* sysdeps/ieee754/ldbl-96/e_lgammal_r.c: New file.
Contributed by Stephen L. Moshier <moshier@na-net.ornl.gov>.
* sysdeps/ieee754/ldbl-96/e_gammal_r.c: Fix handling of boundary cases.
* sysdeps/ieee754/dbl-64/e_gamma_r.c: Always initialize sign variable.
* sysdeps/ieee754/flt-32/e_gammaf_r.c: Likewise.
* sysdeps/i386/fpu/libm-test-ulps: Adjust after addition of lgammal.
* sysdeps/ieee754/ldbl-96/e_jnl.c: New file.
Contributed by Stephen L. Moshier <moshier@na-net.ornl.gov>.
* sysdeps/i386/fpu/libm-test-ulps: Update for jnl and ynl introduction.
* sysdeps/ia64/fpu/libm-test-ulps: Likewise.
* sysdeps/ieee754/ldbl-96/e_j1l.c: New file.
Contributed by Stephen L. Moshier <moshier@na-net.ornl.gov>.
* sysdeps/i386/fpu/libm-test-ulps: Adjust error values for j1 and y1.
* sysdeps/ia64/fpu/libm-test-ulps: Adjust error values for y1.
* math/libm-test.inc (j1_test): Mark constants as long double.
(jn_test): Likewise.
(y1_test): Likewise.
(yn_test): Likewise.
* sysdeps/i386/fpu/libm-test-ulps: Adjust for errors in y0l.
* math/libm-test.inc (j0_test): Make expected results long double
constants.
(y0_test): Likewise.
2000-12-21 H.J. Lu <hjl@gnu.org>
* sysdeps/ia64/fpu/math_ldbl.h (ieee_long_double_shape_type):
Make sign_exponent element signed.
* sysdeps/ieee754/ldbl-96/math_ldbl.h (ieee_long_double_shape_type):
Fix a typo.
* sysdeps/i386/fpu/s_nextafterl.c (__nextafterl): Remove dead branch.
2000-12-20 Ulrich Drepper <drepper@redhat.com>
* math/test-misc.c: Add more tests for nextafter functions.
* sysdeps/i386/fpu/s_nextafterl.c: Handle change from denormal to
normal correctly. Correct test for sign.
Based on a patch by HJ Lu.
* math/test-misc.c: Add more tests for nextafter.
* sysdeps/i386/fpu/s_nextafterl.c: Handle decrement for x<0 correctly.
* sysdeps/ieee754/ldbl-96/math_ldbl.h
(ieee_long_double_shape_type): Make sign_exponent element signed.
* sysdeps/i386/fpu/e_scalb.S: Handle NaN as first parameter correctly.
* sysdeps/i386/fpu/e_scalbf.S: Likewise.
* sysdeps/i386/fpu/e_scalbl.S: Likewise.
* math/w_scalb.c: Don't use matherr except in SVID mode.
* math/w_scalbf.c: Likewise.
* math/w_scalbl.c: Likewise.
* math/test-misc.c: Add test for NaN and scalbl.
Reported by Fred J. Tydeman <tydeman@tybor.com>.
2000-12-04 Ulrich Drepper <drepper@redhat.com>
* sysdeps/i386/fpu/s_nextafterl.c (__nextafterl): Hopefully fix
the handling of denormalized numbers.
* math/test-misc.c (main): Test for fpclassify and pseudo denormal
numbers on x86.
* sysdeps/i386/fpu/fpclassifyl.c: New file.
2000-12-03 Ulrich Drepper <drepper@redhat.com>
* math/test-misc.c (main): Add tests for frexp.
Reported by Fred J. Tydeman <tydeman@tybor.com>.
* sysdeps/i386/fpu/s_frexpl.S: Don't overflow during the computation.
2000-11-28 Ulrich Drepper <drepper@redhat.com>
* sysdeps/i386/fpu/libm-test-ulps: Adjust some values for the
crappy i486 FPU.
Reported by Michael Deutschmann <michael@talamasca.ocis.net>.
building with the default options.
2000-10-13 Geoffrey Keating <geoffk@cygnus.com>
* sysdeps/i386/fpu/libm-test-ulps: Update ULPs for my Pentium III
building with the default options.
* stdlib/Makefile (headers): Add bits/wchar.h.
* sysdeps/unix/sysv/linux/i386/bits/wchar.h: New file.
* sysdeps/generic/bits/wchar.h: New file.
* sysdeps/generic/stdint.h: Include <bits/wchar.h>. Use values in this
file to define WCHAR_MIN and WCHAR_MAX.
Change UINT*8_MAX and UIN*16_MAX value to type int.
* wcsmbs/wchar.h: Include <bits/wchar.h>. Use values in this
file to define WCHAR_MIN and WCHAR_MAX.
* sysdeps/alpha/fpu/bits/mathdef.h: Make FP_ILOGB0 and FP_ILOGBNAN
int values.
* sysdeps/arm/fpu/bits/mathdef.h: Likewise.
* sysdeps/generic/bits/mathdef.h: Likewise.
* sysdeps/i386/fpu/bits/mathdef.h: Likewise.
* sysdeps/m68k/fpu/bits/mathdef.h: Likewise.
* sysdeps/powerpc/fpu/bits/mathdef.h: Likewise.
* sysdeps/sparc/fpu/bits/mathdef.h: Likewise.
* locale/locale.h (struct lconv): Add __ prefix to the new members
if not C99.
* sysdeps/generic/inttypes.h: Remove unnecessary imaxabs and
imaxdiv optimizations.
* sysdeps/wordsize-32/lldiv.c: Add hack to make alias work.
* sysdeps/wordsize-64/ldiv.c: Likewise.
* sysdeps/alpha/fpu/bits/fenv.h: Change type of FE_*_ENV macros to
const fenv_t*.
* sysdeps/generic/bits/fenv.h: Likewise.
* sysdeps/i386/fpu/bits/fenv.h: Likewise.
* sysdeps/ia64/bits/fenv.h: Likewise.
* sysdeps/m68k/fpu/bits/fenv.h: Likewise.
* sysdeps/mips/bits/fenv.h: Likewise.
* sysdeps/s390/fpu/bits/fenv.h: Likewise.
* sysdeps/sparc/fpu/bits/fenv.h: Likewise.
Reported by Joseph S. Myers <jsm28@cam.ac.uk>.
* catgets/open_catalog.c: Likewise.
2000-08-01 Mark Kettenis <kettenis@gnu.org>
* sysdeps/unix/sysv/linux/i386/sys/elf.h (elf_greg_t, ELF_NGREG,
elf_gregset_t, elf_fpregset_t, elf_fpxregset_t): Moved to
sysdeps/unix/sysv/linux/i386/sys/procfs.h.
Remove the junk that's only relevant in the Linux kernel itself.
Add a warning saying that this file is now obsolete, and include
<sys/procfs.h>.
* sysdeps/unix/sysv/linux/i386/sys/procfs.h (elf_greg_t,
ELF_NGREG, elf_gregset_t, elf_fpregset_t, elf_fpxregset_t): Moved
here from sysdeps/unix/sysv/linux/i386/sys/elf.h. Don't include
<signal.h>, <sys/ucontext.h> and <sys/elf.h>. Remove parts of
`struct elf_prstatus' that are #ifdef'ed out. Improve comments
and add the notice that the file is primarily there for GDB.
* sysdeps/unix/sysv/linux/i386/sys/user.h: Fix comment.
2000-08-09 Jakub Jelinek <jakub@redhat.com>
* time/tzfile.c (__tzfile_compute): __tzstring zones
from zone_names.
2000-08-09 Jakub Jelinek <jakub@redhat.com>
* resolv/resolv.h (__res_state): Surround prototype with
__BEGIN_DECLS/__END_DECLS.
* sysdeps/unix/sysv/linux/m68k/chown.c: Include kernel-features.h.
2000-08-04 Jes Sorensen <jes@linuxcare.com>
* sysdeps/unix/sysv/linux/ia64/bits/resource.h (__rlimit_resource):
Make independant of asm/resource.h from the Linux kernel.
2000-08-06 Bruno Haible <haible@clisp.cons.org>
* iconv/gconv_conf.c (__gconv_read_conf): Don't call realpath.
2000-08-06 Bruno Haible <haible@clisp.cons.org>
* intl/Makefile (CPPFLAGS): Remove $(i18ndir) from LOCALE_ALIAS_PATH.
* locale/Makefile (locale-CPPFLAGS): Likewise.
2000-08-09 Ulrich Drepper <drepper@redhat.com>
* sysdeps/i386/fpu/libm-test-ulps: Increase several error values
for AMD processors.
2000-08-03 Jakub Jelinek <jakub@redhat.com>
* misc/Versions (mincore): Export at GLIBC_2.2.
2000-08-03 Bruno Haible <haible@clisp.cons.org>
* intl/libintl.h (gettext, dgettext, dcgettext): Add back __THROW.
2000-06-01 Jakub Jelinek <jakub@redhat.com>
* libc/sysdeps/alpha/fpu/bits/mathinline.h: Add __THROW to all
inlines to match prototypes in mathcalls.h.
* libc/sysdeps/i386/fpu/bits/mathinline.h: Likewise.
* libc/sysdeps/m68k/fpu/bits/mathinline.h: Likewise.
* libc/sysdeps/powerpc/fpu/bits/mathinline.h: Likewise.
* libc/sysdeps/sparc/fpu/bits/mathinline.h: Likewise.
* sysdeps/powerpc/dl-start.S: Close open comment.
2000-05-22 Andreas Jaeger <aj@suse.de>
* math/Makefile: Add -D__FAST_MATH__ to CFLAGS-test-ifloat.c,
CFLAGS-test-idouble.c, CFLAGS-test-ildoubl.c.
* manual/math.texi (FP Function Optimizations): Document gcc
-ffast-math behaviour with mathinlines.
* sysdeps/i386/fpu/bits/mathinline.h: Only use save inline
functions unless -ffast-math is given to gcc.
2000-03-07 Andreas Jaeger <aj@suse.de>
* resolv/netdb.h: The descriptions of rresvport and rresvport_af
were swapped.
Patch by YOSHIFUJI Hideaki <yoshfuji@ecei.tohoku.ac.jp>, closes PR
libc/1636.
* sysdeps/i386/fpu/bits/mathinline.h: Fix union definition error
in __sgn1l, otherwise g++ fails to parse this.
Reported by Sean Chen <sean.chen@turbolinux.com>.
2000-02-15 Andreas Jaeger <aj@suse.de>
* sysdeps/i386/fpu/libm-test-ulps: Add one value for AMD K7.
2000-02-15 Andreas Jaeger <aj@suse.de>
* math/libm-test.inc (nearbyint_test): Add some tests.
(rint_test): Likewise.
2000-02-14 Ulrich Drepper <drepper@redhat.com>
* sysdeps/posix/sigwait.c: Don't look for signal 0 since it does
not exist. Patch by Miloslav Trmac <mitr@volny.cz> (PR libc/1601).
* sysdeps/i386/fpu/s_nearbyint.S: Disable raising exception, not
enable it. Patch by Miloslav Trmac <mitr@volny.cz> (PR libc/1600).
* sysdeps/i386/fpu/s_nearbyintf.S: Likewise.
* sysdeps/i386/fpu/s_nearbyintf.l: Likewise.
* sysdeps/i386/fpu/s_cosf.S: Domain of opcode is not large enough
so test for overflow and handle it.
* sysdeps/i386/fpu/s_sinf.S: Likewise.
* sysdeps/i386/fpu/s_tanf.S: Likewise.
Patch by Miloslav Trmac <mitr@volny.cz> (PR libc/1563).
