Kernel error message

Hello,
We have a new x-series 360 IBM e-server with 4 xeon processors. We got some errors that suggested we send mail to this address. I am hoping you can provide some useful feedback. Please let me know if you need more information.

~Evan Stevens
National Safty Associates

we loaded
Red Hat Linux release 7.3 (Valhalla)
Kernel 2.4.18-3smp on an i686

After several days of uptime, we received a message on all tty's:
Assertion failure in journal_commit_transaction() at commit.c:535:
"buffer_jdirty(bh)"

and this appeared on the console and messages file:

Oct 28 10:00:31 fronttest kernel: Assertion failure in
journal_commit_transaction() at commit.c:535: "buffer_jdirty(bh)"
Oct 28 10:00:31 fronttest kernel: ------------[ cut here ]------------
Oct 28 10:00:31 fronttest kernel: kernel BUG at commit.c:535!
Oct 28 10:00:31 fronttest kernel: invalid operand: 0000
Oct 28 10:00:31 fronttest kernel: autofs eepro100 st usb-ohci usbcore
ext3 jbd aic7xxx sd_mod scsi_mod  
Oct 28 10:00:31 fronttest kernel: CPU:    4
Oct 28 10:00:31 fronttest kernel: EIP:    0010:[<f88560e4>]    Not
tainted
Oct 28 10:00:31 fronttest kernel: EFLAGS: 00010286
Oct 28 10:00:31 fronttest kernel: 
Oct 28 10:00:31 fronttest kernel: EIP is at journal_commit_transaction
[jbd] 0xb04 (2.4.18-3smp)
Oct 28 10:00:31 fronttest kernel: eax: 0000001c   ebx: 0000000a   ecx:
c02eee60   edx: 000054da
Oct 28 10:00:31 fronttest kernel: esi: e4b15880   edi: f7577740   ebp:
f7586000   esp: f7587e78
Oct 28 10:00:31 fronttest kernel: ds: 0018   es: 0018   ss: 0018
Oct 28 10:00:31 fronttest kernel: Process kjournald (pid: 24,
stackpage=f7587000)
Oct 28 10:00:31 fronttest kernel: Stack: f885ceee 00000217 00000000
00000000 00000f2c e99620d4 00000000 e8dcd7e0 
Oct 28 10:00:31 fronttest kernel:        e4e81850 000008e5 37363534
42413938 46454443 4a494847 e9ae4620 e2224320 
Oct 28 10:00:31 fronttest kernel:        e2e95cc0 e2e95d80 e2e95de0
e2e95e40 e2e95ea0 e2e95f00 e2e95f60 e2fb20e0 
Oct 28 10:00:31 fronttest kernel: Call Trace: [<f885ceee>]
.rodata.str1.1 [jbd] 0x26e 
Oct 28 10:00:31 fronttest kernel: [<c0119048>] schedule [kernel] 0x348 
Oct 28 10:00:31 fronttest kernel: [<f88587d6>] kjournald [jbd] 0x136 
Oct 28 10:00:31 fronttest kernel: [<f8858680>] commit_timeout [jbd] 0x0 
Oct 28 10:00:31 fronttest kernel: [<c0107286>] kernel_thread [kernel]
0x26 
Oct 28 10:00:31 fronttest kernel: [<f88586a0>] kjournald [jbd] 0x0 
Oct 28 10:00:31 fronttest kernel: 
Oct 28 10:00:31 fronttest kernel: 
Oct 28 10:00:31 fronttest kernel: Code: 0f 0b 5a 59 6a 04 8b 44 24 18 50

56 e8 4b f1 ff ff 8d 47 48

Some dmesgs we are getting include:

-- Snip --
Total of 8 processors activated (22380.54 BogoMIPS). cpu_sibling_map[0] = 1
cpu_sibling_map[1] = 0
cpu_sibling_map[2] = 3
cpu_sibling_map[3] = 2
cpu_sibling_map[4] = 5
cpu_sibling_map[5] = 4
cpu_sibling_map[6] = 7
cpu_sibling_map[7] = 6
ENABLING IO-APIC IRQs
Setting 14 in the phys_id_present_map
...changing IO-APIC physical APIC ID to 14 ... ok.
init IO_APIC IRQs
IO-APIC (apicid-pin) 14-0, 14-3, 14-5, 14-11, 14-13, 14-15, 14-18,
14-39, 14-40, 14-46, 14-47, 14-48, 14-49, 14-50 not connected .
..TIMER: vector=0x31 pin1=2 pin2=-1
number of MP IRQ sources: 37.
number of IO-APIC #14 registers: 51.
testing the IO APIC.......................

IO APIC #14......
.... register #00: 0E000000
....... : physical APIC id: 0E
.... register #01: 00320011
....... : max redirection entries: 0032
WARNING: unexpected IO-APIC, please mail
to [email protected]
....... : PRQ implemented: 0
....... : IO APIC version: 0011
.... register #02: 00000000
....... : arbitration: 00
.... IRQ redirection table:

-- Snip --

calibrating APIC timer ...
..... CPU clock speed is 1400.2158 MHz.
..... host bus clock speed is 100.0153 MHz.
cpu: 0, clocks: 1000153, slice: 111128 CPU0<T0:1000144,T1:889008,D:8,S:111128,C:1000153>
cpu: 5, clocks: 1000153, slice: 111128
cpu: 6, clocks: 1000153, slice: 111128
cpu: 1, clocks: 1000153, slice: 111128
cpu: 7, clocks: 1000153, slice: 111128
cpu: 3, clocks: 1000153, slice: 111128
cpu: 4, clocks: 1000153, slice: 111128
cpu: 2, clocks: 1000153, slice: 111128 CPU1<T0:1000144,T1:777888,D:0,S:111128,C:1000153> CPU2<T0:1000144,T1:666752,D:8,S:111128,C:1000153> CPU3<T0:1000144,T1:555632,D:0,S:111128,C:1000153> CPU4<T0:1000144,T1:444480,D:24,S:111128,C:1000153> CPU5<T0:1000144,T1:333376,D:0,S:111128,C:1000153> CPU6<T0:1000144,T1:222240,D:8,S:111128,C:1000153> CPU7<T0:1000144,T1:111120,D:0,S:111128,C:1000153> checking TSC synchronization across CPUs:
BIOS BUG: CPU#6 improperly initialized, has 2 usecs TSC skew! FIXED.
PCI: PCI BIOS revision 2.10 entry at 0xfd31c, last bus=14
PCI: Using configuration type 1
PCI: Probing PCI hardware
PCI: Discovered peer bus 01

From the messages file:

Oct 28 13:06:10 fronttest kernel: hm, page 0009e000 reserved twice.
Oct 28 13:06:10 fronttest kernel: hm, page 0009f000 reserved twice.
Oct 28 13:06:10 fronttest kernel: WARNING: MP table in the EBDA can be
UNSAFE, contact [email protected] if you experienc
e SMP problems!
Oct 28 13:06:10 fronttest kernel: On node 0 totalpages: 258280
Oct 28 13:06:10 fronttest kernel: zone(0): 4096 pages.

O

-- Snip --

ct 28 12:55:25 fronttest kernel: cpu_sibling_map[7] = 6

Oct 28 12:55:25 fronttest kernel: ENABLING IO-APIC IRQs
Oct 28 12:55:25 fronttest kernel: Setting 14 in the phys_id_present_map
Oct 28 12:55:25 fronttest kernel: ...changing IO-APIC physical APIC ID
to 14 ... ok.
Oct 28 12:55:25 fronttest kernel: ..TIMER: vector=0x31 pin1=2 pin2=-1
Oct 28 12:55:26 fronttest kernel: testing the IO
APIC.......................
Oct 28 12:55:26 fronttest kernel: 
Oct 28 12:55:26 fronttest kernel:  WARNING: unexpected IO-APIC, please
mail
Oct 28 12:55:26 fronttest kernel:           to [email protected]
Oct 28 12:55:26 fronttest xinetd[748]: xinetd Version 2002.03.28 started
with libwrap options compiled in.
Oct 28 12:55:26 fronttest xinetd[748]: Started working: 8 available
services
Oct 28 12:55:26 fronttest kernel: ....................................
done.
Oct 28 12:55:26 fronttest kernel: Using local APIC timer interrupts.
Oct 28 12:55:26 fronttest kernel: calibrating APIC timer ...
Oct 28 12:55:26 fronttest kernel: ..... CPU clock speed is 1400.1570
MHz.
Oct 28 12:55:26 fronttest kernel: ..... host bus clock speed is 100.0110

MHz.

Please get the Red Hat errata kernel. One of the things it fixed might be causing this.