循环展开(loop unrolling / loop unwinding)

最新推荐文章于 2024-09-01 22:36:27 发布

zhuliting

最新推荐文章于 2024-09-01 22:36:27 发布

阅读量5.3k

点赞数 1

分类专栏：并行/高性能计算文章标签： transformation delete loops branch delay each

本文链接：https://blog.csdn.net/zhuliting/article/details/6073037

版权

并行/高性能计算专栏收录该内容

22 篇文章 3 订阅

订阅专栏

Loop unwinding, also known as loop unrolling, is a loop transformation technique that attempts to optimize a program's execution speed at the expense of its binary size (space-time tradeoff). The transformation can be undertaken manually by the programmer or by an optimizing compiler.

The goal of loop unwinding is to increase a program's speed by reducing (or eliminating) instructions that control the loop, such as pointer arithmetic and "end of loop" tests on each iteration;reducing branch penalties; as well as "hiding latencies, in particular, the delay in reading data from memory".

Loops can be re-written instead as a repeated sequence of similar independent statements eliminating this overhead.

gcc 优化指令：-funroll-loops

A simple manual example in C Language

Normal loop	After loop unrolling
int x; for (x = 0; x < 100; x++) { delete(x); } //. //. //. //.	int x; for (x = 0; x < 100; x += 5) { delete(x); delete(x+1); delete(x+2); delete(x+3); delete(x+4); }