[llvm] r347573 - AMDGPU: Don't optimize exec masks at -O0
Matt Arsenault via llvm-commits
llvm-commits at lists.llvm.org
Mon Nov 26 09:02:03 PST 2018
Author: arsenm
Date: Mon Nov 26 09:02:02 2018
New Revision: 347573
URL: http://llvm.org/viewvc/llvm-project?rev=347573&view=rev
Log:
AMDGPU: Don't optimize exec masks at -O0
Modified:
llvm/trunk/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp
Modified: llvm/trunk/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp?rev=347573&r1=347572&r2=347573&view=diff
==============================================================================
--- llvm/trunk/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp (original)
+++ llvm/trunk/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp Mon Nov 26 09:02:02 2018
@@ -880,7 +880,8 @@ void GCNPassConfig::addOptimizedRegAlloc
void GCNPassConfig::addPostRegAlloc() {
addPass(&SIFixVGPRCopiesID);
- addPass(&SIOptimizeExecMaskingID);
+ if (getOptLevel() > CodeGenOpt::None)
+ addPass(&SIOptimizeExecMaskingID);
TargetPassConfig::addPostRegAlloc();
}
More information about the llvm-commits
mailing list