#include "llvm/ADT/SmallPtrSet.h"
#include "llvm/ADT/SmallSet.h"
#include "llvm/ADT/Statistic.h"
+#include "llvm/Analysis/GlobalsModRef.h"
#include "llvm/Analysis/CFG.h"
+#include "llvm/Analysis/BlockFrequencyInfo.h"
+#include "llvm/Analysis/BlockFrequencyInfoImpl.h"
+#include "llvm/Analysis/BranchProbabilityInfo.h"
#include "llvm/Analysis/ConstantFolding.h"
#include "llvm/Analysis/InstructionSimplify.h"
#include "llvm/Analysis/LazyValueInfo.h"
#include "llvm/Analysis/Loads.h"
+#include "llvm/Analysis/LoopInfo.h"
#include "llvm/Analysis/TargetLibraryInfo.h"
+#include "llvm/Analysis/ValueTracking.h"
#include "llvm/IR/DataLayout.h"
#include "llvm/IR/IntrinsicInst.h"
#include "llvm/IR/LLVMContext.h"
+#include "llvm/IR/MDBuilder.h"
#include "llvm/IR/Metadata.h"
#include "llvm/IR/ValueHandle.h"
#include "llvm/Pass.h"
#include "llvm/Transforms/Utils/BasicBlockUtils.h"
#include "llvm/Transforms/Utils/Local.h"
#include "llvm/Transforms/Utils/SSAUpdater.h"
+#include <algorithm>
+#include <memory>
using namespace llvm;
#define DEBUG_TYPE "jump-threading"
cl::desc("Max block size to duplicate for jump threading"),
cl::init(6), cl::Hidden);
+static cl::opt<unsigned>
+ImplicationSearchThreshold(
+ "jump-threading-implication-search-threshold",
+ cl::desc("The number of predecessors to search for a stronger "
+ "condition to use to thread over a weaker condition"),
+ cl::init(3), cl::Hidden);
+
namespace {
// These are at global scope so static functions can use them too.
typedef SmallVectorImpl<std::pair<Constant*, BasicBlock*> > PredValueInfo;
class JumpThreading : public FunctionPass {
TargetLibraryInfo *TLI;
LazyValueInfo *LVI;
+ std::unique_ptr<BlockFrequencyInfo> BFI;
+ std::unique_ptr<BranchProbabilityInfo> BPI;
+ bool HasProfileData;
#ifdef NDEBUG
SmallPtrSet<BasicBlock*, 16> LoopHeaders;
#else
void getAnalysisUsage(AnalysisUsage &AU) const override {
AU.addRequired<LazyValueInfo>();
AU.addPreserved<LazyValueInfo>();
+ AU.addPreserved<GlobalsAAWrapperPass>();
AU.addRequired<TargetLibraryInfoWrapperPass>();
}
+ void releaseMemory() override {
+ BFI.reset();
+ BPI.reset();
+ }
+
void FindLoopHeaders(Function &F);
bool ProcessBlock(BasicBlock *BB);
bool ThreadEdge(BasicBlock *BB, const SmallVectorImpl<BasicBlock*> &PredBBs,
bool ProcessBranchOnPHI(PHINode *PN);
bool ProcessBranchOnXOR(BinaryOperator *BO);
+ bool ProcessImpliedCondition(BasicBlock *BB);
bool SimplifyPartiallyRedundantLoad(LoadInst *LI);
bool TryToUnfoldSelect(CmpInst *CondCmp, BasicBlock *BB);
+
+ private:
+ BasicBlock *SplitBlockPreds(BasicBlock *BB, ArrayRef<BasicBlock *> Preds,
+ const char *Suffix);
+ void UpdateBlockFreqAndEdgeWeight(BasicBlock *PredBB, BasicBlock *BB,
+ BasicBlock *NewBB, BasicBlock *SuccBB);
};
}
DEBUG(dbgs() << "Jump threading on function '" << F.getName() << "'\n");
TLI = &getAnalysis<TargetLibraryInfoWrapperPass>().getTLI();
LVI = &getAnalysis<LazyValueInfo>();
+ BFI.reset();
+ BPI.reset();
+ // When profile data is available, we need to update edge weights after
+ // successful jump threading, which requires both BPI and BFI being available.
+ HasProfileData = F.getEntryCount().hasValue();
+ if (HasProfileData) {
+ LoopInfo LI{DominatorTree(F)};
+ BPI.reset(new BranchProbabilityInfo(F, LI));
+ BFI.reset(new BlockFrequencyInfo(F, *BPI, LI));
+ }
// Remove unreachable blocks from function as they may result in infinite
// loop. We do threading if we found something profitable. Jump threading a
// branch can create other opportunities. If these opportunities form a cycle
- // i.e. if any jump treading is undoing previous threading in the path, then
+ // i.e. if any jump threading is undoing previous threading in the path, then
// we will loop forever. We take care of this issue by not jump threading for
// back edges. This works for normal cases but not for unreachable blocks as
// they may have cycle with no back edge.
do {
Changed = false;
for (Function::iterator I = F.begin(), E = F.end(); I != E;) {
- BasicBlock *BB = I;
+ BasicBlock *BB = &*I;
// Thread all of the branches we can over this block.
while (ProcessBlock(BB))
Changed = true;
static unsigned getJumpThreadDuplicationCost(const BasicBlock *BB,
unsigned Threshold) {
/// Ignore PHI nodes, these will be flattened when duplication happens.
- BasicBlock::const_iterator I = BB->getFirstNonPHI();
+ BasicBlock::const_iterator I(BB->getFirstNonPHI());
// FIXME: THREADING will delete values that are just used to compute the
// branch, so they shouldn't count against the duplication cost.
if (isa<BitCastInst>(I) && I->getType()->isPointerTy())
continue;
+ // Bail out if this instruction gives back a token type, it is not possible
+ // to duplicate it if it is used outside this BB.
+ if (I->getType()->isTokenTy() && I->isUsedOutsideOfBlock(BB))
+ return ~0U;
+
// All other instructions count for at least one unit.
++Size;
// as having cost of 2 total, and if they are a vector intrinsic, we model
// them as having cost 1.
if (const CallInst *CI = dyn_cast<CallInst>(I)) {
- if (CI->cannotDuplicate())
+ if (CI->cannotDuplicate() || CI->isConvergent())
// Blocks with NoDuplicate are modelled as having infinite cost, so they
// are never duplicated.
return ~0U;
// because now the condition in this block can be threaded through
// predecessors of our predecessor block.
if (BasicBlock *SinglePred = BB->getSinglePredecessor()) {
- if (SinglePred->getTerminator()->getNumSuccessors() == 1 &&
+ const TerminatorInst *TI = SinglePred->getTerminator();
+ if (!TI->isExceptional() && TI->getNumSuccessors() == 1 &&
SinglePred != BB && !hasAddressTakenAndUsed(BB)) {
// If SinglePred was a loop header, BB becomes one.
if (LoopHeaders.erase(SinglePred))
if (CmpInst *CondCmp = dyn_cast<CmpInst>(CondInst)) {
- // For a comparison where the LHS is outside this block, it's possible
- // that we've branched on it before. Used LVI to see if we can simplify
- // the branch based on that.
+ // If we're branching on a conditional, LVI might be able to determine
+ // it's value at the branch instruction. We only handle comparisons
+ // against a constant at this time.
+ // TODO: This should be extended to handle switches as well.
BranchInst *CondBr = dyn_cast<BranchInst>(BB->getTerminator());
Constant *CondConst = dyn_cast<Constant>(CondCmp->getOperand(1));
- pred_iterator PI = pred_begin(BB), PE = pred_end(BB);
- if (CondBr && CondConst && CondBr->isConditional() && PI != PE &&
- (!isa<Instruction>(CondCmp->getOperand(0)) ||
- cast<Instruction>(CondCmp->getOperand(0))->getParent() != BB)) {
- // For predecessor edge, determine if the comparison is true or false
- // on that edge. If they're all true or all false, we can simplify the
- // branch.
- // FIXME: We could handle mixed true/false by duplicating code.
- LazyValueInfo::Tristate Baseline =
- LVI->getPredicateOnEdge(CondCmp->getPredicate(), CondCmp->getOperand(0),
- CondConst, *PI, BB, CondCmp);
- if (Baseline != LazyValueInfo::Unknown) {
- // Check that all remaining incoming values match the first one.
- while (++PI != PE) {
- LazyValueInfo::Tristate Ret =
- LVI->getPredicateOnEdge(CondCmp->getPredicate(),
- CondCmp->getOperand(0), CondConst, *PI, BB,
- CondCmp);
- if (Ret != Baseline) break;
- }
-
- // If we terminated early, then one of the values didn't match.
- if (PI == PE) {
- unsigned ToRemove = Baseline == LazyValueInfo::True ? 1 : 0;
- unsigned ToKeep = Baseline == LazyValueInfo::True ? 0 : 1;
- CondBr->getSuccessor(ToRemove)->removePredecessor(BB, true);
- BranchInst::Create(CondBr->getSuccessor(ToKeep), CondBr);
- CondBr->eraseFromParent();
- if (CondCmp->use_empty())
- CondCmp->eraseFromParent();
- else if (CondCmp->getParent() == BB) {
- // If the fact we just learned is true for all uses of the
- // condition, replace it with a constant value
- auto *CI = Baseline == LazyValueInfo::True ?
- ConstantInt::getTrue(CondCmp->getType()) :
- ConstantInt::getFalse(CondCmp->getType());
- CondCmp->replaceAllUsesWith(CI);
- CondCmp->eraseFromParent();
- }
- return true;
- }
- }
-
- } else if (CondBr && CondConst && CondBr->isConditional()) {
- // There might be an invariant in the same block with the conditional
- // that can determine the predicate.
-
+ if (CondBr && CondConst && CondBr->isConditional()) {
LazyValueInfo::Tristate Ret =
LVI->getPredicateAt(CondCmp->getPredicate(), CondCmp->getOperand(0),
- CondConst, CondCmp);
+ CondConst, CondBr);
if (Ret != LazyValueInfo::Unknown) {
unsigned ToRemove = Ret == LazyValueInfo::True ? 1 : 0;
unsigned ToKeep = Ret == LazyValueInfo::True ? 0 : 1;
CondBr->getSuccessor(ToRemove)->removePredecessor(BB, true);
BranchInst::Create(CondBr->getSuccessor(ToKeep), CondBr);
CondBr->eraseFromParent();
+ if (CondCmp->use_empty())
+ CondCmp->eraseFromParent();
+ else if (CondCmp->getParent() == BB) {
+ // If the fact we just learned is true for all uses of the
+ // condition, replace it with a constant value
+ auto *CI = Ret == LazyValueInfo::True ?
+ ConstantInt::getTrue(CondCmp->getType()) :
+ ConstantInt::getFalse(CondCmp->getType());
+ CondCmp->replaceAllUsesWith(CI);
+ CondCmp->eraseFromParent();
+ }
return true;
}
}
CondInst->getParent() == BB && isa<BranchInst>(BB->getTerminator()))
return ProcessBranchOnXOR(cast<BinaryOperator>(CondInst));
+ // Search for a stronger dominating condition that can be used to simplify a
+ // conditional branch leaving BB.
+ if (ProcessImpliedCondition(BB))
+ return true;
+
+ return false;
+}
+
+bool JumpThreading::ProcessImpliedCondition(BasicBlock *BB) {
+ auto *BI = dyn_cast<BranchInst>(BB->getTerminator());
+ if (!BI || !BI->isConditional())
+ return false;
- // TODO: If we have: "br (X > 0)" and we have a predecessor where we know
- // "(X == 4)", thread through this block.
+ Value *Cond = BI->getCondition();
+ BasicBlock *CurrentBB = BB;
+ BasicBlock *CurrentPred = BB->getSinglePredecessor();
+ unsigned Iter = 0;
+
+ auto &DL = BB->getModule()->getDataLayout();
+
+ while (CurrentPred && Iter++ < ImplicationSearchThreshold) {
+ auto *PBI = dyn_cast<BranchInst>(CurrentPred->getTerminator());
+ if (!PBI || !PBI->isConditional() || PBI->getSuccessor(0) != CurrentBB)
+ return false;
+
+ if (isImpliedCondition(PBI->getCondition(), Cond, DL)) {
+ BI->getSuccessor(1)->removePredecessor(BB);
+ BranchInst::Create(BI->getSuccessor(0), BI);
+ BI->eraseFromParent();
+ return true;
+ }
+ CurrentBB = CurrentPred;
+ CurrentPred = CurrentBB->getSinglePredecessor();
+ }
return false;
}
if (LoadBB->getSinglePredecessor())
return false;
- // If the load is defined in a landing pad, it can't be partially redundant,
- // because the edges between the invoke and the landing pad cannot have other
+ // If the load is defined in an EH pad, it can't be partially redundant,
+ // because the edges between the invoke and the EH pad cannot have other
// instructions between them.
- if (LoadBB->isLandingPad())
+ if (LoadBB->isEHPad())
return false;
Value *LoadedPtr = LI->getOperand(0);
// Scan a few instructions up from the load, to see if it is obviously live at
// the entry to its block.
- BasicBlock::iterator BBIt = LI;
+ BasicBlock::iterator BBIt(LI);
if (Value *AvailableVal =
- FindAvailableLoadedValue(LoadedPtr, LoadBB, BBIt, 6)) {
- // If the value if the load is locally available within the block, just use
+ FindAvailableLoadedValue(LoadedPtr, LoadBB, BBIt, DefMaxInstsToScan)) {
+ // If the value of the load is locally available within the block, just use
// it. This frequently occurs for reg2mem'd allocas.
//cerr << "LOAD ELIMINATED:\n" << *BBIt << *LI << "\n";
// Scan the predecessor to see if the value is available in the pred.
BBIt = PredBB->end();
AAMDNodes ThisAATags;
- Value *PredAvailable = FindAvailableLoadedValue(LoadedPtr, PredBB, BBIt, 6,
+ Value *PredAvailable = FindAvailableLoadedValue(LoadedPtr, PredBB, BBIt,
+ DefMaxInstsToScan,
nullptr, &ThisAATags);
if (!PredAvailable) {
OneUnavailablePred = PredBB;
}
// Split them out to their own block.
- UnavailablePred =
- SplitBlockPredecessors(LoadBB, PredsToSplit, "thread-pre-split");
+ UnavailablePred = SplitBlockPreds(LoadBB, PredsToSplit, "thread-pre-split");
}
// If the value isn't available in all predecessors, then there will be
// Create a PHI node at the start of the block for the PRE'd load value.
pred_iterator PB = pred_begin(LoadBB), PE = pred_end(LoadBB);
PHINode *PN = PHINode::Create(LI->getType(), std::distance(PB, PE), "",
- LoadBB->begin());
+ &LoadBB->front());
PN->takeName(LI);
PN->setDebugLoc(LI->getDebugLoc());
// Into:
// BB':
// %Y = icmp ne i32 %A, %B
- // br i1 %Z, ...
+ // br i1 %Y, ...
PredValueInfoTy XorOpValues;
bool isLHS = true;
return false;
}
- // And finally, do it! Start by factoring the predecessors is needed.
+ // And finally, do it! Start by factoring the predecessors if needed.
BasicBlock *PredBB;
if (PredBBs.size() == 1)
PredBB = PredBBs[0];
else {
DEBUG(dbgs() << " Factoring out " << PredBBs.size()
<< " common predecessors.\n");
- PredBB = SplitBlockPredecessors(BB, PredBBs, ".thr_comm");
+ PredBB = SplitBlockPreds(BB, PredBBs, ".thr_comm");
}
// And finally, do it!
BB->getParent(), BB);
NewBB->moveAfter(PredBB);
+ // Set the block frequency of NewBB.
+ if (HasProfileData) {
+ auto NewBBFreq =
+ BFI->getBlockFreq(PredBB) * BPI->getEdgeProbability(PredBB, BB);
+ BFI->setBlockFreq(NewBB, NewBBFreq.getFrequency());
+ }
+
BasicBlock::iterator BI = BB->begin();
for (; PHINode *PN = dyn_cast<PHINode>(BI); ++BI)
ValueMapping[PN] = PN->getIncomingValueForBlock(PredBB);
Instruction *New = BI->clone();
New->setName(BI->getName());
NewBB->getInstList().push_back(New);
- ValueMapping[BI] = New;
+ ValueMapping[&*BI] = New;
// Remap operands to patch up intra-block references.
for (unsigned i = 0, e = New->getNumOperands(); i != e; ++i)
// We didn't copy the terminator from BB over to NewBB, because there is now
// an unconditional jump to SuccBB. Insert the unconditional jump.
- BranchInst *NewBI =BranchInst::Create(SuccBB, NewBB);
+ BranchInst *NewBI = BranchInst::Create(SuccBB, NewBB);
NewBI->setDebugLoc(BB->getTerminator()->getDebugLoc());
// Check to see if SuccBB has PHI nodes. If so, we need to add entries to the
// its block to be uses of the appropriate PHI node etc. See ValuesInBlocks
// with the two values we know.
SSAUpdate.Initialize(I->getType(), I->getName());
- SSAUpdate.AddAvailableValue(BB, I);
- SSAUpdate.AddAvailableValue(NewBB, ValueMapping[I]);
+ SSAUpdate.AddAvailableValue(BB, &*I);
+ SSAUpdate.AddAvailableValue(NewBB, ValueMapping[&*I]);
while (!UsesToRename.empty())
SSAUpdate.RewriteUse(*UsesToRename.pop_back_val());
// frequently happens because of phi translation.
SimplifyInstructionsInBlock(NewBB, TLI);
+ // Update the edge weight from BB to SuccBB, which should be less than before.
+ UpdateBlockFreqAndEdgeWeight(PredBB, BB, NewBB, SuccBB);
+
// Threaded an edge!
++NumThreads;
return true;
}
+/// Create a new basic block that will be the predecessor of BB and successor of
+/// all blocks in Preds. When profile data is availble, update the frequency of
+/// this new block.
+BasicBlock *JumpThreading::SplitBlockPreds(BasicBlock *BB,
+ ArrayRef<BasicBlock *> Preds,
+ const char *Suffix) {
+ // Collect the frequencies of all predecessors of BB, which will be used to
+ // update the edge weight on BB->SuccBB.
+ BlockFrequency PredBBFreq(0);
+ if (HasProfileData)
+ for (auto Pred : Preds)
+ PredBBFreq += BFI->getBlockFreq(Pred) * BPI->getEdgeProbability(Pred, BB);
+
+ BasicBlock *PredBB = SplitBlockPredecessors(BB, Preds, Suffix);
+
+ // Set the block frequency of the newly created PredBB, which is the sum of
+ // frequencies of Preds.
+ if (HasProfileData)
+ BFI->setBlockFreq(PredBB, PredBBFreq.getFrequency());
+ return PredBB;
+}
+
+/// Update the block frequency of BB and branch weight and the metadata on the
+/// edge BB->SuccBB. This is done by scaling the weight of BB->SuccBB by 1 -
+/// Freq(PredBB->BB) / Freq(BB->SuccBB).
+void JumpThreading::UpdateBlockFreqAndEdgeWeight(BasicBlock *PredBB,
+ BasicBlock *BB,
+ BasicBlock *NewBB,
+ BasicBlock *SuccBB) {
+ if (!HasProfileData)
+ return;
+
+ assert(BFI && BPI && "BFI & BPI should have been created here");
+
+ // As the edge from PredBB to BB is deleted, we have to update the block
+ // frequency of BB.
+ auto BBOrigFreq = BFI->getBlockFreq(BB);
+ auto NewBBFreq = BFI->getBlockFreq(NewBB);
+ auto BB2SuccBBFreq = BBOrigFreq * BPI->getEdgeProbability(BB, SuccBB);
+ auto BBNewFreq = BBOrigFreq - NewBBFreq;
+ BFI->setBlockFreq(BB, BBNewFreq.getFrequency());
+
+ // Collect updated outgoing edges' frequencies from BB and use them to update
+ // edge weights.
+ SmallVector<uint64_t, 4> BBSuccFreq;
+ for (auto I = succ_begin(BB), E = succ_end(BB); I != E; ++I) {
+ auto SuccFreq = (*I == SuccBB)
+ ? BB2SuccBBFreq - NewBBFreq
+ : BBOrigFreq * BPI->getEdgeProbability(BB, *I);
+ BBSuccFreq.push_back(SuccFreq.getFrequency());
+ }
+
+ // Normalize edge weights in Weights64 so that the sum of them can fit in
+ BranchProbability::normalizeEdgeWeights(BBSuccFreq.begin(), BBSuccFreq.end());
+
+ SmallVector<uint32_t, 4> Weights;
+ for (auto Freq : BBSuccFreq)
+ Weights.push_back(static_cast<uint32_t>(Freq));
+
+ // Update edge weights in BPI.
+ for (int I = 0, E = Weights.size(); I < E; I++)
+ BPI->setEdgeWeight(BB, I, Weights[I]);
+
+ if (Weights.size() >= 2) {
+ auto TI = BB->getTerminator();
+ TI->setMetadata(
+ LLVMContext::MD_prof,
+ MDBuilder(TI->getParent()->getContext()).createBranchWeights(Weights));
+ }
+}
+
/// DuplicateCondBranchOnPHIIntoPred - PredBB contains an unconditional branch
/// to BB which contains an i1 PHI node and a conditional branch on that PHI.
/// If we can duplicate the contents of BB up into PredBB do so now, this
return false;
}
- // And finally, do it! Start by factoring the predecessors is needed.
+ // And finally, do it! Start by factoring the predecessors if needed.
BasicBlock *PredBB;
if (PredBBs.size() == 1)
PredBB = PredBBs[0];
else {
DEBUG(dbgs() << " Factoring out " << PredBBs.size()
<< " common predecessors.\n");
- PredBB = SplitBlockPredecessors(BB, PredBBs, ".thr_comm");
+ PredBB = SplitBlockPreds(BB, PredBBs, ".thr_comm");
}
// Okay, we decided to do this! Clone all the instructions in BB onto the end
if (Value *IV =
SimplifyInstruction(New, BB->getModule()->getDataLayout())) {
delete New;
- ValueMapping[BI] = IV;
+ ValueMapping[&*BI] = IV;
} else {
// Otherwise, insert the new instruction into the block.
New->setName(BI->getName());
- PredBB->getInstList().insert(OldPredBranch, New);
- ValueMapping[BI] = New;
+ PredBB->getInstList().insert(OldPredBranch->getIterator(), New);
+ ValueMapping[&*BI] = New;
}
}
// its block to be uses of the appropriate PHI node etc. See ValuesInBlocks
// with the two values we know.
SSAUpdate.Initialize(I->getType(), I->getName());
- SSAUpdate.AddAvailableValue(BB, I);
- SSAUpdate.AddAvailableValue(PredBB, ValueMapping[I]);
+ SSAUpdate.AddAvailableValue(BB, &*I);
+ SSAUpdate.AddAvailableValue(PredBB, ValueMapping[&*I]);
while (!UsesToRename.empty())
SSAUpdate.RewriteUse(*UsesToRename.pop_back_val());
OSDN Git Service
