Efficient Interprocedural Data-Flow Analysis Using Treedepth and Treewidth

We consider interprocedural data-flow analysis as formalized by the standard IFDS framework, which can express many widely-used static analyses such reaching definitions, live variables, and null-pointer. focus on well-studied on-demand setting in queries arrive one-by-one a stream each query should be answered fast possible. While classical algorithm provides polynomial-time solution for this problem, it is not scalable practice. More specifically, will either require quadratic-time preprocessing phase or takes linear time per query, both of are untenable modern huge codebases with hundreds thousands lines. Previous works have already shown that parameterizing problem treewidth program’s control-flow graph promising lead to significant gains efficiency. Unfortunately, these results were only applicable limited special case same-context queries. In work, we obtain speedups general necessarily same-context. This achieved exploiting new sparsity parameter, namely treedepth call graph. Our approach first exploit graphs at same parameterize treedepth. an answer constant wrt size input. Finally, our experimental demonstrate significantly outperforms its variant.