[alsa-devel] [PATCH] STAC92xx : add support for missing features

[Maxim Levitsky]

On Tuesday 28 August 2007 16:55:36 Tobin Davis wrote:
> I plan on testing these later this week.  Will let you know what I find
> then.
> 
> I have systems with a 3-stack 9227 configuration, and I might be able to
> get on a 5 stack system as well.
> 
> Tobin
> 
> On Tue, 2007-08-28 at 12:45 +0200, Takashi Iwai wrote:
> 
> > At Tue, 28 Aug 2007 04:54:29 +0300,
> > Maxim Levitsky wrote:
> > > 
> > > Hi,
> > > 
> > > As promised, those are patches that extend patch_sigmatel.c driver to add 
> > > missing functionality
> > 
> > Thanks for the patches!
> > 
> > > They all are runtime tested with my STAC9227, and datasheet tested with others
> > > Note that I did my best to ensure that even if a codec doesn't support a 
> > > feature (and I don't know sbout that), maximum impact will be a new, 
> > > non-working mixer control
> > > 
> > > STAC927x are almost exactly the same as STAC922x, but have DAC4
> > > 
> > > 1) Center/LFE swap :
> > > It can be sometimes useful to swap those channels, since they are on same 
> > > line.
> > > 
> > > I add that control only in case Center/LFE DAC is there, and supports L/R 
> > > swap, it should be very safe.
> > > 
> > > 
> > > 2) Analog loopback:
> > > Added for 922x, 927x, and for 9204/5 which has a different verb, that offers 
> > > same analog loopback
> > > 
> > > 3) Volume Knob:
> > > Added for 9204/5, 9220/1/3, 922x 92-7x
> > > I made it a master volume
> > > 
> > > 4) Fixed support for codecs that have 2 or 3 ADC:
> > > I added separate mixer controls for all 2/3 ADCs, and made PCM capture device 
> > > sub-streams count to depend on actual number of ADCs
> > > 
> > > I am looking for your comments,
> > > 	Best regards,
> > > 		Maxim Levitsky
> > 
> > All patches except 3 look almost good to me.  I'll give review back to
> > each post.  For the next version, could you give the changelog text
> > for each patch (not only the subject line) together with your
> > sign-off?  Then I can merge to HG tree.
I haven't written them :-), I will add changelogs
> > 
> > About 3 (volume-knob), we'd need a bit more checks and tests.  Does
> > the volume knob widget work always as a mater control regardless
> > whetehr the real hardware component exists or not?  I.e. can we use it
> > blindly?
According to datasheet, it works, also driver already assumes it there, and sets it to maximum volume
see hda_verb stac922x_core_init, hda_verb d965_core_init, hda_verb stac927x_core_init,hda_verb stac9205_core_init
> > 
> > Tobin, Matt, could you test these patches, especially patch 3, with
> > your hardwares?  Anyone who has STAC92xx hardwares are welcome for
> > testing it, too!
> > 
> > 
> > Thanks,
> > 
> > Takashi
> 

Hi, 

Thanks for review, I will fix all things you pointed out, but
I am coding against latest linus' kernel tree, and snd_ctl_boolean_mono_info() isn't there, and thus I didn't use it, 
I will download latest alsa tree, and compile against it

But I have a question: can I only complile snd-hda-intel from that tree, and leave kernel's sound core modules?

Best regards,
	Maxim Levitsky