From owner-freebsd-current Thu Dec 3 23:47:57 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id XAA03812 for freebsd-current-outgoing; Thu, 3 Dec 1998 23:47:57 -0800 (PST) (envelope-from owner-freebsd-current@FreeBSD.ORG) Received: from verdi.nethelp.no (verdi.nethelp.no [158.36.41.162]) by hub.freebsd.org (8.8.8/8.8.8) with SMTP id XAA03807 for ; Thu, 3 Dec 1998 23:47:55 -0800 (PST) (envelope-from sthaug@nethelp.no) From: sthaug@nethelp.no Received: (qmail 27361 invoked by uid 1001); 4 Dec 1998 07:47:39 +0000 (GMT) To: dg@root.com Cc: cdillon@wolves.k12.mo.us, jkh@zippy.cdrom.com, sgk@troutmask.apl.washington.edu, bford@uop.cs.uop.edu, freebsd-current@FreeBSD.ORG Subject: Re: panic: ffs_blkfree: bad size In-Reply-To: Your message of "Thu, 03 Dec 1998 18:38:05 -0800" References: <199812040238.SAA27887@root.com> X-Mailer: Mew version 1.05+ on Emacs 19.34.2 Mime-Version: 1.0 Content-Type: Text/Plain; charset=us-ascii Date: Fri, 04 Dec 1998 08:47:39 +0100 Message-ID: <27359.912757659@verdi.nethelp.no> Sender: owner-freebsd-current@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG > >I posted a message a while back when this bug first started to bite. I > >have a good coredump or two (using debugging kernel and all) if anyone > >wants them. Kirk told me he would look at Greg's dump first and get > >back to me if he needed more samples, but I haven't heard from him. I > >can build a more recent world (mine is a couple weeks old now) and turn > >on vfs.ffs.doreallocblks again to get a more recent dump. It occurrs > >without having to do much work. My system has been rock solid ever > >since I set vfs.ffs.doreallocblks=0 at boot, so it isn't hardware > >instability like I thought it may have been. > > There was a bug found in the original code that has been fixed. What we're > interested in now is continued problems *after* the bugfix (about Nov 18th). As mentioned before, I get this panic on an 3.0-19981123-SNAP system (vfs_cluster.c 1.74), SMP, no softupdates, with vfs.ffs.doreallocblks=1. The panic happens during heavy disk *and* network activity. If anybody wants backtraces or wants to play with the core dumps, they can be made available (warning: the machine has 512 MB of memory, so the core dump is rather big). I've sent a message to Kirk McKusick asking if he wants any of this, but haven't received an answer. Steinar Haug, Nethelp consulting, sthaug@nethelp.no ---------------------------------------------------------------------- #0 boot (howto=256) at ../../kern/kern_shutdown.c:285 #1 0xf0134b2e in panic (fmt=0xf01c5d81 "ffs_blkfree: bad size") at ../../kern/kern_shutdown.c:446 #2 0xf01c5ede in ffs_blkfree (ip=0xf1569b00, bno=6442, size=8192) at ../../ufs/ffs/ffs_alloc.c:1300 #3 0xf01c45a4 in ffs_reallocblks (ap=0xf95ffde4) at ../../ufs/ffs/ffs_alloc.c:523 #4 0xf015216b in cluster_write (bp=0xf6496350, filesize=105562112) at vnode_if.h:1035 #5 0xf01cba67 in ffs_write (ap=0xf95ffebc) at ../../ufs/ufs/ufs_readwrite.c:368 #6 0xf015c14b in vn_write (fp=0xf15686c0, uio=0xf95fff40, cred=0xf1560280) at vnode_if.h:331 #7 0xf013d803 in writev (p=0xf95a3840, uap=0xf95fff94) at ../../kern/sys_generic.c:356 #8 0xf01f5d0b in syscall (frame={tf_es = 39, tf_ds = 39, tf_edi = 714452012, tf_esi = 2553, tf_ebp = 714452172, tf_isp = -111149084, tf_ebx = 0, tf_edx = 16, tf_ecx = 135401472, tf_eax = 121, tf_trapno = 12, tf_err = 2, tf_eip = 135040412, tf_cs = 31, tf_eflags = 514, tf_esp = 714451972, tf_ss = 39}) at ../../i386/i386/trap.c:1031 #9 0xf01e4e1c in Xint0x80_syscall () #10 0x809c311 in ?? () #11 0x809c2b6 in ?? () #12 0x807ce62 in ?? () #13 0x805cc46 in ?? () #14 0x805a882 in ?? () #15 0x8093da1 in ?? () #16 0x8093c61 in ?? () #17 0x80b063c in ?? () To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message