#
#asahi-dev · 2026-01-03 23:51
fl0_idI could do a video just to be sure
#asahi-dev · 2026-01-03 23:51
fl0_idwhen experimenting sometimes I got the rainbow apple logo
#asahi-dev · 2026-01-03 23:51
fl0_idI think first I see the asahi log with some logs. then m1n1 restarts
#asahi-dev · 2026-01-03 23:50
fl0_idright. thanks about clarifying.
#asahi-dev · 2026-01-03 23:47
integralpilot*EL1 m1n1
#asahi-dev · 2026-01-03 23:47
integralpilotbtw, when the error happens, do you see an asahi logo with some logs, or a rainbow apple logo? wondering if m1n1 has loaded yet or not
#asahi-dev · 2026-01-03 23:41
integralpilotwould that change anything though?
#asahi-dev · 2026-01-03 23:37
jannauwe probably want m1n1 to switch to cpu0 in kboot.c
#asahi-dev · 2026-01-03 23:35
integralpilotsorry, rather, that's the one iBoot loaded m1n1 on
#asahi-dev · 2026-01-03 23:35
integralpilotalso, the reason it's happening on cpu4 is because that's just the one m1n1 picked to boot off (based on the ADT). linux will also probably be started on cpu4 by m1n1 as that's it's primary one. because linux only sees one CPU in it's FDT (cpu0) it will not try starting any other CPUs and just stay on cpu4. technically you could change the one CPU in the linux FDT to cpu4 but that wouldn't change anything
#asahi-dev · 2026-01-03 23:33
integralpilotm1n1 doesn't use your linux FDT as the source of truth, it uses the ADT, so modifying the linux device tree won't change anything
#asahi-dev · 2026-01-03 20:38
fl0_id_no, seems to be correct.
#asahi-dev · 2026-01-03 20:31
fl0_id_the t6031 chickens were originally added by dberlin, in the commit message they also said the early uart base changed, but this is exactly where it's currently failing for me. could it be possible that this is not always the same? maybe I'll just try some other ones or it's because of some other issue...
#asahi-dev · 2026-01-03 20:16
fl0_id_playing around a bit more with everest chickens, nothing else seems to effect this so far. neither positively or negatively, even without any chickens it still gets to the same place in the startup sequence
#asahi-dev · 2026-01-03 19:55
fl0_id_but prob these are not related
#asahi-dev · 2026-01-03 19:54
fl0_id_as this seems to fall within VUART, could it help to add uart to the device tree?
#asahi-dev · 2026-01-03 19:42
fl0_id_now it fails after some pmgr stuff and with SERROR
#asahi-dev · 2026-01-03 19:41
fl0_id_hey, so I commented the two t6031 specific chickens before the init_common_everest and that got me quite a bit further afaict https://www.pastery.net/gpnjmq/
#asahi-dev · 2026-01-03 19:14
chaos_princesslet's go
#asahi-dev · 2026-01-03 17:42
svenstill great to get started 😁
#asahi-dev · 2026-01-03 17:22
jannaunot that there is yet much to upstream without gem-shmem and gpuvm
#asahi-dev · 2026-01-03 17:21
jannauI guess I'll have to start upstreaming the driver now
#asahi-dev · 2026-01-03 17:19
jannaum1n1 #466 (gpu initdata creation) merged, thanks chaos_princess
#asahi-dev · 2026-01-03 15:06
chaos_princess14.3.X may reduce the patches needed
#asahi-dev · 2026-01-03 14:48
fl0_idwell that too (bth I didn't think about what it was patching)
#asahi-dev · 2026-01-03 14:45
noopwafelright, patching macos not very relevant for linux boot I guess :D
#asahi-dev · 2026-01-03 14:44
fl0_idthis is also without binary patching, as it was said this was only necessary for macos HV iirc
#asahi-dev · 2026-01-03 14:44
noopwafelfl0_id: are you doing the binary patching (w/updated offsets), or the same situation?
#asahi-dev · 2026-01-03 14:44
fl0_idin case you were talking to me...
#asahi-dev · 2026-01-03 14:43
fl0_idchaos_princess this should be 14.8.3, the one recommended by integralpilot. (as I couldn't get 14.7.7, before it was 14.6.x or so)
#asahi-dev · 2026-01-03 14:42
noopwafelI was trying 14.8.3 (without the binary patching from integralpilot's notes, so p'rhaps expected, we were just trying to confirm what we could trap)
#asahi-dev · 2026-01-03 14:35
chaos_princessjust to make sure - which macos?
#asahi-dev · 2026-01-03 14:35
noopwafelelr 0x200 looks very healthy.. yesterday we were reliably getting corrupted code in arm_init on an M3 pro, but it failed considerably later
#asahi-dev · 2026-01-03 14:31
fl0_id(or actually I think I checked an old branch, so the current next register may be different...will check properly later)
#asahi-dev · 2026-01-03 14:26
fl0_idgonna look into that more later today
#asahi-dev · 2026-01-03 14:26
fl0_idassuming the HID14 is reg_set(SYS_IMP_APL_HID14, HID14_ENABLE_NEX_POWER_GATING); then I assume one of the chickens fails, msr(s3_1_c15_c1_5, 0x3uL); would probably be next
#asahi-dev · 2026-01-03 14:16
fl0_idjannau right that makes sense. wasn't sure about the numbering
#asahi-dev · 2026-01-03 13:26
jannaum3 max has 4 efficiency cores? M3* dev start on the first performance core which would be cpu4 (from m1n1's pov). linux defines the boot cpu as cpu0 independent on device tree entries
#asahi-dev · 2026-01-03 13:20
fl0_id(I don'T mean the uarterror at the bottom)
#asahi-dev · 2026-01-03 13:20
fl0_idabout booting m3 max - is it normal that if a device tree only has cpu0, that an exception can still occur on cpu4? or could it very likely also be a HV error? (https://www.pastery.net/erjhwn/)
#asahi-dev · 2026-01-03 13:18
jannauyes, it's likely that it'll work. There were some changes necessary to get it working on M1 Pro and Max but not much between M1* and M2* (ignoring dptx for the hdmi port on desktop devices)
#asahi-dev · 2026-01-03 12:56
chadmedthats not guaranteed, but we dont expect there to be any
#asahi-dev · 2026-01-03 12:56
chadmedas long as there have been no insane soc specific changes, yes
#asahi-dev · 2026-01-03 12:37
integralpilotdo you think supporting 14.x DCP firmware for M1/M2 will work for M3 as well, like how my newer SMC firmware support for M3 works on M1/M2 as well, or would have apple changed the DCP and firmware API in M3 even on the same base macOS version?
#asahi-dev · 2026-01-03 12:33
JamesCalligeros[m]Having said that i wouldnt say no to a PR
#asahi-dev · 2026-01-03 12:32
JamesCalligeros[m]Yeah exactly, its a job for when a firmware upgrade becomes necessary (again)
#asahi-dev · 2026-01-03 12:25
chaos_princessShouldn't actually be that hard to do, as we can run a part of the install flow, then "repair m1n1", and it should work. But no real reason to bother writing it rn
#asahi-dev · 2026-01-03 12:05
chadmedmaybe once the installer can in-place upgrade an apfs container we can drop it
#asahi-dev · 2026-01-03 12:05
nickchanalarm is alt distro now
#asahi-dev · 2026-01-03 12:04
chadmedeven then it will just be old old alarm users

plus ›