Bootstrapping: a technique for scalable flow and context-sensitive pointer alias analysis

• Authors: Vineet Kahlon
• Venue: Proceedings of the 2008 ACM SIGPLAN conference on Programming language design and implementation
• Year: 2008

We propose a framework for improving both the scalability as well as the accuracy of pointer alias analysis, irrespective of its flow or context-sensitivities, by leveraging a three-pronged strategy that effectively combines divide and conquer, parallelization and function summarization. A key step in our approach is to first identify small subsets of pointers such that the problem of computing aliases of any pointer can be reduced to computing them in these small subsets instead of the entire program. In order to identify these subsets, we first apply a series of increasingly accurate but highly scalable (context and flow-insensitive) alias analyses in a cascaded fashion such that each analysis Ai works on the subsets generated by the previous one Ai-1. Restricting the application of Ai to subsets generated by Ai-1, instead of the entire program, improves it scalability, i.e., Ai is bootstrapped by Ai-1. Once these small subsets have been computed, in order to make our overall analysis accurate, we employ our new summarization-based flow and context-sensitive alias analysis. The small size of each subset offsets the higher computational complexity of the context-sensitive analysis. An important feature of our framework is that the analysis for each of the subsets can be carried out independently of others thereby allowing us to leverage parallelization further improving scalability.