Extensive improvements in benchmark tool

[GiovanniBussi]

Added:

• stop after a maximum time
• benchmark domain decomposition runs
• comparative analysis on different versions
• sleep to simulate MD code calculation

See documentation

@carlocamilloni could you share the SAXS test case you were running on MPI? I want to test it with this benchmark tool, to see if by running multiple versions at a time we get more precise results.

I would need:

• plumed input file
• all accessory files (pdb etc) if needed

no trajectory should be needed

[codecov-commenter]

Codecov Report

Attention: Patch coverage is 2.70270% with 108 lines in your changes are missing coverage. Please review.

Project coverage is 83.24%. Comparing base (0c8be3e) to head (6869c9f).

:exclamation: Current head 6869c9f differs from pull request most recent head 27e789c. Consider uploading reports for the commit 27e789c to get more accurate results

Files	Patch %	Lines
src/cltools/Benchmark.cpp	2.70%	108 Missing :warning:

:exclamation: Your organization needs to install the Codecov GitHub app to enable full functionality.

Additional details and impacted files

```diff @@ Coverage Diff @@ ## master #1042 +/- ## ========================================== - Coverage 83.39% 83.24% -0.15% ========================================== Files 619 619 Lines 59130 59221 +91 ========================================== - Hits 49312 49300 -12 - Misses 9818 9921 +103 ```

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.

[carlocamilloni]

@GiovanniBussi here it is:

Archive.zip