Software and Computational Systems Lab

Algorithms for analysis of C programs are usually designed to verify a concrete class of properties of the C programs. Many of the properties of C programs that are checked by verifiers have a similar form. Memory-Safety and Reach-Safety property are both safety properties of the program with different atomic propositions. By instrumenting C programs, we can convert the Memory-Safety property to the Reach-Safety property, which opens new options for Memory-Safety analysis using algorithms that are used for Reach-safety tasks. After having solved the transformed task, we want to transform the witness i.e. the information the verifier provided for the verdict of the program to a witness for the original program. The goal of the thesis is to implement such a transformation of witnesses.

Algorithms for analysis of C programs are usually designed to verify a concrete class of properties of the C programs. Many of the properties of C programs that are checked by verifiers have a similar form. No-Datarace and Reach-Safety property are both safety properties of the program with different atomic propositions. By instrumenting C programs, we can convert the No-Datarace property to the Reach-Safety property, which opens new options for No-Datarace analysis using algorithms that are used for Reach-safety tasks. The goal of the thesis is to develop an idea of transformation C programs so that Reach-Safety analysis proves No-Datarace, implement the transformation in a tool and observe whether it is more efficient to solve No-Overflow using Reach-Safety analysis.

Algorithms for analysis of C programs are usually designed to verify a concrete class of properties of the C programs. Many of the properties of C programs that are checked by verifiers have a similar form. Memory-Safety and Reach-Safety property are both safety properties of the program with different atomic propositions. By instrumenting C programs, we can convert the Memory-Safety property to the Reach-Safety property, which opens new options for Memory-Safety analysis using algorithms that are used for Reach-safety tasks. The goal of the thesis is to develop an idea of transformation C programs so that Reach-Safety analysis proves Memory-Safety, implement the transformation in a tool and observe whether it is more efficient to solve Memory-Safety using Reach-Safety analysis.

Algorithms for analysis of C programs are usually designed to verify a concrete class of properties of the C programs. Many of the properties of C programs that are checked by verifiers have a similar form. Thread-Liveness property is a liveness property, whereas Reach-Safety is a safety property. It is well-known that liveness can be transformed to safety. By instrumenting C programs, we can convert the Thread-Liveness property to the Reach-Safety property, which opens new options for Thread-Liveness analysis using algorithms that are used for Reach-safety tasks. The goal of the thesis is to implement the transformation in a tool and observe whether it is more efficient to solve Thread-Liveness using Reach-Safety analysis.