#
✕ nick:povik
#asahi-dev · 2022-01-30 22:42
poviki guess the gpio line is untouched before some driver calls gpiod_get_optional
#asahi-dev · 2022-01-30 22:40
povikif the codec isn't in reset to begin with
#asahi-dev · 2022-01-30 22:40
povikyeah, i wonder if it's possible that lines 2087 to 2096 generate a narrow reset pulse
#asahi-dev · 2022-01-30 22:39
poviklet's see the driver code again if that's possible
#asahi-dev · 2022-01-30 22:39
povikwhich means the part is in a weird state
#asahi-dev · 2022-01-30 22:39
povikand then the driver pulls the reset for too short
#asahi-dev · 2022-01-30 22:39
povikthe only idea i had is that the cs42l wasn't in reset when linux booted because of the state of the respective gpio
#asahi-dev · 2022-01-30 22:36
povikyeah post that dmesg please
#asahi-dev · 2022-01-30 22:31
poviknot that i don't want to fix the issue, even if it scarcely manifests
#asahi-dev · 2022-01-30 22:31
povikweird with the speakers. could we have been so unlucky on the first try?
#asahi-dev · 2022-01-30 22:31
povikthanks :)
#asahi-dev · 2022-01-30 22:21
poviki tuned the mmio power-domain sweep a bit and saved it on wiki: https://github.com/AsahiLinux/docs/wiki/HW:Memory-map#mmio-ranges-that-respond-to-probe-reads-t8103
#asahi-dev · 2022-01-29 23:52
povikbut it's there... :)
#asahi-dev · 2022-01-29 23:52
povikpoint was to find AOP's MCA, because i didn't see it in DART nor DAPF
#asahi-dev · 2022-01-29 23:51
povikhttps://tpaste.us/vgYZ
#asahi-dev · 2022-01-29 23:51
poviki had bit of fun finding which mmio ranges come alive as power domains get enabled one-by-one in a topological order
#asahi-dev · 2022-01-29 16:12
povikanywhere where there's a cluster SEL it is the cluster no. plus one i think
#asahi-dev · 2022-01-29 16:11
povikbut it should essentially be the same as in sound/soc/apple/mca.c, module the RX serdes registers
#asahi-dev · 2022-01-29 16:10
povikkettenis: i am battling git, here it is in paste instead: https://tpaste.us/qP6B
#asahi-dev · 2022-01-29 16:01
povikon it
#asahi-dev · 2022-01-29 15:55
poviki plan to push it together with demo script for capture, once that's ready, but if it would be useful to you i can push it now
#asahi-dev · 2022-01-29 15:55
povikkettenis: i have m1n1.hw.mca with register definitions
#asahi-dev · 2022-01-29 15:52
poviki am now experimenting with audio capture, there's two masks too, here it may be obvious what they do
#asahi-dev · 2022-01-29 15:52
povikso in my local tree the MCA driver clears the second mask
#asahi-dev · 2022-01-29 15:51
povikand switching the speaker to be left/right based on the TDM slots it's sensitive to
#asahi-dev · 2022-01-29 15:51
poviki ran into the same when simulating macbook air's speakers on mini
#asahi-dev · 2022-01-29 15:49
povikthat's what happened, right?
#asahi-dev · 2022-01-29 15:49
povikso to get stereo on mca1, you had to clear a bit in the second mask that is otherwise left set by iboot
#asahi-dev · 2022-01-29 15:48
povikor rather recently i learned that it behaves as if the actual mask was AND of the two masks
#asahi-dev · 2022-01-29 15:48
povikkettenis: i know about the mask but i am not sure about its meaning
#asahi-dev · 2022-01-29 09:54
povik(wanted to look at it so i looked it up)
#asahi-dev · 2022-01-29 09:54
povikhere maybe: https://lore.kernel.org/linux-iommu/20211019163737.46269-1-sven@svenpeter.dev/
#asahi-dev · 2022-01-28 21:09
povikimprobable there will be a surprise but worth checking
#asahi-dev · 2022-01-28 21:09
povikthat's the cs42l reset gpio
#asahi-dev · 2022-01-28 21:09
povikalyssa: when you have the chance please try p.read32(0x23d1f002c) after fresh boot and send me the value
#asahi-dev · 2022-01-28 20:28
povikanyway will switch to usleep_range then
#asahi-dev · 2022-01-28 20:27
poviki wonder what that unused TAS2770_MDELAY was there for in the driver
#asahi-dev · 2022-01-28 19:34
poviki assume it's guaranteed to be at least 1 ms
#asahi-dev · 2022-01-28 19:30
povikhttps://tpaste.us/0wM8
#asahi-dev · 2022-01-28 19:11
povikdoesn't help us with the cs42l42 probe unfortunatelly
#asahi-dev · 2022-01-28 19:10
povikyeah, that's not happening here
#asahi-dev · 2022-01-28 19:09
povikfrom tas2770 datasheet
#asahi-dev · 2022-01-28 19:09
povik"Once all supplies are stable the SDZ pin can be set high to initialize the part. After a hardware or software reset additional commands to the device should be delayed for 1 mS to allow the OTP to load."
#asahi-dev · 2022-01-28 19:04
povikwho wants to bet it's this write
#asahi-dev · 2022-01-28 19:04
povikhttps://github.com/torvalds/linux/blob/871bfa02d08d9c0ed981c50082b7afd367d3700b/sound/soc/codecs/tas2770.c#L43
#asahi-dev · 2022-01-28 18:54
povikfigured
#asahi-dev · 2022-01-28 18:53
poviki wanted to chase that the other day but didn't get any
#asahi-dev · 2022-01-28 18:53
povik> [ 1.379196] tas2770 1-0031: ASoC: error at soc_component_write_no_lock on tas2770.1-0031 for register: [0x00000001] -6
#asahi-dev · 2022-01-28 18:53
poviklook who showed up
#asahi-dev · 2022-01-28 18:53
povikanyway

plus ›