Possible race condition in libSplash #262
Description
There is one failed test for the "Fix compile errors" commit in #259 with a assertion failed - Expression: sum == old_sum in only one compilation (https://travis-ci.org/ComputationalRadiationPhysics/libSplash/jobs/243867208) but it is gone in the next commit which did not change anything at all (just re-added a function but did not use it)
In #248 I (very rarely) got a failure in closing the HDF5 file (HandleMgr) but can't really reproduce this.