[Spice-devel] [PATCH spice-gtk v2 9/8] main: don't update display timer for unchanged config

[Jonathon Jongsma]

On Thu, 2016-03-24 at 12:18 +0100, Marc-André Lureau wrote:
> Hi Jonathon
> 
> On Wed, Mar 23, 2016 at 8:58 PM, Jonathon Jongsma <jjongsma at redhat.com> wrote:
> > > It also solves/avoids problems with a guest running on wayland when the
> > > "resize-guest" property is TRUE.
> > > See bug https://bugzilla.redhat.com/show_bug.cgi?id=1266484 ;
> > > It avoids destroying the primary surface when the display configuration
> > > has not changed
> > 
> > hmm.  so, the suggestion from my previous email (only compare with previous
> > config and return early from update_display() if there's currently a pending
> > config message scheduled) would probably break this case again (if I
> > understand
> > the situation correctly). Do you know *why* we're repeatedly updating to the
> > same monitor config under wayland?
> 
> I don't fully understand your hypotetical issue. You said in practice
> it cannot happen. Do you want to investigate it further before we
> apply this fix? I'd apply now, as it fixes 2 existing problems ;)
> 
> thanks

Yes, it does seem to solve some bugs and the issue is basically hypothetical. So
as I said in my previous email, it probably shouldn't hold up the patch.

you can consider this an ACK.

Jonathon