aboutsummaryrefslogtreecommitdiffstats
path: root/contrib/libs/llvm12/include/llvm/Bitstream/BitstreamReader.h
blob: d97a995f83a79513cec9e92642cca4617ba67302 (plain) (blame)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
#pragma once

#ifdef __GNUC__
#pragma GCC diagnostic push
#pragma GCC diagnostic ignored "-Wunused-parameter"
#endif

//===- BitstreamReader.h - Low-level bitstream reader interface -*- C++ -*-===//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
//
//===----------------------------------------------------------------------===//
//
// This header defines the BitstreamReader class.  This class can be used to
// read an arbitrary bitstream, regardless of its contents.
//
//===----------------------------------------------------------------------===//

#ifndef LLVM_BITSTREAM_BITSTREAMREADER_H
#define LLVM_BITSTREAM_BITSTREAMREADER_H

#include "llvm/ADT/ArrayRef.h"
#include "llvm/ADT/SmallVector.h"
#include "llvm/Bitstream/BitCodes.h"
#include "llvm/Support/Endian.h"
#include "llvm/Support/Error.h"
#include "llvm/Support/ErrorHandling.h"
#include "llvm/Support/MathExtras.h"
#include "llvm/Support/MemoryBuffer.h"
#include <algorithm>
#include <cassert>
#include <climits>
#include <cstddef>
#include <cstdint>
#include <memory>
#include <string>
#include <utility>
#include <vector>

namespace llvm {

/// This class maintains the abbreviations read from a block info block.
class BitstreamBlockInfo {
public:
  /// This contains information emitted to BLOCKINFO_BLOCK blocks. These
  /// describe abbreviations that all blocks of the specified ID inherit.
  struct BlockInfo {
    unsigned BlockID = 0;
    std::vector<std::shared_ptr<BitCodeAbbrev>> Abbrevs;
    std::string Name;
    std::vector<std::pair<unsigned, std::string>> RecordNames;
  };

private:
  std::vector<BlockInfo> BlockInfoRecords;

public:
  /// If there is block info for the specified ID, return it, otherwise return
  /// null.
  const BlockInfo *getBlockInfo(unsigned BlockID) const {
    // Common case, the most recent entry matches BlockID.
    if (!BlockInfoRecords.empty() && BlockInfoRecords.back().BlockID == BlockID)
      return &BlockInfoRecords.back();

    for (unsigned i = 0, e = static_cast<unsigned>(BlockInfoRecords.size());
         i != e; ++i)
      if (BlockInfoRecords[i].BlockID == BlockID)
        return &BlockInfoRecords[i];
    return nullptr;
  }

  BlockInfo &getOrCreateBlockInfo(unsigned BlockID) {
    if (const BlockInfo *BI = getBlockInfo(BlockID))
      return *const_cast<BlockInfo*>(BI);

    // Otherwise, add a new record.
    BlockInfoRecords.emplace_back();
    BlockInfoRecords.back().BlockID = BlockID;
    return BlockInfoRecords.back();
  }
};

/// This represents a position within a bitstream. There may be multiple
/// independent cursors reading within one bitstream, each maintaining their
/// own local state.
class SimpleBitstreamCursor {
  ArrayRef<uint8_t> BitcodeBytes;
  size_t NextChar = 0;

public:
  /// This is the current data we have pulled from the stream but have not
  /// returned to the client. This is specifically and intentionally defined to
  /// follow the word size of the host machine for efficiency. We use word_t in
  /// places that are aware of this to make it perfectly explicit what is going
  /// on.
  using word_t = size_t;

private:
  word_t CurWord = 0;

  /// This is the number of bits in CurWord that are valid. This is always from
  /// [0...bits_of(size_t)-1] inclusive.
  unsigned BitsInCurWord = 0;

public:
  static const constexpr size_t MaxChunkSize = sizeof(word_t) * 8;

  SimpleBitstreamCursor() = default;
  explicit SimpleBitstreamCursor(ArrayRef<uint8_t> BitcodeBytes)
      : BitcodeBytes(BitcodeBytes) {}
  explicit SimpleBitstreamCursor(StringRef BitcodeBytes)
      : BitcodeBytes(arrayRefFromStringRef(BitcodeBytes)) {}
  explicit SimpleBitstreamCursor(MemoryBufferRef BitcodeBytes)
      : SimpleBitstreamCursor(BitcodeBytes.getBuffer()) {}

  bool canSkipToPos(size_t pos) const {
    // pos can be skipped to if it is a valid address or one byte past the end.
    return pos <= BitcodeBytes.size();
  }

  bool AtEndOfStream() {
    return BitsInCurWord == 0 && BitcodeBytes.size() <= NextChar;
  }

  /// Return the bit # of the bit we are reading.
  uint64_t GetCurrentBitNo() const {
    return NextChar*CHAR_BIT - BitsInCurWord;
  }

  // Return the byte # of the current bit.
  uint64_t getCurrentByteNo() const { return GetCurrentBitNo() / 8; }

  ArrayRef<uint8_t> getBitcodeBytes() const { return BitcodeBytes; }

  /// Reset the stream to the specified bit number.
  Error JumpToBit(uint64_t BitNo) {
    size_t ByteNo = size_t(BitNo/8) & ~(sizeof(word_t)-1);
    unsigned WordBitNo = unsigned(BitNo & (sizeof(word_t)*8-1));
    assert(canSkipToPos(ByteNo) && "Invalid location");

    // Move the cursor to the right word.
    NextChar = ByteNo;
    BitsInCurWord = 0;

    // Skip over any bits that are already consumed.
    if (WordBitNo) {
      if (Expected<word_t> Res = Read(WordBitNo))
        return Error::success();
      else
        return Res.takeError();
    }

    return Error::success();
  }

  /// Get a pointer into the bitstream at the specified byte offset.
  const uint8_t *getPointerToByte(uint64_t ByteNo, uint64_t NumBytes) {
    return BitcodeBytes.data() + ByteNo;
  }

  /// Get a pointer into the bitstream at the specified bit offset.
  ///
  /// The bit offset must be on a byte boundary.
  const uint8_t *getPointerToBit(uint64_t BitNo, uint64_t NumBytes) {
    assert(!(BitNo % 8) && "Expected bit on byte boundary");
    return getPointerToByte(BitNo / 8, NumBytes);
  }

  Error fillCurWord() {
    if (NextChar >= BitcodeBytes.size())
      return createStringError(std::errc::io_error,
                               "Unexpected end of file reading %u of %u bytes",
                               NextChar, BitcodeBytes.size());

    // Read the next word from the stream.
    const uint8_t *NextCharPtr = BitcodeBytes.data() + NextChar;
    unsigned BytesRead;
    if (BitcodeBytes.size() >= NextChar + sizeof(word_t)) {
      BytesRead = sizeof(word_t);
      CurWord =
          support::endian::read<word_t, support::little, support::unaligned>(
              NextCharPtr);
    } else {
      // Short read.
      BytesRead = BitcodeBytes.size() - NextChar;
      CurWord = 0;
      for (unsigned B = 0; B != BytesRead; ++B)
        CurWord |= uint64_t(NextCharPtr[B]) << (B * 8);
    }
    NextChar += BytesRead;
    BitsInCurWord = BytesRead * 8;
    return Error::success();
  }

  Expected<word_t> Read(unsigned NumBits) {
    static const unsigned BitsInWord = MaxChunkSize;

    assert(NumBits && NumBits <= BitsInWord &&
           "Cannot return zero or more than BitsInWord bits!");

    static const unsigned Mask = sizeof(word_t) > 4 ? 0x3f : 0x1f;

    // If the field is fully contained by CurWord, return it quickly.
    if (BitsInCurWord >= NumBits) {
      word_t R = CurWord & (~word_t(0) >> (BitsInWord - NumBits));

      // Use a mask to avoid undefined behavior.
      CurWord >>= (NumBits & Mask);

      BitsInCurWord -= NumBits;
      return R;
    }

    word_t R = BitsInCurWord ? CurWord : 0;
    unsigned BitsLeft = NumBits - BitsInCurWord;

    if (Error fillResult = fillCurWord())
      return std::move(fillResult);

    // If we run out of data, abort.
    if (BitsLeft > BitsInCurWord)
      return createStringError(std::errc::io_error,
                               "Unexpected end of file reading %u of %u bits",
                               BitsInCurWord, BitsLeft);

    word_t R2 = CurWord & (~word_t(0) >> (BitsInWord - BitsLeft));

    // Use a mask to avoid undefined behavior.
    CurWord >>= (BitsLeft & Mask);

    BitsInCurWord -= BitsLeft;

    R |= R2 << (NumBits - BitsLeft);

    return R;
  }

  Expected<uint32_t> ReadVBR(unsigned NumBits) {
    Expected<unsigned> MaybeRead = Read(NumBits);
    if (!MaybeRead)
      return MaybeRead;
    uint32_t Piece = MaybeRead.get();

    if ((Piece & (1U << (NumBits-1))) == 0)
      return Piece;

    uint32_t Result = 0;
    unsigned NextBit = 0;
    while (true) {
      Result |= (Piece & ((1U << (NumBits-1))-1)) << NextBit;

      if ((Piece & (1U << (NumBits-1))) == 0)
        return Result;

      NextBit += NumBits-1;
      MaybeRead = Read(NumBits);
      if (!MaybeRead)
        return MaybeRead;
      Piece = MaybeRead.get();
    }
  }

  // Read a VBR that may have a value up to 64-bits in size. The chunk size of
  // the VBR must still be <= 32 bits though.
  Expected<uint64_t> ReadVBR64(unsigned NumBits) {
    Expected<uint64_t> MaybeRead = Read(NumBits);
    if (!MaybeRead)
      return MaybeRead;
    uint32_t Piece = MaybeRead.get();

    if ((Piece & (1U << (NumBits-1))) == 0)
      return uint64_t(Piece);

    uint64_t Result = 0;
    unsigned NextBit = 0;
    while (true) {
      Result |= uint64_t(Piece & ((1U << (NumBits-1))-1)) << NextBit;

      if ((Piece & (1U << (NumBits-1))) == 0)
        return Result;

      NextBit += NumBits-1;
      MaybeRead = Read(NumBits);
      if (!MaybeRead)
        return MaybeRead;
      Piece = MaybeRead.get();
    }
  }

  void SkipToFourByteBoundary() {
    // If word_t is 64-bits and if we've read less than 32 bits, just dump
    // the bits we have up to the next 32-bit boundary.
    if (sizeof(word_t) > 4 &&
        BitsInCurWord >= 32) {
      CurWord >>= BitsInCurWord-32;
      BitsInCurWord = 32;
      return;
    }

    BitsInCurWord = 0;
  }

  /// Return the size of the stream in bytes.
  size_t SizeInBytes() const { return BitcodeBytes.size(); }

  /// Skip to the end of the file.
  void skipToEnd() { NextChar = BitcodeBytes.size(); }
};

/// When advancing through a bitstream cursor, each advance can discover a few
/// different kinds of entries:
struct BitstreamEntry {
  enum {
    Error,    // Malformed bitcode was found.
    EndBlock, // We've reached the end of the current block, (or the end of the
              // file, which is treated like a series of EndBlock records.
    SubBlock, // This is the start of a new subblock of a specific ID.
    Record    // This is a record with a specific AbbrevID.
  } Kind;

  unsigned ID;

  static BitstreamEntry getError() {
    BitstreamEntry E; E.Kind = Error; return E;
  }

  static BitstreamEntry getEndBlock() {
    BitstreamEntry E; E.Kind = EndBlock; return E;
  }

  static BitstreamEntry getSubBlock(unsigned ID) {
    BitstreamEntry E; E.Kind = SubBlock; E.ID = ID; return E;
  }

  static BitstreamEntry getRecord(unsigned AbbrevID) {
    BitstreamEntry E; E.Kind = Record; E.ID = AbbrevID; return E;
  }
};

/// This represents a position within a bitcode file, implemented on top of a
/// SimpleBitstreamCursor.
///
/// Unlike iterators, BitstreamCursors are heavy-weight objects that should not
/// be passed by value.
class BitstreamCursor : SimpleBitstreamCursor {
  // This is the declared size of code values used for the current block, in
  // bits.
  unsigned CurCodeSize = 2;

  /// Abbrevs installed at in this block.
  std::vector<std::shared_ptr<BitCodeAbbrev>> CurAbbrevs;

  struct Block {
    unsigned PrevCodeSize;
    std::vector<std::shared_ptr<BitCodeAbbrev>> PrevAbbrevs;

    explicit Block(unsigned PCS) : PrevCodeSize(PCS) {}
  };

  /// This tracks the codesize of parent blocks.
  SmallVector<Block, 8> BlockScope;

  BitstreamBlockInfo *BlockInfo = nullptr;

public:
  static const size_t MaxChunkSize = sizeof(word_t) * 8;

  BitstreamCursor() = default;
  explicit BitstreamCursor(ArrayRef<uint8_t> BitcodeBytes)
      : SimpleBitstreamCursor(BitcodeBytes) {}
  explicit BitstreamCursor(StringRef BitcodeBytes)
      : SimpleBitstreamCursor(BitcodeBytes) {}
  explicit BitstreamCursor(MemoryBufferRef BitcodeBytes)
      : SimpleBitstreamCursor(BitcodeBytes) {}

  using SimpleBitstreamCursor::AtEndOfStream;
  using SimpleBitstreamCursor::canSkipToPos;
  using SimpleBitstreamCursor::fillCurWord;
  using SimpleBitstreamCursor::getBitcodeBytes;
  using SimpleBitstreamCursor::GetCurrentBitNo;
  using SimpleBitstreamCursor::getCurrentByteNo;
  using SimpleBitstreamCursor::getPointerToByte;
  using SimpleBitstreamCursor::JumpToBit;
  using SimpleBitstreamCursor::Read;
  using SimpleBitstreamCursor::ReadVBR;
  using SimpleBitstreamCursor::ReadVBR64;
  using SimpleBitstreamCursor::SizeInBytes;
  using SimpleBitstreamCursor::skipToEnd;

  /// Return the number of bits used to encode an abbrev #.
  unsigned getAbbrevIDWidth() const { return CurCodeSize; }

  /// Flags that modify the behavior of advance().
  enum {
    /// If this flag is used, the advance() method does not automatically pop
    /// the block scope when the end of a block is reached.
    AF_DontPopBlockAtEnd = 1,

    /// If this flag is used, abbrev entries are returned just like normal
    /// records.
    AF_DontAutoprocessAbbrevs = 2
  };

  /// Advance the current bitstream, returning the next entry in the stream.
  Expected<BitstreamEntry> advance(unsigned Flags = 0) {
    while (true) {
      if (AtEndOfStream())
        return BitstreamEntry::getError();

      Expected<unsigned> MaybeCode = ReadCode();
      if (!MaybeCode)
        return MaybeCode.takeError();
      unsigned Code = MaybeCode.get();

      if (Code == bitc::END_BLOCK) {
        // Pop the end of the block unless Flags tells us not to.
        if (!(Flags & AF_DontPopBlockAtEnd) && ReadBlockEnd())
          return BitstreamEntry::getError();
        return BitstreamEntry::getEndBlock();
      }

      if (Code == bitc::ENTER_SUBBLOCK) {
        if (Expected<unsigned> MaybeSubBlock = ReadSubBlockID())
          return BitstreamEntry::getSubBlock(MaybeSubBlock.get());
        else
          return MaybeSubBlock.takeError();
      }

      if (Code == bitc::DEFINE_ABBREV &&
          !(Flags & AF_DontAutoprocessAbbrevs)) {
        // We read and accumulate abbrev's, the client can't do anything with
        // them anyway.
        if (Error Err = ReadAbbrevRecord())
          return std::move(Err);
        continue;
      }

      return BitstreamEntry::getRecord(Code);
    }
  }

  /// This is a convenience function for clients that don't expect any
  /// subblocks. This just skips over them automatically.
  Expected<BitstreamEntry> advanceSkippingSubblocks(unsigned Flags = 0) {
    while (true) {
      // If we found a normal entry, return it.
      Expected<BitstreamEntry> MaybeEntry = advance(Flags);
      if (!MaybeEntry)
        return MaybeEntry;
      BitstreamEntry Entry = MaybeEntry.get();

      if (Entry.Kind != BitstreamEntry::SubBlock)
        return Entry;

      // If we found a sub-block, just skip over it and check the next entry.
      if (Error Err = SkipBlock())
        return std::move(Err);
    }
  }

  Expected<unsigned> ReadCode() { return Read(CurCodeSize); }

  // Block header:
  //    [ENTER_SUBBLOCK, blockid, newcodelen, <align4bytes>, blocklen]

  /// Having read the ENTER_SUBBLOCK code, read the BlockID for the block.
  Expected<unsigned> ReadSubBlockID() { return ReadVBR(bitc::BlockIDWidth); }

  /// Having read the ENTER_SUBBLOCK abbrevid and a BlockID, skip over the body
  /// of this block.
  Error SkipBlock() {
    // Read and ignore the codelen value.
    if (Expected<uint32_t> Res = ReadVBR(bitc::CodeLenWidth))
      ; // Since we are skipping this block, we don't care what code widths are
        // used inside of it.
    else
      return Res.takeError();

    SkipToFourByteBoundary();
    Expected<unsigned> MaybeNum = Read(bitc::BlockSizeWidth);
    if (!MaybeNum)
      return MaybeNum.takeError();
    size_t NumFourBytes = MaybeNum.get();

    // Check that the block wasn't partially defined, and that the offset isn't
    // bogus.
    size_t SkipTo = GetCurrentBitNo() + NumFourBytes * 4 * 8;
    if (AtEndOfStream())
      return createStringError(std::errc::illegal_byte_sequence,
                               "can't skip block: already at end of stream");
    if (!canSkipToPos(SkipTo / 8))
      return createStringError(std::errc::illegal_byte_sequence,
                               "can't skip to bit %zu from %" PRIu64, SkipTo,
                               GetCurrentBitNo());

    if (Error Res = JumpToBit(SkipTo))
      return Res;

    return Error::success();
  }

  /// Having read the ENTER_SUBBLOCK abbrevid, and enter the block.
  Error EnterSubBlock(unsigned BlockID, unsigned *NumWordsP = nullptr);

  bool ReadBlockEnd() {
    if (BlockScope.empty()) return true;

    // Block tail:
    //    [END_BLOCK, <align4bytes>]
    SkipToFourByteBoundary();

    popBlockScope();
    return false;
  }

private:
  void popBlockScope() {
    CurCodeSize = BlockScope.back().PrevCodeSize;

    CurAbbrevs = std::move(BlockScope.back().PrevAbbrevs);
    BlockScope.pop_back();
  }

  //===--------------------------------------------------------------------===//
  // Record Processing
  //===--------------------------------------------------------------------===//

public:
  /// Return the abbreviation for the specified AbbrevId.
  const BitCodeAbbrev *getAbbrev(unsigned AbbrevID) {
    unsigned AbbrevNo = AbbrevID - bitc::FIRST_APPLICATION_ABBREV;
    if (AbbrevNo >= CurAbbrevs.size())
      report_fatal_error("Invalid abbrev number");
    return CurAbbrevs[AbbrevNo].get();
  }

  /// Read the current record and discard it, returning the code for the record.
  Expected<unsigned> skipRecord(unsigned AbbrevID);

  Expected<unsigned> readRecord(unsigned AbbrevID,
                                SmallVectorImpl<uint64_t> &Vals,
                                StringRef *Blob = nullptr);

  //===--------------------------------------------------------------------===//
  // Abbrev Processing
  //===--------------------------------------------------------------------===//
  Error ReadAbbrevRecord();

  /// Read and return a block info block from the bitstream. If an error was
  /// encountered, return None.
  ///
  /// \param ReadBlockInfoNames Whether to read block/record name information in
  /// the BlockInfo block. Only llvm-bcanalyzer uses this.
  Expected<Optional<BitstreamBlockInfo>>
  ReadBlockInfoBlock(bool ReadBlockInfoNames = false);

  /// Set the block info to be used by this BitstreamCursor to interpret
  /// abbreviated records.
  void setBlockInfo(BitstreamBlockInfo *BI) { BlockInfo = BI; }
};

} // end llvm namespace

#endif // LLVM_BITSTREAM_BITSTREAMREADER_H

#ifdef __GNUC__
#pragma GCC diagnostic pop
#endif