On Tue, Sep 02, 2008 at 10:02:13AM +0200, Harald Radke wrote:
Hm, a great thing would be to have something like "codec_instanciate()" where u could select a codec device and an array with control names and values for items to be deactivated and get the "modified" codec...but I guess that would be ways to high level for the sound driver structure (;
For the greatest flexibility you really want the activation and deactivation to be doable via user space - many devices (especially complex ones like smartphones) can run with several different audio setups and need to change between them on the fly at run time. Trying to encode all the possible scenarios in the kernel makes for a relatively large amount of work when there are a lot of options.
Liam has been working on an API to facilitate that sort of scenario setup: