RE: c/6861: Conflicting type name allowed

["Steve Cornett" <stevec at bullseye dot com>]

The following reply was made to PR c/6861; it has been noted by GNATS.

From: "Steve Cornett" <stevec@bullseye.com>
To: "Andrew Pinski" <pinskia@physics.uc.edu>
Cc: <gcc-gnats@gcc.gnu.org>
Subject: RE: c/6861: Conflicting type name allowed
Date: Wed, 29 May 2002 15:16:55 -0700

 Andrew,
 
 Thanks for your reply, however I disagree with your finding.
 
 A warning is too weak, and very much too weak if -pedantic is required to
 get the warning.  This syntax is not just a little bit off, it is completely
 invalid.  Is the type of "a" unsigned char, unsigned short or (signed)
 short?  Any of these seems plausible to me.  The bug is not that you cannot
 get a warning, but rather that a meaningless, erroneous syntax produces no
 error message.
 
 Thanks,
 Steve
 
 -----Original Message-----
 From: Andrew Pinski [mailto:pinskia@physics.uc.edu]
 Sent: Wednesday, May 29, 2002 3:00 PM
 To: stevec@bullseye.com
 Cc: gcc-gnats@gcc.gnu.org
 Subject: Re: c/6861: Conflicting type name allowed
 
 
 you can use `gcc -Wall -pedantic -c testil.c', which warns about it.
 testil.c:2: warning: long, short, signed or unsigned used invalidly for
 `a'
 collect2: ld returned 1 exit status
 
 On Wednesday, May 29, 2002, at 05:54 , stevec@bullseye.com wrote:
 
 > typedef unsigned char T;
 > T short a;
 >> How-To-Repeat:
 > gcc -c bug.c