Kill testing processes if the test takes too long

[typeless]

A temporary band-aid to avoid resource-wasting CI build.
It prints the backtrace in case of timeout.

For #5172

[codecov-io]

Codecov Report

❗ No coverage uploaded for pull request base (master@abf6e0d). Click here to learn what that means.
The diff coverage is n/a.

@@            Coverage Diff            @@
##             master    #5174   +/-   ##
=========================================
  Coverage          ?   37.52%           
=========================================
  Files             ?      309           
  Lines             ?    45820           
  Branches          ?        0           
=========================================
  Hits              ?    17195           
  Misses            ?    26163           
  Partials          ?     2462

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update abf6e0d...cf2dce0. Read the comment docs.

[lunny]

@typeless please resolve the conflict

[typeless]

@lunny Done.

[lunny]

Could you reenable test-sqlite3 on this PR?

[typeless]

@lunny Sure.

[lunny]

I restarted the CI 3 times and everytime CI passed and there is no timeout.

[lunny]

And the killing maybe unnecessary.

[typeless]

@lunny If there are no concerns, I would like to keep it longer or even apply to all the database tests in the hope that some mysterious, sporadic hanging bug can be spotted (which could be very hard to reproduce).

[lunny]

Yes, ten minutes maybe too short.