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
<adrian_>
Is there any NDA'ed datasheets for the SoC components like the clock, reset, TLMM bits?
<adrian_>
or am I gonna have to poke at the linux source? :P
<freekurt[m]>
adrian_: I don't know the answer to that question, but just wanted to say thanks for doing that work.
<adrian_>
thanks
<adrian_>
yeah I .. ugh, know too much about QCA and snapdragon SoCs. :p
jayreap57[m] has joined #aarch64-laptops
rz_ has joined #aarch64-laptops
<agl>
steev: Did you patch the newest kernel (6.13.2) for the x13s and have you forgotten to push it on your web site?
rz has quit [Ping timeout: 480 seconds]
<agl>
steev: Have you now the T14s?
jayreap57[m] has quit [Excess Flood]
<steev>
yes i got the t14s, i haven't done anything with it yet (working on work things)
<steev>
i don't know if i did 6.13.2 kernel yet, is there something in it that you need?
<robclark>
hmm, all of a sudden I've got no wifi.. ath12k_pci is failing to probe, `of_irq_parse_pci: failed with rc=134`.. does that ring a bell for anyone? Kernel hasn't really changed
<robclark>
that error seems to be `PCIBIOS_DEVICE_NOT_FOUND`
<robclark>
which.. we shouldn't really get there if pci_device_to_OF_node() succeeds.. I think
<agl>
steev: I don't know if I need any patch exactly. I just installed your kernels because they always worked. I am a user who only uses a USB-C box (from ACASIS) with different interfaces and two SSD, mouse and Ethernet interface and turns on the "RNDIS_HOST" module (which automatically turns on the whole USB-C framework in menuconfig). The RNDIS_HOST module allows me to do USB thethering with my cell phone via USB-C. I don't use external displays yet,
<agl>
but I might want to buy one in the near future. Otherwise, I work directly on the x13s or via Rustdesk when I establish a connection from the clubhouse to home via VPN/Wireguard (Pi-VPN), for example. I'm usually on IRC via the x13s. Other than programming in my Z80 emulator (https://yaze-ag.de) and otherwise programming/compiling on Debian level (e.g. palemoon) I do not use the x13s.
<steev>
i just built 6.13.2 but as i'm waiting on a work related compile, i can't reboot to test it yet
<steev>
did have to drop bamse's usb flattening work from it, but that's okay, i believe there were changes requested
<steev>
or maybe i was still using the old ones
<steev>
ah no, its a patch that was backported
jglathe_ has quit [Remote host closed the connection]
jglathe_ has joined #aarch64-laptops
<clover[m]>
steev: initial thoughts on t14s?
<steev>
the fn and ctrl keys are swapped from the x13s :(
<steev>
the amount of times i hit fn+t to open a new tab is too damn high
<clover[m]>
lol
<steev>
but we're about to be pushing towards release so i haven't used it except the windows (and wsl)
<steev>
build quality is solid as well
<clover[m]>
is it fast?
<steev>
it does seem to beat the X13s for the kernel compiles (but i'm building 6.6 kernels, not the thinkpad kernels)
<clover[m]>
hows the screen?
<steev>
i got the oled, and i really like it
<clover[m]>
cool
<steev>
i also wanted to try out this anythingllm app because supposedly the qualcomm npu can be used from it, but it only ever says its booting up and/or to restart
tobhe_ has joined #aarch64-laptops
tobhe has quit [Ping timeout: 480 seconds]
hexdump0815 has quit [Ping timeout: 480 seconds]
<steev>
well, that was... a thinng
<steev>
i couldn't figure out how to download a different model and the model it uses, the training was 2021 so, nah
<steev>
agl: pushed
<steev>
also the fuck for anythingllm was that i had to run it as admin, and i'm just not comfortable with that
<steev>
the fix*
<JensGlathe[m]>
heh
<steev>
sorry, these days, wtf is my default reply to *waves hands around as an american*
<pengyu[m]>
<Jasper[m]> "Where did you get it from if I..." <- A used pne from a rich girl, with brand new pen and keyboard, SSD power hours is just 100h. She said she(or her families) has so many devices that she spent less time on a single device, this one was idle for a long time, lmao. She might purchase this one from Kickstarter.
<Jasper[m]>
Ahhh well they're out of stock everywhere so that's a shame
<pengyu[m]>
BTW, this device is pretty good on hardware, I disabled hibernation, it costs 2% for a 9 hours' s0 standby. As for software, it sucks. And there is no driver update from windows update(verified with UUPdownload).
<pengyu[m]>
<Jasper[m]> "Ahhh well they're out of stock..." <- Yeah, I think the company is/was bankrupt, the founder seems to be the former founder of Xiaomi laptop.
<pengyu[m]>
* s0 standby(they said they optimize the pcb to lower idle power). As
icecream95 has quit [Ping timeout: 480 seconds]
srinik has joined #aarch64-laptops
pstef has joined #aarch64-laptops
pstef_ has quit [Ping timeout: 480 seconds]
srini_ has joined #aarch64-laptops
srinik has quit [Remote host closed the connection]
craftyguy has quit [Remote host closed the connection]
craftyguy has joined #aarch64-laptops
craftyguy has quit [Remote host closed the connection]
craftyguy has joined #aarch64-laptops
<jhovold>
robclark: I've seen that ath12k error a couple of times too with 6.13, didn't have time to investigate so just made a note of it
<agl>
steev: I have compiled and installed your kernel 6.13.2. I'am at this Moment in the Clubhouse. Later when I'am at home I will test the rest.
<robclark>
jhovold: so, there is a "pci17cb,1107" in the 7x dts.. so AFAIU pci_device_to_OF_node() should find a dt node?
SpieringsAE has joined #aarch64-laptops
<SpieringsAE>
robclark: anthony25: adding those render options to disable explicit sync seems to have helped here too, I can actually use halloy on hyprland now lol
SpieringsAE has quit [Remote host closed the connection]
<jhovold>
robclark: yeah, it's odd indeed, but that rc=134 appears to come from pci_read_config_byte()
<anthony25>
robclark: thank you so much!
<jhovold>
robclark: nevermind, there's no interrupt specifier in that node so pci_read_config_byte() is being called
<jhovold>
smells like another pwrseq regression
<jhovold>
or just pci
<robclark>
jhovold: yeah, I think the problem is that it even calls pci_read_config_byte() in the first place, rather then of_irq_parse_one(). (But I'm just assuming I know how this is _supposed_ to work ;-))
<maz>
but if that's a PCI device, what's do you call of_irq_parse_one() on? there shoudn't be any node for a PCI device.
<maz>
what is this "pci17cb,1107" describing?
<maz>
if the device is using INTx signalling, the PCI read is the correct way to do it (will return the virtual INTx line, and that gets used by the PCIe RC).
<robclark>
maz, see the pci_device_to_OF_node() case in of_irq_parse_pci()... it seems like some kinda "partial" pci device? But I'm not sure how this is supposed to work... there is something in dts, but presumably it is still a normal pci device?
<robclark>
that said, the dts node specifies some regulators, but not an irq
<robclark>
so maybe it is supposed to go the pci_read_config_byte() route
<anthony25>
robclark: so if I understand correctly, a compositor should disable explicit sync if the driver doesn't support DRM_CAP_SYNCOBJ_TIMELINE?
<anthony25>
(which would explain why it is disabled in Gnome and Sway)
<robclark>
tbh, I don't know why the newer protocol when the syncobj timeline approach... seems kinda like solving an imaginary problem, but wtvr
<anthony25>
and why doesn't msm explicitely support syncobj_timeline if just adding it in the capabilities work?
<anthony25>
(not saying that it's just about exposing the capability, more that the driver seems to be supporting it already)
<robclark>
anthony25: the kernel hadn't exposed the cap yet because that would break old mesa (prior to a bit over a yr ago when I fixed some timeline syncobj issues in turnip)
<anthony25>
oh ok
<robclark>
maybe we've waited enough time, and should just go ahead and enable it now.. but would be good to get some more testing
<anthony25>
I'll leave it enabled then :p
<adrian_>
hm, is there a dev board equivalent for the snapdragon X1 SoC ? that's in this here thinkpad I'm doing bring-up on?
<adrian_>
(so i can get easier access to peripherals to stick scopes on and such)
<JensGlathe[m]>
adrian_: what thinkpad is it
<adrian_>
t14s gen6
<adrian_>
the X1E-78-100 cpu
<JensGlathe[m]>
bringup? There's a dtb for this already, or is it x1p42100?
<HdkR>
adrian_: Sounds like what the "Snapdragon Dev Kit" was supposed to be
<adrian_>
oh freebsd is booting on it already
<adrian_>
minus all the SoC dts peripherals
<adrian_>
I'm gonna write / port them
<adrian_>
it's easier to do that on a dev kit versus a laptop :P
jhovold has quit [Quit: WeeChat 4.4.3]
<adrian_>
HdkR: I have a LOT of older snapdragon dev kits at work and personally, hence why I'm asking if there is one for this chipset. :P
<JensGlathe[m]>
ok. There is the Dev Kit, and several reference designs
<adrian_>
that's publicly available
<adrian_>
ah what should I search for?
<HdkR>
adrian_: Sadly the dev kit didn't really ship
<adrian_>
(background: i'm adrian@freebsd.org, i worked at atheros/QC for a couple years, and I work at Meta on their snapdragon products, so I technically know the chipsets)
<adrian_>
:/
<adrian_>
ugh did they ever partner with someone for a reference design I can buy as an individual?