#
✕ nick:povik
#asahi-dev · 2023-08-14 11:21
povikit's controlled with a bitmask in an `apple_quirks` parameter on the tas2764 module
#asahi-dev · 2023-08-14 11:21
povikhttps://github.com/AsahiLinux/linux/blob/sound/ivsense-improvements/sound/soc/codecs/tas2764-quirks.h
#asahi-dev · 2023-08-14 11:21
povikin the ivsense branches the linux driver can mimick the apple quirks
#asahi-dev · 2023-08-14 11:20
povikfwiw apple configures some undocumented registers on the codec, we don't know to what end
#asahi-dev · 2023-08-14 11:17
povikhow does making it sound good at full volume work around the clicks if we don't know what causes them?
#asahi-dev · 2023-08-14 11:13
povikhm, no faults reported from the codec? (in dmesg)
#asahi-dev · 2023-08-14 11:11
povikchadmed: elaborate on the popping noise, i don't understand
#asahi-dev · 2023-08-14 09:29
povikno we haven't
#asahi-dev · 2023-08-14 09:27
povikhaven't we merged that yet?
#asahi-dev · 2023-08-13 11:30
povikand also mostly ortohogonal to the safety work...
#asahi-dev · 2023-08-04 11:16
povikthey are upstreaming a driver before they even have it listed on their website
#asahi-dev · 2023-08-04 11:16
povikbut it's not ~cs42l84
#asahi-dev · 2023-08-04 11:15
povikthere are one or two similarities to cs42l84 in the register interface
#asahi-dev · 2023-08-04 11:15
povikhttps://lore.kernel.org/all/20230804104602.395892-1-ckeepax@opensource.cirrus.com/
#asahi-dev · 2023-08-04 11:15
povikbabe wake up, new cirrus codec just dropped
#asahi-dev · 2023-07-18 19:59
povikyou mean there's another way? :D
#asahi-dev · 2023-07-18 12:29
poviktimeless
#asahi-dev · 2023-07-18 12:27
poviki see some filtering and sample arithmetic there :)
#asahi-dev · 2023-07-18 12:27
povikha, that already does something
#asahi-dev · 2023-07-18 12:26
poviki was wondering if it would be useful if i put together a jupyter notebook demonstrator of the algorithm we want
#asahi-dev · 2023-07-18 12:25
poviki saw there's someone written there as a contact, is it being worked on?
#asahi-dev · 2023-07-18 12:25
povikchadmed: btw what's the status of the bass yak?
#asahi-dev · 2023-07-18 10:34
povikeiln: congratulations on making a friend!
#asahi-dev · 2023-07-12 09:47
povikthe driver made it to Linuses tree but the maintainers entries didn't
#asahi-dev · 2023-07-12 09:47
povikit's there when Mark merges it: https://github.com/torvalds/linux/commit/7590c6a89
#asahi-dev · 2023-07-12 09:47
povikhuh, the MAINTAINERS entries for ssm3515 got lost somewhere while stepping up the merge ladder up to Linus
#asahi-dev · 2023-06-21 11:56
povikbut the code in ca9fa7 looks good
#asahi-dev · 2023-06-21 11:56
poviki think if there's one or few registers that we need to move up in the restore order, there's a regcache_sync variant we can use for that
#asahi-dev · 2023-06-21 11:52
povikand i have a feeling we were thinking that out once before, not sure what we arrived at
#asahi-dev · 2023-06-21 11:51
povikone potential source of issues is the order in which the registers are restored
#asahi-dev · 2023-06-21 11:49
povikif it wasn't, the ordinary bootup wouldn't be reliable too
#asahi-dev · 2023-06-21 11:48
povikif you do the same delay that's there for the ordinary bootup path then that should be enough
#asahi-dev · 2023-06-07 16:34
povikbut maybe it wouldn't be an issue in practice
#asahi-dev · 2023-06-07 16:34
povikthere's things that would make me uneasy about this, like the AOP then thinking it owns some of the pstate regs involved, which it could reset at any point
#asahi-dev · 2023-06-07 16:31
poviki wasn't considering having a mixed scenario, in which we do boot the AOP and circumvent it for audio still
#asahi-dev · 2023-06-07 16:31
povik(since presumably an AOP-based implementation of the same would be a soundcard of its own)
#asahi-dev · 2023-06-07 16:31
povike.g. making sure the macaudio driver and the use-case configuration falls back nicely if the mic input pcm isn't there
#asahi-dev · 2023-06-07 16:30
poviki mean, i was planning on not closing the doors if we need to walk this decision back
#asahi-dev · 2023-06-07 15:25
povikat least now when the work involved in figuring out the DSP thing has been done...
#asahi-dev · 2023-06-07 15:25
poviki think a good case can be made for it
#asahi-dev · 2023-06-07 15:25
povikmarcan: that's the approach i took with the kernel driver, yeah
#asahi-dev · 2023-05-27 00:48
povikbut even then we want a userspace demonstrator first, that shows the benefit
#asahi-dev · 2023-05-27 00:47
povikwe can move array processing off to the firmware, that's actually part of the value proposition here
#asahi-dev · 2023-05-27 00:45
povikcy8aer: that might come to be useful
#asahi-dev · 2023-05-27 00:44
povikalyssa: glad you get a kick out of it :-p
#asahi-dev · 2023-05-26 05:50
povikhttps://github.com/AsahiLinux/m1n1/pull/314
#asahi-dev · 2023-05-26 05:50
povikalso check out my lucky m1n1 pr number
#asahi-dev · 2023-05-26 05:49
povikflushed a bunch of LEAP/microphone work: https://github.com/AsahiLinux/linux/pull/151
#asahi-dev · 2023-05-22 10:36
povikah, gaming
#asahi-dev · 2023-05-22 10:36
povikwill test next time i boot to macos

plus ›