摘要
代码优化是提高Java虚拟机执行速度的重要手段之一。本文提出了一种基于预处理的代码优化方法。预处理模块与嵌入式Java虚拟机并行工作 ,为其标注基本块、循环等结构 ,并通过数据流分析收集各方法、各变量的引用次数。Java虚拟机利用预处理模块所收集的信息 ,判定各块程序的执行频率 ,对高频部分进行内联、寄存器分配等优化处理。
The optimization of code is an important approach to improve the performance of embedded Java virtual machine, i.e. Kilos Virtual Machine.In the paper, a kind of the pretreatment-based optimization technology is presented (KVM). The pretreatment module concurrently works with Kilos Virtual Machine(KVM).The optimization can search and label some program structures such as block and circle.It also counts the number of times for methods and variants to be referenced.Making using of the result of pretreatment, KVM can find the most frequently executed blocks,and optimize them.Optimization includes inline and register assigning and the like.
出处
《河南科技大学学报（自然科学版）》
CAS
2003年第1期55-58,共4页
Journal of Henan University of Science And Technology:Natural Science
基金
洛阳工学院科研基金资助项目 (2 0 0 10 7)
