Re: NSStepperCell bug (fwd)

[Gregory John Casamento]

This bug has been addressed.   The issue was with NSStepperCell's initWithCoder
method.   This problem was apparent on systems using gcc 3.4.x and on
architectures other than intel using any gcc compiler version.

The specific problem was that in NSStepperCell an integer (32 bit) was being
unarchived into a BOOL (which is an unsigned char) which is always 8 bits, so
it was overrunning the memory buffer and corrupting something in the activation
record.  

Please update NSStepperCell.m (which is all you should really need to get past
this) or update GNUstep from CVS as a whole.

Kind thanks for the bug report.

Thanks, GJC

--- Nicola Pero <nicola@brainstorm.co.uk> wrote:

> James, thanks for your bug report!  It's probably something for the Gorm
> maintainer (Gregory) to look at.
> 
> Generally, the best way to report bugs is the page
> 
> http://savannah.gnu.org/bugs/?func=additem&group=gnustep
> 
> I suppose we need an easy link to that page on the web site, I agree it's
> difficult to find how to report bugs.
> 
> Thanks
> 
> ---------- Forwarded message ----------
> Date: Fri, 29 Oct 2004 11:04:45 -0500
> From: James Pelletier <jamesp@trdlnk.com>
> To: nicola@brainstorm.co.uk, richard@brainstorm.co.uk, fedor@gnu.org
> Subject: NSStepperCell bug
> 
> Greetings!
> 
> I'm trying to install GNUstep and I ran into a problem when trying to 
> run Gorm. I couldn't figure out how to report a bug, so I thought I'd 
> send you an email.
> 
> Basically, Gorm kept seg faulting when trying to load the controls 
> palette. After some digging I foud out that it was dying when decoding 
> the NSStepperCell. I looked at -[NSStepperCell initWithCoder:] and I 
> think that the type of tmp1 and tmp2 should be int and not BOOL. I think 
> that this was causing stack corruption, which explained my seg fault. 
> Once I changed the type, everything started working without any problem.
> I'm currently running on Solaris9/intel and using gcc 3.4.0, for what 
> it's worth.
> 
> -[NSStepperCell encodeWithCoder:] would probably need the same fix.
> 
> By the way, thanks for all the work you guys have done so far!
> 
> James Pelletier
> 
> 
> 
> _______________________________________________
> Bug-gnustep mailing list
> Bug-gnustep@gnu.org
> http://lists.gnu.org/mailman/listinfo/bug-gnustep
> 

=====
Gregory John Casamento 
-- CEO/President Open Logic Corp. (A Maryland Corporation)
#### Maintainer of Gorm for GNUstep.