基于混合模型的持续二进制优化框架

Continuous binary optimization framework based on hybrid model

提出了一项基于混合模型的二进制优化框架,能在运行时和运行后进行持续的程序优化。该框架集成了控制流分析等常用的二进制优化功能,并且提供了安全的编程接口以供其他研究者实现自定义的优化模块。该框架被设计为集成到操作系统内核中,并提供透明和自适应的优化服务,从而在没有用户交互的前提下,利用富余的计算资源帮助二进制程序自适应计算环境,达到加速的目的。描述了该框架的设计与实现以及关键问题的解决方法。

A binary optimization framework based on hybrid model is proposed to enable continuous runtime and post-runtime optimization.Common modules like control flow analysis are embedded in the framework,and safe programming interfaces are provided to facilitate other researchers implementing their own optimization modules.The framework is designed to be integrated into operating systems and transparent and adaptive services are provided,so that the binary programs can adapt to the environment and speedup without any end user assistance.The design and implementation of the framework are presented.