mirror of
https://sourceware.org/git/binutils-gdb.git
synced 2024-12-15 04:31:49 +08:00
dda8d76d0d
Hi Guys, I am applying the rather large patch attached to this email to enhance the readelf and objdump programs so that they now have the ability to follow links to separate debug info files. (As requested by PR 15152). So for example whereas before we had this output: $ readelf -wi main.exe Contents of the .debug_info section: [...] <15> DW_AT_comp_dir : (alt indirect string, offset: 0x30c) [...] With the new option enabled we get: $ readelf -wiK main.exe main.exe: Found separate debug info file: dwz.debug Contents of the .debug_info section (loaded from main.exe): [...] <15> DW_AT_comp_dir : (alt indirect string, offset: 0x30c) /home/nickc/Downloads/dwzm [...] The link following feature also means that we can get two lots of output if the same section exists in both the main file and the separate debug info file: $ readelf -wiK main.exe main.exe: Found separate debug info file: dwz.debug Contents of the .debug_info section (loaded from main.exe): [...] Contents of the .debug_info section (loaded from dwz.debug): [...] The patch also adds the ability to display the contents of debuglink sections: $ readelf -wk main.exe Contents of the .gnu_debugaltlink section: Separate debug info file: dwz.debug Build-ID (0x14 bytes): c4 a8 89 8d 64 cf 70 8a 35 68 21 f2 ed 24 45 3e 18 7a 7a 93 Naturally there are long versions of these options (=follow-links and =links). The documentation has been updated as well, and since both readelf and objdump use the same set of debug display options, I have moved the text into a separate file. There are also a couple of new binutils tests to exercise the new behaviour. There are a couple of missing features in the current patch however, although I do intend to address them in follow up submissions: Firstly the code does not check the build-id inside separate debug info files when it is searching for a file specified by a .gnu_debugaltlink section. It just assumes that if the file is there, then it contains the information being sought. Secondly I have not checked the DWARF-5 version of these link features, so there will probably be code to add there. Thirdly I have only implemented link following for the DW_FORM_GNU_strp_alt format. Other alternate formats (eg DW_FORM_GNU_ref_alt) have yet to be implemented. Lastly, whilst implementing this feature I found it necessary to move some of the global variables used by readelf (eg section_headers) into a structure that can be passed around. I have moved all of the global variables that were necessary to get the patch working, but I need to complete the operation and move the remaining, file-specific variables (eg dynamic_strings). Cheers Nick binutils PR 15152 * dwarf.h (enum dwarf_section_display_enum): Add gnu_debuglink, gnu_debugaltlink and separate_debug_str. (struct dwarf_section): Add filename field. Add prototypes for load_separate_debug_file, close_debug_file and open_debug_file. * dwarf.c (do_debug_links): New. (do_follow_links): New. (separate_debug_file, separate_debug_filename): New. (fetch_alt_indirect_string): New function. Retrieves a string from the debug string table in the separate debug info file. (read_and_display_attr_value): Use it with DW_FORM_GNU_strp_alt. (load_debug_section_with_follow): New function. Like load_debug_section, but if the first attempt fails, then tries again in the separate debug info file. (introduce): New function. (process_debug_info): Use load_debug_section_with_follow and introduce. (load_debug_info): Likewise. (display_debug_lines_raw): Likewise. (display_debug_lines_decoded): Likewise. (display_debug_macinfo): Likewise. (display_debug_macro): Likewise. (display_debug_abbrev): Likewise. (display_debug_loc): Likewise. (display_debug_str): Likewise. (display_debug_aranges): Likewise. (display_debug_addr); Likewise. (display_debug_frames): Likewise. (display_gdb_index): Likewise. (process_cu_tu_index): Likewise. (load_cu_tu_indexes): Likewise. (display_debug_links): New function. Displays the contents of a .gnu_debuglink or .gnu_debugaltlink section. (calc_gnu_debuglink_ctc32):New function. Calculates a CRC32 value. (check_gnu_debuglink): New function. Checks the CRC of a potential separate debug info file. (parse_gnu_debuglink): New function. Reads a CRC value out of a .gnu_debuglink section. (check_gnu_debugaltlink): New function. (parse_gnu_debugaltlink): New function. Reads the build-id value out of a .gnu_debugaltlink section. (load_separate_debug_info): New function. Finds and loads a separate debug info file. (load_separate_debug_file): New function. Attempts to find and follow a link to a separate debug info file. (free_debug_memory): Free the separate debug info file information. (opts_table): Add "follow-links" and "links". (dwarf_select_sections_by_letters): Add "k" and "K". (debug_displays): Reformat. Add .gnu-debuglink and .gnu_debugaltlink. Add an extra entry for .debug_str in a separate debug info file. * doc/binutils.texi: Move description of debug dump features common to both readelf and objdump into... * objdump.c (usage): Add -Wk and -WK. (load_specific_debug_section): Initialise the filename field in the dwarf_section structure. (close_debug_file): New function. (open_debug_file): New function. (dump_dwarf): Load and dump the separate debug info sections. * readelf.c (struct filedata): New structure. Contains various variables that used to be global: (current_file_size, string_table, string_table_length, elf_header) (section_headers, program_headers, dump_sects, num_dump_sects): Move into filedata structure. (cmdline): New global variable. Contains list of sections to dump by number, as specified on the command line. Add filedata parameter to most functions. (load_debug_section): Load the string table if it has not already been retrieved. (close_file): New function. (close_debug_file): New function. (open_file): New function. (open_debug_file): New function. (process_object): Process sections in any separate debug info files. * doc/debug.options.texi: New file. Add description of =links and =follow-links options. * NEWS: Mention the new feature. * elfcomm.c: Have the byte gte functions take a const pointer. * elfcomm.h: Update prototypes. * testsuite/binutils-all/dw5.W: Update expected output. * testsuite/binutils-all/objdump.WL: Update expected output. * testsuite/binutils-all/objdump.exp: Add test of -WK and -Wk. * testsuite/binutils-all/readelf.exp: Add test of -wK and -wk. * testsuite/binutils-all/readelf.k: New file. * testsuite/binutils-all/objdump.Wk: New file. * testsuite/binutils-all/objdump.WK2: New file. * testsuite/binutils-all/linkdebug.s: New file. * testsuite/binutils-all/debuglink.s: New file. gas * testsuite/gas/avr/large-debug-line-table.d: Update expected output. * testsuite/gas/elf/dwarf2-11.d: Likewise. * testsuite/gas/elf/dwarf2-12.d: Likewise. * testsuite/gas/elf/dwarf2-13.d: Likewise. * testsuite/gas/elf/dwarf2-14.d: Likewise. * testsuite/gas/elf/dwarf2-15.d: Likewise. * testsuite/gas/elf/dwarf2-16.d: Likewise. * testsuite/gas/elf/dwarf2-17.d: Likewise. * testsuite/gas/elf/dwarf2-18.d: Likewise. * testsuite/gas/elf/dwarf2-5.d: Likewise. * testsuite/gas/elf/dwarf2-6.d: Likewise. * testsuite/gas/elf/dwarf2-7.d: Likewise. ld * testsuite/ld-avr/gc-section-debugline.d: Update expected output.
930 lines
26 KiB
C
930 lines
26 KiB
C
/* elfcomm.c -- common code for ELF format file.
|
|
Copyright (C) 2010-2017 Free Software Foundation, Inc.
|
|
|
|
Originally developed by Eric Youngdale <eric@andante.jic.com>
|
|
Modifications by Nick Clifton <nickc@redhat.com>
|
|
|
|
This file is part of GNU Binutils.
|
|
|
|
This program is free software; you can redistribute it and/or modify
|
|
it under the terms of the GNU General Public License as published by
|
|
the Free Software Foundation; either version 3 of the License, or
|
|
(at your option) any later version.
|
|
|
|
This program is distributed in the hope that it will be useful,
|
|
but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
GNU General Public License for more details.
|
|
|
|
You should have received a copy of the GNU General Public License
|
|
along with this program; if not, write to the Free Software
|
|
Foundation, Inc., 51 Franklin Street - Fifth Floor, Boston, MA
|
|
02110-1301, USA. */
|
|
|
|
#include "sysdep.h"
|
|
#include "libiberty.h"
|
|
#include "filenames.h"
|
|
#include "bfd.h"
|
|
#include "aout/ar.h"
|
|
#include "bucomm.h"
|
|
#include "elfcomm.h"
|
|
#include <assert.h>
|
|
|
|
void
|
|
error (const char *message, ...)
|
|
{
|
|
va_list args;
|
|
|
|
/* Try to keep error messages in sync with the program's normal output. */
|
|
fflush (stdout);
|
|
|
|
va_start (args, message);
|
|
fprintf (stderr, _("%s: Error: "), program_name);
|
|
vfprintf (stderr, message, args);
|
|
va_end (args);
|
|
}
|
|
|
|
void
|
|
warn (const char *message, ...)
|
|
{
|
|
va_list args;
|
|
|
|
/* Try to keep warning messages in sync with the program's normal output. */
|
|
fflush (stdout);
|
|
|
|
va_start (args, message);
|
|
fprintf (stderr, _("%s: Warning: "), program_name);
|
|
vfprintf (stderr, message, args);
|
|
va_end (args);
|
|
}
|
|
|
|
void (*byte_put) (unsigned char *, elf_vma, int);
|
|
|
|
void
|
|
byte_put_little_endian (unsigned char * field, elf_vma value, int size)
|
|
{
|
|
switch (size)
|
|
{
|
|
case 8:
|
|
field[7] = (((value >> 24) >> 24) >> 8) & 0xff;
|
|
field[6] = ((value >> 24) >> 24) & 0xff;
|
|
field[5] = ((value >> 24) >> 16) & 0xff;
|
|
field[4] = ((value >> 24) >> 8) & 0xff;
|
|
/* Fall through. */
|
|
case 4:
|
|
field[3] = (value >> 24) & 0xff;
|
|
/* Fall through. */
|
|
case 3:
|
|
field[2] = (value >> 16) & 0xff;
|
|
/* Fall through. */
|
|
case 2:
|
|
field[1] = (value >> 8) & 0xff;
|
|
/* Fall through. */
|
|
case 1:
|
|
field[0] = value & 0xff;
|
|
break;
|
|
|
|
default:
|
|
error (_("Unhandled data length: %d\n"), size);
|
|
abort ();
|
|
}
|
|
}
|
|
|
|
void
|
|
byte_put_big_endian (unsigned char * field, elf_vma value, int size)
|
|
{
|
|
switch (size)
|
|
{
|
|
case 8:
|
|
field[7] = value & 0xff;
|
|
field[6] = (value >> 8) & 0xff;
|
|
field[5] = (value >> 16) & 0xff;
|
|
field[4] = (value >> 24) & 0xff;
|
|
value >>= 16;
|
|
value >>= 16;
|
|
/* Fall through. */
|
|
case 4:
|
|
field[3] = value & 0xff;
|
|
value >>= 8;
|
|
/* Fall through. */
|
|
case 3:
|
|
field[2] = value & 0xff;
|
|
value >>= 8;
|
|
/* Fall through. */
|
|
case 2:
|
|
field[1] = value & 0xff;
|
|
value >>= 8;
|
|
/* Fall through. */
|
|
case 1:
|
|
field[0] = value & 0xff;
|
|
break;
|
|
|
|
default:
|
|
error (_("Unhandled data length: %d\n"), size);
|
|
abort ();
|
|
}
|
|
}
|
|
|
|
elf_vma (*byte_get) (const unsigned char *, int);
|
|
|
|
elf_vma
|
|
byte_get_little_endian (const unsigned char *field, int size)
|
|
{
|
|
switch (size)
|
|
{
|
|
case 1:
|
|
return *field;
|
|
|
|
case 2:
|
|
return ((unsigned int) (field[0]))
|
|
| (((unsigned int) (field[1])) << 8);
|
|
|
|
case 3:
|
|
return ((unsigned long) (field[0]))
|
|
| (((unsigned long) (field[1])) << 8)
|
|
| (((unsigned long) (field[2])) << 16);
|
|
|
|
case 4:
|
|
return ((unsigned long) (field[0]))
|
|
| (((unsigned long) (field[1])) << 8)
|
|
| (((unsigned long) (field[2])) << 16)
|
|
| (((unsigned long) (field[3])) << 24);
|
|
|
|
case 5:
|
|
if (sizeof (elf_vma) == 8)
|
|
return ((elf_vma) (field[0]))
|
|
| (((elf_vma) (field[1])) << 8)
|
|
| (((elf_vma) (field[2])) << 16)
|
|
| (((elf_vma) (field[3])) << 24)
|
|
| (((elf_vma) (field[4])) << 32);
|
|
else if (sizeof (elf_vma) == 4)
|
|
/* We want to extract data from an 8 byte wide field and
|
|
place it into a 4 byte wide field. Since this is a little
|
|
endian source we can just use the 4 byte extraction code. */
|
|
return ((unsigned long) (field[0]))
|
|
| (((unsigned long) (field[1])) << 8)
|
|
| (((unsigned long) (field[2])) << 16)
|
|
| (((unsigned long) (field[3])) << 24);
|
|
/* Fall through. */
|
|
|
|
case 6:
|
|
if (sizeof (elf_vma) == 8)
|
|
return ((elf_vma) (field[0]))
|
|
| (((elf_vma) (field[1])) << 8)
|
|
| (((elf_vma) (field[2])) << 16)
|
|
| (((elf_vma) (field[3])) << 24)
|
|
| (((elf_vma) (field[4])) << 32)
|
|
| (((elf_vma) (field[5])) << 40);
|
|
else if (sizeof (elf_vma) == 4)
|
|
/* We want to extract data from an 8 byte wide field and
|
|
place it into a 4 byte wide field. Since this is a little
|
|
endian source we can just use the 4 byte extraction code. */
|
|
return ((unsigned long) (field[0]))
|
|
| (((unsigned long) (field[1])) << 8)
|
|
| (((unsigned long) (field[2])) << 16)
|
|
| (((unsigned long) (field[3])) << 24);
|
|
/* Fall through. */
|
|
|
|
case 7:
|
|
if (sizeof (elf_vma) == 8)
|
|
return ((elf_vma) (field[0]))
|
|
| (((elf_vma) (field[1])) << 8)
|
|
| (((elf_vma) (field[2])) << 16)
|
|
| (((elf_vma) (field[3])) << 24)
|
|
| (((elf_vma) (field[4])) << 32)
|
|
| (((elf_vma) (field[5])) << 40)
|
|
| (((elf_vma) (field[6])) << 48);
|
|
else if (sizeof (elf_vma) == 4)
|
|
/* We want to extract data from an 8 byte wide field and
|
|
place it into a 4 byte wide field. Since this is a little
|
|
endian source we can just use the 4 byte extraction code. */
|
|
return ((unsigned long) (field[0]))
|
|
| (((unsigned long) (field[1])) << 8)
|
|
| (((unsigned long) (field[2])) << 16)
|
|
| (((unsigned long) (field[3])) << 24);
|
|
/* Fall through. */
|
|
|
|
case 8:
|
|
if (sizeof (elf_vma) == 8)
|
|
return ((elf_vma) (field[0]))
|
|
| (((elf_vma) (field[1])) << 8)
|
|
| (((elf_vma) (field[2])) << 16)
|
|
| (((elf_vma) (field[3])) << 24)
|
|
| (((elf_vma) (field[4])) << 32)
|
|
| (((elf_vma) (field[5])) << 40)
|
|
| (((elf_vma) (field[6])) << 48)
|
|
| (((elf_vma) (field[7])) << 56);
|
|
else if (sizeof (elf_vma) == 4)
|
|
/* We want to extract data from an 8 byte wide field and
|
|
place it into a 4 byte wide field. Since this is a little
|
|
endian source we can just use the 4 byte extraction code. */
|
|
return ((unsigned long) (field[0]))
|
|
| (((unsigned long) (field[1])) << 8)
|
|
| (((unsigned long) (field[2])) << 16)
|
|
| (((unsigned long) (field[3])) << 24);
|
|
/* Fall through. */
|
|
|
|
default:
|
|
error (_("Unhandled data length: %d\n"), size);
|
|
abort ();
|
|
}
|
|
}
|
|
|
|
elf_vma
|
|
byte_get_big_endian (const unsigned char *field, int size)
|
|
{
|
|
switch (size)
|
|
{
|
|
case 1:
|
|
return *field;
|
|
|
|
case 2:
|
|
return ((unsigned int) (field[1])) | (((int) (field[0])) << 8);
|
|
|
|
case 3:
|
|
return ((unsigned long) (field[2]))
|
|
| (((unsigned long) (field[1])) << 8)
|
|
| (((unsigned long) (field[0])) << 16);
|
|
|
|
case 4:
|
|
return ((unsigned long) (field[3]))
|
|
| (((unsigned long) (field[2])) << 8)
|
|
| (((unsigned long) (field[1])) << 16)
|
|
| (((unsigned long) (field[0])) << 24);
|
|
|
|
case 5:
|
|
if (sizeof (elf_vma) == 8)
|
|
return ((elf_vma) (field[4]))
|
|
| (((elf_vma) (field[3])) << 8)
|
|
| (((elf_vma) (field[2])) << 16)
|
|
| (((elf_vma) (field[1])) << 24)
|
|
| (((elf_vma) (field[0])) << 32);
|
|
else if (sizeof (elf_vma) == 4)
|
|
{
|
|
/* Although we are extracting data from an 8 byte wide field,
|
|
we are returning only 4 bytes of data. */
|
|
field += 1;
|
|
return ((unsigned long) (field[3]))
|
|
| (((unsigned long) (field[2])) << 8)
|
|
| (((unsigned long) (field[1])) << 16)
|
|
| (((unsigned long) (field[0])) << 24);
|
|
}
|
|
/* Fall through. */
|
|
|
|
case 6:
|
|
if (sizeof (elf_vma) == 8)
|
|
return ((elf_vma) (field[5]))
|
|
| (((elf_vma) (field[4])) << 8)
|
|
| (((elf_vma) (field[3])) << 16)
|
|
| (((elf_vma) (field[2])) << 24)
|
|
| (((elf_vma) (field[1])) << 32)
|
|
| (((elf_vma) (field[0])) << 40);
|
|
else if (sizeof (elf_vma) == 4)
|
|
{
|
|
/* Although we are extracting data from an 8 byte wide field,
|
|
we are returning only 4 bytes of data. */
|
|
field += 2;
|
|
return ((unsigned long) (field[3]))
|
|
| (((unsigned long) (field[2])) << 8)
|
|
| (((unsigned long) (field[1])) << 16)
|
|
| (((unsigned long) (field[0])) << 24);
|
|
}
|
|
/* Fall through. */
|
|
|
|
case 7:
|
|
if (sizeof (elf_vma) == 8)
|
|
return ((elf_vma) (field[6]))
|
|
| (((elf_vma) (field[5])) << 8)
|
|
| (((elf_vma) (field[4])) << 16)
|
|
| (((elf_vma) (field[3])) << 24)
|
|
| (((elf_vma) (field[2])) << 32)
|
|
| (((elf_vma) (field[1])) << 40)
|
|
| (((elf_vma) (field[0])) << 48);
|
|
else if (sizeof (elf_vma) == 4)
|
|
{
|
|
/* Although we are extracting data from an 8 byte wide field,
|
|
we are returning only 4 bytes of data. */
|
|
field += 3;
|
|
return ((unsigned long) (field[3]))
|
|
| (((unsigned long) (field[2])) << 8)
|
|
| (((unsigned long) (field[1])) << 16)
|
|
| (((unsigned long) (field[0])) << 24);
|
|
}
|
|
/* Fall through. */
|
|
|
|
case 8:
|
|
if (sizeof (elf_vma) == 8)
|
|
return ((elf_vma) (field[7]))
|
|
| (((elf_vma) (field[6])) << 8)
|
|
| (((elf_vma) (field[5])) << 16)
|
|
| (((elf_vma) (field[4])) << 24)
|
|
| (((elf_vma) (field[3])) << 32)
|
|
| (((elf_vma) (field[2])) << 40)
|
|
| (((elf_vma) (field[1])) << 48)
|
|
| (((elf_vma) (field[0])) << 56);
|
|
else if (sizeof (elf_vma) == 4)
|
|
{
|
|
/* Although we are extracting data from an 8 byte wide field,
|
|
we are returning only 4 bytes of data. */
|
|
field += 4;
|
|
return ((unsigned long) (field[3]))
|
|
| (((unsigned long) (field[2])) << 8)
|
|
| (((unsigned long) (field[1])) << 16)
|
|
| (((unsigned long) (field[0])) << 24);
|
|
}
|
|
/* Fall through. */
|
|
|
|
default:
|
|
error (_("Unhandled data length: %d\n"), size);
|
|
abort ();
|
|
}
|
|
}
|
|
|
|
elf_vma
|
|
byte_get_signed (const unsigned char *field, int size)
|
|
{
|
|
elf_vma x = byte_get (field, size);
|
|
|
|
switch (size)
|
|
{
|
|
case 1:
|
|
return (x ^ 0x80) - 0x80;
|
|
case 2:
|
|
return (x ^ 0x8000) - 0x8000;
|
|
case 3:
|
|
return (x ^ 0x800000) - 0x800000;
|
|
case 4:
|
|
return (x ^ 0x80000000) - 0x80000000;
|
|
case 5:
|
|
case 6:
|
|
case 7:
|
|
case 8:
|
|
/* Reads of 5-, 6-, and 7-byte numbers are the result of
|
|
trying to read past the end of a buffer, and will therefore
|
|
not have meaningful values, so we don't try to deal with
|
|
the sign in these cases. */
|
|
return x;
|
|
default:
|
|
abort ();
|
|
}
|
|
}
|
|
|
|
/* Return the high-order 32-bits and the low-order 32-bits
|
|
of an 8-byte value separately. */
|
|
|
|
void
|
|
byte_get_64 (const unsigned char *field, elf_vma *high, elf_vma *low)
|
|
{
|
|
if (byte_get == byte_get_big_endian)
|
|
{
|
|
*high = byte_get_big_endian (field, 4);
|
|
*low = byte_get_big_endian (field + 4, 4);
|
|
}
|
|
else
|
|
{
|
|
*high = byte_get_little_endian (field + 4, 4);
|
|
*low = byte_get_little_endian (field, 4);
|
|
}
|
|
return;
|
|
}
|
|
|
|
/* Return the path name for a proxy entry in a thin archive, adjusted
|
|
relative to the path name of the thin archive itself if necessary.
|
|
Always returns a pointer to malloc'ed memory. */
|
|
|
|
char *
|
|
adjust_relative_path (const char *file_name, const char *name,
|
|
unsigned long name_len)
|
|
{
|
|
char * member_file_name;
|
|
const char * base_name = lbasename (file_name);
|
|
size_t amt;
|
|
|
|
/* This is a proxy entry for a thin archive member.
|
|
If the extended name table contains an absolute path
|
|
name, or if the archive is in the current directory,
|
|
use the path name as given. Otherwise, we need to
|
|
find the member relative to the directory where the
|
|
archive is located. */
|
|
if (IS_ABSOLUTE_PATH (name) || base_name == file_name)
|
|
{
|
|
amt = name_len + 1;
|
|
if (amt == 0)
|
|
return NULL;
|
|
member_file_name = (char *) malloc (amt);
|
|
if (member_file_name == NULL)
|
|
{
|
|
error (_("Out of memory\n"));
|
|
return NULL;
|
|
}
|
|
memcpy (member_file_name, name, name_len);
|
|
member_file_name[name_len] = '\0';
|
|
}
|
|
else
|
|
{
|
|
/* Concatenate the path components of the archive file name
|
|
to the relative path name from the extended name table. */
|
|
size_t prefix_len = base_name - file_name;
|
|
|
|
amt = prefix_len + name_len + 1;
|
|
/* PR 17531: file: 2896dc8b
|
|
Catch wraparound. */
|
|
if (amt < prefix_len || amt < name_len)
|
|
{
|
|
error (_("Abnormal length of thin archive member name: %lx\n"),
|
|
name_len);
|
|
return NULL;
|
|
}
|
|
|
|
member_file_name = (char *) malloc (amt);
|
|
if (member_file_name == NULL)
|
|
{
|
|
error (_("Out of memory\n"));
|
|
return NULL;
|
|
}
|
|
memcpy (member_file_name, file_name, prefix_len);
|
|
memcpy (member_file_name + prefix_len, name, name_len);
|
|
member_file_name[prefix_len + name_len] = '\0';
|
|
}
|
|
return member_file_name;
|
|
}
|
|
|
|
/* Processes the archive index table and symbol table in ARCH.
|
|
Entries in the index table are SIZEOF_AR_INDEX bytes long.
|
|
Fills in ARCH->next_arhdr_offset and ARCH->arhdr.
|
|
If READ_SYMBOLS is true then fills in ARCH->index_num, ARCH->index_array,
|
|
ARCH->sym_size and ARCH->sym_table.
|
|
It is the caller's responsibility to free ARCH->index_array and
|
|
ARCH->sym_table.
|
|
Returns TRUE upon success, FALSE otherwise.
|
|
If failure occurs an error message is printed. */
|
|
|
|
static bfd_boolean
|
|
process_archive_index_and_symbols (struct archive_info * arch,
|
|
unsigned int sizeof_ar_index,
|
|
bfd_boolean read_symbols)
|
|
{
|
|
size_t got;
|
|
unsigned long size;
|
|
char fmag_save;
|
|
|
|
fmag_save = arch->arhdr.ar_fmag[0];
|
|
arch->arhdr.ar_fmag[0] = 0;
|
|
size = strtoul (arch->arhdr.ar_size, NULL, 10);
|
|
arch->arhdr.ar_fmag[0] = fmag_save;
|
|
/* PR 17531: file: 912bd7de. */
|
|
if ((signed long) size < 0)
|
|
{
|
|
error (_("%s: invalid archive header size: %ld\n"),
|
|
arch->file_name, size);
|
|
return FALSE;
|
|
}
|
|
|
|
size = size + (size & 1);
|
|
|
|
arch->next_arhdr_offset += sizeof arch->arhdr + size;
|
|
|
|
if (! read_symbols)
|
|
{
|
|
if (fseek (arch->file, size, SEEK_CUR) != 0)
|
|
{
|
|
error (_("%s: failed to skip archive symbol table\n"),
|
|
arch->file_name);
|
|
return FALSE;
|
|
}
|
|
}
|
|
else
|
|
{
|
|
unsigned long i;
|
|
/* A buffer used to hold numbers read in from an archive index.
|
|
These are always SIZEOF_AR_INDEX bytes long and stored in
|
|
big-endian format. */
|
|
unsigned char integer_buffer[sizeof arch->index_num];
|
|
unsigned char * index_buffer;
|
|
|
|
assert (sizeof_ar_index <= sizeof integer_buffer);
|
|
|
|
/* Check the size of the archive index. */
|
|
if (size < sizeof_ar_index)
|
|
{
|
|
error (_("%s: the archive index is empty\n"), arch->file_name);
|
|
return FALSE;
|
|
}
|
|
|
|
/* Read the number of entries in the archive index. */
|
|
got = fread (integer_buffer, 1, sizeof_ar_index, arch->file);
|
|
if (got != sizeof_ar_index)
|
|
{
|
|
error (_("%s: failed to read archive index\n"), arch->file_name);
|
|
return FALSE;
|
|
}
|
|
|
|
arch->index_num = byte_get_big_endian (integer_buffer, sizeof_ar_index);
|
|
size -= sizeof_ar_index;
|
|
|
|
if (size < arch->index_num * sizeof_ar_index
|
|
/* PR 17531: file: 585515d1. */
|
|
|| size < arch->index_num)
|
|
{
|
|
error (_("%s: the archive index is supposed to have 0x%lx entries of %d bytes, but the size is only 0x%lx\n"),
|
|
arch->file_name, (long) arch->index_num, sizeof_ar_index, size);
|
|
return FALSE;
|
|
}
|
|
|
|
/* Read in the archive index. */
|
|
index_buffer = (unsigned char *)
|
|
malloc (arch->index_num * sizeof_ar_index);
|
|
if (index_buffer == NULL)
|
|
{
|
|
error (_("Out of memory whilst trying to read archive symbol index\n"));
|
|
return FALSE;
|
|
}
|
|
|
|
got = fread (index_buffer, sizeof_ar_index, arch->index_num, arch->file);
|
|
if (got != arch->index_num)
|
|
{
|
|
free (index_buffer);
|
|
error (_("%s: failed to read archive index\n"), arch->file_name);
|
|
return FALSE;
|
|
}
|
|
|
|
size -= arch->index_num * sizeof_ar_index;
|
|
|
|
/* Convert the index numbers into the host's numeric format. */
|
|
arch->index_array = (elf_vma *)
|
|
malloc (arch->index_num * sizeof (* arch->index_array));
|
|
if (arch->index_array == NULL)
|
|
{
|
|
free (index_buffer);
|
|
error (_("Out of memory whilst trying to convert the archive symbol index\n"));
|
|
return FALSE;
|
|
}
|
|
|
|
for (i = 0; i < arch->index_num; i++)
|
|
arch->index_array[i] =
|
|
byte_get_big_endian ((unsigned char *) (index_buffer + (i * sizeof_ar_index)),
|
|
sizeof_ar_index);
|
|
free (index_buffer);
|
|
|
|
/* The remaining space in the header is taken up by the symbol table. */
|
|
if (size < 1)
|
|
{
|
|
error (_("%s: the archive has an index but no symbols\n"),
|
|
arch->file_name);
|
|
return FALSE;
|
|
}
|
|
|
|
arch->sym_table = (char *) malloc (size);
|
|
if (arch->sym_table == NULL)
|
|
{
|
|
error (_("Out of memory whilst trying to read archive index symbol table\n"));
|
|
return FALSE;
|
|
}
|
|
|
|
arch->sym_size = size;
|
|
got = fread (arch->sym_table, 1, size, arch->file);
|
|
if (got != size)
|
|
{
|
|
error (_("%s: failed to read archive index symbol table\n"),
|
|
arch->file_name);
|
|
return FALSE;
|
|
}
|
|
}
|
|
|
|
/* Read the next archive header. */
|
|
got = fread (&arch->arhdr, 1, sizeof arch->arhdr, arch->file);
|
|
if (got != sizeof arch->arhdr && got != 0)
|
|
{
|
|
error (_("%s: failed to read archive header following archive index\n"),
|
|
arch->file_name);
|
|
return FALSE;
|
|
}
|
|
|
|
return TRUE;
|
|
}
|
|
|
|
/* Read the symbol table and long-name table from an archive. */
|
|
|
|
int
|
|
setup_archive (struct archive_info *arch, const char *file_name,
|
|
FILE *file, bfd_boolean is_thin_archive,
|
|
bfd_boolean read_symbols)
|
|
{
|
|
size_t got;
|
|
|
|
arch->file_name = strdup (file_name);
|
|
arch->file = file;
|
|
arch->index_num = 0;
|
|
arch->index_array = NULL;
|
|
arch->sym_table = NULL;
|
|
arch->sym_size = 0;
|
|
arch->longnames = NULL;
|
|
arch->longnames_size = 0;
|
|
arch->nested_member_origin = 0;
|
|
arch->is_thin_archive = is_thin_archive;
|
|
arch->uses_64bit_indicies = FALSE;
|
|
arch->next_arhdr_offset = SARMAG;
|
|
|
|
/* Read the first archive member header. */
|
|
if (fseek (file, SARMAG, SEEK_SET) != 0)
|
|
{
|
|
error (_("%s: failed to seek to first archive header\n"), file_name);
|
|
return 1;
|
|
}
|
|
got = fread (&arch->arhdr, 1, sizeof arch->arhdr, file);
|
|
if (got != sizeof arch->arhdr)
|
|
{
|
|
if (got == 0)
|
|
return 0;
|
|
|
|
error (_("%s: failed to read archive header\n"), file_name);
|
|
return 1;
|
|
}
|
|
|
|
/* See if this is the archive symbol table. */
|
|
if (const_strneq (arch->arhdr.ar_name, "/ "))
|
|
{
|
|
if (! process_archive_index_and_symbols (arch, 4, read_symbols))
|
|
return 1;
|
|
}
|
|
else if (const_strneq (arch->arhdr.ar_name, "/SYM64/ "))
|
|
{
|
|
arch->uses_64bit_indicies = TRUE;
|
|
if (! process_archive_index_and_symbols (arch, 8, read_symbols))
|
|
return 1;
|
|
}
|
|
else if (read_symbols)
|
|
printf (_("%s has no archive index\n"), file_name);
|
|
|
|
if (const_strneq (arch->arhdr.ar_name, "// "))
|
|
{
|
|
/* This is the archive string table holding long member names. */
|
|
char fmag_save = arch->arhdr.ar_fmag[0];
|
|
arch->arhdr.ar_fmag[0] = 0;
|
|
arch->longnames_size = strtoul (arch->arhdr.ar_size, NULL, 10);
|
|
arch->arhdr.ar_fmag[0] = fmag_save;
|
|
/* PR 17531: file: 01068045. */
|
|
if (arch->longnames_size < 8)
|
|
{
|
|
error (_("%s: long name table is too small, (size = %ld)\n"),
|
|
file_name, arch->longnames_size);
|
|
return 1;
|
|
}
|
|
/* PR 17531: file: 639d6a26. */
|
|
if ((signed long) arch->longnames_size < 0)
|
|
{
|
|
error (_("%s: long name table is too big, (size = 0x%lx)\n"),
|
|
file_name, arch->longnames_size);
|
|
return 1;
|
|
}
|
|
|
|
arch->next_arhdr_offset += sizeof arch->arhdr + arch->longnames_size;
|
|
|
|
/* Plus one to allow for a string terminator. */
|
|
arch->longnames = (char *) malloc (arch->longnames_size + 1);
|
|
if (arch->longnames == NULL)
|
|
{
|
|
error (_("Out of memory reading long symbol names in archive\n"));
|
|
return 1;
|
|
}
|
|
|
|
if (fread (arch->longnames, arch->longnames_size, 1, file) != 1)
|
|
{
|
|
free (arch->longnames);
|
|
arch->longnames = NULL;
|
|
error (_("%s: failed to read long symbol name string table\n"),
|
|
file_name);
|
|
return 1;
|
|
}
|
|
|
|
if ((arch->longnames_size & 1) != 0)
|
|
getc (file);
|
|
|
|
arch->longnames[arch->longnames_size] = 0;
|
|
}
|
|
|
|
return 0;
|
|
}
|
|
|
|
/* Open and setup a nested archive, if not already open. */
|
|
|
|
int
|
|
setup_nested_archive (struct archive_info *nested_arch,
|
|
const char *member_file_name)
|
|
{
|
|
FILE * member_file;
|
|
|
|
/* Have we already setup this archive? */
|
|
if (nested_arch->file_name != NULL
|
|
&& streq (nested_arch->file_name, member_file_name))
|
|
return 0;
|
|
|
|
/* Close previous file and discard cached information. */
|
|
if (nested_arch->file != NULL)
|
|
fclose (nested_arch->file);
|
|
release_archive (nested_arch);
|
|
|
|
member_file = fopen (member_file_name, "rb");
|
|
if (member_file == NULL)
|
|
return 1;
|
|
return setup_archive (nested_arch, member_file_name, member_file,
|
|
FALSE, FALSE);
|
|
}
|
|
|
|
/* Release the memory used for the archive information. */
|
|
|
|
void
|
|
release_archive (struct archive_info * arch)
|
|
{
|
|
if (arch->file_name != NULL)
|
|
free (arch->file_name);
|
|
if (arch->index_array != NULL)
|
|
free (arch->index_array);
|
|
if (arch->sym_table != NULL)
|
|
free (arch->sym_table);
|
|
if (arch->longnames != NULL)
|
|
free (arch->longnames);
|
|
}
|
|
|
|
/* Get the name of an archive member from the current archive header.
|
|
For simple names, this will modify the ar_name field of the current
|
|
archive header. For long names, it will return a pointer to the
|
|
longnames table. For nested archives, it will open the nested archive
|
|
and get the name recursively. NESTED_ARCH is a single-entry cache so
|
|
we don't keep rereading the same information from a nested archive. */
|
|
|
|
char *
|
|
get_archive_member_name (struct archive_info *arch,
|
|
struct archive_info *nested_arch)
|
|
{
|
|
unsigned long j, k;
|
|
|
|
if (arch->arhdr.ar_name[0] == '/')
|
|
{
|
|
/* We have a long name. */
|
|
char *endp;
|
|
char *member_file_name;
|
|
char *member_name;
|
|
char fmag_save;
|
|
|
|
if (arch->longnames == NULL || arch->longnames_size == 0)
|
|
{
|
|
error (_("Archive member uses long names, but no longname table found\n"));
|
|
return NULL;
|
|
}
|
|
|
|
arch->nested_member_origin = 0;
|
|
fmag_save = arch->arhdr.ar_fmag[0];
|
|
arch->arhdr.ar_fmag[0] = 0;
|
|
k = j = strtoul (arch->arhdr.ar_name + 1, &endp, 10);
|
|
if (arch->is_thin_archive && endp != NULL && * endp == ':')
|
|
arch->nested_member_origin = strtoul (endp + 1, NULL, 10);
|
|
arch->arhdr.ar_fmag[0] = fmag_save;
|
|
|
|
if (j > arch->longnames_size)
|
|
{
|
|
error (_("Found long name index (%ld) beyond end of long name table\n"),j);
|
|
return NULL;
|
|
}
|
|
while ((j < arch->longnames_size)
|
|
&& (arch->longnames[j] != '\n')
|
|
&& (arch->longnames[j] != '\0'))
|
|
j++;
|
|
if (j > 0 && arch->longnames[j-1] == '/')
|
|
j--;
|
|
if (j > arch->longnames_size)
|
|
j = arch->longnames_size;
|
|
arch->longnames[j] = '\0';
|
|
|
|
if (!arch->is_thin_archive || arch->nested_member_origin == 0)
|
|
return arch->longnames + k;
|
|
|
|
/* PR 17531: file: 2896dc8b. */
|
|
if (k >= j)
|
|
{
|
|
error (_("Invalid Thin archive member name\n"));
|
|
return NULL;
|
|
}
|
|
|
|
/* This is a proxy for a member of a nested archive.
|
|
Find the name of the member in that archive. */
|
|
member_file_name = adjust_relative_path (arch->file_name,
|
|
arch->longnames + k, j - k);
|
|
if (member_file_name != NULL
|
|
&& setup_nested_archive (nested_arch, member_file_name) == 0)
|
|
{
|
|
member_name = get_archive_member_name_at (nested_arch,
|
|
arch->nested_member_origin,
|
|
NULL);
|
|
if (member_name != NULL)
|
|
{
|
|
free (member_file_name);
|
|
return member_name;
|
|
}
|
|
}
|
|
free (member_file_name);
|
|
|
|
/* Last resort: just return the name of the nested archive. */
|
|
return arch->longnames + k;
|
|
}
|
|
|
|
/* We have a normal (short) name. */
|
|
for (j = 0; j < sizeof (arch->arhdr.ar_name); j++)
|
|
if (arch->arhdr.ar_name[j] == '/')
|
|
{
|
|
arch->arhdr.ar_name[j] = '\0';
|
|
return arch->arhdr.ar_name;
|
|
}
|
|
|
|
/* The full ar_name field is used. Don't rely on ar_date starting
|
|
with a zero byte. */
|
|
{
|
|
char *name = xmalloc (sizeof (arch->arhdr.ar_name) + 1);
|
|
memcpy (name, arch->arhdr.ar_name, sizeof (arch->arhdr.ar_name));
|
|
name[sizeof (arch->arhdr.ar_name)] = '\0';
|
|
return name;
|
|
}
|
|
}
|
|
|
|
/* Get the name of an archive member at a given OFFSET within an archive
|
|
ARCH. */
|
|
|
|
char *
|
|
get_archive_member_name_at (struct archive_info *arch,
|
|
unsigned long offset,
|
|
struct archive_info *nested_arch)
|
|
{
|
|
size_t got;
|
|
|
|
if (fseek (arch->file, offset, SEEK_SET) != 0)
|
|
{
|
|
error (_("%s: failed to seek to next file name\n"), arch->file_name);
|
|
return NULL;
|
|
}
|
|
got = fread (&arch->arhdr, 1, sizeof arch->arhdr, arch->file);
|
|
if (got != sizeof arch->arhdr)
|
|
{
|
|
error (_("%s: failed to read archive header\n"), arch->file_name);
|
|
return NULL;
|
|
}
|
|
if (memcmp (arch->arhdr.ar_fmag, ARFMAG, 2) != 0)
|
|
{
|
|
error (_("%s: did not find a valid archive header\n"),
|
|
arch->file_name);
|
|
return NULL;
|
|
}
|
|
|
|
return get_archive_member_name (arch, nested_arch);
|
|
}
|
|
|
|
/* Construct a string showing the name of the archive member, qualified
|
|
with the name of the containing archive file. For thin archives, we
|
|
use square brackets to denote the indirection. For nested archives,
|
|
we show the qualified name of the external member inside the square
|
|
brackets (e.g., "thin.a[normal.a(foo.o)]"). */
|
|
|
|
char *
|
|
make_qualified_name (struct archive_info * arch,
|
|
struct archive_info * nested_arch,
|
|
const char *member_name)
|
|
{
|
|
const char * error_name = _("<corrupt>");
|
|
size_t len;
|
|
char * name;
|
|
|
|
len = strlen (arch->file_name) + strlen (member_name) + 3;
|
|
if (arch->is_thin_archive
|
|
&& arch->nested_member_origin != 0)
|
|
{
|
|
/* PR 15140: Allow for corrupt thin archives. */
|
|
if (nested_arch->file_name)
|
|
len += strlen (nested_arch->file_name) + 2;
|
|
else
|
|
len += strlen (error_name) + 2;
|
|
}
|
|
|
|
name = (char *) malloc (len);
|
|
if (name == NULL)
|
|
{
|
|
error (_("Out of memory\n"));
|
|
return NULL;
|
|
}
|
|
|
|
if (arch->is_thin_archive
|
|
&& arch->nested_member_origin != 0)
|
|
{
|
|
if (nested_arch->file_name)
|
|
snprintf (name, len, "%s[%s(%s)]", arch->file_name,
|
|
nested_arch->file_name, member_name);
|
|
else
|
|
snprintf (name, len, "%s[%s(%s)]", arch->file_name,
|
|
error_name, member_name);
|
|
}
|
|
else if (arch->is_thin_archive)
|
|
snprintf (name, len, "%s[%s]", arch->file_name, member_name);
|
|
else
|
|
snprintf (name, len, "%s(%s)", arch->file_name, member_name);
|
|
|
|
return name;
|
|
}
|