[PATCH] D83850: [AMDGPU] Translate s_and/s_andn2 to s_mov in vcc optimisation

[Carl Ritson via Phabricator via llvm-commits]

critson created this revision.
critson added reviewers: rampitec, arsenm.
Herald added subscribers: llvm-commits, kerbowa, hiraditya, t-tye, tpr, dstuttard, yaxunl, nhaehnle, wdng, jvesely, kzhuravl.
Herald added a project: LLVM.

When SCC is dead, but VCC is required then replace s_and / s_andn2
with s_mov into VCC when mask value is 0 or -1.


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D83850

Files:
  llvm/lib/Target/AMDGPU/SIPreEmitPeephole.cpp
  llvm/test/CodeGen/AMDGPU/infinite-loop.ll
  llvm/test/CodeGen/AMDGPU/insert-skip-from-vcc.mir

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D83850.278073.patch
Type: text/x-patch
Size: 4693 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20200715/a8ed0a3c/attachment.bin>