__global__ void gpu_fun(T* data, int n, T* result, CUDA_GLOBAL_CALLBACK<T> func)
{
for (int i = threadIdx.x; i < n; i += blockDim.x)
{
result[i] = func(data[i]);
}
}
大家好,经过一翻搜索,终于能把指针传递到核函数了template __global__ void gpu_fun(T* data, int n, T* result, CUDA_GLOBAL_CALLBACK func) { for (int i = threadIdx.x; i < n; i += blockDim.x) { result[i] = func(data[i]); } } 但是这段代码在debug模式下运行正常,在release模式下,计算出错,甚至