You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I'm still trying to figure out exactly what is causing this. I've noticed it testing pop-os/cosmic-workspaces-epoch#101. It seems to work fine on anvil and sway (using the mock-backend feature) so this appears to be an issue in cosmic-comp. With Smithay/client-toolkit#477 I'm able to reproduce this in a more minimal case; again working on other compositors.
This only happens with multiple monitors, doesn't happen in anvil, and is inconsistent, so I think it's some race condition related to multiple kms threads trying to draw the same surface.
COSMIC_DISABLE_DIRECT_SCANOUT=1 seemed to fix this in cosmic-workspaces, but in the sctk example I've seen it with that as well. So maybe that's just changing the timing of a race condition...
In the sctk example, when a drag doesn't show a drag surface, moving the cursor to another output sometimes, but not always, makes it appear. Regardless, the drag works, it's just the surface isn't rendering.
This doesn't seem to be related to the damage returned by the WaylandSurfaceRenderElement, since I still see it after changing that to always be non-empty.
The text was updated successfully, but these errors were encountered:
I'm still trying to figure out exactly what is causing this. I've noticed it testing pop-os/cosmic-workspaces-epoch#101. It seems to work fine on anvil and sway (using the
mock-backend
feature) so this appears to be an issue in cosmic-comp. With Smithay/client-toolkit#477 I'm able to reproduce this in a more minimal case; again working on other compositors.This only happens with multiple monitors, doesn't happen in anvil, and is inconsistent, so I think it's some race condition related to multiple kms threads trying to draw the same surface.
COSMIC_DISABLE_DIRECT_SCANOUT=1
seemed to fix this in cosmic-workspaces, but in the sctk example I've seen it with that as well. So maybe that's just changing the timing of a race condition...In the sctk example, when a drag doesn't show a drag surface, moving the cursor to another output sometimes, but not always, makes it appear. Regardless, the drag works, it's just the surface isn't rendering.
This doesn't seem to be related to the damage returned by the
WaylandSurfaceRenderElement
, since I still see it after changing that to always be non-empty.The text was updated successfully, but these errors were encountered: