aboutsummaryrefslogtreecommitdiff
path: root/src/common/dwarf/bytereader.cc
blob: 95193834f9e56bc3e7a5b43233bbb8d6c9c44555 (plain)
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
// Copyright (c) 2010 Google Inc. All Rights Reserved.
//
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
//     * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
//     * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following disclaimer
// in the documentation and/or other materials provided with the
// distribution.
//     * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived from
// this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

#include <assert.h>
#include <stdlib.h>

#include "common/dwarf/bytereader-inl.h"
#include "common/dwarf/bytereader.h"

namespace dwarf2reader {

ByteReader::ByteReader(enum Endianness endian)
    :offset_reader_(NULL), address_reader_(NULL), endian_(endian),
     address_size_(0), offset_size_(0),
     have_section_base_(), have_text_base_(), have_data_base_(),
     have_function_base_() { }

ByteReader::~ByteReader() { }

void ByteReader::SetOffsetSize(uint8 size) {
  offset_size_ = size;
  assert(size == 4 || size == 8);
  if (size == 4) {
    this->offset_reader_ = &ByteReader::ReadFourBytes;
  } else {
    this->offset_reader_ = &ByteReader::ReadEightBytes;
  }
}

void ByteReader::SetAddressSize(uint8 size) {
  address_size_ = size;
  assert(size == 4 || size == 8);
  if (size == 4) {
    this->address_reader_ = &ByteReader::ReadFourBytes;
  } else {
    this->address_reader_ = &ByteReader::ReadEightBytes;
  }
}

uint64 ByteReader::ReadInitialLength(const char* start, size_t* len) {
  const uint64 initial_length = ReadFourBytes(start);
  start += 4;

  // In DWARF2/3, if the initial length is all 1 bits, then the offset
  // size is 8 and we need to read the next 8 bytes for the real length.
  if (initial_length == 0xffffffff) {
    SetOffsetSize(8);
    *len = 12;
    return ReadOffset(start);
  } else {
    SetOffsetSize(4);
    *len = 4;
  }
  return initial_length;
}

bool ByteReader::ValidEncoding(DwarfPointerEncoding encoding) const {
  if (encoding == DW_EH_PE_omit) return true;
  if (encoding == DW_EH_PE_aligned) return true;
  if ((encoding & 0x7) > DW_EH_PE_udata8)
    return false;
  if ((encoding & 0x70) > DW_EH_PE_funcrel)
    return false;
  return true;
}

bool ByteReader::UsableEncoding(DwarfPointerEncoding encoding) const {
  switch (encoding & 0x70) {
    case DW_EH_PE_absptr:  return true;
    case DW_EH_PE_pcrel:   return have_section_base_;
    case DW_EH_PE_textrel: return have_text_base_;
    case DW_EH_PE_datarel: return have_data_base_;
    case DW_EH_PE_funcrel: return have_function_base_;
    default:               return false;
  }
}

uint64 ByteReader::ReadEncodedPointer(const char *buffer,
                                      DwarfPointerEncoding encoding,
                                      size_t *len) const {
  // UsableEncoding doesn't approve of DW_EH_PE_omit, so we shouldn't
  // see it here.
  assert(encoding != DW_EH_PE_omit);

  // The Linux Standards Base 4.0 does not make this clear, but the
  // GNU tools (gcc/unwind-pe.h; readelf/dwarf.c; gdb/dwarf2-frame.c)
  // agree that aligned pointers are always absolute, machine-sized,
  // machine-signed pointers.
  if (encoding == DW_EH_PE_aligned) {
    assert(have_section_base_);

    // We don't need to align BUFFER in *our* address space. Rather, we
    // need to find the next position in our buffer that would be aligned
    // when the .eh_frame section the buffer contains is loaded into the
    // program's memory. So align assuming that buffer_base_ gets loaded at
    // address section_base_, where section_base_ itself may or may not be
    // aligned.

    // First, find the offset to START from the closest prior aligned
    // address.
    uint64_t skew = section_base_ & (AddressSize() - 1);
    // Now find the offset from that aligned address to buffer.
    uint64_t offset = skew + (buffer - buffer_base_);
    // Round up to the next boundary.
    uint64_t aligned = (offset + AddressSize() - 1) & -AddressSize();
    // Convert back to a pointer.
    const char *aligned_buffer = buffer_base_ + (aligned - skew);
    // Finally, store the length and actually fetch the pointer.
    *len = aligned_buffer - buffer + AddressSize();
    return ReadAddress(aligned_buffer);
  }

  // Extract the value first, ignoring whether it's a pointer or an
  // offset relative to some base.
  uint64 offset;
  switch (encoding & 0x0f) {
    case DW_EH_PE_absptr:
      // DW_EH_PE_absptr is weird, as it is used as a meaningful value for
      // both the high and low nybble of encoding bytes. When it appears in
      // the high nybble, it means that the pointer is absolute, not an
      // offset from some base address. When it appears in the low nybble,
      // as here, it means that the pointer is stored as a normal
      // machine-sized and machine-signed address. A low nybble of
      // DW_EH_PE_absptr does not imply that the pointer is absolute; it is
      // correct for us to treat the value as an offset from a base address
      // if the upper nybble is not DW_EH_PE_absptr.
      offset = ReadAddress(buffer);
      *len = AddressSize();
      break;

    case DW_EH_PE_uleb128:
      offset = ReadUnsignedLEB128(buffer, len);
      break;
      
    case DW_EH_PE_udata2:
      offset = ReadTwoBytes(buffer);
      *len = 2;
      break;

    case DW_EH_PE_udata4:
      offset = ReadFourBytes(buffer);
      *len = 4;
      break;

    case DW_EH_PE_udata8:
      offset = ReadEightBytes(buffer);
      *len = 8;
      break;

    case DW_EH_PE_sleb128:
      offset = ReadSignedLEB128(buffer, len);
      break;

    case DW_EH_PE_sdata2:
      offset = ReadTwoBytes(buffer);
      // Sign-extend from 16 bits.
      offset = (offset ^ 0x8000) - 0x8000;
      *len = 2;
      break;

    case DW_EH_PE_sdata4:
      offset = ReadFourBytes(buffer);
      // Sign-extend from 32 bits.
      offset = (offset ^ 0x80000000ULL) - 0x80000000ULL;
      *len = 4;
      break;

    case DW_EH_PE_sdata8:
      // No need to sign-extend; this is the full width of our type.
      offset = ReadEightBytes(buffer);
      *len = 8;
      break;

    default:
      abort();
  }

  // Find the appropriate base address.
  uint64 base;
  switch (encoding & 0x70) {
    case DW_EH_PE_absptr:
      base = 0;
      break;

    case DW_EH_PE_pcrel:
      assert(have_section_base_);
      base = section_base_ + (buffer - buffer_base_);
      break;

    case DW_EH_PE_textrel:
      assert(have_text_base_);
      base = text_base_;
      break;

    case DW_EH_PE_datarel:
      assert(have_data_base_);
      base = data_base_;
      break;

    case DW_EH_PE_funcrel:
      assert(have_function_base_);
      base = function_base_;
      break;

    default:
      abort();
  }

  uint64 pointer = base + offset;

  // Remove inappropriate upper bits.
  if (AddressSize() == 4)
    pointer = pointer & 0xffffffff;
  else
    assert(AddressSize() == sizeof(uint64));

  return pointer;
}

}  // namespace dwarf2reader