Many Scheduler Updates In The Linux 6.15 Kernel
Merged last week for the Linux 6.15 merge window were a big set of scheduler updates.
The kernel scheduler updates this cycle were scattered across the board with a lot of material touching different pieces but no single overly exciting addition. Of course, all the small scheduler optimizations and efficiency improvements add up and overall the Linux kernel scheduler code continues certainly moving in the right direction especially with today's high core count processors as well as the hybrid core designs.
Per the since-merged pull request, scheduler feature changes for Linux 6.15 include:
The kernel scheduler updates this cycle were scattered across the board with a lot of material touching different pieces but no single overly exciting addition. Of course, all the small scheduler optimizations and efficiency improvements add up and overall the Linux kernel scheduler code continues certainly moving in the right direction especially with today's high core count processors as well as the hybrid core designs.
Per the since-merged pull request, scheduler feature changes for Linux 6.15 include:
"Core & fair scheduler changes:
- Cancel the slice protection of the idle entity
- Reduce the default slice to avoid tasks getting an extra tick
- Force propagating min_slice of cfs_rq when {en,de}queue tasks
- Refactor can_migrate_task() to [eliminate] looping
- Add unlikey branch hints to several system calls
- Optimize current_clr_polling() on certain architectures
Deadline scheduler:
- Remove redundant dl_clear_root_domain call
- Move dl_rebuild_rd_accounting to cpuset.h
Uclamp:
- Use the uclamp_is_used() helper instead of open-coding it
- Optimize sched_uclamp_used static key enabling
Scheduler topology support:
- Ignore special tasks when rebuilding domains
- Add wrappers for sched_domains_mutex
- Generalize unique visiting of root domains
- Rebuild root domain accounting after every update
- Remove partition_and_rebuild_sched_domains
- Stop exposing partition_sched_domains_locked
RSEQ:
- Update kernel fields in lockstep with CONFIG_DEBUG_RSEQ=y
- Fix segfault on registration when rseq_cs is non-zero
- selftests: Add rseq syscall errors test
- selftests: Ensure the rseq ABI TLS is actually 1024 bytes
Membarriers:
- Fix redundant load of membarrier_state
Scheduler debugging:
- Introduce and use preempt_model_str()
- Make CONFIG_SCHED_DEBUG unconditional
Fixes and cleanups:
- Always save/restore x86 TSC sched_clock() on suspend/resume
- Misc fixes and cleanups"