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

751 lines
26 KiB
C++
Raw Normal View History

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
2009-05-23 04:36:31 +08:00
//===-- DwarfEHPrepare - Prepare exception handling for code generation ---===//
//
// The LLVM Compiler Infrastructure
//
// This file is distributed under the University of Illinois Open Source
// License. See LICENSE.TXT for details.
//
//===----------------------------------------------------------------------===//
//
// 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
2009-05-23 04:36:31 +08:00
//
//===----------------------------------------------------------------------===//
#define DEBUG_TYPE "dwarfehprepare"
#include "llvm/Function.h"
#include "llvm/Instructions.h"
#include "llvm/IntrinsicInst.h"
#include "llvm/Module.h"
#include "llvm/Pass.h"
#include "llvm/ADT/Statistic.h"
#include "llvm/Analysis/Dominators.h"
#include "llvm/CodeGen/Passes.h"
#include "llvm/MC/MCAsmInfo.h"
#include "llvm/Support/CallSite.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
2009-05-23 04:36:31 +08:00
#include "llvm/Target/TargetLowering.h"
#include "llvm/Transforms/Utils/BasicBlockUtils.h"
#include "llvm/Transforms/Utils/PromoteMemToReg.h"
using namespace llvm;
STATISTIC(NumLandingPadsSplit, "Number of landing pads split");
STATISTIC(NumUnwindsLowered, "Number of unwind instructions lowered");
STATISTIC(NumExceptionValuesMoved, "Number of eh.exception calls moved");
STATISTIC(NumStackTempsIntroduced, "Number of stack temporaries introduced");
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
2009-05-23 04:36:31 +08:00
namespace {
class DwarfEHPrepare : public FunctionPass {
const TargetMachine *TM;
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
2009-05-23 04:36:31 +08:00
const TargetLowering *TLI;
bool CompileFast;
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
2009-05-23 04:36:31 +08:00
// The eh.exception intrinsic.
Function *ExceptionValueIntrinsic;
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
2009-05-23 04:36:31 +08:00
// The eh.selector intrinsic.
Function *SelectorIntrinsic;
// _Unwind_Resume_or_Rethrow call.
Constant *URoR;
// The EH language-specific catch-all type.
GlobalVariable *EHCatchAllValue;
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
2009-05-23 04:36:31 +08:00
// _Unwind_Resume or the target equivalent.
Constant *RewindFunction;
// Dominator info is used when turning stack temporaries into registers.
DominatorTree *DT;
DominanceFrontier *DF;
// The function we are running on.
Function *F;
// The landing pads for this function.
typedef SmallPtrSet<BasicBlock*, 8> BBSet;
BBSet LandingPads;
// Stack temporary used to hold eh.exception values.
AllocaInst *ExceptionValueVar;
bool NormalizeLandingPads();
bool LowerUnwinds();
bool MoveExceptionValueCalls();
bool FinishStackTemporaries();
bool PromoteStackTemporaries();
Instruction *CreateExceptionValueCall(BasicBlock *BB);
Instruction *CreateValueLoad(BasicBlock *BB);
/// CreateReadOfExceptionValue - Return the result of the eh.exception
/// intrinsic by calling the intrinsic if in a landing pad, or loading it
/// from the exception value variable otherwise.
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
2009-05-23 04:36:31 +08:00
Instruction *CreateReadOfExceptionValue(BasicBlock *BB) {
return LandingPads.count(BB) ?
CreateExceptionValueCall(BB) : CreateValueLoad(BB);
}
/// CleanupSelectors - Any remaining eh.selector intrinsic calls which still
2010-04-15 23:14:46 +08:00
/// use the ".llvm.eh.catch.all.value" call need to convert to using its
/// initializer instead.
bool CleanupSelectors(SmallPtrSet<IntrinsicInst*, 32> &Sels);
bool HasCatchAllInSelector(IntrinsicInst *);
/// FindAllCleanupSelectors - Find all eh.selector calls that are clean-ups.
void FindAllCleanupSelectors(SmallPtrSet<IntrinsicInst*, 32> &Sels,
SmallPtrSet<IntrinsicInst*, 32> &CatchAllSels);
/// FindAllURoRInvokes - Find all URoR invokes in the function.
void FindAllURoRInvokes(SmallPtrSet<InvokeInst*, 32> &URoRInvokes);
/// HandleURoRInvokes - Handle invokes of "_Unwind_Resume_or_Rethrow"
/// calls. The "unwind" part of these invokes jump to a landing pad within
/// the current function. This is a candidate to merge the selector
/// associated with the URoR invoke with the one from the URoR's landing
/// pad.
bool HandleURoRInvokes();
/// FindSelectorAndURoR - Find the eh.selector call and URoR call associated
/// with the eh.exception call. This recursively looks past instructions
/// which don't change the EH pointer value, like casts or PHI nodes.
bool FindSelectorAndURoR(Instruction *Inst, bool &URoRInvoke,
SmallPtrSet<IntrinsicInst*, 8> &SelCalls);
/// DoMem2RegPromotion - Take an alloca call and promote it from memory to a
/// register.
bool DoMem2RegPromotion(Value *V) {
AllocaInst *AI = dyn_cast<AllocaInst>(V);
if (!AI || !isAllocaPromotable(AI)) return false;
// Turn the alloca into a register.
std::vector<AllocaInst*> Allocas(1, AI);
PromoteMemToReg(Allocas, *DT, *DF);
return true;
}
/// PromoteStoreInst - Perform Mem2Reg on a StoreInst.
bool PromoteStoreInst(StoreInst *SI) {
if (!SI || !DT || !DF) return false;
if (DoMem2RegPromotion(SI->getOperand(1)))
return true;
return false;
}
/// PromoteEHPtrStore - Promote the storing of an EH pointer into a
/// register. This should get rid of the store and subsequent loads.
bool PromoteEHPtrStore(IntrinsicInst *II) {
if (!DT || !DF) return false;
bool Changed = false;
StoreInst *SI;
while (1) {
SI = 0;
for (Value::use_iterator
I = II->use_begin(), E = II->use_end(); I != E; ++I) {
SI = dyn_cast<StoreInst>(I);
if (SI) break;
}
if (!PromoteStoreInst(SI))
break;
Changed = true;
}
return Changed;
}
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
2009-05-23 04:36:31 +08:00
public:
static char ID; // Pass identification, replacement for typeid.
DwarfEHPrepare(const TargetMachine *tm, bool fast) :
FunctionPass(&ID), TM(tm), TLI(TM->getTargetLowering()),
CompileFast(fast),
ExceptionValueIntrinsic(0), SelectorIntrinsic(0),
URoR(0), EHCatchAllValue(0), RewindFunction(0) {}
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
2009-05-23 04:36:31 +08:00
virtual bool runOnFunction(Function &Fn);
// getAnalysisUsage - We need dominance frontiers for memory promotion.
virtual void getAnalysisUsage(AnalysisUsage &AU) const {
if (!CompileFast)
AU.addRequired<DominatorTree>();
AU.addPreserved<DominatorTree>();
if (!CompileFast)
AU.addRequired<DominanceFrontier>();
AU.addPreserved<DominanceFrontier>();
}
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
2009-05-23 04:36:31 +08:00
const char *getPassName() const {
return "Exception handling preparation";
}
};
} // end anonymous namespace
char DwarfEHPrepare::ID = 0;
FunctionPass *llvm::createDwarfEHPass(const TargetMachine *tm, bool fast) {
return new DwarfEHPrepare(tm, fast);
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
2009-05-23 04:36:31 +08:00
}
/// HasCatchAllInSelector - Return true if the intrinsic instruction has a
/// catch-all.
bool DwarfEHPrepare::HasCatchAllInSelector(IntrinsicInst *II) {
if (!EHCatchAllValue) return false;
unsigned ArgIdx = II->getNumArgOperands() - 1;
GlobalVariable *GV = dyn_cast<GlobalVariable>(II->getArgOperand(ArgIdx));
return GV == EHCatchAllValue;
}
/// FindAllCleanupSelectors - Find all eh.selector calls that are clean-ups.
void DwarfEHPrepare::
FindAllCleanupSelectors(SmallPtrSet<IntrinsicInst*, 32> &Sels,
SmallPtrSet<IntrinsicInst*, 32> &CatchAllSels) {
for (Value::use_iterator
I = SelectorIntrinsic->use_begin(),
E = SelectorIntrinsic->use_end(); I != E; ++I) {
IntrinsicInst *II = cast<IntrinsicInst>(I);
if (II->getParent()->getParent() != F)
continue;
if (!HasCatchAllInSelector(II))
Sels.insert(II);
else
CatchAllSels.insert(II);
}
}
/// FindAllURoRInvokes - Find all URoR invokes in the function.
void DwarfEHPrepare::
FindAllURoRInvokes(SmallPtrSet<InvokeInst*, 32> &URoRInvokes) {
for (Value::use_iterator
I = URoR->use_begin(),
E = URoR->use_end(); I != E; ++I) {
if (InvokeInst *II = dyn_cast<InvokeInst>(I))
URoRInvokes.insert(II);
}
}
/// CleanupSelectors - Any remaining eh.selector intrinsic calls which still use
2010-04-15 23:14:46 +08:00
/// the ".llvm.eh.catch.all.value" call need to convert to using its
/// initializer instead.
bool DwarfEHPrepare::CleanupSelectors(SmallPtrSet<IntrinsicInst*, 32> &Sels) {
if (!EHCatchAllValue) return false;
if (!SelectorIntrinsic) {
SelectorIntrinsic =
Intrinsic::getDeclaration(F->getParent(), Intrinsic::eh_selector);
if (!SelectorIntrinsic) return false;
}
bool Changed = false;
for (SmallPtrSet<IntrinsicInst*, 32>::iterator
I = Sels.begin(), E = Sels.end(); I != E; ++I) {
IntrinsicInst *Sel = *I;
// Index of the ".llvm.eh.catch.all.value" variable.
2010-06-29 21:03:46 +08:00
unsigned OpIdx = Sel->getNumArgOperands() - 1;
GlobalVariable *GV = dyn_cast<GlobalVariable>(Sel->getArgOperand(OpIdx));
if (GV != EHCatchAllValue) continue;
2010-06-29 21:03:46 +08:00
Sel->setArgOperand(OpIdx, EHCatchAllValue->getInitializer());
Changed = true;
}
return Changed;
}
/// FindSelectorAndURoR - Find the eh.selector call associated with the
/// eh.exception call. And indicate if there is a URoR "invoke" associated with
/// the eh.exception call. This recursively looks past instructions which don't
/// change the EH pointer value, like casts or PHI nodes.
bool
DwarfEHPrepare::FindSelectorAndURoR(Instruction *Inst, bool &URoRInvoke,
SmallPtrSet<IntrinsicInst*, 8> &SelCalls) {
SmallPtrSet<PHINode*, 32> SeenPHIs;
bool Changed = false;
restart:
for (Value::use_iterator
I = Inst->use_begin(), E = Inst->use_end(); I != E; ++I) {
Instruction *II = dyn_cast<Instruction>(I);
if (!II || II->getParent()->getParent() != F) continue;
if (IntrinsicInst *Sel = dyn_cast<IntrinsicInst>(II)) {
if (Sel->getIntrinsicID() == Intrinsic::eh_selector)
SelCalls.insert(Sel);
} else if (InvokeInst *Invoke = dyn_cast<InvokeInst>(II)) {
if (Invoke->getCalledFunction() == URoR)
URoRInvoke = true;
} else if (CastInst *CI = dyn_cast<CastInst>(II)) {
Changed |= FindSelectorAndURoR(CI, URoRInvoke, SelCalls);
} else if (StoreInst *SI = dyn_cast<StoreInst>(II)) {
if (!PromoteStoreInst(SI)) continue;
Changed = true;
SeenPHIs.clear();
goto restart; // Uses may have changed, restart loop.
} else if (PHINode *PN = dyn_cast<PHINode>(II)) {
if (SeenPHIs.insert(PN))
// Don't process a PHI node more than once.
Changed |= FindSelectorAndURoR(PN, URoRInvoke, SelCalls);
}
}
return Changed;
}
/// HandleURoRInvokes - Handle invokes of "_Unwind_Resume_or_Rethrow" calls. The
/// "unwind" part of these invokes jump to a landing pad within the current
/// function. This is a candidate to merge the selector associated with the URoR
/// invoke with the one from the URoR's landing pad.
bool DwarfEHPrepare::HandleURoRInvokes() {
if (!EHCatchAllValue) {
EHCatchAllValue =
F->getParent()->getNamedGlobal(".llvm.eh.catch.all.value");
if (!EHCatchAllValue) return false;
}
if (!SelectorIntrinsic) {
SelectorIntrinsic =
Intrinsic::getDeclaration(F->getParent(), Intrinsic::eh_selector);
if (!SelectorIntrinsic) return false;
}
SmallPtrSet<IntrinsicInst*, 32> Sels;
SmallPtrSet<IntrinsicInst*, 32> CatchAllSels;
FindAllCleanupSelectors(Sels, CatchAllSels);
if (!DT)
// We require DominatorTree information.
return CleanupSelectors(CatchAllSels);
if (!URoR) {
URoR = F->getParent()->getFunction("_Unwind_Resume_or_Rethrow");
if (!URoR) return CleanupSelectors(CatchAllSels);
}
SmallPtrSet<InvokeInst*, 32> URoRInvokes;
FindAllURoRInvokes(URoRInvokes);
SmallPtrSet<IntrinsicInst*, 32> SelsToConvert;
for (SmallPtrSet<IntrinsicInst*, 32>::iterator
SI = Sels.begin(), SE = Sels.end(); SI != SE; ++SI) {
const BasicBlock *SelBB = (*SI)->getParent();
for (SmallPtrSet<InvokeInst*, 32>::iterator
UI = URoRInvokes.begin(), UE = URoRInvokes.end(); UI != UE; ++UI) {
const BasicBlock *URoRBB = (*UI)->getParent();
if (SelBB == URoRBB || DT->dominates(SelBB, URoRBB)) {
SelsToConvert.insert(*SI);
break;
}
}
}
bool Changed = false;
if (Sels.size() != SelsToConvert.size()) {
// If we haven't been able to convert all of the clean-up selectors, then
// loop through the slow way to see if they still need to be converted.
if (!ExceptionValueIntrinsic) {
ExceptionValueIntrinsic =
Intrinsic::getDeclaration(F->getParent(), Intrinsic::eh_exception);
if (!ExceptionValueIntrinsic)
return CleanupSelectors(CatchAllSels);
}
for (Value::use_iterator
I = ExceptionValueIntrinsic->use_begin(),
E = ExceptionValueIntrinsic->use_end(); I != E; ++I) {
IntrinsicInst *EHPtr = dyn_cast<IntrinsicInst>(I);
if (!EHPtr || EHPtr->getParent()->getParent() != F) continue;
Changed |= PromoteEHPtrStore(EHPtr);
bool URoRInvoke = false;
SmallPtrSet<IntrinsicInst*, 8> SelCalls;
Changed |= FindSelectorAndURoR(EHPtr, URoRInvoke, SelCalls);
if (URoRInvoke) {
// This EH pointer is being used by an invoke of an URoR instruction and
// an eh.selector intrinsic call. If the eh.selector is a 'clean-up', we
// need to convert it to a 'catch-all'.
for (SmallPtrSet<IntrinsicInst*, 8>::iterator
SI = SelCalls.begin(), SE = SelCalls.end(); SI != SE; ++SI)
if (!HasCatchAllInSelector(*SI))
SelsToConvert.insert(*SI);
}
}
}
if (!SelsToConvert.empty()) {
// Convert all clean-up eh.selectors, which are associated with "invokes" of
// URoR calls, into catch-all eh.selectors.
Changed = true;
for (SmallPtrSet<IntrinsicInst*, 8>::iterator
SI = SelsToConvert.begin(), SE = SelsToConvert.end();
SI != SE; ++SI) {
IntrinsicInst *II = *SI;
// Use the exception object pointer and the personality function
// from the original selector.
CallSite CS(II);
IntrinsicInst::op_iterator I = CS.arg_begin();
IntrinsicInst::op_iterator E = CS.arg_end();
IntrinsicInst::op_iterator B = prior(E);
// Exclude last argument if it is an integer.
if (isa<ConstantInt>(B)) E = B;
// Add exception object pointer (front).
// Add personality function (next).
// Add in any filter IDs (rest).
SmallVector<Value*, 8> Args(I, E);
Args.push_back(EHCatchAllValue->getInitializer()); // Catch-all indicator.
CallInst *NewSelector =
CallInst::Create(SelectorIntrinsic, Args.begin(), Args.end(),
"eh.sel.catch.all", II);
NewSelector->setTailCall(II->isTailCall());
NewSelector->setAttributes(II->getAttributes());
NewSelector->setCallingConv(II->getCallingConv());
II->replaceAllUsesWith(NewSelector);
II->eraseFromParent();
}
}
Changed |= CleanupSelectors(CatchAllSels);
return Changed;
}
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
2009-05-23 04:36:31 +08:00
/// NormalizeLandingPads - Normalize and discover landing pads, noting them
/// in the LandingPads set. A landing pad is normal if the only CFG edges
2009-09-16 05:56:46 +08:00
/// that end at it are unwind edges from invoke instructions. If we inlined
/// through an invoke we could have a normal branch from the previous
/// unwind block through to the landing pad for the original invoke.
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
2009-05-23 04:36:31 +08:00
/// Abnormal landing pads are fixed up by redirecting all unwind edges to
/// a new basic block which falls through to the original.
bool DwarfEHPrepare::NormalizeLandingPads() {
bool Changed = false;
const MCAsmInfo *MAI = TM->getMCAsmInfo();
bool usingSjLjEH = MAI->getExceptionHandlingType() == ExceptionHandling::SjLj;
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
2009-05-23 04:36:31 +08:00
for (Function::iterator I = F->begin(), E = F->end(); I != E; ++I) {
TerminatorInst *TI = I->getTerminator();
if (!isa<InvokeInst>(TI))
continue;
BasicBlock *LPad = TI->getSuccessor(1);
// Skip landing pads that have already been normalized.
if (LandingPads.count(LPad))
continue;
// Check that only invoke unwind edges end at the landing pad.
bool OnlyUnwoundTo = true;
bool SwitchOK = usingSjLjEH;
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
2009-05-23 04:36:31 +08:00
for (pred_iterator PI = pred_begin(LPad), PE = pred_end(LPad);
PI != PE; ++PI) {
TerminatorInst *PT = (*PI)->getTerminator();
// The SjLj dispatch block uses a switch instruction. This is effectively
// an unwind edge, so we can disregard it here. There will only ever
// be one dispatch, however, so if there are multiple switches, one
// of them truly is a normal edge, not an unwind edge.
if (SwitchOK && isa<SwitchInst>(PT)) {
SwitchOK = false;
continue;
}
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
2009-05-23 04:36:31 +08:00
if (!isa<InvokeInst>(PT) || LPad == PT->getSuccessor(0)) {
OnlyUnwoundTo = false;
break;
}
}
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
2009-05-23 04:36:31 +08:00
if (OnlyUnwoundTo) {
// Only unwind edges lead to the landing pad. Remember the landing pad.
LandingPads.insert(LPad);
continue;
}
// At least one normal edge ends at the landing pad. Redirect the unwind
// edges to a new basic block which falls through into this one.
// Create the new basic block.
BasicBlock *NewBB = BasicBlock::Create(F->getContext(),
LPad->getName() + "_unwind_edge");
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
2009-05-23 04:36:31 +08:00
// Insert it into the function right before the original landing pad.
LPad->getParent()->getBasicBlockList().insert(LPad, NewBB);
// Redirect unwind edges from the original landing pad to NewBB.
for (pred_iterator PI = pred_begin(LPad), PE = pred_end(LPad); PI != PE; ) {
TerminatorInst *PT = (*PI++)->getTerminator();
if (isa<InvokeInst>(PT) && PT->getSuccessor(1) == LPad)
// Unwind to the new block.
PT->setSuccessor(1, NewBB);
}
// If there are any PHI nodes in LPad, we need to update them so that they
// merge incoming values from NewBB instead.
for (BasicBlock::iterator II = LPad->begin(); isa<PHINode>(II); ++II) {
PHINode *PN = cast<PHINode>(II);
pred_iterator PB = pred_begin(NewBB), PE = pred_end(NewBB);
// Check to see if all of the values coming in via unwind edges are the
// same. If so, we don't need to create a new PHI node.
Value *InVal = PN->getIncomingValueForBlock(*PB);
for (pred_iterator PI = PB; PI != PE; ++PI) {
if (PI != PB && InVal != PN->getIncomingValueForBlock(*PI)) {
InVal = 0;
break;
}
}
if (InVal == 0) {
// Different unwind edges have different values. Create a new PHI node
// in NewBB.
PHINode *NewPN = PHINode::Create(PN->getType(), PN->getName()+".unwind",
NewBB);
// Add an entry for each unwind edge, using the value from the old PHI.
for (pred_iterator PI = PB; PI != PE; ++PI)
NewPN->addIncoming(PN->getIncomingValueForBlock(*PI), *PI);
// Now use this new PHI as the common incoming value for NewBB in PN.
InVal = NewPN;
}
// Revector exactly one entry in the PHI node to come from NewBB
// and delete all other entries that come from unwind edges. If
// there are both normal and unwind edges from the same predecessor,
// this leaves an entry for the normal edge.
for (pred_iterator PI = PB; PI != PE; ++PI)
PN->removeIncomingValue(*PI);
PN->addIncoming(InVal, NewBB);
}
// Add a fallthrough from NewBB to the original landing pad.
BranchInst::Create(LPad, NewBB);
// Now update DominatorTree and DominanceFrontier analysis information.
if (DT)
DT->splitBlock(NewBB);
if (DF)
DF->splitBlock(NewBB);
// Remember the newly constructed landing pad. The original landing pad
// LPad is no longer a landing pad now that all unwind edges have been
// revectored to NewBB.
LandingPads.insert(NewBB);
++NumLandingPadsSplit;
Changed = true;
}
return Changed;
}
/// LowerUnwinds - Turn unwind instructions into calls to _Unwind_Resume,
/// rethrowing any previously caught exception. This will crash horribly
/// at runtime if there is no such exception: using unwind to throw a new
/// exception is currently not supported.
bool DwarfEHPrepare::LowerUnwinds() {
SmallVector<TerminatorInst*, 16> UnwindInsts;
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
2009-05-23 04:36:31 +08:00
for (Function::iterator I = F->begin(), E = F->end(); I != E; ++I) {
TerminatorInst *TI = I->getTerminator();
if (isa<UnwindInst>(TI))
UnwindInsts.push_back(TI);
}
if (UnwindInsts.empty()) return false;
// Find the rewind function if we didn't already.
if (!RewindFunction) {
LLVMContext &Ctx = UnwindInsts[0]->getContext();
std::vector<const Type*>
Params(1, Type::getInt8PtrTy(Ctx));
FunctionType *FTy = FunctionType::get(Type::getVoidTy(Ctx),
Params, false);
const char *RewindName = TLI->getLibcallName(RTLIB::UNWIND_RESUME);
RewindFunction = F->getParent()->getOrInsertFunction(RewindName, FTy);
}
bool Changed = false;
for (SmallVectorImpl<TerminatorInst*>::iterator
I = UnwindInsts.begin(), E = UnwindInsts.end(); I != E; ++I) {
TerminatorInst *TI = *I;
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
2009-05-23 04:36:31 +08:00
// Replace the unwind instruction with a call to _Unwind_Resume (or the
// appropriate target equivalent) followed by an UnreachableInst.
// Create the call...
CallInst *CI = CallInst::Create(RewindFunction,
CreateReadOfExceptionValue(TI->getParent()),
"", TI);
CI->setCallingConv(TLI->getLibcallCallingConv(RTLIB::UNWIND_RESUME));
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
2009-05-23 04:36:31 +08:00
// ...followed by an UnreachableInst.
new UnreachableInst(TI->getContext(), TI);
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
2009-05-23 04:36:31 +08:00
// Nuke the unwind instruction.
TI->eraseFromParent();
++NumUnwindsLowered;
Changed = true;
}
return Changed;
}
/// MoveExceptionValueCalls - Ensure that eh.exception is only ever called from
/// landing pads by replacing calls outside of landing pads with loads from a
/// stack temporary. Move eh.exception calls inside landing pads to the start
/// of the landing pad (optional, but may make things simpler for later passes).
bool DwarfEHPrepare::MoveExceptionValueCalls() {
// If the eh.exception intrinsic is not declared in the module then there is
// nothing to do. Speed up compilation by checking for this common case.
if (!ExceptionValueIntrinsic &&
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
2009-05-23 04:36:31 +08:00
!F->getParent()->getFunction(Intrinsic::getName(Intrinsic::eh_exception)))
return false;
bool Changed = false;
for (Function::iterator BB = F->begin(), E = F->end(); BB != E; ++BB) {
for (BasicBlock::iterator II = BB->begin(), E = BB->end(); II != E;)
if (IntrinsicInst *CI = dyn_cast<IntrinsicInst>(II++))
if (CI->getIntrinsicID() == Intrinsic::eh_exception) {
if (!CI->use_empty()) {
Value *ExceptionValue = CreateReadOfExceptionValue(BB);
if (CI == ExceptionValue) {
// The call was at the start of a landing pad - leave it alone.
assert(LandingPads.count(BB) &&
"Created eh.exception call outside landing pad!");
continue;
}
CI->replaceAllUsesWith(ExceptionValue);
}
CI->eraseFromParent();
++NumExceptionValuesMoved;
Changed = true;
}
}
return Changed;
}
/// FinishStackTemporaries - If we introduced a stack variable to hold the
/// exception value then initialize it in each landing pad.
bool DwarfEHPrepare::FinishStackTemporaries() {
if (!ExceptionValueVar)
// Nothing to do.
return false;
bool Changed = false;
// Make sure that there is a store of the exception value at the start of
// each landing pad.
for (BBSet::iterator LI = LandingPads.begin(), LE = LandingPads.end();
LI != LE; ++LI) {
Instruction *ExceptionValue = CreateReadOfExceptionValue(*LI);
Instruction *Store = new StoreInst(ExceptionValue, ExceptionValueVar);
Store->insertAfter(ExceptionValue);
Changed = true;
}
return Changed;
}
/// PromoteStackTemporaries - Turn any stack temporaries we introduced into
/// registers if possible.
bool DwarfEHPrepare::PromoteStackTemporaries() {
if (ExceptionValueVar && DT && DF && isAllocaPromotable(ExceptionValueVar)) {
// Turn the exception temporary into registers and phi nodes if possible.
std::vector<AllocaInst*> Allocas(1, ExceptionValueVar);
PromoteMemToReg(Allocas, *DT, *DF);
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
2009-05-23 04:36:31 +08:00
return true;
}
return false;
}
/// CreateExceptionValueCall - Insert a call to the eh.exception intrinsic at
/// the start of the basic block (unless there already is one, in which case
/// the existing call is returned).
Instruction *DwarfEHPrepare::CreateExceptionValueCall(BasicBlock *BB) {
Instruction *Start = BB->getFirstNonPHIOrDbg();
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
2009-05-23 04:36:31 +08:00
// Is this a call to eh.exception?
if (IntrinsicInst *CI = dyn_cast<IntrinsicInst>(Start))
if (CI->getIntrinsicID() == Intrinsic::eh_exception)
// Reuse the existing call.
return Start;
// Find the eh.exception intrinsic if we didn't already.
if (!ExceptionValueIntrinsic)
ExceptionValueIntrinsic = Intrinsic::getDeclaration(F->getParent(),
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
2009-05-23 04:36:31 +08:00
Intrinsic::eh_exception);
// Create the call.
return CallInst::Create(ExceptionValueIntrinsic, "eh.value.call", Start);
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
2009-05-23 04:36:31 +08:00
}
/// CreateValueLoad - Insert a load of the exception value stack variable
/// (creating it if necessary) at the start of the basic block (unless
/// there already is a load, in which case the existing load is returned).
Instruction *DwarfEHPrepare::CreateValueLoad(BasicBlock *BB) {
Instruction *Start = BB->getFirstNonPHIOrDbg();
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
2009-05-23 04:36:31 +08:00
// Is this a load of the exception temporary?
if (ExceptionValueVar)
if (LoadInst* LI = dyn_cast<LoadInst>(Start))
if (LI->getPointerOperand() == ExceptionValueVar)
// Reuse the existing load.
return Start;
// Create the temporary if we didn't already.
if (!ExceptionValueVar) {
ExceptionValueVar = new AllocaInst(PointerType::getUnqual(
Type::getInt8Ty(BB->getContext())), "eh.value", F->begin()->begin());
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
2009-05-23 04:36:31 +08:00
++NumStackTempsIntroduced;
}
// Load the value.
return new LoadInst(ExceptionValueVar, "eh.value.load", Start);
}
bool DwarfEHPrepare::runOnFunction(Function &Fn) {
bool Changed = false;
// Initialize internal state.
DT = getAnalysisIfAvailable<DominatorTree>();
DF = getAnalysisIfAvailable<DominanceFrontier>();
ExceptionValueVar = 0;
F = &Fn;
// Ensure that only unwind edges end at landing pads (a landing pad is a
// basic block where an invoke unwind edge ends).
Changed |= NormalizeLandingPads();
// Turn unwind instructions into libcalls.
Changed |= LowerUnwinds();
// TODO: Move eh.selector calls to landing pads and combine them.
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
2009-05-23 04:36:31 +08:00
// Move eh.exception calls to landing pads.
Changed |= MoveExceptionValueCalls();
// Initialize any stack temporaries we introduced.
Changed |= FinishStackTemporaries();
// Turn any stack temporaries into registers if possible.
if (!CompileFast)
Changed |= PromoteStackTemporaries();
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
2009-05-23 04:36:31 +08:00
Changed |= HandleURoRInvokes();
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
2009-05-23 04:36:31 +08:00
LandingPads.clear();
return Changed;
}