- Alsa-devel - mailman.alsa-project.org

DMA Transfer Synchronization Issue in Out-of-Tree Sound Card Driver
by Louis Chauvet 24 May '24

24 May '24

Hello everyone, I am currently developing an out-of-tree driver for a sound card that utilizes XDMA for sample transfers. I am currently using a kernel 6.5 with the latest xdma driver cherry-picked, but I don't think any changes since 6.5 is addressing my issue. My initial issue pertains to the completion of DMA transfers between a start and a stop command in ALSA. If the interval is too brief, the transfer does not conclude properly, leading to distorted samples. A straightforward solution to this problem was to adequately wait for the transfer to finish upon the stop, ie. sleeping until we know the hardware is done with the transfert (the XDMA controller does not support stopping in the middle of a transfer). To address this DMA issue, I have created a patch [1] that guarantees the completion of the DMA transfer upon the return of xdma_synchronize. This means xdma_synchronize now sleeps, but looking at other drivers around it appears expected to be able to do so. Regarding the audio implementation, the following patch enforces the synchronization: int playback_trigger(struct snd_pcm_substream *substream, int command) { struct my_dev *my_dev = snd_pcm_substream_chip(substream); switch (command) { case SNDRV_PCM_TRIGGER_START: /* Synchronize on start, because the trigger stop is called from an IRQ context */ if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK) dmaengine_synchronize(my_dev->playback_dma_chan); pipe_start(&my_dev->playback, substream); break; case SNDRV_PCM_TRIGGER_STOP: pipe_stop(&my_dev->playback, substream); break; default: return -EINVAL; } return 0; } In order for a sleepable function like dmaengine_synchronize() to work in the trigger callbacks, the audio card nonatomic flag had to be set. It basically leads the sound core towards the use of a mutex instead of a spinlock. static int probe([...]) { struct snd_pcm *pcm; [...] /* This flag is needed to be able to sleep in start/stop callbacks */ pcm->nonatomic = true; [...] snd_pcm_set_managed_buffer_all(pcm, [...]); } This approach generally works well, but leads to "scheduling while atomic" errors. Indeed, the IRQ handler from the XDMA driver invokes a function within the sound subsystem, which subsequently acquires a mutex... At the moment, the only solution I've found is to replace the wait_for_completion() in the XDMA driver [2] with a busy wait loop. However, this approach seems incorrect, as all other synchronization functions in other DMA drivers are sleeping, which should not cause an issue. The problem might be related to the sound driver. Should I avoid manually using dmaengine_synchronize? How to achieve the same effect in this case? Perhaps there is a more traditional way to properly clean the stream in the sound subsystem which I overlooked? Could someone please provide guidance on how to resolve this issue? Thanks, Louis Chauvet [1]: https://lore.kernel.org/all/20240327-digigram-xdma-fixes-v1-2-45f4a52c0283@… [2]: https://elixir.bootlin.com/linux/latest/source/drivers/dma/xilinx/xdma.c#L5… -- Louis Chauvet, Bootlin Embedded Linux and Kernel engineering https://bootlin.com

3 3

[PATCH v4 0/3] ASoC: grace time for DPCM cleanup
by Kuninori Morimoto 24 May '24

24 May '24

Hi Mark, Pierre-Louis, Jerome Cc each ASoC driver maintainer This is v4 of DPCM cleanup As we discussed in [1], we don't need to use dpcm_playback/capture flag, so we remove it. But we have been using it for 10 years, some driver might get damage. The most likely case is that the device/driver can use both playback/capture, but have only one flag, and not using xxx_only flag. [1/3] patch indicates warning in such case. These adds grace time for DPCM cleanup. I'm not sure when dpcm_xxx will be removed, and Codec check bypass will be error, but maybe v6.11 or v6.12 ? Please check each driver by that time. Previous patch-set try to check both CPU and Codec in DPCM, but we noticed that there are some special DAI which we can't handle today [2]. So I will escape it in this patch-set. [1] https://lore.kernel.org/r/87edaym2cg.wl-kuninori.morimoto.gx@renesas.com [2] https://lore.kernel.org/all/3e67d62d-fe08-4f55-ab5b-ece8a57154f9@linux.inte… Link: https://lore.kernel.org/r/87edaym2cg.wl-kuninori.morimoto.gx@renesas.com Link: https://lore.kernel.org/r/87wmo6dyxg.wl-kuninori.morimoto.gx@renesas.com Link: https://lore.kernel.org/r/87msole5wc.wl-kuninori.morimoto.gx@renesas.com v3 -> v4 - don't check Codec on DPCM - include Jerome's dpcm_xxx update patch v2 -> v3 - tidyup typo (reuqsts -> requests) - add Tested-by on git-log v1 -> v2 - tidyup Codec check warning output condition Jerome Brunet (1): ASoC: amlogic: do not use dpcm_playback/capture flags Kuninori Morimoto (2): ASoC: soc-pcm: Indicate warning if dpcm_playback/capture were used for availability limition ASoC: remove snd_soc_dai_link_set_capabilities() include/sound/soc-dai.h | 1 - include/sound/soc.h | 1 + sound/soc/fsl/imx-card.c | 3 -- sound/soc/generic/audio-graph-card.c | 2 - sound/soc/generic/audio-graph-card2.c | 2 - sound/soc/generic/simple-card.c | 2 - sound/soc/meson/axg-card.c | 11 +++-- sound/soc/meson/gx-card.c | 1 - sound/soc/meson/meson-card-utils.c | 4 +- sound/soc/qcom/common.c | 1 - sound/soc/soc-dai.c | 38 ---------------- sound/soc/soc-pcm.c | 65 +++++++++++++++------------ 12 files changed, 45 insertions(+), 86 deletions(-) -- 2.25.1

2 5

[PATCH v3 0/3] ASoC: grace time for DPCM cleanup
by Kuninori Morimoto 24 May '24

24 May '24

Hi Mark, Pierre-Louis, Jerome Cc each ASoC driver maintainer This is v3 of DPCM cleanup As we discussed in [1], we don't need to use dpcm_playback/capture flag, so we remove it. But we have been using it for 10 years, some driver might get damage. The most likely case is that the device/driver can use both playback/capture, but have only one flag, and not using xxx_only flag. [1/3] patch indicates warning in such case. And because of its history, DPCM has been checking CPU side only. But it should check Codec side too same as non-DPCM. Some device/driver has been bypassed this check. It should be error. [2/3] patch indicates warning in such case. Because dpcm_xxx flag is no longer used by [1/3] patch, snd_soc_dai_link_set_capabilities() is no longer needed. [3/3] patch remove it. These adds grace time for DPCM cleanup. I'm not sure when dpcm_xxx will be removed, and Codec check bypass will be error, but maybe v6.11 or v6.12 ? Please check each driver by that time. [1] https://lore.kernel.org/r/87edaym2cg.wl-kuninori.morimoto.gx@renesas.com Link: https://lore.kernel.org/r/87edaym2cg.wl-kuninori.morimoto.gx@renesas.com Link: https://lore.kernel.org/r/87wmo6dyxg.wl-kuninori.morimoto.gx@renesas.com v2 -> v3 - tidyup typo (reuqsts -> requests) - add Tested-by on git-log v1 -> v2 - tidyup Codec check warning output condition Kuninori Morimoto (3): ASoC: soc-pcm: Indicate warning if dpcm_playback/capture were used for availability limition ASoC: soc-pcm: Indicate warning if CPU / Codec availability mismatch ASoC: remove snd_soc_dai_link_set_capabilities() include/sound/soc-dai.h | 1 - include/sound/soc.h | 1 + sound/soc/fsl/imx-card.c | 3 - sound/soc/generic/audio-graph-card.c | 2 - sound/soc/generic/audio-graph-card2.c | 2 - sound/soc/generic/simple-card.c | 2 - sound/soc/meson/axg-card.c | 1 - sound/soc/meson/gx-card.c | 1 - sound/soc/qcom/common.c | 1 - sound/soc/soc-dai.c | 38 --------- sound/soc/soc-pcm.c | 110 +++++++++++++++++++------- 11 files changed, 81 insertions(+), 81 deletions(-) -- 2.25.1

3 17

[GIT PULL] ASoC fixes for v6.10-merge-window
by Mark Brown 23 May '24

23 May '24

The following changes since commit c3c5ac4bd7d7019f2e3ad1720572d53226fe656e: ASoC: Intel: updates for 6.10 - part7 (2024-05-10 12:30:12 +0100) are available in the Git repository at: https://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound.git tags/asoc-fix-v6.10-merge-window for you to fetch changes up to 737ce4fb96206f999ddea7530145fc0e8abd5d31: ASoC: dt-bindings: stm32: Ensure compatible pattern matches whole string (2024-05-21 12:35:12 +0100) ---------------------------------------------------------------- ASoC: Fixes for v6.10 A bunch of fixes that came in during the merge window, all driver specific and none of them especially remarkable. ---------------------------------------------------------------- Jack Yu (1): ASoC: rt5645: mic-in detection threshold modification Peter Ujfalusi (1): ASoC: Intel: sof_sdw_rt_sdca_jack_common: Use name_prefix for `-sdca` detection Rob Herring (Arm) (1): ASoC: dt-bindings: stm32: Ensure compatible pattern matches whole string Shenghao Ding (2): ASoC: tas2552: Add TX path for capturing AUDIO-OUT data ASoC: tas2781: Fix wrong loading calibrated data sequence Zhang Yi (1): ASoC: codecs: ES8326: solve hp and button detect issue .../devicetree/bindings/sound/st,stm32-sai.yaml | 2 +- include/sound/tas2781-dsp.h | 7 +- sound/soc/codecs/es8326.c | 6 +- sound/soc/codecs/rt5645.c | 1 + sound/soc/codecs/tas2552.c | 15 ++- sound/soc/codecs/tas2781-fmwlib.c | 103 ++++++--------------- sound/soc/codecs/tas2781-i2c.c | 4 +- .../soc/intel/boards/sof_sdw_rt_sdca_jack_common.c | 2 +- 8 files changed, 51 insertions(+), 89 deletions(-)

2 1

[PATCH v1] ASoc: tas2781: Support still work when only RCA binary loading well without dsp firmware loading.
by Shenghao Ding 23 May '24

23 May '24

In only RCA binary loading case, only default dsp program will be work inside the chip. Signed-off-by: Shenghao Ding <shenghao-ding(a)ti.com> --- v1: - Split out the different logical changes into different patches. - rename tasdevice_dsp_fw_state -> tasdevice_fw_state, the fw are not only dsp fw, but also RCA(Reconfigurable data, such as acoustic data and register setting, etc). - Add TASDEVICE_RCA_FW_OK in tasdevice_fw_state to identify the state that only RCA binary file has been download successfully, but dsp fw is not loaded or loading failure. - Add the this strategy into tasdevice_tuning_switch. - If one side of the if/else has a braces both should in tasdevice_tuning_switch. - Identify whehter both RCA and DSP have been loaded or only RCA has been loaded in tasdevice_fw_ready. - Add check fw load status in tasdevice_startup. - remove ret in tasdevice_startup to make th code neater. --- include/sound/tas2781-dsp.h | 3 ++- sound/soc/codecs/tas2781-fmwlib.c | 14 ++++++++++---- sound/soc/codecs/tas2781-i2c.c | 30 ++++++++++++++++++------------ 3 files changed, 30 insertions(+), 17 deletions(-) diff --git a/include/sound/tas2781-dsp.h b/include/sound/tas2781-dsp.h index 7fba7ea26a4b..92d68ca5eafb 100644 --- a/include/sound/tas2781-dsp.h +++ b/include/sound/tas2781-dsp.h @@ -117,10 +117,11 @@ struct tasdevice_fw { struct device *dev; }; -enum tasdevice_dsp_fw_state { +enum tasdevice_fw_state { TASDEVICE_DSP_FW_NONE = 0, TASDEVICE_DSP_FW_PENDING, TASDEVICE_DSP_FW_FAIL, + TASDEVICE_RCA_FW_OK, TASDEVICE_DSP_FW_ALL_OK, }; diff --git a/sound/soc/codecs/tas2781-fmwlib.c b/sound/soc/codecs/tas2781-fmwlib.c index 265a8ca25cbb..cfa022ef4a59 100644 --- a/sound/soc/codecs/tas2781-fmwlib.c +++ b/sound/soc/codecs/tas2781-fmwlib.c @@ -2324,13 +2324,18 @@ void tasdevice_tuning_switch(void *context, int state) struct tasdevice_fw *tas_fmw = tas_priv->fmw; int profile_cfg_id = tas_priv->rcabin.profile_cfg_id; - if (tas_priv->fw_state == TASDEVICE_DSP_FW_FAIL) { - dev_err(tas_priv->dev, "DSP bin file not loaded\n"); + /* + * Only RCA file loaded can still work with default dsp program inside + * the chip? + */ + if (!(tas_priv->fw_state == TASDEVICE_RCA_FW_OK || + tas_priv->fw_state == TASDEVICE_DSP_FW_ALL_OK)) { + dev_err(tas_priv->dev, "No firmware loaded\n"); return; } if (state == 0) { - if (tas_priv->cur_prog < tas_fmw->nr_programs) { + if (tas_fmw && tas_priv->cur_prog < tas_fmw->nr_programs) { /*dsp mode or tuning mode*/ profile_cfg_id = tas_priv->rcabin.profile_cfg_id; tasdevice_select_tuningprm_cfg(tas_priv, @@ -2340,9 +2345,10 @@ void tasdevice_tuning_switch(void *context, int state) tasdevice_select_cfg_blk(tas_priv, profile_cfg_id, TASDEVICE_BIN_BLK_PRE_POWER_UP); - } else + } else { tasdevice_select_cfg_blk(tas_priv, profile_cfg_id, TASDEVICE_BIN_BLK_PRE_SHUTDOWN); + } } EXPORT_SYMBOL_NS_GPL(tasdevice_tuning_switch, SND_SOC_TAS2781_FMWLIB); diff --git a/sound/soc/codecs/tas2781-i2c.c b/sound/soc/codecs/tas2781-i2c.c index 9350972dfefe..ccb9313ada9b 100644 --- a/sound/soc/codecs/tas2781-i2c.c +++ b/sound/soc/codecs/tas2781-i2c.c @@ -380,23 +380,30 @@ static void tasdevice_fw_ready(const struct firmware *fmw, mutex_lock(&tas_priv->codec_lock); ret = tasdevice_rca_parser(tas_priv, fmw); - if (ret) + if (ret) { + tasdevice_config_info_remove(tas_priv); goto out; + } tasdevice_create_control(tas_priv); tasdevice_dsp_remove(tas_priv); tasdevice_calbin_remove(tas_priv); - tas_priv->fw_state = TASDEVICE_DSP_FW_PENDING; + tas_priv->fw_state = TASDEVICE_RCA_FW_OK; scnprintf(tas_priv->coef_binaryname, 64, "%s_coef.bin", tas_priv->dev_name); + ret = tasdevice_dsp_parser(tas_priv); if (ret) { dev_err(tas_priv->dev, "dspfw load %s error\n", tas_priv->coef_binaryname); - tas_priv->fw_state = TASDEVICE_DSP_FW_FAIL; goto out; } - tasdevice_dsp_create_ctrls(tas_priv); + + ret = tasdevice_dsp_create_ctrls(tas_priv); + if (ret) { + dev_err(tas_priv->dev, "dsp controls error\n"); + goto out; + } tas_priv->fw_state = TASDEVICE_DSP_FW_ALL_OK; @@ -417,9 +424,8 @@ static void tasdevice_fw_ready(const struct firmware *fmw, tasdevice_prmg_load(tas_priv, 0); tas_priv->cur_prog = 0; out: - if (tas_priv->fw_state == TASDEVICE_DSP_FW_FAIL) { - /*If DSP FW fail, kcontrol won't be created */ - tasdevice_config_info_remove(tas_priv); + if (tas_priv->fw_state == TASDEVICE_RCA_FW_OK) { + /*If DSP FW fail, DSP kcontrol won't be created */ tasdevice_dsp_remove(tas_priv); } mutex_unlock(&tas_priv->codec_lock); @@ -466,14 +472,14 @@ static int tasdevice_startup(struct snd_pcm_substream *substream, { struct snd_soc_component *codec = dai->component; struct tasdevice_priv *tas_priv = snd_soc_component_get_drvdata(codec); - int ret = 0; - if (tas_priv->fw_state != TASDEVICE_DSP_FW_ALL_OK) { - dev_err(tas_priv->dev, "DSP bin file not loaded\n"); - ret = -EINVAL; + if (!(tas_priv->fw_state == TASDEVICE_DSP_FW_ALL_OK || + tas_priv->fw_state == TASDEVICE_RCA_FW_OK)) { + dev_err(tas_priv->dev, "Bin file not loaded\n"); + return -EINVAL; } - return ret; + return 0; } static int tasdevice_hw_params(struct snd_pcm_substream *substream, -- 2.34.1

1 0

[PATCH v2 0/2] Add PDM controller for StarFive JH8100 SoC
by Xingyu Wu 23 May '24

23 May '24

The Pulse Density Modulation (PDM) controller is a digital PDM microphone interface controller and decoder that supports both mono/stereo PDM format, and a Inter-IC Sound (I2S) transmitter that outputs standard stereo audio data to another device. The PDM controller includes two PDM blocks, each PDM block can drive one bitstream sampling clock and two bitstream coming data (mono/stereo) with sampling clock rising and falling edge. The first patch adds documentation to describe PDM controller bindings. And the second patch adds PDM driver for the StarFive JH8100 SoC. The DTS patch of PDM controller will be submitted after the patchs of JH8100 DTS are accepted. Changes since v1: - Changed the subject prefix of the bindings. - Dropped the confused property in the bindings. - Added the $ref to dai-common. - Changed the kcontrol names and put it into ops. - Added the reset when device resume. - Modified two blocks into a whole PDM which supports up to 4 channels of recorder. v1: https://lore.kernel.org/all/20240307033708.139535-1-xingyu.wu@starfivetech.… Xingyu Wu (2): ASoC: dt-bindings: Add PDM controller for the StarFive JH8100 SoC ASoC: starfive: Add PDM controller support .../bindings/sound/starfive,jh8100-pdm.yaml | 82 ++++ MAINTAINERS | 7 + sound/soc/starfive/Kconfig | 7 + sound/soc/starfive/Makefile | 1 + sound/soc/starfive/jh8100_pdm.c | 447 ++++++++++++++++++ 5 files changed, 544 insertions(+) create mode 100644 Documentation/devicetree/bindings/sound/starfive,jh8100-pdm.yaml create mode 100644 sound/soc/starfive/jh8100_pdm.c -- 2.25.1

3 9

topology: pre-processor: Introduce extends/overrides keywords for classes
by GitHub pull_request - edited 23 May '24

23 May '24

alsa-project/alsa-utils pull request #268 was edited from ranj063: Introduce a new keyword "Subtreecopy" for extending existing conf nodes with additional nodes. This feature is useful for extending previous pipeline class definitions with the addition of one or more widgets without having to duplicate everything in the new class definition. For example: Consider a pipeline class definition as below. Note that only the widgets & routes are shown here. ``` Class.Pipeline.mixout-gain-dai-copier-playback { Object.Widget { mixout."1" {} dai-copier."1" {} gain."1" {} pipeline."1" {} } Object.Base { !route [ { source mixout.$index.1 sink gain.$index.1 } ] } } ``` If we want to extend this pipeline with the addition of an eqiir/eqfir, we can create a Subtreecopy node with type extend as follows: ``` Subtreecopy.mixout-gain-eqiir-eqfir-dai-copier-playback { source "Class.Pipeline.mixout-gain-dai-copier-playback" target "Class.Pipeline" type extend Object.Widget { eqiir.1 {} eqfir.1 {} } Object.Base { !route [ { source gain.$index.1 sink eqiir.$index.1 } { source eqiir.$index.1 sink eqfir.$index.1 } ] } } ``` But if we want to modify an existing pipeline class while modifying the order of widgets and/or inserting new widgets, we should use the type "override" instead. This allows for adding new widgets to the list of widgets in the base class definition while also allowing overriding the routes to allow inserting the new widgets and reordering the widgets in the base class. For example, if we want to add a drc widget between the gain and the eqiir modules in the above class, we can do the following: ``` Subtreecopy.mixout-efx-dai-copier-playback { source "Class.Pipeline.mixout-gain-eqiir-eqfir-dai-copier-playback" target "Class.Pipeline" type override Object.Widget { drc.1 {} } Object.Base { !route [ { source mixout.$index.1 sink gain.$index.1 } { source gain.$index.1 sink drc.$index.1 } { source drc.$index.1 sink eqiir.$index.1 } { source eqiir.$index.1 sink eqfir.$index.1 } ] } } ``` Note that the routes array contains all the routes in the new subtreecopy definition. This subtreecopy feature can also be used to copy normal config blocks without having the need to duplicate information. For example, if all the widgets in a pipeline support the same audio formats, we can define it as follows: ``` Class.Pipeline.dai-copier-eqiir-module-copier-capture { formats { num_input_audio_formats 2 num_output_audio_formats 2 Object.Base.input_audio_format [ { in_bit_depth 32 in_valid_bit_depth 32 } { in_channels 4 in_bit_depth 32 in_valid_bit_depth 32 in_ch_cfg $CHANNEL_CONFIG_3_POINT_1 in_ch_map $CHANNEL_MAP_3_POINT_1 } ] Object.Base.output_audio_format [ { out_bit_depth 32 out_valid_bit_depth 32 } { out_channels 4 out_bit_depth 32 out_valid_bit_depth 32 out_ch_cfg $CHANNEL_CONFIG_3_POINT_1 out_ch_map $CHANNEL_MAP_3_POINT_1 } ] } Object.Widget { dai-copier."1" { type dai_out num_output_pins 1 Subtreecopy.formats { source "Class.Pipeline.dai-copier-eqiir-module-copier-capture.formats" type override } } eqiir."1" { Subtreecopy.formats { source "Class.Pipeline.dai-copier-eqiir-module-copier-capture.formats" type override } Object.Control.bytes."1" { IncludeByKey.DMIC0_DAI_EQIIR { "passthrough" "include/components/eqiir/passthrough.conf" "highpass_40hz_0db" "include/components/eqiir/highpass_40hz_0db_48khz.conf" "highpass_40hz_20db" "include/components/eqiir/highpass_40hz_20db_48khz.conf" } } } module-copier."2" { Subtreecopy.formats { source "Class.Pipeline.dai-copier-eqiir-module-copier-capture.formats" type override } } pipeline."1" { priority 0 lp_mode 0 } } } ``` The subtreecopy node in each widget ensures that the pipeline formats are applied to all widgets without the need for duplicating them for each widget. Link: https://github.com/thesofproject/sof/issues/9082 Request URL : https://github.com/alsa-project/alsa-utils/pull/268 Patch URL : https://github.com/alsa-project/alsa-utils/pull/268.patch Repository URL: https://github.com/alsa-project/alsa-utils

1 0

topology: pre-processor: Introduce extends/overrides keywords for classes
by GitHub pull_request - edited 23 May '24

23 May '24

alsa-project/alsa-utils pull request #268 was edited from ranj063: Introduce a new keyword "Subtreecopy" for extending existing conf nodes with additional nodes. This feature is useful for extending previous pipeline class definitions with the addition of one or more widgets without having to duplicate everything in the new class definition. For example: Consider a pipeline class definition as below. Note that only the widgets & routes are shown here. Class.Pipeline.mixout-gain-dai-copier-playback { Object.Widget { mixout."1" {} dai-copier."1" {} gain."1" {} pipeline."1" {} } Object.Base { !route [ { source mixout.$index.1 sink gain.$index.1 } ] } } If we want to extend this pipeline with the addition of an eqiir/eqfir, we can create a Subtreecopy node with type extend as follows: Subtreecopy.mixout-gain-eqiir-eqfir-dai-copier-playback { source "Class.Pipeline.mixout-gain-dai-copier-playback" target "Class.Pipeline" type extend Object.Widget { eqiir.1 {} eqfir.1 {} } Object.Base { !route [ { source gain.$index.1 sink eqiir.$index.1 } { source eqiir.$index.1 sink eqfir.$index.1 } ] } } But if we want to modify an existing pipeline class while modifying the order of widgets and/or inserting new widgets, we should use the type "override" instead. This allows for adding new widgets to the list of widgets in the base class definition while also allowing overriding the routes to allow inserting the new widgets and reordering the widgets in the base class. For example, if we want to add a drc widget between the gain and the eqiir modules in the above class, we can do the following: Subtreecopy.mixout-efx-dai-copier-playback { source "Class.Pipeline.mixout-gain-eqiir-eqfir-dai-copier-playback" target "Class.Pipeline" type override Object.Widget { drc.1 {} } Object.Base { !route [ { source mixout.$index.1 sink gain.$index.1 } { source gain.$index.1 sink drc.$index.1 } { source drc.$index.1 sink eqiir.$index.1 } { source eqiir.$index.1 sink eqfir.$index.1 } ] } } Note that the routes array contains all the routes in the new subtreecopy definition. This subtreecopy feature can also be used to copy normal config blocks without having the need to duplicate information. For example, if all the widgets in a pipeline support the same audio formats, we can define it as follows: Class.Pipeline.dai-copier-eqiir-module-copier-capture { formats { num_input_audio_formats 2 num_output_audio_formats 2 Object.Base.input_audio_format [ { in_bit_depth 32 in_valid_bit_depth 32 } { in_channels 4 in_bit_depth 32 in_valid_bit_depth 32 in_ch_cfg $CHANNEL_CONFIG_3_POINT_1 in_ch_map $CHANNEL_MAP_3_POINT_1 } ] Object.Base.output_audio_format [ { out_bit_depth 32 out_valid_bit_depth 32 } { out_channels 4 out_bit_depth 32 out_valid_bit_depth 32 out_ch_cfg $CHANNEL_CONFIG_3_POINT_1 out_ch_map $CHANNEL_MAP_3_POINT_1 } ] } Object.Widget { dai-copier."1" { type dai_out num_output_pins 1 Subtreecopy.formats { source "Class.Pipeline.dai-copier-eqiir-module-copier-capture.formats" type override } } eqiir."1" { Subtreecopy.formats { source "Class.Pipeline.dai-copier-eqiir-module-copier-capture.formats" type override } Object.Control.bytes."1" { IncludeByKey.DMIC0_DAI_EQIIR { "passthrough" "include/components/eqiir/passthrough.conf" "highpass_40hz_0db" "include/components/eqiir/highpass_40hz_0db_48khz.conf" "highpass_40hz_20db" "include/components/eqiir/highpass_40hz_20db_48khz.conf" } } } module-copier."2" { Subtreecopy.formats { source "Class.Pipeline.dai-copier-eqiir-module-copier-capture.formats" type override } } pipeline."1" { priority 0 lp_mode 0 } } } The subtreecopy node in each widget ensures that the pipeline formats are applied to all widgets without the need for duplicating them for each widget. Link: https://github.com/thesofproject/sof/issues/9082 Request URL : https://github.com/alsa-project/alsa-utils/pull/268 Patch URL : https://github.com/alsa-project/alsa-utils/pull/268.patch Repository URL: https://github.com/alsa-project/alsa-utils

1 0

topology: pre-processor: Introduce extends/overrides keywords for classes
by GitHub pull_request - edited 22 May '24

22 May '24

alsa-project/alsa-utils pull request #268 was edited from ranj063: Add 2 new keywords for class definitions to allow new pipeline classes extend the definitions in the base class. This feature is useful for extending previous pipeline class definitions with the addition of one or more widgets without having to duplicate everything in the new class definition. For example: Consider a pipeline class definition as below. Note that only the widgets & routes are shown here. ``` Class.Pipeline.mixout-gain-dai-copier-playback { Object.Widget { mixout."1" {} dai-copier."1" {} gain."1" {} pipeline."1" {} } Object.Base { !route [ { source mixout.$index.1 sink gain.$index.1 } ] } } ``` If we want to extend this pipeline with the addition of an eqiir/eqfir, the extends keyword can be used as below: ``` Class.Pipeline.mixout-gain-eqiir-eqfir-dai-copier-playback { extends "mixout-gain-dai-copier-playback" Object.Widget { eqiir.1 {} eqfir.1 {} } Object.Base { !route [ { source gain.$index.1 sink eqiir.$index.1 } { source eqiir.$index.1 sink eqfir.$index.1 } ] } } ``` This allows for defining a new class without having to duplicate everything in the base class and just adding the new widgets/routes that extend the current pipeline definition in the base class. The extends keyword is useful when extending a pipeline while keeping the routes in the base class intact and adding new widgets at the end of the pipeline. But if we want to modify an existing pipeline class while modifying the order of widgets and/or inserting new widgets, we should use the overridess keyword instead. This allows for adding new widgets to the list of widgets in the base class definition while also allowing overriding the routes to allow inserting the new widgets and reordering the widgets in the base class. For example, if we want to add a drc widget between the gain and the eqiir modules in the above class, we can do the following: ``` Class.Pipeline.mixout-efx-dai-copier-playback { overrides "mixout-gain-eqiir-eqfir-dai-copier-playback" Object.Widget { drc.1 {} } Object.Base { !route [ { source mixout.$index.1 sink gain.$index.1 } { source gain.$index.1 sink drc.$index.1 } { source drc.$index.1 sink eqiir.$index.1 } { source eqiir.$index.1 sink eqfir.$index.1 } ] } } ``` Nore that the routes array contains all the routes in the new class definition. Link: https://github.com/thesofproject/sof/issues/9082 Request URL : https://github.com/alsa-project/alsa-utils/pull/268 Patch URL : https://github.com/alsa-project/alsa-utils/pull/268.patch Repository URL: https://github.com/alsa-project/alsa-utils

1 0

topology: pre-processor: Introduce extends/overrides keywords for classes
by GitHub pull_request - opened 22 May '24

22 May '24

alsa-project/alsa-utils pull request #268 was opened from ranj063: Add 2 new keywords for class definitions to allow new pipeline classes extend the definitions in the base class. This feature is useful for extending previous pipeline class definitions with the addition of one or more widgets without having to duplicate everything in the new class definition. For example: Consider a pipeline class definition as below. Note that only the widgets & routes are shown here. Class.Pipeline.mixout-gain-dai-copier-playback { Object.Widget { mixout."1" {} dai-copier."1" {} gain."1" {} pipeline."1" {} } Object.Base { !route [ { source mixout.$index.1 sink gain.$index.1 } ] } } If we want to extend this pipeline with the addition of an eqiir/eqfir, the extends keyword can be used as below: Class.Pipeline.mixout-gain-eqiir-eqfir-dai-copier-playback { extends "mixout-gain-dai-copier-playback" Object.Widget { eqiir.1 {} eqfir.1 {} } Object.Base { !route [ { source gain.$index.1 sink eqiir.$index.1 } { source eqiir.$index.1 sink eqfir.$index.1 } ] } } This allows for defining a new class without having to duplicate everything in the base class and just adding the new widgets/routes that extend the current pipeline definition in the base class. The extends keyword is useful when extending a pipeline while keeping the routes in the base class intact and adding new widgets at the end of the pipeline. But if we want to modify an existing pipeline class while modifying the order of widgets and/or inserting new widgets, we should use the overridess keyword instead. This allows for adding new widgets to the list of widgets in the base class definition while also allowing overriding the routes to allow inserting the new widgets and reordering the widgets in the base class. For example, if we want to add a drc widget between the gain and the eqiir modules in the above class, we can do the following: Class.Pipeline.mixout-efx-dai-copier-playback { overrides "mixout-gain-eqiir-eqfir-dai-copier-playback" Object.Widget { drc.1 {} } Object.Base { !route [ { source mixout.$index.1 sink gain.$index.1 } { source gain.$index.1 sink drc.$index.1 } { source drc.$index.1 sink eqiir.$index.1 } { source eqiir.$index.1 sink eqfir.$index.1 } ] } } Nore that the routess array contains all the routes in the new class definition. Link: https://github.com/thesofproject/sof/issues/9082 Request URL : https://github.com/alsa-project/alsa-utils/pull/268 Patch URL : https://github.com/alsa-project/alsa-utils/pull/268.patch Repository URL: https://github.com/alsa-project/alsa-utils

1 0