Vendor import of llvm-project main llvmorg-15-init-15358-g53dc0f107877.vendor/llvm-project/llvmorg-15-init-15358-g53dc0f107877

1 files changed, 44 insertions, 20 deletions

diff --git a/llvm/lib/Target/AMDGPU/AMDGPUPromoteKernelArguments.cpp b/llvm/lib/Target/AMDGPU/AMDGPUPromoteKernelArguments.cpp
index 01d03d17ec47..ed450f59e4b3 100644
--- a/llvm/lib/Target/AMDGPU/AMDGPUPromoteKernelArguments.cpp
+++ b/llvm/lib/Target/AMDGPU/AMDGPUPromoteKernelArguments.cpp
@@ -16,7 +16,9 @@
 //===----------------------------------------------------------------------===//
 
 #include "AMDGPU.h"
+#include "Utils/AMDGPUMemoryUtils.h"
 #include "llvm/ADT/SmallVector.h"
+#include "llvm/Analysis/AliasAnalysis.h"
 #include "llvm/Analysis/MemorySSA.h"
 #include "llvm/IR/IRBuilder.h"
 #include "llvm/InitializePasses.h"
@@ -30,6 +32,8 @@ namespace {
 class AMDGPUPromoteKernelArguments : public FunctionPass {
   MemorySSA *MSSA;
 
+  AliasAnalysis *AA;
+
   Instruction *ArgCastInsertPt;
 
   SmallVector<Value *> Ptrs;
@@ -38,16 +42,19 @@ class AMDGPUPromoteKernelArguments : public FunctionPass {
 
   bool promotePointer(Value *Ptr);
 
+  bool promoteLoad(LoadInst *LI);
+
 public:
   static char ID;
 
   AMDGPUPromoteKernelArguments() : FunctionPass(ID) {}
 
-  bool run(Function &F, MemorySSA &MSSA);
+  bool run(Function &F, MemorySSA &MSSA, AliasAnalysis &AA);
 
   bool runOnFunction(Function &F) override;
 
   void getAnalysisUsage(AnalysisUsage &AU) const override {
+    AU.addRequired<AAResultsWrapperPass>();
     AU.addRequired<MemorySSAWrapperPass>();
     AU.setPreservesAll();
   }
@@ -68,17 +75,10 @@ void AMDGPUPromoteKernelArguments::enqueueUsers(Value *Ptr) {
       break;
     case Instruction::Load: {
       LoadInst *LD = cast<LoadInst>(U);
-      PointerType *PT = dyn_cast<PointerType>(LD->getType());
-      if (!PT ||
-          (PT->getAddressSpace() != AMDGPUAS::FLAT_ADDRESS &&
-           PT->getAddressSpace() != AMDGPUAS::GLOBAL_ADDRESS &&
-           PT->getAddressSpace() != AMDGPUAS::CONSTANT_ADDRESS) ||
-          LD->getPointerOperand()->stripInBoundsOffsets() != Ptr)
-        break;
-      const MemoryAccess *MA = MSSA->getWalker()->getClobberingMemoryAccess(LD);
-      // TODO: This load poprobably can be promoted to constant address space.
-      if (MSSA->isLiveOnEntryDef(MA))
+      if (LD->getPointerOperand()->stripInBoundsOffsets() == Ptr &&
+          !AMDGPU::isClobberedInFunction(LD, MSSA, AA))
         Ptrs.push_back(LD);
+
       break;
     }
     case Instruction::GetElementPtr:
@@ -92,15 +92,26 @@ void AMDGPUPromoteKernelArguments::enqueueUsers(Value *Ptr) {
 }
 
 bool AMDGPUPromoteKernelArguments::promotePointer(Value *Ptr) {
-  enqueueUsers(Ptr);
+  bool Changed = false;
+
+  LoadInst *LI = dyn_cast<LoadInst>(Ptr);
+  if (LI)
+    Changed |= promoteLoad(LI);
+
+  PointerType *PT = dyn_cast<PointerType>(Ptr->getType());
+  if (!PT)
+    return Changed;
+
+  if (PT->getAddressSpace() == AMDGPUAS::FLAT_ADDRESS ||
+      PT->getAddressSpace() == AMDGPUAS::GLOBAL_ADDRESS ||
+      PT->getAddressSpace() == AMDGPUAS::CONSTANT_ADDRESS)
+    enqueueUsers(Ptr);
 
-  PointerType *PT = cast<PointerType>(Ptr->getType());
   if (PT->getAddressSpace() != AMDGPUAS::FLAT_ADDRESS)
-    return false;
+    return Changed;
 
-  bool IsArg = isa<Argument>(Ptr);
-  IRBuilder<> B(IsArg ? ArgCastInsertPt
-                      : &*std::next(cast<Instruction>(Ptr)->getIterator()));
+  IRBuilder<> B(LI ? &*std::next(cast<Instruction>(Ptr)->getIterator())
+                   : ArgCastInsertPt);
 
   // Cast pointer to global address space and back to flat and let
   // Infer Address Spaces pass to do all necessary rewriting.
@@ -116,6 +127,14 @@ bool AMDGPUPromoteKernelArguments::promotePointer(Value *Ptr) {
   return true;
 }
 
+bool AMDGPUPromoteKernelArguments::promoteLoad(LoadInst *LI) {
+  if (!LI->isSimple())
+    return false;
+
+  LI->setMetadata("amdgpu.noclobber", MDNode::get(LI->getContext(), {}));
+  return true;
+}
+
 // skip allocas
 static BasicBlock::iterator getInsertPt(BasicBlock &BB) {
   BasicBlock::iterator InsPt = BB.getFirstInsertionPt();
@@ -131,7 +150,8 @@ static BasicBlock::iterator getInsertPt(BasicBlock &BB) {
   return InsPt;
 }
 
-bool AMDGPUPromoteKernelArguments::run(Function &F, MemorySSA &MSSA) {
+bool AMDGPUPromoteKernelArguments::run(Function &F, MemorySSA &MSSA,
+                                       AliasAnalysis &AA) {
   if (skipFunction(F))
     return false;
 
@@ -141,6 +161,7 @@ bool AMDGPUPromoteKernelArguments::run(Function &F, MemorySSA &MSSA) {
 
   ArgCastInsertPt = &*getInsertPt(*F.begin());
   this->MSSA = &MSSA;
+  this->AA = &AA;
 
   for (Argument &Arg : F.args()) {
     if (Arg.use_empty())
@@ -166,11 +187,13 @@ bool AMDGPUPromoteKernelArguments::run(Function &F, MemorySSA &MSSA) {
 
 bool AMDGPUPromoteKernelArguments::runOnFunction(Function &F) {
   MemorySSA &MSSA = getAnalysis<MemorySSAWrapperPass>().getMSSA();
-  return run(F, MSSA);
+  AliasAnalysis &AA = getAnalysis<AAResultsWrapperPass>().getAAResults();
+  return run(F, MSSA, AA);
 }
 
 INITIALIZE_PASS_BEGIN(AMDGPUPromoteKernelArguments, DEBUG_TYPE,
                       "AMDGPU Promote Kernel Arguments", false, false)
+INITIALIZE_PASS_DEPENDENCY(AAResultsWrapperPass)
 INITIALIZE_PASS_DEPENDENCY(MemorySSAWrapperPass)
 INITIALIZE_PASS_END(AMDGPUPromoteKernelArguments, DEBUG_TYPE,
                     "AMDGPU Promote Kernel Arguments", false, false)
@@ -185,7 +208,8 @@ PreservedAnalyses
 AMDGPUPromoteKernelArgumentsPass::run(Function &F,
                                       FunctionAnalysisManager &AM) {
   MemorySSA &MSSA = AM.getResult<MemorySSAAnalysis>(F).getMSSA();
-  if (AMDGPUPromoteKernelArguments().run(F, MSSA)) {
+  AliasAnalysis &AA = AM.getResult<AAManager>(F);
+  if (AMDGPUPromoteKernelArguments().run(F, MSSA, AA)) {
     PreservedAnalyses PA;
     PA.preserveSet<CFGAnalyses>();
     PA.preserve<MemorySSAAnalysis>();