summaryrefslogtreecommitdiff
path: root/math
diff options
context:
space:
mode:
authorUlrich Drepper <drepper@redhat.com>1997-03-08 05:28:20 +0000
committerUlrich Drepper <drepper@redhat.com>1997-03-08 05:28:20 +0000
commit4cca6b868f08e4f95a4c3786a69135676b086c50 (patch)
treeea8f4fbb6972588bc85c5979175c1e5cb6734993 /math
parentf5f52655ceb5152d79ca88db5293fa1136969303 (diff)
downloadglibc-4cca6b868f08e4f95a4c3786a69135676b086c50.tar.gz
1997-03-08 05:30 Ulrich Drepper <drepper@cygnus.com> * argp/Makefile (routines): Add argp-eexst. * argp/argp-ba.c (argp_program_bug_address): Make `const'. * argp/argp-eexst.c: New file. Define exit status variable. * argp/argp-help.c (__argp_state_help): Use argp_err_exit_status variable instead of always exiting with value 1. * argp/argp-pv.c (argp_program_version): Make `const'. * argp/argp.h: Declare argp_err_exit_status. Patches by Miles Bader <miles@gnu.ai.mit.edu>. * locale/programs/localedef.c: Use argp_err_exit_status to terminate with correct value. * inet/rcmd.c (iruserok): Use euidaccess instead of half-hearted switching of UID before opening .rhosts. * libio/vsnprintf.c: Change implementation to follow ISO C 9X proposal. The return value now is always the number of characters which would be written if enough space is available. * manual/stdio.texi: Update description for new behaviour. * locale/locale.h (__locale_t): Don't use __locale_t for struct name and pointer to struct since old gccs cannot keep the namespaces apart. Rename struct to __locale_struct. * locale/duplocale.h: Likewise. * locale/newlocale.h: Likewise. * math/Makefile (headers): Add complex.h and cmathcalls.h. (routines): Remove w_cabs. Add conj, cimag, creal, and cabs. * math/math.h: Undefine __MATHDECL_1, __MATHDECL, and __MATHCALL after use. (signgam): Move declaration to here from mathcalls.h. * math/mathcalls.h (cabs, __cabs_complex, signgam): Remove definition. Correct comment for fmin and fmax. Start implementing complex math function from ISO C 9X. * math/complex.h: New file. * math/cmathcalls.h: New file. * math/cabs.c: New file. * math/cabsf.c: New file. * math/cabsl.c: New file. * math/cimag.c: New file. * math/cimagf.c: New file. * math/cimagl.c: New file. * math/conj.c: New file. * math/conjf.c: New file. * math/conjl.c: New file. * math/creal.c: New file. * math/crealf.c: New file. * math/creall.c: New file. * sysdeps/libm-ieee754/w_cabs.c: Removed. * sysdeps/libm-ieee754/w_cabsf.c: Removed. * sysdeps/libm-ieee754/w_cabsl.c: Removed. * posix/regex.c: Merge with regex sources from Arnold Robbins' version in GNU awk. * posix/regex.h: Likewise. Add regex test suite by Tom Lord. * posix/Makefile (distribute): Add TESTS, TESTS2C.sed, and testcases.h. (tests): Add runtests. (before-compile): Add testcases.h. (testcases.h): New rule to generate header with tests. * posix/TESTS: New file. * posix/TESTS2C.sed.: New file. * posix/runtests.c: New file. * posix/testcases.h: New file. * sysdeps/unix/sysv/linux/poll.c: Test whether poll syscall is available and use fall-back implementation if not. * sysdeps/unix/sysv/linux/syscalls.list: Sort entries. * time/ialloc.c: Update from tzcode1997c. * time/private.h: Likewise. * time/scheck.c: Likewise. * time/tzselect.ksh: Likewise. * time/zdump.c: Likewise. * time/zic.c: Likewise. * time/tzfile.c: Pretty print. 1997-03-06 07:37 Geoff Keating <geoffk@ozemail.com.au> Port to powerpc-*-linux-gnu. Slightly tested, under MkLinux, on a 601. * sysdeps/powerpc/Implies: Added. * sysdeps/powerpc/__longjmp.S: Added. * sysdeps/powerpc/__math.h: Added. * sysdeps/powerpc/bsd-_setjmp.S: Added. * sysdeps/powerpc/bsd-setjmp.S: Added. * sysdeps/powerpc/dl-machine.h: Added. * sysdeps/powerpc/ffs.c: Added. * sysdeps/powerpc/fpu_control.h: Added. * sysdeps/powerpc/jmp_buf.h: Added. * sysdeps/powerpc/setjmp.S: Added. * sysdeps/powerpc/strlen.s: Added. * sysdeps/powerpc/elf/start.c: Added. * sysdeps/powerpc/fpu_control.h: Added. * sysdeps/powerpc/jmp_buf.h: Added. * sysdeps/unix/sysv/linux/powerpc/Dist: Added. * sysdeps/unix/sysv/linux/powerpc/_exit.S: Added. * sysdeps/unix/sysv/linux/powerpc/brk.c: Added. * sysdeps/unix/sysv/linux/powerpc/clone.S: Added. * sysdeps/unix/sysv/linux/powerpc/init-first.h: Added. * sysdeps/unix/sysv/linux/powerpc/ioctl-types.h: Added. * sysdeps/unix/sysv/linux/powerpc/profil.c: Added. * sysdeps/unix/sysv/linux/powerpc/sigreturn.S: Added. * sysdeps/unix/sysv/linux/powerpc/socket.S: Added. * sysdeps/unix/sysv/linux/powerpc/syscall.S: Added. * sysdeps/unix/sysv/linux/powerpc/syscalls.list: Added. * sysdeps/unix/sysv/linux/powerpc/sysdep.c: Added. * sysdeps/unix/sysv/linux/powerpc/sysdep.h: Added. * sysdeps/unix/sysv/linux/powerpc/termbits.h: Added. * sysdeps/unix/sysv/linux/powerpc/sys/syscall.h: Added. 1997-03-05 05:24 Geoff Keating <geoffk@ozemail.com.au> * elf/dl-runtime.c (fixup): Add ELF_FIXUP_RETURNS_ADDRESS switch, because knowing the first instruction of a PowerPC PLT trampoline is not very helpful. 1997-03-04 08:04 Geoff Keating <geoffk@ozemail.com.au> * elf/dl-load.c (ELF_PREFERRED_ADDRESS, ELF_PREFERRED_ADDRESS_DATA, ELF_FIXED_ADDRESS): Added. These are for dl-machine.h to indicate a preference as to where executables should be loaded. 1997-02-28 08:50 Geoff Keating <geoffk@ozemail.com.au> * elf/elf.h: Add in all those PowerPC reloc types. 1997-02-24 07:12 Geoff Keating <geoffk@ozemail.com.au> * stdio-common/vfscanf.c: Use __va_copy if available. 1997-03-06 13:50 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * malloc/malloc.c (mprotect) [_LIBC]: Define as __mprotect to clean up namespace. 1997-03-07 14:27 Ulrich Drepper <drepper@cygnus.com> * sysdeps/libm-ieee754/s_fpclassify.c (__fpclassify): Don't use non-existing GET_WORDS macro, use EXTRACT_WORDS. Reported by Andres Schwab <schwab@issan.informatik.uni-dortmund.de>. 1997-03-07 05:27 Ulrich Drepper <drepper@cygnus.com> * sysdeps/unix/sysv/linux/i386/sigaction.c: The user- and kernel-level sigaction structure is different. handle this correctly. Bug report by Andres Jaeger <jaeger@informatik.uni-kl.de>. 1997-03-06 05:55 David Engel <david@sw.ods.com> * sysdeps/generic/dl-cache.c (_dl_load_cache_lookup): Also recognize cache entries with flag == 3. 1997-03-06 01:05 Ulrich Drepper <drepper@cygnus.com> * sysdeps/generic/netinet/in.h (INADDR_NONE): Cast to unsigned int. (INADDR_LOOPBACK): Likewise. * sysdeps/unix/sysv/linux/netinet/in.h: Likewise. * manual/socket.texi: Correct types of INADDR_* constants from unsigned long to unsigned. 1997-03-05 23:14 Ulrich Drepper <drepper@cygnus.com> * time/antarctica: Update from tzdata1997c. * time/asia: Likewise. * time/zone.tab: Likewise. 1997-03-05 00:43 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * configure.in (libc_cv_asm_symver_directive, libc_cv_ld_version_script_option): Remove unknown instruction from assembler test file. 1997-03-04 19:14 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * sysdeps/m68k/fpu/__math.h: Update feature tests to use __USE_ISOC9X. [__USE_ISOC9X]: Define __log2 and __exp2 inlines. (fabs): Remove defininition, it is a gcc builtin. (sqrt, __sqrt): Remove definition, they have wrappers in libm. 1997-03-04 10:11 H.J. Lu <hjl@lucon.org> * sunrpc/Makefile ($(objpfx)rpc-proto.c): Create subdir before generating file. 1997-03-05 03:59 Ulrich Drepper <drepper@cygnus.com> * inet/rcmd.c (__ivaliduser): Don't use getline, but __getline instead to avoid namespace problems. Reported by David Mosberger-Tang <davidm@AZStarNet.COM>. 1997-03-03 19:01 Andreas Schwab <schwab@issan.informatik.uni-dortmund.de> * login/utmp_file.c: Fix previous change. Make portable by checking the _HAVE_UT_* feature macros. * libio/fileops.c (_IO_file_finish): Likewise.
Diffstat (limited to 'math')
-rw-r--r--math/Makefile8
-rw-r--r--math/cabs.c33
-rw-r--r--math/cabsf.c29
-rw-r--r--math/cabsl.c29
-rw-r--r--math/cimag.c32
-rw-r--r--math/cimagf.c28
-rw-r--r--math/cimagl.c28
-rw-r--r--math/cmathcalls.h152
-rw-r--r--math/complex.h102
-rw-r--r--math/conj.c32
-rw-r--r--math/conjf.c28
-rw-r--r--math/conjl.c28
-rw-r--r--math/creal.c32
-rw-r--r--math/crealf.c28
-rw-r--r--math/creall.c28
-rw-r--r--math/math.h9
-rw-r--r--math/mathcalls.h15
17 files changed, 625 insertions, 16 deletions
diff --git a/math/Makefile b/math/Makefile
index b8964baa88..c4ed312c2b 100644
--- a/math/Makefile
+++ b/math/Makefile
@@ -21,7 +21,8 @@
subdir := math
# Installed header files.
-headers := math.h mathcalls.h __math.h huge_val.h nan.h fpu_control.h
+headers := math.h mathcalls.h __math.h huge_val.h nan.h \
+ fpu_control.h complex.h cmathcalls.h
# Internal header files.
distribute := math_private.h machine/asm.h machine/endian.h
@@ -43,10 +44,11 @@ libm-calls = e_acos e_acosh e_asin e_atan2 e_atanh e_cosh e_exp e_fmod \
s_floor s_frexp s_ilogb s_ldexp s_log1p s_logb \
s_modf s_nextafter s_rint s_scalbn s_significand \
s_sin s_tan s_tanh w_acos w_acosh w_asin \
- w_atan2 w_atanh w_cabs w_cosh w_drem w_exp w_fmod w_gamma \
+ w_atan2 w_atanh w_cosh w_drem w_exp w_fmod w_gamma \
w_gamma_r w_hypot w_j0 w_j1 w_jn w_lgamma w_lgamma_r \
w_log w_log10 w_pow w_remainder w_scalb w_sinh w_sqrt \
- s_signbit s_fpclassify s_fmax s_fmin s_fdim
+ s_signbit s_fpclassify s_fmax s_fmin s_fdim \
+ conj cimag creal cabs
libm-routines = $(libm-support) $(libm-calls) \
$(patsubst %_rf,%f_r,$(libm-calls:=f)) \
$(long-m-$(long-double-fcts))
diff --git a/math/cabs.c b/math/cabs.c
new file mode 100644
index 0000000000..cf775ded95
--- /dev/null
+++ b/math/cabs.c
@@ -0,0 +1,33 @@
+/* Return the complex absolute value of double complex value.
+ Copyright (C) 1997 Free Software Foundation, Inc.
+ 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 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. */
+
+#include <complex.h>
+#include <math.h>
+
+double
+__cabs (double complex z)
+{
+ return __hypot (__real__ z, __imag__ z);
+}
+weak_alias (__cabs, cabs)
+#ifdef NO_LONG_DOUBLE
+strong_alias (__cabs, __cabsl)
+weak_alias (__cabs, cabsl)
+#endif
diff --git a/math/cabsf.c b/math/cabsf.c
new file mode 100644
index 0000000000..2ea2660d97
--- /dev/null
+++ b/math/cabsf.c
@@ -0,0 +1,29 @@
+/* Return the complex absolute value of float complex value.
+ Copyright (C) 1997 Free Software Foundation, Inc.
+ 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 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. */
+
+#include <complex.h>
+#include <math.h>
+
+float
+__cabsf (float complex z)
+{
+ return __hypotf (__real__ z, __imag__ z);
+}
+weak_alias (__cabsf, cabsf)
diff --git a/math/cabsl.c b/math/cabsl.c
new file mode 100644
index 0000000000..1f3554cef2
--- /dev/null
+++ b/math/cabsl.c
@@ -0,0 +1,29 @@
+/* Return the complex absolute value of long double complex value.
+ Copyright (C) 1997 Free Software Foundation, Inc.
+ 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 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. */
+
+#include <complex.h>
+#include <math.h>
+
+long double
+__cabsl (long double complex z)
+{
+ return __hypotl (__real__ z, __imag__ z);
+}
+weak_alias (__cabsl, cabsl)
diff --git a/math/cimag.c b/math/cimag.c
new file mode 100644
index 0000000000..13efa1476e
--- /dev/null
+++ b/math/cimag.c
@@ -0,0 +1,32 @@
+/* Return imaginary part of complex double value.
+ Copyright (C) 1997 Free Software Foundation, Inc.
+ 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 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. */
+
+#include <complex.h>
+
+double
+__cimag (double complex z)
+{
+ return __imag__ z;
+}
+weak_alias (__cimag, cimag)
+#ifdef NO_LONG_DOUBLE
+strong_alias (__cimag, __cimagl)
+weak_alias (__cimag, cimagl)
+#endif
diff --git a/math/cimagf.c b/math/cimagf.c
new file mode 100644
index 0000000000..8edb4e613e
--- /dev/null
+++ b/math/cimagf.c
@@ -0,0 +1,28 @@
+/* Return imaginary part of complex float value.
+ Copyright (C) 1997 Free Software Foundation, Inc.
+ 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 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. */
+
+#include <complex.h>
+
+float
+__cimagf (float complex z)
+{
+ return __imag__ z;
+}
+weak_alias (__cimagf, cimagf)
diff --git a/math/cimagl.c b/math/cimagl.c
new file mode 100644
index 0000000000..9c58899274
--- /dev/null
+++ b/math/cimagl.c
@@ -0,0 +1,28 @@
+/* Return imaginary part of complex long double value.
+ Copyright (C) 1997 Free Software Foundation, Inc.
+ 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 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. */
+
+#include <complex.h>
+
+long double
+__cimagl (long double complex z)
+{
+ return __imag__ z;
+}
+weak_alias (__cimagl, cimagl)
diff --git a/math/cmathcalls.h b/math/cmathcalls.h
new file mode 100644
index 0000000000..658c3a729d
--- /dev/null
+++ b/math/cmathcalls.h
@@ -0,0 +1,152 @@
+/* Prototype declarations for complex math functions;
+ helper file for <complex.h>.
+ 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. */
+
+/* NOTE: Because of the special way this file is used by <math.h>, this
+ file must NOT be protected from multiple inclusion as header files
+ usually are.
+
+ This file provides prototype declarations for the math functions.
+ Most functions are declared using the macro:
+
+ __MATHCALL (NAME, (ARGS...));
+
+ This means there is a function `NAME' returning `double' and a function
+ `NAMEf' returning `float'. Each place `_Mdouble_' appears in the
+ prototype, that is actually `double' in the prototype for `NAME' and
+ `float' in the prototype for `NAMEf'. Reentrant variant functions are
+ called `NAME_r' and `NAMEf_r'.
+
+ Functions returning other types like `int' are declared using the macro:
+
+ __MATHDECL (TYPE, NAME, (ARGS...));
+
+ This is just like __MATHCALL but for a function returning `TYPE'
+ instead of `_Mdouble_'. In all of these cases, there is still
+ both a `NAME' and a `NAMEf' that takes `float' arguments. */
+
+#ifndef _COMPLEX_H
+ #error "Never include cmathcalls.h directly; include <complex.h> instead."
+#endif
+
+#define _Mdouble_complex_ _Mdouble_ complex
+
+
+/* Trigonometric functions. */
+
+/* Arc cosine of Z. */
+__MATHCALL (cacos, (_Mdouble_complex_ __z));
+/* Arc sine of Z. */
+__MATHCALL (casin, (_Mdouble_complex_ __z));
+/* Arc tangent of Z. */
+__MATHCALL (catan, (_Mdouble_complex_ __z));
+
+/* Cosine of Z. */
+__MATHCALL (ccos, (_Mdouble_complex_ __z));
+/* Sine of Z. */
+__MATHCALL (csin, (_Mdouble_complex_ __z));
+/* Tangent of Z. */
+__MATHCALL (ctan, (_Mdouble_complex_ __z));
+
+
+/* Hyperbolic functions. */
+
+/* Hyperbolic arc cosine of Z. */
+__MATHCALL (cacosh, (_Mdouble_complex_ __z));
+/* Hyperbolic arc sine of Z. */
+__MATHCALL (casinh, (_Mdouble_complex_ __z));
+/* Hyperbolic arc tangent of Z. */
+__MATHCALL (catanh, (_Mdouble_complex_ __z));
+
+/* Hyperbolic cosine of Z. */
+__MATHCALL (ccosh, (_Mdouble_complex_ __z));
+/* Hyperbolic sine of Z. */
+__MATHCALL (csinh, (_Mdouble_complex_ __z));
+/* Hyperbolic tangent of Z. */
+__MATHCALL (ctanh, (_Mdouble_complex_ __z));
+
+
+/* Exponential and logarithmic functions. */
+
+/* Exponential function of Z. */
+__MATHCALL (cexp, (_Mdouble_complex_ __z));
+
+/* Natural logarithm of Z. */
+__MATHCALL (clog, (_Mdouble_complex_ __z));
+
+
+/* Power functions. */
+
+/* Return X to the Y power. */
+__MATHCALL (cpow, (_Mdouble_complex_ __x, _Mdouble_complex_ __y));
+
+/* Return the square root of Z. */
+__MATHCALL (csqrt, (_Mdouble_complex_ __z));
+
+
+/* Absolute value, projections, conjugates, and projection. */
+
+/* Absolute value of Z. */
+__MATHDECL (_Mdouble_, cabs, (_Mdouble_complex_ __z));
+
+/* Argument value of Z. */
+__MATHDECL (_Mdouble_, carg, (_Mdouble_complex_ __z));
+
+/* Complex conjugate of Z. */
+__MATHCALL (conj, (_Mdouble_complex_ __z));
+
+/* Projection of Z onto the Riemann sphere. */
+__MATHCALL (cproj, (_Mdouble_complex_ __z));
+
+
+/* Decomposing complex values. */
+
+/* Imaginary part of Z. */
+__MATHDECL (_Mdouble_, cimag, (_Mdouble_complex_ __z));
+
+/* Real part of Z. */
+__MATHDECL (_Mdouble_, creal, (_Mdouble_complex_ __z));
+
+
+/* Now some optimized versions. GCC has handy notations for these
+ functions. */
+#if defined __GNUC__ && defined __OPTIMIZE__
+
+/* Imaginary part of Z. */
+extern __inline _Mdouble_
+__MATH_PRECNAME(cimag) (_Mdouble_complex_ __z)
+{
+ return __imag__ __z;
+}
+
+/* Real part of Z. */
+extern __inline _Mdouble_
+__MATH_PRECNAME(creal) (_Mdouble_complex_ __z)
+{
+ return __real__ __z;
+}
+
+/* Complex conjugate of Z. */
+extern __inline _Mdouble_complex_
+__MATH_PRECNAME(conj) (_Mdouble_complex_ __z)
+{
+ return ~__z;
+}
+
+#endif
diff --git a/math/complex.h b/math/complex.h
new file mode 100644
index 0000000000..6fc5c1975f
--- /dev/null
+++ b/math/complex.h
@@ -0,0 +1,102 @@
+/* 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. */
+
+/*
+ * ISO C 9X: 7.8 Complex arithmetic <complex.h>
+ */
+
+#ifndef _COMPLEX_H
+#define _COMPLEX_H 1
+
+/* We might need to add support for more compilers here. But once ISO
+ C 9X is out hopefully all maintained compilers will provide the data
+ types `float complex' and `double complex'. */
+#if (__GNUC__ == 2 && __GNUC_MINOR__ >= 7) || __GNUC__ > 2
+# define complex __complex__
+#endif
+
+
+/* Narrowest imaginary unit. This depends on the floating-point
+ evaluation method.
+ XXX This probably has to go into a gcc related file. */
+#define _Imaginary_I (DBL_EPSISON * 1.0i)
+
+/* Another more descriptive name is `I'. */
+#undef I
+#define I _Imaginary_I
+
+
+/* Optimization aids. This is not yet implemented in gcc and yonce it
+ is this will probably be available in a gcc header. */
+#define CX_LIMITED_RANGE_ON
+#define CX_LIMITED_RANGE_OFF
+#define CX_LIMITED_RANGE_DEFAULT
+
+
+/* The file <cmathcalls.h> contains the prototypes for all the actual
+ math functions. These macros are used for those prototypes, so
+ we can easily declare each function as both `name' and `__name',
+ and can declare the float versions `namef' and `__namef'. */
+
+#define __MATHCALL(function, args) \
+ __MATHDECL (_Mdouble_complex_,function, args)
+#define __MATHDECL(type, function, args) \
+ __MATHDECL_1(type, function, args); \
+ __MATHDECL_1(type, __CONCAT(__,function), args)
+#define __MATHDECL_1(type, function, args) \
+ extern type __MATH_PRECNAME(function) args
+
+#define _Mdouble_ double
+#define __MATH_PRECNAME(name) name
+#include <cmathcalls.h>
+#undef _Mdouble_
+#undef __MATH_PRECNAME
+
+/* Now the float versions. */
+#ifndef _Mfloat_
+#define _Mfloat_ float
+#endif
+#define _Mdouble_ _Mfloat_
+#ifdef __STDC__
+#define __MATH_PRECNAME(name) name##f
+#else
+#define __MATH_PRECNAME(name) name/**/f
+#endif
+#include <cmathcalls.h>
+#undef _Mdouble_
+#undef __MATH_PRECNAME
+
+/* And the long double versions. It is non-critical to define them
+ here unconditionally since `long double' is required in ISO C 9X. */
+#ifndef _Mlong_double_
+#define _Mlong_double_ long double
+#endif
+#define _Mdouble_ _Mlong_double_
+#ifdef __STDC__
+#define __MATH_PRECNAME(name) name##l
+#else
+#define __MATH_PRECNAME(name) name/**/l
+#endif
+#include <cmathcalls.h>
+#undef _Mdouble_
+#undef __MATH_PRECNAME
+#undef __MATHDECL_1
+#undef __MATHDECL
+#undef __MATHCALL
+
+#endif /* complex.h */
diff --git a/math/conj.c b/math/conj.c
new file mode 100644
index 0000000000..c15c417e4c
--- /dev/null
+++ b/math/conj.c
@@ -0,0 +1,32 @@
+/* Return complex conjugate of complex double value.
+ Copyright (C) 1997 Free Software Foundation, Inc.
+ 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 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. */
+
+#include <complex.h>
+
+double complex
+__conj (double complex z)
+{
+ return ~z;
+}
+weak_alias (__conj, conj)
+#ifdef NO_LONG_DOUBLE
+strong_alias (__conj, __conjl)
+weak_alias (__conj, conjl)
+#endif
diff --git a/math/conjf.c b/math/conjf.c
new file mode 100644
index 0000000000..828edaf2b7
--- /dev/null
+++ b/math/conjf.c
@@ -0,0 +1,28 @@
+/* Return complex conjugate of complex float value.
+ Copyright (C) 1997 Free Software Foundation, Inc.
+ 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 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. */
+
+#include <complex.h>
+
+float complex
+__conjf (float complex z)
+{
+ return ~z;
+}
+weak_alias (__conjf, conjf)
diff --git a/math/conjl.c b/math/conjl.c
new file mode 100644
index 0000000000..9e44dc32a1
--- /dev/null
+++ b/math/conjl.c
@@ -0,0 +1,28 @@
+/* Return complex conjugate of complex long double value.
+ Copyright (C) 1997 Free Software Foundation, Inc.
+ 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 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. */
+
+#include <complex.h>
+
+long double complex
+__conjl (long double complex z)
+{
+ return ~z;
+}
+weak_alias (__conjl, conjl)
diff --git a/math/creal.c b/math/creal.c
new file mode 100644
index 0000000000..c3ec4f7802
--- /dev/null
+++ b/math/creal.c
@@ -0,0 +1,32 @@
+/* Return real part of complex double value.
+ Copyright (C) 1997 Free Software Foundation, Inc.
+ 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 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. */
+
+#include <complex.h>
+
+double
+__creal (double complex z)
+{
+ return __real__ z;
+}
+weak_alias (__creal, creal)
+#ifdef NO_LONG_DOUBLE
+strong_alias (__creal, __creall)
+weak_alias (__creal, creall)
+#endif
diff --git a/math/crealf.c b/math/crealf.c
new file mode 100644
index 0000000000..bb9c2c1418
--- /dev/null
+++ b/math/crealf.c
@@ -0,0 +1,28 @@
+/* Return real part of complex float value.
+ Copyright (C) 1997 Free Software Foundation, Inc.
+ 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 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. */
+
+#include <complex.h>
+
+float
+__crealf (float complex z)
+{
+ return __real__ z;
+}
+weak_alias (__crealf, crealf)
diff --git a/math/creall.c b/math/creall.c
new file mode 100644
index 0000000000..0ab1aab2cb
--- /dev/null
+++ b/math/creall.c
@@ -0,0 +1,28 @@
+/* Return real part of complex long double value.
+ Copyright (C) 1997 Free Software Foundation, Inc.
+ 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 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. */
+
+#include <complex.h>
+
+long double
+__creall (long double complex z)
+{
+ return __real__ z;
+}
+weak_alias (__creall, creall)
diff --git a/math/math.h b/math/math.h
index e08d210be8..22274f023e 100644
--- a/math/math.h
+++ b/math/math.h
@@ -96,6 +96,15 @@ __BEGIN_DECLS
#endif /* __STDC__ || __GNUC__ */
#endif /* Use misc or ISO C 9X. */
+#undef __MATHDECL_1
+#undef __MATHDECL
+#undef __MATHCALL
+
+
+#if defined __USE_MISC || defined __USE_XOPEN || defined __USE_ISOC9X
+/* This variable is used by `gamma' and `lgamma'. */
+extern int signgam;
+#endif
/* ISO C 9X defines some generic macros which work on any data type. */
diff --git a/math/mathcalls.h b/math/mathcalls.h
index fd8ed86e81..d529c1a8e2 100644
--- a/math/mathcalls.h
+++ b/math/mathcalls.h
@@ -178,14 +178,6 @@ __MATHCALL (scalbn,, (_Mdouble_ __x, int __n));
/* Return the remainder of X/Y. */
__MATHCALL (drem,, (_Mdouble_ __x, _Mdouble_ __y));
-struct __MATH_PRECNAME(__cabs_complex,)
-{
- _Mdouble_ x, y;
-};
-
-/* Return `sqrt(X*X + Y*Y)'. */
-__MATHCALL (cabs,, (struct __MATH_PRECNAME(__cabs_complex,)));
-
/* Return the fractional part of X after dividing out `ilogb (X)'. */
__MATHCALL (significand,, (_Mdouble_ __x));
@@ -226,9 +218,6 @@ __MATHCALL (erfc,, (_Mdouble_));
__MATHCALL (gamma,, (_Mdouble_));
__MATHCALL (lgamma,, (_Mdouble_));
-/* This variable is used by `gamma' and `lgamma'. */
-extern int signgam;
-
#ifdef __USE_MISC
/* Reentrant versions of gamma and lgamma. Those functions use the global
variable `signgam'. The reentrant versions instead take a pointer and
@@ -279,10 +268,10 @@ __MATHCALL (remquo,, (_Mdouble_ __x, _Mdouble_ __y, int *__quo));
/* Return positive difference between X and Y. */
__MATHCALL (fdim,, (_Mdouble_ __x, _Mdouble_ __y));
-/* Return minimum numeric value from X and Y. */
+/* Return maximum numeric value from X and Y. */
__MATHCALL (fmax,, (_Mdouble_ __x, _Mdouble_ __y));
-/* Return maximum numeric value from X and Y. */
+/* Return minimum numeric value from X and Y. */
__MATHCALL (fmin,, (_Mdouble_ __x, _Mdouble_ __y));