00:00 <jiegec> Yet the internal touchpad required HID-over-SPI driver instead of SAM, I am trying to fix it, but it was not working yet.

00:01 <jiegec> GPU acceleration was not working yet: It failed when loading firmware. But using llvmpipe on a performant core is okay for development uses.

00:34 <robclark> "qcom/x1e80100/LENOVO/83ED/qcdxkmsuc8380.mbn")

00:34 <robclark> jiegec: for gpu you'll have to find and copy qcdxkmsuc8380.mbn from windows partition.. see arch/arm64/boot/dts/qcom/x1e80100-lenovo-yoga-slim7x.dts for ex, the firmware-name property should be something like (I guess) "qcom/x1e80100/microsoft/$something/qcdxkmsuc8380.mbn" where $something is based on the DMI string (ie for yoga 7x the dmi is `DMI: LENOVO 83ED/LNVNB161216, BIOS NHCN36WW 05/23/2024` so the path becomes

00:34 <robclark> with that, plus ToT mesa, gpu should be working

00:36 <robclark> there is integration branches with extra patches which are needed for some things, ie. https://github.com/jhovold/linux/tree/wip/x1e80100-6.11-rc1

00:49 <jiegec> Thanks, i will try that. I downloaded the firmware from linux-firmware, but maybe a specific firmware is required.

01:01 <louist103> so... regarding my issue above, I didn't have an initramfs....

01:01 <louist103> I forgot that dracut failed right after the kernel build. Issue now is that grub-mkconfig doesn't see the initramfs in the bootfolder

01:57 <robclark> jiegec: the "zap" fw (ie. qcdxkmsuc8380.mbn) is signed by the OEM.. lenovo pushed it to l-f for x13s but I think so far no on has for x1e so for now you need to copy that one from windows partition..

01:59 <robclark> the sqe and gmu fw.. not sure if latest is in l-f yet (but I guess it will be eventually).. the OEM signed zap fw, idk

02:19 <louist103> Update2... got the DTB installed and read from in GRUB but I just get 'qcom_aoss_qmp_c300000.power-management: failed to acquire IPC mailbox" and on the next line: arm-smmu 5040000.iommu: probe with driver arm-smmu failed with error -110. I have no idea how to troubleshoot that one...

02:21 <louist103> Could it be that I didn't install the firmware from the windows side yet? Last time I did that after I booted into the real system

02:48 <louist103> I set the log level to 5. Hoping this helps https://ibb.co/r2JMPRH

02:56 <jiegec> @robclack thanks for suggestion, I will try the firmwares from windows partition.

02:58 <jiegec> On Surface Laptop 7 I also encountered rfkill hard-blocked Wi-Fi one morning, but one day before it was working... No idea. It still works if I boot to Windows, but back to Linux it is still blocked.

02:59 <steev> robclark: that patchset, unfortunately, doesn't seem to help here

03:04 <steev> louist103: that looks like you're missing a bunch of modules in the initramfs - https://github.com/jhovold/linux/wiki/X13s make sure your command line has the options johan lists here, as well as making sure the initramfs has the modules listed in the commit for johan's defconfig (c.f. https://github.com/steev/linux/commit/95346caa14eeb6d4f6aeabbf1ea599ff43adf639 ) - additionally, make sure qnoc-sc8280xp is in there (this is x13s?)

03:04 <steev> or is this the c630?

03:09 <louist103> Its the C630. I used the config you linked yesterday

03:10 <louist103> I do have these options though "clk_ignore_unused pd_ignore_unused arm64.nopauth efi=noruntime"

04:11 <HdkR> https://github.com/Sonicadvance1/wfe_mutex?tab=readme-ov-file#spurious-wake-up-benchmark Spurious wake-up numbers are looking better on all platforms except for Oryon-1. Not sure what's up with Oryon when Linux is booted :)

05:01 <JensGlathe[m]> Jos Dehaes: okay, thank you. But it starts with wl

06:28 <JosDehaes[m]> Jens Glathe: yes, right now it's `wlP4p1s0`, but I've also seen `wlP5p1s0` IIRC

06:59 <JensGlathe[m]> iwconfig will be the way to go to find out for bootmac

07:35 <bumble[m]> using x13s here; wireplumber `wpctl status` only lists a "Dummy Output" Audio would anyone have any suggestion for getting the audio device to appear there? (system is postmarketos edge)

10:36 <Jasper[m]> <jiegec> "On Surface Laptop 7 I also..." <- It might be if you warmboot Linux that it's still loaded in from Windows

10:45 <lollaritits[m]> <bumble[m]> "using x13s here; wireplumber `..." <- alsa installed?

12:48 <abby> bumble[m]: got alsa-ucm-conf?

15:50 <bumble[m]> @_oftc_abby:matrix.org I do not have alsa-ucm-conf. I will try installing it...

15:53 <bumble[m]> @abby @lollar:catgirl.cloud it works! thank you!

16:22 <colemickens> Any update on the udev/module/reboot thing? I've only got a couple more days with the Yoga, might have a lead on 24.2/tip of mesa building, it'd be so cool to see it light up 🤞

16:32 <robclark> I don't think so.. I've just been working around it by commenting the problematic udev rule

16:40 <JensGlathe[m]> colemickens: I have prepared a bootable image for Volterra /X13s /Yoga 7 / Vivobook S15, it's still a bit rough, but should boot up (have no X1E hw to test though). Do you want to test if this comes up?

16:50 <colemickens> If you want a guinea pig I can try it. But mostly trying to get something workable for NixOS users so I can "announce and abandon" it in a close to usable state.

16:51 <JensGlathe[m]> Guinea Pig sounds fine. I've tested it on Volterra / X13s.

16:52 <colemickens> Sure, do you use magic-wormhole? I'll probably transfer it to my fast server and then rsync it slowly to this laptop and burn it. Jens Glathe

16:52 <JensGlathe[m]> google drive

16:53 <colemickens> that works too

16:53 <colemickens> well, :/ idk if it will play nice with my very slow internet, but I'll probably go to a coffee shop in a bit anyway.

16:55 <JensGlathe[m]> uploading.

16:55 <JensGlathe[m]> In this [directory](https://drive.google.com/drive/folders/1sc_CpqOMTJNljfvRyLG-xdwB0yduje_O?usp=sharing)

16:57 <colemickens> lmk filename to try when it's done uploading, I'll try it this afternoon along with maybe one more attempt on my side

16:57 <colemickens> I'm going to try to mask the udev rule again and then force uas/usb_storage and see if I can get it to come up.

16:57 <JensGlathe[m]> It's a raw image, just write with disks or dd onto a USB stick or SSD

16:57 <colemickens> * this afternoon (2-4 hours from now) along with

16:59 <JensGlathe[m]> Ubuntu_Desktop_24.04_x1e_6.11rc.img.xz

17:11 <JensGlathe[m]> It's up, 3,88GB

17:13 <JensGlathe[m]> Will write up some doc on it later, it is a start but no perfect image

20:10 <JensGlathe[m]> Some writeup. To be continued. https://github.com/jglathe/linux_ms_dev_kit/wiki/Bootable-Image-for-multiple-Snapdragon-(SC8280XP)-and-Snapdragon-X-Elite-(X1E80100)-laptops

20:56 <robclark> colemickens: I think the "udev reboot" is something remoteproc related? With CONFIG_DRIVER_DEBUG=y this is the last I see.. https://photos.app.goo.gl/B2XEzcGxtxqZASi49

21:19 <colemickens> robclark: hmph.

21:20 <colemickens> Jens Glathe: `035aad34b3d638109f31bf05574ab31a797ad61e932565985b55558b7091b901 Ubuntu_Desktop_24.04_x1e_6.11rc.img.xz`

21:25 <robclark> or.. maybe not.. still reboots after removing remoteproc modules from initrd and rootfs

21:30 <JosDehaes[m]> Jens Glathe: I tried booting your image on yoga 7x, it boots, but it reboots before reaching desktop

21:47 <robclark> JosDehaes[m]: yeah, I guess this is the "udev reboot", ie. reboot that happens if you don't remove udev role that auto-loads kernel modules

21:59 <JosDehaes[m]> strange I didn't have this issue before, I installed Arch and Debian on the yoga

22:00 <robclark> workaround is to comment a line:

22:00 <robclark> https://www.irccloud.com/pastebin/M7UTmLx6/

22:01 <robclark> in /lib/udev/rules.d/80-drivers.rules

22:01 <JosDehaes[m]> thx, I can try that

22:01 <robclark> the root issue might come down to the order modules are probed

22:01 <robclark> I don't quite understand what is triggering the reboot yet

22:03 <colemickens> someone supposedly got the _pas module loaded after probing some other modules, not sure if related.

22:05 <robclark> it _might_ be.. but OTOH I was still hitting this if I removed the remoteproc moduels altogether

22:06 <robclark> with remoteproc out of the picture, this is the last thing before reboot: https://photos.app.goo.gl/3p9qDQcjpuuFdK5j7

22:07 <robclark> looks like the last thing that probed without defer was qcom-bwmon

22:16 <JosDehaes[m]> robclark: in the rootfs, this udev rule is not there, is it in the initrd?

22:17 <JosDehaes[m]> nope, the initrd only has firmware and modules

22:41 <robclark> hmm, it could be somewhere else with different version of udev? I know it is the same btwn debian and fedora at least