- benchmarks ghost layer field communication
- direct vs. buffered, various field & block sizes
- multiple blocks per process
- extensive database output

option is intended for benchmarking, for production runs leave the
local communication optimization switched on

introduce mesa_pd module to official waLBerla

See merge request walberla/walberla!210

Make waLBerla fit for SuperMUC-NG

See merge request walberla/walberla!207

- workaround seems to be not necessary any more, for recent IntelMPI versions
- the workaround did not play with CMake's try_compile function, which tries to compile some test programs to check for compiler features
  The workaround added '-include mpi.h' to the compile flags, which are also used by try_compile, but try_compile does not set include directories with "-I"
  This lead to the problem, that for example the OpenMP flag was not correctly determined, since there the try_compile mechanism is used

Add directory 'executiontree' to .clang-format

See merge request walberla/walberla!206

Fix too narrow CLI boundary condition assertion

See merge request walberla/walberla!205

add Visual Studio 2019 to the ci

See merge request walberla/walberla!204

Don't use __has_include macro if it's not defined

See merge request walberla/walberla!203

- MPI barrier between runs
- bugfix in overlap_benchmark config

- more overlap strategies
- inner/outer split width is configurable

- fix in time measurement
- options to disable boundaries and/or communication
- priority for communication & outer part of kernel