Question

The question at strlen in assembly
is about strlen in assembly, but that seems to be written in AT&T syntax, which unfortunately isn’t supported by my assembler.

I’m just curious how would you translate the syntax [eax + ebx] to the intel’s syntax?

So

4(%eax)     # offset 4

is equivalent to

[eax + 4]

Suppose the offset value is stored in %ebx.
What’s the equivalence of

[eax + ebx]

?

%ebx(%eax)     # doesn't work!

Thanks,

Answer 1

[eax+ebx] should be translated to (%eax,%ebx).

The full syntax for memory addresses in AT&T syntax is:

offset(%reg1,%reg2,scale)

which is equivalent to:

[reg1 + reg2*scale + offset]

offset and scale are constant values, and scale must be 1, 2, 4, or 8. Defaults are 0 for offset and 1 for scale.