#
#asahi-dev · 2026-08-02 18:57
unifiedgravity[m]Hm force of habit I guess, never forced to use UTM at work
#asahi-dev · 2026-08-02 18:57
svenyou can just start a headless one with UTM
#asahi-dev · 2026-08-02 18:57
svenmy vms are usually all that as well without the docker cruft inbetween
#asahi-dev · 2026-08-02 18:57
svenuh
#asahi-dev · 2026-08-02 18:56
unifiedgravity[m]Exactly
#asahi-dev · 2026-08-02 18:56
svendocker on macos is just a vm in disguise
#asahi-dev · 2026-08-02 18:56
unifiedgravity[m]It's not going to be that terrible though, in reality I'd spin up a docker container and ssh into it. No GUI, mouse issues, can just scp stuff in and out with a script. Disposable.
#asahi-dev · 2026-08-02 18:55
sven*:D
#asahi-dev · 2026-08-02 18:55
svenor, well, it would if the kernel wasn't so annoying to compile on macos :d
#asahi-dev · 2026-08-02 18:55
unifiedgravity[m]Yeah, it's a "CBF setting up a VM" kinda thing
#asahi-dev · 2026-08-02 18:54
svenit's a bit less painful than working in a vm
#asahi-dev · 2026-08-02 18:53
JPeisach1is there benefit to building on macOS? i guess if the system is unbootable?
#asahi-dev · 2026-08-02 18:52
unifiedgravity[m]Ta, interesting
#asahi-dev · 2026-08-02 18:50
jannauhttps://github.com/bee-headers/homebrew-bee-headers#how-to-use-bee-headers-to-build-the-linux-kernel-in-macos---quick-version might work to build linux on macOS, linux vm is probably more robust
#asahi-dev · 2026-08-02 18:47
unifiedgravity[m]And by rock solid I mean I can't crash or stall the thing and can't kill the kernel or userspace with surprise removal or add. Still got a small number of xruns/pops when stop/start or transition system power management states.
#asahi-dev · 2026-08-02 18:45
unifiedgravity[m]Cheers for the advice Sven, chaos_princess
#asahi-dev · 2026-08-02 18:44
unifiedgravity[m]tell if pops and bugs are its cstates, or if it is my driver. Fingers crossed.
#asahi-dev · 2026-08-02 18:44
unifiedgravity[m]I have had my Presonus 2626 driver rock solid for a while now but I keep procrastinating submission to the mailing list in part I have an outlook address and had to find another outgoing smtp server to avoid mangling encoding. Would love to test it on Apple Silicon. I only have Intel 1355U and whilst Intel is good at USB4, I don't trust that their power management is any good (and it blasts EM noise through the interface), hard to
#asahi-dev · 2026-08-02 18:41
sveni used that before asahi was usable on this machine
#asahi-dev · 2026-08-02 18:41
svenjust use a vm then
#asahi-dev · 2026-08-02 18:38
unifiedgravity[m]Cheers
#asahi-dev · 2026-08-02 18:37
unifiedgravity[m]Yeah I could install Asahi on my main but at the rate at which I wreck them, maybe it's better that I use a VM
#asahi-dev · 2026-08-02 18:37
svencat build/m1n1.macho Image.gz t8103-j274.dtb > payload.macho and then just chainload the payload.macho
#asahi-dev · 2026-08-02 18:36
svenit's somehow possible on macos but rather painful, i'd just use a vm
#asahi-dev · 2026-08-02 18:36
sven*:D
#asahi-dev · 2026-08-02 18:36
sveni use linux running on my m1 to compile the kernel :d
#asahi-dev · 2026-08-02 18:36
unifiedgravity[m]Can one build a Linux kernel on MacOS? Or do you use VM on MacOS to compile the kernel and then upload it to target?
#asahi-dev · 2026-08-02 18:35
unifiedgravity[m]But I burned so much just getting this far so I got cracking
#asahi-dev · 2026-08-02 18:35
unifiedgravity[m]Yeah, I was wondering about that flow
#asahi-dev · 2026-08-02 18:34
svenyou really want to just chainload your kernel together with the dtbs it compiled
#asahi-dev · 2026-08-02 18:34
unifiedgravity[m]I thought the thing wasn't coming up but it turns out to get the kernel to take from installed dtbs rather than from the m1n1 underneath it, you have to mess with some stuff, so last time I ran the branch on the M2 and well nothing appeared to happen at all, that'd be why
#asahi-dev · 2026-08-02 18:33
chaos_princessdepending on your firmware ver, you may alternatively get dmesg spam every 100ms or so
#asahi-dev · 2026-08-02 18:33
svenas usualy the second 80% of the project take as much time as first 80% :P
#asahi-dev · 2026-08-02 18:33
chaos_princessdebugging the thing
#asahi-dev · 2026-08-02 18:32
chaos_princessmore like the last 95%
#asahi-dev · 2026-08-02 18:32
sveneverything is up but acio doesn't recognize the connection
#asahi-dev · 2026-08-02 18:32
unifiedgravity[m]What was the last 5% missing; you said you were at 95%?
#asahi-dev · 2026-08-02 18:32
svenand then you should run into the same issue i do on t8103
#asahi-dev · 2026-08-02 18:32
sventbt-wip should only need dt changes for m1 pro
#asahi-dev · 2026-08-02 18:31
unifiedgravity[m]Gold, cheers
#asahi-dev · 2026-08-02 18:31
chaos_princessand then powers it down
#asahi-dev · 2026-08-02 18:31
chaos_princessmacos does bring up the atcphy in what looks like usb3 mode just to probe the port during startup
#asahi-dev · 2026-08-02 18:31
unifiedgravity[m]sorry yeah atcphy
#asahi-dev · 2026-08-02 18:30
sventhat sounds confusing, acio shouldn't even be on for usb3 mode. do you mean atcphy?
#asahi-dev · 2026-08-02 18:29
unifiedgravity[m]When doing acio transitions for example looked like from the mmio trace from macos 14.8.3 that we were going into usb3 mode, issue reset, then enter tbt mode. In Linux was trying to go straight from usb3 to tbt for example
#asahi-dev · 2026-08-02 18:28
sventhen where did you get serrors?
#asahi-dev · 2026-08-02 18:28
unifiedgravity[m]I am on the tbt-wip
#asahi-dev · 2026-08-02 18:27
svenif you use that wip branch you shouldn't get any serrors
#asahi-dev · 2026-08-02 18:27
unifiedgravity[m]Okay, noted thanks
#asahi-dev · 2026-08-02 18:27
unifiedgravity[m]I don't know as much as you by a long shot. Literally just figuring everything out

plus ›