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
<jenneron[m]>
i think it is gnome specific as e.g. plasma should already support it
<steev>
calebccff: one minor thing about it though, you echo that you're setting it to 800 but you're setting it to 900
<jenneron[m]>
> It also triggers a GPU clock boost if deemed necessary by the driver.
<jenneron[m]>
from the MR
systwi has quit [Remote host closed the connection]
Mathew has joined #aarch64-laptops
Mathew has quit [Remote host closed the connection]
<steev>
4 years and still not in
<steev>
speaking of poor performance... wow does the c630 not perform well with LMH not turned on (and i should probably submit a merge request to debian to enable that)
hexdump0815 has joined #aarch64-laptops
hexdump01 has quit [Ping timeout: 480 seconds]
iivanov has joined #aarch64-laptops
iivanov has quit [Quit: Leaving...]
iivanov has joined #aarch64-laptops
<steev>
calebccff: btw, do you happen to see any messages about vblank timeout?
jglathe_ has joined #aarch64-laptops
<steev>
calebccff: ahhh i missed the sensors question... i haven't had a chance to try, i have it loading the slpi firmware, but that's about as far as i have gotten so far
iivanov_ has joined #aarch64-laptops
iivanov has quit [Read error: Connection reset by peer]
jhovold has joined #aarch64-laptops
martiert has quit [Remote host closed the connection]
martiert has joined #aarch64-laptops
shoragan has quit [Quit: quit]
jglathe_ has quit [Remote host closed the connection]
jglathe_ has joined #aarch64-laptops
iivanov_ has quit [Remote host closed the connection]
iivanov has joined #aarch64-laptops
xroumegue has quit [Ping timeout: 480 seconds]
xroumegue has joined #aarch64-laptops
hightower2 has quit [Ping timeout: 480 seconds]
craftyguy has quit [Remote host closed the connection]
craftyguy has joined #aarch64-laptops
<calebccff>
jenneron[m]: right but this will only mask the issue, there will still be input latency on gestures
<jenneron[m]>
steev: can you share some of your firmware files? i would like to check that lenovo flex 5g firmware can be loaded on yoga 5g and vice-versa
systwi has joined #aarch64-laptops
<jenneron[m]>
steev: by any chance, do you have windows on UFS on flex 5g?
enyalios has quit [Remote host closed the connection]
<adamcstephens>
anyone have issues with the x13s trackpoint primary mouse button not always registering? or is this just a me problem?
<adamcstephens>
* registering a click
enyalios has joined #aarch64-laptops
<steev>
jenneron[m]: yeah i still have windows, i don't believe any of those files are on there though; but i can look after work
<steev>
calebccff: https://paste.debian.net/1304175/ here's my "dmesg -tl warn,err" on my c630; the genirq lines are the one that make me wonder why but i'm assuming those are the ones breaking usb
<maz>
steev: can you dump the DT interrupt specifiers for these two interrupts?
<konradybcio>
steev what kernel version is this?
<steev>
maz: afaik they're just the defaults in sdm845.dtsi
<steev>
konradybcio: 6.7.0
<steev>
maz: or do you mean the GIC_SPI 101 IRQ_TYPE_LEVEL_HIGH
<albsen[m]>
are these kernel params are still needed on 6.5 in debian? "iommu.passthrough=0 iommu.strict=0 pcie_aspm.policy=powersupersave"?
<albsen[m]>
albsen[m]: when adding the following my battery percentage doesnt show anymore: pcie_aspm.policy=powersupersave (the rest appear to be working)
<nerdboy>
i think battery is mainly pd-mapper and qcom fw in the initrd
<nerdboy>
enyalios: what profile are on? is it a systemd one?
<nerdboy>
and you do need the pcie_aspm param on 6.5 <= albsen[m]
<nerdboy>
i removed all of those for 6.7
<albsen[m]>
nerdboy: I've been playing around these the params after updating the firmware as suggested. also linked the audioreach fw manually to make sound work. but not working yet. (i'll keep the pcie_aspm param, ok)
<albsen[m]>
this is what I'm currently using as kernel params: "ro arm64.nopauth clk_ignore_unused pd_ignore_unused iommu.passthrough=0 iommu.strict=0 pcie_aspm.policy=powersupersave"
<nerdboy>
i just managed to get my machine happy yesterday
<albsen[m]>
I also had "rd.driver.blacklist=msm" bcs I saw that in ubuntu but it doesnt really make a difference
<albsen[m]>
nerdboy: how?
<nerdboy>
i have the same-ish ones: ro console=tty1 rootfstype=ext4 rootwait net.ifnames=0 pd_ignore_unused clk_ignore_unused arm64.nopauth
<albsen[m]>
why do you have rootwait and net.ifnames=0?
<nerdboy>
for me it was switching from dracut to genkernel to build my initramfs
<nerdboy>
mainly because i normally do embedded stuff and i don't like the nonsense net names
<albsen[m]>
ok
<nerdboy>
i would check the contents of your initramfs and make sure the qcom paths are there
<albsen[m]>
you mean in /lib/firmware/qcom?
<nerdboy>
whw
<nerdboy>
wherever your fw lives
<nerdboy>
i just took a stab at the top-level dirs in mine: FIRMWARE_FILES="qcom/sc8280xp,qca,ath10k,ath11k,ath6k,RTL8192E,regulatory.db"
<nerdboy>
*and the db file
<steev>
albsen[m]: you need newer alsa-ucm-conf than what debian has
<steev>
because upstream hasn't done a release with the fixes yet
<nerdboy>
yeah, the volume fixes are only on upstream master right now
<albsen[m]>
I don't have a device for audio out, nevermind the volume
<nerdboy>
does debian have an lsinitrd tool? if not then you'd need to dump your initrd to a tmp dir
<albsen[m]>
albsen[m]: this one is old ... must be somewhere else
<steev>
albsen[m]: actually... on 6.5 i think audio was different
<nerdboy>
i *think* the debian one is just a compressed cpio archive?
<albsen[m]>
steev: I can move to a newer kernel, I saw that you all were talking about 6.7
<albsen[m]>
nerdboy: idk, will chekc
<steev>
should make sure you have power-domain-mapper installed
<steev>
and started
<steev>
and while i have the firmware in my initrd, it really shouldn't be needed
<nerdboy>
i figured out yseterday that setting the bdaddr is essentially an on/off switch for bluetooth
<steev>
i did that because i was playing with lumag's patch to add pd-mapper into the kernel
<steev>
yes, it needs a valid address
<albsen[m]>
can I pick any address for these? I didn't write down the addresses in windows and now it's on a different drive not installed anymore on my laptop.
<steev>
yeah
<steev>
F4:A8:0D:XX:YY:ZZ if you want one that's sorta valid
<nerdboy>
i used a qualcomm prefix, apparently they have at least 3
<steev>
the MSB/LSB are because... that patch is still... ???
<steev>
so if i'm on a kernel with the LSB patch, i use the LSB style, and MSB for MSB style, though technically either would work for either
iivanov has quit [Remote host closed the connection]
<calebccff>
im booting with systemd-boot and EFISTUB on the c630 and i notice about 1 in 5 boots it will just hang after the EFISTUB loader, no UART output, and the screen doesn't go blank so it's clearly crashed. Anyone else hitting this?
<albsen[m]>
<nerdboy> "does debian have an lsinitrd..." <- so I checked, and the firmware is not listed when running lsinitramfs.
<albsen[m]>
but I have it clearly in /lib/firmware and regenerate it using the same command. odd.
<nerdboy>
how big is the file? with partial firmware all the modules genkernel creae
<nerdboy>
*creates a ~60 MB initrd (compressed with xz even)
<albsen[m]>
50M Jan 15 23:52 initrd.img-6.5.0-5-arm64
<nerdboy>
agaim, i would dump the initramfs to verify what's in there
<albsen[m]>
I even am able to regenerate it using sudo update-initramfs -u -k 6.5.0-5-arm64
<albsen[m]>
hm, u mean, lsinitramfs might miss something?
<albsen[m]>
this is how I was checking: lsinitramfs /boot/initrd.img-6.5.0-5-arm64| grep firm
<nerdboy>
i'm just telling you what worked for me with 6.7 and the other bits, since prior to my hardware failing i was using dracut *without* the firmware
<albsen[m]>
ok
<albsen[m]>
thx, I'll continue to debug
<nerdboy>
and no, i would expect lsinitramfs listing is correct
<nerdboy>
so, now that i have the replacment hw and started from scratch again i could not get the battery status to show in gnome without including the qcom fw bits in my initramfs
<nerdboy>
also i did bump my pd-mapper and qrtr pks from upstream 1.0 tags to master HEAD
<nerdboy>
*pkgs even
<nerdboy>
after that ^^ all the important bits are working => battery status, audio (just alsa/pulse), and USB dock/display
<albsen[m]>
nerdboy: do u get high resolution working on display out?
<nerdboy>
i only have a 1080p 23" or so externel display
<nerdboy>
also i ressurected my HH keyboard so i get a lot of interstioin
<nerdboy>
interesting typos...
<nerdboy>
eg, back-tick key is opposite from where it should be
<albsen[m]>
:)
<nerdboy>
there are at least 3 versions of my cheesy j5 dock; mine has one hdmi and claims 4k support and the other one has dual hdmi ports
<nerdboy>
"Supports 4K HDMI™ or dual 1080p HDMI™ (JTS427) display on compatible USB-C® laptops with
<nerdboy>
DisplayPort™ 1.2 alt mode with MST support* (JTS427)"
<nerdboy>
the dock plus a nice aluminum stand was like $40 so i couldn't resist
<nerdboy>
the other on has ethernet and is less cheap
<steev>
calebccff: i do see something similar but not with systemd-boot, it just hangs when it should be booting but... no idea why
<steev>
albsen[m]: i get 2K on my 2K display, but i *think* up to 4k 30 works?
<steev>
albsen[m]: you're using sid? testing?
<albsen[m]>
yes
<steev>
those are two different suites :P
<steev>
sid should have working power-domain-mapper; but maybe ema can chime in
<nerdboy>
i could swear bookworm just moved to stable
<steev>
it did
<steev>
bookworm was *testing*
<steev>
sid is *unstable*
<nerdboy>
seems like a mixed metaphor?
<steev>
sid and testing are rolling, and when a release is cut, it's based on a snapshot of testing and then fixes applied
<nerdboy>
yes, i am one of those weirdos who only uses those other distros for specific use cases
<nerdboy>
beagleboard stuff makes me touch debian/ubuntu, work stuff makes me touch debian/ubuntu/petalinux/yocto/centos
<nerdboy>
ugh
<nerdboy>
mostly it's enjoyable except the rh2.1 bits... that is painful env after gcc 13 and 6.x kernel
<calebccff>
steev: having now nuked my windows install.... i realised i forgot to yoink the sensor configuration data, do you know where one could maybe find this?
<nerdboy>
heh, my new hw was actually blank
<nerdboy>
although they never asked if that was okay...
<steev>
calebccff: i thought based on my conversations with dylan that i would have to generate them?
<steev>
but, i could probably be convinced to look around on there, since i still have it on my c630 and jenneron[m] wants me to look on the flex5g as well
<jenneron[m]>
steev: yeah, actually i would like you to look into SLPI memory range, because i'm not sure where it should be
<jenneron[m]>
i currently have this diff and it results in -22 when loading firmware