00:18 <deathmist> icecream95: thanks for the suggestions! I can confirm Esc/F2 via a USB keyboard works, I'll look into the other stuff tomorrow.. reminds me I've previously tried messing with the eBPF HID stuff trying to disable the keyboard backlight but never properly got it working so far..

02:38 <machoskraito> hi all cartel here ... . how are you all ... .

03:26 <binarycraft[m]> <deathmist> "icecream95: thanks for the..." <- ebpf program doesn't work with suspend resume scenario, I had a kernel module working with keyboard backlight stuff

04:48 <icecream95> binarycraft[m]: Is suspend/resume a big problem when the backlight is automatically cut off when the lid is closed? (Says someone with a lot of experience of brokenish laptop hinges...)

04:48 <icecream95> But a kernel module could be a good idea anyway, so that it just works without having to enable BPF and setup userspace correctly.

04:48 <icecream95> It could be nice to expose the brightness so that upower can find it, then it can show up in desktop environments:

04:48 <icecream95> https://sources.debian.org/src/upower/1.90.9-1/src/up-kbd-backlight.c/?hl=237#L237

04:56 <binarycraft[m]> <icecream95> "binarycraft: Is suspend/resume a..." <- There are also scenarios you don't close the lid, such as auto suspension

05:31 <icecream95> deathmist: Note that I also wrote a Python script (that works with three incompatible bindings for hidapi) to set the backlight brightness: https://github.com/icecream95/udev-hid-bpf/blob/vivobook/kb-bl.py

09:22 <deathmist> oh wow I could finally disable the vivobook keyboard backlight with the latest changes to that above script and python-hid instead of the seemingly broken-for-me hidapi* modules I tried before. still though my battery drain is at minimum ~7.1W with it disabled and up to ~12.8W at maximum brightness. under windows I've seen down to 2.9W so something is still wrong

09:43 <deathmist> also resettings/crashing the keyboard EC got ESC/F2 on the built-in keyboard working through the i2cget command, wow

09:48 <binarycraft[m]> deathmist: Have you got this tool working on s15: https://github.com/Maccraft123/it8987-qcom-tool/, it outputs nothing for me, which made me double 0x76, i2c5 is the EC

09:48 <binarycraft[m]> * deathmist: Have you

09:49 <binarycraft[m]> > <@_oftc_deathmist:matrix.org> also resettings/crashing the keyboard EC got ESC/F2 on the built-in keyboard working through the i2cget command, wow

09:49 <binarycraft[m]> * deathmist: Have you got this tool working on s15: https://github.com/Maccraft123/it8987-qcom-tool/, it outputs nothing for me, which made me doubt 0x76, i2c5 is the EC

11:37 <icecream95> deathmist: I'm seeing less than 6 W for very dark screen content, and down to 5.2 W at idle, so a bit better than you, but it is still worse than Windows. I guess some of that is the fans, but I wonder know much

12:06 <anthony25> I dont think so: the slim 7x with the driver for the EC can stop its fan, and I also see ~5W power draw with dark content, 3W with the screen off

12:07 <anthony25> We talked about it a bit here before, but PSR enabled could help, for example (it's buggy right now)

12:09 <anthony25> And also deeper powerstate levels for the CPU, and powersaving on the nvme (and probably other things, from what I understood)

13:41 <deathmist> anthony25: I still get ~5.8W drain with display off

13:42 <deathmist> (via brightness set to 0)

13:43 <anthony25> Oh wow

13:57 <deathmist> binarycraft[m]: listen-to-events at least doesn't report much of anything when I try function row stuff etc if that's what you mean

13:58 <deathmist> Fn + F1, Fn + K and such that is

13:58 <binarycraft[m]> deathmist: Same

15:56 <zetam> Hello #aarch64-laptops! Quick update: minimal boot achieved on Dell Latitude 7455. Shell loads but TTY control still elusive. Proof: https://ibb.co/YgMFLpJ. Still hacking with help from you guys and AIs. More soon! ps. deepseek-r1, un genio redactado mensajes irc :D

16:06 <alexVinarskis[m]> zetam: actually Ubuntu Concept's ISO for 25.04 from 20250703 already contains all quirks for Inspiron, Latitude, just not the .dtb itself (awaiting v3 from valpackett before MRing it to Ubuntu repos). If you just grab that installer, and add compiled dtb to /casper/dtbs/ it will boot you into working installer with UI etc. Installation itslef will fail, as it will attempt to install the same device tree which is missing

16:06 <alexVinarskis[m]> in the kernel packaged in the installer, but there is a way to hack that too: modify your DT to change model to already existing device (eg. Dell XPS one), the place it to /casper/dtbs/, boot, install. it will succeed, but install wrong DT obviously. before rebooting, go to /target/boot/... and swap dtb to correct one.

16:09 <kettenis> icecream95: someone should probably tweak drivers/hid/hid-asus.c to support the s15 keyboard

19:44 <valpackett> woah wtf.. rebased to next-20250704 and backlight brightness is super borked, not a smooth 0-100 range but it goes over the actual brightness range many times

19:45 <valpackett> i guess it's the common displayport changes that landed recently

20:02 <valpackett> https://files.catbox.moe/sp1g9v.mp4

20:06 <alexVinarskis[m]> Hmmm i remember having that on xps longgg time ago. Dont recall how it was resolved (not by me)

20:13 <valpackett> ow there's also warning spam: drivers/gpu/drm/drm_gem.c:287 at drm_gem_object_handle_put_unlocked+0xb4/0x100 eevery time gnome-shell frees a framebuffer

20:26 <valpackett> ok yeah the intel patch just increased the dpcd command length to 3 even when luminance is not used

20:26 <valpackett> unsurprisingly not all panels are happy about that

20:28 <alexVinarskis[m]> Ohhh, i got the same warnings on next-20250703 on x86 device few days ago as well

20:29 <alexVinarskis[m]> Guess they will fix within a few days

20:47 <valpackett> brightness fixed: https://lore.kernel.org/all/20250706204446.8918-1-val@packett.cool/

20:57 <alexVinarskis[m]> fast! though, that means drm crashes are not related? just tried, didn't fix it here.

20:59 <valpackett> no the warnings are not related

21:01 <valpackett> and the v3 of the dells: https://lore.kernel.org/all/20250706205723.9790-2-val@packett.cool/

21:03 <valpackett> also for some reason i've been getting higher power consumption recently, pretty much since the bios update that fixed (so far, i think) the random freezes (?)

21:03 <valpackett> i doooon't think it's the removal of the unused i2c busses, pretty sure it started before i did that

21:53 <steev> valpackett: the unlocked one, scroll back

21:53 <steev> i found the fix patch for it

21:54 <steev> https://lore.kernel.org/all/20250704085541.28165-1-tzimmermann@suse.de/ should do it

21:56 <steev> or you can just revert 5307dce878d4 (i think it is, the patch title is drm/gem: Acquire references on GEM handles for framebuffers )

23:24 <osmoz[m]> Hello, everyone, I I’m having problems with a custom Linux (6.16 + patch) for the zenbook a14 x1e the initramfs load and find the boot partition but a get a black screen right after

23:24 <osmoz[m]> Do you know what I should look into

23:32 <steev> do you pass any kernel command line arguments?

23:32 <alexVinarskis[m]> Could you confirm if Ubuntu's latest (plucky-desktop-arm64+x1e-20250703.iso)[1] installer/Live-CD boots up to GUI or not? If it does, there is an issue with you kernel/initrd. If not, there is an issue with devicetree (I tested it on X1 but not X1E), in this case do you perhaps have an external display to attach over Type-C to check if the issue is internal display or something else? [1]:

23:32 <alexVinarskis[m]> https://people.canonical.com/~platform/images/ubuntu-concept/

23:33 <steev> ah, i don't know what one i used because it's just downloaded as "plucky-desktop-arm64+x1e.iso

23:34 <osmoz[m]> Sorry the Ubuntu Qcom-x1e work well, forgot to mention I’m trying to setup Void

23:34 <steev> then probably kernel config and/or kernel command line

23:35 <osmoz[m]> [alexVinarskis](https://matrix.to/#/@anynameaaa:matrix.org): do you think HDMI will work over Type C ? Because I’m traveling right now so I don’t have an display port display (only hotel TVs)

23:36 <osmoz[m]> [steev](https://matrix.to/#/@_oftc_steev:matrix.org): I’m rebooting Ubuntu and will drop the grub config

23:37 <alexVinarskis[m]> steev: its being updated as the 'latest', so if you downlaoded after 3th July, its the latest

23:37 <steev> i think that is the one i used to do my install (but i guess you were asking osmoz, not for someone like me to confirm :) )

23:38 <alexVinarskis[m]> osmoze: the full-sized HDMI connector will not work just yet, it needs additional patches. Type-C to HDMI cable plugged into Type-C should(tm) work.

23:39 <osmoz[m]> Nice I will do an new install with the updated one maybe

23:39 <osmoz[m]> For grub : I have set (following message here) this -> linux /boot/vmlinux-6.16.0-rc2-next-20250620_2 root=UUID=bc3af373-0fb3-49f0-a957-6af71ae4d8f5 ro clk_ignore_unused pd_ignore_unused cma=128M regulator_ignore_unused ram=31G

23:41 <alexVinarskis[m]> steev: ah sorry, got confused since you said you were not sure which one you used. You also have zenbook btw?

23:41 <steev> nah, i have the oled t14s with 64gb ram :)

23:42 <alexVinarskis[m]> also nice

23:42 <alexVinarskis[m]> you mean, 32gb ram /s

23:42 <steev> osmoz[m]: okay nothing seems amiss there, so either missing something in your initramfs or your kernel config - are you just using the defconfig?

23:43 <osmoz[m]> I almost choose the t14s, the choice was hard

23:43 <valpackett> steev: oh cool, thx

23:45 <osmoz[m]> maybe initramfs because for the kernel config I tried to set the same flag as the ubunut concept

23:45 <alexVinarskis[m]> might be wrong as it was a while ago, but I migh've have black screen issue when switched to new kernel but didn't change QCOM's specifc config to generic CONFIG_PHY_SNPS_EUSB2, perhaps you are missing it?

23:45 <alexVinarskis[m]> if 6.14, then you are for sure missing CONFIG_PHY_SNPS_EUSB2.

23:46 <osmoz[m]> Maybe I have it here: https://github.com/Sombre-Osmoze/void-personal-packages/blob/stable/srcpkgs/linux-zenbooka14/files/arm64-dotconfig will check in a few minutes

23:46 <osmoz[m]> I will also upload my dracut conf later

23:46 <alexVinarskis[m]> nope, you only have older CONFIG_PHY_QCOM_SNPS_EUSB2, wont work on 6.16+

23:49 <osmoz[m]> alexVinarskis: sorry maybe I did not explain clearly my situation, I have problem with Void and a 6.16 kernel that I compiled earlier (I'm using Ubu-concept) to do it

23:50 <osmoz[m]> And I have CONFIG_PHY_QCOM_SNPS_EUSB2 set to module (m)

23:50 <osmoz[m]> Or maybe I didn't understand

23:51 <alexVinarskis[m]> CONFIG_PHY_QCOM_SNPS_EUSB2 will not work on 6.16 kernel, you need CONFIG_PHY_SNPS_EUSB2 instead (notice the QCOM part missing). It got renamed somewhere between 6.14 (Ubuntu concept's image, from where you copied .config) and 6.16 (that you are trying to build).

23:52 <osmoz[m]> Ohh understood my bad, thanks, will update I re-build the kernel

23:53 <osmoz[m]> I will either way buy and HDMI cable to check over the Type-C port in I can get to a shell

23:53 <osmoz[m]> It will be easier to debug

23:53 <abby> if you have kconfig changes for void, feel free to make a pr. it'd be nice to support x1e stuff but i haven't the budget to buy any

23:55 <osmoz[m]> abby: aby that's the goal when I'm sure the setup work (I will maintain it), I also have adapted the extraction-firmware script into a void-src config

23:56 <steev> nice

23:59 <alexVinarskis[m]> osmoze: internal screen really should work, even if it was not tested, so maybe you won't need the HDMI cable. Once its working, will appreciate EDID dump. Also WiFi is not added to DT of X1E (its ath12k unlike ath11k in X1 variant) as I could not test it - would be happy to provide changes to test once you there.