Goblint: A Portfolio for Mixed Flow-Sensitive Abstract Interpretation

View at publisher
Authors: Simmo Saan, Ali Rasim Kocal, Michael Petter, Karoliine Holter, Julian Erhard, Michael Schwarz, Vesal Vojdani, Helmut Seidl
Cited by: 1
References: 27

Abstract

Abstract Goblint is a static analyzer for C programs based on abstract interpretation. It provides a flexible framework for combining analyses and specifying what information is tracked flow- and context-sensitively. For SV-COMP 2026, we adopt a sequential portfolio of configurations of increasing precision that exploits Goblint ’s rich family of abstract domains and sensitivity options. We have improved the analyses by adding bit-level and sparse affine equalities domains, new forms of path-sensitivity, and a context gas mechanism that controls context-sensitivity for recursive programs. Goblint continues its unblemished track record of sound verification across all SV-COMP benchmarks.

Fields of Study & Subjects

Logic programming and type systems Formal Methods in Verification Security and Verification in Computing

Publication details

Journal / source
Lecture notes in computer science
Publisher
Springer Nature Switzerland
Pages
516–522
Publication date
1 January 2026
Language
EN