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>
yes you have to use pulseaudio if you want working microphone, see the convo where johan yells at me for breaking people's systems :)
<_[m]123>
I'm sure he didn't intend it negatively
<steev>
oh no, i am saying that tongue in cheek
<steev>
i completely appreciate it, and he's absolutely correct.
<_[m]123>
mullvad still won't start lol but I can do wg up on p2p
<_[m]123>
still some ipv6 shit mm
<_[m]123>
steev glad to read 🙂
<Segfault[m]>
oh adamcstephens you got secure boot working? i tried that at one point but idk if i did it right because when it tried to load any payloads it'd just hang
<Segfault[m]>
does the tpm work with secure boot on or is that still not supported?
<adamcstephens>
I did get nixos+lanzaboote working for secure boot. unfortunately, tpm is still not supported
<Segfault[m]>
<jhovold> "how reproducable is the ffmpeg/..." <- just tried on 6.8-rc3 with your patches not including the shm ones, got the same venus hard reboot with ffmpeg
<Segfault[m]>
ffmpeg version was 6.1.1 and the command was `ffmpeg -i [input] -vf 'format=nv12' -c:v h264_v4l2m2m -c:a copy [output]`
KREYREN_oftc has joined #aarch64-laptops
hexdump01 has joined #aarch64-laptops
<Segfault[m]>
oh and i went through some old chat logs, windows suspend power draw was somewhere in the range of 100-150mW, and iirc it was lower on a clean install rather than using lenovos image
hexdump0815 has quit [Ping timeout: 480 seconds]
jglathe__ has quit [Remote host closed the connection]
jglathe__ has joined #aarch64-laptops
jglathe_volterra has quit [Remote host closed the connection]
jglathe_volterra has joined #aarch64-laptops
<steev>
Segfault[m]: do you have the venus firmware?
<steev>
it's something you have to pull from the windows partition, the filename starts with something like qcvss8280
jglathe_volterra has quit [Remote host closed the connection]
jglathe_volterra has joined #aarch64-laptops
<Segfault[m]>
<steev> "Segfault: do you have the..." <- i'm using the version in ironrobins x13s-alarm repo
<Segfault[m]>
i can extract the one from lenovos latest driver package if that'd help
<steev>
that *should* be fine
<Segfault[m]>
oh yeah the one i was using has a build date of november 2021 i should probably try the newer one
<Segfault[m]>
i should really stop relying on random repos for this stuff lol
<steev>
ehh, it's not really random, clover[m] maintains it
iivanov has joined #aarch64-laptops
<steev>
ideally, the mbn would be upstreamed though, yes
<Segfault[m]>
is there a reason it hasn't been yet?
jglathe_volterra has quit [Remote host closed the connection]
<Jasper[m]>
Burocracy
<jglathe__>
hah, the death of empires
todi has joined #aarch64-laptops
<steev>
i can't say why or why not, but... maybe someone hasn't begged them hard enough to?
<jhovold>
steev, Segfault[m]: I can assure you that I've been reminding Qualcomm about the missing venus firmware (as well as the bluetooth) one repeatedly through multiple channels over the course of the last year
<steev>
oh i was joking, i know how it goes when you ask a vendor to do stuff :) i wasn't trying to slight anyone
<jhovold>
i know, but it's quite frustrating still
<jhovold>
especially since they are very aware of this project and even appears to want to be associated with its success
<steev>
i completely understand :) when i was at genesi, we had to file bugs with freescale, who had to pass them on to amd, who had to pass them on to qualcomm whenever we dealt with the gpu drivers
<jhovold>
hopefully just big corporation bureaucracy
<steev>
legal has to approve everything ;)
<jhovold>
it took a year to get the wifi board file, but then there were actual technical details that needed to be sorted out
<jhovold>
for venus it seems all we need is a stamp of approval to release it
<jhovold>
bluetooth may or may not require something more
<jhovold>
adamcstephens: yes, I tested the efivars support quite a bit a the time too, and never ran into anything like the reboots you see
<jhovold>
and I'm sure qzed did too
<jhovold>
so not sure what changed, or what other conditions are needed to trigger this, hopefully there are some hints in that series (i only skimmed it long ago)
<jhovold>
but even if it's low probability that you hit it, it's really something we want fixed, don't want to have the installer crash on new users
<jhovold>
steev, clover[m]: have you tried playing around with the pipewire configuration to see if there's some easy way to get the microphone to work for pipewire users?
<steev>
i have not yet, no. i was planning to put the 2TB drive in first
<jhovold>
like the reduced quantum size setting that neobrain found that fixed playback?
<steev>
i do have that in
<jhovold>
i onlu confirmed that it did not help with the mic issue
<steev>
but have not looked further at it at all
<jhovold>
robot hiccup :)
<steev>
yeah heh
<steev>
that was the nice response
<steev>
most of them were wtf
<Segfault[m]>
steev: oh do you have any particular drive recommendations for these? i tried a 1TB SN740 and had some pretty serious issues
<steev>
Segfault[m]: i went with a SN770M+2242 extender based on HdkR saying it works good
<Segfault[m]>
hm ok
<jhovold>
sorry, that was exeat who found the pipewire playback config workaround
<steev>
but i have not yet put it in mine as it's the middle of the work week, but my timeframe almost moved up
<steev>
i started creating the windows recovery media on it, and went for a walk and came back to find it booted into it and asking some question
<steev>
i dislike that it does that if the device matches
<Segfault[m]>
damn the newer venus firmware didn't fix the reboot
KREYREN_ has joined #aarch64-laptops
<steev>
well, good news
<jhovold>
Segfault[m]: thanks for the ffmpeg reproducer, it crashes hard here too
KREYREN_oftc has quit [Remote host closed the connection]
<steev>
it reproduces here too
<Segfault[m]>
ah good :P
<steev>
i'm kinda surprised that obs hasn't caused it to crash here
<Segfault[m]>
i've applied the shm patchset, i'll see if i get any more efivars crashes
hightower3 has quit [Ping timeout: 480 seconds]
iivanov has quit [Remote host closed the connection]
<qzed>
yeah, I did test it quite a bit back then and never ran into anything like that... but to be fair, I never really stress-tested it (as in hammering it with EFI calls)
<qzed>
I can maybe try to reproduce it this weekend
<jhovold>
qzed: if there is a minimal fix that can be derived from that scm series, which won't be merged before 6.9, maybe can get something backported too
<jhovold>
iirc, you may be able open code that bufffer allocation which is now being abstracted away
<jhovold>
if that turns out to be the fix
<Segfault[m]>
well no more efivars crashes after leaving efibootmgr running in a loop for half an hour
<Segfault[m]>
seems fixed to me :)
Guest1075 has quit [Quit: WeeChat 4.2.1]
<jhovold>
Segfault[m]: thanks for confirming
jglathe_sdbox2 has joined #aarch64-laptops
<Segfault[m]>
oh yeah since it was mentioned i decided to try firefox, it does report supporting h264 decoding in hardware (although oddly no other codecs) but videos don't actually play, other codecs play fine in software
<Jasper[m]>
Youtube video's?
<jhovold>
ah, thanks for looking into that, I had a feeling something may be off there
<jhovold>
Segfault[m]: which version of firefox are on btw?
<Segfault[m]>
122.0, flatpak from the fedora repo (since it's not in flathub)
<jhovold>
thanks, i believe support for accelerated playback was enabled fairely recently
<jhovold>
using 122 here too
jglathe_sdbox2 has quit [Remote host closed the connection]
<Jasper[m]>
<Segfault[m]> "youtube, reddit, whatever else" <- using h264ify (on yt at least)
<Jasper[m]>
and is your audio working?
<Jasper[m]>
(or is there an audio device present that works)
jglathe_sdbox2 has joined #aarch64-laptops
jglathe_sdbox2 has quit [Remote host closed the connection]
jglathe_sdbox2 has joined #aarch64-laptops
<Segfault[m]>
<Jasper[m]> "using h264ify (on yt at least)" <- yep so i'm using enhanced h264ify for youtube, when i force videos to be h264 they give a playback error, when i let it use newer codecs they work fine but with software decode ofc
<Segfault[m]>
and yes my audio works fine, including with video playback in firefox
<Jasper[m]>
<Segfault[m]> "yep so i'm using enhanced..." <- Nothing about venus in dmesg then?
<Jasper[m]>
And, I can't remember exactly if they exist but, did you enable any settings within firefox to enable v4l2-m2m and video acceleration?
<Jasper[m]>
Oh and did getting the newer venus blob improve results at all?
<Segfault[m]>
nothing in dmesg, i didn't change any settings in firefox but hw accel seems to be enabled by default by the looks of things
<Segfault[m]>
and nah the newer venus firmware made no difference
<_[m]123>
what's it doing anyway?
<Jasper[m]>
<Segfault[m]> "nothing in dmesg, i didn't..." <- What does the HARDWARE_VIDEO_DECODING line say in about:support?
<Segfault[m]>
available
<Jasper[m]>
Just "default available"?
<Segfault[m]>
yep
<Segfault[m]>
and the table of codecs shows h264 as available in the hardware column
<Jasper[m]>
Running webrender?
<Segfault[m]>
yes
<Jasper[m]>
Wayland mode also on?
<Segfault[m]>
yes
<Jasper[m]>
Okay, I've ran out on things to try hahahaha
<Jasper[m]>
Checked fedora's thread on it too
<Jasper[m]>
It was all disabled by default until fairly recently
<janrinze>
jenneron[m]: Serious problems! pmbootstrap has mounted my rootfs as the destination!!! This is what happens with using tools that I don't understand..
steveej[m] has joined #aarch64-laptops
<jenneron[m]>
janrinze: do you mean on your PC? is it because of specifying wrong `--disk`/`--sdcard` or something else?
<travmurav[m]>
pretty sure there are multiple safeguards for this, if it did, it's a big problem...
<janrinze>
jenneron[m]: It was not clear to me that it required a --disk or --sdcard
<jenneron[m]>
these options are the same
<jenneron[m]>
i don't understand what's the problem
<janrinze>
my kernel is gone and i definitely should not reboot this machine until I know how to fix this!!!!!
<jenneron[m]>
did it overwrite your system?
<janrinze>
yup
<jenneron[m]>
what command did you run?
<janrinze>
just pmbootstrap
<jenneron[m]>
the only way to overwrite your system is to specify --disk or --sdcard pointing to storage you're booted from
<jenneron[m]>
otherwise it is probably something else..
<travmurav[m]>
and even then I think pmbootstrap would complain that the disk is >100GiB and would've refused to overwrite a mounted partition
<jenneron[m]>
janrinze: can you show the command you run?
<jenneron[m]>
check bash history
<jenneron[m]>
also send lsblk
<jenneron[m]>
from your PC
<janrinze>
/dev/nvme0n1p5 on / type ext4 (rw,relatime) -> this is my root disk
<jenneron[m]>
please show the command you run and upload lsblk, so it is clearer
<janrinze>
/dev/nvme0n1p5 on /home/janrinze/.local/var/pmbootstrap/chroot_native/var/cache/apk type ext4 (rw,relatime)
<janrinze>
/dev/nvme0n1p5 on /home/janrinze/.local/var/pmbootstrap/chroot_native/mnt/appstream-data type ext4 (rw,relatime)
<janrinze>
/dev/nvme0n1p5 on /home/janrinze/.local/var/pmbootstrap/chroot_native/var/cache/distfiles type ext4 (rw,relatime)
<janrinze>
/dev/nvme0n1p5 on /home/janrinze/.local/var/pmbootstrap/chroot_native/mnt/pmbootstrap/ccache type ext4 (rw,relatime)
<janrinze>
/dev/nvme0n1p5 on /home/janrinze/.local/var/pmbootstrap/chroot_native/mnt/pmbootstrap/git type ext4 (rw,relatime)
<janrinze>
/dev/nvme0n1p5 on /home/janrinze/.local/var/pmbootstrap/chroot_native/mnt/pmbootstrap/go type ext4 (rw,relatime)
<janrinze>
/dev/nvme0n1p5 on /home/janrinze/.local/var/pmbootstrap/chroot_native/mnt/pmbootstrap/rust type ext4 (rw,relatime)
<janrinze>
/dev/nvme0n1p5 on /home/janrinze/.local/var/pmbootstrap/chroot_native/mnt/pmbootstrap/abuild-config type ext4 (rw,relatime)
<janrinze>
/dev/nvme0n1p5 on /home/janrinze/.local/var/pmbootstrap/chroot_native/etc/apk/keys type ext4 (rw,relatime)
<janrinze>
/dev/nvme0n1p5 on /home/janrinze/.local/var/pmbootstrap/chroot_native/mnt/pmbootstrap/sccache type ext4 (rw,relatime)
<janrinze>
/dev/nvme0n1p5 on /home/janrinze/.local/var/pmbootstrap/chroot_native/mnt/pmbootstrap/netboot type ext4 (rw,relatime)
<janrinze>
/dev/nvme0n1p5 on /home/janrinze/.local/var/pmbootstrap/chroot_native/mnt/pmbootstrap/packages type ext4 (rw,relatime)
<jenneron[m]>
there is no problem here
<jenneron[m]>
if these things annoy you, you can run pmbootstrap shutdown
<janrinze>
Okay, it got me seriously worried..
<janrinze>
how do these mounts work?
<Jasper[m]>
pmbootstrap creates several chroots to do build stuff and to set up filesystems for the devices you're flashing
<janrinze>
travmurav[m]: yes I got the >100GiB warning but thought it was because the SDcard is 128GB..
<Jasper[m]>
optionally with qemu-user
<Jasper[m]>
janrinze: What pmbootstrap command did you run before you got that warning?
<janrinze>
Jasper[m]: my memory isn't perfect.. did a lot of pmbootstrap stuf..
<jenneron[m]>
janrinze: are you sure "kernel is gone"?
<Jasper[m]>
janrinze: It should be in your ~/.bash_history
<jenneron[m]>
is there anything else worrying after pmbootstrap shutdown?
<Jasper[m]>
Or if you use your arrow keys
<janrinze>
Jasper[m]: with many different consoles open bash history is sort of defunct..
<jenneron[m]>
janrinze: i don't know specifics, but pmbootstrap does some mounts in runtime between its work directory and chroots, so it can use e.g. built packages, ccache, etc. `pmbootstrap shutdown` gets rid of it, you can run it once you're done with working on pmbootstrap. however, you can keep it, it doesn't harm anything and will be gone after reboot
<_[m]123>
jhovold I see this when the exernal screen is not coming up
<travmurav[m]>
those are probably bind mounts and linux tools suck at showing them properly
<travmurav[m]>
so instead of nvme disk, it's just some dir on the fs in that partition
<jenneron[m]>
yeah most likely
<janrinze>
travmurav[m]: Yes, that's most likely
<janrinze>
jenneron[m]: I solved my mmap problem on Alpine. It appeared to be a mistake in the executable, not the kernel. Fixed it and ran my tests successfully.
<_[m]123>
how can I trigger the dpe?
<_[m]123>
* the dpe on x13s?
<janrinze>
Spoilers: the A78 (stock 2.6GHz) on the is 50% faster than the A76 (RPi5 @2.4GHz) for RISC OS :-D
<_[m]123>
I added this to my grub but lol, seeing the external screen also is on usb it probably prevents the screen from sleeping (I've noticed it staying on even on laptop lid close with the macbook but on linux you have more configs)
<_[m]123>
`usbcore.autosuspend=-`
<_[m]123>
s/=-`/=-1`/
<_[m]123>
I wanted to prevent hte nvme mounted externally disconnectingwhen the cable is slightly disconnecting when I move the laptop..
<_[m]123>
is there like sane kernel config flag list that people refer to btw?
<janrinze>
_[m]123: good reminder.. my rootfs is on USB, havent had a sleep yet because the lappie is connected to power. Not sure if it will wake up after sleep if i disconnect the power.
<janrinze>
_[m]123: LOL.. just checked. yes USB was disconnected and the OS is crumbling on the lappie.. good catch.
<janrinze>
jenneron[m]: do people repartition their chromeos and add a linux partition?
<jenneron[m]>
janrinze: people usually boot from USB and do something like `pmbootstrap insstall --disk /dev/mmcblk0`
<jenneron[m]>
i don't recommend installing it before we merge though
<jenneron[m]>
i'm going to open a merge request today or tomorrow anyway
<janrinze>
jenneron[m]: no worries. my installs are only testing for now.
<janrinze>
The chromeos disks are mtd0 ?
<Segfault[m]>
no that's usually spi flash of some kind
<janrinze>
Okay, so the internal disk is not 'seen' by the kernel.
<jenneron[m]>
janrinze: no, mtd0 is the only thing you should not overwrite
<janrinze>
jenneron[m]: noted!
<jenneron[m]>
but you can repartition eMMC
<janrinze>
eMMC does not show up as far as I can see.
<janrinze>
lsblk
<janrinze>
NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINTS
<janrinze>
├─sda2 8:2 1 256M 0 part /boot
<janrinze>
├─sda1 8:1 1 128M 0 part
<janrinze>
sda 8:0 1 14.8G 0 disk
<janrinze>
└─sda3 8:3 1 14.4G 0 part /
<janrinze>
mtdblock0 31:0 0 8M 0 disk
<janrinze>
nbd0 43:0 0 0B 0 disk
<janrinze>
nbd1 43:32 0 0B 0 disk
<janrinze>
nbd2 43:64 0 0B 0 disk
<janrinze>
nbd3 43:96 0 0B 0 disk
<janrinze>
nbd4 43:128 0 0B 0 disk
<janrinze>
nbd5 43:160 0 0B 0 disk
<janrinze>
nbd6 43:192 0 0B 0 disk
<janrinze>
nbd7 43:224 0 0B 0 disk
<janrinze>
zram0 253:0 0 1.9G 0 disk [SWAP]
<janrinze>
nbd8 43:256 0 0B 0 disk
<janrinze>
nbd9 43:288 0 0B 0 disk
<janrinze>
nbd10 43:320 0 0B 0 disk
<janrinze>
nbd11 43:352 0 0B 0 disk
<janrinze>
nbd12 43:384 0 0B 0 disk
<janrinze>
nbd13 43:416 0 0B 0 disk
<janrinze>
nbd14 43:448 0 0B 0 disk
<janrinze>
nbd15 43:480 0 0B 0 disk
<janrinze>
Oohh.. wow.. not what I expected but it seems the kernel has a tiny amount of knwoledge there..
<janrinze>
jenneron[m]: crossystem tools are nice. Did not have those when porting to older chromeos laptops.
<jenneron[m]>
janrinze: hun interesting
<jenneron[m]>
check dmesg for mmc errors
<_[m]123>
pls use pastebin for long outpus janrinze
<janrinze>
_[m]123: sorry, my bad.
<janrinze>
jenneron[m]: from chromeos the disk is nvme0n1 so it's not mmcblk0.
<janrinze>
different device. maybe UFS?
<jenneron[m]>
maybe we need some config options for nvme
<janrinze>
jenneron[m]: On chromeos the rootfs is mounted on /dev/dm-0
<jenneron[m]>
you said nvme0n1?
<jenneron[m]>
to be clear, my priority is to get things merged in the current state
<jenneron[m]>
then we can try getting internal storage to work
<janrinze>
jenneron[m]: by no means my tinkering should interfere with what you want to do.
<janrinze>
jenneron[m]: Also it's like 48 hours since I bought this lappie so there is a huge amount of learning to do for me.
<jhovold>
_[m]123: not sure what to make of that usb error you see, but I think it means you have a different issue, possibly your HDMI adapter not surviving suspend for some reason
<jhovold>
perhaps you can try a different adapter?
<jhovold>
for kernel config options, use the source or 'make menuconfig' and seach for the options you're want to find a description for
<_[m]123>
ok will try, accidentally ordered a second one anyay
<jhovold>
different brand/model, I assume?
<_[m]123>
> [ 4165.684224] usb 1-1: new full-speed USB device number 6 using xhci-hcd
<_[m]123>
[ 4165.840455] usb 1-1: not running at top speed; connect to a high speed hub
<_[m]123>
yeah, first one is "a brand¨ though
<_[m]123>
no screen now
<jhovold>
the second one doesn't work at all?
<_[m]123>
back to my other adapter, it took 5-10 seconds but it's working now
<_[m]123>
* working now again (but no resume this time ofc)
<jhovold>
which one is working?
<_[m]123>
the first "brand" one
<_[m]123>
which has the resume issue
<jhovold>
ok
<_[m]123>
should I get a cable without adapter?
<_[m]123>
I bought one too but it's too short -_-
<janrinze>
anyone here knows which alternative bootloader can be used for Aarch64 on a chromebook? dev_boot_altfw is an option on this lappie..
<jhovold>
does your monitor have usb-c input?
<jhovold>
otherwise you'll need an adapater (even if it may be built into the cable)
<_[m]123>
you have hdmi -> usb c cables
<_[m]123>
yeah
<jenneron[m]>
janrinze: there are no alternatives right now
<jenneron[m]>
for some older chromebooks it is possible to use u-boot
<janrinze>
just checked u-boot but it does not have any MT8195 support (yet)
<jglathe_sdbox2>
_[m]123 this USB-C oddity is normal I had to learn. When in alt mode (DP over USB-C) only 2 lanes are left for other data, and you get USB-2. Displays with a hub inside like my Iiyama XUB2792QSN do it this way
<jglathe_sdbox2>
the hub hw itself is USB-3 capable but you won't see it using it bc of the DP altmode setup
<janrinze>
jglathe_sdbox2: Good point. noted. I already forgot.
<jglathe_sdbox2>
same on Windows
<janrinze>
jglathe_sdbox2: swapped my dock with a different USB3 hub and connected HDMI with a USBC-HDMI cable now. That will improve disk and network performance on the lappie.
<jglathe__>
Did some experiments with a Lenovo TB4 dock, not very satisfying yet (or I'm too dumb to find the right driver)
<jglathe__>
other I/O is fast over it though, USB3
<janrinze>
my dock does USB-C power pass-through and as such leaves me an extra USB-C available.
<janrinze>
just need to remember that the USB port on the dock is essentially USB-2 so mouse/keyboard etc. are perfectly okay. Should not use it for the root disk :-D
<jglathe__>
Have a Targus PD dock on the X13s for Ubikey and other stuff. USB3, works nice. No video out, though (not tried yet, but USB-C directly on the X13s, as described.
<jenneron[m]>
janrinze: will you be able to test one more image?
<janrinze>
jenneron[m]: sure.
<janrinze>
jenneron[m]: had a reboot with 'Boot partition not found' again.. Interesting.
<jenneron[m]>
janrinze: it is possible that USB takes too long to initialize
<_[m]123>
but the fw was fckd and only upgradable on windows - can't remember but there were issue
<_[m]123>
s/issue/issues/
<janrinze>
jenneron[m]: can live with it (for now ;-) )
<janrinze>
Alpine for Aarch64 only supports Wayland?
<jenneron[m]>
janrinze: no
<jenneron[m]>
we have x11 UIs, you can select them in pmbootstrap init
<jenneron[m]>
but in my experience it doesn't work well with panfrost, at least on older GPUs, results in slower opengl, blinking cursor, etc, but it might be normal x11 experience, i'm not sure
<janrinze>
jenneron[m]: Tried 'setup-desktop xfce' which did some extensive package install. However Xorg can't start, checked the logs and such. lacks the proper driver for the modesetting.
<jenneron[m]>
it has some x11 workarounds for mt8183, maybe it will work on mt8195 as well
<janrinze>
Whoa! that worked
<jenneron[m]>
nice
<jenneron[m]>
building kernel for a new image
<janrinze>
jenneron[m]: OpenGL is faster in X11 than Wayland.
<jenneron[m]>
interesting
<jenneron[m]>
do you compare glmark2 and glmark2-wayland?
<jenneron[m]>
if you run glmark2 under wayland it will end up using xwayland, so use glmark2-wayland
<janrinze>
I use chromium browser with threejs examples
<jenneron[m]>
open chrome://gpu on wayland, make sure it uses wayland and not xwayland
<jenneron[m]>
glmark2 also may be a good thing to compare
<janrinze>
jenneron[m]: on Wayland FPS goes from 45-56 fps whereas on X11 the fps is steady 60.
<jenneron[m]>
if using glmark2, you should probably use glmark2 -b terrain and glmark2-wayland -b terrain
<janrinze>
my laptop just decided to drop the fps to the Wayland fps .. Browsers are unreliable for measurements ;-)
<Segfault[m]>
janrinze: check chrome://gpu while running under wayland, aside from possibly running in xwayland I've also seen chromium fail to use the gpu properly in wayland with some drivers
<Segfault[m]>
panfrost I had a number of issues with although I haven't tried recently
<janrinze>
Segfault[m]: yeah, chromium .. in the past I would tweak the chromium config manually with the flags and such.
<janrinze>
my preferred desktop is LXQT with openbox. Not too much flashy fancy things for me ;-)
<_[m]123>
ah there's a rather recent firmware upgrade for tb4
<jenneron[m]>
janrinze: you can look into sway or if you prefer stacking you can look into labwc, apparently they tried to make something similar to openbox but on wayland
<jenneron[m]>
also, if you want a custom UI setup, you can use "console" UI in pmbootstrap init, then you can install whatever is needed
<janrinze>
been using ARMLinux since 1996.. so yes me == dinosaur..
<jenneron[m]>
the FAQ is probably more for users of Android devices considering pmOS rather than chrome os, but still may be useful
<jenneron[m]>
*rather than for users of chrome os devices
<janrinze>
jenneron[m]: on my other chromebooks i was more of a 'roll-it-yourself' linux person..
<jenneron[m]>
i would like to make it available to less experienced users as well
<janrinze>
jenneron[m]: but pmOS does give a polished user experience out of the box. I can appreciate that.
<jenneron[m]>
unfortunately we lack graphical installer for these devices though, so users have to boot from USB/SD and install with pmbootstrap for now
<jenneron[m]>
when we have graphical installer it should become really nice
<janrinze>
jenneron[m]: yeah, however getting a bit of TCL/TK or python could do that job
<jenneron[m]>
so other distributions should be able to support it later as well
<jenneron[m]>
the point is to make installation process good on phones, and then add chrome os devices support too
<janrinze>
not sure if you need a new framework, it's an installer and should be able to do a few things such as setting up the SD card, fetching and copying stuff..
<janrinze>
jenneron[m]: It's great that there are people who build things like ondev2. I'm sure once that is finished people will enjoy using it.
<jenneron[m]>
janrinze: well there are some important criteria 1) it should be adaptive (for different screen sizes like phones), 2) it should work well with touchscreen 3) it should be able to overwrite itself. we want not only to support SD > eMMC installation, but also SD > SD
<janrinze>
jenneron[m]: NVME support in the kernel : https://wiki.gentoo.org/wiki/NVMe has a good readup. needs a few settings for the block devices etc.
<jenneron[m]>
janrinze: we can take a look at it when it's merged
<jenneron[m]>
i want that image to be tested and right after that i'm opening a MR
<jenneron[m]>
it just changes the way to package kernel and uses newer kernel version, so i want to make sure it still boots
<jenneron[m]>
so we probably just need to enable configs
<janrinze>
jenneron[m]: that's the part i sent about NVME support in the kernel : https://wiki.gentoo.org/wiki/NVMe has a good readup.
<janrinze>
jenneron[m]: I'll do this when i find time.
<jenneron[m]>
sure
<janrinze>
jenneron[m]: want to get back to my attempt at debian. I have disks that are already setup for other chromebooks. They all sort-of work the same once you have a disk image..
<janrinze>
jenneron[m]: P.S. really like the out-of-the-box result with the disk image.
<janrinze>
jenneron[m]: Thanks again for all your work!!
<adamcstephens>
do TB4 docks actually work with the x13s?
<_[m]123>
sure but not by tb4 🙂
<adamcstephens>
right
<jglathe__>
i/o worked, HDMI out didn't, power delivery did. Will do another round when I feel adventurous.
<jglathe__>
Also have a hama usb-c to hdmi adapter, works up to 2560x1440, above it does crazy things
<jglathe__>
_[m]123 yeah yesterday or so (?) I updated Windows on the X13s and it did the same here. Fetched all the sc8280xp firmwares, too, but need to reboot to find out
<adamcstephens>
I have a tb3 dock which obviously doesn’t work at all :)
<Jasper[m]>
<_[m]123> "is this for the tb4 gen2 dock?" <- The second, probably
<Jasper[m]>
It also does firmware updates
<Jasper[m]>
<adamcstephens> "do TB4 docks actually work..." <- Depends on the dock, one of my docks at home works fine as-is but has limited bandwidth on USB
<adamcstephens>
i'm not really worried about USB bandwidth. as long as I can get at least USB2 speeds that's fine. along with 4k@60 and ethernet of course :)
<jglathe_sdbox2>
I had to go to an x86 box for dock fw updates, nice if it works now
todi has quit []
<Jasper[m]>
<adamcstephens> "i'm not really worried about USB..." <- Also depends on how the dock handles it. I know mine doesn't do that very well :^)
<_[m]123>
how come that fw upgrade works on woa? it's emulating x86?
<Jasper[m]>
Probably, not sure what the limitations are with i/o there
<_[m]123>
this fw upgrade is still so fckng fckd, I've ran it 3 times already and still not sure if it worked once
jglathe_sdbox2 has quit [Remote host closed the connection]
<_[m]123>
switched to the other docking station now, probably wont'happen again
<steev>
ema: just went through your install instructions on the wiki, worked a charm
jglathe_sdbox2 has quit [Remote host closed the connection]
<_[m]123>
<steev> "to do what" <- I'm having some bugs on 6.7.3 and not on 6.8.0 but I want to get my vpn (mullvad) working - I figure the official debian trixie kernel config should be quite complete ?
<steev>
aside from new options, yeah, seems odd that mullvad wouldn't work with 6.7.3 though
<dgilmore>
I have my volterra booting Fedora rawhide from a USB stick by adding a dtb and using it to boot. I had to copy the gpu firmware from windows. using bootargs of "arm64.nopauth clk_ignore_unused pd_ignore_unused modprobe.blacklist=qcom_q6v5_pas"
<_[m]123>
<steev> "aside from new options, yeah..." <- ah no there it works but I have some graphical lag
<_[m]123>
your lap is never cold indeed with the laptop 😄
<janrinze>
jenneron[m]: I see you have google-snow and google-nyan too. My google-nyan is actually a nyan-big. But i really should try out pmOS on those :-D
<jenneron[m]>
janrinze: nyan may be quite broken in pmOS
<jenneron[m]>
GPU, audio, maybe wifi/bt are broken
<jenneron[m]>
i don't work on it because my device is in pretty bad shape, maybe i will continue once i get a new device, but i don't really want to spend on it
<janrinze>
jenneron[m]: Okay, I read some stuff that says that the GPU support in Debian was missing a few configs, those will be fixed next week.
<ema>
steev: not if you upgrade your kernel to the one in experimental :D
<jenneron[m]>
janrinze: nouveau is broken on tegra 124
<jenneron[m]>
maybe we can use it with nvk + zink when it gets better
<janrinze>
jenneron[m]: yes, but that seems to have been fixed recently too.
<jenneron[m]>
janrinze: snow is quite well supported though, one of USB ports doesn't work and GPU doesn't work
<janrinze>
maybe over the weekend I might fiddle with snow and nyan-big.
<agl>
_[m]123: wireguard works at me with Clients and Opnsense-Firewall.
hightower2 has quit [Ping timeout: 480 seconds]
hightower2 has joined #aarch64-laptops
hightower2 has quit [Remote host closed the connection]
<_[m]123>
early adoption is super fun
<_[m]123>
which kernel?
<_[m]123>
```
<_[m]123>
[ 83.629096] usb 3-1.1.1.2: 2:2: cannot get freq (v2/v3): err -32
<_[m]123>
[ 83.744849] hid-generic 0003:17EF:30CF.0005: device has no listeners, quitting
<_[m]123>
expensive docking != victory
<_[m]123>
wireguard works but missing kernel modules 😭
<_[m]123>
maybe I can boot the experimental one
<agl>
_[m]123: OPNsense has FreeBSD as OS and for the clients I have X86_64 and ARM64 machines.
<agl>
The clients are all Linux machines.
<agl>
For example the ThinkPad X13s
<_[m]123>
super random started building kernel and the external screen starts working
<_[m]123>
which kernel do you run on the client x13s?
<agl>
The newest from steev
<_[m]123>
which desktopmanager
<agl>
Cinnamon
<_[m]123>
I'm suspecting kde plasma needs some whatever acceleration
<_[m]123>
oh, so you're running X
<agl>
an Wireguird as manager fr the wireguard connections
<agl>
it's a gui
<_[m]123>
that's practical 🙂
<agl>
s(fr/for/
<_[m]123>
I"m gussing not on freebsd though
<_[m]123>
jsut really want to get this p2p wg working before I travel next week
<_[m]123>
so I don't ahve to take so much disks with me haha
<agl>
OPNsense-Firewall have a plugin for wireguard server
<_[m]123>
ah so it's in opnsense gui? nice
<agl>
no, wireguid is in the clients to connect to the wireguard server in OPNsense
<agl>
Under the OPNsense-Firewall was to do a lot to working the wireguard-server
hightower4 has joined #aarch64-laptops
<agl>
but there gives a good tutorial in the Dokumentation of OPNsense
<_[m]123>
so it does the key generation and all?
<_[m]123>
cool
<agl>
yes
<_[m]123>
and what you use the opnsense for? home network?
<agl>
Ist my firewall behind my cable-modem so that there can not to do Connections into my LAN from the Internet and outside from the LAN to the Internet. It have also an TFTP-server and some other what I need.
hightower3 has quit [Ping timeout: 480 seconds]
<agl>
s/Ist/It's/
<janrinze>
Wow.. learning about .its and .itb files. really interesting. I'm going to need to learn a bit more before I can boot my own kernel the hard way ;-)
<agl>
_[m]123: OPNsense handels also the connection to my virtual privat server at the domain in-ulm.de
<_[m]123>
yeah for now I just trust my home router and port forwarding to dedicated raspbi
<_[m]123>
might need to harden that a bit but it's not opnsense for sure
<_[m]123>
though don't need that complexityI think