[all-commits] [llvm/llvm-project] 7fc792: [MLIR] Enable GPU Dialect to SYCL runtime integrat...

[Sang Ik Lee via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 7fc792cba7663b2aa54f259515319d74a5625be0
      https://github.com/llvm/llvm-project/commit/7fc792cba7663b2aa54f259515319d74a5625be0
  Author: Sang Ik Lee <sang.ik.lee at intel.com>
  Date:   2023-12-05 (Tue, 05 Dec 2023)

  Changed paths:
    M mlir/include/mlir/Target/LLVMIR/Dialect/All.h
    A mlir/include/mlir/Target/LLVMIR/Dialect/SPIRV/SPIRVToLLVMIRTranslation.h
    M mlir/lib/Target/LLVMIR/CMakeLists.txt
    M mlir/lib/Target/LLVMIR/Dialect/CMakeLists.txt
    M mlir/lib/Target/LLVMIR/Dialect/GPU/SelectObjectAttr.cpp
    A mlir/lib/Target/LLVMIR/Dialect/SPIRV/CMakeLists.txt
    A mlir/lib/Target/LLVMIR/Dialect/SPIRV/SPIRVToLLVMIRTranslation.cpp
    M mlir/test/CMakeLists.txt
    A mlir/test/Integration/GPU/SYCL/gpu-addf32-to-spirv.mlir
    A mlir/test/Integration/GPU/SYCL/gpu-addi64-to-spirv.mlir
    A mlir/test/Integration/GPU/SYCL/gpu-reluf32-to-spirv.mlir
    A mlir/test/Integration/GPU/SYCL/lit.local.cfg
    M mlir/test/Target/LLVMIR/gpu.mlir
    M mlir/test/lit.cfg.py
    M mlir/test/lit.site.cfg.py.in

  Log Message:
  -----------
  [MLIR] Enable GPU Dialect to SYCL runtime integration (#71430)

GPU Dialect lowering to SYCL runtime is driven by spirv.target_env
attached to gpu.module. As a result of this, spirv.target_env remains as
an input to LLVMIR Translation.
A SPIRVToLLVMIRTranslation without any actual translation is added to
avoid an unregistered error in mlir-cpu-runner.
SelectObjectAttr.cpp is updated to
1) Pass binary size argument to getModuleLoadFn
2) Pass parameter count to getKernelLaunchFn
This change does not impact CUDA and ROCM usage since both
mlir_cuda_runtime and mlir_rocm_runtime are already updated to accept
and ignore the extra arguments.