ChanServ changed the topic of #asahi-dev to: Asahi Linux: porting Linux to Apple Silicon macs | General development | GitHub: https://alx.sh/g | Wiki: https://alx.sh/w | Logs: https://alx.sh/l/asahi-dev
kdrag0n[m] has joined #asahi-dev
unevenrhombus[m] has joined #asahi-dev
casperes1996[m] has joined #asahi-dev
henje[m] has joined #asahi-dev
codegosu[m] has joined #asahi-dev
BrenoMaia[m] has joined #asahi-dev
exhan[m] has joined #asahi-dev
brstream[m] has joined #asahi-dev
IanPlatt[m] has joined #asahi-dev
the_real_briel[m] has joined #asahi-dev
katatafjsh[m] has joined #asahi-dev
JasonAntwi-Appah[m] has joined #asahi-dev
mnc7[m] has joined #asahi-dev
AnushervonTabarov[m] has joined #asahi-dev
RianSouzaSantos[m] has joined #asahi-dev
huddy987[m] has joined #asahi-dev
luxio_39[m] has joined #asahi-dev
izzyisles[m] has joined #asahi-dev
DanStrong[m] has joined #asahi-dev
nametable[m] has joined #asahi-dev
meowmomo1[m] has joined #asahi-dev
pedrohos[m] has joined #asahi-dev
phiologe has joined #asahi-dev
PhilippvK has quit [Ping timeout: 480 seconds]
yuyichao_ has joined #asahi-dev
arnidg[m] has joined #asahi-dev
cadawerum1[m] has joined #asahi-dev
rkjnsn has joined #asahi-dev
AntonioSchettini[m] has joined #asahi-dev
kedde[m] has joined #asahi-dev
butterfly[m] has joined #asahi-dev
minkuu[m] has joined #asahi-dev
ryanhrob1[m] has joined #asahi-dev
alicela1n has joined #asahi-dev
digitalfx[m] has joined #asahi-dev
xf00[m] has joined #asahi-dev
nihilistpc[m] has joined #asahi-dev
steffen[m] has joined #asahi-dev
rohin[m] has joined #asahi-dev
Liam[m] has joined #asahi-dev
happy-dude[m] has joined #asahi-dev
Tom[m]1234 has joined #asahi-dev
kov has quit [Quit: Coyote finally caught me]
ah-[m] has joined #asahi-dev
virmos[m] has joined #asahi-dev
ey3ball[m] has joined #asahi-dev
ogimgd[m] has joined #asahi-dev
unrelentingtech has joined #asahi-dev
boardwalk has quit [Quit: Ping timeout (120 seconds)]
boardwalk has joined #asahi-dev
ivy_[m] has joined #asahi-dev
user974[m] has joined #asahi-dev
lowz[m] has joined #asahi-dev
M0x8FFshulkk[m] has joined #asahi-dev
eeao^ has joined #asahi-dev
the_lanetly_052__ has joined #asahi-dev
roxiun[m] has left #asahi-dev [#asahi-dev]
jluthra has quit [Remote host closed the connection]
jluthra has joined #asahi-dev
akamizu[m] has joined #asahi-dev
<dottedmag> FWIW, what needs to be done in KMS driver so that compositors can handle it (TL;DR: provide fake EDID, as it ended up in uABI): https://paste.debian.net/plain/1240762
<dottedmag> marcan: Does display controller really hide EDID from the CPU? I remember you talked about it, but I may have missed something in meantime
* maz watches the Studio updating macos at a glacial speed... thank fsck this is the last time I'll see that.
<maz> (update failed... good stuff)
<_jannau_> dottedmag: weston / kwin work without problems with the WIP dcp drm/kms driver
<j`ey> maz: you jinxed it
<_jannau_> I'm not sure the data is completely hidden, we might just not yet found it
<maz> j`ey: I'll put it out in the bin... ;-)
<j`ey> maz: I run a bin collection company, whats your address? :P
<_jannau_> dottedmag: the dcp co-processor reports modes which we report to userspace
<psydroid[m]> I see your experience updating macOS on Apple hardware is about as good as mine in QEMU :)
<_jannau_> dottedmag: I think dcp reports all necessary information. make, model and serial number are reported as properties (not parsed atm). Timing and colorimetry modes as well
Lucy[m] has joined #asahi-dev
bisko has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<maz> _jannau_: what's the best way to get started on the t6002? the "old" way (install m1n1 by hand and bootstrap the whole thing), or the Asahi installer?
<dottedmag> _jannau_: yeah, shouldn't be too hard to expose. It is unfortunate that it has to be packed into fake EDID though
the_lanetly_052___ has joined #asahi-dev
<_jannau_> maz: the installer needs trivial changes to work and an updated m1n1 checkout
<_jannau_> so you would have to build it yourself
<maz> building m1n1 is easy. it's the installer itself I'm unsure of.
the_lanetly_052__ has quit [Ping timeout: 480 seconds]
<_jannau_> building the installer itself is not hard either. It builds however m1n1 with chainloading which requires bleeding edge rust
<j`ey> the main thing the installer does is make the partition, but if youre not going to use macOS (dual boot) and you're not using the internal disk, maybe it's not so useful? it does firmware stuff, but you can do that manually too
<jannau> maz: with https://paste.debian.net/1240766/ the stand-alone m1n1 installation should work
<jannau> installer builds with ./build.sh, you have to adjust the host in scripts/bootstrap.sh
<kettenis> or you just copy installer.tar.gz and installer_data.json yourself, untar it and run install.sh
ricekot has joined #asahi-dev
<maz> kettenis: I guess you then need to replace m1n1 in-situ, don't you?
<_jannau_> maz: the self built installer will be fine if you update the submodule to current HEAD
kov has joined #asahi-dev
c10l1 has quit []
c10l1 has joined #asahi-dev
c10l1 is now known as c10l
systwi_ has joined #asahi-dev
systwi has quit [Ping timeout: 480 seconds]
eeao^ has quit [Remote host closed the connection]
bisko has joined #asahi-dev
rdltromw^ has joined #asahi-dev
<maz> _jannau_: does the installer rely on the system rust toolchain? or does it bring its own? I get some funky compilation errors...
<_jannau_> maz: it uses the system one and requires rust nightly
<_jannau_> if marcan doesn't beat me to it I'll prepare a PR for mac studio support in the installer
<_jannau_> but if you plan to boot via usb you can build without chainloading/rust
<maz> _jannau_: I may do that, as I seem to be doing something wrong (I still get the same error despite trying to force the nightly rust).
<j`ey> can you paste the error?
<maz> 'rustup target add aarch64-unknown-none-softfloat' is what I was missing.
<mps> maz: this worked for me about two months ago
<mps> strange it doesn't work now
<maz> mps: as I said above, I didn't have the right toolchain
<mps> ah, ok and sorry
yuyichao_ has quit [Ping timeout: 480 seconds]
yuyichao has joined #asahi-dev
ChaosPrincess has quit [Quit: WeeChat 3.5]
ChaosPrincess has joined #asahi-dev
<maz> hoohoo! I'm at the point where I'm now feeding the box the wrong secondary m1n1/u-boot/DT combo (it is the one from the mini setup).
<maz> serial and reset working.
<_jannau_> maz: working dts is in https://github.com/jannau/linux/tree/asahi-t6002 based on the asahi downstream branch
<maz> _jannau_: yup, spotted it, as well as your u-boot tree.
<_jannau_> u-boot supports t6002 since https://source.denx.de/u-boot/u-boot/-/commit/585fc1c8c535380e1ace681e34bb1e3c9e8bc7ae but master might be still broken
the_lanetly_052___ has quit [Ping timeout: 480 seconds]
TheLink has quit [Quit: ZNC 1.8.2 - https://znc.in]
TheLink has joined #asahi-dev
<maz> _jannau_: OK, I'm in u-boot now. does it need any convincing to probe mass-storage on USB (the C ports, I know the A ports don't work)?
<tpw_rules> no
<tpw_rules> well i can't speak for mac studio. but it should tell you how many of what kind of usb devices it found
<maz> exactly none.
<arnd> maz: have you tried just booting a kernel in a m1n1 vm? that's how I'm still booting at the moment, skipping the u-boot step
<_jannau_> maz: does it probe the 6 usb-c ports?
<maz> arnd: no, I really want to run the full thing. running under the hypervisor isn't an option, for obvious reasons.
<maz> _jannau_: it does probe all 6 ports, but fails to detect anything.
<jannau> maz: ^^^
<jannau> s/might/probably/
<maz> jannau: yeah, looks suspiciously relevant.
<maz> jannau: yup fixes it.
<maz> amusingly, I just had the recovery jumping in and telling me that my "macos" install was corrupted, or something similar.
<maz> had to reselect the m1n1 proxy.
<maz> I guess it doesn't like me resetting the box the hard way... :-/
<_jannau_> yes, iboot does that now if a boot counter is not reset
<arnd> maz: I meant just to get you started with setting up user space. Once it works in the VM, you can obviously boot the same kernel natively from proxyclient, and then move to u-boot the last step. I found the vm proxyclient to be super helpful just getting started when not following the normal install
arti1208 has joined #asahi-dev
<maz> arnd: oh, I have a full userspace ready to be plugged in (my Mini is the donor system). I used proxyclient until about 8 months ago, at which point I switched to u-boot. I now have u-boot correctly probing USB, so it is only a matter of building the kernel, dumping it on the Mini's FS, and moving the disk over.
<maz> _jannau_: I'm really tempted to stick that counter reset in u-boot itself! :D
nicolas17 has joined #asahi-dev
<arnd> maz: makes sense. I wasted a ton of time just finding the right kernel config. I had started with arm64 defconfig and then turned on stuff one option at a time, which meant a lot of reboots until I had something working ;-)
<arnd> I still want to eventually turn on all those options in mainline, so I can just build a defconfig kernel and not think about it much, but there is not much point until enough drivers are there
<sven> _jannau_: wait, if I have a newer iboot I can no longer forcibly reset this thing because I’m too lazy to issue a real reboot?
<_jannau_> sven: linux resets the counters on boot
<sven> I’m not running the asahi branch
<sven> and even if I was right now there are times where it serrors very early because if this tbt mess
<sven> *of
<_jannau_> then you will have to reselect the m1n1 partition every ~10 resets as startup disk
<sven> *sigh*
<sven> guess I’ll be careful to not update iboot then
<maz> this is going to be super annoying, as the box will be in a rack with no way to interact with it other than the serial console...
<_jannau_> I was tempted to add the boot_error_count reset optionally to m1n1
<maz> +1.
<sven> I’d very much like that
<maz> 100% in favor.
<maz> just make it optional.
<sven> yes
<nicolas17> add a remote actuator for the power button like AWS did for their Mac EC2 instances :D
<maz> nicolas17: for a number of reasons, I don't want to do *anything* like AWS does... :-/
<nicolas17> lol
<_jannau_> it would not help at all with the problem
<nicolas17> ah you'd need a full *graphical* console to select the boot partition again?
<_jannau_> yes, it boots into macos recovery
<maz> _jannau_: do you have to select dual mode for DWC3 to work? or host? I have the former in my kernel config, and the ports don't seem to get detected...
<jannau> maz: I think so, host breaks iirc due to typec interaction
<maz> jannau: OK, so I'm missing something else.
<jannau> which kernel? if upstream it's missing support for the iommu variant? otherwise do you have CONFIG_TYPEC_TPS6598X ?
<maz> jannau: using your t6002 branch. I have this module, but the controllers themselves don't seem to appear, and it is like the driver never really probes them.
<maz> feels like a missing dependency.
<jannau> i2c
<maz> CONFIG_I2C_APPLE=y
<maz> and I see a bunch of cd321x devices hanging off the i2c bus.
<jannau> maz: my dev config if you want to compare https://www.jannau.net/asahi-t6002.config
<jannau> cd321x is apple's TPS6598X
<j`ey> jannau: what's different between that and a normal m1 config?
<jannau> it has most stuff built-in for booting via m1n1
<j`ey> well I mean that does the ultra need any extra options?
<jannau> no
<jannau> the asahi distro kernel works with the t6002 dts out of the box
<maz> I have the ugly feeling that I'm missing modules in my initrd.
<maz> they probably are on the mass storage, but since I made the initrd on the Mini, tons of things may be missing from there.
<jannau> hw doesn't differ that much. ethernet and sd card reader are the only things I can think of which use separate module
<maz> my mini uses a DT that doesn't know about the DWC3 ports at all.
<maz> so there is no reason for the initramfs to end-up with these modules...
arti1208 has quit [Quit: Konversation terminated!]
<maz> yup, definitely a module dependency issue. with jannau's config, I can boot by debian install all the way.
<maz> yup, that.
<maz> huh, the transition from USB3 speeds to USB2 is quite a let-dowm...
kit_ty_kate has quit [Quit: WeeChat 3.2.1]
kit_ty_kate has joined #asahi-dev
roxfan2 has quit [Remote host closed the connection]
<sven> if you don't mind hacks you can try https://github.com/AsahiLinux/linux/tree/atc-WIP to get usb3 speeds back
<sven> hrm, although i dunno if someone already found the right fuses and build the devicetree for the ultra
<jannau> the ones from the m1 max should work https://github.com/AsahiLinux/linux/pull/22/files
<maz> ah, I can finally feel a bit of heat coming from the box! :D
roxfan has joined #asahi-dev
<maz> 4:17 for a full debian kernel. not bad. I think I'm limited by the I/O here. already 30% faster than my current box, with only a fraction of the CPUs.
<j`ey> nice!
nico_32_ has joined #asahi-dev
nico_32 has quit [Ping timeout: 480 seconds]
jakebot has quit [Quit: The Lounge - https://thelounge.chat]
rdltromw^ has quit [Remote host closed the connection]
Telvana2 has joined #asahi-dev
Telvana has quit [Ping timeout: 480 seconds]
yuyichao_ has joined #asahi-dev
yuyichao has quit [Ping timeout: 480 seconds]
jakebot has joined #asahi-dev
jakebot has quit []
jakebot has joined #asahi-dev
tanty has quit [Remote host closed the connection]
tanty has joined #asahi-dev
Telvana2 has quit [Ping timeout: 480 seconds]
ricekot has quit [Quit: Connection closed for inactivity]
<Eighth_Doctor> marcan, svenpeter: so I know that linux-5.18 is coming up in a couple of weeks, I was wondering if you folks would be so kind as to be willing to prepare a branch off the 5.18 tag with the remaining non-upstream Asahi stuff that exists so far?
<Eighth_Doctor> that way I can cherry-pick it all onto the Fedora kernel tree once we start our 5.18 rebase
<Eighth_Doctor> for the kernel-asahi package I'm working on
<j`ey> Eighth_Doctor: the current branch is based on 5.18rc4
<Eighth_Doctor> oh it is?
Telvana has joined #asahi-dev
<Eighth_Doctor> so it is (5.18-rc5) as of right now
<Eighth_Doctor> I should have just checked instead of asking a question that apparently already has an answer :)
<j`ey> err rc5. but yeah, I guess it wont be too hard to rebase onto the final release
<Eighth_Doctor> yup
Telvana has quit [Read error: Connection reset by peer]
Telvana has joined #asahi-dev
Telvana has quit [Read error: Connection reset by peer]
Telvana has joined #asahi-dev