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)
shoragan has quit [Quit: quit]
shoragan has joined #aarch64-laptops
<clover[m]> (⊙_◎):you use openSUSE right?
<steev> ?
<steev> the rc7-hot was for someone to play with
<steev> what modules are you missing? it should be kernel 6.4.3
shoragan has quit [Remote host closed the connection]
shoragan has joined #aarch64-laptops
agl7-x13s has quit [Quit: Leaving]
agl7-x13s has joined #aarch64-laptops
<clover[m]> alarm finally got updates again!
agl7-x13s has quit [Remote host closed the connection]
agl7-x13s has joined #aarch64-laptops
<HdkR> Sounds like ALARM needs infrastructure help so the updates keep flowing
<steev> probably needs more than lemmy doing things
<HdkR> Indeed
hexdump01 has joined #aarch64-laptops
hexdump0815 has quit [Ping timeout: 480 seconds]
<clover[m]> some armbian images are available for X13s now: https://imola.armbian.com/dl/thinkpad-x13s/archive/
hexdump01 has quit [Quit: WeeChat 1.9.1]
hexdump0815 has joined #aarch64-laptops
<hexdump0815> ungeskriptet: travmurav[m]: short updated - based on the .config from ungeskriptet i was now able to build a kernel myself which sees the ufs devices
<hexdump0815> next steps are to merge my galaxy book go dts with the ufs additions and to try to figure out which are the relevant kernel config options to base the config reproducable on defconfig
<hexdump0815> more about this hopefully during the next days :)
<gabertron> steev: so update on my sound situation: looks like the typical install process people do either uses dpkg -i linux-*.deb (after doing make deb-pkg or something similar I guess) -- that process kicks off the mkinitramfs I believe, and creates that in the normal way -- that process appears to by default not include /usr/lib/firmware/qcom in the initrd img, and also includes a fairly sparse set of modules..
<gabertron> if I do the normal mkinitramfs process, sound works, but I have been using a method to create a template initrd that has all of the modules built by the kernel build process (then I compress with cpio etc), and I guess in my case sound kernel modules load and it fails to find the firmware for the kernel modules -- vs. the normal initrd appears to not actually try to load sound in the initrd so it doesn't care
<gabertron> that the /usr/lib/firmware/qcom dir is not there
<gabertron> (so I can get my initrd to work if I copy the /usr/lib/firmware/qcom dir into my initrd template dir)
iivanov has joined #aarch64-laptops
<gabertron> and sound works with my template initrd method when that qcom fw is there -- so likely not a process others should follow, but i guess I had a goal to not need to run the manual mkinitramfs process on the target when cross compiling.. but yea probably should be doing things the "right way"
<jhovold> _`[m]: not sure what branch you linked to above, but 6.4.3 and 5.6-rc1 are my latest (for a few more minutes) and have all important fixes
<jhovold> 6.5-rc1*
Lucanis has joined #aarch64-laptops
Lucanis0 has quit [Ping timeout: 480 seconds]
Vectorboost has joined #aarch64-laptops
<jhovold> Here's an updated wip branch for the X13s based on 6.5-rc2:
<jhovold> Changes since the previous rc1 branch include:
<jhovold> - fix bluetooth power consumption when not used
<jhovold> - fix audio gpr port warnings when starting stream
<jhovold> - fix usb controller fw crash
srinik is now known as help
help is now known as IDENTIFY
IDENTIFY has quit [Killed (NickServ (Too many failed password attempts.))]
srinik has joined #aarch64-laptops
<_`[m]> steev I'm seeing this path though on install_modules ` INSTALL /lib/modules/6.4.3+/kernel/net/qrtr/qrtr-mhi.ko
<_`[m]> `
<_`[m]> jhovold I tried your branch too, with defconfig and johan_config
<_`[m]> but I think it was the same issue, with the /lib/modules not being found
<_`[m]> and dunno the dtbs was no symlinked or so
apalos_ has joined #aarch64-laptops
mani_s- has joined #aarch64-laptops
Vectorboost has quit [Quit: Leaving]
<Jasper[m]1> @clover about the sound issues from earlier, I think I figured it out. The pipewire pulse plugin wasn't installed yet, now it's plenty loud and I can control the volume
<Jasper[m]1> Also, interestingly, yt video's work now aswell
<Jasper[m]1> And they're a bit sped up it seems
<Jasper[m]1> Also, audio muted on headphones still leaks a bit of sound funnily enough
<steev> _`[m]: well that would do it - you've got 6.4.3+ but you're passing 6.4.3
<steev> what i do here is pass LOCALVERSION="" so that it doesn't put the + at the end (because we have patches on top of 6.4.3)
<steev> that's during kernel build (and install)
iivanov has quit [Quit: https://quassel-irc.org - Chat comfortably. Anywhere.]
<clover[m]> somethings going on with my bluetooth now
<clover[m]> my override is now timing out
<jhovold> clover[m]: override?
<clover[m]> and when i remove it, set pub address, and try to scan, i get Failed to start discovery: org.bluez.Error.NotReady
<clover[m]> yeah the systemd override
<clover[m]> i dont think its a kernel issue because i rolled my kernel back and still having issues
<clover[m]> hmm it as some rfkill thing
<jhovold> ah, ok, got it
<jhovold> can you still set it manually?
<jhovold> yes, so you wrote above it seems
<steev> clover[m]: new bluez?
<steev> mine still works here
<clover[m]> its possibly the new systemd, it doesnt want to set the bt address anymore. i can do it manually though
<clover[m]> i wonder if i need to do yes | btmgmt or something
<jhovold> perhaps related to the issues that prevented using udev for this, had to break out of some sandbox to work around iirc
<jhovold> if only someone could fix bluetoothd so that it support this out of the box...
<jhovold> hans mentioned the netlink "jail" in his initial report about this:
<jhovold> i guess that may be where you got the inspiration for you systemd service too
<clover[m]> yeah, yes command is needed
<clover[m]> im on btmgmt: 5.68
<jhovold> ah, so you got it working again with yes?
<clover[m]> yeah
<jhovold> nice, do you have a link to the service you're currently using?
<jhovold> what changed in btmgmt? is there really some prompt there now waiting for input?
<jhovold> hmmm. i'm using the same version of btmgmt as you
<jhovold> and it does not seem to require any user input
<jhovold> but something appears to have regressed as it no longer recognises --index:
<jhovold> Unable to open 0: No such file or directory (2)
<clover[m]> here you will find the override config i am currently using
<jhovold> clover[m]: thanks, odd that you'd need that yes command all of a sudden, would be interesting to learn why
<_`[m]> <steev> "that's during kernel build (..." <- tx, rebuilding now
<clover[m]> well arch packages were fixed last night after a few weeks hiatus and we got a flurry of packages that came through. one of them might have carried a regression but i can't pinpoint which
<_`[m]> I'm running ubuntu haha, I'll leave now
<_`[m]> I just hate setting up fde
<agl7-x13s> Hello I'am back from my holidays (only 1 week) :-)
<clover[m]> wb
<agl7-x13s> jhovold: I want to compile the newest linux-kernel 6.4.3 of steev. Is there more necessary to set than only CONFIG_IOMMU_DEFAULT_DMA_LAZY=y ?
<agl7-x13s> Hello clover[m] :-)
<clover[m]> hello, if you use arch, i've already packaged it
<clover[m]> you can find config here, if it helps: https://github.com/ironrobin/x13s-alarm/blob/trunk/linux-x13s/config
<agl7-x13s> clover[m]: I'am use at this moment Debian unstable. In some days I will try to install also ARCH Linux on a USB-stick.
<clover[m]> using my ISO is probably the easiest way to get arch right now
<agl7-x13s> clover[m]: Where can I get your ISO?
<clover[m]> and here is a walkthrough for endeavouros install that should work for arch too (just don't do step 14 basically): https://github.com/ironrobin/archiso-x13s/wiki/EndeavourOS-Tutorial
<agl7-x13s> Oh very good. I prever EndeavourOS. Is there a install programm like the version for x86_64?
<clover[m]> sort of. it's not like x86, its a script you run. it is all there in the tutorial
<agl7-x13s> ok .. I read it.
<agl7-x13s> clover[m]: In your EndeavourOS-Tutorial you use there WLAN. Is it possible to use LAN that must also work, or?
<clover[m]> do you mean an ethernet cable? i've never tried tbh, i've only used wifi on my x13s
<clover[m]> i assume its possible though
<Jasper[m]1> It works on archiso with the average RTL8152 dongle
<agl7-x13s> I have an ethernet cable at an Ethernet-Adapter which is connected to an USBB-Port of a USB-C-Box and the Box is connected to the first USB-C-Port (power) of the x13s. The power adapter is connected to the USB-C-Box. This works very well and at this moment I have this connection to the internet.
<agl7-x13s> Steev's kernel works very well with this USB-C-Box and the Ethernet-Adapter
<agl7-x13s> clover[m]: ?
<agl7-x13s> Jasper[m]1: Oh, I see at this moment your message. Ok, it's an Realtek Chip in the Ethernet-Adapter!
<Jasper[m]1> Yes, usage of RTL8152 and 8153 is very common in those usbc adapters
<agl7-x13s> It's an 2.5 Gbit/s Ethernet-Adapter
<Jasper[m]1> Jasper[m]1: My bad, it was just the 8153
<Jasper[m]1> The 8152 is very slow hahaha
<clover[m]> with my repo we are tracking s t e e v activity, so whatever he uses we also use. i guess we are copying kali linux
<agl7-x13s> clover[m]: Du yo use Steev's kernel?
<clover[m]> of course
<agl7-x13s> ok, then i mean there is no problem to use my 2.5 Gbit/s Ethernet-Adapter
<Jasper[m]1> Shouldn't be as long as the driver is included
<Jasper[m]1> You'll know when you plug it in
<agl7-x13s> Jasper[m]1: I will see. When it doesn't work I can use WLAN. I have also WLAN.
<clover[m]> report any issues here or the github tracker, i will respond to either
<agl7-x13s> clover[m]: OK
<agl7-x13s> clover[m]: I already have experience with Manjaro Linux and EndeavourOS on Raspberry Pi machines. So I know the pacman program. However, I use each there mainly yay to update the system or to install software.
<clover[m]> yay is a good AUR helper which is packaged in the endeavourOS repos
<clover[m]> i also use yay
<clover[m]> manjaro is fine, but i'd rather just use upstream arch and get new packages faster
<_`[m]> <agl7-x13s> "jhovold: I want to compile the..." <- whatnow
<_`[m]> so do any arch users configure full disk encryption?
<_`[m]> anyway apparently LOCALVERSION=""
<Jasper[m]1> _`[m]: I did
<agl7-x13s> _`[m]: I have compiled Steev's newest 6.4.3 kernel, but only with "...DMA_LAZY=y".
<Jasper[m]1> You need some modules in your mkinitcpio.conf. Otherwise it's exactly the same as normal
<_`[m]> I get /lib/module errors bc it sets a suffix in the dirname of version -_-
<_`[m]> Jasper[m]1: except for the extra steps, but I guess that's why you use arch
<steev> _`[m]: well, an easier way to work around your issue would have been to simply match the version with the rest of your commands. e.g. you ran update-initramfs -c -k 6.4.3 but since it had installed 6.4.3+ you would do update-initramfs -c -k 6.4.3+
<steev> i'm not sure what you mean by it sets a suffix in the dirname of version
<Jasper[m]1> _`[m]: Normal for arch I mean hahaha
<_`[m]> it appends dunno what branch name?
<_`[m]> or did I make a typo and put +
<_`[m]> nah
<steev> i do not know, you only provided the previous output
<steev> in the previous output you had simply put "6.4.3" but it showed installing to "6.4.3+"
<agl7-x13s> clover[m]: I use Manjaro Linux on my production machine (x86_64) and I like to wait a bit, as with Manjaro, when software is tested out first before going to repo.
<_`[m]> and what's with these symlinks, is this a convention... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/phPpZaMBMtJnPAWmvPbuzfuu>)
<clover[m]> you poke at arch, but we've had latest steev kernel since two days ago muahaha
<_`[m]> steev: I just run `make ARCH=arm64 CROSS_COMPILE=aarch64-linux-gnu- LOCALVERSION="" laptop_defconfig`
<_`[m]> s/LOCALVERSION=""//
<_`[m]> s/LOCALVERSION=""//
<steev> i do not know how suse does their kernel packaging so i can't answer what the symlinks are or why they do it that way
<_`[m]> it's ubuntu 🙂
<_`[m]> but that sounds like it's not a convention, but I don't see the use
<_`[m]> * the use to replicate such
<steev> xnox: ^^
<steev> are you not just using the bindeb-pkg target?
<steev> should build you a working kernel deb that you can install and the ubuntu stuff should pick things up from there
<_`[m]> I end up with Image
<_`[m]> that sounds sexy though 🙂
<steev> `make ARCH=arm64 CROSS_COMPILE=aarch64-linux-gnu- LOCALVERSION="" laptop_defconfig` and then `make ARCH=arm64 CROSS_COMPILE=aarch64-linux-gnu- LOCALVERSION="" bindeb-pkg`
<steev> the only thing that doesn't work properly there, is the header package is junk
<steev> but if you aren't installing external modules, you should be fine
<clover[m]> aren't those needed for FDE?
<steev> external modules? no
<steev> headers? again, no
<steev> unless you're doing zfs for your rootfs, then, yes, but you get to pick up the pieces every time zfs breaks on a new kernel
<steev> which, from my experience is, every bump
<travmurav[m]> so for the last little while I was occasionally taking stabs at understanding how Hyper-V takes over el2
<travmurav[m]> tl;dr it's vendor lock-in by design
<travmurav[m]> but for more details, I wrote a bit on how that stuff works here: https://github.com/TravMurav/Qcom-Secure-Launch
<travmurav[m]> wasted so much time on this... better spent it looking for a job or something :/
<steev> i appreciate your research into it, will give it a read
<clover[m]> ugh, thats bad news. nice write up though
<steev> just gotta steal microsoft's keys, i guess
<konradybcio> you could probably find a use case for loading hyper-v on linux
<travmurav[m]> there was some abandoned hyper-v "host" support for Linux, x86 only
<travmurav[m]> but I think here you either figure out how to coerce tcblauncher to error out into user code or buy a windows pro license xD
<konradybcio> I've heard bird sing that windows has this nice chain of reaction to loading hyperv.. it then destroys the qualcomm smmu driver and loads a different one
<clover[m]> HdkR: do you have a link to shop for the nvme you are using?
<Jasper[m]1> <clover[m]> "HdkR: do you have a link to shop..." <- For x13s?
<clover[m]> yeah, they updated to a 2 TB drive or something, i wanted to know what brand / type
<clover[m]> s/updated/upgraded/
agl7-x13s has quit [Quit: Leaving]
<konradybcio> i wonder if the power consumption is much different
<Jasper[m]1> clover[m]: Any m.2 2230 drive would be good I guess, at least that's what I'm going for
<Jasper[m]1> Waiting for a small extender from 2230 to 2242 from ali before I order one though
<_`[m]> <steev> "the only thing that doesn't work..." <- ah yeah it's not in the tree with the licenses
<_`[m]> no lazy iommu?
<Jasper[m]1> Btw what's up with 6.5? Any reason to install it over 6.4 currently? Or doesn't it matter since all outstanding patches are in both anyways?
<clover[m]> right now 6.4 is fine, i'll probably install 6.5-rc2 once steev pushes to get this stuff:... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/BbRluHVqYCQbjryVmBXiCEKD>)
sally_ is now known as sally
<Jasper[m]1> gpr thing is already in there
<Jasper[m]1> Don't know about the rest though
<clover[m]> ah ok
<Jasper[m]1> But okay, guess I'll wait a bit
sally_ has joined #aarch64-laptops
sally has quit [Ping timeout: 480 seconds]
sally_ is now known as sally
sally is now known as Sally1
Sally1 has quit []
Sally1 has joined #aarch64-laptops
Sally1 is now known as sally
<hexdump0815> travmurav[m]: thanks a lot for sharing your qcom el2 research
agl7-x13s has joined #aarch64-laptops
sally is now known as Sally1
Sally1 is now known as sally
sally is now known as Sally1
Sally1 is now known as sally
agl7-x13s has quit [Quit: Leaving]
<steev> oh, yeah, i should push that
<steev> i'm testing it currently
<steev> clover[m]: lenovo-x13s-v6.5-rc
<steev> clover[m]: lenovo-x13s-v6.5-rc2
<steev> _`[m]: no, i don't have that enabled in the "laptop_defconfig" - the best way to get something fixed is to have someone be annoyed by it enough that they roll up their sleeves and get to work ;)
<steev> fwiw, i run with `pd_ignore_unused clk_ignore_unused net.ifnames=0 iommu.passthrough=0 iommu.strict=0` in my kernel command line config though
<clover[m]> <steev> "clover: lenovo-x13s-v6.5-rc2" <- pushed for arch folx, dont have to wait anymore Jasper
<steev> what were we waiting on?
<clover[m]> jasper was interested in 6.5-rc2
<steev> oh
<steev> the main thing about 6.5 is that none of the *base* stuff is out of tree
<steev> still various fixes, but, it's all in the tree, which is a nice step
<qzed> jhovold: sry, haven't forgotten your reviews, life has been busy, Thursday maybe
<konradybcio> qzed: addressing review comments mid-house-move like https://memy.memsekcja.pl/images/mSwryhT.500.mem.jpeg
<steev> that's me every day
<qzed> Yeah I have no idea what that means but I have barely managed to get a stable and internet connection last week😕 need to get a washing now, keep social life somewhat in check to not be a complete outcast and then do reviews...
<qzed> *washing machine
<qzed> I thought I would get there faster... Sry