#
✕ nick:_jannau_
#asahi · 2023-03-23 10:09
_jannau_the kernel can not handle it since it has no way to save values so it can load them on the next boot
#asahi · 2023-03-23 09:43
_jannau_we can not make them persistent so it has to be solved by software on the system
#asahi · 2023-03-23 09:40
_jannau_mkurz: can you try to press Esc the next time the display stays black (if you not already tried that)
#asahi · 2023-03-23 09:40
_jannau_we can't do anything about that. easiest solution for now is to have a ssytemd service which writes the start/stop threshold on boot
#asahi · 2023-03-22 09:51
_jannau_those i2c messages should contain cs42l83 or cs42l84? although if you heard something it probably is the speaker? is this a mac mini?
#asahi · 2023-03-22 09:50
_jannau_which device?
#asahi · 2023-03-22 09:49
_jannau_cs42l83 or cs42l84?
#asahi · 2023-03-21 16:39
_jannau_cable on the low resolution image looks also if it has just a single ring, i.e. 3 connectors for stereo audio
#asahi · 2023-03-21 16:32
_jannau_d4ve: the connection for the microphone is not standardized: https://en.wikipedia.org/wiki/Phone_connector_(audio)#TRRS_standards
#asahi · 2023-03-21 16:14
_jannau_that explains why it's not working. are the headset working in macos?
#asahi · 2023-03-21 16:11
_jannau_d4ve: is there a 'cs42l84 .* Detected mic' message in the kernel log when plugging the headset in?
#asahi · 2023-03-21 14:55
_jannau_headset mic works with pipewire as well but might be very quiet. there are two 'jack adc ... gain' controls which are probably at zero. try to raise those with `alsamixer -c 0`
#asahi · 2023-03-21 09:20
_jannau_ruby-bump and mesa-demos are both installing /usr/bin/bump
#asahi · 2023-01-12 14:44
_jannau_it's a known issue that mac os updates sometimes break other os boot policies but it's unclear what exactly triggers it
#asahi · 2023-01-12 14:43
_jannau_most likely fixed by reinstalling the 1st stage m1n1. I think the installer's fix option should be able to handle that
#asahi · 2023-01-09 11:34
_jannau_supporting this in the dcp driver as such is not an issue but figuring out how to expose that to user space is. I don't want to add module parameters for that
#asahi · 2023-01-09 11:31
_jannau_it will be useful to identify runtime propertiees (they are numerical on the iomfb FW interface)
#asahi · 2023-01-09 11:30
_jannau_yes, linux/the dcp driver dowesn't care about those
#asahi · 2022-12-20 15:05
_jannau_no
#asahi · 2022-12-20 15:00
_jannau_the modesetting/apple OutputClass xorg conf fragment is needed. there are configurations were it is not needed, for example dcp built-in and asahi as module
#asahi · 2022-12-12 17:18
_jannau_yes, that sounds exactly like the issue I've seen. should be fixed by commenting XRGB2101010 and ARGB2101010 from the supported color formats
#asahi · 2022-12-12 17:15
_jannau_I tested asahi-wip during the stream and it still looks ok
#asahi · 2022-12-12 17:15
_jannau_colors are too dark / oversatured?
#asahi · 2022-12-12 17:14
_jannau_marcan: can you send me a dcp.log from trace_dcp.py? it sounds like the issue I saw originally with AR30
#asahi · 2022-12-12 16:06
_jannau_already fixed but not pushed
#asahi · 2022-12-12 16:06
_jannau_robher: marcan discovered today that the cpufreq driver leaves the CPUs in a random p-state at suspend and leaves them alone while suspended
#asahi · 2022-12-09 13:49
_jannau_I think the timeout is 3 minutes and might be a dcp problem
#asahi · 2022-12-08 09:46
_jannau_the modeset driver's atomic modeset implementation was/is apparently that broken that it was forcefully disabled on kernel side
#asahi · 2022-12-08 09:44
_jannau_not sure if you have seen https://lore.kernel.org/dri-devel/CAPM=9twc_TBtG_654Hm4SV_G1Ar+PiCuZGg1fV-Zooga+4S0GQ@mail.gmail.com/
#asahi · 2022-12-08 09:43
_jannau_i.e. the workaround will hurt the well behaved clients
#asahi · 2022-12-08 09:42
_jannau_yes, I thought about that but I fear that it leads to a ton of bugs for clients doing swaps
#asahi · 2022-12-08 09:36
_jannau_she removed the drm_vblank initilisation, dcp still emits vblank events after page swap. I'm not sure if those are still registered as vblank irqs
#asahi · 2022-12-08 09:31
_jannau_I would expect Xorg with tearfree to be less broken with dcp. It should receive fake vblank irq on framebuffer swaps
#asahi · 2022-12-07 16:26
_jannau_that sounds like a (fractional) scaling like the one where xdpyinfo reports dimensions being off by one
#asahi · 2022-12-07 13:41
_jannau_macbook pro 13" m2 reports 3W under similar conditions
#asahi · 2022-12-07 13:27
_jannau_I think we might have a power management issue with mbp 14"/16". 9-10W idle with 100 nits display seems too much
#asahi · 2022-12-07 13:26
_jannau_I'm also not satiesfied with the reported battery health of my mbp 14" which was mostly used stationary. I think it improved after a discharge /charge cycle in macos
#asahi · 2022-12-07 10:23
_jannau_morro: on an apple silicon device? I think she's not using that yet. screen capture requires cooperation with the compositor under wayland. please search for "$compositor screen capture" first
#asahi · 2022-12-06 08:28
_jannau_Glanzmann: running (a subset) of phoronix test suite. https://www.phoronix.com/review/apple-m2-december
#asahi · 2022-12-06 08:27
_jannau_running low level benchmarks I doubt any of them would test window resize performance
#asahi · 2022-12-06 08:26
_jannau_the ncnn results look sketchy. 5 benchmarks "regressed" by 30-60% but one "improved" 50%. my first suspicion would be thermal throttling on the mb air m2 used for benchmarking
#asahi · 2022-12-02 09:27
_jannau_err
#asahi · 2022-12-02 09:27
_jannau_ls for dcp
#asahi · 2022-11-30 16:29
_jannau_that's new enough to have have all te bits in place to support multiple installs so it's probably an issue in the debian installer
#asahi · 2022-11-30 16:26
_jannau_janrinze: from when was you asahi linux install? was it the first alpha release or later?
#asahi · 2022-11-30 16:25
_jannau_that sounds like the debian installer is broken and has overwritten files on the asahi ESP
#asahi · 2022-11-24 13:43
_jannau_I think /sys/class/backlight is only for display backlights. the gpio one is/was for the panel backlight and controls the power supply
#asahi · 2022-11-22 16:40
_jannau_also ignoring the obvious problem of the missing support for other OS'
#asahi · 2022-11-18 16:18
_jannau_but the switch works? does the tty update if you type?
#asahi · 2022-11-18 16:16
_jannau_axboe: does it start working if you switch to a tty and back to X11?

plus ›