CUDA中的恒定内存的动态分配 [英] Dynamic Allocation of Constant memory in CUDA

问题描述

我试图利用常量内存，但我很难知道如何嵌套数组。我所拥有的是一个数据数组，它对内部数据进行计数，但对于每个条目，这些数据是不同的。因此，基于以下简化代码，我有两个问题。首先我不知道如何分配我的数据结构的成员指向的数据。第二，因为我不能使用cudaGetSymbolAddress作为常量内存我不知道我是否可以传递全局指针（你不能用普通的__device__内存）。

I'm trying to take advantage of the constant memory, but I'm having a hard time figuring out how to nest arrays. What I have is an array of data that has counts for internal data but those are different for each entry. So based around the following simplified code I have two problems. First I don't know how to allocate the data pointed to by the members of my data structure. Second, since I can't use cudaGetSymbolAddress for constant memory I'm not sure if I can just pass the global pointer (which you cannot do with plain __device__ memory).

struct __align(16)__ data{
int nFiles;
int nNames;
int* files;
int* names;
};

__device__ __constant__ data *mydata;

__host__ void initMemory(...)
{
    cudaMalloc( (void **) &(mydata), sizeof(data)*dynamicsize );
    for(int i=; i lessthan dynamicsize; i++)
    {
        cudaMemcpyToSymbol(mydata, &(nFiles[i]), sizeof(int), sizeof(data)*i, cudaMemcpyHostToDevice);
        //...
        //Problem 1: Allocate & Set mydata[i].files
    }
}

__global__ void myKernel(data *constDataPtr)
{
    //Problem 2: Access constDataPtr[n].files, etc
}

int main()
{
    //...
    myKernel grid, threads (mydata);
}

感谢您提供任何帮助。 ： - ）

Thanks for any help offered. :-)

推荐答案

我认为常数内存是64K，你不能使用CudaMalloc动态分配。必须声明为常数，例如

I think constant memory is 64K and you cannot allocate it dynamically using CudaMalloc. It has to be declared constant, say,

__device__ __constant__ data mydata[100];

同样，你也不需要释放它。此外，你不应该通过指针传递它的引用，只是作为一个全局变量访问它。我试着做一个类似的事情，它给了我segfault（在devicemu）。

Similarly you also don't need to free it. Also, you shouldn't pass the reference to it via pointer, just access it as a global variable. I tried doing a similar thing and it gave me segfault (in devicemu).

这篇关于CUDA中的恒定内存的动态分配的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！