00:04 <craftyguy> nice, thanks for trying it. I was about to but was in the process of saving work elsewhere juuuuust in case 😅

00:10 <steev> luckily, all our stuff is open source, so i can just re-clone after rebuilding :D

00:13 <craftyguy> neato, the update worked for me too

00:14 <craftyguy> ya "saving work elsewhere" is adding commits and pushing heh

00:31 <steev> definitely handy! thanks for pointing that out!

01:42 <freekurt[m]> Wait, how do I upgrade the firmware on my x13s without windows? Run that 2 year old script?

02:19 <robclark> fwupd apparently

02:19 <robclark> I hope that also works on newer things

02:26 <steev> freekurt[m]: i downloaded the exe from lenovo's website, ran `innoextract n3huj22w.exe`, then ran the 2 commands in the link that craftyguy asked about - `fwupdmgr get-devices 2>/dev/null| grep -A 1 "System Firmware"` and then `sudo fwupdtool install-blob 'code$GetExtractPath$/Rfs/Fw/N3HET92W/\$0AN3H00.FL1' b3f2723a04671a24bdf2d087cf9da60dcf5cb886` (where b3f2723.... was the device id that is spit out by the get-devices command

02:52 <steev> could probably pretty it up and awk to only print the device id

02:53 <steev> fwupdmgr get-devices 2>/dev/null | grep -A 1 "System Firmware" | grep "Device ID:" | awk '{print $NF}'

02:57 <dgilmore> b3f2723a04671a24bdf2d087cf9da60dcf5cb886 is probably the device id on all x13s's

03:01 <craftyguy> Maybe, but this doesn't seem like a good time to make assumptions if it could mean bricking something 😅

04:41 <steev> true

06:41 <JensGlathe[m]> rc6 seems better at suspend/resume

06:47 <JeromedeBretagne[m]> alexVinarskis: Thanks a lot for the feedback! Jens Glathe: also told me that the display seems to be using GPIO 3 (instead of 25), now I understand why from the ACPI dumps. I quickly tested it late yesterday, now I don't have the following error message anymore:

06:47 <JeromedeBretagne[m]> ```panel-simple-dp-aux aux-aea0000.displayport-controller: Couldn't power on panel to ID it; using conservative timings: -110 ```

06:48 <JeromedeBretagne[m]> What do you mean by "no gpios node listed" ?

06:51 <alexVinarskis[m]> Pmics with gpios have pmicName_gpios node in dtsi definition. 0th pmic for your platform doesn't. I checked, on Zenbook with x1p there are also x2 pmicgpios defined for display, one seems to not exist (or is something else?) as well, 2nd is enable gpio for the panel.

06:52 <JeromedeBretagne[m]> I'm planning to switch into drm.debug mode to see what's going on, as some other Surface displays need a quirk or 2, or some explicit parameters like link-frequencies

06:53 <alexVinarskis[m]> So no error, but still no image i assume? Did you try adding enable gpio for the panel as per my example? Or you could try to set it with gpioset for the test

06:54 <JensGlathe[m]> The dsdt provided displays have 2 lanes defined (eDP1.2), for 2880x1920 I would either expect HBR (810MHz link speed) or more lanes

06:57 <alexVinarskis[m]> jdb: you may also try reading panel's EDID from linux, eg. via cat /sys/class/drm/card1-eDP-1/edid | edid-decode to confirm that it is indeed powered on.

07:05 <JeromedeBretagne[m]> I don't see the edid file nor the eDP-1 directory yet in /sys/ . I can only see the `card1-DP-[12]` and the edid for my external monitor on the bottom UBS-C port (the only one working for video out so far)

07:05 <JeromedeBretagne[m]> Not having `card1-eDP-1` necessarily means that it is not powered on?

07:08 <JensGlathe[m]> I had the same effect with the newer panel-edp. Maybe it is anothe enable gpio required

07:10 <alexVinarskis[m]> Could be. Jdb: is that with pmicgpio as panel enable gpio as well or just with vreg enable gpio3?

07:12 <JeromedeBretagne[m]> My late night test was with:... (full message at <https://matrix.org/oftc/media/v1/media/download/ATifBlq8vL-BnH9mbBd0eMlWsauKpMF8C6bW8tfstgzw92rW766hp0j-K7yPxe7XNXz-H8O3qKFerCgSGU1d4IJCeXFbwQ6AAG1hdHJpeC5vcmcvTmJacUtDbW1za050U2dvbExRRUNwbEpw>)

07:12 <JeromedeBretagne[m]> Now, I'm on the go for work so my next tests will be later today

07:17 <JeromedeBretagne[m]> alexVinarskis: About your suggestion "Did you try adding enable gpio for the panel as per my example?", do you know if some other devicetrees have it as a reference example? I would need to add it into the `display` node? It doesn't seem very common

07:18 <JensGlathe[m]> The x1e ones have these, for example T14s (oled)

07:19 <JensGlathe[m]> But probably the PMICGPIO needs to be decoded first

07:19 <JeromedeBretagne[m]> Ok, I will have a look at this one, thanks for the pointer Jens!

07:44 <alexVinarskis[m]> <JensGlathe[m]> "But probably the PMICGPIO..." <- `enable-gpios = <&pmc8280c_gpios 8 GPIO ACTIVE_HIGH>`

08:11 <JeromedeBretagne[m]> That's exactly what I had in mind looking at the T14s dt, thanks for the explicit sample code !

09:20 <deathmist> SpieringsAE: how's external display for you on X1E Vivobook? I've got one now and compared to my old Slim 7x I can just get one monitor up at most currently on https://github.com/jglathe/linux_ms_dev_kit/tree/jg/ubuntu-qcom-x1e-6.15rc. https://github.com/SpieringsAE/linux/commit/425aced didn't seem to help with the built-in HDMI port either but maybe I missed something

10:50 <SpieringsAE> deathmist: I have not tried multiple displays yet, the hdmi port I've only managed to get working by booting the laptop while the hdmi cable is already inserted

10:51 <SpieringsAE> hotpluggin seems to be an issue for now

10:51 <deathmist> even with only HDMI plugged in while booting I couldn't get it detected

10:52 <SpieringsAE> hmm, what kind of monitor do you have? it seems to work with my 1440p one, but my 4k screen at work gets very weird with the hdmi port

10:52 <SpieringsAE> any messages in dmesg

10:52 <SpieringsAE> ?

10:53 <SpieringsAE> but yeah the hdmi port is very funky right

10:53 <SpieringsAE> now

11:03 <alexVinarskis[m]> Same hdmi setup also works most of times on coldboot on zenbook with 3440*1440. No hotplug yet, looking into it, but so far not much luck.

13:01 <deathmist> SpieringsAE: it's just some older 144 Hz ViewSonic, on coldplug HDMI boot I do see [drm:msm_dp_ctrl_link_train_1_2 [msm]] *ERROR* link training #1 on phy 0 failed. ret=-6 towards the bottom

13:03 <deathmist> the port also behaved odd under windows and wouldn't calm down messing with blanking my even my main monitor until I reset it back to 60 Hz, then I could put it up to 120 Hz and things seemed ok

13:08 <deathmist> I do wonder if something from the existing the older MST patches can be pieced together for X1E. that way I could use a dock and plug in the original 90W charger to the other USB-C as I've finally confirmed the Vivobook appears to more or less be able to stand all-core compile loads without draining the battery from some brief tests :D

13:54 <davidinux> has anyone tried upgrading from 24.04 concept to 25.04 on a T14s?

13:54 <davidinux> and if so, anything to pay attention to before upgrading, or is the upgrade otherwise smooth

15:20 <alexVinarskis[m]> jdb: please ignore my comment above, its wrong. gpio3 is input pin (could be HPD for example). The only two other gpios mentioned are now defined for backlight enable and backlight PWM, and they seems to be correct. Does backlight work in your current setup, eg. can you see difference when adjusting the brightness?

15:26 <alexVinarskis[m]> weird thing is that one of the pins present for backlight enable on lenovo x13s is not actually listed in your dtst either... I removed stuff related to it, but thats not helping to power it up :D

15:26 <alexVinarskis[m]> here is the diff I reworked from your commit https://pastebin.com/t9WvuHL5, though I guess its very unlikely to work. You could try dumping AeoB files, perhaps they have some usefull TLMMGPIO/PMICGPIO there, if there is a windows driver that enables the panel somehow. Another option is to update BIOS to the latest and dump acpi again - they update these and sometimes add previously missing definitions.

15:40 <deathmist> meanwhile I found out Asus just hates the idea of giving me HDMI through an Aukey CB-C91 which was my previously known-working setup for the other display on Slim 7x. still works on another PC but on the X1E Vivobook it doesn't even work in Windows in neither port in neither orientation

15:44 <bryanodonoghue> alexVinarskis[m] anthony25

15:44 <bryanodonoghue> https://git.codelinaro.org/bryan.odonoghue/kernel/-/tree/linaro/connect-demo-kernel-x14s-sensor-debug?ref_type=heads

15:44 <bryanodonoghue> t14s works - I suspect slim7x will work too

15:44 <bryanodonoghue> data-lanes=<2> in the camera sensor should have been data-lanes = <1 2>

15:44 <bryanodonoghue> please test if you get a chance

15:45 <bryanodonoghue> this branch should work with Inspiron 14p, Dell XPS, Lenovo t14s, CRD and hopefully slim7x now too

15:47 <alexVinarskis[m]> I see this does not include vreg fix that I've shared before - without that one the sensor does not even probe on Yoga 7slim.

15:57 <alexVinarskis[m]> bryanodonoghue tested on Zenbook, it works! the amount of times i went over those lines and missed it... thank you

15:59 <alexVinarskis[m]> though, it still spams dmesg errors i mentioned before (cam_cc_ife_1_gdsc status stuck at 'off' then repeating cam_cc_pll8 failed to enable!). Seems its not critical though.

16:23 <JeromedeBretagne[m]> alexVinarskis: yes, I can see backlight adjusting with my first patch. It's difficult to detect under normal lighting conditions, it's more visible in a dark room, using the backlight +/- buttons from the Surface Typecover

16:52 <alexVinarskis[m]> but then either extra gpios are hidden in AeoB (can check if you dump them), or thats it, there is not explicit enable pin somehow ( try my reduced above)

17:13 <anthony25> bryanodonoghue: it's just the data-lanes to change or there is other things?

17:14 <anthony25> alexVinarskis[m]: I'll try with your vreg fix

17:14 <alexVinarskis[m]> Just data lanes.

18:02 <alexVinarskis> bryanodonoghue: any chance you could share ov02c10's register and bit mask for vflip/hflip/mirror? It seems sensor supports it but its not implemented. In Asus Zenboook camera is upside down (needs vflip). On XPS for whats it worth it would need hflip. Some other sensors like ov02a10 already support this, I wouldn't mind implementing it, but guessing registers and bits is a little annoying....

19:43 <JosDehaes[m]> bryanodonoghue: I get this with slim7x:... (full message at <https://matrix.org/oftc/media/v1/media/download/AeW9T4rAlut0UsoiB_lKhpKgO2YItSvllBW9_ATHNqGYRib2bAMzdkPfExkAFy1WA0ZMmdnyFKDYcUWdpjV5WYtCeXGGwEmAAG1hdHJpeC5vcmcvZW5paVlCcnNncFdXTmhtZWNnZHZVQW5j>)

19:48 <alexVinarskis[m]> could you please try with this patch? https://bpa.st/KDWA

19:54 <JosDehaes[m]> Yes!

19:54 <JosDehaes[m]> [ 0.956381] ov02c10 1-0036: ov02c10_identify_module read chip id 0x00005602 want 0x00005602

19:55 <JosDehaes[m]> with cheese, it says can't open the device, but the Camera app does show the image (upside down and colors very green)

19:56 <JosDehaes[m]> * with cheese, it says there was an error playing video, but the Camera app does show the image (upside down and colors very green)

19:58 <JosDehaes[m]> * with cheese, it says there was an error playing video, but the Camera app does show the image (upside down like you said on zenbook and colors very green)

20:02 <alexVinarskis[m]> Cheese and eg. Firefox will only work when camera is exposed over pipewire, that needs extra steps. But this is awesome.

20:03 <alexVinarskis[m]> Will share a solution for camera rotation in 5 mins, got mine straightened out.

20:03 <JosDehaes[m]> happy to test!

20:15 <alexVinarskis[m]> added ov02c10 rotation support and patch to enable it for zenbook as yoga over here: https://github.com/alexVinarskis/linux-x1e80100-zenbook-a14/pull/1

20:18 <alexVinarskis[m]> probably not the most elegant/ready-to-upstream solution, seems best is to implement vflip and hflip functions, and call both when roatation of 180degs is expected. I have also just realized that camera probably should not be hfliped, as that would be done by client software (eg. video calling webapp).

20:25 <JosDehaes[m]> I just added `rotation = <180>;` and it worked!