mirror of
https://sourceware.org/git/glibc.git
synced 2024-12-25 04:01:10 +00:00
6bc31da01b
Wed Jan 1 15:15:00 1997 Ulrich Drepper <drepper@cygnus.com> * libio/Makefile (routines): Add fcloseall. * libio/fcloseall.c: New file. * libio/stdio.h [__USE_GNU]: Add prototype for fcloseall. * libio/iofclose.c: Don't close all streams when argument is NULL. * stdio/Makefile (routines): Add fcloseall. * stdio/fcloseall.c: New file. * stdio/stdio.h [__USE_GNU]: Add prototype for fcloseall. * stdio/fclose.c: Don't close all streams when argument is NULL. * manual/stdio.texi: Document fcloseall. * sysdeps/generic/abort.c: Call __fcloseall instead of fclose with NULL argument. * socket/sys/un.h (SUN_LEN): Add cast. * sunrpc/auth_unix.c (authunix_create): Change parameter types of uid, gid, and aup_gids to uid_t and gid_t respectively. (authunix_create_default): Likewise. (authunix_refresh): Correct cast for aup_gids assignment. * sunrpc/rpc/auth.h (authunix_create): Correct types in declaration. * sunrpc/rpc/auth_unix.h (struct authunix_parms): Correct field types to uid_t and gid_t respectively. (struct short_hand_verf): Likewise. Reported by Andreas Schwab. * sysdeps/ieee754/Makefile: New file. Make sure ieee754.h gets installed. * sysdeps/ieee754/ieee754.h: Protect against multiple inclusion and use in C++ programs. * sysdeps/libm-i387/e_pow.S: New file. Optimized ix87 assembler version of pow() function. * sysdeps/libm-i387/e_powf.S: New file. Float version. * sysdeps/libm-i387/e_powl.S: New file. Long double version. * sysdeps/libm-i387/s_expm1.S: New file. Optimized ix87 assembler version of expm1() function. * sysdeps/libm-i387/s_expm1f.S: New file. Float version. * sysdeps/libm-i387/s_expm1l.S: New file. Long double version. * sysdeps/libm-i387/e_exp.S: Optimize handling of +-Inf. Better handling of computation stack from current NetBSD sources. * sysdeps/libm-i387/e_expf.S: Likewise. * sysdeps/libm-i387/e_expl.S: Likewise. * sysdeps/libm-i387/e_log.S: Use fstp instead of fincstp. * sysdeps/libm-i387/e_logf.S: Likewise. * sysdeps/libm-i387/e_logl.S: Likewise. * sysdeps/libm-i387/e_log10.S: Likewise. * sysdeps/libm-i387/e_log10f.S: Likewise. * sysdeps/libm-i387/e_log10l.S: Likewise. * sysdeps/libm-i387/s_asinh.S: Likewise. * sysdeps/libm-i387/s_asinhf.S: Likewise. * sysdeps/libm-i387/s_asinhl.S: Likewise. * sysdeps/libm-i387/s_significandl.S: Likewise. * sysdeps/libm-i387/s_ceil.S: Don't create stack frame. Modify stack pointer directly. Use 32 bit instead of 16 bit operations. * sysdeps/libm-i387/s_ceilf.S: Likewise. * sysdeps/libm-i387/s_ceill.S: Likewise. * sysdeps/libm-i387/s_floor.S: Likewise. * sysdeps/libm-i387/s_floorf.S: Likewise. * sysdeps/libm-i387/s_floorl.S: Likewise. * sysdeps/libm-ieee754/e_coshf.c (huge): Don't declare volatile. * sysdeps/libm-ieee754/e_expf.c: Likewise. * sysdeps/libm-ieee754/e_powf.c: Likewise. From current NetBSD sources. * sysdeps/libm-ieee754/e_hypot.c (__ieee754_hypot): Don't initialize `a' and `b' in definition. * sysdeps/libm-ieee754/e_hypotf.c: Likewise. * sysdeps/libm-ieee754/e_hypotl.c: New file. Long double implementation. * sysdeps/libm-ieee754/e_sinhl.c: New file. Long double implementation. * sysdeps/stub/e_hypotl.c: Removed. We have a real version now. * sysdeps/stub/e_sinhl.c: Removed. * sysdeps/posix/sigpause.c (__sigpause): Use sigdelset instead of sigaddset. Reported by Andreas Schwab. (__default_sigpause): New function. Call __sigpause as BSD version. Make this the default function by providing alias sigpause. * sysdeps/stub/sigpause.c (__default_sigpause): New funciton. Simply fail. Provide alias sigpause. Wed Jan 1 12:34:54 1997 Ulrich Drepper <drepper@cygnus.com> * po/es.po: Update for glibc-1.98. * po/fr.po: Likewise. * po/sv.po: New file. Swedish translation. Wed Jan 1 12:18:07 1997 MacGyver <macgyver@tos.net> * values.h: Fix typos: use SHRT_MAX but SHORT_MAX and SHRT_MIN but SHORT_MIN. Tue Dec 24 23:10:21 1996 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * Makefile: Replace `stubs.h' by `gnu/stubs.h'. * features.h: Include <gnu/stubs.h> instead of <stubs.h>. Tue Dec 24 22:16:55 1996 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * Makefile (subdirs): Move `elf' last so that librtld.so isn't relinked in the `make others' pass if an add-on adds something to libc. Mon Dec 23 21:23:16 1996 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/m68k/abort-instr.h: New File. Mon Dec 23 21:16:38 1996 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/m68k/Makefile ($(subdir)=elf): Define CFLAGS-rtld.c to prevent warning. Mon Dec 23 21:14:22 1996 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * resolv/Makefile (subdirs-dirs): Define. Mon Dec 23 20:56:49 1996 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/gnu/utmpbits.h (struct utmp): Fix padding. Mon Dec 30 11:05:11 1996 Ulrich Drepper <drepper@cygnus.com> Update from ADO tzcode1996o. * time/difftime.c: De-ANSI-declfy. Simplify computation of hibit according to ADO version. * time/zic.c: Correct typo is message. Update from ADO tzdata1996n. * time/asia: Updated. * time/australasia: Likewise. * time/europe: Likewise. * time/northamerica: Likewise. * time/southamerica: Likewise.
67 lines
2.7 KiB
C
67 lines
2.7 KiB
C
/* Copyright (C) 1991, 1994, 1996 Free Software Foundation, Inc.
|
|
This file is part of the GNU C Library.
|
|
|
|
The GNU C Library is free software; you can redistribute it and/or
|
|
modify it under the terms of the GNU Library General Public License as
|
|
published by the Free Software Foundation; either version 2 of the
|
|
License, or (at your option) any later version.
|
|
|
|
The GNU C Library is distributed in the hope that it will be useful,
|
|
but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
|
|
Library General Public License for more details.
|
|
|
|
You should have received a copy of the GNU Library General Public
|
|
License along with the GNU C Library; see the file COPYING.LIB. If not,
|
|
write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
|
|
Boston, MA 02111-1307, USA. */
|
|
|
|
#include <time.h>
|
|
#include <values.h>
|
|
|
|
|
|
/* Return the difference between TIME1 and TIME0. */
|
|
double
|
|
difftime (time1, time0)
|
|
time_t time1;
|
|
time_t time0;
|
|
{
|
|
/* Algorithm courtesy Paul Eggert (eggert@twinsun.com). */
|
|
|
|
time_t delta, hibit;
|
|
|
|
if (sizeof (time_t) < sizeof (double))
|
|
return (double) time1 - (double) time0;
|
|
if (sizeof (time_t) < sizeof (long double))
|
|
return (long double) time1 - (long double) time0;
|
|
|
|
if (time1 < time0)
|
|
return - difftime (time0, time1);
|
|
|
|
/* As much as possible, avoid loss of precision by computing the
|
|
difference before converting to double. */
|
|
delta = time1 - time0;
|
|
if (delta >= 0)
|
|
return delta;
|
|
|
|
/* Repair delta overflow. */
|
|
hibit = (~ (time_t) 0) << (_TYPEBITS (time_t) - 1);
|
|
|
|
/* The following expression rounds twice, which means the result may not
|
|
be the closest to the true answer. For example, suppose time_t is
|
|
64-bit signed int, long_double is IEEE 754 double with default
|
|
rounding, time1 = 9223372036854775807 and time0 = -1536. Then the
|
|
true difference is 9223372036854777343, which rounds to
|
|
9223372036854777856 with a total error of 513. But delta overflows to
|
|
-9223372036854774273, which rounds to -9223372036854774784, and
|
|
correcting this by subtracting 2 * (long_double) hibit (i.e. by adding
|
|
2**64 = 18446744073709551616) yields 9223372036854776832, which rounds
|
|
to 9223372036854775808 with a total error of 1535 instead. This
|
|
problem occurs only with very large differences. It's too painful to
|
|
fix this portably. We are not alone in this problem; many C compilers
|
|
round twice when converting large unsigned types to small floating
|
|
types, so if time_t is unsigned the "return delta" above has the same
|
|
double-rounding problem. */
|
|
return delta - 2 * (long double) hibit;
|
|
}
|