[all-commits] [llvm/llvm-project] c0321e: [mlir][gpu] Adding support for transposed mma_load...

[Quinn Dawkins via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: c0321edc26a7c02b46ed18c8f8bdcaaa9d9ce8a2
      https://github.com/llvm/llvm-project/commit/c0321edc26a7c02b46ed18c8f8bdcaaa9d9ce8a2
  Author: Quinn Dawkins <quinn at nod-labs.com>
  Date:   2022-11-29 (Tue, 29 Nov 2022)

  Changed paths:
    M mlir/include/mlir/Dialect/GPU/IR/GPUOps.td
    M mlir/lib/Conversion/GPUToNVVM/WmmaOpsToNvvm.cpp
    M mlir/lib/Conversion/GPUToSPIRV/WmmaOpsToSPIRV.cpp
    M mlir/lib/Conversion/VectorToGPU/VectorToGPU.cpp
    M mlir/test/Conversion/VectorToGPU/vector-to-mma-ops.mlir

  Log Message:
  -----------
  [mlir][gpu] Adding support for transposed mma_load_matrix

Enables transposed gpu.subgroup_mma_load_matrix and updates the lowerings in Vector to GPU and GPU to SPIRV. Needed to enable B transpose matmuls lowering to wmma ops.

Taken over from author: stanley-nod <stanley at nod-labs.com>

Reviewed By: ThomasRaoux, antiagainst

Differential Revision: https://reviews.llvm.org/D138770