>>> Well, please see the updated declaration of snd_hda_ext_bus_init() in
>>> this very patch and then the existing code of
>>> sound/soc/intel/skylake/skl.c - skl_create().
>>> Last argument in updated declaration reads 'modelname'. Skylake-driver
>>> has its own, SOF initializes it differently.
>>
>> Not sure why you have your own?
>>
>
> Not sure I understand the question. If you are talking about changing
> string 'sklbus' to something else, then I don't believe mixing changes:
> update to actual values assigned and assignment relocation in one patch
> is good. I used 'sklbus' as that's what is being currently assigned to
> ->modelname within skl_create(). Such approach makes the change more
> transparent.
What I meant is that this 'modelname' is a module parameter for legacy
and SOF driver, it's attached to the bus, but eventually used by the codec
hda_codec.c: if (codec->bus->modelname) {
hda_codec.c: codec->modelname =
kstrdup(codec->bus->modelname, GFP_KERNEL);
and even further down used to apply board-specific fixups.
"sklbus" doesn't seem to be related to codecs, boards or fixups, so not
sure what this parameter does in the end?
