Geek Answers Handbook

Is "Jump if zero" (jz) faster?

I read this article and I noticed the instructions jz. It made me think:

Will this code in the assembly taken at face value

for (int i=max;i!=0;--i){
    //Some operation
}

outperform this code?

for (int i=0;i<max;++i){
    //Some operation
}

As long as you do not care that your data is processed with an increase i, there is no semantic difference. Cached misses should not suffer either, because they can work consistently anyway.

I am not good at assembly to write examples, but I think that the first example will use only jz. The second will use cmp, then a jg, as well as another variable max. In the first example, only a loop counter is needed, since it is 0implicit.

, , , .

+4
3

, cmp a,b.

( ), , jmp.

, jz, cmp jmp.

+2

. , , , , , , . x86 [-64] , , " " ? ? ?

x86. ,

  • j[e|r]cxz, , .

  • sub add j<cond>, dec inc, .

operation . max ?

for (int i = 0; i < max / 4; i++)
{
    operation;
    operation;
    operation;
    operation;
}

operation (, ), n-way scheduling. , .

- GMP. mpn/x86[-64] . Agner Fog - .

0

, , , ( x86-32, ), . , . SSE/AVX/, ,

for (i=end; i;){
    i-=4;
    val[i+0] = foo(i);
    val[i+1] = foo(i);
    val[i+2] = foo(i);
    val[i+3] = foo(i);
    }

, end%4 == 0, .

0

More articles:


All Articles