Yes, sorry. I misunderstood the statement about "no noticeable difference, about 4% cpu usage with or without threaded-swap-wait.", and then failed to notice your second mention of it with respect to window movement.
I did not include window movement in my testing because the original reporter only mentioned glxgears. I will now retest with window movement, which is also mentioned a major problem for Nvidia users in https://gitlab.gnome.org/GNOME/mutter/merge_requests/168.
Yes, sorry. I misunderstood the statement about "no noticeable difference, about 4% cpu usage with or without threaded- swap-wait. ", and then failed to notice your second mention of it with respect to window movement.
I did not include window movement in my testing because the original reporter only mentioned glxgears. I will now retest with window movement, which is also mentioned a major problem for Nvidia users in https:/ /gitlab. gnome.org/ GNOME/mutter/ merge_requests/ 168.