Re: Does CVS gconf have known problems?
- From: Mark McLoughlin <mark skynet ie>
- To: Malcolm Tredinnick <malcolm commsecure com au>
- Cc: gconf-list gnome org, orbit-list gnome org
- Subject: Re: Does CVS gconf have known problems?
- Date: 07 Apr 2003 16:21:25 +1200
Hey,
On Mon, 2003-04-07 at 15:53, Mark McLoughlin wrote:
> Hi Malcolm,
>
> On Mon, 2003-04-07 at 15:44, Malcolm Tredinnick wrote:
> > On Sun, Apr 06, 2003 at 11:17:57PM -0400, Havoc Pennington wrote:
>
> > > > I have just been building recent versions of everything
> > > > (installed into a clean directories, so I know there is no extraneous
> > > > cruft interfering with things) and gconf is dumping core during the
> > > > install rule of eog. The message in /var/log/messages is telling me to
> > > > report a bug, but just filing "it dumped core" is not really helpful, so
> > > > I'll need to spend a bit of time working out what is going on on.
> > >
> > > If it reliably crashes, just shut it down (gconftool-2 --shutdown)
> > > then run a "gdb gconfd-2" from a terminal and get a backtrace.
> >
> > Yeah, I've got that already. It does not look blindingly helpful,
> > though, so I was going to dig deeper. However, I am not going to have
> > time to do this in the near future, so I have dropped bug #110154 in
> > your pile. I'll keep my b0rked setup around in case you want me to run
> > any tests to narrow this down. I'm sorry not to be able to get a bit
> > closer to the cause of the problem.
>
> Its a recent ORBit2 bug. I just saw it fly past in my bugzilla mailbox:
>
> http://bugzilla.gnome.org/show_bug.cgi?id=110140
>
> There was a recent commit in this area. I'll probably take a look
> shortly ...
Okay, so it seems to just be yet another manifestation of lurking
problems with ORBit2's build ... Michael added a field to an internal
structure and it seems some parts of ORBit2 weren't getting rebuilt to
take into account that change. I was seeing this:
(gdb) p *adaptor
$13 = {parent = {interface = 0x80b3c00, refs = 11}, lock = 0x0,
handle_request = 0x8073fd8 <ORBit_POA_handle_request>, adaptor_key = {_maximum = 20, _length = 134990156,
_buffer = 0x1 <Address 0x1 out of bounds>, _release = 0 '\000'}, thread_hint = ORBIT_THREAD_HINT_NONE}
(gdb) p *(ORBit_ObjectAdaptor) pobj->poa
$14 = {parent = {interface = 0x80b3c00, refs = 11}, handle_request = 0, adaptor_key = {
_maximum = 134692824, _length = 20, _buffer = 0x80bc94c "", _release = 1 '\001'},
thread_hint = ORBIT_THREAD_HINT_NONE}
Note the extra "lock" member in the first dump. So, I dunno how it
happens but if you do a make clean before rebuilding then things should
work just fine ...
Good Luck,
Mark.
[
Date Prev][
Date Next] [
Thread Prev][
Thread Next]
[
Thread Index]
[
Date Index]
[
Author Index]