[RFC] ASoC: SOF: use 32bit valid sample size for I2S gateway #4016
Conversation
FW always generate 32bit valid sample size to dma so align it with driver. Signed-off-by: Rander Wang <[email protected]>
RanderWang
requested review from
lgirdwood,
plbossart,
ranj063,
kv2019i,
dbaluta,
bardliao,
ujfalusi,
ajitkupandey,
kuanhsuncheng and
yaochunhung
as code owners
|
@juimonen please help to review it. Thanks! |
| @@ -1251,7 +1251,7 @@ sof_ipc4_prepare_copier_module(struct snd_sof_widget *swidget, | |||
| ipc4_copier = (struct sof_ipc4_copier *)dai->private; | |||
| copier_data = &ipc4_copier->data; | |||
| available_fmt = &ipc4_copier->available_fmt; | |||
| if (dir == SNDRV_PCM_STREAM_CAPTURE) { | |||
| if (dir == SNDRV_PCM_STREAM_CAPTURE || ipc4_copier->dai_type == SOF_DAI_INTEL_SSP) { | |||
There was a problem hiding this comment.
Commit message: can we use the same term we are using for the components?
ASoC: SOF: ipc4-topology: Use 32bit sample size for copier on Intel SSP
s/FW/The firmware
There was a problem hiding this comment.
You should update the comment down at line 1259 as well.
There was a problem hiding this comment.
@RanderWang the change looks good but do you have an issue that this PR fixes?
There was a problem hiding this comment.
and BTW can this be fixed in the nocodec tplg instead?
There was a problem hiding this comment.
@RanderWang the change looks good but do you have an issue that this PR fixes?
sorry missed the issue in the description
There was a problem hiding this comment.
@ranj063 yong changed nocodec tplg but made no effect thesofproject/sof#6595 (comment).
| @@ -1251,7 +1251,7 @@ sof_ipc4_prepare_copier_module(struct snd_sof_widget *swidget, | |||
| ipc4_copier = (struct sof_ipc4_copier *)dai->private; | |||
| copier_data = &ipc4_copier->data; | |||
| available_fmt = &ipc4_copier->available_fmt; | |||
| if (dir == SNDRV_PCM_STREAM_CAPTURE) { | |||
| if (dir == SNDRV_PCM_STREAM_CAPTURE || ipc4_copier->dai_type == SOF_DAI_INTEL_SSP) { | |||
There was a problem hiding this comment.
I am not getting the explanations in the commit message @RanderWang
Is this really the case that ONLY the SSP gateway assumes 32-bit valid data on playback?
There was a problem hiding this comment.
And all of @ujfalusi 's comments are valid.
There was a problem hiding this comment.
@plbossart It is true for closed source FW. It depends on ssp blob in case of SOF, so I ask @jsarha for help
There was a problem hiding this comment.
Humm, that doesn't sound good at all. This smells of an undocumented dependency between firmware implementation and NHLT blobs. Meh.
There was a problem hiding this comment.
yes, different gateway has different requirement and sof should be compatible with closed source fw.
RanderWang
changed the title
|
Let's do it first in topology. I found something incorrect in yong's test topology |
|
@RanderWang do we still need this PR or is this no longer needed? |
|
NO need now. Close it |
FW always generate 32bit valid sample size to dma so align it with driver.
Fix: thesofproject/sof#6595