update from main archive 970304cvs/libc-970307cvs/libc-970306cvs/libc-970305cvs/libc-970304

1997-03-04 04:31  Ulrich Drepper  <drepper@cygnus.com>

	* Makerules: Add rules to handle versioning.
	* config.h.in (DO_VERSIONING): New macro.
	* config.make.in (versioning): New variable.
	* configure.in: Add checks for .symver directive in gas and
	--version-script option to ld.  Define DO_VERSIONING and
	versioning if appropriate.

	* math/Makefile (routines): Add s_signbit, s_fpclassify, s_fmax,
	s_fmin, and s_fdim.
	* math/math.h: Define ISO C 9X constants, macros and functions.
	* math/mathcalls.h: Likewise.
	* sysdeps/libm-ieee754/s_fdim.c: New file.
	* sysdeps/libm-ieee754/s_fdimf.c: New file.
	* sysdeps/libm-ieee754/s_fdiml.c: New file.
	* sysdeps/libm-ieee754/s_fmax.c: New file.
	* sysdeps/libm-ieee754/s_fmaxf.c: New file.
	* sysdeps/libm-ieee754/s_fmaxl.c: New file.
	* sysdeps/libm-ieee754/s_fmin.c: New file.
	* sysdeps/libm-ieee754/s_fminf.c: New file.
	* sysdeps/libm-ieee754/s_fminl.c: New file.
	* sysdeps/libm-ieee754/s_fpclassify.c: New file.
	* sysdeps/libm-ieee754/s_fpclassifyf.c: New file.
	* sysdeps/libm-ieee754/s_fpclassifyl.c: New file.
	* sysdeps/libm-ieee754/s_signbit.c: New file.
	* sysdeps/libm-ieee754/s_signbitf.c: New file.
	* sysdeps/libm-ieee754/s_signbitl.c: New file.

	* stdio-common/printf_fphex.c: Correct printing of decimal point
	character.
	Simplify conversion of mantissa to string.
	* stdio-common/vfscanf.c: Handle %A format.
	Optimize termination of floating-point scanning.
	* stdio-common/tstscanf.c (main): Add new test to scanf to test
	scanning float values with given width.
	* stdlib/strtod.c: Add handling of floating-point numbers in
	hexadecimal notation.

	* stdlib/stdlib.h: Use __USE_ISOC9X feature macro for new long long
	functions.
	Pretty print #if directives.
	* string/string.h: Pretty print #if directives.

	* sysdeps/ieee754/dbl2mpn.c: Update copyright.
	* sysdeps/ieee754/ldbl2mpn.c: Likewise.
	* sysdeps/ieee754/mpn2dbl.c: Likewise.
	* sysdeps/ieee754/mpn2flt.c: Likewise.
	* sysdeps/ieee754/mpn2ldbl.c: Likewise.

	* sysdeps/unix/sysv/linux/poll.c: Implement poll function by
	falling back to select-based implementation if syscall isn't
	available.
	* sysdeps/unix/sysv/linux/syscalls.list: Add s_poll.

	* time/leapseconds: Update from tzdata1997b.
	* time/zic.c: Update from tzcode1997b.

1997-03-01 15:08  Andreas Schwab  <schwab@issan.informatik.uni-dortmund.de>

	* time/Makefile $(tzfiles:%=$(objpfx)z.%): Remove unneeded
	depedencies between installed $(tzlinks) and $(tzbases) files.

1997-03-01 14:27  Andreas Schwab  <schwab@issan.informatik.uni-dortmund.de>

	* math/math.h: Make compatible with traditional preprocessor;
	requires carefull placement of whitespace in macro arguments.
	Use __CONCAT instead of ##.
	Declare long double functions only if __STDC__ or __GNUC__.
	* math/mathcall.h: Avoid whitespace before argument of macro call
	that is used as function name.

	* sysdeps/m68k/fpu/__math.h: Use __CONCAT instead of ##.
	(__m81_u, __m81_inline): Depend on __LIBC_M81_MATH_INLINES instead
	of __NO_M81_MATH_INLINES.
	[!__LIBC_M81_MATH_INLINES]: Don't define internal functions
	starting with __ieee754.
	[!__NO_MATH_INLINES && __OPTIMIZE__]: Define user visible
	functions as inlines.
	(__m81_defun): Put __attribute__ between return type and function
	name.
	* math/math.h: Include <__math.h> also if __LIBC_M81_MATH_INLINES
	is defined.
	* sysdeps/m68k/fpu/e_acos.c: Define __LIBC_M81_MATH_INLINES
	instead of __NO_M81_MATH_INLINES.
	* sysdeps/m68k/fpu/e_fmod.c: Likewise.
	* sysdeps/m68k/fpu/k_cos.c: Likewise.
	* sysdeps/m68k/fpu/k_sin.c: Likewise.
	* sysdeps/m68k/fpu/k_tan.c: Likewise.
	* sysdeps/m68k/fpu/s_atan.c: Likewise. De-ANSI-declify.
	* sysdeps/m68k/fpu/s_frexp.c: Likewise.
	* sysdeps/m68k/fpu/s_ilogb.c: Likewise.
	* sysdeps/m68k/fpu/s_isinf.c: Likewise.
	* sysdeps/m68k/fpu/s_modf.c: Likewise.
	* sysdeps/m68k/fpu/s_scalbn.c: Likewise.

1997-02-27 21:51  Andreas Schwab  <schwab@issan.informatik.uni-dortmund.de>

	* Makefile (tests): Cope with $PATH not including the current
	directory.

1997-02-27 18:04  Andreas Schwab  <schwab@issan.informatik.uni-dortmund.de>

	* sysdeps/unix/Makefile ($(common-objpfx)mk-local_lim): Use
	$(common-objdir-compile).
	($(common-objpfx)make-ioctls): Likewise.
	(mk-local_lim-CFLAGS): Set this instead of local_lim-CFLAGS.
	($(common-objpfx)sys/param.h): Use $(make-target-directory).
	($(addprefix $(common-objpfx),$(sys/param.h-includes))):
	Likewise.
	($(common-objpfx)sys/syscall.h): Likewise.
	($(common-objpfx)local_lim.h): Let make deal with command
	failure.
	($(common-objpfx)param.h.dep): Use temporary file and update
	target atomically.
	($(common-objpfx)errnos): Avoid the Useless Use of cat Award.
	(include $(common-objpfx)param.h.dep): Ignore error.

	* sysdeps/posix/Makefile ($(common-objpfx)mk-stdiolim): Use
	$(common-objdir-compile).
	(mk-stdiolim-CFLAGS): Renamed from cded-objdir-includes, use
	$(shell pwd) instead of $$cwd.

	* sysdeps/generic/Makefile ($(common-objpfx)det_endian): Use
	$(common-objdir-compile).
	($(objpfx)make_siglist): Use $(native-compile).
	(make_siglist-CFLAGS): New variable.

	* Makerules (ALL_BUILD_CFLAGS): Renamed from BUILD_CFLAGS, leaving
	the old name for the user to pass additional flags to the host
	compiler.  Fix reference to config header.
	(native-compile, common-objdir-compile): Rewritten to make more
	generally usable.
	* sysdeps/unix/sysv/sysv4/solaris2/Makefile: Set ALL_BUILD_CFLAGS
	instead of BUILD_CFLAGS.

	* sysvips/sys/ipc.h: Warn if needed feature select macro are not
	defined.

1997-02-27 17:11  Andreas Schwab  <schwab@issan.informatik.uni-dortmund.de>

	* sunrpc/Makefile ($(objpfx)rpc-proto.d, $(objpfx)rpc-proto.c):
	New rules to generate dependencies for the RPC service objects.

1997-02-27 16:26  Andreas Schwab  <schwab@issan.informatik.uni-dortmund.de>

	* argp/argp-parse.c (parser_finalize): Always set *END_INDEX if
	supplied.

1997-02-28 03:27  Ulrich Drepper  <drepper@cygnus.com>

	* stdlib/strtod.c (STRTOF): Make sure return value is large enough
	so that clearing second word is necessary.

	* sysdeps/unix/sysv/linux/netinet/in_systm.h: Don't use kernel
	header since it is wrong for 64 bit systems.
	Patch by a sun <asun@zoology.washington.edu>.

1997-02-27 10:34:11  Richard Henderson  <rth@tamu.edu>

	* sysdeps/unix/sysv/linux/alpha/brk.S: Support both the Linux/i386
	and OSF/1 style brk syscalls.  We may want to change Linux/Alpha
	for the benefit of running foreign binaries.

1997-03-01 20:21  Miles Bader  <miles@gnu.ai.mit.edu>
1997-02-25 19:42  Miles Bader  <miles@gnu.ai.mit.edu>

1 files changed, 83 insertions, 28 deletions

diff --git a/math/mathcalls.h b/math/mathcalls.h
index ca2fe0c32a..fd8ed86e81 100644
--- a/math/mathcalls.h
+++ b/math/mathcalls.h
@@ -73,7 +73,7 @@ __MATHCALL (sinh,, (_Mdouble_ __x));
 /* Hyperbolic tangent of X.  */
 __MATHCALL (tanh,, (_Mdouble_ __x));
 
-#if defined(__USE_MISC) || defined(__USE_XOPEN_EXTENDED)
+#if defined __USE_MISC || defined __USE_XOPEN_EXTENDED || defined __USE_ISOC9X
 /* Hyperbolic arc cosine of X.  */
 __MATHCALL (acosh,, (_Mdouble_ __x));
 /* Hyperbolic arc sine of X.  */
@@ -99,7 +99,10 @@ __MATHCALL (log,, (_Mdouble_ __x));
 /* Base-ten logarithm of X.  */
 __MATHCALL (log10,, (_Mdouble_ __x));
 
-#if defined(__USE_MISC) || defined(__USE_XOPEN_EXTENDED)
+/* Break VALUE into integral and fractional parts.  */
+__MATHCALL (modf,, (_Mdouble_ __x, _Mdouble_ *__iptr));
+
+#if defined __USE_MISC || defined __USE_XOPEN_EXTENDED || defined __USE_ISOC9X
 /* Return exp(X) - 1.  */
 __MATHCALL (expm1,, (_Mdouble_ __x));
 
@@ -110,8 +113,13 @@ __MATHCALL (log1p,, (_Mdouble_ __x));
 __MATHCALL (logb,, (_Mdouble_ __x));
 #endif
 
-/* Break VALUE into integral and fractional parts.  */
-__MATHCALL (modf,, (_Mdouble_ __x, _Mdouble_ *__iptr));
+#ifdef __USE_ISOC9X
+/* Compute base-2 exponential of X.  */
+__MATHCALL (exp2,, (_Mdouble_ __x));
+
+/* Compute base-2 logarithm of X.  */
+__MATHCALL (log2,, (_Mdouble_ __x));
+#endif
 
 
 /* Power functions.  */
@@ -122,7 +130,12 @@ __MATHCALL (pow,, (_Mdouble_ __x, _Mdouble_ __y));
 /* Return the square root of X.  */
 __MATHCALL (sqrt,, (_Mdouble_ __x));
 
-#if defined(__USE_MISC) || defined(__USE_XOPEN_EXTENDED)
+#if defined __USE_MISC || defined __USE_XOPEN_EXTENDED || defined __USE_ISOC9X
+/* Return `sqrt(X*X + Y*Y)'.  */
+__MATHCALL (hypot,, (_Mdouble_ __x, _Mdouble_ __y));
+#endif
+
+#if defined __USE_MISC || defined __USE_XOPEN_EXTENDED
 /* Return the cube root of X.  */
 __MATHCALL (cbrt,, (_Mdouble_ __x));
 #endif
@@ -144,13 +157,12 @@ __MATHCALL (fmod,, (_Mdouble_ __x, _Mdouble_ __y));
 
 
 #ifdef __USE_MISC
-
 /* Return 0 if VALUE is finite or NaN, +1 if it
    is +Infinity, -1 if it is -Infinity.  */
-__MATHDECL (int, isinf,, (_Mdouble_ __value));
+__MATHDECL (int,isinf,, (_Mdouble_ __value));
 
 /* Return nonzero if VALUE is finite and not NaN.  */
-__MATHDECL (int, finite,, (_Mdouble_ __value));
+__MATHDECL (int,finite,, (_Mdouble_ __value));
 
 /* Deal with an infinite or NaN result.
    If ERROR is ERANGE, result is +Inf;
@@ -160,9 +172,6 @@ __MATHDECL (int, finite,, (_Mdouble_ __value));
    and may return an infinity or NaN, or may do something else.  */
 __MATHCALL (infnan,, (int __error));
 
-/* Return X with its signed changed to Y's.  */
-__MATHCALL (copysign,, (_Mdouble_ __x, _Mdouble_ __y));
-
 /* Return X times (2 to the Nth power).  */
 __MATHCALL (scalbn,, (_Mdouble_ __x, int __n));
 
@@ -182,36 +191,45 @@ __MATHCALL (cabs,, (struct __MATH_PRECNAME(__cabs_complex,)));
 __MATHCALL (significand,, (_Mdouble_ __x));
 #endif /* Use misc.  */
 
+#if defined __USE_MISC || defined __USE_ISOC9X
+/* Return X with its signed changed to Y's.  */
+__MATHCALL (copysign,, (_Mdouble_ __x, _Mdouble_ __y));
+#endif
+
+#ifdef __USE_ISOC9X
+/* Return representation of NaN for double type.  */
+__MATHCALL (nan,, (__const char *__tagb));
+#endif
 
-#if defined(__USE_MISC) || defined(__USE_XOPEN)
 
+#if defined __USE_MISC || defined __USE_XOPEN
 /* Return nonzero if VALUE is not a number.  */
-__MATHDECL (int, isnan,, (_Mdouble_ __value));
+__MATHDECL (int,isnan,, (_Mdouble_ __value));
 
 /* Return the binary exponent of X, which must be nonzero.  */
-__MATHDECL (int, ilogb,, (_Mdouble_ __x));
+__MATHDECL (int,ilogb,, (_Mdouble_ __x));
 
-/* Return `sqrt(X*X + Y*Y)'.  */
-__MATHCALL (hypot,, (_Mdouble_ __x, _Mdouble_ __y));
+/* Bessel functions.  */
+__MATHCALL (j0,, (_Mdouble_));
+__MATHCALL (j1,, (_Mdouble_));
+__MATHCALL (jn,, (int, _Mdouble_));
+__MATHCALL (y0,, (_Mdouble_));
+__MATHCALL (y1,, (_Mdouble_));
+__MATHCALL (yn,, (int, _Mdouble_));
+#endif
 
 
+#if defined __USE_MISC || defined __USE_XOPEN || defined __USE_ISOC9X
 /* Error, gamma, and Bessel functions.  */
 __MATHCALL (erf,, (_Mdouble_));
 __MATHCALL (erfc,, (_Mdouble_));
 __MATHCALL (gamma,, (_Mdouble_));
-__MATHCALL (j0,, (_Mdouble_));
-__MATHCALL (j1,, (_Mdouble_));
-__MATHCALL (jn,, (int, _Mdouble_));
 __MATHCALL (lgamma,, (_Mdouble_));
-__MATHCALL (y0,, (_Mdouble_));
-__MATHCALL (y1,, (_Mdouble_));
-__MATHCALL (yn,, (int, _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
    store the value through it.  */
@@ -219,16 +237,14 @@ __MATHCALL (gamma,_r, (_Mdouble_, int *));
 __MATHCALL (lgamma,_r, (_Mdouble_, int *));
 #endif
 
+#endif /* Use misc or X/Open.  */
 
-#if defined(__USE_MISC) || defined(__USE_XOPEN_EXTENDED)
 
+#if defined __USE_MISC || defined __USE_XOPEN_EXTENDED || defined __USE_ISOC9X
 /* Return the integer nearest X in the direction of the
    prevailing rounding mode.  */
 __MATHCALL (rint,, (_Mdouble_ __x));
 
-/* Return X times (2 to the Nth power).  */
-__MATHCALL (scalb,, (_Mdouble_ __x, _Mdouble_ __n));
-
 /* Return X + epsilon if X < Y, X - epsilon if X > Y.  */
 __MATHCALL (nextafter,, (_Mdouble_ __x, _Mdouble_ __y));
 
@@ -236,4 +252,43 @@ __MATHCALL (nextafter,, (_Mdouble_ __x, _Mdouble_ __y));
 __MATHCALL (remainder,, (_Mdouble_ __x, _Mdouble_ __y));
 #endif
 
-#endif /* Use misc.  */
+#if defined __USE_MISC || defined __USE_XOPEN_EXTENDED
+/* Return X times (2 to the Nth power).  */
+__MATHCALL (scalb,, (_Mdouble_ __x, _Mdouble_ __n));
+#endif
+
+#ifdef __USE_ISOC9X
+/* Round X to integral valuein floating-point format using current
+   rounding direction, but do not raise inexact exception.  */
+__MATHCALL (nearbyint,, (_Mdouble_ __x));
+
+/* Round X to nearest integral value, rounding halfway cases away from
+   zero.  */
+__MATHCALL (round,, (_Mdouble_ __x));
+
+/* Round X to the integral value in floating-point format nearest but
+   not larger in magnitude.  */
+__MATHCALL (trunc,, (_Mdouble_ __x));
+
+/* Compute remainder of X and Y and put in *QUO a value with sign of x/y
+   and magnitude congruent `mod 2^n' to the magnitude of the integral
+   quotient x/y, with n >= 3.  */
+__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.  */
+__MATHCALL (fmax,, (_Mdouble_ __x, _Mdouble_ __y));
+
+/* Return maximum numeric value from X and Y.  */
+__MATHCALL (fmin,, (_Mdouble_ __x, _Mdouble_ __y));
+
+
+/* Classify given number.  */
+__MATHDECL_1 (int, __fpclassify,, (_Mdouble_ __value));
+
+/* Test for negative number.  */
+__MATHDECL_1 (int, __signbit,, (_Mdouble_ __value));
+#endif /* Use ISO C 9X.  */