Improve Upstream Code Quality by Bringing Testing to Patch Level
summary
Downstream testers move regression testing upstream by subscribing to mailing lists, cloning upstream repos, downloading patch series with b4m, applying with git am, compiling and running KVM/kernel tests, and posting results with a tested-by tag. Over two years the team adds tested-by to 144 patches and files 38 upstream reports, catches regressions (including QEMU kernel panics and coredumps) before merge, reduces downstream backports and release risk, and balances effort against limited QE resources.
AI-written summary · about
chapters (7)
- 0:00Upstream testing strategy
- 1:03Downstream bug fix workflow
- 5:23Testing upstream patches
- 11:32Upstream testing tips
- 15:51Upstream testing benefits
- 21:50Balancing upstream workloads
- 22:53Q&A
related talks
19mUpstream kernel CI using patchew and TDC
24mContinuously Integrating the Upstream Linux Kernel on Hardware
28mQuality in embargoed patches
21mThe emerging of the virtual QA team for Linux kernel
34mLinux Kernel Testing: Providing Better Results to the Upstream Community
39mDay-to-day Testing of Linux 'Next' Kernel Branch - My Story
32mReusing the BPF CI
14mThe Power of Backports : Business Impact and Automation Strategies