03:49 <kuruczgy[m]> Good to hear that there are now working camera patches for the slim7x.

03:49 <kuruczgy[m]> Hopefully will have time to also test it out in a couple days.

03:49 <kuruczgy[m]> In the meantime, what's the deal with the green camera images? I read that this is an issue with debayering, correct?

03:49 <kuruczgy[m]> I don't get how that works. If the camera is sending raw sensor information, every pixel is just a raw scalar value at that point, and the concept of color shouldn't even exist at that point?

03:49 <kuruczgy[m]> The image being "green" sound more like some broken debayering to me. Where in the stack is debayering supposed to happen?

06:51 <JosDehaes[m]> Jens Glathe: external screen also works. Tested at home with 1440p60 screen, working fine and mouse/kb pass through same cable. External screen is not scaled correctly though (this may be a GNOME issue?). At work, I could drive 2 1440p60 screens but also here the external screens are scaled like 150% or even 200% even though they are set to 100% in GNOME settings. Changing scaling and applying locks up GNOME settings, and the

06:51 <JosDehaes[m]> scaling is not applied

06:54 <JensGlathe[m]> I see similar issues here if I try to go for fractional scaling.

06:54 <JosDehaes[m]> is it a driver issue you think?

06:55 <JensGlathe[m]> Not sure.

06:57 <alexVinarskis[m]> Fractional scaling works very well here (150% on internal screen, 100% external), but im using Wayland on Gnome with Ubuntu 24.10, 25.04 concept.

06:58 <JosDehaes[m]> I'm also using GNOME on wayland. Arch linux though (but that should not matter)

08:11 <Pengyu[m]> alexVinarskis: I saw your aeob helper occasionally, good jobs. Some suggestions. Actually, there are several aeob blocks in qcdx ext drivers, you can decompile it, it is important for panel bringup. I think these laptops are QRD based, so mtp things can be dropped.

08:20 <JeromedeBretagne[m]> Panel bring-up, I am interested (still for the Surface Pro 9 5G)! Any checks / instructions to run?

08:36 <alexVinarskis[m]> Pengyu: indeed, these are QRD based. Thought to keep everything though just in case. By several blocks ,you mean additional ones that are not yet in the repo? Do you have sample names? Else I could try to decompile every .bin there is haha

08:38 <alexVinarskis[m]> jdb: yes, its the thing i mentioned to you before, check out readme https://github.com/alexVinarskis/qcom-aeob-dumps. If you can open PR with decompiled files, I could help analyze them.

08:42 <Pengyu[m]> <JeromedeBretagne[m]> "Panel bring-up, I am interested..." <- for sc8280xp, check qcdxkmext8280_qrd.bin, splitting it, every block starts with AeoB headers. I can't get a computer nowadays. So you are on your own.

08:43 <Pengyu[m]> After splitting, use tools from gus

08:43 <Pengyu[m]> alexVinarskis: may be qcdxkmext8380_qrd.bin

08:49 <Pengyu[m]> I never checked all binary files. But there may be other grouped AeoB files,

08:50 <alexVinarskis[m]> ohh thats a great tip, thanks. Didn't realize they could combine multiple AeoBs together. Will check that over the weekends.

08:56 <Pengyu[m]> Thanks. Nowadays, I am busy with real life things. Then I am counting on you.☺️

11:37 <JeromedeBretagne[m]> <alexVinarskis[m]> "weird thing is that one of the..." <- alexVinarskis: I can confirm that backlight is still working with a more minimalist patch, when removing the nodes mentioning GPIO 9 (as you had shared on pastebin), good point. I will clean up and simplify my earlier commit

11:37 <JeromedeBretagne[m]> <alexVinarskis[m]> "jdb: yes, its the thing i..." <- I will have a look!

13:13 <efenex> hello, I've recently bought a Zenbook A14 and attempting to install linux on it. I've built the kernel based on https://code.launchpad.net/~alexvinarskis/ubuntu/+source/linux/+git/linux/+ref/qcom-x1e-asus-zenbook-a14 but now trying to find a way to get it onto the installer ISO. I've tried "cubic" but it seems to fail on unpacking the squashfs from the X1E concept ISO, so that won't

13:13 <efenex> work. The ubuntu post indicates livefs-editor should work, but I'm not sure what I need to replace where in order to get the custom kernel to be used. Anyone who's done this before has some pointers or some livefs-editor commands in their history that they can share? :)

14:01 <JeromedeBretagne[m]> <JeromedeBretagne[m]> "alexVinarskis: Thanks a lot for..." <- I must have had an hallucination, now I see again this error message....... (full message at <https://matrix.org/oftc/media/v1/media/download/AUYXaHTjkyP4zRKW8kGoVdv8GZtmXGMslfwrua8IF8QVvr8uJioYhdQMAmSMytvfBMtNmAGT2enkcPDNTpAF5X1CeXIX9q_gAG1hdHJpeC5vcmcvV3VZSGFtQ3NveGVKckZCTnlWVlF4UUhB>)

14:09 <JeromedeBretagne[m]> Here is the reduced patch to enable backlight control on the Surface Pro 9 5G (arcata):

14:09 <JeromedeBretagne[m]> https://github.com/JeromeDeBretagne/linux_surface_pro_9_5G/commit/f03fcd98f393a4dab020bd9bcb185cd0120c30c8

14:10 <JeromedeBretagne[m]> and the work-in-progress patch to enable the built-in display on mdss0_dp3 (not working):

14:10 <JeromedeBretagne[m]> https://github.com/JeromeDeBretagne/linux_surface_pro_9_5G/commit/697fa90584b3e701166e6beccc57e750a90feb5b

14:20 <alexVinarskis[m]> jdb: could you share every .bin file from windows in dm? I can go over the weekends to decompile known AeoB files but also try to get combo AeoB as per Pengyu suggestions.

15:26 <JeromedeBretagne[m]> yes, I was looking at the README.md at the moment to extract the files, I was wondering where they are located on the Windows partition

15:27 <JeromedeBretagne[m]> I have seen some examples of files mentioned here so I will try to find the base directory:

15:27 <JeromedeBretagne[m]> https://github.com/WOA-Project/AeoBUtils

15:29 <alexVinarskis[m]> C:\Windows\System32\ is a good start. You can search for all .bin files, and then just copy them out to different folders. If i understood correctly, the tool mentioned will fail with combo AeoB files... so ill likely have to ask you to partially re-do this in a few days. But even the files already in the repo are a good start!

15:29 <exeat> craftyguy: steev: Do I understand correctly that you have successfully updated the x13s bios with fwupdtool?

15:32 <steev> exeat: si

15:34 <exeat> Nice :)

15:44 <steev> it is quite nice, though i did not check what is actually new in it

16:15 <Pengyu[m]> alexVinarskis: jdb Take x13s as an example, G0MD from qcdx ext binary.

16:26 <alexVinarskis[m]> Do you have a script for splitting multi AeoB into single AoeB files already? Else ill make one

16:27 <alexVinarskis[m]> are these pushed somewhere publically? would be curious to take a look

16:28 <Pengyu[m]> dd if of is a stupid one, I split one block a time.

16:28 <Pengyu[m]> together with the help of xxd

16:29 <craftyguy> has there been any news lately wrt lenovo releasing venus fw for the x13s? or is all hope lost at this point?

16:30 <Pengyu[m]> Writing one to extract all blocks is welcome, dd is inefficient😊

16:30 <craftyguy> exeat: yes I successfully upgraded the x13s fw with fwupdtool, using this process: https://gitlab.com/TheOneWithTheBraid/x13s-firmware-update/-/merge_requests/1/diffs

16:35 <alexVinarskis[m]> pengyu: will add to my repo if I find a better way to do this. If x13s's AeoBs aren't shared yet, mind opening PR to my repo? Would be nice to keep things in one place

16:37 <jhovold> craftyguy: no news about x13s venus fw, I remind them every now and then

16:39 <Pengyu[m]> alexVinarskis: Actually, I don't own a x13s, I downloaded x13s from Lenovo(because it can be compared to the mainline power sequence). You can download it, then extract it with innoextract.

16:39 <Pengyu[m]> https://pcsupport.lenovo.com/us/en/products/laptops-and-netbooks/thinkpad-x-series-laptops/thinkpad-x13s-type-21bx-21by/downloads/driver-list/component?name=Display%20and%20Video%20Graphics&id=635380F4-448F-4129-AA77-265289D96F6F

16:40 <alexVinarskis[m]> ahhh, though you did you for your machine. thanks. you can actually simply 7zip -e filename.exe to extract and flatten on linux directly.

16:42 <Pengyu[m]> Got it👍

16:56 <JeromedeBretagne[m]> alexVinarskis: Here is my pull request :-)

16:56 <JeromedeBretagne[m]> https://github.com/alexVinarskis/qcom-aeob-dumps/pull/1

17:30 <craftyguy> jhovold: I see, thanks for staying on them :)

21:17 <alexVinarskis[m]> Pengyu: added magic AeoB parses, worked like a charm, thanks for the tip! https://github.com/alexVinarskis/qcom-aeob-dumps/pull/2

21:17 <alexVinarskis[m]> parser*

22:12 <alexVinarskis[m]> jdb: added split AeoB for surface here https://github.com/alexVinarskis/qcom-aeob-dumps/commit/447108e27fba23735a67d25342c16b3914fb384f. Found new `gpio36`. You could add it as vreg enable, or just try it on live system via `sudo gpioset gpiochip0 36=1`, or `=0`.

22:19 <JeromedeBretagne[m]> Can you point out in which of the newer files you've find this gpio36 ?

22:22 <alexVinarskis[m]> as per pengyu's suggestion, in one of the qcdxkmext8280..: https://github.com/alexVinarskis/qcom-aeob-dumps/blob/master/microsoft-surface-pro-9-5G-1996/qcdxkmext8280_ARC_part2.json#L156-L168

22:23 <JeromedeBretagne[m]> ok, so 0x24 for gpio36, got it !

22:23 <alexVinarskis[m]> actually you need 2nd instance of that gpio, 1st instance is the off state (LDOs are at 0, gpio at 0), next one LDOs have values, and gpio is 1, so its an active high. yes its just hex->dec.

22:25 <JeromedeBretagne[m]> which line corresponds to which state?

22:26 <JeromedeBretagne[m]> versus... (full message at <https://matrix.org/oftc/media/v1/media/download/ATX2A5rwVByKWKwywClbPCn6etf5rVNtp8eFElDQGtoomt7WakrdkRMR-Bk4rqqdSaSueQ6md4wUYkXKMNW3sf9CeXI02J7AAG1hdHJpeC5vcmcvT2FqRFBFaFhmVEhOcHZZbldJQkhrck54>)

22:26 <alexVinarskis[m]> <Pengyu[m]> "1000078838.png" <- here Pengyu has very nice comments

22:27 <JeromedeBretagne[m]> indeed, I had missed the comments in this picture, I'll need to read quietly tomorrow to see how it transforms into .dts nodes:

22:28 <JeromedeBretagne[m]> s/:/!/

22:48 <JeromedeBretagne[m]> Funny how it corresponds to the dsdt somehow but with some additional info in these files. I guess on Windows, using the .bin files takes precedence over the acpi tables as they are more detailed

22:53 <JeromedeBretagne[m]> Next I need to see how to create 2 instances of the same gpio for the on and off states ("EXIT" and "ENTER"). Need to get some sleep for today. Thanks so much alexVinarskis !