Merge branch 'JK/MPI' into 'develop'

src/Examples/CMakeLists.txt

src/Examples/tnl-optimize-ranks-cuda.cu

src/Examples/tnl-optimize-ranks.cpp

src/Python/pytnl/tnl_mpi/tnl_mpi.cpp

src/TNL/Algorithms/detail/DistributedScan.h

src/TNL/Benchmarks/Utils.h

src/TNL/Containers/DistributedArray.h

src/TNL/Containers/DistributedArray.hpp

src/TNL/Containers/DistributedArrayView.h

src/TNL/Containers/DistributedArrayView.hpp

src/TNL/Containers/DistributedNDArray.h

src/TNL/Containers/DistributedNDArraySynchronizer.h

src/TNL/Containers/DistributedNDArrayView.h

...L/Containers/Expressions/DistributedExpressionTemplates.h

...NL/Containers/Expressions/DistributedVerticalOperations.h

src/TNL/Containers/Partitioner.h

src/TNL/Functions/MeshFunctionIO.h

src/TNL/MPI.h

src/TNL/MPI/Comm.h

src/TNL/MPI/DummyDefs.h