#asahi · 2022-10-24 14:37
marcanYMMV depending on the specific device, I have some that work quite well and others not so well (it's a known issue that we're missing commands to prioritize audio streams properly)#asahi · 2022-10-24 14:37
marcanyes, though wifi coexistence is a bit questionable right now. don't use the LDAC codec if supported, and ideally use 5GHz WiFi simultaneously#asahi · 2022-10-24 14:33
marcanvoltage/current sense#asahi · 2022-10-24 14:32
marcanI had to drag him to the showroom floor and do a side by side comparison with an identical, intact unit#asahi · 2022-10-24 14:32
marcanbecause just woofers sound better than most laptops anyway#asahi · 2022-10-24 14:32
marcanto the point where I showed up at the Apple Store with blown tweeters and the Genius couldn't tell#asahi · 2022-10-24 14:31
marcanapple are pretty much the only laptop vendor actually trying hard with speaker/DSP quality#asahi · 2022-10-24 14:31
marcanalso almost every other laptop sounds worse than these macs#asahi · 2022-10-24 14:25
marcanyes#asahi · 2022-10-24 14:25
marcansafe volume limits would give you crappy speaker volume (especially on the tweeters)#asahi · 2022-10-24 14:25
marcanit's not just volume limits, we need a speaker safety model which computes voice coil temperature based on voltage/current feedback#asahi · 2022-10-24 10:49
marcancapta1nt0ad: not in release kernels yet#asahi · 2022-10-24 04:47
marcanarnd: expect a PR once the DT checker finishes running#asahi · 2022-10-24 04:43
marcanthis is a wake-up call for other packagers: if you want the GPU driver when it's ready, better take care of the rust build infra now ;)#asahi · 2022-10-24 04:42
marcanthen there will be two packages: linux-asahi-dev will have DCP and the rust stuff enabled#asahi · 2022-10-24 04:42
marcanOK, so as for kernels: I'm going to merge in DCP, and also start merging in all the rust-for-linux downstream stuff so we can build with stub rust support#asahi · 2022-10-24 04:39
marcancpufreq v3 sent#asahi · 2022-10-24 03:58
marcanuser982492: yes#asahi · 2022-10-22 08:46
marcanyes#asahi · 2022-10-22 08:45
marcanif you try to remove the GPU node WindowServer just refuses to start#asahi · 2022-10-22 08:45
marcanthe kernel has a framebuffer alright, but the GUI requires GPU acceleration and there is no software renderer for ARM as far as I know (or it is disabled)#asahi · 2022-10-22 08:27
marcanhowever, you're only going to get single user text mode unless you implement a Metal paravirtualized GPU, so it's not very useful if you want macOS#asahi · 2022-10-22 08:26
marcanelvishjerricco: yes, the macOS VM kernel runs in a VM on other ARM hardware, people tried that way way back#asahi · 2022-10-22 08:24
marcanthat's a palm rejection thing in libinput IIRC, or something like that#asahi · 2022-10-17 23:16
marcan(this seems like a problem other platforms would have too)#asahi · 2022-10-17 23:16
marcan(I'm not sure what the right solution is; maybe those drivers need to be changed to work together or have well defined priority)#asahi · 2022-10-17 23:16
marcanpossibly probe order swapped#asahi · 2022-10-17 23:12
marcanI've always told people to disable CONFIG_FB_EFI from the get go... yes, they fight, and then you end up with a random one and they don't behave the same way, and then people complain about xorg/whatever having problems with efifb#asahi · 2022-10-17 15:16
marcanpretty sure the studio also worked#asahi · 2022-10-17 15:15
marcanon an m1pro mbp#asahi · 2022-10-17 15:15
marcan[ 0.339171] simple-framebuffer 107d36f2880.framebuffer: [drm] fb0: simpledrmdrmfb frame buffer device#asahi · 2022-10-17 15:15
marcan[ 0.301214] [drm] Initialized simpledrm 1.0.0 20200625 for 107d36f2880.framebuffer on minor 1#asahi · 2022-10-17 15:13
marcanworks for me?#asahi · 2022-10-17 14:10
marcanother distros with different requirements are free to ship a dts package separately, for example, and have the update-m1n1 script use that#asahi · 2022-10-17 14:09
marcanbut right now, with the incompatible changes, for the reference distro, the 1:1 kernel:dt relationship works better than that will#asahi · 2022-10-17 14:09
marcanwe could certainly have m1n1 ship its own dts, of course, from an autogenerated dt repo (I had plans for that at some point, it's just on the endless back burner because the current solution works)#asahi · 2022-10-17 14:08
marcanthe only problem we have now is that bindings keep changing#asahi · 2022-10-17 14:08
marcanand makes more sense than trying to sync stuff across to some other source of truth#asahi · 2022-10-17 14:07
marcankettenis: using the kernel repo as the canonical dt source and having the latest kernel provide the DTs is fine in principle#asahi · 2022-10-17 14:06
marcanRy_Darcy: you probably didn't update your device trees#asahi · 2022-10-17 07:17
marcanstreaming in a few mins on #asahi-stream as usual#asahi · 2022-10-17 06:41
marcanthis is part of what being on the bleeding edge (i.e. using non-upstream bindings) entails right now#asahi · 2022-10-17 06:41
marcanbecause there will be non-backwards-compatible changes#asahi · 2022-10-17 06:41
marcandoing that properly for other distros is more complicated at this stage, since dtbs are per-kernel so there is no "good" way to pick the right one. for upstream kernels/DTBs, the right choice is the latest installed kernel. for downstream Asahi kernels, there is no good solution right now.#asahi · 2022-10-17 06:40
marcanderzahl: on arch, there is only ever one installed kernel, so it only picks up one set of DTBs#asahi · 2022-10-16 17:10
marcantpw_rules: yeah, I've been thinking of keeping two branches for that purpose#asahi · 2022-10-16 17:08
marcan(like right now)#asahi · 2022-10-16 17:08
marcanthe goal is to not confuse users, and also to not mindlessly encourage people to use asahi-dev because it is broken more often than not#asahi · 2022-10-16 17:07
marcanit's upcoming#asahi · 2022-10-16 17:07
marcanit's fine to talk about unsupported asahi-dev stuff, just don't say it's supported