Re: [alsa-devel] [PATCH] ASoC: cs4270: fix dynamic initialization of register cache

Mark Brown wrote:

Having a driver per device is pretty icky and off the top of my head I'd expect it to cause problems if there are two cs4270 in the system. It would be much nicer and more maintainable to avoid bodging around the API like this.

BTW, why does ASoC even care about the register cache, if I have to have my own functions to read and write it? In other words, why does function cs4270_read_reg_cache() exist at all? What is the point of having ASoC allocate the cache if the driver has to supply functions to read and write it?