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
laine has quit [Ping timeout: 480 seconds]
Ablu is now known as Guest9593
Ablu has joined #aarch64-laptops
<Dantheman825[m]> would anyone else happen to be running FEX on their X13s? It can run Linux games fine on Steam, but I've been having trouble getting Windows games to run
Guest9593 has quit [Ping timeout: 480 seconds]
<Dantheman825[m]> Sometimes the games will be stuck on starting (specifically on games that download VC_Redist or something DirectX related beforehand)
<Dantheman825[m]> The end result is pretty much the same
<Dantheman825[m]> And other games without those dependencies will say their running, but never actually open the game
<Dantheman825[m]> s/VC_Redist/VC\_Redist/, s/their/they're/
<robclark> Dantheman825[m]: I think you probably need https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25534 which might mean (would need to check git history) ToT mesa.. there might be some followon patches to set driconf defaults for specific games.. HdkR might know.. and/or ask in #freedreno
<Dantheman825[m]> hmm, it says this PR is merged
<Dantheman825[m]> I'm on 23.2.1 on both the host system and FEX rootfs
<Dantheman825[m]> should I just compile upstream Mesa?
<robclark> it might be in 23.3, not sure about that.. but probably not 23.2.x
<robclark> if you can use ToT/main mesa build, that is better
hexdump01 has joined #aarch64-laptops
hexdump0815 has quit [Ping timeout: 480 seconds]
<HdkR> We've slowly been trying to figure out the hangs on Discord but it's something I've never seeen before so it's pretty hard
<HdkR> Luckily that PR only affects Half-Life: Alyx, Aperture Desk Job, Counter-Strike 2, and X4: Foundations. Although they would need some driconf
<travmurav[m]> jkm: it was already sorted but you need rmtfs to boot the modem cpu and tqftpserv to let it "download" wifi firmware from the OS. I think pd-mapper is also used but it's not necessary for anything wrt battery/charging since it's handled by the EC on aspire1
<travmurav[m]> HdkR: you can (could?) run HL:A via fex?
<HdkR> travmurav[m]: It runs but it hits a Vulkan binding slot limitation with Mesa. So it needs mesa config option to override it. Then the additional problem is that the USB-C output can't drive an HMD since it can't run at full bandwidth
<HdkR> er, USB-C on the X13s can't run at full bandwidth
<HdkR> No idea what the perf is since I can't plug an HMD in to the thing :P
<travmurav[m]> heh, but it runs via emulated steam-vr or you use thunked monado?
<HdkR> emulated stamvr
<HdkR> steamvr*
<travmurav[m]> ah I see
<travmurav[m]> speaking of vr and vulkan... do you have any other vk+vr games that get in-game?
<HdkR> Beat Saber gets to the menu, but can't control it because no HMD again
<HdkR> SuperHot non-VR runs, so I would expect the VR version to also run
<HdkR> I don't think I tried any other VR games
<travmurav[m]> we were playing with trying to run unity apps built against oculus quest sdk in monado but it seems like there are some gpu artefacts between left and right eyes when running in vk, and we couldn't yet convince it to render gles
<travmurav[m]> seemed like some buffer built from previous eye "bleeds" into the next one, but I can't allow myself to blame mesa since my setup for this is beyond "unusual" lol
<HdkR> Ah, very fancy
<HdkR> `-novr` flag passed of course
<travmurav[m]> looks like it picks the wrong texture or something, the scenery cubes are also broken
<travmurav[m]> and I thiiiink there should be more stuff around...
<HdkR> novr mode makes everything a bit weird
<travmurav[m]> oh you mean it breaks on x86 too?
<HdkR> Yea, that's how it looks on x86 there. The music artist model is working but the rest looks the same
<HdkR> The cubes in the background are just...weird
<travmurav[m]> hm is this the latest game version?
<HdkR> Latest as of about three weeks ago. 2023-11-19
<HdkR> Nearly 300FPS on X13s though. Maybe once the kernel side catches up it can be a good experience :P
<travmurav[m]> this makes me wonder if the emulated vr tracking code won't be an issue with latency etc
<travmurav[m]> But this sounds really cool that it works
<travmurav[m]> HdkR: hm I guess this means that the unity/dotnet magic is fixed now in fex? I recall there was some problem before
<HdkR> Fixed-ish
<travmurav[m]> Cool
<HdkR> There might still be some cases where the garbage collector could get in trouble but that should get solved with time
<travmurav[m]> I guess the dotnet everything is "double vm" now? as in IL is interpreted/JIT'd to x86 just to be interpreted/JIT'd to arm again
<HdkR> Of course
<travmurav[m]> I see, and I guess no way to thunk for most games...
<HdkR> Would be a nightmare job that I wouldn't wish upon my worst enemy :P
<travmurav[m]> I see :D
* travmurav[m] tries to update his beatsaber install to the latest version from the legacy build to see if the weird cubes are their design choice somehow
<HdkR> Might be interesting, I didn't spin up my HMD to see it in a real environment. But it does look like random noise basically
<travmurav[m]> Eh seems like the promo design was either only for one version or it's very broken, I only see the "usual" UI
<travmurav[m]> I guess I could try to downgrade the depot...
<HdkR> hah, dang. I'll need to go back in and see if the rendering bug went away then.
<travmurav[m]> HdkR: eh, it /was/ a promo UI just for one version...
<travmurav[m]> <HdkR> "https://cdn.discordapp.com/..."; <- and this looks correct then
<HdkR> hah, managed to sneak in there and get the weirdo promo version :D
<travmurav[m]> HdkR: Presumably steam/proton supports passing through openxr so I wonder if you can thunk that and use monado with "simulated" hmd and "qwerty" controllers xD
<HdkR> Not sure, I've never used Monado
Caterpillar has joined #aarch64-laptops
<steev> _[m]1: iirc that's ema
<steev> ema: iirc, re: the patches being in next for qseecom, we should be able to put them in the debian 6.6 kernel (they're already in my branch so they definitely work fine with 6.6)
Caterpillar has quit [Ping timeout: 480 seconds]
jhovold has joined #aarch64-laptops
Caterpillar has joined #aarch64-laptops
<Jasper[m]> <HdkR> "Luckily that PR only affects..." <- Whoa, will Source 2 start working on A6XX or did I read the PR incorrectly?
<Jasper[m]> It starts by talking about A7XX, but ends with A6XX getting more descriptor sets too
<HdkR> Jasper[m]: Yea, if you set the driconf option then theoretically latest Source 2 games will work. Never tried it myself
<Jasper[m]> Interesting
<HdkR> Mostly because I haven't spent the time to figure out how to set these options
<HdkR> TU_DEBUG, easy. driconf? I dunno
<Jasper[m]> I might try it out if I ever get to it
<Jasper[m]> *try and figure it out
<cenunix[m]> <jhovold> "i suggest booting into a shell..." <- sorry for the late response but thx for the info, I believe you did tell me before yeah, have been finishing up finals it must have slipped my mind. Thanks for all the work on the wiki! has been immensely helpful
<jhovold> cenunix[m]: thanks, glad to hear you find it useful
<arnd> I wonder if Lenovo is starting to offload their X13s G1 stock in preparation for the newer model, or if this offer is a mistake, but €390 (16GB RAM, 512GB SSD) seems like a great deal either way in case someone is still looking for one https://www.kaufland.de/product/442638800/. Apparently includes free shipping to anywhere in Europe, UK and even USA
<Jasper[m]> arnd: Oh WOW
<Jasper[m]> yeah I mean assuming gen2 is coming soon™ it seems correct that they are
<jhovold> bamse: i tried the ucsi series but see a bunch of of timeout errors during boot, did you test it on the x13s before merging?
<ema> steev: excellent! Do you happen to know in which upstream kernel version they're expected to land?
<jhovold> arnd: sounds like a good deal indeed, with modem too
<steev> ema: assuming it's in next... probably 6.8
<jhovold> steev, ema: if you're talking about efivar support, that is already in 6.7-rc1
<steev> jhovold: fwiw, i don't see any timeouts here, but, i'm not sure i have anything that actually uses ucsi
<steev> jhovold: oh, even better, i was just going off them saying it was in -next :)
<jhovold> steev: you may also not have the driver enabled ;)
<steev> jhovold: no i do lol
<jhovold> ok, good :)
<steev> otherwise efivar -l wouldn't work :)
<steev> 77fa9abd-0359-4d32-bd60-28f4e78f784b-CurrentPolicy
<jhovold> i was talking about the ucsi driver, did you enable that before testing the series?
<steev> oh!
<steev> well, i feel silly
<jhovold> even without my ethernet addapter conencted during boot I see:
<jhovold> ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: GET_CONNECTOR_STATUS failed (-95)
<jhovold> so something's clearly not right
<steev> yeah, i see the same message
Caterpillar has quit [Ping timeout: 480 seconds]
<minecrell> arnd: that price is so ridiculous it feels like a mistake on their side
<minecrell> yeah they now removed the page and cancelled my order, oh well :p
<steev> dang
<arnd> ah, too bad. They are a fairly large online store in Germany, so one would hope that they at least honor your order on a pricing mistake the way that amazon does, but at least there should be no problem in getting an immediate refund
<minecrell> arnd: yeah they cancelled and refunded it within 2 minutes (I guess that's a coincidence, I was probably not the only one to try :p)
<Jasper[m]> <minecrell> "arnd: yeah they cancelled and..." <- It's still up for me interestingly
<Jasper[m]> Otherwise, bummer.
<arnd> I had one in the cart before it got removed. it came back a bit later and ordered one, hasn't gotten canceled yet
<arnd> when I first looked, they had a maximum order of 45 laptops. When I ordered it was at 58, now it's down to 40 again.
<minecrell> yeah I think it just keeps coming back whenever they cancel the orders :p
<arnd> right, and then once the stock sells out they have an actual person look at it and realize they made a mistake
<ardb> tastatur: deutsch :-(
<arnd> I got my refund now, plus a 5€ voucher. I guess that's something ;-)
<Jasper[m]> Free money
<ardb> the x13s is 1300 on lenovo's own website in NL
<minecrell> 1000€ voucher + price reset back to 1300€ would be acceptable :p
<arnd> ardb: for the 5€ I could buy an AZERTY sticker set for when it comes back
<ardb> just a screwdriver to swap the Y and the Z no?
<ardb> the germans are so close to having a normal keyboard ...
<ardb> is at least the enter key the normal shape?
<arnd> DE keyboards have the horizontal enter key unfortunately
<ardb> that's the one i prefer actually
<Jasper[m]> I'm still happy NL does the ANSI with euro key layout
<arnd> I meant vertical, I want the horizonly (ANSI) one
<ardb> yeah exactly - very annoying to hit \ every time you go for the enter key
<ardb> i am fighting internal IT to get a new laptop with sane keyboard layout, they already shipped me an AZERTY one once
<Jasper[m]> ardb: Get the Dutch layout
<ardb> it also has the weird enter key iirc
<ardb> but the x13s on lenovo.nl has the 'EU' english layout, not sure what that means tbh
<Jasper[m]> ardb: No we switched to QWERTY ANSI with a € under Ctrl+Alt+5 X amount of years ago
<arnd> I think that's an ISO layout like UK but with € instead of £
<Jasper[m]> NL doesn't use ISO
<Jasper[m]> Some manufacturers still ship them since we're an exception in the EU afaik (I talked to a Cooler Master employee about this)
<ardb> that explains why there is no NL option, only EU-english
<ardb> they're essentially the same thing
<Jasper[m]> ardb: Yes, the old ISO layout got slowly phased out
<arnd> Ok, I need to remember that for next time I buy a keyboard or a laptop and have the option then. I also realized that this is the layout I'm currently typing on, with a Cherry G80-3000LQCEU, where the final EU means "US plus € symbol"
<Jasper[m]> There are 90's keyboards that do have the "Dutch" layout with ISO
<ardb> yeah those are the ones i remember
<ardb> they have some accented e and i too iirc
<ema> jhovold, steev: even better! I'll give it a try, thank you very much
pbsds9 has quit []
pbsds has joined #aarch64-laptops
<steev> i know i'd love it if it made it into the 6.6 that is gonna make it into testing
<_[m]1> back to arch and reboot loop 👌
<_[m]1> also still seeing the trackpoint (I guess) pushing the mouse forcefully in a direction
<Dantheman825[m]> <robclark> "if you can use ToT/main mesa..." <- Apologies for such the late response, but could you elaborate on the ToT part? Is this a fork I'd be cloning?
<HdkR> Dantheman825[m]: They're meaning to try the current "top of tree" of mesa instead of a release
<Dantheman825[m]> so essentially compiling the latest commit?
<HdkR> Yes
<Dantheman825[m]> gotcha
<HdkR> That's what top of tree means in this context
<Dantheman825[m]> while I'm at it, I'll have to figure out compiling a thunk version too
djakov has quit [Remote host closed the connection]
djakov has joined #aarch64-laptops
<dianders> Hi! Is anyone (maybe bamse?) that uses the PWM on ti-sn65dsi86 able to give a Tested-by to Uwe's patch so I can land it? https://lore.kernel.org/r/20231127101547.734061-2-u.kleine-koenig@pengutronix.de
<travmurav[m]> dianders: does one need to apply the linked other patch before or just this one should be fine to test?
<dianders> travmurav[m]: Doesn't matter. The linked patch is a no-op cleanup, it would just cause a trivial conflict. I've already landed the linked patch in drm-misc, so depending on your tree or not you might need to resolve the trivial conflict.
<travmurav[m]> well, it breaks backlight for me...
<travmurav[m]> will report on the list
<dianders> travmurav[m]: OK, glad I didn't land it then. ;-)
srinik has quit [Killed (NickServ (Too many failed password attempts.))]
srinik has joined #aarch64-laptops
laine has joined #aarch64-laptops
<clover[m]> any update on camera support?
nerdboy_p has joined #aarch64-laptops
nerdboy_ has quit [Ping timeout: 480 seconds]
<steev> bryanodonoghue is the person to ask that :)
rz_ has quit [Remote host closed the connection]
rz has joined #aarch64-laptops
<ajhalaney[m]> the ucsi patches make me realize just how out of touch with anything newer than usb2 :/
<ajhalaney[m]> out of touch I am*
jkm has quit [Quit: leaving]
jhovold has quit [Ping timeout: 480 seconds]
nerdboy_ has joined #aarch64-laptops
nerdboy_p has quit [Ping timeout: 480 seconds]