Assembly Language Absolute addresses and segment registers

问题

a review problem lists these registers in hex:

cs = ????  sp = 0300  ax = a66a  ip = 01cf
ds = 4100  bp = 0003  bx = 1234  
ss = 48ee  si = 0100  cx = 00ff
es = 4cee  di = 1000  dx = 0000

The absolute address of the next instruction to be executed is 40f0f.
```
40f0f  
-01cf
_____
40d40 / 10 = 40d4 = cs
```
Is the size of the data segment in bytes always equal to the stack segment minus the data segment * 10? 48ee - 4100 = 7ee0. Likewise, is the code segment in bytes always equal to the data segment minus the code segment * 10? 48ee - 40d4 = 81a0.

For mov cx,[bx + si], the absolute address of the source operand is 42334.

bx = 1234
si = 0100
_________
     1334

ds = 4100 * 10 = 41000 + 1334 = 42334

For mov cx,[di - 4], the absolute address of the source operand is 41ffc.

di = 1000
     -  4
_________
     0FFC

ds = 4100 * 10 = 41000 + 0ffc = 41ffc

For mov cx,[bp + si - 3], the absolute address of the source operand is 48fe0.

bp = 0003
si = 0100
      - 3
_________
     0100

ss = 48ee * 10 = 48ee0 + 0100 = 48fe0

Am I going about solving these the right way? How do I know when to use the stack segment for these calculations and when to use the data segment?

回答1:

For address calculations involving bp or sp or stack operations like push or pop the segment register is implicitly ss, for other addresses ds. Exception: If you use a string instruction, the destination segment register is implicitly es.

来源：https://stackoverflow.com/questions/8486780/assembly-language-absolute-addresses-and-segment-registers

标签

assembly

stack

segment