#asahi · 2025-08-14 17:24
jannauerr, no. we have `stdout-path = "serial0";` in the dts and u-boot changes it during non-virt boot#asahi · 2025-08-14 17:22
jannaulinux.py works for my dev kernel#asahi · 2025-08-14 17:20
jannaum1n1 sets /chosen/stdout-path to its uart in HV mode#asahi · 2025-08-14 17:13
jannauit's ages since I used linux.py - I wouldn't be surprised if the has bit-rotted or unexpected dependencies#asahi · 2025-08-14 17:12
jannaumy dev kernel has EFI_STUB=y so that works at least with run_guest_kernel.sh#asahi · 2025-08-14 17:10
jannaui.e. it will not work with fedora distro kernel#asahi · 2025-08-14 17:09
jannauI suspect proxyclient/tools/linux.py won't work with an EFI_STUB kernel#asahi · 2025-08-14 06:11
jannauzzzUzz: our u-boot build disables u-boot env files. I'm not sure if that's still necessary after we identify the correct efi system partition but changing this is effort#asahi · 2025-08-13 06:45
jannauTramtrist: it's not yet released but will come before fedora 43.#asahi · 2025-08-04 14:00
jannauthe frankenkernel is at least some HW documentation although certainly partly misleading/wrong#asahi · 2025-07-28 17:37
jannauahepp: you don't need a separate APFS container. a new volume in the existing one is enough#asahi · 2025-07-06 14:27
jannauAndjrfi: a possible workaround would be to apply that workaround only when OS FIRMWARE version is 13.5 or older#asahi · 2025-07-06 14:26
jannauissue is either touching dcp at all when booting or not putting it back into the correct sleep state#asahi · 2025-07-06 14:24
jannauthat's probably caused by m1n1 trying to work around on sequoia bugs. try changing FW_MAX to V15_3 in https://github.com/AsahiLinux/m1n1/commit/277e8e71de92dbb2ff0dba9d49b2b27a97e23a08#asahi · 2025-06-28 20:36
jannaubrltty-dracut is available in fedora but does not seem to be installed by default#asahi · 2025-06-27 07:57
jannaunot sure if we can write m1n1's mode into the DT and have the fb dev use that instead of the preferred mode#asahi · 2025-06-27 07:56
jannaulow priority plan is to add EDID support in the rust rewrite of m1n1's display code. that will eliminate the issue with virtual modes in m1n1#asahi · 2025-06-27 07:52
jannauso just the mode changes but nothing wrong with any of the selected modes itself.#asahi · 2025-06-27 07:19
jannauchadmed: what's the issue? the default mode flickers and if you select a different refresh rate it works? does that happen with other display controllers with the same mode?#asahi · 2025-06-26 13:01
jannauthe preferred mode is based on dcp's scoring. m1n1's mode selection is limited though because the iboot protocol doesn't tell us which modes are real (there are virtual modes) and supported by the display#asahi · 2025-06-26 12:52
jannauselecting larger modes might be buggy (not well tested) since it requires fb reallocation#asahi · 2025-06-26 12:51
jannauchadmed: yes. m1n1 allows even mode selection via the "display" variable: https://github.com/AsahiLinux/m1n1/blob/main/m1n1.conf.example#L17#asahi · 2025-06-26 10:55
jannauthat sounds like it's the switch from simpledrm (boot framebuffer) to dcp (full display driver)#asahi · 2025-06-22 21:36
jannaurandom suspicion from looking at the kernel log. Is NodemManager installed on the arch / ubuntu install?#asahi · 2025-06-22 21:18
jannaucan you try to stop ModemManager?#asahi · 2025-06-20 15:31
jannaulooking at CPU usage while it's played would probably a good indicator as well#asahi · 2025-06-20 14:49
jannauI'd say 2:1 that it works#asahi · 2025-06-20 14:45
jannauflokli: that video is in addition to high resolution high bit bit rate (120mbps). I need 6 performance cores to be faster than realtime decoding (on average). so this video is too much for base m1 and m2#asahi · 2025-06-20 14:22
jannau1080p60 version has in firefox less than 1% dropped frames. not nice but I wouldn't call that struggling performance wise. could be be easily just be an occasional missing vsync#asahi · 2025-06-20 14:17
jannauI see just a few dropped frames even for the 4k60 locally#asahi · 2025-06-20 14:07
jannauchadmed: the 5700k 1080p VP9 version of https://www.youtube.com/watch?v=dQw4w9WgXcQ secodes single threaded with 288fps#asahi · 2025-06-20 14:02
jannauchadmed: vp9 1080p60 should not be a problem (assuming reasonable bitrates). can you share a video which has issues?#asahi · 2025-06-20 12:11
jannauno, no keyboard and no usb#asahi · 2025-06-03 17:37
jannauack, that all looks as expected#asahi · 2025-06-03 17:32
jannau`amixer sget Master` should report at the same time 20%#asahi · 2025-06-03 17:32
jannauaxboe: is in `amixer -D sysdefault:CARD=AppleJ416 | grep -A5 -e ' Speaker' -e ' Amp Gain'` everything on 100% even when the volume in the plasma widget is at 20% (for 16-inch M2 Pro/Max)#asahi · 2025-06-03 17:09
jannauI would describe music at 20% here as background music. clearly audible in a non-busy room but no problem to speak over it and understand each other. I assume that matches your previous loudness levels#asahi · 2025-06-03 17:02
jannauaxboe: can you describe the dropped volume? can you still hear youtube music videos through closed doors in another room (at 100%)#asahi · 2025-06-03 16:53
jannauhmm, that should work fine. Not sure why you would have had 6 channels before. you could try to move ~/.local/state/wireplumber/ out of the way.#asahi · 2025-06-03 16:50
jannauaxboe: how do you control/look at the volume? the setup with the DSP filter chain/wireplumber is more complex than plain alsa devices#asahi · 2025-06-03 16:41
jannauwas any of the mentioned packages updated recently? what about pipewire or wireplumber?#asahi · 2025-06-03 16:39
jannauare asahi-audio, speakersafetyd, bankstown-lv2, triforce-lv2 and alsa-ucm-conf-asahi installed?#asahi · 2025-06-03 16:38
jannauaxboe: just 2 channels (on a macbook pro 14-/16-inch) are expected with the full speaker support. is this trixie?#asahi · 2025-05-27 18:19
jannauaic is there as well and I have chickens#asahi · 2025-05-27 18:17
jannauthat one is special since it still uses i2c tipd#asahi · 2025-05-27 18:17
jannauhttps://github.com/jannau/linux/commits/t8122-j504/ has a little more than a minimal dt for the base m3 macbook pro#asahi · 2025-05-27 15:20
jannauiboot (macOS 14.2 to 14.4* and probably 14.5) is broken with raw boot objects on M4*#asahi · 2025-05-26 16:40
jannaumemzntv: please try after deinstalling box64 (just to test if that's the cause)#asahi · 2025-05-26 16:40
jannauit apparently does#asahi · 2025-05-26 16:38
jannaudoes box64 sets binfmt-misc up?