Collaborative Research: SHF: Small: A General Framework for Responsive Static Analysis

合作研究：SHF：小型：响应式静态分析的通用框架

• 批准号: 2223825
• 负责人: Bor-Yuh Evan Chang
• 金额: $ 30万
• 依托单位: University of Colorado at Boulder
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2022
• 资助国家: 美国
• 起止时间: 2022-10-01 至 2025-09-30
• 项目状态: 未结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=2223825&HistoricalAwards=false
• 关键词: Collaborative; Research; SHF; Small; General

Society increasingly relies on the reliability and security of software. Abstract interpretation is a well-established methodology for proving that software is free of certain classes of bugs. However, for industrial-scale software, standard abstract interpretation techniques may take hours to complete, making them difficult to integrate into modern software development practices. This project develops a framework for responsive static analysis, which retains the power of abstract interpretation while running much more quickly for common use cases. The project's novelties are new algorithms for running abstract interpretation responsively, corresponding mathematical proofs that these algorithms produce the desired, correct results, and working implementations of the algorithms. The project's impacts are greater performance and applicability of powerful abstract interpretation techniques for verifying software correctness, which in turn will yield more reliable and secure software.The project builds on a recently-developed framework for demanded abstract interpretation, a demand-driven and incremental analysis approach based on reifying analysis computations and dependencies in a graph structure. Via generalizations of this approach, this project will extend the framework to handle compositional analysis, essential for efficient analysis of procedure calls, and refinement-based analysis, to enable combining analyses with varying levels of precision and scalability. This generalized framework will facilitate provable guarantees of from-scratch consistency, a crucial property for responsive analysis. The project will also implement the generalized framework and instantiate it with challenging analysis problems, addressing research challenges in making the framework practical.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

社会越来越依赖软件的可靠性和安全性。 抽象解释是一种行之有效的方法，用于证明软件不存在某些类别的错误。 然而，对于工业规模的软件，标准抽象解释技术可能需要几个小时才能完成，这使得它们难以集成到现代软件开发实践中。 该项目开发了一个用于响应式静态分析的框架，该框架保留了抽象解释的能力，同时针对常见用例运行得更快。 该项目的新颖之处在于用于响应地运行抽象解释的新算法、这些算法产生所需的正确结果的相应数学证明以及算法的有效实现。 该项目的影响是验证软件正确性的强大抽象解释技术的更高性能和适用性，这反过来将产生更可靠和安全的软件。该项目建立在最近开发的用于需求抽象解释的框架之上，这是一种需求驱动的增量分析基于图结构中具体化分析计算和依赖关系的方法。通过这种方法的推广，该项目将扩展框架来处理组合分析，这对于过程调用的有效分析和基于细化的分析至关重要，从而能够将不同级别的精度和可扩展性的分析结合起来。这个通用框架将促进从头开始的一致性的可证明保证，这是响应式分析的关键属性。该项目还将实施通用框架，并通过具有挑战性的分析问题对其进行实例化，解决使该框架实用化的研究挑战。该奖项反映了 NSF 的法定使命，并通过使用基金会的智力价值和更广泛的影响审查标准进行评估，被认为值得支持。