From owner-freebsd-current  Thu Sep 11 12:43:05 1997
Return-Path: <owner-freebsd-current>
Received: (from root@localhost)
          by hub.freebsd.org (8.8.7/8.8.7) id MAA12685
          for current-outgoing; Thu, 11 Sep 1997 12:43:05 -0700 (PDT)
Received: from millenia.srrc.usda.gov ([199.78.118.11])
          by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id MAA12677
          for <freebsd-current@freebsd.org>; Thu, 11 Sep 1997 12:42:54 -0700 (PDT)
Received: from nola.srrc.usda.gov (localhost [127.0.0.1])
	by millenia.srrc.usda.gov (8.8.7/8.8.7) with ESMTP id OAA00325
	for <freebsd-current@freebsd.org>; Thu, 11 Sep 1997 14:41:53 -0500 (CDT)
Message-ID: <34184980.49639858@nola.srrc.usda.gov>
Date: Thu, 11 Sep 1997 14:41:52 -0500
From: Glenn Johnson <gjohnson@nola.srrc.usda.gov>
Organization: USDA-ARS-SRRC
X-Mailer: Mozilla 4.03b8 [en] (X11; I; FreeBSD 3.0-CURRENT i386)
MIME-Version: 1.0
To: freebsd-current@freebsd.org
Subject: Strange output in "Top"
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: owner-freebsd-current@freebsd.org
X-Loop: FreeBSD.org
Precedence: bulk

I just did a "make world" on my -current SMP system. I started up a
lengthy (weeks to complete) calculation and then checked 'top'. The WCPU
and CPU fields give me extremely low numbers. When no other processes
are running, the WCPU and CPU field hover around 0-1% for the
calculation process and the %idle field is around 98-99%. Strangely, if
I increase the system load, the WCPU and CPU %values go up, and keep
going up as the load goes up, but not to what I think they should be.
What the values should be is about 50% (100% of a single CPU) with no
other activity. The percentage of this process should decrease as I
increase the load.

The system load numbers are what I would expect and the calculation is
proceeding at its normal speed. So I guess either 'top' is calculating
wrong or is getting the wrong information. This behavior began today for
me with a fresh cvsup, but note that I don't 'make world' every day so I
don't know when the problem might have manifested in the code. Of course
it could be some quirk with my machine.

Thanks.
--
Glenn Johnson
gjohnson@nola.srrc.usda.gov