[committed] Fix bug in unsupported target list in config.gcc

[Nathanael Nerode]

>The same uncertainty works for sysvxxx that *are* supported
>(i.e. aliases for sysv4) also, so I think that's a moot point.  But,
OK.

>I'm only worried about compatibility.  Do we use this construct
>elsewhere?
As far as I can tell, no.  Shall I check this in?

	* config.gcc: Don't use negated character class in shell case
	clause.

Index: config.gcc
===================================================================
RCS file: /cvs/gcc/gcc/gcc/config.gcc,v
retrieving revision 1.347
diff -u -r1.347 config.gcc
--- config.gcc	29 Aug 2003 17:32:17 -0000	1.347
+++ config.gcc	29 Aug 2003 20:31:56 -0000
@@ -246,7 +246,7 @@
 case $machine in
    alpha*-*-linux*libc1*		\
  | i[34567]86-sequent-sysv		\
- | i[34567]86-sequent-sysv[!4]*		\
+ | i[34567]86-sequent-sysv[123]*	\
  | i[34567]86-go32-*			\
  | i[34567]86-*-go32*			\
  | m68k-*-linux*aout*			\

-- 
Nathanael Nerode  <neroden at gcc.gnu.org>
http://home.twcny.rr.com/nerode/neroden/fdl.html