Preserving Timing Anomalies in Pipelines of High-End Processors

Many embedded systems are subject to temporal constraints that require advance guarantees on meeting deadlines. Such systems rely on static analysis to safely bound worst-case execution (WCET) bounds of tasks. Designers of these systems are forced to avoid state-of-the-art processors due to their inherent architectural complexity that results in non-determinism. Such micro-processors are typically tuned to reduce average-case execution times — at the expense of predictability. Dynamic instruction scheduling techniques, such as out-of-order (OOO) execution, are examples of features that reduce average time but are statically unpredictable at large. This work addresses this problem by providing analysis techniques for characterizing the worst-case behavior of real-time systems on modern processor architectures. We propose minor enhancements to processor architectures that, coupled with static analysis techniques, support the derivation of safe WCET bounds. We also introduce novel pipeline analysis techniques for accurately capturing the worst-case behavior of real-time tasks, i.e., methods to capture (“snapshot”) pipeline state and to subsequently perform a “merge” of previously captured snapshots. We prove that our pipeline analysis correctly preserves worst-case timing behavior on OOO processor pipelines. We further specifically show that anomalous pipeline effects, effectively dilating timing, are preserved by our method. To the best of our knowledge, this method of pipeline analysis and interactions between hardware/software for obtaining WCET bounds on OOO processors is the first of its kind.