From nobody@FreeBSD.ORG  Sat Sep  9 14:36:16 2000
Return-Path: <nobody@FreeBSD.ORG>
Received: by hub.freebsd.org (Postfix, from userid 32767)
	id 517EC37B423; Sat,  9 Sep 2000 14:36:16 -0700 (PDT)
Message-Id: <20000909213616.517EC37B423@hub.freebsd.org>
Date: Sat,  9 Sep 2000 14:36:16 -0700 (PDT)
From: riccardo@torrini.org
Sender: nobody@FreeBSD.ORG
To: freebsd-gnats-submit@FreeBSD.org
Subject: Load average (either with uptime both top) go over 1.00 without any load
X-Send-Pr-Version: www-1.0

>Number:         21155
>Category:       kern
>Synopsis:       Load average (either with uptime both top) go over 1.00 without any load
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    freebsd-bugs
>State:          closed
>Quarter:        
>Keywords:       
>Date-Required:  
>Class:          sw-bug
>Submitter-Id:   current-users
>Arrival-Date:   Sat Sep 09 14:40:01 PDT 2000
>Closed-Date:    Wed Aug 15 09:42:58 PDT 2001
>Last-Modified:  Thu Aug 16 04:30:02 PDT 2001
>Originator:     Riccardo Torrini
>Release:        5.0-CURRENT
>Organization:
>Environment:
FreeBSD trudy.home.torrini.org 5.0-CURRENT FreeBSD 5.0-CURRENT #0: Sat Sep  9 10:19:55 CEST 2000     root@trudy.home.torrini.org:/usr/obj/usr/src/sys/TRUDY  i386
>Description:
Load average (either with uptime both top) go over 1.00 without any
program running (I removed also X and tryed on console) but cpu is
really free, as in top, line CPU states:  0.0% user,  0.0% nice,
5.1% system,  0.0% interrupt, 94.9% idle
>How-To-Repeat:
# uptime
11:30PM  up  2:03, 3 users, load averages: 1.26, 1.14, 1.08

(or see first line of top)
>Fix:


>Release-Note:
>Audit-Trail:
State-Changed-From-To: open->closed 
State-Changed-By: mike 
State-Changed-When: Wed Aug 15 09:42:58 PDT 2001 
State-Changed-Why:  

I believe this was a problem with the initial SMPng work.  In any 
case, I haven't seen this happen in 5.0-CURRENT in a long time, so 
I'm closing this PR. 

http://www.FreeBSD.org/cgi/query-pr.cgi?pr=21155 

From: Bruce Evans <bde@zeta.org.au>
To: <mike@FreeBSD.ORG>
Cc: <riccardo@torrini.org>, <freebsd-gnats-submit@FreeBSD.ORG>
Subject: Re: kern/21155: Load average (either with uptime both top) go over
 1.00 without any load
Date: Thu, 16 Aug 2001 21:24:09 +1000 (EST)

 On Wed, 15 Aug 2001 mike@FreeBSD.ORG wrote:
 
 > Synopsis: Load average (either with uptime both top) go over 1.00 without any load
 >
 > State-Changed-From-To: open->closed
 > State-Changed-By: mike
 > State-Changed-When: Wed Aug 15 09:42:58 PDT 2001
 > State-Changed-Why:
 >
 > I believe this was a problem with the initial SMPng work.  In any
 > case, I haven't seen this happen in 5.0-CURRENT in a long time, so
 > I'm closing this PR.
 
 This has been reported again, for 4.3 in PR 27334.  It is believed to
 be caused by a process getting in sync with the load average sampler.
 If this is the cause, then the problem has very little to do with
 SMPng, but later SMPng work may have reduced it by changing the timing
 of rescheduling.
 
 Bruce
 
>Unformatted:
