You are not logged in.
Hi,
I've got a server that has been running Arch Linux for couple of weeks without any major problems. Kernel version is 2.6.18. Unfortunately I can't get a dmesg from the server, as it's crashed at the moment...
Today the server started having weird problems: it runs fine for about 15 minutes, but then a kernel panic strikes:
Nov 20 15:57:43 server01 kernel BUG at lib/radix-tree.c:659!
Nov 20 15:57:43 server01 invalid opcode: 0000 [#1]
Nov 20 15:57:43 server01 PREEMPT SMP
Nov 20 15:57:43 server01 CPU: 0
Nov 20 15:57:43 server01 EIP is at radix_tree_gang_lookup_tag+0x1da/0x200
Nov 20 15:57:43 server01 eax: 00000000 ebx: dff5ca48 ecx: 00000018 edx: 00000010
Nov 20 15:57:43 server01 esi: ff000000 edi: 10000000 ebp: 00000005 esp: dfea7dd4
Nov 20 15:57:43 server01 ds: 007b es: 007b ss: 0068
Nov 20 15:57:43 server01 Process pdflush (pid: 193, ti=dfea6000 task=dfe75ab0 task.ti=dfea6000)
Nov 20 15:57:43 server01 Stack: 01000000 00000000 00000100 dfea7e88 c1b4bcd4 3fffffff 00000000 dff5ca48
Nov 20 15:57:43 server01 00000018 dfea7e88 0000000e c1b4bcd4 00000000 dfea7e88 c1b4bce0 c014d04a
Nov 20 15:57:43 server01 0000000e 00000000 dfea7ec8 dfea7ec8 dfea7e80 dfea7f70 c1b4bcd0 c015459a
Nov 20 15:57:43 server01 Call Trace:
Nov 20 15:57:43 server01 Code: c0 74 ef 8b 6c 24 1c 8b 44 8d 04 85 c0 74 29 89 03 46 83 c3 04 39 74 24 28 75 d7 e9 b4 fe ff ff 31 f6 89 74 24 28 e9 a9 fe ff ff <0f> 0b 93 02 1f a1 35 c0 e9 58 ff ff ff 0f 0b a4 02 1f a1 35 c0 Nov 20 15:57:43 server01 EIP: [<c022158a>] radix_tree_gang_lookup_tag+0x1da/0x200 SS:ESP 0068:dfea7dd4
Is anybody else experiencing the same problem? Any idea what could be causing it?
I don't have any cron jobs or similar running at 15 minute interval, so my only guess is an incoming e-mail message with contents that cause problems for the radix-tree code. This is just a wild guess though...
I noticed there's a newer version of the kernel available at the pacman repository, but I couldn't get it installed as the server crashed again before it was downloaded. (Naturally, the server isn't anywhere near me, and kernal.panic=1 doesn't seem to do nothing: it just hangs and won't reboot automatically...)
All tips of how to start debugging this problem would be warmly welcome.
Thanks!
- Jyri
Offline