[packagekit] Segfault

[Richard Hughes]

On Thu, 2007-11-08 at 18:48 +0000, Grzegorz Dąbrowski wrote:
> On Tue, 2007-11-06 at 07:54 +0000, Richard Hughes wrote:
> > On Tue, 2007-11-06 at 07:34 +0000, Richard Hughes wrote:
> > > It's probably also a good idea to not call Cancel in the GUI tools if we
> > > know we can't support it. I'll do this now.
> > 
> > Actually, I'm not undecided. If the backend can't Cancel, should the
> > PkClient "emulate" the cancel? i.e. ignore all output from the deamon
> > from that tid, and emit a finished canceled signal?
> > 
> > This would mean the thread is still running in the daemon, but the
> > client tools would all do the right thing without lots of if
> > (cancel_supported) calls.
> > 
> > Ideas?
> 
> It make sense only for searching operation.

Sure, see the problem on allow-interrupt in the developer documentation.

> The problem is when you try
> to cancel e.g. package installation. We need to interrupt a thread but I
> don't know how.

Me neither. I guess we can use g_thread_exit, just comparing some sort
of data index. I'll have a play later. Yell if anyone can help.

Richard.