[all-commits] [llvm/llvm-project] 1ac2d1: [mlir][linalg] Add canonicalizers for depthwise conv

[rsuderman via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 1ac2d195ecb5d4c549c11b9c1df00179f5fea7ed
      https://github.com/llvm/llvm-project/commit/1ac2d195ecb5d4c549c11b9c1df00179f5fea7ed
  Author: Rob Suderman <rob.suderman at gmail.com>
  Date:   2021-09-15 (Wed, 15 Sep 2021)

  Changed paths:
    M mlir/lib/Dialect/Linalg/IR/LinalgOps.cpp
    M mlir/test/Dialect/Linalg/canonicalize.mlir

  Log Message:
  -----------
  [mlir][linalg] Add canonicalizers for depthwise conv

There are two main versions of depthwise conv depending whether the multiplier
is 1 or not. In cases where m == 1 we should use the version without the
multiplier channel as it can perform greater optimization.

Add lowering for the quantized/float versions to have a multiplier of one.

Reviewed By: antiagainst

Differential Revision: https://reviews.llvm.org/D108959