An Incremental Points-to Analysis with CFL-Reachability

Abstract

Abstract.Developing scalable and precise points-to analyses is increasingly important for analysing and optimising object-oriented programs where pointers are used pervasively. An incremental analysis for a program updates the existing analysis information after program changes to avoid reanalysing it from scratch. This can be efficiently deployed in software development environments where code changes are often small and frequent. This paper presents an incremental approach for demand-driven context-sensitive points-to analyses based on ContextFree Language (CFL) reachability. By tracing the CFL-reachable paths traversed in computing points-to sets, we can precisely identify and recompute on demand only the points-to sets affected by the program changes made. Combined with a flexible policy for controlling the granularity of traces, our analysis achieves significant speedups with little space overhead over reanalysis from scratch when evaluated with a null dereferencing client using 14 Java benchmarks.

Parallel Pointer Analysis with CFL-Reachability

“…Must-not-alias information obtained during a pre-analysis can be exploited to yield an average speedup of 3X through reducing unnecessary alias-related computations [25]. Incremental techniques [6], [16], which are tailored for scenarios where code changes are small, take advantage of previously computed CFL-reachable paths to avoid unnecessary reanalysis.…”

“…As a result, there has been a recent focus on demanddriven pointer analysis, founded on Context-Free Language (CFL) reachability [15], which only performs the necessary work to compute the points-to information requested in a graph representation of the program [6], [16], [17], [18], [19], [26], [27]. By computing the points-to information of some (instead of all variables as in whole-program Andersen's analysis), demand-driven analysis can be performed both context-and field-sensitively to achieve better precision more scalably than Andersen's analysis, especially for Java programs.…”

“…In the sequential setting, there are efforts for improving its efficiency for Java programs by resorting to refinement [18], [19], summarisation [17], [26], incrementalisation [6], [16], pre-analysis [25] and ad-hoc caching [18], [25]. However, multicore platforms, which are ubiquitous nowadays, have never been exploited for accelerating the CFL-reachability-based analysis.…”

Query-directed adaptive heap cloning for optimizing compilers

“…In the demand-driven alias/pointer analysis based on context-free language reachability [21,[27][28][29], full heap cloning is naturally supported. in future research, whether it scales when performed as a whole-program analysis for large programs remains to be further investigated [36].…”

“…Full heap cloning for C is unscalable for large, heap-intensive programs [22] as the number of abstract heap objects cloned can be prohibitive in terms of both time and space. The demand-driven analysis formulated via context-free-language reachability [21,[27][28][29]38] embraces naturally full heap cloning, but its scalability when deployed as a whole-program pointer analysis remains to be further investigated [36]. BDD-based techniques, while effective for solving context-sensitive pointer analyses [33,39], may not scale for full heap cloning [18,35,36].…”