Справочник Пользователя для AMD 250
Chapter 4
Instruction-Decoding Optimizations
89
Software Optimization Guide for AMD64 Processors
25112
Rev. 3.06
September 2005
4.12
Code Padding with Operand-Size Override and
NOP
NOP
Optimization
Use one or more operand-size overrides (66h) and the NOP instruction (90h) to align code and space
out branches.
out branches.
Application
This optimization applies to:
•
32-bit software
•
64-bit software
Rationale
Occasionally it is necessary to insert neutral code fillers into the code stream (for example, for code-
alignment purposes or to space out branches). Because this filler code is executable, it should take up
as few execution resources as possible, not diminish decode density, and not modify any processor
state other than advancing the instruction pointer (rIP). Although there are several possible multibyte
NOP-equivalent instructions that do not change the processor state (other than rIP), combinations of
the operand-size override and the NOP instruction work best.
alignment purposes or to space out branches). Because this filler code is executable, it should take up
as few execution resources as possible, not diminish decode density, and not modify any processor
state other than advancing the instruction pointer (rIP). Although there are several possible multibyte
NOP-equivalent instructions that do not change the processor state (other than rIP), combinations of
the operand-size override and the NOP instruction work best.
Example
Assign code-padding sequences like these and use them to align code and space out branches. These
sequences are suitable for both 32-bit and 64-bit code, and you can use them on the AMD Athlon 64
and AMD Opteron processors, as well as seventh-generation AMD Athlon processors:
sequences are suitable for both 32-bit and 64-bit code, and you can use them on the AMD Athlon 64
and AMD Opteron processors, as well as seventh-generation AMD Athlon processors:
NOP1_OVERRIDE_NOP TEXTEQU <DB 090h>
NOP2_OVERRIDE_NOP TEXTEQU <DB 066h,090h>
NOP3_OVERRIDE_NOP TEXTEQU <DB 066h,066h,090h>
NOP4_OVERRIDE_NOP TEXTEQU <DB 066h,066h,066h,090h>
NOP5_OVERRIDE_NOP TEXTEQU <DB 066h,066h,090h,066h,090h>
NOP6_OVERRIDE_NOP TEXTEQU <DB 066h,066h,090h,066h,066h,090h>
NOP7_OVERRIDE_NOP TEXTEQU <DB 066h,066h,066h,090h,066h,066h,090h>
NOP8_OVERRIDE_NOP TEXTEQU <DB 066h,066h,066h,090h,066h,066h,066h,090h>
NOP9_OVERRIDE_NOP TEXTEQU <DB 066h,066h,090h,066h,066h,090h,066h,066h,090h>
For x87 floating-point instructions, a better single-byte padding exists. See “Align and Pack
DirectPath x87 Instructions” on page 242.
DirectPath x87 Instructions” on page 242.