#asahi-dev · 2026-09-15 10:16
chadmedmy issue was taking most of today to figure out that the validator doesnt like referencing patternproperties from another file so you basically just have to redefine them#asahi-dev · 2026-09-15 10:15
chaos_princesskinda shameful to admit, but i realized that dt-bindings is just json-schema-but-yaml like less than a month ago, and all my previous attempts at writing them was me figuring out the format from examples :P#asahi-dev · 2026-09-15 10:14
svenlol#asahi-dev · 2026-09-15 10:12
chadmednuke devicetree bindings from orbit seriously#asahi-dev · 2026-09-15 10:10
linkmauveBut these were tiny bugs, not roadblockers.#asahi-dev · 2026-09-15 10:10
linkmauveI’ve only worked against Gstreamer and ffmpeg so far, not Firefox, not Chromium, not Wine, but both had bugs around that which I reported (and maybe fixed, I don’t remember).#asahi-dev · 2026-09-15 10:09
leioHow are userspace things currently faring with encountering a vulkan device that doesn't expose those queues, or that assume the same first iterated device would do VV too#asahi-dev · 2026-09-15 10:09
linkmauve(Due to a bug in Gstreamer or ffmpeg, I don’t remember which, I had to lie and expose a graphics queue, but that’s only because that client couldn’t believe a Vulkan driver could exist without that.)#asahi-dev · 2026-09-15 10:08
linkmauveI have yet to look at how much of the common framework of Mesa I could reuse, since my driver is pretty unique in that it doesn’t expose any graphics or compute queue.#asahi-dev · 2026-09-15 10:05
linkmauveIt’s currently completely separate, but airlied made a compelling argument that it should become part of Mesa, even if only for exposure.#asahi-dev · 2026-09-15 10:05
leiolinkmauve: hey, your v4l2 VV thing, is it on top of mesa or separate?#asahi-dev · 2026-09-15 03:09
chadmeddo we want to keep the M3 audio stuff downstream for a while or will i just squash it all submit it #asahi-dev · 2026-09-15 00:10
catchlinefalsehello world#asahi-dev · 2026-09-14 15:53
adityaI was wondering if there was any update on the upstreaming status of the trackpad driver#asahi-dev · 2026-09-14 15:52
adityaHi @jannau#asahi-dev · 2026-09-14 09:35
chadmedi think im okay with exposing it anyway given that apple does#asahi-dev · 2026-09-14 09:35
chadmeddid anyone test that rgb bt2020 works with an hdmi 2.x display #asahi-dev · 2026-09-14 09:35
chadmeddo we want to expose hdr on hdmi given the weirdness with it trying to use yuv formats#asahi-dev · 2026-09-14 08:40
sofusnice!#asahi-dev · 2026-09-14 08:32
jannausofus: nothing, I have that queued#asahi-dev · 2026-09-14 08:29
chaos_princessyes. Feel free to not bother for stuff that is just in the asahilinux git, and only do it right before sending to lkml#asahi-dev · 2026-09-14 08:25
sofuswhat should i do about "apple,t603[01]-dart" not being documented, simply add it?#asahi-dev · 2026-09-13 12:26
chadmedbut then i think we wouldnt be able to cook the commit timing books #asahi-dev · 2026-09-13 12:25
chadmedhmm i wonder if explicit fencing will let us drop the framebuffer reference hacks#asahi-dev · 2026-09-13 08:01
mihicakhareexit#asahi-dev · 2026-09-13 07:59
mihicakhareopen https://alx.sh/l/asahi-dev#asahi-dev · 2026-09-13 05:50
nickchanbut rather a new kernel#asahi-dev · 2026-09-13 05:50
nickchanthere could be machines still on those versions but when they upgrade i don't think it's going to a new stable revision#asahi-dev · 2026-09-13 05:50
nickchanand i consider that to be fairly unlikely either#asahi-dev · 2026-09-13 05:49
nickchanbut none of that matters because there's no users, am more thinking about possible m1 users on 6.1, 6.6 and 6.12#asahi-dev · 2026-09-13 05:49
nickchan6.12 i think would boot up if specified with the right device tree#asahi-dev · 2026-09-13 05:49
nickchanwell technically the fiq but one gets the point#asahi-dev · 2026-09-13 05:48
nickchanbut like 6.1 and 6.6 wouldn't boot on a11 anyways due to aic#asahi-dev · 2026-09-13 05:48
nickchanthis is what exist in postmarketOS#asahi-dev · 2026-09-13 05:47
nickchannicolas17: the oldest kernel an a11 user would be on is probably 7.0.12#asahi-dev · 2026-09-13 05:32
nicolas17wouldn't any actual A11 user be on bleeding edge kernel?#asahi-dev · 2026-09-13 05:32
nicolas17A11?#asahi-dev · 2026-09-13 05:23
nickchancause the Apple A11 supprot changed quite a bit of indention#asahi-dev · 2026-09-13 05:23
nickchanstable backport because the PRP chicken bit fix depends on Apple A11 support when applied as is#asahi-dev · 2026-09-13 05:22
chadmedstable backports?#asahi-dev · 2026-09-13 05:22
nickchannot enough energy to test and suggest another patch for dropping prp check anyways#asahi-dev · 2026-09-13 05:21
nickchani don't have the energy to care about that when there's probably no users on those kernels#asahi-dev · 2026-09-13 05:21
nickchando i want care about how they want to put the apple a11 nvme support into 6.1, 6.6, and 6.12 cause it's as-is a dependency for the drop prp chicken null bit#asahi-dev · 2026-09-12 09:14
fl0_id_yeah. some ppl did send to the mailing list, but I didn't see any replies#asahi-dev · 2026-09-12 09:13
myria*the last MR I saw was sitting open for a few months and not sure if anything else happened#asahi-dev · 2026-09-12 09:13
myriayeah, I was half tempted to just send a patch to the mailing list, tbe #asahi-dev · 2026-09-12 09:12
fl0_id_myria it seems MRs on sourceforge are most active... but yeah it's bit unclear where activity happens#asahi-dev · 2026-09-12 08:11
myriayeah, now I just need to figure out how to send prs to libpfm4#asahi-dev · 2026-09-12 08:10
JamesCalligeros[m]Yeah hopefully it gets picked up. It is a pretty simple patch after all :p#asahi-dev · 2026-09-12 08:09
myriawell, I sent it upstream anyways so we'll see what happens?