#asahi-dev · 2026-01-07 14:47
chaos_princessand was the second boot (after going to macos) with the display plugged in or not?#asahi-dev · 2026-01-07 14:46
chaos_princess"charger event" is unrelated#asahi-dev · 2026-01-07 14:42
sklugkernel. Now everything seems smooth again.#asahi-dev · 2026-01-07 14:42
sklugduring boot, including some "Unknown charger event". After that, the keyboard wouldn't be usable in grub, u-boot showed different resolution and the luks decryption didn't show a gui anymore. The system would hang after a short while when booting succeeded. Booting to macOS (in the hope to reset something) didn't help. I now reinstalled m1n1 and grub and rolled back to the older#asahi-dev · 2026-01-07 14:42
sklugHi everyone. a short heads up on the fairydust kernel (I used a debian package shared on the ubuntu-asahi irc). I observed a bit of a scary situation. I don't know the root cause, but I thought I'd at least mention it. First DP-altmode was working great. Then, after rebooting while the USB-C display (That also provides charging over USB-C) was still attached, I got lots of errors#asahi-dev · 2026-01-07 13:34
nickchanadds a node for an existing driver#asahi-dev · 2026-01-07 13:34
nickchansven: right that's the only thing left needing attention#asahi-dev · 2026-01-07 12:01
svennickchan: 20251231-b4-j98a-j99a-dwi-bl-v1-1-24793c2b99fc@gmail.com is your only patch left that still needs attention or did i miss something on the list?#asahi-dev · 2026-01-07 11:37
svenI’ll reply on the list in the evening #asahi-dev · 2026-01-07 11:36
svenjannau: merge policy for defconfig is that we pick it up and send it as a separate PR upstream #asahi-dev · 2026-01-07 11:12
jannauoh, there is scripts/dtc/dt_to_config in the kernel. that might be useful and looks worth testing (and fixing if there's something missing)#asahi-dev · 2026-01-07 10:47
integralpilotoh sorry just logged on and saw these! glad it could help with A11 stuff, thank you for looking through it! :)#asahi-dev · 2026-01-07 10:28
nickchanokay spotted some issues with the driver will give a review soon#asahi-dev · 2026-01-07 09:39
nickchanlike how to make the battery actually... charge since it does not auto-charge#asahi-dev · 2026-01-07 09:07
nickchanwould be a while before those changes are ready though there's some sigificant issues#asahi-dev · 2026-01-07 09:00
nickchanthough I have yet to rebase my changes on your patch...#asahi-dev · 2026-01-07 08:59
nickchanchanging it to adding properties makes it more viable to support apple a11#asahi-dev · 2026-01-07 08:58
nickchanthe out of tree driver's way of reducing num_properties just doesn't work well for me on A11 #asahi-dev · 2026-01-07 08:58
nickchanintegralpilot: good work on the smc battery driver!#asahi-dev · 2026-01-07 08:48
chadmedthey still cause very noticeable page flip delay#asahi-dev · 2026-01-07 08:47
chadmedbacklight changes dont feel that much faster on m2 air#asahi-dev · 2026-01-07 06:58
svenand let’s not talk about atcphy :p#asahi-dev · 2026-01-07 06:56
svenSMC took e.g. 2 months or so until the core was upstream #asahi-dev · 2026-01-07 06:56
svenyuuup, and this stuff just takes a while #asahi-dev · 2026-01-07 03:24
JamesCalligeros[m]hence why clearing all this stuff up and moving to upstream-first is important#asahi-dev · 2026-01-07 03:23
JamesCalligeros[m]no one can beat the "ship at all costs" downstream BSP kernel allegations sometimes, not even us #asahi-dev · 2026-01-07 03:23
JamesCalligeros[m]thats okay, its a bit of a theme with some of the drivers#asahi-dev · 2026-01-07 03:21
integralpilotsorry MTP/dockchannel is taking so long, a lot of the code is hacky and not upstreamable so I’m having to rewrite a lot#asahi-dev · 2026-01-06 23:25
JamesCalligeros[m]nice, i wonder if this also fixes m2 mac mini #asahi-dev · 2026-01-06 23:24
jannaualso suspend/resume while hdmi on macbook pros is active is fixed in asahi-wip#asahi-dev · 2026-01-06 23:19
jannauespecially for the type-c ports it would have been nice if we had direct access to the aux channel#asahi-dev · 2026-01-06 23:19
JamesCalligeros[m]right, we probably want to stash the current brightness etc in apple_crtc anyway then#asahi-dev · 2026-01-06 23:18
jannauthere are other commands to route ddc through#asahi-dev · 2026-01-06 23:17
jannauwe have to go through one of the afk endpoints. avserv iirc. the same one which is currently used to get the EDID#asahi-dev · 2026-01-06 23:15
JamesCalligeros[m]istr its just hanging off one of the i2c busses we can access#asahi-dev · 2026-01-06 23:15
JamesCalligeros[m]btw speaking of backlights, do we have direct access to ddc/ci or do we have to go through iomfb#asahi-dev · 2026-01-06 23:09
JamesCalligeros[m]i think it was the former thing. it used to lock up the entire display while rapidly changing the backlight#asahi-dev · 2026-01-06 23:08
jannaubacklight tries to avoid doing a commit to update the brightness so it waits to see if there's a commit/page flip in a reasonable time before it commits itself#asahi-dev · 2026-01-06 23:07
jannauJamesCalligeros[m]: backlight updates might be faster in asahi wip. it was lovking way too many modeset locks. but I suspect you're complaining about the lazy updates#asahi-dev · 2026-01-06 23:05
JamesCalligeros[m]and is the same kb we had to add the usb quirk to u-boot for as it does not respond to hid polling unless a key is pressed#asahi-dev · 2026-01-06 23:05
JamesCalligeros[m]i was just testing with my keychron c2, which reports as an apple aluminium clone#asahi-dev · 2026-01-06 23:04
jannautesting with a network adapter on 14-inch macbook pro and at least in one case the network cable wasn't plugged fully (before I started debugging)#asahi-dev · 2026-01-06 23:02
jannaureproduces way less than 1/3 here. interestingly both ports were affected in my single verified reproduction#asahi-dev · 2026-01-06 23:02
JamesCalligeros[m]could it be a core regression in 6.18? i didnt try grafting the upstream dwc stuff on to 6.17#asahi-dev · 2026-01-06 23:00
jannauno luck with the broken usb-c ports on initial cold/hot plug. from tipd, usb-mux, usb-switch, usb-role-switch, dwc3 every thing looks fine and identical between broken and working state#asahi-dev · 2026-01-06 19:20
fl0_idok sorry#asahi-dev · 2026-01-06 19:18
sven*link#asahi-dev · 2026-01-06 19:18
svenplease don't like slop here when you can't tell yourself if it's actually ccurate#asahi-dev · 2026-01-06 19:08
fl0_idbtw this is for raspberry pi and seems a bit like written by AI, but assuming it's mostly accurate at least is useful to me re early boot https://www.linkedin.com/pulse/comprehensive-guide-early-boot-uart-configuration-tty-david-zhu-2qvec/#asahi-dev · 2026-01-06 16:08
jannauI'm not sure if efi_stub causes issue w.r.t to symbols, it doesn't for scripts/faddr2line