Fuzzing is known to be one of the most effective techniques to uncover security vulnerabilities of large-scale software systems. During fuzzing, it is crucial to distribute the fuzzing resource appropriately so as to achieve the best fuzzing performance under a limited budget. Existing distribution strategies of American Fuzzy Lop (AFL) based greybox fuzzing focus on increasing coverage blindly without considering the metrics of code regions, thus lacking the insight regarding which region is more likely to be vulnerable and deserves more fuzzing resources. We tackle the above drawback by proposing a vulnerable region-aware greybox fuzzing approach. Specifically, we distribute more fuzzing resources towards regions that are more likely to be vulnerable based on four kinds of code metrics. We implemented the approach as an extension to AFL named RegionFuzz. Large-scale experimental evaluations validate the effectiveness and efficiency of RegionFuzz-11 new bugs including three new CVEs are successfully uncovered by RegionFuzz.

Static buffer overflow detection techniques tend to report too many false positives fundamentally due to the lack of software execution information. It is very time consuming to manually inspect all the static warnings. In this paper, we propose BovInspector, a framework for automatically validating static buffer overflow warnings and providing suggestions for automatic repair of true buffer overflow warnings for C programs. Given the program source code and the static buffer overflow warnings, BovInspector first performs warning reachability analysis. Then, BovInspector executes the source code symbolically under the guidance of reachable warnings. Each reachable warning is validated and classified by checking whether all the path conditions and the buffer overflow constraints can be satisfied simultaneously. For each validated true warning, BovInspector provides suggestions to automatically repair it with 11 repair strategies. BovInspector is complementary to prior static buffer overflow discovery schemes. Experimental results on real open source programs show that BovInspector can automatically validate on average 60% of total warnings reported by static tools.

Missing checks for untrusted inputs used in security-sensitive operations is one of the major causes of various vulnerabilities. Efficiently detecting and repairing missing checks are essential for prognosticating potential vulnerabilities and improving code reliability. We propose a systematic static analysis approach to detect missing checks for manipulable data used in security-sensitive operations of C/C₊₊ programs and recommend repair references. First, customized securitysensitive operations are located by lightweight static analysis. Then, the assailability of sensitive data used in securitysensitive operations is determined via taint analysis. And, the existence and the risk degree of missing checks are assessed. Finally, the repair references for high-risk missing checks are recommended. We implemented the approach into an automated and cross-platform tool named Vanguard based on Clang/LLVM 3.6.0. Large-scale experimental evaluation on open-source projects has shown its effectiveness and efficiency. Furthermore, Vanguard has helped us uncover five known vulnerabilities and 12 new bugs.