03:21 <freekurt[m]> I have 100% load on some CPU cores too on my x13s.

03:41 <bamse> freekurt[m]: which kernel version?

03:45 <freekurt[m]> One of the latest kernels that has come out. I use [@jglathe:matrix.org](https://matrix.to/#/@jglathe:matrix.org)'s and update regularly. The issue has been happening for a while though.

04:34 <valpackett> bamse: steev: freekurt[m]: yes finally i'm not alone with that ath12k 100%cpu issue!

04:35 <valpackett> it's bouncing because every 6 seconds networkmanager tries to read wifi status

04:35 <valpackett> and ath12k somehow manages to spin on a spinlock for like a whole second

04:48 <valpackett> it's ath12k_mac_op_cancel_remain_on_channel

06:06 <steev> i use my kernel, as well as the debian kernel on the x13s and have not seen it, but i did not really test 6.15 yet, i saw some stuck clocks and haven't gone back just yet to figure out where the issue is

08:25 <osmoz[m]> Hello again, someone has an idea on which step I forgot to install void Linux on the Asus Zenbook a14 (Qcom X1e-800)... (full message at <https://matrix.org/oftc/media/v1/media/download/AQre-IOHKpoMpRYTArqP28ohr44z-QBBNczxU6KZSVLGHLAOrAH_fgFPZ4UZceA2MuWLA3dbnBNy1Dr_NT8xrPxCeXjHFl-AAG1hdHJpeC5vcmcvcm50QUtrV0N3SmNmSldQTmdQU0RBb0du>)

08:28 <alexVinarskis[m]> Hey!

08:28 <alexVinarskis[m]> Glad someone else got the zenbook. Have you tried booting Ubuntu Concept as live USB stick? Latest ISO should support both x1p/x1e, that would be the simplest way to get started.

08:30 <osmoz[m]> alexVinarskis[m]: > <@anynameaaa:matrix.org> Hey!... (full message at <https://matrix.org/oftc/media/v1/media/download/AQwca0q0V0HoV-PdoYYhSAO2u8XRBUM4qg33kyMoEnr79UlvZE-aHX1PaCvZM5Z5HGc3ya9k6zEh9AbjeYdP7SpCeXjHZpXQAG1hdHJpeC5vcmcvZm9VTWdoVFhPUm1hTkl1ZVhpWUZyc0Vm>)

08:31 <alexVinarskis[m]> Aah okay. So its not the dtb/patches issue.

08:35 <krzk> Can you stop using matrix chat on IRC? Half of these messages are totally unreadable.

08:36 <krzk> Basically there is just "Hey" and some weird link

08:36 <krzk> s/matrix chat/matrix style/

08:37 <osmoz[m]> I will install a client later, I think it’s because I replied to the message

08:39 <krzk> I know. You can use Matrix, just without Matrix features.

08:41 <osmoz[m]> But my first message does include matrix features, maybe it’s the size or the newline with my client

08:58 <osmoz[m]> I don’t think it’s the dtb, but I only used the first one (0012) because there were compile errors when all of them were included and I don’t need sound and media right now

08:59 <osmoz[m]> The kernel seems to start but my issue is when mounting the root partition I will have to research how does the initramfs work and what does it needs

09:02 <jhovold> alexVinarskis[m]: I can confirm that reverting the genirq cpu hotplug commit fixes the hosed nvme after suspend here too

09:04 <jhovold> can you please report this today (to avoid waiting until rc1 is released with this regression on sunday)?

10:02 <alexVinarskis[m]> Thanks for testing! Absolutely, will be on it in maybe an hour.

10:51 <deathmist> osmoz[m]: you could hack together something similar to https://gist.github.com/JamiKettunen/7c2a25688976c09af819eebc9fbb8396 except adapting from initramfs-tools to dracut etc (see e.g. https://github.com/void-linux/void-packages/tree/master/srcpkgs/x13s-base), abby probably can help out if needed :)

10:53 <deathmist> I'd submit my stuff to chimera officially but imo kernel side stuff is still way too much in flux for me to bother while limine (which the aarch64 and most other isos uses as bootloader) is missing stuff to make the live boot experience proper, for some reason it also takes forever to show the menu which I'd have to debug somehow too

11:03 <deathmist> maybe the solution for the iso problem would be to make a separate x1e ones integrating sysd-boot + dtbloader or even just swap the main aarch64 ones to use those instead, will see ig once most stuff on kernel side starts to settle down a bit and the various patches wouldn't need to be dealt with as much on each pkgver bump

12:33 <jhovold> alexVinarskis[m]: thanks for reporting and spotting the autosel selection, just asked stable to drop it for now

12:48 <alexVinarskis[m]> Great, thanks for the followup. Hope they can have it resolved on such short notice...

14:26 <osmoz[m]> deathmist: thanks I will look into it weekend. I will also try to send upstream my changes to void if I can make something clean

14:54 <abby> yeah it'd be cool to support x1e, but i don't have any of the devices

16:01 <freekurt[m]> valpackett: is there a way to fix the ath12k 100% cpu issue? revert firmware version maybe?

17:25 <JosDehaes[m]> Jens Glathe: again same USB issue with yoga 7x with your kernel `jg/ubuntu-qcom-x1e-6.15.0`. No USB works (but DP alt does), lsusb gives an error

17:28 <JensGlathe[m]> Can you check in the dtb if the smb2360 repeaters are explicitly enabled? On the phone, can’t look

17:29 <JosDehaes[m]> it says status=okay

17:31 <JensGlathe[m]> In smb2360_0 , _1, _2 ? Not in the phy node?

17:46 <JensGlathe[m]> no they look okay

17:52 <JensGlathe[m]> lsusb gives permission errors? That's broken since somewhere in 25.04

17:52 <JosDehaes[m]> no it returns 1

17:53 <JosDehaes[m]> this was the same the last time

17:54 <JosDehaes[m]> also, this is arch not ubuntu

17:56 <JosDehaes[m]> hmm, same issue with jhovold's tree

17:59 <JosDehaes[m]> got to go now, will check later

18:01 <alexVinarskis[m]> there was a change of eusb driver ~2 weeks ago, from iirc next-20250515 or next-20250514 CONFIG_PHY_QCOM_SNPS_EUSB2 would become CONFIG_PHY_SNPS_EUSB2, do you have the correct one enabled?

18:02 <JensGlathe[m]> I backed the new synopsys driver out bc of issues, not in my tree/config

18:04 <JosDehaes[m]> no I did not have that one, rebuilding

18:04 <alexVinarskis[m]> since Jens reverted the change, you need the old/CONFIG_PHY_QCOM_SNPS_EUSB2 one

18:09 <JosDehaes[m]> Works, thx 🙏

20:35 <valpackett> freekurt[m]: no, it's not a firmware issue. evidently it's a regression so we need to bisect the kernel to find out when it started happening

20:41 <freekurt[m]> Oh dang. Ok!