[BUG] Glitches in ASRC playback stress test

[singalsu]

Describe the bug
Glitch in playback stress test with sof-diagnostic-driver environment. The issue was created because it is not similar (a much larger discontinuity) that has happened in capture (#2981) and could not be fixed similarly as capture.

To Reproduce
Run the playback test with sufficient number of iterations for 8 to 8 kHz conversion, e.g. 100 iterations.

Reproduction Rate
About 5% occurrence.

Expected behavior
Clean sine waveform output.

Impact
If the glitch would happens in normal audio use case there could be a very short audible corruption happening every few minutes.

Environment

1. Branch name and commit hash of the 2 repositories: sof (firmware/topology) and linux (kernel driver).
   • Kernel: 5.8.0-rc1-test+
   • SOF: git master (7b532d3)
2. Name of the topology file
   • Topology: N/A
3. Name of the platform(s) on which the bug is observed.
   • Platform: CML

Screenshots or console output

Please also include the relevant sections from the firmware log and kernel log in the report (and attach the full logs for complete reference). Kernel log is taken from dmesg and firmware log from sof-logger. See https://thesofproject.github.io/latest/developer_guides/debugability/logger/index.html

[singalsu]

@jdrzaszcz FYI

[slawblauciak]

@singalsu any updates?

[singalsu]

@slawblauciak I have tested a lot of things and work with debugging this nearly all the work time but I haven't found the root cause. I have a draft PR that makes the issue rare but it can still happen.

[singalsu]

@slawblauciak @jdrzaszcz I'm not getting this error with git master firmware if I leave out option to force internal ssp. Now four rounds of 100x stress test passed. I wonder this issue can be closed?

[slawblauciak]

@singalsu what does that option do exactly? When do we use it? Why would we want to use it?

[lgirdwood]

@singalsu ping - any update ?

[lgirdwood]

@singalsu ping - maybe for v1.8 ?

[singalsu]

@lgirdwood Yep, unfortunately v1.8.