性能计数器(performance counter)的API？

cudaEvent_t start,stop;
cudaEventCreate(&start);
cudaEventCreate(&stop);
cudaEventRecord(start,0);
{
//统计的代码段
…………
}
cudaEventRecord(stop,0);
float costtime;
cudaEventElapsedTime(&costtime,start,stop);

楼上的 落了一句：
cudaEventRecord(stop,0)
cudaEventSynchronize(stop); //Waits for an event to complete.Record之前的任务
float costtime;
cudaEventElapsedTime(&costtime,start,stop);