From mboxrd@z Thu Jan 1 00:00:00 1970 From: Dario Faggioli Subject: Re: [PATCH 03 of 10 v2] xen: sched_credit: let the scheduler know about node-affinity Date: Fri, 21 Dec 2012 17:07:58 +0100 Message-ID: <1356106078.15403.59.camel@Abyss> References: <06d2f322a6319d8ba212.1355944039@Solace> <50D3414D.8080901@eu.citrix.com> <50D47235.4090106@eu.citrix.com> Mime-Version: 1.0 Content-Type: multipart/mixed; boundary="===============2358770386315652851==" Return-path: In-Reply-To: <50D47235.4090106@eu.citrix.com> List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Sender: xen-devel-bounces@lists.xen.org Errors-To: xen-devel-bounces@lists.xen.org To: George Dunlap Cc: Marcus Granado , Dan Magenheimer , Ian Campbell , Anil Madhavapeddy , Andrew Cooper , Juergen Gross , Ian Jackson , "xen-devel@lists.xen.org" , Jan Beulich , Daniel De Graaf , Matt Wilson List-Id: xen-devel@lists.xenproject.org --===============2358770386315652851== Content-Type: multipart/signed; micalg=pgp-sha1; protocol="application/pgp-signature"; boundary="=-VHCcvgheZ7clcavNtEty" --=-VHCcvgheZ7clcavNtEty Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable On Fri, 2012-12-21 at 14:29 +0000, George Dunlap wrote:=20 > > Sorry for that, I probably spent so much time buried, as you where > > saying, in the > > various nested loops and calls, that I lost the context a little bit! := -P >=20 > OK, that makes sense -- I figured it was something like that. Don't=20 > feel too bad about missing that connection -- we're all fairly blind to= =20 > our own code, and I only caught it because I was trying to figure out=20 > what was going on. > Yeah, thanks, and no, I won't let this let me down to much, Even if this was quite a big one. After all, that's what we have patch review for! > That's why we do patch review. :-) >=20 Hehe, I see we agree. :-) > Honestly, the whole "steal work" idea seemed a bit backwards to begin=20 > with, but now that we're not just dealing with "possible" and "not=20 > possible", but with "better" and "worse", the work-stealing method of=20 > load balancing sort of falls down. > > [snip] >=20 > But that's kind of a half-baked idea at this point. >=20 Yes, this whole stealing work machinery may be rethought a bit. However, let's get something sane in re NUMA load balancing ASAP, as we planned, and then we'll see whether/how to rework it with both simplicity and effectiveness in mind. > > Ok, I think the problem I was describing is real, and I've seen it happ= ening and > > causing performances degradation. However, as I think a good solution > > is going to > > be more complex than I thought, I'd better repost without this > > function and deal with > > it in a future separate patch (after having figured out the best way > > of doing so). Is > > that fine with you? >=20 > Yes, that's fine. > Ok, I'll sort out all your comments and try to post v3 in early January, so that you'll find it in your inbox as soon as you'll be back from vacations! :-) > Thanks, Dario. >=20 Thanks to you, Dario --=20 <> (Raistlin Majere) ----------------------------------------------------------------- Dario Faggioli, Ph.D, http://retis.sssup.it/people/faggioli Senior Software Engineer, Citrix Systems R&D Ltd., Cambridge (UK) --=-VHCcvgheZ7clcavNtEty 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 v1.4.12 (GNU/Linux) iEYEABECAAYFAlDUiV4ACgkQk4XaBE3IOsSnFgCdE3L7W3LWJbMHf+dht8l/hfty 1swAnjJGfft3xiAYaFLBX6IjJfVozfED =/LeZ -----END PGP SIGNATURE----- --=-VHCcvgheZ7clcavNtEty-- --===============2358770386315652851== Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ Xen-devel mailing list Xen-devel@lists.xen.org http://lists.xen.org/xen-devel --===============2358770386315652851==--