From 577e58bcc754cea2ae0358a5359ac5cc48d8e9d0 Mon Sep 17 00:00:00 2001 From: Wenlei He Date: Sat, 15 Aug 2020 14:52:10 -0700 Subject: [PATCH] [InlineAdvisor] New inliner advisor to replay inlining from optimization remarks This change added a new inline advisor that takes optimization remarks from previous inlining as input, and provides the decision as advice so current inlining can replay inline decisions of a different compilation. Dwarf inline stack with line and discriminator is used as anchor for call sites including call context. The change can be useful for Inliner tuning as it provides a channel to allow external input for tweaking inline decisions. Existing alternatives like alwaysinline attribute is per-function, not per-callsite. Per-callsite inline intrinsic can be another solution (not yet existing), but it's intrusive to implement and also does not differentiate call context. A switch -sample-profile-inline-replay= is added to hook up the new inline advisor with SampleProfileLoader's inline decision for replay. Since SampleProfileLoader does top-down inlining, inline decision can be specialized for each call context, hence we should be able to replay inlining accurately. However with a bottom-up inliner like CGSCC inlining, the replay can be limited due to lack of specialization for different call context. Apart from that limitation, the new inline advisor can still be used by regular CGSCC inliner later if needed for tuning purpose. This is a resubmit of https://reviews.llvm.org/D83743 --- llvm/include/llvm/Analysis/InlineAdvisor.h | 3 + llvm/include/llvm/Analysis/ReplayInlineAdvisor.h | 37 +++++++ llvm/lib/Analysis/CMakeLists.txt | 1 + llvm/lib/Analysis/InlineAdvisor.cpp | 28 ++++- llvm/lib/Analysis/ReplayInlineAdvisor.cpp | 60 ++++++++++ llvm/lib/Transforms/IPO/SampleProfile.cpp | 37 ++++++- .../SampleProfile/Inputs/inline-replay.txt | 2 + .../test/Transforms/SampleProfile/inline-replay.ll | 122 +++++++++++++++++++++ 8 files changed, 285 insertions(+), 5 deletions(-) create mode 100644 llvm/include/llvm/Analysis/ReplayInlineAdvisor.h create mode 100644 llvm/lib/Analysis/ReplayInlineAdvisor.cpp create mode 100644 llvm/test/Transforms/SampleProfile/Inputs/inline-replay.txt create mode 100644 llvm/test/Transforms/SampleProfile/inline-replay.ll diff --git a/llvm/include/llvm/Analysis/InlineAdvisor.h b/llvm/include/llvm/Analysis/InlineAdvisor.h index c0ea40e..f57a609 100644 --- a/llvm/include/llvm/Analysis/InlineAdvisor.h +++ b/llvm/include/llvm/Analysis/InlineAdvisor.h @@ -233,6 +233,9 @@ void emitInlinedInto(OptimizationRemarkEmitter &ORE, DebugLoc DLoc, bool ForProfileContext = false, const char *PassName = nullptr); +/// get call site location as string +std::string getCallSiteLocation(DebugLoc DLoc); + /// Add location info to ORE message. void addLocationToRemarks(OptimizationRemark &Remark, DebugLoc DLoc); diff --git a/llvm/include/llvm/Analysis/ReplayInlineAdvisor.h b/llvm/include/llvm/Analysis/ReplayInlineAdvisor.h new file mode 100644 index 0000000..19e9079 --- /dev/null +++ b/llvm/include/llvm/Analysis/ReplayInlineAdvisor.h @@ -0,0 +1,37 @@ +//===- ReplayInlineAdvisor.h - Replay Inline Advisor interface -*- C++ --*-===// +// +// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. +// See https://llvm.org/LICENSE.txt for license information. +// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception +// +//===----------------------------------------------------------------------===// +// +#ifndef LLVM_ANALYSIS_REPLAYINLINEADVISOR_H +#define LLVM_ANALYSIS_REPLAYINLINEADVISOR_H + +#include "llvm/ADT/StringSet.h" +#include "llvm/Analysis/InlineAdvisor.h" +#include "llvm/IR/LLVMContext.h" + +namespace llvm { +class BasicBlock; +class CallBase; +class Function; +class Module; +class OptimizationRemarkEmitter; + +/// Replay inline advisor that uses optimization remarks from inlining of +/// previous build to guide current inlining. This is useful for inliner tuning. +class ReplayInlineAdvisor : public InlineAdvisor { +public: + ReplayInlineAdvisor(FunctionAnalysisManager &FAM, LLVMContext &Context, + StringRef RemarksFile); + std::unique_ptr getAdvice(CallBase &CB) override; + bool areReplayRemarksLoaded() const { return HasReplayRemarks; } + +private: + StringSet<> InlineSitesFromRemarks; + bool HasReplayRemarks = false; +}; +} // namespace llvm +#endif // LLVM_ANALYSIS_REPLAYINLINEADVISOR_H diff --git a/llvm/lib/Analysis/CMakeLists.txt b/llvm/lib/Analysis/CMakeLists.txt index 8bb3769..f50439b 100644 --- a/llvm/lib/Analysis/CMakeLists.txt +++ b/llvm/lib/Analysis/CMakeLists.txt @@ -102,6 +102,7 @@ add_llvm_component_library(LLVMAnalysis RegionPass.cpp RegionPrinter.cpp ReleaseModeModelRunner.cpp + ReplayInlineAdvisor.cpp ScalarEvolution.cpp ScalarEvolutionAliasAnalysis.cpp ScalarEvolutionDivision.cpp diff --git a/llvm/lib/Analysis/InlineAdvisor.cpp b/llvm/lib/Analysis/InlineAdvisor.cpp index ee92450..960141d 100644 --- a/llvm/lib/Analysis/InlineAdvisor.cpp +++ b/llvm/lib/Analysis/InlineAdvisor.cpp @@ -371,9 +371,35 @@ llvm::shouldInline(CallBase &CB, return IC; } +std::string llvm::getCallSiteLocation(DebugLoc DLoc) { + std::ostringstream CallSiteLoc; + bool First = true; + for (DILocation *DIL = DLoc.get(); DIL; DIL = DIL->getInlinedAt()) { + if (!First) + CallSiteLoc << " @ "; + // Note that negative line offset is actually possible, but we use + // unsigned int to match line offset representation in remarks so + // it's directly consumable by relay advisor. + uint32_t Offset = + DIL->getLine() - DIL->getScope()->getSubprogram()->getLine(); + uint32_t Discriminator = DIL->getBaseDiscriminator(); + StringRef Name = DIL->getScope()->getSubprogram()->getLinkageName(); + if (Name.empty()) + Name = DIL->getScope()->getSubprogram()->getName(); + CallSiteLoc << Name.str() << ":" << llvm::utostr(Offset); + if (Discriminator) { + CallSiteLoc << "." << llvm::utostr(Discriminator); + } + First = false; + } + + return CallSiteLoc.str(); +} + void llvm::addLocationToRemarks(OptimizationRemark &Remark, DebugLoc DLoc) { - if (!DLoc.get()) + if (!DLoc.get()) { return; + } bool First = true; Remark << " at callsite "; diff --git a/llvm/lib/Analysis/ReplayInlineAdvisor.cpp b/llvm/lib/Analysis/ReplayInlineAdvisor.cpp new file mode 100644 index 0000000..d2dd030 --- /dev/null +++ b/llvm/lib/Analysis/ReplayInlineAdvisor.cpp @@ -0,0 +1,60 @@ +//===- ReplayInlineAdvisor.cpp - Replay InlineAdvisor ---------------------===// +// +// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. +// See https://llvm.org/LICENSE.txt for license information. +// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception +// +//===----------------------------------------------------------------------===// +// +// This file implements ReplayInlineAdvisor that replays inline decision based +// on previous inline remarks from optimization remark log. +// +//===----------------------------------------------------------------------===// + +#include "llvm/Analysis/ReplayInlineAdvisor.h" +#include "llvm/IR/DebugInfoMetadata.h" +#include "llvm/IR/Instructions.h" +#include "llvm/Support/LineIterator.h" + +using namespace llvm; + +#define DEBUG_TYPE "inline-replay" + +ReplayInlineAdvisor::ReplayInlineAdvisor(FunctionAnalysisManager &FAM, + LLVMContext &Context, + StringRef RemarksFile) + : InlineAdvisor(FAM), HasReplayRemarks(false) { + auto BufferOrErr = MemoryBuffer::getFileOrSTDIN(RemarksFile); + std::error_code EC = BufferOrErr.getError(); + if (EC) { + Context.emitError("Could not open remarks file: " + EC.message()); + return; + } + + // Example for inline remarks to parse: + // _Z3subii inlined into main [details] at callsite sum:1 @ main:3.1 + // We use the callsite string after `at callsite` to replay inlining. + line_iterator LineIt(*BufferOrErr.get(), /*SkipBlanks=*/true); + for (; !LineIt.is_at_eof(); ++LineIt) { + StringRef Line = *LineIt; + auto Pair = Line.split(" at callsite "); + if (Pair.second.empty()) + continue; + InlineSitesFromRemarks.insert(Pair.second); + } + HasReplayRemarks = true; +} + +std::unique_ptr ReplayInlineAdvisor::getAdvice(CallBase &CB) { + assert(HasReplayRemarks); + + Function &Caller = *CB.getCaller(); + auto &ORE = FAM.getResult(Caller); + + if (InlineSitesFromRemarks.empty()) + return std::make_unique(this, CB, ORE, false); + + std::string CallSiteLoc = getCallSiteLocation(CB.getDebugLoc()); + bool InlineRecommended = InlineSitesFromRemarks.count(CallSiteLoc) > 0; + return std::make_unique(this, CB, ORE, InlineRecommended); +} diff --git a/llvm/lib/Transforms/IPO/SampleProfile.cpp b/llvm/lib/Transforms/IPO/SampleProfile.cpp index 95c4f1c..2c520a1 100644 --- a/llvm/lib/Transforms/IPO/SampleProfile.cpp +++ b/llvm/lib/Transforms/IPO/SampleProfile.cpp @@ -43,6 +43,7 @@ #include "llvm/Analysis/OptimizationRemarkEmitter.h" #include "llvm/Analysis/PostDominators.h" #include "llvm/Analysis/ProfileSummaryInfo.h" +#include "llvm/Analysis/ReplayInlineAdvisor.h" #include "llvm/Analysis/TargetLibraryInfo.h" #include "llvm/Analysis/TargetTransformInfo.h" #include "llvm/IR/BasicBlock.h" @@ -170,6 +171,13 @@ static cl::opt SampleColdCallSiteThreshold( "sample-profile-cold-inline-threshold", cl::Hidden, cl::init(45), cl::desc("Threshold for inlining cold callsites")); +static cl::opt ProfileInlineReplayFile( + "sample-profile-inline-replay", cl::init(""), cl::value_desc("filename"), + cl::desc( + "Optimization remarks file containing inline remarks to be replayed " + "by inlining from sample profile loader."), + cl::Hidden); + namespace { using BlockWeightMap = DenseMap; @@ -319,7 +327,7 @@ public: RemappingFilename(std::string(RemapName)), IsThinLTOPreLink(IsThinLTOPreLink) {} - bool doInitialization(Module &M); + bool doInitialization(Module &M, FunctionAnalysisManager *FAM = nullptr); bool runOnModule(Module &M, ModuleAnalysisManager *AM, ProfileSummaryInfo *_PSI, CallGraph *CG); @@ -473,6 +481,9 @@ protected: // overriden by -profile-sample-accurate or profile-sample-accurate // attribute. bool ProfAccForSymsInList; + + // External inline advisor used to replay inline decision from remarks. + std::unique_ptr ExternalInlineAdvisor; }; class SampleProfileLoaderLegacyPass : public ModulePass { @@ -897,6 +908,16 @@ SampleProfileLoader::findFunctionSamples(const Instruction &Inst) const { } bool SampleProfileLoader::inlineCallInstruction(CallBase &CB) { + if (ExternalInlineAdvisor) { + auto Advice = ExternalInlineAdvisor->getAdvice(CB); + if (!Advice->isInliningRecommended()) { + Advice->recordUnattemptedInlining(); + return false; + } + // Dummy record, we don't use it for replay. + Advice->recordInlining(); + } + Function *CalledFunction = CB.getCalledFunction(); assert(CalledFunction); DebugLoc DLoc = CB.getDebugLoc(); @@ -1006,7 +1027,7 @@ bool SampleProfileLoader::inlineHotFunctions( } } } - if (Hot) { + if (Hot || ExternalInlineAdvisor) { CIS.insert(CIS.begin(), AllCandidates.begin(), AllCandidates.end()); emitOptimizationRemarksForInlineCandidates(AllCandidates, F, true); } else { @@ -1829,7 +1850,8 @@ SampleProfileLoader::buildFunctionOrder(Module &M, CallGraph *CG) { return FunctionOrderList; } -bool SampleProfileLoader::doInitialization(Module &M) { +bool SampleProfileLoader::doInitialization(Module &M, + FunctionAnalysisManager *FAM) { auto &Ctx = M.getContext(); std::unique_ptr RemapReader; @@ -1854,6 +1876,13 @@ bool SampleProfileLoader::doInitialization(Module &M) { NamesInProfile.insert(NameTable->begin(), NameTable->end()); } + if (FAM && !ProfileInlineReplayFile.empty()) { + ExternalInlineAdvisor = std::make_unique( + *FAM, Ctx, ProfileInlineReplayFile); + if (!ExternalInlineAdvisor->areReplayRemarksLoaded()) + ExternalInlineAdvisor.reset(); + } + return true; } @@ -2006,7 +2035,7 @@ PreservedAnalyses SampleProfileLoaderPass::run(Module &M, : ProfileRemappingFileName, IsThinLTOPreLink, GetAssumptionCache, GetTTI, GetTLI); - if (!SampleLoader.doInitialization(M)) + if (!SampleLoader.doInitialization(M, &FAM)) return PreservedAnalyses::all(); ProfileSummaryInfo *PSI = &AM.getResult(M); diff --git a/llvm/test/Transforms/SampleProfile/Inputs/inline-replay.txt b/llvm/test/Transforms/SampleProfile/Inputs/inline-replay.txt new file mode 100644 index 0000000..6842845 --- /dev/null +++ b/llvm/test/Transforms/SampleProfile/Inputs/inline-replay.txt @@ -0,0 +1,2 @@ +remark: calls.cc:10:0: _Z3sumii inlined into main to match profiling context with (cost=45, threshold=337) at callsite main:3.1 +remark: calls.cc:4:0: _Z3subii inlined into main to match profiling context with (cost=-5, threshold=337) at callsite _Z3sumii:1 @ main:3.1 diff --git a/llvm/test/Transforms/SampleProfile/inline-replay.ll b/llvm/test/Transforms/SampleProfile/inline-replay.ll new file mode 100644 index 0000000..a3c4cff --- /dev/null +++ b/llvm/test/Transforms/SampleProfile/inline-replay.ll @@ -0,0 +1,122 @@ +;; Note that this needs new pass manager for now. Passing `-sample-profile-inline-replay` to legacy pass manager is a no-op. + +;; Check baseline inline decisions +; RUN: opt < %s -passes=sample-profile -sample-profile-file=%S/Inputs/inline-topdown.prof -sample-profile-merge-inlinee -sample-profile-top-down-load -pass-remarks=inline -S 2>&1 | FileCheck -check-prefix=DEFAULT %s + +;; Check replay inline decisions +; RUN: opt < %s -passes=sample-profile -sample-profile-file=%S/Inputs/inline-topdown.prof -sample-profile-inline-replay=%S/Inputs/inline-replay.txt -sample-profile-merge-inlinee -sample-profile-top-down-load -pass-remarks=inline -S 2>&1 | FileCheck -check-prefix=REPLAY %s + +@.str = private unnamed_addr constant [11 x i8] c"sum is %d\0A\00", align 1 + +define i32 @_Z3sumii(i32 %x, i32 %y) #0 !dbg !6 { +entry: + %x.addr = alloca i32, align 4 + %y.addr = alloca i32, align 4 + store i32 %x, i32* %x.addr, align 4 + store i32 %y, i32* %y.addr, align 4 + %tmp = load i32, i32* %x.addr, align 4, !dbg !8 + %tmp1 = load i32, i32* %y.addr, align 4, !dbg !8 + %add = add nsw i32 %tmp, %tmp1, !dbg !8 + %tmp2 = load i32, i32* %x.addr, align 4, !dbg !8 + %tmp3 = load i32, i32* %y.addr, align 4, !dbg !8 + %call = call i32 @_Z3subii(i32 %tmp2, i32 %tmp3), !dbg !8 + ret i32 %add, !dbg !8 +} + +define i32 @_Z3subii(i32 %x, i32 %y) #0 !dbg !9 { +entry: + %x.addr = alloca i32, align 4 + %y.addr = alloca i32, align 4 + store i32 %x, i32* %x.addr, align 4 + store i32 %y, i32* %y.addr, align 4 + %tmp = load i32, i32* %x.addr, align 4, !dbg !10 + %tmp1 = load i32, i32* %y.addr, align 4, !dbg !10 + %add = sub nsw i32 %tmp, %tmp1, !dbg !10 + ret i32 %add, !dbg !11 +} + +define i32 @main() #0 !dbg !12 { +entry: + %retval = alloca i32, align 4 + %s = alloca i32, align 4 + %i = alloca i32, align 4 + store i32 0, i32* %retval + store i32 0, i32* %i, align 4, !dbg !13 + br label %while.cond, !dbg !14 + +while.cond: ; preds = %if.end, %entry + %tmp = load i32, i32* %i, align 4, !dbg !15 + %inc = add nsw i32 %tmp, 1, !dbg !15 + store i32 %inc, i32* %i, align 4, !dbg !15 + %cmp = icmp slt i32 %tmp, 400000000, !dbg !15 + br i1 %cmp, label %while.body, label %while.end, !dbg !15 + +while.body: ; preds = %while.cond + %tmp1 = load i32, i32* %i, align 4, !dbg !17 + %cmp1 = icmp ne i32 %tmp1, 100, !dbg !17 + br i1 %cmp1, label %if.then, label %if.else, !dbg !17 + +if.then: ; preds = %while.body + %tmp2 = load i32, i32* %i, align 4, !dbg !19 + %tmp3 = load i32, i32* %s, align 4, !dbg !19 + %call = call i32 @_Z3sumii(i32 %tmp2, i32 %tmp3), !dbg !19 + store i32 %call, i32* %s, align 4, !dbg !19 + br label %if.end, !dbg !19 + +if.else: ; preds = %while.body + store i32 30, i32* %s, align 4, !dbg !21 + br label %if.end + +if.end: ; preds = %if.else, %if.then + br label %while.cond, !dbg !23 + +while.end: ; preds = %while.cond + %tmp4 = load i32, i32* %s, align 4, !dbg !25 + %call2 = call i32 (i8*, ...) @printf(i8* getelementptr inbounds ([11 x i8], [11 x i8]* @.str, i32 0, i32 0), i32 %tmp4), !dbg !25 + ret i32 0, !dbg !26 +} + +declare i32 @printf(i8*, ...) + +attributes #0 = { "use-sample-profile" } + +!llvm.dbg.cu = !{!0} +!llvm.module.flags = !{!3, !4} +!llvm.ident = !{!5} + +!0 = distinct !DICompileUnit(language: DW_LANG_C_plus_plus, file: !1, producer: "clang version 3.5 ", isOptimized: false, runtimeVersion: 0, emissionKind: NoDebug, enums: !2, retainedTypes: !2, globals: !2, imports: !2) +!1 = !DIFile(filename: "calls.cc", directory: ".") +!2 = !{} +!3 = !{i32 2, !"Dwarf Version", i32 4} +!4 = !{i32 1, !"Debug Info Version", i32 3} +!5 = !{!"clang version 3.5 "} +!6 = distinct !DISubprogram(name: "sum", linkageName: "_Z3sumii", scope: !1, file: !1, line: 3, type: !7, scopeLine: 3, virtualIndex: 6, flags: DIFlagPrototyped, spFlags: DISPFlagDefinition, unit: !0, retainedNodes: !2) +!7 = !DISubroutineType(types: !2) +!8 = !DILocation(line: 4, scope: !6) +!9 = distinct !DISubprogram(name: "sub", linkageName: "_Z3subii", scope: !1, file: !1, line: 20, type: !7, scopeLine: 20, virtualIndex: 6, flags: DIFlagPrototyped, spFlags: DISPFlagDefinition, unit: !0, retainedNodes: !2) +!10 = !DILocation(line: 20, scope: !9) +!11 = !DILocation(line: 21, scope: !9) +!12 = distinct !DISubprogram(name: "main", scope: !1, file: !1, line: 7, type: !7, scopeLine: 7, virtualIndex: 6, flags: DIFlagPrototyped, spFlags: DISPFlagDefinition, unit: !0, retainedNodes: !2) +!13 = !DILocation(line: 8, scope: !12) +!14 = !DILocation(line: 9, scope: !12) +!15 = !DILocation(line: 9, scope: !16) +!16 = !DILexicalBlockFile(scope: !12, file: !1, discriminator: 2) +!17 = !DILocation(line: 10, scope: !18) +!18 = distinct !DILexicalBlock(scope: !12, file: !1, line: 10) +!19 = !DILocation(line: 10, scope: !20) +!20 = !DILexicalBlockFile(scope: !18, file: !1, discriminator: 2) +!21 = !DILocation(line: 10, scope: !22) +!22 = !DILexicalBlockFile(scope: !18, file: !1, discriminator: 4) +!23 = !DILocation(line: 10, scope: !24) +!24 = !DILexicalBlockFile(scope: !18, file: !1, discriminator: 6) +!25 = !DILocation(line: 11, scope: !12) +!26 = !DILocation(line: 12, scope: !12) + + +; DEFAULT: _Z3sumii inlined into main +; DEFAULT: _Z3subii inlined into _Z3sumii +; DEFAULT-NOT: _Z3subii inlined into main + +; REPLAY: _Z3sumii inlined into main +; REPLAY: _Z3subii inlined into main +; REPLA-NOT: _Z3subii inlined into _Z3sumii -- 2.7.4