#asahi-dev · 2022-11-09 23:15
povikyeah, seen that once or twice#asahi-dev · 2022-11-04 21:30
povikwait, what's that? how does ane/isp need to know anything about apfs?#asahi-dev · 2022-11-01 17:58
povikgn#asahi-dev · 2022-11-01 17:52
poviklike i said, not important for the release#asahi-dev · 2022-11-01 17:50
povikmost of it on way upstream already, pertains to speakers so low priority#asahi-dev · 2022-11-01 17:50
povikhas https://github.com/povik/linux/commits/audio-fixes#asahi-dev · 2022-11-01 13:16
povikjannau: you get roughly that maximum at x=1.0, as you probably noticed#asahi-dev · 2022-11-01 13:13
poviknot that ~870 amps is sensible either :D#asahi-dev · 2022-11-01 13:12
povikwhat if x plugged into the polynomial is [0.0,1.0] ?#asahi-dev · 2022-10-31 09:03
povikccs1: there was ilya who expressed interest some ten days ago and i gave them a walkthrough of the task. though they didn't have an imac but wanted to try to start writing the driver still. i don't expect them to finish it lacking hardware to test on#asahi-dev · 2022-10-28 07:13
povikit's a module parameter so you can play around with it#asahi-dev · 2022-10-28 07:13
povik * due to DAPM power cycling.#asahi-dev · 2022-10-28 07:13
povik * This is to ensure there are no pops or clicks in between any music tracks#asahi-dev · 2022-10-28 07:13
povik * Power down the audio subsystem pmdown_time msecs after close is called.#asahi-dev · 2022-10-28 07:13
povikto quote:#asahi-dev · 2022-10-28 07:13
poviksee pmdown_time in soc-core.c#asahi-dev · 2022-10-28 07:11
povikchadmed: there's a timer for this in ASoC, let me look it up#asahi-dev · 2022-10-26 15:12
povikyeah, macos host here#asahi-dev · 2022-10-26 10:32
povikso it's unreliable transfer of random data#asahi-dev · 2022-10-26 10:32
povikif i enable it (https://tpaste.us/6Elz) i get checksum errors#asahi-dev · 2022-10-26 10:32
poviki thought it's there still#asahi-dev · 2022-10-26 10:32
povikah checksumming is disabled over USB#asahi-dev · 2022-10-26 10:16
povikwait, now it happens if i put out the memcpy#asahi-dev · 2022-10-26 10:14
povikand most often the start is aligned to 0x10#asahi-dev · 2022-10-26 10:14
povikit's typically 0x40- to 0x180-long chunks containing differing random-looking data at random locations#asahi-dev · 2022-10-26 09:58
poviknah, that's not it. the malloc seems to have some minimum alignment (as i would expect)#asahi-dev · 2022-10-26 09:57
povikwait the malloc might not be aligned -- now that i read the memcpy32 comment in m1n1#asahi-dev · 2022-10-26 09:53
povikaround one in ~50 trials in the loop fails#asahi-dev · 2022-10-26 09:53
povikthought you had an answer for a second there#asahi-dev · 2022-10-26 09:53
povikahh, darn :(#asahi-dev · 2022-10-26 09:52
povikthis is tip m1n1 chainloaded by older m1n1 on t8103#asahi-dev · 2022-10-26 09:52
povikhttps://tpaste.us/PjlY#asahi-dev · 2022-10-26 09:52
povikany idea why shuffling data around in proxyclient wouldn't be reliable?#asahi-dev · 2022-10-25 11:39
povikmarcan: i've invited you to povik/alsa-ucm-conf-asahi so you can push any fixes you have. feel free to recreate the repo under AsahiLinux (or we can do a transfer so there's a redirect, but i would need repo-creation permissions for the ocassion)#asahi-dev · 2022-10-25 08:34
povikalyssa: :D, very nice#asahi-dev · 2022-10-24 12:19
povikonce the time comes i will squash it into base macaudio commit for upstream submission, is that okay by you?#asahi-dev · 2022-10-24 12:19
povikchadmed: no worries. the commit message isn't perfect but we can pick it up all the same#asahi-dev · 2022-10-24 11:58
povikcool#asahi-dev · 2022-10-24 11:49
povikchadmed: looking good apart from that dpcm loop#asahi-dev · 2022-10-24 10:58
povikjust a bunch of constraints#asahi-dev · 2022-10-24 10:58
povikthat fe_chans is meaningless, by .startup stage the PCM doesn't yet have a fixed number of channels#asahi-dev · 2022-10-24 10:57
povikyou should apply an upper bound constraint, as i said#asahi-dev · 2022-10-24 10:57
povikthat's still not okay#asahi-dev · 2022-10-24 10:56
povikor rather the primary frontend, it gets a <2 channels constraint from MCA because of slots#asahi-dev · 2022-10-24 10:55
povikcould it be because you were constraining headphones to 6 channels?#asahi-dev · 2022-10-24 10:55
povikso it knew at that point already the configuration is invalid#asahi-dev · 2022-10-24 10:55
povikthe -EINVAL was probably because there was zero intersection of the applied constraints#asahi-dev · 2022-10-24 10:54
povikso that print doesn't mean anything#asahi-dev · 2022-10-24 10:54
poviki mean you can't be looking into hw_params from .startup callback#asahi-dev · 2022-10-24 10:49
povikyou may as well consider it uninitialized