Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 18 May 2005 11:44:13 -0700
From:      Astrodog <astrodog@gmail.com>
Cc:        freebsd-amd64@freebsd.org
Subject:   Re: FreeBSD on dual core Opterons - stupid buildworld test
Message-ID:  <2fd864e05051811442152fa@mail.gmail.com>
In-Reply-To: <200505181441.39457.jkim@niksun.com>
References:  <428B6FC1.3000907@fsn.hu> <1116437340.69035.0.camel@hood.oook.cz> <20050518180521.GB9719@odin.ac.hmc.edu> <200505181441.39457.jkim@niksun.com>

next in thread | previous in thread | raw e-mail | index | archive | help
On 5/18/05, Jung-uk Kim <jkim@niksun.com> wrote:
> On Wednesday 18 May 2005 02:05 pm, Brooks Davis wrote:
> > On Wed, May 18, 2005 at 07:29:00PM +0200, Pav Lucistnik wrote:
> > > Attila Nagy p=ED?e v st 18. 05. 2005 v 18:39 +0200:
> > > > Hello,
> > > >
> > > > I did some buildworld tests on an HP BL45p with 4 2,2 GHz
> > > > Opterons.
> > > >
> > > > Sequential time make buildworld -jX times (if there is a
> > > > missing iteration, then the build has failed):
> > > >
> > > > http://people.fsn.hu/~bra/bl45p/dmesg.boot
> > > >
> > > >  Hyperthreading: 2 logical CPUs
> > >
> > > So dual-core Opterons report into system as Hyperthreading?
> > > Uncool.
> >
> > What the AMD reps said at a session I attended last year is that
> > they report as supporting hyper threading, but set another bit to
> > say that they are actually real cores.  We probably need to find
> > out what that bit is.
>=20
> FYI, Linux says:
>=20
> http://lxr.linux.no/source/include/asm-x86_64/cpufeature.h#L78
> http://lxr.linux.no/source/arch/x86_64/kernel/setup.c#L834
>=20
> Jung-uk Kim
>=20

Opteron reports Hyperthreading, because it tries to take advantage of
the licencing benifit. Keep in mind, HT P4s are counted as a single
CPU for the most part, in a licence. Opteron may not. By acting as
though its a HT processor, it gets that same benifit.



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?2fd864e05051811442152fa>