Backporting is so 1993
summary
Claim: long-term backporting no longer scales; tracking upstream is the practical way to deliver security fixes. Example: Zephyr releases add ~1,900 commits per cycle, many impossible to cherry-pick; solution: run continuous CI against tip with sanitycheck/qemu and hardware smoke tests, perform regression bisection and bundle stable bite-sized merges, and use kernel-ci/Linux-next style boards. Foundries.io-style pipelines produce actionable upgrade points and faster fixes in the field, and the process demonstrably reduces risky, boot-breaking backports.
AI-written summary · about
chapters (8)
- 0:00Modern IoT maintenance challenges
- 3:53Traditional backporting fails
- 10:19Zephyr upstream case study
- 16:17Linux upstream strategies
- 21:17Rethinking maintenance models
- 30:27Certification and compliance feedback
- 34:02Consumer update expectations
- 42:30Q&A
related talks
36mNo, It's (Still) Never Too Late to Upstream Your Legacy Linux Based Platforms
36mNo, It's Never Too Late to Upstream Your Legacy Linux Based Platform- 36mNo, It's (Still) Never Too Late to Upstream Your Legacy Linux Based Platforms
17mSustaining Yesterday’s Linux: Delivering Security for Today’s Workloads
39mThe Life of a Kernel Bugfix
40mContributing to Zephyr vs (Linux and U-boot)
31mSponsored Session: Can We Do Better than Shipping 4 Year Old Software in Products?
21mFrom CVEs To Compliance: Automating Embedded Linux Kernel Security