CUDA By Examples 7 - 测量GPU运行耗时

测量方法:
1. 使用cudaEventCreate创建event;
2. 使用cudaEventRecord记录;
3. 使用cudaEventSynchronize同步, 等待GPU指令完成. 方便读time stamp.

注意:
1. 不能用于测量device和host的混合代码的用时;
2. 只能用于测量GPU内部kernel执行指令和存储拷贝用时.

#include "cuda.h"
#include "../common/book.h"
#include "../common/cpu_bitmap.h"

#define INF 2e10f
#define rnd( x ) (x * rand() / RAND_MAX)
#define SPHERES 20 //球体数量
#define DIM 1024 // bitmap图大小

struct Sphere{
    float r,b,g;
    float radius;
    float x,y,z; //球心

    //从某一个像素射出的垂直于bitmap平面的射线,和当前球体相交,求出最近的一个交点
    //没有交点就返回负无穷
    __device__ float hit(float ox, float oy, float *n)
    {
        float dx = ox - x;
        float dy = oy - y;
        if(dx*dx + dy*dy < radius*radius)
        {
            float dz = sqrtf(radius*radius - dx*dx - dy*dy);
            *n = dz / sqrtf( radius * radius);
            return dz + z;
        }
        return -INF;
    }
};


// 不是指针
// 常量内存的声明不能放在函数体内部.
__constant__ Sphere s[SPHERES];
__global__ void kernel( unsigned char *ptr)
{
    int x = threadIdx.x + blockIdx.x * blockDim.x;
    int y = threadIdx.y + blockIdx.y * blockDim.y;
    int offset = x + y * blockDim.x * gridDim.x;
    float ox = (x - DIM/2);
    float oy = (y - DIM/2);

    float r=0, g=0, b=0;
    float maxz = -INF;
    for (int i=0; i<SPHERES; i++)
    {
        float n;
        float t = s[i].hit(ox, oy, &n);
        if (t > maxz)
        {
            float fscale = n;
            r = s[i].r * fscale;
            g = s[i].g * fscale;
            b = s[i].b * fscale;
        }
    }
    //四个通道赋值
    ptr[offset*4 + 0] = (int)(r*255);
    ptr[offset*4 + 1] = (int)(g*255);
    ptr[offset*4 + 2] = (int)(b*255);
    ptr[offset*4 + 3] = 255;
}

int main(void)
{
    //创建Event, 记录开始时间.
    cudaEvent_t start, stop;
    HANDLE_ERROR( cudaEventCreate( &start ) );
    HANDLE_ERROR( cudaEventCreate( &stop ) );
    HANDLE_ERROR( cudaEventRecord( start, 0 ) );

    CPUBitmap bitmap(DIM, DIM);
    unsigned char *dev_bitmap;


    HANDLE_ERROR( cudaMalloc( (void**)&dev_bitmap,
                                bitmap.image_size() ) );
    //常量内存不需要为之开辟GPU存储空间.
    //HANDLE_ERROR( cudaMalloc( (void**)&s,
    //                          sizeof(Sphere) * SPHERES ) );

    //球体参数在CPU上初始化
    Sphere *temp_s = (Sphere*)malloc( sizeof(Sphere) * SPHERES );
    for (int i=0; i<SPHERES; i++)
    {
        temp_s[i].r = rnd( 1.0f );
        temp_s[i].g = rnd( 1.0f );
        temp_s[i].b = rnd( 1.0f );
        temp_s[i].x = rnd( 1000.0f ) - 500;
        temp_s[i].y = rnd( 1000.0f ) - 500;
        temp_s[i].z = rnd( 1000.0f ) - 500;
        temp_s[i].radius = rnd( 100.0f ) + 20;
    }
    //把Sphere从主机拷贝到GPU上

    HANDLE_ERROR( cudaMemcpyToSymbol( s, temp_s,
                                        sizeof(Sphere) * SPHERES ) );

    free( temp_s );
    // GPU上做处理
    dim3 grids(DIM/16, DIM/16);
    dim3 threads(16, 16);
    kernel<<<grids,threads>>>(dev_bitmap);

    //从GPU拷贝到Host上
    HANDLE_ERROR( cudaMemcpy( bitmap.get_ptr(),
                                dev_bitmap,
                                bitmap.image_size(),
                                cudaMemcpyDeviceToHost ) );

    //计算耗时
    HANDLE_ERROR( cudaEventRecord( stop, 0 ) );
    HANDLE_ERROR( cudaEventSynchronize( stop ) );
    float elapsedTime;
    HANDLE_ERROR( cudaEventElapsedTime( &elapsedTime, start, stop ) );

    printf( "Time to generate: %3.1f ms\n", elapsedTime );

    HANDLE_ERROR( cudaEventDestroy( start ) );
    HANDLE_ERROR( cudaEventDestroy( stop ) );
    //显示
    bitmap.display_and_exit();

    //free memory
    cudaFree( dev_bitmap );
    //cudaFree( s );

}