Date: Mon, 3 Dec 2001 10:43:35 +0000 From: Josh Paetzel <friar_josh@webwarrior.net> To: Tim <timmcm@optushome.com.au> Cc: freebsd-questions@FreeBSD.ORG Subject: Re: strange kernel panics under high disk i/o - ffs_blkfree Message-ID: <20011203104335.A446@twincat.vladsempire.net> In-Reply-To: <EFEMJNJMJJDEJPHJFNAAMEDMGHAA.timmcm@optushome.com.au>; from timmcm@optushome.com.au on Mon, Dec 03, 2001 at 08:59:59PM %2B1100 References: <EFEMJNJMJJDEJPHJFNAAMEDMGHAA.timmcm@optushome.com.au>
next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, Dec 03, 2001 at 08:59:59PM +1100, Tim wrote: > Hi All, > > Im wondering whether any one has got a solution to this problem > I have found plenty of info about it in google groups etc.. but cant seem to > find a solid fix > > We are running FreeBSD 4.3 with the following hardware: > > Gigabyte GA-6EXDR MB (with on board LSI SCSI card disabled in the bios for > FreeBSD incompatability problems) > Adaptec 29160 SCSI card > 3 X Seagate 73 GIG SCSI ST373405LC HDD > All in a 1ru chasis > > > and get these errors > > dev = #da/0x20004, block = 43584, fs = /disk/00 > panic: ffs_blkfree: freeing free block > mp_lock = 00000001; cpuid = 0; lapic.id = 00000000 > boot() called on cpu#0 > > syncing disks... 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 > giving up on 3 buffers > Uptime: 25d5h24m50s > > > We updated to the latest mother board bios , and the problem is still as > often , but now its a new error > > This is under exactly the same circumsances > > > /kernel: panic: ffs_clusteralloc: map mismatch > /kernel: mp_lock = 00000001; cpuid = 0; lapic.id = 00000000 > /kernel: boot() called on cpu#0 > /kernel: > /kernel: syncing disks... 155 155 74 39 35 35 35 35 35 35 35 35 35 35 35 35 > 35 35 35 35 > /kernel: giving up on 25 buffers > /kernel: Uptime: 2h30m20s > /kernel: Automatic reboot in 15 seconds - press a key on the console to > abort > > This is fairly random (it could be up for 2 hours or 2 days) > It just doesnt happen on one machine it happens on up to 10 seperate we > have (all the same hardware) > But happens always with high SCSI I/O and writing big files > We can simulate it fairly easily testing with Bonnie writing 1-2 gig files > I have tried FreeBSD 4.4 with no difference plus different revisions of > adaptec 29160 card > Also with soft updates on or off > > Any ideas > Please email me if you want further info > > Thanks > Tim Just a hunch here, but I bet it's related to the SMP. If you running these with dual processors and SMP kernels, try doing sysctl machdep.smp_active=0 and run bonnie. If it still panics try booting a generic kernel and testing, and if it still panics, remove a processor, boot generic, and test. My guess is that somewhere down the line in this process the panics will go away. At that point take the info you have collected and do a send-pr. You have multiple machines with reproducable errors, it should be tracked down quickly. You might as well build a kernel with debug symbols in it as well, whoever gets to handle this will probably want a stack trace and kernel dump. Josh To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20011203104335.A446>