[all-commits] [llvm/llvm-project] fa0d04: [mlir] Fix canonicalization of tiled_loop if not a...

[Alexander Belyaev via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: fa0d044c4499535fb7960a5b7053bd043ad09e52
      https://github.com/llvm/llvm-project/commit/fa0d044c4499535fb7960a5b7053bd043ad09e52
  Author: Alexander Belyaev <pifon at google.com>
  Date:   2021-04-28 (Wed, 28 Apr 2021)

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

  Log Message:
  -----------
  [mlir] Fix canonicalization of tiled_loop if not all opresults fold.

The current canonicalization did not remap operation results correctly
and attempted to erase tiledLoop, which is incorrect if not all tensor
results are folded.