[all-commits] [llvm/llvm-project] 203acc: [mlir][Linalg] Improve conv vectorization for the ...

[Nicolas Vasilache via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 203accf0bdde1d276646c79dfa605ee3426f1ca8
      https://github.com/llvm/llvm-project/commit/203accf0bdde1d276646c79dfa605ee3426f1ca8
  Author: Nicolas Vasilache <nicolas.vasilache at gmail.com>
  Date:   2021-10-21 (Thu, 21 Oct 2021)

  Changed paths:
    M mlir/lib/Dialect/Linalg/Transforms/Vectorization.cpp
    M mlir/test/Dialect/Linalg/vectorize-convolution.mlir

  Log Message:
  -----------
  [mlir][Linalg] Improve conv vectorization for the stride==1 case.

In the stride == 1 case, conv1d reads contiguous data along the input dimension. This can be advantageaously used to bulk memory transfers and compute while avoiding unrolling. Experimentally, this can yield speedups of up to 50%.

Reviewed By: antiagainst

Differential Revision: https://reviews.llvm.org/D112139