#include <iostream>
#include <cstdlib>
#include <cstring>
#include <cmath>
#include <ctime>

#include <omp.h>

using namespace std;

float **a; 

float calc(float x) {
  long iter = (long) x; 
  for(long i=0; i<iter; i++) 
    x = sin(x); 
  return(x);
}

int main(int argc, char * argv[]) {

  if(argc == 1) {
    cout << "Usage is:\n" << argv[0] << " veclen numthread " << endl; 
    return(0);
  }

  int n   = atoi(argv[1]); 
  int tn1 = atoi(argv[2]); 
  int tn2 = atoi(argv[3]); 


  float * tim = new float[tn2-tn1+1];

  for(int k=tn1; k<=tn2; k++) {
    a = new float * [n];
    for(int i=0; i<n; i++) 
      a[i] = new float[n];
    
    for(int i=0; i<n; i++) 
      for(int j=0; j<n; j++) 
	a[i][j] = i* 100 + j * 10; 

    float sum = 0.0;
    float ts0 = omp_get_wtime();
#pragma omp parallel for num_threads(k) schedule(SCHED) reduction (+:sum)
    for(int i=0; i<n; i++) 
      for(int j=0; j<n; j++) {
	a[i][j] = calc(a[i][j]);
	sum += a[i][j];
      }
    float ts1 = omp_get_wtime();
    tim[k-1] = ts1 - ts0; 
  }

  for(int i=0; i<tn2-tn1+1; i++) {
    cout << "nt = " << i+1 << "\t tc = " << tim[i] << "\t sp = " << tim[0] / tim[i] << "\t eff = " << 
      (tim[0] / ( (i+1) * tim[i] )) << endl; 
  }
  return(0); 

}