#asahi-dev · 2023-01-14 03:38
marcanand then we can start offering macos-less installs#asahi-dev · 2023-01-14 03:38
marcanwe should figure out the proper SFR update ramdisk stuff some day, after we have some semblance of APFS and nvram support#asahi-dev · 2023-01-14 03:37
marcan(I've been advised not to advocate it for firmware updates for macos-less Asahi installs)#asahi-dev · 2023-01-14 03:37
marcanso it's a pretty big hammer#asahi-dev · 2023-01-14 03:36
marcanand also rolls over the lpnh I think or something like that#asahi-dev · 2023-01-14 03:36
marcanit also reinstalls macOS AIUI #asahi-dev · 2023-01-14 03:36
marcanamarioguy: probably?#asahi-dev · 2023-01-14 03:36
marcantpw_rules: AIUI you can pass in tweak keys per NVMe operation, so if you want that kind of layer you could add it on top and support progressive re-encrypt like LUKS does#asahi-dev · 2023-01-14 03:35
marcan(for a DFU wipe, not revive)#asahi-dev · 2023-01-14 03:35
marcanyes#asahi-dev · 2023-01-14 03:35
marcanso the AP never sees them, ever#asahi-dev · 2023-01-14 03:35
marcanAIUI the keys are transferred through a back channel from SEP into ANS directly#asahi-dev · 2023-01-14 03:34
marcanwas it extended or external?#asahi-dev · 2023-01-14 01:18
marcanj`ey: I do pull via github sometimes, if the stars line up and there wasn't a rebase along the way#asahi-dev · 2023-01-13 12:02
marcanarnd: I think Apple claims TB4 compliance for t600x though? I saw that somewhere#asahi-dev · 2023-01-13 07:51
marcanyeah#asahi-dev · 2023-01-13 07:51
marcanthat's it. that's the reason they aren't.#asahi-dev · 2023-01-13 07:51
marcanwhich hilariously means M1/M2 devices aren't TB4 compliant... because TB4 requires being able to drive two monitors#asahi-dev · 2023-01-13 07:50
marcanapparently TB4 is just extra spec requirements#asahi-dev · 2023-01-13 07:47
marcanheh#asahi-dev · 2023-01-13 07:47
marcanisn't TB4 just USB4 anyway?#asahi-dev · 2023-01-13 07:47
marcan"In addition, it can be used to connect Thunderbolt-enabled displays — such as the Apple Thunderbolt Display and LG Thunderbolt 2 displays — to any of the Thunderbolt 3 (USB-C) / USB 4 ports on your Mac." claims USB4 works#asahi-dev · 2023-01-13 07:46
marcan"However, using this adapter with an Apple Thunderbolt Display requires a power source, because the display does not provide power through the adapter." that makes sense - it doesn't do *reverse* power#asahi-dev · 2023-01-13 07:45
marcanhttps://twitter.com/llllmills/status/921104810304139264#asahi-dev · 2023-01-13 07:44
marcanrkjnsn_: wait what, I thought you could use it with the FireWire adapter (which obviously needs power)?#asahi-dev · 2023-01-13 06:56
marcanright, need to take a look at that#asahi-dev · 2023-01-13 06:25
marcantagged another kernel with the nvme fix hopefully, pushed to asahi-dev#asahi-dev · 2023-01-13 04:56
marcan(I still don't fully agree that this is specifically an Apple thing, the generic code seems brittle right now, but whatever, let's just get it fixed)#asahi-dev · 2023-01-13 04:55
marcanjannau: do you want to resubmit your original NVMe fixup with a big comment about how the spec requires disabling the controller in order to enable it later, and on Apple controllers it makes more sense to do it right after shutdown to simplify the resume codepath?#asahi-dev · 2023-01-11 02:24
marcanand I think there's more than one bug here, in the patchset that caused the regression#asahi-dev · 2023-01-10 23:43
marcanjannau: ohhh that makes sense. thank you for figuring that one out :)#asahi-dev · 2023-01-10 13:34
marcananyway, I should get some sleep, still recovering a bit from sunday :p#asahi-dev · 2023-01-10 13:34
marcan:)#asahi-dev · 2023-01-10 13:32
marcanfirst time I met him IRL too, but he's part of the reason I ended up spending time fixing that nasty OBS audio death bug (that had crashed so many streams before)#asahi-dev · 2023-01-10 13:32
marcantoday I hung out with corporatetool/James Kao from AniCLover/Anikura Unison (which I've helped with in the past, and he was at AnimeltUP! this weekend and is staying a few more days), we talked quite a bit about OBS stories and laser stuff :p#asahi-dev · 2023-01-10 13:29
marcanthat was one of them yes :p#asahi-dev · 2023-01-10 13:29
marcanneed to find some time to look at it (it's been a few busy weeks with irl stuff, should calm down this/next week)#asahi-dev · 2023-01-10 13:28
marcannvme borks#asahi-dev · 2023-01-10 13:28
marcanyes, I know#asahi-dev · 2023-01-10 13:28
marcanI'm getting more and more tired of how broken X is... at some point, if sddm finally does a release with native Wayland support, I might just have to take a page from what half the Linux graphics world is doing and declare X unsupported, because it's not our job to fix it if nobody else cares about it#asahi-dev · 2023-01-10 13:27
marcanit's some stupid X thing, I'm pretty sure#asahi-dev · 2023-01-10 13:27
marcanjannau: no, the arch mesa package does not have any uabi, and mesa explicitly prints a message about failing to load the driver since it isn't actually available#asahi-dev · 2023-01-10 12:42
marcanjannau: no reason why that wouldn't, I think it's the DCP+AGX DRM nodes that break it in some way (but only with llvmpipe)#asahi-dev · 2023-01-10 06:00
marcanthis isn't the race, it consistently fails on arch right now#asahi-dev · 2023-01-10 04:23
marcansigh, I think Xorg is broken (again? or did this never work) on edge without the mesa drivers#asahi-dev · 2023-01-09 16:41
marcan_jannau_: I think that "ack" is more like a state report, and it only gets sent on first init (but you can request another one)#asahi-dev · 2023-01-09 15:38
marcan_jannau_: try sending it msg type 1 (the initcomplete msg)#asahi-dev · 2023-01-09 14:22
marcanchadmed_: nice!#asahi-dev · 2023-01-09 14:22
marcanjannau: stupid idea, does setting it to 0 make it acknowledge it again next time?#asahi-dev · 2023-01-09 04:18
marcanjannau: ha (but also ew, means we need to map it at a fixed address on reload?)