08:22 <KieranBingham[m]> I've installed pd-mapper from the fedora rawhide packages, but it's failing to start saying no pd maps available. Is there some extra files I need to install?

08:22 <travmurav[m]> Kieran Bingham: there should be .jsn files in lib/firmware for it

08:24 <steev> yep

08:24 <steev> should be part of the firmware package(s)

08:31 <travmurav[m]> inb4 they're there but compressed

08:31 <KieranBingham[m]> looks like I have files ... /lib/firmware/qcom/sc8280xp/LENOVO/21BX/adspr.jsn

08:32 <KieranBingham[m]> stracing pd-mapper to see where it's looking

08:32 <steev> that's what i was gonna suggest

08:35 <KieranBingham[m]> hrm ... it#s trying to process /sys/class/remoteproc which is empty..

08:35 <travmurav[m]> ah yeah it tries to read where to find the jsn files from the remoteproc

08:36 <steev> dmesg to check the remote procs

08:41 <steev> which kernel sources are you using?

08:44 <KieranBingham[m]> johans' 6.9-rc7 ... I'll probably rebuild the 6.9 tag next

08:45 <steev> and using his defconfig?

08:45 <steev> i mean the defconfig itself should be fine

08:53 <jhovold> KieranBingham[m]: did you get pd-mapper to start? As travmurav[m] mentioned, I think fedora is using compressed fw so you may need to enable that option in the config if you haven't already

08:55 <jhovold> Segfault[m]: yes, the missing usb-pd bits are supposed to be 6.10, haven't had time to test it myself beyond smoke testing the first two iterations (that failed the test)

08:56 <KieranBingham[m]> ah yes , I went back to Johans' defconfig as part of getting back to a stable build and I probably dropped my additional configs for the fedora ...

08:57 <jhovold> craftyguy: wifi is quite stable here (also when travelling) so I assume it depends a lot on your AP

08:58 <jhovold> lollaritits[m]: pulseaudio handles bluetooth just fine, that's what I've been using

09:00 <lollaritits[m]> jhovold: didnt for me on pmOS

09:01 <lollaritits[m]> Pipewire now does

09:06 <jhovold> must have been missing some package, bluetooth support is a separate package on arch for example

09:08 <travmurav[m]> <travmurav[m]> "inb4 they're there but compresse..." <- (I've actually forgot but pd-mapper nowdays supports .xz compressed .jsn files, but only xz)

09:09 <Jasper[m]> iirc they should be xz

09:09 <Jasper[m]> I had a particularly nasty wifi crash not too long ago

09:09 <Jasper[m]> Have a dmesg, but only from some error messages since ath11k was spamming a lot

09:10 <KieranBingham[m]> travmurav: The json files are not compressed, but the firmware files are and indeed I think this kernel currently doesn't have that enabled. Just trying to rebuild now with my additions to johan_defconfig

09:10 <KieranBingham[m]> so the remoteprocs are not loading.

09:10 <travmurav[m]> ah that would make sense if kernel couldn't boot them at all

09:11 <jhovold> Jasper[m]: sounds like the one I reported upstream, a little worrying that the qualcomm guys have not gotten back to me on that one yet

09:12 <jhovold> https://lore.kernel.org/all/Zh6b-38tGGXo-ee7@hovoldconsulting.com/

09:14 <jhovold> my talk here at linaro connect from yesterday about the (conclusion of) the arm laptop project is now up:

09:14 <jhovold> https://resources.linaro.org/en/resource/1dCZq5rYdNHR6WviFLaPMB

09:16 <steev> conclusion :(

09:16 <jhovold> still hoping that someone will step up and fund a continuation in some form

09:17 <steev> that would be nice :(

09:18 <Jasper[m]> <jhovold> "https://lore.kernel.org/all/Zh6b..."; <- Yep, that's the one

09:18 <Jasper[m]> jhovold: Glad that it started, sad that it ended.

09:18 <Jasper[m]> jhovold: They'll figure it out hopefully, though likely with X Elite

09:19 <jhovold> one of the key points of my talk is that almost all of the remaining issues with the x13s are generic qualcomm issues

09:20 <jhovold> so work should continue on the x13s as it will benenefit later platforms directly

09:20 <jhovold> for when they become available and eventually become as usable as the X13s already is

09:26 <steev> Maybe we need to start a Lasahi Linux gofundme

09:27 <steev> I’ll watch it after I wake up, or maybe to fall asleep to

09:38 <travmurav[m]> I wonder if qcom would bother with remoteprocs for the alleged x elite EL2 mode

09:40 <travmurav[m]> there are definitely a bunch of hexagons in it and I think they boot the same way

09:42 <Jasper[m]> <steev> "I’ll watch it after I wake up..." <- Lo-fi Swedish man talking about computing to work or fall asleep to

09:57 <steev> lol so I ended up watching the whole thing. The most impressive to me is Johan pronouncing my last name correctly

09:58 <jhovold> :)

09:58 <steev> Also hrw’s comments about his Chromebook made me laugh. I was “around” (on irc) for that… good times

11:58 <KieranBingham[m]> hrm ... I'm pretty sure my firmwares are being handled (particularly as wifi is working anyway) ... but /sys/class/remoteproc is still empty. I do'nt see anyhting in menuconfig that would suggest needing an additional config for sysfs interface to remoteproc (there is for cdev) so I guess something else is still going on ... (dmesg for reference https://pastebin.com/raw/4evR2Ukn)

11:59 <travmurav[m]> Kieran Bingham: can you check sys/kernel/debug/devices_deferred ?

12:00 <KieranBingham[m]> 33c0000.pinctrlplatform: wait for supplier /soc@0/remoteproc@3000000/glink-edge/gpr/service@2/clock-controller

12:00 <KieranBingham[m]> soundsnd-sc8280xp: WCD Playback: error getting cpu dai name

12:01 <travmurav[m]> I guess all of this depends on the adsp remoteproc too

12:01 <travmurav[m]> something feels missing for it to appear...

12:01 <travmurav[m]> kernel didn't even try booting it

12:02 <Jasper[m]> I think I've also included some firmware in initramfs before it started working

12:03 <Jasper[m]> You can include the json files aswell

12:03 <Jasper[m]> I think the problem was that it enters some failstate at boot because the firmware wasn't included

12:03 <travmurav[m]> I think it'd still at least register the remoteproc and complain it failed to boot it

12:08 <Jasper[m]> iirc it took a bit of finding out though, so maybe there's nothing there

12:09 <KieranBingham[m]> It does look like my initramfs isn't including the full /usr/lib/firmware/qcom ... I'll see if I can figure out how to tell dracut to rebuild that.

12:12 <Jasper[m]> Should be enough to just include the sc8280xp folder

12:12 <Jasper[m]> And iirc there's a simple "add" command for it

12:17 <KieranBingham[m]> yeah - it' the "how do I tell fedora thigns to include it" is the hard part. Currently all I'm doing is running "make install modules_install" which seems to hook in somewhere to rebuild things.

12:20 <pstef_> if Dracut then maybe https://wiki.gentoo.org/wiki/Lenovo_ThinkPad_X13s#Dracut

12:20 <pstef_> it's what I do under Fedora Rawhide

13:32 <Jasper[m]> <KieranBingham[m]> "yeah - it' the "how do I tell..." <- I'll tell you what I did when I get home. I'm pretty sure it's a permanent solution luckily

14:23 <KieranBingham[m]> Jasper: Thanks, I think I've got the dracut conf. I've done this:... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/cKbCtCANcoOAJQMEhUssDJkV>)

14:48 <Jasper[m]> My bad, something came in the way

14:48 <Jasper[m]> <KieranBingham[m]> "Jasper: Thanks, I think I've got..." <- > <@kbingham:matrix.org> Jasper: Thanks, I think I've got the dracut conf. I've done this:... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/ATtYjEmaimKPrBZrdPWCqzbj>)

14:55 <akaWolf> steev: is there any meaningfull changes from 6.7.10?

15:16 <jglathe_volterra> oh this doesn't look good https://pastebin.com/vFaknEQ8

15:16 <jglathe_volterra> BH-M9A is my bluetooth headset

17:57 <bumble[m]> <jhovold> "https://resources.linaro.org/en..."; <- watched the entire video and enjoyed it, thank you

17:58 <bumble[m]> hoping that x13s will continue to improve for linux

18:01 <bumble[m]> it seemed activity level around x13s was high and one could be hopeful support would reach a high level

18:59 <adamcstephens> :( nixos is now zstd compressing the firmware

18:59 <adamcstephens> so of course pd-mapper is now brokwn

19:05 <HdkR> Looking forward to that in-kernel implementation

19:06 <travmurav[m]> We were discussing using zstd for linux-firmware in alpine and afaiu if one uses the tooling inside it, it would respect whence file and won't compress the .jsn files since they're marked raw

19:07 <travmurav[m]> the tqftpserv would still be sad in this case but that's not needed on x13s ig

19:08 <travmurav[m]> but yeah having it in the kernel would make everyone's life easier xD

19:25 <\[m]> didn't know it was funded by arm

19:25 <\[m]> who initiated it, arm or linaro?

19:27 <HdkR> Who is ARM funding? Netsplit dropped the message

19:32 <steev> HdkR: the laptop project (aka jhovold's work the past 2 years)

19:39 <HdkR> ah

19:57 <jglathe__> Great Job, happy this happens

20:59 <steev> KieranBingham[m]: the other thing i would suggest... make sure you're using the correct dtb?

21:03 <KieranBingham[m]> steev: Yes I believe so. I think my(last) issue really is to get the initramfs generated correctly. When I had to stop earlier I was working with srini to get the adsp modules loaded. I think most of my issues are related to the distro and a big part that stopped me this week was I didn't realise fde was blocking and I had to press "enter" to get a prompt to enter my decrypt password. That was annoying to discover :-) it looked

21:03 <KieranBingham[m]> like it had hung but was just waiting without displaying a prompt!

21:04 <steev> oh wow, that's weird

21:04 <steev> definitely shouldn't require input... to do input lol

22:41 <craftyguy> robclark: are you using startup.nsh to run slbounce on your x13s? (finally looking at the drive you let me copy last week)

22:52 <robclark> yeah.. the slbounce esp has Shell.efi as the bootaa64.efi

22:52 <robclark> then runs startup.nsh

22:53 <robclark> so I don't have to type a lot of things each time I reboot ;-)

22:54 <craftyguy> make sense. I noticed that script is not running the dtbhack thing, is that intentional? I thought it was required for linux boot (or, fixed some things with it?) ?

22:54 <robclark> it is commented out at the momemnt, I just hacked the dts when I build kernel

22:54 <robclark> so dtb is "pre-hacked" ;-)

22:55 <craftyguy> ahh ok

22:55 <craftyguy> do you use that same dtb for booting w/o slbounce?

22:55 <robclark> no

22:55 <robclark> not the final solution

22:56 <robclark> I just haven't gotten around to getting the dtb patching to work properly

22:56 <craftyguy> I see, just curious

22:56 <craftyguy> so for now I'll just use dtbhack since I'm using a different kernel

22:57 <steev> - based on johan's 6.9

22:57 <steev> clover[m]: pushed 6.9.0; there is no laptop_defconfig for now, just johan's defconfig.

22:57 <steev> - include dmitry's in-kernel pd-mapper (enabled as module in johan_defconfig)

22:57 <steev> - konrad's pshold work (and fix so that the x13s shuts down)

22:57 <steev> - srini's dp audio work (needs his latest firmware and alsa-ucm-configs)

22:57 <steev> - dmitry's usb work which should mean no more orientation flipping

22:57 <steev> - my bump to the thermals so it can get hotter

22:57 <steev> - enable OVERLAY_FS as module in johan_defconfig (i needed this for my sbuild chroots)

22:57 <steev> - my patch to bluetooth for using the windows nvm name for the bt firmware

23:00 <akaWolf> steev: are there a lot of important changes since 6.7.10?

23:02 <steev> there've been a few, yeah, but if 6.7.10 works for you, there's no need to feel like you have to jump to the latest

23:02 <akaWolf> audio doesn't work properly, recording too

23:03 <akaWolf> sometimes it rebooting while starting graphical podsystem

23:05 <clover[m]> Steev: interesting! I will build it tonight. I'm sick of work and want something else to do

23:06 <steev> relatable :D i'm building 59 debian testing chroots

23:08 <akaWolf> does audio works well for you?

23:08 <akaWolf> clover[m]: what's reason to have mesa forked from upstream now?

23:12 <KREYREN_oftc> Does anyon know what's the kernel mainline of https://www.nxp.com/part/MIMX8ML8CVNKZAB#/ ?

23:29 <akaWolf> someone using nixos on x13s

23:29 <akaWolf> https://dumpstack.io/1675806876_thinkpad_x13s_nixos.html

23:30 <akaWolf> it can be alternative to arch linux arm

23:35 <akaWolf> steev: https://github.com/steev/linux/commits/lenovo-x13s-linux-6.9.y/ <-- that one branch?