FairFuzz: Targeting Rare Branches to Rapidly Increase Greybox Fuzz Testing Coverage.

摘要: In recent years, fuzz testing has proven itself to be one of the most effective techniques for finding correctness bugs and security vulnerabilities in practice. One particular tool, American Fuzzy Lop or AFL, become popular thanks its ease-of-use bug-finding power. However, AFL remains limited depth program coverage it achieves, because does not consider which parts inputs should mutated order maintain deep coverage. We propose an approach, FairFuzz, that helps alleviate this limitation two key steps. First, FairFuzz automatically prioritizes exercising rare under test. Second, adjusts mutation so are more likely exercise these same program. conduct evaluation on real-world programs against state-of-the-art versions thoroughly repeating experiments get good measures variability. find certain benchmarks shows significant increases after 24 hours compared while others achieves high at a significantly faster rate.