00:00 <cenunix[m]> ah, main reason i'm asking is because im trying to use qrtr and pd-mapper for the battery status on the thinkpad x13s

00:01 <cenunix[m]> pd-mapper relies on qrtr-ns ? at least in the .service file provided in the pd-mapper repo, so i'm unsure how I should proceed to get pd-mapper working as intended

00:01 <bamse> cenunix[m]: should be sufficient to just start the pd-mapper service (or just run pd-mapper in the console)...if you have the relevant firmware files

00:02 <bamse> cenunix[m]: pd-mapper depends on the ns functionality...which these days lives in the kernel

00:04 <cenunix[m]> pd-mapper gives me : no pd maps available

00:05 <cenunix[m]> for reference, im running it on nixos, which does not keep its firmware in /lib/firmware, so I forked pd-mapper and changed the default firmware path from /lib/firmware to /run/current-system/firmware

00:05 <cenunix[m]> it seems as though that part is working, as i'm not getting any errors now regarding not finding the firmware

00:06 <agl7> steev: Ok, I will wait ...

00:06 <bamse> cenunix[m]: the file you need is /lib/firmware/qcom/sc8280xp/LENOVO/21BX/battmgr.jsn

00:07 <bamse> cenunix[m]: pd-mapper typically searches for *.jsn in the directories describes in /sys/class/remoteproc/*/firmware

00:08 <bamse> cenunix[m]: ohh "seems as though ...working"...then you should be fine then :) but make sure that it actually finds the battmgr.jsn and not just one of the other *.jsn

00:09 <bamse> cenunix[m]: also note though that the remoteproc for the adsp must be running, in case you messed around with the firmware paths

00:15 <cenunix[m]> I have battmgr.jsn.xz file in the directory /run/current-system/firmware/qcom/sc8280xp/LENOVO/21BX/

00:16 <bamse> cenunix[m]: have you taught pd-mapper how to read .xz files?

00:17 <cenunix[m]> no, im assuming thats whats causing the issue lol

00:17 <bamse> it sounds like a good lead ;)

00:17 <cenunix[m]> thanks for pointing me in the right direction, i'll get on that :)

00:18 <bamse> perhaps just leave it uncompressed?

00:19 <bamse> cenunix[m]: x13s have been highlighting how annoying the pd-mapper and jsn scheme is...so we've talked about just moving the service into the kernel - just as we did with qrtr-ns

00:20 <steev> what in the world

00:20 <steev> why the hell does the firmware get shoved into /run on nixos?

00:20 <bamse> steev: because it's blazingly fast!

00:20 <steev> apparently so

00:21 <steev> the more i hear about nixos the more i wonder if these people even know what they're doing

00:21 <steev> it's just dumb shit after dumb shit

00:25 <bamse> ugh, looks like this old crap works now

00:25 <steev> Ay, rc6 is out and that means a few more of our patches are in :) and by our, I mean the developers of msm stuffs

00:25 <bamse> been spelunking old trees all afternoon, trying to puzzle together the last pieces to get external display properly up and running on flex 5g

00:26 <steev> 9 or 10 patches that we were carrying

00:26 <steev> Oh spiffy

00:26 <steev> I haven’t revisited mine yet :(

00:26 <bamse> don't bother, yet

00:26 <bamse> linux-next doesn't boot reliably

00:26 <steev> Well the c630 and the x13s are both mostly working so I haven’t needed to

00:26 <bamse> got some fixes coming for that

00:27 <bamse> it's sad how bad the keyboard on the flex5g is though :(

00:28 <steev> I mostly use the flex for its wsl2 testing

00:28 <bamse> i used flex to develop all the external display stuff...

00:29 <bamse> but haven't really touched it since i got the x13s

00:29 <bamse> felt the urge to test it now that i merged the base dts though...and flush out any delta i had laying around

00:34 <steev> makes sense

00:35 <bamse> would be nice to get the prox onboard as well...

00:51 <cenunix[m]> What’s wrong with it being in /run, just deviating from the standard? Genuine question I have no clue lol

00:57 <steev> well, /run is temporary, and firmware doesn't change so often that it needs to be replaced every boot

01:02 <cenunix[m]> Ah, firmware isn’t stored there, it’s technically in /nix/store, it’s symlinked

01:09 <steev> that's a little bit better

01:36 <steev> there is one weird thing i notice between 6.3 and 6.4... for some reason... in 6.4, gdm seems to be at 150%

01:43 <clover[m]> its a feature not a bug

01:46 <steev> seems odd to be at 100% on 6.3 and 150% on 6.4

03:31 <robclark> steev: 100% vs 150% is.. cpu usage? Are things running at same freq in both cases? If so, break out perf?

03:35 <HdkR> I assumed scaling

03:51 <steev> robclark: scaling, not cpu usage, sorry; basically the login screen looks ginormous on 6.4 and normal on 6.3

03:51 <steev> it's looking like it's just a case of gnome devs thinking they know best, again though

03:52 <robclark> ahh.. paste output from running modetest on both kernels? Could be some change in properties reported?

13:59 <clover[m]> is it annoyingly big?

14:01 <steev> yeah, apparently the workaround is to cp ~/.config/monitor.xml /var/lib/gdm3/.config/

16:36 <stirl_> clover[m] -- I didn't get a chance over the weekend to test any of this out but I will try to get the bootloader installed on my lunch break today. The walkthrough looks pretty complete at this point!

16:55 <clover[m]> you'd probably be the first to use it so let us know how it goes

18:44 <stirl_> forgot my live usb at home so I couldn't get it done

20:31 <cenunix[m]> might be the hackiest solution to date, however, battery status is working lol

21:12 <steev> if it works... it works

21:17 <cenunix[m]> Is there a kernel module for sound stuff on the x13s?

21:19 <exeat> Just out of curiosity, where does the kernel normally look for fw on that system?

21:21 <cenunix[m]> I think /run/current-system/firmware

21:23 <exeat> Is that a tmpfs filesystem? If so, how is it populated?

21:29 <cenunix[m]> System is built, everything including fw stored in /nix/store, fw symlinked to /run/current-system/firmware

21:30 <cenunix[m]> I think, honestly haven’t been using NixOS for THAT long and certainly haven’t been messing with fw besides the past week or so

21:46 <exeat> If that's how NixOS does things, it seems to me that the fw-like pd-mapper files should live in /nix/store with the other fw (but be uncompressed)

22:25 <cenunix[m]> Yeah idk the proper way to do it, that’s the reason it’s a hacky workaround. Pd-mapper technically does live in nix/store, but the binary is symlinked to /usr/bin or something like that

23:05 <clover[m]> I did a thing

23:26 <cenunix[m]> wooo

23:43 <clover[m]> robclark: how did you get sound working? I have SC8280XP-LENOVO-X13S-tplg.bin in /usr/lib/firmware/qcom/sc8280xp and I have alsa-ucm installed with dnf

23:43 <robclark> I didn't get sound working yet (or really try to yet)

23:44 <clover[m]> Oh ok

23:46 <clover[m]> ajhalaney: any thoughts on this?

23:55 <ajhalaney[m]> Sound works for me but I'm not really sure _why_ that is I haven't looked into sound much. I thought all that was packaged up proper now but I've polluted my system with various local modifications since purchase so idk. I got a USB stick I was gonna start playing with as a clean install but haven't done much besides verify I could boot once