Re: Instabilities with multi threaded application

On Sat, 2004-07-24 at 15:43, Soeren Sandmann wrote:
> Yes, calling g_idle/timeout_add from another thread is perfectly
> legal. Those functions will do their own locking, so
> gdk_threads_enter/leave() are no not needed. And yes, routing all gui
> related tasks though those functions is acceptable and what I'd
> usually recommend for using threads with GTK+.
> There is no need for gdk_threads_enter/leave() in the callbacks either
> since they will be called from the main gtk+ thread. Of course, if the
> callbacks are accessing data structures that are modified by other
> threads, then locking will be needed,
OK, thank you for your response, then it's the exact way i was doing.

I have one question remaining, will the g_idle_add sources attached in
the context be executed in a FIFO order? It would be nice to have how
they work explicitly in the docs, if you clear it out I can send a patch
to bugzilla.

> > I'm assuming this function came from somewhere else and had the name
> > g_strtokenize, it should be g_strsplit_set. Should i fill a bug report
> > about this?
> Yes, filing a bug at
> is the thing to do, but I'll just go ahead and fix this one. Thanks.
Well I have filled the bug report with the associated patch, don't
forget to mark it as fixed, here's the link:

[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]