[alsa-devel] [PATCH] ASoC: SOF: Intel: hda: Fix SKL dai count

[Pierre-Louis Bossart]

On 1/14/20 5:18 AM, Cezary Rojewski wrote:
> On 2020-01-13 16:30, Pierre-Louis Bossart wrote:
>> On 1/13/20 5:40 AM, Cezary Rojewski wrote:
>>> With fourth pin added for iDisp for skl_dai, update SOF_SKL_DAI_NUM to
>>> account for the change. Without this, dais from the bottom of the list
>>> are skipped. In current state that's the case for 'Alt Analog CPU DAI'.
>>
>> Can you clarify if you are fixing an issue on TGL - which isn't 
>> supported in the mainline? Or fixing an issue with SKL w/ HDaudio 
>> support?
> 
> This is not strictly connected to TGL but to any platform making use of 
> 'skl_dai[]' within /sof/intel. Encountered this issue when playing with 
> probes - my BEs added at the bottom of the list were omitted during 
> component registration. Once debug flags in asoc were enabled, 
> root-cause presented itself clearly.

ok.

Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart at linux.intel.com>

> 
>> FWIW, I couldn't make HDaudio codec support work on KBL, with the 
>> changes in ASoC core, the HDaudio support is broken with the SKL 
>> driver (or depends on an unreleased topology). See traces at:
>>
>> https://github.com/thesofproject/linux/pull/1667/commits/ae584cf633f232d9cf828bb599a1b620cd0b4fe0 
>>
> 
> We do have working cAVS 1.5 machines (e.g.: SKL, KBL, KBL-R, ABL..) with 
> HDA on our end.
> Will checkout the existing "vanilla skylake" to filter the least amount 
> of changes needed to enable it. However, this can lead to topology 
> binary needed to be updated.

yes, that's the problem I faced, and since I don't have a clue how to 
change the topology on SKL I commented out the hard-coded parts in the 
machine driver.