10:54kubast2: https://github.com/skeggsb/nouveau which kernel version can I build this on ?
10:54kubast2: I didn't meant to send this
10:55kubast2: I presume > https://github.com/skeggsb/nouveau < this might not work/build on 5.0.10
10:55kubast2: so I'm grabbing linux-5.2 from skeggsb/nouveau
13:16karolherbst: Lyude: ...... I think I fixed runpm on my laptop....
13:16karolherbst: ohh, no, secboot failed again
14:49karolherbst: Lyude: at least with that pcie link stuff runpm works relibly... just secboot after resuming sometimes fails... but... I have issues with secure boot on that GPU anyway
15:54Lyude: karolherbst: nice! :D
16:04karolherbst: Lyude: I still don't feel entirely comfortable to rely on the pcie link setting code, but I guess we have no choice here ...
16:04karolherbst: especially because that's all kepler1 code
16:06Lyude: tbh, whatever makes it work
16:06Lyude: (also I'll be back in a bit, going to update my server to f30)
16:07karolherbst: Lyude: yeah... I think the code is fairly safe though as the pcie stuff isn't complicated, just a bit overcomplex to get it right
17:39karolherbst: Lyude: do you know how much the difference in reliability in glamor is between 1.19 and 1.20?
20:56karolherbst: Lyude: btw, do you think you'll find some time reviewing and testing my runpm issues? Would be interesting to see if that helps on other devices as well
20:56karolherbst: and doesn't regress already working ones
21:37Lyude: karolherbst: sure thing, remind me on monday and I should have the time to do it
22:19karolherbst: cool, thanks
22:19karolherbst: Lyude: ohh, btw, I will be in boston from the 10th to the 17th :p
23:52kubast2: Yeah I failed building clang, because my llvm-lw-git lacked a binary that is inluced with full blown version of llvm so no mesa-git