include/llvm/ProfileData/InstrProfReader.h

   1 //===- InstrProfReader.h - Instrumented profiling readers -------*- C++ -*-===//
   2 //
   3 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
   4 // See https://llvm.org/LICENSE.txt for license information.
   5 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
   6 //
   7 //===----------------------------------------------------------------------===//
   8 //
   9 // This file contains support for reading profiling data for instrumentation
  10 // based PGO and coverage.
  11 //
  12 //===----------------------------------------------------------------------===//
  13
  14 #ifndef LLVM_PROFILEDATA_INSTRPROFREADER_H
  15 #define LLVM_PROFILEDATA_INSTRPROFREADER_H
  16
  17 #include "llvm/ADT/ArrayRef.h"
  18 #include "llvm/ADT/StringRef.h"
  19 #include "llvm/IR/ProfileSummary.h"
  20 #include "llvm/ProfileData/InstrProf.h"
  21 #include "llvm/Support/Endian.h"
  22 #include "llvm/Support/Error.h"
  23 #include "llvm/Support/LineIterator.h"
  24 #include "llvm/Support/MemoryBuffer.h"
  25 #include "llvm/Support/OnDiskHashTable.h"
  26 #include "llvm/Support/SwapByteOrder.h"
  27 #include <algorithm>
  28 #include <cassert>
  29 #include <cstddef>
  30 #include <cstdint>
  31 #include <iterator>
  32 #include <memory>
  33 #include <utility>
  34 #include <vector>
  35
  36 namespace llvm {
  37
  38 class InstrProfReader;
  39
  40 /// A file format agnostic iterator over profiling data.
  41 class InstrProfIterator : public std::iterator<std::input_iterator_tag,
  42                                                NamedInstrProfRecord> {
  43   InstrProfReader *Reader = nullptr;
  44   value_type Record;
  45
  46   void Increment();
  47
  48 public:
  49   InstrProfIterator() = default;
  50   InstrProfIterator(InstrProfReader *Reader) : Reader(Reader) { Increment(); }
  51
  52   InstrProfIterator &operator++() { Increment(); return *this; }
  53   bool operator==(const InstrProfIterator &RHS) { return Reader == RHS.Reader; }
  54   bool operator!=(const InstrProfIterator &RHS) { return Reader != RHS.Reader; }
  55   value_type &operator*() { return Record; }
  56   value_type *operator->() { return &Record; }
  57 };
  58
  59 /// Base class and interface for reading profiling data of any known instrprof
  60 /// format. Provides an iterator over NamedInstrProfRecords.
  61 class InstrProfReader {
  62   instrprof_error LastError = instrprof_error::success;
  63
  64 public:
  65   InstrProfReader() = default;
  66   virtual ~InstrProfReader() = default;
  67
  68   /// Read the header.  Required before reading first record.
  69   virtual Error readHeader() = 0;
  70
  71   /// Read a single record.
  72   virtual Error readNextRecord(NamedInstrProfRecord &Record) = 0;
  73
  74   /// Iterator over profile data.
  75   InstrProfIterator begin() { return InstrProfIterator(this); }
  76   InstrProfIterator end() { return InstrProfIterator(); }
  77
  78   virtual bool isIRLevelProfile() const = 0;
  79
  80   /// Return the PGO symtab. There are three different readers:
  81   /// Raw, Text, and Indexed profile readers. The first two types
  82   /// of readers are used only by llvm-profdata tool, while the indexed
  83   /// profile reader is also used by llvm-cov tool and the compiler (
  84   /// backend or frontend). Since creating PGO symtab can create
  85   /// significant runtime and memory overhead (as it touches data
  86   /// for the whole program), InstrProfSymtab for the indexed profile
  87   /// reader should be created on demand and it is recommended to be
  88   /// only used for dumping purpose with llvm-proftool, not with the
  89   /// compiler.
  90   virtual InstrProfSymtab &getSymtab() = 0;
  91
  92 protected:
  93   std::unique_ptr<InstrProfSymtab> Symtab;
  94
  95   /// Set the current error and return same.
  96   Error error(instrprof_error Err) {
  97     LastError = Err;
  98     if (Err == instrprof_error::success)
  99       return Error::success();
 100     return make_error<InstrProfError>(Err);
 101   }
 102
 103   Error error(Error &&E) { return error(InstrProfError::take(std::move(E))); }
 104
 105   /// Clear the current error and return a successful one.
 106   Error success() { return error(instrprof_error::success); }
 107
 108 public:
 109   /// Return true if the reader has finished reading the profile data.
 110   bool isEOF() { return LastError == instrprof_error::eof; }
 111
 112   /// Return true if the reader encountered an error reading profiling data.
 113   bool hasError() { return LastError != instrprof_error::success && !isEOF(); }
 114
 115   /// Get the current error.
 116   Error getError() {
 117     if (hasError())
 118       return make_error<InstrProfError>(LastError);
 119     return Error::success();
 120   }
 121
 122   /// Factory method to create an appropriately typed reader for the given
 123   /// instrprof file.
 124   static Expected<std::unique_ptr<InstrProfReader>> create(const Twine &Path);
 125
 126   static Expected<std::unique_ptr<InstrProfReader>>
 127   create(std::unique_ptr<MemoryBuffer> Buffer);
 128 };
 129
 130 /// Reader for the simple text based instrprof format.
 131 ///
 132 /// This format is a simple text format that's suitable for test data. Records
 133 /// are separated by one or more blank lines, and record fields are separated by
 134 /// new lines.
 135 ///
 136 /// Each record consists of a function name, a function hash, a number of
 137 /// counters, and then each counter value, in that order.
 138 class TextInstrProfReader : public InstrProfReader {
 139 private:
 140   /// The profile data file contents.
 141   std::unique_ptr<MemoryBuffer> DataBuffer;
 142   /// Iterator over the profile data.
 143   line_iterator Line;
 144   bool IsIRLevelProfile = false;
 145
 146   Error readValueProfileData(InstrProfRecord &Record);
 147
 148 public:
 149   TextInstrProfReader(std::unique_ptr<MemoryBuffer> DataBuffer_)
 150       : DataBuffer(std::move(DataBuffer_)), Line(*DataBuffer, true, '#') {}
 151   TextInstrProfReader(const TextInstrProfReader &) = delete;
 152   TextInstrProfReader &operator=(const TextInstrProfReader &) = delete;
 153
 154   /// Return true if the given buffer is in text instrprof format.
 155   static bool hasFormat(const MemoryBuffer &Buffer);
 156
 157   bool isIRLevelProfile() const override { return IsIRLevelProfile; }
 158
 159   /// Read the header.
 160   Error readHeader() override;
 161
 162   /// Read a single record.
 163   Error readNextRecord(NamedInstrProfRecord &Record) override;
 164
 165   InstrProfSymtab &getSymtab() override {
 166     assert(Symtab.get());
 167     return *Symtab.get();
 168   }
 169 };
 170
 171 /// Reader for the raw instrprof binary format from runtime.
 172 ///
 173 /// This format is a raw memory dump of the instrumentation-baed profiling data
 174 /// from the runtime.  It has no index.
 175 ///
 176 /// Templated on the unsigned type whose size matches pointers on the platform
 177 /// that wrote the profile.
 178 template <class IntPtrT>
 179 class RawInstrProfReader : public InstrProfReader {
 180 private:
 181   /// The profile data file contents.
 182   std::unique_ptr<MemoryBuffer> DataBuffer;
 183   bool ShouldSwapBytes;
 184   // The value of the version field of the raw profile data header. The lower 56
 185   // bits specifies the format version and the most significant 8 bits specify
 186   // the variant types of the profile.
 187   uint64_t Version;
 188   uint64_t CountersDelta;
 189   uint64_t NamesDelta;
 190   const RawInstrProf::ProfileData<IntPtrT> *Data;
 191   const RawInstrProf::ProfileData<IntPtrT> *DataEnd;
 192   const uint64_t *CountersStart;
 193   const char *NamesStart;
 194   uint64_t NamesSize;
 195   // After value profile is all read, this pointer points to
 196   // the header of next profile data (if exists)
 197   const uint8_t *ValueDataStart;
 198   uint32_t ValueKindLast;
 199   uint32_t CurValueDataSize;
 200
 201 public:
 202   RawInstrProfReader(std::unique_ptr<MemoryBuffer> DataBuffer)
 203       : DataBuffer(std::move(DataBuffer)) {}
 204   RawInstrProfReader(const RawInstrProfReader &) = delete;
 205   RawInstrProfReader &operator=(const RawInstrProfReader &) = delete;
 206
 207   static bool hasFormat(const MemoryBuffer &DataBuffer);
 208   Error readHeader() override;
 209   Error readNextRecord(NamedInstrProfRecord &Record) override;
 210
 211   bool isIRLevelProfile() const override {
 212     return (Version & VARIANT_MASK_IR_PROF) != 0;
 213   }
 214
 215   InstrProfSymtab &getSymtab() override {
 216     assert(Symtab.get());
 217     return *Symtab.get();
 218   }
 219
 220 private:
 221   Error createSymtab(InstrProfSymtab &Symtab);
 222   Error readNextHeader(const char *CurrentPos);
 223   Error readHeader(const RawInstrProf::Header &Header);
 224
 225   template <class IntT> IntT swap(IntT Int) const {
 226     return ShouldSwapBytes ? sys::getSwappedBytes(Int) : Int;
 227   }
 228
 229   support::endianness getDataEndianness() const {
 230     support::endianness HostEndian = getHostEndianness();
 231     if (!ShouldSwapBytes)
 232       return HostEndian;
 233     if (HostEndian == support::little)
 234       return support::big;
 235     else
 236       return support::little;
 237   }
 238
 239   inline uint8_t getNumPaddingBytes(uint64_t SizeInBytes) {
 240     return 7 & (sizeof(uint64_t) - SizeInBytes % sizeof(uint64_t));
 241   }
 242
 243   Error readName(NamedInstrProfRecord &Record);
 244   Error readFuncHash(NamedInstrProfRecord &Record);
 245   Error readRawCounts(InstrProfRecord &Record);
 246   Error readValueProfilingData(InstrProfRecord &Record);
 247   bool atEnd() const { return Data == DataEnd; }
 248
 249   void advanceData() {
 250     Data++;
 251     ValueDataStart += CurValueDataSize;
 252   }
 253
 254   const char *getNextHeaderPos() const {
 255       assert(atEnd());
 256       return (const char *)ValueDataStart;
 257   }
 258
 259   const uint64_t *getCounter(IntPtrT CounterPtr) const {
 260     ptrdiff_t Offset = (swap(CounterPtr) - CountersDelta) / sizeof(uint64_t);
 261     return CountersStart + Offset;
 262   }
 263
 264   StringRef getName(uint64_t NameRef) const {
 265     return Symtab->getFuncName(swap(NameRef));
 266   }
 267 };
 268
 269 using RawInstrProfReader32 = RawInstrProfReader<uint32_t>;
 270 using RawInstrProfReader64 = RawInstrProfReader<uint64_t>;
 271
 272 namespace IndexedInstrProf {
 273
 274 enum class HashT : uint32_t;
 275
 276 } // end namespace IndexedInstrProf
 277
 278 /// Trait for lookups into the on-disk hash table for the binary instrprof
 279 /// format.
 280 class InstrProfLookupTrait {
 281   std::vector<NamedInstrProfRecord> DataBuffer;
 282   IndexedInstrProf::HashT HashType;
 283   unsigned FormatVersion;
 284   // Endianness of the input value profile data.
 285   // It should be LE by default, but can be changed
 286   // for testing purpose.
 287   support::endianness ValueProfDataEndianness = support::little;
 288
 289 public:
 290   InstrProfLookupTrait(IndexedInstrProf::HashT HashType, unsigned FormatVersion)
 291       : HashType(HashType), FormatVersion(FormatVersion) {}
 292
 293   using data_type = ArrayRef<NamedInstrProfRecord>;
 294
 295   using internal_key_type = StringRef;
 296   using external_key_type = StringRef;
 297   using hash_value_type = uint64_t;
 298   using offset_type = uint64_t;
 299
 300   static bool EqualKey(StringRef A, StringRef B) { return A == B; }
 301   static StringRef GetInternalKey(StringRef K) { return K; }
 302   static StringRef GetExternalKey(StringRef K) { return K; }
 303
 304   hash_value_type ComputeHash(StringRef K);
 305
 306   static std::pair<offset_type, offset_type>
 307   ReadKeyDataLength(const unsigned char *&D) {
 308     using namespace support;
 309
 310     offset_type KeyLen = endian::readNext<offset_type, little, unaligned>(D);
 311     offset_type DataLen = endian::readNext<offset_type, little, unaligned>(D);
 312     return std::make_pair(KeyLen, DataLen);
 313   }
 314
 315   StringRef ReadKey(const unsigned char *D, offset_type N) {
 316     return StringRef((const char *)D, N);
 317   }
 318
 319   bool readValueProfilingData(const unsigned char *&D,
 320                               const unsigned char *const End);
 321   data_type ReadData(StringRef K, const unsigned char *D, offset_type N);
 322
 323   // Used for testing purpose only.
 324   void setValueProfDataEndianness(support::endianness Endianness) {
 325     ValueProfDataEndianness = Endianness;
 326   }
 327 };
 328
 329 struct InstrProfReaderIndexBase {
 330   virtual ~InstrProfReaderIndexBase() = default;
 331
 332   // Read all the profile records with the same key pointed to the current
 333   // iterator.
 334   virtual Error getRecords(ArrayRef<NamedInstrProfRecord> &Data) = 0;
 335
 336   // Read all the profile records with the key equal to FuncName
 337   virtual Error getRecords(StringRef FuncName,
 338                                      ArrayRef<NamedInstrProfRecord> &Data) = 0;
 339   virtual void advanceToNextKey() = 0;
 340   virtual bool atEnd() const = 0;
 341   virtual void setValueProfDataEndianness(support::endianness Endianness) = 0;
 342   virtual uint64_t getVersion() const = 0;
 343   virtual bool isIRLevelProfile() const = 0;
 344   virtual Error populateSymtab(InstrProfSymtab &) = 0;
 345 };
 346
 347 using OnDiskHashTableImplV3 =
 348     OnDiskIterableChainedHashTable<InstrProfLookupTrait>;
 349
 350 template <typename HashTableImpl>
 351 class InstrProfReaderItaniumRemapper;
 352
 353 template <typename HashTableImpl>
 354 class InstrProfReaderIndex : public InstrProfReaderIndexBase {
 355 private:
 356   std::unique_ptr<HashTableImpl> HashTable;
 357   typename HashTableImpl::data_iterator RecordIterator;
 358   uint64_t FormatVersion;
 359
 360   friend class InstrProfReaderItaniumRemapper<HashTableImpl>;
 361
 362 public:
 363   InstrProfReaderIndex(const unsigned char *Buckets,
 364                        const unsigned char *const Payload,
 365                        const unsigned char *const Base,
 366                        IndexedInstrProf::HashT HashType, uint64_t Version);
 367   ~InstrProfReaderIndex() override = default;
 368
 369   Error getRecords(ArrayRef<NamedInstrProfRecord> &Data) override;
 370   Error getRecords(StringRef FuncName,
 371                    ArrayRef<NamedInstrProfRecord> &Data) override;
 372   void advanceToNextKey() override { RecordIterator++; }
 373
 374   bool atEnd() const override {
 375     return RecordIterator == HashTable->data_end();
 376   }
 377
 378   void setValueProfDataEndianness(support::endianness Endianness) override {
 379     HashTable->getInfoObj().setValueProfDataEndianness(Endianness);
 380   }
 381
 382   uint64_t getVersion() const override { return GET_VERSION(FormatVersion); }
 383
 384   bool isIRLevelProfile() const override {
 385     return (FormatVersion & VARIANT_MASK_IR_PROF) != 0;
 386   }
 387
 388   Error populateSymtab(InstrProfSymtab &Symtab) override {
 389     return Symtab.create(HashTable->keys());
 390   }
 391 };
 392
 393 /// Name matcher supporting fuzzy matching of symbol names to names in profiles.
 394 class InstrProfReaderRemapper {
 395 public:
 396   virtual ~InstrProfReaderRemapper() {}
 397   virtual Error populateRemappings() { return Error::success(); }
 398   virtual Error getRecords(StringRef FuncName,
 399                            ArrayRef<NamedInstrProfRecord> &Data) = 0;
 400 };
 401
 402 /// Reader for the indexed binary instrprof format.
 403 class IndexedInstrProfReader : public InstrProfReader {
 404 private:
 405   /// The profile data file contents.
 406   std::unique_ptr<MemoryBuffer> DataBuffer;
 407   /// The profile remapping file contents.
 408   std::unique_ptr<MemoryBuffer> RemappingBuffer;
 409   /// The index into the profile data.
 410   std::unique_ptr<InstrProfReaderIndexBase> Index;
 411   /// The profile remapping file contents.
 412   std::unique_ptr<InstrProfReaderRemapper> Remapper;
 413   /// Profile summary data.
 414   std::unique_ptr<ProfileSummary> Summary;
 415   // Index to the current record in the record array.
 416   unsigned RecordIndex;
 417
 418   // Read the profile summary. Return a pointer pointing to one byte past the
 419   // end of the summary data if it exists or the input \c Cur.
 420   const unsigned char *readSummary(IndexedInstrProf::ProfVersion Version,
 421                                    const unsigned char *Cur);
 422
 423 public:
 424   IndexedInstrProfReader(
 425       std::unique_ptr<MemoryBuffer> DataBuffer,
 426       std::unique_ptr<MemoryBuffer> RemappingBuffer = nullptr)
 427       : DataBuffer(std::move(DataBuffer)),
 428         RemappingBuffer(std::move(RemappingBuffer)), RecordIndex(0) {}
 429   IndexedInstrProfReader(const IndexedInstrProfReader &) = delete;
 430   IndexedInstrProfReader &operator=(const IndexedInstrProfReader &) = delete;
 431
 432   /// Return the profile version.
 433   uint64_t getVersion() const { return Index->getVersion(); }
 434   bool isIRLevelProfile() const override { return Index->isIRLevelProfile(); }
 435
 436   /// Return true if the given buffer is in an indexed instrprof format.
 437   static bool hasFormat(const MemoryBuffer &DataBuffer);
 438
 439   /// Read the file header.
 440   Error readHeader() override;
 441   /// Read a single record.
 442   Error readNextRecord(NamedInstrProfRecord &Record) override;
 443
 444   /// Return the NamedInstrProfRecord associated with FuncName and FuncHash
 445   Expected<InstrProfRecord> getInstrProfRecord(StringRef FuncName,
 446                                                uint64_t FuncHash);
 447
 448   /// Fill Counts with the profile data for the given function name.
 449   Error getFunctionCounts(StringRef FuncName, uint64_t FuncHash,
 450                           std::vector<uint64_t> &Counts);
 451
 452   /// Return the maximum of all known function counts.
 453   uint64_t getMaximumFunctionCount() { return Summary->getMaxFunctionCount(); }
 454
 455   /// Factory method to create an indexed reader.
 456   static Expected<std::unique_ptr<IndexedInstrProfReader>>
 457   create(const Twine &Path, const Twine &RemappingPath = "");
 458
 459   static Expected<std::unique_ptr<IndexedInstrProfReader>>
 460   create(std::unique_ptr<MemoryBuffer> Buffer,
 461          std::unique_ptr<MemoryBuffer> RemappingBuffer = nullptr);
 462
 463   // Used for testing purpose only.
 464   void setValueProfDataEndianness(support::endianness Endianness) {
 465     Index->setValueProfDataEndianness(Endianness);
 466   }
 467
 468   // See description in the base class. This interface is designed
 469   // to be used by llvm-profdata (for dumping). Avoid using this when
 470   // the client is the compiler.
 471   InstrProfSymtab &getSymtab() override;
 472   ProfileSummary &getSummary() { return *(Summary.get()); }
 473 };
 474
 475 } // end namespace llvm
 476
 477 #endif // LLVM_PROFILEDATA_INSTRPROFREADER_H