#
✕ nick:sven
#asahi-dev · 2026-07-06 19:04
sventhe efi table i build will only work for directly chainloaded kernels anyway
#asahi-dev · 2026-07-06 19:03
sven*adjust
#asahi-dev · 2026-07-06 19:03
svenokay, good. UEFI technically requires all memory descriptors to be 64KiB aligned but i guess we can get away with 16 KiB then and I don't have to just the ld scripts :)
#asahi-dev · 2026-07-06 19:00
sven is there any M-series chip that support 64 KiB pages?
#asahi-dev · 2026-07-06 08:03
svenprobably a bit more once we can do proper sleep
#asahi-dev · 2026-07-06 08:02
svenit’s maybe 100 instructions
#asahi-dev · 2026-07-06 08:02
svendo you need a mmu for psci stuff?
#asahi-dev · 2026-07-06 08:01
svenno, I mean, for what feature
#asahi-dev · 2026-07-06 08:01
svenwhat do you need 4levels for?
#asahi-dev · 2026-07-06 07:43
svensomeone got macOS running on m2 under sptm by virtualizating everything
#asahi-dev · 2026-07-06 07:15
svenas usual I’m gonna make it work for M1+ and not care about anything before and you get to have that fun then once it’s upstream :)
#asahi-dev · 2026-07-06 07:11
svenFirst goal is to just get an ack on the psci-via-efi approach
#asahi-dev · 2026-07-06 07:10
svenofc
#asahi-dev · 2026-07-05 18:29
svenor someone could figure out what xnu magic we are missing to be able to wake from that deep sleep state again :/
#asahi-dev · 2026-07-05 18:28
svenyeah i guess
#asahi-dev · 2026-07-05 18:28
svenjust catch all cores inside m1n1 first and once we have all of one cluster put them to sleep
#asahi-dev · 2026-07-05 18:27
sventhat's gonna be a bit hacky but it's possible to implement that i think
#asahi-dev · 2026-07-05 18:27
svenwe only have to keep one core alive, don't we?
#asahi-dev · 2026-07-05 18:26
svenwe can turn cores off and then on again, can't we?
#asahi-dev · 2026-07-05 18:25
svenwell, except for one core unless we figure out how to bring that one out of deep sleep :D
#asahi-dev · 2026-07-05 18:25
sventhis also supports idle state that lose all cpu state so we could even implement turning the cores off entirely with this i think
#asahi-dev · 2026-07-05 18:21
sventhis is totally gonna break with u-boot but that can come once this is actually accepted upstream
#asahi-dev · 2026-07-05 18:21
svenalright, suspend works and it ends up in m1n1's deep_wfi(). probably good enough to send this as an RFC for now
#asahi-dev · 2026-07-05 14:31
sventhe joys of mailing list based development :P
#asahi-dev · 2026-07-05 14:16
sventhat's all documented somewhere in Documentation/ iirc
#asahi-dev · 2026-07-05 14:15
svenit doesn't really matter since the dt validation scripts only have to pass on -next and not on each single commit
#asahi-dev · 2026-07-05 14:15
svenbut if there's no driver changes it's just nice to offer to take the change through our tree
#asahi-dev · 2026-07-05 14:15
sventhen the dt-binding change has to go through the respective tree together with the driver that needs the new compatible
#asahi-dev · 2026-07-05 14:01
svenalso please mention in the cover letter that we'd like to take every dt-binding that has an ack through the apple-soc tree since there are no driver changes and we only need them for the dts files themselves
#asahi-dev · 2026-07-05 13:20
sventime to clean up the m1n1 side now
#asahi-dev · 2026-07-05 13:20
svenhttps://git.kernel.org/pub/scm/linux/kernel/git/sven/linux.git/log/?h=efi-psci current kernel code, requires a surprisingly small diffstat
#asahi-dev · 2026-07-05 13:12
svenPSCI: CPU_SUSPEND(power_state=0x0, entry_point=0x0, context_id=0x0) <-- alright, psci-cpuidle seems to work as well and i really don't want debug prints there :D
#asahi-dev · 2026-07-05 13:04
sven:(
#asahi-dev · 2026-07-05 12:27
svensounds good to me
#asahi-dev · 2026-07-04 20:17
svenNice
#asahi-dev · 2026-07-04 16:21
sventhat’s what we get for keeping this downstream for so long :(
#asahi-dev · 2026-07-04 15:24
svenah, lol, because i don't define any idle states in the dt
#asahi-dev · 2026-07-04 15:18
svenso that works. cpuidle doesn't quite like what i do yet though
#asahi-dev · 2026-07-04 15:18
sven...
#asahi-dev · 2026-07-04 15:18
svenPSCI: CPU_ON: releasing CPU 1 (mpidr=0x1) to 0x818bf5510
#asahi-dev · 2026-07-04 15:18
sven[ 0.027371] CPU1: Booted secondary processor 0x0000000001 [0x611f0221]
#asahi-dev · 2026-07-04 15:18
sven[ 0.026227] smp: Bringing up secondary CPUs ...
#asahi-dev · 2026-07-04 14:04
svenhttps://git.kernel.org/pub/scm/linux/kernel/git/sven/linux.git/log/?h=efi-psci linux side, finally got it to stop crashing again
#asahi-dev · 2026-07-04 11:48
svenanyway, first need to finish this psci-via-efi crap. or well, fix it after i broke something after cleaning up :/
#asahi-dev · 2026-07-04 11:47
svenreally needs to get his hands on a m4 or m5 as well
#asahi-dev · 2026-07-04 11:47
svenmakes sense
#asahi-dev · 2026-07-04 11:46
svenprobably needs a small binary that reproduces it
#asahi-dev · 2026-07-04 11:46
sveni'm not even sure how we could file a feedback report for this tbh :/
#asahi-dev · 2026-07-04 11:39
svenodd
#asahi-dev · 2026-07-04 11:36
svenwell, that or somehow power is turned off for random peripherals because something else is missing

plus ›