From owner-freebsd-current@FreeBSD.ORG Sun May 4 19:58:15 2003 Return-Path: Delivered-To: freebsd-current@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id AF69C37B401; Sun, 4 May 2003 19:58:15 -0700 (PDT) Received: from pcat.heimat.gr.jp (catv-129-243.tees.ne.jp [202.216.129.243]) by mx1.FreeBSD.org (Postfix) with ESMTP id 51A5E43F75; Sun, 4 May 2003 19:58:14 -0700 (PDT) (envelope-from nakaji@tutrp.tut.ac.jp) Received: from xa12.heimat.gr.jp (xa12.heimat.gr.jp [202.216.136.35]) by pcat.heimat.gr.jp (8.12.9/8.12.9) with ESMTP id h452wAtV098901; Mon, 5 May 2003 11:58:10 +0900 (JST) (envelope-from nakaji@tutrp.tut.ac.jp) Sender: nakaji@xa12.heimat.gr.jp To: Robert Watson References: <867k96ofmh.fsf@xa12.heimat.gr.jp> From: NAKAJI Hiroyuki MIME-Version: 1.0 (generated by SEMI 1.14.5 - "Awara-Onsen") Content-Type: text/plain; charset=US-ASCII Date: Mon, 05 May 2003 11:58:10 +0900 In-Reply-To: (Robert Watson's message of "Sun, 4 May 2003 12:27:35 -0400 (EDT)") Message-ID: <864r4aje0t.fsf@xa12.heimat.gr.jp> User-Agent: T-gnus/6.16.1 (based on Gnus v5.10.1) (revision 01) SEMI/1.14.5 (Awara-Onsen) FLIM/1.14.5 (Demachiyanagi) APEL/10.4 Emacs/21.3 (i386--freebsd) MULE/5.0 (SAKAKI) X-Virus-Scanned: by AMaViS-perl11-milter (http://amavis.org/) cc: freebsd-current@freebsd.org Subject: Re: panic: ffs_blkfree: freeing free frag X-BeenThere: freebsd-current@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Discussions about the use of FreeBSD-current List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 05 May 2003 02:58:16 -0000 >>>>> In >>>>> Robert Watson wrote: > Could you boot to single user mode, manually fsck all of your file > systems, and then try running your various load tests again? Perhaps a > free list problem persisted on disk after the fix left over from earlier > operation with the bug... The problem still exists unfortunately. Now I switched my kernel to the one of Apr 12 2003. I rebooted my pc98 box to single user mode, ran "fsck" (I have to do this again and again because of panic) and exitted to multi-user mode. At this time no panic happened. Then, I ran "make update buildworld" in /usr/src (/dev/da2s1g) in kterm on X through the night. No panic. After I wake up, I checked my new arriving emails with Emacs and Gnus. Retrieving emails did not cause panic. But while reading emails, because email backend of Gnus is "nnml" like MH, accessing ~/Mail on /dev/da2s1h occured and got panic. And next "make update" in /usr/src was running in background. I cannot beleive these two are the "heavy disk access" but it may be so. Backtrace from vmcore is almost same as I posted before in this thread. -- NAKAJI Hiroyuki