[Spice-devel] [RFC PATCH spice-server 0/2] RHEL7 improvements

[Frediano Ziglio]

> 
> Hi,
> 
> On Tue, Dec 06, 2016 at 12:28:31PM +0000, Frediano Ziglio wrote:
> > These 2 patches attempt to join images split by RHEL7 graphic
> > stack (Mesa) decreasing commands handled by spice-server.
> >
> > You can see the difference between the 2 video:
> > - https://www.youtube.com/watch?v=OarV6zUmUdg (before)
> > - https://www.youtube.com/watch?v=5fTdCCbFeCg (after)
> 
> Yay :)
> 
> > These video are realized with some additional changes:
> > - the statistics from the terminal have some additional
> >   "out_messages" counters. These counters show the messages
> >   sent to the client(s). These changes are part of my "stat"
> >   branch (partially sent couple of days ago);
> > - the replay utility, as you can see, was changed to replay
> >   using the real time to allow the video code (which is dependent
> >   on timing) to work correctly. The patch is currently not in
> >   a good shape (enough to be sent);
> > - the client utility was changed to remove the delay due to
> >   the lip sync. The glitches (present mostly before these patches)
> >   are much reduced.
> >
> > Note the number of commands sent to the client reduced from 6097
> > to 2016 (current year, just a coincidence).
> > The network traffic reduced from 72M to 56M. This is due to the fact
> > that having a single stream (as you can see VP8 codec was used) the
> > compression on the stream is better.
> >
> > These patches fix:
> > - https://bugzilla.redhat.com/show_bug.cgi?id=1158029;
> > - (probably) https://bugzilla.redhat.com/show_bug.cgi?id=1294564.
> 
> OT: I thought it might help with [0] but it does not.
> 
> [0] https://bugs.freedesktop.org/show_bug.cgi?id=94372
> 
> >
> > About the status of these patches they are not far from a good
> > shape however I think would be better to move these stuff into
> > DisplayChannel instead of RedWorker. Also probably would be
> > better to define a constant for the timeout. I think 10ms is
> > fine, perhaps some optimization like not waiting if the joined
> > command end at the screen bottom or if the last command contain
> > a small image (Mesa split at about 64K pixel, 256Kb).
> 
> I tested on Fedora and Windows guest and I think it is very easy to
> see/feel the improvement. Although my tests were simple, by playing
> videos and making some desktop effects that triggers the streaming
> detection.
> 
> I personally don't think this will be enough to enable streaming by
> default again - rhbz#1294564 - mainly because the streaming detection is
> yet trigger on situation that it shouldn't. It causes a noticeable delay
> as well see [1] where I think we could be using cached images instead of
> streaming.
> 
> (This video is without your patches but I reproduce the same think with
> them too)
> [1]
> https://people.freedesktop.org/~victortoso/videos/spice/bad-video-detection.webm
> 

Beside some small delay I don't see much problems in this video.
Yes, one of the patches (on client) removes this delay for gstreamer videos.

> I don't have much background on the qxl/streaming apart from testing it,
> to provide useful feedback in the code.
> 
> Cheers,
>   toso
> 

Frediano