[llvm-branch-commits] [llvm] a9430a1 - AMDGPU: Put inexpensive ops first in AMDGPUAnnotateUniformValues::visitLoadInst

[Hans Wennborg via llvm-branch-commits]

Author: Changpeng Fang
Date: 2020-08-03T16:01:25+02:00
New Revision: a9430a1c9e9c99151361374f0462d751457fa15c

URL: https://github.com/llvm/llvm-project/commit/a9430a1c9e9c99151361374f0462d751457fa15c
DIFF: https://github.com/llvm/llvm-project/commit/a9430a1c9e9c99151361374f0462d751457fa15c.diff

LOG: AMDGPU: Put inexpensive ops first in AMDGPUAnnotateUniformValues::visitLoadInst

Summary:
  This is in response to the review of https://reviews.llvm.org/D84873:
The expensive check should be reordered last

Reviewers:
  arsenm

Differential Revision:
  https://reviews.llvm.org/D84890

(cherry picked from commit 243376cdc7b719d443f42c8c4667e5d96af53dcc)

Added: 
    

Modified: 
    llvm/lib/Target/AMDGPU/AMDGPUAnnotateUniformValues.cpp

Removed: 
    


################################################################################
diff  --git a/llvm/lib/Target/AMDGPU/AMDGPUAnnotateUniformValues.cpp b/llvm/lib/Target/AMDGPU/AMDGPUAnnotateUniformValues.cpp
index b09e92c07f9b..45f515c5115e 100644
--- a/llvm/lib/Target/AMDGPU/AMDGPUAnnotateUniformValues.cpp
+++ b/llvm/lib/Target/AMDGPU/AMDGPUAnnotateUniformValues.cpp
@@ -131,10 +131,20 @@ void AMDGPUAnnotateUniformValues::visitLoadInst(LoadInst &I) {
   // We're tracking up to the Function boundaries, and cannot go beyond because
   // of FunctionPass restrictions. We can ensure that is memory not clobbered
   // for memory operations that are live in to entry points only.
-  bool NotClobbered = isEntryFunc && !isClobberedInFunction(&I);
   Instruction *PtrI = dyn_cast<Instruction>(Ptr);
-  if (!PtrI && NotClobbered && isGlobalLoad(I)) {
-    if (isa<Argument>(Ptr) || isa<GlobalValue>(Ptr)) {
+
+  if (!isEntryFunc) {
+    if (PtrI)
+      setUniformMetadata(PtrI);
+    return;
+  }
+
+  bool NotClobbered = false;
+  if (PtrI)
+    NotClobbered = !isClobberedInFunction(&I);
+  else if (isa<Argument>(Ptr) || isa<GlobalValue>(Ptr)) {
+    if (isGlobalLoad(I) && !isClobberedInFunction(&I)) {
+      NotClobbered = true;
       // Lookup for the existing GEP
       if (noClobberClones.count(Ptr)) {
         PtrI = noClobberClones[Ptr];