From mboxrd@z Thu Jan  1 00:00:00 1970
From: Rik van Riel <riel@redhat.com>
Subject: Re: sysbench throughput degradation in 4.13+
Date: Tue, 03 Oct 2017 12:02:41 -0400
Message-ID: <1507046561.10046.22.camel@redhat.com>
References: <95edafb1-5e9d-8461-db73-bcb002b7ebef@linux.vnet.ibm.com>
         <CAD9gYJJ9nSAbznEn80hfY3=+YjA8cKw6RztpgW6iDm7rQ0EsFg@mail.gmail.com>
         <50a279d3-84eb-3403-f2f0-854934778037@linux.vnet.ibm.com>
         <20170922155348.zujigkn3o5eylctn@hirez.programming.kicks-ass.net>
         <754f5a9f-5332-148d-2631-918fc7a7cfe9@linux.vnet.ibm.com>
         <20170927093530.s3sgdz2vamc5ka4w@hirez.programming.kicks-ass.net>
         <20170927135820.61cd077f@cuia.usersys.redhat.com>
         <20171002225312.GA24578@codeblueprint.co.uk>
         <20171003083932.3qa7jw2spmi5n5pg@hirez.programming.kicks-ass.net>
Mime-Version: 1.0
Content-Type: multipart/signed; micalg="pgp-sha256";
        protocol="application/pgp-signature"; boundary="=-woq1u+T9Pb8E4KdMKN8+"
Cc: Eric Farman <farman@linux.vnet.ibm.com>,
        ????????? <jinpuwang@gmail.com>,
        LKML <linux-kernel@vger.kernel.org>, jhladky@redhat.com,
        Ingo Molnar <mingo@redhat.com>,
        Christian Borntraeger <borntraeger@de.ibm.com>,
        "KVM-ML (kvm@vger.kernel.org)" <kvm@vger.kernel.org>,
        vcaputo@pengaru.com, Matthew Rosato <mjrosato@linux.vnet.ibm.com>
To: Peter Zijlstra <peterz@infradead.org>,
        Matt Fleming <matt@codeblueprint.co.uk>
Return-path: <linux-kernel-owner@vger.kernel.org>
In-Reply-To: <20171003083932.3qa7jw2spmi5n5pg@hirez.programming.kicks-ass.net>
Sender: linux-kernel-owner@vger.kernel.org
List-Id: kvm.vger.kernel.org


--=-woq1u+T9Pb8E4KdMKN8+
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

On Tue, 2017-10-03 at 10:39 +0200, Peter Zijlstra wrote:
> On Mon, Oct 02, 2017 at 11:53:12PM +0100, Matt Fleming wrote:
> > On Wed, 27 Sep, at 01:58:20PM, Rik van Riel wrote:
> > >=20
> > > I like the simplicity of your approach!=C2=A0=C2=A0I hope it does not=
 break
> > > stuff like netperf...
> > >=20
> > > I have been working on the patch below, which is much less
> > > optimistic
> > > about when to do an affine wakeup than before.
> >=20
> > Running netperf for this patch and Peter's patch shows that Peter's
> > comes out on top, with scores pretty close to v4.12 in most places
> > on
> > my 2-NUMA node 48-CPU Xeon box.
> >=20
> > I haven't dug any further into why v4.13-peterz+ is worse than
> > v4.12,
> > but I will next week.
>=20
> So I was waiting for Rik, who promised to run a bunch of NUMA
> workloads
> over the weekend.
>=20
> The trivial thing regresses a wee bit on the overloaded case, I've
> not
> yet tried to fix it.

In Jirka's tests, your simple patch also came out
on top.

--=20
All rights reversed
--=-woq1u+T9Pb8E4KdMKN8+
Content-Type: application/pgp-signature; name="signature.asc"
Content-Description: This is a digitally signed message part
Content-Transfer-Encoding: 7bit

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2

iQEcBAABCAAGBQJZ07ShAAoJEM553pKExN6Dmn8IALGwCgQBKn74fhmScXEJMlTo
KEurC9KkFw9Fe8lr1XEyzHo1TzAOaCt/fSZva6C3vPeS6IJ1biUjZsQ0ko39AId1
qNL93wXDzqvRUPtEvcxXx7VpcdprnRIp0PL2g6HTfyFT8H6c4+61xuXHTSVSPx8p
j49aZgy/DOdekcvQod63NdPezT4OWyWkGxxHnAO73f4XOW0+TzaPHPh6BRKBOW+L
luc7xnEw8zZuMk4E89xea1spii47KzP0KBFNMOqyQ+smbWBD2X9o8uJMhMNR0XpQ
0G0m6ubYG4qyfCdem2cXmG+AVBbCGnRU6q3hFBzRE1TLVjqFaIjHPE1R5i8DG/A=
=WZHH
-----END PGP SIGNATURE-----

--=-woq1u+T9Pb8E4KdMKN8+--