On Sat, Dec 24, 2011 at 09:23:15AM +0100, Hans de Goede wrote: > Hi, > > First of all thanks for the review of this series! > > On 12/23/2011 06:10 PM, Christophe Fergeau wrote: > >Hi, > > > >On Mon, Dec 19, 2011 at 12:24:34PM +0100, Hans de Goede wrote: > >>This is a preparation patch for handling usb packet completion in a > >>separate thread. > > > >I haven't looked at the patches extending this, but I have 2 comments > >already: > >* the xmit_queue_lock + xmit_queue combination looks a lot like GAsyncQueue > > You may be right there, but I also need a lock to protect the variable > which protects the queue from new messages been queued by the usb event > handling > thread after disconnect (and the queue has been cleared). Right now I can > use the same lock for this, so sticking with the current code seems best.
Your call, there are g_async_queue_lock/_unlock functions so the xmit_queue_blocked check could be protected with that. > IOW the only interaction between the thread and the rest of spice-gtk > is the queuing of SpiceMsgOut messages, which is relatively simple and as > such should be quite safe. Yeah, I know the 2 things are quite separated, but now someone hacking on this code has to remember that sometimes we can be in a different thread, in some other parts of the code in a coroutine context. > > >Apart from this, the patch looks good. > > Can I consider that an ack? Yup Christophe
pgpk4u4L6ciXu.pgp
Description: PGP signature
_______________________________________________ Spice-devel mailing list Spice-devel@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/spice-devel
