//===----------------------------------------------------------------------===//
#include "llvm/ProfileData/InstrProfReader.h"
-#include "InstrProfIndexed.h"
#include "llvm/ADT/STLExtras.h"
#include <cassert>
Result.reset(new RawInstrProfReader64(std::move(Buffer)));
else if (RawInstrProfReader32::hasFormat(*Buffer))
Result.reset(new RawInstrProfReader32(std::move(Buffer)));
- else
+ else if (TextInstrProfReader::hasFormat(*Buffer))
Result.reset(new TextInstrProfReader(std::move(Buffer)));
+ else
+ return instrprof_error::unrecognized_format;
// Initialize the reader and return the result.
if (std::error_code EC = initializeReader(*Result))
*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); });
+}
+
std::error_code TextInstrProfReader::readNextRecord(InstrProfRecord &Record) {
// Skip empty lines and comments.
while (!Line.is_at_end() && (Line->empty() || Line->startswith("#")))
return success();
}
-template <class IntPtrT>
-static uint64_t getRawMagic();
-
-template <>
-uint64_t getRawMagic<uint64_t>() {
- return
- uint64_t(255) << 56 |
- uint64_t('l') << 48 |
- uint64_t('p') << 40 |
- uint64_t('r') << 32 |
- uint64_t('o') << 24 |
- uint64_t('f') << 16 |
- uint64_t('r') << 8 |
- uint64_t(129);
-}
-
-template <>
-uint64_t getRawMagic<uint32_t>() {
- return
- uint64_t(255) << 56 |
- uint64_t('l') << 48 |
- uint64_t('p') << 40 |
- uint64_t('r') << 32 |
- uint64_t('o') << 24 |
- uint64_t('f') << 16 |
- uint64_t('R') << 8 |
- uint64_t(129);
-}
-
template <class IntPtrT>
bool RawInstrProfReader<IntPtrT>::hasFormat(const MemoryBuffer &DataBuffer) {
if (DataBuffer.getBufferSize() < sizeof(uint64_t))
return false;
uint64_t Magic =
*reinterpret_cast<const uint64_t *>(DataBuffer.getBufferStart());
- return getRawMagic<IntPtrT>() == Magic ||
- sys::getSwappedBytes(getRawMagic<IntPtrT>()) == Magic;
+ return RawInstrProf::getMagic<IntPtrT>() == Magic ||
+ sys::getSwappedBytes(RawInstrProf::getMagic<IntPtrT>()) == Magic;
}
template <class IntPtrT>
std::error_code RawInstrProfReader<IntPtrT>::readHeader() {
if (!hasFormat(*DataBuffer))
return error(instrprof_error::bad_magic);
- if (DataBuffer->getBufferSize() < sizeof(RawHeader))
+ if (DataBuffer->getBufferSize() < sizeof(RawInstrProf::Header))
return error(instrprof_error::bad_header);
- auto *Header =
- reinterpret_cast<const RawHeader *>(DataBuffer->getBufferStart());
- ShouldSwapBytes = Header->Magic != getRawMagic<IntPtrT>();
+ auto *Header = reinterpret_cast<const RawInstrProf::Header *>(
+ DataBuffer->getBufferStart());
+ ShouldSwapBytes = Header->Magic != RawInstrProf::getMagic<IntPtrT>();
return readHeader(*Header);
}
return 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(RawHeader) > End)
+ if (CurrentPos + sizeof(RawInstrProf::Header) > End)
return instrprof_error::malformed;
// The writer ensures each profile is padded to start at an aligned address.
if (reinterpret_cast<size_t>(CurrentPos) % alignOf<uint64_t>())
return instrprof_error::malformed;
// The magic should have the same byte order as in the previous header.
uint64_t Magic = *reinterpret_cast<const uint64_t *>(CurrentPos);
- if (Magic != swap(getRawMagic<IntPtrT>()))
+ if (Magic != swap(RawInstrProf::getMagic<IntPtrT>()))
return instrprof_error::bad_magic;
// There's another profile to read, so we need to process the header.
- auto *Header = reinterpret_cast<const RawHeader *>(CurrentPos);
+ auto *Header = reinterpret_cast<const RawInstrProf::Header *>(CurrentPos);
return readHeader(*Header);
}
-static uint64_t getRawVersion() {
- return 1;
-}
-
template <class IntPtrT>
-std::error_code
-RawInstrProfReader<IntPtrT>::readHeader(const RawHeader &Header) {
- if (swap(Header.Version) != getRawVersion())
+std::error_code RawInstrProfReader<IntPtrT>::readHeader(
+ const RawInstrProf::Header &Header) {
+ if (swap(Header.Version) != RawInstrProf::Version)
return error(instrprof_error::unsupported_version);
CountersDelta = swap(Header.CountersDelta);
auto CountersSize = swap(Header.CountersSize);
auto NamesSize = swap(Header.NamesSize);
- ptrdiff_t DataOffset = sizeof(RawHeader);
- ptrdiff_t CountersOffset = DataOffset + sizeof(ProfileData) * DataSize;
+ ptrdiff_t DataOffset = sizeof(RawInstrProf::Header);
+ ptrdiff_t CountersOffset =
+ DataOffset + sizeof(RawInstrProf::ProfileData<IntPtrT>) * DataSize;
ptrdiff_t NamesOffset = CountersOffset + sizeof(uint64_t) * CountersSize;
size_t ProfileSize = NamesOffset + sizeof(char) * NamesSize;
if (Start + ProfileSize > DataBuffer->getBufferEnd())
return error(instrprof_error::bad_header);
- Data = reinterpret_cast<const ProfileData *>(Start + DataOffset);
+ Data = reinterpret_cast<const RawInstrProf::ProfileData<IntPtrT> *>(
+ Start + DataOffset);
DataEnd = Data + DataSize;
CountersStart = reinterpret_cast<const uint64_t *>(Start + CountersOffset);
NamesStart = Start + NamesOffset;
}
template <class IntPtrT>
-std::error_code
-RawInstrProfReader<IntPtrT>::readNextRecord(InstrProfRecord &Record) {
- if (Data == DataEnd)
- if (std::error_code EC = readNextHeader(ProfileEnd))
- return EC;
+std::error_code RawInstrProfReader<IntPtrT>::readName(InstrProfRecord &Record) {
+ Record.Name = StringRef(getName(Data->NamePtr), swap(Data->NameSize));
+ if (Record.Name.data() < NamesStart ||
+ Record.Name.data() + Record.Name.size() > DataBuffer->getBufferEnd())
+ return error(instrprof_error::malformed);
- // Get the raw data.
- StringRef RawName(getName(Data->NamePtr), swap(Data->NameSize));
+ return success();
+}
+
+template <class IntPtrT>
+std::error_code RawInstrProfReader<IntPtrT>::readFuncHash(
+ InstrProfRecord &Record) {
+ Record.Hash = swap(Data->FuncHash);
+ return success();
+}
+
+template <class IntPtrT>
+std::error_code RawInstrProfReader<IntPtrT>::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(Data->CounterPtr), NumCounters);
- // Check bounds.
+ auto RawCounts = makeArrayRef(getCounter(CounterPtr), NumCounters);
auto *NamesStartAsCounter = reinterpret_cast<const uint64_t *>(NamesStart);
- if (RawName.data() < NamesStart ||
- RawName.data() + RawName.size() > DataBuffer->getBufferEnd() ||
- RawCounts.data() < CountersStart ||
+
+ // Check bounds.
+ if (RawCounts.data() < CountersStart ||
RawCounts.data() + RawCounts.size() > NamesStartAsCounter)
return error(instrprof_error::malformed);
- // Store the data in Record, byte-swapping as necessary.
- Record.Hash = swap(Data->FuncHash);
- Record.Name = RawName;
if (ShouldSwapBytes) {
Record.Counts.clear();
Record.Counts.reserve(RawCounts.size());
} else
Record.Counts = RawCounts;
+ return success();
+}
+
+template <class IntPtrT>
+std::error_code
+RawInstrProfReader<IntPtrT>::readNextRecord(InstrProfRecord &Record) {
+ if (atEnd())
+ if (std::error_code EC = readNextHeader(ProfileEnd))
+ return EC;
+
+ // Read name ad set it in Record.
+ if (std::error_code EC = readName(Record))
+ return EC;
+
+ // Read FuncHash and set it in Record.
+ if (std::error_code EC = readFuncHash(Record))
+ return EC;
+
+ // Read raw counts and set Record.
+ if (std::error_code EC = readRawCounts(Record))
+ return EC;
+
// Iterate.
- ++Data;
+ advanceData();
return success();
}
bool InstrProfLookupTrait::ReadValueProfilingData(
const unsigned char *&D, const unsigned char *const End) {
+ ErrorOr<std::unique_ptr<IndexedInstrProf::ValueProfData>> VDataPtrOrErr =
+ IndexedInstrProf::ValueProfData::getValueProfData(
+ D, End, ValueProfDataEndianness);
- using namespace support;
- // Read number of value kinds with value sites.
- if (D + sizeof(uint64_t) > End)
+ if (VDataPtrOrErr.getError())
return false;
- uint64_t ValueKindCount = endian::readNext<uint64_t, little, unaligned>(D);
-
- for (uint32_t Kind = 0; Kind < ValueKindCount; ++Kind) {
- // Read value kind and number of value sites for kind.
- if (D + 2 * sizeof(uint64_t) > End)
- return false;
- uint64_t ValueKind = endian::readNext<uint64_t, little, unaligned>(D);
- uint64_t ValueSiteCount = endian::readNext<uint64_t, little, unaligned>(D);
+ VDataPtrOrErr.get()->deserializeTo(DataBuffer.back(), &HashKeys);
+ D += VDataPtrOrErr.get()->TotalSize;
- std::vector<InstrProfValueSiteRecord> &ValueSites =
- DataBuffer.back().getValueSitesForKind(ValueKind);
- ValueSites.reserve(ValueSiteCount);
- for (uint64_t VSite = 0; VSite < ValueSiteCount; ++VSite) {
- // Read number of value data pairs at value site.
- if (D + sizeof(uint64_t) > End)
- return false;
- uint64_t ValueDataCount =
- endian::readNext<uint64_t, little, unaligned>(D);
-
- // Check if there are as many ValueDataPairs as ValueDataCount in memory.
- if (D + (ValueDataCount << 1) * sizeof(uint64_t) > End)
- return false;
-
- InstrProfValueSiteRecord VSiteRecord;
- for (uint64_t VCount = 0; VCount < ValueDataCount; ++VCount) {
- uint64_t Value = endian::readNext<uint64_t, little, unaligned>(D);
- uint64_t NumTaken = endian::readNext<uint64_t, little, unaligned>(D);
- switch (ValueKind) {
- case IPVK_IndirectCallTarget: {
- auto Result =
- std::lower_bound(HashKeys.begin(), HashKeys.end(), Value,
- [](const std::pair<uint64_t, const char *> &LHS,
- uint64_t RHS) { return LHS.first < RHS; });
- assert(Result != HashKeys.end() &&
- "Hash does not match any known keys\n");
- Value = (uint64_t)Result->second;
- break;
- }
- }
- VSiteRecord.ValueData.push_back(std::make_pair(Value, NumTaken));
- }
- ValueSites.push_back(std::move(VSiteRecord));
- }
- }
return true;
}
using namespace support;
const unsigned char *End = D + N;
while (D < End) {
- // Read hash
+ // Read hash.
if (D + sizeof(uint64_t) >= End)
return data_type();
uint64_t Hash = endian::readNext<uint64_t, little, unaligned>(D);
- // Initialize number of counters for FormatVersion == 1
+ // Initialize number of counters for FormatVersion == 1.
uint64_t CountsSize = N / sizeof(uint64_t) - 1;
- // If format version is different then read number of counters
+ // If format version is different then read the number of counters.
if (FormatVersion != 1) {
if (D + sizeof(uint64_t) > End)
return data_type();
CountsSize = endian::readNext<uint64_t, little, unaligned>(D);
}
- // Read counter values
+ // Read counter values.
if (D + CountsSize * sizeof(uint64_t) > End)
return data_type();
for (uint64_t J = 0; J < CountsSize; ++J)
CounterBuffer.push_back(endian::readNext<uint64_t, little, unaligned>(D));
- DataBuffer.push_back(InstrProfRecord(K, Hash, std::move(CounterBuffer)));
+ DataBuffer.emplace_back(K, Hash, std::move(CounterBuffer));
- // Read value profiling data
+ // Read value profiling data.
if (FormatVersion > 2 && !ReadValueProfilingData(D, End)) {
DataBuffer.clear();
return data_type();
return DataBuffer;
}
+std::error_code
+InstrProfReaderIndex::getRecords(StringRef FuncName,
+ ArrayRef<InstrProfRecord> &Data) {
+ auto Iter = Index->find(FuncName);
+ if (Iter == Index->end())
+ return instrprof_error::unknown_function;
+
+ Data = (*Iter);
+ if (Data.empty())
+ return instrprof_error::malformed;
+
+ return instrprof_error::success;
+}
+
+std::error_code InstrProfReaderIndex::getRecords(
+ ArrayRef<InstrProfRecord> &Data) {
+ if (atEnd()) return instrprof_error::eof;
+
+ Data = *RecordIterator;
+
+ if (Data.empty()) return instrprof_error::malformed;
+
+ return instrprof_error::success;
+}
+
+void InstrProfReaderIndex::Init(const unsigned char *Buckets,
+ const unsigned char *const Payload,
+ const unsigned char *const Base,
+ IndexedInstrProf::HashT HashType,
+ uint64_t Version) {
+ FormatVersion = Version;
+ Index.reset(IndexType::Create(Buckets, Payload, Base,
+ InstrProfLookupTrait(HashType, Version)));
+ // Form the map of hash values to const char* keys in profiling data.
+ std::vector<std::pair<uint64_t, const char *>> HashKeys;
+ for (auto Key : Index->keys()) {
+ const char *KeyTableRef = StringTable.insertString(Key);
+ HashKeys.push_back(std::make_pair(ComputeHash(HashType, Key), KeyTableRef));
+ }
+ std::sort(HashKeys.begin(), HashKeys.end(), less_first());
+ HashKeys.erase(std::unique(HashKeys.begin(), HashKeys.end()), HashKeys.end());
+ // Set the hash key map for the InstrLookupTrait
+ Index->getInfoObj().setHashKeys(std::move(HashKeys));
+ RecordIterator = Index->data_begin();
+}
+
bool IndexedInstrProfReader::hasFormat(const MemoryBuffer &DataBuffer) {
if (DataBuffer.getBufferSize() < 8)
return false;
using namespace support;
uint64_t Magic =
endian::read<uint64_t, little, aligned>(DataBuffer.getBufferStart());
+ // Verify that it's magical.
return Magic == IndexedInstrProf::Magic;
}
using namespace support;
+ auto *Header = reinterpret_cast<const IndexedInstrProf::Header *>(Cur);
+ Cur += sizeof(IndexedInstrProf::Header);
+
// Check the magic number.
- uint64_t Magic = endian::readNext<uint64_t, little, unaligned>(Cur);
+ uint64_t Magic = endian::byte_swap<uint64_t, little>(Header->Magic);
if (Magic != IndexedInstrProf::Magic)
return error(instrprof_error::bad_magic);
// Read the version.
- FormatVersion = endian::readNext<uint64_t, little, unaligned>(Cur);
+ uint64_t FormatVersion = endian::byte_swap<uint64_t, little>(Header->Version);
if (FormatVersion > IndexedInstrProf::Version)
return error(instrprof_error::unsupported_version);
// Read the maximal function count.
- MaxFunctionCount = endian::readNext<uint64_t, little, unaligned>(Cur);
+ MaxFunctionCount =
+ endian::byte_swap<uint64_t, little>(Header->MaxFunctionCount);
// Read the hash type and start offset.
IndexedInstrProf::HashT HashType = static_cast<IndexedInstrProf::HashT>(
- endian::readNext<uint64_t, little, unaligned>(Cur));
+ endian::byte_swap<uint64_t, little>(Header->HashType));
if (HashType > IndexedInstrProf::HashT::Last)
return error(instrprof_error::unsupported_hash_type);
- uint64_t HashOffset = endian::readNext<uint64_t, little, unaligned>(Cur);
- // The rest of the file is an on disk hash table.
- Index.reset(InstrProfReaderIndex::Create(
- Start + HashOffset, Cur, Start,
- InstrProfLookupTrait(HashType, FormatVersion)));
+ uint64_t HashOffset = endian::byte_swap<uint64_t, little>(Header->HashOffset);
- // Form the map of hash values to const char* keys in profiling data.
- std::vector<std::pair<uint64_t, const char *>> HashKeys;
- for (auto Key : Index->keys()) {
- const char *KeyTableRef = StringTable.insertString(Key);
- HashKeys.push_back(std::make_pair(ComputeHash(HashType, Key), KeyTableRef));
- }
- std::sort(HashKeys.begin(), HashKeys.end(), less_first());
- HashKeys.erase(std::unique(HashKeys.begin(), HashKeys.end()), HashKeys.end());
- // Set the hash key map for the InstrLookupTrait
- Index->getInfoObj().setHashKeys(std::move(HashKeys));
- // Set up our iterator for readNextRecord.
- RecordIterator = Index->data_begin();
+ // The rest of the file is an on disk hash table.
+ Index.Init(Start + HashOffset, Cur, Start, HashType, FormatVersion);
return success();
}
-std::error_code IndexedInstrProfReader::getFunctionCounts(
- StringRef FuncName, uint64_t FuncHash, std::vector<uint64_t> &Counts) {
- auto Iter = Index->find(FuncName);
- if (Iter == Index->end())
- return error(instrprof_error::unknown_function);
-
+ErrorOr<InstrProfRecord>
+IndexedInstrProfReader::getInstrProfRecord(StringRef FuncName,
+ uint64_t FuncHash) {
+ ArrayRef<InstrProfRecord> Data;
+ std::error_code EC = Index.getRecords(FuncName, Data);
+ if (EC != instrprof_error::success)
+ return EC;
// Found it. Look for counters with the right hash.
- ArrayRef<InstrProfRecord> Data = (*Iter);
- if (Data.empty())
- return error(instrprof_error::malformed);
-
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) {
- Counts = Data[I].Counts;
- return success();
+ return std::move(Data[I]);
}
}
return error(instrprof_error::hash_mismatch);
}
std::error_code
-IndexedInstrProfReader::readNextRecord(InstrProfRecord &Record) {
- // Are we out of records?
- if (RecordIterator == Index->data_end())
- return error(instrprof_error::eof);
+IndexedInstrProfReader::getFunctionCounts(StringRef FuncName, uint64_t FuncHash,
+ std::vector<uint64_t> &Counts) {
+ ErrorOr<InstrProfRecord> Record = getInstrProfRecord(FuncName, FuncHash);
+ if (std::error_code EC = Record.getError())
+ return EC;
- if ((*RecordIterator).empty())
- return error(instrprof_error::malformed);
+ Counts = Record.get().Counts;
+ return success();
+}
+std::error_code IndexedInstrProfReader::readNextRecord(
+ InstrProfRecord &Record) {
static unsigned RecordIndex = 0;
- ArrayRef<InstrProfRecord> Data = (*RecordIterator);
+
+ ArrayRef<InstrProfRecord> Data;
+
+ std::error_code EC = Index.getRecords(Data);
+ if (EC != instrprof_error::success)
+ return error(EC);
+
Record = Data[RecordIndex++];
if (RecordIndex >= Data.size()) {
- ++RecordIterator;
+ Index.advanceToNextKey();
RecordIndex = 0;
}
return success();