Adaptation of the ThreadSanitizer Algorithm for Data Race Detection in an RTOS Kernel

Design and correct implementation of multithreaded synchronization algorithms is a key element in the development of modern real-time operating systems (RTOS). An important step in this direction is to verify that an algorithm is correct within a programming language memory model. This paper integrates ThreadSanitizer—a widely used LLVM tool for data race detection—into an RTOS kernel environment and discusses its advantages and disadvantages over other tools for data race detection. Among other things, the semantics of context switches and interrupt management within the happens-before synchronization model is considered. In conclusion, the results of the ThreadSanitizer integration into an RTOS kernel are presented in comparison with existing approaches to bug detection.