[Sound-open-firmware] [RFC PATCH 0/2] GLK/APL topology edits
Posting this to the mailing list so that its archived and searchable. Basically I am trying to modify the GLK topology to support the ApolloLake Chromebooks. It's only a matter of swapping two SSPs but the changes in the topology files are far from trivial. Feedback welcome.
Pierre-Louis Bossart (2): topology: simplify sof-glk-codec.m4 topology: add sof-apl-da7219 support
topology/Makefile.am | 1 + topology/sof-apl-da7219.m4 | 104 +++++++++++++++++++++++++++++++++++++++++++++ topology/sof-glk-codec.m4 | 30 ++++++------- 3 files changed, 120 insertions(+), 15 deletions(-) create mode 100644 topology/sof-apl-da7219.m4
No need to use PIPELINE_DAI_ADD if DAI_ADD is already used. Also make sure PCM devices start at zero on playback
Signed-off-by: Pierre-Louis Bossart pierre-louis.bossart@linux.intel.com --- topology/sof-glk-codec.m4 | 30 +++++++++++++++--------------- 1 file changed, 15 insertions(+), 15 deletions(-)
diff --git a/topology/sof-glk-codec.m4 b/topology/sof-glk-codec.m4 index d8fe305..98056d5 100644 --- a/topology/sof-glk-codec.m4 +++ b/topology/sof-glk-codec.m4 @@ -22,27 +22,27 @@ include(`platform/intel/dmic.m4') # Define the pipelines # # PCM0 <---- volume <----- DMIC0 (dmic capture) -# PCM1 ----> volume -----> SSP1 (speaker - maxim98357a) -# PCM2 ----> volume -----> SSP2 (headset - da7219) +# PCM0 ----> volume -----> SSP1 (speaker - maxim98357a) +# PCM1 ----> volume -----> SSP2 (headset - da7219) #
# Low Latency capture pipeline 1 on PCM 0 using max 4 channels of s32le. # Schedule 48 frames per 1000us deadline on core 0 with priority 0 -PIPELINE_PCM_DAI_ADD(sof/pipe-volume-capture.m4, +PIPELINE_PCM_ADD(sof/pipe-volume-capture.m4, 1, 0, 4, s32le, - 48, 1000, 0, 0, DMIC, 0, s32le, 2) + 48, 1000, 0, 0)
-# Low Latency playback pipeline 2 on PCM 1 using max 2 channels of s16le. +# Low Latency playback pipeline 2 on PCM 0 using max 2 channels of s16le. # Schedule 48 frames per 1000us deadline on core 0 with priority 0 -PIPELINE_PCM_DAI_ADD(sof/pipe-volume-playback.m4, - 2, 1, 2, s16le, - 48, 1000, 0, 0, SSP, 1, s16le, 2) +PIPELINE_PCM_ADD(sof/pipe-volume-playback.m4, + 2, 0, 2, s16le, + 48, 1000, 0, 0)
-# Low Latency playback pipeline 3 on PCM 2 using max 2 channels of s16le. +# Low Latency playback pipeline 3 on PCM 1 using max 2 channels of s16le. # Schedule 48 frames per 1000us deadline on core 0 with priority 0 -PIPELINE_PCM_DAI_ADD(sof/pipe-volume-playback.m4, - 3, 2, 2, s16le, - 48, 1000, 0, 0, SSP, 2, s16le, 2) +PIPELINE_PCM_ADD(sof/pipe-volume-playback.m4, + 3, 1, 2, s16le, + 48, 1000, 0, 0)
# # DAIs configuration @@ -69,10 +69,10 @@ DAI_ADD(sof/pipe-dai-playback.m4, PIPELINE_SOURCE_3, 2, s16le, 48, 1000, 0, 0)
-# PCM +# PCM (PCM_CAPTURE_ADD(name, pipeline, pcm_id, dai_id, capture)) PCM_CAPTURE_ADD(DMIC0, 1, 0, 0, PIPELINE_PCM_1) -PCM_PLAYBACK_ADD(SSP1, 2, 1, 1, PIPELINE_PCM_2) -PCM_PLAYBACK_ADD(SSP2, 3, 2, 2, PIPELINE_PCM_3) +PCM_PLAYBACK_ADD(SSP1, 2, 0, 1, PIPELINE_PCM_2) +PCM_PLAYBACK_ADD(SSP2, 3, 1, 2, PIPELINE_PCM_3)
# # BE configurations - overrides config in ACPI if present
Add based on sof-glk-codec. In theory this is the same topology with only a swap of SSPs
FIXME: this is a mess between pipeline ID, PCM ID, DAI ID, DAI #. I could use additional eyes to proof-read this...
FIXME: not sure why we would need PCM_CAPTURE_ADD/PCM_PLAYBACK_DAI, this seems to be already done by DAI_ADD?
Signed-off-by: Pierre-Louis Bossart pierre-louis.bossart@linux.intel.com --- topology/Makefile.am | 1 + topology/sof-apl-da7219.m4 | 104 +++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 105 insertions(+) create mode 100644 topology/sof-apl-da7219.m4
diff --git a/topology/Makefile.am b/topology/Makefile.am index 33d0418..396b5b9 100644 --- a/topology/Makefile.am +++ b/topology/Makefile.am @@ -31,6 +31,7 @@ MACHINES = \ sof-hsw-rt5640.tplg \ sof-apl-tdf8532.tplg \ sof-apl-pcm512x.tplg \ + sof-apl-da7219.tplg \ sof-glk-codec.tplg
# Uncomment the following line if you want to debug conf files diff --git a/topology/sof-apl-da7219.m4 b/topology/sof-apl-da7219.m4 new file mode 100644 index 0000000..3934df5 --- /dev/null +++ b/topology/sof-apl-da7219.m4 @@ -0,0 +1,104 @@ +# +# Topology for AppoloLake with headset on SSP1, spk on SSP5 and DMIC capture +# + +# Include topology builder +include(`utils.m4') +include(`dai.m4') +include(`pipeline.m4') +include(`ssp.m4') + +# Include TLV library +include(`common/tlv.m4') + +# Include Token library +include(`sof/tokens.m4') + +# Include bxt DSP configuration +include(`platform/intel/bxt.m4') +include(`platform/intel/dmic.m4') + +# +# Define the pipelines +# +# PCM0 <---- volume <----- DMIC0 (dmic capture) +# PCM0 ----> volume -----> SSP5 (speaker - maxim98357a) +# PCM1 ----> volume -----> SSP1 (headset - da7219) +# + +# Low Latency capture pipeline 1 on PCM 0 using max 4 channels of s32le. +# Schedule 48 frames per 1000us deadline on core 0 with priority 0 +PIPELINE_PCM_ADD(sof/pipe-volume-capture.m4, + 1, 0, 4, s32le, + 48, 1000, 0, 0) + +# Low Latency playback pipeline 2 on PCM 0 using max 2 channels of s16le. +# Schedule 48 frames per 1000us deadline on core 0 with priority 0 +PIPELINE_PCM_ADD(sof/pipe-volume-playback.m4, + 2, 0, 2, s16le, + 48, 1000, 0, 0) + +# Low Latency playback pipeline 3 on PCM 1 using max 2 channels of s16le. +# Schedule 48 frames per 1000us deadline on core 0 with priority 0 +PIPELINE_PCM_ADD(sof/pipe-volume-playback.m4, + 3, 1, 2, s16le, + 48, 1000, 0, 0) + +# +# DAIs configuration +# + +# capture DAI is DMIC0 using 2 periods +# Buffers use s16le format, with 48 frame per 1000us on core 0 with priority 0 +DAI_ADD(sof/pipe-dai-capture.m4, + 1, DMIC, 0, DMIC0, + PIPELINE_SINK_1, 2, s32le, + 48, 1000, 0, 0) + +# playback DAI is SSP5 using 2 periods +# Buffers use s16le format, with 48 frame per 1000us on core 0 with priority 0 +DAI_ADD(sof/pipe-dai-playback.m4, + 2, SSP, 5, SSP5-Codec, + PIPELINE_SOURCE_2, 2, s16le, + 48, 1000, 0, 0) + +# playback DAI is SSP1 using 2 periods +# Buffers use s16le format, with 48 frame per 1000us on core 0 with priority 0 +DAI_ADD(sof/pipe-dai-playback.m4, + 3, SSP, 1, SSP1-Codec, + PIPELINE_SOURCE_3, 2, s16le, + 48, 1000, 0, 0) + +# FIXME: Why is this needed? And is this correct? +# PCM (PCM_CAPTURE_ADD(name, pipeline, pcm_id, dai_id, capture)) +#PCM_CAPTURE_ADD(DMIC0, 1, 0, 0, PIPELINE_PCM_1) +#PCM_PLAYBACK_ADD(SSP5, 2, 0, 5, PIPELINE_PCM_2) +#PCM_PLAYBACK_ADD(SSP1, 3, 1, 1, PIPELINE_PCM_3) + +# +# BE configurations - overrides config in ACPI if present +# +# FIXME: the machine driver uses .name = "dmic01" and .id=2 +# DMIC0 (id: 2) +DAI_CONFIG(DMIC, 0, 2, dmic01, + DMIC_CONFIG(1, 500000, 4800000, 40, 60, 48000, + DMIC_WORD_LENGTH(s32le), DMIC, 0, + PDM_CONFIG(DMIC, 0, FOUR_CH_PDM0_PDM1))) + +# FIXME: the machine driver uses .name = "SSP5-Codec", .id=0, .cpu_name="SSP5 Pin" +#SSP 5 (ID: 0) with 19.2MHz mclk with MCLK_ID 0 +DAI_CONFIG(SSP, 5, 0, SSP5-Codec, + SSP_CONFIG(I2S, SSP_CLOCK(mclk, 19200000, codec_mclk_in), + SSP_CLOCK(bclk, 1920000, codec_slave), + SSP_CLOCK(fsync, 48000, codec_slave), + SSP_TDM(2, 20, 3, 3), + SSP_CONFIG_DATA(SSP, 5, 16, 0))) + +# FIXME: the machine driver uses .name = "SSP1-Codec", .id=1, .cpu_name="SSP1 Pin" +#SSP 1 (ID: 1) with 19.2MHz mclk with MCLK_ID 0 +DAI_CONFIG(SSP, 1, 1, SSP1-Codec, + SSP_CONFIG(I2S, SSP_CLOCK(mclk, 19200000, codec_mclk_in), + SSP_CLOCK(bclk, 1920000, codec_slave), + SSP_CLOCK(fsync, 48000, codec_slave), + SSP_TDM(2, 20, 3, 3), + SSP_CONFIG_DATA(SSP, 1, 16, 0)))
On Tue, 2018-07-10 at 18:15 -0500, Pierre-Louis Bossart wrote:
Add based on sof-glk-codec. In theory this is the same topology with only a swap of SSPs
FIXME: this is a mess between pipeline ID, PCM ID, DAI ID, DAI #. I could use additional eyes to proof-read this...
Looks good to me. I couldnt spot any mistakes. Keyon, Could you please also check?
FIXME: not sure why we would need PCM_CAPTURE_ADD/PCM_PLAYBACK_DAI, this seems to be already done by DAI_ADD?
DAI_ADD only adds the BE DAI's. PCM_CAPTURE_ADD/PLAYBACK_ADD are needed for adding the FE dai links.
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel .com>
topology/Makefile.am | 1 + topology/sof-apl-da7219.m4 | 104 +++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 105 insertions(+) create mode 100644 topology/sof-apl-da7219.m4
diff --git a/topology/Makefile.am b/topology/Makefile.am index 33d0418..396b5b9 100644 --- a/topology/Makefile.am +++ b/topology/Makefile.am @@ -31,6 +31,7 @@ MACHINES = \ sof-hsw-rt5640.tplg \ sof-apl-tdf8532.tplg \ sof-apl-pcm512x.tplg \
- sof-apl-da7219.tplg \ sof-glk-codec.tplg
# Uncomment the following line if you want to debug conf files diff --git a/topology/sof-apl-da7219.m4 b/topology/sof-apl-da7219.m4 new file mode 100644 index 0000000..3934df5 --- /dev/null +++ b/topology/sof-apl-da7219.m4 @@ -0,0 +1,104 @@ +# +# Topology for AppoloLake with headset on SSP1, spk on SSP5 and DMIC capture +#
+# Include topology builder +include(`utils.m4') +include(`dai.m4') +include(`pipeline.m4') +include(`ssp.m4')
+# Include TLV library +include(`common/tlv.m4')
+# Include Token library +include(`sof/tokens.m4')
+# Include bxt DSP configuration +include(`platform/intel/bxt.m4') +include(`platform/intel/dmic.m4')
+# +# Define the pipelines +# +# PCM0 <---- volume <----- DMIC0 (dmic capture) +# PCM0 ----> volume -----> SSP5 (speaker - maxim98357a) +# PCM1 ----> volume -----> SSP1 (headset - da7219) +#
+# Low Latency capture pipeline 1 on PCM 0 using max 4 channels of s32le. +# Schedule 48 frames per 1000us deadline on core 0 with priority 0 +PIPELINE_PCM_ADD(sof/pipe-volume-capture.m4,
- 1, 0, 4, s32le,
- 48, 1000, 0, 0)
+# Low Latency playback pipeline 2 on PCM 0 using max 2 channels of s16le. +# Schedule 48 frames per 1000us deadline on core 0 with priority 0 +PIPELINE_PCM_ADD(sof/pipe-volume-playback.m4,
- 2, 0, 2, s16le,
- 48, 1000, 0, 0)
+# Low Latency playback pipeline 3 on PCM 1 using max 2 channels of s16le. +# Schedule 48 frames per 1000us deadline on core 0 with priority 0 +PIPELINE_PCM_ADD(sof/pipe-volume-playback.m4,
- 3, 1, 2, s16le,
- 48, 1000, 0, 0)
+# +# DAIs configuration +#
+# capture DAI is DMIC0 using 2 periods +# Buffers use s16le format, with 48 frame per 1000us on core 0 with priority 0 +DAI_ADD(sof/pipe-dai-capture.m4,
- 1, DMIC, 0, DMIC0,
- PIPELINE_SINK_1, 2, s32le,
- 48, 1000, 0, 0)
+# playback DAI is SSP5 using 2 periods +# Buffers use s16le format, with 48 frame per 1000us on core 0 with priority 0 +DAI_ADD(sof/pipe-dai-playback.m4,
- 2, SSP, 5, SSP5-Codec,
- PIPELINE_SOURCE_2, 2, s16le,
- 48, 1000, 0, 0)
+# playback DAI is SSP1 using 2 periods +# Buffers use s16le format, with 48 frame per 1000us on core 0 with priority 0 +DAI_ADD(sof/pipe-dai-playback.m4,
- 3, SSP, 1, SSP1-Codec,
- PIPELINE_SOURCE_3, 2, s16le,
- 48, 1000, 0, 0)
+# FIXME: Why is this needed? And is this correct? +# PCM (PCM_CAPTURE_ADD(name, pipeline, pcm_id, dai_id, capture)) +#PCM_CAPTURE_ADD(DMIC0, 1, 0, 0, PIPELINE_PCM_1) +#PCM_PLAYBACK_ADD(SSP5, 2, 0, 5, PIPELINE_PCM_2) +#PCM_PLAYBACK_ADD(SSP1, 3, 1, 1, PIPELINE_PCM_3)
+# +# BE configurations - overrides config in ACPI if present +# +# FIXME: the machine driver uses .name = "dmic01" and .id=2 +# DMIC0 (id: 2) +DAI_CONFIG(DMIC, 0, 2, dmic01,
- DMIC_CONFIG(1, 500000, 4800000, 40, 60, 48000,
DMIC_WORD_LENGTH(s32le), DMIC, 0,PDM_CONFIG(DMIC, 0, FOUR_CH_PDM0_PDM1)))+# FIXME: the machine driver uses .name = "SSP5-Codec", .id=0, .cpu_name="SSP5 Pin" +#SSP 5 (ID: 0) with 19.2MHz mclk with MCLK_ID 0 +DAI_CONFIG(SSP, 5, 0, SSP5-Codec,
- SSP_CONFIG(I2S, SSP_CLOCK(mclk, 19200000, codec_mclk_in),
SSP_CLOCK(bclk, 1920000, codec_slave),SSP_CLOCK(fsync, 48000, codec_slave),SSP_TDM(2, 20, 3, 3),SSP_CONFIG_DATA(SSP, 5, 16, 0)))+# FIXME: the machine driver uses .name = "SSP1-Codec", .id=1, .cpu_name="SSP1 Pin" +#SSP 1 (ID: 1) with 19.2MHz mclk with MCLK_ID 0 +DAI_CONFIG(SSP, 1, 1, SSP1-Codec,
- SSP_CONFIG(I2S, SSP_CLOCK(mclk, 19200000, codec_mclk_in),
SSP_CLOCK(bclk, 1920000, codec_slave),SSP_CLOCK(fsync, 48000, codec_slave),SSP_TDM(2, 20, 3, 3),SSP_CONFIG_DATA(SSP, 1, 16, 0)))
On 07/11/2018 01:23 AM, Ranjani Sridharan wrote:
On Tue, 2018-07-10 at 18:15 -0500, Pierre-Louis Bossart wrote:
Add based on sof-glk-codec. In theory this is the same topology with only a swap of SSPs
FIXME: this is a mess between pipeline ID, PCM ID, DAI ID, DAI #. I could use additional eyes to proof-read this...
Looks good to me. I couldnt spot any mistakes. Keyon, Could you please also check?
FIXME: not sure why we would need PCM_CAPTURE_ADD/PCM_PLAYBACK_DAI, this seems to be already done by DAI_ADD?
DAI_ADD only adds the BE DAI's. PCM_CAPTURE_ADD/PLAYBACK_ADD are needed for adding the FE dai links.
ok, but then why do those macros make references to the DAIs?
+# PCM (PCM_CAPTURE_ADD(name, pipeline, pcm_id, dai_id, capture)) +#PCM_CAPTURE_ADD(DMIC0, 1, 0, 0, PIPELINE_PCM_1) +#PCM_PLAYBACK_ADD(SSP5, 2, 0, 5, PIPELINE_PCM_2) +#PCM_PLAYBACK_ADD(SSP1, 3, 1, 1, PIPELINE_PCM_3)
In addition the information on the PCM # is already provided in PIPELINE_ADD, so this looks redundant on both the FE and BE sides?
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel .com>
topology/Makefile.am | 1 + topology/sof-apl-da7219.m4 | 104 +++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 105 insertions(+) create mode 100644 topology/sof-apl-da7219.m4
diff --git a/topology/Makefile.am b/topology/Makefile.am index 33d0418..396b5b9 100644 --- a/topology/Makefile.am +++ b/topology/Makefile.am @@ -31,6 +31,7 @@ MACHINES = \ sof-hsw-rt5640.tplg \ sof-apl-tdf8532.tplg \ sof-apl-pcm512x.tplg \
sof-apl-da7219.tplg \ sof-glk-codec.tplg
# Uncomment the following line if you want to debug conf files
diff --git a/topology/sof-apl-da7219.m4 b/topology/sof-apl-da7219.m4 new file mode 100644 index 0000000..3934df5 --- /dev/null +++ b/topology/sof-apl-da7219.m4 @@ -0,0 +1,104 @@ +# +# Topology for AppoloLake with headset on SSP1, spk on SSP5 and DMIC+# PCM (PCM_CAPTURE_ADD(name, pipeline, pcm_id, dai_id, capture)) +#PCM_CAPTURE_ADD(DMIC0, 1, 0, 0, PIPELINE_PCM_1) +#PCM_PLAYBACK_ADD(SSP5, 2, 0, 5, PIPELINE_PCM_2) +#PCM_PLAYBACK_ADD(SSP1, 3, 1, 1, PIPELINE_PCM_3) capture +#
+# Include topology builder +include(`utils.m4') +include(`dai.m4') +include(`pipeline.m4') +include(`ssp.m4')
+# Include TLV library +include(`common/tlv.m4')
+# Include Token library +include(`sof/tokens.m4')
+# Include bxt DSP configuration +include(`platform/intel/bxt.m4') +include(`platform/intel/dmic.m4')
+# +# Define the pipelines +# +# PCM0 <---- volume <----- DMIC0 (dmic capture) +# PCM0 ----> volume -----> SSP5 (speaker - maxim98357a) +# PCM1 ----> volume -----> SSP1 (headset - da7219) +#
+# Low Latency capture pipeline 1 on PCM 0 using max 4 channels of s32le. +# Schedule 48 frames per 1000us deadline on core 0 with priority 0 +PIPELINE_PCM_ADD(sof/pipe-volume-capture.m4,
- 1, 0, 4, s32le,
- 48, 1000, 0, 0)
+# Low Latency playback pipeline 2 on PCM 0 using max 2 channels of s16le. +# Schedule 48 frames per 1000us deadline on core 0 with priority 0 +PIPELINE_PCM_ADD(sof/pipe-volume-playback.m4,
- 2, 0, 2, s16le,
- 48, 1000, 0, 0)
+# Low Latency playback pipeline 3 on PCM 1 using max 2 channels of s16le. +# Schedule 48 frames per 1000us deadline on core 0 with priority 0 +PIPELINE_PCM_ADD(sof/pipe-volume-playback.m4,
- 3, 1, 2, s16le,
- 48, 1000, 0, 0)
+# +# DAIs configuration +#
+# capture DAI is DMIC0 using 2 periods +# Buffers use s16le format, with 48 frame per 1000us on core 0 with priority 0 +DAI_ADD(sof/pipe-dai-capture.m4,
- 1, DMIC, 0, DMIC0,
- PIPELINE_SINK_1, 2, s32le,
- 48, 1000, 0, 0)
+# playback DAI is SSP5 using 2 periods +# Buffers use s16le format, with 48 frame per 1000us on core 0 with priority 0 +DAI_ADD(sof/pipe-dai-playback.m4,
- 2, SSP, 5, SSP5-Codec,
- PIPELINE_SOURCE_2, 2, s16le,
- 48, 1000, 0, 0)
+# playback DAI is SSP1 using 2 periods +# Buffers use s16le format, with 48 frame per 1000us on core 0 with priority 0 +DAI_ADD(sof/pipe-dai-playback.m4,
- 3, SSP, 1, SSP1-Codec,
- PIPELINE_SOURCE_3, 2, s16le,
- 48, 1000, 0, 0)
+# FIXME: Why is this needed? And is this correct? +# PCM (PCM_CAPTURE_ADD(name, pipeline, pcm_id, dai_id, capture)) +#PCM_CAPTURE_ADD(DMIC0, 1, 0, 0, PIPELINE_PCM_1) +#PCM_PLAYBACK_ADD(SSP5, 2, 0, 5, PIPELINE_PCM_2) +#PCM_PLAYBACK_ADD(SSP1, 3, 1, 1, PIPELINE_PCM_3)
+# +# BE configurations - overrides config in ACPI if present +# +# FIXME: the machine driver uses .name = "dmic01" and .id=2 +# DMIC0 (id: 2) +DAI_CONFIG(DMIC, 0, 2, dmic01,
- DMIC_CONFIG(1, 500000, 4800000, 40, 60, 48000,
DMIC_WORD_LENGTH(s32le), DMIC, 0,PDM_CONFIG(DMIC, 0, FOUR_CH_PDM0_PDM1)))+# FIXME: the machine driver uses .name = "SSP5-Codec", .id=0, .cpu_name="SSP5 Pin" +#SSP 5 (ID: 0) with 19.2MHz mclk with MCLK_ID 0 +DAI_CONFIG(SSP, 5, 0, SSP5-Codec,
- SSP_CONFIG(I2S, SSP_CLOCK(mclk, 19200000, codec_mclk_in),
SSP_CLOCK(bclk, 1920000, codec_slave),SSP_CLOCK(fsync, 48000, codec_slave),SSP_TDM(2, 20, 3, 3),SSP_CONFIG_DATA(SSP, 5, 16, 0)))+# FIXME: the machine driver uses .name = "SSP1-Codec", .id=1, .cpu_name="SSP1 Pin" +#SSP 1 (ID: 1) with 19.2MHz mclk with MCLK_ID 0 +DAI_CONFIG(SSP, 1, 1, SSP1-Codec,
- SSP_CONFIG(I2S, SSP_CLOCK(mclk, 19200000, codec_mclk_in),
SSP_CLOCK(bclk, 1920000, codec_slave),SSP_CLOCK(fsync, 48000, codec_slave),SSP_TDM(2, 20, 3, 3),SSP_CONFIG_DATA(SSP, 1, 16, 0)))
Sound-open-firmware mailing list Sound-open-firmware@alsa-project.org http://mailman.alsa-project.org/mailman/listinfo/sound-open-firmware
On Wed, 2018-07-11 at 11:55 -0500, Pierre-Louis Bossart wrote:
On 07/11/2018 01:23 AM, Ranjani Sridharan wrote:
On Tue, 2018-07-10 at 18:15 -0500, Pierre-Louis Bossart wrote:
Add based on sof-glk-codec. In theory this is the same topology with only a swap of SSPs
FIXME: this is a mess between pipeline ID, PCM ID, DAI ID, DAI #. I could use additional eyes to proof-read this...
Looks good to me. I couldnt spot any mistakes. Keyon, Could you please also check?
FIXME: not sure why we would need PCM_CAPTURE_ADD/PCM_PLAYBACK_DAI, this seems to be already done by DAI_ADD?
DAI_ADD only adds the BE DAI's. PCM_CAPTURE_ADD/PLAYBACK_ADD are needed for adding the FE dai links.
ok, but then why do those macros make references to the DAIs?
The dai_id is used when creating the cpu_dai for the PCM elems and their ID is set to the dai_id specified here.
I wish I understood cpu dai's better but I couldnt wrap my head around why we do that.
Keyon, any hints?
+# PCM (PCM_CAPTURE_ADD(name, pipeline, pcm_id, dai_id, capture)) +#PCM_CAPTURE_ADD(DMIC0, 1, 0, 0, PIPELINE_PCM_1) +#PCM_PLAYBACK_ADD(SSP5, 2, 0, 5, PIPELINE_PCM_2) +#PCM_PLAYBACK_ADD(SSP1, 3, 1, 1, PIPELINE_PCM_3)
In addition the information on the PCM # is already provided in PIPELINE_ADD, so this looks redundant on both the FE and BE sides?
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.i ntel .com>
topology/Makefile.am | 1 + topology/sof-apl-da7219.m4 | 104 +++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 105 insertions(+) create mode 100644 topology/sof-apl-da7219.m4
diff --git a/topology/Makefile.am b/topology/Makefile.am index 33d0418..396b5b9 100644 --- a/topology/Makefile.am +++ b/topology/Makefile.am @@ -31,6 +31,7 @@ MACHINES = \ sof-hsw-rt5640.tplg \ sof-apl-tdf8532.tplg \ sof-apl-pcm512x.tplg \
sof-apl-da7219.tplg \ sof-glk-codec.tplg
# Uncomment the following line if you want to debug conf files
diff --git a/topology/sof-apl-da7219.m4 b/topology/sof-apl- da7219.m4 new file mode 100644 index 0000000..3934df5 --- /dev/null +++ b/topology/sof-apl-da7219.m4 @@ -0,0 +1,104 @@ +# +# Topology for AppoloLake with headset on SSP1, spk on SSP5 and DMIC+# PCM (PCM_CAPTURE_ADD(name, pipeline, pcm_id, dai_id, capture)) +#PCM_CAPTURE_ADD(DMIC0, 1, 0, 0, PIPELINE_PCM_1) +#PCM_PLAYBACK_ADD(SSP5, 2, 0, 5, PIPELINE_PCM_2) +#PCM_PLAYBACK_ADD(SSP1, 3, 1, 1, PIPELINE_PCM_3) capture +#
+# Include topology builder +include(`utils.m4') +include(`dai.m4') +include(`pipeline.m4') +include(`ssp.m4')
+# Include TLV library +include(`common/tlv.m4')
+# Include Token library +include(`sof/tokens.m4')
+# Include bxt DSP configuration +include(`platform/intel/bxt.m4') +include(`platform/intel/dmic.m4')
+# +# Define the pipelines +# +# PCM0 <---- volume <----- DMIC0 (dmic capture) +# PCM0 ----> volume -----> SSP5 (speaker - maxim98357a) +# PCM1 ----> volume -----> SSP1 (headset - da7219) +#
+# Low Latency capture pipeline 1 on PCM 0 using max 4 channels of s32le. +# Schedule 48 frames per 1000us deadline on core 0 with priority 0 +PIPELINE_PCM_ADD(sof/pipe-volume-capture.m4,
- 1, 0, 4, s32le,
- 48, 1000, 0, 0)
+# Low Latency playback pipeline 2 on PCM 0 using max 2 channels of s16le. +# Schedule 48 frames per 1000us deadline on core 0 with priority 0 +PIPELINE_PCM_ADD(sof/pipe-volume-playback.m4,
- 2, 0, 2, s16le,
- 48, 1000, 0, 0)
+# Low Latency playback pipeline 3 on PCM 1 using max 2 channels of s16le. +# Schedule 48 frames per 1000us deadline on core 0 with priority 0 +PIPELINE_PCM_ADD(sof/pipe-volume-playback.m4,
- 3, 1, 2, s16le,
- 48, 1000, 0, 0)
+# +# DAIs configuration +#
+# capture DAI is DMIC0 using 2 periods +# Buffers use s16le format, with 48 frame per 1000us on core 0 with priority 0 +DAI_ADD(sof/pipe-dai-capture.m4,
- 1, DMIC, 0, DMIC0,
- PIPELINE_SINK_1, 2, s32le,
- 48, 1000, 0, 0)
+# playback DAI is SSP5 using 2 periods +# Buffers use s16le format, with 48 frame per 1000us on core 0 with priority 0 +DAI_ADD(sof/pipe-dai-playback.m4,
- 2, SSP, 5, SSP5-Codec,
- PIPELINE_SOURCE_2, 2, s16le,
- 48, 1000, 0, 0)
+# playback DAI is SSP1 using 2 periods +# Buffers use s16le format, with 48 frame per 1000us on core 0 with priority 0 +DAI_ADD(sof/pipe-dai-playback.m4,
- 3, SSP, 1, SSP1-Codec,
- PIPELINE_SOURCE_3, 2, s16le,
- 48, 1000, 0, 0)
+# FIXME: Why is this needed? And is this correct? +# PCM (PCM_CAPTURE_ADD(name, pipeline, pcm_id, dai_id, capture)) +#PCM_CAPTURE_ADD(DMIC0, 1, 0, 0, PIPELINE_PCM_1) +#PCM_PLAYBACK_ADD(SSP5, 2, 0, 5, PIPELINE_PCM_2) +#PCM_PLAYBACK_ADD(SSP1, 3, 1, 1, PIPELINE_PCM_3)
+# +# BE configurations - overrides config in ACPI if present +# +# FIXME: the machine driver uses .name = "dmic01" and .id=2 +# DMIC0 (id: 2) +DAI_CONFIG(DMIC, 0, 2, dmic01,
- DMIC_CONFIG(1, 500000, 4800000, 40, 60, 48000,
DMIC_WORD_LENGTH(s32le), DMIC, 0,PDM_CONFIG(DMIC, 0, FOUR_CH_PDM0_PDM1)))+# FIXME: the machine driver uses .name = "SSP5-Codec", .id=0, .cpu_name="SSP5 Pin" +#SSP 5 (ID: 0) with 19.2MHz mclk with MCLK_ID 0 +DAI_CONFIG(SSP, 5, 0, SSP5-Codec,
- SSP_CONFIG(I2S, SSP_CLOCK(mclk, 19200000,
codec_mclk_in),
SSP_CLOCK(bclk, 1920000, codec_slave),SSP_CLOCK(fsync, 48000, codec_slave),SSP_TDM(2, 20, 3, 3),SSP_CONFIG_DATA(SSP, 5, 16, 0)))+# FIXME: the machine driver uses .name = "SSP1-Codec", .id=1, .cpu_name="SSP1 Pin" +#SSP 1 (ID: 1) with 19.2MHz mclk with MCLK_ID 0 +DAI_CONFIG(SSP, 1, 1, SSP1-Codec,
- SSP_CONFIG(I2S, SSP_CLOCK(mclk, 19200000,
codec_mclk_in),
SSP_CLOCK(bclk, 1920000, codec_slave),SSP_CLOCK(fsync, 48000, codec_slave),SSP_TDM(2, 20, 3, 3),SSP_CONFIG_DATA(SSP, 1, 16, 0)))
Sound-open-firmware mailing list Sound-open-firmware@alsa-project.org http://mailman.alsa-project.org/mailman/listinfo/sound-open-firmwar e
On 07/11/2018 04:43 PM, Ranjani Sridharan wrote:
On Wed, 2018-07-11 at 11:55 -0500, Pierre-Louis Bossart wrote:
On 07/11/2018 01:23 AM, Ranjani Sridharan wrote:
On Tue, 2018-07-10 at 18:15 -0500, Pierre-Louis Bossart wrote:
Add based on sof-glk-codec. In theory this is the same topology with only a swap of SSPs
FIXME: this is a mess between pipeline ID, PCM ID, DAI ID, DAI #. I could use additional eyes to proof-read this...
Looks good to me. I couldnt spot any mistakes. Keyon, Could you please also check?
FIXME: not sure why we would need PCM_CAPTURE_ADD/PCM_PLAYBACK_DAI, this seems to be already done by DAI_ADD?
DAI_ADD only adds the BE DAI's. PCM_CAPTURE_ADD/PLAYBACK_ADD are needed for adding the FE dai links.
ok, but then why do those macros make references to the DAIs?
The dai_id is used when creating the cpu_dai for the PCM elems and their ID is set to the dai_id specified here.
I wish I understood cpu dai's better but I couldnt wrap my head around why we do that.
Keyon, any hints?
So it's a different dai_id from what is used for back-ends? Looks like a complete mess to me...
+# PCM (PCM_CAPTURE_ADD(name, pipeline, pcm_id, dai_id, capture)) +#PCM_CAPTURE_ADD(DMIC0, 1, 0, 0, PIPELINE_PCM_1) +#PCM_PLAYBACK_ADD(SSP5, 2, 0, 5, PIPELINE_PCM_2) +#PCM_PLAYBACK_ADD(SSP1, 3, 1, 1, PIPELINE_PCM_3)
In addition the information on the PCM # is already provided in PIPELINE_ADD, so this looks redundant on both the FE and BE sides?
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.i ntel .com>
topology/Makefile.am | 1 + topology/sof-apl-da7219.m4 | 104 +++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 105 insertions(+) create mode 100644 topology/sof-apl-da7219.m4
diff --git a/topology/Makefile.am b/topology/Makefile.am index 33d0418..396b5b9 100644 --- a/topology/Makefile.am +++ b/topology/Makefile.am @@ -31,6 +31,7 @@ MACHINES = \ sof-hsw-rt5640.tplg \ sof-apl-tdf8532.tplg \ sof-apl-pcm512x.tplg \
sof-apl-da7219.tplg \ sof-glk-codec.tplg
# Uncomment the following line if you want to debug conf files
diff --git a/topology/sof-apl-da7219.m4 b/topology/sof-apl- da7219.m4 new file mode 100644 index 0000000..3934df5 --- /dev/null +++ b/topology/sof-apl-da7219.m4 @@ -0,0 +1,104 @@ +# +# Topology for AppoloLake with headset on SSP1, spk on SSP5 and DMIC+# PCM (PCM_CAPTURE_ADD(name, pipeline, pcm_id, dai_id, capture)) +#PCM_CAPTURE_ADD(DMIC0, 1, 0, 0, PIPELINE_PCM_1) +#PCM_PLAYBACK_ADD(SSP5, 2, 0, 5, PIPELINE_PCM_2) +#PCM_PLAYBACK_ADD(SSP1, 3, 1, 1, PIPELINE_PCM_3) capture +#
+# Include topology builder +include(`utils.m4') +include(`dai.m4') +include(`pipeline.m4') +include(`ssp.m4')
+# Include TLV library +include(`common/tlv.m4')
+# Include Token library +include(`sof/tokens.m4')
+# Include bxt DSP configuration +include(`platform/intel/bxt.m4') +include(`platform/intel/dmic.m4')
+# +# Define the pipelines +# +# PCM0 <---- volume <----- DMIC0 (dmic capture) +# PCM0 ----> volume -----> SSP5 (speaker - maxim98357a) +# PCM1 ----> volume -----> SSP1 (headset - da7219) +#
+# Low Latency capture pipeline 1 on PCM 0 using max 4 channels of s32le. +# Schedule 48 frames per 1000us deadline on core 0 with priority 0 +PIPELINE_PCM_ADD(sof/pipe-volume-capture.m4,
- 1, 0, 4, s32le,
- 48, 1000, 0, 0)
+# Low Latency playback pipeline 2 on PCM 0 using max 2 channels of s16le. +# Schedule 48 frames per 1000us deadline on core 0 with priority 0 +PIPELINE_PCM_ADD(sof/pipe-volume-playback.m4,
- 2, 0, 2, s16le,
- 48, 1000, 0, 0)
+# Low Latency playback pipeline 3 on PCM 1 using max 2 channels of s16le. +# Schedule 48 frames per 1000us deadline on core 0 with priority 0 +PIPELINE_PCM_ADD(sof/pipe-volume-playback.m4,
- 3, 1, 2, s16le,
- 48, 1000, 0, 0)
+# +# DAIs configuration +#
+# capture DAI is DMIC0 using 2 periods +# Buffers use s16le format, with 48 frame per 1000us on core 0 with priority 0 +DAI_ADD(sof/pipe-dai-capture.m4,
- 1, DMIC, 0, DMIC0,
- PIPELINE_SINK_1, 2, s32le,
- 48, 1000, 0, 0)
+# playback DAI is SSP5 using 2 periods +# Buffers use s16le format, with 48 frame per 1000us on core 0 with priority 0 +DAI_ADD(sof/pipe-dai-playback.m4,
- 2, SSP, 5, SSP5-Codec,
- PIPELINE_SOURCE_2, 2, s16le,
- 48, 1000, 0, 0)
+# playback DAI is SSP1 using 2 periods +# Buffers use s16le format, with 48 frame per 1000us on core 0 with priority 0 +DAI_ADD(sof/pipe-dai-playback.m4,
- 3, SSP, 1, SSP1-Codec,
- PIPELINE_SOURCE_3, 2, s16le,
- 48, 1000, 0, 0)
+# FIXME: Why is this needed? And is this correct? +# PCM (PCM_CAPTURE_ADD(name, pipeline, pcm_id, dai_id, capture)) +#PCM_CAPTURE_ADD(DMIC0, 1, 0, 0, PIPELINE_PCM_1) +#PCM_PLAYBACK_ADD(SSP5, 2, 0, 5, PIPELINE_PCM_2) +#PCM_PLAYBACK_ADD(SSP1, 3, 1, 1, PIPELINE_PCM_3)
+# +# BE configurations - overrides config in ACPI if present +# +# FIXME: the machine driver uses .name = "dmic01" and .id=2 +# DMIC0 (id: 2) +DAI_CONFIG(DMIC, 0, 2, dmic01,
- DMIC_CONFIG(1, 500000, 4800000, 40, 60, 48000,
DMIC_WORD_LENGTH(s32le), DMIC, 0,PDM_CONFIG(DMIC, 0, FOUR_CH_PDM0_PDM1)))+# FIXME: the machine driver uses .name = "SSP5-Codec", .id=0, .cpu_name="SSP5 Pin" +#SSP 5 (ID: 0) with 19.2MHz mclk with MCLK_ID 0 +DAI_CONFIG(SSP, 5, 0, SSP5-Codec,
- SSP_CONFIG(I2S, SSP_CLOCK(mclk, 19200000,
codec_mclk_in),
SSP_CLOCK(bclk, 1920000, codec_slave),SSP_CLOCK(fsync, 48000, codec_slave),SSP_TDM(2, 20, 3, 3),SSP_CONFIG_DATA(SSP, 5, 16, 0)))+# FIXME: the machine driver uses .name = "SSP1-Codec", .id=1, .cpu_name="SSP1 Pin" +#SSP 1 (ID: 1) with 19.2MHz mclk with MCLK_ID 0 +DAI_CONFIG(SSP, 1, 1, SSP1-Codec,
- SSP_CONFIG(I2S, SSP_CLOCK(mclk, 19200000,
codec_mclk_in),
SSP_CLOCK(bclk, 1920000, codec_slave),SSP_CLOCK(fsync, 48000, codec_slave),SSP_TDM(2, 20, 3, 3),SSP_CONFIG_DATA(SSP, 1, 16, 0)))
Sound-open-firmware mailing list Sound-open-firmware@alsa-project.org http://mailman.alsa-project.org/mailman/listinfo/sound-open-firmwar e
On Wed, 2018-07-11 at 17:52 -0500, Pierre-Louis Bossart wrote:
On 07/11/2018 04:43 PM, Ranjani Sridharan wrote:
On Wed, 2018-07-11 at 11:55 -0500, Pierre-Louis Bossart wrote:
On 07/11/2018 01:23 AM, Ranjani Sridharan wrote:
On Tue, 2018-07-10 at 18:15 -0500, Pierre-Louis Bossart wrote:
Add based on sof-glk-codec. In theory this is the same topology with only a swap of SSPs
FIXME: this is a mess between pipeline ID, PCM ID, DAI ID, DAI #. I could use additional eyes to proof-read this...
Looks good to me. I couldnt spot any mistakes. Keyon, Could you please also check?
FIXME: not sure why we would need PCM_CAPTURE_ADD/PCM_PLAYBACK_DAI, this seems to be already done by DAI_ADD?
DAI_ADD only adds the BE DAI's. PCM_CAPTURE_ADD/PLAYBACK_ADD are needed for adding the FE dai links.
ok, but then why do those macros make references to the DAIs?
The dai_id is used when creating the cpu_dai for the PCM elems and their ID is set to the dai_id specified here.
I wish I understood cpu dai's better but I couldnt wrap my head around why we do that.
Keyon, any hints?
So it's a different dai_id from what is used for back-ends? Looks like a complete mess to me...
It is the same ID that is used for the BE DAI, Pierre. Here's a quick look at the all the DAI's that are registered in the case of up2 with the hifiberry dac+ with the test-playback-ssp5*.tplg.
The "Passthrough 0" dai is the last one and has the dai_id of 0 which is the same as the SSP5 id in the topology.
[ 0.076806] NET: Registered protocol family 16 [ 0.176459] Registered efivars operations [ 0.319302] NET: Registered protocol family 2 [ 0.320794] NET: Registered protocol family 1 [ 1.467017] NET: Registered protocol family 10 [ 1.467826] NET: Registered protocol family 17 [ 3.038697] snd-soc-dummy snd-soc-dummy: ASoC: Registered DAI 'snd- soc-dummy-dai' with id 0 [ 3.227853] sof-audio sof-audio: ASoC: Registered platform 'sof- audio' [ 3.227862] sof-audio sof-audio: ASoC: Registered DAI 'SSP0 Pin' with id 0 [ 3.227866] sof-audio sof-audio: ASoC: Registered DAI 'SSP1 Pin' with id 1 [ 3.227869] sof-audio sof-audio: ASoC: Registered DAI 'SSP2 Pin' with id 2 [ 3.227873] sof-audio sof-audio: ASoC: Registered DAI 'SSP3 Pin' with id 3 [ 3.227876] sof-audio sof-audio: ASoC: Registered DAI 'SSP4 Pin' with id 4 [ 3.227879] sof-audio sof-audio: ASoC: Registered DAI 'SSP5 Pin' with id 5 [ 3.227883] sof-audio sof-audio: ASoC: Registered DAI 'DMIC01 Pin' with id 6 [ 3.227886] sof-audio sof-audio: ASoC: Registered DAI 'DMIC16k Pin' with id 7 [ 3.227889] sof-audio sof-audio: ASoC: Registered DAI 'iDisp1 Pin' with id 8 [ 3.227893] sof-audio sof-audio: ASoC: Registered DAI 'iDisp2 Pin' with id 9 [ 3.227896] sof-audio sof-audio: ASoC: Registered DAI 'iDisp3 Pin' with id 10 [ 3.227900] sof-audio sof-audio: ASoC: Registered DAI 'Analog Codec DAI' with id 11 [ 3.227903] sof-audio sof-audio: ASoC: Registered DAI 'Digital Codec DAI' with id 12 [ 3.227906] sof-audio sof-audio: ASoC: Registered DAI 'Alt Analog Codec DAI' with id 13 [ 3.578667] pcm512x i2c-104C5122:00: ASoC: Registered DAI 'pcm512x- hifi' with id 0 [ 4.141636] pstore: Registered efi as persistent store backend [ 4.203432] sof-audio sof-audio: ASoC: Registered DAI 'Passthrough 0' with id 0 [ 4.609670] NET: Registered protocol family 31
+# PCM (PCM_CAPTURE_ADD(name, pipeline, pcm_id, dai_id, capture)) +#PCM_CAPTURE_ADD(DMIC0, 1, 0, 0, PIPELINE_PCM_1) +#PCM_PLAYBACK_ADD(SSP5, 2, 0, 5, PIPELINE_PCM_2) +#PCM_PLAYBACK_ADD(SSP1, 3, 1, 1, PIPELINE_PCM_3)
In addition the information on the PCM # is already provided in PIPELINE_ADD, so this looks redundant on both the FE and BE sides?
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@lin ux.i ntel .com>
topology/Makefile.am | 1 + topology/sof-apl-da7219.m4 | 104 +++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 105 insertions(+) create mode 100644 topology/sof-apl-da7219.m4
diff --git a/topology/Makefile.am b/topology/Makefile.am index 33d0418..396b5b9 100644 --- a/topology/Makefile.am +++ b/topology/Makefile.am @@ -31,6 +31,7 @@ MACHINES = \ sof-hsw-rt5640.tplg \ sof-apl-tdf8532.tplg \ sof-apl-pcm512x.tplg \
sof-apl-da7219.tplg \ sof-glk-codec.tplg
# Uncomment the following line if you want to debug conf
files diff --git a/topology/sof-apl-da7219.m4 b/topology/sof-apl- da7219.m4 new file mode 100644 index 0000000..3934df5 --- /dev/null +++ b/topology/sof-apl-da7219.m4 @@ -0,0 +1,104 @@ +# +# Topology for AppoloLake with headset on SSP1, spk on SSP5 and DMIC+# PCM (PCM_CAPTURE_ADD(name, pipeline, pcm_id, dai_id, capture)) +#PCM_CAPTURE_ADD(DMIC0, 1, 0, 0, PIPELINE_PCM_1) +#PCM_PLAYBACK_ADD(SSP5, 2, 0, 5, PIPELINE_PCM_2) +#PCM_PLAYBACK_ADD(SSP1, 3, 1, 1, PIPELINE_PCM_3) capture +#
+# Include topology builder +include(`utils.m4') +include(`dai.m4') +include(`pipeline.m4') +include(`ssp.m4')
+# Include TLV library +include(`common/tlv.m4')
+# Include Token library +include(`sof/tokens.m4')
+# Include bxt DSP configuration +include(`platform/intel/bxt.m4') +include(`platform/intel/dmic.m4')
+# +# Define the pipelines +# +# PCM0 <---- volume <----- DMIC0 (dmic capture) +# PCM0 ----> volume -----> SSP5 (speaker - maxim98357a) +# PCM1 ----> volume -----> SSP1 (headset - da7219) +#
+# Low Latency capture pipeline 1 on PCM 0 using max 4 channels of s32le. +# Schedule 48 frames per 1000us deadline on core 0 with priority 0 +PIPELINE_PCM_ADD(sof/pipe-volume-capture.m4,
- 1, 0, 4, s32le,
- 48, 1000, 0, 0)
+# Low Latency playback pipeline 2 on PCM 0 using max 2 channels of s16le. +# Schedule 48 frames per 1000us deadline on core 0 with priority 0 +PIPELINE_PCM_ADD(sof/pipe-volume-playback.m4,
- 2, 0, 2, s16le,
- 48, 1000, 0, 0)
+# Low Latency playback pipeline 3 on PCM 1 using max 2 channels of s16le. +# Schedule 48 frames per 1000us deadline on core 0 with priority 0 +PIPELINE_PCM_ADD(sof/pipe-volume-playback.m4,
- 3, 1, 2, s16le,
- 48, 1000, 0, 0)
+# +# DAIs configuration +#
+# capture DAI is DMIC0 using 2 periods +# Buffers use s16le format, with 48 frame per 1000us on core 0 with priority 0 +DAI_ADD(sof/pipe-dai-capture.m4,
- 1, DMIC, 0, DMIC0,
- PIPELINE_SINK_1, 2, s32le,
- 48, 1000, 0, 0)
+# playback DAI is SSP5 using 2 periods +# Buffers use s16le format, with 48 frame per 1000us on core 0 with priority 0 +DAI_ADD(sof/pipe-dai-playback.m4,
- 2, SSP, 5, SSP5-Codec,
- PIPELINE_SOURCE_2, 2, s16le,
- 48, 1000, 0, 0)
+# playback DAI is SSP1 using 2 periods +# Buffers use s16le format, with 48 frame per 1000us on core 0 with priority 0 +DAI_ADD(sof/pipe-dai-playback.m4,
- 3, SSP, 1, SSP1-Codec,
- PIPELINE_SOURCE_3, 2, s16le,
- 48, 1000, 0, 0)
+# FIXME: Why is this needed? And is this correct? +# PCM (PCM_CAPTURE_ADD(name, pipeline, pcm_id, dai_id, capture)) +#PCM_CAPTURE_ADD(DMIC0, 1, 0, 0, PIPELINE_PCM_1) +#PCM_PLAYBACK_ADD(SSP5, 2, 0, 5, PIPELINE_PCM_2) +#PCM_PLAYBACK_ADD(SSP1, 3, 1, 1, PIPELINE_PCM_3)
+# +# BE configurations - overrides config in ACPI if present +# +# FIXME: the machine driver uses .name = "dmic01" and .id=2 +# DMIC0 (id: 2) +DAI_CONFIG(DMIC, 0, 2, dmic01,
- DMIC_CONFIG(1, 500000, 4800000, 40, 60, 48000,
DMIC_WORD_LENGTH(s32le), DMIC, 0,PDM_CONFIG(DMIC, 0, FOUR_CH_PDM0_PDM1)))+# FIXME: the machine driver uses .name = "SSP5-Codec", .id=0, .cpu_name="SSP5 Pin" +#SSP 5 (ID: 0) with 19.2MHz mclk with MCLK_ID 0 +DAI_CONFIG(SSP, 5, 0, SSP5-Codec,
- SSP_CONFIG(I2S, SSP_CLOCK(mclk, 19200000,
codec_mclk_in),
SSP_CLOCK(bclk, 1920000, codec_slave),SSP_CLOCK(fsync, 48000, codec_slave),SSP_TDM(2, 20, 3, 3),SSP_CONFIG_DATA(SSP, 5, 16, 0)))+# FIXME: the machine driver uses .name = "SSP1-Codec", .id=1, .cpu_name="SSP1 Pin" +#SSP 1 (ID: 1) with 19.2MHz mclk with MCLK_ID 0 +DAI_CONFIG(SSP, 1, 1, SSP1-Codec,
- SSP_CONFIG(I2S, SSP_CLOCK(mclk, 19200000,
codec_mclk_in),
SSP_CLOCK(bclk, 1920000, codec_slave),SSP_CLOCK(fsync, 48000, codec_slave),SSP_TDM(2, 20, 3, 3),SSP_CONFIG_DATA(SSP, 1, 16, 0)))
Sound-open-firmware mailing list Sound-open-firmware@alsa-project.org http://mailman.alsa-project.org/mailman/listinfo/sound-open-fir mwar e
Sound-open-firmware mailing list Sound-open-firmware@alsa-project.org http://mailman.alsa-project.org/mailman/listinfo/sound-open-firmware
On 07/11/2018 08:14 PM, Ranjani Sridharan wrote:
On Wed, 2018-07-11 at 17:52 -0500, Pierre-Louis Bossart wrote:
On 07/11/2018 04:43 PM, Ranjani Sridharan wrote:
On Wed, 2018-07-11 at 11:55 -0500, Pierre-Louis Bossart wrote:
On 07/11/2018 01:23 AM, Ranjani Sridharan wrote:
On Tue, 2018-07-10 at 18:15 -0500, Pierre-Louis Bossart wrote:
Add based on sof-glk-codec. In theory this is the same topology with only a swap of SSPs
FIXME: this is a mess between pipeline ID, PCM ID, DAI ID, DAI #. I could use additional eyes to proof-read this...
Looks good to me. I couldnt spot any mistakes. Keyon, Could you please also check?
FIXME: not sure why we would need PCM_CAPTURE_ADD/PCM_PLAYBACK_DAI, this seems to be already done by DAI_ADD?
DAI_ADD only adds the BE DAI's. PCM_CAPTURE_ADD/PLAYBACK_ADD are needed for adding the FE dai links.
ok, but then why do those macros make references to the DAIs?
The dai_id is used when creating the cpu_dai for the PCM elems and their ID is set to the dai_id specified here.
I wish I understood cpu dai's better but I couldnt wrap my head around why we do that.
Keyon, any hints?
So it's a different dai_id from what is used for back-ends? Looks like a complete mess to me...
It is the same ID that is used for the BE DAI, Pierre. Here's a quick look at the all the DAI's that are registered in the case of up2 with the hifiberry dac+ with the test-playback-ssp5*.tplg.
The "Passthrough 0" dai is the last one and has the dai_id of 0 which is the same as the SSP5 id in the topology.
I just don't get all this. I started from the sof-glk-codec.m4 file and both that file and this one are NOT passthrough, they all contain a volume element which *should* decouple front-end from back-end.
[ 0.076806] NET: Registered protocol family 16 [ 0.176459] Registered efivars operations [ 0.319302] NET: Registered protocol family 2 [ 0.320794] NET: Registered protocol family 1 [ 1.467017] NET: Registered protocol family 10 [ 1.467826] NET: Registered protocol family 17 [ 3.038697] snd-soc-dummy snd-soc-dummy: ASoC: Registered DAI 'snd- soc-dummy-dai' with id 0 [ 3.227853] sof-audio sof-audio: ASoC: Registered platform 'sof- audio' [ 3.227862] sof-audio sof-audio: ASoC: Registered DAI 'SSP0 Pin' with id 0 [ 3.227866] sof-audio sof-audio: ASoC: Registered DAI 'SSP1 Pin' with id 1 [ 3.227869] sof-audio sof-audio: ASoC: Registered DAI 'SSP2 Pin' with id 2 [ 3.227873] sof-audio sof-audio: ASoC: Registered DAI 'SSP3 Pin' with id 3 [ 3.227876] sof-audio sof-audio: ASoC: Registered DAI 'SSP4 Pin' with id 4 [ 3.227879] sof-audio sof-audio: ASoC: Registered DAI 'SSP5 Pin' with id 5 [ 3.227883] sof-audio sof-audio: ASoC: Registered DAI 'DMIC01 Pin' with id 6 [ 3.227886] sof-audio sof-audio: ASoC: Registered DAI 'DMIC16k Pin' with id 7 [ 3.227889] sof-audio sof-audio: ASoC: Registered DAI 'iDisp1 Pin' with id 8 [ 3.227893] sof-audio sof-audio: ASoC: Registered DAI 'iDisp2 Pin' with id 9 [ 3.227896] sof-audio sof-audio: ASoC: Registered DAI 'iDisp3 Pin' with id 10 [ 3.227900] sof-audio sof-audio: ASoC: Registered DAI 'Analog Codec DAI' with id 11 [ 3.227903] sof-audio sof-audio: ASoC: Registered DAI 'Digital Codec DAI' with id 12 [ 3.227906] sof-audio sof-audio: ASoC: Registered DAI 'Alt Analog Codec DAI' with id 13 [ 3.578667] pcm512x i2c-104C5122:00: ASoC: Registered DAI 'pcm512x- hifi' with id 0 [ 4.141636] pstore: Registered efi as persistent store backend [ 4.203432] sof-audio sof-audio: ASoC: Registered DAI 'Passthrough 0' with id 0
And I don't get what this means either? What is a Passthrough DAI? it's just not self-explanatory.
[ 4.609670] NET: Registered protocol family 31
+# PCM (PCM_CAPTURE_ADD(name, pipeline, pcm_id, dai_id, capture)) +#PCM_CAPTURE_ADD(DMIC0, 1, 0, 0, PIPELINE_PCM_1) +#PCM_PLAYBACK_ADD(SSP5, 2, 0, 5, PIPELINE_PCM_2) +#PCM_PLAYBACK_ADD(SSP1, 3, 1, 1, PIPELINE_PCM_3)
In addition the information on the PCM # is already provided in PIPELINE_ADD, so this looks redundant on both the FE and BE sides?
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@lin ux.i ntel .com>
topology/Makefile.am | 1 + topology/sof-apl-da7219.m4 | 104+++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 105 insertions(+) create mode 100644 topology/sof-apl-da7219.m4
diff --git a/topology/Makefile.am b/topology/Makefile.am index 33d0418..396b5b9 100644 --- a/topology/Makefile.am +++ b/topology/Makefile.am @@ -31,6 +31,7 @@ MACHINES = \ sof-hsw-rt5640.tplg \ sof-apl-tdf8532.tplg \ sof-apl-pcm512x.tplg \
sof-apl-da7219.tplg \ sof-glk-codec.tplg
# Uncomment the following line if you want to debug conf
files diff --git a/topology/sof-apl-da7219.m4 b/topology/sof-apl- da7219.m4 new file mode 100644 index 0000000..3934df5 --- /dev/null +++ b/topology/sof-apl-da7219.m4 @@ -0,0 +1,104 @@ +# +# Topology for AppoloLake with headset on SSP1, spk on SSP5 and DMIC+# PCM (PCM_CAPTURE_ADD(name, pipeline, pcm_id, dai_id, capture)) +#PCM_CAPTURE_ADD(DMIC0, 1, 0, 0, PIPELINE_PCM_1) +#PCM_PLAYBACK_ADD(SSP5, 2, 0, 5, PIPELINE_PCM_2) +#PCM_PLAYBACK_ADD(SSP1, 3, 1, 1, PIPELINE_PCM_3) capture +#
+# Include topology builder +include(`utils.m4') +include(`dai.m4') +include(`pipeline.m4') +include(`ssp.m4')
+# Include TLV library +include(`common/tlv.m4')
+# Include Token library +include(`sof/tokens.m4')
+# Include bxt DSP configuration +include(`platform/intel/bxt.m4') +include(`platform/intel/dmic.m4')
+# +# Define the pipelines +# +# PCM0 <---- volume <----- DMIC0 (dmic capture) +# PCM0 ----> volume -----> SSP5 (speaker - maxim98357a) +# PCM1 ----> volume -----> SSP1 (headset - da7219) +#
+# Low Latency capture pipeline 1 on PCM 0 using max 4 channels of s32le. +# Schedule 48 frames per 1000us deadline on core 0 with priority 0 +PIPELINE_PCM_ADD(sof/pipe-volume-capture.m4,
- 1, 0, 4, s32le,
- 48, 1000, 0, 0)
+# Low Latency playback pipeline 2 on PCM 0 using max 2 channels of s16le. +# Schedule 48 frames per 1000us deadline on core 0 with priority 0 +PIPELINE_PCM_ADD(sof/pipe-volume-playback.m4,
- 2, 0, 2, s16le,
- 48, 1000, 0, 0)
+# Low Latency playback pipeline 3 on PCM 1 using max 2 channels of s16le. +# Schedule 48 frames per 1000us deadline on core 0 with priority 0 +PIPELINE_PCM_ADD(sof/pipe-volume-playback.m4,
- 3, 1, 2, s16le,
- 48, 1000, 0, 0)
+# +# DAIs configuration +#
+# capture DAI is DMIC0 using 2 periods +# Buffers use s16le format, with 48 frame per 1000us on core 0 with priority 0 +DAI_ADD(sof/pipe-dai-capture.m4,
- 1, DMIC, 0, DMIC0,
- PIPELINE_SINK_1, 2, s32le,
- 48, 1000, 0, 0)
+# playback DAI is SSP5 using 2 periods +# Buffers use s16le format, with 48 frame per 1000us on core 0 with priority 0 +DAI_ADD(sof/pipe-dai-playback.m4,
- 2, SSP, 5, SSP5-Codec,
- PIPELINE_SOURCE_2, 2, s16le,
- 48, 1000, 0, 0)
+# playback DAI is SSP1 using 2 periods +# Buffers use s16le format, with 48 frame per 1000us on core 0 with priority 0 +DAI_ADD(sof/pipe-dai-playback.m4,
- 3, SSP, 1, SSP1-Codec,
- PIPELINE_SOURCE_3, 2, s16le,
- 48, 1000, 0, 0)
+# FIXME: Why is this needed? And is this correct? +# PCM (PCM_CAPTURE_ADD(name, pipeline, pcm_id, dai_id, capture)) +#PCM_CAPTURE_ADD(DMIC0, 1, 0, 0, PIPELINE_PCM_1) +#PCM_PLAYBACK_ADD(SSP5, 2, 0, 5, PIPELINE_PCM_2) +#PCM_PLAYBACK_ADD(SSP1, 3, 1, 1, PIPELINE_PCM_3)
+# +# BE configurations - overrides config in ACPI if present +# +# FIXME: the machine driver uses .name = "dmic01" and .id=2 +# DMIC0 (id: 2) +DAI_CONFIG(DMIC, 0, 2, dmic01,
- DMIC_CONFIG(1, 500000, 4800000, 40, 60, 48000,
DMIC_WORD_LENGTH(s32le), DMIC, 0,PDM_CONFIG(DMIC, 0, FOUR_CH_PDM0_PDM1)))+# FIXME: the machine driver uses .name = "SSP5-Codec", .id=0, .cpu_name="SSP5 Pin" +#SSP 5 (ID: 0) with 19.2MHz mclk with MCLK_ID 0 +DAI_CONFIG(SSP, 5, 0, SSP5-Codec,
- SSP_CONFIG(I2S, SSP_CLOCK(mclk, 19200000,
codec_mclk_in),
SSP_CLOCK(bclk, 1920000, codec_slave),SSP_CLOCK(fsync, 48000, codec_slave),SSP_TDM(2, 20, 3, 3),SSP_CONFIG_DATA(SSP, 5, 16, 0)))+# FIXME: the machine driver uses .name = "SSP1-Codec", .id=1, .cpu_name="SSP1 Pin" +#SSP 1 (ID: 1) with 19.2MHz mclk with MCLK_ID 0 +DAI_CONFIG(SSP, 1, 1, SSP1-Codec,
- SSP_CONFIG(I2S, SSP_CLOCK(mclk, 19200000,
codec_mclk_in),
SSP_CLOCK(bclk, 1920000, codec_slave),SSP_CLOCK(fsync, 48000, codec_slave),SSP_TDM(2, 20, 3, 3),SSP_CONFIG_DATA(SSP, 1, 16, 0)))
Sound-open-firmware mailing list Sound-open-firmware@alsa-project.org http://mailman.alsa-project.org/mailman/listinfo/sound-open-fir mwar e
Sound-open-firmware mailing list Sound-open-firmware@alsa-project.org http://mailman.alsa-project.org/mailman/listinfo/sound-open-firmware
On Wed, 2018-07-11 at 22:13 -0500, Pierre-Louis Bossart wrote:
On 07/11/2018 08:14 PM, Ranjani Sridharan wrote:
On Wed, 2018-07-11 at 17:52 -0500, Pierre-Louis Bossart wrote:
On 07/11/2018 04:43 PM, Ranjani Sridharan wrote:
On Wed, 2018-07-11 at 11:55 -0500, Pierre-Louis Bossart wrote:
On 07/11/2018 01:23 AM, Ranjani Sridharan wrote:
On Tue, 2018-07-10 at 18:15 -0500, Pierre-Louis Bossart wrote: > Add based on sof-glk-codec. In theory this is the same > topology > with only a swap of SSPs > > FIXME: this is a mess between pipeline ID, PCM ID, DAI > ID, > DAI #. > I could use additional eyes to proof-read this...
Looks good to me. I couldnt spot any mistakes. Keyon, Could you please also check? > FIXME: not sure why we would need > PCM_CAPTURE_ADD/PCM_PLAYBACK_DAI, > this seems to be already done by DAI_ADD?
DAI_ADD only adds the BE DAI's. PCM_CAPTURE_ADD/PLAYBACK_ADD are needed for adding the FE dai links.
ok, but then why do those macros make references to the DAIs?
The dai_id is used when creating the cpu_dai for the PCM elems and their ID is set to the dai_id specified here.
I wish I understood cpu dai's better but I couldnt wrap my head around why we do that.
Keyon, any hints?
So it's a different dai_id from what is used for back-ends? Looks like a complete mess to me...
It is the same ID that is used for the BE DAI, Pierre. Here's a quick look at the all the DAI's that are registered in the case of up2 with the hifiberry dac+ with the test-playback- ssp5*.tplg.
The "Passthrough 0" dai is the last one and has the dai_id of 0 which is the same as the SSP5 id in the topology.
I just don't get all this. I started from the sof-glk-codec.m4 file and both that file and this one are NOT passthrough, they all contain a volume element which *should* decouple front-end from back-end.
[ 0.076806] NET: Registered protocol family 16 [ 0.176459] Registered efivars operations [ 0.319302] NET: Registered protocol family 2 [ 0.320794] NET: Registered protocol family 1 [ 1.467017] NET: Registered protocol family 10 [ 1.467826] NET: Registered protocol family 17 [ 3.038697] snd-soc-dummy snd-soc-dummy: ASoC: Registered DAI 'snd- soc-dummy-dai' with id 0 [ 3.227853] sof-audio sof-audio: ASoC: Registered platform 'sof- audio' [ 3.227862] sof-audio sof-audio: ASoC: Registered DAI 'SSP0 Pin' with id 0 [ 3.227866] sof-audio sof-audio: ASoC: Registered DAI 'SSP1 Pin' with id 1 [ 3.227869] sof-audio sof-audio: ASoC: Registered DAI 'SSP2 Pin' with id 2 [ 3.227873] sof-audio sof-audio: ASoC: Registered DAI 'SSP3 Pin' with id 3 [ 3.227876] sof-audio sof-audio: ASoC: Registered DAI 'SSP4 Pin' with id 4 [ 3.227879] sof-audio sof-audio: ASoC: Registered DAI 'SSP5 Pin' with id 5 [ 3.227883] sof-audio sof-audio: ASoC: Registered DAI 'DMIC01 Pin' with id 6 [ 3.227886] sof-audio sof-audio: ASoC: Registered DAI 'DMIC16k Pin' with id 7 [ 3.227889] sof-audio sof-audio: ASoC: Registered DAI 'iDisp1 Pin' with id 8 [ 3.227893] sof-audio sof-audio: ASoC: Registered DAI 'iDisp2 Pin' with id 9 [ 3.227896] sof-audio sof-audio: ASoC: Registered DAI 'iDisp3 Pin' with id 10 [ 3.227900] sof-audio sof-audio: ASoC: Registered DAI 'Analog Codec DAI' with id 11 [ 3.227903] sof-audio sof-audio: ASoC: Registered DAI 'Digital Codec DAI' with id 12 [ 3.227906] sof-audio sof-audio: ASoC: Registered DAI 'Alt Analog Codec DAI' with id 13 [ 3.578667] pcm512x i2c-104C5122:00: ASoC: Registered DAI 'pcm512x- hifi' with id 0 [ 4.141636] pstore: Registered efi as persistent store backend [ 4.203432] sof-audio sof-audio: ASoC: Registered DAI 'Passthrough 0' with id 0
And I don't get what this means either? What is a Passthrough DAI? it's just not self-explanatory.
Sadly, it is all a bit messed up with a lot of cut-copy-paste's. If it isnt urgent, I can work on this next week.
[ 4.609670] NET: Registered protocol family 31
+# PCM (PCM_CAPTURE_ADD(name, pipeline, pcm_id, dai_id, capture)) +#PCM_CAPTURE_ADD(DMIC0, 1, 0, 0, PIPELINE_PCM_1) +#PCM_PLAYBACK_ADD(SSP5, 2, 0, 5, PIPELINE_PCM_2) +#PCM_PLAYBACK_ADD(SSP1, 3, 1, 1, PIPELINE_PCM_3)
In addition the information on the PCM # is already provided in PIPELINE_ADD, so this looks redundant on both the FE and BE sides?
> Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart > @lin > ux.i > ntel > .com> > --- > topology/Makefile.am | 1 + > topology/sof-apl-da7219.m4 | 104 > +++++++++++++++++++++++++++++++++++++++++++++ > 2 files changed, 105 insertions(+) > create mode 100644 topology/sof-apl-da7219.m4 > > diff --git a/topology/Makefile.am b/topology/Makefile.am > index 33d0418..396b5b9 100644 > --- a/topology/Makefile.am > +++ b/topology/Makefile.am > @@ -31,6 +31,7 @@ MACHINES = \ > sof-hsw-rt5640.tplg \ > sof-apl-tdf8532.tplg \ > sof-apl-pcm512x.tplg \ > + sof-apl-da7219.tplg \ > sof-glk-codec.tplg > > # Uncomment the following line if you want to debug > conf > files > diff --git a/topology/sof-apl-da7219.m4 b/topology/sof- > apl- > da7219.m4 > new file mode 100644 > index 0000000..3934df5 > --- /dev/null > +++ b/topology/sof-apl-da7219.m4 > @@ -0,0 +1,104 @@ > +# > +# Topology for AppoloLake with headset on SSP1, spk on > SSP5 > and > DMIC+# PCM (PCM_CAPTURE_ADD(name, pipeline, pcm_id, > dai_id, > capture)) > +#PCM_CAPTURE_ADD(DMIC0, 1, 0, 0, PIPELINE_PCM_1) > +#PCM_PLAYBACK_ADD(SSP5, 2, 0, 5, PIPELINE_PCM_2) > +#PCM_PLAYBACK_ADD(SSP1, 3, 1, 1, PIPELINE_PCM_3) > capture > +# > + > +# Include topology builder > +include(`utils.m4') > +include(`dai.m4') > +include(`pipeline.m4') > +include(`ssp.m4') > + > +# Include TLV library > +include(`common/tlv.m4') > + > +# Include Token library > +include(`sof/tokens.m4') > + > +# Include bxt DSP configuration > +include(`platform/intel/bxt.m4') > +include(`platform/intel/dmic.m4') > + > +# > +# Define the pipelines > +# > +# PCM0 <---- volume <----- DMIC0 (dmic capture) > +# PCM0 ----> volume -----> SSP5 (speaker - maxim98357a) > +# PCM1 ----> volume -----> SSP1 (headset - da7219) > +# > + > +# Low Latency capture pipeline 1 on PCM 0 using max 4 > channels > of > s32le. > +# Schedule 48 frames per 1000us deadline on core 0 with > priority > 0 > +PIPELINE_PCM_ADD(sof/pipe-volume-capture.m4, > + 1, 0, 4, s32le, > + 48, 1000, 0, 0) > + > +# Low Latency playback pipeline 2 on PCM 0 using max 2 > channels > of > s16le. > +# Schedule 48 frames per 1000us deadline on core 0 with > priority > 0 > +PIPELINE_PCM_ADD(sof/pipe-volume-playback.m4, > + 2, 0, 2, s16le, > + 48, 1000, 0, 0) > + > +# Low Latency playback pipeline 3 on PCM 1 using max 2 > channels > of > s16le. > +# Schedule 48 frames per 1000us deadline on core 0 with > priority > 0 > +PIPELINE_PCM_ADD(sof/pipe-volume-playback.m4, > + 3, 1, 2, s16le, > + 48, 1000, 0, 0) > + > +# > +# DAIs configuration > +# > + > +# capture DAI is DMIC0 using 2 periods > +# Buffers use s16le format, with 48 frame per 1000us on > core > 0 > with > priority 0 > +DAI_ADD(sof/pipe-dai-capture.m4, > + 1, DMIC, 0, DMIC0, > + PIPELINE_SINK_1, 2, s32le, > + 48, 1000, 0, 0) > + > +# playback DAI is SSP5 using 2 periods > +# Buffers use s16le format, with 48 frame per 1000us on > core > 0 > with > priority 0 > +DAI_ADD(sof/pipe-dai-playback.m4, > + 2, SSP, 5, SSP5-Codec, > + PIPELINE_SOURCE_2, 2, s16le, > + 48, 1000, 0, 0) > + > +# playback DAI is SSP1 using 2 periods > +# Buffers use s16le format, with 48 frame per 1000us on > core > 0 > with > priority 0 > +DAI_ADD(sof/pipe-dai-playback.m4, > + 3, SSP, 1, SSP1-Codec, > + PIPELINE_SOURCE_3, 2, s16le, > + 48, 1000, 0, 0) > + > +# FIXME: Why is this needed? And is this correct? > +# PCM (PCM_CAPTURE_ADD(name, pipeline, pcm_id, dai_id, > capture)) > +#PCM_CAPTURE_ADD(DMIC0, 1, 0, 0, PIPELINE_PCM_1) > +#PCM_PLAYBACK_ADD(SSP5, 2, 0, 5, PIPELINE_PCM_2) > +#PCM_PLAYBACK_ADD(SSP1, 3, 1, 1, PIPELINE_PCM_3) > + > +# > +# BE configurations - overrides config in ACPI if > present > +# > +# FIXME: the machine driver uses .name = "dmic01" and > .id=2 > +# DMIC0 (id: 2) > +DAI_CONFIG(DMIC, 0, 2, dmic01, > + DMIC_CONFIG(1, 500000, 4800000, 40, 60, 48000, > + DMIC_WORD_LENGTH(s32le), DMIC, 0, > + PDM_CONFIG(DMIC, 0, FOUR_CH_PDM0_PDM1))) > + > +# FIXME: the machine driver uses .name = "SSP5-Codec", > .id=0, > .cpu_name="SSP5 Pin" > +#SSP 5 (ID: 0) with 19.2MHz mclk with MCLK_ID 0 > +DAI_CONFIG(SSP, 5, 0, SSP5-Codec, > + SSP_CONFIG(I2S, SSP_CLOCK(mclk, 19200000, > codec_mclk_in), > + SSP_CLOCK(bclk, 1920000, codec_slave), > + SSP_CLOCK(fsync, 48000, codec_slave), > + SSP_TDM(2, 20, 3, 3), > + SSP_CONFIG_DATA(SSP, 5, 16, 0))) > + > +# FIXME: the machine driver uses .name = "SSP1-Codec", > .id=1, > .cpu_name="SSP1 Pin" > +#SSP 1 (ID: 1) with 19.2MHz mclk with MCLK_ID 0 > +DAI_CONFIG(SSP, 1, 1, SSP1-Codec, > + SSP_CONFIG(I2S, SSP_CLOCK(mclk, 19200000, > codec_mclk_in), > + SSP_CLOCK(bclk, 1920000, codec_slave), > + SSP_CLOCK(fsync, 48000, codec_slave), > + SSP_TDM(2, 20, 3, 3), > + SSP_CONFIG_DATA(SSP, 1, 16, 0)))
Sound-open-firmware mailing list Sound-open-firmware@alsa-project.org http://mailman.alsa-project.org/mailman/listinfo/sound-open -fir mwar e
Sound-open-firmware mailing list Sound-open-firmware@alsa-project.org http://mailman.alsa-project.org/mailman/listinfo/sound-open-firmw are
Sound-open-firmware mailing list Sound-open-firmware@alsa-project.org http://mailman.alsa-project.org/mailman/listinfo/sound-open-firmware
On 2018年07月12日 09:14, Ranjani Sridharan wrote:
On Wed, 2018-07-11 at 17:52 -0500, Pierre-Louis Bossart wrote:
On 07/11/2018 04:43 PM, Ranjani Sridharan wrote:
On Wed, 2018-07-11 at 11:55 -0500, Pierre-Louis Bossart wrote:
On 07/11/2018 01:23 AM, Ranjani Sridharan wrote:
On Tue, 2018-07-10 at 18:15 -0500, Pierre-Louis Bossart wrote:
Add based on sof-glk-codec. In theory this is the same topology with only a swap of SSPs
FIXME: this is a mess between pipeline ID, PCM ID, DAI ID, DAI #. I could use additional eyes to proof-read this...
Looks good to me. I couldnt spot any mistakes. Keyon, Could you please also check?
FIXME: not sure why we would need PCM_CAPTURE_ADD/PCM_PLAYBACK_DAI, this seems to be already done by DAI_ADD?
DAI_ADD only adds the BE DAI's. PCM_CAPTURE_ADD/PLAYBACK_ADD are needed for adding the FE dai links.
ok, but then why do those macros make references to the DAIs?
The dai_id is used when creating the cpu_dai for the PCM elems and their ID is set to the dai_id specified here.
I wish I understood cpu dai's better but I couldnt wrap my head around why we do that.
Keyon, any hints?
So it's a different dai_id from what is used for back-ends? Looks like a complete mess to me...
It is the same ID that is used for the BE DAI, Pierre. Here's a quick look at the all the DAI's that are registered in the case of up2 with the hifiberry dac+ with the test-playback-ssp5*.tplg.
The "Passthrough 0" dai is the last one and has the dai_id of 0 which is the same as the SSP5 id in the topology.
[ 0.076806] NET: Registered protocol family 16 [ 0.176459] Registered efivars operations [ 0.319302] NET: Registered protocol family 2 [ 0.320794] NET: Registered protocol family 1 [ 1.467017] NET: Registered protocol family 10 [ 1.467826] NET: Registered protocol family 17 [ 3.038697] snd-soc-dummy snd-soc-dummy: ASoC: Registered DAI 'snd- soc-dummy-dai' with id 0 [ 3.227853] sof-audio sof-audio: ASoC: Registered platform 'sof- audio' [ 3.227862] sof-audio sof-audio: ASoC: Registered DAI 'SSP0 Pin' with id 0 [ 3.227866] sof-audio sof-audio: ASoC: Registered DAI 'SSP1 Pin' with id 1 [ 3.227869] sof-audio sof-audio: ASoC: Registered DAI 'SSP2 Pin' with id 2 [ 3.227873] sof-audio sof-audio: ASoC: Registered DAI 'SSP3 Pin' with id 3 [ 3.227876] sof-audio sof-audio: ASoC: Registered DAI 'SSP4 Pin' with id 4 [ 3.227879] sof-audio sof-audio: ASoC: Registered DAI 'SSP5 Pin' with id 5 [ 3.227883] sof-audio sof-audio: ASoC: Registered DAI 'DMIC01 Pin' with id 6 [ 3.227886] sof-audio sof-audio: ASoC: Registered DAI 'DMIC16k Pin' with id 7 [ 3.227889] sof-audio sof-audio: ASoC: Registered DAI 'iDisp1 Pin' with id 8 [ 3.227893] sof-audio sof-audio: ASoC: Registered DAI 'iDisp2 Pin' with id 9 [ 3.227896] sof-audio sof-audio: ASoC: Registered DAI 'iDisp3 Pin' with id 10 [ 3.227900] sof-audio sof-audio: ASoC: Registered DAI 'Analog Codec DAI' with id 11 [ 3.227903] sof-audio sof-audio: ASoC: Registered DAI 'Digital Codec DAI' with id 12
These are all BE cpu_dais.
[ 3.227906] sof-audio sof-audio: ASoC: Registered DAI 'Alt Analog Codec DAI' with id 1 [ 3.578667] pcm512x i2c-104C5122:00: ASoC: Registered DAI 'pcm512x- hifi' with id 0 [ 4.141636] pstore: Registered efi as persistent store backend [ 4.203432] sof-audio sof-audio: ASoC: Registered DAI 'Passthrough 0' with id 0
What's this? it should be FE cpu_dai IMO.
~Keyon
[ 4.609670] NET: Registered protocol family 31
+# PCM (PCM_CAPTURE_ADD(name, pipeline, pcm_id, dai_id, capture)) +#PCM_CAPTURE_ADD(DMIC0, 1, 0, 0, PIPELINE_PCM_1) +#PCM_PLAYBACK_ADD(SSP5, 2, 0, 5, PIPELINE_PCM_2) +#PCM_PLAYBACK_ADD(SSP1, 3, 1, 1, PIPELINE_PCM_3)
In addition the information on the PCM # is already provided in PIPELINE_ADD, so this looks redundant on both the FE and BE sides?
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@lin ux.i ntel .com>
topology/Makefile.am | 1 + topology/sof-apl-da7219.m4 | 104+++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 105 insertions(+) create mode 100644 topology/sof-apl-da7219.m4
diff --git a/topology/Makefile.am b/topology/Makefile.am index 33d0418..396b5b9 100644 --- a/topology/Makefile.am +++ b/topology/Makefile.am @@ -31,6 +31,7 @@ MACHINES = \ sof-hsw-rt5640.tplg \ sof-apl-tdf8532.tplg \ sof-apl-pcm512x.tplg \
sof-apl-da7219.tplg \ sof-glk-codec.tplg
# Uncomment the following line if you want to debug conf
files diff --git a/topology/sof-apl-da7219.m4 b/topology/sof-apl- da7219.m4 new file mode 100644 index 0000000..3934df5 --- /dev/null +++ b/topology/sof-apl-da7219.m4 @@ -0,0 +1,104 @@ +# +# Topology for AppoloLake with headset on SSP1, spk on SSP5 and DMIC+# PCM (PCM_CAPTURE_ADD(name, pipeline, pcm_id, dai_id, capture)) +#PCM_CAPTURE_ADD(DMIC0, 1, 0, 0, PIPELINE_PCM_1) +#PCM_PLAYBACK_ADD(SSP5, 2, 0, 5, PIPELINE_PCM_2) +#PCM_PLAYBACK_ADD(SSP1, 3, 1, 1, PIPELINE_PCM_3) capture +#
+# Include topology builder +include(`utils.m4') +include(`dai.m4') +include(`pipeline.m4') +include(`ssp.m4')
+# Include TLV library +include(`common/tlv.m4')
+# Include Token library +include(`sof/tokens.m4')
+# Include bxt DSP configuration +include(`platform/intel/bxt.m4') +include(`platform/intel/dmic.m4')
+# +# Define the pipelines +# +# PCM0 <---- volume <----- DMIC0 (dmic capture) +# PCM0 ----> volume -----> SSP5 (speaker - maxim98357a) +# PCM1 ----> volume -----> SSP1 (headset - da7219) +#
+# Low Latency capture pipeline 1 on PCM 0 using max 4 channels of s32le. +# Schedule 48 frames per 1000us deadline on core 0 with priority 0 +PIPELINE_PCM_ADD(sof/pipe-volume-capture.m4,
- 1, 0, 4, s32le,
- 48, 1000, 0, 0)
+# Low Latency playback pipeline 2 on PCM 0 using max 2 channels of s16le. +# Schedule 48 frames per 1000us deadline on core 0 with priority 0 +PIPELINE_PCM_ADD(sof/pipe-volume-playback.m4,
- 2, 0, 2, s16le,
- 48, 1000, 0, 0)
+# Low Latency playback pipeline 3 on PCM 1 using max 2 channels of s16le. +# Schedule 48 frames per 1000us deadline on core 0 with priority 0 +PIPELINE_PCM_ADD(sof/pipe-volume-playback.m4,
- 3, 1, 2, s16le,
- 48, 1000, 0, 0)
+# +# DAIs configuration +#
+# capture DAI is DMIC0 using 2 periods +# Buffers use s16le format, with 48 frame per 1000us on core 0 with priority 0 +DAI_ADD(sof/pipe-dai-capture.m4,
- 1, DMIC, 0, DMIC0,
- PIPELINE_SINK_1, 2, s32le,
- 48, 1000, 0, 0)
+# playback DAI is SSP5 using 2 periods +# Buffers use s16le format, with 48 frame per 1000us on core 0 with priority 0 +DAI_ADD(sof/pipe-dai-playback.m4,
- 2, SSP, 5, SSP5-Codec,
- PIPELINE_SOURCE_2, 2, s16le,
- 48, 1000, 0, 0)
+# playback DAI is SSP1 using 2 periods +# Buffers use s16le format, with 48 frame per 1000us on core 0 with priority 0 +DAI_ADD(sof/pipe-dai-playback.m4,
- 3, SSP, 1, SSP1-Codec,
- PIPELINE_SOURCE_3, 2, s16le,
- 48, 1000, 0, 0)
+# FIXME: Why is this needed? And is this correct? +# PCM (PCM_CAPTURE_ADD(name, pipeline, pcm_id, dai_id, capture)) +#PCM_CAPTURE_ADD(DMIC0, 1, 0, 0, PIPELINE_PCM_1) +#PCM_PLAYBACK_ADD(SSP5, 2, 0, 5, PIPELINE_PCM_2) +#PCM_PLAYBACK_ADD(SSP1, 3, 1, 1, PIPELINE_PCM_3)
+# +# BE configurations - overrides config in ACPI if present +# +# FIXME: the machine driver uses .name = "dmic01" and .id=2 +# DMIC0 (id: 2) +DAI_CONFIG(DMIC, 0, 2, dmic01,
- DMIC_CONFIG(1, 500000, 4800000, 40, 60, 48000,
DMIC_WORD_LENGTH(s32le), DMIC, 0,PDM_CONFIG(DMIC, 0, FOUR_CH_PDM0_PDM1)))+# FIXME: the machine driver uses .name = "SSP5-Codec", .id=0, .cpu_name="SSP5 Pin" +#SSP 5 (ID: 0) with 19.2MHz mclk with MCLK_ID 0 +DAI_CONFIG(SSP, 5, 0, SSP5-Codec,
- SSP_CONFIG(I2S, SSP_CLOCK(mclk, 19200000,
codec_mclk_in),
SSP_CLOCK(bclk, 1920000, codec_slave),SSP_CLOCK(fsync, 48000, codec_slave),SSP_TDM(2, 20, 3, 3),SSP_CONFIG_DATA(SSP, 5, 16, 0)))+# FIXME: the machine driver uses .name = "SSP1-Codec", .id=1, .cpu_name="SSP1 Pin" +#SSP 1 (ID: 1) with 19.2MHz mclk with MCLK_ID 0 +DAI_CONFIG(SSP, 1, 1, SSP1-Codec,
- SSP_CONFIG(I2S, SSP_CLOCK(mclk, 19200000,
codec_mclk_in),
SSP_CLOCK(bclk, 1920000, codec_slave),SSP_CLOCK(fsync, 48000, codec_slave),SSP_TDM(2, 20, 3, 3),SSP_CONFIG_DATA(SSP, 1, 16, 0)))
Sound-open-firmware mailing list Sound-open-firmware@alsa-project.org http://mailman.alsa-project.org/mailman/listinfo/sound-open-fir mwar e
Sound-open-firmware mailing list Sound-open-firmware@alsa-project.org http://mailman.alsa-project.org/mailman/listinfo/sound-open-firmware
participants (3)
-
Keyon Jie -
Pierre-Louis Bossart -
Ranjani Sridharan