On Mon, 24 Nov 2008 14:33:04 +0000 "ext Mark Brown" <broonie@xxxxxxxxxxxxx> wrote: > On Mon, Nov 24, 2008 at 03:45:33PM +0200, Jarkko Nikula wrote: > > My problem was this kind of switching correct to do in machine driver? > > > cpu_dai->playback.channels_min = 1; > > cpu_dai->playback.channels_max = 1; > > cpu_dai->capture.channels_min = 1; > > cpu_dai->capture.channels_max = 1; > > No, these should be constant. Any dynamic constraints should be applied > with the constraints API. > Oh my, now feeling like not done my homework. Thanks Mark and sorry about the line noise :-) Will send soon patchset adding snd_pcm_hw_constraint_minmax(..., SNDRV_PCM_HW_PARAM_CHANNELS to existing OMAP machine drivers + updated version of this patch. > So, this is a bit messy at the moment since we don't really have any > DAPM-style routing on the digital side yet. What most of these codecs > should probably be doing is declaring a 2 channel minimuim - normally > that's what they require as a wire format even if they only actually use > the data from one of those channels. > I think it's not worth to change existing codecs now. Keeping 1 channel minimum on codec side allow fascinating hackish DAI configurations like using only one channel from stereo codec or to operate with some intelligent CPU DAI able to do mono->stereo and stereo->mono mixing internally. Yet another line noise from me... Jarkko _______________________________________________ Alsa-devel mailing list Alsa-devel@xxxxxxxxxxxxxxxx http://mailman.alsa-project.org/mailman/listinfo/alsa-devel
