AMD 250 manual Loop Instruction, 141

25112 Rev. 3.06 September 2005 Software Optimization Guide for AMD64 Processors

6.8The LOOP Instruction

Optimization

Avoid using the LOOP instruction.

Application

This optimization applies to:

•32-bit software

•64-bit software

Rationale

The LOOP instruction has a latency of at least 8 cycles.

Example

Avoid code like this, which uses the LOOP instruction:

label:

...

loop label

Instead, replace the loop instruction with a DEC and a JNZ:

label:

...

dec rcx jnz label

Chapter 6 Branch Optimizations 141