[PATCH] D115747: [AMDGPU] Flush the vmcnt counter in loop preheader when necessary

[Baptiste Saleil via Phabricator via llvm-commits]

bsaleil updated this revision to Diff 412764.
bsaleil retitled this revision from "[AMDGPU] Hoist waitcnt out of loops when they unecessarily wait for stores" to "[AMDGPU] Flush the vmcnt counter in loop preheader when necessary".
bsaleil edited the summary of this revision.
bsaleil added a comment.
Herald added a project: All.

@foad I updated the patch. It is a lot simpler than the previous, and it fixes both the GFX9 and GFX10 cases. But it may still have a significant impact on compile time, I cannot think of another way to do that without visiting all the instructions from the loops :(


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D115747/new/

https://reviews.llvm.org/D115747

Files:
  llvm/lib/Target/AMDGPU/SIInsertWaitcnts.cpp
  llvm/test/CodeGen/AMDGPU/cvt_f32_ubyte.ll
  llvm/test/CodeGen/AMDGPU/llc-pipeline.ll
  llvm/test/CodeGen/AMDGPU/waitcnt-vmcnt-loop.mir

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D115747.412764.patch
Type: text/x-patch
Size: 27632 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20220303/fa87113f/attachment-0001.bin>