Thu Mar 14 06:01:07 1996 Roland McGrath <roland@charlie-brown.gnu.ai.mit.edu>

	* string/strnlen.c: New file.
	* string/Makefile (routines): Add strnlen.
	* string/string.h [__USE_GNU] (strnlen): Declare new function.
	[__OPTIMIZE__]: Define extern inline implementation of it.

	* stdlib/erand48_r.c: Use __mpn_construct_double instead of ldexp and
	addition, to avoid using anything from -lm.

Mon Mar  4 21:57:14 1996  Andreas Schwab  <schwab@issan.informatik.uni-dortmund.de>

	* sysdeps/unix/sysv/linux/m68k/Makefile: New file.

	* sysdeps/unix/sysv/linux/m68k/sysdep.h (SYS_ify): Redefine.
	(CALL_MCOUNT): New macro, empty unless [PROF].
	(ENTRY): Do CALL_MCOUNT just after the label.
	(JUMPTARGET): New macro.
	(SYSCALL_ERROR_HANDLER): Fix syntax.

Thu Mar 14 04:20:48 1996  Roland McGrath  <roland@charlie-brown.gnu.ai.mit.edu>

	* sysdeps/unix/sysv/linux/init-first.c (init): Call `__personality'
	function instead of using inline asm i386 syscall.
	* sysdeps/unix/sysv/linux/syscalls.list: Add personality syscall.

	* posix/unistd.h [__USE_BSD]: Declare usleep.

	* misc/efgcvt.c (MAXDIG): New macro computed from <floats.h> constants.
	(fcvt, ecvt): Use it for buffer size.

	* stdlib/drand48-iter.c (__drand48_iterate): Use u_int64_t instead of
	conditionalizing long vs long long.

	* stdlib/drand48-iter.c (__drand48_iterate): Don't check for null
	pointers; never return EFAULT.
	* stdlib/drand48_r.c (drand48_r): Likewise.
	* stdlib/erand48_r.c (erand48_r): Likewise.

1 files changed, 6 insertions, 3 deletions

diff --git a/misc/efgcvt.c b/misc/efgcvt.c
index 95b0b0d570..e8a05176e6 100644
--- a/misc/efgcvt.c
+++ b/misc/efgcvt.c
@@ -1,5 +1,5 @@
 /* [efg]cvt -- compatibility functions for floating point formatting.
-Copyright (C) 1995 Free Software Foundation, Inc.
+Copyright (C) 1995, 1996 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
@@ -19,13 +19,16 @@ Cambridge, MA 02139, USA.  */
 
 #include <stdio.h>
 #include <stdlib.h>
+#include <float.h>
+
+#define MAXDIG (DBL_DIG + DBL_MAX_10_EXP)
 
 char *
 fcvt (value, ndigit, decpt, sign)
      double value;
      int ndigit, *decpt, *sign;
 {
-  static char buf[100];
+  static char buf[MAXDIG];
 
   (void) fcvt_r (value, ndigit, decpt, sign, buf, sizeof buf);
 
@@ -37,7 +40,7 @@ ecvt (value, ndigit, decpt, sign)
      double value;
      int ndigit, *decpt, *sign;
 {
-  static char buf[100];
+  static char buf[MAXDIG];
 
   (void) ecvt_r (value, ndigit, decpt, sign, buf, sizeof buf);