Chapter 1. Introduction

Table of Contents

1. Comparison between using SLO and more traditional cache profilers

SLO helps you to identify the source code constructs that generate poor temporal data locality. Poor temporal data locality leads to many data cache misses and slow program execution. SLO aims to be a better cache analysis tool. Most other cache profilers indicate the source code lines that generate most cache misses. However, eliminating the cache misses often requires code changes (refactorings) in completely different statements, functions or even source files. In contrast, SLO highlights the loops and other code constructs that must be refactored to improve the data locality so that the cache misses are turned into cache hits.

Basically, SLO analyzes for each data reuse, which code is executed between use and reuse. Based on this analysis, it suggests an appropriate refactoring to improve the locality.

Please report any problems or feature requests to , or use the bug report and feature request features at the web-site