[PATCH] D25998: AMDGPU/SI: Don't use non-0 waitcnt values when waiting on Flat instructions

[Tom Stellard via llvm-commits]

tstellarAMD updated this revision to Diff 76033.
tstellarAMD added a comment.

Only treat flat operations as unordered if they access the flat address space.


https://reviews.llvm.org/D25998

Files:
  lib/Target/AMDGPU/SIInsertWaits.cpp
  lib/Target/AMDGPU/SIInstrInfo.cpp
  lib/Target/AMDGPU/SIInstrInfo.h
  test/CodeGen/MIR/AMDGPU/waitcnt.mir

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D25998.76033.patch
Type: text/x-patch
Size: 4885 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20161027/c39838b9/attachment.bin>