Skip to content

Revert pipeline stop changes - #4920

Merged
keyonjie merged 4 commits into
thesofproject:mainfrom
ranj063:fix/revert_pipe_stop
Oct 27, 2021
Merged

keyonjie merged 4 commits into
thesofproject:mainfrom
ranj063:fix/revert_pipe_stop

Conversation

@ranj063

@ranj063 ranj063 commented Oct 27, 2021

Copy link
Copy Markdown
Collaborator

No description provided.

This reverts commit b5506ed.

Signed-off-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
This reverts commit b6f26b0.

Signed-off-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
This reverts commit 0fe263d.

Signed-off-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
This reverts commit 756f47a.

Signed-off-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
@XiaoyunWu6666

XiaoyunWu6666 commented Oct 27, 2021 •

Copy link
Copy Markdown
Contributor

hi @ranj063 , thanks for your quick response .

Ran some basic aplay arecord test , the result is PASS .[see inner test 7699 7698]

@keqiaozhang

keqiaozhang commented Oct 27, 2021 •

Copy link
Copy Markdown
Collaborator

Bisect shows that the DSP panic issue is caused by:
756f47a

pipeline: move STOP / PAUSE trigger to the pipeline task A recent commit moved START and RESUME trigger processing to the pipeline task for timer domain pipelines. This patch does the same for STOP and PAUSE.

Signed-off-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>

Checked this commit with yesterday kernel sof-dev/ 3f5049f6eed8, issue is still reproducible. So this issue has nothing to do with kernel.

@aiChaoSONG

Copy link
Copy Markdown
Collaborator

@ranj063 Do we still need to revert all four commits rather than the single 756f47a which causes failure?

@XiaoyunWu6666

XiaoyunWu6666 commented Oct 27, 2021 •

Copy link
Copy Markdown
Contributor

We should rebuild the SOF for daily test after this PR gets merged today, otherwises the kernel PR test will always fail.

Should we revert all 4 relevant commits or just remove the rootcause one ?

@ranj063

ranj063 commented Oct 27, 2021

Copy link
Copy Markdown
Collaborator Author

@ranj063 Do we still need to revert all four commits rather than the single 756f47a which causes failure?

@aiChaoSONG that commit alone cannot be cleanly reverted. So I reverted all

@ujfalusi
ujfalusi requested a review from lyakh October 27, 2021 04:17
@marc-hb

marc-hb commented Oct 27, 2021

Copy link
Copy Markdown
Collaborator

@aiChaoSONG that commit alone cannot be cleanly reverted. So I reverted all

Looking at the original #4724 there are 1 functional change + 3 "cosmetic" commits. The 3 "cosmetic" commits are after the functional change = hence the revert conflicts

@keqiaozhang

Copy link
Copy Markdown
Collaborator

The root cause is that there’re some relations between PR #4724 and PR #4885.

The PR test for #4724 was based on sof commit 1ef769d, at that time #4885 had not been merged,which make sense that the PR test passed.

Then #4724 merged before #4885 and issue happened.
I tried to revert #4885 on top of main branch, issue cannot be reproduced any more.

@aiChaoSONG

Copy link
Copy Markdown
Collaborator

Is everyone good with the revert? can some one with merge right merge this, or all of our test will fail.

@keyonjie

Copy link
Copy Markdown
Contributor

Is everyone good with the revert? can some one with merge right merge this, or all of our test will fail.

Doing that now.

@keyonjie
keyonjie merged commit 01a9662 into thesofproject:main Oct 27, 2021
@ranj063
ranj063 deleted the fix/revert_pipe_stop branch November 1, 2021 22:57
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.

7 participants