One of the niftiest kernel innovations to be upstreamed into Linux over the past year was [1]sched_ext for extensible scheduler behavior in allowing kernel schedulers to be implemented via BPG programs. Sched_ext can allow for interesting scheduler improvements [2]with a variety of use-cases and showed much potential even before being upstreamed. The work on sched_ext isn't yet over though and yet more improvements landed for Linux 6.16.

Last week Tejun Heo submitted all of the sched_ext feature improvements for the Linux 6.16 kernel. This cycle sched_ext is bringing more in-kernel CPU selection improvements with better CPU topology awareness and other enhancements. There is also ongoing work around hierarchical scheduler support within sched_ext.

Tejun Heo summed up the sched_ext changes in Linux 6.16 within [3]the pull request as:

- More in-kernel idle CPU selection improvements. Expand topology awareness coverage add scx_bpf_select_cpu_and() to allow more flexibility. The idle CPU selection kfuncs can now be called from unlocked contexts too.

- A bunch of reorganization changes to lay the foundation for multiple hierarchical scheduler support. This isn't ready yet and the included changes don't make meaningful behavior differences. One notable change is replacing some static_key tests with dynamic tests as the test results may differ depending on the scheduler instance. This isn't expected to cause meaningful performance difference.

- Other minor and doc updates.

We are just getting started in realizing all the possibilities afforded to Linux by sched_ext so it will be very interesting to see what more innovations come around this extensible scheduler class over the months or perhaps even year(s) ahead.



[1] https://www.phoronix.com/search/sched_ext

[2] https://www.phoronix.com/news/NVIDIA-Talks-Up-Sched-Ext

[3] https://lore.kernel.org/lkml/aDYZnzyBFzUn_EYA@slm.duckdns.org/