[PATCH] D128800: [AMDGPU] Fix liveness for loops in si-optimize-exec-masking-pre-ra

[Matt Arsenault via Phabricator via llvm-commits]

arsenm accepted this revision.
arsenm added a comment.
This revision is now accepted and ready to land.

I think D128315 <https://reviews.llvm.org/D128315> D128110 <https://reviews.llvm.org/D128110> would also avoid this. I assumed cndmask in a different block was not useful, but presumably you found this in the wild?



================
Comment at: llvm/lib/Target/AMDGPU/SIOptimizeExecMaskingPreRA.cpp:218
+    // where the result of the v_cndmask_b32 was live (e.g. loops).
+    // This could yield worse register allocation is rare edge cases.
+    SlotIndex EndIdx = AndIdx.getRegSlot();
----------------
s/is/in


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D128800/new/

https://reviews.llvm.org/D128800