robclark changed the topic of #aarch64-laptops to: Linux support for AArch64 Laptops (Chrome OS Trogdor Devices - Asus NovaGo TP370QL - HP Envy x2 - Lenovo Mixx 630 - Lenovo Yoga C630 - Lenovo ThinkPad X13s - and various other snapdragon laptops) - https://oftc.irclog.whitequark.org/aarch64-laptops
<steev> you can always set them to that yourself
<steev> that's what i do, but again, we don't yet know where it's storing that info
<nerdboy> crap, on 6.7 i'm not seeing battery, usb display, or audio device
<nerdboy> i have alsa-ucm-conf master, which is several commits beyond the volume fixes
<nerdboy> i have linux-firmware master
<nerdboy> usb devices and kernel modules looks correct for the usb display stuff
<nerdboy> are there still missing firmware bits to add?
Ablu has quit [Read error: Connection reset by peer]
Ablu has joined #aarch64-laptops
<enyalios> nerdboy: im running steevs 6.7.0 with laptop_defconfig, linux-firmware-20231211, alsa-ucm-conf-1.2.10 + 335 patch
<enyalios> and i have working battery and audio, but i dont have a usb display to test with
hexdump01 has joined #aarch64-laptops
hexdump0815 has quit [Ping timeout: 480 seconds]
<steev> do you have pd-mapper running?
<steev> without that, battery and audio aren't gonna
<steev> while i do have lumag's pd mapper kernel modules, they don't *quite* work 100% on the x13s yet
<steev> it's the "userspace dependencies" section of johan's wiki
<nerdboy> it claims to be running...
<nerdboy> lemme go poke at it
<nerdboy> it doesn't appear to do anything, but i'm running the latest "version" so maybe i need newer?
<steev> what is in your /lib/firmware/qcom/sc8280xp/LENOVO/21BX/ directory
<steev> it doesn't, afaik, provide much output
<steev> also, dmesg output please
<nerdboy> dmesg: https://bpa.st/JMVA
<steev> [ 1.810111] remoteproc remoteproc0: Direct firmware load for qcom/sc8280xp/LENOVO/21BX/qcadsp8280.mbn failed with error -2
<steev> you're missing the firmwares?
<nerdboy> i have later than the version in the wiki?
<nerdboy> qcadsp8280.mbn is in the listing
<steev> try putting them into the initrd
<steev> remoteproc is built in
<steev> because i'm an idiot
<steev> i'm not entirely sure though, but your system is definitely acting like the remote procs aren't running
<nerdboy> any other /lib/fw dirs besides qcom and qca?
<nerdboy> dracut seems a little wonky but i got those ^^ 2 dirs in there
<nerdboy> <sigh> only some of the files actually made it
<steev> should only need the mbns because pd-mapper won't run til userspace
<jglathe> the mbns need to be in the initramfs, with a hook https://drive.google.com/file/d/1yvtVdx9t6qVNcIhw1i__qz122yHQAGUQ/view?usp=sharing
jglathe_ has joined #aarch64-laptops
jglathe_ has quit [Remote host closed the connection]
<jenneron[m]> bamse: i'm wondering, is there any more work planned for sc8180x and lenovo flex 5g from your side?
<_[m]1> hm still no audio nor wifi on my x13s with debian
<_[m]1> I have latest protection-domain-mapper
<_[m]1> I already had this symlink https://bugs.debian.org/1057040 and I tried (and probably failed) to put latest alsa ucm myself
<_[m]1> I also have experimental packages enabled for trixie repo
<_[m]1> is this somehow wrong
<_[m]1> `lrwxrwxrwx 1 root root 45 Dec 7 23:12 qcom/sc8280xp/SC8280XP-LENOVO-X13S-tplg.bin -> qcom/sc8280xp/LENOVO/21BX/audioreach-tplg.bin
<_[m]1> `
<_[m]1> damn it seems so, made it to point to LENOVO/21BX/audioreach-tplg.bin
<_[m]1> yay sound 🎉
<_[m]1> still crackling though
<_[m]1> * crackling though and playing too fast mm
<calebccff> steev: even with your kernel and your defconfig i still get no display on my c630... do you include display modules in the initramfs?
<_[m]1> still got some fw issues seems https://paste.debian.net/1304074/
<_[m]1> building 6.7 rc8
<_[m]1> * building steevs 6.7 rc8
<jenneron[m]> steev: pushed 6.7 stable to master branch. UFS, bluetooth (with bootmac), headphones, touchscreen are fine
<jenneron[m]> also, disabled headset keys since KEY_PLAYPAUSE randomly occurs and it's annoying..
<jenneron[m]> touchpad issue still happens though
hightower2 has quit [Ping timeout: 480 seconds]
hightower2 has joined #aarch64-laptops
<_[m]1> rc8 -> no more error in dmesg but still no wifi, feels like it's a userland config 🤔
<_[m]1> in the end when I set it manage by NM and do nmcli dev wifi connect [...] it connects 😲
<jglathe> the a660 files come from linux-firmware, the *8280.mbn files should be in linux-firmware, too. Anyway, I copied them from the windows installation
<_[m]1> <jglathe> "@_[m]1 do you have the fw..." <- I'm assuming it's in steevs branch(es) yes
<_[m]1> or I suck too much in wpa_supplicant config or nmcli did some magic but it works now
<jglathe> I guess it isn't. I'm on ubuntu mantic here, linux-firmware should have them
<_[m]1> I have 20230625-2
<_[m]1> trying steam again, seems to work fine but doesn't launch the GUI in the end 🤔
<jglathe> 20230919.git3672ccab-0ubuntu2.5
<jglathe> linux-firmware: /lib/firmware/qcom/sc8280xp/LENOVO/21BX/qcdxkmsuc8280.mbn.zst
<bamse> jenneron[m]: "planned" is a strong word...
<bamse> jenneron[m]: i think the only planned work i have on sc8180x is that i intend to get it into kernelci...which might result in some unplanned work
<bamse> jenneron[m]: is there anything in particular that you're looking for?
<jenneron[m]> bamse: well it would be good to have support for speakers (soundwire fails for me), mics (needs routing in dts + ucm) and venus video decoder
<jenneron[m]> these at least don't require as much work as e.g. fingerprint
<jenneron[m]> i mostly gave up on working on audio after getting headphones to work, so i'm just wondering if it can ever happen
<jenneron[m]> also touchpad issue is quite annoying, still have no idea how to fix it
<bamse> jenneron[m]: i'm afraid i don't understand how audio works...but vkoul| was looking at this at some point, iirc it's quite similar to other already supported platforms
<bamse> jenneron[m]: what's up with the touchpad?
<jenneron[m]> audio is pretty similar to c630, but soundwire fails, thus speakers don't work
<bamse> nice with the touchscreen, if you haven't already, please add a commit message and send upstream
<bamse> jenneron[m]: vkoul| had problems with soundwire as well, but i thought srinik fixed a very similar issue when bringing up audio on x13s
<jenneron[m]> bamse: if you spam click e.g. middle click (with 3 fingers) at some point it may become stuck on 3 fingers. you won't be able to move cursor
<jenneron[m]> stuck on middle click* it's better to say it this way
<jenneron[m]> sometimes it gets stuck on right click as well
<jenneron[m]> if you tap with 3 fingers again it gets back to normal state
<jenneron[m]> so my way to reproduce it: press middle click 3 times very quickly, then check whether it's stuck. in one of 5-10 cases it gets stuck for me
<bamse> jenneron[m]: that issue i've actually seen...but presumably only on flex
<bamse> jenneron[m]: feels like it would be a firmware issue then
<jenneron[m]> iirc it works fine on windows though...
<jenneron[m]> i'm not sure how we would upgrade firmware. fwupd detects only camera and UFS
<jenneron[m]> *firmware of touchpad
<bamse> okay, then perhaps it's not firmware after all
<bamse> but it's deeper into the hid driver than i've been
<jenneron[m]> <bamse> "nice with the touchscreen, if..." <- i need to figure out how to upstream patches first of all, heh. but i will definitely try upstreaming touchscreen binding
<jenneron[m]> bamse: since we can reproduce it on spam-clicking, i thought that the problem might be interrupt. i tried setting it to edge falling, but i'm not sure it helped, since it still happens. edge falling is also against i2c hid specification. i'm not sure what else we could do with interrupt
<jenneron[m]> maybe we should configure i2c for a higher frequency
<bamse> for the longest time i had one of those interrupt flags wrong, so we effectiely had an irq storm...which made all other interrupts sluggish, but i believe we fixed that before pushing upstream, right?
<bamse> jenneron[m]: i don't remember if the i2c bus speed comes from acpi, of i just picked a standard number...so definitely worth investigating
<jenneron[m]> apparently we can set 400 and 1000 kHz i2c frequency https://gitlab.com/sc8180x-mainline/linux/-/blob/master/drivers/i2c/busses/i2c-qcom-geni.c#L150-152
<jenneron[m]> i will try setting it to 1k
<bamse> jenneron[m]: sounds good, but please do take a look at b4 prep / b4 send and push that touchscreen patch (and other things that you might come up with)
<jenneron[m]> yeah i definitely should look into it
<bamse> jenneron[m]: there are a few other users out there whom's interest you might stirr up by sending a few patches...
<jenneron[m]> bamse: i don't think i can upstream anything other than touchscreen
<bamse> jenneron[m]: start with that :)
<jenneron[m]> we have camera working, but it depends on usb-mp patches, headphones depend on sm8150 driver being developed by someone else, i just picked it to my tree and did a few modifications
<jenneron[m]> unfortunately these things are not upstreamable from my side right now
<jenneron[m]> but i'm going to start with touchscreen
<bamse> don't we still have a fair amount of warnings when you run make dtbs_check (or make sc8180x-lenogo... CHECK_DTBS=1) ?
<bamse> not sure if that's just cosmetic cleanup, or if there are any issues hidden in there
<bamse> either way, contributions for that is definitely welcome
<jenneron[m]> bamse: i think i can introduce a few more by adding touchscreen, heh
<jenneron[m]> since it uses dt bindings recently added to the driver, but not added to the yaml documentation
<bamse> jenneron[m]: the binding should have been updated accordingly...
<bamse> jenneron[m]: and if it hasn't, then you have another opportunity to contribute
<bamse> anyway, start with figuring out how to send that one patch
<steev> jenneron[m]: well, fun fact, the c630 doesn't work anymore either for audio :) there's some soundwire unflow stuff
<steev> calebccff: it does look like i have msm, and ti-sn65dsi86 in the initrd, and i do include the a630 sqe and gmu fw as well
<jenneron[m]> it seems that i discovered something about UFS dying on high load
<jenneron[m]> 6.7 = dies
<jenneron[m]> 6.7 + 2 patchsets from konradybcio = doesn't die
<jenneron[m]> 6.7 + 2 patchsets from konradybcio + 1k kHz i2c = dies
<jenneron[m]> however it needs much higher load than before..
<jenneron[m]> and i'm not 100% sure that i2c affects it
<calebccff> steev: hmm, i wouldn't think firmware should make a different, i also have the modules but i'll double check
<steev> firmware shouldn't, since it's one of those things
<steev> that it will retry every time the gpu tries to open
<calebccff> right
<steev> does seem to load at 1.7 seconds in so
<calebccff> steev: can you make any sense of this?... https://p.calebs.dev/77e871@raw
<calebccff> turned on a bunch of driver core debugging
<calebccff> it's constantly trying to probe the backlight device and the bridge but deferring on both
<calebccff> and then idk what goes wrong with switch_root...
<steev> maybe i'm not seeing it because i have i2c built in?
<calebccff> ohh dude i2c is built as a module in my config,, i never noticed :sob:
VintourZ2 has joined #aarch64-laptops
VintourZ2 has quit []
<calebccff> finally, feels good to get this device booting again
<calebccff> gonna be a great little tablet for running rnote
<nerdboy> enyalios: how are you generating your initramfs and are you including stuff from /lib/firmware ?
<nerdboy> i was using dracut before the hw died but not having any luck now
<steev> calebccff: oh good to hear yours is booting again :D
<steev> now fix the audio :P
<calebccff> steev: i feel like im probably being dumb, but even with debug logs turned on i couldn't clearly see why the bridge driver kept probe deferring... the fact that it's parent device driver was an unavailable module surely ought to be something the kernel screams louder about?!
<calebccff> lol audio
<steev> calebccff: well the display is i2c so the messages about i2c were what made me think of it
* nerdboy spinning a clean 6.7.0 laptop_defconfig
<calebccff> steev: yeah fair enough x3. btw how does your partition table look on the c630? did you just nuke the whole LUN and repartition?
<steev> i (sadly) still keep windows around on it for some reason
<calebccff> right, this one was the same
<calebccff> trying to figure out a good installation process for pmOS
<steev> well, the qseecom that is in 6.7 doesn't seem to like the c630, but i haven't looked into why at all
<calebccff> ohno :(
<calebccff> have you had any luck with sensors?
<enyalios> nerdboy: echo 'force_drivers+=" nvme phy_qcom_qmp_pcie pcie_qcom i2c_hid_of i2c_qcom_geni leds_qcom_lpg pwm_bl qrtr pmic_glink_altmode gpio_sbu_mux phy_qcom_qmp_combo panel-edp msm phy_qcom_edp "' > /etc/dracut.conf.d/20_thinkpad-x13s.conf && dracut --kver $VERSION -m "kernel-modules rootfs-block base"
<nerdboy> got any corrections for the wiki page?
Mathew has joined #aarch64-laptops
mcbridematt has quit [Read error: Connection reset by peer]
<enyalios> yeah i have to post those soon, i have stuff to make vid accel work
<enyalios> and how i made audio work
<enyalios> i couldnt get it working with just plain alsa, but set it up so things work with pipewire
<enyalios> ill try to get those put in there soon
user__ has joined #aarch64-laptops
jglathe_ has joined #aarch64-laptops
user__ has quit [Remote host closed the connection]
<jglathe_> got the MACs stable on my x13 with bootmac, finally https://drive.google.com/file/d/1x4dMlGo-gxzWo2TjenKDpG3PRGlXtWhX/view?usp=sharing
<jglathe_> there is a small time window after boot where this works without flickering in the GUI -2..4s
<jglathe_> @reboot sleep 3 && /usr/bin/bootmac -b -W wlP6p1s0 -p F4A80D
<calebccff> how does bootmac cause the gui to flicker?!
<jglathe_> if down/up operations apparrently, looks weird. Same happens when BT connection to mouse fails, like when out of range
<jglathe_> its only the icons in the status bar, but enough to intrigue/annoy me
<jglathe_> this thing _is_ fast, so... better find out
<nerdboy> enyalios: are you using the overlay for pd-mapper/qrtr or something else?
<jglathe_> overlay?
<nerdboy> portage overlay
<jglathe_> using pd-mapper and qrtr ofc, started as systemd services
<jglathe_> doubt it, this is plain boring ubuntu
<jglathe> seen it's not me, sorry
<enyalios> yeah, using the pd-mapper from your overlay
<enyalios> nerdboy: ^
<nerdboy> there are commits after the latest 1.0 tag so i pushed some live ebuilds yesterday
jglathe_ has quit [Remote host closed the connection]
jglathe_ has joined #aarch64-laptops
jglathe_ has quit [Remote host closed the connection]
jglathe_ has joined #aarch64-laptops
jglathe_ has quit [Remote host closed the connection]
<jenneron[m]> bamse: steev: i can't reproduce the touchpad issue when i2c is 10 kHz
<jenneron[m]> seems fine, but i will be testing it for a week or several days to make sure
<jenneron[m]> also it might be placebo, but touchpad feels more responsive than before, less input lag
<jenneron[m]> maybe steev wants to try it, i'm using 6.1.69 because mainline still has UFS issues https://gitlab.com/sc8180x-mainline/linux/-/commits/v6.1.x/?ref_type=heads
<steev> i'll take a look at it on mine a bit later, it's football day
<jenneron[m]> if it passes my daily-usage-testing i will upstream it along with touchscreen
Mathew has quit [Quit: Leaving]
mcbridematt has joined #aarch64-laptops
<jenneron[m]> <steev> "jenneron: well, fun fact, the c6..." <- interesting, we probably have the same issue (you can check dmesg on lenovo flex 5g)
<steev> jenneron[m]: which ufs issues are you seeing? because i am booted from the ufs?
<jenneron[m]> steev: run gnome-disks benchmark with these settings
<nerdboy> well i'll be a dirty...
<jenneron[m]> (screenshot)
<steev> jenneron[m]: right, i did that before and it was fine?
<nerdboy> all 3 of those things work fine with big fat genkernel initramfs
<jenneron[m]> steev: yes, the same for me, but i increased the amount of samples and it sometimes fails now
<nerdboy> including j5 dock/display
<steev> jenneron[m]: hm, okay i'll see here by increasing size as well
<jenneron[m]> steev: also, i may be going nuts, but it seems that with https://gitlab.com/sc8180x-mainline/linux/-/commit/a1aa78804b81cb14f2cc580340d518477b73aa9d applied UFS issue appears more
<jenneron[m]> i will check things more later, since it's 1 am here
<jenneron[m]> but i doubt
<steev> hm, i thought i was already applying... oohhh it's only primus in konradybcio's patches
<jenneron[m]> yep
<jenneron[m]> will send this one upstream as well
<steev> i thought you were going to bed
<jenneron[m]> not right now, lol
<jenneron[m]> i noticed this stuff missing, and i did the patch right now because i can forget about it, but i'm not sending it right now
<jenneron[m]> i also want to test touchpad a bit before upstreaming, but it looks promising
Caterpillar has quit [Quit: Konversation terminated!]
<steev> i definitely know how that is... the wakeup sources i actually wrote like 2 years ago for the c630 just forgot to send
mcbridematt has quit [Ping timeout: 480 seconds]