An Algorithmic Methodology for Validation of Mixed Hardware/Software Systems

现代集成系统通常是硬件和软件组件的混合体。调试和验证这些嵌入式计算机系统需要集成硬件和软件验证机制。本研究通过验证嵌入式系统(硬件和软件)设计的描述是否具有一系列输入激励所引发的适当行为，来研究检查嵌入式系统(硬件和软件)设计正确性的方法。这种方法被称为覆盖导向验证，并且正在设计覆盖度量。这些测量由一组功能仿真向量提供的设计验证的程度，这些功能仿真向量计算语句执行计数(可控性信息)，并检查是否可以在电路输出端观察到由程序刺激激活的可能错误的影响(可观性信息)。特别是，该项目正在探索有效的方法来计算基于可观测性的代码覆盖指标，这些指标既可以用于嵌入式软件，也可以用于硬件描述语言的覆盖。

Modern integrated systems are often a mixture of hardware and softwarecomponents. Debugging and verifying these embedded computer systemsrequires the integration of hardware and software verification mechanisms.This research is investigating methods for checking correctness of anembedded system (hardware and software) design by verifying that adescription of the design has the proper behavior as elicited by a seriesof input stimuli. The approach is called coverage directed validation, andcoverage metrics are being devised. These measure the extent of designverification provided by a set of functional simulation vectors, whichcompute statement execution counts (controllability information), and checkto see whether effects of possible errors activated by program stimuli canbe observed at the circuit outputs (observability information). In particular, the project is exploring efficient methods to compute observability-basedcode coverage metrics that can be used for both embedded software andHDL coverage.