[DebugInfo][RemoveDIs] Suppress getNextNonDebugInfoInstruction #144383
+68
−90
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
There are no longer debug-info instructions, thust we don't need this skipping.
llvmbot
added
clang
|
@llvm/pr-subscribers-vectorizers @llvm/pr-subscribers-backend-x86 Author: Jeremy Morse (jmorse) ChangesThere are no longer debug-info instructions, thus we don't need this skipping. Horray! CC @WenleiHe , Patch is 34.17 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/144383.diff 30 Files Affected:
diff --git a/clang/lib/CodeGen/CGBlocks.cpp b/clang/lib/CodeGen/CGBlocks.cpp
index 729758ddce560..b5466bb6e346f 100644
--- a/clang/lib/CodeGen/CGBlocks.cpp
+++ b/clang/lib/CodeGen/CGBlocks.cpp
@@ -1532,8 +1532,8 @@ llvm::Function *CodeGenFunction::GenerateBlockFunction(
llvm::BasicBlock *resume = Builder.GetInsertBlock();
// Go back to the entry.
- if (entry_ptr->getNextNonDebugInstruction())
- entry_ptr = entry_ptr->getNextNonDebugInstruction()->getIterator();
+ if (entry_ptr->getNextNode())
+ entry_ptr = entry_ptr->getNextNode()->getIterator();
else
entry_ptr = entry->end();
Builder.SetInsertPoint(entry, entry_ptr);
diff --git a/libc/docs/configure.rst b/libc/docs/configure.rst
index 8d53390ae19bf..109412225634f 100644
--- a/libc/docs/configure.rst
+++ b/libc/docs/configure.rst
@@ -29,7 +29,7 @@ to learn about the defaults for your platform and target.
- ``LIBC_CONF_ENABLE_STRONG_STACK_PROTECTOR``: Enable -fstack-protector-strong to defend against stack smashing attack.
- ``LIBC_CONF_KEEP_FRAME_POINTER``: Keep frame pointer in functions for better debugging experience.
* **"errno" options**
- - ``LIBC_CONF_ERRNO_MODE``: The implementation used for errno, acceptable values are LIBC_ERRNO_MODE_DEFAULT, LIBC_ERRNO_MODE_UNDEFINED, LIBC_ERRNO_MODE_THREAD_LOCAL, LIBC_ERRNO_MODE_SHARED, LIBC_ERRNO_MODE_EXTERNAL, and LIBC_ERRNO_MODE_SYSTEM.
+ - ``LIBC_CONF_ERRNO_MODE``: The implementation used for errno, acceptable values are LIBC_ERRNO_MODE_DEFAULT, LIBC_ERRNO_MODE_UNDEFINED, LIBC_ERRNO_MODE_THREAD_LOCAL, LIBC_ERRNO_MODE_SHARED, LIBC_ERRNO_MODE_EXTERNAL, LIBC_ERRNO_MODE_SYSTEM, and LIBC_ERRNO_MODE_SYSTEM_INLINE.
* **"general" options**
- ``LIBC_ADD_NULL_CHECKS``: Add nullptr checks in the library's implementations to some functions for which passing nullptr is undefined behavior.
* **"math" options**
diff --git a/llvm/include/llvm/IR/Instruction.h b/llvm/include/llvm/IR/Instruction.h
index 10fc9c1298607..9904e7bc4b299 100644
--- a/llvm/include/llvm/IR/Instruction.h
+++ b/llvm/include/llvm/IR/Instruction.h
@@ -898,17 +898,6 @@ class Instruction : public User,
/// Return true if the instruction is a DbgInfoIntrinsic or PseudoProbeInst.
LLVM_ABI bool isDebugOrPseudoInst() const LLVM_READONLY;
- /// Return a pointer to the next non-debug instruction in the same basic
- /// block as 'this', or nullptr if no such instruction exists. Skip any pseudo
- /// operations if \c SkipPseudoOp is true.
- LLVM_ABI const Instruction *
- getNextNonDebugInstruction(bool SkipPseudoOp = false) const;
- Instruction *getNextNonDebugInstruction(bool SkipPseudoOp = false) {
- return const_cast<Instruction *>(
- static_cast<const Instruction *>(this)->getNextNonDebugInstruction(
- SkipPseudoOp));
- }
-
/// Return a pointer to the previous non-debug instruction in the same basic
/// block as 'this', or nullptr if no such instruction exists. Skip any pseudo
/// operations if \c SkipPseudoOp is true.
diff --git a/llvm/include/llvm/Transforms/Utils/LockstepReverseIterator.h b/llvm/include/llvm/Transforms/Utils/LockstepReverseIterator.h
index 1b6309c7fb1a4..cd525a9710103 100644
--- a/llvm/include/llvm/Transforms/Utils/LockstepReverseIterator.h
+++ b/llvm/include/llvm/Transforms/Utils/LockstepReverseIterator.h
@@ -133,7 +133,7 @@ class LockstepReverseIterator
return *this;
SmallVector<Instruction *, 4> NewInsts;
for (Instruction *Inst : Insts) {
- Instruction *Next = Inst->getNextNonDebugInstruction();
+ Instruction *Next = Inst->getNextNode();
// Already at end of block.
if (!Next) {
Fail = true;
diff --git a/llvm/lib/Analysis/MemoryDependenceAnalysis.cpp b/llvm/lib/Analysis/MemoryDependenceAnalysis.cpp
index f062189bac6a0..3709081e257be 100644
--- a/llvm/lib/Analysis/MemoryDependenceAnalysis.cpp
+++ b/llvm/lib/Analysis/MemoryDependenceAnalysis.cpp
@@ -356,7 +356,7 @@ static bool canSkipClobberingStore(const StoreInst *SI,
if (BatchAA.alias(MemoryLocation::get(LI), MemLoc) != AliasResult::MustAlias)
return false;
unsigned NumVisitedInsts = 0;
- for (const Instruction *I = LI; I != SI; I = I->getNextNonDebugInstruction())
+ for (const Instruction *I = LI; I != SI; I = I->getNextNode())
if (++NumVisitedInsts > ScanLimit ||
isModSet(BatchAA.getModRefInfo(I, MemLoc)))
return false;
diff --git a/llvm/lib/CodeGen/CodeGenPrepare.cpp b/llvm/lib/CodeGen/CodeGenPrepare.cpp
index 3792b456c836e..af6306c126136 100644
--- a/llvm/lib/CodeGen/CodeGenPrepare.cpp
+++ b/llvm/lib/CodeGen/CodeGenPrepare.cpp
@@ -7327,7 +7327,7 @@ bool CodeGenPrepare::optimizeLoadExt(LoadInst *Load) {
!TLI->isLoadExtLegal(ISD::ZEXTLOAD, LoadResultVT, TruncVT))
return false;
- IRBuilder<> Builder(Load->getNextNonDebugInstruction());
+ IRBuilder<> Builder(Load->getNextNode());
auto *NewAnd = cast<Instruction>(
Builder.CreateAnd(Load, ConstantInt::get(Ctx, DemandBits)));
// Mark this instruction as "inserted by CGP", so that other
diff --git a/llvm/lib/IR/Instruction.cpp b/llvm/lib/IR/Instruction.cpp
index 1b60caab6c11a..8c71d077ed67b 100644
--- a/llvm/lib/IR/Instruction.cpp
+++ b/llvm/lib/IR/Instruction.cpp
@@ -1235,14 +1235,6 @@ bool Instruction::isDebugOrPseudoInst() const {
return isa<DbgInfoIntrinsic>(this) || isa<PseudoProbeInst>(this);
}
-const Instruction *
-Instruction::getNextNonDebugInstruction(bool SkipPseudoOp) const {
- for (const Instruction *I = getNextNode(); I; I = I->getNextNode())
- if (!isa<DbgInfoIntrinsic>(I) && !(SkipPseudoOp && isa<PseudoProbeInst>(I)))
- return I;
- return nullptr;
-}
-
const Instruction *
Instruction::getPrevNonDebugInstruction(bool SkipPseudoOp) const {
for (const Instruction *I = getPrevNode(); I; I = I->getPrevNode())
@@ -1252,9 +1244,6 @@ Instruction::getPrevNonDebugInstruction(bool SkipPseudoOp) const {
}
const DebugLoc &Instruction::getStableDebugLoc() const {
- if (isa<DbgInfoIntrinsic>(this))
- if (const Instruction *Next = getNextNonDebugInstruction())
- return Next->getDebugLoc();
return getDebugLoc();
}
diff --git a/llvm/lib/IR/Verifier.cpp b/llvm/lib/IR/Verifier.cpp
index 1f1041b259736..a3a3e3f4e9551 100644
--- a/llvm/lib/IR/Verifier.cpp
+++ b/llvm/lib/IR/Verifier.cpp
@@ -6455,7 +6455,7 @@ void Verifier::visitIntrinsicCall(Intrinsic::ID ID, CallBase &Call) {
Check(!Call.paramHasAttr(3, Attribute::InReg),
"VGPR arguments must not have the `inreg` attribute", &Call);
- auto *Next = Call.getNextNonDebugInstruction();
+ auto *Next = Call.getNextNode();
bool IsAMDUnreachable = Next && isa<IntrinsicInst>(Next) &&
cast<IntrinsicInst>(Next)->getIntrinsicID() ==
Intrinsic::amdgcn_unreachable;
diff --git a/llvm/lib/Target/AArch64/AArch64TargetTransformInfo.cpp b/llvm/lib/Target/AArch64/AArch64TargetTransformInfo.cpp
index 0232ac421aeda..396c8a725b5bd 100644
--- a/llvm/lib/Target/AArch64/AArch64TargetTransformInfo.cpp
+++ b/llvm/lib/Target/AArch64/AArch64TargetTransformInfo.cpp
@@ -2674,14 +2674,14 @@ static std::optional<Instruction *> instCombineSVEInsr(InstCombiner &IC,
static std::optional<Instruction *> instCombineDMB(InstCombiner &IC,
IntrinsicInst &II) {
// If this barrier is post-dominated by identical one we can remove it
- auto *NI = II.getNextNonDebugInstruction();
+ auto *NI = II.getNextNode();
unsigned LookaheadThreshold = DMBLookaheadThreshold;
auto CanSkipOver = [](Instruction *I) {
return !I->mayReadOrWriteMemory() && !I->mayHaveSideEffects();
};
while (LookaheadThreshold-- && CanSkipOver(NI)) {
auto *NIBB = NI->getParent();
- NI = NI->getNextNonDebugInstruction();
+ NI = NI->getNextNode();
if (!NI) {
if (auto *SuccBB = NIBB->getUniqueSuccessor())
NI = &*SuccBB->getFirstNonPHIOrDbgOrLifetime();
diff --git a/llvm/lib/Target/X86/X86WinEHState.cpp b/llvm/lib/Target/X86/X86WinEHState.cpp
index 27111fce45662..a650f6f069e5f 100644
--- a/llvm/lib/Target/X86/X86WinEHState.cpp
+++ b/llvm/lib/Target/X86/X86WinEHState.cpp
@@ -811,7 +811,7 @@ void WinEHStatePass::updateEspForInAllocas(Function &F) {
if (auto *Alloca = dyn_cast<AllocaInst>(&I)) {
if (Alloca->isStaticAlloca())
continue;
- IRBuilder<> Builder(Alloca->getNextNonDebugInstruction());
+ IRBuilder<> Builder(Alloca->getNextNode());
// SavedESP = llvm.stacksave()
Value *SP = Builder.CreateStackSave();
Builder.CreateStore(SP, Builder.CreateStructGEP(RegNodeTy, RegNode, 0));
@@ -820,7 +820,7 @@ void WinEHStatePass::updateEspForInAllocas(Function &F) {
if (auto *II = dyn_cast<IntrinsicInst>(&I)) {
if (II->getIntrinsicID() != Intrinsic::stackrestore)
continue;
- IRBuilder<> Builder(II->getNextNonDebugInstruction());
+ IRBuilder<> Builder(II->getNextNode());
// SavedESP = llvm.stacksave()
Value *SP = Builder.CreateStackSave();
Builder.CreateStore(SP, Builder.CreateStructGEP(RegNodeTy, RegNode, 0));
diff --git a/llvm/lib/Transforms/Coroutines/CoroSplit.cpp b/llvm/lib/Transforms/Coroutines/CoroSplit.cpp
index 8813f91e9060c..1dc49e221b6ff 100644
--- a/llvm/lib/Transforms/Coroutines/CoroSplit.cpp
+++ b/llvm/lib/Transforms/Coroutines/CoroSplit.cpp
@@ -807,7 +807,7 @@ static void updateScopeLine(Instruction *ActiveSuspend,
return;
// No subsequent instruction -> fallback to the location of ActiveSuspend.
- if (!ActiveSuspend->getNextNonDebugInstruction()) {
+ if (!ActiveSuspend->getNextNode()) {
if (auto DL = ActiveSuspend->getDebugLoc())
if (SPToUpdate.getFile() == DL->getFile())
SPToUpdate.setScopeLine(DL->getLine());
@@ -815,7 +815,7 @@ static void updateScopeLine(Instruction *ActiveSuspend,
}
BasicBlock::iterator Successor =
- ActiveSuspend->getNextNonDebugInstruction()->getIterator();
+ ActiveSuspend->getNextNode()->getIterator();
// Corosplit splits the BB around ActiveSuspend, so the meaningful
// instructions are not in the same BB.
if (auto *Branch = dyn_cast_or_null<BranchInst>(Successor);
diff --git a/llvm/lib/Transforms/IPO/Attributor.cpp b/llvm/lib/Transforms/IPO/Attributor.cpp
index 050eed376ed3f..df8c7b680cd5f 100644
--- a/llvm/lib/Transforms/IPO/Attributor.cpp
+++ b/llvm/lib/Transforms/IPO/Attributor.cpp
@@ -795,7 +795,7 @@ isPotentiallyReachable(Attributor &A, const Instruction &FromI,
if (isa<InvokeInst>(CB))
return false;
- Instruction *Inst = CB->getNextNonDebugInstruction();
+ Instruction *Inst = CB->getNextNode();
Worklist.push_back(Inst);
return true;
};
diff --git a/llvm/lib/Transforms/IPO/AttributorAttributes.cpp b/llvm/lib/Transforms/IPO/AttributorAttributes.cpp
index 3799a696f67af..4f5ba74e53f4b 100644
--- a/llvm/lib/Transforms/IPO/AttributorAttributes.cpp
+++ b/llvm/lib/Transforms/IPO/AttributorAttributes.cpp
@@ -1774,7 +1774,7 @@ ChangeStatus AAPointerInfoFloating::updateImpl(Attributor &A) {
do {
if (FromI->mayWriteToMemory() && !IsAssumption(*FromI))
return true;
- FromI = FromI->getNextNonDebugInstruction();
+ FromI = FromI->getNextNode();
} while (FromI && FromI != ToI);
return false;
};
@@ -1785,7 +1785,7 @@ ChangeStatus AAPointerInfoFloating::updateImpl(Attributor &A) {
return false;
BasicBlock *IntrBB = IntrI.getParent();
if (IntrI.getParent() == BB) {
- if (IsImpactedInRange(LoadI->getNextNonDebugInstruction(), &IntrI))
+ if (IsImpactedInRange(LoadI->getNextNode(), &IntrI))
return false;
} else {
auto PredIt = pred_begin(IntrBB);
@@ -1802,7 +1802,7 @@ ChangeStatus AAPointerInfoFloating::updateImpl(Attributor &A) {
continue;
return false;
}
- if (IsImpactedInRange(LoadI->getNextNonDebugInstruction(),
+ if (IsImpactedInRange(LoadI->getNextNode(),
BB->getTerminator()))
return false;
if (IsImpactedInRange(&IntrBB->front(), &IntrI))
diff --git a/llvm/lib/Transforms/IPO/GlobalOpt.cpp b/llvm/lib/Transforms/IPO/GlobalOpt.cpp
index 7db0586386506..f830e53cfc532 100644
--- a/llvm/lib/Transforms/IPO/GlobalOpt.cpp
+++ b/llvm/lib/Transforms/IPO/GlobalOpt.cpp
@@ -1880,7 +1880,7 @@ static void RemovePreallocated(Function *F) {
Builder.SetInsertPoint(PreallocatedSetup);
auto *StackSave = Builder.CreateStackSave();
- Builder.SetInsertPoint(NewCB->getNextNonDebugInstruction());
+ Builder.SetInsertPoint(NewCB->getNextNode());
Builder.CreateStackRestore(StackSave);
// Replace @llvm.call.preallocated.arg() with alloca.
@@ -1904,7 +1904,7 @@ static void RemovePreallocated(Function *F) {
auto AddressSpace = UseCall->getType()->getPointerAddressSpace();
auto *ArgType =
UseCall->getFnAttr(Attribute::Preallocated).getValueAsType();
- auto *InsertBefore = PreallocatedSetup->getNextNonDebugInstruction();
+ auto *InsertBefore = PreallocatedSetup->getNextNode();
Builder.SetInsertPoint(InsertBefore);
auto *Alloca =
Builder.CreateAlloca(ArgType, AddressSpace, nullptr, "paarg");
diff --git a/llvm/lib/Transforms/IPO/IROutliner.cpp b/llvm/lib/Transforms/IPO/IROutliner.cpp
index cb18b55ae2183..bdb16ae812cf3 100644
--- a/llvm/lib/Transforms/IPO/IROutliner.cpp
+++ b/llvm/lib/Transforms/IPO/IROutliner.cpp
@@ -268,7 +268,7 @@ void OutlinableRegion::splitCandidate() {
// instruction. If they do not match, there could be problems in rewriting
// the program after outlining, so we ignore it.
if (!BackInst->isTerminator() &&
- EndInst != BackInst->getNextNonDebugInstruction())
+ EndInst != BackInst->getNextNode())
return;
Instruction *StartInst = (*Candidate->begin()).Inst;
@@ -2340,7 +2340,7 @@ static bool nextIRInstructionDataMatchesNextInst(IRInstructionData &ID) {
Instruction *NextIDLInst = NextIDIt->Inst;
Instruction *NextModuleInst = nullptr;
if (!ID.Inst->isTerminator())
- NextModuleInst = ID.Inst->getNextNonDebugInstruction();
+ NextModuleInst = ID.Inst->getNextNode();
else if (NextIDLInst != nullptr)
NextModuleInst =
&*NextIDIt->Inst->getParent()->instructionsWithoutDebug().begin();
@@ -2367,7 +2367,7 @@ bool IROutliner::isCompatibleWithAlreadyOutlinedCode(
// if it does not, we fix it in the InstructionDataList.
if (!Region.Candidate->backInstruction()->isTerminator()) {
Instruction *NewEndInst =
- Region.Candidate->backInstruction()->getNextNonDebugInstruction();
+ Region.Candidate->backInstruction()->getNextNode();
assert(NewEndInst && "Next instruction is a nullptr?");
if (Region.Candidate->end()->Inst != NewEndInst) {
IRInstructionDataList *IDL = Region.Candidate->front()->IDL;
diff --git a/llvm/lib/Transforms/IPO/OpenMPOpt.cpp b/llvm/lib/Transforms/IPO/OpenMPOpt.cpp
index dd7ae7e66e350..5de2285c2d2e3 100644
--- a/llvm/lib/Transforms/IPO/OpenMPOpt.cpp
+++ b/llvm/lib/Transforms/IPO/OpenMPOpt.cpp
@@ -2856,7 +2856,7 @@ struct AAExecutionDomainFunction : public AAExecutionDomain {
if (!It->getSecond().IsReachingAlignedBarrierOnly)
ForwardIsOk = false;
break;
- } while ((CurI = CurI->getNextNonDebugInstruction()));
+ } while ((CurI = CurI->getNextNode()));
if (!CurI && !BEDMap.lookup(I.getParent()).IsReachingAlignedBarrierOnly)
ForwardIsOk = false;
diff --git a/llvm/lib/Transforms/InstCombine/InstCombineCalls.cpp b/llvm/lib/Transforms/InstCombine/InstCombineCalls.cpp
index c169ab25b2106..1886ee781508e 100644
--- a/llvm/lib/Transforms/InstCombine/InstCombineCalls.cpp
+++ b/llvm/lib/Transforms/InstCombine/InstCombineCalls.cpp
@@ -3243,7 +3243,7 @@ Instruction *InstCombinerImpl::visitCallInst(CallInst &CI) {
// Remove an assume if it is followed by an identical assume.
// TODO: Do we need this? Unless there are conflicting assumptions, the
// computeKnownBits(IIOperand) below here eliminates redundant assumes.
- Instruction *Next = II->getNextNonDebugInstruction();
+ Instruction *Next = II->getNextNode();
if (match(Next, m_Intrinsic<Intrinsic::assume>(m_Specific(IIOperand))))
return RemoveConditionFromAssume(Next);
@@ -3416,12 +3416,12 @@ Instruction *InstCombinerImpl::visitCallInst(CallInst &CI) {
// Is this guard followed by another guard? We scan forward over a small
// fixed window of instructions to handle common cases with conditions
// computed between guards.
- Instruction *NextInst = II->getNextNonDebugInstruction();
+ Instruction *NextInst = II->getNextNode();
for (unsigned i = 0; i < GuardWideningWindow; i++) {
// Note: Using context-free form to avoid compile time blow up
if (!isSafeToSpeculativelyExecute(NextInst))
break;
- NextInst = NextInst->getNextNonDebugInstruction();
+ NextInst = NextInst->getNextNode();
}
Value *NextCond = nullptr;
if (match(NextInst,
@@ -3431,10 +3431,10 @@ Instruction *InstCombinerImpl::visitCallInst(CallInst &CI) {
// Remove a guard that it is immediately preceded by an identical guard.
// Otherwise canonicalize guard(a); guard(b) -> guard(a & b).
if (CurrCond != NextCond) {
- Instruction *MoveI = II->getNextNonDebugInstruction();
+ Instruction *MoveI = II->getNextNode();
while (MoveI != NextInst) {
auto *Temp = MoveI;
- MoveI = MoveI->getNextNonDebugInstruction();
+ MoveI = MoveI->getNextNode();
Temp->moveBefore(II->getIterator());
}
replaceOperand(*II, 0, Builder.CreateAnd(CurrCond, NextCond));
@@ -3871,7 +3871,7 @@ Instruction *InstCombinerImpl::visitCallInst(CallInst &CI) {
// Fence instruction simplification
Instruction *InstCombinerImpl::visitFenceInst(FenceInst &FI) {
- auto *NFI = dyn_cast<FenceInst>(FI.getNextNonDebugInstruction());
+ auto *NFI = dyn_cast<FenceInst>(FI.getNextNode());
// This check is solely here to handle arbitrary target-dependent syncscopes.
// TODO: Can remove if does not matter in practice.
if (NFI && FI.isIdenticalTo(NFI))
diff --git a/llvm/lib/Transforms/InstCombine/InstructionCombining.cpp b/llvm/lib/Transforms/InstCombine/InstructionCombining.cpp
index e261807bbc035..9c9dddb73318f 100644
--- a/llvm/lib/Transforms/InstCombine/InstructionCombining.cpp
+++ b/llvm/lib/Transforms/InstCombine/InstructionCombining.cpp
@@ -4760,7 +4760,7 @@ bool InstCombinerImpl::freezeOtherUses(FreezeInst &FI) {
// Don't move to the position of a debug intrinsic.
if (isa<DbgInfoIntrinsic>(MoveBefore))
- MoveBefore = MoveBefore->getNextNonDebugInstruction()->getIterator();
+ MoveBefore = MoveBefore->getNextNode()->getIterator();
// Re-point iterator to come after any debug-info records, if we're
// running in "RemoveDIs" mode
MoveBefore.setHeadBit(false);
diff --git a/llvm/lib/Transforms/Instrumentation/AddressSanitizer.cpp b/llvm/lib/Transforms/Instrumentation/AddressSanitizer.cpp
index 840a5e3f31dfd..dfbe4f8172066 100644
--- a/llvm/lib/Transforms/Instrumentation/AddressSanitizer.cpp
+++ b/llvm/lib/Transforms/Instrumentation/AddressSanitizer.cpp
@@ -3396,8 +3396,8 @@ void FunctionStackPoisoner::processDynamicAllocas() {
static void findStoresToUninstrumentedArgAllocas(
AddressSanitizer &ASan, Instruction &InsBefore,
SmallVectorImpl<Instruction *> &InitInsts) {
- Instruction *Start = InsBefore.getNextNonDebugInstruction();
- for (Instruction *It = Start; It; It = It->getNextNonDebugInstruction()) {
+ Instruction *Start = InsBefore.getNextNode();
+ for (Instruction *It = Start; It; It = It->getNextNode()) {
// Argument initialization looks like:
// 1) store <Argument>, <Alloca> OR
// 2) <CastArgument> = cast <Argument> to ...
diff --git a/llvm/lib/Transforms/Instrumentation/HWAddressSanitizer.cpp b/llvm/lib/Transforms/Instrumentation/HWAddressSanitizer.cpp
index 77db686f8229c..2c34bf2157cdd 100644
--- a/llvm/lib/Transforms/Instrumentation/HWAddressSanitizer.cpp
+++ b/llvm/lib/Transforms/Instrumentation/HWAddressSanitizer.cpp
@@ -1421,7 +1421,7 @@ void HWAddressSanitizer::emitPrologue(IRBuilder<> &IRB, bool WithFrameRecord) {
bool HWAddressSanitizer::instrum...
[truncated]
|
|
@llvm/pr-subscribers-clang-codegen Author: Jeremy Morse (jmorse) ChangesThere are no longer debug-info instructions, thus we don't need this skipping. Horray! CC @WenleiHe , Patch is 34.17 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/144383.diff 30 Files Affected:
diff --git a/clang/lib/CodeGen/CGBlocks.cpp b/clang/lib/CodeGen/CGBlocks.cpp
index 729758ddce560..b5466bb6e346f 100644
--- a/clang/lib/CodeGen/CGBlocks.cpp
+++ b/clang/lib/CodeGen/CGBlocks.cpp
@@ -1532,8 +1532,8 @@ llvm::Function *CodeGenFunction::GenerateBlockFunction(
llvm::BasicBlock *resume = Builder.GetInsertBlock();
// Go back to the entry.
- if (entry_ptr->getNextNonDebugInstruction())
- entry_ptr = entry_ptr->getNextNonDebugInstruction()->getIterator();
+ if (entry_ptr->getNextNode())
+ entry_ptr = entry_ptr->getNextNode()->getIterator();
else
entry_ptr = entry->end();
Builder.SetInsertPoint(entry, entry_ptr);
diff --git a/libc/docs/configure.rst b/libc/docs/configure.rst
index 8d53390ae19bf..109412225634f 100644
--- a/libc/docs/configure.rst
+++ b/libc/docs/configure.rst
@@ -29,7 +29,7 @@ to learn about the defaults for your platform and target.
- ``LIBC_CONF_ENABLE_STRONG_STACK_PROTECTOR``: Enable -fstack-protector-strong to defend against stack smashing attack.
- ``LIBC_CONF_KEEP_FRAME_POINTER``: Keep frame pointer in functions for better debugging experience.
* **"errno" options**
- - ``LIBC_CONF_ERRNO_MODE``: The implementation used for errno, acceptable values are LIBC_ERRNO_MODE_DEFAULT, LIBC_ERRNO_MODE_UNDEFINED, LIBC_ERRNO_MODE_THREAD_LOCAL, LIBC_ERRNO_MODE_SHARED, LIBC_ERRNO_MODE_EXTERNAL, and LIBC_ERRNO_MODE_SYSTEM.
+ - ``LIBC_CONF_ERRNO_MODE``: The implementation used for errno, acceptable values are LIBC_ERRNO_MODE_DEFAULT, LIBC_ERRNO_MODE_UNDEFINED, LIBC_ERRNO_MODE_THREAD_LOCAL, LIBC_ERRNO_MODE_SHARED, LIBC_ERRNO_MODE_EXTERNAL, LIBC_ERRNO_MODE_SYSTEM, and LIBC_ERRNO_MODE_SYSTEM_INLINE.
* **"general" options**
- ``LIBC_ADD_NULL_CHECKS``: Add nullptr checks in the library's implementations to some functions for which passing nullptr is undefined behavior.
* **"math" options**
diff --git a/llvm/include/llvm/IR/Instruction.h b/llvm/include/llvm/IR/Instruction.h
index 10fc9c1298607..9904e7bc4b299 100644
--- a/llvm/include/llvm/IR/Instruction.h
+++ b/llvm/include/llvm/IR/Instruction.h
@@ -898,17 +898,6 @@ class Instruction : public User,
/// Return true if the instruction is a DbgInfoIntrinsic or PseudoProbeInst.
LLVM_ABI bool isDebugOrPseudoInst() const LLVM_READONLY;
- /// Return a pointer to the next non-debug instruction in the same basic
- /// block as 'this', or nullptr if no such instruction exists. Skip any pseudo
- /// operations if \c SkipPseudoOp is true.
- LLVM_ABI const Instruction *
- getNextNonDebugInstruction(bool SkipPseudoOp = false) const;
- Instruction *getNextNonDebugInstruction(bool SkipPseudoOp = false) {
- return const_cast<Instruction *>(
- static_cast<const Instruction *>(this)->getNextNonDebugInstruction(
- SkipPseudoOp));
- }
-
/// Return a pointer to the previous non-debug instruction in the same basic
/// block as 'this', or nullptr if no such instruction exists. Skip any pseudo
/// operations if \c SkipPseudoOp is true.
diff --git a/llvm/include/llvm/Transforms/Utils/LockstepReverseIterator.h b/llvm/include/llvm/Transforms/Utils/LockstepReverseIterator.h
index 1b6309c7fb1a4..cd525a9710103 100644
--- a/llvm/include/llvm/Transforms/Utils/LockstepReverseIterator.h
+++ b/llvm/include/llvm/Transforms/Utils/LockstepReverseIterator.h
@@ -133,7 +133,7 @@ class LockstepReverseIterator
return *this;
SmallVector<Instruction *, 4> NewInsts;
for (Instruction *Inst : Insts) {
- Instruction *Next = Inst->getNextNonDebugInstruction();
+ Instruction *Next = Inst->getNextNode();
// Already at end of block.
if (!Next) {
Fail = true;
diff --git a/llvm/lib/Analysis/MemoryDependenceAnalysis.cpp b/llvm/lib/Analysis/MemoryDependenceAnalysis.cpp
index f062189bac6a0..3709081e257be 100644
--- a/llvm/lib/Analysis/MemoryDependenceAnalysis.cpp
+++ b/llvm/lib/Analysis/MemoryDependenceAnalysis.cpp
@@ -356,7 +356,7 @@ static bool canSkipClobberingStore(const StoreInst *SI,
if (BatchAA.alias(MemoryLocation::get(LI), MemLoc) != AliasResult::MustAlias)
return false;
unsigned NumVisitedInsts = 0;
- for (const Instruction *I = LI; I != SI; I = I->getNextNonDebugInstruction())
+ for (const Instruction *I = LI; I != SI; I = I->getNextNode())
if (++NumVisitedInsts > ScanLimit ||
isModSet(BatchAA.getModRefInfo(I, MemLoc)))
return false;
diff --git a/llvm/lib/CodeGen/CodeGenPrepare.cpp b/llvm/lib/CodeGen/CodeGenPrepare.cpp
index 3792b456c836e..af6306c126136 100644
--- a/llvm/lib/CodeGen/CodeGenPrepare.cpp
+++ b/llvm/lib/CodeGen/CodeGenPrepare.cpp
@@ -7327,7 +7327,7 @@ bool CodeGenPrepare::optimizeLoadExt(LoadInst *Load) {
!TLI->isLoadExtLegal(ISD::ZEXTLOAD, LoadResultVT, TruncVT))
return false;
- IRBuilder<> Builder(Load->getNextNonDebugInstruction());
+ IRBuilder<> Builder(Load->getNextNode());
auto *NewAnd = cast<Instruction>(
Builder.CreateAnd(Load, ConstantInt::get(Ctx, DemandBits)));
// Mark this instruction as "inserted by CGP", so that other
diff --git a/llvm/lib/IR/Instruction.cpp b/llvm/lib/IR/Instruction.cpp
index 1b60caab6c11a..8c71d077ed67b 100644
--- a/llvm/lib/IR/Instruction.cpp
+++ b/llvm/lib/IR/Instruction.cpp
@@ -1235,14 +1235,6 @@ bool Instruction::isDebugOrPseudoInst() const {
return isa<DbgInfoIntrinsic>(this) || isa<PseudoProbeInst>(this);
}
-const Instruction *
-Instruction::getNextNonDebugInstruction(bool SkipPseudoOp) const {
- for (const Instruction *I = getNextNode(); I; I = I->getNextNode())
- if (!isa<DbgInfoIntrinsic>(I) && !(SkipPseudoOp && isa<PseudoProbeInst>(I)))
- return I;
- return nullptr;
-}
-
const Instruction *
Instruction::getPrevNonDebugInstruction(bool SkipPseudoOp) const {
for (const Instruction *I = getPrevNode(); I; I = I->getPrevNode())
@@ -1252,9 +1244,6 @@ Instruction::getPrevNonDebugInstruction(bool SkipPseudoOp) const {
}
const DebugLoc &Instruction::getStableDebugLoc() const {
- if (isa<DbgInfoIntrinsic>(this))
- if (const Instruction *Next = getNextNonDebugInstruction())
- return Next->getDebugLoc();
return getDebugLoc();
}
diff --git a/llvm/lib/IR/Verifier.cpp b/llvm/lib/IR/Verifier.cpp
index 1f1041b259736..a3a3e3f4e9551 100644
--- a/llvm/lib/IR/Verifier.cpp
+++ b/llvm/lib/IR/Verifier.cpp
@@ -6455,7 +6455,7 @@ void Verifier::visitIntrinsicCall(Intrinsic::ID ID, CallBase &Call) {
Check(!Call.paramHasAttr(3, Attribute::InReg),
"VGPR arguments must not have the `inreg` attribute", &Call);
- auto *Next = Call.getNextNonDebugInstruction();
+ auto *Next = Call.getNextNode();
bool IsAMDUnreachable = Next && isa<IntrinsicInst>(Next) &&
cast<IntrinsicInst>(Next)->getIntrinsicID() ==
Intrinsic::amdgcn_unreachable;
diff --git a/llvm/lib/Target/AArch64/AArch64TargetTransformInfo.cpp b/llvm/lib/Target/AArch64/AArch64TargetTransformInfo.cpp
index 0232ac421aeda..396c8a725b5bd 100644
--- a/llvm/lib/Target/AArch64/AArch64TargetTransformInfo.cpp
+++ b/llvm/lib/Target/AArch64/AArch64TargetTransformInfo.cpp
@@ -2674,14 +2674,14 @@ static std::optional<Instruction *> instCombineSVEInsr(InstCombiner &IC,
static std::optional<Instruction *> instCombineDMB(InstCombiner &IC,
IntrinsicInst &II) {
// If this barrier is post-dominated by identical one we can remove it
- auto *NI = II.getNextNonDebugInstruction();
+ auto *NI = II.getNextNode();
unsigned LookaheadThreshold = DMBLookaheadThreshold;
auto CanSkipOver = [](Instruction *I) {
return !I->mayReadOrWriteMemory() && !I->mayHaveSideEffects();
};
while (LookaheadThreshold-- && CanSkipOver(NI)) {
auto *NIBB = NI->getParent();
- NI = NI->getNextNonDebugInstruction();
+ NI = NI->getNextNode();
if (!NI) {
if (auto *SuccBB = NIBB->getUniqueSuccessor())
NI = &*SuccBB->getFirstNonPHIOrDbgOrLifetime();
diff --git a/llvm/lib/Target/X86/X86WinEHState.cpp b/llvm/lib/Target/X86/X86WinEHState.cpp
index 27111fce45662..a650f6f069e5f 100644
--- a/llvm/lib/Target/X86/X86WinEHState.cpp
+++ b/llvm/lib/Target/X86/X86WinEHState.cpp
@@ -811,7 +811,7 @@ void WinEHStatePass::updateEspForInAllocas(Function &F) {
if (auto *Alloca = dyn_cast<AllocaInst>(&I)) {
if (Alloca->isStaticAlloca())
continue;
- IRBuilder<> Builder(Alloca->getNextNonDebugInstruction());
+ IRBuilder<> Builder(Alloca->getNextNode());
// SavedESP = llvm.stacksave()
Value *SP = Builder.CreateStackSave();
Builder.CreateStore(SP, Builder.CreateStructGEP(RegNodeTy, RegNode, 0));
@@ -820,7 +820,7 @@ void WinEHStatePass::updateEspForInAllocas(Function &F) {
if (auto *II = dyn_cast<IntrinsicInst>(&I)) {
if (II->getIntrinsicID() != Intrinsic::stackrestore)
continue;
- IRBuilder<> Builder(II->getNextNonDebugInstruction());
+ IRBuilder<> Builder(II->getNextNode());
// SavedESP = llvm.stacksave()
Value *SP = Builder.CreateStackSave();
Builder.CreateStore(SP, Builder.CreateStructGEP(RegNodeTy, RegNode, 0));
diff --git a/llvm/lib/Transforms/Coroutines/CoroSplit.cpp b/llvm/lib/Transforms/Coroutines/CoroSplit.cpp
index 8813f91e9060c..1dc49e221b6ff 100644
--- a/llvm/lib/Transforms/Coroutines/CoroSplit.cpp
+++ b/llvm/lib/Transforms/Coroutines/CoroSplit.cpp
@@ -807,7 +807,7 @@ static void updateScopeLine(Instruction *ActiveSuspend,
return;
// No subsequent instruction -> fallback to the location of ActiveSuspend.
- if (!ActiveSuspend->getNextNonDebugInstruction()) {
+ if (!ActiveSuspend->getNextNode()) {
if (auto DL = ActiveSuspend->getDebugLoc())
if (SPToUpdate.getFile() == DL->getFile())
SPToUpdate.setScopeLine(DL->getLine());
@@ -815,7 +815,7 @@ static void updateScopeLine(Instruction *ActiveSuspend,
}
BasicBlock::iterator Successor =
- ActiveSuspend->getNextNonDebugInstruction()->getIterator();
+ ActiveSuspend->getNextNode()->getIterator();
// Corosplit splits the BB around ActiveSuspend, so the meaningful
// instructions are not in the same BB.
if (auto *Branch = dyn_cast_or_null<BranchInst>(Successor);
diff --git a/llvm/lib/Transforms/IPO/Attributor.cpp b/llvm/lib/Transforms/IPO/Attributor.cpp
index 050eed376ed3f..df8c7b680cd5f 100644
--- a/llvm/lib/Transforms/IPO/Attributor.cpp
+++ b/llvm/lib/Transforms/IPO/Attributor.cpp
@@ -795,7 +795,7 @@ isPotentiallyReachable(Attributor &A, const Instruction &FromI,
if (isa<InvokeInst>(CB))
return false;
- Instruction *Inst = CB->getNextNonDebugInstruction();
+ Instruction *Inst = CB->getNextNode();
Worklist.push_back(Inst);
return true;
};
diff --git a/llvm/lib/Transforms/IPO/AttributorAttributes.cpp b/llvm/lib/Transforms/IPO/AttributorAttributes.cpp
index 3799a696f67af..4f5ba74e53f4b 100644
--- a/llvm/lib/Transforms/IPO/AttributorAttributes.cpp
+++ b/llvm/lib/Transforms/IPO/AttributorAttributes.cpp
@@ -1774,7 +1774,7 @@ ChangeStatus AAPointerInfoFloating::updateImpl(Attributor &A) {
do {
if (FromI->mayWriteToMemory() && !IsAssumption(*FromI))
return true;
- FromI = FromI->getNextNonDebugInstruction();
+ FromI = FromI->getNextNode();
} while (FromI && FromI != ToI);
return false;
};
@@ -1785,7 +1785,7 @@ ChangeStatus AAPointerInfoFloating::updateImpl(Attributor &A) {
return false;
BasicBlock *IntrBB = IntrI.getParent();
if (IntrI.getParent() == BB) {
- if (IsImpactedInRange(LoadI->getNextNonDebugInstruction(), &IntrI))
+ if (IsImpactedInRange(LoadI->getNextNode(), &IntrI))
return false;
} else {
auto PredIt = pred_begin(IntrBB);
@@ -1802,7 +1802,7 @@ ChangeStatus AAPointerInfoFloating::updateImpl(Attributor &A) {
continue;
return false;
}
- if (IsImpactedInRange(LoadI->getNextNonDebugInstruction(),
+ if (IsImpactedInRange(LoadI->getNextNode(),
BB->getTerminator()))
return false;
if (IsImpactedInRange(&IntrBB->front(), &IntrI))
diff --git a/llvm/lib/Transforms/IPO/GlobalOpt.cpp b/llvm/lib/Transforms/IPO/GlobalOpt.cpp
index 7db0586386506..f830e53cfc532 100644
--- a/llvm/lib/Transforms/IPO/GlobalOpt.cpp
+++ b/llvm/lib/Transforms/IPO/GlobalOpt.cpp
@@ -1880,7 +1880,7 @@ static void RemovePreallocated(Function *F) {
Builder.SetInsertPoint(PreallocatedSetup);
auto *StackSave = Builder.CreateStackSave();
- Builder.SetInsertPoint(NewCB->getNextNonDebugInstruction());
+ Builder.SetInsertPoint(NewCB->getNextNode());
Builder.CreateStackRestore(StackSave);
// Replace @llvm.call.preallocated.arg() with alloca.
@@ -1904,7 +1904,7 @@ static void RemovePreallocated(Function *F) {
auto AddressSpace = UseCall->getType()->getPointerAddressSpace();
auto *ArgType =
UseCall->getFnAttr(Attribute::Preallocated).getValueAsType();
- auto *InsertBefore = PreallocatedSetup->getNextNonDebugInstruction();
+ auto *InsertBefore = PreallocatedSetup->getNextNode();
Builder.SetInsertPoint(InsertBefore);
auto *Alloca =
Builder.CreateAlloca(ArgType, AddressSpace, nullptr, "paarg");
diff --git a/llvm/lib/Transforms/IPO/IROutliner.cpp b/llvm/lib/Transforms/IPO/IROutliner.cpp
index cb18b55ae2183..bdb16ae812cf3 100644
--- a/llvm/lib/Transforms/IPO/IROutliner.cpp
+++ b/llvm/lib/Transforms/IPO/IROutliner.cpp
@@ -268,7 +268,7 @@ void OutlinableRegion::splitCandidate() {
// instruction. If they do not match, there could be problems in rewriting
// the program after outlining, so we ignore it.
if (!BackInst->isTerminator() &&
- EndInst != BackInst->getNextNonDebugInstruction())
+ EndInst != BackInst->getNextNode())
return;
Instruction *StartInst = (*Candidate->begin()).Inst;
@@ -2340,7 +2340,7 @@ static bool nextIRInstructionDataMatchesNextInst(IRInstructionData &ID) {
Instruction *NextIDLInst = NextIDIt->Inst;
Instruction *NextModuleInst = nullptr;
if (!ID.Inst->isTerminator())
- NextModuleInst = ID.Inst->getNextNonDebugInstruction();
+ NextModuleInst = ID.Inst->getNextNode();
else if (NextIDLInst != nullptr)
NextModuleInst =
&*NextIDIt->Inst->getParent()->instructionsWithoutDebug().begin();
@@ -2367,7 +2367,7 @@ bool IROutliner::isCompatibleWithAlreadyOutlinedCode(
// if it does not, we fix it in the InstructionDataList.
if (!Region.Candidate->backInstruction()->isTerminator()) {
Instruction *NewEndInst =
- Region.Candidate->backInstruction()->getNextNonDebugInstruction();
+ Region.Candidate->backInstruction()->getNextNode();
assert(NewEndInst && "Next instruction is a nullptr?");
if (Region.Candidate->end()->Inst != NewEndInst) {
IRInstructionDataList *IDL = Region.Candidate->front()->IDL;
diff --git a/llvm/lib/Transforms/IPO/OpenMPOpt.cpp b/llvm/lib/Transforms/IPO/OpenMPOpt.cpp
index dd7ae7e66e350..5de2285c2d2e3 100644
--- a/llvm/lib/Transforms/IPO/OpenMPOpt.cpp
+++ b/llvm/lib/Transforms/IPO/OpenMPOpt.cpp
@@ -2856,7 +2856,7 @@ struct AAExecutionDomainFunction : public AAExecutionDomain {
if (!It->getSecond().IsReachingAlignedBarrierOnly)
ForwardIsOk = false;
break;
- } while ((CurI = CurI->getNextNonDebugInstruction()));
+ } while ((CurI = CurI->getNextNode()));
if (!CurI && !BEDMap.lookup(I.getParent()).IsReachingAlignedBarrierOnly)
ForwardIsOk = false;
diff --git a/llvm/lib/Transforms/InstCombine/InstCombineCalls.cpp b/llvm/lib/Transforms/InstCombine/InstCombineCalls.cpp
index c169ab25b2106..1886ee781508e 100644
--- a/llvm/lib/Transforms/InstCombine/InstCombineCalls.cpp
+++ b/llvm/lib/Transforms/InstCombine/InstCombineCalls.cpp
@@ -3243,7 +3243,7 @@ Instruction *InstCombinerImpl::visitCallInst(CallInst &CI) {
// Remove an assume if it is followed by an identical assume.
// TODO: Do we need this? Unless there are conflicting assumptions, the
// computeKnownBits(IIOperand) below here eliminates redundant assumes.
- Instruction *Next = II->getNextNonDebugInstruction();
+ Instruction *Next = II->getNextNode();
if (match(Next, m_Intrinsic<Intrinsic::assume>(m_Specific(IIOperand))))
return RemoveConditionFromAssume(Next);
@@ -3416,12 +3416,12 @@ Instruction *InstCombinerImpl::visitCallInst(CallInst &CI) {
// Is this guard followed by another guard? We scan forward over a small
// fixed window of instructions to handle common cases with conditions
// computed between guards.
- Instruction *NextInst = II->getNextNonDebugInstruction();
+ Instruction *NextInst = II->getNextNode();
for (unsigned i = 0; i < GuardWideningWindow; i++) {
// Note: Using context-free form to avoid compile time blow up
if (!isSafeToSpeculativelyExecute(NextInst))
break;
- NextInst = NextInst->getNextNonDebugInstruction();
+ NextInst = NextInst->getNextNode();
}
Value *NextCond = nullptr;
if (match(NextInst,
@@ -3431,10 +3431,10 @@ Instruction *InstCombinerImpl::visitCallInst(CallInst &CI) {
// Remove a guard that it is immediately preceded by an identical guard.
// Otherwise canonicalize guard(a); guard(b) -> guard(a & b).
if (CurrCond != NextCond) {
- Instruction *MoveI = II->getNextNonDebugInstruction();
+ Instruction *MoveI = II->getNextNode();
while (MoveI != NextInst) {
auto *Temp = MoveI;
- MoveI = MoveI->getNextNonDebugInstruction();
+ MoveI = MoveI->getNextNode();
Temp->moveBefore(II->getIterator());
}
replaceOperand(*II, 0, Builder.CreateAnd(CurrCond, NextCond));
@@ -3871,7 +3871,7 @@ Instruction *InstCombinerImpl::visitCallInst(CallInst &CI) {
// Fence instruction simplification
Instruction *InstCombinerImpl::visitFenceInst(FenceInst &FI) {
- auto *NFI = dyn_cast<FenceInst>(FI.getNextNonDebugInstruction());
+ auto *NFI = dyn_cast<FenceInst>(FI.getNextNode());
// This check is solely here to handle arbitrary target-dependent syncscopes.
// TODO: Can remove if does not matter in practice.
if (NFI && FI.isIdenticalTo(NFI))
diff --git a/llvm/lib/Transforms/InstCombine/InstructionCombining.cpp b/llvm/lib/Transforms/InstCombine/InstructionCombining.cpp
index e261807bbc035..9c9dddb73318f 100644
--- a/llvm/lib/Transforms/InstCombine/InstructionCombining.cpp
+++ b/llvm/lib/Transforms/InstCombine/InstructionCombining.cpp
@@ -4760,7 +4760,7 @@ bool InstCombinerImpl::freezeOtherUses(FreezeInst &FI) {
// Don't move to the position of a debug intrinsic.
if (isa<DbgInfoIntrinsic>(MoveBefore))
- MoveBefore = MoveBefore->getNextNonDebugInstruction()->getIterator();
+ MoveBefore = MoveBefore->getNextNode()->getIterator();
// Re-point iterator to come after any debug-info records, if we're
// running in "RemoveDIs" mode
MoveBefore.setHeadBit(false);
diff --git a/llvm/lib/Transforms/Instrumentation/AddressSanitizer.cpp b/llvm/lib/Transforms/Instrumentation/AddressSanitizer.cpp
index 840a5e3f31dfd..dfbe4f8172066 100644
--- a/llvm/lib/Transforms/Instrumentation/AddressSanitizer.cpp
+++ b/llvm/lib/Transforms/Instrumentation/AddressSanitizer.cpp
@@ -3396,8 +3396,8 @@ void FunctionStackPoisoner::processDynamicAllocas() {
static void findStoresToUninstrumentedArgAllocas(
AddressSanitizer &ASan, Instruction &InsBefore,
SmallVectorImpl<Instruction *> &InitInsts) {
- Instruction *Start = InsBefore.getNextNonDebugInstruction();
- for (Instruction *It = Start; It; It = It->getNextNonDebugInstruction()) {
+ Instruction *Start = InsBefore.getNextNode();
+ for (Instruction *It = Start; It; It = It->getNextNode()) {
// Argument initialization looks like:
// 1) store <Argument>, <Alloca> OR
// 2) <CastArgument> = cast <Argument> to ...
diff --git a/llvm/lib/Transforms/Instrumentation/HWAddressSanitizer.cpp b/llvm/lib/Transforms/Instrumentation/HWAddressSanitizer.cpp
index 77db686f8229c..2c34bf2157cdd 100644
--- a/llvm/lib/Transforms/Instrumentation/HWAddressSanitizer.cpp
+++ b/llvm/lib/Transforms/Instrumentation/HWAddressSanitizer.cpp
@@ -1421,7 +1421,7 @@ void HWAddressSanitizer::emitPrologue(IRBuilder<> &IRB, bool WithFrameRecord) {
bool HWAddressSanitizer::instrum...
[truncated]
|
|
@llvm/pr-subscribers-backend-aarch64 Author: Jeremy Morse (jmorse) ChangesThere are no longer debug-info instructions, thus we don't need this skipping. Horray! CC @WenleiHe , Patch is 34.17 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/144383.diff 30 Files Affected:
diff --git a/clang/lib/CodeGen/CGBlocks.cpp b/clang/lib/CodeGen/CGBlocks.cpp
index 729758ddce560..b5466bb6e346f 100644
--- a/clang/lib/CodeGen/CGBlocks.cpp
+++ b/clang/lib/CodeGen/CGBlocks.cpp
@@ -1532,8 +1532,8 @@ llvm::Function *CodeGenFunction::GenerateBlockFunction(
llvm::BasicBlock *resume = Builder.GetInsertBlock();
// Go back to the entry.
- if (entry_ptr->getNextNonDebugInstruction())
- entry_ptr = entry_ptr->getNextNonDebugInstruction()->getIterator();
+ if (entry_ptr->getNextNode())
+ entry_ptr = entry_ptr->getNextNode()->getIterator();
else
entry_ptr = entry->end();
Builder.SetInsertPoint(entry, entry_ptr);
diff --git a/libc/docs/configure.rst b/libc/docs/configure.rst
index 8d53390ae19bf..109412225634f 100644
--- a/libc/docs/configure.rst
+++ b/libc/docs/configure.rst
@@ -29,7 +29,7 @@ to learn about the defaults for your platform and target.
- ``LIBC_CONF_ENABLE_STRONG_STACK_PROTECTOR``: Enable -fstack-protector-strong to defend against stack smashing attack.
- ``LIBC_CONF_KEEP_FRAME_POINTER``: Keep frame pointer in functions for better debugging experience.
* **"errno" options**
- - ``LIBC_CONF_ERRNO_MODE``: The implementation used for errno, acceptable values are LIBC_ERRNO_MODE_DEFAULT, LIBC_ERRNO_MODE_UNDEFINED, LIBC_ERRNO_MODE_THREAD_LOCAL, LIBC_ERRNO_MODE_SHARED, LIBC_ERRNO_MODE_EXTERNAL, and LIBC_ERRNO_MODE_SYSTEM.
+ - ``LIBC_CONF_ERRNO_MODE``: The implementation used for errno, acceptable values are LIBC_ERRNO_MODE_DEFAULT, LIBC_ERRNO_MODE_UNDEFINED, LIBC_ERRNO_MODE_THREAD_LOCAL, LIBC_ERRNO_MODE_SHARED, LIBC_ERRNO_MODE_EXTERNAL, LIBC_ERRNO_MODE_SYSTEM, and LIBC_ERRNO_MODE_SYSTEM_INLINE.
* **"general" options**
- ``LIBC_ADD_NULL_CHECKS``: Add nullptr checks in the library's implementations to some functions for which passing nullptr is undefined behavior.
* **"math" options**
diff --git a/llvm/include/llvm/IR/Instruction.h b/llvm/include/llvm/IR/Instruction.h
index 10fc9c1298607..9904e7bc4b299 100644
--- a/llvm/include/llvm/IR/Instruction.h
+++ b/llvm/include/llvm/IR/Instruction.h
@@ -898,17 +898,6 @@ class Instruction : public User,
/// Return true if the instruction is a DbgInfoIntrinsic or PseudoProbeInst.
LLVM_ABI bool isDebugOrPseudoInst() const LLVM_READONLY;
- /// Return a pointer to the next non-debug instruction in the same basic
- /// block as 'this', or nullptr if no such instruction exists. Skip any pseudo
- /// operations if \c SkipPseudoOp is true.
- LLVM_ABI const Instruction *
- getNextNonDebugInstruction(bool SkipPseudoOp = false) const;
- Instruction *getNextNonDebugInstruction(bool SkipPseudoOp = false) {
- return const_cast<Instruction *>(
- static_cast<const Instruction *>(this)->getNextNonDebugInstruction(
- SkipPseudoOp));
- }
-
/// Return a pointer to the previous non-debug instruction in the same basic
/// block as 'this', or nullptr if no such instruction exists. Skip any pseudo
/// operations if \c SkipPseudoOp is true.
diff --git a/llvm/include/llvm/Transforms/Utils/LockstepReverseIterator.h b/llvm/include/llvm/Transforms/Utils/LockstepReverseIterator.h
index 1b6309c7fb1a4..cd525a9710103 100644
--- a/llvm/include/llvm/Transforms/Utils/LockstepReverseIterator.h
+++ b/llvm/include/llvm/Transforms/Utils/LockstepReverseIterator.h
@@ -133,7 +133,7 @@ class LockstepReverseIterator
return *this;
SmallVector<Instruction *, 4> NewInsts;
for (Instruction *Inst : Insts) {
- Instruction *Next = Inst->getNextNonDebugInstruction();
+ Instruction *Next = Inst->getNextNode();
// Already at end of block.
if (!Next) {
Fail = true;
diff --git a/llvm/lib/Analysis/MemoryDependenceAnalysis.cpp b/llvm/lib/Analysis/MemoryDependenceAnalysis.cpp
index f062189bac6a0..3709081e257be 100644
--- a/llvm/lib/Analysis/MemoryDependenceAnalysis.cpp
+++ b/llvm/lib/Analysis/MemoryDependenceAnalysis.cpp
@@ -356,7 +356,7 @@ static bool canSkipClobberingStore(const StoreInst *SI,
if (BatchAA.alias(MemoryLocation::get(LI), MemLoc) != AliasResult::MustAlias)
return false;
unsigned NumVisitedInsts = 0;
- for (const Instruction *I = LI; I != SI; I = I->getNextNonDebugInstruction())
+ for (const Instruction *I = LI; I != SI; I = I->getNextNode())
if (++NumVisitedInsts > ScanLimit ||
isModSet(BatchAA.getModRefInfo(I, MemLoc)))
return false;
diff --git a/llvm/lib/CodeGen/CodeGenPrepare.cpp b/llvm/lib/CodeGen/CodeGenPrepare.cpp
index 3792b456c836e..af6306c126136 100644
--- a/llvm/lib/CodeGen/CodeGenPrepare.cpp
+++ b/llvm/lib/CodeGen/CodeGenPrepare.cpp
@@ -7327,7 +7327,7 @@ bool CodeGenPrepare::optimizeLoadExt(LoadInst *Load) {
!TLI->isLoadExtLegal(ISD::ZEXTLOAD, LoadResultVT, TruncVT))
return false;
- IRBuilder<> Builder(Load->getNextNonDebugInstruction());
+ IRBuilder<> Builder(Load->getNextNode());
auto *NewAnd = cast<Instruction>(
Builder.CreateAnd(Load, ConstantInt::get(Ctx, DemandBits)));
// Mark this instruction as "inserted by CGP", so that other
diff --git a/llvm/lib/IR/Instruction.cpp b/llvm/lib/IR/Instruction.cpp
index 1b60caab6c11a..8c71d077ed67b 100644
--- a/llvm/lib/IR/Instruction.cpp
+++ b/llvm/lib/IR/Instruction.cpp
@@ -1235,14 +1235,6 @@ bool Instruction::isDebugOrPseudoInst() const {
return isa<DbgInfoIntrinsic>(this) || isa<PseudoProbeInst>(this);
}
-const Instruction *
-Instruction::getNextNonDebugInstruction(bool SkipPseudoOp) const {
- for (const Instruction *I = getNextNode(); I; I = I->getNextNode())
- if (!isa<DbgInfoIntrinsic>(I) && !(SkipPseudoOp && isa<PseudoProbeInst>(I)))
- return I;
- return nullptr;
-}
-
const Instruction *
Instruction::getPrevNonDebugInstruction(bool SkipPseudoOp) const {
for (const Instruction *I = getPrevNode(); I; I = I->getPrevNode())
@@ -1252,9 +1244,6 @@ Instruction::getPrevNonDebugInstruction(bool SkipPseudoOp) const {
}
const DebugLoc &Instruction::getStableDebugLoc() const {
- if (isa<DbgInfoIntrinsic>(this))
- if (const Instruction *Next = getNextNonDebugInstruction())
- return Next->getDebugLoc();
return getDebugLoc();
}
diff --git a/llvm/lib/IR/Verifier.cpp b/llvm/lib/IR/Verifier.cpp
index 1f1041b259736..a3a3e3f4e9551 100644
--- a/llvm/lib/IR/Verifier.cpp
+++ b/llvm/lib/IR/Verifier.cpp
@@ -6455,7 +6455,7 @@ void Verifier::visitIntrinsicCall(Intrinsic::ID ID, CallBase &Call) {
Check(!Call.paramHasAttr(3, Attribute::InReg),
"VGPR arguments must not have the `inreg` attribute", &Call);
- auto *Next = Call.getNextNonDebugInstruction();
+ auto *Next = Call.getNextNode();
bool IsAMDUnreachable = Next && isa<IntrinsicInst>(Next) &&
cast<IntrinsicInst>(Next)->getIntrinsicID() ==
Intrinsic::amdgcn_unreachable;
diff --git a/llvm/lib/Target/AArch64/AArch64TargetTransformInfo.cpp b/llvm/lib/Target/AArch64/AArch64TargetTransformInfo.cpp
index 0232ac421aeda..396c8a725b5bd 100644
--- a/llvm/lib/Target/AArch64/AArch64TargetTransformInfo.cpp
+++ b/llvm/lib/Target/AArch64/AArch64TargetTransformInfo.cpp
@@ -2674,14 +2674,14 @@ static std::optional<Instruction *> instCombineSVEInsr(InstCombiner &IC,
static std::optional<Instruction *> instCombineDMB(InstCombiner &IC,
IntrinsicInst &II) {
// If this barrier is post-dominated by identical one we can remove it
- auto *NI = II.getNextNonDebugInstruction();
+ auto *NI = II.getNextNode();
unsigned LookaheadThreshold = DMBLookaheadThreshold;
auto CanSkipOver = [](Instruction *I) {
return !I->mayReadOrWriteMemory() && !I->mayHaveSideEffects();
};
while (LookaheadThreshold-- && CanSkipOver(NI)) {
auto *NIBB = NI->getParent();
- NI = NI->getNextNonDebugInstruction();
+ NI = NI->getNextNode();
if (!NI) {
if (auto *SuccBB = NIBB->getUniqueSuccessor())
NI = &*SuccBB->getFirstNonPHIOrDbgOrLifetime();
diff --git a/llvm/lib/Target/X86/X86WinEHState.cpp b/llvm/lib/Target/X86/X86WinEHState.cpp
index 27111fce45662..a650f6f069e5f 100644
--- a/llvm/lib/Target/X86/X86WinEHState.cpp
+++ b/llvm/lib/Target/X86/X86WinEHState.cpp
@@ -811,7 +811,7 @@ void WinEHStatePass::updateEspForInAllocas(Function &F) {
if (auto *Alloca = dyn_cast<AllocaInst>(&I)) {
if (Alloca->isStaticAlloca())
continue;
- IRBuilder<> Builder(Alloca->getNextNonDebugInstruction());
+ IRBuilder<> Builder(Alloca->getNextNode());
// SavedESP = llvm.stacksave()
Value *SP = Builder.CreateStackSave();
Builder.CreateStore(SP, Builder.CreateStructGEP(RegNodeTy, RegNode, 0));
@@ -820,7 +820,7 @@ void WinEHStatePass::updateEspForInAllocas(Function &F) {
if (auto *II = dyn_cast<IntrinsicInst>(&I)) {
if (II->getIntrinsicID() != Intrinsic::stackrestore)
continue;
- IRBuilder<> Builder(II->getNextNonDebugInstruction());
+ IRBuilder<> Builder(II->getNextNode());
// SavedESP = llvm.stacksave()
Value *SP = Builder.CreateStackSave();
Builder.CreateStore(SP, Builder.CreateStructGEP(RegNodeTy, RegNode, 0));
diff --git a/llvm/lib/Transforms/Coroutines/CoroSplit.cpp b/llvm/lib/Transforms/Coroutines/CoroSplit.cpp
index 8813f91e9060c..1dc49e221b6ff 100644
--- a/llvm/lib/Transforms/Coroutines/CoroSplit.cpp
+++ b/llvm/lib/Transforms/Coroutines/CoroSplit.cpp
@@ -807,7 +807,7 @@ static void updateScopeLine(Instruction *ActiveSuspend,
return;
// No subsequent instruction -> fallback to the location of ActiveSuspend.
- if (!ActiveSuspend->getNextNonDebugInstruction()) {
+ if (!ActiveSuspend->getNextNode()) {
if (auto DL = ActiveSuspend->getDebugLoc())
if (SPToUpdate.getFile() == DL->getFile())
SPToUpdate.setScopeLine(DL->getLine());
@@ -815,7 +815,7 @@ static void updateScopeLine(Instruction *ActiveSuspend,
}
BasicBlock::iterator Successor =
- ActiveSuspend->getNextNonDebugInstruction()->getIterator();
+ ActiveSuspend->getNextNode()->getIterator();
// Corosplit splits the BB around ActiveSuspend, so the meaningful
// instructions are not in the same BB.
if (auto *Branch = dyn_cast_or_null<BranchInst>(Successor);
diff --git a/llvm/lib/Transforms/IPO/Attributor.cpp b/llvm/lib/Transforms/IPO/Attributor.cpp
index 050eed376ed3f..df8c7b680cd5f 100644
--- a/llvm/lib/Transforms/IPO/Attributor.cpp
+++ b/llvm/lib/Transforms/IPO/Attributor.cpp
@@ -795,7 +795,7 @@ isPotentiallyReachable(Attributor &A, const Instruction &FromI,
if (isa<InvokeInst>(CB))
return false;
- Instruction *Inst = CB->getNextNonDebugInstruction();
+ Instruction *Inst = CB->getNextNode();
Worklist.push_back(Inst);
return true;
};
diff --git a/llvm/lib/Transforms/IPO/AttributorAttributes.cpp b/llvm/lib/Transforms/IPO/AttributorAttributes.cpp
index 3799a696f67af..4f5ba74e53f4b 100644
--- a/llvm/lib/Transforms/IPO/AttributorAttributes.cpp
+++ b/llvm/lib/Transforms/IPO/AttributorAttributes.cpp
@@ -1774,7 +1774,7 @@ ChangeStatus AAPointerInfoFloating::updateImpl(Attributor &A) {
do {
if (FromI->mayWriteToMemory() && !IsAssumption(*FromI))
return true;
- FromI = FromI->getNextNonDebugInstruction();
+ FromI = FromI->getNextNode();
} while (FromI && FromI != ToI);
return false;
};
@@ -1785,7 +1785,7 @@ ChangeStatus AAPointerInfoFloating::updateImpl(Attributor &A) {
return false;
BasicBlock *IntrBB = IntrI.getParent();
if (IntrI.getParent() == BB) {
- if (IsImpactedInRange(LoadI->getNextNonDebugInstruction(), &IntrI))
+ if (IsImpactedInRange(LoadI->getNextNode(), &IntrI))
return false;
} else {
auto PredIt = pred_begin(IntrBB);
@@ -1802,7 +1802,7 @@ ChangeStatus AAPointerInfoFloating::updateImpl(Attributor &A) {
continue;
return false;
}
- if (IsImpactedInRange(LoadI->getNextNonDebugInstruction(),
+ if (IsImpactedInRange(LoadI->getNextNode(),
BB->getTerminator()))
return false;
if (IsImpactedInRange(&IntrBB->front(), &IntrI))
diff --git a/llvm/lib/Transforms/IPO/GlobalOpt.cpp b/llvm/lib/Transforms/IPO/GlobalOpt.cpp
index 7db0586386506..f830e53cfc532 100644
--- a/llvm/lib/Transforms/IPO/GlobalOpt.cpp
+++ b/llvm/lib/Transforms/IPO/GlobalOpt.cpp
@@ -1880,7 +1880,7 @@ static void RemovePreallocated(Function *F) {
Builder.SetInsertPoint(PreallocatedSetup);
auto *StackSave = Builder.CreateStackSave();
- Builder.SetInsertPoint(NewCB->getNextNonDebugInstruction());
+ Builder.SetInsertPoint(NewCB->getNextNode());
Builder.CreateStackRestore(StackSave);
// Replace @llvm.call.preallocated.arg() with alloca.
@@ -1904,7 +1904,7 @@ static void RemovePreallocated(Function *F) {
auto AddressSpace = UseCall->getType()->getPointerAddressSpace();
auto *ArgType =
UseCall->getFnAttr(Attribute::Preallocated).getValueAsType();
- auto *InsertBefore = PreallocatedSetup->getNextNonDebugInstruction();
+ auto *InsertBefore = PreallocatedSetup->getNextNode();
Builder.SetInsertPoint(InsertBefore);
auto *Alloca =
Builder.CreateAlloca(ArgType, AddressSpace, nullptr, "paarg");
diff --git a/llvm/lib/Transforms/IPO/IROutliner.cpp b/llvm/lib/Transforms/IPO/IROutliner.cpp
index cb18b55ae2183..bdb16ae812cf3 100644
--- a/llvm/lib/Transforms/IPO/IROutliner.cpp
+++ b/llvm/lib/Transforms/IPO/IROutliner.cpp
@@ -268,7 +268,7 @@ void OutlinableRegion::splitCandidate() {
// instruction. If they do not match, there could be problems in rewriting
// the program after outlining, so we ignore it.
if (!BackInst->isTerminator() &&
- EndInst != BackInst->getNextNonDebugInstruction())
+ EndInst != BackInst->getNextNode())
return;
Instruction *StartInst = (*Candidate->begin()).Inst;
@@ -2340,7 +2340,7 @@ static bool nextIRInstructionDataMatchesNextInst(IRInstructionData &ID) {
Instruction *NextIDLInst = NextIDIt->Inst;
Instruction *NextModuleInst = nullptr;
if (!ID.Inst->isTerminator())
- NextModuleInst = ID.Inst->getNextNonDebugInstruction();
+ NextModuleInst = ID.Inst->getNextNode();
else if (NextIDLInst != nullptr)
NextModuleInst =
&*NextIDIt->Inst->getParent()->instructionsWithoutDebug().begin();
@@ -2367,7 +2367,7 @@ bool IROutliner::isCompatibleWithAlreadyOutlinedCode(
// if it does not, we fix it in the InstructionDataList.
if (!Region.Candidate->backInstruction()->isTerminator()) {
Instruction *NewEndInst =
- Region.Candidate->backInstruction()->getNextNonDebugInstruction();
+ Region.Candidate->backInstruction()->getNextNode();
assert(NewEndInst && "Next instruction is a nullptr?");
if (Region.Candidate->end()->Inst != NewEndInst) {
IRInstructionDataList *IDL = Region.Candidate->front()->IDL;
diff --git a/llvm/lib/Transforms/IPO/OpenMPOpt.cpp b/llvm/lib/Transforms/IPO/OpenMPOpt.cpp
index dd7ae7e66e350..5de2285c2d2e3 100644
--- a/llvm/lib/Transforms/IPO/OpenMPOpt.cpp
+++ b/llvm/lib/Transforms/IPO/OpenMPOpt.cpp
@@ -2856,7 +2856,7 @@ struct AAExecutionDomainFunction : public AAExecutionDomain {
if (!It->getSecond().IsReachingAlignedBarrierOnly)
ForwardIsOk = false;
break;
- } while ((CurI = CurI->getNextNonDebugInstruction()));
+ } while ((CurI = CurI->getNextNode()));
if (!CurI && !BEDMap.lookup(I.getParent()).IsReachingAlignedBarrierOnly)
ForwardIsOk = false;
diff --git a/llvm/lib/Transforms/InstCombine/InstCombineCalls.cpp b/llvm/lib/Transforms/InstCombine/InstCombineCalls.cpp
index c169ab25b2106..1886ee781508e 100644
--- a/llvm/lib/Transforms/InstCombine/InstCombineCalls.cpp
+++ b/llvm/lib/Transforms/InstCombine/InstCombineCalls.cpp
@@ -3243,7 +3243,7 @@ Instruction *InstCombinerImpl::visitCallInst(CallInst &CI) {
// Remove an assume if it is followed by an identical assume.
// TODO: Do we need this? Unless there are conflicting assumptions, the
// computeKnownBits(IIOperand) below here eliminates redundant assumes.
- Instruction *Next = II->getNextNonDebugInstruction();
+ Instruction *Next = II->getNextNode();
if (match(Next, m_Intrinsic<Intrinsic::assume>(m_Specific(IIOperand))))
return RemoveConditionFromAssume(Next);
@@ -3416,12 +3416,12 @@ Instruction *InstCombinerImpl::visitCallInst(CallInst &CI) {
// Is this guard followed by another guard? We scan forward over a small
// fixed window of instructions to handle common cases with conditions
// computed between guards.
- Instruction *NextInst = II->getNextNonDebugInstruction();
+ Instruction *NextInst = II->getNextNode();
for (unsigned i = 0; i < GuardWideningWindow; i++) {
// Note: Using context-free form to avoid compile time blow up
if (!isSafeToSpeculativelyExecute(NextInst))
break;
- NextInst = NextInst->getNextNonDebugInstruction();
+ NextInst = NextInst->getNextNode();
}
Value *NextCond = nullptr;
if (match(NextInst,
@@ -3431,10 +3431,10 @@ Instruction *InstCombinerImpl::visitCallInst(CallInst &CI) {
// Remove a guard that it is immediately preceded by an identical guard.
// Otherwise canonicalize guard(a); guard(b) -> guard(a & b).
if (CurrCond != NextCond) {
- Instruction *MoveI = II->getNextNonDebugInstruction();
+ Instruction *MoveI = II->getNextNode();
while (MoveI != NextInst) {
auto *Temp = MoveI;
- MoveI = MoveI->getNextNonDebugInstruction();
+ MoveI = MoveI->getNextNode();
Temp->moveBefore(II->getIterator());
}
replaceOperand(*II, 0, Builder.CreateAnd(CurrCond, NextCond));
@@ -3871,7 +3871,7 @@ Instruction *InstCombinerImpl::visitCallInst(CallInst &CI) {
// Fence instruction simplification
Instruction *InstCombinerImpl::visitFenceInst(FenceInst &FI) {
- auto *NFI = dyn_cast<FenceInst>(FI.getNextNonDebugInstruction());
+ auto *NFI = dyn_cast<FenceInst>(FI.getNextNode());
// This check is solely here to handle arbitrary target-dependent syncscopes.
// TODO: Can remove if does not matter in practice.
if (NFI && FI.isIdenticalTo(NFI))
diff --git a/llvm/lib/Transforms/InstCombine/InstructionCombining.cpp b/llvm/lib/Transforms/InstCombine/InstructionCombining.cpp
index e261807bbc035..9c9dddb73318f 100644
--- a/llvm/lib/Transforms/InstCombine/InstructionCombining.cpp
+++ b/llvm/lib/Transforms/InstCombine/InstructionCombining.cpp
@@ -4760,7 +4760,7 @@ bool InstCombinerImpl::freezeOtherUses(FreezeInst &FI) {
// Don't move to the position of a debug intrinsic.
if (isa<DbgInfoIntrinsic>(MoveBefore))
- MoveBefore = MoveBefore->getNextNonDebugInstruction()->getIterator();
+ MoveBefore = MoveBefore->getNextNode()->getIterator();
// Re-point iterator to come after any debug-info records, if we're
// running in "RemoveDIs" mode
MoveBefore.setHeadBit(false);
diff --git a/llvm/lib/Transforms/Instrumentation/AddressSanitizer.cpp b/llvm/lib/Transforms/Instrumentation/AddressSanitizer.cpp
index 840a5e3f31dfd..dfbe4f8172066 100644
--- a/llvm/lib/Transforms/Instrumentation/AddressSanitizer.cpp
+++ b/llvm/lib/Transforms/Instrumentation/AddressSanitizer.cpp
@@ -3396,8 +3396,8 @@ void FunctionStackPoisoner::processDynamicAllocas() {
static void findStoresToUninstrumentedArgAllocas(
AddressSanitizer &ASan, Instruction &InsBefore,
SmallVectorImpl<Instruction *> &InitInsts) {
- Instruction *Start = InsBefore.getNextNonDebugInstruction();
- for (Instruction *It = Start; It; It = It->getNextNonDebugInstruction()) {
+ Instruction *Start = InsBefore.getNextNode();
+ for (Instruction *It = Start; It; It = It->getNextNode()) {
// Argument initialization looks like:
// 1) store <Argument>, <Alloca> OR
// 2) <CastArgument> = cast <Argument> to ...
diff --git a/llvm/lib/Transforms/Instrumentation/HWAddressSanitizer.cpp b/llvm/lib/Transforms/Instrumentation/HWAddressSanitizer.cpp
index 77db686f8229c..2c34bf2157cdd 100644
--- a/llvm/lib/Transforms/Instrumentation/HWAddressSanitizer.cpp
+++ b/llvm/lib/Transforms/Instrumentation/HWAddressSanitizer.cpp
@@ -1421,7 +1421,7 @@ void HWAddressSanitizer::emitPrologue(IRBuilder<> &IRB, bool WithFrameRecord) {
bool HWAddressSanitizer::instrum...
[truncated]
|
You can test this locally with the following command:git-clang-format --diff HEAD~1 HEAD --extensions cpp,h,inc -- clang/lib/CodeGen/CGBlocks.cpp llvm/include/llvm/IR/Instruction.h llvm/include/llvm/Transforms/Utils/LockstepReverseIterator.h llvm/lib/Analysis/MemoryDependenceAnalysis.cpp llvm/lib/CodeGen/CodeGenPrepare.cpp llvm/lib/IR/Instruction.cpp llvm/lib/IR/Verifier.cpp llvm/lib/Target/AArch64/AArch64TargetTransformInfo.cpp llvm/lib/Target/X86/X86WinEHState.cpp llvm/lib/Transforms/Coroutines/CoroSplit.cpp llvm/lib/Transforms/IPO/Attributor.cpp llvm/lib/Transforms/IPO/AttributorAttributes.cpp llvm/lib/Transforms/IPO/GlobalOpt.cpp llvm/lib/Transforms/IPO/IROutliner.cpp llvm/lib/Transforms/IPO/OpenMPOpt.cpp llvm/lib/Transforms/InstCombine/InstCombineCalls.cpp llvm/lib/Transforms/InstCombine/InstructionCombining.cpp llvm/lib/Transforms/Instrumentation/AddressSanitizer.cpp llvm/lib/Transforms/Instrumentation/HWAddressSanitizer.cpp llvm/lib/Transforms/Instrumentation/ValueProfilePlugins.inc llvm/lib/Transforms/Scalar/LoopStrengthReduce.cpp llvm/lib/Transforms/Scalar/MemCpyOptimizer.cpp llvm/lib/Transforms/Utils/Local.cpp llvm/lib/Transforms/Utils/ScalarEvolutionExpander.cpp llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp llvm/unittests/Frontend/OpenMPIRBuilderTest.cpp llvm/unittests/FuzzMutate/RandomIRBuilderTest.cpp llvm/unittests/IR/InstructionsTest.cpp llvm/unittests/Transforms/Utils/LocalTest.cppView the diff from clang-format here.diff --git a/llvm/lib/IR/Instruction.cpp b/llvm/lib/IR/Instruction.cpp
index c6dca727e..dcede5b6c 100644
--- a/llvm/lib/IR/Instruction.cpp
+++ b/llvm/lib/IR/Instruction.cpp
@@ -1243,9 +1243,7 @@ Instruction::getPrevNonDebugInstruction(bool SkipPseudoOp) const {
return nullptr;
}
-const DebugLoc &Instruction::getStableDebugLoc() const {
- return getDebugLoc();
-}
+const DebugLoc &Instruction::getStableDebugLoc() const { return getDebugLoc(); }
bool Instruction::isAssociative() const {
if (auto *II = dyn_cast<IntrinsicInst>(this))
diff --git a/llvm/lib/Transforms/Coroutines/CoroSplit.cpp b/llvm/lib/Transforms/Coroutines/CoroSplit.cpp
index 1dc49e221..39e2ec529 100644
--- a/llvm/lib/Transforms/Coroutines/CoroSplit.cpp
+++ b/llvm/lib/Transforms/Coroutines/CoroSplit.cpp
@@ -814,8 +814,7 @@ static void updateScopeLine(Instruction *ActiveSuspend,
return;
}
- BasicBlock::iterator Successor =
- ActiveSuspend->getNextNode()->getIterator();
+ BasicBlock::iterator Successor = ActiveSuspend->getNextNode()->getIterator();
// Corosplit splits the BB around ActiveSuspend, so the meaningful
// instructions are not in the same BB.
if (auto *Branch = dyn_cast_or_null<BranchInst>(Successor);
diff --git a/llvm/lib/Transforms/IPO/AttributorAttributes.cpp b/llvm/lib/Transforms/IPO/AttributorAttributes.cpp
index 4f5ba74e5..9c45e1e3b 100644
--- a/llvm/lib/Transforms/IPO/AttributorAttributes.cpp
+++ b/llvm/lib/Transforms/IPO/AttributorAttributes.cpp
@@ -1802,8 +1802,7 @@ ChangeStatus AAPointerInfoFloating::updateImpl(Attributor &A) {
continue;
return false;
}
- if (IsImpactedInRange(LoadI->getNextNode(),
- BB->getTerminator()))
+ if (IsImpactedInRange(LoadI->getNextNode(), BB->getTerminator()))
return false;
if (IsImpactedInRange(&IntrBB->front(), &IntrI))
return false;
diff --git a/llvm/lib/Transforms/IPO/IROutliner.cpp b/llvm/lib/Transforms/IPO/IROutliner.cpp
index bdb16ae81..25a0804a0 100644
--- a/llvm/lib/Transforms/IPO/IROutliner.cpp
+++ b/llvm/lib/Transforms/IPO/IROutliner.cpp
@@ -267,8 +267,7 @@ void OutlinableRegion::splitCandidate() {
// region is the same as the recorded instruction following the last
// instruction. If they do not match, there could be problems in rewriting
// the program after outlining, so we ignore it.
- if (!BackInst->isTerminator() &&
- EndInst != BackInst->getNextNode())
+ if (!BackInst->isTerminator() && EndInst != BackInst->getNextNode())
return;
Instruction *StartInst = (*Candidate->begin()).Inst;
diff --git a/llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp b/llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp
index c68d9249b..b58bdae6b 100644
--- a/llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp
+++ b/llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp
@@ -16162,9 +16162,8 @@ void BoUpSLP::setInsertPointAfterBundle(const TreeEntry *E) {
} else {
// Set the insertion point after the last instruction in the bundle. Set the
// debug location to Front.
- Builder.SetInsertPoint(
- LastInst->getParent(),
- LastInst->getNextNode()->getIterator());
+ Builder.SetInsertPoint(LastInst->getParent(),
+ LastInst->getNextNode()->getIterator());
}
Builder.SetCurrentDebugLocation(Front->getDebugLoc());
}
diff --git a/llvm/unittests/Frontend/OpenMPIRBuilderTest.cpp b/llvm/unittests/Frontend/OpenMPIRBuilderTest.cpp
index 558af6900..2e2044397 100644
--- a/llvm/unittests/Frontend/OpenMPIRBuilderTest.cpp
+++ b/llvm/unittests/Frontend/OpenMPIRBuilderTest.cpp
@@ -7425,8 +7425,7 @@ TEST_F(OpenMPIRBuilderTest, CreateTaskgroup) {
// Checking the general structure of the IR generated is same as expected.
Instruction *GeneratedStoreInst = TaskgroupCall->getNextNode();
EXPECT_EQ(GeneratedStoreInst, InternalStoreInst);
- Instruction *GeneratedLoad32 =
- GeneratedStoreInst->getNextNode();
+ Instruction *GeneratedLoad32 = GeneratedStoreInst->getNextNode();
EXPECT_EQ(GeneratedLoad32, InternalLoad32);
Instruction *GeneratedLoad128 = GeneratedLoad32->getNextNode();
EXPECT_EQ(GeneratedLoad128, InternalLoad128);
|
OCHyams
approved these changes
Jun 17, 2025
| @@ -29,7 +29,7 @@ to learn about the defaults for your platform and target. | |||
| - ``LIBC_CONF_ENABLE_STRONG_STACK_PROTECTOR``: Enable -fstack-protector-strong to defend against stack smashing attack. | |||
| - ``LIBC_CONF_KEEP_FRAME_POINTER``: Keep frame pointer in functions for better debugging experience. | |||
| * **"errno" options** | |||
| - ``LIBC_CONF_ERRNO_MODE``: The implementation used for errno, acceptable values are LIBC_ERRNO_MODE_DEFAULT, LIBC_ERRNO_MODE_UNDEFINED, LIBC_ERRNO_MODE_THREAD_LOCAL, LIBC_ERRNO_MODE_SHARED, LIBC_ERRNO_MODE_EXTERNAL, and LIBC_ERRNO_MODE_SYSTEM. | |||
| - ``LIBC_CONF_ERRNO_MODE``: The implementation used for errno, acceptable values are LIBC_ERRNO_MODE_DEFAULT, LIBC_ERRNO_MODE_UNDEFINED, LIBC_ERRNO_MODE_THREAD_LOCAL, LIBC_ERRNO_MODE_SHARED, LIBC_ERRNO_MODE_EXTERNAL, LIBC_ERRNO_MODE_SYSTEM, and LIBC_ERRNO_MODE_SYSTEM_INLINE. | |||
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.
There are no longer debug-info instructions, thus we don't need this skipping. Horray!
CC @WenleiHe ,
getNextNonDebugInfoInstructionhas a "skip pseudo-probes" argument that appears to be unused throughout all of upstream LLVM. We'd really like to get rid of everything to do with debug-intrinsics to reduce maintenance burden everywhere -- please let us know if this is going to interfere with anything you've got doing on downstream.