Messages in this thread | | | Date | Tue, 17 May 2011 16:02:33 +0200 | From | Tejun Heo <> | Subject | Re: [patch V3] percpu_counter: scalability works |
| |
Hey,
On Tue, May 17, 2011 at 08:55:33AM -0500, Christoph Lameter wrote: > Well lets just adopt the system that vm statistics use. Bound the error by > time and batch and allow the user to change the batch if more accuracy is > desired.
Yeah, that would be the better long term solution.
> The _sum function is optional and should it should be explained that the > result *could* be better but dont count on it.
We can implement force-flush for dire situations, similar to expedited RCU flush thing.
Thanks.
-- tejun
| |