binutils-gdb/binutils/size.c

661 lines
16 KiB
C
Raw Normal View History

1999-05-03 09:29:11 +02:00
/* size.c -- report size of various sections of an executable file.
Copyright (C) 1991-2019 Free Software Foundation, Inc.
1999-05-03 09:29:11 +02:00
2002-01-25 16:37:04 +01:00
This file is part of GNU Binutils.
1999-05-03 09:29:11 +02:00
2002-01-25 16:37:04 +01:00
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
2007-07-05 18:54:46 +02:00
the Free Software Foundation; either version 3 of the License, or
2002-01-25 16:37:04 +01:00
(at your option) any later version.
1999-05-03 09:29:11 +02:00
2002-01-25 16:37:04 +01:00
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.
1999-05-03 09:29:11 +02:00
2002-01-25 16:37:04 +01:00
You should have received a copy of the GNU General Public License
along with this program; if not, write to the Free Software
2007-07-05 18:54:46 +02:00
Foundation, Inc., 51 Franklin Street - Fifth Floor, Boston,
MA 02110-1301, USA. */
1999-05-03 09:29:11 +02:00
/* Extensions/incompatibilities:
o - BSD output has filenames at the end.
o - BSD output can appear in different radicies.
o - SysV output has less redundant whitespace. Filename comes at end.
o - SysV output doesn't show VMA which is always the same as the PMA.
o - We also handle core files.
o - We also handle archives.
If you write shell scripts which manipulate this info then you may be
2002-01-25 16:37:04 +01:00
out of luck; there's no --compatibility or --pedantic option. */
1999-05-03 09:29:11 +02:00
#include "sysdep.h"
1999-05-03 09:29:11 +02:00
#include "bfd.h"
#include "libiberty.h"
#include "getopt.h"
#include "bucomm.h"
1999-05-03 09:29:11 +02:00
#ifndef BSD_DEFAULT
#define BSD_DEFAULT 1
#endif
/* Program options. */
2007-07-05 11:01:28 +02:00
static enum
1999-05-03 09:29:11 +02:00
{
decimal, octal, hex
2002-01-25 16:37:04 +01:00
}
radix = decimal;
binutils: Add new GNU format mode to `size` utility The size tool currently defaults to berkeley format output. However, this output format has a weird quirk, read-only data is counted against the text sections, not the data sections. The code offers no real explanation for why this is, but I'm reluctant to change it for two reasons, first, I'm assuming it probably makes sense in some case that I'm not thinking of (maybe a target where sections are not marked executable, and so there's no distinction between read-only data and code), and second, the code has been this way for at least 20 years, I worry that changing things now might cause more confusion than it solves. This commit then introduces a new output format for the size tool, this new format displays the results in a similar manor to the berkeley format, but counts read-only data in the data column, and only executable sections are counted in the text column. Given that this is a brand new output format I've gone ahead and simplified things a little, while the berkeley format displays the total twice, once in decimal and once in hex, the new display format just displays the total in decimal. Of course, there's still the '--radix' option which can be used to display all the results in hexadecimal or octal. I've called the new format 'gnu', so '--format=gnu' or '-G' are used to access it. binutils/ChangeLog: * size.c (berkeley_format): Delete. (enum output_format): New enum. (selected_output_format): New variable. (usage): Update to mention GNU format. (main): Update to extract options, and select format as needed. Handle GNU format where needed. (berkeley_sum): Renamed to... (berkeley_or_gnu_sum): ...this, and updated to handle both formats. (berkeley_format): Renamed to... (berkeley_or_gnu_format): ...this, and updated to handle both formats. (print_sizes): Handle GNU format. * doc/binutils.texi (size): Document new GNU format. * testsuite/binutils-all/size.exp: Add test of extended functionality. * NEWS: Mention new functionality.
2019-01-24 15:27:27 +01:00
/* Select the desired output format. */
enum output_format
{
FORMAT_BERKLEY,
FORMAT_SYSV,
FORMAT_GNU
};
static enum output_format selected_output_format =
#if BSD_DEFAULT
FORMAT_BERKLEY
#else
FORMAT_SYSV
#endif
;
2007-07-05 11:01:28 +02:00
static int show_version = 0;
static int show_help = 0;
static int show_totals = 0;
static int show_common = 0;
2002-01-25 16:37:04 +01:00
2007-07-05 11:01:28 +02:00
static bfd_size_type common_size;
2002-01-25 16:37:04 +01:00
static bfd_size_type total_bsssize;
static bfd_size_type total_datasize;
static bfd_size_type total_textsize;
1999-05-03 09:29:11 +02:00
/* Program exit status. */
2007-07-05 11:01:28 +02:00
static int return_code = 0;
1999-05-03 09:29:11 +02:00
static char *target = NULL;
2007-07-05 11:01:28 +02:00
/* Forward declarations. */
1999-05-03 09:29:11 +02:00
static void display_file (char *);
static void rprint_number (int, bfd_size_type);
static void print_sizes (bfd * file);
1999-05-03 09:29:11 +02:00
static void
usage (FILE *stream, int status)
1999-05-03 09:29:11 +02:00
{
2002-01-23 17:12:56 +01:00
fprintf (stream, _("Usage: %s [option(s)] [file(s)]\n"), program_name);
fprintf (stream, _(" Displays the sizes of sections inside binary files\n"));
fprintf (stream, _(" If no input file(s) are specified, a.out is assumed\n"));
2002-01-23 17:12:56 +01:00
fprintf (stream, _(" The options are:\n\
binutils: Add new GNU format mode to `size` utility The size tool currently defaults to berkeley format output. However, this output format has a weird quirk, read-only data is counted against the text sections, not the data sections. The code offers no real explanation for why this is, but I'm reluctant to change it for two reasons, first, I'm assuming it probably makes sense in some case that I'm not thinking of (maybe a target where sections are not marked executable, and so there's no distinction between read-only data and code), and second, the code has been this way for at least 20 years, I worry that changing things now might cause more confusion than it solves. This commit then introduces a new output format for the size tool, this new format displays the results in a similar manor to the berkeley format, but counts read-only data in the data column, and only executable sections are counted in the text column. Given that this is a brand new output format I've gone ahead and simplified things a little, while the berkeley format displays the total twice, once in decimal and once in hex, the new display format just displays the total in decimal. Of course, there's still the '--radix' option which can be used to display all the results in hexadecimal or octal. I've called the new format 'gnu', so '--format=gnu' or '-G' are used to access it. binutils/ChangeLog: * size.c (berkeley_format): Delete. (enum output_format): New enum. (selected_output_format): New variable. (usage): Update to mention GNU format. (main): Update to extract options, and select format as needed. Handle GNU format where needed. (berkeley_sum): Renamed to... (berkeley_or_gnu_sum): ...this, and updated to handle both formats. (berkeley_format): Renamed to... (berkeley_or_gnu_format): ...this, and updated to handle both formats. (print_sizes): Handle GNU format. * doc/binutils.texi (size): Document new GNU format. * testsuite/binutils-all/size.exp: Add test of extended functionality. * NEWS: Mention new functionality.
2019-01-24 15:27:27 +01:00
-A|-B|-G --format={sysv|berkeley|gnu} Select output style (default is %s)\n\
-o|-d|-x --radix={8|10|16} Display numbers in octal, decimal or hex\n\
2002-01-25 16:37:04 +01:00
-t --totals Display the total sizes (Berkeley only)\n\
2007-07-05 11:01:28 +02:00
--common Display total size for *COM* syms\n\
2002-01-23 17:12:56 +01:00
--target=<bfdname> Set the binary file format\n\
@<file> Read options from <file>\n\
2002-01-23 17:12:56 +01:00
-h --help Display this information\n\
-v --version Display the program's version\n\
\n"),
1999-05-03 09:29:11 +02:00
#if BSD_DEFAULT
2002-01-23 17:12:56 +01:00
"berkeley"
1999-05-03 09:29:11 +02:00
#else
2002-01-23 17:12:56 +01:00
"sysv"
1999-05-03 09:29:11 +02:00
#endif
2002-01-23 17:12:56 +01:00
);
1999-05-03 09:29:11 +02:00
list_supported_targets (program_name, stream);
top level: 2007-02-17 Mark Mitchell <mark@codesourcery.com> Nathan Sidwell <nathan@codesourcery.com> Vladimir Prus <vladimir@codesourcery.com Joseph Myers <joseph@codesourcery.com> * configure.ac (TOPLEVEL_CONFIGURE_ARGUMENTS): Fix quoting. * configure: Regenerate. bfd: 2007-02-17 Mark Mitchell <mark@codesourcery.com> Nathan Sidwell <nathan@codesourcery.com> Vladimir Prus <vladimir@codesourcery.com Joseph Myers <joseph@codesourcery.com> * configure.in (--with-pkgversion): New option. * configure: Regenerate. * Makefile.am (bfdver.h): Substitute for @bfd_version_package@. * Makefile.in: Regenerate. * version.h (BFD_VERSION_STRING): Define using @bfd_version_package@. bfd/doc: 2007-02-17 Mark Mitchell <mark@codesourcery.com> Nathan Sidwell <nathan@codesourcery.com> Vladimir Prus <vladimir@codesourcery.com Joseph Myers <joseph@codesourcery.com> * Makefile.in: Regenerate. binutils: 2007-02-17 Mark Mitchell <mark@codesourcery.com> Nathan Sidwell <nathan@codesourcery.com> Vladimir Prus <vladimir@codesourcery.com Joseph Myers <joseph@codesourcery.com> * configure.in (--with-bugurl): New option. * configure: Regenerate. * Makefile.am (REPORT_BUGS_TO): Define. (INCLUDES): Define REPORT_BUGS_TO. Regenerate dependencies. * Makefile.in: Regenerate. * doc/Makefile.in: Regenerate. * bucomm.h: Remove include of bin-bugs.h. * addr2line.c (usage): Don't print empty REPORT_BUGS_TO. * ar.c (usage): Pass s to list_supported_targets. Don't print empty REPORT_BUGS_TO. * coffdump.c (show_usage): Don't print empty REPORT_BUGS_TO. * cxxfilt.c (usage): Print bug url when giving help. * dlltool.c (usage): Likewise. * dllwrap.c (usage): Likewise. * nlmconv.c (show_usage): Don't print empty REPORT_BUGS_TO. * nm.c (usage): Likewise. * objcopy.c (copy_usage, strip_usage): Likewise. * objdump.c (usage): Likewise. * readelf.c ((usage): Likewise. Add STREAM argument. Adjust callers. * size.c (usage): Don't print empty REPORT_BUGS_TO. * srconv.c (show_usage): Likewise. * strings.c (usage): Likewise. * sysdymp.c (show_usage): Likewise. * windres.c (usage): Likewise. gas: 2007-02-17 Mark Mitchell <mark@codesourcery.com> Nathan Sidwell <nathan@codesourcery.com> Vladimir Prus <vladimir@codesourcery.com Joseph Myers <joseph@codesourcery.com> * configure.in (--with-bugurl): New option. * configure: Regenerate. * dep-in.sed: Remove bin-bugs.h. * Makefile.am (REPORT_BUGS_TO): Define. (INCLUDES): Define REPORT_BUGS_TO. (DEP_INCLUDES): Likewise. ($(OBJS)): No longer depend on bin-bugs.h. * Makefile.in: Regenerate. * doc/Makefile.in: Regenerate. * as.c (show_usage): Don't print empty REPORT_BUGS_TO. * as.h: Remove include of bin-bugs.h. gprof: 2007-02-17 Mark Mitchell <mark@codesourcery.com> Nathan Sidwell <nathan@codesourcery.com> Vladimir Prus <vladimir@codesourcery.com Joseph Myers <joseph@codesourcery.com> * configure.in (--with-pkgversion, --with-bugurl): New options. * configure: Regenerate. * Makefile.am (PKGVERSION, REPORT_BUGS_TO): Define. (INCLUDES): Define PKGVERSION and REPORT_BUGS_TO. Regenerate dependencies. * Makefile.in: Regenerate. * gprof.c (usage): Don't print empty REPORT_BUGS_TO. (main): Include PKGVERSION in version output. * gprof.h: Remove include of bin-bugs.h. include: 2007-02-17 Mark Mitchell <mark@codesourcery.com> Nathan Sidwell <nathan@codesourcery.com> Vladimir Prus <vladimir@codesourcery.com Joseph Myers <joseph@codesourcery.com> * bin-bugs.h: Remove. ld: 2007-02-17 Mark Mitchell <mark@codesourcery.com> Nathan Sidwell <nathan@codesourcery.com> Vladimir Prus <vladimir@codesourcery.com Joseph Myers <joseph@codesourcery.com> * configure.in (--with-bugurl): New option. * configure: Regenerate. * Makefile.am (REPORT_BUGS_TO): Define. (INCLUDES): Define REPORT_BUGS_TO. Regenerate dependencies. * Makefile.in: Regenerate. * ld.h: Remove include of bin-bugs.h. * lexsup.c (help): Don't print empty REPORT_BUGS_TO.
2007-02-17 14:33:57 +01:00
if (REPORT_BUGS_TO[0] && status == 0)
fprintf (stream, _("Report bugs to %s\n"), REPORT_BUGS_TO);
1999-05-03 09:29:11 +02:00
exit (status);
}
2007-07-05 11:01:28 +02:00
#define OPTION_FORMAT (200)
#define OPTION_RADIX (OPTION_FORMAT + 1)
#define OPTION_TARGET (OPTION_RADIX + 1)
static struct option long_options[] =
1999-05-03 09:29:11 +02:00
{
2007-07-05 11:01:28 +02:00
{"common", no_argument, &show_common, 1},
{"format", required_argument, 0, OPTION_FORMAT},
{"radix", required_argument, 0, OPTION_RADIX},
{"target", required_argument, 0, OPTION_TARGET},
2002-01-25 16:37:04 +01:00
{"totals", no_argument, &show_totals, 1},
1999-05-03 09:29:11 +02:00
{"version", no_argument, &show_version, 1},
{"help", no_argument, &show_help, 1},
{0, no_argument, 0, 0}
};
int main (int, char **);
1999-05-03 09:29:11 +02:00
int
main (int argc, char **argv)
1999-05-03 09:29:11 +02:00
{
int temp;
int c;
#if defined (HAVE_SETLOCALE) && defined (HAVE_LC_MESSAGES)
setlocale (LC_MESSAGES, "");
#endif
#if defined (HAVE_SETLOCALE)
setlocale (LC_CTYPE, "");
1999-05-03 09:29:11 +02:00
#endif
bindtextdomain (PACKAGE, LOCALEDIR);
textdomain (PACKAGE);
program_name = *argv;
xmalloc_set_program_name (program_name);
Fix memory access violations triggered by running strip on fuzzed binaries. PR binutils/17512 * coffcode.h (coff_set_arch_mach_hook): Check return value from bfd_malloc. (coff_slurp_line_table): Return FALSE if the line number information was corrupt. (coff_slurp_symbol_table): Return FALSE if the symbol information was corrupt. * mach-o.c (bfd_mach_o_bfd_copy_private_header_data): Always initialise the fields of the dyld_info structure. (bfd_mach_o_build_exec_seg_command): Replace assertion with an error message and a return value. (bfd_mach_o_layout_commands): Change the function to boolean. Return FALSE if the function fails. (bfd_mach_o_build_commands): Fail if bfd_mach_o_layout_commands fails. (bfd_mach_o_read_command): Fail if an unrecognised command is encountered. * peXXigen.c (_bfd_XXi_swap_aouthdr_in): Set bfd_error if the read fails. (slurp_symtab): Check the return from bfd_malloc. (_bfd_XX_bfd_copy_private_bfd_data_common): Fail if the copy encountered an error. (_bfd_XXi_final_link_postscript): Fail if a section could not be copied. * peicode.h (pe_bfd_object_p): Fail if the header could not be swapped in. * tekhex.c (first_phase): Fail if the section is too big. * versados.c (struct esdid): Add content_size field. (process_otr): Use and check the new field. (versados_get_section_contents): Check that the section exists and that the requested data is available. PR binutils/17512 * addr2line.c (main): Call bfd_set_error_program_name. * ar.c (main): Likewise. * coffdump.c (main): Likewise. * cxxfilt.c (main): Likewise. * dlltool.c (main): Likewise. * nlmconv.c (main): Likewise. * nm.c (main): Likewise. * objdump.c (main): Likewise. * size.c (main): Likewise. * srconv.c (main): Likewise. * strings.c (main): Likewise. * sysdump.c (main): Likewise. * windmc.c (main): Likewise. * windres.c (main): Likewise. * objcopy.c (main): Likewise. (copy_relocations_in_section): Check for relocs without associated symbol pointers.
2015-01-21 18:37:23 +01:00
bfd_set_error_program_name (program_name);
1999-05-03 09:29:11 +02:00
expandargv (&argc, &argv);
if (bfd_init () != BFD_INIT_MAGIC)
fatal (_("fatal error: libbfd ABI mismatch"));
1999-05-03 09:29:11 +02:00
set_default_bfd_target ();
binutils: Add new GNU format mode to `size` utility The size tool currently defaults to berkeley format output. However, this output format has a weird quirk, read-only data is counted against the text sections, not the data sections. The code offers no real explanation for why this is, but I'm reluctant to change it for two reasons, first, I'm assuming it probably makes sense in some case that I'm not thinking of (maybe a target where sections are not marked executable, and so there's no distinction between read-only data and code), and second, the code has been this way for at least 20 years, I worry that changing things now might cause more confusion than it solves. This commit then introduces a new output format for the size tool, this new format displays the results in a similar manor to the berkeley format, but counts read-only data in the data column, and only executable sections are counted in the text column. Given that this is a brand new output format I've gone ahead and simplified things a little, while the berkeley format displays the total twice, once in decimal and once in hex, the new display format just displays the total in decimal. Of course, there's still the '--radix' option which can be used to display all the results in hexadecimal or octal. I've called the new format 'gnu', so '--format=gnu' or '-G' are used to access it. binutils/ChangeLog: * size.c (berkeley_format): Delete. (enum output_format): New enum. (selected_output_format): New variable. (usage): Update to mention GNU format. (main): Update to extract options, and select format as needed. Handle GNU format where needed. (berkeley_sum): Renamed to... (berkeley_or_gnu_sum): ...this, and updated to handle both formats. (berkeley_format): Renamed to... (berkeley_or_gnu_format): ...this, and updated to handle both formats. (print_sizes): Handle GNU format. * doc/binutils.texi (size): Document new GNU format. * testsuite/binutils-all/size.exp: Add test of extended functionality. * NEWS: Mention new functionality.
2019-01-24 15:27:27 +01:00
while ((c = getopt_long (argc, argv, "ABGHhVvdfotx", long_options,
1999-05-03 09:29:11 +02:00
(int *) 0)) != EOF)
switch (c)
{
2007-07-05 11:01:28 +02:00
case OPTION_FORMAT:
1999-05-03 09:29:11 +02:00
switch (*optarg)
{
case 'B':
case 'b':
binutils: Add new GNU format mode to `size` utility The size tool currently defaults to berkeley format output. However, this output format has a weird quirk, read-only data is counted against the text sections, not the data sections. The code offers no real explanation for why this is, but I'm reluctant to change it for two reasons, first, I'm assuming it probably makes sense in some case that I'm not thinking of (maybe a target where sections are not marked executable, and so there's no distinction between read-only data and code), and second, the code has been this way for at least 20 years, I worry that changing things now might cause more confusion than it solves. This commit then introduces a new output format for the size tool, this new format displays the results in a similar manor to the berkeley format, but counts read-only data in the data column, and only executable sections are counted in the text column. Given that this is a brand new output format I've gone ahead and simplified things a little, while the berkeley format displays the total twice, once in decimal and once in hex, the new display format just displays the total in decimal. Of course, there's still the '--radix' option which can be used to display all the results in hexadecimal or octal. I've called the new format 'gnu', so '--format=gnu' or '-G' are used to access it. binutils/ChangeLog: * size.c (berkeley_format): Delete. (enum output_format): New enum. (selected_output_format): New variable. (usage): Update to mention GNU format. (main): Update to extract options, and select format as needed. Handle GNU format where needed. (berkeley_sum): Renamed to... (berkeley_or_gnu_sum): ...this, and updated to handle both formats. (berkeley_format): Renamed to... (berkeley_or_gnu_format): ...this, and updated to handle both formats. (print_sizes): Handle GNU format. * doc/binutils.texi (size): Document new GNU format. * testsuite/binutils-all/size.exp: Add test of extended functionality. * NEWS: Mention new functionality.
2019-01-24 15:27:27 +01:00
selected_output_format = FORMAT_BERKLEY;
1999-05-03 09:29:11 +02:00
break;
case 'S':
case 's':
binutils: Add new GNU format mode to `size` utility The size tool currently defaults to berkeley format output. However, this output format has a weird quirk, read-only data is counted against the text sections, not the data sections. The code offers no real explanation for why this is, but I'm reluctant to change it for two reasons, first, I'm assuming it probably makes sense in some case that I'm not thinking of (maybe a target where sections are not marked executable, and so there's no distinction between read-only data and code), and second, the code has been this way for at least 20 years, I worry that changing things now might cause more confusion than it solves. This commit then introduces a new output format for the size tool, this new format displays the results in a similar manor to the berkeley format, but counts read-only data in the data column, and only executable sections are counted in the text column. Given that this is a brand new output format I've gone ahead and simplified things a little, while the berkeley format displays the total twice, once in decimal and once in hex, the new display format just displays the total in decimal. Of course, there's still the '--radix' option which can be used to display all the results in hexadecimal or octal. I've called the new format 'gnu', so '--format=gnu' or '-G' are used to access it. binutils/ChangeLog: * size.c (berkeley_format): Delete. (enum output_format): New enum. (selected_output_format): New variable. (usage): Update to mention GNU format. (main): Update to extract options, and select format as needed. Handle GNU format where needed. (berkeley_sum): Renamed to... (berkeley_or_gnu_sum): ...this, and updated to handle both formats. (berkeley_format): Renamed to... (berkeley_or_gnu_format): ...this, and updated to handle both formats. (print_sizes): Handle GNU format. * doc/binutils.texi (size): Document new GNU format. * testsuite/binutils-all/size.exp: Add test of extended functionality. * NEWS: Mention new functionality.
2019-01-24 15:27:27 +01:00
selected_output_format = FORMAT_SYSV;
break;
case 'G':
case 'g':
selected_output_format = FORMAT_GNU;
1999-05-03 09:29:11 +02:00
break;
default:
non_fatal (_("invalid argument to --format: %s"), optarg);
1999-05-03 09:29:11 +02:00
usage (stderr, 1);
}
break;
2007-07-05 11:01:28 +02:00
case OPTION_TARGET:
1999-05-03 09:29:11 +02:00
target = optarg;
break;
2007-07-05 11:01:28 +02:00
case OPTION_RADIX:
1999-05-03 09:29:11 +02:00
#ifdef ANSI_LIBRARIES
temp = strtol (optarg, NULL, 10);
#else
temp = atol (optarg);
#endif
switch (temp)
{
case 10:
radix = decimal;
break;
case 8:
radix = octal;
break;
case 16:
radix = hex;
break;
default:
non_fatal (_("Invalid radix: %s\n"), optarg);
1999-05-03 09:29:11 +02:00
usage (stderr, 1);
}
break;
case 'A':
binutils: Add new GNU format mode to `size` utility The size tool currently defaults to berkeley format output. However, this output format has a weird quirk, read-only data is counted against the text sections, not the data sections. The code offers no real explanation for why this is, but I'm reluctant to change it for two reasons, first, I'm assuming it probably makes sense in some case that I'm not thinking of (maybe a target where sections are not marked executable, and so there's no distinction between read-only data and code), and second, the code has been this way for at least 20 years, I worry that changing things now might cause more confusion than it solves. This commit then introduces a new output format for the size tool, this new format displays the results in a similar manor to the berkeley format, but counts read-only data in the data column, and only executable sections are counted in the text column. Given that this is a brand new output format I've gone ahead and simplified things a little, while the berkeley format displays the total twice, once in decimal and once in hex, the new display format just displays the total in decimal. Of course, there's still the '--radix' option which can be used to display all the results in hexadecimal or octal. I've called the new format 'gnu', so '--format=gnu' or '-G' are used to access it. binutils/ChangeLog: * size.c (berkeley_format): Delete. (enum output_format): New enum. (selected_output_format): New variable. (usage): Update to mention GNU format. (main): Update to extract options, and select format as needed. Handle GNU format where needed. (berkeley_sum): Renamed to... (berkeley_or_gnu_sum): ...this, and updated to handle both formats. (berkeley_format): Renamed to... (berkeley_or_gnu_format): ...this, and updated to handle both formats. (print_sizes): Handle GNU format. * doc/binutils.texi (size): Document new GNU format. * testsuite/binutils-all/size.exp: Add test of extended functionality. * NEWS: Mention new functionality.
2019-01-24 15:27:27 +01:00
selected_output_format = FORMAT_SYSV;
1999-05-03 09:29:11 +02:00
break;
case 'B':
binutils: Add new GNU format mode to `size` utility The size tool currently defaults to berkeley format output. However, this output format has a weird quirk, read-only data is counted against the text sections, not the data sections. The code offers no real explanation for why this is, but I'm reluctant to change it for two reasons, first, I'm assuming it probably makes sense in some case that I'm not thinking of (maybe a target where sections are not marked executable, and so there's no distinction between read-only data and code), and second, the code has been this way for at least 20 years, I worry that changing things now might cause more confusion than it solves. This commit then introduces a new output format for the size tool, this new format displays the results in a similar manor to the berkeley format, but counts read-only data in the data column, and only executable sections are counted in the text column. Given that this is a brand new output format I've gone ahead and simplified things a little, while the berkeley format displays the total twice, once in decimal and once in hex, the new display format just displays the total in decimal. Of course, there's still the '--radix' option which can be used to display all the results in hexadecimal or octal. I've called the new format 'gnu', so '--format=gnu' or '-G' are used to access it. binutils/ChangeLog: * size.c (berkeley_format): Delete. (enum output_format): New enum. (selected_output_format): New variable. (usage): Update to mention GNU format. (main): Update to extract options, and select format as needed. Handle GNU format where needed. (berkeley_sum): Renamed to... (berkeley_or_gnu_sum): ...this, and updated to handle both formats. (berkeley_format): Renamed to... (berkeley_or_gnu_format): ...this, and updated to handle both formats. (print_sizes): Handle GNU format. * doc/binutils.texi (size): Document new GNU format. * testsuite/binutils-all/size.exp: Add test of extended functionality. * NEWS: Mention new functionality.
2019-01-24 15:27:27 +01:00
selected_output_format = FORMAT_BERKLEY;
break;
case 'G':
selected_output_format = FORMAT_GNU;
1999-05-03 09:29:11 +02:00
break;
2002-01-23 17:12:56 +01:00
case 'v':
1999-05-03 09:29:11 +02:00
case 'V':
show_version = 1;
break;
case 'd':
radix = decimal;
break;
case 'x':
radix = hex;
break;
case 'o':
radix = octal;
break;
2002-01-25 16:37:04 +01:00
case 't':
show_totals = 1;
break;
case 'f': /* FIXME : For sysv68, `-f' means `full format', i.e.
`[fname:] M(.text) + N(.data) + O(.bss) + P(.comment) = Q'
where `fname: ' appears only if there are >= 2 input files,
and M, N, O, P, Q are expressed in decimal by default,
hexa or octal if requested by `-x' or `-o'.
Just to make things interesting, Solaris also accepts -f,
which prints out the size of each allocatable section, the
name of the section, and the total of the section sizes. */
/* For the moment, accept `-f' silently, and ignore it. */
break;
1999-05-03 09:29:11 +02:00
case 0:
break;
2002-01-23 17:12:56 +01:00
case 'h':
case 'H':
1999-05-03 09:29:11 +02:00
case '?':
usage (stderr, 1);
}
if (show_version)
print_version ("size");
if (show_help)
usage (stdout, 0);
if (optind == argc)
display_file ("a.out");
else
for (; optind < argc;)
display_file (argv[optind++]);
binutils: Add new GNU format mode to `size` utility The size tool currently defaults to berkeley format output. However, this output format has a weird quirk, read-only data is counted against the text sections, not the data sections. The code offers no real explanation for why this is, but I'm reluctant to change it for two reasons, first, I'm assuming it probably makes sense in some case that I'm not thinking of (maybe a target where sections are not marked executable, and so there's no distinction between read-only data and code), and second, the code has been this way for at least 20 years, I worry that changing things now might cause more confusion than it solves. This commit then introduces a new output format for the size tool, this new format displays the results in a similar manor to the berkeley format, but counts read-only data in the data column, and only executable sections are counted in the text column. Given that this is a brand new output format I've gone ahead and simplified things a little, while the berkeley format displays the total twice, once in decimal and once in hex, the new display format just displays the total in decimal. Of course, there's still the '--radix' option which can be used to display all the results in hexadecimal or octal. I've called the new format 'gnu', so '--format=gnu' or '-G' are used to access it. binutils/ChangeLog: * size.c (berkeley_format): Delete. (enum output_format): New enum. (selected_output_format): New variable. (usage): Update to mention GNU format. (main): Update to extract options, and select format as needed. Handle GNU format where needed. (berkeley_sum): Renamed to... (berkeley_or_gnu_sum): ...this, and updated to handle both formats. (berkeley_format): Renamed to... (berkeley_or_gnu_format): ...this, and updated to handle both formats. (print_sizes): Handle GNU format. * doc/binutils.texi (size): Document new GNU format. * testsuite/binutils-all/size.exp: Add test of extended functionality. * NEWS: Mention new functionality.
2019-01-24 15:27:27 +01:00
if (show_totals && (selected_output_format == FORMAT_BERKLEY
|| selected_output_format == FORMAT_GNU))
2002-01-25 16:37:04 +01:00
{
bfd_size_type total = total_textsize + total_datasize + total_bsssize;
binutils: Add new GNU format mode to `size` utility The size tool currently defaults to berkeley format output. However, this output format has a weird quirk, read-only data is counted against the text sections, not the data sections. The code offers no real explanation for why this is, but I'm reluctant to change it for two reasons, first, I'm assuming it probably makes sense in some case that I'm not thinking of (maybe a target where sections are not marked executable, and so there's no distinction between read-only data and code), and second, the code has been this way for at least 20 years, I worry that changing things now might cause more confusion than it solves. This commit then introduces a new output format for the size tool, this new format displays the results in a similar manor to the berkeley format, but counts read-only data in the data column, and only executable sections are counted in the text column. Given that this is a brand new output format I've gone ahead and simplified things a little, while the berkeley format displays the total twice, once in decimal and once in hex, the new display format just displays the total in decimal. Of course, there's still the '--radix' option which can be used to display all the results in hexadecimal or octal. I've called the new format 'gnu', so '--format=gnu' or '-G' are used to access it. binutils/ChangeLog: * size.c (berkeley_format): Delete. (enum output_format): New enum. (selected_output_format): New variable. (usage): Update to mention GNU format. (main): Update to extract options, and select format as needed. Handle GNU format where needed. (berkeley_sum): Renamed to... (berkeley_or_gnu_sum): ...this, and updated to handle both formats. (berkeley_format): Renamed to... (berkeley_or_gnu_format): ...this, and updated to handle both formats. (print_sizes): Handle GNU format. * doc/binutils.texi (size): Document new GNU format. * testsuite/binutils-all/size.exp: Add test of extended functionality. * NEWS: Mention new functionality.
2019-01-24 15:27:27 +01:00
int col_width = (selected_output_format == FORMAT_BERKLEY) ? 7 : 10;
char sep_char = (selected_output_format == FORMAT_BERKLEY) ? '\t' : ' ';
rprint_number (col_width, total_textsize);
putchar(sep_char);
rprint_number (col_width, total_datasize);
putchar(sep_char);
rprint_number (col_width, total_bsssize);
putchar(sep_char);
if (selected_output_format == FORMAT_BERKLEY)
printf (((radix == octal) ? "%7lo\t%7lx" : "%7lu\t%7lx"),
(unsigned long) total, (unsigned long) total);
else
rprint_number (col_width, total);
putchar(sep_char);
2002-01-25 16:37:04 +01:00
fputs ("(TOTALS)\n", stdout);
}
1999-05-03 09:29:11 +02:00
return return_code;
}
2007-07-05 11:01:28 +02:00
/* Total size required for common symbols in ABFD. */
static void
calculate_common_size (bfd *abfd)
{
asymbol **syms = NULL;
long storage, symcount;
common_size = 0;
if ((bfd_get_file_flags (abfd) & (EXEC_P | DYNAMIC | HAS_SYMS)) != HAS_SYMS)
return;
storage = bfd_get_symtab_upper_bound (abfd);
if (storage < 0)
bfd_fatal (bfd_get_filename (abfd));
if (storage)
Updated soruces in binutils/* to compile cleanly with -Wc++-compat. * binutils/addr2line.c (slurp_symtab): Fix casts. Introduce variable minisyms to avoid aliasing varning. * binutils/ar.c: Add casts. (normalize): Use name del instead of delete. (display_target_list,display_info_table): Change loop counter variable a to int. * binutils/bucomm.c: Add casts. * binutils/debug.c: Update function to use new names. (struct debug_baseclass): Rename member from virtual to is_virtual. (struct debug_type_s,struct debug_field_s,struct debug_baseclass_s,struct debug_method_s,struct debug_method_variant_s,struct debug_type_s): Rename struct from avoid name collision. * /binutils/debug.h: Use new struct names. * binutils/dwarf.c: Add casts. (free_debug_memory): Change loop counter variable a to int. * binutils/ieee.c: Add casts. (enum ieee_var_kind): Move to top level. (ieee_class_baseclass): Rename parameter virtual to is_virtual. (ieee_class_method_var): Rename variable virtual to is_virtual. * binutils/nm.c: Add casts. * binutils/objcopy.c: Add casts. (copy_archive): Rename variable delete to del. * binutils/objdump.c: Add casts. (dump_dwarf_section): Change loop counter variable i to int. * binutils/prdbg.c: Add casts. (pr_class_baseclass,tg_class_baseclass): Rename parameters virtual to is_virtual. * binutils/readelf.c: Add casts. (struct ia64_unw_table_entry,struct hppa_unw_table_entry): Move to top level. * binutils/size.c: Add casts. * binutils/stabs.c (parse_stab_type, parse_stab_range_type) (parse_stab_cpp_abbrev): Rename parameter from typename to type_name. (parse_stab_baseclasses): Rename variable virtual to is_virtual. * binutils/strings.c: Add casts. * binutils/wrstabs.c (stab_class_baseclass): Rename parameter virtual to is_virtual.
2009-09-10 15:40:44 +02:00
syms = (asymbol **) xmalloc (storage);
2007-07-05 11:01:28 +02:00
symcount = bfd_canonicalize_symtab (abfd, syms);
if (symcount < 0)
bfd_fatal (bfd_get_filename (abfd));
while (--symcount >= 0)
{
asymbol *sym = syms[symcount];
if (bfd_is_com_section (sym->section)
&& (sym->flags & BSF_SECTION_SYM) == 0)
common_size += sym->value;
}
free (syms);
}
1999-05-03 09:29:11 +02:00
/* Display stats on file or archive member ABFD. */
static void
display_bfd (bfd *abfd)
1999-05-03 09:29:11 +02:00
{
char **matching;
if (bfd_check_format (abfd, bfd_archive))
/* An archive within an archive. */
return;
if (bfd_check_format_matches (abfd, bfd_object, &matching))
{
print_sizes (abfd);
printf ("\n");
return;
}
if (bfd_get_error () == bfd_error_file_ambiguously_recognized)
{
bfd_nonfatal (bfd_get_filename (abfd));
list_matching_formats (matching);
free (matching);
return_code = 3;
return;
}
if (bfd_check_format_matches (abfd, bfd_core, &matching))
{
2002-01-25 16:37:04 +01:00
const char *core_cmd;
1999-05-03 09:29:11 +02:00
print_sizes (abfd);
fputs (" (core file", stdout);
core_cmd = bfd_core_file_failing_command (abfd);
if (core_cmd)
printf (" invoked as %s", core_cmd);
puts (")\n");
return;
}
bfd_nonfatal (bfd_get_filename (abfd));
if (bfd_get_error () == bfd_error_file_ambiguously_recognized)
{
list_matching_formats (matching);
free (matching);
}
return_code = 3;
}
static void
display_archive (bfd *file)
1999-05-03 09:29:11 +02:00
{
bfd *arfile = (bfd *) NULL;
bfd *last_arfile = (bfd *) NULL;
1999-05-03 09:29:11 +02:00
for (;;)
{
bfd_set_error (bfd_error_no_error);
arfile = bfd_openr_next_archived_file (file, arfile);
if (arfile == NULL)
{
if (bfd_get_error () != bfd_error_no_more_archived_files)
{
bfd_nonfatal (bfd_get_filename (file));
return_code = 2;
}
break;
}
display_bfd (arfile);
if (last_arfile != NULL)
{
bfd_close (last_arfile);
/* PR 17512: file: a244edbc. */
if (last_arfile == arfile)
return;
}
last_arfile = arfile;
1999-05-03 09:29:11 +02:00
}
if (last_arfile != NULL)
bfd_close (last_arfile);
1999-05-03 09:29:11 +02:00
}
static void
display_file (char *filename)
1999-05-03 09:29:11 +02:00
{
bfd *file;
2002-01-25 16:37:04 +01:00
if (get_file_size (filename) < 1)
{
return_code = 1;
return;
}
file = bfd_openr (filename, target);
1999-05-03 09:29:11 +02:00
if (file == NULL)
{
bfd_nonfatal (filename);
return_code = 1;
return;
}
if (bfd_check_format (file, bfd_archive))
1999-05-03 09:29:11 +02:00
display_archive (file);
else
display_bfd (file);
if (!bfd_close (file))
1999-05-03 09:29:11 +02:00
{
bfd_nonfatal (filename);
return_code = 1;
return;
}
}
static int
size_number (bfd_size_type num)
1999-05-03 09:29:11 +02:00
{
char buffer[40];
2002-01-25 16:37:04 +01:00
1999-05-03 09:29:11 +02:00
sprintf (buffer,
(radix == decimal ? "%" BFD_VMA_FMT "u" :
((radix == octal) ? "0%" BFD_VMA_FMT "o" : "0x%" BFD_VMA_FMT "x")),
num);
1999-05-03 09:29:11 +02:00
return strlen (buffer);
}
static void
rprint_number (int width, bfd_size_type num)
1999-05-03 09:29:11 +02:00
{
char buffer[40];
2002-01-25 16:37:04 +01:00
1999-05-03 09:29:11 +02:00
sprintf (buffer,
(radix == decimal ? "%" BFD_VMA_FMT "u" :
((radix == octal) ? "0%" BFD_VMA_FMT "o" : "0x%" BFD_VMA_FMT "x")),
num);
1999-05-03 09:29:11 +02:00
printf ("%*s", width, buffer);
}
static bfd_size_type bsssize;
static bfd_size_type datasize;
static bfd_size_type textsize;
static void
binutils: Add new GNU format mode to `size` utility The size tool currently defaults to berkeley format output. However, this output format has a weird quirk, read-only data is counted against the text sections, not the data sections. The code offers no real explanation for why this is, but I'm reluctant to change it for two reasons, first, I'm assuming it probably makes sense in some case that I'm not thinking of (maybe a target where sections are not marked executable, and so there's no distinction between read-only data and code), and second, the code has been this way for at least 20 years, I worry that changing things now might cause more confusion than it solves. This commit then introduces a new output format for the size tool, this new format displays the results in a similar manor to the berkeley format, but counts read-only data in the data column, and only executable sections are counted in the text column. Given that this is a brand new output format I've gone ahead and simplified things a little, while the berkeley format displays the total twice, once in decimal and once in hex, the new display format just displays the total in decimal. Of course, there's still the '--radix' option which can be used to display all the results in hexadecimal or octal. I've called the new format 'gnu', so '--format=gnu' or '-G' are used to access it. binutils/ChangeLog: * size.c (berkeley_format): Delete. (enum output_format): New enum. (selected_output_format): New variable. (usage): Update to mention GNU format. (main): Update to extract options, and select format as needed. Handle GNU format where needed. (berkeley_sum): Renamed to... (berkeley_or_gnu_sum): ...this, and updated to handle both formats. (berkeley_format): Renamed to... (berkeley_or_gnu_format): ...this, and updated to handle both formats. (print_sizes): Handle GNU format. * doc/binutils.texi (size): Document new GNU format. * testsuite/binutils-all/size.exp: Add test of extended functionality. * NEWS: Mention new functionality.
2019-01-24 15:27:27 +01:00
berkeley_or_gnu_sum (bfd *abfd ATTRIBUTE_UNUSED, sec_ptr sec,
void *ignore ATTRIBUTE_UNUSED)
1999-05-03 09:29:11 +02:00
{
flagword flags;
bfd_size_type size;
flags = bfd_get_section_flags (abfd, sec);
if ((flags & SEC_ALLOC) == 0)
return;
size = bfd_get_section_size (sec);
binutils: Add new GNU format mode to `size` utility The size tool currently defaults to berkeley format output. However, this output format has a weird quirk, read-only data is counted against the text sections, not the data sections. The code offers no real explanation for why this is, but I'm reluctant to change it for two reasons, first, I'm assuming it probably makes sense in some case that I'm not thinking of (maybe a target where sections are not marked executable, and so there's no distinction between read-only data and code), and second, the code has been this way for at least 20 years, I worry that changing things now might cause more confusion than it solves. This commit then introduces a new output format for the size tool, this new format displays the results in a similar manor to the berkeley format, but counts read-only data in the data column, and only executable sections are counted in the text column. Given that this is a brand new output format I've gone ahead and simplified things a little, while the berkeley format displays the total twice, once in decimal and once in hex, the new display format just displays the total in decimal. Of course, there's still the '--radix' option which can be used to display all the results in hexadecimal or octal. I've called the new format 'gnu', so '--format=gnu' or '-G' are used to access it. binutils/ChangeLog: * size.c (berkeley_format): Delete. (enum output_format): New enum. (selected_output_format): New variable. (usage): Update to mention GNU format. (main): Update to extract options, and select format as needed. Handle GNU format where needed. (berkeley_sum): Renamed to... (berkeley_or_gnu_sum): ...this, and updated to handle both formats. (berkeley_format): Renamed to... (berkeley_or_gnu_format): ...this, and updated to handle both formats. (print_sizes): Handle GNU format. * doc/binutils.texi (size): Document new GNU format. * testsuite/binutils-all/size.exp: Add test of extended functionality. * NEWS: Mention new functionality.
2019-01-24 15:27:27 +01:00
if ((flags & SEC_CODE) != 0
|| (selected_output_format == FORMAT_BERKLEY
&& (flags & SEC_READONLY) != 0))
1999-05-03 09:29:11 +02:00
textsize += size;
else if ((flags & SEC_HAS_CONTENTS) != 0)
datasize += size;
else
bsssize += size;
}
static void
binutils: Add new GNU format mode to `size` utility The size tool currently defaults to berkeley format output. However, this output format has a weird quirk, read-only data is counted against the text sections, not the data sections. The code offers no real explanation for why this is, but I'm reluctant to change it for two reasons, first, I'm assuming it probably makes sense in some case that I'm not thinking of (maybe a target where sections are not marked executable, and so there's no distinction between read-only data and code), and second, the code has been this way for at least 20 years, I worry that changing things now might cause more confusion than it solves. This commit then introduces a new output format for the size tool, this new format displays the results in a similar manor to the berkeley format, but counts read-only data in the data column, and only executable sections are counted in the text column. Given that this is a brand new output format I've gone ahead and simplified things a little, while the berkeley format displays the total twice, once in decimal and once in hex, the new display format just displays the total in decimal. Of course, there's still the '--radix' option which can be used to display all the results in hexadecimal or octal. I've called the new format 'gnu', so '--format=gnu' or '-G' are used to access it. binutils/ChangeLog: * size.c (berkeley_format): Delete. (enum output_format): New enum. (selected_output_format): New variable. (usage): Update to mention GNU format. (main): Update to extract options, and select format as needed. Handle GNU format where needed. (berkeley_sum): Renamed to... (berkeley_or_gnu_sum): ...this, and updated to handle both formats. (berkeley_format): Renamed to... (berkeley_or_gnu_format): ...this, and updated to handle both formats. (print_sizes): Handle GNU format. * doc/binutils.texi (size): Document new GNU format. * testsuite/binutils-all/size.exp: Add test of extended functionality. * NEWS: Mention new functionality.
2019-01-24 15:27:27 +01:00
print_berkeley_or_gnu_format (bfd *abfd)
1999-05-03 09:29:11 +02:00
{
static int files_seen = 0;
bfd_size_type total;
binutils: Add new GNU format mode to `size` utility The size tool currently defaults to berkeley format output. However, this output format has a weird quirk, read-only data is counted against the text sections, not the data sections. The code offers no real explanation for why this is, but I'm reluctant to change it for two reasons, first, I'm assuming it probably makes sense in some case that I'm not thinking of (maybe a target where sections are not marked executable, and so there's no distinction between read-only data and code), and second, the code has been this way for at least 20 years, I worry that changing things now might cause more confusion than it solves. This commit then introduces a new output format for the size tool, this new format displays the results in a similar manor to the berkeley format, but counts read-only data in the data column, and only executable sections are counted in the text column. Given that this is a brand new output format I've gone ahead and simplified things a little, while the berkeley format displays the total twice, once in decimal and once in hex, the new display format just displays the total in decimal. Of course, there's still the '--radix' option which can be used to display all the results in hexadecimal or octal. I've called the new format 'gnu', so '--format=gnu' or '-G' are used to access it. binutils/ChangeLog: * size.c (berkeley_format): Delete. (enum output_format): New enum. (selected_output_format): New variable. (usage): Update to mention GNU format. (main): Update to extract options, and select format as needed. Handle GNU format where needed. (berkeley_sum): Renamed to... (berkeley_or_gnu_sum): ...this, and updated to handle both formats. (berkeley_format): Renamed to... (berkeley_or_gnu_format): ...this, and updated to handle both formats. (print_sizes): Handle GNU format. * doc/binutils.texi (size): Document new GNU format. * testsuite/binutils-all/size.exp: Add test of extended functionality. * NEWS: Mention new functionality.
2019-01-24 15:27:27 +01:00
int col_width = (selected_output_format == FORMAT_BERKLEY) ? 7 : 10;
char sep_char = (selected_output_format == FORMAT_BERKLEY) ? '\t' : ' ';
1999-05-03 09:29:11 +02:00
bsssize = 0;
datasize = 0;
textsize = 0;
binutils: Add new GNU format mode to `size` utility The size tool currently defaults to berkeley format output. However, this output format has a weird quirk, read-only data is counted against the text sections, not the data sections. The code offers no real explanation for why this is, but I'm reluctant to change it for two reasons, first, I'm assuming it probably makes sense in some case that I'm not thinking of (maybe a target where sections are not marked executable, and so there's no distinction between read-only data and code), and second, the code has been this way for at least 20 years, I worry that changing things now might cause more confusion than it solves. This commit then introduces a new output format for the size tool, this new format displays the results in a similar manor to the berkeley format, but counts read-only data in the data column, and only executable sections are counted in the text column. Given that this is a brand new output format I've gone ahead and simplified things a little, while the berkeley format displays the total twice, once in decimal and once in hex, the new display format just displays the total in decimal. Of course, there's still the '--radix' option which can be used to display all the results in hexadecimal or octal. I've called the new format 'gnu', so '--format=gnu' or '-G' are used to access it. binutils/ChangeLog: * size.c (berkeley_format): Delete. (enum output_format): New enum. (selected_output_format): New variable. (usage): Update to mention GNU format. (main): Update to extract options, and select format as needed. Handle GNU format where needed. (berkeley_sum): Renamed to... (berkeley_or_gnu_sum): ...this, and updated to handle both formats. (berkeley_format): Renamed to... (berkeley_or_gnu_format): ...this, and updated to handle both formats. (print_sizes): Handle GNU format. * doc/binutils.texi (size): Document new GNU format. * testsuite/binutils-all/size.exp: Add test of extended functionality. * NEWS: Mention new functionality.
2019-01-24 15:27:27 +01:00
bfd_map_over_sections (abfd, berkeley_or_gnu_sum, NULL);
1999-05-03 09:29:11 +02:00
2007-07-05 11:01:28 +02:00
bsssize += common_size;
1999-05-03 09:29:11 +02:00
if (files_seen++ == 0)
binutils: Add new GNU format mode to `size` utility The size tool currently defaults to berkeley format output. However, this output format has a weird quirk, read-only data is counted against the text sections, not the data sections. The code offers no real explanation for why this is, but I'm reluctant to change it for two reasons, first, I'm assuming it probably makes sense in some case that I'm not thinking of (maybe a target where sections are not marked executable, and so there's no distinction between read-only data and code), and second, the code has been this way for at least 20 years, I worry that changing things now might cause more confusion than it solves. This commit then introduces a new output format for the size tool, this new format displays the results in a similar manor to the berkeley format, but counts read-only data in the data column, and only executable sections are counted in the text column. Given that this is a brand new output format I've gone ahead and simplified things a little, while the berkeley format displays the total twice, once in decimal and once in hex, the new display format just displays the total in decimal. Of course, there's still the '--radix' option which can be used to display all the results in hexadecimal or octal. I've called the new format 'gnu', so '--format=gnu' or '-G' are used to access it. binutils/ChangeLog: * size.c (berkeley_format): Delete. (enum output_format): New enum. (selected_output_format): New variable. (usage): Update to mention GNU format. (main): Update to extract options, and select format as needed. Handle GNU format where needed. (berkeley_sum): Renamed to... (berkeley_or_gnu_sum): ...this, and updated to handle both formats. (berkeley_format): Renamed to... (berkeley_or_gnu_format): ...this, and updated to handle both formats. (print_sizes): Handle GNU format. * doc/binutils.texi (size): Document new GNU format. * testsuite/binutils-all/size.exp: Add test of extended functionality. * NEWS: Mention new functionality.
2019-01-24 15:27:27 +01:00
{
if (selected_output_format == FORMAT_BERKLEY)
puts ((radix == octal) ? " text\t data\t bss\t oct\t hex\tfilename" :
" text\t data\t bss\t dec\t hex\tfilename");
else
puts (" text data bss total filename");
}
1999-05-03 09:29:11 +02:00
total = textsize + datasize + bsssize;
2002-01-25 16:37:04 +01:00
if (show_totals)
{
total_textsize += textsize;
total_datasize += datasize;
total_bsssize += bsssize;
}
binutils: Add new GNU format mode to `size` utility The size tool currently defaults to berkeley format output. However, this output format has a weird quirk, read-only data is counted against the text sections, not the data sections. The code offers no real explanation for why this is, but I'm reluctant to change it for two reasons, first, I'm assuming it probably makes sense in some case that I'm not thinking of (maybe a target where sections are not marked executable, and so there's no distinction between read-only data and code), and second, the code has been this way for at least 20 years, I worry that changing things now might cause more confusion than it solves. This commit then introduces a new output format for the size tool, this new format displays the results in a similar manor to the berkeley format, but counts read-only data in the data column, and only executable sections are counted in the text column. Given that this is a brand new output format I've gone ahead and simplified things a little, while the berkeley format displays the total twice, once in decimal and once in hex, the new display format just displays the total in decimal. Of course, there's still the '--radix' option which can be used to display all the results in hexadecimal or octal. I've called the new format 'gnu', so '--format=gnu' or '-G' are used to access it. binutils/ChangeLog: * size.c (berkeley_format): Delete. (enum output_format): New enum. (selected_output_format): New variable. (usage): Update to mention GNU format. (main): Update to extract options, and select format as needed. Handle GNU format where needed. (berkeley_sum): Renamed to... (berkeley_or_gnu_sum): ...this, and updated to handle both formats. (berkeley_format): Renamed to... (berkeley_or_gnu_format): ...this, and updated to handle both formats. (print_sizes): Handle GNU format. * doc/binutils.texi (size): Document new GNU format. * testsuite/binutils-all/size.exp: Add test of extended functionality. * NEWS: Mention new functionality.
2019-01-24 15:27:27 +01:00
rprint_number (col_width, textsize);
putchar (sep_char);
rprint_number (col_width, datasize);
putchar (sep_char);
rprint_number (col_width, bsssize);
putchar (sep_char);
1999-05-03 09:29:11 +02:00
binutils: Add new GNU format mode to `size` utility The size tool currently defaults to berkeley format output. However, this output format has a weird quirk, read-only data is counted against the text sections, not the data sections. The code offers no real explanation for why this is, but I'm reluctant to change it for two reasons, first, I'm assuming it probably makes sense in some case that I'm not thinking of (maybe a target where sections are not marked executable, and so there's no distinction between read-only data and code), and second, the code has been this way for at least 20 years, I worry that changing things now might cause more confusion than it solves. This commit then introduces a new output format for the size tool, this new format displays the results in a similar manor to the berkeley format, but counts read-only data in the data column, and only executable sections are counted in the text column. Given that this is a brand new output format I've gone ahead and simplified things a little, while the berkeley format displays the total twice, once in decimal and once in hex, the new display format just displays the total in decimal. Of course, there's still the '--radix' option which can be used to display all the results in hexadecimal or octal. I've called the new format 'gnu', so '--format=gnu' or '-G' are used to access it. binutils/ChangeLog: * size.c (berkeley_format): Delete. (enum output_format): New enum. (selected_output_format): New variable. (usage): Update to mention GNU format. (main): Update to extract options, and select format as needed. Handle GNU format where needed. (berkeley_sum): Renamed to... (berkeley_or_gnu_sum): ...this, and updated to handle both formats. (berkeley_format): Renamed to... (berkeley_or_gnu_format): ...this, and updated to handle both formats. (print_sizes): Handle GNU format. * doc/binutils.texi (size): Document new GNU format. * testsuite/binutils-all/size.exp: Add test of extended functionality. * NEWS: Mention new functionality.
2019-01-24 15:27:27 +01:00
if (selected_output_format == FORMAT_BERKLEY)
printf (((radix == octal) ? "%7lo\t%7lx" : "%7lu\t%7lx"),
(unsigned long) total, (unsigned long) total);
else
rprint_number (col_width, total);
putchar (sep_char);
1999-05-03 09:29:11 +02:00
fputs (bfd_get_filename (abfd), stdout);
2002-01-25 16:37:04 +01:00
if (abfd->my_archive)
printf (" (ex %s)", bfd_get_filename (abfd->my_archive));
1999-05-03 09:29:11 +02:00
}
/* I REALLY miss lexical functions! */
bfd_size_type svi_total = 0;
bfd_vma svi_maxvma = 0;
int svi_namelen = 0;
int svi_vmalen = 0;
int svi_sizelen = 0;
static void
sysv_internal_sizer (bfd *file ATTRIBUTE_UNUSED, sec_ptr sec,
void *ignore ATTRIBUTE_UNUSED)
1999-05-03 09:29:11 +02:00
{
bfd_size_type size = bfd_section_size (file, sec);
2002-01-25 16:37:04 +01:00
if ( ! bfd_is_abs_section (sec)
&& ! bfd_is_com_section (sec)
&& ! bfd_is_und_section (sec))
1999-05-03 09:29:11 +02:00
{
int namelen = strlen (bfd_section_name (file, sec));
2002-01-25 16:37:04 +01:00
1999-05-03 09:29:11 +02:00
if (namelen > svi_namelen)
svi_namelen = namelen;
svi_total += size;
2002-01-25 16:37:04 +01:00
1999-05-03 09:29:11 +02:00
if (bfd_section_vma (file, sec) > svi_maxvma)
svi_maxvma = bfd_section_vma (file, sec);
}
}
2007-07-05 11:01:28 +02:00
static void
sysv_one_line (const char *name, bfd_size_type size, bfd_vma vma)
{
printf ("%-*s ", svi_namelen, name);
rprint_number (svi_sizelen, size);
printf (" ");
rprint_number (svi_vmalen, vma);
printf ("\n");
}
1999-05-03 09:29:11 +02:00
static void
sysv_internal_printer (bfd *file ATTRIBUTE_UNUSED, sec_ptr sec,
void *ignore ATTRIBUTE_UNUSED)
1999-05-03 09:29:11 +02:00
{
bfd_size_type size = bfd_section_size (file, sec);
2002-01-25 16:37:04 +01:00
if ( ! bfd_is_abs_section (sec)
&& ! bfd_is_com_section (sec)
&& ! bfd_is_und_section (sec))
1999-05-03 09:29:11 +02:00
{
svi_total += size;
2007-07-05 11:01:28 +02:00
sysv_one_line (bfd_section_name (file, sec),
size,
bfd_section_vma (file, sec));
1999-05-03 09:29:11 +02:00
}
}
static void
print_sysv_format (bfd *file)
1999-05-03 09:29:11 +02:00
{
2002-01-25 16:37:04 +01:00
/* Size all of the columns. */
1999-05-03 09:29:11 +02:00
svi_total = 0;
svi_maxvma = 0;
svi_namelen = 0;
bfd_map_over_sections (file, sysv_internal_sizer, NULL);
2007-07-05 11:01:28 +02:00
if (show_common)
{
if (svi_namelen < (int) sizeof ("*COM*") - 1)
svi_namelen = sizeof ("*COM*") - 1;
svi_total += common_size;
}
1999-05-03 09:29:11 +02:00
svi_vmalen = size_number ((bfd_size_type)svi_maxvma);
2002-01-25 16:37:04 +01:00
1999-05-03 09:29:11 +02:00
if ((size_t) svi_vmalen < sizeof ("addr") - 1)
svi_vmalen = sizeof ("addr")-1;
svi_sizelen = size_number (svi_total);
if ((size_t) svi_sizelen < sizeof ("size") - 1)
svi_sizelen = sizeof ("size")-1;
svi_total = 0;
printf ("%s ", bfd_get_filename (file));
2002-01-25 16:37:04 +01:00
if (file->my_archive)
printf (" (ex %s)", bfd_get_filename (file->my_archive));
1999-05-03 09:29:11 +02:00
printf (":\n%-*s %*s %*s\n", svi_namelen, "section",
svi_sizelen, "size", svi_vmalen, "addr");
2002-01-25 16:37:04 +01:00
bfd_map_over_sections (file, sysv_internal_printer, NULL);
2007-07-05 11:01:28 +02:00
if (show_common)
{
svi_total += common_size;
sysv_one_line ("*COM*", common_size, 0);
}
1999-05-03 09:29:11 +02:00
printf ("%-*s ", svi_namelen, "Total");
rprint_number (svi_sizelen, svi_total);
printf ("\n\n");
}
static void
print_sizes (bfd *file)
1999-05-03 09:29:11 +02:00
{
2007-07-05 11:01:28 +02:00
if (show_common)
calculate_common_size (file);
binutils: Add new GNU format mode to `size` utility The size tool currently defaults to berkeley format output. However, this output format has a weird quirk, read-only data is counted against the text sections, not the data sections. The code offers no real explanation for why this is, but I'm reluctant to change it for two reasons, first, I'm assuming it probably makes sense in some case that I'm not thinking of (maybe a target where sections are not marked executable, and so there's no distinction between read-only data and code), and second, the code has been this way for at least 20 years, I worry that changing things now might cause more confusion than it solves. This commit then introduces a new output format for the size tool, this new format displays the results in a similar manor to the berkeley format, but counts read-only data in the data column, and only executable sections are counted in the text column. Given that this is a brand new output format I've gone ahead and simplified things a little, while the berkeley format displays the total twice, once in decimal and once in hex, the new display format just displays the total in decimal. Of course, there's still the '--radix' option which can be used to display all the results in hexadecimal or octal. I've called the new format 'gnu', so '--format=gnu' or '-G' are used to access it. binutils/ChangeLog: * size.c (berkeley_format): Delete. (enum output_format): New enum. (selected_output_format): New variable. (usage): Update to mention GNU format. (main): Update to extract options, and select format as needed. Handle GNU format where needed. (berkeley_sum): Renamed to... (berkeley_or_gnu_sum): ...this, and updated to handle both formats. (berkeley_format): Renamed to... (berkeley_or_gnu_format): ...this, and updated to handle both formats. (print_sizes): Handle GNU format. * doc/binutils.texi (size): Document new GNU format. * testsuite/binutils-all/size.exp: Add test of extended functionality. * NEWS: Mention new functionality.
2019-01-24 15:27:27 +01:00
if (selected_output_format == FORMAT_SYSV)
1999-05-03 09:29:11 +02:00
print_sysv_format (file);
binutils: Add new GNU format mode to `size` utility The size tool currently defaults to berkeley format output. However, this output format has a weird quirk, read-only data is counted against the text sections, not the data sections. The code offers no real explanation for why this is, but I'm reluctant to change it for two reasons, first, I'm assuming it probably makes sense in some case that I'm not thinking of (maybe a target where sections are not marked executable, and so there's no distinction between read-only data and code), and second, the code has been this way for at least 20 years, I worry that changing things now might cause more confusion than it solves. This commit then introduces a new output format for the size tool, this new format displays the results in a similar manor to the berkeley format, but counts read-only data in the data column, and only executable sections are counted in the text column. Given that this is a brand new output format I've gone ahead and simplified things a little, while the berkeley format displays the total twice, once in decimal and once in hex, the new display format just displays the total in decimal. Of course, there's still the '--radix' option which can be used to display all the results in hexadecimal or octal. I've called the new format 'gnu', so '--format=gnu' or '-G' are used to access it. binutils/ChangeLog: * size.c (berkeley_format): Delete. (enum output_format): New enum. (selected_output_format): New variable. (usage): Update to mention GNU format. (main): Update to extract options, and select format as needed. Handle GNU format where needed. (berkeley_sum): Renamed to... (berkeley_or_gnu_sum): ...this, and updated to handle both formats. (berkeley_format): Renamed to... (berkeley_or_gnu_format): ...this, and updated to handle both formats. (print_sizes): Handle GNU format. * doc/binutils.texi (size): Document new GNU format. * testsuite/binutils-all/size.exp: Add test of extended functionality. * NEWS: Mention new functionality.
2019-01-24 15:27:27 +01:00
else
print_berkeley_or_gnu_format (file);
1999-05-03 09:29:11 +02:00
}