#asahi · 2023-11-18 17:55
povikhm#asahi · 2023-11-18 17:54
poviksorry, should have been `modprobe snd_soc_macaudio please_blow_up_my_speakers=1`#asahi · 2023-11-18 17:54
povikwhat are you doing prior to modprobe `snd_soc_macaudio please_blow_up_my_speakers=1` that makes the module unloaded to start with?#asahi · 2023-11-18 17:54
povikis there anything else you should point out about your setup?#asahi · 2023-11-18 17:52
povikreverting that won't fix anything#asahi · 2023-11-18 17:52
povikyeah, don't bother#asahi · 2023-11-18 17:52
povikhere's what i reserved the 2 % for :p#asahi · 2023-11-18 17:52
povikah#asahi · 2023-11-18 17:51
povikyou have it applied and still get that error?#asahi · 2023-11-18 17:51
povikno i don't#asahi · 2023-11-18 17:47
poviki am 98 % certain that's a fix for what you have#asahi · 2023-08-24 17:24
povik:)#asahi · 2023-08-24 17:24
povikit does?#asahi · 2023-08-11 16:09
povikhandlerug: please if you get an estimate for how often it's occurring let me know, also on what device#asahi · 2023-08-11 15:04
povikthe fact you are hitting this on a stock fedora asahi is concerning#asahi · 2023-08-11 15:04
povikit seems like a race between some iommu init and alsa init#asahi · 2023-08-11 14:42
povikis this stock asahi, or some custom kernel config?#asahi · 2023-08-11 14:41
poviki have never replicated it, i think#asahi · 2023-08-11 14:41
povikhandlerug: it's a failure mode we have seen a couple of times in the past, never got to the root cause#asahi · 2023-05-29 20:56
povikfwiw we do ship audio support on the headphone jack#asahi · 2023-05-15 09:48
povikyeah, it's not in for that tag#asahi · 2023-05-15 09:46
poviklet's see, 6.2.0-asahi-11-1#asahi · 2023-05-15 09:46
povikhttps://github.com/AsahiLinux/linux/commit/a83cba6338dc1a40eaff3acf45fadfff71e6da6b#asahi · 2023-05-15 09:46
povikthis should have made it go away#asahi · 2023-04-24 14:07
povikyeah, and if the controls are not enough, there are options on the codec to boost the signal even further, though they are not supported by the kernel driver yet#asahi · 2023-04-20 20:24
povikis looking forward to the next one#asahi · 2023-04-20 19:47
povik*necessarily#asahi · 2023-04-20 19:46
povikbut that's not necessary in conflict with your telling of the story...#asahi · 2023-04-20 19:46
povikand the only reason i didn't go ahead with grabbing the responses (yet) is because you beat me to it in constructing an EQ setup anew by ear#asahi · 2023-04-20 19:41
povikalso how we source the DSP data is orthogonal to whether we apply it in kernel or userspace#asahi · 2023-04-20 19:41
povikyeah, i don't think we were ever considering doing anything but grabbing the impulse response of macOS by tracing in hypervisor#asahi · 2023-04-16 21:22
povikyes#asahi · 2023-04-16 20:27
povikbut you know what *does* set it? the dummy codec driver that's involved in the link because of DPCM nuances#asahi · 2023-04-16 20:27
povikat first i was puzzled since neither MCA nor the codec drivers set the use_pmdown_time flag#asahi · 2023-04-16 20:27
povikhttps://github.com/torvalds/linux/blob/6c538e1adbfc696ac4747fb10d63e704344f763d/sound/soc/soc-pcm.c#L318#asahi · 2023-04-16 20:26
povikon a technical side note, i am reading the relevant ASoC code that decides whether to do the pmdown deferral#asahi · 2023-04-16 20:20
povikoh, sorry, that should be zerdox:#asahi · 2023-04-16 20:20
povikzeromind: putting snd-soc-core.pmdown_time=0 into your kernel bootargs might be a workaround, if you are up for trying things before there's a proper fix#asahi · 2023-04-16 20:07
povikthere should be a fix in a kernel release soon#asahi · 2023-04-16 20:07
povikzerdox: that's a known issue some people were hitting that we finally got to understand today#asahi · 2023-03-22 10:54
povikhuh#asahi · 2023-03-06 21:40
poviksven: to clarify, you meant to write it's *not* about poking hardware, right?#asahi · 2023-03-06 13:49
povik:D#asahi · 2023-03-06 13:47
povikis that some python binding, or you wrote ctypes wrappers?#asahi · 2023-03-05 14:19
povikhttps://github.com/AsahiLinux/linux/blob/2fb9510f953092b1cc94e0cb26713f2f152754f8/drivers/input/misc/macsmc-hid.c#L85#asahi · 2023-03-05 14:19
povikindeed, here's the kernel driver code#asahi · 2023-02-23 20:08
poviksetting me up for a guiltless squashing one day...#asahi · 2023-02-23 20:08
povik:)#asahi · 2023-02-23 20:00
povik(otherwise there would be too many)#asahi · 2023-02-23 20:00
povikthough when it will get upstreamed it will get squashed and you will probably get no attribution