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)
<bamse>
steev: cool, thanks
<steev>
still doing the c630 stuff too
<steev>
the wakeup sources do work though
<bamse>
i recall someone adding the sleep stat driver to 845, did that get picked up and does it indicate how deep you sleep?
<steev>
uhhh
<steev>
is it somewhere in sysfs?
<bamse>
drivers/soc/qcom/qcom_stats.c seems to throw them in debugfs/qcom_stat
<bamse>
and at least the tree i have doesn't have sdm845 defined, so then that's still on the list
<steev>
hm, i *thought* i had pulled that in to test already
<bamse>
need to read the discuss with Caleb one more time though
<steev>
based on b4's clean commits though, apparently i did not
<bamse>
but first i need to cook some dinner
<steev>
and i should enable qcom_stat
<steev>
bamse: it does show up but i'm getting 0's (aosd and cxsd)
<steev>
Maybe Iām missing some of abelvesa ās other patches that add the deep sleep
<HdkR>
I ran out of memory again nooo
<HdkR>
I can't find userspace bugs when I'm busy hardlocking my device when it runs out of ram :P
<steev>
oh, reading through the thread... maybe i should try stephen's patch
<steev>
oh, no, that'll break suspend
<bamse>
steev: it's expected that we're holding on to various resources which prevents the rpmh from taking us to the actual sleep states
<bamse>
HdkR: which device?
<HdkR>
bamse: HDK888
<steev>
well it works, in that it shows us 0's :D
<bamse>
steev: :(
<bamse>
well, a little bit :) ;)
<HdkR>
Just waiting for that X13s GPU to light up so I can use the 32GB of RAM there and ignore the HDK again :)
<steev>
hm
<steev>
i think i need to get better at lei
<steev>
this doesn't seem right
<steev>
lei q -o ~/Mail/overlay -I https://lore.kernel.org/all -t 'dfn:drivers/soundwire/qcom.c AND rt:1.month.ago..' is giving me a LOT more than just 1 month ago
<steev>
# 3133 written to /home/steev/Mail/overlay/ (3133 matches)
<steev>
pretty sure there haven't been 3000 patches to soundwire/qcom.c in the past month... but maybe
<bamse>
i was quite interested in lieer and notmuch as a way to search all my email
<bamse>
unfortunately lieer doesn't complete downloading my gmail box
<bamse>
get some error after fetching emails for 12 hours
<leezu>
Has anyone extracted a aarch64 libwidevinecdm.so for playing DRM protected content from CrOS?
<steev>
do they actually have a 64bit version?
<bamse>
sweet, i can report that x13s is happy to drive my two 4k monitors :)
<bamse>
still only 30hz though, need to fix that
<steev>
nice!
<clover[m]>
Coming from a pinebook I'm just excited for 16 GB of RAM
<clover[m]>
Nice bamse
<bamse>
and earlier today, i drove 15 miles home and realized that my ssh connection over lte/5g was still up :)
<bamse>
the future is here!
<clover[m]>
Wow
<clover[m]>
So modemmanager just workz?
<bamse>
yes, we needed a patch int he kernel to add the pci device, but that's landed already
<bamse>
and i need libqmi from the git branch...
<bamse>
there's a change in there which hasn't yet reached arch linux (last time i checked)
<bamse>
but with that in place, you just start the modemmanager and networkmanager services, poke nmcli (or presumably the gui) and you have internet
<steev>
so libqmi changes really
<bamse>
there's an updated fcc-unlock mechanism that was "recently" merged
<bamse>
nice, we got battery status on the flex 5g as well
<steev>
oh heck yeah
<steev>
so that can be submitted soon? or is video still flakey?
<steev>
okay, what are the points of the soundwire compatibles, am i blind, because they're not used anywhere? or is it something internal to the iommu
<steev>
wtf
<bamse>
have you seen any video flakiness?
<steev>
bamse: you were having problems getting video to come up
<bamse>
hmm, i thought that was you?
<steev>
i thought you'd said it wasn't working at all at one point... for me it *usually* comes up now, maybe 1 in 5 boots fails
<bamse>
well, i do have a problem on the flex 5g...i need to ssh in and manually bind the edp phy device to its driver
<steev>
ah right, that's the issue you had
<steev>
also i need to remember that dtsi files are a thing when looking
<steev>
[ 5.308786] qcom-soundwire wcd934x-soundwire.6.auto: Failed to get audio_cgcr reset required for soundwire-v1.6.0
<steev>
oh, yeah, it's not protected like the other one is
<steev>
will test fix and see if it works
<bamse>
steev: forgot to answer your question, i think it works fine, just need a dt binding
<bamse>
almost done with that...
<steev>
bamse: speaking of dt bindings.... sc8280xp lpass stuff was posted i see, are we gonna have audio soon?
<bamse>
ohh cool, i had missed that
<bamse>
srinik joined our weekly meeting last week using native audio on his x13s, so yes
<steev>
oh nice!
<bamse>
btw, i think the hpbtfw21.tlv firmware should work btw, but the bt driver is unhappy about the hci messages coming in...need to find someone who can help me understand what's going on there
<steev>
hm
<steev>
now i get deferred device again
<steev>
i think i was too judicious
<bamse>
steev: there, now one "soon" has passed
<bamse>
and the battery driver etc is in the intertubes
<steev>
woop
<travmurav[m]>
steev: I don't really know the mutter codebase but I'd expect this to be some "device has appeared with this type" handler and in an old kernel a "USB" will never appear.
<steev>
travmurav[m]: yeah, that's what they think too, so, will see :)
<travmurav[m]>
tbh I was just trying to run the session without gdm and it said like "libmutter:<file>:<line>: unreachable code"
<bamse>
it was actually introduced in v5.13...
<travmurav[m]>
So I just opened the file and it was kinda obvious what's going on, since I remember being impressed the kernel named the port as USB after I added the connector-bridge sruff for my ec
<steev>
hmm
<steev>
so, i have a patch, and i get my device but i'm not getting audio
<bamse>
steev: speaking of audio and progress, vkoul has partial audio on flex 5g as well...
<steev>
oh nice!
<steev>
hm
<steev>
i wonder
<steev>
it's the same check twice so what if we just punt it
<steev>
bamse: is it okay to go over the 75 characters when it's in the commit message? checkpatch is pitching a fit because i'm mentioning 2 different commits and when i try to move them to a separate line, it just looks terrible
hexdump01 has joined #aarch64-laptops
hexdump0815 has quit [Ping timeout: 480 seconds]
<steev>
nevermind, i'm just dumb
<steev>
it was literally telling me to write 'commit <12+sha> "msg"' and i was reading it as commit '<12+sha> "msg"' so crisis averted
carlosstive[m] has quit [autokilled: Spambot. Mail support@oftc.net if you think this is in error. (2022-08-18 04:12:09)]
<szclsya[m]>
just got 6.0.0rc1 up and running
<szclsya[m]>
power supply report works, but (as expected) i3status-rust don't know how to handle the data format
<steev>
ah crud, can you tell it to use upower?
<steev>
it should be able to use upower, yeah
* travmurav[m]
wonders what's this special data format
<travmurav[m]>
does it report the battery via something other than a power_supply type object?
<steev>
i think it's because it looks for BAT or CMB
<szclsya[m]>
also charge_* stuff gives "No data available"
<travmurav[m]>
ah it assumes the name is relevant instead of checking the "links" to other objects
<steev>
the charge stuff does indeed
<travmurav[m]>
the i3status-rust that is
<steev>
i haven't looked to see the new things that bamse pushed just a bit ago
<travmurav[m]>
I should probably learn better how the device type detection works in i.e. upower as when I naively registered input device for the lid switch from my ec driver, upower was like "ah so it's your lid that has the battery in it, not the laptop"
<szclsya[m]>
btw acpi -b works and gives a reading
<szclsya[m]>
wpa3 still freaks out. wonder if the lenovo firmware drop contains some proper wifi firmware?
<steev>
no
<steev>
we're waiting on the board-2.bin from atheros/qualcomm (and the ath11k driver to understand the new firmware format)
Penguinpee has quit [Read error: Connection reset by peer]
Votes78 has joined #aarch64-laptops
_Votes78 has quit [Ping timeout: 480 seconds]
_Votes78 has joined #aarch64-laptops
Votes78 has quit [Ping timeout: 480 seconds]
Votes78 has joined #aarch64-laptops
zekra has joined #aarch64-laptops
<zekra>
By the name of Allah most great all able MALIK ALMOLOK almofadda - Allah is the only true God - Sunni Islam is the right religion - By the Greatest authority of Allah most great all capable
<zekra>
It is requested to leave fighting islam - leave harming muslim sunnis - not allter and divert islamic sources and books and religious heritage - leave harming poor people - don't revenge after reading these messages - don't make white/black people trouble to make wars between them - under consequences of falling down hard from Allah the most great all capable addayan
_Votes78 has quit [Ping timeout: 480 seconds]
zekra has left #aarch64-laptops [#aarch64-laptops]
neobrain has quit [Remote host closed the connection]
pierro78 has joined #aarch64-laptops
<steev>
clover[m]: does it have the wifi firmware in the right place?
<steev>
that shouldn't have changed
<clover[m]>
all i know is the old one boots and this one doesn't
<clover[m]>
trying to collect more info
<clover[m]>
maybe some kernel config changed to not support erofs
flowriser has joined #aarch64-laptops
<hexdump01>
travmurav[m]: do you have any pointer about how you extracted the dts info for your acer a1 from the acpi info?
hexdump01 has quit [Quit: WeeChat 1.9.1]
<travmurav[m]>
hexdump01: I mostly wrote it by hand, using acpi and, in bigger part, device schematics as a reference
hexdump0815 has joined #aarch64-laptops
<hexdump0815>
travmurav[m]: do you have any pointer about how you extracted the dts info for your acer a1 from the acpi info?
<hexdump0815>
background: i have a samsung galaxy book go here and would like to see if i can bring it a bit forward - sadly without too much low level know how
<hexdump0815>
i built a kernel from your tree and using your adcer a1 dtb as its as close as one can get for now as a woa sc7180 device and it at least boots
<hexdump0815>
but most things do not work and i think to get started the most relevant would be keyboard and usb
<travmurav[m]>
I'd guess the device would be somewhat similar to the trogdor chromebook too, but defenitely not the same and unlikely to match the acer/quanta design as well, tho I believe they're all based on the same reference
<hexdump0815>
i was thinking maybe a good starting point would be to try to get the kdb working to at least be able to interact with the device at all
<travmurav[m]>
the keyboard/trackpad is probably the easier thing to find in the dsdt
<hexdump0815>
how did you find and and map the info for your keyboard in your dsdt file - i looked at your dsdt file and your dts, but was not able to see how you got from one to the other :)
<travmurav[m]>
I think the labels were somewhat obvious there for kb/touchpad, they also reference the i2c line and are of a generic hid i2c type
<clover[m]>
Steev: tried a different USB and it's booting now. Go figure
<steev>
hexdump0815: someone else had a similar samsung galaxy book device, and said they needed qzed's usb mp work to get their keyboard and such to work, perhaps yours is similar?
<travmurav[m]>
Mine didn't need anything special for kb/touchpad
<hexdump0815>
i think i found the kdb in those acpi files - looks like I2C7 to me, but i have no idea how to get from there to the reg, hid-descr-addr and interrupt info for the dts
<hexdump0815>
steev: i think this was for a galaxy book s, which is 8cx so i guess quite different
<hexdump0815>
or maybe it was 850 even
<jenneron[m]>
8cx
<steev>
it was jenneron
<hexdump0815>
travmurav[m]: to me it looks like for usb you just use the default setup like its used for trogdor, so i guess usb worked out of the box for you or did you have to do any extra things to bring usb forward?
<steev>
15:11:15 <jenneron[m]> qzed: thank you, disabling pmic@a and pmic@2 does the trick
<steev>
15:11:44 <jenneron[m]> also, your usb-mp patchset gives working keyboard and touchpad on samsung galaxy book s
<steev>
ah i was in scrollback
<steev>
could be quite a bit different... or they could still use the same connectors
<hexdump0815>
steev: yes - i looked at those usb-mp patches and to me it looks like they are very 8cx specific and not really applicable to sc7180
<travmurav[m]>
hexdump0815: the trogdor and my laptop (presumably yours too, but I think I couldnt find fccid) all use qsip7180 module so this part is same
<travmurav[m]>
this includes some usb stuff as well
<jenneron[m]>
hexdump0815: are you sure you have multi port usb on galaxy book go?
<travmurav[m]>
but anything that is outside may be different, like the EC (keyboard, battery management), touchpad, etc
<travmurav[m]>
I think 7c has only one usb used (available even)?
<travmurav[m]>
the SoC that is
<travmurav[m]>
they use a hub with a single upstream from the SoC in mine
<jenneron[m]>
at least EC and touchpad seem to be on hid over i2c on galaxy book go
<jenneron[m]>
by EC they may mean keyboard
<travmurav[m]>
ECKB is the kb via ec-embedded scanner I suppose
<clover[m]>
steev: any way to check battery from command line?
<travmurav[m]>
clover: I usually cat /sys/class/power_supply/.../uevent
<szclsya[m]>
upower --dump
<travmurav[m]>
hexdump0815: it might actually be easier to get usb up and ssh to the device
<hexdump0815>
travmurav[m]: i would be very find with this as well - any step forward would be good :)
<travmurav[m]>
You say you used my dt, did the usb ports work?
<jenneron[m]>
hexdump0815: try to bind hid-over-i2c on i2c7 with 0x05 address with 0x140 active-low pulled-up interrupt and 0x20 descriptor address
<travmurav[m]>
i think the acpi i2c numbers didn't match with the real ones for me actually
<hexdump0815>
janrinze: ok will give it a try - have to leave in a few mins, but will report later what comes out
<jenneron[m]>
what i've just mentioned should be keyboard
<hexdump0815>
travmurav[m]: usb looked very dead to me and it cannot find the rootfs on usb which is known to be working in general with my images
<travmurav[m]>
not even power?
<travmurav[m]>
is this usb-a or some type-c only?
<hexdump0815>
sorry - jenneron[m] i ment you and not janrinze
<hexdump0815>
travmurav[m]: there are two usb-c and one usb-a ports - i think i tried all of them and none worked
<hexdump0815>
also from waht i was able to see from dmesg there were no traces about usb ... i guess plugging in a device should at least make some noise with high log levels - ?
<hexdump0815>
jenneron[m]: yes will try what you suggested for kbd
<travmurav[m]>
hexdump0815: you should probably drop most of stuff from the dts tbh, you may have some gpio misconfigured for my board which may e.g. disable usb power
<travmurav[m]>
or in the worst case may even damage the hardware
<jenneron[m]>
hexdump0815: and touchpad is on i2c11 with 0x40 address, 0x0180 active-low pulled-up interrupt, and 0x0E descriptor
<jenneron[m]>
touchpad also uses ldo10a and seems to have a reset pin
<hexdump0815>
travmurav[m]: that sounds like a good idea: strip down the dtb and add things one by one ... problem is that i would need the minimal required dts - would this basically be an empty one only including sc7180.dtsi?
<travmurav[m]>
hexdump0815: there is some stuff that you can expect to be similar, i.e. pmic regulators (tho also need to take care) and stuff like usb phys, wifi, ... power should be inside the qsip
<travmurav[m]>
keyboard, touchpad, ec definitely goes out, the display bridge and panel too
<travmurav[m]>
all the fixed regulators and anything that uses them
Votes78 has quit [Ping timeout: 480 seconds]
<hexdump0815>
i'll have to leave now, but will read the backlog for sure in case you have any other suggestions or ideas
<hexdump0815>
thanks a lot so far
<hexdump0815>
leezu: btw. do you still have your galaxy book go and if yes: did you do any further experiments with it?
Votes78 has joined #aarch64-laptops
<travmurav[m]>
hexdump0815: do you have the fccid for this laptop?
<steev>
clover[m]: i use gwolf's ruby script he wrote (i modified it slightly for the thinkpad since it was written for the c630) - https://paste.debian.net/1250914
<clover[m]>
wow
<steev>
i don't like re-inventing the wheel all the time :)
<clover[m]>
steev: do we do entropy with x13s?
<clover[m]>
i was having some package signing issues
<clover[m]>
i tried installing haveged but it doesn't seem to work on recent kernel versions
<steev>
that's odd that it doesn't
<steev>
you might need to pass an option to it, but i thought upstream fixed that
<clover[m]>
X13s is going to earn this tux key sticker today!
Votes78 has quit [autokilled: This host violated network policy and has been banned. Mail support@oftc.net if you think this is in error. (2022-08-18 20:57:18)]
Votes78 has joined #aarch64-laptops
Votes78 has quit [autokilled: This host violated network policy and has been banned. Mail support@oftc.net if you think this is in error. (2022-08-18 20:58:35)]
Penguinpee has quit [Quit: Leaving]
<HdkR>
ooo, I see patches on the ML to improve GPU recovery
<clover[m]>
Leo Shen: did you set up swap partition? and if so, how big?
<szclsya[m]>
I haven't do swap on x13s, but usually I use a swap partition the same size as the system ram to allow hibernation
megatradeusa[m] has quit [Excess Flood]
<clover[m]>
pacstrapping now wish me luck
megatradeusa[m] has joined #aarch64-laptops
<clover[m]>
hwclock cannot access the Hardware Clock via any known method. hmm
<hexdump0815>
travmurav[m]: fccid of the galaxy book go is a3lnp345xla
<steev>
clover[m]: that's unfortunately, correct
<clover[m]>
steev
<steev>
hwclock on qualcomm is a pita
<clover[m]>
do i still need to do the PARTUUID= thing instead of just UUID= in fstab?
<steev>
afaik, we can only read it, not set it
<steev>
uhhh
<steev>
if you aren't using an initrd, yes
<steev>
wait
<steev>
in the fstab you can use the UUID
<steev>
but the kernel cmdline requires PARTUUID if you don't have an initrd
<clover[m]>
does this look ok to you? for dual boot
<clover[m]>
steev: does PARTUUID and UUID have different values?
<steev>
yes
<steev>
i suppose, if whatever kernel install-y tool you use, if it updates the rootfs via fstab, then you'd want to use PARTUUID in there, but i'd assume that type of stuff would only happen if it's also generating an initrd
<clover[m]>
when i did bootctl install i got failed to write 'LoaderSystemToken' EFI variable: invalid argument
<clover[m]>
maybe its that LoaderSystemToken thing
pierro78 has quit [Remote host closed the connection]
<steev>
bamse: actually, i did end up getting adsp in the qcom_stats
<steev>
qzed: thanks for the pointer, i'll submit a patch because that does appear to fix it up
<steev>
steev@limitless:~$ ls /sys/class/power_supply/
<steev>
yoga-c630-adapter yoga-c630-battery
<szclsya[m]>
hmm seems the new kernel sometimes makes the screen dark after boot
<qzed>
huh, what did I do?
<szclsya[m]>
*after cold boot
<megatradeusa[m]>
Everyday is a winning day we make thousand of dollars! Get started today and enjoy daily profits. I'll help 10 people to earn $30k within 72 hours but you will pay me 10% of your profit when you receive it. Inbox to get started now click on the link below. https://t.me/+_PQUJRTG9QE2Y2U0