#asahi · 2024-09-09 20:25
jannauin the amd DP edid? my edid-decode can't parse it#asahi · 2024-09-09 20:23
jannauI guess you want the DisplayID block from the first edid in https://bbs.archlinux.org/viewtopic.php?pid=2184209#p2184209#asahi · 2024-09-09 20:22
jannaufrom the edid editor? Ignore it#asahi · 2024-09-09 19:10
jannauthe soc's cache and memory system is likely part of the problem#asahi · 2024-09-09 19:07
jannaumkurz: it's not just suspend. the m1 pro/max have too high idle power use#asahi · 2024-09-09 17:56
jannaunot sure if the display exports a specific EDID in that situation#asahi · 2024-09-09 17:55
jannauHDMI on AMD should be limited to 60Hz since there is no HDMI 2.1 support#asahi · 2024-09-09 17:54
jannauthat was amd via DP?#asahi · 2024-09-09 17:52
jannauthe DisplayID extension block contains a timing descriptor#asahi · 2024-09-09 17:42
jannaubase64 pastebin#asahi · 2024-09-09 17:42
jannauleio: usb storage, email, google drive?#asahi · 2024-09-09 17:30
jannau/sys/class/drm/card?-{HDMI,DP}-*/edid#asahi · 2024-09-09 17:27
jannauand yes, edid from nvidia, intel or amd GPU would be interesting#asahi · 2024-09-09 17:25
jannauhave you tried to read the edid over DP/usb-c?#asahi · 2024-09-09 17:24
jannauit contains 2 DTD but only 2560x1440 120/60 Hz. curiously with numbers 2 and 3 so dcp might redact things#asahi · 2024-09-09 17:23
jannauyes it can parse VTDB. That are either DTD (Detailed Timing Descriptor) or CVT (Coordinated Video Timings) descriptor in the CTA-861 Extension Block#asahi · 2024-09-09 17:15
jannauuse edid-decode / https://gitlab.freedesktop.org/emersion/libdisplay-info/#asahi · 2024-09-09 16:59
jannaudo you happen to have seen EDIDs from displays which support 240Hz on apple silicon?#asahi · 2024-09-09 16:58
jannauleio: I didn't see anything in the edid which suggest the display supports 240Hz except for the max VRR value#asahi · 2024-09-03 06:18
jannauxal: please file an issue in https://github.com/AsahiLinux/linux/ not sure if there are tools to help debug this. A trace of the opengl usage to that time would be nice but probably hard to get#asahi · 2024-08-27 14:20
jannauleio: yes, ideally the raw dump but the edid-decode output should allow creating the binary edid. no 240Hz mode though, maybe that's only offered on the HDMI 2.1 inputs?#asahi · 2024-08-27 14:09
jannau+the#asahi · 2024-08-27 14:09
jannauif you find EDID the monitor reports we could try what modes dcp offers#asahi · 2024-08-27 14:05
jannauleio: if there is a way to support the monitor on macos we should be able to support that as well. it won't be a priority for me though. also 4k 240Hz 10-bit RGB is quite beyond the data rate of 4 DP 1.4 lanes#asahi · 2024-08-27 10:52
jannauI believe it's possible to inject custom modes / edid (BetterDisplay on macOS might support it) but that has to be implemented in the driver#asahi · 2024-08-27 10:46
jannauleio: no, we are limited to what the dcp firmware detects as valid modes. In this regard we're strictly worse than apple since we can't easily update the dcp firmware. So even if Apple fixes this it would remain broken in the macos 13.5 DCP firmware we are using#asahi · 2024-08-25 17:12
jannauIsaw, I'll answer there#asahi · 2024-08-25 17:03
jannauallmazz: which kernel? what does /proc/device-tree/chosen/asahi,iboot1-version contain?#asahi · 2024-08-25 13:22
jannauI assumed kf5-baloo-file would have dependencies which would be removed as well. if it works now nop need to stop/do anything else#asahi · 2024-08-25 13:21
jannauyes, typo, `--skip-broken`, try adding ``--allowerasing` and be watchful what it wants to deinstall#asahi · 2024-08-25 13:17
jannauonce you've upgraded you can try to deinstall kf5-baloo-file#asahi · 2024-08-25 13:14
jannauvmk: try adding `--skiop-broken` there will be probably nothing in f40 which needs kf5-baloo*#asahi · 2024-08-25 08:21
jannauit might be an option but it doesn't look like an easy problem free one#asahi · 2024-08-25 08:19
jannauwe do not support out of tree modules like openzfs and keeping kernel/openzfs release in sync will be effort#asahi · 2024-08-24 21:42
jannauI think so too but I'm not sure 100% with the different ACE / TI type-c controllers with spmi interface on the m3 pro/max#asahi · 2024-08-24 21:38
jannauafter that serial over usb acm and finally serial over emulated tty via m1n1's hypervisor#asahi · 2024-08-24 21:37
jannauserial via a second mac or a usb-pd (https://github.com/AsahiLinux/macvdmtool) chip (https://github.com/AsahiLinux/vdmtool) or something like Central Scrutinizer https://git.kernel.org/pub/scm/linux/kernel/git/maz/cs-hw.git/tree/README.txt#asahi · 2024-08-24 18:26
jannaudid you maybe inadvertently muted vlc as application specific audio setting?#asahi · 2024-08-24 18:20
jannaufor some reason (trying to play mono on a stereo output maybe) `vlc /usr/share/sounds/alsa/Front_Left.wav` only works on the mac studios speakers but not on the headphone out#asahi · 2024-08-24 09:31
jannaucan you provide more details? do other applications work? does the test in the audio settings work?#asahi · 2024-08-24 09:30
jannauno issues here with fully upgraded fedora 40 systems (speaker and headphones)#asahi · 2024-08-22 05:21
jannausupporting macos style external disk boot is not out of the question but someone has to add USB host / storage support to m1n1#asahi · 2024-08-22 05:20
jannaunote that you must not suspend if the rootfs is on usb storage (should be only the usb-c usb4 ports) as those usb ports reset after suspend#asahi · 2024-08-14 05:57
jannauleio: I don't know but my best guess is that we would see the same results as on macos. 4k 240hz is well beyond of what 4 lanes of DP 1.4 can deliver. you end up with DSC and 11 bits per pixel. I'm still surprised that Apple advertises it#asahi · 2024-08-11 08:04
jannauI think the only part which assumes btrfs is the swap creation script (should fail gracefully). please install fedora-asahi-remix-scripts, asahi-platform-metapackage. the rest should be already installed. you will have to regenerate the initramfs after installing fedora-asahi-remix-scripts#asahi · 2024-08-11 07:42
jannauwhat does `rpm -q fedora-asahi-remix-scripts` report?#asahi · 2024-08-11 07:42
jannaufelixphew: does /usr/lib/udev/hwdb.d/65-autosuspend-override-asahi.hwdb exists?#asahi · 2024-08-11 06:42
jannaufelixphew: sounds like the sdcard reader autosuspend issue. Does `systemd-hwdb query pci:v000017A0d00009755` report "ID_AUTOSUSPEND=0"? looks like we might have some issue getting that info into the initramfs#asahi · 2024-08-05 18:14
jannauwe might allow running the fan on full blast via hwmon in the future#asahi · 2024-08-05 18:12
jannauif you do longer multi core benchmarks you might want to to reduce max frequency