Hi,

On Thu, Feb 13, 2020 at 05:34:54PM -0800, Tony Lindgren wrote:
> * Sebastian Reichel <sre@kernel.org> [200214 00:35]:
> > On Wed, Feb 12, 2020 at 06:35:43AM -0800, Tony Lindgren wrote:
> > > Yes this should follow the audio-graph-card.txt example. We end up with
> > > mcbsp3 dts node as below on droid4:
> > > 
> > > &mcbsp3 {
> > >         #sound-dai-cells = <0>;
> > >         pinctrl-names = "default";
> > >         pinctrl-0 = <&mcbsp3_pins>;
> > >         status = "okay";
> > > 
> > >         ports {
> > >                 mcbsp3_port: port@0 {
> > >                         #address-cells = <1>;
> > >                         #size-cells = <0>;
> > > 
> > >                         cpu_dai3: endpoint@0 {
> > 
> > cpu_dai3_cpcap
> > 
> > >                                 reg = <0>;
> > >                                 dai-format = "dsp_a";
> > >                                 frame-master = <&cpcap_audio_codec1>;
> > >                                 bitclock-master = <&cpcap_audio_codec1>;
> > >                                 remote-endpoint = <&cpcap_audio_codec1>;
> > >                         };
> > > 
> > >                         cpu_dai_mdm: endpoint@1 {
> > 
> > cpu_dai3_mdm
> 
> OK
> 
> > >                                 reg = <1>;
> > >                                 dai-format = "dsp_a";
> > >                                 frame-master = <&cpcap_audio_codec1>;
> > >                                 bitclock-master = <&cpcap_audio_codec1>;
> > >                                 remote-endpoint = <&mot_mdm6600_audio_codec0>;
> > >                         };
> > >                 };
> > >         };
> > > };
> > > 
> > > That is pretty much the same as the 'Multi DAI with DPCM' example, with
> > > dne dai, and multiple endpoints. I think we still have just one port
> > > for one i2s transport on the mcbsp :)
> > > 
> > > Does the above look as what you would expect based on the binding?
> > 
> > I haven't had a look at this for quite some time. I suppose the
> > cpcap voice DAI and the modem will also have two endpoints? So
> > once the BT support is added it will looks like this [simplified]?
> 
> Well it will be even simpler, no need for extra endpoints at
> the codecs, see below.
>  
> > &mcbsp3 {
> >     ports {
> >         port@0 {
> >             cpu_dai3_cpcap: endpoint@0 {};
> >             cpu_dai3_modem: endpoint@1 {};
> >             cpu_dai3_bt: endpoint@2 {};
> >         };
> >     };
> > };
> 
> But yes, bluetooth would be just added as above under mcbsp3.
> 
[...]
>
> Then the modem codec looks like this:
> 
> mot_mdm6600_audio: audio-codec {
> 	#address-cells = <1>;
> 	#size-cells = <0>;
> 	#sound-dai-cells = <1>;
> 
> 	port@0 {
> 		mot_mdm6600_audio_codec0: endpoint {
> 			remote-endpoint = <&cpu_dai_mdm>;
> 		};
> 	};
> };
> 
> > &bluetooth {
> >     ports {
> >         port@0 {
> >             bt_dai_cpu: endpoint@0 {};
> >             bt_dai_modem: endpoint@1 {};
> >             bt_dai_cpcap: endpoint@2 {};
> >         };
> >     };
> > };
> 
> And bluetooth would be similar to cpcap_audio and mot_mdm6600_audio
> above.

My understanding is, that CPU is not involved for calls (except for
setting up cpcap registers correctly). Basically McBSP3 should
remain idle for a call and data goes directly from modem to cpcap.
The same should work for modem <-> BT, except that CPCAP seems to
always provide the clock. That would imply a direct link between
modem and codec / BT?

> My guess is that only cpcap registers and clock rate needs to be
> changed for bluetooth audio BTW, so if somebody havs a bluetooth
> headset just do the following in Android:
> 
> # cpcaprw --all > /tmp/before
> configure bluetooth headset for audio in android and start
> playing some music or make a phone call
> ...
> # cpcaprw --all > /tmp/after
> stop playing music or phone call
> ...
> diff -u /tmp/before /tmp/after
> 
> The registers will be different for a bluetooth phone call and
> playing music.

I can provider register values once I find some time.

-- Sebastian