Efficient detection and exploitation of infeasible paths for software timing analysis

Abstract

Accurate estimation of the worst-case execution time (WCET) of a program is important for real-time embedded software. Static WCET estimation involves program path analysis and architectural modeling. Path analysis is complex due to the inherent difficulty in detecting and exploiting infeasible paths in a program's control flow graph. In this paper, we propose an efficient method to exploit infeasible path information for WCET estimation without resorting to exhaustive path enumeration. We demonstrate the efficiency of our approach for some real-life control-intensive applications. Copyright 2006 ACM.