+1997-04-02 16:28 Ulrich Drepper <drepper@cygnus.com>
+
+ * grp/fgetgrent.c: Don't use fixed buffer length. Allow dynamic
+ adjustment if the underlying function say it is too small.
+ * pwd/fgetpwent.c: Likewise.
+ * shadow/fgetspent.c: Likewise.
+ * shadow/sgetspent.c: Likewise.
+
+ * grp/fgetgrent_r.c (__fgetgrent_r): If `parse_result' returns -1
+ this means there was not enough room for the gr_mem array.
+
+ * nss/nss_files/files-XXX.c: Leave loop and return error if
+ `parse_result' returns -1.
+
+ * nss/nss_files/files-parse.c (parse_line): Return -1 if parsing
+ of trailing list fails because of low memory.
+ (parse_list): Rewrite. The logic was completly incorrect since
+ the checks for overflowing the array were not executed when
+ needed.
+
+ * shadow/sgetspent_r.c: Only a return value > 0 from parse_line
+ signals success.
+
+ * nss/getXXbyYY.c: Make sure NULL is returned if we ran out of memory.
+ * nss/getXXent.c: Likewise.
+
+ * stdlib/strtod.c (SET_MANTISSA): Make sure that at least one bit
+ is set in mantissa.
+ * stdlib/strtof.c: Likewise.
+ * stdlib/strtold.c: Likewise.
+ * wcsmbs/wcstof.c: Likewise.
+ * wcsmbs/wcstold.c: Likewise.
+
+ * sysdeps/libm-ieee754/s_nan.c: Use strtod is parameter is not empty
+ string.
+ * sysdeps/libm-ieee754/s_nanf.c: Likewise.
+ * sysdeps/libm-ieee754/s_nanld.c: Likewise.
+
+1997-04-02 13:56 Ulrich Drepper <drepper@cygnus.com>
+
+ * manual/errno.texi: Add description of strerror_r.
+
+1997-04-01 19:07 H.J. Lu (hjl@gnu.ai.mit.edu)
+
+ * manual/errno.texi: Add ENOMEDIUM and EMEDIUMTYPE.
+
+1997-04-02 13:55 Ulrich Drepper <drepper@cygnus.com>
+
+ * sysdeps/gnu/errlist.awk: The Texinfo file is named errno.texi.
+
+1997-03-30 10:59 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de>
+
+ * sysdeps/m68k/fpu/s_cexp.c: New file.
+ * sysdeps/m68k/fpu/s_cexpf.c: New file.
+ * sysdeps/m68k/fpu/s_cexpl.c: New file.
+ * sysdeps/m68k/fpu/s_nearbyint.c: New file.
+ * sysdeps/m68k/fpu/s_nearbyintf.c: New file.
+ * sysdeps/m68k/fpu/s_nearbyintl.c: New file.
+ * sysdeps/m68k/fpu/s_trunc.c: New file.
+ * sysdeps/m68k/fpu/s_truncf.c: New file.
+ * sysdeps/m68k/fpu/s_truncl.c: New file.
+
+ * sysdeps/m68k/fpu/s_modf.c: Rewritten.
+ * sysdeps/m68k/fpu/s_modff.c: Adapted.
+ * sysdeps/m68k/fpu/s_modfl.c: Likewise.
+
+ * sysdeps/m68k/fpu/e_atan2.c: Use __signbit inline.
+ * sysdeps/m68k/fpu/e_pow.c: Likewise. Write 0.0/0.0 for NaN. Fix
+ typo in test for `y = 0.5'.
+
+ * sysdeps/m68k/fpu/__math.h: Define __trunc, __signbit,
+ __nearbyint. Remove __modf.
+
+ * sysdeps/m68k/Makefile (CFLAGS-s_copysignl.c) [$(subdir)=math]:
+ Define for gcc bug workaround.
+
+ * math/libm-test.c (log2_test): Add epsilon for long double to
+ "log2 (e)" and "log2 (16)".
+ (modf_test): Fix test name.
+ (sqrt_test): Add missing FUNC.
+
+1997-03-29 16:24 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de>
+
+ * sunrpc/Makefile (CFLAGS-xrquota.c): Fix typo in variable name.
+
+1997-03-28 20:56 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de>
+
+ * stdlib/stdlib.h: Declare __clearenv.
+
+1997-03-28 18:03 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de>
+
+ * sysdeps/libm-ieee754/s_asinhl.c: Fix Ulrich Dreppers mangling of
+ the last change.
+
+1997-03-28 17:55 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de>
+
+ * configure.in: Call AC_CANONICAL_BUILD before referencing
+ $build. Fix test for BUILD_CC.
+
+1997-04-02 01:29 Philip Blundell <pjb27@cam.ac.uk>
+
+ * sysdeps/libm-ieee754/s_log2f.c (__log2f): Fix typo.
+
+1997-03-31 18:53 Philip Blundell <pjb27@cam.ac.uk>
+
+ * sysdeps/unix/sysv/linux/net/route.h: Add new IPv6 definitions.
+
+1997-04-01 14:08 Ulrich Drepper <drepper@cygnus.com>
+
+ * sysdeps/i386/memset.c: Fix typo.
+ * sysdeps/stub/fchmod.c: Fix typo.
+ * sysdeps/stub/fchdir.c: Fix typo.
+ Reported by Fila Kolodny <fila@ibi.com>.
+
1997-04-01 00:37 Ulrich Drepper <drepper@cygnus.com>
* math/Makefile: Prevent inlining for test-{float,double,longdouble}.
[Q21] ``On Linux I've got problems with the declarations in Linux
kernel headers.''
+
+[Q22] ``When I try to compile code which uses IPv6 header and
+ definitions on my Linux 2.x.y system I am in trouble.
+ Nothing seems to work.''
+
+[Q23] ``When compiling GNU libc I get lots of errors saying functions
+ in glibc are duplicated in libgcc.''
\f
~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~
[Q1] ``What systems does the GNU C Library run on?''
~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~
+[Q22] ``When I try to compile code which uses IPv6 header and
+ definitions on my Linux 2.x.y system I am in trouble.
+ Nothing seems to work.''
+
+[A22] {UD} The problem is that the IPv6 development still has not reached
+a point where it is stable. There are still lots of incompatible changes
+made and the libc headers have to follow.
+
+Currently (as of 970401) according to Philip Blundell <pjb27@cam.ac.uk>
+the required kernel version is 2.1.30.
+
+
+~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~
+[Q23] ``When compiling GNU libc I get lots of errors saying functions
+ in glibc are duplicated in libgcc.''
+
+[A23] {EY} This is *exactly* the same problem that I was having. The
+problem was due to the fact that the autoconfigure didn't correctly
+detect that linker flag --no-whole-archive was supported in my linker.
+In my case it was because I had run ./configure with bogus CFLAGS, and
+the test failed.
+
+One thing that is particularly annoying about this problem is that
+once this is misdetected, running configure again won't fix it unless
+you first delete config.cache.
+
+{UD} Starting with glibc-2.0.3 there should be a better test to avoid
+some problems of this kind. The setting of CFLAGS is checked at the
+very beginning and if it is not usable `configure' will bark.
+
+~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~
\f
Answers were given by:
{UD} Ulrich Drepper, <drepper@cygnus.com>
{RM} Roland McGrath, <roland@gnu.ai.mit.edu>
{HJL} H.J. Lu, <hjl@gnu.ai.mit.edu>
{AJ} Andreas Jaeger, <aj@arthur.pfalz.de>
+{EY} Eric Youngdale, <eric@andante.jic.com>
\f
Local Variables:
mode:text
test "${CFLAGS+set}" = set || CFLAGS="-g"
fi
+echo $ac_n "checking build system type""... $ac_c" 1>&6
+echo "configure:1295: checking build system type" >&5
+
+build_alias=$build
+case "$build_alias" in
+NONE)
+ case $nonopt in
+ NONE) build_alias=$host_alias ;;
+ *) build_alias=$nonopt ;;
+ esac ;;
+esac
+
+build=`$ac_config_sub $build_alias`
+build_cpu=`echo $build | sed 's/^\([^-]*\)-\([^-]*\)-\(.*\)$/\1/'`
+build_vendor=`echo $build | sed 's/^\([^-]*\)-\([^-]*\)-\(.*\)$/\2/'`
+build_os=`echo $build | sed 's/^\([^-]*\)-\([^-]*\)-\(.*\)$/\3/'`
+echo "$ac_t""$build" 1>&6
+
if test $host != $build; then
- # Extract the first word of "gcc cc", so it can be a program name with args.
-set dummy gcc cc; ac_word=$2
+ for ac_prog in gcc cc
+do
+# Extract the first word of "$ac_prog", so it can be a program name with args.
+set dummy $ac_prog; ac_word=$2
echo $ac_n "checking for $ac_word""... $ac_c" 1>&6
-echo "configure:1298: checking for $ac_word" >&5
+echo "configure:1318: checking for $ac_word" >&5
if eval "test \"`echo '$''{'ac_cv_prog_BUILD_CC'+set}'`\" = set"; then
echo $ac_n "(cached) $ac_c" 1>&6
else
for ac_dir in $PATH; do
test -z "$ac_dir" && ac_dir=.
if test -f $ac_dir/$ac_word; then
- ac_cv_prog_BUILD_CC=""
+ ac_cv_prog_BUILD_CC="$ac_prog"
break
fi
done
echo "$ac_t""no" 1>&6
fi
+test -n "$BUILD_CC" && break
+done
+
fi
echo $ac_n "checking how to run the C preprocessor""... $ac_c" 1>&6
-echo "configure:1325: checking how to run the C preprocessor" >&5
+echo "configure:1348: checking how to run the C preprocessor" >&5
# On Suns, sometimes $CPP names a directory.
if test -n "$CPP" && test -d "$CPP"; then
CPP=
# On the NeXT, cc -E runs the code through the compiler's parser,
# not just through cpp.
cat > conftest.$ac_ext <<EOF
-#line 1340 "configure"
+#line 1363 "configure"
#include "confdefs.h"
#include <assert.h>
Syntax Error
EOF
ac_try="$ac_cpp conftest.$ac_ext >/dev/null 2>conftest.out"
-{ (eval echo configure:1346: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }
+{ (eval echo configure:1369: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }
ac_err=`grep -v '^ *+' conftest.out`
if test -z "$ac_err"; then
:
rm -rf conftest*
CPP="${CC-cc} -E -traditional-cpp"
cat > conftest.$ac_ext <<EOF
-#line 1357 "configure"
+#line 1380 "configure"
#include "confdefs.h"
#include <assert.h>
Syntax Error
EOF
ac_try="$ac_cpp conftest.$ac_ext >/dev/null 2>conftest.out"
-{ (eval echo configure:1363: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }
+{ (eval echo configure:1386: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }
ac_err=`grep -v '^ *+' conftest.out`
if test -z "$ac_err"; then
:
fi
echo "$ac_t""$CPP" 1>&6
-echo $ac_n "checking build system type""... $ac_c" 1>&6
-echo "configure:1386: checking build system type" >&5
-
-build_alias=$build
-case "$build_alias" in
-NONE)
- case $nonopt in
- NONE) build_alias=$host_alias ;;
- *) build_alias=$nonopt ;;
- esac ;;
-esac
-
-build=`$ac_config_sub $build_alias`
-build_cpu=`echo $build | sed 's/^\([^-]*\)-\([^-]*\)-\(.*\)$/\1/'`
-build_vendor=`echo $build | sed 's/^\([^-]*\)-\([^-]*\)-\(.*\)$/\2/'`
-build_os=`echo $build | sed 's/^\([^-]*\)-\([^-]*\)-\(.*\)$/\3/'`
-echo "$ac_t""$build" 1>&6
-
if test $host != $build; then
ac_tool_prefix=${host_alias}-
else
# Extract the first word of "${ac_tool_prefix}ar", so it can be a program name with args.
set dummy ${ac_tool_prefix}ar; ac_word=$2
echo $ac_n "checking for $ac_word""... $ac_c" 1>&6
-echo "configure:1412: checking for $ac_word" >&5
+echo "configure:1417: checking for $ac_word" >&5
if eval "test \"`echo '$''{'ac_cv_prog_AR'+set}'`\" = set"; then
echo $ac_n "(cached) $ac_c" 1>&6
else
# Extract the first word of "${ac_tool_prefix}ranlib", so it can be a program name with args.
set dummy ${ac_tool_prefix}ranlib; ac_word=$2
echo $ac_n "checking for $ac_word""... $ac_c" 1>&6
-echo "configure:1443: checking for $ac_word" >&5
+echo "configure:1448: checking for $ac_word" >&5
if eval "test \"`echo '$''{'ac_cv_prog_RANLIB'+set}'`\" = set"; then
echo $ac_n "(cached) $ac_c" 1>&6
else
# Extract the first word of "ranlib", so it can be a program name with args.
set dummy ranlib; ac_word=$2
echo $ac_n "checking for $ac_word""... $ac_c" 1>&6
-echo "configure:1474: checking for $ac_word" >&5
+echo "configure:1479: checking for $ac_word" >&5
if eval "test \"`echo '$''{'ac_cv_prog_RANLIB'+set}'`\" = set"; then
echo $ac_n "(cached) $ac_c" 1>&6
else
# Extract the first word of "bash", so it can be a program name with args.
set dummy bash; ac_word=$2
echo $ac_n "checking for $ac_word""... $ac_c" 1>&6
-echo "configure:1509: checking for $ac_word" >&5
+echo "configure:1514: checking for $ac_word" >&5
if eval "test \"`echo '$''{'ac_cv_path_BASH'+set}'`\" = set"; then
echo $ac_n "(cached) $ac_c" 1>&6
else
# Extract the first word of "ksh", so it can be a program name with args.
set dummy ksh; ac_word=$2
echo $ac_n "checking for $ac_word""... $ac_c" 1>&6
-echo "configure:1550: checking for $ac_word" >&5
+echo "configure:1555: checking for $ac_word" >&5
if eval "test \"`echo '$''{'ac_cv_path_KSH'+set}'`\" = set"; then
echo $ac_n "(cached) $ac_c" 1>&6
else
echo $ac_n "checking for signed size_t type""... $ac_c" 1>&6
-echo "configure:1592: checking for signed size_t type" >&5
+echo "configure:1597: checking for signed size_t type" >&5
if eval "test \"`echo '$''{'libc_cv_signed_size_t'+set}'`\" = set"; then
echo $ac_n "(cached) $ac_c" 1>&6
else
fi
echo $ac_n "checking for libc-friendly stddef.h""... $ac_c" 1>&6
-echo "configure:1616: checking for libc-friendly stddef.h" >&5
+echo "configure:1621: checking for libc-friendly stddef.h" >&5
if eval "test \"`echo '$''{'libc_cv_friendly_stddef'+set}'`\" = set"; then
echo $ac_n "(cached) $ac_c" 1>&6
else
cat > conftest.$ac_ext <<EOF
-#line 1621 "configure"
+#line 1626 "configure"
#include "confdefs.h"
#define __need_size_t
#define __need_wchar_t
if (&size == NULL || &wchar == NULL) abort ();
; return 0; }
EOF
-if { (eval echo configure:1636: \"$ac_compile\") 1>&5; (eval $ac_compile) 2>&5; }; then
+if { (eval echo configure:1641: \"$ac_compile\") 1>&5; (eval $ac_compile) 2>&5; }; then
rm -rf conftest*
libc_cv_friendly_stddef=yes
else
fi
echo $ac_n "checking whether we need to use -P to assemble .S files""... $ac_c" 1>&6
-echo "configure:1655: checking whether we need to use -P to assemble .S files" >&5
+echo "configure:1660: checking whether we need to use -P to assemble .S files" >&5
if eval "test \"`echo '$''{'libc_cv_need_minus_P'+set}'`\" = set"; then
echo $ac_n "(cached) $ac_c" 1>&6
else
fi
echo $ac_n "checking for assembler global-symbol directive""... $ac_c" 1>&6
-echo "configure:1678: checking for assembler global-symbol directive" >&5
+echo "configure:1683: checking for assembler global-symbol directive" >&5
if eval "test \"`echo '$''{'libc_cv_asm_global_directive'+set}'`\" = set"; then
echo $ac_n "(cached) $ac_c" 1>&6
else
fi
echo $ac_n "checking for .set assembler directive""... $ac_c" 1>&6
-echo "configure:1708: checking for .set assembler directive" >&5
+echo "configure:1713: checking for .set assembler directive" >&5
if eval "test \"`echo '$''{'libc_cv_asm_set_directive'+set}'`\" = set"; then
echo $ac_n "(cached) $ac_c" 1>&6
else
fi
echo $ac_n "checking for .symver assembler directive""... $ac_c" 1>&6
-echo "configure:1742: checking for .symver assembler directive" >&5
+echo "configure:1747: checking for .symver assembler directive" >&5
if eval "test \"`echo '$''{'libc_cv_asm_symver_directive'+set}'`\" = set"; then
echo $ac_n "(cached) $ac_c" 1>&6
else
echo "$ac_t""$libc_cv_asm_symver_directive" 1>&6
echo $ac_n "checking for ld --version-script""... $ac_c" 1>&6
-echo "configure:1761: checking for ld --version-script" >&5
+echo "configure:1766: checking for ld --version-script" >&5
if eval "test \"`echo '$''{'libc_cv_ld_version_script_option'+set}'`\" = set"; then
echo $ac_n "(cached) $ac_c" 1>&6
else
EOF
if ${CC-cc} -c $CFLAGS $CPPFLAGS conftest.s 1>&5 2>&5; then
if { ac_try='${CC-cc} $CFLAGS --shared -o conftest.so conftest.o
- -Wl,--version-script,conftest.map'; { (eval echo configure:1778: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }; };
+ -Wl,--version-script,conftest.map'; { (eval echo configure:1783: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }; };
then
libc_cv_ld_version_script_option=yes
else
if test $elf = yes; then
echo $ac_n "checking for .previous assembler directive""... $ac_c" 1>&6
-echo "configure:1808: checking for .previous assembler directive" >&5
+echo "configure:1813: checking for .previous assembler directive" >&5
if eval "test \"`echo '$''{'libc_cv_asm_previous_directive'+set}'`\" = set"; then
echo $ac_n "(cached) $ac_c" 1>&6
else
.section foo_section
.previous
EOF
- if { ac_try='${CC-cc} -c $CFLAGS conftest.s 1>&5'; { (eval echo configure:1816: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }; }; then
+ if { ac_try='${CC-cc} -c $CFLAGS conftest.s 1>&5'; { (eval echo configure:1821: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }; }; then
libc_cv_asm_previous_directive=yes
else
libc_cv_asm_previous_directive=no
else
echo $ac_n "checking for .popsection assembler directive""... $ac_c" 1>&6
-echo "configure:1832: checking for .popsection assembler directive" >&5
+echo "configure:1837: checking for .popsection assembler directive" >&5
if eval "test \"`echo '$''{'libc_cv_asm_popsection_directive'+set}'`\" = set"; then
echo $ac_n "(cached) $ac_c" 1>&6
else
.pushsection foo_section
.popsection
EOF
- if { ac_try='${CC-cc} -c $CFLAGS conftest.s 1>&5'; { (eval echo configure:1840: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }; }; then
+ if { ac_try='${CC-cc} -c $CFLAGS conftest.s 1>&5'; { (eval echo configure:1845: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }; }; then
libc_cv_asm_popsection_directive=yes
else
libc_cv_asm_popsection_directive=no
if test $elf != yes; then
echo $ac_n "checking for .init and .fini sections""... $ac_c" 1>&6
-echo "configure:1860: checking for .init and .fini sections" >&5
+echo "configure:1865: checking for .init and .fini sections" >&5
if eval "test \"`echo '$''{'libc_cv_have_initfini'+set}'`\" = set"; then
echo $ac_n "(cached) $ac_c" 1>&6
else
cat > conftest.$ac_ext <<EOF
-#line 1865 "configure"
+#line 1870 "configure"
#include "confdefs.h"
int main() {
asm (".text");
; return 0; }
EOF
-if { (eval echo configure:1874: \"$ac_compile\") 1>&5; (eval $ac_compile) 2>&5; }; then
+if { (eval echo configure:1879: \"$ac_compile\") 1>&5; (eval $ac_compile) 2>&5; }; then
rm -rf conftest*
libc_cv_have_initfini=yes
else
libc_cv_asm_underscores=no
else
echo $ac_n "checking for _ prefix on C symbol names""... $ac_c" 1>&6
-echo "configure:1899: checking for _ prefix on C symbol names" >&5
+echo "configure:1904: checking for _ prefix on C symbol names" >&5
if eval "test \"`echo '$''{'libc_cv_asm_underscores'+set}'`\" = set"; then
echo $ac_n "(cached) $ac_c" 1>&6
else
cat > conftest.$ac_ext <<EOF
-#line 1904 "configure"
+#line 1909 "configure"
#include "confdefs.h"
asm ("_glibc_foobar:");
int main() {
glibc_foobar ();
; return 0; }
EOF
-if { (eval echo configure:1911: \"$ac_link\") 1>&5; (eval $ac_link) 2>&5; } && test -s conftest; then
+if { (eval echo configure:1916: \"$ac_link\") 1>&5; (eval $ac_link) 2>&5; } && test -s conftest; then
rm -rf conftest*
libc_cv_asm_underscores=yes
else
libc_cv_asm_weakext_directive=no
else
echo $ac_n "checking for assembler .weak directive""... $ac_c" 1>&6
-echo "configure:1938: checking for assembler .weak directive" >&5
+echo "configure:1943: checking for assembler .weak directive" >&5
if eval "test \"`echo '$''{'libc_cv_asm_weak_directive'+set}'`\" = set"; then
echo $ac_n "(cached) $ac_c" 1>&6
else
if test $libc_cv_asm_weak_directive = no; then
echo $ac_n "checking for assembler .weakext directive""... $ac_c" 1>&6
-echo "configure:1961: checking for assembler .weakext directive" >&5
+echo "configure:1966: checking for assembler .weakext directive" >&5
if eval "test \"`echo '$''{'libc_cv_asm_weakext_directive'+set}'`\" = set"; then
echo $ac_n "(cached) $ac_c" 1>&6
else
fi
echo $ac_n "checking for ld --no-whole-archive""... $ac_c" 1>&6
-echo "configure:1998: checking for ld --no-whole-archive" >&5
+echo "configure:2003: checking for ld --no-whole-archive" >&5
if eval "test \"`echo '$''{'libc_cv_ld_no_whole_archive'+set}'`\" = set"; then
echo $ac_n "(cached) $ac_c" 1>&6
else
EOF
if { ac_try='${CC-cc} $CFLAGS
-nostdlib -nostartfiles -Wl,--no-whole-archive
- -o conftest conftest.c'; { (eval echo configure:2009: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }; }; then
+ -o conftest conftest.c'; { (eval echo configure:2014: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }; }; then
libc_cv_ld_no_whole_archive=yes
else
libc_cv_ld_no_whole_archive=no
echo "$ac_t""$libc_cv_ld_no_whole_archive" 1>&6
echo $ac_n "checking for gcc -fno-exceptions""... $ac_c" 1>&6
-echo "configure:2020: checking for gcc -fno-exceptions" >&5
+echo "configure:2025: checking for gcc -fno-exceptions" >&5
if eval "test \"`echo '$''{'libc_cv_gcc_no_exceptions'+set}'`\" = set"; then
echo $ac_n "(cached) $ac_c" 1>&6
else
EOF
if { ac_try='${CC-cc} $CFLAGS
-nostdlib -nostartfiles -fno-exceptions
- -o conftest conftest.c'; { (eval echo configure:2031: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }; }; then
+ -o conftest conftest.c'; { (eval echo configure:2036: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }; }; then
libc_cv_gcc_no_exceptions=yes
else
libc_cv_gcc_no_exceptions=no
fi
echo $ac_n "checking OS release for uname""... $ac_c" 1>&6
-echo "configure:2083: checking OS release for uname" >&5
+echo "configure:2088: checking OS release for uname" >&5
if eval "test \"`echo '$''{'libc_cv_uname_release'+set}'`\" = set"; then
echo $ac_n "(cached) $ac_c" 1>&6
else
uname_release="$libc_cv_uname_release"
echo $ac_n "checking OS version for uname""... $ac_c" 1>&6
-echo "configure:2105: checking OS version for uname" >&5
+echo "configure:2110: checking OS version for uname" >&5
if eval "test \"`echo '$''{'libc_cv_uname_version'+set}'`\" = set"; then
echo $ac_n "(cached) $ac_c" 1>&6
else
fi
echo $ac_n "checking stdio selection""... $ac_c" 1>&6
-echo "configure:2127: checking stdio selection" >&5
+echo "configure:2132: checking stdio selection" >&5
case $stdio in
libio) cat >> confdefs.h <<\EOF
s%@LN_S@%$LN_S%g
s%@MSGFMT@%$MSGFMT%g
s%@CC@%$CC%g
-s%@BUILD_CC@%$BUILD_CC%g
-s%@CPP@%$CPP%g
s%@build@%$build%g
s%@build_alias@%$build_alias%g
s%@build_cpu@%$build_cpu%g
s%@build_vendor@%$build_vendor%g
s%@build_os@%$build_os%g
+s%@BUILD_CC@%$BUILD_CC%g
+s%@CPP@%$CPP%g
s%@AR@%$AR%g
s%@RANLIB@%$RANLIB%g
s%@BASH@%$BASH%g
AC_CHECK_PROGS(MSGFMT, msgfmt gmsgfmt, :)
AC_PROG_CC
+AC_CANONICAL_BUILD
if test $host != $build; then
- AC_CHECK_PROG(BUILD_CC, gcc cc)
+ AC_CHECK_PROGS(BUILD_CC, gcc cc)
fi
AC_PROG_CPP
AC_CHECK_TOOL(AR, ar)
write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
Boston, MA 02111-1307, USA. */
+#include <errno.h>
#include <grp.h>
-#include <stdio.h>
+#include <libc-lock.h>
+#include <stdlib.h>
+/* We need to protect the dynamic buffer handling. */
+__libc_lock_define_initialized (static, lock);
+
/* Read one entry from the given stream. */
struct group *
fgetgrent (FILE *stream)
{
- static char buffer[BUFSIZ];
+ static char *buffer;
+ static size_t buffer_size;
static struct group resbuf;
struct group *result;
+ int save;
+
+ /* Get lock. */
+ __libc_lock_lock (lock);
+
+ /* Allocate buffer if not yet available. */
+ if (buffer == NULL)
+ {
+ buffer_size = NSS_BUFLEN_GROUP;
+ buffer = malloc (buffer_size);
+ }
+
+ while (buffer != NULL
+ && __fgetgrent_r (stream, &resbuf, buffer, buffer_size, &result) != 0
+ && errno == ERANGE)
+ {
+ char *new_buf;
+ buffer_size += NSS_BUFLEN_GROUP;
+ new_buf = realloc (buffer, buffer_size);
+ if (new_buf == NULL)
+ {
+ /* We are out of memory. Free the current buffer so that the
+ process gets a chance for a normal termination. */
+ save = errno;
+ free (buffer);
+ __set_errno (save);
+ }
+ buffer = new_buf;
+ }
+
+ if (buffer == NULL)
+ result = NULL;
- if (__fgetgrent_r (stream, &resbuf, buffer, sizeof buffer, &result) != 0)
- return NULL;
+ /* Release lock. Preserve error value. */
+ save = errno;
+ __libc_lock_unlock (lock);
+ __set_errno (save);
return result;
}
-/* Copyright (C) 1991, 1996 Free Software Foundation, Inc.
+/* Copyright (C) 1991, 1996, 1997 Free Software Foundation, Inc.
This file is part of the GNU C Library.
The GNU C Library is free software; you can redistribute it and/or
struct group **result)
{
char *p;
+ int parse_result;
do
{
/* Skip leading blanks. */
while (isspace (*p))
++p;
- } while (*p == '\0' || *p == '#' || /* Ignore empty and comment lines. */
+ } while (*p == '\0' || *p == '#' /* Ignore empty and comment lines. */
/* Parse the line. If it is invalid, loop to
get the next line of the file to parse. */
- ! parse_line (p, resbuf, (void *) buffer, buflen));
+ || ! (parse_result = parse_line (p, resbuf,
+ (void *) buffer, buflen)));
+
+ if (parse_result == -1)
+ {
+ /* The parser ran out of space. */
+ *result = NULL;
+ return errno;
+ }
*result = resbuf;
return 0;
@comment errno ???/121
@end deftypevr
+@comment errno.h
+@comment Linux???: No medium found
+@deftypevr Macro int ENOMEDIUM
+@comment errno ???/???
+@end deftypevr
+
+@comment errno.h
+@comment Linux???: Wrong medium type
+@deftypevr Macro int EMEDIUMTYPE
+@comment errno ???/???
+@end deftypevr
+
@node Error Messages, , Error Codes, Error Reporting
@section Error Messages
The function @code{strerror} is declared in @file{string.h}.
@end deftypefun
+@comment string.h
+@comment GNU
+@deftypefun {char *} strerror_r (int @var{errnum}, char *@var{buf}, size_t @var{n})
+The @code{strerror_r} function works like @code{strerror} but instead of
+returning the error message in a statically allocated buffer shared by
+all threads in the process, it writes the message string in the user
+supplied buffer starting at @var{buf} with the length of @var{n} bytes.
+
+At most @var{n} characters are written (including the NUL byte) so it is
+up to the user to select the buffer large enough.
+
+This function should always be used in multi-threaded programs since
+there is no way to guarantee the string returned by @code{strerror}
+really belongs to the last call of the current thread.
+
+This function @code{strerror_r} is a GNU extension and it is declared in
+@file{string.h}.
+@end deftypefun
+
@comment stdio.h
@comment ISO
@deftypefun void perror (const char *@var{message})
check_isinfp ("log2 (+inf) == +inf", FUNC(log2) (plus_infty));
- check ("log2 (e) == M_LOG2E", FUNC(log2) (M_E), M_LOG2E);
+ check_eps ("log2 (e) == M_LOG2E", FUNC(log2) (M_E), M_LOG2E,
+ CHOOSE (1e-18L, 0, 0));
check ("log2 (2) == 1", FUNC(log2) (2.0), 1);
- check ("log2 (16) == 4", FUNC(log2) (16.0), 4);
+ check_eps ("log2 (16) == 4", FUNC(log2) (16.0), 4, CHOOSE (1e-18L, 0, 0));
check ("log2 (256) == 8", FUNC(log2) (256.0), 8);
}
result = FUNC(modf) (nan_value, &intpart);
check_isnan ("modf (NaN, &x) returns NaN", result);
- check_isnan ("modf (-inf, &x) sets x to NaN", intpart);
+ check_isnan ("modf (NaN, &x) sets x to NaN", intpart);
result = FUNC(modf) (0, &intpart);
check ("modf (0, &x) returns 0", result, 0);
check_isinfp ("sqrt (+inf) == +inf", FUNC(sqrt) (plus_infty));
x = random_value (0, 10000);
- check_ext ("sqrt (x*x) == x", sqrt (x*x), x, x);
+ check_ext ("sqrt (x*x) == x", FUNC(sqrt) (x*x), x, x);
check ("sqrt (4) == 2", FUNC(sqrt) (4), 2);
}
buffer = new_buf;
}
+ if (buffer == NULL)
+ result = NULL;
+
#ifdef HANDLE_DIGITS_DOTS
done:
#endif
-/* Copyright (C) 1996 Free Software Foundation, Inc.
+/* Copyright (C) 1996, 1997 Free Software Foundation, Inc.
This file is part of the GNU C Library.
The GNU C Library is free software; you can redistribute it and/or
static char *buffer;
static size_t buffer_size;
static LOOKUP_TYPE resbuf;
- LOOKUP_TYPE *result = NULL;
+ LOOKUP_TYPE *result;
int save;
/* Get lock. */
buffer = new_buf;
}
+ if (buffer == NULL)
+ result = NULL;
+
/* Release lock. Preserve error value. */
save = errno;
__libc_lock_unlock (lock);
/* Common code for file-based databases in nss_files module.
- Copyright (C) 1996 Free Software Foundation, Inc.
+ Copyright (C) 1996, 1997 Free Software Foundation, Inc.
This file is part of the GNU C Library.
The GNU C Library is free software; you can redistribute it and/or
char *p;
struct parser_data *data = (void *) buffer;
int linebuflen = buffer + buflen - data->linebuffer;
+ int parse_result;
if (buflen < (int) sizeof *data + 1)
{
while (*p == '\0' || *p == '#' /* Ignore empty and comment lines. */
/* Parse the line. If it is invalid, loop to get the next
line of the file to parse. */
- || ! parse_line (p, result, data, buflen));
+ || ! (parse_result = parse_line (p, result, data, buflen)));
/* Filled in RESULT with the next entry from the database file. */
- return NSS_STATUS_SUCCESS;
+ return parse_result == -1 ? NSS_STATUS_TRYAGAIN : NSS_STATUS_SUCCESS;
}
/* Common code for file-based database parsers in nss_files module.
- Copyright (C) 1996 Free Software Foundation, Inc.
+ Copyright (C) 1996, 1997 Free Software Foundation, Inc.
This file is part of the GNU C Library.
The GNU C Library is free software; you can redistribute it and/or
if (list) \
result->TRAILING_LIST_MEMBER = list; \
else \
- return 0; \
+ return -1; /* -1 indicates we ran out of space. */ \
}
static inline char **
if (*line == '\0')
break;
+ /* Skip leading white space. This might not be portable but useful. */
+ while (isspace (*line))
+ ++line;
+
elt = line;
while (1)
{
- if (TRAILING_LIST_SEPARATOR_P (*line))
- {
- *p++ = elt;
- *line = '\0';
- do
- ++line;
- while (isspace (*line));
- elt = line;
- }
- else if (*line == '\0')
+ if (*line == '\0' || TRAILING_LIST_SEPARATOR_P (*line))
{
- /* End of the line. */
+ /* End of the next entry. */
if (line > elt)
- /* Last element. */
+ /* We really found some data. */
*p++ = elt;
- *line = '\0';
+
+ /* Terminate string if necessary. */
+ if (*line != '\0')
+ *line++ = '\0';
break;
}
- else
- ++line;
+ ++line;
}
}
*p = NULL;
write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
Boston, MA 02111-1307, USA. */
+#include <errno.h>
+#include <libc-lock.h>
#include <pwd.h>
-#include <stdio.h>
+#include <stdlib.h>
+/* We need to protect the dynamic buffer handling. */
+__libc_lock_define_initialized (static, lock);
+
/* Read one entry from the given stream. */
struct passwd *
fgetpwent (FILE *stream)
{
- static char buffer[BUFSIZ];
+ static char *buffer;
+ static size_t buffer_size;
static struct passwd resbuf;
struct passwd *result;
+ int save;
+
+ /* Get lock. */
+ __libc_lock_lock (lock);
+
+ /* Allocate buffer if not yet available. */
+ if (buffer == NULL)
+ {
+ buffer_size = NSS_BUFLEN_PASSWD;
+ buffer = malloc (buffer_size);
+ }
+
+ while (buffer != NULL
+ && __fgetpwent_r (stream, &resbuf, buffer, buffer_size, &result) != 0
+ && errno == ERANGE)
+ {
+ char *new_buf;
+ buffer_size += NSS_BUFLEN_PASSWD;
+ new_buf = realloc (buffer, buffer_size);
+ if (new_buf == NULL)
+ {
+ /* We are out of memory. Free the current buffer so that the
+ process gets a chance for a normal termination. */
+ save = errno;
+ free (buffer);
+ __set_errno (save);
+ }
+ buffer = new_buf;
+ }
+
+ if (buffer == NULL)
+ result = NULL;
- if (__fgetpwent_r (stream, &resbuf, buffer, sizeof buffer, &result) != 0)
- return NULL;
+ /* Release lock. Preserve error value. */
+ save = errno;
+ __libc_lock_unlock (lock);
+ __set_errno (save);
return result;
}
write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
Boston, MA 02111-1307, USA. */
+#include <errno.h>
+#include <libc-lock.h>
#include <shadow.h>
-#include <stdio.h>
+#include <stdlib.h>
+/* A reasonable size for a buffer to start with. */
+#define BUFLEN_SPWD 1024
+
+/* We need to protect the dynamic buffer handling. */
+__libc_lock_define_initialized (static, lock);
+
/* Read one shadow entry from the given stream. */
struct spwd *
fgetspent (FILE *stream)
{
- static char buffer[BUFSIZ];
+ static char *buffer;
+ static size_t buffer_size;
static struct spwd resbuf;
struct spwd *result;
+ int save;
+
+ /* Get lock. */
+ __libc_lock_lock (lock);
+
+ /* Allocate buffer if not yet available. */
+ if (buffer == NULL)
+ {
+ buffer_size = BUFLEN_SPWD;
+ buffer = malloc (buffer_size);
+ }
+
+ while (buffer != NULL
+ && __fgetspent_r (stream, &resbuf, buffer, buffer_size, &result) != 0
+ && errno == ERANGE)
+ {
+ char *new_buf;
+ buffer_size += BUFLEN_SPWD;
+ new_buf = realloc (buffer, buffer_size);
+ if (new_buf == NULL)
+ {
+ /* We are out of memory. Free the current buffer so that the
+ process gets a chance for a normal termination. */
+ save = errno;
+ free (buffer);
+ __set_errno (save);
+ }
+ buffer = new_buf;
+ }
+
+ if (buffer == NULL)
+ result = NULL;
+
+ /* Release lock. Preserve error value. */
+ save = errno;
+ __libc_lock_unlock (lock);
+ __set_errno (save);
- return __fgetspent_r (stream, &resbuf, buffer, sizeof buffer, &result)
- ? NULL : result;
+ return result;
}
write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
Boston, MA 02111-1307, USA. */
+#include <errno.h>
+#include <libc-lock.h>
#include <shadow.h>
-#include <stdio.h>
#include <stdlib.h>
-#include <string.h>
-#define MAX(a, b) ((a) > (b) ? (a) : (b))
+/* A reasonable size for a buffer to start with. */
+#define BUFLEN_SPWD 1024
+
+/* We need to protect the dynamic buffer handling. */
+__libc_lock_define_initialized (static, lock);
/* Read one shadow entry from the given stream. */
struct spwd *
sgetspent (const char *string)
{
+ static char *buffer;
+ static size_t buffer_size;
static struct spwd resbuf;
- static int max_size = 0;
- static char *buffer = NULL;
struct spwd *result;
- int len;
+ int save;
+
+ /* Get lock. */
+ __libc_lock_lock (lock);
- len = strlen (string) + 1;
- if (len > max_size)
+ /* Allocate buffer if not yet available. */
+ if (buffer == NULL)
{
- max_size = MAX (128, len + 32);
- buffer = realloc (buffer, max_size);
- if (buffer == NULL)
- return NULL;
+ buffer_size = BUFLEN_SPWD;
+ buffer = malloc (buffer_size);
}
- return __sgetspent_r (string, &resbuf, buffer, max_size, &result)
- ? NULL : result;
+ while (buffer != NULL
+ && __sgetspent_r (string, &resbuf, buffer, buffer_size, &result) != 0
+ && errno == ERANGE)
+ {
+ char *new_buf;
+ buffer_size += BUFLEN_SPWD;
+ new_buf = realloc (buffer, buffer_size);
+ if (new_buf == NULL)
+ {
+ /* We are out of memory. Free the current buffer so that the
+ process gets a chance for a normal termination. */
+ save = errno;
+ free (buffer);
+ __set_errno (save);
+ }
+ buffer = new_buf;
+ }
+
+ if (buffer == NULL)
+ result = NULL;
+
+ /* Release lock. Preserve error value. */
+ save = errno;
+ __libc_lock_unlock (lock);
+ __set_errno (save);
+
+ return result;
}
__sgetspent_r (const char *string, struct spwd *resbuf, char *buffer,
size_t buflen, struct spwd **result)
{
- *result = parse_line (strncpy (buffer, string, buflen), resbuf, NULL, 0)
- ? resbuf : NULL;
+ int parse_result = parse_line (strncpy (buffer, string, buflen),
+ resbuf, NULL, 0);
+ *result = parse_result > 0 ? resbuf : NULL;
return *result == NULL ? errno : 0;
}
/* The `clearenv' was planned to be added to POSIX.1 but probably
never made it. Nevertheless the POSIX.9 standard (POSIX bindings
for Fortran 77) requires this function. */
+extern int __clearenv __P ((void));
extern int clearenv __P ((void));
#endif
# define SET_MANTISSA(flt, mant) \
do { union ieee754_double u; \
u.d = (flt); \
+ if ((mant & 0xfffffffffffffULL) == 0) \
+ mant = 0x8000000000000ULL; \
u.ieee.mantissa0 = ((mant) >> 32) & 0xfffff; \
u.ieee.mantissa1 = (mant) & 0xffffffff; \
} while (0)
expp = cp;
/* Read exponent. */
- if (TOLOWER (c) == (base == 16 ? L_('p') : L_('e')))
+ if ((base == 16 && TOLOWER (c) == L_('p'))
+ || (base != 16 && TOLOWER (c) == L_('e')))
{
int exp_negative = 0;
#define SET_MANTISSA(flt, mant) \
do { union ieee754_float u; \
u.f = (flt); \
+ if ((mant & 0x7fffff) == 0) \
+ mant = 0x400000; \
u.ieee.mantissa = (mant) & 0x7fffff; \
(flt) = u.f; \
} while (0)
#define SET_MANTISSA(flt, mant) \
do { union ieee854_long_double u; \
u.d = (flt); \
- u.ieee.mantissa0 = ((mant) >> 32) & 0x7fffffff; \
+ if ((mant & 0x7fffffffffffffffULL) == 0) \
+ mant = 0x4000000000000000ULL; \
+ u.ieee.mantissa0 = (((mant) >> 32) & 0x7fffffff) | 0x80000000; \
u.ieee.mantissa1 = (mant) & 0xffffffff; \
(flt) = u.d; \
} while (0)
CFLAGS-xrusers.c = -Wno-unused
CFLAGS-xspray.c = -Wno-unused
CFLAGS-xnfs_prot.c = -Wno-unused
-CFLAGS-xrquote.c = -Wno-unused
+CFLAGS-xrquota.c = -Wno-unused
include ../Rules
# not, write to the Free Software Foundation, Inc., 675 Mass Ave,
# Cambridge, MA 02139, USA.
-# errno.texinfo contains lines like:
+# errno.texi contains lines like:
# @comment errno.h
# @comment POSIX.1: Function not implemented
# @deftypevr Macro int ENOSYS
/* */
[EREMOTEIO] = N_("Remote I/O error"),
#endif
+#ifdef ENOMEDIUM
+/* */
+ [ENOMEDIUM] = N_("No medium found"),
+#endif
+#ifdef EMEDIUMTYPE
+/* */
+ [EMEDIUMTYPE] = N_("Wrong medium type"),
+#endif
};
const int _sys_nerr = sizeof _sys_errlist / sizeof _sys_errlist[0];
#ifdef __GNUC__
-vod *
+void *
memset (void *dstpp, int c, size_t len)
{
unsigned long int dstp = (unsigned long int) dstpp;
t = x*x;
w =__log1pl(fabsl(x)+t/(one+__ieee754_sqrtl(one+t)));
}
- if(hx>0x8000) return -w; else return w;
+ if(hx&0x8000) return -w; else return w;
}
weak_alias (__asinhl, asinhl)
if((0x007fffff&(15+ix))<16) { /* |f| < 2**-20 */
if(f==zero) return dk;
R = f*f*((float)0.5-(float)0.33333333333333333*f);
- return dk-(R-f);}
+ return dk-(R-f);
}
s = f/((float)2.0+f);
z = s*s;
write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
Boston, MA 02111-1307, USA. */
-/* XXX The ISO C 9X standard mentions a `n-char-sequence' which is provided
- as the argument to this function but I have no clue what this means.
- Perhaps it is a description of the bits set in the mantissa. */
#include <math.h>
-#ifdef HANDLE_TAGP
-# include <stdlib.h>
-# include <string.h>
-#else
-# include <ieee754.h>
-#endif
+#include <stdlib.h>
+#include <string.h>
+#include <ieee754.h>
double
__nan (const char *tagp)
{
-#ifdef HANDLE_TAGP
- /* If we ever should have use of the TAGP parameter we will use the
- strtod function to analyze it. */
- char buf[6 + strlen (tagp)];
- sprintf (buf, "NAN(%s)", tagp);
- return strtod (buf, NULL);
-#else
static const union ieee754_double nan_value =
- { ieee: { mantissa1: 0x1, mantissa0: 0x0, exponent: 0x7ff, negative: 0 } };
+ { ieee: { mantissa1: 0, mantissa0: 0x80000, exponent: 0x7ff, negative: 0 } };
+
+ if (tagp[0] != '\0')
+ {
+ char buf[6 + strlen (tagp)];
+ sprintf (buf, "NAN(%s)", tagp);
+ return strtod (buf, NULL);
+ }
+
return nan_value.d;
-#endif
}
weak_alias (__nan, nan)
write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
Boston, MA 02111-1307, USA. */
-/* XXX The ISO C 9X standard mentions a `n-char-sequence' which is provided
- as the argument to this function but I have no clue what this means.
- Perhaps it is a description of the bits set in the mantissa. */
#include <math.h>
-#ifdef HANDLE_TAGP
-# include <stdlib.h>
-# include <string.h>
-#else
-# include <ieee754.h>
-#endif
+#include <stdlib.h>
+#include <string.h>
+#include <ieee754.h>
float
__nanf (const char *tagp)
{
-#ifdef HANDLE_TAGP
- /* If we ever should have use of the TAGP parameter we will use the
- strtod function to analyze it. */
- char buf[6 + strlen (tagp)];
- sprintf (buf, "NAN(%s)", tagp);
- return strtof (buf, NULL);
-#else
static const union ieee754_float nan_value =
- { ieee: { mantissa: 0x1, exponent: 0xff, negative: 0 } };
+ { ieee: { mantissa: 0x400000, exponent: 0xff, negative: 0 } };
+
+ if (tagp[0] != '\0')
+ {
+ char buf[6 + strlen (tagp)];
+ sprintf (buf, "NAN(%s)", tagp);
+ return strtof (buf, NULL);
+ }
+
return nan_value.f;
-#endif
}
weak_alias (__nanf, nanf)
write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
Boston, MA 02111-1307, USA. */
-/* XXX The ISO C 9X standard mentions a `n-char-sequence' which is provided
- as the argument to this function but I have no clue what this means.
- Perhaps it is a description of the bits set in the mantissa. */
#include <math.h>
-#ifdef HANDLE_TAGP
-# include <stdlib.h>
-# include <string.h>
-#else
-# include <ieee754.h>
-#endif
+#include <stdlib.h>
+#include <string.h>
+#include <ieee754.h>
long double
__nanl (const char *tagp)
{
-#ifdef HANDLE_TAGP
- /* If we ever should have use of the TAGP parameter we will use the
- strtod function to analyze it. */
- char buf[6 + strlen (tagp)];
- sprintf (buf, "NAN(%s)", tagp);
- return strtold (buf, NULL);
-#else
static const union ieee854_long_double nan_value =
- { ieee: { mantissa1: 0x1, mantissa0: 0x80000000,
+ { ieee: { mantissa1: 0, mantissa0: 0xc0000000,
exponent: 0x7fff, negative: 0 } };
+
+ if (tagp[0] != '\0')
+ {
+ char buf[6 + strlen (tagp)];
+ sprintf (buf, "NAN(%s)", tagp);
+ return strtold (buf, NULL);
+ }
+
return nan_value.d;
-#endif
}
weak_alias (__nanl, nanl)
ifeq ($(subdir),elf)
CFLAGS-rtld.c += -Wno-uninitialized -Wno-unused
endif
+
+ifeq ($(subdir),math)
+# Avoid a bug in gcc
+CFLAGS-s_copysignl.c += -mnobitfield
+endif
__inline_mathop(__log2, log2)
__inline_mathop(__exp2, twotox)
+__inline_mathop(__trunc, intrz)
#if !defined __NO_MATH_INLINES && defined __OPTIMIZE__
#ifdef __USE_ISOC9X
__inline_mathop(log2, log2)
__inline_mathop(exp2, twotox)
+__inline_mathop(trunc, intrz)
#endif
#endif /* !__NO_MATH_INLINES && __OPTIMIZE__ */
/* The rest of the functions are available to the user. */
-#define __inline_functions(float_type, s) \
-__m81_inline float_type \
-__m81_u(__CONCAT(__frexp,s))(float_type __value, int *__expptr) \
-{ \
- float_type __mantissa, __exponent; \
- int __iexponent; \
- if (__value == 0.0) \
- { \
- *__expptr = 0; \
- return __value; \
- } \
- __asm("fgetexp%.x %1, %0" : "=f" (__exponent) : "f" (__value)); \
- __iexponent = (int) __exponent + 1; \
- *__expptr = __iexponent; \
- __asm("fscale%.l %2, %0" : "=f" (__mantissa) \
- : "0" (__value), "dmi" (-__iexponent)); \
- return __mantissa; \
-} \
- \
-__m81_defun (float_type, __CONCAT(__floor,s), (float_type __x)) \
-{ \
- float_type __result; \
- unsigned long int __ctrl_reg; \
- __asm __volatile__ ("fmove%.l %!, %0" : "=dm" (__ctrl_reg)); \
- /* Set rounding towards negative infinity. */ \
- __asm __volatile__ ("fmove%.l %0, %!" : /* No outputs. */ \
- : "dmi" ((__ctrl_reg & ~0x10) | 0x20)); \
- /* Convert X to an integer, using -Inf rounding. */ \
- __asm __volatile__ ("fint%.x %1, %0" : "=f" (__result) : "f" (__x)); \
- /* Restore the previous rounding mode. */ \
- __asm __volatile__ ("fmove%.l %0, %!" : /* No outputs. */ \
- : "dmi" (__ctrl_reg)); \
- return __result; \
-} \
- \
-__m81_defun (float_type, __CONCAT(__ceil,s), (float_type __x)) \
-{ \
- float_type __result; \
- unsigned long int __ctrl_reg; \
- __asm __volatile__ ("fmove%.l %!, %0" : "=dm" (__ctrl_reg)); \
- /* Set rounding towards positive infinity. */ \
- __asm __volatile__ ("fmove%.l %0, %!" : /* No outputs. */ \
- : "dmi" (__ctrl_reg | 0x30)); \
- /* Convert X to an integer, using +Inf rounding. */ \
- __asm __volatile__ ("fint%.x %1, %0" : "=f" (__result) : "f" (__x)); \
- /* Restore the previous rounding mode. */ \
- __asm __volatile__ ("fmove%.l %0, %!" : /* No outputs. */ \
- : "dmi" (__ctrl_reg)); \
- return __result; \
-} \
- \
-__m81_inline float_type \
-__m81_u(__CONCAT(__modf,s))(float_type __value, float_type *__iptr) \
-{ \
- float_type __modf_int; \
- __asm ("fintrz%.x %1, %0" : "=f" (__modf_int) : "f" (__value)); \
- *__iptr = __modf_int; \
- return __value - __modf_int; \
-} \
- \
-__m81_defun (int, __CONCAT(__isinf,s), (float_type __value)) \
-{ \
- /* There is no branch-condition for infinity, \
- so we must extract and examine the condition codes manually. */ \
- unsigned long int __fpsr; \
- __asm("ftst%.x %1\n" \
- "fmove%.l %/fpsr, %0" : "=dm" (__fpsr) : "f" (__value)); \
- return (__fpsr & (2 << 24)) ? (__fpsr & (8 << 24) ? -1 : 1) : 0; \
-} \
- \
-__m81_defun (int, __CONCAT(__isnan,s), (float_type __value)) \
-{ \
- char __result; \
- __asm("ftst%.x %1\n" \
- "fsun %0" : "=dm" (__result) : "f" (__value)); \
- return __result; \
-} \
- \
-__m81_defun (int, __CONCAT(__finite,s), (float_type __value)) \
-{ \
- /* There is no branch-condition for infinity, so we must extract and \
- examine the condition codes manually. */ \
- unsigned long int __fpsr; \
- __asm ("ftst%.x %1\n" \
- "fmove%.l %/fpsr, %0" : "=dm" (__fpsr) : "f" (__value)); \
- return (__fpsr & (3 << 24)) == 0; \
-} \
- \
-__m81_defun (int, __CONCAT(__ilogb,s), (float_type __x)) \
-{ \
- float_type __result; \
- if (__x == 0.0) \
- return 0x80000001; \
- __asm("fgetexp%.x %1, %0" : "=f" (__result) : "f" (__x)); \
- return (int) __result; \
-} \
- \
-__m81_defun (float_type, __CONCAT(__scalbn,s), (float_type __x, int __n)) \
-{ \
- float_type __result; \
- __asm ("fscale%.l %1, %0" : "=f" (__result) : "dmi" (__n), "0" (__x)); \
- return __result; \
+#define __inline_functions(float_type, s) \
+__m81_inline float_type \
+__m81_u(__CONCAT(__frexp,s))(float_type __value, int *__expptr) \
+{ \
+ float_type __mantissa, __exponent; \
+ int __iexponent; \
+ if (__value == 0.0) \
+ { \
+ *__expptr = 0; \
+ return __value; \
+ } \
+ __asm("fgetexp%.x %1, %0" : "=f" (__exponent) : "f" (__value)); \
+ __iexponent = (int) __exponent + 1; \
+ *__expptr = __iexponent; \
+ __asm("fscale%.l %2, %0" : "=f" (__mantissa) \
+ : "0" (__value), "dmi" (-__iexponent)); \
+ return __mantissa; \
+} \
+ \
+__m81_defun (float_type, __CONCAT(__floor,s), (float_type __x)) \
+{ \
+ float_type __result; \
+ unsigned long int __ctrl_reg; \
+ __asm __volatile__ ("fmove%.l %!, %0" : "=dm" (__ctrl_reg)); \
+ /* Set rounding towards negative infinity. */ \
+ __asm __volatile__ ("fmove%.l %0, %!" : /* No outputs. */ \
+ : "dmi" ((__ctrl_reg & ~0x10) | 0x20)); \
+ /* Convert X to an integer, using -Inf rounding. */ \
+ __asm __volatile__ ("fint%.x %1, %0" : "=f" (__result) : "f" (__x)); \
+ /* Restore the previous rounding mode. */ \
+ __asm __volatile__ ("fmove%.l %0, %!" : /* No outputs. */ \
+ : "dmi" (__ctrl_reg)); \
+ return __result; \
+} \
+ \
+__m81_defun (float_type, __CONCAT(__ceil,s), (float_type __x)) \
+{ \
+ float_type __result; \
+ unsigned long int __ctrl_reg; \
+ __asm __volatile__ ("fmove%.l %!, %0" : "=dm" (__ctrl_reg)); \
+ /* Set rounding towards positive infinity. */ \
+ __asm __volatile__ ("fmove%.l %0, %!" : /* No outputs. */ \
+ : "dmi" (__ctrl_reg | 0x30)); \
+ /* Convert X to an integer, using +Inf rounding. */ \
+ __asm __volatile__ ("fint%.x %1, %0" : "=f" (__result) : "f" (__x)); \
+ /* Restore the previous rounding mode. */ \
+ __asm __volatile__ ("fmove%.l %0, %!" : /* No outputs. */ \
+ : "dmi" (__ctrl_reg)); \
+ return __result; \
+} \
+ \
+__m81_defun (int, __CONCAT(__isinf,s), (float_type __value)) \
+{ \
+ /* There is no branch-condition for infinity, \
+ so we must extract and examine the condition codes manually. */ \
+ unsigned long int __fpsr; \
+ __asm("ftst%.x %1\n" \
+ "fmove%.l %/fpsr, %0" : "=dm" (__fpsr) : "f" (__value)); \
+ return (__fpsr & (2 << 24)) ? (__fpsr & (8 << 24) ? -1 : 1) : 0; \
+} \
+ \
+__m81_defun (int, __CONCAT(__isnan,s), (float_type __value)) \
+{ \
+ char __result; \
+ __asm("ftst%.x %1\n" \
+ "fsun %0" : "=dm" (__result) : "f" (__value)); \
+ return __result; \
+} \
+ \
+__m81_defun (int, __CONCAT(__finite,s), (float_type __value)) \
+{ \
+ /* There is no branch-condition for infinity, so we must extract and \
+ examine the condition codes manually. */ \
+ unsigned long int __fpsr; \
+ __asm ("ftst%.x %1\n" \
+ "fmove%.l %/fpsr, %0" : "=dm" (__fpsr) : "f" (__value)); \
+ return (__fpsr & (3 << 24)) == 0; \
+} \
+ \
+__m81_defun (int, __CONCAT(__signbit,s), (float_type __value)) \
+{ \
+ /* There is no branch-condition for the sign bit, so we must extract \
+ and examine the condition codes manually. */ \
+ unsigned long int __fpsr; \
+ __asm ("ftst%.x %1\n" \
+ "fmove%.l %/fpsr, %0" : "=dm" (__fpsr) : "f" (__value)); \
+ return (__fpsr >> 27) & 1; \
+} \
+ \
+__m81_defun (int, __CONCAT(__ilogb,s), (float_type __x)) \
+{ \
+ float_type __result; \
+ if (__x == 0.0) \
+ return 0x80000001; \
+ __asm("fgetexp%.x %1, %0" : "=f" (__result) : "f" (__x)); \
+ return (int) __result; \
+} \
+ \
+__m81_defun (float_type, __CONCAT(__scalbn,s), (float_type __x, int __n)) \
+{ \
+ float_type __result; \
+ __asm ("fscale%.l %1, %0" : "=f" (__result) : "dmi" (__n), "0" (__x)); \
+ return __result; \
+} \
+ \
+__m81_defun (float_type, __CONCAT(__nearbyint,s), (float_type __x)) \
+{ \
+ float_type __result; \
+ unsigned long int __ctrl_reg; \
+ __asm __volatile__ ("fmove%.l %!, %0" : "=dm" (__ctrl_reg)); \
+ /* Temporarily disable the inexact exception. */ \
+ __asm __volatile__ ("fmove%.l %0, %!" : /* No outputs. */ \
+ : "dmi" (__ctrl_reg & ~0x200)); \
+ __asm __volatile__ ("fint%.x %1, %0" : "=f" (__result) : "f" (__x)); \
+ __asm __volatile__ ("fmove%.l %0, %!" : /* No outputs. */ \
+ : "dmi" (__ctrl_reg)); \
+ return __result; \
}
/* This defines the three variants of the inline functions. */
(__value, __expptr))
__inline_forward_c(double,floor, (double __x), (__x))
__inline_forward_c(double,ceil, (double __x), (__x))
-__inline_forward(double,modf, (double __value, double *__iptr),
- (__value, __iptr))
#ifdef __USE_MISC
__inline_forward_c(int,isinf, (double __value), (__value))
__inline_forward_c(int,finite, (double __value), (__value))
#endif
__inline_forward_c(int,ilogb, (double __value), (__value))
#endif
+#ifdef __USE_ISOC9X
+__inline_forward_c(double,nearbyint, (double __value), (__value))
+#endif
#if defined __USE_MISC || defined __USE_ISOC9X
(__value, __expptr))
__inline_forward_c(float,floorf, (float __x), (__x))
__inline_forward_c(float,ceilf, (float __x), (__x))
-__inline_forward(float,modff, (float __value, float *__iptr),
- (__value, __iptr))
#ifdef __USE_MISC
__inline_forward_c(int,isinff, (float __value), (__value))
__inline_forward_c(int,finitef, (float __value), (__value))
__inline_forward_c(int,isnanf, (float __value), (__value))
__inline_forward_c(int,ilogbf, (float __value), (__value))
#endif
+#ifdef __USE_ISOC9X
+__inline_forward_c(float,nearbyintf, (float __value), (__value))
+#endif
__inline_forward(long double,frexpl, (long double __value, int *__expptr),
(__value, __expptr))
__inline_forward_c(long double,floorl, (long double __x), (__x))
__inline_forward_c(long double,ceill, (long double __x), (__x))
-__inline_forward(long double,modfl,
- (long double __value, long double *__iptr),
- (__value, __iptr))
#ifdef __USE_MISC
__inline_forward_c(int,isinfl, (long double __value), (__value))
__inline_forward_c(int,finitel, (long double __value), (__value))
__inline_forward_c(int,isnanl, (long double __value), (__value))
__inline_forward_c(int,ilogbl, (long double __value), (__value))
#endif
+#ifdef __USE_ISOC9X
+__inline_forward_c(long double,nearbyintl, (long double __value), (__value))
+#endif
#endif /* Use misc or ISO C9X */
#define float_type double
#endif
-#define __CONCATX(a,b) __CONCAT(a,b)
+#define CONCATX(a,b) __CONCAT(a,b)
+#define s(name) CONCATX(name,SUFF)
+#define m81(func) __m81_u(s(func))
float_type
-__CONCATX(__ieee754_atan2,SUFF) (float_type y, float_type x)
+s(__ieee754_atan2) (float_type y, float_type x)
{
float_type pi, pi_2, z;
z = x + y;
else if (y == 0)
{
- if (signbit (x))
- z = signbit (y) ? -pi : pi;
+ if (m81(__signbit) (x))
+ z = m81(__signbit) (y) ? -pi : pi;
else
z = y;
}
- else if (__m81_u(__CONCATX(__isinf,SUFF)) (x))
+ else if (m81(__isinf) (x))
{
- if (__m81_u(__CONCATX(__isinf,SUFF)) (y))
+ if (m81(__isinf) (y))
{
float_type pi_4;
- __asm ("fscale%.w %#-1, %0" : "=f" (pi_4) : "0" (pi_2));
+ __asm ("fscale%.w %#-2, %0" : "=f" (pi_4) : "0" (pi));
z = x > 0 ? pi_4 : 3 * pi_4;
}
else
z = x > 0 ? 0 : pi;
- if (signbit (y))
+ if (m81(__signbit) (y))
z = -z;
}
- else if (__m81_u(__CONCATX(__isinf,SUFF)) (y))
+ else if (m81(__isinf) (y))
z = y > 0 ? pi_2 : -pi_2;
else if (x > 0)
{
if (y > 0)
{
if (x > y)
- z = __m81_u(__CONCATX(__atan,SUFF)) (y / x);
+ z = m81(__atan) (y / x);
else
- z = pi_2 - __m81_u(__CONCATX(__atan,SUFF)) (x / y);
+ z = pi_2 - m81(__atan) (x / y);
}
else
{
if (x > -y)
- z = __m81_u(__CONCATX(__atan,SUFF)) (y / x);
+ z = m81(__atan) (y / x);
else
- z = -pi_2 - __m81_u(__CONCATX(__atan,SUFF)) (x / y);
+ z = -pi_2 - m81(__atan) (x / y);
}
}
else
if (y < 0)
{
if (-x > y)
- z = -pi + __m81_u(__CONCATX(__atan,SUFF)) (y / x);
+ z = -pi + m81(__atan) (y / x);
else
- z = -pi_2 - __m81_u(__CONCATX(__atan,SUFF)) (x / y);
+ z = -pi_2 - m81(__atan) (x / y);
}
else
{
if (-x > y)
- z = pi + __m81_u(__CONCATX(__atan,SUFF)) (y / x);
+ z = pi + m81(__atan) (y / x);
else
- z = pi_2 - __m81_u(__CONCATX(__atan,SUFF)) (x / y);
+ z = pi_2 - m81(__atan) (x / y);
}
}
return z;
#define float_type double
#endif
-#define __CONCATX(a,b) __CONCAT(a,b)
+#define CONCATX(a,b) __CONCAT(a,b)
+#define s(name) CONCATX(name,SUFF)
+#define m81(func) __m81_u(s(func))
float_type
-__CONCATX(__ieee754_pow,SUFF) (float_type x, float_type y)
+s(__ieee754_pow) (float_type x, float_type y)
{
float_type z;
float_type ax;
if (x != x || y != y)
return x + y;
- if (__m81_u(__CONCATX(__isinf,SUFF)) (y))
+ if (m81(__isinf) (y))
{
- ax = __CONCATX(fabs,SUFF) (x);
+ ax = s(fabs) (x);
if (ax == 1)
- return y - y;
+ return 0.0/0.0;
if (ax > 1)
return y > 0 ? y : 0;
else
return y < 0 ? -y : 0;
}
- if (__CONCATX(fabs,SUFF) (y) == 1)
+ if (s(fabs) (y) == 1)
return y > 0 ? x : 1 / x;
if (y == 2)
return x * x;
- if (y == 0 && x >= 0)
- return __m81_u(__CONCATX(__ieee754_sqrt,SUFF)) (x);
+ if (y == 0.5 && x >= 0)
+ return m81(__ieee754_sqrt) (x);
if (x == 10.0)
{
return z;
}
- ax = __CONCATX(fabs,SUFF) (x);
- if (__m81_u(__CONCATX(__isinf,SUFF)) (x) || x == 0 || ax == 1)
+ ax = s(fabs) (x);
+ if (m81(__isinf) (x) || x == 0 || ax == 1)
{
z = ax;
if (y < 0)
z = 1 / z;
- if (signbit (x))
+ if (m81(__signbit) (x))
{
- float_type temp = __m81_u (__CONCATX(__rint,SUFF)) (y);
+ float_type temp = m81(__rint) (y);
if (y != temp)
{
if (x == -1)
- z = (z - z) / (z - z);
+ z = 0.0/0.0;
}
else
{
if (x < 0.0)
{
- float_type temp = __m81_u (__CONCATX(__rint,SUFF)) (y);
+ float_type temp = m81(__rint) (y);
if (y == temp)
{
long long i = (long long) y;
- z = (__m81_u(__CONCATX(__ieee754_exp,SUFF))
- (y * __m81_u(__CONCATX(__ieee754_log,SUFF)) (-x)));
+ z = m81(__ieee754_exp) (y * m81(__ieee754_log) (-x));
if (sizeof (float_type) == sizeof (float))
{
long i = (long) y;
}
}
else
- z = (x - x) / (x - x);
+ z = 0.0/0.0;
}
else
- z = (__m81_u(__CONCATX(__ieee754_exp,SUFF))
- (y * __m81_u(__CONCATX(__ieee754_log,SUFF)) (x)));
+ z = m81(__ieee754_exp) (y * m81(__ieee754_log) (x));
return z;
}
--- /dev/null
+/* Complex exponential function. m68k fpu version
+ Copyright (C) 1997 Free Software Foundation, Inc.
+ This file is part of the GNU C Library.
+ Contributed by Andreas Schwab <schwab@issan.informatik.uni-dortmund.de>
+
+ The GNU C Library is free software; you can redistribute it and/or
+ modify it under the terms of the GNU Library General Public License as
+ published by the Free Software Foundation; either version 2 of the
+ License, or (at your option) any later version.
+
+ The GNU C Library is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ Library General Public License for more details.
+
+ You should have received a copy of the GNU Library General Public
+ License along with the GNU C Library; see the file COPYING.LIB. If not,
+ write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
+ Boston, MA 02111-1307, USA. */
+
+#define __LIBC_M81_MATH_INLINES
+#include <complex.h>
+#include <math.h>
+
+#ifndef SUFF
+#define SUFF
+#endif
+#ifndef huge_val
+#define huge_val HUGE_VAL
+#endif
+#ifndef float_type
+#define float_type double
+#endif
+
+#define CONCATX(a,b) __CONCAT(a,b)
+#define s(name) CONCATX(name,SUFF)
+#define m81(func) __m81_u(s(func))
+
+__complex__ float_type
+s(__cexp) (__complex__ float_type x)
+{
+ __complex__ float_type retval;
+ float_type sin_ix, cos_ix;
+
+ if (m81(__finite) (__real__ x))
+ {
+ if (m81(__finite) (__imag__ x))
+ {
+ float_type exp_val = s(__exp) (__real__ x);
+
+ __asm ("fsincos%.x %2,%1:%0" : "=f" (sin_ix), "=f" (cos_ix)
+ : "f" (__imag__ x));
+ __real__ retval = exp_val * cos_ix;
+ __imag__ retval = exp_val * sin_ix;
+ }
+ else
+ /* If the imaginary part is +-inf or NaN and the real part is
+ not +-inf the result is NaN + iNaN. */
+ __real__ retval = __imag__ retval = 0.0/0.0;
+ }
+ else if (m81(__isinf) (__real__ x))
+ {
+ if (m81(__finite) (__imag__ x))
+ {
+ if (m81(__signbit) (__real__ x) == 0 && __imag__ x == 0.0)
+ retval = huge_val;
+ else
+ {
+ float_type value = m81(__signbit) (__real__ x) ? 0.0 : huge_val;
+
+ __asm ("fsincos%.x %2,%1:%0" : "=f" (sin_ix), "=f" (cos_ix)
+ : "f" (__imag__ x));
+ __real__ retval = value * cos_ix;
+ __imag__ retval = value * sin_ix;
+ }
+ }
+ else if (m81(__signbit) (__real__ x) == 0)
+ {
+ __real__ retval = huge_val;
+ __imag__ retval = 0.0/0.0;
+ }
+ else
+ retval = 0.0;
+ }
+ else
+ /* If the real part is NaN the result is NaN + iNaN. */
+ __real__ retval = __imag__ retval = 0.0/0.0;
+
+ return retval;
+}
+#define weak_aliasx(a,b) weak_alias(a,b)
+weak_aliasx (s(__cexp), s(cexp))
--- /dev/null
+#define SUFF f
+#define huge_val HUGE_VALF
+#define float_type float
+#include <s_cexp.c>
--- /dev/null
+#define SUFF l
+#define huge_val HUGE_VALL
+#define float_type long double
+#include <s_cexp.c>
#define __LIBC_M81_MATH_INLINES
#include <math.h>
-#ifndef FUNC
-#define FUNC modf
+#ifndef SUFF
+#define SUFF
#endif
#ifndef float_type
#define float_type double
#endif
-#define __CONCATX(a,b) __CONCAT(a,b)
+#define CONCATX(a,b) __CONCAT(a,b)
+#define s(name) CONCATX(name,SUFF)
+#define m81(func) __m81_u(s(func))
float_type
-__CONCATX(__,FUNC) (x, iptr)
- float_type x;
- float_type *iptr;
+s(__modf) (float_type x, float_type *iptr)
{
- return __m81_u(__CONCATX(__,FUNC))(x, iptr);
+ float_type x_int, result;
+ __asm ("fintrz%.x %1, %0" : "=f" (x_int) : "f" (x));
+ *iptr = x_int;
+ if (m81(__isinf) (x))
+ {
+ result = 0;
+ if (x < 0)
+ result = -result;
+ }
+ else if (x == 0)
+ result = x;
+ else
+ result = x - x_int;
+ return result;
}
#define weak_aliasx(a,b) weak_alias(a,b)
-weak_aliasx(__CONCATX(__,FUNC), FUNC)
+weak_aliasx(s(__modf), s(modf))
-#define FUNC modff
+#define SUFF f
#define float_type float
#include <s_modf.c>
-#define FUNC modfl
+#define SUFF l
#define float_type long double
#include <s_modf.c>
--- /dev/null
+#define FUNC nearbyint
+#include <s_atan.c>
--- /dev/null
+#define FUNC nearbyintf
+#include <s_atanf.c>
--- /dev/null
+#define FUNC nearbyintl
+#include <s_atanl.c>
--- /dev/null
+#define FUNC trunc
+#include <s_atan.c>
--- /dev/null
+#define FUNC truncf
+#include <s_atanf.c>
--- /dev/null
+#define FUNC truncl
+#include <s_atanl.c>
--- /dev/null
+ # Optimized strlen implementation for PowerPC.
+ # Copyright (C) 1997 Free Software Foundation, Inc.
+ # This file is part of the GNU C Library.
+ #
+ # The GNU C Library is free software; you can redistribute it and/or
+ # modify it under the terms of the GNU Library General Public License as
+ # published by the Free Software Foundation; either version 2 of the
+ # License, or (at your option) any later version.
+ #
+ # The GNU C Library is distributed in the hope that it will be useful,
+ # but WITHOUT ANY WARRANTY; without even the implied warranty of
+ # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ # Library General Public License for more details.
+ #
+ # You should have received a copy of the GNU Library General Public
+ # License along with the GNU C Library; see the file COPYING.LIB. If not,
+ # write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
+ # Boston, MA 02111-1307, USA.
+
+ # The algorithm here uses the following techniques:
+ #
+ # 1) Given a word 'x', we can test to see if it contains any 0 bytes
+ # by subtracting 0x01010101, and seeing if any of the high bits of each
+ # byte changed from 0 to 1. This works because the least significant
+ # 0 byte must have had no incoming carry (otherwise it's not the least
+ # significant), so it is 0x00 - 0x01 == 0xff. For all other
+ # byte values, either they have the high bit set initially, or when
+ # 1 is subtracted you get a value in the range 0x00-0x7f, none of which
+ # have their high bit set. The expression here is
+ # (x + 0xfefefeff) & ~(x | 0x7f7f7f7f), which gives 0x00000000 when
+ # there were no 0x00 bytes in the word.
+ #
+ # 2) Given a word 'x', we can test to see _which_ byte was zero by
+ # calculating ~(((x & 0x7f7f7f7f) + 0x7f7f7f7f) | x | 0x7f7f7f7f).
+ # This produces 0x80 in each byte that was zero, and 0x00 in all
+ # the other bytes. The '| 0x7f7f7f7f' clears the low 7 bits in each
+ # byte, and the '| x' part ensures that bytes with the high bit set
+ # produce 0x00. The addition will carry into the high bit of each byte
+ # iff that byte had one of its low 7 bits set. We can then just see
+ # which was the most significant bit set and divide by 8 to find how
+ # many to add to the index.
+ # This is from the book 'The PowerPC Compiler Writer's Guide',
+ # by Steve Hoxey, Faraydon Karim, Bill Hay and Hank Warren.
+ #
+ # We deal with strings not aligned to a word boundary by taking the
+ # first word and ensuring that bytes not part of the string
+ # are treated as nonzero. To allow for memory latency, we unroll the
+ # loop a few times, being careful to ensure that we do not read ahead
+ # across cache line boundaries.
+ #
+ # Questions to answer:
+ # 1) How long are strings passed to strlen? If they're often really long,
+ # we should probably use cache management instructions and/or unroll the
+ # loop more. If they're often quite short, it might be better to use
+ # fact (2) in the inner loop than have to recalculate it.
+ # 2) How popular are bytes with the high bit set? If they are very rare,
+ # on some processors it might be useful to use the simpler expression
+ # ~((x - 0x01010101) | 0x7f7f7f7f) (that is, on processors with only one
+ # ALU), but this fails when any character has its high bit set.
+
+ # Some notes on register usage: Under the SVR4 ABI, we can use registers
+ # 0 and 3 through 12 (so long as we don't call any procedures) without
+ # saving them. We can also use registers 14 through 31 if we save them.
+ # We can't use r1 (it's the stack pointer), nor r2 or r13 because the user
+ # program may expect them to be hold their usual value if we get sent
+ # a signal. Integer parameters are passed in r3 through r10.
+ # We can use condition registers cr0, cr1, cr5, cr6, and cr7 without saving
+ # them, the others we must save.
+
+ .section ".text"
+ .align 3
+ .globl strlen
+ .type strlen,@function
+strlen:
+ # On entry, r3 points to the string, and it's left that way.
+ # We use r6 to store 0x01010101, and r7 to store 0x7f7f7f7f.
+ # r4 is used to keep the current index into the string; r5 holds
+ # the number of padding bits we prepend to the string to make it
+ # start at a word boundary. r8 holds the 'current' word.
+ # r9-12 are temporaries. r0 is used as a temporary and for discarded
+ # results.
+ clrrwi 4,3,2
+ lis 6,0xfeff
+ lis 7,0x7f7f
+ rlwinm 10,3,0,29,29
+ lwz 8,0(4)
+ addi 7,7,0x7f7f
+ rlwinm 5,3,3,27,28
+ cmpwi 1,10,0
+ li 9,-1
+ # That's the setup done, now do the first pair of words.
+ # We make an exception and use method (2) on the first two words, to reduce
+ # overhead.
+ srw 9,9,5
+ and 0,7,8
+ or 10,7,8
+ add 0,0,7
+ nor 0,10,0
+ and. 8,0,9
+ bne done0
+ # Handle second word of pair. Put addi between branches to avoid hurting
+ # branch prediction.
+ addi 6,6,0xfffffeff
+
+ bne 1,loop
+ lwzu 8,4(4)
+ and 0,7,8
+ or 10,7,8
+ add 0,0,7
+ nor. 0,10,0
+ bne done0
+
+ # The loop.
+
+loop: lwz 8,4(4)
+ lwzu 9,8(4)
+ add 0,6,8
+ nor 10,7,8
+ and. 0,0,10
+ add 11,6,9
+ nor 12,7,9
+ bne done1
+ and. 0,11,12
+ beq loop
+
+ and 0,7,9
+ or 10,7,9
+ b done2
+
+done1: addi 4,4,-4
+ and 0,7,9
+ or 10,7,9
+done2: add 0,0,7
+ nor 0,10,0
+
+ # When we get to here, r4 points to the first word in the string that
+ # contains a zero byte, and the most significant set bit in r8 is in that
+ # byte.
+done0: cntlzw 11,8
+ subf 0,3,4
+ srwi 11,11,3
+ add 3,0,11
+ blr
+0:
+ .size strlen,0b-strlen
-/* Copyright (C) 1991, 1995, 1996 Free Software Foundation, Inc.
-This file is part of the GNU C Library.
+/* Copyright (C) 1991, 1995, 1996, 1997 Free Software Foundation, Inc.
+ This file is part of the GNU C Library.
-The GNU C Library is free software; you can redistribute it and/or
-modify it under the terms of the GNU Library General Public License as
-published by the Free Software Foundation; either version 2 of the
-License, or (at your option) any later version.
+ The GNU C Library is free software; you can redistribute it and/or
+ modify it under the terms of the GNU Library General Public License as
+ published by the Free Software Foundation; either version 2 of the
+ License, or (at your option) any later version.
-The GNU C Library is distributed in the hope that it will be useful,
-but WITHOUT ANY WARRANTY; without even the implied warranty of
-MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
-Library General Public License for more details.
+ The GNU C Library is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ Library General Public License for more details.
-You should have received a copy of the GNU Library General Public
-License along with the GNU C Library; see the file COPYING.LIB. If
-not, write to the Free Software Foundation, Inc., 675 Mass Ave,
-Cambridge, MA 02139, USA. */
+ You should have received a copy of the GNU Library General Public
+ License along with the GNU C Library; see the file COPYING.LIB. If not,
+ write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
+ Boston, MA 02111-1307, USA. */
#include <errno.h>
#include <stddef.h>
fchdir (fd)
int fd;
{
- __set_errno (ENOSYS(;
+ __set_errno (ENOSYS);
return -1;
}
-/* Copyright (C) 1991, 1992, 1995, 1996 Free Software Foundation, Inc.
-This file is part of the GNU C Library.
+/* Copyright (C) 1991, 1992, 1995, 1996, 1997 Free Software Foundation, Inc.
+ This file is part of the GNU C Library.
-The GNU C Library is free software; you can redistribute it and/or
-modify it under the terms of the GNU Library General Public License as
-published by the Free Software Foundation; either version 2 of the
-License, or (at your option) any later version.
+ The GNU C Library is free software; you can redistribute it and/or
+ modify it under the terms of the GNU Library General Public License as
+ published by the Free Software Foundation; either version 2 of the
+ License, or (at your option) any later version.
-The GNU C Library is distributed in the hope that it will be useful,
-but WITHOUT ANY WARRANTY; without even the implied warranty of
-MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
-Library General Public License for more details.
+ The GNU C Library is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ Library General Public License for more details.
-You should have received a copy of the GNU Library General Public
-License along with the GNU C Library; see the file COPYING.LIB. If
-not, write to the Free Software Foundation, Inc., 675 Mass Ave,
-Cambridge, MA 02139, USA. */
+ You should have received a copy of the GNU Library General Public
+ License along with the GNU C Library; see the file COPYING.LIB. If not,
+ write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
+ Boston, MA 02111-1307, USA. */
#include <errno.h>
#include <stddef.h>
{
if (fd < 0)
{
- __set_errno (EINVAL(;
+ __set_errno (EINVAL);
return -1;
}
#include <sys/socket.h>
#include <sys/types.h>
+#include <netinet/in.h>
/* This structure gets passed by the SIOCADDRT and SIOCDELRT calls. */
#define rt_mss rt_mtu
+struct in6_rtmsg
+ {
+ struct in6_addr rtmsg_dst;
+ struct in6_addr rtmsg_src;
+ struct in6_addr rtmsg_gateway;
+ unsigned long int rtmsg_type;
+ unsigned short int rtmsg_dst_len;
+ unsigned short int rtmsg_src_len;
+ unsigned long int rtmsg_metric;
+ unsigned long int rtmsg_info;
+ unsigned long int rtmsg_flags;
+ int rtmsg_ifindex;
+ };
+
+
#define RTF_UP 0x0001 /* Route usable. */
#define RTF_GATEWAY 0x0002 /* Destination is a gateway. */
#define RTF_THROW 0x2000 /* Go to next class. */
#define RTF_NOPMTUDISC 0x4000 /* Do not send packets with DF. */
-/* Bad idea. IPv6 should not use broken IPv4 interface. */
-
-#define RTF_ADDRCONF 0x0800 /* Announced on link prefix. */
-#define RTF_INVALID 0x1000
-#define RTF_DCACHE 0x2000
-#define RTF_DEFAULT 0x4000 /* Route is a default route. */
-#define RTF_NEXTHOP 0x8000 /* Non gateway route with nexthop. */
+/* for IPv6 */
+#define RTF_DEFAULT 0x00010000 /* default - learned via ND */
+#define RTF_ALLONLINK 0x00020000 /* fallback, no routers on link */
+#define RTF_ADDRCONF 0x00040000 /* addrconf route - RA */
+#define RTF_LINKRT 0x00100000 /* link specific - device match */
+#define RTF_NONEXTHOP 0x00200000 /* route with no nexthop */
-#define RTF_MAGIC 0x10000 /* Route added/deleted automatically,
- when interface changes its state. */
+#define RTF_CACHE 0x01000000 /* cache entry */
+#define RTF_FLOW 0x02000000 /* flow significant route */
+#define RTF_POLICY 0x04000000 /* policy route */
#define RTCF_VALVE 0x00200000
#define RTCF_MASQ 0x00400000
#define SET_MANTISSA(flt, mant) \
do { union ieee754_float u; \
u.f = (flt); \
+ if ((mant & 0x7fffff) == 0) \
+ mant = 0x400000; \
u.ieee.mantissa = (mant) & 0x7fffff; \
(flt) = u.f; \
} while (0)
#define SET_MANTISSA(flt, mant) \
do { union ieee854_long_double u; \
u.d = (flt); \
- u.ieee.mantissa0 = ((mant) >> 32) & 0x7fffffff; \
+ if ((mant & 0x7fffffffffffffffULL) == 0) \
+ mant = 0x4000000000000000ULL; \
+ u.ieee.mantissa0 = (((mant) >> 32) & 0x7fffffff) | 0x80000000; \
u.ieee.mantissa1 = (mant) & 0xffffffff; \
(flt) = u.d; \
} while (0)
projects / platform / upstream / glibc.git / commitdiff