Logging of initialization time is broken

[Francesco Casella]

Since a few days ago, I am getting this kind of results out of simulations logs:

LOG_STATS         | info    | ### STATISTICS ###
|                 | |       | | timer
|                 | |       | | |     0.121037s          reading init.xml
|                 | |       | | |   0.00554686s          reading info.xml
|                 | |       | | |   0.00598597s          pre-initialization
|                 | |       | | |       136438s [2475022.5%] initialization
|                 | |       | | |   0.00273595s [  0.0%] steps
|                 | |       | | |    0.0125231s [  0.2%] creating output-file
|                 | |       | | |    0.0051622s [  0.1%] event-handling
|                 | |       | | |   0.00267392s [  0.0%] overhead
|                 | |       | | |      -136432s [-2474923.0%] simulation
|                 | |       | | |      5.51258s [100.0%] total

There is clearly something badly wrong with the measurement of the time spent doing initialization. Can you please fix that?

[Lennart Ochel]

Fixed with ​OMCompiler#870. Unfortunately, we cannot catch those regressions with the test suite.

[Martin Sjölund]

Milestone deleted