view example/fft/cuda/transpose.cu @ 2048:6796d85f3d6b draft

remove error
author Masataka Kohagura <kohagura@cr.ie.u-ryukyu.ac.jp>
date Thu, 28 Jan 2016 00:05:49 +0900
parents 4cf85b48ab9e
children
line wrap: on
line source

extern "C" {
    __global__ void 
    transpose(long* param, float* src, float* dst)
    {
        unsigned long xgid = blockIdx.x*blockDim.x+threadIdx.x; // (unsigned long)s->get_param(0);
        unsigned long ygid = blockIdx.y*blockDim.y*threadIdx.y; // (unsigned long)s->get_param(1);

        long n = param[0];

        unsigned int iid = ygid * n + xgid;
        unsigned int oid = xgid * n + ygid;

        dst[2*oid] = src[2*iid];
        dst[2*oid+1] = src[2*iid+1];
    }
}