Re: [RFC PATCH v2 3/3] sched: introduce synchronized idle injection

[Jacob Pan]

On Tue, 10 Nov 2015 18:00:10 +0100
Peter Zijlstra <peterz@xxxxxxxxxxxxx> wrote:

> > Different per CPU timer may intercept parameter changes at slightly
> > different time, so there is a race condition such that some CPUs may
> > catch the period change later by one period, which results in a
> > correct period change but at a different time, i.e. out of sync.  
> 
> Ah yes. So if the locking hurts I can come up with a lockless
> algorithm for this. Shouldn't be too hard.
Great! the current patch was designed to be lockless but a
little awkward. My idea was to have a common start time, then we don't
need to worry about CPUs out of sync, at most they will be off by one
period then catch up.

Thanks,

Jacob
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/