[all-commits] [llvm/llvm-project] cff4a0: AMDGPU: Fix runtime unrolling when cascaded GEPs p...

[macurtis-amd via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: cff4a00d3f7d91c0dd3a93eb81db66be178273d3
      https://github.com/llvm/llvm-project/commit/cff4a00d3f7d91c0dd3a93eb81db66be178273d3
  Author: macurtis-amd <macurtis at amd.com>
  Date:   2025-07-10 (Thu, 10 Jul 2025)

  Changed paths:
    M llvm/lib/Target/AMDGPU/AMDGPUTargetTransformInfo.cpp
    A llvm/test/Transforms/LoopUnroll/AMDGPU/unroll-runtime.ll

  Log Message:
  -----------
  AMDGPU: Fix runtime unrolling when cascaded GEPs present (#147700)

Cascaded GEP (i.e. GEP of GEP) are not handled when determining if it is
ok to runtime unroll loops.

This change simply uses `getUnderlyingObjects` to look through cascaded
GEPs.



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