Apply time benchmarking

[sisuresh]

Description

Resolves #4391.

Checklist

• Reviewed the contributing document
• Rebased on top of master (no merge commits)
• Ran clang-format v8.0.0 (via make format or the Visual Studio extension)
• Compiles
• Ran all tests
• If change impacts performance, include supporting evidence per the performance document

[dmkozh]

I haven't looked too deeply into implementation, but looks good overall! (and seems already usable, would be interesting to see the results for our current configuration) Just left a few nit-picks/questions.

[dmkozh]

LGTM, but I'm not sure if we want to include this in p22 build. On one hand, this is not risky from the protocol standpoint, on the other hand we're modifying loadgen and that has a bit of risk for supercluster tests. I think we still have a bit of time before cutting the release, so we should have a run with just these changes. WDYT?

[dmkozh]

Actually there are some test failures, so we need to address these first.

[sisuresh]

LGTM, but I'm not sure if we want to include this in p22 build. On one hand, this is not risky from the protocol standpoint, on the other hand we're modifying loadgen and that has a bit of risk for supercluster tests. I think we still have a bit of time before cutting the release, so we should have a run with just these changes. WDYT?

Yeah that's a good point. The good thing is that if this change did break loadgen, we'd find out pretty quickly in SSC, so I think it's fine to merge. Worst case we'll just have to revert it.