[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <c9462148-2131-4660-9f65-b53f22979387@linaro.org>
Date: Tue, 7 May 2024 20:15:49 +0200
From: Krzysztof Kozlowski <krzysztof.kozlowski@...aro.org>
To: Srinivas Kandagatla <srinivas.kandagatla@...aro.org>,
James Schulman <james.schulman@...rus.com>,
David Rhodes <david.rhodes@...rus.com>,
Richard Fitzgerald <rf@...nsource.cirrus.com>,
Jaroslav Kysela <perex@...ex.cz>, Takashi Iwai <tiwai@...e.com>,
Liam Girdwood <lgirdwood@...il.com>, Mark Brown <broonie@...nel.org>,
Lars-Peter Clausen <lars@...afoo.de>, Nuno Sá
<nuno.sa@...log.com>, Banajit Goswami <bgoswami@...cinc.com>
Cc: alsa-devel@...a-project.org, patches@...nsource.cirrus.com,
linux-sound@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH 4/4] ASoC: qcom: x1e80100: Correct channel mapping
On 07/05/2024 15:20, Srinivas Kandagatla wrote:
> Thanks Krzystof for the patch.
>
> On 07/05/2024 11:27, Krzysztof Kozlowski wrote:
>> X1E80100 CRD board comes with four speakers arranged as left front+back
>> and then right front+back. Using default channel mapping causes front
>> right speaker to play left back stream.
>>
>> Adjust the channel maps for frontend DAIs to fix stereo and four-channel
>> playback.
>>
>> Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@...aro.org>
>> ---
>> sound/soc/qcom/x1e80100.c | 37 +++++++++++++++++++++++++++++++++++--
>> 1 file changed, 35 insertions(+), 2 deletions(-)
>>
>> diff --git a/sound/soc/qcom/x1e80100.c b/sound/soc/qcom/x1e80100.c
>> index c3c8bf7ffb5b..e90c68815b5c 100644
>> --- a/sound/soc/qcom/x1e80100.c
>> +++ b/sound/soc/qcom/x1e80100.c
>> @@ -12,6 +12,7 @@
>>
>> #include "common.h"
>> #include "qdsp6/q6afe.h"
>> +#include "qdsp6/q6dsp-common.h"
>> #include "sdw.h"
>>
>> struct x1e80100_snd_data {
>> @@ -74,7 +75,7 @@ static int x1e80100_snd_hw_params(struct snd_pcm_substream *substream,
>> return qcom_snd_sdw_hw_params(substream, params, &data->sruntime[cpu_dai->id]);
>> }
>>
>> -static int x1e80100_snd_prepare(struct snd_pcm_substream *substream)
>> +static int x1e80100_snd_be_prepare(struct snd_pcm_substream *substream)
>> {
>> struct snd_soc_pcm_runtime *rtd = substream->private_data;
>> struct snd_soc_dai *cpu_dai = snd_soc_rtd_to_cpu(rtd, 0);
>> @@ -96,12 +97,34 @@ static int x1e80100_snd_hw_free(struct snd_pcm_substream *substream)
>> &data->stream_prepared[cpu_dai->id]);
>> }
>>
>> +static int x1e80100_snd_fe_prepare(struct snd_pcm_substream *substream)
>> +{
>> + struct snd_soc_pcm_runtime *rtd = snd_soc_substream_to_rtd(substream);
>> + struct snd_soc_dai *cpu_dai = snd_soc_rtd_to_cpu(rtd, 0);
>> +
>> + if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK) {
>> + const unsigned int rx_slot[4] = { PCM_CHANNEL_FL,
>> + PCM_CHANNEL_LB,
>> + PCM_CHANNEL_FR,
>> + PCM_CHANNEL_RB };
>> +
>> + snd_soc_dai_set_channel_map(cpu_dai, 0, NULL, ARRAY_SIZE(rx_slot),
>> + rx_slot);
>
> Channel mapping are specific to backend dais rather than front end pcm dais.
>
> This will set all the playback pcms with this channel maps, which is a
> problem.
>
> example the 2 channel headset we will endup with data of front channel
> and zeros on the right channel, however a speaker might work as you have
> 4 speakers in your system.
>
>
> So No for this approach.
OK, I'll go with setting channels for MFC (and expecting MFC being part
of backend).
Best regards,
Krzysztof
Powered by blists - more mailing lists