[Sound-open-firmware] [PATCH] topology: AIF directions are codec centric wrt PCMs
Commit 0a1a9bbc492e4eb22173ed57c7f5499c8f106289 discovered a bug in the direction of AIF widgets in relation to PCMs. AIF widgets are codec centric in ASoC so topology must align.
Signed-off-by: Liam Girdwood liam.r.girdwood@linux.intel.com --- topology/m4/local.m4 | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/topology/m4/local.m4 b/topology/m4/local.m4 index 08425ed..c0ffef2 100644 --- a/topology/m4/local.m4 +++ b/topology/m4/local.m4 @@ -96,7 +96,7 @@ define(`W_PCM_PLAYBACK', `}' `SectionWidget."'N_PCMP`" {' ` index "'PIPELINE_ID`"' -` type "aif_out"' +` type "aif_in"' ` no_pm "true"' ` stream_name "'$1`"' ` data [' @@ -131,7 +131,7 @@ define(`W_PCM_CAPTURE', `}' `SectionWidget."'N_PCMC`" {' ` index "'PIPELINE_ID`"' -` type "aif_in"' +` type "aif_out"' ` no_pm "true"' ` stream_name "'$1`"' ` data ['
On 2017年12月14日 05:10, Liam Girdwood wrote:
Commit 0a1a9bbc492e4eb22173ed57c7f5499c8f106289 discovered a bug in the direction of AIF widgets in relation to PCMs. AIF widgets are codec centric in ASoC so topology must align.
Signed-off-by: Liam Girdwood liam.r.girdwood@linux.intel.com
topology/m4/local.m4 | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/topology/m4/local.m4 b/topology/m4/local.m4 index 08425ed..c0ffef2 100644 --- a/topology/m4/local.m4 +++ b/topology/m4/local.m4 @@ -96,7 +96,7 @@ define(`W_PCM_PLAYBACK', `}' `SectionWidget."'N_PCMP`" {' ` index "'PIPELINE_ID`"' -` type "aif_out"' +` type "aif_in"'
This is opposite to what the driver mapping based on our test.
` no_pm "true"' ` stream_name "'$1`"' ` data [' @@ -131,7 +131,7 @@ define(`W_PCM_CAPTURE', `}' `SectionWidget."'N_PCMC`" {' ` index "'PIPELINE_ID`"' -` type "aif_in"' +` type "aif_out"'
ditto.
Thanks, ~Keyon
` no_pm "true"' ` stream_name "'$1`"' ` data ['
On Thu, 2017-12-14 at 09:26 +0800, Keyon Jie wrote:
On 2017年12月14日 05:10, Liam Girdwood wrote:
Commit 0a1a9bbc492e4eb22173ed57c7f5499c8f106289 discovered a bug in the direction of AIF widgets in relation to PCMs. AIF widgets are codec centric in ASoC so topology must align.
Signed-off-by: Liam Girdwood liam.r.girdwood@linux.intel.com
topology/m4/local.m4 | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/topology/m4/local.m4 b/topology/m4/local.m4 index 08425ed..c0ffef2 100644 --- a/topology/m4/local.m4 +++ b/topology/m4/local.m4 @@ -96,7 +96,7 @@ define(`W_PCM_PLAYBACK', `}' `SectionWidget."'N_PCMP`" {' ` index "'PIPELINE_ID`"' -` type "aif_out"' +` type "aif_in"'
This is opposite to what the driver mapping based on our test.
Are you sure ? Most (all ?) of the ASoC/Topology logic is codec centric (i.e. directions are with respect to the codec). Without this patch playback/capture (using std topology) would use the incorrect IPC configuration data.
i.e. Playback host DMA would use the capture hopst DMA config (with sink periods == 0, but source_periods == 2)
Liam
` no_pm "true"' ` stream_name "'$1`"' ` data [' @@ -131,7 +131,7 @@ define(`W_PCM_CAPTURE', `}' `SectionWidget."'N_PCMC`" {' ` index "'PIPELINE_ID`"' -` type "aif_in"' +` type "aif_out"'
ditto.
Thanks, ~Keyon
` no_pm "true"' ` stream_name "'$1`"' ` data ['
Sound-open-firmware mailing list Sound-open-firmware@alsa-project.org http://mailman.alsa-project.org/mailman/listinfo/sound-open-firmware
On 2017年12月14日 16:18, Liam Girdwood wrote:
On Thu, 2017-12-14 at 09:26 +0800, Keyon Jie wrote:
On 2017年12月14日 05:10, Liam Girdwood wrote:
Commit 0a1a9bbc492e4eb22173ed57c7f5499c8f106289 discovered a bug in the direction of AIF widgets in relation to PCMs. AIF widgets are codec centric in ASoC so topology must align.
Signed-off-by: Liam Girdwood liam.r.girdwood@linux.intel.com
topology/m4/local.m4 | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/topology/m4/local.m4 b/topology/m4/local.m4 index 08425ed..c0ffef2 100644 --- a/topology/m4/local.m4 +++ b/topology/m4/local.m4 @@ -96,7 +96,7 @@ define(`W_PCM_PLAYBACK', `}' `SectionWidget."'N_PCMP`" {' ` index "'PIPELINE_ID`"' -` type "aif_out"' +` type "aif_in"'
This is opposite to what the driver mapping based on our test.
Are you sure ? Most (all ?) of the ASoC/Topology logic is codec centric (i.e. directions are with respect to the codec). Without this patch playback/capture (using std topology) would use the incorrect IPC configuration data.
i.e. Playback host DMA would use the capture hopst DMA config (with sink periods == 0, but source_periods == 2)
Agree with this naming.
Maybe we could keep this and change the wrong mapping in other place.
Thanks, ~Keyon
Liam
` no_pm "true"' ` stream_name "'$1`"' ` data [' @@ -131,7 +131,7 @@ define(`W_PCM_CAPTURE', `}' `SectionWidget."'N_PCMC`" {' ` index "'PIPELINE_ID`"' -` type "aif_in"' +` type "aif_out"'
ditto.
Thanks, ~Keyon
` no_pm "true"' ` stream_name "'$1`"' ` data ['
Sound-open-firmware mailing list Sound-open-firmware@alsa-project.org http://mailman.alsa-project.org/mailman/listinfo/sound-open-firmware
Sound-open-firmware mailing list Sound-open-firmware@alsa-project.org http://mailman.alsa-project.org/mailman/listinfo/sound-open-firmware
participants (2)
-
Keyon Jie -
Liam Girdwood