Biblio

Lock design is an important mechanism for scheduling management and security protection in operating systems. However, there is no effective way to identify the differences and connections among lock models, and users need to spend considerable time to understand different lock architectures. In this paper, we propose a classification scheme that abstracts lock design into three types of models: basic spinlock, semaphore amount extension, lock chain structure, and verify the effectiveness of these three types of lock models in the context of current mainstream applications. We also investigate the specific details of applying this classification method, which can be used as a reference for developers to design lock models, thus shorten the software development cycle.

The semaphore mechanism is an important part of the embedded operating system. Therefore, it is very necessary to ensure its safety. Traditional software testing methods are difficult to ensure 100% coverage of the program. Therefore, it is necessary to adopt a formal verfication method which proves the correctness of the program theoretically. This paper proposes a proof framework based on the theorem proof tool Coq: modeling the semaphore mechanism, extracting important properties from the requirement documents, and finally verifying that the semaphore mechanism can meet these properties, which means the correctness of the semaphore mechanism is proved and also illustrates the feasibility of the verification framework proposed in this paper, which lays a foundation for the verification of other modules of operating systems.