Hypertesting : The Case for Automated Testing of Hyperproperties
Hypertesting : The Case for Automated Testing of Hyperproperties
复制标题
超级测试：超级属性的自动化测试案例
DOI：
10.1109/ase51524.2021.9678758
复制
发表时间：
2015
期刊：
影响因子：
--
通讯作者：
Johannes Kinder
中科院分区：
文献类型：
--
作者：
Johannes Kinder
证明系统提供绝对的保证，但众所周知，对于非专家来说，很难使用bug找到的工具。错误查找是确定的。通常，发现虫子的违规行为不能用单一的测试案例来表达，因此现有的错误查找方法不适用。验证技术和概述了下一代的错误查找系统的概念和工具。要么对系统提供一定程度的信心，要么为高杂质提供了反陈述，我们会林于林木自动化的测试，以违反安全信息流以及程序的数字和加密性能。
Proof systems give absolute guarantees but are notoriously difficult to use for non-experts. Bug-finding tools make no completeness guarantees but offer a high degree of automation and are relatively easy to use for developers. For safety properties, the effectiveness of automatic test generation and bug finding is well established. For security properties like non-interference, which cannot be expressed as properties of a single program execution (i.e., hyperproperties), methods for testing and bug finding are in their infancy. In general, violations of hyperproperties cannot be expressed with a single test case like safety properties, so existing bug finding methods do not apply. This paper takes the position that we should fill this gap in the arsenal of verification technology and outlines concepts and tools for the next generation of bug finding systems. In particular, we aim to establish a generalized concept for the generation of “hypertests”, sets of tests that either provide some level of confidence in the system or give counterexamples to hyperproperties. As concrete instances of hypertesting, we foresee automated testing for violations of secure information flow and of numeric and cryptographic properties of programs.