#
✕ nick:marcan
#asahi-dev · 2022-12-04 17:37
marcantagged asahi-6.1-rc7-1
#asahi-dev · 2022-12-04 15:46
marcanrobher: we're already putting firmware versions for specific devices in those nodes, but this is more like overall bootloader versions and things like that
#asahi-dev · 2022-12-04 13:27
marcan(wrangling the kernel right now)
#asahi-dev · 2022-12-04 13:27
marcanjannau: fixed, thanks
#asahi-dev · 2022-12-04 13:26
marcanjannau: derp.
#asahi-dev · 2022-12-04 11:36
marcanjannau: I don't use vdmtool, I use my arduino thing (which has its own reliability problems, with everything)
#asahi-dev · 2022-12-04 08:04
marcannevermind, reverting
#asahi-dev · 2022-12-04 08:04
marcanah yeah, it's just copying the cryptexes too
#asahi-dev · 2022-12-04 08:02
marcanah wait, maybe I know what this is
#asahi-dev · 2022-12-04 07:58
marcanFYI: I've increased installer stub size to 3GB (in dev only). There was a bug report that 13.0 failed to install due to running out of space. That's all unsupported and I think we can get a lot of leeway if we figure out how to use APFS compression for the recoveryOS image, but I feel uneasy about 2.5GB if we're so close to the edge, so let's play it safe for new installs.
#asahi-dev · 2022-12-04 07:22
marcanI just added a pile of string props to pass various firmware and bootloader versions upward, mostly for documentation / so we can ask for and gather that info with bug reports
#asahi-dev · 2022-12-04 07:21
marcanwe've been using the asahi, prefix for stuff we just stash in there like firmware versions and the keyboard language code (the latter is the only thing used by userspace setup scripts, and we probably need a more standard binding for that going in the keyboard node; that yak is in the back burner)
#asahi-dev · 2022-12-04 07:20
marcanrobher: Is documenting /chosen bindings a thing? We've been adding random stuff there, some of which might make sense to document in the sense of promising stability in the future (or just getting review). None of it is used by the kernel itself right now (the aforementioned version thing is in the device node, that will be reviewed along with that binding).
#asahi-dev · 2022-12-04 07:17
marcan(m1n1 passes version <12 4 0> and compat <12 3 0> for 12.4)
#asahi-dev · 2022-12-04 07:16
marcanthe DCP driver should probably refuse to probe if firmware-compat isn't <12 3 0>
#asahi-dev · 2022-12-04 07:16
marcanjannau: I added firmware version formwarding into the DT, in the style of what Lina proposed for GPU (apple,firmware-version and apple,firmware-compat, containing something like <12 3 0>)
#asahi-dev · 2022-12-02 06:02
marcanthat's the same
#asahi-dev · 2022-12-02 06:02
marcanthere's a reason I ranted about Manjaro shipping a random kernel of ours
#asahi-dev · 2022-12-02 06:02
marcanbecause users *will* file bugs upstream or generally clog the communications channels
#asahi-dev · 2022-12-02 06:01
marcanI expect non-reference distributions not to ship anything newer than what we ship
#asahi-dev · 2022-12-02 06:01
marcan...which doesn't mean it's a free-for-all for packaging unstable software
#asahi-dev · 2022-12-02 05:55
marcan(of some opaque value)
#asahi-dev · 2022-12-02 05:55
marcancould just be a mixer control write
#asahi-dev · 2022-12-02 05:55
marcan(so no other app can trigger it)
#asahi-dev · 2022-12-02 05:55
marcanso it doesn't have to be anything complicated, just something explicit
#asahi-dev · 2022-12-02 05:55
marcanthe only security required is that regular users should not be able to blow up the speakers, and that it should never happen "by accident"
#asahi-dev · 2022-12-02 05:55
marcanit's about the handshake, and yes root
#asahi-dev · 2022-12-02 05:53
marcanv/i separately is not enough to calculate power because you don't know the power factor
#asahi-dev · 2022-12-02 05:52
marcanyou actually multiply v*i sample-wise
#asahi-dev · 2022-12-02 05:52
marcanah yeah, that makes no sense, it's audio-rate data
#asahi-dev · 2022-12-02 05:52
marcansince they come in the same way as input audio data as far as the hardware is concerned, I don't see much reason to reinvent that wheel
#asahi-dev · 2022-12-02 05:51
marcanprobably a dedicated PCM for that
#asahi-dev · 2022-12-02 05:51
marcanI assume so
#asahi-dev · 2022-12-02 05:49
marcanmy current idea is we want the safety capture PCM to always be available, but simply cease to deliver data when the output is idle, and the daemon should compensate for that algorithm-wise
#asahi-dev · 2022-12-02 05:48
marcanthere will be a keepalive of some sort, though that keepalive may be implicit in "is actively reading the PCM data". I need to see how that would work, and also how it will all work when the TX side PCM is closed or idle
#asahi-dev · 2022-12-02 05:48
marcanyes
#asahi-dev · 2022-12-02 05:48
marcan(yes this will require core ALSA changes, we'll bikeshed it I'm sure, but I will not ship a solution that does not have that level of safety in some form)
#asahi-dev · 2022-12-02 05:47
marcanyou should not need to implement any emergency measures in userspace
#asahi-dev · 2022-12-02 05:47
marcanI absolutely want this to have things like only allowing a trusted/special process to raise levels above the max and resetting when that fd is closed or when a keepalive fails
#asahi-dev · 2022-12-02 05:47
marcanchadmed: don't bother with panic hooks, all that will be handled in the kernel
#asahi-dev · 2022-12-01 16:58
marcanand yes we'll need a libinput quirk similarly, to keep it from being interpreted as a regular touchscreen
#asahi-dev · 2022-12-01 16:57
marcanbut I expect we'll have a daemon doing uinput to emulate F keys and that's that for a while
#asahi-dev · 2022-12-01 16:57
marcanusing those flags to keep it out of compositor's reach sounds like a plan (or maybe new flags or something, maybe discuss with #dri-devel?)
#asahi-dev · 2022-12-01 16:56
marcandsharshakov: it's DRM, it has to be DRM (the legacy framebuffer stuff is deprecated)
#asahi-dev · 2022-12-01 16:54
marcansven: you've got mail :)
#asahi-dev · 2022-11-30 05:57
marcansven: I'll write and send you the t600x DT patch for review if you don't beat me to it :)
#asahi-dev · 2022-11-30 05:56
marcanI'll send out the DT changes tonight
#asahi-dev · 2022-11-30 05:56
marcanlooks like cpufreq made it in \o/
#asahi-dev · 2022-11-30 04:47
marcanmaybe we should too
#asahi-dev · 2022-11-30 04:47
marcanbut yeah, the Apple driver does a readback check I think

plus ›