ChanServ 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
alpernebbi has quit [Ping timeout: 480 seconds]
alpernebbi has joined #aarch64-laptops
<dgilmore> ath11k on 6.11 is a mess
<dgilmore> on the x13s
hexdump01 has quit [Ping timeout: 480 seconds]
<steev> i was hoping the powersequencing stuff would help, but it doesn't seem to
jhovold has joined #aarch64-laptops
iivanov has joined #aarch64-laptops
martiert has quit [Quit: WeeChat 4.4.1]
martiert has joined #aarch64-laptops
<Jasper[m]> Out of interest, can someone with a 12 core X Elite run unixbench?
<Jasper[m]> Specifically whets and dhry
jhovold has quit [Read error: Connection reset by peer]
<HdkR> Jasper[m]: How does one run unixbench?
jhovold has joined #aarch64-laptops
<Jasper[m]> HdkR: What distro are you running?
<HdkR> Ubuntu mostly
<Jasper[m]> HdkR: It's in the snapstore :^)
<HdkR> gross
<HdkR> `Can't locate Time/HiRes.pm in @INC (yo` perl fail
<Jasper[m]> rip
iivanov has quit [Read error: Connection reset by peer]
iivanov has joined #aarch64-laptops
xroumegue has joined #aarch64-laptops
<kuruczgy[m]> Does anyone know perhaps how to get the TPM working on X1E? I don't see a device tree binding for it.
<kuruczgy[m]> How does the TPM work on the X1E? Is it using some standard interface, and only the device tree binding is missing, or is it something more complicated that needs a custom driver?
<Jasper[m]> Haven't heard about X1E specifics, but on the X13s it was accessed through Trustzone calls which haven't been implemented (yet?)
<Jasper[m]> Maybe if your device has a separate chip you can access it like any other device, but idk about that
<kuruczgy[m]> Jasper[m]: I am pretty sure it is Microsoft Pluton which is integrated in the SoC, not a dTPM.
svarbanov has joined #aarch64-laptops
<abelvesa> robclark: pacman says 24.1.6
<jhovold> Here's an updated wip branch for the X13s:
<jhovold> Changes include:
<jhovold> - fix in-kernel pd-mapper use-after-free when stopping remoteproc
<jhovold> Here's an updated wip branch for x1e80100:
<jhovold> Changes include:
<jhovold> - fix scmi perf domain debugfs registration
<jhovold> - describe usb multiport controller
<kuruczgy[m]> jhovold: Thanks, will try it out soon!
<kuruczgy[m]> Any update on the x1e pd-mapper issue? (For now the userspace one seems to be working fine though.)
<jhovold> robclark: I decided too keep the cpuid hack for now to avoid breaking people's setups until the new mesa hits the distros
<jhovold> iiuc, it only affects how the gpu is reported (as A740) so there should be no rush to drop it
<abelvesa> jhovold: so no weird display artifacts on your side then ?
<abelvesa> jhovold: I remember seeing them on CRD as well
<abelvesa> jhovold: you're on wayland though, right ?
<jhovold> kuruczgy[m]: no, the in-kernel pd-mapper is still broken and there's currently no plan on how to fix it
<jhovold> I'll send a patch to disable it for all SoCs if there's no progress in the next few weeks
<jhovold> abelvesa: no, nothing I've noticed, but I mostly use the crd over ssh
<jhovold> also on wayland, yes
<jhovold> same mesa as you
<jhovold> with zink emulation
<abelvesa> then you will probably see them with gnome as well
<jhovold> i don't use gnome, no issues with dwl
<abelvesa> oh, wait, I'm on xorg, maybe that is it then
<jhovold> hopefully goes away when you switch to freedreno
<jhovold> that may be it, yes
<jhovold> xorg works fine here too, still no gnome(s)
<anonymix007[m]> jhovold: do USB-A ports work with multiport patch? Or is there any additional work that needs to be done?
<abelvesa> anonymix007[m]: AFAIK, the multiport patch only adds the controllers and PHYs, one would still need to add the board specific DT stuff
<abelvesa> anonymix007[m]: which device do you want multiport on? I'm trying to look into adding it on t14s, maybe that will help.
<jhovold> anonymix007[m]: what abel said, the patch is just adding the controller to the soc dtsi, and is not usable on its own
<anonymix007[m]> jhovold: I have T14s
<anonymix007[m]> abelvesa: where can I find your patches for T14s?
<HdkR> Jasper[m]: Would you be interested in Bytemark results?
<abelvesa> anonymix007[m]: I run with jhovold's branch + t14s dt patches from konradybcio (for now)
<abelvesa> anonymix007[m]: intend to push a branch based -next with everything I have (even WIP patches) to my usual codelinaro tree, hopefully today
<Jasper[m]> HdkR: I specifically wanted to ask for unixbench because of the pmOS wiki page. It has a bunch of results
<HdkR> ah
<Jasper[m]> Wanted to see for fun if the sc8280xp got passed (I assume so, unless core scaling is broken somehow)
<HdkR> In bytemark the results are hit or miss. Sometimes X13s wins, sometimes Yoga 7x wins
<anonymix007[m]> abelvesa: I actually do the same and even set up an arch repo for easier upgrades (since the kernel can be cross-built): https://github.com/anonymix007/t14s-alarm
<abelvesa> anonymix007[m]: BTW, you might want to look into the new "pacman-pkg" kernel target ;)
<anonymix007[m]> Does it work with cross-compilation?
<abelvesa> yes it does
<abelvesa> anonymix007[m]: but you will also want to apply this patch before testing pacman-pkg: https://lore.kernel.org/all/20240819-uncompressed-distro-packages-v1-1-c8accc8bc9ea@quicinc.com/
<abelvesa> specially if you run with systemd-boot
<anonymix007[m]> I probably don't need it since I run grub
<abelvesa> well, until yesterday I was running my t14s with grub as well, but managed to get systemd-boot + uki to work properly
<anonymix007[m]> How does that pacman-pkg handle initramfs? I have a separate mkinitcpio config file in the same package (basically, the same approach what clover uses in his linux-x13s package)
<abelvesa> well, mkinitcpio.conf will have to be on your target anyway
<abelvesa> it's not part of the package
<abelvesa> so when you run pacman -U linux-upstream-...tar.xz, it will run mkinitcpio with whatever mkinitcpio.conf you have in /etc
<anonymix007[m]> BTW, are the key mappings on T14s looking strange for you as well? Namely, Fn (which always behaves as XF86TouchpadOff) and Copilot (which appears to be a key combination) keys
<travmurav[m]> perhaps EC needs some flags set to switch to the "OS" layout from "UEFI" layout
<travmurav[m]> on aspire1 ec forces media keys to be always f1 f2... (so you can always use setup menu even if you press fn I guess), and one needs to notify it that 'rich os" has booted so it unlocks media keys
<travmurav[m]> But copilot being a combo is not even surprising I guess?
<travmurav[m]> Probably just issues whatever is the default combo to open it on windows?
<anonymix007[m]> So there's no way to get it mapped back to Super_R?
<travmurav[m]> Unless there is something in firmware setup to let ec know you don't want any pilots, I doubt it, ec firmware probably just issues that combo
<robclark> abelvesa, jhovold: you need newer mesa and to drop the zink override
<jhovold> zink override works ok here for now (e.g. until arch pushed new mesa)
<jhovold> may be related to abel's problem though, as he said gnome appears to fall back to sw rendering
<jhovold> only thing I've noticed (in my light wayland use) is video stuttering in firefox, hoping switching to freedreno will fix that too
<anonymix007[m]> Was there any progress made on hw video decoding for x1e? The latest iris driver patchset is 9 months old and only supports sm8550
<HdkR> robclark: Alright, updating to 24.2.0 fixes a bunch of the random bugs I was having. Will need to test some applications to see if I can still repro any instability
<abelvesa> robclark: switch to newer meta, artifacts gone
<abelvesa> robclark: thanks!
<abelvesa> newer mesa
<abelvesa> gnome on wayland, at least
<abelvesa> trying on x next
<abelvesa> yep, x works as well
<robclark> \o/
<robclark> yeah, 24.2.0 is the first mesa version that is expected to work
<HdkR> But random commit is so much more fun. Filled with adventure
<robclark> heheh
freekurt[m] has joined #aarch64-laptops
<freekurt[m]> I want to run Debian on my x13s with a working camera and video acceleration. I tried armbian's latest build, but these two things aren't working on it. Does anyone have suggestions on what I can tell the devs there to do to fix these things? https://github.com/armbian/build/blob/main/config/boards/thinkpad-x13s.wip
<freekurt[m]> Or, alternatively, is there a non-armbian Debian image out there that has these things working? Debian's "how to install" wiki seems overly complicated and probably outdated in certain places.
<FarchordSteveCossette[m]> Video acceleration right now afaik is hit or miss. Because of several things.
<FarchordSteveCossette[m]> I haven't heard anything about camera. And for me, audio is also non functional
<FarchordSteveCossette[m]> I personally decided to put a hold on this until there'S some more progress. That, and traveling.
<jhovold> freekurt[m]: camera and video accel is not in mainline yet, but you can use my wip branches, the user space deps are also listed here:
<jhovold> there are some bits missing in user space for video acceleration, so you can't expect any app to use it, and the fw hasn't been released yet (copy from windows)
<jhovold> similar for camera, you should be able to get it to work with pipewire and firefox now, but not sure anyone has really tested it
<jhovold> and pipewire breaks audio...
<jhovold> FarchordSteveCossette[m]: are you also having trouble with debian specifically? audio has been working since 6.5, so any issues should be in your distro or setup
<jhovold> well, except for the pipewire issues...
<FarchordSteveCossette[m]> jhovold: Oh sorry, I missed that he was talking about the X13s....
<FarchordSteveCossette[m]> not the x1e devices. My bad.
<FarchordSteveCossette[m]> Lenovo and their naming schemes lol
<jhovold> ah, ok, thanks for clarifying that :)
<steev> camera works with cheese, but sometimes you have to restart cheese a few times
<steev> that's assuming the wip branches
<jhovold> works fine with qcam, no need to restart that I've noticed
<steev> also tested in firefox and it works (at least on the webcam test websites)
<jhovold> but perhaps taking selfies isn't all that interesting, I guess video conf is what we all want
<steev> i've never used qcam so dunno there, i just know what comes with gnome
<jhovold> yeah, it's supposed to be support with firefox 129, as long as you have all the pipewire deps
<jhovold> if only someone could fix pipewire audio on these machines...
<steev> in good news, these kinda growing pains aren't specific to us :D apparently the sun4i-drm driver has been broken since like 6.8 for the pinebook
<steev> well, anything newer than 6.1 really, but sometimes it worked and sometimes it didn't
<jhovold> camera in firefox still doesn't work out-of-the box on arch at least, not sure what is wrong
<jhovold> camera still showing up as [v4l2] in wpctl, last time I had to rebuild pipewire to get it to be managed by libcamera
<jhovold> but libcamera and pipewire should be recent enough now
<steev> in debian, it just shows up as "Built-in Front Camera"
<steev> at least in cheese
<jhovold> cheese works, I'm was talking about firefox and pipewire
<steev> firefox also calls it that here (firefox 129.0.2)
<jhovold> what does it say after the camera devices if you run 'wpctl status'
<jhovold> should say libcamera or similar if it works in ff
<steev> 85. ov5675 [libcamera]
alfredo has joined #aarch64-laptops
alfredo has quit [Quit: alfredo]
rz has joined #aarch64-laptops
rz_ has quit [Ping timeout: 480 seconds]
rz has quit [Remote host closed the connection]
jhovold has quit [Ping timeout: 480 seconds]
<JensGlathe[m]> Updated rc5 packages with the patches of @_oftc_jhovold:matrix.org worked in: https://drive.google.com/drive/folders/1_ZQ-TwoQckFfxqHCvVB1ShVtsFu1mhi0?usp=drive_link
<JensGlathe[m]> And new Ubuntu bootable image v5 for sc8280xp / X1E80100 Yoga 7 slim and Vivobook S 15 is up: https://drive.google.com/file/d/1QjI3Qyy_P-A2GtSi09H77clDB5JHDmeY/view?usp=sharing
<JensGlathe[m]> Same kernel, does module blacklisting via kernel parameters
phire_ has joined #aarch64-laptops
phire is now known as Guest1571
phire_ is now known as phire
Guest1571 has quit [Ping timeout: 480 seconds]