diff options
Diffstat (limited to 'llvm/lib/Transforms')
-rw-r--r-- | llvm/lib/Transforms/Scalar/LoopUnrollPass.cpp | 340 |
1 files changed, 141 insertions, 199 deletions
diff --git a/llvm/lib/Transforms/Scalar/LoopUnrollPass.cpp b/llvm/lib/Transforms/Scalar/LoopUnrollPass.cpp index bc0f1e9f400..8551862f8dc 100644 --- a/llvm/lib/Transforms/Scalar/LoopUnrollPass.cpp +++ b/llvm/lib/Transforms/Scalar/LoopUnrollPass.cpp @@ -39,16 +39,16 @@ using namespace llvm; #define DEBUG_TYPE "loop-unroll" static cl::opt<unsigned> - UnrollThreshold("unroll-threshold", cl::init(150), cl::Hidden, + UnrollThreshold("unroll-threshold", 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, + "unroll-percent-dynamic-cost-saved-threshold", 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, + "unroll-dynamic-cost-savings-discount", 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).")); @@ -59,17 +59,17 @@ static cl::opt<unsigned> UnrollMaxIterationsCountToAnalyze( "iterations when checking full unroll profitability")); static cl::opt<unsigned> -UnrollCount("unroll-count", cl::init(0), cl::Hidden, +UnrollCount("unroll-count", cl::Hidden, cl::desc("Use this unroll count for all loops including those with " "unroll_count pragma values, for testing purposes")); static cl::opt<bool> -UnrollAllowPartial("unroll-allow-partial", cl::init(false), cl::Hidden, +UnrollAllowPartial("unroll-allow-partial", cl::Hidden, cl::desc("Allows loops to be partially unrolled until " "-unroll-threshold loop size is reached.")); static cl::opt<bool> -UnrollRuntime("unroll-runtime", cl::ZeroOrMore, cl::init(false), cl::Hidden, +UnrollRuntime("unroll-runtime", cl::ZeroOrMore, cl::Hidden, cl::desc("Unroll loops with run-time trip counts")); static cl::opt<unsigned> @@ -77,59 +77,114 @@ PragmaUnrollThreshold("pragma-unroll-threshold", cl::init(16 * 1024), cl::Hidden cl::desc("Unrolled size limit for loops with an unroll(full) or " "unroll_count pragma.")); + +/// A magic value for use with the Threshold parameter to indicate +/// that the loop unroll should be performed regardless of how much +/// code expansion would result. +static const unsigned NoThreshold = UINT_MAX; + +/// Default unroll count for loops with run-time trip count if +/// -unroll-count is not set +static const unsigned DefaultUnrollRuntimeCount = 8; + +/// Gather the various unrolling parameters based on the defaults, compiler +/// flags, TTI overrides, pragmas, and user specified parameters. +static TargetTransformInfo::UnrollingPreferences gatherUnrollingPreferences( + Loop *L, const TargetTransformInfo &TTI, Optional<unsigned> UserThreshold, + Optional<unsigned> UserCount, Optional<bool> UserAllowPartial, + Optional<bool> UserRuntime, unsigned PragmaCount, bool PragmaFullUnroll, + bool PragmaEnableUnroll, unsigned TripCount) { + TargetTransformInfo::UnrollingPreferences UP; + + // Set up the defaults + UP.Threshold = 150; + UP.PercentDynamicCostSavedThreshold = 20; + UP.DynamicCostSavingsDiscount = 2000; + UP.OptSizeThreshold = 50; + UP.PartialThreshold = UP.Threshold; + UP.PartialOptSizeThreshold = UP.OptSizeThreshold; + UP.Count = 0; + UP.MaxCount = UINT_MAX; + UP.Partial = false; + UP.Runtime = false; + UP.AllowExpensiveTripCount = false; + + // Override with any target specific settings + TTI.getUnrollingPreferences(L, UP); + + // Apply size attributes + if (L->getHeader()->getParent()->optForSize()) { + UP.Threshold = UP.OptSizeThreshold; + UP.PartialThreshold = UP.PartialOptSizeThreshold; + } + + // Apply unroll count pragmas + if (PragmaCount) + UP.Count = PragmaCount; + else if (PragmaFullUnroll) + UP.Count = TripCount; + + // Apply any user values specified by cl::opt + if (UnrollThreshold.getNumOccurrences() > 0) { + UP.Threshold = UnrollThreshold; + UP.PartialThreshold = UnrollThreshold; + } + if (UnrollPercentDynamicCostSavedThreshold.getNumOccurrences() > 0) + UP.PercentDynamicCostSavedThreshold = + UnrollPercentDynamicCostSavedThreshold; + if (UnrollDynamicCostSavingsDiscount.getNumOccurrences() > 0) + UP.DynamicCostSavingsDiscount = UnrollDynamicCostSavingsDiscount; + if (UnrollCount.getNumOccurrences() > 0) + UP.Count = UnrollCount; + if (UnrollAllowPartial.getNumOccurrences() > 0) + UP.Partial = UnrollAllowPartial; + if (UnrollRuntime.getNumOccurrences() > 0) + UP.Runtime = UnrollRuntime; + + // Apply user values provided by argument + if (UserThreshold.hasValue()) { + UP.Threshold = *UserThreshold; + UP.PartialThreshold = *UserThreshold; + } + if (UserCount.hasValue()) + UP.Count = *UserCount; + if (UserAllowPartial.hasValue()) + UP.Partial = *UserAllowPartial; + if (UserRuntime.hasValue()) + UP.Runtime = *UserRuntime; + + if (PragmaCount > 0 || + ((PragmaFullUnroll || PragmaEnableUnroll) && TripCount != 0)) { + // If the loop has an unrolling pragma, we want to be more aggressive with + // unrolling limits. Set thresholds to at least the PragmaTheshold value + // which is larger than the default limits. + if (UP.Threshold != NoThreshold) + UP.Threshold = std::max<unsigned>(UP.Threshold, PragmaUnrollThreshold); + if (UP.PartialThreshold != NoThreshold) + UP.PartialThreshold = + std::max<unsigned>(UP.PartialThreshold, PragmaUnrollThreshold); + } + + return UP; +} + namespace { class LoopUnroll : public LoopPass { public: 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); - UserCount = (C != -1) || (UnrollCount.getNumOccurrences() > 0); - + LoopUnroll(Optional<unsigned> Threshold = None, + Optional<unsigned> Count = None, + Optional<bool> AllowPartial = None, + Optional<bool> Runtime = None) + : LoopPass(ID), ProvidedCount(Count), ProvidedThreshold(Threshold), + ProvidedAllowPartial(AllowPartial), ProvidedRuntime(Runtime) { initializeLoopUnrollPass(*PassRegistry::getPassRegistry()); } - /// A magic value for use with the Threshold parameter to indicate - /// that the loop unroll should be performed regardless of how much - /// code expansion would result. - static const unsigned NoThreshold = UINT_MAX; - - // Threshold to use when optsize is specified (and there is no - // explicit -unroll-threshold). - static const unsigned OptSizeUnrollThreshold = 50; - - // Default unroll count for loops with run-time trip count if - // -unroll-count is not set - static const unsigned UnrollRuntimeCount = 8; - - unsigned CurrentCount; - unsigned CurrentThreshold; - 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; + Optional<unsigned> ProvidedCount; + Optional<unsigned> ProvidedThreshold; + Optional<bool> ProvidedAllowPartial; + Optional<bool> ProvidedRuntime; bool runOnLoop(Loop *L, LPPassManager &) override; @@ -156,75 +211,6 @@ namespace { AU.addPreserved<GlobalsAAWrapperPass>(); } - // Fill in the UnrollingPreferences parameter with values from the - // TargetTransformationInfo. - 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; - UP.Count = CurrentCount; - UP.MaxCount = UINT_MAX; - UP.Partial = CurrentAllowPartial; - UP.Runtime = CurrentRuntime; - UP.AllowExpensiveTripCount = false; - TTI.getUnrollingPreferences(L, UP); - } - - // Select and return an unroll count based on parameters from - // user, unroll preferences, unroll pragmas, or a heuristic. - // SetExplicitly is set to true if the unroll count is is set by - // the user or a pragma rather than selected heuristically. - unsigned - selectUnrollCount(const Loop *L, unsigned TripCount, bool PragmaFullUnroll, - unsigned PragmaCount, - const TargetTransformInfo::UnrollingPreferences &UP, - bool &SetExplicitly); - - // Select threshold values used to limit unrolling based on a - // total unrolled size. Parameters Threshold and PartialThreshold - // are set to the maximum unrolled size for fully and partially - // unrolled loops respectively. - void selectThresholds(const Loop *L, bool UsePragmaThreshold, - const TargetTransformInfo::UnrollingPreferences &UP, - 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 - // optimize-for-size, and the unroll threshold was not user - // 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()->optForSize()) { - Threshold = UP.OptSizeThreshold; - PartialThreshold = UP.PartialOptSizeThreshold; - } - if (UsePragmaThreshold) { - // If the loop has an unrolling pragma, we want to be more - // aggressive with unrolling limits. Set thresholds to at - // least the PragmaTheshold value which is larger than the - // default limits. - if (Threshold != NoThreshold) - Threshold = std::max<unsigned>(Threshold, PragmaUnrollThreshold); - if (PartialThreshold != NoThreshold) - PartialThreshold = - std::max<unsigned>(PartialThreshold, PragmaUnrollThreshold); - } - } bool canUnrollCompletely(Loop *L, unsigned Threshold, unsigned PercentDynamicCostSavedThreshold, unsigned DynamicCostSavingsDiscount, @@ -245,7 +231,14 @@ INITIALIZE_PASS_END(LoopUnroll, "loop-unroll", "Unroll loops", false, false) Pass *llvm::createLoopUnrollPass(int Threshold, int Count, int AllowPartial, int Runtime) { - return new LoopUnroll(Threshold, Count, AllowPartial, Runtime); + // TODO: It would make more sense for this function to take the optionals + // directly, but that's dangerous since it would silently break out of tree + // callers. + return new LoopUnroll(Threshold == -1 ? None : Optional<unsigned>(Threshold), + Count == -1 ? None : Optional<unsigned>(Count), + AllowPartial == -1 ? None + : Optional<bool>(AllowPartial), + Runtime == -1 ? None : Optional<bool>(Runtime)); } Pass *llvm::createSimpleLoopUnrollPass() { @@ -843,46 +836,6 @@ bool LoopUnroll::canUnrollCompletely(Loop *L, unsigned Threshold, return false; } -unsigned LoopUnroll::selectUnrollCount( - const Loop *L, unsigned TripCount, bool PragmaFullUnroll, - unsigned PragmaCount, const TargetTransformInfo::UnrollingPreferences &UP, - bool &SetExplicitly) { - SetExplicitly = true; - - // User-specified count (either as a command-line option or - // constructor parameter) has highest precedence. - unsigned Count = UserCount ? CurrentCount : 0; - - // If there is no user-specified count, unroll pragmas have the next - // highest precedence. - if (Count == 0) { - if (PragmaCount) { - Count = PragmaCount; - } else if (PragmaFullUnroll) { - Count = TripCount; - } - } - - if (Count == 0) - Count = UP.Count; - - if (Count == 0) { - SetExplicitly = false; - if (TripCount == 0) - // Runtime trip count. - Count = UnrollRuntimeCount; - else - // Conservative heuristic: if we know the trip count, see if we can - // completely unroll (subject to the threshold, checked below); otherwise - // try to find greatest modulo of the trip count which is still under - // threshold value. - Count = TripCount; - } - if (TripCount && Count > TripCount) - return TripCount; - return Count; -} - bool LoopUnroll::runOnLoop(Loop *L, LPPassManager &) { if (skipOptnoneFunction(L)) return false; @@ -909,9 +862,6 @@ bool LoopUnroll::runOnLoop(Loop *L, LPPassManager &) { unsigned PragmaCount = UnrollCountPragmaValue(L); bool HasPragma = PragmaFullUnroll || PragmaEnableUnroll || PragmaCount > 0; - TargetTransformInfo::UnrollingPreferences UP; - getUnrollingPreferences(L, TTI, UP); - // Find trip count and trip multiple if count is not available unsigned TripCount = 0; unsigned TripMultiple = 1; @@ -926,11 +876,18 @@ bool LoopUnroll::runOnLoop(Loop *L, LPPassManager &) { TripMultiple = SE->getSmallConstantTripMultiple(L, ExitingBlock); } - // Select an initial unroll count. This may be reduced later based - // on size thresholds. - bool CountSetExplicitly; - unsigned Count = selectUnrollCount(L, TripCount, PragmaFullUnroll, - PragmaCount, UP, CountSetExplicitly); + TargetTransformInfo::UnrollingPreferences UP = gatherUnrollingPreferences( + L, TTI, ProvidedThreshold, ProvidedCount, ProvidedAllowPartial, + ProvidedRuntime, PragmaCount, PragmaFullUnroll, PragmaEnableUnroll, + TripCount); + + unsigned Count = UP.Count; + bool CountSetExplicitly = Count != 0; + // Use a heuristic count if we didn't set anything explicitly. + if (!CountSetExplicitly) + Count = TripCount == 0 ? DefaultUnrollRuntimeCount : TripCount; + if (TripCount && Count > TripCount) + Count = TripCount; unsigned NumInlineCandidates; bool notDuplicatable; @@ -952,21 +909,6 @@ bool LoopUnroll::runOnLoop(Loop *L, LPPassManager &) { return false; } - unsigned Threshold, PartialThreshold; - unsigned PercentDynamicCostSavedThreshold; - unsigned DynamicCostSavingsDiscount; - // Only use the high pragma threshold when we have a target unroll factor such - // as with "#pragma unroll N" or a pragma indicating full unrolling and the - // trip count is known. Otherwise we rely on the standard threshold to - // heuristically select a reasonable unroll count. - bool UsePragmaThreshold = - PragmaCount > 0 || - ((PragmaFullUnroll || PragmaEnableUnroll) && TripCount != 0); - - selectThresholds(L, UsePragmaThreshold, 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 }; @@ -974,19 +916,20 @@ bool LoopUnroll::runOnLoop(Loop *L, LPPassManager &) { if (TripCount && Count == TripCount) { Unrolling = Partial; // If the loop is really small, we don't need to run an expensive analysis. - if (canUnrollCompletely(L, Threshold, 100, DynamicCostSavingsDiscount, + if (canUnrollCompletely(L, UP.Threshold, 100, UP.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, DT, *SE, TTI, - Threshold + DynamicCostSavingsDiscount)) - if (canUnrollCompletely(L, Threshold, PercentDynamicCostSavedThreshold, - DynamicCostSavingsDiscount, Cost->UnrolledCost, - Cost->RolledDynamicCost)) { + if (Optional<EstimatedUnrollCost> Cost = analyzeLoopUnrollCost( + L, TripCount, DT, *SE, TTI, + UP.Threshold + UP.DynamicCostSavingsDiscount)) + if (canUnrollCompletely(L, UP.Threshold, + UP.PercentDynamicCostSavedThreshold, + UP.DynamicCostSavingsDiscount, + Cost->UnrolledCost, Cost->RolledDynamicCost)) { Unrolling = Full; } } @@ -998,23 +941,22 @@ bool LoopUnroll::runOnLoop(Loop *L, LPPassManager &) { // Reduce count based on the type of unrolling and the threshold values. unsigned OriginalCount = Count; - bool AllowRuntime = PragmaEnableUnroll || (PragmaCount > 0) || - (UserRuntime ? CurrentRuntime : UP.Runtime); + bool AllowRuntime = PragmaEnableUnroll || (PragmaCount > 0) || UP.Runtime; // Don't unroll a runtime trip count loop with unroll full pragma. if (HasRuntimeUnrollDisablePragma(L) || PragmaFullUnroll) { AllowRuntime = false; } if (Unrolling == Partial) { - bool AllowPartial = PragmaEnableUnroll || - (UserAllowPartial ? CurrentAllowPartial : UP.Partial); + bool AllowPartial = PragmaEnableUnroll || UP.Partial; if (!AllowPartial && !CountSetExplicitly) { DEBUG(dbgs() << " will not try to unroll partially because " << "-unroll-allow-partial not given\n"); return false; } - if (PartialThreshold != NoThreshold && UnrolledSize > PartialThreshold) { + if (UP.PartialThreshold != NoThreshold && + UnrolledSize > UP.PartialThreshold) { // Reduce unroll count to be modulo of TripCount for partial unrolling. - Count = (std::max(PartialThreshold, 3u)-2) / (LoopSize-2); + Count = (std::max(UP.PartialThreshold, 3u) - 2) / (LoopSize - 2); while (Count != 0 && TripCount % Count != 0) Count--; } @@ -1026,7 +968,7 @@ bool LoopUnroll::runOnLoop(Loop *L, LPPassManager &) { } // Reduce unroll count to be the largest power-of-two factor of // the original count which satisfies the threshold limit. - while (Count != 0 && UnrolledSize > PartialThreshold) { + while (Count != 0 && UnrolledSize > UP.PartialThreshold) { Count >>= 1; UnrolledSize = (LoopSize-2) * Count + 2; } |