Thread-Modular Correctness Witnesses and their Validation: An Abstract Interpretation Perspective

Anonymous artifact corresponding to an earlier revision of

Saan, S., Schwarz, M., Erhard, J., Seidl, H., Tilscher, S., & Vojdani, V. (2023). Correctness Witness Validation by Abstract Interpretation. Zenodo. https://doi.org/10.5281/zenodo.8253000

Layout

• autoactive/plain — programs and GraphML witnesses from AutoActive artifact.
• autoactive/generated_yaml — converted YAML witnesses.
• bench — this repository.
• benchexec — BenchExec version.
• CPAchecker-2.2-unix — CPAchecker from SV-COMP 2023 archives.
• ourtool — our tool, including source code. NB! Although git authorship has been stripped, the contents of this directory do reveal the unanonymized tool name.
• results — results directory for scripts.
• sv-benchmarks — SV-COMP 2022 benchmarks, at least ReachSafety-Loops.
• UAutomizer-linux — Ultimate Automizer from SV-COMP 2023 archives.

Scripts

• ./st-same/run.sh — single-threaded programs, same-framework.
• ./st-cross/witnessToYaml.sh — convert single-threaded cross-framework witnesses from GraphML to YAML (very slow!).
• ./st-cross/run.sh — single-threaded programs, cross-framework.
• ./st-lit/run.sh — single-threaded programs from literature.
• ./mt/run.sh — multi-threaded programs, same- and cross-framework.