#asahi-dev · 2025-05-04 18:01
svenhttps://github.com/AsahiLinux/m1n1/pull/460 which cleans up chickens.c could also use some testing as stage1 on various devices if anyone is bored#asahi-dev · 2025-05-04 18:00
svenwould be nice if someone could test https://github.com/AsahiLinux/m1n1/pull/447 on linux. seems to work fine on macos and with that i can finally get rid of my local patches that i keep stashing/unstashing to make proxyclient work here :D#asahi-dev · 2025-05-04 17:40
fl0_idsven thx. let's take it one step at a time ;)#asahi-dev · 2025-05-04 17:40
fl0_idsven ok#asahi-dev · 2025-05-04 17:24
nickchanerm I mean mailboxes for the cortex-a7 cores#asahi-dev · 2025-05-04 17:24
nickchancurrently calling the mailboxes "apple,t8015-marin-mailbox", "apple,marin-mailbox" since the mbox for the kingfisher cores is most definitely different#asahi-dev · 2025-05-04 17:20
nickchanand see which bits can be set#asahi-dev · 2025-05-04 17:20
nickchansart is something that could be figured out just by reading out what iboot configured the register as and then try writing 0xffffffff to it#asahi-dev · 2025-05-04 17:19
svenah well, don't think we'll ever support that one#asahi-dev · 2025-05-04 17:19
nickchana14 is v1 appearently #asahi-dev · 2025-05-04 17:18
sveni think M1 was already v2#asahi-dev · 2025-05-04 17:18
svenv0? huh#asahi-dev · 2025-05-04 17:16
nickchanmailbox - i2a and a2i control has different offsets that's about it; sart is sartv0 and uses a different set of allow flags and 36 bit pa#asahi-dev · 2025-05-04 17:12
nickchanworking now:)#asahi-dev · 2025-05-04 17:07
svenmight make sense to create apple-iphone.c or something like that then#asahi-dev · 2025-05-04 17:06
nickchanit's nvme but less bastardized than m1#asahi-dev · 2025-05-04 17:06
svenhow does command submission work then? more similar to regular nvme?#asahi-dev · 2025-05-04 17:05
svenugh, that's gonna require some invasive changes to nvme/apple.c then#asahi-dev · 2025-05-04 17:04
nickchanwhich will also make it upstreamable#asahi-dev · 2025-05-04 17:04
nickchanafter that i will fix some tech debt like usb not having phy and using bootloader state for bus configuration and such#asahi-dev · 2025-05-04 17:03
nickchan(in linux)#asahi-dev · 2025-05-04 17:03
nickchanspeaking of 8015 ans2 fw already successfully booted now dealing with some command_id/tag weirdness, and there's no nvmmu or linear sq#asahi-dev · 2025-05-04 17:00
svenagreed#asahi-dev · 2025-05-04 17:00
svenyeah#asahi-dev · 2025-05-04 17:00
sveni'd also drop pinctrl there#asahi-dev · 2025-05-04 17:00
nickchanDon't really need pinctrl for bringup I think but wdt can be good#asahi-dev · 2025-05-04 16:58
nickchannote we use clkref now not clk24#asahi-dev · 2025-05-04 16:58
nickchanthis one is newer https://github.com/AsahiLinux/linux/commit/d23457abac1e2f309404e63c7a390d9a32bce9e6#asahi-dev · 2025-05-04 16:48
svenonce something like that boots i'll even walk you through upstreaming that to LKML if you want to :)#asahi-dev · 2025-05-04 16:46
svenhttps://github.com/AsahiLinux/linux/commit/7d2d16ccf15d8eb84accfaf44a0b324f36e39588 that's how the original m1 mini one looked like#asahi-dev · 2025-05-04 16:45
svenoh, and framebuffer maybe if you want penguins#asahi-dev · 2025-05-04 16:45
svenalso skip all the pmgr nodes#asahi-dev · 2025-05-04 16:44
sveninitially you should only need aic and uart plus dependencies I think#asahi-dev · 2025-05-04 16:44
sventhat device tree already does way too much tbh :D#asahi-dev · 2025-05-04 16:42
fl0_idok thx for the pointers. I would have to get a diff device tree then the one linked above anyway, as I only have the m3 max#asahi-dev · 2025-05-04 16:42
fl0_idyeah understanding would be good#asahi-dev · 2025-05-04 16:41
fl0_idsven i haven't tried that yet but could maybe do that this week. right now I am just kind of sick and trying to distract myself haha#asahi-dev · 2025-05-04 16:40
svenand if that change is confirmed like that probably just gate it on the pmu compatible instead of isinstance(..., int) if possible#asahi-dev · 2025-05-04 16:39
svenand nothing's stopping you from playing around with m3, just add the device tree locally#asahi-dev · 2025-05-04 16:39
svenideally i'd like to understand what actually changed before merging that#asahi-dev · 2025-05-04 16:39
svenwhat if you just skip writing to that reset counter? does that change any behavior?#asahi-dev · 2025-05-04 16:38
fl0_idsven for now I would just be happy to play around with it haha#asahi-dev · 2025-05-04 16:37
fl0_idsven I have "confirmed" it in so far that with that, the dumping and other things work in m1n1. But I don't know enough to say if that introduces some weird errors or similar, that's why I am asking. If I remember correctly, mischa85 who did some m4 work also had to do a similar fix.#asahi-dev · 2025-05-04 16:35
svenwe're rebasing on upstream each release though so if you want m3 device tree in our kernel better send them to LKML ;)#asahi-dev · 2025-05-04 16:34
fl0_idsven yeah that's what I thought#asahi-dev · 2025-05-04 16:24
svenhow have you confirmed that you changes w.r.t. that pmu thing are correct?#asahi-dev · 2025-05-04 16:23
svenwe won't merge any m3 support into our kernel trees for now#asahi-dev · 2025-05-04 16:22
fl0_id?#asahi-dev · 2025-05-04 16:22
fl0_idalso wanted to check out booting a basic linux - but m3 device trees like this one https://github.com/a-ramses/linux/commit/f069376766475460a58c66d029c5dc7bb9086c6a haven't been merged in asahi-linux kernel yet, right=#asahi-dev · 2025-05-04 16:21
fl0_idgetting back to this - https://github.com/AsahiLinux/m1n1/issues/449 does it make sense to do a PR for this, or does this need some other ADT processing patches?