#include <iostream>
#include <cstdlib>
#include <vector>
#include <omp.h>
#include <complex>
#include <chrono>

#define N 10000;

using namespace std;

__global__ void check(int *array, int *result){

    int tid = threadIdx.x + blockIdx.x *blockDim.x;

    if(tid < N){
        for(int i = 2; i < sqrt((double)array[tid]); i++){
            if(array[tid] % i == 0){
                result[tid] = 0;
            }
        }
    }
}

int main(){
    srand(time(nullptr));
    int tab[N];
//    int sum = 0;

    for(int i = 0; i < N; i++){
        tab[i] = rand()%N;
    }

    cudaEvent_t start, end;
    float time;

    cudaEventCreate(&start);
    cudaEventCreate(&end);

    int *array= nullptr;
    int *check = nullptr;

    cudaMalloc(&array, N * sizeof(int));
    cudaMalloc(&check, N * sizeof(int));

    cudaMemcpy(array, tab, sizeof(int)*N, cudaMemcpyHostToDevice);

    cudaEventRecord(start);
    check<<<N, 1>>>(array, check);
    cudaEventRecord(end);

    int result[N];
    cudaMemcpy(result, check, N * sizeof(int), cudaMemcpyDeviceToHost);
    cudaEventSynchronize(end);
    cudaEventElapsedTime(&time, start, end);

//    for(int i = 0; i < N; i++){
//        sum = sum + tab[i];
//    }

    cudaFree(array);
    cudaFree(check);

    cout << sum << endl;
    cout << time << endl;

    return 0;
}