Analysis of error propagation between software processes

Abstract

To some extent all software systems contain faults that can potentially propagate from one process to another in different ways. Previous work on error propagation analysis is on probabilistic models. This paper presents a method of analysis to identify mechanisms for error propagation between processes in software that can be detected using static analysis of source code. The method examines the operating system to identify mechanisms for error propagation mechanisms between processes. The obtained results are then analyzed in detail to recognize characteristics that can be detected in source code. © 2007 Taylor & Francis Group.