Function Ptr Cuda Kernel

vec_op.cu

//  CUDA function ptr example
//  For basic vector operations
//  Based on Chapter 2 of Programming Massively Parallel Processors 3rd Edition Kirk & Hwu

#include <stdio.h>
#include <cuda.h>

__global__ void addKernel(float * a, float * b, float * c, int n) {
    int i = blockDim.x * blockIdx.x  + threadIdx.x; 
    if(i < n) {
        c[i] = a[i] + b[i];
    }
}

// Exercises:
// Subtraction
// Dot Product
// Add two indices simultaneously

void vec_op(void (*kernel)(float*, float*, float*, int), float * h_a, float * h_b, float * h_c, int n) {
    int size = n * sizeof(float);
    float *d_a, *d_b, *d_c;
    
    cudaMalloc((void**) &d_a, size);
    cudaMemcpy(d_a, h_a, size, cudaMemcpyHostToDevice);
    cudaMalloc((void**) &d_b, size);
    cudaMemcpy(d_b, h_b, size, cudaMemcpyHostToDevice);
   
    cudaMalloc((void**) &d_c, size);
    
    kernel<<<ceil(n/256.0),256>>>(d_a, d_b, d_c, n);
   
    cudaMemcpy(h_c, d_c, size, cudaMemcpyDeviceToHost);
 
    cudaFree(d_a);
    cudaFree(d_b);
    cudaFree(d_c);
}

int main(void) {
    int n = 100;
    float * a_h = (float*) malloc(sizeof(float) * n);    

    float * b_h = (float*) malloc(sizeof(float) * n);

    // randomize vector creation
    for(int i = 0; i < n; i++) {
        a_h[i] = 1.5*i;
        b_h[i] = 2.0*i;
    }
    float * c_h = (float*) malloc(sizeof(float) * n);
    vec_op(&addKernel, a_h, b_h, c_h, n);

    for(int i = 0; i < n; i++) {
        printf("%f \n", c_h[i] );
    }
    return 0;
}

// By Joshua Mathews//  CUDA function ptr example
//  For basic vector operations
//  Based on Chapter 2 of Programming Massively Parallel Processors 3rd Edition Kirk & Hwu

#include <stdio.h>
#include <cuda.h>

__global__ void addKernel(float * a, float * b, float * c, int n) {
    int i = blockDim.x * blockIdx.x  + threadIdx.x; 
    if(i < n) {
        c[i] = a[i] + b[i];
    }
}

// Exercises:
// Subtraction
// Dot Product
// Add two indices simultaneously

void vec_op(void (*kernel)(float*, float*, float*, int), float * h_a, float * h_b, float * h_c, int n) {
    int size = n * sizeof(float);
    float *d_a, *d_b, *d_c;
    
    cudaMalloc((void**) &d_a, size);
    cudaMemcpy(d_a, h_a, size, cudaMemcpyHostToDevice);
    cudaMalloc((void**) &d_b, size);
    cudaMemcpy(d_b, h_b, size, cudaMemcpyHostToDevice);
   
    cudaMalloc((void**) &d_c, size);
    
    kernel<<<ceil(n/256.0),256>>>(d_a, d_b, d_c, n);
   
    cudaMemcpy(h_c, d_c, size, cudaMemcpyDeviceToHost);
 
    cudaFree(d_a);
    cudaFree(d_b);
    cudaFree(d_c);
}

int main(void) {
    int n = 100;
    float * a_h = (float*) malloc(sizeof(float) * n);    

    float * b_h = (float*) malloc(sizeof(float) * n);

    // TODO: randomize vector creation
    for(int i = 0; i < n; i++) {
        a_h[i] = 1.5*i;
        b_h[i] = 2.0*i;
    }
    float * c_h = (float*) malloc(sizeof(float) * n);
    vec_op(&addKernel, a_h, b_h, c_h, n);

    for(int i = 0; i < n; i++) {
        printf("%f \n", c_h[i] );
    }
    return 0;
}

// By Joshua Mathews