为什么 GCC 先将整数加载到 eax，然后再加载到其他一些相关寄存器? [英] Why GCC loads integer first to eax and then to some other relevant registers?

问题描述

使用这个站点，我为一个简单程序生成了汇编代码.

Using this site, I generated assembly code for a simple program.

main:
        push    rbp
        mov     rbp, rsp
        sub     rsp, 16
        mov     DWORD PTR [rbp-4], 5
        mov     eax, DWORD PTR [rbp-4]
        mov     edi, eax
        mov     eax, 0
        call    ifunc
        mov     eax, 0
        leave
        ret

这里 a 的值从内存加载到 eax 中，然后移动到 edi 传递给函数.但是我写了一个类似的汇编代码，直接将a的值加载到edi中.两者都工作正常.GCC 是否遵循此进行某种优化?先加载eax有什么好处?

Here value of a is loaded from memory into eax and then it is moved to edi to pass to function. But I wrote a similar assembly code and directly loaded value of a into edi. Both works fine. Does GCC follows this for some kind of optimization? What is the benefit of loading eax first?

推荐答案

GCC 是否遵循此进行某种优化?

Does GCC follows this for some kind of optimization?

恰恰相反.您正在编译没有优化，因此 gcc 不会进行优化，该优化会删除多余的 mov 和不必要的 a 作为堆栈内存中的变量.

Just the opposite. You're compiling without optimizations so gcc doesn't do the optimization that would have removed the redundant movs and the unnecessary allocation of a as a variable in stack memory.

先加载 eax 有什么好处?

What is the benefit of loading eax first?

它没有，只是多余的.未经优化的编译器代码包含这样看起来很愚蠢的东西是很常见的.

It has none and is just redundant. It is very common for unoptimized compiler code to contain stupid-looking stuff like this.

这篇关于为什么 GCC 先将整数加载到 eax，然后再加载到其他一些相关寄存器?的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！