update from main archive 960909cvs/libc-960910

Mon Sep  9 20:31:27 1996  Ulrich Drepper  <drepper@cygnus.com>

	* version.c (banner): Report to bug-glibc@prep not @gnu.
	Reported by Andreas Jaeger.

	* libio/stdio.h [!__USE_REENTRANT]: Don't define getc as
 	_IO_getc_unlocked.
	[__USE_REENTRANT]: Don't define

Mon Sep  9 15:59:32 1996  Ulrich Drepper  <drepper@cygnus.com>

	* sysdeps/i386/strrchr.c: Fix bug where NUL byte is not recognized
	when it is the third byte in the string.  Reported by
	NIIBE Yutaka.
	* string/tester.c: Add tests for above bug to strlen, strchr, and
	strrchr tests.

Fri Sep  6 21:23:33 1996  Andreas Schwab  <schwab@issan.informatik.uni-dortmund.de>

	* sysdeps/m68k/fpu/e_acos.c: Deansideclized.  Include
	"math_private.h" to get prototype.
	* 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/k_sin.c: Remove uneeded indirection of
	__m81_u.
	* sysdeps/m68k/fpu/k_tan.c: Likewise.

Sun Sep  1 18:01:35 1996  Andreas Schwab  <schwab@issan.informatik.uni-dortmund.de>

	* sysdeps/m68k/fpu/__math.h: Check for __NO_M81_MATH_INLINES
	instead of __NO_MATH_INLINES.
	(__M81_MATH_INLINES): Renamed from __MATH_INLINES.
	* sysdeps/m68k/fpu/e_acos.c: Define __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.
	* 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_ldexp.c: Likewise.
	* sysdeps/m68k/fpu/s_modf.c: Likewise.
	* sysdeps/m68k/fpu/k_cos.c: Use inline version of __cos.
	* sysdeps/m68k/fpu/k_sin.c: Likewise.
	* sysdeps/m68k/fpu/k_tan.c: Likewise.

Mon Sep  9 12:55:14 1996  Ulrich Drepper  <drepper@cygnus.com>

	* nss/nss_files/files-parse.c (INT_FIELD_MAYBE_NULL): Check for
	available character before converting number and return with
	error if none is available.

Fri Sep  6 22:09:08 1996  Andreas Schwab  <schwab@issan.informatik.uni-dortmund.de>

	* sysdeps/unix/sysv/linux/init-first.h: New file, generic
	implementation for architectures where parameters are passed on
	stack.
	* sysdeps/unix/sysv/linux/m68k/init-first.h: Removed.
	* sysdeps/unix/sysv/linux/i386/init-first.h: Removed.

Fri Sep  6 22:05:32 1996  Andreas Schwab  <schwab@issan.informatik.uni-dortmund.de>

	* configure.in: Remove unneeded sysnames frobbing.

Mon Sep  9 05:29:09 1996  Ulrich Drepper  <drepper@cygnus.com>

	* time/zic.c: Update from ADO 96l.
	* time/africa: Update from ADO 96k.
	* time/antarctica: Likewise.
	* time/asia: Likewise.
	* time/australia: Likewise.
	* time/europe: Likewise.
	* time/northamerica: Likewise.
	* time/southamerica: Likewise.

Mon Sep  9 05:03:47 1996  NIIBE Yutaka  <gniibe@mri.co.jp>

	* sysdeps/unix/sysv/linux/i386/init-first.h: Work around buggy
	gcc <= 2.7.2.1 which optimizes away address operations on weak
	objects.

	* nss/nss_files/files-parse.c (INT_FIELD_MAYBE_NULL): New macro.
	Like INT_FIELD but also accept non-existing field.
	* shadow/sgetspent_r.c (LINE_PARSER): Use INT_FIELD_MAYBE_NULL
	for last four fields.
	* shadow/putspent.c (putspent): Print `:' after numeric value.

Mon Sep  9 02:42:48 1996  Richard Henderson  <rth@tamu.edu>

	* libio/stdio.h (BUFSIZ): Define using _IO_BUFSIZ.

Sun Sep  8 16:43:56 1996  Ulrich Drepper  <drepper@cygnus.com>

	* Make-dist: Filter out $(add-ons) from $(subdirs).
	Don't distribute TAGS files.
	($(tardir).tar): Remove dist.tar when all is done.
	* manual/Makefile (distribute): Add libc.info*.
	* libio/Makefile (distribute): Add Banner.
	* inet/Makefile (headers): Add netinet/tcp.h.

	* resolv/Makefile (libresolv-routines): Add getnetnamadr.

	* crypt-README: Removed.
	* gnu-stabs.h: Removed.

	* sysdeps/alpha/copysign.S: Renamed to...
	* sysdeps/alpha/s_copysign.S: ...this.
	* sysdeps/alpha/fabs.S: Renamed to...
	* sysdeps/alpha/s_fabs.S: ...this.

	* sysdeps/m68k/isinfl.c: Renamed to...
	* sysdeps/m68k/s_isinfl.c: ...this.
	* sysdeps/m68k/isnanl.c: Renamed to...
	* sysdeps/m68k/s_isnanl.c: ...this.

	* sysdeps/sparc/sqrt.c: Renamed to...
	* sysdeps/sparc/e_sqrt.c: ...this.  Function name now is
	__ieee754_sqrt.

	* sysdeps/generic/get_str.c: Removed.

	* sysdeps/ieee754/cbrt.c: Removed.
	* sysdeps/ieee754/drem.c: Removed.
	* sysdeps/ieee754/logb.c: Removed.
	* sysdeps/ieee754/sqrt.c: Removed.

	* sysdeps/stub/cbrt.c: Removed.
	* sysdeps/stub/cos.c: Removed.
	* sysdeps/stub/drem.c: Removed.
	* sysdeps/stub/isinf.c: Removed.
	* sysdeps/stub/isinfl.c: Removed.
	* sysdeps/stub/isnanl.c: Removed.
	* sysdeps/stub/logb.c: Removed.
	* sysdeps/stub/sin.c: Removed.
	* sysdeps/stub/sqrt.c: Removed.

	* sysdeps/tahoe/log10.c: Removed.

	* sysdeps/vax/index.s: Removed.
	* sysdeps/vax/rindex.s: Removed.
	* sysdeps/vax/bcmp.s: Removed.
	* sysdeps/vax/log10.c: Removed.
	* sysdeps/vax/infnan.c: Removed.
	* sysdeps/vax/Dist: Add fl.h.

	* sysdeps/unix/sysv/linux/alpha/Dist: Add llseek.S.

	* inet/rcmd.c (rcmd): Make messages more uniform.

Sun Sep  8 14:15:42 1996  Ulrich Drepper  <drepper@cygnus.com>

	* po/de.po: Update.
	* po/es.po: Update.
	* po/ko.po: Update.
	* po/pl.po: New file.
	* po/nl.po: New file.

	take care for a possible clash.  Nobody will name the domain

1 files changed, 0 insertions, 213 deletions

diff --git a/sysdeps/generic/get_str.c b/sysdeps/generic/get_str.c
deleted file mode 100644
index 182815ee18..0000000000
--- a/sysdeps/generic/get_str.c
+++ /dev/null
@@ -1,213 +0,0 @@
-/* __mpn_get_str -- Convert a MSIZE long limb vector pointed to by MPTR
-   to a printable string in STR in base BASE.
-
-Copyright (C) 1991, 1992, 1993 Free Software Foundation, Inc.
-
-
-This file is part of the GNU C Library.  Its master source is NOT part of
-the C library, however.  This file is in fact copied from the GNU MP
-Library and its source lives there.
-
-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., 675 Mass Ave,
-Cambridge, MA 02139, USA.  */
-
-#include "gmp.h"
-#include "gmp-impl.h"
-
-/* Convert the limb vector pointed to by MPTR and MSIZE long to a
-   char array, using base BASE for the result array.  Store the
-   result in the character array STR.  STR must point to an array with
-   space for the largest possible number represented by a MSIZE long
-   limb vector + 1 extra character.
-
-   The result is NOT in Ascii, to convert it to printable format, add
-   '0' or 'A' depending on the base and range.
-
-   Return the number of digits in the result string.
-   This may include some leading zeros.
-
-   The limb vector pointed to by MPTR is clobbered.  */
-
-size_t
-__mpn_get_str (str, base, mptr, msize)
-     unsigned char *str;
-     int base;
-     mp_ptr mptr;
-     mp_size_t msize;
-{
-  mp_limb big_base;
-#if UDIV_NEEDS_NORMALIZATION || UDIV_TIME > 2 * UMUL_TIME
-  int normalization_steps;
-#endif
-#if UDIV_TIME > 2 * UMUL_TIME
-  mp_limb big_base_inverted;
-#endif
-  unsigned int dig_per_u;
-  mp_size_t out_len;
-  register unsigned char *s;
-
-  big_base = __mp_bases[base].big_base;
-
-  s = str;
-
-  /* Special case zero, as the code below doesn't handle it.  */
-  if (msize == 0)
-    {
-      s[0] = 0;
-      return 1;
-    }
-
-  if ((base & (base - 1)) == 0)
-    {
-      /* The base is a power of 2.  Make conversion from most
-	 significant side.  */
-      mp_limb n1, n0;
-      register int bits_per_digit = big_base;
-      register int x;
-      register int bit_pos;
-      register int i;
-
-      n1 = mptr[msize - 1];
-      count_leading_zeros (x, n1);
-
-	/* BIT_POS should be R when input ends in least sign. nibble,
-	   R + bits_per_digit * n when input ends in n:th least significant
-	   nibble. */
-
-      {
-	int bits;
-
-	bits = BITS_PER_MP_LIMB * msize - x;
-	x = bits % bits_per_digit;
-	if (x != 0)
-	  bits += bits_per_digit - x;
-	bit_pos = bits - (msize - 1) * BITS_PER_MP_LIMB;
-      }
-
-      /* Fast loop for bit output.  */
-      i = msize - 1;
-      for (;;)
-	{
-	  bit_pos -= bits_per_digit;
-	  while (bit_pos >= 0)
-	    {
-	      *s++ = (n1 >> bit_pos) & ((1 << bits_per_digit) - 1);
-	      bit_pos -= bits_per_digit;
-	    }
-	  i--;
-	  if (i < 0)
-	    break;
-	  n0 = (n1 << -bit_pos) & ((1 << bits_per_digit) - 1);
-	  n1 = mptr[i];
-	  bit_pos += BITS_PER_MP_LIMB;
-	  *s++ = n0 | (n1 >> bit_pos);
-	}
-
-      *s = 0;
-
-      return s - str;
-    }
-  else
-    {
-      /* General case.  The base is not a power of 2.  Make conversion
-	 from least significant end.  */
-
-      /* If udiv_qrnnd only handles divisors with the most significant bit
-	 set, prepare BIG_BASE for being a divisor by shifting it to the
-	 left exactly enough to set the most significant bit.  */
-#if UDIV_NEEDS_NORMALIZATION || UDIV_TIME > 2 * UMUL_TIME
-      count_leading_zeros (normalization_steps, big_base);
-      big_base <<= normalization_steps;
-#if UDIV_TIME > 2 * UMUL_TIME
-      /* Get the fixed-point approximation to 1/(BIG_BASE << NORMALIZATION_STEPS).  */
-      big_base_inverted = __mp_bases[base].big_base_inverted;
-#endif
-#endif
-
-      dig_per_u = __mp_bases[base].chars_per_limb;
-      out_len = ((size_t) msize * BITS_PER_MP_LIMB
-		 * __mp_bases[base].chars_per_bit_exactly) + 1;
-      s += out_len;
-
-      while (msize != 0)
-	{
-	  int i;
-	  mp_limb n0, n1;
-
-#if UDIV_NEEDS_NORMALIZATION || UDIV_TIME > 2 * UMUL_TIME
-	  /* If we shifted BIG_BASE above, shift the dividend too, to get
-	     the right quotient.  We need to do this every loop,
-	     since the intermediate quotients are OK, but the quotient from
-	     one turn in the loop is going to be the dividend in the
-	     next turn, and the dividend needs to be up-shifted.  */
-	  if (normalization_steps != 0)
-	    {
-	      n0 = __mpn_lshift (mptr, mptr, msize, normalization_steps);
-
-	      /* If the shifting gave a carry out limb, store it and
-		 increase the length.  */
-	      if (n0 != 0)
-		{
-		  mptr[msize] = n0;
-		  msize++;
-		}
-	    }
-#endif
-
-	  /* Divide the number at TP with BIG_BASE to get a quotient and a
-	     remainder.  The remainder is our new digit in base BIG_BASE.  */
-	  i = msize - 1;
-	  n1 = mptr[i];
-
-	  if (n1 >= big_base)
-	    n1 = 0;
-	  else
-	    {
-	      msize--;
-	      i--;
-	    }
-
-	  for (; i >= 0; i--)
-	    {
-	      n0 = mptr[i];
-#if UDIV_TIME > 2 * UMUL_TIME
-	      udiv_qrnnd_preinv (mptr[i], n1, n1, n0, big_base, big_base_inverted);
-#else
-	      udiv_qrnnd (mptr[i], n1, n1, n0, big_base);
-#endif
-	    }
-
-#if UDIV_NEEDS_NORMALIZATION || UDIV_TIME > 2 * UMUL_TIME
-	  /* If we shifted above (at previous UDIV_NEEDS_NORMALIZATION tests)
-	     the remainder will be up-shifted here.  Compensate.  */
-	  n1 >>= normalization_steps;
-#endif
-
-	  /* Convert N1 from BIG_BASE to a string of digits in BASE
-	     using single precision operations.  */
-	  for (i = dig_per_u - 1; i >= 0; i--)
-	    {
-	      *--s = n1 % base;
-	      n1 /= base;
-	      if (n1 == 0 && msize == 0)
-		break;
-	    }
-	}
-
-      while (s != str)
-	*--s = 0;
-      return out_len;
-    }
-}