#asahi-dev · 2024-10-15 08:42
ariejanHi. New here. I know microphone support for M1 Max mbps is WIP - but where? I'm not sure which repo to look. Looking to see if I can lend a hand in some awy. #asahi-dev · 2024-10-14 14:18
nickchanelse it won't work (so linux needs to be modified)#asahi-dev · 2024-10-14 14:18
nickchancommand is 64-bit while status is 32-bit#asahi-dev · 2024-10-14 14:18
nickchancommand = 0x202220020, status = 0x202220050#asahi-dev · 2024-10-14 14:17
nickchanfound a9 cpufreq#asahi-dev · 2024-10-14 12:57
nickchanCan\'t find a PCore frequency which is lower than the Ecore frequency we are trying to achieve\n#asahi-dev · 2024-10-14 12:57
nickchan*strings*#asahi-dev · 2024-10-14 12:57
nickchana11 cpufreq confirmed to be cursed#asahi-dev · 2024-10-12 12:00
nickchanpmgr adt on older socs seems to have a slightly different struct i will have to fix that#asahi-dev · 2024-10-10 19:30
beastdSorry for the noise, but the gaming capabilities announced in the blog today sound nearly to good to be true. Big thanks to all devs involved!#asahi-dev · 2024-10-10 14:02
nickchanonly needs to be 8 big though, aic1 will never be bigger than that#asahi-dev · 2024-10-10 13:37
nickchani will just add a logical map then#asahi-dev · 2024-10-10 13:34
nickchanin aic 1#asahi-dev · 2024-10-10 13:34
nickchanokay wait it is also used for affinities#asahi-dev · 2024-10-10 13:24
nickchanmaybe it is fine if I just if the warning on !use_fast_ipi#asahi-dev · 2024-10-10 13:24
nickchanIn fact the warning is only there for AIC version 1#asahi-dev · 2024-10-10 13:23
nickchanthe system i am talking about is fast ipi capable#asahi-dev · 2024-10-10 13:23
nickchanonly in slow ipi aic gets the logical cpu id#asahi-dev · 2024-10-10 13:23
nickchanin fast ipi aic gets a value based on the cpu mpidr#asahi-dev · 2024-10-10 13:21
nickchanI looked at the aic drivers again it seems the mismatch only mattered if slow ipi is being used#asahi-dev · 2024-10-10 13:19
nickchanall I need is a mapping between AIC_WHOAMI and smp_processor_id#asahi-dev · 2024-10-10 13:18
nickchanokay i think a10 wouldn't explode yet#asahi-dev · 2024-10-10 12:56
nickchanokay, not right now because cpufreq doesn't exist#asahi-dev · 2024-10-10 12:53
nickchanso I guess I need to add another hack#asahi-dev · 2024-10-10 12:53
nickchanmaz: if i fix this then a10 would explode#asahi-dev · 2024-10-10 12:50
maz*logical* CPU0. normally, this should be indirected via the cpu_logical_map array, and that's missing.#asahi-dev · 2024-10-10 12:38
nickchanwhich causes issues#asahi-dev · 2024-10-10 12:38
nickchanthe kernel always believes it boots on cpu0#asahi-dev · 2024-10-10 12:38
nickchannot sure how to resolve it though#asahi-dev · 2024-10-10 12:38
nickchanI am running into this warning because the system boots on CPU4 (pcluster core 0)#asahi-dev · 2024-10-10 12:37
nickchanhttps://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/drivers/irqchip/irq-apple-aic.c#n865#asahi-dev · 2024-10-08 08:56
maznickchan: my comment was about M1+ CPUs, indicating why CMOs don't need to do much.#asahi-dev · 2024-10-08 08:32
nickchanso what I see is XNU set the bit to avoid problems but when doing cache maintanence unset it, do the instructions, then set it again#asahi-dev · 2024-10-08 08:32
nickchanBut cache maintenance on a7 needs the bit unset it#asahi-dev · 2024-10-08 08:31
nickchanso it is more likely that something is going to be broken without the bit set#asahi-dev · 2024-10-08 08:30
nickchanmaz: these bits are set by xnu on every core as well even on a7#asahi-dev · 2024-10-08 08:24
mazmay also snoop non-cacheable accesses.#asahi-dev · 2024-10-08 08:23
maznickchan: could be related to the fact that recent SoCs don't actually need CMOs to the PoC as all devices are cache coherent.#asahi-dev · 2024-10-08 05:59
nickchanon all cores#asahi-dev · 2024-10-08 05:58
nickchanmarcan: so... why is HID4_DISABLE_DC_MVA set by default anyways#asahi-dev · 2024-10-08 05:36
nickchanin summary: dc cvac does nothing and dc civac does nothing unless a bit in a register is flipped#asahi-dev · 2024-10-08 05:29
nickchanI *don't* think this needs a workaround in linux yet#asahi-dev · 2024-10-08 05:18
nickchanI wish#asahi-dev · 2024-10-08 05:18
nickchan> Cyclone UM 5.3.8.3#asahi-dev · 2024-10-08 05:18
nickchanhttps://github.com/asdfugil/m1n1-idevice/blob/dddc18bca071fd98dd47ec6f63c7ca042f5007a1/src/utils.c#L213#asahi-dev · 2024-10-08 04:58
nickchannot just at init#asahi-dev · 2024-10-08 04:58
nickchanthat needs runtime tunable flipping#asahi-dev · 2024-10-08 04:57
nickchanokay yeah it's an erratum#asahi-dev · 2024-10-08 04:38
nickchana lot of corellium-isms turns out to be required... on these old cores#asahi-dev · 2024-10-08 02:41
nickchancurrently dealing with some cache issues on A7 that might be a Cyclone erratum...