X86 Addressing Under the Hood

x86 instruction encoding is best viewed in octal --- not just the ModRM, but the primary opcode too:<p><a href="http:&#x2F;&#x2F;www.dabo.de&#x2F;ccc99&#x2F;www.camp.ccc.de&#x2F;radio&#x2F;help.txt" rel="nofollow">http:&#x2F;&#x2F;www.dabo.de&#x2F;ccc99&#x2F;www.camp.ccc.de&#x2F;radio&#x2F;help.txt</a><p>You can mentally assemble&#x2F;disassemble the bulk of the commonly encountered instructions by memorising a few tables (in octal), the addressing modes being one of them. In 16-bit the memory addressing modes can be described as &quot;one or more of {displacement}{BX,BP}{SI,DI}&quot; and 32-bit &quot;one or more of {displacement}{register}{scaled register}&quot; with (e)BP as a special case.

Good homework is to implement an x86-64 mod&#x2F;rm encoder.<p>Quite a few corner cases. Sometimes you&#x27;ll need to up-convert input parameters, because some combinations can&#x27;t use 8-bit offset, for example, but require a 32-bit value. RSP can&#x27;t use scaling, etc.<p>There&#x27;s also plenty of redundancy: <a href="https:&#x2F;&#x2F;www.strchr.com&#x2F;machine_code_redundancy" rel="nofollow">https:&#x2F;&#x2F;www.strchr.com&#x2F;machine_code_redundancy</a>

I once had to multiply the `rsi` register by 3 so you bet I used `imul rsi, 3`, right? Nope. This does the exact same thing with an unknown or small performance advantage:<p>lea rsi, [rsi + rsi*2]