#asahi-dev · 2021-12-14 14:32
marcanthe ID register properly says CS42L84 for this one, for one#asahi-dev · 2021-12-14 14:31
marcanyeah#asahi-dev · 2021-12-14 14:30
marcanand there's stuff I can't find a match for in '42, though it does seem apple may be making up register names#asahi-dev · 2021-12-14 14:30
marcansome register field names do match 42l42, so it's not like a total from-scratch chip, but if nothing else the addressing is different for one#asahi-dev · 2021-12-14 14:26
marcan84 is very different from 83#asahi-dev · 2021-12-14 14:26
marcanpovik: yeah, this doesn't look like anything I can find, and stringsing the macos driver points in that direction too#asahi-dev · 2021-12-14 14:14
marcanyeah I get the feeling this is one of those "apple got first dibs on a new chip or got them to make something custom" things#asahi-dev · 2021-12-14 13:59
marcanlet me see if I can find a codec they do with 16-bit addrs...#asahi-dev · 2021-12-14 13:59
marcancs42l92 uses *32-bit* addresses so that is precedent for cirrus doing this #asahi-dev · 2021-12-14 13:58
marcanyeah, I know#asahi-dev · 2021-12-14 13:58
marcancs42l83 instead seems to use 16-bits of address outright#asahi-dev · 2021-12-14 13:58
marcancs42l42 (aka cs42l83, apparently) uses a weird banking thing with 8 bits of page address and 8 bits of register address#asahi-dev · 2021-12-14 13:57
marcanyes#asahi-dev · 2021-12-14 13:54
marcanyeah, like 16-bit addresses, not the banking thing...#asahi-dev · 2021-12-14 13:52
marcanthis almost looks like the i2c addressing mechanism is different from what I expect#asahi-dev · 2021-12-14 13:48
marcanmaybe it's time for gpiola...#asahi-dev · 2021-12-14 13:48
marcansomething is really messed up with the i2c interface#asahi-dev · 2021-12-14 13:48
marcanthis looks like the device ID, but it's in the wrong place#asahi-dev · 2021-12-14 13:48
marcan70: 42 a8 4a 01 00 00 00 00 00 00 00 00 00 00 00 00 B?J?............#asahi-dev · 2021-12-14 13:41
marcanah wait, this is something weird with i2c#asahi-dev · 2021-12-14 13:31
marcanit's 42 forever, even subsequent reads of 00#asahi-dev · 2021-12-14 13:31
marcanand then it sticks there#asahi-dev · 2021-12-14 13:31
marcanso the first register read returns 00, the rest return 42#asahi-dev · 2021-12-14 13:31
marcanlol.#asahi-dev · 2021-12-14 13:31
marcan00: 00 42 42 42 42 42 42 42 42 42 42 42 42 42 42 42 .BBBBBBBBBBBBBBB#asahi-dev · 2021-12-14 13:31
marcan10: 42 42 42 42 42 42 42 42 42 42 42 42 42 42 42 42 BBBBBBBBBBBBBBBB#asahi-dev · 2021-12-14 13:31
marcan 0 1 2 3 4 5 6 7 8 9 a b c d e f 0123456789abcdef#asahi-dev · 2021-12-14 13:31
marcanContinue? [Y/n] #asahi-dev · 2021-12-14 13:26
marcanevery read the driver did returned 0 except the 2nd one#asahi-dev · 2021-12-14 13:25
marcanI feel like this codec is pretty dead#asahi-dev · 2021-12-14 13:24
marcanis that expected with the i2s clock issue?#asahi-dev · 2021-12-14 13:24
marcanfwiw alsamixer fails even before I try to play any audio#asahi-dev · 2021-12-14 13:23
marcan[ 19.795376] apple-mca 39b600000.mca: unsupported SERDES configuration requested (mask=0x3 slots=2 slot_width=64)#asahi-dev · 2021-12-14 13:17
marcans/route1/route/, that was just a test I just did#asahi-dev · 2021-12-14 13:16
marcanhttps://mrcn.st/p/JUN0Xxpf#asahi-dev · 2021-12-14 13:16
marcanlet me paste you my diff#asahi-dev · 2021-12-14 13:16
marcanyes#asahi-dev · 2021-12-14 13:16
marcan(using sox's play)#asahi-dev · 2021-12-14 13:15
marcanplayback does appear to progress at the right rate#asahi-dev · 2021-12-14 13:12
marcanIIRC that was preconfigured by the bootloader and we don't touch those regs in the end?#asahi-dev · 2021-12-14 13:11
marcanhow is the mca clock -> nco mapping defined?#asahi-dev · 2021-12-14 13:10
marcanalsamixer is still broken though :-)#asahi-dev · 2021-12-14 13:09
marcanah, ok#asahi-dev · 2021-12-14 13:09
marcanor rather the 2nd input#asahi-dev · 2021-12-14 13:09
marcannco2 has a different ref though#asahi-dev · 2021-12-14 13:08
marcanthat's for nco0 and 1#asahi-dev · 2021-12-14 13:08
marcannco clock ref is different.#asahi-dev · 2021-12-14 13:08
marcanteehee.#asahi-dev · 2021-12-14 13:08
marcan User: /device-tree/arm-io/nco.clk[0]#asahi-dev · 2021-12-14 13:08
marcan# 84: 1068000000 0 0xa4/0x28e038028: regval: 0x85100100