[llvm-profdata] Profile dump for compact binary format

[wenlei]

Summary

Fix "llvm-profdata show" so it can work with compact binary format profile. The change is to mark all functions "used" so SampleProfileReaderCompactBinary::read will read in all profiles available for dumping. The function names will be MD5 hash for compact binary format.

[wmi]

Thanks for the patch.

llvm/include/llvm/ProfileData/SampleProfReader.h

284 ↗

(On Diff #211344)

We may not need markAllFuncToUse. We can just set UseAllFuncs to False inside of collectFuncsToUse.

A related thing is that is it better to set UseAllFuncs to True by default and only set UseAllFuncs to False in collectFuncsToUse? The reason is we have some out-of-tree tools which needs to read profiles of all functions, so it will be easier for them by setting UseAllFuncs to True by default.

[wenlei]

address feedback

[wenlei]

llvm/include/llvm/ProfileData/SampleProfReader.h

284 ↗

(On Diff #211344)

Thanks for the suggestion, Wei. Agreed it's better to have the flag default to true. Updated.

[wmi]

LGTM. Thanks!

llvm/trunk/include/llvm/ProfileData/SampleProfReader.h

 private:
   /// Function name table.
   std::vector<std::string> NameTable;
   /// The table mapping from function name to the offset of its FunctionSample
   /// towards file start.
   DenseMap<StringRef, uint64_t> FuncOffsetTable;
   /// The set containing the functions to use when compiling a module.
   DenseSet<StringRef> FuncsToUse;
+  /// Use all functions from the input profile.
+  bool UseAllFuncs = true;
   virtual std::error_code verifySPMagic(uint64_t Magic) override;
   virtual std::error_code readNameTable() override;
   /// Read a string indirectly via the name table.
   virtual ErrorOr<StringRef> readStringFromTable() override;
   virtual std::error_code readHeader() override;
   std::error_code readFuncOffsetTable();
 
 public:

llvm/trunk/lib/ProfileData/SampleProfReader.cpp

while (!at_eof()) {
     if (std::error_code EC = readFuncProfile())
       return EC;
   }
 
   return sampleprof_error::success;
 }
 
 std::error_code SampleProfileReaderCompactBinary::read() {
+  std::vector<uint64_t> OffsetsToUse;
+  if (UseAllFuncs) {
+    for (auto FuncEntry : FuncOffsetTable) {
+      OffsetsToUse.push_back(FuncEntry.second);
+    }
+  }
+  else {
     for (auto Name : FuncsToUse) {
       auto GUID = std::to_string(MD5Hash(Name));
       auto iter = FuncOffsetTable.find(StringRef(GUID));
       if (iter == FuncOffsetTable.end())
         continue;
+      OffsetsToUse.push_back(iter->second);
+    }
+  }
+
+  for (auto Offset : OffsetsToUse) {
     const uint8_t *SavedData = Data;
     Data = reinterpret_cast<const uint8_t *>(Buffer->getBufferStart()) +
-           iter->second;
+           Offset;
     if (std::error_code EC = readFuncProfile())
       return EC;
     Data = SavedData;
   }
   return sampleprof_error::success;
 }
 
 std::error_code SampleProfileReaderRawBinary::verifySPMagic(uint64_t Magic) {
for (uint32_t I = 0; I < *Size; ++I) {
     FuncOffsetTable[*FName] = *Offset;
   }
   End = TableStart;
   Data = SavedData;
   return sampleprof_error::success;
 }
 
 void SampleProfileReaderCompactBinary::collectFuncsToUse(const Module &M) {
+  UseAllFuncs = false;
   FuncsToUse.clear();
   for (auto &F : M) {
     StringRef CanonName = FunctionSamples::getCanonicalFnName(F);
     FuncsToUse.insert(CanonName);
   }
 }
 
 std::error_code SampleProfileReaderBinary::readSummaryEntry(

llvm/trunk/test/tools/llvm-profdata/compact-sample.proftext

+# Make sure "llvm-profdata show" works for sample profile in binary compact format
+
+# RUN: llvm-profdata show -sample %S/Inputs/compat-sample.profdata  | FileCheck %s 
+
+# CHECK: Function: 15822663052811949562: 17, 0, 6 sampled lines 
+# CHECK-NEXT: Samples collected in the function's body {
+# CHECK: Samples collected in inlined callsites {
+# CHECK-NEXT: 1: inlined callee: 6309742469962978389: 17, 0, 1