-
Notifications
You must be signed in to change notification settings - Fork 44
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.
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
Multi-playback/capture-pipeline always iterates in the same order #495
Comments
Not following @ranj063 the test result link shows a failure for kmod load / unload. Can you clarify how you made things fail? |
@plbossart yes, thatethe point. The multi-pipeline-playback.sh test for example on BYT starts PCM 0 followed by PCM 1. And stops PCM 0 followed by PCM 1. But if you were to stop PCM1 before PCM0, it would have failed. |
It still don't get how you found that out, do you have a patch for that script? |
Seconding @plbossart : can you please submit a draft and never merged PR with that simple change? A separate sof or linux github about the actual bug (as opposed to this test coverage gap) may also be needed. |
@plbossart I didnt use a script. On BYT, there are exactly 2 playback pipelines and luckily for me I was testing multi pipeline playback start and stop in that order and it failed. @marc-hb What I am asking is not a one line change in the script especially with 4 active PCMs. I would have submitted the code if it were that simple |
You sound like you just entered some commands in a shell to reproduce this issue, no? So that is basically a script. If what you did to reproduce is more complicated than entering a few commands then I think it's missing from the description.
I did understand that a proper test solution wasn't easy, otherwise you would have submitted a PR and not this bug. However compared to a proper solution it's often much easier to make some horrible, throw-away hacks in a test script purely to demonstrate some issue. Think of it as forcibly shoehorning the interactive commands you used into whichever script is the closest to what you want to do. Do not feel obligated by any coding standard. Considering how short and simple your test description sounds so far, I'm doubting we have no script remotely closed to what you did - but please correct me in the description. |
@marc-hb i will attempt to make a script with what I have in my mind. stay tuned! |
This is a pretty old ticket, and it's marked with "good first issue." Does Chao's PR close this ticket? I'm not sure I understand what the next steps are, here. |
The multi pipeline tests start and stop PCM in the same order. It doesnt cover the case where the PCMs are started in one order and stopped in another order. For example:
https://sof-ci.01.org/linuxpr/PR2371/build4845/devicetest/
In the above device test result, the multi-playback-pipeline.sh test passes on BYT/BDW/BSW because the PCMs are started in the order 0 followed by 1 and stopped in the order 0 followed by 1.
If I switch the order for stopping to 1 followed by 0, the test fails on my device.
cc:
The text was updated successfully, but these errors were encountered: