[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Mon, 04 Apr 2016 15:45:16 -0400
From: Rik van Riel <riel@...hat.com>
To: Peter Zijlstra <peterz@...radead.org>,
Ingo Molnar <mingo@...nel.org>
Cc: Jiri Olsa <jolsa@...hat.com>, James Hartsock <hartsjc@...hat.com>,
Srivatsa Vaddagiri <vatsa@...ux.vnet.ibm.com>,
Kirill Tkhai <ktkhai@...allels.com>,
linux-kernel@...r.kernel.org
Subject: Re: [RFC] sched: unused cpu in affine workload
On Mon, 2016-04-04 at 15:23 +0200, Peter Zijlstra wrote:
> On Mon, Apr 04, 2016 at 11:38:44AM +0200, Ingo Molnar wrote:
> >
> > We'd upgrade that to O(nr_cpus^2), which is totally unrealistic
> > with 16,000 CPUs
> > even in a slowpath - but it would probably cause problems even with
> > 120 CPUs. It
> > will get quadratically worse as the number of CPUs in a system
> > increases on its
> > current exponential trajectory ...
> The arbitrary affinity thing is I think a packing problem, which is
> NP
> hard IIRC.
An optimal solution is NP hard.
Heuristics that "move tasks with pressure" may be
much more doable, and lead to perfectly satisfactory
results, especially if most migrations happen within
a socket (and the same shared L3 cache).
--
All Rights Reversed.
Download attachment "signature.asc" of type "application/pgp-signature" (474 bytes)
Powered by blists - more mailing lists