rdtscp_test.cpp

#include <cstdint>
#include <iostream>
//#include <linux/preempt.h>

__attribute__((always_inline)) uint64_t perf_counter(void)
{
  uint32_t lo, hi;
  // take time stamp counter, rdtscp does serialize by itself, and is much cheaper than using CPUID
  __asm__ __volatile__ (
      "rdtscp" : "=a"(lo), "=d"(hi)
      );
  return ((uint64_t)lo) | (((uint64_t)hi) << 32);
}

__attribute__((always_inline)) uint64_t tacc_rdtscp(int *chip, int *core)
{
   unsigned long int x;
   unsigned a, d, c;
   __asm__ volatile("rdtscp" : "=a" (a), "=d" (d), "=c" (c));
       *chip = (c & 0xFFF000)>>12;
       *core = c & 0xFFF;
   return ((uint64_t)a) | (((uint64_t)d) << 32);;

}

int main()
{
  //preempt_disable();
  uint64_t t1 = perf_counter();
  uint64_t t2 = perf_counter();
  //preempt_enable();

  std::cout << t1 << " - " << t2 << ", diff =  " << t2 - t1 << std::endl;
  int chip = 0;
  int core = 0;
  uint64_t t3 = tacc_rdtscp(&chip, &core);
  std::cout << "chip = " << chip << ", core = " << core << std::endl;

  return 0;
}