#asahi-dev · 2023-06-17 04:56
marcanshipping overlays is a minefield, because they will break us any time by doing dependency bumps (not that they don't break themselves too, but still)#asahi-dev · 2023-06-17 04:56
marcanjannau: at this point the answer might be "don't bother with touchbar support on ALARM" tbh...#asahi-dev · 2023-06-16 07:47
marcansven: awesome!#asahi-dev · 2023-06-15 12:30
marcanso maybe we should try merging just the core USB3 rework soonish#asahi-dev · 2023-06-15 12:30
marcanpeople keep reporting fail related to that, but it sounds like a waste of time to debug the current state of affairs#asahi-dev · 2023-06-15 12:30
marcansven: what's the state of your USB3/ATCPHY rework, ignoring alt modes?#asahi-dev · 2023-06-15 11:37
marcanno complaints, guessing it's safe to push :p#asahi-dev · 2023-06-15 04:43
marcanpackages in asahi-dev now#asahi-dev · 2023-06-15 04:25
marcanI think I'll push the kernel straight to stable since it's a strict bugfix; mesa will go to asahi-dev for a bit, please smoke test and I'll push it to stable later today#asahi-dev · 2023-06-15 04:25
marcanbuilding new kernel/mesa with the fixes from alyssa/lina from yesterday#asahi-dev · 2023-06-13 13:14
marcanI'm not particularly interested in lemmy#asahi-dev · 2023-06-13 13:14
marcanwe're going to be on some Discourse instance most likely#asahi-dev · 2023-06-10 15:25
marcanjannau: awesome :D#asahi-dev · 2023-06-09 13:27
marcanand t602x just never gets it#asahi-dev · 2023-06-09 13:27
marcanwe should probably just make m1n1 remove that from the DT on t8112 if the version is new enough#asahi-dev · 2023-06-09 13:27
marcanyeah#asahi-dev · 2023-06-09 13:24
marcanso that suggests they added hardware sequencing or something to t8112 to avoid having to deal with this, and only started using it later, maybe#asahi-dev · 2023-06-09 13:24
marcant8103 still has it#asahi-dev · 2023-06-09 13:20
marcanlet me look at t8103...#asahi-dev · 2023-06-09 13:19
marcanwhich suggests they also removed it on plain M2#asahi-dev · 2023-06-09 13:19
marcansven: performEFuseProgramming is empty on 13.2 for t8112#asahi-dev · 2023-06-09 13:13
marcan(hypervisor)#asahi-dev · 2023-06-09 13:13
marcansven: macos doesn't read them#asahi-dev · 2023-06-09 13:11
marcansven: https://mrcn.st/p/sAzIpA3d#asahi-dev · 2023-06-09 10:14
marcanI wouldn't be surprised if it requires hardware sequencing or something we don't have on t8103 though#asahi-dev · 2023-06-09 10:13
marcansven: yes#asahi-dev · 2023-06-08 00:13
marcantpw_rules: it is not#asahi-dev · 2023-06-07 16:48
marcanI ask this because someone mentioned the ambient light sensor, and I noticed it's under AOP#asahi-dev · 2023-06-07 16:47
marcanas long as we keep both options open#asahi-dev · 2023-06-07 16:47
marcanpovik: fair#asahi-dev · 2023-06-07 16:02
marcanbut that's something we need to validate#asahi-dev · 2023-06-07 16:01
marcanif you can still boot AOP and explicitly choose not to touch the audio stuff and that still works though, then it's fine#asahi-dev · 2023-06-07 16:00
marcanI'm perfectly okay with bypassing firmware where it only does one thing, but I don't think we want to paint ourselves into the corner of bypassing firmware in a way that locks us out of multiple things we don't know about yet...#asahi-dev · 2023-06-07 15:59
marcane.g. the ambient light sensor is hooked up there, and it's entirely possible we need AOP to make full system suspend work#asahi-dev · 2023-06-07 15:58
marcanbecause if so... I think that's not going to work#asahi-dev · 2023-06-07 15:58
marcandoes turning it on interfere with what you did?#asahi-dev · 2023-06-07 15:58
marcanpovik: so the issue is AOP does more things#asahi-dev · 2023-06-07 14:40
marcanpovik: you weren't planning on bringing up AOP at all, just bypass it and talk straight to the DSP thing?#asahi-dev · 2023-06-07 11:28
marcanI just cannot believe how woefully undermaintained it is, given the maintainer works for Broadcom#asahi-dev · 2023-06-07 11:27
marcanthat driver is just cursed#asahi-dev · 2023-06-07 11:27
marcanI should've known better too, since I'd already fixed the race condition before... before implementing suspend regressed that by changing how IRQs work. I should've fixed it the equivalent way instead of trying something different.#asahi-dev · 2023-06-07 11:26
marcanit was a last minute race condition fix for the wifi driver and I didn't test suspend...#asahi-dev · 2023-06-07 11:26
marcanI broke it and I fixed it ;)#asahi-dev · 2023-06-07 00:36
marcanok, I think I fixed the wifi thing properly now#asahi-dev · 2023-06-07 00:10
marcanalso Pro/Max GPU support is not released anyway#asahi-dev · 2023-06-07 00:10
marcan*kaazoo#asahi-dev · 2023-06-07 00:10
marcankaprests: those are set by m1n1#asahi-dev · 2023-06-07 00:10
marcanand look at it later#asahi-dev · 2023-06-07 00:10
marcanI'm just going to revert that#asahi-dev · 2023-06-07 00:09
marcanI probably regressed it with the race bugfix... sigh.