01:00 <mashrafi> @chadmed Are the new M1 GPU drivers working on Gentoo? https://asahilinux.org/2023/03/road-to-vulkan/

07:46 <chadmed> mashrafi: yup, theyre behind ~arm64 keyword for now in keeping with the upstream -edge semantics

07:46 <chadmed> once the dcp pm issue is sorted ill probably stabilise this release since ive been running it all naughtily with no issues for a while now

07:49 <jannau> chadmed: did you run into the same issue as lina? https://github.com/AsahiLinux/linux/commit/bebe37f8e5dbc66fe9dba6667edf0dec623eccc8

07:50 <chadmed> if dcp shuts off (e.g. plasma turns off the screen due to inactivity) its dead until the next reboot

07:52 <jannau> it's the same tags as in asahi, I used ~arm64 as per (old?) gentoo practice to make version bumps unstable

07:54 <jannau> chadmed: so the same issue I tried to debug with mkurz in #asahi 11 hours ago?

07:57 <jannau> I've never seen that. Do you use the macbook without power connected?

08:12 <chadmed> jannau: sorry it happened again :D

08:13 <chadmed> looks like it might be a kde issue, i did loginctl reboot via ssh and the screen came back alive once plasma was killed

08:13 <chadmed> very odd

08:14 <chadmed> i had it plugged into power today yeah but it seems to happen on batt too

08:14 <chadmed> it also seems to depend on how long the screen is off for. it's definitely not sleeping because ive explicitly told the machine not to sleep

08:14 <jannau> mkurz said that `kwin_wayland --replace` fixes it as well

08:15 <chadmed> hrrrmmm

08:15 <chadmed> ill try that next time i notice it

08:15 <chadmed> im home from work now so it wont happen :P

08:15 <chadmed> and yes ~arm64 was the right keyword to use for now, if we dont get any complaints about broken stuff from anyone ill shift it to arm64 to track what's going on with the upstream asahi repo

08:16 <jannau> I'm still on kwin 5.26.5 on j314

08:16 <chadmed> it only started when i upgraded to 5.27.3, however that did coincide with when i started testing new kernels pre-release

08:17 <chadmed> im pretty sure its kwin at this stage though, given that the screen comes back when it's killed by logind

08:17 <jannau> it's a little strange that switching to a different VT doesn't fix it

08:17 <chadmed> it might be kwin hogging input events? i couldnt even "ghost" change to a new tty

08:21 <jannau> if the vt switch doesn't work there's at least an explanation why that doesn't bring the display

08:21 <jannau> have you tried loginctl lock/unlock session?

08:31 <jannau> can you try pressing esc the next time it happens. My mbp 13" with kwin 5.27.3 just woke up with a black lock screen but with a visible mouse pointer and I could still switch VTs. restored after loginctl lock-session, loginctl unlock and pressing esc once or twice

08:35 <jannau> not sure if it's config or new behaviour. kde lock screen shuts the display down on esc with kwin-5.27.3

08:42 <jannau> strace or gdb on kwin the next time this happens might be interesting as well

09:05 <chadmed> pressing esc reliably brings the screen back up

09:08 <chadmed> after ~1hr of xonotic on the mini and leaving the macbook aline

09:08 <chadmed> alone*

09:09 <_jannau_> but other input events did not?

09:11 <chadmed> didnt try that time, ill let it lock again and come back in a bit

11:19 <chadmed> i cant seem to repro it at all now?

17:23 <jannau> chadmed: reproduced locally with kwin 5.27.3. Esc itself doesn't help

17:31 <jannau> looks very much like an kwin issue, possible a regression in 5.27.3

20:40 <ahdeb[m]> Glanzmann (IRC): If it's not too much trouble I'd appreciate if you could add CONFIG_RTW88_8821CU=m to future kernel builds (maybe enable other rtw88 usb variants while at it, they are new in 6.2).

20:43 <ahdeb[m]> CONFIG_RTW88_8822CU=m actually.