[alsa-devel] Generic clock divider indices

[Mark Brown]

On Thu, Jun 06, 2013 at 03:44:04PM +0200, Daniel Mack wrote:

> Would it be an idea to define some commonly used dividers in the core?
> If they start at 0x80000000, they won't collide with existing ones, and
> we could simply add them as alternative case statements to existing drivers.

I don't think this is a terribly sensible idea; as soon as you start
relying on these dividers in machine code you're going to run into
drivers that just don't implement them either due to hardware or due to
them being able to figure things out by themselves and...

> Some commonly used ones I can think of are

> 	MCLK / BCLK
> 	MCLK / LRCLK
> 	BCLK / LRCLK

> That way, platform code could just pass the values dows to the codec
> drivers, whether or not they know what to do with it.

...frankly the above all look like dividers that the drivers should just
be figuring out all by themselves anyway without bothering the machine
driver.  If that's the sort of stuff you're having to do it seems more
generally useful to work on making the relevant drivers require less
manual control by the machine driver.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 836 bytes
Desc: Digital signature
URL: <http://mailman.alsa-project.org/pipermail/alsa-devel/attachments/20130606/95081295/attachment.sig>