Date: Mon, 13 Mar 2006 10:41:37 -0500 From: Bill Moran <wmoran@collaborativefusion.com> To: Pete French <petefrench@ticketswitch.com> Cc: bseklecki@collaborativefusion.com, freebsd-amd64@freebsd.org Subject: Re: How is hyperthreading handled on amd64? Message-ID: <20060313104137.5fcc80ea.wmoran@collaborativefusion.com> In-Reply-To: <E1FIomN-00015b-4x@dilbert.firstcallgroup.co.uk> References: <20060313094356.d7f0988a.wmoran@collaborativefusion.com> <E1FIomN-00015b-4x@dilbert.firstcallgroup.co.uk>
next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, 13 Mar 2006 15:16:15 +0000 Pete French <petefrench@ticketswitch.com> wrote: > > I don't see this behaviour on i386 - my desktop, for example uses only > > CPU 0 if the sysctl is 0, and switches between 0 and 1 when the sysctl > > is 1. > > Uh, a typo on my part - substitute *machine* for *kernel*. Sorry! > > A machine with one physical processor works fine - can enable and disable > hyperthreading as per the manual. A machine with two physical processors > always gets hyperthreading enabled though. All of these machines are > running 6.0 currently - though I will verify that the bug is still there > on 6.1 when I get a moment to do an update on them. Again, not what I'm seeing here. It appears as if the machine behaves identically no matter what machdep.hyperthreading_allowed is set to. The machine I'm testing on only has a single physical processor. I've been running ubench multiple times to test this, and I see processes on CPUs 1 and 0 no matter what I set the above sysctl to. I also don't see any difference in ubench's performance numbers. -- Bill Moran Collaborative Fusion Inc.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20060313104137.5fcc80ea.wmoran>