Question

以下是使用Cuda编译工具，版本6.5，V6.5.12编译的cudaInvalidSymbol sync_and_check("cudaMemcpyToSymbol")的最小示例：

#include<stdio.h>

__inline __host__ void gpuAssert(cudaError_t code, const char* command, const char *file, int line, 
                 bool abort=true)
{
   if (code != cudaSuccess) 
   {
      fprintf(stderr,"Error while calling %s in %s line %d: %s\n", command,
          file, line, cudaGetErrorString(code));
      if (abort) exit(code);
   }
}

#define sync_and_check(command) { gpuAssert(cudaGetLastError(), command, __FILE__, __LINE__); }

struct S
{
    float one;
};

__device__ __constant__ S d_s;

__global__ void kernel(float* f)
{
    (*f) = d_s.one;
}

int main()
{
    cudaGetLastError();

    S s;
    s.one = 1.f;
    cudaMemcpyToSymbol(&d_s, &s, sizeof(S));
    sync_and_check("cudaMemcpyToSymbol");

    float* d_f;
    cudaMalloc(&d_f, sizeof(float));
    sync_and_check("cudaMalloc");

    dim3 dimGrid(1), dimBlock(32);
    kernel<<<dimGrid, dimBlock>>>(d_f);
    sync_and_check("kernel");

    cudaFree(d_f);
    sync_and_check("cudaFree");
}

Answer 1

问题在于引用。正确的调用是

cudaMemcpyToSymbol(d_s, &s, sizeof(S));

我很困惑，因为手册页上写着

cudaError_t cudaMemcpyToSymbol (const void * symbol, const void * src,
   size_t count, size_t offset = 0, enum cudaMemcpyKind kind =
   cudaMemcpyHostToDevice)
   Copies count bytes from the memory area pointed to by src to the memory
   area pointed to by offset bytes from the start of symbol symbol. The
   memory areas may not overlap. symbol is a variable that resides in
   global or constant memory space. kind can be either
   cudaMemcpyHostToDevice or cudaMemcpyDeviceToDevice.

   Parameters:
       symbol - Device symbol address

无论他们的意思是什么＆＃34;符号＆＃34; :)复制到struct元素

    cudaMemcpyToSymbol(d_s.one, &(s.one), sizeof(S));

同样适用。

cudaInvalidSymbol由于cudaMemcpyToSymbol到结构

1 个答案: