Panel Discussion: Super Long Term Kernel Maintenance
summary
Long-term kernels run on a time-based release (one per year) and are currently supported for six years thanks to Google and Linaro funding extended testing; kernelci becomes the central test hub with Canonical/CIP integrations and vendor labs adding hardware. Auto-selection ML tools flag likely stable fixes (research from Sasha/Julia), EU funding backs a regression-tracking pilot, and Renesas-style upstream-first device trees prove rolling to new mainline kernels is feasible. Vendors must send downstream patches upstream to reduce huge backport queues.
AI-written summary · about
chapters (11)
- 0:00Introduction
- 2:12How longterm maintenance was introduced
- 5:41Experience with longterm maintenance
- 11:02Chris
- 20:07Areas for improvement
- 26:29Testing coverage
- 30:34Improving deployment
- 34:15Extended LTS support
- 39:52Recruiting new maintainers
- 47:28AI for longterm maintenance
- 51:19Future super longterm maintenance
related talks
40mFinding the Best Way to Use Linux Kernel - Status Update
36mNo, It's (Still) Never Too Late to Upstream Your Legacy Linux Based Platforms
09mAspects of Dependable Linux Systems
41mIs Linux Kernel Development Good Enough to Make Your Life Depend on It?
30mKeynote: Dirk Hohndel, VP, VMware, in Conversation with Linux Creator Linus Torvalds
25m[PATCH v22] Landing a Large Series
28mYou may be a Linux Kernel Maintainer - and Not Know It!
33mScaling KVM and Its Community