diff --git a/llvm/lib/Passes/PassBuilderPipelines.cpp b/llvm/lib/Passes/PassBuilderPipelines.cpp --- a/llvm/lib/Passes/PassBuilderPipelines.cpp +++ b/llvm/lib/Passes/PassBuilderPipelines.cpp @@ -134,11 +134,11 @@ "enable-ml-inliner", cl::init(InliningAdvisorMode::Default), cl::Hidden, cl::desc("Enable ML policy for inliner. Currently trained for -Oz only"), cl::values(clEnumValN(InliningAdvisorMode::Default, "default", - "Heuristics-based inliner version."), + "Heuristics-based inliner version"), clEnumValN(InliningAdvisorMode::Development, "development", - "Use development mode (runtime-loadable model)."), + "Use development mode (runtime-loadable model)"), clEnumValN(InliningAdvisorMode::Release, "release", - "Use release mode (AOT-compiled model)."))); + "Use release mode (AOT-compiled model)"))); static cl::opt EnableSyntheticCounts( "enable-npm-synthetic-counts", cl::Hidden, @@ -161,7 +161,7 @@ static cl::opt PerformMandatoryInliningsFirst( "mandatory-inlining-first", cl::init(true), cl::Hidden, cl::desc("Perform mandatory inlinings module-wide, before performing " - "inlining.")); + "inlining")); static cl::opt EnableO3NonTrivialUnswitching( "enable-npm-O3-nontrivial-unswitch", cl::init(true), cl::Hidden, @@ -190,6 +190,99 @@ "enable-global-analyses", cl::init(true), cl::Hidden, cl::desc("Enable inter-procedural analyses")); +static cl::opt + RunPartialInlining("enable-partial-inlining", cl::init(false), cl::Hidden, + cl::desc("Run Partial inlinining pass")); + +static cl::opt ExtraVectorizerPasses( + "extra-vectorizer-passes", cl::init(false), cl::Hidden, + cl::desc("Run cleanup optimization passes after vectorization")); + +static cl::opt RunNewGVN("enable-newgvn", cl::init(false), cl::Hidden, + cl::desc("Run the NewGVN pass")); + +static cl::opt EnableLoopInterchange( + "enable-loopinterchange", cl::init(false), cl::Hidden, + cl::desc("Enable the experimental LoopInterchange Pass")); + +static cl::opt EnableUnrollAndJam("enable-unroll-and-jam", + cl::init(false), cl::Hidden, + cl::desc("Enable Unroll And Jam Pass")); + +static cl::opt EnableLoopFlatten("enable-loop-flatten", cl::init(false), + cl::Hidden, + cl::desc("Enable the LoopFlatten Pass")); + +static cl::opt + EnableDFAJumpThreading("enable-dfa-jump-thread", + cl::desc("Enable DFA jump threading"), + cl::init(false), cl::Hidden); + +static cl::opt + EnableHotColdSplit("hot-cold-split", + cl::desc("Enable hot-cold splitting pass")); + +static cl::opt EnableIROutliner("ir-outliner", cl::init(false), + cl::Hidden, + cl::desc("Enable ir outliner pass")); + +static cl::opt + DisablePreInliner("disable-preinline", cl::init(false), cl::Hidden, + cl::desc("Disable pre-instrumentation inliner")); + +static cl::opt PreInlineThreshold( + "preinline-threshold", cl::Hidden, cl::init(75), + cl::desc("Control the amount of inlining in pre-instrumentation inliner " + "(default = 75)")); + +static cl::opt + EnableGVNHoist("enable-gvn-hoist", + cl::desc("Enable the GVN hoisting pass (default = off)")); + +static cl::opt + EnableGVNSink("enable-gvn-sink", + cl::desc("Enable the GVN sinking pass (default = off)")); + +// This option is used in simplifying testing SampleFDO optimizations for +// profile loading. +static cl::opt + EnableCHR("enable-chr", cl::init(true), cl::Hidden, + cl::desc("Enable control height reduction optimization (CHR)")); + +static cl::opt FlattenedProfileUsed( + "flattened-profile-used", cl::init(false), cl::Hidden, + cl::desc("Indicate the sample profile being used is flattened, i.e., " + "no inline hierachy exists in the profile")); + +static cl::opt EnableOrderFileInstrumentation( + "enable-order-file-instrumentation", cl::init(false), cl::Hidden, + cl::desc("Enable order file instrumentation (default = off)")); + +static cl::opt + EnableMatrix("enable-matrix", cl::init(false), cl::Hidden, + cl::desc("Enable lowering of the matrix intrinsics")); + +static cl::opt EnableConstraintElimination( + "enable-constraint-elimination", cl::init(false), cl::Hidden, + cl::desc( + "Enable pass to eliminate conditions based on linear constraints")); + +static cl::opt EnableFunctionSpecialization( + "enable-function-specialization", cl::init(false), cl::Hidden, + cl::desc("Enable Function Specialization pass")); + +static cl::opt AttributorRun( + "attributor-enable", cl::Hidden, cl::init(AttributorRunOption::NONE), + cl::desc("Enable the attributor inter-procedural deduction pass"), + cl::values(clEnumValN(AttributorRunOption::ALL, "all", + "enable all attributor runs"), + clEnumValN(AttributorRunOption::MODULE, "module", + "enable module-wide attributor runs"), + clEnumValN(AttributorRunOption::CGSCC, "cgscc", + "enable call graph SCC attributor runs"), + clEnumValN(AttributorRunOption::NONE, "none", + "disable attributor runs"))); + PipelineTuningOptions::PipelineTuningOptions() { LoopInterleaving = true; LoopVectorization = true; @@ -206,27 +299,7 @@ namespace llvm { extern cl::opt MaxDevirtIterations; -extern cl::opt EnableConstraintElimination; -extern cl::opt EnableFunctionSpecialization; -extern cl::opt EnableGVNHoist; -extern cl::opt EnableGVNSink; -extern cl::opt EnableHotColdSplit; -extern cl::opt EnableIROutliner; -extern cl::opt EnableOrderFileInstrumentation; -extern cl::opt EnableCHR; -extern cl::opt EnableLoopInterchange; -extern cl::opt EnableUnrollAndJam; -extern cl::opt EnableLoopFlatten; -extern cl::opt EnableDFAJumpThreading; -extern cl::opt RunNewGVN; -extern cl::opt RunPartialInlining; -extern cl::opt ExtraVectorizerPasses; -extern cl::opt FlattenedProfileUsed; -extern cl::opt AttributorRun; extern cl::opt EnableKnowledgeRetention; -extern cl::opt EnableMatrix; -extern cl::opt DisablePreInliner; -extern cl::opt PreInlineThreshold; } // namespace llvm void PassBuilder::invokePeepholeEPCallbacks(FunctionPassManager &FPM, diff --git a/llvm/lib/Transforms/IPO/PassManagerBuilder.cpp b/llvm/lib/Transforms/IPO/PassManagerBuilder.cpp --- a/llvm/lib/Transforms/IPO/PassManagerBuilder.cpp +++ b/llvm/lib/Transforms/IPO/PassManagerBuilder.cpp @@ -43,129 +43,6 @@ using namespace llvm; -namespace llvm { -cl::opt RunPartialInlining("enable-partial-inlining", cl::Hidden, - cl::desc("Run Partial inlinining pass")); - -static cl::opt -UseGVNAfterVectorization("use-gvn-after-vectorization", - cl::init(false), cl::Hidden, - cl::desc("Run GVN instead of Early CSE after vectorization passes")); - -cl::opt ExtraVectorizerPasses( - "extra-vectorizer-passes", cl::init(false), cl::Hidden, - cl::desc("Run cleanup optimization passes after vectorization.")); - -static cl::opt -RunLoopRerolling("reroll-loops", cl::Hidden, - cl::desc("Run the loop rerolling pass")); - -cl::opt RunNewGVN("enable-newgvn", cl::init(false), cl::Hidden, - cl::desc("Run the NewGVN pass")); - -// Experimental option to use CFL-AA -static cl::opt<::CFLAAType> - UseCFLAA("use-cfl-aa", cl::init(::CFLAAType::None), cl::Hidden, - cl::desc("Enable the new, experimental CFL alias analysis"), - cl::values(clEnumValN(::CFLAAType::None, "none", "Disable CFL-AA"), - clEnumValN(::CFLAAType::Steensgaard, "steens", - "Enable unification-based CFL-AA"), - clEnumValN(::CFLAAType::Andersen, "anders", - "Enable inclusion-based CFL-AA"), - clEnumValN(::CFLAAType::Both, "both", - "Enable both variants of CFL-AA"))); - -cl::opt EnableLoopInterchange( - "enable-loopinterchange", cl::init(false), cl::Hidden, - cl::desc("Enable the experimental LoopInterchange Pass")); - -cl::opt EnableUnrollAndJam("enable-unroll-and-jam", cl::init(false), - cl::Hidden, - cl::desc("Enable Unroll And Jam Pass")); - -cl::opt EnableLoopFlatten("enable-loop-flatten", cl::init(false), - cl::Hidden, - cl::desc("Enable the LoopFlatten Pass")); - -cl::opt EnableDFAJumpThreading("enable-dfa-jump-thread", - cl::desc("Enable DFA jump threading."), - cl::init(false), cl::Hidden); - -cl::opt EnableHotColdSplit("hot-cold-split", - cl::desc("Enable hot-cold splitting pass")); - -cl::opt EnableIROutliner("ir-outliner", cl::init(false), cl::Hidden, - cl::desc("Enable ir outliner pass")); - -static cl::opt UseLoopVersioningLICM( - "enable-loop-versioning-licm", cl::init(false), cl::Hidden, - cl::desc("Enable the experimental Loop Versioning LICM pass")); - -cl::opt - DisablePreInliner("disable-preinline", cl::init(false), cl::Hidden, - cl::desc("Disable pre-instrumentation inliner")); - -cl::opt PreInlineThreshold( - "preinline-threshold", cl::Hidden, cl::init(75), - cl::desc("Control the amount of inlining in pre-instrumentation inliner " - "(default = 75)")); - -cl::opt - EnableGVNHoist("enable-gvn-hoist", - cl::desc("Enable the GVN hoisting pass (default = off)")); - -static cl::opt - DisableLibCallsShrinkWrap("disable-libcalls-shrinkwrap", cl::init(false), - cl::Hidden, - cl::desc("Disable shrink-wrap library calls")); - -cl::opt - EnableGVNSink("enable-gvn-sink", - cl::desc("Enable the GVN sinking pass (default = off)")); - -// This option is used in simplifying testing SampleFDO optimizations for -// profile loading. -cl::opt - EnableCHR("enable-chr", cl::init(true), cl::Hidden, - cl::desc("Enable control height reduction optimization (CHR)")); - -cl::opt FlattenedProfileUsed( - "flattened-profile-used", cl::init(false), cl::Hidden, - cl::desc("Indicate the sample profile being used is flattened, i.e., " - "no inline hierachy exists in the profile. ")); - -cl::opt EnableOrderFileInstrumentation( - "enable-order-file-instrumentation", cl::init(false), cl::Hidden, - cl::desc("Enable order file instrumentation (default = off)")); - -cl::opt EnableMatrix( - "enable-matrix", cl::init(false), cl::Hidden, - cl::desc("Enable lowering of the matrix intrinsics")); - -cl::opt EnableConstraintElimination( - "enable-constraint-elimination", cl::init(false), cl::Hidden, - cl::desc( - "Enable pass to eliminate conditions based on linear constraints.")); - -cl::opt EnableFunctionSpecialization( - "enable-function-specialization", cl::init(false), cl::Hidden, - cl::desc("Enable Function Specialization pass")); - -cl::opt AttributorRun( - "attributor-enable", cl::Hidden, cl::init(AttributorRunOption::NONE), - cl::desc("Enable the attributor inter-procedural deduction pass."), - cl::values(clEnumValN(AttributorRunOption::ALL, "all", - "enable all attributor runs"), - clEnumValN(AttributorRunOption::MODULE, "module", - "enable module-wide attributor runs"), - clEnumValN(AttributorRunOption::CGSCC, "cgscc", - "enable call graph SCC attributor runs"), - clEnumValN(AttributorRunOption::NONE, "none", - "disable attributor runs"))); - -extern cl::opt EnableKnowledgeRetention; -} // namespace llvm - PassManagerBuilder::PassManagerBuilder() { OptLevel = 2; SizeLevel = 0; @@ -175,8 +52,6 @@ SLPVectorize = false; LoopVectorize = true; LoopsInterleaved = true; - RerollLoops = RunLoopRerolling; - NewGVN = RunNewGVN; LicmMssaOptCap = SetLicmMssaOptCap; LicmMssaNoAccForPromotionCap = SetLicmMssaNoAccForPromotionCap; DisableGVNLoadPRE = false; @@ -253,21 +128,6 @@ void PassManagerBuilder::addInitialAliasAnalysisPasses( legacy::PassManagerBase &PM) const { - switch (UseCFLAA) { - case ::CFLAAType::Steensgaard: - PM.add(createCFLSteensAAWrapperPass()); - break; - case ::CFLAAType::Andersen: - PM.add(createCFLAndersAAWrapperPass()); - break; - case ::CFLAAType::Both: - PM.add(createCFLSteensAAWrapperPass()); - PM.add(createCFLAndersAAWrapperPass()); - break; - default: - break; - } - // Add TypeBasedAliasAnalysis before BasicAliasAnalysis so that // BasicAliasAnalysis wins if they disagree. This is intended to help // support "obvious" type-punning idioms. @@ -283,13 +143,6 @@ if (LibraryInfo) FPM.add(new TargetLibraryInfoWrapperPass(*LibraryInfo)); - // The backends do not handle matrix intrinsics currently. - // Make sure they are also lowered in O0. - // FIXME: A lightweight version of the pass should run in the backend - // pipeline on demand. - if (EnableMatrix && OptLevel == 0) - FPM.add(createLowerMatrixIntrinsicsMinimalPass()); - if (OptLevel == 0) return; addInitialAliasAnalysisPasses(FPM); @@ -309,21 +162,6 @@ assert(OptLevel >= 1 && "Calling function optimizer with no optimization level!"); MPM.add(createSROAPass()); MPM.add(createEarlyCSEPass(true /* Enable mem-ssa. */)); // Catch trivial redundancies - if (EnableKnowledgeRetention) - MPM.add(createAssumeSimplifyPass()); - - if (OptLevel > 1) { - if (EnableGVNHoist) - MPM.add(createGVNHoistPass()); - if (EnableGVNSink) { - MPM.add(createGVNSinkPass()); - MPM.add(createCFGSimplificationPass( - SimplifyCFGOptions().convertSwitchRangeToICmp(true))); - } - } - - if (EnableConstraintElimination) - MPM.add(createConstraintEliminationPass()); if (OptLevel > 1) { // Speculative execution if the target has divergent branches; otherwise nop. @@ -339,7 +177,7 @@ if (OptLevel > 2) MPM.add(createAggressiveInstCombinerPass()); MPM.add(createInstructionCombiningPass()); - if (SizeLevel == 0 && !DisableLibCallsShrinkWrap) + if (SizeLevel == 0) MPM.add(createLibCallsShrinkWrapPass()); addExtensionsToPM(EP_Peephole, MPM); @@ -351,11 +189,6 @@ true))); // Merge & remove BBs MPM.add(createReassociatePass()); // Reassociate expressions - // The matrix extension can introduce large vector operations early, which can - // benefit from running vector-combine early on. - if (EnableMatrix) - MPM.add(createVectorCombinePass()); - // Begin the loop pass pipeline. // The simple loop unswitch pass relies on separate cleanup passes. Schedule @@ -385,18 +218,11 @@ SimplifyCFGOptions().convertSwitchRangeToICmp(true))); MPM.add(createInstructionCombiningPass()); // We resume loop passes creating a second loop pipeline here. - if (EnableLoopFlatten) { - MPM.add(createLoopFlattenPass()); // Flatten loops - MPM.add(createLoopSimplifyCFGPass()); - } MPM.add(createLoopIdiomPass()); // Recognize idioms like memset. MPM.add(createIndVarSimplifyPass()); // Canonicalize indvars addExtensionsToPM(EP_LateLoopOptimizations, MPM); MPM.add(createLoopDeletionPass()); // Delete dead loops - if (EnableLoopInterchange) - MPM.add(createLoopInterchangePass()); // Interchange loops - // Unroll small loops and perform peeling. MPM.add(createSimpleLoopUnrollPass(OptLevel, DisableUnrollLoops, ForgetAllSCEVInLoopUnroll)); @@ -413,9 +239,6 @@ } MPM.add(createSCCPPass()); // Constant prop with SCCP - if (EnableConstraintElimination) - MPM.add(createConstraintEliminationPass()); - // Delete dead bit computations (instcombine runs after to fold away the dead // computations, and then ADCE will run later to exploit any new DCE // opportunities that creates). @@ -426,9 +249,6 @@ MPM.add(createInstructionCombiningPass()); addExtensionsToPM(EP_Peephole, MPM); if (OptLevel > 1) { - if (EnableDFAJumpThreading && SizeLevel == 0) - MPM.add(createDFAJumpThreadingPass()); - MPM.add(createJumpThreadingPass()); // Thread jumps MPM.add(createCorrelatedValuePropagationPass()); } @@ -444,9 +264,6 @@ addExtensionsToPM(EP_ScalarOptimizerLate, MPM); - if (RerollLoops) - MPM.add(createLoopRerollPass()); - // Merge & remove BBs and sink & hoist common instructions. MPM.add(createCFGSimplificationPass( SimplifyCFGOptions().hoistCommonInsts(true).sinkCommonInsts(true))); @@ -468,9 +285,6 @@ // FIXME: It would be really good to use a loop-integrated instruction // combiner for cleanup here so that the unrolling and LICM can be pipelined // across the loop nests. - // We do UnrollAndJam in a separate LPM to ensure it happens before unroll - if (EnableUnrollAndJam && !DisableUnrollLoops) - PM.add(createLoopUnrollAndJamPass(OptLevel)); PM.add(createLoopUnrollPass(OptLevel, DisableUnrollLoops, ForgetAllSCEVInLoopUnroll)); PM.add(createWarnMissedTransformationsPass()); @@ -484,24 +298,6 @@ // Cleanup after the loop optimization passes. PM.add(createInstructionCombiningPass()); - if (OptLevel > 1 && ExtraVectorizerPasses) { - // At higher optimization levels, try to clean up any runtime overlap and - // alignment checks inserted by the vectorizer. We want to track correlated - // runtime checks for two inner loops in the same outer loop, fold any - // common computations, hoist loop-invariant aspects out of any outer loop, - // and unswitch the runtime checks if possible. Once hoisted, we may have - // dead (or speculatable) control flows or more combining opportunities. - PM.add(createEarlyCSEPass()); - PM.add(createCorrelatedValuePropagationPass()); - PM.add(createInstructionCombiningPass()); - PM.add(createLICMPass(LicmMssaOptCap, LicmMssaNoAccForPromotionCap, - /*AllowSpeculation=*/true)); - PM.add(createSimpleLoopUnswitchLegacyPass()); - PM.add(createCFGSimplificationPass( - SimplifyCFGOptions().convertSwitchRangeToICmp(true))); - PM.add(createInstructionCombiningPass()); - } - // Now that we've formed fast to execute loop structures, we do further // optimizations. These are run afterward as they might block doing complex // analyses and transforms such as what are needed for loop vectorization. @@ -528,8 +324,6 @@ // Optimize parallel scalar instruction chains into SIMD instructions. if (SLPVectorize) { PM.add(createSLPVectorizerPass()); - if (OptLevel > 1 && ExtraVectorizerPasses) - PM.add(createEarlyCSEPass()); } // Enhance/cleanup vector code. @@ -539,13 +333,6 @@ addExtensionsToPM(EP_Peephole, PM); PM.add(createInstructionCombiningPass()); - if (EnableUnrollAndJam && !DisableUnrollLoops) { - // Unroll and Jam. We do this before unroll but need to be in a separate - // loop pass manager in order for the outer loop to be processed by - // unroll and jam before the inner loop is unrolled. - PM.add(createLoopUnrollAndJamPass(OptLevel)); - } - // Unroll small loops PM.add(createLoopUnrollPass(OptLevel, DisableUnrollLoops, ForgetAllSCEVInLoopUnroll)); @@ -613,19 +400,11 @@ // Infer attributes about declarations if possible. MPM.add(createInferFunctionAttrsLegacyPass()); - // Infer attributes on declarations, call sites, arguments, etc. - if (AttributorRun & AttributorRunOption::MODULE) - MPM.add(createAttributorLegacyPass()); - addExtensionsToPM(EP_ModuleOptimizerEarly, MPM); if (OptLevel > 2) MPM.add(createCallSiteSplittingPass()); - // Propage constant function arguments by specializing the functions. - if (OptLevel > 2 && EnableFunctionSpecialization) - MPM.add(createFunctionSpecializationPass()); - MPM.add(createIPSCCPPass()); // IP SCCP MPM.add(createCalledValuePropagationPass()); @@ -654,10 +433,6 @@ RunInliner = true; } - // Infer attributes on declarations, call sites, arguments, etc. for an SCC. - if (AttributorRun & AttributorRunOption::CGSCC) - MPM.add(createAttributorCGSCCLegacyPass()); - // Try to perform OpenMP specific optimizations. This is a (quick!) no-op if // there are no OpenMP runtime calls present in the module. if (OptLevel > 1) @@ -673,9 +448,6 @@ // we must insert a no-op module pass to reset the pass manager. MPM.add(createBarrierNoopPass()); - if (RunPartialInlining) - MPM.add(createPartialInliningPass()); - if (OptLevel > 1) // Remove avail extern fns and globals definitions if we aren't // compiling an object file for later LTO. For LTO we want to preserve @@ -701,17 +473,6 @@ MPM.add(createGlobalDCEPass()); } - // Scheduling LoopVersioningLICM when inlining is over, because after that - // we may see more accurate aliasing. Reason to run this late is that too - // early versioning may prevent further inlining due to increase of code - // size. By placing it just after inlining other optimizations which runs - // later might get benefit of no-alias assumption in clone loop. - if (UseLoopVersioningLICM) { - MPM.add(createLoopVersioningLICMPass()); // Do LoopVersioningLICM - MPM.add(createLICMPass(LicmMssaOptCap, LicmMssaNoAccForPromotionCap, - /*AllowSpeculation=*/true)); - } - // We add a fresh GlobalsModRef run at this point. This is particularly // useful as the above will have inlined, DCE'ed, and function-attr // propagated everything. We should at this point have a reasonably minimal @@ -732,14 +493,6 @@ MPM.add(createFloat2IntPass()); MPM.add(createLowerConstantIntrinsicsPass()); - if (EnableMatrix) { - MPM.add(createLowerMatrixIntrinsicsPass()); - // CSE the pointer arithmetic of the column vectors. This allows alias - // analysis to establish no-aliasing between loads and stores of different - // columns of the same matrix. - MPM.add(createEarlyCSEPass(false)); - } - addExtensionsToPM(EP_VectorizerStart, MPM); // Re-rotate loops in all our loop nests. These may have fallout out of @@ -765,14 +518,6 @@ MPM.add(createConstantMergePass()); // Merge dup global constants } - // See comment in the new PM for justification of scheduling splitting at - // this stage (\ref buildModuleSimplificationPipeline). - if (EnableHotColdSplit) - MPM.add(createHotColdSplittingPass()); - - if (EnableIROutliner) - MPM.add(createIROutlinerPass()); - if (MergeFunctions) MPM.add(createMergeFunctionsPass());