10:18 <z3ntu> bamse: looks like you're poking around the Tachyon, anything further interesting to share already? I haven't booted mine up yet, not super excited about an Ubuntu 20.04 image... and I haven't seen kernel sources for that build either but maybe I haven't looked around enough

14:06 <bamse> z3ntu: the information i got from the forum was that they are working on ubuntu 24.04 and when that is done they will release the source code...but i don't want to wait and i don't want ubuntu

14:06 <z3ntu> Same

14:09 <bamse> z3ntu: i have some basic things enabled (storage, pcie, wifi, usb, display, gpu etc)...sufficient to browse the particle forum etc...

14:10 <bamse> it's idling at around 50C and i probably have the wrong wifi calibration data...

16:39 <z3ntu> Hi, anybody have an idea what's going wrong with sound sm8250.c and qcom_snd_sdw_startup leading to null pointer dereference in snd_soc_dai_set_stream? My sound setup right now is just MI2S and no soundwire stuff (no wcd configured yet)

16:40 <aklimov> z3ntu: it goes to NULL pointer dereference?

16:40 <aklimov> z3ntu: i can try to cook simple patch to fix that

16:40 <z3ntu> Trying to debug why dai->driver->ops is null in the last bit, but I don't quite get why it's trying to setup soundwire/sdw bits for MI2S in the first place

16:40 <z3ntu> [ 32.738424] DBG snd_soc_dai_set_stream:558 ffff000081279880 ffffd49a7af5d6f8 0000000000000000

16:41 <aklimov> z3ntu: because it is broken?

16:41 <z3ntu> printk(KERN_ERR "DBG %s:%d %px %px %px\n", __func__, __LINE__, dai, dai->driver, dai->driver->ops);

16:41 <z3ntu> 

16:41 <aklimov> it should just ignore sdw when it is not present/compiled/there

16:41 <z3ntu> pretty sure swr stuff is all enabled in kconfig, just not configured in dts

16:42 <z3ntu> but yeah, if you have an idea, would be nice if you could make something there

16:42 <aklimov> still may fail

16:42 <aklimov> i had some local changes to workaround that, when i was debugging compress playback via hdmi/i2s

16:43 <z3ntu> also a stupid workaround would be fine for now, just trying to actually get to what I was trying to do, this is not going upstream anytime soon anyways because it probably won't work

16:48 <z3ntu> aklimov: I'll stop trying to debug this then since you seem to have something, thanks

16:55 <aklimov> z3ntu: let me search my local files, i'll get back to you

16:59 <z3ntu> aklimov: appreciate it

18:02 <aklimov> z3ntu: can you access this https://pastebin.linaro.org/view/6865b0f0 ?

18:02 <aklimov> so I think i saw this set_stream deref and qcom_snd_sdw_startup() panics on RB3 and RB2

18:03 <aklimov> i think you might need if (dai->driver->ops && dai->driver->ops->set_stream) from that URL

18:03 <aklimov> and for sm8250 maybe something like this might do the job: https://pastebin.linaro.org/view/d433c916

18:04 <z3ntu> aklimov: nope, I get a login screen

18:04 <krzk> z3ntu: aklimov: could be 8167f4f425728? (I love the comment, though: "Unfortunately, current code is broken to start with.")

18:05 <krzk> z3ntu: aklimov: but I would also seriously consideer my commits, including db7e59e6a39a4d3d54ca8197c796557e6d480b0d and alike :)

18:07 <krzk> similar bug was actually fixed by aklimov for sdm845

18:10 <aklimov> z3ntu: is it visible here https://pastes.io/rb3-set-stream-and-qcom-snd-sdw-startup-deref ?

18:11 <aklimov> that's what i used for RB3 but iirc i saw similar thingies on RB2 (sm8250 sndcard)

18:11 <z3ntu> aklimov: yep, that link is accessible

18:13 <z3ntu> yes I need sound/soc/qcom/sm8250.c but should be simple enough to copy

18:15 <aklimov> z3ntu: something like this maybe https://pastes.io/sm8250-ignore-sdw-startup-release-for-i2s-dais

18:15 <aklimov> z3ntu: and if (dai->driver->ops && dai->driver->ops->set_stream) from first link

18:17 <aklimov> krzk: db7e59e6a39a4d3d54ca8197c796557e6d480b0d -- that works fine iirc on sm8250, it just needs to ignore swd calls when dealing with i2s and i2s-based DAIs

18:24 <z3ntu> aklimov: thanks, seems to work :) Now it's back to the amplifier driver complaining about something not being correct

22:24 <aklimov> z3ntu: is it an amp from Texas Instruments? :)

22:24 <aklimov> z3ntu: you're welcome