Exploring Function-Level Code Metrics and Developer Attributes for Linux Kernel Vulnerabilities
summary
Dataset covers about 11,000 kernel CVEs and roughly 8,700 VFC‑VIC pairs collected from NVD and upstream patches. Analysis finds that authors of VICs have lower prior activity and lower code familiarity at file, subdirectory and kernel levels, while maintainers are unexpectedly overrepresented among VICs. Function-level structural metrics are higher in fixes (cyclomatic complexity, max control depth, fan‑out), memory‑operation differences are mixed, and most CVEs are memory safety, null‑dereference or concurrency bugs concentrated in drivers, networking and filesystems.
AI-written summary · about
chapters (7)
- 0:00Linux kernel vulnerability research
- 1:45Research contributions and data collection
- 3:37Developer metrics analysis
- 6:51Code metrics results
- 12:57Vulnerability distribution
- 15:49Key findings and conclusions
- 16:20Q&A
related talks
05mLightning Talk: Pruning the Kernel CVEs With Code Reachability Analysis
20mPrioritizing the Linux OS Hardening and CVE Mitigation
36mLinux Kernel CVEs: What Has Caused So Many to Suddenly Show Up?
28mKernel CVEs are Alive, but Do Not Panic!
43mDo Certain Types Of Developers Or Teams Write More Secure Code?
32mPruning Kernel CVEs With Code Reachability Analysis
26mConfig Based CVE Matching for Linux Kernel
23mSoK: Automating Kernel Vulnerability Discovery and Exploit Generation