[all-commits] [llvm/llvm-project] 95c64b: AMDGPU: Reduce readfirstlane for single demanded v...

[Matt Arsenault via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 95c64b7ee6158a8a4e90638af383ab8826b03a14
      https://github.com/llvm/llvm-project/commit/95c64b7ee6158a8a4e90638af383ab8826b03a14
  Author: Matt Arsenault <Matthew.Arsenault at amd.com>
  Date:   2025-03-05 (Wed, 05 Mar 2025)

  Changed paths:
    M llvm/lib/Target/AMDGPU/AMDGPUInstCombineIntrinsic.cpp
    M llvm/lib/Target/AMDGPU/AMDGPUTargetTransformInfo.h
    M llvm/test/Transforms/InstCombine/AMDGPU/simplify-demanded-vector-elts-lane-intrinsics.ll

  Log Message:
  -----------
  AMDGPU: Reduce readfirstlane for single demanded vector element (#128647)

If we are only extracting a single element, rewrite the intrinsic call
to use the element type. We should extend this to arbitrary extract
shuffles.



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