Unified Runtime

Table of contents

• Unified Runtime
  • Table of contents
  • Contents of the repo
  • Integration
    • Weekly tags
  • Third-Party tools
  • Building
    • Requirements
    • Windows
    • Linux
    • CMake standard options
    • Additional make targets
  • Contributions
    • Adapter naming convention
    • Source code generation
    • Documentation
  • Release Process

Contents of the repo

This repo contains the following:

• API specification in YaML
• API programming guide in RST
• Loader and a null adapter implementation (partially generated)
• Example applications
• API C/C++ header files (generated)
• API Python module (generated)
• Sample C++ wrapper (generated)
• Sample C/C++ import library (generated)

Integration

The recommended way to integrate this project into another is via CMake's FetchContent, for example:

include(FetchContent)

FetchContent_Declare(
    unified-runtime
    GIT_REPOSITORY https://github.com/oneapi-src/unified-runtime.git
    GIT_TAG main  # This will pull the latest changes from the main branch.
)
FetchContent_MakeAvailable(unified-runtime)

add_executable(example example.cpp)
target_link_libraries(example PUBLIC unified-runtime::headers)

Weekly tags

Each Friday at 23:00 UTC time a prerelease tag is created which takes the form weekly-YYYY-MM-DD. These tags should be used by downstream projects which intend to track development closely but maintain a fixed point in history to avoid pulling potentially breaking changes from the main branch.

Third-Party tools

The recommended method to install the third-party tools is using a Python virtual environment, for example:

$ python -m venv .venv
$ source .venv/bin/activate
$ pip install -r third_party/requirements.txt

Building

The requirements and instructions below are for building the project from source without any modifications. To make modifications to the specification, please see the Contribution Guide for more detailed instructions on the correct setup.

Requirements

Required packages:

• C++ compiler with C++17 support
• CMake >= 3.20.0
• Python v3.6.6 or later

Windows

Generating Visual Studio Project. EXE and binaries will be in build/bin/{build_config}

$ mkdir build
$ cd build
$ cmake {path_to_source_dir} -G "Visual Studio 15 2017 Win64"

Linux

Executable and binaries will be in build/bin

$ mkdir build
$ cd build
$ cmake {path_to_source_dir}
$ make

CMake standard options

List of options provided by CMake:

Name	Description	Values	Default
UR_BUILD_EXAMPLES	Build example applications	ON/OFF	ON
UR_BUILD_TESTS	Build the tests	ON/OFF	ON
UR_BUILD_TOOLS	Build tools	ON/OFF	ON
UR_FORMAT_CPP_STYLE	Format code style	ON/OFF	OFF
UR_DEVELOPER_MODE	Treat warnings as errors	ON/OFF	OFF
UR_ENABLE_FAST_SPEC_MODE	Enable fast specification generation mode	ON/OFF	OFF
UR_USE_ASAN	Enable AddressSanitizer	ON/OFF	OFF
UR_USE_TSAN	Enable ThreadSanitizer	ON/OFF	OFF
UR_USE_UBSAN	Enable UndefinedBehavior Sanitizer	ON/OFF	OFF
UR_USE_MSAN	Enable MemorySanitizer (clang only)	ON/OFF	OFF
UR_USE_CFI	Enable Control Flow Integrity checks (clang only, also enables lto)	ON/OFF	OFF
UR_ENABLE_TRACING	Enable XPTI-based tracing layer	ON/OFF	OFF
UR_ENABLE_SANITIZER	Enable device sanitizer layer	ON/OFF	ON
UR_CONFORMANCE_TARGET_TRIPLES	SYCL triples to build CTS device binaries for	Comma-separated list	spir64
UR_CONFORMANCE_AMD_ARCH	AMD device target ID to build CTS binaries for	string	""
UR_CONFORMANCE_ENABLE_MATCH_FILES	Enable CTS match files	ON/OFF	ON
UR_CONFORMANCE_TEST_LOADER	Additionally build and run "loader" tests for the CTS	ON/OFF	OFF
UR_BUILD_ADAPTER_L0	Build the Level-Zero adapter	ON/OFF	OFF
UR_BUILD_ADAPTER_OPENCL	Build the OpenCL adapter	ON/OFF	OFF
UR_BUILD_ADAPTER_CUDA	Build the CUDA adapter	ON/OFF	OFF
UR_BUILD_ADAPTER_HIP	Build the HIP adapter	ON/OFF	OFF
UR_BUILD_ADAPTER_NATIVE_CPU	Build the Native-CPU adapter	ON/OFF	OFF
UR_BUILD_ADAPTER_ALL	Build all currently supported adapters	ON/OFF	OFF
UR_BUILD_ADAPTER_L0_V2	Build the (experimental) Level-Zero v2 adapter	ON/OFF	OFF
UR_STATIC_ADAPTER_L0	Build the Level-Zero adapter as static and embed in the loader	ON/OFF	OFF
UR_HIP_PLATFORM	Build HIP adapter for AMD or NVIDIA platform	AMD/NVIDIA	AMD
UR_ENABLE_COMGR	Enable comgr lib usage	AMD/NVIDIA	AMD
UR_DPCXX	Path of the DPC++ compiler executable to build CTS device binaries	File path	""
UR_DEVICE_CODE_EXTRACTOR	Path of the clang-offload-extract executable from the DPC++ package, required for CTS device binaries	File path	"${dirname(UR_DPCXX)}/clang-offload-extract"
UR_DPCXX_BUILD_FLAGS	Build flags to pass to DPC++ when compiling device programs	Space-separated options list	""
UR_SYCL_LIBRARY_DIR	Path of the SYCL runtime library directory to build CTS device binaries	Directory path	""
UR_HIP_ROCM_DIR	Path of the default ROCm HIP installation	Directory path	$ENV{ROCM_PATH} or /opt/rocm
UR_HIP_INCLUDE_DIR	Path of the ROCm HIP include directory	Directory path	${UR_HIP_ROCM_DIR}/include
UR_HIP_HSA_INCLUDE_DIRS	Path of the ROCm HSA include directory	Directory path	${UR_HIP_ROCM_DIR}/hsa/include;${UR_HIP_ROCM_DIR}/include
UR_HIP_LIB_DIR	Path of the ROCm HIP library directory	Directory path	${UR_HIP_ROCM_DIR}/lib

Additional make targets

To run tests, do the following:

$ make
$ make test

To run automated code formatting, configure CMake with UR_FORMAT_CPP_STYLE option and then run a custom cppformat target:

$ make cppformat

If you've made modifications to the specification, you can also run a custom generate target prior to building. It will generate the source code and run automated code formatting:

$ make generate

This target has additional dependencies which are described in the Build Environment section of the Contribution Guide.

Contributions

For those who intend to make a contribution to the project please read our Contribution Guide for more information.

Adapter naming convention

To maintain consistency and clarity in naming adapter libraries, it is recommended to use the following naming convention:

• On Linux platforms, use libur_adapter_[name].so.
• On Windows platforms, use ur_adapter_[name].dll.

Source code generation

Code is generated using included Python scripts.

Documentation

Documentation is generated from source code using Sphinx - see scripts dir for details.

Release Process

Unified Runtime releases are aligned with oneAPI releases. Once all changes planned for a release have been accepted, the release process is defined as:

1. Create a new release branch based on the main branch taking the form v<major>.<minor>.x where x is a placeholder for the patch version. This branch will always contain the latest patch version for a given release.
2. Create a PR to increment the CMake project version on the main and merge before accepting any other changes.
3. Create a new tag based on the latest commit on the release branch taking the form v<major>.<minor>.<patch>.
4. Create a new GitHub release using the tag created in the previous step.
   • Prior to version 1.0, check the Set as a pre-release tick box.
5. Update downstream projects to utilize the release tag. If any issues arise from integration, apply any necessary hot fixes to v<major>.<minor>.x branch and go back to step 3.