[alsa-devel] [PATCH] ALSA: hda: add Intel DSP configuration / probe code

[Pierre-Louis Bossart]

>>> +#if IS_ENABLED(CONFIG_SND_SOC_SOF_COMETLAKE_LP)
>>> +	0x02c8,	/* Cometlake-LP */
>>> +#endif
>>> +#if IS_ENABLED(CONFIG_SND_SOC_SOF_COMETLAKE_H)
>>> +	0x06c8,	/* Cometlake-H */
>>> +#endif
>>> +#if IS_ENABLED(CONFIG_SND_SOC_SOF_GEMINILAKE)
>>> +	0x3198,	/* Geminilake */
>>> +#endif
>>> +#if IS_ENABLED(CONFIG_SND_SOC_SOF_APOLLOLAKE)
>>> +	0x5a98,	/* Broxton-P (Appololake) */
>>> +#endif
>>> +#if IS_ENABLED(CONFIG_SND_SOC_SOF_CANNONLAKE)
>>> +	0x9dc8, /* Cannonlake */
>>> +#endif
>>> +#if IS_ENABLED(CONFIG_SND_SOC_SOF_COFFEELAKE)
>>> +	0xa348, /* Coffelake */
>>> +#endif
>>
>> What about all the other PCI IDs that SOF handles, e.g. TigerLake, etc?
> 
> There is no PCI ID clash, only one driver will call the DSP probe and
> SND_INTEL_DSP_DRIVER_ANY will be returned in this case.

If we are talking about conflicts between the HDaudio legacy driver and 
SOF, I think all PCI IDs should be there. Not every OEM, or even all 
skews from a single OEM, will transition to DMICs and DSP enabled, and 
the legacy driver will still be required for many years.

> 
>> Also how do you deal with SKL/KBL cases with DMICs? They would need to
>> use the SST driver (for Chromebooks mainly)
> 
> As I noted in the comment, we can add dmi quirks on top. I just do not have
> enough information - can I take the hints from the pull request (your code)
> you already mentioned?

Sure. They should cover the main conflicts.

>> Even for APL, the 'official' driver is still SST for Chromebooks. SOF
>> should work but there will probably be missing firmware/topology files.
> 
> I can rework this part of course. I'll send v2 patch.

I see it now, will comment on the v2.