2012-05-26 23:14:36 +00:00
|
|
|
|
/* Convert a 'struct tm' to a time_t value.
|
2016-01-01 09:12:52 +00:00
|
|
|
|
Copyright (C) 1993-2016 Free Software Foundation, Inc.
|
2011-01-09 06:57:07 +00:00
|
|
|
|
This file is part of the GNU C Library.
|
|
|
|
|
Contributed by Paul Eggert <eggert@twinsun.com>.
|
|
|
|
|
|
2012-05-26 23:14:36 +00:00
|
|
|
|
The GNU C Library 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.
|
2011-01-09 06:57:07 +00:00
|
|
|
|
|
2012-05-26 23:14:36 +00:00
|
|
|
|
The GNU C Library is distributed in the hope that it will be useful,
|
2011-01-09 06:57:07 +00:00
|
|
|
|
but WITHOUT ANY WARRANTY; without even the implied warranty of
|
2012-05-26 23:14:36 +00:00
|
|
|
|
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
|
|
|
|
|
General Public License for more details.
|
2011-01-09 06:57:07 +00:00
|
|
|
|
|
2012-05-26 23:14:36 +00:00
|
|
|
|
You should have received a copy of the GNU General Public
|
|
|
|
|
License along with the GNU C Library; if not, see
|
|
|
|
|
<http://www.gnu.org/licenses/>. */
|
2011-01-09 06:57:07 +00:00
|
|
|
|
|
|
|
|
|
/* Define this to have a standalone program to test this implementation of
|
|
|
|
|
mktime. */
|
Merge from gnulib
This incorporates:
2016-01-24 openat_proc_name: fix last '/' overwritten on OS/2 kLIBC
2016-01-24 closedir, dirfd, opendir: port to OpenSolaris 5.10
2016-01-15 detect utimes() correctly on OS/2 kLIBC
2016-01-15 openat_proc_name: port to OS/2 kLIBC
2016-01-14 stdint: check _INTPTR_T_DECLARED for intptr_t etc.
2016-01-14 opendir, closedir, dirfd, fdopendir: port to OS/2 kLIBC
2016-01-14 dup, dup2, fcntl: support a directory fd on OS/2 kLIBC
2016-01-14 binary-io: don't put fd in binary mode if a console on EMX
2016-01-14 sig2str: list all signals on FreeBSD >= 7
2016-01-13 acl-permissions: port to USE_ACL==0 platforms
2016-01-12 mktime: rename macro to avoid glibc clash
2016-01-12 Port "$@" to OpenIndiana ksh93
2016-01-12 Port Universal Time settings to strict POSIX
* build-aux/gitlog-to-changelog, build-aux/update-copyright:
* doc/misc/texinfo.tex, lib/acl-internal.c, lib/acl-internal.h:
* lib/binary-io.h, lib/dirent.in.h, lib/dirfd.c, lib/dup2.c:
* lib/fcntl.c, lib/fdopendir.c, lib/mktime.c, lib/openat-proc.c:
* lib/sig2str.h, lib/stdint.in.h, m4/dirfd.m4, m4/dup2.m4:
* m4/fcntl.m4, m4/utimes.m4:
Copy from gnulib.
* m4/gnulib-comp.m4: Regenerate.
2016-01-24 22:50:47 +00:00
|
|
|
|
/* #define DEBUG_MKTIME 1 */
|
2011-01-09 06:57:07 +00:00
|
|
|
|
|
|
|
|
|
#ifndef _LIBC
|
|
|
|
|
# include <config.h>
|
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
/* Assume that leap seconds are possible, unless told otherwise.
|
2012-05-26 23:14:36 +00:00
|
|
|
|
If the host has a 'zic' command with a '-L leapsecondfilename' option,
|
2011-01-09 06:57:07 +00:00
|
|
|
|
then it supports leap seconds; otherwise it probably doesn't. */
|
|
|
|
|
#ifndef LEAP_SECONDS_POSSIBLE
|
|
|
|
|
# define LEAP_SECONDS_POSSIBLE 1
|
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
#include <time.h>
|
|
|
|
|
|
|
|
|
|
#include <limits.h>
|
|
|
|
|
|
2012-05-26 23:14:36 +00:00
|
|
|
|
#include <string.h> /* For the real memcpy prototype. */
|
2011-01-09 06:57:07 +00:00
|
|
|
|
|
Merge from gnulib
This incorporates:
2016-01-24 openat_proc_name: fix last '/' overwritten on OS/2 kLIBC
2016-01-24 closedir, dirfd, opendir: port to OpenSolaris 5.10
2016-01-15 detect utimes() correctly on OS/2 kLIBC
2016-01-15 openat_proc_name: port to OS/2 kLIBC
2016-01-14 stdint: check _INTPTR_T_DECLARED for intptr_t etc.
2016-01-14 opendir, closedir, dirfd, fdopendir: port to OS/2 kLIBC
2016-01-14 dup, dup2, fcntl: support a directory fd on OS/2 kLIBC
2016-01-14 binary-io: don't put fd in binary mode if a console on EMX
2016-01-14 sig2str: list all signals on FreeBSD >= 7
2016-01-13 acl-permissions: port to USE_ACL==0 platforms
2016-01-12 mktime: rename macro to avoid glibc clash
2016-01-12 Port "$@" to OpenIndiana ksh93
2016-01-12 Port Universal Time settings to strict POSIX
* build-aux/gitlog-to-changelog, build-aux/update-copyright:
* doc/misc/texinfo.tex, lib/acl-internal.c, lib/acl-internal.h:
* lib/binary-io.h, lib/dirent.in.h, lib/dirfd.c, lib/dup2.c:
* lib/fcntl.c, lib/fdopendir.c, lib/mktime.c, lib/openat-proc.c:
* lib/sig2str.h, lib/stdint.in.h, m4/dirfd.m4, m4/dup2.m4:
* m4/fcntl.m4, m4/utimes.m4:
Copy from gnulib.
* m4/gnulib-comp.m4: Regenerate.
2016-01-24 22:50:47 +00:00
|
|
|
|
#if defined DEBUG_MKTIME && DEBUG_MKTIME
|
2011-01-09 06:57:07 +00:00
|
|
|
|
# include <stdio.h>
|
|
|
|
|
# include <stdlib.h>
|
|
|
|
|
/* Make it work even if the system's libc has its own mktime routine. */
|
gnulib: import mktime and move-if-change fixes from gnulib
* configure: Regenerate from the following.
2011-01-30 Paul Eggert <eggert@cs.ucla.edu>
mktime: clarify long_int width checking
* lib/mktime.c (long_int_is_wide_enough): Move this assertion to
the top level, to make it clearer that the assumption about
long_int width is being checked. See
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00554.html>.
2011-01-29 Paul Eggert <eggert@cs.ucla.edu>
TYPE_MAXIMUM: avoid theoretically undefined behavior
* lib/intprops.h (TYPE_MINIMUM, TYPE_MAXIMUM): Do not shift a
negative number, which the C Standard says has undefined behavior.
In practice this is not a problem, but might as well do it by the book.
Reported by Rich Felker and Eric Blake; see
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00493.html>.
* m4/mktime.m4 (AC_FUNC_MKTIME): Likewise.
* lib/mktime.c (TYPE_MAXIMUM): Redo slightly to match the others.
mktime: #undef mktime before #defining it
* lib/mktime.c (mktime) [DEBUG]: #undef mktime before #defining it.
mktime: systematically normalize tm_isdst comparisons
* lib/mktime.c (isdst_differ): New function.
(__mktime_internal): Use it systematically for all isdst comparisons.
This completes the fix for libc BZ #6723, and removes the need for
normalizing tm_isdst. See
<http://sourceware.org/bugzilla/show_bug.cgi?id=6723>
(not_equal_tm) [DEBUG]: Use isdst_differ here, too.
mktime: fix some integer overflow issues and sidestep the rest
This was prompted by a bug report by Benjamin Lindner for MinGW
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00472.html>.
His bug is due to signed integer overflow (0 - INT_MIN), and I
I scanned through mktime.c looking for other integer overflow
problems, fixing all the bugs I found.
Although the C Standard says the resulting code is still not safe
in the presence of integer overflow, in practice it should be good
enough for all real-world two's-complement implementations, except
for debugging environments that deliberately trap on integer
overflow (e.g., gcc -ftrapv).
* lib/mktime.c (WRAPV): New macro.
(SHR): Also check that long_int and time_t shift right in the
usual way, before using the fast-but-unportable method.
(TYPE_ONES_COMPLEMENT, TYPE_SIGNED_MAGNITUDE): Remove, no longer
used. The code already assumed two's complement, so there's
no need to test for alternatives. All uses removed.
(TYPE_MAXIMUM): Don't rely here on overflow behavior not defined by
the C standard. Problem reported by Rich Felker in
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00488.html>.
(twos_complement_arithmetic): Also check long_int and time_t.
(time_t_avg, time_t_add_ok, time_t_int_add_ok): New functions.
(guess_time_tm, ranged_convert, __mktime_internal): Use them.
(__mktime_internal): Avoid integer overflow with unary subtraction
in two instances where -1 - X is an adequate replacement for -X,
since the calculations are approximate.
2011-01-29 Eric Blake <eblake@redhat.com>
mktime: avoid infinite loop
* m4/mktime.m4 (AC_FUNC_MKTIME): Avoid overflow on possibly-signed
type; behavior is still undefined but portable to all known targets.
Reported by Rich Felker.
2011-01-28 Paul Eggert <eggert@cs.ucla.edu>
mktime: avoid problems on NetBSD 5 / i386
* lib/mktime.c (long_int): New type. This works around a problem
on NetBSD 5 / i386, where 'long int' and 'int' are both 32 bits
but time_t is 64 bits, and where I expect the existing code is
wrong in some cases.
(leapyear, ydhms_diff, guess_time_tm, __mktime_internal): Use it.
(ydhms_diff): Bring back the compile-time check for wide-enough
year and yday.
mktime: fix misspelling in comment
* lib/mktime.c (__mktime_internal): Fix misspelling in comment.
This merges all recent glibc changes of importance.
2011-01-28 Ralf Wildenhues <Ralf.Wildenhues@gmx.de>
move-if-change: cope with concurrent mv of identical file.
* move-if-change (CMPPROG): Accept environment
variable as an override for `cmp'.
(usage): Document CMPPROG.
Adjust comparison to drop stdout. Cope with failure of mv if
the target file exists and is identical to the source, for
parallel builds.
Report from H.J. Lu against binutils in PR binutils/12283.
2011-01-30 19:22:02 +00:00
|
|
|
|
# undef mktime
|
2011-01-09 06:57:07 +00:00
|
|
|
|
# define mktime my_mktime
|
Merge from gnulib
This incorporates:
2016-01-24 openat_proc_name: fix last '/' overwritten on OS/2 kLIBC
2016-01-24 closedir, dirfd, opendir: port to OpenSolaris 5.10
2016-01-15 detect utimes() correctly on OS/2 kLIBC
2016-01-15 openat_proc_name: port to OS/2 kLIBC
2016-01-14 stdint: check _INTPTR_T_DECLARED for intptr_t etc.
2016-01-14 opendir, closedir, dirfd, fdopendir: port to OS/2 kLIBC
2016-01-14 dup, dup2, fcntl: support a directory fd on OS/2 kLIBC
2016-01-14 binary-io: don't put fd in binary mode if a console on EMX
2016-01-14 sig2str: list all signals on FreeBSD >= 7
2016-01-13 acl-permissions: port to USE_ACL==0 platforms
2016-01-12 mktime: rename macro to avoid glibc clash
2016-01-12 Port "$@" to OpenIndiana ksh93
2016-01-12 Port Universal Time settings to strict POSIX
* build-aux/gitlog-to-changelog, build-aux/update-copyright:
* doc/misc/texinfo.tex, lib/acl-internal.c, lib/acl-internal.h:
* lib/binary-io.h, lib/dirent.in.h, lib/dirfd.c, lib/dup2.c:
* lib/fcntl.c, lib/fdopendir.c, lib/mktime.c, lib/openat-proc.c:
* lib/sig2str.h, lib/stdint.in.h, m4/dirfd.m4, m4/dup2.m4:
* m4/fcntl.m4, m4/utimes.m4:
Copy from gnulib.
* m4/gnulib-comp.m4: Regenerate.
2016-01-24 22:50:47 +00:00
|
|
|
|
#endif /* DEBUG_MKTIME */
|
2011-01-09 06:57:07 +00:00
|
|
|
|
|
2012-05-26 23:14:36 +00:00
|
|
|
|
/* Some of the code in this file assumes that signed integer overflow
|
|
|
|
|
silently wraps around. This assumption can't easily be programmed
|
|
|
|
|
around, nor can it be checked for portably at compile-time or
|
|
|
|
|
easily eliminated at run-time.
|
|
|
|
|
|
|
|
|
|
Define WRAPV to 1 if the assumption is valid and if
|
|
|
|
|
#pragma GCC optimize ("wrapv")
|
|
|
|
|
does not trigger GCC bug 51793
|
|
|
|
|
<http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51793>.
|
|
|
|
|
Otherwise, define it to 0; this forces the use of slower code that,
|
|
|
|
|
while not guaranteed by the C Standard, works on all production
|
|
|
|
|
platforms that we know about. */
|
|
|
|
|
#ifndef WRAPV
|
|
|
|
|
# if (((__GNUC__ == 4 && 4 <= __GNUC_MINOR__) || 4 < __GNUC__) \
|
|
|
|
|
&& defined __GLIBC__)
|
|
|
|
|
# pragma GCC optimize ("wrapv")
|
|
|
|
|
# define WRAPV 1
|
|
|
|
|
# else
|
|
|
|
|
# define WRAPV 0
|
|
|
|
|
# endif
|
|
|
|
|
#endif
|
|
|
|
|
|
gnulib: import mktime and move-if-change fixes from gnulib
* configure: Regenerate from the following.
2011-01-30 Paul Eggert <eggert@cs.ucla.edu>
mktime: clarify long_int width checking
* lib/mktime.c (long_int_is_wide_enough): Move this assertion to
the top level, to make it clearer that the assumption about
long_int width is being checked. See
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00554.html>.
2011-01-29 Paul Eggert <eggert@cs.ucla.edu>
TYPE_MAXIMUM: avoid theoretically undefined behavior
* lib/intprops.h (TYPE_MINIMUM, TYPE_MAXIMUM): Do not shift a
negative number, which the C Standard says has undefined behavior.
In practice this is not a problem, but might as well do it by the book.
Reported by Rich Felker and Eric Blake; see
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00493.html>.
* m4/mktime.m4 (AC_FUNC_MKTIME): Likewise.
* lib/mktime.c (TYPE_MAXIMUM): Redo slightly to match the others.
mktime: #undef mktime before #defining it
* lib/mktime.c (mktime) [DEBUG]: #undef mktime before #defining it.
mktime: systematically normalize tm_isdst comparisons
* lib/mktime.c (isdst_differ): New function.
(__mktime_internal): Use it systematically for all isdst comparisons.
This completes the fix for libc BZ #6723, and removes the need for
normalizing tm_isdst. See
<http://sourceware.org/bugzilla/show_bug.cgi?id=6723>
(not_equal_tm) [DEBUG]: Use isdst_differ here, too.
mktime: fix some integer overflow issues and sidestep the rest
This was prompted by a bug report by Benjamin Lindner for MinGW
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00472.html>.
His bug is due to signed integer overflow (0 - INT_MIN), and I
I scanned through mktime.c looking for other integer overflow
problems, fixing all the bugs I found.
Although the C Standard says the resulting code is still not safe
in the presence of integer overflow, in practice it should be good
enough for all real-world two's-complement implementations, except
for debugging environments that deliberately trap on integer
overflow (e.g., gcc -ftrapv).
* lib/mktime.c (WRAPV): New macro.
(SHR): Also check that long_int and time_t shift right in the
usual way, before using the fast-but-unportable method.
(TYPE_ONES_COMPLEMENT, TYPE_SIGNED_MAGNITUDE): Remove, no longer
used. The code already assumed two's complement, so there's
no need to test for alternatives. All uses removed.
(TYPE_MAXIMUM): Don't rely here on overflow behavior not defined by
the C standard. Problem reported by Rich Felker in
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00488.html>.
(twos_complement_arithmetic): Also check long_int and time_t.
(time_t_avg, time_t_add_ok, time_t_int_add_ok): New functions.
(guess_time_tm, ranged_convert, __mktime_internal): Use them.
(__mktime_internal): Avoid integer overflow with unary subtraction
in two instances where -1 - X is an adequate replacement for -X,
since the calculations are approximate.
2011-01-29 Eric Blake <eblake@redhat.com>
mktime: avoid infinite loop
* m4/mktime.m4 (AC_FUNC_MKTIME): Avoid overflow on possibly-signed
type; behavior is still undefined but portable to all known targets.
Reported by Rich Felker.
2011-01-28 Paul Eggert <eggert@cs.ucla.edu>
mktime: avoid problems on NetBSD 5 / i386
* lib/mktime.c (long_int): New type. This works around a problem
on NetBSD 5 / i386, where 'long int' and 'int' are both 32 bits
but time_t is 64 bits, and where I expect the existing code is
wrong in some cases.
(leapyear, ydhms_diff, guess_time_tm, __mktime_internal): Use it.
(ydhms_diff): Bring back the compile-time check for wide-enough
year and yday.
mktime: fix misspelling in comment
* lib/mktime.c (__mktime_internal): Fix misspelling in comment.
This merges all recent glibc changes of importance.
2011-01-28 Ralf Wildenhues <Ralf.Wildenhues@gmx.de>
move-if-change: cope with concurrent mv of identical file.
* move-if-change (CMPPROG): Accept environment
variable as an override for `cmp'.
(usage): Document CMPPROG.
Adjust comparison to drop stdout. Cope with failure of mv if
the target file exists and is identical to the source, for
parallel builds.
Report from H.J. Lu against binutils in PR binutils/12283.
2011-01-30 19:22:02 +00:00
|
|
|
|
/* Verify a requirement at compile-time (unlike assert, which is runtime). */
|
|
|
|
|
#define verify(name, assertion) struct name { char a[(assertion) ? 1 : -1]; }
|
|
|
|
|
|
|
|
|
|
/* A signed type that is at least one bit wider than int. */
|
|
|
|
|
#if INT_MAX <= LONG_MAX / 2
|
|
|
|
|
typedef long int long_int;
|
|
|
|
|
#else
|
|
|
|
|
typedef long long int long_int;
|
|
|
|
|
#endif
|
|
|
|
|
verify (long_int_is_wide_enough, INT_MAX == INT_MAX * (long_int) 2 / 2);
|
|
|
|
|
|
2011-01-09 06:57:07 +00:00
|
|
|
|
/* Shift A right by B bits portably, by dividing A by 2**B and
|
|
|
|
|
truncating towards minus infinity. A and B should be free of side
|
|
|
|
|
effects, and B should be in the range 0 <= B <= INT_BITS - 2, where
|
|
|
|
|
INT_BITS is the number of useful bits in an int. GNU code can
|
|
|
|
|
assume that INT_BITS is at least 32.
|
|
|
|
|
|
|
|
|
|
ISO C99 says that A >> B is implementation-defined if A < 0. Some
|
|
|
|
|
implementations (e.g., UNICOS 9.0 on a Cray Y-MP EL) don't shift
|
|
|
|
|
right in the usual way when A < 0, so SHR falls back on division if
|
|
|
|
|
ordinary A >> B doesn't seem to be the usual signed shift. */
|
gnulib: import mktime and move-if-change fixes from gnulib
* configure: Regenerate from the following.
2011-01-30 Paul Eggert <eggert@cs.ucla.edu>
mktime: clarify long_int width checking
* lib/mktime.c (long_int_is_wide_enough): Move this assertion to
the top level, to make it clearer that the assumption about
long_int width is being checked. See
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00554.html>.
2011-01-29 Paul Eggert <eggert@cs.ucla.edu>
TYPE_MAXIMUM: avoid theoretically undefined behavior
* lib/intprops.h (TYPE_MINIMUM, TYPE_MAXIMUM): Do not shift a
negative number, which the C Standard says has undefined behavior.
In practice this is not a problem, but might as well do it by the book.
Reported by Rich Felker and Eric Blake; see
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00493.html>.
* m4/mktime.m4 (AC_FUNC_MKTIME): Likewise.
* lib/mktime.c (TYPE_MAXIMUM): Redo slightly to match the others.
mktime: #undef mktime before #defining it
* lib/mktime.c (mktime) [DEBUG]: #undef mktime before #defining it.
mktime: systematically normalize tm_isdst comparisons
* lib/mktime.c (isdst_differ): New function.
(__mktime_internal): Use it systematically for all isdst comparisons.
This completes the fix for libc BZ #6723, and removes the need for
normalizing tm_isdst. See
<http://sourceware.org/bugzilla/show_bug.cgi?id=6723>
(not_equal_tm) [DEBUG]: Use isdst_differ here, too.
mktime: fix some integer overflow issues and sidestep the rest
This was prompted by a bug report by Benjamin Lindner for MinGW
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00472.html>.
His bug is due to signed integer overflow (0 - INT_MIN), and I
I scanned through mktime.c looking for other integer overflow
problems, fixing all the bugs I found.
Although the C Standard says the resulting code is still not safe
in the presence of integer overflow, in practice it should be good
enough for all real-world two's-complement implementations, except
for debugging environments that deliberately trap on integer
overflow (e.g., gcc -ftrapv).
* lib/mktime.c (WRAPV): New macro.
(SHR): Also check that long_int and time_t shift right in the
usual way, before using the fast-but-unportable method.
(TYPE_ONES_COMPLEMENT, TYPE_SIGNED_MAGNITUDE): Remove, no longer
used. The code already assumed two's complement, so there's
no need to test for alternatives. All uses removed.
(TYPE_MAXIMUM): Don't rely here on overflow behavior not defined by
the C standard. Problem reported by Rich Felker in
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00488.html>.
(twos_complement_arithmetic): Also check long_int and time_t.
(time_t_avg, time_t_add_ok, time_t_int_add_ok): New functions.
(guess_time_tm, ranged_convert, __mktime_internal): Use them.
(__mktime_internal): Avoid integer overflow with unary subtraction
in two instances where -1 - X is an adequate replacement for -X,
since the calculations are approximate.
2011-01-29 Eric Blake <eblake@redhat.com>
mktime: avoid infinite loop
* m4/mktime.m4 (AC_FUNC_MKTIME): Avoid overflow on possibly-signed
type; behavior is still undefined but portable to all known targets.
Reported by Rich Felker.
2011-01-28 Paul Eggert <eggert@cs.ucla.edu>
mktime: avoid problems on NetBSD 5 / i386
* lib/mktime.c (long_int): New type. This works around a problem
on NetBSD 5 / i386, where 'long int' and 'int' are both 32 bits
but time_t is 64 bits, and where I expect the existing code is
wrong in some cases.
(leapyear, ydhms_diff, guess_time_tm, __mktime_internal): Use it.
(ydhms_diff): Bring back the compile-time check for wide-enough
year and yday.
mktime: fix misspelling in comment
* lib/mktime.c (__mktime_internal): Fix misspelling in comment.
This merges all recent glibc changes of importance.
2011-01-28 Ralf Wildenhues <Ralf.Wildenhues@gmx.de>
move-if-change: cope with concurrent mv of identical file.
* move-if-change (CMPPROG): Accept environment
variable as an override for `cmp'.
(usage): Document CMPPROG.
Adjust comparison to drop stdout. Cope with failure of mv if
the target file exists and is identical to the source, for
parallel builds.
Report from H.J. Lu against binutils in PR binutils/12283.
2011-01-30 19:22:02 +00:00
|
|
|
|
#define SHR(a, b) \
|
|
|
|
|
((-1 >> 1 == -1 \
|
|
|
|
|
&& (long_int) -1 >> 1 == -1 \
|
|
|
|
|
&& ((time_t) -1 >> 1 == -1 || ! TYPE_SIGNED (time_t))) \
|
|
|
|
|
? (a) >> (b) \
|
2011-01-09 06:57:07 +00:00
|
|
|
|
: (a) / (1 << (b)) - ((a) % (1 << (b)) < 0))
|
|
|
|
|
|
|
|
|
|
/* The extra casts in the following macros work around compiler bugs,
|
|
|
|
|
e.g., in Cray C 5.0.3.0. */
|
|
|
|
|
|
|
|
|
|
/* True if the arithmetic type T is an integer type. bool counts as
|
|
|
|
|
an integer. */
|
|
|
|
|
#define TYPE_IS_INTEGER(t) ((t) 1.5 == 1)
|
|
|
|
|
|
|
|
|
|
/* True if negative values of the signed integer type T use two's
|
gnulib: import mktime and move-if-change fixes from gnulib
* configure: Regenerate from the following.
2011-01-30 Paul Eggert <eggert@cs.ucla.edu>
mktime: clarify long_int width checking
* lib/mktime.c (long_int_is_wide_enough): Move this assertion to
the top level, to make it clearer that the assumption about
long_int width is being checked. See
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00554.html>.
2011-01-29 Paul Eggert <eggert@cs.ucla.edu>
TYPE_MAXIMUM: avoid theoretically undefined behavior
* lib/intprops.h (TYPE_MINIMUM, TYPE_MAXIMUM): Do not shift a
negative number, which the C Standard says has undefined behavior.
In practice this is not a problem, but might as well do it by the book.
Reported by Rich Felker and Eric Blake; see
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00493.html>.
* m4/mktime.m4 (AC_FUNC_MKTIME): Likewise.
* lib/mktime.c (TYPE_MAXIMUM): Redo slightly to match the others.
mktime: #undef mktime before #defining it
* lib/mktime.c (mktime) [DEBUG]: #undef mktime before #defining it.
mktime: systematically normalize tm_isdst comparisons
* lib/mktime.c (isdst_differ): New function.
(__mktime_internal): Use it systematically for all isdst comparisons.
This completes the fix for libc BZ #6723, and removes the need for
normalizing tm_isdst. See
<http://sourceware.org/bugzilla/show_bug.cgi?id=6723>
(not_equal_tm) [DEBUG]: Use isdst_differ here, too.
mktime: fix some integer overflow issues and sidestep the rest
This was prompted by a bug report by Benjamin Lindner for MinGW
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00472.html>.
His bug is due to signed integer overflow (0 - INT_MIN), and I
I scanned through mktime.c looking for other integer overflow
problems, fixing all the bugs I found.
Although the C Standard says the resulting code is still not safe
in the presence of integer overflow, in practice it should be good
enough for all real-world two's-complement implementations, except
for debugging environments that deliberately trap on integer
overflow (e.g., gcc -ftrapv).
* lib/mktime.c (WRAPV): New macro.
(SHR): Also check that long_int and time_t shift right in the
usual way, before using the fast-but-unportable method.
(TYPE_ONES_COMPLEMENT, TYPE_SIGNED_MAGNITUDE): Remove, no longer
used. The code already assumed two's complement, so there's
no need to test for alternatives. All uses removed.
(TYPE_MAXIMUM): Don't rely here on overflow behavior not defined by
the C standard. Problem reported by Rich Felker in
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00488.html>.
(twos_complement_arithmetic): Also check long_int and time_t.
(time_t_avg, time_t_add_ok, time_t_int_add_ok): New functions.
(guess_time_tm, ranged_convert, __mktime_internal): Use them.
(__mktime_internal): Avoid integer overflow with unary subtraction
in two instances where -1 - X is an adequate replacement for -X,
since the calculations are approximate.
2011-01-29 Eric Blake <eblake@redhat.com>
mktime: avoid infinite loop
* m4/mktime.m4 (AC_FUNC_MKTIME): Avoid overflow on possibly-signed
type; behavior is still undefined but portable to all known targets.
Reported by Rich Felker.
2011-01-28 Paul Eggert <eggert@cs.ucla.edu>
mktime: avoid problems on NetBSD 5 / i386
* lib/mktime.c (long_int): New type. This works around a problem
on NetBSD 5 / i386, where 'long int' and 'int' are both 32 bits
but time_t is 64 bits, and where I expect the existing code is
wrong in some cases.
(leapyear, ydhms_diff, guess_time_tm, __mktime_internal): Use it.
(ydhms_diff): Bring back the compile-time check for wide-enough
year and yday.
mktime: fix misspelling in comment
* lib/mktime.c (__mktime_internal): Fix misspelling in comment.
This merges all recent glibc changes of importance.
2011-01-28 Ralf Wildenhues <Ralf.Wildenhues@gmx.de>
move-if-change: cope with concurrent mv of identical file.
* move-if-change (CMPPROG): Accept environment
variable as an override for `cmp'.
(usage): Document CMPPROG.
Adjust comparison to drop stdout. Cope with failure of mv if
the target file exists and is identical to the source, for
parallel builds.
Report from H.J. Lu against binutils in PR binutils/12283.
2011-01-30 19:22:02 +00:00
|
|
|
|
complement, or if T is an unsigned integer type. */
|
2011-01-09 06:57:07 +00:00
|
|
|
|
#define TYPE_TWOS_COMPLEMENT(t) ((t) ~ (t) 0 == (t) -1)
|
|
|
|
|
|
|
|
|
|
/* True if the arithmetic type T is signed. */
|
|
|
|
|
#define TYPE_SIGNED(t) (! ((t) 0 < (t) -1))
|
|
|
|
|
|
|
|
|
|
/* The maximum and minimum values for the integer type T. These
|
|
|
|
|
macros have undefined behavior if T is signed and has padding bits.
|
|
|
|
|
If this is a problem for you, please let us know how to fix it for
|
|
|
|
|
your host. */
|
|
|
|
|
#define TYPE_MINIMUM(t) \
|
|
|
|
|
((t) (! TYPE_SIGNED (t) \
|
2012-05-26 23:14:36 +00:00
|
|
|
|
? (t) 0 \
|
|
|
|
|
: ~ TYPE_MAXIMUM (t)))
|
2011-01-09 06:57:07 +00:00
|
|
|
|
#define TYPE_MAXIMUM(t) \
|
|
|
|
|
((t) (! TYPE_SIGNED (t) \
|
2012-05-26 23:14:36 +00:00
|
|
|
|
? (t) -1 \
|
|
|
|
|
: ((((t) 1 << (sizeof (t) * CHAR_BIT - 2)) - 1) * 2 + 1)))
|
2011-01-09 06:57:07 +00:00
|
|
|
|
|
|
|
|
|
#ifndef TIME_T_MIN
|
|
|
|
|
# define TIME_T_MIN TYPE_MINIMUM (time_t)
|
|
|
|
|
#endif
|
|
|
|
|
#ifndef TIME_T_MAX
|
|
|
|
|
# define TIME_T_MAX TYPE_MAXIMUM (time_t)
|
|
|
|
|
#endif
|
|
|
|
|
#define TIME_T_MIDPOINT (SHR (TIME_T_MIN + TIME_T_MAX, 1) + 1)
|
|
|
|
|
|
|
|
|
|
verify (time_t_is_integer, TYPE_IS_INTEGER (time_t));
|
gnulib: import mktime and move-if-change fixes from gnulib
* configure: Regenerate from the following.
2011-01-30 Paul Eggert <eggert@cs.ucla.edu>
mktime: clarify long_int width checking
* lib/mktime.c (long_int_is_wide_enough): Move this assertion to
the top level, to make it clearer that the assumption about
long_int width is being checked. See
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00554.html>.
2011-01-29 Paul Eggert <eggert@cs.ucla.edu>
TYPE_MAXIMUM: avoid theoretically undefined behavior
* lib/intprops.h (TYPE_MINIMUM, TYPE_MAXIMUM): Do not shift a
negative number, which the C Standard says has undefined behavior.
In practice this is not a problem, but might as well do it by the book.
Reported by Rich Felker and Eric Blake; see
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00493.html>.
* m4/mktime.m4 (AC_FUNC_MKTIME): Likewise.
* lib/mktime.c (TYPE_MAXIMUM): Redo slightly to match the others.
mktime: #undef mktime before #defining it
* lib/mktime.c (mktime) [DEBUG]: #undef mktime before #defining it.
mktime: systematically normalize tm_isdst comparisons
* lib/mktime.c (isdst_differ): New function.
(__mktime_internal): Use it systematically for all isdst comparisons.
This completes the fix for libc BZ #6723, and removes the need for
normalizing tm_isdst. See
<http://sourceware.org/bugzilla/show_bug.cgi?id=6723>
(not_equal_tm) [DEBUG]: Use isdst_differ here, too.
mktime: fix some integer overflow issues and sidestep the rest
This was prompted by a bug report by Benjamin Lindner for MinGW
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00472.html>.
His bug is due to signed integer overflow (0 - INT_MIN), and I
I scanned through mktime.c looking for other integer overflow
problems, fixing all the bugs I found.
Although the C Standard says the resulting code is still not safe
in the presence of integer overflow, in practice it should be good
enough for all real-world two's-complement implementations, except
for debugging environments that deliberately trap on integer
overflow (e.g., gcc -ftrapv).
* lib/mktime.c (WRAPV): New macro.
(SHR): Also check that long_int and time_t shift right in the
usual way, before using the fast-but-unportable method.
(TYPE_ONES_COMPLEMENT, TYPE_SIGNED_MAGNITUDE): Remove, no longer
used. The code already assumed two's complement, so there's
no need to test for alternatives. All uses removed.
(TYPE_MAXIMUM): Don't rely here on overflow behavior not defined by
the C standard. Problem reported by Rich Felker in
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00488.html>.
(twos_complement_arithmetic): Also check long_int and time_t.
(time_t_avg, time_t_add_ok, time_t_int_add_ok): New functions.
(guess_time_tm, ranged_convert, __mktime_internal): Use them.
(__mktime_internal): Avoid integer overflow with unary subtraction
in two instances where -1 - X is an adequate replacement for -X,
since the calculations are approximate.
2011-01-29 Eric Blake <eblake@redhat.com>
mktime: avoid infinite loop
* m4/mktime.m4 (AC_FUNC_MKTIME): Avoid overflow on possibly-signed
type; behavior is still undefined but portable to all known targets.
Reported by Rich Felker.
2011-01-28 Paul Eggert <eggert@cs.ucla.edu>
mktime: avoid problems on NetBSD 5 / i386
* lib/mktime.c (long_int): New type. This works around a problem
on NetBSD 5 / i386, where 'long int' and 'int' are both 32 bits
but time_t is 64 bits, and where I expect the existing code is
wrong in some cases.
(leapyear, ydhms_diff, guess_time_tm, __mktime_internal): Use it.
(ydhms_diff): Bring back the compile-time check for wide-enough
year and yday.
mktime: fix misspelling in comment
* lib/mktime.c (__mktime_internal): Fix misspelling in comment.
This merges all recent glibc changes of importance.
2011-01-28 Ralf Wildenhues <Ralf.Wildenhues@gmx.de>
move-if-change: cope with concurrent mv of identical file.
* move-if-change (CMPPROG): Accept environment
variable as an override for `cmp'.
(usage): Document CMPPROG.
Adjust comparison to drop stdout. Cope with failure of mv if
the target file exists and is identical to the source, for
parallel builds.
Report from H.J. Lu against binutils in PR binutils/12283.
2011-01-30 19:22:02 +00:00
|
|
|
|
verify (twos_complement_arithmetic,
|
2012-05-26 23:14:36 +00:00
|
|
|
|
(TYPE_TWOS_COMPLEMENT (int)
|
|
|
|
|
&& TYPE_TWOS_COMPLEMENT (long_int)
|
|
|
|
|
&& TYPE_TWOS_COMPLEMENT (time_t)));
|
2011-01-09 06:57:07 +00:00
|
|
|
|
|
|
|
|
|
#define EPOCH_YEAR 1970
|
|
|
|
|
#define TM_YEAR_BASE 1900
|
|
|
|
|
verify (base_year_is_a_multiple_of_100, TM_YEAR_BASE % 100 == 0);
|
|
|
|
|
|
|
|
|
|
/* Return 1 if YEAR + TM_YEAR_BASE is a leap year. */
|
2012-08-19 23:31:24 +00:00
|
|
|
|
static int
|
gnulib: import mktime and move-if-change fixes from gnulib
* configure: Regenerate from the following.
2011-01-30 Paul Eggert <eggert@cs.ucla.edu>
mktime: clarify long_int width checking
* lib/mktime.c (long_int_is_wide_enough): Move this assertion to
the top level, to make it clearer that the assumption about
long_int width is being checked. See
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00554.html>.
2011-01-29 Paul Eggert <eggert@cs.ucla.edu>
TYPE_MAXIMUM: avoid theoretically undefined behavior
* lib/intprops.h (TYPE_MINIMUM, TYPE_MAXIMUM): Do not shift a
negative number, which the C Standard says has undefined behavior.
In practice this is not a problem, but might as well do it by the book.
Reported by Rich Felker and Eric Blake; see
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00493.html>.
* m4/mktime.m4 (AC_FUNC_MKTIME): Likewise.
* lib/mktime.c (TYPE_MAXIMUM): Redo slightly to match the others.
mktime: #undef mktime before #defining it
* lib/mktime.c (mktime) [DEBUG]: #undef mktime before #defining it.
mktime: systematically normalize tm_isdst comparisons
* lib/mktime.c (isdst_differ): New function.
(__mktime_internal): Use it systematically for all isdst comparisons.
This completes the fix for libc BZ #6723, and removes the need for
normalizing tm_isdst. See
<http://sourceware.org/bugzilla/show_bug.cgi?id=6723>
(not_equal_tm) [DEBUG]: Use isdst_differ here, too.
mktime: fix some integer overflow issues and sidestep the rest
This was prompted by a bug report by Benjamin Lindner for MinGW
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00472.html>.
His bug is due to signed integer overflow (0 - INT_MIN), and I
I scanned through mktime.c looking for other integer overflow
problems, fixing all the bugs I found.
Although the C Standard says the resulting code is still not safe
in the presence of integer overflow, in practice it should be good
enough for all real-world two's-complement implementations, except
for debugging environments that deliberately trap on integer
overflow (e.g., gcc -ftrapv).
* lib/mktime.c (WRAPV): New macro.
(SHR): Also check that long_int and time_t shift right in the
usual way, before using the fast-but-unportable method.
(TYPE_ONES_COMPLEMENT, TYPE_SIGNED_MAGNITUDE): Remove, no longer
used. The code already assumed two's complement, so there's
no need to test for alternatives. All uses removed.
(TYPE_MAXIMUM): Don't rely here on overflow behavior not defined by
the C standard. Problem reported by Rich Felker in
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00488.html>.
(twos_complement_arithmetic): Also check long_int and time_t.
(time_t_avg, time_t_add_ok, time_t_int_add_ok): New functions.
(guess_time_tm, ranged_convert, __mktime_internal): Use them.
(__mktime_internal): Avoid integer overflow with unary subtraction
in two instances where -1 - X is an adequate replacement for -X,
since the calculations are approximate.
2011-01-29 Eric Blake <eblake@redhat.com>
mktime: avoid infinite loop
* m4/mktime.m4 (AC_FUNC_MKTIME): Avoid overflow on possibly-signed
type; behavior is still undefined but portable to all known targets.
Reported by Rich Felker.
2011-01-28 Paul Eggert <eggert@cs.ucla.edu>
mktime: avoid problems on NetBSD 5 / i386
* lib/mktime.c (long_int): New type. This works around a problem
on NetBSD 5 / i386, where 'long int' and 'int' are both 32 bits
but time_t is 64 bits, and where I expect the existing code is
wrong in some cases.
(leapyear, ydhms_diff, guess_time_tm, __mktime_internal): Use it.
(ydhms_diff): Bring back the compile-time check for wide-enough
year and yday.
mktime: fix misspelling in comment
* lib/mktime.c (__mktime_internal): Fix misspelling in comment.
This merges all recent glibc changes of importance.
2011-01-28 Ralf Wildenhues <Ralf.Wildenhues@gmx.de>
move-if-change: cope with concurrent mv of identical file.
* move-if-change (CMPPROG): Accept environment
variable as an override for `cmp'.
(usage): Document CMPPROG.
Adjust comparison to drop stdout. Cope with failure of mv if
the target file exists and is identical to the source, for
parallel builds.
Report from H.J. Lu against binutils in PR binutils/12283.
2011-01-30 19:22:02 +00:00
|
|
|
|
leapyear (long_int year)
|
2011-01-09 06:57:07 +00:00
|
|
|
|
{
|
|
|
|
|
/* Don't add YEAR to TM_YEAR_BASE, as that might overflow.
|
|
|
|
|
Also, work even if YEAR is negative. */
|
|
|
|
|
return
|
|
|
|
|
((year & 3) == 0
|
|
|
|
|
&& (year % 100 != 0
|
2012-05-26 23:14:36 +00:00
|
|
|
|
|| ((year / 100) & 3) == (- (TM_YEAR_BASE / 100) & 3)));
|
2011-01-09 06:57:07 +00:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* How many days come before each month (0-12). */
|
|
|
|
|
#ifndef _LIBC
|
|
|
|
|
static
|
|
|
|
|
#endif
|
|
|
|
|
const unsigned short int __mon_yday[2][13] =
|
|
|
|
|
{
|
|
|
|
|
/* Normal years. */
|
|
|
|
|
{ 0, 31, 59, 90, 120, 151, 181, 212, 243, 273, 304, 334, 365 },
|
|
|
|
|
/* Leap years. */
|
|
|
|
|
{ 0, 31, 60, 91, 121, 152, 182, 213, 244, 274, 305, 335, 366 }
|
|
|
|
|
};
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
#ifndef _LIBC
|
|
|
|
|
/* Portable standalone applications should supply a <time.h> that
|
|
|
|
|
declares a POSIX-compliant localtime_r, for the benefit of older
|
|
|
|
|
implementations that lack localtime_r or have a nonstandard one.
|
|
|
|
|
See the gnulib time_r module for one way to implement this. */
|
|
|
|
|
# undef __localtime_r
|
|
|
|
|
# define __localtime_r localtime_r
|
|
|
|
|
# define __mktime_internal mktime_internal
|
|
|
|
|
# include "mktime-internal.h"
|
|
|
|
|
#endif
|
|
|
|
|
|
gnulib: import mktime and move-if-change fixes from gnulib
* configure: Regenerate from the following.
2011-01-30 Paul Eggert <eggert@cs.ucla.edu>
mktime: clarify long_int width checking
* lib/mktime.c (long_int_is_wide_enough): Move this assertion to
the top level, to make it clearer that the assumption about
long_int width is being checked. See
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00554.html>.
2011-01-29 Paul Eggert <eggert@cs.ucla.edu>
TYPE_MAXIMUM: avoid theoretically undefined behavior
* lib/intprops.h (TYPE_MINIMUM, TYPE_MAXIMUM): Do not shift a
negative number, which the C Standard says has undefined behavior.
In practice this is not a problem, but might as well do it by the book.
Reported by Rich Felker and Eric Blake; see
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00493.html>.
* m4/mktime.m4 (AC_FUNC_MKTIME): Likewise.
* lib/mktime.c (TYPE_MAXIMUM): Redo slightly to match the others.
mktime: #undef mktime before #defining it
* lib/mktime.c (mktime) [DEBUG]: #undef mktime before #defining it.
mktime: systematically normalize tm_isdst comparisons
* lib/mktime.c (isdst_differ): New function.
(__mktime_internal): Use it systematically for all isdst comparisons.
This completes the fix for libc BZ #6723, and removes the need for
normalizing tm_isdst. See
<http://sourceware.org/bugzilla/show_bug.cgi?id=6723>
(not_equal_tm) [DEBUG]: Use isdst_differ here, too.
mktime: fix some integer overflow issues and sidestep the rest
This was prompted by a bug report by Benjamin Lindner for MinGW
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00472.html>.
His bug is due to signed integer overflow (0 - INT_MIN), and I
I scanned through mktime.c looking for other integer overflow
problems, fixing all the bugs I found.
Although the C Standard says the resulting code is still not safe
in the presence of integer overflow, in practice it should be good
enough for all real-world two's-complement implementations, except
for debugging environments that deliberately trap on integer
overflow (e.g., gcc -ftrapv).
* lib/mktime.c (WRAPV): New macro.
(SHR): Also check that long_int and time_t shift right in the
usual way, before using the fast-but-unportable method.
(TYPE_ONES_COMPLEMENT, TYPE_SIGNED_MAGNITUDE): Remove, no longer
used. The code already assumed two's complement, so there's
no need to test for alternatives. All uses removed.
(TYPE_MAXIMUM): Don't rely here on overflow behavior not defined by
the C standard. Problem reported by Rich Felker in
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00488.html>.
(twos_complement_arithmetic): Also check long_int and time_t.
(time_t_avg, time_t_add_ok, time_t_int_add_ok): New functions.
(guess_time_tm, ranged_convert, __mktime_internal): Use them.
(__mktime_internal): Avoid integer overflow with unary subtraction
in two instances where -1 - X is an adequate replacement for -X,
since the calculations are approximate.
2011-01-29 Eric Blake <eblake@redhat.com>
mktime: avoid infinite loop
* m4/mktime.m4 (AC_FUNC_MKTIME): Avoid overflow on possibly-signed
type; behavior is still undefined but portable to all known targets.
Reported by Rich Felker.
2011-01-28 Paul Eggert <eggert@cs.ucla.edu>
mktime: avoid problems on NetBSD 5 / i386
* lib/mktime.c (long_int): New type. This works around a problem
on NetBSD 5 / i386, where 'long int' and 'int' are both 32 bits
but time_t is 64 bits, and where I expect the existing code is
wrong in some cases.
(leapyear, ydhms_diff, guess_time_tm, __mktime_internal): Use it.
(ydhms_diff): Bring back the compile-time check for wide-enough
year and yday.
mktime: fix misspelling in comment
* lib/mktime.c (__mktime_internal): Fix misspelling in comment.
This merges all recent glibc changes of importance.
2011-01-28 Ralf Wildenhues <Ralf.Wildenhues@gmx.de>
move-if-change: cope with concurrent mv of identical file.
* move-if-change (CMPPROG): Accept environment
variable as an override for `cmp'.
(usage): Document CMPPROG.
Adjust comparison to drop stdout. Cope with failure of mv if
the target file exists and is identical to the source, for
parallel builds.
Report from H.J. Lu against binutils in PR binutils/12283.
2011-01-30 19:22:02 +00:00
|
|
|
|
/* Return 1 if the values A and B differ according to the rules for
|
|
|
|
|
tm_isdst: A and B differ if one is zero and the other positive. */
|
|
|
|
|
static int
|
|
|
|
|
isdst_differ (int a, int b)
|
|
|
|
|
{
|
2012-05-26 23:14:36 +00:00
|
|
|
|
return (!a != !b) && (0 <= a) && (0 <= b);
|
gnulib: import mktime and move-if-change fixes from gnulib
* configure: Regenerate from the following.
2011-01-30 Paul Eggert <eggert@cs.ucla.edu>
mktime: clarify long_int width checking
* lib/mktime.c (long_int_is_wide_enough): Move this assertion to
the top level, to make it clearer that the assumption about
long_int width is being checked. See
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00554.html>.
2011-01-29 Paul Eggert <eggert@cs.ucla.edu>
TYPE_MAXIMUM: avoid theoretically undefined behavior
* lib/intprops.h (TYPE_MINIMUM, TYPE_MAXIMUM): Do not shift a
negative number, which the C Standard says has undefined behavior.
In practice this is not a problem, but might as well do it by the book.
Reported by Rich Felker and Eric Blake; see
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00493.html>.
* m4/mktime.m4 (AC_FUNC_MKTIME): Likewise.
* lib/mktime.c (TYPE_MAXIMUM): Redo slightly to match the others.
mktime: #undef mktime before #defining it
* lib/mktime.c (mktime) [DEBUG]: #undef mktime before #defining it.
mktime: systematically normalize tm_isdst comparisons
* lib/mktime.c (isdst_differ): New function.
(__mktime_internal): Use it systematically for all isdst comparisons.
This completes the fix for libc BZ #6723, and removes the need for
normalizing tm_isdst. See
<http://sourceware.org/bugzilla/show_bug.cgi?id=6723>
(not_equal_tm) [DEBUG]: Use isdst_differ here, too.
mktime: fix some integer overflow issues and sidestep the rest
This was prompted by a bug report by Benjamin Lindner for MinGW
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00472.html>.
His bug is due to signed integer overflow (0 - INT_MIN), and I
I scanned through mktime.c looking for other integer overflow
problems, fixing all the bugs I found.
Although the C Standard says the resulting code is still not safe
in the presence of integer overflow, in practice it should be good
enough for all real-world two's-complement implementations, except
for debugging environments that deliberately trap on integer
overflow (e.g., gcc -ftrapv).
* lib/mktime.c (WRAPV): New macro.
(SHR): Also check that long_int and time_t shift right in the
usual way, before using the fast-but-unportable method.
(TYPE_ONES_COMPLEMENT, TYPE_SIGNED_MAGNITUDE): Remove, no longer
used. The code already assumed two's complement, so there's
no need to test for alternatives. All uses removed.
(TYPE_MAXIMUM): Don't rely here on overflow behavior not defined by
the C standard. Problem reported by Rich Felker in
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00488.html>.
(twos_complement_arithmetic): Also check long_int and time_t.
(time_t_avg, time_t_add_ok, time_t_int_add_ok): New functions.
(guess_time_tm, ranged_convert, __mktime_internal): Use them.
(__mktime_internal): Avoid integer overflow with unary subtraction
in two instances where -1 - X is an adequate replacement for -X,
since the calculations are approximate.
2011-01-29 Eric Blake <eblake@redhat.com>
mktime: avoid infinite loop
* m4/mktime.m4 (AC_FUNC_MKTIME): Avoid overflow on possibly-signed
type; behavior is still undefined but portable to all known targets.
Reported by Rich Felker.
2011-01-28 Paul Eggert <eggert@cs.ucla.edu>
mktime: avoid problems on NetBSD 5 / i386
* lib/mktime.c (long_int): New type. This works around a problem
on NetBSD 5 / i386, where 'long int' and 'int' are both 32 bits
but time_t is 64 bits, and where I expect the existing code is
wrong in some cases.
(leapyear, ydhms_diff, guess_time_tm, __mktime_internal): Use it.
(ydhms_diff): Bring back the compile-time check for wide-enough
year and yday.
mktime: fix misspelling in comment
* lib/mktime.c (__mktime_internal): Fix misspelling in comment.
This merges all recent glibc changes of importance.
2011-01-28 Ralf Wildenhues <Ralf.Wildenhues@gmx.de>
move-if-change: cope with concurrent mv of identical file.
* move-if-change (CMPPROG): Accept environment
variable as an override for `cmp'.
(usage): Document CMPPROG.
Adjust comparison to drop stdout. Cope with failure of mv if
the target file exists and is identical to the source, for
parallel builds.
Report from H.J. Lu against binutils in PR binutils/12283.
2011-01-30 19:22:02 +00:00
|
|
|
|
}
|
|
|
|
|
|
2011-01-09 06:57:07 +00:00
|
|
|
|
/* Return an integer value measuring (YEAR1-YDAY1 HOUR1:MIN1:SEC1) -
|
|
|
|
|
(YEAR0-YDAY0 HOUR0:MIN0:SEC0) in seconds, assuming that the clocks
|
|
|
|
|
were not adjusted between the time stamps.
|
|
|
|
|
|
|
|
|
|
The YEAR values uses the same numbering as TP->tm_year. Values
|
|
|
|
|
need not be in the usual range. However, YEAR1 must not be less
|
|
|
|
|
than 2 * INT_MIN or greater than 2 * INT_MAX.
|
|
|
|
|
|
|
|
|
|
The result may overflow. It is the caller's responsibility to
|
|
|
|
|
detect overflow. */
|
|
|
|
|
|
2012-08-19 23:31:24 +00:00
|
|
|
|
static time_t
|
gnulib: import mktime and move-if-change fixes from gnulib
* configure: Regenerate from the following.
2011-01-30 Paul Eggert <eggert@cs.ucla.edu>
mktime: clarify long_int width checking
* lib/mktime.c (long_int_is_wide_enough): Move this assertion to
the top level, to make it clearer that the assumption about
long_int width is being checked. See
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00554.html>.
2011-01-29 Paul Eggert <eggert@cs.ucla.edu>
TYPE_MAXIMUM: avoid theoretically undefined behavior
* lib/intprops.h (TYPE_MINIMUM, TYPE_MAXIMUM): Do not shift a
negative number, which the C Standard says has undefined behavior.
In practice this is not a problem, but might as well do it by the book.
Reported by Rich Felker and Eric Blake; see
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00493.html>.
* m4/mktime.m4 (AC_FUNC_MKTIME): Likewise.
* lib/mktime.c (TYPE_MAXIMUM): Redo slightly to match the others.
mktime: #undef mktime before #defining it
* lib/mktime.c (mktime) [DEBUG]: #undef mktime before #defining it.
mktime: systematically normalize tm_isdst comparisons
* lib/mktime.c (isdst_differ): New function.
(__mktime_internal): Use it systematically for all isdst comparisons.
This completes the fix for libc BZ #6723, and removes the need for
normalizing tm_isdst. See
<http://sourceware.org/bugzilla/show_bug.cgi?id=6723>
(not_equal_tm) [DEBUG]: Use isdst_differ here, too.
mktime: fix some integer overflow issues and sidestep the rest
This was prompted by a bug report by Benjamin Lindner for MinGW
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00472.html>.
His bug is due to signed integer overflow (0 - INT_MIN), and I
I scanned through mktime.c looking for other integer overflow
problems, fixing all the bugs I found.
Although the C Standard says the resulting code is still not safe
in the presence of integer overflow, in practice it should be good
enough for all real-world two's-complement implementations, except
for debugging environments that deliberately trap on integer
overflow (e.g., gcc -ftrapv).
* lib/mktime.c (WRAPV): New macro.
(SHR): Also check that long_int and time_t shift right in the
usual way, before using the fast-but-unportable method.
(TYPE_ONES_COMPLEMENT, TYPE_SIGNED_MAGNITUDE): Remove, no longer
used. The code already assumed two's complement, so there's
no need to test for alternatives. All uses removed.
(TYPE_MAXIMUM): Don't rely here on overflow behavior not defined by
the C standard. Problem reported by Rich Felker in
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00488.html>.
(twos_complement_arithmetic): Also check long_int and time_t.
(time_t_avg, time_t_add_ok, time_t_int_add_ok): New functions.
(guess_time_tm, ranged_convert, __mktime_internal): Use them.
(__mktime_internal): Avoid integer overflow with unary subtraction
in two instances where -1 - X is an adequate replacement for -X,
since the calculations are approximate.
2011-01-29 Eric Blake <eblake@redhat.com>
mktime: avoid infinite loop
* m4/mktime.m4 (AC_FUNC_MKTIME): Avoid overflow on possibly-signed
type; behavior is still undefined but portable to all known targets.
Reported by Rich Felker.
2011-01-28 Paul Eggert <eggert@cs.ucla.edu>
mktime: avoid problems on NetBSD 5 / i386
* lib/mktime.c (long_int): New type. This works around a problem
on NetBSD 5 / i386, where 'long int' and 'int' are both 32 bits
but time_t is 64 bits, and where I expect the existing code is
wrong in some cases.
(leapyear, ydhms_diff, guess_time_tm, __mktime_internal): Use it.
(ydhms_diff): Bring back the compile-time check for wide-enough
year and yday.
mktime: fix misspelling in comment
* lib/mktime.c (__mktime_internal): Fix misspelling in comment.
This merges all recent glibc changes of importance.
2011-01-28 Ralf Wildenhues <Ralf.Wildenhues@gmx.de>
move-if-change: cope with concurrent mv of identical file.
* move-if-change (CMPPROG): Accept environment
variable as an override for `cmp'.
(usage): Document CMPPROG.
Adjust comparison to drop stdout. Cope with failure of mv if
the target file exists and is identical to the source, for
parallel builds.
Report from H.J. Lu against binutils in PR binutils/12283.
2011-01-30 19:22:02 +00:00
|
|
|
|
ydhms_diff (long_int year1, long_int yday1, int hour1, int min1, int sec1,
|
2012-05-26 23:14:36 +00:00
|
|
|
|
int year0, int yday0, int hour0, int min0, int sec0)
|
2011-01-09 06:57:07 +00:00
|
|
|
|
{
|
|
|
|
|
verify (C99_integer_division, -1 / 2 == 0);
|
|
|
|
|
|
|
|
|
|
/* Compute intervening leap days correctly even if year is negative.
|
|
|
|
|
Take care to avoid integer overflow here. */
|
|
|
|
|
int a4 = SHR (year1, 2) + SHR (TM_YEAR_BASE, 2) - ! (year1 & 3);
|
|
|
|
|
int b4 = SHR (year0, 2) + SHR (TM_YEAR_BASE, 2) - ! (year0 & 3);
|
|
|
|
|
int a100 = a4 / 25 - (a4 % 25 < 0);
|
|
|
|
|
int b100 = b4 / 25 - (b4 % 25 < 0);
|
|
|
|
|
int a400 = SHR (a100, 2);
|
|
|
|
|
int b400 = SHR (b100, 2);
|
|
|
|
|
int intervening_leap_days = (a4 - b4) - (a100 - b100) + (a400 - b400);
|
|
|
|
|
|
|
|
|
|
/* Compute the desired time in time_t precision. Overflow might
|
|
|
|
|
occur here. */
|
|
|
|
|
time_t tyear1 = year1;
|
|
|
|
|
time_t years = tyear1 - year0;
|
|
|
|
|
time_t days = 365 * years + yday1 - yday0 + intervening_leap_days;
|
|
|
|
|
time_t hours = 24 * days + hour1 - hour0;
|
|
|
|
|
time_t minutes = 60 * hours + min1 - min0;
|
|
|
|
|
time_t seconds = 60 * minutes + sec1 - sec0;
|
|
|
|
|
return seconds;
|
|
|
|
|
}
|
|
|
|
|
|
gnulib: import mktime and move-if-change fixes from gnulib
* configure: Regenerate from the following.
2011-01-30 Paul Eggert <eggert@cs.ucla.edu>
mktime: clarify long_int width checking
* lib/mktime.c (long_int_is_wide_enough): Move this assertion to
the top level, to make it clearer that the assumption about
long_int width is being checked. See
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00554.html>.
2011-01-29 Paul Eggert <eggert@cs.ucla.edu>
TYPE_MAXIMUM: avoid theoretically undefined behavior
* lib/intprops.h (TYPE_MINIMUM, TYPE_MAXIMUM): Do not shift a
negative number, which the C Standard says has undefined behavior.
In practice this is not a problem, but might as well do it by the book.
Reported by Rich Felker and Eric Blake; see
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00493.html>.
* m4/mktime.m4 (AC_FUNC_MKTIME): Likewise.
* lib/mktime.c (TYPE_MAXIMUM): Redo slightly to match the others.
mktime: #undef mktime before #defining it
* lib/mktime.c (mktime) [DEBUG]: #undef mktime before #defining it.
mktime: systematically normalize tm_isdst comparisons
* lib/mktime.c (isdst_differ): New function.
(__mktime_internal): Use it systematically for all isdst comparisons.
This completes the fix for libc BZ #6723, and removes the need for
normalizing tm_isdst. See
<http://sourceware.org/bugzilla/show_bug.cgi?id=6723>
(not_equal_tm) [DEBUG]: Use isdst_differ here, too.
mktime: fix some integer overflow issues and sidestep the rest
This was prompted by a bug report by Benjamin Lindner for MinGW
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00472.html>.
His bug is due to signed integer overflow (0 - INT_MIN), and I
I scanned through mktime.c looking for other integer overflow
problems, fixing all the bugs I found.
Although the C Standard says the resulting code is still not safe
in the presence of integer overflow, in practice it should be good
enough for all real-world two's-complement implementations, except
for debugging environments that deliberately trap on integer
overflow (e.g., gcc -ftrapv).
* lib/mktime.c (WRAPV): New macro.
(SHR): Also check that long_int and time_t shift right in the
usual way, before using the fast-but-unportable method.
(TYPE_ONES_COMPLEMENT, TYPE_SIGNED_MAGNITUDE): Remove, no longer
used. The code already assumed two's complement, so there's
no need to test for alternatives. All uses removed.
(TYPE_MAXIMUM): Don't rely here on overflow behavior not defined by
the C standard. Problem reported by Rich Felker in
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00488.html>.
(twos_complement_arithmetic): Also check long_int and time_t.
(time_t_avg, time_t_add_ok, time_t_int_add_ok): New functions.
(guess_time_tm, ranged_convert, __mktime_internal): Use them.
(__mktime_internal): Avoid integer overflow with unary subtraction
in two instances where -1 - X is an adequate replacement for -X,
since the calculations are approximate.
2011-01-29 Eric Blake <eblake@redhat.com>
mktime: avoid infinite loop
* m4/mktime.m4 (AC_FUNC_MKTIME): Avoid overflow on possibly-signed
type; behavior is still undefined but portable to all known targets.
Reported by Rich Felker.
2011-01-28 Paul Eggert <eggert@cs.ucla.edu>
mktime: avoid problems on NetBSD 5 / i386
* lib/mktime.c (long_int): New type. This works around a problem
on NetBSD 5 / i386, where 'long int' and 'int' are both 32 bits
but time_t is 64 bits, and where I expect the existing code is
wrong in some cases.
(leapyear, ydhms_diff, guess_time_tm, __mktime_internal): Use it.
(ydhms_diff): Bring back the compile-time check for wide-enough
year and yday.
mktime: fix misspelling in comment
* lib/mktime.c (__mktime_internal): Fix misspelling in comment.
This merges all recent glibc changes of importance.
2011-01-28 Ralf Wildenhues <Ralf.Wildenhues@gmx.de>
move-if-change: cope with concurrent mv of identical file.
* move-if-change (CMPPROG): Accept environment
variable as an override for `cmp'.
(usage): Document CMPPROG.
Adjust comparison to drop stdout. Cope with failure of mv if
the target file exists and is identical to the source, for
parallel builds.
Report from H.J. Lu against binutils in PR binutils/12283.
2011-01-30 19:22:02 +00:00
|
|
|
|
/* Return the average of A and B, even if A + B would overflow. */
|
|
|
|
|
static time_t
|
|
|
|
|
time_t_avg (time_t a, time_t b)
|
|
|
|
|
{
|
|
|
|
|
return SHR (a, 1) + SHR (b, 1) + (a & b & 1);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Return 1 if A + B does not overflow. If time_t is unsigned and if
|
|
|
|
|
B's top bit is set, assume that the sum represents A - -B, and
|
|
|
|
|
return 1 if the subtraction does not wrap around. */
|
|
|
|
|
static int
|
|
|
|
|
time_t_add_ok (time_t a, time_t b)
|
|
|
|
|
{
|
|
|
|
|
if (! TYPE_SIGNED (time_t))
|
|
|
|
|
{
|
|
|
|
|
time_t sum = a + b;
|
|
|
|
|
return (sum < a) == (TIME_T_MIDPOINT <= b);
|
|
|
|
|
}
|
|
|
|
|
else if (WRAPV)
|
|
|
|
|
{
|
|
|
|
|
time_t sum = a + b;
|
|
|
|
|
return (sum < a) == (b < 0);
|
|
|
|
|
}
|
|
|
|
|
else
|
|
|
|
|
{
|
|
|
|
|
time_t avg = time_t_avg (a, b);
|
|
|
|
|
return TIME_T_MIN / 2 <= avg && avg <= TIME_T_MAX / 2;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Return 1 if A + B does not overflow. */
|
|
|
|
|
static int
|
|
|
|
|
time_t_int_add_ok (time_t a, int b)
|
|
|
|
|
{
|
|
|
|
|
verify (int_no_wider_than_time_t, INT_MAX <= TIME_T_MAX);
|
|
|
|
|
if (WRAPV)
|
|
|
|
|
{
|
|
|
|
|
time_t sum = a + b;
|
|
|
|
|
return (sum < a) == (b < 0);
|
|
|
|
|
}
|
|
|
|
|
else
|
|
|
|
|
{
|
|
|
|
|
int a_odd = a & 1;
|
|
|
|
|
time_t avg = SHR (a, 1) + (SHR (b, 1) + (a_odd & b));
|
|
|
|
|
return TIME_T_MIN / 2 <= avg && avg <= TIME_T_MAX / 2;
|
|
|
|
|
}
|
|
|
|
|
}
|
2011-01-09 06:57:07 +00:00
|
|
|
|
|
|
|
|
|
/* Return a time_t value corresponding to (YEAR-YDAY HOUR:MIN:SEC),
|
|
|
|
|
assuming that *T corresponds to *TP and that no clock adjustments
|
|
|
|
|
occurred between *TP and the desired time.
|
|
|
|
|
If TP is null, return a value not equal to *T; this avoids false matches.
|
|
|
|
|
If overflow occurs, yield the minimal or maximal value, except do not
|
|
|
|
|
yield a value equal to *T. */
|
|
|
|
|
static time_t
|
gnulib: import mktime and move-if-change fixes from gnulib
* configure: Regenerate from the following.
2011-01-30 Paul Eggert <eggert@cs.ucla.edu>
mktime: clarify long_int width checking
* lib/mktime.c (long_int_is_wide_enough): Move this assertion to
the top level, to make it clearer that the assumption about
long_int width is being checked. See
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00554.html>.
2011-01-29 Paul Eggert <eggert@cs.ucla.edu>
TYPE_MAXIMUM: avoid theoretically undefined behavior
* lib/intprops.h (TYPE_MINIMUM, TYPE_MAXIMUM): Do not shift a
negative number, which the C Standard says has undefined behavior.
In practice this is not a problem, but might as well do it by the book.
Reported by Rich Felker and Eric Blake; see
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00493.html>.
* m4/mktime.m4 (AC_FUNC_MKTIME): Likewise.
* lib/mktime.c (TYPE_MAXIMUM): Redo slightly to match the others.
mktime: #undef mktime before #defining it
* lib/mktime.c (mktime) [DEBUG]: #undef mktime before #defining it.
mktime: systematically normalize tm_isdst comparisons
* lib/mktime.c (isdst_differ): New function.
(__mktime_internal): Use it systematically for all isdst comparisons.
This completes the fix for libc BZ #6723, and removes the need for
normalizing tm_isdst. See
<http://sourceware.org/bugzilla/show_bug.cgi?id=6723>
(not_equal_tm) [DEBUG]: Use isdst_differ here, too.
mktime: fix some integer overflow issues and sidestep the rest
This was prompted by a bug report by Benjamin Lindner for MinGW
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00472.html>.
His bug is due to signed integer overflow (0 - INT_MIN), and I
I scanned through mktime.c looking for other integer overflow
problems, fixing all the bugs I found.
Although the C Standard says the resulting code is still not safe
in the presence of integer overflow, in practice it should be good
enough for all real-world two's-complement implementations, except
for debugging environments that deliberately trap on integer
overflow (e.g., gcc -ftrapv).
* lib/mktime.c (WRAPV): New macro.
(SHR): Also check that long_int and time_t shift right in the
usual way, before using the fast-but-unportable method.
(TYPE_ONES_COMPLEMENT, TYPE_SIGNED_MAGNITUDE): Remove, no longer
used. The code already assumed two's complement, so there's
no need to test for alternatives. All uses removed.
(TYPE_MAXIMUM): Don't rely here on overflow behavior not defined by
the C standard. Problem reported by Rich Felker in
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00488.html>.
(twos_complement_arithmetic): Also check long_int and time_t.
(time_t_avg, time_t_add_ok, time_t_int_add_ok): New functions.
(guess_time_tm, ranged_convert, __mktime_internal): Use them.
(__mktime_internal): Avoid integer overflow with unary subtraction
in two instances where -1 - X is an adequate replacement for -X,
since the calculations are approximate.
2011-01-29 Eric Blake <eblake@redhat.com>
mktime: avoid infinite loop
* m4/mktime.m4 (AC_FUNC_MKTIME): Avoid overflow on possibly-signed
type; behavior is still undefined but portable to all known targets.
Reported by Rich Felker.
2011-01-28 Paul Eggert <eggert@cs.ucla.edu>
mktime: avoid problems on NetBSD 5 / i386
* lib/mktime.c (long_int): New type. This works around a problem
on NetBSD 5 / i386, where 'long int' and 'int' are both 32 bits
but time_t is 64 bits, and where I expect the existing code is
wrong in some cases.
(leapyear, ydhms_diff, guess_time_tm, __mktime_internal): Use it.
(ydhms_diff): Bring back the compile-time check for wide-enough
year and yday.
mktime: fix misspelling in comment
* lib/mktime.c (__mktime_internal): Fix misspelling in comment.
This merges all recent glibc changes of importance.
2011-01-28 Ralf Wildenhues <Ralf.Wildenhues@gmx.de>
move-if-change: cope with concurrent mv of identical file.
* move-if-change (CMPPROG): Accept environment
variable as an override for `cmp'.
(usage): Document CMPPROG.
Adjust comparison to drop stdout. Cope with failure of mv if
the target file exists and is identical to the source, for
parallel builds.
Report from H.J. Lu against binutils in PR binutils/12283.
2011-01-30 19:22:02 +00:00
|
|
|
|
guess_time_tm (long_int year, long_int yday, int hour, int min, int sec,
|
2012-05-26 23:14:36 +00:00
|
|
|
|
const time_t *t, const struct tm *tp)
|
2011-01-09 06:57:07 +00:00
|
|
|
|
{
|
|
|
|
|
if (tp)
|
|
|
|
|
{
|
|
|
|
|
time_t d = ydhms_diff (year, yday, hour, min, sec,
|
2012-05-26 23:14:36 +00:00
|
|
|
|
tp->tm_year, tp->tm_yday,
|
|
|
|
|
tp->tm_hour, tp->tm_min, tp->tm_sec);
|
gnulib: import mktime and move-if-change fixes from gnulib
* configure: Regenerate from the following.
2011-01-30 Paul Eggert <eggert@cs.ucla.edu>
mktime: clarify long_int width checking
* lib/mktime.c (long_int_is_wide_enough): Move this assertion to
the top level, to make it clearer that the assumption about
long_int width is being checked. See
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00554.html>.
2011-01-29 Paul Eggert <eggert@cs.ucla.edu>
TYPE_MAXIMUM: avoid theoretically undefined behavior
* lib/intprops.h (TYPE_MINIMUM, TYPE_MAXIMUM): Do not shift a
negative number, which the C Standard says has undefined behavior.
In practice this is not a problem, but might as well do it by the book.
Reported by Rich Felker and Eric Blake; see
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00493.html>.
* m4/mktime.m4 (AC_FUNC_MKTIME): Likewise.
* lib/mktime.c (TYPE_MAXIMUM): Redo slightly to match the others.
mktime: #undef mktime before #defining it
* lib/mktime.c (mktime) [DEBUG]: #undef mktime before #defining it.
mktime: systematically normalize tm_isdst comparisons
* lib/mktime.c (isdst_differ): New function.
(__mktime_internal): Use it systematically for all isdst comparisons.
This completes the fix for libc BZ #6723, and removes the need for
normalizing tm_isdst. See
<http://sourceware.org/bugzilla/show_bug.cgi?id=6723>
(not_equal_tm) [DEBUG]: Use isdst_differ here, too.
mktime: fix some integer overflow issues and sidestep the rest
This was prompted by a bug report by Benjamin Lindner for MinGW
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00472.html>.
His bug is due to signed integer overflow (0 - INT_MIN), and I
I scanned through mktime.c looking for other integer overflow
problems, fixing all the bugs I found.
Although the C Standard says the resulting code is still not safe
in the presence of integer overflow, in practice it should be good
enough for all real-world two's-complement implementations, except
for debugging environments that deliberately trap on integer
overflow (e.g., gcc -ftrapv).
* lib/mktime.c (WRAPV): New macro.
(SHR): Also check that long_int and time_t shift right in the
usual way, before using the fast-but-unportable method.
(TYPE_ONES_COMPLEMENT, TYPE_SIGNED_MAGNITUDE): Remove, no longer
used. The code already assumed two's complement, so there's
no need to test for alternatives. All uses removed.
(TYPE_MAXIMUM): Don't rely here on overflow behavior not defined by
the C standard. Problem reported by Rich Felker in
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00488.html>.
(twos_complement_arithmetic): Also check long_int and time_t.
(time_t_avg, time_t_add_ok, time_t_int_add_ok): New functions.
(guess_time_tm, ranged_convert, __mktime_internal): Use them.
(__mktime_internal): Avoid integer overflow with unary subtraction
in two instances where -1 - X is an adequate replacement for -X,
since the calculations are approximate.
2011-01-29 Eric Blake <eblake@redhat.com>
mktime: avoid infinite loop
* m4/mktime.m4 (AC_FUNC_MKTIME): Avoid overflow on possibly-signed
type; behavior is still undefined but portable to all known targets.
Reported by Rich Felker.
2011-01-28 Paul Eggert <eggert@cs.ucla.edu>
mktime: avoid problems on NetBSD 5 / i386
* lib/mktime.c (long_int): New type. This works around a problem
on NetBSD 5 / i386, where 'long int' and 'int' are both 32 bits
but time_t is 64 bits, and where I expect the existing code is
wrong in some cases.
(leapyear, ydhms_diff, guess_time_tm, __mktime_internal): Use it.
(ydhms_diff): Bring back the compile-time check for wide-enough
year and yday.
mktime: fix misspelling in comment
* lib/mktime.c (__mktime_internal): Fix misspelling in comment.
This merges all recent glibc changes of importance.
2011-01-28 Ralf Wildenhues <Ralf.Wildenhues@gmx.de>
move-if-change: cope with concurrent mv of identical file.
* move-if-change (CMPPROG): Accept environment
variable as an override for `cmp'.
(usage): Document CMPPROG.
Adjust comparison to drop stdout. Cope with failure of mv if
the target file exists and is identical to the source, for
parallel builds.
Report from H.J. Lu against binutils in PR binutils/12283.
2011-01-30 19:22:02 +00:00
|
|
|
|
if (time_t_add_ok (*t, d))
|
2012-05-26 23:14:36 +00:00
|
|
|
|
return *t + d;
|
2011-01-09 06:57:07 +00:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Overflow occurred one way or another. Return the nearest result
|
|
|
|
|
that is actually in range, except don't report a zero difference
|
|
|
|
|
if the actual difference is nonzero, as that would cause a false
|
|
|
|
|
match; and don't oscillate between two values, as that would
|
|
|
|
|
confuse the spring-forward gap detector. */
|
|
|
|
|
return (*t < TIME_T_MIDPOINT
|
2012-05-26 23:14:36 +00:00
|
|
|
|
? (*t <= TIME_T_MIN + 1 ? *t + 1 : TIME_T_MIN)
|
|
|
|
|
: (TIME_T_MAX - 1 <= *t ? *t - 1 : TIME_T_MAX));
|
2011-01-09 06:57:07 +00:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Use CONVERT to convert *T to a broken down time in *TP.
|
|
|
|
|
If *T is out of range for conversion, adjust it so that
|
|
|
|
|
it is the nearest in-range value and then convert that. */
|
|
|
|
|
static struct tm *
|
|
|
|
|
ranged_convert (struct tm *(*convert) (const time_t *, struct tm *),
|
2012-05-26 23:14:36 +00:00
|
|
|
|
time_t *t, struct tm *tp)
|
2011-01-09 06:57:07 +00:00
|
|
|
|
{
|
|
|
|
|
struct tm *r = convert (t, tp);
|
|
|
|
|
|
|
|
|
|
if (!r && *t)
|
|
|
|
|
{
|
|
|
|
|
time_t bad = *t;
|
|
|
|
|
time_t ok = 0;
|
|
|
|
|
|
|
|
|
|
/* BAD is a known unconvertible time_t, and OK is a known good one.
|
2012-05-26 23:14:36 +00:00
|
|
|
|
Use binary search to narrow the range between BAD and OK until
|
|
|
|
|
they differ by 1. */
|
2011-01-09 06:57:07 +00:00
|
|
|
|
while (bad != ok + (bad < 0 ? -1 : 1))
|
2012-05-26 23:14:36 +00:00
|
|
|
|
{
|
|
|
|
|
time_t mid = *t = time_t_avg (ok, bad);
|
|
|
|
|
r = convert (t, tp);
|
|
|
|
|
if (r)
|
|
|
|
|
ok = mid;
|
|
|
|
|
else
|
|
|
|
|
bad = mid;
|
|
|
|
|
}
|
2011-01-09 06:57:07 +00:00
|
|
|
|
|
|
|
|
|
if (!r && ok)
|
2012-05-26 23:14:36 +00:00
|
|
|
|
{
|
|
|
|
|
/* The last conversion attempt failed;
|
|
|
|
|
revert to the most recent successful attempt. */
|
|
|
|
|
*t = ok;
|
|
|
|
|
r = convert (t, tp);
|
|
|
|
|
}
|
2011-01-09 06:57:07 +00:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return r;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
/* Convert *TP to a time_t value, inverting
|
|
|
|
|
the monotonic and mostly-unit-linear conversion function CONVERT.
|
|
|
|
|
Use *OFFSET to keep track of a guess at the offset of the result,
|
|
|
|
|
compared to what the result would be for UTC without leap seconds.
|
|
|
|
|
If *OFFSET's guess is correct, only one CONVERT call is needed.
|
|
|
|
|
This function is external because it is used also by timegm.c. */
|
|
|
|
|
time_t
|
|
|
|
|
__mktime_internal (struct tm *tp,
|
2012-05-26 23:14:36 +00:00
|
|
|
|
struct tm *(*convert) (const time_t *, struct tm *),
|
|
|
|
|
time_t *offset)
|
2011-01-09 06:57:07 +00:00
|
|
|
|
{
|
|
|
|
|
time_t t, gt, t0, t1, t2;
|
|
|
|
|
struct tm tm;
|
|
|
|
|
|
|
|
|
|
/* The maximum number of probes (calls to CONVERT) should be enough
|
|
|
|
|
to handle any combinations of time zone rule changes, solar time,
|
|
|
|
|
leap seconds, and oscillations around a spring-forward gap.
|
|
|
|
|
POSIX.1 prohibits leap seconds, but some hosts have them anyway. */
|
|
|
|
|
int remaining_probes = 6;
|
|
|
|
|
|
|
|
|
|
/* Time requested. Copy it in case CONVERT modifies *TP; this can
|
|
|
|
|
occur if TP is localtime's returned value and CONVERT is localtime. */
|
|
|
|
|
int sec = tp->tm_sec;
|
|
|
|
|
int min = tp->tm_min;
|
|
|
|
|
int hour = tp->tm_hour;
|
|
|
|
|
int mday = tp->tm_mday;
|
|
|
|
|
int mon = tp->tm_mon;
|
|
|
|
|
int year_requested = tp->tm_year;
|
gnulib: import mktime and move-if-change fixes from gnulib
* configure: Regenerate from the following.
2011-01-30 Paul Eggert <eggert@cs.ucla.edu>
mktime: clarify long_int width checking
* lib/mktime.c (long_int_is_wide_enough): Move this assertion to
the top level, to make it clearer that the assumption about
long_int width is being checked. See
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00554.html>.
2011-01-29 Paul Eggert <eggert@cs.ucla.edu>
TYPE_MAXIMUM: avoid theoretically undefined behavior
* lib/intprops.h (TYPE_MINIMUM, TYPE_MAXIMUM): Do not shift a
negative number, which the C Standard says has undefined behavior.
In practice this is not a problem, but might as well do it by the book.
Reported by Rich Felker and Eric Blake; see
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00493.html>.
* m4/mktime.m4 (AC_FUNC_MKTIME): Likewise.
* lib/mktime.c (TYPE_MAXIMUM): Redo slightly to match the others.
mktime: #undef mktime before #defining it
* lib/mktime.c (mktime) [DEBUG]: #undef mktime before #defining it.
mktime: systematically normalize tm_isdst comparisons
* lib/mktime.c (isdst_differ): New function.
(__mktime_internal): Use it systematically for all isdst comparisons.
This completes the fix for libc BZ #6723, and removes the need for
normalizing tm_isdst. See
<http://sourceware.org/bugzilla/show_bug.cgi?id=6723>
(not_equal_tm) [DEBUG]: Use isdst_differ here, too.
mktime: fix some integer overflow issues and sidestep the rest
This was prompted by a bug report by Benjamin Lindner for MinGW
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00472.html>.
His bug is due to signed integer overflow (0 - INT_MIN), and I
I scanned through mktime.c looking for other integer overflow
problems, fixing all the bugs I found.
Although the C Standard says the resulting code is still not safe
in the presence of integer overflow, in practice it should be good
enough for all real-world two's-complement implementations, except
for debugging environments that deliberately trap on integer
overflow (e.g., gcc -ftrapv).
* lib/mktime.c (WRAPV): New macro.
(SHR): Also check that long_int and time_t shift right in the
usual way, before using the fast-but-unportable method.
(TYPE_ONES_COMPLEMENT, TYPE_SIGNED_MAGNITUDE): Remove, no longer
used. The code already assumed two's complement, so there's
no need to test for alternatives. All uses removed.
(TYPE_MAXIMUM): Don't rely here on overflow behavior not defined by
the C standard. Problem reported by Rich Felker in
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00488.html>.
(twos_complement_arithmetic): Also check long_int and time_t.
(time_t_avg, time_t_add_ok, time_t_int_add_ok): New functions.
(guess_time_tm, ranged_convert, __mktime_internal): Use them.
(__mktime_internal): Avoid integer overflow with unary subtraction
in two instances where -1 - X is an adequate replacement for -X,
since the calculations are approximate.
2011-01-29 Eric Blake <eblake@redhat.com>
mktime: avoid infinite loop
* m4/mktime.m4 (AC_FUNC_MKTIME): Avoid overflow on possibly-signed
type; behavior is still undefined but portable to all known targets.
Reported by Rich Felker.
2011-01-28 Paul Eggert <eggert@cs.ucla.edu>
mktime: avoid problems on NetBSD 5 / i386
* lib/mktime.c (long_int): New type. This works around a problem
on NetBSD 5 / i386, where 'long int' and 'int' are both 32 bits
but time_t is 64 bits, and where I expect the existing code is
wrong in some cases.
(leapyear, ydhms_diff, guess_time_tm, __mktime_internal): Use it.
(ydhms_diff): Bring back the compile-time check for wide-enough
year and yday.
mktime: fix misspelling in comment
* lib/mktime.c (__mktime_internal): Fix misspelling in comment.
This merges all recent glibc changes of importance.
2011-01-28 Ralf Wildenhues <Ralf.Wildenhues@gmx.de>
move-if-change: cope with concurrent mv of identical file.
* move-if-change (CMPPROG): Accept environment
variable as an override for `cmp'.
(usage): Document CMPPROG.
Adjust comparison to drop stdout. Cope with failure of mv if
the target file exists and is identical to the source, for
parallel builds.
Report from H.J. Lu against binutils in PR binutils/12283.
2011-01-30 19:22:02 +00:00
|
|
|
|
int isdst = tp->tm_isdst;
|
2011-01-09 06:57:07 +00:00
|
|
|
|
|
|
|
|
|
/* 1 if the previous probe was DST. */
|
|
|
|
|
int dst2;
|
|
|
|
|
|
|
|
|
|
/* Ensure that mon is in range, and set year accordingly. */
|
|
|
|
|
int mon_remainder = mon % 12;
|
|
|
|
|
int negative_mon_remainder = mon_remainder < 0;
|
|
|
|
|
int mon_years = mon / 12 - negative_mon_remainder;
|
gnulib: import mktime and move-if-change fixes from gnulib
* configure: Regenerate from the following.
2011-01-30 Paul Eggert <eggert@cs.ucla.edu>
mktime: clarify long_int width checking
* lib/mktime.c (long_int_is_wide_enough): Move this assertion to
the top level, to make it clearer that the assumption about
long_int width is being checked. See
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00554.html>.
2011-01-29 Paul Eggert <eggert@cs.ucla.edu>
TYPE_MAXIMUM: avoid theoretically undefined behavior
* lib/intprops.h (TYPE_MINIMUM, TYPE_MAXIMUM): Do not shift a
negative number, which the C Standard says has undefined behavior.
In practice this is not a problem, but might as well do it by the book.
Reported by Rich Felker and Eric Blake; see
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00493.html>.
* m4/mktime.m4 (AC_FUNC_MKTIME): Likewise.
* lib/mktime.c (TYPE_MAXIMUM): Redo slightly to match the others.
mktime: #undef mktime before #defining it
* lib/mktime.c (mktime) [DEBUG]: #undef mktime before #defining it.
mktime: systematically normalize tm_isdst comparisons
* lib/mktime.c (isdst_differ): New function.
(__mktime_internal): Use it systematically for all isdst comparisons.
This completes the fix for libc BZ #6723, and removes the need for
normalizing tm_isdst. See
<http://sourceware.org/bugzilla/show_bug.cgi?id=6723>
(not_equal_tm) [DEBUG]: Use isdst_differ here, too.
mktime: fix some integer overflow issues and sidestep the rest
This was prompted by a bug report by Benjamin Lindner for MinGW
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00472.html>.
His bug is due to signed integer overflow (0 - INT_MIN), and I
I scanned through mktime.c looking for other integer overflow
problems, fixing all the bugs I found.
Although the C Standard says the resulting code is still not safe
in the presence of integer overflow, in practice it should be good
enough for all real-world two's-complement implementations, except
for debugging environments that deliberately trap on integer
overflow (e.g., gcc -ftrapv).
* lib/mktime.c (WRAPV): New macro.
(SHR): Also check that long_int and time_t shift right in the
usual way, before using the fast-but-unportable method.
(TYPE_ONES_COMPLEMENT, TYPE_SIGNED_MAGNITUDE): Remove, no longer
used. The code already assumed two's complement, so there's
no need to test for alternatives. All uses removed.
(TYPE_MAXIMUM): Don't rely here on overflow behavior not defined by
the C standard. Problem reported by Rich Felker in
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00488.html>.
(twos_complement_arithmetic): Also check long_int and time_t.
(time_t_avg, time_t_add_ok, time_t_int_add_ok): New functions.
(guess_time_tm, ranged_convert, __mktime_internal): Use them.
(__mktime_internal): Avoid integer overflow with unary subtraction
in two instances where -1 - X is an adequate replacement for -X,
since the calculations are approximate.
2011-01-29 Eric Blake <eblake@redhat.com>
mktime: avoid infinite loop
* m4/mktime.m4 (AC_FUNC_MKTIME): Avoid overflow on possibly-signed
type; behavior is still undefined but portable to all known targets.
Reported by Rich Felker.
2011-01-28 Paul Eggert <eggert@cs.ucla.edu>
mktime: avoid problems on NetBSD 5 / i386
* lib/mktime.c (long_int): New type. This works around a problem
on NetBSD 5 / i386, where 'long int' and 'int' are both 32 bits
but time_t is 64 bits, and where I expect the existing code is
wrong in some cases.
(leapyear, ydhms_diff, guess_time_tm, __mktime_internal): Use it.
(ydhms_diff): Bring back the compile-time check for wide-enough
year and yday.
mktime: fix misspelling in comment
* lib/mktime.c (__mktime_internal): Fix misspelling in comment.
This merges all recent glibc changes of importance.
2011-01-28 Ralf Wildenhues <Ralf.Wildenhues@gmx.de>
move-if-change: cope with concurrent mv of identical file.
* move-if-change (CMPPROG): Accept environment
variable as an override for `cmp'.
(usage): Document CMPPROG.
Adjust comparison to drop stdout. Cope with failure of mv if
the target file exists and is identical to the source, for
parallel builds.
Report from H.J. Lu against binutils in PR binutils/12283.
2011-01-30 19:22:02 +00:00
|
|
|
|
long_int lyear_requested = year_requested;
|
|
|
|
|
long_int year = lyear_requested + mon_years;
|
2011-01-09 06:57:07 +00:00
|
|
|
|
|
|
|
|
|
/* The other values need not be in range:
|
|
|
|
|
the remaining code handles minor overflows correctly,
|
|
|
|
|
assuming int and time_t arithmetic wraps around.
|
|
|
|
|
Major overflows are caught at the end. */
|
|
|
|
|
|
|
|
|
|
/* Calculate day of year from year, month, and day of month.
|
|
|
|
|
The result need not be in range. */
|
|
|
|
|
int mon_yday = ((__mon_yday[leapyear (year)]
|
2012-05-26 23:14:36 +00:00
|
|
|
|
[mon_remainder + 12 * negative_mon_remainder])
|
|
|
|
|
- 1);
|
gnulib: import mktime and move-if-change fixes from gnulib
* configure: Regenerate from the following.
2011-01-30 Paul Eggert <eggert@cs.ucla.edu>
mktime: clarify long_int width checking
* lib/mktime.c (long_int_is_wide_enough): Move this assertion to
the top level, to make it clearer that the assumption about
long_int width is being checked. See
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00554.html>.
2011-01-29 Paul Eggert <eggert@cs.ucla.edu>
TYPE_MAXIMUM: avoid theoretically undefined behavior
* lib/intprops.h (TYPE_MINIMUM, TYPE_MAXIMUM): Do not shift a
negative number, which the C Standard says has undefined behavior.
In practice this is not a problem, but might as well do it by the book.
Reported by Rich Felker and Eric Blake; see
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00493.html>.
* m4/mktime.m4 (AC_FUNC_MKTIME): Likewise.
* lib/mktime.c (TYPE_MAXIMUM): Redo slightly to match the others.
mktime: #undef mktime before #defining it
* lib/mktime.c (mktime) [DEBUG]: #undef mktime before #defining it.
mktime: systematically normalize tm_isdst comparisons
* lib/mktime.c (isdst_differ): New function.
(__mktime_internal): Use it systematically for all isdst comparisons.
This completes the fix for libc BZ #6723, and removes the need for
normalizing tm_isdst. See
<http://sourceware.org/bugzilla/show_bug.cgi?id=6723>
(not_equal_tm) [DEBUG]: Use isdst_differ here, too.
mktime: fix some integer overflow issues and sidestep the rest
This was prompted by a bug report by Benjamin Lindner for MinGW
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00472.html>.
His bug is due to signed integer overflow (0 - INT_MIN), and I
I scanned through mktime.c looking for other integer overflow
problems, fixing all the bugs I found.
Although the C Standard says the resulting code is still not safe
in the presence of integer overflow, in practice it should be good
enough for all real-world two's-complement implementations, except
for debugging environments that deliberately trap on integer
overflow (e.g., gcc -ftrapv).
* lib/mktime.c (WRAPV): New macro.
(SHR): Also check that long_int and time_t shift right in the
usual way, before using the fast-but-unportable method.
(TYPE_ONES_COMPLEMENT, TYPE_SIGNED_MAGNITUDE): Remove, no longer
used. The code already assumed two's complement, so there's
no need to test for alternatives. All uses removed.
(TYPE_MAXIMUM): Don't rely here on overflow behavior not defined by
the C standard. Problem reported by Rich Felker in
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00488.html>.
(twos_complement_arithmetic): Also check long_int and time_t.
(time_t_avg, time_t_add_ok, time_t_int_add_ok): New functions.
(guess_time_tm, ranged_convert, __mktime_internal): Use them.
(__mktime_internal): Avoid integer overflow with unary subtraction
in two instances where -1 - X is an adequate replacement for -X,
since the calculations are approximate.
2011-01-29 Eric Blake <eblake@redhat.com>
mktime: avoid infinite loop
* m4/mktime.m4 (AC_FUNC_MKTIME): Avoid overflow on possibly-signed
type; behavior is still undefined but portable to all known targets.
Reported by Rich Felker.
2011-01-28 Paul Eggert <eggert@cs.ucla.edu>
mktime: avoid problems on NetBSD 5 / i386
* lib/mktime.c (long_int): New type. This works around a problem
on NetBSD 5 / i386, where 'long int' and 'int' are both 32 bits
but time_t is 64 bits, and where I expect the existing code is
wrong in some cases.
(leapyear, ydhms_diff, guess_time_tm, __mktime_internal): Use it.
(ydhms_diff): Bring back the compile-time check for wide-enough
year and yday.
mktime: fix misspelling in comment
* lib/mktime.c (__mktime_internal): Fix misspelling in comment.
This merges all recent glibc changes of importance.
2011-01-28 Ralf Wildenhues <Ralf.Wildenhues@gmx.de>
move-if-change: cope with concurrent mv of identical file.
* move-if-change (CMPPROG): Accept environment
variable as an override for `cmp'.
(usage): Document CMPPROG.
Adjust comparison to drop stdout. Cope with failure of mv if
the target file exists and is identical to the source, for
parallel builds.
Report from H.J. Lu against binutils in PR binutils/12283.
2011-01-30 19:22:02 +00:00
|
|
|
|
long_int lmday = mday;
|
|
|
|
|
long_int yday = mon_yday + lmday;
|
2011-01-09 06:57:07 +00:00
|
|
|
|
|
|
|
|
|
time_t guessed_offset = *offset;
|
|
|
|
|
|
|
|
|
|
int sec_requested = sec;
|
|
|
|
|
|
|
|
|
|
if (LEAP_SECONDS_POSSIBLE)
|
|
|
|
|
{
|
|
|
|
|
/* Handle out-of-range seconds specially,
|
2012-05-26 23:14:36 +00:00
|
|
|
|
since ydhms_tm_diff assumes every minute has 60 seconds. */
|
2011-01-09 06:57:07 +00:00
|
|
|
|
if (sec < 0)
|
2012-05-26 23:14:36 +00:00
|
|
|
|
sec = 0;
|
2011-01-09 06:57:07 +00:00
|
|
|
|
if (59 < sec)
|
2012-05-26 23:14:36 +00:00
|
|
|
|
sec = 59;
|
2011-01-09 06:57:07 +00:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Invert CONVERT by probing. First assume the same offset as last
|
|
|
|
|
time. */
|
|
|
|
|
|
|
|
|
|
t0 = ydhms_diff (year, yday, hour, min, sec,
|
2012-05-26 23:14:36 +00:00
|
|
|
|
EPOCH_YEAR - TM_YEAR_BASE, 0, 0, 0, - guessed_offset);
|
2011-01-09 06:57:07 +00:00
|
|
|
|
|
|
|
|
|
if (TIME_T_MAX / INT_MAX / 366 / 24 / 60 / 60 < 3)
|
|
|
|
|
{
|
|
|
|
|
/* time_t isn't large enough to rule out overflows, so check
|
2012-05-26 23:14:36 +00:00
|
|
|
|
for major overflows. A gross check suffices, since if t0
|
|
|
|
|
has overflowed, it is off by a multiple of TIME_T_MAX -
|
|
|
|
|
TIME_T_MIN + 1. So ignore any component of the difference
|
|
|
|
|
that is bounded by a small value. */
|
2011-01-09 06:57:07 +00:00
|
|
|
|
|
|
|
|
|
/* Approximate log base 2 of the number of time units per
|
2012-05-26 23:14:36 +00:00
|
|
|
|
biennium. A biennium is 2 years; use this unit instead of
|
|
|
|
|
years to avoid integer overflow. For example, 2 average
|
|
|
|
|
Gregorian years are 2 * 365.2425 * 24 * 60 * 60 seconds,
|
|
|
|
|
which is 63113904 seconds, and rint (log2 (63113904)) is
|
|
|
|
|
26. */
|
2011-01-09 06:57:07 +00:00
|
|
|
|
int ALOG2_SECONDS_PER_BIENNIUM = 26;
|
|
|
|
|
int ALOG2_MINUTES_PER_BIENNIUM = 20;
|
|
|
|
|
int ALOG2_HOURS_PER_BIENNIUM = 14;
|
|
|
|
|
int ALOG2_DAYS_PER_BIENNIUM = 10;
|
|
|
|
|
int LOG2_YEARS_PER_BIENNIUM = 1;
|
|
|
|
|
|
|
|
|
|
int approx_requested_biennia =
|
2012-05-26 23:14:36 +00:00
|
|
|
|
(SHR (year_requested, LOG2_YEARS_PER_BIENNIUM)
|
|
|
|
|
- SHR (EPOCH_YEAR - TM_YEAR_BASE, LOG2_YEARS_PER_BIENNIUM)
|
|
|
|
|
+ SHR (mday, ALOG2_DAYS_PER_BIENNIUM)
|
|
|
|
|
+ SHR (hour, ALOG2_HOURS_PER_BIENNIUM)
|
|
|
|
|
+ SHR (min, ALOG2_MINUTES_PER_BIENNIUM)
|
|
|
|
|
+ (LEAP_SECONDS_POSSIBLE
|
|
|
|
|
? 0
|
|
|
|
|
: SHR (sec, ALOG2_SECONDS_PER_BIENNIUM)));
|
2011-01-09 06:57:07 +00:00
|
|
|
|
|
|
|
|
|
int approx_biennia = SHR (t0, ALOG2_SECONDS_PER_BIENNIUM);
|
|
|
|
|
int diff = approx_biennia - approx_requested_biennia;
|
2012-05-26 23:14:36 +00:00
|
|
|
|
int approx_abs_diff = diff < 0 ? -1 - diff : diff;
|
2011-01-09 06:57:07 +00:00
|
|
|
|
|
gnulib: import mktime and move-if-change fixes from gnulib
* configure: Regenerate from the following.
2011-01-30 Paul Eggert <eggert@cs.ucla.edu>
mktime: clarify long_int width checking
* lib/mktime.c (long_int_is_wide_enough): Move this assertion to
the top level, to make it clearer that the assumption about
long_int width is being checked. See
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00554.html>.
2011-01-29 Paul Eggert <eggert@cs.ucla.edu>
TYPE_MAXIMUM: avoid theoretically undefined behavior
* lib/intprops.h (TYPE_MINIMUM, TYPE_MAXIMUM): Do not shift a
negative number, which the C Standard says has undefined behavior.
In practice this is not a problem, but might as well do it by the book.
Reported by Rich Felker and Eric Blake; see
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00493.html>.
* m4/mktime.m4 (AC_FUNC_MKTIME): Likewise.
* lib/mktime.c (TYPE_MAXIMUM): Redo slightly to match the others.
mktime: #undef mktime before #defining it
* lib/mktime.c (mktime) [DEBUG]: #undef mktime before #defining it.
mktime: systematically normalize tm_isdst comparisons
* lib/mktime.c (isdst_differ): New function.
(__mktime_internal): Use it systematically for all isdst comparisons.
This completes the fix for libc BZ #6723, and removes the need for
normalizing tm_isdst. See
<http://sourceware.org/bugzilla/show_bug.cgi?id=6723>
(not_equal_tm) [DEBUG]: Use isdst_differ here, too.
mktime: fix some integer overflow issues and sidestep the rest
This was prompted by a bug report by Benjamin Lindner for MinGW
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00472.html>.
His bug is due to signed integer overflow (0 - INT_MIN), and I
I scanned through mktime.c looking for other integer overflow
problems, fixing all the bugs I found.
Although the C Standard says the resulting code is still not safe
in the presence of integer overflow, in practice it should be good
enough for all real-world two's-complement implementations, except
for debugging environments that deliberately trap on integer
overflow (e.g., gcc -ftrapv).
* lib/mktime.c (WRAPV): New macro.
(SHR): Also check that long_int and time_t shift right in the
usual way, before using the fast-but-unportable method.
(TYPE_ONES_COMPLEMENT, TYPE_SIGNED_MAGNITUDE): Remove, no longer
used. The code already assumed two's complement, so there's
no need to test for alternatives. All uses removed.
(TYPE_MAXIMUM): Don't rely here on overflow behavior not defined by
the C standard. Problem reported by Rich Felker in
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00488.html>.
(twos_complement_arithmetic): Also check long_int and time_t.
(time_t_avg, time_t_add_ok, time_t_int_add_ok): New functions.
(guess_time_tm, ranged_convert, __mktime_internal): Use them.
(__mktime_internal): Avoid integer overflow with unary subtraction
in two instances where -1 - X is an adequate replacement for -X,
since the calculations are approximate.
2011-01-29 Eric Blake <eblake@redhat.com>
mktime: avoid infinite loop
* m4/mktime.m4 (AC_FUNC_MKTIME): Avoid overflow on possibly-signed
type; behavior is still undefined but portable to all known targets.
Reported by Rich Felker.
2011-01-28 Paul Eggert <eggert@cs.ucla.edu>
mktime: avoid problems on NetBSD 5 / i386
* lib/mktime.c (long_int): New type. This works around a problem
on NetBSD 5 / i386, where 'long int' and 'int' are both 32 bits
but time_t is 64 bits, and where I expect the existing code is
wrong in some cases.
(leapyear, ydhms_diff, guess_time_tm, __mktime_internal): Use it.
(ydhms_diff): Bring back the compile-time check for wide-enough
year and yday.
mktime: fix misspelling in comment
* lib/mktime.c (__mktime_internal): Fix misspelling in comment.
This merges all recent glibc changes of importance.
2011-01-28 Ralf Wildenhues <Ralf.Wildenhues@gmx.de>
move-if-change: cope with concurrent mv of identical file.
* move-if-change (CMPPROG): Accept environment
variable as an override for `cmp'.
(usage): Document CMPPROG.
Adjust comparison to drop stdout. Cope with failure of mv if
the target file exists and is identical to the source, for
parallel builds.
Report from H.J. Lu against binutils in PR binutils/12283.
2011-01-30 19:22:02 +00:00
|
|
|
|
/* IRIX 4.0.5 cc miscalculates TIME_T_MIN / 3: it erroneously
|
2012-05-26 23:14:36 +00:00
|
|
|
|
gives a positive value of 715827882. Setting a variable
|
|
|
|
|
first then doing math on it seems to work.
|
|
|
|
|
(ghazi@caip.rutgers.edu) */
|
2011-01-09 06:57:07 +00:00
|
|
|
|
time_t time_t_max = TIME_T_MAX;
|
|
|
|
|
time_t time_t_min = TIME_T_MIN;
|
|
|
|
|
time_t overflow_threshold =
|
2012-05-26 23:14:36 +00:00
|
|
|
|
(time_t_max / 3 - time_t_min / 3) >> ALOG2_SECONDS_PER_BIENNIUM;
|
|
|
|
|
|
|
|
|
|
if (overflow_threshold < approx_abs_diff)
|
|
|
|
|
{
|
|
|
|
|
/* Overflow occurred. Try repairing it; this might work if
|
|
|
|
|
the time zone offset is enough to undo the overflow. */
|
|
|
|
|
time_t repaired_t0 = -1 - t0;
|
|
|
|
|
approx_biennia = SHR (repaired_t0, ALOG2_SECONDS_PER_BIENNIUM);
|
|
|
|
|
diff = approx_biennia - approx_requested_biennia;
|
|
|
|
|
approx_abs_diff = diff < 0 ? -1 - diff : diff;
|
|
|
|
|
if (overflow_threshold < approx_abs_diff)
|
|
|
|
|
return -1;
|
|
|
|
|
guessed_offset += repaired_t0 - t0;
|
|
|
|
|
t0 = repaired_t0;
|
|
|
|
|
}
|
2011-01-09 06:57:07 +00:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Repeatedly use the error to improve the guess. */
|
|
|
|
|
|
|
|
|
|
for (t = t1 = t2 = t0, dst2 = 0;
|
|
|
|
|
(gt = guess_time_tm (year, yday, hour, min, sec, &t,
|
2012-05-26 23:14:36 +00:00
|
|
|
|
ranged_convert (convert, &t, &tm)),
|
|
|
|
|
t != gt);
|
2011-01-09 06:57:07 +00:00
|
|
|
|
t1 = t2, t2 = t, t = gt, dst2 = tm.tm_isdst != 0)
|
|
|
|
|
if (t == t1 && t != t2
|
2012-05-26 23:14:36 +00:00
|
|
|
|
&& (tm.tm_isdst < 0
|
|
|
|
|
|| (isdst < 0
|
|
|
|
|
? dst2 <= (tm.tm_isdst != 0)
|
|
|
|
|
: (isdst != 0) != (tm.tm_isdst != 0))))
|
2011-01-09 06:57:07 +00:00
|
|
|
|
/* We can't possibly find a match, as we are oscillating
|
2012-05-26 23:14:36 +00:00
|
|
|
|
between two values. The requested time probably falls
|
|
|
|
|
within a spring-forward gap of size GT - T. Follow the common
|
|
|
|
|
practice in this case, which is to return a time that is GT - T
|
|
|
|
|
away from the requested time, preferring a time whose
|
|
|
|
|
tm_isdst differs from the requested value. (If no tm_isdst
|
|
|
|
|
was requested and only one of the two values has a nonzero
|
|
|
|
|
tm_isdst, prefer that value.) In practice, this is more
|
|
|
|
|
useful than returning -1. */
|
2011-01-09 06:57:07 +00:00
|
|
|
|
goto offset_found;
|
|
|
|
|
else if (--remaining_probes == 0)
|
|
|
|
|
return -1;
|
|
|
|
|
|
|
|
|
|
/* We have a match. Check whether tm.tm_isdst has the requested
|
|
|
|
|
value, if any. */
|
gnulib: import mktime and move-if-change fixes from gnulib
* configure: Regenerate from the following.
2011-01-30 Paul Eggert <eggert@cs.ucla.edu>
mktime: clarify long_int width checking
* lib/mktime.c (long_int_is_wide_enough): Move this assertion to
the top level, to make it clearer that the assumption about
long_int width is being checked. See
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00554.html>.
2011-01-29 Paul Eggert <eggert@cs.ucla.edu>
TYPE_MAXIMUM: avoid theoretically undefined behavior
* lib/intprops.h (TYPE_MINIMUM, TYPE_MAXIMUM): Do not shift a
negative number, which the C Standard says has undefined behavior.
In practice this is not a problem, but might as well do it by the book.
Reported by Rich Felker and Eric Blake; see
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00493.html>.
* m4/mktime.m4 (AC_FUNC_MKTIME): Likewise.
* lib/mktime.c (TYPE_MAXIMUM): Redo slightly to match the others.
mktime: #undef mktime before #defining it
* lib/mktime.c (mktime) [DEBUG]: #undef mktime before #defining it.
mktime: systematically normalize tm_isdst comparisons
* lib/mktime.c (isdst_differ): New function.
(__mktime_internal): Use it systematically for all isdst comparisons.
This completes the fix for libc BZ #6723, and removes the need for
normalizing tm_isdst. See
<http://sourceware.org/bugzilla/show_bug.cgi?id=6723>
(not_equal_tm) [DEBUG]: Use isdst_differ here, too.
mktime: fix some integer overflow issues and sidestep the rest
This was prompted by a bug report by Benjamin Lindner for MinGW
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00472.html>.
His bug is due to signed integer overflow (0 - INT_MIN), and I
I scanned through mktime.c looking for other integer overflow
problems, fixing all the bugs I found.
Although the C Standard says the resulting code is still not safe
in the presence of integer overflow, in practice it should be good
enough for all real-world two's-complement implementations, except
for debugging environments that deliberately trap on integer
overflow (e.g., gcc -ftrapv).
* lib/mktime.c (WRAPV): New macro.
(SHR): Also check that long_int and time_t shift right in the
usual way, before using the fast-but-unportable method.
(TYPE_ONES_COMPLEMENT, TYPE_SIGNED_MAGNITUDE): Remove, no longer
used. The code already assumed two's complement, so there's
no need to test for alternatives. All uses removed.
(TYPE_MAXIMUM): Don't rely here on overflow behavior not defined by
the C standard. Problem reported by Rich Felker in
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00488.html>.
(twos_complement_arithmetic): Also check long_int and time_t.
(time_t_avg, time_t_add_ok, time_t_int_add_ok): New functions.
(guess_time_tm, ranged_convert, __mktime_internal): Use them.
(__mktime_internal): Avoid integer overflow with unary subtraction
in two instances where -1 - X is an adequate replacement for -X,
since the calculations are approximate.
2011-01-29 Eric Blake <eblake@redhat.com>
mktime: avoid infinite loop
* m4/mktime.m4 (AC_FUNC_MKTIME): Avoid overflow on possibly-signed
type; behavior is still undefined but portable to all known targets.
Reported by Rich Felker.
2011-01-28 Paul Eggert <eggert@cs.ucla.edu>
mktime: avoid problems on NetBSD 5 / i386
* lib/mktime.c (long_int): New type. This works around a problem
on NetBSD 5 / i386, where 'long int' and 'int' are both 32 bits
but time_t is 64 bits, and where I expect the existing code is
wrong in some cases.
(leapyear, ydhms_diff, guess_time_tm, __mktime_internal): Use it.
(ydhms_diff): Bring back the compile-time check for wide-enough
year and yday.
mktime: fix misspelling in comment
* lib/mktime.c (__mktime_internal): Fix misspelling in comment.
This merges all recent glibc changes of importance.
2011-01-28 Ralf Wildenhues <Ralf.Wildenhues@gmx.de>
move-if-change: cope with concurrent mv of identical file.
* move-if-change (CMPPROG): Accept environment
variable as an override for `cmp'.
(usage): Document CMPPROG.
Adjust comparison to drop stdout. Cope with failure of mv if
the target file exists and is identical to the source, for
parallel builds.
Report from H.J. Lu against binutils in PR binutils/12283.
2011-01-30 19:22:02 +00:00
|
|
|
|
if (isdst_differ (isdst, tm.tm_isdst))
|
2011-01-09 06:57:07 +00:00
|
|
|
|
{
|
|
|
|
|
/* tm.tm_isdst has the wrong value. Look for a neighboring
|
2012-05-26 23:14:36 +00:00
|
|
|
|
time with the right value, and use its UTC offset.
|
2011-01-09 06:57:07 +00:00
|
|
|
|
|
2012-05-26 23:14:36 +00:00
|
|
|
|
Heuristic: probe the adjacent timestamps in both directions,
|
|
|
|
|
looking for the desired isdst. This should work for all real
|
|
|
|
|
time zone histories in the tz database. */
|
2011-01-09 06:57:07 +00:00
|
|
|
|
|
|
|
|
|
/* Distance between probes when looking for a DST boundary. In
|
2012-05-26 23:14:36 +00:00
|
|
|
|
tzdata2003a, the shortest period of DST is 601200 seconds
|
|
|
|
|
(e.g., America/Recife starting 2000-10-08 01:00), and the
|
|
|
|
|
shortest period of non-DST surrounded by DST is 694800
|
|
|
|
|
seconds (Africa/Tunis starting 1943-04-17 01:00). Use the
|
|
|
|
|
minimum of these two values, so we don't miss these short
|
|
|
|
|
periods when probing. */
|
2011-01-09 06:57:07 +00:00
|
|
|
|
int stride = 601200;
|
|
|
|
|
|
|
|
|
|
/* The longest period of DST in tzdata2003a is 536454000 seconds
|
2012-05-26 23:14:36 +00:00
|
|
|
|
(e.g., America/Jujuy starting 1946-10-01 01:00). The longest
|
|
|
|
|
period of non-DST is much longer, but it makes no real sense
|
|
|
|
|
to search for more than a year of non-DST, so use the DST
|
|
|
|
|
max. */
|
2011-01-09 06:57:07 +00:00
|
|
|
|
int duration_max = 536454000;
|
|
|
|
|
|
|
|
|
|
/* Search in both directions, so the maximum distance is half
|
2012-05-26 23:14:36 +00:00
|
|
|
|
the duration; add the stride to avoid off-by-1 problems. */
|
2011-01-09 06:57:07 +00:00
|
|
|
|
int delta_bound = duration_max / 2 + stride;
|
|
|
|
|
|
|
|
|
|
int delta, direction;
|
|
|
|
|
|
|
|
|
|
for (delta = stride; delta < delta_bound; delta += stride)
|
2012-05-26 23:14:36 +00:00
|
|
|
|
for (direction = -1; direction <= 1; direction += 2)
|
|
|
|
|
if (time_t_int_add_ok (t, delta * direction))
|
|
|
|
|
{
|
|
|
|
|
time_t ot = t + delta * direction;
|
|
|
|
|
struct tm otm;
|
|
|
|
|
ranged_convert (convert, &ot, &otm);
|
|
|
|
|
if (! isdst_differ (isdst, otm.tm_isdst))
|
|
|
|
|
{
|
|
|
|
|
/* We found the desired tm_isdst.
|
|
|
|
|
Extrapolate back to the desired time. */
|
|
|
|
|
t = guess_time_tm (year, yday, hour, min, sec, &ot, &otm);
|
|
|
|
|
ranged_convert (convert, &t, &tm);
|
|
|
|
|
goto offset_found;
|
|
|
|
|
}
|
|
|
|
|
}
|
2011-01-09 06:57:07 +00:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
offset_found:
|
|
|
|
|
*offset = guessed_offset + t - t0;
|
|
|
|
|
|
|
|
|
|
if (LEAP_SECONDS_POSSIBLE && sec_requested != tm.tm_sec)
|
|
|
|
|
{
|
|
|
|
|
/* Adjust time to reflect the tm_sec requested, not the normalized value.
|
2012-05-26 23:14:36 +00:00
|
|
|
|
Also, repair any damage from a false match due to a leap second. */
|
2011-01-09 06:57:07 +00:00
|
|
|
|
int sec_adjustment = (sec == 0 && tm.tm_sec == 60) - sec;
|
gnulib: import mktime and move-if-change fixes from gnulib
* configure: Regenerate from the following.
2011-01-30 Paul Eggert <eggert@cs.ucla.edu>
mktime: clarify long_int width checking
* lib/mktime.c (long_int_is_wide_enough): Move this assertion to
the top level, to make it clearer that the assumption about
long_int width is being checked. See
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00554.html>.
2011-01-29 Paul Eggert <eggert@cs.ucla.edu>
TYPE_MAXIMUM: avoid theoretically undefined behavior
* lib/intprops.h (TYPE_MINIMUM, TYPE_MAXIMUM): Do not shift a
negative number, which the C Standard says has undefined behavior.
In practice this is not a problem, but might as well do it by the book.
Reported by Rich Felker and Eric Blake; see
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00493.html>.
* m4/mktime.m4 (AC_FUNC_MKTIME): Likewise.
* lib/mktime.c (TYPE_MAXIMUM): Redo slightly to match the others.
mktime: #undef mktime before #defining it
* lib/mktime.c (mktime) [DEBUG]: #undef mktime before #defining it.
mktime: systematically normalize tm_isdst comparisons
* lib/mktime.c (isdst_differ): New function.
(__mktime_internal): Use it systematically for all isdst comparisons.
This completes the fix for libc BZ #6723, and removes the need for
normalizing tm_isdst. See
<http://sourceware.org/bugzilla/show_bug.cgi?id=6723>
(not_equal_tm) [DEBUG]: Use isdst_differ here, too.
mktime: fix some integer overflow issues and sidestep the rest
This was prompted by a bug report by Benjamin Lindner for MinGW
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00472.html>.
His bug is due to signed integer overflow (0 - INT_MIN), and I
I scanned through mktime.c looking for other integer overflow
problems, fixing all the bugs I found.
Although the C Standard says the resulting code is still not safe
in the presence of integer overflow, in practice it should be good
enough for all real-world two's-complement implementations, except
for debugging environments that deliberately trap on integer
overflow (e.g., gcc -ftrapv).
* lib/mktime.c (WRAPV): New macro.
(SHR): Also check that long_int and time_t shift right in the
usual way, before using the fast-but-unportable method.
(TYPE_ONES_COMPLEMENT, TYPE_SIGNED_MAGNITUDE): Remove, no longer
used. The code already assumed two's complement, so there's
no need to test for alternatives. All uses removed.
(TYPE_MAXIMUM): Don't rely here on overflow behavior not defined by
the C standard. Problem reported by Rich Felker in
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00488.html>.
(twos_complement_arithmetic): Also check long_int and time_t.
(time_t_avg, time_t_add_ok, time_t_int_add_ok): New functions.
(guess_time_tm, ranged_convert, __mktime_internal): Use them.
(__mktime_internal): Avoid integer overflow with unary subtraction
in two instances where -1 - X is an adequate replacement for -X,
since the calculations are approximate.
2011-01-29 Eric Blake <eblake@redhat.com>
mktime: avoid infinite loop
* m4/mktime.m4 (AC_FUNC_MKTIME): Avoid overflow on possibly-signed
type; behavior is still undefined but portable to all known targets.
Reported by Rich Felker.
2011-01-28 Paul Eggert <eggert@cs.ucla.edu>
mktime: avoid problems on NetBSD 5 / i386
* lib/mktime.c (long_int): New type. This works around a problem
on NetBSD 5 / i386, where 'long int' and 'int' are both 32 bits
but time_t is 64 bits, and where I expect the existing code is
wrong in some cases.
(leapyear, ydhms_diff, guess_time_tm, __mktime_internal): Use it.
(ydhms_diff): Bring back the compile-time check for wide-enough
year and yday.
mktime: fix misspelling in comment
* lib/mktime.c (__mktime_internal): Fix misspelling in comment.
This merges all recent glibc changes of importance.
2011-01-28 Ralf Wildenhues <Ralf.Wildenhues@gmx.de>
move-if-change: cope with concurrent mv of identical file.
* move-if-change (CMPPROG): Accept environment
variable as an override for `cmp'.
(usage): Document CMPPROG.
Adjust comparison to drop stdout. Cope with failure of mv if
the target file exists and is identical to the source, for
parallel builds.
Report from H.J. Lu against binutils in PR binutils/12283.
2011-01-30 19:22:02 +00:00
|
|
|
|
if (! time_t_int_add_ok (t, sec_requested))
|
2012-05-26 23:14:36 +00:00
|
|
|
|
return -1;
|
2011-01-09 06:57:07 +00:00
|
|
|
|
t1 = t + sec_requested;
|
gnulib: import mktime and move-if-change fixes from gnulib
* configure: Regenerate from the following.
2011-01-30 Paul Eggert <eggert@cs.ucla.edu>
mktime: clarify long_int width checking
* lib/mktime.c (long_int_is_wide_enough): Move this assertion to
the top level, to make it clearer that the assumption about
long_int width is being checked. See
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00554.html>.
2011-01-29 Paul Eggert <eggert@cs.ucla.edu>
TYPE_MAXIMUM: avoid theoretically undefined behavior
* lib/intprops.h (TYPE_MINIMUM, TYPE_MAXIMUM): Do not shift a
negative number, which the C Standard says has undefined behavior.
In practice this is not a problem, but might as well do it by the book.
Reported by Rich Felker and Eric Blake; see
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00493.html>.
* m4/mktime.m4 (AC_FUNC_MKTIME): Likewise.
* lib/mktime.c (TYPE_MAXIMUM): Redo slightly to match the others.
mktime: #undef mktime before #defining it
* lib/mktime.c (mktime) [DEBUG]: #undef mktime before #defining it.
mktime: systematically normalize tm_isdst comparisons
* lib/mktime.c (isdst_differ): New function.
(__mktime_internal): Use it systematically for all isdst comparisons.
This completes the fix for libc BZ #6723, and removes the need for
normalizing tm_isdst. See
<http://sourceware.org/bugzilla/show_bug.cgi?id=6723>
(not_equal_tm) [DEBUG]: Use isdst_differ here, too.
mktime: fix some integer overflow issues and sidestep the rest
This was prompted by a bug report by Benjamin Lindner for MinGW
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00472.html>.
His bug is due to signed integer overflow (0 - INT_MIN), and I
I scanned through mktime.c looking for other integer overflow
problems, fixing all the bugs I found.
Although the C Standard says the resulting code is still not safe
in the presence of integer overflow, in practice it should be good
enough for all real-world two's-complement implementations, except
for debugging environments that deliberately trap on integer
overflow (e.g., gcc -ftrapv).
* lib/mktime.c (WRAPV): New macro.
(SHR): Also check that long_int and time_t shift right in the
usual way, before using the fast-but-unportable method.
(TYPE_ONES_COMPLEMENT, TYPE_SIGNED_MAGNITUDE): Remove, no longer
used. The code already assumed two's complement, so there's
no need to test for alternatives. All uses removed.
(TYPE_MAXIMUM): Don't rely here on overflow behavior not defined by
the C standard. Problem reported by Rich Felker in
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00488.html>.
(twos_complement_arithmetic): Also check long_int and time_t.
(time_t_avg, time_t_add_ok, time_t_int_add_ok): New functions.
(guess_time_tm, ranged_convert, __mktime_internal): Use them.
(__mktime_internal): Avoid integer overflow with unary subtraction
in two instances where -1 - X is an adequate replacement for -X,
since the calculations are approximate.
2011-01-29 Eric Blake <eblake@redhat.com>
mktime: avoid infinite loop
* m4/mktime.m4 (AC_FUNC_MKTIME): Avoid overflow on possibly-signed
type; behavior is still undefined but portable to all known targets.
Reported by Rich Felker.
2011-01-28 Paul Eggert <eggert@cs.ucla.edu>
mktime: avoid problems on NetBSD 5 / i386
* lib/mktime.c (long_int): New type. This works around a problem
on NetBSD 5 / i386, where 'long int' and 'int' are both 32 bits
but time_t is 64 bits, and where I expect the existing code is
wrong in some cases.
(leapyear, ydhms_diff, guess_time_tm, __mktime_internal): Use it.
(ydhms_diff): Bring back the compile-time check for wide-enough
year and yday.
mktime: fix misspelling in comment
* lib/mktime.c (__mktime_internal): Fix misspelling in comment.
This merges all recent glibc changes of importance.
2011-01-28 Ralf Wildenhues <Ralf.Wildenhues@gmx.de>
move-if-change: cope with concurrent mv of identical file.
* move-if-change (CMPPROG): Accept environment
variable as an override for `cmp'.
(usage): Document CMPPROG.
Adjust comparison to drop stdout. Cope with failure of mv if
the target file exists and is identical to the source, for
parallel builds.
Report from H.J. Lu against binutils in PR binutils/12283.
2011-01-30 19:22:02 +00:00
|
|
|
|
if (! time_t_int_add_ok (t1, sec_adjustment))
|
2012-05-26 23:14:36 +00:00
|
|
|
|
return -1;
|
2011-01-09 06:57:07 +00:00
|
|
|
|
t2 = t1 + sec_adjustment;
|
gnulib: import mktime and move-if-change fixes from gnulib
* configure: Regenerate from the following.
2011-01-30 Paul Eggert <eggert@cs.ucla.edu>
mktime: clarify long_int width checking
* lib/mktime.c (long_int_is_wide_enough): Move this assertion to
the top level, to make it clearer that the assumption about
long_int width is being checked. See
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00554.html>.
2011-01-29 Paul Eggert <eggert@cs.ucla.edu>
TYPE_MAXIMUM: avoid theoretically undefined behavior
* lib/intprops.h (TYPE_MINIMUM, TYPE_MAXIMUM): Do not shift a
negative number, which the C Standard says has undefined behavior.
In practice this is not a problem, but might as well do it by the book.
Reported by Rich Felker and Eric Blake; see
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00493.html>.
* m4/mktime.m4 (AC_FUNC_MKTIME): Likewise.
* lib/mktime.c (TYPE_MAXIMUM): Redo slightly to match the others.
mktime: #undef mktime before #defining it
* lib/mktime.c (mktime) [DEBUG]: #undef mktime before #defining it.
mktime: systematically normalize tm_isdst comparisons
* lib/mktime.c (isdst_differ): New function.
(__mktime_internal): Use it systematically for all isdst comparisons.
This completes the fix for libc BZ #6723, and removes the need for
normalizing tm_isdst. See
<http://sourceware.org/bugzilla/show_bug.cgi?id=6723>
(not_equal_tm) [DEBUG]: Use isdst_differ here, too.
mktime: fix some integer overflow issues and sidestep the rest
This was prompted by a bug report by Benjamin Lindner for MinGW
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00472.html>.
His bug is due to signed integer overflow (0 - INT_MIN), and I
I scanned through mktime.c looking for other integer overflow
problems, fixing all the bugs I found.
Although the C Standard says the resulting code is still not safe
in the presence of integer overflow, in practice it should be good
enough for all real-world two's-complement implementations, except
for debugging environments that deliberately trap on integer
overflow (e.g., gcc -ftrapv).
* lib/mktime.c (WRAPV): New macro.
(SHR): Also check that long_int and time_t shift right in the
usual way, before using the fast-but-unportable method.
(TYPE_ONES_COMPLEMENT, TYPE_SIGNED_MAGNITUDE): Remove, no longer
used. The code already assumed two's complement, so there's
no need to test for alternatives. All uses removed.
(TYPE_MAXIMUM): Don't rely here on overflow behavior not defined by
the C standard. Problem reported by Rich Felker in
<http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00488.html>.
(twos_complement_arithmetic): Also check long_int and time_t.
(time_t_avg, time_t_add_ok, time_t_int_add_ok): New functions.
(guess_time_tm, ranged_convert, __mktime_internal): Use them.
(__mktime_internal): Avoid integer overflow with unary subtraction
in two instances where -1 - X is an adequate replacement for -X,
since the calculations are approximate.
2011-01-29 Eric Blake <eblake@redhat.com>
mktime: avoid infinite loop
* m4/mktime.m4 (AC_FUNC_MKTIME): Avoid overflow on possibly-signed
type; behavior is still undefined but portable to all known targets.
Reported by Rich Felker.
2011-01-28 Paul Eggert <eggert@cs.ucla.edu>
mktime: avoid problems on NetBSD 5 / i386
* lib/mktime.c (long_int): New type. This works around a problem
on NetBSD 5 / i386, where 'long int' and 'int' are both 32 bits
but time_t is 64 bits, and where I expect the existing code is
wrong in some cases.
(leapyear, ydhms_diff, guess_time_tm, __mktime_internal): Use it.
(ydhms_diff): Bring back the compile-time check for wide-enough
year and yday.
mktime: fix misspelling in comment
* lib/mktime.c (__mktime_internal): Fix misspelling in comment.
This merges all recent glibc changes of importance.
2011-01-28 Ralf Wildenhues <Ralf.Wildenhues@gmx.de>
move-if-change: cope with concurrent mv of identical file.
* move-if-change (CMPPROG): Accept environment
variable as an override for `cmp'.
(usage): Document CMPPROG.
Adjust comparison to drop stdout. Cope with failure of mv if
the target file exists and is identical to the source, for
parallel builds.
Report from H.J. Lu against binutils in PR binutils/12283.
2011-01-30 19:22:02 +00:00
|
|
|
|
if (! convert (&t2, &tm))
|
2012-05-26 23:14:36 +00:00
|
|
|
|
return -1;
|
2011-01-09 06:57:07 +00:00
|
|
|
|
t = t2;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
*tp = tm;
|
|
|
|
|
return t;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
/* FIXME: This should use a signed type wide enough to hold any UTC
|
|
|
|
|
offset in seconds. 'int' should be good enough for GNU code. We
|
|
|
|
|
can't fix this unilaterally though, as other modules invoke
|
|
|
|
|
__mktime_internal. */
|
|
|
|
|
static time_t localtime_offset;
|
|
|
|
|
|
|
|
|
|
/* Convert *TP to a time_t value. */
|
|
|
|
|
time_t
|
|
|
|
|
mktime (struct tm *tp)
|
|
|
|
|
{
|
|
|
|
|
#ifdef _LIBC
|
|
|
|
|
/* POSIX.1 8.1.1 requires that whenever mktime() is called, the
|
2012-05-26 23:14:36 +00:00
|
|
|
|
time zone names contained in the external variable 'tzname' shall
|
2011-01-09 06:57:07 +00:00
|
|
|
|
be set as if the tzset() function had been called. */
|
|
|
|
|
__tzset ();
|
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
return __mktime_internal (tp, __localtime_r, &localtime_offset);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
#ifdef weak_alias
|
|
|
|
|
weak_alias (mktime, timelocal)
|
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
#ifdef _LIBC
|
|
|
|
|
libc_hidden_def (mktime)
|
|
|
|
|
libc_hidden_weak (timelocal)
|
|
|
|
|
#endif
|
|
|
|
|
|
Merge from gnulib
This incorporates:
2016-01-24 openat_proc_name: fix last '/' overwritten on OS/2 kLIBC
2016-01-24 closedir, dirfd, opendir: port to OpenSolaris 5.10
2016-01-15 detect utimes() correctly on OS/2 kLIBC
2016-01-15 openat_proc_name: port to OS/2 kLIBC
2016-01-14 stdint: check _INTPTR_T_DECLARED for intptr_t etc.
2016-01-14 opendir, closedir, dirfd, fdopendir: port to OS/2 kLIBC
2016-01-14 dup, dup2, fcntl: support a directory fd on OS/2 kLIBC
2016-01-14 binary-io: don't put fd in binary mode if a console on EMX
2016-01-14 sig2str: list all signals on FreeBSD >= 7
2016-01-13 acl-permissions: port to USE_ACL==0 platforms
2016-01-12 mktime: rename macro to avoid glibc clash
2016-01-12 Port "$@" to OpenIndiana ksh93
2016-01-12 Port Universal Time settings to strict POSIX
* build-aux/gitlog-to-changelog, build-aux/update-copyright:
* doc/misc/texinfo.tex, lib/acl-internal.c, lib/acl-internal.h:
* lib/binary-io.h, lib/dirent.in.h, lib/dirfd.c, lib/dup2.c:
* lib/fcntl.c, lib/fdopendir.c, lib/mktime.c, lib/openat-proc.c:
* lib/sig2str.h, lib/stdint.in.h, m4/dirfd.m4, m4/dup2.m4:
* m4/fcntl.m4, m4/utimes.m4:
Copy from gnulib.
* m4/gnulib-comp.m4: Regenerate.
2016-01-24 22:50:47 +00:00
|
|
|
|
#if defined DEBUG_MKTIME && DEBUG_MKTIME
|
2011-01-09 06:57:07 +00:00
|
|
|
|
|
|
|
|
|
static int
|
|
|
|
|
not_equal_tm (const struct tm *a, const struct tm *b)
|
|
|
|
|
{
|
|
|
|
|
return ((a->tm_sec ^ b->tm_sec)
|
2012-05-26 23:14:36 +00:00
|
|
|
|
| (a->tm_min ^ b->tm_min)
|
|
|
|
|
| (a->tm_hour ^ b->tm_hour)
|
|
|
|
|
| (a->tm_mday ^ b->tm_mday)
|
|
|
|
|
| (a->tm_mon ^ b->tm_mon)
|
|
|
|
|
| (a->tm_year ^ b->tm_year)
|
|
|
|
|
| (a->tm_yday ^ b->tm_yday)
|
|
|
|
|
| isdst_differ (a->tm_isdst, b->tm_isdst));
|
2011-01-09 06:57:07 +00:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static void
|
|
|
|
|
print_tm (const struct tm *tp)
|
|
|
|
|
{
|
|
|
|
|
if (tp)
|
|
|
|
|
printf ("%04d-%02d-%02d %02d:%02d:%02d yday %03d wday %d isdst %d",
|
2012-05-26 23:14:36 +00:00
|
|
|
|
tp->tm_year + TM_YEAR_BASE, tp->tm_mon + 1, tp->tm_mday,
|
|
|
|
|
tp->tm_hour, tp->tm_min, tp->tm_sec,
|
|
|
|
|
tp->tm_yday, tp->tm_wday, tp->tm_isdst);
|
2011-01-09 06:57:07 +00:00
|
|
|
|
else
|
|
|
|
|
printf ("0");
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static int
|
|
|
|
|
check_result (time_t tk, struct tm tmk, time_t tl, const struct tm *lt)
|
|
|
|
|
{
|
|
|
|
|
if (tk != tl || !lt || not_equal_tm (&tmk, lt))
|
|
|
|
|
{
|
|
|
|
|
printf ("mktime (");
|
|
|
|
|
print_tm (lt);
|
|
|
|
|
printf (")\nyields (");
|
|
|
|
|
print_tm (&tmk);
|
|
|
|
|
printf (") == %ld, should be %ld\n", (long int) tk, (long int) tl);
|
|
|
|
|
return 1;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
int
|
|
|
|
|
main (int argc, char **argv)
|
|
|
|
|
{
|
|
|
|
|
int status = 0;
|
|
|
|
|
struct tm tm, tmk, tml;
|
|
|
|
|
struct tm *lt;
|
|
|
|
|
time_t tk, tl, tl1;
|
|
|
|
|
char trailer;
|
|
|
|
|
|
|
|
|
|
if ((argc == 3 || argc == 4)
|
|
|
|
|
&& (sscanf (argv[1], "%d-%d-%d%c",
|
2012-05-26 23:14:36 +00:00
|
|
|
|
&tm.tm_year, &tm.tm_mon, &tm.tm_mday, &trailer)
|
|
|
|
|
== 3)
|
2011-01-09 06:57:07 +00:00
|
|
|
|
&& (sscanf (argv[2], "%d:%d:%d%c",
|
2012-05-26 23:14:36 +00:00
|
|
|
|
&tm.tm_hour, &tm.tm_min, &tm.tm_sec, &trailer)
|
|
|
|
|
== 3))
|
2011-01-09 06:57:07 +00:00
|
|
|
|
{
|
|
|
|
|
tm.tm_year -= TM_YEAR_BASE;
|
|
|
|
|
tm.tm_mon--;
|
|
|
|
|
tm.tm_isdst = argc == 3 ? -1 : atoi (argv[3]);
|
|
|
|
|
tmk = tm;
|
|
|
|
|
tl = mktime (&tmk);
|
|
|
|
|
lt = localtime (&tl);
|
|
|
|
|
if (lt)
|
2012-05-26 23:14:36 +00:00
|
|
|
|
{
|
|
|
|
|
tml = *lt;
|
|
|
|
|
lt = &tml;
|
|
|
|
|
}
|
2011-01-09 06:57:07 +00:00
|
|
|
|
printf ("mktime returns %ld == ", (long int) tl);
|
|
|
|
|
print_tm (&tmk);
|
|
|
|
|
printf ("\n");
|
|
|
|
|
status = check_result (tl, tmk, tl, lt);
|
|
|
|
|
}
|
|
|
|
|
else if (argc == 4 || (argc == 5 && strcmp (argv[4], "-") == 0))
|
|
|
|
|
{
|
|
|
|
|
time_t from = atol (argv[1]);
|
|
|
|
|
time_t by = atol (argv[2]);
|
|
|
|
|
time_t to = atol (argv[3]);
|
|
|
|
|
|
|
|
|
|
if (argc == 4)
|
2012-05-26 23:14:36 +00:00
|
|
|
|
for (tl = from; by < 0 ? to <= tl : tl <= to; tl = tl1)
|
|
|
|
|
{
|
|
|
|
|
lt = localtime (&tl);
|
|
|
|
|
if (lt)
|
|
|
|
|
{
|
|
|
|
|
tmk = tml = *lt;
|
|
|
|
|
tk = mktime (&tmk);
|
|
|
|
|
status |= check_result (tk, tmk, tl, &tml);
|
|
|
|
|
}
|
|
|
|
|
else
|
|
|
|
|
{
|
|
|
|
|
printf ("localtime (%ld) yields 0\n", (long int) tl);
|
|
|
|
|
status = 1;
|
|
|
|
|
}
|
|
|
|
|
tl1 = tl + by;
|
|
|
|
|
if ((tl1 < tl) != (by < 0))
|
|
|
|
|
break;
|
|
|
|
|
}
|
2011-01-09 06:57:07 +00:00
|
|
|
|
else
|
2012-05-26 23:14:36 +00:00
|
|
|
|
for (tl = from; by < 0 ? to <= tl : tl <= to; tl = tl1)
|
|
|
|
|
{
|
|
|
|
|
/* Null benchmark. */
|
|
|
|
|
lt = localtime (&tl);
|
|
|
|
|
if (lt)
|
|
|
|
|
{
|
|
|
|
|
tmk = tml = *lt;
|
|
|
|
|
tk = tl;
|
|
|
|
|
status |= check_result (tk, tmk, tl, &tml);
|
|
|
|
|
}
|
|
|
|
|
else
|
|
|
|
|
{
|
|
|
|
|
printf ("localtime (%ld) yields 0\n", (long int) tl);
|
|
|
|
|
status = 1;
|
|
|
|
|
}
|
|
|
|
|
tl1 = tl + by;
|
|
|
|
|
if ((tl1 < tl) != (by < 0))
|
|
|
|
|
break;
|
|
|
|
|
}
|
2011-01-09 06:57:07 +00:00
|
|
|
|
}
|
|
|
|
|
else
|
|
|
|
|
printf ("Usage:\
|
|
|
|
|
\t%s YYYY-MM-DD HH:MM:SS [ISDST] # Test given time.\n\
|
|
|
|
|
\t%s FROM BY TO # Test values FROM, FROM+BY, ..., TO.\n\
|
|
|
|
|
\t%s FROM BY TO - # Do not test those values (for benchmark).\n",
|
2012-05-26 23:14:36 +00:00
|
|
|
|
argv[0], argv[0], argv[0]);
|
2011-01-09 06:57:07 +00:00
|
|
|
|
|
|
|
|
|
return status;
|
|
|
|
|
}
|
|
|
|
|
|
Merge from gnulib
This incorporates:
2016-01-24 openat_proc_name: fix last '/' overwritten on OS/2 kLIBC
2016-01-24 closedir, dirfd, opendir: port to OpenSolaris 5.10
2016-01-15 detect utimes() correctly on OS/2 kLIBC
2016-01-15 openat_proc_name: port to OS/2 kLIBC
2016-01-14 stdint: check _INTPTR_T_DECLARED for intptr_t etc.
2016-01-14 opendir, closedir, dirfd, fdopendir: port to OS/2 kLIBC
2016-01-14 dup, dup2, fcntl: support a directory fd on OS/2 kLIBC
2016-01-14 binary-io: don't put fd in binary mode if a console on EMX
2016-01-14 sig2str: list all signals on FreeBSD >= 7
2016-01-13 acl-permissions: port to USE_ACL==0 platforms
2016-01-12 mktime: rename macro to avoid glibc clash
2016-01-12 Port "$@" to OpenIndiana ksh93
2016-01-12 Port Universal Time settings to strict POSIX
* build-aux/gitlog-to-changelog, build-aux/update-copyright:
* doc/misc/texinfo.tex, lib/acl-internal.c, lib/acl-internal.h:
* lib/binary-io.h, lib/dirent.in.h, lib/dirfd.c, lib/dup2.c:
* lib/fcntl.c, lib/fdopendir.c, lib/mktime.c, lib/openat-proc.c:
* lib/sig2str.h, lib/stdint.in.h, m4/dirfd.m4, m4/dup2.m4:
* m4/fcntl.m4, m4/utimes.m4:
Copy from gnulib.
* m4/gnulib-comp.m4: Regenerate.
2016-01-24 22:50:47 +00:00
|
|
|
|
#endif /* DEBUG_MKTIME */
|
2011-01-09 06:57:07 +00:00
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
Local Variables:
|
Merge from gnulib
This incorporates:
2016-01-24 openat_proc_name: fix last '/' overwritten on OS/2 kLIBC
2016-01-24 closedir, dirfd, opendir: port to OpenSolaris 5.10
2016-01-15 detect utimes() correctly on OS/2 kLIBC
2016-01-15 openat_proc_name: port to OS/2 kLIBC
2016-01-14 stdint: check _INTPTR_T_DECLARED for intptr_t etc.
2016-01-14 opendir, closedir, dirfd, fdopendir: port to OS/2 kLIBC
2016-01-14 dup, dup2, fcntl: support a directory fd on OS/2 kLIBC
2016-01-14 binary-io: don't put fd in binary mode if a console on EMX
2016-01-14 sig2str: list all signals on FreeBSD >= 7
2016-01-13 acl-permissions: port to USE_ACL==0 platforms
2016-01-12 mktime: rename macro to avoid glibc clash
2016-01-12 Port "$@" to OpenIndiana ksh93
2016-01-12 Port Universal Time settings to strict POSIX
* build-aux/gitlog-to-changelog, build-aux/update-copyright:
* doc/misc/texinfo.tex, lib/acl-internal.c, lib/acl-internal.h:
* lib/binary-io.h, lib/dirent.in.h, lib/dirfd.c, lib/dup2.c:
* lib/fcntl.c, lib/fdopendir.c, lib/mktime.c, lib/openat-proc.c:
* lib/sig2str.h, lib/stdint.in.h, m4/dirfd.m4, m4/dup2.m4:
* m4/fcntl.m4, m4/utimes.m4:
Copy from gnulib.
* m4/gnulib-comp.m4: Regenerate.
2016-01-24 22:50:47 +00:00
|
|
|
|
compile-command: "gcc -DDEBUG_MKTIME -I. -Wall -W -O2 -g mktime.c -o mktime"
|
2011-01-09 06:57:07 +00:00
|
|
|
|
End:
|
|
|
|
|
*/
|