OpenCL 内核中的自定义类型

发布于 2024-08-25 05:55:06 字数 496 浏览 6 评论 0原文

是否可以在 OpenCL 内核中使用自定义类型，例如 gmp 类型（mpz_t、mpq_t，...）？

要拥有这样的东西（这个内核不只是因为#include而构建）：

#include <gmp.h>
__kernel square(
   __global mpz_t* input,
   __global mpz_t number,
   __global int* output,
   const unsigned int count)
{
   int i = get_global_id(0);
   if(i < count)
       output[i] = mpz_divisible_p(number,input[i]);
}

也许通过向第四个参数添加不同的参数（选项 ) 的 clBuildProgram ？

或者 OpenCL 是否已经拥有可以处理大数字的类型？

原文

Is it possible to use custom types in OpenCL kernel like gmp types (mpz_t, mpq_t, …) ?

To have something like this (this kernel doesn't build just because of #include <gmp.h>) :

#include <gmp.h>
__kernel square(
   __global mpz_t* input,
   __global mpz_t number,
   __global int* output,
   const unsigned int count)
{
   int i = get_global_id(0);
   if(i < count)
       output[i] = mpz_divisible_p(number,input[i]);
}

Maybe by adding different arguments to the fourth parameter (options) of clBuildProgram ?

Or does OpenCL already have types that can handle large numbers ?

分享到QQ

分享到微博

如果你对这篇内容有疑问，欢迎到本站社区发帖提问参与讨论，获取更多帮助，或者扫码二维码加入 Web 技术交流群。

发布评论

需要登录才能够评论，你可以免费注册一个本站的账号。

云胡 2024-09-01 05:55:06

一般来说，您可以在 OpenCL 程序中使用任何类型。但由于导入不起作用，您必须在同一程序中重新定义它们。例如：

typedef char my_char[8];

typedef struct tag_my_struct
{
    long int        id;
    my_char         chars[2];
    int             numerics[4]
    float           decimals[4];
} my_struct;

__kernel void foo(__global my_struct * input,
                  __global int * output)
{
    int gid = get_global_id(0);
    output[gid] = input[gid].numerics[3]== 2 ? 1 : 0;
}

但是，您显然需要保持 OpenCL 内部和外部的定义相同。还要确保该类型在设备和主机上具有相同的大小（使用 sizeof(my_struct) 应该可以解决问题）。在某些情况下，我必须调整定义，以获得匹配的大小。

Generally you can use any types in an OpenCL program. But since imports do not work, you have to re-define them within the same program. For example:

typedef char my_char[8];

typedef struct tag_my_struct
{
    long int        id;
    my_char         chars[2];
    int             numerics[4]
    float           decimals[4];
} my_struct;

__kernel void foo(__global my_struct * input,
                  __global int * output)
{
    int gid = get_global_id(0);
    output[gid] = input[gid].numerics[3]== 2 ? 1 : 0;
}

However, you obviously need to keep the definitions within and outside OpenCL the same. Also make sure the type has the same size on both device and host (using a sizeof(my_struct) should do the trick). In some cases I had to adjust the definitions, to have matching sizes.

回复收藏 0 原文

忆伤 2024-09-01 05:55:06

我使用 VHristov 的答案和 Dietr 的评论来让我的工作正常进行。此代码适用于 OpenCL 1.2

kernel

typedef struct tag_my_struct{
  int a;
  char b;
}my_struct;

__kernel void myKernel(__global my_struct *myStruct)
{
    int gid = get_global_id(0);
    (myStruct+gid)->a = gid;
    (myStruct+gid)->b = gid + 1;
}

host

typedef struct tag_my_struct{
  cl_int a;
  cl_char b;
}my_struct;

void runCode() 
{
    cl_int status = 0;
    my_struct* ms = new my_struct[5];

    cl_mem mem = clCreateBuffer(*context, 0, sizeof(my_struct)*5, NULL, &status);
    clEnqueueWriteBuffer(*queue, mem, CL_TRUE, 0, sizeof(my_struct)*5, &ms, 0, NULL, NULL);

    status = clSetKernelArg(*kernel, 0, sizeof(ms), &mem);

    size_t global[] = {5};
    status = clEnqueueNDRangeKernel(*queue, *kernel, 1, NULL, global, NULL, 0, NULL, NULL);

    status = clEnqueueReadBuffer(*queue, mem, CL_TRUE, 0, sizeof(my_struct)*5, ms, 0, NULL, NULL);

    for(int i = 0; i < 5; i++)
        cout << (ms+i)->a << " " << (ms+i)->b << endl;
}

output

0 ☺

1 ☻

2 ♥

3 ♦

4 ♣

I used VHristov's answer and dietr's comment to get mine working. This code works for me in OpenCL 1.2

kernel

typedef struct tag_my_struct{
  int a;
  char b;
}my_struct;

__kernel void myKernel(__global my_struct *myStruct)
{
    int gid = get_global_id(0);
    (myStruct+gid)->a = gid;
    (myStruct+gid)->b = gid + 1;
}

host

typedef struct tag_my_struct{
  cl_int a;
  cl_char b;
}my_struct;

void runCode() 
{
    cl_int status = 0;
    my_struct* ms = new my_struct[5];

    cl_mem mem = clCreateBuffer(*context, 0, sizeof(my_struct)*5, NULL, &status);
    clEnqueueWriteBuffer(*queue, mem, CL_TRUE, 0, sizeof(my_struct)*5, &ms, 0, NULL, NULL);

    status = clSetKernelArg(*kernel, 0, sizeof(ms), &mem);

    size_t global[] = {5};
    status = clEnqueueNDRangeKernel(*queue, *kernel, 1, NULL, global, NULL, 0, NULL, NULL);

    status = clEnqueueReadBuffer(*queue, mem, CL_TRUE, 0, sizeof(my_struct)*5, ms, 0, NULL, NULL);

    for(int i = 0; i < 5; i++)
        cout << (ms+i)->a << " " << (ms+i)->b << endl;
}

output

0 ☺

1 ☻

2 ♥

3 ♦

4 ♣

回复收藏 0 原文