```python
    def to_gpu(gpu_version, cpu_version):
        gpu_version = gpu_version.boundary_object_to_index_list
        cpu_version = cpu_version.boundary_object_to_index_list

>       if isinstance(self.data_handling, ParallelDataHandling):
E       TypeError: isinstance() arg 2 must be a type or tuple of types

../pystencils/pystencils/boundaries/boundaryhandling.py:102: TypeError
```

None is not a type! ParallelDataHandling is None if `waLBerla` is not
installed

- Type does violate the sympy invariant obj.func(*obj.args) == obj
- This seems to be ok when the object is an atom

Fix failing interpolation tests on newer sympy versions

See merge request pycodegen/pystencils!126

SerialDataHandling: synchronization_function for tensor fields

See merge request pycodegen/pystencils!122

Fix {PyCuda,PyOpenCl}ArrayHandler.{zeros,ones,randn}

See merge request pycodegen/pystencils!125

Add dtype to DataHandling.add_arrays

See merge request pycodegen/pystencils!124

Make create_staggered_kernel work with OpenMP

See merge request pycodegen/pystencils!123

Add DataHandling.add_arrays

See merge request pycodegen/pystencils!120

Automatic Finite Volume and Volume-of-Fluid discretization

See merge request pycodegen/pystencils!121

This adds a function for lazy people who want to write

```python
dh = create_data_handling((20, 30))
dh.add_arrays('x, y(9), z')
```

instead of

```python
dh = create_data_handling((20, 30))
dh.add_array('x')
dh.add_array('y', values_per_cell=9)
dh.add_array('z')
```

because in most cases you want to use more than one array.

Opencl datahandling

See merge request pycodegen/pystencils!119

this to your children or students)

PointerType should be rendered differently in OpenCL backend