CUDA与C++

1.使用thrust:

http://stackoverflow.com/questions/13185221/cuda-host-object-to-device

http://blog.csdn.net/shenlan282/article/details/8237576

http://blog.csdn.net/shenlan282/article/details/8237586

2.使用条件编译

http://stackoverflow.com/questions/6978643/cuda-and-classes

Define the class in a header that you #include, just like in C++.

Any method that must be called from device code should be defined with both __device__ and__host__ declspecs, including the constructor and destructor if you plan to use new/delete on the device (note new/delete require CUDA 4.0 and a compute capability 2.0 or higher GPU).

You probably want to define a macro like

#ifdef __CUDACC__ #define CUDA_CALLABLE_MEMBER __host__ __device__ #else #define CUDA_CALLABLE_MEMBER #endif 

Then use this macro on your member functions

class Foo { public: CUDA_CALLABLE_MEMBER Foo() {} CUDA_CALLABLE_MEMBER ~Foo() {} CUDA_CALLABLE_MEMBER void aMethod() {} };

The reason for this is that only the CUDA compiler knows __device__ and __host__ -- your host C++ compiler will raise an error.