From ec818d7fb3c468c9ab46663873e79772ccc77523 Mon Sep 17 00:00:00 2001 From: Hideki Saito Date: Tue, 6 Aug 2019 06:08:48 +0000 Subject: [PATCH] [LV][NFC] Share the LV illegality reporting with LoopVectorize. Reviewers: hsaito, fhahn, rengolin Reviewed By: rengolin Patch by psamolysov, thanks! Differential Revision: https://reviews.llvm.org/D62997 llvm-svn: 367980 --- .../Vectorize/LoopVectorizationLegality.h | 20 --- .../llvm/Transforms/Vectorize/LoopVectorize.h | 8 ++ .../Vectorize/LoopVectorizationLegality.cpp | 111 +++++---------- llvm/lib/Transforms/Vectorize/LoopVectorize.cpp | 157 +++++++++++++-------- .../Transforms/LoopVectorize/nofloat-report.ll | 27 ++++ 5 files changed, 170 insertions(+), 153 deletions(-) create mode 100644 llvm/test/Transforms/LoopVectorize/nofloat-report.ll diff --git a/llvm/include/llvm/Transforms/Vectorize/LoopVectorizationLegality.h b/llvm/include/llvm/Transforms/Vectorize/LoopVectorizationLegality.h index 8316677..3b0b1ee 100644 --- a/llvm/include/llvm/Transforms/Vectorize/LoopVectorizationLegality.h +++ b/llvm/include/llvm/Transforms/Vectorize/LoopVectorizationLegality.h @@ -33,18 +33,6 @@ namespace llvm { -/// Create an analysis remark that explains why vectorization failed -/// -/// \p PassName is the name of the pass (e.g. can be AlwaysPrint). \p -/// RemarkName is the identifier for the remark. If \p I is passed it is an -/// instruction that prevents vectorization. Otherwise \p TheLoop is used for -/// the location of the remark. \return the remark object that can be -/// streamed to. -OptimizationRemarkAnalysis createLVMissedAnalysis(const char *PassName, - StringRef RemarkName, - Loop *TheLoop, - Instruction *I = nullptr); - /// Utility class for getting and setting loop vectorizer hints in the form /// of loop metadata. /// This class keeps a number of loop annotations locally (as member variables) @@ -387,14 +375,6 @@ private: return LAI ? &LAI->getSymbolicStrides() : nullptr; } - /// Reports a vectorization illegality: print \p DebugMsg for debugging - /// purposes along with the corresponding optimization remark \p RemarkName. - /// If \p I is passed it is an instruction that prevents vectorization. - /// Otherwise the loop is used for the location of the remark. - void reportVectorizationFailure(const StringRef DebugMsg, - const StringRef OREMsg, const StringRef ORETag, - Instruction *I = nullptr) const; - /// The loop that we evaluate. Loop *TheLoop; diff --git a/llvm/include/llvm/Transforms/Vectorize/LoopVectorize.h b/llvm/include/llvm/Transforms/Vectorize/LoopVectorize.h index d1ec06a..d824e29 100644 --- a/llvm/include/llvm/Transforms/Vectorize/LoopVectorize.h +++ b/llvm/include/llvm/Transforms/Vectorize/LoopVectorize.h @@ -155,6 +155,14 @@ struct LoopVectorizePass : public PassInfoMixin { bool processLoop(Loop *L); }; +/// Reports a vectorization failure: print \p DebugMsg for debugging +/// purposes along with the corresponding optimization remark \p RemarkName. +/// If \p I is passed, it is an instruction that prevents vectorization. +/// Otherwise, the loop \p TheLoop is used for the location of the remark. +void reportVectorizationFailure(const StringRef DebugMsg, + const StringRef OREMsg, const StringRef ORETag, + OptimizationRemarkEmitter *ORE, Loop *TheLoop, Instruction *I = nullptr); + } // end namespace llvm #endif // LLVM_TRANSFORMS_VECTORIZE_LOOPVECTORIZE_H diff --git a/llvm/lib/Transforms/Vectorize/LoopVectorizationLegality.cpp b/llvm/lib/Transforms/Vectorize/LoopVectorizationLegality.cpp index cede505..8b311fa 100644 --- a/llvm/lib/Transforms/Vectorize/LoopVectorizationLegality.cpp +++ b/llvm/lib/Transforms/Vectorize/LoopVectorizationLegality.cpp @@ -13,6 +13,7 @@ // pass. It should be easy to create an analysis pass around it if there // is a need (but D45420 needs to happen first). // +#include "llvm/Transforms/Vectorize/LoopVectorize.h" #include "llvm/Transforms/Vectorize/LoopVectorizationLegality.h" #include "llvm/Analysis/VectorUtils.h" #include "llvm/IR/IntrinsicInst.h" @@ -47,38 +48,6 @@ static const unsigned MaxInterleaveFactor = 16; namespace llvm { -#ifndef NDEBUG -static void debugVectorizationFailure(const StringRef DebugMsg, - Instruction *I) { - dbgs() << "LV: Not vectorizing: " << DebugMsg; - if (I != nullptr) - dbgs() << " " << *I; - else - dbgs() << '.'; - dbgs() << '\n'; -} -#endif - -OptimizationRemarkAnalysis createLVMissedAnalysis(const char *PassName, - StringRef RemarkName, - Loop *TheLoop, - Instruction *I) { - Value *CodeRegion = TheLoop->getHeader(); - DebugLoc DL = TheLoop->getStartLoc(); - - if (I) { - CodeRegion = I->getParent(); - // If there is no debug location attached to the instruction, revert back to - // using the loop's. - if (I->getDebugLoc()) - DL = I->getDebugLoc(); - } - - OptimizationRemarkAnalysis R(PassName, RemarkName, DL, CodeRegion); - R << "loop not vectorized: "; - return R; -} - bool LoopVectorizeHints::Hint::validate(unsigned Val) { switch (Kind) { case HK_WIDTH: @@ -448,14 +417,6 @@ bool LoopVectorizationLegality::isUniform(Value *V) { return LAI->isUniform(V); } -void LoopVectorizationLegality::reportVectorizationFailure( - const StringRef DebugMsg, const StringRef OREMsg, - const StringRef ORETag, Instruction *I) const { - LLVM_DEBUG(debugVectorizationFailure(DebugMsg, I)); - ORE->emit(createLVMissedAnalysis(Hints->vectorizeAnalysisPassName(), - ORETag, TheLoop, I) << OREMsg); -} - bool LoopVectorizationLegality::canVectorizeOuterLoop() { assert(!TheLoop->empty() && "We are not vectorizing an outer loop."); // Store the result and return it at the end instead of exiting early, in case @@ -470,7 +431,7 @@ bool LoopVectorizationLegality::canVectorizeOuterLoop() { if (!Br) { reportVectorizationFailure("Unsupported basic block terminator", "loop control flow is not understood by vectorizer", - "CFGNotUnderstood"); + "CFGNotUnderstood", ORE, TheLoop); if (DoExtraAnalysis) Result = false; else @@ -489,7 +450,7 @@ bool LoopVectorizationLegality::canVectorizeOuterLoop() { !LI->isLoopHeader(Br->getSuccessor(1))) { reportVectorizationFailure("Unsupported conditional branch", "loop control flow is not understood by vectorizer", - "CFGNotUnderstood"); + "CFGNotUnderstood", ORE, TheLoop); if (DoExtraAnalysis) Result = false; else @@ -503,7 +464,7 @@ bool LoopVectorizationLegality::canVectorizeOuterLoop() { TheLoop /*context outer loop*/)) { reportVectorizationFailure("Outer loop contains divergent loops", "loop control flow is not understood by vectorizer", - "CFGNotUnderstood"); + "CFGNotUnderstood", ORE, TheLoop); if (DoExtraAnalysis) Result = false; else @@ -514,7 +475,7 @@ bool LoopVectorizationLegality::canVectorizeOuterLoop() { if (!setupOuterLoopInductions()) { reportVectorizationFailure("Unsupported outer loop Phi(s)", "Unsupported outer loop Phi(s)", - "UnsupportedPhi"); + "UnsupportedPhi", ORE, TheLoop); if (DoExtraAnalysis) Result = false; else @@ -621,7 +582,7 @@ bool LoopVectorizationLegality::canVectorizeInstrs() { !PhiTy->isPointerTy()) { reportVectorizationFailure("Found a non-int non-pointer PHI", "loop control flow is not understood by vectorizer", - "CFGNotUnderstood"); + "CFGNotUnderstood", ORE, TheLoop); return false; } @@ -641,7 +602,7 @@ bool LoopVectorizationLegality::canVectorizeInstrs() { if (Phi->getNumIncomingValues() != 2) { reportVectorizationFailure("Found an invalid PHI", "loop control flow is not understood by vectorizer", - "CFGNotUnderstood", Phi); + "CFGNotUnderstood", ORE, TheLoop, Phi); return false; } @@ -693,7 +654,7 @@ bool LoopVectorizationLegality::canVectorizeInstrs() { reportVectorizationFailure("Found an unidentified PHI", "value that could not be identified as " "reduction is used outside the loop", - "NonReductionValueUsedOutsideLoop", Phi); + "NonReductionValueUsedOutsideLoop", ORE, TheLoop, Phi); return false; } // end of PHI handling @@ -724,11 +685,11 @@ bool LoopVectorizationLegality::canVectorizeInstrs() { "library call cannot be vectorized. " "Try compiling with -fno-math-errno, -ffast-math, " "or similar flags", - "CantVectorizeLibcall", CI); + "CantVectorizeLibcall", ORE, TheLoop, CI); } else { reportVectorizationFailure("Found a non-intrinsic callsite", "call instruction cannot be vectorized", - "CantVectorizeLibcall", CI); + "CantVectorizeLibcall", ORE, TheLoop, CI); } return false; } @@ -743,7 +704,7 @@ bool LoopVectorizationLegality::canVectorizeInstrs() { if (!SE->isLoopInvariant(PSE.getSCEV(CI->getOperand(i)), TheLoop)) { reportVectorizationFailure("Found unvectorizable intrinsic", "intrinsic instruction cannot be vectorized", - "CantVectorizeIntrinsic", CI); + "CantVectorizeIntrinsic", ORE, TheLoop, CI); return false; } } @@ -756,7 +717,7 @@ bool LoopVectorizationLegality::canVectorizeInstrs() { isa(I)) { reportVectorizationFailure("Found unvectorizable type", "instruction return type cannot be vectorized", - "CantVectorizeInstructionReturnType", &I); + "CantVectorizeInstructionReturnType", ORE, TheLoop, &I); return false; } @@ -766,7 +727,7 @@ bool LoopVectorizationLegality::canVectorizeInstrs() { if (!VectorType::isValidElementType(T)) { reportVectorizationFailure("Store instruction cannot be vectorized", "store instruction cannot be vectorized", - "CantVectorizeStore", ST); + "CantVectorizeStore", ORE, TheLoop, ST); return false; } @@ -781,7 +742,7 @@ bool LoopVectorizationLegality::canVectorizeInstrs() { reportVectorizationFailure( "nontemporal store instruction cannot be vectorized", "nontemporal store instruction cannot be vectorized", - "CantVectorizeNontemporalStore", ST); + "CantVectorizeNontemporalStore", ORE, TheLoop, ST); return false; } } @@ -797,7 +758,7 @@ bool LoopVectorizationLegality::canVectorizeInstrs() { reportVectorizationFailure( "nontemporal load instruction cannot be vectorized", "nontemporal load instruction cannot be vectorized", - "CantVectorizeNontemporalLoad", LD); + "CantVectorizeNontemporalLoad", ORE, TheLoop, LD); return false; } } @@ -826,7 +787,7 @@ bool LoopVectorizationLegality::canVectorizeInstrs() { } reportVectorizationFailure("Value cannot be used outside the loop", "value cannot be used outside the loop", - "ValueUsedOutsideLoop", &I); + "ValueUsedOutsideLoop", ORE, TheLoop, &I); return false; } } // next instr. @@ -836,12 +797,12 @@ bool LoopVectorizationLegality::canVectorizeInstrs() { if (Inductions.empty()) { reportVectorizationFailure("Did not find one integer induction var", "loop induction variable could not be identified", - "NoInductionVariable"); + "NoInductionVariable", ORE, TheLoop); return false; } else if (!WidestIndTy) { reportVectorizationFailure("Did not find one integer induction var", "integer loop induction variable could not be identified", - "NoIntegerInductionVariable"); + "NoIntegerInductionVariable", ORE, TheLoop); return false; } else { LLVM_DEBUG(dbgs() << "LV: Did not find one integer induction var.\n"); @@ -872,7 +833,7 @@ bool LoopVectorizationLegality::canVectorizeMemory() { if (LAI->hasDependenceInvolvingLoopInvariantAddress()) { reportVectorizationFailure("Stores to a uniform address", "write to a loop invariant address could not be vectorized", - "CantVectorizeStoreToLoopInvariantAddress"); + "CantVectorizeStoreToLoopInvariantAddress", ORE, TheLoop); return false; } Requirements->addRuntimePointerChecks(LAI->getNumRuntimePointerChecks()); @@ -956,7 +917,8 @@ bool LoopVectorizationLegality::canVectorizeWithIfConvert() { if (!EnableIfConversion) { reportVectorizationFailure("If-conversion is disabled", "if-conversion is disabled", - "IfConversionDisabled"); + "IfConversionDisabled", + ORE, TheLoop); return false; } @@ -982,7 +944,8 @@ bool LoopVectorizationLegality::canVectorizeWithIfConvert() { if (!isa(BB->getTerminator())) { reportVectorizationFailure("Loop contains a switch statement", "loop contains a switch statement", - "LoopContainsSwitch", BB->getTerminator()); + "LoopContainsSwitch", ORE, TheLoop, + BB->getTerminator()); return false; } @@ -992,14 +955,16 @@ bool LoopVectorizationLegality::canVectorizeWithIfConvert() { reportVectorizationFailure( "Control flow cannot be substituted for a select", "control flow cannot be substituted for a select", - "NoCFGForSelect", BB->getTerminator()); + "NoCFGForSelect", ORE, TheLoop, + BB->getTerminator()); return false; } } else if (BB != Header && !canIfConvertPHINodes(BB)) { reportVectorizationFailure( "Control flow cannot be substituted for a select", "control flow cannot be substituted for a select", - "NoCFGForSelect", BB->getTerminator()); + "NoCFGForSelect", ORE, TheLoop, + BB->getTerminator()); return false; } } @@ -1029,7 +994,7 @@ bool LoopVectorizationLegality::canVectorizeLoopCFG(Loop *Lp, if (!Lp->getLoopPreheader()) { reportVectorizationFailure("Loop doesn't have a legal pre-header", "loop control flow is not understood by vectorizer", - "CFGNotUnderstood"); + "CFGNotUnderstood", ORE, TheLoop); if (DoExtraAnalysis) Result = false; else @@ -1040,7 +1005,7 @@ bool LoopVectorizationLegality::canVectorizeLoopCFG(Loop *Lp, if (Lp->getNumBackEdges() != 1) { reportVectorizationFailure("The loop must have a single backedge", "loop control flow is not understood by vectorizer", - "CFGNotUnderstood"); + "CFGNotUnderstood", ORE, TheLoop); if (DoExtraAnalysis) Result = false; else @@ -1051,7 +1016,7 @@ bool LoopVectorizationLegality::canVectorizeLoopCFG(Loop *Lp, if (!Lp->getExitingBlock()) { reportVectorizationFailure("The loop must have an exiting block", "loop control flow is not understood by vectorizer", - "CFGNotUnderstood"); + "CFGNotUnderstood", ORE, TheLoop); if (DoExtraAnalysis) Result = false; else @@ -1064,7 +1029,7 @@ bool LoopVectorizationLegality::canVectorizeLoopCFG(Loop *Lp, if (Lp->getExitingBlock() != Lp->getLoopLatch()) { reportVectorizationFailure("The exiting block is not the loop latch", "loop control flow is not understood by vectorizer", - "CFGNotUnderstood"); + "CFGNotUnderstood", ORE, TheLoop); if (DoExtraAnalysis) Result = false; else @@ -1127,7 +1092,8 @@ bool LoopVectorizationLegality::canVectorize(bool UseVPlanNativePath) { if (!canVectorizeOuterLoop()) { reportVectorizationFailure("Unsupported outer loop", "unsupported outer loop", - "UnsupportedOuterLoop"); + "UnsupportedOuterLoop", + ORE, TheLoop); // TODO: Implement DoExtraAnalysis when subsequent legal checks support // outer loops. return false; @@ -1179,7 +1145,7 @@ bool LoopVectorizationLegality::canVectorize(bool UseVPlanNativePath) { if (PSE.getUnionPredicate().getComplexity() > SCEVThreshold) { reportVectorizationFailure("Too many SCEV checks needed", "Too many SCEV assumptions need to be made and checked at runtime", - "TooManySCEVRunTimeChecks"); + "TooManySCEVRunTimeChecks", ORE, TheLoop); if (DoExtraAnalysis) Result = false; else @@ -1202,7 +1168,7 @@ bool LoopVectorizationLegality::canFoldTailByMasking() { "No primary induction, cannot fold tail by masking", "Missing a primary induction variable in the loop, which is " "needed in order to fold tail by masking as required.", - "NoPrimaryInduction"); + "NoPrimaryInduction", ORE, TheLoop); return false; } @@ -1211,7 +1177,7 @@ bool LoopVectorizationLegality::canFoldTailByMasking() { reportVectorizationFailure( "Loop has reductions, cannot fold tail by masking", "Cannot fold tail by masking in the presence of reductions.", - "ReductionFoldingTailByMasking"); + "ReductionFoldingTailByMasking", ORE, TheLoop); return false; } @@ -1225,7 +1191,7 @@ bool LoopVectorizationLegality::canFoldTailByMasking() { reportVectorizationFailure( "Cannot fold tail by masking, loop has an outside user for", "Cannot fold tail by masking in the presence of live outs.", - "LiveOutFoldingTailByMasking", UI); + "LiveOutFoldingTailByMasking", ORE, TheLoop, UI); return false; } } @@ -1240,7 +1206,8 @@ bool LoopVectorizationLegality::canFoldTailByMasking() { reportVectorizationFailure( "Cannot fold tail by masking as required", "control flow cannot be substituted for a select", - "NoCFGForSelect", BB->getTerminator()); + "NoCFGForSelect", ORE, TheLoop, + BB->getTerminator()); return false; } } diff --git a/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp b/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp index 5a0ce75..dd28b18 100644 --- a/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp +++ b/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp @@ -795,6 +795,59 @@ void InnerLoopVectorizer::setDebugLocFromInst(IRBuilder<> &B, const Value *Ptr) B.SetCurrentDebugLocation(DebugLoc()); } +/// Write a record \p DebugMsg about vectorization failure to the debug +/// output stream. If \p I is passed, it is an instruction that prevents +/// vectorization. +#ifndef NDEBUG +static void debugVectorizationFailure(const StringRef DebugMsg, + Instruction *I) { + dbgs() << "LV: Not vectorizing: " << DebugMsg; + if (I != nullptr) + dbgs() << " " << *I; + else + dbgs() << '.'; + dbgs() << '\n'; +} +#endif + +/// Create an analysis remark that explains why vectorization failed +/// +/// \p PassName is the name of the pass (e.g. can be AlwaysPrint). \p +/// RemarkName is the identifier for the remark. If \p I is passed it is an +/// instruction that prevents vectorization. Otherwise \p TheLoop is used for +/// the location of the remark. \return the remark object that can be +/// streamed to. +static OptimizationRemarkAnalysis createLVAnalysis(const char *PassName, + StringRef RemarkName, Loop *TheLoop, Instruction *I) { + Value *CodeRegion = TheLoop->getHeader(); + DebugLoc DL = TheLoop->getStartLoc(); + + if (I) { + CodeRegion = I->getParent(); + // If there is no debug location attached to the instruction, revert back to + // using the loop's. + if (I->getDebugLoc()) + DL = I->getDebugLoc(); + } + + OptimizationRemarkAnalysis R(PassName, RemarkName, DL, CodeRegion); + R << "loop not vectorized: "; + return R; +} + +namespace llvm { + +void reportVectorizationFailure(const StringRef DebugMsg, + const StringRef OREMsg, const StringRef ORETag, + OptimizationRemarkEmitter *ORE, Loop *TheLoop, Instruction *I) { + LLVM_DEBUG(debugVectorizationFailure(DebugMsg, I)); + LoopVectorizeHints Hints(TheLoop, true /* doesn't matter */, *ORE); + ORE->emit(createLVAnalysis(Hints.vectorizeAnalysisPassName(), + ORETag, TheLoop, I) << OREMsg); +} + +} // end namespace llvm + #ifndef NDEBUG /// \return string containing a file name and a line # for the given loop. static std::string getDebugLocString(const Loop *L) { @@ -1274,15 +1327,6 @@ private: /// should be used. bool useEmulatedMaskMemRefHack(Instruction *I); - /// Create an analysis remark that explains why vectorization failed - /// - /// \p RemarkName is the identifier for the remark. \return the remark object - /// that can be streamed to. - OptimizationRemarkAnalysis createMissedAnalysis(StringRef RemarkName) { - return createLVMissedAnalysis(Hints->vectorizeAnalysisPassName(), - RemarkName, TheLoop); - } - /// Map of scalar integer values to the smallest bitwidth they can be legally /// represented as. The vector equivalents of these values should be truncated /// to this type. @@ -4707,36 +4751,30 @@ bool LoopVectorizationCostModel::runtimeChecksRequired() { LLVM_DEBUG(dbgs() << "LV: Performing code size checks.\n"); if (Legal->getRuntimePointerChecking()->Need) { - ORE->emit(createMissedAnalysis("CantVersionLoopWithOptForSize") - << "runtime pointer checks needed. Enable vectorization of this " - "loop with '#pragma clang loop vectorize(enable)' when " - "compiling with -Os/-Oz"); - LLVM_DEBUG( - dbgs() - << "LV: Aborting. Runtime ptr check is required with -Os/-Oz.\n"); + reportVectorizationFailure("Runtime ptr check is required with -Os/-Oz", + "runtime pointer checks needed. Enable vectorization of this " + "loop with '#pragma clang loop vectorize(enable)' when " + "compiling with -Os/-Oz", + "CantVersionLoopWithOptForSize", ORE, TheLoop); return true; } if (!PSE.getUnionPredicate().getPredicates().empty()) { - ORE->emit(createMissedAnalysis("CantVersionLoopWithOptForSize") - << "runtime SCEV checks needed. Enable vectorization of this " - "loop with '#pragma clang loop vectorize(enable)' when " - "compiling with -Os/-Oz"); - LLVM_DEBUG( - dbgs() - << "LV: Aborting. Runtime SCEV check is required with -Os/-Oz.\n"); + reportVectorizationFailure("Runtime SCEV check is required with -Os/-Oz", + "runtime SCEV checks needed. Enable vectorization of this " + "loop with '#pragma clang loop vectorize(enable)' when " + "compiling with -Os/-Oz", + "CantVersionLoopWithOptForSize", ORE, TheLoop); return true; } // FIXME: Avoid specializing for stride==1 instead of bailing out. if (!Legal->getLAI()->getSymbolicStrides().empty()) { - ORE->emit(createMissedAnalysis("CantVersionLoopWithOptForSize") - << "runtime stride == 1 checks needed. Enable vectorization of " - "this loop with '#pragma clang loop vectorize(enable)' when " - "compiling with -Os/-Oz"); - LLVM_DEBUG( - dbgs() - << "LV: Aborting. Runtime stride check is required with -Os/-Oz.\n"); + reportVectorizationFailure("Runtime stride check is required with -Os/-Oz", + "runtime stride == 1 checks needed. Enable vectorization of " + "this loop with '#pragma clang loop vectorize(enable)' when " + "compiling with -Os/-Oz", + "CantVersionLoopWithOptForSize", ORE, TheLoop); return true; } @@ -4747,22 +4785,19 @@ Optional LoopVectorizationCostModel::computeMaxVF() { if (Legal->getRuntimePointerChecking()->Need && TTI.hasBranchDivergence()) { // TODO: It may by useful to do since it's still likely to be dynamically // uniform if the target can skip. - LLVM_DEBUG( - dbgs() << "LV: Not inserting runtime ptr check for divergent target"); - - ORE->emit( - createMissedAnalysis("CantVersionLoopWithDivergentTarget") - << "runtime pointer checks needed. Not enabled for divergent target"); - + reportVectorizationFailure( + "Not inserting runtime ptr check for divergent target", + "runtime pointer checks needed. Not enabled for divergent target", + "CantVersionLoopWithDivergentTarget", ORE, TheLoop); return None; } unsigned TC = PSE.getSE()->getSmallConstantTripCount(TheLoop); LLVM_DEBUG(dbgs() << "LV: Found trip count: " << TC << '\n'); if (TC == 1) { - ORE->emit(createMissedAnalysis("SingleIterationLoop") - << "loop trip count is one, irrelevant for vectorization"); - LLVM_DEBUG(dbgs() << "LV: Aborting, single iteration (non) loop.\n"); + reportVectorizationFailure("Single iteration (non) loop", + "loop trip count is one, irrelevant for vectorization", + "SingleIterationLoop", ORE, TheLoop); return None; } @@ -4816,16 +4851,19 @@ Optional LoopVectorizationCostModel::computeMaxVF() { } if (TC == 0) { - ORE->emit( - createMissedAnalysis("UnknownLoopCountComplexCFG") - << "unable to calculate the loop count due to complex control flow"); + reportVectorizationFailure( + "Unable to calculate the loop count due to complex control flow", + "unable to calculate the loop count due to complex control flow", + "UnknownLoopCountComplexCFG", ORE, TheLoop); return None; } - ORE->emit(createMissedAnalysis("NoTailLoopWithOptForSize") - << "cannot optimize for size and vectorize at the same time. " - "Enable vectorization of this loop with '#pragma clang loop " - "vectorize(enable)' when compiling with -Os/-Oz"); + reportVectorizationFailure( + "Cannot optimize for size and vectorize at the same time.", + "cannot optimize for size and vectorize at the same time. " + "Enable vectorization of this loop with '#pragma clang loop " + "vectorize(enable)' when compiling with -Os/-Oz", + "NoTailLoopWithOptForSize", ORE, TheLoop); return None; } @@ -4936,10 +4974,9 @@ LoopVectorizationCostModel::selectVectorizationFactor(unsigned MaxVF) { } if (!EnableCondStoresVectorization && NumPredStores) { - ORE->emit(createMissedAnalysis("ConditionalStore") - << "store that is conditionally executed prevents vectorization"); - LLVM_DEBUG( - dbgs() << "LV: No vectorization. There are conditional stores.\n"); + reportVectorizationFailure("There are conditional stores.", + "store that is conditionally executed prevents vectorization", + "ConditionalStore", ORE, TheLoop); Width = 1; Cost = ScalarCost; } @@ -7423,11 +7460,10 @@ bool LoopVectorizePass::processLoop(Loop *L) { // an integer loop and the vector instructions selected are purely integer // vector instructions? if (F->hasFnAttribute(Attribute::NoImplicitFloat)) { - LLVM_DEBUG(dbgs() << "LV: Can't vectorize when the NoImplicitFloat" - "attribute is used.\n"); - ORE->emit(createLVMissedAnalysis(Hints.vectorizeAnalysisPassName(), - "NoImplicitFloat", L) - << "loop not vectorized due to NoImplicitFloat attribute"); + reportVectorizationFailure( + "Can't vectorize when the NoImplicitFloat attribute is used", + "loop not vectorized due to NoImplicitFloat attribute", + "NoImplicitFloat", ORE, L); Hints.emitRemarkWithHints(); return false; } @@ -7438,11 +7474,10 @@ bool LoopVectorizePass::processLoop(Loop *L) { // additional fp-math flags can help. if (Hints.isPotentiallyUnsafe() && TTI->isFPVectorizationPotentiallyUnsafe()) { - LLVM_DEBUG( - dbgs() << "LV: Potentially unsafe FP op prevents vectorization.\n"); - ORE->emit( - createLVMissedAnalysis(Hints.vectorizeAnalysisPassName(), "UnsafeFP", L) - << "loop not vectorized due to unsafe FP support."); + reportVectorizationFailure( + "Potentially unsafe FP op prevents vectorization", + "loop not vectorized due to unsafe FP support.", + "UnsafeFP", ORE, L); Hints.emitRemarkWithHints(); return false; } diff --git a/llvm/test/Transforms/LoopVectorize/nofloat-report.ll b/llvm/test/Transforms/LoopVectorize/nofloat-report.ll new file mode 100644 index 0000000..39ba1e9 --- /dev/null +++ b/llvm/test/Transforms/LoopVectorize/nofloat-report.ll @@ -0,0 +1,27 @@ +; REQUIRES: asserts +; RUN: opt -loop-vectorize -force-vector-interleave=1 -force-vector-width=4 -debug-only=loop-vectorize\ +; RUN: -disable-output -S %s 2>&1 | FileCheck %s + +; Make sure that we report about not vectorizing functions with 'noimplicitfloat' attributes + +target datalayout = "e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128-n8:16:32:64-S128" + +@a = common global [2048 x i32] zeroinitializer, align 16 + +; CHECK: LV: Not vectorizing: Can't vectorize when the NoImplicitFloat attribute is used +define void @example_nofloat() noimplicitfloat { ; <--------- "noimplicitfloat" attribute here! + br label %1 + +;