A new approach to detect potential race conditions in component-based systems

Abstract

When programming software applications, developers have to deal with many functional and non-functional requirements. During the last decade, especially in the augmented reality field of research, many frameworks have been developed using a component-based approach in order to fulfil the non-functional requirements. In this paper, we focus on such a specific requirement: race conditions issues in component-based systems. We present a heuristic that analyses data flows and detects components that may be subject to race conditions. A toy example introducing the problem and the solution is developed and implemented under the ARCS (for Augmented Reality Component System) framework. We also show the results of our algorithm on real size applications using up to 70 components and compare those results with some obtained by developers who had to make exactly the same work by hand.

Invariant-Based Safety Assessment of FPGA Projects: Conception and Technique

“…The method of model verification and abstract interpretation is investigated in [35]. In [36] heuristic methods are used to identify potential places in a race, a review of which is given in [37]. For parallel constructions, it is necessary to perform appropriate verification for the permutability property [38].…”