[all-commits] [llvm/llvm-project] 894a59: [mlir][nvgpu] Move mma.sync and ldmatrix in nvgpu ...

[Thomas via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 894a591cf6fc542e6fc5d84222c839495a3d832f
      https://github.com/llvm/llvm-project/commit/894a591cf6fc542e6fc5d84222c839495a3d832f
  Author: Thomas Raoux <thomasraoux at google.com>
  Date:   2022-04-14 (Thu, 14 Apr 2022)

  Changed paths:
    A mlir/include/mlir/Conversion/NVGPUToNVVM/NVGPUToNVVM.h
    M mlir/include/mlir/Conversion/Passes.h
    M mlir/include/mlir/Conversion/Passes.td
    M mlir/include/mlir/Dialect/GPU/GPUOps.td
    M mlir/include/mlir/Dialect/NVGPU/NVGPU.td
    M mlir/lib/Conversion/CMakeLists.txt
    M mlir/lib/Conversion/GPUToNVVM/LowerGpuOpsToNVVMOps.cpp
    A mlir/lib/Conversion/NVGPUToNVVM/CMakeLists.txt
    A mlir/lib/Conversion/NVGPUToNVVM/NVGPUToNVVM.cpp
    R mlir/test/Conversion/GPUToNVVM/mma-sync-to-nvvm.mlir
    A mlir/test/Conversion/NVGPUToNVVM/mma-sync-to-nvvm.mlir
    M mlir/test/Dialect/NVGPU/roundtrip.mlir

  Log Message:
  -----------
  [mlir][nvgpu] Move mma.sync and ldmatrix in nvgpu dialect

Move gpu operation mma.sync and ldmatrix in nvgpu as they are specific
to nvidia target.

Differential Revision: https://reviews.llvm.org/D123824