"Nice" report for failed forks/benchmarks

[Vladimir Sitnikov]

> failing benchmark should be examined by looking at the execution
logs

I do not follow you.
Do you suggest to kill all the results for a particular benchmark even
if there is just a single failure?

Well, that's debatable, however I see no need in digging deeper.

Killing all the results for a failed benchmark is OK for me provided
there is a line in report that tells us that the benchmark did fail.

Vladimir