Remove pystencils.GPU_DEVICE

• SerialDataHandling now performs the device selection upon construction. It can also be constructed with an explicit device number to deviate from the default selection.
• For ParallelDataHandling, the assignment of devices to MPI ranks should be handled by Walberla by calling cudaSetDevice(). It has selectDeviceBasedOnMpiRank for this purpose. I am not sure it actually calls it -- I think it should be called from MPIManager::initializeMPI. Right now everything probably just ends up on the first GPU.
• The kernel wrapper now determines the correct device by inspecting the fields.
• gpu_indexing_params needs an explicit device number, I don't think any kind of default is reasonable.
• Some tests now iterate over all devices instead of using a default device. This is actually the right thing to do because it tests whether the device selection works correctly.

lbmpy's test_gpu_block_size_limiting.py::test_gpu_block_size_limiting fails since !335 (merged), but that is due to an error in the test, which lbmpy!146 (merged) fixes.

pystencils/boundaries/boundaryhandling.py

@@ -18,6 +18,7 @@ try:
     import waLBerla as wlb
     if wlb.cpp_available:
         from pystencils.datahandling.parallel_datahandling import ParallelDataHandling
+        import cupy.cuda.runtime
     else:
         ParallelDataHandling = None
 except ImportError:
@@ -100,7 +101,7 @@ class BoundaryHandling:
         self.flag_interface = fi if fi is not None else FlagInterface(data_handling, name + "Flags")
 
         if ParallelDataHandling and isinstance(self.data_handling, ParallelDataHandling):
-            array_handler = GPUArrayHandler()
+            array_handler = GPUArrayHandler(cupy.cuda.runtime.getDevice())
         else:
             array_handler = self.data_handling.array_handler