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)
<szclsya[m]> as far as it's known and a fix is on the way I'm happy
<steev> it's known, dunno if a fix is on the way
<clover[m]> :^)
<steev> ayy, congrats
<szclsya[m]> ayy the efivars is working
<clover[m]> It booted!!
<szclsya[m]> hmm the cpu scaling seems to work just fine. I wonder what's using so many energy
<clover[m]> i have no wlan0 on my install hmm
<clover[m]> x13s-firmware package is installed
<szclsya[m]> check if hw2.1 is a symlink in /lib/firmware/ath11k/WCN6855/?
<szclsya[m]> hmm that looks alright
<szclsya[m]> check dmesg | grep ath11k ?
<szclsya[m]> there you go, it is renamed for some reason
<clover[m]> thats my ethernet
<szclsya[m]> the wlP6p1s0 one?
<clover[m]> nvm its working
<clover[m]> my bad
<szclsya[m]> phew, I thought I fscked up the firmware package
<clover[m]> nope just me being a noob :)
<clover[m]> now the big question, gnome or plasma
<szclsya[m]> sway :)
<szclsya[m]> sway's software renderer doesn't like fractional scaling tough. have to use 2x scaling and everything is super big
<steev> szclsya[m]: well it's doing all software rendering via llvmpipe
<szclsya[m]> yeah, I'm surprised it's even usable
<clover[m]> I chose gnome sorry 😞
<steev> gnome works fine
<steev> well no
<steev> if arch has the patch, gnome should work fine
<clover[m]> Lol
<steev> the mutter patch otherwise if you plug in external display it'll crash
<clover[m]> Ok good to know
<clover[m]> No battery appearing
<clover[m]> No touchscreen either, and seems to hang when opening gnome-terminal
<clover[m]> Can't open any app it seems
<clover[m]> Some apps open. Some don't. Maybe they are trying to use gpu?
<szclsya[m]> it should still work since mesa will fall back to llvmpipe
<clover[m]> steev could this be related to mutter issue?
<clover[m]> overall, a productive evening. ttyl guys
<steev> no
<steev> clover[m]: do you have anything in /sys/class/power_supply ?
<clover[m]> steev what terminal emulator do you use?
<steev> okay the battery is there
<steev> i use alacritty myself
<steev> gnome terminal works fine here too though
<clover[m]> i must be missing some relevant power package maybe
<steev> upowerd running?
<steev> i woulda thought gnome would start it
<clover[m]> ooo service not found
<steev> service should be upower
<steev> upowerd is the executable
<leezu> hexdump0815: I've returned the Galaxy Book Go. Mainly as Lazor come with dts thanks to CrOS but I also found it has a better build quality and is cheaper. The only downside is 64GB emmc vs the 128GB ufs.
hexdump01 has joined #aarch64-laptops
hexdump0815 has quit [Ping timeout: 480 seconds]
derzahl has quit [Remote host closed the connection]
<clover[m]> yeah even with the service still shows 0% batter steev
<steev> And you have the firmware files in /lib/firmware/qcom/LENOVO/21BX/ ?
<steev> Can you pastebin your dmesg output?
richfree has joined #aarch64-laptops
<clover[m]> alacritty is weird on this, it doesnt refresh the last thing you typed until you type the next thing lol
<clover[m]> also hello from x13s linux :)
<szclsya[m]> yay
<steev> still need the other bits though
<steev> szclsya[m]: that's part of my "i'll do a -1" :)
<steev> or probably roll it in when rc2 arrives
<steev> idk if force pushing will break the tag
<clover[m]> wow wifi is dropping quite a bit 😅
<steev> yeah :/
<steev> hopefully kvalo gets us working stuff soon
<steev> doesn't look like anything has hit the linux wireless mailing list yet though
SallyAhaj has quit [Remote host closed the connection]
<steev> or, more importantly... his firmware repo
richfree has left #aarch64-laptops [Leaving]
<steev> as far as i can tell... it should be working - did you try that ruby script i pasted earlier?
<steev> bamse: https://lore.kernel.org/linux-input/20220819023924.3272-1-steev@kali.org/T/#u - i wasn't sure if i should cc you on it or not, but there's that
<mahmoudajawad[m]> <clover[m]> "also hello from x13s linux :)" <- I'm late to Party, clover. What distro are you on?
<bamse> steev: nice, i don't know much about the hid stuff, so i probably wouldn't be much help
<steev> bamse: it just gets rid of that hid device showing up under /sys/class/power_supply
chuckz has joined #aarch64-laptops
<steev> mahmoudajawad[m]: i believe they're using arch btw(tm)
<steev> bamse: also, just for giggles, https://github.com/steev/linux/commits/sc8280xp-next-20220818 i shoved this one out there which uses your mailing list stuff over what johan had, as well as pulling in srini's stuff
<bamse> steev: and it's not actually showing the battery of the stylus if you have one of those closeby?
<steev> bamse: i don't have a stylus at all (that i'm aware of)
<mahmoudajawad[m]> I see. I was assuming manjaro for a second and wanted to confirm. Great news, as I'm monitoring the development of Linux for x13s. I'm hoping to get one if basic functions become available for it.
<steev> mahmoudajawad[m]: i'm using debian stable on mine, and it's my daily driver
<steev> i basically yanked out the other stuff, put the mailing stuff in, and removed all the stuff that was applied and then reverted later
chuckz has left #aarch64-laptops [Leaving]
<steev> so i'm assuming ci is gonna mail me a few times yelling :P
<mahmoudajawad[m]> steev: beside gpu, what doesn't work for you?
<steev> audio, gpu
<steev> wifi is a hack, but we're at the mercy of qualcomm there
<mahmoudajawad[m]> I see. I was under assumption regular distros still don't go beyond boot screen on x13s. I was wrong. Maybe I should consider getting one already.
<bamse> bt
<steev> oh right, no bluetooth
<steev> bluetooth, audio, gpu, but gpu and audio are close, with audio being closest
<mahmoudajawad[m]> I see.
<mahmoudajawad[m]> The device has sound jack, right? I can live without bt then.
<bamse> steev: audio is up and running, so that's a matter of getting it upstream
<bamse> steev: gpu is not there
<szclsya[m]> mahmoudajawad: if you can do vanilla arch I have a repo with kernel and firmware for x13s https://github.com/szclsya/x13s-alarm
<steev> bamse: that's fair, but it feels close :(
<bamse> yeah, that's the annoying part
<mahmoudajawad[m]> Leo Shen: thanks for sharing. I sure will keep it in mind to test it once I get the device.
<bamse> probably just some detail we're missing somewhere
<steev> maybe it's the const missing
<steev> i know it's not, but i was going over the reviews of your stuff as i was applying them
<bamse> i think it's something that's not clocked, reset or powered...
<travmurav[m]> hexdump0815: The pictures are rather bad unluckily, but one obvious thing is that Samsung decided to _not_ use the qsip module
<travmurav[m]> I suppose they decided to leverage their mobile team
<travmurav[m]> hexdump0815: I would still mostly expect the qsip related power connections to be the same but it's not guaranteed
<steev> bamse: also... earlier i had an adsp in qcom_stats but i can't, for the life of me, seem to get it again
<bamse> steev: what if you unbind/bind qcom_stat once the adsp is up?
<steev> heh, i was just doing that
<steev> oh, now i have more
<steev> adsp aosd cdsp cxsd modem
<bamse> as i replied in the thread, we should be able to register for notifications and add those dynamically
<steev> yeah, i saw that
<steev> i believe abelvesa sent something towards that end, but noticed a bug after submitting, so waiting for him to resend that set
SallyAhaj has joined #aarch64-laptops
SallyAhaj has quit [Remote host closed the connection]
<steev> https://paste.debian.net/1250957/ i have no idea if those numbers look good or not
<steev> modem makes sense because i'm not using it at all
<bamse> i guess it says that everyone else went to sleep...but you
SallyAhaj has joined #aarch64-laptops
<bamse> and from the tips and tricks corner... you can do "grep '' *" instead of that look
<steev> the battery performance on this thing is still ridiculous
<bamse> s/look/loop
<steev> oh, nice, i like that output better to
<bamse> e.g. grep '' /sys/class/power_supply/*/uevent
<bamse> yeah, it's easier to write and the output is much easier to read :)
<steev> agree there
<steev> bamse: i just noticed, you pushed the EC nodes for dts to the list, but i don't see the actual EC driver pushed to the list for c630?
<bamse> steev: i was under the impression that i didn't push the dt nodes to the list(?)
<bamse> nice, i thought that was still on my todo list...
<steev> ohh, it didn't go to msm
<bamse> cool
<steev> got it
<bamse> on 8cx i pushed the driver code, but none of the dts pieces (iirc)
<steev> got ya
<steev> right
<bamse> have received some feedback, so i'll respin that tomorrow and might push out some of the dts as well
<steev> can we not query the actual battery info on the c630?
<steev> i mean, model/make, not the other info
<bamse> it's hard coded like that in the acpi dsdt
<steev> oh, interesting
<bamse> so perhaps it should have been omitted instead
<bamse> let's see if the maintainer has an opinion
hexdump0815 has joined #aarch64-laptops
hexdump01 has quit [Ping timeout: 480 seconds]
klardotsh has joined #aarch64-laptops
systwi has quit [Ping timeout: 480 seconds]
systwi has joined #aarch64-laptops
Guest479 is now known as Penguinpee
<HdkR> https://cdn.discordapp.com/attachments/765304672579092511/1010113819160350761/Screenshot_2022-08-19_02-10-33.png Confirmed that it is RAM problems causing my HDK888 to crash games. Destroyed steamwebhelper to free up a couple gigs of RAM and it played through a full match.
<HdkR> Really need that X13s ram capacity :)
hexdump0815 has quit [Ping timeout: 480 seconds]
flowriser has quit [Quit: Konversation terminated!]
hexdump0815 has joined #aarch64-laptops
hexdump01 has joined #aarch64-laptops
hexdump0815 has quit [Ping timeout: 480 seconds]
Lucanis has quit [Ping timeout: 480 seconds]
<hexdump01> jenneron[m]: travmurav[m]: i tried the mentioned values for the i2c kbd but nothing changes - i'm slowly coming to the conclusion that without some docs, schematics or sources it might be quite hard to move forward with the galaxy book go if one is not an re expert like the asahi experts :)
hexdump01 has quit [Quit: WeeChat 1.9.1]
hexdump0815 has joined #aarch64-laptops
<travmurav[m]> hexdump01: I think acpi has i2c lines off by 1 according to the device addresses in acpi vs sc7180.dtsi
<travmurav[m]> but mine had completely different line in dsdt compared to the reality
<travmurav[m]> so worst case you could make 12 keyboard nodes and see which one probes :P
<travmurav[m]> but yes, without schematics the task is significantly harder IMO
<hexdump0815> travmurav[m]: i tried i2c7 and i2c6
<travmurav[m]> theoretically the dsdt, w10 driver files and preferably high resolution photos of the mainboard would be enough to advance pretty far
<hexdump0815> jenneron[m]: where did you have those reg and irq values from which you suggested to me?
<travmurav[m]> (with the latter serving as a reference of what components exist)
<travmurav[m]> hexdump0815: I think acpi spec defines how the function is called and what are the arguments, returns
<clover[m]> <mahmoudajawad[m]> "I see. I was assuming manjaro..." <- I'm considering adding x13s as a community device to manjaro arm. Their community repository has a lot of useful stuff.
Lucanis has joined #aarch64-laptops
<clover[m]> ok kitty is working much better than alacritty on gnome so far
<clover[m]> <steev> "And you have the firmware..." <- [alex@naga ~]$ ls -l /lib/firmware/qcom/LENOVO/21BX/... (full message at https://matrix.org/_matrix/media/r0/download/matrix.org/TSnJZNHVxdmXYspXZHkCwQGZ)
<steev> i've no idea; maybe it's something with newer gnome
<steev> i'm using debian stable which is 3.38 and it works great here
<steev> well, for certain definitions of great
<clover[m]> [alex@naga ~]$ ./battery
<clover[m]> ./battery:26:in `to_i': NaN (FloatDomainError)
<clover[m]> from ./battery:26:in `<main>'
<clover[m]> from the ruby script
<steev> it's pointing at ENERGY_NOW and ENERGY_FULL in yours?
<steev> it's gonna be annoying if yours points at CHARGE_NOW and CHARGE_FULL unlike... everyone else that gets 0's out of them
<clover[m]> meetings, i will have to check later
<steev> or give me the output of grep '' /sys/class/power_supply/qcom-battmgr-bat/uevent
<clover[m]> POWER_SUPPLY_NAME=qcom-battmgr-bat
<clover[m]> POWER_SUPPLY_TYPE=Battery
<bamse> is the adsp running and is pd-mapper in place?
<clover[m]> how can i tell sir
<bamse> what do you get if you run "qrtr-lookup" any reported services on Node 5?
pierro78 has joined #aarch64-laptops
<clover[m]> um, what
<bamse> the tool qrtr-lookup will list "qmi" services in the system...node address 5 is the entity that runs the battery firmware, so a quick check to see if the remoteproc is up is to run that and check for any services registered from that node
<clover[m]> does it come with this? https://aur.archlinux.org/packages/qrtr-git
<bamse> yes
<clover[m]> :^) k installing
Lucanis has quit [Ping timeout: 480 seconds]
<clover[m]> bamse:
<bamse> hmm, so the firmware is running
<bamse> do you have the pd-loader service running in Linux?
<clover[m]> Unit pd-loader.service could not be found.
<clover[m]> what package do i need to install?
<bamse> https://aur.archlinux.org/packages/pd-mapper-git that would probably do the trick
<bamse> and we need to get rid of that dependency somehow...
<clover[m]> should i start the pd-mapper service? i dont see a pd-loader service
<bamse> that's me multitasking
<bamse> yes, pd-mapper should be started :)
<clover[m]> my battery works now
<bamse> nice
<clover[m]> ty!
<clover[m]> documenting these dependencies now
<steev> silly arch users :P
<clover[m]> :O
<bamse> steev: i don't see you complain when we leave a trail of bread crumbs on the mailing list ;)
<clover[m]> steev: any idea on touchscreen?
<clover[m]> i have a touch screen, but it does not work on linux
<bamse> hmm, you're right...it's not working
<clover[m]> is this a regression?
<bamse> no, we only recently got hold of devices with touchscreen...and i haven't tried poking my screen before
<clover[m]> happy to report that the nipple just works, albeit scrolling is inverted
<bamse> clover[m]: perhaps the i2c address is wrong on the touchscreen...could you run "i2cdetect -y -r 1" (assuming that 990000.i2c is bus 1)
<clover[m]> Error: Could not open file `/dev/i2c-1' or `/dev/i2c/1': No such file or directory
<steev> clover[m]: modprobe i2c-dev
<bamse> modprobe i2c-dev
<bamse> :)
<steev> i don't have a touchscreen here :(
<steev> welll, not really a :( since i purposely bought it because it didn't
<bamse> nice, so there is someone at address $10
<bamse> anything written about it in the kernel log during boot?
<clover[m]> Will check
<clover[m]> https://pastebin.com/FU6RYhhZ <-- dmesg
<bamse> [ 6.399099] i2c_hid_of 1-0010: supply vddl not found, using dummy regulator
<bamse> looks good
<bamse> but then it doesn't say anything more, not sure why
<steev> maybe we don't have the touchscreen driver enabled in the kernel if it doesn't use a generic one?
<steev> hdkr said his worked, once upon a time
<clover[m]> yeah i vaguely remember that, that's why i figured it was a regression
<steev> very well could be
<steev> let me looksie
<steev> i need a break from autopkgtest anyway
<steev> configwise, 5.19->6 doesn't show any difference
<steev> at least, not one that would affect touchscreen
<bamse> clover[m]: grep gpio99 /sys/kernel/debug/gpio
<bamse> please
<clover[m]> nothing
<bamse> then you spelled it wrong, or something
<clover[m]> gpio99 : out high func0 16mA pull down
<clover[m]> i think it needed quotes
<bamse> ahh, i just spotted that johan has that output-high
<bamse> that's a reset# line for the touchscreen...so high is good
<bamse> although it's unlikely that i2cdetect would have detected anything if it was wrong
<clover[m]> ok so kernel driver is there, but maybe missing something in userspace?
<steev> it should report more in dmesg if it's there though
<bamse> the kernel should tell you something about what it found on 1-0010
<bamse> there's no driver bound to 1-0010 (on my device), so i2c_hid_core_probe() returns an error without telling us why
<clover[m]> is there a way i can call that function on mine?
<bamse> looking at the code, there's probably a "nothing at this address" printed if dev_dbg() got a saying
<bamse> it is being called, but it for some reason decided to fail
<clover[m]> ah ok
<clover[m]> putting that down as "not yet supported" on my list
<clover[m]> getting estimated 8:57 hours battery life on gnome at full capacity
<HdkR> steev: I haven't touched the laptop since testing the touchscreen, so it should still be working
Lucanis has joined #aarch64-laptops
<jenneron[m]> hexdump0815: i've got that info from acpi table
<jenneron[m]> i also work this way on galaxy book s, i don't have schematics
<jenneron[m]> btw, bamse , do you know what can cause this? https://dpaste.com/8WF3MSVHS
<jenneron[m]> or anyone else
<bamse> jenneron[m]: isn't just simply saying that you don't have a aux-bus defined in the edp controller node in dt?
<bamse> jenneron[m]: it should look like this:
<bamse> that said, i didn't think it was _required_...
<jenneron[m]> so i need to move panel binding there
<bamse> apparently yes
<bamse> i mean, it should be there...but apparently it's required to be there
<jenneron[m]> lenovo flex 5g in the same tree doesn't have it
<bamse> jenneron[m]: it's a fairly recent creation
<clover[m]> <HdkR> "steev: I haven't touched the..." <- weird that yours works and mine doesn't, i guess it's because we could have different panels?
<clover[m]> i figured out why gnome-terminal wasn't lanching. i forgot to set up my locale. this fixed a lot of things lmao
Lucanis0 has joined #aarch64-laptops
<steev> jenneron[m]: i have it somewhere, i think
<steev> for flex5g i mean
<steev> i need to go revisit it, i seem to have lost most of my work... somewhere
<steev> there it was
Lucanis has quit [Ping timeout: 480 seconds]
<HdkR> clover[m]: Maybe, I don't use touchscreens so I wouldn't care either way :)
<jenneron[m]> thanks, panel over drm works now
<jenneron[m]> one more thing, flex 5g has no-hpd binding, how do i know whether i need it or not?
<clover[m]> Leo does yours work?
<jenneron[m]> bamse: ^
<szclsya[m]> clover[m]: nope
<bamse> jenneron[m]: it should reflect if the hpd signal is connected or not in your device
<bamse> jenneron[m]: presumably though no-hpd should just work
<jenneron[m]> it works without no-hpd
<bamse> it might be that there's no difference in the qualcomm dp driver at the moment...
<jenneron[m]> panel-edp.c uses some additional delays for hdp on this panel, though i use "wrong" compatible which refers to another panel with same timings for now
<bamse> using the generic edp-panel compatible gives you the benefit that you only need to care about the timing of the power sequence...not the video timing
<jenneron[m]> does it parse timings from edid?
hexdump01 has joined #aarch64-laptops
<qzed> the video timings, yes
<qzed> the power sequence timings aren't part of edid
hexdump0815 has quit [Ping timeout: 480 seconds]
hexdump01 has quit [Ping timeout: 480 seconds]
hexdump01 has joined #aarch64-laptops
Lucanis has joined #aarch64-laptops
Lucanis0 has quit [Ping timeout: 480 seconds]
Penguinpee_ has joined #aarch64-laptops
Penguinpee is now known as Guest580
Penguinpee_ is now known as Penguinpee
<leezu> robclark: Do you know if CrOS verifies/uses the hw video encoder on lazor? It turns out that ffmpeg hw accelerated encoding works on the qcom rb5, but hangs on lazor. So I wonder if the issue may not be with ffmpeg. On lazor, triggering the encoding process increases the venus /proc/interrupts by 16, then ffmpeg hangs, and upon hard exiting after Ctrl+C, venus interrupts
<leezu> increase by another 8.
<robclark> we defn use both hw vid enc and dec..
<robclark> but I don't have something handy atm with combo of upstream kernel and CrOS userspace
<robclark> (but chromeos-5.15 kernel should be pretty close to upstream and video was working there)
Lucanis has quit [Ping timeout: 480 seconds]
Penguinpee has quit [Quit: Leaving]
Penguinpee has joined #aarch64-laptops
pierro78 has quit [Remote host closed the connection]
<harvestz[m]> is anyone using an external monitor with the x13s? I haven't kept up with all of the updates recently
Penguinpee has quit [Quit: Leaving]
<steev> harvestz[m]: yes, though if you're using gnome, you need to patch mutter and rebuild it
<steev> rebuild mutter, not all of gnome