[InstCombine] Iterative replacement in PtrReplacer #137215
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
llvmbot
added
backend:AMDGPU
llvm:instcombine
|
@llvm/pr-subscribers-llvm-transforms @llvm/pr-subscribers-backend-amdgpu Author: Anshil Gandhi (gandhi56) ChangesThis patch enhances the PtrReplacer as follows:
This patch specifically fixes the case when an incoming value of a PHI is addrspacecasted. So it is an NFC if the data layout is unknown. Full diff: https://github.com/llvm/llvm-project/pull/137215.diff 2 Files Affected:
diff --git a/llvm/lib/Transforms/InstCombine/InstCombineLoadStoreAlloca.cpp b/llvm/lib/Transforms/InstCombine/InstCombineLoadStoreAlloca.cpp
index c29cba6f675c5..f069685a9a2ce 100644
--- a/llvm/lib/Transforms/InstCombine/InstCombineLoadStoreAlloca.cpp
+++ b/llvm/lib/Transforms/InstCombine/InstCombineLoadStoreAlloca.cpp
@@ -244,11 +244,10 @@ class PointerReplacer {
void replacePointer(Value *V);
private:
- bool collectUsersRecursive(Instruction &I);
void replace(Instruction *I);
- Value *getReplacement(Value *I);
+ Value *getReplacement(Value *V) const { return WorkMap.lookup(V); }
bool isAvailable(Instruction *I) const {
- return I == &Root || Worklist.contains(I);
+ return I == &Root || UsersToReplace.contains(I);
}
bool isEqualOrValidAddrSpaceCast(const Instruction *I,
@@ -260,8 +259,7 @@ class PointerReplacer {
return (FromAS == ToAS) || IC.isValidAddrSpaceCast(FromAS, ToAS);
}
- SmallPtrSet<Instruction *, 32> ValuesToRevisit;
- SmallSetVector<Instruction *, 4> Worklist;
+ SmallSetVector<Instruction *, 4> UsersToReplace;
MapVector<Value *, Value *> WorkMap;
InstCombinerImpl &IC;
Instruction &Root;
@@ -270,80 +268,131 @@ class PointerReplacer {
} // end anonymous namespace
bool PointerReplacer::collectUsers() {
- if (!collectUsersRecursive(Root))
- return false;
-
- // Ensure that all outstanding (indirect) users of I
- // are inserted into the Worklist. Return false
- // otherwise.
- return llvm::set_is_subset(ValuesToRevisit, Worklist);
-}
+ SmallVector<Instruction *> Worklist;
+ SmallSetVector<Instruction *, 4> ValuesToRevisit;
+
+ auto PushUsersToWorklist = [&](Instruction *Inst) {
+ for (auto *U : Inst->users()) {
+ if (auto *I = dyn_cast<Instruction>(U)) {
+ if (!isAvailable(I) && !ValuesToRevisit.contains(I))
+ Worklist.emplace_back(I);
+ }
+ }
+ };
-bool PointerReplacer::collectUsersRecursive(Instruction &I) {
- for (auto *U : I.users()) {
- auto *Inst = cast<Instruction>(&*U);
+ PushUsersToWorklist(&Root);
+ while (!Worklist.empty()) {
+ auto *Inst = Worklist.pop_back_val();
+ if (!Inst)
+ return false;
+ if (isAvailable(Inst))
+ continue;
if (auto *Load = dyn_cast<LoadInst>(Inst)) {
if (Load->isVolatile())
return false;
- Worklist.insert(Load);
+ UsersToReplace.insert(Load);
} else if (auto *PHI = dyn_cast<PHINode>(Inst)) {
// All incoming values must be instructions for replacability
if (any_of(PHI->incoming_values(),
[](Value *V) { return !isa<Instruction>(V); }))
return false;
- // If at least one incoming value of the PHI is not in Worklist,
- // store the PHI for revisiting and skip this iteration of the
- // loop.
- if (any_of(PHI->incoming_values(), [this](Value *V) {
- return !isAvailable(cast<Instruction>(V));
- })) {
- ValuesToRevisit.insert(Inst);
+ // If all incoming values are available, mark this PHI as
+ // replacable and push it's users into the worklist.
+ if (all_of(PHI->incoming_values(),
+ [&](Value *V) { return isAvailable(cast<Instruction>(V)); })) {
+ UsersToReplace.insert(PHI);
+ PushUsersToWorklist(PHI);
continue;
}
- Worklist.insert(PHI);
- if (!collectUsersRecursive(*PHI))
- return false;
- } else if (auto *SI = dyn_cast<SelectInst>(Inst)) {
- if (!isa<Instruction>(SI->getTrueValue()) ||
- !isa<Instruction>(SI->getFalseValue()))
+ // Not all incoming values are available. If this PHI was already
+ // visited prior to this iteration, return false.
+ if (!ValuesToRevisit.insert(PHI))
return false;
- if (!isAvailable(cast<Instruction>(SI->getTrueValue())) ||
- !isAvailable(cast<Instruction>(SI->getFalseValue()))) {
- ValuesToRevisit.insert(Inst);
- continue;
+ // Push PHI back into the stack, followed by unavailable
+ // incoming values.
+ Worklist.emplace_back(PHI);
+ for (unsigned Idx = 0; Idx < PHI->getNumIncomingValues(); ++Idx) {
+ auto *IncomingValue = cast<Instruction>(PHI->getIncomingValue(Idx));
+ if (UsersToReplace.contains(IncomingValue))
+ continue;
+ if (!ValuesToRevisit.insert(IncomingValue))
+ return false;
+ Worklist.emplace_back(IncomingValue);
}
- Worklist.insert(SI);
- if (!collectUsersRecursive(*SI))
- return false;
- } else if (isa<GetElementPtrInst>(Inst)) {
- Worklist.insert(Inst);
- if (!collectUsersRecursive(*Inst))
+ } else if (auto *SI = dyn_cast<SelectInst>(Inst)) {
+ auto *TrueInst = dyn_cast<Instruction>(SI->getTrueValue());
+ auto *FalseInst = dyn_cast<Instruction>(SI->getFalseValue());
+ if (!TrueInst || !FalseInst)
return false;
+
+ UsersToReplace.insert(SI);
+ PushUsersToWorklist(SI);
+ } else if (auto *GEP = dyn_cast<GetElementPtrInst>(Inst)) {
+ UsersToReplace.insert(GEP);
+ PushUsersToWorklist(GEP);
} else if (auto *MI = dyn_cast<MemTransferInst>(Inst)) {
if (MI->isVolatile())
return false;
- Worklist.insert(Inst);
+ UsersToReplace.insert(Inst);
} else if (isEqualOrValidAddrSpaceCast(Inst, FromAS)) {
- Worklist.insert(Inst);
- if (!collectUsersRecursive(*Inst))
- return false;
+ UsersToReplace.insert(Inst);
+ PushUsersToWorklist(Inst);
} else if (Inst->isLifetimeStartOrEnd()) {
continue;
} else {
// TODO: For arbitrary uses with address space mismatches, should we check
// if we can introduce a valid addrspacecast?
- LLVM_DEBUG(dbgs() << "Cannot handle pointer user: " << *U << '\n');
+ LLVM_DEBUG(dbgs() << "Cannot handle pointer user: " << *Inst << '\n');
return false;
}
}
- return true;
+ return llvm::set_is_subset(ValuesToRevisit, UsersToReplace);
}
-Value *PointerReplacer::getReplacement(Value *V) { return WorkMap.lookup(V); }
+void PointerReplacer::replacePointer(Value *V) {
+#ifndef NDEBUG
+ auto *PT = cast<PointerType>(Root.getType());
+ auto *NT = cast<PointerType>(V->getType());
+ assert(PT != NT && "Invalid usage");
+#endif
+
+ WorkMap[&Root] = V;
+ SmallVector<Instruction *> Worklist;
+ SetVector<Instruction *> PostOrderWorklist;
+ SmallPtrSet<Instruction *, 4> Visited;
+
+ // Perform a postorder traversal of the users of Root.
+ Worklist.push_back(&Root);
+ while (!Worklist.empty()) {
+ Instruction *I = Worklist.back();
+
+ // If I has not been processed before, push each of its
+ // replacable users into the worklist.
+ if (Visited.insert(I).second) {
+ for (auto *U : I->users()) {
+ assert(isa<Instruction>(U) &&
+ "User should not have been"
+ " collected as it is not an instruction.");
+ auto *UserInst = cast<Instruction>(U);
+ if (UsersToReplace.contains(UserInst))
+ Worklist.push_back(UserInst);
+ }
+ // Otherwise, users of I have already been pushed into
+ // the PostOrderWorklist. Push I as well.
+ } else {
+ PostOrderWorklist.insert(I);
+ Worklist.pop_back();
+ }
+ }
+
+ // Replace pointers in reverse-postorder.
+ for (Instruction *I : llvm::reverse(PostOrderWorklist))
+ replace(I);
+}
void PointerReplacer::replace(Instruction *I) {
if (getReplacement(I))
@@ -365,12 +414,20 @@ void PointerReplacer::replace(Instruction *I) {
// replacement (new value).
WorkMap[NewI] = NewI;
} else if (auto *PHI = dyn_cast<PHINode>(I)) {
- Type *NewTy = getReplacement(PHI->getIncomingValue(0))->getType();
- auto *NewPHI = PHINode::Create(NewTy, PHI->getNumIncomingValues(),
- PHI->getName(), PHI->getIterator());
+ // Create a new PHI by replacing any incoming value that is a user of the
+ // root pointer and has a replacement.
+ SmallVector<Value *, 4> IncomingValues;
+ for (unsigned int I = 0; I < PHI->getNumIncomingValues(); ++I) {
+ Value *V = getReplacement(PHI->getIncomingValue(I));
+ if (!V)
+ V = PHI->getIncomingValue(I);
+ IncomingValues.push_back(V);
+ }
+ auto *NewPHI = PHINode::Create(IncomingValues[0]->getType(),
+ PHI->getNumIncomingValues(),
+ PHI->getName() + ".r", PHI->getIterator());
for (unsigned int I = 0; I < PHI->getNumIncomingValues(); ++I)
- NewPHI->addIncoming(getReplacement(PHI->getIncomingValue(I)),
- PHI->getIncomingBlock(I));
+ NewPHI->addIncoming(IncomingValues[I], PHI->getIncomingBlock(I));
WorkMap[PHI] = NewPHI;
} else if (auto *GEP = dyn_cast<GetElementPtrInst>(I)) {
auto *V = getReplacement(GEP->getPointerOperand());
@@ -431,22 +488,12 @@ void PointerReplacer::replace(Instruction *I) {
}
} else {
+ dbgs() << "Instruction " << *I
+ << " is not supported in PointerReplacer::replace\n";
llvm_unreachable("should never reach here");
}
}
-void PointerReplacer::replacePointer(Value *V) {
-#ifndef NDEBUG
- auto *PT = cast<PointerType>(Root.getType());
- auto *NT = cast<PointerType>(V->getType());
- assert(PT != NT && "Invalid usage");
-#endif
- WorkMap[&Root] = V;
-
- for (Instruction *Workitem : Worklist)
- replace(Workitem);
-}
-
Instruction *InstCombinerImpl::visitAllocaInst(AllocaInst &AI) {
if (auto *I = simplifyAllocaArraySize(*this, AI, DT))
return I;
diff --git a/llvm/test/Transforms/InstCombine/AMDGPU/ptr-replace-alloca.ll b/llvm/test/Transforms/InstCombine/AMDGPU/ptr-replace-alloca.ll
new file mode 100644
index 0000000000000..1a7a961a9b93a
--- /dev/null
+++ b/llvm/test/Transforms/InstCombine/AMDGPU/ptr-replace-alloca.ll
@@ -0,0 +1,48 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 5
+; RUN: opt -mtriple=amdgcn-amd-amdhsa -passes=instcombine -S < %s | FileCheck %s
+
+%struct.type = type { [256 x <2 x i64>] }
+@g1 = external hidden addrspace(3) global %struct.type, align 16
+
+; This test requires the PtrReplacer to replace users in an RPO traversal.
+; Furthermore, %ptr.else need not to be replaced so it must be retained in
+; %ptr.sink.
+define <2 x i64> @func(ptr addrspace(4) byref(%struct.type) align 16 %0, i1 %cmp.0) {
+; CHECK-LABEL: define <2 x i64> @func(
+; CHECK-SAME: ptr addrspace(4) byref([[STRUCT_TYPE:%.*]]) align 16 [[TMP0:%.*]], i1 [[CMP_0:%.*]]) {
+; CHECK-NEXT: [[ENTRY:.*:]]
+; CHECK-NEXT: br i1 [[CMP_0]], label %[[IF_THEN:.*]], label %[[IF_ELSE:.*]]
+; CHECK: [[IF_THEN]]:
+; CHECK-NEXT: [[VAL_THEN:%.*]] = addrspacecast ptr addrspace(4) [[TMP0]] to ptr
+; CHECK-NEXT: br label %[[SINK:.*]]
+; CHECK: [[IF_ELSE]]:
+; CHECK-NEXT: [[PTR_ELSE:%.*]] = load ptr, ptr addrspace(3) getelementptr inbounds nuw (i8, ptr addrspace(3) @g1, i32 32), align 16
+; CHECK-NEXT: br label %[[SINK]]
+; CHECK: [[SINK]]:
+; CHECK-NEXT: [[PTR_SINK:%.*]] = phi ptr [ [[PTR_ELSE]], %[[IF_ELSE]] ], [ [[VAL_THEN]], %[[IF_THEN]] ]
+; CHECK-NEXT: [[VAL_SINK:%.*]] = load <2 x i64>, ptr [[PTR_SINK]], align 16
+; CHECK-NEXT: ret <2 x i64> [[VAL_SINK]]
+;
+entry:
+ %coerce = alloca %struct.type, align 16, addrspace(5)
+ call void @llvm.memcpy.p5.p4.i64(ptr addrspace(5) align 16 %coerce, ptr addrspace(4) align 16 %0, i64 4096, i1 false)
+ br i1 %cmp.0, label %if.then, label %if.else
+
+if.then: ; preds = %entry
+ %ptr.then = getelementptr inbounds i8, ptr addrspace(5) %coerce, i64 0
+ %val.then = addrspacecast ptr addrspace(5) %ptr.then to ptr
+ br label %sink
+
+if.else: ; preds = %entry
+ %ptr.else = load ptr, ptr addrspace(3) getelementptr inbounds nuw (i8, ptr addrspace(3) @g1, i32 32), align 16
+ %val.else = getelementptr inbounds nuw i8, ptr %ptr.else, i64 0
+ br label %sink
+
+sink:
+ %ptr.sink = phi ptr [ %val.else, %if.else ], [ %val.then, %if.then ]
+ %val.sink = load <2 x i64>, ptr %ptr.sink, align 16
+ ret <2 x i64> %val.sink
+}
+
+; Function Attrs: nocallback nofree nounwind willreturn memory(argmem: readwrite)
+declare void @llvm.memcpy.p5.p4.i64(ptr addrspace(5) noalias writeonly captures(none), ptr addrspace(4) noalias readonly captures(none), i64, i1 immarg) #0
|
|
I tried to break this PR into two parts: user collection and pointer replacement. The changes are interleaved and there were test failures as well, unfortunately. |
|
ping |
arsenm
reviewed
May 5, 2025
arsenm
reviewed
May 5, 2025
gandhi56
force-pushed
the
ptrreplacer-refactor
branch
from
4fe4e68 to
f86efad
Compare
|
ping |
jayfoad
reviewed
May 15, 2025
gandhi56
force-pushed
the
ptrreplacer-refactor
branch
from
f86efad to
898db7b
Compare
arsenm
reviewed
May 21, 2025
gandhi56
force-pushed
the
ptrreplacer-refactor
branch
2 times, most recently
from
5348733 to
f57e87c
Compare
arsenm
reviewed
May 22, 2025
gandhi56
force-pushed
the
ptrreplacer-refactor
branch
from
f57e87c to
2d3f74c
Compare
arsenm
reviewed
May 22, 2025
arsenm
reviewed
May 22, 2025
gandhi56
force-pushed
the
ptrreplacer-refactor
branch
from
e552ce0 to
c5c2b97
Compare
|
ping |
This patch enhances the PtrReplacer as follows: 1. Users are now collected iteratively to be generous on the stack. In the case of PHIs with incoming values which have not yet been visited, they are pushed back into the stack for reconsideration. 2. Replace users of the pointer root in a reverse-postorder traversal, instead of a simple traversal over the collected users. This reordering ensures that the uses of an instruction are replaced before replacing the instruction itself. 3. During the replacement of PHI, use the same incoming value if it does not have a replacement. This patch specifically fixes the case when an incoming value of a PHI is addrspacecasted.
gandhi56
force-pushed
the
ptrreplacer-refactor
branch
from
c5c2b97 to
f4bc253
Compare
gandhi56
commented
Jun 9, 2025
| auto *IncomingValue = cast<Instruction>(PHI->getIncomingValue(Idx)); | ||
| if (UsersToReplace.contains(IncomingValue)) | ||
| continue; | ||
| if (!ValuesToRevisit.insert(IncomingValue)) |
There was a problem hiding this comment.
If a value is inserted into ValuesToRevisit and is not replaceable, it will be caught here and the PointerReplacer will return false. Hence, this function does not need to check for unreplaceable elements in ValuesToRevisit at the time of returning.
shiltian
approved these changes
Jun 9, 2025
|
ping, in case there are any additional comments. |
|
Thanks for all of your feedback, I will merge this PR as it seems there is no further comment. |
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/10/builds/7442 Here is the relevant piece of the build log for the reference |
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/11/builds/17409 Here is the relevant piece of the build log for the reference |
gandhi56
added a commit
that referenced
this pull request
Jun 16, 2025
Reverts #137215 This commit caused a failure in the LLVM CI: https://lab.llvm.org/buildbot/#/builders/10/builds/7442
llvm-sync bot
pushed a commit
to arm/arm-toolchain
that referenced
this pull request
Jun 16, 2025
…" (#144394) Reverts llvm/llvm-project#137215 This commit caused a failure in the LLVM CI: https://lab.llvm.org/buildbot/#/builders/10/builds/7442
fschlimb
pushed a commit
to fschlimb/llvm-project
that referenced
this pull request
Jun 18, 2025
This patch enhances the PtrReplacer as follows: 1. Users are now collected iteratively to be generous on the stack. In the case of PHIs with incoming values which have not yet been visited, they are pushed back into the stack for reconsideration. 2. Replace users of the pointer root in a reverse-postorder traversal, instead of a simple traversal over the collected users. This reordering ensures that the operands of an instruction are replaced before replacing the instruction itself. 3. During the replacement of PHI, use the same incoming value if it does not have a replacement. This patch specifically fixes the case when an incoming value of a PHI is addrspacecasted.
fschlimb
pushed a commit
to fschlimb/llvm-project
that referenced
this pull request
Jun 18, 2025
) Reverts llvm#137215 This commit caused a failure in the LLVM CI: https://lab.llvm.org/buildbot/#/builders/10/builds/7442
gandhi56
added a commit
that referenced
this pull request
Jun 20, 2025
This patch enhances the PtrReplacer as follows: 1. Users are now collected iteratively to be generous on the stack. In the case of PHIs with incoming values which have not yet been visited, they are pushed back into the stack for reconsideration. 2. Replace users of the pointer root in a reverse-postorder traversal, instead of a simpletraversal over the collected users. This reordering ensures that the uses of an instruction are replaced before replacing the instruction itself. 3. During the replacement of PHI, use the same incoming value if it does not have a replacement. This patch specifically fixes the case when an incoming value of a PHI is addrspacecasted. This is a reland of #137215.
llvm-sync bot
pushed a commit
to arm/arm-toolchain
that referenced
this pull request
Jun 20, 2025
… (#144626) This patch enhances the PtrReplacer as follows: 1. Users are now collected iteratively to be generous on the stack. In the case of PHIs with incoming values which have not yet been visited, they are pushed back into the stack for reconsideration. 2. Replace users of the pointer root in a reverse-postorder traversal, instead of a simpletraversal over the collected users. This reordering ensures that the uses of an instruction are replaced before replacing the instruction itself. 3. During the replacement of PHI, use the same incoming value if it does not have a replacement. This patch specifically fixes the case when an incoming value of a PHI is addrspacecasted. This is a reland of llvm/llvm-project#137215.
Jaddyen
pushed a commit
to Jaddyen/llvm-project
that referenced
this pull request
Jun 23, 2025
This patch enhances the PtrReplacer as follows: 1. Users are now collected iteratively to be generous on the stack. In the case of PHIs with incoming values which have not yet been visited, they are pushed back into the stack for reconsideration. 2. Replace users of the pointer root in a reverse-postorder traversal, instead of a simpletraversal over the collected users. This reordering ensures that the uses of an instruction are replaced before replacing the instruction itself. 3. During the replacement of PHI, use the same incoming value if it does not have a replacement. This patch specifically fixes the case when an incoming value of a PHI is addrspacecasted. This is a reland of llvm#137215.
gandhi56
added a commit
to gandhi56/llvm-project
that referenced
this pull request
Jun 23, 2025
This patch enhances the PtrReplacer as follows: 1. Users are now collected iteratively to be generous on the stack. In the case of PHIs with incoming values which have not yet been visited, they are pushed back into the stack for reconsideration. 2. Replace users of the pointer root in a reverse-postorder traversal, instead of a simple traversal over the collected users. This reordering ensures that the uses of an instruction are replaced before replacing the instruction itself. 3. During the replacement of PHI, use the same incoming value if it does not have a replacement. This patch specifically fixes the case when an incoming value of a PHI is addrspacecasted. This reland PR includes a fix for an assertion failure caused by llvm#137215, which was reverted. The failing test involved a phi and gep depending on each other, in which case the PtrReplacer did not order them correctly for replacement. This patch fixes it by adding a check during the population of PostOrderWorklist.
gandhi56
added a commit
to gandhi56/llvm-project
that referenced
this pull request
Jun 23, 2025
This patch enhances the PtrReplacer as follows: 1. Users are now collected iteratively to be generous on the stack. In the case of PHIs with incoming values which have not yet been visited, they are pushed back into the stack for reconsideration. 2. Replace users of the pointer root in a reverse-postorder traversal, instead of a simple traversal over the collected users. This reordering ensures that the uses of an instruction are replaced before replacing the instruction itself. 3. During the replacement of PHI, use the same incoming value if it does not have a replacement. This patch specifically fixes the case when an incoming value of a PHI is addrspacecasted. This reland PR includes a fix for an assertion failure caused by llvm#137215, which was reverted. The failing test involved a phi and gep depending on each other, in which case the PtrReplacer did not order them correctly for replacement. This patch fixes it by adding a check during the population of PostOrderWorklist.
gandhi56
added a commit
that referenced
this pull request
Jun 24, 2025
This patch enhances the PtrReplacer as follows: 1. Users are now collected iteratively to be generous on the stack. In the case of PHIs with incoming values which have not yet been visited, they are pushed back into the stack for reconsideration. 2. Replace users of the pointer root in a reverse-postorder traversal, instead of a simple traversal over the collected users. This reordering ensures that the uses of an instruction are replaced before replacing the instruction itself. 3. During the replacement of PHI, use the same incoming value if it does not have a replacement. This patch specifically fixes the case when an incoming value of a PHI is addrspacecasted. This reland PR includes a fix for an assertion failure caused by #137215, which was reverted. The failing test involved a phi and gep depending on each other, in which case the PtrReplacer did not order them correctly for replacement. This patch fixes it by adding a check during the definition of `PostOrderWorklist`.
llvm-sync bot
pushed a commit
to arm/arm-toolchain
that referenced
this pull request
Jun 24, 2025
… (#145410) This patch enhances the PtrReplacer as follows: 1. Users are now collected iteratively to be generous on the stack. In the case of PHIs with incoming values which have not yet been visited, they are pushed back into the stack for reconsideration. 2. Replace users of the pointer root in a reverse-postorder traversal, instead of a simple traversal over the collected users. This reordering ensures that the uses of an instruction are replaced before replacing the instruction itself. 3. During the replacement of PHI, use the same incoming value if it does not have a replacement. This patch specifically fixes the case when an incoming value of a PHI is addrspacecasted. This reland PR includes a fix for an assertion failure caused by llvm/llvm-project#137215, which was reverted. The failing test involved a phi and gep depending on each other, in which case the PtrReplacer did not order them correctly for replacement. This patch fixes it by adding a check during the definition of `PostOrderWorklist`.
akuhlens
pushed a commit
to akuhlens/llvm-project
that referenced
this pull request
Jun 24, 2025
This patch enhances the PtrReplacer as follows: 1. Users are now collected iteratively to be generous on the stack. In the case of PHIs with incoming values which have not yet been visited, they are pushed back into the stack for reconsideration. 2. Replace users of the pointer root in a reverse-postorder traversal, instead of a simple traversal over the collected users. This reordering ensures that the operands of an instruction are replaced before replacing the instruction itself. 3. During the replacement of PHI, use the same incoming value if it does not have a replacement. This patch specifically fixes the case when an incoming value of a PHI is addrspacecasted.
akuhlens
pushed a commit
to akuhlens/llvm-project
that referenced
this pull request
Jun 24, 2025
) Reverts llvm#137215 This commit caused a failure in the LLVM CI: https://lab.llvm.org/buildbot/#/builders/10/builds/7442
DrSergei
pushed a commit
to DrSergei/llvm-project
that referenced
this pull request
Jun 24, 2025
This patch enhances the PtrReplacer as follows: 1. Users are now collected iteratively to be generous on the stack. In the case of PHIs with incoming values which have not yet been visited, they are pushed back into the stack for reconsideration. 2. Replace users of the pointer root in a reverse-postorder traversal, instead of a simple traversal over the collected users. This reordering ensures that the uses of an instruction are replaced before replacing the instruction itself. 3. During the replacement of PHI, use the same incoming value if it does not have a replacement. This patch specifically fixes the case when an incoming value of a PHI is addrspacecasted. This reland PR includes a fix for an assertion failure caused by llvm#137215, which was reverted. The failing test involved a phi and gep depending on each other, in which case the PtrReplacer did not order them correctly for replacement. This patch fixes it by adding a check during the definition of `PostOrderWorklist`.
anthonyhatran
pushed a commit
to anthonyhatran/llvm-project
that referenced
this pull request
Jun 26, 2025
This patch enhances the PtrReplacer as follows: 1. Users are now collected iteratively to be generous on the stack. In the case of PHIs with incoming values which have not yet been visited, they are pushed back into the stack for reconsideration. 2. Replace users of the pointer root in a reverse-postorder traversal, instead of a simple traversal over the collected users. This reordering ensures that the uses of an instruction are replaced before replacing the instruction itself. 3. During the replacement of PHI, use the same incoming value if it does not have a replacement. This patch specifically fixes the case when an incoming value of a PHI is addrspacecasted. This reland PR includes a fix for an assertion failure caused by llvm#137215, which was reverted. The failing test involved a phi and gep depending on each other, in which case the PtrReplacer did not order them correctly for replacement. This patch fixes it by adding a check during the definition of `PostOrderWorklist`.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This patch enhances the PtrReplacer as follows:
This patch specifically fixes the case when an incoming value of a PHI is addrspacecasted.