Hello,
I have been also working on the twl4030 codec driver for some time now. Altrough my patchset is not yet ready, but I'd like to be heard early enough: My plan is to add _all_ (ok, almost all) the features we can get with TWLx030... So instead of having 'only' one possible route from DACl2/DACR2, I have implemented in a reverse order: For all outputs (HSOL/R, PreDriveL/R, HFL/R IHFL/R, EAR) one can select from where the audio is routed. For example: PreDriveL can output audio from Voice, DACL1, DACL2, DACR2 PreDriveR can output audio from Voice, DACR1, DACR2, DACL2 HSOL can output audio from Voice, DACL1, DACL2 HSOR can output audio from Voice, DACR1, DACR2 and the list goes on... With this mode one can enable whatever combination of routing he/she needs from TWL.
Since these channels are really 'independent' inside of TWL, making the routing/mixing flexible (you might say extremely flexible) could make sense.
CCing: Steve Sakoman who actually wrote the driver - could cc Mark Brown as well, but it seams that he is reading alsa-devel quite frequently ;)
What do you think?