Re: Re: [PATCH v4 11/23] ASoC: simple-card: Loop over all children for 'mclk-fs'

[Sameer Pujar]

On 6/30/2020 4:25 PM, Mark Brown wrote:
> On Tue, Jun 30, 2020 at 09:53:13AM +0530, Sameer Pujar wrote:
> > On 6/30/2020 7:38 AM, Kuninori Morimoto wrote:
> > > External email: Use caution opening links or attachments
> > > > > > +     if (cpu != codec)
> > > > > > +             of_property_read_u32(codec, prop, &props->mclk_fs);
> > Sorry if I was not clear before.
> I agree with Moromito-san that the new code (especially the above bit)
> is very confusing, I'm not sure how the reader is supposed to figure out
> what the purpose of the check is or how the CPU could ever be the CODEC.
>
> > simple_parse_mclk_fs() is used by both simple_dai_link_of_dpcm() and
> > simple_dai_link_of(). To make the same API work for both the cases, I had to
> > use (A) in DPCM function. Now (B) does not get used for
> > simple_dai_link_of_dpcm(), but will get used by simple_dai_link_of().
> > If it is simpler I will just avoid 'cpu != codec' check and keep the
> > function simple_parse_mclk_fs() as it is.
> That'd definitely be simpler, or supporting this with a CPU node as
> well.

OK. I will keep it simple.