TODOs

RFCs

• CPU reclaiming for SCHED_DEADLINE (Luca Abeni): implements CPU reclaiming (using the GRUB algorithm) for SCHED_DEADLINE. This feature allows SCHED_DEADLINE tasks to consume more than their reserved runtime, up to a maximum fraction of the CPU time (so that other tasks are left some spare CPU time to execute), if this does not break the guarantees of other SCHED_DEADLINE tasks.
• v1 - https://lwn.net/Articles/671929/
• v2 - https://lwn.net/Articles/682222/
• v3 - https://lwn.net/Articles/704405/
• v4 - https://marc.info/?l=linux-kernel&m=148309763721395

Known Issues

• admission control vs cpuset/hotplug
• when hotplug events destroy root_domain, information about admitted reservations
• proper PI implementation
• fix deadline inheritance
• implement bandwidth inheritance (BWI)
• M-BWI

Missing features

• group scheduling (cgroups)
• hierarchical DEADLINE server for FIFO/RR
• possibly CFS as well (more tricky)
• make SCHED_DEADLINE usable for unpriv users
• reclaim
• stochastic
• limits to avoid DoS
• adding single CPUs affinities
• schedutil (cpufreq) integration
• DEADLINE and energy aware scheduling
• capacity awareness (asym topologies)
• full energy model
• work conserving scheduler (possible alternatives)
• soft-CBS
• reclaiming by demotion to lower prio classes
• GRUB (see RFCs above)
• self-suspending tasks (alternative CBS wakeup rule)

Testing

• refine testing infrastructure (using LISA ?)

RFD

• behaviour of fork/clone