glibc/sysdeps/ieee754/dbl-64/e_gamma_r.c

221 lines
6.0 KiB
C
Raw Normal View History

Update. 1997-06-19 19:38 Ulrich Drepper <drepper@cygnus.com> * features.h: Define __STDC_IEC_559__ and _STDC_IEC_559_COMPLEX__. * elf/dl-minimal.c (__dcgettext): Remove assertion. * inet/rcmd.c: Correct a few typos. Reported by Eric Troan. * manual/Makefile (distribute): Add dir. * manual/dir: New file. * math/Makefile (libm-support): Rename s_rinttol, s_rinttoll, s_roundtol, and s_roundtoll to s_lrint, s_llrint, s_lround, and s_llround respectively. (libm-calls): Add e_gamma_r. * math/libm-test.c (check_int_exc): New function. (signbit_test): Remove test for sign of NaN. (gamma_test): Clear exception after test of existence. Correct tests and and epsilons. (lgamma_test): Likewise. (ilogb_test): Correct all tests. (scalb_test): Rewrite. (rinttol_test): Rename to lrint_test and correct tests. (rinttoll_test): Likewise. (roundtol_test): Likewise. (roundtoll_test): Likewise. (main): Call lrint/lround functions instead of rinttol/roundtol. * math/math.h: Change prototypes for rinttol/roundtol. * math/mathcalls.h: Rearrange prototypes according to ISO C9X draft. * sysdeps/generic/mathbits.h: Define FP_ILOGB0 and FP_ILOGBNAN. * sysdeps/i386/mathbits.h: Likewise. * sysdeps/libm-i387/e_scalb.S: Handle special cases correctly. * sysdeps/libm-i387/e_scalbf.S: Likewise. * sysdeps/libm-i387/e_scalbl.S: Likewise. * sysdeps/libm-i387/s_asinh.S: Handle -inf correctly. * sysdeps/libm-i387/s_asinhf.S: Likewise. * sysdeps/libm-i387/s_asinhl.S: Likewise. * sysdeps/libm-i387/s_ilogb.S: Optimize. * sysdeps/libm-i387/s_ilogbf.S: Likewise. * sysdeps/libm-i387/s_ilogbl.S: Likewise. * sysdeps/libm-i387/s_rinttol.S: Rename to... * sysdeps/libm-i387/s_lrint.S: ...this. * sysdeps/libm-i387/s_rinttoll.S: Rename to... * sysdeps/libm-i387/s_llrint.S: ...this. * sysdeps/libm-i387/s_remquo.S: Correctly set sign of remainder. * sysdeps/libm-i387/s_remquof.S: Likewise. * sysdeps/libm-i387/s_remquol.S: Likewise. * sysdeps/libm-i387/e_gamma_r.c: New file. Implementation of gamma function according to ISO C. * sysdeps/libm-i387/e_gammaf_r.c: New file. * sysdeps/libm-i387/e_gammal_r.c: New file. * sysdeps/libm-i387/e_lgamma_r.c: Don't let optimize compile the generation of exceptions away. * sysdeps/libm-i387/e_lgammaf_r.c: Likewise. * sysdeps/libm-i387/k_standard.c: Correct return value for infinity points of gamma function when not SVID mode. * sysdeps/libm-i387/s_rinttoll.c: Renamed to... * sysdeps/libm-i387/s_llrint.c: ...this. * sysdeps/libm-i387/s_rinttol.c: Renamed to... * sysdeps/libm-i387/s_lrint.c: ...this. * sysdeps/libm-i387/s_roundtoll.c: Renamed to... * sysdeps/libm-i387/s_llround.c: ...this. * sysdeps/libm-i387/s_roundtol.c: Renamed to... * sysdeps/libm-i387/s_lround.c: ..this. * sysdeps/libm-i387/s_scalbn.c: Change second parameter according to ISO C. * sysdeps/libm-i387/s_scalbnf.c: Likewise. * sysdeps/libm-i387/s_scalbnl.c: Likewise. * sysdeps/libm-i387/w_gamma.c: Call __ieee754_gamma_r if library mode is not _SVID_. * sysdeps/libm-i387/w_gammaf.c: Likewise. * sysdeps/libm-i387/w_gammal.c: Likewise. * sysdeps/m68k/fpu/__math.h: Rename __rinttol to __lrint and rinttol to lrint. * sysdeps/m68k/fpu/s_rinttol.c: Renamed to... * sysdeps/m68k/fpu/s_lrint.c: ...this. * sysdeps/m68k/fpu/s_rinttoll.c: Renamed to... * sysdeps/m68k/fpu/s_llrint.c: ...this. * md5-crypt/Makefile: Link md5test program with md5.o. * stdio-common/temptest.c: Don't use __stdio_gen_tempname which is not exported by the libc.so. * stdio-common/vfscanf.c: Correct scanning of strings after last change. * sysdeps/unix/sysv/linux/i386/sysdep.S: Use .comm to define errno. 1997-06-19 07:37 H.J. Lu <hjl@gnu.ai.mit.edu> * time/tzfile.c (__tzfile_read): Store getc () return in int. 1997-06-13 Miles Bader <miles@gnu.ai.mit.edu> * argp/argp-parse.c (argp_version_parser): Include `(PROGRAM ERROR)' in the no-version error text to indicate that something's fucked. [!_] (N_): New macro. (argp_default_options, argp_version_options): Wrap doc strings in N_(). 1997-06-12 Miles Bader <miles@gnu.ai.mit.edu> * argp/argp-parse.c (parser_parse_opt): Detect and report unhandled options here. (parser_parse_arg): Handle ARGP_KEY_ARGS here. Adjust NEXT pointer back if we fail to parse anything. (parser_parse_next): Simplify arg code. Leave state NEXT frobbing to parser_parse_arg. 1997-06-11 Miles Bader <miles@gnu.ai.mit.edu> * argp/argp.h (ARGP_KEY_ARGS, ARGP_KEY_FINI): New macros. * argp/argp-parse.c (parser_finalize): Do another pass over the parsers with ARGP_KEY_FINI. 1997-06-18 Miles Bader <miles@gnu.ai.mit.edu> * string/Makefile (routines): Add argz-replace. 1997-06-16 00:16 Miles Bader <miles@gnu.ai.mit.edu> * manual/string.texi (Argz Functions): Document argz_replace. 1997-06-12 Miles Bader <miles@gnu.ai.mit.edu> * string/argz.h (__argz_replace, argz_replace): New declarations. * string/argz-replace.c: New file. 1997-06-11 Miles Bader <miles@gnu.ai.mit.edu> * manual/argp.texi (Argp Special Keys): Document ARGP_KEY_ARGS. 1997-06-16 23:08 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * Makeconfig (libc-map): Remove definition. * Makerules (libc-map): Define it here, using the full name. (load-map-file): Remove case for empty $(..). ($(common-objpfx)libc.so): Revert last change. 1997-06-17 22:18 Mark Kettenis <kettenis@phys.uva.nl> * login/programs/utmpd.c (main): Improve signal handling. * login/programs/request.c (do_setutent, do_updwtmp): Allow arbitrary length filenames. * login/programs/utmpd.h (setutent_request, updwtmp_request): Get rid of fixed length file field. * login/utmp_daemon.c (do_setutent, do_updwtmp): Allow arbitrary length filenames. * login/programs/request.c (do_pututline): Don't fail if connection->position is -1 on entry. 1997-06-15 16:32 Mark Kettenis <kettenis@phys.uva.nl> * login/utmp_file.c (updwtmp_file): Use the same method for appending an entry as in pututline_file. 1997-06-11 18:59 Mark Kettenis <kettenis@phys.uva.nl> * login/utmp_file.c (getutent_r_file): Use read lock instead of write lock. (getutline_r_file, internal_getut_r): Lock utmp file. (updwtmp_file): Use fcntl to lock file instead of flock. 1997-06-18 00:11 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * stdio-common/vfscanf.c (inchar, ungetc): Don't count EOF as character read in/put back. * stdio-common/tstscanf.c: Add test case for this. 1997-06-17 22:17 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * libc.map: Add more symbols. 1997-06-18 12:01 Ulrich Drepper <drepper@cygnus.com> * manual/Makefile (install): Make sure `dir' file exists if we use install-info. 1997-06-17 19:32 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * manual/Makefile (info): Depend on dir-add.info. 1997-06-17 17:19 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * elf/ldd.bash.in: Fix spacing in message. 1997-06-17 14:28 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * elf/do-rel.h (elf_dynamic_do_rel): Always use version information if available. 1997-06-17 11:34 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sunrpc/Makefile ($(rpcsvc:%.x=$(objpfx)rpcsvc/%.h)): Make command non-empty to force make to recheck modification time. ($(rpcsvc:%.x=$(objpfx)x%.c)): Likewise. 1997-06-17 00:26 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/stub/e_acoshl.c: Set errno to ENOSYS. * sysdeps/stub/e_acosl.c: Likewise. * sysdeps/stub/e_asinl.c: Likewise. * sysdeps/stub/e_atan2l.c: Likewise. * sysdeps/stub/e_expl.c: Likewise. * sysdeps/stub/e_fmodl.c: Likewise. * sysdeps/stub/e_j0l.c: Likewise. * sysdeps/stub/e_j1l.c: Likewise. * sysdeps/stub/e_jnl.c: Likewise. * sysdeps/stub/e_lgammal_r.c: Likewise. * sysdeps/stub/e_log10l.c: Likewise. * sysdeps/stub/e_logl.c: Likewise. * sysdeps/stub/e_powl.c: Likewise. * sysdeps/stub/e_rem_pio2l.c: Likewise. * sysdeps/stub/e_sqrtl.c: Likewise. * sysdeps/stub/k_cosl.c: Likewise. * sysdeps/stub/k_rem_pio2l.c: Likewise. * sysdeps/stub/k_sinl.c: Likewise. * sysdeps/stub/k_tanl.c: Likewise. * sysdeps/stub/s_atanl.c: Likewise. * sysdeps/stub/s_erfl.c: Likewise. * sysdeps/stub/s_exp2.c: Likewise. * sysdeps/stub/s_exp2f.c: Likewise. * sysdeps/stub/s_exp2l.c: Likewise. * sysdeps/stub/s_expm1l.c: Likewise. * sysdeps/stub/s_log1pl.c: Likewise. * sysdeps/stub/s_log2l.c: Likewise. 1997-06-18 11:46 Ulrich Drepper <drepper@cygnus.com> * sysdeps/libm-ieee754/w_gamma.c: If _LIB_VERSION is _SVID_ compute result as before last change. * sysdeps/libm-ieee754/w_gammaf.c: Likewise. * sysdeps/libm-ieee754/w_gammal.c: Likewise. 1997-06-16 23:37 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/libm-ieee754/s_remquo.c: Fix off-by-one when computing quotient. * sysdeps/libm-ieee754/s_remquof.c: Likewise. * sysdeps/libm-ieee754/s_remquol.c: Likewise. * sysdeps/m68k/fpu/s_remquo.c: Remove FIXME and special case for quotient. * sysdeps/libm-ieee754/w_gamma.c: Add missing call to exp function. Don't use global signgam. * sysdeps/libm-ieee754/w_gammaf.c: Likewise. * sysdeps/libm-ieee754/w_gammal.c: Likewise. * math/Makefile (libm-calls): Remove w_gamma_r. * sysdeps/libm-ieee754/w_gamma_r.c: Remove file. * sysdeps/libm-ieee754/w_gammaf_r.c: Remove file. * sysdeps/libm-ieee754/w_gammal_r.c: Remove file. * math/libm-test.c (atanh_test): Declare x only if needed. (signbit_test): Fix typo. (gamma_test): Check whether function is implemented. Add epsilons. (lgamma_test): Likewise. (fmod_test): Add epsilons. (exp2_test): Use right function for existence test. 1997-06-07 09:20 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/powerpc/Dist: Add fenv_const.c. * sysdeps/unix/sysv/linux/Dist: Add net/if_slip.h. * sysdeps/unix/sysv/linux/powerpc/Dist: Add init-first.h and syscall.h. * sysdeps/unix/sysv/linux/sparc/Dist: Add init-first.h. * string/Makefile (distribute): Add tst-svc.expect. * nis/Makefile (distribute): Add nis_intern.h and Banner. * elf/Makefile (distribute): Add dl-hash.h. * Rules (subdir_echo-distinfo): Add headers from $(distribute). * login/Makefile (others): Add utmpdump. * sysdeps/unix/sysv/linux/Makefile (sysdep_headers): Add net/if_slip.h. * manual/Makefile (dir-add.texi): Also look in indirectly included files. 1997-06-16 23:15 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/unix/sysv/linux/configure.in: Remove commands to generate stdio_lim.h. * sysdeps/unix/sysv/linux/mk-stdiolim.c: Remove. * sysdeps/unix/sysv/linux/Makefile: Add rules to generate stdio_lim.h here. (common-generated): Add generated files. (inhibit-stdio_lim): Define. * sysdeps/posix/Makefile [$(inhibit-stdio_lim)=yes]: Disable rules to generate stdio_lim.h. * sysdeps/unix/sysv/linux/Makefile: Suppress inclusion of dependecy files if no_deps is set. ($(objpfx)syscall-%.d): Add header file as target to dependency generation. 1997-06-14 19:19 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/unix/sysv/linux/arm/Dist: New file. 1997-06-14 17:59 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * extra-lib.mk (others): Depend on versioned shared library, not the unversioned one. * Makerules (build-shlib): Don't make the version link here. ($(common-objpfx)libc.so$(libc.so-version)): New rule for libc version link. 1997-06-16 03:07 Ulrich Drepper <drepper@cygnus.com> * sysdeps/libm-ieee754/k_standard.c: Undo change of Tue Aug 6 01:13:56 1996. * argp/argp-help.c (argp_args_usage): Supply correct argp to filter_doc. * argp/argp-help.c (hol_add_cluster): Initialize CL->depth. * argp/argp-help.c (_help): Supply STATE to argp_args_usage. * argp/argp.h (ARGP_KEY_HELP_ARGS_DOC): New macro. * argp/argp-fmtstream.c: Add casts to prevent warnings. * argp/argp.h (OPTION_NO_USAGE): New macro. * argp/argp-help.c (usage_long_opt, usage_argful_short_opt, * argp/argp-fmtstream.c (__argp_fmtstream_update): Account for case * argp/argp-help.c <stddef.h>: New include. * argp/argp.h (argp_state_help, __argp_state_help, argp_usage, * argp/argp.h (argp_program_bug_address): Make const. * argp/argp-parse.c (argp_default_parser): Set STATE->name for OPT_PROGNAME. * argp/argp-help.c (__argp_error, __argp_failure, __argp_state_help): * argp/argp-parse.c (parser_init): Set PARSER->state.flags.
1997-06-19 19:11:22 +00:00
/* Implementation of gamma function according to ISO C.
Copyright (C) 1997-2017 Free Software Foundation, Inc.
Update. 1997-04-09 01:24 Ulrich Drepper <drepper@cygnus.com> * rellns-sh: Rewrite to work also in presence of symlinks. * arpg/argp-fmtstream.c: Add casts to prevent warnings. * argp/argp-fmtstream.h: Likewise. * argp/argp-help.c: Likewise. * elf/dl-minimal.c: Add definition of calloc. * elf/version.c: Add casts to prevent warnings. (_dl_check_map_versions): Use calloc instead of malloc+memset. * locale/setlocale.c (_nl_current): Add element with index LC_ALL. Reported by Greg McGary <gkm@eng.ascend.com>. * manual/libc.texinfo: Update malloc documentation for new malloc. * manual/memory.texi: Likewise. Patch by Wolfram Gloger <wmglo@dent.med.uni-muenchen.de>. * math/libm-test.c (check_long): New function. (check_longlong): New function. (rinttol_test): New function. (rinttoll_test): New function. * nis/nss_compat/compat-grp.c (in_blacklist): Improve a bit. * nis/nss_compat/compat-pwd.c: Likewise. * nis/nss_compat/compat-spwd.c: Likewise. * stdlib/erand48_r.c (erand48_r): Build double value using ieee754_double union and use random bits in different order to increase effect of seed. Reported by David Mosberger-Tang <davidm@AZStarNet.com>. * sunrpc/svc_auth.c: Moved to ... * sysdeps/generic/svc_auth.c: ...here. * time/time.h: Pretty print. 1997-04-08 07:19 H.J. Lu <hjl@gnu.ai.mit.edu> * libio/genops.c (_IO_flush_all_linebuffered): don't flush on a read-only stream. 1997-04-09 01:19 Ulrich Drepper <drepper@cygnus.com> * malloc/malloc.c (mALLOC_STATs) [MALLOC_DEBUG>1]: Put declaration in correct place. Patch by Marcus G. Daniels <marcus@cathcart.sysc.pdx.edu>. 1997-04-07 15:34 Ulrich Drepper <drepper@cygnus.com> * stdio-common/Makefile (tests): Add tst-ferror. * stdio-common/tst-ferror.c: New file. Some tests for error indicator of streams. * stdio-common/tst-ferror.input: New file. * isomac.c: Let tests not fail because the compiler defines itself symbols which violate the name space rules. gcc defines symbols for the architecture which are not protected by an underscore character. * math/Makefile (libm-support): Add s_rinttol and s_rinttoll. (libm-calls): Add s_clog. * sysdeps/libm-ieee754/s_clog.c: New file. Implementation of logarithm of complex value. * sysdeps/libm-ieee754/s_clogf.c: New file. * sysdeps/libm-ieee754/s_clogl.c: New file. * math/libm-test.c (clog_test): Compile this function. Fix a few typos. (main): Call clog_test. * sysdeps/libm-ieee754/s_rinttol.c: New file. Round long double value to long int. * sysdeps/libm-i387/s_rinttol.S: New file. * sysdeps/libm-ieee754/s_rinttoll.c: new file. Round long double value to long long int. * sysdeps/libm-i387/s_rinttoll.S: New file. * sysdeps/libm-ieee754/s_rintl.c: Many corrections. The previous version was full of errors. * math/math.h (rinttol): Argument is of type `long double' not `double'. (rinttoll): Likewise. (roundtol): Likewise. (roundtoll): Likewise. 1997-04-06 11:32 H.J. Lu <hjl@gnu.ai.mit.edu> * posix/getopt.c (_getopt_initialize): Preserve optind. (_getopt_internal): Set optind to 1 if optind == 0 before calling _getopt_initialize (). 1997-04-05 16:45 Thorsten Kukuk <kukuk@vt.uni-paderborn.de> * nis/rpcsvc/nislib.h: Change const nis_name to new type const_nis_name. * nis/nis_intern.c: Likewise. * nis/nis_intern.h: Likewise. * nis/nis_server.c: Likewise. * nis/nis_subr.c: Likewise. * nis/nis_table.c: Likewise. * nis/nis_names.c: Likewise. Fill out ns_request structure in nis_add(). * nis/nss_compat/compat-pwd.c: Use reentrant netgroup functions. * nis/nss_compat/compat-spwd.c: Likewise. 1997-03-27 07:37 H.J. Lu <hjl@gnu.ai.mit.edu> * libio/fileops.c (_IO_file_overflow): Set error when try to write on a read-only stream. * sysdeps/gnu/utmpbits.h (ut_xtime): New symbol. (ut_time): Define it only if _NO_UT_TIME is not defined. 1997-04-06 00:42 Ulrich Drepper <drepper@cygnus.com> * misc/tst-tsearch.c: Include <string.h>. Define _GNU_SOURCE only if not already defined. 1997-04-05 16:14 Ulrich Drepper <drepper@cygnus.com> * sysdeps/unix/sysv/linux/netatalk/at.h: Include <sys/socket.h> to get definition of sa_family_t for <linux/atalk.h>. Reported by a sun <asun@zoology.washington.edu>. * malloc/malloc.c (cALLOc): Little optimization.
1997-04-08 23:42:08 +00:00
This file is part of the GNU C Library.
Contributed by Ulrich Drepper <drepper@cygnus.com>, 1997.
The GNU C Library is free software; you can redistribute it and/or
modify it under the terms of the GNU Lesser General Public
License as published by the Free Software Foundation; either
version 2.1 of the License, or (at your option) any later version.
Update. 1997-04-09 01:24 Ulrich Drepper <drepper@cygnus.com> * rellns-sh: Rewrite to work also in presence of symlinks. * arpg/argp-fmtstream.c: Add casts to prevent warnings. * argp/argp-fmtstream.h: Likewise. * argp/argp-help.c: Likewise. * elf/dl-minimal.c: Add definition of calloc. * elf/version.c: Add casts to prevent warnings. (_dl_check_map_versions): Use calloc instead of malloc+memset. * locale/setlocale.c (_nl_current): Add element with index LC_ALL. Reported by Greg McGary <gkm@eng.ascend.com>. * manual/libc.texinfo: Update malloc documentation for new malloc. * manual/memory.texi: Likewise. Patch by Wolfram Gloger <wmglo@dent.med.uni-muenchen.de>. * math/libm-test.c (check_long): New function. (check_longlong): New function. (rinttol_test): New function. (rinttoll_test): New function. * nis/nss_compat/compat-grp.c (in_blacklist): Improve a bit. * nis/nss_compat/compat-pwd.c: Likewise. * nis/nss_compat/compat-spwd.c: Likewise. * stdlib/erand48_r.c (erand48_r): Build double value using ieee754_double union and use random bits in different order to increase effect of seed. Reported by David Mosberger-Tang <davidm@AZStarNet.com>. * sunrpc/svc_auth.c: Moved to ... * sysdeps/generic/svc_auth.c: ...here. * time/time.h: Pretty print. 1997-04-08 07:19 H.J. Lu <hjl@gnu.ai.mit.edu> * libio/genops.c (_IO_flush_all_linebuffered): don't flush on a read-only stream. 1997-04-09 01:19 Ulrich Drepper <drepper@cygnus.com> * malloc/malloc.c (mALLOC_STATs) [MALLOC_DEBUG>1]: Put declaration in correct place. Patch by Marcus G. Daniels <marcus@cathcart.sysc.pdx.edu>. 1997-04-07 15:34 Ulrich Drepper <drepper@cygnus.com> * stdio-common/Makefile (tests): Add tst-ferror. * stdio-common/tst-ferror.c: New file. Some tests for error indicator of streams. * stdio-common/tst-ferror.input: New file. * isomac.c: Let tests not fail because the compiler defines itself symbols which violate the name space rules. gcc defines symbols for the architecture which are not protected by an underscore character. * math/Makefile (libm-support): Add s_rinttol and s_rinttoll. (libm-calls): Add s_clog. * sysdeps/libm-ieee754/s_clog.c: New file. Implementation of logarithm of complex value. * sysdeps/libm-ieee754/s_clogf.c: New file. * sysdeps/libm-ieee754/s_clogl.c: New file. * math/libm-test.c (clog_test): Compile this function. Fix a few typos. (main): Call clog_test. * sysdeps/libm-ieee754/s_rinttol.c: New file. Round long double value to long int. * sysdeps/libm-i387/s_rinttol.S: New file. * sysdeps/libm-ieee754/s_rinttoll.c: new file. Round long double value to long long int. * sysdeps/libm-i387/s_rinttoll.S: New file. * sysdeps/libm-ieee754/s_rintl.c: Many corrections. The previous version was full of errors. * math/math.h (rinttol): Argument is of type `long double' not `double'. (rinttoll): Likewise. (roundtol): Likewise. (roundtoll): Likewise. 1997-04-06 11:32 H.J. Lu <hjl@gnu.ai.mit.edu> * posix/getopt.c (_getopt_initialize): Preserve optind. (_getopt_internal): Set optind to 1 if optind == 0 before calling _getopt_initialize (). 1997-04-05 16:45 Thorsten Kukuk <kukuk@vt.uni-paderborn.de> * nis/rpcsvc/nislib.h: Change const nis_name to new type const_nis_name. * nis/nis_intern.c: Likewise. * nis/nis_intern.h: Likewise. * nis/nis_server.c: Likewise. * nis/nis_subr.c: Likewise. * nis/nis_table.c: Likewise. * nis/nis_names.c: Likewise. Fill out ns_request structure in nis_add(). * nis/nss_compat/compat-pwd.c: Use reentrant netgroup functions. * nis/nss_compat/compat-spwd.c: Likewise. 1997-03-27 07:37 H.J. Lu <hjl@gnu.ai.mit.edu> * libio/fileops.c (_IO_file_overflow): Set error when try to write on a read-only stream. * sysdeps/gnu/utmpbits.h (ut_xtime): New symbol. (ut_time): Define it only if _NO_UT_TIME is not defined. 1997-04-06 00:42 Ulrich Drepper <drepper@cygnus.com> * misc/tst-tsearch.c: Include <string.h>. Define _GNU_SOURCE only if not already defined. 1997-04-05 16:14 Ulrich Drepper <drepper@cygnus.com> * sysdeps/unix/sysv/linux/netatalk/at.h: Include <sys/socket.h> to get definition of sa_family_t for <linux/atalk.h>. Reported by a sun <asun@zoology.washington.edu>. * malloc/malloc.c (cALLOc): Little optimization.
1997-04-08 23:42:08 +00:00
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
Lesser General Public License for more details.
Update. 1997-04-09 01:24 Ulrich Drepper <drepper@cygnus.com> * rellns-sh: Rewrite to work also in presence of symlinks. * arpg/argp-fmtstream.c: Add casts to prevent warnings. * argp/argp-fmtstream.h: Likewise. * argp/argp-help.c: Likewise. * elf/dl-minimal.c: Add definition of calloc. * elf/version.c: Add casts to prevent warnings. (_dl_check_map_versions): Use calloc instead of malloc+memset. * locale/setlocale.c (_nl_current): Add element with index LC_ALL. Reported by Greg McGary <gkm@eng.ascend.com>. * manual/libc.texinfo: Update malloc documentation for new malloc. * manual/memory.texi: Likewise. Patch by Wolfram Gloger <wmglo@dent.med.uni-muenchen.de>. * math/libm-test.c (check_long): New function. (check_longlong): New function. (rinttol_test): New function. (rinttoll_test): New function. * nis/nss_compat/compat-grp.c (in_blacklist): Improve a bit. * nis/nss_compat/compat-pwd.c: Likewise. * nis/nss_compat/compat-spwd.c: Likewise. * stdlib/erand48_r.c (erand48_r): Build double value using ieee754_double union and use random bits in different order to increase effect of seed. Reported by David Mosberger-Tang <davidm@AZStarNet.com>. * sunrpc/svc_auth.c: Moved to ... * sysdeps/generic/svc_auth.c: ...here. * time/time.h: Pretty print. 1997-04-08 07:19 H.J. Lu <hjl@gnu.ai.mit.edu> * libio/genops.c (_IO_flush_all_linebuffered): don't flush on a read-only stream. 1997-04-09 01:19 Ulrich Drepper <drepper@cygnus.com> * malloc/malloc.c (mALLOC_STATs) [MALLOC_DEBUG>1]: Put declaration in correct place. Patch by Marcus G. Daniels <marcus@cathcart.sysc.pdx.edu>. 1997-04-07 15:34 Ulrich Drepper <drepper@cygnus.com> * stdio-common/Makefile (tests): Add tst-ferror. * stdio-common/tst-ferror.c: New file. Some tests for error indicator of streams. * stdio-common/tst-ferror.input: New file. * isomac.c: Let tests not fail because the compiler defines itself symbols which violate the name space rules. gcc defines symbols for the architecture which are not protected by an underscore character. * math/Makefile (libm-support): Add s_rinttol and s_rinttoll. (libm-calls): Add s_clog. * sysdeps/libm-ieee754/s_clog.c: New file. Implementation of logarithm of complex value. * sysdeps/libm-ieee754/s_clogf.c: New file. * sysdeps/libm-ieee754/s_clogl.c: New file. * math/libm-test.c (clog_test): Compile this function. Fix a few typos. (main): Call clog_test. * sysdeps/libm-ieee754/s_rinttol.c: New file. Round long double value to long int. * sysdeps/libm-i387/s_rinttol.S: New file. * sysdeps/libm-ieee754/s_rinttoll.c: new file. Round long double value to long long int. * sysdeps/libm-i387/s_rinttoll.S: New file. * sysdeps/libm-ieee754/s_rintl.c: Many corrections. The previous version was full of errors. * math/math.h (rinttol): Argument is of type `long double' not `double'. (rinttoll): Likewise. (roundtol): Likewise. (roundtoll): Likewise. 1997-04-06 11:32 H.J. Lu <hjl@gnu.ai.mit.edu> * posix/getopt.c (_getopt_initialize): Preserve optind. (_getopt_internal): Set optind to 1 if optind == 0 before calling _getopt_initialize (). 1997-04-05 16:45 Thorsten Kukuk <kukuk@vt.uni-paderborn.de> * nis/rpcsvc/nislib.h: Change const nis_name to new type const_nis_name. * nis/nis_intern.c: Likewise. * nis/nis_intern.h: Likewise. * nis/nis_server.c: Likewise. * nis/nis_subr.c: Likewise. * nis/nis_table.c: Likewise. * nis/nis_names.c: Likewise. Fill out ns_request structure in nis_add(). * nis/nss_compat/compat-pwd.c: Use reentrant netgroup functions. * nis/nss_compat/compat-spwd.c: Likewise. 1997-03-27 07:37 H.J. Lu <hjl@gnu.ai.mit.edu> * libio/fileops.c (_IO_file_overflow): Set error when try to write on a read-only stream. * sysdeps/gnu/utmpbits.h (ut_xtime): New symbol. (ut_time): Define it only if _NO_UT_TIME is not defined. 1997-04-06 00:42 Ulrich Drepper <drepper@cygnus.com> * misc/tst-tsearch.c: Include <string.h>. Define _GNU_SOURCE only if not already defined. 1997-04-05 16:14 Ulrich Drepper <drepper@cygnus.com> * sysdeps/unix/sysv/linux/netatalk/at.h: Include <sys/socket.h> to get definition of sa_family_t for <linux/atalk.h>. Reported by a sun <asun@zoology.washington.edu>. * malloc/malloc.c (cALLOc): Little optimization.
1997-04-08 23:42:08 +00:00
You should have received a copy of the GNU Lesser General Public
License along with the GNU C Library; if not, see
<http://www.gnu.org/licenses/>. */
Update. 1997-04-09 01:24 Ulrich Drepper <drepper@cygnus.com> * rellns-sh: Rewrite to work also in presence of symlinks. * arpg/argp-fmtstream.c: Add casts to prevent warnings. * argp/argp-fmtstream.h: Likewise. * argp/argp-help.c: Likewise. * elf/dl-minimal.c: Add definition of calloc. * elf/version.c: Add casts to prevent warnings. (_dl_check_map_versions): Use calloc instead of malloc+memset. * locale/setlocale.c (_nl_current): Add element with index LC_ALL. Reported by Greg McGary <gkm@eng.ascend.com>. * manual/libc.texinfo: Update malloc documentation for new malloc. * manual/memory.texi: Likewise. Patch by Wolfram Gloger <wmglo@dent.med.uni-muenchen.de>. * math/libm-test.c (check_long): New function. (check_longlong): New function. (rinttol_test): New function. (rinttoll_test): New function. * nis/nss_compat/compat-grp.c (in_blacklist): Improve a bit. * nis/nss_compat/compat-pwd.c: Likewise. * nis/nss_compat/compat-spwd.c: Likewise. * stdlib/erand48_r.c (erand48_r): Build double value using ieee754_double union and use random bits in different order to increase effect of seed. Reported by David Mosberger-Tang <davidm@AZStarNet.com>. * sunrpc/svc_auth.c: Moved to ... * sysdeps/generic/svc_auth.c: ...here. * time/time.h: Pretty print. 1997-04-08 07:19 H.J. Lu <hjl@gnu.ai.mit.edu> * libio/genops.c (_IO_flush_all_linebuffered): don't flush on a read-only stream. 1997-04-09 01:19 Ulrich Drepper <drepper@cygnus.com> * malloc/malloc.c (mALLOC_STATs) [MALLOC_DEBUG>1]: Put declaration in correct place. Patch by Marcus G. Daniels <marcus@cathcart.sysc.pdx.edu>. 1997-04-07 15:34 Ulrich Drepper <drepper@cygnus.com> * stdio-common/Makefile (tests): Add tst-ferror. * stdio-common/tst-ferror.c: New file. Some tests for error indicator of streams. * stdio-common/tst-ferror.input: New file. * isomac.c: Let tests not fail because the compiler defines itself symbols which violate the name space rules. gcc defines symbols for the architecture which are not protected by an underscore character. * math/Makefile (libm-support): Add s_rinttol and s_rinttoll. (libm-calls): Add s_clog. * sysdeps/libm-ieee754/s_clog.c: New file. Implementation of logarithm of complex value. * sysdeps/libm-ieee754/s_clogf.c: New file. * sysdeps/libm-ieee754/s_clogl.c: New file. * math/libm-test.c (clog_test): Compile this function. Fix a few typos. (main): Call clog_test. * sysdeps/libm-ieee754/s_rinttol.c: New file. Round long double value to long int. * sysdeps/libm-i387/s_rinttol.S: New file. * sysdeps/libm-ieee754/s_rinttoll.c: new file. Round long double value to long long int. * sysdeps/libm-i387/s_rinttoll.S: New file. * sysdeps/libm-ieee754/s_rintl.c: Many corrections. The previous version was full of errors. * math/math.h (rinttol): Argument is of type `long double' not `double'. (rinttoll): Likewise. (roundtol): Likewise. (roundtoll): Likewise. 1997-04-06 11:32 H.J. Lu <hjl@gnu.ai.mit.edu> * posix/getopt.c (_getopt_initialize): Preserve optind. (_getopt_internal): Set optind to 1 if optind == 0 before calling _getopt_initialize (). 1997-04-05 16:45 Thorsten Kukuk <kukuk@vt.uni-paderborn.de> * nis/rpcsvc/nislib.h: Change const nis_name to new type const_nis_name. * nis/nis_intern.c: Likewise. * nis/nis_intern.h: Likewise. * nis/nis_server.c: Likewise. * nis/nis_subr.c: Likewise. * nis/nis_table.c: Likewise. * nis/nis_names.c: Likewise. Fill out ns_request structure in nis_add(). * nis/nss_compat/compat-pwd.c: Use reentrant netgroup functions. * nis/nss_compat/compat-spwd.c: Likewise. 1997-03-27 07:37 H.J. Lu <hjl@gnu.ai.mit.edu> * libio/fileops.c (_IO_file_overflow): Set error when try to write on a read-only stream. * sysdeps/gnu/utmpbits.h (ut_xtime): New symbol. (ut_time): Define it only if _NO_UT_TIME is not defined. 1997-04-06 00:42 Ulrich Drepper <drepper@cygnus.com> * misc/tst-tsearch.c: Include <string.h>. Define _GNU_SOURCE only if not already defined. 1997-04-05 16:14 Ulrich Drepper <drepper@cygnus.com> * sysdeps/unix/sysv/linux/netatalk/at.h: Include <sys/socket.h> to get definition of sa_family_t for <linux/atalk.h>. Reported by a sun <asun@zoology.washington.edu>. * malloc/malloc.c (cALLOc): Little optimization.
1997-04-08 23:42:08 +00:00
Update. 1997-06-19 19:38 Ulrich Drepper <drepper@cygnus.com> * features.h: Define __STDC_IEC_559__ and _STDC_IEC_559_COMPLEX__. * elf/dl-minimal.c (__dcgettext): Remove assertion. * inet/rcmd.c: Correct a few typos. Reported by Eric Troan. * manual/Makefile (distribute): Add dir. * manual/dir: New file. * math/Makefile (libm-support): Rename s_rinttol, s_rinttoll, s_roundtol, and s_roundtoll to s_lrint, s_llrint, s_lround, and s_llround respectively. (libm-calls): Add e_gamma_r. * math/libm-test.c (check_int_exc): New function. (signbit_test): Remove test for sign of NaN. (gamma_test): Clear exception after test of existence. Correct tests and and epsilons. (lgamma_test): Likewise. (ilogb_test): Correct all tests. (scalb_test): Rewrite. (rinttol_test): Rename to lrint_test and correct tests. (rinttoll_test): Likewise. (roundtol_test): Likewise. (roundtoll_test): Likewise. (main): Call lrint/lround functions instead of rinttol/roundtol. * math/math.h: Change prototypes for rinttol/roundtol. * math/mathcalls.h: Rearrange prototypes according to ISO C9X draft. * sysdeps/generic/mathbits.h: Define FP_ILOGB0 and FP_ILOGBNAN. * sysdeps/i386/mathbits.h: Likewise. * sysdeps/libm-i387/e_scalb.S: Handle special cases correctly. * sysdeps/libm-i387/e_scalbf.S: Likewise. * sysdeps/libm-i387/e_scalbl.S: Likewise. * sysdeps/libm-i387/s_asinh.S: Handle -inf correctly. * sysdeps/libm-i387/s_asinhf.S: Likewise. * sysdeps/libm-i387/s_asinhl.S: Likewise. * sysdeps/libm-i387/s_ilogb.S: Optimize. * sysdeps/libm-i387/s_ilogbf.S: Likewise. * sysdeps/libm-i387/s_ilogbl.S: Likewise. * sysdeps/libm-i387/s_rinttol.S: Rename to... * sysdeps/libm-i387/s_lrint.S: ...this. * sysdeps/libm-i387/s_rinttoll.S: Rename to... * sysdeps/libm-i387/s_llrint.S: ...this. * sysdeps/libm-i387/s_remquo.S: Correctly set sign of remainder. * sysdeps/libm-i387/s_remquof.S: Likewise. * sysdeps/libm-i387/s_remquol.S: Likewise. * sysdeps/libm-i387/e_gamma_r.c: New file. Implementation of gamma function according to ISO C. * sysdeps/libm-i387/e_gammaf_r.c: New file. * sysdeps/libm-i387/e_gammal_r.c: New file. * sysdeps/libm-i387/e_lgamma_r.c: Don't let optimize compile the generation of exceptions away. * sysdeps/libm-i387/e_lgammaf_r.c: Likewise. * sysdeps/libm-i387/k_standard.c: Correct return value for infinity points of gamma function when not SVID mode. * sysdeps/libm-i387/s_rinttoll.c: Renamed to... * sysdeps/libm-i387/s_llrint.c: ...this. * sysdeps/libm-i387/s_rinttol.c: Renamed to... * sysdeps/libm-i387/s_lrint.c: ...this. * sysdeps/libm-i387/s_roundtoll.c: Renamed to... * sysdeps/libm-i387/s_llround.c: ...this. * sysdeps/libm-i387/s_roundtol.c: Renamed to... * sysdeps/libm-i387/s_lround.c: ..this. * sysdeps/libm-i387/s_scalbn.c: Change second parameter according to ISO C. * sysdeps/libm-i387/s_scalbnf.c: Likewise. * sysdeps/libm-i387/s_scalbnl.c: Likewise. * sysdeps/libm-i387/w_gamma.c: Call __ieee754_gamma_r if library mode is not _SVID_. * sysdeps/libm-i387/w_gammaf.c: Likewise. * sysdeps/libm-i387/w_gammal.c: Likewise. * sysdeps/m68k/fpu/__math.h: Rename __rinttol to __lrint and rinttol to lrint. * sysdeps/m68k/fpu/s_rinttol.c: Renamed to... * sysdeps/m68k/fpu/s_lrint.c: ...this. * sysdeps/m68k/fpu/s_rinttoll.c: Renamed to... * sysdeps/m68k/fpu/s_llrint.c: ...this. * md5-crypt/Makefile: Link md5test program with md5.o. * stdio-common/temptest.c: Don't use __stdio_gen_tempname which is not exported by the libc.so. * stdio-common/vfscanf.c: Correct scanning of strings after last change. * sysdeps/unix/sysv/linux/i386/sysdep.S: Use .comm to define errno. 1997-06-19 07:37 H.J. Lu <hjl@gnu.ai.mit.edu> * time/tzfile.c (__tzfile_read): Store getc () return in int. 1997-06-13 Miles Bader <miles@gnu.ai.mit.edu> * argp/argp-parse.c (argp_version_parser): Include `(PROGRAM ERROR)' in the no-version error text to indicate that something's fucked. [!_] (N_): New macro. (argp_default_options, argp_version_options): Wrap doc strings in N_(). 1997-06-12 Miles Bader <miles@gnu.ai.mit.edu> * argp/argp-parse.c (parser_parse_opt): Detect and report unhandled options here. (parser_parse_arg): Handle ARGP_KEY_ARGS here. Adjust NEXT pointer back if we fail to parse anything. (parser_parse_next): Simplify arg code. Leave state NEXT frobbing to parser_parse_arg. 1997-06-11 Miles Bader <miles@gnu.ai.mit.edu> * argp/argp.h (ARGP_KEY_ARGS, ARGP_KEY_FINI): New macros. * argp/argp-parse.c (parser_finalize): Do another pass over the parsers with ARGP_KEY_FINI. 1997-06-18 Miles Bader <miles@gnu.ai.mit.edu> * string/Makefile (routines): Add argz-replace. 1997-06-16 00:16 Miles Bader <miles@gnu.ai.mit.edu> * manual/string.texi (Argz Functions): Document argz_replace. 1997-06-12 Miles Bader <miles@gnu.ai.mit.edu> * string/argz.h (__argz_replace, argz_replace): New declarations. * string/argz-replace.c: New file. 1997-06-11 Miles Bader <miles@gnu.ai.mit.edu> * manual/argp.texi (Argp Special Keys): Document ARGP_KEY_ARGS. 1997-06-16 23:08 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * Makeconfig (libc-map): Remove definition. * Makerules (libc-map): Define it here, using the full name. (load-map-file): Remove case for empty $(..). ($(common-objpfx)libc.so): Revert last change. 1997-06-17 22:18 Mark Kettenis <kettenis@phys.uva.nl> * login/programs/utmpd.c (main): Improve signal handling. * login/programs/request.c (do_setutent, do_updwtmp): Allow arbitrary length filenames. * login/programs/utmpd.h (setutent_request, updwtmp_request): Get rid of fixed length file field. * login/utmp_daemon.c (do_setutent, do_updwtmp): Allow arbitrary length filenames. * login/programs/request.c (do_pututline): Don't fail if connection->position is -1 on entry. 1997-06-15 16:32 Mark Kettenis <kettenis@phys.uva.nl> * login/utmp_file.c (updwtmp_file): Use the same method for appending an entry as in pututline_file. 1997-06-11 18:59 Mark Kettenis <kettenis@phys.uva.nl> * login/utmp_file.c (getutent_r_file): Use read lock instead of write lock. (getutline_r_file, internal_getut_r): Lock utmp file. (updwtmp_file): Use fcntl to lock file instead of flock. 1997-06-18 00:11 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * stdio-common/vfscanf.c (inchar, ungetc): Don't count EOF as character read in/put back. * stdio-common/tstscanf.c: Add test case for this. 1997-06-17 22:17 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * libc.map: Add more symbols. 1997-06-18 12:01 Ulrich Drepper <drepper@cygnus.com> * manual/Makefile (install): Make sure `dir' file exists if we use install-info. 1997-06-17 19:32 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * manual/Makefile (info): Depend on dir-add.info. 1997-06-17 17:19 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * elf/ldd.bash.in: Fix spacing in message. 1997-06-17 14:28 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * elf/do-rel.h (elf_dynamic_do_rel): Always use version information if available. 1997-06-17 11:34 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sunrpc/Makefile ($(rpcsvc:%.x=$(objpfx)rpcsvc/%.h)): Make command non-empty to force make to recheck modification time. ($(rpcsvc:%.x=$(objpfx)x%.c)): Likewise. 1997-06-17 00:26 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/stub/e_acoshl.c: Set errno to ENOSYS. * sysdeps/stub/e_acosl.c: Likewise. * sysdeps/stub/e_asinl.c: Likewise. * sysdeps/stub/e_atan2l.c: Likewise. * sysdeps/stub/e_expl.c: Likewise. * sysdeps/stub/e_fmodl.c: Likewise. * sysdeps/stub/e_j0l.c: Likewise. * sysdeps/stub/e_j1l.c: Likewise. * sysdeps/stub/e_jnl.c: Likewise. * sysdeps/stub/e_lgammal_r.c: Likewise. * sysdeps/stub/e_log10l.c: Likewise. * sysdeps/stub/e_logl.c: Likewise. * sysdeps/stub/e_powl.c: Likewise. * sysdeps/stub/e_rem_pio2l.c: Likewise. * sysdeps/stub/e_sqrtl.c: Likewise. * sysdeps/stub/k_cosl.c: Likewise. * sysdeps/stub/k_rem_pio2l.c: Likewise. * sysdeps/stub/k_sinl.c: Likewise. * sysdeps/stub/k_tanl.c: Likewise. * sysdeps/stub/s_atanl.c: Likewise. * sysdeps/stub/s_erfl.c: Likewise. * sysdeps/stub/s_exp2.c: Likewise. * sysdeps/stub/s_exp2f.c: Likewise. * sysdeps/stub/s_exp2l.c: Likewise. * sysdeps/stub/s_expm1l.c: Likewise. * sysdeps/stub/s_log1pl.c: Likewise. * sysdeps/stub/s_log2l.c: Likewise. 1997-06-18 11:46 Ulrich Drepper <drepper@cygnus.com> * sysdeps/libm-ieee754/w_gamma.c: If _LIB_VERSION is _SVID_ compute result as before last change. * sysdeps/libm-ieee754/w_gammaf.c: Likewise. * sysdeps/libm-ieee754/w_gammal.c: Likewise. 1997-06-16 23:37 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/libm-ieee754/s_remquo.c: Fix off-by-one when computing quotient. * sysdeps/libm-ieee754/s_remquof.c: Likewise. * sysdeps/libm-ieee754/s_remquol.c: Likewise. * sysdeps/m68k/fpu/s_remquo.c: Remove FIXME and special case for quotient. * sysdeps/libm-ieee754/w_gamma.c: Add missing call to exp function. Don't use global signgam. * sysdeps/libm-ieee754/w_gammaf.c: Likewise. * sysdeps/libm-ieee754/w_gammal.c: Likewise. * math/Makefile (libm-calls): Remove w_gamma_r. * sysdeps/libm-ieee754/w_gamma_r.c: Remove file. * sysdeps/libm-ieee754/w_gammaf_r.c: Remove file. * sysdeps/libm-ieee754/w_gammal_r.c: Remove file. * math/libm-test.c (atanh_test): Declare x only if needed. (signbit_test): Fix typo. (gamma_test): Check whether function is implemented. Add epsilons. (lgamma_test): Likewise. (fmod_test): Add epsilons. (exp2_test): Use right function for existence test. 1997-06-07 09:20 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/powerpc/Dist: Add fenv_const.c. * sysdeps/unix/sysv/linux/Dist: Add net/if_slip.h. * sysdeps/unix/sysv/linux/powerpc/Dist: Add init-first.h and syscall.h. * sysdeps/unix/sysv/linux/sparc/Dist: Add init-first.h. * string/Makefile (distribute): Add tst-svc.expect. * nis/Makefile (distribute): Add nis_intern.h and Banner. * elf/Makefile (distribute): Add dl-hash.h. * Rules (subdir_echo-distinfo): Add headers from $(distribute). * login/Makefile (others): Add utmpdump. * sysdeps/unix/sysv/linux/Makefile (sysdep_headers): Add net/if_slip.h. * manual/Makefile (dir-add.texi): Also look in indirectly included files. 1997-06-16 23:15 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/unix/sysv/linux/configure.in: Remove commands to generate stdio_lim.h. * sysdeps/unix/sysv/linux/mk-stdiolim.c: Remove. * sysdeps/unix/sysv/linux/Makefile: Add rules to generate stdio_lim.h here. (common-generated): Add generated files. (inhibit-stdio_lim): Define. * sysdeps/posix/Makefile [$(inhibit-stdio_lim)=yes]: Disable rules to generate stdio_lim.h. * sysdeps/unix/sysv/linux/Makefile: Suppress inclusion of dependecy files if no_deps is set. ($(objpfx)syscall-%.d): Add header file as target to dependency generation. 1997-06-14 19:19 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/unix/sysv/linux/arm/Dist: New file. 1997-06-14 17:59 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * extra-lib.mk (others): Depend on versioned shared library, not the unversioned one. * Makerules (build-shlib): Don't make the version link here. ($(common-objpfx)libc.so$(libc.so-version)): New rule for libc version link. 1997-06-16 03:07 Ulrich Drepper <drepper@cygnus.com> * sysdeps/libm-ieee754/k_standard.c: Undo change of Tue Aug 6 01:13:56 1996. * argp/argp-help.c (argp_args_usage): Supply correct argp to filter_doc. * argp/argp-help.c (hol_add_cluster): Initialize CL->depth. * argp/argp-help.c (_help): Supply STATE to argp_args_usage. * argp/argp.h (ARGP_KEY_HELP_ARGS_DOC): New macro. * argp/argp-fmtstream.c: Add casts to prevent warnings. * argp/argp.h (OPTION_NO_USAGE): New macro. * argp/argp-help.c (usage_long_opt, usage_argful_short_opt, * argp/argp-fmtstream.c (__argp_fmtstream_update): Account for case * argp/argp-help.c <stddef.h>: New include. * argp/argp.h (argp_state_help, __argp_state_help, argp_usage, * argp/argp.h (argp_program_bug_address): Make const. * argp/argp-parse.c (argp_default_parser): Set STATE->name for OPT_PROGNAME. * argp/argp-help.c (__argp_error, __argp_failure, __argp_state_help): * argp/argp-parse.c (parser_init): Set PARSER->state.flags.
1997-06-19 19:11:22 +00:00
#include <math.h>
#include <math_private.h>
#include <float.h>
Update. 1997-06-19 19:38 Ulrich Drepper <drepper@cygnus.com> * features.h: Define __STDC_IEC_559__ and _STDC_IEC_559_COMPLEX__. * elf/dl-minimal.c (__dcgettext): Remove assertion. * inet/rcmd.c: Correct a few typos. Reported by Eric Troan. * manual/Makefile (distribute): Add dir. * manual/dir: New file. * math/Makefile (libm-support): Rename s_rinttol, s_rinttoll, s_roundtol, and s_roundtoll to s_lrint, s_llrint, s_lround, and s_llround respectively. (libm-calls): Add e_gamma_r. * math/libm-test.c (check_int_exc): New function. (signbit_test): Remove test for sign of NaN. (gamma_test): Clear exception after test of existence. Correct tests and and epsilons. (lgamma_test): Likewise. (ilogb_test): Correct all tests. (scalb_test): Rewrite. (rinttol_test): Rename to lrint_test and correct tests. (rinttoll_test): Likewise. (roundtol_test): Likewise. (roundtoll_test): Likewise. (main): Call lrint/lround functions instead of rinttol/roundtol. * math/math.h: Change prototypes for rinttol/roundtol. * math/mathcalls.h: Rearrange prototypes according to ISO C9X draft. * sysdeps/generic/mathbits.h: Define FP_ILOGB0 and FP_ILOGBNAN. * sysdeps/i386/mathbits.h: Likewise. * sysdeps/libm-i387/e_scalb.S: Handle special cases correctly. * sysdeps/libm-i387/e_scalbf.S: Likewise. * sysdeps/libm-i387/e_scalbl.S: Likewise. * sysdeps/libm-i387/s_asinh.S: Handle -inf correctly. * sysdeps/libm-i387/s_asinhf.S: Likewise. * sysdeps/libm-i387/s_asinhl.S: Likewise. * sysdeps/libm-i387/s_ilogb.S: Optimize. * sysdeps/libm-i387/s_ilogbf.S: Likewise. * sysdeps/libm-i387/s_ilogbl.S: Likewise. * sysdeps/libm-i387/s_rinttol.S: Rename to... * sysdeps/libm-i387/s_lrint.S: ...this. * sysdeps/libm-i387/s_rinttoll.S: Rename to... * sysdeps/libm-i387/s_llrint.S: ...this. * sysdeps/libm-i387/s_remquo.S: Correctly set sign of remainder. * sysdeps/libm-i387/s_remquof.S: Likewise. * sysdeps/libm-i387/s_remquol.S: Likewise. * sysdeps/libm-i387/e_gamma_r.c: New file. Implementation of gamma function according to ISO C. * sysdeps/libm-i387/e_gammaf_r.c: New file. * sysdeps/libm-i387/e_gammal_r.c: New file. * sysdeps/libm-i387/e_lgamma_r.c: Don't let optimize compile the generation of exceptions away. * sysdeps/libm-i387/e_lgammaf_r.c: Likewise. * sysdeps/libm-i387/k_standard.c: Correct return value for infinity points of gamma function when not SVID mode. * sysdeps/libm-i387/s_rinttoll.c: Renamed to... * sysdeps/libm-i387/s_llrint.c: ...this. * sysdeps/libm-i387/s_rinttol.c: Renamed to... * sysdeps/libm-i387/s_lrint.c: ...this. * sysdeps/libm-i387/s_roundtoll.c: Renamed to... * sysdeps/libm-i387/s_llround.c: ...this. * sysdeps/libm-i387/s_roundtol.c: Renamed to... * sysdeps/libm-i387/s_lround.c: ..this. * sysdeps/libm-i387/s_scalbn.c: Change second parameter according to ISO C. * sysdeps/libm-i387/s_scalbnf.c: Likewise. * sysdeps/libm-i387/s_scalbnl.c: Likewise. * sysdeps/libm-i387/w_gamma.c: Call __ieee754_gamma_r if library mode is not _SVID_. * sysdeps/libm-i387/w_gammaf.c: Likewise. * sysdeps/libm-i387/w_gammal.c: Likewise. * sysdeps/m68k/fpu/__math.h: Rename __rinttol to __lrint and rinttol to lrint. * sysdeps/m68k/fpu/s_rinttol.c: Renamed to... * sysdeps/m68k/fpu/s_lrint.c: ...this. * sysdeps/m68k/fpu/s_rinttoll.c: Renamed to... * sysdeps/m68k/fpu/s_llrint.c: ...this. * md5-crypt/Makefile: Link md5test program with md5.o. * stdio-common/temptest.c: Don't use __stdio_gen_tempname which is not exported by the libc.so. * stdio-common/vfscanf.c: Correct scanning of strings after last change. * sysdeps/unix/sysv/linux/i386/sysdep.S: Use .comm to define errno. 1997-06-19 07:37 H.J. Lu <hjl@gnu.ai.mit.edu> * time/tzfile.c (__tzfile_read): Store getc () return in int. 1997-06-13 Miles Bader <miles@gnu.ai.mit.edu> * argp/argp-parse.c (argp_version_parser): Include `(PROGRAM ERROR)' in the no-version error text to indicate that something's fucked. [!_] (N_): New macro. (argp_default_options, argp_version_options): Wrap doc strings in N_(). 1997-06-12 Miles Bader <miles@gnu.ai.mit.edu> * argp/argp-parse.c (parser_parse_opt): Detect and report unhandled options here. (parser_parse_arg): Handle ARGP_KEY_ARGS here. Adjust NEXT pointer back if we fail to parse anything. (parser_parse_next): Simplify arg code. Leave state NEXT frobbing to parser_parse_arg. 1997-06-11 Miles Bader <miles@gnu.ai.mit.edu> * argp/argp.h (ARGP_KEY_ARGS, ARGP_KEY_FINI): New macros. * argp/argp-parse.c (parser_finalize): Do another pass over the parsers with ARGP_KEY_FINI. 1997-06-18 Miles Bader <miles@gnu.ai.mit.edu> * string/Makefile (routines): Add argz-replace. 1997-06-16 00:16 Miles Bader <miles@gnu.ai.mit.edu> * manual/string.texi (Argz Functions): Document argz_replace. 1997-06-12 Miles Bader <miles@gnu.ai.mit.edu> * string/argz.h (__argz_replace, argz_replace): New declarations. * string/argz-replace.c: New file. 1997-06-11 Miles Bader <miles@gnu.ai.mit.edu> * manual/argp.texi (Argp Special Keys): Document ARGP_KEY_ARGS. 1997-06-16 23:08 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * Makeconfig (libc-map): Remove definition. * Makerules (libc-map): Define it here, using the full name. (load-map-file): Remove case for empty $(..). ($(common-objpfx)libc.so): Revert last change. 1997-06-17 22:18 Mark Kettenis <kettenis@phys.uva.nl> * login/programs/utmpd.c (main): Improve signal handling. * login/programs/request.c (do_setutent, do_updwtmp): Allow arbitrary length filenames. * login/programs/utmpd.h (setutent_request, updwtmp_request): Get rid of fixed length file field. * login/utmp_daemon.c (do_setutent, do_updwtmp): Allow arbitrary length filenames. * login/programs/request.c (do_pututline): Don't fail if connection->position is -1 on entry. 1997-06-15 16:32 Mark Kettenis <kettenis@phys.uva.nl> * login/utmp_file.c (updwtmp_file): Use the same method for appending an entry as in pututline_file. 1997-06-11 18:59 Mark Kettenis <kettenis@phys.uva.nl> * login/utmp_file.c (getutent_r_file): Use read lock instead of write lock. (getutline_r_file, internal_getut_r): Lock utmp file. (updwtmp_file): Use fcntl to lock file instead of flock. 1997-06-18 00:11 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * stdio-common/vfscanf.c (inchar, ungetc): Don't count EOF as character read in/put back. * stdio-common/tstscanf.c: Add test case for this. 1997-06-17 22:17 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * libc.map: Add more symbols. 1997-06-18 12:01 Ulrich Drepper <drepper@cygnus.com> * manual/Makefile (install): Make sure `dir' file exists if we use install-info. 1997-06-17 19:32 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * manual/Makefile (info): Depend on dir-add.info. 1997-06-17 17:19 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * elf/ldd.bash.in: Fix spacing in message. 1997-06-17 14:28 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * elf/do-rel.h (elf_dynamic_do_rel): Always use version information if available. 1997-06-17 11:34 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sunrpc/Makefile ($(rpcsvc:%.x=$(objpfx)rpcsvc/%.h)): Make command non-empty to force make to recheck modification time. ($(rpcsvc:%.x=$(objpfx)x%.c)): Likewise. 1997-06-17 00:26 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/stub/e_acoshl.c: Set errno to ENOSYS. * sysdeps/stub/e_acosl.c: Likewise. * sysdeps/stub/e_asinl.c: Likewise. * sysdeps/stub/e_atan2l.c: Likewise. * sysdeps/stub/e_expl.c: Likewise. * sysdeps/stub/e_fmodl.c: Likewise. * sysdeps/stub/e_j0l.c: Likewise. * sysdeps/stub/e_j1l.c: Likewise. * sysdeps/stub/e_jnl.c: Likewise. * sysdeps/stub/e_lgammal_r.c: Likewise. * sysdeps/stub/e_log10l.c: Likewise. * sysdeps/stub/e_logl.c: Likewise. * sysdeps/stub/e_powl.c: Likewise. * sysdeps/stub/e_rem_pio2l.c: Likewise. * sysdeps/stub/e_sqrtl.c: Likewise. * sysdeps/stub/k_cosl.c: Likewise. * sysdeps/stub/k_rem_pio2l.c: Likewise. * sysdeps/stub/k_sinl.c: Likewise. * sysdeps/stub/k_tanl.c: Likewise. * sysdeps/stub/s_atanl.c: Likewise. * sysdeps/stub/s_erfl.c: Likewise. * sysdeps/stub/s_exp2.c: Likewise. * sysdeps/stub/s_exp2f.c: Likewise. * sysdeps/stub/s_exp2l.c: Likewise. * sysdeps/stub/s_expm1l.c: Likewise. * sysdeps/stub/s_log1pl.c: Likewise. * sysdeps/stub/s_log2l.c: Likewise. 1997-06-18 11:46 Ulrich Drepper <drepper@cygnus.com> * sysdeps/libm-ieee754/w_gamma.c: If _LIB_VERSION is _SVID_ compute result as before last change. * sysdeps/libm-ieee754/w_gammaf.c: Likewise. * sysdeps/libm-ieee754/w_gammal.c: Likewise. 1997-06-16 23:37 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/libm-ieee754/s_remquo.c: Fix off-by-one when computing quotient. * sysdeps/libm-ieee754/s_remquof.c: Likewise. * sysdeps/libm-ieee754/s_remquol.c: Likewise. * sysdeps/m68k/fpu/s_remquo.c: Remove FIXME and special case for quotient. * sysdeps/libm-ieee754/w_gamma.c: Add missing call to exp function. Don't use global signgam. * sysdeps/libm-ieee754/w_gammaf.c: Likewise. * sysdeps/libm-ieee754/w_gammal.c: Likewise. * math/Makefile (libm-calls): Remove w_gamma_r. * sysdeps/libm-ieee754/w_gamma_r.c: Remove file. * sysdeps/libm-ieee754/w_gammaf_r.c: Remove file. * sysdeps/libm-ieee754/w_gammal_r.c: Remove file. * math/libm-test.c (atanh_test): Declare x only if needed. (signbit_test): Fix typo. (gamma_test): Check whether function is implemented. Add epsilons. (lgamma_test): Likewise. (fmod_test): Add epsilons. (exp2_test): Use right function for existence test. 1997-06-07 09:20 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/powerpc/Dist: Add fenv_const.c. * sysdeps/unix/sysv/linux/Dist: Add net/if_slip.h. * sysdeps/unix/sysv/linux/powerpc/Dist: Add init-first.h and syscall.h. * sysdeps/unix/sysv/linux/sparc/Dist: Add init-first.h. * string/Makefile (distribute): Add tst-svc.expect. * nis/Makefile (distribute): Add nis_intern.h and Banner. * elf/Makefile (distribute): Add dl-hash.h. * Rules (subdir_echo-distinfo): Add headers from $(distribute). * login/Makefile (others): Add utmpdump. * sysdeps/unix/sysv/linux/Makefile (sysdep_headers): Add net/if_slip.h. * manual/Makefile (dir-add.texi): Also look in indirectly included files. 1997-06-16 23:15 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/unix/sysv/linux/configure.in: Remove commands to generate stdio_lim.h. * sysdeps/unix/sysv/linux/mk-stdiolim.c: Remove. * sysdeps/unix/sysv/linux/Makefile: Add rules to generate stdio_lim.h here. (common-generated): Add generated files. (inhibit-stdio_lim): Define. * sysdeps/posix/Makefile [$(inhibit-stdio_lim)=yes]: Disable rules to generate stdio_lim.h. * sysdeps/unix/sysv/linux/Makefile: Suppress inclusion of dependecy files if no_deps is set. ($(objpfx)syscall-%.d): Add header file as target to dependency generation. 1997-06-14 19:19 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/unix/sysv/linux/arm/Dist: New file. 1997-06-14 17:59 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * extra-lib.mk (others): Depend on versioned shared library, not the unversioned one. * Makerules (build-shlib): Don't make the version link here. ($(common-objpfx)libc.so$(libc.so-version)): New rule for libc version link. 1997-06-16 03:07 Ulrich Drepper <drepper@cygnus.com> * sysdeps/libm-ieee754/k_standard.c: Undo change of Tue Aug 6 01:13:56 1996. * argp/argp-help.c (argp_args_usage): Supply correct argp to filter_doc. * argp/argp-help.c (hol_add_cluster): Initialize CL->depth. * argp/argp-help.c (_help): Supply STATE to argp_args_usage. * argp/argp.h (ARGP_KEY_HELP_ARGS_DOC): New macro. * argp/argp-fmtstream.c: Add casts to prevent warnings. * argp/argp.h (OPTION_NO_USAGE): New macro. * argp/argp-help.c (usage_long_opt, usage_argful_short_opt, * argp/argp-fmtstream.c (__argp_fmtstream_update): Account for case * argp/argp-help.c <stddef.h>: New include. * argp/argp.h (argp_state_help, __argp_state_help, argp_usage, * argp/argp.h (argp_program_bug_address): Make const. * argp/argp-parse.c (argp_default_parser): Set STATE->name for OPT_PROGNAME. * argp/argp-help.c (__argp_error, __argp_failure, __argp_state_help): * argp/argp-parse.c (parser_init): Set PARSER->state.flags.
1997-06-19 19:11:22 +00:00
/* Coefficients B_2k / 2k(2k-1) of x^-(2k-1) inside exp in Stirling's
approximation to gamma function. */
static const double gamma_coeff[] =
{
0x1.5555555555555p-4,
-0xb.60b60b60b60b8p-12,
0x3.4034034034034p-12,
-0x2.7027027027028p-12,
0x3.72a3c5631fe46p-12,
-0x7.daac36664f1f4p-12,
};
#define NCOEFF (sizeof (gamma_coeff) / sizeof (gamma_coeff[0]))
/* Return gamma (X), for positive X less than 184, in the form R *
2^(*EXP2_ADJ), where R is the return value and *EXP2_ADJ is set to
avoid overflow or underflow in intermediate calculations. */
static double
gamma_positive (double x, int *exp2_adj)
{
int local_signgam;
if (x < 0.5)
{
*exp2_adj = 0;
return __ieee754_exp (__ieee754_lgamma_r (x + 1, &local_signgam)) / x;
}
else if (x <= 1.5)
{
*exp2_adj = 0;
return __ieee754_exp (__ieee754_lgamma_r (x, &local_signgam));
}
else if (x < 6.5)
{
/* Adjust into the range for using exp (lgamma). */
*exp2_adj = 0;
double n = __ceil (x - 1.5);
double x_adj = x - n;
double eps;
double prod = __gamma_product (x_adj, 0, n, &eps);
return (__ieee754_exp (__ieee754_lgamma_r (x_adj, &local_signgam))
* prod * (1.0 + eps));
}
else
{
double eps = 0;
double x_eps = 0;
double x_adj = x;
double prod = 1;
if (x < 12.0)
{
/* Adjust into the range for applying Stirling's
approximation. */
double n = __ceil (12.0 - x);
Use math_narrow_eval more consistently. Where glibc code needs to avoid excess range and precision in floating-point arithmetic, code variously uses either asms or volatile to force the results of that arithmetic to memory; mostly this is conditional on FLT_EVAL_METHOD, but in the case of lrint / llrint functions some use of volatile is unconditional (and is present unnecessarily in versions for long double). This patch make such code use the recently-added math_narrow_eval macro consistently, removing the unnecessary uses of volatile in long double lrint / llrint implementations completely. Tested for x86_64, x86, mips64 and powerpc. * math/s_nexttowardf.c (__nexttowardf): Use math_narrow_eval. * stdlib/strtod_l.c: Include <math_private.h>. (overflow_value): Use math_narrow_eval. (underflow_value): Likewise. * sysdeps/i386/fpu/s_nexttoward.c (__nexttoward): Likewise. * sysdeps/i386/fpu/s_nexttowardf.c (__nexttowardf): Likewise. * sysdeps/ieee754/dbl-64/e_gamma_r.c (gamma_positive): Likewise. (__ieee754_gamma_r): Likewise. * sysdeps/ieee754/dbl-64/gamma_productf.c (__gamma_productf): Likewise. * sysdeps/ieee754/dbl-64/k_rem_pio2.c (__kernel_rem_pio2): Likewise. * sysdeps/ieee754/dbl-64/lgamma_neg.c (__lgamma_neg): Likewise. * sysdeps/ieee754/dbl-64/s_erf.c (__erfc): Likewise. * sysdeps/ieee754/dbl-64/s_llrint.c (__llrint): Likewise. * sysdeps/ieee754/dbl-64/s_lrint.c (__lrint): Likewise. * sysdeps/ieee754/flt-32/e_gammaf_r.c (gammaf_positive): Likewise. (__ieee754_gammaf_r): Likewise. * sysdeps/ieee754/flt-32/k_rem_pio2f.c (__kernel_rem_pio2f): Likewise. * sysdeps/ieee754/flt-32/lgamma_negf.c (__lgamma_negf): Likewise. * sysdeps/ieee754/flt-32/s_erff.c (__erfcf): Likewise. * sysdeps/ieee754/flt-32/s_llrintf.c (__llrintf): Likewise. * sysdeps/ieee754/flt-32/s_lrintf.c (__lrintf): Likewise. * sysdeps/ieee754/ldbl-128/s_llrintl.c (__llrintl): Do not use volatile. * sysdeps/ieee754/ldbl-128/s_lrintl.c (__lrintl): Likewise. * sysdeps/ieee754/ldbl-128/s_nexttoward.c (__nexttoward): Use math_narrow_eval. * sysdeps/ieee754/ldbl-128ibm/s_nexttoward.c (__nexttoward): Likewise. * sysdeps/ieee754/ldbl-128ibm/s_nexttowardf.c (__nexttowardf): Likewise. * sysdeps/ieee754/ldbl-96/gamma_product.c (__gamma_product): Likewise. * sysdeps/ieee754/ldbl-96/s_llrintl.c (__llrintl): Do not use volatile. * sysdeps/ieee754/ldbl-96/s_lrintl.c (__lrintl): Likewise. * sysdeps/ieee754/ldbl-96/s_nexttoward.c (__nexttoward): Use math_narrow_eval. * sysdeps/ieee754/ldbl-96/s_nexttowardf.c (__nexttowardf): Likewise. * sysdeps/ieee754/ldbl-opt/s_nexttowardfd.c (__nldbl_nexttowardf): Likewise.
2015-09-23 18:14:57 +00:00
x_adj = math_narrow_eval (x + n);
x_eps = (x - (x_adj - n));
prod = __gamma_product (x_adj - n, x_eps, n, &eps);
}
/* The result is now gamma (X_ADJ + X_EPS) / (PROD * (1 + EPS)).
Compute gamma (X_ADJ + X_EPS) using Stirling's approximation,
starting by computing pow (X_ADJ, X_ADJ) with a power of 2
factored out. */
double exp_adj = -eps;
double x_adj_int = __round (x_adj);
double x_adj_frac = x_adj - x_adj_int;
int x_adj_log2;
double x_adj_mant = __frexp (x_adj, &x_adj_log2);
if (x_adj_mant < M_SQRT1_2)
{
x_adj_log2--;
x_adj_mant *= 2.0;
}
*exp2_adj = x_adj_log2 * (int) x_adj_int;
double ret = (__ieee754_pow (x_adj_mant, x_adj)
* __ieee754_exp2 (x_adj_log2 * x_adj_frac)
* __ieee754_exp (-x_adj)
* __ieee754_sqrt (2 * M_PI / x_adj)
/ prod);
Improve tgamma accuracy (bug 18613). In non-default rounding modes, tgamma can be slightly less accurate than permitted by glibc's accuracy goals. Part of the problem is error accumulation, addressed in this patch by setting round-to-nearest for internal computations. However, there was also a bug in the code dealing with computing pow (x + n, x + n) where x + n is not exactly representable, providing another source of error even in round-to-nearest mode; it was necessary to address both bugs to get errors for all testcases within glibc's accuracy goals. Given this second fix, accuracy in round-to-nearest mode is also improved (hence regeneration of ulps for tgamma should be from scratch - truncate libm-test-ulps or at least remove existing tgamma entries - so that the expected ulps can be reduced). Some additional complications also arose. Certain tgamma tests should strictly, according to IEEE semantics, overflow or not depending on the rounding mode; this is beyond the scope of glibc's accuracy goals for any function without exactly-determined results, but gen-auto-libm-tests doesn't handle being lax there as it does for underflow. (libm-test.inc also doesn't handle being lax about whether the result in cases very close to the overflow threshold is infinity or a finite value close to overflow, but that doesn't cause problems in this case though I've seen it cause problems with random test generation for some functions.) Thus, spurious-overflow markings, with a comment, are added to auto-libm-test-in (no bug in Bugzilla because the issue is with the testsuite, not a user-visible bug in glibc). And on x86, after the patch I saw ERANGE issues as previously reported by Carlos (see my commentary in <https://sourceware.org/ml/libc-alpha/2015-01/msg00485.html>), which needed addressing by ensuring excess range and precision were eliminated at various points if FLT_EVAL_METHOD != 0. I also noticed and fixed a cosmetic issue where 1.0f was used in long double functions and should have been 1.0L. This completes the move of all functions to testing in all rounding modes with ALL_RM_TEST, so gen-libm-have-vector-test.sh is updated to remove the workaround for some functions not using ALL_RM_TEST. Tested for x86_64, x86, mips64 and powerpc. [BZ #18613] * sysdeps/ieee754/dbl-64/e_gamma_r.c (gamma_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gamma_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. * sysdeps/ieee754/flt-32/e_gammaf_r.c (gammaf_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammaf_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. * sysdeps/ieee754/ldbl-128/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * sysdeps/ieee754/ldbl-128ibm/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * sysdeps/ieee754/ldbl-96/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * math/libm-test.inc (tgamma_test_data): Remove one test. Moved to auto-libm-test-in. (tgamma_test): Use ALL_RM_TEST. * math/auto-libm-test-in: Add one test of tgamma. Mark some other tests of tgamma with spurious-overflow. * math/auto-libm-test-out: Regenerated. * math/gen-libm-have-vector-test.sh: Do not check for START. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-06-29 23:29:35 +00:00
exp_adj += x_eps * __ieee754_log (x_adj);
double bsum = gamma_coeff[NCOEFF - 1];
double x_adj2 = x_adj * x_adj;
for (size_t i = 1; i <= NCOEFF - 1; i++)
bsum = bsum / x_adj2 + gamma_coeff[NCOEFF - 1 - i];
exp_adj += bsum / x_adj;
return ret + ret * __expm1 (exp_adj);
}
}
Update. 1997-06-19 19:38 Ulrich Drepper <drepper@cygnus.com> * features.h: Define __STDC_IEC_559__ and _STDC_IEC_559_COMPLEX__. * elf/dl-minimal.c (__dcgettext): Remove assertion. * inet/rcmd.c: Correct a few typos. Reported by Eric Troan. * manual/Makefile (distribute): Add dir. * manual/dir: New file. * math/Makefile (libm-support): Rename s_rinttol, s_rinttoll, s_roundtol, and s_roundtoll to s_lrint, s_llrint, s_lround, and s_llround respectively. (libm-calls): Add e_gamma_r. * math/libm-test.c (check_int_exc): New function. (signbit_test): Remove test for sign of NaN. (gamma_test): Clear exception after test of existence. Correct tests and and epsilons. (lgamma_test): Likewise. (ilogb_test): Correct all tests. (scalb_test): Rewrite. (rinttol_test): Rename to lrint_test and correct tests. (rinttoll_test): Likewise. (roundtol_test): Likewise. (roundtoll_test): Likewise. (main): Call lrint/lround functions instead of rinttol/roundtol. * math/math.h: Change prototypes for rinttol/roundtol. * math/mathcalls.h: Rearrange prototypes according to ISO C9X draft. * sysdeps/generic/mathbits.h: Define FP_ILOGB0 and FP_ILOGBNAN. * sysdeps/i386/mathbits.h: Likewise. * sysdeps/libm-i387/e_scalb.S: Handle special cases correctly. * sysdeps/libm-i387/e_scalbf.S: Likewise. * sysdeps/libm-i387/e_scalbl.S: Likewise. * sysdeps/libm-i387/s_asinh.S: Handle -inf correctly. * sysdeps/libm-i387/s_asinhf.S: Likewise. * sysdeps/libm-i387/s_asinhl.S: Likewise. * sysdeps/libm-i387/s_ilogb.S: Optimize. * sysdeps/libm-i387/s_ilogbf.S: Likewise. * sysdeps/libm-i387/s_ilogbl.S: Likewise. * sysdeps/libm-i387/s_rinttol.S: Rename to... * sysdeps/libm-i387/s_lrint.S: ...this. * sysdeps/libm-i387/s_rinttoll.S: Rename to... * sysdeps/libm-i387/s_llrint.S: ...this. * sysdeps/libm-i387/s_remquo.S: Correctly set sign of remainder. * sysdeps/libm-i387/s_remquof.S: Likewise. * sysdeps/libm-i387/s_remquol.S: Likewise. * sysdeps/libm-i387/e_gamma_r.c: New file. Implementation of gamma function according to ISO C. * sysdeps/libm-i387/e_gammaf_r.c: New file. * sysdeps/libm-i387/e_gammal_r.c: New file. * sysdeps/libm-i387/e_lgamma_r.c: Don't let optimize compile the generation of exceptions away. * sysdeps/libm-i387/e_lgammaf_r.c: Likewise. * sysdeps/libm-i387/k_standard.c: Correct return value for infinity points of gamma function when not SVID mode. * sysdeps/libm-i387/s_rinttoll.c: Renamed to... * sysdeps/libm-i387/s_llrint.c: ...this. * sysdeps/libm-i387/s_rinttol.c: Renamed to... * sysdeps/libm-i387/s_lrint.c: ...this. * sysdeps/libm-i387/s_roundtoll.c: Renamed to... * sysdeps/libm-i387/s_llround.c: ...this. * sysdeps/libm-i387/s_roundtol.c: Renamed to... * sysdeps/libm-i387/s_lround.c: ..this. * sysdeps/libm-i387/s_scalbn.c: Change second parameter according to ISO C. * sysdeps/libm-i387/s_scalbnf.c: Likewise. * sysdeps/libm-i387/s_scalbnl.c: Likewise. * sysdeps/libm-i387/w_gamma.c: Call __ieee754_gamma_r if library mode is not _SVID_. * sysdeps/libm-i387/w_gammaf.c: Likewise. * sysdeps/libm-i387/w_gammal.c: Likewise. * sysdeps/m68k/fpu/__math.h: Rename __rinttol to __lrint and rinttol to lrint. * sysdeps/m68k/fpu/s_rinttol.c: Renamed to... * sysdeps/m68k/fpu/s_lrint.c: ...this. * sysdeps/m68k/fpu/s_rinttoll.c: Renamed to... * sysdeps/m68k/fpu/s_llrint.c: ...this. * md5-crypt/Makefile: Link md5test program with md5.o. * stdio-common/temptest.c: Don't use __stdio_gen_tempname which is not exported by the libc.so. * stdio-common/vfscanf.c: Correct scanning of strings after last change. * sysdeps/unix/sysv/linux/i386/sysdep.S: Use .comm to define errno. 1997-06-19 07:37 H.J. Lu <hjl@gnu.ai.mit.edu> * time/tzfile.c (__tzfile_read): Store getc () return in int. 1997-06-13 Miles Bader <miles@gnu.ai.mit.edu> * argp/argp-parse.c (argp_version_parser): Include `(PROGRAM ERROR)' in the no-version error text to indicate that something's fucked. [!_] (N_): New macro. (argp_default_options, argp_version_options): Wrap doc strings in N_(). 1997-06-12 Miles Bader <miles@gnu.ai.mit.edu> * argp/argp-parse.c (parser_parse_opt): Detect and report unhandled options here. (parser_parse_arg): Handle ARGP_KEY_ARGS here. Adjust NEXT pointer back if we fail to parse anything. (parser_parse_next): Simplify arg code. Leave state NEXT frobbing to parser_parse_arg. 1997-06-11 Miles Bader <miles@gnu.ai.mit.edu> * argp/argp.h (ARGP_KEY_ARGS, ARGP_KEY_FINI): New macros. * argp/argp-parse.c (parser_finalize): Do another pass over the parsers with ARGP_KEY_FINI. 1997-06-18 Miles Bader <miles@gnu.ai.mit.edu> * string/Makefile (routines): Add argz-replace. 1997-06-16 00:16 Miles Bader <miles@gnu.ai.mit.edu> * manual/string.texi (Argz Functions): Document argz_replace. 1997-06-12 Miles Bader <miles@gnu.ai.mit.edu> * string/argz.h (__argz_replace, argz_replace): New declarations. * string/argz-replace.c: New file. 1997-06-11 Miles Bader <miles@gnu.ai.mit.edu> * manual/argp.texi (Argp Special Keys): Document ARGP_KEY_ARGS. 1997-06-16 23:08 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * Makeconfig (libc-map): Remove definition. * Makerules (libc-map): Define it here, using the full name. (load-map-file): Remove case for empty $(..). ($(common-objpfx)libc.so): Revert last change. 1997-06-17 22:18 Mark Kettenis <kettenis@phys.uva.nl> * login/programs/utmpd.c (main): Improve signal handling. * login/programs/request.c (do_setutent, do_updwtmp): Allow arbitrary length filenames. * login/programs/utmpd.h (setutent_request, updwtmp_request): Get rid of fixed length file field. * login/utmp_daemon.c (do_setutent, do_updwtmp): Allow arbitrary length filenames. * login/programs/request.c (do_pututline): Don't fail if connection->position is -1 on entry. 1997-06-15 16:32 Mark Kettenis <kettenis@phys.uva.nl> * login/utmp_file.c (updwtmp_file): Use the same method for appending an entry as in pututline_file. 1997-06-11 18:59 Mark Kettenis <kettenis@phys.uva.nl> * login/utmp_file.c (getutent_r_file): Use read lock instead of write lock. (getutline_r_file, internal_getut_r): Lock utmp file. (updwtmp_file): Use fcntl to lock file instead of flock. 1997-06-18 00:11 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * stdio-common/vfscanf.c (inchar, ungetc): Don't count EOF as character read in/put back. * stdio-common/tstscanf.c: Add test case for this. 1997-06-17 22:17 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * libc.map: Add more symbols. 1997-06-18 12:01 Ulrich Drepper <drepper@cygnus.com> * manual/Makefile (install): Make sure `dir' file exists if we use install-info. 1997-06-17 19:32 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * manual/Makefile (info): Depend on dir-add.info. 1997-06-17 17:19 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * elf/ldd.bash.in: Fix spacing in message. 1997-06-17 14:28 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * elf/do-rel.h (elf_dynamic_do_rel): Always use version information if available. 1997-06-17 11:34 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sunrpc/Makefile ($(rpcsvc:%.x=$(objpfx)rpcsvc/%.h)): Make command non-empty to force make to recheck modification time. ($(rpcsvc:%.x=$(objpfx)x%.c)): Likewise. 1997-06-17 00:26 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/stub/e_acoshl.c: Set errno to ENOSYS. * sysdeps/stub/e_acosl.c: Likewise. * sysdeps/stub/e_asinl.c: Likewise. * sysdeps/stub/e_atan2l.c: Likewise. * sysdeps/stub/e_expl.c: Likewise. * sysdeps/stub/e_fmodl.c: Likewise. * sysdeps/stub/e_j0l.c: Likewise. * sysdeps/stub/e_j1l.c: Likewise. * sysdeps/stub/e_jnl.c: Likewise. * sysdeps/stub/e_lgammal_r.c: Likewise. * sysdeps/stub/e_log10l.c: Likewise. * sysdeps/stub/e_logl.c: Likewise. * sysdeps/stub/e_powl.c: Likewise. * sysdeps/stub/e_rem_pio2l.c: Likewise. * sysdeps/stub/e_sqrtl.c: Likewise. * sysdeps/stub/k_cosl.c: Likewise. * sysdeps/stub/k_rem_pio2l.c: Likewise. * sysdeps/stub/k_sinl.c: Likewise. * sysdeps/stub/k_tanl.c: Likewise. * sysdeps/stub/s_atanl.c: Likewise. * sysdeps/stub/s_erfl.c: Likewise. * sysdeps/stub/s_exp2.c: Likewise. * sysdeps/stub/s_exp2f.c: Likewise. * sysdeps/stub/s_exp2l.c: Likewise. * sysdeps/stub/s_expm1l.c: Likewise. * sysdeps/stub/s_log1pl.c: Likewise. * sysdeps/stub/s_log2l.c: Likewise. 1997-06-18 11:46 Ulrich Drepper <drepper@cygnus.com> * sysdeps/libm-ieee754/w_gamma.c: If _LIB_VERSION is _SVID_ compute result as before last change. * sysdeps/libm-ieee754/w_gammaf.c: Likewise. * sysdeps/libm-ieee754/w_gammal.c: Likewise. 1997-06-16 23:37 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/libm-ieee754/s_remquo.c: Fix off-by-one when computing quotient. * sysdeps/libm-ieee754/s_remquof.c: Likewise. * sysdeps/libm-ieee754/s_remquol.c: Likewise. * sysdeps/m68k/fpu/s_remquo.c: Remove FIXME and special case for quotient. * sysdeps/libm-ieee754/w_gamma.c: Add missing call to exp function. Don't use global signgam. * sysdeps/libm-ieee754/w_gammaf.c: Likewise. * sysdeps/libm-ieee754/w_gammal.c: Likewise. * math/Makefile (libm-calls): Remove w_gamma_r. * sysdeps/libm-ieee754/w_gamma_r.c: Remove file. * sysdeps/libm-ieee754/w_gammaf_r.c: Remove file. * sysdeps/libm-ieee754/w_gammal_r.c: Remove file. * math/libm-test.c (atanh_test): Declare x only if needed. (signbit_test): Fix typo. (gamma_test): Check whether function is implemented. Add epsilons. (lgamma_test): Likewise. (fmod_test): Add epsilons. (exp2_test): Use right function for existence test. 1997-06-07 09:20 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/powerpc/Dist: Add fenv_const.c. * sysdeps/unix/sysv/linux/Dist: Add net/if_slip.h. * sysdeps/unix/sysv/linux/powerpc/Dist: Add init-first.h and syscall.h. * sysdeps/unix/sysv/linux/sparc/Dist: Add init-first.h. * string/Makefile (distribute): Add tst-svc.expect. * nis/Makefile (distribute): Add nis_intern.h and Banner. * elf/Makefile (distribute): Add dl-hash.h. * Rules (subdir_echo-distinfo): Add headers from $(distribute). * login/Makefile (others): Add utmpdump. * sysdeps/unix/sysv/linux/Makefile (sysdep_headers): Add net/if_slip.h. * manual/Makefile (dir-add.texi): Also look in indirectly included files. 1997-06-16 23:15 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/unix/sysv/linux/configure.in: Remove commands to generate stdio_lim.h. * sysdeps/unix/sysv/linux/mk-stdiolim.c: Remove. * sysdeps/unix/sysv/linux/Makefile: Add rules to generate stdio_lim.h here. (common-generated): Add generated files. (inhibit-stdio_lim): Define. * sysdeps/posix/Makefile [$(inhibit-stdio_lim)=yes]: Disable rules to generate stdio_lim.h. * sysdeps/unix/sysv/linux/Makefile: Suppress inclusion of dependecy files if no_deps is set. ($(objpfx)syscall-%.d): Add header file as target to dependency generation. 1997-06-14 19:19 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/unix/sysv/linux/arm/Dist: New file. 1997-06-14 17:59 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * extra-lib.mk (others): Depend on versioned shared library, not the unversioned one. * Makerules (build-shlib): Don't make the version link here. ($(common-objpfx)libc.so$(libc.so-version)): New rule for libc version link. 1997-06-16 03:07 Ulrich Drepper <drepper@cygnus.com> * sysdeps/libm-ieee754/k_standard.c: Undo change of Tue Aug 6 01:13:56 1996. * argp/argp-help.c (argp_args_usage): Supply correct argp to filter_doc. * argp/argp-help.c (hol_add_cluster): Initialize CL->depth. * argp/argp-help.c (_help): Supply STATE to argp_args_usage. * argp/argp.h (ARGP_KEY_HELP_ARGS_DOC): New macro. * argp/argp-fmtstream.c: Add casts to prevent warnings. * argp/argp.h (OPTION_NO_USAGE): New macro. * argp/argp-help.c (usage_long_opt, usage_argful_short_opt, * argp/argp-fmtstream.c (__argp_fmtstream_update): Account for case * argp/argp-help.c <stddef.h>: New include. * argp/argp.h (argp_state_help, __argp_state_help, argp_usage, * argp/argp.h (argp_program_bug_address): Make const. * argp/argp-parse.c (argp_default_parser): Set STATE->name for OPT_PROGNAME. * argp/argp-help.c (__argp_error, __argp_failure, __argp_state_help): * argp/argp-parse.c (parser_init): Set PARSER->state.flags.
1997-06-19 19:11:22 +00:00
double
__ieee754_gamma_r (double x, int *signgamp)
{
int32_t hx;
Consistently use uintN_t not u_intN_t in libm. This patch changes libm code to make consistent use of C99 uintN_t types instead of sometimes using those and sometimes using the older nonstandard u_intN_t names. This makes sense as a cleanup in its own right, and also facilitates merges to GCC's libquadmath (which gets the types from stdint.h and so may not have u_intN_t available at all). Tested for x86_64, and with build-many-glibcs.py. * math/s_nextafter.c (__nextafter): Use uintN_t instead of u_intN_t. * math/s_nexttowardf.c (__nexttowardf): Likewise. * sysdeps/generic/math_private.h (ieee_double_shape_type): Likewise. (ieee_float_shape_type): Likewise. * sysdeps/i386/fpu/s_fpclassifyl.c (__fpclassifyl): Likewise. * sysdeps/i386/fpu/s_isnanl.c (__isnanl): Likewise. * sysdeps/i386/fpu/s_nextafterl.c (__nextafterl): Likewise. * sysdeps/i386/fpu/s_nexttoward.c (__nexttoward): Likewise. * sysdeps/i386/fpu/s_nexttowardf.c (__nexttowardf): Likewise. * sysdeps/ieee754/dbl-64/e_acosh.c (__ieee754_acosh): Likewise. * sysdeps/ieee754/dbl-64/e_cosh.c (__ieee754_cosh): Likewise. * sysdeps/ieee754/dbl-64/e_fmod.c (__ieee754_fmod): Likewise. * sysdeps/ieee754/dbl-64/e_gamma_r.c (__ieee754_gamma_r): Likewise. * sysdeps/ieee754/dbl-64/e_hypot.c (__ieee754_hypot): Likewise. * sysdeps/ieee754/dbl-64/e_jn.c (__ieee754_jn): Likewise. (__ieee754_yn): Likewise. * sysdeps/ieee754/dbl-64/e_log10.c (__ieee754_log10): Likewise. * sysdeps/ieee754/dbl-64/e_log2.c (__ieee754_log2): Likewise. * sysdeps/ieee754/dbl-64/e_rem_pio2.c (__ieee754_rem_pio2): Likewise. * sysdeps/ieee754/dbl-64/e_sinh.c (__ieee754_sinh): Likewise. * sysdeps/ieee754/dbl-64/s_ceil.c (__ceil): Likewise. * sysdeps/ieee754/dbl-64/s_copysign.c (__copysign): Likewise. * sysdeps/ieee754/dbl-64/s_erf.c (__erf): Likewise. (__erfc): Likewise. * sysdeps/ieee754/dbl-64/s_expm1.c (__expm1): Likewise. * sysdeps/ieee754/dbl-64/s_finite.c (FINITE): Likewise. * sysdeps/ieee754/dbl-64/s_floor.c (__floor): Likewise. * sysdeps/ieee754/dbl-64/s_fpclassify.c (__fpclassify): Likewise. * sysdeps/ieee754/dbl-64/s_isnan.c (__isnan): Likewise. * sysdeps/ieee754/dbl-64/s_issignaling.c (__issignaling): Likewise. * sysdeps/ieee754/dbl-64/s_llrint.c (__llrint): Likewise. * sysdeps/ieee754/dbl-64/s_llround.c (__llround): Likewise. * sysdeps/ieee754/dbl-64/s_lrint.c (__lrint): Likewise. * sysdeps/ieee754/dbl-64/s_lround.c (__lround): Likewise. * sysdeps/ieee754/dbl-64/s_modf.c (__modf): Likewise. * sysdeps/ieee754/dbl-64/s_nextup.c (__nextup): Likewise. * sysdeps/ieee754/dbl-64/s_remquo.c (__remquo): Likewise. * sysdeps/ieee754/dbl-64/s_round.c (__round): Likewise. * sysdeps/ieee754/dbl-64/s_trunc.c (__trunc): Likewise. * sysdeps/ieee754/dbl-64/wordsize-64/s_issignaling.c (__issignaling): Likewise. * sysdeps/ieee754/flt-32/e_atan2f.c (__ieee754_atan2f): Likewise. * sysdeps/ieee754/flt-32/e_fmodf.c (__ieee754_fmodf): Likewise. * sysdeps/ieee754/flt-32/e_gammaf_r.c (__ieee754_gammaf_r): Likewise. * sysdeps/ieee754/flt-32/e_jnf.c (__ieee754_ynf): Likewise. * sysdeps/ieee754/flt-32/e_log10f.c (__ieee754_log10f): Likewise. * sysdeps/ieee754/flt-32/e_powf.c (__ieee754_powf): Likewise. * sysdeps/ieee754/flt-32/e_rem_pio2f.c (__ieee754_rem_pio2f): Likewise. * sysdeps/ieee754/flt-32/e_remainderf.c (__ieee754_remainderf): Likewise. * sysdeps/ieee754/flt-32/e_sqrtf.c (__ieee754_sqrtf): Likewise. * sysdeps/ieee754/flt-32/s_ceilf.c (__ceilf): Likewise. * sysdeps/ieee754/flt-32/s_copysignf.c (__copysignf): Likewise. * sysdeps/ieee754/flt-32/s_erff.c (__erff): Likewise. (__erfcf): Likewise. * sysdeps/ieee754/flt-32/s_expm1f.c (__expm1f): Likewise. * sysdeps/ieee754/flt-32/s_finitef.c (FINITEF): Likewise. * sysdeps/ieee754/flt-32/s_floorf.c (__floorf): Likewise. * sysdeps/ieee754/flt-32/s_fpclassifyf.c (__fpclassifyf): Likewise. * sysdeps/ieee754/flt-32/s_isnanf.c (__isnanf): Likewise. * sysdeps/ieee754/flt-32/s_issignalingf.c (__issignalingf): Likewise. * sysdeps/ieee754/flt-32/s_llrintf.c (__llrintf): Likewise. * sysdeps/ieee754/flt-32/s_llroundf.c (__llroundf): Likewise. * sysdeps/ieee754/flt-32/s_lrintf.c (__lrintf): Likewise. * sysdeps/ieee754/flt-32/s_lroundf.c (__lroundf): Likewise. * sysdeps/ieee754/flt-32/s_modff.c (__modff): Likewise. * sysdeps/ieee754/flt-32/s_remquof.c (__remquof): Likewise. * sysdeps/ieee754/flt-32/s_roundf.c (__roundf): Likewise. * sysdeps/ieee754/ldbl-128/e_acoshl.c (__ieee754_acoshl): Likewise. * sysdeps/ieee754/ldbl-128/e_atan2l.c (__ieee754_atan2l): Likewise. * sysdeps/ieee754/ldbl-128/e_atanhl.c (__ieee754_atanhl): Likewise. * sysdeps/ieee754/ldbl-128/e_fmodl.c (__ieee754_fmodl): Likewise. * sysdeps/ieee754/ldbl-128/e_gammal_r.c (__ieee754_gammal_r): Likewise. * sysdeps/ieee754/ldbl-128/e_hypotl.c (__ieee754_hypotl): Likewise. * sysdeps/ieee754/ldbl-128/e_jnl.c (__ieee754_jnl): Likewise. (__ieee754_ynl): Likewise. * sysdeps/ieee754/ldbl-128/e_powl.c (__ieee754_powl): Likewise. * sysdeps/ieee754/ldbl-128/e_rem_pio2l.c (__ieee754_rem_pio2l): Likewise. * sysdeps/ieee754/ldbl-128/e_remainderl.c (__ieee754_remainderl): Likewise. * sysdeps/ieee754/ldbl-128/e_sinhl.c (__ieee754_sinhl): Likewise. * sysdeps/ieee754/ldbl-128/k_cosl.c (__kernel_cosl): Likewise. * sysdeps/ieee754/ldbl-128/k_sincosl.c (__kernel_sincosl): Likewise. * sysdeps/ieee754/ldbl-128/k_sinl.c (__kernel_sinl): Likewise. * sysdeps/ieee754/ldbl-128/s_ceill.c (__ceill): Likewise. * sysdeps/ieee754/ldbl-128/s_copysignl.c (__copysignl): Likewise. * sysdeps/ieee754/ldbl-128/s_erfl.c (__erfcl): Likewise. * sysdeps/ieee754/ldbl-128/s_fabsl.c (__fabsl): Likewise. * sysdeps/ieee754/ldbl-128/s_finitel.c (__finitel): Likewise. * sysdeps/ieee754/ldbl-128/s_floorl.c (__floorl): Likewise. * sysdeps/ieee754/ldbl-128/s_fpclassifyl.c (__fpclassifyl): Likewise. * sysdeps/ieee754/ldbl-128/s_frexpl.c (__frexpl): Likewise. * sysdeps/ieee754/ldbl-128/s_isnanl.c (__isnanl): Likewise. * sysdeps/ieee754/ldbl-128/s_issignalingl.c (__issignalingl): Likewise. * sysdeps/ieee754/ldbl-128/s_llrintl.c (__llrintl): Likewise. * sysdeps/ieee754/ldbl-128/s_llroundl.c (__llroundl): Likewise. * sysdeps/ieee754/ldbl-128/s_lrintl.c (__lrintl): Likewise. * sysdeps/ieee754/ldbl-128/s_lroundl.c (__lroundl): Likewise. * sysdeps/ieee754/ldbl-128/s_modfl.c (__modfl): Likewise. * sysdeps/ieee754/ldbl-128/s_nearbyintl.c (__nearbyintl): Likewise. * sysdeps/ieee754/ldbl-128/s_nextafterl.c (__nextafterl): Likewise. * sysdeps/ieee754/ldbl-128/s_nexttoward.c (__nexttoward): Likewise. * sysdeps/ieee754/ldbl-128/s_nexttowardf.c (__nexttowardf): Likewise. * sysdeps/ieee754/ldbl-128/s_nextupl.c (__nextupl): Likewise. * sysdeps/ieee754/ldbl-128/s_remquol.c (__remquol): Likewise. * sysdeps/ieee754/ldbl-128/s_rintl.c (__rintl): Likewise. * sysdeps/ieee754/ldbl-128/s_roundl.c (__roundl): Likewise. * sysdeps/ieee754/ldbl-128/s_tanhl.c (__tanhl): Likewise. * sysdeps/ieee754/ldbl-128/s_truncl.c (__truncl): Likewise. * sysdeps/ieee754/ldbl-128ibm/e_fmodl.c (__ieee754_fmodl): Likewise. * sysdeps/ieee754/ldbl-128ibm/e_gammal_r.c (__ieee754_gammal_r): Likewise. * sysdeps/ieee754/ldbl-128ibm/e_powl.c (__ieee754_powl): Likewise. * sysdeps/ieee754/ldbl-128ibm/e_rem_pio2l.c (__ieee754_rem_pio2l): Likewise. * sysdeps/ieee754/ldbl-128ibm/e_remainderl.c (__ieee754_remainderl): Likewise. * sysdeps/ieee754/ldbl-128ibm/k_cosl.c (__kernel_cosl): Likewise. * sysdeps/ieee754/ldbl-128ibm/k_sinl.c (__kernel_sinl): Likewise. * sysdeps/ieee754/ldbl-128ibm/s_fabsl.c (__fabsl): Likewise. * sysdeps/ieee754/ldbl-128ibm/s_fpclassifyl.c (___fpclassifyl): Likewise. * sysdeps/ieee754/ldbl-128ibm/s_modfl.c (__modfl): Likewise. * sysdeps/ieee754/ldbl-128ibm/s_nexttowardf.c (__nexttowardf): Likewise. * sysdeps/ieee754/ldbl-128ibm/s_remquol.c (__remquol): Likewise. * sysdeps/ieee754/ldbl-96/e_acoshl.c (__ieee754_acoshl): Likewise. * sysdeps/ieee754/ldbl-96/e_asinl.c (__ieee754_asinl): Likewise. * sysdeps/ieee754/ldbl-96/e_atanhl.c (__ieee754_atanhl): Likewise. * sysdeps/ieee754/ldbl-96/e_coshl.c (__ieee754_coshl): Likewise. * sysdeps/ieee754/ldbl-96/e_gammal_r.c (__ieee754_gammal_r): Likewise. * sysdeps/ieee754/ldbl-96/e_hypotl.c (__ieee754_hypotl): Likewise. * sysdeps/ieee754/ldbl-96/e_j0l.c (__ieee754_j0l): Likewise. (__ieee754_y0l): Likewise. (pzero): Likewise. (qzero): Likewise. * sysdeps/ieee754/ldbl-96/e_j1l.c (__ieee754_j1l): Likewise. (__ieee754_y1l): Likewise. (pone): Likewise. (qone): Likewise. * sysdeps/ieee754/ldbl-96/e_jnl.c (__ieee754_jnl): Likewise. (__ieee754_ynl): Likewise. * sysdeps/ieee754/ldbl-96/e_lgammal_r.c (sin_pi): Likewise. (__ieee754_lgammal_r): Likewise. * sysdeps/ieee754/ldbl-96/e_rem_pio2l.c (__ieee754_rem_pio2l): Likewise. * sysdeps/ieee754/ldbl-96/e_sinhl.c (__ieee754_sinhl): Likewise. * sysdeps/ieee754/ldbl-96/s_copysignl.c (__copysignl): Likewise. * sysdeps/ieee754/ldbl-96/s_erfl.c (__erfl): Likewise. (__erfcl): Likewise. * sysdeps/ieee754/ldbl-96/s_frexpl.c (__frexpl): Likewise. * sysdeps/ieee754/ldbl-96/s_issignalingl.c (__issignalingl): Likewise. * sysdeps/ieee754/ldbl-96/s_llrintl.c (__llrintl): Likewise. * sysdeps/ieee754/ldbl-96/s_llroundl.c (__llroundl): Likewise. * sysdeps/ieee754/ldbl-96/s_lrintl.c (__lrintl): Likewise. * sysdeps/ieee754/ldbl-96/s_lroundl.c (__lroundl): Likewise. * sysdeps/ieee754/ldbl-96/s_modfl.c (__modfl): Likewise. * sysdeps/ieee754/ldbl-96/s_nexttoward.c (__nexttoward): Likewise. * sysdeps/ieee754/ldbl-96/s_nexttowardf.c (__nexttowardf): Likewise. * sysdeps/ieee754/ldbl-96/s_nextupl.c (__nextupl): Likewise. * sysdeps/ieee754/ldbl-96/s_remquol.c (__remquol): Likewise. * sysdeps/ieee754/ldbl-96/s_roundl.c (__roundl): Likewise. * sysdeps/ieee754/ldbl-96/s_tanhl.c (__tanhl): Likewise. * sysdeps/ieee754/ldbl-opt/s_nexttowardfd.c (__nldbl_nexttowardf): Likewise. * sysdeps/m68k/m680x0/fpu/e_pow.c (s(__ieee754_pow)): Likewise. * sysdeps/m68k/m680x0/fpu/s_fpclassifyl.c (__fpclassifyl): Likewise. * sysdeps/m68k/m680x0/fpu/s_llrint.c (__llrint): Likewise. * sysdeps/m68k/m680x0/fpu/s_llrintf.c (__llrintf): Likewise. * sysdeps/m68k/m680x0/fpu/s_llrintl.c (__llrintl): Likewise. * sysdeps/m68k/m680x0/fpu/s_nextafterl.c (__nextafterl): Likewise. * sysdeps/x86/fpu/powl_helper.c (__powl_helper): Likewise.
2017-08-03 19:55:04 +00:00
uint32_t lx;
Improve tgamma accuracy (bug 18613). In non-default rounding modes, tgamma can be slightly less accurate than permitted by glibc's accuracy goals. Part of the problem is error accumulation, addressed in this patch by setting round-to-nearest for internal computations. However, there was also a bug in the code dealing with computing pow (x + n, x + n) where x + n is not exactly representable, providing another source of error even in round-to-nearest mode; it was necessary to address both bugs to get errors for all testcases within glibc's accuracy goals. Given this second fix, accuracy in round-to-nearest mode is also improved (hence regeneration of ulps for tgamma should be from scratch - truncate libm-test-ulps or at least remove existing tgamma entries - so that the expected ulps can be reduced). Some additional complications also arose. Certain tgamma tests should strictly, according to IEEE semantics, overflow or not depending on the rounding mode; this is beyond the scope of glibc's accuracy goals for any function without exactly-determined results, but gen-auto-libm-tests doesn't handle being lax there as it does for underflow. (libm-test.inc also doesn't handle being lax about whether the result in cases very close to the overflow threshold is infinity or a finite value close to overflow, but that doesn't cause problems in this case though I've seen it cause problems with random test generation for some functions.) Thus, spurious-overflow markings, with a comment, are added to auto-libm-test-in (no bug in Bugzilla because the issue is with the testsuite, not a user-visible bug in glibc). And on x86, after the patch I saw ERANGE issues as previously reported by Carlos (see my commentary in <https://sourceware.org/ml/libc-alpha/2015-01/msg00485.html>), which needed addressing by ensuring excess range and precision were eliminated at various points if FLT_EVAL_METHOD != 0. I also noticed and fixed a cosmetic issue where 1.0f was used in long double functions and should have been 1.0L. This completes the move of all functions to testing in all rounding modes with ALL_RM_TEST, so gen-libm-have-vector-test.sh is updated to remove the workaround for some functions not using ALL_RM_TEST. Tested for x86_64, x86, mips64 and powerpc. [BZ #18613] * sysdeps/ieee754/dbl-64/e_gamma_r.c (gamma_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gamma_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. * sysdeps/ieee754/flt-32/e_gammaf_r.c (gammaf_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammaf_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. * sysdeps/ieee754/ldbl-128/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * sysdeps/ieee754/ldbl-128ibm/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * sysdeps/ieee754/ldbl-96/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * math/libm-test.inc (tgamma_test_data): Remove one test. Moved to auto-libm-test-in. (tgamma_test): Use ALL_RM_TEST. * math/auto-libm-test-in: Add one test of tgamma. Mark some other tests of tgamma with spurious-overflow. * math/auto-libm-test-out: Regenerated. * math/gen-libm-have-vector-test.sh: Do not check for START. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-06-29 23:29:35 +00:00
double ret;
Update. 1997-06-19 19:38 Ulrich Drepper <drepper@cygnus.com> * features.h: Define __STDC_IEC_559__ and _STDC_IEC_559_COMPLEX__. * elf/dl-minimal.c (__dcgettext): Remove assertion. * inet/rcmd.c: Correct a few typos. Reported by Eric Troan. * manual/Makefile (distribute): Add dir. * manual/dir: New file. * math/Makefile (libm-support): Rename s_rinttol, s_rinttoll, s_roundtol, and s_roundtoll to s_lrint, s_llrint, s_lround, and s_llround respectively. (libm-calls): Add e_gamma_r. * math/libm-test.c (check_int_exc): New function. (signbit_test): Remove test for sign of NaN. (gamma_test): Clear exception after test of existence. Correct tests and and epsilons. (lgamma_test): Likewise. (ilogb_test): Correct all tests. (scalb_test): Rewrite. (rinttol_test): Rename to lrint_test and correct tests. (rinttoll_test): Likewise. (roundtol_test): Likewise. (roundtoll_test): Likewise. (main): Call lrint/lround functions instead of rinttol/roundtol. * math/math.h: Change prototypes for rinttol/roundtol. * math/mathcalls.h: Rearrange prototypes according to ISO C9X draft. * sysdeps/generic/mathbits.h: Define FP_ILOGB0 and FP_ILOGBNAN. * sysdeps/i386/mathbits.h: Likewise. * sysdeps/libm-i387/e_scalb.S: Handle special cases correctly. * sysdeps/libm-i387/e_scalbf.S: Likewise. * sysdeps/libm-i387/e_scalbl.S: Likewise. * sysdeps/libm-i387/s_asinh.S: Handle -inf correctly. * sysdeps/libm-i387/s_asinhf.S: Likewise. * sysdeps/libm-i387/s_asinhl.S: Likewise. * sysdeps/libm-i387/s_ilogb.S: Optimize. * sysdeps/libm-i387/s_ilogbf.S: Likewise. * sysdeps/libm-i387/s_ilogbl.S: Likewise. * sysdeps/libm-i387/s_rinttol.S: Rename to... * sysdeps/libm-i387/s_lrint.S: ...this. * sysdeps/libm-i387/s_rinttoll.S: Rename to... * sysdeps/libm-i387/s_llrint.S: ...this. * sysdeps/libm-i387/s_remquo.S: Correctly set sign of remainder. * sysdeps/libm-i387/s_remquof.S: Likewise. * sysdeps/libm-i387/s_remquol.S: Likewise. * sysdeps/libm-i387/e_gamma_r.c: New file. Implementation of gamma function according to ISO C. * sysdeps/libm-i387/e_gammaf_r.c: New file. * sysdeps/libm-i387/e_gammal_r.c: New file. * sysdeps/libm-i387/e_lgamma_r.c: Don't let optimize compile the generation of exceptions away. * sysdeps/libm-i387/e_lgammaf_r.c: Likewise. * sysdeps/libm-i387/k_standard.c: Correct return value for infinity points of gamma function when not SVID mode. * sysdeps/libm-i387/s_rinttoll.c: Renamed to... * sysdeps/libm-i387/s_llrint.c: ...this. * sysdeps/libm-i387/s_rinttol.c: Renamed to... * sysdeps/libm-i387/s_lrint.c: ...this. * sysdeps/libm-i387/s_roundtoll.c: Renamed to... * sysdeps/libm-i387/s_llround.c: ...this. * sysdeps/libm-i387/s_roundtol.c: Renamed to... * sysdeps/libm-i387/s_lround.c: ..this. * sysdeps/libm-i387/s_scalbn.c: Change second parameter according to ISO C. * sysdeps/libm-i387/s_scalbnf.c: Likewise. * sysdeps/libm-i387/s_scalbnl.c: Likewise. * sysdeps/libm-i387/w_gamma.c: Call __ieee754_gamma_r if library mode is not _SVID_. * sysdeps/libm-i387/w_gammaf.c: Likewise. * sysdeps/libm-i387/w_gammal.c: Likewise. * sysdeps/m68k/fpu/__math.h: Rename __rinttol to __lrint and rinttol to lrint. * sysdeps/m68k/fpu/s_rinttol.c: Renamed to... * sysdeps/m68k/fpu/s_lrint.c: ...this. * sysdeps/m68k/fpu/s_rinttoll.c: Renamed to... * sysdeps/m68k/fpu/s_llrint.c: ...this. * md5-crypt/Makefile: Link md5test program with md5.o. * stdio-common/temptest.c: Don't use __stdio_gen_tempname which is not exported by the libc.so. * stdio-common/vfscanf.c: Correct scanning of strings after last change. * sysdeps/unix/sysv/linux/i386/sysdep.S: Use .comm to define errno. 1997-06-19 07:37 H.J. Lu <hjl@gnu.ai.mit.edu> * time/tzfile.c (__tzfile_read): Store getc () return in int. 1997-06-13 Miles Bader <miles@gnu.ai.mit.edu> * argp/argp-parse.c (argp_version_parser): Include `(PROGRAM ERROR)' in the no-version error text to indicate that something's fucked. [!_] (N_): New macro. (argp_default_options, argp_version_options): Wrap doc strings in N_(). 1997-06-12 Miles Bader <miles@gnu.ai.mit.edu> * argp/argp-parse.c (parser_parse_opt): Detect and report unhandled options here. (parser_parse_arg): Handle ARGP_KEY_ARGS here. Adjust NEXT pointer back if we fail to parse anything. (parser_parse_next): Simplify arg code. Leave state NEXT frobbing to parser_parse_arg. 1997-06-11 Miles Bader <miles@gnu.ai.mit.edu> * argp/argp.h (ARGP_KEY_ARGS, ARGP_KEY_FINI): New macros. * argp/argp-parse.c (parser_finalize): Do another pass over the parsers with ARGP_KEY_FINI. 1997-06-18 Miles Bader <miles@gnu.ai.mit.edu> * string/Makefile (routines): Add argz-replace. 1997-06-16 00:16 Miles Bader <miles@gnu.ai.mit.edu> * manual/string.texi (Argz Functions): Document argz_replace. 1997-06-12 Miles Bader <miles@gnu.ai.mit.edu> * string/argz.h (__argz_replace, argz_replace): New declarations. * string/argz-replace.c: New file. 1997-06-11 Miles Bader <miles@gnu.ai.mit.edu> * manual/argp.texi (Argp Special Keys): Document ARGP_KEY_ARGS. 1997-06-16 23:08 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * Makeconfig (libc-map): Remove definition. * Makerules (libc-map): Define it here, using the full name. (load-map-file): Remove case for empty $(..). ($(common-objpfx)libc.so): Revert last change. 1997-06-17 22:18 Mark Kettenis <kettenis@phys.uva.nl> * login/programs/utmpd.c (main): Improve signal handling. * login/programs/request.c (do_setutent, do_updwtmp): Allow arbitrary length filenames. * login/programs/utmpd.h (setutent_request, updwtmp_request): Get rid of fixed length file field. * login/utmp_daemon.c (do_setutent, do_updwtmp): Allow arbitrary length filenames. * login/programs/request.c (do_pututline): Don't fail if connection->position is -1 on entry. 1997-06-15 16:32 Mark Kettenis <kettenis@phys.uva.nl> * login/utmp_file.c (updwtmp_file): Use the same method for appending an entry as in pututline_file. 1997-06-11 18:59 Mark Kettenis <kettenis@phys.uva.nl> * login/utmp_file.c (getutent_r_file): Use read lock instead of write lock. (getutline_r_file, internal_getut_r): Lock utmp file. (updwtmp_file): Use fcntl to lock file instead of flock. 1997-06-18 00:11 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * stdio-common/vfscanf.c (inchar, ungetc): Don't count EOF as character read in/put back. * stdio-common/tstscanf.c: Add test case for this. 1997-06-17 22:17 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * libc.map: Add more symbols. 1997-06-18 12:01 Ulrich Drepper <drepper@cygnus.com> * manual/Makefile (install): Make sure `dir' file exists if we use install-info. 1997-06-17 19:32 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * manual/Makefile (info): Depend on dir-add.info. 1997-06-17 17:19 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * elf/ldd.bash.in: Fix spacing in message. 1997-06-17 14:28 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * elf/do-rel.h (elf_dynamic_do_rel): Always use version information if available. 1997-06-17 11:34 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sunrpc/Makefile ($(rpcsvc:%.x=$(objpfx)rpcsvc/%.h)): Make command non-empty to force make to recheck modification time. ($(rpcsvc:%.x=$(objpfx)x%.c)): Likewise. 1997-06-17 00:26 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/stub/e_acoshl.c: Set errno to ENOSYS. * sysdeps/stub/e_acosl.c: Likewise. * sysdeps/stub/e_asinl.c: Likewise. * sysdeps/stub/e_atan2l.c: Likewise. * sysdeps/stub/e_expl.c: Likewise. * sysdeps/stub/e_fmodl.c: Likewise. * sysdeps/stub/e_j0l.c: Likewise. * sysdeps/stub/e_j1l.c: Likewise. * sysdeps/stub/e_jnl.c: Likewise. * sysdeps/stub/e_lgammal_r.c: Likewise. * sysdeps/stub/e_log10l.c: Likewise. * sysdeps/stub/e_logl.c: Likewise. * sysdeps/stub/e_powl.c: Likewise. * sysdeps/stub/e_rem_pio2l.c: Likewise. * sysdeps/stub/e_sqrtl.c: Likewise. * sysdeps/stub/k_cosl.c: Likewise. * sysdeps/stub/k_rem_pio2l.c: Likewise. * sysdeps/stub/k_sinl.c: Likewise. * sysdeps/stub/k_tanl.c: Likewise. * sysdeps/stub/s_atanl.c: Likewise. * sysdeps/stub/s_erfl.c: Likewise. * sysdeps/stub/s_exp2.c: Likewise. * sysdeps/stub/s_exp2f.c: Likewise. * sysdeps/stub/s_exp2l.c: Likewise. * sysdeps/stub/s_expm1l.c: Likewise. * sysdeps/stub/s_log1pl.c: Likewise. * sysdeps/stub/s_log2l.c: Likewise. 1997-06-18 11:46 Ulrich Drepper <drepper@cygnus.com> * sysdeps/libm-ieee754/w_gamma.c: If _LIB_VERSION is _SVID_ compute result as before last change. * sysdeps/libm-ieee754/w_gammaf.c: Likewise. * sysdeps/libm-ieee754/w_gammal.c: Likewise. 1997-06-16 23:37 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/libm-ieee754/s_remquo.c: Fix off-by-one when computing quotient. * sysdeps/libm-ieee754/s_remquof.c: Likewise. * sysdeps/libm-ieee754/s_remquol.c: Likewise. * sysdeps/m68k/fpu/s_remquo.c: Remove FIXME and special case for quotient. * sysdeps/libm-ieee754/w_gamma.c: Add missing call to exp function. Don't use global signgam. * sysdeps/libm-ieee754/w_gammaf.c: Likewise. * sysdeps/libm-ieee754/w_gammal.c: Likewise. * math/Makefile (libm-calls): Remove w_gamma_r. * sysdeps/libm-ieee754/w_gamma_r.c: Remove file. * sysdeps/libm-ieee754/w_gammaf_r.c: Remove file. * sysdeps/libm-ieee754/w_gammal_r.c: Remove file. * math/libm-test.c (atanh_test): Declare x only if needed. (signbit_test): Fix typo. (gamma_test): Check whether function is implemented. Add epsilons. (lgamma_test): Likewise. (fmod_test): Add epsilons. (exp2_test): Use right function for existence test. 1997-06-07 09:20 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/powerpc/Dist: Add fenv_const.c. * sysdeps/unix/sysv/linux/Dist: Add net/if_slip.h. * sysdeps/unix/sysv/linux/powerpc/Dist: Add init-first.h and syscall.h. * sysdeps/unix/sysv/linux/sparc/Dist: Add init-first.h. * string/Makefile (distribute): Add tst-svc.expect. * nis/Makefile (distribute): Add nis_intern.h and Banner. * elf/Makefile (distribute): Add dl-hash.h. * Rules (subdir_echo-distinfo): Add headers from $(distribute). * login/Makefile (others): Add utmpdump. * sysdeps/unix/sysv/linux/Makefile (sysdep_headers): Add net/if_slip.h. * manual/Makefile (dir-add.texi): Also look in indirectly included files. 1997-06-16 23:15 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/unix/sysv/linux/configure.in: Remove commands to generate stdio_lim.h. * sysdeps/unix/sysv/linux/mk-stdiolim.c: Remove. * sysdeps/unix/sysv/linux/Makefile: Add rules to generate stdio_lim.h here. (common-generated): Add generated files. (inhibit-stdio_lim): Define. * sysdeps/posix/Makefile [$(inhibit-stdio_lim)=yes]: Disable rules to generate stdio_lim.h. * sysdeps/unix/sysv/linux/Makefile: Suppress inclusion of dependecy files if no_deps is set. ($(objpfx)syscall-%.d): Add header file as target to dependency generation. 1997-06-14 19:19 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/unix/sysv/linux/arm/Dist: New file. 1997-06-14 17:59 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * extra-lib.mk (others): Depend on versioned shared library, not the unversioned one. * Makerules (build-shlib): Don't make the version link here. ($(common-objpfx)libc.so$(libc.so-version)): New rule for libc version link. 1997-06-16 03:07 Ulrich Drepper <drepper@cygnus.com> * sysdeps/libm-ieee754/k_standard.c: Undo change of Tue Aug 6 01:13:56 1996. * argp/argp-help.c (argp_args_usage): Supply correct argp to filter_doc. * argp/argp-help.c (hol_add_cluster): Initialize CL->depth. * argp/argp-help.c (_help): Supply STATE to argp_args_usage. * argp/argp.h (ARGP_KEY_HELP_ARGS_DOC): New macro. * argp/argp-fmtstream.c: Add casts to prevent warnings. * argp/argp.h (OPTION_NO_USAGE): New macro. * argp/argp-help.c (usage_long_opt, usage_argful_short_opt, * argp/argp-fmtstream.c (__argp_fmtstream_update): Account for case * argp/argp-help.c <stddef.h>: New include. * argp/argp.h (argp_state_help, __argp_state_help, argp_usage, * argp/argp.h (argp_program_bug_address): Make const. * argp/argp-parse.c (argp_default_parser): Set STATE->name for OPT_PROGNAME. * argp/argp-help.c (__argp_error, __argp_failure, __argp_state_help): * argp/argp-parse.c (parser_init): Set PARSER->state.flags.
1997-06-19 19:11:22 +00:00
EXTRACT_WORDS (hx, lx, x);
if (__glibc_unlikely (((hx & 0x7fffffff) | lx) == 0))
{
/* Return value for x == 0 is Inf with divide by zero exception. */
*signgamp = 0;
return 1.0 / x;
}
if (__builtin_expect (hx < 0, 0)
Consistently use uintN_t not u_intN_t in libm. This patch changes libm code to make consistent use of C99 uintN_t types instead of sometimes using those and sometimes using the older nonstandard u_intN_t names. This makes sense as a cleanup in its own right, and also facilitates merges to GCC's libquadmath (which gets the types from stdint.h and so may not have u_intN_t available at all). Tested for x86_64, and with build-many-glibcs.py. * math/s_nextafter.c (__nextafter): Use uintN_t instead of u_intN_t. * math/s_nexttowardf.c (__nexttowardf): Likewise. * sysdeps/generic/math_private.h (ieee_double_shape_type): Likewise. (ieee_float_shape_type): Likewise. * sysdeps/i386/fpu/s_fpclassifyl.c (__fpclassifyl): Likewise. * sysdeps/i386/fpu/s_isnanl.c (__isnanl): Likewise. * sysdeps/i386/fpu/s_nextafterl.c (__nextafterl): Likewise. * sysdeps/i386/fpu/s_nexttoward.c (__nexttoward): Likewise. * sysdeps/i386/fpu/s_nexttowardf.c (__nexttowardf): Likewise. * sysdeps/ieee754/dbl-64/e_acosh.c (__ieee754_acosh): Likewise. * sysdeps/ieee754/dbl-64/e_cosh.c (__ieee754_cosh): Likewise. * sysdeps/ieee754/dbl-64/e_fmod.c (__ieee754_fmod): Likewise. * sysdeps/ieee754/dbl-64/e_gamma_r.c (__ieee754_gamma_r): Likewise. * sysdeps/ieee754/dbl-64/e_hypot.c (__ieee754_hypot): Likewise. * sysdeps/ieee754/dbl-64/e_jn.c (__ieee754_jn): Likewise. (__ieee754_yn): Likewise. * sysdeps/ieee754/dbl-64/e_log10.c (__ieee754_log10): Likewise. * sysdeps/ieee754/dbl-64/e_log2.c (__ieee754_log2): Likewise. * sysdeps/ieee754/dbl-64/e_rem_pio2.c (__ieee754_rem_pio2): Likewise. * sysdeps/ieee754/dbl-64/e_sinh.c (__ieee754_sinh): Likewise. * sysdeps/ieee754/dbl-64/s_ceil.c (__ceil): Likewise. * sysdeps/ieee754/dbl-64/s_copysign.c (__copysign): Likewise. * sysdeps/ieee754/dbl-64/s_erf.c (__erf): Likewise. (__erfc): Likewise. * sysdeps/ieee754/dbl-64/s_expm1.c (__expm1): Likewise. * sysdeps/ieee754/dbl-64/s_finite.c (FINITE): Likewise. * sysdeps/ieee754/dbl-64/s_floor.c (__floor): Likewise. * sysdeps/ieee754/dbl-64/s_fpclassify.c (__fpclassify): Likewise. * sysdeps/ieee754/dbl-64/s_isnan.c (__isnan): Likewise. * sysdeps/ieee754/dbl-64/s_issignaling.c (__issignaling): Likewise. * sysdeps/ieee754/dbl-64/s_llrint.c (__llrint): Likewise. * sysdeps/ieee754/dbl-64/s_llround.c (__llround): Likewise. * sysdeps/ieee754/dbl-64/s_lrint.c (__lrint): Likewise. * sysdeps/ieee754/dbl-64/s_lround.c (__lround): Likewise. * sysdeps/ieee754/dbl-64/s_modf.c (__modf): Likewise. * sysdeps/ieee754/dbl-64/s_nextup.c (__nextup): Likewise. * sysdeps/ieee754/dbl-64/s_remquo.c (__remquo): Likewise. * sysdeps/ieee754/dbl-64/s_round.c (__round): Likewise. * sysdeps/ieee754/dbl-64/s_trunc.c (__trunc): Likewise. * sysdeps/ieee754/dbl-64/wordsize-64/s_issignaling.c (__issignaling): Likewise. * sysdeps/ieee754/flt-32/e_atan2f.c (__ieee754_atan2f): Likewise. * sysdeps/ieee754/flt-32/e_fmodf.c (__ieee754_fmodf): Likewise. * sysdeps/ieee754/flt-32/e_gammaf_r.c (__ieee754_gammaf_r): Likewise. * sysdeps/ieee754/flt-32/e_jnf.c (__ieee754_ynf): Likewise. * sysdeps/ieee754/flt-32/e_log10f.c (__ieee754_log10f): Likewise. * sysdeps/ieee754/flt-32/e_powf.c (__ieee754_powf): Likewise. * sysdeps/ieee754/flt-32/e_rem_pio2f.c (__ieee754_rem_pio2f): Likewise. * sysdeps/ieee754/flt-32/e_remainderf.c (__ieee754_remainderf): Likewise. * sysdeps/ieee754/flt-32/e_sqrtf.c (__ieee754_sqrtf): Likewise. * sysdeps/ieee754/flt-32/s_ceilf.c (__ceilf): Likewise. * sysdeps/ieee754/flt-32/s_copysignf.c (__copysignf): Likewise. * sysdeps/ieee754/flt-32/s_erff.c (__erff): Likewise. (__erfcf): Likewise. * sysdeps/ieee754/flt-32/s_expm1f.c (__expm1f): Likewise. * sysdeps/ieee754/flt-32/s_finitef.c (FINITEF): Likewise. * sysdeps/ieee754/flt-32/s_floorf.c (__floorf): Likewise. * sysdeps/ieee754/flt-32/s_fpclassifyf.c (__fpclassifyf): Likewise. * sysdeps/ieee754/flt-32/s_isnanf.c (__isnanf): Likewise. * sysdeps/ieee754/flt-32/s_issignalingf.c (__issignalingf): Likewise. * sysdeps/ieee754/flt-32/s_llrintf.c (__llrintf): Likewise. * sysdeps/ieee754/flt-32/s_llroundf.c (__llroundf): Likewise. * sysdeps/ieee754/flt-32/s_lrintf.c (__lrintf): Likewise. * sysdeps/ieee754/flt-32/s_lroundf.c (__lroundf): Likewise. * sysdeps/ieee754/flt-32/s_modff.c (__modff): Likewise. * sysdeps/ieee754/flt-32/s_remquof.c (__remquof): Likewise. * sysdeps/ieee754/flt-32/s_roundf.c (__roundf): Likewise. * sysdeps/ieee754/ldbl-128/e_acoshl.c (__ieee754_acoshl): Likewise. * sysdeps/ieee754/ldbl-128/e_atan2l.c (__ieee754_atan2l): Likewise. * sysdeps/ieee754/ldbl-128/e_atanhl.c (__ieee754_atanhl): Likewise. * sysdeps/ieee754/ldbl-128/e_fmodl.c (__ieee754_fmodl): Likewise. * sysdeps/ieee754/ldbl-128/e_gammal_r.c (__ieee754_gammal_r): Likewise. * sysdeps/ieee754/ldbl-128/e_hypotl.c (__ieee754_hypotl): Likewise. * sysdeps/ieee754/ldbl-128/e_jnl.c (__ieee754_jnl): Likewise. (__ieee754_ynl): Likewise. * sysdeps/ieee754/ldbl-128/e_powl.c (__ieee754_powl): Likewise. * sysdeps/ieee754/ldbl-128/e_rem_pio2l.c (__ieee754_rem_pio2l): Likewise. * sysdeps/ieee754/ldbl-128/e_remainderl.c (__ieee754_remainderl): Likewise. * sysdeps/ieee754/ldbl-128/e_sinhl.c (__ieee754_sinhl): Likewise. * sysdeps/ieee754/ldbl-128/k_cosl.c (__kernel_cosl): Likewise. * sysdeps/ieee754/ldbl-128/k_sincosl.c (__kernel_sincosl): Likewise. * sysdeps/ieee754/ldbl-128/k_sinl.c (__kernel_sinl): Likewise. * sysdeps/ieee754/ldbl-128/s_ceill.c (__ceill): Likewise. * sysdeps/ieee754/ldbl-128/s_copysignl.c (__copysignl): Likewise. * sysdeps/ieee754/ldbl-128/s_erfl.c (__erfcl): Likewise. * sysdeps/ieee754/ldbl-128/s_fabsl.c (__fabsl): Likewise. * sysdeps/ieee754/ldbl-128/s_finitel.c (__finitel): Likewise. * sysdeps/ieee754/ldbl-128/s_floorl.c (__floorl): Likewise. * sysdeps/ieee754/ldbl-128/s_fpclassifyl.c (__fpclassifyl): Likewise. * sysdeps/ieee754/ldbl-128/s_frexpl.c (__frexpl): Likewise. * sysdeps/ieee754/ldbl-128/s_isnanl.c (__isnanl): Likewise. * sysdeps/ieee754/ldbl-128/s_issignalingl.c (__issignalingl): Likewise. * sysdeps/ieee754/ldbl-128/s_llrintl.c (__llrintl): Likewise. * sysdeps/ieee754/ldbl-128/s_llroundl.c (__llroundl): Likewise. * sysdeps/ieee754/ldbl-128/s_lrintl.c (__lrintl): Likewise. * sysdeps/ieee754/ldbl-128/s_lroundl.c (__lroundl): Likewise. * sysdeps/ieee754/ldbl-128/s_modfl.c (__modfl): Likewise. * sysdeps/ieee754/ldbl-128/s_nearbyintl.c (__nearbyintl): Likewise. * sysdeps/ieee754/ldbl-128/s_nextafterl.c (__nextafterl): Likewise. * sysdeps/ieee754/ldbl-128/s_nexttoward.c (__nexttoward): Likewise. * sysdeps/ieee754/ldbl-128/s_nexttowardf.c (__nexttowardf): Likewise. * sysdeps/ieee754/ldbl-128/s_nextupl.c (__nextupl): Likewise. * sysdeps/ieee754/ldbl-128/s_remquol.c (__remquol): Likewise. * sysdeps/ieee754/ldbl-128/s_rintl.c (__rintl): Likewise. * sysdeps/ieee754/ldbl-128/s_roundl.c (__roundl): Likewise. * sysdeps/ieee754/ldbl-128/s_tanhl.c (__tanhl): Likewise. * sysdeps/ieee754/ldbl-128/s_truncl.c (__truncl): Likewise. * sysdeps/ieee754/ldbl-128ibm/e_fmodl.c (__ieee754_fmodl): Likewise. * sysdeps/ieee754/ldbl-128ibm/e_gammal_r.c (__ieee754_gammal_r): Likewise. * sysdeps/ieee754/ldbl-128ibm/e_powl.c (__ieee754_powl): Likewise. * sysdeps/ieee754/ldbl-128ibm/e_rem_pio2l.c (__ieee754_rem_pio2l): Likewise. * sysdeps/ieee754/ldbl-128ibm/e_remainderl.c (__ieee754_remainderl): Likewise. * sysdeps/ieee754/ldbl-128ibm/k_cosl.c (__kernel_cosl): Likewise. * sysdeps/ieee754/ldbl-128ibm/k_sinl.c (__kernel_sinl): Likewise. * sysdeps/ieee754/ldbl-128ibm/s_fabsl.c (__fabsl): Likewise. * sysdeps/ieee754/ldbl-128ibm/s_fpclassifyl.c (___fpclassifyl): Likewise. * sysdeps/ieee754/ldbl-128ibm/s_modfl.c (__modfl): Likewise. * sysdeps/ieee754/ldbl-128ibm/s_nexttowardf.c (__nexttowardf): Likewise. * sysdeps/ieee754/ldbl-128ibm/s_remquol.c (__remquol): Likewise. * sysdeps/ieee754/ldbl-96/e_acoshl.c (__ieee754_acoshl): Likewise. * sysdeps/ieee754/ldbl-96/e_asinl.c (__ieee754_asinl): Likewise. * sysdeps/ieee754/ldbl-96/e_atanhl.c (__ieee754_atanhl): Likewise. * sysdeps/ieee754/ldbl-96/e_coshl.c (__ieee754_coshl): Likewise. * sysdeps/ieee754/ldbl-96/e_gammal_r.c (__ieee754_gammal_r): Likewise. * sysdeps/ieee754/ldbl-96/e_hypotl.c (__ieee754_hypotl): Likewise. * sysdeps/ieee754/ldbl-96/e_j0l.c (__ieee754_j0l): Likewise. (__ieee754_y0l): Likewise. (pzero): Likewise. (qzero): Likewise. * sysdeps/ieee754/ldbl-96/e_j1l.c (__ieee754_j1l): Likewise. (__ieee754_y1l): Likewise. (pone): Likewise. (qone): Likewise. * sysdeps/ieee754/ldbl-96/e_jnl.c (__ieee754_jnl): Likewise. (__ieee754_ynl): Likewise. * sysdeps/ieee754/ldbl-96/e_lgammal_r.c (sin_pi): Likewise. (__ieee754_lgammal_r): Likewise. * sysdeps/ieee754/ldbl-96/e_rem_pio2l.c (__ieee754_rem_pio2l): Likewise. * sysdeps/ieee754/ldbl-96/e_sinhl.c (__ieee754_sinhl): Likewise. * sysdeps/ieee754/ldbl-96/s_copysignl.c (__copysignl): Likewise. * sysdeps/ieee754/ldbl-96/s_erfl.c (__erfl): Likewise. (__erfcl): Likewise. * sysdeps/ieee754/ldbl-96/s_frexpl.c (__frexpl): Likewise. * sysdeps/ieee754/ldbl-96/s_issignalingl.c (__issignalingl): Likewise. * sysdeps/ieee754/ldbl-96/s_llrintl.c (__llrintl): Likewise. * sysdeps/ieee754/ldbl-96/s_llroundl.c (__llroundl): Likewise. * sysdeps/ieee754/ldbl-96/s_lrintl.c (__lrintl): Likewise. * sysdeps/ieee754/ldbl-96/s_lroundl.c (__lroundl): Likewise. * sysdeps/ieee754/ldbl-96/s_modfl.c (__modfl): Likewise. * sysdeps/ieee754/ldbl-96/s_nexttoward.c (__nexttoward): Likewise. * sysdeps/ieee754/ldbl-96/s_nexttowardf.c (__nexttowardf): Likewise. * sysdeps/ieee754/ldbl-96/s_nextupl.c (__nextupl): Likewise. * sysdeps/ieee754/ldbl-96/s_remquol.c (__remquol): Likewise. * sysdeps/ieee754/ldbl-96/s_roundl.c (__roundl): Likewise. * sysdeps/ieee754/ldbl-96/s_tanhl.c (__tanhl): Likewise. * sysdeps/ieee754/ldbl-opt/s_nexttowardfd.c (__nldbl_nexttowardf): Likewise. * sysdeps/m68k/m680x0/fpu/e_pow.c (s(__ieee754_pow)): Likewise. * sysdeps/m68k/m680x0/fpu/s_fpclassifyl.c (__fpclassifyl): Likewise. * sysdeps/m68k/m680x0/fpu/s_llrint.c (__llrint): Likewise. * sysdeps/m68k/m680x0/fpu/s_llrintf.c (__llrintf): Likewise. * sysdeps/m68k/m680x0/fpu/s_llrintl.c (__llrintl): Likewise. * sysdeps/m68k/m680x0/fpu/s_nextafterl.c (__nextafterl): Likewise. * sysdeps/x86/fpu/powl_helper.c (__powl_helper): Likewise.
2017-08-03 19:55:04 +00:00
&& (uint32_t) hx < 0xfff00000 && __rint (x) == x)
Update. 1997-06-19 19:38 Ulrich Drepper <drepper@cygnus.com> * features.h: Define __STDC_IEC_559__ and _STDC_IEC_559_COMPLEX__. * elf/dl-minimal.c (__dcgettext): Remove assertion. * inet/rcmd.c: Correct a few typos. Reported by Eric Troan. * manual/Makefile (distribute): Add dir. * manual/dir: New file. * math/Makefile (libm-support): Rename s_rinttol, s_rinttoll, s_roundtol, and s_roundtoll to s_lrint, s_llrint, s_lround, and s_llround respectively. (libm-calls): Add e_gamma_r. * math/libm-test.c (check_int_exc): New function. (signbit_test): Remove test for sign of NaN. (gamma_test): Clear exception after test of existence. Correct tests and and epsilons. (lgamma_test): Likewise. (ilogb_test): Correct all tests. (scalb_test): Rewrite. (rinttol_test): Rename to lrint_test and correct tests. (rinttoll_test): Likewise. (roundtol_test): Likewise. (roundtoll_test): Likewise. (main): Call lrint/lround functions instead of rinttol/roundtol. * math/math.h: Change prototypes for rinttol/roundtol. * math/mathcalls.h: Rearrange prototypes according to ISO C9X draft. * sysdeps/generic/mathbits.h: Define FP_ILOGB0 and FP_ILOGBNAN. * sysdeps/i386/mathbits.h: Likewise. * sysdeps/libm-i387/e_scalb.S: Handle special cases correctly. * sysdeps/libm-i387/e_scalbf.S: Likewise. * sysdeps/libm-i387/e_scalbl.S: Likewise. * sysdeps/libm-i387/s_asinh.S: Handle -inf correctly. * sysdeps/libm-i387/s_asinhf.S: Likewise. * sysdeps/libm-i387/s_asinhl.S: Likewise. * sysdeps/libm-i387/s_ilogb.S: Optimize. * sysdeps/libm-i387/s_ilogbf.S: Likewise. * sysdeps/libm-i387/s_ilogbl.S: Likewise. * sysdeps/libm-i387/s_rinttol.S: Rename to... * sysdeps/libm-i387/s_lrint.S: ...this. * sysdeps/libm-i387/s_rinttoll.S: Rename to... * sysdeps/libm-i387/s_llrint.S: ...this. * sysdeps/libm-i387/s_remquo.S: Correctly set sign of remainder. * sysdeps/libm-i387/s_remquof.S: Likewise. * sysdeps/libm-i387/s_remquol.S: Likewise. * sysdeps/libm-i387/e_gamma_r.c: New file. Implementation of gamma function according to ISO C. * sysdeps/libm-i387/e_gammaf_r.c: New file. * sysdeps/libm-i387/e_gammal_r.c: New file. * sysdeps/libm-i387/e_lgamma_r.c: Don't let optimize compile the generation of exceptions away. * sysdeps/libm-i387/e_lgammaf_r.c: Likewise. * sysdeps/libm-i387/k_standard.c: Correct return value for infinity points of gamma function when not SVID mode. * sysdeps/libm-i387/s_rinttoll.c: Renamed to... * sysdeps/libm-i387/s_llrint.c: ...this. * sysdeps/libm-i387/s_rinttol.c: Renamed to... * sysdeps/libm-i387/s_lrint.c: ...this. * sysdeps/libm-i387/s_roundtoll.c: Renamed to... * sysdeps/libm-i387/s_llround.c: ...this. * sysdeps/libm-i387/s_roundtol.c: Renamed to... * sysdeps/libm-i387/s_lround.c: ..this. * sysdeps/libm-i387/s_scalbn.c: Change second parameter according to ISO C. * sysdeps/libm-i387/s_scalbnf.c: Likewise. * sysdeps/libm-i387/s_scalbnl.c: Likewise. * sysdeps/libm-i387/w_gamma.c: Call __ieee754_gamma_r if library mode is not _SVID_. * sysdeps/libm-i387/w_gammaf.c: Likewise. * sysdeps/libm-i387/w_gammal.c: Likewise. * sysdeps/m68k/fpu/__math.h: Rename __rinttol to __lrint and rinttol to lrint. * sysdeps/m68k/fpu/s_rinttol.c: Renamed to... * sysdeps/m68k/fpu/s_lrint.c: ...this. * sysdeps/m68k/fpu/s_rinttoll.c: Renamed to... * sysdeps/m68k/fpu/s_llrint.c: ...this. * md5-crypt/Makefile: Link md5test program with md5.o. * stdio-common/temptest.c: Don't use __stdio_gen_tempname which is not exported by the libc.so. * stdio-common/vfscanf.c: Correct scanning of strings after last change. * sysdeps/unix/sysv/linux/i386/sysdep.S: Use .comm to define errno. 1997-06-19 07:37 H.J. Lu <hjl@gnu.ai.mit.edu> * time/tzfile.c (__tzfile_read): Store getc () return in int. 1997-06-13 Miles Bader <miles@gnu.ai.mit.edu> * argp/argp-parse.c (argp_version_parser): Include `(PROGRAM ERROR)' in the no-version error text to indicate that something's fucked. [!_] (N_): New macro. (argp_default_options, argp_version_options): Wrap doc strings in N_(). 1997-06-12 Miles Bader <miles@gnu.ai.mit.edu> * argp/argp-parse.c (parser_parse_opt): Detect and report unhandled options here. (parser_parse_arg): Handle ARGP_KEY_ARGS here. Adjust NEXT pointer back if we fail to parse anything. (parser_parse_next): Simplify arg code. Leave state NEXT frobbing to parser_parse_arg. 1997-06-11 Miles Bader <miles@gnu.ai.mit.edu> * argp/argp.h (ARGP_KEY_ARGS, ARGP_KEY_FINI): New macros. * argp/argp-parse.c (parser_finalize): Do another pass over the parsers with ARGP_KEY_FINI. 1997-06-18 Miles Bader <miles@gnu.ai.mit.edu> * string/Makefile (routines): Add argz-replace. 1997-06-16 00:16 Miles Bader <miles@gnu.ai.mit.edu> * manual/string.texi (Argz Functions): Document argz_replace. 1997-06-12 Miles Bader <miles@gnu.ai.mit.edu> * string/argz.h (__argz_replace, argz_replace): New declarations. * string/argz-replace.c: New file. 1997-06-11 Miles Bader <miles@gnu.ai.mit.edu> * manual/argp.texi (Argp Special Keys): Document ARGP_KEY_ARGS. 1997-06-16 23:08 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * Makeconfig (libc-map): Remove definition. * Makerules (libc-map): Define it here, using the full name. (load-map-file): Remove case for empty $(..). ($(common-objpfx)libc.so): Revert last change. 1997-06-17 22:18 Mark Kettenis <kettenis@phys.uva.nl> * login/programs/utmpd.c (main): Improve signal handling. * login/programs/request.c (do_setutent, do_updwtmp): Allow arbitrary length filenames. * login/programs/utmpd.h (setutent_request, updwtmp_request): Get rid of fixed length file field. * login/utmp_daemon.c (do_setutent, do_updwtmp): Allow arbitrary length filenames. * login/programs/request.c (do_pututline): Don't fail if connection->position is -1 on entry. 1997-06-15 16:32 Mark Kettenis <kettenis@phys.uva.nl> * login/utmp_file.c (updwtmp_file): Use the same method for appending an entry as in pututline_file. 1997-06-11 18:59 Mark Kettenis <kettenis@phys.uva.nl> * login/utmp_file.c (getutent_r_file): Use read lock instead of write lock. (getutline_r_file, internal_getut_r): Lock utmp file. (updwtmp_file): Use fcntl to lock file instead of flock. 1997-06-18 00:11 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * stdio-common/vfscanf.c (inchar, ungetc): Don't count EOF as character read in/put back. * stdio-common/tstscanf.c: Add test case for this. 1997-06-17 22:17 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * libc.map: Add more symbols. 1997-06-18 12:01 Ulrich Drepper <drepper@cygnus.com> * manual/Makefile (install): Make sure `dir' file exists if we use install-info. 1997-06-17 19:32 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * manual/Makefile (info): Depend on dir-add.info. 1997-06-17 17:19 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * elf/ldd.bash.in: Fix spacing in message. 1997-06-17 14:28 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * elf/do-rel.h (elf_dynamic_do_rel): Always use version information if available. 1997-06-17 11:34 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sunrpc/Makefile ($(rpcsvc:%.x=$(objpfx)rpcsvc/%.h)): Make command non-empty to force make to recheck modification time. ($(rpcsvc:%.x=$(objpfx)x%.c)): Likewise. 1997-06-17 00:26 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/stub/e_acoshl.c: Set errno to ENOSYS. * sysdeps/stub/e_acosl.c: Likewise. * sysdeps/stub/e_asinl.c: Likewise. * sysdeps/stub/e_atan2l.c: Likewise. * sysdeps/stub/e_expl.c: Likewise. * sysdeps/stub/e_fmodl.c: Likewise. * sysdeps/stub/e_j0l.c: Likewise. * sysdeps/stub/e_j1l.c: Likewise. * sysdeps/stub/e_jnl.c: Likewise. * sysdeps/stub/e_lgammal_r.c: Likewise. * sysdeps/stub/e_log10l.c: Likewise. * sysdeps/stub/e_logl.c: Likewise. * sysdeps/stub/e_powl.c: Likewise. * sysdeps/stub/e_rem_pio2l.c: Likewise. * sysdeps/stub/e_sqrtl.c: Likewise. * sysdeps/stub/k_cosl.c: Likewise. * sysdeps/stub/k_rem_pio2l.c: Likewise. * sysdeps/stub/k_sinl.c: Likewise. * sysdeps/stub/k_tanl.c: Likewise. * sysdeps/stub/s_atanl.c: Likewise. * sysdeps/stub/s_erfl.c: Likewise. * sysdeps/stub/s_exp2.c: Likewise. * sysdeps/stub/s_exp2f.c: Likewise. * sysdeps/stub/s_exp2l.c: Likewise. * sysdeps/stub/s_expm1l.c: Likewise. * sysdeps/stub/s_log1pl.c: Likewise. * sysdeps/stub/s_log2l.c: Likewise. 1997-06-18 11:46 Ulrich Drepper <drepper@cygnus.com> * sysdeps/libm-ieee754/w_gamma.c: If _LIB_VERSION is _SVID_ compute result as before last change. * sysdeps/libm-ieee754/w_gammaf.c: Likewise. * sysdeps/libm-ieee754/w_gammal.c: Likewise. 1997-06-16 23:37 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/libm-ieee754/s_remquo.c: Fix off-by-one when computing quotient. * sysdeps/libm-ieee754/s_remquof.c: Likewise. * sysdeps/libm-ieee754/s_remquol.c: Likewise. * sysdeps/m68k/fpu/s_remquo.c: Remove FIXME and special case for quotient. * sysdeps/libm-ieee754/w_gamma.c: Add missing call to exp function. Don't use global signgam. * sysdeps/libm-ieee754/w_gammaf.c: Likewise. * sysdeps/libm-ieee754/w_gammal.c: Likewise. * math/Makefile (libm-calls): Remove w_gamma_r. * sysdeps/libm-ieee754/w_gamma_r.c: Remove file. * sysdeps/libm-ieee754/w_gammaf_r.c: Remove file. * sysdeps/libm-ieee754/w_gammal_r.c: Remove file. * math/libm-test.c (atanh_test): Declare x only if needed. (signbit_test): Fix typo. (gamma_test): Check whether function is implemented. Add epsilons. (lgamma_test): Likewise. (fmod_test): Add epsilons. (exp2_test): Use right function for existence test. 1997-06-07 09:20 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/powerpc/Dist: Add fenv_const.c. * sysdeps/unix/sysv/linux/Dist: Add net/if_slip.h. * sysdeps/unix/sysv/linux/powerpc/Dist: Add init-first.h and syscall.h. * sysdeps/unix/sysv/linux/sparc/Dist: Add init-first.h. * string/Makefile (distribute): Add tst-svc.expect. * nis/Makefile (distribute): Add nis_intern.h and Banner. * elf/Makefile (distribute): Add dl-hash.h. * Rules (subdir_echo-distinfo): Add headers from $(distribute). * login/Makefile (others): Add utmpdump. * sysdeps/unix/sysv/linux/Makefile (sysdep_headers): Add net/if_slip.h. * manual/Makefile (dir-add.texi): Also look in indirectly included files. 1997-06-16 23:15 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/unix/sysv/linux/configure.in: Remove commands to generate stdio_lim.h. * sysdeps/unix/sysv/linux/mk-stdiolim.c: Remove. * sysdeps/unix/sysv/linux/Makefile: Add rules to generate stdio_lim.h here. (common-generated): Add generated files. (inhibit-stdio_lim): Define. * sysdeps/posix/Makefile [$(inhibit-stdio_lim)=yes]: Disable rules to generate stdio_lim.h. * sysdeps/unix/sysv/linux/Makefile: Suppress inclusion of dependecy files if no_deps is set. ($(objpfx)syscall-%.d): Add header file as target to dependency generation. 1997-06-14 19:19 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/unix/sysv/linux/arm/Dist: New file. 1997-06-14 17:59 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * extra-lib.mk (others): Depend on versioned shared library, not the unversioned one. * Makerules (build-shlib): Don't make the version link here. ($(common-objpfx)libc.so$(libc.so-version)): New rule for libc version link. 1997-06-16 03:07 Ulrich Drepper <drepper@cygnus.com> * sysdeps/libm-ieee754/k_standard.c: Undo change of Tue Aug 6 01:13:56 1996. * argp/argp-help.c (argp_args_usage): Supply correct argp to filter_doc. * argp/argp-help.c (hol_add_cluster): Initialize CL->depth. * argp/argp-help.c (_help): Supply STATE to argp_args_usage. * argp/argp.h (ARGP_KEY_HELP_ARGS_DOC): New macro. * argp/argp-fmtstream.c: Add casts to prevent warnings. * argp/argp.h (OPTION_NO_USAGE): New macro. * argp/argp-help.c (usage_long_opt, usage_argful_short_opt, * argp/argp-fmtstream.c (__argp_fmtstream_update): Account for case * argp/argp-help.c <stddef.h>: New include. * argp/argp.h (argp_state_help, __argp_state_help, argp_usage, * argp/argp.h (argp_program_bug_address): Make const. * argp/argp-parse.c (argp_default_parser): Set STATE->name for OPT_PROGNAME. * argp/argp-help.c (__argp_error, __argp_failure, __argp_state_help): * argp/argp-parse.c (parser_init): Set PARSER->state.flags.
1997-06-19 19:11:22 +00:00
{
/* Return value for integer x < 0 is NaN with invalid exception. */
*signgamp = 0;
Update. 1997-06-19 19:38 Ulrich Drepper <drepper@cygnus.com> * features.h: Define __STDC_IEC_559__ and _STDC_IEC_559_COMPLEX__. * elf/dl-minimal.c (__dcgettext): Remove assertion. * inet/rcmd.c: Correct a few typos. Reported by Eric Troan. * manual/Makefile (distribute): Add dir. * manual/dir: New file. * math/Makefile (libm-support): Rename s_rinttol, s_rinttoll, s_roundtol, and s_roundtoll to s_lrint, s_llrint, s_lround, and s_llround respectively. (libm-calls): Add e_gamma_r. * math/libm-test.c (check_int_exc): New function. (signbit_test): Remove test for sign of NaN. (gamma_test): Clear exception after test of existence. Correct tests and and epsilons. (lgamma_test): Likewise. (ilogb_test): Correct all tests. (scalb_test): Rewrite. (rinttol_test): Rename to lrint_test and correct tests. (rinttoll_test): Likewise. (roundtol_test): Likewise. (roundtoll_test): Likewise. (main): Call lrint/lround functions instead of rinttol/roundtol. * math/math.h: Change prototypes for rinttol/roundtol. * math/mathcalls.h: Rearrange prototypes according to ISO C9X draft. * sysdeps/generic/mathbits.h: Define FP_ILOGB0 and FP_ILOGBNAN. * sysdeps/i386/mathbits.h: Likewise. * sysdeps/libm-i387/e_scalb.S: Handle special cases correctly. * sysdeps/libm-i387/e_scalbf.S: Likewise. * sysdeps/libm-i387/e_scalbl.S: Likewise. * sysdeps/libm-i387/s_asinh.S: Handle -inf correctly. * sysdeps/libm-i387/s_asinhf.S: Likewise. * sysdeps/libm-i387/s_asinhl.S: Likewise. * sysdeps/libm-i387/s_ilogb.S: Optimize. * sysdeps/libm-i387/s_ilogbf.S: Likewise. * sysdeps/libm-i387/s_ilogbl.S: Likewise. * sysdeps/libm-i387/s_rinttol.S: Rename to... * sysdeps/libm-i387/s_lrint.S: ...this. * sysdeps/libm-i387/s_rinttoll.S: Rename to... * sysdeps/libm-i387/s_llrint.S: ...this. * sysdeps/libm-i387/s_remquo.S: Correctly set sign of remainder. * sysdeps/libm-i387/s_remquof.S: Likewise. * sysdeps/libm-i387/s_remquol.S: Likewise. * sysdeps/libm-i387/e_gamma_r.c: New file. Implementation of gamma function according to ISO C. * sysdeps/libm-i387/e_gammaf_r.c: New file. * sysdeps/libm-i387/e_gammal_r.c: New file. * sysdeps/libm-i387/e_lgamma_r.c: Don't let optimize compile the generation of exceptions away. * sysdeps/libm-i387/e_lgammaf_r.c: Likewise. * sysdeps/libm-i387/k_standard.c: Correct return value for infinity points of gamma function when not SVID mode. * sysdeps/libm-i387/s_rinttoll.c: Renamed to... * sysdeps/libm-i387/s_llrint.c: ...this. * sysdeps/libm-i387/s_rinttol.c: Renamed to... * sysdeps/libm-i387/s_lrint.c: ...this. * sysdeps/libm-i387/s_roundtoll.c: Renamed to... * sysdeps/libm-i387/s_llround.c: ...this. * sysdeps/libm-i387/s_roundtol.c: Renamed to... * sysdeps/libm-i387/s_lround.c: ..this. * sysdeps/libm-i387/s_scalbn.c: Change second parameter according to ISO C. * sysdeps/libm-i387/s_scalbnf.c: Likewise. * sysdeps/libm-i387/s_scalbnl.c: Likewise. * sysdeps/libm-i387/w_gamma.c: Call __ieee754_gamma_r if library mode is not _SVID_. * sysdeps/libm-i387/w_gammaf.c: Likewise. * sysdeps/libm-i387/w_gammal.c: Likewise. * sysdeps/m68k/fpu/__math.h: Rename __rinttol to __lrint and rinttol to lrint. * sysdeps/m68k/fpu/s_rinttol.c: Renamed to... * sysdeps/m68k/fpu/s_lrint.c: ...this. * sysdeps/m68k/fpu/s_rinttoll.c: Renamed to... * sysdeps/m68k/fpu/s_llrint.c: ...this. * md5-crypt/Makefile: Link md5test program with md5.o. * stdio-common/temptest.c: Don't use __stdio_gen_tempname which is not exported by the libc.so. * stdio-common/vfscanf.c: Correct scanning of strings after last change. * sysdeps/unix/sysv/linux/i386/sysdep.S: Use .comm to define errno. 1997-06-19 07:37 H.J. Lu <hjl@gnu.ai.mit.edu> * time/tzfile.c (__tzfile_read): Store getc () return in int. 1997-06-13 Miles Bader <miles@gnu.ai.mit.edu> * argp/argp-parse.c (argp_version_parser): Include `(PROGRAM ERROR)' in the no-version error text to indicate that something's fucked. [!_] (N_): New macro. (argp_default_options, argp_version_options): Wrap doc strings in N_(). 1997-06-12 Miles Bader <miles@gnu.ai.mit.edu> * argp/argp-parse.c (parser_parse_opt): Detect and report unhandled options here. (parser_parse_arg): Handle ARGP_KEY_ARGS here. Adjust NEXT pointer back if we fail to parse anything. (parser_parse_next): Simplify arg code. Leave state NEXT frobbing to parser_parse_arg. 1997-06-11 Miles Bader <miles@gnu.ai.mit.edu> * argp/argp.h (ARGP_KEY_ARGS, ARGP_KEY_FINI): New macros. * argp/argp-parse.c (parser_finalize): Do another pass over the parsers with ARGP_KEY_FINI. 1997-06-18 Miles Bader <miles@gnu.ai.mit.edu> * string/Makefile (routines): Add argz-replace. 1997-06-16 00:16 Miles Bader <miles@gnu.ai.mit.edu> * manual/string.texi (Argz Functions): Document argz_replace. 1997-06-12 Miles Bader <miles@gnu.ai.mit.edu> * string/argz.h (__argz_replace, argz_replace): New declarations. * string/argz-replace.c: New file. 1997-06-11 Miles Bader <miles@gnu.ai.mit.edu> * manual/argp.texi (Argp Special Keys): Document ARGP_KEY_ARGS. 1997-06-16 23:08 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * Makeconfig (libc-map): Remove definition. * Makerules (libc-map): Define it here, using the full name. (load-map-file): Remove case for empty $(..). ($(common-objpfx)libc.so): Revert last change. 1997-06-17 22:18 Mark Kettenis <kettenis@phys.uva.nl> * login/programs/utmpd.c (main): Improve signal handling. * login/programs/request.c (do_setutent, do_updwtmp): Allow arbitrary length filenames. * login/programs/utmpd.h (setutent_request, updwtmp_request): Get rid of fixed length file field. * login/utmp_daemon.c (do_setutent, do_updwtmp): Allow arbitrary length filenames. * login/programs/request.c (do_pututline): Don't fail if connection->position is -1 on entry. 1997-06-15 16:32 Mark Kettenis <kettenis@phys.uva.nl> * login/utmp_file.c (updwtmp_file): Use the same method for appending an entry as in pututline_file. 1997-06-11 18:59 Mark Kettenis <kettenis@phys.uva.nl> * login/utmp_file.c (getutent_r_file): Use read lock instead of write lock. (getutline_r_file, internal_getut_r): Lock utmp file. (updwtmp_file): Use fcntl to lock file instead of flock. 1997-06-18 00:11 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * stdio-common/vfscanf.c (inchar, ungetc): Don't count EOF as character read in/put back. * stdio-common/tstscanf.c: Add test case for this. 1997-06-17 22:17 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * libc.map: Add more symbols. 1997-06-18 12:01 Ulrich Drepper <drepper@cygnus.com> * manual/Makefile (install): Make sure `dir' file exists if we use install-info. 1997-06-17 19:32 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * manual/Makefile (info): Depend on dir-add.info. 1997-06-17 17:19 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * elf/ldd.bash.in: Fix spacing in message. 1997-06-17 14:28 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * elf/do-rel.h (elf_dynamic_do_rel): Always use version information if available. 1997-06-17 11:34 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sunrpc/Makefile ($(rpcsvc:%.x=$(objpfx)rpcsvc/%.h)): Make command non-empty to force make to recheck modification time. ($(rpcsvc:%.x=$(objpfx)x%.c)): Likewise. 1997-06-17 00:26 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/stub/e_acoshl.c: Set errno to ENOSYS. * sysdeps/stub/e_acosl.c: Likewise. * sysdeps/stub/e_asinl.c: Likewise. * sysdeps/stub/e_atan2l.c: Likewise. * sysdeps/stub/e_expl.c: Likewise. * sysdeps/stub/e_fmodl.c: Likewise. * sysdeps/stub/e_j0l.c: Likewise. * sysdeps/stub/e_j1l.c: Likewise. * sysdeps/stub/e_jnl.c: Likewise. * sysdeps/stub/e_lgammal_r.c: Likewise. * sysdeps/stub/e_log10l.c: Likewise. * sysdeps/stub/e_logl.c: Likewise. * sysdeps/stub/e_powl.c: Likewise. * sysdeps/stub/e_rem_pio2l.c: Likewise. * sysdeps/stub/e_sqrtl.c: Likewise. * sysdeps/stub/k_cosl.c: Likewise. * sysdeps/stub/k_rem_pio2l.c: Likewise. * sysdeps/stub/k_sinl.c: Likewise. * sysdeps/stub/k_tanl.c: Likewise. * sysdeps/stub/s_atanl.c: Likewise. * sysdeps/stub/s_erfl.c: Likewise. * sysdeps/stub/s_exp2.c: Likewise. * sysdeps/stub/s_exp2f.c: Likewise. * sysdeps/stub/s_exp2l.c: Likewise. * sysdeps/stub/s_expm1l.c: Likewise. * sysdeps/stub/s_log1pl.c: Likewise. * sysdeps/stub/s_log2l.c: Likewise. 1997-06-18 11:46 Ulrich Drepper <drepper@cygnus.com> * sysdeps/libm-ieee754/w_gamma.c: If _LIB_VERSION is _SVID_ compute result as before last change. * sysdeps/libm-ieee754/w_gammaf.c: Likewise. * sysdeps/libm-ieee754/w_gammal.c: Likewise. 1997-06-16 23:37 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/libm-ieee754/s_remquo.c: Fix off-by-one when computing quotient. * sysdeps/libm-ieee754/s_remquof.c: Likewise. * sysdeps/libm-ieee754/s_remquol.c: Likewise. * sysdeps/m68k/fpu/s_remquo.c: Remove FIXME and special case for quotient. * sysdeps/libm-ieee754/w_gamma.c: Add missing call to exp function. Don't use global signgam. * sysdeps/libm-ieee754/w_gammaf.c: Likewise. * sysdeps/libm-ieee754/w_gammal.c: Likewise. * math/Makefile (libm-calls): Remove w_gamma_r. * sysdeps/libm-ieee754/w_gamma_r.c: Remove file. * sysdeps/libm-ieee754/w_gammaf_r.c: Remove file. * sysdeps/libm-ieee754/w_gammal_r.c: Remove file. * math/libm-test.c (atanh_test): Declare x only if needed. (signbit_test): Fix typo. (gamma_test): Check whether function is implemented. Add epsilons. (lgamma_test): Likewise. (fmod_test): Add epsilons. (exp2_test): Use right function for existence test. 1997-06-07 09:20 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/powerpc/Dist: Add fenv_const.c. * sysdeps/unix/sysv/linux/Dist: Add net/if_slip.h. * sysdeps/unix/sysv/linux/powerpc/Dist: Add init-first.h and syscall.h. * sysdeps/unix/sysv/linux/sparc/Dist: Add init-first.h. * string/Makefile (distribute): Add tst-svc.expect. * nis/Makefile (distribute): Add nis_intern.h and Banner. * elf/Makefile (distribute): Add dl-hash.h. * Rules (subdir_echo-distinfo): Add headers from $(distribute). * login/Makefile (others): Add utmpdump. * sysdeps/unix/sysv/linux/Makefile (sysdep_headers): Add net/if_slip.h. * manual/Makefile (dir-add.texi): Also look in indirectly included files. 1997-06-16 23:15 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/unix/sysv/linux/configure.in: Remove commands to generate stdio_lim.h. * sysdeps/unix/sysv/linux/mk-stdiolim.c: Remove. * sysdeps/unix/sysv/linux/Makefile: Add rules to generate stdio_lim.h here. (common-generated): Add generated files. (inhibit-stdio_lim): Define. * sysdeps/posix/Makefile [$(inhibit-stdio_lim)=yes]: Disable rules to generate stdio_lim.h. * sysdeps/unix/sysv/linux/Makefile: Suppress inclusion of dependecy files if no_deps is set. ($(objpfx)syscall-%.d): Add header file as target to dependency generation. 1997-06-14 19:19 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/unix/sysv/linux/arm/Dist: New file. 1997-06-14 17:59 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * extra-lib.mk (others): Depend on versioned shared library, not the unversioned one. * Makerules (build-shlib): Don't make the version link here. ($(common-objpfx)libc.so$(libc.so-version)): New rule for libc version link. 1997-06-16 03:07 Ulrich Drepper <drepper@cygnus.com> * sysdeps/libm-ieee754/k_standard.c: Undo change of Tue Aug 6 01:13:56 1996. * argp/argp-help.c (argp_args_usage): Supply correct argp to filter_doc. * argp/argp-help.c (hol_add_cluster): Initialize CL->depth. * argp/argp-help.c (_help): Supply STATE to argp_args_usage. * argp/argp.h (ARGP_KEY_HELP_ARGS_DOC): New macro. * argp/argp-fmtstream.c: Add casts to prevent warnings. * argp/argp.h (OPTION_NO_USAGE): New macro. * argp/argp-help.c (usage_long_opt, usage_argful_short_opt, * argp/argp-fmtstream.c (__argp_fmtstream_update): Account for case * argp/argp-help.c <stddef.h>: New include. * argp/argp.h (argp_state_help, __argp_state_help, argp_usage, * argp/argp.h (argp_program_bug_address): Make const. * argp/argp-parse.c (argp_default_parser): Set STATE->name for OPT_PROGNAME. * argp/argp-help.c (__argp_error, __argp_failure, __argp_state_help): * argp/argp-parse.c (parser_init): Set PARSER->state.flags.
1997-06-19 19:11:22 +00:00
return (x - x) / (x - x);
}
if (__glibc_unlikely ((unsigned int) hx == 0xfff00000 && lx == 0))
{
/* x == -Inf. According to ISO this is NaN. */
*signgamp = 0;
return x - x;
}
if (__glibc_unlikely ((hx & 0x7ff00000) == 0x7ff00000))
{
/* Positive infinity (return positive infinity) or NaN (return
NaN). */
*signgamp = 0;
return x + x;
}
Update. 1997-06-19 19:38 Ulrich Drepper <drepper@cygnus.com> * features.h: Define __STDC_IEC_559__ and _STDC_IEC_559_COMPLEX__. * elf/dl-minimal.c (__dcgettext): Remove assertion. * inet/rcmd.c: Correct a few typos. Reported by Eric Troan. * manual/Makefile (distribute): Add dir. * manual/dir: New file. * math/Makefile (libm-support): Rename s_rinttol, s_rinttoll, s_roundtol, and s_roundtoll to s_lrint, s_llrint, s_lround, and s_llround respectively. (libm-calls): Add e_gamma_r. * math/libm-test.c (check_int_exc): New function. (signbit_test): Remove test for sign of NaN. (gamma_test): Clear exception after test of existence. Correct tests and and epsilons. (lgamma_test): Likewise. (ilogb_test): Correct all tests. (scalb_test): Rewrite. (rinttol_test): Rename to lrint_test and correct tests. (rinttoll_test): Likewise. (roundtol_test): Likewise. (roundtoll_test): Likewise. (main): Call lrint/lround functions instead of rinttol/roundtol. * math/math.h: Change prototypes for rinttol/roundtol. * math/mathcalls.h: Rearrange prototypes according to ISO C9X draft. * sysdeps/generic/mathbits.h: Define FP_ILOGB0 and FP_ILOGBNAN. * sysdeps/i386/mathbits.h: Likewise. * sysdeps/libm-i387/e_scalb.S: Handle special cases correctly. * sysdeps/libm-i387/e_scalbf.S: Likewise. * sysdeps/libm-i387/e_scalbl.S: Likewise. * sysdeps/libm-i387/s_asinh.S: Handle -inf correctly. * sysdeps/libm-i387/s_asinhf.S: Likewise. * sysdeps/libm-i387/s_asinhl.S: Likewise. * sysdeps/libm-i387/s_ilogb.S: Optimize. * sysdeps/libm-i387/s_ilogbf.S: Likewise. * sysdeps/libm-i387/s_ilogbl.S: Likewise. * sysdeps/libm-i387/s_rinttol.S: Rename to... * sysdeps/libm-i387/s_lrint.S: ...this. * sysdeps/libm-i387/s_rinttoll.S: Rename to... * sysdeps/libm-i387/s_llrint.S: ...this. * sysdeps/libm-i387/s_remquo.S: Correctly set sign of remainder. * sysdeps/libm-i387/s_remquof.S: Likewise. * sysdeps/libm-i387/s_remquol.S: Likewise. * sysdeps/libm-i387/e_gamma_r.c: New file. Implementation of gamma function according to ISO C. * sysdeps/libm-i387/e_gammaf_r.c: New file. * sysdeps/libm-i387/e_gammal_r.c: New file. * sysdeps/libm-i387/e_lgamma_r.c: Don't let optimize compile the generation of exceptions away. * sysdeps/libm-i387/e_lgammaf_r.c: Likewise. * sysdeps/libm-i387/k_standard.c: Correct return value for infinity points of gamma function when not SVID mode. * sysdeps/libm-i387/s_rinttoll.c: Renamed to... * sysdeps/libm-i387/s_llrint.c: ...this. * sysdeps/libm-i387/s_rinttol.c: Renamed to... * sysdeps/libm-i387/s_lrint.c: ...this. * sysdeps/libm-i387/s_roundtoll.c: Renamed to... * sysdeps/libm-i387/s_llround.c: ...this. * sysdeps/libm-i387/s_roundtol.c: Renamed to... * sysdeps/libm-i387/s_lround.c: ..this. * sysdeps/libm-i387/s_scalbn.c: Change second parameter according to ISO C. * sysdeps/libm-i387/s_scalbnf.c: Likewise. * sysdeps/libm-i387/s_scalbnl.c: Likewise. * sysdeps/libm-i387/w_gamma.c: Call __ieee754_gamma_r if library mode is not _SVID_. * sysdeps/libm-i387/w_gammaf.c: Likewise. * sysdeps/libm-i387/w_gammal.c: Likewise. * sysdeps/m68k/fpu/__math.h: Rename __rinttol to __lrint and rinttol to lrint. * sysdeps/m68k/fpu/s_rinttol.c: Renamed to... * sysdeps/m68k/fpu/s_lrint.c: ...this. * sysdeps/m68k/fpu/s_rinttoll.c: Renamed to... * sysdeps/m68k/fpu/s_llrint.c: ...this. * md5-crypt/Makefile: Link md5test program with md5.o. * stdio-common/temptest.c: Don't use __stdio_gen_tempname which is not exported by the libc.so. * stdio-common/vfscanf.c: Correct scanning of strings after last change. * sysdeps/unix/sysv/linux/i386/sysdep.S: Use .comm to define errno. 1997-06-19 07:37 H.J. Lu <hjl@gnu.ai.mit.edu> * time/tzfile.c (__tzfile_read): Store getc () return in int. 1997-06-13 Miles Bader <miles@gnu.ai.mit.edu> * argp/argp-parse.c (argp_version_parser): Include `(PROGRAM ERROR)' in the no-version error text to indicate that something's fucked. [!_] (N_): New macro. (argp_default_options, argp_version_options): Wrap doc strings in N_(). 1997-06-12 Miles Bader <miles@gnu.ai.mit.edu> * argp/argp-parse.c (parser_parse_opt): Detect and report unhandled options here. (parser_parse_arg): Handle ARGP_KEY_ARGS here. Adjust NEXT pointer back if we fail to parse anything. (parser_parse_next): Simplify arg code. Leave state NEXT frobbing to parser_parse_arg. 1997-06-11 Miles Bader <miles@gnu.ai.mit.edu> * argp/argp.h (ARGP_KEY_ARGS, ARGP_KEY_FINI): New macros. * argp/argp-parse.c (parser_finalize): Do another pass over the parsers with ARGP_KEY_FINI. 1997-06-18 Miles Bader <miles@gnu.ai.mit.edu> * string/Makefile (routines): Add argz-replace. 1997-06-16 00:16 Miles Bader <miles@gnu.ai.mit.edu> * manual/string.texi (Argz Functions): Document argz_replace. 1997-06-12 Miles Bader <miles@gnu.ai.mit.edu> * string/argz.h (__argz_replace, argz_replace): New declarations. * string/argz-replace.c: New file. 1997-06-11 Miles Bader <miles@gnu.ai.mit.edu> * manual/argp.texi (Argp Special Keys): Document ARGP_KEY_ARGS. 1997-06-16 23:08 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * Makeconfig (libc-map): Remove definition. * Makerules (libc-map): Define it here, using the full name. (load-map-file): Remove case for empty $(..). ($(common-objpfx)libc.so): Revert last change. 1997-06-17 22:18 Mark Kettenis <kettenis@phys.uva.nl> * login/programs/utmpd.c (main): Improve signal handling. * login/programs/request.c (do_setutent, do_updwtmp): Allow arbitrary length filenames. * login/programs/utmpd.h (setutent_request, updwtmp_request): Get rid of fixed length file field. * login/utmp_daemon.c (do_setutent, do_updwtmp): Allow arbitrary length filenames. * login/programs/request.c (do_pututline): Don't fail if connection->position is -1 on entry. 1997-06-15 16:32 Mark Kettenis <kettenis@phys.uva.nl> * login/utmp_file.c (updwtmp_file): Use the same method for appending an entry as in pututline_file. 1997-06-11 18:59 Mark Kettenis <kettenis@phys.uva.nl> * login/utmp_file.c (getutent_r_file): Use read lock instead of write lock. (getutline_r_file, internal_getut_r): Lock utmp file. (updwtmp_file): Use fcntl to lock file instead of flock. 1997-06-18 00:11 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * stdio-common/vfscanf.c (inchar, ungetc): Don't count EOF as character read in/put back. * stdio-common/tstscanf.c: Add test case for this. 1997-06-17 22:17 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * libc.map: Add more symbols. 1997-06-18 12:01 Ulrich Drepper <drepper@cygnus.com> * manual/Makefile (install): Make sure `dir' file exists if we use install-info. 1997-06-17 19:32 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * manual/Makefile (info): Depend on dir-add.info. 1997-06-17 17:19 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * elf/ldd.bash.in: Fix spacing in message. 1997-06-17 14:28 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * elf/do-rel.h (elf_dynamic_do_rel): Always use version information if available. 1997-06-17 11:34 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sunrpc/Makefile ($(rpcsvc:%.x=$(objpfx)rpcsvc/%.h)): Make command non-empty to force make to recheck modification time. ($(rpcsvc:%.x=$(objpfx)x%.c)): Likewise. 1997-06-17 00:26 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/stub/e_acoshl.c: Set errno to ENOSYS. * sysdeps/stub/e_acosl.c: Likewise. * sysdeps/stub/e_asinl.c: Likewise. * sysdeps/stub/e_atan2l.c: Likewise. * sysdeps/stub/e_expl.c: Likewise. * sysdeps/stub/e_fmodl.c: Likewise. * sysdeps/stub/e_j0l.c: Likewise. * sysdeps/stub/e_j1l.c: Likewise. * sysdeps/stub/e_jnl.c: Likewise. * sysdeps/stub/e_lgammal_r.c: Likewise. * sysdeps/stub/e_log10l.c: Likewise. * sysdeps/stub/e_logl.c: Likewise. * sysdeps/stub/e_powl.c: Likewise. * sysdeps/stub/e_rem_pio2l.c: Likewise. * sysdeps/stub/e_sqrtl.c: Likewise. * sysdeps/stub/k_cosl.c: Likewise. * sysdeps/stub/k_rem_pio2l.c: Likewise. * sysdeps/stub/k_sinl.c: Likewise. * sysdeps/stub/k_tanl.c: Likewise. * sysdeps/stub/s_atanl.c: Likewise. * sysdeps/stub/s_erfl.c: Likewise. * sysdeps/stub/s_exp2.c: Likewise. * sysdeps/stub/s_exp2f.c: Likewise. * sysdeps/stub/s_exp2l.c: Likewise. * sysdeps/stub/s_expm1l.c: Likewise. * sysdeps/stub/s_log1pl.c: Likewise. * sysdeps/stub/s_log2l.c: Likewise. 1997-06-18 11:46 Ulrich Drepper <drepper@cygnus.com> * sysdeps/libm-ieee754/w_gamma.c: If _LIB_VERSION is _SVID_ compute result as before last change. * sysdeps/libm-ieee754/w_gammaf.c: Likewise. * sysdeps/libm-ieee754/w_gammal.c: Likewise. 1997-06-16 23:37 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/libm-ieee754/s_remquo.c: Fix off-by-one when computing quotient. * sysdeps/libm-ieee754/s_remquof.c: Likewise. * sysdeps/libm-ieee754/s_remquol.c: Likewise. * sysdeps/m68k/fpu/s_remquo.c: Remove FIXME and special case for quotient. * sysdeps/libm-ieee754/w_gamma.c: Add missing call to exp function. Don't use global signgam. * sysdeps/libm-ieee754/w_gammaf.c: Likewise. * sysdeps/libm-ieee754/w_gammal.c: Likewise. * math/Makefile (libm-calls): Remove w_gamma_r. * sysdeps/libm-ieee754/w_gamma_r.c: Remove file. * sysdeps/libm-ieee754/w_gammaf_r.c: Remove file. * sysdeps/libm-ieee754/w_gammal_r.c: Remove file. * math/libm-test.c (atanh_test): Declare x only if needed. (signbit_test): Fix typo. (gamma_test): Check whether function is implemented. Add epsilons. (lgamma_test): Likewise. (fmod_test): Add epsilons. (exp2_test): Use right function for existence test. 1997-06-07 09:20 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/powerpc/Dist: Add fenv_const.c. * sysdeps/unix/sysv/linux/Dist: Add net/if_slip.h. * sysdeps/unix/sysv/linux/powerpc/Dist: Add init-first.h and syscall.h. * sysdeps/unix/sysv/linux/sparc/Dist: Add init-first.h. * string/Makefile (distribute): Add tst-svc.expect. * nis/Makefile (distribute): Add nis_intern.h and Banner. * elf/Makefile (distribute): Add dl-hash.h. * Rules (subdir_echo-distinfo): Add headers from $(distribute). * login/Makefile (others): Add utmpdump. * sysdeps/unix/sysv/linux/Makefile (sysdep_headers): Add net/if_slip.h. * manual/Makefile (dir-add.texi): Also look in indirectly included files. 1997-06-16 23:15 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/unix/sysv/linux/configure.in: Remove commands to generate stdio_lim.h. * sysdeps/unix/sysv/linux/mk-stdiolim.c: Remove. * sysdeps/unix/sysv/linux/Makefile: Add rules to generate stdio_lim.h here. (common-generated): Add generated files. (inhibit-stdio_lim): Define. * sysdeps/posix/Makefile [$(inhibit-stdio_lim)=yes]: Disable rules to generate stdio_lim.h. * sysdeps/unix/sysv/linux/Makefile: Suppress inclusion of dependecy files if no_deps is set. ($(objpfx)syscall-%.d): Add header file as target to dependency generation. 1997-06-14 19:19 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/unix/sysv/linux/arm/Dist: New file. 1997-06-14 17:59 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * extra-lib.mk (others): Depend on versioned shared library, not the unversioned one. * Makerules (build-shlib): Don't make the version link here. ($(common-objpfx)libc.so$(libc.so-version)): New rule for libc version link. 1997-06-16 03:07 Ulrich Drepper <drepper@cygnus.com> * sysdeps/libm-ieee754/k_standard.c: Undo change of Tue Aug 6 01:13:56 1996. * argp/argp-help.c (argp_args_usage): Supply correct argp to filter_doc. * argp/argp-help.c (hol_add_cluster): Initialize CL->depth. * argp/argp-help.c (_help): Supply STATE to argp_args_usage. * argp/argp.h (ARGP_KEY_HELP_ARGS_DOC): New macro. * argp/argp-fmtstream.c: Add casts to prevent warnings. * argp/argp.h (OPTION_NO_USAGE): New macro. * argp/argp-help.c (usage_long_opt, usage_argful_short_opt, * argp/argp-fmtstream.c (__argp_fmtstream_update): Account for case * argp/argp-help.c <stddef.h>: New include. * argp/argp.h (argp_state_help, __argp_state_help, argp_usage, * argp/argp.h (argp_program_bug_address): Make const. * argp/argp-parse.c (argp_default_parser): Set STATE->name for OPT_PROGNAME. * argp/argp-help.c (__argp_error, __argp_failure, __argp_state_help): * argp/argp-parse.c (parser_init): Set PARSER->state.flags.
1997-06-19 19:11:22 +00:00
if (x >= 172.0)
{
/* Overflow. */
*signgamp = 0;
Use math_narrow_eval more consistently. Where glibc code needs to avoid excess range and precision in floating-point arithmetic, code variously uses either asms or volatile to force the results of that arithmetic to memory; mostly this is conditional on FLT_EVAL_METHOD, but in the case of lrint / llrint functions some use of volatile is unconditional (and is present unnecessarily in versions for long double). This patch make such code use the recently-added math_narrow_eval macro consistently, removing the unnecessary uses of volatile in long double lrint / llrint implementations completely. Tested for x86_64, x86, mips64 and powerpc. * math/s_nexttowardf.c (__nexttowardf): Use math_narrow_eval. * stdlib/strtod_l.c: Include <math_private.h>. (overflow_value): Use math_narrow_eval. (underflow_value): Likewise. * sysdeps/i386/fpu/s_nexttoward.c (__nexttoward): Likewise. * sysdeps/i386/fpu/s_nexttowardf.c (__nexttowardf): Likewise. * sysdeps/ieee754/dbl-64/e_gamma_r.c (gamma_positive): Likewise. (__ieee754_gamma_r): Likewise. * sysdeps/ieee754/dbl-64/gamma_productf.c (__gamma_productf): Likewise. * sysdeps/ieee754/dbl-64/k_rem_pio2.c (__kernel_rem_pio2): Likewise. * sysdeps/ieee754/dbl-64/lgamma_neg.c (__lgamma_neg): Likewise. * sysdeps/ieee754/dbl-64/s_erf.c (__erfc): Likewise. * sysdeps/ieee754/dbl-64/s_llrint.c (__llrint): Likewise. * sysdeps/ieee754/dbl-64/s_lrint.c (__lrint): Likewise. * sysdeps/ieee754/flt-32/e_gammaf_r.c (gammaf_positive): Likewise. (__ieee754_gammaf_r): Likewise. * sysdeps/ieee754/flt-32/k_rem_pio2f.c (__kernel_rem_pio2f): Likewise. * sysdeps/ieee754/flt-32/lgamma_negf.c (__lgamma_negf): Likewise. * sysdeps/ieee754/flt-32/s_erff.c (__erfcf): Likewise. * sysdeps/ieee754/flt-32/s_llrintf.c (__llrintf): Likewise. * sysdeps/ieee754/flt-32/s_lrintf.c (__lrintf): Likewise. * sysdeps/ieee754/ldbl-128/s_llrintl.c (__llrintl): Do not use volatile. * sysdeps/ieee754/ldbl-128/s_lrintl.c (__lrintl): Likewise. * sysdeps/ieee754/ldbl-128/s_nexttoward.c (__nexttoward): Use math_narrow_eval. * sysdeps/ieee754/ldbl-128ibm/s_nexttoward.c (__nexttoward): Likewise. * sysdeps/ieee754/ldbl-128ibm/s_nexttowardf.c (__nexttowardf): Likewise. * sysdeps/ieee754/ldbl-96/gamma_product.c (__gamma_product): Likewise. * sysdeps/ieee754/ldbl-96/s_llrintl.c (__llrintl): Do not use volatile. * sysdeps/ieee754/ldbl-96/s_lrintl.c (__lrintl): Likewise. * sysdeps/ieee754/ldbl-96/s_nexttoward.c (__nexttoward): Use math_narrow_eval. * sysdeps/ieee754/ldbl-96/s_nexttowardf.c (__nexttowardf): Likewise. * sysdeps/ieee754/ldbl-opt/s_nexttowardfd.c (__nldbl_nexttowardf): Likewise.
2015-09-23 18:14:57 +00:00
ret = math_narrow_eval (DBL_MAX * DBL_MAX);
Improve tgamma accuracy (bug 18613). In non-default rounding modes, tgamma can be slightly less accurate than permitted by glibc's accuracy goals. Part of the problem is error accumulation, addressed in this patch by setting round-to-nearest for internal computations. However, there was also a bug in the code dealing with computing pow (x + n, x + n) where x + n is not exactly representable, providing another source of error even in round-to-nearest mode; it was necessary to address both bugs to get errors for all testcases within glibc's accuracy goals. Given this second fix, accuracy in round-to-nearest mode is also improved (hence regeneration of ulps for tgamma should be from scratch - truncate libm-test-ulps or at least remove existing tgamma entries - so that the expected ulps can be reduced). Some additional complications also arose. Certain tgamma tests should strictly, according to IEEE semantics, overflow or not depending on the rounding mode; this is beyond the scope of glibc's accuracy goals for any function without exactly-determined results, but gen-auto-libm-tests doesn't handle being lax there as it does for underflow. (libm-test.inc also doesn't handle being lax about whether the result in cases very close to the overflow threshold is infinity or a finite value close to overflow, but that doesn't cause problems in this case though I've seen it cause problems with random test generation for some functions.) Thus, spurious-overflow markings, with a comment, are added to auto-libm-test-in (no bug in Bugzilla because the issue is with the testsuite, not a user-visible bug in glibc). And on x86, after the patch I saw ERANGE issues as previously reported by Carlos (see my commentary in <https://sourceware.org/ml/libc-alpha/2015-01/msg00485.html>), which needed addressing by ensuring excess range and precision were eliminated at various points if FLT_EVAL_METHOD != 0. I also noticed and fixed a cosmetic issue where 1.0f was used in long double functions and should have been 1.0L. This completes the move of all functions to testing in all rounding modes with ALL_RM_TEST, so gen-libm-have-vector-test.sh is updated to remove the workaround for some functions not using ALL_RM_TEST. Tested for x86_64, x86, mips64 and powerpc. [BZ #18613] * sysdeps/ieee754/dbl-64/e_gamma_r.c (gamma_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gamma_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. * sysdeps/ieee754/flt-32/e_gammaf_r.c (gammaf_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammaf_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. * sysdeps/ieee754/ldbl-128/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * sysdeps/ieee754/ldbl-128ibm/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * sysdeps/ieee754/ldbl-96/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * math/libm-test.inc (tgamma_test_data): Remove one test. Moved to auto-libm-test-in. (tgamma_test): Use ALL_RM_TEST. * math/auto-libm-test-in: Add one test of tgamma. Mark some other tests of tgamma with spurious-overflow. * math/auto-libm-test-out: Regenerated. * math/gen-libm-have-vector-test.sh: Do not check for START. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-06-29 23:29:35 +00:00
return ret;
}
Improve tgamma accuracy (bug 18613). In non-default rounding modes, tgamma can be slightly less accurate than permitted by glibc's accuracy goals. Part of the problem is error accumulation, addressed in this patch by setting round-to-nearest for internal computations. However, there was also a bug in the code dealing with computing pow (x + n, x + n) where x + n is not exactly representable, providing another source of error even in round-to-nearest mode; it was necessary to address both bugs to get errors for all testcases within glibc's accuracy goals. Given this second fix, accuracy in round-to-nearest mode is also improved (hence regeneration of ulps for tgamma should be from scratch - truncate libm-test-ulps or at least remove existing tgamma entries - so that the expected ulps can be reduced). Some additional complications also arose. Certain tgamma tests should strictly, according to IEEE semantics, overflow or not depending on the rounding mode; this is beyond the scope of glibc's accuracy goals for any function without exactly-determined results, but gen-auto-libm-tests doesn't handle being lax there as it does for underflow. (libm-test.inc also doesn't handle being lax about whether the result in cases very close to the overflow threshold is infinity or a finite value close to overflow, but that doesn't cause problems in this case though I've seen it cause problems with random test generation for some functions.) Thus, spurious-overflow markings, with a comment, are added to auto-libm-test-in (no bug in Bugzilla because the issue is with the testsuite, not a user-visible bug in glibc). And on x86, after the patch I saw ERANGE issues as previously reported by Carlos (see my commentary in <https://sourceware.org/ml/libc-alpha/2015-01/msg00485.html>), which needed addressing by ensuring excess range and precision were eliminated at various points if FLT_EVAL_METHOD != 0. I also noticed and fixed a cosmetic issue where 1.0f was used in long double functions and should have been 1.0L. This completes the move of all functions to testing in all rounding modes with ALL_RM_TEST, so gen-libm-have-vector-test.sh is updated to remove the workaround for some functions not using ALL_RM_TEST. Tested for x86_64, x86, mips64 and powerpc. [BZ #18613] * sysdeps/ieee754/dbl-64/e_gamma_r.c (gamma_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gamma_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. * sysdeps/ieee754/flt-32/e_gammaf_r.c (gammaf_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammaf_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. * sysdeps/ieee754/ldbl-128/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * sysdeps/ieee754/ldbl-128ibm/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * sysdeps/ieee754/ldbl-96/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * math/libm-test.inc (tgamma_test_data): Remove one test. Moved to auto-libm-test-in. (tgamma_test): Use ALL_RM_TEST. * math/auto-libm-test-in: Add one test of tgamma. Mark some other tests of tgamma with spurious-overflow. * math/auto-libm-test-out: Regenerated. * math/gen-libm-have-vector-test.sh: Do not check for START. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-06-29 23:29:35 +00:00
else
{
Improve tgamma accuracy (bug 18613). In non-default rounding modes, tgamma can be slightly less accurate than permitted by glibc's accuracy goals. Part of the problem is error accumulation, addressed in this patch by setting round-to-nearest for internal computations. However, there was also a bug in the code dealing with computing pow (x + n, x + n) where x + n is not exactly representable, providing another source of error even in round-to-nearest mode; it was necessary to address both bugs to get errors for all testcases within glibc's accuracy goals. Given this second fix, accuracy in round-to-nearest mode is also improved (hence regeneration of ulps for tgamma should be from scratch - truncate libm-test-ulps or at least remove existing tgamma entries - so that the expected ulps can be reduced). Some additional complications also arose. Certain tgamma tests should strictly, according to IEEE semantics, overflow or not depending on the rounding mode; this is beyond the scope of glibc's accuracy goals for any function without exactly-determined results, but gen-auto-libm-tests doesn't handle being lax there as it does for underflow. (libm-test.inc also doesn't handle being lax about whether the result in cases very close to the overflow threshold is infinity or a finite value close to overflow, but that doesn't cause problems in this case though I've seen it cause problems with random test generation for some functions.) Thus, spurious-overflow markings, with a comment, are added to auto-libm-test-in (no bug in Bugzilla because the issue is with the testsuite, not a user-visible bug in glibc). And on x86, after the patch I saw ERANGE issues as previously reported by Carlos (see my commentary in <https://sourceware.org/ml/libc-alpha/2015-01/msg00485.html>), which needed addressing by ensuring excess range and precision were eliminated at various points if FLT_EVAL_METHOD != 0. I also noticed and fixed a cosmetic issue where 1.0f was used in long double functions and should have been 1.0L. This completes the move of all functions to testing in all rounding modes with ALL_RM_TEST, so gen-libm-have-vector-test.sh is updated to remove the workaround for some functions not using ALL_RM_TEST. Tested for x86_64, x86, mips64 and powerpc. [BZ #18613] * sysdeps/ieee754/dbl-64/e_gamma_r.c (gamma_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gamma_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. * sysdeps/ieee754/flt-32/e_gammaf_r.c (gammaf_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammaf_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. * sysdeps/ieee754/ldbl-128/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * sysdeps/ieee754/ldbl-128ibm/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * sysdeps/ieee754/ldbl-96/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * math/libm-test.inc (tgamma_test_data): Remove one test. Moved to auto-libm-test-in. (tgamma_test): Use ALL_RM_TEST. * math/auto-libm-test-in: Add one test of tgamma. Mark some other tests of tgamma with spurious-overflow. * math/auto-libm-test-out: Regenerated. * math/gen-libm-have-vector-test.sh: Do not check for START. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-06-29 23:29:35 +00:00
SET_RESTORE_ROUND (FE_TONEAREST);
if (x > 0.0)
{
*signgamp = 0;
int exp2_adj;
double tret = gamma_positive (x, &exp2_adj);
ret = __scalbn (tret, exp2_adj);
}
else if (x >= -DBL_EPSILON / 4.0)
{
*signgamp = 0;
ret = 1.0 / x;
}
else
{
double tx = __trunc (x);
*signgamp = (tx == 2.0 * __trunc (tx / 2.0)) ? -1 : 1;
if (x <= -184.0)
/* Underflow. */
ret = DBL_MIN * DBL_MIN;
else
{
double frac = tx - x;
if (frac > 0.5)
frac = 1.0 - frac;
double sinpix = (frac <= 0.25
? __sin (M_PI * frac)
: __cos (M_PI * (0.5 - frac)));
int exp2_adj;
double tret = M_PI / (-x * sinpix
* gamma_positive (-x, &exp2_adj));
ret = __scalbn (tret, -exp2_adj);
Refactor code forcing underflow exceptions. Various floating-point functions have code to force underflow exceptions if a tiny result was computed in a way that might not have resulted in such exceptions even though the result is inexact. This typically uses math_force_eval to ensure that the underflowing expression is evaluated, but sometimes uses volatile. This patch refactors such code to use three new macros math_check_force_underflow, math_check_force_underflow_nonneg and math_check_force_underflow_complex (which in turn use math_force_eval). In the limited number of cases not suited to a simple conversion to these macros, existing uses of volatile are changed to use math_force_eval instead. The converted code does not always execute exactly the same sequence of operations as the original code, but the overall effects should be the same. Tested for x86_64, x86, mips64 and powerpc. * sysdeps/generic/math_private.h (fabs_tg): New macro. (min_of_type): Likewise. (math_check_force_underflow): Likewise. (math_check_force_underflow_nonneg): Likewise. (math_check_force_underflow_complex): Likewise. * math/e_exp2l.c (__ieee754_exp2l): Use math_check_force_underflow_nonneg. * math/k_casinh.c (__kernel_casinh): Likewise. * math/k_casinhf.c (__kernel_casinhf): Likewise. * math/k_casinhl.c (__kernel_casinhl): Likewise. * math/s_catan.c (__catan): Use math_check_force_underflow_complex. * math/s_catanf.c (__catanf): Likewise. * math/s_catanh.c (__catanh): Likewise. * math/s_catanhf.c (__catanhf): Likewise. * math/s_catanhl.c (__catanhl): Likewise. * math/s_catanl.c (__catanl): Likewise. * math/s_ccosh.c (__ccosh): Likewise. * math/s_ccoshf.c (__ccoshf): Likewise. * math/s_ccoshl.c (__ccoshl): Likewise. * math/s_cexp.c (__cexp): Likewise. * math/s_cexpf.c (__cexpf): Likewise. * math/s_cexpl.c (__cexpl): Likewise. * math/s_clog.c (__clog): Use math_check_force_underflow_nonneg. * math/s_clog10.c (__clog10): Likewise. * math/s_clog10f.c (__clog10f): Likewise. * math/s_clog10l.c (__clog10l): Likewise. * math/s_clogf.c (__clogf): Likewise. * math/s_clogl.c (__clogl): Likewise. * math/s_csin.c (__csin): Use math_check_force_underflow_complex. * math/s_csinf.c (__csinf): Likewise. * math/s_csinh.c (__csinh): Likewise. * math/s_csinhf.c (__csinhf): Likewise. * math/s_csinhl.c (__csinhl): Likewise. * math/s_csinl.c (__csinl): Likewise. * math/s_csqrt.c (__csqrt): Use math_check_force_underflow. * math/s_csqrtf.c (__csqrtf): Likewise. * math/s_csqrtl.c (__csqrtl): Likewise. * math/s_ctan.c (__ctan): Use math_check_force_underflow_complex. * math/s_ctanf.c (__ctanf): Likewise. * math/s_ctanh.c (__ctanh): Likewise. * math/s_ctanhf.c (__ctanhf): Likewise. * math/s_ctanhl.c (__ctanhl): Likewise. * math/s_ctanl.c (__ctanl): Likewise. * stdlib/strtod_l.c (round_and_return): Use math_force_eval instead of volatile. * sysdeps/ieee754/dbl-64/e_asin.c (__ieee754_asin): Use math_check_force_underflow. * sysdeps/ieee754/dbl-64/e_atanh.c (__ieee754_atanh): Likewise. * sysdeps/ieee754/dbl-64/e_exp.c (__ieee754_exp): Do not use volatile when forcing underflow. * sysdeps/ieee754/dbl-64/e_exp2.c (__ieee754_exp2): Use math_check_force_underflow_nonneg. * sysdeps/ieee754/dbl-64/e_gamma_r.c (__ieee754_gamma_r): Likewise. * sysdeps/ieee754/dbl-64/e_j1.c (__ieee754_j1): Use math_check_force_underflow. * sysdeps/ieee754/dbl-64/e_jn.c (__ieee754_jn): Likewise. * sysdeps/ieee754/dbl-64/e_sinh.c (__ieee754_sinh): Likewise. * sysdeps/ieee754/dbl-64/s_asinh.c (__asinh): Likewise. * sysdeps/ieee754/dbl-64/s_atan.c (atan): Use math_check_force_underflow_nonneg. * sysdeps/ieee754/dbl-64/s_erf.c (__erf): Use math_check_force_underflow. * sysdeps/ieee754/dbl-64/s_expm1.c (__expm1): Likewise. * sysdeps/ieee754/dbl-64/s_fma.c (__fma): Use math_force_eval instead of volatile. * sysdeps/ieee754/dbl-64/s_log1p.c (__log1p): Use math_check_force_underflow. * sysdeps/ieee754/dbl-64/s_sin.c (__sin): Likewise. * sysdeps/ieee754/dbl-64/s_tan.c (tan): Use math_check_force_underflow_nonneg. * sysdeps/ieee754/dbl-64/s_tanh.c (__tanh): Use math_check_force_underflow. * sysdeps/ieee754/flt-32/e_asinf.c (__ieee754_asinf): Likewise. * sysdeps/ieee754/flt-32/e_atanhf.c (__ieee754_atanhf): Likewise. * sysdeps/ieee754/flt-32/e_exp2f.c (__ieee754_exp2f): Use math_check_force_underflow_nonneg. * sysdeps/ieee754/flt-32/e_gammaf_r.c (__ieee754_gammaf_r): Likewise. * sysdeps/ieee754/flt-32/e_j1f.c (__ieee754_j1f): Use math_check_force_underflow. * sysdeps/ieee754/flt-32/e_jnf.c (__ieee754_jnf): Likewise. * sysdeps/ieee754/flt-32/e_sinhf.c (__ieee754_sinhf): Likewise. * sysdeps/ieee754/flt-32/k_sinf.c (__kernel_sinf): Likewise. * sysdeps/ieee754/flt-32/k_tanf.c (__kernel_tanf): Likewise. * sysdeps/ieee754/flt-32/s_asinhf.c (__asinhf): Likewise. * sysdeps/ieee754/flt-32/s_atanf.c (__atanf): Likewise. * sysdeps/ieee754/flt-32/s_erff.c (__erff): Likewise. * sysdeps/ieee754/flt-32/s_expm1f.c (__expm1f): Likewise. * sysdeps/ieee754/flt-32/s_log1pf.c (__log1pf): Likewise. * sysdeps/ieee754/flt-32/s_tanhf.c (__tanhf): Likewise. * sysdeps/ieee754/ldbl-128/e_asinl.c (__ieee754_asinl): Likewise. * sysdeps/ieee754/ldbl-128/e_atanhl.c (__ieee754_atanhl): Likewise. * sysdeps/ieee754/ldbl-128/e_expl.c (__ieee754_expl): Use math_check_force_underflow_nonneg. * sysdeps/ieee754/ldbl-128/e_gammal_r.c (__ieee754_gammal_r): Likewise. * sysdeps/ieee754/ldbl-128/e_j1l.c (__ieee754_j1l): Use math_check_force_underflow. * sysdeps/ieee754/ldbl-128/e_jnl.c (__ieee754_jnl): Likewise. * sysdeps/ieee754/ldbl-128/e_sinhl.c (__ieee754_sinhl): Likewise. * sysdeps/ieee754/ldbl-128/k_sincosl.c (__kernel_sincosl): Likewise. * sysdeps/ieee754/ldbl-128/k_sinl.c (__kernel_sinl): Likewise. * sysdeps/ieee754/ldbl-128/k_tanl.c (__kernel_tanl): Likewise. * sysdeps/ieee754/ldbl-128/s_asinhl.c (__asinhl): Likewise. * sysdeps/ieee754/ldbl-128/s_atanl.c (__atanl): Likewise. * sysdeps/ieee754/ldbl-128/s_erfl.c (__erfl): Likewise. * sysdeps/ieee754/ldbl-128/s_expm1l.c (__expm1l): Likewise. * sysdeps/ieee754/ldbl-128/s_fmal.c (__fmal): Use math_force_eval instead of volatile. * sysdeps/ieee754/ldbl-128/s_log1pl.c (__log1pl): Use math_check_force_underflow. * sysdeps/ieee754/ldbl-128/s_tanhl.c (__tanhl): Likewise. * sysdeps/ieee754/ldbl-128ibm/e_asinl.c (__ieee754_asinl): Use math_check_force_underflow. * sysdeps/ieee754/ldbl-128ibm/e_atanhl.c (__ieee754_atanhl): Likewise. * sysdeps/ieee754/ldbl-128ibm/e_gammal_r.c (__ieee754_gammal_r): Use math_check_force_underflow_nonneg. * sysdeps/ieee754/ldbl-128ibm/e_jnl.c (__ieee754_jnl): Use math_check_force_underflow. * sysdeps/ieee754/ldbl-128ibm/e_sinhl.c (__ieee754_sinhl): Likewise. * sysdeps/ieee754/ldbl-128ibm/k_sincosl.c (__kernel_sincosl): Likewise. * sysdeps/ieee754/ldbl-128ibm/k_sinl.c (__kernel_sinl): Likewise. * sysdeps/ieee754/ldbl-128ibm/k_tanl.c (__kernel_tanl): Likewise. * sysdeps/ieee754/ldbl-128ibm/s_asinhl.c (__asinhl): Likewise. * sysdeps/ieee754/ldbl-128ibm/s_atanl.c (__atanl): Likewise. * sysdeps/ieee754/ldbl-128ibm/s_erfl.c (__erfl): Likewise. * sysdeps/ieee754/ldbl-128ibm/s_tanhl.c (__tanhl): Likewise. * sysdeps/ieee754/ldbl-96/e_asinl.c (__ieee754_asinl): Likewise. * sysdeps/ieee754/ldbl-96/e_atanhl.c (__ieee754_atanhl): Likewise. * sysdeps/ieee754/ldbl-96/e_gammal_r.c (__ieee754_gammal_r): Use math_check_force_underflow_nonneg. * sysdeps/ieee754/ldbl-96/e_j1l.c (__ieee754_j1l): Use math_check_force_underflow. * sysdeps/ieee754/ldbl-96/e_jnl.c (__ieee754_jnl): Likewise. * sysdeps/ieee754/ldbl-96/e_sinhl.c (__ieee754_sinhl): Likewise. * sysdeps/ieee754/ldbl-96/k_sinl.c (__kernel_sinl): Likewise. * sysdeps/ieee754/ldbl-96/k_tanl.c (__kernel_tanl): Use math_check_force_underflow_nonneg. * sysdeps/ieee754/ldbl-96/s_asinhl.c (__asinhl): Use math_check_force_underflow. * sysdeps/ieee754/ldbl-96/s_erfl.c (__erfl): Likewise. * sysdeps/ieee754/ldbl-96/s_fmal.c (__fmal): Use math_force_eval instead of volatile. * sysdeps/ieee754/ldbl-96/s_tanhl.c (__tanhl): Use math_check_force_underflow.
2015-09-23 22:42:30 +00:00
math_check_force_underflow_nonneg (ret);
Improve tgamma accuracy (bug 18613). In non-default rounding modes, tgamma can be slightly less accurate than permitted by glibc's accuracy goals. Part of the problem is error accumulation, addressed in this patch by setting round-to-nearest for internal computations. However, there was also a bug in the code dealing with computing pow (x + n, x + n) where x + n is not exactly representable, providing another source of error even in round-to-nearest mode; it was necessary to address both bugs to get errors for all testcases within glibc's accuracy goals. Given this second fix, accuracy in round-to-nearest mode is also improved (hence regeneration of ulps for tgamma should be from scratch - truncate libm-test-ulps or at least remove existing tgamma entries - so that the expected ulps can be reduced). Some additional complications also arose. Certain tgamma tests should strictly, according to IEEE semantics, overflow or not depending on the rounding mode; this is beyond the scope of glibc's accuracy goals for any function without exactly-determined results, but gen-auto-libm-tests doesn't handle being lax there as it does for underflow. (libm-test.inc also doesn't handle being lax about whether the result in cases very close to the overflow threshold is infinity or a finite value close to overflow, but that doesn't cause problems in this case though I've seen it cause problems with random test generation for some functions.) Thus, spurious-overflow markings, with a comment, are added to auto-libm-test-in (no bug in Bugzilla because the issue is with the testsuite, not a user-visible bug in glibc). And on x86, after the patch I saw ERANGE issues as previously reported by Carlos (see my commentary in <https://sourceware.org/ml/libc-alpha/2015-01/msg00485.html>), which needed addressing by ensuring excess range and precision were eliminated at various points if FLT_EVAL_METHOD != 0. I also noticed and fixed a cosmetic issue where 1.0f was used in long double functions and should have been 1.0L. This completes the move of all functions to testing in all rounding modes with ALL_RM_TEST, so gen-libm-have-vector-test.sh is updated to remove the workaround for some functions not using ALL_RM_TEST. Tested for x86_64, x86, mips64 and powerpc. [BZ #18613] * sysdeps/ieee754/dbl-64/e_gamma_r.c (gamma_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gamma_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. * sysdeps/ieee754/flt-32/e_gammaf_r.c (gammaf_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammaf_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. * sysdeps/ieee754/ldbl-128/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * sysdeps/ieee754/ldbl-128ibm/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * sysdeps/ieee754/ldbl-96/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * math/libm-test.inc (tgamma_test_data): Remove one test. Moved to auto-libm-test-in. (tgamma_test): Use ALL_RM_TEST. * math/auto-libm-test-in: Add one test of tgamma. Mark some other tests of tgamma with spurious-overflow. * math/auto-libm-test-out: Regenerated. * math/gen-libm-have-vector-test.sh: Do not check for START. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-06-29 23:29:35 +00:00
}
}
Use math_narrow_eval more consistently. Where glibc code needs to avoid excess range and precision in floating-point arithmetic, code variously uses either asms or volatile to force the results of that arithmetic to memory; mostly this is conditional on FLT_EVAL_METHOD, but in the case of lrint / llrint functions some use of volatile is unconditional (and is present unnecessarily in versions for long double). This patch make such code use the recently-added math_narrow_eval macro consistently, removing the unnecessary uses of volatile in long double lrint / llrint implementations completely. Tested for x86_64, x86, mips64 and powerpc. * math/s_nexttowardf.c (__nexttowardf): Use math_narrow_eval. * stdlib/strtod_l.c: Include <math_private.h>. (overflow_value): Use math_narrow_eval. (underflow_value): Likewise. * sysdeps/i386/fpu/s_nexttoward.c (__nexttoward): Likewise. * sysdeps/i386/fpu/s_nexttowardf.c (__nexttowardf): Likewise. * sysdeps/ieee754/dbl-64/e_gamma_r.c (gamma_positive): Likewise. (__ieee754_gamma_r): Likewise. * sysdeps/ieee754/dbl-64/gamma_productf.c (__gamma_productf): Likewise. * sysdeps/ieee754/dbl-64/k_rem_pio2.c (__kernel_rem_pio2): Likewise. * sysdeps/ieee754/dbl-64/lgamma_neg.c (__lgamma_neg): Likewise. * sysdeps/ieee754/dbl-64/s_erf.c (__erfc): Likewise. * sysdeps/ieee754/dbl-64/s_llrint.c (__llrint): Likewise. * sysdeps/ieee754/dbl-64/s_lrint.c (__lrint): Likewise. * sysdeps/ieee754/flt-32/e_gammaf_r.c (gammaf_positive): Likewise. (__ieee754_gammaf_r): Likewise. * sysdeps/ieee754/flt-32/k_rem_pio2f.c (__kernel_rem_pio2f): Likewise. * sysdeps/ieee754/flt-32/lgamma_negf.c (__lgamma_negf): Likewise. * sysdeps/ieee754/flt-32/s_erff.c (__erfcf): Likewise. * sysdeps/ieee754/flt-32/s_llrintf.c (__llrintf): Likewise. * sysdeps/ieee754/flt-32/s_lrintf.c (__lrintf): Likewise. * sysdeps/ieee754/ldbl-128/s_llrintl.c (__llrintl): Do not use volatile. * sysdeps/ieee754/ldbl-128/s_lrintl.c (__lrintl): Likewise. * sysdeps/ieee754/ldbl-128/s_nexttoward.c (__nexttoward): Use math_narrow_eval. * sysdeps/ieee754/ldbl-128ibm/s_nexttoward.c (__nexttoward): Likewise. * sysdeps/ieee754/ldbl-128ibm/s_nexttowardf.c (__nexttowardf): Likewise. * sysdeps/ieee754/ldbl-96/gamma_product.c (__gamma_product): Likewise. * sysdeps/ieee754/ldbl-96/s_llrintl.c (__llrintl): Do not use volatile. * sysdeps/ieee754/ldbl-96/s_lrintl.c (__lrintl): Likewise. * sysdeps/ieee754/ldbl-96/s_nexttoward.c (__nexttoward): Use math_narrow_eval. * sysdeps/ieee754/ldbl-96/s_nexttowardf.c (__nexttowardf): Likewise. * sysdeps/ieee754/ldbl-opt/s_nexttowardfd.c (__nldbl_nexttowardf): Likewise.
2015-09-23 18:14:57 +00:00
ret = math_narrow_eval (ret);
}
Improve tgamma accuracy (bug 18613). In non-default rounding modes, tgamma can be slightly less accurate than permitted by glibc's accuracy goals. Part of the problem is error accumulation, addressed in this patch by setting round-to-nearest for internal computations. However, there was also a bug in the code dealing with computing pow (x + n, x + n) where x + n is not exactly representable, providing another source of error even in round-to-nearest mode; it was necessary to address both bugs to get errors for all testcases within glibc's accuracy goals. Given this second fix, accuracy in round-to-nearest mode is also improved (hence regeneration of ulps for tgamma should be from scratch - truncate libm-test-ulps or at least remove existing tgamma entries - so that the expected ulps can be reduced). Some additional complications also arose. Certain tgamma tests should strictly, according to IEEE semantics, overflow or not depending on the rounding mode; this is beyond the scope of glibc's accuracy goals for any function without exactly-determined results, but gen-auto-libm-tests doesn't handle being lax there as it does for underflow. (libm-test.inc also doesn't handle being lax about whether the result in cases very close to the overflow threshold is infinity or a finite value close to overflow, but that doesn't cause problems in this case though I've seen it cause problems with random test generation for some functions.) Thus, spurious-overflow markings, with a comment, are added to auto-libm-test-in (no bug in Bugzilla because the issue is with the testsuite, not a user-visible bug in glibc). And on x86, after the patch I saw ERANGE issues as previously reported by Carlos (see my commentary in <https://sourceware.org/ml/libc-alpha/2015-01/msg00485.html>), which needed addressing by ensuring excess range and precision were eliminated at various points if FLT_EVAL_METHOD != 0. I also noticed and fixed a cosmetic issue where 1.0f was used in long double functions and should have been 1.0L. This completes the move of all functions to testing in all rounding modes with ALL_RM_TEST, so gen-libm-have-vector-test.sh is updated to remove the workaround for some functions not using ALL_RM_TEST. Tested for x86_64, x86, mips64 and powerpc. [BZ #18613] * sysdeps/ieee754/dbl-64/e_gamma_r.c (gamma_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gamma_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. * sysdeps/ieee754/flt-32/e_gammaf_r.c (gammaf_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammaf_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. * sysdeps/ieee754/ldbl-128/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * sysdeps/ieee754/ldbl-128ibm/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * sysdeps/ieee754/ldbl-96/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * math/libm-test.inc (tgamma_test_data): Remove one test. Moved to auto-libm-test-in. (tgamma_test): Use ALL_RM_TEST. * math/auto-libm-test-in: Add one test of tgamma. Mark some other tests of tgamma with spurious-overflow. * math/auto-libm-test-out: Regenerated. * math/gen-libm-have-vector-test.sh: Do not check for START. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-06-29 23:29:35 +00:00
if (isinf (ret) && x != 0)
{
Improve tgamma accuracy (bug 18613). In non-default rounding modes, tgamma can be slightly less accurate than permitted by glibc's accuracy goals. Part of the problem is error accumulation, addressed in this patch by setting round-to-nearest for internal computations. However, there was also a bug in the code dealing with computing pow (x + n, x + n) where x + n is not exactly representable, providing another source of error even in round-to-nearest mode; it was necessary to address both bugs to get errors for all testcases within glibc's accuracy goals. Given this second fix, accuracy in round-to-nearest mode is also improved (hence regeneration of ulps for tgamma should be from scratch - truncate libm-test-ulps or at least remove existing tgamma entries - so that the expected ulps can be reduced). Some additional complications also arose. Certain tgamma tests should strictly, according to IEEE semantics, overflow or not depending on the rounding mode; this is beyond the scope of glibc's accuracy goals for any function without exactly-determined results, but gen-auto-libm-tests doesn't handle being lax there as it does for underflow. (libm-test.inc also doesn't handle being lax about whether the result in cases very close to the overflow threshold is infinity or a finite value close to overflow, but that doesn't cause problems in this case though I've seen it cause problems with random test generation for some functions.) Thus, spurious-overflow markings, with a comment, are added to auto-libm-test-in (no bug in Bugzilla because the issue is with the testsuite, not a user-visible bug in glibc). And on x86, after the patch I saw ERANGE issues as previously reported by Carlos (see my commentary in <https://sourceware.org/ml/libc-alpha/2015-01/msg00485.html>), which needed addressing by ensuring excess range and precision were eliminated at various points if FLT_EVAL_METHOD != 0. I also noticed and fixed a cosmetic issue where 1.0f was used in long double functions and should have been 1.0L. This completes the move of all functions to testing in all rounding modes with ALL_RM_TEST, so gen-libm-have-vector-test.sh is updated to remove the workaround for some functions not using ALL_RM_TEST. Tested for x86_64, x86, mips64 and powerpc. [BZ #18613] * sysdeps/ieee754/dbl-64/e_gamma_r.c (gamma_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gamma_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. * sysdeps/ieee754/flt-32/e_gammaf_r.c (gammaf_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammaf_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. * sysdeps/ieee754/ldbl-128/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * sysdeps/ieee754/ldbl-128ibm/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * sysdeps/ieee754/ldbl-96/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * math/libm-test.inc (tgamma_test_data): Remove one test. Moved to auto-libm-test-in. (tgamma_test): Use ALL_RM_TEST. * math/auto-libm-test-in: Add one test of tgamma. Mark some other tests of tgamma with spurious-overflow. * math/auto-libm-test-out: Regenerated. * math/gen-libm-have-vector-test.sh: Do not check for START. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-06-29 23:29:35 +00:00
if (*signgamp < 0)
{
Use math_narrow_eval more consistently. Where glibc code needs to avoid excess range and precision in floating-point arithmetic, code variously uses either asms or volatile to force the results of that arithmetic to memory; mostly this is conditional on FLT_EVAL_METHOD, but in the case of lrint / llrint functions some use of volatile is unconditional (and is present unnecessarily in versions for long double). This patch make such code use the recently-added math_narrow_eval macro consistently, removing the unnecessary uses of volatile in long double lrint / llrint implementations completely. Tested for x86_64, x86, mips64 and powerpc. * math/s_nexttowardf.c (__nexttowardf): Use math_narrow_eval. * stdlib/strtod_l.c: Include <math_private.h>. (overflow_value): Use math_narrow_eval. (underflow_value): Likewise. * sysdeps/i386/fpu/s_nexttoward.c (__nexttoward): Likewise. * sysdeps/i386/fpu/s_nexttowardf.c (__nexttowardf): Likewise. * sysdeps/ieee754/dbl-64/e_gamma_r.c (gamma_positive): Likewise. (__ieee754_gamma_r): Likewise. * sysdeps/ieee754/dbl-64/gamma_productf.c (__gamma_productf): Likewise. * sysdeps/ieee754/dbl-64/k_rem_pio2.c (__kernel_rem_pio2): Likewise. * sysdeps/ieee754/dbl-64/lgamma_neg.c (__lgamma_neg): Likewise. * sysdeps/ieee754/dbl-64/s_erf.c (__erfc): Likewise. * sysdeps/ieee754/dbl-64/s_llrint.c (__llrint): Likewise. * sysdeps/ieee754/dbl-64/s_lrint.c (__lrint): Likewise. * sysdeps/ieee754/flt-32/e_gammaf_r.c (gammaf_positive): Likewise. (__ieee754_gammaf_r): Likewise. * sysdeps/ieee754/flt-32/k_rem_pio2f.c (__kernel_rem_pio2f): Likewise. * sysdeps/ieee754/flt-32/lgamma_negf.c (__lgamma_negf): Likewise. * sysdeps/ieee754/flt-32/s_erff.c (__erfcf): Likewise. * sysdeps/ieee754/flt-32/s_llrintf.c (__llrintf): Likewise. * sysdeps/ieee754/flt-32/s_lrintf.c (__lrintf): Likewise. * sysdeps/ieee754/ldbl-128/s_llrintl.c (__llrintl): Do not use volatile. * sysdeps/ieee754/ldbl-128/s_lrintl.c (__lrintl): Likewise. * sysdeps/ieee754/ldbl-128/s_nexttoward.c (__nexttoward): Use math_narrow_eval. * sysdeps/ieee754/ldbl-128ibm/s_nexttoward.c (__nexttoward): Likewise. * sysdeps/ieee754/ldbl-128ibm/s_nexttowardf.c (__nexttowardf): Likewise. * sysdeps/ieee754/ldbl-96/gamma_product.c (__gamma_product): Likewise. * sysdeps/ieee754/ldbl-96/s_llrintl.c (__llrintl): Do not use volatile. * sysdeps/ieee754/ldbl-96/s_lrintl.c (__lrintl): Likewise. * sysdeps/ieee754/ldbl-96/s_nexttoward.c (__nexttoward): Use math_narrow_eval. * sysdeps/ieee754/ldbl-96/s_nexttowardf.c (__nexttowardf): Likewise. * sysdeps/ieee754/ldbl-opt/s_nexttowardfd.c (__nldbl_nexttowardf): Likewise.
2015-09-23 18:14:57 +00:00
ret = math_narrow_eval (-__copysign (DBL_MAX, ret) * DBL_MAX);
Improve tgamma accuracy (bug 18613). In non-default rounding modes, tgamma can be slightly less accurate than permitted by glibc's accuracy goals. Part of the problem is error accumulation, addressed in this patch by setting round-to-nearest for internal computations. However, there was also a bug in the code dealing with computing pow (x + n, x + n) where x + n is not exactly representable, providing another source of error even in round-to-nearest mode; it was necessary to address both bugs to get errors for all testcases within glibc's accuracy goals. Given this second fix, accuracy in round-to-nearest mode is also improved (hence regeneration of ulps for tgamma should be from scratch - truncate libm-test-ulps or at least remove existing tgamma entries - so that the expected ulps can be reduced). Some additional complications also arose. Certain tgamma tests should strictly, according to IEEE semantics, overflow or not depending on the rounding mode; this is beyond the scope of glibc's accuracy goals for any function without exactly-determined results, but gen-auto-libm-tests doesn't handle being lax there as it does for underflow. (libm-test.inc also doesn't handle being lax about whether the result in cases very close to the overflow threshold is infinity or a finite value close to overflow, but that doesn't cause problems in this case though I've seen it cause problems with random test generation for some functions.) Thus, spurious-overflow markings, with a comment, are added to auto-libm-test-in (no bug in Bugzilla because the issue is with the testsuite, not a user-visible bug in glibc). And on x86, after the patch I saw ERANGE issues as previously reported by Carlos (see my commentary in <https://sourceware.org/ml/libc-alpha/2015-01/msg00485.html>), which needed addressing by ensuring excess range and precision were eliminated at various points if FLT_EVAL_METHOD != 0. I also noticed and fixed a cosmetic issue where 1.0f was used in long double functions and should have been 1.0L. This completes the move of all functions to testing in all rounding modes with ALL_RM_TEST, so gen-libm-have-vector-test.sh is updated to remove the workaround for some functions not using ALL_RM_TEST. Tested for x86_64, x86, mips64 and powerpc. [BZ #18613] * sysdeps/ieee754/dbl-64/e_gamma_r.c (gamma_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gamma_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. * sysdeps/ieee754/flt-32/e_gammaf_r.c (gammaf_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammaf_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. * sysdeps/ieee754/ldbl-128/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * sysdeps/ieee754/ldbl-128ibm/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * sysdeps/ieee754/ldbl-96/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * math/libm-test.inc (tgamma_test_data): Remove one test. Moved to auto-libm-test-in. (tgamma_test): Use ALL_RM_TEST. * math/auto-libm-test-in: Add one test of tgamma. Mark some other tests of tgamma with spurious-overflow. * math/auto-libm-test-out: Regenerated. * math/gen-libm-have-vector-test.sh: Do not check for START. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-06-29 23:29:35 +00:00
ret = -ret;
}
else
Use math_narrow_eval more consistently. Where glibc code needs to avoid excess range and precision in floating-point arithmetic, code variously uses either asms or volatile to force the results of that arithmetic to memory; mostly this is conditional on FLT_EVAL_METHOD, but in the case of lrint / llrint functions some use of volatile is unconditional (and is present unnecessarily in versions for long double). This patch make such code use the recently-added math_narrow_eval macro consistently, removing the unnecessary uses of volatile in long double lrint / llrint implementations completely. Tested for x86_64, x86, mips64 and powerpc. * math/s_nexttowardf.c (__nexttowardf): Use math_narrow_eval. * stdlib/strtod_l.c: Include <math_private.h>. (overflow_value): Use math_narrow_eval. (underflow_value): Likewise. * sysdeps/i386/fpu/s_nexttoward.c (__nexttoward): Likewise. * sysdeps/i386/fpu/s_nexttowardf.c (__nexttowardf): Likewise. * sysdeps/ieee754/dbl-64/e_gamma_r.c (gamma_positive): Likewise. (__ieee754_gamma_r): Likewise. * sysdeps/ieee754/dbl-64/gamma_productf.c (__gamma_productf): Likewise. * sysdeps/ieee754/dbl-64/k_rem_pio2.c (__kernel_rem_pio2): Likewise. * sysdeps/ieee754/dbl-64/lgamma_neg.c (__lgamma_neg): Likewise. * sysdeps/ieee754/dbl-64/s_erf.c (__erfc): Likewise. * sysdeps/ieee754/dbl-64/s_llrint.c (__llrint): Likewise. * sysdeps/ieee754/dbl-64/s_lrint.c (__lrint): Likewise. * sysdeps/ieee754/flt-32/e_gammaf_r.c (gammaf_positive): Likewise. (__ieee754_gammaf_r): Likewise. * sysdeps/ieee754/flt-32/k_rem_pio2f.c (__kernel_rem_pio2f): Likewise. * sysdeps/ieee754/flt-32/lgamma_negf.c (__lgamma_negf): Likewise. * sysdeps/ieee754/flt-32/s_erff.c (__erfcf): Likewise. * sysdeps/ieee754/flt-32/s_llrintf.c (__llrintf): Likewise. * sysdeps/ieee754/flt-32/s_lrintf.c (__lrintf): Likewise. * sysdeps/ieee754/ldbl-128/s_llrintl.c (__llrintl): Do not use volatile. * sysdeps/ieee754/ldbl-128/s_lrintl.c (__lrintl): Likewise. * sysdeps/ieee754/ldbl-128/s_nexttoward.c (__nexttoward): Use math_narrow_eval. * sysdeps/ieee754/ldbl-128ibm/s_nexttoward.c (__nexttoward): Likewise. * sysdeps/ieee754/ldbl-128ibm/s_nexttowardf.c (__nexttowardf): Likewise. * sysdeps/ieee754/ldbl-96/gamma_product.c (__gamma_product): Likewise. * sysdeps/ieee754/ldbl-96/s_llrintl.c (__llrintl): Do not use volatile. * sysdeps/ieee754/ldbl-96/s_lrintl.c (__lrintl): Likewise. * sysdeps/ieee754/ldbl-96/s_nexttoward.c (__nexttoward): Use math_narrow_eval. * sysdeps/ieee754/ldbl-96/s_nexttowardf.c (__nexttowardf): Likewise. * sysdeps/ieee754/ldbl-opt/s_nexttowardfd.c (__nldbl_nexttowardf): Likewise.
2015-09-23 18:14:57 +00:00
ret = math_narrow_eval (__copysign (DBL_MAX, ret) * DBL_MAX);
Improve tgamma accuracy (bug 18613). In non-default rounding modes, tgamma can be slightly less accurate than permitted by glibc's accuracy goals. Part of the problem is error accumulation, addressed in this patch by setting round-to-nearest for internal computations. However, there was also a bug in the code dealing with computing pow (x + n, x + n) where x + n is not exactly representable, providing another source of error even in round-to-nearest mode; it was necessary to address both bugs to get errors for all testcases within glibc's accuracy goals. Given this second fix, accuracy in round-to-nearest mode is also improved (hence regeneration of ulps for tgamma should be from scratch - truncate libm-test-ulps or at least remove existing tgamma entries - so that the expected ulps can be reduced). Some additional complications also arose. Certain tgamma tests should strictly, according to IEEE semantics, overflow or not depending on the rounding mode; this is beyond the scope of glibc's accuracy goals for any function without exactly-determined results, but gen-auto-libm-tests doesn't handle being lax there as it does for underflow. (libm-test.inc also doesn't handle being lax about whether the result in cases very close to the overflow threshold is infinity or a finite value close to overflow, but that doesn't cause problems in this case though I've seen it cause problems with random test generation for some functions.) Thus, spurious-overflow markings, with a comment, are added to auto-libm-test-in (no bug in Bugzilla because the issue is with the testsuite, not a user-visible bug in glibc). And on x86, after the patch I saw ERANGE issues as previously reported by Carlos (see my commentary in <https://sourceware.org/ml/libc-alpha/2015-01/msg00485.html>), which needed addressing by ensuring excess range and precision were eliminated at various points if FLT_EVAL_METHOD != 0. I also noticed and fixed a cosmetic issue where 1.0f was used in long double functions and should have been 1.0L. This completes the move of all functions to testing in all rounding modes with ALL_RM_TEST, so gen-libm-have-vector-test.sh is updated to remove the workaround for some functions not using ALL_RM_TEST. Tested for x86_64, x86, mips64 and powerpc. [BZ #18613] * sysdeps/ieee754/dbl-64/e_gamma_r.c (gamma_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gamma_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. * sysdeps/ieee754/flt-32/e_gammaf_r.c (gammaf_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammaf_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. * sysdeps/ieee754/ldbl-128/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * sysdeps/ieee754/ldbl-128ibm/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * sysdeps/ieee754/ldbl-96/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * math/libm-test.inc (tgamma_test_data): Remove one test. Moved to auto-libm-test-in. (tgamma_test): Use ALL_RM_TEST. * math/auto-libm-test-in: Add one test of tgamma. Mark some other tests of tgamma with spurious-overflow. * math/auto-libm-test-out: Regenerated. * math/gen-libm-have-vector-test.sh: Do not check for START. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-06-29 23:29:35 +00:00
return ret;
}
Improve tgamma accuracy (bug 18613). In non-default rounding modes, tgamma can be slightly less accurate than permitted by glibc's accuracy goals. Part of the problem is error accumulation, addressed in this patch by setting round-to-nearest for internal computations. However, there was also a bug in the code dealing with computing pow (x + n, x + n) where x + n is not exactly representable, providing another source of error even in round-to-nearest mode; it was necessary to address both bugs to get errors for all testcases within glibc's accuracy goals. Given this second fix, accuracy in round-to-nearest mode is also improved (hence regeneration of ulps for tgamma should be from scratch - truncate libm-test-ulps or at least remove existing tgamma entries - so that the expected ulps can be reduced). Some additional complications also arose. Certain tgamma tests should strictly, according to IEEE semantics, overflow or not depending on the rounding mode; this is beyond the scope of glibc's accuracy goals for any function without exactly-determined results, but gen-auto-libm-tests doesn't handle being lax there as it does for underflow. (libm-test.inc also doesn't handle being lax about whether the result in cases very close to the overflow threshold is infinity or a finite value close to overflow, but that doesn't cause problems in this case though I've seen it cause problems with random test generation for some functions.) Thus, spurious-overflow markings, with a comment, are added to auto-libm-test-in (no bug in Bugzilla because the issue is with the testsuite, not a user-visible bug in glibc). And on x86, after the patch I saw ERANGE issues as previously reported by Carlos (see my commentary in <https://sourceware.org/ml/libc-alpha/2015-01/msg00485.html>), which needed addressing by ensuring excess range and precision were eliminated at various points if FLT_EVAL_METHOD != 0. I also noticed and fixed a cosmetic issue where 1.0f was used in long double functions and should have been 1.0L. This completes the move of all functions to testing in all rounding modes with ALL_RM_TEST, so gen-libm-have-vector-test.sh is updated to remove the workaround for some functions not using ALL_RM_TEST. Tested for x86_64, x86, mips64 and powerpc. [BZ #18613] * sysdeps/ieee754/dbl-64/e_gamma_r.c (gamma_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gamma_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. * sysdeps/ieee754/flt-32/e_gammaf_r.c (gammaf_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammaf_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. * sysdeps/ieee754/ldbl-128/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * sysdeps/ieee754/ldbl-128ibm/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * sysdeps/ieee754/ldbl-96/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * math/libm-test.inc (tgamma_test_data): Remove one test. Moved to auto-libm-test-in. (tgamma_test): Use ALL_RM_TEST. * math/auto-libm-test-in: Add one test of tgamma. Mark some other tests of tgamma with spurious-overflow. * math/auto-libm-test-out: Regenerated. * math/gen-libm-have-vector-test.sh: Do not check for START. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-06-29 23:29:35 +00:00
else if (ret == 0)
{
Improve tgamma accuracy (bug 18613). In non-default rounding modes, tgamma can be slightly less accurate than permitted by glibc's accuracy goals. Part of the problem is error accumulation, addressed in this patch by setting round-to-nearest for internal computations. However, there was also a bug in the code dealing with computing pow (x + n, x + n) where x + n is not exactly representable, providing another source of error even in round-to-nearest mode; it was necessary to address both bugs to get errors for all testcases within glibc's accuracy goals. Given this second fix, accuracy in round-to-nearest mode is also improved (hence regeneration of ulps for tgamma should be from scratch - truncate libm-test-ulps or at least remove existing tgamma entries - so that the expected ulps can be reduced). Some additional complications also arose. Certain tgamma tests should strictly, according to IEEE semantics, overflow or not depending on the rounding mode; this is beyond the scope of glibc's accuracy goals for any function without exactly-determined results, but gen-auto-libm-tests doesn't handle being lax there as it does for underflow. (libm-test.inc also doesn't handle being lax about whether the result in cases very close to the overflow threshold is infinity or a finite value close to overflow, but that doesn't cause problems in this case though I've seen it cause problems with random test generation for some functions.) Thus, spurious-overflow markings, with a comment, are added to auto-libm-test-in (no bug in Bugzilla because the issue is with the testsuite, not a user-visible bug in glibc). And on x86, after the patch I saw ERANGE issues as previously reported by Carlos (see my commentary in <https://sourceware.org/ml/libc-alpha/2015-01/msg00485.html>), which needed addressing by ensuring excess range and precision were eliminated at various points if FLT_EVAL_METHOD != 0. I also noticed and fixed a cosmetic issue where 1.0f was used in long double functions and should have been 1.0L. This completes the move of all functions to testing in all rounding modes with ALL_RM_TEST, so gen-libm-have-vector-test.sh is updated to remove the workaround for some functions not using ALL_RM_TEST. Tested for x86_64, x86, mips64 and powerpc. [BZ #18613] * sysdeps/ieee754/dbl-64/e_gamma_r.c (gamma_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gamma_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. * sysdeps/ieee754/flt-32/e_gammaf_r.c (gammaf_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammaf_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. * sysdeps/ieee754/ldbl-128/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * sysdeps/ieee754/ldbl-128ibm/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * sysdeps/ieee754/ldbl-96/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * math/libm-test.inc (tgamma_test_data): Remove one test. Moved to auto-libm-test-in. (tgamma_test): Use ALL_RM_TEST. * math/auto-libm-test-in: Add one test of tgamma. Mark some other tests of tgamma with spurious-overflow. * math/auto-libm-test-out: Regenerated. * math/gen-libm-have-vector-test.sh: Do not check for START. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-06-29 23:29:35 +00:00
if (*signgamp < 0)
{
Use math_narrow_eval more consistently. Where glibc code needs to avoid excess range and precision in floating-point arithmetic, code variously uses either asms or volatile to force the results of that arithmetic to memory; mostly this is conditional on FLT_EVAL_METHOD, but in the case of lrint / llrint functions some use of volatile is unconditional (and is present unnecessarily in versions for long double). This patch make such code use the recently-added math_narrow_eval macro consistently, removing the unnecessary uses of volatile in long double lrint / llrint implementations completely. Tested for x86_64, x86, mips64 and powerpc. * math/s_nexttowardf.c (__nexttowardf): Use math_narrow_eval. * stdlib/strtod_l.c: Include <math_private.h>. (overflow_value): Use math_narrow_eval. (underflow_value): Likewise. * sysdeps/i386/fpu/s_nexttoward.c (__nexttoward): Likewise. * sysdeps/i386/fpu/s_nexttowardf.c (__nexttowardf): Likewise. * sysdeps/ieee754/dbl-64/e_gamma_r.c (gamma_positive): Likewise. (__ieee754_gamma_r): Likewise. * sysdeps/ieee754/dbl-64/gamma_productf.c (__gamma_productf): Likewise. * sysdeps/ieee754/dbl-64/k_rem_pio2.c (__kernel_rem_pio2): Likewise. * sysdeps/ieee754/dbl-64/lgamma_neg.c (__lgamma_neg): Likewise. * sysdeps/ieee754/dbl-64/s_erf.c (__erfc): Likewise. * sysdeps/ieee754/dbl-64/s_llrint.c (__llrint): Likewise. * sysdeps/ieee754/dbl-64/s_lrint.c (__lrint): Likewise. * sysdeps/ieee754/flt-32/e_gammaf_r.c (gammaf_positive): Likewise. (__ieee754_gammaf_r): Likewise. * sysdeps/ieee754/flt-32/k_rem_pio2f.c (__kernel_rem_pio2f): Likewise. * sysdeps/ieee754/flt-32/lgamma_negf.c (__lgamma_negf): Likewise. * sysdeps/ieee754/flt-32/s_erff.c (__erfcf): Likewise. * sysdeps/ieee754/flt-32/s_llrintf.c (__llrintf): Likewise. * sysdeps/ieee754/flt-32/s_lrintf.c (__lrintf): Likewise. * sysdeps/ieee754/ldbl-128/s_llrintl.c (__llrintl): Do not use volatile. * sysdeps/ieee754/ldbl-128/s_lrintl.c (__lrintl): Likewise. * sysdeps/ieee754/ldbl-128/s_nexttoward.c (__nexttoward): Use math_narrow_eval. * sysdeps/ieee754/ldbl-128ibm/s_nexttoward.c (__nexttoward): Likewise. * sysdeps/ieee754/ldbl-128ibm/s_nexttowardf.c (__nexttowardf): Likewise. * sysdeps/ieee754/ldbl-96/gamma_product.c (__gamma_product): Likewise. * sysdeps/ieee754/ldbl-96/s_llrintl.c (__llrintl): Do not use volatile. * sysdeps/ieee754/ldbl-96/s_lrintl.c (__lrintl): Likewise. * sysdeps/ieee754/ldbl-96/s_nexttoward.c (__nexttoward): Use math_narrow_eval. * sysdeps/ieee754/ldbl-96/s_nexttowardf.c (__nexttowardf): Likewise. * sysdeps/ieee754/ldbl-opt/s_nexttowardfd.c (__nldbl_nexttowardf): Likewise.
2015-09-23 18:14:57 +00:00
ret = math_narrow_eval (-__copysign (DBL_MIN, ret) * DBL_MIN);
Improve tgamma accuracy (bug 18613). In non-default rounding modes, tgamma can be slightly less accurate than permitted by glibc's accuracy goals. Part of the problem is error accumulation, addressed in this patch by setting round-to-nearest for internal computations. However, there was also a bug in the code dealing with computing pow (x + n, x + n) where x + n is not exactly representable, providing another source of error even in round-to-nearest mode; it was necessary to address both bugs to get errors for all testcases within glibc's accuracy goals. Given this second fix, accuracy in round-to-nearest mode is also improved (hence regeneration of ulps for tgamma should be from scratch - truncate libm-test-ulps or at least remove existing tgamma entries - so that the expected ulps can be reduced). Some additional complications also arose. Certain tgamma tests should strictly, according to IEEE semantics, overflow or not depending on the rounding mode; this is beyond the scope of glibc's accuracy goals for any function without exactly-determined results, but gen-auto-libm-tests doesn't handle being lax there as it does for underflow. (libm-test.inc also doesn't handle being lax about whether the result in cases very close to the overflow threshold is infinity or a finite value close to overflow, but that doesn't cause problems in this case though I've seen it cause problems with random test generation for some functions.) Thus, spurious-overflow markings, with a comment, are added to auto-libm-test-in (no bug in Bugzilla because the issue is with the testsuite, not a user-visible bug in glibc). And on x86, after the patch I saw ERANGE issues as previously reported by Carlos (see my commentary in <https://sourceware.org/ml/libc-alpha/2015-01/msg00485.html>), which needed addressing by ensuring excess range and precision were eliminated at various points if FLT_EVAL_METHOD != 0. I also noticed and fixed a cosmetic issue where 1.0f was used in long double functions and should have been 1.0L. This completes the move of all functions to testing in all rounding modes with ALL_RM_TEST, so gen-libm-have-vector-test.sh is updated to remove the workaround for some functions not using ALL_RM_TEST. Tested for x86_64, x86, mips64 and powerpc. [BZ #18613] * sysdeps/ieee754/dbl-64/e_gamma_r.c (gamma_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gamma_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. * sysdeps/ieee754/flt-32/e_gammaf_r.c (gammaf_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammaf_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. * sysdeps/ieee754/ldbl-128/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * sysdeps/ieee754/ldbl-128ibm/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * sysdeps/ieee754/ldbl-96/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * math/libm-test.inc (tgamma_test_data): Remove one test. Moved to auto-libm-test-in. (tgamma_test): Use ALL_RM_TEST. * math/auto-libm-test-in: Add one test of tgamma. Mark some other tests of tgamma with spurious-overflow. * math/auto-libm-test-out: Regenerated. * math/gen-libm-have-vector-test.sh: Do not check for START. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-06-29 23:29:35 +00:00
ret = -ret;
}
else
Use math_narrow_eval more consistently. Where glibc code needs to avoid excess range and precision in floating-point arithmetic, code variously uses either asms or volatile to force the results of that arithmetic to memory; mostly this is conditional on FLT_EVAL_METHOD, but in the case of lrint / llrint functions some use of volatile is unconditional (and is present unnecessarily in versions for long double). This patch make such code use the recently-added math_narrow_eval macro consistently, removing the unnecessary uses of volatile in long double lrint / llrint implementations completely. Tested for x86_64, x86, mips64 and powerpc. * math/s_nexttowardf.c (__nexttowardf): Use math_narrow_eval. * stdlib/strtod_l.c: Include <math_private.h>. (overflow_value): Use math_narrow_eval. (underflow_value): Likewise. * sysdeps/i386/fpu/s_nexttoward.c (__nexttoward): Likewise. * sysdeps/i386/fpu/s_nexttowardf.c (__nexttowardf): Likewise. * sysdeps/ieee754/dbl-64/e_gamma_r.c (gamma_positive): Likewise. (__ieee754_gamma_r): Likewise. * sysdeps/ieee754/dbl-64/gamma_productf.c (__gamma_productf): Likewise. * sysdeps/ieee754/dbl-64/k_rem_pio2.c (__kernel_rem_pio2): Likewise. * sysdeps/ieee754/dbl-64/lgamma_neg.c (__lgamma_neg): Likewise. * sysdeps/ieee754/dbl-64/s_erf.c (__erfc): Likewise. * sysdeps/ieee754/dbl-64/s_llrint.c (__llrint): Likewise. * sysdeps/ieee754/dbl-64/s_lrint.c (__lrint): Likewise. * sysdeps/ieee754/flt-32/e_gammaf_r.c (gammaf_positive): Likewise. (__ieee754_gammaf_r): Likewise. * sysdeps/ieee754/flt-32/k_rem_pio2f.c (__kernel_rem_pio2f): Likewise. * sysdeps/ieee754/flt-32/lgamma_negf.c (__lgamma_negf): Likewise. * sysdeps/ieee754/flt-32/s_erff.c (__erfcf): Likewise. * sysdeps/ieee754/flt-32/s_llrintf.c (__llrintf): Likewise. * sysdeps/ieee754/flt-32/s_lrintf.c (__lrintf): Likewise. * sysdeps/ieee754/ldbl-128/s_llrintl.c (__llrintl): Do not use volatile. * sysdeps/ieee754/ldbl-128/s_lrintl.c (__lrintl): Likewise. * sysdeps/ieee754/ldbl-128/s_nexttoward.c (__nexttoward): Use math_narrow_eval. * sysdeps/ieee754/ldbl-128ibm/s_nexttoward.c (__nexttoward): Likewise. * sysdeps/ieee754/ldbl-128ibm/s_nexttowardf.c (__nexttowardf): Likewise. * sysdeps/ieee754/ldbl-96/gamma_product.c (__gamma_product): Likewise. * sysdeps/ieee754/ldbl-96/s_llrintl.c (__llrintl): Do not use volatile. * sysdeps/ieee754/ldbl-96/s_lrintl.c (__lrintl): Likewise. * sysdeps/ieee754/ldbl-96/s_nexttoward.c (__nexttoward): Use math_narrow_eval. * sysdeps/ieee754/ldbl-96/s_nexttowardf.c (__nexttowardf): Likewise. * sysdeps/ieee754/ldbl-opt/s_nexttowardfd.c (__nldbl_nexttowardf): Likewise.
2015-09-23 18:14:57 +00:00
ret = math_narrow_eval (__copysign (DBL_MIN, ret) * DBL_MIN);
Improve tgamma accuracy (bug 18613). In non-default rounding modes, tgamma can be slightly less accurate than permitted by glibc's accuracy goals. Part of the problem is error accumulation, addressed in this patch by setting round-to-nearest for internal computations. However, there was also a bug in the code dealing with computing pow (x + n, x + n) where x + n is not exactly representable, providing another source of error even in round-to-nearest mode; it was necessary to address both bugs to get errors for all testcases within glibc's accuracy goals. Given this second fix, accuracy in round-to-nearest mode is also improved (hence regeneration of ulps for tgamma should be from scratch - truncate libm-test-ulps or at least remove existing tgamma entries - so that the expected ulps can be reduced). Some additional complications also arose. Certain tgamma tests should strictly, according to IEEE semantics, overflow or not depending on the rounding mode; this is beyond the scope of glibc's accuracy goals for any function without exactly-determined results, but gen-auto-libm-tests doesn't handle being lax there as it does for underflow. (libm-test.inc also doesn't handle being lax about whether the result in cases very close to the overflow threshold is infinity or a finite value close to overflow, but that doesn't cause problems in this case though I've seen it cause problems with random test generation for some functions.) Thus, spurious-overflow markings, with a comment, are added to auto-libm-test-in (no bug in Bugzilla because the issue is with the testsuite, not a user-visible bug in glibc). And on x86, after the patch I saw ERANGE issues as previously reported by Carlos (see my commentary in <https://sourceware.org/ml/libc-alpha/2015-01/msg00485.html>), which needed addressing by ensuring excess range and precision were eliminated at various points if FLT_EVAL_METHOD != 0. I also noticed and fixed a cosmetic issue where 1.0f was used in long double functions and should have been 1.0L. This completes the move of all functions to testing in all rounding modes with ALL_RM_TEST, so gen-libm-have-vector-test.sh is updated to remove the workaround for some functions not using ALL_RM_TEST. Tested for x86_64, x86, mips64 and powerpc. [BZ #18613] * sysdeps/ieee754/dbl-64/e_gamma_r.c (gamma_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gamma_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. * sysdeps/ieee754/flt-32/e_gammaf_r.c (gammaf_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammaf_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. * sysdeps/ieee754/ldbl-128/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * sysdeps/ieee754/ldbl-128ibm/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * sysdeps/ieee754/ldbl-96/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * math/libm-test.inc (tgamma_test_data): Remove one test. Moved to auto-libm-test-in. (tgamma_test): Use ALL_RM_TEST. * math/auto-libm-test-in: Add one test of tgamma. Mark some other tests of tgamma with spurious-overflow. * math/auto-libm-test-out: Regenerated. * math/gen-libm-have-vector-test.sh: Do not check for START. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-06-29 23:29:35 +00:00
return ret;
}
Improve tgamma accuracy (bug 18613). In non-default rounding modes, tgamma can be slightly less accurate than permitted by glibc's accuracy goals. Part of the problem is error accumulation, addressed in this patch by setting round-to-nearest for internal computations. However, there was also a bug in the code dealing with computing pow (x + n, x + n) where x + n is not exactly representable, providing another source of error even in round-to-nearest mode; it was necessary to address both bugs to get errors for all testcases within glibc's accuracy goals. Given this second fix, accuracy in round-to-nearest mode is also improved (hence regeneration of ulps for tgamma should be from scratch - truncate libm-test-ulps or at least remove existing tgamma entries - so that the expected ulps can be reduced). Some additional complications also arose. Certain tgamma tests should strictly, according to IEEE semantics, overflow or not depending on the rounding mode; this is beyond the scope of glibc's accuracy goals for any function without exactly-determined results, but gen-auto-libm-tests doesn't handle being lax there as it does for underflow. (libm-test.inc also doesn't handle being lax about whether the result in cases very close to the overflow threshold is infinity or a finite value close to overflow, but that doesn't cause problems in this case though I've seen it cause problems with random test generation for some functions.) Thus, spurious-overflow markings, with a comment, are added to auto-libm-test-in (no bug in Bugzilla because the issue is with the testsuite, not a user-visible bug in glibc). And on x86, after the patch I saw ERANGE issues as previously reported by Carlos (see my commentary in <https://sourceware.org/ml/libc-alpha/2015-01/msg00485.html>), which needed addressing by ensuring excess range and precision were eliminated at various points if FLT_EVAL_METHOD != 0. I also noticed and fixed a cosmetic issue where 1.0f was used in long double functions and should have been 1.0L. This completes the move of all functions to testing in all rounding modes with ALL_RM_TEST, so gen-libm-have-vector-test.sh is updated to remove the workaround for some functions not using ALL_RM_TEST. Tested for x86_64, x86, mips64 and powerpc. [BZ #18613] * sysdeps/ieee754/dbl-64/e_gamma_r.c (gamma_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gamma_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. * sysdeps/ieee754/flt-32/e_gammaf_r.c (gammaf_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammaf_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. * sysdeps/ieee754/ldbl-128/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * sysdeps/ieee754/ldbl-128ibm/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * sysdeps/ieee754/ldbl-96/e_gammal_r.c (gammal_positive): Take log of X_ADJ not X when adjusting exponent. (__ieee754_gammal_r): Do intermediate computations in round-to-nearest then adjust overflowing and underflowing results as needed. Use 1.0L not 1.0f as numerator of division. * math/libm-test.inc (tgamma_test_data): Remove one test. Moved to auto-libm-test-in. (tgamma_test): Use ALL_RM_TEST. * math/auto-libm-test-in: Add one test of tgamma. Mark some other tests of tgamma with spurious-overflow. * math/auto-libm-test-out: Regenerated. * math/gen-libm-have-vector-test.sh: Do not check for START. * sysdeps/i386/fpu/libm-test-ulps: Update. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
2015-06-29 23:29:35 +00:00
else
return ret;
Update. 1997-06-19 19:38 Ulrich Drepper <drepper@cygnus.com> * features.h: Define __STDC_IEC_559__ and _STDC_IEC_559_COMPLEX__. * elf/dl-minimal.c (__dcgettext): Remove assertion. * inet/rcmd.c: Correct a few typos. Reported by Eric Troan. * manual/Makefile (distribute): Add dir. * manual/dir: New file. * math/Makefile (libm-support): Rename s_rinttol, s_rinttoll, s_roundtol, and s_roundtoll to s_lrint, s_llrint, s_lround, and s_llround respectively. (libm-calls): Add e_gamma_r. * math/libm-test.c (check_int_exc): New function. (signbit_test): Remove test for sign of NaN. (gamma_test): Clear exception after test of existence. Correct tests and and epsilons. (lgamma_test): Likewise. (ilogb_test): Correct all tests. (scalb_test): Rewrite. (rinttol_test): Rename to lrint_test and correct tests. (rinttoll_test): Likewise. (roundtol_test): Likewise. (roundtoll_test): Likewise. (main): Call lrint/lround functions instead of rinttol/roundtol. * math/math.h: Change prototypes for rinttol/roundtol. * math/mathcalls.h: Rearrange prototypes according to ISO C9X draft. * sysdeps/generic/mathbits.h: Define FP_ILOGB0 and FP_ILOGBNAN. * sysdeps/i386/mathbits.h: Likewise. * sysdeps/libm-i387/e_scalb.S: Handle special cases correctly. * sysdeps/libm-i387/e_scalbf.S: Likewise. * sysdeps/libm-i387/e_scalbl.S: Likewise. * sysdeps/libm-i387/s_asinh.S: Handle -inf correctly. * sysdeps/libm-i387/s_asinhf.S: Likewise. * sysdeps/libm-i387/s_asinhl.S: Likewise. * sysdeps/libm-i387/s_ilogb.S: Optimize. * sysdeps/libm-i387/s_ilogbf.S: Likewise. * sysdeps/libm-i387/s_ilogbl.S: Likewise. * sysdeps/libm-i387/s_rinttol.S: Rename to... * sysdeps/libm-i387/s_lrint.S: ...this. * sysdeps/libm-i387/s_rinttoll.S: Rename to... * sysdeps/libm-i387/s_llrint.S: ...this. * sysdeps/libm-i387/s_remquo.S: Correctly set sign of remainder. * sysdeps/libm-i387/s_remquof.S: Likewise. * sysdeps/libm-i387/s_remquol.S: Likewise. * sysdeps/libm-i387/e_gamma_r.c: New file. Implementation of gamma function according to ISO C. * sysdeps/libm-i387/e_gammaf_r.c: New file. * sysdeps/libm-i387/e_gammal_r.c: New file. * sysdeps/libm-i387/e_lgamma_r.c: Don't let optimize compile the generation of exceptions away. * sysdeps/libm-i387/e_lgammaf_r.c: Likewise. * sysdeps/libm-i387/k_standard.c: Correct return value for infinity points of gamma function when not SVID mode. * sysdeps/libm-i387/s_rinttoll.c: Renamed to... * sysdeps/libm-i387/s_llrint.c: ...this. * sysdeps/libm-i387/s_rinttol.c: Renamed to... * sysdeps/libm-i387/s_lrint.c: ...this. * sysdeps/libm-i387/s_roundtoll.c: Renamed to... * sysdeps/libm-i387/s_llround.c: ...this. * sysdeps/libm-i387/s_roundtol.c: Renamed to... * sysdeps/libm-i387/s_lround.c: ..this. * sysdeps/libm-i387/s_scalbn.c: Change second parameter according to ISO C. * sysdeps/libm-i387/s_scalbnf.c: Likewise. * sysdeps/libm-i387/s_scalbnl.c: Likewise. * sysdeps/libm-i387/w_gamma.c: Call __ieee754_gamma_r if library mode is not _SVID_. * sysdeps/libm-i387/w_gammaf.c: Likewise. * sysdeps/libm-i387/w_gammal.c: Likewise. * sysdeps/m68k/fpu/__math.h: Rename __rinttol to __lrint and rinttol to lrint. * sysdeps/m68k/fpu/s_rinttol.c: Renamed to... * sysdeps/m68k/fpu/s_lrint.c: ...this. * sysdeps/m68k/fpu/s_rinttoll.c: Renamed to... * sysdeps/m68k/fpu/s_llrint.c: ...this. * md5-crypt/Makefile: Link md5test program with md5.o. * stdio-common/temptest.c: Don't use __stdio_gen_tempname which is not exported by the libc.so. * stdio-common/vfscanf.c: Correct scanning of strings after last change. * sysdeps/unix/sysv/linux/i386/sysdep.S: Use .comm to define errno. 1997-06-19 07:37 H.J. Lu <hjl@gnu.ai.mit.edu> * time/tzfile.c (__tzfile_read): Store getc () return in int. 1997-06-13 Miles Bader <miles@gnu.ai.mit.edu> * argp/argp-parse.c (argp_version_parser): Include `(PROGRAM ERROR)' in the no-version error text to indicate that something's fucked. [!_] (N_): New macro. (argp_default_options, argp_version_options): Wrap doc strings in N_(). 1997-06-12 Miles Bader <miles@gnu.ai.mit.edu> * argp/argp-parse.c (parser_parse_opt): Detect and report unhandled options here. (parser_parse_arg): Handle ARGP_KEY_ARGS here. Adjust NEXT pointer back if we fail to parse anything. (parser_parse_next): Simplify arg code. Leave state NEXT frobbing to parser_parse_arg. 1997-06-11 Miles Bader <miles@gnu.ai.mit.edu> * argp/argp.h (ARGP_KEY_ARGS, ARGP_KEY_FINI): New macros. * argp/argp-parse.c (parser_finalize): Do another pass over the parsers with ARGP_KEY_FINI. 1997-06-18 Miles Bader <miles@gnu.ai.mit.edu> * string/Makefile (routines): Add argz-replace. 1997-06-16 00:16 Miles Bader <miles@gnu.ai.mit.edu> * manual/string.texi (Argz Functions): Document argz_replace. 1997-06-12 Miles Bader <miles@gnu.ai.mit.edu> * string/argz.h (__argz_replace, argz_replace): New declarations. * string/argz-replace.c: New file. 1997-06-11 Miles Bader <miles@gnu.ai.mit.edu> * manual/argp.texi (Argp Special Keys): Document ARGP_KEY_ARGS. 1997-06-16 23:08 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * Makeconfig (libc-map): Remove definition. * Makerules (libc-map): Define it here, using the full name. (load-map-file): Remove case for empty $(..). ($(common-objpfx)libc.so): Revert last change. 1997-06-17 22:18 Mark Kettenis <kettenis@phys.uva.nl> * login/programs/utmpd.c (main): Improve signal handling. * login/programs/request.c (do_setutent, do_updwtmp): Allow arbitrary length filenames. * login/programs/utmpd.h (setutent_request, updwtmp_request): Get rid of fixed length file field. * login/utmp_daemon.c (do_setutent, do_updwtmp): Allow arbitrary length filenames. * login/programs/request.c (do_pututline): Don't fail if connection->position is -1 on entry. 1997-06-15 16:32 Mark Kettenis <kettenis@phys.uva.nl> * login/utmp_file.c (updwtmp_file): Use the same method for appending an entry as in pututline_file. 1997-06-11 18:59 Mark Kettenis <kettenis@phys.uva.nl> * login/utmp_file.c (getutent_r_file): Use read lock instead of write lock. (getutline_r_file, internal_getut_r): Lock utmp file. (updwtmp_file): Use fcntl to lock file instead of flock. 1997-06-18 00:11 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * stdio-common/vfscanf.c (inchar, ungetc): Don't count EOF as character read in/put back. * stdio-common/tstscanf.c: Add test case for this. 1997-06-17 22:17 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * libc.map: Add more symbols. 1997-06-18 12:01 Ulrich Drepper <drepper@cygnus.com> * manual/Makefile (install): Make sure `dir' file exists if we use install-info. 1997-06-17 19:32 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * manual/Makefile (info): Depend on dir-add.info. 1997-06-17 17:19 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * elf/ldd.bash.in: Fix spacing in message. 1997-06-17 14:28 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * elf/do-rel.h (elf_dynamic_do_rel): Always use version information if available. 1997-06-17 11:34 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sunrpc/Makefile ($(rpcsvc:%.x=$(objpfx)rpcsvc/%.h)): Make command non-empty to force make to recheck modification time. ($(rpcsvc:%.x=$(objpfx)x%.c)): Likewise. 1997-06-17 00:26 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/stub/e_acoshl.c: Set errno to ENOSYS. * sysdeps/stub/e_acosl.c: Likewise. * sysdeps/stub/e_asinl.c: Likewise. * sysdeps/stub/e_atan2l.c: Likewise. * sysdeps/stub/e_expl.c: Likewise. * sysdeps/stub/e_fmodl.c: Likewise. * sysdeps/stub/e_j0l.c: Likewise. * sysdeps/stub/e_j1l.c: Likewise. * sysdeps/stub/e_jnl.c: Likewise. * sysdeps/stub/e_lgammal_r.c: Likewise. * sysdeps/stub/e_log10l.c: Likewise. * sysdeps/stub/e_logl.c: Likewise. * sysdeps/stub/e_powl.c: Likewise. * sysdeps/stub/e_rem_pio2l.c: Likewise. * sysdeps/stub/e_sqrtl.c: Likewise. * sysdeps/stub/k_cosl.c: Likewise. * sysdeps/stub/k_rem_pio2l.c: Likewise. * sysdeps/stub/k_sinl.c: Likewise. * sysdeps/stub/k_tanl.c: Likewise. * sysdeps/stub/s_atanl.c: Likewise. * sysdeps/stub/s_erfl.c: Likewise. * sysdeps/stub/s_exp2.c: Likewise. * sysdeps/stub/s_exp2f.c: Likewise. * sysdeps/stub/s_exp2l.c: Likewise. * sysdeps/stub/s_expm1l.c: Likewise. * sysdeps/stub/s_log1pl.c: Likewise. * sysdeps/stub/s_log2l.c: Likewise. 1997-06-18 11:46 Ulrich Drepper <drepper@cygnus.com> * sysdeps/libm-ieee754/w_gamma.c: If _LIB_VERSION is _SVID_ compute result as before last change. * sysdeps/libm-ieee754/w_gammaf.c: Likewise. * sysdeps/libm-ieee754/w_gammal.c: Likewise. 1997-06-16 23:37 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/libm-ieee754/s_remquo.c: Fix off-by-one when computing quotient. * sysdeps/libm-ieee754/s_remquof.c: Likewise. * sysdeps/libm-ieee754/s_remquol.c: Likewise. * sysdeps/m68k/fpu/s_remquo.c: Remove FIXME and special case for quotient. * sysdeps/libm-ieee754/w_gamma.c: Add missing call to exp function. Don't use global signgam. * sysdeps/libm-ieee754/w_gammaf.c: Likewise. * sysdeps/libm-ieee754/w_gammal.c: Likewise. * math/Makefile (libm-calls): Remove w_gamma_r. * sysdeps/libm-ieee754/w_gamma_r.c: Remove file. * sysdeps/libm-ieee754/w_gammaf_r.c: Remove file. * sysdeps/libm-ieee754/w_gammal_r.c: Remove file. * math/libm-test.c (atanh_test): Declare x only if needed. (signbit_test): Fix typo. (gamma_test): Check whether function is implemented. Add epsilons. (lgamma_test): Likewise. (fmod_test): Add epsilons. (exp2_test): Use right function for existence test. 1997-06-07 09:20 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/powerpc/Dist: Add fenv_const.c. * sysdeps/unix/sysv/linux/Dist: Add net/if_slip.h. * sysdeps/unix/sysv/linux/powerpc/Dist: Add init-first.h and syscall.h. * sysdeps/unix/sysv/linux/sparc/Dist: Add init-first.h. * string/Makefile (distribute): Add tst-svc.expect. * nis/Makefile (distribute): Add nis_intern.h and Banner. * elf/Makefile (distribute): Add dl-hash.h. * Rules (subdir_echo-distinfo): Add headers from $(distribute). * login/Makefile (others): Add utmpdump. * sysdeps/unix/sysv/linux/Makefile (sysdep_headers): Add net/if_slip.h. * manual/Makefile (dir-add.texi): Also look in indirectly included files. 1997-06-16 23:15 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/unix/sysv/linux/configure.in: Remove commands to generate stdio_lim.h. * sysdeps/unix/sysv/linux/mk-stdiolim.c: Remove. * sysdeps/unix/sysv/linux/Makefile: Add rules to generate stdio_lim.h here. (common-generated): Add generated files. (inhibit-stdio_lim): Define. * sysdeps/posix/Makefile [$(inhibit-stdio_lim)=yes]: Disable rules to generate stdio_lim.h. * sysdeps/unix/sysv/linux/Makefile: Suppress inclusion of dependecy files if no_deps is set. ($(objpfx)syscall-%.d): Add header file as target to dependency generation. 1997-06-14 19:19 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/unix/sysv/linux/arm/Dist: New file. 1997-06-14 17:59 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * extra-lib.mk (others): Depend on versioned shared library, not the unversioned one. * Makerules (build-shlib): Don't make the version link here. ($(common-objpfx)libc.so$(libc.so-version)): New rule for libc version link. 1997-06-16 03:07 Ulrich Drepper <drepper@cygnus.com> * sysdeps/libm-ieee754/k_standard.c: Undo change of Tue Aug 6 01:13:56 1996. * argp/argp-help.c (argp_args_usage): Supply correct argp to filter_doc. * argp/argp-help.c (hol_add_cluster): Initialize CL->depth. * argp/argp-help.c (_help): Supply STATE to argp_args_usage. * argp/argp.h (ARGP_KEY_HELP_ARGS_DOC): New macro. * argp/argp-fmtstream.c: Add casts to prevent warnings. * argp/argp.h (OPTION_NO_USAGE): New macro. * argp/argp-help.c (usage_long_opt, usage_argful_short_opt, * argp/argp-fmtstream.c (__argp_fmtstream_update): Account for case * argp/argp-help.c <stddef.h>: New include. * argp/argp.h (argp_state_help, __argp_state_help, argp_usage, * argp/argp.h (argp_program_bug_address): Make const. * argp/argp-parse.c (argp_default_parser): Set STATE->name for OPT_PROGNAME. * argp/argp-help.c (__argp_error, __argp_failure, __argp_state_help): * argp/argp-parse.c (parser_init): Set PARSER->state.flags.
1997-06-19 19:11:22 +00:00
}
strong_alias (__ieee754_gamma_r, __gamma_r_finite)