[10:26] <alan_g> alf: I've got lost in our kms code (it all looks unfamiliar). Any chance you could take a quick look at lp:1489806 in case there's something obvious to you about what happens?
[10:27] <alf> alan_g: let me see if I can reproduce
[10:45] <alf> alan_g: OK, I can reproduce. Removing the fatal error, doesn't help, since we don't crash but don't render either, which indicates that this is not a soft failure.
[10:46] <alan_g> alf: I got something: KMSPageFlipper::schedule_flip() is failing with "Permission denied"
[10:54] <alf> alan_g: let me check something ...
[10:55] <alan_g> alf: thanks
[11:12] <alf> alan_g: Found a workaround, trying to figure out the core cause. Do you want me to take over?
[11:13] <alan_g> alf: if you've got the time
[11:14] <alan_g> You've seen the link to lp:1489689?
[11:15] <alf> alan_g: Yes, can't try it out that moment though. From what I have seen, it doesn't seem related.
[11:16] <alan_g> np just thought there might be a clue in there
[15:36] <anpok> alan_g, alf: o_O it happens for me too
[15:36] <anpok> but --launch-client works btw..
[15:36] <anpok> whats the difference in behavior
[15:37] <alan_g> --launch-client uses system()
[15:37] <alan_g> anpok: what driver is that?
[15:39] <anpok> i915 kernel and i965 user space
[15:55] <alan_g> :(  - I was hoping for a new data point.
[15:56] <alan_g> alf anpok could someone confirm lp:1489806?
[15:57] <alf> alan_g: done