aboutsummaryrefslogtreecommitdiff
path: root/llvm/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp
diff options
context:
space:
mode:
Diffstat (limited to 'llvm/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp')
-rw-r--r--llvm/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp7
1 files changed, 4 insertions, 3 deletions
diff --git a/llvm/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp b/llvm/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp
index 6214f4d..75a94ac 100644
--- a/llvm/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp
+++ b/llvm/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp
@@ -619,6 +619,7 @@ extern "C" LLVM_ABI LLVM_EXTERNAL_VISIBILITY void LLVMInitializeAMDGPUTarget() {
initializeAMDGPUPreloadKernArgPrologLegacyPass(*PR);
initializeAMDGPUWaitSGPRHazardsLegacyPass(*PR);
initializeAMDGPUPreloadKernelArgumentsLegacyPass(*PR);
+ initializeAMDGPUUniformIntrinsicCombineLegacyPass(*PR);
}
static std::unique_ptr<TargetLoweringObjectFile> createTLOF(const Triple &TT) {
@@ -887,9 +888,6 @@ void AMDGPUTargetMachine::registerPassBuilderCallbacks(PassBuilder &PB) {
if (EarlyInlineAll && !EnableFunctionCalls)
PM.addPass(AMDGPUAlwaysInlinePass());
-
- if (EnableUniformIntrinsicCombine)
- PM.addPass(AMDGPUUniformIntrinsicCombinePass());
});
PB.registerPeepholeEPCallback(
@@ -900,6 +898,9 @@ void AMDGPUTargetMachine::registerPassBuilderCallbacks(PassBuilder &PB) {
FPM.addPass(AMDGPUUseNativeCallsPass());
if (EnableLibCallSimplify)
FPM.addPass(AMDGPUSimplifyLibCallsPass());
+
+ if (EnableUniformIntrinsicCombine)
+ FPM.addPass(AMDGPUUniformIntrinsicCombinePass());
});
PB.registerCGSCCOptimizerLateEPCallback(