Binary compatibility for embedded systems using greedy subgraph mapping

Abstract

We propose a novel lightweight code generation algorithm GSM (Greedy Subgraph Mapping), which can generate compact code with low overhead using many-to-one mapping. GSM is implemented and evaluated in a dynamic binary translation prototype system called TransARM. Experimental results demonstrate that GSM generates higher quality target code than a conventional implementation, which brings an average code expansion rate close to 1.3 for the selected 11 benchmarks. Moreover, GSM causes slightly extra overhead and negligible slowdown of translation and enables 10% performance improvement for target code execution.