LART: Compiled Abstract Execution (Competition Contribution)

Varování

Publikace nespadá pod Ústav výpočetní techniky, ale pod Fakultu informatiky. Oficiální stránka publikace je na webu muni.cz.
Autoři

LAUKO Henrich ROČKAI Petr

Rok publikování 2022
Druh Článek ve sborníku
Konference TACAS 2022: Tools and Algorithms for the Construction and Analysis of Systems
Fakulta / Pracoviště MU

Fakulta informatiky

Citace
Doi http://dx.doi.org/10.1007/978-3-030-99527-0_31
Klíčová slova abstraction;abstract interpretation;abstract execution;compilation;compilation-based abstraction;LLVM;LART;formal verification;symbolic execution
Popis LART – LLVM Abstraction & Refinement Tool – originates from the divine model-checker, in which it was employed as an abstraction toolchain for the LLVM interpreter. In this contribution, we present a stand-alone tool that does not need a verification backend but performs the verification natively. The core idea is to instrument abstract semantics directly into the program and compile it into a native binary that performs program analysis. This approach provides a performance gain of native execution over the interpreted analysis and allows compiler optimizations to be employed on abstracted code, further extending the analysis efficiency. Compilation-based abstraction introduces new challenges solved by LART, like domain interaction of concrete and abstract values simulation of nondeterministic runtime or constraint propagation.
Související projekty:

Používáte starou verzi internetového prohlížeče. Doporučujeme aktualizovat Váš prohlížeč na nejnovější verzi.

Další info