|File: [cvs.NetBSD.org] / src / lib / libc / time / difftime.c (download)
Revision 1.5, Sun Jul 13 20:26:48 1997 UTC (23 years ago) by christos
CVS Tags: netbsd-1-3-base, netbsd-1-3-RELEASE, netbsd-1-3-PATCH003-CANDIDATE2, netbsd-1-3-PATCH003-CANDIDATE1, netbsd-1-3-PATCH003-CANDIDATE0, netbsd-1-3-PATCH003, netbsd-1-3-PATCH002, netbsd-1-3-PATCH001, netbsd-1-3-BETA, netbsd-1-3
Changes since 1.4: +6 -1
/* $NetBSD: difftime.c,v 1.5 1997/07/13 20:26:48 christos Exp $ */
** This file is in the public domain, so clarified as of
** June 5, 1996 by Arthur David Olson (firstname.lastname@example.org).
static char elsieid = "@(#)difftime.c 7.7";
__RCSID("$NetBSD: difftime.c,v 1.5 1997/07/13 20:26:48 christos Exp $");
#endif /* !defined NOID */
#endif /* !defined lint */
** Algorithm courtesy Paul Eggert (email@example.com).
#define long_double long double
#endif /* defined HAVE_LONG_DOUBLE */
#define long_double double
#endif /* !defined HAVE_LONG_DOUBLE */
const time_t time1;
const time_t time0;
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)
** Repair delta overflow.
hibit = (~ (time_t) 0) << (TYPE_BIT(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;
** some 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 with those compilers.
return delta - 2 * (long_double) hibit;