09:36lplc: I have a question about "pixel blend mode" plane property. How should we handle PREMULTI when a display controller does not support it? AFAIS in the code, PREMULTI needs to be present and default, so we cannot have just PIXEL_NONE and COVERAGE... Any hints?
11:22mlankhorst: airlied: I also changed dmem code slightly to allow passing a cgroup, to ensure a BO only has 1 cgroup
12:41mlankhorst: airlied: https://gitlab.freedesktop.org/mlankhorst/kernel/-/commits/ttm-memcg-nouveau
12:46mlankhorst: Oops unrelated hunk, let me remove it
13:04jani: sima: airlied: this series https://lore.kernel.org/r/20250407-asyn-v13-0-b93ef83076c5@intel.com was merged to drm-misc-next just a day after the last drm-misc-next pull request. it's got i915 changes in it too. would it be okay for mlankhorst to send one more feature pull with that this round, so we could backmerge to i915. otherwise we'll have to wait until -rc1/-rc2 or so to sync them
14:39zamundaaa[m]: zzxyb: try to create an egl context and check if it's a software renderer I guess?