Re: [PATCH v9 3/4] ASoC: tda998x: add a codec to the HDMI transmitter

From: Jean-Francois Moine
Date: Tue Jan 13 2015 - 02:39:26 EST


On Sun, 11 Jan 2015 23:03:14 +0200
Jyri Sarha <jsarha@xxxxxx> wrote:

> Some more comments based on my - finally successful - attempt to use
> this code with BBB HDMI audio.
>
> On 01/07/2015 12:51 PM, Jean-Francois Moine wrote:
> > This patch adds a CODEC to the NXP TDA998x HDMI transmitter.
[snip]
> > @@ -47,6 +52,10 @@ struct tda998x_priv {
> > struct drm_encoder *encoder;
> >
> > u8 audio_ports[2];
> > +#ifdef WITH_CODEC
> > + u8 sad[3]; /* Short Audio Descriptor */
> > + struct snd_pcm_hw_constraint_list rate_constraints;
> > +#endif
>
> It feels a bit backwards to me to store these audio side variables here
> in DRM side driver - especially the rate_constraint - and provide a
> function (tda998x_get_audio_var()) for getting their individual
> addresses to ASoC side. Wouldn't it be more straight forward to provide
> functions for setting and getting the audio side data from a void
> pointer in DRM side device drvdata?

Good idea. The rate_constraints would be allocated by the codec and set by a transmitter function.

[snip]
> > diff --git a/include/sound/tda998x.h b/include/sound/tda998x.h
> > new file mode 100644
> > index 0000000..97cff30
> > --- /dev/null
> > +++ b/include/sound/tda998x.h
> > @@ -0,0 +1,23 @@
> > +#ifndef SND_TDA998X_H
> > +#define SND_TDA998X_H
> > +
> > +#include <sound/pcm.h>
> > +
> > +/* port indexes */
> > +#define PORT_NONE (-1)
> > +#define PORT_SPDIF 0
> > +#define PORT_I2S 1
> > +
> > +typedef int (*get_audio_var_t)(struct device *dev,
> > + u8 **sad,
> > + struct snd_pcm_hw_constraint_list **rate_constraints);
> > +typedef int (*set_audio_input_t)(struct device *dev,
> > + int port_index,
> > + unsigned sample_rate,
> > + int sample_format);
> > +
>
> Why don't you simply declare tda998x_get_audio_var() and
> tda998x_set_audio_input() here and export them in DRM side driver? This
> is a tda998x specific codec after all. I see no point to this this
> function pointer typedef game, when the pointers are anyway stored to
> global variables in ASoC the side module.

There cannot be both ways of function call with modules:
the transmitter may call exported functions of the codec,
but the codec cannot call exported functions of the transmitter.

[snip]
> > + snd_pcm_hw_constraint_minmax(runtime,
> > + SNDRV_PCM_HW_PARAM_CHANNELS,
> > + 1, sad[SAD_MX_CHAN_I]);
>
> SAD byte 1 bits 2-0 provides the maximum channel count _minus_one_. You
> should _add_one_ to sad[SAD_MX_CHAN_I] when setting the constraint.

Right. Thanks.

[snip]
> > +static struct snd_soc_dai_driver tda998x_dais[] = {
> > + {
> > + .name = "spdif-hifi",
> > + .id = PORT_SPDIF,
> > + .playback = {
> > + .stream_name = "HDMI SPDIF Playback",
> > + .channels_min = 1,
> > + .channels_max = 2,
> > + .rates = SNDRV_PCM_RATE_CONTINUOUS,
> > + .rate_min = 22050,
> > + .rate_max = 192000,
> > + .formats = TDA998X_FORMATS,
> > + },
> > + .ops = &tda998x_codec_ops,
> > + },
> > + {
> > + .name = "i2s-hifi",
> > + .id = PORT_I2S,
> > + .playback = {
> > + .stream_name = "HDMI I2S Playback",
> > + .channels_min = 1,
> > + .channels_max = 8,
> > + .rates = SNDRV_PCM_RATE_CONTINUOUS,
> > + .rate_min = 5512,
> > + .rate_max = 192000,
> > + .formats = TDA998X_FORMATS,
> > + },
> > + .ops = &tda998x_codec_ops,
> > + },
> > +};
>
> Why do you use SNDRV_PCM_RATE_CONTINUOUS, when HDMI standard only
> supports 32000, 44100, 48000, 88200, 96000, 176400, and 192000 Hz-rates?

Before I treated the EDID, I directly used these definitions,
and my TV display accepted many more rates as 22050 with S/PDIF input
or 7875 with I2S input.

> > +
> > +static const struct snd_soc_dapm_widget tda998x_widgets[] = {
> > + SND_SOC_DAPM_OUTPUT("hdmi-out"),
> > +};
> > +static const struct snd_soc_dapm_route tda998x_routes[] = {
> > + { "hdmi-out", NULL, "HDMI I2S Playback" },
> > + { "hdmi-out", NULL, "HDMI SPDIF Playback" },
> > +};
> > +
> > +static struct snd_soc_codec_driver tda998x_codec_drv = {
> > + .dapm_widgets = tda998x_widgets,
> > + .num_dapm_widgets = ARRAY_SIZE(tda998x_widgets),
> > + .dapm_routes = tda998x_routes,
> > + .num_dapm_routes = ARRAY_SIZE(tda998x_routes),
> > + .ignore_pmdown_time = true,
> > +};
> > +
> > +int tda9998x_codec_register(struct device *dev,
> > + get_audio_var_t get_audio_var_i,
> > + set_audio_input_t set_audio_input_i)
> > +{
> > + get_audio_var = get_audio_var_i;
> > + set_audio_input = set_audio_input_i;
> > + return snd_soc_register_codec(dev,
> > + &tda998x_codec_drv,
> > + tda998x_dais, ARRAY_SIZE(tda998x_dais));
>
> So this always registers a two DAI codec whether or not both the i2s and
> spdif is used. The DT binding document could state that the DAI index 0
> is always the spdif DAI and index 1 is the i2s DAI, or even better you
> could #define them under include/dt-bindings/.
>
> While you are at it, you could also mention the DAPM output name
> "hdmi-out" for the codec in the DT-binding document.
[snip]

This will be changed when using the graph of ports: the DAIs will be
dynamically created from the DT.

--
Ken ar c'hentaà | ** Breizh ha Linux atav! **
Jef | http://moinejf.free.fr/
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/