#
…

2026-10-04

00:13maettu102 has quit [Ping timeout: 480 seconds]
00:15maettu102 has joined #asahi-dev
00:25hightower4 has joined #asahi-dev
00:32hightower3 has quit [Ping timeout: 480 seconds]
00:57jcbm has quit [Remote host closed the connection]
00:58jcbm has joined #asahi-dev
01:23linkmauve has quit [Remote host closed the connection]
01:26linkmauve has joined #asahi-dev
02:00yurix has quit [Remote host closed the connection]
02:00yurix has joined #asahi-dev
02:26jcbm has quit [Remote host closed the connection]
02:27jcbm has joined #asahi-dev
02:47bennofs3 has quit [Ping timeout: 480 seconds]
02:53tobhe_ has joined #asahi-dev
02:55jcbm has quit [Remote host closed the connection]
02:56jcbm has joined #asahi-dev
03:00tobhe has quit [Ping timeout: 480 seconds]
03:09ddxtanx has joined #asahi-dev
03:09bennofs3 has joined #asahi-dev
03:11gargantua has joined #asahi-dev
03:38gargantua has quit [Ping timeout: 480 seconds]
03:43nora_ has joined #asahi-dev
03:45nora has quit [Ping timeout: 480 seconds]
03:55catred has quit [Remote host closed the connection]
03:58bennofs3 has quit [Ping timeout: 480 seconds]
04:10bennofs3 has joined #asahi-dev
05:01chadmed has quit [Read error: Connection reset by peer]
05:01chadmed has joined #asahi-dev
05:18gargantua has joined #asahi-dev
05:18roxfan2 has joined #asahi-dev
05:24roxfan has quit [Ping timeout: 480 seconds]
05:25pb17 has quit [Ping timeout: 480 seconds]
05:25vimproved has quit [Remote host closed the connection]
05:26vimproved has joined #asahi-dev
05:35Yushi has quit [Remote host closed the connection]
05:36Yushi has joined #asahi-dev
05:41chadmed has quit [Quit: Konversation terminated!]
05:54betoissues has quit [Quit: bye bye.]
05:55betoissues has joined #asahi-dev
05:56betoissues has quit []
05:56betoissues has joined #asahi-dev
06:04pb17 has joined #asahi-dev
06:19JamesCalligeros[m]oh cool the robot found another useful devicetree mistake thats nice
06:19JamesCalligeros[m]definitely worth the worst el nino ever recorded
06:20JamesCalligeros[m]it still hasnt learnt about shared gpio though apparently
06:26sven has quit [Quit: The Lounge - https://thelounge.chat]
06:26sven has joined #asahi-dev
06:38mattia013 has quit [Quit: mattia013]
06:43betoissues has quit [Quit: bye bye.]
06:45betoissues has joined #asahi-dev
06:49betoissues has quit []
06:53betoissues has joined #asahi-dev
07:19voidah has joined #asahi-dev
07:28voidah has quit [Ping timeout: 480 seconds]
07:29lacamar[m] has joined #asahi-dev
07:29lacamar[m]Asahi australia conference would be sick
07:29lacamar[m]I’m no longer in Aus though
07:34bennofs3 has quit [Ping timeout: 480 seconds]
07:37tg473 has joined #asahi-dev
07:49bennofs3 has joined #asahi-dev
08:07okb has joined #asahi-dev
08:13gargantua has quit [Ping timeout: 480 seconds]
08:36ddxtanx has quit [Remote host closed the connection]
08:43hello0092 has joined #asahi-dev
08:51pb17 has quit [Ping timeout: 480 seconds]
08:54okb has quit [Ping timeout: 480 seconds]
08:58hightower3 has joined #asahi-dev
09:05hightower4 has quit [Ping timeout: 480 seconds]
09:12mini has quit [Quit: The Lounge - https://thelounge.chat]
09:12mini has joined #asahi-dev
09:25pb17 has joined #asahi-dev
09:43allusivedev has joined #asahi-dev
09:44allusivedevHello! I'd love to get involved with this project, I've got an M4 MacBook Pro and lots of software engineering experience though probs not as low level as this project is.
09:46chaos_princesshttps://asahilinux.org/docs/sw/tethered-boot/ and other m1n1 related pages should be a good start
09:46allusivedevHey thanks so much I'll take a look at that. This is my first time using IRC lol let me know if I'ved efd anything up
09:50sobek[m](Also, make sure to read the contributor, copyright and AI policy, Asahi work has to be extra careful in this regard not to get angry lawyers shutting down the projet, keep stuff clean for licenses and upstreaming, etc).
09:51chaos_princessThe crash course to irc is - set up join/part hiding, do not paste multi-line things directly into irc and use a pastebin.
09:51allusivedevOk awesome thanks so much guys
09:57allusivedevSo i'm a tad confused but i'm assuming I should build m1n1 locally and install it from the user guide instructions?
09:59allusivedevOk i'll proceed and be back if I need anymore help, thank you!
10:00allusivedev has quit [Remote host closed the connection]
10:07MikeM64 has quit [Read error: Connection reset by peer]
10:08MikeM64 has joined #asahi-dev
10:09gargantua has joined #asahi-dev
10:20allusivedev has joined #asahi-dev
10:20allusivedevNot sure what I did wrong here. https://pastebin.com/HiPdwZxd
10:21JamesCalligeros[m]Add ARCH=aarch64-unknown-none to your make invocation
10:28allusivedevSo that got me so far but then I needed cargo installed so I thought I'd try the docker container build method but I'm getting this error? https://pastebin.com/ggshPEME
10:28JamesCalligeros[m]The docker container has probably bitrotted to oblivion
10:28MichaelReeves[m]thats a known issue with 1.99 rustc we were talking about it earlier in #asahi-alt:matrix.org
10:28allusivedevnot sure what you mean?
10:29MichaelReeves[m]add RUSTFLAGS="-C llvm-args=-disable-loop-idiom-wcslen" or update to 1.100 beta or nightly
10:29JamesCalligeros[m]Oh its the wcslen thing again, yeah ^
10:38chaos_princessnative macos builds should work btw, no need to mess with containers
10:38allusivedevYeah i would just need to install cargo i guess.
10:39allusivedevStrangely neither updating to nightly or adding those rustflags has fixed the issue
10:40allusivedevconfirmed by checking rustc -v and echoing the container's rustflags
10:41allusivedev*typo meant to write hasn't fixed the issue
10:42MichaelReeves[m]try without docker so theres less moving parts
10:42MichaelReeves[m]installing cargo is easy its just https://rustup.rs
10:42allusivedevok cool, and nightly yeah?
10:43MichaelReeves[m]yep. or you can use stable with the flag. it doesn't really matter.
10:43MichaelReeves[m](doesn't matter which one you chose)
10:47allusivedevOk that actually successully built!
10:48allusivedevthanks so much for the help!
10:49allusivedevI'm so confused reading the installation step
10:50chadmed has joined #asahi-dev
10:50allusivedevI don't know what 1TR is, what my OS volume should be or where m1n1-stage1.bin is
10:50MichaelReeves[m]1TR = "one true recovery"
10:50chadmedwhich document are you reading
10:50fl0_idallusivedev just get rustup and get nightly, but still have to add the right target
10:50chadmedtheres a good one i recently overhauled and a less good one that is ~4 years old
10:51chaos_princess1tr - shutdown the machine fully, power it on by holding the power button so you get the boot picker and then enter "options"
10:51MichaelReeves[m]basically just turn your Mac all the way off. and hold the startup button for 10 seconds, and then press options
10:51MichaelReeves[m]oops jinx chaos_princess
10:51chadmedoh i need to put instructions for getting to 1TR in there too
10:52jannauor add links to the glossary
10:52allusivedevOh ok I'm starting to understand.
10:52allusivedevAnd so since M4 isn't technically supported yet I have to build the stage1.bin myself?
10:53chadmedthe stage 1 bin is _just_ m1n1
10:53chadmedwith instructions to chainload from an ESP if you want to boot a "production" linux
10:54MichaelReeves[m]basically, what you need to do is just: 1. copy m1n1.bin into somewhere accessible i.e. ~/Users/shared 2. reboot into 1TR (as said above) 3. go to the terminal and then you need to run a couple commands
10:55allusivedevAh gotcha, I was wondering how I'd get the bin from 1tr
10:56chadmedjust to be clear, you do need a second machine to act as the m1n1 "host" to do anything useful
10:56allusivedevA second mac or just any other machine?
10:56chadmedusually what youd do is build m1n1 on that one and serve it up to 1TR via some mechanism (python -m http.server for example)
10:56chadmedit can be any linux machine
10:57chadmeda second apple silicon mac is/was useful for very low level serial debug but is not necessary
10:57allusivedevOh ok. I guess I should probably confirm now but to install m1n1 and continue will I lose any data on my current macos install?
10:58chadmedyou wont, you can put macos back into full security mode at any time
10:58MichaelReeves[m]it just changes the boot object and you an change it back to macOS later
10:58chadmedbut thats not a guarantee. if you break anything you get to keep the pieces
10:58chaos_princessyou are supposed to install m1n1 over a second macos install, not your primary one
10:58chadmedyou should probably read the guide i linked above in its entirety, as the Best Practice(tm) way of setting up m1n1 is to use a second install of macos to somewhat protect your main one if it has stuff you care about on it
10:59allusivedevYeah I noticed about a second copy of macOS in the guide. I'll have a further read of it and the rest of the docs and come back to it tomorrow. I'm in australia and its late
10:59allusivedevThanks again so much for the help guys, looking forward to helping anyway I can.
11:00MichaelReeves[m]and then there were 4...
11:00MichaelReeves[m]anyway glad it went well and good night! :)
11:00chadmedraaahhhh lets go aussie aussie aussie
11:00myriaWell, I'm here :P
11:00myriaJust installing linux on macbook 3
11:00chadmedim going to my local charcoal chicken tomorrow and eating a double chicken burger just as spiritual revenge for tonight
11:01myria(How does this happen to us all)
11:01chadmedall my homies hate the chooks
11:01allusivedev has quit [Quit: allusivedev]
11:02chadmedoh yeah its daylight savings down south now no wonder it's "getting late"
11:03MichaelReeves[m]I should probably go to bed LOL (school starts again tomorrow also...) but going to bed means accepting the terms starting again so im trying to delay it as much as possible :P
11:03myriaYeah, mood
11:04chadmedpublic holiday up here tomorrow >:3
11:04myriaWait, no public holiday for you tomorrow?
11:04chadmedi think victoria still has weird labour day/kings birthday dates
11:04MichaelReeves[m]no public holiday :(
11:05myriaOh, oops, I got the M5 release years wrong...
11:06chadmedive got to split out the zensical build and the pages push into separate flows so we can run a test build on every PR too
11:06chadmedhad to fix a bunch of YAML nonsense this morning as the build failed when i merged your changes
11:06chadmed(not your fault at all, it was other unrelated pages)
11:07myriaAh, sounds like lots of fun
11:10myriaSent the year updates
11:13allusivedev has joined #asahi-dev
11:13allusivedevhey what version of macos do you guys reccomend I install for my second copy (for m1n1)?
11:13myriaUhhh, on M4?
11:13chadmedprobably some 26.x
11:14myriaLatest 26 I think has all the fixes? So either that or 27
11:14chadmedanything older is riddled with bugs
11:14myriayuka if she's around might have a better idea
11:15myriaYeah, latest 26 was what was recommended here a day ago
11:15yukaI would use 26.6.2
11:25allusivedevthanks, is `softwareupdate` meant to take forever to scan for the installer?
11:26MichaelReeves[m]you can use Mist if it's taking too long, it's what I use
11:29chadmedsofus: were 32x32 tiles *not* the right size for 10-bit luma interchange?
11:29chadmedi am going to resend the interchange modifier patch but if the description needs to change to be accurate for 10-bit formats lmk before i do so
11:33allusivedevHey guys for the purpose of trying to help with the M4 support for asahi should I get the production kernel or KDK?
11:33allusivedevor does it not matter?
11:33chadmedits not important, the production kernel is easier to get
11:34allusivedevok sweet thanks!
11:36yuka> the M4 Pro mini uses only dcpext* for its 3 supported displays
11:36yukais that also the case on the macbooks?
11:37yukado I delete the disp domains in t6040.dtsi or in the Mac mini / Mac Studio dts?
11:45jannauI'd be surprised if mbp did not use dcp0 so just remove it for the mac mini/studio
11:49allusivedev has quit [Quit: allusivedev]
11:50jannauthe macOS 27 template adt has dcp. annoying if that requires device ADT
11:50hardy94 has joined #asahi-dev
11:51yukayou mean that iboot removes the power domains from the adt pmgr devices?
11:52yukacan easily check that
11:53yukaindeed
11:53yukaDISP_CPU no_ps = True
11:54jannauno, just the dcp related device node, the power-domains are still there
11:56yuka> annoying if that requires device ADT
11:56yukaI don't think I understand what requires device ADT
11:58jannaufinding out that dcp/disp0 is not used. The j773sap template ADT from the ipsw has those nodes but the adt retrieved via `python3 -m m1n1.adt -r ...` when connected to j773s does not
12:00jannaudrop the m4 max mac studio, that should use dcp it supports up to 5 displays
12:06MichaelReeves[m]would aicv3 have changed anything that breaks emu?
12:06MichaelReeves[m]someone reported it on reddit, and I can reproduce. but if I set its=off, then it works, so I think it's something wit virtual PCI interrupts
12:09sofuschadmed: description is still correct, the issue was bpp
12:11jannauMichaelReeves[m]: we have MIDR based workarounds, for example e910baa9c1efdf7634519c135c6723b0fd499683 and possible impdef_pmuv3_cpus in arch/arm64/kernel/cpu_errata.c
12:11myriaqemu_kvm I'm assuming? What is the guest?
12:11myriaWindows ARM may run into pmuv3 issues
12:12JamesCalligeros[m]sofus: thanks, will send with only the typo fix then :)
12:12MichaelReeves[m]myria: its linux guest
12:13MichaelReeves[m]and yes with KVM.
12:13MichaelReeves[m]software emulation doesn't run into the issue
12:13MichaelReeves[m]jannau: thanks. I will look into it
12:13myriaHmmmm, then yeah, probably the midr based workarounds
12:25mmediouni[m]Apple deliberately only exposes CPU model 0x0
12:25mmediouni[m]but that's more linked to being heterogenous and macOS not supporting core pinning
12:26mmediouni[m]so hvf just zeroes it out (and that's why kernel-irqchip=off,gic-version=3,its=on works fine on a Mac host)
12:27mmediouni[m]But filtering by EL would also be wrong. What's conventionally done by other platforms is to filter by PSCI SoC ID but not an option here...
12:38MichaelReeves[m]lvm-arm.vgic_v3_group1_trap=1 fixes it
12:49rany has quit [Remote host closed the connection]
12:50rany has joined #asahi-dev
12:51mattia013 has joined #asahi-dev
12:53myriaMichaelReeves[m] if I send you a patch for the host, can you test if it works?
12:59myriahttps://github.com/sempiternal-aurora/linux/tree/fix-vgic-m3 <- this might fix it? Might need to also disable nv1, but let me know what happens
12:59integralpilot has joined #asahi-dev
12:59integralpilotmyria: thanks, i'll try that
13:00integralpilotim trying https://pastebin.com/4fhjyLnG now (which, given the kernel cmdline thing fixes it, i think will fix it)
13:00myriaAh true, I guess if only the trap 1 fails...
13:04integralpilotyeah, have to see
13:04integralpilotgoing to bed now but will sort out tmrw, thanks for helping!!
13:04integralpilot has quit [Quit: Lost terminal]
13:05myriaNo problem
13:12revadam has joined #asahi-dev
13:13revadam has quit []
13:29yukat6050 has no smc-gpio o.O
13:33jannauhow is the hdmi hdp handled and possible the power enable for the sd card reader? apple's wifi probbably doesn't need pwren gpios
13:40marvin24 has joined #asahi-dev
13:58bennofs3 has quit [Ping timeout: 480 seconds]
13:59pb17 has quit [Ping timeout: 480 seconds]
14:03bennofs3 has joined #asahi-dev
14:05voidah has joined #asahi-dev
14:11pb17 has joined #asahi-dev
14:11gargantua has quit [Ping timeout: 480 seconds]
14:13yukathere is an dp2hdmi-gpio0
14:14yukabut we already have that on M3
14:15voidah has quit [Ping timeout: 480 seconds]
14:16yukajannau: smc-gpio is the adt name. what I mean is pinctrl_smc
14:16yukathe pwr_en fields still point to the smc-pmu phandle
14:17yukaso no change there
14:17yukaI don't know if pinctrl_smc was ever used for anything?
14:17jannauI was wondering what it uses as for function-*. but my mistake was smc-gpio vs. gpio over smc keys
14:17yukaI think we can mostly just ignore it
14:17yuka(we still need to figure out the spmi gen4 irqs though to do smc stuff)
14:18yuka(ah, but different smc stuff)
14:19yukasomehow I can't get i2c8 working properly
14:19yukait just gets stuck trying to i2cdetect
14:19yuka0-7 work
14:21yukaI'll probably batch M4 Pro/Max and M5 minimal device trees when I send them out. M5 Pro/Max is a bit more involved since that will also have multi die stuff.
14:21jannaujust ignore i2c8 for the dts if it is unused
14:21yukathere is a pcon0 subnode on i2c8
14:21yukano idea what that is
14:25myriaintegralpilot: belay that patch, I don't understand the GICv3 stuff enough to be confident it'll help, might just do nothing at all
14:25myriaBut the MIDR definitions are there if you need them
14:28mmediouni[m]and fyi, I did deprecate its= in qemu a while back, the new syntax is msi= (to allow for things like GICv3 + GICv2m)
14:31hightower4 has joined #asahi-dev
14:38hightower3 has quit [Ping timeout: 480 seconds]
14:40meow676767676 has quit [Read error: Connection reset by peer]
14:50meow676767676 has joined #asahi-dev
14:51bennofs3 has quit [Ping timeout: 480 seconds]
14:53bennofs3 has joined #asahi-dev
15:15MichaelRostom[m] has left #asahi-dev [User left]
15:40aissen_ has quit []
15:42okb has joined #asahi-dev
15:43Guest1217 has quit [Quit: Bridge terminating on SIGTERM]
15:43rhysmdnz has quit [Quit: Bridge terminating on SIGTERM]
15:48rhysmdnz has joined #asahi-dev
15:48Guest1658 has joined #asahi-dev
15:58pb17 has quit [Ping timeout: 480 seconds]
15:59pb17 has joined #asahi-dev
16:09hello0092 has quit [Ping timeout: 480 seconds]
16:10hardy94 has quit [Remote host closed the connection]
16:19bennofs3 has quit [Ping timeout: 480 seconds]
16:22bennofs3 has joined #asahi-dev
16:23hightower3 has joined #asahi-dev
16:28bdmg_ has joined #asahi-dev
16:30hightower4 has quit [Ping timeout: 480 seconds]
16:31yukarhlvista[m] your methodology means I can not tell what you changed after copying the device tree :/
16:32yukathis is not helpful
16:37hello0092 has joined #asahi-dev
16:40marvin24 has quit [Quit: https://quassel-irc.org - Komfortabler Chat. Überall. ]
16:41voidah has joined #asahi-dev
16:43hello0092 has quit [Read error: Connection reset by peer]
16:49voidah has quit [Ping timeout: 480 seconds]
17:00okb has quit [Ping timeout: 480 seconds]
17:08Hel3zi has joined #asahi-dev
17:08Hel3zi has quit [autokilled: This host violated network policy. Mail support@oftc.net if you think this is in error. (2026-10-04 17:08:06)]
17:12Hel3zi has joined #asahi-dev
17:12Hel3zichaos_princess: is running the hv on sequoia a bad idea, or does it work fine on m4?
17:12chaos_princesssequoia on m4 is 100% broken
17:13chaos_princessthere is like one version that can boot one heavily broken core in raw boot mode, and every other version crashes before even reaching m1n1
17:15Hel3zidamn, good thing i asked lol. my main install is actually tahoe, sequoia was just what i had in mind for testing. is tahoe in a much better spot for the hv?
17:15yukahel3zi do not expect to run the hv on M4 at this point
17:16yukas-ven has it working on one machine but I think the changes are still slowly trickling into the m1n1 repo
17:16yukaand that's with some 26/27 version
17:19svenyeah, probably 1-2 more weeks until the entire thing is upstream
17:20hello0092 has joined #asahi-dev
17:20svenit‘s still a it of mess and requires binary patches to xnu and SPTM and I hope to get rid of most of them
17:22Hel3ziout of curiosity, what specifically is missing in the current upstream code that prevents it from booting right now? is it mostly the sptm/el2 bringup or other components?
17:24svenmostly SPTM stuff
17:24sventhe SPRR/GXF emulator also needs a few fixes
17:25svenjust need to clean all that stuff up and wait for j-annau to review it
17:26Hel3zii saw allusivedev earlier and thought you guys were already booting it, and honestly how could i resist a good dose of painful tinkering haha
17:27yukayou can boot Linux if you want
17:27yukabut that's only useful to some degree with the current state
17:28Hel3ziwait, really?
17:28Hel3zihow do i set that up?
17:29yukaRead all of these:
17:29yukait was a base M4 right?
17:29yukathat one already has the basic device tree in asahi-wip
17:31Hel3zino, m4 prщ
17:31yukamacbook or mac mini?
17:32yukaIt requires quite some tinkering, you get started by installing m1n1 and getting a serial connection as described in the articles
17:32yukawhen you have that, you compile the kernel and boot it using m1n1/proxyclient/tools/linux.py
17:32yuka(that is, you load it over usb from a second machine)
17:34yukaIf you really want to take some time to install m1n1 and tinker with the device, I would appreciate if you can test some MacBook specific things (keyboard backlight, framebuffer for the internal display) with my device trees
17:35yukabut basically read the guides, install m1n1, and then come back ;)
17:37Hel3ziabsolutely, i'd be glad to contribute even in a small way! going to read through everything now and get m1n1 up and running. thanks yuka!
17:40yukasome things which I don't know if they are necessary but I always do them and they are not mentioned in the docs: disable file vault on the installation whose boot entry you want to install the custom boot entry to (I think it will complain loudly on csrutil disable if you have it enabled), and bputil -nkas (_before_ csrutil disable) disables some security features
17:41svenno need to disable FileVault, you just have to drop one of the flags from bputil then
17:41svenit’ll tell you which one
17:43hardy94 has joined #asahi-dev
17:51Hel3zi has quit [Remote host closed the connection]
17:53bennofs3 has quit [Ping timeout: 480 seconds]
17:54hardy94 has quit [Remote host closed the connection]
18:05bennofs3 has joined #asahi-dev
18:07okb has joined #asahi-dev
18:14yukafunny... my t6050 thinks it's always at 0 seconds uptime
18:15yukaI would think it had a rudimentary clock even if no rtc
18:16Hel3zi has joined #asahi-dev
18:17Hel3zifetching the 26.6.2 installer right now, will have the test volume ready shortly
18:17Hel3zi has quit [Remote host closed the connection]
18:45iwuvmika has joined #asahi-dev
18:46chadmed has quit [Remote host closed the connection]
18:46chadmed has joined #asahi-dev
18:47rhlvista[m]yuka: i tested the commit it doesnt boot
18:48yukayou just tested it?
18:49yukaI created this commit after your previous report, so just making sure
18:51yukathat is oddly specific. how did you arrive at those changes?
18:54chadmed has quit [Read error: Connection reset by peer]
18:54chadmed has joined #asahi-dev
18:55rhlvista[m]how i found out the uart was the problem i commended out everything everything except the framebuffer and AIC
18:55rhlvista[m]then i took my pmgr compared yours with it changed it enough where it got to shell from trial and error
18:55mini has quit [Quit: The Lounge - https://thelounge.chat]
18:55mini has joined #asahi-dev
18:55rhlvista[m]how i found out the uart was the problem i committed out everything everything except the framebuffer and AIC
18:55rhlvista[m]then i took my pmgr compared yours with it changed it enough where it got to shell from trial and error
18:56rany has quit [Remote host closed the connection]
18:57rany has joined #asahi-dev
18:58yukaok but the uart is fixed in my new version, right? or did I miss another uart thing besides the 0x3ad... -> 0x385...
18:58yuka> the uart also breaks due [...]
18:58yukawhy or how does it break?
18:59rhlvista[m]no funny enough uart works before you fix commit and after the commit it breaks(as of crashing)
18:59rhlvista[m]it breaks as of crashing the laptop
19:00rhlvista[m]this log showes the uart breaking
19:00yukabut this is before my change
19:01rhlvista[m]the pastbin i linked was after the change the other pastbins I just undid the change
19:01yukain the crash you see 3ad... and in the commit I linked it's 3a5200000
19:01yukaSo I don't think you're running the commit I sent in that latest log?
19:03rhlvista[m]not in that bin i can give me a second but the commit for "tty: serial: samsung: Fix S5L on Apple A18 Pro" breaks it on M5
19:03yukaOh huh
19:04yukathat is odd, it works on M4, A18 Pro & M5 Pro
19:04yukaand M4 Pro too
19:05yuka(but admittedly, the newer serials are all somewhat broken depending on what is the other side, so it might continue working with the central-scrutinizer but stop working with some other adapter or mac on the other side?)
19:10okb has quit [Ping timeout: 480 seconds]
19:11rhlvista[m]i tested with the serial Samsung fix its working there is a good chance i did something wrong wile trying to fix the device tree and i didnt think of it
19:11yukaOk good!
19:12rhlvista[m]sorry for the scare :(
19:12yukaYou said "disable ps_gpio due to it crashing the pc", I can imagine what this is. Can you provide the crash logs (with my branch, the DEBUG commit in there is important)
19:13hello0092 has quit [Ping timeout: 480 seconds]
19:17yukaI'm guessing it's some specific places which we can not read (just like on T8140 and T6050), so we need to add them to gpio-reserved-ranges
19:17yukaon pinctrl_ap
19:23rhlvista[m]i believe it haves to do with the glitchless = 1 in the adt
19:23rhlvista[m]but on your kernel it does not crash witch is good
19:23rhlvista[m]it crashes on the asahi kernel so when i was messing with the pmgr it was easier to just disabled it from there
19:26rhlvista[m]i didnt enable pinctrl_apple_gpio in config
19:26hello0092 has joined #asahi-dev
19:28rhlvista[m]it does crash
19:29yukaOk yes that's exactly it
19:29yuka> FAR: 0x39a000138
19:30yukahex(0x138//4) => 0x4e
19:30yukawhich is listed in no-restore-gpio-pins
19:31yukaadd this to pinctrl_ap: pinctrl@39a000000: gpio-reserved-ranges = <78 1>;
19:31yukait will probably crash again with a different FAR
19:34rhlvista[m]TTY> FAR: 0x39a00032c
19:35yukagpio-reserved-ranges = <78 1>, <203 1>;
19:38ipatch has joined #asahi-dev
19:38ipatch_ has quit [Read error: Connection reset by peer]
19:39rhlvista[m]i have shell
19:39yukaI assume it will probably reboot after a few minutes?
19:40hello0092 has quit [Ping timeout: 480 seconds]
19:40hello0092 has joined #asahi-dev
19:41rhlvista[m]make sure you change the pmgr because it influences almost everything
19:41ipatch has quit [Read error: Connection reset by peer]
19:41yukachange what exactly?
19:41ipatch has joined #asahi-dev
19:46yukaI want to be entirely sure that I understand all of these changes, and that it's the minimal set of changes to get the device to boot and stay on
19:48yukaps_soc_dpe already has always-on on most other devices, so that is plausible I guess. But what happens when you do all the changes except that one?
19:48hello0092 has quit [Ping timeout: 480 seconds]
19:48rhlvista[m]yea thats minimal ammount of change for nvme to work and shell to work
19:49yukaalways-on on ps_ans makes sense to me, the M4+ devices do not like the ans power domain going down ime
19:49yukaps_soc_dpe and ps_st0_aus5 I would like more information on
19:49yukawhat happens without the changes to them?
19:49yukadeleting ps_pms_fpwm1 is odd as well
19:50rhlvista[m]anything fpwm1 is realy bad for M5 for some reason
19:50yukaokay... for which reason?
19:51yukadoes it crash with an SError trying to touch that?
19:51yukathere is some precedent for that, disp*_cpu are the same
19:51rhlvista[m]i beleve the disp*_cpu stuff is normal
19:51yukabut then I'd like to see the crash log with the exact FAR or L2C_ERR_STS to understand what is happening
19:55yukaadded: status = "disabled"; /* Locked, don't touch */
19:56yukathis is more proper than always-on because with always-on it might still try to write it in some circumstances
19:58yukathe ps_soc_dpe paste is actually crashing on fpwm1 still
19:58pb17 has quit [Ping timeout: 480 seconds]
19:58yuka> L2C_ERR_ADR: 0x282500380700150
19:58yukathe other one as well
19:58yukaso maybe those changes are not actually needed?
19:59rhlvista[m]wops forgot to delete gonna redo the hypervisor stuff again
19:59rhlvista[m]good chance there are needed
20:00yukasomething something peer review, since one can always make mistakes (all of us do) :-)
20:02hello0092 has joined #asahi-dev
20:04rhlvista[m]ps_st0_aus5: makes system not post to shell (aka rootfs does not run) https://pastebin.com/evjSheiq
20:04rhlvista[m]ps_soc_dpe: does not need apple,always-on; (post to rootfs)
20:05yukahm, ps_st0_aus5 does not cause a crash or anything?
20:05yukait just hangs?
20:05yukawhat happens if you wait 5 or 10 minutes?
20:07chadmed has quit [Quit: Konversation terminated!]
20:08chadmed has joined #asahi-dev
20:08rhlvista[m]didn't test it yet in the log it hangs. ill test for 10 minutes
20:08rhlvista[m]and ans i believe without it the laptop crashes if you wait for 4/8 minutes
20:08rhlvista[m]testing ps_st0_aus5
20:10chadmed has quit [Read error: Connection reset by peer]
20:10chadmed has joined #asahi-dev
20:13rhlvista[m]ps_st0_aus5 crashes nvme
20:13rhlvista[m]?? this laptop is confusing me
20:15yukaAha, wait!
20:15yukathis is interesting
20:16hello0092 has quit [Ping timeout: 480 seconds]
20:16yukaps_aphy_st0_aus5_a indirectly depends on ps_apcie_phy_sw
20:16yukaso I think the apcie0 power-domain of the nvme node is just wrong
20:17yukayou're using your version again, right?
20:18yukaso the apcie0 power domain of the nvme node is ps_apcie_phy_sw?
20:19rhlvista[m]its ps_apcie_phy_sw
20:19iwuvmika has quit [Ping timeout: 480 seconds]
20:20marvin24 has joined #asahi-dev
20:21pb17 has joined #asahi-dev
20:23rhlvista[m]why did ps_st0_aus5 crash everything even though its not connected to nothing?
20:24marvin24 has quit []
20:24hello0092 has joined #asahi-dev
20:24marvin24 has joined #asahi-dev
20:34yukaI'm out for today but thank you for the debugging session
20:36allusivedev has joined #asahi-dev
20:39rhlvista[m]cool btw ps_apcie_phy_sw isnt in the nvme it crashes. thanks for helping with the nvme!
20:47ipatch has quit [Quit: WeeChat 4.10.1]
20:48ipatch has joined #asahi-dev
20:50allusivedev has quit [Remote host closed the connection]
21:00Hel3zi has joined #asahi-dev
21:00Hel3zim1n1 booted on m4 pro! internal display framebuffer works and the asahi logo showed up, currently sitting at "Running proxy..”
21:01Hel3zi has quit [Remote host closed the connection]
21:05allusivedev has joined #asahi-dev
21:05allusivedev has quit []
21:12yukaHel3zi try to check out https://cyberchaos.dev/yuka/linux branch "all" and build it
21:13hightower3 has quit [Ping timeout: 480 seconds]
21:18hightower2 has joined #asahi-dev
21:24Hel3zi has joined #asahi-dev
21:30rosalyntg has joined #asahi-dev
21:30gargantua has joined #asahi-dev
21:31rosalyntg_ has quit [Read error: Connection reset by peer]
21:36gargantu1 has joined #asahi-dev
21:37Hel3ziPage size (16KB)?
21:39gargantua has quit [Ping timeout: 480 seconds]
21:40iwuvmika has joined #asahi-dev
21:40mini has quit [Quit: The Lounge - https://thelounge.chat]
21:40mini has joined #asahi-dev
21:47yukadefconfig + 16KB pagesize works but it may be a bit large (longer compile time)
21:49iwuvmika has quit [Ping timeout: 480 seconds]
21:52allusivedev has joined #asahi-dev
21:52allusivedevHey all I'm back
21:52jannaueasiest way to get rid of some unused drivers is to deselect every platform except ARCH_APPLE
21:52allusivedevI just got the m1n1 hypervisor booted and I'm unsure of what to do now.
21:52yukayes, though that's still quite a lot of drivers
21:53yukaallusivedev what device do you have again?
21:53allusivedevM4 macbook pro base model (24GB ram)
21:55allusivedevI'm up to my Mac showing a rainbow apple logo, and the run_guest scripts last log line is "jumping to entry point ..."
21:55yukawhat payload did you send?
21:55yukaor rather, what are you trying to boot in the hv
21:55yukawe could try to get keyboard working with integralpilot's new dockchannel-hid patches
21:56allusivedevi just followed the hypervisor booted guide. So the kernelcache.macho?
21:56yukaAh well that won't work currently
21:56yukamacOS in hv on M4 is not done yet
21:57allusivedevI see, what do you reccomend?
21:59yukadepends on what do you want to achieve?
22:00allusivedevI just want to help with development really. I thought ive got an m4 mbp I'd like to "offer" for testing.
22:01yukaWell there needs to be something which can be tested first :D
22:01yukaif it's already known working, there is no point in testing it
22:01yukamost things the developers can test themselves
22:02allusivedevOh ok, is there any ways else I can help then?
22:02yukaYou can try to boot a linux kernel and minimal initrd
22:03yukaThis should pretty much work, but then you can test the keyboard when I write the device trees for that (a few days from now)
22:03allusivedevOk sure, is that just via the tethered boot
22:03yukayou use linux.py from m1n1
22:03yukato upload the kernel and initrd to the proxy
22:04yuka(or run it in the hypervisor, doesn't matter too much for most things)
22:04yuka(I only have an M4 Mac mini, so can't test the keyboard device tree nodes ;))
22:05allusivedevOk cool, so since ive already got the hypervisor running do i jusr use the run_guest_kernel tool?
22:05yukayou can use that, yeah. doesn't make too much of a difference.
22:05yukathe important part is that you get the kernel building and an initrd from somewhere.
22:06allusivedevSure, whereabouts should I get those?
22:07Hel3zi has quit [Remote host closed the connection]
22:07JamesCalligeros[m]You need to build them
22:08allusivedevOk and I need to do that from a working macos install on the device?
22:08yukafrom a linux
22:08yukaon a second machine
22:09allusivedevRight ok, sorry I havent built the kernel in a long time I'm not too sure what I'm doing.
22:10yukayou go figure that out, something something make defconfig && make -jsomething ;)
22:10Hel3zi has joined #asahi-dev
22:11yukathen you should get a arch/arm64/boot/Image.gz and arm64/boot/dts/apple/t8132-j604.dtb
22:12allusivedevOk I'll give it a try
22:13Hel3zi has quit [Remote host closed the connection]
22:13allusivedev has quit [Read error: Connection reset by peer]
22:15voidah has joined #asahi-dev
22:21allusivedev has joined #asahi-dev
22:22Hel3zi has joined #asahi-dev
22:22Hel3ziyuka: tested your kernel on t6040 (j614s). It didn't reach Linux, got SError in m1n1 at EL2 right after dapf: Initialized /arm-io/dart-mtp. TTY> Exception: SError TTY> Exception taken from EL2h TTY> MPIDR: 0x80010100
22:23voidah has quit [Ping timeout: 480 seconds]
22:24yukaAh yes you just comment it out like this: https://termbin.com/t3f3
22:24yukauntil we fully understand it
22:26allusivedev_ has joined #asahi-dev
22:26allusivedev_ has quit [Remote host closed the connection]
22:27gargantu1 has quit [Ping timeout: 480 seconds]
22:30Hel3zi has quit [Remote host closed the connection]
22:31ninelore has quit []
22:32allusivedev has quit [Ping timeout: 480 seconds]
22:51Hel3zi has joined #asahi-dev
22:51Hel3ziyuka: applied your dapf.c patch and built m1n1. When using chainload.py, it reconnects (Connected), but then Mac immediately reboots (UartTimeout on iface.nop). Does chainload work on t6040, or should I update m1n1 on disk / use hypervisor?
22:52Hel3zi has quit [Remote host closed the connection]
22:52Hel3zi has joined #asahi-dev
22:53yukawhat are you chainloading
22:53yukachainloading a new version of m1n1 should work
22:54yukayou load linux with linux.py or run_guest.py -r
22:54yukathis is my usual command:
22:54yuka"make -j10 && proxyclient/tools/chainload.py -r build/m1n1.bin && python3 proxyclient/tools/linux.py ~/proj/linux/arch/arm64/boot/{Image.gz,dts/apple/t6050-j714s.dtb} initrd -b "earlycon console=ttySAC0 loglevel=7""
22:55Hel3ziI'm chainloading build/m1n1.macho, not Linux. But in chainload.py I had to skip setting secondary CPU RVBARs because writing to cpu5 [0x211150000] caused an SError. With that skipped, it loads into the stub, says "Connected", but then immediately times out on iface.nop() and the Mac reboots.
22:55yukause build/m1n1.bin, not macho
22:55yukayou're using a very old version then
22:55yukathe RVBARs issue is long fixed
22:56voidah has joined #asahi-dev
22:59Hel3zi has quit [Remote host closed the connection]
23:00iwuvmika has joined #asahi-dev
23:02hello0092 has quit [Ping timeout: 480 seconds]
23:03Hel3zi has joined #asahi-dev
23:03Hel3ziI tried chainload.py -r build/m1n1.bin. It connects after the stub, but then immediately reboots with UartTimeout. I'm currently on AsahiLinux/m1n1 main branch. Which repo and branch should I use?
23:03allusivedev has joined #asahi-dev
23:04ninelore has joined #asahi-dev
23:04Hel3zi has quit [Remote host closed the connection]
23:04ninelore has quit [Read error: Connection reset by peer]
23:04ninelore has joined #asahi-dev
23:05yukado you have a serial console?
23:05yukawithout it, this is very hard to debug
23:06yukaAsahiLinux/m1n1 main is correct
23:06allusivedevSorry to interject out of curiosity how close is asahi to officially supporting m4 series and running like fedora?
23:06yukawrong question
23:06Hel3zi has joined #asahi-dev
23:07Hel3ziJust a regular USB-C cable between the Mac and the PC (I get /dev/ttyACM0 and /dev/ttyACM1 when m1n1 is booted). No dedicated hardware serial cable.
23:07yukayeah you won't see exactly what is happening
23:07yukayou just see that the connection drops
23:07yukaso that could be anything
23:07Hel3zi has quit [Remote host closed the connection]
23:08allusivedev_ has joined #asahi-dev
23:09allusivedev_ has quit [Read error: Connection reset by peer]
23:11allusivedev_ has joined #asahi-dev
23:11allusivedev_What's the right question 😅
23:13Hel3zi has joined #asahi-dev
23:13Hel3ziI guess there's not much else I can help with for now? Did I even help at all though, feels like I just took up your time haha
23:13Hel3zi has quit [Remote host closed the connection]
23:15allusivedev has quit [Ping timeout: 480 seconds]
23:18hello0092 has joined #asahi-dev
23:25voidah has quit [Ping timeout: 480 seconds]
23:27rany has quit [Remote host closed the connection]
23:28rany has joined #asahi-dev
23:29yuka@hel3zi taking up time is okay to some degree. even if it did not work out this time. if you can find another mac you can give it another shot with the serial console :)
23:33Mayanico13 has joined #asahi-dev
23:33voidah has joined #asahi-dev
23:33JamesCalligeros[m]allusivedev_: there isnt a right question really per the link above
23:33allusivedev_ has quit [Ping timeout: 480 seconds]
23:35Mayanico13 has quit []
23:35JamesCalligeros[m]i hacked gentoo support up and was dailying it before we even had speakers
23:41voidah has quit [Ping timeout: 480 seconds]
23:53voidah has joined #asahi-dev
23:54yukasame but s/gentoo/nixos/
…