audio: copier: use v_index for DAI index in I2S link creation #8980

dnikodem · 2024-03-25T10:15:39Z

This commit updates the copier DAI creation logic for I2S links to use the v_index field from the node_id structure as the DAI index. This change ensures that the correct DAI index is used when creating I2S links, aligning with the new SSP link management mechanism.

The following PR must be merged into the Zephyr repository before we can merge this PR:

driver: ssp: update Intel SSP DAI driver to support dynamic SSP link management zephyrproject-rtos/zephyr#70660

dnikodem · 2024-03-25T13:51:20Z

SOFCI TEST

plbossart · 2024-03-25T14:34:00Z

src/audio/copier/copier_dai.c

 	case ipc4_i2s_link_output_class:
 	case ipc4_i2s_link_input_class:
-		dai_index[dai_count - 1] = (dai_index[dai_count - 1] >> 4) & 0xF;
+		dai_index[dai_count - 1] = node_id.f.v_index;


line 247 I can see this:

dai_index[dai_count - 1] = node_id.f.v_index;

So this new assignment is redundant?

correct, this is redundant. I will remove it in the next update (This PR was created mainly to check the CI results - I missed this). Thanks

dnikodem · 2024-03-26T07:14:54Z

This PR is related to zephyrproject-rtos/zephyr#70660 - Still in development

dnikodem · 2024-03-27T12:59:57Z

SOFCI TEST

plbossart · 2024-03-27T15:26:42Z

before merging we need to make sure that all 'nocodec' kernel tests pass, that's where we test the SSP mostly along with the topology definitions/blobs..

dnikodem · 2024-03-27T15:56:51Z

before merging we need to make sure that all 'nocodec' kernel tests pass, that's where we test the SSP mostly along with the topology definitions/blobs..

Of course, I agree with that. I am currently trying to understand why some of the 'nocodec' type tests are not starting.

dnikodem · 2024-03-28T17:37:21Z

SOFCI TEST

lgirdwood · 2024-04-16T08:15:33Z

@dnikodem @plbossart looks like all nocodec tests are working - does this mean we are ready now for non draft ?

dnikodem · 2024-04-16T12:16:11Z

@dnikodem @plbossart looks like all nocodec tests are working - does this mean we are ready now for non draft ?

For my part - I am waiting for the "final" confirmation from the validation team (this is related to the recent changes in the zephyr part). The results should be available tomorrow.

dnikodem · 2024-04-19T14:34:29Z

Referring to Wojciech's comment from the Zephyr part:

wszypelt commented 2 days ago •
@dnikodem @kv2019i Based on the SOF PR 8980, I checked the changes for MTL and LNL, in MTL everything works properly, in LNL I have to wait a while until all the tests are done, however, I can tell you that so far everything looks good. I will write again as soon as all tests are completed

wszypelt commented yesterday
@dnikodem @kv2019i all LNL tests look good :)

dnikodem · 2024-04-22T08:09:37Z

Rebased because the number of SSP devices was reduced with other PR: #9053

lgirdwood

LGTM, @plbossart @ujfalusi pls confirm no driver impact. Thanks !

plbossart · 2024-04-22T16:00:05Z

Can I ask where this 'v_index' is set?

topology?
NHLT blob inserted in topology?

in the latter case, @singalsu @jsarha can you confirm alsa-utils does set this field in the blob?

lgirdwood · 2024-04-23T11:14:48Z

Can I ask where this 'v_index' is set?

topology? NHLT blob inserted in topology?

in the latter case, @singalsu @jsarha can you confirm alsa-utils does set this field in the blob?

@dnikodem can you confirm where v_index comes from, e.g. IPC, NHLT, topology ? Thanks !

ujfalusi · 2024-04-23T12:31:08Z

Can I ask where this 'v_index' is set?
topology? NHLT blob inserted in topology?
in the latter case, @singalsu @jsarha can you confirm alsa-utils does set this field in the blob?

@dnikodem can you confirm where v_index comes from, e.g. IPC, NHLT, topology ? Thanks !

The v_index is coming from topology (it is essentially the SOF_TKN_DAI_INDEX token of a DAI copier.
For SSP we set it into the node_id |= dai_index << 4; and bit0-3 is unused for SSP.

In future platforms we might pass additional information for SSP selection via bit0-3 and this is what this patch is preparing for.
Note: Linux will always pass 0 as bit0-3 for now, which might not need to be changed at all.

plbossart · 2024-04-23T18:39:26Z

@dnikodem rebase needed and draft status to be changed.

dnikodem · 2024-04-24T07:54:55Z

@dnikodem rebase needed and draft status to be changed.

Sure, I'll do it, but before we merge it, we need to merge the zephyr part: zephyrproject-rtos/zephyr#70660

This commit updates the copier DAI creation logic for I2S links to use the `v_index` field from the `node_id` structure as the DAI index. This change ensures that the correct DAI index is used when creating I2S links, aligning with the new SSP link management mechanism. West update to necessary SSP changes: driver: ssp: update Intel SSP DAI driver to support dynamic SSP management Signed-off-by: Damian Nikodem <damian.nikodem@intel.com>

dnikodem · 2024-04-25T14:21:19Z

It seems that CI failures are not related to these PR changes.

lgirdwood · 2024-04-25T14:36:26Z

It seems that CI failures are not related to these PR changes.

@wszypelt @lrudyX safe to merge ?

lrudyX · 2024-04-25T14:54:07Z

It seems that CI failures are not related to these PR changes.

@wszypelt @lrudyX safe to merge ?

This PR was actually validated as green, CI failed after testing.

lgirdwood · 2024-04-25T14:58:14Z

It seems that CI failures are not related to these PR changes.

@wszypelt @lrudyX safe to merge ?

This PR was actually validated as green, CI failed after testing.

Thanks @lrudyX, I will merge.

dnikodem force-pushed the update_ssp_copier_init_for_intel_platform branch 2 times, most recently from 63e6366 to ab09ce4 Compare March 25, 2024 11:54

dnikodem force-pushed the update_ssp_copier_init_for_intel_platform branch from ab09ce4 to ebd532e Compare March 25, 2024 14:09

plbossart reviewed Mar 25, 2024

View reviewed changes

dnikodem force-pushed the update_ssp_copier_init_for_intel_platform branch 2 times, most recently from 1183253 to 562e0d7 Compare March 27, 2024 11:04

dnikodem force-pushed the update_ssp_copier_init_for_intel_platform branch from 562e0d7 to 1654f97 Compare March 28, 2024 15:16

dnikodem force-pushed the update_ssp_copier_init_for_intel_platform branch 4 times, most recently from 75b0b68 to 279d17b Compare April 15, 2024 16:43

marc-hb mentioned this pull request Apr 17, 2024

driver: ssp: update Intel SSP DAI driver to support dynamic SSP link management zephyrproject-rtos/zephyr#70660

Merged

1 task

dnikodem force-pushed the update_ssp_copier_init_for_intel_platform branch from 279d17b to 7ee1cda Compare April 22, 2024 08:06

lgirdwood reviewed Apr 22, 2024

View reviewed changes

dnikodem force-pushed the update_ssp_copier_init_for_intel_platform branch from 7ee1cda to 550c778 Compare April 25, 2024 10:10

dnikodem marked this pull request as ready for review April 25, 2024 10:10

dnikodem requested review from RanderWang, abonislawski and pblaszko as code owners April 25, 2024 10:10

dnikodem force-pushed the update_ssp_copier_init_for_intel_platform branch from 550c778 to 95c4e8c Compare April 25, 2024 12:34

dnikodem requested review from dbaluta, kv2019i, lbetlej and mmaka1 as code owners April 25, 2024 12:34

dnikodem changed the title ~~[DNM] audio: copier: use v_index for DAI index in I2S link creation~~ audio: copier: use v_index for DAI index in I2S link creation Apr 25, 2024

kv2019i added the Zephyr Issues only observed with Zephyr integrated label Apr 25, 2024

lgirdwood approved these changes Apr 25, 2024

View reviewed changes

kv2019i approved these changes Apr 25, 2024

View reviewed changes

iganakov approved these changes Apr 25, 2024

View reviewed changes

lgirdwood merged commit e603d8f into thesofproject:main Apr 25, 2024

audio: copier: use v_index for DAI index in I2S link creation #8980

audio: copier: use v_index for DAI index in I2S link creation #8980

Uh oh!

Conversation

dnikodem commented Mar 25, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

dnikodem commented Mar 25, 2024

Uh oh!

plbossart Mar 25, 2024

Choose a reason for hiding this comment

Uh oh!

dnikodem Mar 26, 2024

Choose a reason for hiding this comment

Uh oh!

dnikodem commented Mar 26, 2024

Uh oh!

dnikodem commented Mar 27, 2024

Uh oh!

plbossart commented Mar 27, 2024

Uh oh!

dnikodem commented Mar 27, 2024

Uh oh!

dnikodem commented Mar 28, 2024

Uh oh!

lgirdwood commented Apr 16, 2024

Uh oh!

dnikodem commented Apr 16, 2024

Uh oh!

dnikodem commented Apr 19, 2024

Uh oh!

dnikodem commented Apr 22, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

lgirdwood left a comment

Choose a reason for hiding this comment

Uh oh!

plbossart commented Apr 22, 2024

Uh oh!

lgirdwood commented Apr 23, 2024

Uh oh!

ujfalusi commented Apr 23, 2024

Uh oh!

plbossart commented Apr 23, 2024

Uh oh!

dnikodem commented Apr 24, 2024

Uh oh!

dnikodem commented Apr 25, 2024

Uh oh!

lgirdwood commented Apr 25, 2024 • edited by dnikodem Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

lrudyX commented Apr 25, 2024

Uh oh!

lgirdwood commented Apr 25, 2024

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

7 participants

dnikodem commented Mar 25, 2024 •

edited

Loading

dnikodem commented Apr 22, 2024 •

edited

Loading

lgirdwood commented Apr 25, 2024 •

edited by dnikodem

Loading