[all-commits] [llvm/llvm-project] d05634: [VectorCombine] Fold bitwise operations of bitcast...

[Narayan via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: d05634d5cdc55f0a637259b10941a8591aa665e3
      https://github.com/llvm/llvm-project/commit/d05634d5cdc55f0a637259b10941a8591aa665e3
  Author: Narayan <nsreekumar6 at gmail.com>
  Date:   2025-06-26 (Thu, 26 Jun 2025)

  Changed paths:
    M llvm/lib/Transforms/Vectorize/VectorCombine.cpp
    M llvm/test/Transforms/PhaseOrdering/X86/blendv-select.ll
    M llvm/test/Transforms/VectorCombine/AArch64/shrink-types.ll

  Log Message:
  -----------
  [VectorCombine] Fold bitwise operations of bitcasts into bitcast of bitwise operation (#137322)

Currently, LLVM fails to convert certain pblendvb intrinsics into select
instructions when the blend mask is derived from complex boolean logic
operations. This occurs even when the mask is ultimately based on
sign-extended comparison results, preventing further optimization
opportunities.

Fixes #66513

---------

Co-authored-by: Simon Pilgrim <llvm-dev at redking.me.uk>



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