[Bug 755554] gtk: Wait at most 5s for calling something on the main thread

[GStreamer (GNOME Bugzilla)]

https://bugzilla.gnome.org/show_bug.cgi?id=755554

--- Comment #3 from Sebastian Dröge (slomo) <slomo at coaxion.net> ---
(In reply to Nicolas Dufresne (stormer) from comment #2)
> Review of attachment 312076 [details] [review]:
> 
> ::: ext/gtk/gstgtkbasesink.c
> @@ +327,3 @@
> +  gboolean timeout, ret;
> +
> +  ret = ! !gst_gtk_invoke_on_main ((GThreadFunc)
> 
> Was there before, but why would you want to use ! ! obscureness here, while
> you can write readable code ?

Boolean normalization.

> ::: ext/gtk/gstgtkutils.c
> @@ +29,3 @@
>    GCond cond;
>    gboolean fired;
> +  gboolean timed_out;
> 
> Consistency, timeout in basesink vs timed_out here.

Why does basesink matter here? :) And timeout is already the variable we use as
out parameter.

> @@ +52,3 @@
> +    g_mutex_clear (&info->lock);
> +    g_cond_clear (&info->cond);
> +    g_free (info);
> 
> Using g_main_context_invoke_full() with a destroy notify would avoid the
> duplication.

No, because otherwise the struct is already freed after g_main_context_invoke()
while we still want to get the result from it.

> @@ +71,1 @@
> +  info = g_new0 (struct invoke_context, 1);
> 
> Slice new please. I'd suggest having a new/free which will be clean with the
> invoke_full(),

No, GSlice is deprecated in newer GLib :)

-- 
You are receiving this mail because:
You are the QA Contact for the bug.
You are the assignee for the bug.