[PATCH v2] ASoC: bdw-rt5650: remove 3-channel capture support

[PATCH v2] ASoC: bdw-rt5650: remove 3-channel capture support

[Brent Lu]

Implement a constrain to exclude 3-channel capture since only 2 and 4
channel capture are supported on the platform.

Signed-off-by: Brent Lu <brent.lu@intel.com>
---
 sound/soc/intel/boards/bdw-rt5650.c | 31 +++++++++++++++++++++++++++++++
 1 file changed, 31 insertions(+)

diff --git a/sound/soc/intel/boards/bdw-rt5650.c b/sound/soc/intel/boards/bdw-rt5650.c
index af2f502..eedbdad 100644
--- a/sound/soc/intel/boards/bdw-rt5650.c
+++ b/sound/soc/intel/boards/bdw-rt5650.c
@@ -83,6 +83,36 @@ static struct snd_soc_jack_pin mic_jack_pin = {
 	.mask	= SND_JACK_MICROPHONE,
 };
 
+static const unsigned int channels[] = {
+	2, 4,
+};
+
+static const struct snd_pcm_hw_constraint_list constraints_channels = {
+	.count = ARRAY_SIZE(channels),
+	.list = channels,
+	.mask = 0,
+};
+
+static int bdw_fe_startup(struct snd_pcm_substream *substream)
+{
+	struct snd_pcm_runtime *runtime = substream->runtime;
+
+	/*
+	 * On this platform for PCM device we support,
+	 * 2 or 4 channel capture
+	 */
+	if (substream->stream == SNDRV_PCM_STREAM_CAPTURE)
+		snd_pcm_hw_constraint_list(runtime, 0,
+					   SNDRV_PCM_HW_PARAM_CHANNELS,
+					   &constraints_channels);
+
+	return 0;
+}
+
+static const struct snd_soc_ops bdw_rt5650_fe_ops = {
+	.startup = bdw_fe_startup,
+};
+
 static int broadwell_ssp0_fixup(struct snd_soc_pcm_runtime *rtd,
 			struct snd_pcm_hw_params *params)
 {
@@ -234,6 +264,7 @@ static struct snd_soc_dai_link bdw_rt5650_dais[] = {
 		.name = "System PCM",
 		.stream_name = "System Playback",
 		.dynamic = 1,
+		.ops = &bdw_rt5650_fe_ops,
 #if !IS_ENABLED(CONFIG_SND_SOC_SOF_BROADWELL)
 		.init = bdw_rt5650_rtd_init,
 #endif
-- 
2.7.4

Re: [PATCH v2] ASoC: bdw-rt5650: remove 3-channel capture support

[Pierre-Louis Bossart]

On 4/13/20 3:28 AM, Brent Lu wrote:
> Implement a constrain to exclude 3-channel capture since only 2 and 4
> channel capture are supported on the platform.

That looks like an error caught by the ALSA conformance tool?

What are the odds that we have a similar issue with the other broadwell 
drivers which don't have a constraint on the number of channels either 
on their 'System PCM' dailink?

Thanks
-Pierre

> 
> Signed-off-by: Brent Lu <brent.lu@intel.com>
> ---
>   sound/soc/intel/boards/bdw-rt5650.c | 31 +++++++++++++++++++++++++++++++
>   1 file changed, 31 insertions(+)
> 
> diff --git a/sound/soc/intel/boards/bdw-rt5650.c b/sound/soc/intel/boards/bdw-rt5650.c
> index af2f502..eedbdad 100644
> --- a/sound/soc/intel/boards/bdw-rt5650.c
> +++ b/sound/soc/intel/boards/bdw-rt5650.c
> @@ -83,6 +83,36 @@ static struct snd_soc_jack_pin mic_jack_pin = {
>   	.mask	= SND_JACK_MICROPHONE,
>   };
>   
> +static const unsigned int channels[] = {
> +	2, 4,
> +};
> +
> +static const struct snd_pcm_hw_constraint_list constraints_channels = {
> +	.count = ARRAY_SIZE(channels),
> +	.list = channels,
> +	.mask = 0,
> +};
> +
> +static int bdw_fe_startup(struct snd_pcm_substream *substream)
> +{
> +	struct snd_pcm_runtime *runtime = substream->runtime;
> +
> +	/*
> +	 * On this platform for PCM device we support,
> +	 * 2 or 4 channel capture
> +	 */
> +	if (substream->stream == SNDRV_PCM_STREAM_CAPTURE)
> +		snd_pcm_hw_constraint_list(runtime, 0,
> +					   SNDRV_PCM_HW_PARAM_CHANNELS,
> +					   &constraints_channels);
> +
> +	return 0;
> +}
> +
> +static const struct snd_soc_ops bdw_rt5650_fe_ops = {
> +	.startup = bdw_fe_startup,
> +};
> +
>   static int broadwell_ssp0_fixup(struct snd_soc_pcm_runtime *rtd,
>   			struct snd_pcm_hw_params *params)
>   {
> @@ -234,6 +264,7 @@ static struct snd_soc_dai_link bdw_rt5650_dais[] = {
>   		.name = "System PCM",
>   		.stream_name = "System Playback",
>   		.dynamic = 1,
> +		.ops = &bdw_rt5650_fe_ops,
>   #if !IS_ENABLED(CONFIG_SND_SOC_SOF_BROADWELL)
>   		.init = bdw_rt5650_rtd_init,
>   #endif
> 

RE: [PATCH v2] ASoC: bdw-rt5650: remove 3-channel capture support

[Lu, Brent]

> 
> That looks like an error caught by the ALSA conformance tool?
> 
> What are the odds that we have a similar issue with the other broadwell
> drivers which don't have a constraint on the number of channels either on
> their 'System PCM' dailink?
> 
> Thanks
> -Pierre
> 

Yes. That's why I am sending patch for this old BDW platform...

So far as I know only 'buddy' supports 2/4-channel recording while other BDW
Chrome products should support stereo recording only. Therefore, this defect
should only be triggered by the ALSA conformance tool.

I am think about implementing the constraint in FE DAI's startup() callback
instead of DAI Link's callback. Since the channels_max is 4 for the capture
stream, ALSA conformance tool will always test 3-channel recording on any
platforms using this driver. Does it make sense to you?

sound/soc/intel/haswell/sst-haswell-pcm.c:
static struct snd_soc_dai_driver hsw_dais[] = {
	{
		.name  = "System Pin",
		.id = HSW_PCM_DAI_ID_SYSTEM,
		.playback = {
			.stream_name = "System Playback",
			.channels_min = 2,
			.channels_max = 2,
			.rates = SNDRV_PCM_RATE_48000,
			.formats = SNDRV_PCM_FMTBIT_S24_LE | SNDRV_PCM_FMTBIT_S16_LE,
		},
		.capture = {
			.stream_name = "Analog Capture",
			.channels_min = 2,
			.channels_max = 4,
			.rates = SNDRV_PCM_RATE_48000,
			.formats = SNDRV_PCM_FMTBIT_S24_LE | SNDRV_PCM_FMTBIT_S16_LE,
		},
	},


Regards,
Brent

Re: [PATCH v2] ASoC: bdw-rt5650: remove 3-channel capture support

[Pierre-Louis Bossart]

On 4/13/20 9:29 AM, Lu, Brent wrote:
>>
>> That looks like an error caught by the ALSA conformance tool?
>>
>> What are the odds that we have a similar issue with the other broadwell
>> drivers which don't have a constraint on the number of channels either on
>> their 'System PCM' dailink?
>>
>> Thanks
>> -Pierre
>>
> 
> Yes. That's why I am sending patch for this old BDW platform...
> 
> So far as I know only 'buddy' supports 2/4-channel recording while other BDW
> Chrome products should support stereo recording only. Therefore, this defect
> should only be triggered by the ALSA conformance tool.
> 
> I am think about implementing the constraint in FE DAI's startup() callback
> instead of DAI Link's callback. Since the channels_max is 4 for the capture
> stream, ALSA conformance tool will always test 3-channel recording on any
> platforms using this driver. Does it make sense to you?

Looking back at previous threads, you indicated that the number of 
channels supported in propagated from BE to FE, so a similar patch to 
add 2ch constraints for bwd-rt5677 was dropped  ("ASoC: bdw-rt5677: 
channel constraint support")

Actually I am not sure it was dropped since later you submitted another 
patch ("ASoC: bdw-rt5677: enable runtime channel merge"), and my 
feedback was that it seemed simpler to add constraints on all machine 
drivers.

And now this patch only addresses bdw-rt5650.c but with the initial 
solution suggested for bdw-rt5677.c

It seems like a generic problem on all Broadwell devices so let's solve 
with one a single patchset.

Shouldn't we just add the 2ch constraints for broadwell.c and 
bdw-rt5677.c, and the 2 or 4ch constraint for bdw-rt5650.c? Would this 
work for you?

Thanks
-Pierre

RE: [PATCH v2] ASoC: bdw-rt5650: remove 3-channel capture support

[Lu, Brent]

> 
> It seems like a generic problem on all Broadwell devices so let's solve with
> one a single patchset.
> 
> Shouldn't we just add the 2ch constraints for broadwell.c and bdw-rt5677.c,
> and the 2 or 4ch constraint for bdw-rt5650.c? Would this work for you?
> 
> Thanks
> -Pierre

Hi Pierre,

Are you saying submitting a new patch to add constraints to all three broadwell.c,
bdw-rt5650.c, and bdw-rt5677.c?


Regards,
Brent

Re: [PATCH v2] ASoC: bdw-rt5650: remove 3-channel capture support

[Cezary Rojewski]

On 2020-04-17 03:32, Lu, Brent wrote:
>>
>> It seems like a generic problem on all Broadwell devices so let's solve with
>> one a single patchset.
>>
>> Shouldn't we just add the 2ch constraints for broadwell.c and bdw-rt5677.c,
>> and the 2 or 4ch constraint for bdw-rt5650.c? Would this work for you?
>>
>> Thanks
>> -Pierre
> 
> Hi Pierre,
> 
> Are you saying submitting a new patch to add constraints to all three broadwell.c,
> bdw-rt5650.c, and bdw-rt5677.c?
> 
> 
> Regards,
> Brent
> 

What Pierre suggested is that you submit a series of patches instead - 
one for each of BDW machine boards. If the same problem exists on all of 
them, there is no reason to left other boards unattended.

Czarek

Re: [PATCH v2] ASoC: bdw-rt5650: remove 3-channel capture support

[Pierre-Louis Bossart]

On 4/17/20 2:59 AM, Cezary Rojewski wrote:
> On 2020-04-17 03:32, Lu, Brent wrote:
>>>
>>> It seems like a generic problem on all Broadwell devices so let's 
>>> solve with
>>> one a single patchset.
>>>
>>> Shouldn't we just add the 2ch constraints for broadwell.c and 
>>> bdw-rt5677.c,
>>> and the 2 or 4ch constraint for bdw-rt5650.c? Would this work for you?
>>>
>>> Thanks
>>> -Pierre
>>
>> Hi Pierre,
>>
>> Are you saying submitting a new patch to add constraints to all three 
>> broadwell.c,
>> bdw-rt5650.c, and bdw-rt5677.c?
>>
>>
>> Regards,
>> Brent
>>
> 
> What Pierre suggested is that you submit a series of patches instead - 
> one for each of BDW machine boards. If the same problem exists on all of 
> them, there is no reason to left other boards unattended.

Yes, a series of 3 patches with the same code pattern, the only 
variation being 4 ch allowed for bdw-rt5650.c in addition to the default 
2ch for all 3 boards.

Thanks
-Pierre