Use level instead of dridNumber in NodeIndicesMultiGPU
and instead of i in loops in MultipleGridBuilder

CUDA doesn't support std::vector

No need to pass node types anymore

simplified the kernel factory infrastructure by omit the enum (Kerneltype) -> adding new kernels to the factory should be straightforward now

Fix various msvc compiler warnings. With msvc the compiler flags are explicit passed to nvcc with -Xcompiler.

As thrust is used in the gpu code, CUDA c++14 is required. Besides fixed compiler warnings in Parameter.cpp.

Remove VirtualFluidsDefinitions.h header file.
Remove redundant compiler definitions from gpu.cmake.

Make the cmake project more flexible while removing the CMAKE_SOURCE_DIR usage.

Make the cmake project more flexible while removing the CMAKE_SOURCE_DIR usage.

Remove unused stuff and the global include from VirtualFluidsDefinition.h. Adds missing export includes.

Warning: [-Wbackslash-newline-escape]
https://stackoverflow.com/questions/31673945/clang-backslash-and-newline-separated-by-space-in-single-line-comment