On Wed, Feb 08, 2012 at 01:27:30PM +0000, Liam Girdwood wrote:
> On Wed, 2012-02-08 at 11:20 +0000, Mark Brown wrote:

> > I'd generally expect to see new machine drivers doing this stuff in
> > set_bias_level() and possibly also having additional startup code in
> > hw_params() depending on how flexible the clocking is.

> The scope of set_bias_level() is now starting to increase to more than
> just the card bias power. It may be worthwhile at some point to rename
> it to cover all possibilities.

I'm really not sure it's worth bothering going round and changing
everything, and obviously the main candidate name is power which I'd
worry is just going to confuse people even more into thinking they don't
need to bother with DAPM when really it'd be helpful to made DAPM
mandatory.

For many of the things where it's important to have clocks for longer
the clocks are being used as part of the biasing - the main reason we
need clocks more these days is for things like the charge pumps which
are not a million miles off what VMID was on older CODECs.