summaryrefslogtreecommitdiff
path: root/tools/binman/elf.py
blob: de1ce73f2ae9bdd86856bbe10411a103196f186e (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
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
# SPDX-License-Identifier: GPL-2.0+
# Copyright (c) 2016 Google, Inc
# Written by Simon Glass <sjg@chromium.org>
#
# Handle various things related to ELF images
#

from __future__ import print_function

from collections import namedtuple, OrderedDict
import command
import io
import os
import re
import shutil
import struct
import tempfile

import tools
import tout

ELF_TOOLS = True
try:
    from elftools.elf.elffile import ELFFile
    from elftools.elf.sections import SymbolTableSection
except:  # pragma: no cover
    ELF_TOOLS = False

Symbol = namedtuple('Symbol', ['section', 'address', 'size', 'weak'])

# Information about an ELF file:
#    data: Extracted program contents of ELF file (this would be loaded by an
#           ELF loader when reading this file
#    load: Load address of code
#    entry: Entry address of code
#    memsize: Number of bytes in memory occupied by loading this ELF file
ElfInfo = namedtuple('ElfInfo', ['data', 'load', 'entry', 'memsize'])


def GetSymbols(fname, patterns):
    """Get the symbols from an ELF file

    Args:
        fname: Filename of the ELF file to read
        patterns: List of regex patterns to search for, each a string

    Returns:
        None, if the file does not exist, or Dict:
          key: Name of symbol
          value: Hex value of symbol
    """
    stdout = tools.Run('objdump', '-t', fname)
    lines = stdout.splitlines()
    if patterns:
        re_syms = re.compile('|'.join(patterns))
    else:
        re_syms = None
    syms = {}
    syms_started = False
    for line in lines:
        if not line or not syms_started:
            if 'SYMBOL TABLE' in line:
                syms_started = True
            line = None  # Otherwise code coverage complains about 'continue'
            continue
        if re_syms and not re_syms.search(line):
            continue

        space_pos = line.find(' ')
        value, rest = line[:space_pos], line[space_pos + 1:]
        flags = rest[:7]
        parts = rest[7:].split()
        section, size =  parts[:2]
        if len(parts) > 2:
            name = parts[2] if parts[2] != '.hidden' else parts[3]
            syms[name] = Symbol(section, int(value, 16), int(size,16),
                                flags[1] == 'w')

    # Sort dict by address
    return OrderedDict(sorted(syms.items(), key=lambda x: x[1].address))

def GetSymbolAddress(fname, sym_name):
    """Get a value of a symbol from an ELF file

    Args:
        fname: Filename of the ELF file to read
        patterns: List of regex patterns to search for, each a string

    Returns:
        Symbol value (as an integer) or None if not found
    """
    syms = GetSymbols(fname, [sym_name])
    sym = syms.get(sym_name)
    if not sym:
        return None
    return sym.address

def LookupAndWriteSymbols(elf_fname, entry, section):
    """Replace all symbols in an entry with their correct values

    The entry contents is updated so that values for referenced symbols will be
    visible at run time. This is done by finding out the symbols offsets in the
    entry (using the ELF file) and replacing them with values from binman's data
    structures.

    Args:
        elf_fname: Filename of ELF image containing the symbol information for
            entry
        entry: Entry to process
        section: Section which can be used to lookup symbol values
    """
    fname = tools.GetInputFilename(elf_fname)
    syms = GetSymbols(fname, ['image', 'binman'])
    if not syms:
        return
    base = syms.get('__image_copy_start')
    if not base:
        return
    for name, sym in syms.items():
        if name.startswith('_binman'):
            msg = ("Section '%s': Symbol '%s'\n   in entry '%s'" %
                   (section.GetPath(), name, entry.GetPath()))
            offset = sym.address - base.address
            if offset < 0 or offset + sym.size > entry.contents_size:
                raise ValueError('%s has offset %x (size %x) but the contents '
                                 'size is %x' % (entry.GetPath(), offset,
                                                 sym.size, entry.contents_size))
            if sym.size == 4:
                pack_string = '<I'
            elif sym.size == 8:
                pack_string = '<Q'
            else:
                raise ValueError('%s has size %d: only 4 and 8 are supported' %
                                 (msg, sym.size))

            # Look up the symbol in our entry tables.
            value = section.LookupSymbol(name, sym.weak, msg, base.address)
            if value is None:
                value = -1
                pack_string = pack_string.lower()
            value_bytes = struct.pack(pack_string, value)
            tout.Debug('%s:\n   insert %s, offset %x, value %x, length %d' %
                       (msg, name, offset, value, len(value_bytes)))
            entry.data = (entry.data[:offset] + value_bytes +
                        entry.data[offset + sym.size:])

def MakeElf(elf_fname, text, data):
    """Make an elf file with the given data in a single section

    The output file has a several section including '.text' and '.data',
    containing the info provided in arguments.

    Args:
        elf_fname: Output filename
        text: Text (code) to put in the file's .text section
        data: Data to put in the file's .data section
    """
    outdir = tempfile.mkdtemp(prefix='binman.elf.')
    s_file = os.path.join(outdir, 'elf.S')

    # Spilt the text into two parts so that we can make the entry point two
    # bytes after the start of the text section
    text_bytes1 = ['\t.byte\t%#x' % tools.ToByte(byte) for byte in text[:2]]
    text_bytes2 = ['\t.byte\t%#x' % tools.ToByte(byte) for byte in text[2:]]
    data_bytes = ['\t.byte\t%#x' % tools.ToByte(byte) for byte in data]
    with open(s_file, 'w') as fd:
        print('''/* Auto-generated C program to produce an ELF file for testing */

.section .text
.code32
.globl _start
.type _start, @function
%s
_start:
%s
.ident "comment"

.comm fred,8,4

.section .empty
.globl _empty
_empty:
.byte 1

.globl ernie
.data
.type ernie, @object
.size ernie, 4
ernie:
%s
''' % ('\n'.join(text_bytes1), '\n'.join(text_bytes2), '\n'.join(data_bytes)),
        file=fd)
    lds_file = os.path.join(outdir, 'elf.lds')

    # Use a linker script to set the alignment and text address.
    with open(lds_file, 'w') as fd:
        print('''/* Auto-generated linker script to produce an ELF file for testing */

PHDRS
{
    text PT_LOAD ;
    data PT_LOAD ;
    empty PT_LOAD FLAGS ( 6 ) ;
    note PT_NOTE ;
}

SECTIONS
{
    . = 0xfef20000;
    ENTRY(_start)
    .text . : SUBALIGN(0)
    {
        *(.text)
    } :text
    .data : {
        *(.data)
    } :data
    _bss_start = .;
    .empty : {
        *(.empty)
    } :empty
    /DISCARD/ : {
        *(.note.gnu.property)
    }
    .note : {
        *(.comment)
    } :note
    .bss _bss_start  (OVERLAY) : {
        *(.bss)
    }
}
''', file=fd)
    # -static: Avoid requiring any shared libraries
    # -nostdlib: Don't link with C library
    # -Wl,--build-id=none: Don't generate a build ID, so that we just get the
    #   text section at the start
    # -m32: Build for 32-bit x86
    # -T...: Specifies the link script, which sets the start address
    stdout = command.Output('cc', '-static', '-nostdlib', '-Wl,--build-id=none',
                            '-m32','-T', lds_file, '-o', elf_fname, s_file)
    shutil.rmtree(outdir)

def DecodeElf(data, location):
    """Decode an ELF file and return information about it

    Args:
        data: Data from ELF file
        location: Start address of data to return

    Returns:
        ElfInfo object containing information about the decoded ELF file
    """
    file_size = len(data)
    with io.BytesIO(data) as fd:
        elf = ELFFile(fd)
        data_start = 0xffffffff;
        data_end = 0;
        mem_end = 0;
        virt_to_phys = 0;

        for i in range(elf.num_segments()):
            segment = elf.get_segment(i)
            if segment['p_type'] != 'PT_LOAD' or not segment['p_memsz']:
                skipped = 1  # To make code-coverage see this line
                continue
            start = segment['p_paddr']
            mend = start + segment['p_memsz']
            rend = start + segment['p_filesz']
            data_start = min(data_start, start)
            data_end = max(data_end, rend)
            mem_end = max(mem_end, mend)
            if not virt_to_phys:
                virt_to_phys = segment['p_paddr'] - segment['p_vaddr']

        output = bytearray(data_end - data_start)
        for i in range(elf.num_segments()):
            segment = elf.get_segment(i)
            if segment['p_type'] != 'PT_LOAD' or not segment['p_memsz']:
                skipped = 1  # To make code-coverage see this line
                continue
            start = segment['p_paddr']
            offset = 0
            if start < location:
                offset = location - start
                start = location
            # A legal ELF file can have a program header with non-zero length
            # but zero-length file size and a non-zero offset which, added
            # together, are greater than input->size (i.e. the total file size).
            #  So we need to not even test in the case that p_filesz is zero.
            # Note: All of this code is commented out since we don't have a test
            # case for it.
            size = segment['p_filesz']
            #if not size:
                #continue
            #end = segment['p_offset'] + segment['p_filesz']
            #if end > file_size:
                #raise ValueError('Underflow copying out the segment. File has %#x bytes left, segment end is %#x\n',
                                 #file_size, end)
            output[start - data_start:start - data_start + size] = (
                segment.data()[offset:])
    return ElfInfo(output, data_start, elf.header['e_entry'] + virt_to_phys,
                   mem_end - data_start)