[SYCL][Graph] Explicit update with index only #350

reble · 2024-01-12T19:20:02Z

This patch removes the match by value functionality,
thus an index is always required to update a kernel parameter.

sycl/doc/extensions/experimental/sycl_ext_oneapi_graph.asciidoc

…attribute (#77426) This patch reworks RISCVTargetInfo::initFeatureMap to fix the issue described in llvm/llvm-project#74889 (review) (and is an alternative to #75804) When a full arch string is specified, a "full" list of extensions is now passed after the __RISCV_TargetAttrNeedOverride marker feature, which includes any negative features that disable ISA extensions. In initFeatureMap, there are now two code paths: 1. If the arch string was overriden, use the "full" list of override features, only adding back any non-isa features that were specified. Using the full list of positive and negative features will mean that the target-cpu will have no effect on the final arch, e.g. __attribute__((target("arch=rv64i"))) with -mcpu=sifive-x280 will have the features for rv64i, not a mix of both. 2. Otherwise, parse and *append* the list of implied features. By appending, we turn back on any features that might have been disabled by a negative extension, i.e. this handles the case fixed in #74889.

## Why The documentation is written relatively to `clang-llvm`, not the root repository directory. However, some steps in the documentation are relative to the repository root, which is not correct. ## What Documentation steps have been modified to make them correct and outdated ones were updated. Some details: * Correct paths in documentation * Change `bootstrap.py` -> `configure.py` since bootstraping Ninja has [slightly changed](https://github.com/ninja-build/ninja/tree/master?tab=readme-ov-file#python)

…8137)

This removes the entire modules testing infrastructure. The current infrastructure uses CMake to generate the std and std.compat module. This requires quite a bit of plumbing and uses CMake. Since CMake introduced module support in CMake 3.26, modules have a higher CMake requirement than the rest of the LLVM project. (The LLVM project requires 3.20.) The main motivation for this approach was how libc++ generated its modules. Every header had its own module partition. This was changed to improve performance and now only two modules remain. The code to build these can be manually crafted. A followup patch will reenable testing modules, using a different approach.

This adds a new module test infrastructure. This requires tagging tests using modules. The test runner uses this information to determine the compiler flags needed to build and use the module. Currently modules are build per test, which allows testing them for tests with ADDITIONAL_COMPILE_FLAGS. At the moment only 4 tests use modules. Therefore the performance penalty is not measurable. If in the future more tests use modules it would be good to measure the overhead and determine whether it's acceptable.

As suggested in #71438 we should use export import std; in the std.compat module. Testing this locally failed when building with the clang-tidy-17 plugin. The std module was considered corrupt in the test libcxx/test/libcxx/module_std_compat.gen.py however the test libcxx/test/libcxx/module_std.gen.py passed. Both test generated identical std.pcm files. Using the clang-tidy-18 plugin solves the issue.

Clang-tidy 18 no longer has false positives with the spaceship operator. Note that I'm quite sure there are more occurrences in our headers that are not caught.

As suggested in #71438 we should use export import std; in the std.compat module. Using this exports some named declarations from functions and records, adding them to the global namespace. Clang correctly does not export these and it's an issue in the declaration filtering. Declarations in function or record context are not considered a global named declaration.

This commit simplifies ASan helper functions in `std::vector` by removing arguments which can be calculated later. Short term it improves readability of helper functions in `std::vector`. Long term it aims to help with a bigger refactor of container annotations.

… reference (#77886) Printing the raw symbol is useful in inline asm (e.g. getting the C++ mangled name, referencing a symbol in a custom way while ensuring it is not optimized out even if internal). Similar constraints are available in other targets (e.g. "S" for aarch64/riscv, "Cs" for m68k). ``` namespace ns { extern int var, a[4]; } void foo() { asm(".pushsection .xxx,\"aw\"; .dc.a %p0; .popsection" :: "Ws"(&ns::var)); asm(".reloc ., BFD_RELOC_NONE, %p0" :: "Ws"(&ns::a[3])); } ``` Link: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105576

… NFC This illustrates #78352. There are other tests for gathers/scatters and masked load/stores which likely have the same issue, but I've left them out for now since they don't seem to be affected by fixing RISCVTTIImpl::getMemoryOpCost.

In https://reviews.llvm.org/D151292 I added the ability to track address masks separately for high and low memory addresses, a capability of AArch64. I did my testing with manual address mask settings (via target.process.highmem-virtual-addressable-bits) but didn't have a real corefile that included this metadata and required it. My intention is that when the high address mask isn't specified, by the user (via the setting) or the Process plugin, we fall back to using the low address mask. The low and high address mask is the same for almost all environments. But the patch I wrote never uses the Process plugin high address mask if it was set, e.g. from corefile metadata. This patch corrects that. I also have an old patch in Phabractor that was approved to add FixAddress methods to SBProcess; I need to pick that patch up and finish it (I wanted to add an enum to specify which mask is being requested iirc), so I can do address masks tests in API tests. rdar://120926000

…R<->FPR moves The fmv can be removed through appropriate logic in RISCVInstrInfo::foldMemoryOperandImpl.

…… (#74125) …tructions into account Hint instructions like G_ASSERT_ZEXT cann be viewed as a copy. Including this fact into the combiner allows the match more patterns involving such instructions.

BasicAA currently says that any Constant cannot alias an identified local object. This is not correct if the local object escaped, as it's possible to create a pointer to the escaped object using an inttoptr constant expression base. To compensate for this, make sure that inttoptr constant expressions are treated as escape sources, just like inttoptr instructions. This ensures that the optimization can still be applied if the local object is non-escaping. This is sufficient to still optimize the original motivating case from c53e2ec. Fixes llvm/llvm-project#76789.

Add a debug counter for DAGCombine. This can help with bisecting which DAG combine introduced a miscompile.

Add a debug counter that allows forcing an sdag fallback after a certain number of functions. The intended use-case is to bisect which function gets miscompiled by global isel using `-debug-counter=globalisel-count=N` (in cases where sdag doesn't also miscompile it, of course). The "falling back" debug line is printed unconditionally, because using `-debug-only` is usually too spammy for the intended purpose.

We currently require that all referenced globals have an explicit declaration or definition in the IR. For intrinsics, this requirement is redundant, because they cannot be called indirectly (including "direct" calls with mismatched function type). The function type used in the call directly determines the function type of the intrinsic declaration. Relax this requirement, and implicitly declare any intrinsics that do not have an explicit declaration. This will remove a common annoyance when writing tests and alive2 proofs. (I also plan to introduce a mode where declarations for all missing symbols will be automatically added, to make working with incomplete IR easier -- but that will be behind a default-disabled flag.)

CONFLICT (content): Merge conflict in clang/lib/Driver/ToolChains/Clang.cpp

Remove Python 2 support and clean up code that conditions based on version. Issue #76664.

This enables the libclang python binding test to check the oldest version of Python supported in addition to the normal python version. It is important to check this for issue #76664, since many new mainstream python type annotation features and best practices are not compatible with older versions of python. Additionally, frustration around ever increasing platform dependencies and versions has been raised. This will help ensure that python maintains reasonable backwards compatibility. Adding this additional build step will increase the run time, but this should always be minimal, since the additional libclang compilation should see 100% cache hit rate. Issue #76664. Fixes #76601.

The wording "fails silently" has been sometimes used to indicate that a silenceable failure was emitted by the operation. The meaning is exactly the opposite: silenceable failure is _not_ silent unless silenced.

…#78409) This reverts commit 020ea3e. This seems to break `test/libcxx/clang_tidy.gen.py/regex.sh.cpp` (in C++03 mode for some reason): ``` 2024-01-17T07:32:22.1759374Z # RUN: at line 12 2024-01-17T07:32:22.1773919Z clang-tidy-18 /home/runner/_work/llvm-project/llvm-project/build/generic-cxx03/test/libcxx/clang_tidy.gen.py/regex.sh.cpp --warnings-as-errors=* -header-filter=.* --config-file=/home/runner/_work/llvm-project/llvm-project/libcxx/.clang-tidy -- -Wweak-vtables -nostdinc++ -I /home/runner/_work/llvm-project/llvm-project/build/generic-cxx03/include/c++/v1 -I /home/runner/_work/llvm-project/llvm-project/build/generic-cxx03/include/c++/v1 -I /home/runner/_work/llvm-project/llvm-project/libcxx/test/support -std=c++03 -Werror -Wall -Wctad-maybe-unsupported -Wextra -Wshadow -Wundef -Wunused-template -Wno-unused-command-line-argument -Wno-attributes -Wno-pessimizing-move -Wno-noexcept-type -Wno-atomic-alignment -Wno-reserved-module-identifier -Wdeprecated-copy -Wdeprecated-copy-dtor -Wno-user-defined-literals -Wno-tautological-compare -Wsign-compare -Wunused-variable -Wunused-parameter -Wunreachable-code -Wno-unused-local-typedef -Wno-local-type-template-args -Wno-c++11-extensions -Wno-unknown-pragmas -Wno-pass-failed -Wno-mismatched-new-delete -Wno-redundant-move -Wno-self-move -D_LIBCPP_HAS_NO_PRAGMA_SYSTEM_HEADER -D_LIBCPP_ENABLE_EXPERIMENTAL -D_LIBCPP_HARDENING_MODE=_LIBCPP_HARDENING_MODE_NONE -Werror=thread-safety -Wuser-defined-warnings -fno-modules 2024-01-17T07:32:22.1803227Z # executed command: clang-tidy-18 /home/runner/_work/llvm-project/llvm-project/build/generic-cxx03/test/libcxx/clang_tidy.gen.py/regex.sh.cpp '--warnings-as-errors=*' '-header-filter=.*' --config-file=/home/runner/_work/llvm-project/llvm-project/libcxx/.clang-tidy -- -Wweak-vtables -nostdinc++ -I /home/runner/_work/llvm-project/llvm-project/build/generic-cxx03/include/c++/v1 -I /home/runner/_work/llvm-project/llvm-project/build/generic-cxx03/include/c++/v1 -I /home/runner/_work/llvm-project/llvm-project/libcxx/test/support -std=c++03 -Werror -Wall -Wctad-maybe-unsupported -Wextra -Wshadow -Wundef -Wunused-template -Wno-unused-command-line-argument -Wno-attributes -Wno-pessimizing-move -Wno-noexcept-type -Wno-atomic-alignment -Wno-reserved-module-identifier -Wdeprecated-copy -Wdeprecated-copy-dtor -Wno-user-defined-literals -Wno-tautological-compare -Wsign-compare -Wunused-variable -Wunused-parameter -Wunreachable-code -Wno-unused-local-typedef -Wno-local-type-template-args -Wno-c++11-extensions -Wno-unknown-pragmas -Wno-pass-failed -Wno-mismatched-new-delete -Wno-redundant-move -Wno-self-move -D_LIBCPP_HAS_NO_PRAGMA_SYSTEM_HEADER -D_LIBCPP_ENABLE_EXPERIMENTAL -D_LIBCPP_HARDENING_MODE=_LIBCPP_HARDENING_MODE_NONE -Werror=thread-safety -Wuser-defined-warnings -fno-modules 2024-01-17T07:32:22.1817757Z # .---command stdout------------ 2024-01-17T07:32:22.1820160Z # | /home/runner/_work/llvm-project/llvm-project/build/generic-cxx03/include/c++/v1/__iterator/iterator_traits.h:124:50: error: use nullptr [modernize-use-nullptr,-warnings-as-errors] 2024-01-17T07:32:22.1822498Z # | 124 | static const bool value = decltype(__test<_Tp>(0, 0, 0, 0, 0))::value; 2024-01-17T07:32:22.1823337Z # | | ^ 2024-01-17T07:32:22.1824052Z # | | nullptr 2024-01-17T07:32:22.1826651Z # | /home/runner/_work/llvm-project/llvm-project/build/generic-cxx03/include/c++/v1/__iterator/iterator_traits.h:124:53: error: use nullptr [modernize-use-nullptr,-warnings-as-errors] 2024-01-17T07:32:22.1829325Z # | 124 | static const bool value = decltype(__test<_Tp>(0, 0, 0, 0, 0))::value; 2024-01-17T07:32:22.1831137Z # | | ^ 2024-01-17T07:32:22.1831963Z # | | nullptr 2024-01-17T07:32:22.1834149Z # | /home/runner/_work/llvm-project/llvm-project/build/generic-cxx03/include/c++/v1/__iterator/iterator_traits.h:124:56: error: use nullptr [modernize-use-nullptr,-warnings-as-errors] 2024-01-17T07:32:22.1848711Z # | 124 | static const bool value = decltype(__test<_Tp>(0, 0, 0, 0, 0))::value; 2024-01-17T07:32:22.1849506Z # | | ^ 2024-01-17T07:32:22.1849997Z # | | nullptr 2024-01-17T07:32:22.1851528Z # | /home/runner/_work/llvm-project/llvm-project/build/generic-cxx03/include/c++/v1/__iterator/iterator_traits.h:124:59: error: use nullptr [modernize-use-nullptr,-warnings-as-errors] 2024-01-17T07:32:22.1853391Z # | 124 | static const bool value = decltype(__test<_Tp>(0, 0, 0, 0, 0))::value; 2024-01-17T07:32:22.1854277Z # | | ^ 2024-01-17T07:32:22.1854841Z # | | nullptr 2024-01-17T07:32:22.1856314Z # | /home/runner/_work/llvm-project/llvm-project/build/generic-cxx03/include/c++/v1/__iterator/iterator_traits.h:124:62: error: use nullptr [modernize-use-nullptr,-warnings-as-errors] 2024-01-17T07:32:22.1857768Z # | 124 | static const bool value = decltype(__test<_Tp>(0, 0, 0, 0, 0))::value; 2024-01-17T07:32:22.1858493Z # | | ^ 2024-01-17T07:32:22.1858998Z # | | nullptr 2024-01-17T07:32:22.1859503Z # `----------------------------- 2024-01-17T07:32:22.1859900Z # .---command stderr------------ 2024-01-17T07:32:22.1860259Z # | 43 warnings generated. 2024-01-17T07:32:22.1860710Z # | Suppressed 33 warnings (33 in non-user code). 2024-01-17T07:32:22.1861809Z # | Use -header-filter=.* to display errors from all non-system headers. Use -system-headers to display errors from system headers as well. 2024-01-17T07:32:22.1862750Z # | 5 warnings treated as errors 2024-01-17T07:32:22.1863229Z # `----------------------------- 2024-01-17T07:32:22.1863622Z # error: command failed with exit status: 1 ``` (see https://github.com/llvm/llvm-project/actions/runs/7552394884/job/20561192096)

…#76641) Fixes llvm/llvm-project#76549 The cause of the optimization miss was - 1. `optimizePow` converting almost integer FP exponents to integer, and turning `pow` to `powi`. 2. `optimizeLog` not accepting `Intrinsic::powi` as a target. This patch converts constantInt back to constantFP where applicable and adds a test.

This fixes #71892 allowing us to check magled names in the IR verifier.

…. (#71556) Substitute with zero-extended to i64 ballot.i32 intrinsic.

EwanC · 2024-01-19T09:14:16Z

I think we need to change how the dynamic_parameter class is templated to as suggested by Greg in [SYCL][Graph] Add API for updating single node parameters #340 (comment)

template<typename T>
class dynamic_parameter {
 public:
  dynamic_parameter(const T& param, command_graph<graph_state::modifiable> graph, std::string label = "");
  register(const handler &cgh);
  update(const T& newval, command_graph<graph_state::executable> execGraph);
};

I think we need to remove wording about changing concurrent execution management and revisit this later. We haven't fully thought this through and the implementation change will drag out the time to implement this spec change

If `graph` is submitted multiple times and
`property::graph::updatable_graph` is not set, dependencies are automatically
added by the runtime to prevent concurrent executions of an identical graph.

This PR preserves unsigned return type of image read and unsigned texel type of image write builtins as ZeroExtend image operand in SPIRV. Original commit: KhronosGroup/SPIRV-LLVM-Translator@0aab124

…ntel#2300) Original commit: KhronosGroup/SPIRV-LLVM-Translator@2108fc5

This functionality was added in SPIR-V 1.2 and allows using an <id> to set the execution modes SubgroupsPerWorkgroupId, LocalSizeId, and LocalSizeHintI, and others. Original commit: KhronosGroup/SPIRV-LLVM-Translator@10b0aab

Due to 158d72d

intel#11840 is resolved by usptreaming the fix in intel#12068 to llvm/llvm-project#74437.

This reverts commit 8c2b0d4.

sycl/doc/extensions/experimental/sycl_ext_oneapi_graph.asciidoc

EwanC · 2024-01-23T09:11:33Z

sycl/doc/extensions/experimental/sycl_ext_oneapi_graph.asciidoc

+Having a more unified approach to updating nodes/graphs may provide a more
+concise and easy to use API.


Been thinking about this since we chatted yesterday, and I'm not mad about going forward with an API that we think we could do better with and will change very soon after delivering. If it'll only take a day or two to modify this so that nd_range is updated on the node, and nodes are passed to a graph to update, then it might be worth doing.

new commit looks good, think we can remove this issue item

…ntel#12433) When a host-task is submitted to in-order queue, dependency between this host-task and the successor is explicitly handled. However, when we record an in-order queue, the recorded CG are not part of the regular in-order queue execution sequence. But inter-CG dependencies are managed by the graph implementation. This PR implements this point and ensures that recording an in-order does not impact the normal execution sequence. Tests (e2e and unitest) have been added to check it. --------- Co-authored-by: Marcos Maronas <maarquitos14@users.noreply.github.com>

The method was deprecated in favor of `StringRef::starts_with` in llvm/llvm-project#75491

pre-commit Pr for oneapi-src/unified-runtime#1135 --------- Signed-off-by: Spruit, Neil R <neil.r.spruit@intel.com> Co-authored-by: Kenneth Benzie (Benie) <k.benzie@codeplay.com>

Bensuo · 2024-01-23T14:46:33Z

I've pushed a commit wihch proposes some API changes to (IMO) improve the usage of the API, and make things a bit clearer in what's happening.

The main change is the ND-ranges are updated on the Node itself, and that updates to the executable graph are triggered by explicitly passing updates nodes to the graph. This to me serves two purposes:

Unifies how Node param and ND-range updates are applied to the executable graph
Makes explicit for the user where the update actually occurs. Previously dynamic_parameter had a lot of responsibility, both updating nodes and also applying those updates to the executable graph. IMO this is clearer and easier to use.

EwanC

Comments are minor nits, LGTM

sycl/doc/extensions/experimental/sycl_ext_oneapi_graph.asciidoc

EwanC · 2024-01-23T16:15:19Z

sycl/doc/extensions/experimental/sycl_ext_oneapi_graph.asciidoc

+Having a more unified approach to updating nodes/graphs may provide a more
+concise and easy to use API.


new commit looks good, think we can remove this issue item

…erations with simd vectors (intel#12089)" (intel#12462) This reverts commit 8c92df9 to address test failures

LLVM: llvm/llvm-project@110e1717b31e SPIRV-LLVM-Translator: KhronosGroup/SPIRV-LLVM-Translator@10b0aab

…l#12414) Signed-off-by: Vyacheslav N Klochkov <vyacheslav.n.klochkov@intel.com>

…ntel#12470) Those test cases fail even with the most recent GPU drivers, it seems the issue with such N was fixed for slm_block_load(), but not for slm_block_store(). Signed-off-by: Klochkov, Vyacheslav N <vyacheslav.n.klochkov@intel.com>

…l name (intel#12393) Implement a custom string class (backed by a `DynArray<char>`, hence owning its memory) to represent kernel names on the fusion interface. *This PR is part of a series of changes to remove uses of STL classes in the kernel fusion interface to prevent ABI issues in the future.* Signed-off-by: Julian Oppermann <julian.oppermann@codeplay.com>

CI testing for oneapi-src/unified-runtime#1070 --------- Co-authored-by: Kenneth Benzie (Benie) <k.benzie@codeplay.com>

…intel#12445) Untangle the interaction between options and the `Config` class, and hide `Config` (which needs to store options in a map) from the KF interface. The idea is to introduce free functions to interact with the `thread_local` instance of `Config` held by the existing `ConfigHelper` class, instead of letting the client construct the `Config` object and handing it to over `ConfigHelper` for storing it. _This PR is part of a series of changes to remove uses of STL classes in the kernel fusion interface to prevent ABI issues in the future._ Signed-off-by: Julian Oppermann <julian.oppermann@codeplay.com>

reble · 2024-01-24T15:49:08Z

sycl/doc/extensions/experimental/sycl_ext_oneapi_graph.asciidoc

+Dynamic parameters are arguments to a node's command-group which can be updated
+by the user after the node has been added to a graph. Updating the value of a
+dynamic parameter will be reflected in the modifiable graph which contains this
+node. Executable graphs can then be updated with these update nodes


Suggested change

node. Executable graphs can then be updated with these update nodes

node.

Sentence doesn't make sense to me. Maybe we can be more specific or just remove it.

I've tried to rework this sentence to make it read better. I think it's important to keep something here that explains how the update makes it's way to the executable graph in the end though.

Scheduled drivers uplift Co-authored-by: GitHub Actions <actions@github.com>

It's failing in postcommit, we need this to fix the CI. Signed-off-by: Sarnie, Nick <nick.sarnie@intel.com>

- Adds an API to the spec for updating graph nodes using explicit indices - Also includes functionality for updating ND-range of kernel nodes - Note: Current design is only for kernel execution nodes

Bensuo · 2024-01-24T18:35:55Z

Closing in favour of upstream PR: intel#12486

reble commented Jan 12, 2024

View reviewed changes

reble commented Jan 16, 2024

View reviewed changes

sycl/doc/extensions/experimental/sycl_ext_oneapi_graph.asciidoc Outdated Show resolved Hide resolved

lukel97 and others added 28 commits January 17, 2024 13:32

[mlir][Tosa]: Add folder to ReciprocalOp of splat constant inputs (#7…

d89a0a6

…8137)

[libc++] Clang-tidy enable modernize-use-nullptr. (#76659)

020ea3e

Clang-tidy 18 no longer has false positives with the spaceship operator. Note that I'm quite sure there are more occurrences in our headers that are not caught.

[RISCV][test] Test showing missed optimisation for spills/fills of GP…

da0755f

…R<->FPR moves The fmv can be removed through appropriate logic in RISCVInstrInfo::foldMemoryOperandImpl.

[GlobalISel] Improve combines for extend operation by taking hint ins…

55172b7

…… (#74125) …tructions into account Hint instructions like G_ASSERT_ZEXT cann be viewed as a copy. Including this fact into the combiner allows the match more patterns involving such instructions.

[DAGCombine] Add debug counter (#78259)

cde780c

Add a debug counter for DAGCombine. This can help with bisecting which DAG combine introduced a miscompile.

Merge from 'main' to 'sycl-web' (3 commits)

ee7fbbb

CONFLICT (content): Merge conflict in clang/lib/Driver/ToolChains/Clang.cpp

[libclang/python] Bump minimum compatibility to Python 3.6 (#77228)

75f2321

Remove Python 2 support and clean up code that conditions based on version. Issue #76664.

[mlir] fix wording in transform dialect docs

baa39b7

The wording "fails silently" has been sometimes used to indicate that a silenceable failure was emitted by the operation. The meaning is exactly the opposite: silenceable failure is _not_ silent unless silenced.

AMDGPU: Remove fixed fixme from a test

53a3c73

[VFABI] Move the Vector ABI demangling utility to LLVMCore. (#77513)

92289db

This fixes #71892 allowing us to check magled names in the IR verifier.

[AArch64] Use parseArchExtension function. NFC (#78158)

7f7bbb9

[AMDGPU] Add InstCombine rule for ballot.i64 intrinsic in wave32 mode…

57b50ef

…. (#71556) Substitute with zero-extended to i64 ballot.i32 intrinsic.

sys-ce-bb and others added 7 commits January 19, 2024 07:06

Merge commit '110e1717b31e' into llvmspirv_pulldown

8cae952

Fix image builtin unsigned type not preserved (intel#2286)

fd5820b

This PR preserves unsigned return type of image read and unsigned texel type of image write builtins as ZeroExtend image operand in SPIRV. Original commit: KhronosGroup/SPIRV-LLVM-Translator@0aab124

Fix GEP generation for reverse translation of OpCompositeConstruct (i…

28a8b6b

…ntel#2300) Original commit: KhronosGroup/SPIRV-LLVM-Translator@2108fc5

Update lit test with new dead_on_unwind/writable attr

c28044f

Due to 158d72d

Remove XFAIL as fixed by aa23e49

69c77fe

intel#11840 is resolved by usptreaming the fix in intel#12068 to llvm/llvm-project#74437.

Revert "[Clang] Fix dependency of SourceLocExpr. (#78436)"

80f470c

This reverts commit 8c2b0d4.

Bensuo changed the base branch from ben/explicit-update to sycl January 22, 2024 18:13

EwanC reviewed Jan 23, 2024

View reviewed changes

mfrancepillois and others added 3 commits January 23, 2024 13:15

[SYCL][NFC] Modernize use of StringRef::startswith (intel#12406)

00754b7

The method was deprecated in favor of `StringRef::starts_with` in llvm/llvm-project#75491

[L0] Add support for multi-device kernel compilation (intel#12032)

dc37ee4

pre-commit Pr for oneapi-src/unified-runtime#1135 --------- Signed-off-by: Spruit, Neil R <neil.r.spruit@intel.com> Co-authored-by: Kenneth Benzie (Benie) <k.benzie@codeplay.com>

EwanC approved these changes Jan 23, 2024

View reviewed changes

fineg74 and others added 8 commits January 23, 2024 11:55

Revert "[SYCL][ESIMD]Limit bfloat16 operators to scalars to enable op…

30a2dae

…erations with simd vectors (intel#12089)" (intel#12462) This reverts commit 8c92df9 to address test failures

LLVM and SPIRV-LLVM-Translator pulldown (WW03)

ae24111

LLVM: llvm/llvm-project@110e1717b31e SPIRV-LLVM-Translator: KhronosGroup/SPIRV-LLVM-Translator@10b0aab

[ESIMD] Implement gather(acc) accepting compile-time properties (inte…

655ab10

…l#12414) Signed-off-by: Vyacheslav N Klochkov <vyacheslav.n.klochkov@intel.com>

[NFCI][SYCL] Refactor duplicated queue shortcut code (intel#12474)

45e49cf

[UR][CUDA] Dynamically load the CUPTI library when tracing (intel#11952)

acf89a6

CI testing for oneapi-src/unified-runtime#1070 --------- Co-authored-by: Kenneth Benzie (Benie) <k.benzie@codeplay.com>

reble commented Jan 24, 2024

View reviewed changes

bb-sycl and others added 3 commits January 24, 2024 08:15

[GHA] Uplift Linux GPU RT version to 23.43.27642.18 (intel#12463)

02dee98

Scheduled drivers uplift Co-authored-by: GitHub Actions <actions@github.com>

[SYCL][ESIMD][E2E] Disable matrix_transpose_glb.cpp (intel#12482)

2414a15

It's failing in postcommit, we need this to fix the CI. Signed-off-by: Sarnie, Nick <nick.sarnie@intel.com>

[SYCL][Graph] API for explicit update using indices

d7f5d51

- Adds an API to the spec for updating graph nodes using explicit indices - Also includes functionality for updating ND-range of kernel nodes - Note: Current design is only for kernel execution nodes

Bensuo force-pushed the pablo/explicit-update branch from dff51db to d7f5d51 Compare January 24, 2024 18:30

Bensuo closed this Jan 24, 2024

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[SYCL][Graph] Explicit update with index only #350

[SYCL][Graph] Explicit update with index only #350

reble commented Jan 12, 2024 •

edited

Loading

EwanC commented Jan 19, 2024

EwanC Jan 23, 2024

EwanC Jan 23, 2024

Bensuo commented Jan 23, 2024

EwanC left a comment

EwanC Jan 23, 2024

reble Jan 24, 2024

Bensuo Jan 24, 2024

Bensuo commented Jan 24, 2024

		Having a more unified approach to updating nodes/graphs may provide a more
		concise and easy to use API.

	node. Executable graphs can then be updated with these update nodes
	node.

[SYCL][Graph] Explicit update with index only #350

[SYCL][Graph] Explicit update with index only #350

Conversation

reble commented Jan 12, 2024 • edited Loading

EwanC commented Jan 19, 2024

EwanC Jan 23, 2024

Choose a reason for hiding this comment

EwanC Jan 23, 2024

Choose a reason for hiding this comment

Bensuo commented Jan 23, 2024

EwanC left a comment

Choose a reason for hiding this comment

EwanC Jan 23, 2024

Choose a reason for hiding this comment

reble Jan 24, 2024

Choose a reason for hiding this comment

Bensuo Jan 24, 2024

Choose a reason for hiding this comment

Bensuo commented Jan 24, 2024

reble commented Jan 12, 2024 •

edited

Loading