by This error message is occurring with kernel 2.4.9-31smp, particularly
on older hardware. I am seeing the following kernel panic with
regularity:
Unable to handle kernel paging request at virtual address 0a1a77da
printing eip:
0a1a77da
*pde = 00000000
Oops: 0000
Kernel 2.4.9-31smp
CPU: 0
EIP: 0010:[<0a1a77da>] Tainted: P
EFLAGS: 00010202
EIP is at Using_Versions [] 0xa1a77d9
eax: 0a1a77da ebx: d112b040 ecx: d112b050 edx: e0e1e183
esi: f74f62e0 edi: e0e1e183 ebp: fffffff6 esp: c22dbf50
ds: 0018 es: 0018 ss: 0018
Process kswapd (pid: 5, stackpage=c22db000)
Stack: c0153d4b d112b040 c211c3a0 c02de878 00000000 c22da000 f74f62f8
f74f62e0
d112b040 c01515eb d112b040 d112b040 c22da000 00000000 c1deb944
00000000
c1eb1250 c1deb944 00000000 c0135c45 00000000 00036a22 000000c0
000000c0
Call Trace: [<c0153d4b>] iput [kernel] 0x3b
[<c01515eb>] prune_dcache [kernel] 0x10b
EIP is at Using_Versions [] 0xa1a77d9
eax: 0a1a77da ebx: d112b040 ecx: d112b050 edx: e0e1e183
esi: f74f62e0 edi: e0e1e183 ebp: fffffff6 esp: c22dbf50
ds: 0018 es: 0018 ss: 0018
Process kswapd (pid: 5, stackpage=c22db000)
Stack: c0153d4b d112b040 c211c3a0 c02de878 00000000 c22da000 f74f62f8
f74f62e0
d112b040 c01515eb d112b040 d112b040 c22da000 00000000 c1deb944
00000000
c1eb1250 c1deb944 00000000 c0135c45 00000000 00036a22 000000c0
000000c0
Call Trace: [<c0153d4b>] iput [kernel] 0x3b
[<c01515eb>] prune_dcache [kernel] 0x10b
[<c0135c45>] page_launder [kernel] 0x955
[<c0151a11>] shrink_dcache_memory [kernel] 0x21
[<c013604b>] do_try_to_free_pages [kernel] 0x1b
[<c01360d5>] kswapd [kernel] 0x55
[<c0105000>] stext [kernel] 0x0
[<c0105866>] kernel_thread [kernel] 0x26
[<c0136080>] kswapd [kernel] 0x0
Code: Bad EIP value.
------------------------------------------------------------------
The Google searches I have done up until now reveal that many
people have seen this problem but nobody has found a solution as yet.
When the kernel panic happens, the machine does not crash. It stays up,
but then kswapd shows as <defunct> in the ps listing.
We have seen this problem two or three times per node on the 20
nodes in question. These nodes are doing a number of swaps usually
at the time the error happens.
Any idea what might be going on?
Steve Timm
------------------------------------------------------------------
Fermilab Computing Division/Operating Systems Support
Scientific Computing Support Group--Computing Farms Operations
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/