From owner-freebsd-smp Mon Feb 3 22: 7:26 2003 Delivered-To: freebsd-smp@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 7D54337B401 for ; Mon, 3 Feb 2003 22:07:24 -0800 (PST) Received: from humphrey.dyndns.org (12-230-212-231.client.attbi.com [12.230.212.231]) by mx1.FreeBSD.org (Postfix) with ESMTP id 0D59043F79 for ; Mon, 3 Feb 2003 22:07:20 -0800 (PST) (envelope-from brendy33@attbi.com) Received: from bigboot.humphrey.world (unknown [192.168.8.3]) by humphrey.dyndns.org (Postfix) with ESMTP id 2A47343A25B; Mon, 3 Feb 2003 23:00:17 -0800 (PST) Content-Type: text/plain; charset="iso-8859-1" From: Brendon and Wendy To: Chris Luke Subject: Re: mp_machdep.c patch Date: Mon, 3 Feb 2003 22:07:19 -0800 User-Agent: KMail/1.4.3 References: <200302032001.23528.brendy33@attbi.com> <20030204041753.GA17107@flix.net> In-Reply-To: <20030204041753.GA17107@flix.net> Cc: freebsd-smp@freebsd.org MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Message-Id: <200302032207.19474.brendy33@attbi.com> Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org Thanks for the reply Chris. Based on your "top" output Id have to say that HT is not working right on= this=20 machine at least. I see all proceses on cpus 0 and 1, exactly the same as= =20 prior to cvsuping and rebuilding. In fact I wonder whether the kernel has= =20 only seeing one physical CPU and is doing HT on that CPU - completely=20 ignoring the second. Cheers, Brendon On Monday 03 February 2003 08:17 pm, you wrote: > Brendon and Wendy wrote (on Feb 04): > > Question - how does one determine whether the hyperthreading working?= My > > DMESG is below, as well as output from top. Booting with HTT disabled > > looks the same to me. > > Simple visual method: run "top". On SMP systems with SMP active it > displays an extra column displaying the CPU number that a given process > last ran or is running on. > > Each HT instance gets its own CPU number, as listed in dmesg. For > example (from an old Quad Xeon): > > PID USERNAME PRI NICE SIZE RES STATE C TIME WCPU CPU COMM= AND > 332 mysql 2 0 32032K 13316K poll 2 42.6H 2.83% 2.83% mysq= ld > 12808 chrisy 28 0 1968K 1148K CPU3 1 0:00 14.86% 1.42% top > 12805 chrisy 18 0 1320K 1004K opause 0 0:00 1.75% 0.24% tcsh > 12802 root 2 0 4852K 1448K sbwait 3 0:00 0.66% 0.15% sshd > 194 namedb 2 0 14564K 13832K select 3 98:23 0.00% 0.00% name= d > > Note the "C" column. > > I'm finding that unlike on MP boards, where you see a more or less > even distribution of processes across the procesors, on a HT > system, mostly stuff sits on CPU0 with occasional showings of CPU1. > (This is alongside my 2nd CPU not firing up, but I'll followup with > more on that tomorrow). > > Chris. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message