Efficient analysis of pipeline models for WCET computation

Worst-case execution time (WCET) prediction for mod-ern CPU’s cannot make local assumptions about the im-pact of input information on the global worst-case because of the existence of timing anomalies. Therefore, static anal-yses on the hardware level must consider a large subset of the reachable states of the underlying hardware model. As the number of states grows, WCET prediction can become infeasible because of the increase in computation time and memory consumption. This paper presents a solution for this problem by defining the static analysis of processor pipelines for WCET computation in terms of operations on binary decision diagrams (BDD’s)