Mainline RT-preempt patchset


 * Summary: Mainline RT-preempt patchset


 * Proposer: Tim Bird

Description
The RT-preempt patch set provides real-time scheduling performance for Linux. Parts of this patch set have been mainlined over time, but there are still significant parts that remain out-of-tree.

This project would consist of mainlining the remaining portions of the RT-preempt patch set, to avoid integration costs for those companies that use this patchset, and to decrease the maintenance cost of the patchset.

Related work
various kernel versions of the patchset tree regarding the status of the RT-preempt patch
 * http://www.kernel.org/pub/linux/kernel/projects/rt/ - index of patchsets for
 * https://lkml.org/lkml/2010/8/3/133 - announcement on LKML of version 2.6.33.7-rt29
 * https://lkml.org/lkml/2010/12/21/110 - announcement on LKML of version 2.6.33.7.2-rt30 of the patchset and the hope that the next version will be based on 2.6.37
 * Real-Time Linux Wiki
 * See especially https://rt.wiki.kernel.org/index.php/RT_PREEMPT_HOWTO
 * http://lwn.net/Articles/319544/ - Interview: the return of the realtime preemption
 * This is an interview with Ingo Molnar and Thomas Gleixner, from February 2009,
 * Realtime Preemption - unfortunately out-of-date wiki page on this work

Scope
Probably measured in man-years. Would be good to define intermediate milestones for progress, and work towards those.

Contractor Candidates
List possible contractors for this work (can be yourself)

Comments
Wolfgang Denk writes: It might make sense to co-ordinate such activities with the OSADL who are already dedicated to that very task.