Accelerating C/C++ Pointer Analysis via Compiler-Based Offline Simplifications

2026-08-05

AI Summary

This work addresses the performance limitations of traditional pointer analysis by proposing a decoupled acceleration paradigm. Instead of tightly coupling simplification rules with the analysis itself—a common drawback of existing offline approaches—the method applies general-purpose, semantics-preserving compiler optimizations to the intermediate representation (IR) prior to analysis. This modular, analysis-agnostic strategy enhances efficiency without requiring modifications to the pointer analysis algorithm, thereby supporting seamless integration with diverse analyses. Empirical evaluation across multiple benchmark programs and three mainstream pointer analyses demonstrates speedups of up to 3.14× and memory reductions of up to 1.94×, all while largely preserving precision.

Abstract

Pointer analysis is a cornerstone of numerous static analysis applications, including compiler optimizations, slicing, bug detection, and verification. While offline simplification is a common approach to boosting performance, existing methods are often tightly coupled to specific analysis algorithms and limited to a set of simplification rules. This paper explores a new perspective: applying semantic-preserving compiler optimizations directly to intermediate representation (IR) before pointer analysis. This strategy is modular, analysis-agnostic, and easily integrates with existing tools. We conduct an empirical study using diverse programs and three pointer analyses. The results show substantial performance gains---up to 3.14x speedup and 1.94x memory reduction---while precision remains largely unchanged. We also analyze the trade-offs between optimization overhead and analysis speedup, quantify changes in IR structure, assess the characteristics of optimization configurations, and identify promising directions for future research.

Problem

• pointer analysis
• offline simplification
• compiler optimization
• intermediate representation
• static analysis

Innovation

• pointer analysis
• compiler optimization
• intermediate representation
• offline simplification
• analysis-agnostic