Sample file showing timings for several alignment implementations.

align.c

// gcc  -fno-inline -std=gnu99 -Wall -O3 align.c -o align -lm -DLIKWID -llikwid -lpthread
// objdump -d align | less (to confirm that the code hasn't been optimized out)
// likwid -m -C2 -g BRANCH align

#include <math.h>
#include <stdio.h>
#include <stdlib.h>

#ifndef ALIGN
#define ALIGN 8
#endif

#ifndef REPEAT
#define REPEAT 1000000
#endif

unsigned int baseline(unsigned int x) {
  return x;
}

unsigned int align_1(unsigned int x) {
  return (x + ALIGN - 1) & ~(ALIGN - 1);
}

unsigned int align_2(unsigned int x) {
  unsigned int boundary = ALIGN;
  while (x > boundary) {
    boundary += ALIGN;
  }
  return boundary;
}

unsigned int align_3(unsigned int x) {
  return ALIGN * ceil((double)x / ALIGN);
}


#ifdef LIKWID
#include <likwid.h>
#else
#define likwid_markerInit()
#define likwid_markerThreadInit()
#define likwid_markerStartRegion(name)
#define likwid_markerStopRegion(name)
#define likwid_markerClose()
#endif // LIKWID


int main(int argc, char **argv) {
  likwid_markerInit();
  likwid_markerThreadInit();

  char *name = "baseline";
  likwid_markerStartRegion(name);
  for (int i = 1; i <= REPEAT; i++) {
    if (baseline(i) == 0) {
      printf("Test failed!");
      exit(1);
    }
  }
  likwid_markerStopRegion(name);

  name = "align_1";
  likwid_markerStartRegion(name);
  for (int i = 1; i <= REPEAT; i++) {
    if (align_1(i) == 0) {
      printf("Test failed!");
      exit(1);
    }
  }
  likwid_markerStopRegion(name);

  name = "align_2";
  likwid_markerStartRegion(name);
  for (int i = 1; i <= REPEAT; i++) {
   if (align_2(i) == 0) {
      printf("Test failed!");
      exit(1);
    }
  }
  likwid_markerStopRegion(name);

  name = "align_3";
  likwid_markerStartRegion(name);
  for (int i = 1; i <= REPEAT; i++) {
    if (align_3(i) == 0) {
      printf("Test failed!");
      exit(1);
    }
  }
  likwid_markerStopRegion(name);

  likwid_markerClose();

  exit(0);
}


#if LIKWID_RESULTS
-------------------------------------------------------------
-------------------------------------------------------------
CPU type:	Intel Core Haswell processor
CPU clock:	3.39 GHz
Measuring group BRANCH
-------------------------------------------------------------
align

=====================
Region: baseline
=====================
+-------------------+------------+
|    Region Info    |   core 2   |
+-------------------+------------+
| RDTSC Runtime [s] | 0.00157839 |
|    call count     |     1      |
+-------------------+------------+
+------------------------------+-------------+
|            Event             |   core 2    |
+------------------------------+-------------+
|      INSTR_RETIRED_ANY       | 9.00103e+06 |
|    CPU_CLK_UNHALTED_CORE     | 5.3404e+06  |
|     CPU_CLK_UNHALTED_REF     | 5.34018e+06 |
| BR_INST_RETIRED_ALL_BRANCHES | 4.00032e+06 |
| BR_MISP_RETIRED_ALL_BRANCHES |     47      |
+------------------------------+-------------+
+----------------------------+-------------+
|           Metric           |   core 2    |
+----------------------------+-------------+
|    Runtime (RDTSC) [s]     | 0.00157839  |
|    Runtime unhalted [s]    | 0.00157435  |
|        Clock [MHz]         |   3392.28   |
|            CPI             |   0.59331   |
|        Branch rate         |  0.444429   |
| Branch misprediction rate  | 5.22163e-06 |
| Branch misprediction ratio | 1.17491e-05 |
|  Instructions per branch   |   2.25008   |
+----------------------------+-------------+

=====================
Region: align_1
=====================
+-------------------+-----------+
|    Region Info    |  core 2   |
+-------------------+-----------+
| RDTSC Runtime [s] | 0.0014741 |
|    call count     |     1     |
+-------------------+-----------+
+------------------------------+-------------+
|            Event             |   core 2    |
+------------------------------+-------------+
|      INSTR_RETIRED_ANY       | 1.0001e+07  |
|    CPU_CLK_UNHALTED_CORE     | 5.00746e+06 |
|     CPU_CLK_UNHALTED_REF     | 5.00752e+06 |
| BR_INST_RETIRED_ALL_BRANCHES | 4.00031e+06 |
| BR_MISP_RETIRED_ALL_BRANCHES |     40      |
+------------------------------+-------------+
+----------------------------+-------------+
|           Metric           |   core 2    |
+----------------------------+-------------+
|    Runtime (RDTSC) [s]     |  0.0014741  |
|    Runtime unhalted [s]    |  0.0014762  |
|        Clock [MHz]         |   3392.1    |
|            CPI             |  0.500695   |
|        Branch rate         |  0.399991   |
| Branch misprediction rate  | 3.99959e-06 |
| Branch misprediction ratio | 9.99922e-06 |
|  Instructions per branch   |   2.50006   |
+----------------------------+-------------+

=====================
Region: align_2
=====================
+-------------------+---------+
|    Region Info    | core 2  |
+-------------------+---------+
| RDTSC Runtime [s] | 18.4446 |
|    call count     |    1    |
+-------------------+---------+
+------------------------------+-------------+
|            Event             |   core 2    |
+------------------------------+-------------+
|      INSTR_RETIRED_ANY       | 1.8751e+11  |
|    CPU_CLK_UNHALTED_CORE     | 6.25503e+10 |
|     CPU_CLK_UNHALTED_REF     | 6.25503e+10 |
| BR_INST_RETIRED_ALL_BRANCHES | 6.25045e+10 |
| BR_MISP_RETIRED_ALL_BRANCHES | 1.0001e+06  |
+------------------------------+-------------+
+----------------------------+-------------+
|           Metric           |   core 2    |
+----------------------------+-------------+
|    Runtime (RDTSC) [s]     |   18.4446   |
|    Runtime unhalted [s]    |   18.4398   |
|        Clock [MHz]         |   3392.14   |
|            CPI             |  0.333583   |
|        Branch rate         |  0.333339   |
| Branch misprediction rate  | 5.33357e-06 |
| Branch misprediction ratio | 1.60004e-05 |
|  Instructions per branch   |   2.99995   |
+----------------------------+-------------+

=====================
Region: align_3
=====================
+-------------------+------------+
|    Region Info    |   core 2   |
+-------------------+------------+
| RDTSC Runtime [s] | 0.00590322 |
|    call count     |     1      |
+-------------------+------------+
+------------------------------+-------------+
|            Event             |   core 2    |
+------------------------------+-------------+
|      INSTR_RETIRED_ANY       | 1.90017e+07 |
|    CPU_CLK_UNHALTED_CORE     | 2.00174e+07 |
|     CPU_CLK_UNHALTED_REF     | 2.00175e+07 |
| BR_INST_RETIRED_ALL_BRANCHES | 7.00045e+06 |
| BR_MISP_RETIRED_ALL_BRANCHES |     64      |
+------------------------------+-------------+
+----------------------------+-------------+
|           Metric           |   core 2    |
+----------------------------+-------------+
|    Runtime (RDTSC) [s]     | 0.00590322  |
|    Runtime unhalted [s]    |  0.0059011  |
|        Clock [MHz]         |   3392.12   |
|            CPI             |   1.05345   |
|        Branch rate         |  0.368412   |
| Branch misprediction rate  | 3.36812e-06 |
| Branch misprediction ratio | 9.14227e-06 |
|  Instructions per branch   |   2.71435   |
+----------------------------+-------------+


#endif