Hi Chen-Yu,
On Mon, Jan 17, 2022 at 05:01:52PM +0800, Chen-Yu Tsai wrote:
> On Sat, Jan 15, 2022 at 7:03 AM Brian Norris <briannorris@xxxxxxxxxxxx> wrote:
> >
> > Now that the cdn-dp driver supports plug-change callbacks, let's wire it
> > up.
> >
> > Signed-off-by: Brian Norris <briannorris@xxxxxxxxxxxx>
> > ---
> >
> > (no changes since v1)
> >
> > sound/soc/rockchip/rk3399_gru_sound.c | 20 ++++++++++++++++++++
> > 1 file changed, 20 insertions(+)
> >
> > diff --git a/sound/soc/rockchip/rk3399_gru_sound.c b/sound/soc/rockchip/rk3399_gru_sound.c
> > index e2d52d8d0ff9..eeef3ed70037 100644
> > --- a/sound/soc/rockchip/rk3399_gru_sound.c
> > +++ b/sound/soc/rockchip/rk3399_gru_sound.c
> > @@ -164,6 +164,25 @@ static int rockchip_sound_da7219_hw_params(struct snd_pcm_substream *substream,
> > return 0;
> > }
> >
> > +static struct snd_soc_jack cdn_dp_card_jack;
> > +
> > +static int rockchip_sound_cdndp_init(struct snd_soc_pcm_runtime *rtd)
> > +{
> > + struct snd_soc_component *component = asoc_rtd_to_codec(rtd, 0)->component;
>
> Using snd_soc_card_get_codec_dai() might be a better choice throughout this
> driver. While it will work for the cdn_dp case, because it is the first DAI
> in |rockchip_dais[]|, all the invocations for the other codecs are likely
> returning the wrong DAI.
I'll admit, I'm not very familiar with the ASoC object model, so you may
well be correct that there's something fishy in here. But I did trace
through the objects involved here, and we *are* getting the correct DAI
for both this case and the DA7219 case (preexisting code).
It looks like we actually have a new runtime for each of our static
dai_links:
devm_snd_soc_register_card()
...
for_each_card_prelinks()
snd_soc_add_pcm_runtime()
So I think this is valid to keep as-is.
> For this particular patch it works either way, so
>
> Reviewed-by: Chen-Yu Tsai <wenst@xxxxxxxxxxxx>
Thanks for looking!
Brian
