GNOME's Dynamic Triple Buffering Now Latency Optimized For Raspberry Pi & X.Org

While back in December the GNOME dynamic triple buffering was self-proclaimed to be "ready to merge", so far that hasn't happened yet. With the GNOME 46 feature freeze scheduled for 10 February, it remains to be seen if this long-worked-on dynamic triple/double buffering will be ready in time for this six month release. In any event, this past week saw a new optimization queued for this code.

Ubuntu desktop engineer Daniel van Vugt with Canonical has added a new optimization for the dynamic triple buffering feature. The new optimization is for avoiding excessive latency under the X.Org session and also benefits some niche platforms like GNOME on the Raspberry Pi single board computers.

The optimization is around optimizing latency for platforms lacking TIMESTAMP_QUERY support. Up to now this could lead to triple buffering to be used when not needed for platforms lacking TIMESTAMP_QUERY support. Van Vugt noted in that new patch:

"This makes a visible difference to the latency when dragging windows in Xorg, but will also help Wayland sessions on platforms lacking TIMESTAMP_QUERY such as Raspberry Pi."

That optimization patch is now part of the three-year-old merge request for adding the dynamic triple buffering support to Mutter. We'll see if by chance the feature is merged in time for GNOME 46 prior to 10 February or if it will be held off another cycle and remain patched-in for the likes of Ubuntu 24.04 LTS this spring.