VRR breakers
#7386
Replies: 2 comments
-
Thanks a bunch for the info (and the PR!). How would one deduce whether the cursor is locked or whether they render it themselves? |
Beta Was this translation helpful? Give feedback.
0 replies
-
With |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Multi-monitor
Multi-monitor doesn't work with nvidia (vrr activates but refresh rate is stuck at max) https://forums.developer.nvidia.com/t/feature-g-sync-freesync-under-wayland-session/220822/116?page=2. Others should work fine.
Animations
Any running animation will break VRR.
Cursor
DIfferent cursor setting combinations might break VRR. Frames scheduled by HL and AQ should be handled by
cursor:no_break_fs_vrr
. The breakage comes from excessive_CWlSurfaceCommit
calls. Those extra calls are observed when pointer frame is sent with lock constraint active or a HW cursor buffer is present with atomic modesetting.Games that use cursor lock usually hide the compositor cursor. They might still render cursor themselves. This is still counts as an invisible cursor.
Some games might switch between different modes, e.g. locked invisible during gameplay and unlocked visible in menus.
I don't know any games that might have locked != invisible.
Assuming
cursor:no_break_fs_vrr = true
andcursor:min_refresh_rate
less than the target framerate:Does anyone has a different experience? Are there any wayland compositors that have a working VRR with HW cursors and atomic modesetting?
Beta Was this translation helpful? Give feedback.
All reactions