[alsa-devel] [PATCH 1/2] ASoC: soc-compress: add a config item for soc-compress

[Jie, Yang]

> -----Original Message-----
> From: Koul, Vinod
> Sent: Tuesday, June 16, 2015 4:14 PM
> To: Jie, Yang
> Cc: Mark Brown; ramesh.babu at linux.intel.com; alsa-devel at alsa-project.org;
> tiwai at suse.de; Zhang, Vivian; vinod.koul at linux.intel.com; Girdwood, Liam R
> Subject: Re: [alsa-devel] [PATCH 1/2] ASoC: soc-compress: add a config item
> for soc-compress
> 
> On Fri, Jun 12, 2015 at 01:33:25PM +0000, Jie, Yang wrote:
> > > -----Original Message-----
> > > From: Mark Brown [mailto:broonie at kernel.org]
> > > Sent: Friday, June 12, 2015 9:00 PM
> > > To: Jie, Yang
> > > Cc: tiwai at suse.de; alsa-devel at alsa-project.org; Girdwood, Liam R;
> > > vinod.koul at linux.intel.com; ramesh.babu at linux.intel.com; Zhang,
> > > Vivian
> > > Subject: Re: [PATCH 1/2] ASoC: soc-compress: add a config item for
> > > soc- compress
> Okay didnt realise that its posted to alsa :( I need to remove all ref to @linux
> id as am not using that anymore
 
Seems that's a big patch for replacing. :)

> 
> > >
> > > On Fri, Jun 12, 2015 at 12:31:25PM +0000, Jie, Yang wrote:
> > >
> > > > > here to make the drivers that do use compressed support select
> > > > > this symbol
> > >
> > > > As mentioned in commit message, we need add 'select
> > > SND_SOC_COMPRESS'
> > > > to driver  Kconfig(e.g. sound/soc/intel/Kconfig) when it is needed.
> > > >
> > > > > and since there's no config text there's no way for users to
> > > > > enable this on a per system basis so this breaks all existing
> > > > > compressed audio
> > > support.
> > > >
> > > > Do you mean I should add this message to 'Help' for this config item?
> > > > Per my understanding, no existing driver really uses this
> > > > compressed audio support currently, is that true?
> > >
> > > Some of the existing Intel machine drivers have references to
> > > compressed audio, it certainly looks like they're trying to use it.
> > > Are you saying that this is not actually the case?
> >
> > Let me double confirm if the atom drivers are using it. Seems mfld
> > platform is using that.
> Yes and we will get runtime errors on all mfld-atom based drivers so please
> add that on all atom machines
> 
> Since the dummy symbols are present so we wont get build errors. If a driver
> is using the compressed API maybe we should not add dummy so that we
> catch ones which are not including this config option, so for that point I would
> say removal of dummay handlers looks best to me.
 
Hi, Vinod, as I mentioned in replying to v2 patch in another thread,  here
cpu_dai->driver->compress_dai is a runtime value, which means we don't know
if we need compress API(for our driver) or not at compiling stage. 
So, reporting error at compiling stage looks not easy to be implemented.

At the same time, we return -EPERM to return and tell caller that compress 
operation is not permitted.

Anyway, let's discuss further on v2 patches thread.

Thanks,
~Keyon

> 
> --
> ~Vinod