Merge branch 'main' into alignas-order

Author: DenisGZM

.github/CODEOWNERS

@@ -128,7 +128,7 @@
 /mlir/**/Transforms/SROA.* @moxinilian
 
 # BOLT
-/bolt/ @aaupov @maksfb @rafaelauler @ayermolo @dcci @yota9
+/bolt/ @aaupov @maksfb @rafaelauler @ayermolo @yota9
 
 # Bazel build system.
 /utils/bazel/ @rupprecht @keith @aaronmondal

.github/workflows/commit-access-greeter.yml

@@ -12,6 +12,7 @@ jobs:
   commit-access-greeter:
     permissions:
       issues: write
+      pull-requests: read
     if: >-
       github.repository_owner == 'llvm' &&
       github.event.label.name == 'infra:commit-access-request'

bolt/docs/CMakeLists.txt

@@ -48,14 +48,6 @@ if (LLVM_ENABLE_DOXYGEN)
     set(bolt_doxygen_qhp_cust_filter_attrs "")
   endif()
 
-  option(LLVM_DOXYGEN_SVG
-    "Use svg instead of png files for doxygen graphs." OFF)
-  if (LLVM_DOXYGEN_SVG)
-    set(DOT_IMAGE_FORMAT "svg")
-  else()
-    set(DOT_IMAGE_FORMAT "png")
-  endif()
-
   configure_file(${CMAKE_CURRENT_SOURCE_DIR}/doxygen.cfg.in
     ${CMAKE_CURRENT_BINARY_DIR}/doxygen.cfg @ONLY)
 
@@ -73,7 +65,6 @@ if (LLVM_ENABLE_DOXYGEN)
   set(bolt_doxygen_qhelpgenerator_path)
   set(bolt_doxygen_qhp_cust_filter_name)
   set(bolt_doxygen_qhp_cust_filter_attrs)
-  set(DOT_IMAGE_FORMAT)
 
   add_custom_target(doxygen-bolt
     COMMAND ${DOXYGEN_EXECUTABLE} ${CMAKE_CURRENT_BINARY_DIR}/doxygen.cfg

bolt/docs/doxygen.cfg.in

@@ -2210,7 +2210,7 @@ DIRECTORY_GRAPH        = YES
 # The default value is: png.
 # This tag requires that the tag HAVE_DOT is set to YES.
 
-DOT_IMAGE_FORMAT       = @DOT_IMAGE_FORMAT@
+DOT_IMAGE_FORMAT       = svg
 
 # If DOT_IMAGE_FORMAT is set to svg, then this option can be set to YES to
 # enable generation of interactive SVG images that allow zooming and panning.

bolt/include/bolt/Profile/DataAggregator.h

@@ -223,14 +223,16 @@ class DataAggregator : public DataReader {
   bool recordExit(BinaryFunction &BF, uint64_t From, bool Mispred,
                   uint64_t Count = 1) const;
 
-  /// Aggregation statistics
+  /// Branch stacks aggregation statistics
+  uint64_t NumTraces{0};
   uint64_t NumInvalidTraces{0};
   uint64_t NumLongRangeTraces{0};
   /// Specifies how many samples were recorded in cold areas if we are dealing
   /// with profiling data collected in a bolted binary. For LBRs, incremented
   /// for the source of the branch to avoid counting cold activity twice (one
   /// for source and another for destination).
   uint64_t NumColdSamples{0};
+  uint64_t NumTotalSamples{0};
 
   /// Looks into system PATH for Linux Perf and set up the aggregator to use it
   void findPerfExecutable();
@@ -327,8 +329,8 @@ class DataAggregator : public DataReader {
   /// Parse a single LBR entry as output by perf script -Fbrstack
   ErrorOr<LBREntry> parseLBREntry();
 
-  /// Parse LBR sample, returns the number of traces.
-  uint64_t parseLBRSample(const PerfBranchSample &Sample, bool NeedsSkylakeFix);
+  /// Parse LBR sample.
+  void parseLBRSample(const PerfBranchSample &Sample, bool NeedsSkylakeFix);
 
   /// Parse and pre-aggregate branch events.
   std::error_code parseBranchEvents();
@@ -487,6 +489,13 @@ class DataAggregator : public DataReader {
   void dump(const PerfBranchSample &Sample) const;
   void dump(const PerfMemSample &Sample) const;
 
+  /// Profile diagnostics print methods
+  void printColdSamplesDiagnostic() const;
+  void printLongRangeTracesDiagnostic() const;
+  void printBranchSamplesDiagnostics() const;
+  void printBasicSamplesDiagnostics(uint64_t OutOfRangeSamples) const;
+  void printBranchStacksDiagnostics(uint64_t IgnoredSamples) const;
+
 public:
   /// If perf.data was collected without build ids, the buildid-list may contain
   /// incomplete entries. Return true if the buffer containing

bolt/lib/Core/BinaryBasicBlock.cpp

@@ -372,8 +372,7 @@ void BinaryBasicBlock::updateJumpTableSuccessors() {
              [](const BinaryBasicBlock *BB1, const BinaryBasicBlock *BB2) {
                return BB1->getInputOffset() < BB2->getInputOffset();
              });
-  SuccessorBBs.erase(std::unique(SuccessorBBs.begin(), SuccessorBBs.end()),
-                     SuccessorBBs.end());
+  SuccessorBBs.erase(llvm::unique(SuccessorBBs), SuccessorBBs.end());
 
   for (BinaryBasicBlock *BB : SuccessorBBs)
     addSuccessor(BB);

bolt/lib/Core/BinaryFunction.cpp

@@ -2376,7 +2376,7 @@ Error BinaryFunction::buildCFG(MCPlusBuilder::AllocatorIdTy AllocatorId) {
   // Without doing jump table value profiling we don't have a use for extra
   // (duplicate) branches.
   llvm::sort(TakenBranches);
-  auto NewEnd = std::unique(TakenBranches.begin(), TakenBranches.end());
+  auto NewEnd = llvm::unique(TakenBranches);
   TakenBranches.erase(NewEnd, TakenBranches.end());
 
   for (std::pair<uint32_t, uint32_t> &Branch : TakenBranches) {

bolt/lib/Core/DebugNames.cpp

@@ -440,8 +440,7 @@ void DWARF5AcceleratorTable::computeBucketCount() {
   for (const auto &E : Entries)
     Uniques.push_back(E.second.HashValue);
   array_pod_sort(Uniques.begin(), Uniques.end());
-  std::vector<uint32_t>::iterator P =
-      std::unique(Uniques.begin(), Uniques.end());
+  std::vector<uint32_t>::iterator P = llvm::unique(Uniques);
 
   UniqueHashCount = std::distance(Uniques.begin(), P);
 

bolt/lib/Core/MCPlusBuilder.cpp

@@ -442,10 +442,10 @@ void MCPlusBuilder::getUsedRegs(const MCInst &Inst, BitVector &Regs) const {
   for (MCPhysReg ImplicitUse : InstInfo.implicit_uses())
     Regs |= getAliases(ImplicitUse, /*OnlySmaller=*/true);
 
-  for (unsigned I = 0, E = Inst.getNumOperands(); I != E; ++I) {
-    if (!Inst.getOperand(I).isReg())
+  for (const MCOperand &Operand : useOperands(Inst)) {
+    if (!Operand.isReg())
       continue;
-    Regs |= getAliases(Inst.getOperand(I).getReg(), /*OnlySmaller=*/true);
+    Regs |= getAliases(Operand.getReg(), /*OnlySmaller=*/true);
   }
 }