From owner-freebsd-hackers@FreeBSD.ORG Wed Apr 28 14:40:46 2004 Return-Path: Delivered-To: freebsd-hackers@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 4116E16A4CE for ; Wed, 28 Apr 2004 14:40:46 -0700 (PDT) Received: from alpha.xerox.com (alpha.Xerox.COM [13.1.64.93]) by mx1.FreeBSD.org (Postfix) with SMTP id 312DB43D49 for ; Wed, 28 Apr 2004 14:40:46 -0700 (PDT) (envelope-from juhlig@parc.com) Received: from parc.com ([13.1.103.2]) by alpha.xerox.com with SMTP id <512525(1)>; Wed, 28 Apr 2004 14:40:38 PDT Message-ID: <409024D6.7090800@parc.com> Date: Wed, 28 Apr 2004 14:40:38 PDT From: John Uhlig User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.2.1) Gecko/20030225 X-Accept-Language: en-us, en MIME-Version: 1.0 To: hackers@freebsd.org Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit cc: sunsa@parc.com Subject: kmem_malloc crashes running FreeBSD 5.2.1-RELEASE-p5 X-BeenThere: freebsd-hackers@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Technical Discussions relating to FreeBSD List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 28 Apr 2004 21:40:46 -0000 Hi. We are running FreeBSD 5.2.1-RELEASE-p5 with a fairly generic kernel on a Dell PowerEdge 2650 with 1/2TB of disk RAID arrays. It recently crashed at about 3AM while running the daily /etc/periodic/security/100.chksetuid script. I am able to crash this system at will (takes about 5-10 minutes by running the security scripts in a loop and using gtar to copy files to the server at the same time. I have included output from one core dump below. The other core dumps I have so far are almost exactly the same e.g. "Error accessing memory address 0xf8b7983c" is the same in all core dumps so far. On the web, freebsd mailing lists and bug lists, I have seen existing thread about "kmem_malloc and kmem_map too small" problems - but no evidence of a fix or patch. We experienced the same problem running 4.9 but were able to fix it by setting the MAXMEM option in our kernel conf file to a value 1GB less than actual physical memory size. This does not help with 5.2.1-RELEASE-p5. My group is under pressure to bring 3 of these servers on-line for our user community by the end of the month and would greatly appreciate any help in resolving this problem. thanks, John Uhlig. (coredump output)--------------------------------------------------------------------------- lifeboat# gdb -k kernel.debug /crash/vmcore.7 GNU gdb 5.2.1 (FreeBSD) Copyright 2002 Free Software Foundation, Inc. GDB is free software, covered by the GNU General Public License, and you are welcome to change it and/or distribute copies of it under certain conditions Type "show copying" to see the conditions. There is absolutely no warranty for GDB. Type "show warranty" for details. This GDB was configured as "i386-unknown-freebsd"... panic: kmem_malloc(4096): kmem_map too small: 275251200 total allocated panic messages: --- dmesg: kvm_read: --- warning: failed to read linker file data at 0xca29b100 #0 doadump () at /usr/src/sys/kern/kern_shutdown.c:240 240 dumping++; (kgdb) backtrace #0 doadump () at /usr/src/sys/kern/kern_shutdown.c:240 can not access 0xf8b7983c, invalid address (f8b7983c) can not access 0xf8b7983c, invalid address (f8b7983c) Error accessing memory address 0xf8b7983c: Invalid argument. (kgdb) list *0xf8b6783c No source file for address 0xf8b6783c.