From owner-freebsd-current  Fri Nov  8 09:17:37 1996
Return-Path: owner-current
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.5/8.7.3) id JAA23286
          for current-outgoing; Fri, 8 Nov 1996 09:17:37 -0800 (PST)
Received: from troutmask.apl.washington.edu (troutmask.apl.washington.edu [128.95.76.54])
          by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id JAA23280;
          Fri, 8 Nov 1996 09:17:34 -0800 (PST)
Received: (from sgk@localhost) by troutmask.apl.washington.edu (8.7.6/8.7.3) id JAA02016; Fri, 8 Nov 1996 09:18:14 -0800 (PST)
From: "Steven G. Kargl" <sgk@troutmask.apl.washington.edu>
Message-Id: <199611081718.JAA02016@troutmask.apl.washington.edu>
Subject: Re: cvs commit: src/usr.bin/systat vmstat.c
In-Reply-To: <199611081338.VAA27430@spinner.DIALix.COM> from Peter Wemm at "Nov 8, 96 09:38:43 pm"
To: peter@spinner.DIALix.COM (Peter Wemm)
Date: Fri, 8 Nov 1996 09:18:14 -0800 (PST)
Cc: jkh@freefall.freebsd.org, freebsd-current@freefall.freebsd.org
X-Mailer: ELM [version 2.4ME+ PL28 (25)]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Sender: owner-current@FreeBSD.ORG
X-Loop: FreeBSD.org
Precedence: bulk

[cc: to freebsd-current]

According to Peter Wemm:
> "Jordan K. Hubbard" wrote:
> >   Log:
> >   Wow, make the vmstat display actually usable!  This is a cosmetic
> >   fix, but low-impact and worth the comeasurate risk.  I [jkh] can't even
> >   test it in -current since systat's vmstat display is completely broken
> >   there at the moment, printing out: "The alternate system clock has died!"
> 
> Show this to Bruce..  According to him, this cannot happen.  The 
> "alternate system clock" is the rtc (irq 8) and it has stopped sending the 
> 128Hz statistics and/or 1024Hz profiling clock.  It's not just vmstat that 
> is affected, check 'ps' and 'top'..  you'll also find that anything cpu 
> intensive through time(1) gets all it's time credited to "system" rather 
> than part "user" and part "system".
> 

I saw the same behaviour while profiling some code.  My routines
were running infinitely fast, and the system was taking all the
time during execution.  Not exactly the expected behaviour.
See the thread "Is profiling broken" on the freebsd-current mailing
list.  Bruce suggested that I may have some hardware problem,
but a "make world" and new kernel seems to fix the problem.

-- 
Steve