[PGO] Supporting code for always instrumenting entry block

[xur]

Summary

This patch is a split of D83024 (https://reviews.llvm.org/D83024)

It includes all the supporting code without flipping the default behavior.
It adds a variant bit in the profile version, adds new directives in text profile format, and changes llvm-profdata tool accrodingly.

Many test changes from D83024 are also included. This should be OK as I explicitly specify the options.

[davidxl]

clang/test/CodeGenCXX/Inputs/profile-remap.proftext

3 ↗

(On Diff #279601)

is this change needed?

llvm/lib/ProfileData/InstrProfReader.cpp

188

The parser should probably be made such that the order of the directive does not matter.

[xur]

clang/test/CodeGenCXX/Inputs/profile-remap.proftext

3 ↗

(On Diff #279601)

They are not needed.

This actually tests the new proftext file.

We can remove it and the change in gcc-flag-compatibility_IR.proftext
Or add update the test to use the new proftext.

llvm/lib/ProfileData/InstrProfReader.cpp

188

That can be done.

How about duplicates? if the order does not matter, I assume we can have multiple directive and the last one rules?
Or we need to throw an error when seeing duplicates?

[xur]

Addressed review comments from David.

[davidxl]

lgtm (with adding a not_entry_first test case).

[kpdev42]

llvm/include/llvm/ProfileData/InstrProfData.inc

676

This revision is closed, so excuse me for the question:
Files ./llvm/include/llvm/ProfileData/InstrProfData.inc and ./compiler-rt/include/profile/InstrProfData.inc should be two identical copies, as stated in their description.

* The file has two identical copies. The master copy lives in LLVM and
* the other one  sits in compiler-rt/lib/profile directory. To make changes
* in this file, first modify the master copy and copy it over to compiler-rt.
* Testing of any change in this file can start only after the two copies are
* synced up.

Should we add VARIANT_MASK_INSTR_ENTRY to compiler-rt or change description?

clang/test/CodeGenCXX/Inputs/profile-remap_entry.proftext

+:ir
+:entry_first
+_ZN3Foo8functionENS_1XE
[phone]
+2
+100
+90
+

clang/test/CodeGenCXX/profile-remap.cpp

 // REQUIRES: x86-registered-target
 //
 // RUN: %clang_cc1 -triple x86_64-linux-gnu -fprofile-sample-use=%S/Inputs/profile-remap.samples -fprofile-remapping-file=%S/Inputs/profile-remap.map -fexperimental-new-pass-manager -O2 %s -emit-llvm -o - | FileCheck %s --check-prefixes=CHECK,CHECK-SAMPLES
 // RUN: llvm-profdata merge -output %T.profdata %S/Inputs/profile-remap.proftext
 // RUN: %clang_cc1 -triple x86_64-linux-gnu -fprofile-instrument-use-path=%T.profdata -fprofile-remapping-file=%S/Inputs/profile-remap.map -fexperimental-new-pass-manager -O2 %s -emit-llvm -o - | FileCheck %s --check-prefixes=CHECK,CHECK-INSTR
+// RUN: llvm-profdata merge -output %T.profdata %S/Inputs/profile-remap_entry.proftext
+// RUN: %clang_cc1 -triple x86_64-linux-gnu -fprofile-instrument-use-path=%T.profdata -fprofile-remapping-file=%S/Inputs/profile-remap.map -fexperimental-new-pass-manager -O2 %s -emit-llvm -o - | FileCheck %s --check-prefixes=CHECK,CHECK-INSTR
 
 namespace Foo {
   struct X {};
   bool cond();
   void bar();
   void baz();
   void function(X x) {
     if (cond())

clang/test/Profile/Inputs/gcc-flag-compatibility_IR_entry.proftext

+# IR level Instrumentation Flag
+:ir
+:entry_first
+main
+# Func Hash:
[phone]
+# Num Counters:
+2
+# Counter Values:
+1
+100
+

clang/test/Profile/gcc-flag-compatibility.c

 
 // Check that -fprofile-use=some/path/file.prof reads some/path/file.prof
 // This uses LLVM IR format profile.
 // RUN: rm -rf %t.dir
 // RUN: mkdir -p %t.dir/some/path
 // RUN: llvm-profdata merge %S/Inputs/gcc-flag-compatibility_IR.proftext -o %t.dir/some/path/file.prof
 // RUN: %clang %s -o - -emit-llvm -S -fprofile-use=%t.dir/some/path/file.prof -fno-experimental-new-pass-manager | FileCheck -check-prefix=PROFILE-USE-IR %s
 // RUN: %clang %s -o - -emit-llvm -S -fprofile-use=%t.dir/some/path/file.prof -fexperimental-new-pass-manager | FileCheck -check-prefix=PROFILE-USE-IR %s
+//
+// RUN: llvm-profdata merge %S/Inputs/gcc-flag-compatibility_IR_entry.proftext -o %t.dir/some/path/file.prof
+// RUN: %clang %s -o - -emit-llvm -S -fprofile-use=%t.dir/some/path/file.prof -fno-experimental-new-pass-manager | FileCheck -check-prefix=PROFILE-USE-IR %s
+// RUN: %clang %s -o - -emit-llvm -S -fprofile-use=%t.dir/some/path/file.prof -fexperimental-new-pass-manager | FileCheck -check-prefix=PROFILE-USE-IR %s
 
 // PROFILE-USE-IR: = !{!"branch_weights", i32 100, i32 1}
 
 int X = 0;
 
 int main() {
   int i;
   for (i = 0; i < 100; i++)
     X += i;
   return 0;
 }

llvm/include/llvm/ProfileData/InstrProf.h

 } // end namespace RawInstrProf
 
 // Parse MemOP Size range option.
 void getMemOPSizeRangeFromOption(StringRef Str, int64_t &RangeStart,
                                  int64_t &RangeLast);
 
 // Create a COMDAT variable INSTR_PROF_RAW_VERSION_VAR to make the runtime
 // aware this is an ir_level profile so it can set the version flag.
-void createIRLevelProfileFlagVar(Module &M, bool IsCS);
+void createIRLevelProfileFlagVar(Module &M, bool IsCS,
+                                 bool InstrEntryBBEnabled);
 
 // Create the variable for the profile file name.
 void createProfileFileNameVar(Module &M, StringRef InstrProfileOutput);
 
 // Whether to compress function names in profile records, and filenames in
 // code coverage mappings. Used by the Instrumentation library and unit tests.
 extern cl::opt<bool> DoInstrProfNameCompression;
 
 } // end namespace llvm
 #endif // LLVM_PROFILEDATA_INSTRPROF_H

llvm/include/llvm/ProfileData/InstrProfData.inc

  * bits (i.e. bit 56) to 1 to indicate if this is an IR-level instrumentaiton
  * generated profile, and 0 if this is a Clang FE generated profile.
  * 1 in bit 57 indicates there are context-sensitive records in the profile.
  */
 #define VARIANT_MASKS_ALL 0xff00000000000000ULL
 #define GET_VERSION(V) ((V) & ~VARIANT_MASKS_ALL)
 #define VARIANT_MASK_IR_PROF (0x1ULL << 56)
 #define VARIANT_MASK_CSIR_PROF (0x1ULL << 57)
+#define VARIANT_MASK_INSTR_ENTRY (0x1ULL << 58)
 #define INSTR_PROF_RAW_VERSION_VAR __llvm_profile_raw_version
 #define INSTR_PROF_PROFILE_RUNTIME_VAR __llvm_profile_runtime
 
 /* The variable that holds the name of the profile data
  * specified via command line. */
 #define INSTR_PROF_PROFILE_NAME_VAR __llvm_profile_filename
 
 /* section name strings common to all targets other

llvm/include/llvm/ProfileData/InstrProfReader.h

public:
   /// Iterator over profile data.
   InstrProfIterator begin() { return InstrProfIterator(this); }
   InstrProfIterator end() { return InstrProfIterator(); }
 
   virtual bool isIRLevelProfile() const = 0;
 
   virtual bool hasCSIRLevelProfile() const = 0;
 
+  virtual bool instrEntryBBEnabled() const = 0;
+
   /// Return the PGO symtab. There are three different readers:
   /// Raw, Text, and Indexed profile readers. The first two types
   /// of readers are used only by llvm-profdata tool, while the indexed
   /// profile reader is also used by llvm-cov tool and the compiler (
   /// backend or frontend). Since creating PGO symtab can create
   /// significant runtime and memory overhead (as it touches data
   /// for the whole program), InstrProfSymtab for the indexed profile
   /// reader should be created on demand and it is recommended to be

 class TextInstrProfReader : public InstrProfReader {
 private:
   /// The profile data file contents.
   std::unique_ptr<MemoryBuffer> DataBuffer;
   /// Iterator over the profile data.
   line_iterator Line;
   bool IsIRLevelProfile = false;
   bool HasCSIRLevelProfile = false;
+  bool InstrEntryBBEnabled = false;
 
   Error readValueProfileData(InstrProfRecord &Record);
 
 public:
   TextInstrProfReader(std::unique_ptr<MemoryBuffer> DataBuffer_)
       : DataBuffer(std::move(DataBuffer_)), Line(*DataBuffer, true, '#') {}
   TextInstrProfReader(const TextInstrProfReader &) = delete;
   TextInstrProfReader &operator=(const TextInstrProfReader &) = delete;
 
   /// Return true if the given buffer is in text instrprof format.
   static bool hasFormat(const MemoryBuffer &Buffer);
 
   bool isIRLevelProfile() const override { return IsIRLevelProfile; }
 
   bool hasCSIRLevelProfile() const override { return HasCSIRLevelProfile; }
 
+  bool instrEntryBBEnabled() const override { return InstrEntryBBEnabled; }
+
   /// Read the header.
   Error readHeader() override;
 
   /// Read a single record.
   Error readNextRecord(NamedInstrProfRecord &Record) override;
 
   InstrProfSymtab &getSymtab() override {
     assert(Symtab.get());
public:
   bool isIRLevelProfile() const override {
     return (Version & VARIANT_MASK_IR_PROF) != 0;
   }
 
   bool hasCSIRLevelProfile() const override {
     return (Version & VARIANT_MASK_CSIR_PROF) != 0;
   }
 
+  bool instrEntryBBEnabled() const override {
+    return (Version & VARIANT_MASK_INSTR_ENTRY) != 0;
+  }
+
   InstrProfSymtab &getSymtab() override {
     assert(Symtab.get());
     return *Symtab.get();
   }
 
 private:
   Error createSymtab(InstrProfSymtab &Symtab);
   Error readNextHeader(const char *CurrentPos);
struct InstrProfReaderIndexBase {
   virtual Error getRecords(StringRef FuncName,
                                      ArrayRef<NamedInstrProfRecord> &Data) = 0;
   virtual void advanceToNextKey() = 0;
   virtual bool atEnd() const = 0;
   virtual void setValueProfDataEndianness(support::endianness Endianness) = 0;
   virtual uint64_t getVersion() const = 0;
   virtual bool isIRLevelProfile() const = 0;
   virtual bool hasCSIRLevelProfile() const = 0;
+  virtual bool instrEntryBBEnabled() const = 0;
   virtual Error populateSymtab(InstrProfSymtab &) = 0;
 };
 
 using OnDiskHashTableImplV3 =
     OnDiskIterableChainedHashTable<InstrProfLookupTrait>;
 
 template <typename HashTableImpl>
 class InstrProfReaderItaniumRemapper;
public:
   bool isIRLevelProfile() const override {
     return (FormatVersion & VARIANT_MASK_IR_PROF) != 0;
   }
 
   bool hasCSIRLevelProfile() const override {
     return (FormatVersion & VARIANT_MASK_CSIR_PROF) != 0;
   }
 
+  bool instrEntryBBEnabled() const override {
+    return (FormatVersion & VARIANT_MASK_INSTR_ENTRY) != 0;
+  }
+
   Error populateSymtab(InstrProfSymtab &Symtab) override {
     return Symtab.create(HashTable->keys());
   }
 };
 
 /// Name matcher supporting fuzzy matching of symbol names to names in profiles.
 class InstrProfReaderRemapper {
 public:
public:
 
   /// Return the profile version.
   uint64_t getVersion() const { return Index->getVersion(); }
   bool isIRLevelProfile() const override { return Index->isIRLevelProfile(); }
   bool hasCSIRLevelProfile() const override {
     return Index->hasCSIRLevelProfile();
   }
 
+  bool instrEntryBBEnabled() const override {
+    return Index->instrEntryBBEnabled();
+  }
+
   /// Return true if the given buffer is in an indexed instrprof format.
   static bool hasFormat(const MemoryBuffer &DataBuffer);
 
   /// Read the file header.
   Error readHeader() override;
   /// Read a single record.
   Error readNextRecord(NamedInstrProfRecord &Record) override;
 

llvm/include/llvm/ProfileData/InstrProfWriter.h

public:
   using ProfilingData = SmallDenseMap<uint64_t, InstrProfRecord>;
   // PF_IRLevelWithCS is the profile from context sensitive IR instrumentation.
   enum ProfKind { PF_Unknown = 0, PF_FE, PF_IRLevel, PF_IRLevelWithCS };
 
 private:
   bool Sparse;
   StringMap<ProfilingData> FunctionData;
   ProfKind ProfileKind = PF_Unknown;
+  bool InstrEntryBBEnabled;
   // Use raw pointer here for the incomplete type object.
   InstrProfRecordWriterTrait *InfoObj;
 
 public:
-  InstrProfWriter(bool Sparse = false);
+  InstrProfWriter(bool Sparse = false, bool InstrEntryBBEnabled = false);
   ~InstrProfWriter();
 
   /// Add function counts for the given function. If there are already counts
   /// for this function and the hash and number of counts match, each counter is
   /// summed. Optionally scale counts by \p Weight.
   void addRecord(NamedInstrProfRecord &&I, uint64_t Weight,
                  function_ref<void(Error)> Warn);
   void addRecord(NamedInstrProfRecord &&I, function_ref<void(Error)> Warn) {
Error setIsIRLevelProfile(bool IsIRLevel, bool WithCS) {
     // When merging a context-sensitive profile (WithCS == true) with an IRLevel
     // profile, set the kind to PF_IRLevelWithCS.
     if (ProfileKind == PF_IRLevel && WithCS)
       ProfileKind = PF_IRLevelWithCS;
 
     return Error::success();
   }
 
+  void setInstrEntryBBEnabled(bool Enabled) { InstrEntryBBEnabled = Enabled; }
   // Internal interface for testing purpose only.
   void setValueProfDataEndianness(support::endianness Endianness);
   void setOutputSparse(bool Sparse);
   // Compute the overlap b/w this object and Other. Program level result is
   // stored in Overlap and function level result is stored in FuncLevelOverlap.
   void overlapRecord(NamedInstrProfRecord &&Other, OverlapStats &Overlap,
                      OverlapStats &FuncLevelOverlap,
                      const OverlapFuncFilters &FuncFilter);

llvm/lib/ProfileData/InstrProf.cpp

if (!MemOPSizeRange.empty()) {
     } else
       MemOPSizeRange.getAsInteger(10, RangeLast);
   }
   assert(RangeLast >= RangeStart);
 }
 
 // Create a COMDAT variable INSTR_PROF_RAW_VERSION_VAR to make the runtime
 // aware this is an ir_level profile so it can set the version flag.
-void createIRLevelProfileFlagVar(Module &M, bool IsCS) {
+void createIRLevelProfileFlagVar(Module &M, bool IsCS,
+                                 bool InstrEntryBBEnabled) {
   const StringRef VarName(INSTR_PROF_QUOTE(INSTR_PROF_RAW_VERSION_VAR));
   Type *IntTy64 = Type::getInt64Ty(M.getContext());
   uint64_t ProfileVersion = (INSTR_PROF_RAW_VERSION | VARIANT_MASK_IR_PROF);
   if (IsCS)
     ProfileVersion |= VARIANT_MASK_CSIR_PROF;
+  if (InstrEntryBBEnabled)
+    ProfileVersion |= VARIANT_MASK_INSTR_ENTRY;
   auto IRLevelVersionVariable = new GlobalVariable(
       M, IntTy64, true, GlobalValue::WeakAnyLinkage,
       Constant::getIntegerValue(IntTy64, APInt(64, ProfileVersion)), VarName);
   IRLevelVersionVariable->setVisibility(GlobalValue::DefaultVisibility);
   Triple TT(M.getTargetTriple());
   if (TT.supportsCOMDAT()) {
     IRLevelVersionVariable->setLinkage(GlobalValue::ExternalLinkage);
     IRLevelVersionVariable->setComdat(M.getOrInsertComdat(VarName));

llvm/lib/ProfileData/InstrProfReader.cpp

 }
 
 // Read the profile variant flag from the header: ":FE" means this is a FE
 // generated profile. ":IR" means this is an IR level profile. Other strings
 // with a leading ':' will be reported an error format.
 Error TextInstrProfReader::readHeader() {
   Symtab.reset(new InstrProfSymtab());
   bool IsIRInstr = false;
-  if (!Line->startswith(":")) {
-    IsIRLevelProfile = false;
-    return success();
-  }
-  StringRef Str = (Line)->substr(1);
+  bool IsEntryFirst = false;
+  bool IsCS = false;
+
+  while (Line->startswith(":")) {
+    StringRef Str = Line->substr(1);
     if (Str.equals_lower("ir"))
       IsIRInstr = true;
     else if (Str.equals_lower("fe"))
       IsIRInstr = false;
     else if (Str.equals_lower("csir")) {
       IsIRInstr = true;
-    HasCSIRLevelProfile = true;
-  } else
+      IsCS = true;
+    } else if (Str.equals_lower("entry_first"))
+      IsEntryFirst = true;
+    else if (Str.equals_lower("not_entry_first"))
+      IsEntryFirst = false;
+    else
       return error(instrprof_error::bad_header);
-
     ++Line;
+  }
   IsIRLevelProfile = IsIRInstr;
+  InstrEntryBBEnabled = IsEntryFirst;
+  HasCSIRLevelProfile = IsCS;
   return success();
 }
 
 Error
 TextInstrProfReader::readValueProfileData(InstrProfRecord &Record) {
 
 #define CHECK_LINE_END(Line)                                                   \
   if (Line.is_at_end())                                                        \
     return error(instrprof_error::truncated);
 #define READ_NUM(Str, Dst)                                                     \
   if ((Str).getAsInteger(10, (Dst)))                                           \
     return error(instrprof_error::malformed);
 #define VP_READ_ADVANCE(Val)                                                   \
   CHECK_LINE_END(Line);                                                        \
   uint32_t Val;                                                                \
   READ_NUM((*Line), (Val));                                                    \
   Line++;

llvm/lib/ProfileData/InstrProfWriter.cpp

for (const auto &ProfileData : *V) {
       VDataPtr->swapBytesFromHost(ValueProfDataEndianness);
       Out.write((const char *)VDataPtr.get(), S);
     }
   }
 };
 
 } // end namespace llvm
 
-InstrProfWriter::InstrProfWriter(bool Sparse)
-    : Sparse(Sparse), InfoObj(new InstrProfRecordWriterTrait()) {}
+InstrProfWriter::InstrProfWriter(bool Sparse, bool InstrEntryBBEnabled)
+    : Sparse(Sparse), InstrEntryBBEnabled(InstrEntryBBEnabled),
+      InfoObj(new InstrProfRecordWriterTrait()) {}
 
 InstrProfWriter::~InstrProfWriter() { delete InfoObj; }
 
 // Internal interface for testing purpose only.
 void InstrProfWriter::setValueProfDataEndianness(
     support::endianness Endianness) {
   InfoObj->ValueProfDataEndianness = Endianness;
 }
void InstrProfWriter::writeImpl(ProfOStream &OS) {
   Header.Magic = IndexedInstrProf::Magic;
   Header.Version = IndexedInstrProf::ProfVersion::CurrentVersion;
   if (ProfileKind == PF_IRLevel)
     Header.Version |= VARIANT_MASK_IR_PROF;
   if (ProfileKind == PF_IRLevelWithCS) {
     Header.Version |= VARIANT_MASK_IR_PROF;
     Header.Version |= VARIANT_MASK_CSIR_PROF;
   }
+  if (InstrEntryBBEnabled)
+    Header.Version |= VARIANT_MASK_INSTR_ENTRY;
+
   Header.Unused = 0;
   Header.HashType = static_cast<uint64_t>(IndexedInstrProf::HashType);
   Header.HashOffset = 0;
   int N = sizeof(IndexedInstrProf::Header) / sizeof(uint64_t);
 
   // Only write out all the fields except 'HashOffset'. We need
   // to remember the offset of that field to allow back patching
   // later.
void InstrProfWriter::writeRecordInText(StringRef Name, uint64_t Hash,
   OS << "\n";
 }
 
 Error InstrProfWriter::writeText(raw_fd_ostream &OS) {
   if (ProfileKind == PF_IRLevel)
     OS << "# IR level Instrumentation Flag\n:ir\n";
   else if (ProfileKind == PF_IRLevelWithCS)
     OS << "# CSIR level Instrumentation Flag\n:csir\n";
+  if (InstrEntryBBEnabled)
+    OS << "# Always instrument the function entry block\n:entry_first\n";
   InstrProfSymtab Symtab;
 
   using FuncPair = detail::DenseMapPair<uint64_t, InstrProfRecord>;
   using RecordType = std::pair<StringRef, FuncPair>;
   SmallVector<RecordType, 4> OrderedFuncData;
 
   for (const auto &I : FunctionData) {
     if (shouldEncodeData(I.getValue())) {

llvm/lib/Transforms/Instrumentation/CFGMST.h

 #define LLVM_LIB_TRANSFORMS_INSTRUMENTATION_CFGMST_H
 
 #include "llvm/ADT/DenseMap.h"
 #include "llvm/ADT/STLExtras.h"
 #include "llvm/Analysis/BlockFrequencyInfo.h"
 #include "llvm/Analysis/BranchProbabilityInfo.h"
 #include "llvm/Analysis/CFG.h"
 #include "llvm/Support/BranchProbability.h"
-#include "llvm/Support/CommandLine.h"
 #include "llvm/Support/Debug.h"
 #include "llvm/Support/raw_ostream.h"
 #include "llvm/Transforms/Utils/BasicBlockUtils.h"
 #include <utility>
 #include <vector>
 
 #define DEBUG_TYPE "cfgmst"
 
 using namespace llvm;
-static cl::opt<bool> PGOInstrumentEntry(
-    "pgo-instrument-entry", cl::init(false), cl::Hidden,
-    cl::desc("Force to instrument function entry basicblock."));
 
 namespace llvm {
 
 /// An union-find based Minimum Spanning Tree for CFG
 ///
 /// Implements a Union-find algorithm to compute Minimum Spanning Tree
 /// for a given CFG.
 template <class Edge, class BBInfo> class CFGMST {
public:
   // Edges with large weight will be put into MST first so they are less likely
   // to be instrumented.
   void buildEdges() {
     LLVM_DEBUG(dbgs() << "Build Edge on " << F.getName() << "\n");
 
     const BasicBlock *Entry = &(F.getEntryBlock());
     uint64_t EntryWeight = (BFI != nullptr ? BFI->getEntryFreq() : 2);
     // If we want to instrument the entry count, lower the weight to 0.
-    if (PGOInstrumentEntry)
+    if (InstrumentFuncEntry)
       EntryWeight = 0;
     Edge *EntryIncoming = nullptr, *EntryOutgoing = nullptr,
          *ExitOutgoing = nullptr, *ExitIncoming = nullptr;
     uint64_t MaxEntryOutWeight = 0, MaxExitOutWeight = 0, MaxExitInWeight = 0;
 
     // Add a fake edge to the entry.
     EntryIncoming = &addEdge(nullptr, Entry, EntryWeight);
     LLVM_DEBUG(dbgs() << "  Edge: from fake node to " << Entry->getName()
if (Inserted)
       Iter->second = std::move(std::make_unique<BBInfo>(Index));
     AllEdges.emplace_back(new Edge(Src, Dest, W));
     return *AllEdges.back();
   }
 
   BranchProbabilityInfo *BPI;
   BlockFrequencyInfo *BFI;
 
+  // If function entry will be always instrumented.
+  bool InstrumentFuncEntry;
+
 public:
-  CFGMST(Function &Func, BranchProbabilityInfo *BPI_ = nullptr,
+  CFGMST(Function &Func, bool InstrumentFuncEntry_,
+         BranchProbabilityInfo *BPI_ = nullptr,
          BlockFrequencyInfo *BFI_ = nullptr)
-      : F(Func), BPI(BPI_), BFI(BFI_) {
+      : F(Func), BPI(BPI_), BFI(BFI_),
+        InstrumentFuncEntry(InstrumentFuncEntry_) {
     buildEdges();
     sortEdgesByWeight();
     computeMinimumSpanningTree();
-    if (PGOInstrumentEntry && (AllEdges.size() > 1))
+    if (AllEdges.size() > 1 && InstrumentFuncEntry)
       std::iter_swap(std::move(AllEdges.begin()),
                      std::move(AllEdges.begin() + AllEdges.size() - 1));
   }
 };
 
 } // end namespace llvm
 
 #undef DEBUG_TYPE // "cfgmst"
 
 #endif // LLVM_LIB_TRANSFORMS_INSTRUMENTATION_CFGMST_H

llvm/lib/Transforms/Instrumentation/PGOInstrumentation.cpp

 // Emit branch probability as optimization remarks.
 static cl::opt<bool>
     EmitBranchProbability("pgo-emit-branch-prob", cl::init(false), cl::Hidden,
                           cl::desc("When this option is on, the annotated "
                                    "branch probability will be emitted as "
                                    "optimization remarks: -{Rpass|"
                                    "pass-remarks}=pgo-instrumentation"));
 
+static cl::opt<bool> PGOInstrumentEntry(
+    "pgo-instrument-entry", cl::init(false), cl::Hidden,
+    cl::desc("Force to instrument function entry basicblock."));
+
 // Command line option to turn on CFG dot dump after profile annotation.
 // Defined in Analysis/BlockFrequencyInfo.cpp:  -pgo-view-counts
 extern cl::opt<PGOViewCountsType> PGOViewCounts;
 
 // Command line option to specify the name of the function for CFG dump
 // Defined in Analysis/BlockFrequencyInfo.cpp:  -view-bfi-func-name=
 extern cl::opt<std::string> ViewBlockFreqFuncName;
 
PGOInstrumentationGenCreateVarLegacyPass(std::string CSInstrName = "")
       : ModulePass(ID), InstrProfileOutput(CSInstrName) {
     initializePGOInstrumentationGenCreateVarLegacyPassPass(
         *PassRegistry::getPassRegistry());
   }
 
 private:
   bool runOnModule(Module &M) override {
     createProfileFileNameVar(M, InstrProfileOutput);
-    createIRLevelProfileFlagVar(M, true);
+    createIRLevelProfileFlagVar(M, /* IsCS */ true, PGOInstrumentEntry);
     return false;
   }
   std::string InstrProfileOutput;
 };
 
 } // end anonymous namespace
 
 char PGOInstrumentationGenLegacyPass::ID = 0;

 // This class implements the CFG edges. Note the CFG can be a multi-graph.
 template <class Edge, class BBInfo> class FuncPGOInstrumentation {
 private:
   Function &F;
 
   // Is this is context-sensitive instrumentation.
   bool IsCS;
 
+  // If we instrument function entry BB by default.
+  bool InstrumentFuncEntry;
+
   // A map that stores the Comdat group in function F.
   std::unordered_multimap<Comdat *, GlobalValue *> &ComdatMembers;
 
   ValueProfileCollector VPC;
 
   void computeCFGHash();
   void renameComdatFunction();
 
void dumpInfo(std::string Str = "") const {
     MST.dumpEdges(dbgs(), Twine("Dump Function ") + FuncName + " Hash: " +
                               Twine(FunctionHash) + "\t" + Str);
   }
 
   FuncPGOInstrumentation(
       Function &Func, TargetLibraryInfo &TLI,
       std::unordered_multimap<Comdat *, GlobalValue *> &ComdatMembers,
       bool CreateGlobalVar = false, BranchProbabilityInfo *BPI = nullptr,
-      BlockFrequencyInfo *BFI = nullptr, bool IsCS = false)
+      BlockFrequencyInfo *BFI = nullptr, bool IsCS = false,
+      bool InstrumentFuncEntry = true)
       : F(Func), IsCS(IsCS), ComdatMembers(ComdatMembers), VPC(Func, TLI),
-        ValueSites(IPVK_Last + 1), SIVisitor(Func), MST(F, BPI, BFI) {
+        ValueSites(IPVK_Last + 1), SIVisitor(Func),
+        MST(F, InstrumentFuncEntry, BPI, BFI) {
     // This should be done before CFG hash computation.
     SIVisitor.countSelects(Func);
     ValueSites[IPVK_MemOPSize] = VPC.get(IPVK_MemOPSize);
     if (!IsCS) {
       NumOfPGOSelectInsts += SIVisitor.getNumOfSelectInsts();
       NumOfPGOMemIntrinsics += ValueSites[IPVK_MemOPSize].size();
       NumOfPGOBB += MST.BBInfos.size();
       ValueSites[IPVK_IndirectCallTarget] = VPC.get(IPVK_IndirectCallTarget);
static void instrumentOneFunc(
     Function &F, Module *M, TargetLibraryInfo &TLI, BranchProbabilityInfo *BPI,
     BlockFrequencyInfo *BFI,
     std::unordered_multimap<Comdat *, GlobalValue *> &ComdatMembers,
     bool IsCS) {
   // Split indirectbr critical edges here before computing the MST rather than
   // later in getInstrBB() to avoid invalidating it.
   SplitIndirectBrCriticalEdges(F, BPI, BFI);
 
-  FuncPGOInstrumentation<PGOEdge, BBInfo> FuncInfo(F, TLI, ComdatMembers, true,
-                                                   BPI, BFI, IsCS);
+  FuncPGOInstrumentation<PGOEdge, BBInfo> FuncInfo(
+      F, TLI, ComdatMembers, true, BPI, BFI, IsCS, PGOInstrumentEntry);
   std::vector<BasicBlock *> InstrumentBBs;
   FuncInfo.getInstrumentBBs(InstrumentBBs);
   unsigned NumCounters =
       InstrumentBBs.size() + FuncInfo.SIVisitor.getNumOfSelectInsts();
 
   uint32_t I = 0;
   Type *I8PtrTy = Type::getInt8PtrTy(M->getContext());
   for (auto *InstrBB : InstrumentBBs) {

 
 namespace {
 
 class PGOUseFunc {
 public:
   PGOUseFunc(Function &Func, Module *Modu, TargetLibraryInfo &TLI,
              std::unordered_multimap<Comdat *, GlobalValue *> &ComdatMembers,
              BranchProbabilityInfo *BPI, BlockFrequencyInfo *BFIin,
-             ProfileSummaryInfo *PSI, bool IsCS)
+             ProfileSummaryInfo *PSI, bool IsCS, bool InstrumentFuncEntry)
       : F(Func), M(Modu), BFI(BFIin), PSI(PSI),
-        FuncInfo(Func, TLI, ComdatMembers, false, BPI, BFIin, IsCS),
+        FuncInfo(Func, TLI, ComdatMembers, false, BPI, BFIin, IsCS,
+                 InstrumentFuncEntry),
         FreqAttr(FFA_Normal), IsCS(IsCS) {}
 
   // Read counts for the instrumented BB from profile.
   bool readCounters(IndexedInstrProfReader *PGOReader, bool &AllZeros);
 
   // Populate the counts for all BBs.
   void populateCounters();
 

 
 static bool InstrumentAllFunctions(
     Module &M, function_ref<TargetLibraryInfo &(Function &)> LookupTLI,
     function_ref<BranchProbabilityInfo *(Function &)> LookupBPI,
     function_ref<BlockFrequencyInfo *(Function &)> LookupBFI, bool IsCS) {
   // For the context-sensitve instrumentation, we should have a separated pass
   // (before LTO/ThinLTO linking) to create these variables.
   if (!IsCS)
-    createIRLevelProfileFlagVar(M, /* IsCS */ false);
+    createIRLevelProfileFlagVar(M, /* IsCS */ false, PGOInstrumentEntry);
   std::unordered_multimap<Comdat *, GlobalValue *> ComdatMembers;
   collectComdatMembers(M, ComdatMembers);
 
   for (auto &F : M) {
     if (F.isDeclaration())
       continue;
     auto &TLI = LookupTLI(F);
     auto *BPI = LookupBPI(F);
     auto *BFI = LookupBFI(F);
     instrumentOneFunc(F, &M, TLI, BPI, BFI, ComdatMembers, IsCS);
   }
   return true;
 }
 
 PreservedAnalyses
 PGOInstrumentationGenCreateVar::run(Module &M, ModuleAnalysisManager &AM) {
   createProfileFileNameVar(M, CSInstrName);
-  createIRLevelProfileFlagVar(M, /* IsCS */ true);
+  createIRLevelProfileFlagVar(M, /* IsCS */ true, PGOInstrumentEntry);
   return PreservedAnalyses::all();
 }
 
 bool PGOInstrumentationGenLegacyPass::runOnModule(Module &M) {
   if (skipModule(M))
     return false;
 
   auto LookupTLI = [this](Function &F) -> TargetLibraryInfo & {
M.setProfileSummary(PGOReader->getSummary(IsCS).getMD(M.getContext()),
                       IsCS ? ProfileSummary::PSK_CSInstr
                            : ProfileSummary::PSK_Instr);
   PSI->refresh();
 
   std::unordered_multimap<Comdat *, GlobalValue *> ComdatMembers;
   collectComdatMembers(M, ComdatMembers);
   std::vector<Function *> HotFunctions;
   std::vector<Function *> ColdFunctions;
+
+  // If the profile marked as always instrument the entry BB, do the
+  // same. Note this can be overwritten by the internal option in CFGMST.h
+  bool InstrumentFuncEntry = PGOReader->instrEntryBBEnabled();
+  if (PGOInstrumentEntry.getNumOccurrences() > 0)
+    InstrumentFuncEntry = PGOInstrumentEntry;
   for (auto &F : M) {
     if (F.isDeclaration())
       continue;
     auto &TLI = LookupTLI(F);
     auto *BPI = LookupBPI(F);
     auto *BFI = LookupBFI(F);
     // Split indirectbr critical edges here before computing the MST rather than
     // later in getInstrBB() to avoid invalidating it.
     SplitIndirectBrCriticalEdges(F, BPI, BFI);
-    PGOUseFunc Func(F, &M, TLI, ComdatMembers, BPI, BFI, PSI, IsCS);
+    PGOUseFunc Func(F, &M, TLI, ComdatMembers, BPI, BFI, PSI, IsCS,
+                    InstrumentFuncEntry);
     bool AllZeros = false;
     if (!Func.readCounters(PGOReader.get(), AllZeros))
       continue;
     if (AllZeros) {
       F.setEntryCount(ProfileCount(0, Function::PCT_Real));
       if (Func.getProgramMaxCount() != 0)
         ColdFunctions.push_back(&F);
       continue;

llvm/test/Transforms/PGOProfile/Inputs/branch2_entry.proftext

+# :ir is the flag to indicate this is IR level profile.
+:ir
+:entry_first
+test_br_2
[phone]
+2
+2
+1
+

llvm/test/Transforms/PGOProfile/Inputs/criticaledge_entry.proftext

+# :ir is the flag to indicate this is IR level profile.
+:ir
+:entry_first
+test_criticalEdge
[phone]
+8
+7
+2
+1
+2
+0
+1
+2
+1
+
+<stdin>:bar
[phone]
+1
+7
+

llvm/test/Transforms/PGOProfile/Inputs/func_entry.proftext

 # IR level Instrumentation Flag
 :ir
+:entry_first
 hot
 # Func Hash:
 12884901887
 # Num Counters:
 1
 # Counter Values:
 9000
 

llvm/test/Transforms/PGOProfile/Inputs/indirectbr_entry.proftext

+# IR level Instrumentation Flag
+:ir
+:entry_first
+foo
+# Func Hash:
[phone]
+# Num Counters:
+4
+# Counter Values:
+202
+20
+5
+63

llvm/test/Transforms/PGOProfile/Inputs/irreducible_entry.proftext

+:ir
+:entry_first
+_Z11irreducibleii
+# Func Hash:
[phone]
+# Num Counters:
+6
+# Counter Values:
+1
+1000
+950
+100
+373
+0
+
+_Z11irreduciblePh
+# Func Hash:
[phone]
+# Num Counters:
+9
+# Counter Values:
+1
+100
+300
+99
+300
+201
+1
+0
+0

llvm/test/Transforms/PGOProfile/Inputs/landingpad_entry.proftext

+# :ir is the flag to indicate this is IR level profile.
+:ir
+:entry_first
+foo
[phone]
+4
+5
+1
+2
+0
+
+bar
[phone]
+2
+3
+2
+

llvm/test/Transforms/PGOProfile/Inputs/loop1_entry.proftext

+# :ir is the flag to indicate this is IR level profile.
+:ir
+:entry_first
+test_simple_for
[phone]
+2
+4
+96
+

llvm/test/Transforms/PGOProfile/Inputs/loop2_entry.proftext

+# :ir is the flag to indicate this is IR level profile.
+:ir
+:entry_first
+test_nested_for
[phone]
+3
+6
+33
+10
+

llvm/test/Transforms/PGOProfile/Inputs/misexpect-branch_entry.proftext

+# IR level Instrumentation Flag
+:ir
+:entry_first
+bar
+# Func Hash:
[phone]
+# Num Counters:
+2
+# Counter Values:
+2000000
+1600332
+
+baz
+# Func Hash:
[phone]
+# Num Counters:
+1
+# Counter Values:
+399668
+
+foo
+# Func Hash:
[phone]
+# Num Counters:
+2
+# Counter Values:
[phone]
+1600332
+
+main
+# Func Hash:
[phone]
+# Num Counters:
+3
+# Counter Values:
+2000000
+2000
+1
+

llvm/test/Transforms/PGOProfile/Inputs/misexpect-switch-correct_entry.proftext

+# IR level Instrumentation Flag
+:ir
+:entry_first
+main
+# Func Hash:
[phone]
+# Num Counters:
+7
+# Counter Values:
+1
+0
+0
+20000
+0
+0
+0
+

llvm/test/Transforms/PGOProfile/Inputs/misexpect-switch_entry.proftext

+# IR level Instrumentation Flag
+:ir
+:entry_first
+main
+# Func Hash:
[phone]
+# Num Counters:
+7
+# Counter Values:
+11889
+3973
+3970
+0
+8111
+1
+0
+

llvm/test/Transforms/PGOProfile/Inputs/select1.proftext

 :ir
+:entry_first
 test_br_2
 72057623705475732
 3
-4
+5
 1
 1
 

llvm/test/Transforms/PGOProfile/Inputs/select2.proftext

 # IR level Instrumentation Flag
 :ir
+:entry_first
 foo
 # Func Hash:
 72057628175588252
 # Num Counters:
 3
 # Counter Values:
-800
 3
+800
 300

llvm/test/Transforms/PGOProfile/Inputs/switch_entry.proftext

+# :ir is the flag to indicate this is IR level profile.
+:ir
+:entry_first
+test_switch
[phone]
+4
+10
+5
+2
+3
+

llvm/test/Transforms/PGOProfile/branch2.ll

-; RUN: opt < %s -pgo-instr-gen -S | FileCheck %s --check-prefix=GEN
-; RUN: opt < %s -passes=pgo-instr-gen -S | FileCheck %s --check-prefix=GEN
+; RUN: opt < %s -pgo-instr-gen -pgo-instrument-entry=false -S | FileCheck %s --check-prefixes=GEN,NOTENTRY
+; RUN: opt < %s -passes=pgo-instr-gen -pgo-instrument-entry=false -S | FileCheck %s --check-prefixes=GEN,NOTENTRY
 ; RUN: llvm-profdata merge %S/Inputs/branch2.proftext -o %t.profdata
-; RUN: opt < %s -pgo-instr-use -pgo-test-profile-file=%t.profdata -S | FileCheck %s --check-prefix=USE
-; RUN: opt < %s -passes=pgo-instr-use -pgo-test-profile-file=%t.profdata -S | FileCheck %s --check-prefix=USE
+; RUN: opt < %s -pgo-instr-use -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -S | FileCheck %s --check-prefix=USE
+; RUN: opt < %s -passes=pgo-instr-use -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -S | FileCheck %s --check-prefix=USE
+
+; RUN: opt < %s -pgo-instr-gen -pgo-instrument-entry=true -S | FileCheck %s --check-prefixes=GEN,ENTRY
+; RUN: opt < %s -passes=pgo-instr-gen -pgo-instrument-entry=true  -S | FileCheck %s --check-prefixes=GEN,ENTRY
+; RUN: llvm-profdata merge %S/Inputs/branch2_entry.proftext -o %t.profdata
+; RUN: opt < %s -pgo-instr-use  -pgo-instrument-entry=true -pgo-test-profile-file=%t.profdata -S | FileCheck %s --check-prefix=USE
+; RUN: opt < %s -passes=pgo-instr-use  -pgo-instrument-entry=true -pgo-test-profile-file=%t.profdata -S | FileCheck %s --check-prefix=USE
 target datalayout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128"
 target triple = "x86_64-unknown-linux-gnu"
 
 ; GEN: $__llvm_profile_raw_version = comdat any
 ; GEN: @__llvm_profile_raw_version = constant i64 {{[0-9]+}}, comdat
 ; GEN: @__profn_test_br_2 = private constant [9 x i8] c"test_br_2"
 
 define i32 @test_br_2(i32 %i) {
 entry:
 ; GEN: entry:
-; GEN-NOT: llvm.instrprof.increment
+; NOTENTRY-NOT: llvm.instrprof.increment
+; ENTRY: call void @llvm.instrprof.increment(i8* getelementptr inbounds ([9 x i8], [9 x i8]* @__profn_test_br_2, i32 0, i32 0), i64 29667547796, i32 2, i32 0)
   %cmp = icmp sgt i32 %i, 0
   br i1 %cmp, label %if.then, label %if.else
 ; USE: br i1 %cmp, label %if.then, label %if.else
 ; USE-SAME: !prof ![[BW_ENTRY:[0-9]+]]
 ; USE: ![[BW_ENTRY]] = !{!"branch_weights", i32 1, i32 1}
 
 if.then:
 ; GEN: if.then:
-; GEN: call void @llvm.instrprof.increment(i8* getelementptr inbounds ([9 x i8], [9 x i8]* @__profn_test_br_2, i32 0, i32 0), i64 29667547796, i32 2, i32 0)
+; NOTENTRY: call void @llvm.instrprof.increment(i8* getelementptr inbounds ([9 x i8], [9 x i8]* @__profn_test_br_2, i32 0, i32 0), i64 29667547796, i32 2, i32 0)
+; ENTRY-NOT: llvm.instrprof.increment
   %add = add nsw i32 %i, 2
   br label %if.end
 
 if.else:
 ; GEN: if.else:
 ; GEN: call void @llvm.instrprof.increment(i8* getelementptr inbounds ([9 x i8], [9 x i8]* @__profn_test_br_2, i32 0, i32 0), i64 29667547796, i32 2, i32 1)
   %sub = sub nsw i32 %i, 2
   br label %if.end
 
 if.end:
 ; GEN: if.end:
 ; GEN-NOT: llvm.instrprof.increment
   %retv = phi i32 [ %add, %if.then ], [ %sub, %if.else ]
   ret i32 %retv
 ; GEN: ret
 }

llvm/test/Transforms/PGOProfile/counter_promo_exit_catchswitch.ll

 ; Test that instrumentation counter promotion for loops does not fail during
 ; compilation for loops that exit to a catchswitch block. In this case, counters
 ; do not get promoted out of the loop body.
 
-; RUN: opt < %s -pgo-instr-gen -instrprof -do-counter-promotion=true -S | FileCheck %s
-; RUN: opt < %s -passes=pgo-instr-gen,instrprof -do-counter-promotion=true -S | FileCheck %s
+; RUN: opt < %s -pgo-instr-gen -instrprof -pgo-instrument-entry=false -do-counter-promotion=true -S | FileCheck %s --check-prefixes=CHECK,NOTENTRY
+; RUN: opt < %s -passes=pgo-instr-gen,instrprof -pgo-instrument-entry=false -do-counter-promotion=true -S | FileCheck %s --check-prefixes=CHECK,NOTENTRY
+; RUN: opt < %s -pgo-instr-gen -instrprof -pgo-instrument-entry=true -do-counter-promotion=true -S | FileCheck %s --check-prefixes=CHECK,ENTRY
+; RUN: opt < %s -passes=pgo-instr-gen,instrprof -pgo-instrument-entry=true -do-counter-promotion=true -S | FileCheck %s --check-prefixes=CHECK,ENTRY
 
 ; Source used to create test:
 ;
 ; extern void may_throw(int);
 ; char buffer[200];
 ; void run(int count) {
 ;   try {
 ;    for (int i = 0; i < count; ++i) {

 
 for.cond:                                         ; preds = %for.inc, %entry
   %i.0 = phi i32 [ 0, %entry ], [ %inc, %for.inc ]
   %cmp = icmp slt i32 %i.0, %count
   br i1 %cmp, label %for.body, label %cleanup
 
 for.body:                                         ; preds = %for.cond
 ; CHECK: for.body:
-; CHECK: %pgocount1 = load i64, i64* getelementptr inbounds ([3 x i64], [3 x i64]* @"__profc_?run@@YAXH@Z", i64 0, i64 0)
+; NOTENTRY: %pgocount1 = load i64, i64* getelementptr inbounds ([3 x i64], [3 x i64]* @"__profc_?run@@YAXH@Z", i64 0, i64 0)
+; TENTRY: %pgocount1 = load i64, i64* getelementptr inbounds ([3 x i64], [3 x i64]* @"__profc_?run@@YAXH@Z", i64 0, i64 1)
 ; CHECK: %1 = add i64 %pgocount1, 1
-; CHECK: store i64 %1, i64* getelementptr inbounds ([3 x i64], [3 x i64]* @"__profc_?run@@YAXH@Z", i64 0, i64 0)
+; NOTENTRY: store i64 %1, i64* getelementptr inbounds ([3 x i64], [3 x i64]* @"__profc_?run@@YAXH@Z", i64 0, i64 0)
+; ENTRY: store i64 %1, i64* getelementptr inbounds ([3 x i64], [3 x i64]* @"__profc_?run@@YAXH@Z", i64 0, i64 1)
   %idxprom = zext i32 %i.0 to i64
   %arrayidx = getelementptr inbounds [200 x i8], [200 x i8]* @"?buffer@@3PADA", i64 0, i64 %idxprom
   %0 = load i8, i8* %arrayidx, align 1
   %cmp1 = icmp eq i8 %0, 0
   br i1 %cmp1, label %cleanup, label %if.end
 
 if.end:                                           ; preds = %for.body
   invoke void @"?may_throw@@YAXH@Z"(i32 %i.0)
           to label %for.inc unwind label %catch.dispatch
 
 for.inc:                                          ; preds = %if.end
 ; CHECK: for.inc:
-; CHECK: %pgocount2 = load i64, i64* getelementptr inbounds ([3 x i64], [3 x i64]* @"__profc_?run@@YAXH@Z", i64 0, i64 1)
+; NOTENTRY: %pgocount2 = load i64, i64* getelementptr inbounds ([3 x i64], [3 x i64]* @"__profc_?run@@YAXH@Z", i64 0, i64 1)
+; ENTRY: %pgocount2 = load i64, i64* getelementptr inbounds ([3 x i64], [3 x i64]* @"__profc_?run@@YAXH@Z", i64 0, i64 2)
 ; CHECK: %3 = add i64 %pgocount2, 1
-; CHECK: store i64 %3, i64* getelementptr inbounds ([3 x i64], [3 x i64]* @"__profc_?run@@YAXH@Z", i64 0, i64 1)
+; NOTENTRY: store i64 %3, i64* getelementptr inbounds ([3 x i64], [3 x i64]* @"__profc_?run@@YAXH@Z", i64 0, i64 1)
+; ENTRY: store i64 %3, i64* getelementptr inbounds ([3 x i64], [3 x i64]* @"__profc_?run@@YAXH@Z", i64 0, i64 2)
   %inc = add nuw nsw i32 %i.0, 1
   br label %for.cond
 
 cleanup:                                          ; preds = %for.body, %for.cond
   ret void
 
 catch.dispatch:                                   ; preds = %if.end
   %1 = catchswitch within none [label %catch] unwind to caller

llvm/test/Transforms/PGOProfile/criticaledge.ll

-; RUN: opt < %s -pgo-instr-gen -S | FileCheck %s --check-prefix=GEN
-; RUN: opt < %s -passes=pgo-instr-gen -S | FileCheck %s --check-prefix=GEN
+; RUN: opt < %s -pgo-instr-gen -pgo-instrument-entry=false -S | FileCheck %s --check-prefix=GEN
+; RUN: opt < %s -passes=pgo-instr-gen -pgo-instrument-entry=false -S | FileCheck %s --check-prefix=GEN
 ; RUN: llvm-profdata merge %S/Inputs/criticaledge.proftext -o %t.profdata
-; RUN: opt < %s -pgo-instr-use -pgo-test-profile-file=%t.profdata -S | FileCheck %s --check-prefix=USE
-; RUN: opt < %s -passes=pgo-instr-use -pgo-test-profile-file=%t.profdata -S | FileCheck %s --check-prefix=USE
+; RUN: opt < %s -pgo-instr-use -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -S | FileCheck %s --check-prefix=USE
+; RUN: opt < %s -passes=pgo-instr-use -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -S | FileCheck %s --check-prefix=USE
+;
+; RUN: opt < %s -pgo-instr-gen -pgo-instrument-entry=true -S | FileCheck %s --check-prefix=GEN
+; RUN: opt < %s -passes=pgo-instr-gen -pgo-instrument-entry=true -S | FileCheck %s --check-prefix=GEN
+; RUN: llvm-profdata merge %S/Inputs/criticaledge_entry.proftext -o %t2.profdata
+; RUN: opt < %s -pgo-instr-use -pgo-instrument-entry=true -pgo-test-profile-file=%t2.profdata -S | FileCheck %s --check-prefix=USE
+; RUN: opt < %s -passes=pgo-instr-use -pgo-instrument-entry=true -pgo-test-profile-file=%t2.profdata -S | FileCheck %s --check-prefix=USE
 target datalayout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128"
 target triple = "x86_64-unknown-linux-gnu"
 
 ; GEN: $__llvm_profile_raw_version = comdat any
 ; GEN: @__llvm_profile_raw_version = constant i64 {{[0-9]+}}, comdat
 ; GEN: @__profn_test_criticalEdge = private constant [17 x i8] c"test_criticalEdge"
 ; GEN: @__profn__stdin__bar = private constant [11 x i8] c"<stdin>:bar"
 
 define i32 @test_criticalEdge(i32 %i, i32 %j) {
 entry:
 ; CHECK: entry:
-; GEN-NOT: call void @llvm.instrprof.increment
+; NOTENTRY-NOT: call void @llvm.instrprof.increment
+; ENTRY:   call void @llvm.instrprof.increment(i8* getelementptr inbounds ([17 x i8], [17 x i8]* @__profn_test_criticalEdge, i32 0, i32 0), i64 82323253069, i32 8, i32 0)
   switch i32 %i, label %sw.default [
     i32 1, label %sw.bb
     i32 2, label %sw.bb1
     i32 3, label %sw.bb2
     i32 4, label %sw.bb2
 ; CHECK:    i32 3, label %entry.sw.bb2_crit_edge
 ; CHECK:    i32 4, label %entry.sw.bb2_crit_edge1
     i32 5, label %sw.bb2
   ]
 ; USE: ]
 ; USE-SAME: !prof ![[BW_SWITCH:[0-9]+]]
 
 ; CHECK: entry.sw.bb2_crit_edge1:
-; GEN:   call void @llvm.instrprof.increment(i8* getelementptr inbounds ([17 x i8], [17 x i8]* @__profn_test_criticalEdge, i32 0, i32 0), i64 82323253069, i32 8, i32 1)
+; NOTENTRY:   call void @llvm.instrprof.increment(i8* getelementptr inbounds ([17 x i8], [17 x i8]* @__profn_test_criticalEdge, i32 0, i32 0), i64 82323253069, i32 8, i32 1)
+; ENTRY: call void @llvm.instrprof.increment(i8* getelementptr inbounds ([17 x i8], [17 x i8]* @__profn_test_criticalEdge, i32 0, i32 0), i64 82323253069, i32 8, i32 2)
 ; CHECK:   br label %sw.bb2
 
 ; CHECK: entry.sw.bb2_crit_edge:
-; GEN:   call void @llvm.instrprof.increment(i8* getelementptr inbounds ([17 x i8], [17 x i8]* @__profn_test_criticalEdge, i32 0, i32 0), i64 82323253069, i32 8, i32 0)
+; NOTENTRY:   call void @llvm.instrprof.increment(i8* getelementptr inbounds ([17 x i8], [17 x i8]* @__profn_test_criticalEdge, i32 0, i32 0), i64 82323253069, i32 8, i32 0)
+; TENTRY:   call void @llvm.instrprof.increment(i8* getelementptr inbounds ([17 x i8], [17 x i8]* @__profn_test_criticalEdge, i32 0, i32 0), i64 82323253069, i32 8, i32 1)
 ; CHECK:   br label %sw.bb2
 
 sw.bb:
 ; GEN: sw.bb:
 ; GEN: call void @llvm.instrprof.increment(i8* getelementptr inbounds ([17 x i8], [17 x i8]* @__profn_test_criticalEdge, i32 0, i32 0), i64 82323253069, i32 8, i32 5)
   %call = call i32 @bar(i32 2)
   br label %sw.epilog
 
 sw.bb1:
 ; GEN: sw.bb1:
 ; GEN: call void @llvm.instrprof.increment(i8* getelementptr inbounds ([17 x i8], [17 x i8]* @__profn_test_criticalEdge, i32 0, i32 0), i64 82323253069, i32 8, i32 4)
   %call2 = call i32 @bar(i32 1024)
   br label %sw.epilog
 
 sw.bb2:
 ; GEN: sw.bb2:
 ; GEN-NOT: call void @llvm.instrprof.increment
   %cmp = icmp eq i32 %j, 2
   br i1 %cmp, label %if.then, label %if.end
 ; USE: br i1 %cmp, label %if.then, label %if.end
 ; USE-SAME: !prof ![[BW_SW_BB2:[0-9]+]]
 
 if.then:
 ; GEN: if.then:
-; GEN: call void @llvm.instrprof.increment(i8* getelementptr inbounds ([17 x i8], [17 x i8]* @__profn_test_criticalEdge, i32 0, i32 0), i64 82323253069, i32 8, i32 2)
+; NOTENTRY: call void @llvm.instrprof.increment(i8* getelementptr inbounds ([17 x i8], [17 x i8]* @__profn_test_criticalEdge, i32 0, i32 0), i64 82323253069, i32 8, i32 2)
+; ENTRY-NOT: call void @llvm.instrprof.increment
   %call4 = call i32 @bar(i32 4)
   br label %return
 
 if.end:
 ; GEN: if.end:
 ; GEN: call void @llvm.instrprof.increment(i8* getelementptr inbounds ([17 x i8], [17 x i8]* @__profn_test_criticalEdge, i32 0, i32 0), i64 82323253069, i32 8, i32 3)
   %call5 = call i32 @bar(i32 8)
   br label %sw.epilog

llvm/test/Transforms/PGOProfile/cspgo_profile_summary.ll

 ; Test the profile summary for context sensitive PGO (CSPGO)
 
 ; RUN: llvm-profdata merge %S/Inputs/cspgo.proftext -o %t.profdata
-; RUN: opt < %s -O2 -disable-preinline -pgo-kind=pgo-instr-use-pipeline -profile-file=%t.profdata -S | FileCheck %s --check-prefix=PGOSUMMARY
-; RUN: opt < %s -O2 -disable-preinline -pgo-kind=pgo-instr-use-pipeline -profile-file=%t.profdata -S -cspgo-kind=cspgo-instr-use-pipeline| FileCheck %s --check-prefix=CSPGOSUMMARY
+; RUN: opt < %s -O2 -disable-preinline -pgo-instrument-entry=false -pgo-kind=pgo-instr-use-pipeline -profile-file=%t.profdata -S | FileCheck %s --check-prefix=PGOSUMMARY
+; RUN: opt < %s -O2 -disable-preinline -pgo-instrument-entry=false -pgo-kind=pgo-instr-use-pipeline -profile-file=%t.profdata -S -cspgo-kind=cspgo-instr-use-pipeline| FileCheck %s --check-prefix=CSPGOSUMMARY
 
 target datalayout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128"
 target triple = "x86_64-unknown-linux-gnu"
 
 @odd = common dso_local global i32 0, align 4
 @even = common dso_local global i32 0, align 4
 @not_six = common dso_local global i32 0, align 4
 

llvm/test/Transforms/PGOProfile/indirectbr.ll

 ; RUN: llvm-profdata merge %S/Inputs/indirectbr.proftext -o %t.profdata
-; RUN: opt < %s -pgo-instr-use -pgo-test-profile-file=%t.profdata -S | FileCheck %s --check-prefix=USE
+; RUN: opt < %s -pgo-instr-use -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -S | FileCheck %s --check-prefix=USE
 ; New PM
-; RUN: opt < %s -passes=pgo-instr-use -pgo-test-profile-file=%t.profdata -S | FileCheck %s --check-prefix=USE
-; RUN: opt < %s -passes=pgo-instr-use -pgo-test-profile-file=%t.profdata -S | opt -S -analyze -branch-prob | FileCheck %s --check-prefix=BRANCHPROB
+; RUN: opt < %s -passes=pgo-instr-use -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -S | FileCheck %s --check-prefix=USE
+; RUN: opt < %s -passes=pgo-instr-use -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -S | opt -S -analyze -branch-prob | FileCheck %s --check-prefix=BRANCHPROB
+; RUN: llvm-profdata merge %S/Inputs/indirectbr_entry.proftext -o %t2.profdata
+; RUN: opt < %s -pgo-instr-use -pgo-instrument-entry=true -pgo-test-profile-file=%t2.profdata -S | FileCheck %s --check-prefix=USE
+; New PM
+; RUN: opt < %s -passes=pgo-instr-use -pgo-instrument-entry=true -pgo-test-profile-file=%t2.profdata -S | FileCheck %s --check-prefix=USE
+; RUN: opt < %s -passes=pgo-instr-use -pgo-instrument-entry=true -pgo-test-profile-file=%t2.profdata -S | opt -S -analyze -branch-prob | FileCheck %s --check-prefix=BRANCHPROB
 
 target datalayout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128"
 target triple = "x86_64-unknown-linux-gnu"
 
 @foo.table = internal unnamed_addr constant [3 x i8*] [i8* blockaddress(@foo, %return), i8* blockaddress(@foo, %label2), i8* blockaddress(@foo, %label3)], align 16
 
 define i32 @foo(i32 %i) {
 entry:

llvm/test/Transforms/PGOProfile/irreducible.ll

 ; RUN: llvm-profdata merge %S/Inputs/irreducible.proftext -o %t.profdata
-; RUN: opt < %s -pgo-instr-use -pgo-test-profile-file=%t.profdata -S | FileCheck %s --check-prefix=USE
-; RUN: opt < %s -passes=pgo-instr-use -pgo-test-profile-file=%t.profdata -S | FileCheck %s --check-prefix=USE
+; RUN: opt < %s -pgo-instr-use -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -S | FileCheck %s --check-prefix=USE
+; RUN: opt < %s -passes=pgo-instr-use -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -S | FileCheck %s --check-prefix=USE
+; RUN: llvm-profdata merge %S/Inputs/irreducible_entry.proftext -o %t2.profdata
+; RUN: opt < %s -pgo-instr-use -pgo-instrument-entry=true -pgo-test-profile-file=%t2.profdata -S | FileCheck %s --check-prefix=USE
+; RUN: opt < %s -passes=pgo-instr-use -pgo-instrument-entry=true -pgo-test-profile-file=%t2.profdata -S | FileCheck %s --check-prefix=USE
 
 ; GEN: $__llvm_profile_raw_version = comdat any
 
 ; Function Attrs: noinline norecurse nounwind readnone uwtable
 define i32 @_Z11irreducibleii(i32 %iter_outer, i32 %iter_inner) local_unnamed_addr #0 {
 entry:
   %cmp24 = icmp sgt i32 %iter_outer, 0
   br i1 %cmp24, label %for.body, label %entry.for.cond.cleanup_crit_edge

llvm/test/Transforms/PGOProfile/landingpad.ll

-; RUN: opt < %s -pgo-instr-gen -S | FileCheck %s --check-prefix=GEN
-; RUN: opt < %s -passes=pgo-instr-gen -S | FileCheck %s --check-prefix=GEN
+; RUN: opt < %s -pgo-instr-gen -pgo-instrument-entry=false -S | FileCheck %s --check-prefixes=GEN,NOTENTRY
+; RUN: opt < %s -passes=pgo-instr-gen -pgo-instrument-entry=false -S | FileCheck %s --check-prefixes=GEN,NOTENTRY
 ; RUN: llvm-profdata merge %S/Inputs/landingpad.proftext -o %t.profdata
-; RUN: opt < %s -pgo-instr-use -pgo-test-profile-file=%t.profdata -S | FileCheck %s --check-prefix=USE
-; RUN: opt < %s -passes=pgo-instr-use -pgo-test-profile-file=%t.profdata -S | FileCheck %s --check-prefix=USE
+; RUN: opt < %s -pgo-instr-use -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -S | FileCheck %s --check-prefix=USE
+; RUN: opt < %s -passes=pgo-instr-use -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -S | FileCheck %s --check-prefix=USE
+;
+; RUN: opt < %s -pgo-instr-gen -pgo-instrument-entry=true -S | FileCheck %s --check-prefixes=GEN,ENTRY
+; RUN: opt < %s -passes=pgo-instr-gen -pgo-instrument-entry=true -S | FileCheck %s --check-prefixes=GEN,ENTRY
+; RUN: llvm-profdata merge %S/Inputs/landingpad_entry.proftext -o %t2.profdata
+; RUN: opt < %s -pgo-instr-use -pgo-instrument-entry=true -pgo-test-profile-file=%t2.profdata -S | FileCheck %s --check-prefix=USE
+; RUN: opt < %s -passes=pgo-instr-use -pgo-instrument-entry=true -pgo-test-profile-file=%t2.profdata -S | FileCheck %s --check-prefix=USE
 target datalayout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128"
 target triple = "x86_64-unknown-linux-gnu"
 
 @val = global i32 0, align 4
 @_ZTIi = external constant i8*
 ; GEN: $__llvm_profile_raw_version = comdat any
 ; GEN: @__llvm_profile_raw_version = constant i64 {{[0-9]+}}, comdat
 ; GEN: @__profn_bar = private constant [3 x i8] c"bar"

 
 declare i8* @__cxa_allocate_exception(i64)
 
 declare void @__cxa_throw(i8*, i8*, i8*)
 
 define i32 @foo(i32 %i) personality i8* bitcast (i32 (...)* @__gxx_personality_v0 to i8*) {
 entry:
 ; GEN: entry:
-; GEN-NOT: call void @llvm.instrprof.increment
+; NOTENTRY-NOT: call void @llvm.instrprof.increment
+; ENTRY: call void @llvm.instrprof.increment(i8* getelementptr inbounds ([3 x i8], [3 x i8]* @__profn_foo, i32 0, i32 0), i64 59130013419, i32 4, i32 0)
   %rem = srem i32 %i, 2
   %tobool = icmp ne i32 %rem, 0
   br i1 %tobool, label %if.then, label %if.end
 ; USE: br i1 %tobool, label %if.then, label %if.end
 ; USE-SAME: !prof ![[BW_FOO_ENTRY:[0-9]+]]
 
 if.then:
 ; GEN: if.then:

 
 try.cont:
 ; GEN: try.cont:
 ; GEN-NOT: call void @llvm.instrprof.increment
   ret i32 -1
 
 if.end:
 ; GEN: if.end:
-; GEN: call void @llvm.instrprof.increment(i8* getelementptr inbounds ([3 x i8], [3 x i8]* @__profn_foo, i32 0, i32 0), i64 59130013419, i32 4, i32 0)
+; NOTENTRY: call void @llvm.instrprof.increment(i8* getelementptr inbounds ([3 x i8], [3 x i8]* @__profn_foo, i32 0, i32 0), i64 59130013419, i32 4, i32 0)
+; ENTRY-NOT: call void @llvm.instrprof.increment
   %tmp8 = load i32, i32* @val, align 4
   %add = add nsw i32 %tmp8, %i
   store i32 %add, i32* @val, align 4
   br label %try.cont
 
 eh.resume:
 ; GEN: eh.resume:
 ; GEN: call void @llvm.instrprof.increment(i8* getelementptr inbounds ([3 x i8], [3 x i8]* @__profn_foo, i32 0, i32 0), i64 59130013419, i32 4, i32 3)

llvm/test/Transforms/PGOProfile/loop1.ll

-; RUN: opt < %s -pgo-instr-gen -S | FileCheck %s --check-prefix=GEN
-; RUN: opt < %s -passes=pgo-instr-gen -S | FileCheck %s --check-prefix=GEN
+; RUN: opt < %s -pgo-instr-gen -pgo-instrument-entry=false -S | FileCheck %s --check-prefixes=GEN,NOTENTRY
+; RUN: opt < %s -passes=pgo-instr-gen -pgo-instrument-entry=false -S | FileCheck %s --check-prefixes=GEN,NOTENTRY
 ; RUN: llvm-profdata merge %S/Inputs/loop1.proftext -o %t.profdata
-; RUN: opt < %s -pgo-instr-use -pgo-test-profile-file=%t.profdata -S | FileCheck %s --check-prefix=USE
-; RUN: opt < %s -passes=pgo-instr-use -pgo-test-profile-file=%t.profdata -S | FileCheck %s --check-prefix=USE
+; RUN: opt < %s -pgo-instr-use -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -S | FileCheck %s --check-prefix=USE
+; RUN: opt < %s -passes=pgo-instr-use -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -S | FileCheck %s --check-prefix=USE
+; RUN: opt < %s -pgo-instr-gen -pgo-instrument-entry=true -S | FileCheck %s --check-prefixes=GEN,ENTRY
+; RUN: opt < %s -passes=pgo-instr-gen -pgo-instrument-entry=true -S | FileCheck %s --check-prefixes=GEN,ENTRY
+; RUN: llvm-profdata merge %S/Inputs/loop1_entry.proftext -o %t.profdata
+; RUN: opt < %s -pgo-instr-use -pgo-instrument-entry=true -pgo-test-profile-file=%t.profdata -S | FileCheck %s --check-prefix=USE
+; RUN: opt < %s -passes=pgo-instr-use -pgo-instrument-entry=true -pgo-test-profile-file=%t.profdata -S | FileCheck %s --check-prefix=USE
+
 target datalayout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128"
 target triple = "x86_64-unknown-linux-gnu"
 
 ; GEN: $__llvm_profile_raw_version = comdat any
 ; GEN: @__llvm_profile_raw_version = constant i64 {{[0-9]+}}, comdat
 ; GEN: @__profn_test_simple_for = private constant [15 x i8] c"test_simple_for"
 
 define i32 @test_simple_for(i32 %n) {
 entry:
 ; GEN: entry:
-; GEN: call void @llvm.instrprof.increment(i8* getelementptr inbounds ([15 x i8], [15 x i8]* @__profn_test_simple_for, i32 0, i32 0), i64 34137660316, i32 2, i32 1)
+; NOTENTRY: call void @llvm.instrprof.increment(i8* getelementptr inbounds ([15 x i8], [15 x i8]* @__profn_test_simple_for, i32 0, i32 0), i64 34137660316, i32 2, i32 1)
+; ENTRY: call void @llvm.instrprof.increment(i8* getelementptr inbounds ([15 x i8], [15 x i8]* @__profn_test_simple_for, i32 0, i32 0), i64 34137660316, i32 2, i32 0)
   br label %for.cond
 
 for.cond:
 ; GEN: for.cond:
 ; GEN-NOT: call void @llvm.instrprof.increment
   %i = phi i32 [ 0, %entry ], [ %inc1, %for.inc ]
   %sum = phi i32 [ 1, %entry ], [ %inc, %for.inc ]
   %cmp = icmp slt i32 %i, %n
   br i1 %cmp, label %for.body, label %for.end
 ; USE: br i1 %cmp, label %for.body, label %for.end
 ; USE-SAME: !prof ![[BW_FOR_COND:[0-9]+]]
 ; USE: ![[BW_FOR_COND]] = !{!"branch_weights", i32 96, i32 4}
 
 for.body:
 ; GEN: for.body:
 ; GEN-NOT: call void @llvm.instrprof.increment
   %inc = add nsw i32 %sum, 1
   br label %for.inc
 
 for.inc:
 ; GEN: for.inc:
-; GEN: call void @llvm.instrprof.increment(i8* getelementptr inbounds ([15 x i8], [15 x i8]* @__profn_test_simple_for, i32 0, i32 0), i64 34137660316, i32 2, i32 0)
+; NOTENTRY: call void @llvm.instrprof.increment(i8* getelementptr inbounds ([15 x i8], [15 x i8]* @__profn_test_simple_for, i32 0, i32 0), i64 34137660316, i32 2, i32 0)
+; ENTRY: call void @llvm.instrprof.increment(i8* getelementptr inbounds ([15 x i8], [15 x i8]* @__profn_test_simple_for, i32 0, i32 0), i64 34137660316, i32 2, i32 1)
   %inc1 = add nsw i32 %i, 1
   br label %for.cond
 
 for.end:
 ; GEN: for.end:
 ; GEN-NOT: call void @llvm.instrprof.increment
 ; GEN: ret i32
   ret i32 %sum
 }

llvm/test/Transforms/PGOProfile/loop2.ll

-; RUN: opt < %s -pgo-instr-gen -S | FileCheck %s --check-prefix=GEN
-; RUN: opt < %s -passes=pgo-instr-gen -S | FileCheck %s --check-prefix=GEN
+; RUN: opt < %s -pgo-instr-gen -pgo-instrument-entry=false -S | FileCheck %s --check-prefixes=GEN,NOTENTRY
+; RUN: opt < %s -passes=pgo-instr-gen -pgo-instrument-entry=false -S | FileCheck %s --check-prefixes=GEN,NOTENTRY
 ; RUN: llvm-profdata merge %S/Inputs/loop2.proftext -o %t.profdata
-; RUN: opt < %s -pgo-instr-use -pgo-test-profile-file=%t.profdata -S | FileCheck %s --check-prefix=USE
-; RUN: opt < %s -passes=pgo-instr-use -pgo-test-profile-file=%t.profdata -S | FileCheck %s --check-prefix=USE
+; RUN: opt < %s -pgo-instr-use -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -S | FileCheck %s --check-prefix=USE
+; RUN: opt < %s -passes=pgo-instr-use -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -S | FileCheck %s --check-prefix=USE
+; RUN: opt < %s -pgo-instr-gen -pgo-instrument-entry=true -S | FileCheck %s --check-prefixes=GEN,ENTRY
+; RUN: opt < %s -passes=pgo-instr-gen -pgo-instrument-entry=true -S | FileCheck %s --check-prefixes=GEN,ENTRY
+; RUN: llvm-profdata merge %S/Inputs/loop2_entry.proftext -o %t.profdata
+; RUN: opt < %s -pgo-instr-use -pgo-instrument-entry=true -pgo-test-profile-file=%t.profdata -S | FileCheck %s --check-prefix=USE
+; RUN: opt < %s -passes=pgo-instr-use -pgo-instrument-entry=true -pgo-test-profile-file=%t.profdata -S | FileCheck %s --check-prefix=USE
 target datalayout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128"
 target triple = "x86_64-unknown-linux-gnu"
 
 ; GEN: $__llvm_profile_raw_version = comdat any
 ; GEN: @__llvm_profile_raw_version = constant i64 {{[0-9]+}}, comdat
 ; GEN: @__profn_test_nested_for = private constant [15 x i8] c"test_nested_for"
 
 define i32 @test_nested_for(i32 %r, i32 %s) {
 entry:
 ; GEN: entry:
-; GEN: call void @llvm.instrprof.increment(i8* getelementptr inbounds ([15 x i8], [15 x i8]* @__profn_test_nested_for, i32 0, i32 0), i64 53929068288, i32 3, i32 2)
+; NOTENTRY: call void @llvm.instrprof.increment(i8* getelementptr inbounds ([15 x i8], [15 x i8]* @__profn_test_nested_for, i32 0, i32 0), i64 53929068288, i32 3, i32 2)
+; ENTRY: call void @llvm.instrprof.increment(i8* getelementptr inbounds ([15 x i8], [15 x i8]* @__profn_test_nested_for, i32 0, i32 0), i64 53929068288, i32 3, i32 0)
   br label %for.cond.outer
 
 for.cond.outer:
 ; GEN: for.cond.outer:
 ; GEN-NOT: call void @llvm.instrprof.increment
   %i.0 = phi i32 [ 0, %entry ], [ %inc.2, %for.inc.outer ]
   %sum.0 = phi i32 [ 1, %entry ], [ %sum.1, %for.inc.outer ]
   %cmp = icmp slt i32 %i.0, %r

 for.body.inner:
 ; GEN: for.body.inner:
 ; GEN-NOT: call void @llvm.instrprof.increment
   %inc = add nsw i32 %sum.1, 1
   br label %for.inc.inner
 
 for.inc.inner:
 ; GEN: for.inc.inner:
-; GEN: call void @llvm.instrprof.increment(i8* getelementptr inbounds ([15 x i8], [15 x i8]* @__profn_test_nested_for, i32 0, i32 0), i64 53929068288, i32 3, i32 0)
+; NOTENTRY: call void @llvm.instrprof.increment(i8* getelementptr inbounds ([15 x i8], [15 x i8]* @__profn_test_nested_for, i32 0, i32 0), i64 53929068288, i32 3, i32 0)
+; ENTRY: call void @llvm.instrprof.increment(i8* getelementptr inbounds ([15 x i8], [15 x i8]* @__profn_test_nested_for, i32 0, i32 0), i64 53929068288, i32 3, i32 1)
   %inc.1 = add nsw i32 %j.0, 1
   br label %for.cond.inner
 
 for.end.inner:
 ; GEN: for.end.inner:
   br label %for.inc.outer
 
 for.inc.outer:
 ; GEN: for.inc.outer:
-; GEN: call void @llvm.instrprof.increment(i8* getelementptr inbounds ([15 x i8], [15 x i8]* @__profn_test_nested_for, i32 0, i32 0), i64 53929068288, i32 3, i32 1)
+; NOTENTRY: call void @llvm.instrprof.increment(i8* getelementptr inbounds ([15 x i8], [15 x i8]* @__profn_test_nested_for, i32 0, i32 0), i64 53929068288, i32 3, i32 1)
+; ENTRY: call void @llvm.instrprof.increment(i8* getelementptr inbounds ([15 x i8], [15 x i8]* @__profn_test_nested_for, i32 0, i32 0), i64 53929068288, i32 3, i32 2)
   %inc.2 = add nsw i32 %i.0, 1
   br label %for.cond.outer
 
 for.end.outer:
 ; GEN: for.end.outer:
 ; GEN-NOT: call void @llvm.instrprof.increment
 ; GEN: ret i32
   ret i32 %sum.0
 }
 
 ; USE-DAG: ![[BW_FOR_COND_OUTER]] = !{!"branch_weights", i32 10, i32 6}
 ; USE-DAG: ![[BW_FOR_COND_INNER]] = !{!"branch_weights", i32 33, i32 10}
 

llvm/test/Transforms/PGOProfile/misexpect-branch-stripped.ll

-
 ; RUN: llvm-profdata merge %S/Inputs/misexpect-branch.proftext -o %t.profdata
 
-
-; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-test-profile-file=%t.profdata -S -pgo-warn-misexpect 2>&1 | FileCheck %s --check-prefix=WARNING
-; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-test-profile-file=%t.profdata -S -pass-remarks=misexpect 2>&1 | FileCheck %s --check-prefix=REMARK
-; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-test-profile-file=%t.profdata -S -pgo-warn-misexpect -pass-remarks=misexpect 2>&1 | FileCheck %s --check-prefix=BOTH
-; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-test-profile-file=%t.profdata -S 2>&1 | FileCheck %s --check-prefix=DISABLED
+; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -S -pgo-warn-misexpect 2>&1 | FileCheck %s --check-prefix=WARNING
+; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -S -pass-remarks=misexpect 2>&1 | FileCheck %s --check-prefix=REMARK
+; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -S -pgo-warn-misexpect -pass-remarks=misexpect 2>&1 | FileCheck %s --check-prefix=BOTH
+; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -S 2>&1 | FileCheck %s --check-prefix=DISABLED
 
 ; New PM
-; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-test-profile-file=%t.profdata -pgo-warn-misexpect -S 2>&1 | FileCheck %s --check-prefix=WARNING
-; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-test-profile-file=%t.profdata -pass-remarks=misexpect -S 2>&1 | FileCheck %s --check-prefix=REMARK
-; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-test-profile-file=%t.profdata -pgo-warn-misexpect -pass-remarks=misexpect -S 2>&1 | FileCheck %s --check-prefix=BOTH
-; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-test-profile-file=%t.profdata -S 2>&1 | FileCheck %s --check-prefix=DISABLED
-
-
+; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -pgo-warn-misexpect -S 2>&1 | FileCheck %s --check-prefix=WARNING
+; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -pass-remarks=misexpect -S 2>&1 | FileCheck %s --check-prefix=REMARK
+; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -pgo-warn-misexpect -pass-remarks=misexpect -S 2>&1 | FileCheck %s --check-prefix=BOTH
+; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -S 2>&1 | FileCheck %s --check-prefix=DISABLED
+
+; RUN: llvm-profdata merge %S/Inputs/misexpect-branch_entry.proftext -o %t2.profdata
+
+; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-instrument-entry=true -pgo-test-profile-file=%t2.profdata -S -pgo-warn-misexpect 2>&1 | FileCheck %s --check-prefix=WARNING
+; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-instrument-entry=true -pgo-test-profile-file=%t2.profdata -S -pass-remarks=misexpect 2>&1 | FileCheck %s --check-prefix=REMARK
+; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-instrument-entry=true -pgo-test-profile-file=%t2.profdata -S -pgo-warn-misexpect -pass-remarks=misexpect 2>&1 | FileCheck %s --check-prefix=BOTH
+; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-instrument-entry=true -pgo-test-profile-file=%t2.profdata -S 2>&1 | FileCheck %s --check-prefix=DISABLED
 
+; New PM
+; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-instrument-entry=true -pgo-test-profile-file=%t2.profdata -pgo-warn-misexpect -S 2>&1 | FileCheck %s --check-prefix=WARNING
+; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-instrument-entry=true -pgo-test-profile-file=%t2.profdata -pass-remarks=misexpect -S 2>&1 | FileCheck %s --check-prefix=REMARK
+; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-instrument-entry=true -pgo-test-profile-file=%t2.profdata -pgo-warn-misexpect -pass-remarks=misexpect -S 2>&1 | FileCheck %s --check-prefix=BOTH
+; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-instrument-entry=true -pgo-test-profile-file=%t2.profdata -S 2>&1 | FileCheck %s --check-prefix=DISABLED
 ; WARNING-DAG: warning: <unknown>:0:0: 19.98%
 ; WARNING-NOT: remark: <unknown>:0:0: Potential performance regression from use of the llvm.expect intrinsic: Annotation was correct on 19.98% (399668 / 2000000) of profiled executions.
 
 ; REMARK-NOT: warning: <unknown>:0:0: 19.98%
 ; REMARK-DAG: remark: <unknown>:0:0: Potential performance regression from use of the llvm.expect intrinsic: Annotation was correct on 19.98% (399668 / 2000000) of profiled executions.
 
 ; BOTH-DAG: warning: <unknown>:0:0: 19.98%
 ; BOTH-DAG: remark: <unknown>:0:0: Potential performance regression from use of the llvm.expect intrinsic: Annotation was correct on 19.98% (399668 / 2000000) of profiled executions.

llvm/test/Transforms/PGOProfile/misexpect-branch.ll

-
 ; RUN: llvm-profdata merge %S/Inputs/misexpect-branch.proftext -o %t.profdata
 
-
-; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-test-profile-file=%t.profdata -S -pgo-warn-misexpect 2>&1 | FileCheck %s --check-prefix=WARNING
-; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-test-profile-file=%t.profdata -S -pass-remarks=misexpect 2>&1 | FileCheck %s --check-prefix=REMARK
-; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-test-profile-file=%t.profdata -S -pgo-warn-misexpect -pass-remarks=misexpect 2>&1 | FileCheck %s --check-prefix=BOTH
-; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-test-profile-file=%t.profdata -S 2>&1 | FileCheck %s --check-prefix=DISABLED
+; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -S -pgo-warn-misexpect 2>&1 | FileCheck %s --check-prefix=WARNING
+; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -S -pass-remarks=misexpect 2>&1 | FileCheck %s --check-prefix=REMARK
+; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -S -pgo-warn-misexpect -pass-remarks=misexpect 2>&1 | FileCheck %s --check-prefix=BOTH
+; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -S 2>&1 | FileCheck %s --check-prefix=DISABLED
 
 ; New PM
-; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-test-profile-file=%t.profdata -pgo-warn-misexpect -S 2>&1 | FileCheck %s --check-prefix=WARNING
-; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-test-profile-file=%t.profdata -pass-remarks=misexpect -S 2>&1 | FileCheck %s --check-prefix=REMARK
-; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-test-profile-file=%t.profdata -pgo-warn-misexpect -pass-remarks=misexpect -S 2>&1 | FileCheck %s --check-prefix=BOTH
-; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-test-profile-file=%t.profdata -S 2>&1 | FileCheck %s --check-prefix=DISABLED
-
+; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -pgo-warn-misexpect -S 2>&1 | FileCheck %s --check-prefix=WARNING
+; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -pass-remarks=misexpect -S 2>&1 | FileCheck %s --check-prefix=REMARK
+; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -pgo-warn-misexpect -pass-remarks=misexpect -S 2>&1 | FileCheck %s --check-prefix=BOTH
+; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -S 2>&1 | FileCheck %s --check-prefix=DISABLED
+
+; RUN: llvm-profdata merge %S/Inputs/misexpect-branch_entry.proftext -o %t2.profdata
+
+; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-instrument-entry=true -pgo-test-profile-file=%t2.profdata -S -pgo-warn-misexpect 2>&1 | FileCheck %s --check-prefix=WARNING
+; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-instrument-entry=true -pgo-test-profile-file=%t2.profdata -S -pass-remarks=misexpect 2>&1 | FileCheck %s --check-prefix=REMARK
+; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-instrument-entry=true -pgo-test-profile-file=%t2.profdata -S -pgo-warn-misexpect -pass-remarks=misexpect 2>&1 | FileCheck %s --check-prefix=BOTH
+; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-instrument-entry=true -pgo-test-profile-file=%t2.profdata -S 2>&1 | FileCheck %s --check-prefix=DISABLED
 
+; New PM
+; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-instrument-entry=true -pgo-test-profile-file=%t2.profdata -pgo-warn-misexpect -S 2>&1 | FileCheck %s --check-prefix=WARNING
+; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-instrument-entry=true -pgo-test-profile-file=%t2.profdata -pass-remarks=misexpect -S 2>&1 | FileCheck %s --check-prefix=REMARK
+; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-instrument-entry=true -pgo-test-profile-file=%t2.profdata -pgo-warn-misexpect -pass-remarks=misexpect -S 2>&1 | FileCheck %s --check-prefix=BOTH
+; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-instrument-entry=true -pgo-test-profile-file=%t2.profdata -S 2>&1 | FileCheck %s --check-prefix=DISABLED
 
 ; WARNING-DAG: warning: misexpect-branch.c:22:0: 19.98%
 ; WARNING-NOT: remark: misexpect-branch.c:22:0: Potential performance regression from use of the llvm.expect intrinsic: Annotation was correct on 19.98% (399668 / 2000000) of profiled executions.
 
 ; REMARK-NOT: warning: misexpect-branch.c:22:0: 19.98%
 ; REMARK-DAG: remark: misexpect-branch.c:22:0: Potential performance regression from use of the llvm.expect intrinsic: Annotation was correct on 19.98% (399668 / 2000000) of profiled executions.
 
 ; BOTH-DAG: warning: misexpect-branch.c:22:0: 19.98%

llvm/test/Transforms/PGOProfile/misexpect-switch-default.ll

-
 ; RUN: llvm-profdata merge %S/Inputs/misexpect-switch.proftext -o %t.profdata
 
-; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-test-profile-file=%t.profdata -S -pgo-warn-misexpect 2>&1 | FileCheck %s --check-prefix=WARNING
-; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-test-profile-file=%t.profdata -S -pass-remarks=misexpect 2>&1 | FileCheck %s --check-prefix=REMARK
-; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-test-profile-file=%t.profdata -S -pgo-warn-misexpect -pass-remarks=misexpect 2>&1 | FileCheck %s --check-prefix=BOTH
-; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-test-profile-file=%t.profdata -S 2>&1 | FileCheck %s --check-prefix=DISABLED
+; RUN: opt < %s -lower-expect -pgo-instrument-entry=false -pgo-instr-use -pgo-test-profile-file=%t.profdata -S -pgo-warn-misexpect 2>&1 | FileCheck %s --check-prefix=WARNING
+; RUN: opt < %s -lower-expect -pgo-instrument-entry=false -pgo-instr-use -pgo-test-profile-file=%t.profdata -S -pass-remarks=misexpect 2>&1 | FileCheck %s --check-prefix=REMARK
+; RUN: opt < %s -lower-expect -pgo-instrument-entry=false -pgo-instr-use -pgo-test-profile-file=%t.profdata -S -pgo-warn-misexpect -pass-remarks=misexpect 2>&1 | FileCheck %s --check-prefix=BOTH
+; RUN: opt < %s -lower-expect -pgo-instrument-entry=false -pgo-instr-use -pgo-test-profile-file=%t.profdata -S 2>&1 | FileCheck %s --check-prefix=DISABLED
 
 ; New PM
-; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-test-profile-file=%t.profdata -pgo-warn-misexpect -S 2>&1 | FileCheck %s --check-prefix=WARNING
-; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-test-profile-file=%t.profdata -pass-remarks=misexpect -S 2>&1 | FileCheck %s --check-prefix=REMARK
-; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-test-profile-file=%t.profdata -pgo-warn-misexpect -pass-remarks=misexpect -S 2>&1 | FileCheck %s --check-prefix=BOTH
-; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-test-profile-file=%t.profdata -S 2>&1 | FileCheck %s --check-prefix=DISABLED
+; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -pgo-warn-misexpect -S 2>&1 | FileCheck %s --check-prefix=WARNING
+; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -pass-remarks=misexpect -S 2>&1 | FileCheck %s --check-prefix=REMARK
+; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -pgo-warn-misexpect -pass-remarks=misexpect -S 2>&1 | FileCheck %s --check-prefix=BOTH
+; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -S 2>&1 | FileCheck %s --check-prefix=DISABLED
 
 ; WARNING-DAG: warning: <unknown>:0:0: 0.00%
 ; WARNING-NOT: remark: <unknown>:0:0: Potential performance regression from use of the llvm.expect intrinsic: Annotation was correct on 0.00% (0 / 27943) of profiled executions.
 
 ; REMARK-NOT: warning: <unknown>:0:0: 0.00%
 ; REMARK-DAG: remark: <unknown>:0:0: Potential performance regression from use of the llvm.expect intrinsic: Annotation was correct on 0.00% (0 / 27943) of profiled executions.
 
 ; BOTH-DAG: warning: <unknown>:0:0: 0.00%
 ; BOTH-DAG: remark: <unknown>:0:0: Potential performance regression from use of the llvm.expect intrinsic: Annotation was correct on 0.00% (0 / 27943) of profiled executions.
 
 ; DISABLED-NOT: warning: <unknown>:0:0: 0.00%
 ; DISABLED-NOT: remark: <unknown>:0:0: Potential performance regression from use of the llvm.expect intrinsic: Annotation was correct on 0.00% (0 / 27943) of profiled executions.
 
 ; DISABLED-NOT: warning: <unknown>:0:0: 0.00%
 ; DISABLED-NOT: remark: <unknown>:0:0: Potential performance regression from use of the llvm.expect intrinsic: Annotation was correct on 0.00% (0 / 27943) of profiled executions.
 
 ; CORRECT-NOT: warning: {{.*}}
 ; CORRECT-NOT: remark: {{.*}}
 ; CHECK-DAG: !{!"misexpect", i64 0, i64 2000, i64 1}
 
 
-
 ; ModuleID = 'misexpect-switch.c'
 source_filename = "misexpect-switch.c"
 target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
 target triple = "x86_64-unknown-linux-gnu"
 
 @inner_loop = dso_local constant i32 1000, align 4
 @outer_loop = dso_local constant i32 20, align 4
 @arry_size = dso_local constant i32 25, align 4

llvm/test/Transforms/PGOProfile/misexpect-switch.ll

-
 ; RUN: llvm-profdata merge %S/Inputs/misexpect-switch.proftext -o %t.profdata
 ; RUN: llvm-profdata merge %S/Inputs/misexpect-switch-correct.proftext -o %t.c.profdata
 
-; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-test-profile-file=%t.profdata -S -pgo-warn-misexpect 2>&1 | FileCheck %s --check-prefix=WARNING
-; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-test-profile-file=%t.profdata -S -pass-remarks=misexpect 2>&1 | FileCheck %s --check-prefix=REMARK
-; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-test-profile-file=%t.profdata -S -pgo-warn-misexpect -pass-remarks=misexpect 2>&1 | FileCheck %s --check-prefix=BOTH
-; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-test-profile-file=%t.profdata -S 2>&1 | FileCheck %s --check-prefix=DISABLED
+; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -S -pgo-warn-misexpect 2>&1 | FileCheck %s --check-prefix=WARNING
+; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -S -pass-remarks=misexpect 2>&1 | FileCheck %s --check-prefix=REMARK
+; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -S -pgo-warn-misexpect -pass-remarks=misexpect 2>&1 | FileCheck %s --check-prefix=BOTH
+; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -S 2>&1 | FileCheck %s --check-prefix=DISABLED
 
 ; New PM
-; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-test-profile-file=%t.profdata -pgo-warn-misexpect -S 2>&1 | FileCheck %s --check-prefix=WARNING
-; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-test-profile-file=%t.profdata -pass-remarks=misexpect -S 2>&1 | FileCheck %s --check-prefix=REMARK
-; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-test-profile-file=%t.profdata -pgo-warn-misexpect -pass-remarks=misexpect -S 2>&1 | FileCheck %s --check-prefix=BOTH
-; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-test-profile-file=%t.profdata -S 2>&1 | FileCheck %s --check-prefix=DISABLED
+; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -pgo-warn-misexpect -S 2>&1 | FileCheck %s --check-prefix=WARNING
+; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -pass-remarks=misexpect -S 2>&1 | FileCheck %s --check-prefix=REMARK
+; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -pgo-warn-misexpect -pass-remarks=misexpect -S 2>&1 | FileCheck %s --check-prefix=BOTH
+; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -S 2>&1 | FileCheck %s --check-prefix=DISABLED
+
+; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-instrument-entry=false -pgo-test-profile-file=%t.c.profdata -S -pgo-warn-misexpect -pass-remarks=misexpect 2>&1 | FileCheck %s --check-prefix=CORRECT
+; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-instrument-entry=false -pgo-test-profile-file=%t.c.profdata -pgo-warn-misexpect -pass-remarks=misexpect -S 2>&1 | FileCheck %s --check-prefix=CORRECT
+
+; RUN: llvm-profdata merge %S/Inputs/misexpect-switch_entry.proftext -o %t2.profdata
+; RUN: llvm-profdata merge %S/Inputs/misexpect-switch-correct_entry.proftext -o %t2.c.profdata
+
+; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-instrument-entry=true -pgo-test-profile-file=%t2.profdata -S -pgo-warn-misexpect 2>&1 | FileCheck %s --check-prefix=WARNING
+; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-instrument-entry=true -pgo-test-profile-file=%t2.profdata -S -pass-remarks=misexpect 2>&1 | FileCheck %s --check-prefix=REMARK
+; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-instrument-entry=true -pgo-test-profile-file=%t2.profdata -S -pgo-warn-misexpect -pass-remarks=misexpect 2>&1 | FileCheck %s --check-prefix=BOTH
+; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-instrument-entry=true -pgo-test-profile-file=%t2.profdata -S 2>&1 | FileCheck %s --check-prefix=DISABLED
 
-; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-test-profile-file=%t.c.profdata -S -pgo-warn-misexpect -pass-remarks=misexpect 2>&1 | FileCheck %s --check-prefix=CORRECT
-; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-test-profile-file=%t.c.profdata -pgo-warn-misexpect -pass-remarks=misexpect -S 2>&1 | FileCheck %s --check-prefix=CORRECT
+; New PM
+; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-instrument-entry=true -pgo-test-profile-file=%t2.profdata -pgo-warn-misexpect -S 2>&1 | FileCheck %s --check-prefix=WARNING
+; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-instrument-entry=true -pgo-test-profile-file=%t2.profdata -pass-remarks=misexpect -S 2>&1 | FileCheck %s --check-prefix=REMARK
+; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-instrument-entry=true -pgo-test-profile-file=%t2.profdata -pgo-warn-misexpect -pass-remarks=misexpect -S 2>&1 | FileCheck %s --check-prefix=BOTH
+; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-instrument-entry=true -pgo-test-profile-file=%t2.profdata -S 2>&1 | FileCheck %s --check-prefix=DISABLED
 
+; RUN: opt < %s -lower-expect -pgo-instr-use -pgo-instrument-entry=true -pgo-test-profile-file=%t2.c.profdata -S -pgo-warn-misexpect -pass-remarks=misexpect 2>&1 | FileCheck %s --check-prefix=CORRECT
+; RUN: opt < %s -passes="function(lower-expect),pgo-instr-use" -pgo-instrument-entry=true -pgo-test-profile-file=%t2.c.profdata -pgo-warn-misexpect -pass-remarks=misexpect -S 2>&1 | FileCheck %s --check-prefix=CORRECT
 ; WARNING-DAG: warning: misexpect-switch.c:26:5: 0.00%
 ; WARNING-NOT: remark: misexpect-switch.c:26:5: Potential performance regression from use of the llvm.expect intrinsic: Annotation was correct on 0.00% (0 / 8112) of profiled executions.
 
 ; REMARK-NOT: warning: misexpect-switch.c:26:5: 0.00%
 ; REMARK-DAG: remark: misexpect-switch.c:26:5: Potential performance regression from use of the llvm.expect intrinsic: Annotation was correct on 0.00% (0 / 8112) of profiled executions.
 
 ; BOTH-DAG: warning: misexpect-switch.c:26:5: 0.00%
 ; BOTH-DAG: remark: misexpect-switch.c:26:5: Potential performance regression from use of the llvm.expect intrinsic: Annotation was correct on 0.00% (0 / 8112) of profiled executions.
 
 ; DISABLED-NOT: warning: misexpect-switch.c:26:5: 0.00%
 ; DISABLED-NOT: remark: misexpect-switch.c:26:5: Potential performance regression from use of the llvm.expect intrinsic: Annotation was correct on 0.00% (0 / 8112) of profiled executions.
 
 ; DISABLED-NOT: warning: misexpect-switch.c:26:5: 0.00%
 ; DISABLED-NOT: remark: misexpect-switch.c:26:5: Potential performance regression from use of the llvm.expect intrinsic: Annotation was correct on 0.00% (0 / 8112) of profiled executions.
 
 ; CORRECT-NOT: warning: {{.*}}
 ; CORRECT-NOT: remark: {{.*}}
 ; CHECK-DAG: !{!"misexpect", i64 0, i64 2000, i64 1}
 
-
-
 ; ModuleID = 'misexpect-switch.c'
 source_filename = "misexpect-switch.c"
 target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
 target triple = "x86_64-unknown-linux-gnu"
 
 @inner_loop = dso_local constant i32 1000, align 4, !dbg !0
 @outer_loop = dso_local constant i32 20, align 4, !dbg !6
 @arry_size = dso_local constant i32 25, align 4, !dbg !10

llvm/test/Transforms/PGOProfile/switch.ll

-; RUN: opt < %s -pgo-instr-gen -S | FileCheck %s --check-prefix=GEN
-; RUN: opt < %s -passes=pgo-instr-gen -S | FileCheck %s --check-prefix=GEN
+; RUN: opt < %s -pgo-instr-gen -pgo-instrument-entry=false -S | FileCheck %s --check-prefixes=GEN,NOTENTRY
+; RUN: opt < %s -passes=pgo-instr-gen -pgo-instrument-entry=false -S | FileCheck %s --check-prefixes=GEN,NOTENTRY
 ; RUN: llvm-profdata merge %S/Inputs/switch.proftext -o %t.profdata
-; RUN: opt < %s -pgo-instr-use -pgo-test-profile-file=%t.profdata -S | FileCheck %s --check-prefix=USE
-; RUN: opt < %s -passes=pgo-instr-use -pgo-test-profile-file=%t.profdata -S | FileCheck %s --check-prefix=USE
+; RUN: opt < %s -pgo-instr-use -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -S | FileCheck %s --check-prefix=USE
+; RUN: opt < %s -passes=pgo-instr-use -pgo-instrument-entry=false -pgo-test-profile-file=%t.profdata -S | FileCheck %s --check-prefix=USE
+; RUN: opt < %s -pgo-instr-gen -pgo-instrument-entry=true -S | FileCheck %s --check-prefixes=GEN,ENTRY
+; RUN: opt < %s -passes=pgo-instr-gen -pgo-instrument-entry=true -S | FileCheck %s --check-prefixes=GEN,ENTRY
+; RUN: llvm-profdata merge %S/Inputs/switch_entry.proftext -o %t.profdata
+; RUN: opt < %s -pgo-instr-use -pgo-instrument-entry=true -pgo-test-profile-file=%t.profdata -S | FileCheck %s --check-prefix=USE
+; RUN: opt < %s -passes=pgo-instr-use -pgo-instrument-entry=true -pgo-test-profile-file=%t.profdata -S | FileCheck %s --check-prefix=USE
 target datalayout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128"
 target triple = "x86_64-unknown-linux-gnu"
 
 ; GEN: $__llvm_profile_raw_version = comdat any
 ; GEN: @__llvm_profile_raw_version = constant i64 {{[0-9]+}}, comdat
 ; GEN: @__profn_test_switch = private constant [11 x i8] c"test_switch"
 
 define void @test_switch(i32 %i) {
 entry:
 ; GEN: entry:
-; GEN-NOT: call void @llvm.instrprof.increment
+; NOTENTRY-NOT: call void @llvm.instrprof.increment
+; ENTRY: call void @llvm.instrprof.increment(i8* getelementptr inbounds ([11 x i8], [11 x i8]* @__profn_test_switch, i32 0, i32 0), i64 46200943743, i32 4, i32 0)
   switch i32 %i, label %sw.default [
     i32 1, label %sw.bb
     i32 2, label %sw.bb1
     i32 3, label %sw.bb2
   ]
 ; USE: ]
 ; USE-SAME: !prof ![[BW_SWITCH:[0-9]+]]
 ; USE: ![[BW_SWITCH]] = !{!"branch_weights", i32 3, i32 2, i32 0, i32 5}
 
 sw.bb:
 ; GEN: sw.bb:
 ; GEN: call void @llvm.instrprof.increment(i8* getelementptr inbounds ([11 x i8], [11 x i8]* @__profn_test_switch, i32 0, i32 0), i64 46200943743, i32 4, i32 2)
   br label %sw.epilog
 
 sw.bb1:
 ; GEN: sw.bb1:
-; GEN: call void @llvm.instrprof.increment(i8* getelementptr inbounds ([11 x i8], [11 x i8]* @__profn_test_switch, i32 0, i32 0), i64 46200943743, i32 4, i32 0)
+; NOTENTRY: call void @llvm.instrprof.increment(i8* getelementptr inbounds ([11 x i8], [11 x i8]* @__profn_test_switch, i32 0, i32 0), i64 46200943743, i32 4, i32 0)
+; ENTRY-NOT: call void @llvm.instrprof.increment
   br label %sw.epilog
 
 sw.bb2:
 ; GEN: sw.bb2:
 ; GEN: call void @llvm.instrprof.increment(i8* getelementptr inbounds ([11 x i8], [11 x i8]* @__profn_test_switch, i32 0, i32 0), i64 46200943743, i32 4, i32 1)
   br label %sw.epilog
 
 sw.default:

llvm/test/Transforms/PGOProfile/thinlto_cspgo_use.ll

 ; REQUIRES: x86-registered-target
 
 ; RUN: opt -module-summary %s -o %t1.bc
 ; RUN: opt -module-summary %S/Inputs/thinlto_cspgo_bar_use.ll -o %t2.bc
 ; RUN: llvm-profdata merge %S/Inputs/thinlto_cs.proftext -o %t3.profdata
-; RUN: llvm-lto2 run -lto-cspgo-profile-file=%t3.profdata -save-temps -o %t %t1.bc %t2.bc \
+; RUN: llvm-lto2 run -lto-cspgo-profile-file=%t3.profdata -pgo-instrument-entry=false -save-temps -o %t %t1.bc %t2.bc \
 ; RUN:   -r=%t1.bc,foo,pl \
 ; RUN:   -r=%t1.bc,bar,l \
 ; RUN:   -r=%t1.bc,main,plx \
 ; RUN:   -r=%t2.bc,bar,pl \
 ; RUN:   -r=%t2.bc,odd,pl \
 ; RUN:   -r=%t2.bc,even,pl
 ; RUN: llvm-dis %t.1.4.opt.bc -o - | FileCheck %s --check-prefix=CSUSE
 

llvm/tools/llvm-profdata/llvm-profdata.cpp

static void loadInput(const WeightedFile &Input, SymbolRemapper *Remapper,
   if (WC->Writer.setIsIRLevelProfile(IsIRProfile, HasCSIRProfile)) {
     WC->Errors.emplace_back(
         make_error<StringError>(
             "Merge IR generated profile with Clang generated profile.",
             std::error_code()),
         Filename);
     return;
   }
+  WC->Writer.setInstrEntryBBEnabled(Reader->instrEntryBBEnabled());
 
   for (auto &I : *Reader) {
     if (Remapper)
       I.Name = (*Remapper)(I.Name);
     const StringRef FuncName = I.Name;
     bool Reported = false;
     WC->Writer.addRecord(std::move(I), Input.Weight, [&](Error E) {
       if (Reported) {
for (const auto &Func : *Reader) {
     }
   }
   if (Reader->hasError())
     exitWithError(Reader->getError(), Filename);
 
   if (TextFormat)
     return 0;
   std::unique_ptr<ProfileSummary> PS(Builder.getSummary());
-  OS << "Instrumentation level: "
-     << (Reader->isIRLevelProfile() ? "IR" : "Front-end") << "\n";
+  bool IsIR = Reader->isIRLevelProfile();
+  OS << "Instrumentation level: " << (IsIR ? "IR" : "Front-end");
+  if (IsIR)
+    OS << "  entry_first = " << Reader->instrEntryBBEnabled();
+  OS << "\n";
   if (ShowAllFunctions || !ShowFunction.empty())
     OS << "Functions shown: " << ShownFunctions << "\n";
   OS << "Total functions: " << PS->getNumFunctions() << "\n";
   if (ValueCutoff > 0) {
     OS << "Number of functions with maximum count (< " << ValueCutoff
        << "): " << BelowCutoffFunctions << "\n";
     OS << "Number of functions with maximum count (>= " << ValueCutoff
        << "): " << PS->getNumFunctions() - BelowCutoffFunctions << "\n";