This master's thesis deals with the analysis of dynamic linked data structures using shape analysis used in the Predator tool. It describes the chosen abstract domain for heap representation - symbolic memory graphs. It deals with the design of framework for the development of static analyzers based on Clang/LLVM. The main contribution is implementing and testing LLVM's transformation passes that simplify the LLVM IR. Second contribution is the optimization of parameters for parallel run of several variants of the Predator tool. Parameters are tuned for benchmark from SV-COMP'16, where our tool won gold medal in Heap Data Structures category. Last contribution is the design of verification core with the focus on the SMG domain.
| Identifer | oai:union.ndltd.org:nusl.cz/oai:invenio.nusl.cz:255420 |
| Date | January 2016 |
| Creators | Šoková, Veronika |
| Contributors | Rogalewicz, Adam, Vojnar, Tomáš |
| Publisher | Vysoké učení technické v Brně. Fakulta informačních technologií |
| Source Sets | Czech ETDs |
| Language | Czech |
| Detected Language | English |
| Type | info:eu-repo/semantics/masterThesis |
| Rights | info:eu-repo/semantics/restrictedAccess |
Page generated in 0.0023 seconds