On Wed, Mar 30, 2016 at 03:52:39PM +0100, Lee Jones wrote:
> On Wed, 30 Mar 2016, Thor Thayer wrote:

> > The read register has a different address than the write register
> > which is handled in this function with the masks (read address is
> > odd, write address is even).

> Mark, do we have an API which handled such a configuration?

No, but it sounds like this is a regmap with seven bit register values
and a read/write bit which should be using read_flag_mask rather than
trying to treat these as separate registers.