Add correction when using refinement and boundary setup with meshes

See merge request walberla/walberla!107

Add fine/coarse overlap regions boundaries and fluid cells have to be setup according to the coarse grid resolution.

Don’t call MPI_Type_free after MPI_Finalize

See merge request walberla/walberla!106

Vector2, Vector3 and Matrix3 define their own MPITrait::type functions that return an MPI_Datatype for them. It is created the first time the function is called and put into a static variable. It is therefore destroyed after the main() function exits, which is usually after MPI_Finalize has been called. Freeing the MPI_Datatype is therefore not possible and mpi::datatype cannot be used.

Updates for ExprTk.

See merge request walberla/walberla!102

[API] Galerkin coarsening for Multigrid

See merge request walberla/walberla!103

Dominik Bartuschat authored 6 years ago and Michael Kuron committed 6 years ago

unified prolongate and restrict operators to work for GCA and DCA

Resolve "missing periodic intersection volume function for AABBs"

Closes #56

See merge request walberla/walberla!104

[API] changed return type to std array for improved usability

See merge request walberla/walberla!101

Mesh init

See merge request walberla/walberla!99

Replace deprecated -openmp flag in Intel 17 compiler

Closes #15

See merge request walberla/walberla!1

[API] Use ParserUBB in ExtendedBoundaryHandlingFactory

See merge request walberla/walberla!94

Refinement support for BoundaryFromDomainBorder

See merge request walberla/walberla!95

BoundarySetter should only call Boundary::createConfiguration once

See merge request walberla/walberla!97

Resolve "There are "gaps" created by the SFC partitionier"

Closes #54

See merge request walberla/walberla!96

- NearestNeighborFieldInterpolator::get needs to break out of all loops, just not the current one. While the test seemed to succeed anyway, it crashed with corrupted memory during exit when compiled with Intel 17 in Release mode.
- FftTest should not use thread-unsafe RNG in parallel section. Otherwise the Intel 16 compiler crashes.
- Revert to -openmp on Intel < 16.0.3, which crashes while processing some of our OpenMP loops (e.g. BoundaryHandling::treatDirection).

Also adds OpenMP support for macOS (requires Xcode 7.0, CMake 3.12 and separate libomp from MacPorts/Homebrew)