Intel Inline Assembly loop benchmark

dm-test.c

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

int main()
{
    int j=0;

    int a=0,b=0,temp=0;

    srand(time(0));
    time_t t1=0;
    time_t t2=0;


    t1=clock();
    for(int i=0; i<200000000; i++)
    {
        temp=a;//instruction 1
        a=b;//instruction 2
        b=temp;//3 instructions total
    }
    t2=clock();
    printf("\n time of for-loop(cycles) %li  \n",(t2-t1));


    t1=clock();
    while(j<200000000)
    {
        temp=a;//again it is three instructions
        a=b;
        b=temp;
        j++;
    }
    t2=clock();
    printf("\n time of while-loop(cycles)  %li  \n",(t2-t1));


    t1=clock();
    j=200000000;//setting the count
    __asm
    {
        pushf           //backup
        push eax        //backup
        push ebx        //backup
        push ecx        //backup
        push edx        //backup

        mov ecx,0       //init of loop range(0 to 200000000)
        mov edx,j

        do_it_again:    //begin to loop


        mov eax,a       //basic swap steps between cpu and mem(cache)
        mov ebx,b
        mov b,eax
        mov a,ebx       //four instructions total

        inc ecx         // j++
        cmp ecx,edx     //i<200000000  ?
        jb do_it_again  // end of loop block

        pop edx     //rolling back to history
        pop ecx
        pop ebx
        pop eax
        popf
    }

    t2=clock();
    printf("\n time of custom-loop-1(cycles)   %li   \n",(t2-t1));

    t1=clock();
    j=200000000;//setting the count
    __asm
    {
        pushf           //backup
            push eax        
            push ebx        
            push ecx        
            push edx        

            mov ecx,0       //init of loop range(0 to 200000000)
            mov edx,j

            mov eax,a       //getting variables to registers
            mov ebx,b

            do_it_again2:   //begin to loop

            //swapping with using only 2 variables(only in cpu)
            sub eax,ebx         //a is now a-b
            add ebx,eax         //b is now a
            sub eax,ebx         //a is now -b
            xor eax,80000000h   //a is now b and four instructions total

            inc ecx         // j++
            cmp ecx,edx     //i<200000000  ?
            jb do_it_again2  // end of loop block

            pop edx         //rollback
            pop ecx         
            pop ebx         
            pop eax         
            popf            
    }

    t2=clock();
    printf("\n time of custom-loop-2(cycles)  %li   \n",(t2-t1));

    t1=clock();
    j=200000000;//setting the count
    __asm
    {
        xor ecx,j     //init of loop range(200000000 to 0)

        mov eax,a     //getting variables to registers
        mov ebx,b

    do_it_again3:   //begin to loop

        //swapping with using only 2 variables(only in cpu)
        xor eax,ebx
        xor ebx,eax         
        xor eax,ebx         

        mov a,eax
        mov b,ebx

        dec ecx         // j--
        jnz do_it_again3  // end of loop block
    }

    t2=clock();
    printf("\n time of custom-loop-3(cycles)  %li   \n",(t2-t1));

    return 0;

}

Makefile

all: test # dm-test

CPPFLAGS+=-g -O3 -Wall -std=c99 -masm=intel

%:%.c
	$(CC) $(CPPFLAGS) $^ -o $@

dm-test: dm-test.c
	dmc -O $^

test.c

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

int j=0;
int a=0,b=0,temp=0;
time_t t1=0;
time_t t2=0;
int main()
{

    srand(time(0));

    t1=clock();
    for(int i=0; i<200000000; i++)
    {
        temp=a;
        a=b;
        b=temp;
    }
    t2=clock();
    printf("\n time of for-loop(cycles) %li  \n",(t2-t1));


    t1=clock();
    while(j<200000000)
    {
        temp=a;
        a=b;
        b=temp;
        j++;
    }
    t2=clock();
    printf("\n time of while-loop(cycles)  %li  \n",(t2-t1));


    t1=clock();
    j=200000000;
#define asm __asm
    {
        asm(".intel_syntax noprefix\n");
        ////asm("pushf\n");
        ////asm("push eax\n");
        ////asm("push ebx\n");
        ////asm("push ecx\n");
        ////asm("push edx\n");

        asm("mov ecx,0\n");
        asm("mov edx,j\n");

        asm("do_it_again:\n");


        asm("mov eax,a\n");
        asm("mov ebx,b\n");
        asm("mov b,eax\n");
        asm("mov a,ebx\n");

        asm("inc ecx\n");
        asm("cmp ecx,edx\n");
        asm("jb do_it_again\n");

        //asm("pop edx\n");
        //asm("pop ecx\n");
        //asm("pop ebx\n");
        //asm("pop eax\n");
        //asm("popf\n");
    }

    t2=clock();
    printf("\n time of custom-loop-1(cycles)   %li   \n",(t2-t1));

    t1=clock();
    j=200000000;
    {
        asm(".intel_syntax noprefix\n");
        ////asm("pushf\n");
        ////asm("push eax\n");
        ////asm("push ebx\n");
        ////asm("push ecx\n");
        ////asm("push edx\n");

        asm("mov ecx,0\n");
        asm("mov edx,j\n");

        asm("mov eax,a\n");
        asm("mov ebx,b\n");

        asm("do_it_again2:\n");


        asm("sub eax,ebx\n");
        asm("add ebx,eax\n");
        asm("sub eax,ebx\n");
        asm("xor eax,0x80000000\n");

        asm("inc ecx\n");
        asm("cmp ecx,edx\n");
        asm("jb do_it_again2\n");

        //asm("pop edx\n");
        //asm("pop ecx\n");
        //asm("pop ebx\n");
        //asm("pop eax\n");
        //asm("popf\n");
    }

    t2=clock();
    printf("\n time of custom-loop-2(cycles)  %li   \n",(t2-t1));

    t1=clock();
    j=200000000;
    {
        asm(".intel_syntax noprefix\n");
        asm("xor ecx,j\n");

        asm("mov eax,a\n");
        asm("mov ebx,b\n");

        asm("do_it_again3:\n");


        asm("xor eax,ebx\n");
        asm("xor ebx,eax\n");
        asm("xor eax,ebx\n");

        asm("mov a,eax\n");
        asm("mov b,ebx\n");

        asm("dec ecx\n");
        asm("jnz do_it_again3\n");
    }

    t2=clock();
    printf("\n time of custom-loop-3(cycles)  %li   \n",(t2-t1));

    return 0;

}