[all-commits] [llvm/llvm-project] 3b0993: AMDGPU: Fix treating unknown mem operands as unifo...

[Matt Arsenault via All-commits]

  Branch: refs/heads/users/arsenm/arsenm/fix-treating-unknown-mmo-uniform-2
  Home:   https://github.com/llvm/llvm-project
  Commit: 3b09937869dc420a740a8df80fbc424b4639f2fe
      https://github.com/llvm/llvm-project/commit/3b09937869dc420a740a8df80fbc424b4639f2fe
  Author: Matt Arsenault <Matthew.Arsenault at amd.com>
  Date:   2025-12-02 (Tue, 02 Dec 2025)

  Changed paths:
    M llvm/lib/Target/AMDGPU/AMDGPUISelDAGToDAG.cpp
    M llvm/lib/Target/AMDGPU/AMDGPUInstrInfo.cpp
    M llvm/lib/Target/AMDGPU/AMDGPULegalizerInfo.cpp
    M llvm/test/CodeGen/AMDGPU/GlobalISel/extractelement.i128.ll
    M llvm/test/CodeGen/AMDGPU/GlobalISel/implicit-kernarg-backend-usage-global-isel.ll

  Log Message:
  -----------
  AMDGPU: Fix treating unknown mem operands as uniform (#168980)

The test changes are mostly GlobalISel specific regressions.
GlobalISel is still relying on isUniformMMO, but it doesn't really
have an excuse for doing so. These should be avoidable with new
regbankselect.

There is an additional regression for addrspacecast for cov4. We
probably ought to be using a separate PseudoSourceValue for the
access of the queue pointer.



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