glibc/math/math.h
Ulrich Drepper 0d8733c4fc Update.
1997-03-16 18:43  Ulrich Drepper  <drepper@cygnus.com>

	* manual/filesys.texi: Add documentation for scandir and alphasort.

	* math/math.c (fpclassify): Correct stupid typos.

	* math/libm-test.c: New file.  libm test suite by Andreas Jaeger.

	* nss/nss_files/files-hosts.c: Add gethostbyname2 imlementation.

	* posix/Makefile (routines): Add bsd-getpgrp.
	* posix/bsd-getpgrp.c: New file.
	* posix/unistd.h [__FAVOR_BSD]: Define macro getpgrp which maps
	calls to __bsd_getpgrp.

	* sysdeps/generic/getpgrp.c: De-ANSI-declfy.

	* sysdeps/i386/huge_val.h: New file.  ix87 specific infinity values.
	* sysdeps/m68k/huge_val.h: New file.  m68k specific infinity values.
	* sysdeps/generic/huge_val.h: Remove definition of long double
	definition.  Make it the same as the double definition.

	* sysdeps/libm-i387/e_acos.S: Fix bug in FPU stack handling.
	* sysdeps/libm-i387/e_acosf.S: Likewise.
	* sysdeps/libm-i387/e_acosl.S: Likewise.
	* sysdeps/libm-i387/e_asin.S: Likewise.
	* sysdeps/libm-i387/e_asinf.S: Likewise.
	* sysdeps/libm-i387/e_asinl.S: Likewise.
	* sysdeps/libm-i387/e_exp.S: Likewise.
	* sysdeps/libm-i387/e_expf.S: Likewise.
	* sysdeps/libm-i387/e_expl.S: Likewise.
	* sysdeps/libm-i387/e_scalbn.S: Likewise.
	* sysdeps/libm-i387/e_scalbnf.S: Likewise.
	* sysdeps/libm-i387/e_scalbnl.S: Likewise.

	* sysdeps/libm-i387/e_log.S: Optimize branch code.
	* 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/e_pow.S: Major rewrite to handle special cases.
	* sysdeps/libm-i387/e_powf.S: Likewise.
	* sysdeps/libm-i387/e_powl.S: Likewise.

	* sysdeps/libm-i387/e_expm1.S: Change return value for -inf
	argument to -1.0.
	* sysdeps/libm-i387/e_expm1f.S: Likewise.
	* sysdeps/libm-i387/e_expm1l.S: Likewise.

	* sysdeps/libm-i387/e_isinfl.c: Return -1 for -inf.

	* sysdeps/libm-i387/e_logbl.S: Correct return value.  Discard first
	stack element after fxtract.

	* sysdeps/libm-ieee754/e_atan2l.c: New file.  `long double'
	implementation for atan2 function.

	* sysdeps/libm-ieee754/k_standard.c: Return NAN for libm not in
	_SVID_ mode when acos, asin, atan2, log, log10 is called with
	argument out of range.
	Add new error case for pow(+0,neg).

	* sysdeps/libm-ieee754/s_fpclassifyf.c: Correct recognition of
	NaN and +-inf.
	* sysdeps/libm-ieee754/s_fpclassifyl.c: Mask out explicit leading
	digit in stupid 80 bit formats.

	* sysdeps/libm-ieee754/s_isinf.c: Rewrite to return -1 for -inf.
	* sysdeps/libm-ieee754/s_isinff.c: Likewise.
	* sysdeps/libm-ieee754/s_isinfl.c: Likewise.

	* sysdeps/libm-ieee754/s_scalbnl.c (huge, tiny): Adapt values for
	long double type.

	* sysdeps/libm-ieee754/w_atan2.c: Do not raise exception expect when
	in SVID mode.
	* sysdeps/libm-ieee754/w_atan2f.c: Likewise.
	* sysdeps/libm-ieee754/w_atan2l.c: Likewise.

	* sysdeps/libm-ieee754/w_pow.c: Distinguish error cases for x is +0
	or -0.

	* sysdeps/posix/isfdtype.c: Add cast to prevent warning.

	* sysdeps/stub/fcntlbits.h: Update copyright.
	* sysdeps/unix/bsd/fcntlbits.h: Likewise.
	* sysdeps/unix/bsd/bsd4.4/fcntlbits.h: Likewise.
	* sysdeps/unix/bsd/sun/sunos4/fcntlbits.h: Likewise.
	* sysdeps/unix/bsd/ultrix4/fcntlbits.h: Likewise.
	* sysdeps/unix/common/fcntlbits.h: Likewise.
	* sysdeps/unix/sysv/fcntlbits.h: Likewise.  Define O_FSYNC as alias
	of O_SYNC.  Add BSD compatibility macros FAPPEND, FFSYNC, FNONBLOCK,
	and FNDELAY.
	* sysdeps/unix/sysv/irix4/fcntlbits.h: Likewise.

	* sysdeps/unix/readdir_r.c: Don't copy whole `struct dirent' record,
	only reclen bytes.

	* sysdeps/unix/sysv/linux/fcntlbits.h [__USE_GNU]: Add O_READ, O_WRITE
	and O_NORW.
	* sysdeps/unix/sysv/linux/alpha/fcntlbits.h: Likewise.

	* sysdeps/unix/sysv/linux/init-first.h: Add copyright.

	* sysdeps/unix/sysv/linux/fxstat.c: New file.  Rewrite kernel-level
	struct stat to user-level form.
	* sysdeps/unix/sysv/linux/lxstat: New file.
	* sysdeps/unix/sysv/linux/xstat: New file.
	* sysdeps/unix/sysv/linux/kernel_stat.h: Define struct stat used in
	kernel.
	* sysdeps/unix/sysv/linux/statbuf.h (struct stat): Change definition
	to use prescribed types for elements.
	(_STAT_VER): Change to value 3.
	* sysdeps/unix/sysv/linux/alph/statbuf.h: Likewise.
	* sysdeps/unix/sysv/linux/Dist: Add kernel_stat.h.
	* sysdeps/unix/sysv/linux/alpha/Dist: Likewise.

	* time/Makefile: Correct dependencies for test-tz.

1997-03-16 14:59  Philip Blundell  <phil@london.uk.eu.org>

	* resolv/netdb.h: Add prototypes for gai_strerror and getnameinfo
	(needed for IPv6 basic sockets API).

1997-03-16 15:02  a sun  <asun@zoology.washington.edu>

	* sysdeps/unix/sysv/linux/net/if_ppp.h: Don't use incompatible
	kernel header.
	* sysdeps/unix/sysv/linux/net/ppp_defs.h: Likewise.

1997-03-14 17:15  Ulrich Drepper  <drepper@cygnus.com>

	* db/hash/hash_bigkey.c (__big_delete): Don't call __free_ovflpage
	without testing for last_bfp to be NULL.
	Reported by fabsoft@fabserver1.zarm.uni-bremen.de.

1997-03-13 11:42  Jim Meyering  <meyering@asic.sc.ti.com>

	* time/mktime.c (TIME_T_MIN): Work around a bug in Cray C 5.0.3.0.

1997-03-14 04:00  Kurt Garloff  <garloff@kg1.ping.de>

	* sysdeps/unix/sysv/linux/fcntlbits.h (O_FSYNC): Make alias for O_SYNC.
	(FASYNC): Move to __USE_BSD section.  Create new macro O_ASYNC.

1997-03-14 02:50  Ulrich Drepper  <drepper@cygnus.com>

	* nis/nss_nis/nis-hosts.c (_nss_nis_gethostbyname2_r): New
	functions.  Compare result for correct address type.
	(_nss_nis_gethostbyname_r): Use _nss_nis_gethostbyname2_r.
	Reported by Mirko Streckenbach <mirko@marian.hil.de>.

1997-02-17 01:40  Zlatko Calusic  <zcalusic@srce.hr>

	* time/strptime.c (recursive): Return rp to caller.
	(strptime_internal): First check for long names, then abbreviated
	(month & weekday).

1997-03-10 19:44  Andreas Schwab  <schwab@issan.informatik.uni-dortmund.de>

	* Makeconfig: Remove useless definitions of ASFLAGS-%.
	* config.make.in (ASFLAGS-.so): Remove.
	* configure.in: Don't substitute ASFLAGS_SO.
	* sysdeps/sparc/configure.in: Remove file.
	* sysdeps/sparc/Makefile (ASFLAGS-.so): Define.

1997-03-11 17:00  Andreas Schwab  <schwab@issan.informatik.uni-dortmund.de>

	* time/strptime.c (strptime_internal) [case 'Y']: Always subtract
	1900 from year, regardless of century.

1997-03-12 05:43  Ulrich Drepper  <drepper@cygnus.com>

	* stdlib/strtod.c (_tens_in_limb) [BITS_PER_MP_LIMB > 32]: Make
	all numbers unsigned to make buggy gccs happy.
	Patch by Bryan W. Headley <bheadley@interaccess.com>.

	* sysdeps/unix/sysv/linux/netinet/ip.h: Add backward-compatibility
	definitions.  Patch by a sun <asun@zoology.washington.edu>.
	Pretty print header.

	* Makerules (build-shlib): Also create symlink if library is versioned.
	based on a patch by H.J. Lu <hjl@gnu.ai.mit.edu>.
	Remove special rule to libc.so symlink.

1997-03-11 20:16  Andreas Jaeger  <aj@arthur.pfalz.de>

	* manual/math.texi (Domain and Range Errors): Change descriptions
	according to recent changes for ISO C 9X.

1997-03-11 22:39  Ulrich Drepper  <drepper@cygnus.com>

	* sysdeps/libm-ieee754/k_standard.c (__kernel_standard): Correct
	return values for acos, asin, and atan2.
	Reported by Andreas Jaeger <aj@arthur.pfalz.de>.

1997-03-10 18:16 Thorsten Kukuk  <kukuk@vt.uni-paderborn.de>

	* ypclnt.c (__yp_bind): Fix possible buffer overflow.

1997-03-10 18:06  Bernd Schmidt  <crux@Pool.Informatik.RWTH-Aachen.DE>

	* dirent/alphasort.c (alphasort): Interpret arguments as pointers
	to pointers to directory entries so that alphasort really can be
	used as argument for scandir.

1997-03-09 23:33  Andreas Jaeger  <aj@arthur.pfalz.de>

	* string/strdup.c: Declare memcpy if !(_LIBC || STDC_HEADERS)
	instead of strcpy.

1997-03-10 03:34  Ulrich Drepper  <drepper@cygnus.com>

	* catgets/catgets.c (catopen): Always add NLSPATH to search path for
	catalogs, not only if the envvar NLSPATH is not available.
	Reported by Andries.Brouwer@cwi.nl.

1997-03-10 02:46  Ulrich Drepper  <drepper@cygnus.com>

	* Makeconfig (localtime-file): Don't define using installation
	directory.
	(inst_localtime-file): New variable.
	* time/Makefile (installed-localtime-file): Use inst_localtime-file.
	Reported by Edward Seidl <seidl@janed.com>.

1997-03-10 02:31  H.J. Lu  <hjl@gnu.ai.mit.edu>

	* time/Makefile: Add source files to dependencies for test data.

1997-03-09 22:53  Thorsten Kukuk  <kukuk@weber.uni-paderborn.de>

	* nis/nss_nis/nis-ethers.c: Don't ignore return value of yp_all.
	* nis/nss_nis/nis-proto.c: Likewise.
	* nis/nss_nis/nis-rpc.c: Likewise.
	* nis/nss_nis/nis-service.c: Likewise.

1997-03-08 14:37  Miguel de Icaza  <miguel@nuclecu.unam.mx>

	* sysdeps/sparc/dl-machine.h (elf_machine_rela): Upgrade to
	versioning;  Added missing R_SPARC_WDISP30 handling.
	(RTLD_START): Implement it.

	* sysdeps/unix/sysv/linux/sparc/brk.c: Fix.

	* sysdeps/unix/sysv/linux/sparc/start.c: Startup code for
	Linux/SPARC.

1997-03-02 18:06  Miguel de Icaza  <miguel@nuclecu.unam.mx>

	* sysdeps/sparc/dl-machine.h (RTLD_START): Make arg as expected by
	the dynamic linker instead of having a new conditional define.
	Thanks to Richard Henderson for pointing this out.
	* elf/rtld.c: Remove usage of ELF_ADJUST_ARG.

1997-03-20 20:44  Thomas Bushnell, n/BSG  <thomas@gnu.ai.mit.edu>

	* sysdeps/mach/hurd/euidaccess.c: Define as __euidaccess and make
	euidaccess weak alias.

1997-03-07 10:30  Thomas Bushnell, n/BSG  <thomas@gnu.ai.mit.edu>

	* stdio-common/printf_fphex.c (MIN): New macro.

	* sysdeps/generic/netinet/in.h: Include <sys/types.h>.

	* sysdeps/generic/sys/mman.h (msync): Mention third arg.

	* sysdeps/generic/netinet/in.h: Add definitions for IPv6 basic
	API.  (See change by Philip Blundell on Feb 16, 1997.)

1997-03-05 10:40  Thomas Bushnell, n/BSG  <thomas@gnu.ai.mit.edu>

	* hurd/hurd.h (vpprintf): Include <stdarg.h>.  New declaration.

	* hurd/set-host.c (_hurd_set_host_config): Cast second arg to
	__file_name_split.

	* mach/mach_error.c (mach_error_string_int): Give full prototype.
	* mach/errstring.c (mach_error_string_int): Likewise.
	* mach/error_compat.c (__mach_error_map_compat): Likewise.
	* hurd/vpprintf.c (pwrite, vpprintf): Likewise.
	* stdio/vasprintf.c (vasprintf): Likewise.

	* mach/mach/mach_traps.h: Include <mach/kern_return.h>.

	* mach/spin-solid.c: Include <mach/mach_traps.h>.
	* mach/spin-solid.c (__spin_lock_solid): Provide arg to
	swtch_pri.

	* mach/mach_init.c: Include <mach/mig_support.h>.

	* mach/mach_error.h (mach_error_string, mach_error,
	mach_error_type): Always provide prototypes.

	* mach/mach/error.h (mach_error_fn_t): Comment out declaration; it
	appears to be entirely unused dead code.

	* stdio/stdio.h (freopen): Fix spelling error.

1997-03-02 13:38  Miles Bader  <miles@gnu.ai.mit.edu>

	* string/argz.h (__need_error_t): New macro, before including <errno.h>
	[!__const] (__const): New macro.
	[!__error_t_defined] (error_t): New typedef.

	* sysdeps/generic/socketbits.h: Add PF_FILE as synonym for PF_LOCAL
	* sysdeps/unix/sysv/linux/socketbits.h: Likewise.
1997-03-16 20:28:07 +00:00

298 lines
8.7 KiB
C

/* Declarations for math functions.
Copyright (C) 1991, 92, 93, 95, 96, 97 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. */
/*
* ISO C Standard: 4.5 MATHEMATICS <math.h>
*/
#ifndef _MATH_H
#define _MATH_H 1
#include <features.h>
__BEGIN_DECLS
/* Get machine-dependent HUGE_VAL value (returned on overflow).
On all IEEE754 machines, this is +Infinity. */
#include <huge_val.h>
/* Get machine-dependent NAN value (returned for some domain errors). */
#ifdef __USE_GNU
#include <nan.h>
#endif
/* The file <mathcalls.h> contains the prototypes for all the actual
math functions. These macros are used for those prototypes, so
we can easily declare each function as both `name' and `__name',
and can declare the float versions `namef' and `__namef'. */
#define __MATHCALL(function,suffix, args) \
__MATHDECL (_Mdouble_,function,suffix, args)
#define __MATHDECL(type, function,suffix, args) \
__MATHDECL_1(type, function,suffix, args); \
__MATHDECL_1(type, __CONCAT(__,function),suffix, args)
#define __MATHDECL_1(type, function,suffix, args) \
extern type __MATH_PRECNAME(function,suffix) args
#define _Mdouble_ double
#define __MATH_PRECNAME(name,r) __CONCAT(name,r)
#include <mathcalls.h>
#undef _Mdouble_
#undef __MATH_PRECNAME
#if defined __USE_MISC || defined __USE_ISOC9X
/* Include the file of declarations again, this time using `float'
instead of `double' and appending f to each function name. */
#ifndef _Mfloat_
#define _Mfloat_ float
#endif
#define _Mdouble_ _Mfloat_
#ifdef __STDC__
#define __MATH_PRECNAME(name,r) name##f##r
#else
#define __MATH_PRECNAME(name,r) name/**/f/**/r
#endif
#include <mathcalls.h>
#undef _Mdouble_
#undef __MATH_PRECNAME
#if __STDC__ - 0 || __GNUC__ - 0
/* Include the file of declarations again, this time using `long double'
instead of `double' and appending l to each function name. */
#ifndef _Mlong_double_
#define _Mlong_double_ long double
#endif
#define _Mdouble_ _Mlong_double_
#ifdef __STDC__
#define __MATH_PRECNAME(name,r) name##l##r
#else
#define __MATH_PRECNAME(name,r) name/**/l/**/r
#endif
#include <mathcalls.h>
#undef _Mdouble_
#undef __MATH_PRECNAME
#endif /* __STDC__ || __GNUC__ */
#endif /* Use misc or ISO C 9X. */
#undef __MATHDECL_1
#undef __MATHDECL
#undef __MATHCALL
#if defined __USE_MISC || defined __USE_XOPEN || defined __USE_ISOC9X
/* This variable is used by `gamma' and `lgamma'. */
extern int signgam;
#endif
/* ISO C 9X defines some generic macros which work on any data type. */
#if __USE_ISOC9X
/* All floating-point numbers can be put in one of these categories. */
enum
{
FP_NAN,
#define FP_NAN FP_NAN
FP_INFINITE,
#define FP_INFINITE FP_INFINITE
FP_ZERO,
#define FP_ZERO FP_ZERO
FP_SUBNORMAL,
#define FP_SUBNORMAL FP_SUBNORMAL
FP_NORMAL
#define FP_NORMAL FP_NORMAL
};
/* Return number of classification appropriate for X. */
#define fpclassify(x) \
(sizeof (x) == sizeof (float) ? \
__fpclassifyf (x) \
: sizeof (x) == sizeof (double) ? \
__fpclassify (x) : __fpclassifyl (x))
/* Return nonzero value if sign of X is negative. */
#define signbit(x) \
(sizeof (x) == sizeof (float) ? \
__signbitf (x) \
: sizeof (x) == sizeof (double) ? \
__signbit (x) : __signbitl (x))
/* Return nonzero value if X is not +-Inf or NaN. */
#define isfinite(x) (fpclassify (x) >= FP_ZERO)
/* Return nonzero value if X is neither zero, subnormal, Inf, nor NaN. */
#define isnormal(x) (fpclassify (x) == FP_NORMAL)
/* Return nonzero value if X is a NaN. We could use `fpclassify' but
we already have this functions `__isnan' and it is faster. */
#define isnan(x) \
(sizeof (x) == sizeof (float) ? \
__isnanf (x) \
: sizeof (x) == sizeof (double) ? \
__isnan (x) : __isnanl (x))
/* Conversion functions. */
/* Round X to nearest integral value according to current rounding
direction. */
extern long int rinttol __P ((double __x));
extern long long int rinttoll __P ((double __x));
/* Round X to nearest integral value, rounding halfway cases away from
zero. */
extern long int roundtol __P ((double __x));
extern long long int roundtoll __P ((double __x));
/* Comparison macros. */
/* Return nonzero value if X is greater than Y. */
#define isgreater(x, y) (!isunordered ((x), (y)) && (x) > (y))
/* Return nonzero value if X is greater than or equal to Y. */
#define isgreaterequal(x, y) (!isunordered ((x), (y)) && (x) >= (y))
/* Return nonzero value if X is less than Y. */
#define isless(x, y) (!isunordered ((x), (y)) && (x) < (y))
/* Return nonzero value if X is less than or equal to Y. */
#define islessequal(x, y) (!isunordered ((x), (y)) && (x) <= (y))
/* Return nonzero value if either X is less than Y or Y is less than X. */
#define islessgreater(x, y) \
(!isunordered ((x), (y)) && ((x) < (y) || (y) < (x)))
/* Return nonzero value if arguments are unordered. */
#define isunordered(x, y) \
(fpclassify (x) == FP_NAN || fpclassify (y) == FP_NAN)
#endif /* Use ISO C 9X. */
#ifdef __USE_MISC
/* Support for various different standard error handling behaviors. */
typedef enum { _IEEE_ = -1, _SVID_, _XOPEN_, _POSIX_ } _LIB_VERSION_TYPE;
/* This variable can be changed at run-time to any of the values above to
affect floating point error handling behavior (it may also be necessary
to change the hardware FPU exception settings). */
extern _LIB_VERSION_TYPE _LIB_VERSION;
#endif
#ifdef __USE_SVID
/* In SVID error handling, `matherr' is called with this description
of the exceptional condition.
We have a problem when using C++ since `exception' is reserved in
C++. */
#ifdef __cplusplus
struct __exception
#else
struct exception
#endif
{
int type;
char *name;
double arg1;
double arg2;
double retval;
};
#ifdef __cplusplus
extern int __matherr __P ((struct __exception *));
extern int matherr __P ((struct __exception *));
#else
extern int __matherr __P ((struct exception *));
extern int matherr __P ((struct exception *));
#endif
#define X_TLOSS 1.41484755040568800000e+16
/* Types of exceptions in the `type' field. */
#define DOMAIN 1
#define SING 2
#define OVERFLOW 3
#define UNDERFLOW 4
#define TLOSS 5
#define PLOSS 6
/* SVID mode specifies returning this large value instead of infinity. */
#define HUGE FLT_MAX
#include <float.h> /* Defines FLT_MAX. */
#else /* !SVID */
#ifdef __USE_XOPEN
/* X/Open wants another strange constant. */
#define MAXFLOAT FLT_MAX
#include <float.h>
#endif
#endif /* SVID */
#ifdef __USE_BSD
/* Some useful constants. */
#define M_E _Mldbl(2.7182818284590452354) /* e */
#define M_LOG2E _Mldbl(1.4426950408889634074) /* log 2e */
#define M_LOG10E _Mldbl(0.43429448190325182765) /* log 10e */
#define M_LN2 _Mldbl(0.69314718055994530942) /* log e2 */
#define M_LN10 _Mldbl(2.30258509299404568402) /* log e10 */
#define M_PI _Mldbl(3.14159265358979323846) /* pi */
#define M_PI_2 _Mldbl(1.57079632679489661923) /* pi/2 */
#define M_PI_4 _Mldbl(0.78539816339744830962) /* pi/4 */
#define M_1_PI _Mldbl(0.31830988618379067154) /* 1/pi */
#define M_2_PI _Mldbl(0.63661977236758134308) /* 2/pi */
#define M_2_SQRTPI _Mldbl(1.12837916709551257390) /* 2/sqrt(pi) */
#define M_SQRT2 _Mldbl(1.41421356237309504880) /* sqrt(2) */
#define M_SQRT1_2 _Mldbl(0.70710678118654752440) /* 1/sqrt(2) */
/* Our constants might specify more precision than `double' can represent.
Use `long double' constants in standard and GNU C, where they are
supported and the cast to `double'. */
#if __STDC__ - 0 || __GNUC__ - 0
#define _Mldbl(x) x##L
#else /* Traditional C. */
#define _Mldbl(x) x
#endif /* Standard or GNU C. */
#endif
/* Get machine-dependent inline versions (if there are any). */
#if (!defined __NO_MATH_INLINES && defined __OPTIMIZE__) \
|| defined __LIBC_M81_MATH_INLINES
#include <__math.h>
#endif
__END_DECLS
#endif /* math.h */