On Wed, Mar 23, 2016 at 02:45:05PM +0100, Victor Toso wrote: > I would rather that we wait to see if we find more issues. At the > moment, we don't have any. The coroutine context is the only one so far > that it was necessary to complete-in-idle. The other on > spice_usbredir_channel_open_acl_cb() was quite simple (finalize before > calling the callback). > > Not having any issue to address and including this g-task-helper does > not make much sense for me. > Ah, I thought there was still some breakage without this patch. If everything seems fine at this point (including in valgrind), maybe it's indeed better to wait and see if we hit more issues before pushing this. Alternatively, we could introduce the helper now, but only use it in vmcstream.c to replace the current idle queueing. Christophe
Attachment:
signature.asc
Description: PGP signature
_______________________________________________ Spice-devel mailing list Spice-devel@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/spice-devel