03:38 <jenneron[m]> aceridus: can you push your kernel changes for galaxy book go somewhere?

14:25 <abelvesa> steev: I intend to resend that soon

14:25 <abelvesa> it kinda got ignored the first time

14:30 <steev> yeah, mostly :( hopefully round two will be better, and i'll definitely test it :)

14:56 <ajhalaney[m]> steev: Finally got around to looking at why that 6.0.3 tag failed to boot for me. pmic_glink got deferred, and it oops'ed, and somehow that made the display unhappy (unsure of the relationship there to be honest). This commit[0] on my branch[1] (which enables a few configs needed for fedora) makes things kosher again:

14:56 <ajhalaney[m]> [0] https://gitlab.com/ahalaney/linux/-/commit/1819fbccd03de430d9fd4c58ded35f5be83e9aa8

14:56 <ajhalaney[m]> [1] https://gitlab.com/ahalaney/linux/-/tree/lenovo-x13s-6.0.3-halaney

15:02 <javierm> ajhalaney[m]: in https://gitlab.com/ahalaney/linux/-/commit/0ac292859789d6ef0fbc007341dec7616a504503 you mention CONFIG_FW_LOADER_COMPRESS, but it's not enabled?

15:03 <javierm> or am I blind?

15:04 <ajhalaney[m]> I may have failed to push it up lemme double check that bit just made that update (when I tried that in the past I failed to boot, turns out it was that pmic_glink oops all along)

15:05 <steev> ajhalaney[m]: oh good to know, i'll pull in too

15:05 <steev> (eventually)

15:08 <ajhalaney[m]> javierm: thanks, guess who forgot to git add and amend? :D The links are all still good though, but not the laptop_defconfig changes are actually present. And awesome thanks steev

15:08 <ajhalaney[m]> s/not/now/ jeez.

15:09 <javierm> ajhalaney[m]: cool. I think you also need CONFIG_FW_LOADER_COMPRESS_XZ=y ?

15:09 <steev> ajhalaney[m]: i *think* [0] should go to the list? not sure if the glink stuff has already gone in

15:10 <ajhalaney[m]> steev: I'll follow up and see if jhovold already covered it in the review with bamse.. I think that patch series is still under review but I'll add to my to-do list today so it doesn't sneak away from us

15:12 <ajhalaney[m]> javierm that's set when you do a `make laptop_defconfig` (default y in Kconfig)

15:13 <javierm> ajhalaney[m]: yeah, but having it will save you to choose each time

15:13 <ajhalaney[m]> javierm long term me/Brian/Eric hope to have a kernel-ark based branch with configs and stuff ready to rock but I haven't spent the time yet to do that

15:13 <ajhalaney[m]> It won't make you choose it

15:14 <ajhalaney[m]> javierm:... (full message at <https://matrix.org/_matrix/media/r0/download/matrix.org/EkeMUMBQAZJxQubzRvnkyksE>)

15:15 <steev> please don't abuse cat

15:15 <ajhalaney[m]> took me a sec to grok that :P

15:15 <steev> this is joke, i habitually cat when i don't need to too

15:17 <javierm> ajhalaney[m]: ah! I see that the default y is fairly recent, commit 07953c54a105 ("firmware_loader: enable XZ by default if compressed support is enabled")

15:18 <ajhalaney[m]> javierm: I appreciate you looking out regardless :)

15:19 <javierm> ajhalaney[m]: I wondered why I had to choose on my v5.18 branch :)

18:36 <Leandro[m]> <jenneron[m]> "aceridus: can you push your..." <- Would be epic

18:37 <aceridus> jenneron[m]: Leandro[m]: Update from over the weekend is that I tried the 4 gpio configurations for the touchpad vdd (31 high, 31 low, 32 high, 32 low). None produced a working touchpad. It maybe a fluke or not relevant at all but when set to 32 active high I sometimes see dmesg lines related to i2c-10 setup failing. Current dst: https://dpaste.com/DL4DT62UT

18:40 <aceridus> Leandro[m]: I am happy to hear someone else wants to jump onboard and help with the same machine. I bought one off ebay recently simply because I knew it didn't work fully with Linux yet and figured it would be a good attempt at understanding device tree more. I am very much not a kernel or DT developer, but I can sort of find my way around them at a novice level...

18:41 <Leandro[m]> I'm probably not much better than you lol

18:41 <Leandro[m]> But now I have an excuse to get a cheap not-17" laptop

18:43 <aceridus> Leandro[m]: Let's figure it out together then once you get your hands on one. I have a fairly streamlined process where I rebuild the kernel/dts/rootfs in an LXD container and then move it over to a USB thumb drive. I am happy to share the bootable image if you want to get you kickstarted

18:50 <aceridus> jenneron[m]: Leandro[m]: The keyboard works, but periodically dmesg gets flooded with 'geni_i2c a80000.i2c: Bus arbitration lost, clock line undriveable' messages. Possibly due to inactivity? dmesg: https://dpaste.com/H75F5PG84

20:14 <aceridus> jenneron[m]: Leandro[m]: Here is a dmesg log when touchpad is set to GPIO 32 active-high and gets i2c_hid_of 10-0040 errors related to failing to change power settings and failing to reset the device. Not sure if this means I am on the correct pin, or maybe it means I am on the wrong pin? https://dpaste.com/3BD368AKD

20:32 <aceridus> It is also worth noting that I don't always see the i2c_hid_of 10-0040 errors, it seems more often than not I don't see any messages related to i2c_hid_of 10-0040

20:35 <hexdump0815> aceridus: jenneron[m]: doing the same correlation like https://oftc.irclog.whitequark.org/aarch64-laptops/2022-10-21#31542172 for I2C8 from the dsdt can it be that it should be i2c7 and not i2c10 in linux?

20:41 <hexdump0815> oh i think i was wrong I2C8 seems to be something else

20:42 <hexdump0815> the other hid protocol device seems to be IC11 in the dsdt which maps to i2c10, so looks like that one is correct ... please forget what i mentioned before :)

20:45 <aceridus> Yea, I think the added complexity for the touchpad is it has an external regulator that needs to be enabled. I am working off of jenneron's recommendations and examples, so I don't know that I am getting the vdd regulator gpio correct...

21:42 <jenneron[m]> I would hardcode this power sequence in hid over i2c driver

21:42 <jenneron[m]> to test

21:46 <jenneron[m]> also please push your changes so it will be easier to look at it

21:56 <ajhalaney[m]> steev: posted on the pmic_glink patch upstream fyi https://lore.kernel.org/linux-arm-msm/20221024200222.vg7vhu7dfd3wlnkc@halaney-x13s/T/#u

21:57 <steev> i'll have to try that out on the flex 5g as well

22:03 <steev> ajhalaney[m]: ah, yeah, the audio stuff went away because there's newer versions posted to the ML but i havent' pulled them in, sorry

22:04 <steev> there are more bits required anyway

22:11 <aceridus> jenneron[m]: push my changes where?

22:11 <jenneron[m]> aceridus: gitlab, github or whatever

23:30 <Leandro[m]> It would be cool if you chose gitlab, no more Microsoft shit than needed then