ChanServ changed the topic of #aarch64-laptops to: Linux support for AArch64 Laptops (Asus NovaGo TP370QL - HP Envy x2 - Lenovo Mixx 630 - Lenovo Yoga C630)
<gwolf> ohai (-:
<gwolf> Please... who controls the linaro repositories we use for the aarch64-laptops bits?
<gwolf> I'm trying to get sound working again, and I found the version in Sid is quite newer than the Linaro one...
<gwolf> (wait, got a phone call)
<gwolf> Right... Well, alsa-ucm-conf is currently at 1.2.7.2-1, while Linaro's package is 1.2.5.1-1+linaro1
<bamse> qzed: thanks, it reports an "int rate"...which after above discussion seems to be measured in mWh
<gwolf> I have set it to be pinned, but it would be somewhat cleaner if its diff could be forward-ported...
<bamse> qzed: scaling it to uWh and presenting it as POWER_NOW seems like a reasonable thing to do
<gwolf> (while we are at it -- I remember I had working sound quite a bit ago, but the volume was too loud... and could not be modulated down. Now, it's the exact opposite: Even when setting pulseaudio to the maximum, 150%, it's barely audible
<gwolf> ... But FWIW, I do have one more specific question, now that I have started ;-) I have an external USB sound interface, so I can use my laptop for videoconferences.
<gwolf> It used to work simply and reliably... But since I switched to 5.19 + the patches for enabling the DP video out, I get the following:
<gwolf> has anybody seen anything like that?
<gwolf> FWIW, I do get a tree of devices properly created AFAICT under /sys/devices... but it seems input-remapper-control gets confused
<gwolf> I checked the relevant device does exist, but input-remapper rather finds:
<gwolf> "USB PnP Audio Device", "Lenovo-YOGA-C630-13Q50 Headset Jack", "hid-over-i2c 04F3:2800", "hid-over-i2c 04F3:30BC Mouse", "pm8941_pwrkey", "hid-over-i2c 6243:0001 Keyboard"
<gwolf> I don't know how to proceed from here :-\
<steev> gwolf: https://salsa.debian.org/steev/alsa-ucm-conf (yell at me if i didn't push the 1.2.7.2 bump)
<steev> oh, nevermind, i didn't
<bamse> we need to figure out how to get the ucm changes for c630 landed in the ucm project...they have been sitting there forever
* gwolf stops yelling halfway :-)
<steev> i just need to remember what system i actually have that on... hmm
<steev> gwolf: pushed
<steev> and yeah, getting them upstream would be nice, it only been like, a year sitting in upstream's pr
<gwolf> steev: pushed... to Linaro's repo?
<gwolf> Don't see it there yet
<steev> gwolf: to my salsa
<steev> i don't work at linaro :(
<gwolf> oh :-] OK, I'll rebuild from there
<gwolf> yes, that's why I started asking who handles that :-]
<gwolf> Anyway, I can at least ping Jordi Mallach, the Debian maintainer for it...
<steev> si, it's frustrating, but, i track it in salsa myself because... well, i like working audio (which doesn't work in 6.0 rc :( )
<gwolf> do you have a usable audio out level?
<steev> with pulseaudio, yes, with pipewire, no
<gwolf> ok, that's a data point... At some point I set up pipewire to be able to screenscrap or something
<HdkR> Ever since some of my devices updated to Pipewire it has broken the pulseaudio server service :(
<steev> HdkR: do you have pipewire-pulse installed?
<HdkR> yep
<steev> supposed to use that, i thought. i dunno, i just know that the pipewire transition has *not* been smooth
<gwolf> ok, uninstalling pipewire made it a bit better
<gwolf> yes, welcome back to the ugly sound quality :-)
<HdkR> On one device that I use pipewire with bluetooth on I've lost BT latency stats as well. Just isn't wired up :/
<steev> i really can't say because as mentioned, audio is borked again in 6.0 rc. it's something in soundwire so i'm assuming a patch will be inc
<steev> otoh, i have a patch applied already that is supposed to fix the breakage but it seems there is more, so i'll just wait
<gwolf> steev: I guess you didn't push all of your branches?
<gwolf> (I am missing your pristine-tar)
<steev> oh... i didn't do that properly :(
<steev> i just run debuild -uc -us -b in the directory
<gwolf> Ok, i'll do that :-)
<gwolf> Hey! We got headphones playback! That's good.. And mic! That's great!
<gwolf> (even if the quality is just what it is... It's advancing in the right direction!)
<steev> you can thank srinik for that :) it's all his work
<gwolf> srinik: Thanks! :-D
<gwolf> (still, audio levels both in and out are still way too low to be useful :-( )
<steev> here, with pulse audio, it's quite usable
<gwolf> I have no resource but to scratch my head even harder
<gwolf> Anyway, I ordered another external USB interface (they are so cheap I see no reason not to try!)
<gwolf> USB-audio, I mean
<gwolf> I just filed #1017424 FWIW
<gwolf> I know the Debian maintainers are likely to say "we only ship what upstream has blessed
<gwolf> but there is a chance they will push the diff to upstream
<gwolf> and there is a chance upstream will react favorably :-)
<steev> worst that can happen is they say no :)
<steev> and at some point, i should look into packaging pd-mapper, since that seems to be the only thing missing from the debianonmobile team
<bamse> steev: can i interest you in some freshly bakes patches?
<bamse> "make laptop_defconfig && make" and you should have a build with battery support and external display
<steev> oh heck yes
<bamse> if it says -11 something in dmesg when you connect the dp cable, turn it around
<steev> building now
<bamse> exciting
<steev> heck yeah
<bamse> build faster!
<clover[m]> oh my god
<bamse> did you see that Josh merged the x13s linux-firmware patch this morning btw?
<clover[m]> <bamse> ""make laptop_defconfig && make..." <- will it support 2 1440p displays?
<bamse> clover[m]: yes, definitely
<bamse> unfortunately we don't support 4k@60 just yet, but i'm driving my two 4k monitors at 30hz on the sc8180x today...
<bamse> there's some work in progress for 4k@60 though, so hopefully we'll see that soon
<clover[m]> ok, but 2k@60 it will work?
<bamse> yes, that should work
<clover[m]> nice
<clover[m]> Leo time to update lol
<bamse> i don't remember exactly where the threshold is where you need to go to 4-lanes or compression
<bamse> i think i had to go 4 lane to drive my 2560x1600 monitors...
<clover[m]> bamse: do you think a displaylink dock works?
<bamse> clover[m]: it should...so if nothing else i appreciate the testing :)
<steev> this is what i get for building on the box :(
<steev> oh and that's awesome that the x13s firmware went in
<bamse> sounds like we need that bwmon enabled as well ;)
<steev> i guess we also need to fixup the dts for the correct path now
<bamse> yes
hexdump01 has joined #aarch64-laptops
<szclsya[m]> wow
<szclsya[m]> does that mean gpu is also enabled?
hexdump0815 has quit [Ping timeout: 480 seconds]
<bamse> no :(
<steev> 72% battery
<bamse> i hope you're happy with that estimate?
<steev> hm
<steev> plugging in the pinephone usbc thing made it angry
<szclsya[m]> welp
<bamse> steev: computer says no?
<steev> that was plugging it in to the port closest to the back of the laptop
<bamse> i tried my tv in both...
<steev> gonna try one thats a bit more polished
<bamse> but did it tell you anything useful?
<steev> i wasn't ssh'd in, i'm doing that now
<steev> if it will let me
<steev> wifi started acting up earlier and my normal network kept just disconnecting or saying it was connected but no traffic
<steev> okay got -11 with the apple adapter, flipped it - it still makes gnome unhappy, and i end up with a cursor in the upper left, kinda like the c630 did... buuuuut, the second monitor also shows the cursor in the upper left, so it's definitely working
<steev> gonna try rebooting with it plugged in
<steev> didn't help, but i do think this is a gnome issue
<bamse> interesting
<bamse> time to get some sleep here though, will do some more testing tomorrow
<steev> do it
<steev> this is awesome
<steev> yeah, it's definitely gnome crashing, because the extensions all got disabled
<steev> bamse: for tomorrow - https://paste.debian.net/1250568/ - for some reason... it's just refusing to give me video out now
<steev> and i just checked in windows, and the battery percentage, at least, is spot on
<HdkR> ooo, x13s in firmware repo? Good job getting that through so quick, holy crap
alpernebbi has quit [Quit: alpernebbi]
alpernebbi has joined #aarch64-laptops
alpernebbi has quit [Quit: alpernebbi]
CosmicPenguin_ has left #aarch64-laptops [#aarch64-laptops]
CosmicPenguin has joined #aarch64-laptops
<bamse> steev: that's with laptop_defconfig?
<bamse> steev: we don't handle probe deferal in the dp driver gracefully, not sure if we fail to clean up the irq or if there's an issue with how we handle the irqs...
<steev> bamse: it was, but it doesn't seem to happen every boot
<steev> sorry - by video out, i meant display, i've slept now
<steev> that was booting it without any usb-c plugged in
<bamse> steev: rigth, there's some timing thing which causes the dp driver to sometimes probe defer, in which case it goes belly up on the second attempt
<steev> ah, got ya
<steev> i think qzed added something for dealing with that on the spx
<qzed> steev: you mean me reverting a bunch of stuff? not really ideal, but it kinda works... still need to check on 6.0-rc1 though
<qzed> ah, right
<steev> i might try that at some point, but gotta get work done :)
<qzed> although I wonder if I need to change some of the panel delays to be a bit longer
<bamse> qzed: making sure the panel_edp is builtin is another good way to trick it into submission, but this need to be corrected...it's not optimal that it has a tendency to miss the panel, but it should survive it
<qzed> yeah, I do that already, but it still failed sporadically
<steev> 10 hours 44 minutes is the estimate at full :)
<bamse> steev: the estimate in gnome seems trustworthy?
<steev> well you say we get about 10 hours currently, so does seem so
<bamse> steev: in i3status it jumps between 30 minutes and 24 hours, rather sporadically...hence my conclusion that it seems like we're not in an agreement on how to interpret the power supply abi
<steev> we use upower
<steev> in gnome
<bamse> where you able to try the external display again btw?
<steev> let me try without gdm running
<steev> without gnome/gdm3 running, it works great
<bamse> strange, why would it not work with gnome?
<steev> i do not know, because it does on the c630 with gnome
<steev> as soon as i start gdm3... it locks up and the screen just becomes the cursor
<bamse> too bad that it's what we get with the installer...
<steev> heh
<bamse> it's one thing to try to convince you to shift to a proper desktop environment... ;)
<steev> i'd say we could always switch to xfce but the default debian xfce is.... not very pretty
<steev> it does *recognize* that i'm unplugging the usb, because the screen blinks
<steev> so it's not locked up, i guess, but i can't ssh in anymore
<bamse> you can ssh i before and cat /proc/kmsg while doing that?
<steev> gonna do that now
<steev> well
<steev> that's unhelpful
<steev> <3>[ 54.327707] [drm:dp_panel_read_sink_caps [msm]] *ERROR* no downstream ports connected
<steev> that's the only error, the rest is just usb kicking in
<steev> oh, it didn't lock up
<steev> let me see what gnome is doing
<clover[m]> KDE plasma comes with the default installer for asahi and it works well
<bamse> clover[m]: any userspace should work, simply trying to figure out what interaction it is that doesn't work...
<steev> weird
<bamse> e.g. fbcon breaks when i connect/disconnect my two external monitors a few times
<steev> the only thing in the journal is... pulseaudio complaining that there was no reply
<bamse> steev: what's the visual result of the exercise? anything on the monitor? does edp still work?
<bamse> for more useful debugging of DP you poke 256 into /sys/module/drm/parameters/debug btw
<steev> lcd of the thinkpad shows a blinking cursor in upper left corner. likewise on the monitor
<steev> it's just weird because, gnome has absolutely crashed, but according to the logs... it hasn't
<steev> i hit ctrl+alt+fn+fX a few times, and at some point, i did get a login prompt again
<bamse> that's a console cursor, a graphical cursor?
<steev> console cursor
<steev> [ 396.371086] msm_dpu ae01000.mdp: [drm:dp_bridge_detect [msm]] is_connected = false
<steev> [ 396.371129] msm_dpu ae01000.mdp: [drm:dp_bridge_hpd_notify [msm]] DP0: status: 2 connected: 0 is_connected: 0 connector: hpd 2
<steev> [ 396.371199] msm_dpu ae01000.mdp: [drm:dp_bridge_hpd_notify [msm]] DP1: status: 1 connected: 0 is_connected: 1 connector: hpd 1
<steev> [ 396.371178] msm_dpu ae01000.mdp: [drm:dp_bridge_detect [msm]] is_connected = true
Penguinpee_ has joined #aarch64-laptops
<steev> unplugged the usb-c adapter and
<bamse> looks like test-ok to me ;)
<steev> yeah...
<steev> and plugging it back in
<steev> this may be stupid, but numbers scare me, is it doing something like trying to set the laptop display to the same resolution (2K) as the monitor?
Penguinpee is now known as Guest223
Penguinpee_ is now known as Penguinpee
<bamse> because it's trying to mirror the displays?
<steev> yeah
<steev> hm, no, it literally says 1920x1200 in the output
<bamse> steev: do we have a msbtfw21.mbn somewhere?
<steev> assuming that's supposed to be the bluetooth firmware, no
<steev> we have hpbtfwX0.bXX
<bamse> and windows has hpbtfw21.tlv
<bamse> pointing the driver at that did not result in a stellar outcome
<steev> maybe try the
<steev> i have a few tlvs here
<steev> 10, 20, and 21
<bamse> same here
<bamse> i do get a lot of Frame reassembly failed (-84) in the logs
<bamse> wonder if that means that my serial bus isn't up to snuff
<bamse> nah, seems to receive data
alpernebbi has joined #aarch64-laptops
<steev> bamse: brought back from suspend and 13 hours :D
Penguinpee has quit [Quit: Leaving]
Penguinpee has joined #aarch64-laptops
<travmurav[m]> steev: if you were wondering, I think I found what causes gnome to crash when an external display is plugged. It doesn't know that the port can be called "USB"
<travmurav[m]> Sadly haven't found anything for my issue
<steev> hmm
<steev> good point
<travmurav[m]> they have a big enum that mirrors the kernel but was not updated in a long time https://gitlab.gnome.org/GNOME/mutter/-/blob/main/src/backends/meta-output.h#L62
Penguinpee has quit [Quit: Leaving]
Guest223 is now known as Penguinpee
<steev> gwolf: here's a question i've always had... what's a good way to deal with installing and uninstalling build-dep
<steev> build deps for mutter are ridiculous, and i don't want to pull them all in and actually leave them installed
<steev> travmurav[m]: ayy
<travmurav[m]> was it indeed the issue of the crash?
<steev> indeed it was
<travmurav[m]> Cool
<steev> https://paste.debian.net/1250650 is the patch i wrote and applied (basically just took what you said about it not being synced in a while)
<travmurav[m]> I couldn't try it myself as my alt mode adapter decided to die this morning (about after I found that while looking at the logs) and I spent the whole day trying to fix it :/
<steev> oh no :(
<steev> bamse: with mutter being fixed, i can say, it 100% works :D
<bamse> steev: what needed to be fixed?
<steev> mutter
<steev> it doesn't know about usb type monitors
<bamse> ahh, read instead of skimmed through those few messages
<bamse> nice, yeah i do think i'm supposed to report USB as DRM type
<bamse> nice!
<steev> now i need to fix mutter on the c630
<bamse> so debian will be broken for the next 14 years?
<steev> pretty much, unless we convince them to backport the patch, which i guess i should submit
<bamse> i approve of said patch :)
<bamse> steev: thanks for the testing
<bamse> now if i only could figure out how to get bt working i think i can drop in the x13s for my desktop
<bamse> alternatively i need to find a bt/usb dongle...
<steev> bamse: is there a specific reason we don't have wakeup-source for the c630's keyboard/touchpad?
<steev> i've no problem submitting, but if it's something like, keyboard is on usb bus, and we disable usb bus during suspend....
<gwolf> steev: Right... There might be a decent way for dealing with build-dep installing and removing
<steev> gwolf: oh, found one - mk-buiddep-deb or something like that
<gwolf> I just take note of the packages and purge afterwards
<steev> it builds a fake package and you can purge that and it removes them
<steev> alas, the thinkpad still has all of them, so that's gonna be a pita
<gwolf> Oh, interesting!
<gwolf> mk-build-deps, nice
<steev> gonna test that this fixes things on debian testing as well before i go about submitting upstream
<bamse> steev: we where years away from suspending when i wrote that...patches are very welcome though :)
<steev> fair enough :) i wasn't sure if we left it off on purpose because usb bus or whatever :) will give it a test and if so, submit...
<steev> also would it be considered churn if i made the c630 look a bit more like the others, in that we do keyboard@ instead of tsel or whatever
<steev> i always have to look through wtf dmesg reports them as when i wanna know which one to do
<steev> i suppose maybe we should do touch screen for if someone is in convertible mode and it sleeps, i've not tested that
<steev> bamse: and yes, with mutter fixed on c630, i don't have the weird suspend crash either
<steev> also random fun fact, on the c630, the apple adapter must be plugged in "right side up"
<steev> it still crashes gnome, out to gdm, but i'm gonna chalk that up to something else, probably in mutter
<steev> gwolf: also, a little one liner (obviously you'd change the date to whatever...) - for x in $(ls -1t /var/log/dpkg.log*); do zcat -f $x |tac |grep -e " install " -e " upgrade "; done |awk -F ":a" '{print $1 " :a" $2}' |column -t | grep 2022-08-16 | awk '{ print $4 }'
Votes78 has quit [Killed (NickServ (Too many failed password attempts.))]
Votes78 has joined #aarch64-laptops
<bamse> steev: sounds like a good plan
<steev> bamse: actually... i do still get the unbalanced irq on the c630
<steev> when bringing it back from suspend
<steev> not always though
<steev> will wait for msm-fixes to merge into rc1 (since that's what i'm based on) and see if it still is around
<gwolf> steev: Right, I have used similar one-liners with varying amounts of brittleness over time ;-)
gwolf has quit [Ping timeout: 480 seconds]
clararussell[m] has joined #aarch64-laptops
carlosstive[m] has joined #aarch64-laptops
billytoken[m] has joined #aarch64-laptops
<steev> oh ffs
<amstan> anyone going to LPC/ELCE in Dublin?
hutchinson70[m] has joined #aarch64-laptops
Donaldbtc[m] has joined #aarch64-laptops
<javierm> amstan: I'm going to LPC
Donaldbtc[m] has quit [autokilled: This host violated network policy. Mail support@oftc.net if you feel this is in error. (2022-08-16 23:51:39)]