From owner-freebsd-stable  Fri Sep 21 10:41:41 2001
Delivered-To: freebsd-stable@freebsd.org
Received: from mail-blue.research.att.com (mail-blue.research.att.com [135.207.30.102])
	by hub.freebsd.org (Postfix) with ESMTP
	id E0AED37B411; Fri, 21 Sep 2001 10:41:36 -0700 (PDT)
Received: from alliance.research.att.com (alliance.research.att.com [135.207.26.26])
	by mail-blue.research.att.com (Postfix) with ESMTP
	id 386194CE73; Fri, 21 Sep 2001 13:41:36 -0400 (EDT)
Received: from chips.research.att.com (chips.research.att.com [135.207.27.139])
	by alliance.research.att.com (8.8.7/8.8.7) with ESMTP id NAA14218;
	Fri, 21 Sep 2001 13:41:35 -0400 (EDT)
Received: (from chuck@localhost)
	by chips.research.att.com (SGI-8.9.3/8.8.5) id NAA81917;
	Fri, 21 Sep 2001 13:41:34 -0400 (EDT)
Date: Fri, 21 Sep 2001 13:41:34 -0400
From: Chuck Cranor <chuck@research.att.com>
To: freebsd-stable@FreeBSD.org
Cc: mike@FreeBSD.org
Subject: more on 4.4-RELEASE kernel deadlock ("FFS node")
Message-ID: <20010921134134.T7319@chips.research.att.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
User-Agent: Mutt/1.2.5i
Organization: AT&T Labs-Research
Sender: owner-freebsd-stable@FreeBSD.ORG
Precedence: bulk
List-ID: <freebsd-stable.FreeBSD.ORG>
List-Archive: <http://docs.freebsd.org/mail/> (Web Archive)
List-Help: <mailto:majordomo@FreeBSD.ORG?subject=help> (List Instructions)
List-Subscribe: <mailto:majordomo@FreeBSD.ORG?subject=subscribe%20freebsd-stable>
List-Unsubscribe: <mailto:majordomo@FreeBSD.ORG?subject=unsubscribe%20freebsd-stable>
X-Loop: FreeBSD.ORG

this is a followup to my previous message:

first, the problem i am seeing is almost identical to the one reported
in kern/19479 (see http://www.FreeBSD.org/cgi/query-pr.cgi?pr=19479) 
which was closed by mike@freebsd back in july.   my data would suggest
that the PR should be re-opened, because it doesn't seem fixed?


i reran a "make release" while having a "vmstat -m | grep 'FFS node'"
running every three seconds in another window.  the memory allocation
for "FFS node" is exploding during the "make release"... here is the
last bit of vmstat -m output before the kernel deadlocked:

Memory statistics by type                          Type  Kern
        Type  InUse MemUse HighUse  Limit Requests Limit Limit Size(s)
....
     FFS node199352 99676K  99677K102400K   648929    0     0  512
     FFS node199871 99936K  99936K102400K   649930    0     0  512
     FFS node200457100229K 100229K102400K   651733    0     0  512
     FFS node200794100397K 100398K102400K   653082    0     0  512
     FFS node200875100438K 100440K102400K   653902    0     0  512
     FFS node201466100733K 100733K102400K   655886    0     0  512
     FFS node202732101366K 101367K102400K   658611    0     0  512
     FFS node203300101650K 101651K102400K   660008    0     0  512
     FFS node203851101926K 101926K102400K   661718    0     0  512
     FFS node203923101962K 101962K102400K   661927    0     0  512
     FFS node204489102245K 102245K102400K   663650    0     0  512
<deadlock here!>

102245K seems like an awful lot of memory for FFS nodes?
what would cause this?   does anyone else see very large FFS node
memory allocation during "make release"?

i got a stack trace of the process stuck in "FFS node"... it is
basically:
	open
	vn_open
	namei
	lookup
	ufs_vnoperate
	vfs_cache_lookup
	ufs_vnoperate
	ufs_lookup
	ffs_vget +0x8b
	malloc + 0x1a3
	tsleep

clearly we are reaching the malloc limit for "FFS node" and hanging.


chuck

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-stable" in the body of the message