main.cu

 
#ifdef __NVCC__
 
#include "Vector/vector_dist.hpp"
#include "util/cuda_util.hpp"
 
 
__global__ void print_data_particle_50(float * scalar, float * vector, float * tensor, size_t capacity)
{
    int p = threadIdx.x + blockIdx.x * blockDim.x;
 
    if (p == 50)
    {
        printf("Scalar particle %d = %f\n",p,scalar[p]);
 
        printf("Vector particle %d = %f\n",p,vector[p]);
        printf("Vector particle %d = %f\n",p,vector[p + capacity]);
 
        printf("Tensor particle %d = %f\n",p,tensor[p + (0*2 + 0)*capacity]);
        printf("Tensor particle %d = %f\n",p,tensor[p + (0*2 + 1)*capacity]);
        printf("Tensor particle %d = %f\n",p,tensor[p + (1*2 + 0)*capacity]);
        printf("Tensor particle %d = %f\n",p,tensor[p + (1*2 + 1)*capacity]);
    }
}
 
 
int main(int argc, char* argv[])
{
    // initialize the library
    openfpm_init(&argc,&argv);
 
    // Here we define our domain a 2D box with internals from 0 to 1.0 for x and y
    Box<2,float> domain({0.0,0.0},{1.0,1.0});
 
    // Here we define the boundary conditions of our problem
    size_t bc[2]={PERIODIC,PERIODIC};
 
    // extended boundary around the domain, and the processor domain
    Ghost<2,float> g(0.05);
 
    vector_dist_gpu<2,float, aggregate<float,float[2],float[2][2]> > vd(100,domain,bc,g);
 
    auto it = vd.getDomainIterator();
 
    while (it.isNext())
    {
        auto p = it.get();
 
        // we define x, assign a random position between 0.0 and 1.0
        vd.getPos(p)[0] = (float)rand() / RAND_MAX;
 
        // we define y, assign a random position between 0.0 and 1.0
        vd.getPos(p)[1] = (float)rand() / RAND_MAX;
 
        vd.template getProp<0>(p) = vd.getPos(p)[0] + vd.getPos(p)[1];
 
        vd.template getProp<1>(p)[0] = vd.getPos(p)[0];
        vd.template getProp<1>(p)[1] = vd.getPos(p)[1];
 
        vd.template getProp<2>(p)[0][0] = vd.getPos(p)[0];
        vd.template getProp<2>(p)[0][1] = vd.getPos(p)[1];
        vd.template getProp<2>(p)[1][0] = vd.getPos(p)[0] + vd.getPos(p)[1];
        vd.template getProp<2>(p)[1][1] = vd.getPos(p)[1] - vd.getPos(p)[0];
 
        // next particle
        ++it;
    }
 
    vd.map();
 
 
 
    std::cout << "First particle property 0, address: " << &vd.template getProp<0>(0) << std::endl;
    std::cout << "First particle property 1, address: " << &vd.template getProp<1>(0)[0] << std::endl;
    std::cout << "First particle property 2, address: " << &vd.template getProp<2>(0)[0][0] << std::endl;
 
 
 
    std::cout << "Capacity internal vector: " << vd.getPropVector().capacity() << std::endl;
 
    std::cout << "First particle property 1 component 0, address: " << &vd.template getProp<1>(0)[0] << std::endl;
    std::cout << "First particle property 1 component 1, address: " << &vd.template getProp<1>(0)[1] << std::endl;
 
    std::cout << "First particle property 2 component 00, address: " << &vd.template getProp<2>(0)[0][0] << std::endl;
    std::cout << "First particle property 2 component 01, address: " << &vd.template getProp<2>(0)[0][1] << std::endl;
    std::cout << "First particle property 2 component 10, address: " << &vd.template getProp<2>(0)[1][0] << std::endl;
    std::cout << "First particle property 2 component 11, address: " << &vd.template getProp<2>(0)[1][1] << std::endl;
 
    std::cout << "Second particle property 1 component 0, address: " << &vd.template getProp<1>(1)[0] << std::endl;
    std::cout << "Second particle property 1 component 1, address: " << &vd.template getProp<1>(1)[1] << std::endl;
 
    std::cout << "Second particle property 2 component 00, address: " << &vd.template getProp<2>(1)[0][0] << std::endl;
    std::cout << "Second particle property 2 component 01, address: " << &vd.template getProp<2>(1)[0][1] << std::endl;
    std::cout << "Second particle property 2 component 10, address: " << &vd.template getProp<2>(1)[1][0] << std::endl;
    std::cout << "Second particle property 2 component 11, address: " << &vd.template getProp<2>(1)[1][1] << std::endl;
 
 
    std::cout << std::endl;
 
 
    vd.template hostToDeviceProp<0,1,2>();
 
    CUDA_LAUNCH_DIM3(print_data_particle_50,100,1,(float *)vd.getPropVector().template getDeviceBuffer<0>(),
                           (float *)vd.getPropVector().template getDeviceBuffer<1>(),
                           (float *)vd.getPropVector().template getDeviceBuffer<2>(),
                           vd.getPropVector().capacity());
 
 
    openfpm_finalize();
}
 
#else
 
int main(int argc, char* argv[])
{
        return 0;
}
 
#endif