Re: mutt_free_header -> free -> mutt_sort_headers -> segfault
Hello.
On Fri, Nov 07, 2003 at 01:50:19PM +0100, Mads Martin Joergensen wrote:
> * Thomas Roessler <roessler@xxxxxxxxxxxxxxxxxx> [Nov 07. 2003 10:59]:
> > On 2003-11-05 20:09:18 -0000, q4xk3j002@xxxxxxxxxxxxxx wrote:
> >
> > > what next?
> >
> > Good question. At this point, the same code produces a NULL pointer
> > dereference, but works for me. That doesn't make it terribly easy
> > to debug.
> >
> > Is anyone else able to reproduce the problem and narrow it down?
IIRC, you'll have better chance to reproduce this bug if you set
sorting mode to threads, because this bug has been there since
incremental-threading code came in (note: I'm not trying to blame
the incremental-threading code or the person who implemented it).
But I haven't managed to fix this anyway.
> Did anyone produce a mailbox to reproduce with, or do we have to use gdb
> still?
I believe it's not a mailbox-specific, and once you managed to get a
usable core dump,
$ gdb /path/to/mutt/with/debug/symbol core