vector addition in OPENCL /C

Question

I hope you are well, I have a problem with an opencl program, I execute the following program of the vector addition

#define CL_USE_DEPRECATED_OPENCL_1_2APIS
#include <stdio.h>
#include <stdlib.h>
 
#include <CL/cl.h>

#define MAX_SOURCE_SIZE (0x100000)
 
int main(void) {
    // Create the two input vectors
    int i;
    const int LIST_SIZE = 10;
    int *A = (int*)malloc(sizeof(int)*LIST_SIZE);
    int *B = (int*)malloc(sizeof(int)*LIST_SIZE);
    for(i = 0; i < LIST_SIZE; i++) {
        A[i] = i;
        B[i] = LIST_SIZE - i;
    }
 
    // Load the kernel source code into the array source_str
    FILE *fp;
    char *source_str;
    size_t source_size;
 
    fp = fopen("vector_add_kernel.cl", "r");
    if (!fp) {
        fprintf(stderr, "Failed to load kernel.\n");
        exit(1);
    }
    source_str = (char*)malloc(MAX_SOURCE_SIZE);
    source_size = fread( source_str, 1, MAX_SOURCE_SIZE, fp);
    fclose( fp );
 
    // Get platform and device information
    cl_platform_id platform_id = NULL;
    cl_device_id device_id = NULL;   
    cl_uint ret_num_devices;
    cl_uint ret_num_platforms;
    cl_int ret = clGetPlatformIDs(1, &platform_id, &ret_num_platforms);
    ret = clGetDeviceIDs( platform_id, CL_DEVICE_TYPE_CPU, 1, 
            &device_id, &ret_num_devices);
 
    // Create an OpenCL context
    cl_context context = clCreateContext( NULL, 1, &device_id, NULL, NULL, &ret);
 
    // Create a command queue
    cl_command_queue command_queue = clCreateCommandQueueWithProperties(context, device_id, 0, &ret);
 
    // Create memory buffers on the device for each vector 
    cl_mem a_mem_obj = clCreateBuffer(context, CL_MEM_READ_ONLY, 
            LIST_SIZE * sizeof(int), NULL, &ret);
    cl_mem b_mem_obj = clCreateBuffer(context, CL_MEM_READ_ONLY,
            LIST_SIZE * sizeof(int), NULL, &ret);
    cl_mem c_mem_obj = clCreateBuffer(context, CL_MEM_WRITE_ONLY, 
            LIST_SIZE * sizeof(int), NULL, &ret);
 
    // Copy the lists A and B to their respective memory buffers
    ret = clEnqueueWriteBuffer(command_queue, a_mem_obj, CL_TRUE, 0,
            LIST_SIZE * sizeof(int), A, 0, NULL, NULL);
    ret = clEnqueueWriteBuffer(command_queue, b_mem_obj, CL_TRUE, 0, 
            LIST_SIZE * sizeof(int), B, 0, NULL, NULL);
 
    // Create a program from the kernel source
    cl_program program = clCreateProgramWithSource(context, 1, 
            (const char **)&source_str, (const size_t *)&source_size, &ret);
 
    // Build the program
    ret = clBuildProgram(program, 1, &device_id, NULL, NULL, NULL);
 
    // Create the OpenCL kernel
    cl_kernel kernel = clCreateKernel(program, "vector_add", &ret);
 
    // Set the arguments of the kernel
 
   ret = clSetKernelArg(kernel, 0, sizeof(cl_mem), (void *)&a_mem_obj);
    ret = clSetKernelArg(kernel, 1, sizeof(cl_mem), (void *)&b_mem_obj);
    ret = clSetKernelArg(kernel, 2, sizeof(cl_mem), (void *)&c_mem_obj);
 
    // Execute the OpenCL kernel on the list
    size_t global_item_size = LIST_SIZE; // Process the entire lists
    size_t local_item_size = 64; // Divide work items into groups of 64
    ret = clEnqueueNDRangeKernel(command_queue, kernel, 1, NULL, 
            &global_item_size, &local_item_size, 0, NULL, NULL);
 
    // Read the memory buffer C on the device to the local variable C
    int *C = (int*)malloc(sizeof(int)*LIST_SIZE);
    ret = clEnqueueReadBuffer(command_queue, c_mem_obj, CL_TRUE, 0, 
            LIST_SIZE * sizeof(int), C, 0, NULL, NULL);
 
    // Display the result to the screen
    for(i = 0; i < LIST_SIZE; i++)
        printf("%d + %d = %d\n", A[i], B[i], C[i]);
 
    // Clean up
    ret = clFlush(command_queue);
    ret = clFinish(command_queue);
    ret = clReleaseKernel(kernel);
    ret = clReleaseProgram(program);
    ret = clReleaseMemObject(a_mem_obj);
    ret = clReleaseMemObject(b_mem_obj);
    ret = clReleaseMemObject(c_mem_obj);
    ret = clReleaseCommandQueue(command_queue);
    ret = clReleaseContext(context);
    free(A);
    free(B);
    free(C);
    return 0;
}

With the following kernel:

__kernel void vector_add(__global int *A, __global int *B, __global int *C) {
    
    // Get the index of the current element
    int i = get_global_id(0);

    // Do the operation
    C[i] = A[i] + B[i];
    printf("calcule effectué");
}

While running a C program, I'm getting the following result and I can't seem to figure out why

0 + 10 = 714121520

1 + 9 = 21995

2 + 8 = 0

3 + 7 = 0

4 + 6 = 1852255608

5 + 5 = 1768697717

6 + 4 = 1932425826

7 + 3 = 3223151

8 + 2 = 1919885413

9 + 1 = 1953459744

I don't know what is the problem, please any help !!!!

Answer 1

I also get wrong results, and moreover the results are random for each execution. This means that the values in the C array are never overwritten, and the uninitialized, random values in that allocated memory location are printed.

First step to debug is to print ret and see where things go wrong:

• After clGetDeviceIDs I got error -1 (device not found). CL_DEVICE_TYPE_ALL cleared that; then the selected device was my Nvidia GPU.
• After clCreateProgramWithSource I got error -6 (out of host memory). To fix this, use cl_program program = clCreateProgramWithSource(context, 1, (const char**)&source_str, NULL, &ret);.
• After clEnqueueNDRangeKernel I got error -54 (invalid workgroup size). To fix this, make global_item_size a multiple of local_item_size: size_t global_item_size = ((LIST_SIZE+local_item_size-1)/local_item_size)*local_item_size;. Since your intended global size (or rather size of allocated buffers) is not a multiple if the workgroup size, you should also put a guard clause if(i>=10) return; in the kernel; otherwise the kernel may write values in undefined memory space which can cause crashes.

Then, everything works as intended. Here is the entire fixed code:

#define CL_USE_DEPRECATED_OPENCL_1_2APIS
#include <stdio.h>
#include <stdlib.h>
#include <iostream>
#include <CL/cl.h>

#define MAX_SOURCE_SIZE (0x100000)

int main(void) {
    // Create the two input vectors
    int i;
    const int LIST_SIZE = 10;
    int* A = (int*)malloc(sizeof(int)*LIST_SIZE);
    int* B = (int*)malloc(sizeof(int)*LIST_SIZE);
    int* C = (int*)malloc(sizeof(int)*LIST_SIZE);
    for(i = 0; i < LIST_SIZE; i++) {
        A[i] = i;
        B[i] = LIST_SIZE - i;
        C[i] = 0;
    }

    // Load the kernel source code into the array source_str
    size_t source_size;

    const char* source_str =
        "__kernel void vector_add(__global int *A, __global int *B, __global int *C) {\n"
        "    int i = get_global_id(0);\n"
        "    if(i>=10) return;\n"
        "    C[i] = A[i]+B[i];\n"
        "}"
    ;

    // Get platform and device information
    cl_platform_id platform_id = NULL;
    cl_device_id device_id = NULL;
    cl_uint ret_num_devices;
    cl_uint ret_num_platforms;
    cl_int ret = clGetPlatformIDs(1, &platform_id, &ret_num_platforms);
    ret = clGetDeviceIDs(platform_id, CL_DEVICE_TYPE_ALL, 1,
        &device_id, &ret_num_devices);

    // Create an OpenCL context
    cl_context context = clCreateContext(NULL, 1, &device_id, NULL, NULL, &ret);
    std::cout << "context " << ret << std::endl;

    // Create a command queue
    cl_command_queue command_queue = clCreateCommandQueue(context, device_id, 0, &ret);
    std::cout << "queue " << ret << std::endl;

    // Create memory buffers on the device for each vector 
    cl_mem a_mem_obj = clCreateBuffer(context, CL_MEM_READ_ONLY,
        LIST_SIZE * sizeof(int), NULL, &ret);
    cl_mem b_mem_obj = clCreateBuffer(context, CL_MEM_READ_ONLY,
        LIST_SIZE * sizeof(int), NULL, &ret);
    cl_mem c_mem_obj = clCreateBuffer(context, CL_MEM_WRITE_ONLY,
        LIST_SIZE * sizeof(int), NULL, &ret);
    std::cout << "buffer " << ret << std::endl;

    // Copy the lists A and B to their respective memory buffers
    ret = clEnqueueWriteBuffer(command_queue, a_mem_obj, CL_TRUE, 0,
        LIST_SIZE * sizeof(int), A, 0, NULL, NULL);
    ret = clEnqueueWriteBuffer(command_queue, b_mem_obj, CL_TRUE, 0,
        LIST_SIZE * sizeof(int), B, 0, NULL, NULL);
    std::cout << "write " << ret << std::endl;

    // Create a program from the kernel source
    cl_program program = clCreateProgramWithSource(context, 1, (const char**)&source_str, NULL, &ret);
    //cl_program program = clCreateProgramWithSource(context, 1, (const char**)&source_str, (const size_t *)&source_size, &ret);
    std::cout << "program " << ret << std::endl;

    // Build the program
    ret = clBuildProgram(program, 1, &device_id, NULL, NULL, NULL);

    // Create the OpenCL kernel
    cl_kernel kernel = clCreateKernel(program, "vector_add", &ret);
    std::cout << "kernel " << ret << std::endl;

    // Set the arguments of the kernel

    ret = clSetKernelArg(kernel, 0, sizeof(cl_mem), (void*)&a_mem_obj);
    ret = clSetKernelArg(kernel, 1, sizeof(cl_mem), (void*)&b_mem_obj);
    ret = clSetKernelArg(kernel, 2, sizeof(cl_mem), (void*)&c_mem_obj);
    std::cout << "args " << ret << std::endl;

    // Execute the OpenCL kernel on the list
    size_t local_item_size = 64; // Divide work items into groups of 64
    size_t global_item_size = ((LIST_SIZE+local_item_size-1)/local_item_size)*local_item_size; // make global range a multiple of local range
    ret = clEnqueueNDRangeKernel(command_queue, kernel, 1, NULL, &global_item_size, &local_item_size, 0, NULL, NULL);
    std::cout << "run " << ret << std::endl;
    clFinish(command_queue);

    // Read the memory buffer C on the device to the local variable C
    ret = clEnqueueReadBuffer(command_queue, c_mem_obj, CL_TRUE, 0,
        LIST_SIZE * sizeof(int), C, 0, NULL, NULL);
    std::cout << "read " << ret << std::endl;

    // Display the result to the screen
    for(i = 0; i < LIST_SIZE; i++)
        printf("%d + %d = %d\n", A[i], B[i], C[i]);

    // Clean up
    ret = clFlush(command_queue);
    ret = clFinish(command_queue);
    ret = clReleaseKernel(kernel);
    ret = clReleaseProgram(program);
    ret = clReleaseMemObject(a_mem_obj);
    ret = clReleaseMemObject(b_mem_obj);
    ret = clReleaseMemObject(c_mem_obj);
    ret = clReleaseCommandQueue(command_queue);
    ret = clReleaseContext(context);
    free(A);
    free(B);
    free(C);
    return 0;
}

---

Debugging that was quite a hassle. Do yourself a favor and use this OpenCL-Wrapper with C++. This eliminates all the code overhead and the countless possibilities for errors that come with it.