On 3/9/21 4:55 AM, Hans de Goede wrote:
The SST firmware's media and deep-buffer inputs are hardcoded to S16LE, the corresponding DAIs don't have a hw_params callback and their prepare callback also does not take the format into account.
So far the advertising of non working S24LE support has not caused issues because pulseaudio defaults to S16LE, but changing pulse-audio's config to use S24LE will result in broken sound.
Pipewire is replacing pulse now and pipewire prefers S24LE over S16LE when available, causing the problem of the broken S24LE support to come to the surface now.
Cc: stable@vger.kernel.org BugLink: https://gitlab.freedesktop.org/pipewire/pipewire/-/issues/866 Signed-off-by: Hans de Goede hdegoede@redhat.com
Humm, that is strange. I can't recall such limitations in the firmware, and the SSP support does make use of 24 bits. Please give me a couple of days to double-check what's missing.
sound/soc/intel/atom/sst-mfld-platform-pcm.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/sound/soc/intel/atom/sst-mfld-platform-pcm.c b/sound/soc/intel/atom/sst-mfld-platform-pcm.c index 9e9b05883557..aa5dd590ddd5 100644 --- a/sound/soc/intel/atom/sst-mfld-platform-pcm.c +++ b/sound/soc/intel/atom/sst-mfld-platform-pcm.c @@ -488,14 +488,14 @@ static struct snd_soc_dai_driver sst_platform_dai[] = { .channels_min = SST_STEREO, .channels_max = SST_STEREO, .rates = SNDRV_PCM_RATE_44100|SNDRV_PCM_RATE_48000,
.formats = SNDRV_PCM_FMTBIT_S16_LE | SNDRV_PCM_FMTBIT_S24_LE,
.formats = SNDRV_PCM_FMTBIT_S16_LE,
.formats = SNDRV_PCM_FMTBIT_S16_LE | SNDRV_PCM_FMTBIT_S24_LE,
.formats = SNDRV_PCM_FMTBIT_S16_LE,@@ -506,7 +506,7 @@ static struct snd_soc_dai_driver sst_platform_dai[] = { .channels_min = SST_STEREO, .channels_max = SST_STEREO, .rates = SNDRV_PCM_RATE_44100|SNDRV_PCM_RATE_48000,
.formats = SNDRV_PCM_FMTBIT_S16_LE | SNDRV_PCM_FMTBIT_S24_LE,
.formats = SNDRV_PCM_FMTBIT_S16_LE,