[PATCH] D118185: [AMDGPU] Mark v_cmp* convergent

[Stanislav Mekhanoshin via Phabricator via llvm-commits]

rampitec created this revision.
rampitec added reviewers: foad, ruiling, sebastian-ne, arsenm.
Herald added subscribers: kerbowa, hiraditya, t-tye, tpr, dstuttard, yaxunl, nhaehnle, jvesely, kzhuravl, MatzeB.
rampitec requested review of this revision.
Herald added a subscriber: wdng.
Herald added a project: LLVM.

This replaces readsExecAsData check with the convergent attribute
to prevent hoisting or sinking which seems to be illegal for vector
compares.


https://reviews.llvm.org/D118185

Files:
  llvm/lib/Target/AMDGPU/SIInstrInfo.cpp
  llvm/lib/Target/AMDGPU/VOPCInstructions.td
  llvm/test/CodeGen/AMDGPU/undefined-subreg-liverange.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D118185.403023.patch
Type: text/x-patch
Size: 4183 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20220125/5717f781/attachment-0001.bin>