Unifying and improving test regression reporting and tracking
summary
A proof-of-concept ingests CarCI and other kernel CI outputs into the KCB aggregator, post-processes test histories and logs to filter infrastructure noise and group similar failures, and feeds curated issues into Regzbot. Experiments over a few months automatically classify many false positives, surface device-specific failure rates and enable targeted re-runs with extra tracing, and show that flaky cases sometimes need ~200 runs to produce stable failure signatures for correlation.
AI-written summary · about
chapters (5)
- 0:00Introduction to regression tracking
- 2:29Current regression reporting gaps
- 7:46Improving regression data quality
- 14:51Community feedback and solutions
- 22:50Q&A and future directions
related talks
33mKernel CI – How Far Can It Go?
37mKernel CI – How Far Can It Go?
40mKernelCI for Embedded: Enhancing Reliability through Scalable Test Automation
40mOne year into the new KernelCI
01h 00mContinuous Regression Testing for Safer and Faster Refactoring in C++
49mCrash and test failure analysis with AI... no wait, ehm, with regexes!
34mLinux Kernel Testing: Providing Better Results to the Upstream Community
34mMerging your Kernel Testing Code into KernelCI