//===- InstrProfReader.cpp - Instrumented profiling reader ----------------===// // // The LLVM Compiler Infrastructure // // This file is distributed under the University of Illinois Open Source // License. See LICENSE.TXT for details. // //===----------------------------------------------------------------------===// // // This file contains support for reading profiling data for clang's // instrumentation based PGO and coverage. // //===----------------------------------------------------------------------===// #include "llvm/ProfileData/InstrProfReader.h" #include "llvm/ADT/ArrayRef.h" #include "llvm/ADT/STLExtras.h" #include "llvm/ADT/StringRef.h" #include "llvm/IR/ProfileSummary.h" #include "llvm/ProfileData/InstrProf.h" #include "llvm/ProfileData/ProfileCommon.h" #include "llvm/Support/Endian.h" #include "llvm/Support/Error.h" #include "llvm/Support/ErrorOr.h" #include "llvm/Support/MemoryBuffer.h" #include "llvm/Support/SwapByteOrder.h" #include #include #include #include #include #include #include #include #include using namespace llvm; static Expected> setupMemoryBuffer(const Twine &Path) { ErrorOr> BufferOrErr = MemoryBuffer::getFileOrSTDIN(Path); if (std::error_code EC = BufferOrErr.getError()) return errorCodeToError(EC); return std::move(BufferOrErr.get()); } static Error initializeReader(InstrProfReader &Reader) { return Reader.readHeader(); } Expected> InstrProfReader::create(const Twine &Path) { // Set up the buffer to read. auto BufferOrError = setupMemoryBuffer(Path); if (Error E = BufferOrError.takeError()) return std::move(E); return InstrProfReader::create(std::move(BufferOrError.get())); } Expected> InstrProfReader::create(std::unique_ptr Buffer) { // Sanity check the buffer. if (uint64_t(Buffer->getBufferSize()) > std::numeric_limits::max()) return make_error(instrprof_error::too_large); if (Buffer->getBufferSize() == 0) return make_error(instrprof_error::empty_raw_profile); std::unique_ptr Result; // Create the reader. if (IndexedInstrProfReader::hasFormat(*Buffer)) Result.reset(new IndexedInstrProfReader(std::move(Buffer))); else if (RawInstrProfReader64::hasFormat(*Buffer)) Result.reset(new RawInstrProfReader64(std::move(Buffer))); else if (RawInstrProfReader32::hasFormat(*Buffer)) Result.reset(new RawInstrProfReader32(std::move(Buffer))); else if (TextInstrProfReader::hasFormat(*Buffer)) Result.reset(new TextInstrProfReader(std::move(Buffer))); else return make_error(instrprof_error::unrecognized_format); // Initialize the reader and return the result. if (Error E = initializeReader(*Result)) return std::move(E); return std::move(Result); } Expected> IndexedInstrProfReader::create(const Twine &Path) { // Set up the buffer to read. auto BufferOrError = setupMemoryBuffer(Path); if (Error E = BufferOrError.takeError()) return std::move(E); return IndexedInstrProfReader::create(std::move(BufferOrError.get())); } Expected> IndexedInstrProfReader::create(std::unique_ptr Buffer) { // Sanity check the buffer. if (uint64_t(Buffer->getBufferSize()) > std::numeric_limits::max()) return make_error(instrprof_error::too_large); // Create the reader. if (!IndexedInstrProfReader::hasFormat(*Buffer)) return make_error(instrprof_error::bad_magic); auto Result = llvm::make_unique(std::move(Buffer)); // Initialize the reader and return the result. if (Error E = initializeReader(*Result)) return std::move(E); return std::move(Result); } void InstrProfIterator::Increment() { if (auto E = Reader->readNextRecord(Record)) { // Handle errors in the reader. InstrProfError::take(std::move(E)); *this = InstrProfIterator(); } } bool TextInstrProfReader::hasFormat(const MemoryBuffer &Buffer) { // Verify that this really looks like plain ASCII text by checking a // 'reasonable' number of characters (up to profile magic size). size_t count = std::min(Buffer.getBufferSize(), sizeof(uint64_t)); StringRef buffer = Buffer.getBufferStart(); return count == 0 || std::all_of(buffer.begin(), buffer.begin() + count, [](char c) { return ::isprint(c) || ::isspace(c); }); } // 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); if (Str.equals_lower("ir")) IsIRInstr = true; else if (Str.equals_lower("fe")) IsIRInstr = false; else return error(instrprof_error::bad_header); ++Line; IsIRLevelProfile = IsIRInstr; 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++; if (Line.is_at_end()) return success(); uint32_t NumValueKinds; if (Line->getAsInteger(10, NumValueKinds)) { // No value profile data return success(); } if (NumValueKinds == 0 || NumValueKinds > IPVK_Last + 1) return error(instrprof_error::malformed); Line++; for (uint32_t VK = 0; VK < NumValueKinds; VK++) { VP_READ_ADVANCE(ValueKind); if (ValueKind > IPVK_Last) return error(instrprof_error::malformed); VP_READ_ADVANCE(NumValueSites); if (!NumValueSites) continue; Record.reserveSites(VK, NumValueSites); for (uint32_t S = 0; S < NumValueSites; S++) { VP_READ_ADVANCE(NumValueData); std::vector CurrentValues; for (uint32_t V = 0; V < NumValueData; V++) { CHECK_LINE_END(Line); std::pair VD = Line->rsplit(':'); uint64_t TakenCount, Value; if (ValueKind == IPVK_IndirectCallTarget) { if (Error E = Symtab->addFuncName(VD.first)) return E; Value = IndexedInstrProf::ComputeHash(VD.first); } else { READ_NUM(VD.first, Value); } READ_NUM(VD.second, TakenCount); CurrentValues.push_back({Value, TakenCount}); Line++; } Record.addValueData(ValueKind, S, CurrentValues.data(), NumValueData, nullptr); } } return success(); #undef CHECK_LINE_END #undef READ_NUM #undef VP_READ_ADVANCE } Error TextInstrProfReader::readNextRecord(NamedInstrProfRecord &Record) { // Skip empty lines and comments. while (!Line.is_at_end() && (Line->empty() || Line->startswith("#"))) ++Line; // If we hit EOF while looking for a name, we're done. if (Line.is_at_end()) { Symtab->finalizeSymtab(); return error(instrprof_error::eof); } // Read the function name. Record.Name = *Line++; if (Error E = Symtab->addFuncName(Record.Name)) return E; // Read the function hash. if (Line.is_at_end()) return error(instrprof_error::truncated); if ((Line++)->getAsInteger(0, Record.Hash)) return error(instrprof_error::malformed); // Read the number of counters. uint64_t NumCounters; if (Line.is_at_end()) return error(instrprof_error::truncated); if ((Line++)->getAsInteger(10, NumCounters)) return error(instrprof_error::malformed); if (NumCounters == 0) return error(instrprof_error::malformed); // Read each counter and fill our internal storage with the values. Record.Clear(); Record.Counts.reserve(NumCounters); for (uint64_t I = 0; I < NumCounters; ++I) { if (Line.is_at_end()) return error(instrprof_error::truncated); uint64_t Count; if ((Line++)->getAsInteger(10, Count)) return error(instrprof_error::malformed); Record.Counts.push_back(Count); } // Check if value profile data exists and read it if so. if (Error E = readValueProfileData(Record)) return E; // This is needed to avoid two pass parsing because llvm-profdata // does dumping while reading. Symtab->finalizeSymtab(); return success(); } template bool RawInstrProfReader::hasFormat(const MemoryBuffer &DataBuffer) { if (DataBuffer.getBufferSize() < sizeof(uint64_t)) return false; uint64_t Magic = *reinterpret_cast(DataBuffer.getBufferStart()); return RawInstrProf::getMagic() == Magic || sys::getSwappedBytes(RawInstrProf::getMagic()) == Magic; } template Error RawInstrProfReader::readHeader() { if (!hasFormat(*DataBuffer)) return error(instrprof_error::bad_magic); if (DataBuffer->getBufferSize() < sizeof(RawInstrProf::Header)) return error(instrprof_error::bad_header); auto *Header = reinterpret_cast( DataBuffer->getBufferStart()); ShouldSwapBytes = Header->Magic != RawInstrProf::getMagic(); return readHeader(*Header); } template Error RawInstrProfReader::readNextHeader(const char *CurrentPos) { const char *End = DataBuffer->getBufferEnd(); // Skip zero padding between profiles. while (CurrentPos != End && *CurrentPos == 0) ++CurrentPos; // If there's nothing left, we're done. if (CurrentPos == End) return make_error(instrprof_error::eof); // If there isn't enough space for another header, this is probably just // garbage at the end of the file. if (CurrentPos + sizeof(RawInstrProf::Header) > End) return make_error(instrprof_error::malformed); // The writer ensures each profile is padded to start at an aligned address. if (reinterpret_cast(CurrentPos) % alignof(uint64_t)) return make_error(instrprof_error::malformed); // The magic should have the same byte order as in the previous header. uint64_t Magic = *reinterpret_cast(CurrentPos); if (Magic != swap(RawInstrProf::getMagic())) return make_error(instrprof_error::bad_magic); // There's another profile to read, so we need to process the header. auto *Header = reinterpret_cast(CurrentPos); return readHeader(*Header); } template Error RawInstrProfReader::createSymtab(InstrProfSymtab &Symtab) { if (Error E = Symtab.create(StringRef(NamesStart, NamesSize))) return error(std::move(E)); for (const RawInstrProf::ProfileData *I = Data; I != DataEnd; ++I) { const IntPtrT FPtr = swap(I->FunctionPointer); if (!FPtr) continue; Symtab.mapAddress(FPtr, I->NameRef); } Symtab.finalizeSymtab(); return success(); } template Error RawInstrProfReader::readHeader( const RawInstrProf::Header &Header) { Version = swap(Header.Version); if (GET_VERSION(Version) != RawInstrProf::Version) return error(instrprof_error::unsupported_version); CountersDelta = swap(Header.CountersDelta); NamesDelta = swap(Header.NamesDelta); auto DataSize = swap(Header.DataSize); auto CountersSize = swap(Header.CountersSize); NamesSize = swap(Header.NamesSize); ValueKindLast = swap(Header.ValueKindLast); auto DataSizeInBytes = DataSize * sizeof(RawInstrProf::ProfileData); auto PaddingSize = getNumPaddingBytes(NamesSize); ptrdiff_t DataOffset = sizeof(RawInstrProf::Header); ptrdiff_t CountersOffset = DataOffset + DataSizeInBytes; ptrdiff_t NamesOffset = CountersOffset + sizeof(uint64_t) * CountersSize; ptrdiff_t ValueDataOffset = NamesOffset + NamesSize + PaddingSize; auto *Start = reinterpret_cast(&Header); if (Start + ValueDataOffset > DataBuffer->getBufferEnd()) return error(instrprof_error::bad_header); Data = reinterpret_cast *>( Start + DataOffset); DataEnd = Data + DataSize; CountersStart = reinterpret_cast(Start + CountersOffset); NamesStart = Start + NamesOffset; ValueDataStart = reinterpret_cast(Start + ValueDataOffset); std::unique_ptr NewSymtab = make_unique(); if (Error E = createSymtab(*NewSymtab.get())) return E; Symtab = std::move(NewSymtab); return success(); } template Error RawInstrProfReader::readName(NamedInstrProfRecord &Record) { Record.Name = getName(Data->NameRef); return success(); } template Error RawInstrProfReader::readFuncHash(NamedInstrProfRecord &Record) { Record.Hash = swap(Data->FuncHash); return success(); } template Error RawInstrProfReader::readRawCounts( InstrProfRecord &Record) { uint32_t NumCounters = swap(Data->NumCounters); IntPtrT CounterPtr = Data->CounterPtr; if (NumCounters == 0) return error(instrprof_error::malformed); auto RawCounts = makeArrayRef(getCounter(CounterPtr), NumCounters); auto *NamesStartAsCounter = reinterpret_cast(NamesStart); // Check bounds. if (RawCounts.data() < CountersStart || RawCounts.data() + RawCounts.size() > NamesStartAsCounter) return error(instrprof_error::malformed); if (ShouldSwapBytes) { Record.Counts.clear(); Record.Counts.reserve(RawCounts.size()); for (uint64_t Count : RawCounts) Record.Counts.push_back(swap(Count)); } else Record.Counts = RawCounts; return success(); } template Error RawInstrProfReader::readValueProfilingData( InstrProfRecord &Record) { Record.clearValueData(); CurValueDataSize = 0; // Need to match the logic in value profile dumper code in compiler-rt: uint32_t NumValueKinds = 0; for (uint32_t I = 0; I < IPVK_Last + 1; I++) NumValueKinds += (Data->NumValueSites[I] != 0); if (!NumValueKinds) return success(); Expected> VDataPtrOrErr = ValueProfData::getValueProfData( ValueDataStart, (const unsigned char *)DataBuffer->getBufferEnd(), getDataEndianness()); if (Error E = VDataPtrOrErr.takeError()) return E; // Note that besides deserialization, this also performs the conversion for // indirect call targets. The function pointers from the raw profile are // remapped into function name hashes. VDataPtrOrErr.get()->deserializeTo(Record, &Symtab->getAddrHashMap()); CurValueDataSize = VDataPtrOrErr.get()->getSize(); return success(); } template Error RawInstrProfReader::readNextRecord(NamedInstrProfRecord &Record) { if (atEnd()) // At this point, ValueDataStart field points to the next header. if (Error E = readNextHeader(getNextHeaderPos())) return E; // Read name ad set it in Record. if (Error E = readName(Record)) return E; // Read FuncHash and set it in Record. if (Error E = readFuncHash(Record)) return E; // Read raw counts and set Record. if (Error E = readRawCounts(Record)) return E; // Read value data and set Record. if (Error E = readValueProfilingData(Record)) return E; // Iterate. advanceData(); return success(); } namespace llvm { template class RawInstrProfReader; template class RawInstrProfReader; } // end namespace llvm InstrProfLookupTrait::hash_value_type InstrProfLookupTrait::ComputeHash(StringRef K) { return IndexedInstrProf::ComputeHash(HashType, K); } using data_type = InstrProfLookupTrait::data_type; using offset_type = InstrProfLookupTrait::offset_type; bool InstrProfLookupTrait::readValueProfilingData( const unsigned char *&D, const unsigned char *const End) { Expected> VDataPtrOrErr = ValueProfData::getValueProfData(D, End, ValueProfDataEndianness); if (VDataPtrOrErr.takeError()) return false; VDataPtrOrErr.get()->deserializeTo(DataBuffer.back(), nullptr); D += VDataPtrOrErr.get()->TotalSize; return true; } data_type InstrProfLookupTrait::ReadData(StringRef K, const unsigned char *D, offset_type N) { using namespace support; // Check if the data is corrupt. If so, don't try to read it. if (N % sizeof(uint64_t)) return data_type(); DataBuffer.clear(); std::vector CounterBuffer; const unsigned char *End = D + N; while (D < End) { // Read hash. if (D + sizeof(uint64_t) >= End) return data_type(); uint64_t Hash = endian::readNext(D); // Initialize number of counters for GET_VERSION(FormatVersion) == 1. uint64_t CountsSize = N / sizeof(uint64_t) - 1; // If format version is different then read the number of counters. if (GET_VERSION(FormatVersion) != IndexedInstrProf::ProfVersion::Version1) { if (D + sizeof(uint64_t) > End) return data_type(); CountsSize = endian::readNext(D); } // Read counter values. if (D + CountsSize * sizeof(uint64_t) > End) return data_type(); CounterBuffer.clear(); CounterBuffer.reserve(CountsSize); for (uint64_t J = 0; J < CountsSize; ++J) CounterBuffer.push_back(endian::readNext(D)); DataBuffer.emplace_back(K, Hash, std::move(CounterBuffer)); // Read value profiling data. if (GET_VERSION(FormatVersion) > IndexedInstrProf::ProfVersion::Version2 && !readValueProfilingData(D, End)) { DataBuffer.clear(); return data_type(); } } return DataBuffer; } template Error InstrProfReaderIndex::getRecords( StringRef FuncName, ArrayRef &Data) { auto Iter = HashTable->find(FuncName); if (Iter == HashTable->end()) return make_error(instrprof_error::unknown_function); Data = (*Iter); if (Data.empty()) return make_error(instrprof_error::malformed); return Error::success(); } template Error InstrProfReaderIndex::getRecords( ArrayRef &Data) { if (atEnd()) return make_error(instrprof_error::eof); Data = *RecordIterator; if (Data.empty()) return make_error(instrprof_error::malformed); return Error::success(); } template InstrProfReaderIndex::InstrProfReaderIndex( const unsigned char *Buckets, const unsigned char *const Payload, const unsigned char *const Base, IndexedInstrProf::HashT HashType, uint64_t Version) { FormatVersion = Version; HashTable.reset(HashTableImpl::Create( Buckets, Payload, Base, typename HashTableImpl::InfoType(HashType, Version))); RecordIterator = HashTable->data_begin(); } bool IndexedInstrProfReader::hasFormat(const MemoryBuffer &DataBuffer) { using namespace support; if (DataBuffer.getBufferSize() < 8) return false; uint64_t Magic = endian::read(DataBuffer.getBufferStart()); // Verify that it's magical. return Magic == IndexedInstrProf::Magic; } const unsigned char * IndexedInstrProfReader::readSummary(IndexedInstrProf::ProfVersion Version, const unsigned char *Cur) { using namespace IndexedInstrProf; using namespace support; if (Version >= IndexedInstrProf::Version4) { const IndexedInstrProf::Summary *SummaryInLE = reinterpret_cast(Cur); uint64_t NFields = endian::byte_swap(SummaryInLE->NumSummaryFields); uint64_t NEntries = endian::byte_swap(SummaryInLE->NumCutoffEntries); uint32_t SummarySize = IndexedInstrProf::Summary::getSize(NFields, NEntries); std::unique_ptr SummaryData = IndexedInstrProf::allocSummary(SummarySize); const uint64_t *Src = reinterpret_cast(SummaryInLE); uint64_t *Dst = reinterpret_cast(SummaryData.get()); for (unsigned I = 0; I < SummarySize / sizeof(uint64_t); I++) Dst[I] = endian::byte_swap(Src[I]); SummaryEntryVector DetailedSummary; for (unsigned I = 0; I < SummaryData->NumCutoffEntries; I++) { const IndexedInstrProf::Summary::Entry &Ent = SummaryData->getEntry(I); DetailedSummary.emplace_back((uint32_t)Ent.Cutoff, Ent.MinBlockCount, Ent.NumBlocks); } // initialize InstrProfSummary using the SummaryData from disk. this->Summary = llvm::make_unique( ProfileSummary::PSK_Instr, DetailedSummary, SummaryData->get(Summary::TotalBlockCount), SummaryData->get(Summary::MaxBlockCount), SummaryData->get(Summary::MaxInternalBlockCount), SummaryData->get(Summary::MaxFunctionCount), SummaryData->get(Summary::TotalNumBlocks), SummaryData->get(Summary::TotalNumFunctions)); return Cur + SummarySize; } else { // For older version of profile data, we need to compute on the fly: using namespace IndexedInstrProf; InstrProfSummaryBuilder Builder(ProfileSummaryBuilder::DefaultCutoffs); // FIXME: This only computes an empty summary. Need to call addRecord for // all NamedInstrProfRecords to get the correct summary. this->Summary = Builder.getSummary(); return Cur; } } Error IndexedInstrProfReader::readHeader() { using namespace support; const unsigned char *Start = (const unsigned char *)DataBuffer->getBufferStart(); const unsigned char *Cur = Start; if ((const unsigned char *)DataBuffer->getBufferEnd() - Cur < 24) return error(instrprof_error::truncated); auto *Header = reinterpret_cast(Cur); Cur += sizeof(IndexedInstrProf::Header); // Check the magic number. uint64_t Magic = endian::byte_swap(Header->Magic); if (Magic != IndexedInstrProf::Magic) return error(instrprof_error::bad_magic); // Read the version. uint64_t FormatVersion = endian::byte_swap(Header->Version); if (GET_VERSION(FormatVersion) > IndexedInstrProf::ProfVersion::CurrentVersion) return error(instrprof_error::unsupported_version); Cur = readSummary((IndexedInstrProf::ProfVersion)FormatVersion, Cur); // Read the hash type and start offset. IndexedInstrProf::HashT HashType = static_cast( endian::byte_swap(Header->HashType)); if (HashType > IndexedInstrProf::HashT::Last) return error(instrprof_error::unsupported_hash_type); uint64_t HashOffset = endian::byte_swap(Header->HashOffset); // The rest of the file is an on disk hash table. InstrProfReaderIndexBase *IndexPtr = nullptr; IndexPtr = new InstrProfReaderIndex( Start + HashOffset, Cur, Start, HashType, FormatVersion); Index.reset(IndexPtr); return success(); } InstrProfSymtab &IndexedInstrProfReader::getSymtab() { if (Symtab.get()) return *Symtab.get(); std::unique_ptr NewSymtab = make_unique(); if (Error E = Index->populateSymtab(*NewSymtab.get())) { consumeError(error(InstrProfError::take(std::move(E)))); } Symtab = std::move(NewSymtab); return *Symtab.get(); } Expected IndexedInstrProfReader::getInstrProfRecord(StringRef FuncName, uint64_t FuncHash) { ArrayRef Data; Error Err = Index->getRecords(FuncName, Data); if (Err) return std::move(Err); // Found it. Look for counters with the right hash. for (unsigned I = 0, E = Data.size(); I < E; ++I) { // Check for a match and fill the vector if there is one. if (Data[I].Hash == FuncHash) { return std::move(Data[I]); } } return error(instrprof_error::hash_mismatch); } Error IndexedInstrProfReader::getFunctionCounts(StringRef FuncName, uint64_t FuncHash, std::vector &Counts) { Expected Record = getInstrProfRecord(FuncName, FuncHash); if (Error E = Record.takeError()) return error(std::move(E)); Counts = Record.get().Counts; return success(); } Error IndexedInstrProfReader::readNextRecord(NamedInstrProfRecord &Record) { ArrayRef Data; Error E = Index->getRecords(Data); if (E) return error(std::move(E)); Record = Data[RecordIndex++]; if (RecordIndex >= Data.size()) { Index->advanceToNextKey(); RecordIndex = 0; } return success(); }