Flow-sensitive points-to analysis for Java programs using BDDs

Abstract

Doing a flow-sensitive points-to analysis benefits many program analyses which need precise results, however, many prefer to do a flow-insensitive analysis to gain speed and overcome the memory limitations of a flow-sensitive points-to analysis. We are able to overcome these limitations by representing and manipulating points-to sets more efficiently. Binary Decision Diagrams (BDDs) have been shown to be a very efficient representation of points-to sets. A reasonable formulation of the solution is of great importance to achieve this efficiency. In this work, we formulate and employ BDDs to represent points-to sets in flow-sensitive points-to analysis for Java programs. Our method was compared with default points-to set in Soot program analysis framework (hybrid points-to set). The results are very promising and show the effectiveness of our method for sufficiently large programs.