Re: [PATCH RFC V2 0/5] another generic audio hdmi codec proposal

[Arnaud Pouliquen <arnaud.pouliquen@st.com>]

On 10/06/2015 04:33 PM, Jean-Francois Moine wrote:
> On Tue, 6 Oct 2015 11:23:03 +0200
> Arnaud Pouliquen <arnaud.pouliquen@st.com> wrote:
> 	[snip]
>> As API is defined in DRM, it seems more logical to match it with the one
>> defined for video. From my windows, i didn't see any blocking point to
>> connect codec callback with this API.
>> But anyway, this API is not freezed, it could be improved with your help.
>
> Arnaud, your implementation seems more heavy than Jyri's, and I don't
> feel the DRM bridge.
>
> For HDMI, the exchanges between DRM and ALSA are only:
> - DRM -> ALSA
>    - device connection with the audio constraints
>    - device disconnection
> - ALSA -> DRM
>    - start audio with the chosen audio parameters
>    - stop audio
>
> and, in the system, the HDMI devices are seen as DRM connectors (video
> view) and as ASoC CODECs (audio view).
>
> We just need a link between these entities.
> I don't think that the bridge offers this.
Not so heavy...
In fact another way to differentiate the twice approaches could be to
answer to this question:
Should ALSA bypasses or not the DRM core/API to access to DRM HDMI 
connector?

Concerning bridge usage, I tried to found a link to address DRM driver 
trough DRM API. Bridge seemed a good candidate. Perhaps the DRM encoder 
API could be an other option to study...

Ken ar wech all/ best Regards
Arnaud