example3.c

#include <omp.h>
#include <stdio.h>

// as code has repeated in two parallel regions,
// we use threadprivate to make single variable accessible accross multiple parallel regions
int tid;
#pragma omp threadprivate(tid)

int main() {
  omp_set_num_threads(4);
  int numt;
  #pragma omp parallel default(shared)
  {
    tid = omp_get_thread_num();
    if(tid==0){
      for(int j=0; j<100000000; j++);
      numt = omp_get_num_threads();
    } 
  }
  #pragma omp parallel default(shared)
  {
    //int tid = omp_get_thread_num();
    printf("Thread ID: %d of %d\n", tid, numt);
  }
}