Date: Tue, 20 Apr 2004 01:58:37 -0700 From: Kris Kennaway <kris@obsecurity.org> To: Rumen Telbizov <altares@e-card.bg> Cc: Kris Kennaway <kris@obsecurity.org> Subject: Re: HTT and SMP question Message-ID: <20040420085837.GA57184@xor.obsecurity.org> In-Reply-To: <20040420085322.GV32493@e-card.bg> References: <20040419085841.GB64662@freenix.no> <20040419090049.GA51659@chihiro.leafy.idv.tw> <20040419092052.GD64662@freenix.no> <20040419093523.GA34419@xor.obsecurity.org> <20040420073837.GS32493@e-card.bg> <20040420081334.GA56291@xor.obsecurity.org> <20040420085322.GV32493@e-card.bg>
next in thread | previous in thread | raw e-mail | index | archive | help
--Dxnq1zWXvFF0Q93v Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable On Tue, Apr 20, 2004 at 11:53:22AM +0300, Rumen Telbizov wrote: > > > Would you please point out some cases when HTT with machdep.hlt_logic= al_cpus=3D0 > > > causes worse performance than machdep.hlt_logical_cpus=3D1? > > > I am using an HTT Xeon since recently and in my tests it showed that > > > machdep.hlt_logical_cpus=3D0 is better! > > > A simple ubench (spawning 2 processes) gives better results in 0 mode! > >=20 > > My parallel package builds are slower on build machines with HTT. > > It's not a magic bullet. > >=20 > > Kris >=20 > Hmm ... interesting. My simple/stupid synthetic tests > showed just the opposite. Anyway. Unfortunatelly I didn't > try to buildworld/kernel with -j2 (say). Shame on me! >=20 > Would you point to a case where machdep.hlt_logical_cpus=3D0 does > show better performance results? httpd ? >=20 > Is it not that exactly _parallel processing_ should be faster with HTT > than whithout ? No, because HTT is not an extra CPU. Kris --Dxnq1zWXvFF0Q93v Content-Type: application/pgp-signature Content-Disposition: inline -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.4 (FreeBSD) iD8DBQFAhOY9Wry0BWjoQKURAq+uAKD88qQMW1aBvAkMCtOnFktA6V0/RwCeNDIj NPOIgto4SZfQCwPbFHMVo7w= =3FzU -----END PGP SIGNATURE----- --Dxnq1zWXvFF0Q93v--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20040420085837.GA57184>