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)
<HdkR>
I'll be generating a Fedora one once I get some more time
<clover[m]>
Oooh
<clover[m]>
Will test tomorrow if time permits
<HdkR>
Seemed to work here, but I only tested in a VM. I'm not an avid Arch user after all
chesterlintw has joined #aarch64-laptops
<chesterlintw>
Hi, I have a question about the Linux boot option offered by Lenovo X13s UEFI FW since FW-ver 1.49. Did anyone test it? How does it work?
<chesterlintw>
I tried editing the boot entry of Linaro customized debian (Upgraded to debian12) by pointing it back to the grubaa64.efi rather than "DtbLoader.efi" but it seems not working.
<chesterlintw>
Not sure how and where to put the upstream DTB so I simply copied the sc8280xp-lenovo-thinkpad-x13s.dtb in both EFI partition's /dtb and /efi/debian/
<chesterlintw>
Is the CHID is required for Lenovo's Linux Boot option in UEFI?
<chesterlintw>
Sorry for asking lots of questions since there's no information offered by Lenovo's release note. I am currently working on openSUSE enablement on LenovoX13S.
<chesterlintw>
I have a customized kernel based on SUSE's kernel tree (master v6.4-rc4 + kernel-default flavor) and it can boot through the debian boot path, for example:
<chesterlintw>
Lenovo UEFI -> DtbLoader.efi + CHID dtb for x13s (replaced with upstream dtb) -> Linaro customized Debian's grub2 -> test suse kernel
hexdump0815 has joined #aarch64-laptops
<chesterlintw>
But I am still trying to enable the pure opensue boot path since somehow the opensuse boot path always triggers CPU reset although I have manaully compiled and placed the DtbLoader.efi before booting into bootaa64.efi (grub2)
hexdump01 has quit [Ping timeout: 480 seconds]
<chesterlintw>
Anyway that's another path compared to Lenovo's "Linux Boot" option.
iivanov has joined #aarch64-laptops
<clover[m]>
<chesterlintw> "Hi, I have a question about..." <- It removes the need for the dtb= kernel parameter. At least when booting from nvme. It doesn't seem to find your dtb if it's on a usb. Just keep it in the top of the ESP
HaikuUser has joined #aarch64-laptops
HaikuUser has left #aarch64-laptops [#aarch64-laptops]
<chesterlintw>
<clover> Thanks for your guidance. By keeping the .dtb file in the top of ESP now I can see the "DEVICE TREE" table in efisystab.
xroumegue has joined #aarch64-laptops
iivanov has quit [Remote host closed the connection]
iivanov has joined #aarch64-laptops
laine_ has joined #aarch64-laptops
laine has quit [Ping timeout: 480 seconds]
Segfault11 has joined #aarch64-laptops
<Segfault11>
hi i hear this is the place to ask about various qualcomm laptops? i'm having some issues with getting linux going on a galaxy book go 5g, mainly related to the pmics afaict
<krzk>
Segfault11: do they run mainline Linux? If not, then no - any weird vendor kernel means you talk to weird vendor :)
<Segfault11>
this is mainline :)
<Segfault11>
although it's one of the machines that comes preloaded with windows
<krzk>
git grep for galaxy book did not point anything on "book go"
<krzk>
Segfault11: ah, so you mean a new DTS. Sure, that's one of places.
<Segfault11>
awesome, so the main problem i seem to be running into right now is that for some of the regulators i can't specify the voltage because when i do i get a "Failed to get the current voltage" error (-ENOTRECOVERABLE)
<Segfault11>
and if i simply don't include the voltage that's fine at first but then it seems like anything relying on those regulators starts to be problematic?
<Segfault11>
idk it's super weird
<Segfault11>
i'm getting it on s5a, l2a, l17a, s6c and s7c
<krzk>
Segfault11: maybe you are mixing PMICs? E.g. set regulators from PMIC C to PMIC A
systwi_ has joined #aarch64-laptops
<Segfault11>
nah i've made absolutely sure to check that i'm not mixing things up
<krzk>
Look at current Lenovo Flex - from mailing lists
<krzk>
Segfault11: there is PMIC A and B (both PMC8180), but you have only one
<krzk>
Segfault11: then how are you sure that things are not mixed? From where do you take the DTS?
<Segfault11>
i'm getting data from the decompiled DSDT
systwi has quit [Ping timeout: 480 seconds]
<Segfault11>
and looking at the flex 5g i'm seeing regulators from pmics A, C, and E if that's what you're talking about? if there's something saying otherwise on the mailing list could you link to it? i'm not very good at browsing them
<jenneron[m]>
i have the same problem on galaxy book s, with the regulator responsible for touchscreen
<jenneron[m]>
i also have this problem on lg v35 phone with the regulator responsible for DAC apparently, it is most likely some problem in mainline, because there is this regulator defined in downstream for this phone
<jenneron[m]>
i don't know how to solve it unfortunately
<krzk>
Segfault11: yes, that's what I am talking about. You don't have C there...
<krzk>
Or you moved it somewhere deep...
<Segfault11>
oh it's below e, i forgot to move it up
laine_ is now known as laine
agl7-Galaxy has joined #aarch64-laptops
<clover[m]>
chesterlintw: Happy to help. good luck with SUSE :)
agl7-Galaxy has quit [Ping timeout: 480 seconds]
<steev>
clover[m]: soon(tm) sorry
<steev>
i got caught up with a new vidya game release
<clover[m]>
TOTK? :)
<steev>
i would... if the a690 vulkan driver worked :( or i knew a hack to make yuzu think it works ;)
<steev>
(it's not mesa fault)
<clover[m]>
i wonder if anyone has made an issue on the yuzu GitHub
<steev>
i think you and i might be the only people who tried it... and if you didn't... then probably not, because i didn't
<clover[m]>
whats a vulkan game that should work on the X13s? i remember there being one with shooting crabs or you are a crab shooting other crabs
<HdkR>
Hollow Knight
<HdkR>
Ender Lilies
<HdkR>
Hades
<quinine>
<clover[m]> "whats a vulkan game that..." <- I tried the bevy-engine based game and it usually runs very well.
<clover[m]>
how are you recording your screen HdkR?
<HdkR>
HDMI capture with screen duplication
<HdkR>
Accidentally uploaded that at 30fps though
<HdkR>
I always capture at 60 though
<clover[m]>
do you need separate hardware for HDMI capture?
<HdkR>
Yea, my desktop has an HDMI capture card in it
<HdkR>
Oh wait, that video says it has a 60 option, probably one of the previous videos I hecked up
<HdkR>
I don't want to do on-device capture because the overhead would be terrible
<HdkR>
Screen mirroring over HDMI is pretty low impact. I just wish HDMI audio worked so audio latency was less terrible
<HdkR>
pulse network audio routing is something like 30-100ms of latency?
<steev>
clover[m]: pushed rc5, it's johan's stuff, with my stuff on top (nothing new on my side from rc4)
<steev>
though v3 of lpass was posted today, i could have grabbed that but didn't; i'll try to get to it after dinner
<HdkR>
steev: I assume there's no DP audio patches in there? :P
<steev>
nope, it's just lpass
<steev>
oh, actually, i pulled in ajhalaney[m]'s patch for usb to use dev_err_probe, but that's a minor thing that i don't expect most end users to see
<steev>
clover[m]: i haven't added the wwan stuff yet to the defconfig either, but i did add the PM8008 to it
<steev>
not sure if everything is in the kernel config that would be needed for the fingerprint reader (i don't do biometrics on any of my devices)
<ajhalaney[m]>
steev: nice, I ran into that when I was missing an interconnect in the initramfs, was getting spammed to death :P
<steev>
yeah... some qcom things are quite spammy
<steev>
like the various gpr ports not being registered - no idea what they are... or how to register them, or if they should (but if they shouldn't be, then it shouldn't complain that they aren't)
<clover[m]>
steev gotcha, generated a new config based on laptop defconfig to pull the changes (and threw in wwan stuff on top for people who use that)
<clover[m]>
steev w.r.t. firmware, is hpnv21.8c now called hpnv21.b8c?
<steev>
it was originally called b8c, you renamed it to 8c
<steev>
it should be b8c (imo)
<clover[m]>
must have been my typo. will fix that
<steev>
no, it wasn't a typo
<steev>
it was on purpose because the qca2066 stuff that generates the name somehow drops it, i dunno why, i added in a hack that does that (but if you don't use the hack, it doesn't bother to check for that file at all)
<clover[m]>
what does PM8008 do?
chesterlintw has quit [Ping timeout: 480 seconds]
<clover[m]>
whoa ok i am seeing a new usb device
<clover[m]>
Bus 005 Device 002: ID 06cb:00fc Synaptics, Inc.
<steev>
that
<steev>
i think?
<stirl>
clover[m] -- I got up to the bootloader install on my lunch break. If I boot back into the live env and mount/chroot back in will I be able to finish the install or do I need to start over from scratch
<steev>
or maybe the pm8008 is part of the camera work
<clover[m]>
got fingerprint scanner working with gdm
<clover[m]>
super cool
<clover[m]>
stirl: the chroot should be permanent, but anything you do on the live usb will be undone each boot
<clover[m]>
should be permanent if you mounted something from internal disk and did work there. stirl you might want to wait since i am working with eos devs to clean up the tutorial. but do what you want :)
<stirl>
I'll go for it because if I have time tonight but I won't have any time over the weekend really. Everything seemed to go pretty smooth from the tutorial. Then only weird thing that happened was after installing all of the packages in chroot I didn't have access to genfstab or vim for some reason. I just ran the install command for those 2 things and everything was fine but they didn't get installed the first time somehow
<clover[m]>
Genfstab should work out of the chroot. You can install vim with pacman
<jhovold>
steev: pm8008 is only needed for camera, yes
<clover[m]>
cheese is still unable to find camera
<steev>
i didn't say it works, i said it's part of it
stirl has quit [Ping timeout: 480 seconds]
stirl has joined #aarch64-laptops
stirl has quit [Remote host closed the connection]