#
✕ nick:marcan
#asahi-alt · 2023-12-22 17:57
marcanthis mechanism is not working for you for some reason
#asahi-alt · 2023-12-22 17:57
marcanthat is set by this udev rule in speakersafetyd: KERNEL=="pcmC*D2c", ATTRS{id}=="AppleJ*", ENV{ACP_IGNORE}="1"
#asahi-alt · 2023-12-22 17:57
marcanI fixed this with logic in PipeWire to check for a udev env var and ignore that subdevice
#asahi-alt · 2023-12-22 17:57
marcanPipeWire has a check that refuses to open any cards with busy subdevices. speakersafetyd keeps a subdevice busy.
#asahi-alt · 2023-12-22 17:56
marcanyour pipewire is too old or your udev rules are not working properly.
#asahi-alt · 2023-12-22 17:56
marcantpw_rules: > if it starts during boot, then plasma shows me no audio devices until i stop it.
#asahi-alt · 2023-12-15 08:11
marcanI need to figure out what to do about the alt routing modes, but honestly I'm tempted to just remove all that, it just makes everything more complicated once speakersafetyd is in the mix
#asahi-alt · 2023-12-15 08:08
marcannote that with bare ALSA it's expected not to work out of the box right now, you need to manually poke the routing settings in alsamixer first (UCM does this when using PW)
#asahi-alt · 2023-12-15 08:08
marcan< yuka> I updated to kernel asahi-6.6-9, speakersafetyd 0.1.7, asahi-audio e1ccf5f. With a freshly booted system the speakers are unlocked, then once I start playing something they immediately become "Speaker volumes locked: Lock timeout" <- IIRC this happens when the routing is configured wrong, which means PW/ALSA UCM didn't do its job.
#asahi-alt · 2023-12-15 08:08
marcan< yuka> also when I unload and load the snd_soc_macaudio again my system sometimes locks up <- unloading modules is basically completely untested and expected to be broken in all sorts of ways :p
#asahi-alt · 2023-11-13 10:01
marcanthe packaging is trivial, modulo distro/package manager specific details that would vary anyway
#asahi-alt · 2023-11-13 10:01
marcanall the audio stuff is "just packages"
#asahi-alt · 2023-11-13 10:01
marcanbut also now that the core m1n1/boot/initramfs type stuff is taken care of, there isn't much special sauce left packaging-wise really
#asahi-alt · 2023-11-13 10:00
marcantpw_rules: https://src.fedoraproject.org/group/asahi-sig this is probably the easiest way to find all the upstream RPMs we're involved with
#asahi-alt · 2023-11-12 15:41
marcanI could smell them while playing that back
#asahi-alt · 2023-11-12 15:41
marcanjn: I destroyed them with a sine sweep actually, but then I made a point of cooking them completely with IWTLW before taking it to the Apple store (and still had trouble convincing the guy that they were damaged)
#asahi-alt · 2023-11-12 10:46
marcanI'm not sure how worthwhile it is right now though, maybe it makes more sense once it evolves into a GUI installer? at that point we'd be shipping third-party modules and there's stuff to make i18n a lot less painful there.
#asahi-alt · 2023-11-12 10:46
marcanyou'd need to hook up some kind of gettext thing, possibly using the built-in gettext module and a bunch of scaffolding. there's no i18n support in the installer right now, so that's a bunch of refactoring that would be needed.
#asahi-alt · 2023-11-12 10:39
marcanGlanzmann: please send github PRs
#asahi-alt · 2023-11-12 10:38
marcanit will only start on j313 because that is the supported model
#asahi-alt · 2023-08-03 06:35
marcanjannau: ack
#asahi-alt · 2023-06-11 07:13
marcanit literally presents an uninitialized frame, purple is what that shows up as on AGX
#asahi-alt · 2023-06-11 07:12
marcanchadmed: purple flashes on firefox startup (the first frame) are a known firefox bug
#asahi-alt · 2023-05-20 10:54
marcanif you purchase an unsupported device that's on *you*, and every time you annoy the developers you are making us even less likely to prioritize supporting your device
#asahi-alt · 2023-05-20 10:53
marcannot even, a bit less
#asahi-alt · 2023-05-20 10:53
marcanrelease was 4 months ago, for a start
#asahi-alt · 2023-04-27 17:18
marcanqdot: ^
#asahi-alt · 2023-04-27 17:18
marcanif those two desync it'll refuse to initialize
#asahi-alt · 2023-04-27 17:17
marcanalso for mesa, use the latest versioned tag I made, *not* any live branch. mesa and kernel live branches are not kept in sync. generally the latest mesa tag will be in sync with the latest kernel tag and usually also the latest live kernel (except when I'm about to do a release)
#asahi-alt · 2023-04-27 17:16
marcan(because it breaks with newer LLVM)
#asahi-alt · 2023-04-27 17:16
marcanwhich is not what the kernel wants but it's a version that will actually work (the one the kernel wants won't)
#asahi-alt · 2023-04-27 17:16
marcanyou want 0.62.0
#asahi-alt · 2023-04-13 12:11
marcanhttps://github.com/AsahiLinux/PKGBUILDs there
#asahi-alt · 2023-04-13 12:09
marcanbut I might as well clarify in case someone thinks they do
#asahi-alt · 2023-04-13 12:09
marcanyeah, agreed
#asahi-alt · 2023-04-13 12:04
marcanand the remaining contributions are few and very trivial, so I think I can just do this
#asahi-alt · 2023-04-13 12:03
marcanwas about to ask
#asahi-alt · 2023-04-13 12:01
marcanand linux-asahi and mesa-asahi which will probably live forever as downstream bleeding edge packages if nothing else
#asahi-alt · 2023-04-13 12:01
marcanwell this includes uboot-asahi which won't
#asahi-alt · 2023-04-13 12:01
marcanlet me add something
#asahi-alt · 2023-04-13 12:00
marcanif we didn't have derivative PKGBUILDs in there I would add a blanket MIT-0 to the repo
#asahi-alt · 2023-04-13 11:58
marcanI don't think we have a license on that repo explicitly, we probably should, but I basically consider everything in there (except for patches subject to upstream licensing) public domain for any distro to adapt as they wish
#asahi-alt · 2023-04-13 11:58
marcanfeel free to note the origin of the script if you want, though I personally don't particularly care
#asahi-alt · 2023-04-13 11:57
marcanmps: maintainer != author, you're the maintainer for the APKBUILD, not me
#asahi-alt · 2023-03-12 18:01
marcananyway, I need to wake up early tomorrow, good night :p
#asahi-alt · 2023-03-12 17:59
marcan(and my maintainer role is trivial enough - the soc stuff - that I don't need tooling)
#asahi-alt · 2023-03-12 17:59
marcanhonestly I think what we're doing here is surprisingly unique as far as kernel dev goes
#asahi-alt · 2023-03-12 17:58
marcanor people overseeing a bunch of others sending patches on their own
#asahi-alt · 2023-03-12 17:58
marcanthere doesn't seem to be much oriented towards submitters *sending* patches
#asahi-alt · 2023-03-12 17:58
marcanmost of these tools seem oriented towards maintainers, *receiving* patches

plus ›