#
#asahi-dev · 2026-06-25 11:57
chadmedlmao
#asahi-dev · 2026-06-25 11:56
sofusdont waste your time i found the error
#asahi-dev · 2026-06-25 11:55
sofusoh god
#asahi-dev · 2026-06-25 11:54
chadmedon it
#asahi-dev · 2026-06-25 11:52
sofuson the default branch
#asahi-dev · 2026-06-25 11:51
sofuschadmed: can you test adding an goto done; after the 'invalid fifo slot:' err
#asahi-dev · 2026-06-25 11:34
sofusits only for v4l2 reference stuff, but thats not whats crashing avd
#asahi-dev · 2026-06-25 11:32
sofusit converts an (wrapping) buffer index to an timestamp
#asahi-dev · 2026-06-25 11:32
sofusit could be different for other forks https://code.ffmpeg.org/FFmpeg/FFmpeg/src/commit/3ac4e07593eeeeac8be43e3f6353161a8d1ce8be/libavcodec/v4l2_request.c#L51
#asahi-dev · 2026-06-25 11:29
chadmedin reality very very few videos outside of the itu conformance stress tests wont have them
#asahi-dev · 2026-06-25 11:28
chadmedthis video has timestamps so ffmpeg isnt making up bogus ones
#asahi-dev · 2026-06-25 11:28
sofususing ffmpeg stil makes the hw crash with some videos
#asahi-dev · 2026-06-25 11:27
sofusi disabled some things, if its multi slice then it wont work
#asahi-dev · 2026-06-25 11:26
chadmedhellsing ultimate abridged did not work however
#asahi-dev · 2026-06-25 11:26
sofusalthough the status registers are clearly wrong
#asahi-dev · 2026-06-25 11:25
chadmedindeed it did
#asahi-dev · 2026-06-25 11:25
sofusthat looks like it worked?
#asahi-dev · 2026-06-25 11:24
chadmedhttps://paste.debian.net/hidden/460570c3
#asahi-dev · 2026-06-25 11:23
sofusyes
#asahi-dev · 2026-06-25 11:23
chadmedhow much of the instruction stream do you need? just one frame?
#asahi-dev · 2026-06-25 11:21
chadmedone sec
#asahi-dev · 2026-06-25 11:21
sofuschadmed: is this AUD_MW_E? The instructions look very wrong
#asahi-dev · 2026-06-25 11:13
chadmedi will do t6k tomorrow morning aus/brisbane time
#asahi-dev · 2026-06-25 11:13
chadmedsofus: from t8103 https://paste.debian.net/hidden/e6057765
#asahi-dev · 2026-06-25 10:54
txmxn[m]maettu102: No I mean just the XKB layout. Also I'm not doing anything related to moving the Fn key functionality from the hid_apple kernel driver to userspace XKB, I'm just updating the german Mac layout in XKB.
#asahi-dev · 2026-06-25 10:37
svenwe could also try finding the interrupts
#asahi-dev · 2026-06-25 09:36
integralpilotyuka: i wrote it for ones with interrupts, but you should be able to adapt it to ones without them by busy-polling TX_FREE before writes RX_COUNT before reads and making the interrupt binding (and platform_get_irq etc.) optional
#asahi-dev · 2026-06-25 09:03
yukaintegralpilot: ty, I saw some dockchannel uarts don't have an interrupt in the adt, is this a problem?
#asahi-dev · 2026-06-25 09:02
sofuswe can enable more but yes
#asahi-dev · 2026-06-25 09:01
JamesCalligeros[m]are the two interrupt lines to aic just mbox0 not empty and mbox1 not empty?
#asahi-dev · 2026-06-25 08:59
sofusi forward all unkown irq's: https://github.com/sofus13/avd-fw/blob/eb0d094265933cd359fe0678ca6276fea90ee487/src/irq.c#L14
#asahi-dev · 2026-06-25 08:56
chaos_princessin asahi firmware that is
#asahi-dev · 2026-06-25 08:55
chaos_princessbtw, have you considered just covering every single irq handler in logs to AP
#asahi-dev · 2026-06-25 08:54
sofusfrom what i have seen, t6000 has the same irq handlers and registers as m2
#asahi-dev · 2026-06-25 08:53
barusuintegralpilot: thanks a lot. I'll look at it.
#asahi-dev · 2026-06-25 08:51
sofusJamesCalligeros[m]: yeah, im using eiln's emu. Figuring out what each irq handler does is hard since the firmware is statefull
#asahi-dev · 2026-06-25 08:46
integralpilotah yeah true, don't really have the time today though. thank you for testing, hopefully SMC isn't more broken then before (fingers crossed!)
#asahi-dev · 2026-06-25 08:42
chaos_princessyou actually can, m3-tier is surprisingly usable. but no worries, i will just throw b2 on mine later today
#asahi-dev · 2026-06-25 08:42
integralpilotoh, did you make the fix? people were reporting it was "fixed with the update" but might have had it the wrong way. thanks for making it!
#asahi-dev · 2026-06-25 08:41
integralpilotah, not sure in that case sorry, as i have just seen people reporting that the SMC issue persists. and can't test it myself as i don't have the hardware
#asahi-dev · 2026-06-25 08:41
chaos_princessand it does not fix the boot menu issue, unless "it" refers to me :P
#asahi-dev · 2026-06-25 08:40
chaos_princessi have b1 smc firmware on mine, haven't tried b2 yet
#asahi-dev · 2026-06-25 08:40
chaos_princessi mean does it break smc further compared to b1
#asahi-dev · 2026-06-25 08:40
integralpilotsome unsuspecting people on the reddit have for us lol! i believe it fixes the boot menu issue but still has the battery / SMC firmware change that causes our driver to emergency shutdown
#asahi-dev · 2026-06-25 08:39
chaos_princessbtw, has anyone tested 27b2 on m1/m2?
#asahi-dev · 2026-06-25 08:36
integralpilotbut anyway hope it helps!
#asahi-dev · 2026-06-25 08:36
integralpilotand the coding style is kinda bad and rough too
#asahi-dev · 2026-06-25 08:35
integralpilotjust a few cavets in that: a) its really rough and quick cause i just wanted to prove the mailbox concept and i only tested it a tiny bit and it might not actually work fully, b) it uses the dockchannel through mailbox subsystem, which isn't in downstream, c) there was a warning about using the wrong workqueue that happens when you send data to it, but i didn't look further into it as it appeared to work
#asahi-dev · 2026-06-25 08:34
JamesCalligeros[m](or any of the other cm3 emulators)
#asahi-dev · 2026-06-25 08:34
JamesCalligeros[m]sofus: can you chuck the apple firmware in qemu and see what it does differently for a given input?

plus ›