Messages in this thread | | | Subject | Re: [RFC][PATCH] sched: Optimize cgroup pick_next_task_fair | From | Peter Zijlstra <> | Date | Sat, 11 Feb 2012 16:02:17 +0100 |
| |
On Sat, 2012-02-11 at 07:56 +0100, Mike Galbraith wrote: > On Sat, 2012-02-11 at 06:05 +0100, Peter Zijlstra wrote: > > Since commit 2f36825b1 ("sched: Next buddy hint on sleep and preempt > > path") it is likely we pick a new task from the same cgroup, doing a put > > and then set on all intermediate entities is a waste of time, so try to > > avoid this. > > Good idea, we need to lose some weight.
I actually set out to do something else, but then ended up here.. weird how that works ;-)
My initial idea was to try and get rid of the pick_next_entity() downward walk by keeping that data up-to-date when we dequeue/enqueue.
That all got a tad convoluted, I'll have to try it again. Anyway, that ended up needing to rework the put/set stuff, and that part wasn't as bad.
| |