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
sri has quit [Ping timeout: 480 seconds]
hipboi has joined #aarch64-laptops
ellyq__ has joined #aarch64-laptops
ellyq_ has quit [Ping timeout: 480 seconds]
tobhe_ has joined #aarch64-laptops
hipboi has quit [Quit: hipboi]
tobhe has quit [Ping timeout: 480 seconds]
hipboi has joined #aarch64-laptops
indy_ has joined #aarch64-laptops
indy has quit [Ping timeout: 480 seconds]
hipboi has quit [Quit: hipboi]
actrawave has joined #aarch64-laptops
actrawave has quit []
hexdump01 has quit [Ping timeout: 480 seconds]
<gwolf>
steev: I get the following in devices_deferred:
<gwolf>
aux_bridge.aux_bridge.0aux_bridge.aux_bridge: failed to acquire drm_bridge
hipboi has joined #aarch64-laptops
hipboi has quit []
alfredo has joined #aarch64-laptops
alfredo has quit [Remote host closed the connection]
hipboi has joined #aarch64-laptops
hipboi has quit []
jhovold has joined #aarch64-laptops
hipboi has joined #aarch64-laptops
DanaG has joined #aarch64-laptops
<DanaG>
Well, I gave in and got one of the Snapdragon X1 laptops (specifically, the Surface Pro 11 with 5G). I see a devicetree for Surface Laptop 7 on the kernel mailing list, but no such luck for the Surface Pro 11. The most I can get with Ubuntu's image (and an edited grub.cfg) is a boot to a console with nonworking keyboard.
<DanaG>
Incidentally, vulkan works decently (tested with Veloren) on Windows, and even works in WSL... but the "dzn" driver doesn't seem to have an arm32 or i386 build to make 32-bit games work under fex-emu.
<DanaG>
The DMI name for the board is "Microsoft Surface Pro with 5G, 11th Edition" (that's rather wordy). It's a Snapdragon X Plus, rather than Elite.
<DanaG>
correction: s/X1/X/ in my first message there.
<JensGlathe[m]>
Is there an UEFI upgrade for the Surface 11 5G? as macc24 showed it can be a source for dt information. internal kb support may not be far off then. Which dt are you using?
<DanaG>
I've tried a few: x1e78100-lenovo-thinkpad-t14s, x1e80100-crd, x1e80100-qcp, x1e80100-dell-xps13-9345
<JensGlathe[m]>
ok and which worked
<DanaG>
If I add break=top, I can get it not to blank out the screen, but none of the usb devices or the keyboard are recognized. Then it spews a bunch of errors about qnoc "sync_state() pending" and a bunch of errors about suppliers. I don't recall if external outputs work during boot; if they do, I'll have to try grabbing something using a pikvm.
<DanaG>
Each of them had about the same degree of brokenness, it felt like. None of them gave working USB, and until I added break=premount, they all seemed to make the backlight turn off.
<DanaG>
The oibaf ppa seems to enable dzn for arm64 and x86-64, but not arm(32) and i386.
<HdkR>
well, arm32 doesn't matter since the hardware doesn't support it
<HdkR>
and fex doesn't thunk 32-bit GL/Vulkan, so only what is in the fex rootfs matters there
<DanaG>
Anyway, it's a bit silly to run Windows games inside fex, inside WSL2, when I could just run them on Windows. But I just had to try, for some reason.
<DanaG>
I think I found the uefi .bin file; added it to that onedrive.
<HdkR>
Definitely a bit silly, unless you're trying to run a game that requires AVX, since XTA/Prism still doesn't support that :P
<DanaG>
I don't recall which partition the "firmware-vol" stuff was from, and the files are in some unknown binary format. There's also an "acpi.lzma" file that contains... not sure what about ACPI, might be a binary of the dsdt, or might be more than that.
<DanaG>
I wish there was a "dd" tool for uefi, that could write each of those partitions to a file on disk... because the OS doesn't seem to have access to them. Oddly, one of the firmware volumes has a "System Volume Information" directory on it. Was that added by running Windows, or by whoever built that image? Beats me.
hipboi has quit [Quit: hipboi]
<DanaG>
Binwalk of the firmware .bin file shows flattened device tree, but trying "binwalk -dd "Flattened device tree" <filename> doesn't seem to work.
<DanaG>
But binwalk for *tree* did seem to work.
rfs613 has quit [Ping timeout: 480 seconds]
<JensGlathe[m]>
the largest file contains i2c definitions with some info where devices are.
<DanaG>
The UEFI.bin file's embedded dtb files seem to have info about a bunch of things.
<DanaG>
I hope there'll be a way to get not just USB host mode, but also dual-role, peripheral mode.
alfredo has joined #aarch64-laptops
rfs613 has joined #aarch64-laptops
<JensGlathe[m]>
what do you want to do
<DanaG>
It would be cool to be able to use MTP, like you can with Steam Deck's dual-role controller.
<macc24>
<JensGlathe[m]> "Is there an UEFI upgrade for the..." <- i still think it's better to look at dsdt
<DanaG>
The "acpiview.txt" in my OneDrive link has hexdumps of all the tables, but those might be annoying to recombine.
hipboi has joined #aarch64-laptops
<steev>
gwolf: okay that looks fine
<steev>
i don't actually know what it is or why, but maybe lumag does
<DanaG>
ah, found it, there's an acpidump for Windows.
<steev>
gwolf: you could try stopping the alsa-restore service and removing the asound.state file and then rebooting (so it doesn't recreate the asound.state file with the current settings)
iivanov has joined #aarch64-laptops
<JensGlathe[m]>
you get the data with acpi, too, but afaiu its sort of easier for the GPIO interrupts used by the i2c devices.
<DanaG>
Okay, I dumped the acpi tables on the machine. Seems like they might be the same as the non-5G? The device upstream of the modem, the qmux, seems to be there on the non-5G too.
<DanaG>
Or rather, it's there in the dsdt code, no idea if it's there in Device Manager.
<DanaG>
Incidentally, if you need 10GbE on one of these things, a ConnectX-4 Lx (or likely anything that generation or newer) works out of the box in a Thunderbolt 3 enclosure.
<DanaG>
Another hidden thing on this device: there must be a serial UART somewhere, because there's an SPCR (Serial Port Console Redirection) table, but I wonder where it's connected? Through a USB-C debug accessory, or through some test points, maybe?
<HdkR>
Looking forward to when thunderbolt works in Linux
<macc24>
<DanaG> "Another hidden thing on this..." <- probably either bluetooth or debug uart
<JensGlathe[m]>
you often find uart21 as debug
<macc24>
microsoft seems to have used uart2 instead
<macc24>
on other surface laptops
<JensGlathe[m]>
MSFT 🤷♂️
<DanaG>
Is the debug UART the same thing as debug via the XHCI debug protocol? usbtreeview doesn't seem to show that capability on any hub ports.
<DanaG>
Then again, maybe usbtreeview doesn't support showing that specific attribute?
<DanaG>
usbtreeview is a third-party utility based on the first-party usbview.
hipboi has quit [Quit: hipboi]
<DanaG>
On my AMD desktop, a bunch of the controllers have multiple ports as debug-capable. The Qualcomm, nope, none.
alfredo has quit [Ping timeout: 480 seconds]
srinik has joined #aarch64-laptops
DanaG_ has joined #aarch64-laptops
hipboi has joined #aarch64-laptops
DanaG__ has joined #aarch64-laptops
alfredo has joined #aarch64-laptops
DanaG_ has quit [Ping timeout: 480 seconds]
<DanaG__>
i also still have my Surface Pro X with 5G, I wonder if that will be usable with Linux before the 11, or after?
<macc24>
<DanaG> "Is the debug UART the same thing..." <- no
<macc24>
debug uart is a literal serial port like in the 90's, but on 1v8
<macc24>
xhci debug protocol runs over xhci iirc
<maz>
and hopefully, EFI knows about the UART and can directly use it.
<DanaG__>
On Chromebooks, the AP UART leads to the CR50 chip, and is exposed as a USB endpoint over CCD. I wonder if Microsoft has some similar UART on the Surface Pro 11? The SPCR table seems to point the UART to irq 182, but neither irq 182 nor decimal 386 (hex 0x182) exist.
<DanaG__>
But maybe iasl is just parsing it wrong.
pneuhardt has joined #aarch64-laptops
DanaG__ has quit [Remote host closed the connection]
DanaG__ has joined #aarch64-laptops
<DanaG__>
Actually, the acpiview.txt dump from UEFI shell says the same number. It also mentions lots of errors, such as: ERROR: The cache line size must be between 16 and 2048 bytes on ARM Platforms.
DanaG__ has quit [Remote host closed the connection]
DanaG__ has joined #aarch64-laptops
DanaG__ has quit [Remote host closed the connection]
DanaG__ has joined #aarch64-laptops
DanaG_ has joined #aarch64-laptops
DanaG__ has quit [Ping timeout: 480 seconds]
DanaG_ has quit [Ping timeout: 480 seconds]
alfredo1 has joined #aarch64-laptops
alfredo has quit [Ping timeout: 480 seconds]
alfredo1 is now known as alfredo
DanaG has quit [Ping timeout: 480 seconds]
hipboi has quit [Quit: hipboi]
iivanov has quit [Remote host closed the connection]
alfredo has quit [Quit: alfredo]
hipboi has joined #aarch64-laptops
<abelvesa>
tobhe_: any reasons why the other dtbs are not added to this installer?
<tobhe_>
Oh we do add more device trees actually, we just haven't had a chance to test everything yet
<tobhe_>
and we need to add a few tweaks to flash-kernel and the livecd grub config to make things work seamlessly
<tobhe_>
afaik the slim 7x should work but there is a flutter bug breaking the graphical installer
<\[m]>
<HdkR> "Looking forward to when thunderb..." <- will it ever?
hipboi has quit [Quit: hipboi]
<HdkR>
\[m]: Why wouldn't it?
indy_ is now known as indy
<\[m]>
it's proprietary no?
<macc24>
usb4 isn't
<\[m]>
are those compatible?
<\[m]>
I don't even know
<macc24>
yes they are compatible
<macc24>
usb4 is based on thunderbolt spec
<\[m]>
I might not have paid attention but I think the t14s is the first device I've seen with usb4 😲
<macc24>
it's been around for quite some time, my 2022 laptop theoretically has usb4
<\[m]>
lulz - rm -rf /etc/intel-license.txt
<\[m]>
it's been around but I've never actually seen it in a laptop that entered my interest sphere
<\[m]>
I never got that slow adoption, manufacturers like to pay licenses? or maybe just long lived contracts or some other leverage - I guess when your cpu is not intel maybe you can opt out more easily?
iivanov has joined #aarch64-laptops
<travmurav[m]>
I wonder if it's diminishing returns kind of a situation: if you have 5gbps usb3.0, how much better, realistically, 40gbps usb4 is?
Jasper[m] has joined #aarch64-laptops
<Jasper[m]>
travmurav[m]: pcie
<Jasper[m]>
That's it
<travmurav[m]>
well yes, and that's niche usecase I think
<macc24>
Jasper[m]: how many people are even going to use that?
<travmurav[m]>
so I'm very much not surprised we don't see usb4 everywhere :D
<Jasper[m]>
travmurav[m]: If you want anything above 10gbps (not including overhead) you can use it
<travmurav[m]>
I myself don't even have almost any usb3.0 things lol
<Jasper[m]>
Fast storage, eGPU, 10gig+ ethernet
<travmurav[m]>
well yes but actually saturating 10gbps or storage beyond 5gbps is how often? xD
<macc24>
average consumer doesn't see the difference between 480mbps and 5gbps
<Jasper[m]>
travmurav[m]: > <@travmurav:matrix.org> well yes but actually saturating 10gbps or storage beyond 5gbps is how often? xD
<Jasper[m]>
>
<Jasper[m]>
pcie3 nvme drives basically
<Jasper[m]>
It's the difference between 1GB/s and 3.5
<travmurav[m]>
well yes, but still that implies the thing is saturated
<Jasper[m]>
travmurav[m]: The bus, not the drive
<\[m]>
it's not that I was seeing usb3.2 whatnot but tb3/4
<\[m]>
(are replies in matrix irc firendly I forgot)
<Jasper[m]>
Depends on the length of the reply chain hahahahaha
<\[m]>
users might not "see" it, but they will be affected and just accept, since "it's slow, I don't know" and move on 🙂
<\[m]>
@jasper
<\[m]>
ugh can't eidt either 😄
<Jasper[m]>
Both edits and replies will show up, just not good
<\[m]>
I know - so I try to avoid it
<Jasper[m]>
They'll be prepended by a > and only part of the message
<Jasper[m]>
edits will resend the message again with an asterisk in front of it
<macc24>
hi irc people
<macc24>
* by editing you can also send messages that are visible only to irc users :D
<macc24>
unless matrix folks see the edit history
<travmurav[m]>
edit history is visible, deletes are not
<travmurav[m]>
but the bridge won't tell anyone that you deleted the message
<travmurav[m]>
so it can be extra awkward xD
alfredo has joined #aarch64-laptops
hipboi has joined #aarch64-laptops
alfredo has quit [Read error: Connection reset by peer]
SpieringsAE has joined #aarch64-laptops
<SpieringsAE>
I guess quite a bit of that 40gbit/s would go to a high quality video link no? or is that added ontop?
<travmurav[m]>
if you mean displayport, it's an alternate mode anyway
<SpieringsAE>
but with a doc it can be in parallel no?
<SpieringsAE>
s/doc/dock
<travmurav[m]>
I think with alternate modes some lanes are for dp and some are for usb, not sure what usb4 does in that case, maybe if dp is encapsulated somehow
<Jasper[m]>
travmurav[m]: iirc yes
<Jasper[m]>
to the last part
<\[m]>
is the tb4 stuff depending on the usb c stuff irt kernel dev?
<travmurav[m]>
ah fun so in case of usb4 it's not raw lines but tunneling, right
<SpieringsAE>
I would guess so, that provides the most flexibility
<SpieringsAE>
from what I read both alt mode and tunneling are supported for DP
<SpieringsAE>
both seem to be mandatory in the spec
<SpieringsAE>
PCIe tunneling does seem to be optional though
<SpieringsAE>
tunneling is based on 1.4a (so I guess also lower ones?) and alt mode can go to DP 2.0
<jhovold>
Here's an updated wip branch for the X13s:
<travmurav[m]>
(well it definitely omits any non x things)
craftyguy has quit [Remote host closed the connection]
craftyguy has joined #aarch64-laptops
<\[m]>
Jasper - how doyou know?
<Jasper[m]>
Precedent
<JensGlathe[m]>
Re faster chip: x1e-78-100 is amazingly fast, any more is bought with more power consumption
<Jasper[m]>
Book Go, Book2 Go, the normal galaxy books
<\[m]>
maybe this time it's different suchmuchhope
<\[m]>
anyway I can't, I hate the keyboard layout
<travmurav[m]>
also isnt samsung is the only one that still uses ufs?
<\[m]>
contrary to what? nvme?
<travmurav[m]>
yes
<Jasper[m]>
travmurav[m]: from what I heard yeah
<travmurav[m]>
> 512GB (eUFS)
<travmurav[m]>
lol
<SpieringsAE>
jeah samsung storage is trash
<SpieringsAE>
it why I didnt want it
<SpieringsAE>
they wanted to give a like a free tablet along with it, but hell no
<travmurav[m]>
it's probably similar perf to a random single-chip ssd but can't replace it and probably extra mess for firmware on it
<SpieringsAE>
its very nice quality wise, but I want to do yocto stuff, I need a big ssd
<\[m]>
more etrash
<\[m]>
it's better to get smallest ssd and put one yo.urself no - at least for lenovo it doesn't void warranty
<SpieringsAE>
multi core performance is still pretty much just the same as the x1e-78-100 models, the gpu gets quite a bit more though
<macc24>
<Jasper[m]> "They also have some fun EC stuff..." <- "fun"
<macc24>
literally the first time i cried while programming was when poking at the ec in slim7x
<SpieringsAE>
the asus proart is also still missing, though I doubt many people with that one have a lot of interest in linux. Its also very overpriced IMO
<macc24>
<SpieringsAE> "multi core performance is..." <- multi core performance depends mostly on cooling nowadays
<SpieringsAE>
Jeff Geerlings devkit geekbench clang score is quite literally the same as the asus vivobook I had
<SpieringsAE>
I was hoping its less limited thermal/power budget would push it a bit but nope
<JensGlathe[m]>
No idea what they reserved for the GPU/NPU part
<\[m]>
it actually started, but the main menu is known for somehow being a real resource hog so it might already freeze there
<tobhe_>
abelvesa: yeah it is having problems today for some reason... will hopefully be back soon
tobhe_ is now known as tobhe
<\[m]>
when I got to ingame - got some funky graphics
<Jasper[m]>
Interesting that it booted at all
chrisl has quit [Ping timeout: 480 seconds]
<Jasper[m]>
Source 2 had some driver thing I keep forgetting the specifics about
<Jasper[m]>
Where it didn't work on the adreno in the sc8280xp
<\[m]>
dota2 has linux support since forever
<Jasper[m]>
Yes, but I'm sure there was a thing with Turnip that the sc8280xp didn't have enough of
<Jasper[m]>
So source 2 games didn't work
<Jasper[m]>
Might be a vulkan only thing?
<\[m]>
ok fck the taxes I'm getting one 😄
<macc24>
hey fun fact: you can mess up the speaker config on slim7x so bad that it doesn't get fixed even by physically removing the battery
<konradybcio>
macc24 windows?
<macc24>
still messed up
<konradybcio>
oh wow what do you think caused this?
<macc24>
poking stuff in alsamixer in hopes of finding the switch that would make it work better
<konradybcio>
Did it end with a literal bang? Do y think they're fried?
<macc24>
left woofer plays sound that sounds like garbled, right one is ok
<macc24>
tweeters don't work
<macc24>
it sounds like a combination of issues when using jos's ucm2 config and krzk's script
<\[m]>
t14s incoming 🙂
<tobhe>
32 or 64GB ?
<JosDehaes[m]>
on my yoga, sound in linux is also bad (even with direct aplay), but in windows ok, so speakers are fine, it's just some params are just not right yet
<\[m]>
32 - no choice
<\[m]>
also still a minor chance that I'll break this machine trying to reball it with the dev kit cpu 😄
hipboi has quit [Quit: hipboi]
<konradybcio>
macc it is very possible you fried one of them then.. Did you adjust analog gain?
<macc24>
:(
<robclark>
Jasper[m]: the issue was # of bindless descriptors.. x1/a7xx has more (but idk if we've enabled that yet)... there is a driconf override that you can use on the older gpus
<Jasper[m]>
Not sure if @[\] messed with that toogle
<Jasper[m]>
*toggle
<Jasper[m]>
They only have sc8280xp though, that's why I found it weird
<Jasper[m]>
@robclark is that only when using Vulkan combined with S2 or also OpenGL? IIRC they still support that
<robclark>
only for vk
<robclark>
gl uses the bindless descriptors in a different way, one per shader stage (so 5 is enough).. tu only exposes 4 because it needs one for it's own internal use
<macc24>
are i2c phys on x1e 3v3 tolerant?
<travmurav[m]>
I think for the longest time it's all 1v8
<travmurav[m]>
so the only 3v3 tolerant pins are probably the special sdc2 ones
<macc24>
sdmmc controller?
<travmurav[m]>
yes
<travmurav[m]>
the mmc has dedicated part in tlmm I think for 1v8/3v0 modes
<travmurav[m]>
but otherwise I think qcom didn't have any 3v3 tolerant io domains for at least a decade
<travmurav[m]>
as in, on paper they're not designed for 3v3, in practice it may survive
<gwolf>
steev: Well... My laptop seems to be convinced it is outputting some sound... I just have to disagree with it, as I cannot hear it yet.
iivanov has quit [Remote host closed the connection]
iivanov has joined #aarch64-laptops
<tobhe>
has anyone seen crashes with a blue screen on the XPS 13 before?
SpieringsAE has quit [Quit: Leaving]
<jhovold>
ok, turns out that x1e rc3 branch broke usb-c hotplugging, good news is that the fix also fixed coldplug orientation detection
<jhovold>
I've forced pushed an updated branch that addresses also a related NVMe regulator glitch during boot
<jhovold>
Here's an updated wip branch for x1e80100:
<hexdump0815>
travmurav[m]: looks like your assumption is correct and the x1p seems to boot just fine, i.e. seems to be very close to the x1e
<travmurav[m]>
oh
<travmurav[m]>
yeah that's nice to confirm indeed
craftyguy has quit [Remote host closed the connection]
<travmurav[m]>
but hm
<travmurav[m]>
I think surfaces are x1e-6x so partial-goods hamoa and not 8 core purwa
craftyguy has joined #aarch64-laptops
craftyguy has quit [Remote host closed the connection]
<travmurav[m]>
not sure if there is 4x surfaces
craftyguy has joined #aarch64-laptops
<hexdump0815>
travmurav[m]: i think the acpi talbes got uploaded as well iirc, so should be easy to check
<zeph[m]>
@tobhe nope, same... attempted online an with brtfs this time, instead of ext4... same
<zeph[m]>
s/online/offline
<Jasper[m]>
travmurav[m]: Only one I've seen is the thinkbook
<Jasper[m]>
And a book4 edge
<tobhe>
zeph[m]: could you upload the log in launchpad?
<zeph[m]>
hah, i reported via the tracking/bug report automated system... ok, let me get back there
<tobhe>
that should be enough, I'll find it :)
<zeph[m]>
cool
<hexdump0815>
there are many x1p devices coming up these days: thinkbook, dell inspiron 14, lenovo ideapad slim 5x, some asus i think as well etc.
<travmurav[m]>
hexdump0815: the ones on the onedrive say hamoa so it's 10 core I guess
<travmurav[m]>
and yeah, dumb naming scheme strikes back
<travmurav[m]>
again
<travmurav[m]>
there is x1e full chip, x1p binned x1e and x1p that is a different silicon :/
<hexdump0815>
ah ok - so no proof yet - right the 10 core one is called plus as well
<travmurav[m]>
and even the first digit is 8, 7, 6, 4
<travmurav[m]>
with 8 and 7 meaning x1e but different bin?
<travmurav[m]>
this is a mess
<travmurav[m]>
at least there is not-intended-to-use codenames "hamoa" and "purwa" but can we really use them beyond the circle of people who deeply knows the platform :/
<tobhe>
zeph[m]: I take that back, can't find it :( if it isn't too much of a hassle it would be nice if you could attach it to the existing bug
<zeph[m]>
ok, i give it a try :)
<zeph[m]>
let me grab a mouse or i go nuts (i hate the trackpad of this laptop)
bigio has quit [Quit: Page closed]
<zeph[m]>
ok, added
<zeph[m]>
hah, it says it cannot stat a .dts file (i could not spot it looking into the install window console)
<tobhe>
that is most certainly a typo :) i have an idea where
srinik has quit [Ping timeout: 480 seconds]
iivanov has joined #aarch64-laptops
<tobhe>
fix is in the works, might take a few hours to rebuild everything
ellyq has quit [Ping timeout: 480 seconds]
<zeph[m]>
cool
<zeph[m]>
also... i kept a partition free for the installation, but the tooling refuses to take it over to install alongside of windows... ideas? it expects an empty partition or?
<zeph[m]>
i have to give the partition as is, aim it as ext4, map as / ... but i miss the encryption
<zeph[m]>
(maybe i shall simply google aside, but since we started...)
<tobhe>
no idea how the new installer decides where to install. I too wish there was a way to get the more fancy features with manual partitioning...
<zeph[m]>
happy we stumbled on the same xD
SpieringsAE has joined #aarch64-laptops
<SpieringsAE>
hexdump0815: yeah asus has the proart, and there is a plus version of the vivobook which funnily enough is the exact same price as the elite version over here
<SpieringsAE>
I wonder if the plus could have a better battery life, I doubt it though, because otherwise, why would one ever buy the plus version with a smaller ssd for the same price?
<macc24>
my logic analyzer says this about the pin
<macc24>
first bunch of data is me pressing enter in systemd-boot
<macc24>
i'll need to hook up more channels tomorrow
SpieringsAE has quit [Remote host closed the connection]
chrisl has joined #aarch64-laptops
iivanov has quit [Remote host closed the connection]
<tobhe>
zeph[m]: new image is up
* whiskey9_
excitedly downloads it. I'm running into the exact same problems that zeph reports on the Ubuntu discourse, and was attempting to work around it by hand. So, thank you!!!!
<JosDehaes[m]>
I've had lots of sudden random reboots on my yoga 7x, since windows updated the firmware/bios. Not sure that is it, but I have it with several kernels, didn't before. Anyone else?
jhovold has quit [Ping timeout: 480 seconds]
<macc24>
which version are you on right now? can you try capping highest cpu frequency state?
<JosDehaes[m]>
how can I check (from linux)?
<JosDehaes[m]>
(running git fsck now)
<macc24>
dmidecode
<JosDehaes[m]>
Vendor: LENOVO
<JosDehaes[m]>
Version: NHCN53WW
<JosDehaes[m]>
Release Date: 08/02/2024
<robclark>
JosDehaes[m]: I'm on the same fw (since Monday).. no problem _so far_
<macc24>
i also haven't noticed any problems
<tobhe>
whiskey9_: thx, let me know how it went!
<nirik>
Is there anywhere to download it yet? :) I'm still on 06/22/2024... I guess I could swap windows nvme back in, but kinda a pain with all those clips.
<whiskey9_>
tobhe: So far so good! The installation completed fine and am booted into Ubuntu 24.10. I'm currently working to copy the firmware from Windows land over.
<whiskey9_>
tobhe: I think the only wierdness had to do with the installer and the lack of accellerated graphics. The installer was unusable (like, it was running, but you could only see the bottom left sixth of it.). Simply adjusting the display scaling from 200% to 100% solved that.
iivanov has joined #aarch64-laptops
DanaG has joined #aarch64-laptops
<JosDehaes[m]>
ok, it's apparently lazygit (opened on a kernel repo) that is triggering the reboots 😲
<robclark>
huh
<JosDehaes[m]>
it reads the entire git log to show the commits pane, not sure how it's able to crash the system though
iivanov has quit [Ping timeout: 480 seconds]
<robclark>
that sounds painful to use on kernel (or any large git tree)
<JosDehaes[m]>
yes, gitk also struggles with the kernel repo
chrisl has quit [Remote host closed the connection]
chrisl has joined #aarch64-laptops
<robclark>
is lazygit a curses type thing? I guess you could try running it from fbcon to rule out anything gpu related
<JosDehaes[m]>
yes, it's a terminal app, not sure if curses based
<JosDehaes[m]>
how can I try from fbcon?
<JosDehaes[m]>
disable the GPU driver?
<macc24>
nah, just ctrl+alt+f2, login as yourself and use it
<JosDehaes[m]>
also happens on text console
iivanov has joined #aarch64-laptops
<tobhe>
whiskey9_: have you tried qualcomm-firmware-extract?