Re: [RFC/RFA] GDB crash when using command lines due to memory corruption

[Daniel Jacobowitz <drow@mvista.com>]

On Thu, Dec 12, 2002 at 11:04:06AM +0100, Joel Brobecker wrote:
> > A safer change for 5.3 might be the patch I submitted on October 30th.
> > 
> > 	http://sources.redhat.com/ml/gdb-patches/2002-10/msg00586.html
> 
> For the record, this patch fixes a few regressions (apart from the
> original problem).  With GNAT as the compiler (still 2.8.1 based):
> 
> Before:
> 
>         # of expected passes            8305
>         # of unexpected failures        47
>         # of unexpected successes       10
>         # of expected failures          167
>         # of untested testcases         3
>         # of unsupported tests          2
> 
> After:
> 
>         # of expected passes            8309
>         # of unexpected failures        38
>         # of unexpected successes       10
>         # of expected failures          172
>         # of untested testcases         3
>         # of unsupported tests          2
> 
> (thanks Klee for pointing out at your patch)

I can't think what in our testsuite could be affected by this.  Are you
sure they're real changes?  Which tests?  Also note that both passes
and expected failures went up, which is fishy.

-- 
Daniel Jacobowitz
MontaVista Software                         Debian GNU/Linux Developer