ChanServ changed the topic of #asahi-dev to: Asahi Linux: porting Linux to Apple Silicon macs | Non-development talk: #asahi | General development | GitHub: https://alx.sh/g | Wiki: https://alx.sh/w | Logs: https://alx.sh/l/asahi-dev
pb17 has joined #asahi-dev
chrisl has joined #asahi-dev
lynndotpy has quit [Quit: Ping timeout (120 seconds)]
lynndotpy has joined #asahi-dev
chrisl has quit [Ping timeout: 480 seconds]
buroa2354 has joined #asahi-dev
tbodt has quit []
tbodt has joined #asahi-dev
buroa235 has quit [Ping timeout: 480 seconds]
buroa2354 is now known as buroa235
tbodt_ has joined #asahi-dev
tbodt has quit [Ping timeout: 480 seconds]
svl has quit []
chrisl has joined #asahi-dev
chrisl has quit [Ping timeout: 480 seconds]
sand4rt has quit [Ping timeout: 480 seconds]
svl has joined #asahi-dev
chrisl has joined #asahi-dev
chrisl has quit [Ping timeout: 480 seconds]
pb17 has quit [Ping timeout: 480 seconds]
chadmed_ has quit [Quit: Konversation terminated!]
pb17 has joined #asahi-dev
svl has quit [Remote host closed the connection]
hightower4 has joined #asahi-dev
chrisl has joined #asahi-dev
hightower3 has quit [Ping timeout: 480 seconds]
chrisl has quit [Ping timeout: 480 seconds]
tobhe_ has joined #asahi-dev
HitokiriKincho has quit [Remote host closed the connection]
HitokiriKincho has joined #asahi-dev
tobhe has quit [Ping timeout: 480 seconds]
HitokiriKincho has quit [Ping timeout: 480 seconds]
pb17 has quit [Ping timeout: 480 seconds]
pb17 has joined #asahi-dev
HitokiriKincho has joined #asahi-dev
<nickchan>
my cpufreq status on A11 (T8015): state switch in m1n1 works perfectly for pcore and ecore using the m1 logic, however, once booted into linux the same logic wouldn't work anymore, whether from linux or from an external debugger
<nickchan>
wouldn't work = APPLE_DVFS_CMD_BUSY never clears
<nickchan>
default governor is set to userspace to make sure the kernel never does anything on its own
nora has joined #asahi-dev
nora_ has quit [Ping timeout: 480 seconds]
jnn has joined #asahi-dev
jn has quit [Ping timeout: 480 seconds]
<nickchan>
and I made sure I did not use any boost states as configured by iBoot
<nickchan>
only the normal states
<nickchan>
notably 'ppt-thrtl' is in the apple device tree but i did not do anything about it
<nickchan>
not sure if it could cause wedging in this way
chrisl has joined #asahi-dev
chrisl has quit [Ping timeout: 480 seconds]
HitokiriKincho has quit [Ping timeout: 480 seconds]
chrisl has joined #asahi-dev
chrisl has quit [Ping timeout: 480 seconds]
HitokiriKincho has joined #asahi-dev
JayBeeFOSS has quit [Remote host closed the connection]
JayBeeFOSS has joined #asahi-dev
jn has joined #asahi-dev
jnn has quit [Ping timeout: 480 seconds]
pb17 has quit [Ping timeout: 480 seconds]
kokoschmoko has joined #asahi-dev
pb17 has joined #asahi-dev
kokoschmoko has quit [Ping timeout: 480 seconds]
chrisl has joined #asahi-dev
chrisl has quit [Ping timeout: 480 seconds]
<chadmed>
jannau: do you remember which igt test failures you were concerned about?
chrisl has joined #asahi-dev
chrisl has quit [Ping timeout: 480 seconds]
chrisl has joined #asahi-dev
jnn has joined #asahi-dev
jn has quit [Ping timeout: 480 seconds]
jn has joined #asahi-dev
jnn has quit [Ping timeout: 480 seconds]
<MikeWyxtt>
Got the build script for my M3 MBA going, i am stumped on 1 issue: line 244 of asahi-installer/src/stub.py throws an exception "Failed to locate a usable build identity for this device") upon attempting to match some values in BuildManifest.plist (i think thats what is going on anyways..). I don't see this file anywhere, the only reference to it is in this source file. I googled it
<MikeWyxtt>
and most of what comes up is results pertaining to the IPSW files. I'm going to RTFM some more and search around, but does anyone know what causes this error/what is going on in the installl_files() method?
chrisl has quit [Ping timeout: 480 seconds]
<MikeWyxtt>
traced it back to load_ipsw(), disregard
chrisl has joined #asahi-dev
pb17 has quit [Ping timeout: 480 seconds]
chrisl has quit [Ping timeout: 480 seconds]
<yuka>
is it allowed to share findings based on grepping through XNU source dumps here?
<yuka>
I'm not sure if the apple source license is giving any problems
<yuka>
I was looking at why the cpu_init function hangs the t8132
pb17 has joined #asahi-dev
jnn has joined #asahi-dev
jn has quit [Ping timeout: 480 seconds]
mps has quit [Ping timeout: 480 seconds]
jn has joined #asahi-dev
jnn has quit [Ping timeout: 480 seconds]
<chaos_princess>
Here - you probably shouldn't. But should be fine in #asahi-re
chrisl has joined #asahi-dev
<MikeWyxtt>
chaos, when you had mentioned to me earlier that the scripts need patching, did anyone ever go through and fully patch it? I have it patched now to use the 14.6.1 IPSW, added the device/cpu id stuff so it just runs and works. Well, it created a bootable USB stick with no errors that is
<MikeWyxtt>
this is regarding resolving the errors the script throws when attempting to run it on an M3 MBA
<chaos_princess>
it created a what?
<MikeWyxtt>
i have a usb flash drive, i had the script use that as the target media
<chaos_princess>
which script
<MikeWyxtt>
it booted up and the boot option is showing m1n1 proxy :)
<chaos_princess>
i am not sure as to how you installed it to usb, since you are only supposed to install to internal ssd, but sure?
chrisl has quit [Ping timeout: 480 seconds]
<chaos_princess>
the "installation method" that people are using for development, is to install a second copy of macos using a full installer, and then to "overwrite" the macos kernel with m1n1 using kmutil
<chaos_princess>
installer kinda does the same thing, but only installs macos firmware, instead of full macos
<MikeWyxtt>
i dunno, i just had my fd plugged in and it gave me the option to so i tried it.. i am referring to the asahi-install util as a whole. i modified a few files to get it to run
<MikeWyxtt>
now that it runs i intened to follow the developer guide
<chaos_princess>
weird, it should not allow you to install to external, but sure
<MikeWyxtt>
when it asks for a volume, i pressed 'd' and selected my drive
<chaos_princess>
i mean, if it is not due to you patching out a check or something - it is a bug, m1n1 does not have usb driver and will not be able to load stage 2
<chaos_princess>
also, you did the part that is done after the reboot, right? the 1tr bit
<MikeWyxtt>
part of my goal today was to not remove anything as to make it something that would be accepted upstream
<MikeWyxtt>
the only things i added was a new IPSW_VERSION entry in main.py, the entry in DEVICES in main.py
<chaos_princess>
ah, you are running expert mode
<chaos_princess>
ok, its fine then
<MikeWyxtt>
and to get the bootstrap.sh script to work in scripts/ i changed the https://localhost:5050 to file://
<MikeWyxtt>
yeah i seen marcan said this kinda stuff has to be flagged expert only
<MikeWyxtt>
so i set the boolean to true for everything i did
<chaos_princess>
i would suggest wiping and re-installing to internal, unless you like having more random issues than neccessary
<MikeWyxtt>
no not at all lol, just wanted to see if it would do anything at all
<MikeWyxtt>
it just says Running proxy... after it rebooted btw, not sure if that answered your question
<chaos_princess>
if it has the m1n1 logo, some debug spew and running proxy - it means it worked.
<MikeWyxtt>
yes it does. says no valid payload found, some stuff about the serial link and then i believe its just waiting for me to attach my host device to it
<MikeWyxtt>
idek how to use m1n1 so thats my next move digging into the wiki
<chaos_princess>
pretty much. you can now use proxyclient from another computer to poke at the hardware, or load a kernel under a debugger
<MikeWyxtt>
yes. super excited for that. should i submit a PR for the changes i made to get the installer script to work like this? or no?
<MikeWyxtt>
and ty!
<chaos_princess>
unless it is something other than adding machines - no. if a machine is listed as a supported in the installer, it implies that it is ready to be used by a normal user, which is very much not the case for m3+
<chaos_princess>
also, not quite sure, but you _might_ want to reinstall using specifically firmware 15.1
<MikeWyxtt>
how do you mean? i set it up so that unless the user is in expert mode it fails and says unsupported? Is that acceptanle? or this needs to sit in my github until a later date
<chaos_princess>
lets just say that expert mode is hidden for a reason, there are too many people who think of themselves as experts :P
<MikeWyxtt>
i tried 15.1 but they removed the 'RestoreBundlePath' key from one of the plists in the IPSW so i just used 14.6.1 because i didn;t have to modify anything out of the ordinary
<chaos_princess>
depending on what hw you want to poke at 14.6 might be fine
<MikeWyxtt>
also, the values in min SFR and iboot version are garbage no clue what to put there
mps has joined #asahi-dev
<chaos_princess>
just don't bother for now.
<MikeWyxtt>
okay
<chaos_princess>
when "official" support comes out, it will most likely be based on at least 15.2 (depends on whether apple does certain things in it, and when exactly it will be ready)
<MikeWyxtt>
ahh that makes sense. i wasn't really sure how the versions were picked, i assumed perhaps it was the version the device was shipped with or something
<MikeWyxtt>
well for what its worth they shuffled around the bootcaches plist in usr/standalone
<MikeWyxtt>
in 15.1.1
<chaos_princess>
in fact - the version that is shipped with the device should to be avoided, since there are two revisions of 15.1 - the one for m4, and the one for everything else. it was previously done that way for m2, which had a special version of 12.4, and i think, people do not want to repeat that
<MikeWyxtt>
yes i noticed that on mrmachintosh when i was picking thru the versions. some were marked M3 ONLY or M4 ONLY. very odd and way over my comprehension
<MikeWyxtt>
my logic doesnt make sense because i picked 14.6.1 so yeah ill just leave it lol
<chaos_princess>
its just apple being apple, and being secretive about specific changes in their new products even from other teams that work on their os
<MikeWyxtt>
yes. the work thats been done so far in spite of that fact by people with no docs or anything to reference is fascinating
opticron has quit [Read error: Connection reset by peer]
opticron has joined #asahi-dev
kokoschmoko has joined #asahi-dev
<MikeWyxtt>
tethered boot setup is a go. last question, is there a particular branch of the asahi kernel fork i should clone and build? or just clone main and go?
<j`ey>
MikeWyxtt: there's tags made
<chaos_princess>
asahi-wip
<MikeWyxtt>
ok i think that makes sense. thx
pb17 has quit [Ping timeout: 480 seconds]
<mps>
just for info, yesterday when building mesa-asahi 20241128 I've got this warning https://tpaste.us/oxvg
<mps>
though it works fine
pb17 has joined #asahi-dev
chrisl has joined #asahi-dev
chrisl has quit [Ping timeout: 480 seconds]
<jannau>
MikeWyxtt: I'd recommend 14.x for m3 for now. It's easier for now since the SSD has a incompatible firmware change in 15.1. as long as you don't start working on firmware driven HW support (dcp, gpu, isp) 14.6 is fine
<jannau>
that tree boots on the m3 14" macbook pro. There is some issue with the display power domains. the display will stay on with "pd_ignore_unused" on the kernel command line and removing the power-domain from chosen.framebuffer in the device tree
hightower3 has joined #asahi-dev
<MikeWyxtt>
I see. I have a 13" M3 MBA
<jannau>
has that still i2c hpm (usb-c PD/port controller (ACE)) or spmi ones?
<MikeWyxtt>
let me check
hightower2 has quit [Ping timeout: 480 seconds]
<MikeWyxtt>
jannau: How do you find that? i dont see that in system profiler or from dmesg
<jannau>
the ADT, apple's devicetree
<MikeWyxtt>
dumping it, one moment
kokoschmoko has quit [Ping timeout: 480 seconds]
<MikeWyxtt>
i have the adt dumped now from ioreg. not as suggestive as i would have liked, what im i searching for here specifically?
chrisl has joined #asahi-dev
chrisl has quit [Ping timeout: 480 seconds]
<chadmed>
dump it from m1n1
<chadmed>
youll get more readable and useful information since m1n1 handily offsets addresses etc for you
<jannau>
search for hpm or usbc,sn201202x,spmi which answers the question already
<chadmed>
do we expect some m3 devices to still have the old i2c chip?
<jannau>
I was thinking of the adt from the device via `python3 -m m1n1.adt -r adt.bin` or the templates from the IPSW Firmware/all_flash/DeviceTree.j613ap.im4p
<jannau>
chadmed: not expect, j504 (14-inch M3 mbp) still has the i2c ones
<chadmed>
oh wow okay
<chadmed>
so we know which devices get finalised first in the product cycle :p
<jannau>
that one shipped with macos 13.5
<chadmed>
[16:21] <chadmed> jannau: do you remember which igt test failures you were concerned about? <- also sorry for the bump but im not gonna have time to look at this again for a while
<chadmed>
the failures i got are all red herrings (tests that should have been skips or tests that failed because igt tried to swap a 0x0 rect which we can never support)
<jannau>
I don't remember exactly. it was a basic overlay test and I was mostly concerned about the overlay parameters. iirc it well of a reasonable display area. and our off active area broke that obviously
<jannau>
did I add the face CRC support to bits/200-dcp?
<chadmed>
doesnt look like it
<chadmed>
last bit of work was stabilising edid audio
<chadmed>
edid and audio*
sand4rt has joined #asahi-dev
pb17 has quit [Ping timeout: 480 seconds]
chrisl has joined #asahi-dev
pb17 has joined #asahi-dev
chrisl has quit [Ping timeout: 480 seconds]
greguu has quit [Ping timeout: 480 seconds]
greguu has joined #asahi-dev
hightower4 has quit [Remote host closed the connection]
V has joined #asahi-dev
hightower2 has joined #asahi-dev
hightower3 has quit [Ping timeout: 480 seconds]
V has quit [Ping timeout: 480 seconds]
<yuka>
Mary: with a few changes, like adding the early return to cpu_init, adding an appropriate code path for t8132 in mcc.c, and commenting out the gxf_init, the usb driver did "just work"
<jannau>
I think Daniel already skipped the i2c init for m3 pro/max devs with spmi usb-c prot controller and usb2 just works
<Mary>
nice, well I guess the custom usb pd breakout board I just receive today will be useless then ^^
<Mary>
not too sure what to tackle anyway, maybe it would make more sense if I could help on m1 stuffs
<yuka>
I couldn't get chainloading another m1n1 to work yet