Re: [PATCH 2/2] ALSA: Integrate control based jack reporting with core jack reporting

[Takashi Iwai <tiwai@suse.de>]

At Wed, 22 Feb 2012 17:18:44 +0000,
Mark Brown wrote:
> 
> On Wed, Feb 22, 2012 at 05:52:17PM +0100, Takashi Iwai wrote:
> > Mark Brown wrote:
> 
> > > Now that the kctl jacks are there I'm getting people asking me about it
> > > often enough so I'd like to see it merged.
> 
> > Yeah, if things were easy, I'd be happy to merge.
> > But, judging from the situation, I see no big reason to hurry too
> > much.
> 
> I'm not sure what problems you see here - all the issues that are being
> discussed here are about the kctl interface to applications, there's no
> issues I can see with the in-kernel interfaces.

First of all, we don't agree yet which naming rule to be applied.
Since your patch assumes the case B, i.e. constant names corresponding
only to the key type (HEADPHONE, etc), it's incompatible with the
current implementation in HD-audio.

That is, once when the patch is merged, the kctl expression will be
forcibly to case B but without TLV yet, because kctls will be created
automatically when the jack instance is created.

In other words, if I merge your patch now, the only solution for
HD-audio side for the time being is to disable
CONFIG_SND_HDA_INPUT_JACK.  That's why I hesitate to merge it now.
And, it's why I prefer defining the naming rule at first, thus
refining the implementation not to conflict with the existing one.


Takashi