diff options
Diffstat (limited to 'contrib/llvm/lib/Transforms/Scalar/LoopUnrollPass.cpp')
-rw-r--r-- | contrib/llvm/lib/Transforms/Scalar/LoopUnrollPass.cpp | 516 |
1 files changed, 454 insertions, 62 deletions
diff --git a/contrib/llvm/lib/Transforms/Scalar/LoopUnrollPass.cpp b/contrib/llvm/lib/Transforms/Scalar/LoopUnrollPass.cpp index fef5210..4ccbfc9 100644 --- a/contrib/llvm/lib/Transforms/Scalar/LoopUnrollPass.cpp +++ b/contrib/llvm/lib/Transforms/Scalar/LoopUnrollPass.cpp @@ -13,15 +13,18 @@ //===----------------------------------------------------------------------===// #include "llvm/Transforms/Scalar.h" +#include "llvm/ADT/SetVector.h" #include "llvm/Analysis/AssumptionCache.h" #include "llvm/Analysis/CodeMetrics.h" -#include "llvm/Analysis/FunctionTargetTransformInfo.h" +#include "llvm/Analysis/InstructionSimplify.h" #include "llvm/Analysis/LoopPass.h" #include "llvm/Analysis/ScalarEvolution.h" +#include "llvm/Analysis/ScalarEvolutionExpressions.h" #include "llvm/Analysis/TargetTransformInfo.h" #include "llvm/IR/DataLayout.h" #include "llvm/IR/DiagnosticInfo.h" #include "llvm/IR/Dominators.h" +#include "llvm/IR/InstVisitor.h" #include "llvm/IR/IntrinsicInst.h" #include "llvm/IR/Metadata.h" #include "llvm/Support/CommandLine.h" @@ -35,8 +38,24 @@ using namespace llvm; #define DEBUG_TYPE "loop-unroll" static cl::opt<unsigned> -UnrollThreshold("unroll-threshold", cl::init(150), cl::Hidden, - cl::desc("The cut-off point for automatic loop unrolling")); + UnrollThreshold("unroll-threshold", cl::init(150), cl::Hidden, + cl::desc("The baseline cost threshold for loop unrolling")); + +static cl::opt<unsigned> UnrollPercentDynamicCostSavedThreshold( + "unroll-percent-dynamic-cost-saved-threshold", cl::init(20), cl::Hidden, + cl::desc("The percentage of estimated dynamic cost which must be saved by " + "unrolling to allow unrolling up to the max threshold.")); + +static cl::opt<unsigned> UnrollDynamicCostSavingsDiscount( + "unroll-dynamic-cost-savings-discount", cl::init(2000), cl::Hidden, + cl::desc("This is the amount discounted from the total unroll cost when " + "the unrolled form has a high dynamic cost savings (triggered by " + "the '-unroll-perecent-dynamic-cost-saved-threshold' flag).")); + +static cl::opt<unsigned> UnrollMaxIterationsCountToAnalyze( + "unroll-max-iteration-count-to-analyze", cl::init(0), cl::Hidden, + cl::desc("Don't allow loop unrolling to simulate more than this number of" + "iterations when checking full unroll profitability")); static cl::opt<unsigned> UnrollCount("unroll-count", cl::init(0), cl::Hidden, @@ -63,11 +82,18 @@ namespace { static char ID; // Pass ID, replacement for typeid LoopUnroll(int T = -1, int C = -1, int P = -1, int R = -1) : LoopPass(ID) { CurrentThreshold = (T == -1) ? UnrollThreshold : unsigned(T); + CurrentPercentDynamicCostSavedThreshold = + UnrollPercentDynamicCostSavedThreshold; + CurrentDynamicCostSavingsDiscount = UnrollDynamicCostSavingsDiscount; CurrentCount = (C == -1) ? UnrollCount : unsigned(C); CurrentAllowPartial = (P == -1) ? UnrollAllowPartial : (bool)P; CurrentRuntime = (R == -1) ? UnrollRuntime : (bool)R; UserThreshold = (T != -1) || (UnrollThreshold.getNumOccurrences() > 0); + UserPercentDynamicCostSavedThreshold = + (UnrollPercentDynamicCostSavedThreshold.getNumOccurrences() > 0); + UserDynamicCostSavingsDiscount = + (UnrollDynamicCostSavingsDiscount.getNumOccurrences() > 0); UserAllowPartial = (P != -1) || (UnrollAllowPartial.getNumOccurrences() > 0); UserRuntime = (R != -1) || (UnrollRuntime.getNumOccurrences() > 0); @@ -91,12 +117,18 @@ namespace { unsigned CurrentCount; unsigned CurrentThreshold; - bool CurrentAllowPartial; - bool CurrentRuntime; - bool UserCount; // CurrentCount is user-specified. - bool UserThreshold; // CurrentThreshold is user-specified. - bool UserAllowPartial; // CurrentAllowPartial is user-specified. - bool UserRuntime; // CurrentRuntime is user-specified. + unsigned CurrentPercentDynamicCostSavedThreshold; + unsigned CurrentDynamicCostSavingsDiscount; + bool CurrentAllowPartial; + bool CurrentRuntime; + + // Flags for whether the 'current' settings are user-specified. + bool UserCount; + bool UserThreshold; + bool UserPercentDynamicCostSavedThreshold; + bool UserDynamicCostSavingsDiscount; + bool UserAllowPartial; + bool UserRuntime; bool runOnLoop(Loop *L, LPPassManager &LPM) override; @@ -105,16 +137,15 @@ namespace { /// void getAnalysisUsage(AnalysisUsage &AU) const override { AU.addRequired<AssumptionCacheTracker>(); - AU.addRequired<LoopInfo>(); - AU.addPreserved<LoopInfo>(); + AU.addRequired<LoopInfoWrapperPass>(); + AU.addPreserved<LoopInfoWrapperPass>(); AU.addRequiredID(LoopSimplifyID); AU.addPreservedID(LoopSimplifyID); AU.addRequiredID(LCSSAID); AU.addPreservedID(LCSSAID); AU.addRequired<ScalarEvolution>(); AU.addPreserved<ScalarEvolution>(); - AU.addRequired<TargetTransformInfo>(); - AU.addRequired<FunctionTargetTransformInfo>(); + AU.addRequired<TargetTransformInfoWrapperPass>(); // FIXME: Loop unroll requires LCSSA. And LCSSA requires dom info. // If loop unroll does not preserve dom info then LCSSA pass on next // loop will receive invalid dom info. @@ -124,9 +155,12 @@ namespace { // Fill in the UnrollingPreferences parameter with values from the // TargetTransformationInfo. - void getUnrollingPreferences(Loop *L, const FunctionTargetTransformInfo &FTTI, + void getUnrollingPreferences(Loop *L, const TargetTransformInfo &TTI, TargetTransformInfo::UnrollingPreferences &UP) { UP.Threshold = CurrentThreshold; + UP.PercentDynamicCostSavedThreshold = + CurrentPercentDynamicCostSavedThreshold; + UP.DynamicCostSavingsDiscount = CurrentDynamicCostSavingsDiscount; UP.OptSizeThreshold = OptSizeUnrollThreshold; UP.PartialThreshold = CurrentThreshold; UP.PartialOptSizeThreshold = OptSizeUnrollThreshold; @@ -134,7 +168,8 @@ namespace { UP.MaxCount = UINT_MAX; UP.Partial = CurrentAllowPartial; UP.Runtime = CurrentRuntime; - FTTI.getUnrollingPreferences(L, UP); + UP.AllowExpensiveTripCount = false; + TTI.getUnrollingPreferences(L, UP); } // Select and return an unroll count based on parameters from @@ -153,7 +188,9 @@ namespace { // unrolled loops respectively. void selectThresholds(const Loop *L, bool HasPragma, const TargetTransformInfo::UnrollingPreferences &UP, - unsigned &Threshold, unsigned &PartialThreshold) { + unsigned &Threshold, unsigned &PartialThreshold, + unsigned &PercentDynamicCostSavedThreshold, + unsigned &DynamicCostSavingsDiscount) { // Determine the current unrolling threshold. While this is // normally set from UnrollThreshold, it is overridden to a // smaller value if the current function is marked as @@ -161,10 +198,17 @@ namespace { // specified. Threshold = UserThreshold ? CurrentThreshold : UP.Threshold; PartialThreshold = UserThreshold ? CurrentThreshold : UP.PartialThreshold; + PercentDynamicCostSavedThreshold = + UserPercentDynamicCostSavedThreshold + ? CurrentPercentDynamicCostSavedThreshold + : UP.PercentDynamicCostSavedThreshold; + DynamicCostSavingsDiscount = UserDynamicCostSavingsDiscount + ? CurrentDynamicCostSavingsDiscount + : UP.DynamicCostSavingsDiscount; + if (!UserThreshold && - L->getHeader()->getParent()->getAttributes(). - hasAttribute(AttributeSet::FunctionIndex, - Attribute::OptimizeForSize)) { + L->getHeader()->getParent()->hasFnAttribute( + Attribute::OptimizeForSize)) { Threshold = UP.OptSizeThreshold; PartialThreshold = UP.PartialOptSizeThreshold; } @@ -180,15 +224,18 @@ namespace { std::max<unsigned>(PartialThreshold, PragmaUnrollThreshold); } } + bool canUnrollCompletely(Loop *L, unsigned Threshold, + unsigned PercentDynamicCostSavedThreshold, + unsigned DynamicCostSavingsDiscount, + uint64_t UnrolledCost, uint64_t RolledDynamicCost); }; } char LoopUnroll::ID = 0; INITIALIZE_PASS_BEGIN(LoopUnroll, "loop-unroll", "Unroll loops", false, false) -INITIALIZE_AG_DEPENDENCY(TargetTransformInfo) +INITIALIZE_PASS_DEPENDENCY(TargetTransformInfoWrapperPass) INITIALIZE_PASS_DEPENDENCY(AssumptionCacheTracker) -INITIALIZE_PASS_DEPENDENCY(FunctionTargetTransformInfo) -INITIALIZE_PASS_DEPENDENCY(LoopInfo) +INITIALIZE_PASS_DEPENDENCY(LoopInfoWrapperPass) INITIALIZE_PASS_DEPENDENCY(LoopSimplify) INITIALIZE_PASS_DEPENDENCY(LCSSA) INITIALIZE_PASS_DEPENDENCY(ScalarEvolution) @@ -203,6 +250,295 @@ Pass *llvm::createSimpleLoopUnrollPass() { return llvm::createLoopUnrollPass(-1, -1, 0, 0); } +namespace { +// This class is used to get an estimate of the optimization effects that we +// could get from complete loop unrolling. It comes from the fact that some +// loads might be replaced with concrete constant values and that could trigger +// a chain of instruction simplifications. +// +// E.g. we might have: +// int a[] = {0, 1, 0}; +// v = 0; +// for (i = 0; i < 3; i ++) +// v += b[i]*a[i]; +// If we completely unroll the loop, we would get: +// v = b[0]*a[0] + b[1]*a[1] + b[2]*a[2] +// Which then will be simplified to: +// v = b[0]* 0 + b[1]* 1 + b[2]* 0 +// And finally: +// v = b[1] +class UnrolledInstAnalyzer : private InstVisitor<UnrolledInstAnalyzer, bool> { + typedef InstVisitor<UnrolledInstAnalyzer, bool> Base; + friend class InstVisitor<UnrolledInstAnalyzer, bool>; + struct SimplifiedAddress { + Value *Base = nullptr; + ConstantInt *Offset = nullptr; + }; + +public: + UnrolledInstAnalyzer(unsigned Iteration, + DenseMap<Value *, Constant *> &SimplifiedValues, + const Loop *L, ScalarEvolution &SE) + : Iteration(Iteration), SimplifiedValues(SimplifiedValues), L(L), SE(SE) { + IterationNumber = SE.getConstant(APInt(64, Iteration)); + } + + // Allow access to the initial visit method. + using Base::visit; + +private: + /// \brief A cache of pointer bases and constant-folded offsets corresponding + /// to GEP (or derived from GEP) instructions. + /// + /// In order to find the base pointer one needs to perform non-trivial + /// traversal of the corresponding SCEV expression, so it's good to have the + /// results saved. + DenseMap<Value *, SimplifiedAddress> SimplifiedAddresses; + + /// \brief Number of currently simulated iteration. + /// + /// If an expression is ConstAddress+Constant, then the Constant is + /// Start + Iteration*Step, where Start and Step could be obtained from + /// SCEVGEPCache. + unsigned Iteration; + + /// \brief SCEV expression corresponding to number of currently simulated + /// iteration. + const SCEV *IterationNumber; + + /// \brief A Value->Constant map for keeping values that we managed to + /// constant-fold on the given iteration. + /// + /// While we walk the loop instructions, we build up and maintain a mapping + /// of simplified values specific to this iteration. The idea is to propagate + /// any special information we have about loads that can be replaced with + /// constants after complete unrolling, and account for likely simplifications + /// post-unrolling. + DenseMap<Value *, Constant *> &SimplifiedValues; + + const Loop *L; + ScalarEvolution &SE; + + /// \brief Try to simplify instruction \param I using its SCEV expression. + /// + /// The idea is that some AddRec expressions become constants, which then + /// could trigger folding of other instructions. However, that only happens + /// for expressions whose start value is also constant, which isn't always the + /// case. In another common and important case the start value is just some + /// address (i.e. SCEVUnknown) - in this case we compute the offset and save + /// it along with the base address instead. + bool simplifyInstWithSCEV(Instruction *I) { + if (!SE.isSCEVable(I->getType())) + return false; + + const SCEV *S = SE.getSCEV(I); + if (auto *SC = dyn_cast<SCEVConstant>(S)) { + SimplifiedValues[I] = SC->getValue(); + return true; + } + + auto *AR = dyn_cast<SCEVAddRecExpr>(S); + if (!AR) + return false; + + const SCEV *ValueAtIteration = AR->evaluateAtIteration(IterationNumber, SE); + // Check if the AddRec expression becomes a constant. + if (auto *SC = dyn_cast<SCEVConstant>(ValueAtIteration)) { + SimplifiedValues[I] = SC->getValue(); + return true; + } + + // Check if the offset from the base address becomes a constant. + auto *Base = dyn_cast<SCEVUnknown>(SE.getPointerBase(S)); + if (!Base) + return false; + auto *Offset = + dyn_cast<SCEVConstant>(SE.getMinusSCEV(ValueAtIteration, Base)); + if (!Offset) + return false; + SimplifiedAddress Address; + Address.Base = Base->getValue(); + Address.Offset = Offset->getValue(); + SimplifiedAddresses[I] = Address; + return true; + } + + /// Base case for the instruction visitor. + bool visitInstruction(Instruction &I) { + return simplifyInstWithSCEV(&I); + } + + /// TODO: Add visitors for other instruction types, e.g. ZExt, SExt. + + /// Try to simplify binary operator I. + /// + /// TODO: Probaly it's worth to hoist the code for estimating the + /// simplifications effects to a separate class, since we have a very similar + /// code in InlineCost already. + bool visitBinaryOperator(BinaryOperator &I) { + Value *LHS = I.getOperand(0), *RHS = I.getOperand(1); + if (!isa<Constant>(LHS)) + if (Constant *SimpleLHS = SimplifiedValues.lookup(LHS)) + LHS = SimpleLHS; + if (!isa<Constant>(RHS)) + if (Constant *SimpleRHS = SimplifiedValues.lookup(RHS)) + RHS = SimpleRHS; + + Value *SimpleV = nullptr; + const DataLayout &DL = I.getModule()->getDataLayout(); + if (auto FI = dyn_cast<FPMathOperator>(&I)) + SimpleV = + SimplifyFPBinOp(I.getOpcode(), LHS, RHS, FI->getFastMathFlags(), DL); + else + SimpleV = SimplifyBinOp(I.getOpcode(), LHS, RHS, DL); + + if (Constant *C = dyn_cast_or_null<Constant>(SimpleV)) + SimplifiedValues[&I] = C; + + if (SimpleV) + return true; + return Base::visitBinaryOperator(I); + } + + /// Try to fold load I. + bool visitLoad(LoadInst &I) { + Value *AddrOp = I.getPointerOperand(); + + auto AddressIt = SimplifiedAddresses.find(AddrOp); + if (AddressIt == SimplifiedAddresses.end()) + return false; + ConstantInt *SimplifiedAddrOp = AddressIt->second.Offset; + + auto *GV = dyn_cast<GlobalVariable>(AddressIt->second.Base); + // We're only interested in loads that can be completely folded to a + // constant. + if (!GV || !GV->hasInitializer()) + return false; + + ConstantDataSequential *CDS = + dyn_cast<ConstantDataSequential>(GV->getInitializer()); + if (!CDS) + return false; + + int ElemSize = CDS->getElementType()->getPrimitiveSizeInBits() / 8U; + assert(SimplifiedAddrOp->getValue().getActiveBits() < 64 && + "Unexpectedly large index value."); + int64_t Index = SimplifiedAddrOp->getSExtValue() / ElemSize; + if (Index >= CDS->getNumElements()) { + // FIXME: For now we conservatively ignore out of bound accesses, but + // we're allowed to perform the optimization in this case. + return false; + } + + Constant *CV = CDS->getElementAsConstant(Index); + assert(CV && "Constant expected."); + SimplifiedValues[&I] = CV; + + return true; + } +}; +} // namespace + + +namespace { +struct EstimatedUnrollCost { + /// \brief The estimated cost after unrolling. + unsigned UnrolledCost; + + /// \brief The estimated dynamic cost of executing the instructions in the + /// rolled form. + unsigned RolledDynamicCost; +}; +} + +/// \brief Figure out if the loop is worth full unrolling. +/// +/// Complete loop unrolling can make some loads constant, and we need to know +/// if that would expose any further optimization opportunities. This routine +/// estimates this optimization. It assigns computed number of instructions, +/// that potentially might be optimized away, to +/// NumberOfOptimizedInstructions, and total number of instructions to +/// UnrolledLoopSize (not counting blocks that won't be reached, if we were +/// able to compute the condition). +/// \returns false if we can't analyze the loop, or if we discovered that +/// unrolling won't give anything. Otherwise, returns true. +Optional<EstimatedUnrollCost> +analyzeLoopUnrollCost(const Loop *L, unsigned TripCount, ScalarEvolution &SE, + const TargetTransformInfo &TTI, + unsigned MaxUnrolledLoopSize) { + // We want to be able to scale offsets by the trip count and add more offsets + // to them without checking for overflows, and we already don't want to + // analyze *massive* trip counts, so we force the max to be reasonably small. + assert(UnrollMaxIterationsCountToAnalyze < (INT_MAX / 2) && + "The unroll iterations max is too large!"); + + // Don't simulate loops with a big or unknown tripcount + if (!UnrollMaxIterationsCountToAnalyze || !TripCount || + TripCount > UnrollMaxIterationsCountToAnalyze) + return None; + + SmallSetVector<BasicBlock *, 16> BBWorklist; + DenseMap<Value *, Constant *> SimplifiedValues; + + // The estimated cost of the unrolled form of the loop. We try to estimate + // this by simplifying as much as we can while computing the estimate. + unsigned UnrolledCost = 0; + // We also track the estimated dynamic (that is, actually executed) cost in + // the rolled form. This helps identify cases when the savings from unrolling + // aren't just exposing dead control flows, but actual reduced dynamic + // instructions due to the simplifications which we expect to occur after + // unrolling. + unsigned RolledDynamicCost = 0; + + // Simulate execution of each iteration of the loop counting instructions, + // which would be simplified. + // Since the same load will take different values on different iterations, + // we literally have to go through all loop's iterations. + for (unsigned Iteration = 0; Iteration < TripCount; ++Iteration) { + SimplifiedValues.clear(); + UnrolledInstAnalyzer Analyzer(Iteration, SimplifiedValues, L, SE); + + BBWorklist.clear(); + BBWorklist.insert(L->getHeader()); + // Note that we *must not* cache the size, this loop grows the worklist. + for (unsigned Idx = 0; Idx != BBWorklist.size(); ++Idx) { + BasicBlock *BB = BBWorklist[Idx]; + + // Visit all instructions in the given basic block and try to simplify + // it. We don't change the actual IR, just count optimization + // opportunities. + for (Instruction &I : *BB) { + unsigned InstCost = TTI.getUserCost(&I); + + // Visit the instruction to analyze its loop cost after unrolling, + // and if the visitor returns false, include this instruction in the + // unrolled cost. + if (!Analyzer.visit(I)) + UnrolledCost += InstCost; + + // Also track this instructions expected cost when executing the rolled + // loop form. + RolledDynamicCost += InstCost; + + // If unrolled body turns out to be too big, bail out. + if (UnrolledCost > MaxUnrolledLoopSize) + return None; + } + + // Add BB's successors to the worklist. + for (BasicBlock *Succ : successors(BB)) + if (L->contains(Succ)) + BBWorklist.insert(Succ); + } + + // If we found no optimization opportunities on the first iteration, we + // won't find them on later ones too. + if (UnrolledCost == RolledDynamicCost) + return None; + } + return {{UnrolledCost, RolledDynamicCost}}; +} + /// ApproximateLoopSize - Approximate the size of the loop. static unsigned ApproximateLoopSize(const Loop *L, unsigned &NumCalls, bool &NotDuplicatable, @@ -234,44 +570,31 @@ static unsigned ApproximateLoopSize(const Loop *L, unsigned &NumCalls, // Returns the loop hint metadata node with the given name (for example, // "llvm.loop.unroll.count"). If no such metadata node exists, then nullptr is // returned. -static const MDNode *GetUnrollMetadata(const Loop *L, StringRef Name) { - MDNode *LoopID = L->getLoopID(); - if (!LoopID) - return nullptr; - - // First operand should refer to the loop id itself. - assert(LoopID->getNumOperands() > 0 && "requires at least one operand"); - assert(LoopID->getOperand(0) == LoopID && "invalid loop id"); - - for (unsigned i = 1, e = LoopID->getNumOperands(); i < e; ++i) { - const MDNode *MD = dyn_cast<MDNode>(LoopID->getOperand(i)); - if (!MD) - continue; - - const MDString *S = dyn_cast<MDString>(MD->getOperand(0)); - if (!S) - continue; - - if (Name.equals(S->getString())) - return MD; - } +static MDNode *GetUnrollMetadataForLoop(const Loop *L, StringRef Name) { + if (MDNode *LoopID = L->getLoopID()) + return GetUnrollMetadata(LoopID, Name); return nullptr; } // Returns true if the loop has an unroll(full) pragma. static bool HasUnrollFullPragma(const Loop *L) { - return GetUnrollMetadata(L, "llvm.loop.unroll.full"); + return GetUnrollMetadataForLoop(L, "llvm.loop.unroll.full"); } // Returns true if the loop has an unroll(disable) pragma. static bool HasUnrollDisablePragma(const Loop *L) { - return GetUnrollMetadata(L, "llvm.loop.unroll.disable"); + return GetUnrollMetadataForLoop(L, "llvm.loop.unroll.disable"); +} + +// Returns true if the loop has an runtime unroll(disable) pragma. +static bool HasRuntimeUnrollDisablePragma(const Loop *L) { + return GetUnrollMetadataForLoop(L, "llvm.loop.unroll.runtime.disable"); } // If loop has an unroll_count pragma return the (necessarily // positive) value from the pragma. Otherwise return 0. static unsigned UnrollCountPragmaValue(const Loop *L) { - const MDNode *MD = GetUnrollMetadata(L, "llvm.loop.unroll.count"); + MDNode *MD = GetUnrollMetadataForLoop(L, "llvm.loop.unroll.count"); if (MD) { assert(MD->getNumOperands() == 2 && "Unroll count hint metadata should have two operands."); @@ -319,6 +642,59 @@ static void SetLoopAlreadyUnrolled(Loop *L) { L->setLoopID(NewLoopID); } +bool LoopUnroll::canUnrollCompletely(Loop *L, unsigned Threshold, + unsigned PercentDynamicCostSavedThreshold, + unsigned DynamicCostSavingsDiscount, + uint64_t UnrolledCost, + uint64_t RolledDynamicCost) { + + if (Threshold == NoThreshold) { + DEBUG(dbgs() << " Can fully unroll, because no threshold is set.\n"); + return true; + } + + if (UnrolledCost <= Threshold) { + DEBUG(dbgs() << " Can fully unroll, because unrolled cost: " + << UnrolledCost << "<" << Threshold << "\n"); + return true; + } + + assert(UnrolledCost && "UnrolledCost can't be 0 at this point."); + assert(RolledDynamicCost >= UnrolledCost && + "Cannot have a higher unrolled cost than a rolled cost!"); + + // Compute the percentage of the dynamic cost in the rolled form that is + // saved when unrolled. If unrolling dramatically reduces the estimated + // dynamic cost of the loop, we use a higher threshold to allow more + // unrolling. + unsigned PercentDynamicCostSaved = + (uint64_t)(RolledDynamicCost - UnrolledCost) * 100ull / RolledDynamicCost; + + if (PercentDynamicCostSaved >= PercentDynamicCostSavedThreshold && + (int64_t)UnrolledCost - (int64_t)DynamicCostSavingsDiscount <= + (int64_t)Threshold) { + DEBUG(dbgs() << " Can fully unroll, because unrolling will reduce the " + "expected dynamic cost by " << PercentDynamicCostSaved + << "% (threshold: " << PercentDynamicCostSavedThreshold + << "%)\n" + << " and the unrolled cost (" << UnrolledCost + << ") is less than the max threshold (" + << DynamicCostSavingsDiscount << ").\n"); + return true; + } + + DEBUG(dbgs() << " Too large to fully unroll:\n"); + DEBUG(dbgs() << " Threshold: " << Threshold << "\n"); + DEBUG(dbgs() << " Max threshold: " << DynamicCostSavingsDiscount << "\n"); + DEBUG(dbgs() << " Percent cost saved threshold: " + << PercentDynamicCostSavedThreshold << "%\n"); + DEBUG(dbgs() << " Unrolled cost: " << UnrolledCost << "\n"); + DEBUG(dbgs() << " Rolled dynamic cost: " << RolledDynamicCost << "\n"); + DEBUG(dbgs() << " Percent cost saved: " << PercentDynamicCostSaved + << "\n"); + return false; +} + unsigned LoopUnroll::selectUnrollCount( const Loop *L, unsigned TripCount, bool PragmaFullUnroll, unsigned PragmaCount, const TargetTransformInfo::UnrollingPreferences &UP, @@ -363,13 +739,13 @@ bool LoopUnroll::runOnLoop(Loop *L, LPPassManager &LPM) { if (skipOptnoneFunction(L)) return false; - LoopInfo *LI = &getAnalysis<LoopInfo>(); + Function &F = *L->getHeader()->getParent(); + + LoopInfo *LI = &getAnalysis<LoopInfoWrapperPass>().getLoopInfo(); ScalarEvolution *SE = &getAnalysis<ScalarEvolution>(); - const TargetTransformInfo &TTI = getAnalysis<TargetTransformInfo>(); - const FunctionTargetTransformInfo &FTTI = - getAnalysis<FunctionTargetTransformInfo>(); - auto &AC = getAnalysis<AssumptionCacheTracker>().getAssumptionCache( - *L->getHeader()->getParent()); + const TargetTransformInfo &TTI = + getAnalysis<TargetTransformInfoWrapperPass>().getTTI(F); + auto &AC = getAnalysis<AssumptionCacheTracker>().getAssumptionCache(F); BasicBlock *Header = L->getHeader(); DEBUG(dbgs() << "Loop Unroll: F[" << Header->getParent()->getName() @@ -383,7 +759,7 @@ bool LoopUnroll::runOnLoop(Loop *L, LPPassManager &LPM) { bool HasPragma = PragmaFullUnroll || PragmaCount > 0; TargetTransformInfo::UnrollingPreferences UP; - getUnrollingPreferences(L, FTTI, UP); + getUnrollingPreferences(L, TTI, UP); // Find trip count and trip multiple if count is not available unsigned TripCount = 0; @@ -426,20 +802,33 @@ bool LoopUnroll::runOnLoop(Loop *L, LPPassManager &LPM) { } unsigned Threshold, PartialThreshold; - selectThresholds(L, HasPragma, UP, Threshold, PartialThreshold); + unsigned PercentDynamicCostSavedThreshold; + unsigned DynamicCostSavingsDiscount; + selectThresholds(L, HasPragma, UP, Threshold, PartialThreshold, + PercentDynamicCostSavedThreshold, + DynamicCostSavingsDiscount); // Given Count, TripCount and thresholds determine the type of // unrolling which is to be performed. enum { Full = 0, Partial = 1, Runtime = 2 }; int Unrolling; if (TripCount && Count == TripCount) { - if (Threshold != NoThreshold && UnrolledSize > Threshold) { - DEBUG(dbgs() << " Too large to fully unroll with count: " << Count - << " because size: " << UnrolledSize << ">" << Threshold - << "\n"); - Unrolling = Partial; - } else { + Unrolling = Partial; + // If the loop is really small, we don't need to run an expensive analysis. + if (canUnrollCompletely(L, Threshold, 100, DynamicCostSavingsDiscount, + UnrolledSize, UnrolledSize)) { Unrolling = Full; + } else { + // The loop isn't that small, but we still can fully unroll it if that + // helps to remove a significant number of instructions. + // To check that, run additional analysis on the loop. + if (Optional<EstimatedUnrollCost> Cost = analyzeLoopUnrollCost( + L, TripCount, *SE, TTI, Threshold + DynamicCostSavingsDiscount)) + if (canUnrollCompletely(L, Threshold, PercentDynamicCostSavedThreshold, + DynamicCostSavingsDiscount, Cost->UnrolledCost, + Cost->RolledDynamicCost)) { + Unrolling = Full; + } } } else if (TripCount && Count < TripCount) { Unrolling = Partial; @@ -450,6 +839,9 @@ bool LoopUnroll::runOnLoop(Loop *L, LPPassManager &LPM) { // Reduce count based on the type of unrolling and the threshold values. unsigned OriginalCount = Count; bool AllowRuntime = UserRuntime ? CurrentRuntime : UP.Runtime; + if (HasRuntimeUnrollDisablePragma(L)) { + AllowRuntime = false; + } if (Unrolling == Partial) { bool AllowPartial = UserAllowPartial ? CurrentAllowPartial : UP.Partial; if (!AllowPartial && !CountSetExplicitly) { @@ -518,8 +910,8 @@ bool LoopUnroll::runOnLoop(Loop *L, LPPassManager &LPM) { } // Unroll the loop. - if (!UnrollLoop(L, Count, TripCount, AllowRuntime, TripMultiple, LI, this, - &LPM, &AC)) + if (!UnrollLoop(L, Count, TripCount, AllowRuntime, UP.AllowExpensiveTripCount, + TripMultiple, LI, this, &LPM, &AC)) return false; return true; |