llvm-project/llvm/lib/CodeGen/DwarfEHPrepare.cpp

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

356 lines
11 KiB
C++
Raw Normal View History

//===- DwarfEHPrepare - Prepare exception handling for code generation ----===//
Add a new codegen pass that normalizes dwarf exception handling code in preparation for code generation. The main thing it does is handle the case when eh.exception calls (and, in a future patch, eh.selector calls) are far away from landing pads. Right now in practice you only find eh.exception calls close to landing pads: either in a landing pad (the common case) or in a landing pad successor, due to loop passes shifting them about. However future exception handling improvements will result in calls far from landing pads: (1) Inlining of rewinds. Consider the following case: In function @f: ... invoke @g to label %normal unwind label %unwinds ... unwinds: %ex = call i8* @llvm.eh.exception() ... In function @g: ... invoke @something to label %continue unwind label %handler ... handler: %ex = call i8* @llvm.eh.exception() ... perform cleanups ... "rethrow exception" Now inline @g into @f. Currently this is turned into: In function @f: ... invoke @something to label %continue unwind label %handler ... handler: %ex = call i8* @llvm.eh.exception() ... perform cleanups ... invoke "rethrow exception" to label %normal unwind label %unwinds unwinds: %ex = call i8* @llvm.eh.exception() ... However we would like to simplify invoke of "rethrow exception" into a branch to the %unwinds label. Then %unwinds is no longer a landing pad, and the eh.exception call there is then far away from any landing pads. (2) Using the unwind instruction for cleanups. It would be nice to have codegen handle the following case: invoke @something to label %continue unwind label %run_cleanups ... handler: ... perform cleanups ... unwind This requires turning "unwind" into a library call, which necessarily takes a pointer to the exception as an argument (this patch also does this unwind lowering). But that means you are using eh.exception again far from a landing pad. (3) Bugpoint simplifications. When bugpoint is simplifying exception handling code it often generates eh.exception calls far from a landing pad, which then causes codegen to assert. Bugpoint then latches on to this assertion and loses sight of the original problem. Note that it is currently rare for this pass to actually do anything. And in fact it normally shouldn't do anything at all given the code coming out of llvm-gcc! But it does fire a few times in the testsuite. As far as I can see this is almost always due to the LoopStrengthReduce codegen pass introducing pointless loop preheader blocks which are landing pads and only contain a branch to another block. This other block contains an eh.exception call. So probably by tweaking LoopStrengthReduce a bit this can be avoided. llvm-svn: 72276
2009-05-22 20:36:31 +00:00
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
Add a new codegen pass that normalizes dwarf exception handling code in preparation for code generation. The main thing it does is handle the case when eh.exception calls (and, in a future patch, eh.selector calls) are far away from landing pads. Right now in practice you only find eh.exception calls close to landing pads: either in a landing pad (the common case) or in a landing pad successor, due to loop passes shifting them about. However future exception handling improvements will result in calls far from landing pads: (1) Inlining of rewinds. Consider the following case: In function @f: ... invoke @g to label %normal unwind label %unwinds ... unwinds: %ex = call i8* @llvm.eh.exception() ... In function @g: ... invoke @something to label %continue unwind label %handler ... handler: %ex = call i8* @llvm.eh.exception() ... perform cleanups ... "rethrow exception" Now inline @g into @f. Currently this is turned into: In function @f: ... invoke @something to label %continue unwind label %handler ... handler: %ex = call i8* @llvm.eh.exception() ... perform cleanups ... invoke "rethrow exception" to label %normal unwind label %unwinds unwinds: %ex = call i8* @llvm.eh.exception() ... However we would like to simplify invoke of "rethrow exception" into a branch to the %unwinds label. Then %unwinds is no longer a landing pad, and the eh.exception call there is then far away from any landing pads. (2) Using the unwind instruction for cleanups. It would be nice to have codegen handle the following case: invoke @something to label %continue unwind label %run_cleanups ... handler: ... perform cleanups ... unwind This requires turning "unwind" into a library call, which necessarily takes a pointer to the exception as an argument (this patch also does this unwind lowering). But that means you are using eh.exception again far from a landing pad. (3) Bugpoint simplifications. When bugpoint is simplifying exception handling code it often generates eh.exception calls far from a landing pad, which then causes codegen to assert. Bugpoint then latches on to this assertion and loses sight of the original problem. Note that it is currently rare for this pass to actually do anything. And in fact it normally shouldn't do anything at all given the code coming out of llvm-gcc! But it does fire a few times in the testsuite. As far as I can see this is almost always due to the LoopStrengthReduce codegen pass introducing pointless loop preheader blocks which are landing pads and only contain a branch to another block. This other block contains an eh.exception call. So probably by tweaking LoopStrengthReduce a bit this can be avoided. llvm-svn: 72276
2009-05-22 20:36:31 +00:00
//
//===----------------------------------------------------------------------===//
//
// This pass mulches exception handling code into a form adapted to code
// generation. Required if using dwarf exception handling.
Add a new codegen pass that normalizes dwarf exception handling code in preparation for code generation. The main thing it does is handle the case when eh.exception calls (and, in a future patch, eh.selector calls) are far away from landing pads. Right now in practice you only find eh.exception calls close to landing pads: either in a landing pad (the common case) or in a landing pad successor, due to loop passes shifting them about. However future exception handling improvements will result in calls far from landing pads: (1) Inlining of rewinds. Consider the following case: In function @f: ... invoke @g to label %normal unwind label %unwinds ... unwinds: %ex = call i8* @llvm.eh.exception() ... In function @g: ... invoke @something to label %continue unwind label %handler ... handler: %ex = call i8* @llvm.eh.exception() ... perform cleanups ... "rethrow exception" Now inline @g into @f. Currently this is turned into: In function @f: ... invoke @something to label %continue unwind label %handler ... handler: %ex = call i8* @llvm.eh.exception() ... perform cleanups ... invoke "rethrow exception" to label %normal unwind label %unwinds unwinds: %ex = call i8* @llvm.eh.exception() ... However we would like to simplify invoke of "rethrow exception" into a branch to the %unwinds label. Then %unwinds is no longer a landing pad, and the eh.exception call there is then far away from any landing pads. (2) Using the unwind instruction for cleanups. It would be nice to have codegen handle the following case: invoke @something to label %continue unwind label %run_cleanups ... handler: ... perform cleanups ... unwind This requires turning "unwind" into a library call, which necessarily takes a pointer to the exception as an argument (this patch also does this unwind lowering). But that means you are using eh.exception again far from a landing pad. (3) Bugpoint simplifications. When bugpoint is simplifying exception handling code it often generates eh.exception calls far from a landing pad, which then causes codegen to assert. Bugpoint then latches on to this assertion and loses sight of the original problem. Note that it is currently rare for this pass to actually do anything. And in fact it normally shouldn't do anything at all given the code coming out of llvm-gcc! But it does fire a few times in the testsuite. As far as I can see this is almost always due to the LoopStrengthReduce codegen pass introducing pointless loop preheader blocks which are landing pads and only contain a branch to another block. This other block contains an eh.exception call. So probably by tweaking LoopStrengthReduce a bit this can be avoided. llvm-svn: 72276
2009-05-22 20:36:31 +00:00
//
//===----------------------------------------------------------------------===//
#include "llvm/ADT/BitVector.h"
#include "llvm/ADT/SmallVector.h"
#include "llvm/ADT/Statistic.h"
#include "llvm/Analysis/CFG.h"
#include "llvm/Analysis/DomTreeUpdater.h"
#include "llvm/Analysis/EHPersonalities.h"
#include "llvm/Analysis/TargetTransformInfo.h"
#include "llvm/CodeGen/RuntimeLibcalls.h"
#include "llvm/CodeGen/TargetLowering.h"
#include "llvm/CodeGen/TargetPassConfig.h"
#include "llvm/CodeGen/TargetSubtargetInfo.h"
#include "llvm/IR/BasicBlock.h"
#include "llvm/IR/Constants.h"
#include "llvm/IR/DerivedTypes.h"
#include "llvm/IR/Dominators.h"
#include "llvm/IR/Function.h"
#include "llvm/IR/Instructions.h"
#include "llvm/IR/Module.h"
#include "llvm/IR/Type.h"
#include "llvm/InitializePasses.h"
Add a new codegen pass that normalizes dwarf exception handling code in preparation for code generation. The main thing it does is handle the case when eh.exception calls (and, in a future patch, eh.selector calls) are far away from landing pads. Right now in practice you only find eh.exception calls close to landing pads: either in a landing pad (the common case) or in a landing pad successor, due to loop passes shifting them about. However future exception handling improvements will result in calls far from landing pads: (1) Inlining of rewinds. Consider the following case: In function @f: ... invoke @g to label %normal unwind label %unwinds ... unwinds: %ex = call i8* @llvm.eh.exception() ... In function @g: ... invoke @something to label %continue unwind label %handler ... handler: %ex = call i8* @llvm.eh.exception() ... perform cleanups ... "rethrow exception" Now inline @g into @f. Currently this is turned into: In function @f: ... invoke @something to label %continue unwind label %handler ... handler: %ex = call i8* @llvm.eh.exception() ... perform cleanups ... invoke "rethrow exception" to label %normal unwind label %unwinds unwinds: %ex = call i8* @llvm.eh.exception() ... However we would like to simplify invoke of "rethrow exception" into a branch to the %unwinds label. Then %unwinds is no longer a landing pad, and the eh.exception call there is then far away from any landing pads. (2) Using the unwind instruction for cleanups. It would be nice to have codegen handle the following case: invoke @something to label %continue unwind label %run_cleanups ... handler: ... perform cleanups ... unwind This requires turning "unwind" into a library call, which necessarily takes a pointer to the exception as an argument (this patch also does this unwind lowering). But that means you are using eh.exception again far from a landing pad. (3) Bugpoint simplifications. When bugpoint is simplifying exception handling code it often generates eh.exception calls far from a landing pad, which then causes codegen to assert. Bugpoint then latches on to this assertion and loses sight of the original problem. Note that it is currently rare for this pass to actually do anything. And in fact it normally shouldn't do anything at all given the code coming out of llvm-gcc! But it does fire a few times in the testsuite. As far as I can see this is almost always due to the LoopStrengthReduce codegen pass introducing pointless loop preheader blocks which are landing pads and only contain a branch to another block. This other block contains an eh.exception call. So probably by tweaking LoopStrengthReduce a bit this can be avoided. llvm-svn: 72276
2009-05-22 20:36:31 +00:00
#include "llvm/Pass.h"
#include "llvm/Support/Casting.h"
#include "llvm/Target/TargetMachine.h"
#include "llvm/Transforms/Utils/Local.h"
#include <cstddef>
Add a new codegen pass that normalizes dwarf exception handling code in preparation for code generation. The main thing it does is handle the case when eh.exception calls (and, in a future patch, eh.selector calls) are far away from landing pads. Right now in practice you only find eh.exception calls close to landing pads: either in a landing pad (the common case) or in a landing pad successor, due to loop passes shifting them about. However future exception handling improvements will result in calls far from landing pads: (1) Inlining of rewinds. Consider the following case: In function @f: ... invoke @g to label %normal unwind label %unwinds ... unwinds: %ex = call i8* @llvm.eh.exception() ... In function @g: ... invoke @something to label %continue unwind label %handler ... handler: %ex = call i8* @llvm.eh.exception() ... perform cleanups ... "rethrow exception" Now inline @g into @f. Currently this is turned into: In function @f: ... invoke @something to label %continue unwind label %handler ... handler: %ex = call i8* @llvm.eh.exception() ... perform cleanups ... invoke "rethrow exception" to label %normal unwind label %unwinds unwinds: %ex = call i8* @llvm.eh.exception() ... However we would like to simplify invoke of "rethrow exception" into a branch to the %unwinds label. Then %unwinds is no longer a landing pad, and the eh.exception call there is then far away from any landing pads. (2) Using the unwind instruction for cleanups. It would be nice to have codegen handle the following case: invoke @something to label %continue unwind label %run_cleanups ... handler: ... perform cleanups ... unwind This requires turning "unwind" into a library call, which necessarily takes a pointer to the exception as an argument (this patch also does this unwind lowering). But that means you are using eh.exception again far from a landing pad. (3) Bugpoint simplifications. When bugpoint is simplifying exception handling code it often generates eh.exception calls far from a landing pad, which then causes codegen to assert. Bugpoint then latches on to this assertion and loses sight of the original problem. Note that it is currently rare for this pass to actually do anything. And in fact it normally shouldn't do anything at all given the code coming out of llvm-gcc! But it does fire a few times in the testsuite. As far as I can see this is almost always due to the LoopStrengthReduce codegen pass introducing pointless loop preheader blocks which are landing pads and only contain a branch to another block. This other block contains an eh.exception call. So probably by tweaking LoopStrengthReduce a bit this can be avoided. llvm-svn: 72276
2009-05-22 20:36:31 +00:00
using namespace llvm;
#define DEBUG_TYPE "dwarfehprepare"
STATISTIC(NumResumesLowered, "Number of resume calls lowered");
STATISTIC(NumCleanupLandingPadsUnreachable,
"Number of cleanup landing pads found unreachable");
STATISTIC(NumCleanupLandingPadsRemaining,
"Number of cleanup landing pads remaining");
STATISTIC(NumNoUnwind, "Number of functions with nounwind");
STATISTIC(NumUnwind, "Number of functions with unwind");
Add a new codegen pass that normalizes dwarf exception handling code in preparation for code generation. The main thing it does is handle the case when eh.exception calls (and, in a future patch, eh.selector calls) are far away from landing pads. Right now in practice you only find eh.exception calls close to landing pads: either in a landing pad (the common case) or in a landing pad successor, due to loop passes shifting them about. However future exception handling improvements will result in calls far from landing pads: (1) Inlining of rewinds. Consider the following case: In function @f: ... invoke @g to label %normal unwind label %unwinds ... unwinds: %ex = call i8* @llvm.eh.exception() ... In function @g: ... invoke @something to label %continue unwind label %handler ... handler: %ex = call i8* @llvm.eh.exception() ... perform cleanups ... "rethrow exception" Now inline @g into @f. Currently this is turned into: In function @f: ... invoke @something to label %continue unwind label %handler ... handler: %ex = call i8* @llvm.eh.exception() ... perform cleanups ... invoke "rethrow exception" to label %normal unwind label %unwinds unwinds: %ex = call i8* @llvm.eh.exception() ... However we would like to simplify invoke of "rethrow exception" into a branch to the %unwinds label. Then %unwinds is no longer a landing pad, and the eh.exception call there is then far away from any landing pads. (2) Using the unwind instruction for cleanups. It would be nice to have codegen handle the following case: invoke @something to label %continue unwind label %run_cleanups ... handler: ... perform cleanups ... unwind This requires turning "unwind" into a library call, which necessarily takes a pointer to the exception as an argument (this patch also does this unwind lowering). But that means you are using eh.exception again far from a landing pad. (3) Bugpoint simplifications. When bugpoint is simplifying exception handling code it often generates eh.exception calls far from a landing pad, which then causes codegen to assert. Bugpoint then latches on to this assertion and loses sight of the original problem. Note that it is currently rare for this pass to actually do anything. And in fact it normally shouldn't do anything at all given the code coming out of llvm-gcc! But it does fire a few times in the testsuite. As far as I can see this is almost always due to the LoopStrengthReduce codegen pass introducing pointless loop preheader blocks which are landing pads and only contain a branch to another block. This other block contains an eh.exception call. So probably by tweaking LoopStrengthReduce a bit this can be avoided. llvm-svn: 72276
2009-05-22 20:36:31 +00:00
namespace {
class DwarfEHPrepare {
CodeGenOpt::Level OptLevel;
// RewindFunction - _Unwind_Resume or the target equivalent.
FunctionCallee &RewindFunction;
Function &F;
const TargetLowering &TLI;
DomTreeUpdater *DTU;
const TargetTransformInfo *TTI;
/// Return the exception object from the value passed into
/// the 'resume' instruction (typically an aggregate). Clean up any dead
/// instructions, including the 'resume' instruction.
Value *GetExceptionObject(ResumeInst *RI);
/// Replace resumes that are not reachable from a cleanup landing pad with
/// unreachable and then simplify those blocks.
size_t
pruneUnreachableResumes(SmallVectorImpl<ResumeInst *> &Resumes,
SmallVectorImpl<LandingPadInst *> &CleanupLPads);
/// Convert the ResumeInsts that are still present
/// into calls to the appropriate _Unwind_Resume function.
bool InsertUnwindResumeCalls();
public:
DwarfEHPrepare(CodeGenOpt::Level OptLevel_, FunctionCallee &RewindFunction_,
Function &F_, const TargetLowering &TLI_, DomTreeUpdater *DTU_,
const TargetTransformInfo *TTI_)
: OptLevel(OptLevel_), RewindFunction(RewindFunction_), F(F_), TLI(TLI_),
DTU(DTU_), TTI(TTI_) {}
bool run();
};
} // namespace
Add a new codegen pass that normalizes dwarf exception handling code in preparation for code generation. The main thing it does is handle the case when eh.exception calls (and, in a future patch, eh.selector calls) are far away from landing pads. Right now in practice you only find eh.exception calls close to landing pads: either in a landing pad (the common case) or in a landing pad successor, due to loop passes shifting them about. However future exception handling improvements will result in calls far from landing pads: (1) Inlining of rewinds. Consider the following case: In function @f: ... invoke @g to label %normal unwind label %unwinds ... unwinds: %ex = call i8* @llvm.eh.exception() ... In function @g: ... invoke @something to label %continue unwind label %handler ... handler: %ex = call i8* @llvm.eh.exception() ... perform cleanups ... "rethrow exception" Now inline @g into @f. Currently this is turned into: In function @f: ... invoke @something to label %continue unwind label %handler ... handler: %ex = call i8* @llvm.eh.exception() ... perform cleanups ... invoke "rethrow exception" to label %normal unwind label %unwinds unwinds: %ex = call i8* @llvm.eh.exception() ... However we would like to simplify invoke of "rethrow exception" into a branch to the %unwinds label. Then %unwinds is no longer a landing pad, and the eh.exception call there is then far away from any landing pads. (2) Using the unwind instruction for cleanups. It would be nice to have codegen handle the following case: invoke @something to label %continue unwind label %run_cleanups ... handler: ... perform cleanups ... unwind This requires turning "unwind" into a library call, which necessarily takes a pointer to the exception as an argument (this patch also does this unwind lowering). But that means you are using eh.exception again far from a landing pad. (3) Bugpoint simplifications. When bugpoint is simplifying exception handling code it often generates eh.exception calls far from a landing pad, which then causes codegen to assert. Bugpoint then latches on to this assertion and loses sight of the original problem. Note that it is currently rare for this pass to actually do anything. And in fact it normally shouldn't do anything at all given the code coming out of llvm-gcc! But it does fire a few times in the testsuite. As far as I can see this is almost always due to the LoopStrengthReduce codegen pass introducing pointless loop preheader blocks which are landing pads and only contain a branch to another block. This other block contains an eh.exception call. So probably by tweaking LoopStrengthReduce a bit this can be avoided. llvm-svn: 72276
2009-05-22 20:36:31 +00:00
Value *DwarfEHPrepare::GetExceptionObject(ResumeInst *RI) {
Value *V = RI->getOperand(0);
Value *ExnObj = nullptr;
InsertValueInst *SelIVI = dyn_cast<InsertValueInst>(V);
LoadInst *SelLoad = nullptr;
InsertValueInst *ExcIVI = nullptr;
bool EraseIVIs = false;
if (SelIVI) {
if (SelIVI->getNumIndices() == 1 && *SelIVI->idx_begin() == 1) {
ExcIVI = dyn_cast<InsertValueInst>(SelIVI->getOperand(0));
if (ExcIVI && isa<UndefValue>(ExcIVI->getOperand(0)) &&
ExcIVI->getNumIndices() == 1 && *ExcIVI->idx_begin() == 0) {
ExnObj = ExcIVI->getOperand(1);
SelLoad = dyn_cast<LoadInst>(SelIVI->getOperand(1));
EraseIVIs = true;
}
}
}
if (!ExnObj)
ExnObj = ExtractValueInst::Create(RI->getOperand(0), 0, "exn.obj", RI);
RI->eraseFromParent();
if (EraseIVIs) {
if (SelIVI->use_empty())
SelIVI->eraseFromParent();
if (ExcIVI->use_empty())
ExcIVI->eraseFromParent();
if (SelLoad && SelLoad->use_empty())
SelLoad->eraseFromParent();
}
return ExnObj;
}
size_t DwarfEHPrepare::pruneUnreachableResumes(
SmallVectorImpl<ResumeInst *> &Resumes,
SmallVectorImpl<LandingPadInst *> &CleanupLPads) {
assert(DTU && "Should have DomTreeUpdater here.");
BitVector ResumeReachable(Resumes.size());
size_t ResumeIndex = 0;
for (auto *RI : Resumes) {
for (auto *LP : CleanupLPads) {
if (isPotentiallyReachable(LP, RI, nullptr, &DTU->getDomTree())) {
ResumeReachable.set(ResumeIndex);
break;
}
}
++ResumeIndex;
}
// If everything is reachable, there is no change.
if (ResumeReachable.all())
return Resumes.size();
LLVMContext &Ctx = F.getContext();
// Otherwise, insert unreachable instructions and call simplifycfg.
size_t ResumesLeft = 0;
for (size_t I = 0, E = Resumes.size(); I < E; ++I) {
ResumeInst *RI = Resumes[I];
if (ResumeReachable[I]) {
Resumes[ResumesLeft++] = RI;
} else {
BasicBlock *BB = RI->getParent();
new UnreachableInst(Ctx, RI);
RI->eraseFromParent();
simplifyCFG(BB, *TTI, DTU);
}
}
Resumes.resize(ResumesLeft);
return ResumesLeft;
}
bool DwarfEHPrepare::InsertUnwindResumeCalls() {
SmallVector<ResumeInst *, 16> Resumes;
SmallVector<LandingPadInst *, 16> CleanupLPads;
if (F.doesNotThrow())
NumNoUnwind++;
else
NumUnwind++;
for (BasicBlock &BB : F) {
if (auto *RI = dyn_cast<ResumeInst>(BB.getTerminator()))
Resumes.push_back(RI);
if (auto *LP = BB.getLandingPadInst())
if (LP->isCleanup())
CleanupLPads.push_back(LP);
}
NumCleanupLandingPadsRemaining += CleanupLPads.size();
if (Resumes.empty())
return false;
// Check the personality, don't do anything if it's scope-based.
EHPersonality Pers = classifyEHPersonality(F.getPersonalityFn());
if (isScopedEHPersonality(Pers))
return false;
LLVMContext &Ctx = F.getContext();
size_t ResumesLeft = Resumes.size();
if (OptLevel != CodeGenOpt::None) {
ResumesLeft = pruneUnreachableResumes(Resumes, CleanupLPads);
#if LLVM_ENABLE_STATS
unsigned NumRemainingLPs = 0;
for (BasicBlock &BB : F) {
if (auto *LP = BB.getLandingPadInst())
if (LP->isCleanup())
NumRemainingLPs++;
}
NumCleanupLandingPadsUnreachable += CleanupLPads.size() - NumRemainingLPs;
NumCleanupLandingPadsRemaining -= CleanupLPads.size() - NumRemainingLPs;
#endif
}
if (ResumesLeft == 0)
return true; // We pruned them all.
// Find the rewind function if we didn't already.
if (!RewindFunction) {
FunctionType *FTy =
FunctionType::get(Type::getVoidTy(Ctx), Type::getInt8PtrTy(Ctx), false);
const char *RewindName = TLI.getLibcallName(RTLIB::UNWIND_RESUME);
RewindFunction = F.getParent()->getOrInsertFunction(RewindName, FTy);
}
// Create the basic block where the _Unwind_Resume call will live.
if (ResumesLeft == 1) {
// Instead of creating a new BB and PHI node, just append the call to
// _Unwind_Resume to the end of the single resume block.
ResumeInst *RI = Resumes.front();
BasicBlock *UnwindBB = RI->getParent();
Value *ExnObj = GetExceptionObject(RI);
// Call the _Unwind_Resume function.
CallInst *CI = CallInst::Create(RewindFunction, ExnObj, "", UnwindBB);
CI->setCallingConv(TLI.getLibcallCallingConv(RTLIB::UNWIND_RESUME));
// We never expect _Unwind_Resume to return.
CI->setDoesNotReturn();
new UnreachableInst(Ctx, UnwindBB);
return true;
}
std::vector<DominatorTree::UpdateType> Updates;
Updates.reserve(Resumes.size());
BasicBlock *UnwindBB = BasicBlock::Create(Ctx, "unwind_resume", &F);
PHINode *PN = PHINode::Create(Type::getInt8PtrTy(Ctx), ResumesLeft, "exn.obj",
UnwindBB);
// Extract the exception object from the ResumeInst and add it to the PHI node
// that feeds the _Unwind_Resume call.
for (ResumeInst *RI : Resumes) {
BasicBlock *Parent = RI->getParent();
BranchInst::Create(UnwindBB, Parent);
Updates.push_back({DominatorTree::Insert, Parent, UnwindBB});
Value *ExnObj = GetExceptionObject(RI);
PN->addIncoming(ExnObj, Parent);
++NumResumesLowered;
}
// Call the function.
CallInst *CI = CallInst::Create(RewindFunction, PN, "", UnwindBB);
CI->setCallingConv(TLI.getLibcallCallingConv(RTLIB::UNWIND_RESUME));
// We never expect _Unwind_Resume to return.
CI->setDoesNotReturn();
new UnreachableInst(Ctx, UnwindBB);
if (DTU)
DTU->applyUpdates(Updates);
return true;
}
bool DwarfEHPrepare::run() {
bool Changed = InsertUnwindResumeCalls();
return Changed;
}
static bool prepareDwarfEH(CodeGenOpt::Level OptLevel,
FunctionCallee &RewindFunction, Function &F,
const TargetLowering &TLI, DominatorTree *DT,
const TargetTransformInfo *TTI) {
DomTreeUpdater DTU(DT, DomTreeUpdater::UpdateStrategy::Lazy);
return DwarfEHPrepare(OptLevel, RewindFunction, F, TLI, DT ? &DTU : nullptr,
TTI)
.run();
}
namespace {
class DwarfEHPrepareLegacyPass : public FunctionPass {
// RewindFunction - _Unwind_Resume or the target equivalent.
FunctionCallee RewindFunction = nullptr;
CodeGenOpt::Level OptLevel;
public:
static char ID; // Pass identification, replacement for typeid.
DwarfEHPrepareLegacyPass(CodeGenOpt::Level OptLevel = CodeGenOpt::Default)
: FunctionPass(ID), OptLevel(OptLevel) {}
bool runOnFunction(Function &F) override {
const TargetMachine &TM =
getAnalysis<TargetPassConfig>().getTM<TargetMachine>();
const TargetLowering &TLI = *TM.getSubtargetImpl(F)->getTargetLowering();
DominatorTree *DT = nullptr;
const TargetTransformInfo *TTI = nullptr;
if (auto *DTWP = getAnalysisIfAvailable<DominatorTreeWrapperPass>())
DT = &DTWP->getDomTree();
if (OptLevel != CodeGenOpt::None) {
if (!DT)
DT = &getAnalysis<DominatorTreeWrapperPass>().getDomTree();
TTI = &getAnalysis<TargetTransformInfoWrapperPass>().getTTI(F);
}
return prepareDwarfEH(OptLevel, RewindFunction, F, TLI, DT, TTI);
}
bool doFinalization(Module &M) override {
RewindFunction = nullptr;
return false;
Re-apply the fix on DwarfEHPrepare and add a test This patch re-introduces the fix in the commit https://github.com/llvm/llvm-project/commit/66b0cebf7f736 by @yrnkrn > In DwarfEHPrepare, after all passes are run, RewindFunction may be a dangling > > pointer to a dead function. To make sure it's valid, doFinalization nullptrs > RewindFunction just like the constructor and so it will be found on next run. > > llvm-svn: 217737 It seems that the fix was not migrated to `DwarfEHPrepareLegacyPass`. This patch also updates `llvm/test/CodeGen/X86/dwarf-eh-prepare.ll` to include `-run-twice` to exercise the cleanup. Without this patch `llvm-lit -v llvm/test/CodeGen/X86/dwarf-eh-prepare.ll` fails with ``` -- Testing: 1 tests, 1 workers -- FAIL: LLVM :: CodeGen/X86/dwarf-eh-prepare.ll (1 of 1) ******************** TEST 'LLVM :: CodeGen/X86/dwarf-eh-prepare.ll' FAILED ******************** Script: -- : 'RUN: at line 1'; /home/arakaki/build/llvm-project/main/bin/opt -mtriple=x86_64-linux-gnu -dwarfehprepare -simplifycfg-require-and-preserve-domtree=1 -run-twice < /home/arakaki/repos/watch/llvm-project/llvm/test/CodeGen/X86/dwarf-eh-prepare.ll -S | /home/arakaki/build/llvm-project/main/bin/FileCheck /home/arakaki/repos/watch/llvm-project/llvm/test/CodeGen/X86/dwarf-eh-prepare.ll -- Exit Code: 2 Command Output (stderr): -- Referencing function in another module! call void @_Unwind_Resume(i8* %ehptr) #1 ; ModuleID = '<stdin>' void (i8*)* @_Unwind_Resume ; ModuleID = '<stdin>' in function simple_cleanup_catch LLVM ERROR: Broken function found, compilation aborted! PLEASE submit a bug report to https://bugs.llvm.org/ and include the crash backtrace. Stack dump: 0. Program arguments: /home/arakaki/build/llvm-project/main/bin/opt -mtriple=x86_64-linux-gnu -dwarfehprepare -simplifycfg-require-and-preserve-domtree=1 -run-twice -S 1. Running pass 'Function Pass Manager' on module '<stdin>'. 2. Running pass 'Module Verifier' on function '@simple_cleanup_catch' #0 0x000056121b570a2c llvm::sys::PrintStackTrace(llvm::raw_ostream&, int) /home/arakaki/repos/watch/llvm-project/llvm/lib/Support/Unix/Signals.inc:569:0 #1 0x000056121b56eb64 llvm::sys::RunSignalHandlers() /home/arakaki/repos/watch/llvm-project/llvm/lib/Support/Signals.cpp:97:0 #2 0x000056121b56f28e SignalHandler(int) /home/arakaki/repos/watch/llvm-project/llvm/lib/Support/Unix/Signals.inc:397:0 #3 0x00007fc7e9b22980 __restore_rt (/lib/x86_64-linux-gnu/libpthread.so.0+0x12980) #4 0x00007fc7e87d3fb7 raise /build/glibc-S7xCS9/glibc-2.27/signal/../sysdeps/unix/sysv/linux/raise.c:51:0 #5 0x00007fc7e87d5921 abort /build/glibc-S7xCS9/glibc-2.27/stdlib/abort.c:81:0 #6 0x000056121b4e1386 llvm::raw_svector_ostream::raw_svector_ostream(llvm::SmallVectorImpl<char>&) /home/arakaki/repos/watch/llvm-project/llvm/include/llvm/Support/raw_ostream.h:674:0 #7 0x000056121b4e1386 llvm::report_fatal_error(llvm::Twine const&, bool) /home/arakaki/repos/watch/llvm-project/llvm/lib/Support/ErrorHandling.cpp:114:0 #8 0x000056121b4e1528 (/home/arakaki/build/llvm-project/main/bin/opt+0x29e3528) #9 0x000056121adfd03f llvm::raw_ostream::operator<<(llvm::StringRef) /home/arakaki/repos/watch/llvm-project/llvm/include/llvm/Support/raw_ostream.h:218:0 FileCheck error: '<stdin>' is empty. FileCheck command line: /home/arakaki/build/llvm-project/main/bin/FileCheck /home/arakaki/repos/watch/llvm-project/llvm/test/CodeGen/X86/dwarf-eh-prepare.ll -- ******************** ******************** Failed Tests (1): LLVM :: CodeGen/X86/dwarf-eh-prepare.ll Testing Time: 0.22s Failed: 1 ``` Reviewed By: loladiro Differential Revision: https://reviews.llvm.org/D110979
2021-10-02 21:31:59 -04:00
}
void getAnalysisUsage(AnalysisUsage &AU) const override {
AU.addRequired<TargetPassConfig>();
AU.addRequired<TargetTransformInfoWrapperPass>();
if (OptLevel != CodeGenOpt::None) {
AU.addRequired<DominatorTreeWrapperPass>();
AU.addRequired<TargetTransformInfoWrapperPass>();
}
AU.addPreserved<DominatorTreeWrapperPass>();
}
StringRef getPassName() const override {
return "Exception handling preparation";
}
};
} // end anonymous namespace
char DwarfEHPrepareLegacyPass::ID = 0;
INITIALIZE_PASS_BEGIN(DwarfEHPrepareLegacyPass, DEBUG_TYPE,
"Prepare DWARF exceptions", false, false)
INITIALIZE_PASS_DEPENDENCY(DominatorTreeWrapperPass)
INITIALIZE_PASS_DEPENDENCY(TargetPassConfig)
INITIALIZE_PASS_DEPENDENCY(TargetTransformInfoWrapperPass)
INITIALIZE_PASS_END(DwarfEHPrepareLegacyPass, DEBUG_TYPE,
"Prepare DWARF exceptions", false, false)
FunctionPass *llvm::createDwarfEHPass(CodeGenOpt::Level OptLevel) {
return new DwarfEHPrepareLegacyPass(OptLevel);
Add a new codegen pass that normalizes dwarf exception handling code in preparation for code generation. The main thing it does is handle the case when eh.exception calls (and, in a future patch, eh.selector calls) are far away from landing pads. Right now in practice you only find eh.exception calls close to landing pads: either in a landing pad (the common case) or in a landing pad successor, due to loop passes shifting them about. However future exception handling improvements will result in calls far from landing pads: (1) Inlining of rewinds. Consider the following case: In function @f: ... invoke @g to label %normal unwind label %unwinds ... unwinds: %ex = call i8* @llvm.eh.exception() ... In function @g: ... invoke @something to label %continue unwind label %handler ... handler: %ex = call i8* @llvm.eh.exception() ... perform cleanups ... "rethrow exception" Now inline @g into @f. Currently this is turned into: In function @f: ... invoke @something to label %continue unwind label %handler ... handler: %ex = call i8* @llvm.eh.exception() ... perform cleanups ... invoke "rethrow exception" to label %normal unwind label %unwinds unwinds: %ex = call i8* @llvm.eh.exception() ... However we would like to simplify invoke of "rethrow exception" into a branch to the %unwinds label. Then %unwinds is no longer a landing pad, and the eh.exception call there is then far away from any landing pads. (2) Using the unwind instruction for cleanups. It would be nice to have codegen handle the following case: invoke @something to label %continue unwind label %run_cleanups ... handler: ... perform cleanups ... unwind This requires turning "unwind" into a library call, which necessarily takes a pointer to the exception as an argument (this patch also does this unwind lowering). But that means you are using eh.exception again far from a landing pad. (3) Bugpoint simplifications. When bugpoint is simplifying exception handling code it often generates eh.exception calls far from a landing pad, which then causes codegen to assert. Bugpoint then latches on to this assertion and loses sight of the original problem. Note that it is currently rare for this pass to actually do anything. And in fact it normally shouldn't do anything at all given the code coming out of llvm-gcc! But it does fire a few times in the testsuite. As far as I can see this is almost always due to the LoopStrengthReduce codegen pass introducing pointless loop preheader blocks which are landing pads and only contain a branch to another block. This other block contains an eh.exception call. So probably by tweaking LoopStrengthReduce a bit this can be avoided. llvm-svn: 72276
2009-05-22 20:36:31 +00:00
}