#
✕ nick:povik
#asahi · 2022-01-20 08:51
povikjoske: in the dmesg it looks like your system tries to suspend. that could be the freeze and hopefully that's not related to audio stuff
#asahi · 2022-01-20 00:40
poviki should really let it go for today
#asahi · 2022-01-20 00:40
povikmaybe this helps
#asahi · 2022-01-20 00:40
povikpatch here: https://tpaste.us/5nly
#asahi · 2022-01-20 00:29
povikreading the datasheet it looks like that can spill over to failed register accesses
#asahi · 2022-01-20 00:29
povik(the clock)
#asahi · 2022-01-20 00:28
povik* it's ready
#asahi · 2022-01-20 00:28
povikthe codec may be upset because it transitions to an external MCLK clock before it
#asahi · 2022-01-20 00:28
povikoh, the i2c issues
#asahi · 2022-01-20 00:27
povikhm?
#asahi · 2022-01-20 00:09
povikanyway i may have found where the i2c issues lie
#asahi · 2022-01-20 00:09
povikyeah, that's right, i meant MIXER knob on the low-level audio card
#asahi · 2022-01-19 23:14
povikhttps://github.com/torvalds/linux/blob/112450df61b7373529b0fe4c122ad13b89d80a8a/drivers/i2c/i2c-core-base.c#L2353
#asahi · 2022-01-19 23:13
poviki would not have guessed that linux by itself scans addresses on i2c buses
#asahi · 2022-01-19 21:41
poviksettled for asm ("hvc %0" : : "I" (42));
#asahi · 2022-01-19 21:25
poviki see
#asahi · 2022-01-19 21:22
povikhm, what would be a nice way to break to hypervisor from linux
#asahi · 2022-01-19 21:20
povikseems they do, good
#asahi · 2022-01-19 21:19
povik(the gpiola traces)
#asahi · 2022-01-19 21:18
poviki wonder if the traces have timestamping in actual time units
#asahi · 2022-01-19 21:17
povikwhy don't i use that to shed some light?
#asahi · 2022-01-19 21:17
povikwhoa, there's whole i2c tracing infrastructure in m1n1 now
#asahi · 2022-01-19 19:45
povik(it's in asahi branch)
#asahi · 2022-01-19 19:45
povikGlanzmann: you are already testing that patch ;)
#asahi · 2022-01-19 19:22
povikthere goes the time spent chasing an impossible error code...
#asahi · 2022-01-19 19:21
povikjust learned that errno numbers are arch-dependent in linux
#asahi · 2022-01-19 19:21
povikoh no
#asahi · 2022-01-19 19:18
povikhttps://github.com/AsahiLinux/linux/commit/ba4afcc99432e9fe141fe19ce4851e63b3a87d9d
#asahi · 2022-01-19 19:18
povikthere's that hack, but still
#asahi · 2022-01-19 18:46
povikseems we have the failure modes pretty documented now
#asahi · 2022-01-19 18:37
povikthere you go
#asahi · 2022-01-19 18:37
povikjoske: snd-soc-apple-mca?
#asahi · 2022-01-19 18:37
povikyou got the gist
#asahi · 2022-01-19 18:37
povik*drivers for platform devices
#asahi · 2022-01-19 18:36
povik:D
#asahi · 2022-01-19 18:36
poviksven: true?
#asahi · 2022-01-19 18:36
poviki assume platform drivers are modprobed on demand autoamtically for platform device drivers
#asahi · 2022-01-19 18:36
povikhm, module name, let's see how do i find out
#asahi · 2022-01-19 18:33
povikjoske: as to your case, i don't know what happens there
#asahi · 2022-01-19 18:32
povikthe mca driver is so far oblivious to power management
#asahi · 2022-01-19 18:31
povikat least now i can reproduce
#asahi · 2022-01-19 18:31
povikshrug, guess i am not doing something right in the driver
#asahi · 2022-01-19 18:30
povikGlanzmann: great, that seems to confirm that pwrstate issue is related to drivers being modules
#asahi · 2022-01-19 18:29
povikjoske: do you have CONFIG_COMMON_CLK_APPLE_NCO?
#asahi · 2022-01-19 18:26
povik:)
#asahi · 2022-01-19 18:22
povikyes, that's how mps had it compiled i think
#asahi · 2022-01-19 18:19
povikjoske: do you maybe see ^ when compiled as module? (if you still have the dmesg)
#asahi · 2022-01-19 18:19
povikapple-pmgr-pwrstate 23b700000.power-management:power-controller@2c0: PS mca1: Failed to reach power state 0xf (now: 0x24f)
#asahi · 2022-01-19 16:57
povikaaaaand... the regcache issue disappeared
#asahi · 2022-01-19 16:50
poviki should probably focus on the regcache issue, which is something i can tinker with, which may have the same root cause as the breakage on Glanzmann's air

plus ›