Ext3 users - Jun 2002 - ext3+raid 1: Assertion failure in journal_commit_transaction()

We're getting the below errors about once a day on a system we're trying
to
set up with RedHat 7.3. This has happened to multiple filesystems on
multiple physical and logical disks (basically we've got 4 drives as 2 sets
of RAID 1 arrays, details below).

Until a week ago, this box was a high-volume IMAP server running RedHat 6.2
with uptimes in the 200-day range, so I don't believe it's a hardware
problem.

Has anyone seen this? If someone will give us some pointers, we are willing
to put the time into a debugging-enabled kernel.

I see various complaints about ext3+software RAID 1 in the archives, but
they all seem to come down to older versions or hardware (mostly RAM)
problems.

Assertion failure in journal_commit_transaction() at commit.c:535:\
 "buffer_jdirty(bh)"
------------[ cut here ]------------
kernel BUG at commit.c:535!
invalid operand: 0000
autofs eepro100 ext3 jbd raid1 aic7xxx sd_mod scsi_mod
CPU:    0
EIP:    0010:[<f885b0e4>]    Not tainted
EFLAGS: 00010286

EIP is at journal_commit_transaction [jbd] 0xb04 (2.4.18-3smp)
eax: 0000001c   ebx: 0000000a   ecx: c02eee60   edx: 00004cb7
esi: dd966910   edi: efe712c0   ebp: f6a64000   esp: f6a65e78
ds: 0018   es: 0018   ss: 0018
Process kjournald (pid: 165, stackpage=f6a65000)
Stack: f8861eee 00000217 f776f2b8 00000000 00000fcc f3aea034 00000000 f7685cc0
       f3632850 00001e1e 00000001 f882ad2f f776f200 00000008 efadbf60 f1a00300
       f1a00ea0 f131dd20 eefe10c0 eefe1de0 eefe1f00 f17ca360 f115f200 eefe18a0
Call Trace: [<f8861eee>] .rodata.str1.1 [jbd] 0x26e
[<f882ad2f>] rw_intr [sd_mod] 0x20f
[<c0124eb5>] update_process_times [kernel] 0x25
[<c0116049>] smp_apic_timer_interrupt [kernel] 0xa9
[<c010a77f>] do_IRQ [kernel] 0xdf
[<c0119048>] schedule [kernel] 0x348
[<f885d7d6>] kjournald [jbd] 0x136
[<f885d680>] commit_timeout [jbd] 0x0
[<c0107286>] kernel_thread [kernel] 0x26
[<f885d6a0>] kjournald [jbd] 0x0

Code: 0f 0b 5a 59 6a 04 8b 44 24 18 50 56 e8 4b f1 ff ff 8d 47 48

[root@arachne root]# rpm -q redhat-release
redhat-release-7.3-1
[root@arachne root]# uname -a
Linux arachne.unet.brandeis.edu 2.4.18-3smp #1 SMP Thu Apr 18 07:27:31 EDT 2002
i686 unknown
[root@arachne root]# fdisk -l

Disk /dev/sda: 255 heads, 63 sectors, 2213 cylinders
Units = cylinders of 16065 * 512 bytes

   Device Boot    Start       End    Blocks   Id  System
/dev/sda1   *         1        64    514048+  fd  Linux raid autodetect
/dev/sda2            65       128    514080   fd  Linux raid autodetect
/dev/sda4           129      2213  16747762+   5  Extended
/dev/sda5           129       141    104391   82  Linux swap
/dev/sda6           142       294   1228941   fd  Linux raid autodetect
/dev/sda7           295       549   2048256   fd  Linux raid autodetect
/dev/sda8           550      2213  13366048+  fd  Linux raid autodetect

Disk /dev/sdb: 255 heads, 63 sectors, 2213 cylinders
Units = cylinders of 16065 * 512 bytes

   Device Boot    Start       End    Blocks   Id  System
/dev/sdb1   *         1        64    514048+  fd  Linux raid autodetect
/dev/sdb2            65       128    514080   fd  Linux raid autodetect
/dev/sdb4           129      2213  16747762+   5  Extended
/dev/sdb5           129       141    104391   82  Linux swap
/dev/sdb6           142       294   1228941   fd  Linux raid autodetect
/dev/sdb7           295       549   2048256   fd  Linux raid autodetect
/dev/sdb8           550      2213  13366048+  fd  Linux raid autodetect

Disk /dev/sdc: 255 heads, 63 sectors, 2213 cylinders
Units = cylinders of 16065 * 512 bytes

   Device Boot    Start       End    Blocks   Id  System
/dev/sdc1   *         1      2213  17775891   fd  Linux raid autodetect

Disk /dev/sdd: 255 heads, 63 sectors, 2213 cylinders
Units = cylinders of 16065 * 512 bytes

   Device Boot    Start       End    Blocks   Id  System
/dev/sdd1   *         1      2213  17775891   fd  Linux raid autodetect
[root@arachne root]# df -k
Filesystem           1k-blocks      Used Available Use% Mounted on
/dev/md1                497765    138991    333075  30% /
none                    515528         0    515528   0% /dev/shm
/dev/md3                497765      8239    463827   2% /tmp
/dev/md4              17496604   5387848  11219968  33% /u1
/dev/md0              13156160   6562828   5925036  53% /u1/www/movies
/dev/md5               1209496    795476    352580  70% /usr
/dev/md2               2015952    106552   1806992   6% /var

Hi,

On Wed, Jun 12, 2002 at 01:32:15PM -0400, Rich Graves wrote:
> We're getting the below errors about once a day on a system we're
trying to
> set up with RedHat 7.3. This has happened to multiple filesystems on
> multiple physical and logical disks (basically we've got 4 drives as 2
sets
> of RAID 1 arrays, details below).
 
> I see various complaints about ext3+software RAID 1 in the archives, but
> they all seem to come down to older versions or hardware (mostly RAM)
> problems.
There are plenty of references to this problem in the archives, too:
it was fixed a few weeks ago.  Fortunately there's an errata out
already:
> Assertion failure in journal_commit_transaction() at commit.c:535:\
>  "buffer_jdirty(bh)"
> Linux arachne.unet.brandeis.edu 2.4.18-3smp #1 SMP Thu Apr 18 07:27:31 EDT
2002 i686 unknown
2.4.18-3smp has the fault.  It's an SMP-only problem which has been
there for ages but is only just showing up recently; I suspect that's
due to timing changes elsewhere.  The Red Hat 2.4.18-4smp errata fixes
it.

Cheers,
 Stephen