//
//===----------------------------------------------------------------------===//
-#define DEBUG_TYPE "licm"
#include "llvm/Transforms/Scalar.h"
#include "llvm/ADT/Statistic.h"
#include "llvm/Analysis/AliasAnalysis.h"
#include "llvm/Analysis/LoopInfo.h"
#include "llvm/Analysis/LoopPass.h"
#include "llvm/Analysis/ScalarEvolution.h"
+#include "llvm/Analysis/TargetLibraryInfo.h"
#include "llvm/Analysis/ValueTracking.h"
+#include "llvm/IR/CFG.h"
#include "llvm/IR/Constants.h"
#include "llvm/IR/DataLayout.h"
#include "llvm/IR/DerivedTypes.h"
#include "llvm/IR/IntrinsicInst.h"
#include "llvm/IR/LLVMContext.h"
#include "llvm/IR/Metadata.h"
-#include "llvm/Support/CFG.h"
+#include "llvm/IR/PredIteratorCache.h"
#include "llvm/Support/CommandLine.h"
#include "llvm/Support/Debug.h"
-#include "llvm/Support/PredIteratorCache.h"
#include "llvm/Support/raw_ostream.h"
-#include "llvm/Target/TargetLibraryInfo.h"
#include "llvm/Transforms/Utils/Local.h"
#include "llvm/Transforms/Utils/LoopUtils.h"
#include "llvm/Transforms/Utils/SSAUpdater.h"
#include <algorithm>
using namespace llvm;
+#define DEBUG_TYPE "licm"
+
STATISTIC(NumSunk , "Number of instructions sunk out of loop");
STATISTIC(NumHoisted , "Number of instructions hoisted out of loop");
STATISTIC(NumMovedLoads, "Number of load insts hoisted or sunk");
DisablePromotion("disable-licm-promotion", cl::Hidden,
cl::desc("Disable memory promotion in LICM pass"));
+static bool inSubLoop(BasicBlock *BB, Loop *CurLoop, LoopInfo *LI);
+static bool isNotUsedInLoop(const Instruction &I, const Loop *CurLoop);
+static bool hoist(Instruction &I, BasicBlock *Preheader);
+static bool sink(Instruction &I, const LoopInfo *LI, const DominatorTree *DT,
+ const Loop *CurLoop, AliasSetTracker *CurAST );
+static bool isGuaranteedToExecute(const Instruction &Inst,
+ const DominatorTree *DT,
+ const Loop *CurLoop,
+ const LICMSafetyInfo *SafetyInfo);
+static bool isSafeToExecuteUnconditionally(const Instruction &Inst,
+ const DominatorTree *DT,
+ const TargetLibraryInfo *TLI,
+ const Loop *CurLoop,
+ const LICMSafetyInfo *SafetyInfo,
+ const Instruction *CtxI = nullptr);
+static bool pointerInvalidatedByLoop(Value *V, uint64_t Size,
+ const AAMDNodes &AAInfo,
+ AliasSetTracker *CurAST);
+static Instruction *CloneInstructionInExitBlock(const Instruction &I,
+ BasicBlock &ExitBlock,
+ PHINode &PN,
+ const LoopInfo *LI);
+static bool canSinkOrHoistInst(Instruction &I, AliasAnalysis *AA,
+ DominatorTree *DT, TargetLibraryInfo *TLI,
+ Loop *CurLoop, AliasSetTracker *CurAST,
+ LICMSafetyInfo *SafetyInfo);
+
namespace {
struct LICM : public LoopPass {
static char ID; // Pass identification, replacement for typeid
initializeLICMPass(*PassRegistry::getPassRegistry());
}
- virtual bool runOnLoop(Loop *L, LPPassManager &LPM);
+ bool runOnLoop(Loop *L, LPPassManager &LPM) override;
/// This transformation requires natural loop information & requires that
/// loop preheaders be inserted into the CFG...
///
- virtual void getAnalysisUsage(AnalysisUsage &AU) const {
+ void getAnalysisUsage(AnalysisUsage &AU) const override {
AU.setPreservesCFG();
AU.addRequired<DominatorTreeWrapperPass>();
- AU.addRequired<LoopInfo>();
+ AU.addRequired<LoopInfoWrapperPass>();
AU.addRequiredID(LoopSimplifyID);
AU.addPreservedID(LoopSimplifyID);
AU.addRequiredID(LCSSAID);
AU.addRequired<AliasAnalysis>();
AU.addPreserved<AliasAnalysis>();
AU.addPreserved<ScalarEvolution>();
- AU.addRequired<TargetLibraryInfo>();
+ AU.addRequired<TargetLibraryInfoWrapperPass>();
}
using llvm::Pass::doFinalization;
- bool doFinalization() {
+ bool doFinalization() override {
assert(LoopToAliasSetMap.empty() && "Didn't free loop alias sets");
return false;
}
LoopInfo *LI; // Current LoopInfo
DominatorTree *DT; // Dominator Tree for the current Loop.
- DataLayout *DL; // DataLayout for constant folding.
TargetLibraryInfo *TLI; // TargetLibraryInfo for constant folding.
// State that is updated as we process loops.
BasicBlock *Preheader; // The preheader block of the current loop...
Loop *CurLoop; // The current loop we are working on...
AliasSetTracker *CurAST; // AliasSet information for the current loop...
- bool MayThrow; // The current loop contains an instruction which
- // may throw, thus preventing code motion of
- // instructions with side effects.
DenseMap<Loop*, AliasSetTracker*> LoopToAliasSetMap;
/// cloneBasicBlockAnalysis - Simple Analysis hook. Clone alias set info.
- void cloneBasicBlockAnalysis(BasicBlock *From, BasicBlock *To, Loop *L);
+ void cloneBasicBlockAnalysis(BasicBlock *From, BasicBlock *To,
+ Loop *L) override;
/// deleteAnalysisValue - Simple Analysis hook. Delete value V from alias
/// set.
- void deleteAnalysisValue(Value *V, Loop *L);
-
- /// SinkRegion - Walk the specified region of the CFG (defined by all blocks
- /// dominated by the specified block, and that are in the current loop) in
- /// reverse depth first order w.r.t the DominatorTree. This allows us to
- /// visit uses before definitions, allowing us to sink a loop body in one
- /// pass without iteration.
- ///
- void SinkRegion(DomTreeNode *N);
-
- /// HoistRegion - Walk the specified region of the CFG (defined by all
- /// blocks dominated by the specified block, and that are in the current
- /// loop) in depth first order w.r.t the DominatorTree. This allows us to
- /// visit definitions before uses, allowing us to hoist a loop body in one
- /// pass without iteration.
- ///
- void HoistRegion(DomTreeNode *N);
-
- /// inSubLoop - Little predicate that returns true if the specified basic
- /// block is in a subloop of the current one, not the current one itself.
- ///
- bool inSubLoop(BasicBlock *BB) {
- assert(CurLoop->contains(BB) && "Only valid if BB is IN the loop");
- return LI->getLoopFor(BB) != CurLoop;
- }
-
- /// sink - When an instruction is found to only be used outside of the loop,
- /// this function moves it to the exit blocks and patches up SSA form as
- /// needed.
- ///
- void sink(Instruction &I);
-
- /// hoist - When an instruction is found to only use loop invariant operands
- /// that is safe to hoist, this instruction is called to do the dirty work.
- ///
- void hoist(Instruction &I);
-
- /// isSafeToExecuteUnconditionally - Only sink or hoist an instruction if it
- /// is not a trapping instruction or if it is a trapping instruction and is
- /// guaranteed to execute.
- ///
- bool isSafeToExecuteUnconditionally(Instruction &I);
-
- /// isGuaranteedToExecute - Check that the instruction is guaranteed to
- /// execute.
- ///
- bool isGuaranteedToExecute(Instruction &I);
-
- /// pointerInvalidatedByLoop - Return true if the body of this loop may
- /// store into the memory location pointed to by V.
- ///
- bool pointerInvalidatedByLoop(Value *V, uint64_t Size,
- const MDNode *TBAAInfo) {
- // Check to see if any of the basic blocks in CurLoop invalidate *V.
- return CurAST->getAliasSetForPointer(V, Size, TBAAInfo).isMod();
- }
-
- bool canSinkOrHoistInst(Instruction &I);
- bool isNotUsedInLoop(Instruction &I);
+ void deleteAnalysisValue(Value *V, Loop *L) override;
- void PromoteAliasSet(AliasSet &AS,
- SmallVectorImpl<BasicBlock*> &ExitBlocks,
- SmallVectorImpl<Instruction*> &InsertPts,
- PredIteratorCache &PIC);
+ /// Simple Analysis hook. Delete loop L from alias set map.
+ void deleteAnalysisLoop(Loop *L) override;
};
}
char LICM::ID = 0;
INITIALIZE_PASS_BEGIN(LICM, "licm", "Loop Invariant Code Motion", false, false)
INITIALIZE_PASS_DEPENDENCY(DominatorTreeWrapperPass)
-INITIALIZE_PASS_DEPENDENCY(LoopInfo)
+INITIALIZE_PASS_DEPENDENCY(LoopInfoWrapperPass)
INITIALIZE_PASS_DEPENDENCY(LoopSimplify)
INITIALIZE_PASS_DEPENDENCY(LCSSA)
INITIALIZE_PASS_DEPENDENCY(ScalarEvolution)
-INITIALIZE_PASS_DEPENDENCY(TargetLibraryInfo)
+INITIALIZE_PASS_DEPENDENCY(TargetLibraryInfoWrapperPass)
INITIALIZE_AG_DEPENDENCY(AliasAnalysis)
INITIALIZE_PASS_END(LICM, "licm", "Loop Invariant Code Motion", false, false)
Changed = false;
// Get our Loop and Alias Analysis information...
- LI = &getAnalysis<LoopInfo>();
+ LI = &getAnalysis<LoopInfoWrapperPass>().getLoopInfo();
AA = &getAnalysis<AliasAnalysis>();
DT = &getAnalysis<DominatorTreeWrapperPass>().getDomTree();
- DL = getAnalysisIfAvailable<DataLayout>();
- TLI = &getAnalysis<TargetLibraryInfo>();
+ TLI = &getAnalysis<TargetLibraryInfoWrapperPass>().getTLI();
assert(L->isLCSSAForm(*DT) && "Loop is not in LCSSA form.");
CurAST->add(*BB); // Incorporate the specified basic block
}
- MayThrow = false;
- // TODO: We've already searched for instructions which may throw in subloops.
- // We may want to reuse this information.
- for (Loop::block_iterator BB = L->block_begin(), BBE = L->block_end();
- (BB != BBE) && !MayThrow ; ++BB)
- for (BasicBlock::iterator I = (*BB)->begin(), E = (*BB)->end();
- (I != E) && !MayThrow; ++I)
- MayThrow |= I->mayThrow();
+ // Compute loop safety information.
+ LICMSafetyInfo SafetyInfo;
+ computeLICMSafetyInfo(&SafetyInfo, CurLoop);
// We want to visit all of the instructions in this loop... that are not parts
// of our subloops (they have already had their invariants hoisted out of
// instructions, we perform another pass to hoist them out of the loop.
//
if (L->hasDedicatedExits())
- SinkRegion(DT->getNode(L->getHeader()));
+ Changed |= sinkRegion(DT->getNode(L->getHeader()), AA, LI, DT, TLI, CurLoop,
+ CurAST, &SafetyInfo);
if (Preheader)
- HoistRegion(DT->getNode(L->getHeader()));
+ Changed |= hoistRegion(DT->getNode(L->getHeader()), AA, LI, DT, TLI,
+ CurLoop, CurAST, &SafetyInfo);
// Now that all loop invariants have been removed from the loop, promote any
// memory references to scalars that we can.
// Loop over all of the alias sets in the tracker object.
for (AliasSetTracker::iterator I = CurAST->begin(), E = CurAST->end();
I != E; ++I)
- PromoteAliasSet(*I, ExitBlocks, InsertPts, PIC);
+ Changed |= promoteLoopAccessesToScalars(*I, ExitBlocks, InsertPts,
+ PIC, LI, DT, CurLoop,
+ CurAST, &SafetyInfo);
// Once we have promoted values across the loop body we have to recursively
// reform LCSSA as any nested loop may now have values defined within the
// SSAUpdater strategy during promotion that was LCSSA aware and reformed
// it as it went.
if (Changed)
- formLCSSARecursively(*L, *DT, getAnalysisIfAvailable<ScalarEvolution>());
+ formLCSSARecursively(*L, *DT, LI,
+ getAnalysisIfAvailable<ScalarEvolution>());
}
// Check that neither this loop nor its parent have had LCSSA broken. LICM is
"Parent loop not left in LCSSA form after LICM!");
// Clear out loops state information for the next iteration
- CurLoop = 0;
- Preheader = 0;
+ CurLoop = nullptr;
+ Preheader = nullptr;
// If this loop is nested inside of another one, save the alias information
// for when we process the outer loop.
return Changed;
}
-/// SinkRegion - Walk the specified region of the CFG (defined by all blocks
-/// dominated by the specified block, and that are in the current loop) in
-/// reverse depth first order w.r.t the DominatorTree. This allows us to visit
-/// uses before definitions, allowing us to sink a loop body in one pass without
-/// iteration.
+/// Walk the specified region of the CFG (defined by all blocks dominated by
+/// the specified block, and that are in the current loop) in reverse depth
+/// first order w.r.t the DominatorTree. This allows us to visit uses before
+/// definitions, allowing us to sink a loop body in one pass without iteration.
///
-void LICM::SinkRegion(DomTreeNode *N) {
- assert(N != 0 && "Null dominator tree node?");
+bool llvm::sinkRegion(DomTreeNode *N, AliasAnalysis *AA, LoopInfo *LI,
+ DominatorTree *DT, TargetLibraryInfo *TLI, Loop *CurLoop,
+ AliasSetTracker *CurAST, LICMSafetyInfo *SafetyInfo) {
+
+ // Verify inputs.
+ assert(N != nullptr && AA != nullptr && LI != nullptr &&
+ DT != nullptr && CurLoop != nullptr && CurAST != nullptr &&
+ SafetyInfo != nullptr && "Unexpected input to sinkRegion");
+
+ // Set changed as false.
+ bool Changed = false;
+ // Get basic block
BasicBlock *BB = N->getBlock();
-
// If this subregion is not in the top level loop at all, exit.
- if (!CurLoop->contains(BB)) return;
+ if (!CurLoop->contains(BB)) return Changed;
// We are processing blocks in reverse dfo, so process children first.
const std::vector<DomTreeNode*> &Children = N->getChildren();
for (unsigned i = 0, e = Children.size(); i != e; ++i)
- SinkRegion(Children[i]);
-
+ Changed |=
+ sinkRegion(Children[i], AA, LI, DT, TLI, CurLoop, CurAST, SafetyInfo);
// Only need to process the contents of this block if it is not part of a
// subloop (which would already have been processed).
- if (inSubLoop(BB)) return;
+ if (inSubLoop(BB,CurLoop,LI)) return Changed;
for (BasicBlock::iterator II = BB->end(); II != BB->begin(); ) {
Instruction &I = *--II;
// outside of the loop. In this case, it doesn't even matter if the
// operands of the instruction are loop invariant.
//
- if (isNotUsedInLoop(I) && canSinkOrHoistInst(I)) {
+ if (isNotUsedInLoop(I, CurLoop) &&
+ canSinkOrHoistInst(I, AA, DT, TLI, CurLoop, CurAST, SafetyInfo)) {
++II;
- sink(I);
+ Changed |= sink(I, LI, DT, CurLoop, CurAST);
}
}
+ return Changed;
}
-/// HoistRegion - Walk the specified region of the CFG (defined by all blocks
-/// dominated by the specified block, and that are in the current loop) in depth
-/// first order w.r.t the DominatorTree. This allows us to visit definitions
-/// before uses, allowing us to hoist a loop body in one pass without iteration.
+/// Walk the specified region of the CFG (defined by all blocks dominated by
+/// the specified block, and that are in the current loop) in depth first
+/// order w.r.t the DominatorTree. This allows us to visit definitions before
+/// uses, allowing us to hoist a loop body in one pass without iteration.
///
-void LICM::HoistRegion(DomTreeNode *N) {
- assert(N != 0 && "Null dominator tree node?");
+bool llvm::hoistRegion(DomTreeNode *N, AliasAnalysis *AA, LoopInfo *LI,
+ DominatorTree *DT, TargetLibraryInfo *TLI, Loop *CurLoop,
+ AliasSetTracker *CurAST, LICMSafetyInfo *SafetyInfo) {
+ // Verify inputs.
+ assert(N != nullptr && AA != nullptr && LI != nullptr &&
+ DT != nullptr && CurLoop != nullptr && CurAST != nullptr &&
+ SafetyInfo != nullptr && "Unexpected input to hoistRegion");
+ // Set changed as false.
+ bool Changed = false;
+ // Get basic block
BasicBlock *BB = N->getBlock();
-
// If this subregion is not in the top level loop at all, exit.
- if (!CurLoop->contains(BB)) return;
-
+ if (!CurLoop->contains(BB)) return Changed;
// Only need to process the contents of this block if it is not part of a
// subloop (which would already have been processed).
- if (!inSubLoop(BB))
+ if (!inSubLoop(BB, CurLoop, LI))
for (BasicBlock::iterator II = BB->begin(), E = BB->end(); II != E; ) {
Instruction &I = *II++;
-
// Try constant folding this instruction. If all the operands are
// constants, it is technically hoistable, but it would be better to just
// fold it.
- if (Constant *C = ConstantFoldInstruction(&I, DL, TLI)) {
+ if (Constant *C = ConstantFoldInstruction(
+ &I, I.getModule()->getDataLayout(), TLI)) {
DEBUG(dbgs() << "LICM folding inst: " << I << " --> " << *C << '\n');
CurAST->copyValue(&I, C);
CurAST->deleteValue(&I);
// if all of the operands of the instruction are loop invariant and if it
// is safe to hoist the instruction.
//
- if (CurLoop->hasLoopInvariantOperands(&I) && canSinkOrHoistInst(I) &&
- isSafeToExecuteUnconditionally(I))
- hoist(I);
+ if (CurLoop->hasLoopInvariantOperands(&I) &&
+ canSinkOrHoistInst(I, AA, DT, TLI, CurLoop, CurAST, SafetyInfo) &&
+ isSafeToExecuteUnconditionally(I, DT, TLI, CurLoop, SafetyInfo,
+ CurLoop->getLoopPreheader()->getTerminator()))
+ Changed |= hoist(I, CurLoop->getLoopPreheader());
}
const std::vector<DomTreeNode*> &Children = N->getChildren();
for (unsigned i = 0, e = Children.size(); i != e; ++i)
- HoistRegion(Children[i]);
+ Changed |=
+ hoistRegion(Children[i], AA, LI, DT, TLI, CurLoop, CurAST, SafetyInfo);
+ return Changed;
+}
+
+/// Computes loop safety information, checks loop body & header
+/// for the possiblity of may throw exception.
+///
+void llvm::computeLICMSafetyInfo(LICMSafetyInfo * SafetyInfo, Loop * CurLoop) {
+ assert(CurLoop != nullptr && "CurLoop cant be null");
+ BasicBlock *Header = CurLoop->getHeader();
+ // Setting default safety values.
+ SafetyInfo->MayThrow = false;
+ SafetyInfo->HeaderMayThrow = false;
+ // Iterate over header and compute dafety info.
+ for (BasicBlock::iterator I = Header->begin(), E = Header->end();
+ (I != E) && !SafetyInfo->HeaderMayThrow; ++I)
+ SafetyInfo->HeaderMayThrow |= I->mayThrow();
+
+ SafetyInfo->MayThrow = SafetyInfo->HeaderMayThrow;
+ // Iterate over loop instructions and compute safety info.
+ for (Loop::block_iterator BB = CurLoop->block_begin(),
+ BBE = CurLoop->block_end(); (BB != BBE) && !SafetyInfo->MayThrow ; ++BB)
+ for (BasicBlock::iterator I = (*BB)->begin(), E = (*BB)->end();
+ (I != E) && !SafetyInfo->MayThrow; ++I)
+ SafetyInfo->MayThrow |= I->mayThrow();
}
/// canSinkOrHoistInst - Return true if the hoister and sinker can handle this
/// instruction.
///
-bool LICM::canSinkOrHoistInst(Instruction &I) {
+bool canSinkOrHoistInst(Instruction &I, AliasAnalysis *AA, DominatorTree *DT,
+ TargetLibraryInfo *TLI, Loop *CurLoop,
+ AliasSetTracker *CurAST, LICMSafetyInfo *SafetyInfo) {
// Loads have extra constraints we have to verify before we can hoist them.
if (LoadInst *LI = dyn_cast<LoadInst>(&I)) {
if (!LI->isUnordered())
// in the same alias set as something that ends up being modified.
if (AA->pointsToConstantMemory(LI->getOperand(0)))
return true;
- if (LI->getMetadata("invariant.load"))
+ if (LI->getMetadata(LLVMContext::MD_invariant_load))
return true;
// Don't hoist loads which have may-aliased stores in loop.
uint64_t Size = 0;
if (LI->getType()->isSized())
Size = AA->getTypeStoreSize(LI->getType());
- return !pointerInvalidatedByLoop(LI->getOperand(0), Size,
- LI->getMetadata(LLVMContext::MD_tbaa));
+
+ AAMDNodes AAInfo;
+ LI->getAAMetadata(AAInfo);
+
+ return !pointerInvalidatedByLoop(LI->getOperand(0), Size, AAInfo, CurAST);
} else if (CallInst *CI = dyn_cast<CallInst>(&I)) {
// Don't sink or hoist dbg info; it's legal, but not useful.
if (isa<DbgInfoIntrinsic>(I))
!isa<InsertValueInst>(I))
return false;
- return isSafeToExecuteUnconditionally(I);
+ // TODO: Plumb the context instruction through to make hoisting and sinking
+ // more powerful. Hoisting of loads already works due to the special casing
+ // above.
+ return isSafeToExecuteUnconditionally(I, DT, TLI, CurLoop, SafetyInfo,
+ nullptr);
}
-/// \brief Returns true if a PHINode is a trivially replaceable with an
+/// Returns true if a PHINode is a trivially replaceable with an
/// Instruction.
+/// This is true when all incoming values are that instruction.
+/// This pattern occurs most often with LCSSA PHI nodes.
///
-/// This is true when all incoming values are that instruction. This pattern
-/// occurs most often with LCSSA PHI nodes.
-static bool isTriviallyReplacablePHI(PHINode &PN, Instruction &I) {
- for (unsigned i = 0, e = PN.getNumIncomingValues(); i != e; ++i)
- if (PN.getIncomingValue(i) != &I)
+static bool isTriviallyReplacablePHI(const PHINode &PN, const Instruction &I) {
+ for (const Value *IncValue : PN.incoming_values())
+ if (IncValue != &I)
return false;
return true;
}
-/// isNotUsedInLoop - Return true if the only users of this instruction are
-/// outside of the loop. If this is true, we can sink the instruction to the
-/// exit blocks of the loop.
+/// Return true if the only users of this instruction are outside of
+/// the loop. If this is true, we can sink the instruction to the exit
+/// blocks of the loop.
///
-bool LICM::isNotUsedInLoop(Instruction &I) {
- for (Value::use_iterator UI = I.use_begin(), E = I.use_end(); UI != E; ++UI) {
- Instruction *User = cast<Instruction>(*UI);
- if (PHINode *PN = dyn_cast<PHINode>(User)) {
+static bool isNotUsedInLoop(const Instruction &I, const Loop *CurLoop) {
+ for (const User *U : I.users()) {
+ const Instruction *UI = cast<Instruction>(U);
+ if (const PHINode *PN = dyn_cast<PHINode>(UI)) {
// A PHI node where all of the incoming values are this instruction are
// special -- they can just be RAUW'ed with the instruction and thus
// don't require a use in the predecessor. This is a particular important
continue;
}
- if (CurLoop->contains(User))
+ if (CurLoop->contains(UI))
return false;
}
return true;
}
-/// sink - When an instruction is found to only be used outside of the loop,
-/// this function moves it to the exit blocks and patches up SSA form as needed.
+static Instruction *CloneInstructionInExitBlock(const Instruction &I,
+ BasicBlock &ExitBlock,
+ PHINode &PN,
+ const LoopInfo *LI) {
+ Instruction *New = I.clone();
+ ExitBlock.getInstList().insert(ExitBlock.getFirstInsertionPt(), New);
+ if (!I.getName().empty()) New->setName(I.getName() + ".le");
+
+ // Build LCSSA PHI nodes for any in-loop operands. Note that this is
+ // particularly cheap because we can rip off the PHI node that we're
+ // replacing for the number and blocks of the predecessors.
+ // OPT: If this shows up in a profile, we can instead finish sinking all
+ // invariant instructions, and then walk their operands to re-establish
+ // LCSSA. That will eliminate creating PHI nodes just to nuke them when
+ // sinking bottom-up.
+ for (User::op_iterator OI = New->op_begin(), OE = New->op_end(); OI != OE;
+ ++OI)
+ if (Instruction *OInst = dyn_cast<Instruction>(*OI))
+ if (Loop *OLoop = LI->getLoopFor(OInst->getParent()))
+ if (!OLoop->contains(&PN)) {
+ PHINode *OpPN =
+ PHINode::Create(OInst->getType(), PN.getNumIncomingValues(),
+ OInst->getName() + ".lcssa", ExitBlock.begin());
+ for (unsigned i = 0, e = PN.getNumIncomingValues(); i != e; ++i)
+ OpPN->addIncoming(OInst, PN.getIncomingBlock(i));
+ *OI = OpPN;
+ }
+ return New;
+}
+
+/// When an instruction is found to only be used outside of the loop, this
+/// function moves it to the exit blocks and patches up SSA form as needed.
/// This method is guaranteed to remove the original instruction from its
/// position, and may either delete it or move it to outside of the loop.
///
-void LICM::sink(Instruction &I) {
+static bool sink(Instruction &I, const LoopInfo *LI, const DominatorTree *DT,
+ const Loop *CurLoop, AliasSetTracker *CurAST ) {
DEBUG(dbgs() << "LICM sinking instruction: " << I << "\n");
-
+ bool Changed = false;
if (isa<LoadInst>(I)) ++NumMovedLoads;
else if (isa<CallInst>(I)) ++NumMovedCalls;
++NumSunk;
#ifndef NDEBUG
SmallVector<BasicBlock *, 32> ExitBlocks;
CurLoop->getUniqueExitBlocks(ExitBlocks);
- SmallPtrSet<BasicBlock *, 32> ExitBlockSet(ExitBlocks.begin(), ExitBlocks.end());
+ SmallPtrSet<BasicBlock *, 32> ExitBlockSet(ExitBlocks.begin(),
+ ExitBlocks.end());
#endif
+ // Clones of this instruction. Don't create more than one per exit block!
+ SmallDenseMap<BasicBlock *, Instruction *, 32> SunkCopies;
+
// If this instruction is only used outside of the loop, then all users are
// PHI nodes in exit blocks due to LCSSA form. Just RAUW them with clones of
// the instruction.
while (!I.use_empty()) {
+ Instruction *User = I.user_back();
+ if (!DT->isReachableFromEntry(User->getParent())) {
+ User->replaceUsesOfWith(&I, UndefValue::get(I.getType()));
+ continue;
+ }
// The user must be a PHI node.
- PHINode *PN = cast<PHINode>(I.use_back());
+ PHINode *PN = cast<PHINode>(User);
BasicBlock *ExitBlock = PN->getParent();
assert(ExitBlockSet.count(ExitBlock) &&
"The LCSSA PHI is not in an exit block!");
- Instruction *New = I.clone();
- ExitBlock->getInstList().insert(ExitBlock->getFirstInsertionPt(), New);
- if (!I.getName().empty())
- New->setName(I.getName() + ".le");
-
- // Build LCSSA PHI nodes for any in-loop operands. Note that this is
- // particularly cheap because we can rip off the PHI node that we're
- // replacing for the number and blocks of the predecessors.
- // OPT: If this shows up in a profile, we can instead finish sinking all
- // invariant instructions, and then walk their operands to re-establish
- // LCSSA. That will eliminate creating PHI nodes just to nuke them when
- // sinking bottom-up.
- for (User::op_iterator OI = New->op_begin(), OE = New->op_end(); OI != OE;
- ++OI)
- if (Instruction *OInst = dyn_cast<Instruction>(*OI))
- if (Loop *OLoop = LI->getLoopFor(OInst->getParent()))
- if (!OLoop->contains(PN)) {
- PHINode *OpPN = PHINode::Create(
- OInst->getType(), PN->getNumIncomingValues(),
- OInst->getName() + ".lcssa", ExitBlock->begin());
- for (unsigned i = 0, e = PN->getNumIncomingValues(); i != e; ++i)
- OpPN->addIncoming(OInst, PN->getIncomingBlock(i));
- *OI = OpPN;
- }
+ Instruction *New;
+ auto It = SunkCopies.find(ExitBlock);
+ if (It != SunkCopies.end())
+ New = It->second;
+ else
+ New = SunkCopies[ExitBlock] =
+ CloneInstructionInExitBlock(I, *ExitBlock, *PN, LI);
PN->replaceAllUsesWith(New);
PN->eraseFromParent();
CurAST->deleteValue(&I);
I.eraseFromParent();
+ return Changed;
}
-/// hoist - When an instruction is found to only use loop invariant operands
-/// that is safe to hoist, this instruction is called to do the dirty work.
+/// When an instruction is found to only use loop invariant operands that
+/// is safe to hoist, this instruction is called to do the dirty work.
///
-void LICM::hoist(Instruction &I) {
+static bool hoist(Instruction &I, BasicBlock *Preheader) {
DEBUG(dbgs() << "LICM hoisting to " << Preheader->getName() << ": "
<< I << "\n");
-
// Move the new node to the Preheader, before its terminator.
I.moveBefore(Preheader->getTerminator());
if (isa<LoadInst>(I)) ++NumMovedLoads;
else if (isa<CallInst>(I)) ++NumMovedCalls;
++NumHoisted;
- Changed = true;
+ return true;
}
-/// isSafeToExecuteUnconditionally - Only sink or hoist an instruction if it is
-/// not a trapping instruction or if it is a trapping instruction and is
-/// guaranteed to execute.
-///
-bool LICM::isSafeToExecuteUnconditionally(Instruction &Inst) {
- // If it is not a trapping instruction, it is always safe to hoist.
- if (isSafeToSpeculativelyExecute(&Inst))
+/// Only sink or hoist an instruction if it is not a trapping instruction,
+/// or if the instruction is known not to trap when moved to the preheader.
+/// or if it is a trapping instruction and is guaranteed to execute.
+static bool isSafeToExecuteUnconditionally(const Instruction &Inst,
+ const DominatorTree *DT,
+ const TargetLibraryInfo *TLI,
+ const Loop *CurLoop,
+ const LICMSafetyInfo *SafetyInfo,
+ const Instruction *CtxI) {
+ if (isSafeToSpeculativelyExecute(&Inst, CtxI, DT, TLI))
return true;
- return isGuaranteedToExecute(Inst);
+ return isGuaranteedToExecute(Inst, DT, CurLoop, SafetyInfo);
}
-bool LICM::isGuaranteedToExecute(Instruction &Inst) {
+static bool isGuaranteedToExecute(const Instruction &Inst,
+ const DominatorTree *DT,
+ const Loop *CurLoop,
+ const LICMSafetyInfo * SafetyInfo) {
- // Somewhere in this loop there is an instruction which may throw and make us
- // exit the loop.
- if (MayThrow)
- return false;
-
- // Otherwise we have to check to make sure that the instruction dominates all
+ // We have to check to make sure that the instruction dominates all
// of the exit blocks. If it doesn't, then there is a path out of the loop
// which does not execute this instruction, so we can't hoist it.
// common), it is always guaranteed to dominate the exit blocks. Since this
// is a common case, and can save some work, check it now.
if (Inst.getParent() == CurLoop->getHeader())
- return true;
+ // If there's a throw in the header block, we can't guarantee we'll reach
+ // Inst.
+ return !SafetyInfo->HeaderMayThrow;
+
+ // Somewhere in this loop there is an instruction which may throw and make us
+ // exit the loop.
+ if (SafetyInfo->MayThrow)
+ return false;
// Get the exit blocks for the current loop.
SmallVector<BasicBlock*, 8> ExitBlocks;
namespace {
class LoopPromoter : public LoadAndStorePromoter {
Value *SomePtr; // Designated pointer to store to.
- SmallPtrSet<Value*, 4> &PointerMustAliases;
+ SmallPtrSetImpl<Value*> &PointerMustAliases;
SmallVectorImpl<BasicBlock*> &LoopExitBlocks;
SmallVectorImpl<Instruction*> &LoopInsertPts;
PredIteratorCache &PredCache;
LoopInfo &LI;
DebugLoc DL;
int Alignment;
- MDNode *TBAATag;
+ AAMDNodes AATags;
Value *maybeInsertLCSSAPHI(Value *V, BasicBlock *BB) const {
if (Instruction *I = dyn_cast<Instruction>(V))
// We need to create an LCSSA PHI node for the incoming value and
// store that.
PHINode *PN = PHINode::Create(
- I->getType(), PredCache.GetNumPreds(BB),
+ I->getType(), PredCache.size(BB),
I->getName() + ".lcssa", BB->begin());
- for (BasicBlock **PI = PredCache.GetPreds(BB); *PI; ++PI)
- PN->addIncoming(I, *PI);
+ for (BasicBlock *Pred : PredCache.get(BB))
+ PN->addIncoming(I, Pred);
return PN;
}
return V;
}
public:
- LoopPromoter(Value *SP, const SmallVectorImpl<Instruction *> &Insts,
- SSAUpdater &S, SmallPtrSet<Value *, 4> &PMA,
+ LoopPromoter(Value *SP,
+ ArrayRef<const Instruction *> Insts,
+ SSAUpdater &S, SmallPtrSetImpl<Value *> &PMA,
SmallVectorImpl<BasicBlock *> &LEB,
SmallVectorImpl<Instruction *> &LIP, PredIteratorCache &PIC,
AliasSetTracker &ast, LoopInfo &li, DebugLoc dl, int alignment,
- MDNode *TBAATag)
+ const AAMDNodes &AATags)
: LoadAndStorePromoter(Insts, S), SomePtr(SP), PointerMustAliases(PMA),
LoopExitBlocks(LEB), LoopInsertPts(LIP), PredCache(PIC), AST(ast),
- LI(li), DL(dl), Alignment(alignment), TBAATag(TBAATag) {}
+ LI(li), DL(dl), Alignment(alignment), AATags(AATags) {}
- virtual bool isInstInList(Instruction *I,
- const SmallVectorImpl<Instruction*> &) const {
+ bool isInstInList(Instruction *I,
+ const SmallVectorImpl<Instruction*> &) const override {
Value *Ptr;
if (LoadInst *LI = dyn_cast<LoadInst>(I))
Ptr = LI->getOperand(0);
return PointerMustAliases.count(Ptr);
}
- virtual void doExtraRewritesBeforeFinalDeletion() const {
+ void doExtraRewritesBeforeFinalDeletion() const override {
// Insert stores after in the loop exit blocks. Each exit block gets a
// store of the live-out values that feed them. Since we've already told
// the SSA updater about the defs in the loop and the preheader
StoreInst *NewSI = new StoreInst(LiveInValue, Ptr, InsertPos);
NewSI->setAlignment(Alignment);
NewSI->setDebugLoc(DL);
- if (TBAATag) NewSI->setMetadata(LLVMContext::MD_tbaa, TBAATag);
+ if (AATags) NewSI->setAAMetadata(AATags);
}
}
- virtual void replaceLoadWithValue(LoadInst *LI, Value *V) const {
+ void replaceLoadWithValue(LoadInst *LI, Value *V) const override {
// Update alias analysis.
AST.copyValue(LI, V);
}
- virtual void instructionDeleted(Instruction *I) const {
+ void instructionDeleted(Instruction *I) const override {
AST.deleteValue(I);
}
};
} // end anon namespace
-/// PromoteAliasSet - Try to promote memory values to scalars by sinking
-/// stores out of the loop and moving loads to before the loop. We do this by
-/// looping over the stores in the loop, looking for stores to Must pointers
-/// which are loop invariant.
+/// Try to promote memory values to scalars by sinking stores out of the
+/// loop and moving loads to before the loop. We do this by looping over
+/// the stores in the loop, looking for stores to Must pointers which are
+/// loop invariant.
///
-void LICM::PromoteAliasSet(AliasSet &AS,
- SmallVectorImpl<BasicBlock*> &ExitBlocks,
- SmallVectorImpl<Instruction*> &InsertPts,
- PredIteratorCache &PIC) {
+bool llvm::promoteLoopAccessesToScalars(AliasSet &AS,
+ SmallVectorImpl<BasicBlock*>&ExitBlocks,
+ SmallVectorImpl<Instruction*>&InsertPts,
+ PredIteratorCache &PIC, LoopInfo *LI,
+ DominatorTree *DT, Loop *CurLoop,
+ AliasSetTracker *CurAST,
+ LICMSafetyInfo * SafetyInfo) {
+ // Verify inputs.
+ assert(LI != nullptr && DT != nullptr &&
+ CurLoop != nullptr && CurAST != nullptr &&
+ SafetyInfo != nullptr &&
+ "Unexpected Input to promoteLoopAccessesToScalars");
+ // Initially set Changed status to false.
+ bool Changed = false;
// We can promote this alias set if it has a store, if it is a "Must" alias
// set, if the pointer is loop invariant, and if we are not eliminating any
// volatile loads or stores.
if (AS.isForwardingAliasSet() || !AS.isMod() || !AS.isMustAlias() ||
AS.isVolatile() || !CurLoop->isLoopInvariant(AS.begin()->getValue()))
- return;
+ return Changed;
assert(!AS.empty() &&
"Must alias set should have at least one pointer element in it!");
+
Value *SomePtr = AS.begin()->getValue();
+ BasicBlock * Preheader = CurLoop->getLoopPreheader();
// It isn't safe to promote a load/store from the loop if the load/store is
// conditional. For example, turning:
// We start with an alignment of one and try to find instructions that allow
// us to prove better alignment.
unsigned Alignment = 1;
- MDNode *TBAATag = 0;
+ AAMDNodes AATags;
+ bool HasDedicatedExits = CurLoop->hasDedicatedExits();
// Check that all of the pointers in the alias set have the same type. We
// cannot (yet) promote a memory location that is loaded and stored in
- // different sizes. While we are at it, collect alignment and TBAA info.
+ // different sizes. While we are at it, collect alignment and AA info.
for (AliasSet::iterator ASI = AS.begin(), E = AS.end(); ASI != E; ++ASI) {
Value *ASIV = ASI->getValue();
PointerMustAliases.insert(ASIV);
// cannot (yet) promote a memory location that is loaded and stored in
// different sizes.
if (SomePtr->getType() != ASIV->getType())
- return;
+ return Changed;
- for (Value::use_iterator UI = ASIV->use_begin(), UE = ASIV->use_end();
- UI != UE; ++UI) {
+ for (User *U : ASIV->users()) {
// Ignore instructions that are outside the loop.
- Instruction *Use = dyn_cast<Instruction>(*UI);
- if (!Use || !CurLoop->contains(Use))
+ Instruction *UI = dyn_cast<Instruction>(U);
+ if (!UI || !CurLoop->contains(UI))
continue;
// If there is an non-load/store instruction in the loop, we can't promote
// it.
- if (LoadInst *load = dyn_cast<LoadInst>(Use)) {
+ if (const LoadInst *load = dyn_cast<LoadInst>(UI)) {
assert(!load->isVolatile() && "AST broken");
if (!load->isSimple())
- return;
- } else if (StoreInst *store = dyn_cast<StoreInst>(Use)) {
+ return Changed;
+ } else if (const StoreInst *store = dyn_cast<StoreInst>(UI)) {
// Stores *of* the pointer are not interesting, only stores *to* the
// pointer.
- if (Use->getOperand(1) != ASIV)
+ if (UI->getOperand(1) != ASIV)
continue;
assert(!store->isVolatile() && "AST broken");
if (!store->isSimple())
- return;
+ return Changed;
+ // Don't sink stores from loops without dedicated block exits. Exits
+ // containing indirect branches are not transformed by loop simplify,
+ // make sure we catch that. An additional load may be generated in the
+ // preheader for SSA updater, so also avoid sinking when no preheader
+ // is available.
+ if (!HasDedicatedExits || !Preheader)
+ return Changed;
// Note that we only check GuaranteedToExecute inside the store case
// so that we do not introduce stores where they did not exist before
// Larger is better, with the exception of 0 being the best alignment.
unsigned InstAlignment = store->getAlignment();
if ((InstAlignment > Alignment || InstAlignment == 0) && Alignment != 0)
- if (isGuaranteedToExecute(*Use)) {
+ if (isGuaranteedToExecute(*UI, DT, CurLoop, SafetyInfo)) {
GuaranteedToExecute = true;
Alignment = InstAlignment;
}
if (!GuaranteedToExecute)
- GuaranteedToExecute = isGuaranteedToExecute(*Use);
+ GuaranteedToExecute = isGuaranteedToExecute(*UI, DT,
+ CurLoop, SafetyInfo);
} else
- return; // Not a load or store.
+ return Changed; // Not a load or store.
- // Merge the TBAA tags.
+ // Merge the AA tags.
if (LoopUses.empty()) {
- // On the first load/store, just take its TBAA tag.
- TBAATag = Use->getMetadata(LLVMContext::MD_tbaa);
- } else if (TBAATag) {
- TBAATag = MDNode::getMostGenericTBAA(TBAATag,
- Use->getMetadata(LLVMContext::MD_tbaa));
+ // On the first load/store, just take its AA tags.
+ UI->getAAMetadata(AATags);
+ } else if (AATags) {
+ UI->getAAMetadata(AATags, /* Merge = */ true);
}
-
- LoopUses.push_back(Use);
+
+ LoopUses.push_back(UI);
}
}
// If there isn't a guaranteed-to-execute instruction, we can't promote.
if (!GuaranteedToExecute)
- return;
+ return Changed;
// Otherwise, this is safe to promote, lets do it!
DEBUG(dbgs() << "LICM: Promoting value stored to in loop: " <<*SomePtr<<'\n');
// We use the SSAUpdater interface to insert phi nodes as required.
SmallVector<PHINode*, 16> NewPHIs;
SSAUpdater SSA(&NewPHIs);
- LoopPromoter Promoter(SomePtr, LoopUses, SSA, PointerMustAliases, ExitBlocks,
- InsertPts, PIC, *CurAST, *LI, DL, Alignment, TBAATag);
+ LoopPromoter Promoter(SomePtr, LoopUses, SSA,
+ PointerMustAliases, ExitBlocks,
+ InsertPts, PIC, *CurAST, *LI, DL, Alignment, AATags);
// Set up the preheader to have a definition of the value. It is the live-out
// value from the preheader that uses in the loop will use.
Preheader->getTerminator());
PreheaderLoad->setAlignment(Alignment);
PreheaderLoad->setDebugLoc(DL);
- if (TBAATag) PreheaderLoad->setMetadata(LLVMContext::MD_tbaa, TBAATag);
+ if (AATags) PreheaderLoad->setAAMetadata(AATags);
SSA.AddAvailableValue(Preheader, PreheaderLoad);
// Rewrite all the loads in the loop and remember all the definitions from
// If the SSAUpdater didn't use the load in the preheader, just zap it now.
if (PreheaderLoad->use_empty())
PreheaderLoad->eraseFromParent();
-}
+ return Changed;
+}
-/// cloneBasicBlockAnalysis - Simple Analysis hook. Clone alias set info.
+/// Simple Analysis hook. Clone alias set info.
+///
void LICM::cloneBasicBlockAnalysis(BasicBlock *From, BasicBlock *To, Loop *L) {
AliasSetTracker *AST = LoopToAliasSetMap.lookup(L);
if (!AST)
AST->copyValue(From, To);
}
-/// deleteAnalysisValue - Simple Analysis hook. Delete value V from alias
-/// set.
+/// Simple Analysis hook. Delete value V from alias set
+///
void LICM::deleteAnalysisValue(Value *V, Loop *L) {
AliasSetTracker *AST = LoopToAliasSetMap.lookup(L);
if (!AST)
AST->deleteValue(V);
}
+
+/// Simple Analysis hook. Delete value L from alias set map.
+///
+void LICM::deleteAnalysisLoop(Loop *L) {
+ AliasSetTracker *AST = LoopToAliasSetMap.lookup(L);
+ if (!AST)
+ return;
+
+ delete AST;
+ LoopToAliasSetMap.erase(L);
+}
+
+
+/// Return true if the body of this loop may store into the memory
+/// location pointed to by V.
+///
+static bool pointerInvalidatedByLoop(Value *V, uint64_t Size,
+ const AAMDNodes &AAInfo,
+ AliasSetTracker *CurAST) {
+ // Check to see if any of the basic blocks in CurLoop invalidate *V.
+ return CurAST->getAliasSetForPointer(V, Size, AAInfo).isMod();
+}
+
+/// Little predicate that returns true if the specified basic block is in
+/// a subloop of the current one, not the current one itself.
+///
+static bool inSubLoop(BasicBlock *BB, Loop *CurLoop, LoopInfo *LI) {
+ assert(CurLoop->contains(BB) && "Only valid if BB is IN the loop");
+ return LI->getLoopFor(BB) != CurLoop;
+}
+