#asahi-dev · 2023-09-25 11:27
jannegI'm not seeing anything which would explain why the headphones fail on some system but work on others#asahi-dev · 2023-09-25 11:14
jannegno, I mean using both speakers and headphones at the same time is a "Pro Audio" feature, or I'm misunderstanding what ConflictingDevice does#asahi-dev · 2023-09-25 11:08
jannegI suppose that would be a use case covered by "Pro Audio"#asahi-dev · 2023-09-25 11:06
jannegchadmed: I don't think you should hardcode the CardId#asahi-dev · 2023-09-25 11:02
jannegchadmed: even more? I already fixed 1 1/5 issues last week breaking speaker detection#asahi-dev · 2023-09-25 11:00
jannegsven: oh. I hope you'll get better soon#asahi-dev · 2023-09-25 07:01
jannegsven: did your kernel recipes talk get rescheduled into the void? I remember seeing it scheduled for today at 16:00 (CEST) but currently I don't see it at all on the schedule page#asahi-dev · 2023-09-24 22:22
jannegdiskutil (command line) should display it#asahi-dev · 2023-09-24 22:20
jannegappend "chosen.asahi,efi-system-partition=${ESP_PART_UUID}\n" to the m1n1.bin you're chainloading. use low case letters in the UUID#asahi-dev · 2023-09-24 22:19
jannegstill using linux.py?#asahi-dev · 2023-09-24 22:15
jannegfor UUIDs you need a warking initrd. use disk/partition numbers instead#asahi-dev · 2023-09-24 22:14
jannegpossible that linux.py can only handle uncompressed initrd's. I only remember there was something special#asahi-dev · 2023-09-24 22:13
jannegbut that sounds that it neither finds the initrd or a rootfs. in the beginning it's easier to boot without initrd and modules and simply build everything into the kernel image#asahi-dev · 2023-09-24 22:11
jannegtry usb-c, usb-a is more complex and needs firmware (and firmware handling in the initrd)#asahi-dev · 2023-09-24 22:08
jannegthe hypervisor has the advantage of proving a virtual serial on the second acm interface making debugging easier#asahi-dev · 2023-09-24 22:03
jannegI'll try to emulate on the m2 pro tomorrow#asahi-dev · 2023-09-24 22:02
jannegyes, in t602x-pmgr.dtsi, for the change you can ignore the DIE_NODE stuff#asahi-dev · 2023-09-24 22:01
jannegthat will solve the double chainload annoyance. you need to chainload once before run_guest_kernel to ensure the HV interfaces matches and run_guest_kernel will exucute m1n1 in the hypervisor#asahi-dev · 2023-09-24 21:59
jannegis the initramfs compressed? I thiink most of use use tools/run_guest_kernel.sh nowadays#asahi-dev · 2023-09-24 21:57
jannegwe will need that anyway#asahi-dev · 2023-09-24 21:56
jannegplease do#asahi-dev · 2023-09-24 21:56
jannegleio: I would have thought that #192 is sufficient but ps_dispext0_cpu0 might need 'apple,min-state = <4>;'#asahi-dev · 2023-09-24 21:38
jannegmodule or built-in for CONFIG_DRM_APPLE shouldn't make a difference but in doubt disable it#asahi-dev · 2023-09-24 21:37
jannegleio: try booting with the kernel cmdline parameter pd_ignore_unused#asahi-dev · 2023-09-24 20:19
jannegyes, you probably wouldn't even have run into the issue on 6.5-3 unless you copied the fedora kernel config#asahi-dev · 2023-09-24 17:42
jannegstek: asahi-${date} tags in https://gitlab.freedesktop.org/asahi/mesa not really a dev topic, in doubt look what the distro packages do#asahi-dev · 2023-09-24 13:00
jannegI would expect that u-boot and linux can use the framebuffer even if there is just a single m1n1 initializing the display#asahi-dev · 2023-09-24 12:58
jannegi.e. it will not be noticeable once the 1st stage m1n1 inits the display#asahi-dev · 2023-09-24 12:57
jannegleio: that's not the "problem". the video.base stays the same all the time. I'm not even sure this behavior differs from from m1 ultra. it's just usually hidden since we have 2 m1n1 instances initializing the display anyway#asahi-dev · 2023-09-24 11:49
jannegOh, I've never looked at the nvram on m2 devices. Apple changed the nvram format: https://github.com/AsahiLinux/linux/issues/193#asahi-dev · 2023-09-23 12:37
jannegthere was a at-spi update recently which might explain the breakage#asahi-dev · 2023-09-23 12:25
jannegnot reproducible here with plasma wayland but the report is pretty useless#asahi-dev · 2023-09-23 12:24
jannegcmp https://www.reddit.com/r/AsahiLinux/comments/16q2pe5/internal_keyboard_not_working_after_pacman_syu/#asahi-dev · 2023-09-23 12:23
jannegmarcan: can we get rid of libinput in the asahi repo, we're behind one version (alarm has 1.24.0) and I suppose as long as it's in asahi it will not switch repos#asahi-dev · 2023-09-22 20:08
jannegrenaming boot.bin seems esier than that#asahi-dev · 2023-09-22 20:07
jannegthe 2nd stage m1n1 falls back to proxy mode sionce there is no devicetree#asahi-dev · 2023-09-22 20:07
jannegyes, but I think it's atm not necessary for the m2 ultra mac studio#asahi-dev · 2023-09-22 20:06
jannegotherwise it will result in a fallback to proxy mode in the 1st stage m1n1#asahi-dev · 2023-09-22 20:05
jannegah, yes not necessary if it misses the dtb for the mac studio anyway#asahi-dev · 2023-09-22 20:04
jannegno boot.bin is the 2nd stage m1n1 holding m1n1, dtbs and u-boot#asahi-dev · 2023-09-22 20:02
jannegafter that rename m1n1/boot.bin on the ESP and then start with the kernel booting sections on the bottom of https://github.com/AsahiLinux/docs/wiki/Tethered-Boot-Setup-(For-Developers) #asahi-dev · 2023-09-22 20:00
jannegI'd recommend to install fedora-asahi-remix, that images already support 13.5#asahi-dev · 2023-09-22 19:57
janneghmm, I think after the display init it should start printing the log from "fb console: max rows ..." on#asahi-dev · 2023-09-22 19:54
jannegbut I don't think there any tricky bits, in doubt read the installer source code#asahi-dev · 2023-09-22 19:52
jannegthere is https://github.com/AsahiLinux/docs/wiki/Distro:Differences-from-Arch-Linux-ARM#image-format with a dead link to https://github.com/AsahiLinux/docs/wiki/Installer%3AOS-Packages#asahi-dev · 2023-09-22 19:47
jannegthe black screen after the first chainload is just a side effect of starting with an uninitilized display and iboot not setting an apple logo in the framebuffer#asahi-dev · 2023-09-22 19:45
janneghmm, no I think#asahi-dev · 2023-09-22 19:33
jannegleio: see https://github.com/AsahiLinux/m1n1/blob/main/udev/80-m1n1.rules for stable symlinks#asahi-dev · 2023-09-22 19:32
jannegyes, you need https://github.com/AsahiLinux/linux/pull/180 and https://github.com/AsahiLinux/linux/pull/192 for the kernel#asahi-dev · 2023-09-22 19:31
janneglooks like display_map_fb() misses to update video.base. I'll prepare a patch to test