Abstract Interpretation of Binary Code with Memory Accesses using Polyhedra

TL;DR

This paper introduces a new static analysis method for binary code using polyhedral abstract interpretation, effectively handling memory accesses to compute loop bounds relevant for real-time system WCET analysis.

Contribution

It presents a novel polyhedral-based abstract domain and mapping functions for binary code analysis, applicable to loop bounds and other static analysis tasks.

Findings

Successfully computes upper bounds for loop iterations

Applicable to Worst-Case Execution Time analysis

General approach for static analysis of binary code

Abstract

In this paper we propose a novel methodology for static analysis of binary code using abstract interpretation. We use an abstract domain based on polyhedra and two mapping functions that associate polyhedra variables with registers and memory. We demonstrate our methodology to the problem of computing upper bounds to loop iterations in the code. This problem is particularly important in the domain of Worst-Case Execution Time (WCET) analysis of safety-critical real-time code. However, our approach is general and it can applied to other static analysis problems.