From owner-freebsd-current@FreeBSD.ORG Fri Jan 14 11:06:48 2005 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 8DA6716A4CE for ; Fri, 14 Jan 2005 11:06:48 +0000 (GMT) Received: from gw.catspoiler.org (217-ip-163.nccn.net [209.79.217.163]) by mx1.FreeBSD.org (Postfix) with ESMTP id 27EE143D41 for ; Fri, 14 Jan 2005 11:06:48 +0000 (GMT) (envelope-from truckman@FreeBSD.org) Received: from FreeBSD.org (mousie.catspoiler.org [192.168.101.2]) by gw.catspoiler.org (8.13.1/8.13.1) with ESMTP id j0EB6eFa032116; Fri, 14 Jan 2005 03:06:44 -0800 (PST) (envelope-from truckman@FreeBSD.org) Message-Id: <200501141106.j0EB6eFa032116@gw.catspoiler.org> Date: Fri, 14 Jan 2005 03:06:40 -0800 (PST) From: Don Lewis To: current@FreeBSD.org In-Reply-To: <200501140326.j0E3PxDf031145@gw.catspoiler.org> MIME-Version: 1.0 Content-Type: TEXT/plain; charset=us-ascii cc: mckusick@mckusick.com Subject: Re: repeatable panic - ffs_blkfree: freeing free block 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: Fri, 14 Jan 2005 11:06:48 -0000 On 13 Jan, Don Lewis wrote: > On 13 Jan, Don Lewis wrote: >> I'm getting a repeatable panic when trying to build openoffice with a >> 2005-01-12 kernel and world. I manually fsck'ed the file system >> immediately before the build. The first panic happened during the >> patching phase of the build. I tried again and this time the panic >> happened during configuration. >> >> dev = da0s2a, block = 5090392, fs = /mnt >> panic: ffs_blkfree: freeing free block >> cpuid = 0 >> KDB: enter: panic >> [thread pid 55 tid 100043 ] >> Stopped at kdb_enter+0x2c: leave >> db> tr >> Tracing pid 55 tid 100043 td 0xc22fa450 >> kdb_enter(c0848a40,100,c22fa450,1361b80,4b5) at kdb_enter+0x2c >> panic(c085f39d,c085f37d,c26cc280,4dac58,0) at panic+0x17f >> ffs_blkfree(c2863800,c2987114,4dac58,0,4000) at ffs_blkfree+0x3a0 >> indir_trunc(136b060,0,0,c,0,e4d5bc4c) at indir_trunc+0x1bf >> handle_workitem_freeblocks(0,0,0,1,0) at handle_workitem_freeblocks+0x2e6 >> process_worklist_item(41e7112d,0,0,c247d2bc,23) at process_worklist_item+0x1c0 >> softdep_process_worklist(0,29,41e7112d,1,0) at softdep_process_worklist+0x66 >> sched_sync(0,e4d5bd48,0,c0677b60,0) at sched_sync+0x3b2 >> fork_exit(c0677b60,0,e4d5bd48) at fork_exit+0x7e >> fork_trampoline() at fork_trampoline+0x8 >> --- trap 0x1, eip = 0, esp = 0xe4d5bd7c, ebp = 0 --- > It looks like the configure script has just been regenerated and a bug > is being triggered by executing the script shortly after the script file > was regenerated and before softupdates has flushed the changes to disk. It appears that executing the configure script isn't necessary. I can execute this: # make extract patch build-depends lib-depends misc-depends \ configure-message pre-configure pre-configure-script patch-autotools and things are fine, but I get the panic shortly after I run this: # make run-autotools Another interesting bit of information is that the problem goes away if I unmount and remount the file system between the patch-autotools and run-autotools steps. I was able to loop over the whole sequence from "make clean" to "make run-autotools" ten times without failure. On the other hand, doing "sync; sleep 60" instead of unmounting and remounting the file system does not seem to help and the system still panics. Doing fsync /mnt/usr/ports/editors/openoffice-1.1/work/config_office/configure doesn't solve the problem, either.