diff options
author | Dehao Chen <dehao@google.com> | 2017-07-07 21:01:00 +0000 |
---|---|---|
committer | Dehao Chen <dehao@google.com> | 2017-07-07 21:01:00 +0000 |
commit | b1f864a94769b7d0fd8f724cc6dbfe5195683d63 (patch) | |
tree | 74e49c46ec14c8c20d2c6085afd9a2c5e23d7884 | |
parent | 59bf0ba31835bd2cad26df09f3a6724ab0869cd1 (diff) |
Increase the import-threshold for crtical functions.
Summary: For interative sample-pgo, if a hot call site is inlined in the profiling binary, we should inline it in before profile annotation in the backend. Before that, the compile phase first collects all GUIDs that needs to be imported and creates virtual "hot" call edge in the summary. However, "hot" is not good enough to guarantee the callsites get inlined. This patch introduces "critical" call edge, and assign much higher importing threshold for those edges.
Reviewers: tejohnson
Reviewed By: tejohnson
Subscribers: sanjoy, mehdi_amini, llvm-commits, eraman
Differential Revision: https://reviews.llvm.org/D35096
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@307439 91177308-0d34-0410-b5e6-96231b3b80d8
5 files changed, 18 insertions, 4 deletions
diff --git a/include/llvm/IR/ModuleSummaryIndex.h b/include/llvm/IR/ModuleSummaryIndex.h index b43d5886586..c043c159ac1 100644 --- a/include/llvm/IR/ModuleSummaryIndex.h +++ b/include/llvm/IR/ModuleSummaryIndex.h @@ -47,7 +47,13 @@ template <typename T> struct MappingTraits; /// \brief Class to accumulate and hold information about a callee. struct CalleeInfo { - enum class HotnessType : uint8_t { Unknown = 0, Cold = 1, None = 2, Hot = 3 }; + enum class HotnessType : uint8_t { + Unknown = 0, + Cold = 1, + None = 2, + Hot = 3, + Critical = 4 + }; HotnessType Hotness = HotnessType::Unknown; CalleeInfo() = default; diff --git a/lib/Analysis/ModuleSummaryAnalysis.cpp b/lib/Analysis/ModuleSummaryAnalysis.cpp index 095647e1bd2..e9e354ebb88 100644 --- a/lib/Analysis/ModuleSummaryAnalysis.cpp +++ b/lib/Analysis/ModuleSummaryAnalysis.cpp @@ -266,7 +266,7 @@ computeFunctionSummary(ModuleSummaryIndex &Index, const Module &M, // sample PGO, to enable the same inlines as the profiled optimized binary. for (auto &I : F.getImportGUIDs()) CallGraphEdges[Index.getOrInsertValueInfo(I)].updateHotness( - CalleeInfo::HotnessType::Hot); + CalleeInfo::HotnessType::Critical); bool NonRenamableLocal = isNonRenamableLocal(F); bool NotEligibleForImport = diff --git a/lib/Transforms/IPO/FunctionImport.cpp b/lib/Transforms/IPO/FunctionImport.cpp index c89e0e8e1bf..233a36d2bc5 100644 --- a/lib/Transforms/IPO/FunctionImport.cpp +++ b/lib/Transforms/IPO/FunctionImport.cpp @@ -64,6 +64,12 @@ static cl::opt<float> ImportHotMultiplier( "import-hot-multiplier", cl::init(3.0), cl::Hidden, cl::value_desc("x"), cl::desc("Multiply the `import-instr-limit` threshold for hot callsites")); +static cl::opt<float> ImportCriticalMultiplier( + "import-critical-multiplier", cl::init(100.0), cl::Hidden, + cl::value_desc("x"), + cl::desc( + "Multiply the `import-instr-limit` threshold for critical callsites")); + // FIXME: This multiplier was not really tuned up. static cl::opt<float> ImportColdMultiplier( "import-cold-multiplier", cl::init(0), cl::Hidden, cl::value_desc("N"), @@ -207,6 +213,8 @@ static void computeImportForFunction( return ImportHotMultiplier; if (Hotness == CalleeInfo::HotnessType::Cold) return ImportColdMultiplier; + if (Hotness == CalleeInfo::HotnessType::Critical) + return ImportCriticalMultiplier; return 1.0; }; diff --git a/test/Bitcode/thinlto-function-summary-callgraph-profile-summary.ll b/test/Bitcode/thinlto-function-summary-callgraph-profile-summary.ll index a4d259add60..86766f19468 100644 --- a/test/Bitcode/thinlto-function-summary-callgraph-profile-summary.ll +++ b/test/Bitcode/thinlto-function-summary-callgraph-profile-summary.ll @@ -29,7 +29,7 @@ ; CHECK-NEXT: <VERSION ; CHECK-NEXT: <VALUE_GUID op0=25 op1=123/> ; op4=hot1 op6=cold op8=hot2 op10=hot4 op12=none1 op14=hot3 op16=none2 op18=none3 op20=123 -; CHECK-NEXT: <PERMODULE_PROFILE {{.*}} op4=1 op5=3 op6=5 op7=1 op8=2 op9=3 op10=4 op11=1 op12=6 op13=2 op14=3 op15=3 op16=7 op17=2 op18=8 op19=2 op20=25 op21=3/> +; CHECK-NEXT: <PERMODULE_PROFILE {{.*}} op4=1 op5=3 op6=5 op7=1 op8=2 op9=3 op10=4 op11=1 op12=6 op13=2 op14=3 op15=3 op16=7 op17=2 op18=8 op19=2 op20=25 op21=4/> ; CHECK-NEXT: </GLOBALVAL_SUMMARY_BLOCK> ; CHECK: <STRTAB_BLOCK diff --git a/test/Bitcode/thinlto-function-summary-callgraph-sample-profile-summary.ll b/test/Bitcode/thinlto-function-summary-callgraph-sample-profile-summary.ll index b62090efe20..09a6bbcb51d 100644 --- a/test/Bitcode/thinlto-function-summary-callgraph-sample-profile-summary.ll +++ b/test/Bitcode/thinlto-function-summary-callgraph-sample-profile-summary.ll @@ -29,7 +29,7 @@ ; CHECK-NEXT: <VERSION ; CHECK-NEXT: <VALUE_GUID op0=25 op1=123/> ; op4=hot1 op6=cold op8=hot2 op10=hot4 op12=none1 op14=hot3 op16=none2 op18=none3 op20=123 -; CHECK-NEXT: <PERMODULE_PROFILE {{.*}} op4=1 op5=3 op6=5 op7=1 op8=2 op9=3 op10=4 op11=3 op12=6 op13=2 op14=3 op15=3 op16=7 op17=2 op18=8 op19=2 op20=25 op21=3/> +; CHECK-NEXT: <PERMODULE_PROFILE {{.*}} op4=1 op5=3 op6=5 op7=1 op8=2 op9=3 op10=4 op11=3 op12=6 op13=2 op14=3 op15=3 op16=7 op17=2 op18=8 op19=2 op20=25 op21=4/> ; CHECK-NEXT: </GLOBALVAL_SUMMARY_BLOCK> ; CHECK: <STRTAB_BLOCK |