On 3/29/16 5:50 PM, Bernhard Übelacker wrote:
Hello Pierre-Louis, thank you for your answers and your efforts.
Kind regards, Bernhard
PS.: Could some info from a preinstalled Android of any help?
What would help is the schematics for your board, there are so many variations out there that it's hard to generalize
on Baytrail devices the SSP2 is typically connected to AIF1 and SSP0 to AIF2. On Baytrail-CR devices the SSP0 is connected to AIF1. SSP1 is usually used for BT/FM. You may want to play with the DAI links to see if you get better luck by swapping the codec connections.
Am 29.03.2016 um 02:04 schrieb Pierre-Louis Bossart:
If you modified the interrupt order then it's a Baytrail-CR device and you will need to use the firmware that makes use of SSP0 (SSP2 is not connected in this package). I am working to enable SSP0 routing directly without the need for a new firmware but I haven't had time to test it.
That said, we've not been successful so far with such Baytrail-CR devices. So far the issue seem to be in the SSP0-AIF2 connection; if you do a loopback within the DSP things work fine. Without access to the schematics and the board information it's not straightforward to solve the problem. Note also that Baytrail-CR devices typically use different microphones and speaker settings (differential mics, mono speakers typically) so you may need to add some quirks.
Just for reference if someone gets here via search engine:
I have tried the firmware [1] and put it under /lib/firmware/intel/fw_sst_0f28.bin with the modified interrupt order and linux 4.5.0-rc7+, entered the amixer sequence and the speaker-test.
I saw with that also no obvious errors in dmesg. Interrupts are also counted. The clicks on device open and close are not to hear anymore. But unfortunately it stays silent.
[1] https://git.kernel.org/cgit/linux/kernel/git/vkoul/firmware.git/plain/intel/...