMP1:%.*]] = load double, ptr [[ARGP_CUR_ALIGNED]], align 8
-// CHECK-NEXT:    store double [[TMP1]], ptr [[V]], align 8
-// CHECK-NEXT:    call void @llvm.va_end(ptr [[VA]])
-// CHECK-NEXT:    [[TMP2:%.*]] = load double, ptr [[V]], align 8
-// CHECK-NEXT:    ret double [[TMP2]]
+// CHECK-ILP32F-LABEL: define dso_local double @f_va_2
+// CHECK-ILP32F-SAME: (ptr noundef [[FMT:%.*]], ...) #[[ATTR0]] {
+// CHECK-ILP32F-NEXT:  entry:
+// CHECK-ILP32F-NEXT:    [[FMT_ADDR:%.*]] = alloca ptr, align 4
+// CHECK-ILP32F-NEXT:    [[VA:%.*]] = alloca ptr, align 4
+// CHECK-ILP32F-NEXT:    [[V:%.*]] = alloca double, align 8
+// CHECK-ILP32F-NEXT:    store ptr [[FMT]], ptr [[FMT_ADDR]], align 4
+// CHECK-ILP32F-NEXT:    call void @llvm.va_start(ptr [[VA]])
+// CHECK-ILP32F-NEXT:    [[ARGP_CUR:%.*]] = load ptr, ptr [[VA]], align 4
+// CHECK-ILP32F-NEXT:    [[TMP0:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR]], i32 7
+// CHECK-ILP32F-NEXT:    [[ARGP_CUR_ALIGNED:%.*]] = call ptr @llvm.ptrmask.p0.i32(ptr [[TMP0]], i32 -8)
+// CHECK-ILP32F-NEXT:    [[ARGP_NEXT:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR_ALIGNED]], i32 8
+// CHECK-ILP32F-NEXT:    store ptr [[ARGP_NEXT]], ptr [[VA]], align 4
+// CHECK-ILP32F-NEXT:    [[TMP1:%.*]] = load double, ptr [[ARGP_CUR_ALIGNED]], align 8
+// CHECK-ILP32F-NEXT:    store double [[TMP1]], ptr [[V]], align 8
+// CHECK-ILP32F-NEXT:    call void @llvm.va_end(ptr [[VA]])
+// CHECK-ILP32F-NEXT:    [[TMP2:%.*]] = load double, ptr [[V]], align 8
+// CHECK-ILP32F-NEXT:    ret double [[TMP2]]
+//
+// CHECK-ILP32D-LABEL: define dso_local double @f_va_2
+// CHECK-ILP32D-SAME: (ptr noundef [[FMT:%.*]], ...) #[[ATTR0]] {
+// CHECK-ILP32D-NEXT:  entry:
+// CHECK-ILP32D-NEXT:    [[FMT_ADDR:%.*]] = alloca ptr, align 4
+// CHECK-ILP32D-NEXT:    [[VA:%.*]] = alloca ptr, align 4
+// CHECK-ILP32D-NEXT:    [[V:%.*]] = alloca double, align 8
+// CHECK-ILP32D-NEXT:    store ptr [[FMT]], ptr [[FMT_ADDR]], align 4
+// CHECK-ILP32D-NEXT:    call void @llvm.va_start(ptr [[VA]])
+// CHECK-ILP32D-NEXT:    [[ARGP_CUR:%.*]] = load ptr, ptr [[VA]], align 4
+// CHECK-ILP32D-NEXT:    [[TMP0:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR]], i32 7
+// CHECK-ILP32D-NEXT:    [[ARGP_CUR_ALIGNED:%.*]] = call ptr @llvm.ptrmask.p0.i32(ptr [[TMP0]], i32 -8)
+// CHECK-ILP32D-NEXT:    [[ARGP_NEXT:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR_ALIGNED]], i32 8
+// CHECK-ILP32D-NEXT:    store ptr [[ARGP_NEXT]], ptr [[VA]], align 4
+// CHECK-ILP32D-NEXT:    [[TMP1:%.*]] = load double, ptr [[ARGP_CUR_ALIGNED]], align 8
+// CHECK-ILP32D-NEXT:    store double [[TMP1]], ptr [[V]], align 8
+// CHECK-ILP32D-NEXT:    call void @llvm.va_end(ptr [[VA]])
+// CHECK-ILP32D-NEXT:    [[TMP2:%.*]] = load double, ptr [[V]], align 8
+// CHECK-ILP32D-NEXT:    ret double [[TMP2]]
+//
+// CHECK-ILP32E-LABEL: define dso_local double @f_va_2
+// CHECK-ILP32E-SAME: (ptr noundef [[FMT:%.*]], ...) #[[ATTR0]] {
+// CHECK-ILP32E-NEXT:  entry:
+// CHECK-ILP32E-NEXT:    [[FMT_ADDR:%.*]] = alloca ptr, align 4
+// CHECK-ILP32E-NEXT:    [[VA:%.*]] = alloca ptr, align 4
+// CHECK-ILP32E-NEXT:    [[V:%.*]] = alloca double, align 8
+// CHECK-ILP32E-NEXT:    store ptr [[FMT]], ptr [[FMT_ADDR]], align 4
+// CHECK-ILP32E-NEXT:    call void @llvm.va_start(ptr [[VA]])
+// CHECK-ILP32E-NEXT:    [[ARGP_CUR:%.*]] = load ptr, ptr [[VA]], align 4
+// CHECK-ILP32E-NEXT:    [[ARGP_NEXT:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR]], i32 8
+// CHECK-ILP32E-NEXT:    store ptr [[ARGP_NEXT]], ptr [[VA]], align 4
+// CHECK-ILP32E-NEXT:    [[TMP0:%.*]] = load double, ptr [[ARGP_CUR]], align 4
+// CHECK-ILP32E-NEXT:    store double [[TMP0]], ptr [[V]], align 8
+// CHECK-ILP32E-NEXT:    call void @llvm.va_end(ptr [[VA]])
+// CHECK-ILP32E-NEXT:    [[TMP1:%.*]] = load double, ptr [[V]], align 8
+// CHECK-ILP32E-NEXT:    ret double [[TMP1]]
 //
 double f_va_2(char *fmt, ...) {
   __builtin_va_list va;
@@ -133,40 +171,106 @@ double f_va_2(char *fmt, ...) {
 
 // Two "aligned" register pairs.
 
-// CHECK-LABEL: define dso_local double @f_va_3
-// CHECK-SAME: (ptr noundef [[FMT:%.*]], ...) #[[ATTR0]] {
-// CHECK-NEXT:  entry:
-// CHECK-NEXT:    [[FMT_ADDR:%.*]] = alloca ptr, align 4
-// CHECK-NEXT:    [[VA:%.*]] = alloca ptr, align 4
-// CHECK-NEXT:    [[V:%.*]] = alloca double, align 8
-// CHECK-NEXT:    [[W:%.*]] = alloca i32, align 4
-// CHECK-NEXT:    [[X:%.*]] = alloca double, align 8
-// CHECK-NEXT:    store ptr [[FMT]], ptr [[FMT_ADDR]], align 4
-// CHECK-NEXT:    call void @llvm.va_start(ptr [[VA]])
-// CHECK-NEXT:    [[ARGP_CUR:%.*]] = load ptr, ptr [[VA]], align 4
-// CHECK-NEXT:    [[TMP0:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR]], i32 7
-// CHECK-NEXT:    [[ARGP_CUR_ALIGNED:%.*]] = call ptr @llvm.ptrmask.p0.i32(ptr [[TMP0]], i32 -8)
-// CHECK-NEXT:    [[ARGP_NEXT:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR_ALIGNED]], i32 8
-// CHECK-NEXT:    store ptr [[ARGP_NEXT]], ptr [[VA]], align 4
-// CHECK-NEXT:    [[TMP1:%.*]] = load double, ptr [[ARGP_CUR_ALIGNED]], align 8
-// CHECK-NEXT:    store double [[TMP1]], ptr [[V]], align 8
-// CHECK-NEXT:    [[ARGP_CUR1:%.*]] = load ptr, ptr [[VA]], align 4
-// CHECK-NEXT:    [[ARGP_NEXT2:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR1]], i32 4
-// CHECK-NEXT:    store ptr [[ARGP_NEXT2]], ptr [[VA]], align 4
-// CHECK-NEXT:    [[TMP2:%.*]] = load i32, ptr [[ARGP_CUR1]], align 4
-// CHECK-NEXT:    store i32 [[TMP2]], ptr [[W]], align 4
-// CHECK-NEXT:    [[ARGP_CUR3:%.*]] = load ptr, ptr [[VA]], align 4
-// CHECK-NEXT:    [[TMP3:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR3]], i32 7
-// CHECK-NEXT:    [[ARGP_CUR3_ALIGNED:%.*]] = call ptr @llvm.ptrmask.p0.i32(ptr [[TMP3]], i32 -8)
-// CHECK-NEXT:    [[ARGP_NEXT4:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR3_ALIGNED]], i32 8
-// CHECK-NEXT:    store ptr [[ARGP_NEXT4]], ptr [[VA]], align 4
-// CHECK-NEXT:    [[TMP4:%.*]] = load double, ptr [[ARGP_CUR3_ALIGNED]], align 8
-// CHECK-NEXT:    store double [[TMP4]], ptr [[X]], align 8
-// CHECK-NEXT:    call void @llvm.va_end(ptr [[VA]])
-// CHECK-NEXT:    [[TMP5:%.*]] = load double, ptr [[V]], align 8
-// CHECK-NEXT:    [[TMP6:%.*]] = load double, ptr [[X]], align 8
-// CHECK-NEXT:    [[ADD:%.*]] = fadd double [[TMP5]], [[TMP6]]
-// CHECK-NEXT:    ret double [[ADD]]
+// CHECK-ILP32F-LABEL: define dso_local double @f_va_3
+// CHECK-ILP32F-SAME: (ptr noundef [[FMT:%.*]], ...) #[[ATTR0]] {
+// CHECK-ILP32F-NEXT:  entry:
+// CHECK-ILP32F-NEXT:    [[FMT_ADDR:%.*]] = alloca ptr, align 4
+// CHECK-ILP32F-NEXT:    [[VA:%.*]] = alloca ptr, align 4
+// CHECK-ILP32F-NEXT:    [[V:%.*]] = alloca double, align 8
+// CHECK-ILP32F-NEXT:    [[W:%.*]] = alloca i32, align 4
+// CHECK-ILP32F-NEXT:    [[X:%.*]] = alloca double, align 8
+// CHECK-ILP32F-NEXT:    store ptr [[FMT]], ptr [[FMT_ADDR]], align 4
+// CHECK-ILP32F-NEXT:    call void @llvm.va_start(ptr [[VA]])
+// CHECK-ILP32F-NEXT:    [[ARGP_CUR:%.*]] = load ptr, ptr [[VA]], align 4
+// CHECK-ILP32F-NEXT:    [[TMP0:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR]], i32 7
+// CHECK-ILP32F-NEXT:    [[ARGP_CUR_ALIGNED:%.*]] = call ptr @llvm.ptrmask.p0.i32(ptr [[TMP0]], i32 -8)
+// CHECK-ILP32F-NEXT:    [[ARGP_NEXT:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR_ALIGNED]], i32 8
+// CHECK-ILP32F-NEXT:    store ptr [[ARGP_NEXT]], ptr [[VA]], align 4
+// CHECK-ILP32F-NEXT:    [[TMP1:%.*]] = load double, ptr [[ARGP_CUR_ALIGNED]], align 8
+// CHECK-ILP32F-NEXT:    store double [[TMP1]], ptr [[V]], align 8
+// CHECK-ILP32F-NEXT:    [[ARGP_CUR1:%.*]] = load ptr, ptr [[VA]], align 4
+// CHECK-ILP32F-NEXT:    [[ARGP_NEXT2:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR1]], i32 4
+// CHECK-ILP32F-NEXT:    store ptr [[ARGP_NEXT2]], ptr [[VA]], align 4
+// CHECK-ILP32F-NEXT:    [[TMP2:%.*]] = load i32, ptr [[ARGP_CUR1]], align 4
+// CHECK-ILP32F-NEXT:    store i32 [[TMP2]], ptr [[W]], align 4
+// CHECK-ILP32F-NEXT:    [[ARGP_CUR3:%.*]] = load ptr, ptr [[VA]], align 4
+// CHECK-ILP32F-NEXT:    [[TMP3:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR3]], i32 7
+// CHECK-ILP32F-NEXT:    [[ARGP_CUR3_ALIGNED:%.*]] = call ptr @llvm.ptrmask.p0.i32(ptr [[TMP3]], i32 -8)
+// CHECK-ILP32F-NEXT:    [[ARGP_NEXT4:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR3_ALIGNED]], i32 8
+// CHECK-ILP32F-NEXT:    store ptr [[ARGP_NEXT4]], ptr [[VA]], align 4
+// CHECK-ILP32F-NEXT:    [[TMP4:%.*]] = load double, ptr [[ARGP_CUR3_ALIGNED]], align 8
+// CHECK-ILP32F-NEXT:    store double [[TMP4]], ptr [[X]], align 8
+// CHECK-ILP32F-NEXT:    call void @llvm.va_end(ptr [[VA]])
+// CHECK-ILP32F-NEXT:    [[TMP5:%.*]] = load double, ptr [[V]], align 8
+// CHECK-ILP32F-NEXT:    [[TMP6:%.*]] = load double, ptr [[X]], align 8
+// CHECK-ILP32F-NEXT:    [[ADD:%.*]] = fadd double [[TMP5]], [[TMP6]]
+// CHECK-ILP32F-NEXT:    ret double [[ADD]]
+//
+// CHECK-ILP32D-LABEL: define dso_local double @f_va_3
+// CHECK-ILP32D-SAME: (ptr noundef [[FMT:%.*]], ...) #[[ATTR0]] {
+// CHECK-ILP32D-NEXT:  entry:
+// CHECK-ILP32D-NEXT:    [[FMT_ADDR:%.*]] = alloca ptr, align 4
+// CHECK-ILP32D-NEXT:    [[VA:%.*]] = alloca ptr, align 4
+// CHECK-ILP32D-NEXT:    [[V:%.*]] = alloca double, align 8
+// CHECK-ILP32D-NEXT:    [[W:%.*]] = alloca i32, align 4
+// CHECK-ILP32D-NEXT:    [[X:%.*]] = alloca double, align 8
+// CHECK-ILP32D-NEXT:    store ptr [[FMT]], ptr [[FMT_ADDR]], align 4
+// CHECK-ILP32D-NEXT:    call void @llvm.va_start(ptr [[VA]])
+// CHECK-ILP32D-NEXT:    [[ARGP_CUR:%.*]] = load ptr, ptr [[VA]], align 4
+// CHECK-ILP32D-NEXT:    [[TMP0:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR]], i32 7
+// CHECK-ILP32D-NEXT:    [[ARGP_CUR_ALIGNED:%.*]] = call ptr @llvm.ptrmask.p0.i32(ptr [[TMP0]], i32 -8)
+// CHECK-ILP32D-NEXT:    [[ARGP_NEXT:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR_ALIGNED]], i32 8
+// CHECK-ILP32D-NEXT:    store ptr [[ARGP_NEXT]], ptr [[VA]], align 4
+// CHECK-ILP32D-NEXT:    [[TMP1:%.*]] = load double, ptr [[ARGP_CUR_ALIGNED]], align 8
+// CHECK-ILP32D-NEXT:    store double [[TMP1]], ptr [[V]], align 8
+// CHECK-ILP32D-NEXT:    [[ARGP_CUR1:%.*]] = load ptr, ptr [[VA]], align 4
+// CHECK-ILP32D-NEXT:    [[ARGP_NEXT2:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR1]], i32 4
+// CHECK-ILP32D-NEXT:    store ptr [[ARGP_NEXT2]], ptr [[VA]], align 4
+// CHECK-ILP32D-NEXT:    [[TMP2:%.*]] = load i32, ptr [[ARGP_CUR1]], align 4
+// CHECK-ILP32D-NEXT:    store i32 [[TMP2]], ptr [[W]], align 4
+// CHECK-ILP32D-NEXT:    [[ARGP_CUR3:%.*]] = load ptr, ptr [[VA]], align 4
+// CHECK-ILP32D-NEXT:    [[TMP3:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR3]], i32 7
+// CHECK-ILP32D-NEXT:    [[ARGP_CUR3_ALIGNED:%.*]] = call ptr @llvm.ptrmask.p0.i32(ptr [[TMP3]], i32 -8)
+// CHECK-ILP32D-NEXT:    [[ARGP_NEXT4:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR3_ALIGNED]], i32 8
+// CHECK-ILP32D-NEXT:    store ptr [[ARGP_NEXT4]], ptr [[VA]], align 4
+// CHECK-ILP32D-NEXT:    [[TMP4:%.*]] = load double, ptr [[ARGP_CUR3_ALIGNED]], align 8
+// CHECK-ILP32D-NEXT:    store double [[TMP4]], ptr [[X]], align 8
+// CHECK-ILP32D-NEXT:    call void @llvm.va_end(ptr [[VA]])
+// CHECK-ILP32D-NEXT:    [[TMP5:%.*]] = load double, ptr [[V]], align 8
+// CHECK-ILP32D-NEXT:    [[TMP6:%.*]] = load double, ptr [[X]], align 8
+// CHECK-ILP32D-NEXT:    [[ADD:%.*]] = fadd double [[TMP5]], [[TMP6]]
+// CHECK-ILP32D-NEXT:    ret double [[ADD]]
+//
+// CHECK-ILP32E-LABEL: define dso_local double @f_va_3
+// CHECK-ILP32E-SAME: (ptr noundef [[FMT:%.*]], ...) #[[ATTR0]] {
+// CHECK-ILP32E-NEXT:  entry:
+// CHECK-ILP32E-NEXT:    [[FMT_ADDR:%.*]] = alloca ptr, align 4
+// CHECK-ILP32E-NEXT:    [[VA:%.*]] = alloca ptr, align 4
+// CHECK-ILP32E-NEXT:    [[V:%.*]] = alloca double, align 8
+// CHECK-ILP32E-NEXT:    [[W:%.*]] = alloca i32, align 4
+// CHECK-ILP32E-NEXT:    [[X:%.*]] = alloca double, align 8
+// CHECK-ILP32E-NEXT:    store ptr [[FMT]], ptr [[FMT_ADDR]], align 4
+// CHECK-ILP32E-NEXT:    call void @llvm.va_start(ptr [[VA]])
+// CHECK-ILP32E-NEXT:    [[ARGP_CUR:%.*]] = load ptr, ptr [[VA]], align 4
+// CHECK-ILP32E-NEXT:    [[ARGP_NEXT:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR]], i32 8
+// CHECK-ILP32E-NEXT:    store ptr [[ARGP_NEXT]], ptr [[VA]], align 4
+// CHECK-ILP32E-NEXT:    [[TMP0:%.*]] = load double, ptr [[ARGP_CUR]], align 4
+// CHECK-ILP32E-NEXT:    store double [[TMP0]], ptr [[V]], align 8
+// CHECK-ILP32E-NEXT:    [[ARGP_CUR1:%.*]] = load ptr, ptr [[VA]], align 4
+// CHECK-ILP32E-NEXT:    [[ARGP_NEXT2:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR1]], i32 4
+// CHECK-ILP32E-NEXT:    store ptr [[ARGP_NEXT2]], ptr [[VA]], align 4
+// CHECK-ILP32E-NEXT:    [[TMP1:%.*]] = load i32, ptr [[ARGP_CUR1]], align 4
+// CHECK-ILP32E-NEXT:    store i32 [[TMP1]], ptr [[W]], align 4
+// CHECK-ILP32E-NEXT:    [[ARGP_CUR3:%.*]] = load ptr, ptr [[VA]], align 4
+// CHECK-ILP32E-NEXT:    [[ARGP_NEXT4:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR3]], i32 8
+// CHECK-ILP32E-NEXT:    store ptr [[ARGP_NEXT4]], ptr [[VA]], align 4
+// CHECK-ILP32E-NEXT:    [[TMP2:%.*]] = load double, ptr [[ARGP_CUR3]], align 4
+// CHECK-ILP32E-NEXT:    store double [[TMP2]], ptr [[X]], align 8
+// CHECK-ILP32E-NEXT:    call void @llvm.va_end(ptr [[VA]])
+// CHECK-ILP32E-NEXT:    [[TMP3:%.*]] = load double, ptr [[V]], align 8
+// CHECK-ILP32E-NEXT:    [[TMP4:%.*]] = load double, ptr [[X]], align 8
+// CHECK-ILP32E-NEXT:    [[ADD:%.*]] = fadd double [[TMP3]], [[TMP4]]
+// CHECK-ILP32E-NEXT:    ret double [[ADD]]
 //
 double f_va_3(char *fmt, ...) {
   __builtin_va_list va;
@@ -180,93 +284,269 @@ double f_va_3(char *fmt, ...) {
   return v + x;
 }
 
-// CHECK-LABEL: define dso_local i32 @f_va_4
-// CHECK-SAME: (ptr noundef [[FMT:%.*]], ...) #[[ATTR0]] {
-// CHECK-NEXT:  entry:
-// CHECK-NEXT:    [[FMT_ADDR:%.*]] = alloca ptr, align 4
-// CHECK-NEXT:    [[VA:%.*]] = alloca ptr, align 4
-// CHECK-NEXT:    [[V:%.*]] = alloca i32, align 4
-// CHECK-NEXT:    [[LD:%.*]] = alloca fp128, align 16
-// CHECK-NEXT:    [[TS:%.*]] = alloca [[STRUCT_TINY:%.*]], align 1
-// CHECK-NEXT:    [[SS:%.*]] = alloca [[STRUCT_SMALL:%.*]], align 4
-// CHECK-NEXT:    [[LS:%.*]] = alloca [[STRUCT_LARGE:%.*]], align 4
-// CHECK-NEXT:    [[RET:%.*]] = alloca i32, align 4
-// CHECK-NEXT:    store ptr [[FMT]], ptr [[FMT_ADDR]], align 4
-// CHECK-NEXT:    call void @llvm.va_start(ptr [[VA]])
-// CHECK-NEXT:    [[ARGP_CUR:%.*]] = load ptr, ptr [[VA]], align 4
-// CHECK-NEXT:    [[ARGP_NEXT:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR]], i32 4
-// CHECK-NEXT:    store ptr [[ARGP_NEXT]], ptr [[VA]], align 4
-// CHECK-NEXT:    [[TMP0:%.*]] = load i32, ptr [[ARGP_CUR]], align 4
-// CHECK-NEXT:    store i32 [[TMP0]], ptr [[V]], align 4
-// CHECK-NEXT:    [[ARGP_CUR1:%.*]] = load ptr, ptr [[VA]], align 4
-// CHECK-NEXT:    [[ARGP_NEXT2:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR1]], i32 4
-// CHECK-NEXT:    store ptr [[ARGP_NEXT2]], ptr [[VA]], align 4
-// CHECK-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[ARGP_CUR1]], align 4
-// CHECK-NEXT:    [[TMP2:%.*]] = load fp128, ptr [[TMP1]], align 16
-// CHECK-NEXT:    store fp128 [[TMP2]], ptr [[LD]], align 16
-// CHECK-NEXT:    [[ARGP_CUR3:%.*]] = load ptr, ptr [[VA]], align 4
-// CHECK-NEXT:    [[ARGP_NEXT4:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR3]], i32 4
-// CHECK-NEXT:    store ptr [[ARGP_NEXT4]], ptr [[VA]], align 4
-// CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[TS]], ptr align 4 [[ARGP_CUR3]], i32 4, i1 false)
-// CHECK-NEXT:    [[ARGP_CUR5:%.*]] = load ptr, ptr [[VA]], align 4
-// CHECK-NEXT:    [[ARGP_NEXT6:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR5]], i32 8
-// CHECK-NEXT:    store ptr [[ARGP_NEXT6]], ptr [[VA]], align 4
-// CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 4 [[SS]], ptr align 4 [[ARGP_CUR5]], i32 8, i1 false)
-// CHECK-NEXT:    [[ARGP_CUR7:%.*]] = load ptr, ptr [[VA]], align 4
-// CHECK-NEXT:    [[ARGP_NEXT8:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR7]], i32 4
-// CHECK-NEXT:    store ptr [[ARGP_NEXT8]], ptr [[VA]], align 4
-// CHECK-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[ARGP_CUR7]], align 4
-// CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 4 [[LS]], ptr align 4 [[TMP3]], i32 16, i1 false)
-// CHECK-NEXT:    call void @llvm.va_end(ptr [[VA]])
-// CHECK-NEXT:    [[TMP4:%.*]] = load i32, ptr [[V]], align 4
-// CHECK-NEXT:    [[CONV:%.*]] = sitofp i32 [[TMP4]] to fp128
-// CHECK-NEXT:    [[TMP5:%.*]] = load fp128, ptr [[LD]], align 16
-// CHECK-NEXT:    [[ADD:%.*]] = fadd fp128 [[CONV]], [[TMP5]]
-// CHECK-NEXT:    [[CONV9:%.*]] = fptosi fp128 [[ADD]] to i32
-// CHECK-NEXT:    store i32 [[CONV9]], ptr [[RET]], align 4
-// CHECK-NEXT:    [[TMP6:%.*]] = load i32, ptr [[RET]], align 4
-// CHECK-NEXT:    [[A:%.*]] = getelementptr inbounds [[STRUCT_TINY]], ptr [[TS]], i32 0, i32 0
-// CHECK-NEXT:    [[TMP7:%.*]] = load i8, ptr [[A]], align 1
-// CHECK-NEXT:    [[CONV10:%.*]] = zext i8 [[TMP7]] to i32
-// CHECK-NEXT:    [[ADD11:%.*]] = add nsw i32 [[TMP6]], [[CONV10]]
-// CHECK-NEXT:    [[B:%.*]] = getelementptr inbounds [[STRUCT_TINY]], ptr [[TS]], i32 0, i32 1
-// CHECK-NEXT:    [[TMP8:%.*]] = load i8, ptr [[B]], align 1
-// CHECK-NEXT:    [[CONV12:%.*]] = zext i8 [[TMP8]] to i32
-// CHECK-NEXT:    [[ADD13:%.*]] = add nsw i32 [[ADD11]], [[CONV12]]
-// CHECK-NEXT:    [[C:%.*]] = getelementptr inbounds [[STRUCT_TINY]], ptr [[TS]], i32 0, i32 2
-// CHECK-NEXT:    [[TMP9:%.*]] = load i8, ptr [[C]], align 1
-// CHECK-NEXT:    [[CONV14:%.*]] = zext i8 [[TMP9]] to i32
-// CHECK-NEXT:    [[ADD15:%.*]] = add nsw i32 [[ADD13]], [[CONV14]]
-// CHECK-NEXT:    [[D:%.*]] = getelementptr inbounds [[STRUCT_TINY]], ptr [[TS]], i32 0, i32 3
-// CHECK-NEXT:    [[TMP10:%.*]] = load i8, ptr [[D]], align 1
-// CHECK-NEXT:    [[CONV16:%.*]] = zext i8 [[TMP10]] to i32
-// CHECK-NEXT:    [[ADD17:%.*]] = add nsw i32 [[ADD15]], [[CONV16]]
-// CHECK-NEXT:    store i32 [[ADD17]], ptr [[RET]], align 4
-// CHECK-NEXT:    [[TMP11:%.*]] = load i32, ptr [[RET]], align 4
-// CHECK-NEXT:    [[A18:%.*]] = getelementptr inbounds [[STRUCT_SMALL]], ptr [[SS]], i32 0, i32 0
-// CHECK-NEXT:    [[TMP12:%.*]] = load i32, ptr [[A18]], align 4
-// CHECK-NEXT:    [[ADD19:%.*]] = add nsw i32 [[TMP11]], [[TMP12]]
-// CHECK-NEXT:    [[B20:%.*]] = getelementptr inbounds [[STRUCT_SMALL]], ptr [[SS]], i32 0, i32 1
-// CHECK-NEXT:    [[TMP13:%.*]] = load ptr, ptr [[B20]], align 4
-// CHECK-NEXT:    [[TMP14:%.*]] = ptrtoint ptr [[TMP13]] to i32
-// CHECK-NEXT:    [[ADD21:%.*]] = add nsw i32 [[ADD19]], [[TMP14]]
-// CHECK-NEXT:    store i32 [[ADD21]], ptr [[RET]], align 4
-// CHECK-NEXT:    [[TMP15:%.*]] = load i32, ptr [[RET]], align 4
-// CHECK-NEXT:    [[A22:%.*]] = getelementptr inbounds [[STRUCT_LARGE]], ptr [[LS]], i32 0, i32 0
-// CHECK-NEXT:    [[TMP16:%.*]] = load i32, ptr [[A22]], align 4
-// CHECK-NEXT:    [[ADD23:%.*]] = add nsw i32 [[TMP15]], [[TMP16]]
-// CHECK-NEXT:    [[B24:%.*]] = getelementptr inbounds [[STRUCT_LARGE]], ptr [[LS]], i32 0, i32 1
-// CHECK-NEXT:    [[TMP17:%.*]] = load i32, ptr [[B24]], align 4
-// CHECK-NEXT:    [[ADD25:%.*]] = add nsw i32 [[ADD23]], [[TMP17]]
-// CHECK-NEXT:    [[C26:%.*]] = getelementptr inbounds [[STRUCT_LARGE]], ptr [[LS]], i32 0, i32 2
-// CHECK-NEXT:    [[TMP18:%.*]] = load i32, ptr [[C26]], align 4
-// CHECK-NEXT:    [[ADD27:%.*]] = add nsw i32 [[ADD25]], [[TMP18]]
-// CHECK-NEXT:    [[D28:%.*]] = getelementptr inbounds [[STRUCT_LARGE]], ptr [[LS]], i32 0, i32 3
-// CHECK-NEXT:    [[TMP19:%.*]] = load i32, ptr [[D28]], align 4
-// CHECK-NEXT:    [[ADD29:%.*]] = add nsw i32 [[ADD27]], [[TMP19]]
-// CHECK-NEXT:    store i32 [[ADD29]], ptr [[RET]], align 4
-// CHECK-NEXT:    [[TMP20:%.*]] = load i32, ptr [[RET]], align 4
-// CHECK-NEXT:    ret i32 [[TMP20]]
+// CHECK-ILP32F-LABEL: define dso_local i32 @f_va_4
+// CHECK-ILP32F-SAME: (ptr noundef [[FMT:%.*]], ...) #[[ATTR0]] {
+// CHECK-ILP32F-NEXT:  entry:
+// CHECK-ILP32F-NEXT:    [[FMT_ADDR:%.*]] = alloca ptr, align 4
+// CHECK-ILP32F-NEXT:    [[VA:%.*]] = alloca ptr, align 4
+// CHECK-ILP32F-NEXT:    [[V:%.*]] = alloca i32, align 4
+// CHECK-ILP32F-NEXT:    [[LD:%.*]] = alloca fp128, align 16
+// CHECK-ILP32F-NEXT:    [[TS:%.*]] = alloca [[STRUCT_TINY:%.*]], align 1
+// CHECK-ILP32F-NEXT:    [[SS:%.*]] = alloca [[STRUCT_SMALL:%.*]], align 4
+// CHECK-ILP32F-NEXT:    [[LS:%.*]] = alloca [[STRUCT_LARGE:%.*]], align 4
+// CHECK-ILP32F-NEXT:    [[RET:%.*]] = alloca i32, align 4
+// CHECK-ILP32F-NEXT:    store ptr [[FMT]], ptr [[FMT_ADDR]], align 4
+// CHECK-ILP32F-NEXT:    call void @llvm.va_start(ptr [[VA]])
+// CHECK-ILP32F-NEXT:    [[ARGP_CUR:%.*]] = load ptr, ptr [[VA]], align 4
+// CHECK-ILP32F-NEXT:    [[ARGP_NEXT:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR]], i32 4
+// CHECK-ILP32F-NEXT:    store ptr [[ARGP_NEXT]], ptr [[VA]], align 4
+// CHECK-ILP32F-NEXT:    [[TMP0:%.*]] = load i32, ptr [[ARGP_CUR]], align 4
+// CHECK-ILP32F-NEXT:    store i32 [[TMP0]], ptr [[V]], align 4
+// CHECK-ILP32F-NEXT:    [[ARGP_CUR1:%.*]] = load ptr, ptr [[VA]], align 4
+// CHECK-ILP32F-NEXT:    [[ARGP_NEXT2:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR1]], i32 4
+// CHECK-ILP32F-NEXT:    store ptr [[ARGP_NEXT2]], ptr [[VA]], align 4
+// CHECK-ILP32F-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[ARGP_CUR1]], align 4
+// CHECK-ILP32F-NEXT:    [[TMP2:%.*]] = load fp128, ptr [[TMP1]], align 16
+// CHECK-ILP32F-NEXT:    store fp128 [[TMP2]], ptr [[LD]], align 16
+// CHECK-ILP32F-NEXT:    [[ARGP_CUR3:%.*]] = load ptr, ptr [[VA]], align 4
+// CHECK-ILP32F-NEXT:    [[ARGP_NEXT4:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR3]], i32 4
+// CHECK-ILP32F-NEXT:    store ptr [[ARGP_NEXT4]], ptr [[VA]], align 4
+// CHECK-ILP32F-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[TS]], ptr align 4 [[ARGP_CUR3]], i32 4, i1 false)
+// CHECK-ILP32F-NEXT:    [[ARGP_CUR5:%.*]] = load ptr, ptr [[VA]], align 4
+// CHECK-ILP32F-NEXT:    [[ARGP_NEXT6:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR5]], i32 8
+// CHECK-ILP32F-NEXT:    store ptr [[ARGP_NEXT6]], ptr [[VA]], align 4
+// CHECK-ILP32F-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 4 [[SS]], ptr align 4 [[ARGP_CUR5]], i32 8, i1 false)
+// CHECK-ILP32F-NEXT:    [[ARGP_CUR7:%.*]] = load ptr, ptr [[VA]], align 4
+// CHECK-ILP32F-NEXT:    [[ARGP_NEXT8:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR7]], i32 4
+// CHECK-ILP32F-NEXT:    store ptr [[ARGP_NEXT8]], ptr [[VA]], align 4
+// CHECK-ILP32F-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[ARGP_CUR7]], align 4
+// CHECK-ILP32F-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 4 [[LS]], ptr align 4 [[TMP3]], i32 16, i1 false)
+// CHECK-ILP32F-NEXT:    call void @llvm.va_end(ptr [[VA]])
+// CHECK-ILP32F-NEXT:    [[TMP4:%.*]] = load i32, ptr [[V]], align 4
+// CHECK-ILP32F-NEXT:    [[CONV:%.*]] = sitofp i32 [[TMP4]] to fp128
+// CHECK-ILP32F-NEXT:    [[TMP5:%.*]] = load fp128, ptr [[LD]], align 16
+// CHECK-ILP32F-NEXT:    [[ADD:%.*]] = fadd fp128 [[CONV]], [[TMP5]]
+// CHECK-ILP32F-NEXT:    [[CONV9:%.*]] = fptosi fp128 [[ADD]] to i32
+// CHECK-ILP32F-NEXT:    store i32 [[CONV9]], ptr [[RET]], align 4
+// CHECK-ILP32F-NEXT:    [[TMP6:%.*]] = load i32, ptr [[RET]], align 4
+// CHECK-ILP32F-NEXT:    [[A:%.*]] = getelementptr inbounds [[STRUCT_TINY]], ptr [[TS]], i32 0, i32 0
+// CHECK-ILP32F-NEXT:    [[TMP7:%.*]] = load i8, ptr [[A]], align 1
+// CHECK-ILP32F-NEXT:    [[CONV10:%.*]] = zext i8 [[TMP7]] to i32
+// CHECK-ILP32F-NEXT:    [[ADD11:%.*]] = add nsw i32 [[TMP6]], [[CONV10]]
+// CHECK-ILP32F-NEXT:    [[B:%.*]] = getelementptr inbounds [[STRUCT_TINY]], ptr [[TS]], i32 0, i32 1
+// CHECK-ILP32F-NEXT:    [[TMP8:%.*]] = load i8, ptr [[B]], align 1
+// CHECK-ILP32F-NEXT:    [[CONV12:%.*]] = zext i8 [[TMP8]] to i32
+// CHECK-ILP32F-NEXT:    [[ADD13:%.*]] = add nsw i32 [[ADD11]], [[CONV12]]
+// CHECK-ILP32F-NEXT:    [[C:%.*]] = getelementptr inbounds [[STRUCT_TINY]], ptr [[TS]], i32 0, i32 2
+// CHECK-ILP32F-NEXT:    [[TMP9:%.*]] = load i8, ptr [[C]], align 1
+// CHECK-ILP32F-NEXT:    [[CONV14:%.*]] = zext i8 [[TMP9]] to i32
+// CHECK-ILP32F-NEXT:    [[ADD15:%.*]] = add nsw i32 [[ADD13]], [[CONV14]]
+// CHECK-ILP32F-NEXT:    [[D:%.*]] = getelementptr inbounds [[STRUCT_TINY]], ptr [[TS]], i32 0, i32 3
+// CHECK-ILP32F-NEXT:    [[TMP10:%.*]] = load i8, ptr [[D]], align 1
+// CHECK-ILP32F-NEXT:    [[CONV16:%.*]] = zext i8 [[TMP10]] to i32
+// CHECK-ILP32F-NEXT:    [[ADD17:%.*]] = add nsw i32 [[ADD15]], [[CONV16]]
+// CHECK-ILP32F-NEXT:    store i32 [[ADD17]], ptr [[RET]], align 4
+// CHECK-ILP32F-NEXT:    [[TMP11:%.*]] = load i32, ptr [[RET]], align 4
+// CHECK-ILP32F-NEXT:    [[A18:%.*]] = getelementptr inbounds [[STRUCT_SMALL]], ptr [[SS]], i32 0, i32 0
+// CHECK-ILP32F-NEXT:    [[TMP12:%.*]] = load i32, ptr [[A18]], align 4
+// CHECK-ILP32F-NEXT:    [[ADD19:%.*]] = add nsw i32 [[TMP11]], [[TMP12]]
+// CHECK-ILP32F-NEXT:    [[B20:%.*]] = getelementptr inbounds [[STRUCT_SMALL]], ptr [[SS]], i32 0, i32 1
+// CHECK-ILP32F-NEXT:    [[TMP13:%.*]] = load ptr, ptr [[B20]], align 4
+// CHECK-ILP32F-NEXT:    [[TMP14:%.*]] = ptrtoint ptr [[TMP13]] to i32
+// CHECK-ILP32F-NEXT:    [[ADD21:%.*]] = add nsw i32 [[ADD19]], [[TMP14]]
+// CHECK-ILP32F-NEXT:    store i32 [[ADD21]], ptr [[RET]], align 4
+// CHECK-ILP32F-NEXT:    [[TMP15:%.*]] = load i32, ptr [[RET]], align 4
+// CHECK-ILP32F-NEXT:    [[A22:%.*]] = getelementptr inbounds [[STRUCT_LARGE]], ptr [[LS]], i32 0, i32 0
+// CHECK-ILP32F-NEXT:    [[TMP16:%.*]] = load i32, ptr [[A22]], align 4
+// CHECK-ILP32F-NEXT:    [[ADD23:%.*]] = add nsw i32 [[TMP15]], [[TMP16]]
+// CHECK-ILP32F-NEXT:    [[B24:%.*]] = getelementptr inbounds [[STRUCT_LARGE]], ptr [[LS]], i32 0, i32 1
+// CHECK-ILP32F-NEXT:    [[TMP17:%.*]] = load i32, ptr [[B24]], align 4
+// CHECK-ILP32F-NEXT:    [[ADD25:%.*]] = add nsw i32 [[ADD23]], [[TMP17]]
+// CHECK-ILP32F-NEXT:    [[C26:%.*]] = getelementptr inbounds [[STRUCT_LARGE]], ptr [[LS]], i32 0, i32 2
+// CHECK-ILP32F-NEXT:    [[TMP18:%.*]] = load i32, ptr [[C26]], align 4
+// CHECK-ILP32F-NEXT:    [[ADD27:%.*]] = add nsw i32 [[ADD25]], [[TMP18]]
+// CHECK-ILP32F-NEXT:    [[D28:%.*]] = getelementptr inbounds [[STRUCT_LARGE]], ptr [[LS]], i32 0, i32 3
+// CHECK-ILP32F-NEXT:    [[TMP19:%.*]] = load i32, ptr [[D28]], align 4
+// CHECK-ILP32F-NEXT:    [[ADD29:%.*]] = add nsw i32 [[ADD27]], [[TMP19]]
+// CHECK-ILP32F-NEXT:    store i32 [[ADD29]], ptr [[RET]], align 4
+// CHECK-ILP32F-NEXT:    [[TMP20:%.*]] = load i32, ptr [[RET]], align 4
+// CHECK-ILP32F-NEXT:    ret i32 [[TMP20]]
+//
+// CHECK-ILP32D-LABEL: define dso_local i32 @f_va_4
+// CHECK-ILP32D-SAME: (ptr noundef [[FMT:%.*]], ...) #[[ATTR0]] {
+// CHECK-ILP32D-NEXT:  entry:
+// CHECK-ILP32D-NEXT:    [[FMT_ADDR:%.*]] = alloca ptr, align 4
+// CHECK-ILP32D-NEXT:    [[VA:%.*]] = alloca ptr, align 4
+// CHECK-ILP32D-NEXT:    [[V:%.*]] = alloca i32, align 4
+// CHECK-ILP32D-NEXT:    [[LD:%.*]] = alloca fp128, align 16
+// CHECK-ILP32D-NEXT:    [[TS:%.*]] = alloca [[STRUCT_TINY:%.*]], align 1
+// CHECK-ILP32D-NEXT:    [[SS:%.*]] = alloca [[STRUCT_SMALL:%.*]], align 4
+// CHECK-ILP32D-NEXT:    [[LS:%.*]] = alloca [[STRUCT_LARGE:%.*]], align 4
+// CHECK-ILP32D-NEXT:    [[RET:%.*]] = alloca i32, align 4
+// CHECK-ILP32D-NEXT:    store ptr [[FMT]], ptr [[FMT_ADDR]], align 4
+// CHECK-ILP32D-NEXT:    call void @llvm.va_start(ptr [[VA]])
+// CHECK-ILP32D-NEXT:    [[ARGP_CUR:%.*]] = load ptr, ptr [[VA]], align 4
+// CHECK-ILP32D-NEXT:    [[ARGP_NEXT:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR]], i32 4
+// CHECK-ILP32D-NEXT:    store ptr [[ARGP_NEXT]], ptr [[VA]], align 4
+// CHECK-ILP32D-NEXT:    [[TMP0:%.*]] = load i32, ptr [[ARGP_CUR]], align 4
+// CHECK-ILP32D-NEXT:    store i32 [[TMP0]], ptr [[V]], align 4
+// CHECK-ILP32D-NEXT:    [[ARGP_CUR1:%.*]] = load ptr, ptr [[VA]], align 4
+// CHECK-ILP32D-NEXT:    [[ARGP_NEXT2:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR1]], i32 4
+// CHECK-ILP32D-NEXT:    store ptr [[ARGP_NEXT2]], ptr [[VA]], align 4
+// CHECK-ILP32D-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[ARGP_CUR1]], align 4
+// CHECK-ILP32D-NEXT:    [[TMP2:%.*]] = load fp128, ptr [[TMP1]], align 16
+// CHECK-ILP32D-NEXT:    store fp128 [[TMP2]], ptr [[LD]], align 16
+// CHECK-ILP32D-NEXT:    [[ARGP_CUR3:%.*]] = load ptr, ptr [[VA]], align 4
+// CHECK-ILP32D-NEXT:    [[ARGP_NEXT4:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR3]], i32 4
+// CHECK-ILP32D-NEXT:    store ptr [[ARGP_NEXT4]], ptr [[VA]], align 4
+// CHECK-ILP32D-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[TS]], ptr align 4 [[ARGP_CUR3]], i32 4, i1 false)
+// CHECK-ILP32D-NEXT:    [[ARGP_CUR5:%.*]] = load ptr, ptr [[VA]], align 4
+// CHECK-ILP32D-NEXT:    [[ARGP_NEXT6:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR5]], i32 8
+// CHECK-ILP32D-NEXT:    store ptr [[ARGP_NEXT6]], ptr [[VA]], align 4
+// CHECK-ILP32D-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 4 [[SS]], ptr align 4 [[ARGP_CUR5]], i32 8, i1 false)
+// CHECK-ILP32D-NEXT:    [[ARGP_CUR7:%.*]] = load ptr, ptr [[VA]], align 4
+// CHECK-ILP32D-NEXT:    [[ARGP_NEXT8:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR7]], i32 4
+// CHECK-ILP32D-NEXT:    store ptr [[ARGP_NEXT8]], ptr [[VA]], align 4
+// CHECK-ILP32D-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[ARGP_CUR7]], align 4
+// CHECK-ILP32D-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 4 [[LS]], ptr align 4 [[TMP3]], i32 16, i1 false)
+// CHECK-ILP32D-NEXT:    call void @llvm.va_end(ptr [[VA]])
+// CHECK-ILP32D-NEXT:    [[TMP4:%.*]] = load i32, ptr [[V]], align 4
+// CHECK-ILP32D-NEXT:    [[CONV:%.*]] = sitofp i32 [[TMP4]] to fp128
+// CHECK-ILP32D-NEXT:    [[TMP5:%.*]] = load fp128, ptr [[LD]], align 16
+// CHECK-ILP32D-NEXT:    [[ADD:%.*]] = fadd fp128 [[CONV]], [[TMP5]]
+// CHECK-ILP32D-NEXT:    [[CONV9:%.*]] = fptosi fp128 [[ADD]] to i32
+// CHECK-ILP32D-NEXT:    store i32 [[CONV9]], ptr [[RET]], align 4
+// CHECK-ILP32D-NEXT:    [[TMP6:%.*]] = load i32, ptr [[RET]], align 4
+// CHECK-ILP32D-NEXT:    [[A:%.*]] = getelementptr inbounds [[STRUCT_TINY]], ptr [[TS]], i32 0, i32 0
+// CHECK-ILP32D-NEXT:    [[TMP7:%.*]] = load i8, ptr [[A]], align 1
+// CHECK-ILP32D-NEXT:    [[CONV10:%.*]] = zext i8 [[TMP7]] to i32
+// CHECK-ILP32D-NEXT:    [[ADD11:%.*]] = add nsw i32 [[TMP6]], [[CONV10]]
+// CHECK-ILP32D-NEXT:    [[B:%.*]] = getelementptr inbounds [[STRUCT_TINY]], ptr [[TS]], i32 0, i32 1
+// CHECK-ILP32D-NEXT:    [[TMP8:%.*]] = load i8, ptr [[B]], align 1
+// CHECK-ILP32D-NEXT:    [[CONV12:%.*]] = zext i8 [[TMP8]] to i32
+// CHECK-ILP32D-NEXT:    [[ADD13:%.*]] = add nsw i32 [[ADD11]], [[CONV12]]
+// CHECK-ILP32D-NEXT:    [[C:%.*]] = getelementptr inbounds [[STRUCT_TINY]], ptr [[TS]], i32 0, i32 2
+// CHECK-ILP32D-NEXT:    [[TMP9:%.*]] = load i8, ptr [[C]], align 1
+// CHECK-ILP32D-NEXT:    [[CONV14:%.*]] = zext i8 [[TMP9]] to i32
+// CHECK-ILP32D-NEXT:    [[ADD15:%.*]] = add nsw i32 [[ADD13]], [[CONV14]]
+// CHECK-ILP32D-NEXT:    [[D:%.*]] = getelementptr inbounds [[STRUCT_TINY]], ptr [[TS]], i32 0, i32 3
+// CHECK-ILP32D-NEXT:    [[TMP10:%.*]] = load i8, ptr [[D]], align 1
+// CHECK-ILP32D-NEXT:    [[CONV16:%.*]] = zext i8 [[TMP10]] to i32
+// CHECK-ILP32D-NEXT:    [[ADD17:%.*]] = add nsw i32 [[ADD15]], [[CONV16]]
+// CHECK-ILP32D-NEXT:    store i32 [[ADD17]], ptr [[RET]], align 4
+// CHECK-ILP32D-NEXT:    [[TMP11:%.*]] = load i32, ptr [[RET]], align 4
+// CHECK-ILP32D-NEXT:    [[A18:%.*]] = getelementptr inbounds [[STRUCT_SMALL]], ptr [[SS]], i32 0, i32 0
+// CHECK-ILP32D-NEXT:    [[TMP12:%.*]] = load i32, ptr [[A18]], align 4
+// CHECK-ILP32D-NEXT:    [[ADD19:%.*]] = add nsw i32 [[TMP11]], [[TMP12]]
+// CHECK-ILP32D-NEXT:    [[B20:%.*]] = getelementptr inbounds [[STRUCT_SMALL]], ptr [[SS]], i32 0, i32 1
+// CHECK-ILP32D-NEXT:    [[TMP13:%.*]] = load ptr, ptr [[B20]], align 4
+// CHECK-ILP32D-NEXT:    [[TMP14:%.*]] = ptrtoint ptr [[TMP13]] to i32
+// CHECK-ILP32D-NEXT:    [[ADD21:%.*]] = add nsw i32 [[ADD19]], [[TMP14]]
+// CHECK-ILP32D-NEXT:    store i32 [[ADD21]], ptr [[RET]], align 4
+// CHECK-ILP32D-NEXT:    [[TMP15:%.*]] = load i32, ptr [[RET]], align 4
+// CHECK-ILP32D-NEXT:    [[A22:%.*]] = getelementptr inbounds [[STRUCT_LARGE]], ptr [[LS]], i32 0, i32 0
+// CHECK-ILP32D-NEXT:    [[TMP16:%.*]] = load i32, ptr [[A22]], align 4
+// CHECK-ILP32D-NEXT:    [[ADD23:%.*]] = add nsw i32 [[TMP15]], [[TMP16]]
+// CHECK-ILP32D-NEXT:    [[B24:%.*]] = getelementptr inbounds [[STRUCT_LARGE]], ptr [[LS]], i32 0, i32 1
+// CHECK-ILP32D-NEXT:    [[TMP17:%.*]] = load i32, ptr [[B24]], align 4
+// CHECK-ILP32D-NEXT:    [[ADD25:%.*]] = add nsw i32 [[ADD23]], [[TMP17]]
+// CHECK-ILP32D-NEXT:    [[C26:%.*]] = getelementptr inbounds [[STRUCT_LARGE]], ptr [[LS]], i32 0, i32 2
+// CHECK-ILP32D-NEXT:    [[TMP18:%.*]] = load i32, ptr [[C26]], align 4
+// CHECK-ILP32D-NEXT:    [[ADD27:%.*]] = add nsw i32 [[ADD25]], [[TMP18]]
+// CHECK-ILP32D-NEXT:    [[D28:%.*]] = getelementptr inbounds [[STRUCT_LARGE]], ptr [[LS]], i32 0, i32 3
+// CHECK-ILP32D-NEXT:    [[TMP19:%.*]] = load i32, ptr [[D28]], align 4
+// CHECK-ILP32D-NEXT:    [[ADD29:%.*]] = add nsw i32 [[ADD27]], [[TMP19]]
+// CHECK-ILP32D-NEXT:    store i32 [[ADD29]], ptr [[RET]], align 4
+// CHECK-ILP32D-NEXT:    [[TMP20:%.*]] = load i32, ptr [[RET]], align 4
+// CHECK-ILP32D-NEXT:    ret i32 [[TMP20]]
+//
+// CHECK-ILP32E-LABEL: define dso_local i32 @f_va_4
+// CHECK-ILP32E-SAME: (ptr noundef [[FMT:%.*]], ...) #[[ATTR0]] {
+// CHECK-ILP32E-NEXT:  entry:
+// CHECK-ILP32E-NEXT:    [[FMT_ADDR:%.*]] = alloca ptr, align 4
+// CHECK-ILP32E-NEXT:    [[VA:%.*]] = alloca ptr, align 4
+// CHECK-ILP32E-NEXT:    [[V:%.*]] = alloca i32, align 4
+// CHECK-ILP32E-NEXT:    [[LD:%.*]] = alloca fp128, align 16
+// CHECK-ILP32E-NEXT:    [[TS:%.*]] = alloca [[STRUCT_TINY:%.*]], align 1
+// CHECK-ILP32E-NEXT:    [[SS:%.*]] = alloca [[STRUCT_SMALL:%.*]], align 4
+// CHECK-ILP32E-NEXT:    [[LS:%.*]] = alloca [[STRUCT_LARGE:%.*]], align 4
+// CHECK-ILP32E-NEXT:    [[RET:%.*]] = alloca i32, align 4
+// CHECK-ILP32E-NEXT:    store ptr [[FMT]], ptr [[FMT_ADDR]], align 4
+// CHECK-ILP32E-NEXT:    call void @llvm.va_start(ptr [[VA]])
+// CHECK-ILP32E-NEXT:    [[ARGP_CUR:%.*]] = load ptr, ptr [[VA]], align 4
+// CHECK-ILP32E-NEXT:    [[ARGP_NEXT:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR]], i32 4
+// CHECK-ILP32E-NEXT:    store ptr [[ARGP_NEXT]], ptr [[VA]], align 4
+// CHECK-ILP32E-NEXT:    [[TMP0:%.*]] = load i32, ptr [[ARGP_CUR]], align 4
+// CHECK-ILP32E-NEXT:    store i32 [[TMP0]], ptr [[V]], align 4
+// CHECK-ILP32E-NEXT:    [[ARGP_CUR1:%.*]] = load ptr, ptr [[VA]], align 4
+// CHECK-ILP32E-NEXT:    [[ARGP_NEXT2:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR1]], i32 4
+// CHECK-ILP32E-NEXT:    store ptr [[ARGP_NEXT2]], ptr [[VA]], align 4
+// CHECK-ILP32E-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[ARGP_CUR1]], align 4
+// CHECK-ILP32E-NEXT:    [[TMP2:%.*]] = load fp128, ptr [[TMP1]], align 4
+// CHECK-ILP32E-NEXT:    store fp128 [[TMP2]], ptr [[LD]], align 16
+// CHECK-ILP32E-NEXT:    [[ARGP_CUR3:%.*]] = load ptr, ptr [[VA]], align 4
+// CHECK-ILP32E-NEXT:    [[ARGP_NEXT4:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR3]], i32 4
+// CHECK-ILP32E-NEXT:    store ptr [[ARGP_NEXT4]], ptr [[VA]], align 4
+// CHECK-ILP32E-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 1 [[TS]], ptr align 4 [[ARGP_CUR3]], i32 4, i1 false)
+// CHECK-ILP32E-NEXT:    [[ARGP_CUR5:%.*]] = load ptr, ptr [[VA]], align 4
+// CHECK-ILP32E-NEXT:    [[ARGP_NEXT6:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR5]], i32 8
+// CHECK-ILP32E-NEXT:    store ptr [[ARGP_NEXT6]], ptr [[VA]], align 4
+// CHECK-ILP32E-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 4 [[SS]], ptr align 4 [[ARGP_CUR5]], i32 8, i1 false)
+// CHECK-ILP32E-NEXT:    [[ARGP_CUR7:%.*]] = load ptr, ptr [[VA]], align 4
+// CHECK-ILP32E-NEXT:    [[ARGP_NEXT8:%.*]] = getelementptr inbounds i8, ptr [[ARGP_CUR7]], i32 4
+// CHECK-ILP32E-NEXT:    store ptr [[ARGP_NEXT8]], ptr [[VA]], align 4
+// CHECK-ILP32E-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[ARGP_CUR7]], align 4
+// CHECK-ILP32E-NEXT:    call void @llvm.memcpy.p0.p0.i32(ptr align 4 [[LS]], ptr align 4 [[TMP3]], i32 16, i1 false)
+// CHECK-ILP32E-NEXT:    call void @llvm.va_end(ptr [[VA]])
+// CHECK-ILP32E-NEXT:    [[TMP4:%.*]] = load i32, ptr [[V]], align 4
+// CHECK-ILP32E-NEXT:    [[CONV:%.*]] = sitofp i32 [[TMP4]] to fp128
+// CHECK-ILP32E-NEXT:    [[TMP5:%.*]] = load fp128, ptr [[LD]], align 16
+// CHECK-ILP32E-NEXT:    [[ADD:%.*]] = fadd fp128 [[CONV]], [[TMP5]]
+// CHECK-ILP32E-NEXT:    [[CONV9:%.*]] = fptosi fp128 [[ADD]] to i32
+// CHECK-ILP32E-NEXT:    store i32 [[CONV9]], ptr [[RET]], align 4
+// CHECK-ILP32E-NEXT:    [[TMP6:%.*]] = load i32, ptr [[RET]], align 4
+// CHECK-ILP32E-NEXT:    [[A:%.*]] = getelementptr inbounds [[STRUCT_TINY]], ptr [[TS]], i32 0, i32 0
+// CHECK-ILP32E-NEXT:    [[TMP7:%.*]] = load i8, ptr [[A]], align 1
+// CHECK-ILP32E-NEXT:    [[CONV10:%.*]] = zext i8 [[TMP7]] to i32
+// CHECK-ILP32E-NEXT:    [[ADD11:%.*]] = add nsw i32 [[TMP6]], [[CONV10]]
+// CHECK-ILP32E-NEXT:    [[B:%.*]] = getelementptr inbounds [[STRUCT_TINY]], ptr [[TS]], i32 0, i32 1
+// CHECK-ILP32E-NEXT:    [[TMP8:%.*]] = load i8, ptr [[B]], align 1
+// CHECK-ILP32E-NEXT:    [[CONV12:%.*]] = zext i8 [[TMP8]] to i32
+// CHECK-ILP32E-NEXT:    [[ADD13:%.*]] = add nsw i32 [[ADD11]], [[CONV12]]
+// CHECK-ILP32E-NEXT:    [[C:%.*]] = getelementptr inbounds [[STRUCT_TINY]], ptr [[TS]], i32 0, i32 2
+// CHECK-ILP32E-NEXT:    [[TMP9:%.*]] = load i8, ptr [[C]], align 1
+// CHECK-ILP32E-NEXT:    [[CONV14:%.*]] = zext i8 [[TMP9]] to i32
+// CHECK-ILP32E-NEXT:    [[ADD15:%.*]] = add nsw i32 [[ADD13]], [[CONV14]]
+// CHECK-ILP32E-NEXT:    [[D:%.*]] = getelementptr inbounds [[STRUCT_TINY]], ptr [[TS]], i32 0, i32 3
+// CHECK-ILP32E-NEXT:    [[TMP10:%.*]] = load i8, ptr [[D]], align 1
+// CHECK-ILP32E-NEXT:    [[CONV16:%.*]] = zext i8 [[TMP10]] to i32
+// CHECK-ILP32E-NEXT:    [[ADD17:%.*]] = add nsw i32 [[ADD15]], [[CONV16]]
+// CHECK-ILP32E-NEXT:    store i32 [[ADD17]], ptr [[RET]], align 4
+// CHECK-ILP32E-NEXT:    [[TMP11:%.*]] = load i32, ptr [[RET]], align 4
+// CHECK-ILP32E-NEXT:    [[A18:%.*]] = getelementptr inbounds [[STRUCT_SMALL]], ptr [[SS]], i32 0, i32 0
+// CHECK-ILP32E-NEXT:    [[TMP12:%.*]] = load i32, ptr [[A18]], align 4
+// CHECK-ILP32E-NEXT:    [[ADD19:%.*]] = add nsw i32 [[TMP11]], [[TMP12]]
+// CHECK-ILP32E-NEXT:    [[B20:%.*]] = getelementptr inbounds [[STRUCT_SMALL]], ptr [[SS]], i32 0, i32 1
+// CHECK-ILP32E-NEXT:    [[TMP13:%.*]] = load ptr, ptr [[B20]], align 4
+// CHECK-ILP32E-NEXT:    [[TMP14:%.*]] = ptrtoint ptr [[TMP13]] to i32
+// CHECK-ILP32E-NEXT:    [[ADD21:%.*]] = add nsw i32 [[ADD19]], [[TMP14]]
+// CHECK-ILP32E-NEXT:    store i32 [[ADD21]], ptr [[RET]], align 4
+// CHECK-ILP32E-NEXT:    [[TMP15:%.*]] = load i32, ptr [[RET]], align 4
+// CHECK-ILP32E-NEXT:    [[A22:%.*]] = getelementptr inbounds [[STRUCT_LARGE]], ptr [[LS]], i32 0, i32 0
+// CHECK-ILP32E-NEXT:    [[TMP16:%.*]] = load i32, ptr [[A22]], align 4
+// CHECK-ILP32E-NEXT:    [[ADD23:%.*]] = add nsw i32 [[TMP15]], [[TMP16]]
+// CHECK-ILP32E-NEXT:    [[B24:%.*]] = getelementptr inbounds [[STRUCT_LARGE]], ptr [[LS]], i32 0, i32 1
+// CHECK-ILP32E-NEXT:    [[TMP17:%.*]] = load i32, ptr [[B24]], align 4
+// CHECK-ILP32E-NEXT:    [[ADD25:%.*]] = add nsw i32 [[ADD23]], [[TMP17]]
+// CHECK-ILP32E-NEXT:    [[C26:%.*]] = getelementptr inbounds [[STRUCT_LARGE]], ptr [[LS]], i32 0, i32 2
+// CHECK-ILP32E-NEXT:    [[TMP18:%.*]] = load i32, ptr [[C26]], align 4
+// CHECK-ILP32E-NEXT:    [[ADD27:%.*]] = add nsw i32 [[ADD25]], [[TMP18]]
+// CHECK-ILP32E-NEXT:    [[D28:%.*]] = getelementptr inbounds [[STRUCT_LARGE]], ptr [[LS]], i32 0, i32 3
+// CHECK-ILP32E-NEXT:    [[TMP19:%.*]] = load i32, ptr [[D28]], align 4
+// CHECK-ILP32E-NEXT:    [[ADD29:%.*]] = add nsw i32 [[ADD27]], [[TMP19]]
+// CHECK-ILP32E-NEXT:    store i32 [[ADD29]], ptr [[RET]], align 4
+// CHECK-ILP32E-NEXT:    [[TMP20:%.*]] = load i32, ptr [[RET]], align 4
+// CHECK-ILP32E-NEXT:    ret i32 [[TMP20]]
 //
 int f_va_4(char *fmt, ...) {
   __builtin_va_list va;
diff --git a/clang/test/CodeGen/RISCV/riscv64-abi.c b/clang/test/CodeGen/RISCV/riscv64-abi.c
index 8c857f86ddfff..021565238904e 100644
--- a/clang/test/CodeGen/RISCV/riscv64-abi.c
+++ b/clang/test/CodeGen/RISCV/riscv64-abi.c
@@ -5,6 +5,8 @@
 // RUN:   | FileCheck -check-prefixes=LP64-LP64F-LP64D,LP64F-LP64D,LP64-LP64F,LP64F %s
 // RUN: %clang_cc1 -triple riscv64 -target-feature +f -target-feature +d -target-abi lp64d -emit-llvm %s -o - \
 // RUN:   | FileCheck -check-prefixes=LP64-LP64F-LP64D,LP64F-LP64D,LP64D %s
+// RUN: %clang_cc1 -triple riscv64 -emit-llvm -target-abi lp64e %s -o - \
+// RUN:     | FileCheck -check-prefixes=LP64-LP64F-LP64D,LP64-LP64F,LP64,LP64E %s
 
 #include 
 #include 
@@ -250,7 +252,7 @@ void f_agg_large(struct large x) {
 // The address where the struct should be written to will be the first
 // argument
 // LP64-LP64F-LP64D-LABEL: define dso_local void @f_agg_large_ret
-// LP64-LP64F-LP64D-SAME: (ptr noalias sret([[STRUCT_LARGE:%.*]]) align 8 [[AGG_RESULT:%.*]], i32 noundef signext [[I:%.*]], i8 noundef signext [[J:%.*]]) #[[ATTR0]] {
+// LP64-LP64F-LP64D-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_LARGE:%.*]]) align 8 [[AGG_RESULT:%.*]], i32 noundef signext [[I:%.*]], i8 noundef signext [[J:%.*]]) #[[ATTR0]] {
 // LP64-LP64F-LP64D:  entry:
 //
 struct large f_agg_large_ret(int32_t i, int8_t j) {
@@ -268,7 +270,7 @@ void f_vec_large_v32i8(v32i8 x) {
 }
 
 // LP64-LP64F-LP64D-LABEL: define dso_local void @f_vec_large_v32i8_ret
-// LP64-LP64F-LP64D-SAME: (ptr noalias sret(<32 x i8>) align 32 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
+// LP64-LP64F-LP64D-SAME: (ptr dead_on_unwind noalias writable sret(<32 x i8>) align 32 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
 // LP64-LP64F-LP64D:  entry:
 //
 v32i8 f_vec_large_v32i8_ret(void) {
@@ -310,7 +312,7 @@ int f_scalar_stack_3(int32_t a, __int128_t b, double c, long double d, v32i8 e,
 // to pass a pointer.
 
 // LP64-LP64F-LP64D-LABEL: define dso_local void @f_scalar_stack_4
-// LP64-LP64F-LP64D-SAME: (ptr noalias sret([[STRUCT_LARGE:%.*]]) align 8 [[AGG_RESULT:%.*]], i32 noundef signext [[A:%.*]], i128 noundef [[B:%.*]], fp128 noundef [[C:%.*]], ptr noundef [[TMP0:%.*]], i8 noundef zeroext [[E:%.*]], i8 noundef signext [[F:%.*]], i8 noundef zeroext [[G:%.*]]) #[[ATTR0]] {
+// LP64-LP64F-LP64D-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_LARGE:%.*]]) align 8 [[AGG_RESULT:%.*]], i32 noundef signext [[A:%.*]], i128 noundef [[B:%.*]], fp128 noundef [[C:%.*]], ptr noundef [[TMP0:%.*]], i8 noundef zeroext [[E:%.*]], i8 noundef signext [[F:%.*]], i8 noundef zeroext [[G:%.*]]) #[[ATTR0]] {
 // LP64-LP64F-LP64D:  entry:
 //
 struct large f_scalar_stack_4(uint32_t a, __int128_t b, long double c, v32i8 d,
@@ -319,7 +321,7 @@ struct large f_scalar_stack_4(uint32_t a, __int128_t b, long double c, v32i8 d,
 }
 
 // LP64-LP64F-LP64D-LABEL: define dso_local void @f_scalar_stack_5
-// LP64-LP64F-LP64D-SAME: (ptr noalias sret([[STRUCT_LARGE:%.*]]) align 8 [[AGG_RESULT:%.*]], double noundef [[A:%.*]], i128 noundef [[B:%.*]], fp128 noundef [[C:%.*]], ptr noundef [[TMP0:%.*]], i8 noundef zeroext [[E:%.*]], i8 noundef signext [[F:%.*]], i8 noundef zeroext [[G:%.*]]) #[[ATTR0]] {
+// LP64-LP64F-LP64D-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_LARGE:%.*]]) align 8 [[AGG_RESULT:%.*]], double noundef [[A:%.*]], i128 noundef [[B:%.*]], fp128 noundef [[C:%.*]], ptr noundef [[TMP0:%.*]], i8 noundef zeroext [[E:%.*]], i8 noundef signext [[F:%.*]], i8 noundef zeroext [[G:%.*]]) #[[ATTR0]] {
 // LP64-LP64F-LP64D:  entry:
 //
 struct large f_scalar_stack_5(double a, __int128_t b, long double c, v32i8 d,
@@ -1444,7 +1446,7 @@ struct int_double_int_s { int a; double b; int c; };
 void f_int_double_int_s_arg(struct int_double_int_s a) {}
 
 // LP64-LP64F-LP64D-LABEL: define dso_local void @f_ret_int_double_int_s
-// LP64-LP64F-LP64D-SAME: (ptr noalias sret([[STRUCT_INT_DOUBLE_INT_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
+// LP64-LP64F-LP64D-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_INT_DOUBLE_INT_S:%.*]]) align 8 [[AGG_RESULT:%.*]]) #[[ATTR0]] {
 // LP64-LP64F-LP64D:  entry:
 //
 struct int_double_int_s f_ret_int_double_int_s(void) {
@@ -2046,3 +2048,5 @@ union float16_u f_ret_float16_u(void) {
   return (union float16_u){1.0};
 }
 
+//// NOTE: These prefixes are unused and the list is autogenerated. Do not add tests below this line:
+// LP64E: {{.*}}
diff --git a/clang/test/CodeGen/RISCV/riscv64-vararg.c b/clang/test/CodeGen/RISCV/riscv64-vararg.c
index 26261f0095ac9..634cde61320cb 100644
--- a/clang/test/CodeGen/RISCV/riscv64-vararg.c
+++ b/clang/test/CodeGen/RISCV/riscv64-vararg.c
@@ -4,6 +4,8 @@
 // RUN:     | FileCheck %s
 // RUN: %clang_cc1 -triple riscv64 -target-feature +d -target-feature +f -target-abi lp64d -emit-llvm %s -o - \
 // RUN:     | FileCheck %s
+// RUN: %clang_cc1 -triple riscv64 -target-abi lp64e -emit-llvm %s -o - \
+// RUN:     | FileCheck %s
 
 #include 
 #include 
diff --git a/clang/test/CodeGen/SystemZ/gnu-atomic-builtins-i128-8Al.c b/clang/test/CodeGen/SystemZ/gnu-atomic-builtins-i128-8Al.c
index e38e6572bd58f..4f6dcbc2c01ec 100644
--- a/clang/test/CodeGen/SystemZ/gnu-atomic-builtins-i128-8Al.c
+++ b/clang/test/CodeGen/SystemZ/gnu-atomic-builtins-i128-8Al.c
@@ -20,10 +20,7 @@ __int128 Des;
 
 // CHECK-LABEL: @f1(
 // CHECK-NEXT:  entry:
-// CHECK-NEXT:    [[ATOMIC_TEMP:%.*]] = alloca i128, align 8
-// CHECK-NEXT:    call void @__atomic_load(i64 noundef 16, ptr noundef nonnull @Ptr, ptr noundef nonnull [[ATOMIC_TEMP]], i32 noundef signext 5)
-// CHECK-NEXT:    [[TMP0:%.*]] = load i128, ptr [[ATOMIC_TEMP]], align 8, !tbaa [[TBAA2:![0-9]+]]
-// CHECK-NEXT:    store i128 [[TMP0]], ptr [[AGG_RESULT:%.*]], align 8, !tbaa [[TBAA2]]
+// CHECK-NEXT:    tail call void @__atomic_load(i64 noundef 16, ptr noundef nonnull @Ptr, ptr noundef nonnull [[AGG_RESULT:%.*]], i32 noundef signext 5)
 // CHECK-NEXT:    ret void
 //
 __int128 f1() {
@@ -33,7 +30,7 @@ __int128 f1() {
 // CHECK-LABEL: @f2(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    tail call void @__atomic_load(i64 noundef 16, ptr noundef nonnull @Ptr, ptr noundef nonnull @Ret, i32 noundef signext 5)
-// CHECK-NEXT:    [[TMP0:%.*]] = load i128, ptr @Ret, align 8, !tbaa [[TBAA2]]
+// CHECK-NEXT:    [[TMP0:%.*]] = load i128, ptr @Ret, align 8, !tbaa [[TBAA2:![0-9]+]]
 // CHECK-NEXT:    store i128 [[TMP0]], ptr [[AGG_RESULT:%.*]], align 8, !tbaa [[TBAA2]]
 // CHECK-NEXT:    ret void
 //
@@ -66,12 +63,9 @@ void f4() {
 // CHECK-LABEL: @f5(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[DOTATOMICTMP:%.*]] = alloca i128, align 8
-// CHECK-NEXT:    [[ATOMIC_TEMP:%.*]] = alloca i128, align 8
 // CHECK-NEXT:    [[TMP0:%.*]] = load i128, ptr @Val, align 8, !tbaa [[TBAA2]]
 // CHECK-NEXT:    store i128 [[TMP0]], ptr [[DOTATOMICTMP]], align 8, !tbaa [[TBAA2]]
-// CHECK-NEXT:    call void @__atomic_exchange(i64 noundef 16, ptr noundef nonnull @Ptr, ptr noundef nonnull [[DOTATOMICTMP]], ptr noundef nonnull [[ATOMIC_TEMP]], i32 noundef signext 5)
-// CHECK-NEXT:    [[TMP1:%.*]] = load i128, ptr [[ATOMIC_TEMP]], align 8, !tbaa [[TBAA2]]
-// CHECK-NEXT:    store i128 [[TMP1]], ptr [[AGG_RESULT:%.*]], align 8, !tbaa [[TBAA2]]
+// CHECK-NEXT:    call void @__atomic_exchange(i64 noundef 16, ptr noundef nonnull @Ptr, ptr noundef nonnull [[DOTATOMICTMP]], ptr noundef nonnull [[AGG_RESULT:%.*]], i32 noundef signext 5)
 // CHECK-NEXT:    ret void
 //
 __int128 f5() {
@@ -119,7 +113,7 @@ _Bool f8() {
 // CHECK-NEXT:    [[INDIRECT_ARG_TEMP:%.*]] = alloca i128, align 8
 // CHECK-NEXT:    [[TMP0:%.*]] = load i128, ptr @Val, align 8, !tbaa [[TBAA2]]
 // CHECK-NEXT:    store i128 [[TMP0]], ptr [[INDIRECT_ARG_TEMP]], align 8, !tbaa [[TBAA2]]
-// CHECK-NEXT:    call void @__atomic_fetch_add_16(ptr nonnull sret(i128) align 8 [[TMP]], ptr noundef nonnull @Ptr, ptr noundef nonnull [[INDIRECT_ARG_TEMP]], i32 noundef signext 5)
+// CHECK-NEXT:    call void @__atomic_fetch_add_16(ptr dead_on_unwind nonnull writable sret(i128) align 8 [[TMP]], ptr noundef nonnull @Ptr, ptr noundef nonnull [[INDIRECT_ARG_TEMP]], i32 noundef signext 5)
 // CHECK-NEXT:    [[TMP1:%.*]] = load i128, ptr [[TMP]], align 8, !tbaa [[TBAA2]]
 // CHECK-NEXT:    [[TMP2:%.*]] = add i128 [[TMP1]], [[TMP0]]
 // CHECK-NEXT:    store i128 [[TMP2]], ptr [[AGG_RESULT:%.*]], align 8, !tbaa [[TBAA2]]
@@ -135,7 +129,7 @@ __int128 f9() {
 // CHECK-NEXT:    [[INDIRECT_ARG_TEMP:%.*]] = alloca i128, align 8
 // CHECK-NEXT:    [[TMP0:%.*]] = load i128, ptr @Val, align 8, !tbaa [[TBAA2]]
 // CHECK-NEXT:    store i128 [[TMP0]], ptr [[INDIRECT_ARG_TEMP]], align 8, !tbaa [[TBAA2]]
-// CHECK-NEXT:    call void @__atomic_fetch_sub_16(ptr nonnull sret(i128) align 8 [[TMP]], ptr noundef nonnull @Ptr, ptr noundef nonnull [[INDIRECT_ARG_TEMP]], i32 noundef signext 5)
+// CHECK-NEXT:    call void @__atomic_fetch_sub_16(ptr dead_on_unwind nonnull writable sret(i128) align 8 [[TMP]], ptr noundef nonnull @Ptr, ptr noundef nonnull [[INDIRECT_ARG_TEMP]], i32 noundef signext 5)
 // CHECK-NEXT:    [[TMP1:%.*]] = load i128, ptr [[TMP]], align 8, !tbaa [[TBAA2]]
 // CHECK-NEXT:    [[TMP2:%.*]] = sub i128 [[TMP1]], [[TMP0]]
 // CHECK-NEXT:    store i128 [[TMP2]], ptr [[AGG_RESULT:%.*]], align 8, !tbaa [[TBAA2]]
@@ -151,7 +145,7 @@ __int128 f10() {
 // CHECK-NEXT:    [[INDIRECT_ARG_TEMP:%.*]] = alloca i128, align 8
 // CHECK-NEXT:    [[TMP0:%.*]] = load i128, ptr @Val, align 8, !tbaa [[TBAA2]]
 // CHECK-NEXT:    store i128 [[TMP0]], ptr [[INDIRECT_ARG_TEMP]], align 8, !tbaa [[TBAA2]]
-// CHECK-NEXT:    call void @__atomic_fetch_and_16(ptr nonnull sret(i128) align 8 [[TMP]], ptr noundef nonnull @Ptr, ptr noundef nonnull [[INDIRECT_ARG_TEMP]], i32 noundef signext 5)
+// CHECK-NEXT:    call void @__atomic_fetch_and_16(ptr dead_on_unwind nonnull writable sret(i128) align 8 [[TMP]], ptr noundef nonnull @Ptr, ptr noundef nonnull [[INDIRECT_ARG_TEMP]], i32 noundef signext 5)
 // CHECK-NEXT:    [[TMP1:%.*]] = load i128, ptr [[TMP]], align 8, !tbaa [[TBAA2]]
 // CHECK-NEXT:    [[TMP2:%.*]] = and i128 [[TMP1]], [[TMP0]]
 // CHECK-NEXT:    store i128 [[TMP2]], ptr [[AGG_RESULT:%.*]], align 8, !tbaa [[TBAA2]]
@@ -167,7 +161,7 @@ __int128 f11() {
 // CHECK-NEXT:    [[INDIRECT_ARG_TEMP:%.*]] = alloca i128, align 8
 // CHECK-NEXT:    [[TMP0:%.*]] = load i128, ptr @Val, align 8, !tbaa [[TBAA2]]
 // CHECK-NEXT:    store i128 [[TMP0]], ptr [[INDIRECT_ARG_TEMP]], align 8, !tbaa [[TBAA2]]
-// CHECK-NEXT:    call void @__atomic_fetch_xor_16(ptr nonnull sret(i128) align 8 [[TMP]], ptr noundef nonnull @Ptr, ptr noundef nonnull [[INDIRECT_ARG_TEMP]], i32 noundef signext 5)
+// CHECK-NEXT:    call void @__atomic_fetch_xor_16(ptr dead_on_unwind nonnull writable sret(i128) align 8 [[TMP]], ptr noundef nonnull @Ptr, ptr noundef nonnull [[INDIRECT_ARG_TEMP]], i32 noundef signext 5)
 // CHECK-NEXT:    [[TMP1:%.*]] = load i128, ptr [[TMP]], align 8, !tbaa [[TBAA2]]
 // CHECK-NEXT:    [[TMP2:%.*]] = xor i128 [[TMP1]], [[TMP0]]
 // CHECK-NEXT:    store i128 [[TMP2]], ptr [[AGG_RESULT:%.*]], align 8, !tbaa [[TBAA2]]
@@ -183,7 +177,7 @@ __int128 f12() {
 // CHECK-NEXT:    [[INDIRECT_ARG_TEMP:%.*]] = alloca i128, align 8
 // CHECK-NEXT:    [[TMP0:%.*]] = load i128, ptr @Val, align 8, !tbaa [[TBAA2]]
 // CHECK-NEXT:    store i128 [[TMP0]], ptr [[INDIRECT_ARG_TEMP]], align 8, !tbaa [[TBAA2]]
-// CHECK-NEXT:    call void @__atomic_fetch_or_16(ptr nonnull sret(i128) align 8 [[TMP]], ptr noundef nonnull @Ptr, ptr noundef nonnull [[INDIRECT_ARG_TEMP]], i32 noundef signext 5)
+// CHECK-NEXT:    call void @__atomic_fetch_or_16(ptr dead_on_unwind nonnull writable sret(i128) align 8 [[TMP]], ptr noundef nonnull @Ptr, ptr noundef nonnull [[INDIRECT_ARG_TEMP]], i32 noundef signext 5)
 // CHECK-NEXT:    [[TMP1:%.*]] = load i128, ptr [[TMP]], align 8, !tbaa [[TBAA2]]
 // CHECK-NEXT:    [[TMP2:%.*]] = or i128 [[TMP1]], [[TMP0]]
 // CHECK-NEXT:    store i128 [[TMP2]], ptr [[AGG_RESULT:%.*]], align 8, !tbaa [[TBAA2]]
@@ -199,7 +193,7 @@ __int128 f13() {
 // CHECK-NEXT:    [[INDIRECT_ARG_TEMP:%.*]] = alloca i128, align 8
 // CHECK-NEXT:    [[TMP0:%.*]] = load i128, ptr @Val, align 8, !tbaa [[TBAA2]]
 // CHECK-NEXT:    store i128 [[TMP0]], ptr [[INDIRECT_ARG_TEMP]], align 8, !tbaa [[TBAA2]]
-// CHECK-NEXT:    call void @__atomic_fetch_nand_16(ptr nonnull sret(i128) align 8 [[TMP]], ptr noundef nonnull @Ptr, ptr noundef nonnull [[INDIRECT_ARG_TEMP]], i32 noundef signext 5)
+// CHECK-NEXT:    call void @__atomic_fetch_nand_16(ptr dead_on_unwind nonnull writable sret(i128) align 8 [[TMP]], ptr noundef nonnull @Ptr, ptr noundef nonnull [[INDIRECT_ARG_TEMP]], i32 noundef signext 5)
 // CHECK-NEXT:    [[TMP1:%.*]] = load i128, ptr [[TMP]], align 8, !tbaa [[TBAA2]]
 // CHECK-NEXT:    [[TMP2:%.*]] = and i128 [[TMP1]], [[TMP0]]
 // CHECK-NEXT:    [[TMP3:%.*]] = xor i128 [[TMP2]], -1
@@ -212,13 +206,10 @@ __int128 f14() {
 
 // CHECK-LABEL: @f15(
 // CHECK-NEXT:  entry:
-// CHECK-NEXT:    [[TMP:%.*]] = alloca i128, align 8
 // CHECK-NEXT:    [[INDIRECT_ARG_TEMP:%.*]] = alloca i128, align 8
 // CHECK-NEXT:    [[TMP0:%.*]] = load i128, ptr @Val, align 8, !tbaa [[TBAA2]]
 // CHECK-NEXT:    store i128 [[TMP0]], ptr [[INDIRECT_ARG_TEMP]], align 8, !tbaa [[TBAA2]]
-// CHECK-NEXT:    call void @__atomic_fetch_add_16(ptr nonnull sret(i128) align 8 [[TMP]], ptr noundef nonnull @Ptr, ptr noundef nonnull [[INDIRECT_ARG_TEMP]], i32 noundef signext 5)
-// CHECK-NEXT:    [[TMP1:%.*]] = load i128, ptr [[TMP]], align 8, !tbaa [[TBAA2]]
-// CHECK-NEXT:    store i128 [[TMP1]], ptr [[AGG_RESULT:%.*]], align 8, !tbaa [[TBAA2]]
+// CHECK-NEXT:    call void @__atomic_fetch_add_16(ptr dead_on_unwind nonnull writable sret(i128) align 8 [[AGG_RESULT:%.*]], ptr noundef nonnull @Ptr, ptr noundef nonnull [[INDIRECT_ARG_TEMP]], i32 noundef signext 5)
 // CHECK-NEXT:    ret void
 //
 __int128 f15() {
@@ -227,13 +218,10 @@ __int128 f15() {
 
 // CHECK-LABEL: @f16(
 // CHECK-NEXT:  entry:
-// CHECK-NEXT:    [[TMP:%.*]] = alloca i128, align 8
 // CHECK-NEXT:    [[INDIRECT_ARG_TEMP:%.*]] = alloca i128, align 8
 // CHECK-NEXT:    [[TMP0:%.*]] = load i128, ptr @Val, align 8, !tbaa [[TBAA2]]
 // CHECK-NEXT:    store i128 [[TMP0]], ptr [[INDIRECT_ARG_TEMP]], align 8, !tbaa [[TBAA2]]
-// CHECK-NEXT:    call void @__atomic_fetch_sub_16(ptr nonnull sret(i128) align 8 [[TMP]], ptr noundef nonnull @Ptr, ptr noundef nonnull [[INDIRECT_ARG_TEMP]], i32 noundef signext 5)
-// CHECK-NEXT:    [[TMP1:%.*]] = load i128, ptr [[TMP]], align 8, !tbaa [[TBAA2]]
-// CHECK-NEXT:    store i128 [[TMP1]], ptr [[AGG_RESULT:%.*]], align 8, !tbaa [[TBAA2]]
+// CHECK-NEXT:    call void @__atomic_fetch_sub_16(ptr dead_on_unwind nonnull writable sret(i128) align 8 [[AGG_RESULT:%.*]], ptr noundef nonnull @Ptr, ptr noundef nonnull [[INDIRECT_ARG_TEMP]], i32 noundef signext 5)
 // CHECK-NEXT:    ret void
 //
 __int128 f16() {
@@ -242,13 +230,10 @@ __int128 f16() {
 
 // CHECK-LABEL: @f17(
 // CHECK-NEXT:  entry:
-// CHECK-NEXT:    [[TMP:%.*]] = alloca i128, align 8
 // CHECK-NEXT:    [[INDIRECT_ARG_TEMP:%.*]] = alloca i128, align 8
 // CHECK-NEXT:    [[TMP0:%.*]] = load i128, ptr @Val, align 8, !tbaa [[TBAA2]]
 // CHECK-NEXT:    store i128 [[TMP0]], ptr [[INDIRECT_ARG_TEMP]], align 8, !tbaa [[TBAA2]]
-// CHECK-NEXT:    call void @__atomic_fetch_and_16(ptr nonnull sret(i128) align 8 [[TMP]], ptr noundef nonnull @Ptr, ptr noundef nonnull [[INDIRECT_ARG_TEMP]], i32 noundef signext 5)
-// CHECK-NEXT:    [[TMP1:%.*]] = load i128, ptr [[TMP]], align 8, !tbaa [[TBAA2]]
-// CHECK-NEXT:    store i128 [[TMP1]], ptr [[AGG_RESULT:%.*]], align 8, !tbaa [[TBAA2]]
+// CHECK-NEXT:    call void @__atomic_fetch_and_16(ptr dead_on_unwind nonnull writable sret(i128) align 8 [[AGG_RESULT:%.*]], ptr noundef nonnull @Ptr, ptr noundef nonnull [[INDIRECT_ARG_TEMP]], i32 noundef signext 5)
 // CHECK-NEXT:    ret void
 //
 __int128 f17() {
@@ -257,13 +242,10 @@ __int128 f17() {
 
 // CHECK-LABEL: @f18(
 // CHECK-NEXT:  entry:
-// CHECK-NEXT:    [[TMP:%.*]] = alloca i128, align 8
 // CHECK-NEXT:    [[INDIRECT_ARG_TEMP:%.*]] = alloca i128, align 8
 // CHECK-NEXT:    [[TMP0:%.*]] = load i128, ptr @Val, align 8, !tbaa [[TBAA2]]
 // CHECK-NEXT:    store i128 [[TMP0]], ptr [[INDIRECT_ARG_TEMP]], align 8, !tbaa [[TBAA2]]
-// CHECK-NEXT:    call void @__atomic_fetch_xor_16(ptr nonnull sret(i128) align 8 [[TMP]], ptr noundef nonnull @Ptr, ptr noundef nonnull [[INDIRECT_ARG_TEMP]], i32 noundef signext 5)
-// CHECK-NEXT:    [[TMP1:%.*]] = load i128, ptr [[TMP]], align 8, !tbaa [[TBAA2]]
-// CHECK-NEXT:    store i128 [[TMP1]], ptr [[AGG_RESULT:%.*]], align 8, !tbaa [[TBAA2]]
+// CHECK-NEXT:    call void @__atomic_fetch_xor_16(ptr dead_on_unwind nonnull writable sret(i128) align 8 [[AGG_RESULT:%.*]], ptr noundef nonnull @Ptr, ptr noundef nonnull [[INDIRECT_ARG_TEMP]], i32 noundef signext 5)
 // CHECK-NEXT:    ret void
 //
 __int128 f18() {
@@ -272,13 +254,10 @@ __int128 f18() {
 
 // CHECK-LABEL: @f19(
 // CHECK-NEXT:  entry:
-// CHECK-NEXT:    [[TMP:%.*]] = alloca i128, align 8
 // CHECK-NEXT:    [[INDIRECT_ARG_TEMP:%.*]] = alloca i128, align 8
 // CHECK-NEXT:    [[TMP0:%.*]] = load i128, ptr @Val, align 8, !tbaa [[TBAA2]]
 // CHECK-NEXT:    store i128 [[TMP0]], ptr [[INDIRECT_ARG_TEMP]], align 8, !tbaa [[TBAA2]]
-// CHECK-NEXT:    call void @__atomic_fetch_or_16(ptr nonnull sret(i128) align 8 [[TMP]], ptr noundef nonnull @Ptr, ptr noundef nonnull [[INDIRECT_ARG_TEMP]], i32 noundef signext 5)
-// CHECK-NEXT:    [[TMP1:%.*]] = load i128, ptr [[TMP]], align 8, !tbaa [[TBAA2]]
-// CHECK-NEXT:    store i128 [[TMP1]], ptr [[AGG_RESULT:%.*]], align 8, !tbaa [[TBAA2]]
+// CHECK-NEXT:    call void @__atomic_fetch_or_16(ptr dead_on_unwind nonnull writable sret(i128) align 8 [[AGG_RESULT:%.*]], ptr noundef nonnull @Ptr, ptr noundef nonnull [[INDIRECT_ARG_TEMP]], i32 noundef signext 5)
 // CHECK-NEXT:    ret void
 //
 __int128 f19() {
@@ -287,13 +266,10 @@ __int128 f19() {
 
 // CHECK-LABEL: @f20(
 // CHECK-NEXT:  entry:
-// CHECK-NEXT:    [[TMP:%.*]] = alloca i128, align 8
 // CHECK-NEXT:    [[INDIRECT_ARG_TEMP:%.*]] = alloca i128, align 8
 // CHECK-NEXT:    [[TMP0:%.*]] = load i128, ptr @Val, align 8, !tbaa [[TBAA2]]
 // CHECK-NEXT:    store i128 [[TMP0]], ptr [[INDIRECT_ARG_TEMP]], align 8, !tbaa [[TBAA2]]
-// CHECK-NEXT:    call void @__atomic_fetch_nand_16(ptr nonnull sret(i128) align 8 [[TMP]], ptr noundef nonnull @Ptr, ptr noundef nonnull [[INDIRECT_ARG_TEMP]], i32 noundef signext 5)
-// CHECK-NEXT:    [[TMP1:%.*]] = load i128, ptr [[TMP]], align 8, !tbaa [[TBAA2]]
-// CHECK-NEXT:    store i128 [[TMP1]], ptr [[AGG_RESULT:%.*]], align 8, !tbaa [[TBAA2]]
+// CHECK-NEXT:    call void @__atomic_fetch_nand_16(ptr dead_on_unwind nonnull writable sret(i128) align 8 [[AGG_RESULT:%.*]], ptr noundef nonnull @Ptr, ptr noundef nonnull [[INDIRECT_ARG_TEMP]], i32 noundef signext 5)
 // CHECK-NEXT:    ret void
 //
 __int128 f20() {
diff --git a/clang/test/CodeGen/SystemZ/systemz-abi-vector.c b/clang/test/CodeGen/SystemZ/systemz-abi-vector.c
index f7606641a3747..6a06d15f22aa3 100644
--- a/clang/test/CodeGen/SystemZ/systemz-abi-vector.c
+++ b/clang/test/CodeGen/SystemZ/systemz-abi-vector.c
@@ -54,91 +54,91 @@ unsigned int align = __alignof__ (v16i8);
 // CHECK-VECTOR: @align ={{.*}} global i32 8
 
 v1i8 pass_v1i8(v1i8 arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_v1i8(ptr noalias sret(<1 x i8>) align 1 %{{.*}}, ptr %0)
+// CHECK-LABEL: define{{.*}} void @pass_v1i8(ptr dead_on_unwind noalias writable sret(<1 x i8>) align 1 %{{.*}}, ptr %0)
 // CHECK-VECTOR-LABEL: define{{.*}} <1 x i8> @pass_v1i8(<1 x i8> %{{.*}})
 
 v2i8 pass_v2i8(v2i8 arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_v2i8(ptr noalias sret(<2 x i8>) align 2 %{{.*}}, ptr %0)
+// CHECK-LABEL: define{{.*}} void @pass_v2i8(ptr dead_on_unwind noalias writable sret(<2 x i8>) align 2 %{{.*}}, ptr %0)
 // CHECK-VECTOR-LABEL: define{{.*}} <2 x i8> @pass_v2i8(<2 x i8> %{{.*}})
 
 v4i8 pass_v4i8(v4i8 arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_v4i8(ptr noalias sret(<4 x i8>) align 4 %{{.*}}, ptr %0)
+// CHECK-LABEL: define{{.*}} void @pass_v4i8(ptr dead_on_unwind noalias writable sret(<4 x i8>) align 4 %{{.*}}, ptr %0)
 // CHECK-VECTOR-LABEL: define{{.*}} <4 x i8> @pass_v4i8(<4 x i8> %{{.*}})
 
 v8i8 pass_v8i8(v8i8 arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_v8i8(ptr noalias sret(<8 x i8>) align 8 %{{.*}}, ptr %0)
+// CHECK-LABEL: define{{.*}} void @pass_v8i8(ptr dead_on_unwind noalias writable sret(<8 x i8>) align 8 %{{.*}}, ptr %0)
 // CHECK-VECTOR-LABEL: define{{.*}} <8 x i8> @pass_v8i8(<8 x i8> %{{.*}})
 
 v16i8 pass_v16i8(v16i8 arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_v16i8(ptr noalias sret(<16 x i8>) align 16 %{{.*}}, ptr %0)
+// CHECK-LABEL: define{{.*}} void @pass_v16i8(ptr dead_on_unwind noalias writable sret(<16 x i8>) align 16 %{{.*}}, ptr %0)
 // CHECK-VECTOR-LABEL: define{{.*}} <16 x i8> @pass_v16i8(<16 x i8> %{{.*}})
 
 v32i8 pass_v32i8(v32i8 arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_v32i8(ptr noalias sret(<32 x i8>) align 32 %{{.*}}, ptr %0)
-// CHECK-VECTOR-LABEL: define{{.*}} void @pass_v32i8(ptr noalias sret(<32 x i8>) align 8 %{{.*}}, ptr %0)
+// CHECK-LABEL: define{{.*}} void @pass_v32i8(ptr dead_on_unwind noalias writable sret(<32 x i8>) align 32 %{{.*}}, ptr %0)
+// CHECK-VECTOR-LABEL: define{{.*}} void @pass_v32i8(ptr dead_on_unwind noalias writable sret(<32 x i8>) align 8 %{{.*}}, ptr %0)
 
 v1i16 pass_v1i16(v1i16 arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_v1i16(ptr noalias sret(<1 x i16>) align 2 %{{.*}}, ptr %0)
+// CHECK-LABEL: define{{.*}} void @pass_v1i16(ptr dead_on_unwind noalias writable sret(<1 x i16>) align 2 %{{.*}}, ptr %0)
 // CHECK-VECTOR-LABEL: define{{.*}} <1 x i16> @pass_v1i16(<1 x i16> %{{.*}})
 
 v2i16 pass_v2i16(v2i16 arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_v2i16(ptr noalias sret(<2 x i16>) align 4 %{{.*}}, ptr %0)
+// CHECK-LABEL: define{{.*}} void @pass_v2i16(ptr dead_on_unwind noalias writable sret(<2 x i16>) align 4 %{{.*}}, ptr %0)
 // CHECK-VECTOR-LABEL: define{{.*}} <2 x i16> @pass_v2i16(<2 x i16> %{{.*}})
 
 v4i16 pass_v4i16(v4i16 arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_v4i16(ptr noalias sret(<4 x i16>) align 8 %{{.*}}, ptr %0)
+// CHECK-LABEL: define{{.*}} void @pass_v4i16(ptr dead_on_unwind noalias writable sret(<4 x i16>) align 8 %{{.*}}, ptr %0)
 // CHECK-VECTOR-LABEL: define{{.*}} <4 x i16> @pass_v4i16(<4 x i16> %{{.*}})
 
 v8i16 pass_v8i16(v8i16 arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_v8i16(ptr noalias sret(<8 x i16>) align 16 %{{.*}}, ptr %0)
+// CHECK-LABEL: define{{.*}} void @pass_v8i16(ptr dead_on_unwind noalias writable sret(<8 x i16>) align 16 %{{.*}}, ptr %0)
 // CHECK-VECTOR-LABEL: define{{.*}} <8 x i16> @pass_v8i16(<8 x i16> %{{.*}})
 
 v1i32 pass_v1i32(v1i32 arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_v1i32(ptr noalias sret(<1 x i32>) align 4 %{{.*}}, ptr %0)
+// CHECK-LABEL: define{{.*}} void @pass_v1i32(ptr dead_on_unwind noalias writable sret(<1 x i32>) align 4 %{{.*}}, ptr %0)
 // CHECK-VECTOR-LABEL: define{{.*}} <1 x i32> @pass_v1i32(<1 x i32> %{{.*}})
 
 v2i32 pass_v2i32(v2i32 arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_v2i32(ptr noalias sret(<2 x i32>) align 8 %{{.*}}, ptr %0)
+// CHECK-LABEL: define{{.*}} void @pass_v2i32(ptr dead_on_unwind noalias writable sret(<2 x i32>) align 8 %{{.*}}, ptr %0)
 // CHECK-VECTOR-LABEL: define{{.*}} <2 x i32> @pass_v2i32(<2 x i32> %{{.*}})
 
 v4i32 pass_v4i32(v4i32 arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_v4i32(ptr noalias sret(<4 x i32>) align 16 %{{.*}}, ptr %0)
+// CHECK-LABEL: define{{.*}} void @pass_v4i32(ptr dead_on_unwind noalias writable sret(<4 x i32>) align 16 %{{.*}}, ptr %0)
 // CHECK-VECTOR-LABEL: define{{.*}} <4 x i32> @pass_v4i32(<4 x i32> %{{.*}})
 
 v1i64 pass_v1i64(v1i64 arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_v1i64(ptr noalias sret(<1 x i64>) align 8 %{{.*}}, ptr %0)
+// CHECK-LABEL: define{{.*}} void @pass_v1i64(ptr dead_on_unwind noalias writable sret(<1 x i64>) align 8 %{{.*}}, ptr %0)
 // CHECK-VECTOR-LABEL: define{{.*}} <1 x i64> @pass_v1i64(<1 x i64> %{{.*}})
 
 v2i64 pass_v2i64(v2i64 arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_v2i64(ptr noalias sret(<2 x i64>) align 16 %{{.*}}, ptr %0)
+// CHECK-LABEL: define{{.*}} void @pass_v2i64(ptr dead_on_unwind noalias writable sret(<2 x i64>) align 16 %{{.*}}, ptr %0)
 // CHECK-VECTOR-LABEL: define{{.*}} <2 x i64> @pass_v2i64(<2 x i64> %{{.*}})
 
 v1i128 pass_v1i128(v1i128 arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_v1i128(ptr noalias sret(<1 x i128>) align 16 %{{.*}}, ptr %0)
+// CHECK-LABEL: define{{.*}} void @pass_v1i128(ptr dead_on_unwind noalias writable sret(<1 x i128>) align 16 %{{.*}}, ptr %0)
 // CHECK-VECTOR-LABEL: define{{.*}} <1 x i128> @pass_v1i128(<1 x i128> %{{.*}})
 
 v1f32 pass_v1f32(v1f32 arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_v1f32(ptr noalias sret(<1 x float>) align 4 %{{.*}}, ptr %0)
+// CHECK-LABEL: define{{.*}} void @pass_v1f32(ptr dead_on_unwind noalias writable sret(<1 x float>) align 4 %{{.*}}, ptr %0)
 // CHECK-VECTOR-LABEL: define{{.*}} <1 x float> @pass_v1f32(<1 x float> %{{.*}})
 
 v2f32 pass_v2f32(v2f32 arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_v2f32(ptr noalias sret(<2 x float>) align 8 %{{.*}}, ptr %0)
+// CHECK-LABEL: define{{.*}} void @pass_v2f32(ptr dead_on_unwind noalias writable sret(<2 x float>) align 8 %{{.*}}, ptr %0)
 // CHECK-VECTOR-LABEL: define{{.*}} <2 x float> @pass_v2f32(<2 x float> %{{.*}})
 
 v4f32 pass_v4f32(v4f32 arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_v4f32(ptr noalias sret(<4 x float>) align 16 %{{.*}}, ptr %0)
+// CHECK-LABEL: define{{.*}} void @pass_v4f32(ptr dead_on_unwind noalias writable sret(<4 x float>) align 16 %{{.*}}, ptr %0)
 // CHECK-VECTOR-LABEL: define{{.*}} <4 x float> @pass_v4f32(<4 x float> %{{.*}})
 
 v1f64 pass_v1f64(v1f64 arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_v1f64(ptr noalias sret(<1 x double>) align 8 %{{.*}}, ptr %0)
+// CHECK-LABEL: define{{.*}} void @pass_v1f64(ptr dead_on_unwind noalias writable sret(<1 x double>) align 8 %{{.*}}, ptr %0)
 // CHECK-VECTOR-LABEL: define{{.*}} <1 x double> @pass_v1f64(<1 x double> %{{.*}})
 
 v2f64 pass_v2f64(v2f64 arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_v2f64(ptr noalias sret(<2 x double>) align 16 %{{.*}}, ptr %0)
+// CHECK-LABEL: define{{.*}} void @pass_v2f64(ptr dead_on_unwind noalias writable sret(<2 x double>) align 16 %{{.*}}, ptr %0)
 // CHECK-VECTOR-LABEL: define{{.*}} <2 x double> @pass_v2f64(<2 x double> %{{.*}})
 
 v1f128 pass_v1f128(v1f128 arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_v1f128(ptr noalias sret(<1 x fp128>) align 16 %{{.*}}, ptr %0)
+// CHECK-LABEL: define{{.*}} void @pass_v1f128(ptr dead_on_unwind noalias writable sret(<1 x fp128>) align 16 %{{.*}}, ptr %0)
 // CHECK-VECTOR-LABEL: define{{.*}} <1 x fp128> @pass_v1f128(<1 x fp128> %{{.*}})
 
 
@@ -146,62 +146,62 @@ v1f128 pass_v1f128(v1f128 arg) { return arg; }
 
 struct agg_v1i8 { v1i8 a; };
 struct agg_v1i8 pass_agg_v1i8(struct agg_v1i8 arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_agg_v1i8(ptr noalias sret(%struct.agg_v1i8) align 1 %{{.*}}, i8 %{{.*}})
-// CHECK-VECTOR-LABEL: define{{.*}} void @pass_agg_v1i8(ptr noalias sret(%struct.agg_v1i8) align 1 %{{.*}}, <1 x i8> %{{.*}})
+// CHECK-LABEL: define{{.*}} void @pass_agg_v1i8(ptr dead_on_unwind noalias writable sret(%struct.agg_v1i8) align 1 %{{.*}}, i8 %{{.*}})
+// CHECK-VECTOR-LABEL: define{{.*}} void @pass_agg_v1i8(ptr dead_on_unwind noalias writable sret(%struct.agg_v1i8) align 1 %{{.*}}, <1 x i8> %{{.*}})
 
 struct agg_v2i8 { v2i8 a; };
 struct agg_v2i8 pass_agg_v2i8(struct agg_v2i8 arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_agg_v2i8(ptr noalias sret(%struct.agg_v2i8) align 2 %{{.*}}, i16 %{{.*}})
-// CHECK-VECTOR-LABEL: define{{.*}} void @pass_agg_v2i8(ptr noalias sret(%struct.agg_v2i8) align 2 %{{.*}}, <2 x i8> %{{.*}})
+// CHECK-LABEL: define{{.*}} void @pass_agg_v2i8(ptr dead_on_unwind noalias writable sret(%struct.agg_v2i8) align 2 %{{.*}}, i16 %{{.*}})
+// CHECK-VECTOR-LABEL: define{{.*}} void @pass_agg_v2i8(ptr dead_on_unwind noalias writable sret(%struct.agg_v2i8) align 2 %{{.*}}, <2 x i8> %{{.*}})
 
 struct agg_v4i8 { v4i8 a; };
 struct agg_v4i8 pass_agg_v4i8(struct agg_v4i8 arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_agg_v4i8(ptr noalias sret(%struct.agg_v4i8) align 4 %{{.*}}, i32 %{{.*}})
-// CHECK-VECTOR-LABEL: define{{.*}} void @pass_agg_v4i8(ptr noalias sret(%struct.agg_v4i8) align 4 %{{.*}}, <4 x i8> %{{.*}})
+// CHECK-LABEL: define{{.*}} void @pass_agg_v4i8(ptr dead_on_unwind noalias writable sret(%struct.agg_v4i8) align 4 %{{.*}}, i32 %{{.*}})
+// CHECK-VECTOR-LABEL: define{{.*}} void @pass_agg_v4i8(ptr dead_on_unwind noalias writable sret(%struct.agg_v4i8) align 4 %{{.*}}, <4 x i8> %{{.*}})
 
 struct agg_v8i8 { v8i8 a; };
 struct agg_v8i8 pass_agg_v8i8(struct agg_v8i8 arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_agg_v8i8(ptr noalias sret(%struct.agg_v8i8) align 8 %{{.*}}, i64 %{{.*}})
-// CHECK-VECTOR-LABEL: define{{.*}} void @pass_agg_v8i8(ptr noalias sret(%struct.agg_v8i8) align 8 %{{.*}}, <8 x i8> %{{.*}})
+// CHECK-LABEL: define{{.*}} void @pass_agg_v8i8(ptr dead_on_unwind noalias writable sret(%struct.agg_v8i8) align 8 %{{.*}}, i64 %{{.*}})
+// CHECK-VECTOR-LABEL: define{{.*}} void @pass_agg_v8i8(ptr dead_on_unwind noalias writable sret(%struct.agg_v8i8) align 8 %{{.*}}, <8 x i8> %{{.*}})
 
 struct agg_v16i8 { v16i8 a; };
 struct agg_v16i8 pass_agg_v16i8(struct agg_v16i8 arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_agg_v16i8(ptr noalias sret(%struct.agg_v16i8) align 16 %{{.*}}, ptr %{{.*}})
-// CHECK-VECTOR-LABEL: define{{.*}} void @pass_agg_v16i8(ptr noalias sret(%struct.agg_v16i8) align 8 %{{.*}}, <16 x i8> %{{.*}})
+// CHECK-LABEL: define{{.*}} void @pass_agg_v16i8(ptr dead_on_unwind noalias writable sret(%struct.agg_v16i8) align 16 %{{.*}}, ptr %{{.*}})
+// CHECK-VECTOR-LABEL: define{{.*}} void @pass_agg_v16i8(ptr dead_on_unwind noalias writable sret(%struct.agg_v16i8) align 8 %{{.*}}, <16 x i8> %{{.*}})
 
 struct agg_v32i8 { v32i8 a; };
 struct agg_v32i8 pass_agg_v32i8(struct agg_v32i8 arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_agg_v32i8(ptr noalias sret(%struct.agg_v32i8) align 32 %{{.*}}, ptr %{{.*}})
-// CHECK-VECTOR-LABEL: define{{.*}} void @pass_agg_v32i8(ptr noalias sret(%struct.agg_v32i8) align 8 %{{.*}}, ptr %{{.*}})
+// CHECK-LABEL: define{{.*}} void @pass_agg_v32i8(ptr dead_on_unwind noalias writable sret(%struct.agg_v32i8) align 32 %{{.*}}, ptr %{{.*}})
+// CHECK-VECTOR-LABEL: define{{.*}} void @pass_agg_v32i8(ptr dead_on_unwind noalias writable sret(%struct.agg_v32i8) align 8 %{{.*}}, ptr %{{.*}})
 
 
 // Verify that the following are *not* vector-like aggregate types
 
 struct agg_novector1 { v4i8 a; v4i8 b; };
 struct agg_novector1 pass_agg_novector1(struct agg_novector1 arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_agg_novector1(ptr noalias sret(%struct.agg_novector1) align 4 %{{.*}}, i64 %{{.*}})
-// CHECK-VECTOR-LABEL: define{{.*}} void @pass_agg_novector1(ptr noalias sret(%struct.agg_novector1) align 4 %{{.*}}, i64 %{{.*}})
+// CHECK-LABEL: define{{.*}} void @pass_agg_novector1(ptr dead_on_unwind noalias writable sret(%struct.agg_novector1) align 4 %{{.*}}, i64 %{{.*}})
+// CHECK-VECTOR-LABEL: define{{.*}} void @pass_agg_novector1(ptr dead_on_unwind noalias writable sret(%struct.agg_novector1) align 4 %{{.*}}, i64 %{{.*}})
 
 struct agg_novector2 { v4i8 a; float b; };
 struct agg_novector2 pass_agg_novector2(struct agg_novector2 arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_agg_novector2(ptr noalias sret(%struct.agg_novector2) align 4 %{{.*}}, i64 %{{.*}})
-// CHECK-VECTOR-LABEL: define{{.*}} void @pass_agg_novector2(ptr noalias sret(%struct.agg_novector2) align 4 %{{.*}}, i64 %{{.*}})
+// CHECK-LABEL: define{{.*}} void @pass_agg_novector2(ptr dead_on_unwind noalias writable sret(%struct.agg_novector2) align 4 %{{.*}}, i64 %{{.*}})
+// CHECK-VECTOR-LABEL: define{{.*}} void @pass_agg_novector2(ptr dead_on_unwind noalias writable sret(%struct.agg_novector2) align 4 %{{.*}}, i64 %{{.*}})
 
 struct agg_novector3 { v4i8 a; int : 0; };
 struct agg_novector3 pass_agg_novector3(struct agg_novector3 arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_agg_novector3(ptr noalias sret(%struct.agg_novector3) align 4 %{{.*}}, i32 %{{.*}})
-// CHECK-VECTOR-LABEL: define{{.*}} void @pass_agg_novector3(ptr noalias sret(%struct.agg_novector3) align 4 %{{.*}}, i32 %{{.*}})
+// CHECK-LABEL: define{{.*}} void @pass_agg_novector3(ptr dead_on_unwind noalias writable sret(%struct.agg_novector3) align 4 %{{.*}}, i32 %{{.*}})
+// CHECK-VECTOR-LABEL: define{{.*}} void @pass_agg_novector3(ptr dead_on_unwind noalias writable sret(%struct.agg_novector3) align 4 %{{.*}}, i32 %{{.*}})
 
 struct agg_novector4 { v4i8 a __attribute__((aligned (8))); };
 struct agg_novector4 pass_agg_novector4(struct agg_novector4 arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_agg_novector4(ptr noalias sret(%struct.agg_novector4) align 8 %{{.*}}, i64 %{{.*}})
-// CHECK-VECTOR-LABEL: define{{.*}} void @pass_agg_novector4(ptr noalias sret(%struct.agg_novector4) align 8 %{{.*}}, i64 %{{.*}})
+// CHECK-LABEL: define{{.*}} void @pass_agg_novector4(ptr dead_on_unwind noalias writable sret(%struct.agg_novector4) align 8 %{{.*}}, i64 %{{.*}})
+// CHECK-VECTOR-LABEL: define{{.*}} void @pass_agg_novector4(ptr dead_on_unwind noalias writable sret(%struct.agg_novector4) align 8 %{{.*}}, i64 %{{.*}})
 
 
 // Accessing variable argument lists
 
 v1i8 va_v1i8(__builtin_va_list l) { return __builtin_va_arg(l, v1i8); }
-// CHECK-LABEL: define{{.*}} void @va_v1i8(ptr noalias sret(<1 x i8>) align 1 %{{.*}}, ptr %{{.*}})
+// CHECK-LABEL: define{{.*}} void @va_v1i8(ptr dead_on_unwind noalias writable sret(<1 x i8>) align 1 %{{.*}}, ptr %{{.*}})
 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, ptr %{{.*}}, i32 0, i32 0
 // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, ptr [[REG_COUNT_PTR]]
 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
@@ -230,7 +230,7 @@ v1i8 va_v1i8(__builtin_va_list l) { return __builtin_va_arg(l, v1i8); }
 // CHECK-VECTOR: ret <1 x i8> [[RET]]
 
 v2i8 va_v2i8(__builtin_va_list l) { return __builtin_va_arg(l, v2i8); }
-// CHECK-LABEL: define{{.*}} void @va_v2i8(ptr noalias sret(<2 x i8>) align 2 %{{.*}}, ptr %{{.*}})
+// CHECK-LABEL: define{{.*}} void @va_v2i8(ptr dead_on_unwind noalias writable sret(<2 x i8>) align 2 %{{.*}}, ptr %{{.*}})
 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, ptr %{{.*}}, i32 0, i32 0
 // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, ptr [[REG_COUNT_PTR]]
 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
@@ -259,7 +259,7 @@ v2i8 va_v2i8(__builtin_va_list l) { return __builtin_va_arg(l, v2i8); }
 // CHECK-VECTOR: ret <2 x i8> [[RET]]
 
 v4i8 va_v4i8(__builtin_va_list l) { return __builtin_va_arg(l, v4i8); }
-// CHECK-LABEL: define{{.*}} void @va_v4i8(ptr noalias sret(<4 x i8>) align 4 %{{.*}}, ptr %{{.*}})
+// CHECK-LABEL: define{{.*}} void @va_v4i8(ptr dead_on_unwind noalias writable sret(<4 x i8>) align 4 %{{.*}}, ptr %{{.*}})
 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, ptr %{{.*}}, i32 0, i32 0
 // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, ptr [[REG_COUNT_PTR]]
 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
@@ -288,7 +288,7 @@ v4i8 va_v4i8(__builtin_va_list l) { return __builtin_va_arg(l, v4i8); }
 // CHECK-VECTOR: ret <4 x i8> [[RET]]
 
 v8i8 va_v8i8(__builtin_va_list l) { return __builtin_va_arg(l, v8i8); }
-// CHECK-LABEL: define{{.*}} void @va_v8i8(ptr noalias sret(<8 x i8>) align 8 %{{.*}}, ptr %{{.*}})
+// CHECK-LABEL: define{{.*}} void @va_v8i8(ptr dead_on_unwind noalias writable sret(<8 x i8>) align 8 %{{.*}}, ptr %{{.*}})
 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, ptr %{{.*}}, i32 0, i32 0
 // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, ptr [[REG_COUNT_PTR]]
 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
@@ -317,7 +317,7 @@ v8i8 va_v8i8(__builtin_va_list l) { return __builtin_va_arg(l, v8i8); }
 // CHECK-VECTOR: ret <8 x i8> [[RET]]
 
 v16i8 va_v16i8(__builtin_va_list l) { return __builtin_va_arg(l, v16i8); }
-// CHECK-LABEL: define{{.*}} void @va_v16i8(ptr noalias sret(<16 x i8>) align 16 %{{.*}}, ptr %{{.*}})
+// CHECK-LABEL: define{{.*}} void @va_v16i8(ptr dead_on_unwind noalias writable sret(<16 x i8>) align 16 %{{.*}}, ptr %{{.*}})
 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, ptr %{{.*}}, i32 0, i32 0
 // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, ptr [[REG_COUNT_PTR]]
 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
@@ -346,7 +346,7 @@ v16i8 va_v16i8(__builtin_va_list l) { return __builtin_va_arg(l, v16i8); }
 // CHECK-VECTOR: ret <16 x i8> [[RET]]
 
 v32i8 va_v32i8(__builtin_va_list l) { return __builtin_va_arg(l, v32i8); }
-// CHECK-LABEL: define{{.*}} void @va_v32i8(ptr noalias sret(<32 x i8>) align 32 %{{.*}}, ptr %{{.*}})
+// CHECK-LABEL: define{{.*}} void @va_v32i8(ptr dead_on_unwind noalias writable sret(<32 x i8>) align 32 %{{.*}}, ptr %{{.*}})
 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, ptr %{{.*}}, i32 0, i32 0
 // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, ptr [[REG_COUNT_PTR]]
 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
@@ -366,7 +366,7 @@ v32i8 va_v32i8(__builtin_va_list l) { return __builtin_va_arg(l, v32i8); }
 // CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi ptr [ [[RAW_REG_ADDR]], %{{.*}} ], [ [[RAW_MEM_ADDR]], %{{.*}} ]
 // CHECK: [[INDIRECT_ARG:%[^ ]+]] = load ptr, ptr [[VA_ARG_ADDR]]
 // CHECK: ret void
-// CHECK-VECTOR-LABEL: define{{.*}} void @va_v32i8(ptr noalias sret(<32 x i8>) align 8 %{{.*}}, ptr %{{.*}})
+// CHECK-VECTOR-LABEL: define{{.*}} void @va_v32i8(ptr dead_on_unwind noalias writable sret(<32 x i8>) align 8 %{{.*}}, ptr %{{.*}})
 // CHECK-VECTOR: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, ptr %{{.*}}, i32 0, i32 0
 // CHECK-VECTOR: [[REG_COUNT:%[^ ]+]] = load i64, ptr [[REG_COUNT_PTR]]
 // CHECK-VECTOR: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
@@ -388,7 +388,7 @@ v32i8 va_v32i8(__builtin_va_list l) { return __builtin_va_arg(l, v32i8); }
 // CHECK-VECTOR: ret void
 
 struct agg_v1i8 va_agg_v1i8(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_v1i8); }
-// CHECK-LABEL: define{{.*}} void @va_agg_v1i8(ptr noalias sret(%struct.agg_v1i8) align 1 %{{.*}}, ptr %{{.*}})
+// CHECK-LABEL: define{{.*}} void @va_agg_v1i8(ptr dead_on_unwind noalias writable sret(%struct.agg_v1i8) align 1 %{{.*}}, ptr %{{.*}})
 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, ptr %{{.*}}, i32 0, i32 0
 // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, ptr [[REG_COUNT_PTR]]
 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
@@ -407,7 +407,7 @@ struct agg_v1i8 va_agg_v1i8(__builtin_va_list l) { return __builtin_va_arg(l, st
 // CHECK: store ptr [[OVERFLOW_ARG_AREA2]], ptr [[OVERFLOW_ARG_AREA_PTR]]
 // CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi ptr [ [[RAW_REG_ADDR]], %{{.*}} ], [ [[RAW_MEM_ADDR]], %{{.*}} ]
 // CHECK: ret void
-// CHECK-VECTOR-LABEL: define{{.*}} void @va_agg_v1i8(ptr noalias sret(%struct.agg_v1i8) align 1 %{{.*}}, ptr %{{.*}})
+// CHECK-VECTOR-LABEL: define{{.*}} void @va_agg_v1i8(ptr dead_on_unwind noalias writable sret(%struct.agg_v1i8) align 1 %{{.*}}, ptr %{{.*}})
 // CHECK-VECTOR: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, ptr %{{.*}}, i32 0, i32 2
 // CHECK-VECTOR: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load ptr, ptr [[OVERFLOW_ARG_AREA_PTR]]
 // CHECK-VECTOR: [[OVERFLOW_ARG_AREA1:%[^ ]+]] = getelementptr i8, ptr [[OVERFLOW_ARG_AREA]], i64 8
@@ -415,7 +415,7 @@ struct agg_v1i8 va_agg_v1i8(__builtin_va_list l) { return __builtin_va_arg(l, st
 // CHECK-VECTOR: ret void
 
 struct agg_v2i8 va_agg_v2i8(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_v2i8); }
-// CHECK-LABEL: define{{.*}} void @va_agg_v2i8(ptr noalias sret(%struct.agg_v2i8) align 2 %{{.*}}, ptr %{{.*}})
+// CHECK-LABEL: define{{.*}} void @va_agg_v2i8(ptr dead_on_unwind noalias writable sret(%struct.agg_v2i8) align 2 %{{.*}}, ptr %{{.*}})
 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, ptr %{{.*}}, i32 0, i32 0
 // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, ptr [[REG_COUNT_PTR]]
 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
@@ -434,7 +434,7 @@ struct agg_v2i8 va_agg_v2i8(__builtin_va_list l) { return __builtin_va_arg(l, st
 // CHECK: store ptr [[OVERFLOW_ARG_AREA2]], ptr [[OVERFLOW_ARG_AREA_PTR]]
 // CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi ptr [ [[RAW_REG_ADDR]], %{{.*}} ], [ [[RAW_MEM_ADDR]], %{{.*}} ]
 // CHECK: ret void
-// CHECK-VECTOR-LABEL: define{{.*}} void @va_agg_v2i8(ptr noalias sret(%struct.agg_v2i8) align 2 %{{.*}}, ptr %{{.*}})
+// CHECK-VECTOR-LABEL: define{{.*}} void @va_agg_v2i8(ptr dead_on_unwind noalias writable sret(%struct.agg_v2i8) align 2 %{{.*}}, ptr %{{.*}})
 // CHECK-VECTOR: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, ptr %{{.*}}, i32 0, i32 2
 // CHECK-VECTOR: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load ptr, ptr [[OVERFLOW_ARG_AREA_PTR]]
 // CHECK-VECTOR: [[OVERFLOW_ARG_AREA1:%[^ ]+]] = getelementptr i8, ptr [[OVERFLOW_ARG_AREA]], i64 8
@@ -442,7 +442,7 @@ struct agg_v2i8 va_agg_v2i8(__builtin_va_list l) { return __builtin_va_arg(l, st
 // CHECK-VECTOR: ret void
 
 struct agg_v4i8 va_agg_v4i8(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_v4i8); }
-// CHECK-LABEL: define{{.*}} void @va_agg_v4i8(ptr noalias sret(%struct.agg_v4i8) align 4 %{{.*}}, ptr %{{.*}})
+// CHECK-LABEL: define{{.*}} void @va_agg_v4i8(ptr dead_on_unwind noalias writable sret(%struct.agg_v4i8) align 4 %{{.*}}, ptr %{{.*}})
 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, ptr %{{.*}}, i32 0, i32 0
 // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, ptr [[REG_COUNT_PTR]]
 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
@@ -461,7 +461,7 @@ struct agg_v4i8 va_agg_v4i8(__builtin_va_list l) { return __builtin_va_arg(l, st
 // CHECK: store ptr [[OVERFLOW_ARG_AREA2]], ptr [[OVERFLOW_ARG_AREA_PTR]]
 // CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi ptr [ [[RAW_REG_ADDR]], %{{.*}} ], [ [[RAW_MEM_ADDR]], %{{.*}} ]
 // CHECK: ret void
-// CHECK-VECTOR-LABEL: define{{.*}} void @va_agg_v4i8(ptr noalias sret(%struct.agg_v4i8) align 4 %{{.*}}, ptr %{{.*}})
+// CHECK-VECTOR-LABEL: define{{.*}} void @va_agg_v4i8(ptr dead_on_unwind noalias writable sret(%struct.agg_v4i8) align 4 %{{.*}}, ptr %{{.*}})
 // CHECK-VECTOR: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, ptr %{{.*}}, i32 0, i32 2
 // CHECK-VECTOR: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load ptr, ptr [[OVERFLOW_ARG_AREA_PTR]]
 // CHECK-VECTOR: [[OVERFLOW_ARG_AREA1:%[^ ]+]] = getelementptr i8, ptr [[OVERFLOW_ARG_AREA]], i64 8
@@ -469,7 +469,7 @@ struct agg_v4i8 va_agg_v4i8(__builtin_va_list l) { return __builtin_va_arg(l, st
 // CHECK-VECTOR: ret void
 
 struct agg_v8i8 va_agg_v8i8(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_v8i8); }
-// CHECK-LABEL: define{{.*}} void @va_agg_v8i8(ptr noalias sret(%struct.agg_v8i8) align 8 %{{.*}}, ptr %{{.*}})
+// CHECK-LABEL: define{{.*}} void @va_agg_v8i8(ptr dead_on_unwind noalias writable sret(%struct.agg_v8i8) align 8 %{{.*}}, ptr %{{.*}})
 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, ptr %{{.*}}, i32 0, i32 0
 // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, ptr [[REG_COUNT_PTR]]
 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
@@ -488,7 +488,7 @@ struct agg_v8i8 va_agg_v8i8(__builtin_va_list l) { return __builtin_va_arg(l, st
 // CHECK: store ptr [[OVERFLOW_ARG_AREA2]], ptr [[OVERFLOW_ARG_AREA_PTR]]
 // CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi ptr [ [[RAW_REG_ADDR]], %{{.*}} ], [ [[RAW_MEM_ADDR]], %{{.*}} ]
 // CHECK: ret void
-// CHECK-VECTOR-LABEL: define{{.*}} void @va_agg_v8i8(ptr noalias sret(%struct.agg_v8i8) align 8 %{{.*}}, ptr %{{.*}})
+// CHECK-VECTOR-LABEL: define{{.*}} void @va_agg_v8i8(ptr dead_on_unwind noalias writable sret(%struct.agg_v8i8) align 8 %{{.*}}, ptr %{{.*}})
 // CHECK-VECTOR: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, ptr %{{.*}}, i32 0, i32 2
 // CHECK-VECTOR: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load ptr, ptr [[OVERFLOW_ARG_AREA_PTR]]
 // CHECK-VECTOR: [[OVERFLOW_ARG_AREA1:%[^ ]+]] = getelementptr i8, ptr [[OVERFLOW_ARG_AREA]], i64 8
@@ -496,7 +496,7 @@ struct agg_v8i8 va_agg_v8i8(__builtin_va_list l) { return __builtin_va_arg(l, st
 // CHECK-VECTOR: ret void
 
 struct agg_v16i8 va_agg_v16i8(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_v16i8); }
-// CHECK-LABEL: define{{.*}} void @va_agg_v16i8(ptr noalias sret(%struct.agg_v16i8) align 16 %{{.*}}, ptr %{{.*}})
+// CHECK-LABEL: define{{.*}} void @va_agg_v16i8(ptr dead_on_unwind noalias writable sret(%struct.agg_v16i8) align 16 %{{.*}}, ptr %{{.*}})
 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, ptr %{{.*}}, i32 0, i32 0
 // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, ptr [[REG_COUNT_PTR]]
 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
@@ -516,7 +516,7 @@ struct agg_v16i8 va_agg_v16i8(__builtin_va_list l) { return __builtin_va_arg(l,
 // CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi ptr [ [[RAW_REG_ADDR]], %{{.*}} ], [ [[RAW_MEM_ADDR]], %{{.*}} ]
 // CHECK: [[INDIRECT_ARG:%[^ ]+]] = load ptr, ptr [[VA_ARG_ADDR]]
 // CHECK: ret void
-// CHECK-VECTOR-LABEL: define{{.*}} void @va_agg_v16i8(ptr noalias sret(%struct.agg_v16i8) align 8 %{{.*}}, ptr %{{.*}})
+// CHECK-VECTOR-LABEL: define{{.*}} void @va_agg_v16i8(ptr dead_on_unwind noalias writable sret(%struct.agg_v16i8) align 8 %{{.*}}, ptr %{{.*}})
 // CHECK-VECTOR: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, ptr %{{.*}}, i32 0, i32 2
 // CHECK-VECTOR: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load ptr, ptr [[OVERFLOW_ARG_AREA_PTR]]
 // CHECK-VECTOR: [[OVERFLOW_ARG_AREA1:%[^ ]+]] = getelementptr i8, ptr [[OVERFLOW_ARG_AREA]], i64 16
@@ -524,7 +524,7 @@ struct agg_v16i8 va_agg_v16i8(__builtin_va_list l) { return __builtin_va_arg(l,
 // CHECK-VECTOR: ret void
 
 struct agg_v32i8 va_agg_v32i8(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_v32i8); }
-// CHECK-LABEL: define{{.*}} void @va_agg_v32i8(ptr noalias sret(%struct.agg_v32i8) align 32 %{{.*}}, ptr %{{.*}})
+// CHECK-LABEL: define{{.*}} void @va_agg_v32i8(ptr dead_on_unwind noalias writable sret(%struct.agg_v32i8) align 32 %{{.*}}, ptr %{{.*}})
 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, ptr %{{.*}}, i32 0, i32 0
 // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, ptr [[REG_COUNT_PTR]]
 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
@@ -544,7 +544,7 @@ struct agg_v32i8 va_agg_v32i8(__builtin_va_list l) { return __builtin_va_arg(l,
 // CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi ptr [ [[RAW_REG_ADDR]], %{{.*}} ], [ [[RAW_MEM_ADDR]], %{{.*}} ]
 // CHECK: [[INDIRECT_ARG:%[^ ]+]] = load ptr, ptr [[VA_ARG_ADDR]]
 // CHECK: ret void
-// CHECK-VECTOR-LABEL: define{{.*}} void @va_agg_v32i8(ptr noalias sret(%struct.agg_v32i8) align 8 %{{.*}}, ptr %{{.*}})
+// CHECK-VECTOR-LABEL: define{{.*}} void @va_agg_v32i8(ptr dead_on_unwind noalias writable sret(%struct.agg_v32i8) align 8 %{{.*}}, ptr %{{.*}})
 // CHECK-VECTOR: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, ptr %{{.*}}, i32 0, i32 0
 // CHECK-VECTOR: [[REG_COUNT:%[^ ]+]] = load i64, ptr [[REG_COUNT_PTR]]
 // CHECK-VECTOR: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
diff --git a/clang/test/CodeGen/SystemZ/systemz-abi.c b/clang/test/CodeGen/SystemZ/systemz-abi.c
index 7b2c04ece185f..65a2bc9bbb680 100644
--- a/clang/test/CodeGen/SystemZ/systemz-abi.c
+++ b/clang/test/CodeGen/SystemZ/systemz-abi.c
@@ -43,7 +43,7 @@ long long pass_longlong(long long arg) { return arg; }
 // CHECK-LABEL: define{{.*}} i64 @pass_longlong(i64 %{{.*}})
 
 __int128 pass_int128(__int128 arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_int128(ptr noalias sret(i128) align 8 %{{.*}}, ptr %0)
+// CHECK-LABEL: define{{.*}} void @pass_int128(ptr dead_on_unwind noalias writable sret(i128) align 8 %{{.*}}, ptr %0)
 
 float pass_float(float arg) { return arg; }
 // CHECK-LABEL: define{{.*}} float @pass_float(float %{{.*}})
@@ -52,125 +52,125 @@ double pass_double(double arg) { return arg; }
 // CHECK-LABEL: define{{.*}} double @pass_double(double %{{.*}})
 
 long double pass_longdouble(long double arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_longdouble(ptr noalias sret(fp128) align 8 %{{.*}}, ptr %0)
+// CHECK-LABEL: define{{.*}} void @pass_longdouble(ptr dead_on_unwind noalias writable sret(fp128) align 8 %{{.*}}, ptr %0)
 
 
 // Complex types
 
 _Complex char pass_complex_char(_Complex char arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_complex_char(ptr noalias sret({ i8, i8 }) align 1 %{{.*}}, ptr %{{.*}}arg)
+// CHECK-LABEL: define{{.*}} void @pass_complex_char(ptr dead_on_unwind noalias writable sret({ i8, i8 }) align 1 %{{.*}}, ptr %{{.*}}arg)
 
 _Complex short pass_complex_short(_Complex short arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_complex_short(ptr noalias sret({ i16, i16 }) align 2 %{{.*}}, ptr %{{.*}}arg)
+// CHECK-LABEL: define{{.*}} void @pass_complex_short(ptr dead_on_unwind noalias writable sret({ i16, i16 }) align 2 %{{.*}}, ptr %{{.*}}arg)
 
 _Complex int pass_complex_int(_Complex int arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_complex_int(ptr noalias sret({ i32, i32 }) align 4 %{{.*}}, ptr %{{.*}}arg)
+// CHECK-LABEL: define{{.*}} void @pass_complex_int(ptr dead_on_unwind noalias writable sret({ i32, i32 }) align 4 %{{.*}}, ptr %{{.*}}arg)
 
 _Complex long pass_complex_long(_Complex long arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_complex_long(ptr noalias sret({ i64, i64 }) align 8 %{{.*}}, ptr %{{.*}}arg)
+// CHECK-LABEL: define{{.*}} void @pass_complex_long(ptr dead_on_unwind noalias writable sret({ i64, i64 }) align 8 %{{.*}}, ptr %{{.*}}arg)
 
 _Complex long long pass_complex_longlong(_Complex long long arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_complex_longlong(ptr noalias sret({ i64, i64 }) align 8 %{{.*}}, ptr %{{.*}}arg)
+// CHECK-LABEL: define{{.*}} void @pass_complex_longlong(ptr dead_on_unwind noalias writable sret({ i64, i64 }) align 8 %{{.*}}, ptr %{{.*}}arg)
 
 _Complex float pass_complex_float(_Complex float arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_complex_float(ptr noalias sret({ float, float }) align 4 %{{.*}}, ptr %{{.*}}arg)
+// CHECK-LABEL: define{{.*}} void @pass_complex_float(ptr dead_on_unwind noalias writable sret({ float, float }) align 4 %{{.*}}, ptr %{{.*}}arg)
 
 _Complex double pass_complex_double(_Complex double arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_complex_double(ptr noalias sret({ double, double }) align 8 %{{.*}}, ptr %{{.*}}arg)
+// CHECK-LABEL: define{{.*}} void @pass_complex_double(ptr dead_on_unwind noalias writable sret({ double, double }) align 8 %{{.*}}, ptr %{{.*}}arg)
 
 _Complex long double pass_complex_longdouble(_Complex long double arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_complex_longdouble(ptr noalias sret({ fp128, fp128 }) align 8 %{{.*}}, ptr %{{.*}}arg)
+// CHECK-LABEL: define{{.*}} void @pass_complex_longdouble(ptr dead_on_unwind noalias writable sret({ fp128, fp128 }) align 8 %{{.*}}, ptr %{{.*}}arg)
 
 
 // Aggregate types
 
 struct agg_1byte { char a[1]; };
 struct agg_1byte pass_agg_1byte(struct agg_1byte arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_agg_1byte(ptr noalias sret(%struct.agg_1byte) align 1 %{{.*}}, i8 %{{.*}})
+// CHECK-LABEL: define{{.*}} void @pass_agg_1byte(ptr dead_on_unwind noalias writable sret(%struct.agg_1byte) align 1 %{{.*}}, i8 %{{.*}})
 
 struct agg_2byte { char a[2]; };
 struct agg_2byte pass_agg_2byte(struct agg_2byte arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_agg_2byte(ptr noalias sret(%struct.agg_2byte) align 1 %{{.*}}, i16 %{{.*}})
+// CHECK-LABEL: define{{.*}} void @pass_agg_2byte(ptr dead_on_unwind noalias writable sret(%struct.agg_2byte) align 1 %{{.*}}, i16 %{{.*}})
 
 struct agg_3byte { char a[3]; };
 struct agg_3byte pass_agg_3byte(struct agg_3byte arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_agg_3byte(ptr noalias sret(%struct.agg_3byte) align 1 %{{.*}}, ptr %{{.*}})
+// CHECK-LABEL: define{{.*}} void @pass_agg_3byte(ptr dead_on_unwind noalias writable sret(%struct.agg_3byte) align 1 %{{.*}}, ptr %{{.*}})
 
 struct agg_4byte { char a[4]; };
 struct agg_4byte pass_agg_4byte(struct agg_4byte arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_agg_4byte(ptr noalias sret(%struct.agg_4byte) align 1 %{{.*}}, i32 %{{.*}})
+// CHECK-LABEL: define{{.*}} void @pass_agg_4byte(ptr dead_on_unwind noalias writable sret(%struct.agg_4byte) align 1 %{{.*}}, i32 %{{.*}})
 
 struct agg_5byte { char a[5]; };
 struct agg_5byte pass_agg_5byte(struct agg_5byte arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_agg_5byte(ptr noalias sret(%struct.agg_5byte) align 1 %{{.*}}, ptr %{{.*}})
+// CHECK-LABEL: define{{.*}} void @pass_agg_5byte(ptr dead_on_unwind noalias writable sret(%struct.agg_5byte) align 1 %{{.*}}, ptr %{{.*}})
 
 struct agg_6byte { char a[6]; };
 struct agg_6byte pass_agg_6byte(struct agg_6byte arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_agg_6byte(ptr noalias sret(%struct.agg_6byte) align 1 %{{.*}}, ptr %{{.*}})
+// CHECK-LABEL: define{{.*}} void @pass_agg_6byte(ptr dead_on_unwind noalias writable sret(%struct.agg_6byte) align 1 %{{.*}}, ptr %{{.*}})
 
 struct agg_7byte { char a[7]; };
 struct agg_7byte pass_agg_7byte(struct agg_7byte arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_agg_7byte(ptr noalias sret(%struct.agg_7byte) align 1 %{{.*}}, ptr %{{.*}})
+// CHECK-LABEL: define{{.*}} void @pass_agg_7byte(ptr dead_on_unwind noalias writable sret(%struct.agg_7byte) align 1 %{{.*}}, ptr %{{.*}})
 
 struct agg_8byte { char a[8]; };
 struct agg_8byte pass_agg_8byte(struct agg_8byte arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_agg_8byte(ptr noalias sret(%struct.agg_8byte) align 1 %{{.*}}, i64 %{{.*}})
+// CHECK-LABEL: define{{.*}} void @pass_agg_8byte(ptr dead_on_unwind noalias writable sret(%struct.agg_8byte) align 1 %{{.*}}, i64 %{{.*}})
 
 struct agg_16byte { char a[16]; };
 struct agg_16byte pass_agg_16byte(struct agg_16byte arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_agg_16byte(ptr noalias sret(%struct.agg_16byte) align 1 %{{.*}}, ptr %{{.*}})
+// CHECK-LABEL: define{{.*}} void @pass_agg_16byte(ptr dead_on_unwind noalias writable sret(%struct.agg_16byte) align 1 %{{.*}}, ptr %{{.*}})
 
 
 // Float-like aggregate types
 
 struct agg_float { float a; };
 struct agg_float pass_agg_float(struct agg_float arg) { return arg; }
-// HARD-FLOAT-LABEL: define{{.*}} void @pass_agg_float(ptr noalias sret(%struct.agg_float) align 4 %{{.*}}, float %{{.*}})
-// SOFT-FLOAT-LABEL: define{{.*}} void @pass_agg_float(ptr noalias sret(%struct.agg_float) align 4 %{{.*}}, i32 %{{.*}})
+// HARD-FLOAT-LABEL: define{{.*}} void @pass_agg_float(ptr dead_on_unwind noalias writable sret(%struct.agg_float) align 4 %{{.*}}, float %{{.*}})
+// SOFT-FLOAT-LABEL: define{{.*}} void @pass_agg_float(ptr dead_on_unwind noalias writable sret(%struct.agg_float) align 4 %{{.*}}, i32 %{{.*}})
 
 struct agg_double { double a; };
 struct agg_double pass_agg_double(struct agg_double arg) { return arg; }
-// HARD-FLOAT-LABEL: define{{.*}} void @pass_agg_double(ptr noalias sret(%struct.agg_double) align 8 %{{.*}}, double %{{.*}})
-// SOFT-FLOAT-LABEL: define{{.*}} void @pass_agg_double(ptr noalias sret(%struct.agg_double) align 8 %{{.*}}, i64 %{{.*}})
+// HARD-FLOAT-LABEL: define{{.*}} void @pass_agg_double(ptr dead_on_unwind noalias writable sret(%struct.agg_double) align 8 %{{.*}}, double %{{.*}})
+// SOFT-FLOAT-LABEL: define{{.*}} void @pass_agg_double(ptr dead_on_unwind noalias writable sret(%struct.agg_double) align 8 %{{.*}}, i64 %{{.*}})
 
 struct agg_longdouble { long double a; };
 struct agg_longdouble pass_agg_longdouble(struct agg_longdouble arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_agg_longdouble(ptr noalias sret(%struct.agg_longdouble) align 8 %{{.*}}, ptr %{{.*}})
+// CHECK-LABEL: define{{.*}} void @pass_agg_longdouble(ptr dead_on_unwind noalias writable sret(%struct.agg_longdouble) align 8 %{{.*}}, ptr %{{.*}})
 
 struct agg_float_a8 { float a __attribute__((aligned (8))); };
 struct agg_float_a8 pass_agg_float_a8(struct agg_float_a8 arg) { return arg; }
-// HARD-FLOAT-LABEL: define{{.*}} void @pass_agg_float_a8(ptr noalias sret(%struct.agg_float_a8) align 8 %{{.*}}, double %{{.*}})
-// SOFT-FLOAT-LABEL: define{{.*}} void @pass_agg_float_a8(ptr noalias sret(%struct.agg_float_a8) align 8 %{{.*}}, i64 %{{.*}})
+// HARD-FLOAT-LABEL: define{{.*}} void @pass_agg_float_a8(ptr dead_on_unwind noalias writable sret(%struct.agg_float_a8) align 8 %{{.*}}, double %{{.*}})
+// SOFT-FLOAT-LABEL: define{{.*}} void @pass_agg_float_a8(ptr dead_on_unwind noalias writable sret(%struct.agg_float_a8) align 8 %{{.*}}, i64 %{{.*}})
 
 struct agg_float_a16 { float a __attribute__((aligned (16))); };
 struct agg_float_a16 pass_agg_float_a16(struct agg_float_a16 arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_agg_float_a16(ptr noalias sret(%struct.agg_float_a16) align 16 %{{.*}}, ptr %{{.*}})
+// CHECK-LABEL: define{{.*}} void @pass_agg_float_a16(ptr dead_on_unwind noalias writable sret(%struct.agg_float_a16) align 16 %{{.*}}, ptr %{{.*}})
 
 
 // Verify that the following are *not* float-like aggregate types
 
 struct agg_nofloat1 { float a; float b; };
 struct agg_nofloat1 pass_agg_nofloat1(struct agg_nofloat1 arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_agg_nofloat1(ptr noalias sret(%struct.agg_nofloat1) align 4 %{{.*}}, i64 %{{.*}})
+// CHECK-LABEL: define{{.*}} void @pass_agg_nofloat1(ptr dead_on_unwind noalias writable sret(%struct.agg_nofloat1) align 4 %{{.*}}, i64 %{{.*}})
 
 struct agg_nofloat2 { float a; int b; };
 struct agg_nofloat2 pass_agg_nofloat2(struct agg_nofloat2 arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_agg_nofloat2(ptr noalias sret(%struct.agg_nofloat2) align 4 %{{.*}}, i64 %{{.*}})
+// CHECK-LABEL: define{{.*}} void @pass_agg_nofloat2(ptr dead_on_unwind noalias writable sret(%struct.agg_nofloat2) align 4 %{{.*}}, i64 %{{.*}})
 
 struct agg_nofloat3 { float a; int : 0; };
 struct agg_nofloat3 pass_agg_nofloat3(struct agg_nofloat3 arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_agg_nofloat3(ptr noalias sret(%struct.agg_nofloat3) align 4 %{{.*}}, i32 %{{.*}})
+// CHECK-LABEL: define{{.*}} void @pass_agg_nofloat3(ptr dead_on_unwind noalias writable sret(%struct.agg_nofloat3) align 4 %{{.*}}, i32 %{{.*}})
 
 
 // Union types likewise are *not* float-like aggregate types
 
 union union_float { float a; };
 union union_float pass_union_float(union union_float arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_union_float(ptr noalias sret(%union.union_float) align 4 %{{.*}}, i32 %{{.*}})
+// CHECK-LABEL: define{{.*}} void @pass_union_float(ptr dead_on_unwind noalias writable sret(%union.union_float) align 4 %{{.*}}, i32 %{{.*}})
 
 union union_double { double a; };
 union union_double pass_union_double(union union_double arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @pass_union_double(ptr noalias sret(%union.union_double) align 8 %{{.*}}, i64 %{{.*}})
+// CHECK-LABEL: define{{.*}} void @pass_union_double(ptr dead_on_unwind noalias writable sret(%union.union_double) align 8 %{{.*}}, i64 %{{.*}})
 
 
 // Accessing variable argument lists
@@ -267,7 +267,7 @@ double va_double(__builtin_va_list l) { return __builtin_va_arg(l, double); }
 // CHECK: ret double [[RET]]
 
 long double va_longdouble(__builtin_va_list l) { return __builtin_va_arg(l, long double); }
-// CHECK-LABEL: define{{.*}} void @va_longdouble(ptr noalias sret(fp128) align 8 %{{.*}}, ptr %{{.*}})
+// CHECK-LABEL: define{{.*}} void @va_longdouble(ptr dead_on_unwind noalias writable sret(fp128) align 8 %{{.*}}, ptr %{{.*}})
 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, ptr %{{.*}}, i32 0, i32 0
 // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, ptr [[REG_COUNT_PTR]]
 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
@@ -291,7 +291,7 @@ long double va_longdouble(__builtin_va_list l) { return __builtin_va_arg(l, long
 // CHECK: ret void
 
 _Complex char va_complex_char(__builtin_va_list l) { return __builtin_va_arg(l, _Complex char); }
-// CHECK-LABEL: define{{.*}} void @va_complex_char(ptr noalias sret({ i8, i8 }) align 1 %{{.*}}, ptr %{{.*}}
+// CHECK-LABEL: define{{.*}} void @va_complex_char(ptr dead_on_unwind noalias writable sret({ i8, i8 }) align 1 %{{.*}}, ptr %{{.*}}
 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, ptr %{{.*}}, i32 0, i32 0
 // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, ptr [[REG_COUNT_PTR]]
 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
@@ -313,7 +313,7 @@ _Complex char va_complex_char(__builtin_va_list l) { return __builtin_va_arg(l,
 // CHECK: ret void
 
 struct agg_1byte va_agg_1byte(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_1byte); }
-// CHECK-LABEL: define{{.*}} void @va_agg_1byte(ptr noalias sret(%struct.agg_1byte) align 1 %{{.*}}, ptr %{{.*}}
+// CHECK-LABEL: define{{.*}} void @va_agg_1byte(ptr dead_on_unwind noalias writable sret(%struct.agg_1byte) align 1 %{{.*}}, ptr %{{.*}}
 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, ptr %{{.*}}, i32 0, i32 0
 // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, ptr [[REG_COUNT_PTR]]
 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
@@ -334,7 +334,7 @@ struct agg_1byte va_agg_1byte(__builtin_va_list l) { return __builtin_va_arg(l,
 // CHECK: ret void
 
 struct agg_2byte va_agg_2byte(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_2byte); }
-// CHECK-LABEL: define{{.*}} void @va_agg_2byte(ptr noalias sret(%struct.agg_2byte) align 1 %{{.*}}, ptr %{{.*}}
+// CHECK-LABEL: define{{.*}} void @va_agg_2byte(ptr dead_on_unwind noalias writable sret(%struct.agg_2byte) align 1 %{{.*}}, ptr %{{.*}}
 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, ptr %{{.*}}, i32 0, i32 0
 // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, ptr [[REG_COUNT_PTR]]
 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
@@ -355,7 +355,7 @@ struct agg_2byte va_agg_2byte(__builtin_va_list l) { return __builtin_va_arg(l,
 // CHECK: ret void
 
 struct agg_3byte va_agg_3byte(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_3byte); }
-// CHECK-LABEL: define{{.*}} void @va_agg_3byte(ptr noalias sret(%struct.agg_3byte) align 1 %{{.*}}, ptr %{{.*}}
+// CHECK-LABEL: define{{.*}} void @va_agg_3byte(ptr dead_on_unwind noalias writable sret(%struct.agg_3byte) align 1 %{{.*}}, ptr %{{.*}}
 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, ptr %{{.*}}, i32 0, i32 0
 // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, ptr [[REG_COUNT_PTR]]
 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
@@ -377,7 +377,7 @@ struct agg_3byte va_agg_3byte(__builtin_va_list l) { return __builtin_va_arg(l,
 // CHECK: ret void
 
 struct agg_4byte va_agg_4byte(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_4byte); }
-// CHECK-LABEL: define{{.*}} void @va_agg_4byte(ptr noalias sret(%struct.agg_4byte) align 1 %{{.*}}, ptr %{{.*}}
+// CHECK-LABEL: define{{.*}} void @va_agg_4byte(ptr dead_on_unwind noalias writable sret(%struct.agg_4byte) align 1 %{{.*}}, ptr %{{.*}}
 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, ptr %{{.*}}, i32 0, i32 0
 // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, ptr [[REG_COUNT_PTR]]
 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
@@ -398,7 +398,7 @@ struct agg_4byte va_agg_4byte(__builtin_va_list l) { return __builtin_va_arg(l,
 // CHECK: ret void
 
 struct agg_8byte va_agg_8byte(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_8byte); }
-// CHECK-LABEL: define{{.*}} void @va_agg_8byte(ptr noalias sret(%struct.agg_8byte) align 1 %{{.*}}, ptr %{{.*}}
+// CHECK-LABEL: define{{.*}} void @va_agg_8byte(ptr dead_on_unwind noalias writable sret(%struct.agg_8byte) align 1 %{{.*}}, ptr %{{.*}}
 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, ptr %{{.*}}, i32 0, i32 0
 // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, ptr [[REG_COUNT_PTR]]
 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
@@ -419,7 +419,7 @@ struct agg_8byte va_agg_8byte(__builtin_va_list l) { return __builtin_va_arg(l,
 // CHECK: ret void
 
 struct agg_float va_agg_float(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_float); }
-// CHECK-LABEL: define{{.*}} void @va_agg_float(ptr noalias sret(%struct.agg_float) align 4 %{{.*}}, ptr %{{.*}}
+// CHECK-LABEL: define{{.*}} void @va_agg_float(ptr dead_on_unwind noalias writable sret(%struct.agg_float) align 4 %{{.*}}, ptr %{{.*}}
 // HARD-FLOAT: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, ptr %{{.*}}, i32 0, i32 1
 // SOFT-FLOAT: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, ptr %{{.*}}, i32 0, i32 0
 // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, ptr [[REG_COUNT_PTR]]
@@ -443,7 +443,7 @@ struct agg_float va_agg_float(__builtin_va_list l) { return __builtin_va_arg(l,
 // CHECK: ret void
 
 struct agg_double va_agg_double(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_double); }
-// CHECK-LABEL: define{{.*}} void @va_agg_double(ptr noalias sret(%struct.agg_double) align 8 %{{.*}}, ptr %{{.*}}
+// CHECK-LABEL: define{{.*}} void @va_agg_double(ptr dead_on_unwind noalias writable sret(%struct.agg_double) align 8 %{{.*}}, ptr %{{.*}}
 // HARD-FLOAT: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, ptr %{{.*}}, i32 0, i32 1
 // SOFT-FLOAT: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, ptr %{{.*}}, i32 0, i32 0
 // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, ptr [[REG_COUNT_PTR]]
@@ -467,7 +467,7 @@ struct agg_double va_agg_double(__builtin_va_list l) { return __builtin_va_arg(l
 // CHECK: ret void
 
 struct agg_longdouble va_agg_longdouble(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_longdouble); }
-// CHECK-LABEL: define{{.*}} void @va_agg_longdouble(ptr noalias sret(%struct.agg_longdouble) align 8 %{{.*}}, ptr %{{.*}}
+// CHECK-LABEL: define{{.*}} void @va_agg_longdouble(ptr dead_on_unwind noalias writable sret(%struct.agg_longdouble) align 8 %{{.*}}, ptr %{{.*}}
 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, ptr %{{.*}}, i32 0, i32 0
 // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, ptr [[REG_COUNT_PTR]]
 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
@@ -489,7 +489,7 @@ struct agg_longdouble va_agg_longdouble(__builtin_va_list l) { return __builtin_
 // CHECK: ret void
 
 struct agg_float_a8 va_agg_float_a8(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_float_a8); }
-// CHECK-LABEL: define{{.*}} void @va_agg_float_a8(ptr noalias sret(%struct.agg_float_a8) align 8 %{{.*}}, ptr %{{.*}}
+// CHECK-LABEL: define{{.*}} void @va_agg_float_a8(ptr dead_on_unwind noalias writable sret(%struct.agg_float_a8) align 8 %{{.*}}, ptr %{{.*}}
 // HARD-FLOAT: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, ptr %{{.*}}, i32 0, i32 1
 // SOFT-FLOAT: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, ptr %{{.*}}, i32 0, i32 0
 // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, ptr [[REG_COUNT_PTR]]
@@ -513,7 +513,7 @@ struct agg_float_a8 va_agg_float_a8(__builtin_va_list l) { return __builtin_va_a
 // CHECK: ret void
 
 struct agg_float_a16 va_agg_float_a16(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_float_a16); }
-// CHECK-LABEL: define{{.*}} void @va_agg_float_a16(ptr noalias sret(%struct.agg_float_a16) align 16 %{{.*}}, ptr %{{.*}}
+// CHECK-LABEL: define{{.*}} void @va_agg_float_a16(ptr dead_on_unwind noalias writable sret(%struct.agg_float_a16) align 16 %{{.*}}, ptr %{{.*}}
 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, ptr %{{.*}}, i32 0, i32 0
 // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, ptr [[REG_COUNT_PTR]]
 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
@@ -535,7 +535,7 @@ struct agg_float_a16 va_agg_float_a16(__builtin_va_list l) { return __builtin_va
 // CHECK: ret void
 
 struct agg_nofloat1 va_agg_nofloat1(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_nofloat1); }
-// CHECK-LABEL: define{{.*}} void @va_agg_nofloat1(ptr noalias sret(%struct.agg_nofloat1) align 4 %{{.*}}, ptr %{{.*}}
+// CHECK-LABEL: define{{.*}} void @va_agg_nofloat1(ptr dead_on_unwind noalias writable sret(%struct.agg_nofloat1) align 4 %{{.*}}, ptr %{{.*}}
 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, ptr %{{.*}}, i32 0, i32 0
 // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, ptr [[REG_COUNT_PTR]]
 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
@@ -556,7 +556,7 @@ struct agg_nofloat1 va_agg_nofloat1(__builtin_va_list l) { return __builtin_va_a
 // CHECK: ret void
 
 struct agg_nofloat2 va_agg_nofloat2(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_nofloat2); }
-// CHECK-LABEL: define{{.*}} void @va_agg_nofloat2(ptr noalias sret(%struct.agg_nofloat2) align 4 %{{.*}}, ptr %{{.*}}
+// CHECK-LABEL: define{{.*}} void @va_agg_nofloat2(ptr dead_on_unwind noalias writable sret(%struct.agg_nofloat2) align 4 %{{.*}}, ptr %{{.*}}
 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, ptr %{{.*}}, i32 0, i32 0
 // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, ptr [[REG_COUNT_PTR]]
 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
@@ -577,7 +577,7 @@ struct agg_nofloat2 va_agg_nofloat2(__builtin_va_list l) { return __builtin_va_a
 // CHECK: ret void
 
 struct agg_nofloat3 va_agg_nofloat3(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_nofloat3); }
-// CHECK-LABEL: define{{.*}} void @va_agg_nofloat3(ptr noalias sret(%struct.agg_nofloat3) align 4 %{{.*}}, ptr %{{.*}}
+// CHECK-LABEL: define{{.*}} void @va_agg_nofloat3(ptr dead_on_unwind noalias writable sret(%struct.agg_nofloat3) align 4 %{{.*}}, ptr %{{.*}}
 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, ptr %{{.*}}, i32 0, i32 0
 // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, ptr [[REG_COUNT_PTR]]
 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
diff --git a/clang/test/CodeGen/SystemZ/systemz-abi.cpp b/clang/test/CodeGen/SystemZ/systemz-abi.cpp
index 08bf08c3803f4..06be85421ba17 100644
--- a/clang/test/CodeGen/SystemZ/systemz-abi.cpp
+++ b/clang/test/CodeGen/SystemZ/systemz-abi.cpp
@@ -6,20 +6,20 @@
 
 class agg_float_class { float a; };
 class agg_float_class pass_agg_float_class(class agg_float_class arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @_Z20pass_agg_float_class15agg_float_class(ptr noalias sret(%class.agg_float_class) align 4 %{{.*}}, float %{{.*}})
-// SOFT-FLOAT-LABEL: define{{.*}} void @_Z20pass_agg_float_class15agg_float_class(ptr noalias sret(%class.agg_float_class) align 4 %{{.*}}, i32 %{{.*}})
+// CHECK-LABEL: define{{.*}} void @_Z20pass_agg_float_class15agg_float_class(ptr dead_on_unwind noalias writable sret(%class.agg_float_class) align 4 %{{.*}}, float %{{.*}})
+// SOFT-FLOAT-LABEL: define{{.*}} void @_Z20pass_agg_float_class15agg_float_class(ptr dead_on_unwind noalias writable sret(%class.agg_float_class) align 4 %{{.*}}, i32 %{{.*}})
 
 class agg_double_class { double a; };
 class agg_double_class pass_agg_double_class(class agg_double_class arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @_Z21pass_agg_double_class16agg_double_class(ptr noalias sret(%class.agg_double_class) align 8 %{{.*}}, double %{{.*}})
-// SOFT-FLOAT-LABEL: define{{.*}} void @_Z21pass_agg_double_class16agg_double_class(ptr noalias sret(%class.agg_double_class) align 8 %{{.*}}, i64 %{{.*}})
+// CHECK-LABEL: define{{.*}} void @_Z21pass_agg_double_class16agg_double_class(ptr dead_on_unwind noalias writable sret(%class.agg_double_class) align 8 %{{.*}}, double %{{.*}})
+// SOFT-FLOAT-LABEL: define{{.*}} void @_Z21pass_agg_double_class16agg_double_class(ptr dead_on_unwind noalias writable sret(%class.agg_double_class) align 8 %{{.*}}, i64 %{{.*}})
 
 
 // This structure is passed in a GPR in C++ (and C, checked in systemz-abi.c).
 struct agg_float_cpp { float a; int : 0; };
 struct agg_float_cpp pass_agg_float_cpp(struct agg_float_cpp arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @_Z18pass_agg_float_cpp13agg_float_cpp(ptr noalias sret(%struct.agg_float_cpp) align 4 %{{.*}}, i32 %{{.*}})
-// SOFT-FLOAT-LABEL:  define{{.*}} void @_Z18pass_agg_float_cpp13agg_float_cpp(ptr noalias sret(%struct.agg_float_cpp) align 4 %{{.*}}, i32 %{{.*}})
+// CHECK-LABEL: define{{.*}} void @_Z18pass_agg_float_cpp13agg_float_cpp(ptr dead_on_unwind noalias writable sret(%struct.agg_float_cpp) align 4 %{{.*}}, i32 %{{.*}})
+// SOFT-FLOAT-LABEL:  define{{.*}} void @_Z18pass_agg_float_cpp13agg_float_cpp(ptr dead_on_unwind noalias writable sret(%struct.agg_float_cpp) align 4 %{{.*}}, i32 %{{.*}})
 
 
 // A field member of empty class type in C++ makes the record nonhomogeneous,
@@ -27,31 +27,31 @@ struct agg_float_cpp pass_agg_float_cpp(struct agg_float_cpp arg) { return arg;
 struct empty { };
 struct agg_nofloat_empty { float a; empty dummy; };
 struct agg_nofloat_empty pass_agg_nofloat_empty(struct agg_nofloat_empty arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @_Z22pass_agg_nofloat_empty17agg_nofloat_empty(ptr noalias sret(%struct.agg_nofloat_empty) align 4 %{{.*}}, i64 %{{.*}})
-// SOFT-FLOAT-LABEL:  define{{.*}} void @_Z22pass_agg_nofloat_empty17agg_nofloat_empty(ptr noalias sret(%struct.agg_nofloat_empty) align 4 %{{.*}}, i64 %{{.*}})
+// CHECK-LABEL: define{{.*}} void @_Z22pass_agg_nofloat_empty17agg_nofloat_empty(ptr dead_on_unwind noalias writable sret(%struct.agg_nofloat_empty) align 4 %{{.*}}, i64 %{{.*}})
+// SOFT-FLOAT-LABEL:  define{{.*}} void @_Z22pass_agg_nofloat_empty17agg_nofloat_empty(ptr dead_on_unwind noalias writable sret(%struct.agg_nofloat_empty) align 4 %{{.*}}, i64 %{{.*}})
 struct agg_float_empty { float a; [[no_unique_address]] empty dummy; };
 struct agg_float_empty pass_agg_float_empty(struct agg_float_empty arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @_Z20pass_agg_float_empty15agg_float_empty(ptr noalias sret(%struct.agg_float_empty) align 4 %{{.*}}, float %{{.*}})
-// SOFT-FLOAT-LABEL:  define{{.*}} void @_Z20pass_agg_float_empty15agg_float_empty(ptr noalias sret(%struct.agg_float_empty) align 4 %{{.*}}, i32 %{{.*}})
+// CHECK-LABEL: define{{.*}} void @_Z20pass_agg_float_empty15agg_float_empty(ptr dead_on_unwind noalias writable sret(%struct.agg_float_empty) align 4 %{{.*}}, float %{{.*}})
+// SOFT-FLOAT-LABEL:  define{{.*}} void @_Z20pass_agg_float_empty15agg_float_empty(ptr dead_on_unwind noalias writable sret(%struct.agg_float_empty) align 4 %{{.*}}, i32 %{{.*}})
 struct agg_nofloat_emptyarray { float a; [[no_unique_address]] empty dummy[3]; };
 struct agg_nofloat_emptyarray pass_agg_nofloat_emptyarray(struct agg_nofloat_emptyarray arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @_Z27pass_agg_nofloat_emptyarray22agg_nofloat_emptyarray(ptr noalias sret(%struct.agg_nofloat_emptyarray) align 4 %{{.*}}, i64 %{{.*}})
-// SOFT-FLOAT-LABEL:  define{{.*}} void @_Z27pass_agg_nofloat_emptyarray22agg_nofloat_emptyarray(ptr noalias sret(%struct.agg_nofloat_emptyarray) align 4 %{{.*}}, i64 %{{.*}})
+// CHECK-LABEL: define{{.*}} void @_Z27pass_agg_nofloat_emptyarray22agg_nofloat_emptyarray(ptr dead_on_unwind noalias writable sret(%struct.agg_nofloat_emptyarray) align 4 %{{.*}}, i64 %{{.*}})
+// SOFT-FLOAT-LABEL:  define{{.*}} void @_Z27pass_agg_nofloat_emptyarray22agg_nofloat_emptyarray(ptr dead_on_unwind noalias writable sret(%struct.agg_nofloat_emptyarray) align 4 %{{.*}}, i64 %{{.*}})
 
 // And likewise for members of base classes.
 struct noemptybase { empty dummy; };
 struct agg_nofloat_emptybase : noemptybase { float a; };
 struct agg_nofloat_emptybase pass_agg_nofloat_emptybase(struct agg_nofloat_emptybase arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @_Z26pass_agg_nofloat_emptybase21agg_nofloat_emptybase(ptr noalias sret(%struct.agg_nofloat_emptybase) align 4 %{{.*}}, i64 %{{.*}})
-// SOFT-FLOAT-LABEL:  define{{.*}} void @_Z26pass_agg_nofloat_emptybase21agg_nofloat_emptybase(ptr noalias sret(%struct.agg_nofloat_emptybase) align 4 %{{.*}}, i64 %{{.*}})
+// CHECK-LABEL: define{{.*}} void @_Z26pass_agg_nofloat_emptybase21agg_nofloat_emptybase(ptr dead_on_unwind noalias writable sret(%struct.agg_nofloat_emptybase) align 4 %{{.*}}, i64 %{{.*}})
+// SOFT-FLOAT-LABEL:  define{{.*}} void @_Z26pass_agg_nofloat_emptybase21agg_nofloat_emptybase(ptr dead_on_unwind noalias writable sret(%struct.agg_nofloat_emptybase) align 4 %{{.*}}, i64 %{{.*}})
 struct emptybase { [[no_unique_address]] empty dummy; };
 struct agg_float_emptybase : emptybase { float a; };
 struct agg_float_emptybase pass_agg_float_emptybase(struct agg_float_emptybase arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @_Z24pass_agg_float_emptybase19agg_float_emptybase(ptr noalias sret(%struct.agg_float_emptybase) align 4 %{{.*}}, float %{{.*}})
-// SOFT-FLOAT-LABEL:  define{{.*}} void @_Z24pass_agg_float_emptybase19agg_float_emptybase(ptr noalias sret(%struct.agg_float_emptybase) align 4 %{{.*}}, i32 %{{.*}})
+// CHECK-LABEL: define{{.*}} void @_Z24pass_agg_float_emptybase19agg_float_emptybase(ptr dead_on_unwind noalias writable sret(%struct.agg_float_emptybase) align 4 %{{.*}}, float %{{.*}})
+// SOFT-FLOAT-LABEL:  define{{.*}} void @_Z24pass_agg_float_emptybase19agg_float_emptybase(ptr dead_on_unwind noalias writable sret(%struct.agg_float_emptybase) align 4 %{{.*}}, i32 %{{.*}})
 struct noemptybasearray { [[no_unique_address]] empty dummy[3]; };
 struct agg_nofloat_emptybasearray : noemptybasearray { float a; };
 struct agg_nofloat_emptybasearray pass_agg_nofloat_emptybasearray(struct agg_nofloat_emptybasearray arg) { return arg; }
-// CHECK-LABEL: define{{.*}} void @_Z31pass_agg_nofloat_emptybasearray26agg_nofloat_emptybasearray(ptr noalias sret(%struct.agg_nofloat_emptybasearray) align 4 %{{.*}}, i64 %{{.*}})
-// SOFT-FLOAT-LABEL:  define{{.*}} void @_Z31pass_agg_nofloat_emptybasearray26agg_nofloat_emptybasearray(ptr noalias sret(%struct.agg_nofloat_emptybasearray) align 4 %{{.*}}, i64 %{{.*}})
+// CHECK-LABEL: define{{.*}} void @_Z31pass_agg_nofloat_emptybasearray26agg_nofloat_emptybasearray(ptr dead_on_unwind noalias writable sret(%struct.agg_nofloat_emptybasearray) align 4 %{{.*}}, i64 %{{.*}})
+// SOFT-FLOAT-LABEL:  define{{.*}} void @_Z31pass_agg_nofloat_emptybasearray26agg_nofloat_emptybasearray(ptr dead_on_unwind noalias writable sret(%struct.agg_nofloat_emptybasearray) align 4 %{{.*}}, i64 %{{.*}})
 
diff --git a/clang/test/CodeGen/SystemZ/systemz-inline-asm.c b/clang/test/CodeGen/SystemZ/systemz-inline-asm.c
index 0d614f8cbb905..e38d37cd345e2 100644
--- a/clang/test/CodeGen/SystemZ/systemz-inline-asm.c
+++ b/clang/test/CodeGen/SystemZ/systemz-inline-asm.c
@@ -123,7 +123,7 @@ double test_f64(double f, double g) {
 long double test_f128(long double f, long double g) {
   asm("axbr %0, %2" : "=f" (f) : "0" (f), "f" (g));
   return f;
-// CHECK: define{{.*}} void @test_f128(ptr noalias nocapture writeonly sret(fp128) align 8 [[DEST:%.*]], ptr nocapture noundef readonly %0, ptr nocapture noundef readonly %1)
+// CHECK: define{{.*}} void @test_f128(ptr dead_on_unwind noalias nocapture writable writeonly sret(fp128) align 8 [[DEST:%.*]], ptr nocapture noundef readonly %0, ptr nocapture noundef readonly %1)
 // CHECK: %f = load fp128, ptr %0
 // CHECK: %g = load fp128, ptr %1
 // CHECK: [[RESULT:%.*]] = tail call fp128 asm "axbr $0, $2", "=f,0,f"(fp128 %f, fp128 %g)
diff --git a/clang/test/CodeGen/WebAssembly/wasm-arguments.c b/clang/test/CodeGen/WebAssembly/wasm-arguments.c
index 1b54b6401db0e..a0914fc768809 100644
--- a/clang/test/CodeGen/WebAssembly/wasm-arguments.c
+++ b/clang/test/CodeGen/WebAssembly/wasm-arguments.c
@@ -25,9 +25,9 @@ typedef struct {
 void struct_arg(s1 i) {}
 
 // Structs should be returned sret and not simplified by the frontend.
-// WEBASSEMBLY32: define void @struct_ret(ptr noalias sret(%struct.s1) align 4 %agg.result)
+// WEBASSEMBLY32: define void @struct_ret(ptr dead_on_unwind noalias writable sret(%struct.s1) align 4 %agg.result)
 // WEBASSEMBLY32: ret void
-// WEBASSEMBLY64: define void @struct_ret(ptr noalias sret(%struct.s1) align 4 %agg.result)
+// WEBASSEMBLY64: define void @struct_ret(ptr dead_on_unwind noalias writable sret(%struct.s1) align 4 %agg.result)
 // WEBASSEMBLY64: ret void
 
 // Except with the experimental multivalue ABI, which returns structs by value
@@ -103,9 +103,9 @@ union simple_union {
 void union_arg(union simple_union s) {}
 
 // Unions should be returned sret and not simplified by the frontend.
-// WEBASSEMBLY32: define void @union_ret(ptr noalias sret(%union.simple_union) align 4 %agg.result)
+// WEBASSEMBLY32: define void @union_ret(ptr dead_on_unwind noalias writable sret(%union.simple_union) align 4 %agg.result)
 // WEBASSEMBLY32: ret void
-// WEBASSEMBLY64: define void @union_ret(ptr noalias sret(%union.simple_union) align 4 %agg.result)
+// WEBASSEMBLY64: define void @union_ret(ptr dead_on_unwind noalias writable sret(%union.simple_union) align 4 %agg.result)
 // WEBASSEMBLY64: ret void
 
 // The experimental multivalue ABI returns them by value, though.
@@ -129,8 +129,8 @@ typedef struct {
 void bitfield_arg(bitfield1 bf1) {}
 
 // And returned via sret pointers.
-// WEBASSEMBLY32: define void @bitfield_ret(ptr noalias sret(%struct.bitfield1) align 4 %agg.result)
-// WEBASSEMBLY64: define void @bitfield_ret(ptr noalias sret(%struct.bitfield1) align 4 %agg.result)
+// WEBASSEMBLY32: define void @bitfield_ret(ptr dead_on_unwind noalias writable sret(%struct.bitfield1) align 4 %agg.result)
+// WEBASSEMBLY64: define void @bitfield_ret(ptr dead_on_unwind noalias writable sret(%struct.bitfield1) align 4 %agg.result)
 
 // Except, of course, in the experimental multivalue ABI
 // EXPERIMENTAL-MV: define %struct.bitfield1 @bitfield_ret()
diff --git a/clang/test/CodeGen/WebAssembly/wasm-varargs.c b/clang/test/CodeGen/WebAssembly/wasm-varargs.c
index da22ad6a1b2e9..c475de19ae448 100644
--- a/clang/test/CodeGen/WebAssembly/wasm-varargs.c
+++ b/clang/test/CodeGen/WebAssembly/wasm-varargs.c
@@ -68,7 +68,7 @@ struct S {
 };
 
 // CHECK-LABEL: define {{[^@]+}}@test_struct
-// CHECK-SAME: (ptr noalias sret([[STRUCT_S:%.*]]) align 4 [[AGG_RESULT:%.*]], ptr noundef [[FMT:%.*]], ...) #[[ATTR0]] {
+// CHECK-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_S:%.*]]) align 4 [[AGG_RESULT:%.*]], ptr noundef [[FMT:%.*]], ...) #[[ATTR0]] {
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[FMT_ADDR:%.*]] = alloca ptr, align 4
 // CHECK-NEXT:    [[VA:%.*]] = alloca ptr, align 4
@@ -96,7 +96,7 @@ struct S test_struct(char *fmt, ...) {
 struct Z {};
 
 // CHECK-LABEL: define {{[^@]+}}@test_empty_struct
-// CHECK-SAME: (ptr noalias sret([[STRUCT_S:%.*]]) align 4 [[AGG_RESULT:%.*]], ptr noundef [[FMT:%.*]], ...) #[[ATTR0]] {
+// CHECK-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_S:%.*]]) align 4 [[AGG_RESULT:%.*]], ptr noundef [[FMT:%.*]], ...) #[[ATTR0]] {
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[FMT_ADDR:%.*]] = alloca ptr, align 4
 // CHECK-NEXT:    [[VA:%.*]] = alloca ptr, align 4
diff --git a/clang/test/CodeGen/X86/avx512-kconstraints-att_inline_asm.c b/clang/test/CodeGen/X86/avx512-kconstraints-att_inline_asm.c
index b4939bfc2ca83..74b6719bf9cfd 100644
--- a/clang/test/CodeGen/X86/avx512-kconstraints-att_inline_asm.c
+++ b/clang/test/CodeGen/X86/avx512-kconstraints-att_inline_asm.c
@@ -41,7 +41,7 @@ __m512i mask_Yk_i64(long long msk, __m512i x, __m512i y){
 }
 
 char k_wise_op_i8(char msk_src1,char msk_src2){
-//CHECK: i8 asm "kandb\09$2, $1, $0", "=k,k,k,~{dirflag},~{fpsr},~{flags}"(i8 %{{.*}}, i8 %{{.*}})
+//CHECK: <8 x i1> asm "kandb\09$2, $1, $0", "=k,k,k,~{dirflag},~{fpsr},~{flags}"(<8 x i1> %{{.*}}, <8 x i1> %{{.*}})
   char msk_dst;
   asm ("kandb\t%2, %1, %0"
        : "=k" (msk_dst)
@@ -50,7 +50,7 @@ char k_wise_op_i8(char msk_src1,char msk_src2){
 }
 
 short k_wise_op_i16(short msk_src1, short msk_src2){
-//CHECK: i16 asm "kandw\09$2, $1, $0", "=k,k,k,~{dirflag},~{fpsr},~{flags}"(i16 %{{.*}}, i16 %{{.*}})
+//CHECK: <16 x i1> asm "kandw\09$2, $1, $0", "=k,k,k,~{dirflag},~{fpsr},~{flags}"(<16 x i1> %{{.*}}, <16 x i1> %{{.*}})
   short msk_dst;
   asm ("kandw\t%2, %1, %0"
        : "=k" (msk_dst)
@@ -59,7 +59,7 @@ short k_wise_op_i16(short msk_src1, short msk_src2){
 }
 
 int k_wise_op_i32(int msk_src1, int msk_src2){
-//CHECK: i32 asm "kandd\09$2, $1, $0", "=k,k,k,~{dirflag},~{fpsr},~{flags}"(i32 %{{.*}}, i32 %{{.*}})
+//CHECK: <32 x i1> asm "kandd\09$2, $1, $0", "=k,k,k,~{dirflag},~{fpsr},~{flags}"(<32 x i1> %{{.*}}, <32 x i1> %{{.*}})
   int msk_dst;
   asm ("kandd\t%2, %1, %0"
        : "=k" (msk_dst)
@@ -68,7 +68,7 @@ int k_wise_op_i32(int msk_src1, int msk_src2){
 }
 
 long long k_wise_op_i64(long long msk_src1, long long msk_src2){
-//CHECK: i64 asm "kandq\09$2, $1, $0", "=k,k,k,~{dirflag},~{fpsr},~{flags}"(i64 %{{.*}}, i64 %{{.*}})
+//CHECK: <64 x i1> asm "kandq\09$2, $1, $0", "=k,k,k,~{dirflag},~{fpsr},~{flags}"(<64 x i1> %{{.*}}, <64 x i1> %{{.*}})
   long long msk_dst;
   asm ("kandq\t%2, %1, %0"
        : "=k" (msk_dst)
diff --git a/clang/test/CodeGen/X86/x86-inline-asm-v-constraint.c b/clang/test/CodeGen/X86/inline-asm-constraints.c
similarity index 79%
rename from clang/test/CodeGen/X86/x86-inline-asm-v-constraint.c
rename to clang/test/CodeGen/X86/inline-asm-constraints.c
index b75a84d7a7bcb..c89d94cab946b 100644
--- a/clang/test/CodeGen/X86/x86-inline-asm-v-constraint.c
+++ b/clang/test/CodeGen/X86/inline-asm-constraints.c
@@ -53,3 +53,14 @@ __m512 testZMM0(void) {
 #endif
   return zmm0;
 }
+
+extern int var, arr[4];
+struct Pair { int a, b; } pair;
+
+// CHECK-LABEL: test_Ws(
+// CHECK:         call void asm sideeffect "// ${0:p} ${1:p} ${2:p}", "^Ws,^Ws,^Ws,~{dirflag},~{fpsr},~{flags}"(ptr @var, ptr getelementptr inbounds ([4 x i32], ptr @arr, i64 0, i64 3), ptr @test_Ws)
+// CHECK:         call void asm sideeffect "// $0", "^Ws,~{dirflag},~{fpsr},~{flags}"(ptr getelementptr inbounds (%struct.Pair, ptr @pair, i32 0, i32 1))
+void test_Ws(void) {
+  asm("// %p0 %p1 %p2" :: "Ws"(&var), "Ws"(&arr[3]), "Ws"(test_Ws));
+  asm("// %0" :: "Ws"(&pair.b));
+}
diff --git a/clang/test/CodeGen/X86/x86_32-arguments-darwin.c b/clang/test/CodeGen/X86/x86_32-arguments-darwin.c
index 69d1156acf7eb..8eeba631157d0 100644
--- a/clang/test/CodeGen/X86/x86_32-arguments-darwin.c
+++ b/clang/test/CodeGen/X86/x86_32-arguments-darwin.c
@@ -71,7 +71,7 @@ struct s10 {
 // Small vectors and 1 x {i64,double} are returned in registers
 
 // CHECK: i32 @f11()
-// CHECK: void @f12(ptr noalias sret(<2 x i32>) align 8 %agg.result)
+// CHECK: void @f12(ptr dead_on_unwind noalias writable sret(<2 x i32>) align 8 %agg.result)
 // CHECK: i64 @f13()
 // CHECK: i64 @f14()
 // CHECK: <2 x i64> @f15()
@@ -93,11 +93,11 @@ T16 f16(void) { while (1) {} }
 // 128-bits).
 
 // CHECK: i32 @f17()
-// CHECK: void @f18(ptr noalias sret(%struct.anon.{{[0-9]+}}) align 8 %agg.result)
-// CHECK: void @f19(ptr noalias sret(%struct.anon.{{[0-9]+}}) align 8 %agg.result)
-// CHECK: void @f20(ptr noalias sret(%struct.anon.{{[0-9]+}}) align 8 %agg.result)
-// CHECK: void @f21(ptr noalias sret(%struct.anon.{{[0-9]+}}) align 16 %agg.result)
-// CHECK: void @f22(ptr noalias sret(%struct.anon.{{[0-9]+}}) align 16 %agg.result)
+// CHECK: void @f18(ptr dead_on_unwind noalias writable sret(%struct.anon.{{[0-9]+}}) align 8 %agg.result)
+// CHECK: void @f19(ptr dead_on_unwind noalias writable sret(%struct.anon.{{[0-9]+}}) align 8 %agg.result)
+// CHECK: void @f20(ptr dead_on_unwind noalias writable sret(%struct.anon.{{[0-9]+}}) align 8 %agg.result)
+// CHECK: void @f21(ptr dead_on_unwind noalias writable sret(%struct.anon.{{[0-9]+}}) align 16 %agg.result)
+// CHECK: void @f22(ptr dead_on_unwind noalias writable sret(%struct.anon.{{[0-9]+}}) align 16 %agg.result)
 struct { T11 a; } f17(void) { while (1) {} }
 struct { T12 a; } f18(void) { while (1) {} }
 struct { T13 a; } f19(void) { while (1) {} }
@@ -116,11 +116,11 @@ struct { struct {} a; struct { float a[1]; } b; } f25(void) { while (1) {} }
 
 // Small structures are handled recursively
 // CHECK: i32 @f26()
-// CHECK: void @f27(ptr noalias sret(%struct.s27) align 1 %agg.result)
+// CHECK: void @f27(ptr dead_on_unwind noalias writable sret(%struct.s27) align 1 %agg.result)
 struct s26 { struct { char a, b; } a; struct { char a, b; } b; } f26(void) { while (1) {} }
 struct s27 { struct { char a, b, c; } a; struct { char a; } b; } f27(void) { while (1) {} }
 
-// CHECK: void @f28(ptr noalias sret(%struct.s28) align 4 %agg.result)
+// CHECK: void @f28(ptr dead_on_unwind noalias writable sret(%struct.s28) align 4 %agg.result)
 struct s28 { int a; int b[]; } f28(void) { while (1) {} }
 
 // CHECK-LABEL: define{{.*}} i16 @f29()
@@ -150,7 +150,7 @@ struct s36 { struct { int : 0; } a[2][10]; char b; char c; } f36(void) { while (
 // CHECK-LABEL: define{{.*}} float @f37()
 struct s37 { float c[1][1]; } f37(void) { while (1) {} }
 
-// CHECK-LABEL: define{{.*}} void @f38(ptr noalias sret(%struct.s38) align 2 %agg.result)
+// CHECK-LABEL: define{{.*}} void @f38(ptr dead_on_unwind noalias writable sret(%struct.s38) align 2 %agg.result)
 struct s38 { char a[3]; short b; } f38(void) { while (1) {} }
 
 // CHECK-LABEL: define{{.*}} void @f39(ptr noundef byval(%struct.s39) align 16 %x)
diff --git a/clang/test/CodeGen/X86/x86_32-arguments-iamcu.c b/clang/test/CodeGen/X86/x86_32-arguments-iamcu.c
index c2b8f23ba5f33..2700bda1e7edf 100644
--- a/clang/test/CodeGen/X86/x86_32-arguments-iamcu.c
+++ b/clang/test/CodeGen/X86/x86_32-arguments-iamcu.c
@@ -58,7 +58,7 @@ st4_t retSmallStruct(st4_t r) { return r; }
 // CHECK-LABEL: define{{.*}} i64 @retPaddedStruct(i32 %r.coerce0, i32 %r.coerce1)
 st5_t retPaddedStruct(st5_t r) { return r; }
 
-// CHECK-LABEL: define{{.*}} void @retLargeStruct(ptr noalias sret(%struct.st12_t) align 4 %agg.result, i32 noundef %i1, ptr noundef byval(%struct.st12_t) align 4 %r)
+// CHECK-LABEL: define{{.*}} void @retLargeStruct(ptr dead_on_unwind noalias writable sret(%struct.st12_t) align 4 %agg.result, i32 noundef %i1, ptr noundef byval(%struct.st12_t) align 4 %r)
 st12_t retLargeStruct(int i1, st12_t r) { return r; }
 
 // CHECK-LABEL: define{{.*}} i32 @varArgs(i32 noundef %i1, ...)
diff --git a/clang/test/CodeGen/X86/x86_64-arguments-nacl.c b/clang/test/CodeGen/X86/x86_64-arguments-nacl.c
index 3f91ed1b1c461..4d820aef8fd2d 100644
--- a/clang/test/CodeGen/X86/x86_64-arguments-nacl.c
+++ b/clang/test/CodeGen/X86/x86_64-arguments-nacl.c
@@ -61,7 +61,7 @@ void f12_1(struct s12 a0) {}
 
 // Check that sret parameter is accounted for when checking available integer
 // registers.
-// CHECK: define{{.*}} void @f13(ptr noalias sret(%struct.s13_0) align 8 %agg.result, i32 noundef %a, i32 noundef %b, i32 noundef %c, i32 noundef %d, ptr noundef byval({{.*}}) align 8 %e, i32 noundef %f)
+// CHECK: define{{.*}} void @f13(ptr dead_on_unwind noalias writable sret(%struct.s13_0) align 8 %agg.result, i32 noundef %a, i32 noundef %b, i32 noundef %c, i32 noundef %d, ptr noundef byval({{.*}}) align 8 %e, i32 noundef %f)
 
 struct s13_0 { long long f0[3]; };
 struct s13_1 { long long f0[2]; };
diff --git a/clang/test/CodeGen/X86/x86_64-arguments-win32.c b/clang/test/CodeGen/X86/x86_64-arguments-win32.c
index 31aa0546a3c86..8768e73a854aa 100644
--- a/clang/test/CodeGen/X86/x86_64-arguments-win32.c
+++ b/clang/test/CodeGen/X86/x86_64-arguments-win32.c
@@ -27,5 +27,5 @@ void f6(_Complex double a) {}
 // CHECK-LABEL: define dso_local i64 @f7()
 _Complex float f7(void) { return 1.0; }
 
-// CHECK-LABEL: define dso_local void @f8(ptr noalias sret({ double, double }) align 8 %agg.result)
+// CHECK-LABEL: define dso_local void @f8(ptr dead_on_unwind noalias writable sret({ double, double }) align 8 %agg.result)
 _Complex double f8(void) { return 1.0; }
diff --git a/clang/test/CodeGen/X86/x86_64-arguments.c b/clang/test/CodeGen/X86/x86_64-arguments.c
index b2c4283b5e6fe..cf5636cfd518b 100644
--- a/clang/test/CodeGen/X86/x86_64-arguments.c
+++ b/clang/test/CodeGen/X86/x86_64-arguments.c
@@ -47,7 +47,7 @@ void f7(e7 a0) {
 
 // Test merging/passing of upper eightbyte with X87 class.
 //
-// CHECK-LABEL: define{{.*}} void @f8_1(ptr noalias sret(%union.u8) align 16 %agg.result)
+// CHECK-LABEL: define{{.*}} void @f8_1(ptr dead_on_unwind noalias writable sret(%union.u8) align 16 %agg.result)
 // CHECK-LABEL: define{{.*}} void @f8_2(ptr noundef byval(%union.u8) align 16 %a0)
 union u8 {
   long double a;
@@ -63,7 +63,7 @@ struct s9 { int a; int b; int : 0; } f9(void) { while (1) {} }
 struct s10 { int a; int b; int : 0; };
 void f10(struct s10 a0) {}
 
-// CHECK-LABEL: define{{.*}} void @f11(ptr noalias sret(%union.anon) align 16 %agg.result)
+// CHECK-LABEL: define{{.*}} void @f11(ptr dead_on_unwind noalias writable sret(%union.anon) align 16 %agg.result)
 union { long double a; float b; } f11(void) { while (1) {} }
 
 // CHECK-LABEL: define{{.*}} i32 @f12_0()
@@ -74,7 +74,7 @@ void f12_1(struct s12 a0) {}
 
 // Check that sret parameter is accounted for when checking available integer
 // registers.
-// CHECK: define{{.*}} void @f13(ptr noalias sret(%struct.s13_0) align 8 %agg.result, i32 noundef %a, i32 noundef %b, i32 noundef %c, i32 noundef %d, ptr noundef byval({{.*}}) align 8 %e, i32 noundef %f)
+// CHECK: define{{.*}} void @f13(ptr dead_on_unwind noalias writable sret(%struct.s13_0) align 8 %agg.result, i32 noundef %a, i32 noundef %b, i32 noundef %c, i32 noundef %d, ptr noundef byval({{.*}}) align 8 %e, i32 noundef %f)
 
 struct s13_0 { long long f0[3]; };
 struct s13_1 { long long f0[2]; };
diff --git a/clang/test/CodeGen/aarch64-branch-protection-attr.c b/clang/test/CodeGen/aarch64-branch-protection-attr.c
index 8ab3e17ade426..f0e1dcccd1e82 100644
--- a/clang/test/CodeGen/aarch64-branch-protection-attr.c
+++ b/clang/test/CodeGen/aarch64-branch-protection-attr.c
@@ -63,29 +63,33 @@ __attribute__ ((target("branch-protection=pac-ret+pc+bti")))
 void pauthlr_bti() {}
 // CHECK: define{{.*}} void @pauthlr_bti()  #[[#PAUTHLR_BTI:]]
 
+__attribute__ ((target("branch-protection=gcs")))
+void gcs() {}
+// CHECK: define{{.*}} void @gcs() #[[#GCS:]]
 
-// CHECK-DAG: attributes #[[#NONE]] = { {{.*}} "branch-target-enforcement"="false" {{.*}} "sign-return-address"="none"
+// CHECK-DAG: attributes #[[#NONE]] = { {{.*}} "branch-target-enforcement"="false" "guarded-control-stack"="false" {{.*}} "sign-return-address"="none"
 
-// CHECK-DAG: attributes #[[#STD]] = { {{.*}} "branch-target-enforcement"="true" {{.*}} "sign-return-address"="non-leaf" "sign-return-address-key"="a_key"
+// CHECK-DAG: attributes #[[#STD]] = { {{.*}} "branch-target-enforcement"="true" "guarded-control-stack"="true" {{.*}} "sign-return-address"="non-leaf" "sign-return-address-key"="a_key"
 
-// CHECK-DAG: attributes #[[#BTI]] = { {{.*}} "branch-target-enforcement"="true" {{.*}} "sign-return-address"="none"
+// CHECK-DAG: attributes #[[#BTI]] = { {{.*}} "branch-target-enforcement"="true" "guarded-control-stack"="false" {{.*}} "sign-return-address"="none"
 
-// CHECK-DAG: attributes #[[#PAC]] = { {{.*}} "branch-target-enforcement"="false" {{.*}} "sign-return-address"="non-leaf" "sign-return-address-key"="a_key"
+// CHECK-DAG: attributes #[[#PAC]] = { {{.*}} "branch-target-enforcement"="false" "guarded-control-stack"="false" {{.*}} "sign-return-address"="non-leaf" "sign-return-address-key"="a_key"
 
-// CHECK-DAG: attributes #[[#PACLEAF]] = { {{.*}} "branch-target-enforcement"="false" {{.*}}"sign-return-address"="all" "sign-return-address-key"="a_key"
+// CHECK-DAG: attributes #[[#PACLEAF]] = { {{.*}} "branch-target-enforcement"="false" "guarded-control-stack"="false" {{.*}}"sign-return-address"="all" "sign-return-address-key"="a_key"
 
-// CHECK-DAG: attributes #[[#PACBKEY]] = { {{.*}}"branch-target-enforcement"="false" {{.*}} "sign-return-address"="non-leaf" "sign-return-address-key"="b_key"
+// CHECK-DAG: attributes #[[#PACBKEY]] = { {{.*}}"branch-target-enforcement"="false" "guarded-control-stack"="false" {{.*}} "sign-return-address"="non-leaf" "sign-return-address-key"="b_key"
 
-// CHECK-DAG: attributes #[[#PACBKEYLEAF]] = { {{.*}} "branch-target-enforcement"="false" {{.*}}"sign-return-address"="all" "sign-return-address-key"="b_key"
+// CHECK-DAG: attributes #[[#PACBKEYLEAF]] = { {{.*}} "branch-target-enforcement"="false" "guarded-control-stack"="false" {{.*}}"sign-return-address"="all" "sign-return-address-key"="b_key"
 
-// CHECK-DAG: attributes #[[#BTIPACLEAF]] = { {{.*}}"branch-target-enforcement"="true" {{.*}} "sign-return-address"="all" "sign-return-address-key"="a_key"
+// CHECK-DAG: attributes #[[#BTIPACLEAF]] = { {{.*}}"branch-target-enforcement"="true" "guarded-control-stack"="false" {{.*}} "sign-return-address"="all" "sign-return-address-key"="a_key"
 
 
-// CHECK-DAG: attributes #[[#PAUTHLR]] = { {{.*}}"branch-protection-pauth-lr"="true" {{.*}}"branch-target-enforcement"="false" {{.*}}"sign-return-address"="non-leaf" "sign-return-address-key"="a_key"
+// CHECK-DAG: attributes #[[#PAUTHLR]] = { {{.*}}"branch-protection-pauth-lr"="true" {{.*}}"branch-target-enforcement"="false" "guarded-control-stack"="false" {{.*}}"sign-return-address"="non-leaf" "sign-return-address-key"="a_key"
 
-// CHECK-DAG: attributes #[[#PAUTHLR_BKEY]] = { {{.*}}"branch-protection-pauth-lr"="true" {{.*}}"branch-target-enforcement"="false" {{.*}}"sign-return-address"="non-leaf" "sign-return-address-key"="b_key"
+// CHECK-DAG: attributes #[[#PAUTHLR_BKEY]] = { {{.*}}"branch-protection-pauth-lr"="true" {{.*}}"branch-target-enforcement"="false" "guarded-control-stack"="false" {{.*}}"sign-return-address"="non-leaf" "sign-return-address-key"="b_key"
 
-// CHECK-DAG: attributes #[[#PAUTHLR_LEAF]] = { {{.*}}"branch-protection-pauth-lr"="true" {{.*}}"branch-target-enforcement"="false" {{.*}}"sign-return-address"="all" "sign-return-address-key"="a_key"
+// CHECK-DAG: attributes #[[#PAUTHLR_LEAF]] = { {{.*}}"branch-protection-pauth-lr"="true" {{.*}}"branch-target-enforcement"="false" "guarded-control-stack"="false" {{.*}}"sign-return-address"="all" "sign-return-address-key"="a_key"
 
-// CHECK-DAG: attributes #[[#PAUTHLR_BTI]] = { {{.*}}"branch-protection-pauth-lr"="true" {{.*}}"branch-target-enforcement"="true" {{.*}}"sign-return-address"="non-leaf" "sign-return-address-key"="a_key"
+// CHECK-DAG: attributes #[[#PAUTHLR_BTI]] = { {{.*}}"branch-protection-pauth-lr"="true" {{.*}}"branch-target-enforcement"="true" "guarded-control-stack"="false" {{.*}}"sign-return-address"="non-leaf" "sign-return-address-key"="a_key"
 
+// CHECK-DAG: attributes #[[#GCS]] = { {{.*}}"branch-target-enforcement"="false" "guarded-control-stack"="true" {{.*}} "sign-return-address"="none"
diff --git a/clang/test/CodeGen/aarch64-sme-intrinsics/aarch64-sme-attrs.cpp b/clang/test/CodeGen/aarch64-sme-intrinsics/aarch64-sme-attrs.cpp
index 0768bfc332387..f69703a8a7d89 100644
--- a/clang/test/CodeGen/aarch64-sme-intrinsics/aarch64-sme-attrs.cpp
+++ b/clang/test/CodeGen/aarch64-sme-intrinsics/aarch64-sme-attrs.cpp
@@ -12,8 +12,8 @@ extern int normal_callee();
 
 int streaming_decl(void) __arm_streaming;
 int streaming_compatible_decl(void) __arm_streaming_compatible;
-int shared_za_decl(void) __arm_shared_za;
-int preserves_za_decl(void) __arm_preserves_za;
+int shared_za_decl(void) __arm_inout("za");
+int preserves_za_decl(void) __arm_preserves("za");
 int private_za_decl(void);
 
 // == FUNCTION DEFINITIONS ==
@@ -78,7 +78,7 @@ __arm_locally_streaming int locally_streaming_callee() {
 // CHECK-SAME: #[[ZA_SHARED:[0-9]+]]
 // CHECK: call i32 @normal_callee()
 //
- int shared_za_caller() __arm_shared_za {
+ int shared_za_caller() __arm_inout("za") {
   return normal_callee();
 }
 
@@ -86,7 +86,7 @@ __arm_locally_streaming int locally_streaming_callee() {
 // CHECK-SAME: #[[ZA_SHARED]]
 // CHECK: call i32 @shared_za_decl() #[[ZA_SHARED_CALL:[0-9]+]]
 //
- int shared_za_callee() __arm_shared_za {
+ int shared_za_callee() __arm_inout("za") {
   return shared_za_decl();
 }
 
@@ -97,7 +97,7 @@ __arm_locally_streaming int locally_streaming_callee() {
 // CHECK-SAME: #[[ZA_PRESERVED:[0-9]+]]
 // CHECK: call i32 @normal_callee()
 //
- int preserves_za_caller() __arm_preserves_za {
+ int preserves_za_caller() __arm_preserves("za") {
   return normal_callee();
 }
 
@@ -105,7 +105,7 @@ __arm_locally_streaming int locally_streaming_callee() {
 // CHECK-SAME: #[[ZA_PRESERVED]]
 // CHECK: call i32 @preserves_za_decl() #[[ZA_PRESERVED_CALL:[0-9]+]]
 //
- int preserves_za_callee() __arm_preserves_za {
+ int preserves_za_callee() __arm_preserves("za") {
   return preserves_za_decl();
 }
 
@@ -116,7 +116,7 @@ __arm_locally_streaming int locally_streaming_callee() {
 // CHECK-SAME: #[[ZA_NEW:[0-9]+]]
 // CHECK: call i32 @normal_callee()
 //
-__arm_new_za int new_za_caller() {
+__arm_new("za") int new_za_caller() {
   return normal_callee();
 }
 
@@ -124,7 +124,7 @@ __arm_new_za int new_za_caller() {
 // CHECK-SAME: #[[ZA_NEW]]
 // CHECK: call i32 @private_za_decl()
 //
-__arm_new_za int new_za_callee() {
+__arm_new("za") int new_za_callee() {
   return private_za_decl();
 }
 
@@ -135,8 +135,8 @@ __arm_new_za int new_za_callee() {
 // and also to callsites.
 typedef void (*s_ptrty) (int, int) __arm_streaming;
 typedef void (*sc_ptrty) (int, int) __arm_streaming_compatible;
-typedef void (*sz_ptrty) (int, int) __arm_shared_za;
-typedef void (*pz_ptrty) (int, int) __arm_preserves_za;
+typedef void (*sz_ptrty) (int, int) __arm_inout("za");
+typedef void (*pz_ptrty) (int, int) __arm_preserves("za");
 
 // CHECK-LABEL: @test_streaming_ptrty(
 // CHECK-SAME: #[[NORMAL_DEF:[0-9]+]]
@@ -152,12 +152,12 @@ void test_streaming_compatible_ptrty(sc_ptrty f, int x, int y) { return f(x, y);
 // CHECK-SAME: #[[ZA_SHARED]]
 // CHECK: call void [[F:%.*]](i32 noundef [[X:%.*]], i32 noundef [[Y:%.*]]) #[[ZA_SHARED_CALL]]
 //
-void  test_shared_za(sz_ptrty f, int x, int y) __arm_shared_za { return f(x, y); }
+void  test_shared_za(sz_ptrty f, int x, int y) __arm_inout("za") { return f(x, y); }
 // CHECK-LABEL: @test_preserved_za(
 // CHECK-SAME: #[[ZA_SHARED]]
 // CHECK: call void [[F:%.*]](i32 noundef [[X:%.*]], i32 noundef [[Y:%.*]]) #[[ZA_PRESERVED_CALL]]
 //
-void  test_preserved_za(pz_ptrty f, int x, int y) __arm_shared_za { return f(x, y); }
+void  test_preserved_za(pz_ptrty f, int x, int y) __arm_inout("za") { return f(x, y); }
 
 // CHECK-LABEL: @test_indirect_streaming_ptrty(
 // CHECK-SAME: #[[NORMAL_DEF:[0-9]+]]
@@ -255,7 +255,7 @@ int call() { return 0; }
 
 template 
 __attribute__((always_inline))
-int call(T f, Other... other) __arm_shared_za {
+int call(T f, Other... other) __arm_inout("za") {
     return f() + call(other...);
 }
 
@@ -270,7 +270,7 @@ int call(T f, Other... other) __arm_shared_za {
 // CHECK-NEXT: add nsw
 // CHECK-NEXT: add nsw
 // CHECK-NEXT: ret
-int test_variadic_template() __arm_shared_za {
+int test_variadic_template() __arm_inout("za") {
   return call(normal_callee,
               streaming_decl,
               streaming_compatible_decl,
@@ -286,18 +286,18 @@ int test_variadic_template() __arm_shared_za {
 // CHECK: attributes #[[SM_BODY]] = { mustprogress noinline nounwind "aarch64_pstate_sm_body" "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-features"="+bf16,+sme" }
 // CHECK: attributes #[[ZA_SHARED]] = { mustprogress noinline nounwind "aarch64_pstate_za_shared" "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-features"="+bf16,+sme" }
 // CHECK: attributes #[[ZA_SHARED_DECL]] = { "aarch64_pstate_za_shared" "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-features"="+bf16,+sme" }
-// CHECK: attributes #[[ZA_PRESERVED]] = { mustprogress noinline nounwind "aarch64_pstate_za_preserved" "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-features"="+bf16,+sme" }
-// CHECK: attributes #[[ZA_PRESERVED_DECL]] = { "aarch64_pstate_za_preserved" "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-features"="+bf16,+sme" }
+// CHECK: attributes #[[ZA_PRESERVED]] = { mustprogress noinline nounwind "aarch64_pstate_za_preserved" "aarch64_pstate_za_shared" "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-features"="+bf16,+sme" }
+// CHECK: attributes #[[ZA_PRESERVED_DECL]] = { "aarch64_pstate_za_preserved" "aarch64_pstate_za_shared" "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-features"="+bf16,+sme" }
 // CHECK: attributes #[[ZA_NEW]] = { mustprogress noinline nounwind "aarch64_pstate_za_new" "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-features"="+bf16,+sme" }
 // CHECK: attributes #[[NORMAL_DEF]] = { mustprogress noinline nounwind "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-features"="+bf16,+sme" }
 // CHECK: attributes #[[SM_ENABLED_CALL]] = { "aarch64_pstate_sm_enabled" }
 // CHECK: attributes #[[SM_COMPATIBLE_CALL]] = { "aarch64_pstate_sm_compatible" }
 // CHECK: attributes #[[SM_BODY_CALL]] = { "aarch64_pstate_sm_body" }
 // CHECK: attributes #[[ZA_SHARED_CALL]] = { "aarch64_pstate_za_shared" }
-// CHECK: attributes #[[ZA_PRESERVED_CALL]] = { "aarch64_pstate_za_preserved" }
+// CHECK: attributes #[[ZA_PRESERVED_CALL]] = { "aarch64_pstate_za_preserved" "aarch64_pstate_za_shared" }
 // CHECK: attributes #[[NOUNWIND_CALL]] = { nounwind }
 // CHECK: attributes #[[NOUNWIND_SM_ENABLED_CALL]] = { nounwind "aarch64_pstate_sm_enabled" }
 // CHECK: attributes #[[NOUNWIND_SM_COMPATIBLE_CALL]] = { nounwind "aarch64_pstate_sm_compatible" }
 // CHECK: attributes #[[NOUNWIND_ZA_SHARED_CALL]] = { nounwind "aarch64_pstate_za_shared" }
-// CHECK: attributes #[[NOUNWIND_ZA_PRESERVED_CALL]] = { nounwind "aarch64_pstate_za_preserved" }
+// CHECK: attributes #[[NOUNWIND_ZA_PRESERVED_CALL]] = { nounwind "aarch64_pstate_za_preserved" "aarch64_pstate_za_shared" }
 
diff --git a/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_add-i32.c b/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_add-i32.c
index 55d2e355897f7..2ee14f6a7e882 100644
--- a/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_add-i32.c
+++ b/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_add-i32.c
@@ -30,7 +30,7 @@
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.addha.nxv4i32(i32 0,  [[TMP0]],  [[TMP1]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svaddha_za32_u32(svbool_t pn, svbool_t pm, svuint32_t zn) __arm_streaming __arm_shared_za {
+void test_svaddha_za32_u32(svbool_t pn, svbool_t pm, svuint32_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svaddha_za32, _u32, _m)(0, pn, pm, zn);
 }
 
@@ -50,7 +50,7 @@ void test_svaddha_za32_u32(svbool_t pn, svbool_t pm, svuint32_t zn) __arm_stream
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.addha.nxv4i32(i32 3,  [[TMP0]],  [[TMP1]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svaddha_za32_u32_1(svbool_t pn, svbool_t pm, svuint32_t zn) __arm_streaming __arm_shared_za {
+void test_svaddha_za32_u32_1(svbool_t pn, svbool_t pm, svuint32_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svaddha_za32, _u32, _m)(3, pn, pm, zn);
 }
 
@@ -70,7 +70,7 @@ void test_svaddha_za32_u32_1(svbool_t pn, svbool_t pm, svuint32_t zn) __arm_stre
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.addha.nxv4i32(i32 0,  [[TMP0]],  [[TMP1]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svaddha_za32_s32(svbool_t pn, svbool_t pm, svint32_t zn) __arm_streaming __arm_shared_za {
+void test_svaddha_za32_s32(svbool_t pn, svbool_t pm, svint32_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svaddha_za32, _s32, _m)(0, pn, pm, zn);
 }
 
@@ -90,7 +90,7 @@ void test_svaddha_za32_s32(svbool_t pn, svbool_t pm, svint32_t zn) __arm_streami
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.addha.nxv4i32(i32 3,  [[TMP0]],  [[TMP1]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svaddha_za32_s32_1(svbool_t pn, svbool_t pm, svint32_t zn) __arm_streaming __arm_shared_za {
+void test_svaddha_za32_s32_1(svbool_t pn, svbool_t pm, svint32_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svaddha_za32, _s32, _m)(3, pn, pm, zn);
 }
 
@@ -110,7 +110,7 @@ void test_svaddha_za32_s32_1(svbool_t pn, svbool_t pm, svint32_t zn) __arm_strea
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.addva.nxv4i32(i32 0,  [[TMP0]],  [[TMP1]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svaddva_za32_u32(svbool_t pn, svbool_t pm, svuint32_t zn) __arm_streaming __arm_shared_za {
+void test_svaddva_za32_u32(svbool_t pn, svbool_t pm, svuint32_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svaddva_za32, _u32, _m)(0, pn, pm, zn);
 }
 
@@ -130,7 +130,7 @@ void test_svaddva_za32_u32(svbool_t pn, svbool_t pm, svuint32_t zn) __arm_stream
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.addva.nxv4i32(i32 3,  [[TMP0]],  [[TMP1]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svaddva_za32_u32_1(svbool_t pn, svbool_t pm, svuint32_t zn) __arm_streaming __arm_shared_za {
+void test_svaddva_za32_u32_1(svbool_t pn, svbool_t pm, svuint32_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svaddva_za32, _u32, _m)(3, pn, pm, zn);
 }
 
@@ -150,7 +150,7 @@ void test_svaddva_za32_u32_1(svbool_t pn, svbool_t pm, svuint32_t zn) __arm_stre
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.addva.nxv4i32(i32 0,  [[TMP0]],  [[TMP1]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svaddva_za32_s32(svbool_t pn, svbool_t pm, svint32_t zn) __arm_streaming __arm_shared_za {
+void test_svaddva_za32_s32(svbool_t pn, svbool_t pm, svint32_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svaddva_za32, _s32, _m)(0, pn, pm, zn);
 }
 
@@ -170,7 +170,7 @@ void test_svaddva_za32_s32(svbool_t pn, svbool_t pm, svint32_t zn) __arm_streami
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.addva.nxv4i32(i32 3,  [[TMP0]],  [[TMP1]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svaddva_za32_s32_1(svbool_t pn, svbool_t pm, svint32_t zn) __arm_streaming __arm_shared_za {
+void test_svaddva_za32_s32_1(svbool_t pn, svbool_t pm, svint32_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svaddva_za32, _s32, _m)(3, pn, pm, zn);
 }
 //// NOTE: These prefixes are unused and the list is autogenerated. Do not add tests below this line:
diff --git a/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_add-i64.c b/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_add-i64.c
index 8e9c2e7da46a3..a0fb9bdb4e25a 100644
--- a/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_add-i64.c
+++ b/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_add-i64.c
@@ -30,7 +30,7 @@
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.addha.nxv2i64(i32 0,  [[TMP0]],  [[TMP1]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svaddha_za64_u64(svbool_t pn, svbool_t pm, svuint64_t zn) __arm_streaming __arm_shared_za {
+void test_svaddha_za64_u64(svbool_t pn, svbool_t pm, svuint64_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svaddha_za64, _u64, _m)(0, pn, pm, zn);
 }
 
@@ -50,7 +50,7 @@ void test_svaddha_za64_u64(svbool_t pn, svbool_t pm, svuint64_t zn) __arm_stream
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.addha.nxv2i64(i32 7,  [[TMP0]],  [[TMP1]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svaddha_za64_u64_1(svbool_t pn, svbool_t pm, svuint64_t zn) __arm_streaming __arm_shared_za {
+void test_svaddha_za64_u64_1(svbool_t pn, svbool_t pm, svuint64_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svaddha_za64, _u64, _m)(7, pn, pm, zn);
 }
 
@@ -70,7 +70,7 @@ void test_svaddha_za64_u64_1(svbool_t pn, svbool_t pm, svuint64_t zn) __arm_stre
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.addha.nxv2i64(i32 0,  [[TMP0]],  [[TMP1]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svaddha_za64_s64(svbool_t pn, svbool_t pm, svint64_t zn) __arm_streaming __arm_shared_za {
+void test_svaddha_za64_s64(svbool_t pn, svbool_t pm, svint64_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svaddha_za64, _s64, _m)(0, pn, pm, zn);
 }
 
@@ -90,7 +90,7 @@ void test_svaddha_za64_s64(svbool_t pn, svbool_t pm, svint64_t zn) __arm_streami
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.addha.nxv2i64(i32 7,  [[TMP0]],  [[TMP1]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svaddha_za64_s64_1(svbool_t pn, svbool_t pm, svint64_t zn) __arm_streaming __arm_shared_za {
+void test_svaddha_za64_s64_1(svbool_t pn, svbool_t pm, svint64_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svaddha_za64, _s64, _m)(7, pn, pm, zn);
 }
 
@@ -110,7 +110,7 @@ void test_svaddha_za64_s64_1(svbool_t pn, svbool_t pm, svint64_t zn) __arm_strea
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.addva.nxv2i64(i32 0,  [[TMP0]],  [[TMP1]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svaddva_za64_u64(svbool_t pn, svbool_t pm, svuint64_t zn) __arm_streaming __arm_shared_za {
+void test_svaddva_za64_u64(svbool_t pn, svbool_t pm, svuint64_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svaddva_za64, _u64, _m)(0, pn, pm, zn);
 }
 
@@ -130,7 +130,7 @@ void test_svaddva_za64_u64(svbool_t pn, svbool_t pm, svuint64_t zn) __arm_stream
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.addva.nxv2i64(i32 7,  [[TMP0]],  [[TMP1]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svaddva_za64_u64_1(svbool_t pn, svbool_t pm, svuint64_t zn) __arm_streaming __arm_shared_za {
+void test_svaddva_za64_u64_1(svbool_t pn, svbool_t pm, svuint64_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svaddva_za64, _u64, _m)(7, pn, pm, zn);
 }
 
@@ -150,7 +150,7 @@ void test_svaddva_za64_u64_1(svbool_t pn, svbool_t pm, svuint64_t zn) __arm_stre
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.addva.nxv2i64(i32 0,  [[TMP0]],  [[TMP1]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svaddva_za64_s64(svbool_t pn, svbool_t pm, svint64_t zn) __arm_streaming __arm_shared_za {
+void test_svaddva_za64_s64(svbool_t pn, svbool_t pm, svint64_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svaddva_za64, _s64, _m)(0, pn, pm, zn);
 }
 
@@ -170,7 +170,7 @@ void test_svaddva_za64_s64(svbool_t pn, svbool_t pm, svint64_t zn) __arm_streami
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.addva.nxv2i64(i32 7,  [[TMP0]],  [[TMP1]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svaddva_za64_s64_1(svbool_t pn, svbool_t pm, svint64_t zn) __arm_streaming __arm_shared_za {
+void test_svaddva_za64_s64_1(svbool_t pn, svbool_t pm, svint64_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svaddva_za64, _s64, _m)(7, pn, pm, zn);
 }
 //// NOTE: These prefixes are unused and the list is autogenerated. Do not add tests below this line:
diff --git a/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_ld1.c b/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_ld1.c
index e17782db222b6..183a3986212bc 100644
--- a/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_ld1.c
+++ b/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_ld1.c
@@ -22,7 +22,7 @@
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.ld1b.horiz( [[PG]], ptr [[PTR]], i32 0, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svld1_hor_za8(uint32_t slice_base, svbool_t pg, const void *ptr) __arm_streaming __arm_shared_za {
+void test_svld1_hor_za8(uint32_t slice_base, svbool_t pg, const void *ptr) __arm_streaming __arm_out("za") {
   svld1_hor_za8(0, slice_base, pg, ptr);
   svld1_hor_za8(0, slice_base + 15, pg, ptr);
 }
@@ -45,7 +45,7 @@ void test_svld1_hor_za8(uint32_t slice_base, svbool_t pg, const void *ptr) __arm
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.ld1h.horiz( [[TMP0]], ptr [[PTR]], i32 1, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svld1_hor_za16(uint32_t slice_base, svbool_t pg, const void *ptr) __arm_streaming __arm_shared_za {
+void test_svld1_hor_za16(uint32_t slice_base, svbool_t pg, const void *ptr) __arm_streaming __arm_out("za") {
   svld1_hor_za16(0, slice_base, pg, ptr);
   svld1_hor_za16(1, slice_base + 7, pg, ptr);
 }
@@ -68,7 +68,7 @@ void test_svld1_hor_za16(uint32_t slice_base, svbool_t pg, const void *ptr) __ar
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.ld1w.horiz( [[TMP0]], ptr [[PTR]], i32 3, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svld1_hor_za32(uint32_t slice_base, svbool_t pg, const void *ptr) __arm_streaming __arm_shared_za {
+void test_svld1_hor_za32(uint32_t slice_base, svbool_t pg, const void *ptr) __arm_streaming __arm_out("za") {
   svld1_hor_za32(0, slice_base, pg, ptr);
   svld1_hor_za32(3, slice_base + 3, pg, ptr);
 }
@@ -91,7 +91,7 @@ void test_svld1_hor_za32(uint32_t slice_base, svbool_t pg, const void *ptr) __ar
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.ld1d.horiz( [[TMP0]], ptr [[PTR]], i32 7, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svld1_hor_za64(uint32_t slice_base, svbool_t pg, const void *ptr) __arm_streaming __arm_shared_za {
+void test_svld1_hor_za64(uint32_t slice_base, svbool_t pg, const void *ptr) __arm_streaming __arm_out("za") {
   svld1_hor_za64(0, slice_base, pg, ptr);
   svld1_hor_za64(7, slice_base + 1, pg, ptr);
 }
@@ -112,7 +112,7 @@ void test_svld1_hor_za64(uint32_t slice_base, svbool_t pg, const void *ptr) __ar
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.ld1q.horiz( [[TMP0]], ptr [[PTR]], i32 15, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svld1_hor_za128(uint32_t slice_base, svbool_t pg, const void *ptr) __arm_streaming __arm_shared_za {
+void test_svld1_hor_za128(uint32_t slice_base, svbool_t pg, const void *ptr) __arm_streaming __arm_out("za") {
   svld1_hor_za128(0, slice_base, pg, ptr);
   svld1_hor_za128(15, slice_base, pg, ptr);
 }
@@ -133,7 +133,7 @@ void test_svld1_hor_za128(uint32_t slice_base, svbool_t pg, const void *ptr) __a
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.ld1b.vert( [[PG]], ptr [[PTR]], i32 0, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svld1_ver_za8(uint32_t slice_base, svbool_t pg, const void *ptr) __arm_streaming __arm_shared_za {
+void test_svld1_ver_za8(uint32_t slice_base, svbool_t pg, const void *ptr) __arm_streaming __arm_out("za") {
   svld1_ver_za8(0, slice_base, pg, ptr);
   svld1_ver_za8(0, slice_base + 15, pg, ptr);
 }
@@ -156,7 +156,7 @@ void test_svld1_ver_za8(uint32_t slice_base, svbool_t pg, const void *ptr) __arm
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.ld1h.vert( [[TMP0]], ptr [[PTR]], i32 1, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svld1_ver_za16(uint32_t slice_base, svbool_t pg, const void *ptr) __arm_streaming __arm_shared_za {
+void test_svld1_ver_za16(uint32_t slice_base, svbool_t pg, const void *ptr) __arm_streaming __arm_out("za") {
   svld1_ver_za16(0, slice_base, pg, ptr);
   svld1_ver_za16(1, slice_base + 7, pg, ptr);
 }
@@ -179,7 +179,7 @@ void test_svld1_ver_za16(uint32_t slice_base, svbool_t pg, const void *ptr) __ar
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.ld1w.vert( [[TMP0]], ptr [[PTR]], i32 3, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svld1_ver_za32(uint32_t slice_base, svbool_t pg, const void *ptr) __arm_streaming __arm_shared_za {
+void test_svld1_ver_za32(uint32_t slice_base, svbool_t pg, const void *ptr) __arm_streaming __arm_out("za") {
   svld1_ver_za32(0, slice_base, pg, ptr);
   svld1_ver_za32(3, slice_base + 3, pg, ptr);
 }
@@ -202,7 +202,7 @@ void test_svld1_ver_za32(uint32_t slice_base, svbool_t pg, const void *ptr) __ar
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.ld1d.vert( [[TMP0]], ptr [[PTR]], i32 7, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svld1_ver_za64(uint32_t slice_base, svbool_t pg, const void *ptr) __arm_streaming __arm_shared_za {
+void test_svld1_ver_za64(uint32_t slice_base, svbool_t pg, const void *ptr) __arm_streaming __arm_out("za") {
   svld1_ver_za64(0, slice_base, pg, ptr);
   svld1_ver_za64(7, slice_base + 1, pg, ptr);
 }
@@ -223,7 +223,7 @@ void test_svld1_ver_za64(uint32_t slice_base, svbool_t pg, const void *ptr) __ar
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.ld1q.vert( [[TMP0]], ptr [[PTR]], i32 15, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svld1_ver_za128(uint32_t slice_base, svbool_t pg, const void *ptr) __arm_streaming __arm_shared_za {
+void test_svld1_ver_za128(uint32_t slice_base, svbool_t pg, const void *ptr) __arm_streaming __arm_out("za") {
   svld1_ver_za128(0, slice_base, pg, ptr);
   svld1_ver_za128(15, slice_base, pg, ptr);
 }
diff --git a/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_ld1_vnum.c b/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_ld1_vnum.c
index 0fa77e1144a7d..68a294b1a237a 100644
--- a/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_ld1_vnum.c
+++ b/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_ld1_vnum.c
@@ -28,7 +28,7 @@
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.ld1b.horiz( [[PG]], ptr [[TMP1]], i32 0, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svld1_hor_vnum_za8(uint32_t slice_base, svbool_t pg, const void *ptr, int64_t vnum) __arm_streaming __arm_shared_za {
+void test_svld1_hor_vnum_za8(uint32_t slice_base, svbool_t pg, const void *ptr, int64_t vnum) __arm_streaming __arm_out("za") {
   svld1_hor_vnum_za8(0, slice_base, pg, ptr, vnum);
   svld1_hor_vnum_za8(0, slice_base + 15, pg, ptr, vnum);
 }
@@ -57,7 +57,7 @@ void test_svld1_hor_vnum_za8(uint32_t slice_base, svbool_t pg, const void *ptr,
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.ld1h.horiz( [[TMP0]], ptr [[TMP2]], i32 1, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svld1_hor_vnum_za16(uint32_t slice_base, svbool_t pg, const void *ptr, int64_t vnum) __arm_streaming __arm_shared_za {
+void test_svld1_hor_vnum_za16(uint32_t slice_base, svbool_t pg, const void *ptr, int64_t vnum) __arm_streaming __arm_out("za") {
   svld1_hor_vnum_za16(0, slice_base, pg, ptr, vnum);
   svld1_hor_vnum_za16(1, slice_base + 7, pg, ptr, vnum);
 }
@@ -86,7 +86,7 @@ void test_svld1_hor_vnum_za16(uint32_t slice_base, svbool_t pg, const void *ptr,
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.ld1w.horiz( [[TMP0]], ptr [[TMP2]], i32 3, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svld1_hor_vnum_za32(uint32_t slice_base, svbool_t pg, const void *ptr, int64_t vnum) __arm_streaming __arm_shared_za {
+void test_svld1_hor_vnum_za32(uint32_t slice_base, svbool_t pg, const void *ptr, int64_t vnum) __arm_streaming __arm_out("za") {
   svld1_hor_vnum_za32(0, slice_base, pg, ptr, vnum);
   svld1_hor_vnum_za32(3, slice_base + 3, pg, ptr, vnum);
 }
@@ -115,7 +115,7 @@ void test_svld1_hor_vnum_za32(uint32_t slice_base, svbool_t pg, const void *ptr,
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.ld1d.horiz( [[TMP0]], ptr [[TMP2]], i32 7, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svld1_hor_vnum_za64(uint32_t slice_base, svbool_t pg, const void *ptr, int64_t vnum) __arm_streaming __arm_shared_za {
+void test_svld1_hor_vnum_za64(uint32_t slice_base, svbool_t pg, const void *ptr, int64_t vnum) __arm_streaming __arm_out("za") {
   svld1_hor_vnum_za64(0, slice_base, pg, ptr, vnum);
   svld1_hor_vnum_za64(7, slice_base + 1, pg, ptr, vnum);
 }
@@ -142,7 +142,7 @@ void test_svld1_hor_vnum_za64(uint32_t slice_base, svbool_t pg, const void *ptr,
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.ld1q.horiz( [[TMP0]], ptr [[TMP2]], i32 15, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svld1_hor_vnum_za128(uint32_t slice_base, svbool_t pg, const void *ptr, int64_t vnum) __arm_streaming __arm_shared_za {
+void test_svld1_hor_vnum_za128(uint32_t slice_base, svbool_t pg, const void *ptr, int64_t vnum) __arm_streaming __arm_out("za") {
   svld1_hor_vnum_za128(0, slice_base, pg, ptr, vnum);
   svld1_hor_vnum_za128(15, slice_base, pg, ptr, vnum);
 }
@@ -169,7 +169,7 @@ void test_svld1_hor_vnum_za128(uint32_t slice_base, svbool_t pg, const void *ptr
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.ld1b.vert( [[PG]], ptr [[TMP1]], i32 0, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svld1_ver_hor_za8(uint32_t slice_base, svbool_t pg, const void *ptr, int64_t vnum) __arm_streaming __arm_shared_za {
+void test_svld1_ver_hor_za8(uint32_t slice_base, svbool_t pg, const void *ptr, int64_t vnum) __arm_streaming __arm_out("za") {
   svld1_ver_vnum_za8(0, slice_base, pg, ptr, vnum);
   svld1_ver_vnum_za8(0, slice_base + 15, pg, ptr, vnum);
 }
@@ -198,7 +198,7 @@ void test_svld1_ver_hor_za8(uint32_t slice_base, svbool_t pg, const void *ptr, i
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.ld1h.vert( [[TMP0]], ptr [[TMP2]], i32 1, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svld1_ver_vnum_za16(uint32_t slice_base, svbool_t pg, const void *ptr, int64_t vnum) __arm_streaming __arm_shared_za {
+void test_svld1_ver_vnum_za16(uint32_t slice_base, svbool_t pg, const void *ptr, int64_t vnum) __arm_streaming __arm_out("za") {
   svld1_ver_vnum_za16(0, slice_base, pg, ptr, vnum);
   svld1_ver_vnum_za16(1, slice_base + 7, pg, ptr, vnum);
 }
@@ -227,7 +227,7 @@ void test_svld1_ver_vnum_za16(uint32_t slice_base, svbool_t pg, const void *ptr,
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.ld1w.vert( [[TMP0]], ptr [[TMP2]], i32 3, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svld1_ver_vnum_za32(uint32_t slice_base, svbool_t pg, const void *ptr, int64_t vnum) __arm_streaming __arm_shared_za {
+void test_svld1_ver_vnum_za32(uint32_t slice_base, svbool_t pg, const void *ptr, int64_t vnum) __arm_streaming __arm_out("za") {
   svld1_ver_vnum_za32(0, slice_base, pg, ptr, vnum);
   svld1_ver_vnum_za32(3, slice_base + 3, pg, ptr, vnum);
 }
@@ -256,7 +256,7 @@ void test_svld1_ver_vnum_za32(uint32_t slice_base, svbool_t pg, const void *ptr,
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.ld1d.vert( [[TMP0]], ptr [[TMP2]], i32 7, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svld1_ver_vnum_za64(uint32_t slice_base, svbool_t pg, const void *ptr, int64_t vnum) __arm_streaming __arm_shared_za {
+void test_svld1_ver_vnum_za64(uint32_t slice_base, svbool_t pg, const void *ptr, int64_t vnum) __arm_streaming __arm_out("za") {
   svld1_ver_vnum_za64(0, slice_base, pg, ptr, vnum);
   svld1_ver_vnum_za64(7, slice_base + 1, pg, ptr, vnum);
 }
@@ -283,7 +283,7 @@ void test_svld1_ver_vnum_za64(uint32_t slice_base, svbool_t pg, const void *ptr,
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.ld1q.vert( [[TMP0]], ptr [[TMP2]], i32 15, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svld1_ver_vnum_za128(uint32_t slice_base, svbool_t pg, const void *ptr, int64_t vnum) __arm_streaming __arm_shared_za {
+void test_svld1_ver_vnum_za128(uint32_t slice_base, svbool_t pg, const void *ptr, int64_t vnum) __arm_streaming __arm_out("za") {
   svld1_ver_vnum_za128(0, slice_base, pg, ptr, vnum);
   svld1_ver_vnum_za128(15, slice_base, pg, ptr, vnum);
 }
diff --git a/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_ldr.c b/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_ldr.c
index 314c9645dd4f7..56eb7f45f6d9a 100644
--- a/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_ldr.c
+++ b/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_ldr.c
@@ -12,7 +12,7 @@
 // CHECK-NEXT:    tail call void @llvm.aarch64.sme.ldr(i32 [[SLICE_BASE:%.*]], ptr [[PTR:%.*]], i32 0)
 // CHECK-NEXT:    ret void
 //
-void test_svldr_vnum_za(uint32_t slice_base, const void *ptr) __arm_shared_za {
+void test_svldr_vnum_za(uint32_t slice_base, const void *ptr) __arm_out("za") {
   svldr_vnum_za(slice_base, ptr, 0);
 }
 
@@ -22,7 +22,7 @@ void test_svldr_vnum_za(uint32_t slice_base, const void *ptr) __arm_shared_za {
 // CHECK-NEXT:    tail call void @llvm.aarch64.sme.ldr(i32 [[SLICE_BASE:%.*]], ptr [[PTR:%.*]], i32 15)
 // CHECK-NEXT:    ret void
 //
-void test_svldr_vnum_za_1(uint32_t slice_base, const void *ptr) __arm_shared_za {
+void test_svldr_vnum_za_1(uint32_t slice_base, const void *ptr) __arm_out("za") {
   svldr_vnum_za(slice_base, ptr, 15);
 }
 
@@ -32,7 +32,7 @@ void test_svldr_vnum_za_1(uint32_t slice_base, const void *ptr) __arm_shared_za
 // CHECK-NEXT:    tail call void @llvm.aarch64.sme.ldr(i32 [[SLICE_BASE:%.*]], ptr [[PTR:%.*]], i32 0)
 // CHECK-NEXT:    ret void
 //
-void test_svldr_za(uint32_t slice_base, const void *ptr) __arm_shared_za {
+void test_svldr_za(uint32_t slice_base, const void *ptr) __arm_out("za") {
   svldr_za(slice_base, ptr);
 }
 
@@ -43,7 +43,7 @@ void test_svldr_za(uint32_t slice_base, const void *ptr) __arm_shared_za {
 // CHECK-NEXT:    tail call void @llvm.aarch64.sme.ldr(i32 [[SLICE_BASE:%.*]], ptr [[PTR:%.*]], i32 [[TMP0:%.*]])
 // CHECK-NEXT:    ret void
 //
-void test_svldr_vnum_za_var(uint32_t slice_base, const void *ptr, int64_t vnum) __arm_shared_za {
+void test_svldr_vnum_za_var(uint32_t slice_base, const void *ptr, int64_t vnum) __arm_out("za") {
   svldr_vnum_za(slice_base, ptr, vnum);
 }
 
@@ -53,6 +53,6 @@ void test_svldr_vnum_za_var(uint32_t slice_base, const void *ptr, int64_t vnum)
 // CHECK-NEXT:    tail call void @llvm.aarch64.sme.ldr(i32 [[SLICE_BASE:%.*]], ptr [[PTR:%.*]], i32 16)
 // CHECK-NEXT:    ret void
 //
-void test_svldr_vnum_za_2(uint32_t slice_base, const void *ptr) __arm_shared_za {
+void test_svldr_vnum_za_2(uint32_t slice_base, const void *ptr) __arm_out("za") {
   svldr_vnum_za(slice_base, ptr, 16);
 }
diff --git a/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_mopa-za32.c b/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_mopa-za32.c
index e84f31c2dfa92..5015f9bc9f41e 100644
--- a/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_mopa-za32.c
+++ b/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_mopa-za32.c
@@ -26,7 +26,7 @@
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.smopa.wide.nxv16i8(i32 0,  [[PN]],  [[PM]],  [[ZN]],  [[ZM]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svmopa_za32_s8(svbool_t pn, svbool_t pm, svint8_t zn, svint8_t zm) __arm_streaming __arm_shared_za {
+void test_svmopa_za32_s8(svbool_t pn, svbool_t pm, svint8_t zn, svint8_t zm) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svmopa_za32, _s8, _m)(0, pn, pm, zn, zm);
 }
 
@@ -42,7 +42,7 @@ void test_svmopa_za32_s8(svbool_t pn, svbool_t pm, svint8_t zn, svint8_t zm) __a
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.umopa.wide.nxv16i8(i32 0,  [[PN]],  [[PM]],  [[ZN]],  [[ZM]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svmopa_za32_u8(svbool_t pn, svbool_t pm, svuint8_t zn, svuint8_t zm) __arm_streaming __arm_shared_za {
+void test_svmopa_za32_u8(svbool_t pn, svbool_t pm, svuint8_t zn, svuint8_t zm) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svmopa_za32, _u8, _m)(0, pn, pm, zn, zm);
 }
 
@@ -62,7 +62,7 @@ void test_svmopa_za32_u8(svbool_t pn, svbool_t pm, svuint8_t zn, svuint8_t zm) _
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.mopa.wide.nxv8bf16(i32 0,  [[TMP0]],  [[TMP1]],  [[ZN]],  [[ZM]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svmopa_za32_bf16(svbool_t pn, svbool_t pm, svbfloat16_t zn, svbfloat16_t zm) __arm_streaming __arm_shared_za {
+void test_svmopa_za32_bf16(svbool_t pn, svbool_t pm, svbfloat16_t zn, svbfloat16_t zm) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svmopa_za32, _bf16, _m)(0, pn, pm, zn, zm);
 }
 
@@ -82,7 +82,7 @@ void test_svmopa_za32_bf16(svbool_t pn, svbool_t pm, svbfloat16_t zn, svbfloat16
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.mopa.wide.nxv8f16(i32 1,  [[TMP0]],  [[TMP1]],  [[ZN]],  [[ZM]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svmopa_za32_f16(svbool_t pn, svbool_t pm, svfloat16_t zn, svfloat16_t zm) __arm_streaming __arm_shared_za {
+void test_svmopa_za32_f16(svbool_t pn, svbool_t pm, svfloat16_t zn, svfloat16_t zm) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svmopa_za32, _f16, _m)(1, pn, pm, zn, zm);
 }
 
@@ -102,7 +102,7 @@ void test_svmopa_za32_f16(svbool_t pn, svbool_t pm, svfloat16_t zn, svfloat16_t
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.mopa.nxv4f32(i32 1,  [[TMP0]],  [[TMP1]],  [[ZN]],  [[ZM]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svmopa_za32_f32(svbool_t pn, svbool_t pm, svfloat32_t zn, svfloat32_t zm) __arm_streaming __arm_shared_za {
+void test_svmopa_za32_f32(svbool_t pn, svbool_t pm, svfloat32_t zn, svfloat32_t zm) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svmopa_za32, _f32, _m)(1, pn, pm, zn, zm);
 }
 
@@ -118,7 +118,7 @@ void test_svmopa_za32_f32(svbool_t pn, svbool_t pm, svfloat32_t zn, svfloat32_t
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.sumopa.wide.nxv16i8(i32 0,  [[PN]],  [[PM]],  [[ZN]],  [[ZM]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svsumopa_za32_s8(svbool_t pn, svbool_t pm, svint8_t zn, svuint8_t zm) __arm_streaming __arm_shared_za {
+void test_svsumopa_za32_s8(svbool_t pn, svbool_t pm, svint8_t zn, svuint8_t zm) __arm_streaming __arm_inout("za") {
  SME_ACLE_FUNC(svsumopa_za32, _s8, _m)(0, pn, pm, zn, zm);
 }
 
@@ -134,7 +134,7 @@ void test_svsumopa_za32_s8(svbool_t pn, svbool_t pm, svint8_t zn, svuint8_t zm)
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.usmopa.wide.nxv16i8(i32 0,  [[PN]],  [[PM]],  [[ZN]],  [[ZM]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svusmopa_za32_u8(svbool_t pn, svbool_t pm, svuint8_t zn, svint8_t zm) __arm_streaming __arm_shared_za {
+void test_svusmopa_za32_u8(svbool_t pn, svbool_t pm, svuint8_t zn, svint8_t zm) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svusmopa_za32, _u8, _m)(0, pn, pm, zn, zm);
 }
 //// NOTE: These prefixes are unused and the list is autogenerated. Do not add tests below this line:
diff --git a/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_mopa-za64.c b/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_mopa-za64.c
index 1b22eb64e9e36..80457359e2a4e 100644
--- a/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_mopa-za64.c
+++ b/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_mopa-za64.c
@@ -30,7 +30,7 @@
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.smopa.wide.nxv8i16(i32 7,  [[TMP0]],  [[TMP1]],  [[ZN]],  [[ZM]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svmopa_za64_s16(svbool_t pn, svbool_t pm, svint16_t zn, svint16_t zm) __arm_streaming __arm_shared_za {
+void test_svmopa_za64_s16(svbool_t pn, svbool_t pm, svint16_t zn, svint16_t zm) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svmopa_za64, _s16, _m)(7, pn, pm, zn, zm);
 }
 
@@ -50,7 +50,7 @@ void test_svmopa_za64_s16(svbool_t pn, svbool_t pm, svint16_t zn, svint16_t zm)
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.umopa.wide.nxv8i16(i32 0,  [[TMP0]],  [[TMP1]],  [[ZN]],  [[ZM]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svmopa_za64_u16(svbool_t pn, svbool_t pm, svuint16_t zn, svuint16_t zm) __arm_streaming __arm_shared_za {
+void test_svmopa_za64_u16(svbool_t pn, svbool_t pm, svuint16_t zn, svuint16_t zm) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svmopa_za64, _u16, _m)(0, pn, pm, zn, zm);
 }
 
@@ -70,7 +70,7 @@ void test_svmopa_za64_u16(svbool_t pn, svbool_t pm, svuint16_t zn, svuint16_t zm
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.mopa.nxv2f64(i32 7,  [[TMP0]],  [[TMP1]],  [[ZN]],  [[ZM]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svmopa_za64_f64(svbool_t pn, svbool_t pm, svfloat64_t zn, svfloat64_t zm) __arm_streaming __arm_shared_za {
+void test_svmopa_za64_f64(svbool_t pn, svbool_t pm, svfloat64_t zn, svfloat64_t zm) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svmopa_za64, _f64, _m)(7, pn, pm, zn, zm);
 }
 
@@ -90,7 +90,7 @@ void test_svmopa_za64_f64(svbool_t pn, svbool_t pm, svfloat64_t zn, svfloat64_t
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.sumopa.wide.nxv8i16(i32 0,  [[TMP0]],  [[TMP1]],  [[ZN]],  [[ZM]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svsumopa_za64_s16(svbool_t pn, svbool_t pm, svint16_t zn, svuint16_t zm) __arm_streaming __arm_shared_za {
+void test_svsumopa_za64_s16(svbool_t pn, svbool_t pm, svint16_t zn, svuint16_t zm) __arm_streaming __arm_inout("za") {
  SME_ACLE_FUNC(svsumopa_za64, _s16, _m)(0, pn, pm, zn, zm);
 }
 
@@ -110,7 +110,7 @@ void test_svsumopa_za64_s16(svbool_t pn, svbool_t pm, svint16_t zn, svuint16_t z
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.usmopa.wide.nxv8i16(i32 7,  [[TMP0]],  [[TMP1]],  [[ZN]],  [[ZM]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svusmopa_za64_u16(svbool_t pn, svbool_t pm, svuint16_t zn, svint16_t zm) __arm_streaming __arm_shared_za {
+void test_svusmopa_za64_u16(svbool_t pn, svbool_t pm, svuint16_t zn, svint16_t zm) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svusmopa_za64, _u16, _m)(7, pn, pm, zn, zm);
 }
 //// NOTE: These prefixes are unused and the list is autogenerated. Do not add tests below this line:
diff --git a/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_mops-za32.c b/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_mops-za32.c
index 0ff97ff92f714..7dad7289d98ad 100644
--- a/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_mops-za32.c
+++ b/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_mops-za32.c
@@ -26,7 +26,7 @@
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.smops.wide.nxv16i8(i32 0,  [[PN]],  [[PM]],  [[ZN]],  [[ZM]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svmops_za32_s8(svbool_t pn, svbool_t pm, svint8_t zn, svint8_t zm) __arm_streaming __arm_shared_za {
+void test_svmops_za32_s8(svbool_t pn, svbool_t pm, svint8_t zn, svint8_t zm) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svmops_za32, _s8, _m)(0, pn, pm, zn, zm);
 }
 
@@ -42,7 +42,7 @@ void test_svmops_za32_s8(svbool_t pn, svbool_t pm, svint8_t zn, svint8_t zm) __a
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.umops.wide.nxv16i8(i32 0,  [[PN]],  [[PM]],  [[ZN]],  [[ZM]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svmops_za32_u8(svbool_t pn, svbool_t pm, svuint8_t zn, svuint8_t zm) __arm_streaming __arm_shared_za {
+void test_svmops_za32_u8(svbool_t pn, svbool_t pm, svuint8_t zn, svuint8_t zm) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svmops_za32, _u8, _m)(0, pn, pm, zn, zm);
 }
 
@@ -62,7 +62,7 @@ void test_svmops_za32_u8(svbool_t pn, svbool_t pm, svuint8_t zn, svuint8_t zm) _
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.mops.wide.nxv8bf16(i32 0,  [[TMP0]],  [[TMP1]],  [[ZN]],  [[ZM]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svmops_za32_bf16(svbool_t pn, svbool_t pm, svbfloat16_t zn, svbfloat16_t zm) __arm_streaming __arm_shared_za {
+void test_svmops_za32_bf16(svbool_t pn, svbool_t pm, svbfloat16_t zn, svbfloat16_t zm) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svmops_za32, _bf16, _m)(0, pn, pm, zn, zm);
 }
 
@@ -82,7 +82,7 @@ void test_svmops_za32_bf16(svbool_t pn, svbool_t pm, svbfloat16_t zn, svbfloat16
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.mops.wide.nxv8f16(i32 1,  [[TMP0]],  [[TMP1]],  [[ZN]],  [[ZM]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svmops_za32_f16(svbool_t pn, svbool_t pm, svfloat16_t zn, svfloat16_t zm) __arm_streaming __arm_shared_za {
+void test_svmops_za32_f16(svbool_t pn, svbool_t pm, svfloat16_t zn, svfloat16_t zm) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svmops_za32, _f16, _m)(1, pn, pm, zn, zm);
 }
 
@@ -102,7 +102,7 @@ void test_svmops_za32_f16(svbool_t pn, svbool_t pm, svfloat16_t zn, svfloat16_t
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.mops.nxv4f32(i32 1,  [[TMP0]],  [[TMP1]],  [[ZN]],  [[ZM]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svmops_za32_f32(svbool_t pn, svbool_t pm, svfloat32_t zn, svfloat32_t zm) __arm_streaming __arm_shared_za {
+void test_svmops_za32_f32(svbool_t pn, svbool_t pm, svfloat32_t zn, svfloat32_t zm) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svmops_za32, _f32, _m)(1, pn, pm, zn, zm);
 }
 
@@ -118,7 +118,7 @@ void test_svmops_za32_f32(svbool_t pn, svbool_t pm, svfloat32_t zn, svfloat32_t
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.sumops.wide.nxv16i8(i32 0,  [[PN]],  [[PM]],  [[ZN]],  [[ZM]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svsumops_za32_s8(svbool_t pn, svbool_t pm, svint8_t zn, svuint8_t zm) __arm_streaming __arm_shared_za {
+void test_svsumops_za32_s8(svbool_t pn, svbool_t pm, svint8_t zn, svuint8_t zm) __arm_streaming __arm_inout("za") {
  SME_ACLE_FUNC(svsumops_za32, _s8, _m)(0, pn, pm, zn, zm);
 }
 
@@ -134,7 +134,7 @@ void test_svsumops_za32_s8(svbool_t pn, svbool_t pm, svint8_t zn, svuint8_t zm)
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.usmops.wide.nxv16i8(i32 0,  [[PN]],  [[PM]],  [[ZN]],  [[ZM]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svusmops_za32_u8(svbool_t pn, svbool_t pm, svuint8_t zn, svint8_t zm) __arm_streaming __arm_shared_za {
+void test_svusmops_za32_u8(svbool_t pn, svbool_t pm, svuint8_t zn, svint8_t zm) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svusmops_za32, _u8, _m)(0, pn, pm, zn, zm);
 }
 //// NOTE: These prefixes are unused and the list is autogenerated. Do not add tests below this line:
diff --git a/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_mops-za64.c b/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_mops-za64.c
index 5b190a7f9b748..e95a5ea2e6941 100644
--- a/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_mops-za64.c
+++ b/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_mops-za64.c
@@ -30,7 +30,7 @@
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.smops.wide.nxv8i16(i32 7,  [[TMP0]],  [[TMP1]],  [[ZN]],  [[ZM]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svmops_za64_s16(svbool_t pn, svbool_t pm, svint16_t zn, svint16_t zm) __arm_streaming __arm_shared_za {
+void test_svmops_za64_s16(svbool_t pn, svbool_t pm, svint16_t zn, svint16_t zm) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svmops_za64, _s16, _m)(7, pn, pm, zn, zm);
 }
 
@@ -50,7 +50,7 @@ void test_svmops_za64_s16(svbool_t pn, svbool_t pm, svint16_t zn, svint16_t zm)
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.umops.wide.nxv8i16(i32 0,  [[TMP0]],  [[TMP1]],  [[ZN]],  [[ZM]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svmops_za64_u16(svbool_t pn, svbool_t pm, svuint16_t zn, svuint16_t zm) __arm_streaming __arm_shared_za {
+void test_svmops_za64_u16(svbool_t pn, svbool_t pm, svuint16_t zn, svuint16_t zm) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svmops_za64, _u16, _m)(0, pn, pm, zn, zm);
 }
 
@@ -70,7 +70,7 @@ void test_svmops_za64_u16(svbool_t pn, svbool_t pm, svuint16_t zn, svuint16_t zm
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.mops.nxv2f64(i32 7,  [[TMP0]],  [[TMP1]],  [[ZN]],  [[ZM]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svmops_za64_f64(svbool_t pn, svbool_t pm, svfloat64_t zn, svfloat64_t zm) __arm_streaming __arm_shared_za {
+void test_svmops_za64_f64(svbool_t pn, svbool_t pm, svfloat64_t zn, svfloat64_t zm) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svmops_za64, _f64, _m)(7, pn, pm, zn, zm);
 }
 
@@ -90,7 +90,7 @@ void test_svmops_za64_f64(svbool_t pn, svbool_t pm, svfloat64_t zn, svfloat64_t
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.sumops.wide.nxv8i16(i32 0,  [[TMP0]],  [[TMP1]],  [[ZN]],  [[ZM]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svsumops_za64_s16(svbool_t pn, svbool_t pm, svint16_t zn, svuint16_t zm) __arm_streaming __arm_shared_za {
+void test_svsumops_za64_s16(svbool_t pn, svbool_t pm, svint16_t zn, svuint16_t zm) __arm_streaming __arm_inout("za") {
  SME_ACLE_FUNC(svsumops_za64, _s16, _m)(0, pn, pm, zn, zm);
 }
 
@@ -110,7 +110,7 @@ void test_svsumops_za64_s16(svbool_t pn, svbool_t pm, svint16_t zn, svuint16_t z
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.usmops.wide.nxv8i16(i32 7,  [[TMP0]],  [[TMP1]],  [[ZN]],  [[ZM]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svusmops_za64_u16(svbool_t pn, svbool_t pm, svuint16_t zn, svint16_t zm) __arm_streaming __arm_shared_za {
+void test_svusmops_za64_u16(svbool_t pn, svbool_t pm, svuint16_t zn, svint16_t zm) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svusmops_za64, _u16, _m)(7, pn, pm, zn, zm);
 }
 //// NOTE: These prefixes are unused and the list is autogenerated. Do not add tests below this line:
diff --git a/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_read.c b/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_read.c
index 843a96da90278..700564c1532e0 100644
--- a/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_read.c
+++ b/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_read.c
@@ -26,7 +26,7 @@
 // CHECK-CXX-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sme.read.horiz.nxv16i8( [[ZD]],  [[PG]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP0]]
 //
-svint8_t test_svread_hor_za8_s8(svint8_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svint8_t test_svread_hor_za8_s8(svint8_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_hor_za8, _s8, _m)(zd, pg, 0, slice_base);
 }
 
@@ -44,7 +44,7 @@ svint8_t test_svread_hor_za8_s8(svint8_t zd, svbool_t pg, uint32_t slice_base) _
 // CHECK-CXX-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sme.read.horiz.nxv16i8( [[ZD]],  [[PG]], i32 0, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret  [[TMP0]]
 //
-svint8_t test_svread_hor_za8_s8_1(svint8_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svint8_t test_svread_hor_za8_s8_1(svint8_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     uint32_t slice = slice_base + 15;
     return SME_ACLE_FUNC(svread_hor_za8, _s8, _m)(zd, pg, 0, slice);
 }
@@ -63,7 +63,7 @@ svint8_t test_svread_hor_za8_s8_1(svint8_t zd, svbool_t pg, uint32_t slice_base)
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.read.horiz.nxv8i16( [[ZD]],  [[TMP0]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svint16_t test_svread_hor_za16_s16(svint16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svint16_t test_svread_hor_za16_s16(svint16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
      return SME_ACLE_FUNC(svread_hor_za16, _s16, _m)(zd, pg, 0, slice_base);
 }
 
@@ -83,7 +83,7 @@ svint16_t test_svread_hor_za16_s16(svint16_t zd, svbool_t pg, uint32_t slice_bas
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.read.horiz.nxv8i16( [[ZD]],  [[TMP0]], i32 1, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svint16_t test_svread_hor_za16_s16_1(svint16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svint16_t test_svread_hor_za16_s16_1(svint16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
      uint32_t slice = slice_base + 7;
      return SME_ACLE_FUNC(svread_hor_za16, _s16, _m)(zd, pg, 1, slice);
 }
@@ -102,7 +102,7 @@ svint16_t test_svread_hor_za16_s16_1(svint16_t zd, svbool_t pg, uint32_t slice_b
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.read.horiz.nxv4i32( [[ZD]],  [[TMP0]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svint32_t test_svread_hor_za32_s32(svint32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svint32_t test_svread_hor_za32_s32(svint32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_hor_za32, _s32, _m)(zd, pg, 0, slice_base);
 }
 
@@ -122,7 +122,7 @@ svint32_t test_svread_hor_za32_s32(svint32_t zd, svbool_t pg, uint32_t slice_bas
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.read.horiz.nxv4i32( [[ZD]],  [[TMP0]], i32 3, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svint32_t test_svread_hor_za32_s32_1(svint32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svint32_t test_svread_hor_za32_s32_1(svint32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     uint32_t slice = slice_base + 3;
     return SME_ACLE_FUNC(svread_hor_za32, _s32, _m)(zd, pg, 3, slice);
 }
@@ -141,7 +141,7 @@ svint32_t test_svread_hor_za32_s32_1(svint32_t zd, svbool_t pg, uint32_t slice_b
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.read.horiz.nxv2i64( [[ZD]],  [[TMP0]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svint64_t test_svread_hor_za64_s64(svint64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svint64_t test_svread_hor_za64_s64(svint64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_hor_za64, _s64, _m)(zd, pg, 0, slice_base);
 }
 
@@ -161,7 +161,7 @@ svint64_t test_svread_hor_za64_s64(svint64_t zd, svbool_t pg, uint32_t slice_bas
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.read.horiz.nxv2i64( [[ZD]],  [[TMP0]], i32 7, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svint64_t test_svread_hor_za64_s64_1(svint64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svint64_t test_svread_hor_za64_s64_1(svint64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     uint32_t slice = slice_base + 1;
     return SME_ACLE_FUNC(svread_hor_za64, _s64, _m)(zd, pg, 7, slice);
 }
@@ -178,7 +178,7 @@ svint64_t test_svread_hor_za64_s64_1(svint64_t zd, svbool_t pg, uint32_t slice_b
 // CHECK-CXX-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sme.read.horiz.nxv16i8( [[ZD]],  [[PG]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP0]]
 //
-svuint8_t test_svread_hor_za8_u8(svuint8_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svuint8_t test_svread_hor_za8_u8(svuint8_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_hor_za8, _u8, _m)(zd, pg, 0, slice_base);
 }
 
@@ -196,7 +196,7 @@ svuint8_t test_svread_hor_za8_u8(svuint8_t zd, svbool_t pg, uint32_t slice_base)
 // CHECK-CXX-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sme.read.horiz.nxv16i8( [[ZD]],  [[PG]], i32 0, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret  [[TMP0]]
 //
-svuint8_t test_svread_hor_za8_u8_1(svuint8_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svuint8_t test_svread_hor_za8_u8_1(svuint8_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     uint32_t slice = slice_base + 15;
     return SME_ACLE_FUNC(svread_hor_za8, _u8, _m)(zd, pg, 0, slice);
 }
@@ -215,7 +215,7 @@ svuint8_t test_svread_hor_za8_u8_1(svuint8_t zd, svbool_t pg, uint32_t slice_bas
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.read.horiz.nxv8i16( [[ZD]],  [[TMP0]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svuint16_t test_svread_hor_za16_u16(svuint16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svuint16_t test_svread_hor_za16_u16(svuint16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_hor_za16, _u16, _m)(zd, pg, 0, slice_base);
 }
 
@@ -235,7 +235,7 @@ svuint16_t test_svread_hor_za16_u16(svuint16_t zd, svbool_t pg, uint32_t slice_b
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.read.horiz.nxv8i16( [[ZD]],  [[TMP0]], i32 1, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svuint16_t test_svread_hor_za16_u16_1(svuint16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svuint16_t test_svread_hor_za16_u16_1(svuint16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     uint32_t slice = slice_base + 7;
     return SME_ACLE_FUNC(svread_hor_za16, _u16, _m)(zd, pg, 1, slice);
 }
@@ -254,7 +254,7 @@ svuint16_t test_svread_hor_za16_u16_1(svuint16_t zd, svbool_t pg, uint32_t slice
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.read.horiz.nxv4i32( [[ZD]],  [[TMP0]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svuint32_t test_svread_hor_za32_u32(svuint32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svuint32_t test_svread_hor_za32_u32(svuint32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_hor_za32, _u32, _m)(zd, pg, 0, slice_base);
 }
 
@@ -274,7 +274,7 @@ svuint32_t test_svread_hor_za32_u32(svuint32_t zd, svbool_t pg, uint32_t slice_b
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.read.horiz.nxv4i32( [[ZD]],  [[TMP0]], i32 3, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svuint32_t test_svread_hor_za32_u32_1(svuint32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svuint32_t test_svread_hor_za32_u32_1(svuint32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     uint32_t slice = slice_base + 3;
     return SME_ACLE_FUNC(svread_hor_za32, _u32, _m)(zd, pg, 3, slice);
 }
@@ -293,7 +293,7 @@ svuint32_t test_svread_hor_za32_u32_1(svuint32_t zd, svbool_t pg, uint32_t slice
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.read.horiz.nxv2i64( [[ZD]],  [[TMP0]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svuint64_t test_svread_hor_za64_u64(svuint64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svuint64_t test_svread_hor_za64_u64(svuint64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_hor_za64, _u64, _m)(zd, pg, 0, slice_base);
 }
 
@@ -313,7 +313,7 @@ svuint64_t test_svread_hor_za64_u64(svuint64_t zd, svbool_t pg, uint32_t slice_b
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.read.horiz.nxv2i64( [[ZD]],  [[TMP0]], i32 7, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svuint64_t test_svread_hor_za64_u64_1(svuint64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svuint64_t test_svread_hor_za64_u64_1(svuint64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     uint32_t slice = slice_base + 1;
     return SME_ACLE_FUNC(svread_hor_za64, _u64, _m)(zd, pg, 7, slice);
 }
@@ -332,7 +332,7 @@ svuint64_t test_svread_hor_za64_u64_1(svuint64_t zd, svbool_t pg, uint32_t slice
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.read.horiz.nxv8f16( [[ZD]],  [[TMP0]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svfloat16_t test_svread_hor_za16_f16(svfloat16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svfloat16_t test_svread_hor_za16_f16(svfloat16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_hor_za16, _f16, _m)(zd, pg, 0, slice_base);
 }
 
@@ -352,7 +352,7 @@ svfloat16_t test_svread_hor_za16_f16(svfloat16_t zd, svbool_t pg, uint32_t slice
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.read.horiz.nxv8f16( [[ZD]],  [[TMP0]], i32 1, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svfloat16_t test_svread_hor_za16_f16_1(svfloat16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svfloat16_t test_svread_hor_za16_f16_1(svfloat16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     uint32_t slice = slice_base + 7;
     return SME_ACLE_FUNC(svread_hor_za16, _f16, _m)(zd, pg, 1, slice);
 }
@@ -371,7 +371,7 @@ svfloat16_t test_svread_hor_za16_f16_1(svfloat16_t zd, svbool_t pg, uint32_t sli
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.read.horiz.nxv8bf16( [[ZD]],  [[TMP0]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svbfloat16_t test_svread_hor_za16_bf16(svbfloat16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svbfloat16_t test_svread_hor_za16_bf16(svbfloat16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_hor_za16, _bf16, _m)(zd, pg, 0, slice_base);
 }
 
@@ -391,7 +391,7 @@ svbfloat16_t test_svread_hor_za16_bf16(svbfloat16_t zd, svbool_t pg, uint32_t sl
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.read.horiz.nxv8bf16( [[ZD]],  [[TMP0]], i32 1, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svbfloat16_t test_svread_hor_za16_bf16_1(svbfloat16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svbfloat16_t test_svread_hor_za16_bf16_1(svbfloat16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     uint32_t slice = slice_base + 7;
     return SME_ACLE_FUNC(svread_hor_za16, _bf16, _m)(zd, pg, 1, slice);
 }
@@ -410,7 +410,7 @@ svbfloat16_t test_svread_hor_za16_bf16_1(svbfloat16_t zd, svbool_t pg, uint32_t
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.read.horiz.nxv4f32( [[ZD]],  [[TMP0]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svfloat32_t test_svread_hor_za32_f32(svfloat32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svfloat32_t test_svread_hor_za32_f32(svfloat32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_hor_za32, _f32, _m)(zd, pg, 0, slice_base);
 }
 
@@ -430,7 +430,7 @@ svfloat32_t test_svread_hor_za32_f32(svfloat32_t zd, svbool_t pg, uint32_t slice
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.read.horiz.nxv4f32( [[ZD]],  [[TMP0]], i32 3, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svfloat32_t test_svread_hor_za32_f32_1(svfloat32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svfloat32_t test_svread_hor_za32_f32_1(svfloat32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     uint32_t slice = slice_base + 3;
     return SME_ACLE_FUNC(svread_hor_za32, _f32, _m)(zd, pg, 3, slice);
 }
@@ -449,7 +449,7 @@ svfloat32_t test_svread_hor_za32_f32_1(svfloat32_t zd, svbool_t pg, uint32_t sli
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.read.horiz.nxv2f64( [[ZD]],  [[TMP0]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svfloat64_t test_svread_hor_za64_f64(svfloat64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svfloat64_t test_svread_hor_za64_f64(svfloat64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_hor_za64, _f64, _m)(zd, pg, 0, slice_base);
 }
 
@@ -469,7 +469,7 @@ svfloat64_t test_svread_hor_za64_f64(svfloat64_t zd, svbool_t pg, uint32_t slice
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.read.horiz.nxv2f64( [[ZD]],  [[TMP0]], i32 7, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svfloat64_t test_svread_hor_za64_f64_1(svfloat64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svfloat64_t test_svread_hor_za64_f64_1(svfloat64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     uint32_t slice = slice_base + 1;
     return SME_ACLE_FUNC(svread_hor_za64, _f64, _m)(zd, pg, 7, slice);
 }
@@ -486,7 +486,7 @@ svfloat64_t test_svread_hor_za64_f64_1(svfloat64_t zd, svbool_t pg, uint32_t sli
 // CHECK-CXX-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sme.readq.horiz.nxv16i8( [[ZD]],  [[PG]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP0]]
 //
-svint8_t test_svread_hor_za128_s8(svint8_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svint8_t test_svread_hor_za128_s8(svint8_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_hor_za128, _s8, _m)(zd, pg, 0, slice_base);
 }
 
@@ -502,7 +502,7 @@ svint8_t test_svread_hor_za128_s8(svint8_t zd, svbool_t pg, uint32_t slice_base)
 // CHECK-CXX-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sme.readq.horiz.nxv16i8( [[ZD]],  [[PG]], i32 15, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP0]]
 //
-svint8_t test_svread_hor_za128_s8_1(svint8_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svint8_t test_svread_hor_za128_s8_1(svint8_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_hor_za128, _s8, _m)(zd, pg, 15, slice_base);
 }
 
@@ -520,7 +520,7 @@ svint8_t test_svread_hor_za128_s8_1(svint8_t zd, svbool_t pg, uint32_t slice_bas
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.readq.horiz.nxv8i16( [[ZD]],  [[TMP0]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svint16_t test_svread_hor_za128_s16(svint16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svint16_t test_svread_hor_za128_s16(svint16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_hor_za128, _s16, _m)(zd, pg, 0, slice_base);
 }
 
@@ -538,7 +538,7 @@ svint16_t test_svread_hor_za128_s16(svint16_t zd, svbool_t pg, uint32_t slice_ba
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.readq.horiz.nxv8i16( [[ZD]],  [[TMP0]], i32 15, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svint16_t test_svread_hor_za128_s16_1(svint16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svint16_t test_svread_hor_za128_s16_1(svint16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_hor_za128, _s16, _m)(zd, pg, 15, slice_base);
 }
 
@@ -556,7 +556,7 @@ svint16_t test_svread_hor_za128_s16_1(svint16_t zd, svbool_t pg, uint32_t slice_
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.readq.horiz.nxv4i32( [[ZD]],  [[TMP0]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svint32_t test_svread_hor_za128_s32(svint32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svint32_t test_svread_hor_za128_s32(svint32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_hor_za128, _s32, _m)(zd, pg, 0, slice_base);
 }
 
@@ -574,7 +574,7 @@ svint32_t test_svread_hor_za128_s32(svint32_t zd, svbool_t pg, uint32_t slice_ba
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.readq.horiz.nxv4i32( [[ZD]],  [[TMP0]], i32 15, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svint32_t test_svread_hor_za128_s32_1(svint32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svint32_t test_svread_hor_za128_s32_1(svint32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_hor_za128, _s32, _m)(zd, pg, 15, slice_base);
 }
 
@@ -592,7 +592,7 @@ svint32_t test_svread_hor_za128_s32_1(svint32_t zd, svbool_t pg, uint32_t slice_
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.readq.horiz.nxv2i64( [[ZD]],  [[TMP0]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svint64_t test_svread_hor_za128_s64(svint64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svint64_t test_svread_hor_za128_s64(svint64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_hor_za128, _s64, _m)(zd, pg, 0, slice_base);
 }
 
@@ -610,7 +610,7 @@ svint64_t test_svread_hor_za128_s64(svint64_t zd, svbool_t pg, uint32_t slice_ba
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.readq.horiz.nxv2i64( [[ZD]],  [[TMP0]], i32 15, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svint64_t test_svread_hor_za128_s64_1(svint64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svint64_t test_svread_hor_za128_s64_1(svint64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_hor_za128, _s64, _m)(zd, pg, 15, slice_base);
 }
 
@@ -626,7 +626,7 @@ svint64_t test_svread_hor_za128_s64_1(svint64_t zd, svbool_t pg, uint32_t slice_
 // CHECK-CXX-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sme.readq.horiz.nxv16i8( [[ZD]],  [[PG]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP0]]
 //
-svuint8_t test_svread_hor_za128_u8(svuint8_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svuint8_t test_svread_hor_za128_u8(svuint8_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_hor_za128, _u8, _m)(zd, pg, 0, slice_base);
 }
 
@@ -642,7 +642,7 @@ svuint8_t test_svread_hor_za128_u8(svuint8_t zd, svbool_t pg, uint32_t slice_bas
 // CHECK-CXX-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sme.readq.horiz.nxv16i8( [[ZD]],  [[PG]], i32 15, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP0]]
 //
-svuint8_t test_svread_hor_za128_u8_1(svuint8_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svuint8_t test_svread_hor_za128_u8_1(svuint8_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_hor_za128, _u8, _m)(zd, pg, 15, slice_base);
 }
 
@@ -660,7 +660,7 @@ svuint8_t test_svread_hor_za128_u8_1(svuint8_t zd, svbool_t pg, uint32_t slice_b
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.readq.horiz.nxv8i16( [[ZD]],  [[TMP0]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svuint16_t test_svread_hor_za128_u16(svuint16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svuint16_t test_svread_hor_za128_u16(svuint16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_hor_za128, _u16, _m)(zd, pg, 0, slice_base);
 }
 
@@ -678,7 +678,7 @@ svuint16_t test_svread_hor_za128_u16(svuint16_t zd, svbool_t pg, uint32_t slice_
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.readq.horiz.nxv8i16( [[ZD]],  [[TMP0]], i32 15, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svuint16_t test_svread_hor_za128_u16_1(svuint16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svuint16_t test_svread_hor_za128_u16_1(svuint16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_hor_za128, _u16, _m)(zd, pg, 15, slice_base);
 }
 
@@ -696,7 +696,7 @@ svuint16_t test_svread_hor_za128_u16_1(svuint16_t zd, svbool_t pg, uint32_t slic
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.readq.horiz.nxv4i32( [[ZD]],  [[TMP0]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svuint32_t test_svread_hor_za128_u32(svuint32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svuint32_t test_svread_hor_za128_u32(svuint32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_hor_za128, _u32, _m)(zd, pg, 0, slice_base);
 }
 
@@ -714,7 +714,7 @@ svuint32_t test_svread_hor_za128_u32(svuint32_t zd, svbool_t pg, uint32_t slice_
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.readq.horiz.nxv4i32( [[ZD]],  [[TMP0]], i32 15, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svuint32_t test_svread_hor_za128_u32_1(svuint32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svuint32_t test_svread_hor_za128_u32_1(svuint32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_hor_za128, _u32, _m)(zd, pg, 15, slice_base);
 }
 
@@ -732,7 +732,7 @@ svuint32_t test_svread_hor_za128_u32_1(svuint32_t zd, svbool_t pg, uint32_t slic
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.readq.horiz.nxv2i64( [[ZD]],  [[TMP0]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svuint64_t test_svread_hor_za128_u64(svuint64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svuint64_t test_svread_hor_za128_u64(svuint64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_hor_za128, _u64, _m)(zd, pg, 0, slice_base);
 }
 
@@ -750,7 +750,7 @@ svuint64_t test_svread_hor_za128_u64(svuint64_t zd, svbool_t pg, uint32_t slice_
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.readq.horiz.nxv2i64( [[ZD]],  [[TMP0]], i32 15, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svuint64_t test_svread_hor_za128_u64_1(svuint64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svuint64_t test_svread_hor_za128_u64_1(svuint64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_hor_za128, _u64, _m)(zd, pg, 15, slice_base);
 }
 
@@ -768,7 +768,7 @@ svuint64_t test_svread_hor_za128_u64_1(svuint64_t zd, svbool_t pg, uint32_t slic
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.readq.horiz.nxv8f16( [[ZD]],  [[TMP0]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svfloat16_t test_svread_hor_za128_f16(svfloat16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svfloat16_t test_svread_hor_za128_f16(svfloat16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_hor_za128, _f16, _m)(zd, pg, 0, slice_base);
 }
 
@@ -786,7 +786,7 @@ svfloat16_t test_svread_hor_za128_f16(svfloat16_t zd, svbool_t pg, uint32_t slic
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.readq.horiz.nxv8f16( [[ZD]],  [[TMP0]], i32 15, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svfloat16_t test_svread_hor_za128_f16_1(svfloat16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svfloat16_t test_svread_hor_za128_f16_1(svfloat16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_hor_za128, _f16, _m)(zd, pg, 15, slice_base);
 }
 
@@ -804,7 +804,7 @@ svfloat16_t test_svread_hor_za128_f16_1(svfloat16_t zd, svbool_t pg, uint32_t sl
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.readq.horiz.nxv8bf16( [[ZD]],  [[TMP0]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svbfloat16_t test_svread_hor_za128_bf16(svbfloat16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svbfloat16_t test_svread_hor_za128_bf16(svbfloat16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_hor_za128, _bf16, _m)(zd, pg, 0, slice_base);
 }
 
@@ -822,7 +822,7 @@ svbfloat16_t test_svread_hor_za128_bf16(svbfloat16_t zd, svbool_t pg, uint32_t s
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.readq.horiz.nxv8bf16( [[ZD]],  [[TMP0]], i32 15, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svbfloat16_t test_svread_hor_za128_bf16_1(svbfloat16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svbfloat16_t test_svread_hor_za128_bf16_1(svbfloat16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_hor_za128, _bf16, _m)(zd, pg, 15, slice_base);
 }
 
@@ -840,7 +840,7 @@ svbfloat16_t test_svread_hor_za128_bf16_1(svbfloat16_t zd, svbool_t pg, uint32_t
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.readq.horiz.nxv4f32( [[ZD]],  [[TMP0]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svfloat32_t test_svread_hor_za128_f32(svfloat32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svfloat32_t test_svread_hor_za128_f32(svfloat32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_hor_za128, _f32, _m)(zd, pg, 0, slice_base);
 }
 
@@ -858,7 +858,7 @@ svfloat32_t test_svread_hor_za128_f32(svfloat32_t zd, svbool_t pg, uint32_t slic
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.readq.horiz.nxv4f32( [[ZD]],  [[TMP0]], i32 15, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svfloat32_t test_svread_hor_za128_f32_1(svfloat32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svfloat32_t test_svread_hor_za128_f32_1(svfloat32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_hor_za128, _f32, _m)(zd, pg, 15, slice_base);
 }
 
@@ -876,7 +876,7 @@ svfloat32_t test_svread_hor_za128_f32_1(svfloat32_t zd, svbool_t pg, uint32_t sl
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.readq.horiz.nxv2f64( [[ZD]],  [[TMP0]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svfloat64_t test_svread_hor_za128_f64(svfloat64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svfloat64_t test_svread_hor_za128_f64(svfloat64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_hor_za128, _f64, _m)(zd, pg, 0, slice_base);
 }
 
@@ -894,7 +894,7 @@ svfloat64_t test_svread_hor_za128_f64(svfloat64_t zd, svbool_t pg, uint32_t slic
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.readq.horiz.nxv2f64( [[ZD]],  [[TMP0]], i32 15, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svfloat64_t test_svread_hor_za128_f64_1(svfloat64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svfloat64_t test_svread_hor_za128_f64_1(svfloat64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_hor_za128, _f64, _m)(zd, pg, 15, slice_base);
 }
 
@@ -910,7 +910,7 @@ svfloat64_t test_svread_hor_za128_f64_1(svfloat64_t zd, svbool_t pg, uint32_t sl
 // CHECK-CXX-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sme.read.vert.nxv16i8( [[ZD]],  [[PG]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP0]]
 //
-svint8_t test_svread_ver_za8_s8(svint8_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svint8_t test_svread_ver_za8_s8(svint8_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_ver_za8, _s8, _m)(zd, pg, 0, slice_base);
 }
 
@@ -928,7 +928,7 @@ svint8_t test_svread_ver_za8_s8(svint8_t zd, svbool_t pg, uint32_t slice_base) _
 // CHECK-CXX-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sme.read.vert.nxv16i8( [[ZD]],  [[PG]], i32 0, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret  [[TMP0]]
 //
-svint8_t test_svread_ver_za8_s8_1(svint8_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svint8_t test_svread_ver_za8_s8_1(svint8_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     uint32_t slice = slice_base + 15;
     return SME_ACLE_FUNC(svread_ver_za8, _s8, _m)(zd, pg, 0, slice);
 }
@@ -947,7 +947,7 @@ svint8_t test_svread_ver_za8_s8_1(svint8_t zd, svbool_t pg, uint32_t slice_base)
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.read.vert.nxv8i16( [[ZD]],  [[TMP0]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svint16_t test_svread_ver_za16_s16(svint16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svint16_t test_svread_ver_za16_s16(svint16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
      return SME_ACLE_FUNC(svread_ver_za16, _s16, _m)(zd, pg, 0, slice_base);
 }
 
@@ -967,7 +967,7 @@ svint16_t test_svread_ver_za16_s16(svint16_t zd, svbool_t pg, uint32_t slice_bas
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.read.vert.nxv8i16( [[ZD]],  [[TMP0]], i32 1, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svint16_t test_svread_ver_za16_s16_1(svint16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svint16_t test_svread_ver_za16_s16_1(svint16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
      uint32_t slice = slice_base + 7;
      return SME_ACLE_FUNC(svread_ver_za16, _s16, _m)(zd, pg, 1, slice);
 }
@@ -986,7 +986,7 @@ svint16_t test_svread_ver_za16_s16_1(svint16_t zd, svbool_t pg, uint32_t slice_b
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.read.vert.nxv4i32( [[ZD]],  [[TMP0]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svint32_t test_svread_ver_za32_s32(svint32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svint32_t test_svread_ver_za32_s32(svint32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_ver_za32, _s32, _m)(zd, pg, 0, slice_base);
 }
 
@@ -1006,7 +1006,7 @@ svint32_t test_svread_ver_za32_s32(svint32_t zd, svbool_t pg, uint32_t slice_bas
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.read.vert.nxv4i32( [[ZD]],  [[TMP0]], i32 3, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svint32_t test_svread_ver_za32_s32_1(svint32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svint32_t test_svread_ver_za32_s32_1(svint32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     uint32_t slice = slice_base + 3;
     return SME_ACLE_FUNC(svread_ver_za32, _s32, _m)(zd, pg, 3, slice);
 }
@@ -1025,7 +1025,7 @@ svint32_t test_svread_ver_za32_s32_1(svint32_t zd, svbool_t pg, uint32_t slice_b
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.read.vert.nxv2i64( [[ZD]],  [[TMP0]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svint64_t test_svread_ver_za64_s64(svint64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svint64_t test_svread_ver_za64_s64(svint64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_ver_za64, _s64, _m)(zd, pg, 0, slice_base);
 }
 
@@ -1045,7 +1045,7 @@ svint64_t test_svread_ver_za64_s64(svint64_t zd, svbool_t pg, uint32_t slice_bas
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.read.vert.nxv2i64( [[ZD]],  [[TMP0]], i32 7, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svint64_t test_svread_ver_za64_s64_1(svint64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svint64_t test_svread_ver_za64_s64_1(svint64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     uint32_t slice = slice_base + 1;
     return SME_ACLE_FUNC(svread_ver_za64, _s64, _m)(zd, pg, 7, slice);
 }
@@ -1062,7 +1062,7 @@ svint64_t test_svread_ver_za64_s64_1(svint64_t zd, svbool_t pg, uint32_t slice_b
 // CHECK-CXX-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sme.read.vert.nxv16i8( [[ZD]],  [[PG]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP0]]
 //
-svuint8_t test_svread_ver_za8_u8(svuint8_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svuint8_t test_svread_ver_za8_u8(svuint8_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_ver_za8, _u8, _m)(zd, pg, 0, slice_base);
 }
 
@@ -1080,7 +1080,7 @@ svuint8_t test_svread_ver_za8_u8(svuint8_t zd, svbool_t pg, uint32_t slice_base)
 // CHECK-CXX-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sme.read.vert.nxv16i8( [[ZD]],  [[PG]], i32 0, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret  [[TMP0]]
 //
-svuint8_t test_svread_ver_za8_u8_1(svuint8_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svuint8_t test_svread_ver_za8_u8_1(svuint8_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     uint32_t slice = slice_base + 15;
     return SME_ACLE_FUNC(svread_ver_za8, _u8, _m)(zd, pg, 0, slice);
 }
@@ -1099,7 +1099,7 @@ svuint8_t test_svread_ver_za8_u8_1(svuint8_t zd, svbool_t pg, uint32_t slice_bas
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.read.vert.nxv8i16( [[ZD]],  [[TMP0]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svuint16_t test_svread_ver_za16_u16(svuint16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svuint16_t test_svread_ver_za16_u16(svuint16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_ver_za16, _u16, _m)(zd, pg, 0, slice_base);
 }
 
@@ -1119,7 +1119,7 @@ svuint16_t test_svread_ver_za16_u16(svuint16_t zd, svbool_t pg, uint32_t slice_b
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.read.vert.nxv8i16( [[ZD]],  [[TMP0]], i32 1, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svuint16_t test_svread_ver_za16_u16_1(svuint16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svuint16_t test_svread_ver_za16_u16_1(svuint16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     uint32_t slice = slice_base + 7;
     return SME_ACLE_FUNC(svread_ver_za16, _u16, _m)(zd, pg, 1, slice);
 }
@@ -1138,7 +1138,7 @@ svuint16_t test_svread_ver_za16_u16_1(svuint16_t zd, svbool_t pg, uint32_t slice
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.read.vert.nxv4i32( [[ZD]],  [[TMP0]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svuint32_t test_svread_ver_za32_u32(svuint32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svuint32_t test_svread_ver_za32_u32(svuint32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_ver_za32, _u32, _m)(zd, pg, 0, slice_base);
 }
 
@@ -1158,7 +1158,7 @@ svuint32_t test_svread_ver_za32_u32(svuint32_t zd, svbool_t pg, uint32_t slice_b
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.read.vert.nxv4i32( [[ZD]],  [[TMP0]], i32 3, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svuint32_t test_svread_ver_za32_u32_1(svuint32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svuint32_t test_svread_ver_za32_u32_1(svuint32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     uint32_t slice = slice_base + 3;
     return SME_ACLE_FUNC(svread_ver_za32, _u32, _m)(zd, pg, 3, slice);
 }
@@ -1177,7 +1177,7 @@ svuint32_t test_svread_ver_za32_u32_1(svuint32_t zd, svbool_t pg, uint32_t slice
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.read.vert.nxv2i64( [[ZD]],  [[TMP0]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svuint64_t test_svread_ver_za64_u64(svuint64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svuint64_t test_svread_ver_za64_u64(svuint64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_ver_za64, _u64, _m)(zd, pg, 0, slice_base);
 }
 
@@ -1197,7 +1197,7 @@ svuint64_t test_svread_ver_za64_u64(svuint64_t zd, svbool_t pg, uint32_t slice_b
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.read.vert.nxv2i64( [[ZD]],  [[TMP0]], i32 7, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svuint64_t test_svread_ver_za64_u64_1(svuint64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svuint64_t test_svread_ver_za64_u64_1(svuint64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     uint32_t slice = slice_base + 1;
     return SME_ACLE_FUNC(svread_ver_za64, _u64, _m)(zd, pg, 7, slice);
 }
@@ -1216,7 +1216,7 @@ svuint64_t test_svread_ver_za64_u64_1(svuint64_t zd, svbool_t pg, uint32_t slice
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.read.vert.nxv8f16( [[ZD]],  [[TMP0]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svfloat16_t test_svread_ver_za16_f16(svfloat16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svfloat16_t test_svread_ver_za16_f16(svfloat16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_ver_za16, _f16, _m)(zd, pg, 0, slice_base);
 }
 
@@ -1236,7 +1236,7 @@ svfloat16_t test_svread_ver_za16_f16(svfloat16_t zd, svbool_t pg, uint32_t slice
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.read.vert.nxv8f16( [[ZD]],  [[TMP0]], i32 1, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svfloat16_t test_svread_ver_za16_f16_1(svfloat16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svfloat16_t test_svread_ver_za16_f16_1(svfloat16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     uint32_t slice = slice_base + 7;
     return SME_ACLE_FUNC(svread_ver_za16, _f16, _m)(zd, pg, 1, slice);
 }
@@ -1255,7 +1255,7 @@ svfloat16_t test_svread_ver_za16_f16_1(svfloat16_t zd, svbool_t pg, uint32_t sli
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.read.vert.nxv8bf16( [[ZD]],  [[TMP0]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svbfloat16_t test_svread_ver_za16_bf16(svbfloat16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svbfloat16_t test_svread_ver_za16_bf16(svbfloat16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_ver_za16, _bf16, _m)(zd, pg, 0, slice_base);
 }
 
@@ -1275,7 +1275,7 @@ svbfloat16_t test_svread_ver_za16_bf16(svbfloat16_t zd, svbool_t pg, uint32_t sl
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.read.vert.nxv8bf16( [[ZD]],  [[TMP0]], i32 1, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svbfloat16_t test_svread_ver_za16_bf16_1(svbfloat16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svbfloat16_t test_svread_ver_za16_bf16_1(svbfloat16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     uint32_t slice = slice_base + 7;
     return SME_ACLE_FUNC(svread_ver_za16, _bf16, _m)(zd, pg, 1, slice);
 }
@@ -1294,7 +1294,7 @@ svbfloat16_t test_svread_ver_za16_bf16_1(svbfloat16_t zd, svbool_t pg, uint32_t
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.read.vert.nxv4f32( [[ZD]],  [[TMP0]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svfloat32_t test_svread_ver_za32_f32(svfloat32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svfloat32_t test_svread_ver_za32_f32(svfloat32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_ver_za32, _f32, _m)(zd, pg, 0, slice_base);
 }
 
@@ -1314,7 +1314,7 @@ svfloat32_t test_svread_ver_za32_f32(svfloat32_t zd, svbool_t pg, uint32_t slice
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.read.vert.nxv4f32( [[ZD]],  [[TMP0]], i32 3, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svfloat32_t test_svread_ver_za32_f32_1(svfloat32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svfloat32_t test_svread_ver_za32_f32_1(svfloat32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     uint32_t slice = slice_base + 3;
     return SME_ACLE_FUNC(svread_ver_za32, _f32, _m)(zd, pg, 3, slice);
 }
@@ -1333,7 +1333,7 @@ svfloat32_t test_svread_ver_za32_f32_1(svfloat32_t zd, svbool_t pg, uint32_t sli
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.read.vert.nxv2f64( [[ZD]],  [[TMP0]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svfloat64_t test_svread_ver_za64_f64(svfloat64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svfloat64_t test_svread_ver_za64_f64(svfloat64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_ver_za64, _f64, _m)(zd, pg, 0, slice_base);
 }
 
@@ -1353,7 +1353,7 @@ svfloat64_t test_svread_ver_za64_f64(svfloat64_t zd, svbool_t pg, uint32_t slice
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.read.vert.nxv2f64( [[ZD]],  [[TMP0]], i32 7, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svfloat64_t test_svread_ver_za64_f64_1(svfloat64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svfloat64_t test_svread_ver_za64_f64_1(svfloat64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     uint32_t slice = slice_base + 1;
     return SME_ACLE_FUNC(svread_ver_za64, _f64, _m)(zd, pg, 7, slice);
 }
@@ -1370,7 +1370,7 @@ svfloat64_t test_svread_ver_za64_f64_1(svfloat64_t zd, svbool_t pg, uint32_t sli
 // CHECK-CXX-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sme.readq.vert.nxv16i8( [[ZD]],  [[PG]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP0]]
 //
-svint8_t test_svread_ver_za128_s8(svint8_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svint8_t test_svread_ver_za128_s8(svint8_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_ver_za128, _s8, _m)(zd, pg, 0, slice_base);
 }
 
@@ -1386,7 +1386,7 @@ svint8_t test_svread_ver_za128_s8(svint8_t zd, svbool_t pg, uint32_t slice_base)
 // CHECK-CXX-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sme.readq.vert.nxv16i8( [[ZD]],  [[PG]], i32 15, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP0]]
 //
-svint8_t test_svread_ver_za128_s8_1(svint8_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svint8_t test_svread_ver_za128_s8_1(svint8_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_ver_za128, _s8, _m)(zd, pg, 15, slice_base);
 }
 
@@ -1404,7 +1404,7 @@ svint8_t test_svread_ver_za128_s8_1(svint8_t zd, svbool_t pg, uint32_t slice_bas
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.readq.vert.nxv8i16( [[ZD]],  [[TMP0]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svint16_t test_svread_ver_za128_s16(svint16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svint16_t test_svread_ver_za128_s16(svint16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_ver_za128, _s16, _m)(zd, pg, 0, slice_base);
 }
 
@@ -1422,7 +1422,7 @@ svint16_t test_svread_ver_za128_s16(svint16_t zd, svbool_t pg, uint32_t slice_ba
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.readq.vert.nxv8i16( [[ZD]],  [[TMP0]], i32 15, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svint16_t test_svread_ver_za128_s16_1(svint16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svint16_t test_svread_ver_za128_s16_1(svint16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_ver_za128, _s16, _m)(zd, pg, 15, slice_base);
 }
 
@@ -1440,7 +1440,7 @@ svint16_t test_svread_ver_za128_s16_1(svint16_t zd, svbool_t pg, uint32_t slice_
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.readq.vert.nxv4i32( [[ZD]],  [[TMP0]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svint32_t test_svread_ver_za128_s32(svint32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svint32_t test_svread_ver_za128_s32(svint32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_ver_za128, _s32, _m)(zd, pg, 0, slice_base);
 }
 
@@ -1458,7 +1458,7 @@ svint32_t test_svread_ver_za128_s32(svint32_t zd, svbool_t pg, uint32_t slice_ba
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.readq.vert.nxv4i32( [[ZD]],  [[TMP0]], i32 15, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svint32_t test_svread_ver_za128_s32_1(svint32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svint32_t test_svread_ver_za128_s32_1(svint32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_ver_za128, _s32, _m)(zd, pg, 15, slice_base);
 }
 
@@ -1476,7 +1476,7 @@ svint32_t test_svread_ver_za128_s32_1(svint32_t zd, svbool_t pg, uint32_t slice_
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.readq.vert.nxv2i64( [[ZD]],  [[TMP0]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svint64_t test_svread_ver_za128_s64(svint64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svint64_t test_svread_ver_za128_s64(svint64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_ver_za128, _s64, _m)(zd, pg, 0, slice_base);
 }
 
@@ -1494,7 +1494,7 @@ svint64_t test_svread_ver_za128_s64(svint64_t zd, svbool_t pg, uint32_t slice_ba
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.readq.vert.nxv2i64( [[ZD]],  [[TMP0]], i32 15, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svint64_t test_svread_ver_za128_s64_1(svint64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svint64_t test_svread_ver_za128_s64_1(svint64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_ver_za128, _s64, _m)(zd, pg, 15, slice_base);
 }
 
@@ -1510,7 +1510,7 @@ svint64_t test_svread_ver_za128_s64_1(svint64_t zd, svbool_t pg, uint32_t slice_
 // CHECK-CXX-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sme.readq.vert.nxv16i8( [[ZD]],  [[PG]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP0]]
 //
-svuint8_t test_svread_ver_za128_u8(svuint8_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svuint8_t test_svread_ver_za128_u8(svuint8_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_ver_za128, _u8, _m)(zd, pg, 0, slice_base);
 }
 
@@ -1526,7 +1526,7 @@ svuint8_t test_svread_ver_za128_u8(svuint8_t zd, svbool_t pg, uint32_t slice_bas
 // CHECK-CXX-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sme.readq.vert.nxv16i8( [[ZD]],  [[PG]], i32 15, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP0]]
 //
-svuint8_t test_svread_ver_za128_u8_1(svuint8_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svuint8_t test_svread_ver_za128_u8_1(svuint8_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_ver_za128, _u8, _m)(zd, pg, 15, slice_base);
 }
 
@@ -1544,7 +1544,7 @@ svuint8_t test_svread_ver_za128_u8_1(svuint8_t zd, svbool_t pg, uint32_t slice_b
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.readq.vert.nxv8i16( [[ZD]],  [[TMP0]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svuint16_t test_svread_ver_za128_u16(svuint16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svuint16_t test_svread_ver_za128_u16(svuint16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_ver_za128, _u16, _m)(zd, pg, 0, slice_base);
 }
 
@@ -1562,7 +1562,7 @@ svuint16_t test_svread_ver_za128_u16(svuint16_t zd, svbool_t pg, uint32_t slice_
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.readq.vert.nxv8i16( [[ZD]],  [[TMP0]], i32 15, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svuint16_t test_svread_ver_za128_u16_1(svuint16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svuint16_t test_svread_ver_za128_u16_1(svuint16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_ver_za128, _u16, _m)(zd, pg, 15, slice_base);
 }
 
@@ -1580,7 +1580,7 @@ svuint16_t test_svread_ver_za128_u16_1(svuint16_t zd, svbool_t pg, uint32_t slic
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.readq.vert.nxv4i32( [[ZD]],  [[TMP0]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svuint32_t test_svread_ver_za128_u32(svuint32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svuint32_t test_svread_ver_za128_u32(svuint32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_ver_za128, _u32, _m)(zd, pg, 0, slice_base);
 }
 
@@ -1598,7 +1598,7 @@ svuint32_t test_svread_ver_za128_u32(svuint32_t zd, svbool_t pg, uint32_t slice_
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.readq.vert.nxv4i32( [[ZD]],  [[TMP0]], i32 15, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svuint32_t test_svread_ver_za128_u32_1(svuint32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svuint32_t test_svread_ver_za128_u32_1(svuint32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_ver_za128, _u32, _m)(zd, pg, 15, slice_base);
 }
 
@@ -1616,7 +1616,7 @@ svuint32_t test_svread_ver_za128_u32_1(svuint32_t zd, svbool_t pg, uint32_t slic
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.readq.vert.nxv2i64( [[ZD]],  [[TMP0]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svuint64_t test_svread_ver_za128_u64(svuint64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svuint64_t test_svread_ver_za128_u64(svuint64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_ver_za128, _u64, _m)(zd, pg, 0, slice_base);
 }
 
@@ -1634,7 +1634,7 @@ svuint64_t test_svread_ver_za128_u64(svuint64_t zd, svbool_t pg, uint32_t slice_
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.readq.vert.nxv2i64( [[ZD]],  [[TMP0]], i32 15, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svuint64_t test_svread_ver_za128_u64_1(svuint64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svuint64_t test_svread_ver_za128_u64_1(svuint64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_ver_za128, _u64, _m)(zd, pg, 15, slice_base);
 }
 
@@ -1652,7 +1652,7 @@ svuint64_t test_svread_ver_za128_u64_1(svuint64_t zd, svbool_t pg, uint32_t slic
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.readq.vert.nxv8f16( [[ZD]],  [[TMP0]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svfloat16_t test_svread_ver_za128_f16(svfloat16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svfloat16_t test_svread_ver_za128_f16(svfloat16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_ver_za128, _f16, _m)(zd, pg, 0, slice_base);
 }
 
@@ -1670,7 +1670,7 @@ svfloat16_t test_svread_ver_za128_f16(svfloat16_t zd, svbool_t pg, uint32_t slic
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.readq.vert.nxv8f16( [[ZD]],  [[TMP0]], i32 15, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svfloat16_t test_svread_ver_za128_f16_1(svfloat16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svfloat16_t test_svread_ver_za128_f16_1(svfloat16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_ver_za128, _f16, _m)(zd, pg, 15, slice_base);
 }
 
@@ -1688,7 +1688,7 @@ svfloat16_t test_svread_ver_za128_f16_1(svfloat16_t zd, svbool_t pg, uint32_t sl
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.readq.vert.nxv8bf16( [[ZD]],  [[TMP0]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svbfloat16_t test_svread_ver_za128_bf16(svbfloat16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svbfloat16_t test_svread_ver_za128_bf16(svbfloat16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_ver_za128, _bf16, _m)(zd, pg, 0, slice_base);
 }
 
@@ -1706,7 +1706,7 @@ svbfloat16_t test_svread_ver_za128_bf16(svbfloat16_t zd, svbool_t pg, uint32_t s
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.readq.vert.nxv8bf16( [[ZD]],  [[TMP0]], i32 15, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svbfloat16_t test_svread_ver_za128_bf16_1(svbfloat16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svbfloat16_t test_svread_ver_za128_bf16_1(svbfloat16_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_ver_za128, _bf16, _m)(zd, pg, 15, slice_base);
 }
 
@@ -1724,7 +1724,7 @@ svbfloat16_t test_svread_ver_za128_bf16_1(svbfloat16_t zd, svbool_t pg, uint32_t
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.readq.vert.nxv4f32( [[ZD]],  [[TMP0]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svfloat32_t test_svread_ver_za128_f32(svfloat32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svfloat32_t test_svread_ver_za128_f32(svfloat32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_ver_za128, _f32, _m)(zd, pg, 0, slice_base);
 }
 
@@ -1742,7 +1742,7 @@ svfloat32_t test_svread_ver_za128_f32(svfloat32_t zd, svbool_t pg, uint32_t slic
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.readq.vert.nxv4f32( [[ZD]],  [[TMP0]], i32 15, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svfloat32_t test_svread_ver_za128_f32_1(svfloat32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svfloat32_t test_svread_ver_za128_f32_1(svfloat32_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_ver_za128, _f32, _m)(zd, pg, 15, slice_base);
 }
 
@@ -1760,7 +1760,7 @@ svfloat32_t test_svread_ver_za128_f32_1(svfloat32_t zd, svbool_t pg, uint32_t sl
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.readq.vert.nxv2f64( [[ZD]],  [[TMP0]], i32 0, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svfloat64_t test_svread_ver_za128_f64(svfloat64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svfloat64_t test_svread_ver_za128_f64(svfloat64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_ver_za128, _f64, _m)(zd, pg, 0, slice_base);
 }
 
@@ -1778,7 +1778,7 @@ svfloat64_t test_svread_ver_za128_f64(svfloat64_t zd, svbool_t pg, uint32_t slic
 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sme.readq.vert.nxv2f64( [[ZD]],  [[TMP0]], i32 15, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret  [[TMP1]]
 //
-svfloat64_t test_svread_ver_za128_f64_1(svfloat64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_shared_za {
+svfloat64_t test_svread_ver_za128_f64_1(svfloat64_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
     return SME_ACLE_FUNC(svread_ver_za128, _f64, _m)(zd, pg, 15, slice_base);
 }
 //// NOTE: These prefixes are unused and the list is autogenerated. Do not add tests below this line:
diff --git a/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_st1.c b/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_st1.c
index 98ebbefc2e74c..97e20ab262902 100644
--- a/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_st1.c
+++ b/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_st1.c
@@ -22,7 +22,7 @@
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.st1b.horiz( [[PG]], ptr [[PTR]], i32 0, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svst1_hor_za8(uint32_t slice_base, svbool_t pg, void *ptr) __arm_streaming __arm_shared_za {
+void test_svst1_hor_za8(uint32_t slice_base, svbool_t pg, void *ptr) __arm_streaming __arm_in("za") {
   svst1_hor_za8(0, slice_base, pg, ptr);
   svst1_hor_za8(0, slice_base + 15, pg, ptr);
 }
@@ -45,7 +45,7 @@ void test_svst1_hor_za8(uint32_t slice_base, svbool_t pg, void *ptr) __arm_strea
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.st1h.horiz( [[TMP0]], ptr [[PTR]], i32 1, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svst1_hor_za16(uint32_t slice_base, svbool_t pg, void *ptr) __arm_streaming __arm_shared_za {
+void test_svst1_hor_za16(uint32_t slice_base, svbool_t pg, void *ptr) __arm_streaming __arm_in("za") {
   svst1_hor_za16(0, slice_base, pg, ptr);
   svst1_hor_za16(1, slice_base + 7, pg, ptr);
 }
@@ -68,7 +68,7 @@ void test_svst1_hor_za16(uint32_t slice_base, svbool_t pg, void *ptr) __arm_stre
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.st1w.horiz( [[TMP0]], ptr [[PTR]], i32 3, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svst1_hor_za32(uint32_t slice_base, svbool_t pg, void *ptr) __arm_streaming __arm_shared_za {
+void test_svst1_hor_za32(uint32_t slice_base, svbool_t pg, void *ptr) __arm_streaming __arm_in("za") {
   svst1_hor_za32(0, slice_base, pg, ptr);
   svst1_hor_za32(3, slice_base + 3, pg, ptr);
 }
@@ -91,7 +91,7 @@ void test_svst1_hor_za32(uint32_t slice_base, svbool_t pg, void *ptr) __arm_stre
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.st1d.horiz( [[TMP0]], ptr [[PTR]], i32 7, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svst1_hor_za64(uint32_t slice_base, svbool_t pg, void *ptr) __arm_streaming __arm_shared_za {
+void test_svst1_hor_za64(uint32_t slice_base, svbool_t pg, void *ptr) __arm_streaming __arm_in("za") {
   svst1_hor_za64(0, slice_base, pg, ptr);
   svst1_hor_za64(7, slice_base + 1, pg, ptr);
 }
@@ -112,7 +112,7 @@ void test_svst1_hor_za64(uint32_t slice_base, svbool_t pg, void *ptr) __arm_stre
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.st1q.horiz( [[TMP0]], ptr [[PTR]], i32 15, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svst1_hor_za128(uint32_t slice_base, svbool_t pg, void *ptr) __arm_streaming __arm_shared_za {
+void test_svst1_hor_za128(uint32_t slice_base, svbool_t pg, void *ptr) __arm_streaming __arm_in("za") {
   svst1_hor_za128(0, slice_base, pg, ptr);
   svst1_hor_za128(15, slice_base, pg, ptr);
 }
@@ -133,7 +133,7 @@ void test_svst1_hor_za128(uint32_t slice_base, svbool_t pg, void *ptr) __arm_str
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.st1b.vert( [[PG]], ptr [[PTR]], i32 0, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svst1_ver_za8(uint32_t slice_base, svbool_t pg, void *ptr) __arm_streaming __arm_shared_za {
+void test_svst1_ver_za8(uint32_t slice_base, svbool_t pg, void *ptr) __arm_streaming __arm_in("za") {
   svst1_ver_za8(0, slice_base, pg, ptr);
   svst1_ver_za8(0, slice_base + 15, pg, ptr);
 }
@@ -156,7 +156,7 @@ void test_svst1_ver_za8(uint32_t slice_base, svbool_t pg, void *ptr) __arm_strea
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.st1h.vert( [[TMP0]], ptr [[PTR]], i32 1, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svst1_ver_za16(uint32_t slice_base, svbool_t pg, void *ptr) __arm_streaming __arm_shared_za {
+void test_svst1_ver_za16(uint32_t slice_base, svbool_t pg, void *ptr) __arm_streaming __arm_in("za") {
   svst1_ver_za16(0, slice_base, pg, ptr);
   svst1_ver_za16(1, slice_base + 7, pg, ptr);
 }
@@ -179,7 +179,7 @@ void test_svst1_ver_za16(uint32_t slice_base, svbool_t pg, void *ptr) __arm_stre
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.st1w.vert( [[TMP0]], ptr [[PTR]], i32 3, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svst1_ver_za32(uint32_t slice_base, svbool_t pg, void *ptr) __arm_streaming __arm_shared_za {
+void test_svst1_ver_za32(uint32_t slice_base, svbool_t pg, void *ptr) __arm_streaming __arm_in("za") {
   svst1_ver_za32(0, slice_base, pg, ptr);
   svst1_ver_za32(3, slice_base + 3, pg, ptr);
 }
@@ -202,7 +202,7 @@ void test_svst1_ver_za32(uint32_t slice_base, svbool_t pg, void *ptr) __arm_stre
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.st1d.vert( [[TMP0]], ptr [[PTR]], i32 7, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svst1_ver_za64(uint32_t slice_base, svbool_t pg, void *ptr) __arm_streaming __arm_shared_za {
+void test_svst1_ver_za64(uint32_t slice_base, svbool_t pg, void *ptr) __arm_streaming __arm_in("za") {
   svst1_ver_za64(0, slice_base, pg, ptr);
   svst1_ver_za64(7, slice_base + 1, pg, ptr);
 }
@@ -223,7 +223,7 @@ void test_svst1_ver_za64(uint32_t slice_base, svbool_t pg, void *ptr) __arm_stre
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.st1q.vert( [[TMP0]], ptr [[PTR]], i32 15, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svst1_ver_za128(uint32_t slice_base, svbool_t pg, void *ptr) __arm_streaming __arm_shared_za {
+void test_svst1_ver_za128(uint32_t slice_base, svbool_t pg, void *ptr) __arm_streaming __arm_in("za") {
   svst1_ver_za128(0, slice_base, pg, ptr);
   svst1_ver_za128(15, slice_base, pg, ptr);
 }
diff --git a/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_st1_vnum.c b/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_st1_vnum.c
index 938e62a15c771..7566ad8889e05 100644
--- a/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_st1_vnum.c
+++ b/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_st1_vnum.c
@@ -28,7 +28,7 @@
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.st1b.horiz( [[PG]], ptr [[TMP1]], i32 0, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svst1_hor_vnum_za8(uint32_t slice_base, svbool_t pg, void *ptr, int64_t vnum) __arm_streaming __arm_shared_za {
+void test_svst1_hor_vnum_za8(uint32_t slice_base, svbool_t pg, void *ptr, int64_t vnum) __arm_streaming __arm_in("za") {
   svst1_hor_vnum_za8(0, slice_base, pg, ptr, vnum);
   svst1_hor_vnum_za8(0, slice_base + 15, pg, ptr, vnum);
 }
@@ -57,7 +57,7 @@ void test_svst1_hor_vnum_za8(uint32_t slice_base, svbool_t pg, void *ptr, int64_
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.st1h.horiz( [[TMP0]], ptr [[TMP2]], i32 1, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svst1_hor_vnum_za16(uint32_t slice_base, svbool_t pg, void *ptr, int64_t vnum) __arm_streaming __arm_shared_za {
+void test_svst1_hor_vnum_za16(uint32_t slice_base, svbool_t pg, void *ptr, int64_t vnum) __arm_streaming __arm_in("za") {
   svst1_hor_vnum_za16(0, slice_base, pg, ptr, vnum);
   svst1_hor_vnum_za16(1, slice_base + 7, pg, ptr, vnum);
 }
@@ -86,7 +86,7 @@ void test_svst1_hor_vnum_za16(uint32_t slice_base, svbool_t pg, void *ptr, int64
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.st1w.horiz( [[TMP0]], ptr [[TMP2]], i32 3, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svst1_hor_vnum_za32(uint32_t slice_base, svbool_t pg, void *ptr, int64_t vnum) __arm_streaming __arm_shared_za {
+void test_svst1_hor_vnum_za32(uint32_t slice_base, svbool_t pg, void *ptr, int64_t vnum) __arm_streaming __arm_in("za") {
   svst1_hor_vnum_za32(0, slice_base, pg, ptr, vnum);
   svst1_hor_vnum_za32(3, slice_base + 3, pg, ptr, vnum);
 }
@@ -115,7 +115,7 @@ void test_svst1_hor_vnum_za32(uint32_t slice_base, svbool_t pg, void *ptr, int64
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.st1d.horiz( [[TMP0]], ptr [[TMP2]], i32 7, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svst1_hor_vnum_za64(uint32_t slice_base, svbool_t pg, void *ptr, int64_t vnum) __arm_streaming __arm_shared_za {
+void test_svst1_hor_vnum_za64(uint32_t slice_base, svbool_t pg, void *ptr, int64_t vnum) __arm_streaming __arm_in("za") {
   svst1_hor_vnum_za64(0, slice_base, pg, ptr, vnum);
   svst1_hor_vnum_za64(7, slice_base + 1, pg, ptr, vnum);
 }
@@ -142,7 +142,7 @@ void test_svst1_hor_vnum_za64(uint32_t slice_base, svbool_t pg, void *ptr, int64
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.st1q.horiz( [[TMP0]], ptr [[TMP2]], i32 15, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svst1_hor_vnum_za128(uint32_t slice_base, svbool_t pg, void *ptr, int64_t vnum) __arm_streaming __arm_shared_za {
+void test_svst1_hor_vnum_za128(uint32_t slice_base, svbool_t pg, void *ptr, int64_t vnum) __arm_streaming __arm_in("za") {
   svst1_hor_vnum_za128(0, slice_base, pg, ptr, vnum);
   svst1_hor_vnum_za128(15, slice_base, pg, ptr, vnum);
 }
@@ -169,7 +169,7 @@ void test_svst1_hor_vnum_za128(uint32_t slice_base, svbool_t pg, void *ptr, int6
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.st1b.vert( [[PG]], ptr [[TMP1]], i32 0, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svst1_ver_vnum_za8(uint32_t slice_base, svbool_t pg, void *ptr, int64_t vnum) __arm_streaming __arm_shared_za {
+void test_svst1_ver_vnum_za8(uint32_t slice_base, svbool_t pg, void *ptr, int64_t vnum) __arm_streaming __arm_in("za") {
   svst1_ver_vnum_za8(0, slice_base, pg, ptr, vnum);
   svst1_ver_vnum_za8(0, slice_base + 15, pg, ptr, vnum);
 }
@@ -198,7 +198,7 @@ void test_svst1_ver_vnum_za8(uint32_t slice_base, svbool_t pg, void *ptr, int64_
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.st1h.vert( [[TMP0]], ptr [[TMP2]], i32 1, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svst1_ver_vnum_za16(uint32_t slice_base, svbool_t pg, void *ptr, int64_t vnum) __arm_streaming __arm_shared_za {
+void test_svst1_ver_vnum_za16(uint32_t slice_base, svbool_t pg, void *ptr, int64_t vnum) __arm_streaming __arm_in("za") {
   svst1_ver_vnum_za16(0, slice_base, pg, ptr, vnum);
   svst1_ver_vnum_za16(1, slice_base + 7, pg, ptr, vnum);
 }
@@ -227,7 +227,7 @@ void test_svst1_ver_vnum_za16(uint32_t slice_base, svbool_t pg, void *ptr, int64
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.st1w.vert( [[TMP0]], ptr [[TMP2]], i32 3, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svst1_ver_vnum_za32(uint32_t slice_base, svbool_t pg, void *ptr, int64_t vnum) __arm_streaming __arm_shared_za {
+void test_svst1_ver_vnum_za32(uint32_t slice_base, svbool_t pg, void *ptr, int64_t vnum) __arm_streaming __arm_in("za") {
   svst1_ver_vnum_za32(0, slice_base, pg, ptr, vnum);
   svst1_ver_vnum_za32(3, slice_base + 3, pg, ptr, vnum);
 }
@@ -256,7 +256,7 @@ void test_svst1_ver_vnum_za32(uint32_t slice_base, svbool_t pg, void *ptr, int64
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.st1d.vert( [[TMP0]], ptr [[TMP2]], i32 7, i32 [[ADD]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svst1_ver_vnum_za64(uint32_t slice_base, svbool_t pg, void *ptr, int64_t vnum) __arm_streaming __arm_shared_za {
+void test_svst1_ver_vnum_za64(uint32_t slice_base, svbool_t pg, void *ptr, int64_t vnum) __arm_streaming __arm_in("za") {
   svst1_ver_vnum_za64(0, slice_base, pg, ptr, vnum);
   svst1_ver_vnum_za64(7, slice_base + 1, pg, ptr, vnum);
 }
@@ -283,7 +283,7 @@ void test_svst1_ver_vnum_za64(uint32_t slice_base, svbool_t pg, void *ptr, int64
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.st1q.vert( [[TMP0]], ptr [[TMP2]], i32 15, i32 [[SLICE_BASE]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svst1_ver_vnum_za128(uint32_t slice_base, svbool_t pg, void *ptr, int64_t vnum) __arm_streaming __arm_shared_za {
+void test_svst1_ver_vnum_za128(uint32_t slice_base, svbool_t pg, void *ptr, int64_t vnum) __arm_streaming __arm_in("za") {
   svst1_ver_vnum_za128(0, slice_base, pg, ptr, vnum);
   svst1_ver_vnum_za128(15, slice_base, pg, ptr, vnum);
 }
diff --git a/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_state_funs.c b/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_state_funs.c
index 282819c8ca350..c3e4967bfe9b1 100644
--- a/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_state_funs.c
+++ b/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_state_funs.c
@@ -66,7 +66,7 @@ bool test_has_sme(void) __arm_streaming_compatible {
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svundef_za(void) __arm_streaming_compatible __arm_shared_za {
+void test_svundef_za(void) __arm_streaming_compatible __arm_out("za") {
   svundef_za();
 }
 
diff --git a/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_str.c b/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_str.c
index bcf368bc8dce4..d21c1ce7a8cd9 100644
--- a/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_str.c
+++ b/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_str.c
@@ -12,7 +12,7 @@
 // CHECK-NEXT:    tail call void @llvm.aarch64.sme.str(i32 [[SLICE_BASE:%.*]], ptr [[PTR:%.*]], i32 0)
 // CHECK-NEXT:    ret void
 //
-void test_svstr_vnum_za(uint32_t slice_base, void *ptr) __arm_shared_za {
+void test_svstr_vnum_za(uint32_t slice_base, void *ptr) __arm_in("za") {
   svstr_vnum_za(slice_base, ptr, 0);
 }
 
@@ -22,7 +22,7 @@ void test_svstr_vnum_za(uint32_t slice_base, void *ptr) __arm_shared_za {
 // CHECK-NEXT:    tail call void @llvm.aarch64.sme.str(i32 [[SLICE_BASE:%.*]], ptr [[PTR:%.*]], i32 15)
 // CHECK-NEXT:    ret void
 //
-void test_svstr_vnum_za_1(uint32_t slice_base, void *ptr) __arm_shared_za {
+void test_svstr_vnum_za_1(uint32_t slice_base, void *ptr) __arm_in("za") {
   svstr_vnum_za(slice_base, ptr, 15);
 }
 
@@ -32,7 +32,7 @@ void test_svstr_vnum_za_1(uint32_t slice_base, void *ptr) __arm_shared_za {
 // CHECK-NEXT:    tail call void @llvm.aarch64.sme.str(i32 [[SLICE_BASE:%.*]], ptr [[PTR:%.*]], i32 0)
 // CHECK-NEXT:    ret void
 //
-void test_svstr_za(uint32_t slice_base, void *ptr) __arm_shared_za {
+void test_svstr_za(uint32_t slice_base, void *ptr) __arm_in("za") {
   svstr_za(slice_base, ptr);
 }
 
@@ -43,7 +43,7 @@ void test_svstr_za(uint32_t slice_base, void *ptr) __arm_shared_za {
 // CHECK-NEXT:    tail call void @llvm.aarch64.sme.str(i32 [[SLICE_BASE:%.*]], ptr [[PTR:%.*]], i32 [[TMP0:%.*]])
 // CHECK-NEXT:    ret void
 //
-void test_svstr_vnum_za_var(uint32_t slice_base, void *ptr, int64_t vnum) __arm_shared_za {
+void test_svstr_vnum_za_var(uint32_t slice_base, void *ptr, int64_t vnum) __arm_in("za") {
   svstr_vnum_za(slice_base, ptr, vnum);
 }
 
@@ -53,6 +53,6 @@ void test_svstr_vnum_za_var(uint32_t slice_base, void *ptr, int64_t vnum) __arm_
 // CHECK-NEXT:    tail call void @llvm.aarch64.sme.str(i32 [[SLICE_BASE:%.*]], ptr [[PTR:%.*]], i32 16)
 // CHECK-NEXT:    ret void
 //
-void test_svstr_vnum_za_2(uint32_t slice_base, void *ptr) __arm_shared_za {
+void test_svstr_vnum_za_2(uint32_t slice_base, void *ptr) __arm_in("za") {
   svstr_vnum_za(slice_base, ptr, 16);
 }
diff --git a/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_write.c b/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_write.c
index 38c8402c3d0fa..5ddd90b95ce89 100644
--- a/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_write.c
+++ b/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_write.c
@@ -26,7 +26,7 @@
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.horiz.nxv16i8(i32 0, i32 [[SLICE_BASE]],  [[PG]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za8_s8(uint32_t slice_base, svbool_t pg, svint8_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za8_s8(uint32_t slice_base, svbool_t pg, svint8_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_hor_za8, _s8, _m)(0, slice_base, pg, zn);
 }
 
@@ -44,7 +44,7 @@ void test_svwrite_hor_za8_s8(uint32_t slice_base, svbool_t pg, svint8_t zn) __ar
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.horiz.nxv16i8(i32 0, i32 [[ADD]],  [[PG]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za8_s8_1(uint32_t slice_base, svbool_t pg, svint8_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za8_s8_1(uint32_t slice_base, svbool_t pg, svint8_t zn) __arm_streaming __arm_inout("za") {
    uint32_t slice = slice_base + 15;
   SME_ACLE_FUNC(svwrite_hor_za8, _s8, _m)(0, slice, pg, zn);
 }
@@ -63,7 +63,7 @@ void test_svwrite_hor_za8_s8_1(uint32_t slice_base, svbool_t pg, svint8_t zn) __
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.horiz.nxv8i16(i32 0, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za16_s16(uint32_t slice_base, svbool_t pg, svint16_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za16_s16(uint32_t slice_base, svbool_t pg, svint16_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_hor_za16, _s16, _m)(0, slice_base, pg, zn);
 }
 
@@ -83,7 +83,7 @@ void test_svwrite_hor_za16_s16(uint32_t slice_base, svbool_t pg, svint16_t zn) _
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.horiz.nxv8i16(i32 1, i32 [[ADD]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za16_s16_1(uint32_t slice_base, svbool_t pg, svint16_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za16_s16_1(uint32_t slice_base, svbool_t pg, svint16_t zn) __arm_streaming __arm_inout("za") {
   uint32_t slice = slice_base + 7;
   SME_ACLE_FUNC(svwrite_hor_za16, _s16, _m)(1, slice, pg, zn);
 }
@@ -102,7 +102,7 @@ void test_svwrite_hor_za16_s16_1(uint32_t slice_base, svbool_t pg, svint16_t zn)
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.horiz.nxv4i32(i32 0, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za32_s32(uint32_t slice_base, svbool_t pg, svint32_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za32_s32(uint32_t slice_base, svbool_t pg, svint32_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_hor_za32, _s32, _m)(0, slice_base, pg, zn);
 }
 
@@ -122,7 +122,7 @@ void test_svwrite_hor_za32_s32(uint32_t slice_base, svbool_t pg, svint32_t zn) _
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.horiz.nxv4i32(i32 3, i32 [[ADD]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za32_s32_1(uint32_t slice_base, svbool_t pg, svint32_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za32_s32_1(uint32_t slice_base, svbool_t pg, svint32_t zn) __arm_streaming __arm_inout("za") {
   uint32_t slice = slice_base + 3;
   SME_ACLE_FUNC(svwrite_hor_za32, _s32, _m)(3, slice, pg, zn);
 }
@@ -141,7 +141,7 @@ void test_svwrite_hor_za32_s32_1(uint32_t slice_base, svbool_t pg, svint32_t zn)
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.horiz.nxv2i64(i32 0, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za64_s64(uint32_t slice_base, svbool_t pg, svint64_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za64_s64(uint32_t slice_base, svbool_t pg, svint64_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_hor_za64, _s64, _m)(0, slice_base, pg, zn);
 }
 
@@ -161,7 +161,7 @@ void test_svwrite_hor_za64_s64(uint32_t slice_base, svbool_t pg, svint64_t zn) _
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.horiz.nxv2i64(i32 7, i32 [[ADD]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za64_s64_1(uint32_t slice_base, svbool_t pg, svint64_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za64_s64_1(uint32_t slice_base, svbool_t pg, svint64_t zn) __arm_streaming __arm_inout("za") {
   uint32_t slice = slice_base + 1;
   SME_ACLE_FUNC(svwrite_hor_za64, _s64, _m)(7, slice, pg, zn);
 }
@@ -178,7 +178,7 @@ void test_svwrite_hor_za64_s64_1(uint32_t slice_base, svbool_t pg, svint64_t zn)
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.horiz.nxv16i8(i32 0, i32 [[SLICE_BASE]],  [[PG]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za8_u8(uint32_t slice_base, svbool_t pg, svuint8_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za8_u8(uint32_t slice_base, svbool_t pg, svuint8_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_hor_za8, _u8, _m)(0, slice_base, pg, zn);
 }
 
@@ -196,7 +196,7 @@ void test_svwrite_hor_za8_u8(uint32_t slice_base, svbool_t pg, svuint8_t zn) __a
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.horiz.nxv16i8(i32 0, i32 [[ADD]],  [[PG]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za8_u8_1(uint32_t slice_base, svbool_t pg, svuint8_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za8_u8_1(uint32_t slice_base, svbool_t pg, svuint8_t zn) __arm_streaming __arm_inout("za") {
   uint32_t slice = slice_base + 15;
   SME_ACLE_FUNC(svwrite_hor_za8, _u8, _m)(0, slice, pg, zn);
 }
@@ -215,7 +215,7 @@ void test_svwrite_hor_za8_u8_1(uint32_t slice_base, svbool_t pg, svuint8_t zn) _
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.horiz.nxv8i16(i32 0, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za16_u16(uint32_t slice_base, svbool_t pg, svuint16_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za16_u16(uint32_t slice_base, svbool_t pg, svuint16_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_hor_za16, _u16, _m)(0, slice_base, pg, zn);
 }
 
@@ -235,7 +235,7 @@ void test_svwrite_hor_za16_u16(uint32_t slice_base, svbool_t pg, svuint16_t zn)
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.horiz.nxv8i16(i32 1, i32 [[ADD]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za16_u16_1(uint32_t slice_base, svbool_t pg, svuint16_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za16_u16_1(uint32_t slice_base, svbool_t pg, svuint16_t zn) __arm_streaming __arm_inout("za") {
   uint32_t slice = slice_base + 7;
   SME_ACLE_FUNC(svwrite_hor_za16, _u16, _m)(1, slice, pg, zn);
 }
@@ -254,7 +254,7 @@ void test_svwrite_hor_za16_u16_1(uint32_t slice_base, svbool_t pg, svuint16_t zn
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.horiz.nxv4i32(i32 0, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za32_u32(uint32_t slice_base, svbool_t pg, svuint32_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za32_u32(uint32_t slice_base, svbool_t pg, svuint32_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_hor_za32, _u32, _m)(0, slice_base, pg, zn);
 }
 
@@ -274,7 +274,7 @@ void test_svwrite_hor_za32_u32(uint32_t slice_base, svbool_t pg, svuint32_t zn)
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.horiz.nxv4i32(i32 3, i32 [[ADD]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za32_u32_1(uint32_t slice_base, svbool_t pg, svuint32_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za32_u32_1(uint32_t slice_base, svbool_t pg, svuint32_t zn) __arm_streaming __arm_inout("za") {
   uint32_t slice = slice_base + 3;
   SME_ACLE_FUNC(svwrite_hor_za32, _u32, _m)(3, slice, pg, zn);
 }
@@ -293,7 +293,7 @@ void test_svwrite_hor_za32_u32_1(uint32_t slice_base, svbool_t pg, svuint32_t zn
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.horiz.nxv2i64(i32 0, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za64_u64(uint32_t slice_base, svbool_t pg, svuint64_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za64_u64(uint32_t slice_base, svbool_t pg, svuint64_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_hor_za64, _u64, _m)(0, slice_base, pg, zn);
 }
 
@@ -313,7 +313,7 @@ void test_svwrite_hor_za64_u64(uint32_t slice_base, svbool_t pg, svuint64_t zn)
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.horiz.nxv2i64(i32 7, i32 [[ADD]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za64_u64_1(uint32_t slice_base, svbool_t pg, svuint64_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za64_u64_1(uint32_t slice_base, svbool_t pg, svuint64_t zn) __arm_streaming __arm_inout("za") {
   uint32_t slice = slice_base + 1;
   SME_ACLE_FUNC(svwrite_hor_za64, _u64, _m)(7, slice, pg, zn);
 }
@@ -332,7 +332,7 @@ void test_svwrite_hor_za64_u64_1(uint32_t slice_base, svbool_t pg, svuint64_t zn
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.horiz.nxv8f16(i32 0, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za16_f16(uint32_t slice_base, svbool_t pg, svfloat16_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za16_f16(uint32_t slice_base, svbool_t pg, svfloat16_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_hor_za16, _f16, _m)(0, slice_base, pg, zn);
 }
 
@@ -352,7 +352,7 @@ void test_svwrite_hor_za16_f16(uint32_t slice_base, svbool_t pg, svfloat16_t zn)
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.horiz.nxv8f16(i32 1, i32 [[ADD]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za16_f16_1(uint32_t slice_base, svbool_t pg, svfloat16_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za16_f16_1(uint32_t slice_base, svbool_t pg, svfloat16_t zn) __arm_streaming __arm_inout("za") {
   uint32_t slice = slice_base + 7;
   SME_ACLE_FUNC(svwrite_hor_za16, _f16, _m)(1, slice, pg, zn);
 }
@@ -371,7 +371,7 @@ void test_svwrite_hor_za16_f16_1(uint32_t slice_base, svbool_t pg, svfloat16_t z
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.horiz.nxv8bf16(i32 0, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za16_bf16(uint32_t slice_base, svbool_t pg, svbfloat16_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za16_bf16(uint32_t slice_base, svbool_t pg, svbfloat16_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_hor_za16, _bf16, _m)(0, slice_base, pg, zn);
 }
 
@@ -391,7 +391,7 @@ void test_svwrite_hor_za16_bf16(uint32_t slice_base, svbool_t pg, svbfloat16_t z
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.horiz.nxv8bf16(i32 1, i32 [[ADD]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za16_bf16_1(uint32_t slice_base, svbool_t pg, svbfloat16_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za16_bf16_1(uint32_t slice_base, svbool_t pg, svbfloat16_t zn) __arm_streaming __arm_inout("za") {
    uint32_t slice = slice_base + 7;
   SME_ACLE_FUNC(svwrite_hor_za16, _bf16, _m)(1, slice, pg, zn);
 }
@@ -410,7 +410,7 @@ void test_svwrite_hor_za16_bf16_1(uint32_t slice_base, svbool_t pg, svbfloat16_t
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.horiz.nxv4f32(i32 0, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za32_f32(uint32_t slice_base, svbool_t pg, svfloat32_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za32_f32(uint32_t slice_base, svbool_t pg, svfloat32_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_hor_za32, _f32, _m)(0, slice_base, pg, zn);
 }
 
@@ -430,7 +430,7 @@ void test_svwrite_hor_za32_f32(uint32_t slice_base, svbool_t pg, svfloat32_t zn)
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.horiz.nxv4f32(i32 3, i32 [[ADD]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za32_f32_1(uint32_t slice_base, svbool_t pg, svfloat32_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za32_f32_1(uint32_t slice_base, svbool_t pg, svfloat32_t zn) __arm_streaming __arm_inout("za") {
   uint32_t slice = slice_base + 3;
   SME_ACLE_FUNC(svwrite_hor_za32, _f32, _m)(3, slice, pg, zn);
 }
@@ -449,7 +449,7 @@ void test_svwrite_hor_za32_f32_1(uint32_t slice_base, svbool_t pg, svfloat32_t z
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.horiz.nxv2f64(i32 0, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za64_f64(uint32_t slice_base, svbool_t pg, svfloat64_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za64_f64(uint32_t slice_base, svbool_t pg, svfloat64_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_hor_za64, _f64, _m)(0, slice_base, pg, zn);
 }
 
@@ -469,7 +469,7 @@ void test_svwrite_hor_za64_f64(uint32_t slice_base, svbool_t pg, svfloat64_t zn)
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.horiz.nxv2f64(i32 7, i32 [[ADD]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za64_f64_1(uint32_t slice_base, svbool_t pg, svfloat64_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za64_f64_1(uint32_t slice_base, svbool_t pg, svfloat64_t zn) __arm_streaming __arm_inout("za") {
   uint32_t slice = slice_base + 1;
   SME_ACLE_FUNC(svwrite_hor_za64, _f64, _m)(7, slice, pg, zn);
 }
@@ -486,7 +486,7 @@ void test_svwrite_hor_za64_f64_1(uint32_t slice_base, svbool_t pg, svfloat64_t z
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.horiz.nxv16i8(i32 0, i32 [[SLICE_BASE]],  [[PG]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za128_s8(uint32_t slice_base, svbool_t pg, svint8_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za128_s8(uint32_t slice_base, svbool_t pg, svint8_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_hor_za128, _s8, _m)(0, slice_base, pg, zn);
 }
 
@@ -502,7 +502,7 @@ void test_svwrite_hor_za128_s8(uint32_t slice_base, svbool_t pg, svint8_t zn) __
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.horiz.nxv16i8(i32 15, i32 [[SLICE_BASE]],  [[PG]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za128_s8_1(uint32_t slice_base, svbool_t pg, svint8_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za128_s8_1(uint32_t slice_base, svbool_t pg, svint8_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_hor_za128, _s8, _m)(15, slice_base, pg, zn);
 }
 
@@ -520,7 +520,7 @@ void test_svwrite_hor_za128_s8_1(uint32_t slice_base, svbool_t pg, svint8_t zn)
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.horiz.nxv8i16(i32 0, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za128_s16(uint32_t slice_base, svbool_t pg, svint16_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za128_s16(uint32_t slice_base, svbool_t pg, svint16_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_hor_za128, _s16, _m)(0, slice_base, pg, zn);
 }
 
@@ -538,7 +538,7 @@ void test_svwrite_hor_za128_s16(uint32_t slice_base, svbool_t pg, svint16_t zn)
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.horiz.nxv8i16(i32 15, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za128_s16_1(uint32_t slice_base, svbool_t pg, svint16_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za128_s16_1(uint32_t slice_base, svbool_t pg, svint16_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_hor_za128, _s16, _m)(15, slice_base, pg, zn);
 }
 
@@ -556,7 +556,7 @@ void test_svwrite_hor_za128_s16_1(uint32_t slice_base, svbool_t pg, svint16_t zn
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.horiz.nxv4i32(i32 0, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za128_s32(uint32_t slice_base, svbool_t pg, svint32_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za128_s32(uint32_t slice_base, svbool_t pg, svint32_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_hor_za128, _s32, _m)(0, slice_base, pg, zn);
 }
 
@@ -574,7 +574,7 @@ void test_svwrite_hor_za128_s32(uint32_t slice_base, svbool_t pg, svint32_t zn)
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.horiz.nxv4i32(i32 15, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za128_s32_1(uint32_t slice_base, svbool_t pg, svint32_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za128_s32_1(uint32_t slice_base, svbool_t pg, svint32_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_hor_za128, _s32, _m)(15, slice_base, pg, zn);
 }
 
@@ -592,7 +592,7 @@ void test_svwrite_hor_za128_s32_1(uint32_t slice_base, svbool_t pg, svint32_t zn
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.horiz.nxv2i64(i32 0, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za128_s64(uint32_t slice_base, svbool_t pg, svint64_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za128_s64(uint32_t slice_base, svbool_t pg, svint64_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_hor_za128, _s64, _m)(0, slice_base, pg, zn);
 }
 
@@ -610,7 +610,7 @@ void test_svwrite_hor_za128_s64(uint32_t slice_base, svbool_t pg, svint64_t zn)
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.horiz.nxv2i64(i32 15, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za128_s64_1(uint32_t slice_base, svbool_t pg, svint64_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za128_s64_1(uint32_t slice_base, svbool_t pg, svint64_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_hor_za128, _s64, _m)(15, slice_base, pg, zn);
 }
 
@@ -626,7 +626,7 @@ void test_svwrite_hor_za128_s64_1(uint32_t slice_base, svbool_t pg, svint64_t zn
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.horiz.nxv16i8(i32 0, i32 [[SLICE_BASE]],  [[PG]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za128_u8(uint32_t slice_base, svbool_t pg, svuint8_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za128_u8(uint32_t slice_base, svbool_t pg, svuint8_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_hor_za128, _u8, _m)(0, slice_base, pg, zn);
 }
 
@@ -642,7 +642,7 @@ void test_svwrite_hor_za128_u8(uint32_t slice_base, svbool_t pg, svuint8_t zn) _
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.horiz.nxv16i8(i32 15, i32 [[SLICE_BASE]],  [[PG]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za128_u8_1(uint32_t slice_base, svbool_t pg, svuint8_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za128_u8_1(uint32_t slice_base, svbool_t pg, svuint8_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_hor_za128, _u8, _m)(15, slice_base, pg, zn);
 }
 
@@ -660,7 +660,7 @@ void test_svwrite_hor_za128_u8_1(uint32_t slice_base, svbool_t pg, svuint8_t zn)
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.horiz.nxv8i16(i32 0, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za128_u16(uint32_t slice_base, svbool_t pg, svuint16_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za128_u16(uint32_t slice_base, svbool_t pg, svuint16_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_hor_za128, _u16, _m)(0, slice_base, pg, zn);
 }
 
@@ -678,7 +678,7 @@ void test_svwrite_hor_za128_u16(uint32_t slice_base, svbool_t pg, svuint16_t zn)
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.horiz.nxv8i16(i32 15, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za128_u16_1(uint32_t slice_base, svbool_t pg, svuint16_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za128_u16_1(uint32_t slice_base, svbool_t pg, svuint16_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_hor_za128, _u16, _m)(15, slice_base, pg, zn);
 }
 
@@ -696,7 +696,7 @@ void test_svwrite_hor_za128_u16_1(uint32_t slice_base, svbool_t pg, svuint16_t z
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.horiz.nxv4i32(i32 0, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za128_u32(uint32_t slice_base, svbool_t pg, svuint32_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za128_u32(uint32_t slice_base, svbool_t pg, svuint32_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_hor_za128, _u32, _m)(0, slice_base, pg, zn);
 }
 
@@ -714,7 +714,7 @@ void test_svwrite_hor_za128_u32(uint32_t slice_base, svbool_t pg, svuint32_t zn)
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.horiz.nxv4i32(i32 15, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za128_u32_1(uint32_t slice_base, svbool_t pg, svuint32_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za128_u32_1(uint32_t slice_base, svbool_t pg, svuint32_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_hor_za128, _u32, _m)(15, slice_base, pg, zn);
 }
 
@@ -732,7 +732,7 @@ void test_svwrite_hor_za128_u32_1(uint32_t slice_base, svbool_t pg, svuint32_t z
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.horiz.nxv2i64(i32 0, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za128_u64(uint32_t slice_base, svbool_t pg, svuint64_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za128_u64(uint32_t slice_base, svbool_t pg, svuint64_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_hor_za128, _u64, _m)(0, slice_base, pg, zn);
 }
 
@@ -750,7 +750,7 @@ void test_svwrite_hor_za128_u64(uint32_t slice_base, svbool_t pg, svuint64_t zn)
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.horiz.nxv2i64(i32 15, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za128_u64_1(uint32_t slice_base, svbool_t pg, svuint64_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za128_u64_1(uint32_t slice_base, svbool_t pg, svuint64_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_hor_za128, _u64, _m)(15, slice_base, pg, zn);
 }
 
@@ -768,7 +768,7 @@ void test_svwrite_hor_za128_u64_1(uint32_t slice_base, svbool_t pg, svuint64_t z
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.horiz.nxv8f16(i32 0, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za128_f16(uint32_t slice_base, svbool_t pg, svfloat16_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za128_f16(uint32_t slice_base, svbool_t pg, svfloat16_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_hor_za128, _f16, _m)(0, slice_base, pg, zn);
 }
 
@@ -786,7 +786,7 @@ void test_svwrite_hor_za128_f16(uint32_t slice_base, svbool_t pg, svfloat16_t zn
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.horiz.nxv8f16(i32 15, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za128_f16_1(uint32_t slice_base, svbool_t pg, svfloat16_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za128_f16_1(uint32_t slice_base, svbool_t pg, svfloat16_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_hor_za128, _f16, _m)(15, slice_base, pg, zn);
 }
 
@@ -804,7 +804,7 @@ void test_svwrite_hor_za128_f16_1(uint32_t slice_base, svbool_t pg, svfloat16_t
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.horiz.nxv8bf16(i32 0, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za128_bf16(uint32_t slice_base, svbool_t pg, svbfloat16_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za128_bf16(uint32_t slice_base, svbool_t pg, svbfloat16_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_hor_za128, _bf16, _m)(0, slice_base, pg, zn);
 }
 
@@ -822,7 +822,7 @@ void test_svwrite_hor_za128_bf16(uint32_t slice_base, svbool_t pg, svbfloat16_t
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.horiz.nxv8bf16(i32 15, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za128_bf16_1(uint32_t slice_base, svbool_t pg, svbfloat16_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za128_bf16_1(uint32_t slice_base, svbool_t pg, svbfloat16_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_hor_za128, _bf16, _m)(15, slice_base, pg, zn);
 }
 
@@ -840,7 +840,7 @@ void test_svwrite_hor_za128_bf16_1(uint32_t slice_base, svbool_t pg, svbfloat16_
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.horiz.nxv4f32(i32 0, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za128_f32(uint32_t slice_base, svbool_t pg, svfloat32_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za128_f32(uint32_t slice_base, svbool_t pg, svfloat32_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_hor_za128, _f32, _m)(0, slice_base, pg, zn);
 }
 
@@ -858,7 +858,7 @@ void test_svwrite_hor_za128_f32(uint32_t slice_base, svbool_t pg, svfloat32_t zn
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.horiz.nxv4f32(i32 15, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za128_f32_1(uint32_t slice_base, svbool_t pg, svfloat32_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za128_f32_1(uint32_t slice_base, svbool_t pg, svfloat32_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_hor_za128, _f32, _m)(15, slice_base, pg, zn);
 }
 
@@ -876,7 +876,7 @@ void test_svwrite_hor_za128_f32_1(uint32_t slice_base, svbool_t pg, svfloat32_t
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.horiz.nxv2f64(i32 0, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za128_f64(uint32_t slice_base, svbool_t pg, svfloat64_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za128_f64(uint32_t slice_base, svbool_t pg, svfloat64_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_hor_za128, _f64, _m)(0, slice_base, pg, zn);
 }
 
@@ -894,7 +894,7 @@ void test_svwrite_hor_za128_f64(uint32_t slice_base, svbool_t pg, svfloat64_t zn
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.horiz.nxv2f64(i32 15, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_hor_za128_f64_1(uint32_t slice_base, svbool_t pg, svfloat64_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za128_f64_1(uint32_t slice_base, svbool_t pg, svfloat64_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_hor_za128, _f64, _m)(15, slice_base, pg, zn);
 }
 
@@ -910,7 +910,7 @@ void test_svwrite_hor_za128_f64_1(uint32_t slice_base, svbool_t pg, svfloat64_t
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.vert.nxv16i8(i32 0, i32 [[SLICE_BASE]],  [[PG]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za8_s8(uint32_t slice_base, svbool_t pg, svint8_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za8_s8(uint32_t slice_base, svbool_t pg, svint8_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_ver_za8, _s8, _m)(0, slice_base, pg, zn);
 }
 
@@ -928,7 +928,7 @@ void test_svwrite_ver_za8_s8(uint32_t slice_base, svbool_t pg, svint8_t zn) __ar
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.vert.nxv16i8(i32 0, i32 [[ADD]],  [[PG]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za8_s8_1(uint32_t slice_base, svbool_t pg, svint8_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za8_s8_1(uint32_t slice_base, svbool_t pg, svint8_t zn) __arm_streaming __arm_inout("za") {
   uint32_t slice = slice_base + 15;
   SME_ACLE_FUNC(svwrite_ver_za8, _s8, _m)(0, slice, pg, zn);
 }
@@ -947,7 +947,7 @@ void test_svwrite_ver_za8_s8_1(uint32_t slice_base, svbool_t pg, svint8_t zn) __
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.vert.nxv8i16(i32 0, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za16_s16(uint32_t slice_base, svbool_t pg, svint16_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za16_s16(uint32_t slice_base, svbool_t pg, svint16_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_ver_za16, _s16, _m)(0, slice_base, pg, zn);
 }
 
@@ -967,7 +967,7 @@ void test_svwrite_ver_za16_s16(uint32_t slice_base, svbool_t pg, svint16_t zn) _
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.vert.nxv8i16(i32 1, i32 [[ADD]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za16_s16_1(uint32_t slice_base, svbool_t pg, svint16_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za16_s16_1(uint32_t slice_base, svbool_t pg, svint16_t zn) __arm_streaming __arm_inout("za") {
   uint32_t slice = slice_base + 7;
   SME_ACLE_FUNC(svwrite_ver_za16, _s16, _m)(1, slice, pg, zn);
 }
@@ -986,7 +986,7 @@ void test_svwrite_ver_za16_s16_1(uint32_t slice_base, svbool_t pg, svint16_t zn)
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.vert.nxv4i32(i32 0, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za32_s32(uint32_t slice_base, svbool_t pg, svint32_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za32_s32(uint32_t slice_base, svbool_t pg, svint32_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_ver_za32, _s32, _m)(0, slice_base, pg, zn);
 }
 
@@ -1006,7 +1006,7 @@ void test_svwrite_ver_za32_s32(uint32_t slice_base, svbool_t pg, svint32_t zn) _
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.vert.nxv4i32(i32 3, i32 [[ADD]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za32_s32_1(uint32_t slice_base, svbool_t pg, svint32_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za32_s32_1(uint32_t slice_base, svbool_t pg, svint32_t zn) __arm_streaming __arm_inout("za") {
   uint32_t slice = slice_base + 3;
   SME_ACLE_FUNC(svwrite_ver_za32, _s32, _m)(3, slice, pg, zn);
 }
@@ -1025,7 +1025,7 @@ void test_svwrite_ver_za32_s32_1(uint32_t slice_base, svbool_t pg, svint32_t zn)
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.vert.nxv2i64(i32 0, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za64_s64(uint32_t slice_base, svbool_t pg, svint64_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za64_s64(uint32_t slice_base, svbool_t pg, svint64_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_ver_za64, _s64, _m)(0, slice_base, pg, zn);
 }
 
@@ -1045,7 +1045,7 @@ void test_svwrite_ver_za64_s64(uint32_t slice_base, svbool_t pg, svint64_t zn) _
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.vert.nxv2i64(i32 7, i32 [[ADD]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za64_s64_1(uint32_t slice_base, svbool_t pg, svint64_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za64_s64_1(uint32_t slice_base, svbool_t pg, svint64_t zn) __arm_streaming __arm_inout("za") {
   uint32_t slice = slice_base + 1;
   SME_ACLE_FUNC(svwrite_ver_za64, _s64, _m)(7, slice, pg, zn);
 }
@@ -1062,7 +1062,7 @@ void test_svwrite_ver_za64_s64_1(uint32_t slice_base, svbool_t pg, svint64_t zn)
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.vert.nxv16i8(i32 0, i32 [[SLICE_BASE]],  [[PG]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za8_u8(uint32_t slice_base, svbool_t pg, svuint8_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za8_u8(uint32_t slice_base, svbool_t pg, svuint8_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_ver_za8, _u8, _m)(0, slice_base, pg, zn);
 }
 
@@ -1080,7 +1080,7 @@ void test_svwrite_ver_za8_u8(uint32_t slice_base, svbool_t pg, svuint8_t zn) __a
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.vert.nxv16i8(i32 0, i32 [[ADD]],  [[PG]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za8_u8_1(uint32_t slice_base, svbool_t pg, svuint8_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za8_u8_1(uint32_t slice_base, svbool_t pg, svuint8_t zn) __arm_streaming __arm_inout("za") {
   uint32_t slice = slice_base + 15;
   SME_ACLE_FUNC(svwrite_ver_za8, _u8, _m)(0, slice, pg, zn);
 }
@@ -1099,7 +1099,7 @@ void test_svwrite_ver_za8_u8_1(uint32_t slice_base, svbool_t pg, svuint8_t zn) _
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.vert.nxv8i16(i32 0, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za16_u16(uint32_t slice_base, svbool_t pg, svuint16_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za16_u16(uint32_t slice_base, svbool_t pg, svuint16_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_ver_za16, _u16, _m)(0, slice_base, pg, zn);
 }
 
@@ -1119,7 +1119,7 @@ void test_svwrite_ver_za16_u16(uint32_t slice_base, svbool_t pg, svuint16_t zn)
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.vert.nxv8i16(i32 1, i32 [[ADD]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za16_u16_1(uint32_t slice_base, svbool_t pg, svuint16_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za16_u16_1(uint32_t slice_base, svbool_t pg, svuint16_t zn) __arm_streaming __arm_inout("za") {
   uint32_t slice = slice_base + 7;
   SME_ACLE_FUNC(svwrite_ver_za16, _u16, _m)(1, slice, pg, zn);
 }
@@ -1138,7 +1138,7 @@ void test_svwrite_ver_za16_u16_1(uint32_t slice_base, svbool_t pg, svuint16_t zn
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.vert.nxv4i32(i32 0, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za32_u32(uint32_t slice_base, svbool_t pg, svuint32_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za32_u32(uint32_t slice_base, svbool_t pg, svuint32_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_ver_za32, _u32, _m)(0, slice_base, pg, zn);
 }
 
@@ -1158,7 +1158,7 @@ void test_svwrite_ver_za32_u32(uint32_t slice_base, svbool_t pg, svuint32_t zn)
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.vert.nxv4i32(i32 3, i32 [[ADD]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za32_u32_1(uint32_t slice_base, svbool_t pg, svuint32_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za32_u32_1(uint32_t slice_base, svbool_t pg, svuint32_t zn) __arm_streaming __arm_inout("za") {
   uint32_t slice = slice_base + 3;
   SME_ACLE_FUNC(svwrite_ver_za32, _u32, _m)(3, slice, pg, zn);
 }
@@ -1177,7 +1177,7 @@ void test_svwrite_ver_za32_u32_1(uint32_t slice_base, svbool_t pg, svuint32_t zn
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.vert.nxv2i64(i32 0, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za64_u64(uint32_t slice_base, svbool_t pg, svuint64_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za64_u64(uint32_t slice_base, svbool_t pg, svuint64_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_ver_za64, _u64, _m)(0, slice_base, pg, zn);
 }
 
@@ -1197,7 +1197,7 @@ void test_svwrite_ver_za64_u64(uint32_t slice_base, svbool_t pg, svuint64_t zn)
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.vert.nxv2i64(i32 7, i32 [[ADD]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za64_u64_1(uint32_t slice_base, svbool_t pg, svuint64_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za64_u64_1(uint32_t slice_base, svbool_t pg, svuint64_t zn) __arm_streaming __arm_inout("za") {
   uint32_t slice = slice_base + 1;
   SME_ACLE_FUNC(svwrite_ver_za64, _u64, _m)(7, slice, pg, zn);
 }
@@ -1216,7 +1216,7 @@ void test_svwrite_ver_za64_u64_1(uint32_t slice_base, svbool_t pg, svuint64_t zn
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.vert.nxv8f16(i32 0, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za16_f16(uint32_t slice_base, svbool_t pg, svfloat16_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za16_f16(uint32_t slice_base, svbool_t pg, svfloat16_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_ver_za16, _f16, _m)(0, slice_base, pg, zn);
 }
 
@@ -1236,7 +1236,7 @@ void test_svwrite_ver_za16_f16(uint32_t slice_base, svbool_t pg, svfloat16_t zn)
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.vert.nxv8f16(i32 1, i32 [[ADD]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za16_f16_1(uint32_t slice_base, svbool_t pg, svfloat16_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za16_f16_1(uint32_t slice_base, svbool_t pg, svfloat16_t zn) __arm_streaming __arm_inout("za") {
   uint32_t slice = slice_base + 7;
   SME_ACLE_FUNC(svwrite_ver_za16, _f16, _m)(1, slice, pg, zn);
 }
@@ -1255,7 +1255,7 @@ void test_svwrite_ver_za16_f16_1(uint32_t slice_base, svbool_t pg, svfloat16_t z
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.vert.nxv8bf16(i32 0, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za16_bf16(uint32_t slice_base, svbool_t pg, svbfloat16_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za16_bf16(uint32_t slice_base, svbool_t pg, svbfloat16_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_ver_za16, _bf16, _m)(0, slice_base, pg, zn);
 }
 
@@ -1275,7 +1275,7 @@ void test_svwrite_ver_za16_bf16(uint32_t slice_base, svbool_t pg, svbfloat16_t z
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.vert.nxv8bf16(i32 1, i32 [[ADD]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za16_bf16_1(uint32_t slice_base, svbool_t pg, svbfloat16_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za16_bf16_1(uint32_t slice_base, svbool_t pg, svbfloat16_t zn) __arm_streaming __arm_inout("za") {
   uint32_t slice = slice_base + 7;
   SME_ACLE_FUNC(svwrite_ver_za16, _bf16, _m)(1, slice, pg, zn);
 }
@@ -1294,7 +1294,7 @@ void test_svwrite_ver_za16_bf16_1(uint32_t slice_base, svbool_t pg, svbfloat16_t
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.vert.nxv4f32(i32 0, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za32_f32(uint32_t slice_base, svbool_t pg, svfloat32_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za32_f32(uint32_t slice_base, svbool_t pg, svfloat32_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_ver_za32, _f32, _m)(0, slice_base, pg, zn);
 }
 
@@ -1314,7 +1314,7 @@ void test_svwrite_ver_za32_f32(uint32_t slice_base, svbool_t pg, svfloat32_t zn)
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.vert.nxv4f32(i32 3, i32 [[ADD]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za32_f32_1(uint32_t slice_base, svbool_t pg, svfloat32_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za32_f32_1(uint32_t slice_base, svbool_t pg, svfloat32_t zn) __arm_streaming __arm_inout("za") {
   uint32_t slice = slice_base + 3;
   SME_ACLE_FUNC(svwrite_ver_za32, _f32, _m)(3, slice, pg, zn);
 }
@@ -1333,7 +1333,7 @@ void test_svwrite_ver_za32_f32_1(uint32_t slice_base, svbool_t pg, svfloat32_t z
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.vert.nxv2f64(i32 0, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za64_f64(uint32_t slice_base, svbool_t pg, svfloat64_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za64_f64(uint32_t slice_base, svbool_t pg, svfloat64_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_ver_za64, _f64, _m)(0, slice_base, pg, zn);
 }
 
@@ -1353,7 +1353,7 @@ void test_svwrite_ver_za64_f64(uint32_t slice_base, svbool_t pg, svfloat64_t zn)
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.write.vert.nxv2f64(i32 7, i32 [[ADD]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za64_f64_1(uint32_t slice_base, svbool_t pg, svfloat64_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za64_f64_1(uint32_t slice_base, svbool_t pg, svfloat64_t zn) __arm_streaming __arm_inout("za") {
   uint32_t slice = slice_base + 1;
   SME_ACLE_FUNC(svwrite_ver_za64, _f64, _m)(7, slice, pg, zn);
 }
@@ -1370,7 +1370,7 @@ void test_svwrite_ver_za64_f64_1(uint32_t slice_base, svbool_t pg, svfloat64_t z
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.vert.nxv16i8(i32 0, i32 [[SLICE_BASE]],  [[PG]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za128_s8(uint32_t slice_base, svbool_t pg, svint8_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za128_s8(uint32_t slice_base, svbool_t pg, svint8_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_ver_za128, _s8, _m)(0, slice_base, pg, zn);
 }
 
@@ -1386,7 +1386,7 @@ void test_svwrite_ver_za128_s8(uint32_t slice_base, svbool_t pg, svint8_t zn) __
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.vert.nxv16i8(i32 15, i32 [[SLICE_BASE]],  [[PG]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za128_s8_1(uint32_t slice_base, svbool_t pg, svint8_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za128_s8_1(uint32_t slice_base, svbool_t pg, svint8_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_ver_za128, _s8, _m)(15, slice_base, pg, zn);
 }
 
@@ -1404,7 +1404,7 @@ void test_svwrite_ver_za128_s8_1(uint32_t slice_base, svbool_t pg, svint8_t zn)
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.vert.nxv8i16(i32 0, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za128_s16(uint32_t slice_base, svbool_t pg, svint16_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za128_s16(uint32_t slice_base, svbool_t pg, svint16_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_ver_za128, _s16, _m)(0, slice_base, pg, zn);
 }
 
@@ -1422,7 +1422,7 @@ void test_svwrite_ver_za128_s16(uint32_t slice_base, svbool_t pg, svint16_t zn)
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.vert.nxv8i16(i32 15, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za128_s16_1(uint32_t slice_base, svbool_t pg, svint16_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za128_s16_1(uint32_t slice_base, svbool_t pg, svint16_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_ver_za128, _s16, _m)(15, slice_base, pg, zn);
 }
 
@@ -1440,7 +1440,7 @@ void test_svwrite_ver_za128_s16_1(uint32_t slice_base, svbool_t pg, svint16_t zn
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.vert.nxv4i32(i32 0, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za128_s32(uint32_t slice_base, svbool_t pg, svint32_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za128_s32(uint32_t slice_base, svbool_t pg, svint32_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_ver_za128, _s32, _m)(0, slice_base, pg, zn);
 }
 
@@ -1458,7 +1458,7 @@ void test_svwrite_ver_za128_s32(uint32_t slice_base, svbool_t pg, svint32_t zn)
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.vert.nxv4i32(i32 15, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za128_s32_1(uint32_t slice_base, svbool_t pg, svint32_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za128_s32_1(uint32_t slice_base, svbool_t pg, svint32_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_ver_za128, _s32, _m)(15, slice_base, pg, zn);
 }
 
@@ -1476,7 +1476,7 @@ void test_svwrite_ver_za128_s32_1(uint32_t slice_base, svbool_t pg, svint32_t zn
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.vert.nxv2i64(i32 0, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za128_s64(uint32_t slice_base, svbool_t pg, svint64_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za128_s64(uint32_t slice_base, svbool_t pg, svint64_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_ver_za128, _s64, _m)(0, slice_base, pg, zn);
 }
 
@@ -1494,7 +1494,7 @@ void test_svwrite_ver_za128_s64(uint32_t slice_base, svbool_t pg, svint64_t zn)
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.vert.nxv2i64(i32 15, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za128_s64_1(uint32_t slice_base, svbool_t pg, svint64_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za128_s64_1(uint32_t slice_base, svbool_t pg, svint64_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_ver_za128, _s64, _m)(15, slice_base, pg, zn);
 }
 
@@ -1510,7 +1510,7 @@ void test_svwrite_ver_za128_s64_1(uint32_t slice_base, svbool_t pg, svint64_t zn
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.vert.nxv16i8(i32 0, i32 [[SLICE_BASE]],  [[PG]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za128_u8(uint32_t slice_base, svbool_t pg, svuint8_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za128_u8(uint32_t slice_base, svbool_t pg, svuint8_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_ver_za128, _u8, _m)(0, slice_base, pg, zn);
 }
 
@@ -1526,7 +1526,7 @@ void test_svwrite_ver_za128_u8(uint32_t slice_base, svbool_t pg, svuint8_t zn) _
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.vert.nxv16i8(i32 15, i32 [[SLICE_BASE]],  [[PG]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za128_u8_1(uint32_t slice_base, svbool_t pg, svuint8_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za128_u8_1(uint32_t slice_base, svbool_t pg, svuint8_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_ver_za128, _u8, _m)(15, slice_base, pg, zn);
 }
 
@@ -1544,7 +1544,7 @@ void test_svwrite_ver_za128_u8_1(uint32_t slice_base, svbool_t pg, svuint8_t zn)
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.vert.nxv8i16(i32 0, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za128_u16(uint32_t slice_base, svbool_t pg, svuint16_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za128_u16(uint32_t slice_base, svbool_t pg, svuint16_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_ver_za128, _u16, _m)(0, slice_base, pg, zn);
 }
 
@@ -1562,7 +1562,7 @@ void test_svwrite_ver_za128_u16(uint32_t slice_base, svbool_t pg, svuint16_t zn)
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.vert.nxv8i16(i32 15, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za128_u16_1(uint32_t slice_base, svbool_t pg, svuint16_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za128_u16_1(uint32_t slice_base, svbool_t pg, svuint16_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_ver_za128, _u16, _m)(15, slice_base, pg, zn);
 }
 
@@ -1580,7 +1580,7 @@ void test_svwrite_ver_za128_u16_1(uint32_t slice_base, svbool_t pg, svuint16_t z
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.vert.nxv4i32(i32 0, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za128_u32(uint32_t slice_base, svbool_t pg, svuint32_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za128_u32(uint32_t slice_base, svbool_t pg, svuint32_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_ver_za128, _u32, _m)(0, slice_base, pg, zn);
 }
 
@@ -1598,7 +1598,7 @@ void test_svwrite_ver_za128_u32(uint32_t slice_base, svbool_t pg, svuint32_t zn)
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.vert.nxv4i32(i32 15, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za128_u32_1(uint32_t slice_base, svbool_t pg, svuint32_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za128_u32_1(uint32_t slice_base, svbool_t pg, svuint32_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_ver_za128, _u32, _m)(15, slice_base, pg, zn);
 }
 
@@ -1616,7 +1616,7 @@ void test_svwrite_ver_za128_u32_1(uint32_t slice_base, svbool_t pg, svuint32_t z
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.vert.nxv2i64(i32 0, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za128_u64(uint32_t slice_base, svbool_t pg, svuint64_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za128_u64(uint32_t slice_base, svbool_t pg, svuint64_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_ver_za128, _u64, _m)(0, slice_base, pg, zn);
 }
 
@@ -1634,7 +1634,7 @@ void test_svwrite_ver_za128_u64(uint32_t slice_base, svbool_t pg, svuint64_t zn)
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.vert.nxv2i64(i32 15, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za128_u64_1(uint32_t slice_base, svbool_t pg, svuint64_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za128_u64_1(uint32_t slice_base, svbool_t pg, svuint64_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_ver_za128, _u64, _m)(15, slice_base, pg, zn);
 }
 
@@ -1652,7 +1652,7 @@ void test_svwrite_ver_za128_u64_1(uint32_t slice_base, svbool_t pg, svuint64_t z
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.vert.nxv8f16(i32 0, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za128_f16(uint32_t slice_base, svbool_t pg, svfloat16_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za128_f16(uint32_t slice_base, svbool_t pg, svfloat16_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_ver_za128, _f16, _m)(0, slice_base, pg, zn);
 }
 
@@ -1670,7 +1670,7 @@ void test_svwrite_ver_za128_f16(uint32_t slice_base, svbool_t pg, svfloat16_t zn
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.vert.nxv8f16(i32 15, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za128_f16_1(uint32_t slice_base, svbool_t pg, svfloat16_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za128_f16_1(uint32_t slice_base, svbool_t pg, svfloat16_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_ver_za128, _f16, _m)(15, slice_base, pg, zn);
 }
 
@@ -1688,7 +1688,7 @@ void test_svwrite_ver_za128_f16_1(uint32_t slice_base, svbool_t pg, svfloat16_t
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.vert.nxv8bf16(i32 0, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za128_bf16(uint32_t slice_base, svbool_t pg, svbfloat16_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za128_bf16(uint32_t slice_base, svbool_t pg, svbfloat16_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_ver_za128, _bf16, _m)(0, slice_base, pg, zn);
 }
 
@@ -1706,7 +1706,7 @@ void test_svwrite_ver_za128_bf16(uint32_t slice_base, svbool_t pg, svbfloat16_t
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.vert.nxv8bf16(i32 15, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za128_bf16_1(uint32_t slice_base, svbool_t pg, svbfloat16_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za128_bf16_1(uint32_t slice_base, svbool_t pg, svbfloat16_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_ver_za128, _bf16, _m)(15, slice_base, pg, zn);
 }
 
@@ -1724,7 +1724,7 @@ void test_svwrite_ver_za128_bf16_1(uint32_t slice_base, svbool_t pg, svbfloat16_
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.vert.nxv4f32(i32 0, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za128_f32(uint32_t slice_base, svbool_t pg, svfloat32_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za128_f32(uint32_t slice_base, svbool_t pg, svfloat32_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_ver_za128, _f32, _m)(0, slice_base, pg, zn);
 }
 
@@ -1742,7 +1742,7 @@ void test_svwrite_ver_za128_f32(uint32_t slice_base, svbool_t pg, svfloat32_t zn
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.vert.nxv4f32(i32 15, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za128_f32_1(uint32_t slice_base, svbool_t pg, svfloat32_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za128_f32_1(uint32_t slice_base, svbool_t pg, svfloat32_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_ver_za128, _f32, _m)(15, slice_base, pg, zn);
 }
 
@@ -1760,7 +1760,7 @@ void test_svwrite_ver_za128_f32_1(uint32_t slice_base, svbool_t pg, svfloat32_t
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.vert.nxv2f64(i32 0, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za128_f64(uint32_t slice_base, svbool_t pg, svfloat64_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za128_f64(uint32_t slice_base, svbool_t pg, svfloat64_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_ver_za128, _f64, _m)(0, slice_base, pg, zn);
 }
 
@@ -1778,7 +1778,7 @@ void test_svwrite_ver_za128_f64(uint32_t slice_base, svbool_t pg, svfloat64_t zn
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.writeq.vert.nxv2f64(i32 15, i32 [[SLICE_BASE]],  [[TMP0]],  [[ZN]])
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svwrite_ver_za128_f64_1(uint32_t slice_base, svbool_t pg, svfloat64_t zn) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za128_f64_1(uint32_t slice_base, svbool_t pg, svfloat64_t zn) __arm_streaming __arm_inout("za") {
   SME_ACLE_FUNC(svwrite_ver_za128, _f64, _m)(15, slice_base, pg, zn);
 }
 //// NOTE: These prefixes are unused and the list is autogenerated. Do not add tests below this line:
diff --git a/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_zero.c b/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_zero.c
index ddd9602369538..658092bc33ff9 100644
--- a/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_zero.c
+++ b/clang/test/CodeGen/aarch64-sme-intrinsics/acle_sme_zero.c
@@ -18,7 +18,7 @@
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.zero(i32 0)
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svzero_mask_za(void) __arm_shared_za {
+void test_svzero_mask_za(void) __arm_inout("za") {
   svzero_mask_za(0);
 }
 
@@ -34,7 +34,7 @@ void test_svzero_mask_za(void) __arm_shared_za {
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.zero(i32 176)
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svzero_mask_za_1(void) __arm_shared_za {
+void test_svzero_mask_za_1(void) __arm_inout("za") {
   svzero_mask_za(176);
 }
 
@@ -50,7 +50,7 @@ void test_svzero_mask_za_1(void) __arm_shared_za {
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.zero(i32 255)
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svzero_mask_za_2(void) __arm_shared_za {
+void test_svzero_mask_za_2(void) __arm_inout("za") {
   svzero_mask_za(255);
 }
 
@@ -66,7 +66,7 @@ void test_svzero_mask_za_2(void) __arm_shared_za {
 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.zero(i32 255)
 // CHECK-CXX-NEXT:    ret void
 //
-void test_svzero_za(void) __arm_shared_za {
+void test_svzero_za(void) __arm_out("za") {
   svzero_za();
 }
 //// NOTE: These prefixes are unused and the list is autogenerated. Do not add tests below this line:
diff --git a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_add.c b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_add.c
index 46b4a0386502e..311f965786137 100644
--- a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_add.c
+++ b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_add.c
@@ -36,7 +36,7 @@
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.add.write.single.za.vg1x2.nxv4i32(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svadd_write_single2_s32(uint32_t slice_base, svint32x2_t zn, svint32_t zm) __arm_streaming __arm_shared_za {
+void test_svadd_write_single2_s32(uint32_t slice_base, svint32x2_t zn, svint32_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svadd_write,_single,_za32,_s32,_vg1x2)(slice_base, zn, zm);
 }
 
@@ -54,7 +54,7 @@ void test_svadd_write_single2_s32(uint32_t slice_base, svint32x2_t zn, svint32_t
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.add.write.single.za.vg1x2.nxv4i32(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svadd_write_single2_u32(uint32_t slice_base, svuint32x2_t zn, svuint32_t zm) __arm_streaming __arm_shared_za {
+void test_svadd_write_single2_u32(uint32_t slice_base, svuint32x2_t zn, svuint32_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svadd_write,_single,_za32,_u32,_vg1x2)(slice_base, zn, zm);
 }
 
@@ -72,7 +72,7 @@ void test_svadd_write_single2_u32(uint32_t slice_base, svuint32x2_t zn, svuint32
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.add.write.single.za.vg1x2.nxv2i64(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svadd_write_single2_s64(uint32_t slice_base, svint64x2_t zn, svint64_t zm) __arm_streaming __arm_shared_za {
+void test_svadd_write_single2_s64(uint32_t slice_base, svint64x2_t zn, svint64_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svadd_write,_single,_za64,_s64,_vg1x2)(slice_base, zn, zm);
 }
 
@@ -90,7 +90,7 @@ void test_svadd_write_single2_s64(uint32_t slice_base, svint64x2_t zn, svint64_t
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.add.write.single.za.vg1x2.nxv2i64(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svadd_write_single2_u64(uint32_t slice_base, svuint64x2_t zn, svuint64_t zm) __arm_streaming __arm_shared_za {
+void test_svadd_write_single2_u64(uint32_t slice_base, svuint64x2_t zn, svuint64_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svadd_write,_single,_za64,_u64,_vg1x2)(slice_base, zn, zm);
 }
 
@@ -114,7 +114,7 @@ void test_svadd_write_single2_u64(uint32_t slice_base, svuint64x2_t zn, svuint64
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.add.write.single.za.vg1x4.nxv4i32(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svadd_write_single4_s32(uint32_t slice_base, svint32x4_t zn, svint32_t zm) __arm_streaming __arm_shared_za {
+void test_svadd_write_single4_s32(uint32_t slice_base, svint32x4_t zn, svint32_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svadd_write,_single,_za32,_s32,_vg1x4)(slice_base, zn, zm);
 }
 
@@ -136,7 +136,7 @@ void test_svadd_write_single4_s32(uint32_t slice_base, svint32x4_t zn, svint32_t
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.add.write.single.za.vg1x4.nxv4i32(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svadd_write_single4_u32(uint32_t slice_base, svuint32x4_t zn, svuint32_t zm) __arm_streaming __arm_shared_za {
+void test_svadd_write_single4_u32(uint32_t slice_base, svuint32x4_t zn, svuint32_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svadd_write,_single,_za32,_u32,_vg1x4)(slice_base, zn, zm);
 }
 
@@ -158,7 +158,7 @@ void test_svadd_write_single4_u32(uint32_t slice_base, svuint32x4_t zn, svuint32
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.add.write.single.za.vg1x4.nxv2i64(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svadd_write_single4_s64(uint32_t slice_base, svint64x4_t zn, svint64_t zm) __arm_streaming __arm_shared_za {
+void test_svadd_write_single4_s64(uint32_t slice_base, svint64x4_t zn, svint64_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svadd_write,_single,_za64,_s64,_vg1x4)(slice_base, zn, zm);
 }
 
@@ -180,7 +180,7 @@ void test_svadd_write_single4_s64(uint32_t slice_base, svint64x4_t zn, svint64_t
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.add.write.single.za.vg1x4.nxv2i64(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svadd_write_single4_u64(uint32_t slice_base, svuint64x4_t zn, svuint64_t zm) __arm_streaming __arm_shared_za {
+void test_svadd_write_single4_u64(uint32_t slice_base, svuint64x4_t zn, svuint64_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svadd_write,_single,_za64,_u64,_vg1x4)(slice_base, zn, zm);
 }
 
@@ -208,7 +208,7 @@ void test_svadd_write_single4_u64(uint32_t slice_base, svuint64x4_t zn, svuint64
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.add.write.za.vg1x2.nxv4i32(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svadd_write_multi2_s32(uint32_t slice_base, svint32x2_t zn, svint32x2_t zm) __arm_streaming __arm_shared_za {
+void test_svadd_write_multi2_s32(uint32_t slice_base, svint32x2_t zn, svint32x2_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svadd_write,,_za32,_s32,_vg1x2)(slice_base, zn, zm);
 }
 
@@ -230,7 +230,7 @@ void test_svadd_write_multi2_s32(uint32_t slice_base, svint32x2_t zn, svint32x2_
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.add.write.za.vg1x2.nxv4i32(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svadd_write_multi2_u32(uint32_t slice_base, svuint32x2_t zn, svuint32x2_t zm) __arm_streaming __arm_shared_za {
+void test_svadd_write_multi2_u32(uint32_t slice_base, svuint32x2_t zn, svuint32x2_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svadd_write,,_za32,_u32,_vg1x2)(slice_base, zn, zm);
 }
 
@@ -252,7 +252,7 @@ void test_svadd_write_multi2_u32(uint32_t slice_base, svuint32x2_t zn, svuint32x
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.add.write.za.vg1x2.nxv2i64(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svadd_write_multi2_s64(uint32_t slice_base, svint64x2_t zn, svint64x2_t zm) __arm_streaming __arm_shared_za {
+void test_svadd_write_multi2_s64(uint32_t slice_base, svint64x2_t zn, svint64x2_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svadd_write,,_za64,_s64,_vg1x2)(slice_base, zn, zm);
 }
 
@@ -274,7 +274,7 @@ void test_svadd_write_multi2_s64(uint32_t slice_base, svint64x2_t zn, svint64x2_
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.add.write.za.vg1x2.nxv2i64(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svadd_write_multi2_u64(uint32_t slice_base, svuint64x2_t zn, svuint64x2_t zm) __arm_streaming __arm_shared_za {
+void test_svadd_write_multi2_u64(uint32_t slice_base, svuint64x2_t zn, svuint64x2_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svadd_write,,_za64,_u64,_vg1x2)(slice_base, zn, zm);
 }
 
@@ -306,7 +306,7 @@ void test_svadd_write_multi2_u64(uint32_t slice_base, svuint64x2_t zn, svuint64x
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.add.write.za.vg1x4.nxv4i32(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[TMP4]],  [[TMP5]],  [[TMP6]],  [[TMP7]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svadd_write_multi4_s32(uint32_t slice_base, svint32x4_t zn, svint32x4_t zm) __arm_streaming __arm_shared_za {
+void test_svadd_write_multi4_s32(uint32_t slice_base, svint32x4_t zn, svint32x4_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svadd_write,,_za32,_s32,_vg1x4)(slice_base, zn, zm);
 }
 
@@ -336,7 +336,7 @@ void test_svadd_write_multi4_s32(uint32_t slice_base, svint32x4_t zn, svint32x4_
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.add.write.za.vg1x4.nxv4i32(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[TMP4]],  [[TMP5]],  [[TMP6]],  [[TMP7]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svadd_write_multi4_u32(uint32_t slice_base, svuint32x4_t zn, svuint32x4_t zm) __arm_streaming __arm_shared_za {
+void test_svadd_write_multi4_u32(uint32_t slice_base, svuint32x4_t zn, svuint32x4_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svadd_write,,_za32,_u32,_vg1x4)(slice_base, zn, zm);
 }
 
@@ -366,7 +366,7 @@ void test_svadd_write_multi4_u32(uint32_t slice_base, svuint32x4_t zn, svuint32x
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.add.write.za.vg1x4.nxv2i64(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[TMP4]],  [[TMP5]],  [[TMP6]],  [[TMP7]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svadd_write_multi4_s64(uint32_t slice_base, svint64x4_t zn, svint64x4_t zm) __arm_streaming __arm_shared_za {
+void test_svadd_write_multi4_s64(uint32_t slice_base, svint64x4_t zn, svint64x4_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svadd_write,,_za64,_s64,_vg1x4)(slice_base, zn, zm);
 }
 
@@ -396,7 +396,7 @@ void test_svadd_write_multi4_s64(uint32_t slice_base, svint64x4_t zn, svint64x4_
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.add.write.za.vg1x4.nxv2i64(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[TMP4]],  [[TMP5]],  [[TMP6]],  [[TMP7]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svadd_write_multi4_u64(uint32_t slice_base, svuint64x4_t zn, svuint64x4_t zm) __arm_streaming __arm_shared_za {
+void test_svadd_write_multi4_u64(uint32_t slice_base, svuint64x4_t zn, svuint64x4_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svadd_write,,_za64,_u64,_vg1x4)(slice_base, zn, zm);
 }
 
@@ -420,7 +420,7 @@ void test_svadd_write_multi4_u64(uint32_t slice_base, svuint64x4_t zn, svuint64x
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.add.za32.vg1x2.nxv4f32(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svadd_za32_vg1x2_f32(uint32_t slice_base, svfloat32x2_t zn) __arm_streaming __arm_shared_za {
+void test_svadd_za32_vg1x2_f32(uint32_t slice_base, svfloat32x2_t zn) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svadd_za32,,_f32,,_vg1x2)(slice_base, zn);
 }
 
@@ -438,7 +438,7 @@ void test_svadd_za32_vg1x2_f32(uint32_t slice_base, svfloat32x2_t zn) __arm_stre
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.add.za32.vg1x2.nxv4i32(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svadd_za32_vg1x2_s32(uint32_t slice_base, svint32x2_t zn) __arm_streaming __arm_shared_za {
+void test_svadd_za32_vg1x2_s32(uint32_t slice_base, svint32x2_t zn) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svadd_za32,,_s32,,_vg1x2)(slice_base , zn);
 }
 
@@ -456,7 +456,7 @@ void test_svadd_za32_vg1x2_s32(uint32_t slice_base, svint32x2_t zn) __arm_stream
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.add.za32.vg1x2.nxv4i32(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svadd_za32_vg1x2_u32(uint32_t slice_base, svuint32x2_t zn) __arm_streaming __arm_shared_za {
+void test_svadd_za32_vg1x2_u32(uint32_t slice_base, svuint32x2_t zn) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svadd_za32,,_u32,,_vg1x2)(slice_base, zn);
 }
 
@@ -474,7 +474,7 @@ void test_svadd_za32_vg1x2_u32(uint32_t slice_base, svuint32x2_t zn) __arm_strea
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.add.za64.vg1x2.nxv2f64(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svadd_za64_vg1x2_f64(uint32_t slice_base, svfloat64x2_t zn) __arm_streaming __arm_shared_za {
+void test_svadd_za64_vg1x2_f64(uint32_t slice_base, svfloat64x2_t zn) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svadd_za64,,_f64,,_vg1x2)(slice_base, zn);
 }
 
@@ -492,7 +492,7 @@ void test_svadd_za64_vg1x2_f64(uint32_t slice_base, svfloat64x2_t zn) __arm_stre
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.add.za64.vg1x2.nxv2i64(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svadd_za64_vg1x2_s64(uint32_t slice_base, svint64x2_t zn) __arm_streaming __arm_shared_za {
+void test_svadd_za64_vg1x2_s64(uint32_t slice_base, svint64x2_t zn) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svadd_za64,,_s64,,_vg1x2)(slice_base, zn);
 }
 
@@ -510,7 +510,7 @@ void test_svadd_za64_vg1x2_s64(uint32_t slice_base, svint64x2_t zn) __arm_stream
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.add.za64.vg1x2.nxv2i64(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svadd_za64_vg1x2_u64(uint32_t slice_base, svuint64x2_t zn) __arm_streaming __arm_shared_za {
+void test_svadd_za64_vg1x2_u64(uint32_t slice_base, svuint64x2_t zn) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svadd_za64,,_u64,,_vg1x2)(slice_base, zn);
 }
 
@@ -534,7 +534,7 @@ void test_svadd_za64_vg1x2_u64(uint32_t slice_base, svuint64x2_t zn) __arm_strea
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.add.za32.vg1x4.nxv4f32(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svadd_za32_vg1x4_f32(uint32_t slice_base, svfloat32x4_t zn) __arm_streaming __arm_shared_za {
+void test_svadd_za32_vg1x4_f32(uint32_t slice_base, svfloat32x4_t zn) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svadd_za32,,_f32,,_vg1x4)(slice_base, zn);
 }
 
@@ -556,7 +556,7 @@ void test_svadd_za32_vg1x4_f32(uint32_t slice_base, svfloat32x4_t zn) __arm_stre
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.add.za32.vg1x4.nxv4i32(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svadd_za32_vg1x4_s32(uint32_t slice_base, svint32x4_t zn) __arm_streaming __arm_shared_za {
+void test_svadd_za32_vg1x4_s32(uint32_t slice_base, svint32x4_t zn) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svadd_za32,,_s32,,_vg1x4)(slice_base, zn);
 }
 
@@ -578,7 +578,7 @@ void test_svadd_za32_vg1x4_s32(uint32_t slice_base, svint32x4_t zn) __arm_stream
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.add.za32.vg1x4.nxv4i32(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svadd_za32_vg1x4_u32(uint32_t slice_base, svuint32x4_t zn) __arm_streaming __arm_shared_za {
+void test_svadd_za32_vg1x4_u32(uint32_t slice_base, svuint32x4_t zn) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svadd_za32,,_u32,,_vg1x4)(slice_base, zn);
 }
 
@@ -600,7 +600,7 @@ void test_svadd_za32_vg1x4_u32(uint32_t slice_base, svuint32x4_t zn) __arm_strea
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.add.za64.vg1x4.nxv2f64(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svadd_za64_vg1x4_f64(uint32_t slice_base, svfloat64x4_t zn) __arm_streaming __arm_shared_za {
+void test_svadd_za64_vg1x4_f64(uint32_t slice_base, svfloat64x4_t zn) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svadd_za64,,_f64,,_vg1x4)(slice_base, zn);
 }
 
@@ -622,7 +622,7 @@ void test_svadd_za64_vg1x4_f64(uint32_t slice_base, svfloat64x4_t zn) __arm_stre
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.add.za64.vg1x4.nxv2i64(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svadd_za64_vg1x4_s64(uint32_t slice_base, svint64x4_t zn) __arm_streaming __arm_shared_za {
+void test_svadd_za64_vg1x4_s64(uint32_t slice_base, svint64x4_t zn) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svadd_za64,,_s64,,_vg1x4)(slice_base, zn);
 }
 
@@ -644,6 +644,6 @@ void test_svadd_za64_vg1x4_s64(uint32_t slice_base, svint64x4_t zn) __arm_stream
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.add.za64.vg1x4.nxv2i64(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svadd_za64_vg1x4_u64(uint32_t slice_base, svuint64x4_t zn) __arm_streaming __arm_shared_za {
+void test_svadd_za64_vg1x4_u64(uint32_t slice_base, svuint64x4_t zn) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svadd_za64,,_u64,,_vg1x4)(slice_base, zn);
 }
diff --git a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_bmop.c b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_bmop.c
index 67a330625884f..e6415c66fbb48 100644
--- a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_bmop.c
+++ b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_bmop.c
@@ -33,7 +33,7 @@
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.bmopa.za32.nxv4i32(i32 3,  [[TMP0]],  [[TMP1]],  [[ZN:%.*]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svbmopa_u32(svbool_t pn, svbool_t pm, svuint32_t zn, svuint32_t zm) __arm_streaming __arm_shared_za {
+void test_svbmopa_u32(svbool_t pn, svbool_t pm, svuint32_t zn, svuint32_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svbmopa_za32,_u32,_m,)(3, pn, pm, zn, zm);
 }
 
@@ -51,7 +51,7 @@ void test_svbmopa_u32(svbool_t pn, svbool_t pm, svuint32_t zn, svuint32_t zm) __
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.bmopa.za32.nxv4i32(i32 3,  [[TMP0]],  [[TMP1]],  [[ZN:%.*]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svbmopa_s32(svbool_t pn, svbool_t pm, svint32_t zn, svint32_t zm) __arm_streaming __arm_shared_za {
+void test_svbmopa_s32(svbool_t pn, svbool_t pm, svint32_t zn, svint32_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svbmopa_za32,_s32,_m,)(3, pn, pm, zn, zm);
 }
 
@@ -71,7 +71,7 @@ void test_svbmopa_s32(svbool_t pn, svbool_t pm, svint32_t zn, svint32_t zm) __ar
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.bmops.za32.nxv4i32(i32 3,  [[TMP0]],  [[TMP1]],  [[ZN:%.*]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svbmops_u32(svbool_t pn, svbool_t pm, svuint32_t zn, svuint32_t zm) __arm_streaming __arm_shared_za {
+void test_svbmops_u32(svbool_t pn, svbool_t pm, svuint32_t zn, svuint32_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svbmops_za32,_u32,_m,)(3, pn, pm, zn, zm);
 }
 
@@ -89,6 +89,6 @@ void test_svbmops_u32(svbool_t pn, svbool_t pm, svuint32_t zn, svuint32_t zm) __
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.bmops.za32.nxv4i32(i32 3,  [[TMP0]],  [[TMP1]],  [[ZN:%.*]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svbmops_s32(svbool_t pn, svbool_t pm, svint32_t zn, svint32_t zm) __arm_streaming __arm_shared_za {
+void test_svbmops_s32(svbool_t pn, svbool_t pm, svint32_t zn, svint32_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svbmops_za32,_s32,_m,)(3, pn, pm, zn, zm);
 }
diff --git a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_cvt.c b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_cvt.c
index a3ee7d2092f79..f596b3167f2b8 100644
--- a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_cvt.c
+++ b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_cvt.c
@@ -59,7 +59,7 @@ svbfloat16_t test_cvt_bf16_x2(svfloat32x2_t zn)  __arm_streaming {
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv4i32.nxv8i32( [[ZN:%.*]], i64 0)
 // CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.vector.extract.nxv4i32.nxv8i32( [[ZN]], i64 4)
-// CHECK-NEXT:    [[TMP2:%.*]] = tail call { ,  } @llvm.aarch64.sve.fcvtu.x2.nxv4f32( [[TMP0]],  [[TMP1]])
+// CHECK-NEXT:    [[TMP2:%.*]] = tail call { ,  } @llvm.aarch64.sve.ucvtf.x2.nxv4f32( [[TMP0]],  [[TMP1]])
 // CHECK-NEXT:    [[TMP3:%.*]] = extractvalue { ,  } [[TMP2]], 0
 // CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv8f32.nxv4f32( poison,  [[TMP3]], i64 0)
 // CHECK-NEXT:    [[TMP5:%.*]] = extractvalue { ,  } [[TMP2]], 1
@@ -70,7 +70,7 @@ svbfloat16_t test_cvt_bf16_x2(svfloat32x2_t zn)  __arm_streaming {
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv4i32.nxv8i32( [[ZN:%.*]], i64 0)
 // CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.vector.extract.nxv4i32.nxv8i32( [[ZN]], i64 4)
-// CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call { ,  } @llvm.aarch64.sve.fcvtu.x2.nxv4f32( [[TMP0]],  [[TMP1]])
+// CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call { ,  } @llvm.aarch64.sve.ucvtf.x2.nxv4f32( [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    [[TMP3:%.*]] = extractvalue { ,  } [[TMP2]], 0
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv8f32.nxv4f32( poison,  [[TMP3]], i64 0)
 // CPP-CHECK-NEXT:    [[TMP5:%.*]] = extractvalue { ,  } [[TMP2]], 1
@@ -85,7 +85,7 @@ svfloat32x2_t test_svcvt_f32_u32_x2(svuint32x2_t zn)  __arm_streaming {
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv4i32.nxv8i32( [[ZN:%.*]], i64 0)
 // CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.vector.extract.nxv4i32.nxv8i32( [[ZN]], i64 4)
-// CHECK-NEXT:    [[TMP2:%.*]] = tail call { ,  } @llvm.aarch64.sve.fcvts.x2.nxv4f32( [[TMP0]],  [[TMP1]])
+// CHECK-NEXT:    [[TMP2:%.*]] = tail call { ,  } @llvm.aarch64.sve.scvtf.x2.nxv4f32( [[TMP0]],  [[TMP1]])
 // CHECK-NEXT:    [[TMP3:%.*]] = extractvalue { ,  } [[TMP2]], 0
 // CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv8f32.nxv4f32( poison,  [[TMP3]], i64 0)
 // CHECK-NEXT:    [[TMP5:%.*]] = extractvalue { ,  } [[TMP2]], 1
@@ -96,7 +96,7 @@ svfloat32x2_t test_svcvt_f32_u32_x2(svuint32x2_t zn)  __arm_streaming {
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv4i32.nxv8i32( [[ZN:%.*]], i64 0)
 // CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.vector.extract.nxv4i32.nxv8i32( [[ZN]], i64 4)
-// CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call { ,  } @llvm.aarch64.sve.fcvts.x2.nxv4f32( [[TMP0]],  [[TMP1]])
+// CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call { ,  } @llvm.aarch64.sve.scvtf.x2.nxv4f32( [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    [[TMP3:%.*]] = extractvalue { ,  } [[TMP2]], 0
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv8f32.nxv4f32( poison,  [[TMP3]], i64 0)
 // CPP-CHECK-NEXT:    [[TMP5:%.*]] = extractvalue { ,  } [[TMP2]], 1
@@ -111,7 +111,7 @@ svfloat32x2_t test_svcvt_f32_s32_x2(svint32x2_t zn) __arm_streaming {
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv4f32.nxv8f32( [[ZN:%.*]], i64 0)
 // CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.vector.extract.nxv4f32.nxv8f32( [[ZN]], i64 4)
-// CHECK-NEXT:    [[TMP2:%.*]] = tail call { ,  } @llvm.aarch64.sve.ucvtf.x2.nxv4f32( [[TMP0]],  [[TMP1]])
+// CHECK-NEXT:    [[TMP2:%.*]] = tail call { ,  } @llvm.aarch64.sve.fcvtu.x2.nxv4f32( [[TMP0]],  [[TMP1]])
 // CHECK-NEXT:    [[TMP3:%.*]] = extractvalue { ,  } [[TMP2]], 0
 // CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv8i32.nxv4i32( poison,  [[TMP3]], i64 0)
 // CHECK-NEXT:    [[TMP5:%.*]] = extractvalue { ,  } [[TMP2]], 1
@@ -122,7 +122,7 @@ svfloat32x2_t test_svcvt_f32_s32_x2(svint32x2_t zn) __arm_streaming {
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv4f32.nxv8f32( [[ZN:%.*]], i64 0)
 // CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.vector.extract.nxv4f32.nxv8f32( [[ZN]], i64 4)
-// CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call { ,  } @llvm.aarch64.sve.ucvtf.x2.nxv4f32( [[TMP0]],  [[TMP1]])
+// CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call { ,  } @llvm.aarch64.sve.fcvtu.x2.nxv4f32( [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    [[TMP3:%.*]] = extractvalue { ,  } [[TMP2]], 0
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv8i32.nxv4i32( poison,  [[TMP3]], i64 0)
 // CPP-CHECK-NEXT:    [[TMP5:%.*]] = extractvalue { ,  } [[TMP2]], 1
@@ -137,7 +137,7 @@ svuint32x2_t test_svcvt_u32_f32_x2(svfloat32x2_t zn)  __arm_streaming {
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv4f32.nxv8f32( [[ZN:%.*]], i64 0)
 // CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.vector.extract.nxv4f32.nxv8f32( [[ZN]], i64 4)
-// CHECK-NEXT:    [[TMP2:%.*]] = tail call { ,  } @llvm.aarch64.sve.scvtf.x2.nxv4f32( [[TMP0]],  [[TMP1]])
+// CHECK-NEXT:    [[TMP2:%.*]] = tail call { ,  } @llvm.aarch64.sve.fcvts.x2.nxv4f32( [[TMP0]],  [[TMP1]])
 // CHECK-NEXT:    [[TMP3:%.*]] = extractvalue { ,  } [[TMP2]], 0
 // CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv8i32.nxv4i32( poison,  [[TMP3]], i64 0)
 // CHECK-NEXT:    [[TMP5:%.*]] = extractvalue { ,  } [[TMP2]], 1
@@ -148,7 +148,7 @@ svuint32x2_t test_svcvt_u32_f32_x2(svfloat32x2_t zn)  __arm_streaming {
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv4f32.nxv8f32( [[ZN:%.*]], i64 0)
 // CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.vector.extract.nxv4f32.nxv8f32( [[ZN]], i64 4)
-// CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call { ,  } @llvm.aarch64.sve.scvtf.x2.nxv4f32( [[TMP0]],  [[TMP1]])
+// CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call { ,  } @llvm.aarch64.sve.fcvts.x2.nxv4f32( [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    [[TMP3:%.*]] = extractvalue { ,  } [[TMP2]], 0
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv8i32.nxv4i32( poison,  [[TMP3]], i64 0)
 // CPP-CHECK-NEXT:    [[TMP5:%.*]] = extractvalue { ,  } [[TMP2]], 1
@@ -166,7 +166,7 @@ svint32x2_t test_svcvt_s32_f32_x2(svfloat32x2_t zn) __arm_streaming {
 // CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.vector.extract.nxv4i32.nxv16i32( [[ZN]], i64 4)
 // CHECK-NEXT:    [[TMP2:%.*]] = tail call  @llvm.vector.extract.nxv4i32.nxv16i32( [[ZN]], i64 8)
 // CHECK-NEXT:    [[TMP3:%.*]] = tail call  @llvm.vector.extract.nxv4i32.nxv16i32( [[ZN]], i64 12)
-// CHECK-NEXT:    [[TMP4:%.*]] = tail call { , , ,  } @llvm.aarch64.sve.fcvtu.x4.nxv4f32( [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
+// CHECK-NEXT:    [[TMP4:%.*]] = tail call { , , ,  } @llvm.aarch64.sve.ucvtf.x4.nxv4f32( [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CHECK-NEXT:    [[TMP5:%.*]] = extractvalue { , , ,  } [[TMP4]], 0
 // CHECK-NEXT:    [[TMP6:%.*]] = tail call  @llvm.vector.insert.nxv16f32.nxv4f32( poison,  [[TMP5]], i64 0)
 // CHECK-NEXT:    [[TMP7:%.*]] = extractvalue { , , ,  } [[TMP4]], 1
@@ -183,7 +183,7 @@ svint32x2_t test_svcvt_s32_f32_x2(svfloat32x2_t zn) __arm_streaming {
 // CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.vector.extract.nxv4i32.nxv16i32( [[ZN]], i64 4)
 // CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call  @llvm.vector.extract.nxv4i32.nxv16i32( [[ZN]], i64 8)
 // CPP-CHECK-NEXT:    [[TMP3:%.*]] = tail call  @llvm.vector.extract.nxv4i32.nxv16i32( [[ZN]], i64 12)
-// CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call { , , ,  } @llvm.aarch64.sve.fcvtu.x4.nxv4f32( [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
+// CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call { , , ,  } @llvm.aarch64.sve.ucvtf.x4.nxv4f32( [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    [[TMP5:%.*]] = extractvalue { , , ,  } [[TMP4]], 0
 // CPP-CHECK-NEXT:    [[TMP6:%.*]] = tail call  @llvm.vector.insert.nxv16f32.nxv4f32( poison,  [[TMP5]], i64 0)
 // CPP-CHECK-NEXT:    [[TMP7:%.*]] = extractvalue { , , ,  } [[TMP4]], 1
@@ -204,7 +204,7 @@ svfloat32x4_t test_svcvt_f32_u32_x4(svuint32x4_t zn) __arm_streaming {
 // CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.vector.extract.nxv4i32.nxv16i32( [[ZN]], i64 4)
 // CHECK-NEXT:    [[TMP2:%.*]] = tail call  @llvm.vector.extract.nxv4i32.nxv16i32( [[ZN]], i64 8)
 // CHECK-NEXT:    [[TMP3:%.*]] = tail call  @llvm.vector.extract.nxv4i32.nxv16i32( [[ZN]], i64 12)
-// CHECK-NEXT:    [[TMP4:%.*]] = tail call { , , ,  } @llvm.aarch64.sve.fcvts.x4.nxv4f32( [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
+// CHECK-NEXT:    [[TMP4:%.*]] = tail call { , , ,  } @llvm.aarch64.sve.scvtf.x4.nxv4f32( [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CHECK-NEXT:    [[TMP5:%.*]] = extractvalue { , , ,  } [[TMP4]], 0
 // CHECK-NEXT:    [[TMP6:%.*]] = tail call  @llvm.vector.insert.nxv16f32.nxv4f32( poison,  [[TMP5]], i64 0)
 // CHECK-NEXT:    [[TMP7:%.*]] = extractvalue { , , ,  } [[TMP4]], 1
@@ -221,7 +221,7 @@ svfloat32x4_t test_svcvt_f32_u32_x4(svuint32x4_t zn) __arm_streaming {
 // CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.vector.extract.nxv4i32.nxv16i32( [[ZN]], i64 4)
 // CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call  @llvm.vector.extract.nxv4i32.nxv16i32( [[ZN]], i64 8)
 // CPP-CHECK-NEXT:    [[TMP3:%.*]] = tail call  @llvm.vector.extract.nxv4i32.nxv16i32( [[ZN]], i64 12)
-// CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call { , , ,  } @llvm.aarch64.sve.fcvts.x4.nxv4f32( [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
+// CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call { , , ,  } @llvm.aarch64.sve.scvtf.x4.nxv4f32( [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    [[TMP5:%.*]] = extractvalue { , , ,  } [[TMP4]], 0
 // CPP-CHECK-NEXT:    [[TMP6:%.*]] = tail call  @llvm.vector.insert.nxv16f32.nxv4f32( poison,  [[TMP5]], i64 0)
 // CPP-CHECK-NEXT:    [[TMP7:%.*]] = extractvalue { , , ,  } [[TMP4]], 1
@@ -242,7 +242,7 @@ svfloat32x4_t test_svcvt_f32_s32_x4(svint32x4_t zn) __arm_streaming {
 // CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.vector.extract.nxv4f32.nxv16f32( [[ZN]], i64 4)
 // CHECK-NEXT:    [[TMP2:%.*]] = tail call  @llvm.vector.extract.nxv4f32.nxv16f32( [[ZN]], i64 8)
 // CHECK-NEXT:    [[TMP3:%.*]] = tail call  @llvm.vector.extract.nxv4f32.nxv16f32( [[ZN]], i64 12)
-// CHECK-NEXT:    [[TMP4:%.*]] = tail call { , , ,  } @llvm.aarch64.sve.ucvtf.x4.nxv4f32( [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
+// CHECK-NEXT:    [[TMP4:%.*]] = tail call { , , ,  } @llvm.aarch64.sve.fcvtu.x4.nxv4f32( [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CHECK-NEXT:    [[TMP5:%.*]] = extractvalue { , , ,  } [[TMP4]], 0
 // CHECK-NEXT:    [[TMP6:%.*]] = tail call  @llvm.vector.insert.nxv16i32.nxv4i32( poison,  [[TMP5]], i64 0)
 // CHECK-NEXT:    [[TMP7:%.*]] = extractvalue { , , ,  } [[TMP4]], 1
@@ -259,7 +259,7 @@ svfloat32x4_t test_svcvt_f32_s32_x4(svint32x4_t zn) __arm_streaming {
 // CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.vector.extract.nxv4f32.nxv16f32( [[ZN]], i64 4)
 // CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call  @llvm.vector.extract.nxv4f32.nxv16f32( [[ZN]], i64 8)
 // CPP-CHECK-NEXT:    [[TMP3:%.*]] = tail call  @llvm.vector.extract.nxv4f32.nxv16f32( [[ZN]], i64 12)
-// CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call { , , ,  } @llvm.aarch64.sve.ucvtf.x4.nxv4f32( [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
+// CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call { , , ,  } @llvm.aarch64.sve.fcvtu.x4.nxv4f32( [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    [[TMP5:%.*]] = extractvalue { , , ,  } [[TMP4]], 0
 // CPP-CHECK-NEXT:    [[TMP6:%.*]] = tail call  @llvm.vector.insert.nxv16i32.nxv4i32( poison,  [[TMP5]], i64 0)
 // CPP-CHECK-NEXT:    [[TMP7:%.*]] = extractvalue { , , ,  } [[TMP4]], 1
@@ -280,7 +280,7 @@ svuint32x4_t test_svcvt_u32_f32_x4(svfloat32x4_t zn) __arm_streaming {
 // CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.vector.extract.nxv4f32.nxv16f32( [[ZN]], i64 4)
 // CHECK-NEXT:    [[TMP2:%.*]] = tail call  @llvm.vector.extract.nxv4f32.nxv16f32( [[ZN]], i64 8)
 // CHECK-NEXT:    [[TMP3:%.*]] = tail call  @llvm.vector.extract.nxv4f32.nxv16f32( [[ZN]], i64 12)
-// CHECK-NEXT:    [[TMP4:%.*]] = tail call { , , ,  } @llvm.aarch64.sve.scvtf.x4.nxv4f32( [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
+// CHECK-NEXT:    [[TMP4:%.*]] = tail call { , , ,  } @llvm.aarch64.sve.fcvts.x4.nxv4f32( [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CHECK-NEXT:    [[TMP5:%.*]] = extractvalue { , , ,  } [[TMP4]], 0
 // CHECK-NEXT:    [[TMP6:%.*]] = tail call  @llvm.vector.insert.nxv16i32.nxv4i32( poison,  [[TMP5]], i64 0)
 // CHECK-NEXT:    [[TMP7:%.*]] = extractvalue { , , ,  } [[TMP4]], 1
@@ -297,7 +297,7 @@ svuint32x4_t test_svcvt_u32_f32_x4(svfloat32x4_t zn) __arm_streaming {
 // CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.vector.extract.nxv4f32.nxv16f32( [[ZN]], i64 4)
 // CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call  @llvm.vector.extract.nxv4f32.nxv16f32( [[ZN]], i64 8)
 // CPP-CHECK-NEXT:    [[TMP3:%.*]] = tail call  @llvm.vector.extract.nxv4f32.nxv16f32( [[ZN]], i64 12)
-// CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call { , , ,  } @llvm.aarch64.sve.scvtf.x4.nxv4f32( [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
+// CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call { , , ,  } @llvm.aarch64.sve.fcvts.x4.nxv4f32( [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    [[TMP5:%.*]] = extractvalue { , , ,  } [[TMP4]], 0
 // CPP-CHECK-NEXT:    [[TMP6:%.*]] = tail call  @llvm.vector.insert.nxv16i32.nxv4i32( poison,  [[TMP5]], i64 0)
 // CPP-CHECK-NEXT:    [[TMP7:%.*]] = extractvalue { , , ,  } [[TMP4]], 1
diff --git a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_fp_dots.c b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_fp_dots.c
index ff4176530710a..c4651f4d83e1f 100644
--- a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_fp_dots.c
+++ b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_fp_dots.c
@@ -35,7 +35,7 @@
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fdot.za32.vg1x2.nxv8f16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_multi_za32_vg1x2_f16(uint32_t slice_base, svfloat16x2_t zn, svfloat16x2_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_multi_za32_vg1x2_f16(uint32_t slice_base, svfloat16x2_t zn, svfloat16x2_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_za32,,,_f16,_vg1x2)(slice_base, zn, zm);
 }
 
@@ -65,7 +65,7 @@ void test_svdot_multi_za32_vg1x2_f16(uint32_t slice_base, svfloat16x2_t zn, svfl
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fdot.za32.vg1x4.nxv8f16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[TMP4]],  [[TMP5]],  [[TMP6]],  [[TMP7]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_multi_za32_vg1x4_f16(uint32_t slice_base, svfloat16x4_t zn, svfloat16x4_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_multi_za32_vg1x4_f16(uint32_t slice_base, svfloat16x4_t zn, svfloat16x4_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_za32,,,_f16,_vg1x4)(slice_base, zn, zm);
 }
 
@@ -86,7 +86,7 @@ void test_svdot_multi_za32_vg1x4_f16(uint32_t slice_base, svfloat16x4_t zn, svfl
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fdot.single.za32.vg1x2.nxv8f16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_single_za32_vg1x2_f16(uint32_t slice_base, svfloat16x2_t zn, svfloat16_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_single_za32_vg1x2_f16(uint32_t slice_base, svfloat16x2_t zn, svfloat16_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_single_za32,,_f16,,_vg1x2)(slice_base, zn, zm);
 }
 
@@ -108,7 +108,7 @@ void test_svdot_single_za32_vg1x2_f16(uint32_t slice_base, svfloat16x2_t zn, svf
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fdot.single.za32.vg1x4.nxv8f16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_single_za32_vg1x4_f16(uint32_t slice_base, svfloat16x4_t zn, svfloat16_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_single_za32_vg1x4_f16(uint32_t slice_base, svfloat16x4_t zn, svfloat16_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_single_za32,,_f16,,_vg1x4)(slice_base, zn, zm);
 }
 
@@ -129,7 +129,7 @@ void test_svdot_single_za32_vg1x4_f16(uint32_t slice_base, svfloat16x4_t zn, svf
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fdot.lane.za32.vg1x2.nxv8f16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]], i32 3)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_lane_za32_vg1x2_f16(uint32_t slice_base, svfloat16x2_t zn, svfloat16_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_lane_za32_vg1x2_f16(uint32_t slice_base, svfloat16x2_t zn, svfloat16_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_lane_za32,,,_f16,_vg1x2)(slice_base, zn, zm, 3);
 }
 
@@ -151,7 +151,7 @@ void test_svdot_lane_za32_vg1x2_f16(uint32_t slice_base, svfloat16x2_t zn, svflo
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fdot.lane.za32.vg1x4.nxv8f16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]], i32 3)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_lane_za32_vg1x4_f16(uint32_t slice_base, svfloat16x4_t zn, svfloat16_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_lane_za32_vg1x4_f16(uint32_t slice_base, svfloat16x4_t zn, svfloat16_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_lane_za32,,,_f16,_vg1x4)(slice_base, zn, zm, 3);
 }
 
@@ -176,7 +176,7 @@ void test_svdot_lane_za32_vg1x4_f16(uint32_t slice_base, svfloat16x4_t zn, svflo
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fdot.za32.vg1x2.nxv8bf16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_multi_za32_vg1x2_bf16(uint32_t slice_base, svbfloat16x2_t zn, svbfloat16x2_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_multi_za32_vg1x2_bf16(uint32_t slice_base, svbfloat16x2_t zn, svbfloat16x2_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_za32,,,_bf16,_vg1x2)(slice_base, zn, zm);
 }
 
@@ -206,7 +206,7 @@ void test_svdot_multi_za32_vg1x2_bf16(uint32_t slice_base, svbfloat16x2_t zn, sv
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fdot.za32.vg1x4.nxv8bf16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[TMP4]],  [[TMP5]],  [[TMP6]],  [[TMP7]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_multi_za32_vg1x4_bf16(uint32_t slice_base, svbfloat16x4_t zn, svbfloat16x4_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_multi_za32_vg1x4_bf16(uint32_t slice_base, svbfloat16x4_t zn, svbfloat16x4_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_za32,,,_bf16,_vg1x4)(slice_base, zn, zm);
 }
 
@@ -227,7 +227,7 @@ void test_svdot_multi_za32_vg1x4_bf16(uint32_t slice_base, svbfloat16x4_t zn, sv
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fdot.single.za32.vg1x2.nxv8bf16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_single_za32_vg1x2_bf16(uint32_t slice_base, svbfloat16x2_t zn, svbfloat16_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_single_za32_vg1x2_bf16(uint32_t slice_base, svbfloat16x2_t zn, svbfloat16_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_single_za32,,_bf16,,_vg1x2)(slice_base, zn, zm);
 }
 
@@ -249,7 +249,7 @@ void test_svdot_single_za32_vg1x2_bf16(uint32_t slice_base, svbfloat16x2_t zn, s
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fdot.single.za32.vg1x4.nxv8bf16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_single_za32_vg1x4_bf16(uint32_t slice_base, svbfloat16x4_t zn, svbfloat16_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_single_za32_vg1x4_bf16(uint32_t slice_base, svbfloat16x4_t zn, svbfloat16_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_single_za32,,_bf16,,_vg1x4)(slice_base, zn, zm);
 }
 
@@ -270,7 +270,7 @@ void test_svdot_single_za32_vg1x4_bf16(uint32_t slice_base, svbfloat16x4_t zn, s
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fdot.lane.za32.vg1x2.nxv8bf16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]], i32 3)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_lane_za32_vg1x2_bf16(uint32_t slice_base, svbfloat16x2_t zn, svbfloat16_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_lane_za32_vg1x2_bf16(uint32_t slice_base, svbfloat16x2_t zn, svbfloat16_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_lane_za32,,_bf16,,_vg1x2)(slice_base, zn, zm, 3);
 }
 
@@ -292,6 +292,6 @@ void test_svdot_lane_za32_vg1x2_bf16(uint32_t slice_base, svbfloat16x2_t zn, svb
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fdot.lane.za32.vg1x4.nxv8bf16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]], i32 3)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_lane_za32_vg1x4_bf16(uint32_t slice_base, svbfloat16x4_t zn, svbfloat16_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_lane_za32_vg1x4_bf16(uint32_t slice_base, svbfloat16x4_t zn, svbfloat16_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_lane_za32,,_bf16,,_vg1x4)(slice_base, zn, zm, 3);
 }
diff --git a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_int_dots.c b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_int_dots.c
index 0d85071b7fc3e..6e0e9433bae2f 100644
--- a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_int_dots.c
+++ b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_int_dots.c
@@ -35,7 +35,7 @@
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.udot.za32.vg1x2.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_multi_za32_vg1x2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16x2_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_multi_za32_vg1x2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16x2_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_za32,,,_u16,_vg1x2)(slice_base, zn, zm);
 }
 
@@ -65,7 +65,7 @@ void test_svdot_multi_za32_vg1x2_u16(uint32_t slice_base, svuint16x2_t zn, svuin
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.udot.za32.vg1x4.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[TMP4]],  [[TMP5]],  [[TMP6]],  [[TMP7]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_multi_za32_vg1x4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16x4_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_multi_za32_vg1x4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16x4_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_za32,,,_u16,_vg1x4)(slice_base, zn, zm);
 }
 
@@ -87,7 +87,7 @@ void test_svdot_multi_za32_vg1x4_u16(uint32_t slice_base, svuint16x4_t zn, svuin
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.udot.za32.vg1x2.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_multi_za32_vg1x2_u8(uint32_t slice_base, svuint8x2_t zn, svuint8x2_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_multi_za32_vg1x2_u8(uint32_t slice_base, svuint8x2_t zn, svuint8x2_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_za32,,,_u8,_vg1x2)(slice_base, zn, zm);
 }
 
@@ -117,7 +117,7 @@ void test_svdot_multi_za32_vg1x2_u8(uint32_t slice_base, svuint8x2_t zn, svuint8
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.udot.za32.vg1x4.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[TMP4]],  [[TMP5]],  [[TMP6]],  [[TMP7]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_multi_za32_vg1x4_u8(uint32_t slice_base, svuint8x4_t zn, svuint8x4_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_multi_za32_vg1x4_u8(uint32_t slice_base, svuint8x4_t zn, svuint8x4_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_za32,,,_u8,_vg1x4)(slice_base, zn, zm);
 }
 
@@ -139,7 +139,7 @@ void test_svdot_multi_za32_vg1x4_u8(uint32_t slice_base, svuint8x4_t zn, svuint8
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.udot.za64.vg1x2.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_multi_za64_vg1x2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16x2_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_multi_za64_vg1x2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16x2_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_za64,,,_u16,_vg1x2)(slice_base, zn, zm);
 }
 
@@ -169,7 +169,7 @@ void test_svdot_multi_za64_vg1x2_u16(uint32_t slice_base, svuint16x2_t zn, svuin
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.udot.za64.vg1x4.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[TMP4]],  [[TMP5]],  [[TMP6]],  [[TMP7]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_multi_za64_vg1x4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16x4_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_multi_za64_vg1x4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16x4_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_za64,,,_u16,_vg1x4)(slice_base, zn, zm);
 }
 
@@ -194,7 +194,7 @@ void test_svdot_multi_za64_vg1x4_u16(uint32_t slice_base, svuint16x4_t zn, svuin
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sdot.za32.vg1x2.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_multi_za32_vg1x2_s16(uint32_t slice_base, svint16x2_t zn, svint16x2_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_multi_za32_vg1x2_s16(uint32_t slice_base, svint16x2_t zn, svint16x2_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_za32,,,_s16,_vg1x2)(slice_base, zn, zm);
 }
 
@@ -224,7 +224,7 @@ void test_svdot_multi_za32_vg1x2_s16(uint32_t slice_base, svint16x2_t zn, svint1
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sdot.za32.vg1x4.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[TMP4]],  [[TMP5]],  [[TMP6]],  [[TMP7]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_multi_za32_vg1x4_s16(uint32_t slice_base, svint16x4_t zn, svint16x4_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_multi_za32_vg1x4_s16(uint32_t slice_base, svint16x4_t zn, svint16x4_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_za32,,,_s16,_vg1x4)(slice_base, zn, zm);
 }
 
@@ -246,7 +246,7 @@ void test_svdot_multi_za32_vg1x4_s16(uint32_t slice_base, svint16x4_t zn, svint1
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sdot.za32.vg1x2.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_multi_za32_vg1x2_s8(uint32_t slice_base, svint8x2_t zn, svint8x2_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_multi_za32_vg1x2_s8(uint32_t slice_base, svint8x2_t zn, svint8x2_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_za32,,,_s8,_vg1x2)(slice_base, zn, zm);
 }
 
@@ -276,7 +276,7 @@ void test_svdot_multi_za32_vg1x2_s8(uint32_t slice_base, svint8x2_t zn, svint8x2
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sdot.za32.vg1x4.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[TMP4]],  [[TMP5]],  [[TMP6]],  [[TMP7]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_multi_za32_vg1x4_s8(uint32_t slice_base, svint8x4_t zn, svint8x4_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_multi_za32_vg1x4_s8(uint32_t slice_base, svint8x4_t zn, svint8x4_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_za32,,,_s8,_vg1x4)(slice_base, zn, zm);
 }
 
@@ -298,7 +298,7 @@ void test_svdot_multi_za32_vg1x4_s8(uint32_t slice_base, svint8x4_t zn, svint8x4
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sdot.za64.vg1x2.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_multi_za64_vg1x2_s16(uint32_t slice_base, svint16x2_t zn, svint16x2_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_multi_za64_vg1x2_s16(uint32_t slice_base, svint16x2_t zn, svint16x2_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_za64,,,_s16,_vg1x2)(slice_base, zn, zm);
 }
 
@@ -328,7 +328,7 @@ void test_svdot_multi_za64_vg1x2_s16(uint32_t slice_base, svint16x2_t zn, svint1
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sdot.za64.vg1x4.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[TMP4]],  [[TMP5]],  [[TMP6]],  [[TMP7]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_multi_za64_vg1x4_s16(uint32_t slice_base, svint16x4_t zn, svint16x4_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_multi_za64_vg1x4_s16(uint32_t slice_base, svint16x4_t zn, svint16x4_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_za64,,,_s16,_vg1x4)(slice_base, zn, zm);
 }
 
@@ -349,7 +349,7 @@ void test_svdot_multi_za64_vg1x4_s16(uint32_t slice_base, svint16x4_t zn, svint1
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.udot.single.za32.vg1x2.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_single_za32_vg1x2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_single_za32_vg1x2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_single_za32,,_u16,,_vg1x2)(slice_base, zn, zm);
 }
 
@@ -371,7 +371,7 @@ void test_svdot_single_za32_vg1x2_u16(uint32_t slice_base, svuint16x2_t zn, svui
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.udot.single.za32.vg1x4.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_single_za32_vg1x4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_single_za32_vg1x4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_single_za32,,_u16,,_vg1x4)(slice_base, zn, zm);
 }
 
@@ -389,7 +389,7 @@ void test_svdot_single_za32_vg1x4_u16(uint32_t slice_base, svuint16x4_t zn, svui
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.udot.single.za32.vg1x2.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_single_za32_vg1x2_u8(uint32_t slice_base, svuint8x2_t zn, svuint8_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_single_za32_vg1x2_u8(uint32_t slice_base, svuint8x2_t zn, svuint8_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_single_za32,,_u8,,_vg1x2)(slice_base, zn, zm);
 }
 
@@ -411,7 +411,7 @@ void test_svdot_single_za32_vg1x2_u8(uint32_t slice_base, svuint8x2_t zn, svuint
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.udot.single.za32.vg1x4.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_single_za32_vg1x4_u8(uint32_t slice_base, svuint8x4_t zn, svuint8_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_single_za32_vg1x4_u8(uint32_t slice_base, svuint8x4_t zn, svuint8_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_single_za32,,_u8,,_vg1x4)(slice_base, zn, zm);
 }
 
@@ -429,7 +429,7 @@ void test_svdot_single_za32_vg1x4_u8(uint32_t slice_base, svuint8x4_t zn, svuint
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.udot.single.za64.vg1x2.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_single_za64_vg1x2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_single_za64_vg1x2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_single_za64,,_u16,,_vg1x2)(slice_base, zn, zm);
 }
 
@@ -451,7 +451,7 @@ void test_svdot_single_za64_vg1x2_u16(uint32_t slice_base, svuint16x2_t zn, svui
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.udot.single.za64.vg1x4.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_single_za64_vg1x4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_single_za64_vg1x4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_single_za64,,_u16,,_vg1x4)(slice_base, zn, zm);
 }
 
@@ -472,7 +472,7 @@ void test_svdot_single_za64_vg1x4_u16(uint32_t slice_base, svuint16x4_t zn, svui
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sdot.single.za32.vg1x2.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_single_za32_vg1x2_s16(uint32_t slice_base, svint16x2_t zn, svint16_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_single_za32_vg1x2_s16(uint32_t slice_base, svint16x2_t zn, svint16_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_single_za32,,_s16,,_vg1x2)(slice_base, zn, zm);
 }
 
@@ -494,7 +494,7 @@ void test_svdot_single_za32_vg1x2_s16(uint32_t slice_base, svint16x2_t zn, svint
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sdot.single.za32.vg1x4.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_single_za32_vg1x4_s16(uint32_t slice_base, svint16x4_t zn, svint16_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_single_za32_vg1x4_s16(uint32_t slice_base, svint16x4_t zn, svint16_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_single_za32,,_s16,,_vg1x4)(slice_base, zn, zm);
 }
 
@@ -512,7 +512,7 @@ void test_svdot_single_za32_vg1x4_s16(uint32_t slice_base, svint16x4_t zn, svint
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sdot.single.za32.vg1x2.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_single_za32_vg1x2_s8(uint32_t slice_base, svint8x2_t zn, svint8_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_single_za32_vg1x2_s8(uint32_t slice_base, svint8x2_t zn, svint8_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_single_za32,,_s8,,_vg1x2)(slice_base, zn, zm);
 }
 
@@ -534,7 +534,7 @@ void test_svdot_single_za32_vg1x2_s8(uint32_t slice_base, svint8x2_t zn, svint8_
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sdot.single.za32.vg1x4.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_single_za32_vg1x4_s8(uint32_t slice_base, svint8x4_t zn, svint8_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_single_za32_vg1x4_s8(uint32_t slice_base, svint8x4_t zn, svint8_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_single_za32,,_s8,,_vg1x4)(slice_base, zn, zm);
 }
 
@@ -552,7 +552,7 @@ void test_svdot_single_za32_vg1x4_s8(uint32_t slice_base, svint8x4_t zn, svint8_
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sdot.single.za64.vg1x2.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_single_za64_vg1x2_s16(uint32_t slice_base, svint16x2_t zn, svint16_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_single_za64_vg1x2_s16(uint32_t slice_base, svint16x2_t zn, svint16_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_single_za64,,_s16,,_vg1x2)(slice_base, zn, zm);
 }
 
@@ -574,7 +574,7 @@ void test_svdot_single_za64_vg1x2_s16(uint32_t slice_base, svint16x2_t zn, svint
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sdot.single.za64.vg1x4.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_single_za64_vg1x4_s16(uint32_t slice_base, svint16x4_t zn, svint16_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_single_za64_vg1x4_s16(uint32_t slice_base, svint16x4_t zn, svint16_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_single_za64,,_s16,,_vg1x4)(slice_base, zn, zm);
 }
 
@@ -594,7 +594,7 @@ void test_svdot_single_za64_vg1x4_s16(uint32_t slice_base, svint16x4_t zn, svint
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.udot.lane.za32.vg1x2.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]], i32 3)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_lane_za32_vg1x2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_lane_za32_vg1x2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_lane_za32,,_u16,,_vg1x2)(slice_base, zn, zm, 3);
 }
 
@@ -616,7 +616,7 @@ void test_svdot_lane_za32_vg1x2_u16(uint32_t slice_base, svuint16x2_t zn, svuint
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.udot.lane.za32.vg1x4.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]], i32 3)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_lane_za32_vg1x4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_lane_za32_vg1x4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_lane_za32,,_u16,,_vg1x4)(slice_base, zn, zm, 3);
 }
 
@@ -634,7 +634,7 @@ void test_svdot_lane_za32_vg1x4_u16(uint32_t slice_base, svuint16x4_t zn, svuint
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.udot.lane.za32.vg1x2.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]], i32 3)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_lane_za32_vg1x2_u8(uint32_t slice_base, svuint8x2_t zn, svuint8_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_lane_za32_vg1x2_u8(uint32_t slice_base, svuint8x2_t zn, svuint8_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_lane_za32,,_u8,,_vg1x2)(slice_base, zn, zm, 3);
 }
 
@@ -656,7 +656,7 @@ void test_svdot_lane_za32_vg1x2_u8(uint32_t slice_base, svuint8x2_t zn, svuint8_
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.udot.lane.za32.vg1x4.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]], i32 3)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_lane_za32_vg1x4_u8(uint32_t slice_base, svuint8x4_t zn, svuint8_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_lane_za32_vg1x4_u8(uint32_t slice_base, svuint8x4_t zn, svuint8_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_lane_za32,,_u8,,_vg1x4)(slice_base, zn, zm, 3);
 }
 
@@ -674,7 +674,7 @@ void test_svdot_lane_za32_vg1x4_u8(uint32_t slice_base, svuint8x4_t zn, svuint8_
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.udot.lane.za64.vg1x2.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]], i32 1)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_lane_za64_vg1x2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_lane_za64_vg1x2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_lane_za64,,,_u16,_vg1x2)(slice_base, zn, zm, 1);
 }
 
@@ -696,7 +696,7 @@ void test_svdot_lane_za64_vg1x2_u16(uint32_t slice_base, svuint16x2_t zn, svuint
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.udot.lane.za64.vg1x4.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]], i32 1)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_lane_za64_vg1x4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_lane_za64_vg1x4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_lane_za64,,,_u16,_vg1x4)(slice_base, zn, zm, 1);
 }
 
@@ -717,7 +717,7 @@ void test_svdot_lane_za64_vg1x4_u16(uint32_t slice_base, svuint16x4_t zn, svuint
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sdot.lane.za32.vg1x2.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]], i32 3)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_lane_za32_vg1x2_s16(uint32_t slice_base, svint16x2_t zn, svint16_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_lane_za32_vg1x2_s16(uint32_t slice_base, svint16x2_t zn, svint16_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_lane_za32,,,_s16,_vg1x2)(slice_base, zn, zm, 3);
 }
 
@@ -739,7 +739,7 @@ void test_svdot_lane_za32_vg1x2_s16(uint32_t slice_base, svint16x2_t zn, svint16
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sdot.lane.za32.vg1x4.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]], i32 3)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_lane_za32_vg1x4_s16(uint32_t slice_base, svint16x4_t zn, svint16_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_lane_za32_vg1x4_s16(uint32_t slice_base, svint16x4_t zn, svint16_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_lane_za32,,,_s16,_vg1x4)(slice_base, zn, zm, 3);
 }
 
@@ -757,7 +757,7 @@ void test_svdot_lane_za32_vg1x4_s16(uint32_t slice_base, svint16x4_t zn, svint16
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sdot.lane.za32.vg1x2.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]], i32 3)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_lane_za32_vg1x2_s8(uint32_t slice_base, svint8x2_t zn, svint8_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_lane_za32_vg1x2_s8(uint32_t slice_base, svint8x2_t zn, svint8_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_lane_za32,,,_s8,_vg1x2)(slice_base, zn, zm, 3);
 }
 
@@ -779,7 +779,7 @@ void test_svdot_lane_za32_vg1x2_s8(uint32_t slice_base, svint8x2_t zn, svint8_t
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sdot.lane.za32.vg1x4.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]], i32 3)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_lane_za32_vg1x4_s8(uint32_t slice_base, svint8x4_t zn, svint8_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_lane_za32_vg1x4_s8(uint32_t slice_base, svint8x4_t zn, svint8_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_lane_za32,,,_s8,_vg1x4)(slice_base, zn, zm, 3);
 }
 
@@ -797,7 +797,7 @@ void test_svdot_lane_za32_vg1x4_s8(uint32_t slice_base, svint8x4_t zn, svint8_t
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sdot.lane.za64.vg1x2.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]], i32 1)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_lane_za64_vg1x2_s16(uint32_t slice_base, svint16x2_t zn, svint16_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_lane_za64_vg1x2_s16(uint32_t slice_base, svint16x2_t zn, svint16_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_lane_za64,,,_s16,_vg1x2)(slice_base, zn, zm, 1);
 }
 
@@ -819,7 +819,7 @@ void test_svdot_lane_za64_vg1x2_s16(uint32_t slice_base, svint16x2_t zn, svint16
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sdot.lane.za64.vg1x4.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]], i32 1)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svdot_lane_za64_vg1x4_s16(uint32_t slice_base, svint16x4_t zn, svint16_t zm) __arm_streaming __arm_shared_za {
+void test_svdot_lane_za64_vg1x4_s16(uint32_t slice_base, svint16x4_t zn, svint16_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svdot_lane_za64,,,_s16,_vg1x4)(slice_base, zn, zm, 1);
 }
 
@@ -844,7 +844,7 @@ void test_svdot_lane_za64_vg1x4_s16(uint32_t slice_base, svint16x4_t zn, svint16
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.usdot.za32.vg1x2.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svusdot_multi_za32_vg1x2_u8(uint32_t slice_base, svuint8x2_t zn, svint8x2_t zm) __arm_streaming __arm_shared_za {
+void test_svusdot_multi_za32_vg1x2_u8(uint32_t slice_base, svuint8x2_t zn, svint8x2_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svusdot_za32,,,_u8,_vg1x2)(slice_base, zn, zm);
 }
 
@@ -874,7 +874,7 @@ void test_svusdot_multi_za32_vg1x2_u8(uint32_t slice_base, svuint8x2_t zn, svint
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.usdot.za32.vg1x4.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[TMP4]],  [[TMP5]],  [[TMP6]],  [[TMP7]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svusdot_multi_za32_vg1x4_u8(uint32_t slice_base, svuint8x4_t zn, svint8x4_t zm) __arm_streaming __arm_shared_za {
+void test_svusdot_multi_za32_vg1x4_u8(uint32_t slice_base, svuint8x4_t zn, svint8x4_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svusdot_za32,,,_u8,_vg1x4)(slice_base, zn, zm);
 }
 
@@ -895,7 +895,7 @@ void test_svusdot_multi_za32_vg1x4_u8(uint32_t slice_base, svuint8x4_t zn, svint
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.usdot.single.za32.vg1x2.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svusdot_single_za32_vg1x2_u8(uint32_t slice_base, svuint8x2_t zn, svint8_t zm) __arm_streaming __arm_shared_za {
+void test_svusdot_single_za32_vg1x2_u8(uint32_t slice_base, svuint8x2_t zn, svint8_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svusdot_single_za32,,_u8,,_vg1x2)(slice_base, zn, zm);
 }
 
@@ -917,7 +917,7 @@ void test_svusdot_single_za32_vg1x2_u8(uint32_t slice_base, svuint8x2_t zn, svin
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.usdot.single.za32.vg1x4.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svusdot_single_za32_vg1x4_u8(uint32_t slice_base, svuint8x4_t zn, svint8_t zm) __arm_streaming __arm_shared_za {
+void test_svusdot_single_za32_vg1x4_u8(uint32_t slice_base, svuint8x4_t zn, svint8_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svusdot_single_za32,,_u8,,_vg1x4)(slice_base, zn, zm);
 }
 
@@ -937,7 +937,7 @@ void test_svusdot_single_za32_vg1x4_u8(uint32_t slice_base, svuint8x4_t zn, svin
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.usdot.lane.za32.vg1x2.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]], i32 3)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svusdot_lane_za32_vg1x2_u8(uint32_t slice_base, svuint8x2_t zn, svint8_t zm) __arm_streaming __arm_shared_za {
+void test_svusdot_lane_za32_vg1x2_u8(uint32_t slice_base, svuint8x2_t zn, svint8_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svusdot_lane_za32,,_u8,,_vg1x2)(slice_base, zn, zm, 3);
 }
 
@@ -959,7 +959,7 @@ void test_svusdot_lane_za32_vg1x2_u8(uint32_t slice_base, svuint8x2_t zn, svint8
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.usdot.lane.za32.vg1x4.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]], i32 3)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svusdot_lane_za32_vg1x4_u8(uint32_t slice_base, svuint8x4_t zn, svint8_t zm) __arm_streaming __arm_shared_za {
+void test_svusdot_lane_za32_vg1x4_u8(uint32_t slice_base, svuint8x4_t zn, svint8_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svusdot_lane_za32,,_u8,,_vg1x4)(slice_base, zn, zm, 3);
 }
 
@@ -980,7 +980,7 @@ void test_svusdot_lane_za32_vg1x4_u8(uint32_t slice_base, svuint8x4_t zn, svint8
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sudot.single.za32.vg1x2.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svsudot_single_za32_vg1x2_s8(uint32_t slice_base, svint8x2_t zn, svuint8_t zm) __arm_streaming __arm_shared_za {
+void test_svsudot_single_za32_vg1x2_s8(uint32_t slice_base, svint8x2_t zn, svuint8_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svsudot_single_za32,,_s8,,_vg1x2)(slice_base, zn, zm);
 }
 
@@ -1002,7 +1002,7 @@ void test_svsudot_single_za32_vg1x2_s8(uint32_t slice_base, svint8x2_t zn, svuin
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sudot.single.za32.vg1x4.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svsudot_single_za32_vg1x4_s8(uint32_t slice_base, svint8x4_t zn, svuint8_t zm) __arm_streaming __arm_shared_za {
+void test_svsudot_single_za32_vg1x4_s8(uint32_t slice_base, svint8x4_t zn, svuint8_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svsudot_single_za32,,_s8,,_vg1x4)(slice_base, zn, zm);
 }
 
@@ -1026,7 +1026,7 @@ void test_svsudot_single_za32_vg1x4_s8(uint32_t slice_base, svint8x4_t zn, svuin
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.usdot.za32.vg1x2.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP2]],  [[TMP3]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svsudot_multi_za32_vg1x2_s8(uint32_t slice_base, svint8x2_t zn, svuint8x2_t zm) __arm_streaming __arm_shared_za {
+void test_svsudot_multi_za32_vg1x2_s8(uint32_t slice_base, svint8x2_t zn, svuint8x2_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svsudot_za32,,_s8,,_vg1x2)(slice_base, zn, zm);
 }
 
@@ -1056,7 +1056,7 @@ void test_svsudot_multi_za32_vg1x2_s8(uint32_t slice_base, svint8x2_t zn, svuint
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.usdot.za32.vg1x4.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP4]],  [[TMP5]],  [[TMP6]],  [[TMP7]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svsudot_multi_za32_vg1x4_s8(uint32_t slice_base, svint8x4_t zn, svuint8x4_t zm) __arm_streaming __arm_shared_za {
+void test_svsudot_multi_za32_vg1x4_s8(uint32_t slice_base, svint8x4_t zn, svuint8x4_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svsudot_za32,,_s8,,_vg1x4)(slice_base, zn, zm);
 }
 
@@ -1076,7 +1076,7 @@ void test_svsudot_multi_za32_vg1x4_s8(uint32_t slice_base, svint8x4_t zn, svuint
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sudot.lane.za32.vg1x2.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]], i32 3)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svsudot_lane_za32_vg1x2_s8(uint32_t slice_base, svint8x2_t zn, svuint8_t zm) __arm_streaming __arm_shared_za {
+void test_svsudot_lane_za32_vg1x2_s8(uint32_t slice_base, svint8x2_t zn, svuint8_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svsudot_lane_za32,,_s8,,_vg1x2)(slice_base, zn, zm, 3);
 }
 
@@ -1098,6 +1098,6 @@ void test_svsudot_lane_za32_vg1x2_s8(uint32_t slice_base, svint8x2_t zn, svuint8
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sudot.lane.za32.vg1x4.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]], i32 3)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svsudot_lane_za32_vg1x4_s8(uint32_t slice_base, svint8x4_t zn, svuint8_t zm) __arm_streaming __arm_shared_za {
+void test_svsudot_lane_za32_vg1x4_s8(uint32_t slice_base, svint8x4_t zn, svuint8_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svsudot_lane_za32,,_s8,,_vg1x4)(slice_base, zn, zm, 3);
 }
diff --git a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_ldr_str_zt.c b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_ldr_str_zt.c
index 83fbd6e5855ca..2e3cbd33d1e2c 100644
--- a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_ldr_str_zt.c
+++ b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_ldr_str_zt.c
@@ -20,7 +20,7 @@
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.ldr.zt(i32 0, ptr [[BASE:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svldr_zt(const void *base) __arm_streaming_compatible __arm_shared_za {
+void test_svldr_zt(const void *base) __arm_streaming_compatible __arm_out("za") {
   svldr_zt(0, base);
 }
 
@@ -36,6 +36,6 @@ void test_svldr_zt(const void *base) __arm_streaming_compatible __arm_shared_za
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.str.zt(i32 0, ptr [[BASE:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svstr_zt(void *base) __arm_streaming_compatible __arm_shared_za __arm_preserves_za {
+void test_svstr_zt(void *base) __arm_streaming_compatible __arm_in("za") {
   svstr_zt(0, base);
 }
diff --git a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_luti2_lane_zt.c b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_luti2_lane_zt.c
index cb34db3695b24..a31c6d982100b 100644
--- a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_luti2_lane_zt.c
+++ b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_luti2_lane_zt.c
@@ -19,7 +19,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sme.luti2.lane.zt.nxv16i8(i32 0,  [[ZN:%.*]], i32 15)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svuint8_t test_svluti2_lane_zt_u8(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint8_t test_svluti2_lane_zt_u8(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti2_lane_zt_u8(0, zn, 15);
 }
 
@@ -34,7 +34,7 @@ svuint8_t test_svluti2_lane_zt_u8(svuint8_t zn) __arm_streaming __arm_shared_za
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sme.luti2.lane.zt.nxv16i8(i32 0,  [[ZN:%.*]], i32 15)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svint8_t test_svluti2_lane_zt_s8(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint8_t test_svluti2_lane_zt_s8(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti2_lane_zt_s8(0, zn, 15);
 }
 
@@ -48,7 +48,7 @@ svint8_t test_svluti2_lane_zt_s8(svuint8_t zn) __arm_streaming __arm_shared_za _
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sme.luti2.lane.zt.nxv8i16(i32 0,  [[ZN:%.*]], i32 15)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svuint16_t test_svluti2_lane_zt_u16(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint16_t test_svluti2_lane_zt_u16(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti2_lane_zt_u16(0, zn, 15);
 }
 
@@ -63,7 +63,7 @@ svuint16_t test_svluti2_lane_zt_u16(svuint8_t zn) __arm_streaming __arm_shared_z
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sme.luti2.lane.zt.nxv8i16(i32 0,  [[ZN:%.*]], i32 15)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svint16_t test_svluti2_lane_zt_s16(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint16_t test_svluti2_lane_zt_s16(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti2_lane_zt_s16(0, zn, 15);
 }
 
@@ -77,7 +77,7 @@ svint16_t test_svluti2_lane_zt_s16(svuint8_t zn) __arm_streaming __arm_shared_za
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sme.luti2.lane.zt.nxv8f16(i32 0,  [[ZN:%.*]], i32 15)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svfloat16_t test_svluti2_lane_zt_f16(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svfloat16_t test_svluti2_lane_zt_f16(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti2_lane_zt_f16(0, zn, 15);
 }
 
@@ -91,7 +91,7 @@ svfloat16_t test_svluti2_lane_zt_f16(svuint8_t zn) __arm_streaming __arm_shared_
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sme.luti2.lane.zt.nxv8bf16(i32 0,  [[ZN:%.*]], i32 15)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svbfloat16_t test_svluti2_lane_zt_bf16(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svbfloat16_t test_svluti2_lane_zt_bf16(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti2_lane_zt_bf16(0, zn, 15);
 }
 
@@ -105,7 +105,7 @@ svbfloat16_t test_svluti2_lane_zt_bf16(svuint8_t zn) __arm_streaming __arm_share
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sme.luti2.lane.zt.nxv4i32(i32 0,  [[ZN:%.*]], i32 15)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svuint32_t test_svluti2_lane_zt_u32(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint32_t test_svluti2_lane_zt_u32(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti2_lane_zt_u32(0, zn, 15);
 }
 
@@ -119,7 +119,7 @@ svuint32_t test_svluti2_lane_zt_u32(svuint8_t zn) __arm_streaming __arm_shared_z
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sme.luti2.lane.zt.nxv4i32(i32 0,  [[ZN:%.*]], i32 15)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svint32_t test_svluti2_lane_zt_s32(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint32_t test_svluti2_lane_zt_s32(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti2_lane_zt_s32(0, zn, 15);
 }
 
@@ -133,6 +133,6 @@ svint32_t test_svluti2_lane_zt_s32(svuint8_t zn) __arm_streaming __arm_shared_za
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sme.luti2.lane.zt.nxv4f32(i32 0,  [[ZN:%.*]], i32 15)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svfloat32_t test_svluti2_lane_zt_f32(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svfloat32_t test_svluti2_lane_zt_f32(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti2_lane_zt_f32(0, zn, 15);
 }
diff --git a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_luti2_lane_zt_x2.c b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_luti2_lane_zt_x2.c
index 04f37af46767a..db44f52a37bf0 100644
--- a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_luti2_lane_zt_x2.c
+++ b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_luti2_lane_zt_x2.c
@@ -26,7 +26,7 @@
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv32i8.nxv16i8( [[TMP2]],  [[TMP3]], i64 16)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svuint8x2_t test_svluti2_lane_zt_u8(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint8x2_t test_svluti2_lane_zt_u8(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti2_lane_zt_u8_x2(0, zn, 7);
 }
 
@@ -49,7 +49,7 @@ svuint8x2_t test_svluti2_lane_zt_u8(svuint8_t zn) __arm_streaming __arm_shared_z
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv32i8.nxv16i8( [[TMP2]],  [[TMP3]], i64 16)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svint8x2_t test_svluti2_lane_zt_s8(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint8x2_t test_svluti2_lane_zt_s8(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti2_lane_zt_s8_x2(0, zn, 7);
 }
 
@@ -71,7 +71,7 @@ svint8x2_t test_svluti2_lane_zt_s8(svuint8_t zn) __arm_streaming __arm_shared_za
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv16i16.nxv8i16( [[TMP2]],  [[TMP3]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svuint16x2_t test_svluti2_lane_zt_u16(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint16x2_t test_svluti2_lane_zt_u16(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti2_lane_zt_u16_x2(0, zn, 7);
 }
 
@@ -94,7 +94,7 @@ svuint16x2_t test_svluti2_lane_zt_u16(svuint8_t zn) __arm_streaming __arm_shared
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv16i16.nxv8i16( [[TMP2]],  [[TMP3]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svint16x2_t test_svluti2_lane_zt_s16(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint16x2_t test_svluti2_lane_zt_s16(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti2_lane_zt_s16_x2(0, zn, 7);
 }
 
@@ -116,7 +116,7 @@ svint16x2_t test_svluti2_lane_zt_s16(svuint8_t zn) __arm_streaming __arm_shared_
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv16f16.nxv8f16( [[TMP2]],  [[TMP3]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svfloat16x2_t test_svluti2_lane_zt_f16(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svfloat16x2_t test_svluti2_lane_zt_f16(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti2_lane_zt_f16_x2(0, zn, 7);
 }
 
@@ -138,7 +138,7 @@ svfloat16x2_t test_svluti2_lane_zt_f16(svuint8_t zn) __arm_streaming __arm_share
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv16bf16.nxv8bf16( [[TMP2]],  [[TMP3]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svbfloat16x2_t test_svluti2_lane_zt_bf16(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svbfloat16x2_t test_svluti2_lane_zt_bf16(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti2_lane_zt_bf16_x2(0, zn, 7);
 }
 
@@ -160,7 +160,7 @@ svbfloat16x2_t test_svluti2_lane_zt_bf16(svuint8_t zn) __arm_streaming __arm_sha
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv8i32.nxv4i32( [[TMP2]],  [[TMP3]], i64 4)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svuint32x2_t test_svluti2_lane_zt_u32(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint32x2_t test_svluti2_lane_zt_u32(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti2_lane_zt_u32_x2(0, zn, 7);
 }
 
@@ -182,7 +182,7 @@ svuint32x2_t test_svluti2_lane_zt_u32(svuint8_t zn) __arm_streaming __arm_shared
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv8i32.nxv4i32( [[TMP2]],  [[TMP3]], i64 4)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svint32x2_t test_svluti2_lane_zt_s32(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint32x2_t test_svluti2_lane_zt_s32(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti2_lane_zt_s32_x2(0, zn, 7);
 }
 
@@ -204,6 +204,6 @@ svint32x2_t test_svluti2_lane_zt_s32(svuint8_t zn) __arm_streaming __arm_shared_
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv8f32.nxv4f32( [[TMP2]],  [[TMP3]], i64 4)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svfloat32x2_t test_svluti2_lane_zt_f32(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svfloat32x2_t test_svluti2_lane_zt_f32(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti2_lane_zt_f32_x2(0, zn, 7);
 }
diff --git a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_luti2_lane_zt_x4.c b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_luti2_lane_zt_x4.c
index 8c38d829a7f4c..23b2c6cc51283 100644
--- a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_luti2_lane_zt_x4.c
+++ b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_luti2_lane_zt_x4.c
@@ -34,7 +34,7 @@
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv64i8.nxv16i8( [[TMP6]],  [[TMP7]], i64 48)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svuint8x4_t test_svluti2_lane_zt_u8(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint8x4_t test_svluti2_lane_zt_u8(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti2_lane_zt_u8_x4(0, zn, 3);
 }
 
@@ -65,7 +65,7 @@ svuint8x4_t test_svluti2_lane_zt_u8(svuint8_t zn) __arm_streaming __arm_shared_z
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv64i8.nxv16i8( [[TMP6]],  [[TMP7]], i64 48)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svint8x4_t test_svluti2_lane_zt_s8(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint8x4_t test_svluti2_lane_zt_s8(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti2_lane_zt_s8_x4(0, zn, 3);
 }
 
@@ -95,7 +95,7 @@ svint8x4_t test_svluti2_lane_zt_s8(svuint8_t zn) __arm_streaming __arm_shared_za
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv32i16.nxv8i16( [[TMP6]],  [[TMP7]], i64 24)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svuint16x4_t test_svluti2_lane_zt_u16(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint16x4_t test_svluti2_lane_zt_u16(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti2_lane_zt_u16_x4(0, zn, 3);
 }
 
@@ -125,7 +125,7 @@ svuint16x4_t test_svluti2_lane_zt_u16(svuint8_t zn) __arm_streaming __arm_shared
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv32i16.nxv8i16( [[TMP6]],  [[TMP7]], i64 24)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svint16x4_t test_svluti2_lane_zt_s16(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint16x4_t test_svluti2_lane_zt_s16(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti2_lane_zt_s16_x4(0, zn, 3);
 }
 
@@ -155,7 +155,7 @@ svint16x4_t test_svluti2_lane_zt_s16(svuint8_t zn) __arm_streaming __arm_shared_
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv32f16.nxv8f16( [[TMP6]],  [[TMP7]], i64 24)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svfloat16x4_t test_svluti2_lane_zt_f16(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svfloat16x4_t test_svluti2_lane_zt_f16(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti2_lane_zt_f16_x4(0, zn, 3);
 }
 
@@ -185,7 +185,7 @@ svfloat16x4_t test_svluti2_lane_zt_f16(svuint8_t zn) __arm_streaming __arm_share
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv32bf16.nxv8bf16( [[TMP6]],  [[TMP7]], i64 24)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svbfloat16x4_t test_svluti2_lane_zt_bf16(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svbfloat16x4_t test_svluti2_lane_zt_bf16(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti2_lane_zt_bf16_x4(0, zn, 3);
 }
 
@@ -215,7 +215,7 @@ svbfloat16x4_t test_svluti2_lane_zt_bf16(svuint8_t zn) __arm_streaming __arm_sha
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv16i32.nxv4i32( [[TMP6]],  [[TMP7]], i64 12)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svuint32x4_t test_svluti2_lane_zt_u32(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint32x4_t test_svluti2_lane_zt_u32(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti2_lane_zt_u32_x4(0, zn, 3);
 }
 
@@ -245,7 +245,7 @@ svuint32x4_t test_svluti2_lane_zt_u32(svuint8_t zn) __arm_streaming __arm_shared
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv16i32.nxv4i32( [[TMP6]],  [[TMP7]], i64 12)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svint32x4_t test_svluti2_lane_zt_s32(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint32x4_t test_svluti2_lane_zt_s32(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti2_lane_zt_s32_x4(0, zn, 3);
 }
 
@@ -275,6 +275,6 @@ svint32x4_t test_svluti2_lane_zt_s32(svuint8_t zn) __arm_streaming __arm_shared_
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv16f32.nxv4f32( [[TMP6]],  [[TMP7]], i64 12)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svfloat32x4_t test_svluti2_lane_zt_f32(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svfloat32x4_t test_svluti2_lane_zt_f32(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti2_lane_zt_f32_x4(0, zn, 3);
 }
diff --git a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_luti4_lane_zt.c b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_luti4_lane_zt.c
index 9815b0e825b30..c949cee1124ec 100644
--- a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_luti4_lane_zt.c
+++ b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_luti4_lane_zt.c
@@ -19,7 +19,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sme.luti4.lane.zt.nxv16i8(i32 0,  [[ZN:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svuint8_t test_svluti4_lane_zt_u8(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint8_t test_svluti4_lane_zt_u8(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti4_lane_zt_u8(0, zn, 7);
 }
 
@@ -34,7 +34,7 @@ svuint8_t test_svluti4_lane_zt_u8(svuint8_t zn) __arm_streaming __arm_shared_za
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sme.luti4.lane.zt.nxv16i8(i32 0,  [[ZN:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svint8_t test_svluti4_lane_zt_s8(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint8_t test_svluti4_lane_zt_s8(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti4_lane_zt_s8(0, zn, 7);
 }
 
@@ -48,7 +48,7 @@ svint8_t test_svluti4_lane_zt_s8(svuint8_t zn) __arm_streaming __arm_shared_za _
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sme.luti4.lane.zt.nxv8i16(i32 0,  [[ZN:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svuint16_t test_svluti4_lane_zt_u16(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint16_t test_svluti4_lane_zt_u16(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti4_lane_zt_u16(0, zn, 7);
 }
 
@@ -62,7 +62,7 @@ svuint16_t test_svluti4_lane_zt_u16(svuint8_t zn) __arm_streaming __arm_shared_z
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sme.luti4.lane.zt.nxv8i16(i32 0,  [[ZN:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svint16_t test_svluti4_lane_zt_s16(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint16_t test_svluti4_lane_zt_s16(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti4_lane_zt_s16(0, zn, 7);
 }
 
@@ -76,7 +76,7 @@ svint16_t test_svluti4_lane_zt_s16(svuint8_t zn) __arm_streaming __arm_shared_za
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sme.luti4.lane.zt.nxv8f16(i32 0,  [[ZN:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svfloat16_t test_svluti4_lane_zt_f16(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svfloat16_t test_svluti4_lane_zt_f16(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti4_lane_zt_f16(0, zn, 7);
 }
 
@@ -90,7 +90,7 @@ svfloat16_t test_svluti4_lane_zt_f16(svuint8_t zn) __arm_streaming __arm_shared_
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sme.luti4.lane.zt.nxv8bf16(i32 0,  [[ZN:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svbfloat16_t test_svluti4_lane_zt_bf16(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svbfloat16_t test_svluti4_lane_zt_bf16(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti4_lane_zt_bf16(0, zn, 7);
 }
 
@@ -104,7 +104,7 @@ svbfloat16_t test_svluti4_lane_zt_bf16(svuint8_t zn) __arm_streaming __arm_share
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sme.luti4.lane.zt.nxv4i32(i32 0,  [[ZN:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svuint32_t test_svluti4_lane_zt_u32(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint32_t test_svluti4_lane_zt_u32(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti4_lane_zt_u32(0, zn, 7);
 }
 
@@ -118,7 +118,7 @@ svuint32_t test_svluti4_lane_zt_u32(svuint8_t zn) __arm_streaming __arm_shared_z
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sme.luti4.lane.zt.nxv4i32(i32 0,  [[ZN:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svint32_t test_svluti4_lane_zt_s32(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint32_t test_svluti4_lane_zt_s32(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti4_lane_zt_s32(0, zn, 7);
 }
 
@@ -132,6 +132,6 @@ svint32_t test_svluti4_lane_zt_s32(svuint8_t zn) __arm_streaming __arm_shared_za
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sme.luti4.lane.zt.nxv4f32(i32 0,  [[ZN:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svfloat32_t test_svluti4_lane_zt_f32(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svfloat32_t test_svluti4_lane_zt_f32(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti4_lane_zt_f32(0, zn, 7);
 }
diff --git a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_luti4_lane_zt_x2.c b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_luti4_lane_zt_x2.c
index 4c181dd9123c5..61affd86d9119 100644
--- a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_luti4_lane_zt_x2.c
+++ b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_luti4_lane_zt_x2.c
@@ -26,7 +26,7 @@
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv32i8.nxv16i8( [[TMP2]],  [[TMP3]], i64 16)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svuint8x2_t test_svluti4_lane_zt_u8(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint8x2_t test_svluti4_lane_zt_u8(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti4_lane_zt_u8_x2(0, zn, 3);
 }
 
@@ -49,7 +49,7 @@ svuint8x2_t test_svluti4_lane_zt_u8(svuint8_t zn) __arm_streaming __arm_shared_z
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv32i8.nxv16i8( [[TMP2]],  [[TMP3]], i64 16)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svint8x2_t test_svluti4_lane_zt_s8(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint8x2_t test_svluti4_lane_zt_s8(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti4_lane_zt_s8_x2(0, zn, 3);
 }
 
@@ -71,7 +71,7 @@ svint8x2_t test_svluti4_lane_zt_s8(svuint8_t zn) __arm_streaming __arm_shared_za
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv16i16.nxv8i16( [[TMP2]],  [[TMP3]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svuint16x2_t test_svluti4_lane_zt_u16(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint16x2_t test_svluti4_lane_zt_u16(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti4_lane_zt_u16_x2(0, zn, 3);
 }
 
@@ -94,7 +94,7 @@ svuint16x2_t test_svluti4_lane_zt_u16(svuint8_t zn) __arm_streaming __arm_shared
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv16i16.nxv8i16( [[TMP2]],  [[TMP3]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svint16x2_t test_svluti4_lane_zt_s16(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint16x2_t test_svluti4_lane_zt_s16(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti4_lane_zt_s16_x2(0, zn, 3);
 }
 
@@ -116,7 +116,7 @@ svint16x2_t test_svluti4_lane_zt_s16(svuint8_t zn) __arm_streaming __arm_shared_
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv16f16.nxv8f16( [[TMP2]],  [[TMP3]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svfloat16x2_t test_svluti4_lane_zt_f16(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svfloat16x2_t test_svluti4_lane_zt_f16(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti4_lane_zt_f16_x2(0, zn, 3);
 }
 
@@ -138,7 +138,7 @@ svfloat16x2_t test_svluti4_lane_zt_f16(svuint8_t zn) __arm_streaming __arm_share
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv16bf16.nxv8bf16( [[TMP2]],  [[TMP3]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svbfloat16x2_t test_svluti4_lane_zt_bf16(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svbfloat16x2_t test_svluti4_lane_zt_bf16(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti4_lane_zt_bf16_x2(0, zn, 3);
 }
 
@@ -160,7 +160,7 @@ svbfloat16x2_t test_svluti4_lane_zt_bf16(svuint8_t zn) __arm_streaming __arm_sha
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv8i32.nxv4i32( [[TMP2]],  [[TMP3]], i64 4)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svuint32x2_t test_svluti4_lane_zt_u32(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint32x2_t test_svluti4_lane_zt_u32(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti4_lane_zt_u32_x2(0, zn, 3);
 }
 
@@ -182,7 +182,7 @@ svuint32x2_t test_svluti4_lane_zt_u32(svuint8_t zn) __arm_streaming __arm_shared
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv8i32.nxv4i32( [[TMP2]],  [[TMP3]], i64 4)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svint32x2_t test_svluti4_lane_zt_s32(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint32x2_t test_svluti4_lane_zt_s32(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti4_lane_zt_s32_x2(0, zn, 3);
 }
 
@@ -204,6 +204,6 @@ svint32x2_t test_svluti4_lane_zt_s32(svuint8_t zn) __arm_streaming __arm_shared_
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv8f32.nxv4f32( [[TMP2]],  [[TMP3]], i64 4)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svfloat32x2_t test_svluti4_lane_zt_f32(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svfloat32x2_t test_svluti4_lane_zt_f32(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti4_lane_zt_f32_x2(0, zn, 3);
 }
diff --git a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_luti4_lane_zt_x4.c b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_luti4_lane_zt_x4.c
index 9baccef888d58..7b478aa2b6ad4 100644
--- a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_luti4_lane_zt_x4.c
+++ b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_luti4_lane_zt_x4.c
@@ -36,7 +36,7 @@
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv32i16.nxv8i16( [[TMP6]],  [[TMP7]], i64 24)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svuint16x4_t test_svluti4_lane_zt_u16(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint16x4_t test_svluti4_lane_zt_u16(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti4_lane_zt_u16_x4(0, zn, 1);
 }
 
@@ -68,7 +68,7 @@ svuint16x4_t test_svluti4_lane_zt_u16(svuint8_t zn) __arm_streaming __arm_shared
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv32f16.nxv8f16( [[TMP6]],  [[TMP7]], i64 24)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svfloat16x4_t test_svluti4_lane_zt_f16(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svfloat16x4_t test_svluti4_lane_zt_f16(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti4_lane_zt_f16_x4(0, zn, 1);
 }
 
@@ -100,7 +100,7 @@ svfloat16x4_t test_svluti4_lane_zt_f16(svuint8_t zn) __arm_streaming __arm_share
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv32bf16.nxv8bf16( [[TMP6]],  [[TMP7]], i64 24)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svbfloat16x4_t test_svluti4_lane_zt_bf16(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svbfloat16x4_t test_svluti4_lane_zt_bf16(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti4_lane_zt_bf16_x4(0, zn, 1);
 }
 
@@ -132,7 +132,7 @@ svbfloat16x4_t test_svluti4_lane_zt_bf16(svuint8_t zn) __arm_streaming __arm_sha
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv32i16.nxv8i16( [[TMP6]],  [[TMP7]], i64 24)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svint16x4_t test_svluti4_lane_zt_s16(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint16x4_t test_svluti4_lane_zt_s16(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti4_lane_zt_s16_x4(0, zn, 1);
 }
 
@@ -164,7 +164,7 @@ svint16x4_t test_svluti4_lane_zt_s16(svuint8_t zn) __arm_streaming __arm_shared_
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv16i32.nxv4i32( [[TMP6]],  [[TMP7]], i64 12)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svuint32x4_t test_svluti4_lane_zt_u32(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint32x4_t test_svluti4_lane_zt_u32(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti4_lane_zt_u32_x4(0, zn, 1);
 }
 
@@ -196,7 +196,7 @@ svuint32x4_t test_svluti4_lane_zt_u32(svuint8_t zn) __arm_streaming __arm_shared
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv16i32.nxv4i32( [[TMP6]],  [[TMP7]], i64 12)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svint32x4_t test_svluti4_lane_zt_s32(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint32x4_t test_svluti4_lane_zt_s32(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti4_lane_zt_s32_x4(0, zn, 1);
 }
 
@@ -228,6 +228,6 @@ svint32x4_t test_svluti4_lane_zt_s32(svuint8_t zn) __arm_streaming __arm_shared_
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv16f32.nxv4f32( [[TMP6]],  [[TMP7]], i64 12)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svfloat32x4_t test_svluti4_lane_zt_f32(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+svfloat32x4_t test_svluti4_lane_zt_f32(svuint8_t zn) __arm_streaming __arm_in("za") {
   return svluti4_lane_zt_f32_x4(0, zn, 1);
 }
diff --git a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_mla.c b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_mla.c
index f52edd9888daa..18c4b77ac3292 100644
--- a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_mla.c
+++ b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_mla.c
@@ -35,7 +35,7 @@
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmla.vg1x2.nxv4f32(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla2_f32(uint32_t slice_base, svfloat32x2_t zn, svfloat32x2_t zm) __arm_streaming __arm_shared_za {
+void test_svmla2_f32(uint32_t slice_base, svfloat32x2_t zn, svfloat32x2_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svmla_za32,_f32,_vg1x2,,)(slice_base, zn, zm);
 }
 
@@ -65,7 +65,7 @@ void test_svmla2_f32(uint32_t slice_base, svfloat32x2_t zn, svfloat32x2_t zm) __
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmla.vg1x4.nxv4f32(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[TMP4]],  [[TMP5]],  [[TMP6]],  [[TMP7]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla4_f32(uint32_t slice_base, svfloat32x4_t zn, svfloat32x4_t zm) __arm_streaming __arm_shared_za {
+void test_svmla4_f32(uint32_t slice_base, svfloat32x4_t zn, svfloat32x4_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svmla_za32,_f32,_vg1x4,,)(slice_base, zn, zm);
 }
 
@@ -85,8 +85,8 @@ void test_svmla4_f32(uint32_t slice_base, svfloat32x4_t zn, svfloat32x4_t zm) __
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmla.single.vg1x2.nxv4f32(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_single2_f32(uint32_t slice_base, svfloat32x2_t zn, svfloat32_t zm) __arm_streaming __arm_shared_za {
-  SVE_ACLE_FUNC(svmla_single_za32,,_f32,,_vg1x2)(slice_base, zn, zm);
+void test_svmla_single2_f32(uint32_t slice_base, svfloat32x2_t zn, svfloat32_t zm) __arm_streaming __arm_inout("za") {
+  SVE_ACLE_FUNC(svmla,_single,_za32,_f32,_vg1x2)(slice_base, zn, zm);
 }
 
 // CHECK-LABEL: @test_svmla_single4_f32(
@@ -107,8 +107,8 @@ void test_svmla_single2_f32(uint32_t slice_base, svfloat32x2_t zn, svfloat32_t z
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmla.single.vg1x4.nxv4f32(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_single4_f32(uint32_t slice_base, svfloat32x4_t zn, svfloat32_t zm) __arm_streaming __arm_shared_za {
-  SVE_ACLE_FUNC(svmla_single_za32,,_f32,,_vg1x4)(slice_base, zn, zm);
+void test_svmla_single4_f32(uint32_t slice_base, svfloat32x4_t zn, svfloat32_t zm) __arm_streaming __arm_inout("za") {
+  SVE_ACLE_FUNC(svmla,_single,_za32,_f32,_vg1x4)(slice_base, zn, zm);
 }
 
 //
@@ -127,7 +127,7 @@ void test_svmla_single4_f32(uint32_t slice_base, svfloat32x4_t zn, svfloat32_t z
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmla.lane.vg1x2.nxv4f32(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]], i32 3)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_lane2_f32(uint32_t slice_base, svfloat32x2_t zn, svfloat32_t zm) __arm_streaming __arm_shared_za {
+void test_svmla_lane2_f32(uint32_t slice_base, svfloat32x2_t zn, svfloat32_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svmla_lane_za32,_f32,_vg1x2,,)(slice_base, zn, zm, 3);
 }
 
@@ -149,7 +149,7 @@ void test_svmla_lane2_f32(uint32_t slice_base, svfloat32x2_t zn, svfloat32_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmla.lane.vg1x4.nxv4f32(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]], i32 3)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_lane4_f32(uint32_t slice_base, svfloat32x4_t zn, svfloat32_t zm) __arm_streaming __arm_shared_za {
+void test_svmla_lane4_f32(uint32_t slice_base, svfloat32x4_t zn, svfloat32_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svmla_lane_za32,_f32,_vg1x4,,)(slice_base, zn, zm, 3);
 }
 
@@ -173,7 +173,7 @@ void test_svmla_lane4_f32(uint32_t slice_base, svfloat32x4_t zn, svfloat32_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmla.vg1x2.nxv2f64(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla2_f64(uint32_t slice_base, svfloat64x2_t zn, svfloat64x2_t zm) __arm_streaming __arm_shared_za {
+void test_svmla2_f64(uint32_t slice_base, svfloat64x2_t zn, svfloat64x2_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svmla_za64,_f64,_vg1x2,,)(slice_base, zn, zm);
 }
 
@@ -203,7 +203,7 @@ void test_svmla2_f64(uint32_t slice_base, svfloat64x2_t zn, svfloat64x2_t zm) __
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmla.vg1x4.nxv2f64(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[TMP4]],  [[TMP5]],  [[TMP6]],  [[TMP7]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla4_f64(uint32_t slice_base, svfloat64x4_t zn, svfloat64x4_t zm) __arm_streaming __arm_shared_za {
+void test_svmla4_f64(uint32_t slice_base, svfloat64x4_t zn, svfloat64x4_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svmla_za64,_f64,_vg1x4,,)(slice_base, zn, zm);
 }
 
@@ -223,8 +223,8 @@ void test_svmla4_f64(uint32_t slice_base, svfloat64x4_t zn, svfloat64x4_t zm) __
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmla.single.vg1x2.nxv2f64(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_single2_f64(uint32_t slice_base, svfloat64x2_t zn, svfloat64_t zm) __arm_streaming __arm_shared_za {
-  SVE_ACLE_FUNC(svmla_single_za64,,_f64,,_vg1x2)(slice_base, zn, zm);
+void test_svmla_single2_f64(uint32_t slice_base, svfloat64x2_t zn, svfloat64_t zm) __arm_streaming __arm_inout("za") {
+  SVE_ACLE_FUNC(svmla,_single,_za64,_f64,_vg1x2)(slice_base, zn, zm);
 }
 
 // CHECK-LABEL: @test_svmla_single4_f64(
@@ -245,8 +245,8 @@ void test_svmla_single2_f64(uint32_t slice_base, svfloat64x2_t zn, svfloat64_t z
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmla.single.vg1x4.nxv2f64(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_single4_f64(uint32_t slice_base, svfloat64x4_t zn, svfloat64_t zm) __arm_streaming __arm_shared_za {
-  SVE_ACLE_FUNC(svmla_single_za64,,_f64,,_vg1x4)(slice_base, zn, zm);
+void test_svmla_single4_f64(uint32_t slice_base, svfloat64x4_t zn, svfloat64_t zm) __arm_streaming __arm_inout("za") {
+  SVE_ACLE_FUNC(svmla,_single,_za64,_f64,_vg1x4)(slice_base, zn, zm);
 }
 
 //
@@ -265,7 +265,7 @@ void test_svmla_single4_f64(uint32_t slice_base, svfloat64x4_t zn, svfloat64_t z
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmla.lane.vg1x2.nxv2f64(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]], i32 1)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_lane2_f64(uint32_t slice_base, svfloat64x2_t zn, svfloat64_t zm) __arm_streaming __arm_shared_za {
+void test_svmla_lane2_f64(uint32_t slice_base, svfloat64x2_t zn, svfloat64_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svmla_lane_za64,_f64,_vg1x2,,)(slice_base, zn, zm, 1);
 }
 
@@ -287,6 +287,6 @@ void test_svmla_lane2_f64(uint32_t slice_base, svfloat64x2_t zn, svfloat64_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmla.lane.vg1x4.nxv2f64(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]], i32 1)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_lane4_f64(uint32_t slice_base, svfloat64x4_t zn, svfloat64_t zm) __arm_streaming __arm_shared_za {
+void test_svmla_lane4_f64(uint32_t slice_base, svfloat64x4_t zn, svfloat64_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svmla_lane_za64,_f64,_vg1x4,,)(slice_base, zn, zm, 1);
 }
diff --git a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_mlal.c b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_mlal.c
index 834ade7535076..4b785cb86df64 100644
--- a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_mlal.c
+++ b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_mlal.c
@@ -35,7 +35,7 @@
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmlal.vg2x2.nxv8f16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla2_f16(uint32_t slice_base, svfloat16x2_t zn, svfloat16x2_t zm) __arm_streaming __arm_shared_za
+void test_svmla2_f16(uint32_t slice_base, svfloat16x2_t zn, svfloat16x2_t zm) __arm_streaming __arm_inout("za")
 {
    SVE_ACLE_FUNC(svmla_za32,_f16,_vg2x2,,)(slice_base, zn, zm);
 }
@@ -58,7 +58,7 @@ void test_svmla2_f16(uint32_t slice_base, svfloat16x2_t zn, svfloat16x2_t zm) __
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmlal.vg2x2.nxv8bf16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla2_bf16(uint32_t slice_base, svbfloat16x2_t zn, svbfloat16x2_t zm) __arm_streaming __arm_shared_za
+void test_svmla2_bf16(uint32_t slice_base, svbfloat16x2_t zn, svbfloat16x2_t zm) __arm_streaming __arm_inout("za")
 {
    SVE_ACLE_FUNC(svmla_za32,_bf16,_vg2x2,,)(slice_base, zn, zm);
 }
@@ -81,7 +81,7 @@ void test_svmla2_bf16(uint32_t slice_base, svbfloat16x2_t zn, svbfloat16x2_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umlal.vg2x2.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16x2_t zm) __arm_streaming __arm_shared_za
+void test_svmla2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16x2_t zm) __arm_streaming __arm_inout("za")
 {
    SVE_ACLE_FUNC(svmla_za32,_u16,_vg2x2,,)(slice_base, zn, zm);
 }
@@ -104,7 +104,7 @@ void test_svmla2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16x2_t zm) __ar
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smlal.vg2x2.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla2_s16(uint32_t slice_base, svint16x2_t zn, svint16x2_t zm) __arm_streaming __arm_shared_za
+void test_svmla2_s16(uint32_t slice_base, svint16x2_t zn, svint16x2_t zm) __arm_streaming __arm_inout("za")
 {
    SVE_ACLE_FUNC(svmla_za32,_s16,_vg2x2,,)(slice_base, zn, zm);
 }
@@ -135,7 +135,7 @@ void test_svmla2_s16(uint32_t slice_base, svint16x2_t zn, svint16x2_t zm) __arm_
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmlal.vg2x4.nxv8f16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[TMP4]],  [[TMP5]],  [[TMP6]],  [[TMP7]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla4_f16(uint32_t slice_base, svfloat16x4_t zn, svfloat16x4_t zm) __arm_streaming __arm_shared_za
+void test_svmla4_f16(uint32_t slice_base, svfloat16x4_t zn, svfloat16x4_t zm) __arm_streaming __arm_inout("za")
 {
    SVE_ACLE_FUNC(svmla_za32,_f16,_vg2x4,,)(slice_base, zn, zm);
 }
@@ -166,7 +166,7 @@ void test_svmla4_f16(uint32_t slice_base, svfloat16x4_t zn, svfloat16x4_t zm) __
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmlal.vg2x4.nxv8bf16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[TMP4]],  [[TMP5]],  [[TMP6]],  [[TMP7]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla4_bf16(uint32_t slice_base, svbfloat16x4_t zn, svbfloat16x4_t zm) __arm_streaming __arm_shared_za
+void test_svmla4_bf16(uint32_t slice_base, svbfloat16x4_t zn, svbfloat16x4_t zm) __arm_streaming __arm_inout("za")
 {
    SVE_ACLE_FUNC(svmla_za32,_bf16,_vg2x4,,)(slice_base, zn, zm);
 }
@@ -197,7 +197,7 @@ void test_svmla4_bf16(uint32_t slice_base, svbfloat16x4_t zn, svbfloat16x4_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umlal.vg2x4.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[TMP4]],  [[TMP5]],  [[TMP6]],  [[TMP7]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16x4_t zm) __arm_streaming __arm_shared_za
+void test_svmla4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16x4_t zm) __arm_streaming __arm_inout("za")
 {
    SVE_ACLE_FUNC(svmla_za32,_u16,_vg2x4,,)(slice_base, zn, zm);
 }
@@ -228,7 +228,7 @@ void test_svmla4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16x4_t zm) __ar
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smlal.vg2x4.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[TMP4]],  [[TMP5]],  [[TMP6]],  [[TMP7]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla4_s16(uint32_t slice_base, svint16x4_t zn, svint16x4_t zm) __arm_streaming __arm_shared_za
+void test_svmla4_s16(uint32_t slice_base, svint16x4_t zn, svint16x4_t zm) __arm_streaming __arm_inout("za")
 {
    SVE_ACLE_FUNC(svmla_za32,_s16,_vg2x4,,)(slice_base, zn, zm);
 }
@@ -245,7 +245,7 @@ void test_svmla4_s16(uint32_t slice_base, svint16x4_t zn, svint16x4_t zm) __arm_
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmlal.single.vg2x1.nxv8f16(i32 [[SLICE_BASE:%.*]],  [[ZN:%.*]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_single1_f16(uint32_t slice_base, svfloat16_t zn, svfloat16_t zm) __arm_streaming __arm_shared_za
+void test_svmla_single1_f16(uint32_t slice_base, svfloat16_t zn, svfloat16_t zm) __arm_streaming __arm_inout("za")
 {
    SVE_ACLE_FUNC(svmla_za32,_f16,_vg2x1,,)(slice_base, zn, zm);
 }
@@ -260,7 +260,7 @@ void test_svmla_single1_f16(uint32_t slice_base, svfloat16_t zn, svfloat16_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmlal.single.vg2x1.nxv8bf16(i32 [[SLICE_BASE:%.*]],  [[ZN:%.*]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_single1_bf16(uint32_t slice_base, svbfloat16_t zn, svbfloat16_t zm) __arm_streaming __arm_shared_za
+void test_svmla_single1_bf16(uint32_t slice_base, svbfloat16_t zn, svbfloat16_t zm) __arm_streaming __arm_inout("za")
 {
    SVE_ACLE_FUNC(svmla_za32,_bf16,_vg2x1,,)(slice_base, zn, zm);
 }
@@ -275,7 +275,7 @@ void test_svmla_single1_bf16(uint32_t slice_base, svbfloat16_t zn, svbfloat16_t
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umlal.single.vg2x1.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[ZN:%.*]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_single1_u16(uint32_t slice_base, svuint16_t zn, svuint16_t zm) __arm_streaming __arm_shared_za
+void test_svmla_single1_u16(uint32_t slice_base, svuint16_t zn, svuint16_t zm) __arm_streaming __arm_inout("za")
 {
    SVE_ACLE_FUNC(svmla_za32,_u16,_vg2x1,,)(slice_base, zn, zm);
 }
@@ -290,7 +290,7 @@ void test_svmla_single1_u16(uint32_t slice_base, svuint16_t zn, svuint16_t zm) _
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smlal.single.vg2x1.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[ZN:%.*]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_single1_s16(uint32_t slice_base, svint16_t zn, svint16_t zm) __arm_streaming __arm_shared_za
+void test_svmla_single1_s16(uint32_t slice_base, svint16_t zn, svint16_t zm) __arm_streaming __arm_inout("za")
 {
    SVE_ACLE_FUNC(svmla_za32,_s16,_vg2x1,,)(slice_base, zn, zm);
 }
@@ -309,9 +309,9 @@ void test_svmla_single1_s16(uint32_t slice_base, svint16_t zn, svint16_t zm) __a
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmlal.single.vg2x2.nxv8f16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_single2_f16(uint32_t slice_base, svfloat16x2_t zn, svfloat16_t zm) __arm_streaming __arm_shared_za
+void test_svmla_single2_f16(uint32_t slice_base, svfloat16x2_t zn, svfloat16_t zm) __arm_streaming __arm_inout("za")
 {
-   SVE_ACLE_FUNC(svmla_single_za32,,_f16,,_vg2x2)(slice_base, zn, zm);
+   SVE_ACLE_FUNC(svmla,_single,_za32,_f16,_vg2x2)(slice_base, zn, zm);
 }
 
 // CHECK-LABEL: @test_svmla_single2_bf16(
@@ -328,9 +328,9 @@ void test_svmla_single2_f16(uint32_t slice_base, svfloat16x2_t zn, svfloat16_t z
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmlal.single.vg2x2.nxv8bf16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_single2_bf16(uint32_t slice_base, svbfloat16x2_t zn, svbfloat16_t zm) __arm_streaming __arm_shared_za
+void test_svmla_single2_bf16(uint32_t slice_base, svbfloat16x2_t zn, svbfloat16_t zm) __arm_streaming __arm_inout("za")
 {
-   SVE_ACLE_FUNC(svmla_single_za32,,_bf16,,_vg2x2)(slice_base, zn, zm);
+   SVE_ACLE_FUNC(svmla,_single,_za32,_bf16,_vg2x2)(slice_base, zn, zm);
 }
 
 // CHECK-LABEL: @test_svmla_single2_u16(
@@ -347,9 +347,9 @@ void test_svmla_single2_bf16(uint32_t slice_base, svbfloat16x2_t zn, svbfloat16_
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umlal.single.vg2x2.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_single2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16_t zm) __arm_streaming __arm_shared_za
+void test_svmla_single2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16_t zm) __arm_streaming __arm_inout("za")
 {
-   SVE_ACLE_FUNC(svmla_single_za32,,_u16,,_vg2x2)(slice_base, zn, zm);
+   SVE_ACLE_FUNC(svmla,_single,_za32,_u16,_vg2x2)(slice_base, zn, zm);
 }
 
 // CHECK-LABEL: @test_svmla_single2_s16(
@@ -366,9 +366,9 @@ void test_svmla_single2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smlal.single.vg2x2.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_single2_s16(uint32_t slice_base, svint16x2_t zn, svint16_t zm) __arm_streaming __arm_shared_za
+void test_svmla_single2_s16(uint32_t slice_base, svint16x2_t zn, svint16_t zm) __arm_streaming __arm_inout("za")
 {
-   SVE_ACLE_FUNC(svmla_single_za32,,_s16,,_vg2x2)(slice_base, zn, zm);
+   SVE_ACLE_FUNC(svmla,_single,_za32,_s16,_vg2x2)(slice_base, zn, zm);
 }
 
 // CHECK-LABEL: @test_svmla_single4_f16(
@@ -389,9 +389,9 @@ void test_svmla_single2_s16(uint32_t slice_base, svint16x2_t zn, svint16_t zm) _
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmlal.single.vg2x4.nxv8f16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_single4_f16(uint32_t slice_base, svfloat16x4_t zn, svfloat16_t zm) __arm_streaming __arm_shared_za
+void test_svmla_single4_f16(uint32_t slice_base, svfloat16x4_t zn, svfloat16_t zm) __arm_streaming __arm_inout("za")
 {
-   SVE_ACLE_FUNC(svmla_single_za32,,_f16,,_vg2x4)(slice_base, zn, zm);
+   SVE_ACLE_FUNC(svmla,_single,_za32,_f16,_vg2x4)(slice_base, zn, zm);
 }
 
 // CHECK-LABEL: @test_svmla_single4_bf16(
@@ -412,9 +412,9 @@ void test_svmla_single4_f16(uint32_t slice_base, svfloat16x4_t zn, svfloat16_t z
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmlal.single.vg2x4.nxv8bf16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_single4_bf16(uint32_t slice_base, svbfloat16x4_t zn, svbfloat16_t zm) __arm_streaming __arm_shared_za
+void test_svmla_single4_bf16(uint32_t slice_base, svbfloat16x4_t zn, svbfloat16_t zm) __arm_streaming __arm_inout("za")
 {
-   SVE_ACLE_FUNC(svmla_single_za32,,_bf16,,_vg2x4)(slice_base, zn, zm);
+   SVE_ACLE_FUNC(svmla,_single,_za32,_bf16,_vg2x4)(slice_base, zn, zm);
 }
 
 // CHECK-LABEL: @test_svmla_single4_u16(
@@ -435,9 +435,9 @@ void test_svmla_single4_bf16(uint32_t slice_base, svbfloat16x4_t zn, svbfloat16_
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umlal.single.vg2x4.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_single4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16_t zm) __arm_streaming __arm_shared_za
+void test_svmla_single4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16_t zm) __arm_streaming __arm_inout("za")
 {
-   SVE_ACLE_FUNC(svmla_single_za32,,_u16,,_vg2x4)(slice_base, zn, zm);
+   SVE_ACLE_FUNC(svmla,_single,_za32,_u16,_vg2x4)(slice_base, zn, zm);
 }
 
 // CHECK-LABEL: @test_svmla_single4_s16(
@@ -458,9 +458,9 @@ void test_svmla_single4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smlal.single.vg2x4.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_single4_s16(uint32_t slice_base, svint16x4_t zn, svint16_t zm) __arm_streaming __arm_shared_za
+void test_svmla_single4_s16(uint32_t slice_base, svint16x4_t zn, svint16_t zm) __arm_streaming __arm_inout("za")
 {
-   SVE_ACLE_FUNC(svmla_single_za32,,_s16,,_vg2x4)(slice_base, zn, zm);
+   SVE_ACLE_FUNC(svmla,_single,_za32,_s16,_vg2x4)(slice_base, zn, zm);
 }
 
 //
@@ -477,9 +477,9 @@ void test_svmla_single4_s16(uint32_t slice_base, svint16x4_t zn, svint16_t zm) _
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmlal.lane.vg2x1.nxv8f16(i32 [[SLICE_BASE:%.*]],  [[ZN:%.*]],  [[ZM:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_lane1_f16(uint32_t slice_base, svfloat16_t zn, svfloat16_t zm) __arm_streaming __arm_shared_za
+void test_svmla_lane1_f16(uint32_t slice_base, svfloat16_t zn, svfloat16_t zm) __arm_streaming __arm_inout("za")
 {
-   SVE_ACLE_FUNC(svmla_lane_za32,,,_f16,_vg2x1)(slice_base, zn, zm, 7);
+   SVE_ACLE_FUNC(svmla_lane_za32,_f16,_vg2x1,,)(slice_base, zn, zm, 7);
 }
 
 // CHECK-LABEL: @test_svmla_lane1_bf16(
@@ -492,9 +492,9 @@ void test_svmla_lane1_f16(uint32_t slice_base, svfloat16_t zn, svfloat16_t zm) _
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmlal.lane.vg2x1.nxv8bf16(i32 [[SLICE_BASE:%.*]],  [[ZN:%.*]],  [[ZM:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_lane1_bf16(uint32_t slice_base, svbfloat16_t zn, svbfloat16_t zm) __arm_streaming __arm_shared_za
+void test_svmla_lane1_bf16(uint32_t slice_base, svbfloat16_t zn, svbfloat16_t zm) __arm_streaming __arm_inout("za")
 {
-   SVE_ACLE_FUNC(svmla_lane_za32,,,_bf16,_vg2x1)(slice_base, zn, zm, 7);
+   SVE_ACLE_FUNC(svmla_lane_za32,_bf16,_vg2x1,,)(slice_base, zn, zm, 7);
 }
 
 // CHECK-LABEL: @test_svmla_lane1_u16(
@@ -507,9 +507,9 @@ void test_svmla_lane1_bf16(uint32_t slice_base, svbfloat16_t zn, svbfloat16_t zm
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umlal.lane.vg2x1.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[ZN:%.*]],  [[ZM:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_lane1_u16(uint32_t slice_base, svuint16_t zn, svuint16_t zm) __arm_streaming __arm_shared_za
+void test_svmla_lane1_u16(uint32_t slice_base, svuint16_t zn, svuint16_t zm) __arm_streaming __arm_inout("za")
 {
-   SVE_ACLE_FUNC(svmla_lane_za32,,,_u16,_vg2x1)(slice_base, zn, zm, 7);
+   SVE_ACLE_FUNC(svmla_lane_za32,_u16,_vg2x1,,)(slice_base, zn, zm, 7);
 }
 
 // CHECK-LABEL: @test_svmla_lane1_s16(
@@ -522,9 +522,9 @@ void test_svmla_lane1_u16(uint32_t slice_base, svuint16_t zn, svuint16_t zm) __a
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smlal.lane.vg2x1.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[ZN:%.*]],  [[ZM:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_lane1_s16(uint32_t slice_base, svint16_t zn, svint16_t zm) __arm_streaming __arm_shared_za
+void test_svmla_lane1_s16(uint32_t slice_base, svint16_t zn, svint16_t zm) __arm_streaming __arm_inout("za")
 {
-   SVE_ACLE_FUNC(svmla_lane_za32,,,_s16,_vg2x1)(slice_base, zn, zm, 7);
+   SVE_ACLE_FUNC(svmla_lane_za32,_s16,_vg2x1,,)(slice_base, zn, zm, 7);
 }
 
 // CHECK-LABEL: @test_svmla_lane2_f16(
@@ -541,9 +541,9 @@ void test_svmla_lane1_s16(uint32_t slice_base, svint16_t zn, svint16_t zm) __arm
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmlal.lane.vg2x2.nxv8f16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_lane2_f16(uint32_t slice_base, svfloat16x2_t zn, svfloat16_t zm) __arm_streaming __arm_shared_za
+void test_svmla_lane2_f16(uint32_t slice_base, svfloat16x2_t zn, svfloat16_t zm) __arm_streaming __arm_inout("za")
 {
-   SVE_ACLE_FUNC(svmla_lane_za32,,,_f16,_vg2x2)(slice_base, zn, zm, 7);
+   SVE_ACLE_FUNC(svmla_lane_za32,_f16,_vg2x2,,)(slice_base, zn, zm, 7);
 }
 
 // CHECK-LABEL: @test_svmla_lane2_bf16(
@@ -560,9 +560,9 @@ void test_svmla_lane2_f16(uint32_t slice_base, svfloat16x2_t zn, svfloat16_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmlal.lane.vg2x2.nxv8bf16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_lane2_bf16(uint32_t slice_base, svbfloat16x2_t zn, svbfloat16_t zm) __arm_streaming __arm_shared_za
+void test_svmla_lane2_bf16(uint32_t slice_base, svbfloat16x2_t zn, svbfloat16_t zm) __arm_streaming __arm_inout("za")
 {
-   SVE_ACLE_FUNC(svmla_lane_za32,,,_bf16,_vg2x2)(slice_base, zn, zm, 7);
+   SVE_ACLE_FUNC(svmla_lane_za32,_bf16,_vg2x2,,)(slice_base, zn, zm, 7);
 }
 
 // CHECK-LABEL: @test_svmla_lane2_u16(
@@ -579,9 +579,9 @@ void test_svmla_lane2_bf16(uint32_t slice_base, svbfloat16x2_t zn, svbfloat16_t
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umlal.lane.vg2x2.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_lane2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16_t zm) __arm_streaming __arm_shared_za
+void test_svmla_lane2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16_t zm) __arm_streaming __arm_inout("za")
 {
-   SVE_ACLE_FUNC(svmla_lane_za32,,,_u16,_vg2x2)(slice_base, zn, zm, 7);
+   SVE_ACLE_FUNC(svmla_lane_za32,_u16,_vg2x2,,)(slice_base, zn, zm, 7);
 }
 
 // CHECK-LABEL: @test_svmla_lane2_s16(
@@ -598,9 +598,9 @@ void test_svmla_lane2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16_t zm) _
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smlal.lane.vg2x2.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_lane2_s16(uint32_t slice_base, svint16x2_t zn, svint16_t zm) __arm_streaming __arm_shared_za
+void test_svmla_lane2_s16(uint32_t slice_base, svint16x2_t zn, svint16_t zm) __arm_streaming __arm_inout("za")
 {
-   SVE_ACLE_FUNC(svmla_lane_za32,,,_s16,_vg2x2)(slice_base, zn, zm, 7);
+   SVE_ACLE_FUNC(svmla_lane_za32,_s16,_vg2x2,,)(slice_base, zn, zm, 7);
 }
 
 // CHECK-LABEL: @test_svmla_lane4_f16(
@@ -621,9 +621,9 @@ void test_svmla_lane2_s16(uint32_t slice_base, svint16x2_t zn, svint16_t zm) __a
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmlal.lane.vg2x4.nxv8f16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_lane4_f16(uint32_t slice_base, svfloat16x4_t zn, svfloat16_t zm) __arm_streaming __arm_shared_za
+void test_svmla_lane4_f16(uint32_t slice_base, svfloat16x4_t zn, svfloat16_t zm) __arm_streaming __arm_inout("za")
 {
-   SVE_ACLE_FUNC(svmla_lane_za32,,,_f16,_vg2x4)(slice_base, zn, zm, 7);
+   SVE_ACLE_FUNC(svmla_lane_za32,_f16,_vg2x4,,)(slice_base, zn, zm, 7);
 }
 
 // CHECK-LABEL: @test_svmla_lane4_bf16(
@@ -644,9 +644,9 @@ void test_svmla_lane4_f16(uint32_t slice_base, svfloat16x4_t zn, svfloat16_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmlal.lane.vg2x4.nxv8bf16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_lane4_bf16(uint32_t slice_base, svbfloat16x4_t zn, svbfloat16_t zm) __arm_streaming __arm_shared_za
+void test_svmla_lane4_bf16(uint32_t slice_base, svbfloat16x4_t zn, svbfloat16_t zm) __arm_streaming __arm_inout("za")
 {
-   SVE_ACLE_FUNC(svmla_lane_za32,,,_bf16,_vg2x4)(slice_base, zn, zm, 7);
+   SVE_ACLE_FUNC(svmla_lane_za32,_bf16,_vg2x4,,)(slice_base, zn, zm, 7);
 }
 
 // CHECK-LABEL: @test_svmla_lane4_u16(
@@ -667,9 +667,9 @@ void test_svmla_lane4_bf16(uint32_t slice_base, svbfloat16x4_t zn, svbfloat16_t
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umlal.lane.vg2x4.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_lane4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16_t zm) __arm_streaming __arm_shared_za
+void test_svmla_lane4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16_t zm) __arm_streaming __arm_inout("za")
 {
-   SVE_ACLE_FUNC(svmla_lane_za32,,,_u16,_vg2x4)(slice_base, zn, zm, 7);
+   SVE_ACLE_FUNC(svmla_lane_za32,_u16,_vg2x4,,)(slice_base, zn, zm, 7);
 }
 
 // CHECK-LABEL: @test_svmla_lane4_s16(
@@ -690,7 +690,7 @@ void test_svmla_lane4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16_t zm) _
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smlal.lane.vg2x4.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_lane4_s16(uint32_t slice_base, svint16x4_t zn, svint16_t zm) __arm_streaming __arm_shared_za
+void test_svmla_lane4_s16(uint32_t slice_base, svint16x4_t zn, svint16_t zm) __arm_streaming __arm_inout("za")
 {
-   SVE_ACLE_FUNC(svmla_lane_za32,,,_s16,_vg2x4)(slice_base, zn, zm, 7);
+   SVE_ACLE_FUNC(svmla_lane_za32,_s16,_vg2x4,,)(slice_base, zn, zm, 7);
 }
diff --git a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_mlall.c b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_mlall.c
index fceb16a482600..50cbe3e4bac5b 100644
--- a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_mlall.c
+++ b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_mlall.c
@@ -31,7 +31,7 @@
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smla.za32.single.vg4x1.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[ZN:%.*]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_single_x1_s8(uint32_t slice_base, svint8_t zn, svint8_t zm) __arm_streaming __arm_shared_za
+void test_svmla_single_x1_s8(uint32_t slice_base, svint8_t zn, svint8_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmla_za32,_s8,_vg4x1,,)(slice_base, zn, zm);
 }
@@ -46,7 +46,7 @@ void test_svmla_single_x1_s8(uint32_t slice_base, svint8_t zn, svint8_t zm) __ar
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smla.za64.single.vg4x1.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[ZN:%.*]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_single_x1_s16(uint32_t slice_base, svint16_t zn, svint16_t zm) __arm_streaming __arm_shared_za
+void test_svmla_single_x1_s16(uint32_t slice_base, svint16_t zn, svint16_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmla_za64,_s16,_vg4x1,,)(slice_base, zn, zm);
 }
@@ -61,7 +61,7 @@ void test_svmla_single_x1_s16(uint32_t slice_base, svint16_t zn, svint16_t zm) _
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umla.za32.single.vg4x1.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[ZN:%.*]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_uvmlal_single_x1_u8(uint32_t slice_base, svuint8_t zn, svuint8_t zm) __arm_streaming __arm_shared_za
+void test_uvmlal_single_x1_u8(uint32_t slice_base, svuint8_t zn, svuint8_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmla_za32,_u8,_vg4x1,,)(slice_base, zn, zm);
 }
@@ -76,7 +76,7 @@ void test_uvmlal_single_x1_u8(uint32_t slice_base, svuint8_t zn, svuint8_t zm) _
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umla.za64.single.vg4x1.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[ZN:%.*]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_uvmlal_single_x1_u16(uint32_t slice_base, svuint16_t zn, svuint16_t zm) __arm_streaming __arm_shared_za
+void test_uvmlal_single_x1_u16(uint32_t slice_base, svuint16_t zn, svuint16_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmla_za64,_u16,_vg4x1,,)(slice_base, zn, zm);
 }
@@ -93,7 +93,7 @@ void test_uvmlal_single_x1_u16(uint32_t slice_base, svuint16_t zn, svuint16_t zm
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smls.za32.single.vg4x1.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[ZN:%.*]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_single_x1_s8(uint32_t slice_base, svint8_t zn, svint8_t zm) __arm_streaming __arm_shared_za
+void test_svmls_single_x1_s8(uint32_t slice_base, svint8_t zn, svint8_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmls_za32,_s8,_vg4x1,,)(slice_base, zn, zm);
 }
@@ -108,7 +108,7 @@ void test_svmls_single_x1_s8(uint32_t slice_base, svint8_t zn, svint8_t zm) __ar
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smls.za64.single.vg4x1.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[ZN:%.*]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_single_x1_s16(uint32_t slice_base, svint16_t zn, svint16_t zm) __arm_streaming __arm_shared_za
+void test_svmls_single_x1_s16(uint32_t slice_base, svint16_t zn, svint16_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmls_za64,_s16,_vg4x1,,)(slice_base, zn, zm);
 }
@@ -123,7 +123,7 @@ void test_svmls_single_x1_s16(uint32_t slice_base, svint16_t zn, svint16_t zm) _
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umls.za32.single.vg4x1.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[ZN:%.*]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_uvmlsl_single_x1_u8(uint32_t slice_base, svuint8_t zn, svuint8_t zm) __arm_streaming __arm_shared_za
+void test_uvmlsl_single_x1_u8(uint32_t slice_base, svuint8_t zn, svuint8_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmls_za32,_u8,_vg4x1,,)(slice_base, zn, zm);
 }
@@ -138,7 +138,7 @@ void test_uvmlsl_single_x1_u8(uint32_t slice_base, svuint8_t zn, svuint8_t zm) _
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umls.za64.single.vg4x1.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[ZN:%.*]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_uvmlsl_single_x1_u16(uint32_t slice_base, svuint16_t zn, svuint16_t zm) __arm_streaming __arm_shared_za
+void test_uvmlsl_single_x1_u16(uint32_t slice_base, svuint16_t zn, svuint16_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmls_za64,_u16,_vg4x1,,)(slice_base, zn, zm);
 }
@@ -155,7 +155,7 @@ void test_uvmlsl_single_x1_u16(uint32_t slice_base, svuint16_t zn, svuint16_t zm
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.usmla.za32.single.vg4x1.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[ZM:%.*]],  [[ZN:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_sumlall_single_x1_s8(uint32_t slice_base, svint8_t zn, svuint8_t zm) __arm_streaming __arm_shared_za
+void test_sumlall_single_x1_s8(uint32_t slice_base, svint8_t zn, svuint8_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svsumla_za32,_s8,_vg4x1,,)(slice_base, zn, zm);
 }
@@ -172,7 +172,7 @@ void test_sumlall_single_x1_s8(uint32_t slice_base, svint8_t zn, svuint8_t zm) _
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.usmla.za32.single.vg4x1.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[ZN:%.*]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_usmlall_single_x1_u8(uint32_t slice_base, svuint8_t zn, svint8_t zm) __arm_streaming __arm_shared_za
+void test_usmlall_single_x1_u8(uint32_t slice_base, svuint8_t zn, svint8_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svusmla_za32,_u8,_vg4x1,,)(slice_base, zn, zm);
 }
@@ -197,9 +197,9 @@ void test_usmlall_single_x1_u8(uint32_t slice_base, svuint8_t zn, svint8_t zm) _
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smla.za32.single.vg4x2.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_single_x2_s8(uint32_t slice_base, svint8x2_t zn, svint8_t zm) __arm_streaming __arm_shared_za
+void test_svmla_single_x2_s8(uint32_t slice_base, svint8x2_t zn, svint8_t zm) __arm_streaming __arm_inout("za")
 {
-  SVE_ACLE_FUNC(svmla_single_za32,,_s8,,_vg4x2)(slice_base, zn, zm);
+  SVE_ACLE_FUNC(svmla,_single,_za32,_s8,_vg4x2)(slice_base, zn, zm);
 }
 
 // CHECK-LABEL: @test_svmla_single_x2_s16(
@@ -216,9 +216,9 @@ void test_svmla_single_x2_s8(uint32_t slice_base, svint8x2_t zn, svint8_t zm) __
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smla.za64.single.vg4x2.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_single_x2_s16(uint32_t slice_base, svint16x2_t zn, svint16_t zm) __arm_streaming __arm_shared_za
+void test_svmla_single_x2_s16(uint32_t slice_base, svint16x2_t zn, svint16_t zm) __arm_streaming __arm_inout("za")
 {
-  SVE_ACLE_FUNC(svmla_single_za64,,_s16,,_vg4x2)(slice_base, zn, zm);
+  SVE_ACLE_FUNC(svmla,_single,_za64,_s16,_vg4x2)(slice_base, zn, zm);
 }
 
 // CHECK-LABEL: @test_svmla_single_x2_u8(
@@ -235,9 +235,9 @@ void test_svmla_single_x2_s16(uint32_t slice_base, svint16x2_t zn, svint16_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umla.za32.single.vg4x2.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_single_x2_u8(uint32_t slice_base, svuint8x2_t zn, svuint8_t zm) __arm_streaming __arm_shared_za
+void test_svmla_single_x2_u8(uint32_t slice_base, svuint8x2_t zn, svuint8_t zm) __arm_streaming __arm_inout("za")
 {
-  SVE_ACLE_FUNC(svmla_single_za32,,_u8,,_vg4x2)(slice_base, zn, zm);
+  SVE_ACLE_FUNC(svmla,_single,_za32,_u8,_vg4x2)(slice_base, zn, zm);
 }
 
 // CHECK-LABEL: @test_svmla_single_x2_u16(
@@ -254,9 +254,9 @@ void test_svmla_single_x2_u8(uint32_t slice_base, svuint8x2_t zn, svuint8_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umla.za64.single.vg4x2.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_single_x2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16_t zm) __arm_streaming __arm_shared_za
+void test_svmla_single_x2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16_t zm) __arm_streaming __arm_inout("za")
 {
-  SVE_ACLE_FUNC(svmla_single_za64,,_u16,,_vg4x2)(slice_base, zn, zm);
+  SVE_ACLE_FUNC(svmla,_single,_za64,_u16,_vg4x2)(slice_base, zn, zm);
 }
 
 // MLSL
@@ -275,9 +275,9 @@ void test_svmla_single_x2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16_t z
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smls.za32.single.vg4x2.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_single_x2_s8(uint32_t slice_base, svint8x2_t zn, svint8_t zm) __arm_streaming __arm_shared_za
+void test_svmls_single_x2_s8(uint32_t slice_base, svint8x2_t zn, svint8_t zm) __arm_streaming __arm_inout("za")
 {
-  SVE_ACLE_FUNC(svmls_single_za32,,_s8,,_vg4x2)(slice_base, zn, zm);
+  SVE_ACLE_FUNC(svmls,_single,_za32,_s8,_vg4x2)(slice_base, zn, zm);
 }
 
 // CHECK-LABEL: @test_svmls_single_x2_s16(
@@ -294,9 +294,9 @@ void test_svmls_single_x2_s8(uint32_t slice_base, svint8x2_t zn, svint8_t zm) __
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smls.za64.single.vg4x2.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_single_x2_s16(uint32_t slice_base, svint16x2_t zn, svint16_t zm) __arm_streaming __arm_shared_za
+void test_svmls_single_x2_s16(uint32_t slice_base, svint16x2_t zn, svint16_t zm) __arm_streaming __arm_inout("za")
 {
-  SVE_ACLE_FUNC(svmls_single_za64,,_s16,,_vg4x2)(slice_base, zn, zm);
+  SVE_ACLE_FUNC(svmls,_single,_za64,_s16,_vg4x2)(slice_base, zn, zm);
 }
 
 // CHECK-LABEL: @test_svmls_single_x2_u8(
@@ -313,9 +313,9 @@ void test_svmls_single_x2_s16(uint32_t slice_base, svint16x2_t zn, svint16_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umls.za32.single.vg4x2.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_single_x2_u8(uint32_t slice_base, svuint8x2_t zn, svuint8_t zm) __arm_streaming __arm_shared_za
+void test_svmls_single_x2_u8(uint32_t slice_base, svuint8x2_t zn, svuint8_t zm) __arm_streaming __arm_inout("za")
 {
-  SVE_ACLE_FUNC(svmls_single_za32,,_u8,,_vg4x2)(slice_base, zn, zm);
+  SVE_ACLE_FUNC(svmls,_single,_za32,_u8,_vg4x2)(slice_base, zn, zm);
 }
 
 // CHECK-LABEL: @test_svmls_single_x2_u16(
@@ -332,9 +332,9 @@ void test_svmls_single_x2_u8(uint32_t slice_base, svuint8x2_t zn, svuint8_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umls.za64.single.vg4x2.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_single_x2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16_t zm) __arm_streaming __arm_shared_za
+void test_svmls_single_x2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16_t zm) __arm_streaming __arm_inout("za")
 {
-  SVE_ACLE_FUNC(svmls_single_za64,,_u16,,_vg4x2)(slice_base, zn, zm);
+  SVE_ACLE_FUNC(svmls,_single,_za64,_u16,_vg4x2)(slice_base, zn, zm);
 }
 
 // SUMLALL
@@ -353,9 +353,9 @@ void test_svmls_single_x2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16_t z
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sumla.za32.single.vg4x2.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svsumla_single_x2_s8(uint32_t slice_base, svint8x2_t zn, svuint8_t zm) __arm_streaming __arm_shared_za
+void test_svsumla_single_x2_s8(uint32_t slice_base, svint8x2_t zn, svuint8_t zm) __arm_streaming __arm_inout("za")
 {
-  SVE_ACLE_FUNC(svsumla_single_za32,,_s8,,_vg4x2)(slice_base, zn, zm);
+  SVE_ACLE_FUNC(svsumla,_single,_za32,_s8,_vg4x2)(slice_base, zn, zm);
 }
 
 // USMLALL
@@ -374,9 +374,9 @@ void test_svsumla_single_x2_s8(uint32_t slice_base, svint8x2_t zn, svuint8_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.usmla.za32.single.vg4x2.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_usmlall_single_x2_u8(uint32_t slice_base, svuint8x2_t zn, svint8_t zm) __arm_streaming __arm_shared_za
+void test_usmlall_single_x2_u8(uint32_t slice_base, svuint8x2_t zn, svint8_t zm) __arm_streaming __arm_inout("za")
 {
-  SVE_ACLE_FUNC(svusmla_single_za32,,_u8,,_vg4x2)(slice_base, zn, zm);
+  SVE_ACLE_FUNC(svusmla,_single,_za32,_u8,_vg4x2)(slice_base, zn, zm);
 }
 
 //
@@ -403,9 +403,9 @@ void test_usmlall_single_x2_u8(uint32_t slice_base, svuint8x2_t zn, svint8_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smla.za32.single.vg4x4.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_single_x4_s8(uint32_t slice_base, svint8x4_t zn, svint8_t zm) __arm_streaming __arm_shared_za
+void test_svmla_single_x4_s8(uint32_t slice_base, svint8x4_t zn, svint8_t zm) __arm_streaming __arm_inout("za")
 {
-  SVE_ACLE_FUNC(svmla_single_za32,,_s8,,_vg4x4)(slice_base, zn, zm);
+  SVE_ACLE_FUNC(svmla,_single,_za32,_s8,_vg4x4)(slice_base, zn, zm);
 }
 
 // CHECK-LABEL: @test_svmla_single_x4_s16(
@@ -426,9 +426,9 @@ void test_svmla_single_x4_s8(uint32_t slice_base, svint8x4_t zn, svint8_t zm) __
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smla.za64.single.vg4x4.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_single_x4_s16(uint32_t slice_base, svint16x4_t zn, svint16_t zm) __arm_streaming __arm_shared_za
+void test_svmla_single_x4_s16(uint32_t slice_base, svint16x4_t zn, svint16_t zm) __arm_streaming __arm_inout("za")
 {
-  SVE_ACLE_FUNC(svmla_single_za64,,_s16,,_vg4x4)(slice_base, zn, zm);
+  SVE_ACLE_FUNC(svmla,_single,_za64,_s16,_vg4x4)(slice_base, zn, zm);
 }
 
 // CHECK-LABEL: @test_svmla_single_x4_u8(
@@ -449,9 +449,9 @@ void test_svmla_single_x4_s16(uint32_t slice_base, svint16x4_t zn, svint16_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umla.za32.single.vg4x4.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_single_x4_u8(uint32_t slice_base, svuint8x4_t zn, svuint8_t zm) __arm_streaming __arm_shared_za
+void test_svmla_single_x4_u8(uint32_t slice_base, svuint8x4_t zn, svuint8_t zm) __arm_streaming __arm_inout("za")
 {
-  SVE_ACLE_FUNC(svmla_single_za32,,_u8,,_vg4x4)(slice_base, zn, zm);
+  SVE_ACLE_FUNC(svmla,_single,_za32,_u8,_vg4x4)(slice_base, zn, zm);
 }
 
 // CHECK-LABEL: @test_svmla_single_x4_u16(
@@ -472,9 +472,9 @@ void test_svmla_single_x4_u8(uint32_t slice_base, svuint8x4_t zn, svuint8_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umla.za64.single.vg4x4.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmla_single_x4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16_t zm) __arm_streaming __arm_shared_za
+void test_svmla_single_x4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16_t zm) __arm_streaming __arm_inout("za")
 {
-  SVE_ACLE_FUNC(svmla_single_za64,,_u16,,_vg4x4)(slice_base, zn, zm);
+  SVE_ACLE_FUNC(svmla,_single,_za64,_u16,_vg4x4)(slice_base, zn, zm);
 }
 
 // MLSL
@@ -497,9 +497,9 @@ void test_svmla_single_x4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16_t z
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smls.za32.single.vg4x4.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_single_x4_s8(uint32_t slice_base, svint8x4_t zn, svint8_t zm) __arm_streaming __arm_shared_za
+void test_svmls_single_x4_s8(uint32_t slice_base, svint8x4_t zn, svint8_t zm) __arm_streaming __arm_inout("za")
 {
-  SVE_ACLE_FUNC(svmls_single_za32,,_s8,,_vg4x4)(slice_base, zn, zm);
+  SVE_ACLE_FUNC(svmls,_single,_za32,_s8,_vg4x4)(slice_base, zn, zm);
 }
 
 // CHECK-LABEL: @test_svmls_single_x4_s16(
@@ -520,9 +520,9 @@ void test_svmls_single_x4_s8(uint32_t slice_base, svint8x4_t zn, svint8_t zm) __
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smls.za64.single.vg4x4.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_single_x4_s16(uint32_t slice_base, svint16x4_t zn, svint16_t zm) __arm_streaming __arm_shared_za
+void test_svmls_single_x4_s16(uint32_t slice_base, svint16x4_t zn, svint16_t zm) __arm_streaming __arm_inout("za")
 {
-  SVE_ACLE_FUNC(svmls_single_za64,,_s16,,_vg4x4)(slice_base, zn, zm);
+  SVE_ACLE_FUNC(svmls,_single,_za64,_s16,_vg4x4)(slice_base, zn, zm);
 }
 
 // CHECK-LABEL: @test_svmls_single_x4_u8(
@@ -543,9 +543,9 @@ void test_svmls_single_x4_s16(uint32_t slice_base, svint16x4_t zn, svint16_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umls.za32.single.vg4x4.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_single_x4_u8(uint32_t slice_base, svuint8x4_t zn, svuint8_t zm) __arm_streaming __arm_shared_za
+void test_svmls_single_x4_u8(uint32_t slice_base, svuint8x4_t zn, svuint8_t zm) __arm_streaming __arm_inout("za")
 {
-  SVE_ACLE_FUNC(svmls_single_za32,,_u8,,_vg4x4)(slice_base, zn, zm);
+  SVE_ACLE_FUNC(svmls,_single,_za32,_u8,_vg4x4)(slice_base, zn, zm);
 }
 
 // CHECK-LABEL: @test_svmls_single_x4_u16(
@@ -566,9 +566,9 @@ void test_svmls_single_x4_u8(uint32_t slice_base, svuint8x4_t zn, svuint8_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umls.za64.single.vg4x4.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_single_x4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16_t zm) __arm_streaming __arm_shared_za
+void test_svmls_single_x4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16_t zm) __arm_streaming __arm_inout("za")
 {
-  SVE_ACLE_FUNC(svmls_single_za64,,_u16,,_vg4x4)(slice_base, zn, zm);
+  SVE_ACLE_FUNC(svmls,_single,_za64,_u16,_vg4x4)(slice_base, zn, zm);
 }
 
 // SUMLALL
@@ -591,9 +591,9 @@ void test_svmls_single_x4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16_t z
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sumla.za32.single.vg4x4.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svsumla_single_x4_s8(uint32_t slice_base, svint8x4_t zn, svuint8_t zm) __arm_streaming __arm_shared_za
+void test_svsumla_single_x4_s8(uint32_t slice_base, svint8x4_t zn, svuint8_t zm) __arm_streaming __arm_inout("za")
 {
-  SVE_ACLE_FUNC(svsumla_single_za32,,_s8,,_vg4x4)(slice_base, zn, zm);
+  SVE_ACLE_FUNC(svsumla,_single,_za32,_s8,_vg4x4)(slice_base, zn, zm);
 }
 
 // USMLALL
@@ -616,9 +616,9 @@ void test_svsumla_single_x4_s8(uint32_t slice_base, svint8x4_t zn, svuint8_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.usmla.za32.single.vg4x4.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_usmlall_single_x4_u8(uint32_t slice_base, svuint8x4_t zn, svint8_t zm) __arm_streaming __arm_shared_za
+void test_usmlall_single_x4_u8(uint32_t slice_base, svuint8x4_t zn, svint8_t zm) __arm_streaming __arm_inout("za")
 {
-  SVE_ACLE_FUNC(svusmla_single_za32,,_u8,,_vg4x4)(slice_base, zn, zm);
+  SVE_ACLE_FUNC(svusmla,_single,_za32,_u8,_vg4x4)(slice_base, zn, zm);
 }
 
 //
@@ -645,7 +645,7 @@ void test_usmlall_single_x4_u8(uint32_t slice_base, svuint8x4_t zn, svint8_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smla.za32.vg4x2.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_mlal_multi_x2_s8(uint32_t slice_base, svint8x2_t zn, svint8x2_t zm) __arm_streaming __arm_shared_za
+void test_mlal_multi_x2_s8(uint32_t slice_base, svint8x2_t zn, svint8x2_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmla_za32,_s8,_vg4x2,,)(slice_base, zn, zm);
 }
@@ -668,7 +668,7 @@ void test_mlal_multi_x2_s8(uint32_t slice_base, svint8x2_t zn, svint8x2_t zm) __
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smla.za64.vg4x2.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_mlal_multi_x2_s16(uint32_t slice_base, svint16x2_t zn, svint16x2_t zm) __arm_streaming __arm_shared_za
+void test_mlal_multi_x2_s16(uint32_t slice_base, svint16x2_t zn, svint16x2_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmla_za64,_s16,_vg4x2,,)(slice_base, zn, zm);
 }
@@ -691,7 +691,7 @@ void test_mlal_multi_x2_s16(uint32_t slice_base, svint16x2_t zn, svint16x2_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umla.za32.vg4x2.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_mlal_multi_x2_u8(uint32_t slice_base, svuint8x2_t zn, svuint8x2_t zm) __arm_streaming __arm_shared_za
+void test_mlal_multi_x2_u8(uint32_t slice_base, svuint8x2_t zn, svuint8x2_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmla_za32,_u8,_vg4x2,,)(slice_base, zn, zm);
 }
@@ -714,7 +714,7 @@ void test_mlal_multi_x2_u8(uint32_t slice_base, svuint8x2_t zn, svuint8x2_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umla.za64.vg4x2.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_mlal_multi_x2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16x2_t zm) __arm_streaming __arm_shared_za
+void test_mlal_multi_x2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16x2_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmla_za64,_u16,_vg4x2,,)(slice_base, zn, zm);
 }
@@ -739,7 +739,7 @@ void test_mlal_multi_x2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16x2_t z
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smls.za32.vg4x2.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_mlsl_multi_x2_s8(uint32_t slice_base, svint8x2_t zn, svint8x2_t zm) __arm_streaming __arm_shared_za
+void test_mlsl_multi_x2_s8(uint32_t slice_base, svint8x2_t zn, svint8x2_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmls_za32,_s8,_vg4x2,,)(slice_base, zn, zm);
 }
@@ -762,7 +762,7 @@ void test_mlsl_multi_x2_s8(uint32_t slice_base, svint8x2_t zn, svint8x2_t zm) __
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smls.za64.vg4x2.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_mlsl_multi_x2_s16(uint32_t slice_base, svint16x2_t zn, svint16x2_t zm) __arm_streaming __arm_shared_za
+void test_mlsl_multi_x2_s16(uint32_t slice_base, svint16x2_t zn, svint16x2_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmls_za64,_s16,_vg4x2,,)(slice_base, zn, zm);
 }
@@ -785,7 +785,7 @@ void test_mlsl_multi_x2_s16(uint32_t slice_base, svint16x2_t zn, svint16x2_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umls.za32.vg4x2.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_mlsl_multi_x2_u8(uint32_t slice_base, svuint8x2_t zn, svuint8x2_t zm) __arm_streaming __arm_shared_za
+void test_mlsl_multi_x2_u8(uint32_t slice_base, svuint8x2_t zn, svuint8x2_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmls_za32,_u8,_vg4x2,,)(slice_base, zn, zm);
 }
@@ -808,7 +808,7 @@ void test_mlsl_multi_x2_u8(uint32_t slice_base, svuint8x2_t zn, svuint8x2_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umls.za64.vg4x2.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_mlsl_multi_x2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16x2_t zm) __arm_streaming __arm_shared_za
+void test_mlsl_multi_x2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16x2_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmls_za64,_u16,_vg4x2,,)(slice_base, zn, zm);
 }
@@ -833,7 +833,7 @@ void test_mlsl_multi_x2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16x2_t z
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.usmla.za32.vg4x2.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP2]],  [[TMP3]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_sumlal_multi_x2_s8(uint32_t slice_base, svint8x2_t zn, svuint8x2_t zm) __arm_streaming __arm_shared_za
+void test_sumlal_multi_x2_s8(uint32_t slice_base, svint8x2_t zn, svuint8x2_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svsumla_za32,_s8,_vg4x2,,)(slice_base, zn, zm);
 }
@@ -858,7 +858,7 @@ void test_sumlal_multi_x2_s8(uint32_t slice_base, svint8x2_t zn, svuint8x2_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.usmla.za32.vg4x2.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_usmlal_multi_x2_u8(uint32_t slice_base, svuint8x2_t zn, svint8x2_t zm) __arm_streaming __arm_shared_za
+void test_usmlal_multi_x2_u8(uint32_t slice_base, svuint8x2_t zn, svint8x2_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svusmla_za32,_u8,_vg4x2,,)(slice_base, zn, zm);
 }
@@ -895,7 +895,7 @@ void test_usmlal_multi_x2_u8(uint32_t slice_base, svuint8x2_t zn, svint8x2_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smla.za32.vg4x4.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[TMP4]],  [[TMP5]],  [[TMP6]],  [[TMP7]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_mlal_multi_x4_s8(uint32_t slice_base, svint8x4_t zn, svint8x4_t zm) __arm_streaming __arm_shared_za
+void test_mlal_multi_x4_s8(uint32_t slice_base, svint8x4_t zn, svint8x4_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmla_za32,_s8,_vg4x4,,)(slice_base, zn, zm);
 }
@@ -926,7 +926,7 @@ void test_mlal_multi_x4_s8(uint32_t slice_base, svint8x4_t zn, svint8x4_t zm) __
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smla.za64.vg4x4.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[TMP4]],  [[TMP5]],  [[TMP6]],  [[TMP7]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_mlal_multi_x4_s16(uint32_t slice_base, svint16x4_t zn, svint16x4_t zm) __arm_streaming __arm_shared_za
+void test_mlal_multi_x4_s16(uint32_t slice_base, svint16x4_t zn, svint16x4_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmla_za64,_s16,_vg4x4,,)(slice_base, zn, zm);
 }
@@ -957,7 +957,7 @@ void test_mlal_multi_x4_s16(uint32_t slice_base, svint16x4_t zn, svint16x4_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umla.za32.vg4x4.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[TMP4]],  [[TMP5]],  [[TMP6]],  [[TMP7]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_mlal_multi_x4_u8(uint32_t slice_base, svuint8x4_t zn, svuint8x4_t zm) __arm_streaming __arm_shared_za
+void test_mlal_multi_x4_u8(uint32_t slice_base, svuint8x4_t zn, svuint8x4_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmla_za32,_u8,_vg4x4,,)(slice_base, zn, zm);
 }
@@ -988,7 +988,7 @@ void test_mlal_multi_x4_u8(uint32_t slice_base, svuint8x4_t zn, svuint8x4_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umla.za64.vg4x4.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[TMP4]],  [[TMP5]],  [[TMP6]],  [[TMP7]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_mlal_multi_x4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16x4_t zm) __arm_streaming __arm_shared_za
+void test_mlal_multi_x4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16x4_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmla_za64,_u16,_vg4x4,,)(slice_base, zn, zm);
 }
@@ -1021,7 +1021,7 @@ void test_mlal_multi_x4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16x4_t z
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smls.za32.vg4x4.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[TMP4]],  [[TMP5]],  [[TMP6]],  [[TMP7]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_mlsl_multi_x4_s8(uint32_t slice_base, svint8x4_t zn, svint8x4_t zm) __arm_streaming __arm_shared_za
+void test_mlsl_multi_x4_s8(uint32_t slice_base, svint8x4_t zn, svint8x4_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmls_za32,_s8,_vg4x4,,)(slice_base, zn, zm);
 }
@@ -1052,7 +1052,7 @@ void test_mlsl_multi_x4_s8(uint32_t slice_base, svint8x4_t zn, svint8x4_t zm) __
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smls.za64.vg4x4.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[TMP4]],  [[TMP5]],  [[TMP6]],  [[TMP7]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_mlsl_multi_x4_s16(uint32_t slice_base, svint16x4_t zn, svint16x4_t zm) __arm_streaming __arm_shared_za
+void test_mlsl_multi_x4_s16(uint32_t slice_base, svint16x4_t zn, svint16x4_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmls_za64,_s16,_vg4x4,,)(slice_base, zn, zm);
 }
@@ -1083,7 +1083,7 @@ void test_mlsl_multi_x4_s16(uint32_t slice_base, svint16x4_t zn, svint16x4_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umls.za32.vg4x4.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[TMP4]],  [[TMP5]],  [[TMP6]],  [[TMP7]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_mlsl_multi_x4_u8(uint32_t slice_base, svuint8x4_t zn, svuint8x4_t zm) __arm_streaming __arm_shared_za
+void test_mlsl_multi_x4_u8(uint32_t slice_base, svuint8x4_t zn, svuint8x4_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmls_za32,_u8,_vg4x4,,)(slice_base, zn, zm);
 }
@@ -1114,7 +1114,7 @@ void test_mlsl_multi_x4_u8(uint32_t slice_base, svuint8x4_t zn, svuint8x4_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umls.za64.vg4x4.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[TMP4]],  [[TMP5]],  [[TMP6]],  [[TMP7]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_mlsl_multi_x4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16x4_t zm) __arm_streaming __arm_shared_za
+void test_mlsl_multi_x4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16x4_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmls_za64,_u16,_vg4x4,,)(slice_base, zn, zm);
 }
@@ -1147,7 +1147,7 @@ void test_mlsl_multi_x4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16x4_t z
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.usmla.za32.vg4x4.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP4]],  [[TMP5]],  [[TMP6]],  [[TMP7]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_sumlal_multi_x4_s8(uint32_t slice_base, svint8x4_t zn, svuint8x4_t zm) __arm_streaming __arm_shared_za
+void test_sumlal_multi_x4_s8(uint32_t slice_base, svint8x4_t zn, svuint8x4_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svsumla_za32,_s8,_vg4x4,,)(slice_base, zn, zm);
 }
@@ -1180,7 +1180,7 @@ void test_sumlal_multi_x4_s8(uint32_t slice_base, svint8x4_t zn, svuint8x4_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.usmla.za32.vg4x4.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[TMP4]],  [[TMP5]],  [[TMP6]],  [[TMP7]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_usmlal_multi_x4_u8(uint32_t slice_base, svuint8x4_t zn, svint8x4_t zm) __arm_streaming __arm_shared_za
+void test_usmlal_multi_x4_u8(uint32_t slice_base, svuint8x4_t zn, svint8x4_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svusmla_za32,_u8,_vg4x4,,)(slice_base, zn, zm);
 }
@@ -1201,7 +1201,7 @@ void test_usmlal_multi_x4_u8(uint32_t slice_base, svuint8x4_t zn, svint8x4_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smla.za32.lane.vg4x1.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[ZN:%.*]],  [[ZM:%.*]], i32 15)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_smlal_lane_x1_s8(uint32_t slice_base, svint8_t zn, svint8_t zm) __arm_streaming __arm_shared_za
+void test_smlal_lane_x1_s8(uint32_t slice_base, svint8_t zn, svint8_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmla_lane_za32,_s8,_vg4x1,,)(slice_base, zn, zm, 15);
 }
@@ -1216,7 +1216,7 @@ void test_smlal_lane_x1_s8(uint32_t slice_base, svint8_t zn, svint8_t zm) __arm_
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smla.za64.lane.vg4x1.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[ZN:%.*]],  [[ZM:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_smlal_lane_x1_s16(uint32_t slice_base, svint16_t zn, svint16_t zm) __arm_streaming __arm_shared_za
+void test_smlal_lane_x1_s16(uint32_t slice_base, svint16_t zn, svint16_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmla_lane_za64,_s16,_vg4x1,,)(slice_base, zn, zm, 7);
 }
@@ -1231,7 +1231,7 @@ void test_smlal_lane_x1_s16(uint32_t slice_base, svint16_t zn, svint16_t zm) __a
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umla.za32.lane.vg4x1.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[ZN:%.*]],  [[ZM:%.*]], i32 15)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_smlal_lane_x1_u8(uint32_t slice_base, svuint8_t zn, svuint8_t zm) __arm_streaming __arm_shared_za
+void test_smlal_lane_x1_u8(uint32_t slice_base, svuint8_t zn, svuint8_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmla_lane_za32,_u8,_vg4x1,,)(slice_base, zn, zm, 15);
 }
@@ -1246,7 +1246,7 @@ void test_smlal_lane_x1_u8(uint32_t slice_base, svuint8_t zn, svuint8_t zm) __ar
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umla.za64.lane.vg4x1.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[ZN:%.*]],  [[ZM:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_smlal_lane_x1_u16(uint32_t slice_base, svuint16_t zn, svuint16_t zm) __arm_streaming __arm_shared_za
+void test_smlal_lane_x1_u16(uint32_t slice_base, svuint16_t zn, svuint16_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmla_lane_za64,_u16,_vg4x1,,)(slice_base, zn, zm, 7);
 }
@@ -1263,7 +1263,7 @@ void test_smlal_lane_x1_u16(uint32_t slice_base, svuint16_t zn, svuint16_t zm) _
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smls.za32.lane.vg4x1.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[ZN:%.*]],  [[ZM:%.*]], i32 15)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_smlsl_lane_x1_s8(uint32_t slice_base, svint8_t zn, svint8_t zm) __arm_streaming __arm_shared_za
+void test_smlsl_lane_x1_s8(uint32_t slice_base, svint8_t zn, svint8_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmls_lane_za32,_s8,_vg4x1,,)(slice_base, zn, zm, 15);
 }
@@ -1278,7 +1278,7 @@ void test_smlsl_lane_x1_s8(uint32_t slice_base, svint8_t zn, svint8_t zm) __arm_
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smls.za64.lane.vg4x1.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[ZN:%.*]],  [[ZM:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_smlsl_lane_x1_s16(uint32_t slice_base, svint16_t zn, svint16_t zm) __arm_streaming __arm_shared_za
+void test_smlsl_lane_x1_s16(uint32_t slice_base, svint16_t zn, svint16_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmls_lane_za64,_s16,_vg4x1,,)(slice_base, zn, zm, 7);
 }
@@ -1293,7 +1293,7 @@ void test_smlsl_lane_x1_s16(uint32_t slice_base, svint16_t zn, svint16_t zm) __a
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umls.za32.lane.vg4x1.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[ZN:%.*]],  [[ZM:%.*]], i32 15)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_smlsl_lane_x1_u8(uint32_t slice_base, svuint8_t zn, svuint8_t zm) __arm_streaming __arm_shared_za
+void test_smlsl_lane_x1_u8(uint32_t slice_base, svuint8_t zn, svuint8_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmls_lane_za32,_u8,_vg4x1,,)(slice_base, zn, zm, 15);
 }
@@ -1308,7 +1308,7 @@ void test_smlsl_lane_x1_u8(uint32_t slice_base, svuint8_t zn, svuint8_t zm) __ar
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umls.za64.lane.vg4x1.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[ZN:%.*]],  [[ZM:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_smlsl_lane_x1_u16(uint32_t slice_base, svuint16_t zn, svuint16_t zm) __arm_streaming __arm_shared_za
+void test_smlsl_lane_x1_u16(uint32_t slice_base, svuint16_t zn, svuint16_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmls_lane_za64,_u16,_vg4x1,,)(slice_base, zn, zm, 7);
 }
@@ -1325,7 +1325,7 @@ void test_smlsl_lane_x1_u16(uint32_t slice_base, svuint16_t zn, svuint16_t zm) _
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sumla.za32.lane.vg4x1.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[ZN:%.*]],  [[ZM:%.*]], i32 15)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_sumlall_lane_x1_s8(uint32_t slice_base, svint8_t zn, svuint8_t zm) __arm_streaming __arm_shared_za
+void test_sumlall_lane_x1_s8(uint32_t slice_base, svint8_t zn, svuint8_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svsumla_lane_za32,_s8,_vg4x1,,)(slice_base, zn, zm, 15);
 }
@@ -1342,7 +1342,7 @@ void test_sumlall_lane_x1_s8(uint32_t slice_base, svint8_t zn, svuint8_t zm) __a
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.usmla.za32.lane.vg4x1.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[ZN:%.*]],  [[ZM:%.*]], i32 15)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_usmlall_lane_x1_u8(uint32_t slice_base, svuint8_t zn, svint8_t zm) __arm_streaming __arm_shared_za
+void test_usmlall_lane_x1_u8(uint32_t slice_base, svuint8_t zn, svint8_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svusmla_lane_za32,_u8,_vg4x1,,)(slice_base, zn, zm, 15);
 }
@@ -1367,7 +1367,7 @@ void test_usmlall_lane_x1_u8(uint32_t slice_base, svuint8_t zn, svint8_t zm) __a
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smla.za32.lane.vg4x2.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]], i32 15)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_smlal_lane_x2_s8(uint32_t slice_base, svint8x2_t zn, svint8_t zm) __arm_streaming __arm_shared_za
+void test_smlal_lane_x2_s8(uint32_t slice_base, svint8x2_t zn, svint8_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmla_lane_za32,_s8,_vg4x2,,)(slice_base, zn, zm, 15);
 }
@@ -1386,7 +1386,7 @@ void test_smlal_lane_x2_s8(uint32_t slice_base, svint8x2_t zn, svint8_t zm) __ar
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smla.za64.lane.vg4x2.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_smlal_lane_x2_s16(uint32_t slice_base, svint16x2_t zn, svint16_t zm) __arm_streaming __arm_shared_za
+void test_smlal_lane_x2_s16(uint32_t slice_base, svint16x2_t zn, svint16_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmla_lane_za64,_s16,_vg4x2,,)(slice_base, zn, zm, 7);
 }
@@ -1405,7 +1405,7 @@ void test_smlal_lane_x2_s16(uint32_t slice_base, svint16x2_t zn, svint16_t zm) _
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umla.za32.lane.vg4x2.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]], i32 15)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_smlal_lane_x2_u8(uint32_t slice_base, svuint8x2_t zn, svuint8_t zm) __arm_streaming __arm_shared_za
+void test_smlal_lane_x2_u8(uint32_t slice_base, svuint8x2_t zn, svuint8_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmla_lane_za32,_u8,_vg4x2,,)(slice_base, zn, zm, 15);
 }
@@ -1424,7 +1424,7 @@ void test_smlal_lane_x2_u8(uint32_t slice_base, svuint8x2_t zn, svuint8_t zm) __
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umla.za64.lane.vg4x2.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_smlal_lane_x2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16_t zm) __arm_streaming __arm_shared_za
+void test_smlal_lane_x2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmla_lane_za64,_u16,_vg4x2,,)(slice_base, zn, zm, 7);
 }
@@ -1445,7 +1445,7 @@ void test_smlal_lane_x2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smls.za32.lane.vg4x2.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]], i32 15)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_smlsl_lane_x2_s8(uint32_t slice_base, svint8x2_t zn, svint8_t zm) __arm_streaming __arm_shared_za
+void test_smlsl_lane_x2_s8(uint32_t slice_base, svint8x2_t zn, svint8_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmls_lane_za32,_s8,_vg4x2,,)(slice_base, zn, zm, 15);
 }
@@ -1464,7 +1464,7 @@ void test_smlsl_lane_x2_s8(uint32_t slice_base, svint8x2_t zn, svint8_t zm) __ar
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smls.za64.lane.vg4x2.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_smlsl_lane_x2_s16(uint32_t slice_base, svint16x2_t zn, svint16_t zm) __arm_streaming __arm_shared_za
+void test_smlsl_lane_x2_s16(uint32_t slice_base, svint16x2_t zn, svint16_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmls_lane_za64,_s16,_vg4x2,,)(slice_base, zn, zm, 7);
 }
@@ -1483,7 +1483,7 @@ void test_smlsl_lane_x2_s16(uint32_t slice_base, svint16x2_t zn, svint16_t zm) _
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umls.za32.lane.vg4x2.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]], i32 15)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_smlsl_lane_x2_u8(uint32_t slice_base, svuint8x2_t zn, svuint8_t zm) __arm_streaming __arm_shared_za
+void test_smlsl_lane_x2_u8(uint32_t slice_base, svuint8x2_t zn, svuint8_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmls_lane_za32,_u8,_vg4x2,,)(slice_base, zn, zm, 15);
 }
@@ -1502,7 +1502,7 @@ void test_smlsl_lane_x2_u8(uint32_t slice_base, svuint8x2_t zn, svuint8_t zm) __
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umls.za64.lane.vg4x2.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_smlsl_lane_x2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16_t zm) __arm_streaming __arm_shared_za
+void test_smlsl_lane_x2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmls_lane_za64,_u16,_vg4x2,,)(slice_base, zn, zm, 7);
 }
@@ -1523,7 +1523,7 @@ void test_smlsl_lane_x2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sumla.za32.lane.vg4x2.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]], i32 15)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_sumlall_lane_x2_s8(uint32_t slice_base, svint8x2_t zn, svuint8_t zm) __arm_streaming __arm_shared_za
+void test_sumlall_lane_x2_s8(uint32_t slice_base, svint8x2_t zn, svuint8_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svsumla_lane_za32,_s8,_vg4x2,,)(slice_base, zn, zm, 15);
 }
@@ -1542,7 +1542,7 @@ void test_sumlall_lane_x2_s8(uint32_t slice_base, svint8x2_t zn, svuint8_t zm) _
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.usmla.za32.lane.vg4x2.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]], i32 15)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_usmlall_lane_x2_u8(uint32_t slice_base, svuint8x2_t zn, svint8_t zm) __arm_streaming __arm_shared_za
+void test_usmlall_lane_x2_u8(uint32_t slice_base, svuint8x2_t zn, svint8_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svusmla_lane_za32,_u8,_vg4x2,,)(slice_base, zn, zm, 15);
 }
@@ -1571,7 +1571,7 @@ void test_usmlall_lane_x2_u8(uint32_t slice_base, svuint8x2_t zn, svint8_t zm) _
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smla.za32.lane.vg4x4.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]], i32 15)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_smlal_lane_x4_s8(uint32_t slice_base, svint8x4_t zn, svint8_t zm) __arm_streaming __arm_shared_za
+void test_smlal_lane_x4_s8(uint32_t slice_base, svint8x4_t zn, svint8_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmla_lane_za32,_s8,_vg4x4,,)(slice_base, zn, zm, 15);
 }
@@ -1594,7 +1594,7 @@ void test_smlal_lane_x4_s8(uint32_t slice_base, svint8x4_t zn, svint8_t zm) __ar
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smla.za64.lane.vg4x4.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_smlal_lane_x4_s16(uint32_t slice_base, svint16x4_t zn, svint16_t zm) __arm_streaming __arm_shared_za
+void test_smlal_lane_x4_s16(uint32_t slice_base, svint16x4_t zn, svint16_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmla_lane_za64,_s16,_vg4x4,,)(slice_base, zn, zm, 7);
 }
@@ -1617,7 +1617,7 @@ void test_smlal_lane_x4_s16(uint32_t slice_base, svint16x4_t zn, svint16_t zm) _
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umla.za32.lane.vg4x4.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]], i32 15)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_smlal_lane_x4_u8(uint32_t slice_base, svuint8x4_t zn, svuint8_t zm) __arm_streaming __arm_shared_za
+void test_smlal_lane_x4_u8(uint32_t slice_base, svuint8x4_t zn, svuint8_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmla_lane_za32,_u8,_vg4x4,,)(slice_base, zn, zm, 15);
 }
@@ -1640,7 +1640,7 @@ void test_smlal_lane_x4_u8(uint32_t slice_base, svuint8x4_t zn, svuint8_t zm) __
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umla.za64.lane.vg4x4.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_smlal_lane_x4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16_t zm) __arm_streaming __arm_shared_za
+void test_smlal_lane_x4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmla_lane_za64,_u16,_vg4x4,,)(slice_base, zn, zm, 7);
 }
@@ -1665,7 +1665,7 @@ void test_smlal_lane_x4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smls.za32.lane.vg4x4.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]], i32 15)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_smlsl_lane_x4_s8(uint32_t slice_base, svint8x4_t zn, svint8_t zm) __arm_streaming __arm_shared_za
+void test_smlsl_lane_x4_s8(uint32_t slice_base, svint8x4_t zn, svint8_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmls_lane_za32,_s8,_vg4x4,,)(slice_base, zn, zm, 15);
 }
@@ -1688,7 +1688,7 @@ void test_smlsl_lane_x4_s8(uint32_t slice_base, svint8x4_t zn, svint8_t zm) __ar
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smls.za64.lane.vg4x4.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_smlsl_lane_x4_s16(uint32_t slice_base, svint16x4_t zn, svint16_t zm) __arm_streaming __arm_shared_za
+void test_smlsl_lane_x4_s16(uint32_t slice_base, svint16x4_t zn, svint16_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmls_lane_za64,_s16,_vg4x4,,)(slice_base, zn, zm, 7);
 }
@@ -1711,7 +1711,7 @@ void test_smlsl_lane_x4_s16(uint32_t slice_base, svint16x4_t zn, svint16_t zm) _
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umls.za32.lane.vg4x4.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]], i32 15)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_smlsl_lane_x4_u8(uint32_t slice_base, svuint8x4_t zn, svuint8_t zm) __arm_streaming __arm_shared_za
+void test_smlsl_lane_x4_u8(uint32_t slice_base, svuint8x4_t zn, svuint8_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmls_lane_za32,_u8,_vg4x4,,)(slice_base, zn, zm, 15);
 }
@@ -1734,7 +1734,7 @@ void test_smlsl_lane_x4_u8(uint32_t slice_base, svuint8x4_t zn, svuint8_t zm) __
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umls.za64.lane.vg4x4.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_smlsl_lane_x4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16_t zm) __arm_streaming __arm_shared_za
+void test_smlsl_lane_x4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svmls_lane_za64,_u16,_vg4x4,,)(slice_base, zn, zm, 7);
 }
@@ -1759,7 +1759,7 @@ void test_smlsl_lane_x4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sumla.za32.lane.vg4x4.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]], i32 15)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_sumlall_lane_x4_s8(uint32_t slice_base, svint8x4_t zn, svuint8_t zm) __arm_streaming __arm_shared_za
+void test_sumlall_lane_x4_s8(uint32_t slice_base, svint8x4_t zn, svuint8_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svsumla_lane_za32,_s8,_vg4x4,,)(slice_base, zn, zm, 15);
 }
@@ -1784,7 +1784,7 @@ void test_sumlall_lane_x4_s8(uint32_t slice_base, svint8x4_t zn, svuint8_t zm) _
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.usmla.za32.lane.vg4x4.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]], i32 15)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_usmlall_lane_x4_s8(uint32_t slice_base, svuint8x4_t zn, svint8_t zm) __arm_streaming __arm_shared_za
+void test_usmlall_lane_x4_s8(uint32_t slice_base, svuint8x4_t zn, svint8_t zm) __arm_streaming __arm_inout("za")
 {
   SVE_ACLE_FUNC(svusmla_lane_za32,_u8,_vg4x4,,)(slice_base, zn, zm, 15);
 }
diff --git a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_mls.c b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_mls.c
index 6830a399e91d6..98eaa929cfb77 100644
--- a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_mls.c
+++ b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_mls.c
@@ -35,8 +35,8 @@
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmls.vg1x2.nxv4f32(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls2_f32(uint32_t slice_base, svfloat32x2_t zn, svfloat32x2_t zm) __arm_streaming __arm_shared_za {
-  SVE_ACLE_FUNC(svmls_za32,,_f32,,_vg1x2)(slice_base, zn, zm);
+void test_svmls2_f32(uint32_t slice_base, svfloat32x2_t zn, svfloat32x2_t zm) __arm_streaming __arm_inout("za") {
+  SVE_ACLE_FUNC(svmls_za32,_f32,_vg1x2,,)(slice_base, zn, zm);
 }
 
 // CHECK-LABEL: @test_svmls4_f32(
@@ -65,8 +65,8 @@ void test_svmls2_f32(uint32_t slice_base, svfloat32x2_t zn, svfloat32x2_t zm) __
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmls.vg1x4.nxv4f32(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[TMP4]],  [[TMP5]],  [[TMP6]],  [[TMP7]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls4_f32(uint32_t slice_base, svfloat32x4_t zn, svfloat32x4_t zm) __arm_streaming __arm_shared_za {
-  SVE_ACLE_FUNC(svmls_za32,,_f32,,_vg1x4)(slice_base, zn, zm);
+void test_svmls4_f32(uint32_t slice_base, svfloat32x4_t zn, svfloat32x4_t zm) __arm_streaming __arm_inout("za") {
+  SVE_ACLE_FUNC(svmls_za32,_f32,_vg1x4,,)(slice_base, zn, zm);
 }
 
 //
@@ -85,8 +85,8 @@ void test_svmls4_f32(uint32_t slice_base, svfloat32x4_t zn, svfloat32x4_t zm) __
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmls.single.vg1x2.nxv4f32(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_single2_f32(uint32_t slice_base, svfloat32x2_t zn, svfloat32_t zm) __arm_streaming __arm_shared_za {
-  SVE_ACLE_FUNC(svmls_single_za32,,_f32,,_vg1x2)(slice_base, zn, zm);
+void test_svmls_single2_f32(uint32_t slice_base, svfloat32x2_t zn, svfloat32_t zm) __arm_streaming __arm_inout("za") {
+  SVE_ACLE_FUNC(svmls,_single,_za32,_f32,_vg1x2)(slice_base, zn, zm);
 }
 
 // CHECK-LABEL: @test_svmls_single4_f32(
@@ -107,8 +107,8 @@ void test_svmls_single2_f32(uint32_t slice_base, svfloat32x2_t zn, svfloat32_t z
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmls.single.vg1x4.nxv4f32(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_single4_f32(uint32_t slice_base, svfloat32x4_t zn, svfloat32_t zm) __arm_streaming __arm_shared_za {
-  SVE_ACLE_FUNC(svmls_single_za32,,_f32,,_vg1x4)(slice_base, zn, zm);
+void test_svmls_single4_f32(uint32_t slice_base, svfloat32x4_t zn, svfloat32_t zm) __arm_streaming __arm_inout("za") {
+  SVE_ACLE_FUNC(svmls,_single,_za32,_f32,_vg1x4)(slice_base, zn, zm);
 }
 
 //
@@ -127,8 +127,8 @@ void test_svmls_single4_f32(uint32_t slice_base, svfloat32x4_t zn, svfloat32_t z
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmls.lane.vg1x2.nxv4f32(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]], i32 3)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_lane2_f32(uint32_t slice_base, svfloat32x2_t zn, svfloat32_t zm) __arm_streaming __arm_shared_za {
-  SVE_ACLE_FUNC(svmls_lane_za32,,_f32,,_vg1x2)(slice_base, zn, zm, 3);
+void test_svmls_lane2_f32(uint32_t slice_base, svfloat32x2_t zn, svfloat32_t zm) __arm_streaming __arm_inout("za") {
+  SVE_ACLE_FUNC(svmls_lane_za32,_f32,_vg1x2,,)(slice_base, zn, zm, 3);
 }
 
 // CHECK-LABEL: @test_svmls_lane4_f32(
@@ -149,8 +149,8 @@ void test_svmls_lane2_f32(uint32_t slice_base, svfloat32x2_t zn, svfloat32_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmls.lane.vg1x4.nxv4f32(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]], i32 3)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_lane4_f32(uint32_t slice_base, svfloat32x4_t zn, svfloat32_t zm) __arm_streaming __arm_shared_za {
-  SVE_ACLE_FUNC(svmls_lane_za32,,_f32,,_vg1x4)(slice_base, zn, zm, 3);
+void test_svmls_lane4_f32(uint32_t slice_base, svfloat32x4_t zn, svfloat32_t zm) __arm_streaming __arm_inout("za") {
+  SVE_ACLE_FUNC(svmls_lane_za32,_f32,_vg1x4,,)(slice_base, zn, zm, 3);
 }
 
 //
@@ -173,8 +173,8 @@ void test_svmls_lane4_f32(uint32_t slice_base, svfloat32x4_t zn, svfloat32_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmls.vg1x2.nxv2f64(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls2_f64(uint32_t slice_base, svfloat64x2_t zn, svfloat64x2_t zm) __arm_streaming __arm_shared_za {
-  SVE_ACLE_FUNC(svmls_za64,,_f64,,_vg1x2)(slice_base, zn, zm);
+void test_svmls2_f64(uint32_t slice_base, svfloat64x2_t zn, svfloat64x2_t zm) __arm_streaming __arm_inout("za") {
+  SVE_ACLE_FUNC(svmls_za64,_f64,_vg1x2,,)(slice_base, zn, zm);
 }
 
 // CHECK-LABEL: @test_svmls4_f64(
@@ -203,8 +203,8 @@ void test_svmls2_f64(uint32_t slice_base, svfloat64x2_t zn, svfloat64x2_t zm) __
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmls.vg1x4.nxv2f64(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[TMP4]],  [[TMP5]],  [[TMP6]],  [[TMP7]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls4_f64(uint32_t slice_base, svfloat64x4_t zn, svfloat64x4_t zm) __arm_streaming __arm_shared_za {
-  SVE_ACLE_FUNC(svmls_za64,,_f64,,_vg1x4)(slice_base, zn, zm);
+void test_svmls4_f64(uint32_t slice_base, svfloat64x4_t zn, svfloat64x4_t zm) __arm_streaming __arm_inout("za") {
+  SVE_ACLE_FUNC(svmls_za64,_f64,_vg1x4,,)(slice_base, zn, zm);
 }
 
 //
@@ -223,8 +223,8 @@ void test_svmls4_f64(uint32_t slice_base, svfloat64x4_t zn, svfloat64x4_t zm) __
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmls.single.vg1x2.nxv2f64(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_single2_f64(uint32_t slice_base, svfloat64x2_t zn, svfloat64_t zm) __arm_streaming __arm_shared_za {
-  SVE_ACLE_FUNC(svmls_single_za64,,_f64,,_vg1x2)(slice_base, zn, zm);
+void test_svmls_single2_f64(uint32_t slice_base, svfloat64x2_t zn, svfloat64_t zm) __arm_streaming __arm_inout("za") {
+  SVE_ACLE_FUNC(svmls,_single,_za64,_f64,_vg1x2)(slice_base, zn, zm);
 }
 
 // CHECK-LABEL: @test_svmls_single4_f64(
@@ -245,8 +245,8 @@ void test_svmls_single2_f64(uint32_t slice_base, svfloat64x2_t zn, svfloat64_t z
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmls.single.vg1x4.nxv2f64(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_single4_f64(uint32_t slice_base, svfloat64x4_t zn, svfloat64_t zm) __arm_streaming __arm_shared_za {
-  SVE_ACLE_FUNC(svmls_single_za64,,_f64,,_vg1x4)(slice_base, zn, zm);
+void test_svmls_single4_f64(uint32_t slice_base, svfloat64x4_t zn, svfloat64_t zm) __arm_streaming __arm_inout("za") {
+  SVE_ACLE_FUNC(svmls,_single,_za64,_f64,_vg1x4)(slice_base, zn, zm);
 }
 
 //
@@ -265,8 +265,8 @@ void test_svmls_single4_f64(uint32_t slice_base, svfloat64x4_t zn, svfloat64_t z
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmls.lane.vg1x2.nxv2f64(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]], i32 1)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_lane2_f64(uint32_t slice_base, svfloat64x2_t zn, svfloat64_t zm) __arm_streaming __arm_shared_za {
-  SVE_ACLE_FUNC(svmls_lane_za64,,_f64,,_vg1x2)(slice_base, zn, zm, 1);
+void test_svmls_lane2_f64(uint32_t slice_base, svfloat64x2_t zn, svfloat64_t zm) __arm_streaming __arm_inout("za") {
+  SVE_ACLE_FUNC(svmls_lane_za64,_f64,_vg1x2,,)(slice_base, zn, zm, 1);
 }
 
 // CHECK-LABEL: @test_svmls_lane4_f64(
@@ -287,6 +287,6 @@ void test_svmls_lane2_f64(uint32_t slice_base, svfloat64x2_t zn, svfloat64_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmls.lane.vg1x4.nxv2f64(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]], i32 1)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_lane4_f64(uint32_t slice_base, svfloat64x4_t zn, svfloat64_t zm) __arm_streaming __arm_shared_za {
-  SVE_ACLE_FUNC(svmls_lane_za64,,_f64,,_vg1x4)(slice_base, zn, zm, 1);
+void test_svmls_lane4_f64(uint32_t slice_base, svfloat64x4_t zn, svfloat64_t zm) __arm_streaming __arm_inout("za") {
+  SVE_ACLE_FUNC(svmls_lane_za64,_f64,_vg1x4,,)(slice_base, zn, zm, 1);
 }
diff --git a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_mlsl.c b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_mlsl.c
index 0a87d97f649c9..ea685079ebbca 100644
--- a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_mlsl.c
+++ b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_mlsl.c
@@ -35,7 +35,7 @@
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmlsl.vg2x2.nxv8f16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls2_f16(uint32_t slice_base, svfloat16x2_t zn, svfloat16x2_t zm) __arm_streaming __arm_shared_za
+void test_svmls2_f16(uint32_t slice_base, svfloat16x2_t zn, svfloat16x2_t zm) __arm_streaming __arm_inout("za")
 {
    SVE_ACLE_FUNC(svmls_za32,_f16,_vg2x2,,)(slice_base, zn, zm);
 }
@@ -58,7 +58,7 @@ void test_svmls2_f16(uint32_t slice_base, svfloat16x2_t zn, svfloat16x2_t zm) __
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmlsl.vg2x2.nxv8bf16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls2_bf16(uint32_t slice_base, svbfloat16x2_t zn, svbfloat16x2_t zm) __arm_streaming __arm_shared_za
+void test_svmls2_bf16(uint32_t slice_base, svbfloat16x2_t zn, svbfloat16x2_t zm) __arm_streaming __arm_inout("za")
 {
    SVE_ACLE_FUNC(svmls_za32,_bf16,_vg2x2,,)(slice_base, zn, zm);
 }
@@ -81,7 +81,7 @@ void test_svmls2_bf16(uint32_t slice_base, svbfloat16x2_t zn, svbfloat16x2_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umlsl.vg2x2.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16x2_t zm) __arm_streaming __arm_shared_za
+void test_svmls2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16x2_t zm) __arm_streaming __arm_inout("za")
 {
    SVE_ACLE_FUNC(svmls_za32,_u16,_vg2x2,,)(slice_base, zn, zm);
 }
@@ -104,7 +104,7 @@ void test_svmls2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16x2_t zm) __ar
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smlsl.vg2x2.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls2_s16(uint32_t slice_base, svint16x2_t zn, svint16x2_t zm) __arm_streaming __arm_shared_za
+void test_svmls2_s16(uint32_t slice_base, svint16x2_t zn, svint16x2_t zm) __arm_streaming __arm_inout("za")
 {
    SVE_ACLE_FUNC(svmls_za32,_s16,_vg2x2,,)(slice_base, zn, zm);
 }
@@ -135,7 +135,7 @@ void test_svmls2_s16(uint32_t slice_base, svint16x2_t zn, svint16x2_t zm) __arm_
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmlsl.vg2x4.nxv8f16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[TMP4]],  [[TMP5]],  [[TMP6]],  [[TMP7]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls4_f16(uint32_t slice_base, svfloat16x4_t zn, svfloat16x4_t zm) __arm_streaming __arm_shared_za
+void test_svmls4_f16(uint32_t slice_base, svfloat16x4_t zn, svfloat16x4_t zm) __arm_streaming __arm_inout("za")
 {
    SVE_ACLE_FUNC(svmls_za32,_f16,_vg2x4,,)(slice_base, zn, zm);
 }
@@ -166,7 +166,7 @@ void test_svmls4_f16(uint32_t slice_base, svfloat16x4_t zn, svfloat16x4_t zm) __
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmlsl.vg2x4.nxv8bf16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[TMP4]],  [[TMP5]],  [[TMP6]],  [[TMP7]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls4_bf16(uint32_t slice_base, svbfloat16x4_t zn, svbfloat16x4_t zm) __arm_streaming __arm_shared_za
+void test_svmls4_bf16(uint32_t slice_base, svbfloat16x4_t zn, svbfloat16x4_t zm) __arm_streaming __arm_inout("za")
 {
    SVE_ACLE_FUNC(svmls_za32,_bf16,_vg2x4,,)(slice_base, zn, zm);
 }
@@ -197,7 +197,7 @@ void test_svmls4_bf16(uint32_t slice_base, svbfloat16x4_t zn, svbfloat16x4_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umlsl.vg2x4.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[TMP4]],  [[TMP5]],  [[TMP6]],  [[TMP7]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16x4_t zm) __arm_streaming __arm_shared_za
+void test_svmls4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16x4_t zm) __arm_streaming __arm_inout("za")
 {
    SVE_ACLE_FUNC(svmls_za32,_u16,_vg2x4,,)(slice_base, zn, zm);
 }
@@ -228,7 +228,7 @@ void test_svmls4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16x4_t zm) __ar
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smlsl.vg2x4.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[TMP4]],  [[TMP5]],  [[TMP6]],  [[TMP7]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls4_s16(uint32_t slice_base, svint16x4_t zn, svint16x4_t zm) __arm_streaming __arm_shared_za
+void test_svmls4_s16(uint32_t slice_base, svint16x4_t zn, svint16x4_t zm) __arm_streaming __arm_inout("za")
 {
    SVE_ACLE_FUNC(svmls_za32,_s16,_vg2x4,,)(slice_base, zn, zm);
 }
@@ -245,7 +245,7 @@ void test_svmls4_s16(uint32_t slice_base, svint16x4_t zn, svint16x4_t zm) __arm_
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmlsl.single.vg2x1.nxv8f16(i32 [[SLICE_BASE:%.*]],  [[ZN:%.*]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_single1_f16(uint32_t slice_base, svfloat16_t zn, svfloat16_t zm) __arm_streaming __arm_shared_za
+void test_svmls_single1_f16(uint32_t slice_base, svfloat16_t zn, svfloat16_t zm) __arm_streaming __arm_inout("za")
 {
    SVE_ACLE_FUNC(svmls_za32,_f16,_vg2x1,,)(slice_base, zn, zm);
 }
@@ -260,7 +260,7 @@ void test_svmls_single1_f16(uint32_t slice_base, svfloat16_t zn, svfloat16_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmlsl.single.vg2x1.nxv8bf16(i32 [[SLICE_BASE:%.*]],  [[ZN:%.*]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_single1_bf16(uint32_t slice_base, svbfloat16_t zn, svbfloat16_t zm) __arm_streaming __arm_shared_za
+void test_svmls_single1_bf16(uint32_t slice_base, svbfloat16_t zn, svbfloat16_t zm) __arm_streaming __arm_inout("za")
 {
    SVE_ACLE_FUNC(svmls_za32,_bf16,_vg2x1,,)(slice_base, zn, zm);
 }
@@ -275,7 +275,7 @@ void test_svmls_single1_bf16(uint32_t slice_base, svbfloat16_t zn, svbfloat16_t
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umlsl.single.vg2x1.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[ZN:%.*]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_single1_u16(uint32_t slice_base, svuint16_t zn, svuint16_t zm) __arm_streaming __arm_shared_za
+void test_svmls_single1_u16(uint32_t slice_base, svuint16_t zn, svuint16_t zm) __arm_streaming __arm_inout("za")
 {
    SVE_ACLE_FUNC(svmls_za32,_u16,_vg2x1,,)(slice_base, zn, zm);
 }
@@ -290,7 +290,7 @@ void test_svmls_single1_u16(uint32_t slice_base, svuint16_t zn, svuint16_t zm) _
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smlsl.single.vg2x1.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[ZN:%.*]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_single1_s16(uint32_t slice_base, svint16_t zn, svint16_t zm) __arm_streaming __arm_shared_za
+void test_svmls_single1_s16(uint32_t slice_base, svint16_t zn, svint16_t zm) __arm_streaming __arm_inout("za")
 {
    SVE_ACLE_FUNC(svmls_za32,_s16,_vg2x1,,)(slice_base, zn, zm);
 }
@@ -309,9 +309,9 @@ void test_svmls_single1_s16(uint32_t slice_base, svint16_t zn, svint16_t zm) __a
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmlsl.single.vg2x2.nxv8f16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_single2_f16(uint32_t slice_base, svfloat16x2_t zn, svfloat16_t zm) __arm_streaming __arm_shared_za
+void test_svmls_single2_f16(uint32_t slice_base, svfloat16x2_t zn, svfloat16_t zm) __arm_streaming __arm_inout("za")
 {
-   SVE_ACLE_FUNC(svmls_single_za32,,_f16,,_vg2x2)(slice_base, zn, zm);
+   SVE_ACLE_FUNC(svmls,_single,_za32,_f16,_vg2x2)(slice_base, zn, zm);
 }
 
 // CHECK-LABEL: @test_svmls_single2_bf16(
@@ -328,9 +328,9 @@ void test_svmls_single2_f16(uint32_t slice_base, svfloat16x2_t zn, svfloat16_t z
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmlsl.single.vg2x2.nxv8bf16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_single2_bf16(uint32_t slice_base, svbfloat16x2_t zn, svbfloat16_t zm) __arm_streaming __arm_shared_za
+void test_svmls_single2_bf16(uint32_t slice_base, svbfloat16x2_t zn, svbfloat16_t zm) __arm_streaming __arm_inout("za")
 {
-   SVE_ACLE_FUNC(svmls_single_za32,,_bf16,,_vg2x2)(slice_base, zn, zm);
+   SVE_ACLE_FUNC(svmls,_single,_za32,_bf16,_vg2x2)(slice_base, zn, zm);
 }
 
 // CHECK-LABEL: @test_svmls_single2_u16(
@@ -347,9 +347,9 @@ void test_svmls_single2_bf16(uint32_t slice_base, svbfloat16x2_t zn, svbfloat16_
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umlsl.single.vg2x2.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_single2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16_t zm) __arm_streaming __arm_shared_za
+void test_svmls_single2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16_t zm) __arm_streaming __arm_inout("za")
 {
-   SVE_ACLE_FUNC(svmls_single_za32,,_u16,,_vg2x2)(slice_base, zn, zm);
+   SVE_ACLE_FUNC(svmls,_single,_za32,_u16,_vg2x2)(slice_base, zn, zm);
 }
 
 // CHECK-LABEL: @test_svmls_single2_s16(
@@ -366,9 +366,9 @@ void test_svmls_single2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smlsl.single.vg2x2.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_single2_s16(uint32_t slice_base, svint16x2_t zn, svint16_t zm) __arm_streaming __arm_shared_za
+void test_svmls_single2_s16(uint32_t slice_base, svint16x2_t zn, svint16_t zm) __arm_streaming __arm_inout("za")
 {
-   SVE_ACLE_FUNC(svmls_single_za32,,_s16,,_vg2x2)(slice_base, zn, zm);
+   SVE_ACLE_FUNC(svmls,_single,_za32,_s16,_vg2x2)(slice_base, zn, zm);
 }
 
 // CHECK-LABEL: @test_svmls_single4_f16(
@@ -389,9 +389,9 @@ void test_svmls_single2_s16(uint32_t slice_base, svint16x2_t zn, svint16_t zm) _
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmlsl.single.vg2x4.nxv8f16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_single4_f16(uint32_t slice_base, svfloat16x4_t zn, svfloat16_t zm) __arm_streaming __arm_shared_za
+void test_svmls_single4_f16(uint32_t slice_base, svfloat16x4_t zn, svfloat16_t zm) __arm_streaming __arm_inout("za")
 {
-   SVE_ACLE_FUNC(svmls_single_za32,,_f16,,_vg2x4)(slice_base, zn, zm);
+   SVE_ACLE_FUNC(svmls,_single,_za32,_f16,_vg2x4)(slice_base, zn, zm);
 }
 
 // CHECK-LABEL: @test_svmls_single4_bf16(
@@ -412,9 +412,9 @@ void test_svmls_single4_f16(uint32_t slice_base, svfloat16x4_t zn, svfloat16_t z
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmlsl.single.vg2x4.nxv8bf16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_single4_bf16(uint32_t slice_base, svbfloat16x4_t zn, svbfloat16_t zm) __arm_streaming __arm_shared_za
+void test_svmls_single4_bf16(uint32_t slice_base, svbfloat16x4_t zn, svbfloat16_t zm) __arm_streaming __arm_inout("za")
 {
-   SVE_ACLE_FUNC(svmls_single_za32,,_bf16,,_vg2x4)(slice_base, zn, zm);
+   SVE_ACLE_FUNC(svmls,_single,_za32,_bf16,_vg2x4)(slice_base, zn, zm);
 }
 
 // CHECK-LABEL: @test_svmls_single4_u16(
@@ -435,9 +435,9 @@ void test_svmls_single4_bf16(uint32_t slice_base, svbfloat16x4_t zn, svbfloat16_
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umlsl.single.vg2x4.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_single4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16_t zm) __arm_streaming __arm_shared_za
+void test_svmls_single4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16_t zm) __arm_streaming __arm_inout("za")
 {
-   SVE_ACLE_FUNC(svmls_single_za32,,_u16,,_vg2x4)(slice_base, zn, zm);
+   SVE_ACLE_FUNC(svmls,_single,_za32,_u16,_vg2x4)(slice_base, zn, zm);
 }
 
 // CHECK-LABEL: @test_svmls_single4_s16(
@@ -458,9 +458,9 @@ void test_svmls_single4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smlsl.single.vg2x4.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_single4_s16(uint32_t slice_base, svint16x4_t zn, svint16_t zm) __arm_streaming __arm_shared_za
+void test_svmls_single4_s16(uint32_t slice_base, svint16x4_t zn, svint16_t zm) __arm_streaming __arm_inout("za")
 {
-   SVE_ACLE_FUNC(svmls_single_za32,,_s16,,_vg2x4)(slice_base, zn, zm);
+   SVE_ACLE_FUNC(svmls,_single,_za32,_s16,_vg2x4)(slice_base, zn, zm);
 }
 
 //
@@ -477,9 +477,9 @@ void test_svmls_single4_s16(uint32_t slice_base, svint16x4_t zn, svint16_t zm) _
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmlsl.lane.vg2x1.nxv8f16(i32 [[SLICE_BASE:%.*]],  [[ZN:%.*]],  [[ZM:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_lane1_f16(uint32_t slice_base, svfloat16_t zn, svfloat16_t zm) __arm_streaming __arm_shared_za
+void test_svmls_lane1_f16(uint32_t slice_base, svfloat16_t zn, svfloat16_t zm) __arm_streaming __arm_inout("za")
 {
-   SVE_ACLE_FUNC(svmls_lane_za32,,_f16,,_vg2x1)(slice_base, zn, zm, 7);
+   SVE_ACLE_FUNC(svmls_lane_za32,_f16,_vg2x1,,)(slice_base, zn, zm, 7);
 }
 
 // CHECK-LABEL: @test_svmls_lane1_bf16(
@@ -492,9 +492,9 @@ void test_svmls_lane1_f16(uint32_t slice_base, svfloat16_t zn, svfloat16_t zm) _
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmlsl.lane.vg2x1.nxv8bf16(i32 [[SLICE_BASE:%.*]],  [[ZN:%.*]],  [[ZM:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_lane1_bf16(uint32_t slice_base, svbfloat16_t zn, svbfloat16_t zm) __arm_streaming __arm_shared_za
+void test_svmls_lane1_bf16(uint32_t slice_base, svbfloat16_t zn, svbfloat16_t zm) __arm_streaming __arm_inout("za")
 {
-   SVE_ACLE_FUNC(svmls_lane_za32,,_bf16,,_vg2x1)(slice_base, zn, zm, 7);
+   SVE_ACLE_FUNC(svmls_lane_za32,_bf16,_vg2x1,,)(slice_base, zn, zm, 7);
 }
 
 // CHECK-LABEL: @test_svmls_lane1_u16(
@@ -507,9 +507,9 @@ void test_svmls_lane1_bf16(uint32_t slice_base, svbfloat16_t zn, svbfloat16_t zm
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umlsl.lane.vg2x1.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[ZN:%.*]],  [[ZM:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_lane1_u16(uint32_t slice_base, svuint16_t zn, svuint16_t zm) __arm_streaming __arm_shared_za
+void test_svmls_lane1_u16(uint32_t slice_base, svuint16_t zn, svuint16_t zm) __arm_streaming __arm_inout("za")
 {
-   SVE_ACLE_FUNC(svmls_lane_za32,,_u16,,_vg2x1)(slice_base, zn, zm, 7);
+   SVE_ACLE_FUNC(svmls_lane_za32,_u16,_vg2x1,,)(slice_base, zn, zm, 7);
 }
 
 // CHECK-LABEL: @test_svmls_lane1_s16(
@@ -522,9 +522,9 @@ void test_svmls_lane1_u16(uint32_t slice_base, svuint16_t zn, svuint16_t zm) __a
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smlsl.lane.vg2x1.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[ZN:%.*]],  [[ZM:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_lane1_s16(uint32_t slice_base, svint16_t zn, svint16_t zm) __arm_streaming __arm_shared_za
+void test_svmls_lane1_s16(uint32_t slice_base, svint16_t zn, svint16_t zm) __arm_streaming __arm_inout("za")
 {
-   SVE_ACLE_FUNC(svmls_lane_za32,,_s16,,_vg2x1)(slice_base, zn, zm, 7);
+   SVE_ACLE_FUNC(svmls_lane_za32,_s16,_vg2x1,,)(slice_base, zn, zm, 7);
 }
 
 // CHECK-LABEL: @test_svmls_lane2_f16(
@@ -541,9 +541,9 @@ void test_svmls_lane1_s16(uint32_t slice_base, svint16_t zn, svint16_t zm) __arm
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmlsl.lane.vg2x2.nxv8f16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_lane2_f16(uint32_t slice_base, svfloat16x2_t zn, svfloat16_t zm) __arm_streaming __arm_shared_za
+void test_svmls_lane2_f16(uint32_t slice_base, svfloat16x2_t zn, svfloat16_t zm) __arm_streaming __arm_inout("za")
 {
-   SVE_ACLE_FUNC(svmls_lane_za32,,_f16,,_vg2x2)(slice_base, zn, zm, 7);
+   SVE_ACLE_FUNC(svmls_lane_za32,_f16,_vg2x2,,)(slice_base, zn, zm, 7);
 }
 
 // CHECK-LABEL: @test_svmls_lane2_bf16(
@@ -560,9 +560,9 @@ void test_svmls_lane2_f16(uint32_t slice_base, svfloat16x2_t zn, svfloat16_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmlsl.lane.vg2x2.nxv8bf16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_lane2_bf16(uint32_t slice_base, svbfloat16x2_t zn, svbfloat16_t zm) __arm_streaming __arm_shared_za
+void test_svmls_lane2_bf16(uint32_t slice_base, svbfloat16x2_t zn, svbfloat16_t zm) __arm_streaming __arm_inout("za")
 {
-   SVE_ACLE_FUNC(svmls_lane_za32,,_bf16,,_vg2x2)(slice_base, zn, zm, 7);
+   SVE_ACLE_FUNC(svmls_lane_za32,_bf16,_vg2x2,,)(slice_base, zn, zm, 7);
 }
 
 // CHECK-LABEL: @test_svmls_lane2_u16(
@@ -579,9 +579,9 @@ void test_svmls_lane2_bf16(uint32_t slice_base, svbfloat16x2_t zn, svbfloat16_t
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umlsl.lane.vg2x2.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_lane2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16_t zm) __arm_streaming __arm_shared_za
+void test_svmls_lane2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16_t zm) __arm_streaming __arm_inout("za")
 {
-   SVE_ACLE_FUNC(svmls_lane_za32,,_u16,,_vg2x2)(slice_base, zn, zm, 7);
+   SVE_ACLE_FUNC(svmls_lane_za32,_u16,_vg2x2,,)(slice_base, zn, zm, 7);
 }
 
 // CHECK-LABEL: @test_svmls_lane2_s16(
@@ -598,9 +598,9 @@ void test_svmls_lane2_u16(uint32_t slice_base, svuint16x2_t zn, svuint16_t zm) _
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smlsl.lane.vg2x2.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_lane2_s16(uint32_t slice_base, svint16x2_t zn, svint16_t zm) __arm_streaming __arm_shared_za
+void test_svmls_lane2_s16(uint32_t slice_base, svint16x2_t zn, svint16_t zm) __arm_streaming __arm_inout("za")
 {
-   SVE_ACLE_FUNC(svmls_lane_za32,,_s16,,_vg2x2)(slice_base, zn, zm, 7);
+   SVE_ACLE_FUNC(svmls_lane_za32,_s16,_vg2x2,,)(slice_base, zn, zm, 7);
 }
 
 // CHECK-LABEL: @test_svmls_lane4_f16(
@@ -621,9 +621,9 @@ void test_svmls_lane2_s16(uint32_t slice_base, svint16x2_t zn, svint16_t zm) __a
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmlsl.lane.vg2x4.nxv8f16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_lane4_f16(uint32_t slice_base, svfloat16x4_t zn, svfloat16_t zm) __arm_streaming __arm_shared_za
+void test_svmls_lane4_f16(uint32_t slice_base, svfloat16x4_t zn, svfloat16_t zm) __arm_streaming __arm_inout("za")
 {
-   SVE_ACLE_FUNC(svmls_lane_za32,,_f16,,_vg2x4)(slice_base, zn, zm, 7);
+   SVE_ACLE_FUNC(svmls_lane_za32,_f16,_vg2x4,,)(slice_base, zn, zm, 7);
 }
 
 // CHECK-LABEL: @test_svmls_lane4_bf16(
@@ -644,9 +644,9 @@ void test_svmls_lane4_f16(uint32_t slice_base, svfloat16x4_t zn, svfloat16_t zm)
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fmlsl.lane.vg2x4.nxv8bf16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_lane4_bf16(uint32_t slice_base, svbfloat16x4_t zn, svbfloat16_t zm) __arm_streaming __arm_shared_za
+void test_svmls_lane4_bf16(uint32_t slice_base, svbfloat16x4_t zn, svbfloat16_t zm) __arm_streaming __arm_inout("za")
 {
-   SVE_ACLE_FUNC(svmls_lane_za32,,_bf16,,_vg2x4)(slice_base, zn, zm, 7);
+   SVE_ACLE_FUNC(svmls_lane_za32,_bf16,_vg2x4,,)(slice_base, zn, zm, 7);
 }
 
 // CHECK-LABEL: @test_svmls_lane4_u16(
@@ -667,9 +667,9 @@ void test_svmls_lane4_bf16(uint32_t slice_base, svbfloat16x4_t zn, svbfloat16_t
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umlsl.lane.vg2x4.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_lane4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16_t zm) __arm_streaming __arm_shared_za
+void test_svmls_lane4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16_t zm) __arm_streaming __arm_inout("za")
 {
-   SVE_ACLE_FUNC(svmls_lane_za32,,_u16,,_vg2x4)(slice_base, zn, zm, 7);
+   SVE_ACLE_FUNC(svmls_lane_za32,_u16,_vg2x4,,)(slice_base, zn, zm, 7);
 }
 
 // CHECK-LABEL: @test_svmls_lane4_s16(
@@ -690,7 +690,7 @@ void test_svmls_lane4_u16(uint32_t slice_base, svuint16x4_t zn, svuint16_t zm) _
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smlsl.lane.vg2x4.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]], i32 7)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmls_lane4_s16(uint32_t slice_base, svint16x4_t zn, svint16_t zm) __arm_streaming __arm_shared_za
+void test_svmls_lane4_s16(uint32_t slice_base, svint16x4_t zn, svint16_t zm) __arm_streaming __arm_inout("za")
 {
-   SVE_ACLE_FUNC(svmls_lane_za32,,_s16,,_vg2x4)(slice_base, zn, zm, 7);
+   SVE_ACLE_FUNC(svmls_lane_za32,_s16,_vg2x4,,)(slice_base, zn, zm, 7);
 }
diff --git a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_mop.c b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_mop.c
index eef99eb1b75a3..a2904dc1a6025 100644
--- a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_mop.c
+++ b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_mop.c
@@ -33,7 +33,7 @@
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smopa.za32.nxv8i16(i32 3,  [[TMP0]],  [[TMP1]],  [[ZN:%.*]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmopa_s16(svbool_t pn, svbool_t pm, svint16_t zn, svint16_t zm) __arm_streaming __arm_shared_za {
+void test_svmopa_s16(svbool_t pn, svbool_t pm, svint16_t zn, svint16_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svmopa_za32,_s16,_m,)(3, pn, pm, zn, zm);
 }
 
@@ -51,7 +51,7 @@ void test_svmopa_s16(svbool_t pn, svbool_t pm, svint16_t zn, svint16_t zm) __arm
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umopa.za32.nxv8i16(i32 3,  [[TMP0]],  [[TMP1]],  [[ZN:%.*]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmopa_u16(svbool_t pn, svbool_t pm, svuint16_t zn, svuint16_t zm) __arm_streaming __arm_shared_za {
+void test_svmopa_u16(svbool_t pn, svbool_t pm, svuint16_t zn, svuint16_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svmopa_za32,_u16,_m,)(3, pn, pm, zn, zm);
 }
 
@@ -71,7 +71,7 @@ void test_svmopa_u16(svbool_t pn, svbool_t pm, svuint16_t zn, svuint16_t zm) __a
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.smops.za32.nxv8i16(i32 3,  [[TMP0]],  [[TMP1]],  [[ZN:%.*]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmops_s16(svbool_t pn, svbool_t pm, svint16_t zn, svint16_t zm) __arm_streaming __arm_shared_za {
+void test_svmops_s16(svbool_t pn, svbool_t pm, svint16_t zn, svint16_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svmops_za32,_s16,_m,)(3, pn, pm, zn, zm);
 }
 
@@ -89,6 +89,6 @@ void test_svmops_s16(svbool_t pn, svbool_t pm, svint16_t zn, svint16_t zm) __arm
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.umops.za32.nxv8i16(i32 3,  [[TMP0]],  [[TMP1]],  [[ZN:%.*]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svmops_u16(svbool_t pn, svbool_t pm, svuint16_t zn, svuint16_t zm) __arm_streaming __arm_shared_za {
+void test_svmops_u16(svbool_t pn, svbool_t pm, svuint16_t zn, svuint16_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svmops_za32,_u16,_m,)(3, pn, pm, zn, zm);
 }
diff --git a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_read.c b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_read.c
index 583a7fc815472..028a3d7b155dd 100644
--- a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_read.c
+++ b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_read.c
@@ -25,7 +25,7 @@
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv32i8.nxv16i8( [[TMP2]],  [[TMP3]], i64 16)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svuint8x2_t test_svread_ver_za8_u8_vg2(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint8x2_t test_svread_ver_za8_u8_vg2(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_ver_za8_u8_vg2(0, base);
 }
 
@@ -47,7 +47,7 @@ svuint8x2_t test_svread_ver_za8_u8_vg2(uint32_t base) __arm_streaming __arm_shar
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv32i8.nxv16i8( [[TMP2]],  [[TMP3]], i64 16)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svint8x2_t test_svread_ver_za8_s8_vg2(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint8x2_t test_svread_ver_za8_s8_vg2(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_ver_za8_s8_vg2(0, base);
 }
 
@@ -69,7 +69,7 @@ svint8x2_t test_svread_ver_za8_s8_vg2(uint32_t base) __arm_streaming __arm_share
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv32i8.nxv16i8( [[TMP2]],  [[TMP3]], i64 16)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svuint8x2_t test_svread_hor_za8_u8_vg2(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint8x2_t test_svread_hor_za8_u8_vg2(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_hor_za8_u8_vg2(0, base);
 }
 
@@ -91,7 +91,7 @@ svuint8x2_t test_svread_hor_za8_u8_vg2(uint32_t base) __arm_streaming __arm_shar
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv32i8.nxv16i8( [[TMP2]],  [[TMP3]], i64 16)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svint8x2_t test_svread_hor_za8_s8_vg2(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint8x2_t test_svread_hor_za8_s8_vg2(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_hor_za8_s8_vg2(0, base);
 }
 
@@ -121,7 +121,7 @@ svint8x2_t test_svread_hor_za8_s8_vg2(uint32_t base) __arm_streaming __arm_share
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv64i8.nxv16i8( [[TMP6]],  [[TMP7]], i64 48)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svuint8x4_t test_svread_hor_za8_u8_vg4(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint8x4_t test_svread_hor_za8_u8_vg4(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_hor_za8_u8_vg4(0, base);
 }
 
@@ -151,7 +151,7 @@ svuint8x4_t test_svread_hor_za8_u8_vg4(uint32_t base) __arm_streaming __arm_shar
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv64i8.nxv16i8( [[TMP6]],  [[TMP7]], i64 48)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svint8x4_t test_svread_hor_za8_s8_vg4(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint8x4_t test_svread_hor_za8_s8_vg4(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_hor_za8_s8_vg4(0, base);
 }
 
@@ -181,7 +181,7 @@ svint8x4_t test_svread_hor_za8_s8_vg4(uint32_t base) __arm_streaming __arm_share
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv64i8.nxv16i8( [[TMP6]],  [[TMP7]], i64 48)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svuint8x4_t test_svread_ver_za8_u8_vg4(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint8x4_t test_svread_ver_za8_u8_vg4(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_ver_za8_u8_vg4(0, base);
 }
 
@@ -211,7 +211,7 @@ svuint8x4_t test_svread_ver_za8_u8_vg4(uint32_t base) __arm_streaming __arm_shar
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv64i8.nxv16i8( [[TMP6]],  [[TMP7]], i64 48)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svint8x4_t test_svread_ver_za8_s8_vg4(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint8x4_t test_svread_ver_za8_s8_vg4(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_ver_za8_s8_vg4(0, base);
 }
 
@@ -233,7 +233,7 @@ svint8x4_t test_svread_ver_za8_s8_vg4(uint32_t base) __arm_streaming __arm_share
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv16i16.nxv8i16( [[TMP2]],  [[TMP3]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svuint16x2_t test_svread_hor_za16_u16_vg2(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint16x2_t test_svread_hor_za16_u16_vg2(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_hor_za16_u16_vg2(1, base);
 }
 
@@ -255,7 +255,7 @@ svuint16x2_t test_svread_hor_za16_u16_vg2(uint32_t base) __arm_streaming __arm_s
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv16bf16.nxv8bf16( [[TMP2]],  [[TMP3]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svbfloat16x2_t test_svread_hor_za16_bf16_vg2(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svbfloat16x2_t test_svread_hor_za16_bf16_vg2(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_hor_za16_bf16_vg2(1, base);
 }
 
@@ -277,7 +277,7 @@ svbfloat16x2_t test_svread_hor_za16_bf16_vg2(uint32_t base) __arm_streaming __ar
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv16f16.nxv8f16( [[TMP2]],  [[TMP3]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svfloat16x2_t test_svread_hor_za16_f16_vg2(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svfloat16x2_t test_svread_hor_za16_f16_vg2(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_hor_za16_f16_vg2(1, base);
 }
 
@@ -299,7 +299,7 @@ svfloat16x2_t test_svread_hor_za16_f16_vg2(uint32_t base) __arm_streaming __arm_
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv16i16.nxv8i16( [[TMP2]],  [[TMP3]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svint16x2_t test_svread_hor_za16_s16_vg2(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint16x2_t test_svread_hor_za16_s16_vg2(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_hor_za16_s16_vg2(1, base);
 }
 
@@ -321,7 +321,7 @@ svint16x2_t test_svread_hor_za16_s16_vg2(uint32_t base) __arm_streaming __arm_sh
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv16i16.nxv8i16( [[TMP2]],  [[TMP3]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svuint16x2_t test_svread_ver_za16_u16_vg2(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint16x2_t test_svread_ver_za16_u16_vg2(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_ver_za16_u16_vg2(1, base);
 }
 
@@ -343,7 +343,7 @@ svuint16x2_t test_svread_ver_za16_u16_vg2(uint32_t base) __arm_streaming __arm_s
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv16bf16.nxv8bf16( [[TMP2]],  [[TMP3]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svbfloat16x2_t test_svread_ver_za16_bf16_vg2(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svbfloat16x2_t test_svread_ver_za16_bf16_vg2(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_ver_za16_bf16_vg2(1, base);
 }
 
@@ -365,7 +365,7 @@ svbfloat16x2_t test_svread_ver_za16_bf16_vg2(uint32_t base) __arm_streaming __ar
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv16f16.nxv8f16( [[TMP2]],  [[TMP3]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svfloat16x2_t test_svread_ver_za16_f16_vg2(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svfloat16x2_t test_svread_ver_za16_f16_vg2(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_ver_za16_f16_vg2(1, base);
 }
 
@@ -387,7 +387,7 @@ svfloat16x2_t test_svread_ver_za16_f16_vg2(uint32_t base) __arm_streaming __arm_
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv16i16.nxv8i16( [[TMP2]],  [[TMP3]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svint16x2_t test_svread_ver_za16_s16_vg2(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint16x2_t test_svread_ver_za16_s16_vg2(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_ver_za16_s16_vg2(1, base);
 }
 
@@ -417,7 +417,7 @@ svint16x2_t test_svread_ver_za16_s16_vg2(uint32_t base) __arm_streaming __arm_sh
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv32i16.nxv8i16( [[TMP6]],  [[TMP7]], i64 24)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svuint16x4_t test_svread_hor_za16_u16_vg4(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint16x4_t test_svread_hor_za16_u16_vg4(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_hor_za16_u16_vg4(1, base);
 }
 
@@ -447,7 +447,7 @@ svuint16x4_t test_svread_hor_za16_u16_vg4(uint32_t base) __arm_streaming __arm_s
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv32bf16.nxv8bf16( [[TMP6]],  [[TMP7]], i64 24)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svbfloat16x4_t test_svread_hor_za16_bf16_vg4(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svbfloat16x4_t test_svread_hor_za16_bf16_vg4(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_hor_za16_bf16_vg4(1, base);
 }
 
@@ -477,7 +477,7 @@ svbfloat16x4_t test_svread_hor_za16_bf16_vg4(uint32_t base) __arm_streaming __ar
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv32f16.nxv8f16( [[TMP6]],  [[TMP7]], i64 24)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svfloat16x4_t test_svread_hor_za16_f16_vg4(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svfloat16x4_t test_svread_hor_za16_f16_vg4(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_hor_za16_f16_vg4(1, base);
 }
 
@@ -507,7 +507,7 @@ svfloat16x4_t test_svread_hor_za16_f16_vg4(uint32_t base) __arm_streaming __arm_
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv32i16.nxv8i16( [[TMP6]],  [[TMP7]], i64 24)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svint16x4_t test_svread_hor_za16_s16_vg4(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint16x4_t test_svread_hor_za16_s16_vg4(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_hor_za16_s16_vg4(1, base);
 }
 
@@ -537,7 +537,7 @@ svint16x4_t test_svread_hor_za16_s16_vg4(uint32_t base) __arm_streaming __arm_sh
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv32i16.nxv8i16( [[TMP6]],  [[TMP7]], i64 24)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svuint16x4_t test_svread_ver_za16_u16_vg4(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint16x4_t test_svread_ver_za16_u16_vg4(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_ver_za16_u16_vg4(1, base);
 }
 
@@ -567,7 +567,7 @@ svuint16x4_t test_svread_ver_za16_u16_vg4(uint32_t base) __arm_streaming __arm_s
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv32bf16.nxv8bf16( [[TMP6]],  [[TMP7]], i64 24)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svbfloat16x4_t test_svread_ver_za16_bf16_vg4(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svbfloat16x4_t test_svread_ver_za16_bf16_vg4(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_ver_za16_bf16_vg4(1, base);
 }
 
@@ -597,7 +597,7 @@ svbfloat16x4_t test_svread_ver_za16_bf16_vg4(uint32_t base) __arm_streaming __ar
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv32f16.nxv8f16( [[TMP6]],  [[TMP7]], i64 24)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svfloat16x4_t test_svread_ver_za16_f16_vg4(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svfloat16x4_t test_svread_ver_za16_f16_vg4(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_ver_za16_f16_vg4(1, base);
 }
 
@@ -627,7 +627,7 @@ svfloat16x4_t test_svread_ver_za16_f16_vg4(uint32_t base) __arm_streaming __arm_
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv32i16.nxv8i16( [[TMP6]],  [[TMP7]], i64 24)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svint16x4_t test_svread_ver_za16_s16_vg4(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint16x4_t test_svread_ver_za16_s16_vg4(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_ver_za16_s16_vg4(1, base);
 }
 
@@ -649,7 +649,7 @@ svint16x4_t test_svread_ver_za16_s16_vg4(uint32_t base) __arm_streaming __arm_sh
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv8i32.nxv4i32( [[TMP2]],  [[TMP3]], i64 4)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svuint32x2_t test_svread_hor_za32_u32_vg2(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint32x2_t test_svread_hor_za32_u32_vg2(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_hor_za32_u32_vg2(3, base);
 }
 
@@ -671,7 +671,7 @@ svuint32x2_t test_svread_hor_za32_u32_vg2(uint32_t base) __arm_streaming __arm_s
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv8f32.nxv4f32( [[TMP2]],  [[TMP3]], i64 4)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svfloat32x2_t test_svread_hor_za32_f32_vg2(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svfloat32x2_t test_svread_hor_za32_f32_vg2(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_hor_za32_f32_vg2(3, base);
 }
 
@@ -693,7 +693,7 @@ svfloat32x2_t test_svread_hor_za32_f32_vg2(uint32_t base) __arm_streaming __arm_
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv8i32.nxv4i32( [[TMP2]],  [[TMP3]], i64 4)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svint32x2_t test_svread_hor_za32_s32_vg2(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint32x2_t test_svread_hor_za32_s32_vg2(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_hor_za32_s32_vg2(3, base);
 }
 
@@ -715,7 +715,7 @@ svint32x2_t test_svread_hor_za32_s32_vg2(uint32_t base) __arm_streaming __arm_sh
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv8i32.nxv4i32( [[TMP2]],  [[TMP3]], i64 4)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svuint32x2_t test_svread_ver_za32_u32_vg2(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint32x2_t test_svread_ver_za32_u32_vg2(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_ver_za32_u32_vg2(3, base);
 }
 
@@ -737,7 +737,7 @@ svuint32x2_t test_svread_ver_za32_u32_vg2(uint32_t base) __arm_streaming __arm_s
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv8f32.nxv4f32( [[TMP2]],  [[TMP3]], i64 4)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svfloat32x2_t test_svread_ver_za32_f32_vg2(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svfloat32x2_t test_svread_ver_za32_f32_vg2(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_ver_za32_f32_vg2(3, base);
 }
 
@@ -759,7 +759,7 @@ svfloat32x2_t test_svread_ver_za32_f32_vg2(uint32_t base) __arm_streaming __arm_
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv8i32.nxv4i32( [[TMP2]],  [[TMP3]], i64 4)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svint32x2_t test_svread_ver_za32_s32_vg2(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint32x2_t test_svread_ver_za32_s32_vg2(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_ver_za32_s32_vg2(3, base);
 }
 
@@ -789,7 +789,7 @@ svint32x2_t test_svread_ver_za32_s32_vg2(uint32_t base) __arm_streaming __arm_sh
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv16i32.nxv4i32( [[TMP6]],  [[TMP7]], i64 12)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svuint32x4_t test_svread_hor_za32_u32_vg4(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint32x4_t test_svread_hor_za32_u32_vg4(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_hor_za32_u32_vg4(3, base);
 }
 
@@ -819,7 +819,7 @@ svuint32x4_t test_svread_hor_za32_u32_vg4(uint32_t base) __arm_streaming __arm_s
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv16f32.nxv4f32( [[TMP6]],  [[TMP7]], i64 12)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svfloat32x4_t test_svread_hor_za32_f32_vg4(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svfloat32x4_t test_svread_hor_za32_f32_vg4(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_hor_za32_f32_vg4(3, base);
 }
 
@@ -849,7 +849,7 @@ svfloat32x4_t test_svread_hor_za32_f32_vg4(uint32_t base) __arm_streaming __arm_
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv16i32.nxv4i32( [[TMP6]],  [[TMP7]], i64 12)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svint32x4_t test_svread_hor_za32_s32_vg4(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint32x4_t test_svread_hor_za32_s32_vg4(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_hor_za32_s32_vg4(3, base);
 }
 
@@ -879,7 +879,7 @@ svint32x4_t test_svread_hor_za32_s32_vg4(uint32_t base) __arm_streaming __arm_sh
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv16i32.nxv4i32( [[TMP6]],  [[TMP7]], i64 12)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svuint32x4_t test_svread_ver_za32_u32_vg4(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint32x4_t test_svread_ver_za32_u32_vg4(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_ver_za32_u32_vg4(3, base);
 }
 
@@ -909,7 +909,7 @@ svuint32x4_t test_svread_ver_za32_u32_vg4(uint32_t base) __arm_streaming __arm_s
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv16f32.nxv4f32( [[TMP6]],  [[TMP7]], i64 12)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svfloat32x4_t test_svread_ver_za32_f32_vg4(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svfloat32x4_t test_svread_ver_za32_f32_vg4(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_ver_za32_f32_vg4(3, base);
 }
 
@@ -939,7 +939,7 @@ svfloat32x4_t test_svread_ver_za32_f32_vg4(uint32_t base) __arm_streaming __arm_
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv16i32.nxv4i32( [[TMP6]],  [[TMP7]], i64 12)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svint32x4_t test_svread_ver_za32_s32_vg4(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint32x4_t test_svread_ver_za32_s32_vg4(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_ver_za32_s32_vg4(3, base);
 }
 
@@ -961,7 +961,7 @@ svint32x4_t test_svread_ver_za32_s32_vg4(uint32_t base) __arm_streaming __arm_sh
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv4i64.nxv2i64( [[TMP2]],  [[TMP3]], i64 2)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svuint64x2_t test_svread_hor_za64_u64_vg2(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint64x2_t test_svread_hor_za64_u64_vg2(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_hor_za64_u64_vg2(7, base);
 }
 
@@ -983,7 +983,7 @@ svuint64x2_t test_svread_hor_za64_u64_vg2(uint32_t base) __arm_streaming __arm_s
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv4f64.nxv2f64( [[TMP2]],  [[TMP3]], i64 2)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svfloat64x2_t test_svread_hor_za64_f64_vg2(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svfloat64x2_t test_svread_hor_za64_f64_vg2(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_hor_za64_f64_vg2(7, base);
 }
 
@@ -1005,7 +1005,7 @@ svfloat64x2_t test_svread_hor_za64_f64_vg2(uint32_t base) __arm_streaming __arm_
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv4i64.nxv2i64( [[TMP2]],  [[TMP3]], i64 2)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svint64x2_t test_svread_hor_za64_s64_vg2(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint64x2_t test_svread_hor_za64_s64_vg2(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_hor_za64_s64_vg2(7, base);
 }
 
@@ -1027,7 +1027,7 @@ svint64x2_t test_svread_hor_za64_s64_vg2(uint32_t base) __arm_streaming __arm_sh
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv4i64.nxv2i64( [[TMP2]],  [[TMP3]], i64 2)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svuint64x2_t test_svread_ver_za64_u64_vg2(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint64x2_t test_svread_ver_za64_u64_vg2(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_ver_za64_u64_vg2(7, base);
 }
 
@@ -1049,7 +1049,7 @@ svuint64x2_t test_svread_ver_za64_u64_vg2(uint32_t base) __arm_streaming __arm_s
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv4f64.nxv2f64( [[TMP2]],  [[TMP3]], i64 2)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svfloat64x2_t test_svread_ver_za64_f64_vg2(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svfloat64x2_t test_svread_ver_za64_f64_vg2(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_ver_za64_f64_vg2(7, base);
 }
 
@@ -1071,7 +1071,7 @@ svfloat64x2_t test_svread_ver_za64_f64_vg2(uint32_t base) __arm_streaming __arm_
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv4i64.nxv2i64( [[TMP2]],  [[TMP3]], i64 2)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svint64x2_t test_svread_ver_za64_s64_vg2(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint64x2_t test_svread_ver_za64_s64_vg2(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_ver_za64_s64_vg2(7, base);
 }
 
@@ -1101,7 +1101,7 @@ svint64x2_t test_svread_ver_za64_s64_vg2(uint32_t base) __arm_streaming __arm_sh
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv8i64.nxv2i64( [[TMP6]],  [[TMP7]], i64 6)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svuint64x4_t test_svread_hor_za64_u64_vg4(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint64x4_t test_svread_hor_za64_u64_vg4(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_hor_za64_u64_vg4(7, base);
 }
 
@@ -1131,7 +1131,7 @@ svuint64x4_t test_svread_hor_za64_u64_vg4(uint32_t base) __arm_streaming __arm_s
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv8f64.nxv2f64( [[TMP6]],  [[TMP7]], i64 6)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svfloat64x4_t test_svread_hor_za64_f64_vg4(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svfloat64x4_t test_svread_hor_za64_f64_vg4(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_hor_za64_f64_vg4(7, base);
 }
 
@@ -1161,7 +1161,7 @@ svfloat64x4_t test_svread_hor_za64_f64_vg4(uint32_t base) __arm_streaming __arm_
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv8i64.nxv2i64( [[TMP6]],  [[TMP7]], i64 6)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svint64x4_t test_svread_hor_za64_s64_vg4(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint64x4_t test_svread_hor_za64_s64_vg4(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_hor_za64_s64_vg4(7, base);
 }
 
@@ -1191,7 +1191,7 @@ svint64x4_t test_svread_hor_za64_s64_vg4(uint32_t base) __arm_streaming __arm_sh
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv8i64.nxv2i64( [[TMP6]],  [[TMP7]], i64 6)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svuint64x4_t test_svread_ver_za64_u64_vg4(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint64x4_t test_svread_ver_za64_u64_vg4(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_ver_za64_u64_vg4(7, base);
 }
 
@@ -1221,7 +1221,7 @@ svuint64x4_t test_svread_ver_za64_u64_vg4(uint32_t base) __arm_streaming __arm_s
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv8f64.nxv2f64( [[TMP6]],  [[TMP7]], i64 6)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svfloat64x4_t test_svread_ver_za64_f64_vg4(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svfloat64x4_t test_svread_ver_za64_f64_vg4(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_ver_za64_f64_vg4(7, base);
 }
 
@@ -1251,7 +1251,7 @@ svfloat64x4_t test_svread_ver_za64_f64_vg4(uint32_t base) __arm_streaming __arm_
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv8i64.nxv2i64( [[TMP6]],  [[TMP7]], i64 6)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svint64x4_t test_svread_ver_za64_s64_vg4(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint64x4_t test_svread_ver_za64_s64_vg4(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_ver_za64_s64_vg4(7, base);
 }
 
@@ -1273,7 +1273,7 @@ svint64x4_t test_svread_ver_za64_s64_vg4(uint32_t base) __arm_streaming __arm_sh
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv32i8.nxv16i8( [[TMP2]],  [[TMP3]], i64 16)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svint8x2_t test_svread_za8_s8_vg1x2(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint8x2_t test_svread_za8_s8_vg1x2(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_za8_s8_vg1x2(base);
 }
 
@@ -1295,7 +1295,7 @@ svint8x2_t test_svread_za8_s8_vg1x2(uint32_t base) __arm_streaming __arm_shared_
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv32i8.nxv16i8( [[TMP2]],  [[TMP3]], i64 16)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svuint8x2_t test_svread_za8_u8_vg1x2(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint8x2_t test_svread_za8_u8_vg1x2(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_za8_u8_vg1x2(base);
 }
 
@@ -1317,7 +1317,7 @@ svuint8x2_t test_svread_za8_u8_vg1x2(uint32_t base) __arm_streaming __arm_shared
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv16i16.nxv8i16( [[TMP2]],  [[TMP3]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svint16x2_t test_svread_za16_s16_vg1x2(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint16x2_t test_svread_za16_s16_vg1x2(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_za16_s16_vg1x2(base);
 }
 
@@ -1339,7 +1339,7 @@ svint16x2_t test_svread_za16_s16_vg1x2(uint32_t base) __arm_streaming __arm_shar
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv16i16.nxv8i16( [[TMP2]],  [[TMP3]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svuint16x2_t test_svread_za16_u16_vg1x2(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint16x2_t test_svread_za16_u16_vg1x2(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_za16_u16_vg1x2(base);
 }
 
@@ -1361,7 +1361,7 @@ svuint16x2_t test_svread_za16_u16_vg1x2(uint32_t base) __arm_streaming __arm_sha
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv16bf16.nxv8bf16( [[TMP2]],  [[TMP3]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svbfloat16x2_t test_svread_za16_bf16_vg1x2(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svbfloat16x2_t test_svread_za16_bf16_vg1x2(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_za16_bf16_vg1x2(base);
 }
 
@@ -1383,7 +1383,7 @@ svbfloat16x2_t test_svread_za16_bf16_vg1x2(uint32_t base) __arm_streaming __arm_
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv16f16.nxv8f16( [[TMP2]],  [[TMP3]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svfloat16x2_t test_svread_za16_f16_vg1x2(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svfloat16x2_t test_svread_za16_f16_vg1x2(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_za16_f16_vg1x2(base);
 }
 
@@ -1405,7 +1405,7 @@ svfloat16x2_t test_svread_za16_f16_vg1x2(uint32_t base) __arm_streaming __arm_sh
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv8i32.nxv4i32( [[TMP2]],  [[TMP3]], i64 4)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svint32x2_t test_svread_za32_s32_vg1x2(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint32x2_t test_svread_za32_s32_vg1x2(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_za32_s32_vg1x2(base);
 }
 
@@ -1427,7 +1427,7 @@ svint32x2_t test_svread_za32_s32_vg1x2(uint32_t base) __arm_streaming __arm_shar
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv8i32.nxv4i32( [[TMP2]],  [[TMP3]], i64 4)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svuint32x2_t test_svread_za32_u32_vg1x2(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint32x2_t test_svread_za32_u32_vg1x2(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_za32_u32_vg1x2(base);
 }
 
@@ -1449,7 +1449,7 @@ svuint32x2_t test_svread_za32_u32_vg1x2(uint32_t base) __arm_streaming __arm_sha
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv8f32.nxv4f32( [[TMP2]],  [[TMP3]], i64 4)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svfloat32x2_t test_svread_za32_f32_vg1x2(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svfloat32x2_t test_svread_za32_f32_vg1x2(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_za32_f32_vg1x2(base);
 }
 
@@ -1471,7 +1471,7 @@ svfloat32x2_t test_svread_za32_f32_vg1x2(uint32_t base) __arm_streaming __arm_sh
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv4i64.nxv2i64( [[TMP2]],  [[TMP3]], i64 2)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svuint64x2_t test_svread_za64_u64_vg1x2(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint64x2_t test_svread_za64_u64_vg1x2(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_za64_u64_vg1x2(base);
 }
 
@@ -1493,7 +1493,7 @@ svuint64x2_t test_svread_za64_u64_vg1x2(uint32_t base) __arm_streaming __arm_sha
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv4f64.nxv2f64( [[TMP2]],  [[TMP3]], i64 2)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svfloat64x2_t test_svread_za64_f64_vg1x2(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svfloat64x2_t test_svread_za64_f64_vg1x2(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_za64_f64_vg1x2(base);
 }
 
@@ -1515,7 +1515,7 @@ svfloat64x2_t test_svread_za64_f64_vg1x2(uint32_t base) __arm_streaming __arm_sh
 // CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call  @llvm.vector.insert.nxv4i64.nxv2i64( [[TMP2]],  [[TMP3]], i64 2)
 // CPP-CHECK-NEXT:    ret  [[TMP4]]
 //
-svint64x2_t test_svread_za64_s64_vg1x2(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint64x2_t test_svread_za64_s64_vg1x2(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_za64_s64_vg1x2(base);
 }
 
@@ -1545,7 +1545,7 @@ svint64x2_t test_svread_za64_s64_vg1x2(uint32_t base) __arm_streaming __arm_shar
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv64i8.nxv16i8( [[TMP6]],  [[TMP7]], i64 48)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svint8x4_t test_svread_za8_s8_vg1x4(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint8x4_t test_svread_za8_s8_vg1x4(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_za8_s8_vg1x4(base);
 }
 
@@ -1575,7 +1575,7 @@ svint8x4_t test_svread_za8_s8_vg1x4(uint32_t base) __arm_streaming __arm_shared_
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv64i8.nxv16i8( [[TMP6]],  [[TMP7]], i64 48)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svuint8x4_t test_svread_za8_u8_vg1x4(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint8x4_t test_svread_za8_u8_vg1x4(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_za8_u8_vg1x4(base);
 }
 
@@ -1605,7 +1605,7 @@ svuint8x4_t test_svread_za8_u8_vg1x4(uint32_t base) __arm_streaming __arm_shared
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv32i16.nxv8i16( [[TMP6]],  [[TMP7]], i64 24)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svint16x4_t test_svread_za16_s16_vg1x4(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint16x4_t test_svread_za16_s16_vg1x4(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_za16_s16_vg1x4(base);
 }
 
@@ -1635,7 +1635,7 @@ svint16x4_t test_svread_za16_s16_vg1x4(uint32_t base) __arm_streaming __arm_shar
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv32i16.nxv8i16( [[TMP6]],  [[TMP7]], i64 24)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svuint16x4_t test_svread_za16_u16_vg1x4(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint16x4_t test_svread_za16_u16_vg1x4(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_za16_u16_vg1x4(base);
 }
 
@@ -1665,7 +1665,7 @@ svuint16x4_t test_svread_za16_u16_vg1x4(uint32_t base) __arm_streaming __arm_sha
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv32bf16.nxv8bf16( [[TMP6]],  [[TMP7]], i64 24)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svbfloat16x4_t test_svread_za16_bf16_vg1x4(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svbfloat16x4_t test_svread_za16_bf16_vg1x4(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_za16_bf16_vg1x4(base);
 }
 
@@ -1695,7 +1695,7 @@ svbfloat16x4_t test_svread_za16_bf16_vg1x4(uint32_t base) __arm_streaming __arm_
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv32f16.nxv8f16( [[TMP6]],  [[TMP7]], i64 24)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svfloat16x4_t test_svread_za16_f16_vg1x4(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svfloat16x4_t test_svread_za16_f16_vg1x4(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_za16_f16_vg1x4(base);
 }
 
@@ -1725,7 +1725,7 @@ svfloat16x4_t test_svread_za16_f16_vg1x4(uint32_t base) __arm_streaming __arm_sh
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv16i32.nxv4i32( [[TMP6]],  [[TMP7]], i64 12)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svint32x4_t test_svread_za32_s32_vg1x4(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint32x4_t test_svread_za32_s32_vg1x4(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_za32_s32_vg1x4(base);
 }
 
@@ -1755,7 +1755,7 @@ svint32x4_t test_svread_za32_s32_vg1x4(uint32_t base) __arm_streaming __arm_shar
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv16i32.nxv4i32( [[TMP6]],  [[TMP7]], i64 12)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svuint32x4_t test_svread_za32_u32_vg1x4(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint32x4_t test_svread_za32_u32_vg1x4(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_za32_u32_vg1x4(base);
 }
 
@@ -1785,7 +1785,7 @@ svuint32x4_t test_svread_za32_u32_vg1x4(uint32_t base) __arm_streaming __arm_sha
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv16f32.nxv4f32( [[TMP6]],  [[TMP7]], i64 12)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svfloat32x4_t test_svread_za32_f32_vg1x4(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svfloat32x4_t test_svread_za32_f32_vg1x4(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_za32_f32_vg1x4(base);
 }
 
@@ -1815,7 +1815,7 @@ svfloat32x4_t test_svread_za32_f32_vg1x4(uint32_t base) __arm_streaming __arm_sh
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv8i64.nxv2i64( [[TMP6]],  [[TMP7]], i64 6)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svuint64x4_t test_svread_za64_u64_vg1x4(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svuint64x4_t test_svread_za64_u64_vg1x4(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_za64_u64_vg1x4(base);
 }
 
@@ -1845,7 +1845,7 @@ svuint64x4_t test_svread_za64_u64_vg1x4(uint32_t base) __arm_streaming __arm_sha
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv8f64.nxv2f64( [[TMP6]],  [[TMP7]], i64 6)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svfloat64x4_t test_svread_za64_f64_vg1x4(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svfloat64x4_t test_svread_za64_f64_vg1x4(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_za64_f64_vg1x4(base);
 }
 
@@ -1875,6 +1875,6 @@ svfloat64x4_t test_svread_za64_f64_vg1x4(uint32_t base) __arm_streaming __arm_sh
 // CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call  @llvm.vector.insert.nxv8i64.nxv2i64( [[TMP6]],  [[TMP7]], i64 6)
 // CPP-CHECK-NEXT:    ret  [[TMP8]]
 //
-svint64x4_t test_svread_za64_s64_vg1x4(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+svint64x4_t test_svread_za64_s64_vg1x4(uint32_t base) __arm_streaming __arm_in("za") {
   return svread_za64_s64_vg1x4(base);
 }
diff --git a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_sub.c b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_sub.c
index 8c9a0a489ebf9..68913c5da4f75 100644
--- a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_sub.c
+++ b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_sub.c
@@ -36,7 +36,7 @@
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sub.write.single.za.vg1x2.nxv4i32(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svsub_write_single2_s32(uint32_t slice_base, svint32x2_t zn, svint32_t zm) __arm_streaming __arm_shared_za {
+void test_svsub_write_single2_s32(uint32_t slice_base, svint32x2_t zn, svint32_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svsub_write,_single,_za32,_s32,_vg1x2)(slice_base, zn, zm);
 }
 
@@ -54,7 +54,7 @@ void test_svsub_write_single2_s32(uint32_t slice_base, svint32x2_t zn, svint32_t
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sub.write.single.za.vg1x2.nxv4i32(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svsub_write_single2_u32(uint32_t slice_base, svuint32x2_t zn, svuint32_t zm) __arm_streaming __arm_shared_za {
+void test_svsub_write_single2_u32(uint32_t slice_base, svuint32x2_t zn, svuint32_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svsub_write,_single,_za32,_u32,_vg1x2)(slice_base, zn, zm);
 }
 
@@ -72,7 +72,7 @@ void test_svsub_write_single2_u32(uint32_t slice_base, svuint32x2_t zn, svuint32
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sub.write.single.za.vg1x2.nxv2i64(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svsub_write_single2_s64(uint32_t slice_base, svint64x2_t zn, svint64_t zm) __arm_streaming __arm_shared_za {
+void test_svsub_write_single2_s64(uint32_t slice_base, svint64x2_t zn, svint64_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svsub_write,_single,_za64,_s64,_vg1x2)(slice_base, zn, zm);
 }
 
@@ -90,7 +90,7 @@ void test_svsub_write_single2_s64(uint32_t slice_base, svint64x2_t zn, svint64_t
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sub.write.single.za.vg1x2.nxv2i64(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svsub_write_single2_u64(uint32_t slice_base, svuint64x2_t zn, svuint64_t zm) __arm_streaming __arm_shared_za {
+void test_svsub_write_single2_u64(uint32_t slice_base, svuint64x2_t zn, svuint64_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svsub_write,_single,_za64,_u64,_vg1x2)(slice_base, zn, zm);
 }
 
@@ -114,7 +114,7 @@ void test_svsub_write_single2_u64(uint32_t slice_base, svuint64x2_t zn, svuint64
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sub.write.single.za.vg1x4.nxv4i32(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svsub_write_single4_s32(uint32_t slice_base, svint32x4_t zn, svint32_t zm) __arm_streaming __arm_shared_za {
+void test_svsub_write_single4_s32(uint32_t slice_base, svint32x4_t zn, svint32_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svsub_write,_single,_za32,_s32,_vg1x4)(slice_base, zn, zm);
 }
 
@@ -136,7 +136,7 @@ void test_svsub_write_single4_s32(uint32_t slice_base, svint32x4_t zn, svint32_t
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sub.write.single.za.vg1x4.nxv4i32(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svsub_write_single4_u32(uint32_t slice_base, svuint32x4_t zn, svuint32_t zm) __arm_streaming __arm_shared_za {
+void test_svsub_write_single4_u32(uint32_t slice_base, svuint32x4_t zn, svuint32_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svsub_write,_single,_za32,_u32,_vg1x4)(slice_base, zn, zm);
 }
 
@@ -158,7 +158,7 @@ void test_svsub_write_single4_u32(uint32_t slice_base, svuint32x4_t zn, svuint32
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sub.write.single.za.vg1x4.nxv2i64(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svsub_write_single4_s64(uint32_t slice_base, svint64x4_t zn, svint64_t zm) __arm_streaming __arm_shared_za {
+void test_svsub_write_single4_s64(uint32_t slice_base, svint64x4_t zn, svint64_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svsub_write,_single,_za64,_s64,_vg1x4)(slice_base, zn, zm);
 }
 
@@ -180,7 +180,7 @@ void test_svsub_write_single4_s64(uint32_t slice_base, svint64x4_t zn, svint64_t
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sub.write.single.za.vg1x4.nxv2i64(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svsub_write_single4_u64(uint32_t slice_base, svuint64x4_t zn, svuint64_t zm) __arm_streaming __arm_shared_za {
+void test_svsub_write_single4_u64(uint32_t slice_base, svuint64x4_t zn, svuint64_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svsub_write,_single,_za64,_u64,_vg1x4)(slice_base, zn, zm);
 }
 
@@ -208,7 +208,7 @@ void test_svsub_write_single4_u64(uint32_t slice_base, svuint64x4_t zn, svuint64
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sub.write.za.vg1x2.nxv4i32(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svsub_write_multi2_s32(uint32_t slice_base, svint32x2_t zn, svint32x2_t zm) __arm_streaming __arm_shared_za {
+void test_svsub_write_multi2_s32(uint32_t slice_base, svint32x2_t zn, svint32x2_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svsub_write,,_za32,_s32,_vg1x2)(slice_base, zn, zm);
 }
 
@@ -230,7 +230,7 @@ void test_svsub_write_multi2_s32(uint32_t slice_base, svint32x2_t zn, svint32x2_
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sub.write.za.vg1x2.nxv4i32(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svsub_write_multi2_u32(uint32_t slice_base, svuint32x2_t zn, svuint32x2_t zm) __arm_streaming __arm_shared_za {
+void test_svsub_write_multi2_u32(uint32_t slice_base, svuint32x2_t zn, svuint32x2_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svsub_write,,_za32,_u32,_vg1x2)(slice_base, zn, zm);
 }
 
@@ -252,7 +252,7 @@ void test_svsub_write_multi2_u32(uint32_t slice_base, svuint32x2_t zn, svuint32x
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sub.write.za.vg1x2.nxv2i64(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svsub_write_multi2_s64(uint32_t slice_base, svint64x2_t zn, svint64x2_t zm) __arm_streaming __arm_shared_za {
+void test_svsub_write_multi2_s64(uint32_t slice_base, svint64x2_t zn, svint64x2_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svsub_write,,_za64,_s64,_vg1x2)(slice_base, zn, zm);
 }
 
@@ -274,7 +274,7 @@ void test_svsub_write_multi2_s64(uint32_t slice_base, svint64x2_t zn, svint64x2_
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sub.write.za.vg1x2.nxv2i64(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svsub_write_multi2_u64(uint32_t slice_base, svuint64x2_t zn, svuint64x2_t zm) __arm_streaming __arm_shared_za {
+void test_svsub_write_multi2_u64(uint32_t slice_base, svuint64x2_t zn, svuint64x2_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svsub_write,,_za64,_u64,_vg1x2)(slice_base, zn, zm);
 }
 
@@ -306,7 +306,7 @@ void test_svsub_write_multi2_u64(uint32_t slice_base, svuint64x2_t zn, svuint64x
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sub.write.za.vg1x4.nxv4i32(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[TMP4]],  [[TMP5]],  [[TMP6]],  [[TMP7]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svsub_write_multi4_s32(uint32_t slice_base, svint32x4_t zn, svint32x4_t zm) __arm_streaming __arm_shared_za {
+void test_svsub_write_multi4_s32(uint32_t slice_base, svint32x4_t zn, svint32x4_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svsub_write,,_za32,_s32,_vg1x4)(slice_base, zn, zm);
 }
 
@@ -336,7 +336,7 @@ void test_svsub_write_multi4_s32(uint32_t slice_base, svint32x4_t zn, svint32x4_
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sub.write.za.vg1x4.nxv4i32(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[TMP4]],  [[TMP5]],  [[TMP6]],  [[TMP7]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svsub_write_multi4_u32(uint32_t slice_base, svuint32x4_t zn, svuint32x4_t zm) __arm_streaming __arm_shared_za {
+void test_svsub_write_multi4_u32(uint32_t slice_base, svuint32x4_t zn, svuint32x4_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svsub_write,,_za32,_u32,_vg1x4)(slice_base, zn, zm);
 }
 
@@ -366,7 +366,7 @@ void test_svsub_write_multi4_u32(uint32_t slice_base, svuint32x4_t zn, svuint32x
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sub.write.za.vg1x4.nxv2i64(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[TMP4]],  [[TMP5]],  [[TMP6]],  [[TMP7]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svsub_write_multi4_s64(uint32_t slice_base, svint64x4_t zn, svint64x4_t zm) __arm_streaming __arm_shared_za {
+void test_svsub_write_multi4_s64(uint32_t slice_base, svint64x4_t zn, svint64x4_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svsub_write,,_za64,_s64,_vg1x4)(slice_base, zn, zm);
 }
 
@@ -396,7 +396,7 @@ void test_svsub_write_multi4_s64(uint32_t slice_base, svint64x4_t zn, svint64x4_
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sub.write.za.vg1x4.nxv2i64(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[TMP4]],  [[TMP5]],  [[TMP6]],  [[TMP7]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svsub_write_multi4_u64(uint32_t slice_base, svuint64x4_t zn, svuint64x4_t zm) __arm_streaming __arm_shared_za {
+void test_svsub_write_multi4_u64(uint32_t slice_base, svuint64x4_t zn, svuint64x4_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svsub_write,,_za64,_u64,_vg1x4)(slice_base, zn, zm);
 }
 
@@ -420,7 +420,7 @@ void test_svsub_write_multi4_u64(uint32_t slice_base, svuint64x4_t zn, svuint64x
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sub.za32.vg1x2.nxv4f32(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svsub_za32_vg1x2_f32(uint32_t slice_base, svfloat32x2_t zn) __arm_streaming __arm_shared_za {
+void test_svsub_za32_vg1x2_f32(uint32_t slice_base, svfloat32x2_t zn) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svsub_za32,,_f32,,_vg1x2)(slice_base, zn);
 }
 
@@ -438,7 +438,7 @@ void test_svsub_za32_vg1x2_f32(uint32_t slice_base, svfloat32x2_t zn) __arm_stre
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sub.za32.vg1x2.nxv4i32(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svsub_za32_vg1x2_s32(uint32_t slice_base, svint32x2_t zn) __arm_streaming __arm_shared_za {
+void test_svsub_za32_vg1x2_s32(uint32_t slice_base, svint32x2_t zn) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svsub_za32,,_s32,,_vg1x2)(slice_base , zn);
 }
 
@@ -456,7 +456,7 @@ void test_svsub_za32_vg1x2_s32(uint32_t slice_base, svint32x2_t zn) __arm_stream
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sub.za32.vg1x2.nxv4i32(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svsub_za32_vg1x2_u32(uint32_t slice_base, svuint32x2_t zn) __arm_streaming __arm_shared_za {
+void test_svsub_za32_vg1x2_u32(uint32_t slice_base, svuint32x2_t zn) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svsub_za32,,_u32,,_vg1x2)(slice_base, zn);
 }
 
@@ -474,7 +474,7 @@ void test_svsub_za32_vg1x2_u32(uint32_t slice_base, svuint32x2_t zn) __arm_strea
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sub.za64.vg1x2.nxv2f64(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svsub_za64_vg1x2_f64(uint32_t slice_base, svfloat64x2_t zn) __arm_streaming __arm_shared_za {
+void test_svsub_za64_vg1x2_f64(uint32_t slice_base, svfloat64x2_t zn) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svsub_za64,,_f64,,_vg1x2)(slice_base, zn);
 }
 
@@ -492,7 +492,7 @@ void test_svsub_za64_vg1x2_f64(uint32_t slice_base, svfloat64x2_t zn) __arm_stre
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sub.za64.vg1x2.nxv2i64(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svsub_za64_vg1x2_s64(uint32_t slice_base, svint64x2_t zn) __arm_streaming __arm_shared_za {
+void test_svsub_za64_vg1x2_s64(uint32_t slice_base, svint64x2_t zn) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svsub_za64,,_s64,,_vg1x2)(slice_base, zn);
 }
 
@@ -510,7 +510,7 @@ void test_svsub_za64_vg1x2_s64(uint32_t slice_base, svint64x2_t zn) __arm_stream
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sub.za64.vg1x2.nxv2i64(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svsub_za64_vg1x2_u64(uint32_t slice_base, svuint64x2_t zn) __arm_streaming __arm_shared_za {
+void test_svsub_za64_vg1x2_u64(uint32_t slice_base, svuint64x2_t zn) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svsub_za64,,_u64,,_vg1x2)(slice_base, zn);
 }
 
@@ -534,7 +534,7 @@ void test_svsub_za64_vg1x2_u64(uint32_t slice_base, svuint64x2_t zn) __arm_strea
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sub.za32.vg1x4.nxv4f32(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svsub_za32_vg1x4_f32(uint32_t slice_base, svfloat32x4_t zn) __arm_streaming __arm_shared_za {
+void test_svsub_za32_vg1x4_f32(uint32_t slice_base, svfloat32x4_t zn) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svsub_za32,,_f32,,_vg1x4)(slice_base, zn);
 }
 
@@ -556,7 +556,7 @@ void test_svsub_za32_vg1x4_f32(uint32_t slice_base, svfloat32x4_t zn) __arm_stre
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sub.za32.vg1x4.nxv4i32(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svsub_za32_vg1x4_s32(uint32_t slice_base, svint32x4_t zn) __arm_streaming __arm_shared_za {
+void test_svsub_za32_vg1x4_s32(uint32_t slice_base, svint32x4_t zn) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svsub_za32,,_s32,,_vg1x4)(slice_base, zn);
 }
 
@@ -578,7 +578,7 @@ void test_svsub_za32_vg1x4_s32(uint32_t slice_base, svint32x4_t zn) __arm_stream
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sub.za32.vg1x4.nxv4i32(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svsub_za32_vg1x4_u32(uint32_t slice_base, svuint32x4_t zn) __arm_streaming __arm_shared_za {
+void test_svsub_za32_vg1x4_u32(uint32_t slice_base, svuint32x4_t zn) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svsub_za32,,_u32,,_vg1x4)(slice_base, zn);
 }
 
@@ -600,7 +600,7 @@ void test_svsub_za32_vg1x4_u32(uint32_t slice_base, svuint32x4_t zn) __arm_strea
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sub.za64.vg1x4.nxv2f64(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svsub_za64_vg1x4_f64(uint32_t slice_base, svfloat64x4_t zn) __arm_streaming __arm_shared_za {
+void test_svsub_za64_vg1x4_f64(uint32_t slice_base, svfloat64x4_t zn) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svsub_za64,,_f64,,_vg1x4)(slice_base, zn);
 }
 
@@ -622,7 +622,7 @@ void test_svsub_za64_vg1x4_f64(uint32_t slice_base, svfloat64x4_t zn) __arm_stre
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sub.za64.vg1x4.nxv2i64(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svsub_za64_vg1x4_s64(uint32_t slice_base, svint64x4_t zn) __arm_streaming __arm_shared_za {
+void test_svsub_za64_vg1x4_s64(uint32_t slice_base, svint64x4_t zn) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svsub_za64,,_s64,,_vg1x4)(slice_base, zn);
 }
 
@@ -644,6 +644,6 @@ void test_svsub_za64_vg1x4_s64(uint32_t slice_base, svint64x4_t zn) __arm_stream
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.sub.za64.vg1x4.nxv2i64(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svsub_za64_vg1x4_u64(uint32_t slice_base, svuint64x4_t zn) __arm_streaming __arm_shared_za {
+void test_svsub_za64_vg1x4_u64(uint32_t slice_base, svuint64x4_t zn) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svsub_za64,,_u64,,_vg1x4)(slice_base, zn);
 }
diff --git a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_vdot.c b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_vdot.c
index fb313d4cebd72..9aa993f68b163 100644
--- a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_vdot.c
+++ b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_vdot.c
@@ -28,7 +28,7 @@
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fvdot.lane.za32.vg1x2.nxv8bf16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]], i32 3)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svvdot_lane_za32_bf16_vg1x2(uint32_t slice_base, svbfloat16x2_t zn, svbfloat16_t zm) __arm_streaming __arm_shared_za {
+void test_svvdot_lane_za32_bf16_vg1x2(uint32_t slice_base, svbfloat16x2_t zn, svbfloat16_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svvdot_lane_za32,_bf16,_vg1x2)(slice_base, zn, zm, 3);
 }
 
@@ -46,7 +46,7 @@ void test_svvdot_lane_za32_bf16_vg1x2(uint32_t slice_base, svbfloat16x2_t zn, sv
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.fvdot.lane.za32.vg1x2.nxv8f16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]], i32 3)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svvdot_lane_za32_f16_vg1x2(uint32_t slice_base, svfloat16x2_t zn, svfloat16_t zm) __arm_streaming __arm_shared_za {
+void test_svvdot_lane_za32_f16_vg1x2(uint32_t slice_base, svfloat16x2_t zn, svfloat16_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svvdot_lane_za32,_f16,_vg1x2)(slice_base, zn, zm, 3);
 }
 
@@ -64,7 +64,7 @@ void test_svvdot_lane_za32_f16_vg1x2(uint32_t slice_base, svfloat16x2_t zn, svfl
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.svdot.lane.za32.vg1x2.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]], i32 3)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svvdot_lane_za32_s16_vg1x2(uint32_t slice_base, svint16x2_t zn, svint16_t zm) __arm_streaming __arm_shared_za {
+void test_svvdot_lane_za32_s16_vg1x2(uint32_t slice_base, svint16x2_t zn, svint16_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svvdot_lane_za32,_s16,_vg1x2)(slice_base, zn, zm, 3);
 }
 
@@ -82,7 +82,7 @@ void test_svvdot_lane_za32_s16_vg1x2(uint32_t slice_base, svint16x2_t zn, svint1
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.uvdot.lane.za32.vg1x2.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[ZM:%.*]], i32 3)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svvdot_lane_za32_u16_vg1x2(uint32_t slice_base, svuint16x2_t zn, svuint16_t zm) __arm_streaming __arm_shared_za {
+void test_svvdot_lane_za32_u16_vg1x2(uint32_t slice_base, svuint16x2_t zn, svuint16_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svvdot_lane_za32,_u16,_vg1x2)(slice_base, zn, zm, 3);
 }
 
@@ -104,7 +104,7 @@ void test_svvdot_lane_za32_u16_vg1x2(uint32_t slice_base, svuint16x2_t zn, svuin
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.svdot.lane.za32.vg1x4.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]], i32 3)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svvdot_lane_za32_s8_vg1x4(uint32_t slice_base, svint8x4_t zn, svint8_t zm) __arm_streaming __arm_shared_za {
+void test_svvdot_lane_za32_s8_vg1x4(uint32_t slice_base, svint8x4_t zn, svint8_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svvdot_lane_za32,_s8,_vg1x4)(slice_base, zn, zm, 3);
 }
 
@@ -126,7 +126,7 @@ void test_svvdot_lane_za32_s8_vg1x4(uint32_t slice_base, svint8x4_t zn, svint8_t
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.uvdot.lane.za32.vg1x4.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]], i32 3)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svvdot_lane_za32_u8_vg1x4(uint32_t slice_base, svuint8x4_t zn, svuint8_t zm) __arm_streaming __arm_shared_za {
+void test_svvdot_lane_za32_u8_vg1x4(uint32_t slice_base, svuint8x4_t zn, svuint8_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svvdot_lane_za32,_u8,_vg1x4)(slice_base, zn, zm, 3);
 }
 
@@ -148,7 +148,7 @@ void test_svvdot_lane_za32_u8_vg1x4(uint32_t slice_base, svuint8x4_t zn, svuint8
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.svdot.lane.za64.vg1x4.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]], i32 1)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svvdot_lane_za64_s16_vg1x4(uint32_t slice_base, svint16x4_t zn, svint16_t zm) __arm_streaming __arm_shared_za {
+void test_svvdot_lane_za64_s16_vg1x4(uint32_t slice_base, svint16x4_t zn, svint16_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svvdot_lane_za64,_s16,_vg1x4)(slice_base, zn, zm, 1);
 }
 
@@ -170,7 +170,7 @@ void test_svvdot_lane_za64_s16_vg1x4(uint32_t slice_base, svint16x4_t zn, svint1
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.uvdot.lane.za64.vg1x4.nxv8i16(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]], i32 1)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svvdot_lane_za64_u16_vg1x4(uint32_t slice_base, svuint16x4_t zn, svuint16_t zm) __arm_streaming __arm_shared_za {
+void test_svvdot_lane_za64_u16_vg1x4(uint32_t slice_base, svuint16x4_t zn, svuint16_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svvdot_lane_za64,_u16,_vg1x4)(slice_base, zn, zm, 1);
 }
 
@@ -193,7 +193,7 @@ void test_svvdot_lane_za64_u16_vg1x4(uint32_t slice_base, svuint16x4_t zn, svuin
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.suvdot.lane.za32.vg1x4.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]], i32 3)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svsuvdot_lane_za32_s8_vg1x4(uint32_t slice_base, svint8x4_t zn, svint8_t zm) __arm_streaming __arm_shared_za {
+void test_svsuvdot_lane_za32_s8_vg1x4(uint32_t slice_base, svint8x4_t zn, svint8_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svsuvdot_lane_za32,_s8,_vg1x4)(slice_base, zn, zm, 3);
 }
 
@@ -216,7 +216,7 @@ void test_svsuvdot_lane_za32_s8_vg1x4(uint32_t slice_base, svint8x4_t zn, svint8
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.usvdot.lane.za32.vg1x4.nxv16i8(i32 [[SLICE_BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]],  [[ZM:%.*]], i32 3)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svusvdot_lane_za32_u8_vg1x4(uint32_t slice_base, svuint8x4_t zn, svuint8_t zm) __arm_streaming __arm_shared_za {
+void test_svusvdot_lane_za32_u8_vg1x4(uint32_t slice_base, svuint8x4_t zn, svuint8_t zm) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svusvdot_lane_za32,_u8,_vg1x4)(slice_base, zn, zm, 3);
 }
 
diff --git a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_write.c b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_write.c
index 14b0371bce574..bf8e12b6b4969 100644
--- a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_write.c
+++ b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_write.c
@@ -30,7 +30,7 @@
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.ver.vg2.nxv16i8(i32 0, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_ver_za8_u8_vg2(uint32_t base, svuint8x2_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za8_u8_vg2(uint32_t base, svuint8x2_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_ver_za8,_u8,_vg2,)(0, base, val);
 }
 
@@ -48,7 +48,7 @@ void test_svwrite_ver_za8_u8_vg2(uint32_t base, svuint8x2_t val) __arm_streaming
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.ver.vg2.nxv16i8(i32 0, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_ver_za8_s8_vg2(uint32_t base, svint8x2_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za8_s8_vg2(uint32_t base, svint8x2_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_ver_za8,_s8,_vg2,)(0, base, val);
 }
 
@@ -66,7 +66,7 @@ void test_svwrite_ver_za8_s8_vg2(uint32_t base, svint8x2_t val) __arm_streaming
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.hor.vg2.nxv16i8(i32 0, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_hor_za8_u8_vg2(uint32_t base, svuint8x2_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za8_u8_vg2(uint32_t base, svuint8x2_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_hor_za8,_u8,_vg2,)(0, base, val);
 }
 
@@ -84,7 +84,7 @@ void test_svwrite_hor_za8_u8_vg2(uint32_t base, svuint8x2_t val) __arm_streaming
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.hor.vg2.nxv16i8(i32 0, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_hor_za8_s8_vg2(uint32_t base, svint8x2_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za8_s8_vg2(uint32_t base, svint8x2_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_hor_za8,_s8,_vg2,)(0, base, val);
 }
 
@@ -106,7 +106,7 @@ void test_svwrite_hor_za8_s8_vg2(uint32_t base, svint8x2_t val) __arm_streaming
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.hor.vg4.nxv16i8(i32 0, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_hor_za8_u8_vg4(uint32_t base, svuint8x4_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za8_u8_vg4(uint32_t base, svuint8x4_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_hor_za8,_u8,_vg4,)(0, base, val);
 }
 
@@ -128,7 +128,7 @@ void test_svwrite_hor_za8_u8_vg4(uint32_t base, svuint8x4_t val) __arm_streaming
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.hor.vg4.nxv16i8(i32 0, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_hor_za8_s8_vg4(uint32_t base, svint8x4_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za8_s8_vg4(uint32_t base, svint8x4_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_hor_za8,_s8,_vg4,)(0, base, val);
 }
 
@@ -150,7 +150,7 @@ void test_svwrite_hor_za8_s8_vg4(uint32_t base, svint8x4_t val) __arm_streaming
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.ver.vg4.nxv16i8(i32 0, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_ver_za8_u8_vg4(uint32_t base, svuint8x4_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za8_u8_vg4(uint32_t base, svuint8x4_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_ver_za8,_u8,_vg4,)(0, base, val);
 }
 
@@ -172,7 +172,7 @@ void test_svwrite_ver_za8_u8_vg4(uint32_t base, svuint8x4_t val) __arm_streaming
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.ver.vg4.nxv16i8(i32 0, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_ver_za8_s8_vg4(uint32_t base, svint8x4_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za8_s8_vg4(uint32_t base, svint8x4_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_ver_za8,_s8,_vg4,)(0, base, val);
 }
 
@@ -190,7 +190,7 @@ void test_svwrite_ver_za8_s8_vg4(uint32_t base, svint8x4_t val) __arm_streaming
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.hor.vg2.nxv8i16(i32 1, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_hor_za16_u16_vg2(uint32_t base, svuint16x2_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za16_u16_vg2(uint32_t base, svuint16x2_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_hor_za16,_u16,_vg2,)(1, base, val);
 }
 
@@ -208,7 +208,7 @@ void test_svwrite_hor_za16_u16_vg2(uint32_t base, svuint16x2_t val) __arm_stream
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.hor.vg2.nxv8bf16(i32 1, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_hor_za16_bf16_vg2(uint32_t base, svbfloat16x2_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za16_bf16_vg2(uint32_t base, svbfloat16x2_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_hor_za16,_bf16,_vg2,)(1, base, val);
 }
 
@@ -226,7 +226,7 @@ void test_svwrite_hor_za16_bf16_vg2(uint32_t base, svbfloat16x2_t val) __arm_str
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.hor.vg2.nxv8f16(i32 1, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_hor_za16_f16_vg2(uint32_t base, svfloat16x2_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za16_f16_vg2(uint32_t base, svfloat16x2_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_hor_za16,_f16,_vg2,)(1, base, val);
 }
 
@@ -244,7 +244,7 @@ void test_svwrite_hor_za16_f16_vg2(uint32_t base, svfloat16x2_t val) __arm_strea
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.hor.vg2.nxv8i16(i32 1, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_hor_za16_s16_vg2(uint32_t base, svint16x2_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za16_s16_vg2(uint32_t base, svint16x2_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_hor_za16,_s16,_vg2,)(1, base, val);
 }
 
@@ -262,7 +262,7 @@ void test_svwrite_hor_za16_s16_vg2(uint32_t base, svint16x2_t val) __arm_streami
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.ver.vg2.nxv8i16(i32 1, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_ver_za16_u16_vg2(uint32_t base, svuint16x2_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za16_u16_vg2(uint32_t base, svuint16x2_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_ver_za16,_u16,_vg2,)(1, base, val);
 }
 
@@ -280,7 +280,7 @@ void test_svwrite_ver_za16_u16_vg2(uint32_t base, svuint16x2_t val) __arm_stream
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.ver.vg2.nxv8bf16(i32 1, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_ver_za16_bf16_vg2(uint32_t base, svbfloat16x2_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za16_bf16_vg2(uint32_t base, svbfloat16x2_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_ver_za16,_bf16,_vg2,)(1, base, val);
 }
 
@@ -298,7 +298,7 @@ void test_svwrite_ver_za16_bf16_vg2(uint32_t base, svbfloat16x2_t val) __arm_str
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.ver.vg2.nxv8f16(i32 1, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_ver_za16_f16_vg2(uint32_t base, svfloat16x2_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za16_f16_vg2(uint32_t base, svfloat16x2_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_ver_za16,_f16,_vg2,)(1, base, val);
 }
 
@@ -316,7 +316,7 @@ void test_svwrite_ver_za16_f16_vg2(uint32_t base, svfloat16x2_t val) __arm_strea
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.ver.vg2.nxv8i16(i32 1, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_ver_za16_s16_vg2(uint32_t base, svint16x2_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za16_s16_vg2(uint32_t base, svint16x2_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_ver_za16,_s16,_vg2,)(1, base, val);
 }
 
@@ -338,7 +338,7 @@ void test_svwrite_ver_za16_s16_vg2(uint32_t base, svint16x2_t val) __arm_streami
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.hor.vg4.nxv8i16(i32 1, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_hor_za16_u16_vg4(uint32_t base, svuint16x4_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za16_u16_vg4(uint32_t base, svuint16x4_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_hor_za16,_u16,_vg4,)(1, base, val);
 }
 
@@ -360,7 +360,7 @@ void test_svwrite_hor_za16_u16_vg4(uint32_t base, svuint16x4_t val) __arm_stream
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.hor.vg4.nxv8bf16(i32 1, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_hor_za16_bf16_vg4(uint32_t base, svbfloat16x4_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za16_bf16_vg4(uint32_t base, svbfloat16x4_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_hor_za16,_bf16,_vg4,)(1, base, val);
 }
 
@@ -382,7 +382,7 @@ void test_svwrite_hor_za16_bf16_vg4(uint32_t base, svbfloat16x4_t val) __arm_str
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.hor.vg4.nxv8f16(i32 1, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_hor_za16_f16_vg4(uint32_t base, svfloat16x4_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za16_f16_vg4(uint32_t base, svfloat16x4_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_hor_za16,_f16,_vg4,)(1, base, val);
 }
 
@@ -404,7 +404,7 @@ void test_svwrite_hor_za16_f16_vg4(uint32_t base, svfloat16x4_t val) __arm_strea
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.hor.vg4.nxv8i16(i32 1, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_hor_za16_s16_vg4(uint32_t base, svint16x4_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za16_s16_vg4(uint32_t base, svint16x4_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_hor_za16,_s16,_vg4,)(1, base, val);
 }
 
@@ -426,7 +426,7 @@ void test_svwrite_hor_za16_s16_vg4(uint32_t base, svint16x4_t val) __arm_streami
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.ver.vg4.nxv8i16(i32 1, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_ver_za16_u16_vg4(uint32_t base, svuint16x4_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za16_u16_vg4(uint32_t base, svuint16x4_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_ver_za16,_u16,_vg4,)(1, base, val);
 }
 
@@ -448,7 +448,7 @@ void test_svwrite_ver_za16_u16_vg4(uint32_t base, svuint16x4_t val) __arm_stream
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.ver.vg4.nxv8bf16(i32 1, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_ver_za16_bf16_vg4(uint32_t base, svbfloat16x4_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za16_bf16_vg4(uint32_t base, svbfloat16x4_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_ver_za16,_bf16,_vg4,)(1, base, val);
 }
 
@@ -470,7 +470,7 @@ void test_svwrite_ver_za16_bf16_vg4(uint32_t base, svbfloat16x4_t val) __arm_str
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.ver.vg4.nxv8f16(i32 1, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_ver_za16_f16_vg4(uint32_t base, svfloat16x4_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za16_f16_vg4(uint32_t base, svfloat16x4_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_ver_za16,_f16,_vg4,)(1, base, val);
 }
 
@@ -492,7 +492,7 @@ void test_svwrite_ver_za16_f16_vg4(uint32_t base, svfloat16x4_t val) __arm_strea
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.ver.vg4.nxv8i16(i32 1, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_ver_za16_s16_vg4(uint32_t base, svint16x4_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za16_s16_vg4(uint32_t base, svint16x4_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_ver_za16,_s16,_vg4,)(1, base, val);
 }
 
@@ -510,7 +510,7 @@ void test_svwrite_ver_za16_s16_vg4(uint32_t base, svint16x4_t val) __arm_streami
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.hor.vg2.nxv4i32(i32 3, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_hor_za32_u32_vg2(uint32_t base, svuint32x2_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za32_u32_vg2(uint32_t base, svuint32x2_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_hor_za32,_u32,_vg2,)(3, base, val);
 }
 
@@ -528,7 +528,7 @@ void test_svwrite_hor_za32_u32_vg2(uint32_t base, svuint32x2_t val) __arm_stream
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.hor.vg2.nxv4f32(i32 3, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_hor_za32_f32_vg2(uint32_t base, svfloat32x2_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za32_f32_vg2(uint32_t base, svfloat32x2_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_hor_za32,_f32,_vg2,)(3, base, val);
 }
 
@@ -546,7 +546,7 @@ void test_svwrite_hor_za32_f32_vg2(uint32_t base, svfloat32x2_t val) __arm_strea
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.hor.vg2.nxv4i32(i32 3, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_hor_za32_s32_vg2(uint32_t base, svint32x2_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za32_s32_vg2(uint32_t base, svint32x2_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_hor_za32,_s32,_vg2,)(3, base, val);
 }
 
@@ -564,7 +564,7 @@ void test_svwrite_hor_za32_s32_vg2(uint32_t base, svint32x2_t val) __arm_streami
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.ver.vg2.nxv4i32(i32 3, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_ver_za32_u32_vg2(uint32_t base, svuint32x2_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za32_u32_vg2(uint32_t base, svuint32x2_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_ver_za32,_u32,_vg2,)(3, base, val);
 }
 
@@ -582,7 +582,7 @@ void test_svwrite_ver_za32_u32_vg2(uint32_t base, svuint32x2_t val) __arm_stream
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.ver.vg2.nxv4f32(i32 3, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_ver_za32_f32_vg2(uint32_t base, svfloat32x2_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za32_f32_vg2(uint32_t base, svfloat32x2_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_ver_za32,_f32,_vg2,)(3, base, val);
 }
 
@@ -600,7 +600,7 @@ void test_svwrite_ver_za32_f32_vg2(uint32_t base, svfloat32x2_t val) __arm_strea
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.ver.vg2.nxv4i32(i32 3, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_ver_za32_s32_vg2(uint32_t base, svint32x2_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za32_s32_vg2(uint32_t base, svint32x2_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_ver_za32,_s32,_vg2,)(3, base, val);
 }
 
@@ -622,7 +622,7 @@ void test_svwrite_ver_za32_s32_vg2(uint32_t base, svint32x2_t val) __arm_streami
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.hor.vg4.nxv4i32(i32 3, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_hor_za32_u32_vg4(uint32_t base, svuint32x4_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za32_u32_vg4(uint32_t base, svuint32x4_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_hor_za32,_u32,_vg4,)(3, base, val);
 }
 
@@ -644,7 +644,7 @@ void test_svwrite_hor_za32_u32_vg4(uint32_t base, svuint32x4_t val) __arm_stream
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.hor.vg4.nxv4f32(i32 3, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_hor_za32_f32_vg4(uint32_t base, svfloat32x4_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za32_f32_vg4(uint32_t base, svfloat32x4_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_hor_za32,_f32,_vg4,)(3, base, val);
 }
 
@@ -666,7 +666,7 @@ void test_svwrite_hor_za32_f32_vg4(uint32_t base, svfloat32x4_t val) __arm_strea
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.hor.vg4.nxv4i32(i32 3, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_hor_za32_s32_vg4(uint32_t base, svint32x4_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za32_s32_vg4(uint32_t base, svint32x4_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_hor_za32,_s32,_vg4,)(3, base, val);
 }
 
@@ -688,7 +688,7 @@ void test_svwrite_hor_za32_s32_vg4(uint32_t base, svint32x4_t val) __arm_streami
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.ver.vg4.nxv4i32(i32 3, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_ver_za32_u32_vg4(uint32_t base, svuint32x4_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za32_u32_vg4(uint32_t base, svuint32x4_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_ver_za32,_u32,_vg4,)(3, base, val);
 }
 
@@ -710,7 +710,7 @@ void test_svwrite_ver_za32_u32_vg4(uint32_t base, svuint32x4_t val) __arm_stream
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.ver.vg4.nxv4f32(i32 3, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_ver_za32_f32_vg4(uint32_t base, svfloat32x4_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za32_f32_vg4(uint32_t base, svfloat32x4_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_ver_za32,_f32,_vg4,)(3, base, val);
 }
 
@@ -732,7 +732,7 @@ void test_svwrite_ver_za32_f32_vg4(uint32_t base, svfloat32x4_t val) __arm_strea
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.ver.vg4.nxv4i32(i32 3, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_ver_za32_s32_vg4(uint32_t base, svint32x4_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za32_s32_vg4(uint32_t base, svint32x4_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_ver_za32,_s32,_vg4,)(3, base, val);
 }
 
@@ -750,7 +750,7 @@ void test_svwrite_ver_za32_s32_vg4(uint32_t base, svint32x4_t val) __arm_streami
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.hor.vg2.nxv2i64(i32 7, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_hor_za64_u64_vg2(uint32_t base, svuint64x2_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za64_u64_vg2(uint32_t base, svuint64x2_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_hor_za64,_u64,_vg2,)(7, base, val);
 }
 
@@ -768,7 +768,7 @@ void test_svwrite_hor_za64_u64_vg2(uint32_t base, svuint64x2_t val) __arm_stream
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.hor.vg2.nxv2f64(i32 7, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_hor_za64_f64_vg2(uint32_t base, svfloat64x2_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za64_f64_vg2(uint32_t base, svfloat64x2_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_hor_za64,_f64,_vg2,)(7, base, val);
 }
 
@@ -786,7 +786,7 @@ void test_svwrite_hor_za64_f64_vg2(uint32_t base, svfloat64x2_t val) __arm_strea
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.hor.vg2.nxv2i64(i32 7, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_hor_za64_s64_vg2(uint32_t base, svint64x2_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za64_s64_vg2(uint32_t base, svint64x2_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_hor_za64,_s64,_vg2,)(7, base, val);
 }
 
@@ -804,7 +804,7 @@ void test_svwrite_hor_za64_s64_vg2(uint32_t base, svint64x2_t val) __arm_streami
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.ver.vg2.nxv2i64(i32 7, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_ver_za64_u64_vg2(uint32_t base, svuint64x2_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za64_u64_vg2(uint32_t base, svuint64x2_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_ver_za64,_u64,_vg2,)(7, base, val);
 }
 
@@ -822,7 +822,7 @@ void test_svwrite_ver_za64_u64_vg2(uint32_t base, svuint64x2_t val) __arm_stream
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.ver.vg2.nxv2f64(i32 7, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_ver_za64_f64_vg2(uint32_t base, svfloat64x2_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za64_f64_vg2(uint32_t base, svfloat64x2_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_ver_za64,_f64,_vg2,)(7, base, val);
 }
 
@@ -840,7 +840,7 @@ void test_svwrite_ver_za64_f64_vg2(uint32_t base, svfloat64x2_t val) __arm_strea
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.ver.vg2.nxv2i64(i32 7, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_ver_za64_s64_vg2(uint32_t base, svint64x2_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za64_s64_vg2(uint32_t base, svint64x2_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_ver_za64,_s64,_vg2,)(7, base, val);
 }
 
@@ -862,7 +862,7 @@ void test_svwrite_ver_za64_s64_vg2(uint32_t base, svint64x2_t val) __arm_streami
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.hor.vg4.nxv2i64(i32 7, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_hor_za64_u64_vg4(uint32_t base, svuint64x4_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za64_u64_vg4(uint32_t base, svuint64x4_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_hor_za64,_u64,_vg4,)(7, base, val);
 }
 
@@ -884,7 +884,7 @@ void test_svwrite_hor_za64_u64_vg4(uint32_t base, svuint64x4_t val) __arm_stream
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.ver.vg4.nxv2i64(i32 7, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_ver_za64_u64_vg4(uint32_t base, svuint64x4_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za64_u64_vg4(uint32_t base, svuint64x4_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_ver_za64,_u64,_vg4,)(7, base, val);
 }
 
@@ -906,7 +906,7 @@ void test_svwrite_ver_za64_u64_vg4(uint32_t base, svuint64x4_t val) __arm_stream
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.hor.vg4.nxv2f64(i32 7, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_hor_za64_f64_vg4(uint32_t base, svfloat64x4_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za64_f64_vg4(uint32_t base, svfloat64x4_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_hor_za64,_f64,_vg4,)(7, base, val);
 }
 
@@ -928,7 +928,7 @@ void test_svwrite_hor_za64_f64_vg4(uint32_t base, svfloat64x4_t val) __arm_strea
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.hor.vg4.nxv2i64(i32 7, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_hor_za64_s64_vg4(uint32_t base, svint64x4_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_hor_za64_s64_vg4(uint32_t base, svint64x4_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_hor_za64,_s64,_vg4,)(7, base, val);
 }
 
@@ -950,7 +950,7 @@ void test_svwrite_hor_za64_s64_vg4(uint32_t base, svint64x4_t val) __arm_streami
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.ver.vg4.nxv2f64(i32 7, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_ver_za64_f64_vg4(uint32_t base, svfloat64x4_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za64_f64_vg4(uint32_t base, svfloat64x4_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_ver_za64,_f64,_vg4,)(7, base, val);
 }
 
@@ -972,7 +972,7 @@ void test_svwrite_ver_za64_f64_vg4(uint32_t base, svfloat64x4_t val) __arm_strea
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.ver.vg4.nxv2i64(i32 7, i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_ver_za64_s64_vg4(uint32_t base, svint64x4_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_ver_za64_s64_vg4(uint32_t base, svint64x4_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_ver_za64,_s64,_vg4,)(7, base, val);
 }
 
@@ -990,7 +990,7 @@ void test_svwrite_ver_za64_s64_vg4(uint32_t base, svint64x4_t val) __arm_streami
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.vg1x2.nxv16i8(i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_za8_s8_vg1x2(uint32_t base, svint8x2_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_za8_s8_vg1x2(uint32_t base, svint8x2_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_za8,_s8,_vg1x2,)(base, val);
 }
 
@@ -1008,7 +1008,7 @@ void test_svwrite_za8_s8_vg1x2(uint32_t base, svint8x2_t val) __arm_streaming __
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.vg1x2.nxv16i8(i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_za8_u8_vg1x2(uint32_t base, svuint8x2_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_za8_u8_vg1x2(uint32_t base, svuint8x2_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_za8,_u8,_vg1x2,)(base, val);
 }
 
@@ -1026,7 +1026,7 @@ void test_svwrite_za8_u8_vg1x2(uint32_t base, svuint8x2_t val) __arm_streaming _
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.vg1x2.nxv8i16(i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_za16_s16_vg1x2(uint32_t base, svint16x2_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_za16_s16_vg1x2(uint32_t base, svint16x2_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_za16,_s16,_vg1x2,)(base, val);
 }
 
@@ -1044,7 +1044,7 @@ void test_svwrite_za16_s16_vg1x2(uint32_t base, svint16x2_t val) __arm_streaming
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.vg1x2.nxv8i16(i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_za16_u16_vg1x2(uint32_t base, svuint16x2_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_za16_u16_vg1x2(uint32_t base, svuint16x2_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_za16,_u16,_vg1x2,)(base, val);
 }
 
@@ -1062,7 +1062,7 @@ void test_svwrite_za16_u16_vg1x2(uint32_t base, svuint16x2_t val) __arm_streamin
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.vg1x2.nxv8bf16(i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_za16_bf16_vg1x2(uint32_t base, svbfloat16x2_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_za16_bf16_vg1x2(uint32_t base, svbfloat16x2_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_za16,_bf16,_vg1x2,)(base, val);
 }
 
@@ -1080,7 +1080,7 @@ void test_svwrite_za16_bf16_vg1x2(uint32_t base, svbfloat16x2_t val) __arm_strea
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.vg1x2.nxv8f16(i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_za16_f16_vg1x2(uint32_t base, svfloat16x2_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_za16_f16_vg1x2(uint32_t base, svfloat16x2_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_za16,_f16,_vg1x2,)(base, val);
 }
 
@@ -1098,7 +1098,7 @@ void test_svwrite_za16_f16_vg1x2(uint32_t base, svfloat16x2_t val) __arm_streami
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.vg1x2.nxv4i32(i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_za32_s32_vg1x2(uint32_t base, svint32x2_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_za32_s32_vg1x2(uint32_t base, svint32x2_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_za32,_s32,_vg1x2,)(base, val);
 }
 
@@ -1116,7 +1116,7 @@ void test_svwrite_za32_s32_vg1x2(uint32_t base, svint32x2_t val) __arm_streaming
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.vg1x2.nxv4i32(i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_za32_u32_vg1x2(uint32_t base, svuint32x2_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_za32_u32_vg1x2(uint32_t base, svuint32x2_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_za32,_u32,_vg1x2,)(base, val);
 }
 
@@ -1134,7 +1134,7 @@ void test_svwrite_za32_u32_vg1x2(uint32_t base, svuint32x2_t val) __arm_streamin
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.vg1x2.nxv4f32(i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_za32_f32_vg1x2(uint32_t base, svfloat32x2_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_za32_f32_vg1x2(uint32_t base, svfloat32x2_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_za32,_f32,_vg1x2,)(base, val);
 }
 
@@ -1152,7 +1152,7 @@ void test_svwrite_za32_f32_vg1x2(uint32_t base, svfloat32x2_t val) __arm_streami
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.vg1x2.nxv2i64(i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_za64_u64_vg1x2(uint32_t base, svuint64x2_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_za64_u64_vg1x2(uint32_t base, svuint64x2_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_za64,_u64,_vg1x2,)(base, val);
 }
 
@@ -1170,7 +1170,7 @@ void test_svwrite_za64_u64_vg1x2(uint32_t base, svuint64x2_t val) __arm_streamin
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.vg1x2.nxv2f64(i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_za64_f64_vg1x2(uint32_t base, svfloat64x2_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_za64_f64_vg1x2(uint32_t base, svfloat64x2_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_za64,_f64,_vg1x2,)(base, val);
 }
 
@@ -1188,7 +1188,7 @@ void test_svwrite_za64_f64_vg1x2(uint32_t base, svfloat64x2_t val) __arm_streami
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.vg1x2.nxv2i64(i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_za64_s64_vg1x2(uint32_t base, svint64x2_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_za64_s64_vg1x2(uint32_t base, svint64x2_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_za64,_s64,_vg1x2,)(base, val);
 }
 
@@ -1210,7 +1210,7 @@ void test_svwrite_za64_s64_vg1x2(uint32_t base, svint64x2_t val) __arm_streaming
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.vg1x4.nxv16i8(i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_za8_s8_vg1x4(uint32_t base, svint8x4_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_za8_s8_vg1x4(uint32_t base, svint8x4_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_za8,_s8,_vg1x4,)(base, val);
 }
 
@@ -1232,7 +1232,7 @@ void test_svwrite_za8_s8_vg1x4(uint32_t base, svint8x4_t val) __arm_streaming __
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.vg1x4.nxv16i8(i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_za8_u8_vg1x4(uint32_t base, svuint8x4_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_za8_u8_vg1x4(uint32_t base, svuint8x4_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_za8,_u8,_vg1x4,)(base, val);
 }
 
@@ -1254,7 +1254,7 @@ void test_svwrite_za8_u8_vg1x4(uint32_t base, svuint8x4_t val) __arm_streaming _
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.vg1x4.nxv8i16(i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_za16_s16_vg1x4(uint32_t base, svint16x4_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_za16_s16_vg1x4(uint32_t base, svint16x4_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_za16,_s16,_vg1x4,)(base, val);
 }
 
@@ -1276,7 +1276,7 @@ void test_svwrite_za16_s16_vg1x4(uint32_t base, svint16x4_t val) __arm_streaming
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.vg1x4.nxv8i16(i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_za16_u16_vg1x4(uint32_t base, svuint16x4_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_za16_u16_vg1x4(uint32_t base, svuint16x4_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_za16,_u16,_vg1x4,)(base, val);
 }
 
@@ -1298,7 +1298,7 @@ void test_svwrite_za16_u16_vg1x4(uint32_t base, svuint16x4_t val) __arm_streamin
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.vg1x4.nxv8bf16(i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_za16_bf16_vg1x4(uint32_t base, svbfloat16x4_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_za16_bf16_vg1x4(uint32_t base, svbfloat16x4_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_za16,_bf16,_vg1x4,)(base, val);
 }
 
@@ -1320,7 +1320,7 @@ void test_svwrite_za16_bf16_vg1x4(uint32_t base, svbfloat16x4_t val) __arm_strea
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.vg1x4.nxv8f16(i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_za16_f16_vg1x4(uint32_t base, svfloat16x4_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_za16_f16_vg1x4(uint32_t base, svfloat16x4_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_za16,_f16,_vg1x4,)(base, val);
 }
 
@@ -1342,7 +1342,7 @@ void test_svwrite_za16_f16_vg1x4(uint32_t base, svfloat16x4_t val) __arm_streami
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.vg1x4.nxv4i32(i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_za32_s32_vg1x4(uint32_t base, svint32x4_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_za32_s32_vg1x4(uint32_t base, svint32x4_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_za32,_s32,_vg1x4,)(base, val);
 }
 
@@ -1364,7 +1364,7 @@ void test_svwrite_za32_s32_vg1x4(uint32_t base, svint32x4_t val) __arm_streaming
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.vg1x4.nxv4i32(i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_za32_u32_vg1x4(uint32_t base, svuint32x4_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_za32_u32_vg1x4(uint32_t base, svuint32x4_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_za32,_u32,_vg1x4,)(base, val);
 }
 
@@ -1386,7 +1386,7 @@ void test_svwrite_za32_u32_vg1x4(uint32_t base, svuint32x4_t val) __arm_streamin
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.vg1x4.nxv4f32(i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_za32_f32_vg1x4(uint32_t base, svfloat32x4_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_za32_f32_vg1x4(uint32_t base, svfloat32x4_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_za32,_f32,_vg1x4,)(base, val);
 }
 
@@ -1408,7 +1408,7 @@ void test_svwrite_za32_f32_vg1x4(uint32_t base, svfloat32x4_t val) __arm_streami
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.vg1x4.nxv2i64(i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_za64_u64_vg1x4(uint32_t base, svuint64x4_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_za64_u64_vg1x4(uint32_t base, svuint64x4_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_za64,_u64,_vg1x4,)(base, val);
 }
 
@@ -1430,7 +1430,7 @@ void test_svwrite_za64_u64_vg1x4(uint32_t base, svuint64x4_t val) __arm_streamin
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.vg1x4.nxv2f64(i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_za64_f64_vg1x4(uint32_t base, svfloat64x4_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_za64_f64_vg1x4(uint32_t base, svfloat64x4_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_za64,_f64,_vg1x4,)(base, val);
 }
 
@@ -1452,6 +1452,6 @@ void test_svwrite_za64_f64_vg1x4(uint32_t base, svfloat64x4_t val) __arm_streami
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.write.vg1x4.nxv2i64(i32 [[BASE:%.*]],  [[TMP0]],  [[TMP1]],  [[TMP2]],  [[TMP3]])
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svwrite_za64_s64_vg1x4(uint32_t base, svint64x4_t val) __arm_streaming __arm_shared_za {
+void test_svwrite_za64_s64_vg1x4(uint32_t base, svint64x4_t val) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svwrite_za64,_s64,_vg1x4,)(base, val);
 }
diff --git a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_zero_zt.c b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_zero_zt.c
index 31e8d6850fb28..081b1a1d2627c 100644
--- a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_zero_zt.c
+++ b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_zero_zt.c
@@ -18,6 +18,6 @@
 // CPP-CHECK-NEXT:    tail call void @llvm.aarch64.sme.zero.zt(i32 0)
 // CPP-CHECK-NEXT:    ret void
 //
-void test_svzero_zt(void) __arm_streaming_compatible __arm_shared_za {
+void test_svzero_zt(void) __arm_streaming_compatible __arm_out("za") {
   svzero_zt(0);
 }
diff --git a/clang/test/CodeGen/aarch64-sve-acle-__ARM_FEATURE_SVE_VECTOR_OPERATORS.c b/clang/test/CodeGen/aarch64-sve-acle-__ARM_FEATURE_SVE_VECTOR_OPERATORS.c
index 9c96dfb9e33ee..1e6a4500cc886 100644
--- a/clang/test/CodeGen/aarch64-sve-acle-__ARM_FEATURE_SVE_VECTOR_OPERATORS.c
+++ b/clang/test/CodeGen/aarch64-sve-acle-__ARM_FEATURE_SVE_VECTOR_OPERATORS.c
@@ -59,7 +59,7 @@ typedef int8_t vec_int8 __attribute__((vector_size(N / 8)));
 // CHECK128-NEXT:    ret <16 x i8> [[CASTFIXEDSVE]]
 
 // CHECK-LABEL: define{{.*}} void @f2(
-// CHECK-SAME:   ptr noalias nocapture writeonly sret(<[[#div(VBITS,8)]] x i8>) align 16 %agg.result, ptr nocapture noundef readonly %0)
+// CHECK-SAME:   ptr dead_on_unwind noalias nocapture writable writeonly sret(<[[#div(VBITS,8)]] x i8>) align 16 %agg.result, ptr nocapture noundef readonly %0)
 // CHECK-NEXT: entry:
 // CHECK-NEXT:   [[X:%.*]] = load <[[#div(VBITS,8)]] x i8>, ptr [[TMP0:%.*]], align 16, [[TBAA6:!tbaa !.*]]
 // CHECK-NEXT:   [[TMP1:%.*]] = tail call  @llvm.aarch64.sve.ptrue.nxv16i1(i32 31)
diff --git a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_create2-bfloat.c b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_create2-bfloat.c
index 4058af0051359..1ed09cc5965fd 100644
--- a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_create2-bfloat.c
+++ b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_create2-bfloat.c
@@ -3,6 +3,7 @@
 // RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +bf16 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
 // RUN: %clang_cc1 -fclang-abi-compat=latest -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +bf16 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s
 // RUN: %clang_cc1 -fclang-abi-compat=latest -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +bf16 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -DTEST_SME -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +sme -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
 
 // REQUIRES: aarch64-registered-target
 
@@ -15,6 +16,12 @@
 #define SVE_ACLE_FUNC(A1,A2,A3,A4) A1##A2##A3##A4
 #endif
 
+#ifndef TEST_SME
+#define ATTR
+#else
+#define ATTR __arm_streaming
+#endif
+
 // CHECK-LABEL: @test_svcreate2_bf16(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv16bf16.nxv8bf16( poison,  [[X0:%.*]], i64 0)
@@ -27,7 +34,7 @@
 // CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.vector.insert.nxv16bf16.nxv8bf16( [[TMP0]],  [[X1:%.*]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP1]]
 //
-svbfloat16x2_t test_svcreate2_bf16(svbfloat16_t x0, svbfloat16_t x1)
+svbfloat16x2_t test_svcreate2_bf16(svbfloat16_t x0, svbfloat16_t x1) ATTR
 {
   return SVE_ACLE_FUNC(svcreate2,_bf16,,)(x0, x1);
 }
diff --git a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_create2.c b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_create2.c
index 5051ab88d9226..1aead4e5572f3 100644
--- a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_create2.c
+++ b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_create2.c
@@ -3,6 +3,7 @@
 // RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu -target-feature +sve -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
 // RUN: %clang_cc1 -fclang-abi-compat=latest -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s
 // RUN: %clang_cc1 -fclang-abi-compat=latest -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -DTEST_SME -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +sme -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
 
 // REQUIRES: aarch64-registered-target
 
@@ -15,6 +16,12 @@
 #define SVE_ACLE_FUNC(A1,A2,A3,A4) A1##A2##A3##A4
 #endif
 
+#ifndef TEST_SME
+#define ATTR
+#else
+#define ATTR __arm_streaming
+#endif
+
 // CHECK-LABEL: @test_svcreate2_s8(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv32i8.nxv16i8( poison,  [[X0:%.*]], i64 0)
@@ -27,7 +34,7 @@
 // CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.vector.insert.nxv32i8.nxv16i8( [[TMP0]],  [[X1:%.*]], i64 16)
 // CPP-CHECK-NEXT:    ret  [[TMP1]]
 //
-svint8x2_t test_svcreate2_s8(svint8_t x0, svint8_t x1)
+svint8x2_t test_svcreate2_s8(svint8_t x0, svint8_t x1) ATTR
 {
   return SVE_ACLE_FUNC(svcreate2,_s8,,)(x0, x1);
 }
@@ -44,7 +51,7 @@ svint8x2_t test_svcreate2_s8(svint8_t x0, svint8_t x1)
 // CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.vector.insert.nxv16i16.nxv8i16( [[TMP0]],  [[X1:%.*]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP1]]
 //
-svint16x2_t test_svcreate2_s16(svint16_t x0, svint16_t x1)
+svint16x2_t test_svcreate2_s16(svint16_t x0, svint16_t x1) ATTR
 {
   return SVE_ACLE_FUNC(svcreate2,_s16,,)(x0, x1);
 }
@@ -61,7 +68,7 @@ svint16x2_t test_svcreate2_s16(svint16_t x0, svint16_t x1)
 // CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.vector.insert.nxv8i32.nxv4i32( [[TMP0]],  [[X1:%.*]], i64 4)
 // CPP-CHECK-NEXT:    ret  [[TMP1]]
 //
-svint32x2_t test_svcreate2_s32(svint32_t x0, svint32_t x1)
+svint32x2_t test_svcreate2_s32(svint32_t x0, svint32_t x1) ATTR
 {
   return SVE_ACLE_FUNC(svcreate2,_s32,,)(x0, x1);
 }
@@ -78,7 +85,7 @@ svint32x2_t test_svcreate2_s32(svint32_t x0, svint32_t x1)
 // CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.vector.insert.nxv4i64.nxv2i64( [[TMP0]],  [[X1:%.*]], i64 2)
 // CPP-CHECK-NEXT:    ret  [[TMP1]]
 //
-svint64x2_t test_svcreate2_s64(svint64_t x0, svint64_t x1)
+svint64x2_t test_svcreate2_s64(svint64_t x0, svint64_t x1) ATTR
 {
   return SVE_ACLE_FUNC(svcreate2,_s64,,)(x0, x1);
 }
@@ -95,7 +102,7 @@ svint64x2_t test_svcreate2_s64(svint64_t x0, svint64_t x1)
 // CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.vector.insert.nxv32i8.nxv16i8( [[TMP0]],  [[X1:%.*]], i64 16)
 // CPP-CHECK-NEXT:    ret  [[TMP1]]
 //
-svuint8x2_t test_svcreate2_u8(svuint8_t x0, svuint8_t x1)
+svuint8x2_t test_svcreate2_u8(svuint8_t x0, svuint8_t x1) ATTR
 {
   return SVE_ACLE_FUNC(svcreate2,_u8,,)(x0, x1);
 }
@@ -112,7 +119,7 @@ svuint8x2_t test_svcreate2_u8(svuint8_t x0, svuint8_t x1)
 // CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.vector.insert.nxv16i16.nxv8i16( [[TMP0]],  [[X1:%.*]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP1]]
 //
-svuint16x2_t test_svcreate2_u16(svuint16_t x0, svuint16_t x1)
+svuint16x2_t test_svcreate2_u16(svuint16_t x0, svuint16_t x1) ATTR
 {
   return SVE_ACLE_FUNC(svcreate2,_u16,,)(x0, x1);
 }
@@ -129,7 +136,7 @@ svuint16x2_t test_svcreate2_u16(svuint16_t x0, svuint16_t x1)
 // CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.vector.insert.nxv8i32.nxv4i32( [[TMP0]],  [[X1:%.*]], i64 4)
 // CPP-CHECK-NEXT:    ret  [[TMP1]]
 //
-svuint32x2_t test_svcreate2_u32(svuint32_t x0, svuint32_t x1)
+svuint32x2_t test_svcreate2_u32(svuint32_t x0, svuint32_t x1) ATTR
 {
   return SVE_ACLE_FUNC(svcreate2,_u32,,)(x0, x1);
 }
@@ -146,7 +153,7 @@ svuint32x2_t test_svcreate2_u32(svuint32_t x0, svuint32_t x1)
 // CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.vector.insert.nxv4i64.nxv2i64( [[TMP0]],  [[X1:%.*]], i64 2)
 // CPP-CHECK-NEXT:    ret  [[TMP1]]
 //
-svuint64x2_t test_svcreate2_u64(svuint64_t x0, svuint64_t x1)
+svuint64x2_t test_svcreate2_u64(svuint64_t x0, svuint64_t x1) ATTR
 {
   return SVE_ACLE_FUNC(svcreate2,_u64,,)(x0, x1);
 }
@@ -163,7 +170,7 @@ svuint64x2_t test_svcreate2_u64(svuint64_t x0, svuint64_t x1)
 // CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.vector.insert.nxv16f16.nxv8f16( [[TMP0]],  [[X1:%.*]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP1]]
 //
-svfloat16x2_t test_svcreate2_f16(svfloat16_t x0, svfloat16_t x1)
+svfloat16x2_t test_svcreate2_f16(svfloat16_t x0, svfloat16_t x1) ATTR
 {
   return SVE_ACLE_FUNC(svcreate2,_f16,,)(x0, x1);
 }
@@ -180,7 +187,7 @@ svfloat16x2_t test_svcreate2_f16(svfloat16_t x0, svfloat16_t x1)
 // CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.vector.insert.nxv8f32.nxv4f32( [[TMP0]],  [[X1:%.*]], i64 4)
 // CPP-CHECK-NEXT:    ret  [[TMP1]]
 //
-svfloat32x2_t test_svcreate2_f32(svfloat32_t x0, svfloat32_t x1)
+svfloat32x2_t test_svcreate2_f32(svfloat32_t x0, svfloat32_t x1) ATTR
 {
   return SVE_ACLE_FUNC(svcreate2,_f32,,)(x0, x1);
 }
@@ -197,7 +204,7 @@ svfloat32x2_t test_svcreate2_f32(svfloat32_t x0, svfloat32_t x1)
 // CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.vector.insert.nxv4f64.nxv2f64( [[TMP0]],  [[X1:%.*]], i64 2)
 // CPP-CHECK-NEXT:    ret  [[TMP1]]
 //
-svfloat64x2_t test_svcreate2_f64(svfloat64_t x0, svfloat64_t x1)
+svfloat64x2_t test_svcreate2_f64(svfloat64_t x0, svfloat64_t x1) ATTR
 {
   return SVE_ACLE_FUNC(svcreate2,_f64,,)(x0, x1);
 }
diff --git a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_create3-bfloat.c b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_create3-bfloat.c
index 85eff36241163..90176ff0dbd48 100644
--- a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_create3-bfloat.c
+++ b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_create3-bfloat.c
@@ -3,6 +3,7 @@
 // RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +bf16 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
 // RUN: %clang_cc1 -fclang-abi-compat=latest -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +bf16 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s
 // RUN: %clang_cc1 -fclang-abi-compat=latest -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +bf16 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -DTEST_SME -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +sme -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
 
 // REQUIRES: aarch64-registered-target
 
@@ -15,6 +16,12 @@
 #define SVE_ACLE_FUNC(A1,A2,A3,A4) A1##A2##A3##A4
 #endif
 
+#ifndef TEST_SME
+#define ATTR
+#else
+#define ATTR __arm_streaming
+#endif
+
 // CHECK-LABEL: @test_svcreate3_bf16(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv24bf16.nxv8bf16( poison,  [[X0:%.*]], i64 0)
@@ -29,7 +36,7 @@
 // CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call  @llvm.vector.insert.nxv24bf16.nxv8bf16( [[TMP1]],  [[X2:%.*]], i64 16)
 // CPP-CHECK-NEXT:    ret  [[TMP2]]
 //
-svbfloat16x3_t test_svcreate3_bf16(svbfloat16_t x0, svbfloat16_t x1, svbfloat16_t x2)
+svbfloat16x3_t test_svcreate3_bf16(svbfloat16_t x0, svbfloat16_t x1, svbfloat16_t x2) ATTR
 {
   return SVE_ACLE_FUNC(svcreate3,_bf16,,)(x0, x1, x2);
 }
diff --git a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_create3.c b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_create3.c
index 3d6f34d5234cf..2fe1a88441b29 100644
--- a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_create3.c
+++ b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_create3.c
@@ -3,6 +3,7 @@
 // RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu -target-feature +sve -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
 // RUN: %clang_cc1 -fclang-abi-compat=latest -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s
 // RUN: %clang_cc1 -fclang-abi-compat=latest -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -DTEST_SME -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +sme -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
 
 // REQUIRES: aarch64-registered-target
 
@@ -15,6 +16,12 @@
 #define SVE_ACLE_FUNC(A1,A2,A3,A4) A1##A2##A3##A4
 #endif
 
+#ifndef TEST_SME
+#define ATTR
+#else
+#define ATTR __arm_streaming
+#endif
+
 // CHECK-LABEL: @test_svcreate3_s8(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv48i8.nxv16i8( poison,  [[X0:%.*]], i64 0)
@@ -29,7 +36,7 @@
 // CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call  @llvm.vector.insert.nxv48i8.nxv16i8( [[TMP1]],  [[X2:%.*]], i64 32)
 // CPP-CHECK-NEXT:    ret  [[TMP2]]
 //
-svint8x3_t test_svcreate3_s8(svint8_t x0, svint8_t x1, svint8_t x2)
+svint8x3_t test_svcreate3_s8(svint8_t x0, svint8_t x1, svint8_t x2) ATTR
 {
   return SVE_ACLE_FUNC(svcreate3,_s8,,)(x0, x1, x2);
 }
@@ -48,7 +55,7 @@ svint8x3_t test_svcreate3_s8(svint8_t x0, svint8_t x1, svint8_t x2)
 // CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call  @llvm.vector.insert.nxv24i16.nxv8i16( [[TMP1]],  [[X2:%.*]], i64 16)
 // CPP-CHECK-NEXT:    ret  [[TMP2]]
 //
-svint16x3_t test_svcreate3_s16(svint16_t x0, svint16_t x1, svint16_t x2)
+svint16x3_t test_svcreate3_s16(svint16_t x0, svint16_t x1, svint16_t x2) ATTR
 {
   return SVE_ACLE_FUNC(svcreate3,_s16,,)(x0, x1, x2);
 }
@@ -67,7 +74,7 @@ svint16x3_t test_svcreate3_s16(svint16_t x0, svint16_t x1, svint16_t x2)
 // CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call  @llvm.vector.insert.nxv12i32.nxv4i32( [[TMP1]],  [[X2:%.*]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP2]]
 //
-svint32x3_t test_svcreate3_s32(svint32_t x0, svint32_t x1, svint32_t x2)
+svint32x3_t test_svcreate3_s32(svint32_t x0, svint32_t x1, svint32_t x2) ATTR
 {
   return SVE_ACLE_FUNC(svcreate3,_s32,,)(x0, x1, x2);
 }
@@ -86,7 +93,7 @@ svint32x3_t test_svcreate3_s32(svint32_t x0, svint32_t x1, svint32_t x2)
 // CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call  @llvm.vector.insert.nxv6i64.nxv2i64( [[TMP1]],  [[X2:%.*]], i64 4)
 // CPP-CHECK-NEXT:    ret  [[TMP2]]
 //
-svint64x3_t test_svcreate3_s64(svint64_t x0, svint64_t x1, svint64_t x2)
+svint64x3_t test_svcreate3_s64(svint64_t x0, svint64_t x1, svint64_t x2) ATTR
 {
   return SVE_ACLE_FUNC(svcreate3,_s64,,)(x0, x1, x2);
 }
@@ -105,7 +112,7 @@ svint64x3_t test_svcreate3_s64(svint64_t x0, svint64_t x1, svint64_t x2)
 // CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call  @llvm.vector.insert.nxv48i8.nxv16i8( [[TMP1]],  [[X2:%.*]], i64 32)
 // CPP-CHECK-NEXT:    ret  [[TMP2]]
 //
-svuint8x3_t test_svcreate3_u8(svuint8_t x0, svuint8_t x1, svuint8_t x2)
+svuint8x3_t test_svcreate3_u8(svuint8_t x0, svuint8_t x1, svuint8_t x2) ATTR
 {
   return SVE_ACLE_FUNC(svcreate3,_u8,,)(x0, x1, x2);
 }
@@ -124,7 +131,7 @@ svuint8x3_t test_svcreate3_u8(svuint8_t x0, svuint8_t x1, svuint8_t x2)
 // CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call  @llvm.vector.insert.nxv24i16.nxv8i16( [[TMP1]],  [[X2:%.*]], i64 16)
 // CPP-CHECK-NEXT:    ret  [[TMP2]]
 //
-svuint16x3_t test_svcreate3_u16(svuint16_t x0, svuint16_t x1, svuint16_t x2)
+svuint16x3_t test_svcreate3_u16(svuint16_t x0, svuint16_t x1, svuint16_t x2) ATTR
 {
   return SVE_ACLE_FUNC(svcreate3,_u16,,)(x0, x1, x2);
 }
@@ -143,7 +150,7 @@ svuint16x3_t test_svcreate3_u16(svuint16_t x0, svuint16_t x1, svuint16_t x2)
 // CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call  @llvm.vector.insert.nxv12i32.nxv4i32( [[TMP1]],  [[X2:%.*]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP2]]
 //
-svuint32x3_t test_svcreate3_u32(svuint32_t x0, svuint32_t x1, svuint32_t x2)
+svuint32x3_t test_svcreate3_u32(svuint32_t x0, svuint32_t x1, svuint32_t x2) ATTR
 {
   return SVE_ACLE_FUNC(svcreate3,_u32,,)(x0, x1, x2);
 }
@@ -162,7 +169,7 @@ svuint32x3_t test_svcreate3_u32(svuint32_t x0, svuint32_t x1, svuint32_t x2)
 // CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call  @llvm.vector.insert.nxv6i64.nxv2i64( [[TMP1]],  [[X2:%.*]], i64 4)
 // CPP-CHECK-NEXT:    ret  [[TMP2]]
 //
-svuint64x3_t test_svcreate3_u64(svuint64_t x0, svuint64_t x1, svuint64_t x2)
+svuint64x3_t test_svcreate3_u64(svuint64_t x0, svuint64_t x1, svuint64_t x2) ATTR
 {
   return SVE_ACLE_FUNC(svcreate3,_u64,,)(x0, x1, x2);
 }
@@ -181,7 +188,7 @@ svuint64x3_t test_svcreate3_u64(svuint64_t x0, svuint64_t x1, svuint64_t x2)
 // CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call  @llvm.vector.insert.nxv24f16.nxv8f16( [[TMP1]],  [[X2:%.*]], i64 16)
 // CPP-CHECK-NEXT:    ret  [[TMP2]]
 //
-svfloat16x3_t test_svcreate3_f16(svfloat16_t x0, svfloat16_t x1, svfloat16_t x2)
+svfloat16x3_t test_svcreate3_f16(svfloat16_t x0, svfloat16_t x1, svfloat16_t x2) ATTR
 {
   return SVE_ACLE_FUNC(svcreate3,_f16,,)(x0, x1, x2);
 }
@@ -200,7 +207,7 @@ svfloat16x3_t test_svcreate3_f16(svfloat16_t x0, svfloat16_t x1, svfloat16_t x2)
 // CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call  @llvm.vector.insert.nxv12f32.nxv4f32( [[TMP1]],  [[X2:%.*]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP2]]
 //
-svfloat32x3_t test_svcreate3_f32(svfloat32_t x0, svfloat32_t x1, svfloat32_t x2)
+svfloat32x3_t test_svcreate3_f32(svfloat32_t x0, svfloat32_t x1, svfloat32_t x2) ATTR
 {
   return SVE_ACLE_FUNC(svcreate3,_f32,,)(x0, x1, x2);
 }
@@ -219,7 +226,7 @@ svfloat32x3_t test_svcreate3_f32(svfloat32_t x0, svfloat32_t x1, svfloat32_t x2)
 // CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call  @llvm.vector.insert.nxv6f64.nxv2f64( [[TMP1]],  [[X2:%.*]], i64 4)
 // CPP-CHECK-NEXT:    ret  [[TMP2]]
 //
-svfloat64x3_t test_svcreate3_f64(svfloat64_t x0, svfloat64_t x1, svfloat64_t x2)
+svfloat64x3_t test_svcreate3_f64(svfloat64_t x0, svfloat64_t x1, svfloat64_t x2) ATTR
 {
   return SVE_ACLE_FUNC(svcreate3,_f64,,)(x0, x1, x2);
 }
diff --git a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_create4-bfloat.c b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_create4-bfloat.c
index 3d372be19c9f0..8ad8019123454 100644
--- a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_create4-bfloat.c
+++ b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_create4-bfloat.c
@@ -3,6 +3,7 @@
 // RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +bf16 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
 // RUN: %clang_cc1 -fclang-abi-compat=latest -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +bf16 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s
 // RUN: %clang_cc1 -fclang-abi-compat=latest -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +bf16 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -DTEST_SME -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +sme -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
 
 // REQUIRES: aarch64-registered-target
 
@@ -15,6 +16,12 @@
 #define SVE_ACLE_FUNC(A1,A2,A3,A4) A1##A2##A3##A4
 #endif
 
+#ifndef TEST_SME
+#define ATTR
+#else
+#define ATTR __arm_streaming
+#endif
+
 // CHECK-LABEL: @test_svcreate4_bf16(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv32bf16.nxv8bf16( poison,  [[X0:%.*]], i64 0)
@@ -31,7 +38,7 @@
 // CPP-CHECK-NEXT:    [[TMP3:%.*]] = tail call  @llvm.vector.insert.nxv32bf16.nxv8bf16( [[TMP2]],  [[X4:%.*]], i64 24)
 // CPP-CHECK-NEXT:    ret  [[TMP3]]
 //
-svbfloat16x4_t test_svcreate4_bf16(svbfloat16_t x0, svbfloat16_t x1, svbfloat16_t x2, svbfloat16_t x4)
+svbfloat16x4_t test_svcreate4_bf16(svbfloat16_t x0, svbfloat16_t x1, svbfloat16_t x2, svbfloat16_t x4) ATTR
 {
   return SVE_ACLE_FUNC(svcreate4,_bf16,,)(x0, x1, x2, x4);
 }
diff --git a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_create4.c b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_create4.c
index ab39e6406dc3d..5953cdd4e207f 100644
--- a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_create4.c
+++ b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_create4.c
@@ -3,6 +3,7 @@
 // RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu -target-feature +sve -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
 // RUN: %clang_cc1 -fclang-abi-compat=latest -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s
 // RUN: %clang_cc1 -fclang-abi-compat=latest -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -DTEST_SME -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +sme -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
 
 // REQUIRES: aarch64-registered-target
 
@@ -15,6 +16,12 @@
 #define SVE_ACLE_FUNC(A1,A2,A3,A4) A1##A2##A3##A4
 #endif
 
+#ifndef TEST_SME
+#define ATTR
+#else
+#define ATTR __arm_streaming
+#endif
+
 // CHECK-LABEL: @test_svcreate4_s8(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv64i8.nxv16i8( poison,  [[X0:%.*]], i64 0)
@@ -31,7 +38,7 @@
 // CPP-CHECK-NEXT:    [[TMP3:%.*]] = tail call  @llvm.vector.insert.nxv64i8.nxv16i8( [[TMP2]],  [[X4:%.*]], i64 48)
 // CPP-CHECK-NEXT:    ret  [[TMP3]]
 //
-svint8x4_t test_svcreate4_s8(svint8_t x0, svint8_t x1, svint8_t x2, svint8_t x4)
+svint8x4_t test_svcreate4_s8(svint8_t x0, svint8_t x1, svint8_t x2, svint8_t x4) ATTR
 {
   return SVE_ACLE_FUNC(svcreate4,_s8,,)(x0, x1, x2, x4);
 }
@@ -52,7 +59,7 @@ svint8x4_t test_svcreate4_s8(svint8_t x0, svint8_t x1, svint8_t x2, svint8_t x4)
 // CPP-CHECK-NEXT:    [[TMP3:%.*]] = tail call  @llvm.vector.insert.nxv32i16.nxv8i16( [[TMP2]],  [[X4:%.*]], i64 24)
 // CPP-CHECK-NEXT:    ret  [[TMP3]]
 //
-svint16x4_t test_svcreate4_s16(svint16_t x0, svint16_t x1, svint16_t x2, svint16_t x4)
+svint16x4_t test_svcreate4_s16(svint16_t x0, svint16_t x1, svint16_t x2, svint16_t x4) ATTR
 {
   return SVE_ACLE_FUNC(svcreate4,_s16,,)(x0, x1, x2, x4);
 }
@@ -73,7 +80,7 @@ svint16x4_t test_svcreate4_s16(svint16_t x0, svint16_t x1, svint16_t x2, svint16
 // CPP-CHECK-NEXT:    [[TMP3:%.*]] = tail call  @llvm.vector.insert.nxv16i32.nxv4i32( [[TMP2]],  [[X4:%.*]], i64 12)
 // CPP-CHECK-NEXT:    ret  [[TMP3]]
 //
-svint32x4_t test_svcreate4_s32(svint32_t x0, svint32_t x1, svint32_t x2, svint32_t x4)
+svint32x4_t test_svcreate4_s32(svint32_t x0, svint32_t x1, svint32_t x2, svint32_t x4) ATTR
 {
   return SVE_ACLE_FUNC(svcreate4,_s32,,)(x0, x1, x2, x4);
 }
@@ -94,7 +101,7 @@ svint32x4_t test_svcreate4_s32(svint32_t x0, svint32_t x1, svint32_t x2, svint32
 // CPP-CHECK-NEXT:    [[TMP3:%.*]] = tail call  @llvm.vector.insert.nxv8i64.nxv2i64( [[TMP2]],  [[X4:%.*]], i64 6)
 // CPP-CHECK-NEXT:    ret  [[TMP3]]
 //
-svint64x4_t test_svcreate4_s64(svint64_t x0, svint64_t x1, svint64_t x2, svint64_t x4)
+svint64x4_t test_svcreate4_s64(svint64_t x0, svint64_t x1, svint64_t x2, svint64_t x4) ATTR
 {
   return SVE_ACLE_FUNC(svcreate4,_s64,,)(x0, x1, x2, x4);
 }
@@ -115,7 +122,7 @@ svint64x4_t test_svcreate4_s64(svint64_t x0, svint64_t x1, svint64_t x2, svint64
 // CPP-CHECK-NEXT:    [[TMP3:%.*]] = tail call  @llvm.vector.insert.nxv64i8.nxv16i8( [[TMP2]],  [[X4:%.*]], i64 48)
 // CPP-CHECK-NEXT:    ret  [[TMP3]]
 //
-svuint8x4_t test_svcreate4_u8(svuint8_t x0, svuint8_t x1, svuint8_t x2, svuint8_t x4)
+svuint8x4_t test_svcreate4_u8(svuint8_t x0, svuint8_t x1, svuint8_t x2, svuint8_t x4) ATTR
 {
   return SVE_ACLE_FUNC(svcreate4,_u8,,)(x0, x1, x2, x4);
 }
@@ -136,7 +143,7 @@ svuint8x4_t test_svcreate4_u8(svuint8_t x0, svuint8_t x1, svuint8_t x2, svuint8_
 // CPP-CHECK-NEXT:    [[TMP3:%.*]] = tail call  @llvm.vector.insert.nxv32i16.nxv8i16( [[TMP2]],  [[X4:%.*]], i64 24)
 // CPP-CHECK-NEXT:    ret  [[TMP3]]
 //
-svuint16x4_t test_svcreate4_u16(svuint16_t x0, svuint16_t x1, svuint16_t x2, svuint16_t x4)
+svuint16x4_t test_svcreate4_u16(svuint16_t x0, svuint16_t x1, svuint16_t x2, svuint16_t x4) ATTR
 {
   return SVE_ACLE_FUNC(svcreate4,_u16,,)(x0, x1, x2, x4);
 }
@@ -157,7 +164,7 @@ svuint16x4_t test_svcreate4_u16(svuint16_t x0, svuint16_t x1, svuint16_t x2, svu
 // CPP-CHECK-NEXT:    [[TMP3:%.*]] = tail call  @llvm.vector.insert.nxv16i32.nxv4i32( [[TMP2]],  [[X4:%.*]], i64 12)
 // CPP-CHECK-NEXT:    ret  [[TMP3]]
 //
-svuint32x4_t test_svcreate4_u32(svuint32_t x0, svuint32_t x1, svuint32_t x2, svuint32_t x4)
+svuint32x4_t test_svcreate4_u32(svuint32_t x0, svuint32_t x1, svuint32_t x2, svuint32_t x4) ATTR
 {
   return SVE_ACLE_FUNC(svcreate4,_u32,,)(x0, x1, x2, x4);
 }
@@ -178,7 +185,7 @@ svuint32x4_t test_svcreate4_u32(svuint32_t x0, svuint32_t x1, svuint32_t x2, svu
 // CPP-CHECK-NEXT:    [[TMP3:%.*]] = tail call  @llvm.vector.insert.nxv8i64.nxv2i64( [[TMP2]],  [[X4:%.*]], i64 6)
 // CPP-CHECK-NEXT:    ret  [[TMP3]]
 //
-svuint64x4_t test_svcreate4_u64(svuint64_t x0, svuint64_t x1, svuint64_t x2, svuint64_t x4)
+svuint64x4_t test_svcreate4_u64(svuint64_t x0, svuint64_t x1, svuint64_t x2, svuint64_t x4) ATTR
 {
   return SVE_ACLE_FUNC(svcreate4,_u64,,)(x0, x1, x2, x4);
 }
@@ -199,7 +206,7 @@ svuint64x4_t test_svcreate4_u64(svuint64_t x0, svuint64_t x1, svuint64_t x2, svu
 // CPP-CHECK-NEXT:    [[TMP3:%.*]] = tail call  @llvm.vector.insert.nxv32f16.nxv8f16( [[TMP2]],  [[X4:%.*]], i64 24)
 // CPP-CHECK-NEXT:    ret  [[TMP3]]
 //
-svfloat16x4_t test_svcreate4_f16(svfloat16_t x0, svfloat16_t x1, svfloat16_t x2, svfloat16_t x4)
+svfloat16x4_t test_svcreate4_f16(svfloat16_t x0, svfloat16_t x1, svfloat16_t x2, svfloat16_t x4) ATTR
 {
   return SVE_ACLE_FUNC(svcreate4,_f16,,)(x0, x1, x2, x4);
 }
@@ -220,7 +227,7 @@ svfloat16x4_t test_svcreate4_f16(svfloat16_t x0, svfloat16_t x1, svfloat16_t x2,
 // CPP-CHECK-NEXT:    [[TMP3:%.*]] = tail call  @llvm.vector.insert.nxv16f32.nxv4f32( [[TMP2]],  [[X4:%.*]], i64 12)
 // CPP-CHECK-NEXT:    ret  [[TMP3]]
 //
-svfloat32x4_t test_svcreate4_f32(svfloat32_t x0, svfloat32_t x1, svfloat32_t x2, svfloat32_t x4)
+svfloat32x4_t test_svcreate4_f32(svfloat32_t x0, svfloat32_t x1, svfloat32_t x2, svfloat32_t x4) ATTR
 {
   return SVE_ACLE_FUNC(svcreate4,_f32,,)(x0, x1, x2, x4);
 }
@@ -241,7 +248,7 @@ svfloat32x4_t test_svcreate4_f32(svfloat32_t x0, svfloat32_t x1, svfloat32_t x2,
 // CPP-CHECK-NEXT:    [[TMP3:%.*]] = tail call  @llvm.vector.insert.nxv8f64.nxv2f64( [[TMP2]],  [[X4:%.*]], i64 6)
 // CPP-CHECK-NEXT:    ret  [[TMP3]]
 //
-svfloat64x4_t test_svcreate4_f64(svfloat64_t x0, svfloat64_t x1, svfloat64_t x2, svfloat64_t x4)
+svfloat64x4_t test_svcreate4_f64(svfloat64_t x0, svfloat64_t x1, svfloat64_t x2, svfloat64_t x4) ATTR
 {
   return SVE_ACLE_FUNC(svcreate4,_f64,,)(x0, x1, x2, x4);
 }
diff --git a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_get2-bfloat.c b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_get2-bfloat.c
index 25dc49a4c2bd3..b9c46b2261f5d 100644
--- a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_get2-bfloat.c
+++ b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_get2-bfloat.c
@@ -3,6 +3,7 @@
 // RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +bf16 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
 // RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +bf16 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s
 // RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +bf16 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -DTEST_SME -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +sme -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
 
 // REQUIRES: aarch64-registered-target
 
@@ -15,6 +16,12 @@
 #define SVE_ACLE_FUNC(A1,A2,A3,A4) A1##A2##A3##A4
 #endif
 
+#ifndef TEST_SME
+#define ATTR
+#else
+#define ATTR __arm_streaming
+#endif
+
 // CHECK-LABEL: @test_svget2_bf16_0(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv8bf16.nxv16bf16( [[TUPLE:%.*]], i64 0)
@@ -25,7 +32,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv8bf16.nxv16bf16( [[TUPLE:%.*]], i64 0)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svbfloat16_t test_svget2_bf16_0(svbfloat16x2_t tuple)
+svbfloat16_t test_svget2_bf16_0(svbfloat16x2_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget2,_bf16,,)(tuple, 0);
 }
@@ -40,7 +47,7 @@ svbfloat16_t test_svget2_bf16_0(svbfloat16x2_t tuple)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv8bf16.nxv16bf16( [[TUPLE:%.*]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svbfloat16_t test_svget2_bf16_1(svbfloat16x2_t tuple)
+svbfloat16_t test_svget2_bf16_1(svbfloat16x2_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget2,_bf16,,)(tuple, 1);
 }
diff --git a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_get2.c b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_get2.c
index 32a84c91b74d4..8cd887aaff407 100644
--- a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_get2.c
+++ b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_get2.c
@@ -5,6 +5,7 @@
 // RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s
 // RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
 // RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
+// RUN: %clang_cc1 -DTEST_SME -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +sme -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
 #include 
 
 #ifdef SVE_OVERLOADED_FORMS
@@ -14,6 +15,12 @@
 #define SVE_ACLE_FUNC(A1,A2,A3,A4) A1##A2##A3##A4
 #endif
 
+#ifndef TEST_SME
+#define ATTR
+#else
+#define ATTR __arm_streaming
+#endif
+
 // CHECK-LABEL: @test_svget2_s8(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv16i8.nxv32i8( [[TUPLE:%.*]], i64 0)
@@ -24,7 +31,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv16i8.nxv32i8( [[TUPLE:%.*]], i64 0)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svint8_t test_svget2_s8(svint8x2_t tuple)
+svint8_t test_svget2_s8(svint8x2_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget2,_s8,,)(tuple, 0);
 }
@@ -39,7 +46,7 @@ svint8_t test_svget2_s8(svint8x2_t tuple)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv8i16.nxv16i16( [[TUPLE:%.*]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svint16_t test_svget2_s16(svint16x2_t tuple)
+svint16_t test_svget2_s16(svint16x2_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget2,_s16,,)(tuple, 1);
 }
@@ -54,7 +61,7 @@ svint16_t test_svget2_s16(svint16x2_t tuple)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv4i32.nxv8i32( [[TUPLE:%.*]], i64 0)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svint32_t test_svget2_s32(svint32x2_t tuple)
+svint32_t test_svget2_s32(svint32x2_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget2,_s32,,)(tuple, 0);
 }
@@ -69,7 +76,7 @@ svint32_t test_svget2_s32(svint32x2_t tuple)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv2i64.nxv4i64( [[TUPLE:%.*]], i64 2)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svint64_t test_svget2_s64(svint64x2_t tuple)
+svint64_t test_svget2_s64(svint64x2_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget2,_s64,,)(tuple, 1);
 }
@@ -84,7 +91,7 @@ svint64_t test_svget2_s64(svint64x2_t tuple)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv16i8.nxv32i8( [[TUPLE:%.*]], i64 0)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svuint8_t test_svget2_u8(svuint8x2_t tuple)
+svuint8_t test_svget2_u8(svuint8x2_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget2,_u8,,)(tuple, 0);
 }
@@ -99,7 +106,7 @@ svuint8_t test_svget2_u8(svuint8x2_t tuple)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv8i16.nxv16i16( [[TUPLE:%.*]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svuint16_t test_svget2_u16(svuint16x2_t tuple)
+svuint16_t test_svget2_u16(svuint16x2_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget2,_u16,,)(tuple, 1);
 }
@@ -114,7 +121,7 @@ svuint16_t test_svget2_u16(svuint16x2_t tuple)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv4i32.nxv8i32( [[TUPLE:%.*]], i64 0)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svuint32_t test_svget2_u32(svuint32x2_t tuple)
+svuint32_t test_svget2_u32(svuint32x2_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget2,_u32,,)(tuple, 0);
 }
@@ -129,7 +136,7 @@ svuint32_t test_svget2_u32(svuint32x2_t tuple)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv2i64.nxv4i64( [[TUPLE:%.*]], i64 2)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svuint64_t test_svget2_u64(svuint64x2_t tuple)
+svuint64_t test_svget2_u64(svuint64x2_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget2,_u64,,)(tuple, 1);
 }
@@ -144,7 +151,7 @@ svuint64_t test_svget2_u64(svuint64x2_t tuple)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv8f16.nxv16f16( [[TUPLE:%.*]], i64 0)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svfloat16_t test_svget2_f16(svfloat16x2_t tuple)
+svfloat16_t test_svget2_f16(svfloat16x2_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget2,_f16,,)(tuple, 0);
 }
@@ -159,7 +166,7 @@ svfloat16_t test_svget2_f16(svfloat16x2_t tuple)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv4f32.nxv8f32( [[TUPLE:%.*]], i64 4)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svfloat32_t test_svget2_f32(svfloat32x2_t tuple)
+svfloat32_t test_svget2_f32(svfloat32x2_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget2,_f32,,)(tuple, 1);
 }
@@ -174,7 +181,7 @@ svfloat32_t test_svget2_f32(svfloat32x2_t tuple)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv2f64.nxv4f64( [[TUPLE:%.*]], i64 0)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svfloat64_t test_svget2_f64(svfloat64x2_t tuple)
+svfloat64_t test_svget2_f64(svfloat64x2_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget2,_f64,,)(tuple, 0);
 }
diff --git a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_get3-bfloat.c b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_get3-bfloat.c
index 47ce6bd19244e..7a991bc7431d0 100644
--- a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_get3-bfloat.c
+++ b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_get3-bfloat.c
@@ -3,6 +3,7 @@
 // RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +bf16 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
 // RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +bf16 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s
 // RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +bf16 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -DTEST_SME -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +sme -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
 
 // REQUIRES: aarch64-registered-target
 
@@ -15,6 +16,12 @@
 #define SVE_ACLE_FUNC(A1,A2,A3,A4) A1##A2##A3##A4
 #endif
 
+#ifndef TEST_SME
+#define ATTR
+#else
+#define ATTR __arm_streaming
+#endif
+
 // CHECK-LABEL: @test_svget3_bf16_0(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv8bf16.nxv24bf16( [[TUPLE:%.*]], i64 0)
@@ -25,7 +32,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv8bf16.nxv24bf16( [[TUPLE:%.*]], i64 0)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svbfloat16_t test_svget3_bf16_0(svbfloat16x3_t tuple)
+svbfloat16_t test_svget3_bf16_0(svbfloat16x3_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget3,_bf16,,)(tuple, 0);
 }
@@ -40,7 +47,7 @@ svbfloat16_t test_svget3_bf16_0(svbfloat16x3_t tuple)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv8bf16.nxv24bf16( [[TUPLE:%.*]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svbfloat16_t test_svget3_bf16_1(svbfloat16x3_t tuple)
+svbfloat16_t test_svget3_bf16_1(svbfloat16x3_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget3,_bf16,,)(tuple, 1);
 }
@@ -55,7 +62,7 @@ svbfloat16_t test_svget3_bf16_1(svbfloat16x3_t tuple)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv8bf16.nxv24bf16( [[TUPLE:%.*]], i64 16)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svbfloat16_t test_svget3_bf16_2(svbfloat16x3_t tuple)
+svbfloat16_t test_svget3_bf16_2(svbfloat16x3_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget3,_bf16,,)(tuple, 2);
 }
diff --git a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_get3.c b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_get3.c
index 54847152dee7c..de7c3c303ffcb 100644
--- a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_get3.c
+++ b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_get3.c
@@ -5,6 +5,8 @@
 // RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s
 // RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
 // RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
+// RUN: %clang_cc1 -DTEST_SME -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +sme -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
+
 #include 
 
 #ifdef SVE_OVERLOADED_FORMS
@@ -14,6 +16,12 @@
 #define SVE_ACLE_FUNC(A1,A2,A3,A4) A1##A2##A3##A4
 #endif
 
+#ifndef TEST_SME
+#define ATTR
+#else
+#define ATTR __arm_streaming
+#endif
+
 // CHECK-LABEL: @test_svget3_s8(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv16i8.nxv48i8( [[TUPLE:%.*]], i64 0)
@@ -24,7 +32,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv16i8.nxv48i8( [[TUPLE:%.*]], i64 0)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svint8_t test_svget3_s8(svint8x3_t tuple)
+svint8_t test_svget3_s8(svint8x3_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget3,_s8,,)(tuple, 0);
 }
@@ -39,7 +47,7 @@ svint8_t test_svget3_s8(svint8x3_t tuple)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv8i16.nxv24i16( [[TUPLE:%.*]], i64 16)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svint16_t test_svget3_s16(svint16x3_t tuple)
+svint16_t test_svget3_s16(svint16x3_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget3,_s16,,)(tuple, 2);
 }
@@ -54,7 +62,7 @@ svint16_t test_svget3_s16(svint16x3_t tuple)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv4i32.nxv12i32( [[TUPLE:%.*]], i64 4)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svint32_t test_svget3_s32(svint32x3_t tuple)
+svint32_t test_svget3_s32(svint32x3_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget3,_s32,,)(tuple, 1);
 }
@@ -69,7 +77,7 @@ svint32_t test_svget3_s32(svint32x3_t tuple)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv2i64.nxv6i64( [[TUPLE:%.*]], i64 0)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svint64_t test_svget3_s64(svint64x3_t tuple)
+svint64_t test_svget3_s64(svint64x3_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget3,_s64,,)(tuple, 0);
 }
@@ -84,7 +92,7 @@ svint64_t test_svget3_s64(svint64x3_t tuple)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv16i8.nxv48i8( [[TUPLE:%.*]], i64 32)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svuint8_t test_svget3_u8(svuint8x3_t tuple)
+svuint8_t test_svget3_u8(svuint8x3_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget3,_u8,,)(tuple, 2);
 }
@@ -99,7 +107,7 @@ svuint8_t test_svget3_u8(svuint8x3_t tuple)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv8i16.nxv24i16( [[TUPLE:%.*]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svuint16_t test_svget3_u16(svuint16x3_t tuple)
+svuint16_t test_svget3_u16(svuint16x3_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget3,_u16,,)(tuple, 1);
 }
@@ -114,7 +122,7 @@ svuint16_t test_svget3_u16(svuint16x3_t tuple)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv4i32.nxv12i32( [[TUPLE:%.*]], i64 0)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svuint32_t test_svget3_u32(svuint32x3_t tuple)
+svuint32_t test_svget3_u32(svuint32x3_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget3,_u32,,)(tuple, 0);
 }
@@ -129,7 +137,7 @@ svuint32_t test_svget3_u32(svuint32x3_t tuple)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv2i64.nxv6i64( [[TUPLE:%.*]], i64 4)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svuint64_t test_svget3_u64(svuint64x3_t tuple)
+svuint64_t test_svget3_u64(svuint64x3_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget3,_u64,,)(tuple, 2);
 }
@@ -144,7 +152,7 @@ svuint64_t test_svget3_u64(svuint64x3_t tuple)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv8f16.nxv24f16( [[TUPLE:%.*]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svfloat16_t test_svget3_f16(svfloat16x3_t tuple)
+svfloat16_t test_svget3_f16(svfloat16x3_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget3,_f16,,)(tuple, 1);
 }
@@ -159,7 +167,7 @@ svfloat16_t test_svget3_f16(svfloat16x3_t tuple)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv4f32.nxv12f32( [[TUPLE:%.*]], i64 0)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svfloat32_t test_svget3_f32(svfloat32x3_t tuple)
+svfloat32_t test_svget3_f32(svfloat32x3_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget3,_f32,,)(tuple, 0);
 }
@@ -174,7 +182,7 @@ svfloat32_t test_svget3_f32(svfloat32x3_t tuple)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv2f64.nxv6f64( [[TUPLE:%.*]], i64 4)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svfloat64_t test_svget3_f64(svfloat64x3_t tuple)
+svfloat64_t test_svget3_f64(svfloat64x3_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget3,_f64,,)(tuple, 2);
 }
diff --git a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_get4-bfloat.c b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_get4-bfloat.c
index 454b3bf38bd31..3a5e282bfdfa3 100644
--- a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_get4-bfloat.c
+++ b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_get4-bfloat.c
@@ -3,6 +3,7 @@
 // RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +bf16 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
 // RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +bf16 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s
 // RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +bf16 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -DTEST_SME -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +sme -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
 
 // REQUIRES: aarch64-registered-target
 
@@ -15,6 +16,12 @@
 #define SVE_ACLE_FUNC(A1,A2,A3,A4) A1##A2##A3##A4
 #endif
 
+#ifndef TEST_SME
+#define ATTR
+#else
+#define ATTR __arm_streaming
+#endif
+
 // CHECK-LABEL: @test_svget4_bf16_0(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv8bf16.nxv32bf16( [[TUPLE:%.*]], i64 0)
@@ -25,7 +32,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv8bf16.nxv32bf16( [[TUPLE:%.*]], i64 0)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svbfloat16_t test_svget4_bf16_0(svbfloat16x4_t tuple)
+svbfloat16_t test_svget4_bf16_0(svbfloat16x4_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget4,_bf16,,)(tuple, 0);
 }
@@ -40,7 +47,7 @@ svbfloat16_t test_svget4_bf16_0(svbfloat16x4_t tuple)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv8bf16.nxv32bf16( [[TUPLE:%.*]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svbfloat16_t test_svget4_bf16_1(svbfloat16x4_t tuple)
+svbfloat16_t test_svget4_bf16_1(svbfloat16x4_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget4,_bf16,,)(tuple, 1);
 }
@@ -55,7 +62,7 @@ svbfloat16_t test_svget4_bf16_1(svbfloat16x4_t tuple)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv8bf16.nxv32bf16( [[TUPLE:%.*]], i64 16)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svbfloat16_t test_svget4_bf16_2(svbfloat16x4_t tuple)
+svbfloat16_t test_svget4_bf16_2(svbfloat16x4_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget4,_bf16,,)(tuple, 2);
 }
@@ -70,7 +77,7 @@ svbfloat16_t test_svget4_bf16_2(svbfloat16x4_t tuple)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv8bf16.nxv32bf16( [[TUPLE:%.*]], i64 24)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svbfloat16_t test_svget4_bf16_3(svbfloat16x4_t tuple)
+svbfloat16_t test_svget4_bf16_3(svbfloat16x4_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget4,_bf16,,)(tuple, 3);
 }
diff --git a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_get4.c b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_get4.c
index 13f8c2a2906ef..9b4f9e5332a57 100644
--- a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_get4.c
+++ b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_get4.c
@@ -5,6 +5,7 @@
 // RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s
 // RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
 // RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
+// RUN: %clang_cc1 -DTEST_SME -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +sme -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
 #include 
 
 #ifdef SVE_OVERLOADED_FORMS
@@ -14,6 +15,12 @@
 #define SVE_ACLE_FUNC(A1,A2,A3,A4) A1##A2##A3##A4
 #endif
 
+#ifndef TEST_SME
+#define ATTR
+#else
+#define ATTR __arm_streaming
+#endif
+
 // NOTE: For these tests clang converts the struct parameter into
 // several parameters, one for each member of the original struct.
 // CHECK-LABEL: @test_svget4_s8(
@@ -26,7 +33,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv16i8.nxv64i8( [[TUPLE:%.*]], i64 0)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svint8_t test_svget4_s8(svint8x4_t tuple)
+svint8_t test_svget4_s8(svint8x4_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget4,_s8,,)(tuple, 0);
 }
@@ -41,7 +48,7 @@ svint8_t test_svget4_s8(svint8x4_t tuple)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv8i16.nxv32i16( [[TUPLE:%.*]], i64 16)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svint16_t test_svget4_s16(svint16x4_t tuple)
+svint16_t test_svget4_s16(svint16x4_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget4,_s16,,)(tuple, 2);
 }
@@ -56,7 +63,7 @@ svint16_t test_svget4_s16(svint16x4_t tuple)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv4i32.nxv16i32( [[TUPLE:%.*]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svint32_t test_svget4_s32(svint32x4_t tuple)
+svint32_t test_svget4_s32(svint32x4_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget4,_s32,,)(tuple, 2);
 }
@@ -71,7 +78,7 @@ svint32_t test_svget4_s32(svint32x4_t tuple)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv2i64.nxv8i64( [[TUPLE:%.*]], i64 6)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svint64_t test_svget4_s64(svint64x4_t tuple)
+svint64_t test_svget4_s64(svint64x4_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget4,_s64,,)(tuple, 3);
 }
@@ -86,7 +93,7 @@ svint64_t test_svget4_s64(svint64x4_t tuple)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv16i8.nxv64i8( [[TUPLE:%.*]], i64 32)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svuint8_t test_svget4_u8(svuint8x4_t tuple)
+svuint8_t test_svget4_u8(svuint8x4_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget4,_u8,,)(tuple, 2);
 }
@@ -101,7 +108,7 @@ svuint8_t test_svget4_u8(svuint8x4_t tuple)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv8i16.nxv32i16( [[TUPLE:%.*]], i64 24)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svuint16_t test_svget4_u16(svuint16x4_t tuple)
+svuint16_t test_svget4_u16(svuint16x4_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget4,_u16,,)(tuple, 3);
 }
@@ -116,7 +123,7 @@ svuint16_t test_svget4_u16(svuint16x4_t tuple)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv4i32.nxv16i32( [[TUPLE:%.*]], i64 0)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svuint32_t test_svget4_u32(svuint32x4_t tuple)
+svuint32_t test_svget4_u32(svuint32x4_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget4,_u32,,)(tuple, 0);
 }
@@ -131,7 +138,7 @@ svuint32_t test_svget4_u32(svuint32x4_t tuple)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv2i64.nxv8i64( [[TUPLE:%.*]], i64 6)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svuint64_t test_svget4_u64(svuint64x4_t tuple)
+svuint64_t test_svget4_u64(svuint64x4_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget4,_u64,,)(tuple, 3);
 }
@@ -146,7 +153,7 @@ svuint64_t test_svget4_u64(svuint64x4_t tuple)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv8f16.nxv32f16( [[TUPLE:%.*]], i64 16)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svfloat16_t test_svget4_f16(svfloat16x4_t tuple)
+svfloat16_t test_svget4_f16(svfloat16x4_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget4,_f16,,)(tuple, 2);
 }
@@ -161,7 +168,7 @@ svfloat16_t test_svget4_f16(svfloat16x4_t tuple)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv4f32.nxv16f32( [[TUPLE:%.*]], i64 0)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svfloat32_t test_svget4_f32(svfloat32x4_t tuple)
+svfloat32_t test_svget4_f32(svfloat32x4_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget4,_f32,,)(tuple, 0);
 }
@@ -176,7 +183,7 @@ svfloat32_t test_svget4_f32(svfloat32x4_t tuple)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv2f64.nxv8f64( [[TUPLE:%.*]], i64 4)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svfloat64_t test_svget4_f64(svfloat64x4_t tuple)
+svfloat64_t test_svget4_f64(svfloat64x4_t tuple) ATTR
 {
   return SVE_ACLE_FUNC(svget4,_f64,,)(tuple, 2);
 }
diff --git a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_set2-bfloat.c b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_set2-bfloat.c
index e54cf82370c3b..8d683784ddf8c 100644
--- a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_set2-bfloat.c
+++ b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_set2-bfloat.c
@@ -3,6 +3,7 @@
 // RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +bf16 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
 // RUN: %clang_cc1 -fclang-abi-compat=latest -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +bf16 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s
 // RUN: %clang_cc1 -fclang-abi-compat=latest -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +bf16 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -DTEST_SME -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +sme -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
 
 // REQUIRES: aarch64-registered-target
 
@@ -15,6 +16,12 @@
 #define SVE_ACLE_FUNC(A1,A2,A3,A4) A1##A2##A3##A4
 #endif
 
+#ifndef TEST_SME
+#define ATTR
+#else
+#define ATTR __arm_streaming
+#endif
+
 // CHECK-LABEL: @test_svset2_bf16_0(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv16bf16.nxv8bf16( [[TUPLE:%.*]],  [[X:%.*]], i64 0)
@@ -25,7 +32,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv16bf16.nxv8bf16( [[TUPLE:%.*]],  [[X:%.*]], i64 0)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svbfloat16x2_t test_svset2_bf16_0(svbfloat16x2_t tuple, svbfloat16_t x)
+svbfloat16x2_t test_svset2_bf16_0(svbfloat16x2_t tuple, svbfloat16_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset2,_bf16,,)(tuple, 0, x);
 }
@@ -40,7 +47,7 @@ svbfloat16x2_t test_svset2_bf16_0(svbfloat16x2_t tuple, svbfloat16_t x)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv16bf16.nxv8bf16( [[TUPLE:%.*]],  [[X:%.*]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svbfloat16x2_t test_svset2_bf16_1(svbfloat16x2_t tuple, svbfloat16_t x)
+svbfloat16x2_t test_svset2_bf16_1(svbfloat16x2_t tuple, svbfloat16_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset2,_bf16,,)(tuple, 1, x);
 }
diff --git a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_set2.c b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_set2.c
index 9ae3011549207..b2bf4ad08aa9e 100644
--- a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_set2.c
+++ b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_set2.c
@@ -3,6 +3,7 @@
 // RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
 // RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s
 // RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -DTEST_SME -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +sme -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
 
 // REQUIRES: aarch64-registered-target
 
@@ -15,6 +16,12 @@
 #define SVE_ACLE_FUNC(A1,A2,A3,A4) A1##A2##A3##A4
 #endif
 
+#ifndef TEST_SME
+#define ATTR
+#else
+#define ATTR __arm_streaming
+#endif
+
 // CHECK-LABEL: @test_svset2_s8(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv32i8.nxv16i8( [[TUPLE:%.*]],  [[X:%.*]], i64 16)
@@ -25,7 +32,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv32i8.nxv16i8( [[TUPLE:%.*]],  [[X:%.*]], i64 16)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svint8x2_t test_svset2_s8(svint8x2_t tuple, svint8_t x)
+svint8x2_t test_svset2_s8(svint8x2_t tuple, svint8_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset2,_s8,,)(tuple, 1, x);
 }
@@ -40,7 +47,7 @@ svint8x2_t test_svset2_s8(svint8x2_t tuple, svint8_t x)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv16i16.nxv8i16( [[TUPLE:%.*]],  [[X:%.*]], i64 0)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svint16x2_t test_svset2_s16(svint16x2_t tuple, svint16_t x)
+svint16x2_t test_svset2_s16(svint16x2_t tuple, svint16_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset2,_s16,,)(tuple, 0, x);
 }
@@ -55,7 +62,7 @@ svint16x2_t test_svset2_s16(svint16x2_t tuple, svint16_t x)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv8i32.nxv4i32( [[TUPLE:%.*]],  [[X:%.*]], i64 4)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svint32x2_t test_svset2_s32(svint32x2_t tuple, svint32_t x)
+svint32x2_t test_svset2_s32(svint32x2_t tuple, svint32_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset2,_s32,,)(tuple, 1, x);
 }
@@ -70,7 +77,7 @@ svint32x2_t test_svset2_s32(svint32x2_t tuple, svint32_t x)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv4i64.nxv2i64( [[TUPLE:%.*]],  [[X:%.*]], i64 0)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svint64x2_t test_svset2_s64(svint64x2_t tuple, svint64_t x)
+svint64x2_t test_svset2_s64(svint64x2_t tuple, svint64_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset2,_s64,,)(tuple, 0, x);
 }
@@ -85,7 +92,7 @@ svint64x2_t test_svset2_s64(svint64x2_t tuple, svint64_t x)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv32i8.nxv16i8( [[TUPLE:%.*]],  [[X:%.*]], i64 16)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svuint8x2_t test_svset2_u8(svuint8x2_t tuple, svuint8_t x)
+svuint8x2_t test_svset2_u8(svuint8x2_t tuple, svuint8_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset2,_u8,,)(tuple, 1, x);
 }
@@ -100,7 +107,7 @@ svuint8x2_t test_svset2_u8(svuint8x2_t tuple, svuint8_t x)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv16i16.nxv8i16( [[TUPLE:%.*]],  [[X:%.*]], i64 0)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svuint16x2_t test_svset2_u16(svuint16x2_t tuple, svuint16_t x)
+svuint16x2_t test_svset2_u16(svuint16x2_t tuple, svuint16_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset2,_u16,,)(tuple, 0, x);
 }
@@ -115,7 +122,7 @@ svuint16x2_t test_svset2_u16(svuint16x2_t tuple, svuint16_t x)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv8i32.nxv4i32( [[TUPLE:%.*]],  [[X:%.*]], i64 4)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svuint32x2_t test_svset2_u32(svuint32x2_t tuple, svuint32_t x)
+svuint32x2_t test_svset2_u32(svuint32x2_t tuple, svuint32_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset2,_u32,,)(tuple, 1, x);
 }
@@ -130,7 +137,7 @@ svuint32x2_t test_svset2_u32(svuint32x2_t tuple, svuint32_t x)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv4i64.nxv2i64( [[TUPLE:%.*]],  [[X:%.*]], i64 0)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svuint64x2_t test_svset2_u64(svuint64x2_t tuple, svuint64_t x)
+svuint64x2_t test_svset2_u64(svuint64x2_t tuple, svuint64_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset2,_u64,,)(tuple, 0, x);
 }
@@ -145,7 +152,7 @@ svuint64x2_t test_svset2_u64(svuint64x2_t tuple, svuint64_t x)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv16f16.nxv8f16( [[TUPLE:%.*]],  [[X:%.*]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svfloat16x2_t test_svset2_f16(svfloat16x2_t tuple, svfloat16_t x)
+svfloat16x2_t test_svset2_f16(svfloat16x2_t tuple, svfloat16_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset2,_f16,,)(tuple, 1, x);
 }
@@ -160,7 +167,7 @@ svfloat16x2_t test_svset2_f16(svfloat16x2_t tuple, svfloat16_t x)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv8f32.nxv4f32( [[TUPLE:%.*]],  [[X:%.*]], i64 0)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svfloat32x2_t test_svset2_f32(svfloat32x2_t tuple, svfloat32_t x)
+svfloat32x2_t test_svset2_f32(svfloat32x2_t tuple, svfloat32_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset2,_f32,,)(tuple, 0, x);
 }
@@ -175,7 +182,7 @@ svfloat32x2_t test_svset2_f32(svfloat32x2_t tuple, svfloat32_t x)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv4f64.nxv2f64( [[TUPLE:%.*]],  [[X:%.*]], i64 2)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svfloat64x2_t test_svset2_f64(svfloat64x2_t tuple, svfloat64_t x)
+svfloat64x2_t test_svset2_f64(svfloat64x2_t tuple, svfloat64_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset2,_f64,,)(tuple, 1, x);
 }
diff --git a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_set3-bfloat.c b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_set3-bfloat.c
index fffdd0cc36e63..d488576c4be7a 100644
--- a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_set3-bfloat.c
+++ b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_set3-bfloat.c
@@ -3,6 +3,7 @@
 // RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +bf16 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
 // RUN: %clang_cc1 -fclang-abi-compat=latest -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +bf16 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s
 // RUN: %clang_cc1 -fclang-abi-compat=latest -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +bf16 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -DTEST_SME -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +sme -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
 
 // REQUIRES: aarch64-registered-target
 
@@ -15,6 +16,11 @@
 #define SVE_ACLE_FUNC(A1,A2,A3,A4) A1##A2##A3##A4
 #endif
 
+#ifndef TEST_SME
+#define ATTR
+#else
+#define ATTR __arm_streaming
+#endif
 
 // CHECK-LABEL: @test_svset3_bf16_0(
 // CHECK-NEXT:  entry:
@@ -26,7 +32,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv24bf16.nxv8bf16( [[TUPLE:%.*]],  [[X:%.*]], i64 0)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svbfloat16x3_t test_svset3_bf16_0(svbfloat16x3_t tuple, svbfloat16_t x)
+svbfloat16x3_t test_svset3_bf16_0(svbfloat16x3_t tuple, svbfloat16_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset3,_bf16,,)(tuple, 0, x);
 }
@@ -41,7 +47,7 @@ svbfloat16x3_t test_svset3_bf16_0(svbfloat16x3_t tuple, svbfloat16_t x)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv24bf16.nxv8bf16( [[TUPLE:%.*]],  [[X:%.*]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svbfloat16x3_t test_svset3_bf16_1(svbfloat16x3_t tuple, svbfloat16_t x)
+svbfloat16x3_t test_svset3_bf16_1(svbfloat16x3_t tuple, svbfloat16_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset3,_bf16,,)(tuple, 1, x);
 }
@@ -56,7 +62,7 @@ svbfloat16x3_t test_svset3_bf16_1(svbfloat16x3_t tuple, svbfloat16_t x)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv24bf16.nxv8bf16( [[TUPLE:%.*]],  [[X:%.*]], i64 16)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svbfloat16x3_t test_svset3_bf16_2(svbfloat16x3_t tuple, svbfloat16_t x)
+svbfloat16x3_t test_svset3_bf16_2(svbfloat16x3_t tuple, svbfloat16_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset3,_bf16,,)(tuple, 2, x);
 }
diff --git a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_set3.c b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_set3.c
index 1b9191cc8a330..9d10e6afca935 100644
--- a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_set3.c
+++ b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_set3.c
@@ -3,6 +3,7 @@
 // RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
 // RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s
 // RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -DTEST_SME -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +sme -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
 
 // REQUIRES: aarch64-registered-target
 
@@ -15,6 +16,11 @@
 #define SVE_ACLE_FUNC(A1,A2,A3,A4) A1##A2##A3##A4
 #endif
 
+#ifndef TEST_SME
+#define ATTR
+#else
+#define ATTR __arm_streaming
+#endif
 
 // NOTE: For these tests clang converts the struct parameter into
 // several parameters, one for each member of the original struct.
@@ -28,7 +34,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv48i8.nxv16i8( [[TUPLE:%.*]],  [[X:%.*]], i64 16)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svint8x3_t test_svset3_s8(svint8x3_t tuple, svint8_t x)
+svint8x3_t test_svset3_s8(svint8x3_t tuple, svint8_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset3,_s8,,)(tuple, 1, x);
 }
@@ -43,7 +49,7 @@ svint8x3_t test_svset3_s8(svint8x3_t tuple, svint8_t x)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv24i16.nxv8i16( [[TUPLE:%.*]],  [[X:%.*]], i64 16)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svint16x3_t test_svset3_s16(svint16x3_t tuple, svint16_t x)
+svint16x3_t test_svset3_s16(svint16x3_t tuple, svint16_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset3,_s16,,)(tuple, 2, x);
 }
@@ -58,7 +64,7 @@ svint16x3_t test_svset3_s16(svint16x3_t tuple, svint16_t x)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv12i32.nxv4i32( [[TUPLE:%.*]],  [[X:%.*]], i64 0)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svint32x3_t test_svset3_s32(svint32x3_t tuple, svint32_t x)
+svint32x3_t test_svset3_s32(svint32x3_t tuple, svint32_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset3,_s32,,)(tuple, 0, x);
 }
@@ -73,7 +79,7 @@ svint32x3_t test_svset3_s32(svint32x3_t tuple, svint32_t x)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv6i64.nxv2i64( [[TUPLE:%.*]],  [[X:%.*]], i64 2)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svint64x3_t test_svset3_s64(svint64x3_t tuple, svint64_t x)
+svint64x3_t test_svset3_s64(svint64x3_t tuple, svint64_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset3,_s64,,)(tuple, 1, x);
 }
@@ -88,7 +94,7 @@ svint64x3_t test_svset3_s64(svint64x3_t tuple, svint64_t x)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv48i8.nxv16i8( [[TUPLE:%.*]],  [[X:%.*]], i64 32)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svuint8x3_t test_svset3_u8(svuint8x3_t tuple, svuint8_t x)
+svuint8x3_t test_svset3_u8(svuint8x3_t tuple, svuint8_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset3,_u8,,)(tuple, 2, x);
 }
@@ -103,7 +109,7 @@ svuint8x3_t test_svset3_u8(svuint8x3_t tuple, svuint8_t x)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv24i16.nxv8i16( [[TUPLE:%.*]],  [[X:%.*]], i64 0)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svuint16x3_t test_svset3_u16(svuint16x3_t tuple, svuint16_t x)
+svuint16x3_t test_svset3_u16(svuint16x3_t tuple, svuint16_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset3,_u16,,)(tuple, 0, x);
 }
@@ -118,7 +124,7 @@ svuint16x3_t test_svset3_u16(svuint16x3_t tuple, svuint16_t x)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv12i32.nxv4i32( [[TUPLE:%.*]],  [[X:%.*]], i64 4)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svuint32x3_t test_svset3_u32(svuint32x3_t tuple, svuint32_t x)
+svuint32x3_t test_svset3_u32(svuint32x3_t tuple, svuint32_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset3,_u32,,)(tuple, 1, x);
 }
@@ -133,7 +139,7 @@ svuint32x3_t test_svset3_u32(svuint32x3_t tuple, svuint32_t x)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv6i64.nxv2i64( [[TUPLE:%.*]],  [[X:%.*]], i64 4)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svuint64x3_t test_svset3_u64(svuint64x3_t tuple, svuint64_t x)
+svuint64x3_t test_svset3_u64(svuint64x3_t tuple, svuint64_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset3,_u64,,)(tuple, 2, x);
 }
@@ -148,7 +154,7 @@ svuint64x3_t test_svset3_u64(svuint64x3_t tuple, svuint64_t x)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv24f16.nxv8f16( [[TUPLE:%.*]],  [[X:%.*]], i64 0)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svfloat16x3_t test_svset3_f16(svfloat16x3_t tuple, svfloat16_t x)
+svfloat16x3_t test_svset3_f16(svfloat16x3_t tuple, svfloat16_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset3,_f16,,)(tuple, 0, x);
 }
@@ -163,7 +169,7 @@ svfloat16x3_t test_svset3_f16(svfloat16x3_t tuple, svfloat16_t x)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv12f32.nxv4f32( [[TUPLE:%.*]],  [[X:%.*]], i64 4)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svfloat32x3_t test_svset3_f32(svfloat32x3_t tuple, svfloat32_t x)
+svfloat32x3_t test_svset3_f32(svfloat32x3_t tuple, svfloat32_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset3,_f32,,)(tuple, 1, x);
 }
@@ -178,7 +184,7 @@ svfloat32x3_t test_svset3_f32(svfloat32x3_t tuple, svfloat32_t x)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv6f64.nxv2f64( [[TUPLE:%.*]],  [[X:%.*]], i64 4)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svfloat64x3_t test_svset3_f64(svfloat64x3_t tuple, svfloat64_t x)
+svfloat64x3_t test_svset3_f64(svfloat64x3_t tuple, svfloat64_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset3,_f64,,)(tuple, 2, x);
 }
diff --git a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_set4-bfloat.c b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_set4-bfloat.c
index d1d5fb02d5463..f7124ac2ac4b7 100644
--- a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_set4-bfloat.c
+++ b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_set4-bfloat.c
@@ -3,6 +3,7 @@
 // RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +bf16 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
 // RUN: %clang_cc1 -fclang-abi-compat=latest -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +bf16 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s
 // RUN: %clang_cc1 -fclang-abi-compat=latest -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +bf16 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -DTEST_SME -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +sme -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
 
 // REQUIRES: aarch64-registered-target
 
@@ -15,6 +16,11 @@
 #define SVE_ACLE_FUNC(A1,A2,A3,A4) A1##A2##A3##A4
 #endif
 
+#ifndef TEST_SME
+#define ATTR
+#else
+#define ATTR __arm_streaming
+#endif
 
 // CHECK-LABEL: @test_svset4_bf16_0(
 // CHECK-NEXT:  entry:
@@ -26,7 +32,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv32bf16.nxv8bf16( [[TUPLE:%.*]],  [[X:%.*]], i64 0)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svbfloat16x4_t test_svset4_bf16_0(svbfloat16x4_t tuple, svbfloat16_t x)
+svbfloat16x4_t test_svset4_bf16_0(svbfloat16x4_t tuple, svbfloat16_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset4,_bf16,,)(tuple, 0, x);
 }
@@ -41,7 +47,7 @@ svbfloat16x4_t test_svset4_bf16_0(svbfloat16x4_t tuple, svbfloat16_t x)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv32bf16.nxv8bf16( [[TUPLE:%.*]],  [[X:%.*]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svbfloat16x4_t test_svset4_bf16_1(svbfloat16x4_t tuple, svbfloat16_t x)
+svbfloat16x4_t test_svset4_bf16_1(svbfloat16x4_t tuple, svbfloat16_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset4,_bf16,,)(tuple, 1, x);
 }
@@ -56,7 +62,7 @@ svbfloat16x4_t test_svset4_bf16_1(svbfloat16x4_t tuple, svbfloat16_t x)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv32bf16.nxv8bf16( [[TUPLE:%.*]],  [[X:%.*]], i64 16)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svbfloat16x4_t test_svset4_bf16_2(svbfloat16x4_t tuple, svbfloat16_t x)
+svbfloat16x4_t test_svset4_bf16_2(svbfloat16x4_t tuple, svbfloat16_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset4,_bf16,,)(tuple, 2, x);
 }
@@ -71,7 +77,7 @@ svbfloat16x4_t test_svset4_bf16_2(svbfloat16x4_t tuple, svbfloat16_t x)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv32bf16.nxv8bf16( [[TUPLE:%.*]],  [[X:%.*]], i64 24)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svbfloat16x4_t test_svset4_bf16_3(svbfloat16x4_t tuple, svbfloat16_t x)
+svbfloat16x4_t test_svset4_bf16_3(svbfloat16x4_t tuple, svbfloat16_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset4,_bf16,,)(tuple, 3, x);
 }
diff --git a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_set4.c b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_set4.c
index e4ece8c2a65ff..ce35bfb83c88d 100644
--- a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_set4.c
+++ b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_set4.c
@@ -3,6 +3,7 @@
 // RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
 // RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s
 // RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -DTEST_SME -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +sme -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
 
 // REQUIRES: aarch64-registered-target
 
@@ -15,6 +16,11 @@
 #define SVE_ACLE_FUNC(A1,A2,A3,A4) A1##A2##A3##A4
 #endif
 
+#ifndef TEST_SME
+#define ATTR
+#else
+#define ATTR __arm_streaming
+#endif
 
 // CHECK-LABEL: @test_svset4_s8(
 // CHECK-NEXT:  entry:
@@ -26,7 +32,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv64i8.nxv16i8( [[TUPLE:%.*]],  [[X:%.*]], i64 16)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svint8x4_t test_svset4_s8(svint8x4_t tuple, svint8_t x)
+svint8x4_t test_svset4_s8(svint8x4_t tuple, svint8_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset4,_s8,,)(tuple, 1, x);
 }
@@ -41,7 +47,7 @@ svint8x4_t test_svset4_s8(svint8x4_t tuple, svint8_t x)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv32i16.nxv8i16( [[TUPLE:%.*]],  [[X:%.*]], i64 24)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svint16x4_t test_svset4_s16(svint16x4_t tuple, svint16_t x)
+svint16x4_t test_svset4_s16(svint16x4_t tuple, svint16_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset4,_s16,,)(tuple, 3, x);
 }
@@ -56,7 +62,7 @@ svint16x4_t test_svset4_s16(svint16x4_t tuple, svint16_t x)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv16i32.nxv4i32( [[TUPLE:%.*]],  [[X:%.*]], i64 4)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svint32x4_t test_svset4_s32(svint32x4_t tuple, svint32_t x)
+svint32x4_t test_svset4_s32(svint32x4_t tuple, svint32_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset4,_s32,,)(tuple, 1, x);
 }
@@ -71,7 +77,7 @@ svint32x4_t test_svset4_s32(svint32x4_t tuple, svint32_t x)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv8i64.nxv2i64( [[TUPLE:%.*]],  [[X:%.*]], i64 2)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svint64x4_t test_svset4_s64(svint64x4_t tuple, svint64_t x)
+svint64x4_t test_svset4_s64(svint64x4_t tuple, svint64_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset4,_s64,,)(tuple, 1, x);
 }
@@ -86,7 +92,7 @@ svint64x4_t test_svset4_s64(svint64x4_t tuple, svint64_t x)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv64i8.nxv16i8( [[TUPLE:%.*]],  [[X:%.*]], i64 48)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svuint8x4_t test_svset4_u8(svuint8x4_t tuple, svuint8_t x)
+svuint8x4_t test_svset4_u8(svuint8x4_t tuple, svuint8_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset4,_u8,,)(tuple, 3, x);
 }
@@ -101,7 +107,7 @@ svuint8x4_t test_svset4_u8(svuint8x4_t tuple, svuint8_t x)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv32i16.nxv8i16( [[TUPLE:%.*]],  [[X:%.*]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svuint16x4_t test_svset4_u16(svuint16x4_t tuple, svuint16_t x)
+svuint16x4_t test_svset4_u16(svuint16x4_t tuple, svuint16_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset4,_u16,,)(tuple, 1, x);
 }
@@ -116,7 +122,7 @@ svuint16x4_t test_svset4_u16(svuint16x4_t tuple, svuint16_t x)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv16i32.nxv4i32( [[TUPLE:%.*]],  [[X:%.*]], i64 4)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svuint32x4_t test_svset4_u32(svuint32x4_t tuple, svuint32_t x)
+svuint32x4_t test_svset4_u32(svuint32x4_t tuple, svuint32_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset4,_u32,,)(tuple, 1, x);
 }
@@ -131,7 +137,7 @@ svuint32x4_t test_svset4_u32(svuint32x4_t tuple, svuint32_t x)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv8i64.nxv2i64( [[TUPLE:%.*]],  [[X:%.*]], i64 6)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svuint64x4_t test_svset4_u64(svuint64x4_t tuple, svuint64_t x)
+svuint64x4_t test_svset4_u64(svuint64x4_t tuple, svuint64_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset4,_u64,,)(tuple, 3, x);
 }
@@ -146,7 +152,7 @@ svuint64x4_t test_svset4_u64(svuint64x4_t tuple, svuint64_t x)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv32f16.nxv8f16( [[TUPLE:%.*]],  [[X:%.*]], i64 8)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svfloat16x4_t test_svset4_f16(svfloat16x4_t tuple, svfloat16_t x)
+svfloat16x4_t test_svset4_f16(svfloat16x4_t tuple, svfloat16_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset4,_f16,,)(tuple, 1, x);
 }
@@ -161,7 +167,7 @@ svfloat16x4_t test_svset4_f16(svfloat16x4_t tuple, svfloat16_t x)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv16f32.nxv4f32( [[TUPLE:%.*]],  [[X:%.*]], i64 4)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svfloat32x4_t test_svset4_f32(svfloat32x4_t tuple, svfloat32_t x)
+svfloat32x4_t test_svset4_f32(svfloat32x4_t tuple, svfloat32_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset4,_f32,,)(tuple, 1, x);
 }
@@ -176,7 +182,7 @@ svfloat32x4_t test_svset4_f32(svfloat32x4_t tuple, svfloat32_t x)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv8f64.nxv2f64( [[TUPLE:%.*]],  [[X:%.*]], i64 6)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svfloat64x4_t test_svset4_f64(svfloat64x4_t tuple, svfloat64_t x)
+svfloat64x4_t test_svset4_f64(svfloat64x4_t tuple, svfloat64_t x) ATTR
 {
   return SVE_ACLE_FUNC(svset4,_f64,,)(tuple, 3, x);
 }
diff --git a/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_create2_bool.c b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_create2_bool.c
index eb5a19b1d9d32..c4cbec11bf324 100644
--- a/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_create2_bool.c
+++ b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_create2_bool.c
@@ -5,9 +5,14 @@
 // RUN: | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
 // RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s \
 // RUN: | opt -S -passes=mem2reg,tailcallelim | FileCheck %s
+// RUN: %clang_cc1 -DTEST_SME2 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s \
+// RUN: | opt -S -passes=mem2reg,tailcallelim | FileCheck %s
 // RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s\
 // RUN: | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -DTEST_SME -triple aarch64-none-linux-gnu -target-feature +sme2 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s \
+// RUN: | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
 // RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
+// RUN: %clang_cc1 -DTEST_SME2 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
 
 // REQUIRES: aarch64-registered-target
 
@@ -20,19 +25,25 @@
 #define SVE_ACLE_FUNC(A1,A2,A3,A4) A1##A2##A3##A4
 #endif
 
-// CHECK-LABEL: @test_svcreate2_s8(
+#ifndef TEST_SME2
+#define ATTR
+#else
+#define ATTR __arm_streaming
+#endif
+
+// CHECK-LABEL: @test_svcreate2_b(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv32i1.nxv16i1( poison,  [[X0:%.*]], i64 0)
 // CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.vector.insert.nxv32i1.nxv16i1( [[TMP0]],  [[X1:%.*]], i64 16)
 // CHECK-NEXT:    ret  [[TMP1]]
 //
-// CPP-CHECK-LABEL: @_Z17test_svcreate2_s8u10__SVBool_tS_(
+// CPP-CHECK-LABEL: @_Z16test_svcreate2_bu10__SVBool_tS_(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv32i1.nxv16i1( poison,  [[X0:%.*]], i64 0)
 // CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.vector.insert.nxv32i1.nxv16i1( [[TMP0]],  [[X1:%.*]], i64 16)
 // CPP-CHECK-NEXT:    ret  [[TMP1]]
 //
-svboolx2_t test_svcreate2_s8(svbool_t x0, svbool_t x1)
+svboolx2_t test_svcreate2_b(svbool_t x0, svbool_t x1) ATTR
 {
-  return SVE_ACLE_FUNC(svcreate2,_b8,,)(x0, x1);
+  return SVE_ACLE_FUNC(svcreate2,_b,,)(x0, x1);
 }
diff --git a/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_create4_bool.c b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_create4_bool.c
index 3f99ee6e79c07..bc889d9f60183 100644
--- a/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_create4_bool.c
+++ b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_create4_bool.c
@@ -5,9 +5,14 @@
 // RUN: | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
 // RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s \
 // RUN: | opt -S -passes=mem2reg,tailcallelim | FileCheck %s
+// RUN: %clang_cc1 -DTEST_SME2 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s \
+// RUN: | opt -S -passes=mem2reg,tailcallelim | FileCheck %s
 // RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s\
 // RUN: | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -DTEST_SME2 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s \
+// RUN: | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
 // RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
+// RUN: %clang_cc1 -DTEST_SME2 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
 
 // REQUIRES: aarch64-registered-target
 
@@ -20,7 +25,13 @@
 #define SVE_ACLE_FUNC(A1,A2,A3,A4) A1##A2##A3##A4
 #endif
 
-// CHECK-LABEL: @test_svcreate4_b8(
+#ifndef TEST_SME2
+#define ATTR
+#else
+#define ATTR __arm_streaming
+#endif
+
+// CHECK-LABEL: @test_svcreate4_b(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv64i1.nxv16i1( poison,  [[X0:%.*]], i64 0)
 // CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.vector.insert.nxv64i1.nxv16i1( [[TMP0]],  [[X1:%.*]], i64 16)
@@ -28,7 +39,7 @@
 // CHECK-NEXT:    [[TMP3:%.*]] = tail call  @llvm.vector.insert.nxv64i1.nxv16i1( [[TMP2]],  [[X4:%.*]], i64 48)
 // CHECK-NEXT:    ret  [[TMP3]]
 //
-// CPP-CHECK-LABEL: @_Z17test_svcreate4_b8u10__SVBool_tS_S_S_(
+// CPP-CHECK-LABEL: @_Z16test_svcreate4_bu10__SVBool_tS_S_S_(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv64i1.nxv16i1( poison,  [[X0:%.*]], i64 0)
 // CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.vector.insert.nxv64i1.nxv16i1( [[TMP0]],  [[X1:%.*]], i64 16)
@@ -36,7 +47,7 @@
 // CPP-CHECK-NEXT:    [[TMP3:%.*]] = tail call  @llvm.vector.insert.nxv64i1.nxv16i1( [[TMP2]],  [[X4:%.*]], i64 48)
 // CPP-CHECK-NEXT:    ret  [[TMP3]]
 //
-svboolx4_t test_svcreate4_b8(svbool_t x0, svbool_t x1, svbool_t x2, svbool_t x4)
+svboolx4_t test_svcreate4_b(svbool_t x0, svbool_t x1, svbool_t x2, svbool_t x4) ATTR
 {
-  return SVE_ACLE_FUNC(svcreate4,_b8,,)(x0, x1, x2, x4);
+  return SVE_ACLE_FUNC(svcreate4,_b,,)(x0, x1, x2, x4);
 }
diff --git a/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_dot.c b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_dot.c
index 035ba244f9441..97099c1f1061c 100644
--- a/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_dot.c
+++ b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_dot.c
@@ -2,11 +2,20 @@
 // REQUIRES: aarch64-registered-target
 // RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s | opt -S -p mem2reg,instcombine,tailcallelim | FileCheck %s
 // RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -p mem2reg,instcombine,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu -target-feature +sme2 -DTEST_SME2 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s | opt -S -p mem2reg,instcombine,tailcallelim | FileCheck %s
+// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu -target-feature +sme2 -DTEST_SME2 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -p mem2reg,instcombine,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
 // RUN: %clang_cc1 -fclang-abi-compat=latest -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s | opt -S -p mem2reg,instcombine,tailcallelim | FileCheck %s
 // RUN: %clang_cc1 -fclang-abi-compat=latest -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -p mem2reg,instcombine,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
 // RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
+// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu -target-feature +sme2 -DTEST_SME2 -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
 #include 
 
+#ifndef TEST_SME2
+#define ATTR
+#else
+#define ATTR __arm_streaming
+#endif
+
 #ifdef SVE_OVERLOADED_FORMS
 // A simple used,unused... macro, long enough to represent any SVE builtin.
 #define SVE_ACLE_FUNC(A1,A2_UNUSED,A3) A1##A3
@@ -24,7 +33,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sve.sdot.x2.nxv4i32( [[OP1:%.*]],  [[OP2:%.*]],  [[OP3:%.*]])
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svint32_t test_svdot_s32_x2(svint32_t op1, svint16_t op2, svint16_t op3)
+svint32_t test_svdot_s32_x2(svint32_t op1, svint16_t op2, svint16_t op3) ATTR
 {
   return SVE_ACLE_FUNC(svdot,_s32_s16,)(op1, op2, op3);
 }
@@ -39,7 +48,7 @@ svint32_t test_svdot_s32_x2(svint32_t op1, svint16_t op2, svint16_t op3)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sve.udot.x2.nxv4i32( [[OP1:%.*]],  [[OP2:%.*]],  [[OP3:%.*]])
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svuint32_t test_svdot_u32_x2(svuint32_t op1, svuint16_t op2, svuint16_t op3)
+svuint32_t test_svdot_u32_x2(svuint32_t op1, svuint16_t op2, svuint16_t op3) ATTR
 {
   return SVE_ACLE_FUNC(svdot,_u32_u16,)(op1, op2, op3);
 }
@@ -54,7 +63,7 @@ svuint32_t test_svdot_u32_x2(svuint32_t op1, svuint16_t op2, svuint16_t op3)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sve.fdot.x2.nxv4f32( [[OP1:%.*]],  [[OP2:%.*]],  [[OP3:%.*]])
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svfloat32_t test_svdot_f32_x2(svfloat32_t op1, svfloat16_t op2, svfloat16_t op3)
+svfloat32_t test_svdot_f32_x2(svfloat32_t op1, svfloat16_t op2, svfloat16_t op3) ATTR
 {
   return SVE_ACLE_FUNC(svdot,_f32_f16,)(op1, op2, op3);
 }
@@ -71,7 +80,7 @@ svfloat32_t test_svdot_f32_x2(svfloat32_t op1, svfloat16_t op2, svfloat16_t op3)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sve.sdot.lane.x2.nxv4i32( [[OP1:%.*]],  [[OP2:%.*]],  [[OP3:%.*]], i32 3)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svint32_t test_svdot_lane_s32_x2(svint32_t op1, svint16_t op2, svint16_t op3)
+svint32_t test_svdot_lane_s32_x2(svint32_t op1, svint16_t op2, svint16_t op3) ATTR
 {
   return SVE_ACLE_FUNC(svdot_lane,_s32_s16,)(op1, op2, op3, 3);
 }
@@ -86,7 +95,7 @@ svint32_t test_svdot_lane_s32_x2(svint32_t op1, svint16_t op2, svint16_t op3)
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sve.udot.lane.x2.nxv4i32( [[OP1:%.*]],  [[OP2:%.*]],  [[OP3:%.*]], i32 3)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svuint32_t test_svdot_lane_u32_x2(svuint32_t op1, svuint16_t op2, svuint16_t op3)
+svuint32_t test_svdot_lane_u32_x2(svuint32_t op1, svuint16_t op2, svuint16_t op3) ATTR
 {
   return SVE_ACLE_FUNC(svdot_lane,_u32_u16,)(op1, op2, op3, 3);
 }
@@ -101,7 +110,7 @@ svuint32_t test_svdot_lane_u32_x2(svuint32_t op1, svuint16_t op2, svuint16_t op3
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sve.fdot.lane.x2.nxv4f32( [[OP1:%.*]],  [[OP2:%.*]],  [[OP3:%.*]], i32 3)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svfloat32_t test_svdot_lane_f32_x2(svfloat32_t op1, svfloat16_t op2, svfloat16_t op3)
+svfloat32_t test_svdot_lane_f32_x2(svfloat32_t op1, svfloat16_t op2, svfloat16_t op3) ATTR
 {
   return SVE_ACLE_FUNC(svdot_lane,_f32_f16,)(op1, op2, op3, 3);
 }
diff --git a/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_get2_bool.c b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_get2_bool.c
index 5d38f72b34b16..35e6f1b84ab1b 100644
--- a/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_get2_bool.c
+++ b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_get2_bool.c
@@ -5,9 +5,14 @@
 // RUN: | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
 // RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s \
 // RUN: | opt -S -passes=mem2reg,tailcallelim | FileCheck %s
+// RUN: %clang_cc1 -DTEST_SME2 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s \
+// RUN: | opt -S -passes=mem2reg,tailcallelim | FileCheck %s
 // RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s\
 // RUN: | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -DTEST_SME2 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s \
+// RUN: | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
 // RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
+// RUN: %clang_cc1 -DTEST_SME2 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
 // REQUIRES: aarch64-registered-target
 #include 
 
@@ -18,32 +23,38 @@
 #define SVE_ACLE_FUNC(A1,A2,A3,A4) A1##A2##A3##A4
 #endif
 
-// CHECK-LABEL: @test_svget2_b8_0(
+#ifndef TEST_SME2
+#define ATTR
+#else
+#define ATTR __arm_streaming
+#endif
+
+// CHECK-LABEL: @test_svget2_b_0(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv16i1.nxv32i1( [[TUPLE:%.*]], i64 0)
 // CHECK-NEXT:    ret  [[TMP0]]
 //
-// CPP-CHECK-LABEL: @_Z16test_svget2_b8_010svboolx2_t(
+// CPP-CHECK-LABEL: @_Z15test_svget2_b_010svboolx2_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv16i1.nxv32i1( [[TUPLE:%.*]], i64 0)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svbool_t test_svget2_b8_0(svboolx2_t tuple)
+svbool_t test_svget2_b_0(svboolx2_t tuple) ATTR
 {
-  return SVE_ACLE_FUNC(svget2,_b8,,)(tuple, 0);
+  return SVE_ACLE_FUNC(svget2,_b,,)(tuple, 0);
 }
 
-// CHECK-LABEL: @test_svget2_b8_1(
+// CHECK-LABEL: @test_svget2_b_1(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv16i1.nxv32i1( [[TUPLE:%.*]], i64 16)
 // CHECK-NEXT:    ret  [[TMP0]]
 //
-// CPP-CHECK-LABEL: @_Z16test_svget2_b8_110svboolx2_t(
+// CPP-CHECK-LABEL: @_Z15test_svget2_b_110svboolx2_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv16i1.nxv32i1( [[TUPLE:%.*]], i64 16)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svbool_t test_svget2_b8_1(svboolx2_t tuple)
+svbool_t test_svget2_b_1(svboolx2_t tuple) ATTR
 {
-  return SVE_ACLE_FUNC(svget2,_b8,,)(tuple, 1);
+  return SVE_ACLE_FUNC(svget2,_b,,)(tuple, 1);
 }
diff --git a/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_get4_bool.c b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_get4_bool.c
index 821a5be3254eb..8a5f9568e3677 100644
--- a/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_get4_bool.c
+++ b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_get4_bool.c
@@ -5,9 +5,14 @@
 // RUN: | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
 // RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s \
 // RUN: | opt -S -passes=mem2reg,tailcallelim | FileCheck %s
+// RUN: %clang_cc1 -DTEST_SME2 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s \
+// RUN: | opt -S -passes=mem2reg,tailcallelim | FileCheck %s
 // RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s\
 // RUN: | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -DTEST_SME2 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s \
+// RUN: | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
 // RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
+// RUN: %clang_cc1 -DTEST_SME2 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
 
 // REQUIRES: aarch64-registered-target
 
@@ -20,53 +25,59 @@
 #define SVE_ACLE_FUNC(A1,A2,A3,A4) A1##A2##A3##A4
 #endif
 
+#ifndef TEST_SME
+#define ATTR
+#else
+#define ATTR __arm_streaming
+#endif
+
 // NOTE: For these tests clang converts the struct parameter into
 // several parameters, one for each member of the original struct.
-// CHECK-LABEL: @test_svget4_b8_0(
+// CHECK-LABEL: @test_svget4_b_0(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv16i1.nxv64i1( [[TUPLE:%.*]], i64 0)
 // CHECK-NEXT:    ret  [[TMP0]]
 //
-// CPP-CHECK-LABEL: @_Z16test_svget4_b8_010svboolx4_t(
+// CPP-CHECK-LABEL: @_Z15test_svget4_b_010svboolx4_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv16i1.nxv64i1( [[TUPLE:%.*]], i64 0)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svbool_t test_svget4_b8_0(svboolx4_t tuple)
+svbool_t test_svget4_b_0(svboolx4_t tuple) ATTR
 {
-  return SVE_ACLE_FUNC(svget4,_b8,,)(tuple, 0);
+  return SVE_ACLE_FUNC(svget4,_b,,)(tuple, 0);
 }
 
 // NOTE: For these tests clang converts the struct parameter into
 // several parameters, one for each member of the original struct.
-// CHECK-LABEL: @test_svget4_b8_1(
+// CHECK-LABEL: @test_svget4_b_1(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv16i1.nxv64i1( [[TUPLE:%.*]], i64 16)
 // CHECK-NEXT:    ret  [[TMP0]]
 //
-// CPP-CHECK-LABEL: @_Z16test_svget4_b8_110svboolx4_t(
+// CPP-CHECK-LABEL: @_Z15test_svget4_b_110svboolx4_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv16i1.nxv64i1( [[TUPLE:%.*]], i64 16)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svbool_t test_svget4_b8_1(svboolx4_t tuple)
+svbool_t test_svget4_b_1(svboolx4_t tuple) ATTR
 {
-  return SVE_ACLE_FUNC(svget4,_b8,,)(tuple, 1);
+  return SVE_ACLE_FUNC(svget4,_b,,)(tuple, 1);
 }
 
 // NOTE: For these tests clang converts the struct parameter into
 // several parameters, one for each member of the original struct.
-// CHECK-LABEL: @test_svget4_b8_3(
+// CHECK-LABEL: @test_svget4_b_3(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv16i1.nxv64i1( [[TUPLE:%.*]], i64 48)
 // CHECK-NEXT:    ret  [[TMP0]]
 //
-// CPP-CHECK-LABEL: @_Z16test_svget4_b8_310svboolx4_t(
+// CPP-CHECK-LABEL: @_Z15test_svget4_b_310svboolx4_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.extract.nxv16i1.nxv64i1( [[TUPLE:%.*]], i64 48)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svbool_t test_svget4_b8_3(svboolx4_t tuple)
+svbool_t test_svget4_b_3(svboolx4_t tuple) ATTR
 {
-  return SVE_ACLE_FUNC(svget4,_b8,,)(tuple, 3);
+  return SVE_ACLE_FUNC(svget4,_b,,)(tuple, 3);
 }
diff --git a/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_psel.c b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_psel.c
index de3f6a9a57bfe..a61d874a1d9e9 100644
--- a/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_psel.c
+++ b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_psel.c
@@ -11,16 +11,16 @@
 // RUN:   -target-feature +sve2p1 -S -O1 -Werror -emit-llvm -o - -x c++ %s | FileCheck %s -check-prefix=CPP-CHECK
 // RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
 // RUN: %clang_cc1 -triple aarch64-none-linux-gnu \
-// RUN:   -target-feature +sme2 -S -DTEST_SME2 -O1 -Werror -emit-llvm -o - -x c++ %s | FileCheck %s -check-prefix=CPP-CHECK
-// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -DTEST_SME2 -disable-O0-optnone -Werror -Wall -o /dev/null %s
+// RUN:   -target-feature +sme -S -DTEST_SME -O1 -Werror -emit-llvm -o - -x c++ %s | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sme -S -DTEST_SME -disable-O0-optnone -Werror -Wall -o /dev/null %s
 // RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
 
 #include 
 
-#ifndef TEST_SME2
+#ifndef TEST_SME
 #define ATTR
 #else
-#define ATTR __arm_streaming_compatible
+#define ATTR __arm_streaming
 #endif
 
 // CHECK-LABEL: @test_svpsel_lane_b8(
@@ -92,89 +92,3 @@ svbool_t test_svpsel_lane_b32(svbool_t p1, svbool_t p2, uint32_t idx) ATTR {
 svbool_t test_svpsel_lane_b64(svbool_t p1, svbool_t p2, uint32_t idx) ATTR {
   return svpsel_lane_b64(p1, p2, idx + 1);
 }
-
-// CHECK-LABEL: @test_svpsel_lane_c8(
-// CHECK-NEXT:  entry:
-// CHECK-NEXT:    [[ADD:%.*]] = add i32 [[IDX:%.*]], 15
-// CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sve.convert.to.svbool.taarch64.svcountt(target("aarch64.svcount") [[P1:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sve.psel.nxv16i1( [[TMP0]],  [[P2:%.*]], i32 [[ADD]])
-// CHECK-NEXT:    [[TMP2:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.convert.from.svbool.taarch64.svcountt( [[TMP1]])
-// CHECK-NEXT:    ret target("aarch64.svcount") [[TMP2]]
-//
-// CPP-CHECK-LABEL: @_Z19test_svpsel_lane_c8u11__SVCount_tu10__SVBool_tj(
-// CPP-CHECK-NEXT:  entry:
-// CPP-CHECK-NEXT:    [[ADD:%.*]] = add i32 [[IDX:%.*]], 15
-// CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sve.convert.to.svbool.taarch64.svcountt(target("aarch64.svcount") [[P1:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sve.psel.nxv16i1( [[TMP0]],  [[P2:%.*]], i32 [[ADD]])
-// CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.convert.from.svbool.taarch64.svcountt( [[TMP1]])
-// CPP-CHECK-NEXT:    ret target("aarch64.svcount") [[TMP2]]
-//
-svcount_t test_svpsel_lane_c8(svcount_t p1, svbool_t p2, uint32_t idx) ATTR {
-  return svpsel_lane_c8(p1, p2, idx + 15);
-}
-
-// CHECK-LABEL: @test_svpsel_lane_c16(
-// CHECK-NEXT:  entry:
-// CHECK-NEXT:    [[ADD:%.*]] = add i32 [[IDX:%.*]], 7
-// CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sve.convert.to.svbool.taarch64.svcountt(target("aarch64.svcount") [[P1:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sve.convert.from.svbool.nxv8i1( [[P2:%.*]])
-// CHECK-NEXT:    [[TMP2:%.*]] = tail call  @llvm.aarch64.sve.psel.nxv8i1( [[TMP0]],  [[TMP1]], i32 [[ADD]])
-// CHECK-NEXT:    [[TMP3:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.convert.from.svbool.taarch64.svcountt( [[TMP2]])
-// CHECK-NEXT:    ret target("aarch64.svcount") [[TMP3]]
-//
-// CPP-CHECK-LABEL: @_Z20test_svpsel_lane_c16u11__SVCount_tu10__SVBool_tj(
-// CPP-CHECK-NEXT:  entry:
-// CPP-CHECK-NEXT:    [[ADD:%.*]] = add i32 [[IDX:%.*]], 7
-// CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sve.convert.to.svbool.taarch64.svcountt(target("aarch64.svcount") [[P1:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sve.convert.from.svbool.nxv8i1( [[P2:%.*]])
-// CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call  @llvm.aarch64.sve.psel.nxv8i1( [[TMP0]],  [[TMP1]], i32 [[ADD]])
-// CPP-CHECK-NEXT:    [[TMP3:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.convert.from.svbool.taarch64.svcountt( [[TMP2]])
-// CPP-CHECK-NEXT:    ret target("aarch64.svcount") [[TMP3]]
-//
-svcount_t test_svpsel_lane_c16(svcount_t p1, svbool_t p2, uint32_t idx) ATTR {
-  return svpsel_lane_c16(p1, p2, idx + 7);
-}
-
-// CHECK-LABEL: @test_svpsel_lane_c32(
-// CHECK-NEXT:  entry:
-// CHECK-NEXT:    [[ADD:%.*]] = add i32 [[IDX:%.*]], 3
-// CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sve.convert.to.svbool.taarch64.svcountt(target("aarch64.svcount") [[P1:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sve.convert.from.svbool.nxv4i1( [[P2:%.*]])
-// CHECK-NEXT:    [[TMP2:%.*]] = tail call  @llvm.aarch64.sve.psel.nxv4i1( [[TMP0]],  [[TMP1]], i32 [[ADD]])
-// CHECK-NEXT:    [[TMP3:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.convert.from.svbool.taarch64.svcountt( [[TMP2]])
-// CHECK-NEXT:    ret target("aarch64.svcount") [[TMP3]]
-//
-// CPP-CHECK-LABEL: @_Z20test_svpsel_lane_c32u11__SVCount_tu10__SVBool_tj(
-// CPP-CHECK-NEXT:  entry:
-// CPP-CHECK-NEXT:    [[ADD:%.*]] = add i32 [[IDX:%.*]], 3
-// CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sve.convert.to.svbool.taarch64.svcountt(target("aarch64.svcount") [[P1:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sve.convert.from.svbool.nxv4i1( [[P2:%.*]])
-// CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call  @llvm.aarch64.sve.psel.nxv4i1( [[TMP0]],  [[TMP1]], i32 [[ADD]])
-// CPP-CHECK-NEXT:    [[TMP3:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.convert.from.svbool.taarch64.svcountt( [[TMP2]])
-// CPP-CHECK-NEXT:    ret target("aarch64.svcount") [[TMP3]]
-//
-svcount_t test_svpsel_lane_c32(svcount_t p1, svbool_t p2, uint32_t idx) ATTR {
-  return svpsel_lane_c32(p1, p2, idx + 3);
-}
-
-// CHECK-LABEL: @test_svpsel_lane_c64(
-// CHECK-NEXT:  entry:
-// CHECK-NEXT:    [[ADD:%.*]] = add i32 [[IDX:%.*]], 1
-// CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sve.convert.to.svbool.taarch64.svcountt(target("aarch64.svcount") [[P1:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sve.convert.from.svbool.nxv2i1( [[P2:%.*]])
-// CHECK-NEXT:    [[TMP2:%.*]] = tail call  @llvm.aarch64.sve.psel.nxv2i1( [[TMP0]],  [[TMP1]], i32 [[ADD]])
-// CHECK-NEXT:    [[TMP3:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.convert.from.svbool.taarch64.svcountt( [[TMP2]])
-// CHECK-NEXT:    ret target("aarch64.svcount") [[TMP3]]
-//
-// CPP-CHECK-LABEL: @_Z20test_svpsel_lane_c64u11__SVCount_tu10__SVBool_tj(
-// CPP-CHECK-NEXT:  entry:
-// CPP-CHECK-NEXT:    [[ADD:%.*]] = add i32 [[IDX:%.*]], 1
-// CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sve.convert.to.svbool.taarch64.svcountt(target("aarch64.svcount") [[P1:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sve.convert.from.svbool.nxv2i1( [[P2:%.*]])
-// CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call  @llvm.aarch64.sve.psel.nxv2i1( [[TMP0]],  [[TMP1]], i32 [[ADD]])
-// CPP-CHECK-NEXT:    [[TMP3:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.convert.from.svbool.taarch64.svcountt( [[TMP2]])
-// CPP-CHECK-NEXT:    ret target("aarch64.svcount") [[TMP3]]
-//
-svcount_t test_svpsel_lane_c64(svcount_t p1, svbool_t p2, uint32_t idx) ATTR {
-  return svpsel_lane_c64(p1, p2, idx + 1);
-}
diff --git a/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_psel_svcount.c b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_psel_svcount.c
new file mode 100644
index 0000000000000..4b8a582be3013
--- /dev/null
+++ b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_psel_svcount.c
@@ -0,0 +1,110 @@
+// NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py
+// REQUIRES: aarch64-registered-target
+// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu \
+// RUN:   -target-feature +sve2p1 -S -O1 -Werror -emit-llvm -o - %s | FileCheck %s
+// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu \
+// RUN:   -target-feature +sve2p1 -S -O1 -Werror -emit-llvm -o - -x c++ %s | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
+// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu \
+// RUN:   -target-feature +sve2p1 -S -O1 -Werror -emit-llvm -o - %s | FileCheck %s
+// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu \
+// RUN:   -target-feature +sve2p1 -S -O1 -Werror -emit-llvm -o - -x c++ %s | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu \
+// RUN:   -target-feature +sme2 -S -DTEST_SME2 -O1 -Werror -emit-llvm -o - -x c++ %s | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -DTEST_SME2 -disable-O0-optnone -Werror -Wall -o /dev/null %s
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
+
+#include 
+
+#ifndef TEST_SME2
+#define ATTR
+#else
+#define ATTR __arm_streaming
+#endif
+
+// CHECK-LABEL: @test_svpsel_lane_c8(
+// CHECK-NEXT:  entry:
+// CHECK-NEXT:    [[ADD:%.*]] = add i32 [[IDX:%.*]], 15
+// CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sve.convert.to.svbool.taarch64.svcountt(target("aarch64.svcount") [[P1:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sve.psel.nxv16i1( [[TMP0]],  [[P2:%.*]], i32 [[ADD]])
+// CHECK-NEXT:    [[TMP2:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.convert.from.svbool.taarch64.svcountt( [[TMP1]])
+// CHECK-NEXT:    ret target("aarch64.svcount") [[TMP2]]
+//
+// CPP-CHECK-LABEL: @_Z19test_svpsel_lane_c8u11__SVCount_tu10__SVBool_tj(
+// CPP-CHECK-NEXT:  entry:
+// CPP-CHECK-NEXT:    [[ADD:%.*]] = add i32 [[IDX:%.*]], 15
+// CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sve.convert.to.svbool.taarch64.svcountt(target("aarch64.svcount") [[P1:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sve.psel.nxv16i1( [[TMP0]],  [[P2:%.*]], i32 [[ADD]])
+// CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.convert.from.svbool.taarch64.svcountt( [[TMP1]])
+// CPP-CHECK-NEXT:    ret target("aarch64.svcount") [[TMP2]]
+//
+svcount_t test_svpsel_lane_c8(svcount_t p1, svbool_t p2, uint32_t idx) ATTR {
+  return svpsel_lane_c8(p1, p2, idx + 15);
+}
+
+// CHECK-LABEL: @test_svpsel_lane_c16(
+// CHECK-NEXT:  entry:
+// CHECK-NEXT:    [[ADD:%.*]] = add i32 [[IDX:%.*]], 7
+// CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sve.convert.to.svbool.taarch64.svcountt(target("aarch64.svcount") [[P1:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sve.convert.from.svbool.nxv8i1( [[P2:%.*]])
+// CHECK-NEXT:    [[TMP2:%.*]] = tail call  @llvm.aarch64.sve.psel.nxv8i1( [[TMP0]],  [[TMP1]], i32 [[ADD]])
+// CHECK-NEXT:    [[TMP3:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.convert.from.svbool.taarch64.svcountt( [[TMP2]])
+// CHECK-NEXT:    ret target("aarch64.svcount") [[TMP3]]
+//
+// CPP-CHECK-LABEL: @_Z20test_svpsel_lane_c16u11__SVCount_tu10__SVBool_tj(
+// CPP-CHECK-NEXT:  entry:
+// CPP-CHECK-NEXT:    [[ADD:%.*]] = add i32 [[IDX:%.*]], 7
+// CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sve.convert.to.svbool.taarch64.svcountt(target("aarch64.svcount") [[P1:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sve.convert.from.svbool.nxv8i1( [[P2:%.*]])
+// CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call  @llvm.aarch64.sve.psel.nxv8i1( [[TMP0]],  [[TMP1]], i32 [[ADD]])
+// CPP-CHECK-NEXT:    [[TMP3:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.convert.from.svbool.taarch64.svcountt( [[TMP2]])
+// CPP-CHECK-NEXT:    ret target("aarch64.svcount") [[TMP3]]
+//
+svcount_t test_svpsel_lane_c16(svcount_t p1, svbool_t p2, uint32_t idx) ATTR {
+  return svpsel_lane_c16(p1, p2, idx + 7);
+}
+
+// CHECK-LABEL: @test_svpsel_lane_c32(
+// CHECK-NEXT:  entry:
+// CHECK-NEXT:    [[ADD:%.*]] = add i32 [[IDX:%.*]], 3
+// CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sve.convert.to.svbool.taarch64.svcountt(target("aarch64.svcount") [[P1:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sve.convert.from.svbool.nxv4i1( [[P2:%.*]])
+// CHECK-NEXT:    [[TMP2:%.*]] = tail call  @llvm.aarch64.sve.psel.nxv4i1( [[TMP0]],  [[TMP1]], i32 [[ADD]])
+// CHECK-NEXT:    [[TMP3:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.convert.from.svbool.taarch64.svcountt( [[TMP2]])
+// CHECK-NEXT:    ret target("aarch64.svcount") [[TMP3]]
+//
+// CPP-CHECK-LABEL: @_Z20test_svpsel_lane_c32u11__SVCount_tu10__SVBool_tj(
+// CPP-CHECK-NEXT:  entry:
+// CPP-CHECK-NEXT:    [[ADD:%.*]] = add i32 [[IDX:%.*]], 3
+// CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sve.convert.to.svbool.taarch64.svcountt(target("aarch64.svcount") [[P1:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sve.convert.from.svbool.nxv4i1( [[P2:%.*]])
+// CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call  @llvm.aarch64.sve.psel.nxv4i1( [[TMP0]],  [[TMP1]], i32 [[ADD]])
+// CPP-CHECK-NEXT:    [[TMP3:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.convert.from.svbool.taarch64.svcountt( [[TMP2]])
+// CPP-CHECK-NEXT:    ret target("aarch64.svcount") [[TMP3]]
+//
+svcount_t test_svpsel_lane_c32(svcount_t p1, svbool_t p2, uint32_t idx) ATTR {
+  return svpsel_lane_c32(p1, p2, idx + 3);
+}
+
+// CHECK-LABEL: @test_svpsel_lane_c64(
+// CHECK-NEXT:  entry:
+// CHECK-NEXT:    [[ADD:%.*]] = add i32 [[IDX:%.*]], 1
+// CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sve.convert.to.svbool.taarch64.svcountt(target("aarch64.svcount") [[P1:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sve.convert.from.svbool.nxv2i1( [[P2:%.*]])
+// CHECK-NEXT:    [[TMP2:%.*]] = tail call  @llvm.aarch64.sve.psel.nxv2i1( [[TMP0]],  [[TMP1]], i32 [[ADD]])
+// CHECK-NEXT:    [[TMP3:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.convert.from.svbool.taarch64.svcountt( [[TMP2]])
+// CHECK-NEXT:    ret target("aarch64.svcount") [[TMP3]]
+//
+// CPP-CHECK-LABEL: @_Z20test_svpsel_lane_c64u11__SVCount_tu10__SVBool_tj(
+// CPP-CHECK-NEXT:  entry:
+// CPP-CHECK-NEXT:    [[ADD:%.*]] = add i32 [[IDX:%.*]], 1
+// CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.aarch64.sve.convert.to.svbool.taarch64.svcountt(target("aarch64.svcount") [[P1:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call  @llvm.aarch64.sve.convert.from.svbool.nxv2i1( [[P2:%.*]])
+// CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call  @llvm.aarch64.sve.psel.nxv2i1( [[TMP0]],  [[TMP1]], i32 [[ADD]])
+// CPP-CHECK-NEXT:    [[TMP3:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.convert.from.svbool.taarch64.svcountt( [[TMP2]])
+// CPP-CHECK-NEXT:    ret target("aarch64.svcount") [[TMP3]]
+//
+svcount_t test_svpsel_lane_c64(svcount_t p1, svbool_t p2, uint32_t idx) ATTR {
+  return svpsel_lane_c64(p1, p2, idx + 1);
+}
diff --git a/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_set2_bool.c b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_set2_bool.c
index 298e5b80ceb00..75c8d035aedb0 100644
--- a/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_set2_bool.c
+++ b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_set2_bool.c
@@ -5,9 +5,14 @@
 // RUN: | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
 // RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s \
 // RUN: | opt -S -passes=mem2reg,tailcallelim | FileCheck %s
+// RUN: %clang_cc1 -DTEST_SME2 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s \
+// RUN: | opt -S -passes=mem2reg,tailcallelim | FileCheck %s
 // RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s\
 // RUN: | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -DTEST_SME2 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s \
+// RUN: | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
 // RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
+// RUN: %clang_cc1 -DTEST_SME2 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
 
 // REQUIRES: aarch64-registered-target
 
@@ -20,33 +25,39 @@
 #define SVE_ACLE_FUNC(A1,A2,A3,A4) A1##A2##A3##A4
 #endif
 
-// CHECK-LABEL: @test_svset2_b8_0(
+#ifndef TEST_SME2
+#define ATTR
+#else
+#define ATTR __arm_streaming
+#endif
+
+// CHECK-LABEL: @test_svset2_b_0(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv32i1.nxv16i1( [[TUPLE:%.*]],  [[X:%.*]], i64 0)
 // CHECK-NEXT:    ret  [[TMP0]]
 //
-// CPP-CHECK-LABEL: @_Z16test_svset2_b8_010svboolx2_tu10__SVBool_t(
+// CPP-CHECK-LABEL: @_Z15test_svset2_b_010svboolx2_tu10__SVBool_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv32i1.nxv16i1( [[TUPLE:%.*]],  [[X:%.*]], i64 0)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svboolx2_t test_svset2_b8_0(svboolx2_t tuple, svbool_t x)
+svboolx2_t test_svset2_b_0(svboolx2_t tuple, svbool_t x) ATTR
 {
-  return SVE_ACLE_FUNC(svset2,_b8,,)(tuple, 0, x);
+  return SVE_ACLE_FUNC(svset2,_b,,)(tuple, 0, x);
 }
 
-// CHECK-LABEL: @test_svset2_b8_1(
+// CHECK-LABEL: @test_svset2_b_1(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv32i1.nxv16i1( [[TUPLE:%.*]],  [[X:%.*]], i64 16)
 // CHECK-NEXT:    ret  [[TMP0]]
 //
-// CPP-CHECK-LABEL: @_Z16test_svset2_b8_110svboolx2_tu10__SVBool_t(
+// CPP-CHECK-LABEL: @_Z15test_svset2_b_110svboolx2_tu10__SVBool_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv32i1.nxv16i1( [[TUPLE:%.*]],  [[X:%.*]], i64 16)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svboolx2_t test_svset2_b8_1(svboolx2_t tuple, svbool_t x)
+svboolx2_t test_svset2_b_1(svboolx2_t tuple, svbool_t x) ATTR
 {
-  return SVE_ACLE_FUNC(svset2,_b8,,)(tuple, 1, x);
+  return SVE_ACLE_FUNC(svset2,_b,,)(tuple, 1, x);
 }
 
diff --git a/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_set4_bool.c b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_set4_bool.c
index 746a8280ac53f..d68810352693b 100644
--- a/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_set4_bool.c
+++ b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_set4_bool.c
@@ -5,9 +5,14 @@
 // RUN: | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
 // RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s \
 // RUN: | opt -S -passes=mem2reg,tailcallelim | FileCheck %s
+// RUN: %clang_cc1 -DTEST_SME2 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s \
+// RUN: | opt -S -passes=mem2reg,tailcallelim | FileCheck %s
 // RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s\
 // RUN: | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -DTEST_SME2 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s \
+// RUN: | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
 // RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
+// RUN: %clang_cc1 -DTEST_SME2 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
 // REQUIRES: aarch64-registered-target
 
 #include 
@@ -19,48 +24,53 @@
 #define SVE_ACLE_FUNC(A1,A2,A3,A4) A1##A2##A3##A4
 #endif
 
+#ifndef TEST_SME2
+#define ATTR
+#else
+#define ATTR __arm_streaming
+#endif
 
-// CHECK-LABEL: @test_svset4_b8_0(
+// CHECK-LABEL: @test_svset4_b_0(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv64i1.nxv16i1( [[TUPLE:%.*]],  [[X:%.*]], i64 0)
 // CHECK-NEXT:    ret  [[TMP0]]
 //
-// CPP-CHECK-LABEL: @_Z16test_svset4_b8_010svboolx4_tu10__SVBool_t(
+// CPP-CHECK-LABEL: @_Z15test_svset4_b_010svboolx4_tu10__SVBool_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv64i1.nxv16i1( [[TUPLE:%.*]],  [[X:%.*]], i64 0)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svboolx4_t test_svset4_b8_0(svboolx4_t tuple, svbool_t x)
+svboolx4_t test_svset4_b_0(svboolx4_t tuple, svbool_t x) ATTR
 {
-  return SVE_ACLE_FUNC(svset4,_b8,,)(tuple, 0, x);
+  return SVE_ACLE_FUNC(svset4,_b,,)(tuple, 0, x);
 }
 
-// CHECK-LABEL: @test_svset4_b8_1(
+// CHECK-LABEL: @test_svset4_b_1(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv64i1.nxv16i1( [[TUPLE:%.*]],  [[X:%.*]], i64 16)
 // CHECK-NEXT:    ret  [[TMP0]]
 //
-// CPP-CHECK-LABEL: @_Z16test_svset4_b8_110svboolx4_tu10__SVBool_t(
+// CPP-CHECK-LABEL: @_Z15test_svset4_b_110svboolx4_tu10__SVBool_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv64i1.nxv16i1( [[TUPLE:%.*]],  [[X:%.*]], i64 16)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svboolx4_t test_svset4_b8_1(svboolx4_t tuple, svbool_t x)
+svboolx4_t test_svset4_b_1(svboolx4_t tuple, svbool_t x) ATTR
 {
-  return SVE_ACLE_FUNC(svset4,_b8,,)(tuple, 1, x);
+  return SVE_ACLE_FUNC(svset4,_b,,)(tuple, 1, x);
 }
 
-// CHECK-LABEL: @test_svset4_b8_3(
+// CHECK-LABEL: @test_svset4_b_3(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv64i1.nxv16i1( [[TUPLE:%.*]],  [[X:%.*]], i64 48)
 // CHECK-NEXT:    ret  [[TMP0]]
 //
-// CPP-CHECK-LABEL: @_Z16test_svset4_b8_310svboolx4_tu10__SVBool_t(
+// CPP-CHECK-LABEL: @_Z15test_svset4_b_310svboolx4_tu10__SVBool_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call  @llvm.vector.insert.nxv64i1.nxv16i1( [[TUPLE:%.*]],  [[X:%.*]], i64 48)
 // CPP-CHECK-NEXT:    ret  [[TMP0]]
 //
-svboolx4_t test_svset4_b8_3(svboolx4_t tuple, svbool_t x)
+svboolx4_t test_svset4_b_3(svboolx4_t tuple, svbool_t x) ATTR
 {
-  return SVE_ACLE_FUNC(svset4,_b8,,)(tuple, 3, x);
+  return SVE_ACLE_FUNC(svset4,_b,,)(tuple, 3, x);
 }
diff --git a/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_undef_bool.c b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_undef_bool.c
new file mode 100644
index 0000000000000..1cc7476ac5386
--- /dev/null
+++ b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_undef_bool.c
@@ -0,0 +1,36 @@
+// NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py UTC_ARGS: --version 4
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -O2 -Werror -Wall -emit-llvm -o - %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -O2 -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -O2 -Werror -Wall -emit-llvm -o - %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -O2 -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
+
+// REQUIRES: aarch64-registered-target
+
+#include 
+
+// CHECK-LABEL: define dso_local  @test_svundef2_b(
+// CPP-CHECK-LABEL: define dso_local  @_Z15test_svundef2_bv(
+//
+// CHECK-NEXT:  entry:
+// CHECK-NEXT:    ret  undef
+//
+// CPP-CHECK-NEXT:  entry:
+// CPP-CHECK-NEXT:    ret  undef
+//
+svboolx2_t test_svundef2_b() {
+    return svundef2_b();
+}
+
+// CHECK-LABEL: define dso_local  @test_svundef4_b(
+// CPP-CHECK-LABEL: define dso_local  @_Z15test_svundef4_bv(
+//
+// CHECK-NEXT:  entry:
+// CHECK-NEXT:    ret  undef
+//
+// CPP-CHECK-NEXT:  entry:
+// CPP-CHECK-NEXT:    ret  undef
+//
+svboolx4_t test_svundef4_b() {
+    return svundef4_b();
+}
diff --git a/clang/test/CodeGen/aarch64-targetattr.c b/clang/test/CodeGen/aarch64-targetattr.c
index 9664b723a2b2c..02da18264da0a 100644
--- a/clang/test/CodeGen/aarch64-targetattr.c
+++ b/clang/test/CodeGen/aarch64-targetattr.c
@@ -97,19 +97,19 @@ void minusarch() {}
 // CHECK: attributes #0 = { {{.*}} "target-features"="+crc,+fp-armv8,+lse,+neon,+ras,+rdm,+v8.1a,+v8.2a,+v8a" }
 // CHECK: attributes #1 = { {{.*}} "target-features"="+crc,+fp-armv8,+fullfp16,+lse,+neon,+ras,+rdm,+sve,+v8.1a,+v8.2a,+v8a" }
 // CHECK: attributes #2 = { {{.*}} "target-features"="+crc,+fp-armv8,+fullfp16,+lse,+neon,+ras,+rdm,+sve,+sve2,+v8.1a,+v8.2a,+v8a" }
-// CHECK: attributes #3 = { {{.*}} "target-features"="+bf16,+crc,+dotprod,+fp-armv8,+fullfp16,+i8mm,+lse,+neon,+ras,+rcpc,+rdm,+sve,+sve2,+v8.1a,+v8.2a,+v8.3a,+v8.4a,+v8.5a,+v8.6a,+v8a" }
-// CHECK: attributes #4 = { {{.*}} "target-cpu"="cortex-a710" "target-features"="+bf16,+crc,+dotprod,+flagm,+fp-armv8,+fp16fml,+fullfp16,+i8mm,+lse,+mte,+neon,+pauth,+ras,+rcpc,+rdm,+sb,+sve,+sve2,+sve2-bitperm" }
+// CHECK: attributes #3 = { {{.*}} "target-features"="+bf16,+complxnum,+crc,+dotprod,+fp-armv8,+fullfp16,+i8mm,+jsconv,+lse,+neon,+ras,+rcpc,+rdm,+sve,+sve2,+v8.1a,+v8.2a,+v8.3a,+v8.4a,+v8.5a,+v8.6a,+v8a" }
+// CHECK: attributes #4 = { {{.*}} "target-cpu"="cortex-a710" "target-features"="+bf16,+complxnum,+crc,+dotprod,+flagm,+fp-armv8,+fp16fml,+fullfp16,+i8mm,+jsconv,+lse,+mte,+neon,+pauth,+ras,+rcpc,+rdm,+sb,+sve,+sve2,+sve2-bitperm" }
 // CHECK: attributes #5 = { {{.*}} "tune-cpu"="cortex-a710" }
 // CHECK: attributes #6 = { {{.*}} "target-cpu"="generic" }
 // CHECK: attributes #7 = { {{.*}} "tune-cpu"="generic" }
 // CHECK: attributes #8 = { {{.*}} "target-cpu"="neoverse-n1" "target-features"="+aes,+crc,+dotprod,+fp-armv8,+fullfp16,+lse,+neon,+ras,+rcpc,+rdm,+sha2,+spe,+ssbs" "tune-cpu"="cortex-a710" }
 // CHECK: attributes #9 = { {{.*}} "target-features"="+fp-armv8,+fullfp16,+neon,+sve" "tune-cpu"="cortex-a710" }
-// CHECK: attributes #10 = { {{.*}} "target-cpu"="neoverse-v1" "target-features"="+aes,+bf16,+crc,+dotprod,+fp-armv8,+fp16fml,+fullfp16,+i8mm,+lse,+neon,+rand,+ras,+rcpc,+rdm,+sha2,+sha3,+sm4,+spe,+ssbs,+sve,+sve2" }
-// CHECK: attributes #11 = { {{.*}} "target-cpu"="neoverse-v1" "target-features"="+aes,+bf16,+crc,+dotprod,+fp-armv8,+fp16fml,+fullfp16,+i8mm,+lse,+neon,+rand,+ras,+rcpc,+rdm,+sha2,+sha3,+sm4,+spe,+ssbs,-sve" }
+// CHECK: attributes #10 = { {{.*}} "target-cpu"="neoverse-v1" "target-features"="+aes,+bf16,+complxnum,+crc,+dotprod,+fp-armv8,+fp16fml,+fullfp16,+i8mm,+jsconv,+lse,+neon,+rand,+ras,+rcpc,+rdm,+sha2,+sha3,+sm4,+spe,+ssbs,+sve,+sve2" }
+// CHECK: attributes #11 = { {{.*}} "target-cpu"="neoverse-v1" "target-features"="+aes,+bf16,+complxnum,+crc,+dotprod,+fp-armv8,+fp16fml,+fullfp16,+i8mm,+jsconv,+lse,+neon,+rand,+ras,+rcpc,+rdm,+sha2,+sha3,+sm4,+spe,+ssbs,-sve" }
 // CHECK: attributes #12 = { {{.*}} "target-features"="+fp-armv8,+fullfp16,+neon,+sve" }
 // CHECK: attributes #13 = { {{.*}} "target-features"="+fp-armv8,+fullfp16,+neon,+sve,-sve2" }
 // CHECK: attributes #14 = { {{.*}} "target-features"="+fullfp16" }
-// CHECK: attributes #15 = { {{.*}} "target-cpu"="neoverse-n1" "target-features"="+aes,+bf16,+crc,+dotprod,+fp-armv8,+fullfp16,+i8mm,+lse,+neon,+ras,+rcpc,+rdm,+sha2,+spe,+ssbs,+sve,+sve2,+v8.1a,+v8.2a,+v8.3a,+v8.4a,+v8.5a,+v8.6a,+v8a" "tune-cpu"="cortex-a710" }
-// CHECK: attributes #16 = { {{.*}} "branch-target-enforcement"="true" {{.*}} "target-features"="+aes,+bf16,+crc,+dotprod,+fp-armv8,+fullfp16,+i8mm,+lse,+neon,+ras,+rcpc,+rdm,+sha2,+spe,+ssbs,+sve,+sve2,+v8.1a,+v8.2a,+v8.3a,+v8.4a,+v8.5a,+v8.6a,+v8a" "tune-cpu"="cortex-a710" }
+// CHECK: attributes #15 = { {{.*}} "target-cpu"="neoverse-n1" "target-features"="+aes,+bf16,+complxnum,+crc,+dotprod,+fp-armv8,+fullfp16,+i8mm,+jsconv,+lse,+neon,+ras,+rcpc,+rdm,+sha2,+spe,+ssbs,+sve,+sve2,+v8.1a,+v8.2a,+v8.3a,+v8.4a,+v8.5a,+v8.6a,+v8a" "tune-cpu"="cortex-a710" }
+// CHECK: attributes #16 = { {{.*}} "branch-target-enforcement"="true" "guarded-control-stack"="true" {{.*}} "target-features"="+aes,+bf16,+complxnum,+crc,+dotprod,+fp-armv8,+fullfp16,+i8mm,+jsconv,+lse,+neon,+ras,+rcpc,+rdm,+sha2,+spe,+ssbs,+sve,+sve2,+v8.1a,+v8.2a,+v8.3a,+v8.4a,+v8.5a,+v8.6a,+v8a" "tune-cpu"="cortex-a710" }
 // CHECK: attributes #17 = { {{.*}} "target-features"="-neon" }
 // CHECK: attributes #18 = { {{.*}} "target-features"="-v9.3a" }
diff --git a/clang/test/CodeGen/aarch64-varargs.c b/clang/test/CodeGen/aarch64-varargs.c
index f2ce7d59a8cce..44b87029e7b3d 100644
--- a/clang/test/CodeGen/aarch64-varargs.c
+++ b/clang/test/CodeGen/aarch64-varargs.c
@@ -601,7 +601,7 @@ typedef struct __attribute__((aligned(32))) {
   __int128 val;
 } overaligned_int128_struct;
 overaligned_int128_struct overaligned_int128_struct_test(void) {
-// CHECK-LABEL: define{{.*}} void @overaligned_int128_struct_test(ptr noalias sret(%struct.overaligned_int128_struct) align 32 %agg.result)
+// CHECK-LABEL: define{{.*}} void @overaligned_int128_struct_test(ptr dead_on_unwind noalias writable sret(%struct.overaligned_int128_struct) align 32 %agg.result)
   return va_arg(the_list, overaligned_int128_struct);
 // CHECK: [[GR_OFFS:%[a-z_0-9]+]] = load i32, ptr getelementptr inbounds (%struct.__va_list, ptr @the_list, i32 0, i32 3)
 // CHECK: [[EARLY_ONSTACK:%[a-z_0-9]+]] = icmp sge i32 [[GR_OFFS]], 0
@@ -804,7 +804,7 @@ typedef struct {
   __int128 val __attribute__((aligned(32)));
 } overaligned_int128_struct_member;
 overaligned_int128_struct_member overaligned_int128_struct_member_test(void) {
-// CHECK-LABEL: define{{.*}} void @overaligned_int128_struct_member_test(ptr noalias sret(%struct.overaligned_int128_struct_member) align 32 %agg.result)
+// CHECK-LABEL: define{{.*}} void @overaligned_int128_struct_member_test(ptr dead_on_unwind noalias writable sret(%struct.overaligned_int128_struct_member) align 32 %agg.result)
   return va_arg(the_list, overaligned_int128_struct_member);
 // CHECK: [[GR_OFFS:%[a-z_0-9]+]] = load i32, ptr getelementptr inbounds (%struct.__va_list, ptr @the_list, i32 0, i32 3)
 // CHECK: [[EARLY_ONSTACK:%[a-z_0-9]+]] = icmp sge i32 [[GR_OFFS]], 0
diff --git a/clang/test/CodeGen/aggregate-assign-call.c b/clang/test/CodeGen/aggregate-assign-call.c
index b5c7f213dcbe2..d6571269456a4 100644
--- a/clang/test/CodeGen/aggregate-assign-call.c
+++ b/clang/test/CodeGen/aggregate-assign-call.c
@@ -63,11 +63,11 @@ struct S baz(int i, volatile int *j) {
     //
     // O1: call void @llvm.lifetime.end.p0({{[^,]*}}, ptr %[[TMP1_ALLOCA]])
     //
-    // O1: call void @foo_int(ptr sret(%struct.S) align 4 %[[TMP1_ALLOCA]],
+    // O1: call void @foo_int(ptr dead_on_unwind writable sret(%struct.S) align 4 %[[TMP1_ALLOCA]],
     // O1: call void @llvm.memcpy
     // O1: call void @llvm.lifetime.end.p0({{[^,]*}}, ptr %[[TMP1_ALLOCA]])
     // O1: call void @llvm.lifetime.start.p0({{[^,]*}}, ptr %[[TMP2_ALLOCA]])
-    // O1: call void @foo_int(ptr sret(%struct.S) align 4 %[[TMP2_ALLOCA]],
+    // O1: call void @foo_int(ptr dead_on_unwind writable sret(%struct.S) align 4 %[[TMP2_ALLOCA]],
     // O1: call void @llvm.memcpy
     // O1: call void @llvm.lifetime.end.p0({{[^,]*}}, ptr %[[TMP2_ALLOCA]])
     r = foo_int(({
diff --git a/clang/test/CodeGen/aligned-sret.c b/clang/test/CodeGen/aligned-sret.c
index 4759d69219a6a..4e1f86e7f07a6 100644
--- a/clang/test/CodeGen/aligned-sret.c
+++ b/clang/test/CodeGen/aligned-sret.c
@@ -4,7 +4,7 @@ typedef __attribute__((__ext_vector_type__(4),__aligned__(16))) double simd_doub
 typedef struct { simd_double4 columns[4]; } simd_double4x4;
 typedef simd_double4x4 matrix_double4x4;
 
-// CHECK: define{{.*}} void @ident(ptr noalias sret(%struct.simd_double4x4) align 16 %agg.result
+// CHECK: define{{.*}} void @ident(ptr dead_on_unwind noalias writable sret(%struct.simd_double4x4) align 16 %agg.result
 matrix_double4x4 ident(matrix_double4x4 x) {
   return x;
 }
diff --git a/clang/test/CodeGen/arc/arguments.c b/clang/test/CodeGen/arc/arguments.c
index a913984e13f85..648a2ea3ce9c1 100644
--- a/clang/test/CodeGen/arc/arguments.c
+++ b/clang/test/CodeGen/arc/arguments.c
@@ -22,7 +22,7 @@ void cf1(cs1 i) {}
 typedef struct {
   int cc;
 } s2;
-// CHECK: define{{.*}} void @f2(ptr noalias sret(%struct.s2) align 4 %agg.result)
+// CHECK: define{{.*}} void @f2(ptr dead_on_unwind noalias writable sret(%struct.s2) align 4 %agg.result)
 s2 f2(void) {
   s2 foo;
   return foo;
@@ -32,7 +32,7 @@ typedef struct {
   int cc;
   int dd;
 } s3;
-// CHECK: define{{.*}} void @f3(ptr noalias sret(%struct.s3) align 4 %agg.result)
+// CHECK: define{{.*}} void @f3(ptr dead_on_unwind noalias writable sret(%struct.s3) align 4 %agg.result)
 s3 f3(void) {
   s3 foo;
   return foo;
@@ -128,8 +128,8 @@ void st3(s16 a, s16 b, s16 c) {}
 
 // 1 sret + 1 i32 + 2*(i32 coerce) + 4*(i32 coerce) + 1 byval
 s16 st4(int x, s8 a, s16 b, s16 c) { return b; }
-// CHECK: define{{.*}} void @st4(ptr noalias sret(%struct.s16) align 4 %agg.result, i32 inreg noundef %x, i32 inreg %a.coerce0, i32 inreg %a.coerce1, i32 inreg %b.coerce0, i32 inreg %b.coerce1, i32 inreg %b.coerce2, i32 inreg %b.coerce3, { i32, i32, i32, i32 } %c.coerce)
+// CHECK: define{{.*}} void @st4(ptr dead_on_unwind noalias writable sret(%struct.s16) align 4 %agg.result, i32 inreg noundef %x, i32 inreg %a.coerce0, i32 inreg %a.coerce1, i32 inreg %b.coerce0, i32 inreg %b.coerce1, i32 inreg %b.coerce2, i32 inreg %b.coerce3, { i32, i32, i32, i32 } %c.coerce)
 
 // 1 sret + 2*(i32 coerce) + 4*(i32 coerce) + 4*(i32 coerce)
 s16 st5(s8 a, s16 b, s16 c) { return b; }
-// CHECK: define{{.*}} void @st5(ptr noalias sret(%struct.s16) align 4 %agg.result, i32 inreg %a.coerce0, i32 inreg %a.coerce1, i32 inreg %b.coerce0, i32 inreg %b.coerce1, i32 inreg %b.coerce2, i32 inreg %b.coerce3, { i32, i32, i32, i32 } %c.coerce)
+// CHECK: define{{.*}} void @st5(ptr dead_on_unwind noalias writable sret(%struct.s16) align 4 %agg.result, i32 inreg %a.coerce0, i32 inreg %a.coerce1, i32 inreg %b.coerce0, i32 inreg %b.coerce1, i32 inreg %b.coerce2, i32 inreg %b.coerce3, { i32, i32, i32, i32 } %c.coerce)
diff --git a/clang/test/CodeGen/arm-aapcs-vfp.c b/clang/test/CodeGen/arm-aapcs-vfp.c
index 5caf93016cd2e..9fae33f476d35 100644
--- a/clang/test/CodeGen/arm-aapcs-vfp.c
+++ b/clang/test/CodeGen/arm-aapcs-vfp.c
@@ -126,7 +126,7 @@ void test_vfp_stack_gpr_split_1(double a, double b, double c, double d, double e
 // CHECK: define{{.*}} arm_aapcs_vfpcc void @test_vfp_stack_gpr_split_2(double noundef %a, double noundef %b, double noundef %c, double noundef %d, double noundef %e, double noundef %f, double noundef %g, double noundef %h, double noundef %i, i32 noundef %j, [2 x i64] %k.coerce)
 void test_vfp_stack_gpr_split_2(double a, double b, double c, double d, double e, double f, double g, double h, double i, int j, struct_long_long_int k) {}
 
-// CHECK: define{{.*}} arm_aapcs_vfpcc void @test_vfp_stack_gpr_split_3(ptr noalias sret(%struct.struct_long_long_int) align 8 %agg.result, double noundef %a, double noundef %b, double noundef %c, double noundef %d, double noundef %e, double noundef %f, double noundef %g, double noundef %h, double noundef %i, [2 x i64] %k.coerce)
+// CHECK: define{{.*}} arm_aapcs_vfpcc void @test_vfp_stack_gpr_split_3(ptr dead_on_unwind noalias writable sret(%struct.struct_long_long_int) align 8 %agg.result, double noundef %a, double noundef %b, double noundef %c, double noundef %d, double noundef %e, double noundef %f, double noundef %g, double noundef %h, double noundef %i, [2 x i64] %k.coerce)
 struct_long_long_int test_vfp_stack_gpr_split_3(double a, double b, double c, double d, double e, double f, double g, double h, double i, struct_long_long_int k) {}
 
 typedef struct { int a; int b:4; int c; } struct_int_bitfield_int;
diff --git a/clang/test/CodeGen/arm-arguments.c b/clang/test/CodeGen/arm-arguments.c
index 17e4d3abd9767..8fe2016315f27 100644
--- a/clang/test/CodeGen/arm-arguments.c
+++ b/clang/test/CodeGen/arm-arguments.c
@@ -29,13 +29,13 @@ struct s4 { struct s4_0 { int f0; } f0; };
 struct s4 f4(void) {}
 
 // APCS-GNU-LABEL: define{{.*}} void @f5(
-// APCS-GNU: ptr noalias sret
+// APCS-GNU: ptr dead_on_unwind noalias writable sret
 // AAPCS-LABEL: define{{.*}} arm_aapcscc i32 @f5()
 struct s5 { struct { } f0; int f1; };
 struct s5 f5(void) {}
 
 // APCS-GNU-LABEL: define{{.*}} void @f6(
-// APCS-GNU: ptr noalias sret
+// APCS-GNU: ptr dead_on_unwind noalias writable sret
 // AAPCS-LABEL: define{{.*}} arm_aapcscc i32 @f6()
 struct s6 { int f0[1]; };
 struct s6 f6(void) {}
@@ -46,7 +46,7 @@ struct s7 { struct { int : 0; } f0; };
 struct s7 f7(void) {}
 
 // APCS-GNU-LABEL: define{{.*}} void @f8(
-// APCS-GNU: ptr noalias sret
+// APCS-GNU: ptr dead_on_unwind noalias writable sret
 // AAPCS-LABEL: define{{.*}} arm_aapcscc void @f8()
 struct s8 { struct { int : 0; } f0[1]; };
 struct s8 f8(void) {}
@@ -62,7 +62,7 @@ struct s10 { int f0; int : 0; int : 0; };
 struct s10 f10(void) {}
 
 // APCS-GNU-LABEL: define{{.*}} void @f11(
-// APCS-GNU: ptr noalias sret
+// APCS-GNU: ptr dead_on_unwind noalias writable sret
 // AAPCS-LABEL: define{{.*}} arm_aapcscc i32 @f11()
 struct s11 { int : 0; int f0; };
 struct s11 f11(void) {}
@@ -73,7 +73,7 @@ union u12 { char f0; short f1; int f2; };
 union u12 f12(void) {}
 
 // APCS-GNU-LABEL: define{{.*}} void @f13(
-// APCS-GNU: ptr noalias sret
+// APCS-GNU: ptr dead_on_unwind noalias writable sret
 
 // FIXME: This should return a float.
 // AAPCS-FIXME: darm_aapcscc efine float @f13()
@@ -81,7 +81,7 @@ struct s13 { float f0; };
 struct s13 f13(void) {}
 
 // APCS-GNU-LABEL: define{{.*}} void @f14(
-// APCS-GNU: ptr noalias sret
+// APCS-GNU: ptr dead_on_unwind noalias writable sret
 // AAPCS-LABEL: define{{.*}} arm_aapcscc i32 @f14()
 union u14 { float f0; };
 union u14 f14(void) {}
@@ -105,13 +105,13 @@ struct s18 { short f0; char f1 : 4; };
 struct s18 f18(void) {}
 
 // APCS-GNU-LABEL: define{{.*}} void @f19(
-// APCS-GNU: ptr noalias sret
+// APCS-GNU: ptr dead_on_unwind noalias writable sret
 // AAPCS-LABEL: define{{.*}} arm_aapcscc i32 @f19()
 struct s19 { int f0; struct s8 f1; };
 struct s19 f19(void) {}
 
 // APCS-GNU-LABEL: define{{.*}} void @f20(
-// APCS-GNU: ptr noalias sret
+// APCS-GNU: ptr dead_on_unwind noalias writable sret
 // AAPCS-LABEL: define{{.*}} arm_aapcscc i32 @f20()
 struct s20 { struct s8 f1; int f0; };
 struct s20 f20(void) {}
@@ -129,10 +129,10 @@ struct s21 f21(void) {}
 // APCS-GNU-LABEL: define{{.*}} i128 @f27()
 // AAPCS-LABEL: define{{.*}} arm_aapcscc i16 @f22()
 // AAPCS-LABEL: define{{.*}} arm_aapcscc i32 @f23()
-// AAPCS: define{{.*}} arm_aapcscc void @f24({{.*}} noalias sret
-// AAPCS: define{{.*}} arm_aapcscc void @f25({{.*}} noalias sret
-// AAPCS: define{{.*}} arm_aapcscc void @f26({{.*}} noalias sret
-// AAPCS: define{{.*}} arm_aapcscc void @f27({{.*}} noalias sret
+// AAPCS: define{{.*}} arm_aapcscc void @f24({{.*}} dead_on_unwind noalias writable sret
+// AAPCS: define{{.*}} arm_aapcscc void @f25({{.*}} dead_on_unwind noalias writable sret
+// AAPCS: define{{.*}} arm_aapcscc void @f26({{.*}} dead_on_unwind noalias writable sret
+// AAPCS: define{{.*}} arm_aapcscc void @f27({{.*}} dead_on_unwind noalias writable sret
 _Complex char       f22(void) {}
 _Complex short      f23(void) {}
 _Complex int        f24(void) {}
@@ -150,8 +150,8 @@ struct s28 f28() {}
 struct s29 { _Complex short f0; };
 struct s29 f29() {}
 
-// APCS-GNU: define{{.*}} void @f30({{.*}} noalias sret
-// AAPCS: define{{.*}} arm_aapcscc void @f30({{.*}} noalias sret
+// APCS-GNU: define{{.*}} void @f30({{.*}} dead_on_unwind noalias writable sret
+// AAPCS: define{{.*}} arm_aapcscc void @f30({{.*}} dead_on_unwind noalias writable sret
 struct s30 { _Complex int f0; };
 struct s30 f30() {}
 
diff --git a/clang/test/CodeGen/arm-homogenous.c b/clang/test/CodeGen/arm-homogenous.c
index dbfb5fb284467..44a539598eb9b 100644
--- a/clang/test/CodeGen/arm-homogenous.c
+++ b/clang/test/CodeGen/arm-homogenous.c
@@ -27,7 +27,7 @@ void test_union_with_first_floats(void) {
 void test_return_union_with_first_floats(void) {
   g_u_f = returns_union_with_first_floats();
 }
-// CHECK: declare arm_aapcs_vfpcc void @returns_union_with_first_floats(ptr sret(%union.union_with_first_floats) align 4)
+// CHECK: declare arm_aapcs_vfpcc void @returns_union_with_first_floats(ptr dead_on_unwind writable sret(%union.union_with_first_floats) align 4)
 
 /* This is not a homogenous aggregate - fundamental types are different */
 typedef union {
@@ -47,7 +47,7 @@ void test_union_with_non_first_floats(void) {
 void test_return_union_with_non_first_floats(void) {
   g_u_nf_f = returns_union_with_non_first_floats();
 }
-// CHECK: declare arm_aapcs_vfpcc void @returns_union_with_non_first_floats(ptr sret(%union.union_with_non_first_floats) align 4)
+// CHECK: declare arm_aapcs_vfpcc void @returns_union_with_non_first_floats(ptr dead_on_unwind writable sret(%union.union_with_non_first_floats) align 4)
 
 /* This is not a homogenous aggregate - fundamental types are different */
 typedef struct {
@@ -67,7 +67,7 @@ void test_struct_with_union_with_first_floats(void) {
 void test_return_struct_with_union_with_first_floats(void) {
   g_s_f = returns_struct_with_union_with_first_floats();
 }
-// CHECK: declare arm_aapcs_vfpcc void @returns_struct_with_union_with_first_floats(ptr sret(%struct.struct_with_union_with_first_floats) align 4)
+// CHECK: declare arm_aapcs_vfpcc void @returns_struct_with_union_with_first_floats(ptr dead_on_unwind writable sret(%struct.struct_with_union_with_first_floats) align 4)
 
 /* This is not a homogenous aggregate - fundamental types are different */
 typedef struct {
@@ -87,7 +87,7 @@ void test_struct_with_union_with_non_first_floats(void) {
 void test_return_struct_with_union_with_non_first_floats(void) {
   g_s_nf_f = returns_struct_with_union_with_non_first_floats();
 }
-// CHECK: declare arm_aapcs_vfpcc void @returns_struct_with_union_with_non_first_floats(ptr sret(%struct.struct_with_union_with_non_first_floats) align 4)
+// CHECK: declare arm_aapcs_vfpcc void @returns_struct_with_union_with_non_first_floats(ptr dead_on_unwind writable sret(%struct.struct_with_union_with_non_first_floats) align 4)
 
 /* Plain array is not a homogenous aggregate */
 extern void takes_array_of_floats(float a[4]);
diff --git a/clang/test/CodeGen/arm-neon-vld.c b/clang/test/CodeGen/arm-neon-vld.c
index 697992edb5f6f..959c07cc2219b 100644
--- a/clang/test/CodeGen/arm-neon-vld.c
+++ b/clang/test/CodeGen/arm-neon-vld.c
@@ -11,7 +11,7 @@
 
 // CHECK-LABEL: @test_vld1_f16_x2(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.float16x4x2_t, align 8
-// CHECK-A32: ptr noalias sret(%struct.float16x4x2_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.float16x4x2_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.float16x4x2_t, align 8
 // CHECK: [[VLD1XN:%.*]] = call { <4 x [[HALF:half|i16]]>, <4 x [[HALF]]> } @llvm.{{aarch64.neon.ld1x2.v4f16.p0|arm.neon.vld1x2.v4i16.p0}}(ptr %a)
 // CHECK: store { <4 x [[HALF]]>, <4 x [[HALF]]> } [[VLD1XN]], ptr [[__RET]]
@@ -25,7 +25,7 @@ float16x4x2_t test_vld1_f16_x2(float16_t const *a) {
 
 // CHECK-LABEL: @test_vld1_f16_x3(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.float16x4x3_t, align 8
-// CHECK-A32: ptr noalias sret(%struct.float16x4x3_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.float16x4x3_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.float16x4x3_t, align 8
 // CHECK: [[VLD1XN:%.*]] = call { <4 x [[HALF:half|i16]]>, <4 x [[HALF]]>, <4 x [[HALF]]> } @llvm.{{aarch64.neon.ld1x3.v4f16.p0|arm.neon.vld1x3.v4i16.p0}}(ptr %a)
 // CHECK: store { <4 x [[HALF]]>, <4 x [[HALF]]>, <4 x [[HALF]]> } [[VLD1XN]], ptr [[__RET]]
@@ -39,7 +39,7 @@ float16x4x3_t test_vld1_f16_x3(float16_t const *a) {
 
 // CHECK-LABEL: @test_vld1_f16_x4(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.float16x4x4_t, align 8
-// CHECK-A32: ptr noalias sret(%struct.float16x4x4_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.float16x4x4_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.float16x4x4_t, align 8
 // CHECK: [[VLD1XN:%.*]] = call { <4 x [[HALF:half|i16]]>, <4 x [[HALF]]>, <4 x [[HALF]]>, <4 x [[HALF]]> } @llvm.{{aarch64.neon.ld1x4.v4f16.p0|arm.neon.vld1x4.v4i16.p0}}(ptr %a)
 // CHECK: store { <4 x [[HALF]]>, <4 x [[HALF]]>, <4 x [[HALF]]>, <4 x [[HALF]]> } [[VLD1XN]], ptr [[__RET]]
@@ -53,7 +53,7 @@ float16x4x4_t test_vld1_f16_x4(float16_t const *a) {
 
 // CHECK-LABEL: @test_vld1_f32_x2(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.float32x2x2_t, align 8
-// CHECK-A32: ptr noalias sret(%struct.float32x2x2_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.float32x2x2_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.float32x2x2_t, align 8
 // CHECK: [[VLD1XN:%.*]] = call { <2 x float>, <2 x float> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v2f32.p0(ptr %a)
 // CHECK: store { <2 x float>, <2 x float> } [[VLD1XN]], ptr [[__RET]]
@@ -67,7 +67,7 @@ float32x2x2_t test_vld1_f32_x2(float32_t const *a) {
 
 // CHECK-LABEL: @test_vld1_f32_x3(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.float32x2x3_t, align 8
-// CHECK-A32: ptr noalias sret(%struct.float32x2x3_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.float32x2x3_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.float32x2x3_t, align 8
 // CHECK: [[VLD1XN:%.*]] = call { <2 x float>, <2 x float>, <2 x float> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v2f32.p0(ptr %a)
 // CHECK: store { <2 x float>, <2 x float>, <2 x float> } [[VLD1XN]], ptr [[__RET]]
@@ -80,7 +80,7 @@ float32x2x3_t test_vld1_f32_x3(float32_t const *a) {
 
 // CHECK-LABEL: @test_vld1_f32_x4(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.float32x2x4_t, align 8
-// CHECK-A32: ptr noalias sret(%struct.float32x2x4_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.float32x2x4_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.float32x2x4_t, align 8
 // CHECK: [[VLD1XN:%.*]] = call { <2 x float>, <2 x float>, <2 x float>, <2 x float> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v2f32.p0(ptr %a)
 // CHECK: store { <2 x float>, <2 x float>, <2 x float>, <2 x float> } [[VLD1XN]], ptr [[__RET]]
@@ -94,7 +94,7 @@ float32x2x4_t test_vld1_f32_x4(float32_t const *a) {
 
 // CHECK-LABEL: @test_vld1_p16_x2(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.poly16x4x2_t, align 8
-// CHECK-A32: ptr noalias sret(%struct.poly16x4x2_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.poly16x4x2_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.poly16x4x2_t, align 8
 // CHECK: [[VLD1XN:%.*]] = call { <4 x i16>, <4 x i16> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v4i16.p0(ptr %a)
 // CHECK: store { <4 x i16>, <4 x i16> } [[VLD1XN]], ptr [[__RET]]
@@ -108,7 +108,7 @@ poly16x4x2_t test_vld1_p16_x2(poly16_t const *a) {
 
 // CHECK-LABEL: @test_vld1_p16_x3(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.poly16x4x3_t, align 8
-// CHECK-A32: ptr noalias sret(%struct.poly16x4x3_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.poly16x4x3_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.poly16x4x3_t, align 8
 // CHECK: [[VLD1XN:%.*]] = call { <4 x i16>, <4 x i16>, <4 x i16> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v4i16.p0(ptr %a)
 // CHECK: store { <4 x i16>, <4 x i16>, <4 x i16> } [[VLD1XN]], ptr [[__RET]]
@@ -122,7 +122,7 @@ poly16x4x3_t test_vld1_p16_x3(poly16_t const *a) {
 
 // CHECK-LABEL: @test_vld1_p16_x4(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.poly16x4x4_t, align 8
-// CHECK-A32: ptr noalias sret(%struct.poly16x4x4_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.poly16x4x4_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.poly16x4x4_t, align 8
 // CHECK: [[VLD1XN:%.*]] = call { <4 x i16>, <4 x i16>, <4 x i16>, <4 x i16> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v4i16.p0(ptr %a)
 // CHECK: store { <4 x i16>, <4 x i16>, <4 x i16>, <4 x i16> } [[VLD1XN]], ptr [[__RET]]
@@ -136,7 +136,7 @@ poly16x4x4_t test_vld1_p16_x4(poly16_t const *a) {
 
 // CHECK-LABEL: @test_vld1_p8_x2(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.poly8x8x2_t, align 8
-// CHECK-A32: ptr noalias sret(%struct.poly8x8x2_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.poly8x8x2_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.poly8x8x2_t, align 8
 // CHECK: [[VLD1XN:%.*]] = call { <8 x i8>, <8 x i8> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v8i8.p0(ptr %a)
 // CHECK: store { <8 x i8>, <8 x i8> } [[VLD1XN]], ptr [[__RET]]
@@ -150,7 +150,7 @@ poly8x8x2_t test_vld1_p8_x2(poly8_t const *a) {
 
 // CHECK-LABEL: @test_vld1_p8_x3(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.poly8x8x3_t, align 8
-// CHECK-A32: ptr noalias sret(%struct.poly8x8x3_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.poly8x8x3_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.poly8x8x3_t, align 8
 // CHECK: [[VLD1XN:%.*]] = call { <8 x i8>, <8 x i8>, <8 x i8> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v8i8.p0(ptr %a)
 // CHECK: store { <8 x i8>, <8 x i8>, <8 x i8> } [[VLD1XN]], ptr [[__RET]]
@@ -164,7 +164,7 @@ poly8x8x3_t test_vld1_p8_x3(poly8_t const *a) {
 
 // CHECK-LABEL: @test_vld1_p8_x4(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.poly8x8x4_t, align 8
-// CHECK-A32: ptr noalias sret(%struct.poly8x8x4_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.poly8x8x4_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.poly8x8x4_t, align 8
 // CHECK: [[VLD1XN:%.*]] = call { <8 x i8>, <8 x i8>, <8 x i8>, <8 x i8> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v8i8.p0(ptr %a)
 // CHECK: store { <8 x i8>, <8 x i8>, <8 x i8>, <8 x i8> } [[VLD1XN]], ptr [[__RET]]
@@ -178,7 +178,7 @@ poly8x8x4_t test_vld1_p8_x4(poly8_t const *a) {
 
 // CHECK-LABEL: @test_vld1_s16_x2(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int16x4x2_t, align 8
-// CHECK-A32: ptr noalias sret(%struct.int16x4x2_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.int16x4x2_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.int16x4x2_t, align 8
 // CHECK: [[VLD1XN:%.*]] = call { <4 x i16>, <4 x i16> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v4i16.p0(ptr %a)
 // CHECK: store { <4 x i16>, <4 x i16> } [[VLD1XN]], ptr [[__RET]]
@@ -192,7 +192,7 @@ int16x4x2_t test_vld1_s16_x2(int16_t const *a) {
 
 // CHECK-LABEL: @test_vld1_s16_x3(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int16x4x3_t, align 8
-// CHECK-A32: ptr noalias sret(%struct.int16x4x3_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.int16x4x3_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.int16x4x3_t, align 8
 // CHECK: [[VLD1XN:%.*]] = call { <4 x i16>, <4 x i16>, <4 x i16> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v4i16.p0(ptr %a)
 // CHECK: store { <4 x i16>, <4 x i16>, <4 x i16> } [[VLD1XN]], ptr [[__RET]]
@@ -206,7 +206,7 @@ int16x4x3_t test_vld1_s16_x3(int16_t const *a) {
 
 // CHECK-LABEL: @test_vld1_s16_x4(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int16x4x4_t, align 8
-// CHECK-A32: ptr noalias sret(%struct.int16x4x4_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.int16x4x4_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.int16x4x4_t, align 8
 // CHECK: [[VLD1XN:%.*]] = call { <4 x i16>, <4 x i16>, <4 x i16>, <4 x i16> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v4i16.p0(ptr %a)
 // CHECK: store { <4 x i16>, <4 x i16>, <4 x i16>, <4 x i16> } [[VLD1XN]], ptr [[__RET]]
@@ -220,7 +220,7 @@ int16x4x4_t test_vld1_s16_x4(int16_t const *a) {
 
 // CHECK-LABEL: @test_vld1_s32_x2(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int32x2x2_t, align 8
-// CHECK-A32: ptr noalias sret(%struct.int32x2x2_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.int32x2x2_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.int32x2x2_t, align 8
 // CHECK: [[VLD1XN:%.*]] = call { <2 x i32>, <2 x i32> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v2i32.p0(ptr %a)
 // CHECK: store { <2 x i32>, <2 x i32> } [[VLD1XN]], ptr [[__RET]]
@@ -234,7 +234,7 @@ int32x2x2_t test_vld1_s32_x2(int32_t const *a) {
 
 // CHECK-LABEL: @test_vld1_s32_x3(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int32x2x3_t, align 8
-// CHECK-A32: ptr noalias sret(%struct.int32x2x3_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.int32x2x3_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.int32x2x3_t, align 8
 // CHECK: [[VLD1XN:%.*]] = call { <2 x i32>, <2 x i32>, <2 x i32> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v2i32.p0(ptr %a)
 // CHECK: store { <2 x i32>, <2 x i32>, <2 x i32> } [[VLD1XN]], ptr [[__RET]]
@@ -248,7 +248,7 @@ int32x2x3_t test_vld1_s32_x3(int32_t const *a) {
 
 // CHECK-LABEL: @test_vld1_s32_x4(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int32x2x4_t, align 8
-// CHECK-A32: ptr noalias sret(%struct.int32x2x4_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.int32x2x4_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.int32x2x4_t, align 8
 // CHECK: [[VLD1XN:%.*]] = call { <2 x i32>, <2 x i32>, <2 x i32>, <2 x i32> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v2i32.p0(ptr %a)
 // CHECK: store { <2 x i32>, <2 x i32>, <2 x i32>, <2 x i32> } [[VLD1XN]], ptr [[__RET]]
@@ -262,7 +262,7 @@ int32x2x4_t test_vld1_s32_x4(int32_t const *a) {
 
 // CHECK-LABEL: @test_vld1_s64_x2(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int64x1x2_t, align 8
-// CHECK-A32: ptr noalias sret(%struct.int64x1x2_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.int64x1x2_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.int64x1x2_t, align 8
 // CHECK: [[VLD1XN:%.*]] = call { <1 x i64>, <1 x i64> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v1i64.p0(ptr %a)
 // CHECK: store { <1 x i64>, <1 x i64> } [[VLD1XN]], ptr [[__RET]]
@@ -276,7 +276,7 @@ int64x1x2_t test_vld1_s64_x2(int64_t const *a) {
 
 // CHECK-LABEL: @test_vld1_s64_x3(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int64x1x3_t, align 8
-// CHECK-A32: ptr noalias sret(%struct.int64x1x3_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.int64x1x3_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.int64x1x3_t, align 8
 // CHECK: [[VLD1XN:%.*]] = call { <1 x i64>, <1 x i64>, <1 x i64> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v1i64.p0(ptr %a)
 // CHECK: store { <1 x i64>, <1 x i64>, <1 x i64> } [[VLD1XN]], ptr [[__RET]]
@@ -290,7 +290,7 @@ int64x1x3_t test_vld1_s64_x3(int64_t const *a) {
 
 // CHECK-LABEL: @test_vld1_s64_x4(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int64x1x4_t, align 8
-// CHECK-A32: ptr noalias sret(%struct.int64x1x4_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.int64x1x4_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.int64x1x4_t, align 8
 // CHECK: [[VLD1XN:%.*]] = call { <1 x i64>, <1 x i64>, <1 x i64>, <1 x i64> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v1i64.p0(ptr %a)
 // CHECK: store { <1 x i64>, <1 x i64>, <1 x i64>, <1 x i64> } [[VLD1XN]], ptr [[__RET]]
@@ -304,7 +304,7 @@ int64x1x4_t test_vld1_s64_x4(int64_t const *a) {
 
 // CHECK-LABEL: @test_vld1_s8_x2(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int8x8x2_t, align 8
-// CHECK-A32: ptr noalias sret(%struct.int8x8x2_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.int8x8x2_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.int8x8x2_t, align 8
 // CHECK: [[VLD1XN:%.*]] = call { <8 x i8>, <8 x i8> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v8i8.p0(ptr %a)
 // CHECK: store { <8 x i8>, <8 x i8> } [[VLD1XN]], ptr [[__RET]]
@@ -318,7 +318,7 @@ int8x8x2_t test_vld1_s8_x2(int8_t const *a) {
 
 // CHECK-LABEL: @test_vld1_s8_x3(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int8x8x3_t, align 8
-// CHECK-A32: ptr noalias sret(%struct.int8x8x3_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.int8x8x3_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.int8x8x3_t, align 8
 // CHECK: [[VLD1XN:%.*]] = call { <8 x i8>, <8 x i8>, <8 x i8> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v8i8.p0(ptr %a)
 // CHECK: store { <8 x i8>, <8 x i8>, <8 x i8> } [[VLD1XN]], ptr [[__RET]]
@@ -332,7 +332,7 @@ int8x8x3_t test_vld1_s8_x3(int8_t const *a) {
 
 // CHECK-LABEL: @test_vld1_s8_x4(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int8x8x4_t, align 8
-// CHECK-A32: ptr noalias sret(%struct.int8x8x4_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.int8x8x4_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.int8x8x4_t, align 8
 // CHECK: [[VLD1XN:%.*]] = call { <8 x i8>, <8 x i8>, <8 x i8>, <8 x i8> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v8i8.p0(ptr %a)
 // CHECK: store { <8 x i8>, <8 x i8>, <8 x i8>, <8 x i8> } [[VLD1XN]], ptr [[__RET]]
@@ -346,7 +346,7 @@ int8x8x4_t test_vld1_s8_x4(int8_t const *a) {
 
 // CHECK-LABEL: @test_vld1_u16_x2(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint16x4x2_t, align 8
-// CHECK-A32: ptr noalias sret(%struct.uint16x4x2_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.uint16x4x2_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.uint16x4x2_t, align 8
 // CHECK: [[VLD1XN:%.*]] = call { <4 x i16>, <4 x i16> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v4i16.p0(ptr %a)
 // CHECK: store { <4 x i16>, <4 x i16> } [[VLD1XN]], ptr [[__RET]]
@@ -360,7 +360,7 @@ uint16x4x2_t test_vld1_u16_x2(uint16_t const *a) {
 
 // CHECK-LABEL: @test_vld1_u16_x3(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint16x4x3_t, align 8
-// CHECK-A32: ptr noalias sret(%struct.uint16x4x3_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.uint16x4x3_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.uint16x4x3_t, align 8
 // CHECK: [[VLD1XN:%.*]] = call { <4 x i16>, <4 x i16>, <4 x i16> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v4i16.p0(ptr %a)
 // CHECK: store { <4 x i16>, <4 x i16>, <4 x i16> } [[VLD1XN]], ptr [[__RET]]
@@ -374,7 +374,7 @@ uint16x4x3_t test_vld1_u16_x3(uint16_t const *a) {
 
 // CHECK-LABEL: @test_vld1_u16_x4(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint16x4x4_t, align 8
-// CHECK-A32: ptr noalias sret(%struct.uint16x4x4_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.uint16x4x4_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.uint16x4x4_t, align 8
 // CHECK: [[VLD1XN:%.*]] = call { <4 x i16>, <4 x i16>, <4 x i16>, <4 x i16> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v4i16.p0(ptr %a)
 // CHECK: store { <4 x i16>, <4 x i16>, <4 x i16>, <4 x i16> } [[VLD1XN]], ptr [[__RET]]
@@ -388,7 +388,7 @@ uint16x4x4_t test_vld1_u16_x4(uint16_t const *a) {
 
 // CHECK-LABEL: @test_vld1_u32_x2(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint32x2x2_t, align 8
-// CHECK-A32: ptr noalias sret(%struct.uint32x2x2_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.uint32x2x2_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.uint32x2x2_t, align 8
 // CHECK: [[VLD1XN:%.*]] = call { <2 x i32>, <2 x i32> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v2i32.p0(ptr %a)
 // CHECK: store { <2 x i32>, <2 x i32> } [[VLD1XN]], ptr [[__RET]]
@@ -402,7 +402,7 @@ uint32x2x2_t test_vld1_u32_x2(uint32_t const *a) {
 
 // CHECK-LABEL: @test_vld1_u32_x3(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint32x2x3_t, align 8
-// CHECK-A32: ptr noalias sret(%struct.uint32x2x3_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.uint32x2x3_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.uint32x2x3_t, align 8
 // CHECK: [[VLD1XN:%.*]] = call { <2 x i32>, <2 x i32>, <2 x i32> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v2i32.p0(ptr %a)
 // CHECK: store { <2 x i32>, <2 x i32>, <2 x i32> } [[VLD1XN]], ptr [[__RET]]
@@ -416,7 +416,7 @@ uint32x2x3_t test_vld1_u32_x3(uint32_t const *a) {
 
 // CHECK-LABEL: @test_vld1_u32_x4(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint32x2x4_t, align 8
-// CHECK-A32: ptr noalias sret(%struct.uint32x2x4_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.uint32x2x4_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.uint32x2x4_t, align 8
 // CHECK: [[VLD1XN:%.*]] = call { <2 x i32>, <2 x i32>, <2 x i32>, <2 x i32> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v2i32.p0(ptr %a)
 // CHECK: store { <2 x i32>, <2 x i32>, <2 x i32>, <2 x i32> } [[VLD1XN]], ptr [[__RET]]
@@ -430,7 +430,7 @@ uint32x2x4_t test_vld1_u32_x4(uint32_t const *a) {
 
 // CHECK-LABEL: @test_vld1_u64_x2(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint64x1x2_t, align 8
-// CHECK-A32: ptr noalias sret(%struct.uint64x1x2_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.uint64x1x2_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.uint64x1x2_t, align 8
 // CHECK: [[VLD1XN:%.*]] = call { <1 x i64>, <1 x i64> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v1i64.p0(ptr %a)
 // CHECK: store { <1 x i64>, <1 x i64> } [[VLD1XN]], ptr [[__RET]]
@@ -444,7 +444,7 @@ uint64x1x2_t test_vld1_u64_x2(uint64_t const *a) {
 
 // CHECK-LABEL: @test_vld1_u64_x3(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint64x1x3_t, align 8
-// CHECK-A32: ptr noalias sret(%struct.uint64x1x3_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.uint64x1x3_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.uint64x1x3_t, align 8
 // CHECK: [[VLD1XN:%.*]] = call { <1 x i64>, <1 x i64>, <1 x i64> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v1i64.p0(ptr %a)
 // CHECK: store { <1 x i64>, <1 x i64>, <1 x i64> } [[VLD1XN]], ptr [[__RET]]
@@ -458,7 +458,7 @@ uint64x1x3_t test_vld1_u64_x3(uint64_t const *a) {
 
 // CHECK-LABEL: @test_vld1_u64_x4(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint64x1x4_t, align 8
-// CHECK-A32: ptr noalias sret(%struct.uint64x1x4_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.uint64x1x4_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.uint64x1x4_t, align 8
 // CHECK: [[VLD1XN:%.*]] = call { <1 x i64>, <1 x i64>, <1 x i64>, <1 x i64> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v1i64.p0(ptr %a)
 // CHECK: store { <1 x i64>, <1 x i64>, <1 x i64>, <1 x i64> } [[VLD1XN]], ptr [[__RET]]
@@ -472,7 +472,7 @@ uint64x1x4_t test_vld1_u64_x4(uint64_t const *a) {
 
 // CHECK-LABEL: @test_vld1_u8_x2(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint8x8x2_t, align 8
-// CHECK-A32: ptr noalias sret(%struct.uint8x8x2_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.uint8x8x2_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.uint8x8x2_t, align 8
 // CHECK: [[VLD1XN:%.*]] = call { <8 x i8>, <8 x i8> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v8i8.p0(ptr %a)
 // CHECK: store { <8 x i8>, <8 x i8> } [[VLD1XN]], ptr [[__RET]]
@@ -486,7 +486,7 @@ uint8x8x2_t test_vld1_u8_x2(uint8_t const *a) {
 
 // CHECK-LABEL: @test_vld1_u8_x3(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint8x8x3_t, align 8
-// CHECK-A32: ptr noalias sret(%struct.uint8x8x3_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.uint8x8x3_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.uint8x8x3_t, align 8
 // CHECK: [[VLD1XN:%.*]] = call { <8 x i8>, <8 x i8>, <8 x i8> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v8i8.p0(ptr %a)
 // CHECK: store { <8 x i8>, <8 x i8>, <8 x i8> } [[VLD1XN]], ptr [[__RET]]
@@ -500,7 +500,7 @@ uint8x8x3_t test_vld1_u8_x3(uint8_t const *a) {
 
 // CHECK-LABEL: @test_vld1_u8_x4(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint8x8x4_t, align 8
-// CHECK-A32: ptr noalias sret(%struct.uint8x8x4_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.uint8x8x4_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.uint8x8x4_t, align 8
 // CHECK: [[VLD1XN:%.*]] = call { <8 x i8>, <8 x i8>, <8 x i8>, <8 x i8> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v8i8.p0(ptr %a)
 // CHECK: store { <8 x i8>, <8 x i8>, <8 x i8>, <8 x i8> } [[VLD1XN]], ptr [[__RET]]
@@ -514,7 +514,7 @@ uint8x8x4_t test_vld1_u8_x4(uint8_t const *a) {
 
 // CHECK-LABEL: @test_vld1q_f16_x2(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.float16x8x2_t, align 16
-// CHECK-A32: ptr noalias sret(%struct.float16x8x2_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.float16x8x2_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.float16x8x2_t, align {{16|8}}
 // CHECK: [[VLD1XN:%.*]] = call { <8 x [[HALF:half|i16]]>, <8 x [[HALF]]> } @llvm.{{aarch64.neon.ld1x2.v8f16.p0|arm.neon.vld1x2.v8i16.p0}}(ptr %a)
 // CHECK: store { <8 x [[HALF]]>, <8 x [[HALF]]> } [[VLD1XN]], ptr [[__RET]]
@@ -528,7 +528,7 @@ float16x8x2_t test_vld1q_f16_x2(float16_t const *a) {
 
 // CHECK-LABEL: @test_vld1q_f16_x3(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.float16x8x3_t, align 16
-// CHECK-A32: ptr noalias sret(%struct.float16x8x3_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.float16x8x3_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.float16x8x3_t, align {{16|8}}
 // CHECK: [[VLD1XN:%.*]] = call { <8 x [[HALF:half|i16]]>, <8 x [[HALF]]>, <8 x [[HALF]]> } @llvm.{{aarch64.neon.ld1x3.v8f16.p0|arm.neon.vld1x3.v8i16.p0}}(ptr %a)
 // CHECK: store { <8 x [[HALF]]>, <8 x [[HALF]]>, <8 x [[HALF]]> } [[VLD1XN]], ptr [[__RET]]
@@ -542,7 +542,7 @@ float16x8x3_t test_vld1q_f16_x3(float16_t const *a) {
 
 // CHECK-LABEL: @test_vld1q_f16_x4(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.float16x8x4_t, align 16
-// CHECK-A32: ptr noalias sret(%struct.float16x8x4_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.float16x8x4_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.float16x8x4_t, align {{16|8}}
 // CHECK: [[VLD1XN:%.*]] = call { <8 x [[HALF:half|i16]]>, <8 x [[HALF]]>, <8 x [[HALF]]>, <8 x [[HALF]]> } @llvm.{{aarch64.neon.ld1x4.v8f16.p0|arm.neon.vld1x4.v8i16.p0}}(ptr %a)
 // CHECK: store { <8 x [[HALF]]>, <8 x [[HALF]]>, <8 x [[HALF]]>, <8 x [[HALF]]> } [[VLD1XN]], ptr [[__RET]]
@@ -556,7 +556,7 @@ float16x8x4_t test_vld1q_f16_x4(float16_t const *a) {
 
 // CHECK-LABEL: @test_vld1q_f32_x2(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.float32x4x2_t, align 16
-// CHECK-A32: ptr noalias sret(%struct.float32x4x2_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.float32x4x2_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.float32x4x2_t, align {{16|8}}
 // CHECK: [[VLD1XN:%.*]] = call { <4 x float>, <4 x float> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v4f32.p0(ptr %a)
 // CHECK: store { <4 x float>, <4 x float> } [[VLD1XN]], ptr [[__RET]]
@@ -570,7 +570,7 @@ float32x4x2_t test_vld1q_f32_x2(float32_t const *a) {
 
 // CHECK-LABEL: @test_vld1q_f32_x3(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.float32x4x3_t, align 16
-// CHECK-A32: ptr noalias sret(%struct.float32x4x3_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.float32x4x3_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.float32x4x3_t, align {{16|8}}
 // CHECK: [[VLD1XN:%.*]] = call { <4 x float>, <4 x float>, <4 x float> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v4f32.p0(ptr %a)
 // CHECK: store { <4 x float>, <4 x float>, <4 x float> } [[VLD1XN]], ptr [[__RET]]
@@ -584,7 +584,7 @@ float32x4x3_t test_vld1q_f32_x3(float32_t const *a) {
 
 // CHECK-LABEL: @test_vld1q_f32_x4(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.float32x4x4_t, align 16
-// CHECK-A32: ptr noalias sret(%struct.float32x4x4_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.float32x4x4_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.float32x4x4_t, align {{16|8}}
 // CHECK: [[VLD1XN:%.*]] = call { <4 x float>, <4 x float>, <4 x float>, <4 x float> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v4f32.p0(ptr %a)
 // CHECK: store { <4 x float>, <4 x float>, <4 x float>, <4 x float> } [[VLD1XN]], ptr [[__RET]]
@@ -598,7 +598,7 @@ float32x4x4_t test_vld1q_f32_x4(float32_t const *a) {
 
 // CHECK-LABEL: @test_vld1q_p16_x2(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.poly16x8x2_t, align 16
-// CHECK-A32: ptr noalias sret(%struct.poly16x8x2_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.poly16x8x2_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.poly16x8x2_t, align {{16|8}}
 // CHECK: [[VLD1XN:%.*]] = call { <8 x i16>, <8 x i16> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v8i16.p0(ptr %a)
 // CHECK: store { <8 x i16>, <8 x i16> } [[VLD1XN]], ptr [[__RET]]
@@ -612,7 +612,7 @@ poly16x8x2_t test_vld1q_p16_x2(poly16_t const *a) {
 
 // CHECK-LABEL: @test_vld1q_p16_x3(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.poly16x8x3_t, align 16
-// CHECK-A32: ptr noalias sret(%struct.poly16x8x3_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.poly16x8x3_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.poly16x8x3_t, align {{16|8}}
 // CHECK: [[VLD1XN:%.*]] = call { <8 x i16>, <8 x i16>, <8 x i16> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v8i16.p0(ptr %a)
 // CHECK: store { <8 x i16>, <8 x i16>, <8 x i16> } [[VLD1XN]], ptr [[__RET]]
@@ -626,7 +626,7 @@ poly16x8x3_t test_vld1q_p16_x3(poly16_t const *a) {
 
 // CHECK-LABEL: @test_vld1q_p16_x4(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.poly16x8x4_t, align 16
-// CHECK-A32: ptr noalias sret(%struct.poly16x8x4_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.poly16x8x4_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.poly16x8x4_t, align {{16|8}}
 // CHECK: [[VLD1XN:%.*]] = call { <8 x i16>, <8 x i16>, <8 x i16>, <8 x i16> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v8i16.p0(ptr %a)
 // CHECK: store { <8 x i16>, <8 x i16>, <8 x i16>, <8 x i16> } [[VLD1XN]], ptr [[__RET]]
@@ -640,7 +640,7 @@ poly16x8x4_t test_vld1q_p16_x4(poly16_t const *a) {
 
 // CHECK-LABEL: @test_vld1q_p8_x2(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.poly8x16x2_t, align 16
-// CHECK-A32: ptr noalias sret(%struct.poly8x16x2_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.poly8x16x2_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.poly8x16x2_t, align {{16|8}}
 // CHECK: [[VLD1XN:%.*]] = call { <16 x i8>, <16 x i8> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v16i8.p0(ptr %a)
 // CHECK: store { <16 x i8>, <16 x i8> } [[VLD1XN]], ptr [[__RET]]
@@ -654,7 +654,7 @@ poly8x16x2_t test_vld1q_p8_x2(poly8_t const *a) {
 
 // CHECK-LABEL: @test_vld1q_p8_x3(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.poly8x16x3_t, align 16
-// CHECK-A32: ptr noalias sret(%struct.poly8x16x3_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.poly8x16x3_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.poly8x16x3_t, align {{16|8}}
 // CHECK: [[VLD1XN:%.*]] = call { <16 x i8>, <16 x i8>, <16 x i8> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v16i8.p0(ptr %a)
 // CHECK: store { <16 x i8>, <16 x i8>, <16 x i8> } [[VLD1XN]], ptr [[__RET]]
@@ -668,7 +668,7 @@ poly8x16x3_t test_vld1q_p8_x3(poly8_t const *a) {
 
 // CHECK-LABEL: @test_vld1q_p8_x4(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.poly8x16x4_t, align 16
-// CHECK-A32: ptr noalias sret(%struct.poly8x16x4_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.poly8x16x4_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.poly8x16x4_t, align {{16|8}}
 // CHECK: [[VLD1XN:%.*]] = call { <16 x i8>, <16 x i8>, <16 x i8>, <16 x i8> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v16i8.p0(ptr %a)
 // CHECK: store { <16 x i8>, <16 x i8>, <16 x i8>, <16 x i8> } [[VLD1XN]], ptr [[__RET]]
@@ -682,7 +682,7 @@ poly8x16x4_t test_vld1q_p8_x4(poly8_t const *a) {
 
 // CHECK-LABEL: @test_vld1q_s16_x2(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int16x8x2_t, align 16
-// CHECK-A32: ptr noalias sret(%struct.int16x8x2_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.int16x8x2_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.int16x8x2_t, align {{16|8}}
 // CHECK: [[VLD1XN:%.*]] = call { <8 x i16>, <8 x i16> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v8i16.p0(ptr %a)
 // CHECK: store { <8 x i16>, <8 x i16> } [[VLD1XN]], ptr [[__RET]]
@@ -696,7 +696,7 @@ int16x8x2_t test_vld1q_s16_x2(int16_t const *a) {
 
 // CHECK-LABEL: @test_vld1q_s16_x3(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int16x8x3_t, align 16
-// CHECK-A32: ptr noalias sret(%struct.int16x8x3_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.int16x8x3_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.int16x8x3_t, align {{16|8}}
 // CHECK: [[VLD1XN:%.*]] = call { <8 x i16>, <8 x i16>, <8 x i16> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v8i16.p0(ptr %a)
 // CHECK: store { <8 x i16>, <8 x i16>, <8 x i16> } [[VLD1XN]], ptr [[__RET]]
@@ -710,7 +710,7 @@ int16x8x3_t test_vld1q_s16_x3(int16_t const *a) {
 
 // CHECK-LABEL: @test_vld1q_s16_x4(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int16x8x4_t, align 16
-// CHECK-A32: ptr noalias sret(%struct.int16x8x4_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.int16x8x4_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.int16x8x4_t, align {{16|8}}
 // CHECK: [[VLD1XN:%.*]] = call { <8 x i16>, <8 x i16>, <8 x i16>, <8 x i16> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v8i16.p0(ptr %a)
 // CHECK: store { <8 x i16>, <8 x i16>, <8 x i16>, <8 x i16> } [[VLD1XN]], ptr [[__RET]]
@@ -724,7 +724,7 @@ int16x8x4_t test_vld1q_s16_x4(int16_t const *a) {
 
 // CHECK-LABEL: @test_vld1q_s32_x2(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int32x4x2_t, align 16
-// CHECK-A32: ptr noalias sret(%struct.int32x4x2_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.int32x4x2_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.int32x4x2_t, align {{16|8}}
 // CHECK: [[VLD1XN:%.*]] = call { <4 x i32>, <4 x i32> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v4i32.p0(ptr %a)
 // CHECK: store { <4 x i32>, <4 x i32> } [[VLD1XN]], ptr [[__RET]]
@@ -738,7 +738,7 @@ int32x4x2_t test_vld1q_s32_x2(int32_t const *a) {
 
 // CHECK-LABEL: @test_vld1q_s32_x3(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int32x4x3_t, align 16
-// CHECK-A32: ptr noalias sret(%struct.int32x4x3_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.int32x4x3_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.int32x4x3_t, align {{16|8}}
 // CHECK: [[VLD1XN:%.*]] = call { <4 x i32>, <4 x i32>, <4 x i32> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v4i32.p0(ptr %a)
 // CHECK: store { <4 x i32>, <4 x i32>, <4 x i32> } [[VLD1XN]], ptr [[__RET]]
@@ -752,7 +752,7 @@ int32x4x3_t test_vld1q_s32_x3(int32_t const *a) {
 
 // CHECK-LABEL: @test_vld1q_s32_x4(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int32x4x4_t, align 16
-// CHECK-A32: ptr noalias sret(%struct.int32x4x4_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.int32x4x4_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.int32x4x4_t, align {{16|8}}
 // CHECK: [[VLD1XN:%.*]] = call { <4 x i32>, <4 x i32>, <4 x i32>, <4 x i32> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v4i32.p0(ptr %a)
 // CHECK: store { <4 x i32>, <4 x i32>, <4 x i32>, <4 x i32> } [[VLD1XN]], ptr [[__RET]]
@@ -766,7 +766,7 @@ int32x4x4_t test_vld1q_s32_x4(int32_t const *a) {
 
 // CHECK-LABEL: @test_vld1q_s64_x2(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int64x2x2_t, align 16
-// CHECK-A32: ptr noalias sret(%struct.int64x2x2_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.int64x2x2_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.int64x2x2_t, align {{16|8}}
 // CHECK: [[VLD1XN:%.*]] = call { <2 x i64>, <2 x i64> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v2i64.p0(ptr %a)
 // CHECK: store { <2 x i64>, <2 x i64> } [[VLD1XN]], ptr [[__RET]]
@@ -780,7 +780,7 @@ int64x2x2_t test_vld1q_s64_x2(int64_t const *a) {
 
 // CHECK-LABEL: @test_vld1q_s64_x3(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int64x2x3_t, align 16
-// CHECK-A32: ptr noalias sret(%struct.int64x2x3_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.int64x2x3_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.int64x2x3_t, align {{16|8}}
 // CHECK: [[VLD1XN:%.*]] = call { <2 x i64>, <2 x i64>, <2 x i64> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v2i64.p0(ptr %a)
 // CHECK: store { <2 x i64>, <2 x i64>, <2 x i64> } [[VLD1XN]], ptr [[__RET]]
@@ -794,7 +794,7 @@ int64x2x3_t test_vld1q_s64_x3(int64_t const *a) {
 
 // CHECK-LABEL: @test_vld1q_s64_x4(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int64x2x4_t, align 16
-// CHECK-A32: ptr noalias sret(%struct.int64x2x4_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.int64x2x4_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.int64x2x4_t, align {{16|8}}
 // CHECK: [[VLD1XN:%.*]] = call { <2 x i64>, <2 x i64>, <2 x i64>, <2 x i64> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v2i64.p0(ptr %a)
 // CHECK: store { <2 x i64>, <2 x i64>, <2 x i64>, <2 x i64> } [[VLD1XN]], ptr [[__RET]]
@@ -808,7 +808,7 @@ int64x2x4_t test_vld1q_s64_x4(int64_t const *a) {
 
 // CHECK-LABEL: @test_vld1q_s8_x2(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int8x16x2_t, align 16
-// CHECK-A32: ptr noalias sret(%struct.int8x16x2_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.int8x16x2_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.int8x16x2_t, align {{16|8}}
 // CHECK: [[VLD1XN:%.*]] = call { <16 x i8>, <16 x i8> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v16i8.p0(ptr %a)
 // CHECK: store { <16 x i8>, <16 x i8> } [[VLD1XN]], ptr [[__RET]]
@@ -822,7 +822,7 @@ int8x16x2_t test_vld1q_s8_x2(int8_t const *a) {
 
 // CHECK-LABEL: @test_vld1q_s8_x3(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int8x16x3_t, align 16
-// CHECK-A32: ptr noalias sret(%struct.int8x16x3_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.int8x16x3_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.int8x16x3_t, align {{16|8}}
 // CHECK: [[VLD1XN:%.*]] = call { <16 x i8>, <16 x i8>, <16 x i8> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v16i8.p0(ptr %a)
 // CHECK: store { <16 x i8>, <16 x i8>, <16 x i8> } [[VLD1XN]], ptr [[__RET]]
@@ -836,7 +836,7 @@ int8x16x3_t test_vld1q_s8_x3(int8_t const *a) {
 
 // CHECK-LABEL: @test_vld1q_s8_x4(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int8x16x4_t, align 16
-// CHECK-A32: ptr noalias sret(%struct.int8x16x4_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.int8x16x4_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.int8x16x4_t, align {{16|8}}
 // CHECK: [[VLD1XN:%.*]] = call { <16 x i8>, <16 x i8>, <16 x i8>, <16 x i8> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v16i8.p0(ptr %a)
 // CHECK: store { <16 x i8>, <16 x i8>, <16 x i8>, <16 x i8> } [[VLD1XN]], ptr [[__RET]]
@@ -850,7 +850,7 @@ int8x16x4_t test_vld1q_s8_x4(int8_t const *a) {
 
 // CHECK-LABEL: @test_vld1q_u16_x2(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint16x8x2_t, align 16
-// CHECK-A32: ptr noalias sret(%struct.uint16x8x2_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.uint16x8x2_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.uint16x8x2_t, align {{16|8}}
 // CHECK: [[VLD1XN:%.*]] = call { <8 x i16>, <8 x i16> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v8i16.p0(ptr %a)
 // CHECK: store { <8 x i16>, <8 x i16> } [[VLD1XN]], ptr [[__RET]]
@@ -864,7 +864,7 @@ uint16x8x2_t test_vld1q_u16_x2(uint16_t const *a) {
 
 // CHECK-LABEL: @test_vld1q_u16_x3(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint16x8x3_t, align 16
-// CHECK-A32: ptr noalias sret(%struct.uint16x8x3_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.uint16x8x3_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.uint16x8x3_t, align {{16|8}}
 // CHECK: [[VLD1XN:%.*]] = call { <8 x i16>, <8 x i16>, <8 x i16> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v8i16.p0(ptr %a)
 // CHECK: store { <8 x i16>, <8 x i16>, <8 x i16> } [[VLD1XN]], ptr [[__RET]]
@@ -878,7 +878,7 @@ uint16x8x3_t test_vld1q_u16_x3(uint16_t const *a) {
 
 // CHECK-LABEL: @test_vld1q_u16_x4(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint16x8x4_t, align 16
-// CHECK-A32: ptr noalias sret(%struct.uint16x8x4_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.uint16x8x4_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.uint16x8x4_t, align {{16|8}}
 // CHECK: [[VLD1XN:%.*]] = call { <8 x i16>, <8 x i16>, <8 x i16>, <8 x i16> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v8i16.p0(ptr %a)
 // CHECK: store { <8 x i16>, <8 x i16>, <8 x i16>, <8 x i16> } [[VLD1XN]], ptr [[__RET]]
@@ -892,7 +892,7 @@ uint16x8x4_t test_vld1q_u16_x4(uint16_t const *a) {
 
 // CHECK-LABEL: @test_vld1q_u32_x2(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint32x4x2_t, align 16
-// CHECK-A32: ptr noalias sret(%struct.uint32x4x2_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.uint32x4x2_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.uint32x4x2_t, align {{16|8}}
 // CHECK: [[VLD1XN:%.*]] = call { <4 x i32>, <4 x i32> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v4i32.p0(ptr %a)
 // CHECK: store { <4 x i32>, <4 x i32> } [[VLD1XN]], ptr [[__RET]]
@@ -906,7 +906,7 @@ uint32x4x2_t test_vld1q_u32_x2(uint32_t const *a) {
 
 // CHECK-LABEL: @test_vld1q_u32_x3(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint32x4x3_t, align 16
-// CHECK-A32: ptr noalias sret(%struct.uint32x4x3_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.uint32x4x3_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.uint32x4x3_t, align {{16|8}}
 // CHECK: [[VLD1XN:%.*]] = call { <4 x i32>, <4 x i32>, <4 x i32> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v4i32.p0(ptr %a)
 // CHECK: store { <4 x i32>, <4 x i32>, <4 x i32> } [[VLD1XN]], ptr [[__RET]]
@@ -920,7 +920,7 @@ uint32x4x3_t test_vld1q_u32_x3(uint32_t const *a) {
 
 // CHECK-LABEL: @test_vld1q_u32_x4(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint32x4x4_t, align 16
-// CHECK-A32: ptr noalias sret(%struct.uint32x4x4_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.uint32x4x4_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.uint32x4x4_t, align {{16|8}}
 // CHECK: [[VLD1XN:%.*]] = call { <4 x i32>, <4 x i32>, <4 x i32>, <4 x i32> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v4i32.p0(ptr %a)
 // CHECK: store { <4 x i32>, <4 x i32>, <4 x i32>, <4 x i32> } [[VLD1XN]], ptr [[__RET]]
@@ -934,7 +934,7 @@ uint32x4x4_t test_vld1q_u32_x4(uint32_t const *a) {
 
 // CHECK-LABEL: @test_vld1q_u64_x2(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint64x2x2_t, align 16
-// CHECK-A32: ptr noalias sret(%struct.uint64x2x2_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.uint64x2x2_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.uint64x2x2_t, align {{16|8}}
 // CHECK: [[VLD1XN:%.*]] = call { <2 x i64>, <2 x i64> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v2i64.p0(ptr %a)
 // CHECK: store { <2 x i64>, <2 x i64> } [[VLD1XN]], ptr [[__RET]]
@@ -948,7 +948,7 @@ uint64x2x2_t test_vld1q_u64_x2(uint64_t const *a) {
 
 // CHECK-LABEL: @test_vld1q_u64_x3(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint64x2x3_t, align 16
-// CHECK-A32: ptr noalias sret(%struct.uint64x2x3_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.uint64x2x3_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.uint64x2x3_t, align {{16|8}}
 // CHECK: [[VLD1XN:%.*]] = call { <2 x i64>, <2 x i64>, <2 x i64> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v2i64.p0(ptr %a)
 // CHECK: store { <2 x i64>, <2 x i64>, <2 x i64> } [[VLD1XN]], ptr [[__RET]]
@@ -962,7 +962,7 @@ uint64x2x3_t test_vld1q_u64_x3(uint64_t const *a) {
 
 // CHECK-LABEL: @test_vld1q_u64_x4(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint64x2x4_t, align 16
-// CHECK-A32: ptr noalias sret(%struct.uint64x2x4_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.uint64x2x4_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.uint64x2x4_t, align {{16|8}}
 // CHECK: [[VLD1XN:%.*]] = call { <2 x i64>, <2 x i64>, <2 x i64>, <2 x i64> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v2i64.p0(ptr %a)
 // CHECK: store { <2 x i64>, <2 x i64>, <2 x i64>, <2 x i64> } [[VLD1XN]], ptr [[__RET]]
@@ -976,7 +976,7 @@ uint64x2x4_t test_vld1q_u64_x4(uint64_t const *a) {
 
 // CHECK-LABEL: @test_vld1q_u8_x2(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint8x16x2_t, align 16
-// CHECK-A32: ptr noalias sret(%struct.uint8x16x2_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.uint8x16x2_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.uint8x16x2_t, align {{16|8}}
 // CHECK: [[VLD1XN:%.*]] = call { <16 x i8>, <16 x i8> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v16i8.p0(ptr %a)
 // CHECK: store { <16 x i8>, <16 x i8> } [[VLD1XN]], ptr [[__RET]]
@@ -990,7 +990,7 @@ uint8x16x2_t test_vld1q_u8_x2(uint8_t const *a) {
 
 // CHECK-LABEL: @test_vld1q_u8_x3(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint8x16x3_t, align 16
-// CHECK-A32: ptr noalias sret(%struct.uint8x16x3_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.uint8x16x3_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.uint8x16x3_t, align {{16|8}}
 // CHECK: [[VLD1XN:%.*]] = call { <16 x i8>, <16 x i8>, <16 x i8> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v16i8.p0(ptr %a)
 // CHECK: store { <16 x i8>, <16 x i8>, <16 x i8> } [[VLD1XN]], ptr [[__RET]]
@@ -1004,7 +1004,7 @@ uint8x16x3_t test_vld1q_u8_x3(uint8_t const *a) {
 
 // CHECK-LABEL: @test_vld1q_u8_x4(
 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint8x16x4_t, align 16
-// CHECK-A32: ptr noalias sret(%struct.uint8x16x4_t) align 8 [[RETVAL:%.*]],
+// CHECK-A32: ptr dead_on_unwind noalias writable sret(%struct.uint8x16x4_t) align 8 [[RETVAL:%.*]],
 // CHECK: [[__RET:%.*]] = alloca %struct.uint8x16x4_t, align {{16|8}}
 // CHECK: [[VLD1XN:%.*]] = call { <16 x i8>, <16 x i8>, <16 x i8>, <16 x i8> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v16i8.p0(ptr %a)
 // CHECK: store { <16 x i8>, <16 x i8>, <16 x i8>, <16 x i8> } [[VLD1XN]], ptr [[__RET]]
diff --git a/clang/test/CodeGen/arm-swiftcall.c b/clang/test/CodeGen/arm-swiftcall.c
index 12bccb5087c8e..9fa607a968ccf 100644
--- a/clang/test/CodeGen/arm-swiftcall.c
+++ b/clang/test/CodeGen/arm-swiftcall.c
@@ -22,7 +22,7 @@ SWIFTCALL int indirect_result_2(OUT int *arg0, OUT float *arg1) {  __builtin_unr
 
 typedef struct { char array[1024]; } struct_reallybig;
 SWIFTCALL struct_reallybig indirect_result_3(OUT int *arg0, OUT float *arg1) { __builtin_unreachable(); }
-// CHECK-LABEL: define{{.*}} void @indirect_result_3(ptr noalias sret(%struct.struct_reallybig) {{.*}}, ptr noalias align 4 dereferenceable(4){{.*}}, ptr noalias align 4 dereferenceable(4){{.*}})
+// CHECK-LABEL: define{{.*}} void @indirect_result_3(ptr dead_on_unwind noalias writable sret(%struct.struct_reallybig) {{.*}}, ptr noalias align 4 dereferenceable(4){{.*}}, ptr noalias align 4 dereferenceable(4){{.*}})
 
 SWIFTCALL void context_1(CONTEXT void *self) {}
 // CHECK-LABEL: define{{.*}} void @context_1(ptr swiftself
@@ -258,7 +258,7 @@ typedef struct {
 } struct_big_1;
 TEST(struct_big_1)
 
-// CHECK-LABEL: define{{.*}} void @return_struct_big_1({{.*}} noalias sret({{.*}})
+// CHECK-LABEL: define{{.*}} void @return_struct_big_1({{.*}} dead_on_unwind noalias writable sret({{.*}})
 
 // Should not be byval.
 // CHECK-LABEL: define{{.*}} void @take_struct_big_1(ptr{{( %.*)?}})
@@ -553,7 +553,7 @@ typedef struct {
   double d4;
 } struct_d5;
 TEST(struct_d5)
-// CHECK: define{{.*}} swiftcc void @return_struct_d5(ptr noalias sret([[STRUCT5:%.*]])
+// CHECK: define{{.*}} swiftcc void @return_struct_d5(ptr dead_on_unwind noalias writable sret([[STRUCT5:%.*]])
 // CHECK: define{{.*}} swiftcc void @take_struct_d5(ptr
 
 typedef struct {
@@ -703,7 +703,7 @@ typedef struct {
   long long l2;
 } struct_l3;
 TEST(struct_l3)
-// CHECK: define{{.*}} swiftcc void @return_struct_l3(ptr noalias sret([[STRUCT:%.*]])
+// CHECK: define{{.*}} swiftcc void @return_struct_l3(ptr dead_on_unwind noalias writable sret([[STRUCT:%.*]])
 // CHECK: define{{.*}} swiftcc void @take_struct_l3(ptr
 
 typedef struct {
@@ -713,7 +713,7 @@ typedef struct {
   long long l3;
 } struct_l4;
 TEST(struct_l4)
-// CHECK: define{{.*}} swiftcc void @return_struct_l4(ptr noalias sret([[STRUCT:%.*]])
+// CHECK: define{{.*}} swiftcc void @return_struct_l4(ptr dead_on_unwind noalias writable sret([[STRUCT:%.*]])
 // CHECK: define{{.*}} swiftcc void @take_struct_l4(ptr
 
 typedef struct {
@@ -724,7 +724,7 @@ typedef struct {
   long long l4;
 } struct_l5;
 TEST(struct_l5)
-// CHECK: define{{.*}} swiftcc void @return_struct_l5(ptr noalias sret([[STRUCT5:%.*]])
+// CHECK: define{{.*}} swiftcc void @return_struct_l5(ptr dead_on_unwind noalias writable sret([[STRUCT5:%.*]])
 // CHECK: define{{.*}} swiftcc void @take_struct_l5(ptr
 
 typedef struct {
@@ -769,7 +769,7 @@ typedef struct {
   char16 c4;
 } struct_vc5;
 TEST(struct_vc5)
-// CHECK: define{{.*}} swiftcc void @return_struct_vc5(ptr noalias sret([[STRUCT:%.*]])
+// CHECK: define{{.*}} swiftcc void @return_struct_vc5(ptr dead_on_unwind noalias writable sret([[STRUCT:%.*]])
 // CHECK: define{{.*}} swiftcc void @take_struct_vc5(ptr
 
 typedef struct {
@@ -814,7 +814,7 @@ typedef struct {
   short8 c4;
 } struct_vs5;
 TEST(struct_vs5)
-// CHECK: define{{.*}} swiftcc void @return_struct_vs5(ptr noalias sret([[STRUCT:%.*]])
+// CHECK: define{{.*}} swiftcc void @return_struct_vs5(ptr dead_on_unwind noalias writable sret([[STRUCT:%.*]])
 // CHECK: define{{.*}} swiftcc void @take_struct_vs5(ptr
 
 typedef struct {
@@ -859,7 +859,7 @@ typedef struct {
   int4 c4;
 } struct_vi5;
 TEST(struct_vi5)
-// CHECK: define{{.*}} swiftcc void @return_struct_vi5(ptr noalias sret([[STRUCT:%.*]])
+// CHECK: define{{.*}} swiftcc void @return_struct_vi5(ptr dead_on_unwind noalias writable sret([[STRUCT:%.*]])
 // CHECK: define{{.*}} swiftcc void @take_struct_vi5(ptr
 
 typedef struct {
@@ -887,7 +887,7 @@ typedef struct {
   long2 c4;
 } struct_vl5;
 TEST(struct_vl5)
-// CHECK: define{{.*}} swiftcc void @return_struct_vl5(ptr noalias sret([[STRUCT:%.*]])
+// CHECK: define{{.*}} swiftcc void @return_struct_vl5(ptr dead_on_unwind noalias writable sret([[STRUCT:%.*]])
 // CHECK: define{{.*}} swiftcc void @take_struct_vl5(ptr
 
 typedef struct {
@@ -915,7 +915,7 @@ typedef struct {
   double2 c4;
 } struct_vd5;
 TEST(struct_vd5)
-// CHECK: define{{.*}} swiftcc void @return_struct_vd5(ptr noalias sret([[STRUCT:%.*]])
+// CHECK: define{{.*}} swiftcc void @return_struct_vd5(ptr dead_on_unwind noalias writable sret([[STRUCT:%.*]])
 // CHECK: define{{.*}} swiftcc void @take_struct_vd5(ptr
 
 typedef struct {
@@ -939,7 +939,7 @@ typedef struct {
   double4 c2;
 } struct_vd43;
 TEST(struct_vd43)
-// CHECK: define{{.*}} swiftcc void @return_struct_vd43(ptr noalias sret([[STRUCT:%.*]])
+// CHECK: define{{.*}} swiftcc void @return_struct_vd43(ptr dead_on_unwind noalias writable sret([[STRUCT:%.*]])
 // CHECK: define{{.*}} swiftcc void @take_struct_vd43(ptr
 
 typedef struct {
@@ -975,7 +975,7 @@ typedef struct {
   float4 c4;
 } struct_vf5;
 TEST(struct_vf5)
-// CHECK: define{{.*}} swiftcc void @return_struct_vf5(ptr noalias sret([[STRUCT:%.*]])
+// CHECK: define{{.*}} swiftcc void @return_struct_vf5(ptr dead_on_unwind noalias writable sret([[STRUCT:%.*]])
 // CHECK: define{{.*}} swiftcc void @take_struct_vf5(ptr
 
 typedef struct {
diff --git a/clang/test/CodeGen/arm-varargs.c b/clang/test/CodeGen/arm-varargs.c
index 17330262e6ad6..f754c7f52e590 100644
--- a/clang/test/CodeGen/arm-varargs.c
+++ b/clang/test/CodeGen/arm-varargs.c
@@ -23,7 +23,7 @@ struct bigstruct {
 };
 
 struct bigstruct simple_struct(void) {
-// CHECK-LABEL: define{{.*}} void @simple_struct(ptr noalias sret(%struct.bigstruct) align 4 %agg.result)
+// CHECK-LABEL: define{{.*}} void @simple_struct(ptr dead_on_unwind noalias writable sret(%struct.bigstruct) align 4 %agg.result)
   return va_arg(the_list, struct bigstruct);
 // CHECK: [[CUR:%[a-z0-9._]+]] = load ptr, ptr @the_list, align 4
 // CHECK: [[NEXT:%[a-z0-9._]+]] = getelementptr inbounds i8, ptr [[CUR]], i32 40
@@ -38,7 +38,7 @@ struct aligned_bigstruct {
 };
 
 struct aligned_bigstruct simple_aligned_struct(void) {
-// CHECK-LABEL: define{{.*}} void @simple_aligned_struct(ptr noalias sret(%struct.aligned_bigstruct) align 8 %agg.result)
+// CHECK-LABEL: define{{.*}} void @simple_aligned_struct(ptr dead_on_unwind noalias writable sret(%struct.aligned_bigstruct) align 8 %agg.result)
   return va_arg(the_list, struct aligned_bigstruct);
 // CHECK: [[CUR:%[a-z0-9._]+]] = load ptr, ptr @the_list, align 4
 // CHECK: [[CUR_INT_ADD:%[a-z0-9._]+]] = getelementptr inbounds i8, ptr [[CUR]], i32 7
@@ -66,7 +66,7 @@ struct hfa {
 };
 
 struct hfa simple_hfa(void) {
-// CHECK-LABEL: define{{.*}} void @simple_hfa(ptr noalias sret(%struct.hfa) align 4 %agg.result)
+// CHECK-LABEL: define{{.*}} void @simple_hfa(ptr dead_on_unwind noalias writable sret(%struct.hfa) align 4 %agg.result)
   return va_arg(the_list, struct hfa);
 // CHECK: [[CUR:%[a-z0-9._]+]] = load ptr, ptr @the_list, align 4
 // CHECK: [[NEXT:%[a-z0-9._]+]] = getelementptr inbounds i8, ptr [[CUR]], i32 8
@@ -159,7 +159,7 @@ typedef struct __attribute__((aligned(16))) {
   int val;
 } overaligned_int_struct;
 overaligned_int_struct overaligned_int_struct_test(void) {
-// CHECK-LABEL: define{{.*}} void @overaligned_int_struct_test(ptr noalias sret(%struct.overaligned_int_struct) align 16 %agg.result)
+// CHECK-LABEL: define{{.*}} void @overaligned_int_struct_test(ptr dead_on_unwind noalias writable sret(%struct.overaligned_int_struct) align 16 %agg.result)
   return va_arg(the_list, overaligned_int_struct);
 // CHECK: [[CUR:%[a-z0-9._]+]] = load ptr, ptr @the_list, align 4
 // CHECK: [[NEXT:%[a-z0-9._]+]] = getelementptr inbounds i8, ptr [[CUR]], i32 16
@@ -172,7 +172,7 @@ typedef struct __attribute__((packed,aligned(2))) {
   long long val;
 } underaligned_long_long_struct;
 underaligned_long_long_struct underaligned_long_long_struct_test(void) {
-// CHECK-LABEL: define{{.*}} void @underaligned_long_long_struct_test(ptr noalias sret(%struct.underaligned_long_long_struct) align 2 %agg.result)
+// CHECK-LABEL: define{{.*}} void @underaligned_long_long_struct_test(ptr dead_on_unwind noalias writable sret(%struct.underaligned_long_long_struct) align 2 %agg.result)
   return va_arg(the_list, underaligned_long_long_struct);
 // CHECK: [[CUR:%[a-z0-9._]+]] = load ptr, ptr @the_list, align 4
 // CHECK: [[NEXT:%[a-z0-9._]+]] = getelementptr inbounds i8, ptr [[CUR]], i32 8
@@ -185,7 +185,7 @@ typedef struct __attribute__((aligned(16))) {
   long long val;
 } overaligned_long_long_struct;
 overaligned_long_long_struct overaligned_long_long_struct_test(void) {
-// CHECK-LABEL: define{{.*}} void @overaligned_long_long_struct_test(ptr noalias sret(%struct.overaligned_long_long_struct) align 16 %agg.result)
+// CHECK-LABEL: define{{.*}} void @overaligned_long_long_struct_test(ptr dead_on_unwind noalias writable sret(%struct.overaligned_long_long_struct) align 16 %agg.result)
   return va_arg(the_list, overaligned_long_long_struct);
 // CHECK: [[CUR:%[a-z0-9._]+]] = load ptr, ptr @the_list, align 4
 // CHECK: [[CUR_INT_ADD:%[a-z0-9._]+]] = getelementptr inbounds i8, ptr [[CUR]], i32 7
@@ -219,7 +219,7 @@ typedef struct {
   int val __attribute__((aligned(16)));
 } overaligned_int_struct_member;
 overaligned_int_struct_member overaligned_int_struct_member_test(void) {
-// CHECK-LABEL: define{{.*}} void @overaligned_int_struct_member_test(ptr noalias sret(%struct.overaligned_int_struct_member) align 16 %agg.result)
+// CHECK-LABEL: define{{.*}} void @overaligned_int_struct_member_test(ptr dead_on_unwind noalias writable sret(%struct.overaligned_int_struct_member) align 16 %agg.result)
   return va_arg(the_list, overaligned_int_struct_member);
 // CHECK: [[CUR:%[a-z0-9._]+]] = load ptr, ptr @the_list, align 4
 // CHECK: [[CUR_INT_ADD:%[a-z0-9._]+]] = getelementptr inbounds i8, ptr [[CUR]], i32 7
@@ -234,7 +234,7 @@ typedef struct {
   long long val __attribute__((packed,aligned(2)));
 } underaligned_long_long_struct_member;
 underaligned_long_long_struct_member underaligned_long_long_struct_member_test(void) {
-// CHECK-LABEL: define{{.*}} void @underaligned_long_long_struct_member_test(ptr noalias sret(%struct.underaligned_long_long_struct_member) align 2 %agg.result)
+// CHECK-LABEL: define{{.*}} void @underaligned_long_long_struct_member_test(ptr dead_on_unwind noalias writable sret(%struct.underaligned_long_long_struct_member) align 2 %agg.result)
   return va_arg(the_list, underaligned_long_long_struct_member);
 // CHECK: [[CUR:%[a-z0-9._]+]] = load ptr, ptr @the_list, align 4
 // CHECK: [[NEXT:%[a-z0-9._]+]] = getelementptr inbounds i8, ptr [[CUR]], i32 8
@@ -247,7 +247,7 @@ typedef struct {
   long long val __attribute__((aligned(16)));
 } overaligned_long_long_struct_member;
 overaligned_long_long_struct_member overaligned_long_long_struct_member_test(void) {
-// CHECK-LABEL: define{{.*}} void @overaligned_long_long_struct_member_test(ptr noalias sret(%struct.overaligned_long_long_struct_member) align 16 %agg.result)
+// CHECK-LABEL: define{{.*}} void @overaligned_long_long_struct_member_test(ptr dead_on_unwind noalias writable sret(%struct.overaligned_long_long_struct_member) align 16 %agg.result)
   return va_arg(the_list, overaligned_long_long_struct_member);
 // CHECK: [[CUR:%[a-z0-9._]+]] = load ptr, ptr @the_list, align 4
 // CHECK: [[CUR_INT_ADD:%[a-z0-9._]+]] = getelementptr inbounds i8, ptr [[CUR]], i32 7
diff --git a/clang/test/CodeGen/arm-vector-arguments.c b/clang/test/CodeGen/arm-vector-arguments.c
index a92e5a84798bc..0e3abfb7e5a06 100644
--- a/clang/test/CodeGen/arm-vector-arguments.c
+++ b/clang/test/CodeGen/arm-vector-arguments.c
@@ -9,7 +9,7 @@
 
 #include 
 
-// CHECK: define{{.*}} void @f0(ptr noalias sret(%struct.int8x16x2_t) align 16 %agg.result, <16 x i8> noundef %{{.*}}, <16 x i8> noundef %{{.*}})
+// CHECK: define{{.*}} void @f0(ptr dead_on_unwind noalias writable sret(%struct.int8x16x2_t) align 16 %agg.result, <16 x i8> noundef %{{.*}}, <16 x i8> noundef %{{.*}})
 int8x16x2_t f0(int8x16_t a0, int8x16_t a1) {
   return vzipq_s8(a0, a1);
 }
@@ -25,7 +25,7 @@ typedef float T_float32x16 __attribute__ ((__vector_size__ (64)));
 T_float32x2 f1_0(T_float32x2 a0) { return a0; }
 // CHECK: define{{.*}} <4 x float> @f1_1(<4 x float> noundef %{{.*}})
 T_float32x4 f1_1(T_float32x4 a0) { return a0; }
-// CHECK: define{{.*}} void @f1_2(ptr noalias sret(<8 x float>) align 32 %{{.*}}, <8 x float> noundef %{{.*}})
+// CHECK: define{{.*}} void @f1_2(ptr dead_on_unwind noalias writable sret(<8 x float>) align 32 %{{.*}}, <8 x float> noundef %{{.*}})
 T_float32x8 f1_2(T_float32x8 a0) { return a0; }
-// CHECK: define{{.*}} void @f1_3(ptr noalias sret(<16 x float>) align 64 %{{.*}}, <16 x float> noundef %{{.*}})
+// CHECK: define{{.*}} void @f1_3(ptr dead_on_unwind noalias writable sret(<16 x float>) align 64 %{{.*}}, <16 x float> noundef %{{.*}})
 T_float32x16 f1_3(T_float32x16 a0) { return a0; }
diff --git a/clang/test/CodeGen/arm-vfp16-arguments.c b/clang/test/CodeGen/arm-vfp16-arguments.c
index 59120c1cc7f14..da034626024f8 100644
--- a/clang/test/CodeGen/arm-vfp16-arguments.c
+++ b/clang/test/CodeGen/arm-vfp16-arguments.c
@@ -71,6 +71,6 @@ void test_hfa(hfa_t a) {}
 
 hfa_t ghfa;
 hfa_t test_ret_hfa(void) { return ghfa; }
-// CHECK-SOFT: define{{.*}} void @test_ret_hfa(ptr noalias nocapture writeonly sret(%struct.hfa_t) align 8 %agg.result)
+// CHECK-SOFT: define{{.*}} void @test_ret_hfa(ptr dead_on_unwind noalias nocapture writable writeonly sret(%struct.hfa_t) align 8 %agg.result)
 // CHECK-HARD: define{{.*}} arm_aapcs_vfpcc [2 x <2 x i32>] @test_ret_hfa()
 // CHECK-FULL: define{{.*}} arm_aapcs_vfpcc %struct.hfa_t @test_ret_hfa()
diff --git a/clang/test/CodeGen/arm-vfp16-arguments2.cpp b/clang/test/CodeGen/arm-vfp16-arguments2.cpp
index 2ba0c0d349ab7..6221e85e856b4 100644
--- a/clang/test/CodeGen/arm-vfp16-arguments2.cpp
+++ b/clang/test/CodeGen/arm-vfp16-arguments2.cpp
@@ -37,27 +37,27 @@ struct S5 : B1 {
   B1 M[1];
 };
 
-// CHECK-SOFT: define{{.*}} void @_Z2f12S1(ptr noalias nocapture writeonly sret(%struct.S1) align 8 %agg.result, [2 x i64] %s1.coerce)
+// CHECK-SOFT: define{{.*}} void @_Z2f12S1(ptr dead_on_unwind noalias nocapture writable writeonly sret(%struct.S1) align 8 %agg.result, [2 x i64] %s1.coerce)
 // CHECK-HARD: define{{.*}} arm_aapcs_vfpcc [2 x <2 x i32>] @_Z2f12S1([2 x <2 x i32>] returned %s1.coerce)
 // CHECK-FULL: define{{.*}} arm_aapcs_vfpcc %struct.S1 @_Z2f12S1(%struct.S1 returned %s1.coerce)
 struct S1 f1(struct S1 s1) { return s1; }
 
-// CHECK-SOFT: define{{.*}} void @_Z2f22S2(ptr noalias nocapture writeonly sret(%struct.S2) align 8 %agg.result, [4 x i32] %s2.coerce)
+// CHECK-SOFT: define{{.*}} void @_Z2f22S2(ptr dead_on_unwind noalias nocapture writable writeonly sret(%struct.S2) align 8 %agg.result, [4 x i32] %s2.coerce)
 // CHECK-HARD: define{{.*}} arm_aapcs_vfpcc [2 x <2 x i32>] @_Z2f22S2([2 x <2 x i32>] returned %s2.coerce)
 // CHECK-FULL: define{{.*}} arm_aapcs_vfpcc %struct.S2 @_Z2f22S2(%struct.S2 returned %s2.coerce)
 struct S2 f2(struct S2 s2) { return s2; }
 
-// CHECK-SOFT: define{{.*}} void @_Z2f32S3(ptr noalias nocapture writeonly sret(%struct.S3) align 8 %agg.result, [2 x i64] %s3.coerce)
+// CHECK-SOFT: define{{.*}} void @_Z2f32S3(ptr dead_on_unwind noalias nocapture writable writeonly sret(%struct.S3) align 8 %agg.result, [2 x i64] %s3.coerce)
 // CHECK-HARD: define{{.*}} arm_aapcs_vfpcc [2 x <2 x i32>] @_Z2f32S3([2 x <2 x i32>] returned %s3.coerce)
 // CHECK-FULL: define{{.*}} arm_aapcs_vfpcc %struct.S3 @_Z2f32S3(%struct.S3 returned %s3.coerce)
 struct S3 f3(struct S3 s3) { return s3; }
 
-// CHECK-SOFT: define{{.*}} void @_Z2f42S4(ptr noalias nocapture writeonly sret(%struct.S4) align 8 %agg.result, [2 x i64] %s4.coerce)
+// CHECK-SOFT: define{{.*}} void @_Z2f42S4(ptr dead_on_unwind noalias nocapture writable writeonly sret(%struct.S4) align 8 %agg.result, [2 x i64] %s4.coerce)
 // CHECK-HARD: define{{.*}} arm_aapcs_vfpcc [2 x <2 x i32>] @_Z2f42S4([2 x <2 x i32>] returned %s4.coerce)
 // CHECK-FULL: define{{.*}} arm_aapcs_vfpcc %struct.S4 @_Z2f42S4(%struct.S4 returned %s4.coerce)
 struct S4 f4(struct S4 s4) { return s4; }
 
-// CHECK-SOFT: define{{.*}} void @_Z2f52S5(ptr noalias nocapture writeonly sret(%struct.S5) align 8 %agg.result, [2 x i64] %s5.coerce)
+// CHECK-SOFT: define{{.*}} void @_Z2f52S5(ptr dead_on_unwind noalias nocapture writable writeonly sret(%struct.S5) align 8 %agg.result, [2 x i64] %s5.coerce)
 // CHECK-HARD: define{{.*}} arm_aapcs_vfpcc %struct.S5 @_Z2f52S5(%struct.S5 returned %s5.coerce)
 // CHECK-FULL: define{{.*}} arm_aapcs_vfpcc %struct.S5 @_Z2f52S5(%struct.S5 returned %s5.coerce)
 struct S5 f5(struct S5 s5) { return s5; }
diff --git a/clang/test/CodeGen/arm64-arguments.c b/clang/test/CodeGen/arm64-arguments.c
index caa71ced0a8ae..8ed9d952f80c8 100644
--- a/clang/test/CodeGen/arm64-arguments.c
+++ b/clang/test/CodeGen/arm64-arguments.c
@@ -226,9 +226,9 @@ T_float32x2 f1_0(T_float32x2 a0) { return a0; }
 // CHECK: define{{.*}} <4 x float> @f1_1(<4 x float> noundef %{{.*}})
 T_float32x4 f1_1(T_float32x4 a0) { return a0; }
 // Vector with length bigger than 16-byte is illegal and is passed indirectly.
-// CHECK: define{{.*}} void @f1_2(ptr noalias sret(<8 x float>) align 16 %{{.*}}, ptr noundef %0)
+// CHECK: define{{.*}} void @f1_2(ptr dead_on_unwind noalias writable sret(<8 x float>) align 16 %{{.*}}, ptr noundef %0)
 T_float32x8 f1_2(T_float32x8 a0) { return a0; }
-// CHECK: define{{.*}} void @f1_3(ptr noalias sret(<16 x float>) align 16 %{{.*}}, ptr noundef %0)
+// CHECK: define{{.*}} void @f1_3(ptr dead_on_unwind noalias writable sret(<16 x float>) align 16 %{{.*}}, ptr noundef %0)
 T_float32x16 f1_3(T_float32x16 a0) { return a0; }
 
 // Testing alignment with aggregates: HFA, aggregates with size <= 16 bytes and
diff --git a/clang/test/CodeGen/arm64-microsoft-arguments.cpp b/clang/test/CodeGen/arm64-microsoft-arguments.cpp
index a9ae6911b16e3..e8309888dcfe2 100644
--- a/clang/test/CodeGen/arm64-microsoft-arguments.cpp
+++ b/clang/test/CodeGen/arm64-microsoft-arguments.cpp
@@ -28,8 +28,8 @@ S2 f2() {
 }
 
 // Pass and return for type size > 16 bytes.
-// CHECK: define {{.*}} void @{{.*}}f3{{.*}}(ptr noalias sret(%struct.S3) align 4 %agg.result)
-// CHECK: call void {{.*}}func3{{.*}}(ptr sret(%struct.S3) align 4 %agg.result, ptr noundef %agg.tmp)
+// CHECK: define {{.*}} void @{{.*}}f3{{.*}}(ptr dead_on_unwind noalias writable sret(%struct.S3) align 4 %agg.result)
+// CHECK: call void {{.*}}func3{{.*}}(ptr dead_on_unwind writable sret(%struct.S3) align 4 %agg.result, ptr noundef %agg.tmp)
 struct S3 {
   int a[5];
 };
@@ -42,8 +42,8 @@ S3 f3() {
 
 // Pass and return aggregate (of size < 16 bytes) with non-trivial destructor.
 // Passed directly but returned indirectly.
-// CHECK: define {{.*}} void {{.*}}f4{{.*}}(ptr inreg noalias sret(%struct.S4) align 4 %agg.result)
-// CHECK: call void {{.*}}func4{{.*}}(ptr inreg sret(%struct.S4) align 4 %agg.result, [2 x i64] %0)
+// CHECK: define {{.*}} void {{.*}}f4{{.*}}(ptr dead_on_unwind inreg noalias writable sret(%struct.S4) align 4 %agg.result)
+// CHECK: call void {{.*}}func4{{.*}}(ptr dead_on_unwind inreg writable sret(%struct.S4) align 4 %agg.result, [2 x i64] %0)
 struct S4 {
   int a[3];
   ~S4();
@@ -56,8 +56,8 @@ S4 f4() {
 }
 
 // Pass and return from instance method called from instance method.
-// CHECK: define {{.*}} void @{{.*}}bar@Q1{{.*}}(ptr {{[^,]*}} %this, ptr inreg noalias sret(%class.P1) align 1 %agg.result)
-// CHECK: call void {{.*}}foo@P1{{.*}}(ptr noundef{{[^,]*}} %ref.tmp, ptr inreg sret(%class.P1) align 1 %agg.result, i8 %0)
+// CHECK: define {{.*}} void @{{.*}}bar@Q1{{.*}}(ptr {{[^,]*}} %this, ptr dead_on_unwind inreg noalias writable sret(%class.P1) align 1 %agg.result)
+// CHECK: call void {{.*}}foo@P1{{.*}}(ptr noundef{{[^,]*}} %ref.tmp, ptr dead_on_unwind inreg writable sret(%class.P1) align 1 %agg.result, i8 %0)
 
 class P1 {
 public:
@@ -76,7 +76,7 @@ P1 Q1::bar() {
 
 // Pass and return from instance method called from free function.
 // CHECK: define {{.*}} void {{.*}}bar{{.*}}()
-// CHECK: call void {{.*}}foo@P2{{.*}}(ptr noundef{{[^,]*}} %ref.tmp, ptr inreg sret(%class.P2) align 1 %retval, i8 %0)
+// CHECK: call void {{.*}}foo@P2{{.*}}(ptr noundef{{[^,]*}} %ref.tmp, ptr dead_on_unwind inreg writable sret(%class.P2) align 1 %retval, i8 %0)
 class P2 {
 public:
   P2 foo(P2 x);
@@ -89,8 +89,8 @@ P2 bar() {
 
 // Pass and return an object with a user-provided constructor (passed directly,
 // returned indirectly)
-// CHECK: define {{.*}} void @{{.*}}f5{{.*}}(ptr inreg noalias sret(%struct.S5) align 4 %agg.result)
-// CHECK: call void {{.*}}func5{{.*}}(ptr inreg sret(%struct.S5) align 4 %agg.result, i64 {{.*}})
+// CHECK: define {{.*}} void @{{.*}}f5{{.*}}(ptr dead_on_unwind inreg noalias writable sret(%struct.S5) align 4 %agg.result)
+// CHECK: call void {{.*}}func5{{.*}}(ptr dead_on_unwind inreg writable sret(%struct.S5) align 4 %agg.result, i64 {{.*}})
 struct S5 {
   S5();
   int x;
@@ -146,8 +146,8 @@ struct S8 {
   int y;
 };
 
-// CHECK: define {{.*}} void {{.*}}?f8{{.*}}(ptr inreg noalias sret(%struct.S8) align 4 {{.*}})
-// CHECK: call void {{.*}}func8{{.*}}(ptr inreg sret(%struct.S8) align 4 {{.*}}, i64 {{.*}})
+// CHECK: define {{.*}} void {{.*}}?f8{{.*}}(ptr dead_on_unwind inreg noalias writable sret(%struct.S8) align 4 {{.*}})
+// CHECK: call void {{.*}}func8{{.*}}(ptr dead_on_unwind inreg writable sret(%struct.S8) align 4 {{.*}}, i64 {{.*}})
 S8 func8(S8 x);
 S8 f8() {
   S8 x;
@@ -157,8 +157,8 @@ S8 f8() {
 
 // Pass and return an object with a non-trivial copy-assignment operator and
 // a trivial copy constructor (passed directly, returned indirectly)
-// CHECK: define {{.*}} void @"?f9@@YA?AUS9@@XZ"(ptr inreg noalias sret(%struct.S9) align 4 {{.*}})
-// CHECK: call void {{.*}}func9{{.*}}(ptr inreg sret(%struct.S9) align 4 {{.*}}, i64 {{.*}})
+// CHECK: define {{.*}} void @"?f9@@YA?AUS9@@XZ"(ptr dead_on_unwind inreg noalias writable sret(%struct.S9) align 4 {{.*}})
+// CHECK: call void {{.*}}func9{{.*}}(ptr dead_on_unwind inreg writable sret(%struct.S9) align 4 {{.*}}, i64 {{.*}})
 struct S9 {
   S9& operator=(const S9&);
   int x;
@@ -174,8 +174,8 @@ S9 f9() {
 
 // Pass and return an object with a base class (passed directly, returned
 // indirectly).
-// CHECK: define dso_local void {{.*}}f10{{.*}}(ptr inreg noalias sret(%struct.S10) align 4 {{.*}})
-// CHECK: call void {{.*}}func10{{.*}}(ptr inreg sret(%struct.S10) align 4 {{.*}}, [2 x i64] {{.*}})
+// CHECK: define dso_local void {{.*}}f10{{.*}}(ptr dead_on_unwind inreg noalias writable sret(%struct.S10) align 4 {{.*}})
+// CHECK: call void {{.*}}func10{{.*}}(ptr dead_on_unwind inreg writable sret(%struct.S10) align 4 {{.*}}, [2 x i64] {{.*}})
 struct S10 : public S1 {
   int x;
 };
@@ -189,8 +189,8 @@ S10 f10() {
 
 // Pass and return a non aggregate object exceeding > 128 bits (passed
 // indirectly, returned indirectly)
-// CHECK: define dso_local void {{.*}}f11{{.*}}(ptr inreg noalias sret(%struct.S11) align 8 {{.*}})
-// CHECK: call void {{.*}}func11{{.*}}(ptr inreg sret(%struct.S11) align 8 {{.*}}, ptr {{.*}})
+// CHECK: define dso_local void {{.*}}f11{{.*}}(ptr dead_on_unwind inreg noalias writable sret(%struct.S11) align 8 {{.*}})
+// CHECK: call void {{.*}}func11{{.*}}(ptr dead_on_unwind inreg writable sret(%struct.S11) align 8 {{.*}}, ptr {{.*}})
 struct S11 {
   virtual void f();
   int a[5];
diff --git a/clang/test/CodeGen/arm64_32.c b/clang/test/CodeGen/arm64_32.c
index 31d94c610704a..f7473610d30d9 100644
--- a/clang/test/CodeGen/arm64_32.c
+++ b/clang/test/CodeGen/arm64_32.c
@@ -27,4 +27,4 @@ long double LongDoubleVar = 0.0;
 
 typedef float __attribute__((ext_vector_type(16))) v16f32;
 v16f32 func(v16f32 in) { return in; }
-// CHECK: define{{.*}} void @func(ptr noalias sret(<16 x float>) align 16 {{%.*}}, <16 x float> noundef {{%.*}})
+// CHECK: define{{.*}} void @func(ptr dead_on_unwind noalias writable sret(<16 x float>) align 16 {{%.*}}, <16 x float> noundef {{%.*}})
diff --git a/clang/test/CodeGen/armv7k-abi.c b/clang/test/CodeGen/armv7k-abi.c
index e070d5a9c7041..fd18dafa7d03f 100644
--- a/clang/test/CodeGen/armv7k-abi.c
+++ b/clang/test/CodeGen/armv7k-abi.c
@@ -42,7 +42,7 @@ typedef struct {
 // CHECK: define{{.*}} void @big_struct_indirect(ptr noundef %b)
 void big_struct_indirect(BigStruct b) {}
 
-// CHECK: define{{.*}} void @return_big_struct_indirect(ptr noalias sret
+// CHECK: define{{.*}} void @return_big_struct_indirect(ptr dead_on_unwind noalias writable sret
 BigStruct return_big_struct_indirect() {}
 
 // Structs smaller than 16 bytes should be passed directly, and coerced to
diff --git a/clang/test/CodeGen/attr-noundef.cpp b/clang/test/CodeGen/attr-noundef.cpp
index a2aa7e313f0e8..d236b35fdfd7d 100644
--- a/clang/test/CodeGen/attr-noundef.cpp
+++ b/clang/test/CodeGen/attr-noundef.cpp
@@ -26,7 +26,7 @@ struct NoCopy {
 };
 NoCopy ret_nocopy() { return {}; }
 void pass_nocopy(NoCopy e) {}
-// CHECK: [[DEF]] void @{{.*}}ret_nocopy{{.*}}(ptr noalias sret({{[^)]+}}) align 4 %
+// CHECK: [[DEF]] void @{{.*}}ret_nocopy{{.*}}(ptr dead_on_unwind noalias writable sret({{[^)]+}}) align 4 %
 // CHECK: [[DEF]] void @{{.*}}pass_nocopy{{.*}}(ptr noundef %
 
 struct Huge {
@@ -34,7 +34,7 @@ struct Huge {
 };
 Huge ret_huge() { return {}; }
 void pass_huge(Huge h) {}
-// CHECK: [[DEF]] void @{{.*}}ret_huge{{.*}}(ptr noalias sret({{[^)]+}}) align 4 %
+// CHECK: [[DEF]] void @{{.*}}ret_huge{{.*}}(ptr dead_on_unwind noalias writable sret({{[^)]+}}) align 4 %
 // CHECK: [[DEF]] void @{{.*}}pass_huge{{.*}}(ptr noundef
 } // namespace check_structs
 
@@ -58,7 +58,7 @@ union NoCopy {
 };
 NoCopy ret_nocopy() { return {}; }
 void pass_nocopy(NoCopy e) {}
-// CHECK: [[DEF]] void @{{.*}}ret_nocopy{{.*}}(ptr noalias sret({{[^)]+}}) align 4 %
+// CHECK: [[DEF]] void @{{.*}}ret_nocopy{{.*}}(ptr dead_on_unwind noalias writable sret({{[^)]+}}) align 4 %
 // CHECK: [[DEF]] void @{{.*}}pass_nocopy{{.*}}(ptr noundef %
 } // namespace check_unions
 
diff --git a/clang/test/CodeGen/blocks.c b/clang/test/CodeGen/blocks.c
index 469cf7cb89a2c..8f947fcdfccb2 100644
--- a/clang/test/CodeGen/blocks.c
+++ b/clang/test/CodeGen/blocks.c
@@ -15,7 +15,7 @@ struct s0 {
   int a[64];
 };
 
-// CHECK: define internal void @__f2_block_invoke(ptr noalias sret(%struct.s0) align 4 {{%.*}}, ptr noundef {{%.*}}, ptr noundef byval(%struct.s0) align 4 {{.*}})
+// CHECK: define internal void @__f2_block_invoke(ptr dead_on_unwind noalias writable sret(%struct.s0) align 4 {{%.*}}, ptr noundef {{%.*}}, ptr noundef byval(%struct.s0) align 4 {{.*}})
 struct s0 f2(struct s0 a0) {
   return ^(struct s0 a1){ return a1; }(a0);
 }
diff --git a/clang/test/CodeGen/c11atomics-ios.c b/clang/test/CodeGen/c11atomics-ios.c
index af489811edc58..bcb6519ab0dc8 100644
--- a/clang/test/CodeGen/c11atomics-ios.c
+++ b/clang/test/CodeGen/c11atomics-ios.c
@@ -178,7 +178,7 @@ void testPromotedStruct(_Atomic(PS) *fp) {
 }
 
 PS test_promoted_load(_Atomic(PS) *addr) {
-  // CHECK-LABEL: @test_promoted_load(ptr noalias sret(%struct.PS) align 2 %agg.result, ptr noundef %addr)
+  // CHECK-LABEL: @test_promoted_load(ptr dead_on_unwind noalias writable sret(%struct.PS) align 2 %agg.result, ptr noundef %addr)
   // CHECK:   [[ADDR_ARG:%.*]] = alloca ptr, align 4
   // CHECK:   [[ATOMIC_RES:%.*]] = alloca { %struct.PS, [2 x i8] }, align 8
   // CHECK:   store ptr %addr, ptr [[ADDR_ARG]], align 4
@@ -209,7 +209,7 @@ void test_promoted_store(_Atomic(PS) *addr, PS *val) {
 }
 
 PS test_promoted_exchange(_Atomic(PS) *addr, PS *val) {
-  // CHECK-LABEL: @test_promoted_exchange(ptr noalias sret(%struct.PS) align 2 %agg.result, ptr noundef %addr, ptr noundef %val)
+  // CHECK-LABEL: @test_promoted_exchange(ptr dead_on_unwind noalias writable sret(%struct.PS) align 2 %agg.result, ptr noundef %addr, ptr noundef %val)
   // CHECK:   [[ADDR_ARG:%.*]] = alloca ptr, align 4
   // CHECK:   [[VAL_ARG:%.*]] = alloca ptr, align 4
   // CHECK:   [[NONATOMIC_TMP:%.*]] = alloca %struct.PS, align 2
diff --git a/clang/test/CodeGen/c11atomics.c b/clang/test/CodeGen/c11atomics.c
index 773ed41991f78..dd1f52f70ae09 100644
--- a/clang/test/CodeGen/c11atomics.c
+++ b/clang/test/CodeGen/c11atomics.c
@@ -338,7 +338,7 @@ void testPromotedStruct(_Atomic(PS) *fp) {
 }
 
 PS test_promoted_load(_Atomic(PS) *addr) {
-  // CHECK-LABEL: @test_promoted_load(ptr noalias sret(%struct.PS) align 2 %agg.result, ptr noundef %addr)
+  // CHECK-LABEL: @test_promoted_load(ptr dead_on_unwind noalias writable sret(%struct.PS) align 2 %agg.result, ptr noundef %addr)
   // CHECK:   [[ADDR_ARG:%.*]] = alloca ptr, align 4
   // CHECK:   [[ATOMIC_RES:%.*]] = alloca { %struct.PS, [2 x i8] }, align 8
   // CHECK:   store ptr %addr, ptr [[ADDR_ARG]], align 4
@@ -368,7 +368,7 @@ void test_promoted_store(_Atomic(PS) *addr, PS *val) {
 }
 
 PS test_promoted_exchange(_Atomic(PS) *addr, PS *val) {
-  // CHECK-LABEL: @test_promoted_exchange(ptr noalias sret(%struct.PS) align 2 %agg.result, ptr noundef %addr, ptr noundef %val)
+  // CHECK-LABEL: @test_promoted_exchange(ptr dead_on_unwind noalias writable sret(%struct.PS) align 2 %agg.result, ptr noundef %addr, ptr noundef %val)
   // CHECK:   [[ADDR_ARG:%.*]] = alloca ptr, align 4
   // CHECK:   [[VAL_ARG:%.*]] = alloca ptr, align 4
   // CHECK:   [[NONATOMIC_TMP:%.*]] = alloca %struct.PS, align 2
diff --git a/clang/test/CodeGen/debug-names-compound-type-units.ll b/clang/test/CodeGen/debug-names-compound-type-units.ll
new file mode 100644
index 0000000000000..62cecfdeaf7f9
--- /dev/null
+++ b/clang/test/CodeGen/debug-names-compound-type-units.ll
@@ -0,0 +1,74 @@
+; REQUIRES: asserts
+; REQUIRES: x86-registered-target
+
+;; Tests that we use correct accelerator table when processing nested TUs.
+;; Assert is not triggered.
+;; File1
+;; struct Foo {
+;;   char f;
+;; };
+;; struct Foo2 {
+;;   char f;
+;;   Foo f1;
+;; };
+;; void fooFunc() {
+;; Foo2 global2;
+;; }
+;; clang++ .cpp -O0 -g2 -fdebug-types-section -gpubnames -S -emit-llvm -o .ll
+
+; RUN: llc -O0 -dwarf-version=5 -generate-type-units -filetype=obj < %s -o %t.o
+; RUN: llvm-readelf --sections %t.o | FileCheck --check-prefix=OBJ %s
+
+; OBJ: debug_names
+
+; ModuleID = 'main.cpp'
+source_filename = "main.cpp"
+target datalayout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-i128:128-f80:128-n8:16:32:64-S128"
+target triple = "x86_64-unknown-linux-gnu"
+
+%struct.Foo2 = type { i8, %struct.Foo }
+%struct.Foo = type { i8 }
+
+; Function Attrs: mustprogress noinline nounwind optnone uwtable
+define dso_local void @_Z7fooFuncv() #0 !dbg !10 {
+entry:
+  %global2 = alloca %struct.Foo2, align 1
+  call void @llvm.dbg.declare(metadata ptr %global2, metadata !14, metadata !DIExpression()), !dbg !23
+  ret void, !dbg !24
+}
+
+; Function Attrs: nocallback nofree nosync nounwind speculatable willreturn memory(none)
+declare void @llvm.dbg.declare(metadata, metadata, metadata) #1
+
+attributes #0 = { mustprogress noinline nounwind optnone uwtable "frame-pointer"="all" "min-legal-vector-width"="0" "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-cpu"="x86-64" "target-features"="+cmov,+cx8,+fxsr,+mmx,+sse,+sse2,+x87" "tune-cpu"="generic" }
+attributes #1 = { nocallback nofree nosync nounwind speculatable willreturn memory(none) }
+
+!llvm.dbg.cu = !{!0}
+!llvm.module.flags = !{!2, !3, !4, !5, !6, !7, !8}
+!llvm.ident = !{!9}
+
+!0 = distinct !DICompileUnit(language: DW_LANG_C_plus_plus_14, file: !1, producer: "clang version 18.0.0git", isOptimized: false, runtimeVersion: 0, emissionKind: FullDebug, splitDebugInlining: false)
+!1 = !DIFile(filename: "main.cpp", directory: "/smallMultipleTUs", checksumkind: CSK_MD5, checksum: "70bff4d50322126f3d8ca4178afad376")
+!2 = !{i32 7, !"Dwarf Version", i32 5}
+!3 = !{i32 2, !"Debug Info Version", i32 3}
+!4 = !{i32 1, !"wchar_size", i32 4}
+!5 = !{i32 8, !"PIC Level", i32 2}
+!6 = !{i32 7, !"PIE Level", i32 2}
+!7 = !{i32 7, !"uwtable", i32 2}
+!8 = !{i32 7, !"frame-pointer", i32 2}
+!9 = !{!"clang version 18.0.0git"}
+!10 = distinct !DISubprogram(name: "fooFunc", linkageName: "_Z7fooFuncv", scope: !1, file: !1, line: 9, type: !11, scopeLine: 9, flags: DIFlagPrototyped, spFlags: DISPFlagDefinition, unit: !0, retainedNodes: !13)
+!11 = !DISubroutineType(types: !12)
+!12 = !{null}
+!13 = !{}
+!14 = !DILocalVariable(name: "global2", scope: !10, file: !1, line: 10, type: !15)
+!15 = distinct !DICompositeType(tag: DW_TAG_structure_type, name: "Foo2", file: !1, line: 4, size: 16, flags: DIFlagTypePassByValue, elements: !16, identifier: "_ZTS4Foo2")
+!16 = !{!17, !19}
+!17 = !DIDerivedType(tag: DW_TAG_member, name: "f", scope: !15, file: !1, line: 5, baseType: !18, size: 8)
+!18 = !DIBasicType(name: "char", size: 8, encoding: DW_ATE_signed_char)
+!19 = !DIDerivedType(tag: DW_TAG_member, name: "f1", scope: !15, file: !1, line: 6, baseType: !20, size: 8, offset: 8)
+!20 = distinct !DICompositeType(tag: DW_TAG_structure_type, name: "Foo", file: !1, line: 1, size: 8, flags: DIFlagTypePassByValue, elements: !21, identifier: "_ZTS3Foo")
+!21 = !{!22}
+!22 = !DIDerivedType(tag: DW_TAG_member, name: "f", scope: !20, file: !1, line: 2, baseType: !18, size: 8)
+!23 = !DILocation(line: 10, column: 6, scope: !10)
+!24 = !DILocation(line: 11, column: 1, scope: !10)
diff --git a/clang/test/CodeGen/ext-int-cc.c b/clang/test/CodeGen/ext-int-cc.c
index bcf29fc1309de..001e866d34b45 100644
--- a/clang/test/CodeGen/ext-int-cc.c
+++ b/clang/test/CodeGen/ext-int-cc.c
@@ -226,104 +226,104 @@ _BitInt(64) ReturnPassing2(void){}
 
 _BitInt(127) ReturnPassing3(void){}
 // LIN64: define{{.*}} { i64, i64 } @ReturnPassing3(
-// WIN64: define dso_local void @ReturnPassing3(ptr noalias sret
-// LIN32: define{{.*}} void @ReturnPassing3(ptr noalias sret
-// WIN32: define dso_local void @ReturnPassing3(ptr noalias sret
-// NACL: define{{.*}} void @ReturnPassing3(ptr noalias sret
+// WIN64: define dso_local void @ReturnPassing3(ptr dead_on_unwind noalias writable sret
+// LIN32: define{{.*}} void @ReturnPassing3(ptr dead_on_unwind noalias writable sret
+// WIN32: define dso_local void @ReturnPassing3(ptr dead_on_unwind noalias writable sret
+// NACL: define{{.*}} void @ReturnPassing3(ptr dead_on_unwind noalias writable sret
 // NVPTX/64 makes the intentional choice to put all return values direct, even
 // large structures, so we do the same here.
 // NVPTX64: define{{.*}} i127 @ReturnPassing3(
 // NVPTX: define{{.*}} i127 @ReturnPassing3(
 // SPARCV9: define{{.*}} i127 @ReturnPassing3(
-// SPARC: define{{.*}} void @ReturnPassing3(ptr noalias sret
+// SPARC: define{{.*}} void @ReturnPassing3(ptr dead_on_unwind noalias writable sret
 // MIPS64: define{{.*}} i127 @ReturnPassing3(
-// MIPS: define{{.*}} void @ReturnPassing3(ptr noalias sret
-// SPIR64: define{{.*}} spir_func void @ReturnPassing3(ptr noalias sret
-// SPIR: define{{.*}} spir_func void @ReturnPassing3(ptr noalias sret
-// HEX: define{{.*}} void @ReturnPassing3(ptr noalias sret
-// LANAI: define{{.*}} void @ReturnPassing3(ptr noalias sret
-// R600: define{{.*}} void @ReturnPassing3(ptr addrspace(5) noalias sret
-// ARC: define{{.*}} void @ReturnPassing3(ptr noalias sret
-// XCORE: define{{.*}} void @ReturnPassing3(ptr noalias sret
+// MIPS: define{{.*}} void @ReturnPassing3(ptr dead_on_unwind noalias writable sret
+// SPIR64: define{{.*}} spir_func void @ReturnPassing3(ptr dead_on_unwind noalias writable sret
+// SPIR: define{{.*}} spir_func void @ReturnPassing3(ptr dead_on_unwind noalias writable sret
+// HEX: define{{.*}} void @ReturnPassing3(ptr dead_on_unwind noalias writable sret
+// LANAI: define{{.*}} void @ReturnPassing3(ptr dead_on_unwind noalias writable sret
+// R600: define{{.*}} void @ReturnPassing3(ptr addrspace(5) dead_on_unwind noalias writable sret
+// ARC: define{{.*}} void @ReturnPassing3(ptr dead_on_unwind noalias writable sret
+// XCORE: define{{.*}} void @ReturnPassing3(ptr dead_on_unwind noalias writable sret
 // RISCV64: define{{.*}} i127 @ReturnPassing3(
-// RISCV32: define{{.*}} void @ReturnPassing3(ptr noalias sret
+// RISCV32: define{{.*}} void @ReturnPassing3(ptr dead_on_unwind noalias writable sret
 // WASM: define{{.*}} i127 @ReturnPassing3(
-// SYSTEMZ: define{{.*}} void @ReturnPassing3(ptr noalias sret
+// SYSTEMZ: define{{.*}} void @ReturnPassing3(ptr dead_on_unwind noalias writable sret
 // PPC64: define{{.*}} i127 @ReturnPassing3(
-// PPC32: define{{.*}} void @ReturnPassing3(ptr noalias sret
+// PPC32: define{{.*}} void @ReturnPassing3(ptr dead_on_unwind noalias writable sret
 // AARCH64: define{{.*}} i127 @ReturnPassing3(
 // AARCH64DARWIN: define{{.*}} i127 @ReturnPassing3(
-// ARM: define{{.*}} arm_aapcscc void @ReturnPassing3(ptr noalias sret
+// ARM: define{{.*}} arm_aapcscc void @ReturnPassing3(ptr dead_on_unwind noalias writable sret
 // LA64: define{{.*}} i127 @ReturnPassing3(
-// LA32: define{{.*}} void @ReturnPassing3(ptr noalias sret
+// LA32: define{{.*}} void @ReturnPassing3(ptr dead_on_unwind noalias writable sret
 
 _BitInt(128) ReturnPassing4(void){}
 // LIN64: define{{.*}} { i64, i64 } @ReturnPassing4(
-// WIN64: define dso_local void @ReturnPassing4(ptr noalias sret
-// LIN32: define{{.*}} void @ReturnPassing4(ptr noalias sret
-// WIN32: define dso_local void @ReturnPassing4(ptr noalias sret
-// NACL: define{{.*}} void @ReturnPassing4(ptr noalias sret
+// WIN64: define dso_local void @ReturnPassing4(ptr dead_on_unwind noalias writable sret
+// LIN32: define{{.*}} void @ReturnPassing4(ptr dead_on_unwind noalias writable sret
+// WIN32: define dso_local void @ReturnPassing4(ptr dead_on_unwind noalias writable sret
+// NACL: define{{.*}} void @ReturnPassing4(ptr dead_on_unwind noalias writable sret
 // NVPTX64: define{{.*}} i128 @ReturnPassing4(
 // NVPTX: define{{.*}} i128 @ReturnPassing4(
 // SPARCV9: define{{.*}} i128 @ReturnPassing4(
-// SPARC: define{{.*}} void @ReturnPassing4(ptr noalias sret
+// SPARC: define{{.*}} void @ReturnPassing4(ptr dead_on_unwind noalias writable sret
 // MIPS64: define{{.*}} i128 @ReturnPassing4(
-// MIPS: define{{.*}} void @ReturnPassing4(ptr noalias sret
-// SPIR64: define{{.*}} spir_func void @ReturnPassing4(ptr noalias sret
-// SPIR: define{{.*}} spir_func void @ReturnPassing4(ptr noalias sret
-// HEX: define{{.*}} void @ReturnPassing4(ptr noalias sret
-// LANAI: define{{.*}} void @ReturnPassing4(ptr noalias sret
-// R600: define{{.*}} void @ReturnPassing4(ptr addrspace(5) noalias sret
-// ARC: define{{.*}} void @ReturnPassing4(ptr noalias sret
-// XCORE: define{{.*}} void @ReturnPassing4(ptr noalias sret
+// MIPS: define{{.*}} void @ReturnPassing4(ptr dead_on_unwind noalias writable sret
+// SPIR64: define{{.*}} spir_func void @ReturnPassing4(ptr dead_on_unwind noalias writable sret
+// SPIR: define{{.*}} spir_func void @ReturnPassing4(ptr dead_on_unwind noalias writable sret
+// HEX: define{{.*}} void @ReturnPassing4(ptr dead_on_unwind noalias writable sret
+// LANAI: define{{.*}} void @ReturnPassing4(ptr dead_on_unwind noalias writable sret
+// R600: define{{.*}} void @ReturnPassing4(ptr addrspace(5) dead_on_unwind noalias writable sret
+// ARC: define{{.*}} void @ReturnPassing4(ptr dead_on_unwind noalias writable sret
+// XCORE: define{{.*}} void @ReturnPassing4(ptr dead_on_unwind noalias writable sret
 // RISCV64: define{{.*}} i128 @ReturnPassing4(
-// RISCV32: define{{.*}} void @ReturnPassing4(ptr noalias sret
+// RISCV32: define{{.*}} void @ReturnPassing4(ptr dead_on_unwind noalias writable sret
 // WASM: define{{.*}} i128 @ReturnPassing4(
-// SYSTEMZ: define{{.*}} void @ReturnPassing4(ptr noalias sret
+// SYSTEMZ: define{{.*}} void @ReturnPassing4(ptr dead_on_unwind noalias writable sret
 // PPC64: define{{.*}} i128 @ReturnPassing4(
-// PPC32: define{{.*}} void @ReturnPassing4(ptr noalias sret
+// PPC32: define{{.*}} void @ReturnPassing4(ptr dead_on_unwind noalias writable sret
 // AARCH64: define{{.*}} i128 @ReturnPassing4(
 // AARCH64DARWIN: define{{.*}} i128 @ReturnPassing4(
-// ARM: define{{.*}} arm_aapcscc void @ReturnPassing4(ptr noalias sret
+// ARM: define{{.*}} arm_aapcscc void @ReturnPassing4(ptr dead_on_unwind noalias writable sret
 // LA64: define{{.*}} i128 @ReturnPassing4(
-// LA32: define{{.*}} void @ReturnPassing4(ptr noalias sret
+// LA32: define{{.*}} void @ReturnPassing4(ptr dead_on_unwind noalias writable sret
 
 #if __BITINT_MAXWIDTH__ > 128
 _BitInt(129) ReturnPassing5(void){}
-// LIN64: define{{.*}} void @ReturnPassing5(ptr noalias sret
-// WIN64: define dso_local void @ReturnPassing5(ptr noalias sret
-// LIN32: define{{.*}} void @ReturnPassing5(ptr noalias sret
-// WIN32: define dso_local void @ReturnPassing5(ptr noalias sret
-// NACL-NOT: define{{.*}} void @ReturnPassing5(ptr noalias sret
+// LIN64: define{{.*}} void @ReturnPassing5(ptr dead_on_unwind noalias writable sret
+// WIN64: define dso_local void @ReturnPassing5(ptr dead_on_unwind noalias writable sret
+// LIN32: define{{.*}} void @ReturnPassing5(ptr dead_on_unwind noalias writable sret
+// WIN32: define dso_local void @ReturnPassing5(ptr dead_on_unwind noalias writable sret
+// NACL-NOT: define{{.*}} void @ReturnPassing5(ptr dead_on_unwind noalias writable sret
 // NVPTX64-NOT: define{{.*}} i129 @ReturnPassing5(
 // NVPTX-NOT: define{{.*}} i129 @ReturnPassing5(
 // SPARCV9-NOT: define{{.*}} i129 @ReturnPassing5(
-// SPARC-NOT: define{{.*}} void @ReturnPassing5(ptr noalias sret
-// MIPS64-NOT: define{{.*}} void @ReturnPassing5(ptr noalias sret
-// MIPS-NOT: define{{.*}} void @ReturnPassing5(ptr noalias sret
-// SPIR64-NOT: define{{.*}} spir_func void @ReturnPassing5(ptr noalias sret
-// SPIR-NOT: define{{.*}} spir_func void @ReturnPassing5(ptr noalias sret
-// HEX-NOT: define{{.*}} void @ReturnPassing5(ptr noalias sret
-// LANAI-NOT: define{{.*}} void @ReturnPassing5(ptr noalias sret
+// SPARC-NOT: define{{.*}} void @ReturnPassing5(ptr dead_on_unwind noalias writable sret
+// MIPS64-NOT: define{{.*}} void @ReturnPassing5(ptr dead_on_unwind noalias writable sret
+// MIPS-NOT: define{{.*}} void @ReturnPassing5(ptr dead_on_unwind noalias writable sret
+// SPIR64-NOT: define{{.*}} spir_func void @ReturnPassing5(ptr dead_on_unwind noalias writable sret
+// SPIR-NOT: define{{.*}} spir_func void @ReturnPassing5(ptr dead_on_unwind noalias writable sret
+// HEX-NOT: define{{.*}} void @ReturnPassing5(ptr dead_on_unwind noalias writable sret
+// LANAI-NOT: define{{.*}} void @ReturnPassing5(ptr dead_on_unwind noalias writable sret
 // R600-NOT: define{{.*}} void @ReturnPassing5(ptr addrspace(5) noalias sret
-// ARC-NOT: define{{.*}} void @ReturnPassing5(ptr inreg noalias sret
-// XCORE-NOT: define{{.*}} void @ReturnPassing5(ptr noalias sret
-// RISCV64-NOT: define{{.*}} void @ReturnPassing5(ptr noalias sret
-// RISCV32-NOT: define{{.*}} void @ReturnPassing5(ptr noalias sret
-// WASM-NOT: define{{.*}} void @ReturnPassing5(ptr noalias sret
-// SYSTEMZ-NOT: define{{.*}} void @ReturnPassing5(ptr noalias sret
-// PPC64-NOT: define{{.*}} void @ReturnPassing5(ptr noalias sret
-// PPC32-NOT: define{{.*}} void @ReturnPassing5(ptr noalias sret
-// AARCH64-NOT: define{{.*}} void @ReturnPassing5(ptr noalias sret
-// AARCH64DARWIN-NOT: define{{.*}} void @ReturnPassing5(ptr noalias sret
-// ARM-NOT: define{{.*}} arm_aapcscc void @ReturnPassing5(ptr noalias sret
-// LA64-NOT: define{{.*}} void @ReturnPassing5(ptr noalias sret
-// LA32-NOT: define{{.*}} void @ReturnPassing5(ptr noalias sret
+// ARC-NOT: define{{.*}} void @ReturnPassing5(ptr dead_on_unwind inreg noalias writable sret
+// XCORE-NOT: define{{.*}} void @ReturnPassing5(ptr dead_on_unwind noalias writable sret
+// RISCV64-NOT: define{{.*}} void @ReturnPassing5(ptr dead_on_unwind noalias writable sret
+// RISCV32-NOT: define{{.*}} void @ReturnPassing5(ptr dead_on_unwind noalias writable sret
+// WASM-NOT: define{{.*}} void @ReturnPassing5(ptr dead_on_unwind noalias writable sret
+// SYSTEMZ-NOT: define{{.*}} void @ReturnPassing5(ptr dead_on_unwind noalias writable sret
+// PPC64-NOT: define{{.*}} void @ReturnPassing5(ptr dead_on_unwind noalias writable sret
+// PPC32-NOT: define{{.*}} void @ReturnPassing5(ptr dead_on_unwind noalias writable sret
+// AARCH64-NOT: define{{.*}} void @ReturnPassing5(ptr dead_on_unwind noalias writable sret
+// AARCH64DARWIN-NOT: define{{.*}} void @ReturnPassing5(ptr dead_on_unwind noalias writable sret
+// ARM-NOT: define{{.*}} arm_aapcscc void @ReturnPassing5(ptr dead_on_unwind noalias writable sret
+// LA64-NOT: define{{.*}} void @ReturnPassing5(ptr dead_on_unwind noalias writable sret
+// LA32-NOT: define{{.*}} void @ReturnPassing5(ptr dead_on_unwind noalias writable sret
 
 // SparcV9 is odd in that it has a return-size limit of 256, not 128 or 64
 // like other platforms, so test to make sure this behavior will still work.
 _BitInt(256) ReturnPassing6(void) {}
 // SPARCV9-NOT: define{{.*}} i256 @ReturnPassing6(
 _BitInt(257) ReturnPassing7(void) {}
-// SPARCV9-NOT: define{{.*}} void @ReturnPassing7(ptr noalias sret
+// SPARCV9-NOT: define{{.*}} void @ReturnPassing7(ptr dead_on_unwind noalias writable sret
 #endif
diff --git a/clang/test/CodeGen/isfpclass.c b/clang/test/CodeGen/isfpclass.c
index 430f5d94211f1..6633db88f71a8 100644
--- a/clang/test/CodeGen/isfpclass.c
+++ b/clang/test/CodeGen/isfpclass.c
@@ -160,7 +160,7 @@ int4 check_isfpclass_nan_strict_v4f32(float4 x) {
 }
 
 // CHECK-LABEL: define dso_local void @check_isfpclass_nan_v4f64
-// CHECK-SAME: (ptr noalias nocapture writeonly sret(<4 x i64>) align 16 [[AGG_RESULT:%.*]], ptr nocapture noundef readonly [[TMP0:%.*]]) local_unnamed_addr #[[ATTR3:[0-9]+]] {
+// CHECK-SAME: (ptr dead_on_unwind noalias nocapture writable writeonly sret(<4 x i64>) align 16 [[AGG_RESULT:%.*]], ptr nocapture noundef readonly [[TMP0:%.*]]) local_unnamed_addr #[[ATTR3:[0-9]+]] {
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[X:%.*]] = load <4 x double>, ptr [[TMP0]], align 16, !tbaa [[TBAA2:![0-9]+]]
 // CHECK-NEXT:    [[TMP1:%.*]] = fcmp uno <4 x double> [[X]], zeroinitializer
diff --git a/clang/test/CodeGen/lanai-arguments.c b/clang/test/CodeGen/lanai-arguments.c
index bc291faa867f5..bfd3b41a91eb2 100644
--- a/clang/test/CodeGen/lanai-arguments.c
+++ b/clang/test/CodeGen/lanai-arguments.c
@@ -16,7 +16,7 @@ void f1(s1 i) {}
 typedef struct {
   int cc;
 } s2;
-// CHECK: define{{.*}} void @f2(ptr noalias sret(%struct.s2) align 4 %agg.result)
+// CHECK: define{{.*}} void @f2(ptr dead_on_unwind noalias writable sret(%struct.s2) align 4 %agg.result)
 s2 f2(void) {
   s2 foo;
   return foo;
@@ -26,7 +26,7 @@ typedef struct {
   int cc;
   int dd;
 } s3;
-// CHECK: define{{.*}} void @f3(ptr noalias sret(%struct.s3) align 4 %agg.result)
+// CHECK: define{{.*}} void @f3(ptr dead_on_unwind noalias writable sret(%struct.s3) align 4 %agg.result)
 s3 f3(void) {
   s3 foo;
   return foo;
diff --git a/clang/test/CodeGen/large-data-threshold.c b/clang/test/CodeGen/large-data-threshold.c
index 29ae19e9b7189..d110ad2125c7b 100644
--- a/clang/test/CodeGen/large-data-threshold.c
+++ b/clang/test/CodeGen/large-data-threshold.c
@@ -2,11 +2,14 @@
 
 // RUN: %clang_cc1 -triple x86_64-unknown-unknown -emit-llvm %s -o - -mcmodel=medium | FileCheck %s --check-prefix=IR-DEFAULT
 // RUN: %clang_cc1 -triple x86_64-unknown-unknown -emit-llvm %s -o - -mcmodel=medium -mlarge-data-threshold=200 | FileCheck %s --check-prefix=IR-CUSTOM
+// RUN: %clang_cc1 -triple x86_64-unknown-unknown -emit-llvm %s -o - -mcmodel=large -mlarge-data-threshold=200 | FileCheck %s --check-prefix=IR-CUSTOM
+// RUN: %clang_cc1 -triple x86_64-unknown-unknown -emit-llvm %s -o - -mcmodel=small -mlarge-data-threshold=200 | FileCheck %s --check-prefix=IR-SMALL
 // RUN: %clang_cc1 -triple x86_64-unknown-unknown -S %s -o - -mcmodel=medium -mlarge-data-threshold=200 | FileCheck %s --check-prefix=ASM-SMALL
 // RUN: %clang_cc1 -triple x86_64-unknown-unknown -S %s -o - -mcmodel=medium -mlarge-data-threshold=2 | FileCheck %s --check-prefix=ASM-LARGE
 
-// IR-DEFAULT: !{i32 1, !"Large Data Threshold", i64 65535}
+// IR-DEFAULT: !{i32 1, !"Large Data Threshold", i64 0}
 // IR-CUSTOM: !{i32 1, !"Large Data Threshold", i64 200}
+// IR-SMALL-NOT: !"Large Data Threshold"
 
 // ASM-SMALL-NOT: movabsq
 // ASM-LARGE: movabsq
diff --git a/clang/test/CodeGen/mcu-struct-return.c b/clang/test/CodeGen/mcu-struct-return.c
index 218f9734c9c32..38a9bc2a36bfe 100644
--- a/clang/test/CodeGen/mcu-struct-return.c
+++ b/clang/test/CodeGen/mcu-struct-return.c
@@ -42,7 +42,7 @@ struct S1 bar1(void) { return s1; }
 struct S2 bar2(void) { return s2; }
 struct S1 bar3(union U1 u) { return s1; }
 // CHECK: define{{.*}} void @foo1()
-// CHECK: define{{.*}} void @foo2(ptr noalias sret([[UNION2_TYPE]]) align 4 %{{.+}})
+// CHECK: define{{.*}} void @foo2(ptr dead_on_unwind noalias writable sret([[UNION2_TYPE]]) align 4 %{{.+}})
 // CHECK: define{{.*}} i32 @foo3()
 // CHECK: define{{.*}} void @bar1()
 // CHECK: define{{.*}} i32 @bar2()
@@ -62,7 +62,7 @@ void run(void) {
   // CHECK: [[Y1:%.+]] = alloca [[STRUCT1_TYPE]]
   // CHECK: [[Y2:%.+]] = alloca [[STRUCT2_TYPE]]
   // CHECK: call void @foo1()
-  // CHECK: call void @foo2(ptr sret([[UNION2_TYPE]]) align 4 [[X2]])
+  // CHECK: call void @foo2(ptr dead_on_unwind writable sret([[UNION2_TYPE]]) align 4 [[X2]])
   // CHECK: {{.+}} = call i32 @foo3()
   // CHECK: call void @bar1()
   // CHECK: {{.+}} = call i32 @bar2()
diff --git a/clang/test/CodeGen/mingw-long-double.c b/clang/test/CodeGen/mingw-long-double.c
index a50f8f0b3b633..4be97526f9631 100644
--- a/clang/test/CodeGen/mingw-long-double.c
+++ b/clang/test/CodeGen/mingw-long-double.c
@@ -32,15 +32,15 @@ long double TestLD(long double x) {
   return x * x;
 }
 // GNU32: define dso_local x86_fp80 @TestLD(x86_fp80 noundef %x)
-// GNU64: define dso_local void @TestLD(ptr noalias sret(x86_fp80) align 16 %agg.result, ptr noundef %0)
+// GNU64: define dso_local void @TestLD(ptr dead_on_unwind noalias writable sret(x86_fp80) align 16 %agg.result, ptr noundef %0)
 // MSC64: define dso_local double @TestLD(double noundef %x)
 
 long double _Complex TestLDC(long double _Complex x) {
   return x * x;
 }
-// GNU32: define dso_local void @TestLDC(ptr noalias sret({ x86_fp80, x86_fp80 }) align 4 %agg.result, ptr noundef byval({ x86_fp80, x86_fp80 }) align 4 %x)
-// GNU64: define dso_local void @TestLDC(ptr noalias sret({ x86_fp80, x86_fp80 }) align 16 %agg.result, ptr noundef %x)
-// MSC64: define dso_local void @TestLDC(ptr noalias sret({ double, double }) align 8 %agg.result, ptr noundef %x)
+// GNU32: define dso_local void @TestLDC(ptr dead_on_unwind noalias writable sret({ x86_fp80, x86_fp80 }) align 4 %agg.result, ptr noundef byval({ x86_fp80, x86_fp80 }) align 4 %x)
+// GNU64: define dso_local void @TestLDC(ptr dead_on_unwind noalias writable sret({ x86_fp80, x86_fp80 }) align 16 %agg.result, ptr noundef %x)
+// MSC64: define dso_local void @TestLDC(ptr dead_on_unwind noalias writable sret({ double, double }) align 8 %agg.result, ptr noundef %x)
 
 // GNU32: declare dso_local void @__mulxc3
 // GNU64: declare dso_local void @__mulxc3
diff --git a/clang/test/CodeGen/mips-vector-return.c b/clang/test/CodeGen/mips-vector-return.c
index b6da2af2ffd66..d2103d1def745 100644
--- a/clang/test/CodeGen/mips-vector-return.c
+++ b/clang/test/CodeGen/mips-vector-return.c
@@ -8,14 +8,14 @@ typedef float  v4sf __attribute__ ((__vector_size__ (16)));
 typedef double v4df __attribute__ ((__vector_size__ (32)));
 typedef int v4i32 __attribute__ ((__vector_size__ (16)));
 
-// O32-LABEL: define{{.*}} void @test_v4sf(ptr noalias nocapture writeonly sret
+// O32-LABEL: define{{.*}} void @test_v4sf(ptr dead_on_unwind noalias nocapture writable writeonly sret
 // N64: define{{.*}} inreg { i64, i64 } @test_v4sf
 v4sf test_v4sf(float a) {
   return (v4sf){0.0f, a, 0.0f, 0.0f};
 }
 
-// O32-LABEL: define{{.*}} void @test_v4df(ptr noalias nocapture writeonly sret
-// N64-LABEL: define{{.*}} void @test_v4df(ptr noalias nocapture writeonly sret
+// O32-LABEL: define{{.*}} void @test_v4df(ptr dead_on_unwind noalias nocapture writable writeonly sret
+// N64-LABEL: define{{.*}} void @test_v4df(ptr dead_on_unwind noalias nocapture writable writeonly sret
 v4df test_v4df(double a) {
   return (v4df){0.0, a, 0.0, 0.0};
 }
diff --git a/clang/test/CodeGen/mips-zero-sized-struct.c b/clang/test/CodeGen/mips-zero-sized-struct.c
index 12bd9abf6cf4c..b40ff59f73fb2 100644
--- a/clang/test/CodeGen/mips-zero-sized-struct.c
+++ b/clang/test/CodeGen/mips-zero-sized-struct.c
@@ -19,7 +19,7 @@
 // RUN: %clang_cc1 -triple mipsisa64r6-unknown-linux-gnuabi64 -S -emit-llvm -o - %s | FileCheck -check-prefix=N64 %s
 // RUN: %clang_cc1 -triple mipsisa64r6el-unknown-linux-gnuabi64 -S -emit-llvm -o - %s | FileCheck -check-prefix=N64 %s
 
-// O32: define{{.*}} void @fn28(ptr noalias sret(%struct.T2) align 1 %agg.result, i8 noundef signext %arg0)
+// O32: define{{.*}} void @fn28(ptr dead_on_unwind noalias writable sret(%struct.T2) align 1 %agg.result, i8 noundef signext %arg0)
 // N32: define{{.*}} void @fn28(i8 noundef signext %arg0)
 // N64: define{{.*}} void @fn28(i8 noundef signext %arg0)
 
diff --git a/clang/test/CodeGen/mips64-nontrivial-return.cpp b/clang/test/CodeGen/mips64-nontrivial-return.cpp
index 0987dcaf355dd..a8fbf4622f803 100644
--- a/clang/test/CodeGen/mips64-nontrivial-return.cpp
+++ b/clang/test/CodeGen/mips64-nontrivial-return.cpp
@@ -10,7 +10,7 @@ class D : public B {
 
 extern D gd0;
 
-// CHECK: _Z4foo1v(ptr noalias nocapture writeonly sret
+// CHECK: _Z4foo1v(ptr dead_on_unwind noalias nocapture writable writeonly sret
 
 D foo1(void) {
   return gd0;
diff --git a/clang/test/CodeGen/mips64-padding-arg.c b/clang/test/CodeGen/mips64-padding-arg.c
index 7fda45f75098f..038103b1df3aa 100644
--- a/clang/test/CodeGen/mips64-padding-arg.c
+++ b/clang/test/CodeGen/mips64-padding-arg.c
@@ -33,9 +33,9 @@ void foo3(int a0, long double a1) {
 
 // Insert padding after hidden argument.
 //
-// N64-LABEL: define{{.*}} void @foo5(ptr noalias sret(%struct.S0) align 16 %agg.result, i64 %0, fp128 noundef %a0)
-// N64: call void @foo6(ptr sret(%struct.S0) align 16 %agg.result, i32 noundef signext 1, i32 noundef signext 2, i64 undef, fp128 noundef %a0)
-// N64: declare void @foo6(ptr sret(%struct.S0) align 16, i32 noundef signext, i32 noundef signext, i64, fp128 noundef)
+// N64-LABEL: define{{.*}} void @foo5(ptr dead_on_unwind noalias writable sret(%struct.S0) align 16 %agg.result, i64 %0, fp128 noundef %a0)
+// N64: call void @foo6(ptr dead_on_unwind writable sret(%struct.S0) align 16 %agg.result, i32 noundef signext 1, i32 noundef signext 2, i64 undef, fp128 noundef %a0)
+// N64: declare void @foo6(ptr dead_on_unwind writable sret(%struct.S0) align 16, i32 noundef signext, i32 noundef signext, i64, fp128 noundef)
 
 extern S0 foo6(int, int, long double);
 
diff --git a/clang/test/CodeGen/ms_abi.c b/clang/test/CodeGen/ms_abi.c
index adc5094267cbd..0fe1741cf9a0a 100644
--- a/clang/test/CodeGen/ms_abi.c
+++ b/clang/test/CodeGen/ms_abi.c
@@ -141,7 +141,7 @@ struct i128 {
 };
 
 __attribute__((ms_abi)) struct i128 f7(struct i128 a) {
-  // WIN64: define dso_local void @f7(ptr noalias sret(%struct.i128) align 8 %agg.result, ptr noundef %a)
-  // FREEBSD: define{{.*}} win64cc void @f7(ptr noalias sret(%struct.i128) align 8 %agg.result, ptr noundef %a)
+  // WIN64: define dso_local void @f7(ptr dead_on_unwind noalias writable sret(%struct.i128) align 8 %agg.result, ptr noundef %a)
+  // FREEBSD: define{{.*}} win64cc void @f7(ptr dead_on_unwind noalias writable sret(%struct.i128) align 8 %agg.result, ptr noundef %a)
   return a;
 }
diff --git a/clang/test/CodeGen/paren-list-agg-init.cpp b/clang/test/CodeGen/paren-list-agg-init.cpp
index 0e68beb5c3706..eb8d91a7f97f6 100644
--- a/clang/test/CodeGen/paren-list-agg-init.cpp
+++ b/clang/test/CodeGen/paren-list-agg-init.cpp
@@ -136,7 +136,7 @@ A foo1() {
   return a1;
 }
 
-// CHECK: define dso_local void @{{.*foo2.*}}(ptr noalias sret([[STRUCT_B]]) align 8 [[AGG_RESULT:%.*]])
+// CHECK: define dso_local void @{{.*foo2.*}}(ptr dead_on_unwind noalias writable sret([[STRUCT_B]]) align 8 [[AGG_RESULT:%.*]])
 // CHECK-NEXT: entry:
 // CHECK-NEXT: call void @llvm.memcpy.p0.p0.i64(ptr align 8 [[AGG_RESULT]], ptr align 8 [[B1]], i64 24, i1 false)
 // CHECK-NEXT: ret void
@@ -144,7 +144,7 @@ B foo2() {
   return b1;
 }
 
-// CHECK: define dso_local void @{{.*foo3.*}}(ptr noalias sret([[STRUCT_C]]) align 8 [[AGG_RESULT:%.*]])
+// CHECK: define dso_local void @{{.*foo3.*}}(ptr dead_on_unwind noalias writable sret([[STRUCT_C]]) align 8 [[AGG_RESULT:%.*]])
 // CHECK-NEXT: entry:
 // CHECK-NEXT: call void @llvm.memcpy.p0.p0.i64(ptr align 8 [[AGG_RESULT]], ptr align 8 [[C1]], i64 48, i1 false)
 // CHECK-NEXT: ret void
@@ -229,7 +229,7 @@ void foo7() {
   D d(A(1, 1), A(11, 11), A(111, 111));
 }
 
-// CHECK: dso_local void @{{.*foo8.*}}(ptr noalias sret([[STRUCT_D]]) align 8 [[AGG_RESULT:%.*]])
+// CHECK: dso_local void @{{.*foo8.*}}(ptr dead_on_unwind noalias writable sret([[STRUCT_D]]) align 8 [[AGG_RESULT:%.*]])
 // CHECK-NEXT: entry:
 // CHECK-NEXT: call void @llvm.memcpy.p0.p0.i64(ptr align 8 [[AGG_RESULT]], ptr align 8 [[D1]], i64 56, i1 false)
 // CHECK-NEXT: ret void
diff --git a/clang/test/CodeGen/regcall2.c b/clang/test/CodeGen/regcall2.c
index 96bc6615012af..c88d4e485b104 100644
--- a/clang/test/CodeGen/regcall2.c
+++ b/clang/test/CodeGen/regcall2.c
@@ -19,7 +19,7 @@ double __regcall bar(__sVector a) {
 }
 
 // FIXME: Do we need to change for Windows?
-// Win: define dso_local x86_regcallcc void @__regcall3__foo(ptr noalias sret(%struct.__sVector) align 64 %agg.result, i32 noundef %a) #0
+// Win: define dso_local x86_regcallcc void @__regcall3__foo(ptr dead_on_unwind noalias writable sret(%struct.__sVector) align 64 %agg.result, i32 noundef %a) #0
 // Win: define dso_local x86_regcallcc double @__regcall3__bar(ptr noundef %a) #0
 // Win: attributes #0 = { noinline nounwind optnone "min-legal-vector-width"="0" "no-builtins" "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-features"="+avx,+avx2,+avx512f,+avx512vl,+crc32,+cx8,+evex512,+f16c,+fma,+mmx,+popcnt,+sse,+sse2,+sse3,+sse4.1,+sse4.2,+ssse3,+x87,+xsave" }
 
diff --git a/clang/test/CodeGen/regparm-struct.c b/clang/test/CodeGen/regparm-struct.c
index 70901ca364d05..a533f6cedbba2 100644
--- a/clang/test/CodeGen/regparm-struct.c
+++ b/clang/test/CodeGen/regparm-struct.c
@@ -159,7 +159,7 @@ void g16(void) {
 }
 
 __attribute__((regparm(3))) struct s12 f17(int a, int b, int c);
-// CHECK: declare void @f17(ptr inreg sret(%struct.s12) align 4, i32 inreg noundef, i32 inreg noundef, i32 noundef)
+// CHECK: declare void @f17(ptr dead_on_unwind inreg writable sret(%struct.s12) align 4, i32 inreg noundef, i32 inreg noundef, i32 noundef)
 void g17(void) {
   f17(41, 42, 43);
 }
diff --git a/clang/test/CodeGen/renderscript.c b/clang/test/CodeGen/renderscript.c
index 8acf16566454c..1629665c1ffb8 100644
--- a/clang/test/CodeGen/renderscript.c
+++ b/clang/test/CodeGen/renderscript.c
@@ -83,15 +83,15 @@ void argLongInt(sLongInt s) {}
 // and coerced to [a x iNN] for 64-bit RenderScript
 // =============================================================================
 
-// CHECK-RS32: void @retShortCharShort(ptr noalias sret(%struct.sShortCharShort) align 2 %agg.result)
+// CHECK-RS32: void @retShortCharShort(ptr dead_on_unwind noalias writable sret(%struct.sShortCharShort) align 2 %agg.result)
 // CHECK-RS64: [3 x i16] @retShortCharShort()
 sShortCharShort retShortCharShort(void) { sShortCharShort r; return r; }
 
-// CHECK-RS32: void @retIntShortChar(ptr noalias sret(%struct.sIntShortChar) align 4 %agg.result)
+// CHECK-RS32: void @retIntShortChar(ptr dead_on_unwind noalias writable sret(%struct.sIntShortChar) align 4 %agg.result)
 // CHECK-RS64: [2 x i32] @retIntShortChar()
 sIntShortChar retIntShortChar(void) { sIntShortChar r; return r; }
 
-// CHECK-RS32: void @retLongInt(ptr noalias sret(%struct.sLongInt) align 8 %agg.result)
+// CHECK-RS32: void @retLongInt(ptr dead_on_unwind noalias writable sret(%struct.sLongInt) align 8 %agg.result)
 // CHECK-RS64: [2 x i64] @retLongInt()
 sLongInt retLongInt(void) { sLongInt r; return r; }
 
@@ -116,12 +116,12 @@ void argLong2Char(sLong2Char s) {}
 // 64-bit RenderScript
 // =============================================================================
 
-// CHECK-RS32: void @retInt5(ptr noalias sret(%struct.sInt5) align 4 %agg.result)
-// CHECK-RS64: void @retInt5(ptr noalias sret(%struct.sInt5) align 4 %agg.result)
+// CHECK-RS32: void @retInt5(ptr dead_on_unwind noalias writable sret(%struct.sInt5) align 4 %agg.result)
+// CHECK-RS64: void @retInt5(ptr dead_on_unwind noalias writable sret(%struct.sInt5) align 4 %agg.result)
 sInt5 retInt5(void) { sInt5 r; return r;}
 
-// CHECK-RS32: void @retLong2Char(ptr noalias sret(%struct.sLong2Char) align 8 %agg.result)
-// CHECK-RS64: void @retLong2Char(ptr noalias sret(%struct.sLong2Char) align 8 %agg.result)
+// CHECK-RS32: void @retLong2Char(ptr dead_on_unwind noalias writable sret(%struct.sLong2Char) align 8 %agg.result)
+// CHECK-RS64: void @retLong2Char(ptr dead_on_unwind noalias writable sret(%struct.sLong2Char) align 8 %agg.result)
 sLong2Char retLong2Char(void) { sLong2Char r; return r;}
 
 // =============================================================================
@@ -135,6 +135,6 @@ typedef struct {long l1, l2, l3, l4, l5, l6, l7, l8, l9; } sLong9;
 // CHECK-RS64: void @argLong9(ptr noundef %s)
 void argLong9(sLong9 s) {}
 
-// CHECK-RS32: void @retLong9(ptr noalias sret(%struct.sLong9) align 8 %agg.result)
-// CHECK-RS64: void @retLong9(ptr noalias sret(%struct.sLong9) align 8 %agg.result)
+// CHECK-RS32: void @retLong9(ptr dead_on_unwind noalias writable sret(%struct.sLong9) align 8 %agg.result)
+// CHECK-RS64: void @retLong9(ptr dead_on_unwind noalias writable sret(%struct.sLong9) align 8 %agg.result)
 sLong9 retLong9(void) { sLong9 r; return r; }
diff --git a/clang/test/CodeGen/sparcv9-abi.c b/clang/test/CodeGen/sparcv9-abi.c
index 2d39c64c1c2ce..5e74a9a883cef 100644
--- a/clang/test/CodeGen/sparcv9-abi.c
+++ b/clang/test/CodeGen/sparcv9-abi.c
@@ -53,7 +53,7 @@ struct large {
   int x;
 };
 
-// CHECK-LABEL: define{{.*}} void @f_large(ptr noalias sret(%struct.large) align 8 %agg.result, ptr noundef %x)
+// CHECK-LABEL: define{{.*}} void @f_large(ptr dead_on_unwind noalias writable sret(%struct.large) align 8 %agg.result, ptr noundef %x)
 struct large f_large(struct large x) {
   x.a += *x.b;
   x.b = 0;
diff --git a/clang/test/CodeGen/sret.c b/clang/test/CodeGen/sret.c
index 548581d3e3dc6..6d905e89b2c6f 100644
--- a/clang/test/CodeGen/sret.c
+++ b/clang/test/CodeGen/sret.c
@@ -9,15 +9,15 @@ struct abc {
 };
 
 struct abc foo1(void);
-// CHECK-DAG: declare {{.*}} @foo1(ptr sret(%struct.abc)
+// CHECK-DAG: declare {{.*}} @foo1(ptr dead_on_unwind writable sret(%struct.abc)
 struct abc foo2();
-// CHECK-DAG: declare {{.*}} @foo2(ptr sret(%struct.abc)
+// CHECK-DAG: declare {{.*}} @foo2(ptr dead_on_unwind writable sret(%struct.abc)
 struct abc foo3(void){}
-// CHECK-DAG: define {{.*}} @foo3(ptr noalias sret(%struct.abc)
+// CHECK-DAG: define {{.*}} @foo3(ptr dead_on_unwind noalias writable sret(%struct.abc)
 
 void bar(void) {
   struct abc dummy1 = foo1();
-  // CHECK-DAG: call {{.*}} @foo1(ptr sret(%struct.abc)
+  // CHECK-DAG: call {{.*}} @foo1(ptr dead_on_unwind writable sret(%struct.abc)
   struct abc dummy2 = foo2();
-  // CHECK-DAG: call {{.*}} @foo2(ptr sret(%struct.abc)
+  // CHECK-DAG: call {{.*}} @foo2(ptr dead_on_unwind writable sret(%struct.abc)
 }
diff --git a/clang/test/CodeGen/target-builtin-noerror.c b/clang/test/CodeGen/target-builtin-noerror.c
index 505f4a3e94565..06bb40a2b71ea 100644
--- a/clang/test/CodeGen/target-builtin-noerror.c
+++ b/clang/test/CodeGen/target-builtin-noerror.c
@@ -82,6 +82,8 @@ void verifyfeaturestrings(void) {
   (void)__builtin_cpu_supports("avx512bitalg");
   (void)__builtin_cpu_supports("avx512bf16");
   (void)__builtin_cpu_supports("avx512vp2intersect");
+  (void)__builtin_cpu_supports("f16c");
+  (void)__builtin_cpu_supports("avx512fp16");
 }
 
 void verifycpustrings(void) {
diff --git a/clang/test/CodeGen/thinlto-debug-names-tu-reuse.ll b/clang/test/CodeGen/thinlto-debug-names-tu-reuse.ll
new file mode 100644
index 0000000000000..8719473820053
--- /dev/null
+++ b/clang/test/CodeGen/thinlto-debug-names-tu-reuse.ll
@@ -0,0 +1,59 @@
+; REQUIRES: asserts
+; REQUIRES: x86-registered-target
+
+;; Tests that accelerator table switches correctly from TU to CU when a top level TU is re-used.
+;; Assert is not triggered.
+;; File1
+;; struct Foo {
+;;   char fChar;
+;; };
+;; Foo fGlobal;
+;; FIle2
+;; struct Foo {
+;;   char fChar;
+;; };
+;; Foo fGlobal2;
+;; clang++ .cpp -O0 -g2 -fdebug-types-section -gpubnames -S -emit-llvm -o .ll
+;; llvm-link file1.ll file2.ll -S -o thinlto-debug-names-tu-reuse.ll
+
+; RUN: llc -O0 -dwarf-version=5 -generate-type-units -filetype=obj < %s -o %t.o
+; RUN: llvm-readelf --sections %t.o | FileCheck --check-prefix=OBJ %s
+
+; OBJ: debug_names
+
+; ModuleID = 'llvm-link'
+source_filename = "llvm-link"
+target datalayout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-i128:128-f80:128-n8:16:32:64-S128"
+target triple = "x86_64-unknown-linux-gnu"
+
+%struct.Foo = type { i8 }
+
+@fGlobal = dso_local global %struct.Foo zeroinitializer, align 1, !dbg !0
+@fGlobal2 = dso_local global %struct.Foo zeroinitializer, align 1, !dbg !9
+
+!llvm.dbg.cu = !{!2, !11}
+!llvm.ident = !{!14, !14}
+!llvm.module.flags = !{!15, !16, !17, !18, !19, !20, !21}
+
+!0 = !DIGlobalVariableExpression(var: !1, expr: !DIExpression())
+!1 = distinct !DIGlobalVariable(name: "fGlobal", scope: !2, file: !3, line: 5, type: !5, isLocal: false, isDefinition: true)
+!2 = distinct !DICompileUnit(language: DW_LANG_C_plus_plus_14, file: !3, producer: "clang version 18.0.0git", isOptimized: false, runtimeVersion: 0, emissionKind: FullDebug, globals: !4, splitDebugInlining: false)
+!3 = !DIFile(filename: "main.cpp", directory: "/smallTUReuse", checksumkind: CSK_MD5, checksum: "4f1831504f0948b03880356fae49cb58")
+!4 = !{!0}
+!5 = distinct !DICompositeType(tag: DW_TAG_structure_type, name: "Foo", file: !3, line: 2, size: 8, flags: DIFlagTypePassByValue, elements: !6, identifier: "_ZTS3Foo")
+!6 = !{!7}
+!7 = !DIDerivedType(tag: DW_TAG_member, name: "fChar", scope: !5, file: !3, line: 3, baseType: !8, size: 8)
+!8 = !DIBasicType(name: "char", size: 8, encoding: DW_ATE_signed_char)
+!9 = !DIGlobalVariableExpression(var: !10, expr: !DIExpression())
+!10 = distinct !DIGlobalVariable(name: "fGlobal2", scope: !11, file: !12, line: 5, type: !5, isLocal: false, isDefinition: true)
+!11 = distinct !DICompileUnit(language: DW_LANG_C_plus_plus_14, file: !12, producer: "clang version 18.0.0git", isOptimized: false, runtimeVersion: 0, emissionKind: FullDebug, globals: !13, splitDebugInlining: false)
+!12 = !DIFile(filename: "helper.cpp", directory: "/smallTUReuse", checksumkind: CSK_MD5, checksum: "014145d46991fd1eb6a2192d382feb75")
+!13 = !{!9}
+!14 = !{!"clang version 18.0.0git"}
+!15 = !{i32 7, !"Dwarf Version", i32 5}
+!16 = !{i32 2, !"Debug Info Version", i32 3}
+!17 = !{i32 1, !"wchar_size", i32 4}
+!18 = !{i32 8, !"PIC Level", i32 2}
+!19 = !{i32 7, !"PIE Level", i32 2}
+!20 = !{i32 7, !"uwtable", i32 2}
+!21 = !{i32 7, !"frame-pointer", i32 2}
diff --git a/clang/test/CodeGen/vectorcall.c b/clang/test/CodeGen/vectorcall.c
index b5f7efec67511..cb53ecc70351d 100644
--- a/clang/test/CodeGen/vectorcall.c
+++ b/clang/test/CodeGen/vectorcall.c
@@ -90,8 +90,8 @@ struct HVA4 __vectorcall hva6(struct HVA4 a, struct HVA4 b) { return b;}
 // X64: define dso_local x86_vectorcallcc %struct.HVA4 @"\01hva6@@128"(%struct.HVA4 inreg %a.coerce, ptr noundef %b)
 
 struct HVA5 __vectorcall hva7(void) {struct HVA5 a = {}; return a;}
-// X86: define dso_local x86_vectorcallcc void @"\01hva7@@0"(ptr inreg noalias sret(%struct.HVA5) align 16 %agg.result)
-// X64: define dso_local x86_vectorcallcc void @"\01hva7@@0"(ptr noalias sret(%struct.HVA5) align 16 %agg.result)
+// X86: define dso_local x86_vectorcallcc void @"\01hva7@@0"(ptr dead_on_unwind inreg noalias writable sret(%struct.HVA5) align 16 %agg.result)
+// X64: define dso_local x86_vectorcallcc void @"\01hva7@@0"(ptr dead_on_unwind noalias writable sret(%struct.HVA5) align 16 %agg.result)
 
 v4f32 __vectorcall hva8(v4f32 a, v4f32 b, v4f32 c, v4f32 d, int e, v4f32 f) {return f;}
 // X86: define dso_local x86_vectorcallcc <4 x float> @"\01hva8@@84"(<4 x float> inreg noundef %a, <4 x float> inreg noundef %b, <4 x float> inreg noundef %c, <4 x float> inreg noundef %d, i32 inreg noundef %e, <4 x float> inreg noundef %f)
diff --git a/clang/test/CodeGen/windows-struct-abi.c b/clang/test/CodeGen/windows-struct-abi.c
index 4431d91b63ca5..5e63c5b3344dd 100644
--- a/clang/test/CodeGen/windows-struct-abi.c
+++ b/clang/test/CodeGen/windows-struct-abi.c
@@ -34,7 +34,7 @@ struct f4 {
 
 struct f4 return_f4(void) { while (1); }
 
-// CHECK: define dso_local void @return_f4(ptr noalias sret(%struct.f4) align 4 %agg.result)
+// CHECK: define dso_local void @return_f4(ptr dead_on_unwind noalias writable sret(%struct.f4) align 4 %agg.result)
 
 void receive_f4(struct f4 a0) { }
 
diff --git a/clang/test/CodeGen/windows-swiftcall.c b/clang/test/CodeGen/windows-swiftcall.c
index 3e5c8a4d4b9d7..6c138a341835d 100644
--- a/clang/test/CodeGen/windows-swiftcall.c
+++ b/clang/test/CodeGen/windows-swiftcall.c
@@ -20,7 +20,7 @@ SWIFTCALL int indirect_result_2(OUT int *arg0, OUT float *arg1) {  __builtin_unr
 
 typedef struct { char array[1024]; } struct_reallybig;
 SWIFTCALL struct_reallybig indirect_result_3(OUT int *arg0, OUT float *arg1) { __builtin_unreachable(); }
-// CHECK-LABEL: define {{.*}} void @indirect_result_3(ptr noalias sret({{.*}}) {{.*}}, ptr noalias noundef align 4 dereferenceable(4){{.*}}, ptr noalias noundef align 4 dereferenceable(4){{.*}})
+// CHECK-LABEL: define {{.*}} void @indirect_result_3(ptr dead_on_unwind noalias writable sret({{.*}}) {{.*}}, ptr noalias noundef align 4 dereferenceable(4){{.*}}, ptr noalias noundef align 4 dereferenceable(4){{.*}})
 
 SWIFTCALL void context_1(CONTEXT void *self) {}
 // CHECK-LABEL: define {{.*}} void @context_1(ptr noundef swiftself
@@ -218,7 +218,7 @@ typedef struct {
 } struct_big_1;
 TEST(struct_big_1)
 
-// CHECK-LABEL: define {{.*}} void @return_struct_big_1({{.*}} noalias sret
+// CHECK-LABEL: define {{.*}} void @return_struct_big_1({{.*}} dead_on_unwind noalias writable sret
 
 // Should not be byval.
 // CHECK-LABEL: define {{.*}} void @take_struct_big_1(ptr noundef{{( %.*)?}})
diff --git a/clang/test/CodeGenCXX/aix-alignment.cpp b/clang/test/CodeGenCXX/aix-alignment.cpp
index a8bb95814b1a4..d4397f3120ec5 100644
--- a/clang/test/CodeGenCXX/aix-alignment.cpp
+++ b/clang/test/CodeGenCXX/aix-alignment.cpp
@@ -29,7 +29,7 @@ typedef struct D {
   ~D(){};
 } D;
 
-// AIX: define void @_Z3foo1D(ptr noalias sret(%struct.D) align 4 %agg.result, ptr noundef %x)
+// AIX: define void @_Z3foo1D(ptr dead_on_unwind noalias writable sret(%struct.D) align 4 %agg.result, ptr noundef %x)
 // AIX32  call void @llvm.memcpy.p0.p0.i32(ptr align 4 %agg.result, ptr align 4 %x, i32 16, i1 false)
 // AIX64: call void @llvm.memcpy.p0.p0.i64(ptr align 4 %agg.result, ptr align 4 %x, i64 16, i1 false)
 D foo(D x) { return x; }
diff --git a/clang/test/CodeGenCXX/arm-cc.cpp b/clang/test/CodeGenCXX/arm-cc.cpp
index 967d10096f72b..68e1b7e4e1e46 100644
--- a/clang/test/CodeGenCXX/arm-cc.cpp
+++ b/clang/test/CodeGenCXX/arm-cc.cpp
@@ -16,5 +16,5 @@ void baz() {
   zed(a);
 }
 
-// CHECK: declare void @_Z3fooPv(ptr sret(%class.SMLoc) align 4, ptr noundef)
+// CHECK: declare void @_Z3fooPv(ptr dead_on_unwind writable sret(%class.SMLoc) align 4, ptr noundef)
 // CHECK: declare void @_Z3zed5SMLoc(ptr noundef)
diff --git a/clang/test/CodeGenCXX/arm-swiftcall.cpp b/clang/test/CodeGenCXX/arm-swiftcall.cpp
index 687e4c8b4dcbe..e60c1482700a4 100644
--- a/clang/test/CodeGenCXX/arm-swiftcall.cpp
+++ b/clang/test/CodeGenCXX/arm-swiftcall.cpp
@@ -105,7 +105,7 @@ struct struct_indirect_1 {
 };
 TEST(struct_indirect_1)
 
-// CHECK-LABEL: define {{.*}} void @return_struct_indirect_1({{.*}} noalias sret
+// CHECK-LABEL: define {{.*}} void @return_struct_indirect_1({{.*}} dead_on_unwind noalias writable sret
 
 // Should not be byval.
 // CHECK-LABEL: define {{.*}} void @take_struct_indirect_1(ptr noundef{{( %.*)?}})
diff --git a/clang/test/CodeGenCXX/attr-cpuspecific-outoflinedefs.cpp b/clang/test/CodeGenCXX/attr-cpuspecific-outoflinedefs.cpp
index ef2498bd7e14c..183eb4fb6ac61 100644
--- a/clang/test/CodeGenCXX/attr-cpuspecific-outoflinedefs.cpp
+++ b/clang/test/CodeGenCXX/attr-cpuspecific-outoflinedefs.cpp
@@ -80,8 +80,8 @@ OutOfLineDefs::foo(int, int, int) {
 // LINUX: define dso_local noundef i32 @_ZN13OutOfLineDefs3fooEiii.S
 // LINUX: define dso_local noundef i32 @_ZN13OutOfLineDefs3fooEiii.R
 // LINUX: define weak_odr ptr @_ZN13OutOfLineDefs3fooEiii.resolver()
-// LINUX: ret ptr @_ZN13OutOfLineDefs3fooEiii.R
 // LINUX: ret ptr @_ZN13OutOfLineDefs3fooEiii.S
+// LINUX: ret ptr @_ZN13OutOfLineDefs3fooEiii.R
 // LINUX: ret ptr @_ZN13OutOfLineDefs3fooEiii.O
 // LINUX: call void @llvm.trap
 // LINUX: define linkonce_odr noundef i32 @_ZN13OutOfLineDefs3fooEiii.O
@@ -89,8 +89,8 @@ OutOfLineDefs::foo(int, int, int) {
 // WINDOWS: define dso_local noundef i32 @"?foo@OutOfLineDefs@@QEAAHHHH@Z.S"
 // WINDOWS: define dso_local noundef i32 @"?foo@OutOfLineDefs@@QEAAHHHH@Z.R"
 // WINDOWS: define weak_odr dso_local i32 @"?foo@OutOfLineDefs@@QEAAHHHH@Z"(ptr %0, i32 %1, i32 %2, i32 %3)
-// WINDOWS: musttail call i32 @"?foo@OutOfLineDefs@@QEAAHHHH@Z.R"(ptr %0, i32 %1, i32 %2, i32 %3)
 // WINDOWS: musttail call i32 @"?foo@OutOfLineDefs@@QEAAHHHH@Z.S"(ptr %0, i32 %1, i32 %2, i32 %3)
+// WINDOWS: musttail call i32 @"?foo@OutOfLineDefs@@QEAAHHHH@Z.R"(ptr %0, i32 %1, i32 %2, i32 %3)
 // WINDOWS: musttail call i32 @"?foo@OutOfLineDefs@@QEAAHHHH@Z.O"(ptr %0, i32 %1, i32 %2, i32 %3)
 // WINDOWS: call void @llvm.trap
 // WINDOWS: define linkonce_odr dso_local noundef i32 @"?foo@OutOfLineDefs@@QEAAHHHH@Z.O"
diff --git a/clang/test/CodeGenCXX/attr-musttail.cpp b/clang/test/CodeGenCXX/attr-musttail.cpp
index b8a1961d5eb77..720e50c5a240f 100644
--- a/clang/test/CodeGenCXX/attr-musttail.cpp
+++ b/clang/test/CodeGenCXX/attr-musttail.cpp
@@ -162,7 +162,7 @@ HasNonTrivialCopyConstructor TestNonElidableCopyConstructor() {
   [[clang::musttail]] return (((ReturnsClassByValue())));
 }
 
-// CHECK: musttail call void @_Z19ReturnsClassByValuev(ptr sret(%struct.HasNonTrivialCopyConstructor) align 1 %agg.result)
+// CHECK: musttail call void @_Z19ReturnsClassByValuev(ptr dead_on_unwind writable sret(%struct.HasNonTrivialCopyConstructor) align 1 %agg.result)
 
 struct HasNonTrivialCopyConstructor2 {
   // Copy constructor works even if it has extra default params.
@@ -191,8 +191,8 @@ LargeWithCopyConstructor TestLargeWithCopyConstructor() {
   [[clang::musttail]] return ReturnsLarge();
 }
 
-// CHECK: define dso_local void @_Z28TestLargeWithCopyConstructorv(ptr noalias sret(%struct.LargeWithCopyConstructor) align 1 %agg.result)
-// CHECK: musttail call void @_Z12ReturnsLargev(ptr sret(%struct.LargeWithCopyConstructor) align 1 %agg.result)
+// CHECK: define dso_local void @_Z28TestLargeWithCopyConstructorv(ptr dead_on_unwind noalias writable sret(%struct.LargeWithCopyConstructor) align 1 %agg.result)
+// CHECK: musttail call void @_Z12ReturnsLargev(ptr dead_on_unwind writable sret(%struct.LargeWithCopyConstructor) align 1 %agg.result)
 
 using IntFunctionType = int();
 IntFunctionType *ReturnsIntFunction();
diff --git a/clang/test/CodeGenCXX/call-with-static-chain.cpp b/clang/test/CodeGenCXX/call-with-static-chain.cpp
index 06c3f15396a86..61011a65e4a2c 100644
--- a/clang/test/CodeGenCXX/call-with-static-chain.cpp
+++ b/clang/test/CodeGenCXX/call-with-static-chain.cpp
@@ -25,8 +25,8 @@ void test() {
   // CHECK64: call i32 @f1(ptr nest noundef @f1
   __builtin_call_with_static_chain(f1(a, a, a, a), f1);
 
-  // CHECK32: call void @f2(ptr sret(%struct.B) align 4 %{{[0-9a-z]+}}, ptr nest noundef @f2)
-  // CHECK64: call void @f2(ptr sret(%struct.B) align 8 %{{[0-9a-z]+}}, ptr nest noundef @f2)
+  // CHECK32: call void @f2(ptr dead_on_unwind writable sret(%struct.B) align 4 %{{[0-9a-z]+}}, ptr nest noundef @f2)
+  // CHECK64: call void @f2(ptr dead_on_unwind writable sret(%struct.B) align 8 %{{[0-9a-z]+}}, ptr nest noundef @f2)
   __builtin_call_with_static_chain(f2(), f2);
 
   // CHECK32: call i64 @f3(ptr nest noundef @f3)
diff --git a/clang/test/CodeGenCXX/conditional-gnu-ext.cpp b/clang/test/CodeGenCXX/conditional-gnu-ext.cpp
index b17e0f7bb9a22..3d3d210f22f69 100644
--- a/clang/test/CodeGenCXX/conditional-gnu-ext.cpp
+++ b/clang/test/CodeGenCXX/conditional-gnu-ext.cpp
@@ -94,7 +94,7 @@ namespace test3 {
   B test1() {
     // CHECK-LABEL:    define{{.*}} void @_ZN5test35test1Ev(
     // CHECK:      [[TEMP:%.*]] = alloca [[B:%.*]],
-    // CHECK:      call  void @_ZN5test312test1_helperEv(ptr sret([[B]]) align 1 [[TEMP]])
+    // CHECK:      call  void @_ZN5test312test1_helperEv(ptr dead_on_unwind writable sret([[B]]) align 1 [[TEMP]])
     // CHECK-NEXT: [[BOOL:%.*]] = call noundef zeroext i1 @_ZN5test31BcvbEv(ptr {{[^,]*}} [[TEMP]])
     // CHECK-NEXT: br i1 [[BOOL]]
     // CHECK:      call void @_ZN5test31BC1ERKS0_(ptr {{[^,]*}} [[RESULT:%.*]], ptr noundef nonnull align {{[0-9]+}} dereferenceable({{[0-9]+}}) [[TEMP]])
@@ -117,7 +117,7 @@ namespace test3 {
     // CHECK-NEXT: [[T0:%.*]] = load ptr, ptr [[X]]
     // CHECK-NEXT: [[BOOL:%.*]] = call noundef zeroext i1 @_ZN5test31BcvbEv(ptr {{[^,]*}} [[T0]])
     // CHECK-NEXT: br i1 [[BOOL]]
-    // CHECK:      call void @_ZN5test31BcvNS_1AEEv(ptr sret([[A:%.*]]) align 1 [[RESULT:%.*]], ptr {{[^,]*}} [[T0]])
+    // CHECK:      call void @_ZN5test31BcvNS_1AEEv(ptr dead_on_unwind writable sret([[A:%.*]]) align 1 [[RESULT:%.*]], ptr {{[^,]*}} [[T0]])
     // CHECK-NEXT: br label
     // CHECK:      call void @_ZN5test31AC1Ev(ptr {{[^,]*}} [[RESULT]])
     // CHECK-NEXT: br label
@@ -128,10 +128,10 @@ namespace test3 {
   A test3() {
     // CHECK-LABEL:    define{{.*}} void @_ZN5test35test3Ev(
     // CHECK:      [[TEMP:%.*]] = alloca [[B]],
-    // CHECK:      call  void @_ZN5test312test3_helperEv(ptr sret([[B]]) align 1 [[TEMP]])
+    // CHECK:      call  void @_ZN5test312test3_helperEv(ptr dead_on_unwind writable sret([[B]]) align 1 [[TEMP]])
     // CHECK-NEXT: [[BOOL:%.*]] = call noundef zeroext i1 @_ZN5test31BcvbEv(ptr {{[^,]*}} [[TEMP]])
     // CHECK-NEXT: br i1 [[BOOL]]
-    // CHECK:      call void @_ZN5test31BcvNS_1AEEv(ptr sret([[A]]) align 1 [[RESULT:%.*]], ptr {{[^,]*}} [[TEMP]])
+    // CHECK:      call void @_ZN5test31BcvNS_1AEEv(ptr dead_on_unwind writable sret([[A]]) align 1 [[RESULT:%.*]], ptr {{[^,]*}} [[TEMP]])
     // CHECK-NEXT: br label
     // CHECK:      call void @_ZN5test31AC1Ev(ptr {{[^,]*}} [[RESULT]])
     // CHECK-NEXT: br label
diff --git a/clang/test/CodeGenCXX/cxx1z-copy-omission.cpp b/clang/test/CodeGenCXX/cxx1z-copy-omission.cpp
index db6671bcd301b..e64d163fd3d6c 100644
--- a/clang/test/CodeGenCXX/cxx1z-copy-omission.cpp
+++ b/clang/test/CodeGenCXX/cxx1z-copy-omission.cpp
@@ -19,7 +19,7 @@ void g() {
   // CHECK: %[[A:.*]] = alloca
   // CHECK-NOT: alloca
   // CHECK-NOT: call
-  // CHECK: call {{.*}} @_Z1fv(ptr sret({{.*}}) align 4 %[[A]])
+  // CHECK: call {{.*}} @_Z1fv(ptr dead_on_unwind writable sret({{.*}}) align 4 %[[A]])
   A a = A( A{ f() } );
   // CHECK-NOT: call
 
@@ -40,7 +40,7 @@ void h() {
   // CHECK-NOT: alloca
   // CHECK-NOT: call
 
-  // CHECK: call {{.*}} @_Z1fv(ptr sret({{.*}}) align 4 %[[A]])
+  // CHECK: call {{.*}} @_Z1fv(ptr dead_on_unwind writable sret({{.*}}) align 4 %[[A]])
   // CHECK-NOT: call
   // CHECK: call {{.*}} @_Z1f1A(ptr noundef %[[A]])
   f(f());
diff --git a/clang/test/CodeGenCXX/cxx1z-lambda-star-this.cpp b/clang/test/CodeGenCXX/cxx1z-lambda-star-this.cpp
index 63bfe5603b694..1388f1b87889a 100644
--- a/clang/test/CodeGenCXX/cxx1z-lambda-star-this.cpp
+++ b/clang/test/CodeGenCXX/cxx1z-lambda-star-this.cpp
@@ -10,7 +10,7 @@ namespace ns1 {
 int X = A{}.foo()();
 } //end ns1
 
-//CHECK: @"?foo@A@@QAE?A?@@XZ"(ptr {{[^,]*}} %this, ptr noalias sret(%class.anon) align 8 %[[A_LAMBDA_RETVAL:.*]])
+//CHECK: @"?foo@A@@QAE?A?@@XZ"(ptr {{[^,]*}} %this, ptr dead_on_unwind noalias writable sret(%class.anon) align 8 %[[A_LAMBDA_RETVAL:.*]])
 // get the first object with the closure type, which is of type 'struct.A'
 //CHECK: %[[I0:.+]] = getelementptr inbounds %[[A_LAMBDA]], ptr %[[A_LAMBDA_RETVAL]], i32 0, i32 0
 // copy the contents ...
@@ -24,6 +24,6 @@ struct B {
 namespace ns2 {
 int X = B{}.bar()();
 }
-//CHECK: @"?bar@B@@QAE?A?@@XZ"(ptr {{[^,]*}} %this, ptr noalias sret(%class.anon.0) align 4 %agg.result)
+//CHECK: @"?bar@B@@QAE?A?@@XZ"(ptr {{[^,]*}} %this, ptr dead_on_unwind noalias writable sret(%class.anon.0) align 4 %agg.result)
 //CHECK: %[[I20:.+]] = getelementptr inbounds %class.anon.0, ptr %agg.result, i32 0, i32 0
 //CHECK: store ptr %this1, ptr %[[I20]], align 4
diff --git a/clang/test/CodeGenCXX/exceptions.cpp b/clang/test/CodeGenCXX/exceptions.cpp
index 483876cc212e3..e8179f9828fb6 100644
--- a/clang/test/CodeGenCXX/exceptions.cpp
+++ b/clang/test/CodeGenCXX/exceptions.cpp
@@ -142,12 +142,12 @@ namespace test1 {
     // CHECK:      [[ACTIVE:%.*]] = alloca i1
     // CHECK:      [[NEW:%.*]] = call noalias nonnull ptr @_Znwm(i64 8)
     // CHECK-NEXT: store i1 true, ptr [[ACTIVE]]
-    // CHECK-NEXT: invoke void @_ZN5test15makeBEv(ptr sret([[B:%.*]]) align 4 [[T0:%.*]])
+    // CHECK-NEXT: invoke void @_ZN5test15makeBEv(ptr dead_on_unwind writable sret([[B:%.*]]) align 4 [[T0:%.*]])
     // CHECK:      [[T1:%.*]] = invoke i32 @_ZN5test11BcviEv(ptr {{[^,]*}} [[T0]])
     // CHECK:      invoke void @_ZN5test11AC1Ei(ptr {{[^,]*}} [[NEW]], i32 [[T1]])
     // CHECK:      store i1 false, ptr [[ACTIVE]]
     // CHECK-NEXT: store ptr [[NEW]], ptr [[X]], align 8
-    // CHECK:      invoke void @_ZN5test15makeBEv(ptr sret([[B:%.*]]) align 4 [[T2:%.*]])
+    // CHECK:      invoke void @_ZN5test15makeBEv(ptr dead_on_unwind writable sret([[B:%.*]]) align 4 [[T2:%.*]])
     // CHECK:      [[RET:%.*]] = load ptr, ptr [[X]], align 8
 
     // CHECK98:      invoke void @_ZN5test11BD1Ev(ptr {{[^,]*}} [[T2]])
@@ -231,7 +231,7 @@ namespace test3 {
     // CHECK-NEXT: store ptr [[NEW]], ptr [[SAVED0]]
     // CHECK-NEXT: store ptr [[FOO]], ptr [[SAVED1]]
     // CHECK-NEXT: store i1 true, ptr [[CLEANUPACTIVE]]
-    // CHECK-NEXT: invoke void @_ZN5test35makeAEv(ptr sret([[A:%.*]]) align 8 [[NEW]])
+    // CHECK-NEXT: invoke void @_ZN5test35makeAEv(ptr dead_on_unwind writable sret([[A:%.*]]) align 8 [[NEW]])
     // CHECK: br label
     //   -> cond.end
             new(foo(),10.0) A(makeA()) :
diff --git a/clang/test/CodeGenCXX/homogeneous-aggregates.cpp b/clang/test/CodeGenCXX/homogeneous-aggregates.cpp
index bd83b9f932aad..63ffc6b5bfac8 100644
--- a/clang/test/CodeGenCXX/homogeneous-aggregates.cpp
+++ b/clang/test/CodeGenCXX/homogeneous-aggregates.cpp
@@ -39,10 +39,10 @@ struct I2 : Base2 {};
 struct I3 : Base2 {};
 struct D5 : I1, I2, I3 {}; // homogeneous aggregate
 
-// PPC: define{{.*}} void @_Z7func_D12D1(ptr noalias sret(%struct.D1) align 8 %agg.result, [3 x i64] %x.coerce)
-// ARM32: define{{.*}} arm_aapcs_vfpcc void @_Z7func_D12D1(ptr noalias sret(%struct.D1) align 8 %agg.result, [3 x i64] %x.coerce)
-// ARM64: define{{.*}} void @_Z7func_D12D1(ptr noalias sret(%struct.D1) align 8 %agg.result, ptr noundef %x)
-// X64: define dso_local x86_vectorcallcc void @"\01_Z7func_D12D1@@24"(ptr noalias sret(%struct.D1) align 8 %agg.result, ptr noundef %x)
+// PPC: define{{.*}} void @_Z7func_D12D1(ptr dead_on_unwind noalias writable sret(%struct.D1) align 8 %agg.result, [3 x i64] %x.coerce)
+// ARM32: define{{.*}} arm_aapcs_vfpcc void @_Z7func_D12D1(ptr dead_on_unwind noalias writable sret(%struct.D1) align 8 %agg.result, [3 x i64] %x.coerce)
+// ARM64: define{{.*}} void @_Z7func_D12D1(ptr dead_on_unwind noalias writable sret(%struct.D1) align 8 %agg.result, ptr noundef %x)
+// X64: define dso_local x86_vectorcallcc void @"\01_Z7func_D12D1@@24"(ptr dead_on_unwind noalias writable sret(%struct.D1) align 8 %agg.result, ptr noundef %x)
 D1 CC func_D1(D1 x) { return x; }
 
 // PPC: define{{.*}} [3 x double] @_Z7func_D22D2([3 x double] %x.coerce)
@@ -51,9 +51,9 @@ D1 CC func_D1(D1 x) { return x; }
 // X64: define dso_local x86_vectorcallcc %struct.D2 @"\01_Z7func_D22D2@@24"(%struct.D2 inreg %x.coerce)
 D2 CC func_D2(D2 x) { return x; }
 
-// PPC: define{{.*}} void @_Z7func_D32D3(ptr noalias sret(%struct.D3) align 8 %agg.result, [4 x i64] %x.coerce)
-// ARM32: define{{.*}} arm_aapcs_vfpcc void @_Z7func_D32D3(ptr noalias sret(%struct.D3) align 8 %agg.result, [4 x i64] %x.coerce)
-// ARM64: define{{.*}} void @_Z7func_D32D3(ptr noalias sret(%struct.D3) align 8 %agg.result, ptr noundef %x)
+// PPC: define{{.*}} void @_Z7func_D32D3(ptr dead_on_unwind noalias writable sret(%struct.D3) align 8 %agg.result, [4 x i64] %x.coerce)
+// ARM32: define{{.*}} arm_aapcs_vfpcc void @_Z7func_D32D3(ptr dead_on_unwind noalias writable sret(%struct.D3) align 8 %agg.result, [4 x i64] %x.coerce)
+// ARM64: define{{.*}} void @_Z7func_D32D3(ptr dead_on_unwind noalias writable sret(%struct.D3) align 8 %agg.result, ptr noundef %x)
 D3 CC func_D3(D3 x) { return x; }
 
 // PPC: define{{.*}} [4 x double] @_Z7func_D42D4([4 x double] %x.coerce)
@@ -133,13 +133,13 @@ struct HasEmptyBase : public Empty {
 struct HasPodBase : public Pod {};
 // WOA64-LABEL: define dso_local %"struct.pr47611::Pod" @"?copy@pr47611@@YA?AUPod@1@PEAU21@@Z"(ptr noundef %x)
 Pod copy(Pod *x) { return *x; } // MSVC: ldp d0,d1,[x0], Clang: ldp d0,d1,[x0]
-// WOA64-LABEL: define dso_local void @"?copy@pr47611@@YA?AUNotCXX14Aggregate@1@PEAU21@@Z"(ptr inreg noalias sret(%"struct.pr47611::NotCXX14Aggregate") align 8 %agg.result, ptr noundef %x)
+// WOA64-LABEL: define dso_local void @"?copy@pr47611@@YA?AUNotCXX14Aggregate@1@PEAU21@@Z"(ptr dead_on_unwind inreg noalias writable sret(%"struct.pr47611::NotCXX14Aggregate") align 8 %agg.result, ptr noundef %x)
 NotCXX14Aggregate copy(NotCXX14Aggregate *x) { return *x; } // MSVC: stp x8,x9,[x0], Clang: str q0,[x0]
 // WOA64-LABEL: define dso_local [2 x i64] @"?copy@pr47611@@YA?AUNotPod@1@PEAU21@@Z"(ptr noundef %x)
 NotPod copy(NotPod *x) { return *x; }
-// WOA64-LABEL: define dso_local void @"?copy@pr47611@@YA?AUHasEmptyBase@1@PEAU21@@Z"(ptr inreg noalias sret(%"struct.pr47611::HasEmptyBase") align 8 %agg.result, ptr noundef %x)
+// WOA64-LABEL: define dso_local void @"?copy@pr47611@@YA?AUHasEmptyBase@1@PEAU21@@Z"(ptr dead_on_unwind inreg noalias writable sret(%"struct.pr47611::HasEmptyBase") align 8 %agg.result, ptr noundef %x)
 HasEmptyBase copy(HasEmptyBase *x) { return *x; }
-// WOA64-LABEL: define dso_local void @"?copy@pr47611@@YA?AUHasPodBase@1@PEAU21@@Z"(ptr inreg noalias sret(%"struct.pr47611::HasPodBase") align 8 %agg.result, ptr noundef %x)
+// WOA64-LABEL: define dso_local void @"?copy@pr47611@@YA?AUHasPodBase@1@PEAU21@@Z"(ptr dead_on_unwind inreg noalias writable sret(%"struct.pr47611::HasPodBase") align 8 %agg.result, ptr noundef %x)
 HasPodBase copy(HasPodBase *x) { return *x; }
 
 void call_copy_pod(Pod *pod) {
@@ -151,7 +151,7 @@ void call_copy_pod(Pod *pod) {
 void call_copy_notcxx14aggregate(NotCXX14Aggregate *notcxx14aggregate) {
   *notcxx14aggregate = copy(notcxx14aggregate);
   // WOA64-LABEL: define dso_local void @"?call_copy_notcxx14aggregate@pr47611@@YAXPEAUNotCXX14Aggregate@1@@Z"
-  // WOA64: call void @"?copy@pr47611@@YA?AUNotCXX14Aggregate@1@PEAU21@@Z"(ptr inreg sret(%"struct.pr47611::NotCXX14Aggregate") align 8 %{{.*}}, ptr noundef %{{.*}})
+  // WOA64: call void @"?copy@pr47611@@YA?AUNotCXX14Aggregate@1@PEAU21@@Z"(ptr dead_on_unwind inreg writable sret(%"struct.pr47611::NotCXX14Aggregate") align 8 %{{.*}}, ptr noundef %{{.*}})
 }
 
 void call_copy_notpod(NotPod *notPod) {
@@ -163,13 +163,13 @@ void call_copy_notpod(NotPod *notPod) {
 void call_copy_hasemptybase(HasEmptyBase *hasEmptyBase) {
   *hasEmptyBase = copy(hasEmptyBase);
   // WOA64-LABEL: define dso_local void @"?call_copy_hasemptybase@pr47611@@YAXPEAUHasEmptyBase@1@@Z"
-  // WOA64: call void @"?copy@pr47611@@YA?AUHasEmptyBase@1@PEAU21@@Z"(ptr inreg sret(%"struct.pr47611::HasEmptyBase") align 8 %{{.*}}, ptr noundef %{{.*}})
+  // WOA64: call void @"?copy@pr47611@@YA?AUHasEmptyBase@1@PEAU21@@Z"(ptr dead_on_unwind inreg writable sret(%"struct.pr47611::HasEmptyBase") align 8 %{{.*}}, ptr noundef %{{.*}})
 }
 
 void call_copy_haspodbase(HasPodBase *hasPodBase) {
   *hasPodBase = copy(hasPodBase);
   // WOA64-LABEL: define dso_local void @"?call_copy_haspodbase@pr47611@@YAXPEAUHasPodBase@1@@Z"
-  // WOA64: call void @"?copy@pr47611@@YA?AUHasPodBase@1@PEAU21@@Z"(ptr inreg sret(%"struct.pr47611::HasPodBase") align 8 %{{.*}}, ptr noundef %{{.*}})
+  // WOA64: call void @"?copy@pr47611@@YA?AUHasPodBase@1@PEAU21@@Z"(ptr dead_on_unwind inreg writable sret(%"struct.pr47611::HasPodBase") align 8 %{{.*}}, ptr noundef %{{.*}})
 }
 } // namespace pr47611
 
@@ -300,5 +300,5 @@ test f(test *x) { return *x; }
 struct base2 { double v; };
 struct test2 : base2 { test2(double); protected: double v2;};
 test2 f(test2 *x) { return *x; }
-// WOA64: define dso_local void @"?f@pr62223@@YA?AUtest2@1@PEAU21@@Z"(ptr inreg noalias sret(%"struct.pr62223::test2") align 8 %{{.*}}, ptr noundef %{{.*}})
+// WOA64: define dso_local void @"?f@pr62223@@YA?AUtest2@1@PEAU21@@Z"(ptr dead_on_unwind inreg noalias writable sret(%"struct.pr62223::test2") align 8 %{{.*}}, ptr noundef %{{.*}})
 }
diff --git a/clang/test/CodeGenCXX/lambda-expressions.cpp b/clang/test/CodeGenCXX/lambda-expressions.cpp
index 3ad982a195cc9..b929aa0c67513 100644
--- a/clang/test/CodeGenCXX/lambda-expressions.cpp
+++ b/clang/test/CodeGenCXX/lambda-expressions.cpp
@@ -193,8 +193,8 @@ namespace pr28595 {
 // CHECK-NEXT: call noundef i32 @"_ZZ1fvENK3$_0clEii"
 // CHECK-NEXT: ret i32
 
-// CHECK-LABEL: define internal void @"_ZZ1hvEN3$_08__invokeEv"(ptr noalias sret(%struct.A) align 1 %agg.result) {{.*}} {
-// CHECK: call void @"_ZZ1hvENK3$_0clEv"(ptr sret(%struct.A) align 1 %agg.result,
+// CHECK-LABEL: define internal void @"_ZZ1hvEN3$_08__invokeEv"(ptr dead_on_unwind noalias writable sret(%struct.A) align 1 %agg.result) {{.*}} {
+// CHECK: call void @"_ZZ1hvENK3$_0clEv"(ptr dead_on_unwind writable sret(%struct.A) align 1 %agg.result,
 // CHECK-NEXT: ret void
 struct A { ~A(); };
 void h() {
diff --git a/clang/test/CodeGenCXX/matrix-casts.cpp b/clang/test/CodeGenCXX/matrix-casts.cpp
index 4369d99a34915..0a946e464adef 100644
--- a/clang/test/CodeGenCXX/matrix-casts.cpp
+++ b/clang/test/CodeGenCXX/matrix-casts.cpp
@@ -324,7 +324,7 @@ class Foo {
 };
 
 Foo class_constructor_matrix_ty(matrix_5_5 m) {
-  // CHECK-LABEL: define void @_Z27class_constructor_matrix_tyu11matrix_typeILm5ELm5EiE(ptr noalias sret(%class.Foo) align 4 %agg.result, <25 x i32> noundef %m)
+  // CHECK-LABEL: define void @_Z27class_constructor_matrix_tyu11matrix_typeILm5ELm5EiE(ptr dead_on_unwind noalias writable sret(%class.Foo) align 4 %agg.result, <25 x i32> noundef %m)
   // CHECK:         [[M:%.*]]  = load <25 x i32>, ptr {{.*}}, align 4
   // CHECK-NEXT:    call void @_ZN3FooC1Eu11matrix_typeILm5ELm5EiE(ptr noundef nonnull align 4 dereferenceable(40) %agg.result, <25 x i32> noundef [[M]])
   // CHECK-NEXT:    ret void
@@ -338,7 +338,7 @@ struct Bar {
 };
 
 Bar struct_constructor_matrix_ty(matrix_4_4 m) {
-  // CHECK-LABEL: define void @_Z28struct_constructor_matrix_tyu11matrix_typeILm4ELm4EfE(ptr noalias sret(%struct.Bar) align 4 %agg.result, <16 x float> noundef %m)
+  // CHECK-LABEL: define void @_Z28struct_constructor_matrix_tyu11matrix_typeILm4ELm4EfE(ptr dead_on_unwind noalias writable sret(%struct.Bar) align 4 %agg.result, <16 x float> noundef %m)
   // CHECK:         [[M:%.*]] = load <16 x float>, ptr {{.*}}, align 4
   // CHECK-NEXT:    call void @_ZN3BarC1Eu11matrix_typeILm4ELm4EfE(ptr noundef nonnull align 4 dereferenceable(40) %agg.result, <16 x float> noundef [[M]])
   // CHECK-NEXT:    ret void
diff --git a/clang/test/CodeGenCXX/matrix-type-builtins.cpp b/clang/test/CodeGenCXX/matrix-type-builtins.cpp
index 732fe1a18db3b..9c334a6858f15 100644
--- a/clang/test/CodeGenCXX/matrix-type-builtins.cpp
+++ b/clang/test/CodeGenCXX/matrix-type-builtins.cpp
@@ -19,7 +19,7 @@ MyMatrix transpose(const MyMatrix &M) {
 
 void test_transpose_template1() {
   // CHECK-LABEL: define{{.*}} void @_Z24test_transpose_template1v()
-  // CHECK:         call void @_Z9transposeIiLj4ELj10EE8MyMatrixIT_XT1_EXT0_EERKS0_IS1_XT0_EXT1_EE(ptr sret(%struct.MyMatrix.0) align 4 %M1_t, ptr noundef nonnull align 4 dereferenceable(160) %M1)
+  // CHECK:         call void @_Z9transposeIiLj4ELj10EE8MyMatrixIT_XT1_EXT0_EERKS0_IS1_XT0_EXT1_EE(ptr dead_on_unwind writable sret(%struct.MyMatrix.0) align 4 %M1_t, ptr noundef nonnull align 4 dereferenceable(160) %M1)
 
   // CHECK-LABEL: define linkonce_odr void @_Z9transposeIiLj4ELj10EE8MyMatrixIT_XT1_EXT0_EERKS0_IS1_XT0_EXT1_EE(
   // CHECK:         [[M:%.*]] = load <40 x i32>, ptr {{.*}}, align 4
@@ -31,9 +31,9 @@ void test_transpose_template1() {
 
 void test_transpose_template2(MyMatrix &M) {
   // CHECK-LABEL: define{{.*}} void @_Z24test_transpose_template2R8MyMatrixIdLj7ELj6EE(
-  // CHECK:         call void @_Z9transposeIdLj7ELj6EE8MyMatrixIT_XT1_EXT0_EERKS0_IS1_XT0_EXT1_EE(ptr sret(%struct.MyMatrix.1) align 8 %ref.tmp1, ptr noundef nonnull align 8 dereferenceable(336) %0)
-  // CHECK-NEXT:    call void @_Z9transposeIdLj6ELj7EE8MyMatrixIT_XT1_EXT0_EERKS0_IS1_XT0_EXT1_EE(ptr sret(%struct.MyMatrix.2) align 8 %ref.tmp, ptr noundef nonnull align 8 dereferenceable(336) %ref.tmp1)
-  // CHECK-NEXT:    call void @_Z9transposeIdLj7ELj6EE8MyMatrixIT_XT1_EXT0_EERKS0_IS1_XT0_EXT1_EE(ptr sret(%struct.MyMatrix.1) align 8 %M2_t, ptr noundef nonnull align 8 dereferenceable(336) %ref.tmp)
+  // CHECK:         call void @_Z9transposeIdLj7ELj6EE8MyMatrixIT_XT1_EXT0_EERKS0_IS1_XT0_EXT1_EE(ptr dead_on_unwind writable sret(%struct.MyMatrix.1) align 8 %ref.tmp1, ptr noundef nonnull align 8 dereferenceable(336) %0)
+  // CHECK-NEXT:    call void @_Z9transposeIdLj6ELj7EE8MyMatrixIT_XT1_EXT0_EERKS0_IS1_XT0_EXT1_EE(ptr dead_on_unwind writable sret(%struct.MyMatrix.2) align 8 %ref.tmp, ptr noundef nonnull align 8 dereferenceable(336) %ref.tmp1)
+  // CHECK-NEXT:    call void @_Z9transposeIdLj7ELj6EE8MyMatrixIT_XT1_EXT0_EERKS0_IS1_XT0_EXT1_EE(ptr dead_on_unwind writable sret(%struct.MyMatrix.1) align 8 %M2_t, ptr noundef nonnull align 8 dereferenceable(336) %ref.tmp)
 
   // CHECK-LABEL: define linkonce_odr void @_Z9transposeIdLj7ELj6EE8MyMatrixIT_XT1_EXT0_EERKS0_IS1_XT0_EXT1_EE(
   // CHECK:         [[M:%.*]] = load <42 x double>, ptr {{.*}}, align 8
diff --git a/clang/test/CodeGenCXX/matrix-type.cpp b/clang/test/CodeGenCXX/matrix-type.cpp
index 79bceabb115f8..d93db29d4d0aa 100644
--- a/clang/test/CodeGenCXX/matrix-type.cpp
+++ b/clang/test/CodeGenCXX/matrix-type.cpp
@@ -127,7 +127,7 @@ void matrix_template_reference(MatrixClassTemplate &a, MatrixCla
 }
 
 MatrixClassTemplate matrix_template_reference_caller(float *Data) {
-  // CHECK-LABEL: define{{.*}} void @_Z32matrix_template_reference_callerPf(ptr noalias sret(%class.MatrixClassTemplate) align 8 %agg.result, ptr %Data
+  // CHECK-LABEL: define{{.*}} void @_Z32matrix_template_reference_callerPf(ptr dead_on_unwind noalias writable sret(%class.MatrixClassTemplate) align 8 %agg.result, ptr %Data
   // CHECK-NEXT:  entry:
   // CHECK-NEXT:    %Data.addr = alloca ptr, align 8
   // CHECK-NEXT:    %Arg = alloca %class.MatrixClassTemplate, align 8
diff --git a/clang/test/CodeGenCXX/microsoft-abi-byval-sret.cpp b/clang/test/CodeGenCXX/microsoft-abi-byval-sret.cpp
index fde27c6ce85db..668a27e9d2623 100644
--- a/clang/test/CodeGenCXX/microsoft-abi-byval-sret.cpp
+++ b/clang/test/CodeGenCXX/microsoft-abi-byval-sret.cpp
@@ -49,7 +49,7 @@ A B::qux(A x) {
 }
 
 // CHECK-LABEL: define dso_local x86_fastcallcc void @"?qux@B@@QAI?AUA@@U2@@Z"
-// CHECK:       (ptr inreg noundef %this, ptr inreg noalias sret(%struct.A) align 4 %agg.result, ptr inalloca(<{ %struct.A }>) %0)
+// CHECK:       (ptr inreg noundef %this, ptr dead_on_unwind inreg noalias writable sret(%struct.A) align 4 %agg.result, ptr inalloca(<{ %struct.A }>) %0)
 // CHECK:   ret void
 
 int main() {
@@ -67,4 +67,4 @@ int main() {
 // CHECK: call x86_stdcallcc ptr @"?baz@B@@QAG?AUA@@U2@@Z"
 // CHECK:       (ptr inalloca(<{ ptr, ptr, %struct.A }>) %{{[^,]*}})
 // CHECK: call x86_fastcallcc void @"?qux@B@@QAI?AUA@@U2@@Z"
-// CHECK:       (ptr inreg noundef %{{[^,]*}}, ptr inreg sret(%struct.A) align 4 %{{.*}}, ptr inalloca(<{ %struct.A }>) %{{[^,]*}})
+// CHECK:       (ptr inreg noundef %{{[^,]*}}, ptr dead_on_unwind inreg writable sret(%struct.A) align 4 %{{.*}}, ptr inalloca(<{ %struct.A }>) %{{[^,]*}})
diff --git a/clang/test/CodeGenCXX/microsoft-abi-byval-thunks.cpp b/clang/test/CodeGenCXX/microsoft-abi-byval-thunks.cpp
index 66ecf500a8693..4c485fc3142ab 100644
--- a/clang/test/CodeGenCXX/microsoft-abi-byval-thunks.cpp
+++ b/clang/test/CodeGenCXX/microsoft-abi-byval-thunks.cpp
@@ -86,10 +86,10 @@ C::C() {} // force emission
 // CHECK32-NEXT: ret ptr %[[rv]]
 
 // CHECK64-LABEL: define linkonce_odr dso_local void @"?foo@C@sret_thunk@@W7EAA?AUAgg@2@U32@@Z"
-// CHECK64:             (ptr noundef %this, ptr noalias sret(%"struct.sret_thunk::Agg") align 4 %agg.result, ptr noundef %x)
+// CHECK64:             (ptr noundef %this, ptr dead_on_unwind noalias writable sret(%"struct.sret_thunk::Agg") align 4 %agg.result, ptr noundef %x)
 // CHECK64:   getelementptr i8, ptr %{{.*}}, i32 -8
 // CHECK64:   call void @"?foo@C@sret_thunk@@UEAA?AUAgg@2@U32@@Z"
-// CHECK64:       (ptr {{[^,]*}} %{{.*}}, ptr sret(%"struct.sret_thunk::Agg") align 4 %agg.result, ptr noundef %x)
+// CHECK64:       (ptr {{[^,]*}} %{{.*}}, ptr dead_on_unwind writable sret(%"struct.sret_thunk::Agg") align 4 %agg.result, ptr noundef %x)
 // CHECK64-NOT: call
 // CHECK64:   ret void
 }
diff --git a/clang/test/CodeGenCXX/microsoft-abi-cdecl-method-sret.cpp b/clang/test/CodeGenCXX/microsoft-abi-cdecl-method-sret.cpp
index b4a6ff8e6d0ff..a1faab25b109c 100644
--- a/clang/test/CodeGenCXX/microsoft-abi-cdecl-method-sret.cpp
+++ b/clang/test/CodeGenCXX/microsoft-abi-cdecl-method-sret.cpp
@@ -19,9 +19,9 @@ S C::variadic_sret(const char *f, ...) { return S(); }
 S C::cdecl_sret() { return S(); }
 S C::byval_and_sret(S a) { return S(); }
 
-// CHECK: define dso_local void @"?variadic_sret@C@@QAA?AUS@@PBDZZ"(ptr {{[^,]*}} %this, ptr noalias sret(%struct.S) align 4 %agg.result, ptr noundef %f, ...)
-// CHECK: define dso_local void @"?cdecl_sret@C@@QAA?AUS@@XZ"(ptr {{[^,]*}} %this, ptr noalias sret(%struct.S) align 4 %agg.result)
-// CHECK: define dso_local void @"?byval_and_sret@C@@QAA?AUS@@U2@@Z"(ptr {{[^,]*}} %this, ptr noalias sret(%struct.S) align 4 %agg.result, ptr noundef byval(%struct.S) align 4 %a)
+// CHECK: define dso_local void @"?variadic_sret@C@@QAA?AUS@@PBDZZ"(ptr {{[^,]*}} %this, ptr dead_on_unwind noalias writable sret(%struct.S) align 4 %agg.result, ptr noundef %f, ...)
+// CHECK: define dso_local void @"?cdecl_sret@C@@QAA?AUS@@XZ"(ptr {{[^,]*}} %this, ptr dead_on_unwind noalias writable sret(%struct.S) align 4 %agg.result)
+// CHECK: define dso_local void @"?byval_and_sret@C@@QAA?AUS@@U2@@Z"(ptr {{[^,]*}} %this, ptr dead_on_unwind noalias writable sret(%struct.S) align 4 %agg.result, ptr noundef byval(%struct.S) align 4 %a)
 
 int main() {
   C c;
@@ -41,4 +41,4 @@ struct A {
 S A::f(int x) {
   return S();
 }
-// CHECK-LABEL: define dso_local x86_fastcallcc void @"?f@A@@QAI?AUS@@H@Z"(ptr inreg noundef nonnull align 1 dereferenceable(1) %this, ptr inreg noalias sret(%struct.S) align 4 %agg.result, i32 noundef %x)
+// CHECK-LABEL: define dso_local x86_fastcallcc void @"?f@A@@QAI?AUS@@H@Z"(ptr inreg noundef nonnull align 1 dereferenceable(1) %this, ptr dead_on_unwind inreg noalias writable sret(%struct.S) align 4 %agg.result, i32 noundef %x)
diff --git a/clang/test/CodeGenCXX/microsoft-abi-eh-cleanups.cpp b/clang/test/CodeGenCXX/microsoft-abi-eh-cleanups.cpp
index 4bff7ccf8986b..f00cf90762756 100644
--- a/clang/test/CodeGenCXX/microsoft-abi-eh-cleanups.cpp
+++ b/clang/test/CodeGenCXX/microsoft-abi-eh-cleanups.cpp
@@ -18,9 +18,9 @@ void HasEHCleanup() {
 // WIN32-LABEL: define dso_local void @"?HasEHCleanup@@YAXXZ"() {{.*}} {
 // WIN32:   %[[base:.*]] = call ptr @llvm.stacksave.p0()
 //    If this call throws, we have to restore the stack.
-// WIN32:   call void @"?getA@@YA?AUA@@XZ"(ptr sret(%struct.A) align 4 %{{.*}})
+// WIN32:   call void @"?getA@@YA?AUA@@XZ"(ptr dead_on_unwind writable sret(%struct.A) align 4 %{{.*}})
 //    If this call throws, we have to cleanup the first temporary.
-// WIN32:   invoke void @"?getA@@YA?AUA@@XZ"(ptr sret(%struct.A) align 4 %{{.*}})
+// WIN32:   invoke void @"?getA@@YA?AUA@@XZ"(ptr dead_on_unwind writable sret(%struct.A) align 4 %{{.*}})
 //    If this call throws, we have to cleanup the stacksave.
 // WIN32:   call noundef i32 @"?TakesTwo@@YAHUA@@0@Z"
 // WIN32:   call void @llvm.stackrestore
@@ -42,8 +42,8 @@ void HasEHCleanupNoexcept() noexcept {
 // With exceptions, we need to clean up at least one of these temporaries.
 // WIN32-LABEL: define dso_local void @"?HasEHCleanupNoexcept@@YAXXZ"() {{.*}} {
 // WIN32:   %[[base:.*]] = call ptr @llvm.stacksave.p0()
-// WIN32:   invoke void @"?getA@@YA?AUA@@XZ"(ptr sret(%struct.A) align 4 %{{.*}})
-// WIN32:   invoke void @"?getA@@YA?AUA@@XZ"(ptr sret(%struct.A) align 4 %{{.*}})
+// WIN32:   invoke void @"?getA@@YA?AUA@@XZ"(ptr dead_on_unwind writable sret(%struct.A) align 4 %{{.*}})
+// WIN32:   invoke void @"?getA@@YA?AUA@@XZ"(ptr dead_on_unwind writable sret(%struct.A) align 4 %{{.*}})
 // WIN32:   invoke noundef i32 @"?TakesTwo@@YAHUA@@0@Z"
 // WIN32:   call void @llvm.stackrestore
 // WIN32:   ret void
diff --git a/clang/test/CodeGenCXX/microsoft-abi-sret-and-byval.cpp b/clang/test/CodeGenCXX/microsoft-abi-sret-and-byval.cpp
index 9d737e3979ddd..585167d0a1423 100644
--- a/clang/test/CodeGenCXX/microsoft-abi-sret-and-byval.cpp
+++ b/clang/test/CodeGenCXX/microsoft-abi-sret-and-byval.cpp
@@ -87,58 +87,58 @@ void call_bools_and_chars() {
 
 // Returning structs that fit into a register.
 Small small_return() { return Small(); }
-// LINUX-LABEL: define{{.*}} void @_Z12small_returnv(ptr noalias sret(%struct.Small) align 4 %agg.result)
+// LINUX-LABEL: define{{.*}} void @_Z12small_returnv(ptr dead_on_unwind noalias writable sret(%struct.Small) align 4 %agg.result)
 // WIN32: define dso_local i32 @"?small_return@@YA?AUSmall@@XZ"()
 // WIN64: define dso_local i32 @"?small_return@@YA?AUSmall@@XZ"()
 // WOA64: define dso_local i32 @"?small_return@@YA?AUSmall@@XZ"()
 
 Medium medium_return() { return Medium(); }
-// LINUX-LABEL: define{{.*}} void @_Z13medium_returnv(ptr noalias sret(%struct.Medium) align 4 %agg.result)
+// LINUX-LABEL: define{{.*}} void @_Z13medium_returnv(ptr dead_on_unwind noalias writable sret(%struct.Medium) align 4 %agg.result)
 // WIN32: define dso_local i64 @"?medium_return@@YA?AUMedium@@XZ"()
 // WIN64: define dso_local i64 @"?medium_return@@YA?AUMedium@@XZ"()
 // WOA64: define dso_local i64 @"?medium_return@@YA?AUMedium@@XZ"()
 
 // Returning structs that fit into a register but are not POD.
 SmallCpp11NotCpp03Pod small_non_pod_return() { return SmallCpp11NotCpp03Pod(); }
-// LINUX-LABEL: define{{.*}} void @_Z20small_non_pod_returnv(ptr noalias sret(%struct.SmallCpp11NotCpp03Pod) align 4 %agg.result)
-// WIN32: define dso_local void @"?small_non_pod_return@@YA?AUSmallCpp11NotCpp03Pod@@XZ"(ptr noalias sret(%struct.SmallCpp11NotCpp03Pod) align 4 %agg.result)
-// WIN64: define dso_local void @"?small_non_pod_return@@YA?AUSmallCpp11NotCpp03Pod@@XZ"(ptr noalias sret(%struct.SmallCpp11NotCpp03Pod) align 4 %agg.result)
-// WOA64: define dso_local void @"?small_non_pod_return@@YA?AUSmallCpp11NotCpp03Pod@@XZ"(ptr inreg noalias sret(%struct.SmallCpp11NotCpp03Pod) align 4 %agg.result)
+// LINUX-LABEL: define{{.*}} void @_Z20small_non_pod_returnv(ptr dead_on_unwind noalias writable sret(%struct.SmallCpp11NotCpp03Pod) align 4 %agg.result)
+// WIN32: define dso_local void @"?small_non_pod_return@@YA?AUSmallCpp11NotCpp03Pod@@XZ"(ptr dead_on_unwind noalias writable sret(%struct.SmallCpp11NotCpp03Pod) align 4 %agg.result)
+// WIN64: define dso_local void @"?small_non_pod_return@@YA?AUSmallCpp11NotCpp03Pod@@XZ"(ptr dead_on_unwind noalias writable sret(%struct.SmallCpp11NotCpp03Pod) align 4 %agg.result)
+// WOA64: define dso_local void @"?small_non_pod_return@@YA?AUSmallCpp11NotCpp03Pod@@XZ"(ptr dead_on_unwind inreg noalias writable sret(%struct.SmallCpp11NotCpp03Pod) align 4 %agg.result)
 
 SmallWithCtor small_with_ctor_return() { return SmallWithCtor(); }
-// LINUX-LABEL: define{{.*}} void @_Z22small_with_ctor_returnv(ptr noalias sret(%struct.SmallWithCtor) align 4 %agg.result)
-// WIN32: define dso_local void @"?small_with_ctor_return@@YA?AUSmallWithCtor@@XZ"(ptr noalias sret(%struct.SmallWithCtor) align 4 %agg.result)
-// WIN64: define dso_local void @"?small_with_ctor_return@@YA?AUSmallWithCtor@@XZ"(ptr noalias sret(%struct.SmallWithCtor) align 4 %agg.result)
+// LINUX-LABEL: define{{.*}} void @_Z22small_with_ctor_returnv(ptr dead_on_unwind noalias writable sret(%struct.SmallWithCtor) align 4 %agg.result)
+// WIN32: define dso_local void @"?small_with_ctor_return@@YA?AUSmallWithCtor@@XZ"(ptr dead_on_unwind noalias writable sret(%struct.SmallWithCtor) align 4 %agg.result)
+// WIN64: define dso_local void @"?small_with_ctor_return@@YA?AUSmallWithCtor@@XZ"(ptr dead_on_unwind noalias writable sret(%struct.SmallWithCtor) align 4 %agg.result)
 // FIXME: The 'sret' mark here doesn't seem to be enough to convince LLVM to
 // preserve the hidden sret pointer in R0 across the function.
-// WOA: define dso_local arm_aapcs_vfpcc void @"?small_with_ctor_return@@YA?AUSmallWithCtor@@XZ"(ptr noalias sret(%struct.SmallWithCtor) align 4 %agg.result)
-// WOA64: define dso_local void @"?small_with_ctor_return@@YA?AUSmallWithCtor@@XZ"(ptr inreg noalias sret(%struct.SmallWithCtor) align 4 %agg.result)
+// WOA: define dso_local arm_aapcs_vfpcc void @"?small_with_ctor_return@@YA?AUSmallWithCtor@@XZ"(ptr dead_on_unwind noalias writable sret(%struct.SmallWithCtor) align 4 %agg.result)
+// WOA64: define dso_local void @"?small_with_ctor_return@@YA?AUSmallWithCtor@@XZ"(ptr dead_on_unwind inreg noalias writable sret(%struct.SmallWithCtor) align 4 %agg.result)
 
 SmallWithDtor small_with_dtor_return() { return SmallWithDtor(); }
-// LINUX-LABEL: define{{.*}} void @_Z22small_with_dtor_returnv(ptr noalias sret(%struct.SmallWithDtor) align 4 %agg.result)
-// WIN32: define dso_local void @"?small_with_dtor_return@@YA?AUSmallWithDtor@@XZ"(ptr noalias sret(%struct.SmallWithDtor) align 4 %agg.result)
-// WIN64: define dso_local void @"?small_with_dtor_return@@YA?AUSmallWithDtor@@XZ"(ptr noalias sret(%struct.SmallWithDtor) align 4 %agg.result)
-// WOA64: define dso_local void @"?small_with_dtor_return@@YA?AUSmallWithDtor@@XZ"(ptr inreg noalias sret(%struct.SmallWithDtor) align 4 %agg.result)
+// LINUX-LABEL: define{{.*}} void @_Z22small_with_dtor_returnv(ptr dead_on_unwind noalias writable sret(%struct.SmallWithDtor) align 4 %agg.result)
+// WIN32: define dso_local void @"?small_with_dtor_return@@YA?AUSmallWithDtor@@XZ"(ptr dead_on_unwind noalias writable sret(%struct.SmallWithDtor) align 4 %agg.result)
+// WIN64: define dso_local void @"?small_with_dtor_return@@YA?AUSmallWithDtor@@XZ"(ptr dead_on_unwind noalias writable sret(%struct.SmallWithDtor) align 4 %agg.result)
+// WOA64: define dso_local void @"?small_with_dtor_return@@YA?AUSmallWithDtor@@XZ"(ptr dead_on_unwind inreg noalias writable sret(%struct.SmallWithDtor) align 4 %agg.result)
 
 SmallWithVftable small_with_vftable_return() { return SmallWithVftable(); }
-// LINUX-LABEL: define{{.*}} void @_Z25small_with_vftable_returnv(ptr noalias sret(%struct.SmallWithVftable) align 4 %agg.result)
-// WIN32: define dso_local void @"?small_with_vftable_return@@YA?AUSmallWithVftable@@XZ"(ptr noalias sret(%struct.SmallWithVftable) align 4 %agg.result)
-// WIN64: define dso_local void @"?small_with_vftable_return@@YA?AUSmallWithVftable@@XZ"(ptr noalias sret(%struct.SmallWithVftable) align 8 %agg.result)
-// WOA64: define dso_local void @"?small_with_vftable_return@@YA?AUSmallWithVftable@@XZ"(ptr inreg noalias sret(%struct.SmallWithVftable) align 8 %agg.result)
+// LINUX-LABEL: define{{.*}} void @_Z25small_with_vftable_returnv(ptr dead_on_unwind noalias writable sret(%struct.SmallWithVftable) align 4 %agg.result)
+// WIN32: define dso_local void @"?small_with_vftable_return@@YA?AUSmallWithVftable@@XZ"(ptr dead_on_unwind noalias writable sret(%struct.SmallWithVftable) align 4 %agg.result)
+// WIN64: define dso_local void @"?small_with_vftable_return@@YA?AUSmallWithVftable@@XZ"(ptr dead_on_unwind noalias writable sret(%struct.SmallWithVftable) align 8 %agg.result)
+// WOA64: define dso_local void @"?small_with_vftable_return@@YA?AUSmallWithVftable@@XZ"(ptr dead_on_unwind inreg noalias writable sret(%struct.SmallWithVftable) align 8 %agg.result)
 
 MediumWithCopyCtor medium_with_copy_ctor_return() { return MediumWithCopyCtor(); }
-// LINUX-LABEL: define{{.*}} void @_Z28medium_with_copy_ctor_returnv(ptr noalias sret(%struct.MediumWithCopyCtor) align 4 %agg.result)
-// WIN32: define dso_local void @"?medium_with_copy_ctor_return@@YA?AUMediumWithCopyCtor@@XZ"(ptr noalias sret(%struct.MediumWithCopyCtor) align 4 %agg.result)
-// WIN64: define dso_local void @"?medium_with_copy_ctor_return@@YA?AUMediumWithCopyCtor@@XZ"(ptr noalias sret(%struct.MediumWithCopyCtor) align 4 %agg.result)
-// WOA: define dso_local arm_aapcs_vfpcc void @"?medium_with_copy_ctor_return@@YA?AUMediumWithCopyCtor@@XZ"(ptr noalias sret(%struct.MediumWithCopyCtor) align 4 %agg.result)
-// WOA64: define dso_local void @"?medium_with_copy_ctor_return@@YA?AUMediumWithCopyCtor@@XZ"(ptr inreg noalias sret(%struct.MediumWithCopyCtor) align 4 %agg.result)
+// LINUX-LABEL: define{{.*}} void @_Z28medium_with_copy_ctor_returnv(ptr dead_on_unwind noalias writable sret(%struct.MediumWithCopyCtor) align 4 %agg.result)
+// WIN32: define dso_local void @"?medium_with_copy_ctor_return@@YA?AUMediumWithCopyCtor@@XZ"(ptr dead_on_unwind noalias writable sret(%struct.MediumWithCopyCtor) align 4 %agg.result)
+// WIN64: define dso_local void @"?medium_with_copy_ctor_return@@YA?AUMediumWithCopyCtor@@XZ"(ptr dead_on_unwind noalias writable sret(%struct.MediumWithCopyCtor) align 4 %agg.result)
+// WOA: define dso_local arm_aapcs_vfpcc void @"?medium_with_copy_ctor_return@@YA?AUMediumWithCopyCtor@@XZ"(ptr dead_on_unwind noalias writable sret(%struct.MediumWithCopyCtor) align 4 %agg.result)
+// WOA64: define dso_local void @"?medium_with_copy_ctor_return@@YA?AUMediumWithCopyCtor@@XZ"(ptr dead_on_unwind inreg noalias writable sret(%struct.MediumWithCopyCtor) align 4 %agg.result)
 
 // Returning a large struct that doesn't fit into a register.
 Big big_return() { return Big(); }
-// LINUX-LABEL: define{{.*}} void @_Z10big_returnv(ptr noalias sret(%struct.Big) align 4 %agg.result)
-// WIN32: define dso_local void @"?big_return@@YA?AUBig@@XZ"(ptr noalias sret(%struct.Big) align 4 %agg.result)
-// WIN64: define dso_local void @"?big_return@@YA?AUBig@@XZ"(ptr noalias sret(%struct.Big) align 4 %agg.result)
-// WOA64: define dso_local void @"?big_return@@YA?AUBig@@XZ"(ptr noalias sret(%struct.Big) align 4 %agg.result)
+// LINUX-LABEL: define{{.*}} void @_Z10big_returnv(ptr dead_on_unwind noalias writable sret(%struct.Big) align 4 %agg.result)
+// WIN32: define dso_local void @"?big_return@@YA?AUBig@@XZ"(ptr dead_on_unwind noalias writable sret(%struct.Big) align 4 %agg.result)
+// WIN64: define dso_local void @"?big_return@@YA?AUBig@@XZ"(ptr dead_on_unwind noalias writable sret(%struct.Big) align 4 %agg.result)
+// WOA64: define dso_local void @"?big_return@@YA?AUBig@@XZ"(ptr dead_on_unwind noalias writable sret(%struct.Big) align 4 %agg.result)
 
 
 void small_arg(Small s) {}
@@ -197,7 +197,7 @@ void small_arg_with_dtor(SmallWithDtor s) {}
 
 // Test that the eligible non-aggregate is passed directly, but returned
 // indirectly on ARM64 Windows.
-// WOA64: define dso_local void @"?small_arg_with_private_member@@YA?AUSmallWithPrivate@@U1@@Z"(ptr inreg noalias sret(%struct.SmallWithPrivate) align 4 %agg.result, i64 %s.coerce) {{.*}} {
+// WOA64: define dso_local void @"?small_arg_with_private_member@@YA?AUSmallWithPrivate@@U1@@Z"(ptr dead_on_unwind inreg noalias writable sret(%struct.SmallWithPrivate) align 4 %agg.result, i64 %s.coerce) {{.*}} {
 SmallWithPrivate small_arg_with_private_member(SmallWithPrivate s) { return s; }
 
 // WOA64: define dso_local i32 @"?small_arg_with_small_struct_with_private_member@@YA?AUSmallWithSmallWithPrivate@@U1@@Z"(i64 %s.coerce) {{.*}} {
@@ -301,27 +301,27 @@ void pass_ref_field() {
 class Class {
  public:
   Small thiscall_method_small() { return Small(); }
-  // LINUX: define {{.*}} void @_ZN5Class21thiscall_method_smallEv(ptr noalias sret(%struct.Small) align 4 %agg.result, ptr {{[^,]*}} %this)
-  // WIN32: define {{.*}} x86_thiscallcc void @"?thiscall_method_small@Class@@QAE?AUSmall@@XZ"(ptr {{[^,]*}} %this, ptr noalias sret(%struct.Small) align 4 %agg.result)
-  // WIN64: define linkonce_odr dso_local void @"?thiscall_method_small@Class@@QEAA?AUSmall@@XZ"(ptr {{[^,]*}} %this, ptr noalias sret(%struct.Small) align 4 %agg.result)
-  // WOA64: define linkonce_odr dso_local void @"?thiscall_method_small@Class@@QEAA?AUSmall@@XZ"(ptr {{[^,]*}} %this, ptr inreg noalias sret(%struct.Small) align 4 %agg.result)
+  // LINUX: define {{.*}} void @_ZN5Class21thiscall_method_smallEv(ptr dead_on_unwind noalias writable sret(%struct.Small) align 4 %agg.result, ptr {{[^,]*}} %this)
+  // WIN32: define {{.*}} x86_thiscallcc void @"?thiscall_method_small@Class@@QAE?AUSmall@@XZ"(ptr {{[^,]*}} %this, ptr dead_on_unwind noalias writable sret(%struct.Small) align 4 %agg.result)
+  // WIN64: define linkonce_odr dso_local void @"?thiscall_method_small@Class@@QEAA?AUSmall@@XZ"(ptr {{[^,]*}} %this, ptr dead_on_unwind noalias writable sret(%struct.Small) align 4 %agg.result)
+  // WOA64: define linkonce_odr dso_local void @"?thiscall_method_small@Class@@QEAA?AUSmall@@XZ"(ptr {{[^,]*}} %this, ptr dead_on_unwind inreg noalias writable sret(%struct.Small) align 4 %agg.result)
 
   SmallWithCtor thiscall_method_small_with_ctor() { return SmallWithCtor(); }
-  // LINUX: define {{.*}} void @_ZN5Class31thiscall_method_small_with_ctorEv(ptr noalias sret(%struct.SmallWithCtor) align 4 %agg.result, ptr {{[^,]*}} %this)
-  // WIN32: define {{.*}} x86_thiscallcc void @"?thiscall_method_small_with_ctor@Class@@QAE?AUSmallWithCtor@@XZ"(ptr {{[^,]*}} %this, ptr noalias sret(%struct.SmallWithCtor) align 4 %agg.result)
-  // WIN64: define linkonce_odr dso_local void @"?thiscall_method_small_with_ctor@Class@@QEAA?AUSmallWithCtor@@XZ"(ptr {{[^,]*}} %this, ptr noalias sret(%struct.SmallWithCtor) align 4 %agg.result)
-  // WOA64: define linkonce_odr dso_local void @"?thiscall_method_small_with_ctor@Class@@QEAA?AUSmallWithCtor@@XZ"(ptr {{[^,]*}} %this, ptr inreg noalias sret(%struct.SmallWithCtor) align 4 %agg.result)
+  // LINUX: define {{.*}} void @_ZN5Class31thiscall_method_small_with_ctorEv(ptr dead_on_unwind noalias writable sret(%struct.SmallWithCtor) align 4 %agg.result, ptr {{[^,]*}} %this)
+  // WIN32: define {{.*}} x86_thiscallcc void @"?thiscall_method_small_with_ctor@Class@@QAE?AUSmallWithCtor@@XZ"(ptr {{[^,]*}} %this, ptr dead_on_unwind noalias writable sret(%struct.SmallWithCtor) align 4 %agg.result)
+  // WIN64: define linkonce_odr dso_local void @"?thiscall_method_small_with_ctor@Class@@QEAA?AUSmallWithCtor@@XZ"(ptr {{[^,]*}} %this, ptr dead_on_unwind noalias writable sret(%struct.SmallWithCtor) align 4 %agg.result)
+  // WOA64: define linkonce_odr dso_local void @"?thiscall_method_small_with_ctor@Class@@QEAA?AUSmallWithCtor@@XZ"(ptr {{[^,]*}} %this, ptr dead_on_unwind inreg noalias writable sret(%struct.SmallWithCtor) align 4 %agg.result)
 
   Small __cdecl cdecl_method_small() { return Small(); }
-  // LINUX: define {{.*}} void @_ZN5Class18cdecl_method_smallEv(ptr noalias sret(%struct.Small) align 4 %agg.result, ptr {{[^,]*}} %this)
-  // WIN32: define {{.*}} void @"?cdecl_method_small@Class@@QAA?AUSmall@@XZ"(ptr {{[^,]*}} %this, ptr noalias sret(%struct.Small) align 4 %agg.result)
-  // WIN64: define linkonce_odr dso_local void @"?cdecl_method_small@Class@@QEAA?AUSmall@@XZ"(ptr {{[^,]*}} %this, ptr noalias sret(%struct.Small) align 4 %agg.result)
+  // LINUX: define {{.*}} void @_ZN5Class18cdecl_method_smallEv(ptr dead_on_unwind noalias writable sret(%struct.Small) align 4 %agg.result, ptr {{[^,]*}} %this)
+  // WIN32: define {{.*}} void @"?cdecl_method_small@Class@@QAA?AUSmall@@XZ"(ptr {{[^,]*}} %this, ptr dead_on_unwind noalias writable sret(%struct.Small) align 4 %agg.result)
+  // WIN64: define linkonce_odr dso_local void @"?cdecl_method_small@Class@@QEAA?AUSmall@@XZ"(ptr {{[^,]*}} %this, ptr dead_on_unwind noalias writable sret(%struct.Small) align 4 %agg.result)
 
   Big __cdecl cdecl_method_big() { return Big(); }
-  // LINUX: define {{.*}} void @_ZN5Class16cdecl_method_bigEv(ptr noalias sret(%struct.Big) align 4 %agg.result, ptr {{[^,]*}} %this)
-  // WIN32: define {{.*}} void @"?cdecl_method_big@Class@@QAA?AUBig@@XZ"(ptr {{[^,]*}} %this, ptr noalias sret(%struct.Big) align 4 %agg.result)
-  // WIN64: define linkonce_odr dso_local void @"?cdecl_method_big@Class@@QEAA?AUBig@@XZ"(ptr {{[^,]*}} %this, ptr noalias sret(%struct.Big) align 4 %agg.result)
-  // WOA64: define linkonce_odr dso_local void @"?cdecl_method_big@Class@@QEAA?AUBig@@XZ"(ptr {{[^,]*}} %this, ptr inreg noalias sret(%struct.Big) align 4 %agg.result)
+  // LINUX: define {{.*}} void @_ZN5Class16cdecl_method_bigEv(ptr dead_on_unwind noalias writable sret(%struct.Big) align 4 %agg.result, ptr {{[^,]*}} %this)
+  // WIN32: define {{.*}} void @"?cdecl_method_big@Class@@QAA?AUBig@@XZ"(ptr {{[^,]*}} %this, ptr dead_on_unwind noalias writable sret(%struct.Big) align 4 %agg.result)
+  // WIN64: define linkonce_odr dso_local void @"?cdecl_method_big@Class@@QEAA?AUBig@@XZ"(ptr {{[^,]*}} %this, ptr dead_on_unwind noalias writable sret(%struct.Big) align 4 %agg.result)
+  // WOA64: define linkonce_odr dso_local void @"?cdecl_method_big@Class@@QEAA?AUBig@@XZ"(ptr {{[^,]*}} %this, ptr dead_on_unwind inreg noalias writable sret(%struct.Big) align 4 %agg.result)
 
   void thiscall_method_arg(Empty s) {}
   // LINUX: define {{.*}} void @_ZN5Class19thiscall_method_argE5Empty(ptr {{[^,]*}} %this)
diff --git a/clang/test/CodeGenCXX/microsoft-abi-unknown-arch.cpp b/clang/test/CodeGenCXX/microsoft-abi-unknown-arch.cpp
index 3045e90777cbc..9e37e71e257fd 100644
--- a/clang/test/CodeGenCXX/microsoft-abi-unknown-arch.cpp
+++ b/clang/test/CodeGenCXX/microsoft-abi-unknown-arch.cpp
@@ -18,4 +18,4 @@ A B::foo(A x) {
   return x;
 }
 
-// CHECK-LABEL: define{{.*}} void @"?foo@B@@QEAA?AUA@@U2@@Z"(ptr {{[^,]*}} %this, ptr noalias sret(%struct.A) align 4 %agg.result, ptr noundef %x)
+// CHECK-LABEL: define{{.*}} void @"?foo@B@@QEAA?AUA@@U2@@Z"(ptr {{[^,]*}} %this, ptr dead_on_unwind noalias writable sret(%struct.A) align 4 %agg.result, ptr noundef %x)
diff --git a/clang/test/CodeGenCXX/microsoft-abi-vmemptr-conflicts.cpp b/clang/test/CodeGenCXX/microsoft-abi-vmemptr-conflicts.cpp
index 2a29a080d025a..615a2e34d7a2b 100644
--- a/clang/test/CodeGenCXX/microsoft-abi-vmemptr-conflicts.cpp
+++ b/clang/test/CodeGenCXX/microsoft-abi-vmemptr-conflicts.cpp
@@ -65,7 +65,7 @@ void f(C *c) {
 
 // CHECK-LABEL: define dso_local void @"?f@sret@@YAXPAUC@1@@Z"(ptr noundef %c)
 // CHECK: call x86_thiscallcc noundef i32 @"??_9C@sret@@$BA@AE"(ptr {{[^,]*}} %{{.*}})
-// CHECK: call x86_thiscallcc void @"??_9C@sret@@$BA@AE"(ptr {{[^,]*}} %{{.*}}, ptr sret(%"struct.sret::Big") align 4 %{{.*}})
+// CHECK: call x86_thiscallcc void @"??_9C@sret@@$BA@AE"(ptr {{[^,]*}} %{{.*}}, ptr dead_on_unwind writable sret(%"struct.sret::Big") align 4 %{{.*}})
 
 // CHECK-LABEL: define linkonce_odr x86_thiscallcc void @"??_9C@sret@@$BA@AE"(ptr noundef %this, ...) {{.*}} comdat
 // CHECK: musttail call x86_thiscallcc void (ptr, ...) %{{.*}}(ptr noundef %{{.*}}, ...)
diff --git a/clang/test/CodeGenCXX/module-initializer-guard-elision.cpp b/clang/test/CodeGenCXX/module-initializer-guard-elision.cpp
index bae7c930b5102..53e4b909ee2a1 100644
--- a/clang/test/CodeGenCXX/module-initializer-guard-elision.cpp
+++ b/clang/test/CodeGenCXX/module-initializer-guard-elision.cpp
@@ -8,9 +8,9 @@
 // RUN:  -o - | FileCheck %s --check-prefix=CHECK-O
 
 // RUN: %clang_cc1 -triple %itanium_abi_triple -std=c++20 P.cpp \
-// RUN:    -emit-module-interface -fmodule-file=O=O.pcm -o P.pcm
+// RUN:    -emit-module-interface -fprebuilt-module-path=%t -o P.pcm
 // RUN: %clang_cc1 -triple %itanium_abi_triple -std=c++20 P.pcm -S -emit-llvm \
-// RUN:  -o - | FileCheck %s --check-prefix=CHECK-P
+// RUN:   -fprebuilt-module-path=%t -o - | FileCheck %s --check-prefix=CHECK-P
 
 // RUN: %clang_cc1 -triple %itanium_abi_triple -std=c++20 Q.cpp \
 // RUN:    -emit-module-interface -o Q.pcm
@@ -18,24 +18,24 @@
 // RUN:    -o - | FileCheck %s --check-prefix=CHECK-Q
 
 // RUN: %clang_cc1 -triple %itanium_abi_triple -std=c++20 R.cpp \
-// RUN:    -emit-module-interface -fmodule-file=Q=Q.pcm -o R.pcm
+// RUN:    -emit-module-interface -fprebuilt-module-path=%t -o R.pcm
 // RUN: %clang_cc1 -triple %itanium_abi_triple -std=c++20 R.pcm -S -emit-llvm \
-// RUN:    -o - | FileCheck %s --check-prefix=CHECK-R
+// RUN:    -fprebuilt-module-path=%t -o - | FileCheck %s --check-prefix=CHECK-R
 
 // RUN: %clang_cc1 -triple %itanium_abi_triple -std=c++20 S.cpp \
-// RUN:    -emit-module-interface -fmodule-file=Q=Q.pcm -fmodule-file=R=R.pcm -o S.pcm
+// RUN:    -emit-module-interface -fprebuilt-module-path=%t -o S.pcm
 // RUN: %clang_cc1 -triple %itanium_abi_triple -std=c++20 S.pcm -S -emit-llvm \
-// RUN:    -o - | FileCheck %s --check-prefix=CHECK-S
+// RUN:    -fprebuilt-module-path=%t -o - | FileCheck %s --check-prefix=CHECK-S
 
 // RUN: %clang_cc1 -triple %itanium_abi_triple -std=c++20 T.cpp \
-// RUN:    -emit-module-interface -fmodule-file=S=S.pcm -fmodule-file=R=R.pcm -o T.pcm
+// RUN:    -emit-module-interface -fprebuilt-module-path=%t -o T.pcm
 // RUN: %clang_cc1 -triple %itanium_abi_triple -std=c++20 T.pcm -S -emit-llvm \
-// RUN:    -o - | FileCheck %s --check-prefix=CHECK-T
+// RUN:    -fprebuilt-module-path=%t -o - | FileCheck %s --check-prefix=CHECK-T
 
 // RUN: %clang_cc1 -triple %itanium_abi_triple -std=c++20 U.cpp \
-// RUN:    -emit-module-interface -fmodule-file=T=T.pcm -fmodule-file=R=R.pcm -o U.pcm
+// RUN:    -emit-module-interface -fprebuilt-module-path=%t -o U.pcm
 // RUN: %clang_cc1 -triple %itanium_abi_triple -std=c++20 U.pcm -S -emit-llvm \
-// RUN:    -o - | FileCheck %s --check-prefix=CHECK-U
+// RUN:    -fprebuilt-module-path=%t -o - | FileCheck %s --check-prefix=CHECK-U
 
 // Testing cases where we can elide the module initializer guard variable.
 
diff --git a/clang/test/CodeGenCXX/module-intializer.cpp b/clang/test/CodeGenCXX/module-intializer.cpp
index d365d180ac59d..8a464ae7403d6 100644
--- a/clang/test/CodeGenCXX/module-intializer.cpp
+++ b/clang/test/CodeGenCXX/module-intializer.cpp
@@ -12,24 +12,23 @@
 // RUN: %clang_cc1 -triple %itanium_abi_triple -std=c++20 O.pcm -S -emit-llvm \
 // RUN:  -o - | FileCheck %s --check-prefix=CHECK-O
 
-// RUN: %clang_cc1 -triple %itanium_abi_triple -std=c++20 M-part.cpp \
-// RUN:    -emit-module-interface -o M-part.pcm
-// RUN: %clang_cc1 -triple %itanium_abi_triple -std=c++20 M-part.pcm -S \
-// RUN: -emit-llvm -o - | FileCheck %s --check-prefix=CHECK-P
+// RUN: %clang_cc1 -triple %itanium_abi_triple -std=c++20 M-Part.cpp \
+// RUN:    -emit-module-interface -o M-Part.pcm
+// RUN: %clang_cc1 -triple %itanium_abi_triple -std=c++20 M-Part.pcm -S \
+// RUN:    -emit-module-interface  -emit-llvm -o - | FileCheck %s --check-prefix=CHECK-P
 
 // RUN: %clang_cc1 -triple %itanium_abi_triple -std=c++20 M.cpp \
-// RUN: -fmodule-file=N=N.pcm -fmodule-file=O=O.pcm -fmodule-file=M:Part=M-part.pcm \
-// RUN:    -emit-module-interface -o M.pcm
+// RUN:    -fprebuilt-module-path=%t -emit-module-interface -o M.pcm
 // RUN: %clang_cc1 -triple %itanium_abi_triple -std=c++20 M.pcm -S -emit-llvm \
-// RUN:  -o - | FileCheck %s --check-prefix=CHECK-M
+// RUN:    -fprebuilt-module-path=%t -o - | FileCheck %s --check-prefix=CHECK-M
 
 // RUN: %clang_cc1 -triple %itanium_abi_triple -std=c++20 useM.cpp \
-// RUN: -fmodule-file=M=M.pcm -S -emit-llvm  -o - \
-// RUN: | FileCheck %s --check-prefix=CHECK-USE
+// RUN:   -fprebuilt-module-path=%t -S -emit-llvm  -o - \
+// RUN:   | FileCheck %s --check-prefix=CHECK-USE
 
 // RUN: %clang_cc1 -triple %itanium_abi_triple -std=c++20 M-impl.cpp \
-// RUN: -fmodule-file=M=M.pcm -S -emit-llvm  -o - \
-// RUN: | FileCheck %s --check-prefix=CHECK-IMPL
+// RUN:   -fprebuilt-module-path=%t -S -emit-llvm  -o - \
+// RUN:   | FileCheck %s --check-prefix=CHECK-IMPL
 
 // RUN: %clang_cc1 -triple %itanium_abi_triple -std=c++20 N.cpp -S -emit-llvm \
 // RUN:   -o - | FileCheck %s --check-prefix=CHECK-N
@@ -37,12 +36,11 @@
 // RUN: %clang_cc1 -triple %itanium_abi_triple -std=c++20 O.cpp -S -emit-llvm \
 // RUN:   -o - | FileCheck %s --check-prefix=CHECK-O
 
-// RUN: %clang_cc1 -triple %itanium_abi_triple -std=c++20 M-part.cpp -S -emit-llvm \
+// RUN: %clang_cc1 -triple %itanium_abi_triple -std=c++20 M-Part.cpp -S -emit-llvm \
 // RUN:   -o - | FileCheck %s --check-prefix=CHECK-P
 
 // RUN: %clang_cc1 -triple %itanium_abi_triple -std=c++20 M.cpp \
-// RUN:   -fmodule-file=N.pcm -fmodule-file=O=O.pcm -fmodule-file=M:Part=M-part.pcm \
-// RUN:   -S -emit-llvm -o - | FileCheck %s --check-prefix=CHECK-M
+// RUN:   -fprebuilt-module-path=%t -S -emit-llvm -o - | FileCheck %s --check-prefix=CHECK-M
 
 //--- N-h.h
 
@@ -112,7 +110,7 @@ struct Croak {
 
 Croak Frog;
 
-//--- M-part.cpp
+//--- M-Part.cpp
 
 module;
 #include "P-h.h"
diff --git a/clang/test/CodeGenCXX/ms-property-new.cpp b/clang/test/CodeGenCXX/ms-property-new.cpp
new file mode 100644
index 0000000000000..bc326dda6c377
--- /dev/null
+++ b/clang/test/CodeGenCXX/ms-property-new.cpp
@@ -0,0 +1,52 @@
+// RUN: %clang_cc1 -emit-llvm -triple=x86_64-pc-win32 -fms-compatibility %s -o - | FileCheck %s
+// RUN: %clang_cc1 -triple=x86_64-pc-win32 -fms-compatibility -emit-pch -o %t %s
+// RUN: %clang_cc1 -emit-llvm -triple=x86_64-pc-win32 -fms-compatibility -include-pch %t -verify %s -o - | FileCheck %s
+// expected-no-diagnostics
+
+#ifndef HEADER
+#define HEADER
+
+struct S {
+  int GetX() { return 42; }
+  __declspec(property(get=GetX)) int x;
+
+  int GetY(int i, int j) { return i+j; }
+  __declspec(property(get=GetY)) int y[];
+
+  void* operator new(__SIZE_TYPE__, int);
+};
+
+template 
+struct TS {
+  T GetT() { return T(); }
+  __declspec(property(get=GetT)) T t;
+
+  T GetR(T i, T j) { return i+j; }
+  __declspec(property(get=GetR)) T r[];
+};
+
+// CHECK-LABEL: main
+int main(int argc, char **argv) {
+  S *s;
+  TS *ts;
+
+  // CHECK: [[X:%.+]] = call noundef i32 @"?GetX@S@@QEAAHXZ"(ptr {{[^,]*}} %{{.+}})
+  // CHECK-NEXT: call noundef ptr @"??2S@@SAPEAX_KH@Z"(i64 noundef 1, i32 noundef [[X]])
+  new (s->x) S;
+
+  // CHECK: [[Y:%.+]] = call noundef i32 @"?GetY@S@@QEAAHHH@Z"(ptr {{[^,]*}} %{{.+}}, i32 noundef 1, i32 noundef 2)
+  // CHECK-NEXT: call noundef ptr @"??2S@@SAPEAX_KH@Z"(i64 noundef 1, i32 noundef [[Y]])
+  new ((s->y)[1][2]) S;
+
+  // CHECK: [[T:%.+]] = call noundef double @"?GetT@?$TS@N@@QEAANXZ"(ptr {{[^,]*}} %{{.+}})
+  // CHECK-NEXT: [[TI:%.+]] = fptosi double [[T]] to i32
+  // CHECK-NEXT: call noundef ptr @"??2S@@SAPEAX_KH@Z"(i64 noundef 1, i32 noundef [[TI]])
+  new (ts->t) S;
+
+  // CHECK: [[R:%.+]] = call noundef double @"?GetR@?$TS@N@@QEAANNN@Z"(ptr {{[^,]*}} %{{.+}}, double {{[^,]*}}, double {{[^,]*}})
+  // CHECK-NEXT: [[RI:%.+]] = fptosi double [[R]] to i32
+  // CHECK-NEXT: call noundef ptr @"??2S@@SAPEAX_KH@Z"(i64 noundef 1, i32 noundef [[RI]])
+  new ((ts->r)[1][2]) S;
+}
+
+#endif
diff --git a/clang/test/CodeGenCXX/ms-thread_local.cpp b/clang/test/CodeGenCXX/ms-thread_local.cpp
index a6c62967f23de..cb0e8720c19b8 100644
--- a/clang/test/CodeGenCXX/ms-thread_local.cpp
+++ b/clang/test/CodeGenCXX/ms-thread_local.cpp
@@ -29,9 +29,9 @@ thread_local A b;
 // CHECK-LD-LABEL: declare dso_local void @__dyn_tls_on_demand_init()
 // CHECK-LEGACY-NOT: declare dso_local void @__dyn_tls_on_demand_init()
 
-// CHECK-LABEL: define dso_local void @"?f@@YA?AUA@@XZ"(ptr noalias sret(%struct.A) align 1 %agg.result)
+// CHECK-LABEL: define dso_local void @"?f@@YA?AUA@@XZ"(ptr dead_on_unwind noalias writable sret(%struct.A) align 1 %agg.result)
 // CHECK: call void @__dyn_tls_on_demand_init()
-// CHECK-LD-LABEL: define dso_local void @"?f@@YA?AUA@@XZ"(ptr noalias sret(%struct.A) align 1 %agg.result)
+// CHECK-LD-LABEL: define dso_local void @"?f@@YA?AUA@@XZ"(ptr dead_on_unwind noalias writable sret(%struct.A) align 1 %agg.result)
 // CHECK-LD: call void @__dyn_tls_on_demand_init()
 // CHECK-LEGACY-NOT: call void @__dyn_tls_on_demand_init()
 
diff --git a/clang/test/CodeGenCXX/nrvo.cpp b/clang/test/CodeGenCXX/nrvo.cpp
index 975cf6ee1b66d..33dc4cf9dbc8d 100644
--- a/clang/test/CodeGenCXX/nrvo.cpp
+++ b/clang/test/CodeGenCXX/nrvo.cpp
@@ -1059,7 +1059,7 @@ X test8(bool b) {
 // CHECK-NEXT:    [[RESULT_PTR:%.*]] = alloca ptr, align 4
 // CHECK-NEXT:    [[TMP:%.*]] = alloca [[STRUCT_Y:%.*]], align 1
 // CHECK-NEXT:    store ptr [[AGG_RESULT:%.*]], ptr [[RESULT_PTR]], align 4
-// CHECK-NEXT:    call void @_ZN1YIiE1fEv(ptr sret([[STRUCT_Y]]) align 1 [[TMP]])
+// CHECK-NEXT:    call void @_ZN1YIiE1fEv(ptr dead_on_unwind writable sret([[STRUCT_Y]]) align 1 [[TMP]])
 // CHECK-NEXT:    call void @llvm.trap()
 // CHECK-NEXT:    unreachable
 //
@@ -1068,7 +1068,7 @@ X test8(bool b) {
 // CHECK-EH-03-NEXT:    [[RESULT_PTR:%.*]] = alloca ptr, align 4
 // CHECK-EH-03-NEXT:    [[TMP:%.*]] = alloca [[STRUCT_Y:%.*]], align 1
 // CHECK-EH-03-NEXT:    store ptr [[AGG_RESULT:%.*]], ptr [[RESULT_PTR]], align 4
-// CHECK-EH-03-NEXT:    call void @_ZN1YIiE1fEv(ptr sret([[STRUCT_Y]]) align 1 [[TMP]])
+// CHECK-EH-03-NEXT:    call void @_ZN1YIiE1fEv(ptr dead_on_unwind writable sret([[STRUCT_Y]]) align 1 [[TMP]])
 // CHECK-EH-03-NEXT:    call void @llvm.trap()
 // CHECK-EH-03-NEXT:    unreachable
 //
@@ -1077,7 +1077,7 @@ X test8(bool b) {
 // CHECK-EH-11-NEXT:    [[RESULT_PTR:%.*]] = alloca ptr, align 4
 // CHECK-EH-11-NEXT:    [[TMP:%.*]] = alloca [[STRUCT_Y:%.*]], align 1
 // CHECK-EH-11-NEXT:    store ptr [[AGG_RESULT:%.*]], ptr [[RESULT_PTR]], align 4
-// CHECK-EH-11-NEXT:    call void @_ZN1YIiE1fEv(ptr sret([[STRUCT_Y]]) align 1 [[TMP]])
+// CHECK-EH-11-NEXT:    call void @_ZN1YIiE1fEv(ptr dead_on_unwind writable sret([[STRUCT_Y]]) align 1 [[TMP]])
 // CHECK-EH-11-NEXT:    call void @llvm.trap()
 // CHECK-EH-11-NEXT:    unreachable
 //
@@ -1915,7 +1915,7 @@ X test15(bool b) { // http://wg21.link/p2025r2#ex-15
 // CHECK-EH-11-NEXT:    call void @_ZN1XC1Ev(ptr noundef nonnull align 1 dereferenceable(1) [[X]])
 // CHECK-EH-11-NEXT:    [[TMP0:%.*]] = getelementptr inbounds [[CLASS_ANON]], ptr [[REF_TMP]], i32 0, i32 0
 // CHECK-EH-11-NEXT:    store ptr [[X]], ptr [[TMP0]], align 4
-// CHECK-EH-11-NEXT:    invoke void @"_ZZ6test16vENK3$_0clEv"(ptr sret([[CLASS_X]]) align 1 [[AGG_TMP]], ptr noundef nonnull align 4 dereferenceable(4) [[REF_TMP]])
+// CHECK-EH-11-NEXT:    invoke void @"_ZZ6test16vENK3$_0clEv"(ptr dead_on_unwind writable sret([[CLASS_X]]) align 1 [[AGG_TMP]], ptr noundef nonnull align 4 dereferenceable(4) [[REF_TMP]])
 // CHECK-EH-11-NEXT:            to label [[INVOKE_CONT:%.*]] unwind label [[LPAD:%.*]]
 // CHECK-EH-11:       invoke.cont:
 // CHECK-EH-11-NEXT:    invoke void @_Z8ConsumeX1X(ptr noundef [[AGG_TMP]])
@@ -2530,7 +2530,7 @@ X test18(int i) { // http://wg21.link/p2025r2#ex-11
 // CHECK-EH-11-NEXT:    call void @_ZN1XC1Ev(ptr noundef nonnull align 1 dereferenceable(1) [[AGG_RESULT]])
 // CHECK-EH-11-NEXT:    [[TMP0:%.*]] = getelementptr inbounds [[CLASS_ANON_0]], ptr [[REF_TMP]], i32 0, i32 0
 // CHECK-EH-11-NEXT:    store ptr [[AGG_RESULT]], ptr [[TMP0]], align 4
-// CHECK-EH-11-NEXT:    invoke void @"_ZZ6test19vENK3$_0clEv"(ptr sret([[CLASS_X]]) align 1 [[L]], ptr noundef nonnull align 4 dereferenceable(4) [[REF_TMP]])
+// CHECK-EH-11-NEXT:    invoke void @"_ZZ6test19vENK3$_0clEv"(ptr dead_on_unwind writable sret([[CLASS_X]]) align 1 [[L]], ptr noundef nonnull align 4 dereferenceable(4) [[REF_TMP]])
 // CHECK-EH-11-NEXT:            to label [[INVOKE_CONT:%.*]] unwind label [[LPAD:%.*]]
 // CHECK-EH-11:       invoke.cont:
 // CHECK-EH-11-NEXT:    store i1 true, ptr [[NRVO]], align 1
@@ -2584,9 +2584,9 @@ X test20() { // http://wg21.link/p2025r2#ex-18
 // CHECK-EH-11-NEXT:  entry:
 // CHECK-EH-11-NEXT:    [[AGG_TMP_ENSURED:%.*]] = alloca [[CLASS_X:%.*]], align 1
 // CHECK-EH-11-NEXT:    [[AGG_TMP_ENSURED1:%.*]] = alloca [[CLASS_X]], align 1
-// CHECK-EH-11-NEXT:    call void @_Z6test20ILb1EE1Xv(ptr sret([[CLASS_X]]) align 1 [[AGG_TMP_ENSURED]])
+// CHECK-EH-11-NEXT:    call void @_Z6test20ILb1EE1Xv(ptr dead_on_unwind writable sret([[CLASS_X]]) align 1 [[AGG_TMP_ENSURED]])
 // CHECK-EH-11-NEXT:    call void @_ZN1XD1Ev(ptr noundef nonnull align 1 dereferenceable(1) [[AGG_TMP_ENSURED]]) #[[ATTR6]]
-// CHECK-EH-11-NEXT:    call void @_Z6test20ILb0EE1Xv(ptr sret([[CLASS_X]]) align 1 [[AGG_TMP_ENSURED1]])
+// CHECK-EH-11-NEXT:    call void @_Z6test20ILb0EE1Xv(ptr dead_on_unwind writable sret([[CLASS_X]]) align 1 [[AGG_TMP_ENSURED1]])
 // CHECK-EH-11-NEXT:    call void @_ZN1XD1Ev(ptr noundef nonnull align 1 dereferenceable(1) [[AGG_TMP_ENSURED1]]) #[[ATTR6]]
 // CHECK-EH-11-NEXT:    ret void
 //
@@ -2934,9 +2934,9 @@ X test25() {
 // CHECK-EH-11-NEXT:  entry:
 // CHECK-EH-11-NEXT:    [[AGG_TMP_ENSURED:%.*]] = alloca [[CLASS_X:%.*]], align 1
 // CHECK-EH-11-NEXT:    [[AGG_TMP_ENSURED1:%.*]] = alloca [[CLASS_X]], align 1
-// CHECK-EH-11-NEXT:    call void @_Z6test25ILb1EE1Xv(ptr sret([[CLASS_X]]) align 1 [[AGG_TMP_ENSURED]])
+// CHECK-EH-11-NEXT:    call void @_Z6test25ILb1EE1Xv(ptr dead_on_unwind writable sret([[CLASS_X]]) align 1 [[AGG_TMP_ENSURED]])
 // CHECK-EH-11-NEXT:    call void @_ZN1XD1Ev(ptr noundef nonnull align 1 dereferenceable(1) [[AGG_TMP_ENSURED]]) #[[ATTR6]]
-// CHECK-EH-11-NEXT:    call void @_Z6test25ILb0EE1Xv(ptr sret([[CLASS_X]]) align 1 [[AGG_TMP_ENSURED1]])
+// CHECK-EH-11-NEXT:    call void @_Z6test25ILb0EE1Xv(ptr dead_on_unwind writable sret([[CLASS_X]]) align 1 [[AGG_TMP_ENSURED1]])
 // CHECK-EH-11-NEXT:    call void @_ZN1XD1Ev(ptr noundef nonnull align 1 dereferenceable(1) [[AGG_TMP_ENSURED1]]) #[[ATTR6]]
 // CHECK-EH-11-NEXT:    ret void
 //
diff --git a/clang/test/CodeGenCXX/partitions.cpp b/clang/test/CodeGenCXX/partitions.cpp
index b58e88fafdc6b..3b3e69271e7c7 100644
--- a/clang/test/CodeGenCXX/partitions.cpp
+++ b/clang/test/CodeGenCXX/partitions.cpp
@@ -4,14 +4,14 @@
 
 // RUN: %clang_cc1 -std=c++20 -emit-module-interface -triple %itanium_abi_triple %t/parta.cppm -o %t/mod-parta.pcm
 // RUN: %clang_cc1 -std=c++20 -emit-module-interface -triple %itanium_abi_triple %t/partb.cppm -o %t/mod-partb.pcm
-// RUN: %clang_cc1 -std=c++20 -emit-module-interface -triple %itanium_abi_triple -fmodule-file=%t/mod-parta.pcm \
-// RUN:     -fmodule-file=%t/mod-partb.pcm %t/mod.cppm -o %t/mod.pcm
+// RUN: %clang_cc1 -std=c++20 -emit-module-interface -triple %itanium_abi_triple %t/mod.cppm \
+// RUN:   -fprebuilt-module-path=%t -o %t/mod.pcm
 // RUN: %clang_cc1 -std=c++20 -triple %itanium_abi_triple %t/mod.pcm -S -emit-llvm -disable-llvm-passes -o - \
-// RUN:     | FileCheck %t/mod.cppm
-// RUN: %clang_cc1 -std=c++20 -O2 -emit-module-interface -triple %itanium_abi_triple -fmodule-file=%t/mod-parta.pcm \
-// RUN:     -fmodule-file=%t/mod-partb.pcm %t/mod.cppm -o %t/mod.pcm
-// RUN: %clang_cc1 -std=c++20 -O2 -triple %itanium_abi_triple %t/mod.pcm -S -emit-llvm -disable-llvm-passes -o - \
-// RUN:     | FileCheck %t/mod.cppm  -check-prefix=CHECK-OPT
+// RUN:   -fprebuilt-module-path=%t | FileCheck %t/mod.cppm
+// RUN: %clang_cc1 -std=c++20 -O2 -emit-module-interface -triple %itanium_abi_triple \
+// RUN:   -fprebuilt-module-path=%t %t/mod.cppm -o %t/mod.pcm
+// RUN: %clang_cc1 -std=c++20 -O2 -triple %itanium_abi_triple %t/mod.pcm -S -emit-llvm \
+// RUN:   -fprebuilt-module-path=%t -disable-llvm-passes -o - | FileCheck %t/mod.cppm  -check-prefix=CHECK-OPT
 
 //--- parta.cppm
 export module mod:parta;
diff --git a/clang/test/CodeGenCXX/pass-by-value-noalias.cpp b/clang/test/CodeGenCXX/pass-by-value-noalias.cpp
index 765a6fb66a72b..773cf6b81c3b2 100644
--- a/clang/test/CodeGenCXX/pass-by-value-noalias.cpp
+++ b/clang/test/CodeGenCXX/pass-by-value-noalias.cpp
@@ -58,8 +58,8 @@ A *p;
 // NO_NOALIAS: define{{.*}} void @_Z4take1A(ptr noundef %arg)
 void take(A arg) {}
 
-// WITH_NOALIAS: define{{.*}} void @_Z7CreateAPP1A(ptr noalias sret(%struct.A) align 1 %agg.result, ptr noundef %where)
-// NO_NOALIAS: define{{.*}} void @_Z7CreateAPP1A(ptr noalias sret(%struct.A) align 1 %agg.result, ptr noundef %where)
+// WITH_NOALIAS: define{{.*}} void @_Z7CreateAPP1A(ptr dead_on_unwind noalias writable sret(%struct.A) align 1 %agg.result, ptr noundef %where)
+// NO_NOALIAS: define{{.*}} void @_Z7CreateAPP1A(ptr dead_on_unwind noalias writable sret(%struct.A) align 1 %agg.result, ptr noundef %where)
 A CreateA(A **where) {
   A justlikethis;
   *where = &justlikethis; //Escaped pointer 2 (should also be UB, then)
diff --git a/clang/test/CodeGenCXX/personality.cpp b/clang/test/CodeGenCXX/personality.cpp
index 1c8baa2c412fe..42ba2c227f028 100644
--- a/clang/test/CodeGenCXX/personality.cpp
+++ b/clang/test/CodeGenCXX/personality.cpp
@@ -15,6 +15,9 @@
 // RUN: %clang_cc1 -triple powerpc-unknown-aix-xcoff -fexceptions -fcxx-exceptions -S -emit-llvm %s -o - | FileCheck %s -check-prefix CHECK-AIX
 // RUN: %clang_cc1 -triple powerpc64-unknown-aix-xcoff -fexceptions -fcxx-exceptions -S -emit-llvm %s -o - | FileCheck %s -check-prefix CHECK-AIX
 
+// RUN: %clang_cc1 -triple s390x-unknown-zos -fexceptions -fcxx-exceptions -S -emit-llvm %s -o - | FileCheck %s -check-prefix CHECK-ZOS
+// RUN: %clang_cc1 -triple systemz-unknown-zos -fexceptions -fcxx-exceptions -S -emit-llvm %s -o - | FileCheck %s -check-prefix CHECK-ZOS
+
 extern void g();
 
 // CHECK-GNU: personality ptr @__gxx_personality_v0
@@ -26,6 +29,8 @@ extern void g();
 
 // CHECK-AIX: personality ptr @__xlcxx_personality_v1
 
+// CHECK-ZOS: personality ptr @__zos_cxx_personality_v2
+
 void f() {
   try {
     g();
diff --git a/clang/test/CodeGenCXX/placement-new-ms-__noop.cpp b/clang/test/CodeGenCXX/placement-new-ms-__noop.cpp
new file mode 100644
index 0000000000000..96149fdff98b4
--- /dev/null
+++ b/clang/test/CodeGenCXX/placement-new-ms-__noop.cpp
@@ -0,0 +1,12 @@
+// RUN: %clang_cc1 -triple x86_64-windows-msvc -fms-extensions -std=c++11 -emit-llvm -o - %s | FileCheck %s
+
+struct S {
+  void* operator new(__SIZE_TYPE__, int);
+};
+
+int main() {
+  // CHECK: call {{.*}} ptr @"??2S@@SAPEAX_KH@Z"(i64 {{.*}} 1, i32 {{.*}} 0)
+  // CHECK: call {{.*}} ptr @"??2S@@SAPEAX_KH@Z"(i64 {{.*}} 1, i32 {{.*}} 0)
+  new (__noop) S;
+  new ((__noop)) S;
+}
diff --git a/clang/test/CodeGenCXX/regcall.cpp b/clang/test/CodeGenCXX/regcall.cpp
index b47a2125d7873..1ff7597e9deef 100644
--- a/clang/test/CodeGenCXX/regcall.cpp
+++ b/clang/test/CodeGenCXX/regcall.cpp
@@ -74,8 +74,8 @@ bool __regcall operator ==(const test_class&, const test_class&){ --x; return fa
 // CHECK-WIN32-DAG: define dso_local x86_regcallcc noundef zeroext i1 @"??8@Yw_NABVtest_class@@0@Z"
 
 test_class __regcall operator""_test_class (unsigned long long) { ++x; return test_class{};}
-// CHECK-LIN64-DAG: define{{.*}} x86_regcallcc void @_Zli11_test_classy(ptr noalias sret(%class.test_class) align 4 %agg.result, i64 noundef %0)
-// CHECK-LIN32-DAG: define{{.*}} x86_regcallcc void @_Zli11_test_classy(ptr inreg noalias sret(%class.test_class) align 4 %agg.result, i64 noundef %0)
+// CHECK-LIN64-DAG: define{{.*}} x86_regcallcc void @_Zli11_test_classy(ptr dead_on_unwind noalias writable sret(%class.test_class) align 4 %agg.result, i64 noundef %0)
+// CHECK-LIN32-DAG: define{{.*}} x86_regcallcc void @_Zli11_test_classy(ptr dead_on_unwind inreg noalias writable sret(%class.test_class) align 4 %agg.result, i64 noundef %0)
 // CHECK-WIN64-DAG: ??__K_test_class@@Yw?AVtest_class@@_K@Z"
 // CHECK-WIN32-DAG: ??__K_test_class@@Yw?AVtest_class@@_K@Z"
 
@@ -99,8 +99,8 @@ void force_gen() {
 long double _Complex __regcall foo(long double _Complex f) {
   return f;
 }
-// CHECK-LIN64-DAG: define{{.*}} x86_regcallcc void @_Z15__regcall3__fooCe(ptr noalias sret({ x86_fp80, x86_fp80 }) align 16 %agg.result, ptr noundef byval({ x86_fp80, x86_fp80 }) align 16 %f)
-// CHECK-LIN32-DAG: define{{.*}} x86_regcallcc void @_Z15__regcall3__fooCe(ptr inreg noalias sret({ x86_fp80, x86_fp80 }) align 4 %agg.result, ptr noundef byval({ x86_fp80, x86_fp80 }) align 4 %f)
+// CHECK-LIN64-DAG: define{{.*}} x86_regcallcc void @_Z15__regcall3__fooCe(ptr dead_on_unwind noalias writable sret({ x86_fp80, x86_fp80 }) align 16 %agg.result, ptr noundef byval({ x86_fp80, x86_fp80 }) align 16 %f)
+// CHECK-LIN32-DAG: define{{.*}} x86_regcallcc void @_Z15__regcall3__fooCe(ptr dead_on_unwind inreg noalias writable sret({ x86_fp80, x86_fp80 }) align 4 %agg.result, ptr noundef byval({ x86_fp80, x86_fp80 }) align 4 %f)
 // CHECK-WIN64-DAG: define dso_local x86_regcallcc noundef { double, double } @"?foo@@YwU?$_Complex@O@__clang@@U12@@Z"(double noundef %f.0, double noundef %f.1)
 // CHECK-WIN32-DAG: define dso_local x86_regcallcc noundef { double, double } @"?foo@@YwU?$_Complex@O@__clang@@U12@@Z"(double noundef %f.0, double noundef %f.1)
 
diff --git a/clang/test/CodeGenCXX/regcall4.cpp b/clang/test/CodeGenCXX/regcall4.cpp
index 7c35db36e1053..2ecd5b4951c75 100644
--- a/clang/test/CodeGenCXX/regcall4.cpp
+++ b/clang/test/CodeGenCXX/regcall4.cpp
@@ -74,8 +74,8 @@ bool __regcall operator ==(const test_class&, const test_class&){ --x; return fa
 // CHECK-WIN32-DAG: define dso_local x86_regcallcc noundef zeroext i1 @"??8@Yx_NABVtest_class@@0@Z"
 
 test_class __regcall operator""_test_class (unsigned long long) { ++x; return test_class{};}
-// CHECK-LIN64-DAG: define{{.*}} x86_regcallcc void @_Zli11_test_classy(ptr noalias sret(%class.test_class) align 4 %agg.result, i64 noundef %0)
-// CHECK-LIN32-DAG: define{{.*}} x86_regcallcc void @_Zli11_test_classy(ptr inreg noalias sret(%class.test_class) align 4 %agg.result, i64 noundef %0)
+// CHECK-LIN64-DAG: define{{.*}} x86_regcallcc void @_Zli11_test_classy(ptr dead_on_unwind noalias writable sret(%class.test_class) align 4 %agg.result, i64 noundef %0)
+// CHECK-LIN32-DAG: define{{.*}} x86_regcallcc void @_Zli11_test_classy(ptr dead_on_unwind inreg noalias writable sret(%class.test_class) align 4 %agg.result, i64 noundef %0)
 // CHECK-WIN64-DAG: ??__K_test_class@@Yx?AVtest_class@@_K@Z"
 // CHECK-WIN32-DAG: ??__K_test_class@@Yx?AVtest_class@@_K@Z"
 
@@ -99,8 +99,8 @@ void force_gen() {
 long double _Complex __regcall foo(long double _Complex f) {
   return f;
 }
-// CHECK-LIN64-DAG: define{{.*}} x86_regcallcc void @_Z15__regcall4__fooCe(ptr noalias sret({ x86_fp80, x86_fp80 }) align 16 %agg.result, ptr noundef byval({ x86_fp80, x86_fp80 }) align 16 %f)
-// CHECK-LIN32-DAG: define{{.*}} x86_regcallcc void @_Z15__regcall4__fooCe(ptr inreg noalias sret({ x86_fp80, x86_fp80 }) align 4 %agg.result, ptr noundef byval({ x86_fp80, x86_fp80 }) align 4 %f)
+// CHECK-LIN64-DAG: define{{.*}} x86_regcallcc void @_Z15__regcall4__fooCe(ptr dead_on_unwind noalias writable sret({ x86_fp80, x86_fp80 }) align 16 %agg.result, ptr noundef byval({ x86_fp80, x86_fp80 }) align 16 %f)
+// CHECK-LIN32-DAG: define{{.*}} x86_regcallcc void @_Z15__regcall4__fooCe(ptr dead_on_unwind inreg noalias writable sret({ x86_fp80, x86_fp80 }) align 4 %agg.result, ptr noundef byval({ x86_fp80, x86_fp80 }) align 4 %f)
 // CHECK-WIN64-DAG: define dso_local x86_regcallcc noundef { double, double } @"?foo@@YxU?$_Complex@O@__clang@@U12@@Z"(double noundef %f.0, double noundef %f.1)
 // CHECK-WIN32-DAG: define dso_local x86_regcallcc noundef { double, double } @"?foo@@YxU?$_Complex@O@__clang@@U12@@Z"(double noundef %f.0, double noundef %f.1)
 
diff --git a/clang/test/CodeGenCXX/stack-reuse-miscompile.cpp b/clang/test/CodeGenCXX/stack-reuse-miscompile.cpp
index dbeea5e32cfb8..50a8d167f5b77 100644
--- a/clang/test/CodeGenCXX/stack-reuse-miscompile.cpp
+++ b/clang/test/CodeGenCXX/stack-reuse-miscompile.cpp
@@ -36,7 +36,7 @@ const char * f(S s)
 // CHECK: call void @llvm.lifetime.start.p0(i64 16, ptr [[T3]])
 // CHECK: [[T5:%.*]] = call noundef ptr @_ZN1TC1E1S(ptr {{[^,]*}} [[T3]], [2 x i32] %{{.*}})
 //
-// CHECK: call void @_ZNK1T6concatERKS_(ptr sret(%class.T) align 4 [[T1]], ptr {{[^,]*}} [[T2]], ptr noundef nonnull align 4 dereferenceable(16) [[T3]])
+// CHECK: call void @_ZNK1T6concatERKS_(ptr dead_on_unwind writable sret(%class.T) align 4 [[T1]], ptr {{[^,]*}} [[T2]], ptr noundef nonnull align 4 dereferenceable(16) [[T3]])
 // CHECK: [[T6:%.*]] = call noundef ptr @_ZNK1T3strEv(ptr {{[^,]*}} [[T1]])
 //
 // CHECK: call void @llvm.lifetime.end.p0(
diff --git a/clang/test/CodeGenCXX/stack-reuse.cpp b/clang/test/CodeGenCXX/stack-reuse.cpp
index e2412a55c1f62..ca73781b79ec1 100644
--- a/clang/test/CodeGenCXX/stack-reuse.cpp
+++ b/clang/test/CodeGenCXX/stack-reuse.cpp
@@ -135,7 +135,7 @@ int large_combiner_test(S_large s) {
 // CHECK: [[T2:%.*]] = alloca %struct.Combiner
 // CHECK: [[T1:%.*]] = alloca %struct.Combiner
 // CHECK: [[T3:%.*]] = call noundef ptr @_ZN8CombinerC1E7S_large(ptr {{[^,]*}} [[T1]], [9 x i32] %s.coerce)
-// CHECK: call void @_ZN8Combiner1fEv(ptr nonnull sret(%struct.Combiner) align 4 [[T2]], ptr {{[^,]*}} [[T1]])
+// CHECK: call void @_ZN8Combiner1fEv(ptr dead_on_unwind nonnull writable sret(%struct.Combiner) align 4 [[T2]], ptr {{[^,]*}} [[T1]])
 // CHECK: [[T5:%.*]] = load i32, ptr [[T2]]
 // CHECK: ret i32 [[T5]]
 
diff --git a/clang/test/CodeGenCXX/temporaries.cpp b/clang/test/CodeGenCXX/temporaries.cpp
index c5adb42a6f173..9f29d3b732783 100644
--- a/clang/test/CodeGenCXX/temporaries.cpp
+++ b/clang/test/CodeGenCXX/temporaries.cpp
@@ -414,13 +414,13 @@ namespace Elision {
     // CHECK-NEXT: call void @_ZN7Elision1AC1Ev(ptr {{[^,]*}} [[I]])
     A i = (foo(), A());
 
-    // CHECK-NEXT: call void @_ZN7Elision4fooAEv(ptr sret([[A]]) align 8 [[T0]])
+    // CHECK-NEXT: call void @_ZN7Elision4fooAEv(ptr dead_on_unwind writable sret([[A]]) align 8 [[T0]])
     // CHECK-NEXT: call void @_ZN7Elision1AC1Ev(ptr {{[^,]*}} [[J]])
     // CHECK-NEXT: call void @_ZN7Elision1AD1Ev(ptr {{[^,]*}} [[T0]])
     A j = (fooA(), A());
 
     // CHECK-NEXT: call void @_ZN7Elision1AC1Ev(ptr {{[^,]*}} [[T1]])
-    // CHECK-NEXT: call void @_ZN7Elision4fooAEv(ptr sret([[A]]) align 8 [[K]])
+    // CHECK-NEXT: call void @_ZN7Elision4fooAEv(ptr dead_on_unwind writable sret([[A]]) align 8 [[K]])
     // CHECK-NEXT: call void @_ZN7Elision1AD1Ev(ptr {{[^,]*}} [[T1]])
     A k = (A(), fooA());
 
@@ -447,7 +447,7 @@ namespace Elision {
     // CHECK-NEXT: call void @_ZN7Elision1AD1Ev(ptr {{[^,]*}} [[I]])
   }
 
-  // CHECK: define{{.*}} void @_ZN7Elision5test2Ev(ptr noalias sret([[A]]) align 8
+  // CHECK: define{{.*}} void @_ZN7Elision5test2Ev(ptr dead_on_unwind noalias writable sret([[A]]) align 8
   A test2() {
     // CHECK:      call void @_ZN7Elision3fooEv()
     // CHECK-NEXT: call void @_ZN7Elision1AC1Ev(ptr {{[^,]*}} [[RET:%.*]])
@@ -455,7 +455,7 @@ namespace Elision {
     return (foo(), A());
   }
 
-  // CHECK: define{{.*}} void @_ZN7Elision5test3EiNS_1AE(ptr noalias sret([[A]]) align 8
+  // CHECK: define{{.*}} void @_ZN7Elision5test3EiNS_1AE(ptr dead_on_unwind noalias writable sret([[A]]) align 8
   A test3(int v, A x) {
     if (v < 5)
     // CHECK:      call void @_ZN7Elision1AC1Ev(ptr {{[^,]*}} [[RET:%.*]])
@@ -495,7 +495,7 @@ namespace Elision {
     // CHECK:      call void @_ZN7Elision1AD1Ev(ptr {{[^,]*}} [[X]])
   }
 
-  // CHECK: define{{.*}} void @_ZN7Elision5test5Ev(ptr noalias sret([[A]]) align 8
+  // CHECK: define{{.*}} void @_ZN7Elision5test5Ev(ptr dead_on_unwind noalias writable sret([[A]]) align 8
   struct B { A a; B(); };
   A test5() {
     // CHECK:      [[AT0:%.*]] = alloca [[A]], align 8
@@ -533,7 +533,7 @@ namespace Elision {
   void test6(const C *x) {
     // CHECK:      [[T0:%.*]] = alloca [[A]], align 8
     // CHECK:      [[X:%.*]] = load ptr, ptr {{%.*}}, align 8
-    // CHECK-NEXT: call void @_ZNK7Elision1CcvNS_1AEEv(ptr sret([[A]]) align 8 [[T0]], ptr {{[^,]*}} [[X]])
+    // CHECK-NEXT: call void @_ZNK7Elision1CcvNS_1AEEv(ptr dead_on_unwind writable sret([[A]]) align 8 [[T0]], ptr {{[^,]*}} [[X]])
     // CHECK-NEXT: call void @_ZNK7Elision1A3fooEv(ptr {{[^,]*}} [[T0]])
     // CHECK-NEXT: call void @_ZN7Elision1AD1Ev(ptr {{[^,]*}} [[T0]])
     // CHECK-NEXT: ret void
diff --git a/clang/test/CodeGenCXX/thiscall-struct-return.cpp b/clang/test/CodeGenCXX/thiscall-struct-return.cpp
index c29ec56b41c1a..7802ed4b18ebd 100644
--- a/clang/test/CodeGenCXX/thiscall-struct-return.cpp
+++ b/clang/test/CodeGenCXX/thiscall-struct-return.cpp
@@ -34,8 +34,8 @@ void test( void ) {
 // CHECK: call void @_ZN1CC1Ev(ptr {{[^,]*}} [[C:%.+]])
   C c;
 
-// CHECK: call x86_thiscallcc void @_ZNK1C5SmallEv(ptr sret(%struct.S) align 4 %{{.+}}, ptr {{[^,]*}} [[C]])
+// CHECK: call x86_thiscallcc void @_ZNK1C5SmallEv(ptr dead_on_unwind writable sret(%struct.S) align 4 %{{.+}}, ptr {{[^,]*}} [[C]])
   (void)c.Small();
-// CHECK: call x86_thiscallcc void @_ZNK1C6MediumEv(ptr sret(%struct.M) align 4 %{{.+}}, ptr {{[^,]*}} [[C]])
+// CHECK: call x86_thiscallcc void @_ZNK1C6MediumEv(ptr dead_on_unwind writable sret(%struct.M) align 4 %{{.+}}, ptr {{[^,]*}} [[C]])
   (void)c.Medium();
 }
diff --git a/clang/test/CodeGenCXX/thunk-returning-memptr.cpp b/clang/test/CodeGenCXX/thunk-returning-memptr.cpp
index aaa5c8605f23e..99ef76f91cecf 100644
--- a/clang/test/CodeGenCXX/thunk-returning-memptr.cpp
+++ b/clang/test/CodeGenCXX/thunk-returning-memptr.cpp
@@ -23,5 +23,5 @@ C::C() {}
 // Because of the tail call, the return value cannot be copied into a local
 // alloca. (PR39901)
 
-// CHECK-LABEL: define linkonce_odr void @_ZThn4_N1C1fEv(ptr noalias sret({ i32, i32 }) align 4 %agg.result, ptr noundef %this)
-// CHECK: tail call void @_ZN1C1fEv(ptr sret({ i32, i32 }) align 4 %agg.result
+// CHECK-LABEL: define linkonce_odr void @_ZThn4_N1C1fEv(ptr dead_on_unwind noalias writable sret({ i32, i32 }) align 4 %agg.result, ptr noundef %this)
+// CHECK: tail call void @_ZN1C1fEv(ptr dead_on_unwind writable sret({ i32, i32 }) align 4 %agg.result
diff --git a/clang/test/CodeGenCXX/trivial_abi.cpp b/clang/test/CodeGenCXX/trivial_abi.cpp
index 3249df129157f..3012b0f2bc33d 100644
--- a/clang/test/CodeGenCXX/trivial_abi.cpp
+++ b/clang/test/CodeGenCXX/trivial_abi.cpp
@@ -151,7 +151,7 @@ void testIgnoredSmall() {
 void testParamLarge(Large a) noexcept {
 }
 
-// CHECK: define{{.*}} void @_Z15testReturnLargev(ptr noalias sret(%[[STRUCT_LARGE]]) align 8 %[[AGG_RESULT:.*]])
+// CHECK: define{{.*}} void @_Z15testReturnLargev(ptr dead_on_unwind noalias writable sret(%[[STRUCT_LARGE]]) align 8 %[[AGG_RESULT:.*]])
 // CHECK: %[[CALL:.*]] = call noundef ptr @_ZN5LargeC1Ev(ptr {{[^,]*}} %[[AGG_RESULT]])
 // CHECK: ret void
 // CHECK: }
@@ -178,7 +178,7 @@ void testCallLarge0() {
 
 // CHECK: define{{.*}} void @_Z14testCallLarge1v()
 // CHECK: %[[AGG_TMP:.*]] = alloca %[[STRUCT_LARGE:.*]], align 8
-// CHECK: call void @_Z15testReturnLargev(ptr sret(%[[STRUCT_LARGE]]) align 8 %[[AGG_TMP]])
+// CHECK: call void @_Z15testReturnLargev(ptr dead_on_unwind writable sret(%[[STRUCT_LARGE]]) align 8 %[[AGG_TMP]])
 // CHECK: call void @_Z14testParamLarge5Large(ptr noundef %[[AGG_TMP]])
 // CHECK: ret void
 // CHECK: }
@@ -189,7 +189,7 @@ void testCallLarge1() {
 
 // CHECK: define{{.*}} void @_Z16testIgnoredLargev()
 // CHECK: %[[AGG_TMP_ENSURED:.*]] = alloca %[[STRUCT_LARGE:.*]], align 8
-// CHECK: call void @_Z15testReturnLargev(ptr sret(%[[STRUCT_LARGE]]) align 8 %[[AGG_TMP_ENSURED]])
+// CHECK: call void @_Z15testReturnLargev(ptr dead_on_unwind writable sret(%[[STRUCT_LARGE]]) align 8 %[[AGG_TMP_ENSURED]])
 // CHECK: %[[CALL:.*]] = call noundef ptr @_ZN5LargeD1Ev(ptr {{[^,]*}} %[[AGG_TMP_ENSURED]])
 // CHECK: ret void
 // CHECK: }
@@ -210,7 +210,7 @@ Trivial testReturnHasTrivial() {
   return t;
 }
 
-// CHECK: define{{.*}} void @_Z23testReturnHasNonTrivialv(ptr noalias sret(%[[STRUCT_NONTRIVIAL:.*]]) align 4 %[[AGG_RESULT:.*]])
+// CHECK: define{{.*}} void @_Z23testReturnHasNonTrivialv(ptr dead_on_unwind noalias writable sret(%[[STRUCT_NONTRIVIAL:.*]]) align 4 %[[AGG_RESULT:.*]])
 // CHECK: %[[CALL:.*]] = call noundef ptr @_ZN10NonTrivialC1Ev(ptr {{[^,]*}} %[[AGG_RESULT]])
 // CHECK: ret void
 // CHECK: }
diff --git a/clang/test/CodeGenCXX/unknown-anytype.cpp b/clang/test/CodeGenCXX/unknown-anytype.cpp
index 862b8fe8b66dd..ade9a6482d75c 100644
--- a/clang/test/CodeGenCXX/unknown-anytype.cpp
+++ b/clang/test/CodeGenCXX/unknown-anytype.cpp
@@ -70,7 +70,7 @@ struct Test7 {
 };
 extern "C" __unknown_anytype test7_any(int);
 Test7 test7() {
-  // COMMON: call void @test7_any(ptr sret({{%.*}}) align 1 {{%.*}}, i32 noundef 5)
+  // COMMON: call void @test7_any(ptr dead_on_unwind writable sret({{%.*}}) align 1 {{%.*}}, i32 noundef 5)
   return (Test7) test7_any(5);
 }
 
diff --git a/clang/test/CodeGenCXX/wasm-args-returns.cpp b/clang/test/CodeGenCXX/wasm-args-returns.cpp
index 090ca9a99756f..e80dfefedece1 100644
--- a/clang/test/CodeGenCXX/wasm-args-returns.cpp
+++ b/clang/test/CodeGenCXX/wasm-args-returns.cpp
@@ -30,52 +30,52 @@ struct two_fields {
   double d, e;
 };
 test(two_fields);
-// CHECK: define void @_Z7forward10two_fields(ptr noalias nocapture writeonly sret(%struct.two_fields) align 8 %{{.*}}, ptr nocapture readonly byval(%struct.two_fields) align 8 %{{.*}})
+// CHECK: define void @_Z7forward10two_fields(ptr dead_on_unwind noalias nocapture writable writeonly sret(%struct.two_fields) align 8 %{{.*}}, ptr nocapture readonly byval(%struct.two_fields) align 8 %{{.*}})
 //
 // CHECK: define void @_Z15test_two_fieldsv()
 // CHECK: %[[tmp:.*]] = alloca %struct.two_fields, align 8
-// CHECK: call void @_Z14def_two_fieldsv(ptr nonnull sret(%struct.two_fields) align 8 %[[tmp]])
+// CHECK: call void @_Z14def_two_fieldsv(ptr dead_on_unwind nonnull writable sret(%struct.two_fields) align 8 %[[tmp]])
 // CHECK: call void @_Z3use10two_fields(ptr nonnull byval(%struct.two_fields) align 8 %[[tmp]])
 // CHECK: ret void
 //
 // CHECK: declare void @_Z3use10two_fields(ptr byval(%struct.two_fields) align 8)
-// CHECK: declare void @_Z14def_two_fieldsv(ptr sret(%struct.two_fields) align 8)
+// CHECK: declare void @_Z14def_two_fieldsv(ptr dead_on_unwind writable sret(%struct.two_fields) align 8)
 
 struct copy_ctor {
   double d;
   copy_ctor(copy_ctor const &);
 };
 test(copy_ctor);
-// CHECK: define void @_Z7forward9copy_ctor(ptr noalias {{[^,]*}} sret(%struct.copy_ctor) align 8 %{{.*}}, ptr nonnull %{{.*}})
+// CHECK: define void @_Z7forward9copy_ctor(ptr dead_on_unwind noalias {{[^,]*}} sret(%struct.copy_ctor) align 8 %{{.*}}, ptr nonnull %{{.*}})
 //
 // CHECK: declare ptr @_ZN9copy_ctorC1ERKS_(ptr {{[^,]*}} returned {{[^,]*}}, ptr nonnull align 8 dereferenceable(8))
 //
 // CHECK: define void @_Z14test_copy_ctorv()
 // CHECK: %[[tmp:.*]] = alloca %struct.copy_ctor, align 8
-// CHECK: call void @_Z13def_copy_ctorv(ptr nonnull sret(%struct.copy_ctor) align 8 %[[tmp]])
+// CHECK: call void @_Z13def_copy_ctorv(ptr dead_on_unwind nonnull writable sret(%struct.copy_ctor) align 8 %[[tmp]])
 // CHECK: call void @_Z3use9copy_ctor(ptr nonnull %[[tmp]])
 // CHECK: ret void
 //
 // CHECK: declare void @_Z3use9copy_ctor(ptr)
-// CHECK: declare void @_Z13def_copy_ctorv(ptr sret(%struct.copy_ctor) align 8)
+// CHECK: declare void @_Z13def_copy_ctorv(ptr dead_on_unwind writable sret(%struct.copy_ctor) align 8)
 
 struct __attribute__((aligned(16))) aligned_copy_ctor {
   double d, e;
   aligned_copy_ctor(aligned_copy_ctor const &);
 };
 test(aligned_copy_ctor);
-// CHECK: define void @_Z7forward17aligned_copy_ctor(ptr noalias {{[^,]*}} sret(%struct.aligned_copy_ctor) align 16 %{{.*}}, ptr nonnull %{{.*}})
+// CHECK: define void @_Z7forward17aligned_copy_ctor(ptr dead_on_unwind noalias {{[^,]*}} sret(%struct.aligned_copy_ctor) align 16 %{{.*}}, ptr nonnull %{{.*}})
 //
 // CHECK: declare ptr @_ZN17aligned_copy_ctorC1ERKS_(ptr {{[^,]*}} returned {{[^,]*}}, ptr nonnull align 16 dereferenceable(16))
 //
 // CHECK: define void @_Z22test_aligned_copy_ctorv()
 // CHECK: %[[tmp:.*]] = alloca %struct.aligned_copy_ctor, align 16
-// CHECK: call void @_Z21def_aligned_copy_ctorv(ptr nonnull sret(%struct.aligned_copy_ctor) align 16 %[[tmp]])
+// CHECK: call void @_Z21def_aligned_copy_ctorv(ptr dead_on_unwind nonnull writable sret(%struct.aligned_copy_ctor) align 16 %[[tmp]])
 // CHECK: call void @_Z3use17aligned_copy_ctor(ptr nonnull %[[tmp]])
 // CHECK: ret void
 //
 // CHECK: declare void @_Z3use17aligned_copy_ctor(ptr)
-// CHECK: declare void @_Z21def_aligned_copy_ctorv(ptr sret(%struct.aligned_copy_ctor) align 16)
+// CHECK: declare void @_Z21def_aligned_copy_ctorv(ptr dead_on_unwind writable sret(%struct.aligned_copy_ctor) align 16)
 
 struct empty {};
 test(empty);
diff --git a/clang/test/CodeGenCXX/x86_32-arguments.cpp b/clang/test/CodeGenCXX/x86_32-arguments.cpp
index 7388b2fcfe6cd..c071548327e8d 100644
--- a/clang/test/CodeGenCXX/x86_32-arguments.cpp
+++ b/clang/test/CodeGenCXX/x86_32-arguments.cpp
@@ -6,7 +6,7 @@ struct S {
   short s;
 };
 
-// CHECK-LABEL: define{{.*}} void @_Z1fv(ptr noalias sret(%struct.S) align 2 %
+// CHECK-LABEL: define{{.*}} void @_Z1fv(ptr dead_on_unwind noalias writable sret(%struct.S) align 2 %
 S f() { return S(); }
 // CHECK-LABEL: define{{.*}} void @_Z1f1S(ptr noundef %0)
 void f(S) { }
@@ -18,7 +18,7 @@ class C {
   double c;
 };
 
-// CHECK-LABEL: define{{.*}} void @_Z1gv(ptr noalias sret(%class.C) align 4 %
+// CHECK-LABEL: define{{.*}} void @_Z1gv(ptr dead_on_unwind noalias writable sret(%class.C) align 4 %
 C g() { return C(); }
 
 // CHECK-LABEL: define{{.*}} void @_Z1f1C(ptr noundef %0)
@@ -103,13 +103,13 @@ struct s7_1 { double x; };
 struct s7 : s7_0, s7_1 { };
 s7 f7() { return s7(); }
 
-// CHECK-LABEL: define{{.*}} void @_Z2f8v(ptr noalias sret(%struct.s8) align 4 %agg.result)
+// CHECK-LABEL: define{{.*}} void @_Z2f8v(ptr dead_on_unwind noalias writable sret(%struct.s8) align 4 %agg.result)
 struct s8_0 { };
 struct s8_1 { double x; };
 struct s8 { s8_0 a; s8_1 b; };
 s8 f8() { return s8(); }
 
-// CHECK-LABEL: define{{.*}} void @_Z2f9v(ptr noalias sret(%struct.s9) align 4 %agg.result)
+// CHECK-LABEL: define{{.*}} void @_Z2f9v(ptr dead_on_unwind noalias writable sret(%struct.s9) align 4 %agg.result)
 struct s9_0 { unsigned : 0; };
 struct s9_1 { double x; };
 struct s9 { s9_0 a; s9_1 b; };
diff --git a/clang/test/CodeGenCXX/x86_64-arguments.cpp b/clang/test/CodeGenCXX/x86_64-arguments.cpp
index d1bbf5d30f59e..ebeba7fd65495 100644
--- a/clang/test/CodeGenCXX/x86_64-arguments.cpp
+++ b/clang/test/CodeGenCXX/x86_64-arguments.cpp
@@ -173,7 +173,7 @@ namespace test9 {
   // CHECK: define{{.*}} void @_ZN5test93fooEPNS_1SEPNS_1TE(ptr %0, ptr %1)
   void foo(S*, T*) {}
 
-  // CHECK: define{{.*}} void @_ZN5test91aEiiiiNS_1TEPv(ptr noalias sret([[S:%.*]]) align 8 {{%.*}}, i32 %0, i32 %1, i32 %2, i32 %3, ptr byval([[T:%.*]]) align 8 %4, ptr %5)
+  // CHECK: define{{.*}} void @_ZN5test91aEiiiiNS_1TEPv(ptr dead_on_unwind noalias writable sret([[S:%.*]]) align 8 {{%.*}}, i32 %0, i32 %1, i32 %2, i32 %3, ptr byval([[T:%.*]]) align 8 %4, ptr %5)
   S a(int, int, int, int, T, void*) {
     return S();
   }
@@ -183,7 +183,7 @@ namespace test9 {
     return sret;
   }
 
-  // CHECK: define{{.*}} void @_ZN5test91cEiiiNS_1TEPv(ptr noalias sret([[S]]) align 8 {{%.*}}, i32 %0, i32 %1, i32 %2, ptr {{%.*}}, ptr {{%.*}}, ptr %3)
+  // CHECK: define{{.*}} void @_ZN5test91cEiiiNS_1TEPv(ptr dead_on_unwind noalias writable sret([[S]]) align 8 {{%.*}}, i32 %0, i32 %1, i32 %2, ptr {{%.*}}, ptr {{%.*}}, ptr %3)
   S c(int, int, int, T, void*) {
     return S();
   }
diff --git a/clang/test/CodeGenCoroutines/coro-await.cpp b/clang/test/CodeGenCoroutines/coro-await.cpp
index e1c7039469afc..dc5a765ccb83d 100644
--- a/clang/test/CodeGenCoroutines/coro-await.cpp
+++ b/clang/test/CodeGenCoroutines/coro-await.cpp
@@ -127,7 +127,7 @@ extern "C" void f1(int) {
   // CHECK: %[[PROMISE:.+]] = alloca %"struct.std::coroutine_traits::promise_type"
   // CHECK: %[[FRAME:.+]] = call ptr @llvm.coro.begin(
   co_yield 42;
-  // CHECK: call void @_ZNSt16coroutine_traitsIJviEE12promise_type11yield_valueEi(ptr sret(%struct.suspend_maybe) align 4 %[[AWAITER:.+]], ptr {{[^,]*}} %[[PROMISE]], i32 42)
+  // CHECK: call void @_ZNSt16coroutine_traitsIJviEE12promise_type11yield_valueEi(ptr dead_on_unwind writable sret(%struct.suspend_maybe) align 4 %[[AWAITER:.+]], ptr {{[^,]*}} %[[PROMISE]], i32 42)
 
   // See if we need to suspend:
   // --------------------------
@@ -194,20 +194,20 @@ extern "C" void UseAggr(Aggr&&);
 extern "C" void TestAggr() {
   UseAggr(co_await AggrAwaiter{});
   Whatever();
-  // CHECK: call void @_ZN11AggrAwaiter12await_resumeEv(ptr sret(%struct.Aggr) align 4 %[[AwaitResume:.+]],
+  // CHECK: call void @_ZN11AggrAwaiter12await_resumeEv(ptr dead_on_unwind writable sret(%struct.Aggr) align 4 %[[AwaitResume:.+]],
   // CHECK: call void @UseAggr(ptr nonnull align 4 dereferenceable(12) %[[AwaitResume]])
   // CHECK: call void @_ZN4AggrD1Ev(ptr {{[^,]*}} %[[AwaitResume]])
   // CHECK: call void @Whatever()
 
   co_await AggrAwaiter{};
   Whatever();
-  // CHECK: call void @_ZN11AggrAwaiter12await_resumeEv(ptr sret(%struct.Aggr) align 4 %[[AwaitResume2:.+]],
+  // CHECK: call void @_ZN11AggrAwaiter12await_resumeEv(ptr dead_on_unwind writable sret(%struct.Aggr) align 4 %[[AwaitResume2:.+]],
   // CHECK: call void @_ZN4AggrD1Ev(ptr {{[^,]*}} %[[AwaitResume2]])
   // CHECK: call void @Whatever()
 
   Aggr Val = co_await AggrAwaiter{};
   Whatever();
-  // CHECK: call void @_ZN11AggrAwaiter12await_resumeEv(ptr sret(%struct.Aggr) align 4 %[[AwaitResume3:.+]],
+  // CHECK: call void @_ZN11AggrAwaiter12await_resumeEv(ptr dead_on_unwind writable sret(%struct.Aggr) align 4 %[[AwaitResume3:.+]],
   // CHECK: call void @Whatever()
   // CHECK: call void @_ZN4AggrD1Ev(ptr {{[^,]*}} %[[AwaitResume3]])
 }
@@ -252,7 +252,7 @@ extern "C" void TestOpAwait() {
 
   co_await MyAgg{};
   // CHECK: call void @_ZN5MyAggawEv(ptr {{[^,]*}} %
-  // CHECK: call void @_ZN11AggrAwaiter12await_resumeEv(ptr sret(%struct.Aggr) align 4 %
+  // CHECK: call void @_ZN11AggrAwaiter12await_resumeEv(ptr dead_on_unwind writable sret(%struct.Aggr) align 4 %
 }
 
 // CHECK-LABEL: EndlessLoop(
diff --git a/clang/test/CodeGenCoroutines/coro-gro2.cpp b/clang/test/CodeGenCoroutines/coro-gro2.cpp
index b0faa66edd333..56299574fbf25 100644
--- a/clang/test/CodeGenCoroutines/coro-gro2.cpp
+++ b/clang/test/CodeGenCoroutines/coro-gro2.cpp
@@ -34,13 +34,13 @@ struct coro {
 };
 
 // Verify that the RVO is applied.
-// CHECK-LABEL: define{{.*}} void @_Z1fi(ptr noalias sret(%struct.coro) align 8 %agg.result, i32 noundef %0)
+// CHECK-LABEL: define{{.*}} void @_Z1fi(ptr dead_on_unwind noalias writable sret(%struct.coro) align 8 %agg.result, i32 noundef %0)
 coro f(int) {
 // CHECK: %call = call noalias noundef nonnull ptr @_Znwm(
 // CHECK-NEXT: br label %[[CoroInit:.*]]
 
 // CHECK: {{.*}}[[CoroInit]]:
-// CHECK: call void @{{.*get_return_objectEv}}(ptr sret(%struct.coro) align 8 %agg.result
+// CHECK: call void @{{.*get_return_objectEv}}(ptr dead_on_unwind writable sret(%struct.coro) align 8 %agg.result
 co_return;
 }
 
@@ -63,7 +63,7 @@ struct coro_two {
 };
 
 // Verify that the RVO is applied.
-// CHECK-LABEL: define{{.*}} void @_Z1hi(ptr noalias sret(%struct.coro_two) align 8 %agg.result, i32 noundef %0)
+// CHECK-LABEL: define{{.*}} void @_Z1hi(ptr dead_on_unwind noalias writable sret(%struct.coro_two) align 8 %agg.result, i32 noundef %0)
 coro_two h(int) {
 
   // CHECK: %call = call noalias noundef ptr @_ZnwmRKSt9nothrow_t
@@ -71,11 +71,11 @@ coro_two h(int) {
   // CHECK-NEXT: br i1 %[[CheckNull]], label %[[InitOnSuccess:.*]], label %[[InitOnFailure:.*]]
 
   // CHECK: {{.*}}[[InitOnFailure]]:
-  // CHECK-NEXT: call void @{{.*get_return_object_on_allocation_failureEv}}(ptr sret(%struct.coro_two) align 8 %agg.result
+  // CHECK-NEXT: call void @{{.*get_return_object_on_allocation_failureEv}}(ptr dead_on_unwind writable sret(%struct.coro_two) align 8 %agg.result
   // CHECK-NEXT: br label %[[RetLabel:.*]]
 
   // CHECK: {{.*}}[[InitOnSuccess]]:
-  // CHECK: call void @{{.*get_return_objectEv}}(ptr sret(%struct.coro_two) align 8 %agg.result
+  // CHECK: call void @{{.*get_return_objectEv}}(ptr dead_on_unwind writable sret(%struct.coro_two) align 8 %agg.result
 
   // CHECK: [[RetLabel]]:
   // CHECK-NEXT: ret void
diff --git a/clang/test/CodeGenHLSL/sret_output.hlsl b/clang/test/CodeGenHLSL/sret_output.hlsl
index 6b15e017a024d..33f88c639525f 100644
--- a/clang/test/CodeGenHLSL/sret_output.hlsl
+++ b/clang/test/CodeGenHLSL/sret_output.hlsl
@@ -10,7 +10,7 @@ struct S {
 
 
 // Make sure sret parameter is generated.
-// CHECK:define internal void @"?ps_main@@YA?AUS@@XZ"(ptr noalias sret(%struct.S) align 4 %agg.result)
+// CHECK:define internal void @"?ps_main@@YA?AUS@@XZ"(ptr dead_on_unwind noalias writable sret(%struct.S) align 4 %agg.result)
 // FIXME: change it to real value instead of poison value once semantic is add to a.
 // Make sure the function with sret is called.
 // CHECK:call void @"?ps_main@@YA?AUS@@XZ"(ptr poison)
diff --git a/clang/test/CodeGenObjC/arc.m b/clang/test/CodeGenObjC/arc.m
index 74356d8895d14..aeead58f8131d 100644
--- a/clang/test/CodeGenObjC/arc.m
+++ b/clang/test/CodeGenObjC/arc.m
@@ -1358,11 +1358,11 @@ void test70(id i) {
 // CHECK-LABEL: define{{.*}} void @test71
 void test71(void) {
   // CHECK: call void @llvm.lifetime.start.p0({{[^,]+}}, ptr %[[T:.*]])
-  // CHECK: call void @getAggDtor(ptr sret(%struct.AggDtor) align 8 %[[T]])
+  // CHECK: call void @getAggDtor(ptr dead_on_unwind writable sret(%struct.AggDtor) align 8 %[[T]])
   // CHECK: call void @__destructor_8_s40(ptr %[[T]])
   // CHECK: call void @llvm.lifetime.end.p0({{[^,]+}}, ptr %[[T]])
   // CHECK: call void @llvm.lifetime.start.p0({{[^,]+}}, ptr %[[T2:.*]])
-  // CHECK: call void @getAggDtor(ptr sret(%struct.AggDtor) align 8 %[[T2]])
+  // CHECK: call void @getAggDtor(ptr dead_on_unwind writable sret(%struct.AggDtor) align 8 %[[T2]])
   // CHECK: call void @__destructor_8_s40(ptr %[[T2]])
   // CHECK: call void @llvm.lifetime.end.p0({{[^,]+}}, ptr %[[T2]])
   getAggDtor();
diff --git a/clang/test/CodeGenObjC/direct-method.m b/clang/test/CodeGenObjC/direct-method.m
index 8a3c2f575d2d3..028a0888d594e 100644
--- a/clang/test/CodeGenObjC/direct-method.m
+++ b/clang/test/CodeGenObjC/direct-method.m
@@ -111,7 +111,7 @@ + (struct my_complex_struct)classGetComplex __attribute__((objc_direct)) {
 
 // CHECK-LABEL: define hidden void @"\01-[Root getAggregate]"(
 - (struct my_aggregate_struct)getAggregate __attribute__((objc_direct)) {
-  // CHECK: ptr noalias sret(%struct.my_aggregate_struct) align 4 [[RETVAL:%[^,]*]],
+  // CHECK: ptr dead_on_unwind noalias writable sret(%struct.my_aggregate_struct) align 4 [[RETVAL:%[^,]*]],
 
   // loading parameters
   // CHECK-LABEL: entry:
diff --git a/clang/test/CodeGenObjC/dllstorage.m b/clang/test/CodeGenObjC/dllstorage.m
index 0dbf1881caa9c..f45eb7bb6aee7 100644
--- a/clang/test/CodeGenObjC/dllstorage.m
+++ b/clang/test/CodeGenObjC/dllstorage.m
@@ -35,7 +35,7 @@ @implementation J {
 
 // CHECK-IR-DAG: @"OBJC_IVAR_$_J._ivar" = global i32
 
-// CHECK-NF-DAG: @"__objc_ivar_offset_J._ivar.\01" = hidden global i32
+// CHECK-NF-DAG: @"__objc_ivar_offset_J._ivar.@" = hidden global i32
 
 @interface K : J
 @end
@@ -56,7 +56,7 @@ @implementation K {
 
 // CHECK-IR-DAG: @"OBJC_IVAR_$_K._ivar" = global i32
 
-// CHECK-NF-DAG: @"__objc_ivar_offset_K._ivar.\01" = hidden global i32
+// CHECK-NF-DAG: @"__objc_ivar_offset_K._ivar.@" = hidden global i32
 
 __declspec(dllexport)
 @interface L : K
@@ -94,11 +94,11 @@ @implementation L {
 // CHECK-IR-DAG: @"OBJC_IVAR_$_L._package" = global i32
 // CHECK-IR-DAG: @"OBJC_IVAR_$_L._private" = global i32
 
-// CHECK-NF-DAG: @"__objc_ivar_offset_L._none.\01" = hidden global i32
-// CHECK-NF-DAG: @"__objc_ivar_offset_L._public.\01" = dso_local dllexport global i32
-// CHECK-NF-DAG: @"__objc_ivar_offset_L._protected.\01" = dso_local dllexport global i32
-// CHECK-NF-DAG: @"__objc_ivar_offset_L._package.\01" = hidden global i32
-// CHECK-NF-DAG: @"__objc_ivar_offset_L._private.\01" = hidden global i32
+// CHECK-NF-DAG: @"__objc_ivar_offset_L._none.@" = hidden global i32
+// CHECK-NF-DAG: @"__objc_ivar_offset_L._public.@" = dso_local dllexport global i32
+// CHECK-NF-DAG: @"__objc_ivar_offset_L._protected.@" = dso_local dllexport global i32
+// CHECK-NF-DAG: @"__objc_ivar_offset_L._package.@" = hidden global i32
+// CHECK-NF-DAG: @"__objc_ivar_offset_L._private.@" = hidden global i32
 
 __declspec(dllimport)
 @interface M : I {
@@ -112,7 +112,7 @@ @interface M : I {
 // CHECK-IR-DAG: @"OBJC_IVAR_$_M._ivar" = external dllimport global i32
 
 // CHECK-NF-DAG: @"$_OBJC_REF_CLASS_M" = external dllimport global ptr
-// CHECK-NF-DAG: @"__objc_ivar_offset_M._ivar.\01" = external global i32
+// CHECK-NF-DAG: @"__objc_ivar_offset_M._ivar.@" = external global i32
 
 __declspec(dllexport)
 __attribute__((__objc_exception__))
@@ -151,7 +151,7 @@ id f(Q *q) {
 
 // CHECK-IR-DAG: @"OBJC_IVAR_$_M._ivar" = external dllimport global i32
 
-// CHECK-NF-DAG: @"__objc_ivar_offset_M._ivar.\01" = external global i32
+// CHECK-NF-DAG: @"__objc_ivar_offset_M._ivar.@" = external global i32
 
 int g(void) {
   @autoreleasepool {
diff --git a/clang/test/CodeGenObjC/encode-test-6.m b/clang/test/CodeGenObjC/encode-test-6.m
index 261eb7fb3368b..c32f8f24c0009 100644
--- a/clang/test/CodeGenObjC/encode-test-6.m
+++ b/clang/test/CodeGenObjC/encode-test-6.m
@@ -1,5 +1,10 @@
-// RUN: %clang_cc1 -triple x86_64-apple-darwin -emit-llvm -o %t %s
-// RUN: FileCheck < %t %s
+// RUN: %clang_cc1 -triple x86_64-apple-darwin -emit-llvm -o - %s | FileCheck -check-prefix CHECK-DWARF %s
+
+// RUN: %clang_cc1 -triple x86_64-w64-windows-gnu -emit-llvm -fobjc-runtime=gnustep-2.0 -o - %s | FileCheck -check-prefix CHECK-MINGW %s
+
+// RUN: %clang_cc1 -triple x86_64-windows-msvc -emit-llvm -fobjc-runtime=gnustep-2.0 -o - %s | FileCheck -check-prefix CHECK-MSVC %s
+
+// RUN: %clang_cc1 -triple x86_64-linux-gnu -emit-llvm -fobjc-runtime=gnustep-2.0 -o - %s | FileCheck -check-prefix CHECK-ELF %s
 
 typedef struct {} Z;
 
@@ -13,8 +18,17 @@ -(void)bar:(Z)a {}
 -(void)foo:(Z)a: (char*)b : (Z)c : (double) d {}
 @end
 
-// CHECK: private unnamed_addr constant [14 x i8] c"v16@0:8{?=}16
-// CHECK: private unnamed_addr constant [26 x i8] c"v32@0:8{?=}16*16{?=}24d24
+// CHECK-DWARF: private unnamed_addr constant [14 x i8] c"v16@0:8{?=}16
+// CHECK-DWARF: private unnamed_addr constant [26 x i8] c"v32@0:8{?=}16*16{?=}24d24
+
+// CHECK-MINGW: @".objc_sel_types_v16@0:8{?\02}16" = linkonce_odr hidden constant [14 x i8] c"v16@0:8{?=}16\00"
+// CHECK-MINGW: @".objc_sel_types_v32@0:8{?\02}16*16{?\02}24d24" = linkonce_odr hidden constant [26 x i8] c"v32@0:8{?=}16*16{?=}24d24\00"
+
+// CHECK-MSVC: @".objc_sel_types_v20@0:8{?\02}16" = linkonce_odr hidden constant [14 x i8] c"v20@0:8{?=}16\00"
+// CHECK-MSVC: @".objc_sel_types_v40@0:8{?\02}16*20{?\02}28d32" = linkonce_odr hidden constant [26 x i8] c"v40@0:8{?=}16*20{?=}28d32\00"
+
+// CHECK-ELF: @".objc_sel_types_v16\010:8{?=}16" = linkonce_odr hidden constant [14 x i8] c"v16@0:8{?=}16\00"
+// CHECK-ELF: @".objc_sel_types_v32\010:8{?=}16*16{?=}24d24" = linkonce_odr hidden constant [26 x i8] c"v32@0:8{?=}16*16{?=}24d24\00"
 
 @interface NSObject @end
 
@@ -31,7 +45,10 @@ @implementation BABugExample
 @synthesize property = _property;
 @end
 
-// CHECK: private unnamed_addr constant [8 x i8] c"@16
+// CHECK-DWARF: private unnamed_addr constant [8 x i8] c"@16
+// CHECK-MINGW: @".objc_sel_types_@16@0:8" = linkonce_odr hidden constant [8 x i8] c"@16@0:8\00"
+// CHECK-MSVC: @".objc_sel_types_@16@0:8" = linkonce_odr hidden constant [8 x i8] c"@16@0:8\00"
+// CHECK-ELF @".objc_sel_types_\0116\010:8" = linkonce_odr hidden constant [8 x i8] c"@16@0:8\00"
 
 @class SCNCamera;
 typedef SCNCamera C3DCamera;
@@ -48,7 +65,10 @@ @implementation SCNCamera
     C3DCameraStorage _storage;
 }
 @end
-// CHECK: private unnamed_addr constant [39 x i8] c"{?=\22presentationInstance\22@\22SCNCamera\22}\00"
+// CHECK-DWARF: private unnamed_addr constant [39 x i8] c"{?=\22presentationInstance\22@\22SCNCamera\22}\00"
+// CHECK-MINGW: @"__objc_ivar_offset_SCNCamera._storage.{?\02@}"
+// CHECK-MSVC: @"__objc_ivar_offset_SCNCamera._storage.{?\02@}"
+// CHECK-ELF: @"__objc_ivar_offset_SCNCamera._storage.{?=\01}"
 
 int i;
 typeof(@encode(typeof(i))) e = @encode(typeof(i));
@@ -56,6 +76,10 @@ @implementation SCNCamera
 {
     return e;
 }
-// CHECK: @e ={{.*}} global [2 x i8] c"i\00", align 1
-// CHECK: define{{.*}} ptr @Test()
-// CHECK: ret ptr @e
+// CHECK-DWARF: @e ={{.*}} global [2 x i8] c"i\00", align 1
+// CHECK-DWARF: define{{.*}} ptr @Test()
+// CHECK-DWARF: ret ptr @e
+
+// CHECK-MSVC: @e = dso_local global [2 x i8] c"i\00", align 1
+// CHECK-MINGW: @e = dso_local global [2 x i8] c"i\00", align 1
+// CHECK-ELF: @e = global [2 x i8] c"i\00", align 1
diff --git a/clang/test/CodeGenObjC/nontrivial-c-struct-exception.m b/clang/test/CodeGenObjC/nontrivial-c-struct-exception.m
index 0159e111d9ad0..d2a954ae26a04 100644
--- a/clang/test/CodeGenObjC/nontrivial-c-struct-exception.m
+++ b/clang/test/CodeGenObjC/nontrivial-c-struct-exception.m
@@ -39,8 +39,8 @@ void testStrongException(void) {
 // CHECK: define{{.*}} void @testWeakException()
 // CHECK: %[[AGG_TMP:.*]] = alloca %[[STRUCT_WEAK]], align 8
 // CHECK: %[[AGG_TMP1:.*]] = alloca %[[STRUCT_WEAK]], align 8
-// CHECK: call void @genWeak(ptr sret(%[[STRUCT_WEAK]]) align 8 %[[AGG_TMP]])
-// CHECK: invoke void @genWeak(ptr sret(%[[STRUCT_WEAK]]) align 8 %[[AGG_TMP1]])
+// CHECK: call void @genWeak(ptr dead_on_unwind writable sret(%[[STRUCT_WEAK]]) align 8 %[[AGG_TMP]])
+// CHECK: invoke void @genWeak(ptr dead_on_unwind writable sret(%[[STRUCT_WEAK]]) align 8 %[[AGG_TMP1]])
 
 // CHECK: call void @calleeWeak(ptr noundef %[[AGG_TMP]], ptr noundef %[[AGG_TMP1]])
 // CHECK: ret void
diff --git a/clang/test/CodeGenObjC/objc-non-trivial-struct-nrvo.m b/clang/test/CodeGenObjC/objc-non-trivial-struct-nrvo.m
index effba8c958441..cc697b0e15373 100644
--- a/clang/test/CodeGenObjC/objc-non-trivial-struct-nrvo.m
+++ b/clang/test/CodeGenObjC/objc-non-trivial-struct-nrvo.m
@@ -37,7 +37,7 @@ Trivial testTrivial(void) {
 
 void func1(TrivialBig *);
 
-// CHECK: define{{.*}} void @testTrivialBig(ptr noalias sret(%[[STRUCT_TRIVIALBIG]]) align 4 %[[AGG_RESULT:.*]])
+// CHECK: define{{.*}} void @testTrivialBig(ptr dead_on_unwind noalias writable sret(%[[STRUCT_TRIVIALBIG]]) align 4 %[[AGG_RESULT:.*]])
 // CHECK: call void @func1(ptr noundef %[[AGG_RESULT]])
 // CHECK-NEXT: ret void
 
@@ -67,7 +67,7 @@ Strong testStrong(void) {
   return a;
 }
 
-// CHECK: define{{.*}} void @testWeak(ptr noalias sret(%[[STRUCT_WEAK]]) align 8 %[[AGG_RESULT:.*]])
+// CHECK: define{{.*}} void @testWeak(ptr dead_on_unwind noalias writable sret(%[[STRUCT_WEAK]]) align 8 %[[AGG_RESULT:.*]])
 // CHECK: %[[NRVO:.*]] = alloca i1, align 1
 // CHECK: call void @__default_constructor_8_w0(ptr %[[AGG_RESULT]])
 // CHECK: store i1 true, ptr %[[NRVO]], align 1
@@ -101,7 +101,7 @@ Weak testWeak2(int c) {
     return b;
 }
 
-// CHECK: define internal void @"\01-[C1 foo1]"(ptr noalias sret(%[[STRUCT_WEAK]]) align 8 %[[AGG_RESULT:.*]], ptr noundef %{{.*}}, ptr noundef %{{.*}})
+// CHECK: define internal void @"\01-[C1 foo1]"(ptr dead_on_unwind noalias writable sret(%[[STRUCT_WEAK]]) align 8 %[[AGG_RESULT:.*]], ptr noundef %{{.*}}, ptr noundef %{{.*}})
 // CHECK: %[[NRVO:.*]] = alloca i1, align 1
 // CHECK: call void @__default_constructor_8_w0(ptr %[[AGG_RESULT]])
 // CHECK: store i1 true, ptr %[[NRVO]], align 1
diff --git a/clang/test/CodeGenObjC/stret-1.m b/clang/test/CodeGenObjC/stret-1.m
index fb1b4abe70f4c..20e26af72c6f9 100644
--- a/clang/test/CodeGenObjC/stret-1.m
+++ b/clang/test/CodeGenObjC/stret-1.m
@@ -13,18 +13,18 @@ int main(int argc, const char **argv)
 {
     struct stret s;
     s = [(id)(argc&~255) method];
-    // CHECK: call void @objc_msgSend(ptr sret(%struct.stret) align 4 [[T0:%[^,]+]]
+    // CHECK: call void @objc_msgSend(ptr dead_on_unwind writable sret(%struct.stret) align 4 [[T0:%[^,]+]]
     // CHECK: call void @llvm.memset.p0.i64(ptr align 4 [[T0]], i8 0, i64 400, i1 false)
 
     s = [Test method];
-    // CHECK: call void @objc_msgSend(ptr sret(%struct.stret) align 4 [[T1:%[^,]+]]
+    // CHECK: call void @objc_msgSend(ptr dead_on_unwind writable sret(%struct.stret) align 4 [[T1:%[^,]+]]
     // CHECK-NOT: call void @llvm.memset.p0.i64(
 
     [(id)(argc&~255) method];
-    // CHECK: call void @objc_msgSend(ptr sret(%struct.stret) align 4 [[T1:%[^,]+]]
+    // CHECK: call void @objc_msgSend(ptr dead_on_unwind writable sret(%struct.stret) align 4 [[T1:%[^,]+]]
     // CHECK-NOT: call void @llvm.memset.p0.i64(
 
     [Test method];
-    // CHECK: call void @objc_msgSend(ptr sret(%struct.stret) align 4 [[T1:%[^,]+]]
+    // CHECK: call void @objc_msgSend(ptr dead_on_unwind writable sret(%struct.stret) align 4 [[T1:%[^,]+]]
     // CHECK-NOT: call void @llvm.memset.p0.i64(
 }
diff --git a/clang/test/CodeGenObjC/stret_lookup.m b/clang/test/CodeGenObjC/stret_lookup.m
index 1a4665d4309db..315a1627da5b7 100644
--- a/clang/test/CodeGenObjC/stret_lookup.m
+++ b/clang/test/CodeGenObjC/stret_lookup.m
@@ -20,8 +20,8 @@ void test0(void) {
 
 // HASSTRET-LABEL: define{{.*}} void @test0()
 // HASSTRET: [[T0:%.*]] = call ptr @objc_msg_lookup_stret(ptr @_OBJC_CLASS_Test0,
-// HASSTRET-NEXT: call void [[T0]](ptr sret(%struct.test) {{.*}}, ptr noundef @_OBJC_CLASS_Test0,
+// HASSTRET-NEXT: call void [[T0]](ptr dead_on_unwind writable sret(%struct.test) {{.*}}, ptr noundef @_OBJC_CLASS_Test0,
 
 // NOSTRET-LABEL: define{{.*}} void @test0()
 // NOSTRET: [[T0:%.*]] = call ptr @objc_msg_lookup(ptr
-// NOSTRET-NEXT: call void [[T0]](ptr sret(%struct.test) {{.*}}, ptr {{.*}}, ptr noundef
+// NOSTRET-NEXT: call void [[T0]](ptr dead_on_unwind writable sret(%struct.test) {{.*}}, ptr {{.*}}, ptr noundef
diff --git a/clang/test/CodeGenObjC/weak-in-c-struct.m b/clang/test/CodeGenObjC/weak-in-c-struct.m
index 3ec08c2b9f18b..be80edd1ff11d 100644
--- a/clang/test/CodeGenObjC/weak-in-c-struct.m
+++ b/clang/test/CodeGenObjC/weak-in-c-struct.m
@@ -149,7 +149,7 @@ void test_argument_Weak(Weak *a) {
   calleeWeak(*a);
 }
 
-// COMMON: define{{.*}} void @test_return_Weak(ptr noalias sret(%[[STRUCT_WEAK]]) align {{.*}} %[[AGG_RESULT:.*]], ptr noundef %[[A:.*]])
+// COMMON: define{{.*}} void @test_return_Weak(ptr dead_on_unwind noalias writable sret(%[[STRUCT_WEAK]]) align {{.*}} %[[AGG_RESULT:.*]], ptr noundef %[[A:.*]])
 // COMMON: %[[A_ADDR:.*]] = alloca ptr
 // COMMON: store ptr %[[A]], ptr %[[A_ADDR]]
 // COMMON: %[[V0:.*]] = load ptr, ptr %[[A_ADDR]]
diff --git a/clang/test/CodeGenObjC/x86_64-struct-return-gc.m b/clang/test/CodeGenObjC/x86_64-struct-return-gc.m
index 6bc2929c7596f..bc0b88ff037ff 100644
--- a/clang/test/CodeGenObjC/x86_64-struct-return-gc.m
+++ b/clang/test/CodeGenObjC/x86_64-struct-return-gc.m
@@ -25,7 +25,7 @@ void Coerce_test(void) {
 void Indirect_test(void) {
   struct Indirect i;
   
-  // CHECK: call void @indirect_func(ptr sret
+  // CHECK: call void @indirect_func(ptr dead_on_unwind writable sret
   // CHECK: call ptr @objc_memmove_collectable(
   i = indirect_func();
 }
diff --git a/clang/test/CodeGenObjCXX/objc-struct-cxx-abi.mm b/clang/test/CodeGenObjCXX/objc-struct-cxx-abi.mm
index ed35eb1a7983a..f9f783ff5caca 100644
--- a/clang/test/CodeGenObjCXX/objc-struct-cxx-abi.mm
+++ b/clang/test/CodeGenObjCXX/objc-struct-cxx-abi.mm
@@ -96,7 +96,7 @@ void testCallStrongWeak(StrongWeak *a) {
   testParamStrongWeak(*a);
 }
 
-// CHECK: define{{.*}} void @_Z20testReturnStrongWeakP10StrongWeak(ptr noalias sret(%[[STRUCT_STRONGWEAK:.*]]) align 8 %[[AGG_RESULT:.*]], ptr noundef %[[A:.*]])
+// CHECK: define{{.*}} void @_Z20testReturnStrongWeakP10StrongWeak(ptr dead_on_unwind noalias writable sret(%[[STRUCT_STRONGWEAK:.*]]) align 8 %[[AGG_RESULT:.*]], ptr noundef %[[A:.*]])
 // CHECK: %[[A_ADDR:a.addr]] = alloca ptr, align 8
 // CHECK: store ptr %[[A]], ptr %[[A_ADDR]], align 8
 // CHECK: %[[V0:.*]] = load ptr, ptr %[[A_ADDR]], align 8
diff --git a/clang/test/CodeGenObjCXX/property-placement-new.mm b/clang/test/CodeGenObjCXX/property-placement-new.mm
new file mode 100644
index 0000000000000..3f4a3a94b7253
--- /dev/null
+++ b/clang/test/CodeGenObjCXX/property-placement-new.mm
@@ -0,0 +1,46 @@
+// RUN: %clang_cc1 -x objective-c++ -std=c++11 %s -triple x86_64-apple-darwin10 -emit-llvm -o - | FileCheck %s
+
+// CHECK: [[NAME:@.*]] = private unnamed_addr constant [9 x i8] c"position\00"
+// CHECK: [[SEL:@.*]] = internal externally_initialized global ptr [[NAME]]
+
+@interface I {
+  int position;
+}
+@property(nonatomic) int position;
+@end
+
+struct S {
+  void *operator new(__SIZE_TYPE__, int);
+};
+
+template 
+struct TS {
+  void *operator new(__SIZE_TYPE__, T);
+};
+
+I *GetI();
+
+int main() {
+  @autoreleasepool {
+    // CHECK: [[I:%.+]] = alloca ptr
+    auto* i = GetI();
+    i.position = 42;
+
+    // This is so we can find the next line more easily.
+    // CHECK: store double
+    double d = 42.0;
+
+    // CHECK: [[I1:%.+]] = load ptr, ptr [[I]]
+    // CHECK-NEXT: [[SEL1:%.+]] = load ptr, ptr [[SEL]]
+    // CHECK-NEXT: [[POS1:%.+]] = call {{.*}} i32 @objc_msgSend(ptr {{.*}} [[I1]], ptr {{.*}} [[SEL1]])
+    // CHECK-NEXT: call {{.*}} ptr @_ZN1SnwEmi(i64 {{.*}} 1, i32 {{.*}} [[POS1]])
+    new (i.position) S;
+
+    // CHECK: [[I2:%.+]] = load ptr, ptr [[I]]
+    // CHECK-NEXT: [[SEL2:%.+]] = load ptr, ptr [[SEL]]
+    // CHECK-NEXT: [[POS2:%.+]] = call {{.*}} i32 @objc_msgSend(ptr {{.*}} [[I2]], ptr {{.*}} [[SEL2]])
+    // CHECK-NEXT: [[DBL:%.+]] = sitofp i32 [[POS2]] to double
+    // CHECK-NEXT: call {{.*}} ptr  @_ZN2TSIdEnwEmd(i64 {{.*}} 1, double {{.*}} [[DBL]])
+    new (i.position) TS;
+  }
+}
diff --git a/clang/test/CodeGenOpenCL/addr-space-struct-arg.cl b/clang/test/CodeGenOpenCL/addr-space-struct-arg.cl
index 054fdc0662228..385f8a753cd8e 100644
--- a/clang/test/CodeGenOpenCL/addr-space-struct-arg.cl
+++ b/clang/test/CodeGenOpenCL/addr-space-struct-arg.cl
@@ -45,7 +45,7 @@ struct LargeStructTwoMember {
 struct LargeStructOneMember g_s;
 #endif
 
-// X86-LABEL: define{{.*}} void @foo(ptr noalias sret(%struct.Mat4X4) align 4 %agg.result, ptr noundef byval(%struct.Mat3X3) align 4 %in)
+// X86-LABEL: define{{.*}} void @foo(ptr dead_on_unwind noalias writable sret(%struct.Mat4X4) align 4 %agg.result, ptr noundef byval(%struct.Mat3X3) align 4 %in)
 // AMDGCN-LABEL: define{{.*}} %struct.Mat4X4 @foo([9 x i32] %in.coerce)
 Mat4X4 __attribute__((noinline)) foo(Mat3X3 in) {
   Mat4X4 out;
@@ -65,8 +65,8 @@ kernel void ker(global Mat3X3 *in, global Mat4X4 *out) {
   out[0] = foo(in[1]);
 }
 
-// X86-LABEL: define{{.*}} void @foo_large(ptr noalias sret(%struct.Mat64X64) align 4 %agg.result, ptr noundef byval(%struct.Mat32X32) align 4 %in)
-// AMDGCN-LABEL: define{{.*}} void @foo_large(ptr addrspace(5) noalias sret(%struct.Mat64X64) align 4 %agg.result, ptr addrspace(5) noundef byref(%struct.Mat32X32) align 4 %{{.*}}
+// X86-LABEL: define{{.*}} void @foo_large(ptr dead_on_unwind noalias writable sret(%struct.Mat64X64) align 4 %agg.result, ptr noundef byval(%struct.Mat32X32) align 4 %in)
+// AMDGCN-LABEL: define{{.*}} void @foo_large(ptr addrspace(5) dead_on_unwind noalias writable sret(%struct.Mat64X64) align 4 %agg.result, ptr addrspace(5) noundef byref(%struct.Mat32X32) align 4 %{{.*}}
 // AMDGCN:       %in = alloca %struct.Mat32X32, align 4, addrspace(5)
 // AMDGCN-NEXT:  call void @llvm.memcpy.p5.p5.i64(ptr addrspace(5) align 4 %in, ptr addrspace(5) align 4 %{{.*}}, i64 4096, i1 false)
 Mat64X64 __attribute__((noinline)) foo_large(Mat32X32 in) {
diff --git a/clang/test/CodeGenOpenCL/amdgpu-abi-struct-arg-byref.cl b/clang/test/CodeGenOpenCL/amdgpu-abi-struct-arg-byref.cl
index 52fad9599de04..fa83a38a01b0a 100644
--- a/clang/test/CodeGenOpenCL/amdgpu-abi-struct-arg-byref.cl
+++ b/clang/test/CodeGenOpenCL/amdgpu-abi-struct-arg-byref.cl
@@ -86,7 +86,7 @@ kernel void ker(global Mat3X3 *in, global Mat4X4 *out) {
 }
 
 // AMDGCN-LABEL: define dso_local void @foo_large
-// AMDGCN-SAME: (ptr addrspace(5) noalias sret([[STRUCT_MAT64X64:%.*]]) align 4 [[AGG_RESULT:%.*]], ptr addrspace(5) noundef byref([[STRUCT_MAT32X32:%.*]]) align 4 [[TMP0:%.*]]) #[[ATTR0]] {
+// AMDGCN-SAME: (ptr addrspace(5) dead_on_unwind noalias writable sret([[STRUCT_MAT64X64:%.*]]) align 4 [[AGG_RESULT:%.*]], ptr addrspace(5) noundef byref([[STRUCT_MAT32X32:%.*]]) align 4 [[TMP0:%.*]]) #[[ATTR0]] {
 // AMDGCN-NEXT:  entry:
 // AMDGCN-NEXT:    [[IN:%.*]] = alloca [[STRUCT_MAT32X32]], align 4, addrspace(5)
 // AMDGCN-NEXT:    call void @llvm.memcpy.p5.p5.i64(ptr addrspace(5) align 4 [[IN]], ptr addrspace(5) align 4 [[TMP0]], i64 4096, i1 false)
@@ -111,7 +111,7 @@ Mat64X64 __attribute__((noinline)) foo_large(Mat32X32 in) {
 // AMDGCN-NEXT:    [[TMP1:%.*]] = load ptr addrspace(1), ptr addrspace(5) [[IN_ADDR]], align 8
 // AMDGCN-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds [[STRUCT_MAT32X32]], ptr addrspace(1) [[TMP1]], i64 1
 // AMDGCN-NEXT:    call void @llvm.memcpy.p5.p1.i64(ptr addrspace(5) align 4 [[BYVAL_TEMP]], ptr addrspace(1) align 4 [[ARRAYIDX1]], i64 4096, i1 false)
-// AMDGCN-NEXT:    call void @foo_large(ptr addrspace(5) sret([[STRUCT_MAT64X64]]) align 4 [[TMP]], ptr addrspace(5) noundef byref([[STRUCT_MAT32X32]]) align 4 [[BYVAL_TEMP]]) #[[ATTR3]]
+// AMDGCN-NEXT:    call void @foo_large(ptr addrspace(5) dead_on_unwind writable sret([[STRUCT_MAT64X64]]) align 4 [[TMP]], ptr addrspace(5) noundef byref([[STRUCT_MAT32X32]]) align 4 [[BYVAL_TEMP]]) #[[ATTR3]]
 // AMDGCN-NEXT:    call void @llvm.memcpy.p1.p5.i64(ptr addrspace(1) align 4 [[ARRAYIDX]], ptr addrspace(5) align 4 [[TMP]], i64 16384, i1 false)
 // AMDGCN-NEXT:    ret void
 //
diff --git a/clang/test/CodeGenOpenCL/amdgpu-abi-struct-coerce.cl b/clang/test/CodeGenOpenCL/amdgpu-abi-struct-coerce.cl
index 665609e54a83e..90e4b65c6f1f7 100644
--- a/clang/test/CodeGenOpenCL/amdgpu-abi-struct-coerce.cl
+++ b/clang/test/CodeGenOpenCL/amdgpu-abi-struct-coerce.cl
@@ -402,14 +402,14 @@ struct_arr16 func_ret_struct_arr16()
   return s;
 }
 
-// CHECK: define{{.*}} void @func_ret_struct_arr32(ptr addrspace(5) noalias nocapture writeonly sret(%struct.struct_arr32) align 4 %agg.result)
+// CHECK: define{{.*}} void @func_ret_struct_arr32(ptr addrspace(5) dead_on_unwind noalias nocapture writable writeonly sret(%struct.struct_arr32) align 4 %agg.result)
 struct_arr32 func_ret_struct_arr32()
 {
   struct_arr32 s = { 0 };
   return s;
 }
 
-// CHECK: define{{.*}} void @func_ret_struct_arr33(ptr addrspace(5) noalias nocapture writeonly sret(%struct.struct_arr33) align 4 %agg.result)
+// CHECK: define{{.*}} void @func_ret_struct_arr33(ptr addrspace(5) dead_on_unwind noalias nocapture writable writeonly sret(%struct.struct_arr33) align 4 %agg.result)
 struct_arr33 func_ret_struct_arr33()
 {
   struct_arr33 s = { 0 };
@@ -438,7 +438,7 @@ different_size_type_pair func_different_size_type_pair_ret()
   return s;
 }
 
-// CHECK: define{{.*}} void @func_flexible_array_ret(ptr addrspace(5) noalias nocapture writeonly sret(%struct.flexible_array) align 4 %agg.result)
+// CHECK: define{{.*}} void @func_flexible_array_ret(ptr addrspace(5) dead_on_unwind noalias nocapture writable writeonly sret(%struct.flexible_array) align 4 %agg.result)
 flexible_array func_flexible_array_ret()
 {
   flexible_array s = { 0 };
diff --git a/clang/test/CodeGenOpenCL/amdgpu-features.cl b/clang/test/CodeGenOpenCL/amdgpu-features.cl
index 8959634572b44..7495bca72a9df 100644
--- a/clang/test/CodeGenOpenCL/amdgpu-features.cl
+++ b/clang/test/CodeGenOpenCL/amdgpu-features.cl
@@ -80,9 +80,9 @@
 // GFX909: "target-features"="+16-bit-insts,+ci-insts,+dpp,+gfx8-insts,+gfx9-insts,+s-memrealtime,+s-memtime-inst,+wavefrontsize64"
 // GFX90A: "target-features"="+16-bit-insts,+atomic-buffer-global-pk-add-f16-insts,+atomic-fadd-rtn-insts,+ci-insts,+dl-insts,+dot1-insts,+dot10-insts,+dot2-insts,+dot3-insts,+dot4-insts,+dot5-insts,+dot6-insts,+dot7-insts,+dpp,+gfx8-insts,+gfx9-insts,+gfx90a-insts,+mai-insts,+s-memrealtime,+s-memtime-inst,+wavefrontsize64"
 // GFX90C: "target-features"="+16-bit-insts,+ci-insts,+dpp,+gfx8-insts,+gfx9-insts,+s-memrealtime,+s-memtime-inst,+wavefrontsize64"
-// GFX940: "target-features"="+16-bit-insts,+atomic-buffer-global-pk-add-f16-insts,+atomic-ds-pk-add-16-insts,+atomic-fadd-rtn-insts,+atomic-flat-pk-add-16-insts,+atomic-global-pk-add-bf16-inst,+ci-insts,+dl-insts,+dot1-insts,+dot10-insts,+dot2-insts,+dot3-insts,+dot4-insts,+dot5-insts,+dot6-insts,+dot7-insts,+dpp,+fp8-insts,+gfx8-insts,+gfx9-insts,+gfx90a-insts,+gfx940-insts,+mai-insts,+s-memrealtime,+s-memtime-inst,+wavefrontsize64"
-// GFX941: "target-features"="+16-bit-insts,+atomic-buffer-global-pk-add-f16-insts,+atomic-ds-pk-add-16-insts,+atomic-fadd-rtn-insts,+atomic-flat-pk-add-16-insts,+atomic-global-pk-add-bf16-inst,+ci-insts,+dl-insts,+dot1-insts,+dot10-insts,+dot2-insts,+dot3-insts,+dot4-insts,+dot5-insts,+dot6-insts,+dot7-insts,+dpp,+fp8-insts,+gfx8-insts,+gfx9-insts,+gfx90a-insts,+gfx940-insts,+mai-insts,+s-memrealtime,+s-memtime-inst,+wavefrontsize64"
-// GFX942: "target-features"="+16-bit-insts,+atomic-buffer-global-pk-add-f16-insts,+atomic-ds-pk-add-16-insts,+atomic-fadd-rtn-insts,+atomic-flat-pk-add-16-insts,+atomic-global-pk-add-bf16-inst,+ci-insts,+dl-insts,+dot1-insts,+dot10-insts,+dot2-insts,+dot3-insts,+dot4-insts,+dot5-insts,+dot6-insts,+dot7-insts,+dpp,+fp8-insts,+gfx8-insts,+gfx9-insts,+gfx90a-insts,+gfx940-insts,+mai-insts,+s-memrealtime,+s-memtime-inst,+wavefrontsize64"
+// GFX940: "target-features"="+16-bit-insts,+atomic-buffer-global-pk-add-f16-insts,+atomic-ds-pk-add-16-insts,+atomic-fadd-rtn-insts,+atomic-flat-pk-add-16-insts,+atomic-global-pk-add-bf16-inst,+ci-insts,+dl-insts,+dot1-insts,+dot10-insts,+dot2-insts,+dot3-insts,+dot4-insts,+dot5-insts,+dot6-insts,+dot7-insts,+dpp,+fp8-conversion-insts,+fp8-insts,+gfx8-insts,+gfx9-insts,+gfx90a-insts,+gfx940-insts,+mai-insts,+s-memrealtime,+s-memtime-inst,+wavefrontsize64"
+// GFX941: "target-features"="+16-bit-insts,+atomic-buffer-global-pk-add-f16-insts,+atomic-ds-pk-add-16-insts,+atomic-fadd-rtn-insts,+atomic-flat-pk-add-16-insts,+atomic-global-pk-add-bf16-inst,+ci-insts,+dl-insts,+dot1-insts,+dot10-insts,+dot2-insts,+dot3-insts,+dot4-insts,+dot5-insts,+dot6-insts,+dot7-insts,+dpp,+fp8-conversion-insts,+fp8-insts,+gfx8-insts,+gfx9-insts,+gfx90a-insts,+gfx940-insts,+mai-insts,+s-memrealtime,+s-memtime-inst,+wavefrontsize64"
+// GFX942: "target-features"="+16-bit-insts,+atomic-buffer-global-pk-add-f16-insts,+atomic-ds-pk-add-16-insts,+atomic-fadd-rtn-insts,+atomic-flat-pk-add-16-insts,+atomic-global-pk-add-bf16-inst,+ci-insts,+dl-insts,+dot1-insts,+dot10-insts,+dot2-insts,+dot3-insts,+dot4-insts,+dot5-insts,+dot6-insts,+dot7-insts,+dpp,+fp8-conversion-insts,+fp8-insts,+gfx8-insts,+gfx9-insts,+gfx90a-insts,+gfx940-insts,+mai-insts,+s-memrealtime,+s-memtime-inst,+wavefrontsize64"
 // GFX1010: "target-features"="+16-bit-insts,+ci-insts,+dl-insts,+dpp,+gfx10-insts,+gfx8-insts,+gfx9-insts,+s-memrealtime,+s-memtime-inst,+wavefrontsize32"
 // GFX1011: "target-features"="+16-bit-insts,+ci-insts,+dl-insts,+dot1-insts,+dot10-insts,+dot2-insts,+dot5-insts,+dot6-insts,+dot7-insts,+dpp,+gfx10-insts,+gfx8-insts,+gfx9-insts,+s-memrealtime,+s-memtime-inst,+wavefrontsize32"
 // GFX1012: "target-features"="+16-bit-insts,+ci-insts,+dl-insts,+dot1-insts,+dot10-insts,+dot2-insts,+dot5-insts,+dot6-insts,+dot7-insts,+dpp,+gfx10-insts,+gfx8-insts,+gfx9-insts,+s-memrealtime,+s-memtime-inst,+wavefrontsize32"
@@ -100,8 +100,8 @@
 // GFX1103: "target-features"="+16-bit-insts,+atomic-fadd-rtn-insts,+ci-insts,+dl-insts,+dot10-insts,+dot5-insts,+dot7-insts,+dot8-insts,+dot9-insts,+dpp,+gfx10-3-insts,+gfx10-insts,+gfx11-insts,+gfx8-insts,+gfx9-insts,+wavefrontsize32"
 // GFX1150: "target-features"="+16-bit-insts,+atomic-fadd-rtn-insts,+ci-insts,+dl-insts,+dot10-insts,+dot5-insts,+dot7-insts,+dot8-insts,+dot9-insts,+dpp,+gfx10-3-insts,+gfx10-insts,+gfx11-insts,+gfx8-insts,+gfx9-insts,+wavefrontsize32"
 // GFX1151: "target-features"="+16-bit-insts,+atomic-fadd-rtn-insts,+ci-insts,+dl-insts,+dot10-insts,+dot5-insts,+dot7-insts,+dot8-insts,+dot9-insts,+dpp,+gfx10-3-insts,+gfx10-insts,+gfx11-insts,+gfx8-insts,+gfx9-insts,+wavefrontsize32"
-// GFX1200: "target-features"="+16-bit-insts,+atomic-fadd-rtn-insts,+ci-insts,+dl-insts,+dot10-insts,+dot5-insts,+dot7-insts,+dot8-insts,+dot9-insts,+dpp,+gfx10-3-insts,+gfx10-insts,+gfx11-insts,+gfx12-insts,+gfx8-insts,+gfx9-insts,+wavefrontsize32"
-// GFX1201: "target-features"="+16-bit-insts,+atomic-fadd-rtn-insts,+ci-insts,+dl-insts,+dot10-insts,+dot5-insts,+dot7-insts,+dot8-insts,+dot9-insts,+dpp,+gfx10-3-insts,+gfx10-insts,+gfx11-insts,+gfx12-insts,+gfx8-insts,+gfx9-insts,+wavefrontsize32"
+// GFX1200: "target-features"="+16-bit-insts,+atomic-buffer-global-pk-add-f16-insts,+atomic-ds-pk-add-16-insts,+atomic-fadd-rtn-insts,+atomic-flat-pk-add-16-insts,+atomic-global-pk-add-bf16-inst,+ci-insts,+dl-insts,+dot10-insts,+dot5-insts,+dot7-insts,+dot8-insts,+dot9-insts,+dpp,+gfx10-3-insts,+gfx10-insts,+gfx11-insts,+gfx12-insts,+gfx8-insts,+gfx9-insts,+wavefrontsize32"
+// GFX1201: "target-features"="+16-bit-insts,+atomic-buffer-global-pk-add-f16-insts,+atomic-ds-pk-add-16-insts,+atomic-fadd-rtn-insts,+atomic-flat-pk-add-16-insts,+atomic-global-pk-add-bf16-inst,+ci-insts,+dl-insts,+dot10-insts,+dot5-insts,+dot7-insts,+dot8-insts,+dot9-insts,+dpp,+gfx10-3-insts,+gfx10-insts,+gfx11-insts,+gfx12-insts,+gfx8-insts,+gfx9-insts,+wavefrontsize32"
 
 // GFX1103-W64: "target-features"="+16-bit-insts,+atomic-fadd-rtn-insts,+ci-insts,+dl-insts,+dot10-insts,+dot5-insts,+dot7-insts,+dot8-insts,+dot9-insts,+dpp,+gfx10-3-insts,+gfx10-insts,+gfx11-insts,+gfx8-insts,+gfx9-insts,+wavefrontsize64"
 
diff --git a/clang/test/CodeGenOpenCL/builtins-amdgcn-dl-insts-err.cl b/clang/test/CodeGenOpenCL/builtins-amdgcn-dl-insts-err.cl
index 6573325150d95..f5317683d0ff9 100644
--- a/clang/test/CodeGenOpenCL/builtins-amdgcn-dl-insts-err.cl
+++ b/clang/test/CodeGenOpenCL/builtins-amdgcn-dl-insts-err.cl
@@ -49,4 +49,9 @@ kernel void builtins_amdgcn_dl_insts_err(
 
   iOut[3] = __builtin_amdgcn_sudot8(false, A, true, B, C, false);    // expected-error {{'__builtin_amdgcn_sudot8' needs target feature dot8-insts}}
   iOut[4] = __builtin_amdgcn_sudot8(true, A, false, B, C, true);     // expected-error {{'__builtin_amdgcn_sudot8' needs target feature dot8-insts}}
+
+  fOut[5] = __builtin_amdgcn_dot4_f32_fp8_bf8(uiA, uiB, fC);        // expected-error {{'__builtin_amdgcn_dot4_f32_fp8_bf8' needs target feature gfx12-insts}}
+  fOut[6] = __builtin_amdgcn_dot4_f32_bf8_fp8(uiA, uiB, fC);        // expected-error {{'__builtin_amdgcn_dot4_f32_bf8_fp8' needs target feature gfx12-insts}}
+  fOut[7] = __builtin_amdgcn_dot4_f32_fp8_fp8(uiA, uiB, fC);        // expected-error {{'__builtin_amdgcn_dot4_f32_fp8_fp8' needs target feature gfx12-insts}}
+  fOut[8] = __builtin_amdgcn_dot4_f32_bf8_bf8(uiA, uiB, fC);        // expected-error {{'__builtin_amdgcn_dot4_f32_bf8_bf8' needs target feature gfx12-insts}}
 }
diff --git a/clang/test/CodeGenOpenCL/builtins-amdgcn-dl-insts-gfx12.cl b/clang/test/CodeGenOpenCL/builtins-amdgcn-dl-insts-gfx12.cl
new file mode 100644
index 0000000000000..087883e9f5608
--- /dev/null
+++ b/clang/test/CodeGenOpenCL/builtins-amdgcn-dl-insts-gfx12.cl
@@ -0,0 +1,20 @@
+// REQUIRES: amdgpu-registered-target
+
+// RUN: %clang_cc1 -triple amdgcn-unknown-unknown -target-cpu gfx1200 -S -emit-llvm -o - %s | FileCheck %s
+
+typedef unsigned int uint;
+
+// CHECK-LABEL: @builtins_amdgcn_dl_insts
+// CHECK: call float @llvm.amdgcn.dot4.f32.fp8.bf8(i32 %uiA, i32 %uiB, float %fC)
+// CHECK: call float @llvm.amdgcn.dot4.f32.bf8.fp8(i32 %uiA, i32 %uiB, float %fC)
+// CHECK: call float @llvm.amdgcn.dot4.f32.fp8.fp8(i32 %uiA, i32 %uiB, float %fC)
+// CHECK: call float @llvm.amdgcn.dot4.f32.bf8.bf8(i32 %uiA, i32 %uiB, float %fC)
+
+#pragma OPENCL EXTENSION cl_khr_fp16 : enable
+kernel void builtins_amdgcn_dl_insts_err(global float *fOut,
+                                         uint uiA, uint uiB, float fC) {
+  fOut[0] = __builtin_amdgcn_dot4_f32_fp8_bf8(uiA, uiB, fC);
+  fOut[1] = __builtin_amdgcn_dot4_f32_bf8_fp8(uiA, uiB, fC);
+  fOut[2] = __builtin_amdgcn_dot4_f32_fp8_fp8(uiA, uiB, fC);
+  fOut[3] = __builtin_amdgcn_dot4_f32_bf8_bf8(uiA, uiB, fC);
+}
diff --git a/clang/test/CodeGenOpenCL/builtins-amdgcn-gfx11-err.cl b/clang/test/CodeGenOpenCL/builtins-amdgcn-gfx11-err.cl
index 00ecf32d94923..622e9dd2eed42 100644
--- a/clang/test/CodeGenOpenCL/builtins-amdgcn-gfx11-err.cl
+++ b/clang/test/CodeGenOpenCL/builtins-amdgcn-gfx11-err.cl
@@ -2,10 +2,7 @@
 
 // RUN: %clang_cc1 -triple amdgcn-unknown-unknown -target-cpu gfx1100 -verify -S -emit-llvm -o - %s
 
-typedef unsigned int uint;
-typedef uint uint2 __attribute__((ext_vector_type(2)));
-typedef uint uint4 __attribute__((ext_vector_type(4)));
-
-kernel void builtins_amdgcn_bvh_err(global uint2* out, uint addr, uint data, uint4 data1, uint offset) {
-  *out = __builtin_amdgcn_ds_bvh_stack_rtn(addr, data, data1, offset); // expected-error {{'__builtin_amdgcn_ds_bvh_stack_rtn' must be a constant integer}}
+void test_s_sleep_var(int d)
+{
+  __builtin_amdgcn_s_sleep_var(d); // expected-error {{'__builtin_amdgcn_s_sleep_var' needs target feature gfx12-insts}}
 }
diff --git a/clang/test/CodeGenOpenCL/builtins-amdgcn-gfx11-param-err.cl b/clang/test/CodeGenOpenCL/builtins-amdgcn-gfx11-param-err.cl
new file mode 100644
index 0000000000000..00ecf32d94923
--- /dev/null
+++ b/clang/test/CodeGenOpenCL/builtins-amdgcn-gfx11-param-err.cl
@@ -0,0 +1,11 @@
+// REQUIRES: amdgpu-registered-target
+
+// RUN: %clang_cc1 -triple amdgcn-unknown-unknown -target-cpu gfx1100 -verify -S -emit-llvm -o - %s
+
+typedef unsigned int uint;
+typedef uint uint2 __attribute__((ext_vector_type(2)));
+typedef uint uint4 __attribute__((ext_vector_type(4)));
+
+kernel void builtins_amdgcn_bvh_err(global uint2* out, uint addr, uint data, uint4 data1, uint offset) {
+  *out = __builtin_amdgcn_ds_bvh_stack_rtn(addr, data, data1, offset); // expected-error {{'__builtin_amdgcn_ds_bvh_stack_rtn' must be a constant integer}}
+}
diff --git a/clang/test/CodeGenOpenCL/builtins-amdgcn-gfx12.cl b/clang/test/CodeGenOpenCL/builtins-amdgcn-gfx12.cl
index 2899d9e5c2889..ebd367bba0cdc 100644
--- a/clang/test/CodeGenOpenCL/builtins-amdgcn-gfx12.cl
+++ b/clang/test/CodeGenOpenCL/builtins-amdgcn-gfx12.cl
@@ -5,6 +5,21 @@
 
 typedef unsigned int uint;
 
+// CHECK-LABEL: @test_s_sleep_var(
+// CHECK-NEXT:  entry:
+// CHECK-NEXT:    [[D_ADDR:%.*]] = alloca i32, align 4, addrspace(5)
+// CHECK-NEXT:    store i32 [[D:%.*]], ptr addrspace(5) [[D_ADDR]], align 4
+// CHECK-NEXT:    [[TMP0:%.*]] = load i32, ptr addrspace(5) [[D_ADDR]], align 4
+// CHECK-NEXT:    call void @llvm.amdgcn.s.sleep.var(i32 [[TMP0]])
+// CHECK-NEXT:    call void @llvm.amdgcn.s.sleep.var(i32 15)
+// CHECK-NEXT:    ret void
+//
+void test_s_sleep_var(int d)
+{
+  __builtin_amdgcn_s_sleep_var(d);
+  __builtin_amdgcn_s_sleep_var(15);
+}
+
 // CHECK-LABEL: @test_permlane16_var(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[OUT_ADDR:%.*]] = alloca ptr addrspace(1), align 8, addrspace(5)
diff --git a/clang/test/CodeGenOpenCL/builtins-amdgcn-global-load-tr-gfx11-err.cl b/clang/test/CodeGenOpenCL/builtins-amdgcn-global-load-tr-gfx11-err.cl
new file mode 100644
index 0000000000000..f7afb7cb97eda
--- /dev/null
+++ b/clang/test/CodeGenOpenCL/builtins-amdgcn-global-load-tr-gfx11-err.cl
@@ -0,0 +1,26 @@
+// RUN: %clang_cc1 -O0 -cl-std=CL2.0 -triple amdgcn-amd-amdhsa -target-cpu gfx1100 -emit-llvm \
+// RUN:   -verify -S -o - %s
+
+// REQUIRES: amdgpu-registered-target
+
+typedef int    v2i   __attribute__((ext_vector_type(2)));
+typedef half   v8h   __attribute__((ext_vector_type(8)));
+typedef short  v8s   __attribute__((ext_vector_type(8)));
+
+typedef half   v4h   __attribute__((ext_vector_type(4)));
+typedef short  v4s   __attribute__((ext_vector_type(4)));
+
+
+
+void amdgcn_global_load_tr(global v2i* v2i_inptr, global v8s* v8s_inptr, global v8h* v8h_inptr,
+                           global int* int_inptr, global v4s* v4s_inptr, global v4h* v4h_inptr)
+{
+  v2i out_1 = __builtin_amdgcn_global_load_tr_v2i32(v2i_inptr); // expected-error{{'__builtin_amdgcn_global_load_tr_v2i32' needs target feature gfx12-insts,wavefrontsize32}}
+  v8s out_2 = __builtin_amdgcn_global_load_tr_v8i16(v8s_inptr); // expected-error{{'__builtin_amdgcn_global_load_tr_v8i16' needs target feature gfx12-insts,wavefrontsize32}}
+  v8h out_3 = __builtin_amdgcn_global_load_tr_v8f16(v8h_inptr); // expected-error{{'__builtin_amdgcn_global_load_tr_v8f16' needs target feature gfx12-insts,wavefrontsize32}}
+
+  int out_4 = __builtin_amdgcn_global_load_tr_i32(int_inptr); // expected-error{{'__builtin_amdgcn_global_load_tr_i32' needs target feature gfx12-insts,wavefrontsize64}}
+  v4s out_5 = __builtin_amdgcn_global_load_tr_v4i16(v4s_inptr); // expected-error{{'__builtin_amdgcn_global_load_tr_v4i16' needs target feature gfx12-insts,wavefrontsize64}}
+  v4h out_6 = __builtin_amdgcn_global_load_tr_v4f16(v4h_inptr); // expected-error{{'__builtin_amdgcn_global_load_tr_v4f16' needs target feature gfx12-insts,wavefrontsize64}}
+}
+
diff --git a/clang/test/CodeGenOpenCL/builtins-amdgcn-global-load-tr-gfx12-w32-err.cl b/clang/test/CodeGenOpenCL/builtins-amdgcn-global-load-tr-gfx12-w32-err.cl
new file mode 100644
index 0000000000000..04ac0a66db7ce
--- /dev/null
+++ b/clang/test/CodeGenOpenCL/builtins-amdgcn-global-load-tr-gfx12-w32-err.cl
@@ -0,0 +1,15 @@
+// RUN: %clang_cc1 -O0 -cl-std=CL2.0 -triple amdgcn-amd-amdhsa -target-cpu gfx1200 -target-feature +wavefrontsize32 -emit-llvm \
+// RUN:   -verify -S -o - %s
+
+// REQUIRES: amdgpu-registered-target
+
+typedef half   v4h   __attribute__((ext_vector_type(4)));
+typedef short  v4s   __attribute__((ext_vector_type(4)));
+
+void amdgcn_global_load_tr(global int* int_inptr, global v4s* v4s_inptr, global v4h* v4h_inptr)
+{
+  int out_4 = __builtin_amdgcn_global_load_tr_i32(int_inptr); // expected-error{{'__builtin_amdgcn_global_load_tr_i32' needs target feature gfx12-insts,wavefrontsize64}}
+  v4s out_5 = __builtin_amdgcn_global_load_tr_v4i16(v4s_inptr); // expected-error{{'__builtin_amdgcn_global_load_tr_v4i16' needs target feature gfx12-insts,wavefrontsize64}}
+  v4h out_6 = __builtin_amdgcn_global_load_tr_v4f16(v4h_inptr); // expected-error{{'__builtin_amdgcn_global_load_tr_v4f16' needs target feature gfx12-insts,wavefrontsize64}}
+}
+
diff --git a/clang/test/CodeGenOpenCL/builtins-amdgcn-global-load-tr-gfx12-w64-err.cl b/clang/test/CodeGenOpenCL/builtins-amdgcn-global-load-tr-gfx12-w64-err.cl
new file mode 100644
index 0000000000000..113b54b853a9f
--- /dev/null
+++ b/clang/test/CodeGenOpenCL/builtins-amdgcn-global-load-tr-gfx12-w64-err.cl
@@ -0,0 +1,16 @@
+// RUN: %clang_cc1 -O0 -cl-std=CL2.0 -triple amdgcn-amd-amdhsa -target-cpu gfx1200 -target-feature +wavefrontsize64 -emit-llvm \
+// RUN:   -verify -S -o - %s
+
+// REQUIRES: amdgpu-registered-target
+
+typedef int    v2i   __attribute__((ext_vector_type(2)));
+typedef half   v8h   __attribute__((ext_vector_type(8)));
+typedef short  v8s   __attribute__((ext_vector_type(8)));
+
+void amdgcn_global_load_tr(global v2i* v2i_inptr, global v8s* v8s_inptr, global v8h* v8h_inptr)
+{
+  v2i out_1 = __builtin_amdgcn_global_load_tr_v2i32(v2i_inptr); // expected-error{{'__builtin_amdgcn_global_load_tr_v2i32' needs target feature gfx12-insts,wavefrontsize32}}
+  v8s out_2 = __builtin_amdgcn_global_load_tr_v8i16(v8s_inptr); // expected-error{{'__builtin_amdgcn_global_load_tr_v8i16' needs target feature gfx12-insts,wavefrontsize32}}
+  v8h out_3 = __builtin_amdgcn_global_load_tr_v8f16(v8h_inptr); // expected-error{{'__builtin_amdgcn_global_load_tr_v8f16' needs target feature gfx12-insts,wavefrontsize32}}
+}
+
diff --git a/clang/test/CodeGenOpenCL/builtins-amdgcn-global-load-tr-w32.cl b/clang/test/CodeGenOpenCL/builtins-amdgcn-global-load-tr-w32.cl
new file mode 100644
index 0000000000000..b5fcad68a4702
--- /dev/null
+++ b/clang/test/CodeGenOpenCL/builtins-amdgcn-global-load-tr-w32.cl
@@ -0,0 +1,48 @@
+// NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py
+// REQUIRES: amdgpu-registered-target
+// RUN: %clang_cc1 -triple amdgcn-unknown-unknown -target-cpu gfx1200 -target-feature +wavefrontsize32 -S -emit-llvm -o - %s | FileCheck %s --check-prefix=CHECK-GFX1200
+
+typedef int    v2i   __attribute__((ext_vector_type(2)));
+typedef half   v8h   __attribute__((ext_vector_type(8)));
+typedef short  v8s   __attribute__((ext_vector_type(8)));
+
+// Wave32
+
+//
+// amdgcn_global_load_tr
+//
+
+// CHECK-GFX1200-LABEL: @test_amdgcn_global_load_tr_v2i32(
+// CHECK-GFX1200-NEXT:  entry:
+// CHECK-GFX1200-NEXT:    [[TMP0:%.*]] = tail call <2 x i32> @llvm.amdgcn.global.load.tr.v2i32(ptr addrspace(1) [[INPTR:%.*]])
+// CHECK-GFX1200-NEXT:    ret <2 x i32> [[TMP0]]
+//
+v2i test_amdgcn_global_load_tr_v2i32(global v2i* inptr)
+{
+  return __builtin_amdgcn_global_load_tr_v2i32(inptr);
+}
+
+//
+// amdgcn_global_load_tr
+//
+
+// CHECK-GFX1200-LABEL: @test_amdgcn_global_load_tr_v8i16(
+// CHECK-GFX1200-NEXT:  entry:
+// CHECK-GFX1200-NEXT:    [[TMP0:%.*]] = tail call <8 x i16> @llvm.amdgcn.global.load.tr.v8i16(ptr addrspace(1) [[INPTR:%.*]])
+// CHECK-GFX1200-NEXT:    ret <8 x i16> [[TMP0]]
+//
+v8s test_amdgcn_global_load_tr_v8i16(global v8s* inptr)
+{
+  return __builtin_amdgcn_global_load_tr_v8i16(inptr);
+}
+
+// CHECK-GFX1200-LABEL: @test_amdgcn_global_load_tr_v8f16(
+// CHECK-GFX1200-NEXT:  entry:
+// CHECK-GFX1200-NEXT:    [[TMP0:%.*]] = tail call <8 x half> @llvm.amdgcn.global.load.tr.v8f16(ptr addrspace(1) [[INPTR:%.*]])
+// CHECK-GFX1200-NEXT:    ret <8 x half> [[TMP0]]
+//
+v8h test_amdgcn_global_load_tr_v8f16(global v8h* inptr)
+{
+  return __builtin_amdgcn_global_load_tr_v8f16(inptr);
+}
+
diff --git a/clang/test/CodeGenOpenCL/builtins-amdgcn-global-load-tr-w64.cl b/clang/test/CodeGenOpenCL/builtins-amdgcn-global-load-tr-w64.cl
new file mode 100644
index 0000000000000..9c48ac071b4d3
--- /dev/null
+++ b/clang/test/CodeGenOpenCL/builtins-amdgcn-global-load-tr-w64.cl
@@ -0,0 +1,47 @@
+// NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py
+// REQUIRES: amdgpu-registered-target
+// RUN: %clang_cc1 -triple amdgcn-unknown-unknown -target-cpu gfx1200 -target-feature +wavefrontsize64 -S -emit-llvm -o - %s | FileCheck %s --check-prefix=CHECK-GFX1200
+
+typedef half   v4h   __attribute__((ext_vector_type(4)));
+typedef short  v4s   __attribute__((ext_vector_type(4)));
+
+// Wave64
+
+//
+// amdgcn_global_load_tr
+//
+
+// CHECK-GFX1200-LABEL: @test_amdgcn_global_load_tr_i32(
+// CHECK-GFX1200-NEXT:  entry:
+// CHECK-GFX1200-NEXT:    [[TMP0:%.*]] = tail call i32 @llvm.amdgcn.global.load.tr.i32(ptr addrspace(1) [[INPTR:%.*]])
+// CHECK-GFX1200-NEXT:    ret i32 [[TMP0]]
+//
+int test_amdgcn_global_load_tr_i32(global int* inptr)
+{
+  return __builtin_amdgcn_global_load_tr_i32(inptr);
+}
+
+//
+// amdgcn_global_load_tr
+//
+
+// CHECK-GFX1200-LABEL: @test_amdgcn_global_load_tr_v4i16(
+// CHECK-GFX1200-NEXT:  entry:
+// CHECK-GFX1200-NEXT:    [[TMP0:%.*]] = tail call <4 x i16> @llvm.amdgcn.global.load.tr.v4i16(ptr addrspace(1) [[INPTR:%.*]])
+// CHECK-GFX1200-NEXT:    ret <4 x i16> [[TMP0]]
+//
+v4s test_amdgcn_global_load_tr_v4i16(global v4s* inptr)
+{
+  return __builtin_amdgcn_global_load_tr_v4i16(inptr);
+}
+
+// CHECK-GFX1200-LABEL: @test_amdgcn_global_load_tr_v4f16(
+// CHECK-GFX1200-NEXT:  entry:
+// CHECK-GFX1200-NEXT:    [[TMP0:%.*]] = tail call <4 x half> @llvm.amdgcn.global.load.tr.v4f16(ptr addrspace(1) [[INPTR:%.*]])
+// CHECK-GFX1200-NEXT:    ret <4 x half> [[TMP0]]
+//
+v4h test_amdgcn_global_load_tr_v4f16(global v4h* inptr)
+{
+  return __builtin_amdgcn_global_load_tr_v4f16(inptr);
+}
+
diff --git a/clang/test/CodeGenOpenCL/builtins-fp-atomics-gfx12.cl b/clang/test/CodeGenOpenCL/builtins-fp-atomics-gfx12.cl
new file mode 100644
index 0000000000000..20ff12c337637
--- /dev/null
+++ b/clang/test/CodeGenOpenCL/builtins-fp-atomics-gfx12.cl
@@ -0,0 +1,92 @@
+// RUN: %clang_cc1 -O0 -cl-std=CL2.0 -triple amdgcn-amd-amdhsa -target-cpu gfx1200 \
+// RUN:   %s -S -emit-llvm -o - | FileCheck %s
+
+// RUN: %clang_cc1 -O0 -cl-std=CL2.0 -triple amdgcn-amd-amdhsa -target-cpu gfx1200 \
+// RUN:   -S -o - %s | FileCheck -check-prefix=GFX12 %s
+
+// REQUIRES: amdgpu-registered-target
+
+typedef half  __attribute__((ext_vector_type(2))) half2;
+typedef short __attribute__((ext_vector_type(2))) short2;
+
+// CHECK-LABEL: test_local_add_2bf16
+// CHECK: call <2 x i16> @llvm.amdgcn.ds.fadd.v2bf16(ptr addrspace(3) %{{.*}}, <2 x i16> %
+// GFX12-LABEL:  test_local_add_2bf16
+// GFX12: ds_pk_add_rtn_bf16
+short2 test_local_add_2bf16(__local short2 *addr, short2 x) {
+  return __builtin_amdgcn_ds_atomic_fadd_v2bf16(addr, x);
+}
+
+// CHECK-LABEL: test_local_add_2bf16_noret
+// CHECK: call <2 x i16> @llvm.amdgcn.ds.fadd.v2bf16(ptr addrspace(3) %{{.*}}, <2 x i16> %
+// GFX12-LABEL:  test_local_add_2bf16_noret
+// GFX12: ds_pk_add_bf16
+void test_local_add_2bf16_noret(__local short2 *addr, short2 x) {
+  __builtin_amdgcn_ds_atomic_fadd_v2bf16(addr, x);
+}
+
+// CHECK-LABEL: test_local_add_2f16
+// CHECK: call <2 x half> @llvm.amdgcn.ds.fadd.v2f16(ptr addrspace(3) %{{.*}}, <2 x half> %
+// GFX12-LABEL:  test_local_add_2f16
+// GFX12: ds_pk_add_rtn_f16
+half2 test_local_add_2f16(__local half2 *addr, half2 x) {
+  return __builtin_amdgcn_ds_atomic_fadd_v2f16(addr, x);
+}
+
+// CHECK-LABEL: test_local_add_2f16_noret
+// CHECK: call <2 x half> @llvm.amdgcn.ds.fadd.v2f16(ptr addrspace(3) %{{.*}}, <2 x half> %
+// GFX12-LABEL:  test_local_add_2f16_noret
+// GFX12: ds_pk_add_f16
+void test_local_add_2f16_noret(__local half2 *addr, half2 x) {
+  __builtin_amdgcn_ds_atomic_fadd_v2f16(addr, x);
+}
+
+// CHECK-LABEL: test_flat_add_2f16
+// CHECK: call <2 x half> @llvm.amdgcn.flat.atomic.fadd.v2f16.p0.v2f16(ptr %{{.*}}, <2 x half> %{{.*}})
+// GFX12-LABEL:  test_flat_add_2f16
+// GFX12: flat_atomic_pk_add_f16
+half2 test_flat_add_2f16(__generic half2 *addr, half2 x) {
+  return __builtin_amdgcn_flat_atomic_fadd_v2f16(addr, x);
+}
+
+// CHECK-LABEL: test_flat_add_2bf16
+// CHECK: call <2 x i16> @llvm.amdgcn.flat.atomic.fadd.v2bf16.p0(ptr %{{.*}}, <2 x i16> %{{.*}})
+// GFX12-LABEL:  test_flat_add_2bf16
+// GFX12: flat_atomic_pk_add_bf16
+short2 test_flat_add_2bf16(__generic short2 *addr, short2 x) {
+  return __builtin_amdgcn_flat_atomic_fadd_v2bf16(addr, x);
+}
+
+// CHECK-LABEL: test_global_add_half2
+// CHECK: call <2 x half> @llvm.amdgcn.global.atomic.fadd.v2f16.p1.v2f16(ptr addrspace(1) %{{.*}}, <2 x half> %{{.*}})
+// GFX12-LABEL:  test_global_add_half2
+// GFX12:  global_atomic_pk_add_f16 v2, v[0:1], v2, off th:TH_ATOMIC_RETURN
+void test_global_add_half2(__global half2 *addr, half2 x) {
+  half2 *rtn;
+  *rtn = __builtin_amdgcn_global_atomic_fadd_v2f16(addr, x);
+}
+
+// CHECK-LABEL: test_global_add_half2_noret
+// CHECK: call <2 x half> @llvm.amdgcn.global.atomic.fadd.v2f16.p1.v2f16(ptr addrspace(1) %{{.*}}, <2 x half> %{{.*}})
+// GFX12-LABEL:  test_global_add_half2_noret
+// GFX12:  global_atomic_pk_add_f16 v[0:1], v2, off
+void test_global_add_half2_noret(__global half2 *addr, half2 x) {
+  __builtin_amdgcn_global_atomic_fadd_v2f16(addr, x);
+}
+
+// CHECK-LABEL: test_global_add_2bf16
+// CHECK: call <2 x i16> @llvm.amdgcn.global.atomic.fadd.v2bf16.p1(ptr addrspace(1) %{{.*}}, <2 x i16> %{{.*}})
+// GFX12-LABEL:  test_global_add_2bf16
+// GFX12: global_atomic_pk_add_bf16 v2, v[0:1], v2, off th:TH_ATOMIC_RETURN
+void test_global_add_2bf16(__global short2 *addr, short2 x) {
+  short2 *rtn;
+  *rtn = __builtin_amdgcn_global_atomic_fadd_v2bf16(addr, x);
+}
+
+// CHECK-LABEL: test_global_add_2bf16_noret
+// CHECK: call <2 x i16> @llvm.amdgcn.global.atomic.fadd.v2bf16.p1(ptr addrspace(1) %{{.*}}, <2 x i16> %{{.*}})
+// GFX12-LABEL:  test_global_add_2bf16_noret
+// GFX12: global_atomic_pk_add_bf16 v[0:1], v2, off
+void test_global_add_2bf16_noret(__global short2 *addr, short2 x) {
+  __builtin_amdgcn_global_atomic_fadd_v2bf16(addr, x);
+}
diff --git a/clang/test/CodeGenOpenCLCXX/addrspace-of-this.clcpp b/clang/test/CodeGenOpenCLCXX/addrspace-of-this.clcpp
index 1c09743d3c8db..2f1b6c196fd51 100644
--- a/clang/test/CodeGenOpenCLCXX/addrspace-of-this.clcpp
+++ b/clang/test/CodeGenOpenCLCXX/addrspace-of-this.clcpp
@@ -111,7 +111,7 @@ __kernel void test__global() {
 // Test the address space of 'this' when invoking the operator+
 // COMMON: [[C1GEN:%[.a-z0-9]+]] = addrspacecast ptr %c1 to ptr addrspace(4)
 // COMMON: [[C2GEN:%[.a-z0-9]+]] = addrspacecast ptr %c2 to ptr addrspace(4)
-// COMMON: call spir_func void @_ZNU3AS41CplERU3AS4KS_(ptr sret(%class.C) align 4 %c3, ptr addrspace(4) {{[^,]*}} [[C1GEN]], ptr addrspace(4) noundef align 4 dereferenceable(4) [[C2GEN]])
+// COMMON: call spir_func void @_ZNU3AS41CplERU3AS4KS_(ptr dead_on_unwind writable sret(%class.C) align 4 %c3, ptr addrspace(4) {{[^,]*}} [[C1GEN]], ptr addrspace(4) noundef align 4 dereferenceable(4) [[C2GEN]])
 
 // Test the address space of 'this' when invoking the move constructor
 // COMMON: [[C4GEN:%[.a-z0-9]+]] = addrspacecast ptr %c4 to ptr addrspace(4)
@@ -127,7 +127,7 @@ __kernel void test__global() {
 
 // Tests address space of inline members
 //COMMON: @_ZNU3AS41C3getEv(ptr addrspace(4) {{[^,]*}} %this)
-//COMMON: @_ZNU3AS41CplERU3AS4KS_(ptr noalias sret(%class.C) align 4 %agg.result, ptr addrspace(4) {{[^,]*}} %this
+//COMMON: @_ZNU3AS41CplERU3AS4KS_(ptr dead_on_unwind noalias writable sret(%class.C) align 4 %agg.result, ptr addrspace(4) {{[^,]*}} %this
 #define TEST(AS)             \
   __kernel void test##AS() { \
     AS C c;                  \
diff --git a/clang/test/CodeGenSYCL/address-space-of-returns.cpp b/clang/test/CodeGenSYCL/address-space-of-returns.cpp
index 35d970af04ab5..37e1b4d194ea0 100644
--- a/clang/test/CodeGenSYCL/address-space-of-returns.cpp
+++ b/clang/test/CodeGenSYCL/address-space-of-returns.cpp
@@ -25,7 +25,7 @@ A ret_agg() {
   A a;
   return a;
 }
-// CHECK: define {{.*}}spir_func void @{{.*}}ret_agg{{.*}}(ptr addrspace(4) noalias sret(%struct{{.*}}.A) align 4 %agg.result)
+// CHECK: define {{.*}}spir_func void @{{.*}}ret_agg{{.*}}(ptr addrspace(4) dead_on_unwind noalias writable sret(%struct{{.*}}.A) align 4 %agg.result)
 
 template 
 __attribute__((sycl_kernel)) void kernel_single_task(const Func &kernelFunc) {
diff --git a/clang/test/CoverageMapping/mcdc-logical-stmt-ids-all.cpp b/clang/test/CoverageMapping/mcdc-logical-stmt-ids-all.cpp
index 5b13cc3507e96..6f47a4b901a8a 100644
--- a/clang/test/CoverageMapping/mcdc-logical-stmt-ids-all.cpp
+++ b/clang/test/CoverageMapping/mcdc-logical-stmt-ids-all.cpp
@@ -129,3 +129,80 @@ bool func_ternary_or(bool a, bool b, bool c, bool d, bool e, bool f) {
 // CHECK:  Branch,File 0, 122:26 -> 122:27 = (#6 - #7), #7 [4,0,3]
 // CHECK:  Branch,File 0, 122:31 -> 122:32 = (#4 - #5), #5 [3,0,2]
 // CHECK:  Branch,File 0, 122:36 -> 122:37 = (#2 - #3), #3 [2,0,0]
+
+bool func_if_nested_if(bool a, bool b, bool c, bool d, bool e) {
+  if (a || (b && c) || d || e)
+    return true;
+  else
+    return false;
+}
+
+// CHECK-LABEL: Decision,File 0, 134:7 -> 134:30 = M:0, C:5
+// CHECK-NEXT:  Branch,File 0, 134:7 -> 134:8 = (#0 - #6), #6 [1,0,4]
+// CHECK:  Branch,File 0, 134:13 -> 134:14 = #7, (#6 - #7) [4,5,3]
+// CHECK:  Branch,File 0, 134:18 -> 134:19 = #8, (#7 - #8) [5,0,3]
+// CHECK:  Branch,File 0, 134:24 -> 134:25 = (#4 - #5), #5 [3,0,2]
+// CHECK:  Branch,File 0, 134:29 -> 134:30 = (#2 - #3), #3 [2,0,0]
+
+bool func_ternary_nested_if(bool a, bool b, bool c, bool d, bool e) {
+  return (a || (b && c) || d || e) ? true : false;
+}
+
+// CHECK-LABEL: Decision,File 0, 148:11 -> 148:34 = M:0, C:5
+// CHECK-NEXT:  Branch,File 0, 148:11 -> 148:12 = (#0 - #6), #6 [1,0,4]
+// CHECK:  Branch,File 0, 148:17 -> 148:18 = #7, (#6 - #7) [4,5,3]
+// CHECK:  Branch,File 0, 148:22 -> 148:23 = #8, (#7 - #8) [5,0,3]
+// CHECK:  Branch,File 0, 148:28 -> 148:29 = (#4 - #5), #5 [3,0,2]
+// CHECK:  Branch,File 0, 148:33 -> 148:34 = (#2 - #3), #3 [2,0,0]
+
+bool func_if_nested_if_2(bool a, bool b, bool c, bool d, bool e) {
+  if (a || ((b && c) || d) && e)
+    return true;
+  else
+    return false;
+}
+
+// CHECK-LABEL: Decision,File 0, 159:7 -> 159:32 = M:0, C:5
+// CHECK-NEXT:  Branch,File 0, 159:7 -> 159:8 = (#0 - #2), #2 [1,0,2]
+// CHECK:  Branch,File 0, 159:14 -> 159:15 = #7, (#2 - #7) [2,5,4]
+// CHECK:  Branch,File 0, 159:19 -> 159:20 = #8, (#7 - #8) [5,3,4]
+// CHECK:  Branch,File 0, 159:25 -> 159:26 = (#5 - #6), #6 [4,3,0]
+// CHECK:  Branch,File 0, 159:31 -> 159:32 = #4, (#3 - #4) [3,0,0]
+
+bool func_ternary_nested_if_2(bool a, bool b, bool c, bool d, bool e) {
+  return (a || ((b && c) || d) && e) ? true : false;
+}
+
+// CHECK-LABEL: Decision,File 0, 173:11 -> 173:36 = M:0, C:5
+// CHECK-NEXT:  Branch,File 0, 173:11 -> 173:12 = (#0 - #2), #2 [1,0,2]
+// CHECK:  Branch,File 0, 173:18 -> 173:19 = #7, (#2 - #7) [2,5,4]
+// CHECK:  Branch,File 0, 173:23 -> 173:24 = #8, (#7 - #8) [5,3,4]
+// CHECK:  Branch,File 0, 173:29 -> 173:30 = (#5 - #6), #6 [4,3,0]
+// CHECK:  Branch,File 0, 173:35 -> 173:36 = #4, (#3 - #4) [3,0,0]
+
+bool func_if_nested_if_3(bool a, bool b, bool c, bool d, bool e, bool f) {
+  if ((a && (b || c) || (d && e)) && f)
+    return true;
+  else
+    return false;
+}
+
+// CHECK-LABEL: Decision,File 0, 184:7 -> 184:39 = M:0, C:6
+// CHECK:  Branch,File 0, 184:8 -> 184:9 = #5, (#0 - #5) [1,4,3]
+// CHECK:  Branch,File 0, 184:14 -> 184:15 = (#5 - #6), #6 [4,2,5]
+// CHECK:  Branch,File 0, 184:19 -> 184:20 = (#6 - #7), #7 [5,2,3]
+// CHECK:  Branch,File 0, 184:26 -> 184:27 = #8, (#4 - #8) [3,6,0]
+// CHECK:  Branch,File 0, 184:31 -> 184:32 = #9, (#8 - #9) [6,2,0]
+// CHECK:  Branch,File 0, 184:38 -> 184:39 = #3, (#2 - #3) [2,0,0]
+
+bool func_ternary_nested_if_3(bool a, bool b, bool c, bool d, bool e, bool f) {
+  return ((a && (b || c) || (d && e)) && f) ? true : false;
+}
+
+// CHECK-LABEL: Decision,File 0, 199:11 -> 199:43 = M:0, C:6
+// CHECK:  Branch,File 0, 199:12 -> 199:13 = #5, (#0 - #5) [1,4,3]
+// CHECK:  Branch,File 0, 199:18 -> 199:19 = (#5 - #6), #6 [4,2,5]
+// CHECK:  Branch,File 0, 199:23 -> 199:24 = (#6 - #7), #7 [5,2,3]
+// CHECK:  Branch,File 0, 199:30 -> 199:31 = #8, (#4 - #8) [3,6,0]
+// CHECK:  Branch,File 0, 199:35 -> 199:36 = #9, (#8 - #9) [6,2,0]
+// CHECK:  Branch,File 0, 199:42 -> 199:43 = #3, (#2 - #3) [2,0,0]
diff --git a/clang/test/Driver/Inputs/basic_android_tree/sysroot/usr/lib/crt_pad_segment.o b/clang/test/Driver/Inputs/basic_android_tree/sysroot/usr/lib/crt_pad_segment.o
new file mode 100644
index 0000000000000..e69de29bb2d1d
diff --git a/clang/test/Driver/Inputs/basic_cross_hurd_tree/usr/lib/gcc/x86_64-gnu/10/crtbegin.o b/clang/test/Driver/Inputs/basic_cross_hurd_tree/usr/lib/gcc/x86_64-gnu/10/crtbegin.o
new file mode 100644
index 0000000000000..e69de29bb2d1d
diff --git a/clang/test/Driver/Inputs/basic_cross_hurd_tree/usr/x86_64-gnu/bin/as b/clang/test/Driver/Inputs/basic_cross_hurd_tree/usr/x86_64-gnu/bin/as
new file mode 100755
index 0000000000000..e69de29bb2d1d
diff --git a/clang/test/Driver/Inputs/basic_cross_hurd_tree/usr/x86_64-gnu/bin/ld b/clang/test/Driver/Inputs/basic_cross_hurd_tree/usr/x86_64-gnu/bin/ld
new file mode 100755
index 0000000000000..e69de29bb2d1d
diff --git a/clang/test/Driver/Inputs/basic_cross_hurd_tree/usr/x86_64-gnu/lib/.keep b/clang/test/Driver/Inputs/basic_cross_hurd_tree/usr/x86_64-gnu/lib/.keep
new file mode 100644
index 0000000000000..e69de29bb2d1d
diff --git a/clang/test/Driver/Inputs/basic_hurd_tree/lib/x86_64-gnu/.keep b/clang/test/Driver/Inputs/basic_hurd_tree/lib/x86_64-gnu/.keep
new file mode 100644
index 0000000000000..e69de29bb2d1d
diff --git a/clang/test/Driver/Inputs/basic_hurd_tree/lib64/.keep b/clang/test/Driver/Inputs/basic_hurd_tree/lib64/.keep
new file mode 100644
index 0000000000000..e69de29bb2d1d
diff --git a/clang/test/Driver/Inputs/basic_hurd_tree/usr/include/x86_64-gnu/.keep b/clang/test/Driver/Inputs/basic_hurd_tree/usr/include/x86_64-gnu/.keep
new file mode 100644
index 0000000000000..e69de29bb2d1d
diff --git a/clang/test/Driver/Inputs/basic_hurd_tree/usr/include/x86_64-gnu/c++/10/.keep b/clang/test/Driver/Inputs/basic_hurd_tree/usr/include/x86_64-gnu/c++/10/.keep
new file mode 100644
index 0000000000000..e69de29bb2d1d
diff --git a/clang/test/Driver/Inputs/basic_hurd_tree/usr/lib/gcc/x86_64-gnu/10/crtbegin.o b/clang/test/Driver/Inputs/basic_hurd_tree/usr/lib/gcc/x86_64-gnu/10/crtbegin.o
new file mode 100644
index 0000000000000..e69de29bb2d1d
diff --git a/clang/test/Driver/Inputs/basic_hurd_tree/usr/lib/gcc/x86_64-gnu/10/crtbeginS.o b/clang/test/Driver/Inputs/basic_hurd_tree/usr/lib/gcc/x86_64-gnu/10/crtbeginS.o
new file mode 100644
index 0000000000000..e69de29bb2d1d
diff --git a/clang/test/Driver/Inputs/basic_hurd_tree/usr/lib/gcc/x86_64-gnu/10/crtbeginT.o b/clang/test/Driver/Inputs/basic_hurd_tree/usr/lib/gcc/x86_64-gnu/10/crtbeginT.o
new file mode 100644
index 0000000000000..e69de29bb2d1d
diff --git a/clang/test/Driver/Inputs/basic_hurd_tree/usr/lib/x86_64-gnu/.keep b/clang/test/Driver/Inputs/basic_hurd_tree/usr/lib/x86_64-gnu/.keep
new file mode 100644
index 0000000000000..e69de29bb2d1d
diff --git a/clang/test/Driver/Inputs/basic_hurd_tree/usr/lib64/.keep b/clang/test/Driver/Inputs/basic_hurd_tree/usr/lib64/.keep
new file mode 100644
index 0000000000000..e69de29bb2d1d
diff --git a/clang/test/Driver/aarch64-bf16.c b/clang/test/Driver/aarch64-bf16.c
index ae6b9faa95d71..c32bd5878c640 100644
--- a/clang/test/Driver/aarch64-bf16.c
+++ b/clang/test/Driver/aarch64-bf16.c
@@ -5,4 +5,4 @@
 // RUN: %clang --target=aarch64 -march=armv8.5a+bf16+nobf16 -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV85A-BF16-NO-BF16 %s
 // GENERICV85A-BF16-NO-BF16: "-target-feature" "-bf16"
 // RUN: %clang --target=aarch64 -march=armv8.5a+bf16+sve -### -c %s 2>&1 | FileCheck -check-prefixes=GENERICV85A-BF16-SVE %s
-// GENERICV85A-BF16-SVE: "-target-feature" "+bf16" "-target-feature" "+sve"
+// GENERICV85A-BF16-SVE: "-target-feature" "+bf16"{{.*}} "-target-feature" "+sve"
diff --git a/clang/test/Driver/aarch64-cortex-a510.c b/clang/test/Driver/aarch64-cortex-a510.c
index 85c40c5fade31..0d8659c46a894 100644
--- a/clang/test/Driver/aarch64-cortex-a510.c
+++ b/clang/test/Driver/aarch64-cortex-a510.c
@@ -5,4 +5,4 @@
 // CORTEX-A510-NOT: "-target-feature" "{{[+-]}}aes"
 // CORTEX-A510-SAME: {{$}}
 // RUN: %clang --target=aarch64 -mcpu=cortex-a510+crypto -### -c %s 2>&1 | FileCheck -check-prefix=CORTEX-A510-CRYPTO %s
-// CORTEX-A510-CRYPTO: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+sm4" "-target-feature" "+sha3" "-target-feature" "+sha2" "-target-feature" "+aes"
+// CORTEX-A510-CRYPTO: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+aes"{{.*}} "-target-feature" "+sha2"{{.*}} "-target-feature" "+sha3"{{.*}} "-target-feature" "+sm4"
diff --git a/clang/test/Driver/aarch64-cortex-a710.c b/clang/test/Driver/aarch64-cortex-a710.c
index d84b5b2ee7e52..b5a6ff93e0767 100644
--- a/clang/test/Driver/aarch64-cortex-a710.c
+++ b/clang/test/Driver/aarch64-cortex-a710.c
@@ -5,4 +5,4 @@
 // CORTEX-A710-NOT: "-target-feature" "{{[+-]}}aes"
 // CORTEX-A710-SAME: {{$}}
 // RUN: %clang --target=aarch64 -mcpu=cortex-a710+crypto -### -c %s 2>&1 | FileCheck -check-prefix=CORTEX-A710-CRYPTO %s
-// CORTEX-A710-CRYPTO: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+sm4" "-target-feature" "+sha3" "-target-feature" "+sha2" "-target-feature" "+aes"
+// CORTEX-A710-CRYPTO: "-cc1"{{.*}} "-triple" "aarch64"{{.*}} "-target-feature" "+aes"{{.*}} "-target-feature" "+sha2" "-target-feature" "+sha3"{{.*}} "-target-feature" "+sm4"
diff --git a/clang/test/Driver/aarch64-cortex-x2.c b/clang/test/Driver/aarch64-cortex-x2.c
index 9c6153e5beb48..2833029ba55e3 100644
--- a/clang/test/Driver/aarch64-cortex-x2.c
+++ b/clang/test/Driver/aarch64-cortex-x2.c
@@ -5,4 +5,4 @@
 // CORTEX-X2-NOT: "-target-feature" "{{[+-]}}aes"
 // CORTEX-X2-SAME: {{$}}
 // RUN: %clang --target=aarch64 -mcpu=cortex-x2+crypto -### -c %s 2>&1 | FileCheck -check-prefix=CORTEX-X2-CRYPTO %s
-// CORTEX-X2-CRYPTO: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+sm4" "-target-feature" "+sha3" "-target-feature" "+sha2" "-target-feature" "+aes"
+// CORTEX-X2-CRYPTO: "-cc1"{{.*}} "-triple" "aarch64"{{.*}} "-target-feature" "+aes"{{.*}} "-target-feature" "+sha2" "-target-feature" "+sha3"{{.*}} "-target-feature" "+sm4"
diff --git a/clang/test/Driver/aarch64-d128.c b/clang/test/Driver/aarch64-d128.c
index e2a87444d5475..5d112c0aad67f 100644
--- a/clang/test/Driver/aarch64-d128.c
+++ b/clang/test/Driver/aarch64-d128.c
@@ -3,8 +3,7 @@
 // FEAT_D128 is optional (off by default) for v9.4a and older, and can be enabled using +d128
 // RUN: %clang -### --target=aarch64-none-elf -march=armv9.4-a        %s 2>&1 | FileCheck %s --check-prefix=NOT_ENABLED
 // RUN: %clang -### --target=aarch64-none-elf -march=armv9.4-a+d128   %s 2>&1 | FileCheck %s --check-prefix=ENABLED
-// RUN: %clang -### --target=aarch64-none-elf -march=armv9.4-a+nod128 %s 2>&1 | FileCheck %s --check-prefix=DISABLED
+// RUN: %clang -### --target=aarch64-none-elf -march=armv9.4-a+nod128 %s 2>&1 | FileCheck %s --check-prefix=NOT_ENABLED
 
 // ENABLED: "-target-feature" "+d128"
 // NOT_ENABLED-NOT: "-target-feature" "+d128"
-// DISABLED: "-target-feature" "-d128"
diff --git a/clang/test/Driver/aarch64-fp16.c b/clang/test/Driver/aarch64-fp16.c
index 451cb6162ee66..4da1834cf6876 100644
--- a/clang/test/Driver/aarch64-fp16.c
+++ b/clang/test/Driver/aarch64-fp16.c
@@ -21,7 +21,7 @@
 
 // RUN: %clang --target=aarch64 -march=armv8a+fp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV8A-FP16FML %s
 // RUN: %clang --target=aarch64 -march=armv8-a+fp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV8A-FP16FML %s
-// GENERICV8A-FP16FML: "-target-feature" "+fp16fml" "-target-feature" "+fullfp16"
+// GENERICV8A-FP16FML: "-target-feature" "+fullfp16" "-target-feature" "+fp16fml"
 
 // RUN: %clang --target=aarch64 -march=armv8.2a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV82A-NO-FP16FML %s
 // RUN: %clang --target=aarch64 -march=armv8.2-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV82A-NO-FP16FML %s
@@ -30,36 +30,36 @@
 
 // RUN: %clang --target=aarch64 -march=armv8.2a+fp16 -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV82A-FP16 %s
 // RUN: %clang --target=aarch64 -march=armv8.2-a+fp16 -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV82A-FP16 %s
-// GENERICV82A-FP16: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v8.2a" "-target-feature" "+fullfp16"
+// GENERICV82A-FP16: "-cc1"{{.*}} "-triple" "aarch64"{{.*}} "-target-cpu" "generic"{{.*}} "-target-feature" "+v8.2a"{{.*}} "-target-feature" "+fullfp16"{{.*}} "-target-feature" "+neon"
 // GENERICV82A-FP16-NOT: "-target-feature" "{{[+-]}}fp16fml"
 // GENERICV82A-FP16-SAME: {{$}}
 
 // RUN: %clang --target=aarch64 -march=armv8.2-a+profile -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV82A-SPE %s
-// GENERICV82A-SPE: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v8.2a" "-target-feature" "+spe"
+// GENERICV82A-SPE: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-cpu" "generic" "-target-feature" "+v8.2a"{{.*}} "-target-feature" "+spe"{{.*}} "-target-feature" "+neon"
 
 // RUN: %clang --target=aarch64 -march=armv8.2a+fp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV82A-FP16FML %s
 // RUN: %clang --target=aarch64 -march=armv8.2-a+fp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV82A-FP16FML %s
-// GENERICV82A-FP16FML: "-target-feature" "+fp16fml" "-target-feature" "+fullfp16"
+// GENERICV82A-FP16FML: "-target-feature" "+fullfp16" "-target-feature" "+fp16fml"
 
 // RUN: %clang --target=aarch64 -march=armv8.2a+fp16+nofp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV82A-FP16-NO-FP16FML %s
 // RUN: %clang --target=aarch64 -march=armv8.2-a+fp16+nofp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV82A-FP16-NO-FP16FML %s
-// GENERICV82A-FP16-NO-FP16FML: "-target-feature" "+fullfp16" "-target-feature" "-fp16fml"
+// GENERICV82A-FP16-NO-FP16FML: "-target-feature" "+fullfp16"
 
 // RUN: %clang --target=aarch64 -march=armv8.2a+nofp16fml+fp16 -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV82A-NO-FP16FML-FP16 %s
 // RUN: %clang --target=aarch64 -march=armv8.2-a+nofp16fml+fp16 -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV82A-NO-FP16FML-FP16 %s
-// GENERICV82A-NO-FP16FML-FP16: "-target-feature" "-fp16fml" "-target-feature" "+fullfp16"
+// GENERICV82A-NO-FP16FML-FP16: "-target-feature" "+fullfp16"
 
 // RUN: %clang --target=aarch64 -march=armv8.2a+fp16fml+nofp16 -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV82A-FP16FML-NO-FP16 %s
 // RUN: %clang --target=aarch64 -march=armv8.2-a+fp16fml+nofp16 -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV82A-FP16FML-NO-FP16 %s
-// GENERICV82A-FP16FML-NO-FP16: "-target-feature" "-fullfp16" "-target-feature" "-fp16fml"
+// GENERICV82A-FP16FML-NO-FP16: "-target-feature" "-fullfp16"{{.*}} "-target-feature" "-fp16fml"
 
 // RUN: %clang --target=aarch64 -march=armv8.2a+nofp16+fp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV82A-NO-FP16-FP16FML %s
 // RUN: %clang --target=aarch64 -march=armv8.2-a+nofp16+fp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV82A-NO-FP16-FP16FML %s
-// GENERICV82A-NO-FP16-FP16FML: "-target-feature" "+fp16fml" "-target-feature" "+fullfp16"
+// GENERICV82A-NO-FP16-FP16FML: "-target-feature" "+fullfp16" "-target-feature" "+fp16fml"
 
 // RUN: %clang --target=aarch64 -march=armv8.2a+fp16+profile -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV82A-FP16-SPE %s
 // RUN: %clang --target=aarch64 -march=armv8.2-a+fp16+profile -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV82A-FP16-SPE %s
-// GENERICV82A-FP16-SPE: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v8.2a" "-target-feature" "+fullfp16" "-target-feature" "+spe"
+// GENERICV82A-FP16-SPE: "-cc1"{{.*}} "-triple" "aarch64{{.*}}"{{.*}} "-target-cpu" "generic"{{.*}} "-target-feature" "+v8.2a"{{.*}}  "-target-feature" "+fullfp16"{{.*}} "-target-feature" "+spe"{{.*}} "-target-feature" "+neon"
 // GENERICV82A-FP16-SPE-NOT:  "-target-feature" "{{[+-]}}fp16fml"
 // GENERICV82A-FP16-SPE-SAME: {{$}}
 
@@ -70,19 +70,21 @@
 
 // RUN: %clang --target=aarch64 -march=armv8.3a+fp16 -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV83A-FP16 %s
 // RUN: %clang --target=aarch64 -march=armv8.3-a+fp16 -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV83A-FP16 %s
-// GENERICV83A-FP16: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v8.3a" "-target-feature" "+fullfp16"
+// GENERICV83A-FP16: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-cpu" "generic"{{.*}} "-target-feature" "+v8.3a"{{.*}} "-target-feature" "+fullfp16"{{.*}} "-target-feature" "+neon"
 
 // RUN: %clang --target=aarch64 -march=armv8.3a+fp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV83A-FP16FML %s
 // RUN: %clang --target=aarch64 -march=armv8.3-a+fp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV83A-FP16FML %s
-// GENERICV83A-FP16FML: "-target-feature" "+fp16fml" "-target-feature" "+fullfp16"
+// GENERICV83A-FP16FML: "-target-feature" "+fullfp16" "-target-feature" "+fp16fml"
 
 // RUN: %clang --target=aarch64 -march=armv8.3a+fp16+nofp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV83A-FP16-NO-FP16FML %s
 // RUN: %clang --target=aarch64 -march=armv8.3-a+fp16+nofp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV83A-FP16-NO-FP16FML %s
-// GENERICV83A-FP16-NO-FP16FML: "-target-feature" "+fullfp16" "-target-feature" "-fp16fml"
+// GENERICV83A-FP16-NO-FP16FML: "-target-feature" "+fullfp16"
+// GENERICV83A-FP16-NO-FP16FML-NOT: fp16fml
 
 // RUN: %clang --target=aarch64 -march=armv8.3a+nofp16fml+fp16 -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV83A-NO-FP16FML-FP16 %s
 // RUN: %clang --target=aarch64 -march=armv8.3-a+nofp16fml+fp16 -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV83A-NO-FP16FML-FP16 %s
-// GENERICV83A-NO-FP16FML-FP16: "-target-feature" "-fp16fml" "-target-feature" "+fullfp16"
+// GENERICV83A-NO-FP16FML-FP16: "-target-feature" "+fullfp16"
+// GENERICV83A-NO-FP16FML-FP16-NOT: fp16fml
 
 // RUN: %clang --target=aarch64 -march=armv8.3a+fp16fml+nofp16 -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV83A-FP16FML-NO-FP16 %s
 // RUN: %clang --target=aarch64 -march=armv8.3-a+fp16fml+nofp16 -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV83A-FP16FML-NO-FP16 %s
@@ -90,7 +92,7 @@
 
 // RUN: %clang --target=aarch64 -march=armv8.3a+nofp16+fp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV83A-NO-FP16-FP16FML %s
 // RUN: %clang --target=aarch64 -march=armv8.3-a+nofp16+fp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV83A-NO-FP16-FP16FML %s
-// GENERICV83A-NO-FP16-FP16FML: "-target-feature" "+fp16fml" "-target-feature" "+fullfp16"
+// GENERICV83A-NO-FP16-FP16FML: "-target-feature" "+fullfp16" "-target-feature" "+fp16fml"
 
 // RUN: %clang --target=aarch64 -march=armv8.4a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV84A-NO-FP16FML %s
 // RUN: %clang --target=aarch64 -march=armv8.4-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV84A-NO-FP16FML %s
@@ -103,7 +105,7 @@
 
 // RUN: %clang --target=aarch64 -march=armv8.4a+fp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV84A-FP16FML %s
 // RUN: %clang --target=aarch64 -march=armv8.4-a+fp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV84A-FP16FML %s
-// GENERICV84A-FP16FML: "-target-feature" "+fp16fml" "-target-feature" "+fullfp16"
+// GENERICV84A-FP16FML: "-target-feature" "+fullfp16" "-target-feature" "+fp16fml"
 
 // RUN: %clang --target=aarch64 -march=armv8.4a+fp16+nofp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV84A-FP16-NO-FP16FML %s
 // RUN: %clang --target=aarch64 -march=armv8.4-a+fp16+nofp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV84A-FP16-NO-FP16FML %s
@@ -119,7 +121,7 @@
 
 // RUN: %clang --target=aarch64 -march=armv8.4a+nofp16+fp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV84A-NO-FP16-FP16FML %s
 // RUN: %clang --target=aarch64 -march=armv8.4-a+nofp16+fp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV84A-NO-FP16-FP16FML %s
-// GENERICV84A-NO-FP16-FP16FML: "-target-feature" "+fp16fml" "-target-feature" "+fullfp16"
+// GENERICV84A-NO-FP16-FP16FML: "-target-feature" "+fullfp16" "-target-feature" "+fp16fml"
 
 // RUN: %clang --target=aarch64 -march=armv8.5a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV85A-NO-FP16FML %s
 // RUN: %clang --target=aarch64 -march=armv8.5-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV85A-NO-FP16FML %s
@@ -132,7 +134,7 @@
 
 // RUN: %clang --target=aarch64 -march=armv8.5a+fp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV85A-FP16FML %s
 // RUN: %clang --target=aarch64 -march=armv8.5-a+fp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV85A-FP16FML %s
-// GENERICV85A-FP16FML: "-target-feature" "+fp16fml" "-target-feature" "+fullfp16"
+// GENERICV85A-FP16FML: "-target-feature" "+fullfp16" "-target-feature" "+fp16fml"
 
 // RUN: %clang --target=aarch64 -march=armv8.5a+fp16+nofp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV85A-FP16-NO-FP16FML %s
 // RUN: %clang --target=aarch64 -march=armv8.5-a+fp16+nofp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV85A-FP16-NO-FP16FML %s
@@ -148,7 +150,7 @@
 
 // RUN: %clang --target=aarch64 -march=armv8.5a+nofp16+fp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV85A-NO-FP16-FP16FML %s
 // RUN: %clang --target=aarch64 -march=armv8.5-a+nofp16+fp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV85A-NO-FP16-FP16FML %s
-// GENERICV85A-NO-FP16-FP16FML: "-target-feature" "+fp16fml" "-target-feature" "+fullfp16"
+// GENERICV85A-NO-FP16-FP16FML: "-target-feature" "+fullfp16" "-target-feature" "+fp16fml"
 
 // RUN: %clang --target=aarch64 -march=armv8.6a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV86A-NO-FP16FML %s
 // RUN: %clang --target=aarch64 -march=armv8.6-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV86A-NO-FP16FML %s
@@ -161,7 +163,7 @@
 
 // RUN: %clang --target=aarch64 -march=armv8.6a+fp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV86A-FP16FML %s
 // RUN: %clang --target=aarch64 -march=armv8.6-a+fp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV86A-FP16FML %s
-// GENERICV86A-FP16FML: "-target-feature" "+fp16fml" "-target-feature" "+fullfp16"
+// GENERICV86A-FP16FML: "-target-feature" "+fullfp16" "-target-feature" "+fp16fml"
 
 // RUN: %clang --target=aarch64 -march=armv8.6a+fp16+nofp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV86A-FP16-NO-FP16FML %s
 // RUN: %clang --target=aarch64 -march=armv8.6-a+fp16+nofp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV86A-FP16-NO-FP16FML %s
@@ -177,7 +179,7 @@
 
 // RUN: %clang --target=aarch64 -march=armv8.6a+nofp16+fp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV86A-NO-FP16-FP16FML %s
 // RUN: %clang --target=aarch64 -march=armv8.6-a+nofp16+fp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV86A-NO-FP16-FP16FML %s
-// GENERICV86A-NO-FP16-FP16FML: "-target-feature" "+fp16fml" "-target-feature" "+fullfp16"
+// GENERICV86A-NO-FP16-FP16FML: "-target-feature" "+fullfp16" "-target-feature" "+fp16fml"
 
 // RUN: %clang --target=aarch64 -march=armv8.7a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV87A-NO-FP16FML %s
 // RUN: %clang --target=aarch64 -march=armv8.7-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV87A-NO-FP16FML %s
@@ -190,7 +192,7 @@
 
 // RUN: %clang --target=aarch64 -march=armv8.7a+fp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV87A-FP16FML %s
 // RUN: %clang --target=aarch64 -march=armv8.7-a+fp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV87A-FP16FML %s
-// GENERICV87A-FP16FML: "-target-feature" "+fp16fml" "-target-feature" "+fullfp16"
+// GENERICV87A-FP16FML: "-target-feature" "+fullfp16" "-target-feature" "+fp16fml"
 
 // RUN: %clang --target=aarch64 -march=armv8.7a+fp16+nofp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV87A-FP16-NO-FP16FML %s
 // RUN: %clang --target=aarch64 -march=armv8.7-a+fp16+nofp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV87A-FP16-NO-FP16FML %s
@@ -206,7 +208,7 @@
 
 // RUN: %clang --target=aarch64 -march=armv8.7a+nofp16+fp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV87A-NO-FP16-FP16FML %s
 // RUN: %clang --target=aarch64 -march=armv8.7-a+nofp16+fp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV87A-NO-FP16-FP16FML %s
-// GENERICV87A-NO-FP16-FP16FML: "-target-feature" "+fp16fml" "-target-feature" "+fullfp16"
+// GENERICV87A-NO-FP16-FP16FML: "-target-feature" "+fullfp16" "-target-feature" "+fp16fml"
 
 // RUN: %clang --target=aarch64 -march=armv8.8a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV88A-NO-FP16FML %s
 // RUN: %clang --target=aarch64 -march=armv8.8-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV88A-NO-FP16FML %s
@@ -219,7 +221,7 @@
 
 // RUN: %clang --target=aarch64 -march=armv8.8a+fp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV88A-FP16FML %s
 // RUN: %clang --target=aarch64 -march=armv8.8-a+fp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV88A-FP16FML %s
-// GENERICV88A-FP16FML: "-target-feature" "+fp16fml" "-target-feature" "+fullfp16"
+// GENERICV88A-FP16FML: "-target-feature" "+fullfp16" "-target-feature" "+fp16fml"
 
 // RUN: %clang --target=aarch64 -march=armv8.8a+fp16+nofp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV88A-FP16-NO-FP16FML %s
 // RUN: %clang --target=aarch64 -march=armv8.8-a+fp16+nofp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV88A-FP16-NO-FP16FML %s
@@ -235,4 +237,4 @@
 
 // RUN: %clang --target=aarch64 -march=armv8.8a+nofp16+fp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV88A-NO-FP16-FP16FML %s
 // RUN: %clang --target=aarch64 -march=armv8.8-a+nofp16+fp16fml -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV88A-NO-FP16-FP16FML %s
-// GENERICV88A-NO-FP16-FP16FML: "-target-feature" "+fp16fml" "-target-feature" "+fullfp16"
+// GENERICV88A-NO-FP16-FP16FML: "-target-feature" "+fullfp16" "-target-feature" "+fp16fml"
diff --git a/clang/test/Driver/aarch64-implied-sme-features.c b/clang/test/Driver/aarch64-implied-sme-features.c
index 2277aceb881d7..67836f42f2c02 100644
--- a/clang/test/Driver/aarch64-implied-sme-features.c
+++ b/clang/test/Driver/aarch64-implied-sme-features.c
@@ -1,44 +1,46 @@
 // RUN: %clang -target aarch64-linux-gnu -march=armv8-a+sme %s -### 2>&1 | FileCheck %s --check-prefix=SME-IMPLY
-// SME-IMPLY: "-target-feature" "+sme" "-target-feature" "+bf16"
+// SME-IMPLY: "-target-feature" "+bf16"{{.*}} "-target-feature" "+sme"
 
 // RUN: %clang -target aarch64-linux-gnu -march=armv8-a+nosme %s -### 2>&1 | FileCheck %s --check-prefix=NOSME
-// NOSME: "-target-feature" "-sme"
+// NOSME-NOT: "-target-feature" "{{\+|-}}sme"
 
 // RUN: %clang -target aarch64-linux-gnu -march=armv8-a+sme+nosme %s -### 2>&1 | FileCheck %s --check-prefix=SME-REVERT
 // SME-REVERT-NOT: "-target-feature" "+sme"
-// SME-REVERT: "-target-feature" "+bf16" "-target-feature" "-sme" "-target-feature" "-sme-f64f64" "-target-feature" "-sme-i16i64"
+// SME-REVERT: "-target-feature" "+bf16"{{.*}} "-target-feature" "-sme"
 
 // RUN: %clang -target aarch64-linux-gnu -march=armv8-a+sme+nobf16 %s -### 2>&1 | FileCheck %s --check-prefix=SME-CONFLICT
 // SME-CONFLICT-NOT: "-target-feature" "+sme"
 // SME-CONFLICT-NOT: "-target-feature" "+bf16"
-// SME-CONFLICT: "-target-feature" "-bf16" "-target-feature" "-sme" "-target-feature" "-sme-f64f64" "-target-feature" "-sme-i16i64"
+// SME-CONFLICT: "-target-feature" "-bf16"{{.*}} "-target-feature" "-sme"
 
 // RUN: %clang -target aarch64-linux-gnu -march=armv8-a+sme-i16i64 %s -### 2>&1 | FileCheck %s --check-prefix=SME-I16I64
-// SME-I16I64: "-target-feature" "+sme-i16i64" "-target-feature" "+sme" "-target-feature" "+bf16"
+// SME-I16I64: "-target-feature" "+bf16"{{.*}} "-target-feature" "+sme-i16i64" "-target-feature" "+sme"
 
 // RUN: %clang -target aarch64-linux-gnu -march=armv8-a+nosme-i16i64 %s -### 2>&1 | FileCheck %s --check-prefix=NOSME-I16I64
 // NOSME-I16I64-NOT: "-target-feature" "+sme-i16i64"
 // NOSME-I16I64-NOT: "-target-feature" "+sme"
 // NOSME-I16I64-NOT: "-target-feature" "+bf16"
-// NOSME-I16I64:     "-target-feature" "-sme-i16i64"
+// NOSME-I16I64-NOT: sme-i16i64"
 
 // RUN: %clang -target aarch64-linux-gnu -march=armv8-a+sme-i16i64+nosme-i16i64 %s -### 2>&1 | FileCheck %s --check-prefix=SME-I16I64-REVERT
-// SME-I16I64-REVERT: "-target-feature" "+sme" "-target-feature" "+bf16" "-target-feature" "-sme-i16i64"
+// SME-I16I64-REVERT: "-target-feature" "+bf16"{{.*}} "-target-feature" "-sme-i16i64" "-target-feature" "+sme"
 
 // RUN: %clang -target aarch64-linux-gnu -march=armv8-a+nosme-f64f64 %s -### 2>&1 | FileCheck %s --check-prefix=NOSME-F64F64
 // NOSME-F64F64-NOT: "-target-feature" "+sme-f64f64"
 // NOSME-F64F64-NOT: "-target-feature" "+sme"
 // NOSME-F64F64-NOT: "-target-feature" "+bf16"
-// NOSME-F64F64:     "-target-feature" "-sme-f64f64"
+// NOSME-F64F64-NOT: sme-f64f64"
 
 // RUN: %clang -target aarch64-linux-gnu -march=armv8-a+sme-f64f64+nosme-f64f64 %s -### 2>&1 | FileCheck %s --check-prefix=SME-F64F64-REVERT
-// SME-F64F64-REVERT: "-target-feature" "+sme" "-target-feature" "+bf16" "-target-feature" "-sme-f64f64"
+// SME-F64F64-REVERT: "-target-feature" "+bf16"{{.*}} "-target-feature" "-sme-f64f64" "-target-feature" "+sme"
 
 // RUN: %clang -target aarch64-linux-gnu -march=armv8-a+sme-f64f64+nosme-i16i64 %s -### 2>&1 | FileCheck %s --check-prefix=SME-SUBFEATURE-MIX
-// SME-SUBFEATURE-MIX: "-target-feature" "+sme-f64f64" "-target-feature" "+sme" "-target-feature" "+bf16" "-target-feature" "-sme-i16i64"
+// SME-SUBFEATURE-MIX-NOT: "+sme-i16i64"
+// SME-SUBFEATURE-MIX: "-target-feature" "+bf16"{{.*}} "-target-feature" "+sme-f64f64" "-target-feature" "+sme"
+// SME-SUBFEATURE-MIX-NOT: "+sme-i16i64"
 
 // RUN: %clang -target aarch64-linux-gnu -march=armv8-a+sme-i16i64+nosme %s -### 2>&1 | FileCheck %s --check-prefix=SME-SUBFEATURE-CONFLICT1
-// SME-SUBFEATURE-CONFLICT1: "-target-feature" "+bf16" "-target-feature" "-sme" "-target-feature" "-sme-f64f64" "-target-feature" "-sme-i16i64"
+// SME-SUBFEATURE-CONFLICT1: "-target-feature" "+bf16"{{.*}} "-target-feature" "-sme-i16i64" "-target-feature" "-sme"
 
 // RUN: %clang -target aarch64-linux-gnu -march=armv8-a+sme-f64f64+nobf16 %s -### 2>&1 | FileCheck %s --check-prefix=SME-SUBFEATURE-CONFLICT2
 // SME-SUBFEATURE-CONFLICT2-NOT: "-target-feature" "+bf16"
@@ -46,4 +48,4 @@
 // SME-SUBFEATURE-CONFLICT2-NOT: "-target-feature" "+sme-f64f64"
 
 // RUN: %clang -target aarch64-linux-gnu -march=armv8-a+nosme+sme-i16i64 %s -### 2>&1 | FileCheck %s --check-prefix=SME-SUBFEATURE-CONFLICT-REV
-// SME-SUBFEATURE-CONFLICT-REV: "-target-feature" "-sme-f64f64" "-target-feature" "+sme-i16i64" "-target-feature" "+sme" "-target-feature" "+bf16"
+// SME-SUBFEATURE-CONFLICT-REV: "-target-feature" "+bf16"{{.*}} "-target-feature" "+sme-i16i64" "-target-feature" "+sme"
diff --git a/clang/test/Driver/aarch64-implied-sve-features.c b/clang/test/Driver/aarch64-implied-sve-features.c
index 02d9cb5b59426..9227cd4981c2e 100644
--- a/clang/test/Driver/aarch64-implied-sve-features.c
+++ b/clang/test/Driver/aarch64-implied-sve-features.c
@@ -2,62 +2,63 @@
 // SVE-ONLY: "-target-feature" "+sve"
 
 // RUN: %clang --target=aarch64-linux-gnu -march=armv8-a+nosve %s -### 2>&1 | FileCheck %s --check-prefix=NOSVE
-// NOSVE: "-target-feature" "-sve"
+// NOSVE-NOT: sve"
 
 // RUN: %clang --target=aarch64-linux-gnu -march=armv8-a+sve+nosve %s -### 2>&1 | FileCheck %s --check-prefix=SVE-REVERT
 // SVE-REVERT-NOT: "-target-feature" "+sve"
 // SVE-REVERT: "-target-feature" "-sve"
 
 // RUN: %clang --target=aarch64-linux-gnu -march=armv8-a+sve2 %s -### 2>&1 | FileCheck %s --check-prefix=SVE2-IMPLY
-// SVE2-IMPLY: "-target-feature" "+sve2" "-target-feature" "+sve"
+// SVE2-IMPLY: "-target-feature" "+sve" "-target-feature" "+sve2"
 
 // RUN: %clang --target=aarch64-linux-gnu -march=armv8-a+sve2+nosve2 %s -### 2>&1 | FileCheck %s --check-prefix=SVE2-REVERT
-// SVE2-REVERT: "-target-feature" "+sve" "-target-feature" "-sve2"
+// SVE2-REVERT: "-target-feature" "+sve"{{.*}} "-target-feature" "-sve2"
 
 // RUN: %clang --target=aarch64-linux-gnu -march=armv8-a+sve2+nosve %s -### 2>&1 | FileCheck %s --check-prefix=SVE2-CONFLICT
-// SVE2-CONFLICT: "-target-feature" "-sve" "-target-feature" "-sve2" "-target-feature" "-sve2-bitperm" "-target-feature" "-sve2-sha3" "-target-feature" "-sve2-aes" "-target-feature" "-sve2-sm4"
+// SVE2-CONFLICT: "-target-feature" "-sve" "-target-feature" "-sve2"
 
 // RUN: %clang --target=aarch64-linux-gnu -march=armv8-a+nosve+sve2 %s -### 2>&1 | FileCheck %s --check-prefix=SVE2-CONFLICT-REV
-// SVE2-CONFLICT-REV: "-target-feature" "+sve2" "-target-feature" "+sve"
+// SVE2-CONFLICT-REV: "-target-feature" "+sve" "-target-feature" "+sve2"
 
 // RUN: %clang --target=aarch64-linux-gnu -march=armv8-a+sve+sve2 %s -### 2>&1 | FileCheck %s --check-prefix=SVE-SVE2
-// SVE-SVE2: "-target-feature" "+sve2" "-target-feature" "+sve"
+// SVE-SVE2: "-target-feature" "+sve" "-target-feature" "+sve2"
 
 // RUN: %clang --target=aarch64-linux-gnu -march=armv8-a+sve2-bitperm %s -### 2>&1 | FileCheck %s --check-prefix=SVE2-BITPERM
-// SVE2-BITPERM: "-target-feature" "+sve2-bitperm" "-target-feature" "+sve" "-target-feature" "+sve2"
+// SVE2-BITPERM: "-target-feature" "+sve" "-target-feature" "+sve2-bitperm" "-target-feature" "+sve2"
 
 // RUN: %clang --target=aarch64-linux-gnu -march=armv8-a+nosve2-bitperm %s -### 2>&1 | FileCheck %s --check-prefix=NOSVE2-BITPERM
 // NOSVE2-BITPERM-NOT: "-target-feature" "+sve2-bitperm"
 // NOSVE2-BITPERM-NOT: "-target-feature" "+sve2"
 // NOSVE2-BITPERM-NOT: "-target-feature" "+sve"
-// NOSVE2-BITPERM: "-target-feature" "-sve2-bitperm"
+// NOSVE2-BITPERM-NOT: sve2-bitperm"
 
 // RUN: %clang --target=aarch64-linux-gnu -march=armv8-a+sve2-bitperm+nosve2-bitperm %s -### 2>&1 | FileCheck %s --check-prefix=SVE2-BITPERM-REVERT
-// SVE2-BITPERM-REVERT: "-target-feature" "+sve" "-target-feature" "+sve2" "-target-feature" "-sve2-bitperm"
+// SVE2-BITPERM-REVERT: "-target-feature" "+sve" "-target-feature" "-sve2-bitperm" "-target-feature" "+sve2"
 
 // RUN: %clang --target=aarch64-linux-gnu -march=armv8-a+sve2-aes+nosve2-aes %s -### 2>&1 | FileCheck %s --check-prefix=SVE2-AES-REVERT
-// SVE2-AES-REVERT: "-target-feature" "+sve" "-target-feature" "+sve2" "-target-feature" "-sve2-aes"
+// SVE2-AES-REVERT: "-target-feature" "+sve" "-target-feature" "-sve2-aes" "-target-feature" "+sve2"
 
 // RUN: %clang --target=aarch64-linux-gnu -march=armv8-a+sve2-sha3+nosve2-sha3 %s -### 2>&1 | FileCheck %s --check-prefix=SVE2-SHA3-REVERT
-// SVE2-SHA3-REVERT: "-target-feature" "+sve" "-target-feature" "+sve2" "-target-feature" "-sve2-sha3"
+// SVE2-SHA3-REVERT: "-target-feature" "+sve" "-target-feature" "-sve2-sha3" "-target-feature" "+sve2"
 
 // RUN: %clang --target=aarch64-linux-gnu -march=armv8-a+sve2-sm4+nosve2-sm4 %s -### 2>&1 | FileCheck %s --check-prefix=SVE2-SM4-REVERT
-// SVE2-SM4-REVERT: "-target-feature" "+sve" "-target-feature" "+sve2" "-target-feature" "-sve2-sm4"
+// SVE2-SM4-REVERT: "-target-feature" "+sve" "-target-feature" "-sve2-sm4" "-target-feature" "+sve2"
 
 // RUN: %clang --target=aarch64-linux-gnu -march=armv8-a+sve2-sha3 %s -### 2>&1 | FileCheck %s --check-prefix=SVE2-SHA3
-// SVE2-SHA3: "-target-feature" "+sve2-sha3" "-target-feature" "+sve" "-target-feature" "+sve2"
+// SVE2-SHA3: "-target-feature" "+sve" "-target-feature" "+sve2-sha3" "-target-feature" "+sve2"
 
 // RUN: %clang --target=aarch64-linux-gnu -march=armv8-a+sve2-aes %s -### 2>&1 | FileCheck %s --check-prefix=SVE2-AES
-// SVE2-AES: "-target-feature" "+sve2-aes" "-target-feature" "+sve" "-target-feature" "+sve2"
+// SVE2-AES: "-target-feature" "+sve" "-target-feature" "+sve2-aes" "-target-feature" "+sve2"
 
 // RUN: %clang --target=aarch64-linux-gnu -march=armv8-a+sve2-sm4 %s -### 2>&1 | FileCheck %s --check-prefix=SVE2-SM4
-// SVE2-SM4: "-target-feature" "+sve2-sm4" "-target-feature" "+sve" "-target-feature" "+sve2"
+// SVE2-SM4: "-target-feature" "+sve" "-target-feature" "+sve2-sm4" "-target-feature" "+sve2"
 
 // RUN: %clang --target=aarch64-linux-gnu -march=armv8-a+sve2-bitperm+nosve2-aes %s -### 2>&1 | FileCheck %s --check-prefix=SVE2-SUBFEATURE-MIX
-// SVE2-SUBFEATURE-MIX: "-target-feature" "+sve2-bitperm" "-target-feature" "+sve" "-target-feature" "+sve2" "-target-feature" "-sve2-aes"
+// SVE2-SUBFEATURE-MIX: "-target-feature" "+sve" "-target-feature" "+sve2-bitperm" "-target-feature" "+sve2"
+// SVE2-SUBFEATURE-NOT: sve2-aes
 
 // RUN: %clang --target=aarch64-linux-gnu -march=armv8-a+sve2-sm4+nosve2 %s -### 2>&1 | FileCheck %s --check-prefix=SVE2-SUBFEATURE-CONFLICT
-// SVE2-SUBFEATURE-CONFLICT: "-target-feature" "+sve" "-target-feature" "-sve2" "-target-feature" "-sve2-bitperm" "-target-feature" "-sve2-sha3" "-target-feature" "-sve2-aes" "-target-feature" "-sve2-sm4"
+// SVE2-SUBFEATURE-CONFLICT: "-target-feature" "+sve" "-target-feature" "-sve2-sm4" "-target-feature" "-sve2"
 
 // RUN: %clang --target=aarch64-linux-gnu -march=armv8-a+sve2-aes+nosve %s -### 2>&1 | FileCheck %s --check-prefix=SVE-SUBFEATURE-CONFLICT
 // SVE-SUBFEATURE-CONFLICT-NOT: "-target-feature" "+sve2-aes"
@@ -65,7 +66,7 @@
 // SVE-SUBFEATURE-CONFLICT-NOT: "-target-feature" "+sve"
 
 // RUN: %clang --target=aarch64-linux-gnu -march=armv8-a+nosve+sve2-aes %s -### 2>&1 | FileCheck %s --check-prefix=SVE-SUBFEATURE-CONFLICT-REV
-// SVE-SUBFEATURE-CONFLICT-REV: "-target-feature" "-sve2-bitperm" "-target-feature" "-sve2-sha3" "-target-feature" "-sve2-sm4" "-target-feature" "+sve2-aes" "-target-feature" "+sve" "-target-feature" "+sve2"
+// SVE-SUBFEATURE-CONFLICT-REV: "-target-feature" "+sve" "-target-feature" "+sve2-aes" "-target-feature" "+sve2"
 
 // RUN: %clang --target=aarch64-linux-gnu -mcpu=neoverse-n2+nosve2 %s -### 2>&1 | FileCheck %s --check-prefix=SVE-MCPU-FEATURES
 // SVE-MCPU-FEATURES-NOT: "-target-feature" "+sve2-bitperm"
@@ -74,5 +75,5 @@
 
 // RUN: %clang --target=aarch64-linux-gnu -mcpu=neoverse-n2+nosve+sve2 %s -### 2>&1 | FileCheck %s --check-prefix=SVE-MCPU-FEATURES-CONFLICT
 // SVE-MCPU-FEATURES-CONFLICT-NOT: "-target-feature" "+sve2-bitperm"
-// SVE-MCPU-FEATURES-CONFLICT: "-target-feature" "+sve2"
 // SVE-MCPU-FEATURES-CONFLICT: "-target-feature" "+sve"
+// SVE-MCPU-FEATURES-CONFLICT-SAME: "-target-feature" "+sve2"
diff --git a/clang/test/Driver/aarch64-ite.c b/clang/test/Driver/aarch64-ite.c
index cc1f700ce0cbf..dde2535872dfd 100644
--- a/clang/test/Driver/aarch64-ite.c
+++ b/clang/test/Driver/aarch64-ite.c
@@ -3,10 +3,10 @@
 // FEAT_ITE is optional (off by default) for v8.9a/9.4a and older, and can be enabled using +ite
 // RUN: %clang -### --target=aarch64-none-elf -march=armv8.8-a       %s 2>&1 | FileCheck %s --check-prefix=NOT_ENABLED
 // RUN: %clang -### --target=aarch64-none-elf -march=armv8.8-a+ite   %s 2>&1 | FileCheck %s --check-prefix=ENABLED
-// RUN: %clang -### --target=aarch64-none-elf -march=armv8.8-a+noite %s 2>&1 | FileCheck %s --check-prefix=DISABLED
+// RUN: %clang -### --target=aarch64-none-elf -march=armv8.8-a+noite %s 2>&1 | FileCheck %s --check-prefix=NOT_ENABLED
 // RUN: %clang -### --target=aarch64-none-elf -march=armv9.3-a       %s 2>&1 | FileCheck %s --check-prefix=NOT_ENABLED
 // RUN: %clang -### --target=aarch64-none-elf -march=armv9.3-a+ite   %s 2>&1 | FileCheck %s --check-prefix=ENABLED
-// RUN: %clang -### --target=aarch64-none-elf -march=armv9.3-a+noite %s 2>&1 | FileCheck %s --check-prefix=DISABLED
+// RUN: %clang -### --target=aarch64-none-elf -march=armv9.3-a+noite %s 2>&1 | FileCheck %s --check-prefix=NOT_ENABLED
 
 // FEAT_ITE is invalid before v8
 // RUN: not %clang -### --target=arm-none-none-eabi -march=armv7-a+ite     %s 2>&1 | FileCheck %s --check-prefix=INVALID
@@ -14,4 +14,3 @@
 // INVALID: error: unsupported argument 'armv7-a+ite' to option '-march='
 // ENABLED: "-target-feature" "+ite"
 // NOT_ENABLED-NOT: "-target-feature" "+ite"
-// DISABLED: "-target-feature" "-ite"
diff --git a/clang/test/Driver/aarch64-lrcpc3.c b/clang/test/Driver/aarch64-lrcpc3.c
index eeec0b137e028..755d6a823ce5f 100644
--- a/clang/test/Driver/aarch64-lrcpc3.c
+++ b/clang/test/Driver/aarch64-lrcpc3.c
@@ -3,18 +3,18 @@
 // FEAT_RCPC3 is optional for v8.2a onwards, and can be enabled with +rcpc3
 // RUN: %clang -### --target=aarch64-none-elf -march=armv8.9-a         %s 2>&1 | FileCheck %s --check-prefix=NOT_ENABLED
 // RUN: %clang -### --target=aarch64-none-elf -march=armv8.9-a+rcpc3   %s 2>&1 | FileCheck %s --check-prefix=ENABLED
-// RUN: %clang -### --target=aarch64-none-elf -march=armv8.9-a+norcpc3 %s 2>&1 | FileCheck %s --check-prefix=DISABLED
+// RUN: %clang -### --target=aarch64-none-elf -march=armv8.9-a+norcpc3 %s 2>&1 | FileCheck %s --check-prefix=NOT_ENABLED
 // RUN: %clang -### --target=aarch64-none-elf -march=armv9.4-a         %s 2>&1 | FileCheck %s --check-prefix=NOT_ENABLED
 // RUN: %clang -### --target=aarch64-none-elf -march=armv9.4-a+rcpc3   %s 2>&1 | FileCheck %s --check-prefix=ENABLED
-// RUN: %clang -### --target=aarch64-none-elf -march=armv9.4-a+norcpc3 %s 2>&1 | FileCheck %s --check-prefix=DISABLED
+// RUN: %clang -### --target=aarch64-none-elf -march=armv9.4-a+norcpc3 %s 2>&1 | FileCheck %s --check-prefix=NOT_ENABLED
 
 // FEAT_RCPC3 is optional (off by default) for v8.8a/9.3a and older, and can be enabled using +rcpc3
 // RUN: %clang -### --target=aarch64-none-elf -march=armv8.2-a         %s 2>&1 | FileCheck %s --check-prefix=NOT_ENABLED
 // RUN: %clang -### --target=aarch64-none-elf -march=armv8.2-a+rcpc3   %s 2>&1 | FileCheck %s --check-prefix=ENABLED
-// RUN: %clang -### --target=aarch64-none-elf -march=armv8.2-a+norcpc3 %s 2>&1 | FileCheck %s --check-prefix=DISABLED
+// RUN: %clang -### --target=aarch64-none-elf -march=armv8.2-a+norcpc3 %s 2>&1 | FileCheck %s --check-prefix=NOT_ENABLED
 // RUN: %clang -### --target=aarch64-none-elf -march=armv9-a         %s 2>&1 | FileCheck %s --check-prefix=NOT_ENABLED
 // RUN: %clang -### --target=aarch64-none-elf -march=armv9-a+rcpc3   %s 2>&1 | FileCheck %s --check-prefix=ENABLED
-// RUN: %clang -### --target=aarch64-none-elf -march=armv9-a+norcpc3 %s 2>&1 | FileCheck %s --check-prefix=DISABLED
+// RUN: %clang -### --target=aarch64-none-elf -march=armv9-a+norcpc3 %s 2>&1 | FileCheck %s --check-prefix=NOT_ENABLED
 
 // FEAT_RCPC3 is invalid before v8
 // RUN: not %clang -### --target=arm-none-none-eabi -march=armv7-a+rcpc3         %s 2>&1 | FileCheck %s --check-prefix=INVALID
@@ -22,5 +22,4 @@
 // INVALID: error: unsupported argument 'armv7-a+rcpc3' to option '-march='
 // ENABLED: "-target-feature" "+rcpc3"
 // NOT_ENABLED-NOT: "-target-feature" "+rcpc3"
-// DISABLED: "-target-feature" "-rcpc3"
 
diff --git a/clang/test/Driver/aarch64-ls64.c b/clang/test/Driver/aarch64-ls64.c
index f9781563a4ab4..7d828a22c60fd 100644
--- a/clang/test/Driver/aarch64-ls64.c
+++ b/clang/test/Driver/aarch64-ls64.c
@@ -3,8 +3,7 @@
 // RUN: %clang -### --target=aarch64-none-elf -march=armv8.7-a+ls64 %s 2>&1 | FileCheck %s
 // CHECK: "-target-feature" "+ls64"
 
-// RUN: %clang -### --target=aarch64-none-elf -march=armv8.7-a+nols64 %s 2>&1 | FileCheck %s --check-prefix=NO_LS64
-// NO_LS64: "-target-feature" "-ls64"
+// RUN: %clang -### --target=aarch64-none-elf -march=armv8.7-a+nols64 %s 2>&1 | FileCheck %s --check-prefix=ABSENT_LS64
 
 // The LD64B/ST64B accelerator extension is disabled by default.
 // RUN: %clang -### --target=aarch64-none-elf                  %s 2>&1 | FileCheck %s --check-prefix=ABSENT_LS64
diff --git a/clang/test/Driver/aarch64-lse128.c b/clang/test/Driver/aarch64-lse128.c
index d0bf31e49c965..612df87f21b35 100644
--- a/clang/test/Driver/aarch64-lse128.c
+++ b/clang/test/Driver/aarch64-lse128.c
@@ -3,8 +3,7 @@
 // FEAT_LSE128 is optional (off by default) for v9.4a and older, and can be enabled using +lse128
 // RUN: %clang -### --target=aarch64-none-elf -march=armv9.4-a          %s 2>&1 | FileCheck %s --check-prefix=NOT_ENABLED
 // RUN: %clang -### --target=aarch64-none-elf -march=armv9.4-a+lse128   %s 2>&1 | FileCheck %s --check-prefix=ENABLED
-// RUN: %clang -### --target=aarch64-none-elf -march=armv9.4-a+nolse128 %s 2>&1 | FileCheck %s --check-prefix=DISABLED
+// RUN: %clang -### --target=aarch64-none-elf -march=armv9.4-a+nolse128 %s 2>&1 | FileCheck %s --check-prefix=NOT_ENABLED
 
 // ENABLED: "-target-feature" "+lse128"
 // NOT_ENABLED-NOT: "-target-feature" "+lse128"
-// DISABLED: "-target-feature" "-lse128"
diff --git a/clang/test/Driver/aarch64-march.c b/clang/test/Driver/aarch64-march.c
index 59be5836b39dc..132cea0e2c624 100644
--- a/clang/test/Driver/aarch64-march.c
+++ b/clang/test/Driver/aarch64-march.c
@@ -1,9 +1,9 @@
 // RUN: %clang --target=aarch64 -### -c %s 2>&1 | FileCheck -check-prefix=GENERIC-V8A %s
 // RUN: %clang --target=aarch64 -march=armv8-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERIC-V8A %s
-// GENERIC-V8A: "-cc1"{{.*}} "-triple" "aarch64{{(--)?}}"{{.*}} "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v8a"
+// GENERIC-V8A: "-cc1"{{.*}} "-triple" "aarch64{{(--)?}}"{{.*}} "-target-cpu" "generic" "-target-feature" "+v8a"
 
 // RUN: %clang --target=aarch64 -march=armv8-r -### -c %s 2>&1 | FileCheck -check-prefix=GENERIC-V8R %s
-// GENERIC-V8R: "-cc1"{{.*}} "-triple" "aarch64{{(--)?}}"{{.*}} "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v8r"
+// GENERIC-V8R: "-cc1"{{.*}} "-triple" "aarch64{{(--)?}}"{{.*}} "-target-cpu" "generic" "-target-feature" "+v8r"
 
 // RUN: %clang --target=aarch64 -march=armv9a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV9A %s
 // RUN: %clang --target=aarch64 -march=armv9-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV9A %s
@@ -11,7 +11,7 @@
 // RUN: %clang --target=aarch64 -mlittle-endian -march=armv9-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV9A %s
 // RUN: %clang --target=aarch64_be -mlittle-endian -march=armv9a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV9A %s
 // RUN: %clang --target=aarch64_be -mlittle-endian -march=armv9-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV9A %s
-// GENERICV9A: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v9a" "-target-feature" "+sve" "-target-feature" "+sve2"
+// GENERICV9A: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-cpu" "generic" "-target-feature" "+v9a"{{.*}} "-target-feature" "+sve" "-target-feature" "+sve2"
 
 // RUN: %clang --target=aarch64_be -march=armv9a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV9A-BE %s
 // RUN: %clang --target=aarch64_be -march=armv9-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV9A-BE %s
@@ -19,7 +19,7 @@
 // RUN: %clang --target=aarch64 -mbig-endian -march=armv9-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV9A-BE %s
 // RUN: %clang --target=aarch64_be -mbig-endian -march=armv9a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV9A-BE %s
 // RUN: %clang --target=aarch64_be -mbig-endian -march=armv9-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV9A-BE %s
-// GENERICV9A-BE: "-cc1"{{.*}} "-triple" "aarch64_be{{.*}}" "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v9a" "-target-feature" "+sve" "-target-feature" "+sve2"
+// GENERICV9A-BE: "-cc1"{{.*}} "-triple" "aarch64_be{{.*}}" "-target-cpu" "generic" "-target-feature" "+v9a"{{.*}} "-target-feature" "+sve" "-target-feature" "+sve2"
 
 // ================== Check whether -march accepts mixed-case values.
 // RUN: %clang --target=aarch64_be -march=ARMV8.1A -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV81A-BE %s
@@ -28,4 +28,4 @@
 // RUN: %clang --target=aarch64 -mbig-endian -march=Armv8.1-A -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV81A-BE %s
 // RUN: %clang --target=aarch64_be -mbig-endian -march=ARMv8.1a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV81A-BE %s
 // RUN: %clang --target=aarch64_be -mbig-endian -march=ARMV8.1-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV81A-BE %s
-// GENERICV81A-BE: "-cc1"{{.*}} "-triple" "aarch64_be{{.*}}" "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v8.1a"
+// GENERICV81A-BE: "-cc1"{{.*}} "-triple" "aarch64_be{{.*}}" "-target-cpu" "generic" "-target-feature" "+v8.1a"
diff --git a/clang/test/Driver/aarch64-mgeneral_regs_only.c b/clang/test/Driver/aarch64-mgeneral_regs_only.c
index 07e837f94975f..17da9c614a8a1 100644
--- a/clang/test/Driver/aarch64-mgeneral_regs_only.c
+++ b/clang/test/Driver/aarch64-mgeneral_regs_only.c
@@ -5,5 +5,4 @@
 // RUN: %clang --target=arm64-linux-eabi -mgeneral-regs-only %s -### 2>&1 \
 // RUN:   | FileCheck --check-prefix=CHECK-NO-FP %s
 // CHECK-NO-FP: "-target-feature" "-fp-armv8"
-// CHECK-NO-FP: "-target-feature" "-crypto"
 // CHECK-NO-FP: "-target-feature" "-neon"
diff --git a/clang/test/Driver/aarch64-mte.c b/clang/test/Driver/aarch64-mte.c
index 7e8112734603f..435ffe0defc4b 100644
--- a/clang/test/Driver/aarch64-mte.c
+++ b/clang/test/Driver/aarch64-mte.c
@@ -1,9 +1,9 @@
 // RUN: %clang -### --target=aarch64-none-elf -march=armv8.4a+memtag %s 2>&1 | FileCheck %s
 // RUN: %clang -### --target=aarch64-none-elf -march=armv8.5a+memtag %s 2>&1 | FileCheck %s
+// RUN: %clang -### --target=aarch64-none-elf -mcpu=cortex-a510 %s 2>&1 | FileCheck %s
 // CHECK: "-target-feature" "+mte"
 
-// RUN: %clang -### --target=aarch64-none-elf -march=armv8.4a+nomemtag %s 2>&1 | FileCheck %s --check-prefix=NOMTE
-// RUN: %clang -### --target=aarch64-none-elf -march=armv8.5a+nomemtag %s 2>&1 | FileCheck %s --check-prefix=NOMTE
+// RUN: %clang -### --target=aarch64-none-elf -mcpu=cortex-a510+nomemtag %s 2>&1 | FileCheck %s --check-prefix=NOMTE
 // NOMTE: "-target-feature" "-mte"
 
 // RUN: %clang -### --target=aarch64-none-elf                 %s 2>&1 | FileCheck %s --check-prefix=ABSENTMTE
diff --git a/clang/test/Driver/aarch64-perfmon.c b/clang/test/Driver/aarch64-perfmon.c
index 1c421a4c3f73e..071be278497e0 100644
--- a/clang/test/Driver/aarch64-perfmon.c
+++ b/clang/test/Driver/aarch64-perfmon.c
@@ -2,12 +2,13 @@
 // RUN: %clang -### --target=aarch64-none-elf -march=armv8.2a+pmuv3 %s 2>&1 | FileCheck --check-prefix=CHECK-PERFMON %s
 // CHECK-PERFMON: "-target-feature" "+perfmon"
 
-// RUN: %clang -### --target=aarch64-none-elf -march=armv8.4a+nopmuv3 %s 2>&1 | FileCheck --check-prefix=CHECK-NOPERFMON %s
-// RUN: %clang -### --target=aarch64-none-elf -march=armv8.2a+nopmuv3 %s 2>&1 | FileCheck --check-prefix=CHECK-NOPERFMON %s
+// RUN: %clang -### --target=aarch64-none-elf -mcpu=cortex-a520+nopmuv3 %s 2>&1 | FileCheck --check-prefix=CHECK-NOPERFMON %s
 // CHECK-NOPERFMON: "-target-feature" "-perfmon"
 
 // RUN: %clang -### --target=aarch64-none-elf                 %s 2>&1 | FileCheck %s --check-prefix=ABSENTPERFMON
 // RUN: %clang -### --target=aarch64-none-elf -march=armv8.4a %s 2>&1 | FileCheck %s --check-prefix=ABSENTPERFMON
 // RUN: %clang -### --target=aarch64-none-elf -march=armv8.2a %s 2>&1 | FileCheck %s --check-prefix=ABSENTPERFMON
+// RUN: %clang -### --target=aarch64-none-elf -march=armv8.4a+nopmuv3 %s 2>&1 | FileCheck --check-prefix=ABSENTPERFMON %s
+// RUN: %clang -### --target=aarch64-none-elf -march=armv8.2a+nopmuv3 %s 2>&1 | FileCheck --check-prefix=ABSENTPERFMON %s
 // ABSENTPERFMON-NOT: "-target-feature" "+perfmon"
-// ABSENTPERFMON-NOT: "-target-feature" "-perfmon"
\ No newline at end of file
+// ABSENTPERFMON-NOT: "-target-feature" "-perfmon"
diff --git a/clang/test/Driver/aarch64-predres.c b/clang/test/Driver/aarch64-predres.c
index 4ed832f29a5c9..afa0830df7dd4 100644
--- a/clang/test/Driver/aarch64-predres.c
+++ b/clang/test/Driver/aarch64-predres.c
@@ -1,8 +1,9 @@
 // RUN: %clang -### --target=aarch64-none-elf -march=armv8a+predres     %s 2>&1 | FileCheck %s
+// RUN: %clang -### --target=aarch64-none-elf -mcpu=cortex-a520         %s 2>&1 | FileCheck %s
 // CHECK: "-target-feature" "+predres"
 // CHECK-NOT: "-target-feature" "-predres"
 
-// RUN: %clang -### --target=aarch64-none-elf -march=armv8.5a+nopredres %s 2>&1 | FileCheck %s --check-prefix=NOPR
+// RUN: %clang -### --target=aarch64-none-elf -mcpu=cortex-a520+nopredres %s 2>&1 | FileCheck %s --check-prefix=NOPR
 // NOPR: "-target-feature" "-predres"
 // NOPR-NOT: "-target-feature" "+predres"
 
diff --git a/clang/test/Driver/aarch64-rand.c b/clang/test/Driver/aarch64-rand.c
index 924ff638458c6..80891ec3f5482 100644
--- a/clang/test/Driver/aarch64-rand.c
+++ b/clang/test/Driver/aarch64-rand.c
@@ -2,8 +2,7 @@
 // RUN: %clang -### --target=aarch64-none-elf -march=armv8.5a+rng %s 2>&1 | FileCheck %s
 // CHECK: "-target-feature" "+rand"
 
-// RUN: %clang -### --target=aarch64-none-elf -march=armv8.4a+norng %s 2>&1 | FileCheck %s --check-prefix=NORAND
-// RUN: %clang -### --target=aarch64-none-elf -march=armv8.5a+norng %s 2>&1 | FileCheck %s --check-prefix=NORAND
+// RUN: %clang -### --target=aarch64-none-elf -mcpu=neoverse-v1+norng %s 2>&1 | FileCheck %s --check-prefix=NORAND
 // NORAND: "-target-feature" "-rand"
 
 // RUN: %clang -### --target=aarch64-none-elf                 %s 2>&1 | FileCheck %s --check-prefix=ABSENTRAND
diff --git a/clang/test/Driver/aarch64-ras.c b/clang/test/Driver/aarch64-ras.c
index f72fa211ff886..c1e1f1b85a171 100644
--- a/clang/test/Driver/aarch64-ras.c
+++ b/clang/test/Driver/aarch64-ras.c
@@ -1,20 +1,21 @@
 // RAS is off by default for v8a, but can be enabled by +ras (this is not architecturally valid)
 // RUN: %clang --target=aarch64-none-elf -march=armv8a+ras -### -c %s 2>&1 | FileCheck --check-prefix=CHECK-RAS %s
+// RUN: %clang --target=aarch64-none-elf -march=armv8.2a+ras -### -c %s 2>&1 | FileCheck --check-prefix=CHECK-RAS %s
 // RUN: %clang --target=aarch64-none-elf -march=armv8-a+ras -### -c %s 2>&1 | FileCheck --check-prefix=CHECK-RAS %s
 // RUN: %clang --target=aarch64-none-elf -mcpu=generic+ras -### -c %s 2>&1 | FileCheck --check-prefix=CHECK-RAS %s
 // RUN: %clang --target=aarch64-none-elf -mcpu=cortex-a75 -### -c %s 2>&1 | FileCheck --check-prefix=CHECK-RAS %s
 // RUN: %clang --target=aarch64-none-elf -mcpu=cortex-a55 -### -c %s 2>&1 | FileCheck --check-prefix=CHECK-RAS %s
 // CHECK-RAS: "-target-feature" "+ras"
 
-// RUN: %clang --target=aarch64-none-elf -march=armv8a+noras -### -c %s 2>&1 | FileCheck --check-prefix=CHECK-NORAS %s
-// RUN: %clang --target=aarch64-none-elf -mcpu=generic+noras -### -c %s 2>&1 | FileCheck --check-prefix=CHECK-NORAS %s
-// CHECK-NORAS: "-target-feature" "-ras"
+// RUN: %clang --target=aarch64-none-elf -march=armv8a+noras -### -c %s 2>&1 | FileCheck --check-prefix=CHECK-ABSENT %s
+// RUN: %clang --target=aarch64-none-elf -mcpu=generic+noras -### -c %s 2>&1 | FileCheck --check-prefix=CHECK-ABSENT %s
+// CHECK-ABSENT-NOT: "-target-feature" ++ras"
 
 // RAS is on by default for v8.2a, but can be disabled by +noras
-// FIXME: in the current implementation, RAS is not on by default at all for v8.2a (the test says it all...)
-// RUN: %clang --target=aarch64 -march=armv8.2a  -### -c %s 2>&1 | FileCheck -check-prefix=V82ARAS %s
-// RUN: %clang --target=aarch64 -march=armv8.2-a -### -c %s 2>&1 | FileCheck -check-prefix=V82ARAS %s
+// RUN: %clang --target=aarch64 -march=armv8.2a  -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-RAS %s
+// RUN: %clang --target=aarch64 -march=armv8.2-a -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-RAS %s
 // V82ARAS-NOT: "-target-feature" "+ras"
 // V82ARAS-NOT: "-target-feature" "-ras"
 // RUN: %clang --target=aarch64 -march=armv8.2a+noras  -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-NORAS %s
 // RUN: %clang --target=aarch64 -march=armv8.2-a+noras -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-NORAS %s
+// CHECK-NORAS: "-target-feature" "-ras"
diff --git a/clang/test/Driver/aarch64-rdm.c b/clang/test/Driver/aarch64-rdm.c
index 5cad9a9973a17..f2542b381e7c2 100644
--- a/clang/test/Driver/aarch64-rdm.c
+++ b/clang/test/Driver/aarch64-rdm.c
@@ -6,4 +6,8 @@
 
 // RUN: %clang --target=aarch64-none-elf -march=armv8a+nordm -### -c %s 2>&1 | FileCheck --check-prefix=CHECK-NORDM %s
 // RUN: %clang --target=aarch64-none-elf -mcpu=generic+nordm -### -c %s 2>&1 | FileCheck --check-prefix=CHECK-NORDM %s
-// CHECK-NORDM: "-target-feature" "-rdm"
+// CHECK-NORDM-NOT: "-target-feature" "+rdm"
+//
+// RUN: %clang --target=aarch64-none-elf -march=armv8.1a -### -c %s 2>&1 | FileCheck --check-prefix=CHECK-RDM %s
+// RUN: %clang --target=aarch64-none-elf -march=armv8.1a+nordm -### -c %s 2>&1 | FileCheck --check-prefix=CHECK-NORDM-DEFAULT %s
+// CHECK-NORDM-DEFAULT: "-target-feature" "-rdm"
diff --git a/clang/test/Driver/aarch64-security-options.c b/clang/test/Driver/aarch64-security-options.c
index e1f3013c41137..146add2d1cf70 100644
--- a/clang/test/Driver/aarch64-security-options.c
+++ b/clang/test/Driver/aarch64-security-options.c
@@ -1,17 +1,17 @@
 // Check the -msign-return-address= option, which has a required argument to
 // select scope.
 // RUN: %clang --target=aarch64 -c %s -### -msign-return-address=none                             2>&1 | \
-// RUN: FileCheck %s --check-prefix=RA-OFF --check-prefix=KEY --check-prefix=BTE-OFF --check-prefix=WARN
+// RUN: FileCheck %s --check-prefix=RA-OFF --check-prefix=KEY --check-prefix=BTE-OFF --check-prefix=GCS-OFF --check-prefix=WARN
 
 // RUN: %clang --target=aarch64 -c %s -### -msign-return-address=non-leaf                         2>&1 | \
-// RUN: FileCheck %s --check-prefix=RA-NON-LEAF --check-prefix=KEY-A --check-prefix=BTE-OFF --check-prefix=WARN
+// RUN: FileCheck %s --check-prefix=RA-NON-LEAF --check-prefix=KEY-A --check-prefix=BTE-OFF --check-prefix=GCS-OFF --check-prefix=WARN
 
 // RUN: %clang --target=aarch64 -c %s -### -msign-return-address=all                              2>&1 | \
-// RUN: FileCheck %s --check-prefix=RA-ALL      --check-prefix=KEY-A --check-prefix=BTE-OFF --check-prefix=WARN
+// RUN: FileCheck %s --check-prefix=RA-ALL      --check-prefix=KEY-A --check-prefix=BTE-OFF --check-prefix=GCS-OFF --check-prefix=WARN
 
 // -mbranch-protection with standard
 // RUN: %clang --target=aarch64 -c %s -### -mbranch-protection=standard                                2>&1 | \
-// RUN: FileCheck %s --check-prefix=RA-NON-LEAF --check-prefix=KEY-A --check-prefix=BTE-ON --check-prefix=WARN
+// RUN: FileCheck %s --check-prefix=RA-NON-LEAF --check-prefix=KEY-A --check-prefix=BTE-ON --check-prefix=GCS-ON --check-prefix=WARN
 
 // If the -msign-return-address and -mbranch-protection are both used, the
 // right-most one controls return address signing.
@@ -42,6 +42,9 @@
 // BTE-OFF-NOT: "-mbranch-target-enforce"
 // BTE-ON: "-mbranch-target-enforce"
 
+// GCS-OFF-NOT: "-mguarded-control-stack"
+// GCS-ON: "-mguarded-control-stack"
+
 // CONFLICT: "-msign-return-address=none"
 
 // BAD-RA-PROTECTION: unsupported argument 'foo' to option '-msign-return-address='
diff --git a/clang/test/Driver/aarch64-ssbs.c b/clang/test/Driver/aarch64-ssbs.c
index 033e3a6471713..62b603a096bc4 100644
--- a/clang/test/Driver/aarch64-ssbs.c
+++ b/clang/test/Driver/aarch64-ssbs.c
@@ -6,7 +6,7 @@
 
 // RUN: %clang -### --target=aarch64-none-elf -march=armv8a+nossbs %s 2>&1 | FileCheck %s --check-prefix=NOSSBS
 // RUN: %clang -### --target=aarch64-none-elf -mcpu=cortex-x1c+nossbs %s 2>&1 | FileCheck %s --check-prefix=NOSSBS
-// NOSSBS: "-target-feature" "-ssbs"
+// NOSSBS-NOT: "-target-feature" "+ssbs"
 
 // RUN: %clang -### --target=aarch64-none-elf                      %s 2>&1 | FileCheck %s --check-prefix=ABSENTSSBS
 // ABSENTSSBS-NOT: "-target-feature" "+ssbs"
diff --git a/clang/test/Driver/aarch64-sve2.c b/clang/test/Driver/aarch64-sve2.c
index 224c89dc3ce11..c801f44d3cb84 100644
--- a/clang/test/Driver/aarch64-sve2.c
+++ b/clang/test/Driver/aarch64-sve2.c
@@ -5,4 +5,4 @@
 // RUN: %clang --target=aarch64 -mlittle-endian -march=armv9-a+nosve2 -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV9A-NOSVE2 %s
 // RUN: %clang --target=aarch64_be -mlittle-endian -march=armv9a+nosve2 -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV9A-NOSVE2 %s
 // RUN: %clang --target=aarch64_be -mlittle-endian -march=armv9-a+nosve2 -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV9A-NOSVE2 %s
-// GENERICV9A-NOSVE2: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v9a" "-target-feature" "+sve" "-target-feature" "-sve2"
+// GENERICV9A-NOSVE2: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-cpu" "generic" "-target-feature" "+v9a"{{.*}} "-target-feature" "+neon" "-target-feature" "+sve" "-target-feature" "-sve2"
diff --git a/clang/test/Driver/aarch64-the.c b/clang/test/Driver/aarch64-the.c
index dcaeae115d2ea..c776cfeb044b3 100644
--- a/clang/test/Driver/aarch64-the.c
+++ b/clang/test/Driver/aarch64-the.c
@@ -22,5 +22,5 @@
 // INVALID: error: unsupported argument 'armv7-a+the' to option '-march='
 // ENABLED: "-target-feature" "+the"
 // NOT_ENABLED-NOT: "-target-feature" "+the"
-// DISABLED: "-target-feature" "-the"
+// DISABLED-NOT: "-target-feature" "{{[-+]}}the"
 
diff --git a/clang/test/Driver/aarch64-v81a.c b/clang/test/Driver/aarch64-v81a.c
index 1f8a707b116ee..e84652ec7f11e 100644
--- a/clang/test/Driver/aarch64-v81a.c
+++ b/clang/test/Driver/aarch64-v81a.c
@@ -4,7 +4,7 @@
 // RUN: %clang --target=aarch64 -mbig-endian -march=armv8.1-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV81A-BE %s
 // RUN: %clang --target=aarch64_be -mbig-endian -march=armv8.1a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV81A-BE %s
 // RUN: %clang --target=aarch64_be -mbig-endian -march=armv8.1-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV81A-BE %s
-// GENERICV81A-BE: "-cc1"{{.*}} "-triple" "aarch64_be{{.*}}" "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v8.1a"
+// GENERICV81A-BE: "-cc1"{{.*}} "-triple" "aarch64_be{{.*}}" "-target-cpu" "generic"{{.*}} "-target-feature" "+v8.1a"{{.*}} "-target-feature" "+neon"
 
 // RUN: %clang --target=aarch64 -march=armv8.1a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV81A %s
 // RUN: %clang --target=aarch64 -march=armv8.1-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV81A %s
@@ -12,10 +12,10 @@
 // RUN: %clang --target=aarch64 -mlittle-endian -march=armv8.1-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV81A %s
 // RUN: %clang --target=aarch64_be -mlittle-endian -march=armv8.1a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV81A %s
 // RUN: %clang --target=aarch64_be -mlittle-endian -march=armv8.1-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV81A %s
-// GENERICV81A: "-cc1"{{.*}} "-triple" "aarch64{{(--)?}}"{{.*}} "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v8.1a"
+// GENERICV81A: "-cc1"{{.*}} "-triple" "aarch64{{(--)?}}"{{.*}} "-target-cpu" "generic"{{.*}} "-target-feature" "+v8.1a"{{.*}} "-target-feature" "+neon"{{.*}}
 
 // RUN: %clang --target=arm64 -march=armv8.1a -### -c %s 2>&1 | FileCheck -check-prefix=ARM64-GENERICV81A %s
 // RUN: %clang --target=arm64 -march=armv8.1-a -### -c %s 2>&1 | FileCheck -check-prefix=ARM64-GENERICV81A %s
 // RUN: %clang --target=arm64 -mlittle-endian -march=armv8.1a -### -c %s 2>&1 | FileCheck -check-prefix=ARM64-GENERICV81A %s
 // RUN: %clang --target=arm64 -mlittle-endian -march=armv8.1-a -### -c %s 2>&1 | FileCheck -check-prefix=ARM64-GENERICV81A %s
-// ARM64-GENERICV81A: "-cc1"{{.*}} "-triple" "arm64{{.*}}" "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v8.1a"
+// ARM64-GENERICV81A: "-cc1"{{.*}} "-triple" "arm64{{.*}}" "-target-cpu" "generic"{{.*}} "-target-feature" "+v8.1a"{{.*}} "-target-feature" "+neon"
diff --git a/clang/test/Driver/aarch64-v82a.c b/clang/test/Driver/aarch64-v82a.c
index ae70255e6a230..9dd355934c105 100644
--- a/clang/test/Driver/aarch64-v82a.c
+++ b/clang/test/Driver/aarch64-v82a.c
@@ -4,7 +4,7 @@
 // RUN: %clang --target=aarch64 -mlittle-endian -march=armv8.2-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV82A %s
 // RUN: %clang --target=aarch64_be -mlittle-endian -march=armv8.2a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV82A %s
 // RUN: %clang --target=aarch64_be -mlittle-endian -march=armv8.2-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV82A %s
-// GENERICV82A: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v8.2a"
+// GENERICV82A: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-cpu" "generic" "-target-feature" "+v8.2a"{{.*}} "-target-feature" "+neon"
 
 // RUN: %clang --target=aarch64_be -march=armv8.2a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV82A-BE %s
 // RUN: %clang --target=aarch64_be -march=armv8.2-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV82A-BE %s
@@ -12,4 +12,4 @@
 // RUN: %clang --target=aarch64 -mbig-endian -march=armv8.2-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV82A-BE %s
 // RUN: %clang --target=aarch64_be -mbig-endian -march=armv8.2a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV82A-BE %s
 // RUN: %clang --target=aarch64_be -mbig-endian -march=armv8.2-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV82A-BE %s
-// GENERICV82A-BE: "-cc1"{{.*}} "-triple" "aarch64_be{{.*}}" "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v8.2a"
+// GENERICV82A-BE: "-cc1"{{.*}} "-triple" "aarch64_be{{.*}}" "-target-cpu" "generic" "-target-feature" "+v8.2a"{{.*}} "-target-feature" "+neon"
diff --git a/clang/test/Driver/aarch64-v83a.c b/clang/test/Driver/aarch64-v83a.c
index 63c47e39f78c2..b0ff9fb3abc24 100644
--- a/clang/test/Driver/aarch64-v83a.c
+++ b/clang/test/Driver/aarch64-v83a.c
@@ -4,7 +4,7 @@
 // RUN: %clang --target=aarch64 -mlittle-endian -march=armv8.3-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV83A %s
 // RUN: %clang --target=aarch64_be -mlittle-endian -march=armv8.3a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV83A %s
 // RUN: %clang --target=aarch64_be -mlittle-endian -march=armv8.3-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV83A %s
-// GENERICV83A: "-cc1"{{.*}} "-triple" "aarch64{{(--)?}}"{{.*}} "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v8.3a"
+// GENERICV83A: "-cc1"{{.*}} "-triple" "aarch64{{(--)?}}"{{.*}} "-target-cpu" "generic" "-target-feature" "+v8.3a"{{.*}} "-target-feature" "+neon"
 
 // RUN: %clang --target=aarch64_be -march=armv8.3a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV83A-BE %s
 // RUN: %clang --target=aarch64_be -march=armv8.3-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV83A-BE %s
@@ -12,4 +12,4 @@
 // RUN: %clang --target=aarch64 -mbig-endian -march=armv8.3-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV83A-BE %s
 // RUN: %clang --target=aarch64_be -mbig-endian -march=armv8.3a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV83A-BE %s
 // RUN: %clang --target=aarch64_be -mbig-endian -march=armv8.3-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV83A-BE %s
-// GENERICV83A-BE: "-cc1"{{.*}} "-triple" "aarch64_be{{.*}}" "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v8.3a"
+// GENERICV83A-BE: "-cc1"{{.*}} "-triple" "aarch64_be{{.*}}" "-target-cpu" "generic" "-target-feature" "+v8.3a"{{.*}} "-target-feature" "+neon"
diff --git a/clang/test/Driver/aarch64-v84a.c b/clang/test/Driver/aarch64-v84a.c
index f6a573bb02831..030990bfe5131 100644
--- a/clang/test/Driver/aarch64-v84a.c
+++ b/clang/test/Driver/aarch64-v84a.c
@@ -4,7 +4,7 @@
 // RUN: %clang --target=aarch64 -mlittle-endian -march=armv8.4-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV84A %s
 // RUN: %clang --target=aarch64_be -mlittle-endian -march=armv8.4a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV84A %s
 // RUN: %clang --target=aarch64_be -mlittle-endian -march=armv8.4-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV84A %s
-// GENERICV84A: "-cc1"{{.*}} "-triple" "aarch64{{(--)?}}"{{.*}} "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v8.4a"
+// GENERICV84A: "-cc1"{{.*}} "-triple" "aarch64{{(--)?}}"{{.*}} "-target-cpu" "generic" "-target-feature" "+v8.4a"{{.*}} "-target-feature" "+neon"
 
 // RUN: %clang --target=aarch64_be -march=armv8.4a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV84A-BE %s
 // RUN: %clang --target=aarch64_be -march=armv8.4-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV84A-BE %s
@@ -12,4 +12,4 @@
 // RUN: %clang --target=aarch64 -mbig-endian -march=armv8.4-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV84A-BE %s
 // RUN: %clang --target=aarch64_be -mbig-endian -march=armv8.4a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV84A-BE %s
 // RUN: %clang --target=aarch64_be -mbig-endian -march=armv8.4-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV84A-BE %s
-// GENERICV84A-BE: "-cc1"{{.*}} "-triple" "aarch64_be{{.*}}" "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v8.4a"
+// GENERICV84A-BE: "-cc1"{{.*}} "-triple" "aarch64_be{{.*}}" "-target-cpu" "generic" "-target-feature" "+v8.4a"{{.*}} "-target-feature" "+neon"
diff --git a/clang/test/Driver/aarch64-v85a.c b/clang/test/Driver/aarch64-v85a.c
index 4ee76faf48c41..3e1e921dcc013 100644
--- a/clang/test/Driver/aarch64-v85a.c
+++ b/clang/test/Driver/aarch64-v85a.c
@@ -4,7 +4,7 @@
 // RUN: %clang --target=aarch64 -mlittle-endian -march=armv8.5-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV85A %s
 // RUN: %clang --target=aarch64_be -mlittle-endian -march=armv8.5a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV85A %s
 // RUN: %clang --target=aarch64_be -mlittle-endian -march=armv8.5-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV85A %s
-// GENERICV85A: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v8.5a"
+// GENERICV85A: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-cpu" "generic" "-target-feature" "+v8.5a"{{.*}} "-target-feature" "+neon"
 
 // RUN: %clang --target=aarch64_be -march=armv8.5a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV85A-BE %s
 // RUN: %clang --target=aarch64_be -march=armv8.5-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV85A-BE %s
@@ -12,4 +12,4 @@
 // RUN: %clang --target=aarch64 -mbig-endian -march=armv8.5-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV85A-BE %s
 // RUN: %clang --target=aarch64_be -mbig-endian -march=armv8.5a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV85A-BE %s
 // RUN: %clang --target=aarch64_be -mbig-endian -march=armv8.5-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV85A-BE %s
-// GENERICV85A-BE: "-cc1"{{.*}} "-triple" "aarch64_be{{.*}}" "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v8.5a"
+// GENERICV85A-BE: "-cc1"{{.*}} "-triple" "aarch64_be{{.*}}" "-target-cpu" "generic" "-target-feature" "+v8.5a"{{.*}} "-target-feature" "+neon"
diff --git a/clang/test/Driver/aarch64-v86a.c b/clang/test/Driver/aarch64-v86a.c
index 970e21c69ce8e..ba2b57979b518 100644
--- a/clang/test/Driver/aarch64-v86a.c
+++ b/clang/test/Driver/aarch64-v86a.c
@@ -4,7 +4,7 @@
 // RUN: %clang --target=aarch64 -mlittle-endian -march=armv8.6-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV86A %s
 // RUN: %clang --target=aarch64_be -mlittle-endian -march=armv8.6a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV86A %s
 // RUN: %clang --target=aarch64_be -mlittle-endian -march=armv8.6-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV86A %s
-// GENERICV86A: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v8.6a"
+// GENERICV86A: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-cpu" "generic" "-target-feature" "+v8.6a"{{.*}} "-target-feature" "+neon"
 
 // RUN: %clang --target=aarch64_be -march=armv8.6a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV86A-BE %s
 // RUN: %clang --target=aarch64_be -march=armv8.6-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV86A-BE %s
@@ -12,4 +12,4 @@
 // RUN: %clang --target=aarch64 -mbig-endian -march=armv8.6-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV86A-BE %s
 // RUN: %clang --target=aarch64_be -mbig-endian -march=armv8.6a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV86A-BE %s
 // RUN: %clang --target=aarch64_be -mbig-endian -march=armv8.6-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV86A-BE %s
-// GENERICV86A-BE: "-cc1"{{.*}} "-triple" "aarch64_be{{.*}}" "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v8.6a"
+// GENERICV86A-BE: "-cc1"{{.*}} "-triple" "aarch64_be{{.*}}" "-target-cpu" "generic" "-target-feature" "+v8.6a"{{.*}} "-target-feature" "+neon"
diff --git a/clang/test/Driver/aarch64-v87a.c b/clang/test/Driver/aarch64-v87a.c
index 4e1d403faada7..ee4b68882739a 100644
--- a/clang/test/Driver/aarch64-v87a.c
+++ b/clang/test/Driver/aarch64-v87a.c
@@ -4,7 +4,7 @@
 // RUN: %clang --target=aarch64 -mlittle-endian -march=armv8.7-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV87A %s
 // RUN: %clang --target=aarch64_be -mlittle-endian -march=armv8.7a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV87A %s
 // RUN: %clang --target=aarch64_be -mlittle-endian -march=armv8.7-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV87A %s
-// GENERICV87A: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v8.7a"
+// GENERICV87A: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-cpu" "generic" "-target-feature" "+v8.7a"{{.*}} "-target-feature" "+neon"
 
 // RUN: %clang --target=aarch64_be -march=armv8.7a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV87A-BE %s
 // RUN: %clang --target=aarch64_be -march=armv8.7-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV87A-BE %s
@@ -12,4 +12,4 @@
 // RUN: %clang --target=aarch64 -mbig-endian -march=armv8.7-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV87A-BE %s
 // RUN: %clang --target=aarch64_be -mbig-endian -march=armv8.7a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV87A-BE %s
 // RUN: %clang --target=aarch64_be -mbig-endian -march=armv8.7-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV87A-BE %s
-// GENERICV87A-BE: "-cc1"{{.*}} "-triple" "aarch64_be{{.*}}" "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v8.7a"
+// GENERICV87A-BE: "-cc1"{{.*}} "-triple" "aarch64_be{{.*}}" "-target-cpu" "generic" "-target-feature" "+v8.7a"{{.*}} "-target-feature" "+neon"
diff --git a/clang/test/Driver/aarch64-v88a.c b/clang/test/Driver/aarch64-v88a.c
index 19db5fb862cba..b680c1f567134 100644
--- a/clang/test/Driver/aarch64-v88a.c
+++ b/clang/test/Driver/aarch64-v88a.c
@@ -4,7 +4,7 @@
 // RUN: %clang --target=aarch64 -mlittle-endian -march=armv8.8-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV88A %s
 // RUN: %clang --target=aarch64_be -mlittle-endian -march=armv8.8a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV88A %s
 // RUN: %clang --target=aarch64_be -mlittle-endian -march=armv8.8-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV88A %s
-// GENERICV88A: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v8.8a"
+// GENERICV88A: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-cpu" "generic" "-target-feature" "+v8.8a"{{.*}} "-target-feature" "+neon"
 
 // RUN: %clang --target=aarch64_be -march=armv8.8a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV88A-BE %s
 // RUN: %clang --target=aarch64_be -march=armv8.8-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV88A-BE %s
@@ -12,4 +12,4 @@
 // RUN: %clang --target=aarch64 -mbig-endian -march=armv8.8-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV88A-BE %s
 // RUN: %clang --target=aarch64_be -mbig-endian -march=armv8.8a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV88A-BE %s
 // RUN: %clang --target=aarch64_be -mbig-endian -march=armv8.8-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV88A-BE %s
-// GENERICV88A-BE: "-cc1"{{.*}} "-triple" "aarch64_be{{.*}}" "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v8.8a"
+// GENERICV88A-BE: "-cc1"{{.*}} "-triple" "aarch64_be{{.*}}" "-target-cpu" "generic" "-target-feature" "+v8.8a"{{.*}} "-target-feature" "+neon"
diff --git a/clang/test/Driver/aarch64-v89a.c b/clang/test/Driver/aarch64-v89a.c
index d803dff269920..903b793d046ba 100644
--- a/clang/test/Driver/aarch64-v89a.c
+++ b/clang/test/Driver/aarch64-v89a.c
@@ -4,11 +4,11 @@
 // RUN: %clang --target=aarch64 -mlittle-endian -march=armv8.9-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV89A %s
 // RUN: %clang --target=aarch64_be -mlittle-endian -march=armv8.9a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV89A %s
 // RUN: %clang --target=aarch64_be -mlittle-endian -march=armv8.9-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV89A %s
-// GENERICV89A: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v8.9a"
+// GENERICV89A: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-cpu" "generic" "-target-feature" "+v8.9a"{{.*}} "-target-feature" "+neon"
 // RUN: %clang --target=aarch64_be -march=armv8.9a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV89A-BE %s
 // RUN: %clang --target=aarch64_be -march=armv8.9-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV89A-BE %s
 // RUN: %clang --target=aarch64 -mbig-endian -march=armv8.9a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV89A-BE %s
 // RUN: %clang --target=aarch64 -mbig-endian -march=armv8.9-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV89A-BE %s
 // RUN: %clang --target=aarch64_be -mbig-endian -march=armv8.9a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV89A-BE %s
 // RUN: %clang --target=aarch64_be -mbig-endian -march=armv8.9-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV89A-BE %s
-// GENERICV89A-BE: "-cc1"{{.*}} "-triple" "aarch64_be{{.*}}" "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v8.9a"
+// GENERICV89A-BE: "-cc1"{{.*}} "-triple" "aarch64_be{{.*}}" "-target-cpu" "generic" "-target-feature" "+v8.9a"{{.*}} "-target-feature" "+neon"
diff --git a/clang/test/Driver/aarch64-v91a.c b/clang/test/Driver/aarch64-v91a.c
index a9e3833c5af9f..80853a59d0153 100644
--- a/clang/test/Driver/aarch64-v91a.c
+++ b/clang/test/Driver/aarch64-v91a.c
@@ -4,7 +4,7 @@
 // RUN: %clang --target=aarch64 -mlittle-endian -march=armv9.1-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV91A %s
 // RUN: %clang --target=aarch64_be -mlittle-endian -march=armv9.1a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV91A %s
 // RUN: %clang --target=aarch64_be -mlittle-endian -march=armv9.1-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV91A %s
-// GENERICV91A: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v9.1a" "-target-feature" "+sve" "-target-feature" "+sve2"
+// GENERICV91A: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-cpu" "generic" "-target-feature" "+v9.1a"{{.*}} "-target-feature" "+sve" "-target-feature" "+sve2"
 
 // RUN: %clang --target=aarch64_be -march=armv9.1a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV91A-BE %s
 // RUN: %clang --target=aarch64_be -march=armv9.1-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV91A-BE %s
@@ -12,4 +12,4 @@
 // RUN: %clang --target=aarch64 -mbig-endian -march=armv9.1-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV91A-BE %s
 // RUN: %clang --target=aarch64_be -mbig-endian -march=armv9.1a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV91A-BE %s
 // RUN: %clang --target=aarch64_be -mbig-endian -march=armv9.1-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV91A-BE %s
-// GENERICV91A-BE: "-cc1"{{.*}} "-triple" "aarch64_be{{.*}}" "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v9.1a" "-target-feature" "+sve" "-target-feature" "+sve2"
+// GENERICV91A-BE: "-cc1"{{.*}} "-triple" "aarch64_be{{.*}}" "-target-cpu" "generic" "-target-feature" "+v9.1a"{{.*}} "-target-feature" "+sve" "-target-feature" "+sve2"
diff --git a/clang/test/Driver/aarch64-v92a.c b/clang/test/Driver/aarch64-v92a.c
index c9eab8475155d..ee644cc6f3c62 100644
--- a/clang/test/Driver/aarch64-v92a.c
+++ b/clang/test/Driver/aarch64-v92a.c
@@ -4,7 +4,7 @@
 // RUN: %clang --target=aarch64 -mlittle-endian -march=armv9.2-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV92A %s
 // RUN: %clang --target=aarch64_be -mlittle-endian -march=armv9.2a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV92A %s
 // RUN: %clang --target=aarch64_be -mlittle-endian -march=armv9.2-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV92A %s
-// GENERICV92A: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v9.2a" "-target-feature" "+sve" "-target-feature" "+sve2"
+// GENERICV92A: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-cpu" "generic" "-target-feature" "+v9.2a"{{.*}} "-target-feature" "+sve" "-target-feature" "+sve2"
 
 // RUN: %clang --target=aarch64_be -march=armv9.2a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV92A-BE %s
 // RUN: %clang --target=aarch64_be -march=armv9.2-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV92A-BE %s
@@ -12,4 +12,4 @@
 // RUN: %clang --target=aarch64 -mbig-endian -march=armv9.2-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV92A-BE %s
 // RUN: %clang --target=aarch64_be -mbig-endian -march=armv9.2a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV92A-BE %s
 // RUN: %clang --target=aarch64_be -mbig-endian -march=armv9.2-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV92A-BE %s
-// GENERICV92A-BE: "-cc1"{{.*}} "-triple" "aarch64_be{{.*}}" "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v9.2a" "-target-feature" "+sve" "-target-feature" "+sve2"
+// GENERICV92A-BE: "-cc1"{{.*}} "-triple" "aarch64_be{{.*}}" "-target-cpu" "generic" "-target-feature" "+v9.2a"{{.*}} "-target-feature" "+sve" "-target-feature" "+sve2"
diff --git a/clang/test/Driver/aarch64-v93a.c b/clang/test/Driver/aarch64-v93a.c
index 9d6e715f3aac2..817559e28ccf4 100644
--- a/clang/test/Driver/aarch64-v93a.c
+++ b/clang/test/Driver/aarch64-v93a.c
@@ -4,7 +4,7 @@
 // RUN: %clang --target=aarch64 -mlittle-endian -march=armv9.3-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV93A %s
 // RUN: %clang --target=aarch64_be -mlittle-endian -march=armv9.3a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV93A %s
 // RUN: %clang --target=aarch64_be -mlittle-endian -march=armv9.3-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV93A %s
-// GENERICV93A: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v9.3a"
+// GENERICV93A: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-cpu" "generic" "-target-feature" "+v9.3a"
 
 // RUN: %clang --target=aarch64_be -march=armv9.3a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV93A-BE %s
 // RUN: %clang --target=aarch64_be -march=armv9.3-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV93A-BE %s
@@ -12,4 +12,4 @@
 // RUN: %clang --target=aarch64 -mbig-endian -march=armv9.3-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV93A-BE %s
 // RUN: %clang --target=aarch64_be -mbig-endian -march=armv9.3a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV93A-BE %s
 // RUN: %clang --target=aarch64_be -mbig-endian -march=armv9.3-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV93A-BE %s
-// GENERICV93A-BE: "-cc1"{{.*}} "-triple" "aarch64_be{{.*}}" "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v9.3a"
+// GENERICV93A-BE: "-cc1"{{.*}} "-triple" "aarch64_be{{.*}}" "-target-cpu" "generic" "-target-feature" "+v9.3a"
diff --git a/clang/test/Driver/aarch64-v94a.c b/clang/test/Driver/aarch64-v94a.c
index 49659ac949b4e..9998cc8a4a216 100644
--- a/clang/test/Driver/aarch64-v94a.c
+++ b/clang/test/Driver/aarch64-v94a.c
@@ -4,7 +4,7 @@
 // RUN: %clang --target=aarch64 -mlittle-endian -march=armv9.4-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV94A %s
 // RUN: %clang --target=aarch64_be -mlittle-endian -march=armv9.4a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV94A %s
 // RUN: %clang --target=aarch64_be -mlittle-endian -march=armv9.4-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV94A %s
-// GENERICV94A: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v9.4a"
+// GENERICV94A: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-cpu" "generic" "-target-feature" "+v9.4a"
 
 // RUN: %clang --target=aarch64_be -march=armv9.4a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV94A-BE %s
 // RUN: %clang --target=aarch64_be -march=armv9.4-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV94A-BE %s
@@ -12,4 +12,4 @@
 // RUN: %clang --target=aarch64 -mbig-endian -march=armv9.4-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV94A-BE %s
 // RUN: %clang --target=aarch64_be -mbig-endian -march=armv9.4a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV94A-BE %s
 // RUN: %clang --target=aarch64_be -mbig-endian -march=armv9.4-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV94A-BE %s
-// GENERICV94A-BE: "-cc1"{{.*}} "-triple" "aarch64_be{{.*}}" "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v9.4a"
+// GENERICV94A-BE: "-cc1"{{.*}} "-triple" "aarch64_be{{.*}}" "-target-cpu" "generic" "-target-feature" "+v9.4a"
diff --git a/clang/test/Driver/aarch64-v95a.c b/clang/test/Driver/aarch64-v95a.c
index 13069c04c8d1c..1037da65c8cb5 100644
--- a/clang/test/Driver/aarch64-v95a.c
+++ b/clang/test/Driver/aarch64-v95a.c
@@ -6,7 +6,7 @@
 // RUN: %clang -target aarch64 -mlittle-endian -march=armv9.5-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV95A %s
 // RUN: %clang -target aarch64_be -mlittle-endian -march=armv9.5a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV95A %s
 // RUN: %clang -target aarch64_be -mlittle-endian -march=armv9.5-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV95A %s
-// GENERICV95A: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v9.5a"
+// GENERICV95A: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-cpu" "generic" "-target-feature" "+v9.5a"
 
 // RUN: %clang -target aarch64_be -march=armv9.5a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV95A-BE %s
 // RUN: %clang -target aarch64_be -march=armv9.5-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV95A-BE %s
@@ -14,18 +14,18 @@
 // RUN: %clang -target aarch64 -mbig-endian -march=armv9.5-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV95A-BE %s
 // RUN: %clang -target aarch64_be -mbig-endian -march=armv9.5a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV95A-BE %s
 // RUN: %clang -target aarch64_be -mbig-endian -march=armv9.5-a -### -c %s 2>&1 | FileCheck -check-prefix=GENERICV95A-BE %s
-// GENERICV95A-BE: "-cc1"{{.*}} "-triple" "aarch64_be{{.*}}" "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v9.5a"
+// GENERICV95A-BE: "-cc1"{{.*}} "-triple" "aarch64_be{{.*}}" "-target-cpu" "generic" "-target-feature" "+v9.5a"
 
 // ===== Features supported on aarch64 =====
 
 // RUN: %clang -target aarch64 -march=armv9.5a+cpa -### -c %s 2>&1 | FileCheck -check-prefix=V95A-CPA %s
 // RUN: %clang -target aarch64 -march=armv9.5-a+cpa -### -c %s 2>&1 | FileCheck -check-prefix=V95A-CPA %s
-// V95A-CPA: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v9.5a" "-target-feature" "+cpa"
+// V95A-CPA: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-cpu" "generic" "-target-feature" "+v9.5a"{{.*}} "-target-feature" "+cpa"
 
 // RUN: %clang -target aarch64 -march=armv9.5a+pauth-lr -### -c %s 2>&1 | FileCheck -check-prefix=V95A-PAUTHLR %s
 // RUN: %clang -target aarch64 -march=armv9.5-a+pauth-lr -### -c %s 2>&1 | FileCheck -check-prefix=V95A-PAUTHLR %s
-// V95A-PAUTHLR: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v9.5a" "-target-feature" "+pauth-lr"
+// V95A-PAUTHLR: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-cpu" "generic" "-target-feature" "+v9.5a"{{.*}} "-target-feature" "+pauth-lr"
 
 // RUN: %clang -target aarch64 -march=armv9.5a+tlbiw -### -c %s 2>&1 | FileCheck -check-prefix=V95A-TLBIW %s
 // RUN: %clang -target aarch64 -march=armv9.5-a+tlbiw -### -c %s 2>&1 | FileCheck -check-prefix=V95A-TLBIW %s
-// V95A-TLBIW: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-cpu" "generic" "-target-feature" "+neon" "-target-feature" "+v9.5a" "-target-feature" "+tlbiw"
+// V95A-TLBIW: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-cpu" "generic" "-target-feature" "+v9.5a"{{.*}} "-target-feature" "+tlbiw"
diff --git a/clang/test/Driver/arm-sb.c b/clang/test/Driver/arm-sb.c
index 9c2bf082d5b02..f2704f33c2711 100644
--- a/clang/test/Driver/arm-sb.c
+++ b/clang/test/Driver/arm-sb.c
@@ -1,14 +1,16 @@
 // RUN: %clang -### -target arm-none-none-eabi -march=armv8a+sb %s 2>&1 | FileCheck %s
 // RUN: %clang -### -target aarch64-none-elf -march=armv8a+sb %s 2>&1 | FileCheck %s
+// RUN: %clang -### -target aarch64-none-elf -mcpu=cortex-a510 %s 2>&1 | FileCheck %s
 // CHECK: "-target-feature" "+sb"
 // CHECK-NOT: "-target-feature" "-sb"
 
 // RUN: %clang -### -target arm-none-none-eabi -march=armv8.5a+nosb %s 2>&1 | FileCheck %s --check-prefix=NOSB
-// RUN: %clang -### -target aarch64-none-elf -march=armv8.5a+nosb %s 2>&1 | FileCheck %s --check-prefix=NOSB
+// RUN: %clang -### -target aarch64-none-elf -mcpu=cortex-a510+nosb %s 2>&1 | FileCheck %s --check-prefix=NOSB
 // NOSB: "-target-feature" "-sb"
 // NOSB-NOT: "-target-feature" "+sb"
 
 // RUN: %clang -### -target arm-none-none-eabi %s 2>&1 | FileCheck %s --check-prefix=ABSENT
 // RUN: %clang -### -target aarch64-none-elf %s 2>&1 | FileCheck %s --check-prefix=ABSENT
+// RUN: %clang -### -target aarch64-none-elf -march=armv8.5a+nosb %s 2>&1 | FileCheck %s --check-prefix=ABSENT
 // ABSENT-NOT: "-target-feature" "+sb"
 // ABSENT-NOT: "-target-feature" "-sb"
diff --git a/clang/test/Driver/fdefine-target-os-macros.c b/clang/test/Driver/fdefine-target-os-macros.c
index 030d4ce34cb28..d7379dd3d5396 100644
--- a/clang/test/Driver/fdefine-target-os-macros.c
+++ b/clang/test/Driver/fdefine-target-os-macros.c
@@ -1,12 +1,11 @@
 // RUN: %clang -### --target=arm64-apple-darwin %s 2>&1 | FileCheck %s --check-prefix=DARWIN-DEFAULT
-// DARWIN-DEFAULT-NOT: "-fdefine-target-os-macros"
+// DARWIN-DEFAULT: "-fdefine-target-os-macros"
 
 // RUN: %clang -### --target=arm-none-linux-gnu %s 2>&1 | FileCheck %s --check-prefix=NON-DARWIN-DEFAULT
 // RUN: %clang -### --target=x86_64-pc-win32 %s 2>&1 | FileCheck %s --check-prefix=NON-DARWIN-DEFAULT
 // NON-DARWIN-DEFAULT-NOT: "-fdefine-target-os-macros"
 
-// RUN: %clang -dM -E --target=arm64-apple-macos \
-// RUN:        -fdefine-target-os-macros %s 2>&1 \
+// RUN: %clang -dM -E --target=arm64-apple-macos %s 2>&1 \
 // RUN: | FileCheck %s -DMAC=1         \
 // RUN:                -DOSX=1         \
 // RUN:                -DIPHONE=0      \
@@ -21,8 +20,7 @@
 // RUN:                -DLINUX=0       \
 // RUN:                -DUNIX=0
 
-// RUN: %clang -dM -E --target=arm64-apple-ios \
-// RUN:        -fdefine-target-os-macros %s 2>&1 \
+// RUN: %clang -dM -E --target=arm64-apple-ios %s 2>&1 \
 // RUN: | FileCheck %s -DMAC=1         \
 // RUN:                -DOSX=0         \
 // RUN:                -DIPHONE=1      \
@@ -37,8 +35,7 @@
 // RUN:                -DLINUX=0       \
 // RUN:                -DUNIX=0
 
-// RUN: %clang -dM -E --target=arm64-apple-ios-macabi \
-// RUN:        -fdefine-target-os-macros %s 2>&1 \
+// RUN: %clang -dM -E --target=arm64-apple-ios-macabi %s 2>&1 \
 // RUN: | FileCheck %s -DMAC=1         \
 // RUN:                -DOSX=0         \
 // RUN:                -DIPHONE=1      \
@@ -53,8 +50,7 @@
 // RUN:                -DLINUX=0       \
 // RUN:                -DUNIX=0
 
-// RUN: %clang -dM -E --target=arm64-apple-ios-simulator \
-// RUN:        -fdefine-target-os-macros %s 2>&1 \
+// RUN: %clang -dM -E --target=arm64-apple-ios-simulator %s 2>&1 \
 // RUN: | FileCheck %s -DMAC=1         \
 // RUN:                -DOSX=0         \
 // RUN:                -DIPHONE=1      \
@@ -69,8 +65,7 @@
 // RUN:                -DLINUX=0       \
 // RUN:                -DUNIX=0
 
-// RUN: %clang -dM -E --target=arm64-apple-tvos \
-// RUN:        -fdefine-target-os-macros %s 2>&1 \
+// RUN: %clang -dM -E --target=arm64-apple-tvos %s 2>&1 \
 // RUN: | FileCheck %s -DMAC=1         \
 // RUN:                -DOSX=0         \
 // RUN:                -DIPHONE=1      \
@@ -85,8 +80,7 @@
 // RUN:                -DLINUX=0       \
 // RUN:                -DUNIX=0
 
-// RUN: %clang -dM -E --target=arm64-apple-tvos-simulator \
-// RUN:        -fdefine-target-os-macros %s 2>&1 \
+// RUN: %clang -dM -E --target=arm64-apple-tvos-simulator %s 2>&1 \
 // RUN: | FileCheck %s -DMAC=1         \
 // RUN:                -DOSX=0         \
 // RUN:                -DIPHONE=1      \
@@ -101,8 +95,7 @@
 // RUN:                -DLINUX=0       \
 // RUN:                -DUNIX=0
 
-// RUN: %clang -dM -E --target=arm64-apple-watchos \
-// RUN:        -fdefine-target-os-macros %s 2>&1 \
+// RUN: %clang -dM -E --target=arm64-apple-watchos %s 2>&1 \
 // RUN: | FileCheck %s -DMAC=1         \
 // RUN:                -DOSX=0         \
 // RUN:                -DIPHONE=1      \
@@ -117,8 +110,7 @@
 // RUN:                -DLINUX=0       \
 // RUN:                -DUNIX=0
 
-// RUN: %clang -dM -E --target=arm64-apple-watchos-simulator \
-// RUN:        -fdefine-target-os-macros %s 2>&1 \
+// RUN: %clang -dM -E --target=arm64-apple-watchos-simulator %s 2>&1 \
 // RUN: | FileCheck %s -DMAC=1         \
 // RUN:                -DOSX=0         \
 // RUN:                -DIPHONE=1      \
@@ -133,8 +125,7 @@
 // RUN:                -DLINUX=0       \
 // RUN:                -DUNIX=0
 
-// RUN: %clang -dM -E --target=arm64-apple-driverkit \
-// RUN:        -fdefine-target-os-macros %s 2>&1 \
+// RUN: %clang -dM -E --target=arm64-apple-driverkit %s 2>&1 \
 // RUN: | FileCheck %s -DMAC=1         \
 // RUN:                -DOSX=0         \
 // RUN:                -DIPHONE=0      \
diff --git a/clang/test/Driver/fsanitize.c b/clang/test/Driver/fsanitize.c
index 84a8e2b6b203d..1671825042c32 100644
--- a/clang/test/Driver/fsanitize.c
+++ b/clang/test/Driver/fsanitize.c
@@ -896,11 +896,6 @@
 // RUN: %clang --target=riscv64-linux-gnu -fsanitize=scudo %s -### 2>&1 | FileCheck %s --check-prefix=CHECK-SCUDO
 // CHECK-SCUDO: "-fsanitize=scudo"
 
-// RUN: %clang --target=x86_64-linux-gnu -fsanitize=scudo %s -### 2>&1 | FileCheck %s --check-prefix=CHECK-SCUDO-PIE
-// RUN: %clang --target=arm-linux-androideabi -fsanitize=scudo %s -### 2>&1 | FileCheck %s --check-prefix=CHECK-SCUDO-PIE
-// CHECK-SCUDO-PIE: "-mrelocation-model" "pic" "-pic-level" "2" "-pic-is-pie"
-// CHECK-SCUDO-PIE: "-pie"
-
 // RUN: %clang --target=x86_64-linux-gnu -fsanitize=scudo,undefined %s -### 2>&1 | FileCheck %s --check-prefix=CHECK-SCUDO-UBSAN
 // CHECK-SCUDO-UBSAN: "-fsanitize={{.*}}scudo"
 
diff --git a/clang/test/Driver/hurd.cpp b/clang/test/Driver/hurd.cpp
index 1c4ba16af063d..03a74049144c7 100644
--- a/clang/test/Driver/hurd.cpp
+++ b/clang/test/Driver/hurd.cpp
@@ -2,22 +2,22 @@
 
 // RUN: %clang -### %s --target=i686-pc-hurd-gnu --sysroot=%S/Inputs/basic_hurd_tree \
 // RUN:   --stdlib=platform 2>&1 | FileCheck --check-prefix=CHECK %s
-// CHECK: "-cc1"
-// CHECK: "-isysroot" "[[SYSROOT:[^"]+]]"
+// CHECK:      "-cc1"
+// CHECK-SAME: "-isysroot" "[[SYSROOT:[^"]+]]"
 // CHECK-SAME: {{^}} "-internal-isystem" "[[SYSROOT]]/usr/lib/gcc/i686-gnu/10/../../../../include/c++/10"
 /// Debian specific - the path component after 'include' is i386-gnu even
 /// though the installation is i686-gnu.
 // CHECK-SAME: {{^}} "-internal-isystem" "[[SYSROOT]]/usr/lib/gcc/i686-gnu/10/../../../../include/i386-gnu/c++/10"
 // CHECK-SAME: {{^}} "-internal-isystem" "[[SYSROOT]]/usr/lib/gcc/i686-gnu/10/../../../../include/c++/10/backward"
 // CHECK-SAME: {{^}} "-internal-isystem" "[[SYSROOT]]/usr/local/include"
-// CHECK:      "-internal-externc-isystem"
+// CHECK-SAME: "-internal-externc-isystem"
 // CHECK-SAME: {{^}} "[[SYSROOT]]/usr/include/i386-gnu"
 // CHECK-SAME: {{^}} "-internal-externc-isystem" "[[SYSROOT]]/include"
 // CHECK-SAME: {{^}} "-internal-externc-isystem" "[[SYSROOT]]/usr/include"
 // CHECK:      "{{.*}}ld" "--sysroot=[[SYSROOT:[^"]+]]"
-// CHECK:      "-dynamic-linker" "/lib/ld.so"
-// CHECK:      "{{.*}}/usr/lib/gcc/i686-gnu/10/crtbegin.o"
-// CHECK:      "-L
+// CHECK-SAME: "-dynamic-linker" "/lib/ld.so"
+// CHECK-SAME: "{{.*}}/usr/lib/gcc/i686-gnu/10/crtbegin.o"
+// CHECK-SAME: "-L
 // CHECK-SAME: {{^}}[[SYSROOT]]/usr/lib/gcc/i686-gnu/10"
 // CHECK-SAME: {{^}} "-L[[SYSROOT]]/usr/lib/gcc/i686-gnu/10/../../../../lib32"
 // CHECK-SAME: {{^}} "-L[[SYSROOT]]/lib/i386-gnu"
@@ -29,23 +29,23 @@
 
 // RUN: %clang -### %s --target=i686-pc-hurd-gnu --sysroot=%S/Inputs/basic_hurd_tree \
 // RUN:   --stdlib=platform -static 2>&1 | FileCheck --check-prefix=CHECK-STATIC %s
-// CHECK-STATIC: "-cc1"
-// CHECK-STATIC: "-static-define"
-// CHECK-STATIC: "-isysroot" "[[SYSROOT:[^"]+]]"
+// CHECK-STATIC:      "-cc1"
+// CHECK-STATIC-SAME: "-static-define"
+// CHECK-STATIC-SAME: "-isysroot" "[[SYSROOT:[^"]+]]"
 // CHECK-STATIC-SAME: {{^}} "-internal-isystem" "[[SYSROOT]]/usr/lib/gcc/i686-gnu/10/../../../../include/c++/10"
 /// Debian specific - the path component after 'include' is i386-gnu even
 /// though the installation is i686-gnu.
 // CHECK-STATIC-SAME: {{^}} "-internal-isystem" "[[SYSROOT]]/usr/lib/gcc/i686-gnu/10/../../../../include/i386-gnu/c++/10"
 // CHECK-STATIC-SAME: {{^}} "-internal-isystem" "[[SYSROOT]]/usr/lib/gcc/i686-gnu/10/../../../../include/c++/10/backward"
 // CHECK-STATIC-SAME: {{^}} "-internal-isystem" "[[SYSROOT]]/usr/local/include"
-// CHECK-STATIC:      "-internal-externc-isystem"
+// CHECK-STATIC-SAME: "-internal-externc-isystem"
 // CHECK-STATIC-SAME: {{^}} "[[SYSROOT]]/usr/include/i386-gnu"
 // CHECK-STATIC-SAME: {{^}} "-internal-externc-isystem" "[[SYSROOT]]/include"
 // CHECK-STATIC-SAME: {{^}} "-internal-externc-isystem" "[[SYSROOT]]/usr/include"
 // CHECK-STATIC:      "{{.*}}ld" "--sysroot=[[SYSROOT:[^"]+]]"
-// CHECK-STATIC:      "-static"
-// CHECK-STATIC:      "{{.*}}/usr/lib/gcc/i686-gnu/10/crtbeginT.o"
-// CHECK-STATIC:      "-L
+// CHECK-STATIC-SAME: "-static"
+// CHECK-STATIC-SAME: "{{.*}}/usr/lib/gcc/i686-gnu/10/crtbeginT.o"
+// CHECK-STATIC-SAME: "-L
 // CHECK-STATIC-SAME: {{^}}[[SYSROOT]]/usr/lib/gcc/i686-gnu/10"
 // CHECK-STATIC-SAME: {{^}} "-L[[SYSROOT]]/usr/lib/gcc/i686-gnu/10/../../../../lib32"
 // CHECK-STATIC-SAME: {{^}} "-L[[SYSROOT]]/lib/i386-gnu"
@@ -57,9 +57,9 @@
 
 // RUN: %clang -### %s --target=i686-pc-hurd-gnu --sysroot=%S/Inputs/basic_hurd_tree \
 // RUN:   -shared 2>&1 | FileCheck --check-prefix=CHECK-SHARED %s
-// CHECK-SHARED: "{{.*}}ld" "--sysroot=[[SYSROOT:[^"]+]]"
-// CHECK-SHARED: "{{.*}}/usr/lib/gcc/i686-gnu/10/crtbeginS.o"
-// CHECK-SHARED: "-L
+// CHECK-SHARED:      "{{.*}}ld" "--sysroot=[[SYSROOT:[^"]+]]"
+// CHECK-SHARED-SAME: "{{.*}}/usr/lib/gcc/i686-gnu/10/crtbeginS.o"
+// CHECK-SHARED-SAME: "-L
 // CHECK-SHARED-SAME: {{^}}[[SYSROOT]]/usr/lib/gcc/i686-gnu/10"
 // CHECK-SHARED-SAME: {{^}} "-L[[SYSROOT]]/usr/lib/gcc/i686-gnu/10/../../../../lib32"
 // CHECK-SHARED-SAME: {{^}} "-L[[SYSROOT]]/lib/i386-gnu"
@@ -78,3 +78,82 @@
 // CHECK-CROSS: "{{.*}}/Inputs/basic_cross_hurd_tree/usr/lib/gcc/i686-gnu/10/../../../../i686-gnu/bin/ld" {{.*}} "-m" "elf_i386"
 // CHECK-CROSS: "{{.*}}/Inputs/basic_cross_hurd_tree/usr/lib/gcc/i686-gnu/10/crtbegin.o"
 // CHECK-CROSS: "-L{{.*}}/Inputs/basic_cross_hurd_tree/usr/lib/gcc/i686-gnu/10/../../../../i686-gnu/lib"
+
+// RUN: %clang -### %s --target=x86_64-pc-hurd-gnu --sysroot=%S/Inputs/basic_hurd_tree \
+// RUN:   --stdlib=platform 2>&1 | FileCheck --check-prefix=CHECK-64 %s
+// CHECK-64:      "-cc1"
+// CHECK-64-SAME: "-isysroot" "[[SYSROOT:[^"]+]]"
+// CHECK-64-SAME: {{^}} "-internal-isystem" "[[SYSROOT]]/usr/lib/gcc/x86_64-gnu/10/../../../../include/c++/10"
+/// Debian specific - the path component after 'include' is x86_64-gnu even
+/// though the installation is x86_64-gnu.
+// CHECK-64-SAME: {{^}} "-internal-isystem" "[[SYSROOT]]/usr/lib/gcc/x86_64-gnu/10/../../../../include/x86_64-gnu/c++/10"
+// CHECK-64-SAME: {{^}} "-internal-isystem" "[[SYSROOT]]/usr/lib/gcc/x86_64-gnu/10/../../../../include/c++/10/backward"
+// CHECK-64-SAME: {{^}} "-internal-isystem" "[[SYSROOT]]/usr/local/include"
+// CHECK-64-SAME: "-internal-externc-isystem"
+// CHECK-64-SAME: {{^}} "[[SYSROOT]]/usr/include/x86_64-gnu"
+// CHECK-64-SAME: {{^}} "-internal-externc-isystem" "[[SYSROOT]]/include"
+// CHECK-64-SAME: {{^}} "-internal-externc-isystem" "[[SYSROOT]]/usr/include"
+// CHECK-64:      "{{.*}}ld" "--sysroot=[[SYSROOT:[^"]+]]"
+// CHECK-64-SAME: "-dynamic-linker" "/lib/ld-x86-64.so.1"
+// CHECK-64-SAME: "{{.*}}/usr/lib/gcc/x86_64-gnu/10/crtbegin.o"
+// CHECK-64-SAME: "-L
+// CHECK-64-SAME: {{^}}[[SYSROOT]]/usr/lib/gcc/x86_64-gnu/10"
+// CHECK-64-SAME: {{^}} "-L[[SYSROOT]]/usr/lib/gcc/x86_64-gnu/10/../../../../lib64"
+// CHECK-64-SAME: {{^}} "-L[[SYSROOT]]/lib/x86_64-gnu"
+// CHECK-64-SAME: {{^}} "-L[[SYSROOT]]/lib/../lib64"
+// CHECK-64-SAME: {{^}} "-L[[SYSROOT]]/usr/lib/x86_64-gnu"
+// CHECK-64-SAME: {{^}} "-L[[SYSROOT]]/usr/lib/../lib64"
+// CHECK-64-SAME: {{^}} "-L[[SYSROOT]]/lib"
+// CHECK-64-SAME: {{^}} "-L[[SYSROOT]]/usr/lib"
+
+// RUN: %clang -### %s --target=x86_64-pc-hurd-gnu --sysroot=%S/Inputs/basic_hurd_tree \
+// RUN:   --stdlib=platform -static 2>&1 | FileCheck --check-prefix=CHECK-64-STATIC %s
+// CHECK-64-STATIC:      "-cc1"
+// CHECK-64-STATIC-SAME: "-static-define"
+// CHECK-64-STATIC-SAME: "-isysroot" "[[SYSROOT:[^"]+]]"
+// CHECK-64-STATIC-SAME: {{^}} "-internal-isystem" "[[SYSROOT]]/usr/lib/gcc/x86_64-gnu/10/../../../../include/c++/10"
+/// Debian specific - the path component after 'include' is x86_64-gnu even
+/// though the installation is x86_64-gnu.
+// CHECK-64-STATIC-SAME: {{^}} "-internal-isystem" "[[SYSROOT]]/usr/lib/gcc/x86_64-gnu/10/../../../../include/x86_64-gnu/c++/10"
+// CHECK-64-STATIC-SAME: {{^}} "-internal-isystem" "[[SYSROOT]]/usr/lib/gcc/x86_64-gnu/10/../../../../include/c++/10/backward"
+// CHECK-64-STATIC-SAME: {{^}} "-internal-isystem" "[[SYSROOT]]/usr/local/include"
+// CHECK-64-STATIC-SAME: "-internal-externc-isystem"
+// CHECK-64-STATIC-SAME: {{^}} "[[SYSROOT]]/usr/include/x86_64-gnu"
+// CHECK-64-STATIC-SAME: {{^}} "-internal-externc-isystem" "[[SYSROOT]]/include"
+// CHECK-64-STATIC-SAME: {{^}} "-internal-externc-isystem" "[[SYSROOT]]/usr/include"
+// CHECK-64-STATIC:      "{{.*}}ld" "--sysroot=[[SYSROOT:[^"]+]]"
+// CHECK-64-STATIC-SAME: "-static"
+// CHECK-64-STATIC-SAME: "{{.*}}/usr/lib/gcc/x86_64-gnu/10/crtbeginT.o"
+// CHECK-64-STATIC-SAME: "-L
+// CHECK-64-STATIC-SAME: {{^}}[[SYSROOT]]/usr/lib/gcc/x86_64-gnu/10"
+// CHECK-64-STATIC-SAME: {{^}} "-L[[SYSROOT]]/usr/lib/gcc/x86_64-gnu/10/../../../../lib64"
+// CHECK-64-STATIC-SAME: {{^}} "-L[[SYSROOT]]/lib/x86_64-gnu"
+// CHECK-64-STATIC-SAME: {{^}} "-L[[SYSROOT]]/lib/../lib64"
+// CHECK-64-STATIC-SAME: {{^}} "-L[[SYSROOT]]/usr/lib/x86_64-gnu"
+// CHECK-64-STATIC-SAME: {{^}} "-L[[SYSROOT]]/usr/lib/../lib64"
+// CHECK-64-STATIC-SAME: {{^}} "-L[[SYSROOT]]/lib"
+// CHECK-64-STATIC-SAME: {{^}} "-L[[SYSROOT]]/usr/lib"
+
+// RUN: %clang -### %s --target=x86_64-pc-hurd-gnu --sysroot=%S/Inputs/basic_hurd_tree \
+// RUN:   -shared 2>&1 | FileCheck --check-prefix=CHECK-64-SHARED %s
+// CHECK-64-SHARED:      "{{.*}}ld" "--sysroot=[[SYSROOT:[^"]+]]"
+// CHECK-64-SHARED-SAME: "{{.*}}/usr/lib/gcc/x86_64-gnu/10/crtbeginS.o"
+// CHECK-64-SHARED-SAME: "-L
+// CHECK-64-SHARED-SAME: {{^}}[[SYSROOT]]/usr/lib/gcc/x86_64-gnu/10"
+// CHECK-64-SHARED-SAME: {{^}} "-L[[SYSROOT]]/usr/lib/gcc/x86_64-gnu/10/../../../../lib64"
+// CHECK-64-SHARED-SAME: {{^}} "-L[[SYSROOT]]/lib/x86_64-gnu"
+// CHECK-64-SHARED-SAME: {{^}} "-L[[SYSROOT]]/lib/../lib64"
+// CHECK-64-SHARED-SAME: {{^}} "-L[[SYSROOT]]/usr/lib/x86_64-gnu"
+// CHECK-64-SHARED-SAME: {{^}} "-L[[SYSROOT]]/usr/lib/../lib64"
+// CHECK-64-SHARED-SAME: {{^}} "-L[[SYSROOT]]/lib"
+// CHECK-64-SHARED-SAME: {{^}} "-L[[SYSROOT]]/usr/lib"
+
+// RUN: %clang -### -o %t %s 2>&1 -no-integrated-as -fuse-ld=ld \
+// RUN:     --gcc-toolchain=%S/Inputs/basic_cross_hurd_tree/usr \
+// RUN:     --target=x86_64-pc-gnu \
+// RUN:   | FileCheck --check-prefix=CHECK-64-CROSS %s
+// CHECK-64-CROSS: "-cc1" "-triple" "x86_64-pc-hurd-gnu"
+// CHECK-64-CROSS: "{{.*}}/Inputs/basic_cross_hurd_tree/usr/lib/gcc/x86_64-gnu/10/../../../../x86_64-gnu/bin/as" "--64"
+// CHECK-64-CROSS: "{{.*}}/Inputs/basic_cross_hurd_tree/usr/lib/gcc/x86_64-gnu/10/../../../../x86_64-gnu/bin/ld" {{.*}} "-m" "elf_x86_64"
+// CHECK-64-CROSS: "{{.*}}/Inputs/basic_cross_hurd_tree/usr/lib/gcc/x86_64-gnu/10/crtbegin.o"
+// CHECK-64-CROSS: "-L{{.*}}/Inputs/basic_cross_hurd_tree/usr/lib/gcc/x86_64-gnu/10/../../../../x86_64-gnu/lib"
diff --git a/clang/test/Driver/large-data-threshold.c b/clang/test/Driver/large-data-threshold.c
index 9914286701c6c..6500edc0fc0ef 100644
--- a/clang/test/Driver/large-data-threshold.c
+++ b/clang/test/Driver/large-data-threshold.c
@@ -1,8 +1,13 @@
+// RUN: %clang --target=x86_64 -### -c -mcmodel=large -mlarge-data-threshold=200 %s 2>&1 | FileCheck --check-prefix=ARG %s
 // RUN: %clang --target=x86_64 -### -c -mcmodel=medium -mlarge-data-threshold=200 %s 2>&1 | FileCheck --check-prefix=ARG %s
-// RUN: %clang --target=x86_64 -### -c -mcmodel=small -mlarge-data-threshold=200 %s 2>&1 | FileCheck --check-prefix=SMALL %s
+// RUN: %clang --target=x86_64 -### -c -mcmodel=large %s 2>&1 | FileCheck --check-prefix=ARG-LARGE-DEFAULT %s
+// RUN: %clang --target=x86_64 -### -c -mcmodel=medium %s 2>&1 | FileCheck --check-prefix=ARG-MEDIUM-DEFAULT %s
+// RUN: %clang --target=x86_64 -### -c -mcmodel=small -mlarge-data-threshold=200 %s 2>&1 | FileCheck --check-prefix=SMALL --implicit-check-not=mlarge-data-threshold %s
 // RUN: not %clang --target=riscv32 -### -c -mcmodel=medium -mlarge-data-threshold=200 %s 2>&1 | FileCheck --check-prefix=ARCH %s
 
 // ARG: "-mlarge-data-threshold=200"
+// ARG-MEDIUM-DEFAULT: "-mlarge-data-threshold=65536"
+// ARG-LARGE-DEFAULT: "-mlarge-data-threshold=0"
 
-// SMALL: 'mlarge-data-threshold=' only applies to medium code model
-// ARCH: unsupported option 'mlarge-data-threshold=' for target 'riscv32'
+// SMALL: 'mlarge-data-threshold=' only applies to medium and large code models
+// ARCH: unsupported option '-mlarge-data-threshold=' for target
diff --git a/clang/test/Driver/linker-wrapper-image.c b/clang/test/Driver/linker-wrapper-image.c
index da3bbc0f7c878..f9df2de51404f 100644
--- a/clang/test/Driver/linker-wrapper-image.c
+++ b/clang/test/Driver/linker-wrapper-image.c
@@ -44,10 +44,6 @@
 // RUN: clang-linker-wrapper --print-wrapped-module --dry-run --host-triple=x86_64-unknown-windows-gnu \
 // RUN:   --linker-path=/usr/bin/ld -- %t.o -o a.out 2>&1 | FileCheck %s --check-prefixes=CUDA,CUDA-COFF
 
-//      CUDA: @.fatbin_image = internal constant [0 x i8] zeroinitializer, section ".nv_fatbin"
-// CUDA-NEXT: @.fatbin_wrapper = internal constant %fatbin_wrapper { i32 1180844977, i32 1, ptr @.fatbin_image, ptr null }, section ".nvFatBinSegment", align 8
-// CUDA-NEXT: @.cuda.binary_handle = internal global ptr null
-
 //      CUDA-ELF: @__start_cuda_offloading_entries = external hidden constant [0 x %struct.__tgt_offload_entry]
 // CUDA-ELF-NEXT: @__stop_cuda_offloading_entries = external hidden constant [0 x %struct.__tgt_offload_entry]
 // CUDA-ELF-NEXT: @__dummy.cuda_offloading_entries = hidden constant [0 x %struct.__tgt_offload_entry] zeroinitializer, section "cuda_offloading_entries"
@@ -55,6 +51,10 @@
 //      CUDA-COFF: @__start_cuda_offloading_entries = hidden constant [0 x %struct.__tgt_offload_entry] zeroinitializer, section "cuda_offloading_entries$OA"
 // CUDA-COFF-NEXT: @__stop_cuda_offloading_entries = hidden constant [0 x %struct.__tgt_offload_entry] zeroinitializer, section "cuda_offloading_entries$OZ"
 
+//      CUDA: @.fatbin_image = internal constant [0 x i8] zeroinitializer, section ".nv_fatbin"
+// CUDA-NEXT: @.fatbin_wrapper = internal constant %fatbin_wrapper { i32 1180844977, i32 1, ptr @.fatbin_image, ptr null }, section ".nvFatBinSegment", align 8
+// CUDA-NEXT: @.cuda.binary_handle = internal global ptr null
+
 // CUDA: @llvm.global_ctors = appending global [1 x { i32, ptr, ptr }] [{ i32, ptr, ptr } { i32 1, ptr @.cuda.fatbin_reg, ptr null }]
 
 //      CUDA: define internal void @.cuda.fatbin_reg() section ".text.startup" {
@@ -144,10 +144,6 @@
 // RUN: clang-linker-wrapper --print-wrapped-module --dry-run --host-triple=x86_64-unknown-windows-gnu \
 // RUN:   --linker-path=/usr/bin/ld -- %t.o -o a.out 2>&1 | FileCheck %s --check-prefixes=HIP,HIP-COFF
 
-//      HIP: @.fatbin_image = internal constant [0 x i8] zeroinitializer, section ".hip_fatbin"
-// HIP-NEXT: @.fatbin_wrapper = internal constant %fatbin_wrapper { i32 1212764230, i32 1, ptr @.fatbin_image, ptr null }, section ".hipFatBinSegment", align 8
-// HIP-NEXT: @.hip.binary_handle = internal global ptr null
-
 //      HIP-ELF: @__start_hip_offloading_entries = external hidden constant [0 x %struct.__tgt_offload_entry]
 // HIP-ELF-NEXT: @__stop_hip_offloading_entries = external hidden constant [0 x %struct.__tgt_offload_entry]
 // HIP-ELF-NEXT: @__dummy.hip_offloading_entries = hidden constant [0 x %struct.__tgt_offload_entry] zeroinitializer, section "hip_offloading_entries"
@@ -155,6 +151,10 @@
 //      HIP-COFF: @__start_hip_offloading_entries = hidden constant [0 x %struct.__tgt_offload_entry] zeroinitializer, section "hip_offloading_entries$OA"
 // HIP-COFF-NEXT: @__stop_hip_offloading_entries = hidden constant [0 x %struct.__tgt_offload_entry] zeroinitializer, section "hip_offloading_entries$OZ"
 
+//      HIP: @.fatbin_image = internal constant [0 x i8] zeroinitializer, section ".hip_fatbin"
+// HIP-NEXT: @.fatbin_wrapper = internal constant %fatbin_wrapper { i32 1212764230, i32 1, ptr @.fatbin_image, ptr null }, section ".hipFatBinSegment", align 8
+// HIP-NEXT: @.hip.binary_handle = internal global ptr null
+
 // HIP: @llvm.global_ctors = appending global [1 x { i32, ptr, ptr }] [{ i32, ptr, ptr } { i32 1, ptr @.hip.fatbin_reg, ptr null }]
 
 //      HIP: define internal void @.hip.fatbin_reg() section ".text.startup" {
diff --git a/clang/test/Driver/linux-ld.c b/clang/test/Driver/linux-ld.c
index d5cc3103a3a74..b8efd64cd91f0 100644
--- a/clang/test/Driver/linux-ld.c
+++ b/clang/test/Driver/linux-ld.c
@@ -1347,7 +1347,24 @@
 // RUN:     --sysroot=%S/Inputs/basic_android_tree/sysroot \
 // RUN:   | FileCheck --check-prefix=CHECK-ANDROID-PTHREAD-LINK %s
 // CHECK-ANDROID-PTHREAD-LINK-NOT: argument unused during compilation: '-pthread'
-//
+
+/// Check -fandroid-pad-segment.
+// RUN: %clang -### %s --target=aarch64-linux-android -rtlib=platform --unwindlib=platform \
+// RUN:   --gcc-toolchain="" --sysroot=%S/Inputs/basic_android_tree/sysroot \
+// RUN:   -fandroid-pad-segment 2>&1 | FileCheck --check-prefix=CHECK-ANDROID-PAD-PHDR %s
+// CHECK-ANDROID-PAD-PHDR: "{{.*}}ld{{(.exe)?}}" "--sysroot=[[SYSROOT:[^"]+]]"
+// CHECK-ANDROID-PAD-PHDR: "[[SYSROOT]]/usr/lib/crtbegin_dynamic.o" "[[SYSROOT]]/usr/lib/crt_pad_segment.o"
+
+// RUN: %clang -### %s --target=aarch64-linux-android -rtlib=platform --unwindlib=platform \
+// RUN:   --gcc-toolchain="" --sysroot=%S/Inputs/basic_android_tree/sysroot \
+// RUN:   -fandroid-pad-segment -fno-android-pad-segment 2>&1 | FileCheck --check-prefix=CHECK-NO-ANDROID-PAD-PHDR %s
+// CHECK-NO-ANDROID-PAD-PHDR: "{{.*}}ld{{(.exe)?}}" "--sysroot=[[SYSROOT:[^"]+]]"
+// CHECK-NO-ANDROID-PAD-PHDR: "[[SYSROOT]]/usr/lib/crtbegin_dynamic.o"
+// CHECK-NO-ANDROID-PAD-PHDR-NOT: crt_pad_segment.o"
+
+// RUN: not %clang -### %s --target=aarch64-linux -fandroid-pad-segment 2>&1 | FileCheck --check-prefix=ERR-ANDROID-PAD-EHDR %s
+// ERR-ANDROID-PAD-EHDR: error: unsupported option '-fandroid-pad-segment' for target 'aarch64-linux'
+
 // Check linker invocation on a Debian LoongArch sysroot.
 // RUN: %clang -### %s -no-pie 2>&1 \
 // RUN:     --target=loongarch64-linux-gnu -rtlib=platform --unwindlib=platform \
diff --git a/clang/test/Driver/matrix.c b/clang/test/Driver/matrix.c
index 15b44ce5a4ec1..4d2624ad39c16 100644
--- a/clang/test/Driver/matrix.c
+++ b/clang/test/Driver/matrix.c
@@ -6,3 +6,15 @@
 // RUN: %clang -flto=thin -fenable-matrix %t.o -### --target=x86_64-unknown-linux 2>&1 \
 // RUN:     | FileCheck %s -check-prefix=CHECK-THINLTO-MATRIX
 // CHECK-THINLTO-MATRIX: "-plugin-opt=-enable-matrix"
+
+// RUN: %clang -flto -ffat-lto-objects -fenable-matrix %t.o -### --target=x86_64-unknown-linux 2>&1 \
+// RUN:     | FileCheck %s -check-prefix=CHECK-THINLTO-MATRIX
+
+// RUN: %clang -flto=thin -ffat-lto-objects -fenable-matrix %t.o -### --target=x86_64-unknown-linux 2>&1 \
+// RUN:     | FileCheck %s -check-prefix=CHECK-THINLTO-MATRIX
+
+// RUN: %clang -flto -funified-lto -fenable-matrix %t.o -### --target=x86_64-unknown-linux 2>&1 \
+// RUN:     | FileCheck %s -check-prefix=CHECK-THINLTO-MATRIX
+
+// RUN: %clang -flto=thin -funified-lto -fenable-matrix %t.o -### --target=x86_64-unknown-linux 2>&1 \
+// RUN:     | FileCheck %s -check-prefix=CHECK-THINLTO-MATRIX
diff --git a/clang/test/Driver/sanitizer-ld.c b/clang/test/Driver/sanitizer-ld.c
index 2831c2dad8d24..79ce70e19ff4f 100644
--- a/clang/test/Driver/sanitizer-ld.c
+++ b/clang/test/Driver/sanitizer-ld.c
@@ -896,7 +896,6 @@
 // RUN:     --sysroot=%S/Inputs/basic_linux_tree \
 // RUN:   | FileCheck --check-prefix=CHECK-SCUDO-LINUX %s
 // CHECK-SCUDO-LINUX: "{{.*}}ld{{(.exe)?}}"
-// CHECK-SCUDO-LINUX: "-pie"
 // CHECK-SCUDO-LINUX: "--whole-archive" "{{.*}}libclang_rt.scudo_standalone-i386.a" "--no-whole-archive"
 // CHECK-SCUDO-LINUX-NOT: "-lstdc++"
 // CHECK-SCUDO-LINUX: "-lpthread"
diff --git a/clang/test/Driver/xros-driver.c b/clang/test/Driver/xros-driver.c
new file mode 100644
index 0000000000000..dd2c55ed4e70c
--- /dev/null
+++ b/clang/test/Driver/xros-driver.c
@@ -0,0 +1,41 @@
+// RUN: %clang -target arm64-apple-xros1 -c -### %s 2>&1 | FileCheck --check-prefix=VERSION1 %s
+// RUN: %clang -target arm64-apple-xros -c -### %s 2>&1 | FileCheck --check-prefix=VERSION1 %s
+
+// RUN: %clang -target arm64-apple-xros1-simulator -c -### %s 2>&1 | FileCheck --check-prefix=VERSION1_ASi %s
+
+// RUN: not %clang -target arm64-apple-xros1000 -c -### %s 2>&1 | FileCheck --check-prefix=INVALID-VERSION %s
+
+// RUN: %clang -target arm64-apple-xros1  -### %s 2>&1 | FileCheck --check-prefix=LINK %s
+// RUN: %clang -target arm64-apple-xros1-simulator  -### %s 2>&1 | FileCheck --check-prefix=LINK-SIM %s
+
+// RUN: %clang -target arm64-apple-xros1 -c -x objective-c -fobjc-arc -### %s 2>&1 | FileCheck --check-prefix=ARC %s
+// RUN: %clang -target arm64-apple-xros -c -x objective-c -fobjc-arc -### %s 2>&1 | FileCheck --check-prefix=OBJC-RUNTIME %s
+// RUN: %clang -target arm64-apple-xros2 -c -x objective-c -fobjc-arc -### %s 2>&1 | FileCheck --check-prefix=OBJC-RUNTIME2 %s
+
+// RUN: %clang -target arm64-apple-xros1 -c -x objective-c -fobjc-arc -### %s 2>&1 | FileCheck --check-prefix=ARC %s
+// RUN: %clang -target arm64-apple-xros1-simulator -c -x objective-c -fobjc-arc -### %s 2>&1 | FileCheck --check-prefix=ARC %s
+
+// RUN: %clang -target arm64-apple-xros -c -### %s 2>&1 | FileCheck --check-prefix=SSP_ON %s
+
+// RUN: %clang -target arm64e-apple-xros -c -### %s 2>&1 | FileCheck --check-prefix=CPU-ARM64E %s
+// RUN: %clang -target arm64-apple-xros -c -### %s 2>&1 | FileCheck --check-prefix=CPU-ARM64 %s
+// RUN: %clang -target arm64-apple-xros-simulator -c -### %s 2>&1 | FileCheck --check-prefix=CPU-ARM64-SIM %s
+
+// VERSION1: "-cc1"{{.*}} "-triple" "arm64-apple-xros1.0.0"
+// VERSION1_ASi: "-cc1"{{.*}} "-triple" "arm64-apple-xros1.0.0-simulator"
+// INVALID-VERSION: error: invalid version number in
+
+// VERSION1-NOT: -faligned-alloc-unavailable
+
+// LINK: "-platform_version" "xros" "1.0.0" "1.0.0"
+// LINK-SIM: "-platform_version" "xros-simulator" "1.0.0" "1.0.0"
+
+// OBJC-RUNTIME: "-fobjc-runtime=ios-17.0.0.0"
+// OBJC-RUNTIME2: "-fobjc-runtime=ios-18.0.0.0"
+// ARC-NOT: error:
+
+// SSP_ON: "-stack-protector" "1"
+
+// CPU-ARM64E: "-cc1"{{.*}} "-triple" "arm64e-apple-xros1.0.0"{{.*}} "-target-cpu" "apple-a12"{{.*}}
+// CPU-ARM64: "-cc1"{{.*}} "-triple" "arm64-apple-xros1.0.0"{{.*}} "-target-cpu" "apple-a12"
+// CPU-ARM64-SIM: "-cc1"{{.*}} "-triple" "arm64-apple-xros1.0.0-simulator"{{.*}} "-target-cpu" "apple-m1"
diff --git a/clang/test/Format/clang-format-ignore.cpp b/clang/test/Format/clang-format-ignore.cpp
index 5a2267b302d22..b4e526463000a 100644
--- a/clang/test/Format/clang-format-ignore.cpp
+++ b/clang/test/Format/clang-format-ignore.cpp
@@ -1,3 +1,4 @@
+// UNSUPPORTED: system-windows
 // RUN: rm -rf %t.dir
 // RUN: mkdir -p %t.dir/level1/level2
 
@@ -6,41 +7,45 @@
 // RUN: echo "level*/*.c*" >> .clang-format-ignore
 // RUN: echo "*/*2/foo.*" >> .clang-format-ignore
 // RUN: touch foo.cc
-// RUN: clang-format -verbose .clang-format-ignore foo.cc 2> %t.stderr
-// RUN: not grep Formatting %t.stderr
+// RUN: clang-format -verbose .clang-format-ignore foo.cc 2>&1 \
+// RUN:   | FileCheck %s -allow-empty
 
 // RUN: cd level1
 // RUN: touch bar.cc baz.c
-// RUN: clang-format -verbose bar.cc baz.c 2> %t.stderr
-// RUN: not grep Formatting %t.stderr
+// RUN: clang-format -verbose bar.cc baz.c 2>&1 | FileCheck %s -allow-empty
 
 // RUN: cd level2
 // RUN: touch foo.c foo.js
-// RUN: clang-format -verbose foo.c foo.js 2> %t.stderr
-// RUN: not grep Formatting %t.stderr
+// RUN: clang-format -verbose foo.c foo.js 2>&1 | FileCheck %s -allow-empty
+
+// CHECK-NOT: Formatting
 
 // RUN: touch .clang-format-ignore
-// RUN: clang-format -verbose foo.c foo.js 2> %t.stderr
-// RUN: grep -Fx "Formatting [1/2] foo.c" %t.stderr
-// RUN: grep -Fx "Formatting [2/2] foo.js" %t.stderr
+// RUN: clang-format -verbose foo.c foo.js 2>&1 \
+// RUN:   | FileCheck %s -check-prefix=CHECK2 -match-full-lines
+// CHECK2: Formatting [1/2] foo.c
+// CHECK2-NEXT: Formatting [2/2] foo.js
 
 // RUN: echo "*.js" > .clang-format-ignore
-// RUN: clang-format -verbose foo.c foo.js 2> %t.stderr
-// RUN: grep -Fx "Formatting [1/2] foo.c" %t.stderr
-// RUN: not grep -F foo.js %t.stderr
+// RUN: clang-format -verbose foo.c foo.js 2>&1 \
+// RUN:   | FileCheck %s -check-prefix=CHECK3 -match-full-lines
+// CHECK3: Formatting [1/2] foo.c
+// CHECK3-NOT: foo.js
 
 // RUN: cd ../..
-// RUN: clang-format -verbose *.cc level1/*.c* level1/level2/foo.* 2> %t.stderr
-// RUN: grep -x "Formatting \[1/5] .*foo\.c" %t.stderr
-// RUN: not grep -F foo.js %t.stderr
+// RUN: clang-format -verbose *.cc level1/*.c* level1/level2/foo.* 2>&1 \
+// RUN:   | FileCheck %s -check-prefix=CHECK4 -match-full-lines
+// CHECK4: {{Formatting \[1/5] .*foo\.c}}
+// CHECK4-NOT: foo.js
 
 // RUN: rm .clang-format-ignore
-// RUN: clang-format -verbose *.cc level1/*.c* level1/level2/foo.* 2> %t.stderr
-// RUN: grep -x "Formatting \[1/5] .*foo\.cc" %t.stderr
-// RUN: grep -x "Formatting \[2/5] .*bar\.cc" %t.stderr
-// RUN: grep -x "Formatting \[3/5] .*baz\.c" %t.stderr
-// RUN: grep -x "Formatting \[4/5] .*foo\.c" %t.stderr
-// RUN: not grep -F foo.js %t.stderr
+// RUN: clang-format -verbose *.cc level1/*.c* level1/level2/foo.* 2>&1 \
+// RUN:   | FileCheck %s -check-prefix=CHECK5 -match-full-lines
+// CHECK5: {{Formatting \[1/5] .*foo\.cc}}
+// CHECK5-NEXT: {{Formatting \[2/5] .*bar\.cc}}
+// CHECK5-NEXT: {{Formatting \[3/5] .*baz\.c}}
+// CHECK5-NEXT: {{Formatting \[4/5] .*foo\.c}}
+// CHECK5-NOT: foo.js
 
 // RUN: cd ..
 // RUN: rm -r %t.dir
diff --git a/clang/test/Frontend/xros-version.c b/clang/test/Frontend/xros-version.c
new file mode 100644
index 0000000000000..68f69f3de7633
--- /dev/null
+++ b/clang/test/Frontend/xros-version.c
@@ -0,0 +1,3 @@
+// RUN: %clang_cc1 -triple arm64-apple-xros1 -dM -E -o - %s | FileCheck %s
+
+// CHECK: __ENVIRONMENT_OS_VERSION_MIN_REQUIRED__ 10000
diff --git a/clang/test/Interpreter/incremental-mode.cpp b/clang/test/Interpreter/incremental-mode.cpp
index e6350d237ef57..d63cee0dd6d15 100644
--- a/clang/test/Interpreter/incremental-mode.cpp
+++ b/clang/test/Interpreter/incremental-mode.cpp
@@ -1,3 +1,4 @@
 // RUN: clang-repl -Xcc -E
-// RUN: clang-repl -Xcc -emit-llvm 
+// RUN: clang-repl -Xcc -emit-llvm
+// RUN: clang-repl -Xcc -xc
 // expected-no-diagnostics
diff --git a/clang/test/Lexer/SourceLocationsOverflow.c b/clang/test/Lexer/SourceLocationsOverflow.c
index c42f648fedf50..f058c09428e6e 100644
--- a/clang/test/Lexer/SourceLocationsOverflow.c
+++ b/clang/test/Lexer/SourceLocationsOverflow.c
@@ -1,6 +1,6 @@
 // RUN: not %clang %s -S -o - 2>&1 | FileCheck %s
 // CHECK: In file included from {{.*}}SourceLocationsOverflow.c
-// CHECK-NEXT: inc1.h{{.*}}: fatal error: sorry, the translation unit is too large for Clang to process: ran out of source locations
+// CHECK-NEXT: inc1.h{{.*}}: fatal error: translation unit is too large for Clang to process: ran out of source locations
 // CHECK-NEXT: #include "inc2.h"
 // CHECK-NEXT:          ^
 // CHECK-NEXT: note: 214{{.......}}B in local locations, 0B in locations loaded from AST files, for a total of 214{{.......}}B (99% of available space)
diff --git a/clang/test/Lexer/unicode.c b/clang/test/Lexer/unicode.c
index d86ac2d5e2604..909b5b4244438 100644
--- a/clang/test/Lexer/unicode.c
+++ b/clang/test/Lexer/unicode.c
@@ -38,9 +38,10 @@ extern int ༀ;
 extern int 𑩐;
 extern int 𐠈;
 extern int ꙮ;
-extern int  \u1B4C;     // BALINESE LETTER ARCHAIC JNYA - Added in Unicode 14
-extern int  \U00016AA2; // TANGSA LETTER GA - Added in Unicode 14
-extern int  \U0001E4D0; // 𞓐 NAG MUNDARI LETTER O - Added in Unicode 15
+extern int \u1B4C;     // BALINESE LETTER ARCHAIC JNYA - Added in Unicode 14
+extern int \U00016AA2; // TANGSA LETTER GA - Added in Unicode 14
+extern int \U0001E4D0; // 𞓐 NAG MUNDARI LETTER O - Added in Unicode 15
+extern int \u{2EBF0}; // CJK UNIFIED IDEOGRAPH-2EBF0 - Added in Unicode 15.1
 extern int a\N{TANGSA LETTER GA};
 extern int a\N{TANGSALETTERGA}; // expected-error {{'TANGSALETTERGA' is not a valid Unicode character name}} \
                                 // expected-error {{expected ';' after top level declarator}} \
@@ -74,7 +75,7 @@ extern int 👷; // expected-error {{unexpected character }} \
 
 extern int 👷‍♀; // expected-warning {{declaration does not declare anything}} \
                   expected-error {{unexpected character }} \
-                  expected-error {{unexpected character }} \
+                  expected-error {{character  not allowed at the start of an identifier}} \
                   expected-error {{unexpected character }}
 #else
 
diff --git a/clang/test/Modules/GH77953.cpp b/clang/test/Modules/GH77953.cpp
new file mode 100644
index 0000000000000..c0c16f00395bc
--- /dev/null
+++ b/clang/test/Modules/GH77953.cpp
@@ -0,0 +1,28 @@
+// From https://github.com/llvm/llvm-project/issues/77953
+// RUN: rm -rf %t
+// RUN: mkdir -p %t
+// RUN: split-file %s %t
+
+// RUN: %clang_cc1 -std=c++20 -emit-module-interface %t/a.cppm -o %t/a.pcm
+// RUN: %clang_cc1 -std=c++20 -fmodule-file=a=%t/a.pcm %t/b.cpp -fsyntax-only -verify
+
+//--- a.cppm
+export module a;
+
+template
+concept c = true;
+
+export template
+struct a {
+	template requires(... and c)
+	friend bool operator==(a, a) {
+		return true;
+	}
+};
+
+template struct a<>;
+
+//--- b.cpp
+// expected-no-diagnostics
+import a;
+template struct a;
diff --git a/clang/test/Modules/aarch64-sme-keywords.cppm b/clang/test/Modules/aarch64-sme-keywords.cppm
index 6784aaa01d219..df4dd32b16cff 100644
--- a/clang/test/Modules/aarch64-sme-keywords.cppm
+++ b/clang/test/Modules/aarch64-sme-keywords.cppm
@@ -13,8 +13,8 @@ export module A;
 
 export void f_streaming(void) __arm_streaming { }
 export void f_streaming_compatible(void) __arm_streaming_compatible { }
-export void f_shared_za(void) __arm_shared_za { }
-export void f_preserves_za(void) __arm_preserves_za { }
+export void f_shared_za(void) __arm_inout("za") { }
+export void f_preserves_za(void) __arm_preserves("za") { }
 
 //--- Use.cpp
 // expected-no-diagnostics
@@ -50,11 +50,11 @@ import A;
 // CHECK-DAG: attributes #[[STREAMING_DECL]] = {{{.*}} "aarch64_pstate_sm_enabled" {{.*}}}
 // CHECK-DAG: attributes #[[STREAMING_COMPATIBLE_DECL]] = {{{.*}} "aarch64_pstate_sm_compatible" {{.*}}}
 // CHECK-DAG: attributes #[[SHARED_ZA_USE]] = { "aarch64_pstate_za_shared" }
-// CHECK-DAG: attributes #[[PRESERVES_ZA_USE]] = { "aarch64_pstate_za_preserved" }
+// CHECK-DAG: attributes #[[PRESERVES_ZA_USE]] = { "aarch64_pstate_za_preserved" "aarch64_pstate_za_shared" }
 // CHECK-DAG: attributes #[[STREAMING_USE]] = { "aarch64_pstate_sm_enabled" }
 // CHECK-DAG: attributes #[[STREAMING_COMPATIBLE_USE]] = { "aarch64_pstate_sm_compatible" }
 
-void f_shared_za_caller(void) __arm_shared_za {
+void f_shared_za_caller(void) __arm_inout("za") {
   f_shared_za();
   f_preserves_za();
 }
diff --git a/clang/test/Modules/cxx20-10-1-ex1.cpp b/clang/test/Modules/cxx20-10-1-ex1.cpp
index b9a5e8023d035..b330e0a6c9a9d 100644
--- a/clang/test/Modules/cxx20-10-1-ex1.cpp
+++ b/clang/test/Modules/cxx20-10-1-ex1.cpp
@@ -9,13 +9,15 @@
 // RUN:  -o %t/A_Internals.pcm
 
 // RUN: %clang_cc1 -std=c++20 -emit-module-interface %t/std10-1-ex1-tu2.cpp \
-// RUN:  -fmodule-file=%t/A_Internals.pcm -o %t/A_Foo.pcm
+// RUN:  -fmodule-file=A:Internals=%t/A_Internals.pcm -o %t/A_Foo.pcm
 
 // RUN: %clang_cc1 -std=c++20 -emit-module-interface %t/std10-1-ex1-tu3.cpp \
-// RUN:  -fmodule-file=%t/A_Foo.pcm -o %t/A.pcm
+// RUN:  -fmodule-file=A:Foo=%t/A_Foo.pcm -fmodule-file=A:Internals=%t/A_Internals.pcm \
+// RUN:  -o %t/A.pcm
 
 // RUN: %clang_cc1 -std=c++20 -emit-obj %t/std10-1-ex1-tu4.cpp \
-// RUN:  -fmodule-file=%t/A.pcm -o %t/ex1.o
+// RUN:  -fmodule-file=A=%t/A.pcm -fmodule-file=A:Foo=%t/A_Foo.pcm \
+// RUN:  -fmodule-file=A:Internals=%t/A_Internals.pcm -o %t/ex1.o
 
 // expected-no-diagnostics
 
diff --git a/clang/test/Modules/cxx20-importing-function-bodies.cppm b/clang/test/Modules/cxx20-importing-function-bodies.cppm
index f1b8f9a8f0c91..c34e48aaa3f67 100644
--- a/clang/test/Modules/cxx20-importing-function-bodies.cppm
+++ b/clang/test/Modules/cxx20-importing-function-bodies.cppm
@@ -9,7 +9,8 @@
 // RUN: %clang_cc1 -std=c++20 -triple %itanium_abi_triple %t/c.cppm \
 // RUN:     -emit-module-interface -fprebuilt-module-path=%t -o %t/c.pcm
 // RUN: %clang_cc1 -std=c++20 -triple %itanium_abi_triple %t/c.pcm -S \
-// RUN:     -emit-llvm -disable-llvm-passes -o - | FileCheck %t/c.cppm
+// RUN:     -fprebuilt-module-path=%t -emit-llvm -disable-llvm-passes -o - \
+// RUN:     | FileCheck %t/c.cppm
 //
 // Be sure that we keep the same behavior as if optization not enabled.
 // RUN: %clang_cc1 -std=c++20 -triple %itanium_abi_triple -O3 %t/a.cppm \
@@ -19,7 +20,8 @@
 // RUN: %clang_cc1 -std=c++20 -triple %itanium_abi_triple -O3 %t/c.cppm \
 // RUN:     -emit-module-interface -fprebuilt-module-path=%t -o %t/c.pcm
 // RUN: %clang_cc1 -std=c++20 -triple %itanium_abi_triple -O3 %t/c.pcm \
-// RUN:     -S -emit-llvm -disable-llvm-passes -o - | FileCheck %t/c.cppm
+// RUN:     -fprebuilt-module-path=%t -S -emit-llvm -disable-llvm-passes \
+// RUN:     -o - | FileCheck %t/c.cppm
 
 //--- a.cppm
 export module a;
diff --git a/clang/test/Modules/cxx20-module-file-info.cpp b/clang/test/Modules/cxx20-module-file-info.cpp
index 99a215645e8fe..9ef061f076330 100644
--- a/clang/test/Modules/cxx20-module-file-info.cpp
+++ b/clang/test/Modules/cxx20-module-file-info.cpp
@@ -14,13 +14,14 @@
 // RUN:  -o %t/B.pcm
 
 // RUN: %clang_cc1 -std=c++20 -module-file-info %t/B.pcm | FileCheck \
-// RUN:  --check-prefix=CHECK-B %s
+// RUN:     --check-prefix=CHECK-B %s
 
 // RUN: %clang_cc1 -std=c++20 -emit-module-interface %t/mod-info-tu3.cpp \
-// RUN:  -fmodule-file=%t/A.pcm -fmodule-file=%t/B.pcm -o %t/Foo.pcm
+// RUN:     -fmodule-file=A=%t/A.pcm -fmodule-file=B=%t/B.pcm -o %t/Foo.pcm
 
-// RUN: %clang_cc1 -std=c++20 -module-file-info %t/Foo.pcm | FileCheck \
-// RUN:  --check-prefix=CHECK-FOO %s
+// RUN: %clang_cc1 -std=c++20 -module-file-info %t/Foo.pcm -fmodule-file=A=%t/A.pcm \
+// RUN:     -fmodule-file=B=%t/B.pcm  | FileCheck \
+// RUN:     --check-prefix=CHECK-FOO %s
 
 // expected-no-diagnostics
 
diff --git a/clang/test/Modules/cxx20-partition-redeclarations.cpp b/clang/test/Modules/cxx20-partition-redeclarations.cpp
index f99ef009a79af..03658ba92534d 100644
--- a/clang/test/Modules/cxx20-partition-redeclarations.cpp
+++ b/clang/test/Modules/cxx20-partition-redeclarations.cpp
@@ -7,10 +7,10 @@
 // RUN: %clang_cc1 -std=c++20 A-interface.cpp -emit-module-interface \
 // RUN:   -fmodule-file=A-PubPart.pcm -o A.pcm
 
-// RUN: %clang_cc1 -std=c++20 A-impl-top.cpp -fsyntax-only -fmodule-file=A.pcm
-// RUN: %clang_cc1 -std=c++20 A-impl-part.cpp -fsyntax-only -fmodule-file=A.pcm
-// RUN: %clang_cc1 -std=c++20 A-impl-1.cpp -fsyntax-only -fmodule-file=A.pcm
-// RUN: %clang_cc1 -std=c++20 A-impl-2.cpp -fsyntax-only -fmodule-file=A.pcm
+// RUN: %clang_cc1 -std=c++20 A-impl-top.cpp -fsyntax-only -fprebuilt-module-path=%t
+// RUN: %clang_cc1 -std=c++20 A-impl-part.cpp -fsyntax-only -fprebuilt-module-path=%t
+// RUN: %clang_cc1 -std=c++20 A-impl-1.cpp -fsyntax-only -fprebuilt-module-path=%t
+// RUN: %clang_cc1 -std=c++20 A-impl-2.cpp -fsyntax-only -fprebuilt-module-path=%t
 
 //--- A-interface.cpp
 export module A;
diff --git a/clang/test/Modules/eagerly-load-cxx-named-modules.cppm b/clang/test/Modules/eagerly-load-cxx-named-modules.cppm
index febda6ef0f573..ab2ac891fb400 100644
--- a/clang/test/Modules/eagerly-load-cxx-named-modules.cppm
+++ b/clang/test/Modules/eagerly-load-cxx-named-modules.cppm
@@ -7,8 +7,8 @@
 // RUN:    2>&1 | FileCheck %t/user.cpp
 // RUN: %clang_cc1 -std=c++20 %t/b.cppm -emit-module-interface -o %t/b.pcm \
 // RUN:    -fprebuilt-module-path=%t
-// RUN: %clang_cc1 -std=c++20 %t/b.pcm -Wno-read-modules-implicitly -S \
-// RUN:    -emit-llvm 2>&1 -o - | FileCheck %t/b.cppm
+// RUN: %clang_cc1 -std=c++20 %t/b.pcm -S \
+// RUN:    -fprebuilt-module-path=%t -emit-llvm 2>&1 -o - | FileCheck %t/b.cppm
 
 //--- a.cppm
 export module a;
diff --git a/clang/test/Modules/implicit-module-with-missing-path.cpp b/clang/test/Modules/implicit-module-with-missing-path.cpp
index 851a4401df0b6..d167f742e7b21 100644
--- a/clang/test/Modules/implicit-module-with-missing-path.cpp
+++ b/clang/test/Modules/implicit-module-with-missing-path.cpp
@@ -6,7 +6,7 @@
 // RUN: echo -e "export module B;\nimport C;" >> %t/B.cppm
 // RUN: %clang_cc1 -std=c++20 -emit-module-interface %t/subdir/C.cppm -o %t/subdir/C.pcm
 // RUN: %clang_cc1 -std=c++20 -emit-module-interface -fprebuilt-module-path=%t/subdir %t/B.cppm -o %t/B.pcm
-// RUN: %clang_cc1 -std=c++20 -fprebuilt-module-path=%t %s -fsyntax-only -verify -Wno-read-modules-implicitly
+// RUN: %clang_cc1 -std=c++20 -fmodule-file=B=%t/B.pcm %s -fsyntax-only -verify
 
-import B;
-import C; // expected-error {{module 'C' is needed but has not been provided, and implicit use of module files is disabled}}
+import B; // expected-error {{failed to find module file for module 'C'}}
+import C; // expected-error {{module 'C' not found}}
diff --git a/clang/test/Modules/module-init-duplicated-import.cppm b/clang/test/Modules/module-init-duplicated-import.cppm
index de0ce1962f100..7adce11779566 100644
--- a/clang/test/Modules/module-init-duplicated-import.cppm
+++ b/clang/test/Modules/module-init-duplicated-import.cppm
@@ -7,7 +7,7 @@
 // RUN: %clang_cc1 -triple %itanium_abi_triple -std=c++20 %t/m.cppm \
 // RUN:      -emit-module-interface -fmodule-file=a=%t/a.pcm -o %t/m.pcm
 // RUN: %clang_cc1 -triple %itanium_abi_triple -std=c++20 %t/m.pcm  \
-// RUN:     -S -emit-llvm -o - | FileCheck %t/m.cppm
+// RUN:      -fmodule-file=a=%t/a.pcm -S -emit-llvm -o - | FileCheck %t/m.cppm
 
 //--- a.cppm
 export module a;
diff --git a/clang/test/Modules/no-duplicate-codegen-in-GMF.cppm b/clang/test/Modules/no-duplicate-codegen-in-GMF.cppm
index e4d06415fabd4..a743b64cb18d6 100644
--- a/clang/test/Modules/no-duplicate-codegen-in-GMF.cppm
+++ b/clang/test/Modules/no-duplicate-codegen-in-GMF.cppm
@@ -7,7 +7,8 @@
 // RUN: %clang_cc1 -std=c++20 -triple %itanium_abi_triple %t/A.cppm -emit-module-interface -o %t/A.pcm
 // RUN: %clang_cc1 -std=c++20 -triple %itanium_abi_triple %t/B.cppm -emit-module-interface -o %t/B.pcm \
 // RUN:     -fprebuilt-module-path=%t
-// RUN: %clang_cc1 -std=c++20 -triple %itanium_abi_triple %t/B.pcm -S -emit-llvm -o - | FileCheck %t/B.cppm
+// RUN: %clang_cc1 -std=c++20 -triple %itanium_abi_triple %t/B.pcm -S -emit-llvm -o - \
+// RUN:     -fprebuilt-module-path=%t | FileCheck %t/B.cppm
 
 //--- foo.h
 
diff --git a/clang/test/Modules/no-implicit-std-cxx-module.cppm b/clang/test/Modules/no-implicit-std-cxx-module.cppm
index 0432159fed180..2df8b5cf5d848 100644
--- a/clang/test/Modules/no-implicit-std-cxx-module.cppm
+++ b/clang/test/Modules/no-implicit-std-cxx-module.cppm
@@ -6,21 +6,6 @@
 // RUN: %clang_cc1 -std=c++20 -emit-module-interface %t/a.cppm -fmodule-file=b=%t/b.pcm \
 // RUN:     -o %t/a.pcm
 // RUN: %clang_cc1 -std=c++20 %t/user.cpp -fmodule-file=a=%t/a.pcm -verify -fsyntax-only
-// RUN: %clang_cc1 -std=c++20 %t/user.cpp -fmodule-file=a=%t/a.pcm -verify -fsyntax-only \
-// RUN:     -Wno-read-modules-implicitly -DNO_DIAG
-// RUN: %clang_cc1 -std=c++20 %t/user.cpp -fmodule-file=a=%t/a.pcm -fmodule-file=b=%t/b.pcm \
-// RUN:     -DNO_DIAG -verify -fsyntax-only
-//
-// RUN: %clang_cc1 -std=c++20 %t/a.pcm -S -emit-llvm -o - 2>&1 | FileCheck %t/a.cppm
-// RUN: %clang_cc1 -std=c++20 %t/a.pcm -fmodule-file=b=%t/b.pcm -S -emit-llvm -o - 2>&1 \
-// RUN:     | FileCheck %t/a.cppm -check-prefix=CHECK-CORRECT
-//
-// RUN: mkdir -p %t/tmp
-// RUN: mv %t/b.pcm %t/tmp/b.pcm
-// RUN: not %clang_cc1 -std=c++20 %t/a.pcm -S -emit-llvm -o - 2>&1 \
-// RUN:     | FileCheck %t/a.cppm -check-prefix=CHECK-ERROR
-// RUN: %clang_cc1 -std=c++20 %t/a.pcm -S -emit-llvm -o - 2>&1 -fmodule-file=b=%t/tmp/b.pcm \
-// RUN:     | FileCheck %t/a.cppm -check-prefix=CHECK-CORRECT
 
 //--- b.cppm
 export module b;
@@ -35,21 +20,8 @@ export int a() {
     return b() + 43;
 }
 
-// CHECK: it is deprecated to read module 'b' implicitly;
-
-// CHECK-CORRECT-NOT: warning
-// CHECK-CORRECT-NOT: error
-
-// CHECK-ERROR: error: module file{{.*}}not found: module file not found
-
-
 //--- user.cpp
-#ifdef NO_DIAG
-// expected-no-diagnostics
-#else
- // expected-warning@+2 {{it is deprecated to read module 'b' implicitly;}}
-#endif
-import a;
+import a; // expected-error {{failed to find module file for module 'b'}}
 int use() {
-    return a();
+    return a(); // expected-error {{use of undeclared identifier 'a'}}
 }
diff --git a/clang/test/Modules/no-import-func-body.cppm b/clang/test/Modules/no-import-func-body.cppm
index 6c7292b1e469b..af7c3a3ad84b3 100644
--- a/clang/test/Modules/no-import-func-body.cppm
+++ b/clang/test/Modules/no-import-func-body.cppm
@@ -9,7 +9,8 @@
 // RUN: %clang_cc1 -std=c++20 -O1 -triple %itanium_abi_triple %t/c.cppm \
 // RUN:     -emit-module-interface -fprebuilt-module-path=%t -o %t/c.pcm
 // RUN: %clang_cc1 -std=c++20 -O1 -triple %itanium_abi_triple %t/c.pcm -S \
-// RUN:     -emit-llvm -disable-llvm-passes -o - | FileCheck %t/c.cppm
+// RUN:     -fprebuilt-module-path=%t -emit-llvm -disable-llvm-passes -o - \
+// RUN:     | FileCheck %t/c.cppm
 
 //--- a.cppm
 export module a;
diff --git a/clang/test/Modules/pr61067.cppm b/clang/test/Modules/pr61067.cppm
index 8469a1db1f1c8..baee4b83de566 100644
--- a/clang/test/Modules/pr61067.cppm
+++ b/clang/test/Modules/pr61067.cppm
@@ -8,7 +8,7 @@
 // RUN: %clang_cc1 -std=c++20 -triple %itanium_abi_triple %t/b.cppm \
 // RUN:     -emit-module-interface -fmodule-file=a=%t/a.pcm -o %t/b.pcm
 // RUN: %clang_cc1 -std=c++20 -triple %itanium_abi_triple %t/b.pcm -S \
-// RUN:     -emit-llvm -disable-llvm-passes -o - | FileCheck %t/b.cppm
+// RUN:     -emit-llvm -fmodule-file=a=%t/a.pcm -disable-llvm-passes -o - | FileCheck %t/b.cppm
 // RUN: %clang_cc1 -std=c++20 -triple %itanium_abi_triple %t/c.cpp -fmodule-file=a=%t/a.pcm \
 // RUN:     -S -emit-llvm -disable-llvm-passes -o - | FileCheck %t/c.cpp
 
diff --git a/clang/test/Modules/pr62705.cppm b/clang/test/Modules/pr62705.cppm
index a09bdf2563e84..00769d2277f4f 100644
--- a/clang/test/Modules/pr62705.cppm
+++ b/clang/test/Modules/pr62705.cppm
@@ -8,7 +8,7 @@
 // RUN:     -emit-module-interface -o %t/b.pcm \
 // RUN:     -fmodule-file=a=%t/a.pcm
 // RUN: %clang_cc1 %t/b.pcm -std=c++20 -triple %itanium_abi_triple \
-// RUN:     -emit-llvm -o - | FileCheck %t/b.cppm
+// RUN:     -fmodule-file=a=%t/a.pcm -emit-llvm -o - | FileCheck %t/b.cppm
 
 //--- foo.h
 namespace n {
diff --git a/clang/test/Modules/pr67893.cppm b/clang/test/Modules/pr67893.cppm
index 7d4e4c1dc5d84..00b024ecc2eb1 100644
--- a/clang/test/Modules/pr67893.cppm
+++ b/clang/test/Modules/pr67893.cppm
@@ -5,9 +5,9 @@
 // RUN: %clang_cc1 -triple %itanium_abi_triple -std=c++20 %t/a.cppm \
 // RUN:      -emit-module-interface -o %t/a.pcm
 // RUN: %clang_cc1 -triple %itanium_abi_triple -std=c++20 %t/m.cppm \
-// RUN:      -emit-module-interface -fmodule-file=a=%t/a.pcm -o %t/m.pcm
+// RUN:      -emit-module-interface -fprebuilt-module-path=%t
 // RUN: %clang_cc1 -triple %itanium_abi_triple -std=c++20 %t/m.pcm  \
-// RUN:     -S -emit-llvm -o - | FileCheck %t/m.cppm
+// RUN:      -fprebuilt-module-path=%t -S -emit-llvm -o - | FileCheck %t/m.cppm
 
 //--- a.cppm
 export module a;
diff --git a/clang/test/Modules/pr68702.cpp b/clang/test/Modules/pr68702.cpp
new file mode 100644
index 0000000000000..d32f946910f4f
--- /dev/null
+++ b/clang/test/Modules/pr68702.cpp
@@ -0,0 +1,65 @@
+// RUN: rm -rf %t
+// RUN: mkdir %t
+// RUN: split-file %s %t
+
+// RUN: %clang_cc1 -std=c++20 -fmodules -fimplicit-module-maps -fmodules-cache-path=%t %t/main.cpp -o %t/main.o
+
+//--- V.h
+#ifndef V_H
+#define V_H
+
+class A {
+public:
+  constexpr A() { }
+  constexpr ~A() { }
+};
+
+template 
+class V {
+public:
+  V() = default;
+
+  constexpr V(int n, const A& a = A()) {}
+};
+
+#endif
+
+//--- inst1.h
+#include "V.h"
+
+static void inst1() {
+  V v;
+}
+
+//--- inst2.h
+#include "V.h"
+
+static void inst2() {
+  V v(100);
+}
+
+//--- module.modulemap
+module "M" {
+  export *
+  module "V.h" {
+    export *
+    header "V.h"
+  }
+  module "inst1.h" {
+    export *
+    header "inst1.h"
+  }
+}
+
+module "inst2.h" {
+  export *
+  header "inst2.h"
+}
+
+//--- main.cpp
+#include "V.h"
+#include "inst2.h"
+
+static void m() {
+  static V v(100);
+}
diff --git a/clang/test/Modules/pr76638.cppm b/clang/test/Modules/pr76638.cppm
new file mode 100644
index 0000000000000..8cc807961421b
--- /dev/null
+++ b/clang/test/Modules/pr76638.cppm
@@ -0,0 +1,69 @@
+// RUN: rm -rf %t
+// RUN: mkdir -p %t
+// RUN: split-file %s %t
+//
+// RUN: %clang_cc1 -std=c++20 %t/mod1.cppm -emit-module-interface -o %t/mod1.pcm
+// RUN: %clang_cc1 -std=c++20 %t/mod2.cppm -fmodule-file=mod1=%t/mod1.pcm \
+// RUN:     -fsyntax-only -verify
+//
+// RUN: %clang_cc1 -std=c++20 %t/mod3.cppm -emit-module-interface -o %t/mod3.pcm
+// RUN: %clang_cc1 -std=c++20 %t/mod4.cppm -fmodule-file=mod3=%t/mod3.pcm \
+// RUN:     -fsyntax-only -verify
+
+//--- size_t.h
+
+extern "C" {
+    typedef unsigned int size_t;
+}
+
+//--- csize_t
+namespace std {
+            using :: size_t;
+}
+
+//--- align.h
+namespace std {
+    enum class align_val_t : size_t {};
+}
+
+//--- mod1.cppm
+module;
+#include "size_t.h"
+#include "align.h"
+export module mod1;
+export using std::align_val_t;
+
+//--- mod2.cppm
+// expected-no-diagnostics
+module;
+#include "size_t.h"
+#include "csize_t"
+#include "align.h"
+export module mod2;
+import mod1;
+export using std::align_val_t;
+
+//--- signed_size_t.h
+// Test that we can still find the case if the underlying type is different
+extern "C" {
+    typedef signed int size_t;
+}
+
+//--- mod3.cppm
+module;
+#include "size_t.h"
+#include "align.h"
+export module mod3;
+export using std::align_val_t;
+
+//--- mod4.cppm
+module;
+#include "signed_size_t.h"
+#include "csize_t"
+#include "align.h"
+export module mod4;
+import mod3;
+export using std::align_val_t;
+
+// expected-error@align.h:* {{'std::align_val_t' has different definitions in different modules; defined here first difference is enum with specified type 'size_t' (aka 'int')}}
+// expected-note@align.h:* {{but in 'mod3.' found enum with specified type 'size_t' (aka 'unsigned int')}}
diff --git a/clang/test/Modules/pr77995.cppm b/clang/test/Modules/pr77995.cppm
new file mode 100644
index 0000000000000..26442dacac4fb
--- /dev/null
+++ b/clang/test/Modules/pr77995.cppm
@@ -0,0 +1,25 @@
+// RUN: rm -rf %t
+// RUN: mkdir -p %t
+// RUN: split-file %s %t
+//
+// RUN: %clang_cc1 -std=c++20 %t/foo.cppm -emit-module-interface -o %t/foo.pcm
+// RUN: %clang_cc1 -std=c++20 %t/use.cpp -fmodule-file=foo=%t/foo.pcm -verify -fsyntax-only
+
+//--- a.hpp
+#pragma once
+#define A 43
+
+//--- foo.cppm
+module;
+#include "a.hpp"
+export module foo;
+export constexpr auto B = A;
+
+//--- use.cpp
+// expected-no-diagnostics
+import foo;
+#include "a.hpp"
+
+static_assert(A == 43);
+static_assert(B == 43);
+
diff --git a/clang/test/Modules/templates.mm b/clang/test/Modules/templates.mm
index 0e66c9843cc09..d7b7d5af030a7 100644
--- a/clang/test/Modules/templates.mm
+++ b/clang/test/Modules/templates.mm
@@ -125,7 +125,7 @@ void testWithAttributes() {
 
 // Check that returnNonTrivial doesn't return Class0 directly in registers.
 
-// CHECK: declare void @_Z16returnNonTrivialv(ptr sret(%struct.Class0) align 8)
+// CHECK: declare void @_Z16returnNonTrivialv(ptr dead_on_unwind writable sret(%struct.Class0) align 8)
 
 @import template_nontrivial0;
 @import template_nontrivial1;
diff --git a/clang/test/OpenMP/irbuilder_for_iterator.cpp b/clang/test/OpenMP/irbuilder_for_iterator.cpp
index 91bbb6fc18ddb..b88416b36c4fa 100644
--- a/clang/test/OpenMP/irbuilder_for_iterator.cpp
+++ b/clang/test/OpenMP/irbuilder_for_iterator.cpp
@@ -159,7 +159,7 @@ extern "C" void workshareloop_iterator(float *a, float *b, float *c) {
 // CHECK-NEXT:    [[TMP2:%.*]] = load i64, ptr [[LOGICAL_ADDR]], align 8
 // CHECK-NEXT:    [[MUL:%.*]] = mul i64 1, [[TMP2]]
 // CHECK-NEXT:    [[CONV:%.*]] = trunc i64 [[MUL]] to i32
-// CHECK-NEXT:    call void @_ZNK10MyIteratorplEj(ptr sret([[STRUCT_MYITERATOR]]) align 1 [[REF_TMP]], ptr noundef nonnull align 1 dereferenceable(1) [[TMP1]], i32 noundef [[CONV]])
+// CHECK-NEXT:    call void @_ZNK10MyIteratorplEj(ptr dead_on_unwind writable sret([[STRUCT_MYITERATOR]]) align 1 [[REF_TMP]], ptr noundef nonnull align 1 dereferenceable(1) [[TMP1]], i32 noundef [[CONV]])
 // CHECK-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[LOOPVAR_ADDR]], align 8
 // CHECK-NEXT:    [[CALL:%.*]] = call noundef nonnull align 1 dereferenceable(1) ptr @_ZN10MyIteratoraSERKS_(ptr noundef nonnull align 1 dereferenceable(1) [[TMP3]], ptr noundef nonnull align 1 dereferenceable(1) [[REF_TMP]])
 // CHECK-NEXT:    ret void
diff --git a/clang/test/OpenMP/irbuilder_for_rangefor.cpp b/clang/test/OpenMP/irbuilder_for_rangefor.cpp
index 3b952d43eb257..6bf91bfda138a 100644
--- a/clang/test/OpenMP/irbuilder_for_rangefor.cpp
+++ b/clang/test/OpenMP/irbuilder_for_rangefor.cpp
@@ -57,9 +57,9 @@ extern "C" void workshareloop_rangefor(float *a, float *b, float *c) {
 // CHECK-NEXT:    call void @_ZN7MyRangeC1Ei(ptr noundef nonnull align 1 dereferenceable(1) [[REF_TMP]], i32 noundef 42)
 // CHECK-NEXT:    store ptr [[REF_TMP]], ptr [[__RANGE2]], align 8
 // CHECK-NEXT:    [[TMP0:%.*]] = load ptr, ptr [[__RANGE2]], align 8
-// CHECK-NEXT:    call void @_ZN7MyRange5beginEv(ptr sret([[STRUCT_MYITERATOR]]) align 1 [[__BEGIN2]], ptr noundef nonnull align 1 dereferenceable(1) [[TMP0]])
+// CHECK-NEXT:    call void @_ZN7MyRange5beginEv(ptr dead_on_unwind writable sret([[STRUCT_MYITERATOR]]) align 1 [[__BEGIN2]], ptr noundef nonnull align 1 dereferenceable(1) [[TMP0]])
 // CHECK-NEXT:    [[TMP1:%.*]] = load ptr, ptr [[__RANGE2]], align 8
-// CHECK-NEXT:    call void @_ZN7MyRange3endEv(ptr sret([[STRUCT_MYITERATOR]]) align 1 [[__END2]], ptr noundef nonnull align 1 dereferenceable(1) [[TMP1]])
+// CHECK-NEXT:    call void @_ZN7MyRange3endEv(ptr dead_on_unwind writable sret([[STRUCT_MYITERATOR]]) align 1 [[__END2]], ptr noundef nonnull align 1 dereferenceable(1) [[TMP1]])
 // CHECK-NEXT:    [[CALL:%.*]] = call noundef i32 @_ZNK10MyIteratordeEv(ptr noundef nonnull align 1 dereferenceable(1) [[__BEGIN2]])
 // CHECK-NEXT:    store i32 [[CALL]], ptr [[I]], align 4
 // CHECK-NEXT:    [[TMP2:%.*]] = getelementptr inbounds [[STRUCT_ANON]], ptr [[AGG_CAPTURED]], i32 0, i32 0
@@ -177,7 +177,7 @@ extern "C" void workshareloop_rangefor(float *a, float *b, float *c) {
 // CHECK-NEXT:    [[TMP2:%.*]] = load i64, ptr [[LOGICAL_ADDR]], align 8
 // CHECK-NEXT:    [[MUL:%.*]] = mul i64 1, [[TMP2]]
 // CHECK-NEXT:    [[CONV:%.*]] = trunc i64 [[MUL]] to i32
-// CHECK-NEXT:    call void @_ZNK10MyIteratorplEj(ptr sret([[STRUCT_MYITERATOR]]) align 1 [[REF_TMP]], ptr noundef nonnull align 1 dereferenceable(1) [[TMP1]], i32 noundef [[CONV]])
+// CHECK-NEXT:    call void @_ZNK10MyIteratorplEj(ptr dead_on_unwind writable sret([[STRUCT_MYITERATOR]]) align 1 [[REF_TMP]], ptr noundef nonnull align 1 dereferenceable(1) [[TMP1]], i32 noundef [[CONV]])
 // CHECK-NEXT:    [[CALL:%.*]] = call noundef i32 @_ZNK10MyIteratordeEv(ptr noundef nonnull align 1 dereferenceable(1) [[REF_TMP]])
 // CHECK-NEXT:    [[TMP3:%.*]] = load ptr, ptr [[LOOPVAR_ADDR]], align 8
 // CHECK-NEXT:    store i32 [[CALL]], ptr [[TMP3]], align 4
diff --git a/clang/test/OpenMP/master_taskloop_in_reduction_codegen.cpp b/clang/test/OpenMP/master_taskloop_in_reduction_codegen.cpp
index 7afaf035988f0..41c086c7e1d06 100644
--- a/clang/test/OpenMP/master_taskloop_in_reduction_codegen.cpp
+++ b/clang/test/OpenMP/master_taskloop_in_reduction_codegen.cpp
@@ -339,7 +339,7 @@ int main(int argc, char **argv) {
 // CHECK1:       omp.arraycpy.body:
 // CHECK1-NEXT:    [[OMP_ARRAYCPY_SRCELEMENTPAST:%.*]] = phi ptr [ [[TMP3]], [[ENTRY:%.*]] ], [ [[OMP_ARRAYCPY_SRC_ELEMENT:%.*]], [[OMP_ARRAYCPY_BODY]] ]
 // CHECK1-NEXT:    [[OMP_ARRAYCPY_DESTELEMENTPAST:%.*]] = phi ptr [ [[TMP2]], [[ENTRY]] ], [ [[OMP_ARRAYCPY_DEST_ELEMENT:%.*]], [[OMP_ARRAYCPY_BODY]] ]
-// CHECK1-NEXT:    call void @_ZplRK1SS1_(ptr sret([[STRUCT_S]]) align 4 [[REF_TMP]], ptr nonnull align 4 dereferenceable(4) [[OMP_ARRAYCPY_DESTELEMENTPAST]], ptr nonnull align 4 dereferenceable(4) [[OMP_ARRAYCPY_SRCELEMENTPAST]])
+// CHECK1-NEXT:    call void @_ZplRK1SS1_(ptr dead_on_unwind writable sret([[STRUCT_S]]) align 4 [[REF_TMP]], ptr nonnull align 4 dereferenceable(4) [[OMP_ARRAYCPY_DESTELEMENTPAST]], ptr nonnull align 4 dereferenceable(4) [[OMP_ARRAYCPY_SRCELEMENTPAST]])
 // CHECK1-NEXT:    [[CALL:%.*]] = call nonnull align 4 dereferenceable(4) ptr @_ZN1SaSERKS_(ptr nonnull align 4 dereferenceable(4) [[OMP_ARRAYCPY_DESTELEMENTPAST]], ptr nonnull align 4 dereferenceable(4) [[REF_TMP]])
 // CHECK1-NEXT:    call void @_ZN1SD1Ev(ptr nonnull align 4 dereferenceable(4) [[REF_TMP]]) #[[ATTR3]]
 // CHECK1-NEXT:    [[OMP_ARRAYCPY_DEST_ELEMENT]] = getelementptr [[STRUCT_S]], ptr [[OMP_ARRAYCPY_DESTELEMENTPAST]], i32 1
@@ -351,7 +351,7 @@ int main(int argc, char **argv) {
 //
 //
 // CHECK1-LABEL: define {{[^@]+}}@_ZplRK1SS1_
-// CHECK1-SAME: (ptr noalias sret([[STRUCT_S:%.*]]) align 4 [[AGG_RESULT:%.*]], ptr nonnull align 4 dereferenceable(4) [[A:%.*]], ptr nonnull align 4 dereferenceable(4) [[B:%.*]]) #[[ATTR1]] {
+// CHECK1-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_S:%.*]]) align 4 [[AGG_RESULT:%.*]], ptr nonnull align 4 dereferenceable(4) [[A:%.*]], ptr nonnull align 4 dereferenceable(4) [[B:%.*]]) #[[ATTR1]] {
 // CHECK1-NEXT:  entry:
 // CHECK1-NEXT:    [[RESULT_PTR:%.*]] = alloca ptr, align 8
 // CHECK1-NEXT:    [[A_ADDR:%.*]] = alloca ptr, align 8
diff --git a/clang/test/OpenMP/master_taskloop_simd_in_reduction_codegen.cpp b/clang/test/OpenMP/master_taskloop_simd_in_reduction_codegen.cpp
index 5d6404cd3e54b..f25fe346c835e 100644
--- a/clang/test/OpenMP/master_taskloop_simd_in_reduction_codegen.cpp
+++ b/clang/test/OpenMP/master_taskloop_simd_in_reduction_codegen.cpp
@@ -339,7 +339,7 @@ int main(int argc, char **argv) {
 // CHECK1:       omp.arraycpy.body:
 // CHECK1-NEXT:    [[OMP_ARRAYCPY_SRCELEMENTPAST:%.*]] = phi ptr [ [[TMP3]], [[ENTRY:%.*]] ], [ [[OMP_ARRAYCPY_SRC_ELEMENT:%.*]], [[OMP_ARRAYCPY_BODY]] ]
 // CHECK1-NEXT:    [[OMP_ARRAYCPY_DESTELEMENTPAST:%.*]] = phi ptr [ [[TMP2]], [[ENTRY]] ], [ [[OMP_ARRAYCPY_DEST_ELEMENT:%.*]], [[OMP_ARRAYCPY_BODY]] ]
-// CHECK1-NEXT:    call void @_ZplRK1SS1_(ptr sret([[STRUCT_S]]) align 4 [[REF_TMP]], ptr nonnull align 4 dereferenceable(4) [[OMP_ARRAYCPY_DESTELEMENTPAST]], ptr nonnull align 4 dereferenceable(4) [[OMP_ARRAYCPY_SRCELEMENTPAST]])
+// CHECK1-NEXT:    call void @_ZplRK1SS1_(ptr dead_on_unwind writable sret([[STRUCT_S]]) align 4 [[REF_TMP]], ptr nonnull align 4 dereferenceable(4) [[OMP_ARRAYCPY_DESTELEMENTPAST]], ptr nonnull align 4 dereferenceable(4) [[OMP_ARRAYCPY_SRCELEMENTPAST]])
 // CHECK1-NEXT:    [[CALL:%.*]] = call nonnull align 4 dereferenceable(4) ptr @_ZN1SaSERKS_(ptr nonnull align 4 dereferenceable(4) [[OMP_ARRAYCPY_DESTELEMENTPAST]], ptr nonnull align 4 dereferenceable(4) [[REF_TMP]])
 // CHECK1-NEXT:    call void @_ZN1SD1Ev(ptr nonnull align 4 dereferenceable(4) [[REF_TMP]]) #[[ATTR3]]
 // CHECK1-NEXT:    [[OMP_ARRAYCPY_DEST_ELEMENT]] = getelementptr [[STRUCT_S]], ptr [[OMP_ARRAYCPY_DESTELEMENTPAST]], i32 1
@@ -351,7 +351,7 @@ int main(int argc, char **argv) {
 //
 //
 // CHECK1-LABEL: define {{[^@]+}}@_ZplRK1SS1_
-// CHECK1-SAME: (ptr noalias sret([[STRUCT_S:%.*]]) align 4 [[AGG_RESULT:%.*]], ptr nonnull align 4 dereferenceable(4) [[A:%.*]], ptr nonnull align 4 dereferenceable(4) [[B:%.*]]) #[[ATTR1]] {
+// CHECK1-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_S:%.*]]) align 4 [[AGG_RESULT:%.*]], ptr nonnull align 4 dereferenceable(4) [[A:%.*]], ptr nonnull align 4 dereferenceable(4) [[B:%.*]]) #[[ATTR1]] {
 // CHECK1-NEXT:  entry:
 // CHECK1-NEXT:    [[RESULT_PTR:%.*]] = alloca ptr, align 8
 // CHECK1-NEXT:    [[A_ADDR:%.*]] = alloca ptr, align 8
diff --git a/clang/test/OpenMP/target_in_reduction_codegen.cpp b/clang/test/OpenMP/target_in_reduction_codegen.cpp
index 2eb6058461244..c6b8afba52ee5 100644
--- a/clang/test/OpenMP/target_in_reduction_codegen.cpp
+++ b/clang/test/OpenMP/target_in_reduction_codegen.cpp
@@ -333,7 +333,7 @@ int main(int argc, char **argv) {
 // CHECK1:       omp.arraycpy.body:
 // CHECK1-NEXT:    [[OMP_ARRAYCPY_SRCELEMENTPAST:%.*]] = phi ptr [ [[TMP5]], [[ENTRY:%.*]] ], [ [[OMP_ARRAYCPY_SRC_ELEMENT:%.*]], [[OMP_ARRAYCPY_BODY]] ]
 // CHECK1-NEXT:    [[OMP_ARRAYCPY_DESTELEMENTPAST:%.*]] = phi ptr [ [[TMP3]], [[ENTRY]] ], [ [[OMP_ARRAYCPY_DEST_ELEMENT:%.*]], [[OMP_ARRAYCPY_BODY]] ]
-// CHECK1-NEXT:    call void @_ZplRK1SS1_(ptr sret([[STRUCT_S]]) align 4 [[REF_TMP]], ptr nonnull align 4 dereferenceable(4) [[OMP_ARRAYCPY_DESTELEMENTPAST]], ptr nonnull align 4 dereferenceable(4) [[OMP_ARRAYCPY_SRCELEMENTPAST]])
+// CHECK1-NEXT:    call void @_ZplRK1SS1_(ptr dead_on_unwind writable sret([[STRUCT_S]]) align 4 [[REF_TMP]], ptr nonnull align 4 dereferenceable(4) [[OMP_ARRAYCPY_DESTELEMENTPAST]], ptr nonnull align 4 dereferenceable(4) [[OMP_ARRAYCPY_SRCELEMENTPAST]])
 // CHECK1-NEXT:    [[CALL:%.*]] = call nonnull align 4 dereferenceable(4) ptr @_ZN1SaSERKS_(ptr nonnull align 4 dereferenceable(4) [[OMP_ARRAYCPY_DESTELEMENTPAST]], ptr nonnull align 4 dereferenceable(4) [[REF_TMP]])
 // CHECK1-NEXT:    call void @_ZN1SD1Ev(ptr nonnull align 4 dereferenceable(4) [[REF_TMP]]) #3
 // CHECK1-NEXT:    [[OMP_ARRAYCPY_DEST_ELEMENT]] = getelementptr [[STRUCT_S]], ptr [[OMP_ARRAYCPY_DESTELEMENTPAST]], i32 1
@@ -345,7 +345,7 @@ int main(int argc, char **argv) {
 //
 //
 // CHECK1-LABEL: define {{[^@]+}}@_ZplRK1SS1_
-// CHECK1-SAME: (ptr noalias sret([[STRUCT_S:%.*]]) align 4 [[AGG_RESULT:%.*]], ptr nonnull align 4 dereferenceable(4) [[A:%.*]], ptr nonnull align 4 dereferenceable(4) [[B:%.*]]) #[[ATTR7:[0-9]+]] {
+// CHECK1-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_S:%.*]]) align 4 [[AGG_RESULT:%.*]], ptr nonnull align 4 dereferenceable(4) [[A:%.*]], ptr nonnull align 4 dereferenceable(4) [[B:%.*]]) #[[ATTR7:[0-9]+]] {
 // CHECK1-NEXT:  entry:
 // CHECK1-NEXT:    [[RESULT_PTR:%.*]] = alloca ptr, align 8
 // CHECK1-NEXT:    [[A_ADDR:%.*]] = alloca ptr, align 8
diff --git a/clang/test/OpenMP/task_in_reduction_codegen.cpp b/clang/test/OpenMP/task_in_reduction_codegen.cpp
index 82d2038943646..348bd7bea3b4a 100644
--- a/clang/test/OpenMP/task_in_reduction_codegen.cpp
+++ b/clang/test/OpenMP/task_in_reduction_codegen.cpp
@@ -361,7 +361,7 @@ int main(int argc, char **argv) {
 // CHECK1:       omp.arraycpy.body:
 // CHECK1-NEXT:    [[OMP_ARRAYCPY_SRCELEMENTPAST:%.*]] = phi ptr [ [[TMP3]], [[ENTRY:%.*]] ], [ [[OMP_ARRAYCPY_SRC_ELEMENT:%.*]], [[OMP_ARRAYCPY_BODY]] ]
 // CHECK1-NEXT:    [[OMP_ARRAYCPY_DESTELEMENTPAST:%.*]] = phi ptr [ [[TMP2]], [[ENTRY]] ], [ [[OMP_ARRAYCPY_DEST_ELEMENT:%.*]], [[OMP_ARRAYCPY_BODY]] ]
-// CHECK1-NEXT:    call void @_ZplRK1SS1_(ptr sret([[STRUCT_S]]) align 4 [[REF_TMP]], ptr nonnull align 4 dereferenceable(4) [[OMP_ARRAYCPY_DESTELEMENTPAST]], ptr nonnull align 4 dereferenceable(4) [[OMP_ARRAYCPY_SRCELEMENTPAST]])
+// CHECK1-NEXT:    call void @_ZplRK1SS1_(ptr dead_on_unwind writable sret([[STRUCT_S]]) align 4 [[REF_TMP]], ptr nonnull align 4 dereferenceable(4) [[OMP_ARRAYCPY_DESTELEMENTPAST]], ptr nonnull align 4 dereferenceable(4) [[OMP_ARRAYCPY_SRCELEMENTPAST]])
 // CHECK1-NEXT:    [[CALL:%.*]] = call nonnull align 4 dereferenceable(4) ptr @_ZN1SaSERKS_(ptr nonnull align 4 dereferenceable(4) [[OMP_ARRAYCPY_DESTELEMENTPAST]], ptr nonnull align 4 dereferenceable(4) [[REF_TMP]])
 // CHECK1-NEXT:    call void @_ZN1SD1Ev(ptr nonnull align 4 dereferenceable(4) [[REF_TMP]]) #[[ATTR3]]
 // CHECK1-NEXT:    [[OMP_ARRAYCPY_DEST_ELEMENT]] = getelementptr [[STRUCT_S]], ptr [[OMP_ARRAYCPY_DESTELEMENTPAST]], i32 1
@@ -373,7 +373,7 @@ int main(int argc, char **argv) {
 //
 //
 // CHECK1-LABEL: define {{[^@]+}}@_ZplRK1SS1_
-// CHECK1-SAME: (ptr noalias sret([[STRUCT_S:%.*]]) align 4 [[AGG_RESULT:%.*]], ptr nonnull align 4 dereferenceable(4) [[A:%.*]], ptr nonnull align 4 dereferenceable(4) [[B:%.*]]) #[[ATTR1]] {
+// CHECK1-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_S:%.*]]) align 4 [[AGG_RESULT:%.*]], ptr nonnull align 4 dereferenceable(4) [[A:%.*]], ptr nonnull align 4 dereferenceable(4) [[B:%.*]]) #[[ATTR1]] {
 // CHECK1-NEXT:  entry:
 // CHECK1-NEXT:    [[RESULT_PTR:%.*]] = alloca ptr, align 8
 // CHECK1-NEXT:    [[A_ADDR:%.*]] = alloca ptr, align 8
diff --git a/clang/test/OpenMP/taskloop_in_reduction_codegen.cpp b/clang/test/OpenMP/taskloop_in_reduction_codegen.cpp
index e48035a71eea9..abf9164b789d3 100644
--- a/clang/test/OpenMP/taskloop_in_reduction_codegen.cpp
+++ b/clang/test/OpenMP/taskloop_in_reduction_codegen.cpp
@@ -339,7 +339,7 @@ int main(int argc, char **argv) {
 // CHECK1:       omp.arraycpy.body:
 // CHECK1-NEXT:    [[OMP_ARRAYCPY_SRCELEMENTPAST:%.*]] = phi ptr [ [[TMP3]], [[ENTRY:%.*]] ], [ [[OMP_ARRAYCPY_SRC_ELEMENT:%.*]], [[OMP_ARRAYCPY_BODY]] ]
 // CHECK1-NEXT:    [[OMP_ARRAYCPY_DESTELEMENTPAST:%.*]] = phi ptr [ [[TMP2]], [[ENTRY]] ], [ [[OMP_ARRAYCPY_DEST_ELEMENT:%.*]], [[OMP_ARRAYCPY_BODY]] ]
-// CHECK1-NEXT:    call void @_ZplRK1SS1_(ptr sret([[STRUCT_S]]) align 4 [[REF_TMP]], ptr nonnull align 4 dereferenceable(4) [[OMP_ARRAYCPY_DESTELEMENTPAST]], ptr nonnull align 4 dereferenceable(4) [[OMP_ARRAYCPY_SRCELEMENTPAST]])
+// CHECK1-NEXT:    call void @_ZplRK1SS1_(ptr dead_on_unwind writable sret([[STRUCT_S]]) align 4 [[REF_TMP]], ptr nonnull align 4 dereferenceable(4) [[OMP_ARRAYCPY_DESTELEMENTPAST]], ptr nonnull align 4 dereferenceable(4) [[OMP_ARRAYCPY_SRCELEMENTPAST]])
 // CHECK1-NEXT:    [[CALL:%.*]] = call nonnull align 4 dereferenceable(4) ptr @_ZN1SaSERKS_(ptr nonnull align 4 dereferenceable(4) [[OMP_ARRAYCPY_DESTELEMENTPAST]], ptr nonnull align 4 dereferenceable(4) [[REF_TMP]])
 // CHECK1-NEXT:    call void @_ZN1SD1Ev(ptr nonnull align 4 dereferenceable(4) [[REF_TMP]]) #[[ATTR3]]
 // CHECK1-NEXT:    [[OMP_ARRAYCPY_DEST_ELEMENT]] = getelementptr [[STRUCT_S]], ptr [[OMP_ARRAYCPY_DESTELEMENTPAST]], i32 1
@@ -351,7 +351,7 @@ int main(int argc, char **argv) {
 //
 //
 // CHECK1-LABEL: define {{[^@]+}}@_ZplRK1SS1_
-// CHECK1-SAME: (ptr noalias sret([[STRUCT_S:%.*]]) align 4 [[AGG_RESULT:%.*]], ptr nonnull align 4 dereferenceable(4) [[A:%.*]], ptr nonnull align 4 dereferenceable(4) [[B:%.*]]) #[[ATTR1]] {
+// CHECK1-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_S:%.*]]) align 4 [[AGG_RESULT:%.*]], ptr nonnull align 4 dereferenceable(4) [[A:%.*]], ptr nonnull align 4 dereferenceable(4) [[B:%.*]]) #[[ATTR1]] {
 // CHECK1-NEXT:  entry:
 // CHECK1-NEXT:    [[RESULT_PTR:%.*]] = alloca ptr, align 8
 // CHECK1-NEXT:    [[A_ADDR:%.*]] = alloca ptr, align 8
diff --git a/clang/test/OpenMP/taskloop_simd_in_reduction_codegen.cpp b/clang/test/OpenMP/taskloop_simd_in_reduction_codegen.cpp
index e4d24fa482440..f279596401a8e 100644
--- a/clang/test/OpenMP/taskloop_simd_in_reduction_codegen.cpp
+++ b/clang/test/OpenMP/taskloop_simd_in_reduction_codegen.cpp
@@ -339,7 +339,7 @@ int main(int argc, char **argv) {
 // CHECK1:       omp.arraycpy.body:
 // CHECK1-NEXT:    [[OMP_ARRAYCPY_SRCELEMENTPAST:%.*]] = phi ptr [ [[TMP3]], [[ENTRY:%.*]] ], [ [[OMP_ARRAYCPY_SRC_ELEMENT:%.*]], [[OMP_ARRAYCPY_BODY]] ]
 // CHECK1-NEXT:    [[OMP_ARRAYCPY_DESTELEMENTPAST:%.*]] = phi ptr [ [[TMP2]], [[ENTRY]] ], [ [[OMP_ARRAYCPY_DEST_ELEMENT:%.*]], [[OMP_ARRAYCPY_BODY]] ]
-// CHECK1-NEXT:    call void @_ZplRK1SS1_(ptr sret([[STRUCT_S]]) align 4 [[REF_TMP]], ptr nonnull align 4 dereferenceable(4) [[OMP_ARRAYCPY_DESTELEMENTPAST]], ptr nonnull align 4 dereferenceable(4) [[OMP_ARRAYCPY_SRCELEMENTPAST]])
+// CHECK1-NEXT:    call void @_ZplRK1SS1_(ptr dead_on_unwind writable sret([[STRUCT_S]]) align 4 [[REF_TMP]], ptr nonnull align 4 dereferenceable(4) [[OMP_ARRAYCPY_DESTELEMENTPAST]], ptr nonnull align 4 dereferenceable(4) [[OMP_ARRAYCPY_SRCELEMENTPAST]])
 // CHECK1-NEXT:    [[CALL:%.*]] = call nonnull align 4 dereferenceable(4) ptr @_ZN1SaSERKS_(ptr nonnull align 4 dereferenceable(4) [[OMP_ARRAYCPY_DESTELEMENTPAST]], ptr nonnull align 4 dereferenceable(4) [[REF_TMP]])
 // CHECK1-NEXT:    call void @_ZN1SD1Ev(ptr nonnull align 4 dereferenceable(4) [[REF_TMP]]) #[[ATTR3]]
 // CHECK1-NEXT:    [[OMP_ARRAYCPY_DEST_ELEMENT]] = getelementptr [[STRUCT_S]], ptr [[OMP_ARRAYCPY_DESTELEMENTPAST]], i32 1
@@ -351,7 +351,7 @@ int main(int argc, char **argv) {
 //
 //
 // CHECK1-LABEL: define {{[^@]+}}@_ZplRK1SS1_
-// CHECK1-SAME: (ptr noalias sret([[STRUCT_S:%.*]]) align 4 [[AGG_RESULT:%.*]], ptr nonnull align 4 dereferenceable(4) [[A:%.*]], ptr nonnull align 4 dereferenceable(4) [[B:%.*]]) #[[ATTR1]] {
+// CHECK1-SAME: (ptr dead_on_unwind noalias writable sret([[STRUCT_S:%.*]]) align 4 [[AGG_RESULT:%.*]], ptr nonnull align 4 dereferenceable(4) [[A:%.*]], ptr nonnull align 4 dereferenceable(4) [[B:%.*]]) #[[ATTR1]] {
 // CHECK1-NEXT:  entry:
 // CHECK1-NEXT:    [[RESULT_PTR:%.*]] = alloca ptr, align 8
 // CHECK1-NEXT:    [[A_ADDR:%.*]] = alloca ptr, align 8
diff --git a/clang/test/PCH/gch-probe.c b/clang/test/PCH/gch-probe.c
index 8b1e1fab5ad97..370abd26d2517 100644
--- a/clang/test/PCH/gch-probe.c
+++ b/clang/test/PCH/gch-probe.c
@@ -1,7 +1,15 @@
+// Unsupported on AIX because we don't support the requisite "__clangast"
+// section in XCOFF yet.
+// UNSUPPORTED: target={{.*}}-aix{{.*}}
+
 // For GCC compatibility, clang should probe also with the .gch extension.
 // RUN: %clang -x c-header -c %s -o %t.h.gch
 // RUN: %clang -fsyntax-only -include %t.h %s
 
+// -gmodules embeds the Clang AST file in an object file.
+// RUN: %clang -x c-header -c %s -gmodules -o %t.h.gch
+// RUN: %clang -fsyntax-only -include %t.h %s
+
 // gch probing should ignore files which are not clang pch files.
 // RUN: %clang -fsyntax-only -include %S/Inputs/gch-probe.h %s 2>&1 | FileCheck %s
 // CHECK: warning: precompiled header '{{.*}}gch-probe.h.gch' was ignored because it is not a clang PCH file
diff --git a/clang/test/Parser/c2x-attribute-keywords.c b/clang/test/Parser/c2x-attribute-keywords.c
index d8291b710e6db..b88d2b9c23e69 100644
--- a/clang/test/Parser/c2x-attribute-keywords.c
+++ b/clang/test/Parser/c2x-attribute-keywords.c
@@ -1,60 +1,64 @@
-// RUN: %clang_cc1 -fsyntax-only -triple aarch64-none-linux-gnu -target-feature +sme -verify=expected,notc2x -Wno-strict-prototypes %s
-// RUN: %clang_cc1 -fsyntax-only -triple aarch64-none-linux-gnu -target-feature +sme -verify=expected,c2x %s
-
-enum __arm_streaming E { // expected-error {{'__arm_streaming' only applies to non-K&R-style functions}}
-  One __arm_streaming, // expected-error {{'__arm_streaming' only applies to non-K&R-style functions}}
+// RUN: sed -e "s@ATTR_USE@__arm_streaming@g" -e "s@ATTR_NAME@__arm_streaming@g" %s > %t
+// RUN: %clang_cc1 -fsyntax-only -triple aarch64-none-linux-gnu -target-feature +sme -verify=expected,notc2x -Wno-strict-prototypes %t
+// RUN: %clang_cc1 -fsyntax-only -triple aarch64-none-linux-gnu -target-feature +sme -verify=expected,c2x %t
+// RUN: sed -e "s@ATTR_USE@__arm_inout\(\"za\"\)@g" -e "s@ATTR_NAME@__arm_inout@g" %s > %t
+// RUN: %clang_cc1 -fsyntax-only -triple aarch64-none-linux-gnu -target-feature +sme -verify=expected,notc2x -Wno-strict-prototypes %t
+// RUN: %clang_cc1 -fsyntax-only -triple aarch64-none-linux-gnu -target-feature +sme -verify=expected,c2x %t
+
+enum ATTR_USE E { // expected-error {{'ATTR_NAME' only applies to non-K&R-style functions}}
+  One ATTR_USE, // expected-error {{'ATTR_NAME' only applies to non-K&R-style functions}}
   Two,
-  Three __arm_streaming // expected-error {{'__arm_streaming' only applies to non-K&R-style functions}}
+  Three ATTR_USE // expected-error {{'ATTR_NAME' only applies to non-K&R-style functions}}
 };
 
-enum __arm_streaming { Four }; // expected-error {{'__arm_streaming' only applies to non-K&R-style functions}}
-__arm_streaming enum E2 { Five }; // expected-error {{misplaced '__arm_streaming'}}
+enum ATTR_USE { Four }; // expected-error {{'ATTR_NAME' only applies to non-K&R-style functions}}
+ATTR_USE enum E2 { Five }; // expected-error {{misplaced 'ATTR_NAME'}}
 
 // FIXME: this diagnostic can be improved.
-enum { __arm_streaming Six }; // expected-error {{expected identifier}}
+enum { ATTR_USE Six }; // expected-error {{expected identifier}}
 
 // FIXME: this diagnostic can be improved.
-enum E3 __arm_streaming { Seven }; // expected-error {{expected identifier or '('}}
-
-struct __arm_streaming S1 { // expected-error {{'__arm_streaming' only applies to non-K&R-style functions}}
-  int i __arm_streaming; // expected-error {{'__arm_streaming' only applies to function types}}
-  int __arm_streaming j; // expected-error {{'__arm_streaming' only applies to function types}}
-  int k[10] __arm_streaming; // expected-error {{'__arm_streaming' only applies to function types}}
-  int l __arm_streaming[10]; // expected-error {{'__arm_streaming' only applies to function types}}
-  __arm_streaming int m, n; // expected-error {{'__arm_streaming' only applies to function types}}
-  int o __arm_streaming : 12; // expected-error {{'__arm_streaming' only applies to function types}}
-  int __arm_streaming : 0; // expected-error {{'__arm_streaming' only applies to function types}}
-  int p, __arm_streaming : 0; // expected-error {{'__arm_streaming' cannot appear here}}
-  int q, __arm_streaming r; // expected-error {{'__arm_streaming' cannot appear here}}
-  __arm_streaming int; // expected-error {{'__arm_streaming' cannot appear here}} \
+enum E3 ATTR_USE { Seven }; // expected-error {{expected identifier or '('}}
+
+struct ATTR_USE S1 { // expected-error {{'ATTR_NAME' only applies to non-K&R-style functions}}
+  int i ATTR_USE; // expected-error {{'ATTR_NAME' only applies to function types}}
+  int ATTR_USE j; // expected-error {{'ATTR_NAME' only applies to function types}}
+  int k[10] ATTR_USE; // expected-error {{'ATTR_NAME' only applies to function types}}
+  int l ATTR_USE[10]; // expected-error {{'ATTR_NAME' only applies to function types}}
+  ATTR_USE int m, n; // expected-error {{'ATTR_NAME' only applies to function types}}
+  int o ATTR_USE : 12; // expected-error {{'ATTR_NAME' only applies to function types}}
+  int ATTR_USE : 0; // expected-error {{'ATTR_NAME' only applies to function types}}
+  int p, ATTR_USE : 0; // expected-error {{'ATTR_NAME' cannot appear here}}
+  int q, ATTR_USE r; // expected-error {{'ATTR_NAME' cannot appear here}}
+  ATTR_USE int; // expected-error {{'ATTR_NAME' cannot appear here}} \
             // expected-warning {{declaration does not declare anything}}
 };
 
-__arm_streaming struct S2 { int a; }; // expected-error {{misplaced '__arm_streaming'}}
-struct S3 __arm_streaming { int a; }; // expected-error {{'__arm_streaming' cannot appear here}} \
-                                         expected-error {{'__arm_streaming' only applies to non-K&R-style functions}}
+ATTR_USE struct S2 { int a; }; // expected-error {{misplaced 'ATTR_NAME'}}
+struct S3 ATTR_USE { int a; }; // expected-error {{'ATTR_NAME' cannot appear here}} \
+                                         expected-error {{'ATTR_NAME' only applies to non-K&R-style functions}}
 
-union __arm_streaming U { // expected-error {{'__arm_streaming' only applies to non-K&R-style functions}}
-  double d __arm_streaming; // expected-error {{'__arm_streaming' only applies to function types; type here is 'double'}}
-  __arm_streaming int i; // expected-error {{'__arm_streaming' only applies to function types; type here is 'int'}}
+union ATTR_USE U { // expected-error {{'ATTR_NAME' only applies to non-K&R-style functions}}
+  double d ATTR_USE; // expected-error {{'ATTR_NAME' only applies to function types; type here is 'double'}}
+  ATTR_USE int i; // expected-error {{'ATTR_NAME' only applies to function types; type here is 'int'}}
 };
 
-__arm_streaming union U2 { double d; }; // expected-error {{misplaced '__arm_streaming'}}
-union U3 __arm_streaming { double d; }; // expected-error {{'__arm_streaming' cannot appear here}} \
-                                           expected-error {{'__arm_streaming' only applies to non-K&R-style functions}}
+ATTR_USE union U2 { double d; }; // expected-error {{misplaced 'ATTR_NAME'}}
+union U3 ATTR_USE { double d; }; // expected-error {{'ATTR_NAME' cannot appear here}} \
+                                           expected-error {{'ATTR_NAME' only applies to non-K&R-style functions}}
 
-struct __arm_streaming IncompleteStruct; // expected-error {{'__arm_streaming' only applies to non-K&R-style functions}}
-union __arm_streaming IncompleteUnion; // expected-error {{'__arm_streaming' only applies to non-K&R-style functions}}
-enum __arm_streaming IncompleteEnum; // expected-error {{'__arm_streaming' only applies to non-K&R-style functions}}
+struct ATTR_USE IncompleteStruct; // expected-error {{'ATTR_NAME' only applies to non-K&R-style functions}}
+union ATTR_USE IncompleteUnion; // expected-error {{'ATTR_NAME' only applies to non-K&R-style functions}}
+enum ATTR_USE IncompleteEnum; // expected-error {{'ATTR_NAME' only applies to non-K&R-style functions}}
 
-__arm_streaming void f1(void); // expected-error {{'__arm_streaming' cannot be applied to a declaration}}
-void __arm_streaming f2(void); // expected-error {{'__arm_streaming' only applies to function types}}
-void f3 __arm_streaming (void); // expected-error {{'__arm_streaming' cannot be applied to a declaration}}
-void f4(void) __arm_streaming;
+ATTR_USE void f1(void); // expected-error {{'ATTR_NAME' cannot be applied to a declaration}}
+void ATTR_USE f2(void); // expected-error {{'ATTR_NAME' only applies to function types}}
+void f3 ATTR_USE (void); // expected-error {{'ATTR_NAME' cannot be applied to a declaration}}
+void f4(void) ATTR_USE;
 
-void f5(int i __arm_streaming, __arm_streaming int j, int __arm_streaming k); // expected-error 3 {{'__arm_streaming' only applies to function types}}
+void f5(int i ATTR_USE, ATTR_USE int j, int ATTR_USE k); // expected-error 3 {{'ATTR_NAME' only applies to function types}}
 
-void f6(a, b) __arm_streaming int a; int b; { // expected-error {{'__arm_streaming' cannot appear here}} \
+void f6(a, b) ATTR_USE int a; int b; { // expected-error {{'ATTR_NAME' cannot appear here}} \
                                                  c2x-warning {{deprecated}}
 }
 
@@ -63,57 +67,74 @@ void f6(a, b) __arm_streaming int a; int b; { // expected-error {{'__arm_streami
 // behavior given that we *don't* want to parse it as part of the K&R parameter
 // declarations. It is disallowed to avoid a parsing ambiguity we already
 // handle well.
-int (*f7(a, b))(int, int) __arm_streaming int a; int b; { // c2x-warning {{deprecated}}
+int (*f7(a, b))(int, int) ATTR_USE int a; int b; { // c2x-warning {{deprecated}}
   return 0;
 }
 
-__arm_streaming int a, b; // expected-error {{'__arm_streaming' only applies to function types}}
-int c __arm_streaming, d __arm_streaming; // expected-error 2 {{'__arm_streaming' only applies to function types}}
+ATTR_USE int a, b; // expected-error {{'ATTR_NAME' only applies to function types}}
+int c ATTR_USE, d ATTR_USE; // expected-error 2 {{'ATTR_NAME' only applies to function types}}
 
-void f8(void) __arm_streaming {
-  __arm_streaming int i, j; // expected-error {{'__arm_streaming' only applies to function types}}
-  int k, l __arm_streaming; // expected-error {{'__arm_streaming' only applies to function types}}
+void f8(void) ATTR_USE {
+  ATTR_USE int i, j; // expected-error {{'ATTR_NAME' only applies to function types}}
+  int k, l ATTR_USE; // expected-error {{'ATTR_NAME' only applies to function types}}
 }
 
-__arm_streaming void f9(void) { // expected-error {{'__arm_streaming' cannot be applied to a declaration}}
-  int i[10] __arm_streaming; // expected-error {{'__arm_streaming' only applies to function types}}
-  int (*fp1)(void)__arm_streaming;
-  int (*fp2 __arm_streaming)(void); // expected-error {{'__arm_streaming' cannot be applied to a declaration}}
+ATTR_USE void f9(void) { // expected-error {{'ATTR_NAME' cannot be applied to a declaration}}
+  int i[10] ATTR_USE; // expected-error {{'ATTR_NAME' only applies to function types}}
+  int (*fp1)(void)ATTR_USE;
+  int (*fp2 ATTR_USE)(void); // expected-error {{'ATTR_NAME' cannot be applied to a declaration}}
 
-  int * __arm_streaming *ipp; // expected-error {{'__arm_streaming' only applies to function types}}
+  int * ATTR_USE *ipp; // expected-error {{'ATTR_NAME' only applies to function types}}
 }
 
-void f10(int j[static 10] __arm_streaming, int k[*] __arm_streaming); // expected-error 2 {{'__arm_streaming' only applies to function types}}
+void f10(int j[static 10] ATTR_USE, int k[*] ATTR_USE); // expected-error 2 {{'ATTR_NAME' only applies to function types}}
 
 void f11(void) {
-  __arm_streaming {} // expected-error {{'__arm_streaming' cannot be applied to a statement}}
-  __arm_streaming if (1) {} // expected-error {{'__arm_streaming' cannot be applied to a statement}}
+  ATTR_USE {} // expected-error {{'ATTR_NAME' cannot be applied to a statement}}
+  ATTR_USE if (1) {} // expected-error {{'ATTR_NAME' cannot be applied to a statement}}
 
-  __arm_streaming switch (1) { // expected-error {{'__arm_streaming' cannot be applied to a statement}}
-  __arm_streaming case 1: __arm_streaming break; // expected-error 2 {{'__arm_streaming' cannot be applied to a statement}}
-  __arm_streaming default: break; // expected-error {{'__arm_streaming' cannot be applied to a statement}}
+  ATTR_USE switch (1) { // expected-error {{'ATTR_NAME' cannot be applied to a statement}}
+  ATTR_USE case 1: ATTR_USE break; // expected-error 2 {{'ATTR_NAME' cannot be applied to a statement}}
+  ATTR_USE default: break; // expected-error {{'ATTR_NAME' cannot be applied to a statement}}
   }
 
   goto foo;
-  __arm_streaming foo: (void)1; // expected-error {{'__arm_streaming' only applies to non-K&R-style functions}}
+  ATTR_USE foo: (void)1; // expected-error {{'ATTR_NAME' only applies to non-K&R-style functions}}
 
-  __arm_streaming for (;;); // expected-error {{'__arm_streaming' cannot be applied to a statement}}
-  __arm_streaming while (1); // expected-error {{'__arm_streaming' cannot be applied to a statement}}
-  __arm_streaming do __arm_streaming { } while(1); // expected-error 2 {{'__arm_streaming' cannot be applied to a statement}}
+  ATTR_USE for (;;); // expected-error {{'ATTR_NAME' cannot be applied to a statement}}
+  ATTR_USE while (1); // expected-error {{'ATTR_NAME' cannot be applied to a statement}}
+  ATTR_USE do ATTR_USE { } while(1); // expected-error 2 {{'ATTR_NAME' cannot be applied to a statement}}
 
-  __arm_streaming (void)1; // expected-error {{'__arm_streaming' cannot be applied to a statement}}
+  ATTR_USE (void)1; // expected-error {{'ATTR_NAME' cannot be applied to a statement}}
 
-  __arm_streaming; // expected-error {{'__arm_streaming' cannot be applied to a statement}}
+  ATTR_USE; // expected-error {{'ATTR_NAME' cannot be applied to a statement}}
 
-  (void)sizeof(int [4]__arm_streaming); // expected-error {{'__arm_streaming' only applies to function types}}
-  (void)sizeof(struct __arm_streaming S3 { int a __arm_streaming; }); // expected-error {{'__arm_streaming' only applies to non-K&R-style functions}} \
-                                                                      // expected-error {{'__arm_streaming' only applies to function types; type here is 'int'}}
+  (void)sizeof(int [4]ATTR_USE); // expected-error {{'ATTR_NAME' only applies to function types}}
+  (void)sizeof(struct ATTR_USE S3 { int a ATTR_USE; }); // expected-error {{'ATTR_NAME' only applies to non-K&R-style functions}} \
+                                                                      // expected-error {{'ATTR_NAME' only applies to function types; type here is 'int'}}
 
-  __arm_streaming return; // expected-error {{'__arm_streaming' cannot be applied to a statement}}
+  ATTR_USE return; // expected-error {{'ATTR_NAME' cannot be applied to a statement}}
 
-  __arm_streaming asm (""); // expected-error {{'__arm_streaming' cannot appear here}}
+  ATTR_USE asm (""); // expected-error {{'ATTR_NAME' cannot appear here}}
 }
 
-struct __arm_streaming S4 *s; // expected-error {{'__arm_streaming' cannot appear here}}
+struct ATTR_USE S4 *s; // expected-error {{'ATTR_NAME' cannot appear here}}
 struct S5 {};
-int c = sizeof(struct __arm_streaming S5); // expected-error {{'__arm_streaming' cannot appear here}}
+int c = sizeof(struct ATTR_USE S5); // expected-error {{'ATTR_NAME' cannot appear here}}
+
+void invalid_parentheses1() __arm_inout; // expected-error {{expected '(' after ''__arm_inout''}}
+void invalid_parentheses2() __arm_inout(; // expected-error {{expected string literal as argument of '__arm_inout' attribute}}
+void invalid_parentheses3() __arm_inout((); // expected-error {{expected string literal as argument of '__arm_inout' attribute}}
+void invalid_parentheses4() __arm_inout); // expected-error {{expected '(' after ''__arm_inout''}} \
+                                          // expected-error {{expected function body after function declarator}}
+void invalid_parentheses5() __arm_inout(());  // expected-error {{expected string literal as argument of '__arm_inout' attribute}}
+void invalid_parentheses6() __arm_inout("za"; // expected-error {{expected ')'}}
+void invalid_parentheses7() __arm_streaming(; // expected-error {{expected parameter declarator}} \
+                                              // expected-error {{expected ')'}} \
+                                              // expected-note {{to match this '('}} \
+                                              // expected-error {{function cannot return function type 'void ()'}} \
+                                              // expected-error {{'__arm_streaming' only applies to function types; type here is 'int ()'}} \
+                                              // expected-warning {{'__arm_streaming' only applies to non-K&R-style functions}}
+void invalid_parentheses8() __arm_streaming();  // expected-error {{function cannot return function type 'void ()'}} \
+                                                // expected-error {{'__arm_streaming' only applies to function types; type here is 'int ()'}} \
+                                                // expected-warning {{'__arm_streaming' only applies to non-K&R-style functions}}
diff --git a/clang/test/Parser/c2x-attribute-keywords.m b/clang/test/Parser/c2x-attribute-keywords.m
index 2296be13cb714..575c88ffffc3d 100644
--- a/clang/test/Parser/c2x-attribute-keywords.m
+++ b/clang/test/Parser/c2x-attribute-keywords.m
@@ -1,6 +1,6 @@
 // RUN: %clang_cc1 -fsyntax-only -triple aarch64-none-linux-gnu -target-feature +sme -verify %s
 
-enum __arm_streaming E1 : int; // expected-error {{'__arm_streaming' only applies to non-K&R-style functions}}
+enum __arm_inout("za") E1 : int; // expected-error {{'__arm_inout' only applies to non-K&R-style functions}}
 
 @interface Base
 @end
@@ -15,5 +15,5 @@ - (S *) foo;
 
 
 void f(T *t) {
-  __arm_streaming[[t foo] bar]; // expected-error {{'__arm_streaming' cannot be applied to a statement}}
+  __arm_inout("za")[[t foo] bar]; // expected-error {{'__arm_inout' cannot be applied to a statement}}
 }
diff --git a/clang/test/Parser/cxx0x-keyword-attributes.cpp b/clang/test/Parser/cxx0x-keyword-attributes.cpp
index 8d31efac53208..be7423cc7ecee 100644
--- a/clang/test/Parser/cxx0x-keyword-attributes.cpp
+++ b/clang/test/Parser/cxx0x-keyword-attributes.cpp
@@ -1,4 +1,7 @@
-// RUN: %clang_cc1 -fcxx-exceptions -fdeclspec -fexceptions -fsyntax-only -verify -std=c++11 -Wc++14-compat -Wc++14-extensions -Wc++17-extensions -triple aarch64-none-linux-gnu -target-feature +sme %s
+// RUN: sed -e "s@ATTR_USE@__arm_streaming@g" -e "s@ATTR_NAME@__arm_streaming@g" %s > %t
+// RUN: %clang_cc1 -fcxx-exceptions -fdeclspec -fexceptions -fsyntax-only -verify -std=c++11 -Wc++14-compat -Wc++14-extensions -Wc++17-extensions -triple aarch64-none-linux-gnu -target-feature +sme -x c++ %t
+// RUN: sed -e "s@ATTR_USE@__arm_inout\(\"za\"\)@g" -e "s@ATTR_NAME@__arm_inout@g" %s > %t
+// RUN: %clang_cc1 -fcxx-exceptions -fdeclspec -fexceptions -fsyntax-only -verify -std=c++11 -Wc++14-compat -Wc++14-extensions -Wc++17-extensions -triple aarch64-none-linux-gnu -target-feature +sme -x c++ %t
 
 // Need std::initializer_list
 namespace std {
@@ -35,136 +38,136 @@ namespace std {
 
 
 // Declaration syntax checks
-__arm_streaming int before_attr; // expected-error {{'__arm_streaming' only applies to function types}}
-int __arm_streaming between_attr; // expected-error {{'__arm_streaming' only applies to function types}}
-const __arm_streaming int between_attr_2 = 0; // expected-error {{'__arm_streaming' cannot appear here}}
-int after_attr __arm_streaming; // expected-error {{'__arm_streaming' only applies to function types}}
-int * __arm_streaming ptr_attr; // expected-error {{'__arm_streaming' only applies to function types}}
-int & __arm_streaming ref_attr = after_attr; // expected-error {{'__arm_streaming' only applies to function types}}
-int && __arm_streaming rref_attr = 0; // expected-error {{'__arm_streaming' only applies to function types}}
-int array_attr [1] __arm_streaming; // expected-error {{'__arm_streaming' only applies to function types}}
-void fn_attr () __arm_streaming;
-void noexcept_fn_attr () noexcept __arm_streaming;
+ATTR_USE int before_attr; // expected-error {{'ATTR_NAME' only applies to function types}}
+int ATTR_USE between_attr; // expected-error {{'ATTR_NAME' only applies to function types}}
+const ATTR_USE int between_attr_2 = 0; // expected-error {{'ATTR_NAME' cannot appear here}}
+int after_attr ATTR_USE; // expected-error {{'ATTR_NAME' only applies to function types}}
+int * ATTR_USE ptr_attr; // expected-error {{'ATTR_NAME' only applies to function types}}
+int & ATTR_USE ref_attr = after_attr; // expected-error {{'ATTR_NAME' only applies to function types}}
+int && ATTR_USE rref_attr = 0; // expected-error {{'ATTR_NAME' only applies to function types}}
+int array_attr [1] ATTR_USE; // expected-error {{'ATTR_NAME' only applies to function types}}
+void fn_attr () ATTR_USE;
+void noexcept_fn_attr () noexcept ATTR_USE;
 struct MemberFnOrder {
-  virtual void f() const volatile && noexcept __arm_streaming final = 0;
+  virtual void f() const volatile && noexcept ATTR_USE final = 0;
 };
-struct __arm_streaming struct_attr; // expected-error {{'__arm_streaming' only applies to non-K&R-style functions}}
-class __arm_streaming class_attr {}; // expected-error {{'__arm_streaming' only applies to non-K&R-style functions}}
-union __arm_streaming union_attr; // expected-error {{'__arm_streaming' only applies to non-K&R-style functions}}
-enum __arm_streaming E { }; // expected-error {{'__arm_streaming' only applies to non-K&R-style functions}}
+struct ATTR_USE struct_attr; // expected-error {{'ATTR_NAME' only applies to non-K&R-style functions}}
+class ATTR_USE class_attr {}; // expected-error {{'ATTR_NAME' only applies to non-K&R-style functions}}
+union ATTR_USE union_attr; // expected-error {{'ATTR_NAME' only applies to non-K&R-style functions}}
+enum ATTR_USE E { }; // expected-error {{'ATTR_NAME' only applies to non-K&R-style functions}}
 namespace test_misplacement {
-__arm_streaming struct struct_attr2;  // expected-error {{misplaced '__arm_streaming'}}
-__arm_streaming class class_attr2; // expected-error {{misplaced '__arm_streaming'}}
-__arm_streaming union union_attr2; // expected-error {{misplaced '__arm_streaming'}}
-__arm_streaming enum  E2 { }; // expected-error {{misplaced '__arm_streaming'}}
+ATTR_USE struct struct_attr2;  // expected-error {{misplaced 'ATTR_NAME'}}
+ATTR_USE class class_attr2; // expected-error {{misplaced 'ATTR_NAME'}}
+ATTR_USE union union_attr2; // expected-error {{misplaced 'ATTR_NAME'}}
+ATTR_USE enum  E2 { }; // expected-error {{misplaced 'ATTR_NAME'}}
 }
 
 // Checks attributes placed at wrong syntactic locations of class specifiers.
-class __arm_streaming __arm_streaming // expected-error 2 {{'__arm_streaming' only applies to non-K&R-style functions}}
-  attr_after_class_name_decl __arm_streaming __arm_streaming; // expected-error {{'__arm_streaming' cannot appear here}} \
-                                                                 expected-error 2 {{'__arm_streaming' only applies to non-K&R-style functions}}
+class ATTR_USE ATTR_USE // expected-error 2 {{'ATTR_NAME' only applies to non-K&R-style functions}}
+  attr_after_class_name_decl ATTR_USE ATTR_USE; // expected-error {{'ATTR_NAME' cannot appear here}} \
+                                                                 expected-error 2 {{'ATTR_NAME' only applies to non-K&R-style functions}}
 
-class __arm_streaming __arm_streaming // expected-error 2 {{'__arm_streaming' only applies to non-K&R-style functions}}
- attr_after_class_name_definition __arm_streaming __arm_streaming __arm_streaming{}; // expected-error {{'__arm_streaming' cannot appear here}} \
-                                                                                        expected-error 3 {{'__arm_streaming' only applies to non-K&R-style functions}}
+class ATTR_USE ATTR_USE // expected-error 2 {{'ATTR_NAME' only applies to non-K&R-style functions}}
+ attr_after_class_name_definition ATTR_USE ATTR_USE ATTR_USE{}; // expected-error {{'ATTR_NAME' cannot appear here}} \
+                                                                                        expected-error 3 {{'ATTR_NAME' only applies to non-K&R-style functions}}
 
-class __arm_streaming c {}; // expected-error {{'__arm_streaming' only applies to non-K&R-style functions}}
-class c __arm_streaming __arm_streaming x; // expected-error 2 {{'__arm_streaming' only applies to function types}}
-class c __arm_streaming __arm_streaming y __arm_streaming __arm_streaming; // expected-error 4 {{'__arm_streaming' only applies to function types}}
+class ATTR_USE c {}; // expected-error {{'ATTR_NAME' only applies to non-K&R-style functions}}
+class c ATTR_USE ATTR_USE x; // expected-error 2 {{'ATTR_NAME' only applies to function types}}
+class c ATTR_USE ATTR_USE y ATTR_USE ATTR_USE; // expected-error 4 {{'ATTR_NAME' only applies to function types}}
 class c final [(int){0}];
 
 class base {};
-class __arm_streaming __arm_streaming final_class // expected-error 2 {{'__arm_streaming' only applies to non-K&R-style functions}}
-  __arm_streaming alignas(float) final // expected-error {{'__arm_streaming' cannot appear here}} \
-                                          expected-error {{'__arm_streaming' only applies to non-K&R-style functions}}
-  __arm_streaming alignas(float) __arm_streaming alignas(float): base{}; // expected-error {{'__arm_streaming' cannot appear here}}
+class ATTR_USE ATTR_USE final_class // expected-error 2 {{'ATTR_NAME' only applies to non-K&R-style functions}}
+  ATTR_USE alignas(float) final // expected-error {{'ATTR_NAME' cannot appear here}} \
+                                          expected-error {{'ATTR_NAME' only applies to non-K&R-style functions}}
+  ATTR_USE alignas(float) ATTR_USE alignas(float): base{}; // expected-error {{'ATTR_NAME' cannot appear here}}
 
-class __arm_streaming __arm_streaming final_class_another // expected-error 2 {{'__arm_streaming' only applies to non-K&R-style functions}}
-  __arm_streaming __arm_streaming alignas(16) final // expected-error {{'__arm_streaming' cannot appear here}} \
-                                                       expected-error 2 {{'__arm_streaming' only applies to non-K&R-style functions}}
-  __arm_streaming __arm_streaming alignas(16) __arm_streaming{}; // expected-error {{'__arm_streaming' cannot appear here}}
+class ATTR_USE ATTR_USE final_class_another // expected-error 2 {{'ATTR_NAME' only applies to non-K&R-style functions}}
+  ATTR_USE ATTR_USE alignas(16) final // expected-error {{'ATTR_NAME' cannot appear here}} \
+                                                       expected-error 2 {{'ATTR_NAME' only applies to non-K&R-style functions}}
+  ATTR_USE ATTR_USE alignas(16) ATTR_USE{}; // expected-error {{'ATTR_NAME' cannot appear here}}
 
-class after_class_close {} __arm_streaming; // expected-error {{'__arm_streaming' cannot appear here, place it after "class" to apply it to the type declaration}}
+class after_class_close {} ATTR_USE; // expected-error {{'ATTR_NAME' cannot appear here, place it after "class" to apply it to the type declaration}}
 
 class C {};
 
-__arm_streaming struct with_init_declarators {} init_declarator; // expected-error {{'__arm_streaming' only applies to function types}}
-__arm_streaming struct no_init_declarators; // expected-error {{misplaced '__arm_streaming'}}
-template __arm_streaming struct no_init_declarators_template; // expected-error {{'__arm_streaming' cannot appear here}}
+ATTR_USE struct with_init_declarators {} init_declarator; // expected-error {{'ATTR_NAME' only applies to function types}}
+ATTR_USE struct no_init_declarators; // expected-error {{misplaced 'ATTR_NAME'}}
+template ATTR_USE struct no_init_declarators_template; // expected-error {{'ATTR_NAME' cannot appear here}}
 void fn_with_structs() {
-  __arm_streaming struct with_init_declarators {} init_declarator; // expected-error {{'__arm_streaming' only applies to function types}}
-  __arm_streaming struct no_init_declarators; // expected-error {{'__arm_streaming' cannot appear here}}
+  ATTR_USE struct with_init_declarators {} init_declarator; // expected-error {{'ATTR_NAME' only applies to function types}}
+  ATTR_USE struct no_init_declarators; // expected-error {{'ATTR_NAME' cannot appear here}}
 }
-__arm_streaming; // expected-error {{'__arm_streaming' only applies to non-K&R-style functions}}
+ATTR_USE; // expected-error {{'ATTR_NAME' only applies to non-K&R-style functions}}
 struct ctordtor {
-  __arm_streaming ctordtor __arm_streaming () __arm_streaming; // expected-error 2 {{'__arm_streaming' cannot be applied to a declaration}}
-  ctordtor (C) __arm_streaming;
-  __arm_streaming ~ctordtor __arm_streaming () __arm_streaming; // expected-error 2 {{'__arm_streaming' cannot be applied to a declaration}}
+  ATTR_USE ctordtor ATTR_USE () ATTR_USE; // expected-error 2 {{'ATTR_NAME' cannot be applied to a declaration}}
+  ctordtor (C) ATTR_USE;
+  ATTR_USE ~ctordtor ATTR_USE () ATTR_USE; // expected-error 2 {{'ATTR_NAME' cannot be applied to a declaration}}
 };
-__arm_streaming ctordtor::ctordtor __arm_streaming () __arm_streaming {} // expected-error 2 {{'__arm_streaming' cannot be applied to a declaration}}
-__arm_streaming ctordtor::ctordtor (C) __arm_streaming try {} catch (...) {} // expected-error {{'__arm_streaming' cannot be applied to a declaration}}
-__arm_streaming ctordtor::~ctordtor __arm_streaming () __arm_streaming {} // expected-error 2 {{'__arm_streaming' cannot be applied to a declaration}}
-extern "C++" __arm_streaming int extern_attr; // expected-error {{'__arm_streaming' only applies to function types}}
-template  __arm_streaming void template_attr (); // expected-error {{'__arm_streaming' cannot be applied to a declaration}}
-__arm_streaming __arm_streaming int __arm_streaming __arm_streaming multi_attr __arm_streaming __arm_streaming; // expected-error 6 {{'__arm_streaming' only applies to function types}}
-
-int (paren_attr) __arm_streaming; // expected-error {{'__arm_streaming' cannot appear here}}
-unsigned __arm_streaming int attr_in_decl_spec; // expected-error {{'__arm_streaming' cannot appear here}}
-unsigned __arm_streaming int __arm_streaming const double_decl_spec = 0; // expected-error 2 {{'__arm_streaming' cannot appear here}}
+ATTR_USE ctordtor::ctordtor ATTR_USE () ATTR_USE {} // expected-error 2 {{'ATTR_NAME' cannot be applied to a declaration}}
+ATTR_USE ctordtor::ctordtor (C) ATTR_USE try {} catch (...) {} // expected-error {{'ATTR_NAME' cannot be applied to a declaration}}
+ATTR_USE ctordtor::~ctordtor ATTR_USE () ATTR_USE {} // expected-error 2 {{'ATTR_NAME' cannot be applied to a declaration}}
+extern "C++" ATTR_USE int extern_attr; // expected-error {{'ATTR_NAME' only applies to function types}}
+template  ATTR_USE void template_attr (); // expected-error {{'ATTR_NAME' cannot be applied to a declaration}}
+ATTR_USE ATTR_USE int ATTR_USE ATTR_USE multi_attr ATTR_USE ATTR_USE; // expected-error 6 {{'ATTR_NAME' only applies to function types}}
+
+int (paren_attr) ATTR_USE; // expected-error {{'ATTR_NAME' cannot appear here}}
+unsigned ATTR_USE int attr_in_decl_spec; // expected-error {{'ATTR_NAME' cannot appear here}}
+unsigned ATTR_USE int ATTR_USE const double_decl_spec = 0; // expected-error 2 {{'ATTR_NAME' cannot appear here}}
 class foo {
-  void const_after_attr () __arm_streaming const; // expected-error {{expected ';'}}
+  void const_after_attr () ATTR_USE const; // expected-error {{expected ';'}}
 };
-extern "C++" __arm_streaming { } // expected-error {{'__arm_streaming' cannot appear here}}
-__arm_streaming extern "C++" { } // expected-error {{'__arm_streaming' cannot appear here}}
-__arm_streaming template  void before_template_attr (); // expected-error {{'__arm_streaming' cannot appear here}}
-__arm_streaming namespace ns { int i; } // expected-error {{'__arm_streaming' cannot appear here}}
-__arm_streaming static_assert(true, ""); //expected-error {{'__arm_streaming' cannot appear here}}
-__arm_streaming asm(""); // expected-error {{'__arm_streaming' cannot appear here}}
-
-__arm_streaming using ns::i; // expected-warning {{ISO C++}} \
-                                expected-error {{'__arm_streaming' only applies to non-K&R-style functions}}
-__arm_streaming using namespace ns; // expected-error {{'__arm_streaming' only applies to non-K&R-style functions}}
-namespace __arm_streaming ns2 {} // expected-warning {{attributes on a namespace declaration are a C++17 extension}} \
-                                    expected-error {{'__arm_streaming' only applies to non-K&R-style functions}}
-
-using __arm_streaming alignas(4)__arm_streaming ns::i;          // expected-warning 2 {{ISO C++}} \
-                                                                   expected-error {{'__arm_streaming' cannot appear here}} \
+extern "C++" ATTR_USE { } // expected-error {{'ATTR_NAME' cannot appear here}}
+ATTR_USE extern "C++" { } // expected-error {{'ATTR_NAME' cannot appear here}}
+ATTR_USE template  void before_template_attr (); // expected-error {{'ATTR_NAME' cannot appear here}}
+ATTR_USE namespace ns { int i; } // expected-error {{'ATTR_NAME' cannot appear here}}
+ATTR_USE static_assert(true, ""); //expected-error {{'ATTR_NAME' cannot appear here}}
+ATTR_USE asm(""); // expected-error {{'ATTR_NAME' cannot appear here}}
+
+ATTR_USE using ns::i; // expected-warning {{ISO C++}} \
+                                expected-error {{'ATTR_NAME' only applies to non-K&R-style functions}}
+ATTR_USE using namespace ns; // expected-error {{'ATTR_NAME' only applies to non-K&R-style functions}}
+namespace ATTR_USE ns2 {} // expected-warning {{attributes on a namespace declaration are a C++17 extension}} \
+                                    expected-error {{'ATTR_NAME' only applies to non-K&R-style functions}}
+
+using ATTR_USE alignas(4)ATTR_USE ns::i;          // expected-warning 2 {{ISO C++}} \
+                                                                   expected-error {{'ATTR_NAME' cannot appear here}} \
                                                                    expected-error {{'alignas' attribute only applies to variables, data members and tag types}} \
                                                                    expected-warning {{ISO C++}} \
-                                                                   expected-error 2 {{'__arm_streaming' only applies to non-K&R-style functions}}
-using __arm_streaming alignas(4) __arm_streaming foobar = int; // expected-error {{'__arm_streaming' cannot appear here}} \
+                                                                   expected-error 2 {{'ATTR_NAME' only applies to non-K&R-style functions}}
+using ATTR_USE alignas(4) ATTR_USE foobar = int; // expected-error {{'ATTR_NAME' cannot appear here}} \
                                                                   expected-error {{'alignas' attribute only applies to}} \
-                                                                  expected-error 2 {{'__arm_streaming' only applies to function types}}
-
-__arm_streaming using T = int; // expected-error {{'__arm_streaming' cannot appear here}}
-using T __arm_streaming = int; // expected-error {{'__arm_streaming' only applies to function types}}
-template using U __arm_streaming = T; // expected-error {{'__arm_streaming' only applies to function types}}
-using ns::i __arm_streaming; // expected-warning {{ISO C++}} \
-                                expected-error {{'__arm_streaming' only applies to non-K&R-style functions}}
-using ns::i __arm_streaming, ns::i __arm_streaming; // expected-warning 2 {{ISO C++}} \
+                                                                  expected-error 2 {{'ATTR_NAME' only applies to function types}}
+
+ATTR_USE using T = int; // expected-error {{'ATTR_NAME' cannot appear here}}
+using T ATTR_USE = int; // expected-error {{'ATTR_NAME' only applies to function types}}
+template using U ATTR_USE = T; // expected-error {{'ATTR_NAME' only applies to function types}}
+using ns::i ATTR_USE; // expected-warning {{ISO C++}} \
+                                expected-error {{'ATTR_NAME' only applies to non-K&R-style functions}}
+using ns::i ATTR_USE, ns::i ATTR_USE; // expected-warning 2 {{ISO C++}} \
                                                        expected-warning {{use of multiple declarators in a single using declaration is a C++17 extension}} \
-                                                       expected-error 2 {{'__arm_streaming' only applies to non-K&R-style functions}}
+                                                       expected-error 2 {{'ATTR_NAME' only applies to non-K&R-style functions}}
 struct using_in_struct_base {
   typedef int i, j, k, l;
 };
 struct using_in_struct : using_in_struct_base {
-  __arm_streaming using using_in_struct_base::i; // expected-warning {{ISO C++}} \
-                                                    expected-error {{'__arm_streaming' only applies to non-K&R-style functions}}
-  using using_in_struct_base::j __arm_streaming; // expected-warning {{ISO C++}} \
-                                                    expected-error {{'__arm_streaming' only applies to non-K&R-style functions}}
-  __arm_streaming using using_in_struct_base::k __arm_streaming, using_in_struct_base::l __arm_streaming; // expected-warning 3 {{ISO C++}} \
+  ATTR_USE using using_in_struct_base::i; // expected-warning {{ISO C++}} \
+                                                    expected-error {{'ATTR_NAME' only applies to non-K&R-style functions}}
+  using using_in_struct_base::j ATTR_USE; // expected-warning {{ISO C++}} \
+                                                    expected-error {{'ATTR_NAME' only applies to non-K&R-style functions}}
+  ATTR_USE using using_in_struct_base::k ATTR_USE, using_in_struct_base::l ATTR_USE; // expected-warning 3 {{ISO C++}} \
                                                                                                              expected-warning {{use of multiple declarators in a single using declaration is a C++17 extension}} \
-                                                                                                             expected-error 4 {{'__arm_streaming' only applies to non-K&R-style functions}}
+                                                                                                             expected-error 4 {{'ATTR_NAME' only applies to non-K&R-style functions}}
 };
-using __arm_streaming ns::i; // expected-warning {{ISO C++}} \
-                                expected-error {{'__arm_streaming' cannot appear here}} \
-                                expected-error {{'__arm_streaming' only applies to non-K&R-style functions}}
-using T __arm_streaming = int; // expected-error {{'__arm_streaming' only applies to function types}}
+using ATTR_USE ns::i; // expected-warning {{ISO C++}} \
+                                expected-error {{'ATTR_NAME' cannot appear here}} \
+                                expected-error {{'ATTR_NAME' only applies to non-K&R-style functions}}
+using T ATTR_USE = int; // expected-error {{'ATTR_NAME' only applies to function types}}
 
-auto trailing() -> __arm_streaming const int; // expected-error {{'__arm_streaming' cannot appear here}}
-auto trailing() -> const __arm_streaming int; // expected-error {{'__arm_streaming' cannot appear here}}
-auto trailing() -> const int __arm_streaming; // expected-error {{'__arm_streaming' only applies to function types}}
-auto trailing_2() -> struct struct_attr __arm_streaming; // expected-error {{'__arm_streaming' only applies to function types}}
+auto trailing() -> ATTR_USE const int; // expected-error {{'ATTR_NAME' cannot appear here}}
+auto trailing() -> const ATTR_USE int; // expected-error {{'ATTR_NAME' cannot appear here}}
+auto trailing() -> const int ATTR_USE; // expected-error {{'ATTR_NAME' only applies to function types}}
+auto trailing_2() -> struct struct_attr ATTR_USE; // expected-error {{'ATTR_NAME' only applies to function types}}
 
 namespace N {
   struct S {};
@@ -172,88 +175,88 @@ namespace N {
 template struct Template {};
 
 // FIXME: Improve this diagnostic
-struct __arm_streaming N::S s; // expected-error {{'__arm_streaming' cannot appear here}}
-struct __arm_streaming Template t; // expected-error {{'__arm_streaming' cannot appear here}}
-struct __arm_streaming ::template Template u; // expected-error {{'__arm_streaming' cannot appear here}}
-template struct __arm_streaming Template; // expected-error {{'__arm_streaming' cannot appear here}}
+struct ATTR_USE N::S s; // expected-error {{'ATTR_NAME' cannot appear here}}
+struct ATTR_USE Template t; // expected-error {{'ATTR_NAME' cannot appear here}}
+struct ATTR_USE ::template Template u; // expected-error {{'ATTR_NAME' cannot appear here}}
+template struct ATTR_USE Template; // expected-error {{'ATTR_NAME' cannot appear here}}
 template struct __attribute__((pure)) Template; // We still allow GNU-style attributes here
-template <> struct __arm_streaming Template; // expected-error {{'__arm_streaming' only applies to non-K&R-style functions}}
-
-enum __arm_streaming E1 {}; // expected-error {{'__arm_streaming' only applies to non-K&R-style functions}}
-enum __arm_streaming E2; // expected-error {{forbids forward references}} \
-                            expected-error {{'__arm_streaming' only applies to non-K&R-style functions}}
-enum __arm_streaming E1; // expected-error {{'__arm_streaming' only applies to non-K&R-style functions}}
-enum __arm_streaming E3 : int; // expected-error {{'__arm_streaming' only applies to non-K&R-style functions}}
-enum __arm_streaming { // expected-error {{'__arm_streaming' only applies to non-K&R-style functions}}
-  k_123 __arm_streaming = 123 // expected-warning {{attributes on an enumerator declaration are a C++17 extension}} \
-                                 expected-error {{'__arm_streaming' only applies to non-K&R-style functions}}
+template <> struct ATTR_USE Template; // expected-error {{'ATTR_NAME' only applies to non-K&R-style functions}}
+
+enum ATTR_USE E1 {}; // expected-error {{'ATTR_NAME' only applies to non-K&R-style functions}}
+enum ATTR_USE E2; // expected-error {{forbids forward references}} \
+                            expected-error {{'ATTR_NAME' only applies to non-K&R-style functions}}
+enum ATTR_USE E1; // expected-error {{'ATTR_NAME' only applies to non-K&R-style functions}}
+enum ATTR_USE E3 : int; // expected-error {{'ATTR_NAME' only applies to non-K&R-style functions}}
+enum ATTR_USE { // expected-error {{'ATTR_NAME' only applies to non-K&R-style functions}}
+  k_123 ATTR_USE = 123 // expected-warning {{attributes on an enumerator declaration are a C++17 extension}} \
+                                 expected-error {{'ATTR_NAME' only applies to non-K&R-style functions}}
 };
-enum __arm_streaming E1 e; // expected-error {{'__arm_streaming' cannot appear here}}
-enum __arm_streaming class E4 { }; // expected-error {{'__arm_streaming' cannot appear here}}
-enum struct __arm_streaming E5; // expected-error {{'__arm_streaming' only applies to non-K&R-style functions}}
-enum E6 {} __arm_streaming; // expected-error {{'__arm_streaming' cannot appear here, place it after "enum" to apply it to the type declaration}}
+enum ATTR_USE E1 e; // expected-error {{'ATTR_NAME' cannot appear here}}
+enum ATTR_USE class E4 { }; // expected-error {{'ATTR_NAME' cannot appear here}}
+enum struct ATTR_USE E5; // expected-error {{'ATTR_NAME' only applies to non-K&R-style functions}}
+enum E6 {} ATTR_USE; // expected-error {{'ATTR_NAME' cannot appear here, place it after "enum" to apply it to the type declaration}}
 
 struct S {
-  friend int f __arm_streaming (); // expected-error {{'__arm_streaming' cannot appear here}} \
-                                      expected-error {{'__arm_streaming' cannot be applied to a declaration}}
-  friend int f2 __arm_streaming () {} // expected-error {{'__arm_streaming' cannot be applied to a declaration}}
-  __arm_streaming friend int g(); // expected-error {{'__arm_streaming' cannot appear here}}
-  __arm_streaming friend int h() { // expected-error {{'__arm_streaming' cannot be applied to a declaration}}
+  friend int f ATTR_USE (); // expected-error {{'ATTR_NAME' cannot appear here}} \
+                                      expected-error {{'ATTR_NAME' cannot be applied to a declaration}}
+  friend int f2 ATTR_USE () {} // expected-error {{'ATTR_NAME' cannot be applied to a declaration}}
+  ATTR_USE friend int g(); // expected-error {{'ATTR_NAME' cannot appear here}}
+  ATTR_USE friend int h() { // expected-error {{'ATTR_NAME' cannot be applied to a declaration}}
   }
-  __arm_streaming friend int f3(), f4(), f5(); // expected-error {{'__arm_streaming' cannot appear here}}
-  friend int f6 __arm_streaming (), f7 __arm_streaming (), f8 __arm_streaming (); // expected-error3 {{'__arm_streaming' cannot appear here}} \
-                                                                                     expected-error 3 {{'__arm_streaming' cannot be applied to a declaration}}
-  friend class __arm_streaming C; // expected-error {{'__arm_streaming' cannot appear here}}
-  __arm_streaming friend class D; // expected-error {{'__arm_streaming' cannot appear here}}
-  __arm_streaming friend int; // expected-error {{'__arm_streaming' cannot appear here}}
+  ATTR_USE friend int f3(), f4(), f5(); // expected-error {{'ATTR_NAME' cannot appear here}}
+  friend int f6 ATTR_USE (), f7 ATTR_USE (), f8 ATTR_USE (); // expected-error3 {{'ATTR_NAME' cannot appear here}} \
+                                                                                     expected-error 3 {{'ATTR_NAME' cannot be applied to a declaration}}
+  friend class ATTR_USE C; // expected-error {{'ATTR_NAME' cannot appear here}}
+  ATTR_USE friend class D; // expected-error {{'ATTR_NAME' cannot appear here}}
+  ATTR_USE friend int; // expected-error {{'ATTR_NAME' cannot appear here}}
 };
 template void tmpl (T) {}
-template __arm_streaming void tmpl(char); // expected-error {{'__arm_streaming' cannot appear here}}
-template void __arm_streaming tmpl(short); // expected-error {{'__arm_streaming' only applies to function types}}
+template ATTR_USE void tmpl(char); // expected-error {{'ATTR_NAME' cannot appear here}}
+template void ATTR_USE tmpl(short); // expected-error {{'ATTR_NAME' only applies to function types}}
 
 // Statement tests
 void foo () {
-  __arm_streaming ; // expected-error {{'__arm_streaming' cannot be applied to a statement}}
-  __arm_streaming { } // expected-error {{'__arm_streaming' cannot be applied to a statement}}
-  __arm_streaming if (0) { } // expected-error {{'__arm_streaming' cannot be applied to a statement}}
-  __arm_streaming for (;;); // expected-error {{'__arm_streaming' cannot be applied to a statement}}
-  __arm_streaming do { // expected-error {{'__arm_streaming' cannot be applied to a statement}}
-    __arm_streaming continue; // expected-error {{'__arm_streaming' cannot be applied to a statement}}
+  ATTR_USE ; // expected-error {{'ATTR_NAME' cannot be applied to a statement}}
+  ATTR_USE { } // expected-error {{'ATTR_NAME' cannot be applied to a statement}}
+  ATTR_USE if (0) { } // expected-error {{'ATTR_NAME' cannot be applied to a statement}}
+  ATTR_USE for (;;); // expected-error {{'ATTR_NAME' cannot be applied to a statement}}
+  ATTR_USE do { // expected-error {{'ATTR_NAME' cannot be applied to a statement}}
+    ATTR_USE continue; // expected-error {{'ATTR_NAME' cannot be applied to a statement}}
   } while (0);
-  __arm_streaming while (0); // expected-error {{'__arm_streaming' cannot be applied to a statement}}
+  ATTR_USE while (0); // expected-error {{'ATTR_NAME' cannot be applied to a statement}}
 
-  __arm_streaming switch (i) { // expected-error {{'__arm_streaming' cannot be applied to a statement}}
-    __arm_streaming case 0: // expected-error {{'__arm_streaming' cannot be applied to a statement}}
-    __arm_streaming default: // expected-error {{'__arm_streaming' cannot be applied to a statement}}
-      __arm_streaming break; // expected-error {{'__arm_streaming' cannot be applied to a statement}}
+  ATTR_USE switch (i) { // expected-error {{'ATTR_NAME' cannot be applied to a statement}}
+    ATTR_USE case 0: // expected-error {{'ATTR_NAME' cannot be applied to a statement}}
+    ATTR_USE default: // expected-error {{'ATTR_NAME' cannot be applied to a statement}}
+      ATTR_USE break; // expected-error {{'ATTR_NAME' cannot be applied to a statement}}
   }
 
-  __arm_streaming goto there; // expected-error {{'__arm_streaming' cannot be applied to a statement}}
-  __arm_streaming there: // expected-error {{'__arm_streaming' only applies to non-K&R-style functions}}
+  ATTR_USE goto there; // expected-error {{'ATTR_NAME' cannot be applied to a statement}}
+  ATTR_USE there: // expected-error {{'ATTR_NAME' only applies to non-K&R-style functions}}
 
-  __arm_streaming try { // expected-error {{'__arm_streaming' cannot be applied to a statement}}
-  } __arm_streaming catch (...) { // expected-error {{'__arm_streaming' cannot appear here}}
+  ATTR_USE try { // expected-error {{'ATTR_NAME' cannot be applied to a statement}}
+  } ATTR_USE catch (...) { // expected-error {{'ATTR_NAME' cannot appear here}}
   }
 
-  void bar __arm_streaming (__arm_streaming int i, __arm_streaming int j); // expected-error 2 {{'__arm_streaming' only applies to function types}} \
-                                                                              expected-error {{'__arm_streaming' cannot be applied to a declaration}}
-  using FuncType = void (__arm_streaming int); // expected-error {{'__arm_streaming' only applies to function types}}
-  void baz(__arm_streaming...); // expected-error {{expected parameter declarator}}
+  void bar ATTR_USE (ATTR_USE int i, ATTR_USE int j); // expected-error 2 {{'ATTR_NAME' only applies to function types}} \
+                                                                              expected-error {{'ATTR_NAME' cannot be applied to a declaration}}
+  using FuncType = void (ATTR_USE int); // expected-error {{'ATTR_NAME' only applies to function types}}
+  void baz(ATTR_USE...); // expected-error {{expected parameter declarator}}
 
-  __arm_streaming return; // expected-error {{'__arm_streaming' cannot be applied to a statement}}
+  ATTR_USE return; // expected-error {{'ATTR_NAME' cannot be applied to a statement}}
 }
 
 // Expression tests
 void bar () {
-  new int[42]__arm_streaming[5]__arm_streaming{}; // expected-error {{'__arm_streaming' only applies to function types}}
+  new int[42]ATTR_USE[5]ATTR_USE{}; // expected-error {{'ATTR_NAME' only applies to function types}}
 }
 
 // Condition tests
 void baz () {
-  if (__arm_streaming bool b = true) { // expected-error {{'__arm_streaming' only applies to function types}}
-    switch (__arm_streaming int n { 42 }) { // expected-error {{'__arm_streaming' only applies to function types}}
+  if (ATTR_USE bool b = true) { // expected-error {{'ATTR_NAME' only applies to function types}}
+    switch (ATTR_USE int n { 42 }) { // expected-error {{'ATTR_NAME' only applies to function types}}
     default:
-      for (__arm_streaming int n = 0; __arm_streaming char b = n < 5; ++b) { // expected-error 2 {{'__arm_streaming' only applies to function types}}
+      for (ATTR_USE int n = 0; ATTR_USE char b = n < 5; ++b) { // expected-error 2 {{'ATTR_NAME' only applies to function types}}
       }
     }
   }
@@ -261,37 +264,37 @@ void baz () {
   // An attribute can be applied to an expression-statement, such as the first
   // statement in a for. But it can't be applied to a condition which is an
   // expression.
-  for (__arm_streaming x = 0; ; ) {} // expected-error {{'__arm_streaming' cannot appear here}}
-  for (; __arm_streaming x < 5; ) {} // expected-error {{'__arm_streaming' cannot appear here}}
-  while (__arm_streaming bool k { false }) { // expected-error {{'__arm_streaming' only applies to function types}}
+  for (ATTR_USE x = 0; ; ) {} // expected-error {{'ATTR_NAME' cannot appear here}}
+  for (; ATTR_USE x < 5; ) {} // expected-error {{'ATTR_NAME' cannot appear here}}
+  while (ATTR_USE bool k { false }) { // expected-error {{'ATTR_NAME' only applies to function types}}
   }
-  while (__arm_streaming true) { // expected-error {{'__arm_streaming' cannot appear here}}
+  while (ATTR_USE true) { // expected-error {{'ATTR_NAME' cannot appear here}}
   }
   do {
-  } while (__arm_streaming false); // expected-error {{'__arm_streaming' cannot appear here}}
+  } while (ATTR_USE false); // expected-error {{'ATTR_NAME' cannot appear here}}
 
-  for (__arm_streaming int n : { 1, 2, 3 }) { // expected-error {{'__arm_streaming' only applies to function types}}
+  for (ATTR_USE int n : { 1, 2, 3 }) { // expected-error {{'ATTR_NAME' only applies to function types}}
   }
 }
 
 enum class __attribute__((visibility("hidden"))) SecretKeepers {
   one, /* rest are deprecated */ two, three
 };
-enum class __arm_streaming EvenMoreSecrets {}; // expected-error {{'__arm_streaming' only applies to non-K&R-style functions}}
+enum class ATTR_USE EvenMoreSecrets {}; // expected-error {{'ATTR_NAME' only applies to non-K&R-style functions}}
 
 // Forbid attributes on decl specifiers.
-unsigned __arm_streaming static int __arm_streaming v1; // expected-error {{'__arm_streaming' only applies to function types}} \
-           expected-error {{'__arm_streaming' cannot appear here}}
-typedef __arm_streaming unsigned long __arm_streaming v2; // expected-error {{'__arm_streaming' only applies to function types}} \
-          expected-error {{'__arm_streaming' cannot appear here}}
-int __arm_streaming foo(int __arm_streaming x); // expected-error 2 {{'__arm_streaming' only applies to function types}}
+unsigned ATTR_USE static int ATTR_USE v1; // expected-error {{'ATTR_NAME' only applies to function types}} \
+           expected-error {{'ATTR_NAME' cannot appear here}}
+typedef ATTR_USE unsigned long ATTR_USE v2; // expected-error {{'ATTR_NAME' only applies to function types}} \
+          expected-error {{'ATTR_NAME' cannot appear here}}
+int ATTR_USE foo(int ATTR_USE x); // expected-error 2 {{'ATTR_NAME' only applies to function types}}
 
-__arm_streaming; // expected-error {{'__arm_streaming' only applies to non-K&R-style functions}}
+ATTR_USE; // expected-error {{'ATTR_NAME' only applies to non-K&R-style functions}}
 
 class A {
-  A(__arm_streaming int a); // expected-error {{'__arm_streaming' only applies to function types}}
+  A(ATTR_USE int a); // expected-error {{'ATTR_NAME' only applies to function types}}
 };
-A::A(__arm_streaming int a) {} // expected-error {{'__arm_streaming' only applies to function types}}
+A::A(ATTR_USE int a) {} // expected-error {{'ATTR_NAME' only applies to function types}}
 
 template struct TemplateStruct {};
 class FriendClassesWithAttributes {
@@ -299,47 +302,47 @@ class FriendClassesWithAttributes {
   template  friend class __attribute__((__type_visibility__("default"))) vector;
   template  friend class __declspec(code_seg("foo,whatever")) vector2;
   // But not C++11 ones
-  template  friend class __arm_streaming vector3;                                         // expected-error {{'__arm_streaming' cannot appear here}}
+  template  friend class ATTR_USE vector3;                                         // expected-error {{'ATTR_NAME' cannot appear here}}
 
   // Also allowed
   friend struct __attribute__((__type_visibility__("default"))) TemplateStruct;
   friend struct __declspec(code_seg("foo,whatever")) TemplateStruct;
-  friend struct __arm_streaming TemplateStruct;                                       // expected-error {{'__arm_streaming' cannot appear here}}
+  friend struct ATTR_USE TemplateStruct;                                       // expected-error {{'ATTR_NAME' cannot appear here}}
 };
 
 // Check ordering: C++11 attributes must appear before GNU attributes.
 class Ordering {
   void f1(
-    int (__arm_streaming __attribute__(()) int n) // expected-error {{'__arm_streaming' only applies to function types}}
+    int (ATTR_USE __attribute__(()) int n) // expected-error {{'ATTR_NAME' only applies to function types}}
   ) {
   }
 
   void f2(
-      int (*)(__arm_streaming __attribute__(()) int n) // expected-error {{'__arm_streaming' only applies to function types}}
+      int (*)(ATTR_USE __attribute__(()) int n) // expected-error {{'ATTR_NAME' only applies to function types}}
   ) {
   }
 
   void f3(
-    int (__attribute__(()) __arm_streaming int n) // expected-error {{'__arm_streaming' cannot appear here}}
+    int (__attribute__(()) ATTR_USE int n) // expected-error {{'ATTR_NAME' cannot appear here}}
   ) {
   }
 
   void f4(
-      int (*)(__attribute__(()) __arm_streaming int n) // expected-error {{'__arm_streaming' cannot appear here}}
+      int (*)(__attribute__(()) ATTR_USE int n) // expected-error {{'ATTR_NAME' cannot appear here}}
   ) {
   }
 };
 
 namespace base_specs {
 struct A {};
-struct B : __arm_streaming A {}; // expected-error {{'__arm_streaming' cannot be applied to a base specifier}}
-struct C : __arm_streaming virtual A {}; // expected-error {{'__arm_streaming' cannot be applied to a base specifier}}
-struct D : __arm_streaming public virtual A {}; // expected-error {{'__arm_streaming' cannot be applied to a base specifier}}
-struct E : public __arm_streaming virtual A {}; // expected-error {{'__arm_streaming' cannot appear here}} \
-                                                   expected-error {{'__arm_streaming' cannot be applied to a base specifier}}
-struct F : virtual __arm_streaming public A {}; // expected-error {{'__arm_streaming' cannot appear here}} \
-                                                   expected-error {{'__arm_streaming' cannot be applied to a base specifier}}
+struct B : ATTR_USE A {}; // expected-error {{'ATTR_NAME' cannot be applied to a base specifier}}
+struct C : ATTR_USE virtual A {}; // expected-error {{'ATTR_NAME' cannot be applied to a base specifier}}
+struct D : ATTR_USE public virtual A {}; // expected-error {{'ATTR_NAME' cannot be applied to a base specifier}}
+struct E : public ATTR_USE virtual A {}; // expected-error {{'ATTR_NAME' cannot appear here}} \
+                                                   expected-error {{'ATTR_NAME' cannot be applied to a base specifier}}
+struct F : virtual ATTR_USE public A {}; // expected-error {{'ATTR_NAME' cannot appear here}} \
+                                                   expected-error {{'ATTR_NAME' cannot be applied to a base specifier}}
 }
 
-namespace __arm_streaming ns_attr {}; // expected-error {{'__arm_streaming' only applies to non-K&R-style functions}} \
+namespace ATTR_USE ns_attr {}; // expected-error {{'ATTR_NAME' only applies to non-K&R-style functions}} \
                                          expected-warning {{attributes on a namespace declaration are a C++17 extension}}
diff --git a/clang/test/ParserOpenACC/parse-cache-construct.c b/clang/test/ParserOpenACC/parse-cache-construct.c
index d54632fc8f466..093587f37df4f 100644
--- a/clang/test/ParserOpenACC/parse-cache-construct.c
+++ b/clang/test/ParserOpenACC/parse-cache-construct.c
@@ -114,6 +114,18 @@ void func() {
     #pragma acc cache(readonly)
   }
 
+  for (int i = 0; i < 10; ++i) {
+    // expected-error@+2{{invalid tag 'devnum' on 'cache' directive}}
+    // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+    #pragma acc cache(devnum:ArrayPtr)
+  }
+
+  for (int i = 0; i < 10; ++i) {
+    // expected-error@+2{{invalid tag 'invalid' on 'cache' directive}}
+    // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+    #pragma acc cache(invalid:ArrayPtr)
+  }
+
   for (int i = 0; i < 10; ++i) {
     // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
     #pragma acc cache(readonly:ArrayPtr)
diff --git a/clang/test/ParserOpenACC/parse-clauses.c b/clang/test/ParserOpenACC/parse-clauses.c
index 11e89d420e6b0..22dce4c4d8cd9 100644
--- a/clang/test/ParserOpenACC/parse-clauses.c
+++ b/clang/test/ParserOpenACC/parse-clauses.c
@@ -54,6 +54,45 @@ void func() {
   // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
 #pragma acc loop seq,
 
+  // expected-error@+2{{expected '('}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc loop collapse
+  for(;;){}
+
+  // expected-error@+2{{expected expression}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc loop collapse()
+  for(;;){}
+
+  // expected-error@+3{{invalid tag 'unknown' on 'collapse' clause}}
+  // expected-error@+2{{expected expression}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc loop collapse(unknown:)
+  for(;;){}
+
+  // expected-error@+2{{expected expression}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc loop collapse(force:)
+  for(;;){}
+
+  // expected-error@+2{{invalid tag 'unknown' on 'collapse' clause}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc loop collapse(unknown:5)
+  for(;;){}
+
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc loop collapse(force:5)
+  for(;;){}
+
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc loop collapse(5)
+  for(;;){}
+
+  // expected-error@+3{{expected ')'}}
+  // expected-note@+2{{to match this '('}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc loop collapse(5, 6)
+  for(;;){}
 }
 
 void DefaultClause() {
@@ -306,6 +345,11 @@ void SyncClause() {
 #pragma acc serial self(i > j, seq
   for(;;){}
 
+  // expected-warning@+2{{left operand of comma operator has no effect}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial self(i, j)
+  for(;;){}
+
   // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
 #pragma acc serial self(i > j)
   for(;;){}
@@ -315,9 +359,373 @@ void SyncClause() {
   for(;;){}
 }
 
+struct Members {
+  int value;
+  char array[5];
+};
+struct HasMembersArray {
+  struct Members MemArr[4];
+};
+
+// On 'update', self behaves differently and requires parens, plus takes a var-list instead.
+void SelfUpdate() {
+  struct Members s;
+
+  // expected-error@+2{{expected '('}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc update self
+  for(;;){}
+
+  // expected-error@+2{{use of undeclared identifier 'zero'}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc update self(zero : s.array[s.value : 5], s.value), seq
+  for(;;){}
+
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc update self(s.array[s.value : 5], s.value), seq
+  for(;;){}
+}
+
+void VarListClauses() {
+  // expected-error@+2{{expected '('}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial copy
+
+  // expected-error@+2{{expected '('}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial copy, seq
+
+  // expected-error@+3{{expected '('}}
+  // expected-error@+2{{expected identifier}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial copy)
+
+  // expected-error@+3{{expected '('}}
+  // expected-error@+2{{expected identifier}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial copy), seq
+
+  // expected-error@+2{{expected expression}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial copy(
+
+  // expected-error@+2{{expected expression}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial copy(, seq
+
+  // expected-error@+2{{expected expression}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial copy()
+
+  // expected-error@+2{{expected expression}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial copy(), seq
+
+  struct Members s;
+  struct HasMembersArray HasMem;
+
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial copy(s.array[s.value]), seq
+
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial copy(s.array[s.value], s.array[s.value :5] ), seq
+
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial copy(HasMem.MemArr[3].array[1]), seq
+
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial copy(HasMem.MemArr[3].array[1:4]), seq
+
+  // expected-error@+2{{OpenMP array section is not allowed here}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial copy(HasMem.MemArr[1:3].array[1]), seq
+
+  // expected-error@+2{{OpenMP array section is not allowed here}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial copy(HasMem.MemArr[1:3].array[1:2]), seq
+
+  // expected-error@+2{{expected expression}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial copy(HasMem.MemArr[:]), seq
+
+  // expected-error@+2{{expected expression}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial copy(HasMem.MemArr[::]), seq
+
+  // expected-error@+4{{expected expression}}
+  // expected-error@+3{{expected ']'}}
+  // expected-note@+2{{to match this '['}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial copy(HasMem.MemArr[: :]), seq
+
+  // expected-error@+2{{expected expression}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial copy(HasMem.MemArr[3:]), seq
+
+  // expected-error@+2{{expected ','}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial use_device(s.array[s.value] s.array[s.value :5] ), seq
+
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial use_device(s.array[s.value : 5]), seq
+
+  // expected-error@+2{{expected ','}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial no_create(s.array[s.value] s.array[s.value :5] ), seq
+
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial no_create(s.array[s.value : 5], s.value), seq
+
+  // expected-error@+2{{expected ','}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial present(s.array[s.value] s.array[s.value :5] ), seq
+
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial present(s.array[s.value : 5], s.value), seq
+
+  // expected-error@+2{{expected ','}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial deviceptr(s.array[s.value] s.array[s.value :5] ), seq
+
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial deviceptr(s.array[s.value : 5], s.value), seq
+
+  // expected-error@+2{{expected ','}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial attach(s.array[s.value] s.array[s.value :5] ), seq
+
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial attach(s.array[s.value : 5], s.value), seq
+
+  // expected-error@+2{{expected ','}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial detach(s.array[s.value] s.array[s.value :5] ), seq
+
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial detach(s.array[s.value : 5], s.value), seq
+
+  // expected-error@+2{{expected ','}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial private(s.array[s.value] s.array[s.value :5] ), seq
+
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial private(s.array[s.value : 5], s.value), seq
+
+  // expected-error@+2{{expected ','}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial firstprivate(s.array[s.value] s.array[s.value :5] ), seq
+
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial firstprivate(s.array[s.value : 5], s.value), seq
+
+  // expected-error@+2{{expected ','}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial delete(s.array[s.value] s.array[s.value :5] ), seq
+
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial delete(s.array[s.value : 5], s.value), seq
+
+  // expected-error@+2{{expected ','}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial use_device(s.array[s.value] s.array[s.value :5] ), seq
+
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial use_device(s.array[s.value : 5], s.value), seq
+
+  // expected-error@+2{{expected ','}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial device_resident(s.array[s.value] s.array[s.value :5] ), seq
+
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial device_resident(s.array[s.value : 5], s.value), seq
+
+  // expected-error@+2{{expected ','}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial link(s.array[s.value] s.array[s.value :5] ), seq
+
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial link(s.array[s.value : 5], s.value), seq
+
+  // expected-error@+2{{expected ','}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial host(s.array[s.value] s.array[s.value :5] ), seq
+
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial host(s.array[s.value : 5], s.value), seq
+
+  // expected-error@+2{{expected ','}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial device(s.array[s.value] s.array[s.value :5] ), seq
+
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial device(s.array[s.value : 5], s.value), seq
+
+  // expected-error@+2{{expected ','}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial copyout(s.array[s.value] s.array[s.value :5] ), seq
+
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial copyout(s.array[s.value : 5], s.value), seq
+
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial copyout(zero:s.array[s.value : 5], s.value), seq
+
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial copyout(zero : s.array[s.value : 5], s.value), seq
+
+  // expected-error@+2{{use of undeclared identifier 'zero'}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial copyout(zero s.array[s.value : 5], s.value), seq
+
+  // expected-error@+2{{invalid tag 'readonly' on 'copyout' clause}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial copyout(readonly:s.array[s.value : 5], s.value), seq
+
+  // expected-error@+2{{invalid tag 'invalid' on 'copyout' clause}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial copyout(invalid:s.array[s.value : 5], s.value), seq
+
+  // expected-error@+2{{invalid tag 'invalid' on 'copyout' clause}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial copyout(invalid:s.array[s.value : 5], s.value), seq
+
+  // expected-error@+2{{use of undeclared identifier 'invalid'}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial copyout(invalid s.array[s.value : 5], s.value), seq
+
+  // expected-error@+2{{expected ','}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial create(s.array[s.value] s.array[s.value :5] ), seq
+
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial create(s.array[s.value : 5], s.value), seq
+
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial create(zero:s.array[s.value : 5], s.value), seq
+
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial create(zero : s.array[s.value : 5], s.value), seq
+
+  // expected-error@+2{{use of undeclared identifier 'zero'}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial create(zero s.array[s.value : 5], s.value), seq
+
+  // expected-error@+2{{invalid tag 'readonly' on 'create' clause}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial create(readonly:s.array[s.value : 5], s.value), seq
+
+  // expected-error@+2{{invalid tag 'invalid' on 'create' clause}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial create(invalid:s.array[s.value : 5], s.value), seq
+
+  // expected-error@+2{{invalid tag 'invalid' on 'create' clause}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial create(invalid:s.array[s.value : 5], s.value), seq
+
+  // expected-error@+2{{use of undeclared identifier 'invalid'}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial create(invalid s.array[s.value : 5], s.value), seq
+
+  // expected-error@+2{{expected ','}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial copyin(s.array[s.value] s.array[s.value :5] ), seq
+
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial copyin(s.array[s.value : 5], s.value), seq
+
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial copyin(readonly:s.array[s.value : 5], s.value), seq
+
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial copyin(readonly : s.array[s.value : 5], s.value), seq
+
+  // expected-error@+2{{use of undeclared identifier 'readonly'}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial copyin(readonly s.array[s.value : 5], s.value), seq
+
+  // expected-error@+2{{invalid tag 'zero' on 'copyin' clause}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial copyin(zero :s.array[s.value : 5], s.value), seq
+
+  // expected-error@+2{{invalid tag 'invalid' on 'copyin' clause}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial copyin(invalid:s.array[s.value : 5], s.value), seq
+
+  // expected-error@+2{{invalid tag 'invalid' on 'copyin' clause}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial copyin(invalid:s.array[s.value : 5], s.value), seq
+
+  // expected-error@+2{{use of undeclared identifier 'invalid'}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial copyin(invalid s.array[s.value : 5], s.value), seq
+}
+
+void ReductionClauseParsing() {
+  char *Begin, *End;
+  // expected-error@+2{{expected '('}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial reduction
+  // expected-error@+3{{missing reduction operator, expected '+', '*', 'max', 'min', '&', '|', '^', '&&', or '||', follwed by a ':'}}
+  // expected-error@+2{{expected expression}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial reduction()
+  // expected-error@+2{{missing reduction operator, expected '+', '*', 'max', 'min', '&', '|', '^', '&&', or '||', follwed by a ':'}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial reduction(Begin)
+  // expected-error@+2{{missing reduction operator, expected '+', '*', 'max', 'min', '&', '|', '^', '&&', or '||', follwed by a ':'}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial reduction(Begin, End)
+  // expected-error@+2{{missing reduction operator, expected '+', '*', 'max', 'min', '&', '|', '^', '&&', or '||', follwed by a ':'}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial reduction(Begin, End)
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial reduction(+:Begin)
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial reduction(+:Begin, End)
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial reduction(*: Begin, End)
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial reduction(max : Begin, End)
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial reduction(min: Begin, End)
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial reduction(&: Begin, End)
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial reduction(|: Begin, End)
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial reduction(^: Begin, End)
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial seq, reduction(&&: Begin, End)
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc serial reduction(||: Begin, End), seq
+}
+
   // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
 #pragma acc routine worker, vector, seq, nohost
 void bar();
 
   // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
 #pragma acc routine(bar) worker, vector, seq, nohost
+
+
+// Bind Clause Parsing.
+
+  // expected-error@+2{{expected '('}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc routine bind
+void BCP1();
+
+  // expected-error@+2{{expected identifier or string literal}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc routine(BCP1) bind()
+
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc routine bind("ReductionClauseParsing")
+void BCP2();
+
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc routine(BCP1) bind(BCP2)
+
+  // expected-error@+2{{use of undeclared identifier 'unknown_thing'}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc routine(BCP1) bind(unknown_thing)
diff --git a/clang/test/ParserOpenACC/parse-clauses.cpp b/clang/test/ParserOpenACC/parse-clauses.cpp
new file mode 100644
index 0000000000000..1b3196094103c
--- /dev/null
+++ b/clang/test/ParserOpenACC/parse-clauses.cpp
@@ -0,0 +1,73 @@
+// RUN: %clang_cc1 %s -verify -fopenacc
+
+template
+void templ() {
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc loop collapse(I)
+  for(;;){}
+
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc loop collapse(T::value)
+  for(;;){}
+}
+
+struct S {
+  static constexpr unsigned value = 5;
+};
+
+void use() {
+  templ<7, S>();
+}
+
+namespace NS {
+void NSFunc();
+
+class RecordTy { // #RecTy
+  static constexpr bool Value = false; // #VAL
+  void priv_mem_function(); // #PrivMemFun
+  public:
+  static constexpr bool ValuePub = true;
+  void mem_function();
+};
+template
+class TemplTy{};
+void function();
+}
+
+
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc routine(use) bind(NS::NSFunc)
+  // expected-error@+3{{'RecordTy' does not refer to a value}}
+  // expected-note@#RecTy{{declared here}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc routine(use) bind(NS::RecordTy)
+  // expected-error@+3{{'Value' is a private member of 'NS::RecordTy'}}
+  // expected-note@#VAL{{implicitly declared private here}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc routine(use) bind(NS::RecordTy::Value)
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc routine(use) bind(NS::RecordTy::ValuePub)
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc routine(use) bind(NS::TemplTy)
+  // expected-error@+2{{no member named 'unknown' in namespace 'NS'}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc routine(use) bind(NS::unknown)
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc routine(use) bind(NS::function)
+  // expected-error@+3{{'priv_mem_function' is a private member of 'NS::RecordTy'}}
+  // expected-note@#PrivMemFun{{implicitly declared private here}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc routine(use) bind(NS::RecordTy::priv_mem_function)
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc routine(use) bind(NS::RecordTy::mem_function)
+
+  // expected-error@+2{{string literal with user-defined suffix cannot be used here}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc routine(use) bind("unknown udl"_UDL)
+
+  // expected-warning@+2{{encoding prefix 'u' on an unevaluated string literal has no effect}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc routine(use) bind(u"16 bits")
+  // expected-warning@+2{{encoding prefix 'U' on an unevaluated string literal has no effect}}
+  // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc routine(use) bind(U"32 bits")
diff --git a/clang/test/ParserOpenACC/parse-constructs.cpp b/clang/test/ParserOpenACC/parse-constructs.cpp
index f26f9aee8546b..abeddd357619e 100644
--- a/clang/test/ParserOpenACC/parse-constructs.cpp
+++ b/clang/test/ParserOpenACC/parse-constructs.cpp
@@ -5,6 +5,12 @@ namespace NS {
 
   template
   void templ(); // expected-note 2{{declared here}}
+
+  class C { // #CDef
+    void private_mem_func(); // #PrivateMemFunc
+    public:
+    void public_mem_func();
+  };
 }
 
 // expected-error@+2{{use of undeclared identifier 'foo'; did you mean 'NS::foo'?}}
@@ -44,3 +50,14 @@ namespace NS {
 // expected-error@+2 {{expected unqualified-id}}
 // expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
 #pragma acc routine(int)
+
+// expected-error@+3{{'C' does not refer to a value}}
+// expected-note@#CDef{{declared here}}
+// expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc routine (NS::C)
+// expected-error@+3{{'private_mem_func' is a private member of 'NS::C'}}
+// expected-note@#PrivateMemFunc{{implicitly declared private here}}
+// expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc routine (NS::C::private_mem_func)
+// expected-warning@+1{{OpenACC directives not yet implemented, pragma ignored}}
+#pragma acc routine (NS::C::public_mem_func)
diff --git a/clang/test/Preprocessor/aarch64-target-features.c b/clang/test/Preprocessor/aarch64-target-features.c
index 96c7e39a18af7..15879da04fcf0 100644
--- a/clang/test/Preprocessor/aarch64-target-features.c
+++ b/clang/test/Preprocessor/aarch64-target-features.c
@@ -50,6 +50,7 @@
 // CHECK-NOT: __ARM_FEATURE_DOTPROD
 // CHECK-NOT: __ARM_FEATURE_PAC_DEFAULT
 // CHECK-NOT: __ARM_FEATURE_BTI_DEFAULT
+// CHECK-NOT: __ARM_FEATURE_GCS_DEFAULT
 // CHECK-NOT: __ARM_BF16_FORMAT_ALTERNATIVE 1
 // CHECK-NOT: __ARM_FEATURE_BF16 1
 // CHECK-NOT: __ARM_FEATURE_BF16_VECTOR_ARITHMETIC 1
@@ -284,7 +285,7 @@
 
 // ================== Check whether -mtune accepts mixed-case features.
 // RUN: %clang -target aarch64 -mtune=CYCLONE -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-MTUNE-CYCLONE %s
-// CHECK-MTUNE-CYCLONE: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+neon" "-target-feature" "+v8a" "-target-feature" "+zcm" "-target-feature" "+zcz"
+// CHECK-MTUNE-CYCLONE: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+zcm" "-target-feature" "+zcz" "-target-feature" "+v8a"
 
 // RUN: %clang -target aarch64 -mcpu=apple-a7 -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-MCPU-APPLE-A7 %s
 // RUN: %clang -target aarch64 -mcpu=apple-a8 -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-MCPU-APPLE-A7 %s
@@ -310,18 +311,18 @@
 // RUN: %clang -target aarch64 -mcpu=thunderx2t99 -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-MCPU-THUNDERX2T99 %s
 // RUN: %clang -target aarch64 -mcpu=a64fx -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-MCPU-A64FX %s
 // RUN: %clang -target aarch64 -mcpu=carmel -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-MCPU-CARMEL %s
-// CHECK-MCPU-APPLE-A7: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8a" "-target-feature" "+aes" "-target-feature" "+fp-armv8" "-target-feature" "+sha2" "-target-feature" "+neon" "-target-feature" "+zcm" "-target-feature" "+zcz"
-// CHECK-MCPU-APPLE-A10: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8a" "-target-feature" "+aes" "-target-feature" "+crc" "-target-feature" "+fp-armv8" "-target-feature" "+rdm" "-target-feature" "+sha2" "-target-feature" "+neon" "-target-feature" "+zcm" "-target-feature" "+zcz"
-// CHECK-MCPU-APPLE-A11: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8.2a" "-target-feature" "+aes" "-target-feature" "+crc" "-target-feature" "+fp-armv8" "-target-feature" "+fullfp16" "-target-feature" "+lse" "-target-feature" "+ras" "-target-feature" "+rdm" "-target-feature" "+sha2" "-target-feature" "+neon" "-target-feature" "+zcm" "-target-feature"
-// CHECK-MCPU-APPLE-A12: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8.3a" "-target-feature" "+aes" "-target-feature" "+crc" "-target-feature" "+fp-armv8" "-target-feature" "+fullfp16" "-target-feature" "+lse" "-target-feature" "+ras" "-target-feature" "+rcpc" "-target-feature" "+rdm" "-target-feature" "+sha2" "-target-feature" "+neon" "-target-feature" "+zcm" "-target-feature" "+zcz"
+// CHECK-MCPU-APPLE-A7: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+zcm" "-target-feature" "+zcz" "-target-feature" "+v8a" "-target-feature" "+aes"{{.*}} "-target-feature" "+fp-armv8" "-target-feature" "+sha2" "-target-feature" "+neon"
+// CHECK-MCPU-APPLE-A10: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+zcm" "-target-feature" "+zcz" "-target-feature" "+v8a" "-target-feature" "+aes"{{.*}} "-target-feature" "+crc" "-target-feature" "+fp-armv8" "-target-feature" "+rdm" "-target-feature" "+sha2" "-target-feature" "+neon"
+// CHECK-MCPU-APPLE-A11: "-cc1"{{.*}} "-triple" "aarch64{{.*}}"{{.*}}"-target-feature" "+zcm" "-target-feature" "+zcz" "-target-feature" "+v8.2a" "-target-feature" "+aes" "-target-feature" "+crc" "-target-feature" "+fp-armv8" "-target-feature" "+fullfp16" "-target-feature" "+lse" "-target-feature" "+ras" "-target-feature" "+rdm" "-target-feature" "+sha2" "-target-feature" "+neon"
+// CHECK-MCPU-APPLE-A12: "-cc1"{{.*}} "-triple" "aarch64"{{.*}} "-target-feature" "+zcm" "-target-feature" "+zcz" "-target-feature" "+v8.3a" "-target-feature" "+aes" "-target-feature" "+crc" "-target-feature" "+complxnum" "-target-feature" "+fp-armv8" "-target-feature" "+fullfp16" "-target-feature" "+jsconv" "-target-feature" "+lse" "-target-feature" "+ras" "-target-feature" "+rcpc" "-target-feature" "+rdm" "-target-feature" "+sha2" "-target-feature" "+neon"
 // CHECK-MCPU-A34: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+aes" "-target-feature" "+crc" "-target-feature" "+fp-armv8" "-target-feature" "+sha2" "-target-feature" "+neon"
-// CHECK-MCPU-APPLE-A13: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8.4a" "-target-feature" "+aes" "-target-feature" "+crc" "-target-feature" "+dotprod" "-target-feature" "+fp-armv8" "-target-feature" "+fp16fml" "-target-feature" "+lse" "-target-feature" "+ras" "-target-feature" "+rcpc" "-target-feature" "+rdm"  "-target-feature" "+sha2" "-target-feature" "+sha3" "-target-feature" "+neon" "-target-feature" "+zcm" "-target-feature" "+zcz" "-target-feature" "+fullfp16"
+// CHECK-MCPU-APPLE-A13: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-cpu" "apple-a13" "-target-feature" "+zcm" "-target-feature" "+zcz" "-target-feature" "+v8.4a" "-target-feature" "+aes" "-target-feature" "+crc" "-target-feature" "+dotprod" "-target-feature" "+complxnum" "-target-feature" "+fp-armv8" "-target-feature" "+fullfp16" "-target-feature" "+fp16fml" "-target-feature" "+jsconv" "-target-feature" "+lse" "-target-feature" "+ras" "-target-feature" "+rcpc" "-target-feature" "+rdm" "-target-feature" "+sha2" "-target-feature" "+sha3" "-target-feature" "+neon"
 // CHECK-MCPU-A35: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8a" "-target-feature" "+aes" "-target-feature" "+crc" "-target-feature" "+fp-armv8" "-target-feature" "+sha2" "-target-feature" "+neon"
 // CHECK-MCPU-A53: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8a" "-target-feature" "+aes" "-target-feature" "+crc" "-target-feature" "+fp-armv8" "-target-feature" "+sha2" "-target-feature" "+neon"
 // CHECK-MCPU-A57: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8a" "-target-feature" "+aes" "-target-feature" "+crc" "-target-feature" "+fp-armv8" "-target-feature" "+sha2" "-target-feature" "+neon"
 // CHECK-MCPU-A72: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8a" "-target-feature" "+aes" "-target-feature" "+crc" "-target-feature" "+fp-armv8" "-target-feature" "+sha2" "-target-feature" "+neon"
 // CHECK-MCPU-CORTEX-A73: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8a" "-target-feature" "+aes" "-target-feature" "+crc" "-target-feature" "+fp-armv8" "-target-feature" "+sha2" "-target-feature" "+neon"
-// CHECK-MCPU-CORTEX-R82: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8r" "-target-feature" "+crc" "-target-feature" "+dotprod" "-target-feature" "+fp-armv8" "-target-feature" "+fp16fml" "-target-feature" "+lse" "-target-feature" "+ras" "-target-feature" "+rcpc" "-target-feature" "+rdm" "-target-feature" "+sb" "-target-feature" "+neon" "-target-feature" "+ssbs" "-target-feature" "+fullfp16"
+// CHECK-MCPU-CORTEX-R82: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8r" "-target-feature" "+crc" "-target-feature" "+dotprod" "-target-feature" "+complxnum" "-target-feature" "+fp-armv8" "-target-feature" "+fullfp16" "-target-feature" "+fp16fml" "-target-feature" "+jsconv" "-target-feature" "+lse" "-target-feature" "+ras" "-target-feature" "+rcpc" "-target-feature" "+rdm" "-target-feature" "+sb" "-target-feature" "+neon" "-target-feature" "+ssbs"
 // CHECK-MCPU-M3: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8a" "-target-feature" "+aes" "-target-feature" "+crc" "-target-feature" "+fp-armv8" "-target-feature" "+sha2" "-target-feature" "+neon"
 // CHECK-MCPU-M4: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8.2a" "-target-feature" "+aes" "-target-feature" "+crc" "-target-feature" "+dotprod" "-target-feature" "+fp-armv8" "-target-feature" "+fullfp16" "-target-feature" "+lse" "-target-feature" "+ras" "-target-feature" "+rdm" "-target-feature" "+sha2" "-target-feature" "+neon"
 // CHECK-MCPU-KRYO: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8a" "-target-feature" "+aes" "-target-feature" "+crc" "-target-feature" "+fp-armv8" "-target-feature" "+sha2" "-target-feature" "+neon"
@@ -330,18 +331,18 @@
 // CHECK-MCPU-CARMEL: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8.2a" "-target-feature" "+aes" "-target-feature" "+crc" "-target-feature" "+fp-armv8" "-target-feature" "+fullfp16" "-target-feature" "+lse" "-target-feature" "+ras" "-target-feature" "+rdm" "-target-feature" "+sha2" "-target-feature" "+neon"
 
 // RUN: %clang -target x86_64-apple-macosx -arch arm64 -### -c %s 2>&1 | FileCheck --check-prefix=CHECK-ARCH-ARM64 %s
-// CHECK-ARCH-ARM64: "-target-cpu" "apple-m1" "-target-feature" "+v8.5a" "-target-feature" "+aes" "-target-feature" "+crc" "-target-feature" "+dotprod" "-target-feature" "+fp-armv8" "-target-feature" "+fp16fml" "-target-feature" "+lse" "-target-feature" "+ras" "-target-feature" "+rcpc" "-target-feature" "+rdm" "-target-feature" "+sha2" "-target-feature" "+sha3" "-target-feature" "+neon" "-target-feature" "+zcm" "-target-feature" "+zcz" "-target-feature" "+fullfp16"
+// CHECK-ARCH-ARM64: "-target-cpu" "apple-m1" "-target-feature" "+zcm" "-target-feature" "+zcz" "-target-feature" "+v8.5a" "-target-feature" "+aes" "-target-feature" "+crc" "-target-feature" "+dotprod" "-target-feature" "+complxnum" "-target-feature" "+fp-armv8" "-target-feature" "+fullfp16" "-target-feature" "+fp16fml" "-target-feature" "+jsconv" "-target-feature" "+lse" "-target-feature" "+ras" "-target-feature" "+rcpc" "-target-feature" "+rdm" "-target-feature" "+sha2" "-target-feature" "+sha3" "-target-feature" "+neon"
 
 // RUN: %clang -target x86_64-apple-macosx -arch arm64_32 -### -c %s 2>&1 | FileCheck --check-prefix=CHECK-ARCH-ARM64_32 %s
-// CHECK-ARCH-ARM64_32: "-target-cpu" "apple-s4" "-target-feature" "+v8.3a" "-target-feature" "+aes" "-target-feature" "+crc" "-target-feature" "+fp-armv8" "-target-feature" "+fullfp16" "-target-feature" "+lse" "-target-feature" "+ras" "-target-feature" "+rcpc" "-target-feature" "+rdm" "-target-feature" "+sha2" "-target-feature" "+neon" "-target-feature" "+zcm" "-target-feature" "+zcz"
+// CHECK-ARCH-ARM64_32: "-target-cpu" "apple-s4" "-target-feature" "+zcm" "-target-feature" "+zcz" "-target-feature" "+v8.3a" "-target-feature" "+aes" "-target-feature" "+crc" "-target-feature" "+complxnum" "-target-feature" "+fp-armv8" "-target-feature" "+fullfp16" "-target-feature" "+jsconv" "-target-feature" "+lse" "-target-feature" "+ras" "-target-feature" "+rcpc" "-target-feature" "+rdm" "-target-feature" "+sha2" "-target-feature" "+neon"
 
 // RUN: %clang -target aarch64 -march=armv8-a+fp+simd+crc+crypto -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-MARCH-1 %s
 // RUN: %clang -target aarch64 -march=armv8-a+nofp+nosimd+nocrc+nocrypto+fp+simd+crc+crypto -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-MARCH-1 %s
 // RUN: %clang -target aarch64 -march=armv8-a+nofp+nosimd+nocrc+nocrypto -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-MARCH-2 %s
 // RUN: %clang -target aarch64 -march=armv8-a+fp+simd+crc+crypto+nofp+nosimd+nocrc+nocrypto -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-MARCH-2 %s
 // RUN: %clang -target aarch64 -march=armv8-a+nosimd -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-MARCH-3 %s
-// CHECK-MARCH-1: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+fp-armv8" "-target-feature" "+neon" "-target-feature" "+crc" "-target-feature" "+crypto"
-// CHECK-MARCH-2: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "-fp-armv8" "-target-feature" "-neon" "-target-feature" "-crc" "-target-feature" "-crypto"
+// CHECK-MARCH-1: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8a" "-target-feature" "+aes" "-target-feature" "+crc" "-target-feature" "+crypto" "-target-feature" "+fp-armv8" "-target-feature" "+sha2" "-target-feature" "+neon"
+// CHECK-MARCH-2: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "-fp-armv8"{{.*}} "-target-feature" "-neon"
 // CHECK-MARCH-3: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "-neon"
 
 // While we're checking +nofp, also make sure it stops defining __ARM_FP
@@ -351,42 +352,42 @@
 // Check +sm4:
 //
 // RUN: %clang -target aarch64 -march=armv8.2a+sm4 -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-SM4 %s
-// CHECK-SM4:  "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8.2a" "-target-feature" "+sm4"
+// CHECK-SM4:  "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8.2a"{{.*}} "-target-feature" "+sm4"
 //
 // Check +sha3:
 //
 // RUN: %clang -target aarch64 -march=armv8.2a+sha3 -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-SHA3 %s
-// CHECK-SHA3: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8.2a" "-target-feature" "+sha3"
+// CHECK-SHA3: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8.2a"{{.*}} "-target-feature" "+sha3"
 //
 // Check +sha2:
 //
 // RUN: %clang -target aarch64 -march=armv8.3a+sha2 -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-SHA2 %s
-// CHECK-SHA2: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8.{{.}}a" "-target-feature" "+sha2"
+// CHECK-SHA2: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8.{{.}}a"{{.*}} "-target-feature" "+sha2"
 //
 // Check +aes:
 //
 // RUN: %clang -target aarch64 -march=armv8.3a+aes -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-AES %s
-// CHECK-AES: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8.{{.}}a" "-target-feature" "+aes"
+// CHECK-AES: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8.{{.}}a"{{.*}} "-target-feature" "+aes"
 //
 // Check -sm4:
 //
 // RUN: %clang -target aarch64 -march=armv8.2a+noSM4 -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-NO-SM4 %s
-// CHECK-NO-SM4:  "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8.2a" "-target-feature" "-sm4"
+// CHECK-NO-SM4-NOT: "-target-feature" "+sm4"
 //
 // Check -sha3:
 //
 // RUN: %clang -target aarch64 -march=armv8.2a+noSHA3 -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-NO-SHA3 %s
-// CHECK-NO-SHA3: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8.2a" "-target-feature" "-sha3"
+// CHECK-NO-SHA3-NOT: "-target-feature" "+sha3"
 //
 // Check -sha2:
 //
 // RUN: %clang -target aarch64 -march=armv8.2a+noSHA2 -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-NO-SHA2 %s
-// CHECK-NO-SHA2: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8.2a" "-target-feature" "-sha2"
+// CHECK-NO-SHA2-NOT: "-target-feature" "+sha2"
 //
 // Check -aes:
 //
 // RUN: %clang -target aarch64 -march=armv8.2a+noAES -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-NO-AES %s
-// CHECK-NO-AES: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8.2a" "-target-feature" "-aes"
+// CHECK-NO-AES-NOT: "-target-feature" "+aes"
 //
 //
 // Arch <= ARMv8.3:  crypto = sha2 + aes
@@ -399,19 +400,19 @@
 // RUN: %clang -target aarch64 -march=armv8.2a+crypto -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-CRYPTO83 %s
 // RUN: %clang -target aarch64 -march=armv8.3a+crypto -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-CRYPTO83 %s
 // RUN: %clang -target aarch64 -march=armv8a+crypto+nocrypto+crypto -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-CRYPTO83 %s
-// CHECK-CRYPTO83: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+crypto" "-target-feature" "+sha2" "-target-feature" "+aes"
+// CHECK-CRYPTO83: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+aes"{{.*}} "-target-feature" "+crypto"{{.*}} "-target-feature" "+sha2"
 //
 // Check -crypto:
 //
 // RUN: %clang -target aarch64 -march=armv8a+nocrypto -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-NOCRYPTO8A %s
-// RUN: %clang -target aarch64 -march=armv8.1a+nocrypto -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-NOCRYPTO81 %s
-// RUN: %clang -target aarch64 -march=armv8.2a+nocrypto -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-NOCRYPTO82 %s
-// RUN: %clang -target aarch64 -march=armv8.3a+nocrypto -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-NOCRYPTO82 %s
-// RUN: %clang -target aarch64 -march=armv8.3a+nocrypto+crypto+nocrypto -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-NOCRYPTO82 %s
+// RUN: %clang -target aarch64 -march=armv8.1a+nocrypto -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-NOCRYPTO8A %s
+// RUN: %clang -target aarch64 -march=armv8.2a+nocrypto -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-NOCRYPTO8A %s
+// RUN: %clang -target aarch64 -march=armv8.3a+nocrypto -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-NOCRYPTO8A %s
+// RUN: %clang -target aarch64 -march=armv8.3a+nocrypto+crypto+nocrypto -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-NOCRYPTO8A %s
 
-// CHECK-NOCRYPTO8A: "-target-feature" "+neon" "-target-feature" "+v8a" "-target-feature" "-crypto" "-target-feature" "-sha2" "-target-feature" "-aes" "-target-abi" "aapcs"
-// CHECK-NOCRYPTO81: "-target-feature" "+neon" "-target-feature" "+v8.1a" "-target-feature" "-crypto" "-target-feature" "-sha2" "-target-feature" "-aes" "-target-abi" "aapcs"
-// CHECK-NOCRYPTO82: "-target-feature" "+neon" "-target-feature" "+v8.{{.}}a" "-target-feature" "-crypto" "-target-feature" "-sha2" "-target-feature" "-aes" "-target-feature" "-sm4" "-target-feature" "-sha3" "-target-abi" "aapcs"
+// CHECK-NOCRYPTO8A-NOT: "-target-feature" "+crypto"
+// CHECK-NOCRYPTO8A-NOT: "-target-feature" "+aes"
+// CHECK-NOCRYPTO8A-NOT: "-target-feature" "+sha2"
 //
 // Check +crypto -sha2 -aes:
 //
@@ -421,7 +422,7 @@
 // Check -crypto +sha2 +aes:
 //
 // RUN: %clang -target aarch64 -march=armv8.1a+nocrypto+sha2+aes -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-NOCRYPTO83-SHA2-AES %s
-// CHECK-NOCRYPTO83-SHA2-AES: "-target-feature" "+sha2" "-target-feature" "+aes"
+// CHECK-NOCRYPTO83-SHA2-AES: "-target-feature" "+aes"{{.*}} "-target-feature" "+sha2"
 //
 //
 // Arch >= ARMv8.4:  crypto = sm4 + sha3 + sha2 + aes
@@ -429,14 +430,11 @@
 //
 // Check +crypto:
 //
-// RUN: %clang -target aarch64 -march=armv8.4a+crypto -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-CRYPTO84 %s
-// RUN: %clang -target aarch64 -march=armv8.5a+crypto -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-CRYPTO85 %s
-// RUN: %clang -target aarch64 -march=armv8.6a+crypto -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-CRYPTO86 %s
-// RUN: %clang -target aarch64 -march=armv8.7a+crypto -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-CRYPTO87 %s
-// CHECK-CRYPTO84: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8.4a"{{.*}} "-target-feature" "+crypto" "-target-feature" "+sm4" "-target-feature" "+sha3" "-target-feature" "+sha2" "-target-feature" "+aes"
-// CHECK-CRYPTO85: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8.5a"{{.*}} "-target-feature" "+crypto" "-target-feature" "+sm4" "-target-feature" "+sha3" "-target-feature" "+sha2" "-target-feature" "+aes"
-// CHECK-CRYPTO86: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8.6a"{{.*}} "-target-feature" "+crypto" "-target-feature" "+sm4" "-target-feature" "+sha3" "-target-feature" "+sha2" "-target-feature" "+aes"
-// CHECK-CRYPTO87: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8.7a"{{.*}} "-target-feature" "+crypto" "-target-feature" "+sm4" "-target-feature" "+sha3" "-target-feature" "+sha2" "-target-feature" "+aes"
+// RUN: %clang -target aarch64 -march=armv8.4a+crypto -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-CRYPTO8_4567 %s
+// RUN: %clang -target aarch64 -march=armv8.5a+crypto -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-CRYPTO8_4567 %s
+// RUN: %clang -target aarch64 -march=armv8.6a+crypto -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-CRYPTO8_4567 %s
+// RUN: %clang -target aarch64 -march=armv8.7a+crypto -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-CRYPTO8_4567 %s
+// CHECK-CRYPTO8_4567: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8.{{[4567]}}a"{{.*}} "-target-feature" "+aes"{{.*}} "-target-feature" "+crypto"{{.*}} "-target-feature" "+sha2"{{.*}} "-target-feature" "+sha3"{{.*}} "-target-feature" "+sm4"
 //
 // Check -crypto:
 //
@@ -446,7 +444,7 @@
 // Check +crypto -sm4 -sha3:
 //
 // RUN: %clang -target aarch64 -march=armv8.4a+crypto+sm4+nosm4+sha3+nosha3 -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-CRYPTO84-NOSMSHA %s
-// CHECK-CRYPTO84-NOSMSHA: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8.4a" "-target-feature" "+crypto" "-target-feature" "-sm4" "-target-feature" "-sha3" "-target-feature" "+sha2" "-target-feature" "+aes"
+// CHECK-CRYPTO84-NOSMSHA: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8.4a"{{.*}} "-target-feature" "+aes"{{.*}} "-target-feature" "+crypto"{{.*}} "-target-feature" "+sha2"{{.*}} "-target-feature" "-sha3"{{.*}} "-target-feature" "-sm4"
 //
 
 // RUN: %clang -target aarch64 -mcpu=cyclone+nocrypto -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-MCPU-1 %s
@@ -460,20 +458,22 @@
 // RUN: %clang -target aarch64 -mcpu=generic+Crc -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-MCPU-2 %s
 // RUN: %clang -target aarch64 -mcpu=GENERIC+nocrc+CRC -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-MCPU-2 %s
 // RUN: %clang -target aarch64 -mcpu=cortex-a53+noSIMD -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-MCPU-3 %s
-// CHECK-MCPU-1: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8a" "-target-feature" "+fp-armv8" "-target-feature" "+neon" "-target-feature" "-crypto" "-target-feature" "+zcm" "-target-feature" "+zcz" "-target-feature" "-sha2" "-target-feature" "-aes"
-// CHECK-MCPU-2: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+neon" "-target-feature" "+crc"
-// CHECK-MCPU-3: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8a" "-target-feature" "+crc" "-target-feature" "+fp-armv8" "-target-feature" "-neon" "-target-feature" "-sha2" "-target-feature" "-aes"
+// CHECK-MCPU-1: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "-aes"{{.*}} "-target-feature" "-sha2"
+// CHECK-MCPU-2: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+crc" "-target-feature" "+fp-armv8" "-target-feature" "+neon"
+// CHECK-MCPU-3: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8a" "-target-feature" "-aes" "-target-feature" "+crc" "-target-feature" "+fp-armv8" "-target-feature" "-sha2" "-target-feature" "-neon"
 
 // RUN: %clang -target aarch64 -mcpu=cyclone+nocrc+nocrypto -march=armv8-a -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-MCPU-MARCH %s
 // RUN: %clang -target aarch64 -march=armv8-a -mcpu=cyclone+nocrc+nocrypto  -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-MCPU-MARCH %s
-// CHECK-MCPU-MARCH: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+neon" "-target-feature" "+v8a" "-target-feature" "+zcm" "-target-feature" "+zcz"
+// CHECK-MCPU-MARCH-NOT: "-target-feature" "+aes"
+// CHECK-MCPU-MARCH-NOT: "-target-feature" "+sha2"
+// CHECK-MCPU-MARCH-NOT: "-target-feature" "+crypto"
 
-// RUN: %clang -target aarch64 -mcpu=cortex-a53 -mtune=cyclone -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-MCPU-MTUNE %s
-// RUN: %clang -target aarch64 -mtune=cyclone -mcpu=cortex-a53  -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-MCPU-MTUNE %s
+// RUN: %clang -target aarch64 -mcpu=cortex-a53 -mtune=cyclone -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-MCPU-MTUNE --ignore-case %s
+// RUN: %clang -target aarch64 -mtune=cyclone -mcpu=cortex-a53  -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-MCPU-MTUNE --ignore-case %s
 // ================== Check whether -mtune accepts mixed-case features.
-// RUN: %clang -target aarch64 -mcpu=cortex-a53 -mtune=CYCLONE -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-MCPU-MTUNE %s
-// RUN: %clang -target aarch64 -mtune=CyclonE -mcpu=cortex-a53  -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-MCPU-MTUNE %s
-// CHECK-MCPU-MTUNE: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+aes" "-target-feature" "+crc" "-target-feature" "+fp-armv8" "-target-feature" "+sha2" "-target-feature" "+neon" "-target-feature" "+zcm" "-target-feature" "+zcz"
+// RUN: %clang -target aarch64 -mcpu=cortex-a53 -mtune=CYCLONE -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-MCPU-MTUNE --ignore-case %s
+// RUN: %clang -target aarch64 -mtune=CyclonE -mcpu=cortex-a53  -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-MCPU-MTUNE --ignore-case %s
+// CHECK-MCPU-MTUNE: "-tune-cpu" "cyclone"
 
 // RUN: not %clang -target aarch64 -mcpu=generic+neon -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-ERROR-NEON %s
 // RUN: not %clang -target aarch64 -mcpu=generic+noneon -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-ERROR-NEON %s
@@ -488,9 +488,9 @@
 // RUN: %clang -target aarch64 -march=ARMV8.1A+CRYPTO -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-V81A-FEATURE-1 %s
 // RUN: %clang -target aarch64 -march=Armv8.1a+NOcrypto -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-V81A-FEATURE-2 %s
 // RUN: %clang -target aarch64 -march=armv8.1a+noSIMD -### -c %s 2>&1 | FileCheck -check-prefix=CHECK-V81A-FEATURE-3 %s
-// CHECK-V81A-FEATURE-1: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+neon" "-target-feature" "+v8.1a" "-target-feature" "+crypto"
-// CHECK-V81A-FEATURE-2: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+neon" "-target-feature" "+v8.1a" "-target-feature" "-crypto"
-// CHECK-V81A-FEATURE-3: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8.1a" "-target-feature" "-neon"
+// CHECK-V81A-FEATURE-1: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8.1a" "-target-feature" "+aes" "-target-feature" "+crc" "-target-feature" "+crypto" "-target-feature" "+fp-armv8" "-target-feature" "+lse" "-target-feature" "+rdm" "-target-feature" "+sha2" "-target-feature" "+neon"
+// CHECK-V81A-FEATURE-2: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8.1a" "-target-feature" "+crc" "-target-feature" "+fp-armv8" "-target-feature" "+lse" "-target-feature" "+rdm" "-target-feature" "+neon"
+// CHECK-V81A-FEATURE-3: "-cc1"{{.*}} "-triple" "aarch64{{.*}}" "-target-feature" "+v8.1a" "-target-feature" "+crc" "-target-feature" "+fp-armv8" "-target-feature" "+lse" "-target-feature" "-rdm" "-target-feature" "-neon"
 
 // ================== Check Memory Tagging Extensions (MTE).
 // RUN: %clang -target arm64-none-linux-gnu -march=armv8.5-a+memtag -x c -E -dM %s -o - 2>&1 | FileCheck -check-prefix=CHECK-MEMTAG %s
@@ -588,6 +588,20 @@
 // CHECK-SYS128: __ARM_FEATURE_SYSREG128 1
 // CHECK-NOSYS128-NOT: __ARM_FEATURE_SYSREG128 1
 
+// ================== Check Armv8.9-A/Armv9.4-A Guarded Control Stack (FEAT_GCS)
+// RUN: %clang -target aarch64-arm-none-eabi -march=armv8.9-a     -x c -E -dM %s -o - | FileCheck --check-prefixes=CHECK-NOGCS,CHECK-NOGCS-DEFAULT %s
+// RUN: %clang -target aarch64-arm-none-eabi -march=armv9.4-a     -x c -E -dM %s -o - | FileCheck --check-prefixes=CHECK-NOGCS,CHECK-NOGCS-DEFAULT %s
+// RUN: %clang -target aarch64-arm-none-eabi -march=armv8.9-a+gcs -x c -E -dM %s -o - | FileCheck --check-prefixes=CHECK-GCS,CHECK-NOGCS-DEFAULT %s
+// RUN: %clang -target aarch64-arm-none-eabi -march=armv9.4-a+gcs -x c -E -dM %s -o - | FileCheck --check-prefixes=CHECK-GCS,CHECK-NOGCS-DEFAULT %s
+// RUN: %clang -target aarch64-arm-none-eabi -march=armv8.9-a     -mbranch-protection=gcs -x c -E -dM %s -o - | FileCheck --check-prefixes=CHECK-NOGCS,CHECK-GCS-DEFAULT %s
+// RUN: %clang -target aarch64-arm-none-eabi -march=armv9.4-a     -mbranch-protection=gcs -x c -E -dM %s -o - | FileCheck --check-prefixes=CHECK-NOGCS,CHECK-GCS-DEFAULT %s
+// RUN: %clang -target aarch64-arm-none-eabi -march=armv8.9-a+gcs -mbranch-protection=gcs -x c -E -dM %s -o - | FileCheck --check-prefixes=CHECK-GCS,CHECK-GCS-DEFAULT %s
+// RUN: %clang -target aarch64-arm-none-eabi -march=armv9.4-a+gcs -mbranch-protection=gcs -x c -E -dM %s -o - | FileCheck --check-prefixes=CHECK-GCS,CHECK-GCS-DEFAULT %s
+// CHECK-GCS: __ARM_FEATURE_GCS 1
+// CHECK-NOGCS-NOT: __ARM_FEATURE_GCS 1
+// CHECK-GCS-DEFAULT: __ARM_FEATURE_GCS_DEFAULT 1
+// CHECK-NOGCS-DEFAULT-NOT: __ARM_FEATURE_GCS_DEFAULT 1
+
 // ================== Check default macros for Armv8.1-A and later
 // RUN: %clang -target aarch64-none-elf -march=armv8.1-a -x c -E -dM %s -o - | FileCheck --check-prefixes=CHECK-V81-OR-LATER,CHECK-BEFORE-V83,CHECK-BEFORE-V85     %s
 // RUN: %clang -target aarch64-none-elf -march=armv8.2-a -x c -E -dM %s -o - | FileCheck --check-prefixes=CHECK-V81-OR-LATER,CHECK-BEFORE-V83,CHECK-BEFORE-V85     %s
diff --git a/clang/test/Preprocessor/riscv-target-features.c b/clang/test/Preprocessor/riscv-target-features.c
index 02d8d34116f80..fd0970bd87fd1 100644
--- a/clang/test/Preprocessor/riscv-target-features.c
+++ b/clang/test/Preprocessor/riscv-target-features.c
@@ -3,6 +3,8 @@
 // RUN: %clang --target=riscv64-unknown-linux-gnu -march=rv64i -x c -E -dM %s \
 // RUN: -o - | FileCheck %s
 
+// CHECK-NOT: __riscv_32e {{.*$}}
+// CHECK-NOT: __riscv_64e {{.*$}}
 // CHECK-NOT: __riscv_a {{.*$}}
 // CHECK-NOT: __riscv_atomic
 // CHECK-NOT: __riscv_c {{.*$}}
@@ -170,6 +172,17 @@
 // CHECK-D-EXT: __riscv_flen 64
 // CHECK-D-EXT: __riscv_fsqrt 1
 
+// RUN: %clang --target=riscv32-unknown-linux-gnu \
+// RUN: -march=rv32e -x c -E -dM %s \
+// RUN: -o - | FileCheck --check-prefixes=CHECK-E-EXT,CHECK-RV32E %s
+// RUN: %clang --target=riscv64-unknown-linux-gnu \
+// RUN: -march=rv64e -x c -E -dM %s \
+// RUN: -o - | FileCheck --check-prefixes=CHECK-E-EXT,CHECK-RV64E %s
+// CHECK-RV32E: __riscv_32e 1
+// CHECK-RV64E: __riscv_64e 1
+// CHECK-E-EXT: __riscv_abi_rve 1
+// CHECK-E-EXT: __riscv_e 2000000{{$}}
+
 // RUN: %clang --target=riscv32-unknown-linux-gnu \
 // RUN: -march=rv32if -x c -E -dM %s \
 // RUN: -o - | FileCheck --check-prefix=CHECK-F-EXT %s
@@ -211,6 +224,15 @@
 // CHECK-DOUBLE-NOT: __riscv_float_abi_soft
 // CHECK-DOUBLE-NOT: __riscv_float_abi_single
 
+// RUN: %clang --target=riscv32-unknown-linux-gnu \
+// RUN: -march=rv32i -mabi=ilp32e -x c -E -dM %s \
+// RUN: -o - | FileCheck --check-prefix=CHECK-ILP32E %s
+// RUN: %clang --target=riscv64-unknown-linux-gnu \
+// RUN: -march=rv64i -mabi=lp64e -x c -E -dM %s \
+// RUN: -o - | FileCheck --check-prefix=CHECK-LP64E %s
+// CHECK-ILP32E: __riscv_abi_rve 1
+// CHECK-LP64E: __riscv_abi_rve 1
+
 // RUN: %clang --target=riscv32-unknown-linux-gnu \
 // RUN: -march=rv32ih -x c -E -dM %s \
 // RUN: -o - | FileCheck --check-prefix=CHECK-H-EXT %s
@@ -1042,12 +1064,12 @@
 // CHECK-ZFA-EXT: __riscv_zfa 1000000{{$}}
 
 // RUN: %clang --target=riscv32 -menable-experimental-extensions \
-// RUN: -march=rv32izfbfmin0p8 -x c -E -dM %s \
+// RUN: -march=rv32izfbfmin1p0 -x c -E -dM %s \
 // RUN: -o - | FileCheck --check-prefix=CHECK-ZFBFMIN-EXT %s
 // RUN: %clang --target=riscv64 -menable-experimental-extensions \
-// RUN: -march=rv64izfbfmin0p8 -x c -E -dM %s \
+// RUN: -march=rv64izfbfmin1p0 -x c -E -dM %s \
 // RUN: -o - | FileCheck --check-prefix=CHECK-ZFBFMIN-EXT %s
-// CHECK-ZFBFMIN-EXT: __riscv_zfbfmin 8000{{$}}
+// CHECK-ZFBFMIN-EXT: __riscv_zfbfmin 1000000{{$}}
 
 // RUN: %clang --target=riscv32 -menable-experimental-extensions \
 // RUN: -march=rv32i_zicfilp0p4 -x c -E -dM %s \
@@ -1106,20 +1128,20 @@
 // CHECK-ZVBC-EXT: __riscv_zvbc  1000000{{$}}
 
 // RUN: %clang --target=riscv32 -menable-experimental-extensions \
-// RUN: -march=rv32ifzvfbfmin0p8 -x c -E -dM %s \
+// RUN: -march=rv32ifzvfbfmin1p0 -x c -E -dM %s \
 // RUN: -o - | FileCheck --check-prefix=CHECK-ZVFBFMIN-EXT %s
 // RUN: %clang --target=riscv64 -menable-experimental-extensions \
-// RUN: -march=rv64ifzvfbfmin0p8 -x c -E -dM %s \
+// RUN: -march=rv64ifzvfbfmin1p0 -x c -E -dM %s \
 // RUN: -o - | FileCheck --check-prefix=CHECK-ZVFBFMIN-EXT %s
-// CHECK-ZVFBFMIN-EXT: __riscv_zvfbfmin 8000{{$}}
+// CHECK-ZVFBFMIN-EXT: __riscv_zvfbfmin 1000000{{$}}
 
 // RUN: %clang --target=riscv32 -menable-experimental-extensions \
-// RUN: -march=rv32ifzvfbfwma0p8 -x c -E -dM %s \
+// RUN: -march=rv32ifzvfbfwma1p0 -x c -E -dM %s \
 // RUN: -o - | FileCheck --check-prefix=CHECK-ZVFBFWMA-EXT %s
 // RUN: %clang --target=riscv64 -menable-experimental-extensions \
-// RUN: -march=rv64ifzvfbfwma0p8 -x c -E -dM %s \
+// RUN: -march=rv64ifzvfbfwma1p0 -x c -E -dM %s \
 // RUN: -o - | FileCheck --check-prefix=CHECK-ZVFBFWMA-EXT %s
-// CHECK-ZVFBFWMA-EXT: __riscv_zvfbfwma 8000{{$}}
+// CHECK-ZVFBFWMA-EXT: __riscv_zvfbfwma 1000000{{$}}
 
 // RUN: %clang --target=riscv32 -menable-experimental-extensions \
 // RUN: -march=rv32i_zve32x_zvkg1p0 -x c -E -dM %s \
diff --git a/clang/test/Sema/aarch64-incompat-sm-builtin-calls.c b/clang/test/Sema/aarch64-incompat-sm-builtin-calls.c
index e63d9f0a84757..476da8534ce76 100644
--- a/clang/test/Sema/aarch64-incompat-sm-builtin-calls.c
+++ b/clang/test/Sema/aarch64-incompat-sm-builtin-calls.c
@@ -23,7 +23,7 @@ int16x8_t incompat_neon_smc(int16x8_t splat) __arm_streaming_compatible {
   return (int16x8_t)__builtin_neon_vqaddq_v((int8x16_t)splat, (int8x16_t)splat, 33);
 }
 
-void incompat_sme_smc(svbool_t pg, void const *ptr) __arm_streaming_compatible __arm_shared_za {
+void incompat_sme_smc(svbool_t pg, void const *ptr) __arm_streaming_compatible __arm_inout("za") {
   // expected-warning@+1 {{builtin call has undefined behaviour when called from a streaming compatible function}}
   return __builtin_sme_svld1_hor_za128(0, 0, pg, ptr);
 }
@@ -58,7 +58,7 @@ svuint32_t incompat_sve2_smc(svbool_t pg, svuint32_t a, int64_t b) __arm_streami
   return __builtin_sve_svldnt1_gather_u32base_index_u32(pg, a, b);
 }
 
-void incompat_sme_sm(svbool_t pn, svbool_t pm, svfloat32_t zn, svfloat32_t zm) __arm_shared_za {
+void incompat_sme_sm(svbool_t pn, svbool_t pm, svfloat32_t zn, svfloat32_t zm) __arm_inout("za") {
   // expected-warning@+1 {{builtin call has undefined behaviour when called from a non-streaming function}}
   svmops_za32_f32_m(0, pn, pm, zn, zm);
 }
diff --git a/clang/test/Sema/aarch64-sme-func-attrs-without-target-feature.cpp b/clang/test/Sema/aarch64-sme-func-attrs-without-target-feature.cpp
index b59d67f7f57b8..0a54a94f408b7 100644
--- a/clang/test/Sema/aarch64-sme-func-attrs-without-target-feature.cpp
+++ b/clang/test/Sema/aarch64-sme-func-attrs-without-target-feature.cpp
@@ -4,16 +4,16 @@
 
 void streaming_compatible_def() __arm_streaming_compatible {} // OK
 void streaming_def() __arm_streaming { } // expected-error {{function executed in streaming-SVE mode requires 'sme'}}
-void shared_za_def() __arm_shared_za { } // expected-error {{function using ZA state requires 'sme'}}
-__arm_new_za void new_za_def() { } // expected-error {{function using ZA state requires 'sme'}}
+void shared_za_def() __arm_inout("za") { } // expected-error {{function using ZA state requires 'sme'}}
+__arm_new("za") void new_za_def() { } // expected-error {{function using ZA state requires 'sme'}}
 __arm_locally_streaming void locally_streaming_def() { } // expected-error {{function executed in streaming-SVE mode requires 'sme'}}
-void streaming_shared_za_def() __arm_streaming __arm_shared_za { } // expected-error {{function executed in streaming-SVE mode requires 'sme'}}
+void streaming_shared_za_def() __arm_streaming __arm_inout("za") { } // expected-error {{function executed in streaming-SVE mode requires 'sme'}}
 
 // It should work fine when we explicitly add the target("sme") attribute.
 __attribute__((target("sme"))) void streaming_compatible_def_sme_attr() __arm_streaming_compatible {} // OK
 __attribute__((target("sme"))) void streaming_def_sme_attr() __arm_streaming { } // OK
-__attribute__((target("sme"))) void shared_za_def_sme_attr() __arm_shared_za { } // OK
-__arm_new_za __attribute__((target("sme"))) void new_za_def_sme_attr() {} // OK
+__attribute__((target("sme"))) void shared_za_def_sme_attr() __arm_inout("za") { } // OK
+__arm_new("za") __attribute__((target("sme"))) void new_za_def_sme_attr() {} // OK
 __arm_locally_streaming __attribute__((target("sme"))) void locally_streaming_def_sme_attr() {} // OK
 
 // Test that it also works with the target("sme2") attribute.
@@ -22,7 +22,7 @@ __attribute__((target("sme2"))) void streaming_def_sme2_attr() __arm_streaming {
 // No code is generated for declarations, so it should be fine to declare using the attribute.
 void streaming_compatible_decl() __arm_streaming_compatible; // OK
 void streaming_decl() __arm_streaming; // OK
-void shared_za_decl() __arm_shared_za; // OK
+void shared_za_decl() __arm_inout("za"); // OK
 
 void non_streaming_decl();
 void non_streaming_def(void (*streaming_fn_ptr)(void) __arm_streaming,
diff --git a/clang/test/Sema/aarch64-sme-func-attrs.c b/clang/test/Sema/aarch64-sme-func-attrs.c
index 73c0934d689e7..b986b0b3de2e1 100644
--- a/clang/test/Sema/aarch64-sme-func-attrs.c
+++ b/clang/test/Sema/aarch64-sme-func-attrs.c
@@ -6,27 +6,25 @@
 void sme_arm_streaming(void) __arm_streaming;
 void sme_arm_streaming_compatible(void) __arm_streaming_compatible;
 
-__arm_new_za void sme_arm_new_za(void) {}
-void sme_arm_shared_za(void) __arm_shared_za;
-void sme_arm_preserves_za(void) __arm_preserves_za;
+__arm_new("za") void sme_arm_new_za(void) {}
+void sme_arm_shared_za(void) __arm_inout("za");
+void sme_arm_preserves_za(void) __arm_preserves("za");
 
-__arm_new_za void sme_arm_streaming_new_za(void) __arm_streaming {}
-void sme_arm_streaming_shared_za(void) __arm_streaming __arm_shared_za;
-void sme_arm_streaming_preserves_za(void) __arm_streaming __arm_preserves_za;
+__arm_new("za") void sme_arm_streaming_new_za(void) __arm_streaming {}
+void sme_arm_streaming_shared_za(void) __arm_streaming __arm_inout("za");
+void sme_arm_streaming_preserves_za(void) __arm_streaming __arm_preserves("za");
 
-__arm_new_za void sme_arm_sc_new_za(void) __arm_streaming_compatible {}
-void sme_arm_sc_shared_za(void) __arm_streaming_compatible __arm_shared_za;
-void sme_arm_sc_preserves_za(void) __arm_streaming_compatible __arm_preserves_za;
-
-void sme_arm_shared_preserves_za(void) __arm_shared_za __arm_preserves_za;
+__arm_new("za") void sme_arm_sc_new_za(void) __arm_streaming_compatible {}
+void sme_arm_sc_shared_za(void) __arm_streaming_compatible __arm_inout("za");
+void sme_arm_sc_preserves_za(void) __arm_streaming_compatible __arm_preserves("za");
 
 __arm_locally_streaming void sme_arm_locally_streaming(void) { }
 __arm_locally_streaming void sme_arm_streaming_and_locally_streaming(void) __arm_streaming { }
 __arm_locally_streaming void sme_arm_streaming_and_streaming_compatible(void) __arm_streaming_compatible { }
 
-__arm_locally_streaming __arm_new_za void sme_arm_ls_new_za(void) { }
-__arm_locally_streaming void sme_arm_ls_shared_za(void) __arm_shared_za { }
-__arm_locally_streaming void sme_arm_ls_preserves_za(void) __arm_preserves_za { }
+__arm_locally_streaming __arm_new("za") void sme_arm_ls_new_za(void) { }
+__arm_locally_streaming void sme_arm_ls_shared_za(void) __arm_inout("za") { }
+__arm_locally_streaming void sme_arm_ls_preserves_za(void) __arm_preserves("za") { }
 
 // Valid attributes on function pointers
 
@@ -38,18 +36,14 @@ void streaming_compatible_ptr(void) __arm_streaming_compatible;
 typedef void (*fptrty2) (void) __arm_streaming_compatible;
 fptrty2 call_sc_func() { return streaming_compatible_ptr; }
 
-void shared_za_ptr(void) __arm_shared_za;
-typedef void (*fptrty3) (void) __arm_shared_za;
+void shared_za_ptr(void) __arm_inout("za");
+typedef void (*fptrty3) (void) __arm_inout("za");
 fptrty3 call_shared_za_func() { return shared_za_ptr; }
 
-void preserves_za_ptr(void) __arm_preserves_za;
-typedef void (*fptrty4) (void) __arm_preserves_za;
+void preserves_za_ptr(void) __arm_preserves("za");
+typedef void (*fptrty4) (void) __arm_preserves("za");
 fptrty4 call_preserve_za_func() { return preserves_za_ptr; }
 
-void shared_preserves_za_ptr(void) __arm_shared_za __arm_preserves_za;
-typedef void (*fptrty5) (void) __arm_shared_za __arm_preserves_za;
-fptrty5 call_shared_preserve_za_func() { return shared_preserves_za_ptr; }
-
 typedef void (*fptrty6) (void);
 fptrty6 cast_nza_func_to_normal() { return sme_arm_new_za; }
 fptrty6 cast_ls_func_to_normal() { return sme_arm_locally_streaming; }
@@ -68,13 +62,13 @@ void streaming_mode(void) __arm_streaming __arm_streaming_compatible;
 // expected-note@+1 {{conflicting attribute is here}}
 void streaming_compatible(void) __arm_streaming_compatible __arm_streaming;
 
-// expected-cpp-error@+2 {{'__arm_new_za' and '__arm_shared_za' are not compatible}}
-// expected-error@+1 {{'__arm_new_za' and '__arm_shared_za' are not compatible}}
-__arm_new_za void new_shared_za(void) __arm_shared_za {}
+// expected-cpp-error@+2 {{'__arm_new("za")' and '__arm_inout("za")' are not compatible}}
+// expected-error@+1 {{'__arm_new("za")' and '__arm_inout("za")' are not compatible}}
+__arm_new("za") void new_shared_za(void) __arm_inout("za") {}
 
-// expected-cpp-error@+2 {{'__arm_new_za' and '__arm_preserves_za' are not compatible}}
-// expected-error@+1 {{'__arm_new_za' and '__arm_preserves_za' are not compatible}}
-__arm_new_za void new_preserves_za(void) __arm_preserves_za {}
+// expected-cpp-error@+2 {{'__arm_new("za")' and '__arm_preserves("za")' are not compatible}}
+// expected-error@+1 {{'__arm_new("za")' and '__arm_preserves("za")' are not compatible}}
+__arm_new("za") void new_preserves_za(void) __arm_preserves("za") {}
 
 // Invalid attributes on function pointers
 
@@ -125,24 +119,25 @@ sc_ptrty return_invalid_fptr_streaming_compatible_normal(n_ptrty f) { return f;
 // expected-error@+1 {{incompatible function pointer types returning 'sc_ptrty' (aka 'void (*)(void) __arm_streaming_compatible') from a function with result type 'n_ptrty' (aka 'void (*)(void)')}}
 n_ptrty return_invalid_fptr_normal_streaming_compatible(sc_ptrty f) { return f; }
 
-typedef void (*sz_ptrty) (void) __arm_shared_za;
+typedef void (*sz_ptrty) (void) __arm_inout("za");
 sz_ptrty return_valid_shared_za_fptr(sz_ptrty f) { return f; }
 
 
-// expected-cpp-error@+2 {{cannot initialize return object of type 'sz_ptrty' (aka 'void (*)() __arm_shared_za') with an lvalue of type 'n_ptrty' (aka 'void (*)()')}}
-// expected-error@+1 {{incompatible function pointer types returning 'n_ptrty' (aka 'void (*)(void)') from a function with result type 'sz_ptrty' (aka 'void (*)(void) __arm_shared_za')}}
+// expected-cpp-error@+2 {{cannot initialize return object of type 'sz_ptrty' (aka 'void (*)() __arm_inout("za")') with an lvalue of type 'n_ptrty' (aka 'void (*)()')}}
+// expected-error@+1 {{incompatible function pointer types returning 'n_ptrty' (aka 'void (*)(void)') from a function with result type 'sz_ptrty' (aka 'void (*)(void) __arm_inout("za")')}}
 sz_ptrty return_invalid_fptr_shared_za_normal(n_ptrty f) { return f; }
-// expected-cpp-error@+2 {{cannot initialize return object of type 'n_ptrty' (aka 'void (*)()') with an lvalue of type 'sz_ptrty' (aka 'void (*)() __arm_shared_za')}}
-// expected-error@+1 {{incompatible function pointer types returning 'sz_ptrty' (aka 'void (*)(void) __arm_shared_za') from a function with result type 'n_ptrty' (aka 'void (*)(void)')}}
+// expected-cpp-error@+2 {{cannot initialize return object of type 'n_ptrty' (aka 'void (*)()') with an lvalue of type 'sz_ptrty' (aka 'void (*)() __arm_inout("za")')}}
+// expected-error@+1 {{incompatible function pointer types returning 'sz_ptrty' (aka 'void (*)(void) __arm_inout("za")') from a function with result type 'n_ptrty' (aka 'void (*)(void)')}}
 n_ptrty return_invalid_fptr_normal_shared_za(sz_ptrty f) { return f; }
 
-typedef void (*pz_ptrty) (void) __arm_preserves_za;
+typedef void (*pz_ptrty) (void) __arm_preserves("za");
 pz_ptrty return_valid_preserves_za_fptr(pz_ptrty f) { return f; }
 
-// expected-cpp-error@+2 {{cannot initialize return object of type 'pz_ptrty' (aka 'void (*)() __arm_preserves_za') with an lvalue of type 'n_ptrty' (aka 'void (*)()')}}
-// expected-error@+1 {{incompatible function pointer types returning 'n_ptrty' (aka 'void (*)(void)') from a function with result type 'pz_ptrty' (aka 'void (*)(void) __arm_preserves_za')}}
+// expected-cpp-error@+2 {{cannot initialize return object of type 'pz_ptrty' (aka 'void (*)() __arm_preserves("za")') with an lvalue of type 'n_ptrty' (aka 'void (*)()')}}
+// expected-error@+1 {{incompatible function pointer types returning 'n_ptrty' (aka 'void (*)(void)') from a function with result type 'pz_ptrty' (aka 'void (*)(void) __arm_preserves("za")')}}
 pz_ptrty return_invalid_fptr_preserves_za_normal(n_ptrty f) { return f; }
-// No diagnostics, the preserves_za hint should be dropped silently.
+// expected-cpp-error@+2 {{cannot initialize return object of type 'n_ptrty' (aka 'void (*)()') with an lvalue of type 'pz_ptrty' (aka 'void (*)() __arm_preserves("za")')}}
+// expected-error@+1 {{incompatible function pointer types returning 'pz_ptrty' (aka 'void (*)(void) __arm_preserves("za")') from a function with result type 'n_ptrty' (aka 'void (*)(void)')}}
 n_ptrty return_invalid_fptr_normal_preserves_za(pz_ptrty f) { return f; }
 
 // Test template instantiations
@@ -164,21 +159,21 @@ template short templated(short);
 void redecl(void) __arm_streaming;
 void redecl(void) __arm_streaming_compatible { }
 
-// expected-error@+5 {{function declared 'void (void) __arm_shared_za' was previously declared 'void (void) __arm_shared_za __arm_preserves_za', which has different SME function attributes}}
+// expected-error@+5 {{function declared 'void (void)' was previously declared 'void (void) __arm_preserves("za")', which has different SME function attributes}}
 // expected-note@+3 {{previous declaration is here}}
-// expected-cpp-error@+3 {{function declared 'void () __arm_shared_za' was previously declared 'void () __arm_shared_za __arm_preserves_za', which has different SME function attributes}}
+// expected-cpp-error@+3 {{function declared 'void ()' was previously declared 'void () __arm_preserves("za")', which has different SME function attributes}}
 // expected-cpp-note@+1 {{previous declaration is here}}
-void redecl_preserve_za(void) __arm_shared_za __arm_preserves_za;;
-void redecl_preserve_za(void) __arm_shared_za {}
+void redecl_preserve_za(void) __arm_preserves("za");;
+void redecl_preserve_za(void) {}
 
-// expected-error@+5 {{function declared 'void (void) __arm_shared_za __arm_preserves_za' was previously declared 'void (void) __arm_shared_za', which has different SME function attributes}}
+// expected-error@+5 {{function declared 'void (void) __arm_preserves("za")' was previously declared 'void (void)', which has different SME function attributes}}
 // expected-note@+3 {{previous declaration is here}}
-// expected-cpp-error@+3 {{function declared 'void () __arm_shared_za __arm_preserves_za' was previously declared 'void () __arm_shared_za', which has different SME function attributes}}
+// expected-cpp-error@+3 {{function declared 'void () __arm_preserves("za")' was previously declared 'void ()', which has different SME function attributes}}
 // expected-cpp-note@+1 {{previous declaration is here}}
-void redecl_nopreserve_za(void) __arm_shared_za;
-void redecl_nopreserve_za(void) __arm_shared_za __arm_preserves_za {}
+void redecl_nopreserve_za(void);
+void redecl_nopreserve_za(void) __arm_preserves("za") {}
 
-void non_za_definition(void (*shared_za_fn_ptr)(void) __arm_shared_za) {
+void non_za_definition(void (*shared_za_fn_ptr)(void) __arm_inout("za"), void (*preserves_za_fn_ptr)(void) __arm_preserves("za")) {
   sme_arm_new_za(); // OK
   // expected-error@+2 {{call to a shared ZA function requires the caller to have ZA state}}
   // expected-cpp-error@+1 {{call to a shared ZA function requires the caller to have ZA state}}
@@ -186,43 +181,46 @@ void non_za_definition(void (*shared_za_fn_ptr)(void) __arm_shared_za) {
   // expected-error@+2 {{call to a shared ZA function requires the caller to have ZA state}}
   // expected-cpp-error@+1 {{call to a shared ZA function requires the caller to have ZA state}}
   shared_za_fn_ptr();
+  // expected-error@+2 {{call to a shared ZA function requires the caller to have ZA state}}
+  // expected-cpp-error@+1 {{call to a shared ZA function requires the caller to have ZA state}}
+  preserves_za_fn_ptr();
 }
 
-void shared_za_definition(void (*shared_za_fn_ptr)(void) __arm_shared_za) __arm_shared_za {
+void shared_za_definition(void (*shared_za_fn_ptr)(void) __arm_inout("za")) __arm_inout("za") {
   sme_arm_shared_za(); // OK
   shared_za_fn_ptr(); // OK
 }
 
-__arm_new_za void new_za_definition(void (*shared_za_fn_ptr)(void) __arm_shared_za) {
+__arm_new("za") void new_za_definition(void (*shared_za_fn_ptr)(void) __arm_inout("za")) {
   sme_arm_shared_za(); // OK
   shared_za_fn_ptr(); // OK
 }
 
 #ifdef __cplusplus
-int shared_za_initializer(void) __arm_shared_za;
+int shared_za_initializer(void) __arm_inout("za");
 // expected-cpp-error@+1 {{call to a shared ZA function requires the caller to have ZA state}}
 int global = shared_za_initializer();
 
 struct S {
-  virtual void shared_za_memberfn(void) __arm_shared_za;
+  virtual void shared_za_memberfn(void) __arm_inout("za");
 };
 
 struct S2 : public S {
-// expected-cpp-error@+2 {{virtual function 'shared_za_memberfn' has different attributes ('void ()') than the function it overrides (which has 'void () __arm_shared_za')}}
+// expected-cpp-error@+2 {{virtual function 'shared_za_memberfn' has different attributes ('void ()') than the function it overrides (which has 'void () __arm_inout("za")')}}
 // expected-cpp-note@-5 {{overridden virtual function is here}}
-  __arm_new_za void shared_za_memberfn(void) override {}
+  __arm_new("za") void shared_za_memberfn(void) override {}
 };
 
-// The '__arm_preserves_za' property cannot be dropped when overriding a virtual
-// function. It is however fine for the overriding function to be '__arm_preserves_za'
+// The '__arm_preserves("za")' property cannot be dropped when overriding a virtual
+// function. It is however fine for the overriding function to be '__arm_preserves("za")'
 // even though the function that it overrides is not.
 
 struct S_PreservesZA {
-  virtual void memberfn(void) __arm_preserves_za;
+  virtual void memberfn(void) __arm_preserves("za");
 };
 
 struct S_Drop_PreservesZA : S_PreservesZA {
-// expected-cpp-error@+2 {{virtual function 'memberfn' has different attributes ('void ()') than the function it overrides (which has 'void () __arm_preserves_za')}}
+// expected-cpp-error@+2 {{virtual function 'memberfn' has different attributes ('void ()') than the function it overrides (which has 'void () __arm_preserves("za")')}}
 // expected-cpp-note@-5 {{overridden virtual function is here}}
   void memberfn(void) override {}
 };
@@ -230,9 +228,11 @@ struct S_Drop_PreservesZA : S_PreservesZA {
 struct S_NoPreservesZA {
   virtual void memberfn(void);
 };
+
 struct S_AddPreservesZA : S_NoPreservesZA {
-// This is fine, the overridden function just adds more guarantees.
-  void memberfn(void) __arm_preserves_za override {}
+// expected-cpp-error@+2 {{virtual function 'memberfn' has different attributes ('void () __arm_preserves("za")') than the function it overrides (which has 'void ()')}}
+// expected-cpp-note@-5 {{overridden virtual function is here}}
+  void memberfn(void) __arm_preserves("za") override {}
 };
 
 
@@ -258,20 +258,20 @@ struct S3 {
 };
 
 template <>
-struct S3 {
+struct S3 {
   static constexpr int value = 8;
 };
 
 template <>
-struct S3 {
+struct S3 {
   static constexpr int value = 16;
 };
 
 void normal_func(void) {}
 void streaming_func(void) __arm_streaming {}
 void streaming_compatible_func(void) __arm_streaming_compatible {}
-void shared_za_func(void) __arm_shared_za {}
-void preserves_za_func(void) __arm_preserves_za {}
+void shared_za_func(void) __arm_inout("za") {}
+void preserves_za_func(void) __arm_preserves("za") {}
 
 static_assert(S3::value == 1, "why are we picking the wrong specialization?");
 static_assert(S3::value == 2, "why are we picking the wrong specialization?");
@@ -295,8 +295,8 @@ template int test_templated_f(T);
 template<> constexpr int test_templated_f(void(*)(void)) { return 1; }
 template<> constexpr int test_templated_f(void(*)(void)__arm_streaming) { return 2; }
 template<> constexpr int test_templated_f(void(*)(void)__arm_streaming_compatible) { return 4; }
-template<> constexpr int test_templated_f(void(*)(void)__arm_shared_za) { return 8; }
-template<> constexpr int test_templated_f(void(*)(void)__arm_preserves_za) { return 16; }
+template<> constexpr int test_templated_f(void(*)(void)__arm_inout("za")) { return 8; }
+template<> constexpr int test_templated_f(void(*)(void)__arm_preserves("za")) { return 16; }
 
 static_assert(test_templated_f(&normal_func) == 1, "Instantiated to wrong function");
 static_assert(test_templated_f(&streaming_func) == 2, "Instantiated to wrong function");
@@ -312,8 +312,8 @@ int invalid_type_for_attribute __arm_streaming;
 constexpr int overload(void f(void)) { return 1; }
 constexpr int overload(void f(void) __arm_streaming) { return 2; }
 constexpr int overload(void f(void) __arm_streaming_compatible) { return 4; }
-constexpr int overload(void f(void) __arm_shared_za) { return 8; }
-constexpr int overload(void f(void) __arm_preserves_za) { return 16; }
+constexpr int overload(void f(void) __arm_inout("za")) { return 8; }
+constexpr int overload(void f(void) __arm_preserves("za")) { return 16; }
 static_assert(overload(&normal_func) == 1, "Overloaded to wrong function");
 static_assert(overload(&streaming_func) == 2, "Overloaded to wrong function");
 static_assert(overload(&streaming_compatible_func) == 4, "Overloaded to wrong function");
@@ -330,3 +330,73 @@ constexpr X *ptr = 0;
 static_assert(overload_int(ptr->foo) == 2, "Overloaded to the wrong function after implicit instantiation");
 
 #endif // ifdef __cplusplus
+
+// expected-cpp-error@+2 {{unknown state ''}}
+// expected-error@+1 {{unknown state ''}}
+__arm_new("") void invalid_arm_new_empty_string(void);
+// expected-cpp-error@+2 {{expected string literal as argument of '__arm_new' attribute}}
+// expected-error@+1 {{expected string literal as argument of '__arm_new' attribute}}
+__arm_new(0) void invalid_arm_new_non_literal_string(void);
+// expected-cpp-error@+2 {{unknown state 'unknownstate'}}
+// expected-error@+1 {{unknown state 'unknownstate'}}
+__arm_new("unknownstate") void invalid_arm_new_unknown_state(void);
+
+// expected-cpp-error@+2 {{unknown state ''}}
+// expected-error@+1 {{unknown state ''}}
+void invalid_arm_in_empty_string(void) __arm_in("");
+// expected-cpp-error@+2 {{expected string literal as argument of '__arm_in' attribute}}
+// expected-error@+1 {{expected string literal as argument of '__arm_in' attribute}}
+void invalid_arm_in_non_literal_string(void) __arm_in(0);
+// expected-cpp-error@+2 {{unknown state 'unknownstate'}}
+// expected-error@+1 {{unknown state 'unknownstate'}}
+void invalid_arm_in_unknown_state(void) __arm_in("unknownstate");
+
+void valid_state_attrs_in_in1(void) __arm_in("za");
+void valid_state_attrs_in_in2(void) __arm_in("za", "za");
+
+// expected-cpp-error@+2 {{missing state for '__arm_in'}}
+// expected-error@+1 {{missing state for '__arm_in'}}
+void invalid_state_attrs_no_arg1(void) __arm_in();
+// expected-cpp-error@+2 {{missing state for '__arm_new'}}
+// expected-error@+1 {{missing state for '__arm_new'}}
+__arm_new() void invalid_state_attrs_no_arg2(void);
+
+// expected-cpp-error@+2 {{conflicting attributes for state 'za'}}
+// expected-error@+1 {{conflicting attributes for state 'za'}}
+void conflicting_state_attrs_in_out(void) __arm_in("za") __arm_out("za");
+// expected-cpp-error@+2 {{conflicting attributes for state 'za'}}
+// expected-error@+1 {{conflicting attributes for state 'za'}}
+void conflicting_state_attrs_in_inout(void) __arm_in("za") __arm_inout("za");
+// expected-cpp-error@+2 {{conflicting attributes for state 'za'}}
+// expected-error@+1 {{conflicting attributes for state 'za'}}
+void conflicting_state_attrs_in_preserves(void) __arm_in("za") __arm_preserves("za");
+
+// expected-cpp-error@+2 {{conflicting attributes for state 'za'}}
+// expected-error@+1 {{conflicting attributes for state 'za'}}
+void conflicting_state_attrs_out_in(void) __arm_out("za") __arm_in("za");
+// expected-cpp-error@+2 {{conflicting attributes for state 'za'}}
+// expected-error@+1 {{conflicting attributes for state 'za'}}
+void conflicting_state_attrs_out_inout(void) __arm_out("za") __arm_inout("za");
+// expected-cpp-error@+2 {{conflicting attributes for state 'za'}}
+// expected-error@+1 {{conflicting attributes for state 'za'}}
+void conflicting_state_attrs_out_preserves(void) __arm_out("za") __arm_preserves("za");
+
+// expected-cpp-error@+2 {{conflicting attributes for state 'za'}}
+// expected-error@+1 {{conflicting attributes for state 'za'}}
+void conflicting_state_attrs_inout_in(void) __arm_inout("za") __arm_in("za");
+// expected-cpp-error@+2 {{conflicting attributes for state 'za'}}
+// expected-error@+1 {{conflicting attributes for state 'za'}}
+void conflicting_state_attrs_inout_out(void) __arm_inout("za") __arm_out("za");
+// expected-cpp-error@+2 {{conflicting attributes for state 'za'}}
+// expected-error@+1 {{conflicting attributes for state 'za'}}
+void conflicting_state_attrs_inout_preserves(void) __arm_inout("za") __arm_preserves("za");
+
+// expected-cpp-error@+2 {{conflicting attributes for state 'za'}}
+// expected-error@+1 {{conflicting attributes for state 'za'}}
+void conflicting_state_attrs_preserves_in(void) __arm_preserves("za") __arm_in("za");
+// expected-cpp-error@+2 {{conflicting attributes for state 'za'}}
+// expected-error@+1 {{conflicting attributes for state 'za'}}
+void conflicting_state_attrs_preserves_out(void) __arm_preserves("za") __arm_out("za");
+// expected-cpp-error@+2 {{conflicting attributes for state 'za'}}
+// expected-error@+1 {{conflicting attributes for state 'za'}}
+void conflicting_state_attrs_preserves_inout(void) __arm_preserves("za") __arm_inout("za");
diff --git a/clang/test/Sema/aarch64-sme-intrinsics/acle_sme_imm.cpp b/clang/test/Sema/aarch64-sme-intrinsics/acle_sme_imm.cpp
index 529d0d2d1e625..40254a5a0eafa 100644
--- a/clang/test/Sema/aarch64-sme-intrinsics/acle_sme_imm.cpp
+++ b/clang/test/Sema/aarch64-sme-intrinsics/acle_sme_imm.cpp
@@ -12,7 +12,7 @@
 
 #include 
 
-void test_range_0_0(uint32_t slice, svbool_t pg, void *ptr) __arm_streaming __arm_shared_za {
+void test_range_0_0(uint32_t slice, svbool_t pg, void *ptr) __arm_streaming __arm_inout("za") {
   // expected-error@+1 {{argument value 18446744073709551615 is outside the valid range [0, 0]}}
   SVE_ACLE_FUNC(svld1_hor_za8,,,)(-1, slice, pg, ptr);
   // expected-error@+1 {{argument value 1 is outside the valid range [0, 0]}}
@@ -32,7 +32,7 @@ void test_range_0_0(uint32_t slice, svbool_t pg, void *ptr) __arm_streaming __ar
   SVE_ACLE_FUNC(svwrite_ver_za8, _s8, _m,)(1, slice, pg, svundef_s8());
 }
 
-void test_range_0_1(uint32_t slice, svbool_t pg, void *ptr) __arm_streaming __arm_shared_za {
+void test_range_0_1(uint32_t slice, svbool_t pg, void *ptr) __arm_streaming __arm_inout("za") {
   // expected-error@+1 {{argument value 18446744073709551615 is outside the valid range [0, 1]}}
   SVE_ACLE_FUNC(svld1_hor_za16,,,)(-1, slice, pg, ptr);
   // expected-error@+1 {{argument value 2 is outside the valid range [0, 1]}}
@@ -52,7 +52,7 @@ void test_range_0_1(uint32_t slice, svbool_t pg, void *ptr) __arm_streaming __ar
   SVE_ACLE_FUNC(svwrite_ver_za16, _s16, _m,)(2, slice, pg, svundef_s16());
 }
 
-void test_range_0_3(uint32_t slice, svbool_t pg, void *ptr) __arm_streaming __arm_shared_za {
+void test_range_0_3(uint32_t slice, svbool_t pg, void *ptr) __arm_streaming __arm_inout("za") {
   // expected-error@+1 {{argument value 18446744073709551615 is outside the valid range [0, 3]}}
   SVE_ACLE_FUNC(svld1_hor_za32,,,)(-1, slice, pg, ptr);
   // expected-error@+1 {{argument value 4 is outside the valid range [0, 3]}}
@@ -90,7 +90,7 @@ void test_range_0_3(uint32_t slice, svbool_t pg, void *ptr) __arm_streaming __ar
   SVE_ACLE_FUNC(svusmops_za32, _u8, _m,)(-1, pg, pg, svundef_u8(), svundef_s8());
 }
 
-void test_range_0_7(uint32_t slice, svbool_t pg, void *ptr) __arm_streaming __arm_shared_za {
+void test_range_0_7(uint32_t slice, svbool_t pg, void *ptr) __arm_streaming __arm_inout("za") {
   // expected-error@+1 {{argument value 18446744073709551615 is outside the valid range [0, 7]}}
   SVE_ACLE_FUNC(svld1_hor_za64,,,)(-1, slice, pg, ptr);
   // expected-error@+1 {{argument value 8 is outside the valid range [0, 7]}}
@@ -133,7 +133,7 @@ void test_range_0_7(uint32_t slice, svbool_t pg, void *ptr) __arm_streaming __ar
   SVE_ACLE_FUNC(svmops_za64, _f64, _m,)(-1, pg, pg, svundef_f64(), svundef_f64());
 }
 
-void test_range_0_15(uint32_t slice, svbool_t pg, void *ptr) __arm_streaming __arm_shared_za {
+void test_range_0_15(uint32_t slice, svbool_t pg, void *ptr) __arm_streaming __arm_inout("za") {
   // expected-error@+1 {{argument value 18446744073709551615 is outside the valid range [0, 15]}}
   SVE_ACLE_FUNC(svld1_hor_za128,,,)(-1, slice, pg, ptr);
   // expected-error@+1 {{argument value 16 is outside the valid range [0, 15]}}
@@ -153,14 +153,14 @@ void test_range_0_15(uint32_t slice, svbool_t pg, void *ptr) __arm_streaming __a
   SVE_ACLE_FUNC(svwrite_ver_za128, _s8, _m,)(16, slice, pg, svundef_s8());
 }
 
-void test_range_0_255(svbool_t pg, void *ptr) __arm_streaming __arm_shared_za {
+void test_range_0_255(svbool_t pg, void *ptr) __arm_streaming __arm_inout("za") {
   // expected-error@+1 {{argument value 256 is outside the valid range [0, 255]}}
   SVE_ACLE_FUNC(svzero_mask_za,,,)(256);
   // expected-error@+1 {{argument value 18446744073709551615 is outside the valid range [0, 255]}}
   SVE_ACLE_FUNC(svzero_mask_za,,,)(-1);
 }
 
-void test_constant(uint64_t u64, svbool_t pg, void *ptr) __arm_streaming __arm_shared_za {
+void test_constant(uint64_t u64, svbool_t pg, void *ptr) __arm_streaming __arm_inout("za") {
   SVE_ACLE_FUNC(svld1_hor_za8,,,)(u64, u64, pg, ptr);  // expected-error {{argument to 'svld1_hor_za8' must be a constant integer}}
   SVE_ACLE_FUNC(svst1_hor_za32,,,)(u64, 0, pg, ptr); // expected-error {{argument to 'svst1_hor_za32' must be a constant integer}}
   SVE_ACLE_FUNC(svld1_hor_vnum_za8,,,)(u64, 0, pg, ptr, u64);  // expected-error {{argument to 'svld1_hor_vnum_za8' must be a constant integer}}
diff --git a/clang/test/Sema/aarch64-sme-intrinsics/acle_sme_target.c b/clang/test/Sema/aarch64-sme-intrinsics/acle_sme_target.c
index 95bb6be2d2d34..f1e858f819602 100644
--- a/clang/test/Sema/aarch64-sme-intrinsics/acle_sme_target.c
+++ b/clang/test/Sema/aarch64-sme-intrinsics/acle_sme_target.c
@@ -6,21 +6,21 @@
 #include 
 
 __attribute__((target("sme")))
-void test_sme(svbool_t pg, void *ptr) __arm_streaming __arm_shared_za {
+void test_sme(svbool_t pg, void *ptr) __arm_streaming __arm_inout("za") {
   svld1_hor_za8(0, 0, pg, ptr);
 }
 
 __attribute__((target("arch=armv8-a+sme")))
-void test_arch_sme(svbool_t pg, void *ptr) __arm_streaming __arm_shared_za {
+void test_arch_sme(svbool_t pg, void *ptr) __arm_streaming __arm_inout("za") {
   svld1_hor_vnum_za32(0, 0, pg, ptr, 0);
 }
 
 __attribute__((target("+sme")))
-void test_plus_sme(svbool_t pg, void *ptr) __arm_streaming __arm_shared_za {
+void test_plus_sme(svbool_t pg, void *ptr) __arm_streaming __arm_inout("za") {
   svst1_ver_za16(0, 0, pg, ptr);
 }
 
 __attribute__((target("+sme")))
-void undefined(svbool_t pg, void *ptr) __arm_shared_za {
+void undefined(svbool_t pg, void *ptr) __arm_inout("za") {
   svst1_ver_vnum_za64(0, 0, pg, ptr, 0); // expected-warning {{builtin call has undefined behaviour when called from a non-streaming function}}
 }
diff --git a/clang/test/Sema/aarch64-sme2-intrinsics/acle_sme2_imm.cpp b/clang/test/Sema/aarch64-sme2-intrinsics/acle_sme2_imm.cpp
index 5118f743174c2..36103480861c3 100644
--- a/clang/test/Sema/aarch64-sme2-intrinsics/acle_sme2_imm.cpp
+++ b/clang/test/Sema/aarch64-sme2-intrinsics/acle_sme2_imm.cpp
@@ -5,7 +5,7 @@
 
 #include 
 
-void test_multivector_read(uint32_t base) __arm_streaming __arm_shared_za __arm_preserves_za {
+void test_multivector_read(uint32_t base) __arm_streaming __arm_in("za") {
 
   // Test Tile Range
   svread_hor_za8_u8_vg2(1, base); // expected-error {{argument value 1 is outside the valid range [0, 0]}}
@@ -32,7 +32,7 @@ void test_multivector_read(uint32_t base) __arm_streaming __arm_shared_za __arm_
 void test_multivector_write(uint32_t base, svuint8x2_t v8x2, svuint8x4_t v8x4,
                             svuint16x2_t v16x2, svuint16x4_t v16x4,
                             svuint32x2_t v32x2, svuint32x4_t v32x4,
-                            svuint64x2_t v64x2, svuint64x4_t v64x4) __arm_streaming __arm_shared_za {
+                            svuint64x2_t v64x2, svuint64x4_t v64x4) __arm_streaming __arm_inout("za") {
 
   // Test Tile Range
   svwrite_hor_za8_u8_vg2(1, base, v8x2); // expected-error {{argument value 1 is outside the valid range [0, 0]}}
@@ -56,7 +56,7 @@ void test_multivector_write(uint32_t base, svuint8x2_t v8x2, svuint8x4_t v8x4,
   svwrite_ver_za64_u64_vg4(8, base, v64x4); // expected-error {{argument value 8 is outside the valid range [0, 7]}}
 }
 
-void test_outer_product(svbool_t pred, svint16_t s16, svuint16_t u16, svint32_t s32, svuint32_t u32) __arm_streaming __arm_shared_za {
+void test_outer_product(svbool_t pred, svint16_t s16, svuint16_t u16, svint32_t s32, svuint32_t u32) __arm_streaming __arm_inout("za") {
   // Test Tile Range
   svmopa_za32_u16_m(4, pred, pred, u16, u16); // expected-error {{argument value 4 is outside the valid range [0, 3]}}
   svmopa_za32_s16_m(4, pred, pred, s16, s16); // expected-error {{argument value 4 is outside the valid range [0, 3]}}
@@ -71,15 +71,15 @@ void test_outer_product(svbool_t pred, svint16_t s16, svuint16_t u16, svint32_t
   svbmops_za32_s32_m(4, pred, pred, s32, s32); // expected-error {{argument value 4 is outside the valid range [0, 3]}}
 }
 
-void test_ldr_zt(const void *const_base) __arm_streaming_compatible __arm_shared_za {
+void test_ldr_zt(const void *const_base) __arm_streaming_compatible __arm_inout("za") {
   svldr_zt(1, const_base); // expected-error {{argument value 1 is outside the valid range [0, 0]}}
 }
 
-void test_str_zt(void *base) __arm_streaming_compatible __arm_shared_za __arm_preserves_za {
+void test_str_zt(void *base) __arm_streaming_compatible __arm_in("za") {
   svstr_zt(1, base);       // expected-error {{argument value 1 is outside the valid range [0, 0]}}
 }
 
-void test_svluti2_lane_zt_x4(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+void test_svluti2_lane_zt_x4(svuint8_t zn) __arm_streaming __arm_in("za") {
   // Test Reg Offset
   svluti2_lane_zt_u8_x4(1, zn, 0);   // expected-error {{argument value 1 is outside the valid range [0, 0]}}
   // Test index value range
@@ -106,7 +106,7 @@ void test_svluti2_lane_zt_x4(svuint8_t zn) __arm_streaming __arm_shared_za __arm
   svluti2_lane_zt_f32_x4(0, zn, 4);   // expected-error {{argument value 4 is outside the valid range [0, 3]}}
 }
 
-void test_svluti4_lane_zt_x4(svuint8_t zn) __arm_streaming __arm_shared_za __arm_preserves_za {
+void test_svluti4_lane_zt_x4(svuint8_t zn) __arm_streaming __arm_in("za") {
   // Test Reg Offset
   svluti4_lane_zt_u16_x4(1, zn, 0);   // expected-error {{argument value 1 is outside the valid range [0, 0]}}
   // Test index value range
@@ -129,7 +129,7 @@ void test_svluti4_lane_zt_x4(svuint8_t zn) __arm_streaming __arm_shared_za __arm
   svluti4_lane_zt_f32_x4(0, zn, 2);   // expected-error {{argument value 2 is outside the valid range [0, 1]}}
 }
 
-void test_svluti2_lane_zt(svuint8_t zn_u8) __arm_streaming __arm_shared_za __arm_preserves_za {
+void test_svluti2_lane_zt(svuint8_t zn_u8) __arm_streaming __arm_in("za") {
   // Test Reg Offset
   svluti2_lane_zt_u8(1, zn_u8, 2);    // expected-error {{argument value 1 is outside the valid range [0, 0]}}
   // Test index value range
@@ -156,7 +156,7 @@ void test_svluti2_lane_zt(svuint8_t zn_u8) __arm_streaming __arm_shared_za __arm
   svluti2_lane_zt_f32(0, zn_u8, 16); // expected-error {{argument value 16 is outside the valid range [0, 15]}}
 }
 
-void test_svluti4_lane_zt(svuint8_t zn_u8) __arm_streaming __arm_shared_za __arm_preserves_za {
+void test_svluti4_lane_zt(svuint8_t zn_u8) __arm_streaming __arm_in("za") {
   // Test Reg Offset
   svluti4_lane_zt_u8(1, zn_u8, 2);   // expected-error {{argument value 1 is outside the valid range [0, 0]}}
   // Test index value range
@@ -183,7 +183,7 @@ void test_svluti4_lane_zt(svuint8_t zn_u8) __arm_streaming __arm_shared_za __arm
   svluti4_lane_zt_f32(0, zn_u8, 8); // expected-error {{argument value 8 is outside the valid range [0, 7]}}
 }
 
-void test_svluti2_lane_zt_x2(svuint8_t zn_u8) __arm_streaming __arm_shared_za __arm_preserves_za {
+void test_svluti2_lane_zt_x2(svuint8_t zn_u8) __arm_streaming __arm_in("za") {
   // Test Reg Offset
   svluti2_lane_zt_u8_x2(1, zn_u8, 2);    // expected-error {{argument value 1 is outside the valid range [0, 0]}}
   // Test index value range
@@ -210,7 +210,7 @@ void test_svluti2_lane_zt_x2(svuint8_t zn_u8) __arm_streaming __arm_shared_za __
   svluti2_lane_zt_f32_x2(0, zn_u8, 8); // expected-error {{argument value 8 is outside the valid range [0, 7]}}
 }
 
-void test_svluti4_lane_zt_x2(svuint8_t zn_u8) __arm_streaming __arm_shared_za __arm_preserves_za {
+void test_svluti4_lane_zt_x2(svuint8_t zn_u8) __arm_streaming __arm_in("za") {
   // Test Reg Offset
   svluti4_lane_zt_u8_x2(1, zn_u8, 2);   // expected-error {{argument value 1 is outside the valid range [0, 0]}}
   // Test index value range
@@ -245,7 +245,7 @@ void test_bfmlslb_bad_lane(svfloat32_t zda, svbfloat16_t zn, svbfloat16_t zm) __
 void test_multiply_add_sub_long(uint32_t base, svint8_t s8, svuint8_t u8,
                                 svint16_t s16, svuint16_t u16, svint8x2_t s8x2,
                                 svuint8x2_t u8x2, svint16x2_t s16x2, svuint16x2_t u16x2,
-                                svint8x4_t s8x4, svuint8x4_t u8x4, svint16x4_t s16x4, svuint16x4_t u16x4) __arm_streaming __arm_shared_za {
+                                svint8x4_t s8x4, svuint8x4_t u8x4, svint16x4_t s16x4, svuint16x4_t u16x4) __arm_streaming __arm_inout("za") {
 
   svmla_lane_za32_s8_vg4x1(base, s8, s8, 16);   // expected-error {{argument value 16 is outside the valid range [0, 15]}}
   svmla_lane_za32_u8_vg4x1(base, u8, u8, 16);   // expected-error {{argument value 16 is outside the valid range [0, 15]}}
@@ -292,7 +292,7 @@ void test_vertical_dot_product(uint32_t base, svint16x2_t s16x2, svuint16x2_t u1
                                svfloat16x2_t f16x2, svbfloat16x2_t bf16x2,
                                svint16_t s16, svuint16_t u16,
                                svint8_t s8, svuint8_t u8,
-                               svfloat16_t f16, svbfloat16_t b16) __arm_streaming __arm_shared_za {
+                               svfloat16_t f16, svbfloat16_t b16) __arm_streaming __arm_inout("za") {
   // Test lane indices.
   svvdot_lane_za32_s16_vg1x2(base, s16x2, s16, 4);   // expected-error {{argument value 4 is outside the valid range [0, 3]}}
   svvdot_lane_za32_u16_vg1x2(base, u16x2, u16, 4);   // expected-error {{argument value 4 is outside the valid range [0, 3]}}
@@ -309,7 +309,7 @@ void test_vertical_dot_product(uint32_t base, svint16x2_t s16x2, svuint16x2_t u1
 void test_fdot_za32_bad_lane(uint32_t slice_base, svfloat16_t z_f16,
                              svfloat16x2_t z_f16x2, svfloat16x4_t z_f16x4,
                              svbfloat16_t z_bf16, svbfloat16x2_t z_bf16x2,
-                             svbfloat16x4_t z_bf16x4) __arm_streaming __arm_shared_za {
+                             svbfloat16x4_t z_bf16x4) __arm_streaming __arm_inout("za") {
   // 16-bit float
   svdot_lane_za32_f16_vg1x2(slice_base, z_f16x2, z_f16, 4); // expected-error {{argument value 4 is outside the valid range [0, 3]}}
   svdot_lane_za32_f16_vg1x4(slice_base, z_f16x4, z_f16, 4); // expected-error {{argument value 4 is outside the valid range [0, 3]}}
@@ -325,7 +325,7 @@ void test_svdot_multi_za32_bad_lane(uint32_t slice_base, svuint16_t z_u16,
                                     svint16x4_t z_s16x4, svuint8_t z_u8,
                                     svuint8x2_t z_u8x2, svuint8x4_t z_u8x4,
                                     svint8_t z_s8, svint8x2_t z_s8x2,
-                                    svint8x4_t z_s8x4) __arm_streaming __arm_shared_za {
+                                    svint8x4_t z_s8x4) __arm_streaming __arm_inout("za") {
   // Multi, indexed (unsigned)
   svdot_lane_za32_u16_vg1x2(slice_base, z_u16x2, z_u16, 4); // expected-error {{argument value 4 is outside the valid range [0, 3]}}
   svdot_lane_za32_u16_vg1x4(slice_base, z_u16x4, z_u16, 4); // expected-error {{argument value 4 is outside the valid range [0, 3]}}
diff --git a/clang/test/Sema/aarch64-sve2p1-intrinsics/acle_sve2p1_imm.cpp b/clang/test/Sema/aarch64-sve2p1-intrinsics/acle_sve2p1_imm.cpp
index a6154daadeeab..d857608d44155 100644
--- a/clang/test/Sema/aarch64-sve2p1-intrinsics/acle_sve2p1_imm.cpp
+++ b/clang/test/Sema/aarch64-sve2p1-intrinsics/acle_sve2p1_imm.cpp
@@ -173,18 +173,18 @@ void test_svpmov_lane(){
 
 __attribute__((target("+sve2p1")))
 void test_svget_svset_b(uint64_t idx, svboolx2_t tuple2, svboolx4_t tuple4, svbool_t res){
-  svset2(tuple2, -1, res); // expected-error {{argument value 18446744073709551615 is outside the valid range [0, 1]}}
-  svset2(tuple2, 2,  res); // expected-error {{argument value 2 is outside the valid range [0, 1]}}
-  svset4(tuple4, -1, res); // expected-error {{argument value 18446744073709551615 is outside the valid range [0, 3]}}
-  svset4(tuple4, 4,  res); // expected-error {{argument value 4 is outside the valid range [0, 3]}}
-
-  svget2(tuple2, -1); // expected-error {{argument value 18446744073709551615 is outside the valid range [0, 1]}}
-  svget2(tuple2,  2); // expected-error {{argument value 2 is outside the valid range [0, 1]}}
-  svget4(tuple4, -1); // expected-error {{argument value 18446744073709551615 is outside the valid range [0, 3]}}
-  svget4(tuple4,  4); // expected-error {{argument value 4 is outside the valid range [0, 3]}}
-
-  svset2(tuple2, idx, res); // expected-error {{argument to 'svset2' must be a constant integer}}
-  svset4(tuple4, idx, res); // expected-error {{argument to 'svset4' must be a constant integer}}
-  svget2(tuple2, idx); // expected-error {{argument to 'svget2' must be a constant integer}}
-  svget4(tuple4, idx); // expected-error {{argument to 'svget4' must be a constant integer}}
+  svset2_b(tuple2, -1, res); // expected-error {{argument value 18446744073709551615 is outside the valid range [0, 1]}}
+  svset2_b(tuple2, 2,  res); // expected-error {{argument value 2 is outside the valid range [0, 1]}}
+  svset4_b(tuple4, -1, res); // expected-error {{argument value 18446744073709551615 is outside the valid range [0, 3]}}
+  svset4_b(tuple4, 4,  res); // expected-error {{argument value 4 is outside the valid range [0, 3]}}
+
+  svget2_b(tuple2, -1); // expected-error {{argument value 18446744073709551615 is outside the valid range [0, 1]}}
+  svget2_b(tuple2,  2); // expected-error {{argument value 2 is outside the valid range [0, 1]}}
+  svget4_b(tuple4, -1); // expected-error {{argument value 18446744073709551615 is outside the valid range [0, 3]}}
+  svget4_b(tuple4,  4); // expected-error {{argument value 4 is outside the valid range [0, 3]}}
+
+  svset2_b(tuple2, idx, res); // expected-error {{argument to 'svset2_b' must be a constant integer}}
+  svset4_b(tuple4, idx, res); // expected-error {{argument to 'svset4_b' must be a constant integer}}
+  svget2_b(tuple2, idx); // expected-error {{argument to 'svget2_b' must be a constant integer}}
+  svget4_b(tuple4, idx); // expected-error {{argument to 'svget4_b' must be a constant integer}}
 }
diff --git a/clang/test/Sema/inline-asm-validate-x86.c b/clang/test/Sema/inline-asm-validate-x86.c
index 87b60a0955301..d4c10c36f5fbe 100644
--- a/clang/test/Sema/inline-asm-validate-x86.c
+++ b/clang/test/Sema/inline-asm-validate-x86.c
@@ -130,3 +130,11 @@ void pr40890(void) {
   __asm__ __volatile__("\n#define BEEF abcd%0\n" : : "n"((int*)0xdeadbeeeeeef));
 #endif
 }
+
+void test_W(int i) {
+  __asm__("" : : "Wd"(test_W)); // expected-error{{invalid input constraint 'Wd' in asm}}
+
+  __asm__("" : : "Ws"(test_W(0))); // expected-error{{invalid type 'void' in asm input for constraint 'Ws'}}
+  // Codegen error
+  __asm__("" : : "Ws"(i));
+}
diff --git a/clang/test/Sema/ns_error_enum.m b/clang/test/Sema/ns_error_enum.m
index 0b8f8f7cced00..29027f1a54f03 100644
--- a/clang/test/Sema/ns_error_enum.m
+++ b/clang/test/Sema/ns_error_enum.m
@@ -53,7 +53,6 @@ typedef NS_ERROR_ENUM(unsigned char, MyCFTypedefErrorEnum, MyCFTypedefErrorDomai
 
 extern char *const WrongErrorDomainType;
 enum __attribute__((ns_error_domain(WrongErrorDomainType))) MyWrongErrorDomainType { MyWrongErrorDomain };
-// expected-error@-1{{domain argument 'WrongErrorDomainType' does not point to an NSString or CFString constant}}
 
 struct __attribute__((ns_error_domain(MyErrorDomain))) MyStructWithErrorDomain {};
 // expected-error@-1{{'ns_error_domain' attribute only applies to enums}}
@@ -68,7 +67,7 @@ typedef NS_ERROR_ENUM(unsigned char, MyCFTypedefErrorEnum, MyCFTypedefErrorDomai
 // expected-error@-1{{'ns_error_domain' attribute takes one argument}}
 
 typedef NS_ERROR_ENUM(unsigned char, MyErrorEnumInvalid, InvalidDomain) {
-	// expected-error@-1{{use of undeclared identifier 'InvalidDomain'}}
+	// expected-error@-1{{domain argument 'InvalidDomain' does not refer to global constant}}
 	MyErrFirstInvalid,
 	MyErrSecondInvalid,
 };
diff --git a/clang/test/SemaCXX/PR76631.cpp b/clang/test/SemaCXX/PR76631.cpp
new file mode 100644
index 0000000000000..947fa3fc2635e
--- /dev/null
+++ b/clang/test/SemaCXX/PR76631.cpp
@@ -0,0 +1,9 @@
+// RUN: %clang_cc1 -verify -std=c++11 -fsyntax-only %s
+
+[[noreturn]] void throw_int() {
+  throw int(); // expected-error {{cannot use 'throw' with exceptions disabled}}
+}
+
+void throw_int_wrapper() {
+  [[clang::musttail]] return throw_int(); // expected-error {{'musttail' attribute may not be used with no-return-attribute functions}}
+}
diff --git a/clang/test/SemaCXX/builtin-std-move.cpp b/clang/test/SemaCXX/builtin-std-move.cpp
index adad66afc2cbd..e20bf8e156e19 100644
--- a/clang/test/SemaCXX/builtin-std-move.cpp
+++ b/clang/test/SemaCXX/builtin-std-move.cpp
@@ -1,6 +1,6 @@
-// RUN: %clang_cc1 -std=c++17 -verify %s
-// RUN: %clang_cc1 -std=c++17 -verify %s -DNO_CONSTEXPR
-// RUN: %clang_cc1 -std=c++20 -verify %s
+// RUN: %clang_cc1 -std=c++17 -verify=cxx17,expected %s
+// RUN: %clang_cc1 -std=c++17 -verify=cxx17,expected %s -DNO_CONSTEXPR
+// RUN: %clang_cc1 -std=c++20 -verify=cxx20,expected %s
 
 namespace std {
 #ifndef NO_CONSTEXPR
@@ -12,6 +12,7 @@ namespace std {
   template CONSTEXPR T &&move(T &x) {
     static_assert(T::moveable, "instantiated move"); // expected-error {{no member named 'moveable' in 'B'}}
                                                      // expected-error@-1 {{no member named 'moveable' in 'C'}}
+                                                     // expected-error@-2 {{no member named 'moveable' in 'D'}}
     return static_cast(x);
   }
 
@@ -23,6 +24,7 @@ namespace std {
 
   template CONSTEXPR auto move_if_noexcept(T &x) -> typename ref(x)))>::type {
     static_assert(T::moveable, "instantiated move_if_noexcept"); // expected-error {{no member named 'moveable' in 'B'}}
+                                                                 // expected-error@-1 {{no member named 'moveable' in 'D'}}
     return static_cast(x)))>::type>(x);
   }
 
@@ -36,6 +38,7 @@ namespace std {
   template CONSTEXPR T &&forward(typename remove_reference::type &x) {
     static_assert(T::moveable, "instantiated forward"); // expected-error {{no member named 'moveable' in 'B'}}
                                                         // expected-error@-1 {{no member named 'moveable' in 'C'}}
+                                                        // expected-error@-2 {{no member named 'moveable' in 'D'}}
     return static_cast(x);
   }
   template CONSTEXPR T &&forward(typename remove_reference::type &&x) {
@@ -67,21 +70,25 @@ namespace std {
   CONSTEXPR auto forward_like(T &&t) -> ForwardLikeRetType {
     using TT = typename remove_reference::type;
     static_assert(TT::moveable, "instantiated as_const"); // expected-error {{no member named 'moveable' in 'B'}}
+                                                          // expected-error@-1 {{no member named 'moveable' in 'D'}}
     return static_cast>(t);
   }
 
   template CONSTEXPR const T &as_const(T &x) {
     static_assert(T::moveable, "instantiated as_const"); // expected-error {{no member named 'moveable' in 'B'}}
+                                                         // expected-error@-1 {{no member named 'moveable' in 'D'}}
     return x;
   }
 
   template CONSTEXPR T *addressof(T &x) {
     static_assert(T::moveable, "instantiated addressof"); // expected-error {{no member named 'moveable' in 'B'}}
+                                                          // expected-error@-1 {{no member named 'moveable' in 'D'}}
     return __builtin_addressof(x);
   }
 
   template CONSTEXPR T *__addressof(T &x) {
     static_assert(T::moveable, "instantiated __addressof"); // expected-error {{no member named 'moveable' in 'B'}}
+                                                            // expected-error@-1 {{no member named 'moveable' in 'D'}}
     return __builtin_addressof(x);
   }
 }
@@ -116,42 +123,20 @@ A &forward_rval_as_lval() {
 }
 
 struct B {};
-B &&(*pMove)(B&) = std::move; // #1 expected-note {{instantiation of}}
-B &&(*pMoveIfNoexcept)(B&) = &std::move_if_noexcept; // #2 expected-note {{instantiation of}}
-B &&(*pForward)(B&) = &std::forward; // #3 expected-note {{instantiation of}}
-B &&(*pForwardLike)(B&) = &std::forward_like; // #4 expected-note {{instantiation of}}
-const B &(*pAsConst)(B&) = &std::as_const; // #5 expected-note {{instantiation of}}
-B *(*pAddressof)(B&) = &std::addressof; // #6 expected-note {{instantiation of}}
-B *(*pUnderUnderAddressof)(B&) = &std::__addressof; // #7 expected-note {{instantiation of}}
+B &&(*pMove)(B&) = std::move;                            // cxx17-warning {{non-addressable}} cxx20-error {{non-addressable}} expected-note {{instantiation of}}
+B &&(*pMoveIfNoexcept)(B&) = &std::move_if_noexcept;     // cxx17-warning {{non-addressable}} cxx20-error {{non-addressable}} expected-note {{instantiation of}}
+B &&(*pForward)(B&) = &std::forward;                  // cxx17-warning {{non-addressable}} cxx20-error {{non-addressable}} expected-note {{instantiation of}}
+B &&(*pForwardLike)(B&) = &std::forward_like; // cxx17-warning {{non-addressable}} cxx20-error {{non-addressable}} expected-note {{instantiation of}}
+const B &(*pAsConst)(B&) = &std::as_const;               // cxx17-warning {{non-addressable}} cxx20-error {{non-addressable}} expected-note {{instantiation of}}
+B *(*pAddressof)(B&) = &std::addressof;                  // cxx17-warning {{non-addressable}} cxx20-error {{non-addressable}} expected-note {{instantiation of}}
+B *(*pUnderUnderAddressof)(B&) = &std::__addressof;      // cxx17-warning {{non-addressable}} cxx20-error {{non-addressable}} expected-note {{instantiation of}}
 int (*pUnrelatedMove)(B, B) = std::move;
 
 struct C {};
-C &&(&rMove)(C&) = std::move; // #8 expected-note {{instantiation of}}
-C &&(&rForward)(C&) = std::forward; // #9 expected-note {{instantiation of}}
+C &&(&rMove)(C&) = std::move;          // cxx17-warning {{non-addressable}} cxx20-error {{non-addressable}} expected-note {{instantiation of}}
+C &&(&rForward)(C&) = std::forward; // cxx17-warning {{non-addressable}} cxx20-error {{non-addressable}} expected-note {{instantiation of}}
 int (&rUnrelatedMove)(B, B) = std::move;
 
-#if __cplusplus <= 201703L
-// expected-warning@#1 {{non-addressable}}
-// expected-warning@#2 {{non-addressable}}
-// expected-warning@#3 {{non-addressable}}
-// expected-warning@#4 {{non-addressable}}
-// expected-warning@#5 {{non-addressable}}
-// expected-warning@#6 {{non-addressable}}
-// expected-warning@#7 {{non-addressable}}
-// expected-warning@#8 {{non-addressable}}
-// expected-warning@#9 {{non-addressable}}
-#else
-// expected-error@#1 {{non-addressable}}
-// expected-error@#2 {{non-addressable}}
-// expected-error@#3 {{non-addressable}}
-// expected-error@#4 {{non-addressable}}
-// expected-error@#5 {{non-addressable}}
-// expected-error@#6 {{non-addressable}}
-// expected-error@#7 {{non-addressable}}
-// expected-error@#8 {{non-addressable}}
-// expected-error@#9 {{non-addressable}}
-#endif
-
 void attribute_const() {
   int n;
   std::move(n); // expected-warning {{ignoring return value}}
@@ -163,6 +148,22 @@ void attribute_const() {
   std::as_const(n); // expected-warning {{ignoring return value}}
 }
 
+struct D {
+  void* operator new(__SIZE_TYPE__, D&&(*)(D&));
+  void* operator new(__SIZE_TYPE__, D*(*)(D&));
+  void* operator new(__SIZE_TYPE__, const D&(*)(D&));
+};
+
+void placement_new() {
+  new (std::move) D;             // cxx17-warning {{non-addressable}} cxx20-error {{non-addressable}} expected-note {{instantiation of}}
+  new (std::move_if_noexcept) D; // cxx17-warning {{non-addressable}} cxx20-error {{non-addressable}} expected-note {{instantiation of}}
+  new (std::forward) D;          // cxx17-warning {{non-addressable}} cxx20-error {{non-addressable}} expected-note {{instantiation of}}
+  new (std::forward_like) D;     // cxx17-warning {{non-addressable}} cxx20-error {{non-addressable}} expected-note {{instantiation of}}
+  new (std::addressof) D;        // cxx17-warning {{non-addressable}} cxx20-error {{non-addressable}} expected-note {{instantiation of}}
+  new (std::__addressof) D;      // cxx17-warning {{non-addressable}} cxx20-error {{non-addressable}} expected-note {{instantiation of}}
+  new (std::as_const) D;         // cxx17-warning {{non-addressable}} cxx20-error {{non-addressable}} expected-note {{instantiation of}}
+}
+
 namespace std {
   template int &move(T);
 }
diff --git a/clang/test/SemaCXX/constexpr-ackermann.cpp b/clang/test/SemaCXX/constexpr-ackermann.cpp
index c4ea3139483df..9d368adc8f7a2 100644
--- a/clang/test/SemaCXX/constexpr-ackermann.cpp
+++ b/clang/test/SemaCXX/constexpr-ackermann.cpp
@@ -1,4 +1,5 @@
 // RUN: %clang_cc1 -std=c++11 -fsyntax-only %s
+// RUN: %clang_cc1 -std=c++11 -fsyntax-only -fexperimental-new-constant-interpreter %s
 
 constexpr unsigned long long A(unsigned long long m, unsigned long long n) {
   return m == 0 ? n + 1 : n == 0 ? A(m-1, 1) : A(m - 1, A(m, n - 1));
diff --git a/clang/test/SemaCXX/constexpr-default-arg.cpp b/clang/test/SemaCXX/constexpr-default-arg.cpp
index 0cef4aa578af3..7c88369282954 100644
--- a/clang/test/SemaCXX/constexpr-default-arg.cpp
+++ b/clang/test/SemaCXX/constexpr-default-arg.cpp
@@ -1,4 +1,5 @@
 // RUN: %clang_cc1 -std=c++1y -S -o - -emit-llvm -verify %s
+// RUN: %clang_cc1 -std=c++1y -fexperimental-new-constant-interpreter -S -o - -emit-llvm -verify %s
 
 namespace default_arg_temporary {
 
@@ -33,6 +34,6 @@ void test_default_arg2() {
 // Check that multiple CXXDefaultInitExprs don't cause an assertion failure.
 struct A { int &&r = 0; }; // expected-note 2{{default member initializer}}
 struct B { A x, y; };
-B b = {}; // expected-warning 2{{not supported}}
+B b = {}; // expected-warning 2{{lifetime extension of temporary created by aggregate initialization using a default member initializer is not yet supported}}
 
 }
diff --git a/clang/test/SemaCXX/constexpr-late-instantiation.cpp b/clang/test/SemaCXX/constexpr-late-instantiation.cpp
index ec8e071217c1d..9aec0c90e61dc 100644
--- a/clang/test/SemaCXX/constexpr-late-instantiation.cpp
+++ b/clang/test/SemaCXX/constexpr-late-instantiation.cpp
@@ -1,4 +1,5 @@
 // RUN: %clang_cc1 %s -fsyntax-only -verify
+// RUN: %clang_cc1 %s -fexperimental-new-constant-interpreter -fsyntax-only -verify
 
 template 
 constexpr T foo(T a);   // expected-note {{declared here}}
diff --git a/clang/test/SemaCXX/constexpr-nqueens.cpp b/clang/test/SemaCXX/constexpr-nqueens.cpp
index 47133a2934340..b28397bfb2191 100644
--- a/clang/test/SemaCXX/constexpr-nqueens.cpp
+++ b/clang/test/SemaCXX/constexpr-nqueens.cpp
@@ -1,4 +1,5 @@
 // RUN: %clang_cc1 -std=c++11 -fsyntax-only %s
+// RUN: %clang_cc1 -std=c++11 -fsyntax-only -fexperimental-new-constant-interpreter %s
 
 typedef unsigned long uint64_t;
 
diff --git a/clang/test/SemaCXX/constexpr-unsigned-high-bit.cpp b/clang/test/SemaCXX/constexpr-unsigned-high-bit.cpp
index 19d8dcab66073..d85e645509d8e 100644
--- a/clang/test/SemaCXX/constexpr-unsigned-high-bit.cpp
+++ b/clang/test/SemaCXX/constexpr-unsigned-high-bit.cpp
@@ -1,4 +1,5 @@
 // RUN: %clang_cc1 -std=c++14 -fsyntax-only %s
+// RUN: %clang_cc1 -std=c++14 -fsyntax-only -fexperimental-new-constant-interpreter %s
 
 #include 
 
diff --git a/clang/test/SemaCXX/conversion-function.cpp b/clang/test/SemaCXX/conversion-function.cpp
index b6e6142d17906..749e2fc1b452b 100644
--- a/clang/test/SemaCXX/conversion-function.cpp
+++ b/clang/test/SemaCXX/conversion-function.cpp
@@ -475,13 +475,22 @@ struct S {
 
 #if __cplusplus >= 201103L
 namespace dependent_conversion_function_id_lookup {
-  template struct A {
-    operator T();
-  };
-  template struct B : A {
-    template using Lookup = decltype(&B::operator U);
-  };
-  using Result = B::Lookup;
-  using Result = int (A::*)();
+namespace gh77583 {
+struct A1 {
+  operator int();
+};
+template struct C {
+  template  using Lookup = decltype(T{}.operator U());
+};
+C v{};
+}
+template struct A2 {
+  operator T();
+};
+template struct B : A2 {
+  template using Lookup = decltype(&B::operator U);
+};
+using Result = B::Lookup;
+using Result = int (A2::*)();
 }
 #endif
diff --git a/clang/test/SemaCXX/coro-lifetimebound.cpp b/clang/test/SemaCXX/coro-lifetimebound.cpp
index 3fc7ca70a14a1..9e96a296562a0 100644
--- a/clang/test/SemaCXX/coro-lifetimebound.cpp
+++ b/clang/test/SemaCXX/coro-lifetimebound.cpp
@@ -64,6 +64,10 @@ Co bar_coro(const int &b, int c) {
       : bar_coro(0, 1); // expected-warning {{returning address of local temporary object}}
 }
 
+// =============================================================================
+// Lambdas
+// =============================================================================
+namespace lambdas {
 void lambdas() {
   auto unsafe_lambda = [] [[clang::coro_wrapper]] (int b) {
     return foo_coro(b); // expected-warning {{address of stack memory associated with parameter}}
@@ -84,6 +88,51 @@ void lambdas() {
     co_return x + co_await foo_coro(b);
   };
 }
+
+Co lambda_captures() {
+  int a = 1;
+  // Temporary lambda object dies.
+  auto lamb = [a](int x, const int& y) -> Co { // expected-warning {{temporary whose address is used as value of local variable 'lamb'}}
+    co_return x + y + a;
+  }(1, a);
+  // Object dies but it has no capture.
+  auto no_capture = []() -> Co { co_return 1; }();
+  auto bad_no_capture = [](const int& a) -> Co { co_return a; }(1); // expected-warning {{temporary}}
+  // Temporary lambda object with lifetime extension under co_await.
+  int res = co_await [a](int x, const int& y) -> Co {
+    co_return x + y + a;
+  }(1, a);
+  // Lambda object on stack should be fine.
+  auto lamb2 = [a]() -> Co { co_return a; };
+  auto on_stack = lamb2();
+  auto res2 = co_await on_stack;
+  co_return 1;
+}
+} // namespace lambdas
+
+// =============================================================================
+// Member coroutines
+// =============================================================================
+namespace member_coroutines{
+struct S {
+  Co member(const int& a) { co_return a; }
+};
+
+Co use() {
+  S s;
+  int a = 1;
+  auto test1 = s.member(1);  // expected-warning {{temporary whose address is used as value of local variable}}
+  auto test2 = s.member(a);
+  auto test3 = S{}.member(a);  // expected-warning {{temporary whose address is used as value of local variable}}
+  co_return 1;
+}
+
+[[clang::coro_wrapper]] Co wrapper(const int& a) {
+  S s;
+  return s.member(a); // expected-warning {{address of stack memory}}
+}
+} // member_coroutines
+
 // =============================================================================
 // Safe usage when parameters are value
 // =============================================================================
@@ -91,7 +140,7 @@ namespace by_value {
 Co value_coro(int b) { co_return co_await foo_coro(b); }
 [[clang::coro_wrapper]] Co wrapper1(int b) { return value_coro(b); }
 [[clang::coro_wrapper]] Co wrapper2(const int& b) { return value_coro(b); }
-}
+} // namespace by_value
 
 // =============================================================================
 // Lifetime bound but not a Coroutine Return Type: No analysis.
@@ -122,11 +171,29 @@ CoNoCRT bar(int a) {
 namespace disable_lifetimebound {
 Co foo(int x) {  co_return x; }
 
-[[clang::coro_wrapper, clang::coro_disable_lifetimebound]] 
+[[clang::coro_wrapper, clang::coro_disable_lifetimebound]]
 Co foo_wrapper(const int& x) { return foo(x); }
 
 [[clang::coro_wrapper]] Co caller() {
   // The call to foo_wrapper is wrapper is safe.
   return foo_wrapper(1);
 }
+
+struct S{
+[[clang::coro_wrapper, clang::coro_disable_lifetimebound]]
+Co member(const int& x) { return foo(x); }
+};
+
+Co use() {
+  S s;
+  int a = 1;
+  auto test1 = s.member(1); // param is not flagged.
+  auto test2 = S{}.member(a); // 'this' is not flagged.
+  co_return 1;
+}
+
+[[clang::coro_wrapper]] Co return_stack_addr(const int& a) {
+  S s;
+  return s.member(a); // return of stack addr is not flagged.
+}
 } // namespace disable_lifetimebound
diff --git a/clang/test/SemaCXX/coro-return-type-and-wrapper.cpp b/clang/test/SemaCXX/coro-return-type-and-wrapper.cpp
index ac49e03ba9d90..b08e1c9c065a0 100644
--- a/clang/test/SemaCXX/coro-return-type-and-wrapper.cpp
+++ b/clang/test/SemaCXX/coro-return-type-and-wrapper.cpp
@@ -5,11 +5,23 @@ using std::suspend_always;
 using std::suspend_never;
 
 
+namespace std {
+  struct nothrow_t {};
+  constexpr nothrow_t nothrow = {};
+}
+
+using SizeT = decltype(sizeof(int));
+
+void* operator new(SizeT __sz, const std::nothrow_t&) noexcept;
+
 template  struct [[clang::coro_return_type]] Gen {
   struct promise_type {
     Gen get_return_object() {
       return {};
     }
+    static Gen get_return_object_on_allocation_failure() {
+      return {};
+    }
     suspend_always initial_suspend();
     suspend_always final_suspend() noexcept;
     void unhandled_exception();
diff --git a/clang/test/SemaCXX/cxx1z-class-template-argument-deduction.cpp b/clang/test/SemaCXX/cxx1z-class-template-argument-deduction.cpp
index a490d318f54b5..33ed4295c2e48 100644
--- a/clang/test/SemaCXX/cxx1z-class-template-argument-deduction.cpp
+++ b/clang/test/SemaCXX/cxx1z-class-template-argument-deduction.cpp
@@ -645,4 +645,37 @@ namespace undefined_warnings {
     auto test2 = TemplDObj(.0f);
   }
 }
+
+namespace GH51710 {
+template
+struct A {
+  A(T f()) {}
+  A(int f(), T) {}
+
+  A(T array[10]) {}
+  A(int array[10], T) {}
+};
+
+template
+struct B {
+   B(T array[]) {}
+   B(int array[], T) {}
+};
+
+
+int foo();
+
+void bar() {
+  A test1(foo);
+  A test2(foo, 1);
+
+  int array[10];
+  A test3(array);
+  A test4(array, 1);
+
+  B test5(array);
+  B test6(array, 1);
+}
+} // namespace GH51710
+
 #endif
diff --git a/clang/test/SemaCXX/eval-crashes.cpp b/clang/test/SemaCXX/eval-crashes.cpp
index ac04b113f99b7..017df977b26b7 100644
--- a/clang/test/SemaCXX/eval-crashes.cpp
+++ b/clang/test/SemaCXX/eval-crashes.cpp
@@ -29,7 +29,7 @@ namespace pr33140_2 {
   // temporaries.
   struct A { int &&r = 0; }; // expected-note 2{{initializing field 'r' with default member initializer}}
   struct B { A x, y; };
-  B b = {}; // expected-warning 2{{not supported}}
+  B b = {}; // expected-warning 2{{lifetime extension of temporary created by aggregate initialization using a default member initializer is not yet supported}}
 }
 
 namespace pr33140_3 {
diff --git a/clang/test/SemaCXX/literal-type.cpp b/clang/test/SemaCXX/literal-type.cpp
index 14a4094c45a1b..88535c169fe01 100644
--- a/clang/test/SemaCXX/literal-type.cpp
+++ b/clang/test/SemaCXX/literal-type.cpp
@@ -1,4 +1,6 @@
 // RUN: %clang_cc1 -fsyntax-only -verify -std=c++11 %s
+// RUN: %clang_cc1 -fsyntax-only -verify -std=c++20 %s
+
 
 static_assert(__is_literal(int), "fail");
 static_assert(__is_literal_type(int), "fail"); // alternate spelling for GCC
@@ -75,3 +77,34 @@ template  class HasConstExprCtorT {
 static_assert(__is_literal(HasConstExprCtor), "fail");
 static_assert(__is_literal(HasConstExprCtorTemplate), "fail");
 static_assert(__is_literal(HasConstExprCtorT), "fail");
+
+
+#if __cplusplus >= 202003L
+namespace GH77924 {
+
+struct A { A(); };
+template 
+struct opt {
+  union Data {
+      constexpr Data() : x{} {}
+      constexpr ~Data() {}
+      char x;
+      T data;
+  };
+
+  constexpr opt() : data{} {}
+  constexpr ~opt() { if (engaged) data.data.~T(); }
+  Data data;
+  bool engaged = false;
+};
+
+consteval void foo() {
+  opt a;
+}
+
+void test() {
+  foo();
+}
+
+}
+#endif
diff --git a/clang/test/SemaCXX/ms-property-new.cpp b/clang/test/SemaCXX/ms-property-new.cpp
new file mode 100644
index 0000000000000..9c32c3e564262
--- /dev/null
+++ b/clang/test/SemaCXX/ms-property-new.cpp
@@ -0,0 +1,44 @@
+// RUN: %clang_cc1 -ast-print -verify -triple=x86_64-pc-win32 -fms-compatibility %s -o - | FileCheck %s
+// RUN: %clang_cc1 -triple=x86_64-pc-win32 -fms-compatibility -emit-pch -o %t %s
+// RUN: %clang_cc1 -triple=x86_64-pc-win32 -fms-compatibility -include-pch %t -verify %s -ast-print -o - | FileCheck %s
+// expected-no-diagnostics
+
+#ifndef HEADER
+#define HEADER
+
+struct S {
+  int GetX() { return 42; }
+  __declspec(property(get=GetX)) int x;
+
+  int GetY(int i, int j) { return i+j; }
+  __declspec(property(get=GetY)) int y[];
+
+  void* operator new(__SIZE_TYPE__, int);
+};
+
+template 
+struct TS {
+  T GetT() { return T(); }
+  __declspec(property(get=GetT)) T t;
+
+  T GetR(T i, T j) { return i+j; }
+  __declspec(property(get=GetR)) T r[];
+};
+
+int main(int argc, char **argv) {
+  // CHECK: S *s;
+  // CHECK-NEXT: new (s->x) S;
+  // CHECK-NEXT: new ((s->y)[1][2]) S;
+  S *s;
+  new (s->x) S;
+  new ((s->y)[1][2]) S;
+
+  // CHECK-NEXT: TS *ts;
+  // CHECK-NEXT: new (ts->t) S;
+  // CHECK-NEXT: new ((ts->r)[1][2]) S;
+  TS *ts;
+  new (ts->t) S;
+  new ((ts->r)[1][2]) S;
+}
+
+#endif
diff --git a/clang/test/SemaCXX/placement-new-bound-member-function.cpp b/clang/test/SemaCXX/placement-new-bound-member-function.cpp
new file mode 100644
index 0000000000000..5733894253325
--- /dev/null
+++ b/clang/test/SemaCXX/placement-new-bound-member-function.cpp
@@ -0,0 +1,68 @@
+// RUN: %clang_cc1 -fsyntax-only -verify %s -std=c++11
+
+struct S {
+  int NoArgs();
+  int OneArg(int);
+
+  template 
+  T TemplNoArgs(); // expected-note {{possible target for call}} // expected-note {{possible target for call}}
+
+  template 
+  T TemplOneArg(T); // expected-note {{possible target for call}} // expected-note {{possible target for call}}
+
+  void* operator new(__SIZE_TYPE__, int);
+};
+
+S* GetS();
+
+int test() {
+  S s, *ps = GetS();
+  int (S::*pNoArgs)() = &S::NoArgs;
+  int (S::*pOneArg)(int) = &S::OneArg;
+  int (S::*pTemplNoArgs)() = &S::TemplNoArgs;
+  int (S::*pTemplOneArg)(int) = &S::TemplOneArg;
+
+  new (s.NoArgs()) S;
+  new (s.OneArg(1)) S;
+  new (ps->NoArgs()) S;
+  new (ps->OneArg(1)) S;
+  new ((s.*pNoArgs)()) S;
+  new ((s.*pOneArg)(1)) S;
+  new ((ps->*pNoArgs)()) S;
+  new ((ps->*pOneArg)(1)) S;
+
+  new (s.TemplNoArgs()) S;
+  new (s.TemplOneArg(1)) S;
+  new (ps->TemplNoArgs()) S;
+  new (ps->TemplOneArg(1)) S;
+  new ((s.*pTemplNoArgs)()) S;
+  new ((s.*pTemplOneArg)(1)) S;
+  new ((ps->*pTemplNoArgs)()) S;
+  new ((ps->*pTemplOneArg)(1)) S;
+
+  new (s.NoArgs) S; // expected-error {{reference to non-static member function must be called; did you mean to call it with no arguments?}}
+  new (s.OneArg) S; // expected-error {{reference to non-static member function must be called}}
+  new (ps->NoArgs) S; // expected-error {{reference to non-static member function must be called; did you mean to call it with no arguments?}}
+  new (ps->OneArg) S; // expected-error {{reference to non-static member function must be called}}
+  new (s.*pNoArgs) S; // expected-error {{reference to non-static member function must be called; did you mean to call it with no arguments?}}
+  new (s.*pOneArg) S; // expected-error {{reference to non-static member function must be called}}
+  new (ps->*pNoArgs) S; // expected-error {{reference to non-static member function must be called; did you mean to call it with no arguments?}}
+  new (ps->*pOneArg) S; // expected-error {{reference to non-static member function must be called}}
+  new ((s.*pNoArgs)) S; // expected-error {{reference to non-static member function must be called; did you mean to call it with no arguments?}}
+  new ((s.*pOneArg)) S; // expected-error {{reference to non-static member function must be called}}
+  new ((ps->*pNoArgs)) S; // expected-error {{reference to non-static member function must be called; did you mean to call it with no arguments?}}
+  new ((ps->*pOneArg)) S; // expected-error {{reference to non-static member function must be called}}
+
+  new (s.TemplNoArgs) S; // expected-error {{reference to non-static member function must be called; did you mean to call it with no arguments?}}
+  new (s.TemplOneArg) S; // expected-error {{reference to non-static member function must be called}}
+  new (ps->TemplNoArgs) S; // expected-error {{reference to non-static member function must be called; did you mean to call it with no arguments?}}
+  new (ps->TemplOneArg) S; // expected-error {{reference to non-static member function must be called}}
+  new (s.*pTemplNoArgs) S; // expected-error {{reference to non-static member function must be called; did you mean to call it with no arguments?}}
+  new (s.*pTemplOneArg) S; // expected-error {{reference to non-static member function must be called}}
+  new (ps->*pTemplNoArgs) S; // expected-error {{reference to non-static member function must be called; did you mean to call it with no arguments?}}
+  new (ps->*pTemplOneArg) S; // expected-error {{reference to non-static member function must be called}}
+  new ((s.*pTemplNoArgs)) S; // expected-error {{reference to non-static member function must be called; did you mean to call it with no arguments?}}
+  new ((s.*pTemplOneArg)) S; // expected-error {{reference to non-static member function must be called}}
+  new ((ps->*pTemplNoArgs)) S; // expected-error {{reference to non-static member function must be called; did you mean to call it with no arguments?}}
+  new ((ps->*pTemplOneArg)) S; // expected-error {{reference to non-static member function must be called}}
+}
diff --git a/clang/test/SemaCXX/placement-new-builtin.cpp b/clang/test/SemaCXX/placement-new-builtin.cpp
new file mode 100644
index 0000000000000..5776a94dc3ad8
--- /dev/null
+++ b/clang/test/SemaCXX/placement-new-builtin.cpp
@@ -0,0 +1,11 @@
+// RUN: %clang_cc1 -fsyntax-only -verify %s -std=c++11
+
+struct S {
+  void* operator new(__SIZE_TYPE__, char*);
+  void* operator new(__SIZE_TYPE__, __SIZE_TYPE__);
+};
+
+int main() {
+  new (__builtin_strchr) S; // expected-error {{builtin functions must be directly called}}
+  new ((__builtin_strlen)) S; // expected-error {{builtin functions must be directly called}}
+}
diff --git a/clang/test/SemaCXX/placement-new-matrix.cpp b/clang/test/SemaCXX/placement-new-matrix.cpp
new file mode 100644
index 0000000000000..7168e60bdbb00
--- /dev/null
+++ b/clang/test/SemaCXX/placement-new-matrix.cpp
@@ -0,0 +1,26 @@
+// RUN: %clang_cc1 -fenable-matrix -fsyntax-only -verify %s -std=c++11
+
+using Matrix = int __attribute__((matrix_type(4, 3)));
+
+template <__SIZE_TYPE__ a, __SIZE_TYPE__ b>
+using TMatrix = int __attribute__((matrix_type(a, b)));
+
+struct S {
+    void* operator new(__SIZE_TYPE__, int);
+    void* operator new(__SIZE_TYPE__, Matrix);
+    void* operator new(__SIZE_TYPE__, TMatrix<2, 2>);
+};
+
+int main() {
+    Matrix m;
+    TMatrix<2, 2> tm;
+
+    new (m) S {};
+    new (tm) S {};
+
+    new (m[1][1]) S {};
+    new (tm[1][1]) S {};
+
+    new (m[1]) S {}; // expected-error {{single subscript expressions are not allowed for matrix values}}
+    new (tm[1]) S {}; // expected-error {{single subscript expressions are not allowed for matrix values}}
+}
diff --git a/clang/test/SemaCXX/placement-new-ms-__noop.cpp b/clang/test/SemaCXX/placement-new-ms-__noop.cpp
new file mode 100644
index 0000000000000..a251676c00012
--- /dev/null
+++ b/clang/test/SemaCXX/placement-new-ms-__noop.cpp
@@ -0,0 +1,12 @@
+// RUN: %clang_cc1 -fsyntax-only -fms-extensions -verify %s -std=c++11
+// expected-no-diagnostics
+
+struct S {
+  void* operator new(__SIZE_TYPE__, int);
+};
+
+int main() {
+  // MSVC supports __noop with no arguments or (), so we do as well.
+  new (__noop) S;
+  new ((__noop)) S;
+}
diff --git a/clang/test/SemaObjCXX/property-placement-new.mm b/clang/test/SemaObjCXX/property-placement-new.mm
new file mode 100644
index 0000000000000..30d02428a4c33
--- /dev/null
+++ b/clang/test/SemaObjCXX/property-placement-new.mm
@@ -0,0 +1,28 @@
+// RUN: %clang_cc1 -std=c++11 -fsyntax-only -verify -Wno-objc-root-class %s
+// expected-no-diagnostics
+
+@interface I {
+  int position;
+}
+@property(nonatomic) int position;
+@end
+
+struct S {
+  void *operator new(__SIZE_TYPE__, int);
+};
+
+template 
+struct TS {
+  void *operator new(__SIZE_TYPE__, T);
+};
+
+I *GetI();
+
+int main() {
+  @autoreleasepool {
+    auto* i = GetI();
+    i.position = 42;
+    new (i.position) S;
+    new (i.position) TS;
+  }
+}
diff --git a/clang/test/SemaTemplate/GH71595.cpp b/clang/test/SemaTemplate/GH71595.cpp
index 7d34d1bf054e4..daec9410e547a 100644
--- a/clang/test/SemaTemplate/GH71595.cpp
+++ b/clang/test/SemaTemplate/GH71595.cpp
@@ -18,17 +18,17 @@ void f() {
 template
 class temp {
     template T>
-    friend void g();
+    friend void g(); // expected-error {{friend declaration with a constraint that depends on an enclosing template parameter must be a definition}}
 
-    temp(); // expected-note {{implicitly declared private here}}
+    temp();
 };
 
 template> T>
 void g() {
-    auto v = temp(); // expected-error {{calling a private constructor of class 'temp'}}
+    auto v = temp();
 }
 
 void h() {
     f();
-    g(); // expected-note {{in instantiation of function template specialization 'g' requested here}}
+    g();
 }
diff --git a/clang/test/SemaTemplate/GH75426.cpp b/clang/test/SemaTemplate/GH75426.cpp
new file mode 100644
index 0000000000000..faf70699f9c5f
--- /dev/null
+++ b/clang/test/SemaTemplate/GH75426.cpp
@@ -0,0 +1,16 @@
+// RUN: %clang_cc1 -std=c++20 -fsyntax-only -verify %s
+// expected-no-diagnostics
+
+template concept C = true;
+
+struct A {
+    template void f();
+};
+
+auto L = []{};
+
+template
+class Friends {
+    template friend void A::f();
+    template friend void decltype(L)::operator()();
+};
diff --git a/clang/test/SemaTemplate/concepts-friends.cpp b/clang/test/SemaTemplate/concepts-friends.cpp
index 5c4609520a3c7..255b0858917fb 100644
--- a/clang/test/SemaTemplate/concepts-friends.cpp
+++ b/clang/test/SemaTemplate/concepts-friends.cpp
@@ -2,6 +2,7 @@
 
 template 
 concept constraint = false;
+
 namespace temp_friend_9 {
 // A non-template friend declaration with a requires-clause shall be a
 // definition. ...Such a constrained friend function ... does not declare the
@@ -11,6 +12,14 @@ struct NonTemplateFriend {
   friend void foo()
     requires true
   {}
+
+  friend void baz() // expected-error {{non-template friend declaration with a requires clause must be a definition}}
+    requires true;
+};
+
+struct TempP9NotShownIfFunctionWouldBeInvalidAnyway {
+  friend void foo()
+    requires true; // expected-error {{non-templated function cannot have a requires clause}}
 };
 
 // A friend function template with a constraint that depends on a template
@@ -19,6 +28,10 @@ struct NonTemplateFriend {
 // function template as a declaration in any other scope.
 template 
 struct TemplateFromEnclosing {
+  template 
+  friend void bar2() // expected-error {{friend declaration with a constraint that depends on an enclosing template parameter must be a definition}}
+    requires constraint;
+
   template 
   friend void foo()
     requires constraint
diff --git a/clang/test/SemaTemplate/concepts.cpp b/clang/test/SemaTemplate/concepts.cpp
index e98ebcc9203a4..bac209a28da91 100644
--- a/clang/test/SemaTemplate/concepts.cpp
+++ b/clang/test/SemaTemplate/concepts.cpp
@@ -1064,3 +1064,24 @@ void cand(T t)
 
 void test() { cand(42); }
 }
+
+namespace GH63837 {
+
+template concept IsFoo = true;
+
+template struct Struct {
+  template
+  void foo() {}
+
+  template requires (... && IsFoo)
+  void bar() {}
+
+  template
+  static inline int field = 0;
+};
+
+template void Struct::foo<>();
+template void Struct::bar<>();
+template int Struct::field<1, 2>;
+
+}
diff --git a/clang/test/SemaTemplate/temp_arg_nontype_cxx1z.cpp b/clang/test/SemaTemplate/temp_arg_nontype_cxx1z.cpp
index 22788e1c1ffb3..b54b5a8007408 100644
--- a/clang/test/SemaTemplate/temp_arg_nontype_cxx1z.cpp
+++ b/clang/test/SemaTemplate/temp_arg_nontype_cxx1z.cpp
@@ -205,9 +205,9 @@ namespace Auto {
 
     struct Y : X {};
     void type_affects_identity(B<&X::n>) {}
-    void type_affects_identity(B<(int Y::*)&X::n>) {} // FIXME: expected-error {{sorry}}
+    void type_affects_identity(B<(int Y::*)&X::n>) {} // FIXME: expected-error {{non-type template argument of pointer-to-member type}}
     void type_affects_identity(B<(const int X::*)&X::n>) {}
-    void type_affects_identity(B<(const int Y::*)&X::n>) {} // FIXME: expected-error {{sorry}}
+    void type_affects_identity(B<(const int Y::*)&X::n>) {} // FIXME: expected-error {{non-type template argument of pointer-to-member type}}
 
     // A case where we need to do auto-deduction, and check whether the
     // resulting dependent types match during partial ordering. These
diff --git a/clang/test/SemaTemplate/temp_arg_nontype_cxx20.cpp b/clang/test/SemaTemplate/temp_arg_nontype_cxx20.cpp
index 982f6ec221570..e345a1f64b921 100644
--- a/clang/test/SemaTemplate/temp_arg_nontype_cxx20.cpp
+++ b/clang/test/SemaTemplate/temp_arg_nontype_cxx20.cpp
@@ -8,8 +8,8 @@ namespace std {
 
 // floating-point arguments
 template struct Float {};
-using F1 = Float<1.0f>; // FIXME expected-error {{sorry}}
-using F1 = Float<2.0f / 2>; // FIXME expected-error {{sorry}}
+using F1 = Float<1.0f>; // FIXME expected-error {{non-type template argument of type}}
+using F1 = Float<2.0f / 2>; // FIXME expected-error {{non-type template argument of type}}
 
 struct S { int n[3]; } s; // expected-note 1+{{here}}
 union U { int a, b; } u;
@@ -48,12 +48,12 @@ using U1 = Union; // expected-error {{different types}}
 
 // miscellaneous scalar types
 template<_Complex int> struct ComplexInt {};
-using CI = ComplexInt<1 + 3i>; // FIXME: expected-error {{sorry}}
-using CI = ComplexInt<1 + 3i>; // FIXME: expected-error {{sorry}}
+using CI = ComplexInt<1 + 3i>; // FIXME: expected-error {{non-type template argument of type}}
+using CI = ComplexInt<1 + 3i>; // FIXME: expected-error {{non-type template argument of type}}
 
 template<_Complex float> struct ComplexFloat {};
-using CF = ComplexFloat<1.0f + 3.0fi>; // FIXME: expected-error {{sorry}}
-using CF = ComplexFloat<1.0f + 3.0fi>; // FIXME: expected-error {{sorry}}
+using CF = ComplexFloat<1.0f + 3.0fi>; // FIXME: expected-error {{non-type template argument of type}}
+using CF = ComplexFloat<1.0f + 3.0fi>; // FIXME: expected-error {{non-type template argument of type}}
 
 namespace ClassNTTP {
   struct A { // expected-note 2{{candidate}}
diff --git a/clang/tools/c-index-test/c-index-test.c b/clang/tools/c-index-test/c-index-test.c
index 6fa400a0675b7..25235eb7f06d6 100644
--- a/clang/tools/c-index-test/c-index-test.c
+++ b/clang/tools/c-index-test/c-index-test.c
@@ -46,7 +46,14 @@ char *basename(const char* path)
     else if (base2)
         return(base2 + 1);
 
-    return((char*)path);
+#ifdef __clang__
+#pragma clang diagnostic push
+#pragma clang diagnostic ignored "-Wcast-qual"
+#endif
+    return ((char *)path);
+#ifdef __clang__
+#pragma clang diagnostic pop
+#endif
 }
 char *dirname(char* path)
 {
@@ -235,11 +242,16 @@ void free_remapped_files(struct CXUnsavedFile *unsaved_files,
 #ifdef __GNUC__
 #pragma GCC diagnostic push
 #pragma GCC diagnostic ignored "-Wcast-qual"
+#elif defined(__clang__)
+#pragma clang diagnostic push
+#pragma clang diagnostic ignored "-Wcast-qual"
 #endif
     free((char *)unsaved_files[i].Filename);
     free((char *)unsaved_files[i].Contents);
 #ifdef __GNUC__
 #pragma GCC diagnostic pop
+#elif defined(__clang__)
+#pragma clang diagnostic pop
 #endif
   }
   free(unsaved_files);
diff --git a/clang/tools/clang-linker-wrapper/CMakeLists.txt b/clang/tools/clang-linker-wrapper/CMakeLists.txt
index d619049b0eaab..077d617002e94 100644
--- a/clang/tools/clang-linker-wrapper/CMakeLists.txt
+++ b/clang/tools/clang-linker-wrapper/CMakeLists.txt
@@ -28,7 +28,6 @@ endif()
 
 add_clang_tool(clang-linker-wrapper
   ClangLinkerWrapper.cpp
-  OffloadWrapper.cpp
   SYCLOffloadWrapper.cpp
 
   DEPENDS
diff --git a/clang/tools/clang-linker-wrapper/ClangLinkerWrapper.cpp b/clang/tools/clang-linker-wrapper/ClangLinkerWrapper.cpp
index 4dc83d73a41fa..bd177d667bb0e 100644
--- a/clang/tools/clang-linker-wrapper/ClangLinkerWrapper.cpp
+++ b/clang/tools/clang-linker-wrapper/ClangLinkerWrapper.cpp
@@ -14,12 +14,13 @@
 //
 //===---------------------------------------------------------------------===//
 
-#include "OffloadWrapper.h"
 #include "SYCLOffloadWrapper.h"
 #include "clang/Basic/Version.h"
 #include "llvm/BinaryFormat/Magic.h"
 #include "llvm/Bitcode/BitcodeWriter.h"
 #include "llvm/CodeGen/CommandFlags.h"
+#include "llvm/Frontend/Offloading/OffloadWrapper.h"
+#include "llvm/Frontend/Offloading/Utility.h"
 #include "llvm/IR/Constants.h"
 #include "llvm/IR/DiagnosticPrinter.h"
 #include "llvm/IR/Module.h"
@@ -1499,15 +1500,21 @@ wrapDeviceImages(ArrayRef> Buffers,
 
   switch (Kind) {
   case OFK_OpenMP:
-    if (Error Err = wrapOpenMPBinaries(M, BuffersToWrap))
+    if (Error Err = offloading::wrapOpenMPBinaries(
+            M, BuffersToWrap,
+            offloading::getOffloadEntryArray(M, "omp_offloading_entries")))
       return std::move(Err);
     break;
   case OFK_Cuda:
-    if (Error Err = wrapCudaBinary(M, BuffersToWrap.front()))
+    if (Error Err = offloading::wrapCudaBinary(
+            M, BuffersToWrap.front(),
+            offloading::getOffloadEntryArray(M, "cuda_offloading_entries")))
       return std::move(Err);
     break;
   case OFK_HIP:
-    if (Error Err = wrapHIPBinary(M, BuffersToWrap.front()))
+    if (Error Err = offloading::wrapHIPBinary(
+            M, BuffersToWrap.front(),
+            offloading::getOffloadEntryArray(M, "hip_offloading_entries")))
       return std::move(Err);
     break;
   default:
diff --git a/clang/tools/clang-linker-wrapper/OffloadWrapper.h b/clang/tools/clang-linker-wrapper/OffloadWrapper.h
deleted file mode 100644
index 679333975b212..0000000000000
--- a/clang/tools/clang-linker-wrapper/OffloadWrapper.h
+++ /dev/null
@@ -1,28 +0,0 @@
-//===- OffloadWrapper.h --r-------------------------------------*- C++ -*-===//
-//
-// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
-// See https://llvm.org/LICENSE.txt for license information.
-// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
-//
-//===----------------------------------------------------------------------===//
-
-#ifndef LLVM_CLANG_TOOLS_CLANG_LINKER_WRAPPER_OFFLOAD_WRAPPER_H
-#define LLVM_CLANG_TOOLS_CLANG_LINKER_WRAPPER_OFFLOAD_WRAPPER_H
-
-#include "llvm/ADT/ArrayRef.h"
-#include "llvm/IR/Module.h"
-
-/// Wraps the input device images into the module \p M as global symbols and
-/// registers the images with the OpenMP Offloading runtime libomptarget.
-llvm::Error wrapOpenMPBinaries(llvm::Module &M,
-                               llvm::ArrayRef> Images);
-
-/// Wraps the input fatbinary image into the module \p M as global symbols and
-/// registers the images with the CUDA runtime.
-llvm::Error wrapCudaBinary(llvm::Module &M, llvm::ArrayRef Images);
-
-/// Wraps the input bundled image into the module \p M as global symbols and
-/// registers the images with the HIP runtime.
-llvm::Error wrapHIPBinary(llvm::Module &M, llvm::ArrayRef Images);
-
-#endif
diff --git a/clang/tools/clang-scan-deps/ClangScanDeps.cpp b/clang/tools/clang-scan-deps/ClangScanDeps.cpp
index 75aa4ae97c618..fb42dc1085afb 100644
--- a/clang/tools/clang-scan-deps/ClangScanDeps.cpp
+++ b/clang/tools/clang-scan-deps/ClangScanDeps.cpp
@@ -20,7 +20,6 @@
 #include "llvm/Support/CommandLine.h"
 #include "llvm/Support/FileUtilities.h"
 #include "llvm/Support/Format.h"
-#include "llvm/Support/InitLLVM.h"
 #include "llvm/Support/JSON.h"
 #include "llvm/Support/LLVMDriver.h"
 #include "llvm/Support/Program.h"
@@ -699,7 +698,6 @@ static std::string getModuleCachePath(ArrayRef Args) {
 // form specified command line after the positional parameter "--".
 static std::unique_ptr
 getCompilationDataBase(int argc, char **argv, std::string &ErrorMessage) {
-  llvm::InitLLVM X(argc, argv);
   ParseArgs(argc, argv);
 
   if (!CompilationDB.empty())
diff --git a/clang/tools/driver/driver.cpp b/clang/tools/driver/driver.cpp
index 4adc7f7ad0dac..1407c7fcdab76 100644
--- a/clang/tools/driver/driver.cpp
+++ b/clang/tools/driver/driver.cpp
@@ -36,7 +36,6 @@
 #include "llvm/Support/CrashRecoveryContext.h"
 #include "llvm/Support/ErrorHandling.h"
 #include "llvm/Support/FileSystem.h"
-#include "llvm/Support/InitLLVM.h"
 #include "llvm/Support/LLVMDriver.h"
 #include "llvm/Support/Path.h"
 #include "llvm/Support/PrettyStackTrace.h"
@@ -65,7 +64,7 @@ std::string GetExecutablePath(const char *Argv0, bool CanonicalPrefixes) {
       if (llvm::ErrorOr P =
               llvm::sys::findProgramByName(ExecutablePath))
         ExecutablePath = *P;
-    return std::string(ExecutablePath.str());
+    return std::string(ExecutablePath);
   }
 
   // This just needs to be some symbol in the binary; C++ doesn't
@@ -377,7 +376,6 @@ static int ExecuteCC1Tool(SmallVectorImpl &ArgV,
 
 int clang_main(int Argc, char **Argv, const llvm::ToolContext &ToolContext) {
   noteBottomOfStack();
-  llvm::InitLLVM X(Argc, Argv);
   llvm::setBugReportMsg("PLEASE submit a bug report to " BUG_REPORT_URL
                         " and include the crash backtrace, preprocessed "
                         "source, and associated run script.\n");
diff --git a/clang/tools/libclang/CIndex.cpp b/clang/tools/libclang/CIndex.cpp
index 20dcc8b587845..1d25d32e161c5 100644
--- a/clang/tools/libclang/CIndex.cpp
+++ b/clang/tools/libclang/CIndex.cpp
@@ -9012,7 +9012,7 @@ unsigned clang_CXXMethod_isPureVirtual(CXCursor C) {
   const Decl *D = cxcursor::getCursorDecl(C);
   const CXXMethodDecl *Method =
       D ? dyn_cast_or_null(D->getAsFunction()) : nullptr;
-  return (Method && Method->isVirtual() && Method->isPure()) ? 1 : 0;
+  return (Method && Method->isPureVirtual()) ? 1 : 0;
 }
 
 unsigned clang_CXXMethod_isConst(CXCursor C) {
diff --git a/clang/tools/libclang/linker-script-to-export-list.py b/clang/tools/libclang/linker-script-to-export-list.py
index 745996028d835..9c7b6a98a34b5 100644
--- a/clang/tools/libclang/linker-script-to-export-list.py
+++ b/clang/tools/libclang/linker-script-to-export-list.py
@@ -6,6 +6,6 @@
 output_file = open(sys.argv[2], "w")
 
 for line in input_file:
-    m = re.search("^\s+(clang_[^;]+)", line)
+    m = re.search(r"^\s+(clang_[^;]+)", line)
     if m:
         output_file.write(m.group(1) + "\n")
diff --git a/clang/unittests/AST/ASTImporterTest.cpp b/clang/unittests/AST/ASTImporterTest.cpp
index e4bd0d646cc9d..e3f3a1bef656b 100644
--- a/clang/unittests/AST/ASTImporterTest.cpp
+++ b/clang/unittests/AST/ASTImporterTest.cpp
@@ -816,22 +816,15 @@ TEST_P(ImportExpr, ImportSizeOfPackExpr) {
           hasUnqualifiedDesugaredType(constantArrayType(hasSize(7))))))))));
 }
 
-const internal::VariadicDynCastAllOfMatcher cxxFoldExpr;
-
-AST_MATCHER_P(CXXFoldExpr, hasOperator, BinaryOperatorKind, Op) {
-  return Node.getOperator() == Op;
-}
-AST_MATCHER(CXXFoldExpr, hasInit) { return Node.getInit(); }
-AST_MATCHER(CXXFoldExpr, isRightFold) { return Node.isRightFold(); }
-AST_MATCHER(CXXFoldExpr, isLeftFold) { return Node.isLeftFold(); }
-
 TEST_P(ImportExpr, ImportCXXFoldExpr) {
-  auto Match1 =
-      cxxFoldExpr(hasOperator(BO_Add), isLeftFold(), unless(hasInit()));
-  auto Match2 = cxxFoldExpr(hasOperator(BO_Sub), isLeftFold(), hasInit());
-  auto Match3 =
-      cxxFoldExpr(hasOperator(BO_Mul), isRightFold(), unless(hasInit()));
-  auto Match4 = cxxFoldExpr(hasOperator(BO_Div), isRightFold(), hasInit());
+  auto Match1 = cxxFoldExpr(hasOperatorName("+"), isLeftFold(),
+                            unless(hasFoldInit(expr())));
+  auto Match2 =
+      cxxFoldExpr(hasOperatorName("-"), isLeftFold(), hasFoldInit(expr()));
+  auto Match3 = cxxFoldExpr(hasOperatorName("*"), isRightFold(),
+                            unless(hasFoldInit(expr())));
+  auto Match4 =
+      cxxFoldExpr(hasOperatorName("/"), isRightFold(), hasFoldInit(expr()));
 
   MatchVerifier Verifier;
   testImport("template "
@@ -1717,7 +1710,7 @@ TEST_P(ASTImporterOptionSpecificTestBase,
       R"s(
       struct declToImport {
         int a = d;
-        union { 
+        union {
           int b;
           int c;
         };
@@ -4012,7 +4005,7 @@ TEST_P(ImportVariables, ImportBindingDecl) {
         int a[2] = {1,2};
         auto [x1,y1] = a;
         auto& [x2,y2] = a;
-        
+
         struct S {
           mutable int x1 : 2;
           volatile double y1;
@@ -4540,6 +4533,162 @@ TEST_P(ImportFriendClasses, ImportOfClassDefinitionAndFwdFriendShouldBeLinked) {
   EXPECT_EQ(ImportedFwd, ImportedDef->getPreviousDecl());
 }
 
+TEST_P(ImportFriendClasses,
+       ImportFriendTemplatesInDependentContext_DefToFriend) {
+  Decl *ToTU = getToTuDecl(
+      R"(
+      template
+      struct X {
+        template
+        friend struct Y;
+      };
+      )",
+      Lang_CXX03);
+  auto *ToYFriend = FirstDeclMatcher().match(
+      ToTU, classTemplateDecl(hasName("Y")));
+  Decl *FromTU = getTuDecl(
+      R"(
+      template
+      struct Y {};
+      )",
+      Lang_CXX03, "input0.cc");
+  auto *FromYDef = FirstDeclMatcher().match(
+      FromTU, classTemplateDecl(hasName("Y")));
+  auto *ImportedYDef = Import(FromYDef, Lang_CXX03);
+  EXPECT_TRUE(ImportedYDef);
+  EXPECT_FALSE(ImportedYDef->getPreviousDecl());
+  EXPECT_NE(ImportedYDef, ToYFriend);
+}
+
+TEST_P(ImportFriendClasses,
+       ImportFriendTemplatesInDependentContext_DefToFriend_NE) {
+  getToTuDecl(
+      R"(
+      template
+      struct X {
+        template
+        friend struct Y;
+      };
+      )",
+      Lang_CXX03);
+  Decl *FromTU = getTuDecl(
+      R"(
+      template
+      struct Y {};
+      )",
+      Lang_CXX03, "input0.cc");
+  auto *FromYDef = FirstDeclMatcher().match(
+      FromTU, classTemplateDecl(hasName("Y")));
+  auto *ImportedYDef = Import(FromYDef, Lang_CXX03);
+  EXPECT_FALSE(ImportedYDef);
+}
+
+TEST_P(ImportFriendClasses,
+       ImportFriendTemplatesInDependentContext_FriendToFriend) {
+  Decl *ToTU = getToTuDecl(
+      R"(
+      template
+      struct X {
+        template
+        friend struct Y;
+      };
+      )",
+      Lang_CXX03);
+  auto *ToYFriend = FirstDeclMatcher().match(
+      ToTU, classTemplateDecl(hasName("Y")));
+  Decl *FromTU = getTuDecl(
+      R"(
+      template
+      struct X {
+        template
+        friend struct Y;
+      };
+      )",
+      Lang_CXX03, "input0.cc");
+  auto *FromYFriend = FirstDeclMatcher().match(
+      FromTU, classTemplateDecl(hasName("Y")));
+  auto *ImportedYFriend = Import(FromYFriend, Lang_CXX03);
+  EXPECT_TRUE(ImportedYFriend);
+  EXPECT_FALSE(ImportedYFriend->getPreviousDecl());
+  EXPECT_NE(ImportedYFriend, ToYFriend);
+}
+
+TEST_P(ImportFriendClasses,
+       ImportFriendTemplatesInDependentContext_FriendToFriend_NE) {
+  getToTuDecl(
+      R"(
+      template
+      struct X {
+        template
+        friend struct Y;
+      };
+      )",
+      Lang_CXX03);
+  Decl *FromTU = getTuDecl(
+      R"(
+      template
+      struct X {
+        template
+        friend struct Y;
+      };
+      )",
+      Lang_CXX03, "input0.cc");
+  auto *FromYFriend = FirstDeclMatcher().match(
+      FromTU, classTemplateDecl(hasName("Y")));
+  auto *ImportedYFriend = Import(FromYFriend, Lang_CXX03);
+  EXPECT_FALSE(ImportedYFriend);
+}
+
+TEST_P(ImportFriendClasses,
+       ImportFriendTemplatesInDependentContext_FriendToDef) {
+  Decl *ToTU = getToTuDecl(
+      R"(
+      template
+      struct Y {};
+      )",
+      Lang_CXX03);
+  auto *ToYDef = FirstDeclMatcher().match(
+      ToTU, classTemplateDecl(hasName("Y")));
+  Decl *FromTU = getTuDecl(
+      R"(
+      template
+      struct X {
+        template
+        friend struct Y;
+      };
+      )",
+      Lang_CXX03, "input0.cc");
+  auto *FromYFriend = FirstDeclMatcher().match(
+      FromTU, classTemplateDecl(hasName("Y")));
+  auto *ImportedYFriend = Import(FromYFriend, Lang_CXX03);
+  EXPECT_TRUE(ImportedYFriend);
+  EXPECT_FALSE(ImportedYFriend->getPreviousDecl());
+  EXPECT_NE(ImportedYFriend, ToYDef);
+}
+
+TEST_P(ImportFriendClasses,
+       ImportFriendTemplatesInDependentContext_FriendToDef_NE) {
+  getToTuDecl(
+      R"(
+      template
+      struct Y {};
+      )",
+      Lang_CXX03);
+  Decl *FromTU = getTuDecl(
+      R"(
+      template
+      struct X {
+        template
+        friend struct Y;
+      };
+      )",
+      Lang_CXX03, "input0.cc");
+  auto *FromYFriend = FirstDeclMatcher().match(
+      FromTU, classTemplateDecl(hasName("Y")));
+  auto *ImportedYFriend = Import(FromYFriend, Lang_CXX03);
+  EXPECT_FALSE(ImportedYFriend);
+}
+
 TEST_P(ImportFriendClasses, ImportOfRepeatedFriendType) {
   const char *Code =
       R"(
@@ -5063,6 +5212,59 @@ TEST_P(ImportFriendClasses, RecordVarTemplateDecl) {
   EXPECT_EQ(ToTUX, ToX);
 }
 
+TEST_P(ASTImporterOptionSpecificTestBase, VarTemplateDeclConflict) {
+  getToTuDecl(
+      R"(
+      template 
+      constexpr int X = 1;
+      )",
+      Lang_CXX14);
+
+  Decl *FromTU = getTuDecl(
+      R"(
+      template 
+      constexpr int X = 2;
+      )",
+      Lang_CXX14, "input1.cc");
+  auto *FromX = FirstDeclMatcher().match(
+      FromTU, varTemplateDecl(hasName("X")));
+  auto *ToX = Import(FromX, Lang_CXX11);
+  // FIXME: This import should fail.
+  EXPECT_TRUE(ToX);
+}
+
+TEST_P(ASTImporterOptionSpecificTestBase, VarTemplateStaticDefinition) {
+  Decl *ToTU = getToTuDecl(
+      R"(
+      struct A {
+        template 
+        static int X;
+      };
+      )",
+      Lang_CXX14);
+  auto *ToX = FirstDeclMatcher().match(
+      ToTU, varTemplateDecl(hasName("X")));
+  ASSERT_FALSE(ToX->isThisDeclarationADefinition());
+
+  Decl *FromTU = getTuDecl(
+      R"(
+      struct A {
+        template 
+        static int X;
+      };
+      template 
+      int A::X = 2;
+      )",
+      Lang_CXX14, "input1.cc");
+  auto *FromXDef = LastDeclMatcher().match(
+      FromTU, varTemplateDecl(hasName("X")));
+  ASSERT_TRUE(FromXDef->isThisDeclarationADefinition());
+  auto *ToXDef = Import(FromXDef, Lang_CXX14);
+  EXPECT_TRUE(ToXDef);
+  EXPECT_TRUE(ToXDef->isThisDeclarationADefinition());
+  EXPECT_EQ(ToXDef->getPreviousDecl(), ToX);
+}
+
 TEST_P(ASTImporterOptionSpecificTestBase, VarTemplateParameterDeclContext) {
   constexpr auto Code =
       R"(
diff --git a/clang/unittests/AST/StructuralEquivalenceTest.cpp b/clang/unittests/AST/StructuralEquivalenceTest.cpp
index 22c7b82460f0a..91dd717d7b25e 100644
--- a/clang/unittests/AST/StructuralEquivalenceTest.cpp
+++ b/clang/unittests/AST/StructuralEquivalenceTest.cpp
@@ -2275,6 +2275,176 @@ TEST_F(StructuralEquivalenceStmtTest, UnaryOperatorDifferentOps) {
   EXPECT_FALSE(testStructuralMatch(t));
 }
 
+TEST_F(StructuralEquivalenceStmtTest,
+       CXXOperatorCallExprVsUnaryBinaryOperator) {
+  auto t = makeNamedDecls(
+      R"(
+      template 
+      class A;
+      template 
+      void foo(
+        A,
+        A,
+        A,
+        A,
+        A,
+        A,
+        A,
+        A,
+        A,
+        A,
+        A,
+        A,
+        A,
+        A,
+        A y)>,
+        A,
+        A> y)>,
+        A,
+        A,
+        A,
+        A= y>,
+        A y>,
+        A,
+        A,
+        A,
+        A,
+        A,
+        A* y>,
+        A y>
+      );
+      )",
+      R"(
+      struct Bar {
+        Bar& operator=(Bar&);
+        Bar& operator->();
+      };
+
+      Bar& operator+(Bar&, Bar&);
+      Bar& operator+(Bar&);
+      Bar& operator-(Bar&, Bar&);
+      Bar& operator-(Bar&);
+      Bar& operator*(Bar&, Bar&);
+      Bar& operator*(Bar&);
+      Bar& operator/(Bar&, Bar&);
+      Bar& operator%(Bar&, Bar&);
+      Bar& operator^(Bar&, Bar&);
+      Bar& operator&(Bar&, Bar&);
+      Bar& operator&(Bar&);
+      Bar& operator|(Bar&, Bar&);
+      Bar& operator~(Bar&);
+      Bar& operator!(Bar&);
+      Bar& operator<(Bar&, Bar&);
+      Bar& operator>(Bar&, Bar&);
+      Bar& operator+=(Bar&, Bar&);
+      Bar& operator-=(Bar&, Bar&);
+      Bar& operator*=(Bar&, Bar&);
+      Bar& operator/=(Bar&, Bar&);
+      Bar& operator%=(Bar&, Bar&);
+      Bar& operator^=(Bar&, Bar&);
+      Bar& operator&=(Bar&, Bar&);
+      Bar& operator|=(Bar&, Bar&);
+      Bar& operator<<(Bar&, Bar&);
+      Bar& operator>>(Bar&, Bar&);
+      Bar& operator<<=(Bar&, Bar&);
+      Bar& operator>>=(Bar&, Bar&);
+      Bar& operator==(Bar&, Bar&);
+      Bar& operator!=(Bar&, Bar&);
+      Bar& operator<=(Bar&, Bar&);
+      Bar& operator>=(Bar&, Bar&);
+      Bar& operator<=>(Bar&, Bar&);
+      Bar& operator&&(Bar&, Bar&);
+      Bar& operator||(Bar&, Bar&);
+      Bar& operator++(Bar&);
+      Bar& operator--(Bar&);
+      Bar& operator,(Bar&, Bar&);
+      Bar& operator->*(Bar&, Bar&);
+
+      template 
+      class A;
+      template 
+      void foo(
+        A,
+        A,
+        A,
+        A,
+        A,
+        A,
+        A,
+        A,
+        A,
+        A,
+        A,
+        A,
+        A,
+        A,
+        A y)>,
+        A,
+        A> y)>,
+        A,
+        A,
+        A,
+        A= y>,
+        A y>,
+        A,
+        A,
+        A,
+        A,
+        A,
+        A* y>,
+        A y>
+      );
+      )",
+      Lang_CXX20);
+  EXPECT_TRUE(testStructuralMatch(t));
+}
+
+TEST_F(StructuralEquivalenceStmtTest,
+       CXXOperatorCallExprVsUnaryBinaryOperatorNe) {
+  auto t = makeNamedDecls(
+      R"(
+      template 
+      class A;
+      template 
+      void foo(
+        A
+      );
+      )",
+      R"(
+      struct Bar;
+
+      Bar& operator-(Bar&, Bar&);
+
+      template 
+      class A;
+      template 
+      void foo(
+        A
+      );
+      )",
+      Lang_CXX11);
+  EXPECT_FALSE(testStructuralMatch(t));
+}
+
+TEST_F(StructuralEquivalenceStmtTest, NonTypeTemplateParm) {
+  auto t = makeNamedDecls(
+      R"(
+      template 
+      class A;
+      template 
+      void foo(A);
+      )",
+      R"(
+      template 
+      class A;
+      template 
+      void foo(A);
+      )",
+      Lang_CXX11);
+  // FIXME: These should not match,
+  EXPECT_TRUE(testStructuralMatch(t));
+}
+
 TEST_F(StructuralEquivalenceStmtTest, UnresolvedLookupDifferentName) {
   auto t = makeStmts(
       R"(
diff --git a/clang/unittests/ASTMatchers/ASTMatchersNarrowingTest.cpp b/clang/unittests/ASTMatchers/ASTMatchersNarrowingTest.cpp
index d78676fd289d5..edcdae4559d97 100644
--- a/clang/unittests/ASTMatchers/ASTMatchersNarrowingTest.cpp
+++ b/clang/unittests/ASTMatchers/ASTMatchersNarrowingTest.cpp
@@ -4103,15 +4103,102 @@ TEST_P(ASTMatchersTest, IsComparisonOperator) {
       notMatches("void x() { int a; if(a = 0) return; }", BinCompOperator));
 }
 
-TEST_P(ASTMatchersTest, HasInit) {
-  if (!GetParam().isCXX11OrLater()) {
-    // FIXME: Add a test for `hasInit()` that does not depend on C++.
+TEST_P(ASTMatchersTest, isRightFold) {
+  if (!GetParam().isCXX17OrLater()) {
+    return;
+  }
+
+  EXPECT_FALSE(matches("template  auto sum(Args... args) { "
+                       "return (0 + ... + args); }",
+                       cxxFoldExpr(isRightFold())));
+  EXPECT_TRUE(matches("template  auto sum(Args... args) { "
+                      "return (args + ... + 0); }",
+                      cxxFoldExpr(isRightFold())));
+  EXPECT_FALSE(matches("template  auto sum(Args... args) { "
+                       "return (... + args); };",
+                       cxxFoldExpr(isRightFold())));
+  EXPECT_TRUE(matches("template  auto sum(Args... args) { "
+                      "return (args + ...); };",
+                      cxxFoldExpr(isRightFold())));
+}
+
+TEST_P(ASTMatchersTest, isLeftFold) {
+  if (!GetParam().isCXX17OrLater()) {
+    return;
+  }
+
+  EXPECT_TRUE(matches("template  auto sum(Args... args) { "
+                      "return (0 + ... + args); }",
+                      cxxFoldExpr(isLeftFold())));
+  EXPECT_FALSE(matches("template  auto sum(Args... args) { "
+                       "return (args + ... + 0); }",
+                       cxxFoldExpr(isLeftFold())));
+  EXPECT_TRUE(matches("template  auto sum(Args... args) { "
+                      "return (... + args); };",
+                      cxxFoldExpr(isLeftFold())));
+  EXPECT_FALSE(matches("template  auto sum(Args... args) { "
+                       "return (args + ...); };",
+                       cxxFoldExpr(isLeftFold())));
+}
+
+TEST_P(ASTMatchersTest, isUnaryFold) {
+  if (!GetParam().isCXX17OrLater()) {
     return;
   }
 
-  EXPECT_TRUE(matches("int x{0};", initListExpr(hasInit(0, expr()))));
-  EXPECT_FALSE(matches("int x{0};", initListExpr(hasInit(1, expr()))));
-  EXPECT_FALSE(matches("int x;", initListExpr(hasInit(0, expr()))));
+  EXPECT_FALSE(matches("template  auto sum(Args... args) { "
+                       "return (0 + ... + args); }",
+                       cxxFoldExpr(isUnaryFold())));
+  EXPECT_FALSE(matches("template  auto sum(Args... args) { "
+                       "return (args + ... + 0); }",
+                       cxxFoldExpr(isUnaryFold())));
+  EXPECT_TRUE(matches("template  auto sum(Args... args) { "
+                      "return (... + args); };",
+                      cxxFoldExpr(isUnaryFold())));
+  EXPECT_TRUE(matches("template  auto sum(Args... args) { "
+                      "return (args + ...); };",
+                      cxxFoldExpr(isUnaryFold())));
+}
+
+TEST_P(ASTMatchersTest, isBinaryFold) {
+  if (!GetParam().isCXX17OrLater()) {
+    return;
+  }
+
+  EXPECT_TRUE(matches("template  auto sum(Args... args) { "
+                      "return (0 + ... + args); }",
+                      cxxFoldExpr(isBinaryFold())));
+  EXPECT_TRUE(matches("template  auto sum(Args... args) { "
+                      "return (args + ... + 0); }",
+                      cxxFoldExpr(isBinaryFold())));
+  EXPECT_FALSE(matches("template  auto sum(Args... args) { "
+                       "return (... + args); };",
+                       cxxFoldExpr(isBinaryFold())));
+  EXPECT_FALSE(matches("template  auto sum(Args... args) { "
+                       "return (args + ...); };",
+                       cxxFoldExpr(isBinaryFold())));
+}
+
+TEST_P(ASTMatchersTest, hasOperator) {
+  if (!GetParam().isCXX17OrLater()) {
+    return;
+  }
+
+  EXPECT_TRUE(matches("template  auto sum(Args... args) { "
+                      "return (0 + ... + args); }",
+                      cxxFoldExpr(hasOperatorName("+"))));
+  EXPECT_TRUE(matches("template  auto sum(Args... args) { "
+                      "return (... + args); };",
+                      cxxFoldExpr(hasOperatorName("+"))));
+
+  EXPECT_FALSE(
+      matches("template  auto multiply(Args... args) { "
+              "return (0 * ... * args); }",
+              cxxFoldExpr(hasOperatorName("+"))));
+  EXPECT_FALSE(
+      matches("template  auto multiply(Args... args) { "
+              "return (... * args); };",
+              cxxFoldExpr(hasOperatorName("+"))));
 }
 
 TEST_P(ASTMatchersTest, IsMain) {
diff --git a/clang/unittests/ASTMatchers/ASTMatchersNodeTest.cpp b/clang/unittests/ASTMatchers/ASTMatchersNodeTest.cpp
index 8f0dd5602307c..ae30c03126d76 100644
--- a/clang/unittests/ASTMatchers/ASTMatchersNodeTest.cpp
+++ b/clang/unittests/ASTMatchers/ASTMatchersNodeTest.cpp
@@ -471,6 +471,19 @@ TEST_P(ASTMatchersTest, CXXOperatorCallExpr) {
   EXPECT_TRUE(notMatches("int t = 5 << 2;", OpCall));
 }
 
+TEST_P(ASTMatchersTest, FoldExpr) {
+  if (!GetParam().isCXX() || !GetParam().isCXX17OrLater()) {
+    return;
+  }
+
+  EXPECT_TRUE(matches("template  auto sum(Args... args) { "
+                      "return (0 + ... + args); }",
+                      cxxFoldExpr()));
+  EXPECT_TRUE(matches("template  auto sum(Args... args) { "
+                      "return (args + ...); }",
+                      cxxFoldExpr()));
+}
+
 TEST_P(ASTMatchersTest, ThisPointerType) {
   if (!GetParam().isCXX()) {
     return;
diff --git a/clang/unittests/ASTMatchers/ASTMatchersTraversalTest.cpp b/clang/unittests/ASTMatchers/ASTMatchersTraversalTest.cpp
index d4a695b974bf0..6911d7600a718 100644
--- a/clang/unittests/ASTMatchers/ASTMatchersTraversalTest.cpp
+++ b/clang/unittests/ASTMatchers/ASTMatchersTraversalTest.cpp
@@ -658,27 +658,27 @@ void check_match_co_return() {
   co_return 1;
 }
 )cpp";
-  EXPECT_TRUE(matchesConditionally(CoReturnCode, 
-                                   coreturnStmt(isExpansionInMainFile()), 
-                                   true, {"-std=c++20", "-I/"}, M));
+  EXPECT_TRUE(matchesConditionally(CoReturnCode,
+                                   coreturnStmt(isExpansionInMainFile()), true,
+                                   {"-std=c++20", "-I/"}, M));
   StringRef CoAwaitCode = R"cpp(
 #include 
 void check_match_co_await() {
   co_await a;
 }
 )cpp";
-  EXPECT_TRUE(matchesConditionally(CoAwaitCode, 
-                                   coawaitExpr(isExpansionInMainFile()), 
-                                   true, {"-std=c++20", "-I/"}, M));
+  EXPECT_TRUE(matchesConditionally(CoAwaitCode,
+                                   coawaitExpr(isExpansionInMainFile()), true,
+                                   {"-std=c++20", "-I/"}, M));
   StringRef CoYieldCode = R"cpp(
 #include 
 void check_match_co_yield() {
   co_yield 1.0;
 }
 )cpp";
-  EXPECT_TRUE(matchesConditionally(CoYieldCode, 
-                                   coyieldExpr(isExpansionInMainFile()), 
-                                   true, {"-std=c++20", "-I/"}, M));
+  EXPECT_TRUE(matchesConditionally(CoYieldCode,
+                                   coyieldExpr(isExpansionInMainFile()), true,
+                                   {"-std=c++20", "-I/"}, M));
 
   StringRef NonCoroCode = R"cpp(
 #include 
@@ -2000,6 +2000,146 @@ TEST(Matcher, UnaryOperatorTypes) {
       "void x() { A a; !a; }", unaryOperator(hasOperatorName("!"))));
 }
 
+TEST_P(ASTMatchersTest, HasInit) {
+  if (!GetParam().isCXX11OrLater()) {
+    // FIXME: Add a test for `hasInit()` that does not depend on C++.
+    return;
+  }
+
+  EXPECT_TRUE(matches("int x{0};", initListExpr(hasInit(0, expr()))));
+  EXPECT_FALSE(matches("int x{0};", initListExpr(hasInit(1, expr()))));
+  EXPECT_FALSE(matches("int x;", initListExpr(hasInit(0, expr()))));
+}
+
+TEST_P(ASTMatchersTest, HasFoldInit) {
+  if (!GetParam().isCXX17OrLater()) {
+    return;
+  }
+
+  EXPECT_TRUE(matches("template  auto sum(Args... args) { "
+                      "return (0 + ... + args); }",
+                      cxxFoldExpr(hasFoldInit(expr()))));
+  EXPECT_TRUE(matches("template  auto sum(Args... args) { "
+                      "return (args + ... + 0); }",
+                      cxxFoldExpr(hasFoldInit(expr()))));
+  EXPECT_FALSE(matches("template  auto sum(Args... args) { "
+                       "return (... + args); };",
+                       cxxFoldExpr(hasFoldInit(expr()))));
+}
+
+TEST_P(ASTMatchersTest, HasPattern) {
+  if (!GetParam().isCXX17OrLater()) {
+    return;
+  }
+
+  EXPECT_TRUE(matches("template  auto sum(Args... args) { "
+                      "return (0 + ... + args); }",
+                      cxxFoldExpr(hasPattern(expr()))));
+  EXPECT_TRUE(matches("template  auto sum(Args... args) { "
+                      "return (args + ... + 0); }",
+                      cxxFoldExpr(hasPattern(expr()))));
+  EXPECT_TRUE(matches("template  auto sum(Args... args) { "
+                      "return (... + args); };",
+                      cxxFoldExpr(hasPattern(expr()))));
+}
+
+TEST_P(ASTMatchersTest, HasLHSAndHasRHS) {
+  if (!GetParam().isCXX17OrLater()) {
+    return;
+  }
+
+  EXPECT_TRUE(matches("template  auto sum(Args... args) { "
+                      "return (0 + ... + args); }",
+                      cxxFoldExpr(hasLHS(expr()))));
+  EXPECT_TRUE(matches("template  auto sum(Args... args) { "
+                      "return (args + ... + 0); }",
+                      cxxFoldExpr(hasLHS(expr()))));
+  EXPECT_FALSE(matches("template  auto sum(Args... args) { "
+                       "return (... + args); };",
+                       cxxFoldExpr(hasLHS(expr()))));
+  EXPECT_TRUE(matches("template  auto sum(Args... args) { "
+                      "return (args + ...); };",
+                      cxxFoldExpr(hasLHS(expr()))));
+
+  EXPECT_TRUE(matches("template  auto sum(Args... args) { "
+                      "return (0 + ... + args); }",
+                      cxxFoldExpr(hasRHS(expr()))));
+  EXPECT_TRUE(matches("template  auto sum(Args... args) { "
+                      "return (args + ... + 0); }",
+                      cxxFoldExpr(hasRHS(expr()))));
+  EXPECT_TRUE(matches("template  auto sum(Args... args) { "
+                      "return (... + args); };",
+                      cxxFoldExpr(hasRHS(expr()))));
+  EXPECT_FALSE(matches("template  auto sum(Args... args) { "
+                       "return (args + ...); };",
+                       cxxFoldExpr(hasRHS(expr()))));
+}
+
+TEST_P(ASTMatchersTest, HasEitherOperandAndHasOperands) {
+  if (!GetParam().isCXX17OrLater()) {
+    return;
+  }
+
+  EXPECT_TRUE(matches("template  auto sum(Args... args) { "
+                      "return (0 + ... + args); }",
+                      cxxFoldExpr(hasEitherOperand(integerLiteral()))));
+  EXPECT_TRUE(matches("template  auto sum(Args... args) { "
+                      "return (args + ... + 0); }",
+                      cxxFoldExpr(hasEitherOperand(integerLiteral()))));
+
+  EXPECT_TRUE(matches("template  auto sum(Args... args) { "
+                      "return (0 + ... + args); }",
+                      cxxFoldExpr(hasEitherOperand(
+                          declRefExpr(to(namedDecl(hasName("args"))))))));
+  EXPECT_TRUE(matches("template  auto sum(Args... args) { "
+                      "return (args + ... + 0); }",
+                      cxxFoldExpr(hasEitherOperand(
+                          declRefExpr(to(namedDecl(hasName("args"))))))));
+  EXPECT_TRUE(matches("template  auto sum(Args... args) { "
+                      "return (... + args); };",
+                      cxxFoldExpr(hasEitherOperand(
+                          declRefExpr(to(namedDecl(hasName("args"))))))));
+  EXPECT_TRUE(matches("template  auto sum(Args... args) { "
+                      "return (args + ...); };",
+                      cxxFoldExpr(hasEitherOperand(
+                          declRefExpr(to(namedDecl(hasName("args"))))))));
+
+  EXPECT_TRUE(matches(
+      "template  auto sum(Args... args) { "
+      "return (0 + ... + args); }",
+      cxxFoldExpr(hasOperands(declRefExpr(to(namedDecl(hasName("args")))),
+                              integerLiteral()))));
+  EXPECT_TRUE(matches(
+      "template  auto sum(Args... args) { "
+      "return (args + ... + 0); }",
+      cxxFoldExpr(hasOperands(declRefExpr(to(namedDecl(hasName("args")))),
+                              integerLiteral()))));
+  EXPECT_FALSE(matches(
+      "template  auto sum(Args... args) { "
+      "return (... + args); };",
+      cxxFoldExpr(hasOperands(declRefExpr(to(namedDecl(hasName("args")))),
+                              integerLiteral()))));
+  EXPECT_FALSE(matches(
+      "template  auto sum(Args... args) { "
+      "return (args + ...); };",
+      cxxFoldExpr(hasOperands(declRefExpr(to(namedDecl(hasName("args")))),
+                              integerLiteral()))));
+}
+
+TEST_P(ASTMatchersTest, Callee) {
+  if (!GetParam().isCXX17OrLater()) {
+    return;
+  }
+
+  EXPECT_TRUE(matches(
+      "struct Dummy {}; Dummy operator+(Dummy, Dummy); template "
+      " auto sum(Args... args) { return (0 + ... + args); }",
+      cxxFoldExpr(callee(expr()))));
+  EXPECT_FALSE(matches("template  auto sum(Args... args) { "
+                       "return (0 + ... + args); }",
+                       cxxFoldExpr(callee(expr()))));
+}
+
 TEST(ArraySubscriptMatchers, ArrayIndex) {
   EXPECT_TRUE(matches(
     "int i[2]; void f() { i[1] = 1; }",
diff --git a/clang/unittests/Analysis/ExprMutationAnalyzerTest.cpp b/clang/unittests/Analysis/ExprMutationAnalyzerTest.cpp
index f1a1f857a0ee5..a94f857720b03 100644
--- a/clang/unittests/Analysis/ExprMutationAnalyzerTest.cpp
+++ b/clang/unittests/Analysis/ExprMutationAnalyzerTest.cpp
@@ -592,6 +592,26 @@ TEST(ExprMutationAnalyzerTest, ByConstRRefArgument) {
               ElementsAre("static_cast(x)"));
 }
 
+// Section: bindings.
+
+TEST(ExprMutationAnalyzerTest, BindingModifies) {
+  const auto AST =
+      buildASTFromCode("struct Point { int x; int y; };"
+                       "void mod(int&);"
+                       "void f(Point p) { auto& [x, y] = p; mod(x); }");
+  const auto Results =
+      match(withEnclosingCompound(declRefTo("p")), AST->getASTContext());
+  EXPECT_THAT(mutatedBy(Results, AST.get()), ElementsAre("x", "mod(x)"));
+}
+
+TEST(ExprMutationAnalyzerTest, BindingDoesNotModify) {
+  const auto AST = buildASTFromCode("struct Point { int x; int y; };"
+                                    "void f(Point p) { auto& [x, y] = p; x; }");
+  const auto Results =
+      match(withEnclosingCompound(declRefTo("p")), AST->getASTContext());
+  EXPECT_FALSE(isMutated(Results, AST.get()));
+}
+
 // section: explicit std::move and std::forward testing
 
 TEST(ExprMutationAnalyzerTest, Move) {
diff --git a/clang/unittests/Analysis/FlowSensitive/DataflowAnalysisContextTest.cpp b/clang/unittests/Analysis/FlowSensitive/DataflowAnalysisContextTest.cpp
index 88eb11045b9e9..4f7a72c502ccf 100644
--- a/clang/unittests/Analysis/FlowSensitive/DataflowAnalysisContextTest.cpp
+++ b/clang/unittests/Analysis/FlowSensitive/DataflowAnalysisContextTest.cpp
@@ -33,10 +33,34 @@ TEST_F(DataflowAnalysisContextTest, DistinctTopsNotEquivalent) {
   EXPECT_FALSE(Context.equivalentFormulas(X.formula(), Y.formula()));
 }
 
-TEST_F(DataflowAnalysisContextTest, EmptyFlowCondition) {
+TEST_F(DataflowAnalysisContextTest, TautologicalFlowConditionImplies) {
   Atom FC = A.makeFlowConditionToken();
-  auto &C = A.makeAtomRef(A.makeAtom());
-  EXPECT_FALSE(Context.flowConditionImplies(FC, C));
+  EXPECT_TRUE(Context.flowConditionImplies(FC, A.makeLiteral(true)));
+  EXPECT_FALSE(Context.flowConditionImplies(FC, A.makeLiteral(false)));
+  EXPECT_FALSE(Context.flowConditionImplies(FC, A.makeAtomRef(A.makeAtom())));
+}
+
+TEST_F(DataflowAnalysisContextTest, TautologicalFlowConditionAllows) {
+  Atom FC = A.makeFlowConditionToken();
+  EXPECT_TRUE(Context.flowConditionAllows(FC, A.makeLiteral(true)));
+  EXPECT_FALSE(Context.flowConditionAllows(FC, A.makeLiteral(false)));
+  EXPECT_TRUE(Context.flowConditionAllows(FC, A.makeAtomRef(A.makeAtom())));
+}
+
+TEST_F(DataflowAnalysisContextTest, ContradictoryFlowConditionImpliesAnything) {
+  Atom FC = A.makeFlowConditionToken();
+  Context.addFlowConditionConstraint(FC, A.makeLiteral(false));
+  EXPECT_TRUE(Context.flowConditionImplies(FC, A.makeLiteral(true)));
+  EXPECT_TRUE(Context.flowConditionImplies(FC, A.makeLiteral(false)));
+  EXPECT_TRUE(Context.flowConditionImplies(FC, A.makeAtomRef(A.makeAtom())));
+}
+
+TEST_F(DataflowAnalysisContextTest, ContradictoryFlowConditionAllowsNothing) {
+  Atom FC = A.makeFlowConditionToken();
+  Context.addFlowConditionConstraint(FC, A.makeLiteral(false));
+  EXPECT_FALSE(Context.flowConditionAllows(FC, A.makeLiteral(true)));
+  EXPECT_FALSE(Context.flowConditionAllows(FC, A.makeLiteral(false)));
+  EXPECT_FALSE(Context.flowConditionAllows(FC, A.makeAtomRef(A.makeAtom())));
 }
 
 TEST_F(DataflowAnalysisContextTest, AddFlowConditionConstraint) {
diff --git a/clang/unittests/Analysis/FlowSensitive/DataflowEnvironmentTest.cpp b/clang/unittests/Analysis/FlowSensitive/DataflowEnvironmentTest.cpp
index 003434a58b107..8799d03dfd3c5 100644
--- a/clang/unittests/Analysis/FlowSensitive/DataflowEnvironmentTest.cpp
+++ b/clang/unittests/Analysis/FlowSensitive/DataflowEnvironmentTest.cpp
@@ -58,6 +58,52 @@ TEST_F(EnvironmentTest, FlowCondition) {
   EXPECT_FALSE(Env.allows(NotX));
 }
 
+TEST_F(EnvironmentTest, SetAndGetValueOnCfgOmittedNodes) {
+  // Check that we can set a value on an expression that is omitted from the CFG
+  // (see `ignoreCFGOmittedNodes()`), then retrieve that same value from the
+  // expression. This is a regression test; `setValue()` and `getValue()`
+  // previously did not use `ignoreCFGOmittedNodes()` consistently.
+
+  using namespace ast_matchers;
+
+  std::string Code = R"cc(
+    struct S {
+      int f();
+    };
+    void target() {
+      // Method call on a temporary produces an `ExprWithCleanups`.
+      S().f();
+      (1);
+    }
+  )cc";
+
+  auto Unit =
+      tooling::buildASTFromCodeWithArgs(Code, {"-fsyntax-only", "-std=c++17"});
+  auto &Context = Unit->getASTContext();
+
+  ASSERT_EQ(Context.getDiagnostics().getClient()->getNumErrors(), 0U);
+
+  const ExprWithCleanups *WithCleanups = selectFirst(
+      "cleanups",
+      match(exprWithCleanups(hasType(isInteger())).bind("cleanups"), Context));
+  ASSERT_NE(WithCleanups, nullptr);
+
+  const ParenExpr *Paren = selectFirst(
+      "paren", match(parenExpr(hasType(isInteger())).bind("paren"), Context));
+  ASSERT_NE(Paren, nullptr);
+
+  Environment Env(DAContext);
+  IntegerValue *Val1 =
+      cast(Env.createValue(Unit->getASTContext().IntTy));
+  Env.setValue(*WithCleanups, *Val1);
+  EXPECT_EQ(Env.getValue(*WithCleanups), Val1);
+
+  IntegerValue *Val2 =
+      cast(Env.createValue(Unit->getASTContext().IntTy));
+  Env.setValue(*Paren, *Val2);
+  EXPECT_EQ(Env.getValue(*Paren), Val2);
+}
+
 TEST_F(EnvironmentTest, CreateValueRecursiveType) {
   using namespace ast_matchers;
 
diff --git a/clang/unittests/Analysis/FlowSensitive/TransferTest.cpp b/clang/unittests/Analysis/FlowSensitive/TransferTest.cpp
index 056c4f3383d83..d0a0e6d3f5836 100644
--- a/clang/unittests/Analysis/FlowSensitive/TransferTest.cpp
+++ b/clang/unittests/Analysis/FlowSensitive/TransferTest.cpp
@@ -2684,6 +2684,48 @@ TEST(TransferTest, ResultObjectLocation) {
       });
 }
 
+TEST(TransferTest, ResultObjectLocationForDefaultInitExpr) {
+  std::string Code = R"(
+    struct S {};
+    struct target {
+      target () {
+        (void)0;
+        // [[p]]
+      }
+      S s = {};
+    };
+  )";
+
+  using ast_matchers::cxxCtorInitializer;
+  using ast_matchers::match;
+  using ast_matchers::selectFirst;
+  runDataflow(
+      Code,
+      [](const llvm::StringMap> &Results,
+         ASTContext &ASTCtx) {
+        const Environment &Env = getEnvironmentAtAnnotation(Results, "p");
+
+        const ValueDecl *SField = findValueDecl(ASTCtx, "s");
+
+        auto *CtorInit = selectFirst(
+            "ctor_initializer",
+            match(cxxCtorInitializer().bind("ctor_initializer"), ASTCtx));
+        ASSERT_NE(CtorInit, nullptr);
+
+        auto *DefaultInit = cast(CtorInit->getInit());
+
+        RecordStorageLocation &Loc = Env.getResultObjectLocation(*DefaultInit);
+
+        // FIXME: The result object location for the `CXXDefaultInitExpr` should
+        // be the location of the member variable being initialized, but we
+        // don't do this correctly yet; see also comments in
+        // `builtinTransferInitializer()`.
+        // For the time being, we just document the current erroneous behavior
+        // here (this should be `EXPECT_EQ` when the behavior is fixed).
+        EXPECT_NE(&Loc, Env.getThisPointeeStorageLocation()->getChild(*SField));
+      });
+}
+
 TEST(TransferTest, StaticCast) {
   std::string Code = R"(
     void target(int Foo) {
diff --git a/clang/unittests/Analysis/FlowSensitive/ValueTest.cpp b/clang/unittests/Analysis/FlowSensitive/ValueTest.cpp
index c5d18ba74c3ed..b07328d2a5621 100644
--- a/clang/unittests/Analysis/FlowSensitive/ValueTest.cpp
+++ b/clang/unittests/Analysis/FlowSensitive/ValueTest.cpp
@@ -45,7 +45,20 @@ TEST(ValueTest, TopsEquivalent) {
   EXPECT_TRUE(areEquivalentValues(V2, V1));
 }
 
-TEST(ValueTest, EquivalentValuesWithDifferentPropsEquivalent) {
+// The framework does not (currently) consider equivalence for values with
+// properties, leaving such to individual analyses.
+TEST(ValueTest, ValuesWithSamePropsDifferent) {
+  Arena A;
+  TopBoolValue Prop(A.makeAtomRef(Atom(0)));
+  TopBoolValue V1(A.makeAtomRef(Atom(2)));
+  TopBoolValue V2(A.makeAtomRef(Atom(3)));
+  V1.setProperty("foo", Prop);
+  V2.setProperty("foo", Prop);
+  EXPECT_FALSE(areEquivalentValues(V1, V2));
+  EXPECT_FALSE(areEquivalentValues(V2, V1));
+}
+
+TEST(ValueTest, ValuesWithDifferentPropsDifferent) {
   Arena A;
   TopBoolValue Prop1(A.makeAtomRef(Atom(0)));
   TopBoolValue Prop2(A.makeAtomRef(Atom(1)));
@@ -53,8 +66,19 @@ TEST(ValueTest, EquivalentValuesWithDifferentPropsEquivalent) {
   TopBoolValue V2(A.makeAtomRef(Atom(3)));
   V1.setProperty("foo", Prop1);
   V2.setProperty("bar", Prop2);
-  EXPECT_TRUE(areEquivalentValues(V1, V2));
-  EXPECT_TRUE(areEquivalentValues(V2, V1));
+  EXPECT_FALSE(areEquivalentValues(V1, V2));
+  EXPECT_FALSE(areEquivalentValues(V2, V1));
+}
+
+TEST(ValueTest, ValuesWithDifferentNumberPropsDifferent) {
+  Arena A;
+  TopBoolValue Prop(A.makeAtomRef(Atom(0)));
+  TopBoolValue V1(A.makeAtomRef(Atom(2)));
+  TopBoolValue V2(A.makeAtomRef(Atom(3)));
+  // Only set a property on `V1`.
+  V1.setProperty("foo", Prop);
+  EXPECT_FALSE(areEquivalentValues(V1, V2));
+  EXPECT_FALSE(areEquivalentValues(V2, V1));
 }
 
 TEST(ValueTest, DifferentKindsNotEquivalent) {
diff --git a/clang/unittests/Format/ConfigParseTest.cpp b/clang/unittests/Format/ConfigParseTest.cpp
index 18ecba270e345..172aaab5988ce 100644
--- a/clang/unittests/Format/ConfigParseTest.cpp
+++ b/clang/unittests/Format/ConfigParseTest.cpp
@@ -255,6 +255,8 @@ TEST(ConfigParseTest, ParsesConfiguration) {
               PenaltyBreakTemplateDeclaration, 1234u);
   CHECK_PARSE("PenaltyBreakOpenParenthesis: 1234", PenaltyBreakOpenParenthesis,
               1234u);
+  CHECK_PARSE("PenaltyBreakScopeResolution: 1234", PenaltyBreakScopeResolution,
+              1234u);
   CHECK_PARSE("PenaltyExcessCharacter: 1234", PenaltyExcessCharacter, 1234u);
   CHECK_PARSE("PenaltyReturnTypeOnItsOwnLine: 1234",
               PenaltyReturnTypeOnItsOwnLine, 1234u);
@@ -994,6 +996,14 @@ TEST(ConfigParseTest, ParsesConfiguration) {
               FormatStyle::BBNSS_OnlyWithParen);
   CHECK_PARSE("AllowBreakBeforeNoexceptSpecifier: Never",
               AllowBreakBeforeNoexceptSpecifier, FormatStyle::BBNSS_Never);
+
+  Style.SeparateDefinitionBlocks = FormatStyle::SDS_Never;
+  CHECK_PARSE("SeparateDefinitionBlocks: Always", SeparateDefinitionBlocks,
+              FormatStyle::SDS_Always);
+  CHECK_PARSE("SeparateDefinitionBlocks: Leave", SeparateDefinitionBlocks,
+              FormatStyle::SDS_Leave);
+  CHECK_PARSE("SeparateDefinitionBlocks: Never", SeparateDefinitionBlocks,
+              FormatStyle::SDS_Never);
 }
 
 TEST(ConfigParseTest, ParsesConfigurationWithLanguages) {
diff --git a/clang/unittests/Format/FormatTest.cpp b/clang/unittests/Format/FormatTest.cpp
index 8f115fb8cbf0f..c229d9bc56def 100644
--- a/clang/unittests/Format/FormatTest.cpp
+++ b/clang/unittests/Format/FormatTest.cpp
@@ -7,7 +7,6 @@
 //===----------------------------------------------------------------------===//
 
 #include "FormatTestBase.h"
-#include "gmock/gmock.h"
 
 #define DEBUG_TYPE "format-test"
 
@@ -8563,9 +8562,6 @@ TEST_F(FormatTest, BreaksFunctionDeclarationsWithTrailingTokens) {
                "    __attribute__((unused));");
 
   Style = getGoogleStyle();
-  ASSERT_THAT(Style.AttributeMacros,
-              testing::AllOf(testing::Contains("GUARDED_BY"),
-                             testing::Contains("ABSL_GUARDED_BY")));
 
   verifyFormat(
       "bool aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa\n"
@@ -10158,11 +10154,11 @@ TEST_F(FormatTest, ReturnTypeBreakingStyle) {
                getGoogleStyleWithColumns(40));
   verifyFormat("Tttttttttttttttttttttttt ppppppppppppppp\n"
                "    ABSL_GUARDED_BY(mutex1)\n"
-               "    ABSL_GUARDED_BY(mutex2);",
+               "        ABSL_GUARDED_BY(mutex2);",
                getGoogleStyleWithColumns(40));
   verifyFormat("Tttttt f(int a, int b)\n"
                "    ABSL_GUARDED_BY(mutex1)\n"
-               "    ABSL_GUARDED_BY(mutex2);",
+               "        ABSL_GUARDED_BY(mutex2);",
                getGoogleStyleWithColumns(40));
   // * typedefs
   verifyGoogleFormat("typedef ATTR(X) char x;");
@@ -20931,6 +20927,11 @@ TEST_F(FormatTest, CatchAlignArrayOfStructuresRightAlignment) {
       "};",
       Style);
 
+  verifyNoCrash("Foo f[] = {\n"
+                "    [0] = { 1, },\n"
+                "    [i] { 1, },\n"
+                "};",
+                Style);
   verifyNoCrash("Foo foo[] = {\n"
                 "    [0] = {1, 1},\n"
                 "    [1] { 1, 1, },\n"
@@ -21179,6 +21180,11 @@ TEST_F(FormatTest, CatchAlignArrayOfStructuresLeftAlignment) {
       "};",
       Style);
 
+  verifyNoCrash("Foo f[] = {\n"
+                "    [0] = { 1, },\n"
+                "    [i] { 1, },\n"
+                "};",
+                Style);
   verifyNoCrash("Foo foo[] = {\n"
                 "    [0] = {1, 1},\n"
                 "    [1] { 1, 1, },\n"
@@ -21557,6 +21563,19 @@ TEST_F(FormatTest, BreakPenaltyAfterForLoopLParen) {
                Style);
 }
 
+TEST_F(FormatTest, BreakPenaltyScopeResolution) {
+  FormatStyle Style = getLLVMStyle();
+  Style.ColumnLimit = 20;
+  Style.PenaltyExcessCharacter = 100;
+  verifyFormat("unsigned long\n"
+               "foo::bar();",
+               Style);
+  Style.PenaltyBreakScopeResolution = 10;
+  verifyFormat("unsigned long foo::\n"
+               "    bar();",
+               Style);
+}
+
 TEST_F(FormatTest, WorksFor8bitEncodings) {
   // FIXME: unstable test case
   EXPECT_EQ("\"\xce\xe4\xed\xe0\xe6\xe4\xfb \xe2 \"\n"
diff --git a/clang/unittests/Format/FormatTestCSharp.cpp b/clang/unittests/Format/FormatTestCSharp.cpp
index 4a0840d32341e..6f5e1e41ef7e0 100644
--- a/clang/unittests/Format/FormatTestCSharp.cpp
+++ b/clang/unittests/Format/FormatTestCSharp.cpp
@@ -1304,6 +1304,18 @@ TEST_F(FormatTestCSharp, CSharpGenericTypeConstraints) {
                "}",
                Style);
 
+  // When the "where" line is not to be formatted, following lines should not
+  // take on its indentation.
+  verifyFormat("class ItemFactory\n"
+               "    where T : new() {\n"
+               "  int f() {}\n"
+               "}",
+               "class ItemFactory\n"
+               "    where T : new() {\n"
+               "  int f() {}\n"
+               "}",
+               Style, {tooling::Range(43, 13)});
+
   verifyFormat("class Dictionary\n"
                "    where TKey : IComparable\n"
                "    where TVal : IMyInterface {\n"
diff --git a/clang/unittests/Format/FormatTestJava.cpp b/clang/unittests/Format/FormatTestJava.cpp
index 202d603d05779..6da5f4fa25433 100644
--- a/clang/unittests/Format/FormatTestJava.cpp
+++ b/clang/unittests/Format/FormatTestJava.cpp
@@ -603,6 +603,21 @@ TEST_F(FormatTestJava, ShortFunctions) {
                Style);
 }
 
+TEST_F(FormatTestJava, ConfigurableSpacesInSquareBrackets) {
+  FormatStyle Spaces = getLLVMStyle(FormatStyle::LK_Java);
+
+  verifyFormat("Object[] arguments", Spaces);
+  verifyFormat("final Class[] types = new Class[numElements];", Spaces);
+  verifyFormat("types[i] = arguments[i].getClass();", Spaces);
+
+  Spaces.SpacesInSquareBrackets = true;
+
+  verifyFormat("Object[ ] arguments", Spaces);
+  verifyFormat("final Class[ ] types = new Class[ numElements ];",
+               Spaces);
+  verifyFormat("types[ i ] = arguments[ i ].getClass();", Spaces);
+}
+
 } // namespace
 } // namespace test
 } // namespace format
diff --git a/clang/unittests/Format/FormatTestMacroExpansion.cpp b/clang/unittests/Format/FormatTestMacroExpansion.cpp
index 68250234f4201..653ec2a94c64d 100644
--- a/clang/unittests/Format/FormatTestMacroExpansion.cpp
+++ b/clang/unittests/Format/FormatTestMacroExpansion.cpp
@@ -255,6 +255,38 @@ TEST_F(FormatTestMacroExpansion,
                          Style);
 }
 
+TEST_F(FormatTestMacroExpansion, CommaAsOperator) {
+  FormatStyle Style = getGoogleStyleWithColumns(42);
+  Style.Macros.push_back("MACRO(a, b, c)=a=(b); if(x) c");
+  verifyFormat("MACRO(auto a,\n"
+               "      looooongfunction(first, second,\n"
+               "                       third),\n"
+               "      fourth);",
+               Style);
+}
+
+TEST_F(FormatTestMacroExpansion, ForcedBreakDiffers) {
+  FormatStyle Style = getGoogleStyleWithColumns(40);
+  Style.Macros.push_back("MACRO(a, b)=a=(b)");
+  verifyFormat("//\n"
+               "MACRO(const type variable,\n"
+               "      functtioncall(\n"
+               "          first, longsecondarg, third));",
+               Style);
+}
+
+TEST_F(FormatTestMacroExpansion,
+       PreferNotBreakingBetweenReturnTypeAndFunction) {
+  FormatStyle Style = getGoogleStyleWithColumns(22);
+  Style.Macros.push_back("MOCK_METHOD(r, n, a)=r n a");
+  // In the expanded code, we parse a full function signature, and afterwards
+  // know that we prefer not to break before the function name.
+  verifyFormat("MOCK_METHOD(\n"
+               "    type, variable,\n"
+               "    (type));",
+               Style);
+}
+
 } // namespace
 } // namespace test
 } // namespace format
diff --git a/clang/unittests/Format/FormatTestVerilog.cpp b/clang/unittests/Format/FormatTestVerilog.cpp
index fcda05df18268..abebf9f7d4c78 100644
--- a/clang/unittests/Format/FormatTestVerilog.cpp
+++ b/clang/unittests/Format/FormatTestVerilog.cpp
@@ -613,6 +613,17 @@ TEST_F(FormatTestVerilog, Headers) {
                "      (input var x aaaaaaaaaaaaaaa``x, \\\n"
                "                   b);",
                Style);
+  // When the ports line is not to be formatted, following lines should not take
+  // on its indentation.
+  verifyFormat("module x\n"
+               "    (output x);\n"
+               "  assign x = 0;\n"
+               "endmodule",
+               "module x\n"
+               "    (output x);\n"
+               "    assign x = 0;\n"
+               "endmodule",
+               getDefaultStyle(), {tooling::Range(25, 18)});
 }
 
 TEST_F(FormatTestVerilog, Hierarchy) {
diff --git a/clang/unittests/Format/TokenAnnotatorTest.cpp b/clang/unittests/Format/TokenAnnotatorTest.cpp
index 494205a1f2d8e..117d8fe8f7dc1 100644
--- a/clang/unittests/Format/TokenAnnotatorTest.cpp
+++ b/clang/unittests/Format/TokenAnnotatorTest.cpp
@@ -1066,6 +1066,11 @@ TEST_F(TokenAnnotatorTest, UnderstandsRequiresClausesAndConcepts) {
   EXPECT_EQ(Tokens.size(), 17u) << Tokens;
   EXPECT_TOKEN(Tokens[4], tok::amp, TT_PointerOrReference);
   EXPECT_TOKEN(Tokens[5], tok::kw_requires, TT_RequiresClause);
+
+  Tokens = annotate("template \n"
+                    "concept C = (!Foo) && Bar;");
+  ASSERT_EQ(Tokens.size(), 19u) << Tokens;
+  EXPECT_TOKEN(Tokens[15], tok::ampamp, TT_BinaryOperator);
 }
 
 TEST_F(TokenAnnotatorTest, UnderstandsRequiresExpressions) {
@@ -2172,6 +2177,40 @@ TEST_F(TokenAnnotatorTest, UnderstandsVerilogOperators) {
   EXPECT_TOKEN(Tokens[4], tok::string_literal, TT_Unknown);
 }
 
+TEST_F(TokenAnnotatorTest, UnderstandTableGenTokens) {
+  auto Style = getLLVMStyle(FormatStyle::LK_TableGen);
+  ASSERT_TRUE(Style.isTableGen());
+
+  TestLexer Lexer(Allocator, Buffers, Style);
+  AdditionalKeywords Keywords(Lexer.IdentTable);
+  auto Annotate = [&Lexer](llvm::StringRef Code) {
+    return Lexer.annotate(Code);
+  };
+
+  // Additional keywords representation test.
+  auto Tokens = Annotate("def foo : Bar<1>;");
+  ASSERT_TRUE(Keywords.isTableGenKeyword(*Tokens[0]));
+  ASSERT_TRUE(Keywords.isTableGenDefinition(*Tokens[0]));
+  ASSERT_TRUE(Tokens[0]->is(Keywords.kw_def));
+  ASSERT_TRUE(Tokens[1]->is(TT_StartOfName));
+
+  // Code, the multiline string token.
+  Tokens = Annotate("[{ code is multiline string }]");
+  ASSERT_EQ(Tokens.size(), 2u) << Tokens;
+  EXPECT_TOKEN(Tokens[0], tok::string_literal, TT_TableGenMultiLineString);
+  EXPECT_FALSE(Tokens[0]->IsMultiline);
+  // Case with multiple lines.
+  Tokens = Annotate("[{ It can break\n"
+                    "   across lines and the line breaks\n"
+                    "   are retained in \n"
+                    "   the string. }]");
+  ASSERT_EQ(Tokens.size(), 2u) << Tokens;
+  EXPECT_TOKEN(Tokens[0], tok::string_literal, TT_TableGenMultiLineString);
+  EXPECT_EQ(Tokens[0]->ColumnWidth, sizeof("[{ It can break\n") - 1);
+  EXPECT_TRUE(Tokens[0]->IsMultiline);
+  EXPECT_EQ(Tokens[0]->LastLineColumnWidth, sizeof("   the string. }]") - 1);
+}
+
 TEST_F(TokenAnnotatorTest, UnderstandConstructors) {
   auto Tokens = annotate("Class::Class() : BaseClass(), Member() {}");
 
diff --git a/clang/unittests/Interpreter/InterpreterTest.cpp b/clang/unittests/Interpreter/InterpreterTest.cpp
index 5f2911e9a7ada..1e0854b3c4af4 100644
--- a/clang/unittests/Interpreter/InterpreterTest.cpp
+++ b/clang/unittests/Interpreter/InterpreterTest.cpp
@@ -248,28 +248,10 @@ TEST(IncrementalProcessing, FindMangledNameSymbol) {
 #endif // _WIN32
 }
 
-static void *AllocateObject(TypeDecl *TD, Interpreter &Interp) {
+static Value AllocateObject(TypeDecl *TD, Interpreter &Interp) {
   std::string Name = TD->getQualifiedNameAsString();
-  const clang::Type *RDTy = TD->getTypeForDecl();
-  clang::ASTContext &C = Interp.getCompilerInstance()->getASTContext();
-  size_t Size = C.getTypeSize(RDTy);
-  void *Addr = malloc(Size);
-
-  // Tell the interpreter to call the default ctor with this memory. Synthesize:
-  // new (loc) ClassName;
-  static unsigned Counter = 0;
-  std::stringstream SS;
-  SS << "auto _v" << Counter++ << " = "
-     << "new ((void*)"
-     // Windows needs us to prefix the hexadecimal value of a pointer with '0x'.
-     << std::hex << std::showbase << (size_t)Addr << ")" << Name << "();";
-
-  auto R = Interp.ParseAndExecute(SS.str());
-  if (!R) {
-    free(Addr);
-    return nullptr;
-  }
-
+  Value Addr;
+  cantFail(Interp.ParseAndExecute("new " + Name + "()", &Addr));
   return Addr;
 }
 
@@ -317,7 +299,7 @@ TEST(IncrementalProcessing, InstantiateTemplate) {
   }
 
   TypeDecl *TD = cast(LookupSingleName(*Interp, "A"));
-  void *NewA = AllocateObject(TD, *Interp);
+  Value NewA = AllocateObject(TD, *Interp);
 
   // Find back the template specialization
   VarDecl *VD = static_cast(*PTUDeclRange.begin());
@@ -328,8 +310,7 @@ TEST(IncrementalProcessing, InstantiateTemplate) {
   typedef int (*TemplateSpecFn)(void *);
   auto fn =
       cantFail(Interp->getSymbolAddress(MangledName)).toPtr();
-  EXPECT_EQ(42, fn(NewA));
-  free(NewA);
+  EXPECT_EQ(42, fn(NewA.getPtr()));
 }
 
 #ifdef CLANG_INTERPRETER_NO_SUPPORT_EXEC
diff --git a/clang/unittests/Tooling/Syntax/TokensTest.cpp b/clang/unittests/Tooling/Syntax/TokensTest.cpp
index 0c08318a637c0..42f5169713965 100644
--- a/clang/unittests/Tooling/Syntax/TokensTest.cpp
+++ b/clang/unittests/Tooling/Syntax/TokensTest.cpp
@@ -816,6 +816,18 @@ TEST_F(TokenBufferTest, SpelledByExpanded) {
   EXPECT_EQ(Buffer.spelledForExpanded(findExpanded("prev good")), std::nullopt);
 }
 
+TEST_F(TokenBufferTest, NoCrashForEofToken) {
+  recordTokens(R"cpp(
+    int main() {
+  )cpp");
+  ASSERT_TRUE(!Buffer.expandedTokens().empty());
+  ASSERT_EQ(Buffer.expandedTokens().back().kind(), tok::eof);
+  // Expanded range including `eof` is handled gracefully (`eof` is ignored).
+  EXPECT_THAT(
+      Buffer.spelledForExpanded(Buffer.expandedTokens()),
+      ValueIs(SameRange(Buffer.spelledTokens(SourceMgr->getMainFileID()))));
+}
+
 TEST_F(TokenBufferTest, ExpandedTokensForRange) {
   recordTokens(R"cpp(
     #define SIGN(X) X##_washere
diff --git a/clang/utils/TableGen/ClangAttrEmitter.cpp b/clang/utils/TableGen/ClangAttrEmitter.cpp
index f3958aa0c7fd8..cc1b4afb0e823 100644
--- a/clang/utils/TableGen/ClangAttrEmitter.cpp
+++ b/clang/utils/TableGen/ClangAttrEmitter.cpp
@@ -1415,7 +1415,29 @@ namespace {
     }
   };
 
-} // end anonymous namespace
+  class WrappedAttr : public SimpleArgument {
+  public:
+    WrappedAttr(const Record &Arg, StringRef Attr)
+        : SimpleArgument(Arg, Attr, "Attr *") {}
+
+    void writePCHReadDecls(raw_ostream &OS) const override {
+      OS << "    Attr *" << getLowerName() << " = Record.readAttr();";
+    }
+
+    void writePCHWrite(raw_ostream &OS) const override {
+      OS << "    AddAttr(SA->get" << getUpperName() << "());";
+    }
+
+    void writeDump(raw_ostream &OS) const override {}
+
+    void writeDumpChildren(raw_ostream &OS) const override {
+      OS << "    Visit(SA->get" << getUpperName() << "());\n";
+    }
+
+    void writeHasChildren(raw_ostream &OS) const override { OS << "true"; }
+  };
+
+  } // end anonymous namespace
 
 static std::unique_ptr
 createArgument(const Record &Arg, StringRef Attr,
@@ -1474,6 +1496,8 @@ createArgument(const Record &Arg, StringRef Attr,
     Ptr = std::make_unique(Arg, Attr);
   else if (ArgName == "VersionArgument")
     Ptr = std::make_unique(Arg, Attr);
+  else if (ArgName == "WrappedAttr")
+    Ptr = std::make_unique(Arg, Attr);
   else if (ArgName == "OMPTraitInfoArgument")
     Ptr = std::make_unique(Arg, Attr, "OMPTraitInfo *");
   else if (ArgName == "VariadicOMPInteropInfoArgument")
@@ -1726,8 +1750,7 @@ SpellingNamesAreCommon(const std::vector& Spellings) {
   assert(!Spellings.empty() && "An empty list of spellings was provided");
   std::string FirstName =
       std::string(NormalizeNameForSpellingComparison(Spellings.front().name()));
-  for (const auto &Spelling :
-       llvm::make_range(std::next(Spellings.begin()), Spellings.end())) {
+  for (const auto &Spelling : llvm::drop_begin(Spellings)) {
     std::string Name =
         std::string(NormalizeNameForSpellingComparison(Spelling.name()));
     if (Name != FirstName)
@@ -3559,23 +3582,27 @@ static void GenerateHasAttrSpellingStringSwitch(
   OS << "    .Default(0);\n";
 }
 
-// Emits the list of tokens for regular keyword attributes.
-void EmitClangAttrTokenKinds(RecordKeeper &Records, raw_ostream &OS) {
-  emitSourceFileHeader("A list of tokens generated from the attribute"
-                       " definitions",
-                       OS);
+// Emits list of regular keyword attributes with info about their arguments.
+void EmitClangRegularKeywordAttributeInfo(RecordKeeper &Records,
+                                          raw_ostream &OS) {
+  emitSourceFileHeader(
+      "A list of regular keyword attributes generated from the attribute"
+      " definitions",
+      OS);
   // Assume for now that the same token is not used in multiple regular
   // keyword attributes.
   for (auto *R : Records.getAllDerivedDefinitions("Attr"))
-    for (const auto &S : GetFlattenedSpellings(*R))
-      if (isRegularKeywordAttribute(S)) {
-        if (!R->getValueAsListOfDefs("Args").empty())
-          PrintError(R->getLoc(),
-                     "RegularKeyword attributes with arguments are not "
-                     "yet supported");
-        OS << "KEYWORD_ATTRIBUTE("
-           << S.getSpellingRecord().getValueAsString("Name") << ")\n";
-      }
+    for (const auto &S : GetFlattenedSpellings(*R)) {
+      if (!isRegularKeywordAttribute(S))
+        continue;
+      std::vector Args = R->getValueAsListOfDefs("Args");
+      bool HasArgs = llvm::any_of(
+          Args, [](const Record *Arg) { return !Arg->getValueAsBit("Fake"); });
+
+      OS << "KEYWORD_ATTRIBUTE("
+         << S.getSpellingRecord().getValueAsString("Name") << ", "
+         << (HasArgs ? "true" : "false") << ")\n";
+    }
   OS << "#undef KEYWORD_ATTRIBUTE\n";
 }
 
diff --git a/clang/utils/TableGen/RISCVVEmitter.cpp b/clang/utils/TableGen/RISCVVEmitter.cpp
index d570bcae8d863..2ca47f1ba59f3 100644
--- a/clang/utils/TableGen/RISCVVEmitter.cpp
+++ b/clang/utils/TableGen/RISCVVEmitter.cpp
@@ -656,7 +656,7 @@ void RVVEmitter::createRVVIntrinsics(
       RVVRequire RequireExt =
           StringSwitch(RequiredFeature)
               .Case("RV64", RVV_REQ_RV64)
-              .Case("ZvfhminOrZvfh", RVV_REQ_ZvfhminOrZvfh)
+              .Case("Zvfhmin", RVV_REQ_Zvfhmin)
               .Case("Xsfvcp", RVV_REQ_Xsfvcp)
               .Case("Xsfvfnrclipxfqf", RVV_REQ_Xsfvfnrclipxfqf)
               .Case("Xsfvfwmaccqqq", RVV_REQ_Xsfvfwmaccqqq)
diff --git a/clang/utils/TableGen/SveEmitter.cpp b/clang/utils/TableGen/SveEmitter.cpp
index 5de2223e71b04..99b7148c17496 100644
--- a/clang/utils/TableGen/SveEmitter.cpp
+++ b/clang/utils/TableGen/SveEmitter.cpp
@@ -1066,17 +1066,6 @@ void Intrinsic::emitIntrinsic(raw_ostream &OS, SVEEmitter &Emitter,
 
   std::string FullName = mangleName(ClassS);
   std::string ProtoName = mangleName(getClassKind());
-  std::string SMEAttrs = "";
-
-  if (Flags & Emitter.getEnumValueForFlag("IsStreaming"))
-    SMEAttrs += ", arm_streaming";
-  if (Flags & Emitter.getEnumValueForFlag("IsStreamingCompatible"))
-    SMEAttrs += ", arm_streaming_compatible";
-  if (Flags & Emitter.getEnumValueForFlag("IsSharedZA"))
-    SMEAttrs += ", arm_shared_za";
-  if (Flags & Emitter.getEnumValueForFlag("IsPreservesZA"))
-    SMEAttrs += ", arm_preserves_za";
-
   OS << (IsOverloaded ? "__aio " : "__ai ")
      << "__attribute__((__clang_arm_builtin_alias(";
 
@@ -1089,8 +1078,6 @@ void Intrinsic::emitIntrinsic(raw_ostream &OS, SVEEmitter &Emitter,
     break;
   }
 
-  if (!SMEAttrs.empty())
-    OS << SMEAttrs;
   OS << "))\n";
 
   OS << getTypes()[0].str() << " " << ProtoName << "(";
@@ -1619,7 +1606,7 @@ void SVEEmitter::createSMEHeader(raw_ostream &OS) {
   OS << "}\n\n";
 
   OS << "__ai __attribute__((target(\"sme\"))) void svundef_za(void) "
-        "__arm_streaming_compatible __arm_shared_za "
+        "__arm_streaming_compatible __arm_out(\"za\") "
         "{ }\n\n";
 
   createCoreHeaderIntrinsics(OS, *this, ACLEKind::SME);
diff --git a/clang/utils/TableGen/TableGen.cpp b/clang/utils/TableGen/TableGen.cpp
index d3ab1fd451816..9712521bd6e9a 100644
--- a/clang/utils/TableGen/TableGen.cpp
+++ b/clang/utils/TableGen/TableGen.cpp
@@ -37,7 +37,7 @@ enum ActionType {
   GenClangAttrSubjectMatchRuleList,
   GenClangAttrPCHRead,
   GenClangAttrPCHWrite,
-  GenClangAttrTokenKinds,
+  GenClangRegularKeywordAttributeInfo,
   GenClangAttrHasAttributeImpl,
   GenClangAttrSpellingListIndex,
   GenClangAttrASTVisitor,
@@ -151,8 +151,10 @@ cl::opt Action(
                    "Generate clang PCH attribute reader"),
         clEnumValN(GenClangAttrPCHWrite, "gen-clang-attr-pch-write",
                    "Generate clang PCH attribute writer"),
-        clEnumValN(GenClangAttrTokenKinds, "gen-clang-attr-token-kinds",
-                   "Generate a list of attribute-related clang tokens"),
+        clEnumValN(GenClangRegularKeywordAttributeInfo,
+                   "gen-clang-regular-keyword-attr-info",
+                   "Generate a list of regular keyword attributes with info "
+                   "about their arguments"),
         clEnumValN(GenClangAttrHasAttributeImpl,
                    "gen-clang-attr-has-attribute-impl",
                    "Generate a clang attribute spelling list"),
@@ -294,11 +296,14 @@ cl::opt Action(
                    "Generate riscv_vector_builtin_cg.inc for clang"),
         clEnumValN(GenRISCVVectorBuiltinSema, "gen-riscv-vector-builtin-sema",
                    "Generate riscv_vector_builtin_sema.inc for clang"),
-        clEnumValN(GenRISCVSiFiveVectorBuiltins, "gen-riscv-sifive-vector-builtins",
+        clEnumValN(GenRISCVSiFiveVectorBuiltins,
+                   "gen-riscv-sifive-vector-builtins",
                    "Generate riscv_sifive_vector_builtins.inc for clang"),
-        clEnumValN(GenRISCVSiFiveVectorBuiltinCG, "gen-riscv-sifive-vector-builtin-codegen",
+        clEnumValN(GenRISCVSiFiveVectorBuiltinCG,
+                   "gen-riscv-sifive-vector-builtin-codegen",
                    "Generate riscv_sifive_vector_builtin_cg.inc for clang"),
-        clEnumValN(GenRISCVSiFiveVectorBuiltinSema, "gen-riscv-sifive-vector-builtin-sema",
+        clEnumValN(GenRISCVSiFiveVectorBuiltinSema,
+                   "gen-riscv-sifive-vector-builtin-sema",
                    "Generate riscv_sifive_vector_builtin_sema.inc for clang"),
         clEnumValN(GenAttrDocs, "gen-attr-docs",
                    "Generate attribute documentation"),
@@ -358,8 +363,8 @@ bool ClangTableGenMain(raw_ostream &OS, RecordKeeper &Records) {
   case GenClangAttrPCHWrite:
     EmitClangAttrPCHWrite(Records, OS);
     break;
-  case GenClangAttrTokenKinds:
-    EmitClangAttrTokenKinds(Records, OS);
+  case GenClangRegularKeywordAttributeInfo:
+    EmitClangRegularKeywordAttributeInfo(Records, OS);
     break;
   case GenClangAttrHasAttributeImpl:
     EmitClangAttrHasAttrImpl(Records, OS);
diff --git a/clang/utils/TableGen/TableGenBackends.h b/clang/utils/TableGen/TableGenBackends.h
index dad2e2f3e9ba7..665dcf69aa7d3 100644
--- a/clang/utils/TableGen/TableGenBackends.h
+++ b/clang/utils/TableGen/TableGenBackends.h
@@ -53,8 +53,8 @@ void EmitClangAttrSubjectMatchRuleList(llvm::RecordKeeper &Records,
                                        llvm::raw_ostream &OS);
 void EmitClangAttrPCHRead(llvm::RecordKeeper &Records, llvm::raw_ostream &OS);
 void EmitClangAttrPCHWrite(llvm::RecordKeeper &Records, llvm::raw_ostream &OS);
-void EmitClangAttrTokenKinds(llvm::RecordKeeper &Records,
-                             llvm::raw_ostream &OS);
+void EmitClangRegularKeywordAttributeInfo(llvm::RecordKeeper &Records,
+                                          llvm::raw_ostream &OS);
 void EmitClangAttrHasAttrImpl(llvm::RecordKeeper &Records,
                               llvm::raw_ostream &OS);
 void EmitClangAttrSpellingListIndex(llvm::RecordKeeper &Records,
diff --git a/clang/www/c_dr_status.html b/clang/www/c_dr_status.html
index 4fe088e297750..fa2ceb1be58bd 100644
--- a/clang/www/c_dr_status.html
+++ b/clang/www/c_dr_status.html
@@ -12,7 +12,7 @@
     .unreleased { background-color: #FFFF99 }
     .full { background-color: #CCFF99 }
     .na { background-color: #DDDDDD }
-    .unknown { background-color: #FF55FF }
+    .unknown { background-color: #EBCAFE }
     .open * { color: #AAAAAA }
     //.open { filter: opacity(0.2) }
     tr:target { background-color: #FFFFBB }
@@ -35,7 +35,7 @@ 
C defect report implementation status

 
 
The implementation status for defect reports against the C Standard are
 currently under investigation. Any defect report whose status in Clang is
-currently unknown will be marked in magenta.

+currently unknown will be marked in purple.

 
 
The LLVM bug tracker uses
 the "c", "c99", "c11", "c17", and "c23" labels to track known bugs with Clang's language
diff --git a/clang/www/c_status.html b/clang/www/c_status.html
index 47acb1f87b8e3..b9e0650ddca24 100644
--- a/clang/www/c_status.html
+++ b/clang/www/c_status.html
@@ -9,7 +9,7 @@
     .none { background-color: #FFCCCC }
     .partial { background-color: #FFE0B0 }
     .unreleased { background-color: #FFFF99 }
-    .unknown { background-color: #FF55FF }
+    .unknown { background-color: #EBCAFE }
     .full { background-color: #CCFF99 }
     .na { background-color: #DDDDDD }
     :target { background-color: #FFFFBB; outline: #DDDD55 solid thin; }
diff --git a/clang/www/cxx_dr_status.html b/clang/www/cxx_dr_status.html
index 2bded63d5cd41..3ee359945582f 100755
--- a/clang/www/cxx_dr_status.html
+++ b/clang/www/cxx_dr_status.html
@@ -8,6 +8,7 @@