memchr() patched

[Nick Clifton]

Hi Guys,

  This is just a quick note to say that I have applied the enclosed
patch to memchr().  It fixes a bug reported by a customer where they
were passing a large integer value as the character to be sought,
assuming that memchr would only look at the bottom byte of that value.
(This is the documented behaviour).

Cheers

    Nick

1998-12-15  Nick Clifton  <nickc@cygnus.com>

 * libc/string/memchr.c: Treat 'c' argument as an unsigned byte not
 a signed int.


Index: memchr.c
===================================================================
RCS file: /cvs/cvsfiles/devo/newlib/libc/string/memchr.c,v
retrieving revision 1.10.42.1
diff -p -r1.10.42.1 memchr.c
*** memchr.c    1998/06/03 22:05:05     1.10.42.1
--- memchr.c    1998/12/16 01:14:20
*************** _DEFUN (memchr, (src_void, c, length),
*** 74,79 ****
--- 74,81 ----
  #if defined(PREFER_SIZE_OVER_SPEED) || defined(__OPTIMIZE_SIZE__)
    _CONST unsigned char *src = (_CONST unsigned char *) src_void;

+   c &= 0xff;
+
    while (length--)
      {
        if (*src == c)
*************** _DEFUN (memchr, (src_void, c, length),
*** 88,93 ****
--- 90,97 ----
    unsigned long  mask;
    int i, j;

+   c &= 0xff;
+
    /* If the size is small, or src is unaligned, then
       use the bytewise loop.  We can hope this is rare.  */
    if (TOO_SMALL (length) || UNALIGNED (src))