rebase

Created using spr 1.3.4

Author: aaupov

.github/CODEOWNERS

@@ -68,7 +68,7 @@
 /mlir/lib/Dialect/Linalg/Transforms/ElementwiseOpFusion.cpp @MaheshRavishankar @nicolasvasilache
 /mlir/lib/Dialect/Linalg/Transforms/DataLayoutPropagation.cpp @hanhanW @nicolasvasilache
 /mlir/lib/Dialect/Linalg/Transforms/Transforms.cpp @dcaballe @hanhanW @nicolasvasilache
-/mlir/lib/Dialect/Linalg/Transforms/Vectorization.cpp @banach-space @dcaballe @hanhanW @nicolasvasilache
+/mlir/lib/Dialect/Linalg/Transforms/Vectorization.cpp @banach-space @dcaballe @hanhanW @nicolasvasilache @Groverkss
 
 # MemRef Dialect in MLIR.
 /mlir/lib/Dialect/MemRef/Transforms/EmulateNarrowType.cpp @MaheshRavishankar @nicolasvasilache
@@ -82,9 +82,9 @@
 /mlir/**/*VectorToSCF* @banach-space @dcaballe @matthias-springer @nicolasvasilache
 /mlir/**/*VectorToLLVM* @banach-space @dcaballe @nicolasvasilache
 /mlir/**/*X86Vector* @aartbik @dcaballe @nicolasvasilache
-/mlir/include/mlir/Dialect/Vector @banach-space @dcaballe @nicolasvasilache
+/mlir/include/mlir/Dialect/Vector @banach-space @dcaballe @nicolasvasilache @Groverkss
 /mlir/include/mlir/Dialect/Vector/IR @kuhar
-/mlir/lib/Dialect/Vector @banach-space @dcaballe @nicolasvasilache
+/mlir/lib/Dialect/Vector @banach-space @dcaballe @nicolasvasilache @Groverkss
 /mlir/lib/Dialect/Vector/Transforms/* @banach-space @dcaballe @hanhanW @nicolasvasilache
 /mlir/lib/Dialect/Vector/Transforms/VectorEmulateNarrowType.cpp @banach-space @dcaballe @MaheshRavishankar @nicolasvasilache
 /mlir/**/*EmulateNarrowType* @dcaballe @hanhanW
@@ -141,7 +141,7 @@
 /clang/tools/clang-installapi/ @cyndyishida
 
 # ExtractAPI
-/clang/**/ExtractAPI @daniel-grumberg
+/clang/**/ExtractAPI @daniel-grumberg @QuietMisdreavus
 
 # DWARFLinker, dwarfutil, dsymutil
 /llvm/**/DWARFLinker/ @JDevlieghere

.github/workflows/containers/github-action-ci/stage1.Dockerfile

@@ -2,7 +2,7 @@ FROM docker.io/library/ubuntu:22.04 as base
 ENV LLVM_SYSROOT=/opt/llvm
 
 FROM base as stage1-toolchain
-ENV LLVM_VERSION=18.1.8
+ENV LLVM_VERSION=19.1.2
 
 RUN apt-get update && \
     apt-get install -y \

.github/workflows/libcxx-build-and-test.yaml

@@ -49,7 +49,8 @@ env:
 jobs:
   stage1:
     if: github.repository_owner == 'llvm'
-    runs-on: libcxx-runners-8-set
+    runs-on: libcxx-runners-set
+    container: ghcr.io/libcxx/actions-builder:testing-2024-09-21
     continue-on-error: false
     strategy:
       fail-fast: false
@@ -79,12 +80,14 @@ jobs:
           path: |
             **/test-results.xml
             **/*.abilist
+            **/CMakeConfigureLog.yaml
             **/CMakeError.log
             **/CMakeOutput.log
             **/crash_diagnostics/*
   stage2:
     if: github.repository_owner == 'llvm'
-    runs-on: libcxx-runners-8-set
+    runs-on: libcxx-runners-set
+    container: ghcr.io/libcxx/actions-builder:testing-2024-09-21
     needs: [ stage1 ]
     continue-on-error: false
     strategy:
@@ -123,6 +126,7 @@ jobs:
           path: |
             **/test-results.xml
             **/*.abilist
+            **/CMakeConfigureLog.yaml
             **/CMakeError.log
             **/CMakeOutput.log
             **/crash_diagnostics/*
@@ -155,25 +159,23 @@ jobs:
           'generic-no-rtti',
           'generic-optimized-speed',
           'generic-static',
-          # TODO Find a better place for the benchmark and bootstrapping builds to live. They're either very expensive
-          # or don't provide much value since the benchmark run results are too noise on the bots.
-          'benchmarks',
           'bootstrapping-build'
         ]
-        machine: [ 'libcxx-runners-8-set' ]
+        machine: [ 'libcxx-runners-set' ]
         include:
         - config: 'generic-cxx26'
-          machine: libcxx-runners-8-set
+          machine: libcxx-runners-set
         - config: 'generic-asan'
-          machine: libcxx-runners-8-set
+          machine: libcxx-runners-set
         - config: 'generic-tsan'
-          machine: libcxx-runners-8-set
+          machine: libcxx-runners-set
         - config: 'generic-ubsan'
-          machine: libcxx-runners-8-set
+          machine: libcxx-runners-set
         # Use a larger machine for MSAN to avoid timeout and memory allocation issues.
         - config: 'generic-msan'
-          machine: libcxx-runners-8-set
+          machine: libcxx-runners-set
     runs-on: ${{ matrix.machine }}
+    container: ghcr.io/libcxx/actions-builder:testing-2024-09-21
     steps:
       - uses: actions/checkout@v4
       - name: ${{ matrix.config }}
@@ -188,6 +190,7 @@ jobs:
           path: |
             **/test-results.xml
             **/*.abilist
+            **/CMakeConfigureLog.yaml
             **/CMakeError.log
             **/CMakeOutput.log
             **/crash_diagnostics/*
@@ -230,6 +233,7 @@ jobs:
           path: |
             **/test-results.xml
             **/*.abilist
+            **/CMakeConfigureLog.yaml
             **/CMakeError.log
             **/CMakeOutput.log
             **/crash_diagnostics/*

bolt/include/bolt/Core/BinaryContext.h

@@ -229,12 +229,6 @@ class BinaryContext {
   /// Functions injected by BOLT
   std::vector<BinaryFunction *> InjectedBinaryFunctions;
 
-  /// Thunk functions.
-  std::vector<BinaryFunction *> ThunkBinaryFunctions;
-
-  /// Function that precedes thunks in the binary.
-  const BinaryFunction *ThunkLocation{nullptr};
-
   /// Jump tables for all functions mapped by address.
   std::map<uint64_t, JumpTable *> JumpTables;
 
@@ -546,16 +540,6 @@ class BinaryContext {
     return InjectedBinaryFunctions;
   }
 
-  BinaryFunction *createThunkBinaryFunction(const std::string &Name);
-
-  std::vector<BinaryFunction *> &getThunkBinaryFunctions() {
-    return ThunkBinaryFunctions;
-  }
-
-  const BinaryFunction *getThunkLocation() const { return ThunkLocation; }
-
-  void setThunkLocation(const BinaryFunction *BF) { ThunkLocation = BF; }
-
   /// Return vector with all functions, i.e. include functions from the input
   /// binary and functions created by BOLT.
   std::vector<BinaryFunction *> getAllBinaryFunctions();
@@ -1355,10 +1339,6 @@ class BinaryContext {
   uint64_t
   computeInstructionSize(const MCInst &Inst,
                          const MCCodeEmitter *Emitter = nullptr) const {
-    // FIXME: hack for faster size computation on aarch64.
-    if (isAArch64())
-      return MIB->isPseudo(Inst) ? 0 : 4;
-
     if (std::optional<uint32_t> Size = MIB->getSize(Inst))
       return *Size;
 

bolt/include/bolt/Core/BinaryFunction.h

@@ -360,10 +360,6 @@ class BinaryFunction {
   /// True if another function body was merged into this one.
   bool HasFunctionsFoldedInto{false};
 
-  /// True if the function is used for remapping hot text and shall not be
-  /// placed on a huge page.
-  bool IsHotTextMover{false};
-
   /// Name for the section this function code should reside in.
   std::string CodeSectionName;
 
@@ -390,6 +386,9 @@ class BinaryFunction {
   /// Raw branch count for this function in the profile.
   uint64_t RawBranchCount{0};
 
+  /// Dynamically executed function bytes, used for density computation.
+  uint64_t SampleCountInBytes{0};
+
   /// Indicates the type of profile the function is using.
   uint16_t ProfileFlags{PF_NONE};
 
@@ -909,6 +908,10 @@ class BinaryFunction {
     return BB && BB->getOffset() == Offset ? BB : nullptr;
   }
 
+  const BinaryBasicBlock *getBasicBlockAtOffset(uint64_t Offset) const {
+    return const_cast<BinaryFunction *>(this)->getBasicBlockAtOffset(Offset);
+  }
+
   /// Retrieve the landing pad BB associated with invoke instruction \p Invoke
   /// that is in \p BB. Return nullptr if none exists
   BinaryBasicBlock *getLandingPadBBFor(const BinaryBasicBlock &BB,
@@ -1360,8 +1363,6 @@ class BinaryFunction {
   /// Return true if the original entry point was patched.
   bool isPatched() const { return IsPatched; }
 
-  bool isHotTextMover() const { return IsHotTextMover; }
-
   const JumpTable *getJumpTable(const MCInst &Inst) const {
     const uint64_t Address = BC.MIB->getJumpTable(Inst);
     return getJumpTableContainingAddress(Address);
@@ -1714,8 +1715,6 @@ class BinaryFunction {
 
   void setIsPatched(bool V) { IsPatched = V; }
 
-  void setHotTextMover(bool V) { IsHotTextMover = V; }
-
   void setHasIndirectTargetToSplitFragment(bool V) {
     HasIndirectTargetToSplitFragment = V;
   }
@@ -1852,6 +1851,9 @@ class BinaryFunction {
   /// to this function.
   void setRawBranchCount(uint64_t Count) { RawBranchCount = Count; }
 
+  /// Return the number of dynamically executed bytes, from raw perf data.
+  uint64_t getSampleCountInBytes() const { return SampleCountInBytes; }
+
   /// Return the execution count for functions with known profile.
   /// Return 0 if the function has no profile.
   uint64_t getKnownExecutionCount() const {

bolt/include/bolt/Core/DIEBuilder.h

@@ -314,7 +314,7 @@ class DIEBuilder {
 
     BC.errs()
         << "BOLT-ERROR: unable to find TypeUnit for Type Unit at offset 0x"
-        << DU.getOffset() << "\n";
+        << Twine::utohexstr(DU.getOffset()) << "\n";
     return nullptr;
   }
 

bolt/include/bolt/Passes/IdenticalCodeFolding.h

@@ -20,7 +20,15 @@ namespace bolt {
 ///
 class IdenticalCodeFolding : public BinaryFunctionPass {
 protected:
-  bool shouldOptimize(const BinaryFunction &BF) const override;
+  bool shouldOptimize(const BinaryFunction &BF) const override {
+    if (BF.hasUnknownControlFlow())
+      return false;
+    if (BF.isFolded())
+      return false;
+    if (BF.hasSDTMarker())
+      return false;
+    return BinaryFunctionPass::shouldOptimize(BF);
+  }
 
 public:
   explicit IdenticalCodeFolding(const cl::opt<bool> &PrintPass)

bolt/include/bolt/Passes/LongJmp.h

@@ -71,36 +71,11 @@ class LongJmpPass : public BinaryFunctionPass {
   static constexpr size_t LongestJumpBits = 28;
   static constexpr size_t LongestJumpSpan = 1ULL << (LongestJumpBits - 1);
 
-  /// Size of the instruction on AArch64.
-  static constexpr size_t InstSize = 4;
-
   /// Relax all internal function branches including those between fragments.
-  /// Assume fragments are placed in different sections but are within 128MB
-  /// branch span ("medium" code model).
+  /// Assume that fragments are placed in different sections but are within
+  /// 128MB of each other.
   void relaxLocalBranches(BinaryFunction &BF);
 
-  struct FunctionCluster {
-    std::set<BinaryFunction *> Functions;
-
-    // Functions that this cluster of functions is calling. Note that it
-    // excludes all functions in the cluster itself.
-    std::set<BinaryFunction *> Callees;
-
-    uint64_t Size{0};
-
-    // Last function in the cluster.
-    BinaryFunction *LastBF{nullptr};
-  };
-
-  /// Maximum size of the function cluster. Note that it's less than 128MB
-  /// as the size of the cluster plus thunk island should be less than 128MB.
-  static constexpr uint64_t MaxClusterSize = 124 * 1024 * 1024;
-
-  /// Relax calls for medium code model where code is < 256MB.
-  /// A thunk island will be introduced between two clusters of functions to
-  /// enable calls over 128MB.
-  void relaxCalls(BinaryContext &BC);
-
   ///                 -- Layout estimation methods --
   /// Try to do layout before running the emitter, by looking at BinaryFunctions
   /// and MCInsts -- this is an estimation. To be correct for longjmp inserter

bolt/include/bolt/Profile/DataAggregator.h

@@ -266,7 +266,8 @@ class DataAggregator : public DataReader {
                      uint64_t Mispreds);
 
   /// Register a \p Branch.
-  bool doBranch(uint64_t From, uint64_t To, uint64_t Count, uint64_t Mispreds);
+  bool doBranch(uint64_t From, uint64_t To, uint64_t Count, uint64_t Mispreds,
+                bool IsPreagg);
 
   /// Register a trace between two LBR entries supplied in execution order.
   bool doTrace(const LBREntry &First, const LBREntry &Second,

bolt/include/bolt/Profile/YAMLProfileReader.h

@@ -105,7 +105,7 @@ class YAMLProfileReader : public ProfileReaderBase {
   yaml::bolt::BinaryProfile YamlBP;
 
   /// Map a function ID from a YAML profile to a BinaryFunction object.
-  std::vector<BinaryFunction *> YamlProfileToFunction;
+  DenseMap<uint32_t, BinaryFunction *> YamlProfileToFunction;
 
   using FunctionSet = std::unordered_set<const BinaryFunction *>;
   /// To keep track of functions that have a matched profile before the profile
@@ -162,8 +162,6 @@ class YAMLProfileReader : public ProfileReaderBase {
   /// Update matched YAML -> BinaryFunction pair.
   void matchProfileToFunction(yaml::bolt::BinaryFunctionProfile &YamlBF,
                               BinaryFunction &BF) {
-    if (YamlBF.Id >= YamlProfileToFunction.size())
-      YamlProfileToFunction.resize(YamlBF.Id + 1);
     YamlProfileToFunction[YamlBF.Id] = &BF;
     YamlBF.Used = true;