[alsa-devel] [PATCH 4/4] ASoC: cs35l35: Add multi-device synchronisation

[Charles Keepax]

On Tue, May 30, 2017 at 09:51:38AM +0100, Charles Keepax wrote:
> On Mon, May 29, 2017 at 02:53:06PM +0100, Mark Brown wrote:
> > On Thu, May 18, 2017 at 04:32:39PM +0100, Charles Keepax wrote:
> > 
> > > +	SOC_SINGLE_EXT("SYNC Audio", CS35L35_MULT_DEV_SYNCH2, 1, 1, 0,
> > > +		       snd_soc_get_volsw, cs35l35_put_sync),
> > > +	SOC_SINGLE_EXT("SYNC VPBR", CS35L35_MULT_DEV_SYNCH2, 2, 1, 0,
> > > +		       snd_soc_get_volsw, cs35l35_put_sync),
> > > +	SOC_SINGLE_EXT("SYNC OTW", CS35L35_MULT_DEV_SYNCH2, 3, 1, 0,
> > > +		       snd_soc_get_volsw, cs35l35_put_sync),
> > 
> > I can't tell how this works.  It feels like this shouldn't just be being
> > controlled from userspace but rather should be handled in some more
> > standard fashion, or possibly as part of the platform integration but
> > right now it's just some totally undocumented application managed
> > controls.
> 
> These activate the individual types of synchronisation between
> the two stereo amps over a proprietary single wire connection
> between the two amps. The audio option synchronises the group
> delay between the two amps. The VPBR links the brown out on the
> two amps and the OTW links the over temperature warning.
> 
> The issue is really one of it being use-case specific whether the
> amps are being used independently or in a stereo configuration.
> You could have only a single amp in use in which case the sync
> features are best turned off. Or one might even have use-cases
> with both amps where they are being used independently but at the
> same time.
> 
> I guess we could potentially do those as calls from the machine
> driver, although in some cases it might be hard to tell which
> use-case is being used. Alternatively, one could try to actually
> link the two amps in DAPM and control it that way although you
> probably still want some device tree stuff to say which things
> you want to sync and it might cause issues in any cases where you
> had both amps up but didn't want to sync.

Ok turns out there is one other corner case issue here, lets drop
this patch for now and I will try to see if I can come up with
better solution for it.

Thanks,
Charles