Integrating Abstract Caches with Symbolic Pipeline Analysis

Abstract

Static worst-case execution time analysis of real-time tasks is based on abstract models that capture the timing behavior of the processor on which the tasks run. For complex processors, task-level execution time bounds are obtained by a state space exploration which involves the abstract model and the program. Partial state space exploration is not sound. Symbolic methods using binary decision diagrams (BDDs) allow for a full state space exploration of the pipeline, thereby maintaining soundness. Caches are too large to admit an efficient BDD representation. On the other hand, invariants of the cache state can be computed efficiently using abstract interpretation. How to integrate abstract caches with symbolic-state pipeline analysis is an open question. We propose a semi-symbolic domain to solve this problem. Statistical data from industrial-level software and WCET tools indicate that this new domain will enable an efficient analysis.