Prioritizing pointer analysis algorithm based on points-to updating

Pointer analysis is a key technology in data flow analysis, and the result of pointer analysis is the basis of compiler optimization and program transformation. Based on the inclusion-based pointer analysis algorithm research, this paper analyzes the problems of redundant constraints evaluation and computational overhead of priority evaluation model in Narse priority constraints evaluation algorithm. Candidate set of constraint evaluation is determined by points-to set updating information of pointers, and the prioritizing pointer analysis algorithm based on points-to updating is presented. Constraint dependency graph is built by pointer dereference dependence and pointer scalar dependence in constraint statements, and priority of constraint evaluation is determined by the dependencies. Prioritizing algorithm based on constraint dependency graph is presented to simplify the complex priority evaluation model in Narse algorithm, and the overall framework of the optimized algorithm is provided. The experimental results on SPEC 2000/SPEC 2006 benchmark show that the algorithm has a significant performance boost on the time overhead and storage overhead compared with Narse priority algorithm. © 2014 ISCAS.