#
✕ nick:sven
#asahi-dev · 2025-08-22 08:29
svenah okay
#asahi-dev · 2025-08-22 08:29
svenunless you have a gxf bypass that allows pagetable manipulation ofc
#asahi-dev · 2025-08-22 08:29
svencause then stuff like iommus are protected by gl2 and you can map that mmio region
#asahi-dev · 2025-08-22 08:28
svendoes that thing already have gxf?
#asahi-dev · 2025-08-22 08:22
sven„just“ do it like Apple: disable all drm keys if booting unsigned code
#asahi-dev · 2025-08-22 08:00
sven*move
#asahi-dev · 2025-08-22 08:00
svenbut I think I can even more the role switch mess into dwc3-apple.c with that
#asahi-dev · 2025-08-22 07:59
svenalso confused why the dt checks are failing, I remember running them before submitting and I didn’t get the errors :(
#asahi-dev · 2025-08-22 07:59
svenyeah, but that looks like it’ll make everything much cleaner
#asahi-dev · 2025-08-22 07:54
svenbut I guess that would mean dropping the snps,dwc3 compatible
#asahi-dev · 2025-08-22 07:51
svenohh.. that glue.h stuff the dwc3 maintainer suggested looks interesting. maybe I can move all the crap this thing needs to dwc3-apple.c
#asahi-dev · 2025-08-22 07:08
svenand only deal with that situation if it really happens
#asahi-dev · 2025-08-22 07:08
svenbut anyway, I wouldn’t worry about devices that will likely never run Linux
#asahi-dev · 2025-08-22 07:03
svencould be dp over tbt as well
#asahi-dev · 2025-08-22 07:03
svenI assume that is dp altmode but dunno
#asahi-dev · 2025-08-22 07:03
svenmy 15 pro at least does display port via a dell thunderbolt dock
#asahi-dev · 2025-08-22 06:27
svenoh yeah, that sounds like something they would’ve done
#asahi-dev · 2025-08-22 06:27
svensame for the aes block, there’s some script in m1n1 but I never bothered to write a driver because it’s slower than the main cpu with the optimized arm64 implementation Linux has
#asahi-dev · 2025-08-22 06:26
svennever wrote a driver though
#asahi-dev · 2025-08-22 06:25
sveniirc someone (alyssa or maybe eiln) reverse engineered it a long while ago
#asahi-dev · 2025-08-21 20:06
svenseriously though, why not use rust adt if it’s there now and makes the code nicer
#asahi-dev · 2025-08-21 20:06
svenmaybe also riir libfdt while you’re at it and do it completely in rust :D
#asahi-dev · 2025-08-21 17:25
svenso yeah, they are just the parents and the child domain is actually the correct one
#asahi-dev · 2025-08-21 17:24
sveni thought i dropped that line but apparently something went wrong while rebasing and reshuffling commits
#asahi-dev · 2025-08-21 17:23
svenugh, that was an older experiment iirc
#asahi-dev · 2025-08-21 17:13
svenboltctl list -a identifies the m1 host as „Apple iOS“ btw:D
#asahi-dev · 2025-08-21 17:05
svenyeah, the 2018 model is A12 which is no longer exploitable
#asahi-dev · 2025-08-21 17:04
sventhat’s still lightning
#asahi-dev · 2025-08-21 17:04
svenso iPad Pro from 2017
#asahi-dev · 2025-08-21 17:03
sveniirc the latest exploitable SoC is A11
#asahi-dev · 2025-08-21 15:40
svenhttps://lore.kernel.org/asahi/20250821-atcphy-6-17-v1-0-172beda182b8@kernel.org/ and here we go, usb3 RFC series
#asahi-dev · 2025-08-21 10:59
svenok
#asahi-dev · 2025-08-21 10:58
svenand i guess only two are controlled by atcphy
#asahi-dev · 2025-08-21 10:57
svenoh, okay
#asahi-dev · 2025-08-21 10:57
svendoes anyone remember what label = "USB-C Back-right-middle"; means on the imac? i'm a bit confused about the "right-middle" part
#asahi-dev · 2025-08-21 09:51
svenok, great
#asahi-dev · 2025-08-21 09:36
svenwe can also add some code to m1n1 to just add apple,dwc3 for now and keep that for a year or so
#asahi-dev · 2025-08-21 08:30
svengreat. i definitely want to keep that in the binding
#asahi-dev · 2025-08-20 20:15
sventrue
#asahi-dev · 2025-08-20 20:12
svenmaybe I shouldn’t write kernel patches while exhausted from traveling for $work ;)
#asahi-dev · 2025-08-20 20:12
sventhanks :D
#asahi-dev · 2025-08-20 20:12
svenlol, good catch
#asahi-dev · 2025-08-20 16:41
svenin that branch i now use "apple,t8103-dwc3", "snps,dwc3" as the base one and the other SoCs get an additional "apple,tXXXX-dwc3" in front of that
#asahi-dev · 2025-08-20 16:40
svenworst case is we just add back that one inside m1n1
#asahi-dev · 2025-08-20 16:40
svenok, great
#asahi-dev · 2025-08-20 16:10
svenhttps://git.kernel.org/pub/scm/linux/kernel/git/sven/linux.git/log/?h=wip/atcphy& alright, mostly happy with that. needs a cover letter, one final test (don't have my mac mini with me) and then i'll send that our as RFC
#asahi-dev · 2025-08-20 15:39
svenkettenis: does openbsd use apple,dwc3 anywhere? because that's probably going to disappear after that discussion yesterday
#asahi-dev · 2025-08-19 16:38
svensure, I’d wait until we’ve clarified the dt binding submission before you submit v3 fwiw
#asahi-dev · 2025-08-19 16:35
svenoh, to mention it doesn’t work on t6000 anymore?
#asahi-dev · 2025-08-19 16:35
svenhm?

plus ›