result - O0

(pytest)pupssman@dirigible:/trash$ for n in 1 2 4 8; do g++ -Dswapfunc=swap$n -O0 test.cpp && ./a.out; done
Max size is18446744073709551615, will use 1073741824 bytes for test
Took 2541397 microsec for swap1
Max size is18446744073709551615, will use 1073741824 bytes for test
Took 2391349 microsec for swap2
Max size is18446744073709551615, will use 1073741824 bytes for test
Took 2280239 microsec for swap4
Max size is18446744073709551615, will use 1073741824 bytes for test
Took 2297439 microsec for swap8

result - O1

(pytest)pupssman@dirigible:/trash$ for n in 1 2 4 8; do g++ -Dswapfunc=swap$n -O1 test.cpp && ./a.out; done
Max size is18446744073709551615, will use 1073741824 bytes for test
Took 770556 microsec for swap1
Max size is18446744073709551615, will use 1073741824 bytes for test
Took 625807 microsec for swap2
Max size is18446744073709551615, will use 1073741824 bytes for test
Took 537648 microsec for swap4
Max size is18446744073709551615, will use 1073741824 bytes for test
Took 551631 microsec for swap8

result - O2

(pytest)pupssman@dirigible:/trash$ for n in 1 2 4 8; do g++ -Dswapfunc=swap$n -O2 test.cpp && ./a.out; done
Max size is18446744073709551615, will use 1073741824 bytes for test
Took 775289 microsec for swap1
Max size is18446744073709551615, will use 1073741824 bytes for test
Took 532047 microsec for swap2
Max size is18446744073709551615, will use 1073741824 bytes for test
Took 522245 microsec for swap4
Max size is18446744073709551615, will use 1073741824 bytes for test
Took 587368 microsec for swap8

result - O3

(pytest)pupssman@dirigible:/trash$ for n in 1 2 4 8; do g++ -Dswapfunc=swap$n -O3 test.cpp && ./a.out; done
Max size is18446744073709551615, will use 1073741824 bytes for test
Took 281666 microsec for swap1
Max size is18446744073709551615, will use 1073741824 bytes for test
Took 281743 microsec for swap2
Max size is18446744073709551615, will use 1073741824 bytes for test
Took 282381 microsec for swap4
Max size is18446744073709551615, will use 1073741824 bytes for test
Took 536043 microsec for swap8

system props

(pytest)pupssman@dirigible:/trash$ dpkg -l | grep gcc
ii  gcc                                                   4:4.8.2-1ubuntu6                                    amd64        GNU C compiler
ii  gcc-4.7                                               4.7.3-12ubuntu1                                     amd64        GNU C compiler
ii  gcc-4.7-base:amd64                                    4.7.3-12ubuntu1                                     amd64        GCC, the GNU Compiler Collection (base package)
ii  gcc-4.8                                               4.8.2-19ubuntu1                                     amd64        GNU C compiler
ii  gcc-4.8-base:amd64                                    4.8.2-19ubuntu1                                     amd64        GCC, the GNU Compiler Collection (base package)
ii  gcc-4.8-base:i386                                     4.8.2-19ubuntu1                                     i386         GCC, the GNU Compiler Collection (base package)
ii  gcc-4.9-base:amd64                                    4.9-20140406-0ubuntu1                               amd64        GCC, the GNU Compiler Collection (base package)
ii  gcc-4.9-base:i386                                     4.9-20140406-0ubuntu1                               i386         GCC, the GNU Compiler Collection (base package)
ii  libgcc-4.7-dev:amd64                                  4.7.3-12ubuntu1                                     amd64        GCC support library (development files)
ii  libgcc-4.8-dev:amd64                                  4.8.2-19ubuntu1                                     amd64        GCC support library (development files)
ii  libgcc1:amd64                                         1:4.9-20140406-0ubuntu1                             amd64        GCC support library
ii  libgcc1:i386                                          1:4.9-20140406-0ubuntu1                             i386         GCC support library
(pytest)pupssman@dirigible:/trash$ uname -a
Linux dirigible 3.13.0-24-generic #47-Ubuntu SMP Fri May 2 23:30:00 UTC 2014 x86_64 x86_64 x86_64 GNU/Linux
(pytest)pupssman@dirigible:/trash$ 

test.cpp

#define __STDC_LIMIT_MACROS
#include <iostream>
#include <stdlib.h>
#include <stdint.h>
#include <ctime>

#ifndef swapfunc
#define swapfunc swap1
#endif

#define STRINGY(X) #X
#define STRINGYMACRO(X) STRINGY(X)

using namespace std;


void fill(char* start, size_t len) {
    for (int i = 0; i < len; i ++) {
        *(start + i) = 'A';
    }
}

void swap1(char* src, char* dst, size_t len) {for(size_t i = 0; i < len; i += 1) {*dst++ = *src++;}}
void swap2(char* src, char* dst, size_t len) {for(size_t i = 0; i < len; i += 2) {*dst++ = *src++; *dst++ = *src++;}}
void swap4(char* src, char* dst, size_t len) {for(size_t i = 0; i < len; i += 4) {*dst++ = *src++; *dst++ = *src++; *dst++ = *src++; *dst++ = *src++;}}
void swap8(char* src, char* dst, size_t len) {for(size_t i = 0; i < len; i += 8) {*dst++ = *src++; *dst++ = *src++; *dst++ = *src++; *dst++ = *src++; *dst++ = *src++; *dst++ = *src++; *dst++ = *src++; *dst++ = *src++;}}

int main() {
    size_t len = 1024 * 1024 * 1024;
    cout << "Max size is" << SIZE_MAX << ", will use "<< len << " bytes for test" << endl; 
    
    char* src = (char*) malloc(len); 
    char* dst = (char*) malloc(len);


    fill(src, len);
    clock_t start = clock();
    swapfunc(src, dst, len);
    clock_t end = clock();
    cout << "Took " << (unsigned int)(end - start) << " microsec for "<< STRINGYMACRO(swapfunc) << endl; 
}