Skip to content

Add support for hostless and DAIless pipelines#5609

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
bardliao merged 6 commits into from
Dec 22, 2025
Merged

Add support for hostless and DAIless pipelines #5609

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
216feda
uapi: sound: sof: tokens: Add missing token for KCPS
ranj063 Nov 24, 2025
4e33d00
ASoC: Intel: sof_sdw: Add a DAI link for loopback capture
ranj063 Nov 24, 2025
d6ec583
ASoC: SOF: ipc4-topology: Add new tokens for pipeline direction
ranj063 Dec 1, 2025
01ffdb7
ASoC: SOF: ipc4-topology: Add support for process modules with no inp…
ranj063 Dec 1, 2025
7700d9f
ASoC: SOF: sof-audio: Traverse paths with aggregated DAI widgets
ranj063 Dec 1, 2025
ab6ff22
ASoC: SOF: sof-audio: Add support for loopback capture
ranj063 Dec 2, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
3 changes: 3 additions & 0 deletions include/uapi/sound/sof/tokens.h
View file
Open in desktop
Original file line numberDiff line numberDiff line change
Expand Up@@ -56,6 +56,9 @@
#define SOF_TKN_SCHED_LP_MODE 207
#define SOF_TKN_SCHED_MEM_USAGE 208
#define SOF_TKN_SCHED_USE_CHAIN_DMA 209
#define SOF_TKN_SCHED_KCPS 210
#define SOF_TKN_SCHED_DIRECTION 211
#define SOF_TKN_SCHED_DIRECTION_VALID 212
Comment thread
ranj063 marked this conversation as resolved.

/* volume */
#define SOF_TKN_VOLUME_RAMP_STEP_TYPE 250
Expand Down
43 changes: 41 additions & 2 deletions sound/soc/intel/boards/sof_sdw.c
View file
Open in desktop
Original file line numberDiff line numberDiff line change
Expand Up@@ -1196,6 +1196,34 @@ static int create_bt_dailinks(struct snd_soc_card *card,
return 0;
}

static int create_echoref_dailink(struct snd_soc_card *card,
struct snd_soc_dai_link **dai_links, int *be_id)
{
struct device *dev = card->dev;
int ret;
char *name = devm_kasprintf(dev, GFP_KERNEL, "Loopback_Virtual");

if (!name)
return -ENOMEM;

/*
* use dummy DAI names as this won't be connected to an actual DAI but just to establish a
* fe <-> be connection for loopback capture for echo reference
*/
ret = asoc_sdw_init_simple_dai_link(dev, *dai_links, be_id, name,
0, 1, "snd-soc-dummy-dai", "dummy",
snd_soc_dummy_dlc.name, snd_soc_dummy_dlc.dai_name,
1, NULL, NULL);

@ujfalusiujfalusiDec 8, 2025

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Only sdw can have this feature, HDA or SSP machines never?
We cannot add to topology a tone generator for example at will, but we need to change the kernel as well?

@ranj063ranj063Dec 11, 2025

Copy link
Copy Markdown
CollaboratorAuthor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

In theory the same logic would work for all machines and yes you would need the machine driver change in order to use it. But in the case of SSP, we dont need this. We can use the Loopback mode like we've always used. So maybe just the HDA machine change would be enough

@lgirdwoodlgirdwoodDec 19, 2025

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

HDA and SSP not priority.

if (ret)
return ret;

(*dai_links)++;

@ujfalusiujfalusiDec 19, 2025

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

how this will work with existing topologies and with function topology?

@ranj063ranj063Dec 19, 2025

Copy link
Copy Markdown
CollaboratorAuthor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It shouldnt matter if we use function topologies right? We just have a single machine even in that case right?

@ujfalusiujfalusiDec 19, 2025

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think the function topology parsing will fall back to monolithic when it faces with a link of unknown origin/no fragment.
We have now patch to make this not fail when there is no monolithic topology defined, but this (I think) will eventually disable the function topologies?

I'm not sure, but this is my feeling.

@lgirdwoodlgirdwoodDec 19, 2025

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

best to comment what the assumption will be here regarding topology.


dev_dbg(dev, "Added echo reference DAI link\n");

return 0;
}

static int sof_card_dai_links_create(struct snd_soc_card *card)
{
struct device *dev = card->dev;
Expand DownExpand Up@@ -1304,8 +1332,12 @@ static int sof_card_dai_links_create(struct snd_soc_card *card)
goto err_end;
}

/* allocate BE dailinks */
num_links = sdw_be_num + ssp_num + dmic_num + hdmi_num + bt_num;
/*
* allocate BE dailinks, add an extra DAI link for echo reference capture.
* This should be the last DAI link and it is expected both for monolithic
* and functional SOF topologies to support echo reference.
*/
num_links = sdw_be_num + ssp_num + dmic_num + hdmi_num + bt_num + 1;
Comment thread
ranj063 marked this conversation as resolved.

@ujfalusiujfalusiDec 19, 2025

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I guess this is also a chicken and egg problem... This is purely topology construct, but we don't know if the topology have virtual DAI before we pick and parse the topology and for that we need the links in place if the topology do have this defined, right?

@ranj063ranj063Dec 19, 2025

Copy link
Copy Markdown
CollaboratorAuthor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

right

@lgirdwoodlgirdwoodDec 19, 2025

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

ok, lets capture this behavior in the comments.

dai_links = devm_kcalloc(dev, num_links, sizeof(*dai_links), GFP_KERNEL);
if (!dai_links) {
ret = -ENOMEM;
Expand DownExpand Up@@ -1354,6 +1386,13 @@ static int sof_card_dai_links_create(struct snd_soc_card *card)
goto err_end;
}

/* dummy echo ref link. keep this as the last DAI link. The DAI link ID does not matter */
ret = create_echoref_dailink(card, &dai_links, &be_id);
Comment thread
lgirdwood marked this conversation as resolved.
if (ret) {
dev_err(dev, "failed to create echo ref dai link: %d\n", ret);
goto err_end;
}

WARN_ON(codec_conf != card->codec_conf + card->num_configs);
WARN_ON(dai_links != card->dai_link + card->num_links);

Expand Down
62 changes: 49 additions & 13 deletions sound/soc/sof/ipc4-topology.c
View file
Open in desktop
Original file line numberDiff line numberDiff line change
Expand Up@@ -76,6 +76,10 @@ static const struct sof_topology_token ipc4_sched_tokens[] = {
offsetof(struct sof_ipc4_pipeline, core_id)},
{SOF_TKN_SCHED_PRIORITY, SND_SOC_TPLG_TUPLE_TYPE_WORD, get_token_u32,
offsetof(struct sof_ipc4_pipeline, priority)},
{SOF_TKN_SCHED_DIRECTION, SND_SOC_TPLG_TUPLE_TYPE_WORD, get_token_u32,
offsetof(struct sof_ipc4_pipeline, direction)},
{SOF_TKN_SCHED_DIRECTION, SND_SOC_TPLG_TUPLE_TYPE_BOOL, get_token_u16,
offsetof(struct sof_ipc4_pipeline, direction_valid)},
};

static const struct sof_topology_token pipeline_tokens[] = {
Expand DownExpand Up@@ -939,6 +943,10 @@ static int sof_ipc4_widget_setup_comp_pipeline(struct snd_sof_widget *swidget)

swidget->core = pipeline->core_id;
spipe->core_mask |= BIT(pipeline->core_id);
if (pipeline->direction_valid) {
spipe->direction = pipeline->direction;
spipe->direction_valid = true;
}

if (pipeline->use_chain_dma) {
dev_dbg(scomp->dev, "Set up chain DMA for %s\n", swidget->widget->name);
Expand All@@ -954,9 +962,9 @@ static int sof_ipc4_widget_setup_comp_pipeline(struct snd_sof_widget *swidget)
goto err;
}

dev_dbg(scomp->dev, "pipeline '%s': id %d, pri %d, core_id %u, lp mode %d\n",
dev_dbg(scomp->dev, "pipeline '%s': id %d, pri %d, core_id %u, lp mode %d direction %d\n",
swidget->widget->name, swidget->pipeline_id,
pipeline->priority, pipeline->core_id, pipeline->lp_mode);
pipeline->priority, pipeline->core_id, pipeline->lp_mode, pipeline->direction);

swidget->private = pipeline;

Expand DownExpand Up@@ -2745,12 +2753,14 @@ static int sof_ipc4_prepare_process_module(struct snd_sof_widget *swidget,
int input_fmt_index = 0;
int ret;

input_fmt_index = sof_ipc4_init_input_audio_fmt(sdev, swidget,
&process->base_config,
pipeline_params,
available_fmt);
if (input_fmt_index < 0)
return input_fmt_index;
if (available_fmt->num_input_formats) {
input_fmt_index = sof_ipc4_init_input_audio_fmt(sdev, swidget,
&process->base_config,
pipeline_params,
available_fmt);
if (input_fmt_index < 0)
return input_fmt_index;
}

/* Configure output audio format only if the module supports output */
if (available_fmt->num_output_formats) {
Expand All@@ -2759,12 +2769,28 @@ static int sof_ipc4_prepare_process_module(struct snd_sof_widget *swidget,
u32 out_ref_rate, out_ref_channels;
int out_ref_valid_bits, out_ref_type;

in_fmt = &available_fmt->input_pin_fmts[input_fmt_index].audio_fmt;
if (available_fmt->num_input_formats) {
in_fmt = &available_fmt->input_pin_fmts[input_fmt_index].audio_fmt;

out_ref_rate = in_fmt->sampling_frequency;
out_ref_channels = SOF_IPC4_AUDIO_FORMAT_CFG_CHANNELS_COUNT(in_fmt->fmt_cfg);
out_ref_valid_bits = SOF_IPC4_AUDIO_FORMAT_CFG_V_BIT_DEPTH(in_fmt->fmt_cfg);
out_ref_type = sof_ipc4_fmt_cfg_to_type(in_fmt->fmt_cfg);
out_ref_rate = in_fmt->sampling_frequency;
out_ref_channels =
SOF_IPC4_AUDIO_FORMAT_CFG_CHANNELS_COUNT(in_fmt->fmt_cfg);
out_ref_valid_bits =
Comment thread
ranj063 marked this conversation as resolved.
SOF_IPC4_AUDIO_FORMAT_CFG_V_BIT_DEPTH(in_fmt->fmt_cfg);
out_ref_type = sof_ipc4_fmt_cfg_to_type(in_fmt->fmt_cfg);

@ujfalusiujfalusiDec 19, 2025

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

and to add I have this PR for some time which touches the same code and fixes issue when not all parameters are changed by a process module:
#5442

} else {
/* for modules without input formats, use FE params as reference */
out_ref_rate = params_rate(fe_params);
out_ref_channels = params_channels(fe_params);
ret = sof_ipc4_get_sample_type(sdev, fe_params);
if (ret < 0)
Comment thread
ranj063 marked this conversation as resolved.
return ret;
out_ref_type = (u32)ret;

out_ref_valid_bits = sof_ipc4_get_valid_bits(sdev, fe_params);
if (out_ref_valid_bits < 0)
return out_ref_valid_bits;
}

output_fmt_index = sof_ipc4_init_output_audio_fmt(sdev, swidget,
&process->base_config,
Expand DownExpand Up@@ -2792,6 +2818,16 @@ static int sof_ipc4_prepare_process_module(struct snd_sof_widget *swidget,
if (ret)
return ret;
}

/* set base cfg to match the first output format if there are no input formats */
if (!available_fmt->num_input_formats) {
Comment thread
lgirdwood marked this conversation as resolved.
struct sof_ipc4_audio_format *out_fmt;

out_fmt = &available_fmt->output_pin_fmts[0].audio_fmt;

/* copy output format */
memcpy(&process->base_config.audio_fmt, out_fmt, sizeof(*out_fmt));

@ujfalusiujfalusiDec 8, 2025

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

what will happen if the fe_parmas and the first output_fmt of the module does not match?

@ranj063ranj063Dec 10, 2025

Copy link
Copy Markdown
CollaboratorAuthor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

You'll end up with an error that out format match wasnt found on line 2802 isnt it?

}
}

sof_ipc4_dbg_module_audio_format(sdev->dev, swidget, available_fmt,
Expand Down
4 changes: 4 additions & 0 deletions sound/soc/sof/ipc4-topology.h
View file
Open in desktop
Original file line numberDiff line numberDiff line change
Expand Up@@ -150,6 +150,8 @@ struct sof_ipc4_copier_config_set_sink_format {
* @use_chain_dma: flag to indicate if the firmware shall use chained DMA
* @msg: message structure for pipeline
* @skip_during_fe_trigger: skip triggering this pipeline during the FE DAI trigger
* @direction_valid: flag indicating if valid direction is set in topology
* @direction: pipeline direction set in topology if direction_valid is true
*/
struct sof_ipc4_pipeline {
uint32_t priority;
Expand All@@ -160,6 +162,8 @@ struct sof_ipc4_pipeline {
bool use_chain_dma;
struct sof_ipc4_msg msg;
bool skip_during_fe_trigger;
bool direction_valid;
u32 direction;
Comment thread
lgirdwood marked this conversation as resolved.

@ujfalusiujfalusiDec 19, 2025

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

should it be documented what values are valid for direction, like what 0 means, what 1 means, is there a direction == 2...10?

@ranj063ranj063Dec 19, 2025

Copy link
Copy Markdown
CollaboratorAuthor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@ujfalusi these come directly from topology, so are checked for validity there.

@ujfalusiujfalusiDec 19, 2025

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

so, the value for kernel does not matter, only they should match?

@lgirdwoodlgirdwoodDec 19, 2025

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

we can add that context to the comment. Can complain too if out of range.

};

/**
Expand Down
Loading
Loading