[PATCH] D43249: [AMDGPU] Remove non-temporal flag from argument loads

Stanislav Mekhanoshin via Phabricator via llvm-commits llvm-commits at lists.llvm.org
Tue Feb 13 11:21:11 PST 2018


rampitec created this revision.
rampitec added reviewers: t-tye, arsenm, kzhuravl.
Herald added subscribers: tpr, dstuttard, yaxunl, nhaehnle, wdng.

Kernel arguments likely read by all workitems and should not bypass
cache. Fixes performance hit in sub-dword argument loads.


https://reviews.llvm.org/D43249

Files:
  lib/Target/AMDGPU/SIISelLowering.cpp
  test/CodeGen/AMDGPU/kernel-args.ll
  test/CodeGen/AMDGPU/llvm.amdgcn.buffer.store.format.d16.ll
  test/CodeGen/AMDGPU/llvm.amdgcn.image.d16.ll
  test/CodeGen/AMDGPU/llvm.amdgcn.tbuffer.store.d16.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D43249.134087.patch
Type: text/x-patch
Size: 8964 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20180213/550d23ea/attachment.bin>


More information about the llvm-commits mailing list