[llvm] r268095 - AMDGPU/SI: Move post regalloc run of SIShrinkInstructions
Matt Arsenault via llvm-commits
llvm-commits at lists.llvm.org
Fri Apr 29 13:23:43 PDT 2016
Author: arsenm
Date: Fri Apr 29 15:23:42 2016
New Revision: 268095
URL: http://llvm.org/viewvc/llvm-project?rev=268095&view=rev
Log:
AMDGPU/SI: Move post regalloc run of SIShrinkInstructions
Move to addPreEmitPass. This is so it runs after post-RA
scheduling so we can merge s_nops emitted by the scheduler
and hazard recognizer.
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=268095&r1=268094&r2=268095&view=diff
==============================================================================
--- llvm/trunk/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp (original)
+++ llvm/trunk/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp Fri Apr 29 15:23:42 2016
@@ -205,7 +205,6 @@ public:
void addFastRegAlloc(FunctionPass *RegAllocPass) override;
void addOptimizedRegAlloc(FunctionPass *RegAllocPass) override;
void addPreRegAlloc() override;
- void addPostRegAlloc() override;
void addPreSched2() override;
void addPreEmitPass() override;
};
@@ -381,15 +380,12 @@ void GCNPassConfig::addOptimizedRegAlloc
TargetPassConfig::addOptimizedRegAlloc(RegAllocPass);
}
-void GCNPassConfig::addPostRegAlloc() {
- addPass(createSIShrinkInstructionsPass(), false);
-}
-
void GCNPassConfig::addPreSched2() {
}
void GCNPassConfig::addPreEmitPass() {
addPass(createSIInsertWaitsPass(), false);
+ addPass(createSIShrinkInstructionsPass());
addPass(createSILowerControlFlowPass(), false);
addPass(createSIInsertNopsPass(), false);
}
More information about the llvm-commits
mailing list