#asahi-dev · 2026-08-24 16:56
svenalright, with some hacking the m3 nodes pass the checks as well#asahi-dev · 2026-08-24 16:39
svenbut I can't change the numbers because they need to line up with the ids in acio for the out of band notifications i'll need for dp and pcie#asahi-dev · 2026-08-24 16:38
sveni can just add the unused port then if I get complaints#asahi-dev · 2026-08-24 16:27
svenbut maybe i should also run the dt check against the m3 trees#asahi-dev · 2026-08-24 16:27
sveni thought i counted 24#asahi-dev · 2026-08-24 16:26
svenalright, so max only then#asahi-dev · 2026-08-24 16:25
sveni'm expecting complaints about that parent/child thing anyway but I really don't see any other way to model it#asahi-dev · 2026-08-24 16:24
svenack, let me check if that works#asahi-dev · 2026-08-24 16:23
svenso just minItems then?#asahi-dev · 2026-08-24 16:23
svenso the second one has the full example#asahi-dev · 2026-08-24 16:23
svenno, because it's always a child node of the second one#asahi-dev · 2026-08-24 16:19
svenalready added M3 while i was at it#asahi-dev · 2026-08-24 16:18
svendt-bindings: https://github.com/AsahiLinux/linux/commit/b89bff99c31fb9ccec8c1cad3508813fc03ca55b and https://github.com/AsahiLinux/linux/commit/6dba6ccf194a708eb366c714d4ffd6ceb259be24#asahi-dev · 2026-08-24 16:02
svenjannau: yeah, and i'm also not gonna find the time/justification to implement this in linux :D#asahi-dev · 2026-08-24 14:21
sven*paranoia#asahi-dev · 2026-08-24 14:21
svenand the step to 24 is just extra paranoid then: prevent the hosts in j_annau's crazy setup from messing with each other :D#asahi-dev · 2026-08-24 14:20
svenif they used that properly another host couldn't inject anything into the control ring#asahi-dev · 2026-08-24 14:20
svennow that i'm writing dt-bindings i think i just realized why they added the two darts: that allows isolation between control traffic and xdomain/host-to-host traffic#asahi-dev · 2026-08-24 11:34
svenwe don't break it intentionally either, but i always chainload before i run anything else#asahi-dev · 2026-08-24 11:33
svenyeah, we don't make any forward/backward compatible guarantees for the stuff inside proxyclient/#asahi-dev · 2026-08-24 10:36
svenI wonder why #asahi-dev · 2026-08-24 10:36
svenjannau: :(#asahi-dev · 2026-08-24 08:21
svenoh, nice#asahi-dev · 2026-08-24 08:07
svenpity#asahi-dev · 2026-08-24 08:07
svenso remap internally then probably just replays any invalidation across all SID separately#asahi-dev · 2026-08-24 07:58
svenah right, i wanted to check where xnu invalidates because likely we don't need that invalidate after map_pages ever#asahi-dev · 2026-08-24 07:53
svenlet's hope it's actually faster then :D#asahi-dev · 2026-08-24 07:53
svennice!#asahi-dev · 2026-08-24 07:42
svenmight be worth a try, i still think we could add that remap to just apple-dart.c without requiring any device tree changes#asahi-dev · 2026-08-24 07:05
svenregarding those dart streams, did you check if using remap makes it any better?#asahi-dev · 2026-08-24 06:26
svenack#asahi-dev · 2026-08-23 18:21
svenSame old rule for anyone who asks or complains that PCIe or DP doesn’t work yet once it’s merged: I’ll make sure to double the eta for that again!#asahi-dev · 2026-08-23 18:18
svenfeel free to pull it in #asahi-dev · 2026-08-23 17:23
sveni don't wanna do them, i just wanted to suggest to take a look at v2 and copy/paste from there :D#asahi-dev · 2026-08-23 17:21
svendt-bindings are a bit tricky if you haven't done them before, do we have some for pmpv2 already?#asahi-dev · 2026-08-21 20:37
svenhahahaha, amazing :D#asahi-dev · 2026-08-21 20:13
svenhttps://lore.kernel.org/all/2026081530-CVE-2026-72131-91e2@gregkh/ lol#asahi-dev · 2026-08-21 19:22
svenugh, so that's gonna be fun then#asahi-dev · 2026-08-21 19:20
svenhow do you know it's new?#asahi-dev · 2026-08-21 19:19
sveni'll probably get a m4 pro anyway, and m3 ultra is probably rare enough to just not bother#asahi-dev · 2026-08-21 19:18
sveni'll leave that as-is but may as well already add the compatibles for acio/nhi then#asahi-dev · 2026-08-21 19:18
svenmeh, the reset binding is already reviewed and i just sent v2#asahi-dev · 2026-08-21 19:17
svendid any m3 chip already have usb2v4? i suspect that will required changes but it may still be compatible enough#asahi-dev · 2026-08-21 19:16
svenif it works as-is on M3 i'll just add the compatibles to the initial submission as well#asahi-dev · 2026-08-21 18:56
svenwhat's so special about port2? :D#asahi-dev · 2026-08-21 18:50
svendid you need any changes to the tbt driver for m3 btw? or did everything just work#asahi-dev · 2026-08-21 18:39
sven... i just have no idea how to fix that#asahi-dev · 2026-08-21 18:39
svenso the grand parent is freed and then later that deferred works runs and calls that pm_runtime_whatever thing on the grandparent that's arleady free#asahi-dev · 2026-08-21 18:39
sventb_switch_removes runs during unplug and freed the switches which are parents of the xdomain netdev and that xdomain thing only pins its direct parent