On Thu, Aug 28, 2008 at 08:42:52PM +0200, Manuel Lauss wrote:
Attached below. Although I'm no longer convinced it's a good idea after all; I don't see how multiple asoc devices with this codec could be realized (which afaik is planned with asoc2).
You can't have multiple ASoC devices. You also can't have more than one codec within the one ASoC device that you can instantiate at the moment. Loading the codec driver normally can be made to work (the PowerPC guys are doing it at the minute) but it's fragile since it doesn't play with the assumptions the ASoC code is currently making.
I'd be more comfortable with leaving the I2C driver registration done where it is currently until ASoC at least supports dynamic registration of codec drivers explicitly purely on a risk avoidance basis.