[all-commits] [llvm/llvm-project] c657a6: [AMDGPU] Fix selection of s_load_b96 on GFX11 (#10...

[Jay Foad via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: c657a6f6aa7e802d65aba84b8b3fe2eb5e2459d8
      https://github.com/llvm/llvm-project/commit/c657a6f6aa7e802d65aba84b8b3fe2eb5e2459d8
  Author: Jay Foad <jay.foad at amd.com>
  Date:   2024-09-12 (Thu, 12 Sep 2024)

  Changed paths:
    M llvm/lib/Target/AMDGPU/SIISelLowering.cpp
    A llvm/test/CodeGen/AMDGPU/load-constant-always-uniform.ll

  Log Message:
  -----------
  [AMDGPU] Fix selection of s_load_b96 on GFX11 (#108029)

Fix a bug which resulted in selection of s_load_b96 on GFX11, which only
exists in GFX12.

The root cause was a mismatch between legalization and selection. The
condition used to check that the load was uniform in legalization
(SITargetLowering::LowerLOAD) was "!Op->isDivergent()". The condition
used to detect a non-uniform load during selection
(AMDGPUDAGToDAGISel::isUniformLoad()) was
"N->isDivergent() && !AMDGPUInstrInfo::isUniformMMO(MMO)". This makes a
difference when IR uniformity analysis has more information than SDAG's
built in analysis. In the test case this is because IR UA reports that
everything is uniform if isSingleLaneExecution() returns true, e.g. if
the specified max flat workgroup size is 1, but SDAG does not have this
optimization.

The immediate fix is to use the same condition to detect uniform loads
in legalization and selection. In future SDAG should learn about
isSingleLaneExecution(), and then it could probably stop relying on IR
metadata to detect uniform loads.



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