[all-commits] [llvm/llvm-project] 8fc052: [mlir][linalg] Stage application of pad tensor op ...

[Tobias Gysi via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 8fc0525a159b3be3c83f8d80c03585237ea40a73
      https://github.com/llvm/llvm-project/commit/8fc0525a159b3be3c83f8d80c03585237ea40a73
  Author: gysit <gysit at google.com>
  Date:   2021-12-13 (Mon, 13 Dec 2021)

  Changed paths:
    M mlir/lib/Dialect/Linalg/Transforms/LinalgStrategyPasses.cpp
    M mlir/test/Dialect/Linalg/codegen-strategy.mlir
    M mlir/test/lib/Dialect/Linalg/TestLinalgCodegenStrategy.cpp

  Log Message:
  -----------
  [mlir][linalg] Stage application of pad tensor op vectoriztaion.

Adapt the LinalgStrategyVectorizationPattern pass to apply the vectorization patterns in two stages. The change ensures the generic pad tensor op vectorization pattern does not run too early. Additionally, the revision adds the transfer op canonicalization patterns to the set of applied patterns, since they are needed to enable efficient vectorization for rank-reduced convolutions.

Reviewed By: nicolasvasilache

Differential Revision: https://reviews.llvm.org/D115627