#
#asahi-dev · 2025-06-07 21:36
svenright, so we can simulate that situation on m1/m2/m3 already by just disabling it ourselves
#asahi-dev · 2025-06-07 21:35
sven+you
#asahi-dev · 2025-06-07 21:35
amarioguyohhhh i see the dilemma now ugh
#asahi-dev · 2025-06-07 21:35
svenhow would run xnu without gxf/sprr?
#asahi-dev · 2025-06-07 21:35
svenhow?
#asahi-dev · 2025-06-07 21:35
amarioguyik that on m4 apple makes it so that you either run under SPTM at all times if you specify a mach-o to kmutil, or you don't get GXF at all if you do raw binary instead, but wouldn't tracing be possible in the latter case anyways?
#asahi-dev · 2025-06-07 21:34
amarioguyor the opposite, where you need it jettisoned entirely?
#asahi-dev · 2025-06-07 21:33
amarioguysven: for m4 bringup, i assume the problem is mainly that you need GXF for m1n1 to be able to do HV tracing?
#asahi-dev · 2025-06-07 21:25
amarioguyfl0_id: hm - i guess you could try reading ICH_VTR_EL2? (would be a mrs(ICH_VTR_EL2) in m1n1 console) would verify if the CPU interface exists at least
#asahi-dev · 2025-06-07 20:49
svenshould be possible but painful, might need nested virtualization
#asahi-dev · 2025-06-07 20:49
sveni.e. emulate SPRR/GXF and all other apple instructions somehow
#asahi-dev · 2025-06-07 20:49
svenamarioguy: if you (or anyone else) wants to work on m4 support try disabling SPRR/GXF on M1/M2/M3 and make XNU boot under the hv. iirc there's even one kernel that also comes with SPTM
#asahi-dev · 2025-06-07 20:02
fl0_idI can also try updating my stage1 m1n1 / check if there were any changes or generally try to check uart
#asahi-dev · 2025-06-07 20:01
fl0_idamarioguy I guess so? you know more then me tbh. do you think it would make sense to try other registers, to check if it's just sth about this on or sth else?
#asahi-dev · 2025-06-07 19:31
miyuh_sweet thank you
#asahi-dev · 2025-06-07 19:31
chaos_princesssure. anyway, installing would be a good step, here is the support matrix, pick something that you find interesting https://asahilinux.org/docs/platform/feature-support/m1/
#asahi-dev · 2025-06-07 19:30
miyuh_that's fine with me]
#asahi-dev · 2025-06-07 19:29
chaos_princessso, i know it might sound unkind, but maybe pick an easier project? This one is unfortunately one of "you are thrown right into the deep end of the pool, figure it out yourself" ones.
#asahi-dev · 2025-06-07 19:28
miyuh_to be honest i havent even installed it yet so maybe i should start there.. i am interested in getting into driver development though!
#asahi-dev · 2025-06-07 19:26
chaos_princessis there a specific missing feature that you are interested in?
#asahi-dev · 2025-06-07 19:22
miyuh_hey all im new here.. i love linux and open source and i wanna contribute. what's the best way to get started?
#asahi-dev · 2025-06-07 19:21
miyuh_a
#asahi-dev · 2025-06-07 19:03
amarioguyright, heard of that one, for now, i'll probably just hard limit my support to m1/m2 and pro equivalents until i get more time to properly examine the m4 case
#asahi-dev · 2025-06-07 19:01
svendepending on your firmware version it may not boot at all due to an apple bug
#asahi-dev · 2025-06-07 19:01
svendunno
#asahi-dev · 2025-06-07 19:01
amarioguysven: does m4 at least boot to m1n1 shell as of rn?
#asahi-dev · 2025-06-07 18:56
amarioguywhich m1n1 then recovers from
#asahi-dev · 2025-06-07 18:56
amarioguythe uart timeout is concerning tho - because an undefined MSR would throw an exception
#asahi-dev · 2025-06-07 18:55
amarioguyfl0_id: saw the results - that register should be well defined on the platform?
#asahi-dev · 2025-06-07 10:28
fl0_idsven tracing is prob a good idea, but I think I don't have a setup yet for that. (so far I only have avery basic second macos container, which is now set up to boot m1n1 and would have to check how much work it would be to enable a ful hv / tracing setup on that
#asahi-dev · 2025-06-07 10:27
fl0_id*is
#asahi-dev · 2025-06-07 10:27
fl0_idsven ok. it might be, I don't know enough to know which error count that s
#asahi-dev · 2025-06-07 09:51
svenso maybe also do a few boots without resetting it and confirm that it's incremented
#asahi-dev · 2025-06-07 09:51
svenand isn't the error count that thing that increases automatically?
#asahi-dev · 2025-06-07 09:50
sveni'd probably trace the pmu using the hv and see what values macos updates and if they match that new scrpad offset
#asahi-dev · 2025-06-07 09:49
svenah, sure, that might help
#asahi-dev · 2025-06-07 09:48
fl0_idto try to understand the data structure that it's reading before it breaks. to see what's what. Like just locally of course for trying things. sorry if I am misunderstanding something
#asahi-dev · 2025-06-07 09:37
svenadding print statements? how is that going to help?
#asahi-dev · 2025-06-07 09:36
fl0_idsven mmh thx that already helps. though not sure how. maybe for starters I'd just add some print statements...
#asahi-dev · 2025-06-07 09:33
svenfl0_id: you'd need to understand what actually changed and if that new way is indeed correct and then make it conditional on the compatible
#asahi-dev · 2025-06-07 09:25
fl0_id@amarioguy - oh and if I tried to run it again, it would not give an error / timeout but just not respond and I'd have to close the shell forcefully
#asahi-dev · 2025-06-06 23:47
fl0_idalso, I updated m1n1 python console etc to the latest if there were any changes - but it seems that some errors for m3 at least remain. f.e. I had to reapply my hack: https://github.com/AsahiLinux/m1n1/issues/449 - otherwise shell and ishell.py crash. Could I make a PR for my bypass, or can someone point me in the right direction to otherwise fixing it?
#asahi-dev · 2025-06-06 23:34
fl0_idamarioguy I got this as a result, though I have to check if sth else is wrong maybe too. https://www.pastery.net/praejy/
#asahi-dev · 2025-06-05 16:35
nickchanseems to be fine
#asahi-dev · 2025-06-05 14:52
ketteniscan you check that fixes your issue?
#asahi-dev · 2025-06-05 14:52
kettenisnickchan: https://github.com/AsahiLinux/m1n1/pull/473
#asahi-dev · 2025-06-04 14:32
nickchanthe stage prior to m1n1 accepts usb uploads so it's a logical thing to do during development
#asahi-dev · 2025-06-04 14:27
nickchankettenis: using m1n1.bin and loading kernel appended to stage1
#asahi-dev · 2025-06-04 13:38
kettenisah, wait I guess you're loading a kernel directly from the stage1 m1n1
#asahi-dev · 2025-06-04 13:28
kettenisnickchan: are you using the .macho?

plus ›