[all-commits] [llvm/llvm-project] fb7ef6: [mlir][vector][nvgpu] Move MMA contraction prepara...

[Jakub Kuderski via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: fb7ef637a84652dbd3d973a1ba7db9470181b5aa
      https://github.com/llvm/llvm-project/commit/fb7ef637a84652dbd3d973a1ba7db9470181b5aa
  Author: Jakub Kuderski <kubak at google.com>
  Date:   2023-03-09 (Thu, 09 Mar 2023)

  Changed paths:
    M mlir/include/mlir/Dialect/NVGPU/Utils/MMAUtils.h
    M mlir/include/mlir/Dialect/Vector/Transforms/VectorRewritePatterns.h
    M mlir/lib/Conversion/VectorToGPU/VectorToGPU.cpp
    M mlir/lib/Dialect/NVGPU/Utils/MMAUtils.cpp
    M mlir/lib/Dialect/Vector/Transforms/VectorTransforms.cpp
    A mlir/test/Dialect/Vector/vector-contract-matmul-transforms.mlir
    M mlir/test/lib/Dialect/Vector/TestVectorTransforms.cpp

  Log Message:
  -----------
  [mlir][vector][nvgpu] Move MMA contraction preparation to VectorUtils

This pattern is not specific to nvgpu; I intend to use in SPIR-V codegen. `VectorTransforms` seems like a more generally useful place.

In addition:
-  Fix a bug in the second condition (the dimensions were swapped for RHS).
-  Add tests.
-  Add support for externally provided filter functions, similar to other vector transforms.
-  Prefer to transpose before zero/sign-extending inputs.

Reviewed By: ThomasRaoux

Differential Revision: https://reviews.llvm.org/D145638