Skip to content

topology: hdmi-generic: change HDMI PCM id - #6787

Merged
lgirdwood merged 4 commits into
thesofproject:mainfrom
bardliao:hdmi-cpm-id
Jan 11, 2023
Merged

lgirdwood merged 4 commits into
thesofproject:mainfrom
bardliao:hdmi-cpm-id

Conversation

@bardliao

Copy link
Copy Markdown
Collaborator

Define HDMI_PCM_ID macro so that we can have different pcm id for different topology. And change the hdmi pcm id of sof-hda-generic.tplg to align with UCM.

Signed-off-by: Bard Liao yung-chuan.liao@linux.intel.com

Comment thread tools/topology/topology2/sof-hda-generic.conf Outdated
Comment thread tools/topology/topology2/platform/intel/hdmi-default.conf Outdated
Comment thread tools/topology/topology2/platform/intel/hdmi-generic.conf Outdated

@kv2019i kv2019i left a comment

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.

Comments inline.

Comment thread tools/topology/topology2/platform/intel/hdmi-generic.conf Outdated
@plbossart

Copy link
Copy Markdown
Member

@kv2019i for sof-sdw we only have 3 devices in the UCM conf. So there's no issue if we stop using the 4th pipe for audio for new platforms

# Use case Configuration for sof-soundwire

Include.hdmi.File "/codecs/hda/hdmi.conf"

If.hdmi1 {
	Condition { Type AlwaysTrue }
	True.Macro.hdmi1.HDMI {
		Number 1
		Device 5
		Priority 500
	}
}

If.hdmi2 {
	Condition { Type AlwaysTrue }
	True.Macro.hdmi2.HDMI {
		Number 2
		Device 6
		Priority 600
	}
}

If.hdmi3 {
	Condition { Type AlwaysTrue }
	True.Macro.hdmi2.HDMI {
		Number 3
		Device 7
		Priority 700
	}
}

@bardliao

Copy link
Copy Markdown
Collaborator Author

@plbossart I added a commit to change NUM_HDMIS from 4 to 3 for all existing topologies, but keep hdmi-4th.conf in case we want to use it in the future. I also add other commits to fix the pipeline id conflict between sdw speaker and hdmi and add hdmi support to the existing sdw topologies.

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.

@bardliao @jsarha I assume we can change
"copier.host.50.1" to "copier.host.$HDMI1_HOST_PIPELINE_ID.1"
i.e. we dereference the HDMI1_HOST_PIPELINE_ID assignment above ?

Copy link
Copy Markdown
Collaborator Author

Choose a reason for hiding this comment

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

@bardliao @jsarha I assume we can change "copier.host.50.1" to "copier.host.$HDMI1_HOST_PIPELINE_ID.1" i.e. we dereference the HDMI1_HOST_PIPELINE_ID assignment above ?

That is what I was looking for. But according to @ranj063, we don't support it yet.

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.

@bardliao @jsarha I assume we can change "copier.host.50.1" to "copier.host.$HDMI1_HOST_PIPELINE_ID.1" i.e. we dereference the HDMI1_HOST_PIPELINE_ID assignment above ?

That is what I was looking for. But according to @ranj063, we don't support it yet.

ok, got you - @jsarha can you confirm this is in your topology improvement work.

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

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

@bardliao @lgirdwood that is not exactly what I trying to achieve just yet. So far I've been looking for a way to implement variable substitution for alsa control names, not generic widget name substitution.

@plbossart plbossart left a comment

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.

couple of comments below @bardliao

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.

HDMI in capital letters

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.

typo: Playback

Comment thread tools/topology/topology2/platform/intel/hdmi-generic.conf Outdated

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.

this topology name needs to be fixed or this needs to moved to a different develop, it's not used by the kernel.

cavs-sdw.conf uses grouped route and pipeline id. And index 20 based is
for Speaker pcm. That will conflict with the existing hdmi1 pipeline id.
Update hdmi route and pipeline ids with the same formula to avoid the
conflict.

Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Define HDMI_PCM_ID macro so that we can have different pcm id for
different topology. And change the hdmi pcm id of sof-hda-generic.tplg
to align with UCM.

Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Set NUM_HDMIS=3 for all topologies.

Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com>
@bardliao

Copy link
Copy Markdown
Collaborator Author

@kv2019i @miRoox Can you take a look the "topology2: move development topologies to development folder" commit? Some topologies are moved to development folder. I think CI will be impacted.

Some topologies are for development only. Move them to the development
folder.

Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com>
@lgirdwood

Copy link
Copy Markdown
Member

@plbossart good for you ?

@plbossart

Copy link
Copy Markdown
Member

@ranj063 can you please double-check this PR?

@lgirdwood
lgirdwood merged commit fab1376 into thesofproject:main Jan 11, 2023
@marc-hb

marc-hb commented Jan 12, 2023

Copy link
Copy Markdown
Collaborator

This caused regression

Tentative fix that disables HDMI:

@kv2019i

kv2019i commented Jan 13, 2023

Copy link
Copy Markdown
Collaborator

@lgirdwood

Copy link
Copy Markdown
Member

Ok, lets revert. @bardliao I guess you have enough data to root cause.

@marc-hb

marc-hb commented Jan 13, 2023 •

Copy link
Copy Markdown
Collaborator

Revert submitted by @lgirdwood in #6953

Indeed this was merged despite ADLP_RVP_SDW_IPC4ZPH totally failing in https://sof-ci.01.org/sofpr/PR6787/build3161/devicetest/index.html !

@bardliao please convert failing PRs to "draft" to block merges of failing PRs and help overloaded reviewers (also: please submit a revert if there's no quick fix)

@bardliao

Copy link
Copy Markdown
Collaborator Author

Revert submitted by @lgirdwood in #6953

Indeed this was merged despite ADLP_RVP_SDW_IPC4ZPH totally failing in https://sof-ci.01.org/sofpr/PR6787/build3161/devicetest/index.html !

@bardliao please convert failing PRs to "draft" to block merges of failing PRs and help overloaded reviewers (also: please submit a revert if there's no quick fix)

@marc-hb Sorry for the regressions. The problem is that as I mentioned on #6787 (comment), the CI will be impacted, so the PR test result was not reliable for this PR.
I do post fix PRs in some hour when I was informed that this PR causes some regressions and that's why I didn't submit a revert.

@marc-hb

marc-hb commented Jan 17, 2023

Copy link
Copy Markdown
Collaborator

the CI will be impacted, so the PR test result was not reliable for this PR.

In situations with a CI dependency like this we should coordinate and make the CI changes first (after local testing and approvals). Then use "SOFCI TEST" in 6787. If the tests pass then 6787 can be merged, if tests still don't pass then CI changes should be reverted.

Another possibility is to temporarily disable some tests in some configurations.

I do post fix PRs in some hour when I was informed that this PR causes some regressions and that's why I didn't submit a revert.

"Quick" fixes are great except they must be tested, reviewed etc. which means they're almost never quick.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

6 participants