【24h】

Stack Bound Inference for Abstract Java Bytecode

机译:堆栈绑定推理抽象java字节码

获取原文

摘要

Ubiquitous embedded systems are often resource-constrained. Developing software for these systems should take into account resources such as memory space. In this paper, we develop and implement an analysis framework to infer statically stack usage bounds for assembly-level programs in abstract Java Byte code. Our stack bound inference process, extended from a theoretical framework proposed earlier by some of the authors, is composed of deductive inference rules in multiple passes. Based on these rules, a usable tool has been developed for processing programs to capture the stack memory needs of each procedure in terms of the symbolic values of its parameters. The final result contains path-sensitive information to achieve better precision. The tool invokes a Presburger solver to perform fixed point analysis for loops and recursive procedures. Our initial experiments have confirmed the viability and power of the approach.
机译:普遍存在的嵌入式系统通常是资源受限的。开发这些系统的软件应考虑存储空间等资源。在本文中,我们开发并实施分析框架,以在抽象的Java字节代码中推断静态堆叠使用范围。我们的堆栈绑定推理过程从一些作者提出的一个理论框架扩展,由多次通过的Defuctive推理规则组成。基于这些规则,已经开发了一种可用的工具,用于处理程序,以捕获每个过程的堆栈内存需求,以符号的参数的符号值。最终结果包含路径敏感信息,以实现更好的精度。该工具调用Presburger求解器,以对循环和递归程序执行定点分析。我们的初步实验证实了这种方法的可行性和力量。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号