Work-In-Progress: Models and tools to detect Real-Time Scheduling Anomalies
Abstract
This paper deals with scheduling anomalies in real-time systems. Scheduling anomalies jeopardize schedulability analysis made prior to execution. In this paper, we propose a model to specify conditions leading to scheduling anomalies. A scheduling anomaly is modeled as a set of constraints on the architecture. We use this model to detect scheduling anomalies by offline and online analysis. To validate our approach, we implemented the approach as an extension to Cheddar, a schedulability tool. We apply our approach to seven scheduling anomalies and we show that most of these anomalies can be successfully detected.