00:50 <steev> bamse: in your c630 next branch you revert a commit - wasn't there a patch posted (at some point) that fixes that?

00:51 <bamse> steev: the drm scheduler thing?

00:51 <steev> yeah

00:51 <steev> 5.16-rc1 runs into it

00:51 <bamse> okay, i thought there was something on the list for that...but failed to follow up on it

00:52 <steev> i'll search my mail, because i swear i saw it too but figured that was in next for 17

00:59 <steev> ah, somewhere in drm tip

00:59 <steev> it's drm/scheduler: fix drm_sched_job_add_implicit_dependencies harder

01:00 <steev> hm, no

01:01 <steev> i mean, yes, but it needs some other commit too i guess

01:02 <bamse> if you have it infront of you, would you mind sending a reply to that patch saying that it break our boards? i presume you have some oops you could share?

01:02 <steev> i do, i'll have to cut the log down a bit, but i mean that patch is the one that should fix it

01:05 <steev> just testing that first

01:06 <bamse> that would be nice, but if so that needs to end up in 5.16-rc

01:20 <steev> ah, okay, 4eaf02d60 and 963d0b356 in next fix it (applied in that order)

01:21 <bamse> and they are targeting 5.17? or 5.16?

01:21 <steev> i have no idea how to tell that, christian koenig said that they are pushed to drm-misc-fixes in drm-tip but i have no idea what that means

01:22 <steev> so i shall reply to that mail and find out

02:21 <steev> aside from that minor issue, 5.16 rc1 appears to work well (minus audio of course)

12:05 <srinik> calebccff: if you get a chance can you pick last two patches from this https://git.linaro.org/people/srinivas.kandagatla/linux.git/log/?h=c630/v5.15 and try

15:20 <srinik> steev: I got few patches at https://git.linaro.org/people/srinivas.kandagatla/linux.git/log/?h=c630/v5.15 + ucm changes at https://git.linaro.org/people/srinivas.kandagatla/alsa-ucm-conf.git/log/?h=v1.2.5/c630

15:20 <steev> srinik: already pulled in the patches, will try the ucm changes in a bit

15:20 <srinik> steev: do you have a 5.16-rc1 based branch somewhere I can try

15:20 <steev> sure, one moment

15:22 <steev> https://github.com/steev/linux/commits/c630-5.16.0-rc1

15:27 <srinik> steev: thanks

16:01 <steev> srinik: the kernel config i'm using is based off the distro_defconfig (i add some inconsequentials, the new ksmb stuff and damo things)

16:08 <srinik> steev: its good for me as long as I can get UI up.. I tested it, and it works for me

16:17 <steev> srinik: i get.... two audio pops

16:17 <steev> well, audible, not audio

16:18 <steev> and "aplay -D hw:0,1 audio/africa-toto.wav" doesn't error, but there is also no audio output

16:18 <steev> oh

16:19 <steev> https://paste.debian.net/1219959/

16:21 <steev> srinik: i keep getting SLIM_RX2 port is busy

16:27 <srinik> steev: can you delete the asound.state file, alsactl might be configuring RX2 even before ucm starts setting up things

16:29 <steev> sure

16:29 <srinik> steev: new ucm configs have changed a bit so old asound.state might be not a good starting point

16:32 <steev> rm'd /var/lib/asound.state, powered the machine off so it wouldn't save one from the reboot and... aplay plays, however canberra-gtk-play -i desktop-login just does the audio pops

16:33 <steev> i just have an ogg file setup to play at login, mostly so i know that audio works that boot, and if it doesn't play i know to reboot

16:34 <steev> https://paste.debian.net/1219964/

17:00 <srinik> steev: looks like the mixers are not setup, can you try to add this debug https://paste.debian.net/1219969/ patch and see if the mixers are actually getting setup correctly

17:04 <srinik> steev: am also on debian 11, I will try the same usecase of adding a ogg file at login and see if i can reproduce the issue

17:04 <srinik> steev: am sure you replaced the ucm2 files with new ones..

17:05 <steev> si, they're in there, there was a command, how can i tell which one is being used? i don't recall

17:06 <steev> i'm thinking not the right one now, because gnome's sound settings only show speaker output, no mic

17:10 <steev> https://salsa.debian.org/steev/alsa-ucm-conf is the alsa-ucm-conf package i'm using, it's 1.2.5.1 with just the 2 patches from the repo you gave me on top

17:11 <steev> i'm guessing it's the module symlink? lrwxrwxrwx 1 root root 30 Nov 9 16:04 snd_soc_sdm845.conf -> ../Qualcomm/sdm845/sdm845.conf

17:26 <steev> srinik: with patch applied - https://paste.debian.net/1219973/

17:52 <steev> srinik: also, for what it's worth, up until sid/unstable switched over to pipewire/wireplumber, sound was working (mostly) fine

17:59 <ndec> i thought pipewire was supposed to fix all audio problems!

18:27 <steev> it does.... something

19:16 <steev> my understanding was that pipewire was meant to make it so i could screenshare easily in wayland, no idea where audio came in

20:27 <calebccff> srinik: just gave those patches a spin, no more crashes!, just these messages; https://p.calebs.dev/irozinubam.md, is that expected?

20:30 <steev> calebccff: i had that too, see the above where he mentions ditching the existing asound.state

20:31 <calebccff> steev: hmm, I'm running AOSP / tinyalsa on my device

20:31 <steev> do they not have an asound.state? i haven't touched android in quite a few years

20:33 <calebccff> steev: I don't think so, Android does a lot of it's own custom audio stuff

20:33 <steev> https://git.linaro.org/people/srinivas.kandagatla/alsa-ucm-conf.git/commit/?h=v1.2.5/c630&id=093688df3b1926aaba56a8d560648ca1e9b0eba1 maybe that gives some insight?

20:34 <calebccff> i would guess that my UCM configs are mostly utter nonsense, I'll take alook through that thanks

20:35 <steev> sadly... i'm not much help, i can't really wrap my head around all these things. i've been fighting with getting audio working for quite a while with no luck. it works on debian 11, but breaks on newer, which seems to be par for the course with audio, every update breaks something

20:42 <calebccff> steev: yeah I have not had the best of luck with UCM, we've resorted to just shipping our own fork of ucm2 in postmarketOS to avoid breakages, i have noo idea what most of these mixer controls are

20:43 <steev> they are obvious INT0 MIX2 INP, duh

20:44 <steev> the part that frustrates me is when i'm being told "as an end user, you shouldn't be worrying about that"

20:44 <steev> well yes, you're right, i shouldn't as an end user, but if there are so many moving parts, why aren't they documented properly, so that troubleshooting is... much easier

20:46 <calebccff> it seems with a lot of this stuff lack of docs are a running trend, i would expect it's largely because it's Qcom IP

20:47 <steev> i get that, but like, figuring out which ucm config file is in use...

20:48 <steev> that has nothing to do with qcom and everything to do with alsa

20:48 <calebccff> ahh, yeah that's a different issue entirely