如何用Intel X86汇编语言计算员工结构体的薪资总和？

Got it, let's walk through exactly how to calculate the total salary of all employees using Intel x86 assembly. First, let's recap the data we're working with—you've got a Employee struct with a 20-byte name and 16-bit salary, an array of 4 employees, and a double-word variable sum to store the final total.

核心思路 #

The key steps are straightforward:

• Initialize a loop counter to iterate over every employee in the array
• Calculate the memory offset of each employee's salary field
• Accumulate each salary into the sum variable (handling 32-bit storage since sum is a double-word)
• Exit the loop once all employees are processed

16-bit MASM 实现（DOS环境） #

This version works for classic 16-bit DOS assembly, using segment registers and 16-bit operations:

.MODEL SMALL
.STACK 100h

.DATA
    ; 定义题目给出的结构体和数据
    Employee struct
        name DB 20 dup(0)
        salary DW 0
    Employee ends
    nrEmployees DW 4
    Employees Employee {"Alex", 100}, {"Mike", 100}, {"Paul", 300}, {"Ana", 450}
    sum DD 0

.CODE
MAIN PROC
    ; 初始化数据段寄存器（访问数据段的必要步骤）
    MOV AX, @DATA
    MOV DS, AX

    ; 初始化循环计数器和累加相关寄存器
    XOR CX, CX          ; CX = 0，作为循环计数器
    XOR AX, AX          ; AX临时存储单个员工的薪资

sum_loop:
    ; 检查是否遍历完所有员工：CX == nrEmployees？
    CMP CX, nrEmployees
    JE loop_end         ; 相等则跳转到循环结束

    ; 计算当前员工在数组中的偏移地址：CX * 结构体大小
    MOV AX, CX
    MOV BX, SIZEOF Employee  ; 自动获取结构体大小（22字节），避免硬编码
    MUL BX
    MOV BX, AX          ; BX现在是当前员工的起始偏移量

    ; 取出当前员工的salary字段：起始偏移 + salary字段的偏移（20字节）
    MOV AX, [Employees + BX + OFFSET Employee.salary]

    ; 累加薪资到sum（双字变量，分高低16位处理进位）
    ADD WORD PTR sum, AX       ; 先加低16位
    ADC WORD PTR sum + 2, 0    ; 加上进位到高16位，防止溢出

    ; 计数器加1，继续循环
    INC CX
    JMP sum_loop

loop_end:
    ; DOS环境下退出程序
    MOV AH, 4Ch
    INT 21h

MAIN ENDP
END MAIN

32-bit MASM32 实现（平展内存模型） #

If you're working in a 32-bit environment, the code becomes simpler thanks to 32-bit registers and flat memory addressing:

.386
.MODEL FLAT, C

.DATA
    Employee struct
        name DB 20 dup(0)
        salary DW 0
    Employee ends
    nrEmployees DWORD 4
    Employees Employee {"Alex", 100}, {"Mike", 100}, {"Paul", 300}, {"Ana", 450}
    sum DWORD 0

.CODE
_main PROC
    ; 初始化累加器和计数器
    XOR EAX, EAX          ; EAX存储总和，初始为0
    XOR ECX, ECX          ; ECX作为循环计数器

sum_loop:
    CMP ECX, nrEmployees
    JE loop_end

    ; 计算当前员工的内存地址：数组基址 + 计数器*结构体大小
    MOV EDX, ECX
    IMUL EDX, SIZEOF Employee
    MOV BX, [Employees + EDX + OFFSET Employee.salary]

    ; 直接累加薪资到EAX（32位足够存储总和，无需拆分处理）
    ADD EAX, EBX
    INC ECX
    JMP sum_loop

loop_end:
    MOV sum, EAX          ; 把最终总和存入sum变量
    RET
_main ENDP
END

关键注意事项 #

• 结构体大小计算: Always use SIZEOF Employee instead of hardcoding 22 bytes—this makes your code maintainable if the struct changes later.
• 32-bit存储处理: Since sum is a double-word (32 bits), in 16-bit mode you need to use ADC (add with carry) to handle overflow from the low 16 bits to the high 16 bits.
• 循环边界: Make sure the loop runs exactly nrEmployees times—using CMP CX, nrEmployees and JE loop_end ensures you don't under/over iterate.

内容的提问来源于stack exchange，提问作者Alexei Postolachi