OpenCL 简单矩阵乘法没有返回正确的结果

如何解决OpenCL 简单矩阵乘法没有返回正确的结果

我正在尝试使用 OpenCL 和 C++ 主机将 2 个方阵 (32x32) 相乘。我正在尝试从一本书（OpenCL Programming By Example - R Banger,K Bhattacharyya）和 here 的基础中重现结果。然而，结果是错误的。我检查了代码的所有部分，然后推断出“enqueueNDRangeKernel”似乎设置错误。谁能帮我解决这个瓶颈？我使用的是 NVIDIA MX250 显卡，但我猜这段代码是在集成的英特尔 GPU 上运行的。代码如下：

#pragma comment(lib,"OpenCL.lib")

#include <iostream>
#include <CL/cl.hpp>
#include <vector>
#include <chrono>

using namespace std;

int main()
{
    int i;
    int dim = 1024;
    float* A = (float*)malloc(sizeof(float) * dim * dim);
    float* B = (float*)malloc(sizeof(float) * dim * dim);
    float* C = (float*)malloc(sizeof(float) * dim * dim);
    for (i = 0; i < dim * dim; i++)
    {
        A[i] = (float)(rand() % 10);
        B[i] = (float)(rand() % 10);
        C[i] = 0;
    }

    //get all platforms (drivers)
    std::vector<cl::Platform> all_platforms;
    cl::Platform::get(&all_platforms);
    if (all_platforms.size() == 0) {
        std::cout << " No platforms found. Check OpenCL installation!\n";
        exit(1);
    }
    cl::Platform default_platform = all_platforms[0];
    std::cout << "Using platform: " << default_platform.getInfo<CL_PLATFORM_NAME>() << "\n";

    //get default device of the default platform
    std::vector<cl::Device> all_devices;
    default_platform.getDevices(CL_DEVICE_TYPE_ALL,&all_devices);
    if (all_devices.size() == 0) {
        std::cout << " No devices found. Check OpenCL installation!\n";
        exit(1);
    }
    cl::Device default_device = all_devices[0];
    std::cout << "Using device: " << default_device.getInfo<CL_DEVICE_NAME>() << "\n";


    cl::Context context({ default_device });

    cl::Program::Sources sources;

    // kernel calculates for each element C=A+B
    std::string kernel_code =
        "   void kernel simple_add(global const float* A,global const float* B,global float* C,int dim){       "
        "       int iCol = get_global_id(0);                            "
        "       int iRow = get_global_id(1);                            "
        "       float result = 0.0;                                     "
        "       for (int i = 0; i < dim; ++i) "
        "       { "
        "           result += A[iRow * dim + i] * B[i * dim + iCol]; "
        "       }   "
        "           "
        "       C[iRow * dim + iCol] = result; "
        "   }                                                                               ";

    sources.push_back({ kernel_code.c_str(),kernel_code.length() });

    cl::Program program(context,sources);
    if (program.build({ default_device }) != CL_SUCCESS) {
        std::cout << " Error building: " << program.getBuildInfo<CL_PROGRAM_BUILD_LOG>(default_device) << "\n";
        exit(1);
    }


    // create buffers on the device
    cl::Buffer buffer_A(context,CL_MEM_READ_WRITE,sizeof(int) * dim);
    cl::Buffer buffer_B(context,sizeof(int) * dim);
    cl::Buffer buffer_C(context,sizeof(int) * dim);

    //create queue to which we will push commands for the device.
    cl::CommandQueue queue(context,default_device);

    //write arrays A and B to the device
    queue.enqueueWriteBuffer(buffer_A,CL_TRUE,sizeof(float) * dim,A);
    queue.enqueueWriteBuffer(buffer_B,B);

    //run the kernel

    cl::Kernel simple_add(program,"simple_add");
    simple_add.setArg(0,buffer_A);
    simple_add.setArg(1,buffer_B);
    simple_add.setArg(2,buffer_C);
    simple_add.setArg(3,dim);

    cl::NDRange global(32,32);

    queue.enqueueNDRangeKernel(simple_add,cl::NullRange,global,cl::NullRange);
    queue.finish();

    //  float C[10];
        //read result C from the device to array C
    queue.enqueueReadBuffer(buffer_C,C);

    for (int i = 0; i < dim; i++) {
        std::cout << C[i] << " ";
        C[i] = 0.0;
    }
}

解决方法

我终于找到了错误。问题是我传递了dim = 1024，即线性化矩阵（向量）的长度。它应该是 32。以防万一它在将来的某个时候帮助某人。感谢所有观看过尝试回答的人。