[all-commits] [llvm/llvm-project] 8cfd9b: [MLIR] Make generic skip packing init operand when...

[Zhuoran Yin via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 8cfd9b88215acd1bff339c0d4ed60d688dcbcfdd
      https://github.com/llvm/llvm-project/commit/8cfd9b88215acd1bff339c0d4ed60d688dcbcfdd
  Author: Zhuoran Yin <zhuoryin at amd.com>
  Date:   2025-07-01 (Tue, 01 Jul 2025)

  Changed paths:
    M mlir/lib/Dialect/Linalg/Transforms/DataLayoutPropagation.cpp
    M mlir/test/Dialect/Linalg/data-layout-propagation.mlir

  Log Message:
  -----------
  [MLIR] Make generic skip packing init operand when not used in DataLayoutPropagation (#146139)

In both `bubbleUpPackOpThroughGenericOp()` or
`pushDownUnPackOpThroughGenericOp()`, we can simplify the lowered IR by
removing the pack of an empty when the init tensor isn't used in generic
op. Instead of packing an empty tensor, the empty tensor can be
forwarded to the generic output. This allows cleaner result after data
layout propagation.



To unsubscribe from these emails, change your notification settings at https://github.com/llvm/llvm-project/settings/notifications