Simulation with two internal mesh bodies crashes in parallel

Describe the bug
Crash when running an xml with two internal mesh bodies in parallel (any number of ranks >1).

To Reproduce
Steps to reproduce the behavior:

1. Check out feature/andrembcosta/multiBodyVTKOutput branch if https://github.com/GEOSX/GEOSX/pull/2027 is not yet merged, or develop
2. Run mpirun -n 2 geosx -i inputFiles/multipleMeshBodies/testMultipleBodies.xml -x 2
3. Get a crash on rank 1:

Received signal 8: Floating point exception
Floating point exception:

** StackTrace of 9 frames **
Frame 0: /lib/x86_64-linux-gnu/libc.so.6 
Frame 1: void geosx::InternalMeshGenerator::getNodePosition<LvArray::ArraySlice<double, 1, 0, int> >(int const (&) [3], int, LvArray::ArraySlice<double, 1, 0, int>&&) 
Frame 2: geosx::InternalMeshGenerator::generateMesh(geosx::DomainPartition&) 
Frame 3: geosx::MeshManager::generateMeshes(geosx::DomainPartition&) 
Frame 4: geosx::ProblemManager::generateMesh() 
Frame 5: geosx::ProblemManager::problemSetup() 
Frame 6: geosx::GeosxState::initializeDataRepository() 
Frame 7: main 
Frame 8: __libc_start_main 
Frame 9: _start 
=====

--------------------------------------------------------------------------
MPI_ABORT was invoked on rank 1 in communicator MPI_COMM_WORLD
with errorcode 1.

NOTE: invoking MPI_ABORT causes Open MPI to kill all MPI processes.
You may or may not see output from other processes, depending on
exactly when Open MPI kills them.
--------------------------------------------------------------------------

Expected behavior
Should run (it does in serial).

Platform (please complete the following information):

• Machine: desktop, Ubuntu 20.04
• Compiler: gcc 8.4.0
• GEOSX version 0.2.0 (feature/andrembcosta/multiBodyVTKOutput, sha1: 37dc49e6c)

Additional context
I suspect (haven't investigated yet), that this is related to two different instances of InternalMeshGenerator interacting with the same SpatialPartition object. Each one tries to call partition.setSizes with its own bounds. I assume SpatialPartition was not written with multiple mesh bodies in mind and does something invalid. If either of the two meshes is commented out, mesh generation completes without a crash. It's not immediately obvious to me how the domain and each of the bodies should be partitioned in the general case when their physical extents don't match.

Reproduce with mpirun -n 2 geosx -i inputFiles/multipleMeshBodies/testMultipleBodies.xml -x 2 and inspect the stack path through InternalMeshGenerator::getNodePosition, generateMesh, and MeshManager::generateMeshes. Investigate how multiple InternalMeshGenerator instances use the shared SpatialPartition and call setSizes. Done means the two internal mesh bodies complete mesh generation in parallel without the rank-1 floating-point crash.