Hardware-accelerated regular expression matching with overlap handling on IBM PowerEN™ processor

Abstract

Programmable hardware accelerators for regular expression (regex) matching are evolving into increasingly complex stream processors, which involve multiple state machines that operate in parallel, and specialized post-processors that can process instructions dispatched by the state machines. To improve the speed and the storage-efficiency, complex regexs are decomposed into simpler sub expressions, where each sub expression can fire one or more instructions. Although the impact of regex decompositions on the storage efficiency is well-known, little has been done to address the correctness and completeness. We show that regex decompositions can resultin false positives if overlaps between sub expressions are not taken into account. We describe formal methods to recognize various types of sub expression overlaps that can arise in regex decompositions. We also describe efficient post-processing techniques to eliminate the associated false positives. To enable efficient mapping of the decomposed regexs to the post-processors, we propose integer programming based register allocation methods. Our methods pack narrow variables to reduce the register and instruction usage, and take advantage of multi-register reset instructions to reduce the number of instructions that must be executed in parallel. Experiments on regex sets obtained from open-source and proprietary network intrusion detection systems demonstrate orders of magnitude improvement in the storage efficiency over state-of-the-art. © 2013 IEEE.