[PATCH] D41377: [AMDGPU] Turn off MergeConsecutiveStores() before Instruction Selection for AMDGPU
Matt Arsenault via Phabricator via llvm-commits
llvm-commits at lists.llvm.org
Tue Dec 19 10:26:42 PST 2017
arsenm accepted this revision.
arsenm added a comment.
This revision is now accepted and ready to land.
LGTM with minor fixes
================
Comment at: lib/Target/AMDGPU/AMDGPUISelLowering.h:206
+ // FIXME: Turn off MergeConsecutiveStores() before Instruction Selection
+ // for AMDGPU. Commit dbbb6c5fc3642987430866dffdf710df4f616ac7 turned on
+ // MergeConsecutiveStores() before Instruction Selection for all targets.
----------------
Should refer to the svn revision since that is still the canonical name
================
Comment at: test/CodeGen/AMDGPU/memory-legalizer-store-infinite-loop.ll:14
+; GCN-LABEL: {{^}}_Z6brokenPd:
+; GCN: s_endpgm
+define weak_odr amdgpu_kernel void @_Z6brokenPd(double* %arg) {
----------------
Check the stores
================
Comment at: test/CodeGen/AMDGPU/memory-legalizer-store-infinite-loop.ll:15
+; GCN: s_endpgm
+define weak_odr amdgpu_kernel void @_Z6brokenPd(double* %arg) {
+bb:
----------------
You can remove weak_odr
https://reviews.llvm.org/D41377
More information about the llvm-commits
mailing list