[alsa-devel] [PATCH 1/3] ASoC: TWL4030: Enable voice filters for voice sidetone

[Lopez Cruz, Misael]

> On Tuesday 23 June 2009 13:16:16 ext Mark Brown wrote:
> > On Tue, Jun 23, 2009 at 08:38:20AM +0300, Peter Ujfalusi wrote:
> > > Now if you start an audio playback (which will enable the 
> > > ARXL1, even if we are in Option2), than you stop it: the
> > > ARXL1 (which is VRX in Option2) bit will be turned off...
> > > Now I'm not sure how this will be handled by
> > > the ASoC core.
> > > What I mean is: One path enabled this bit, than another 
> > > path also enables the same bit, than it disables it. Will
> > > the first path notice it? Or will it think that the bit
> > > is still enabled?

> > Sounds like a possible case for a supply widget, though the 
> > fact that you've got both option 1 and option 2 is fun.

> Well I use the fun word for a bit different things...
> I have to check what actually the supply widget supposed to 
> be doing and what is it for.

> > If the driver were a proper platform driver you could select
> > between option 1 and option 2 in platform data (assuming
> > that's sensible) and register different widgets
> > depending on which was in use.

If we register widgets based on an option passed through platform
data, then we won't be able to switch between options, right?
I think switching between option1/2 during runtime is needed,
otherwise 4-channels and voice support will be mutually exclusive.

> I have been looking at this. It does not seamed that 
> complicated. I had written some of the code, but than it
> turned out a bit more complicated:
> In twl4030 series the Vibra interface also reside inside of 
> the CODEC part.
> in short: twl4030 is a MFD, than the TWL4030 codec itself 
> also and MFD..
> So in order to get this in a proper way, this should be 
> handled somehow (another 'fake' MFD for the twl4030 codec/vibra
> perhaps).

> This is still in my list, but certainly need more time to get 
> a reasonable good implementation. 

> -- 
> Péter