check-sof-logger: reload drivers first + dma_nudge() workaround

[marc-hb]

As we look for the FW ABI banner we must start from a clean state

Signed-off-by: Marc Herbert marc.herbert@intel.com

[marc-hb]

https://sof-ci.01.org/softestpr/PR710/build741/devicetest/ failed on

• TGLH_RVP_HDA error: Unable to open in file /sys/kernel/debug/sof/trace: No such file or directory
• BDW and BYT: empty DMA trace, see [BUG]Empty or stuck DMA trace - Initial FW ABI banner message not found in logger.data.txt sof#4333

[marc-hb]

TGLH_RVP_HDA error: Unable to open in file /sys/kernel/debug/sof/trace: No such file or directory

Fixed in https://sof-ci.01.org/softestpr/PR710/build743/devicetest/ by adding a delay.

BDW and BYT: empty DMA trace

Unchanged.

[marc-hb]

DMA trace stuck again for BDW and BYT in https://sof-ci.01.org/softestpr/PR710/build744/devicetest/, until a "nudge" gets everything out.

[marc-hb]

#711 fix should go first

[marc-hb]

In https://sof-ci.01.org/softestpr/PR710/build748/devicetest/ bug thesofproject/sof#4333 is reproduced as usual on BDW and BYT and then worked around by this new check-sof-logger "nudge".

[marc-hb]

Thanks @plbossart !

This has been in review for 7+ days and I tested this quite thoroughly. Will merge tomorrow if no objection by then.

To be clear:

• The first commit makes failure 4333 more deterministic
• The second commit "hides" it. The intention is to leave 4333 open until it is fixed (or close as wontfix)