Subject: Help me shed some light on a kernel bug
From: Cedric Veilleux <cveilleux@neopeak.com>
Date: Mon, 02 Feb 2009 23:44:03 -0500
Mon, 02 Feb 2009 23:44:03 -0500
Hi,

I had a server running a vserver patched kernel crash repeatedly (3-4
times) in a 48 hours period.

The only error logged was the one pasted below. There was in fact
sometimes one of these logged, some times many similar ones before a
crash.

So, the log output might be misleading as it did not lead immediately to
a system crash, but everytime it happened, the system would freeze
eventually.

What is very special is that this system had 440 days of uptime before
the first crash. It was running on a Debian Backports kernel labeled
2.6.22-6~bpo40+1, with vserver-info showing:

                   Kernel: 2.6.22-3-vserver-amd64
                   VS-API: 0x00020200
             util-vserver: 0.30.216-pre2772; Nov  8 2008, 21:49:08


So, this might very well not be a vserver related bug at all, but I am
posting here since people more knowledgeable than me can point me in the
right direction.

Also, I would suspect a failing system, maybe some failing system
memory? But the wording of the error message "Kernel BUG at ..." leads
me to believe this is a software error and not hardware related...?

Finally, I since upgraded to 2.6.26-bpo.1-vserver-amd64 and now it has
been running fine for 4 days with no errors.


Let me know what you think.

Thank you

Cedric







Jan 28 19:05:19 mtl1 kernel: ------------[ cut here ]------------
Jan 28 19:05:19 mtl1 kernel: kernel BUG at mm/rmap.c:629!
Jan 28 19:05:19 mtl1 kernel: invalid opcode: 0000 [1] SMP 
Jan 28 19:05:19 mtl1 kernel: CPU 3 
Jan 28 19:05:19 mtl1 kernel: Modules linked in: generic piix ide core ata generic ata
piix libata ehci hcd scsi mod tg3 uhci hcd thermal processor fan
Jan 28 19:05:19 mtl1 kernel: Pid: 706:#0, comm: udevd Not tainted 2.6.22-3-vserver-amd64
#1
Jan 28 19:05:19 mtl1 kernel: RIP: 0010:[<ffffffff8027d383>]  [<ffffffff8027d383>] page
remove rmap+0xed/0x109
Jan 28 19:05:19 mtl1 kernel: RSP: 0018:ffff8101191cfdb8  EFLAGS: 00010246
Jan 28 19:05:19 mtl1 kernel: RAX: 0000000000000000 RBX: ffff81011fb674a0 RCX: 0000000000004e51
Jan 28 19:05:19 mtl1 kernel: RDX: 00000000ffffff01 RSI: ffff81011b4048b8 RDI: ffffffff804c3d5c
Jan 28 19:05:19 mtl1 kernel: RBP: ffff8101196b3348 R08: 000000000000000e R09: 000000007fffffff
Jan 28 19:05:19 mtl1 kernel: R10: 0000000000000000 R11: ffffffff8031e493 R12: 0000000000517000
Jan 28 19:05:19 mtl1 kernel: R13: ffff81011b4048b8 R14: ffff810119160f40 R15: 0000000000552000
Jan 28 19:05:19 mtl1 kernel: FS:  0000000000000000(0000) GS:ffff81011ba10b40(0000) knlGS:0000000000000000
Jan 28 19:05:19 mtl1 kernel: CS:  0010 DS: 0000 ES: 0000 CR0: 000000008005003b
Jan 28 19:05:19 mtl1 kernel: CR2: 00002b8cfc7fb4c8 CR3: 0000000119aef000 CR4: 00000000000006e0
Jan 28 19:05:19 mtl1 kernel: Process udevd (pid: 706[#0], threadinfo ffff8101191ce000,
task ffff81011af18140)
Jan 28 19:05:19 mtl1 kernel: Stack:  000000001ab23020 ffff81011fb674a0 000000000afcc020
ffffffff8027516e
Jan 28 19:05:19 mtl1 kernel:  0000000000000000 ffff8101191cfea8 ffffffffffffffff 0000000000000000
Jan 28 19:05:19 mtl1 kernel:  ffff8101196b3348 ffff8101191cfeb0 00000000003e9ffd 0000000000000000
Jan 28 19:05:19 mtl1 kernel: Call Trace:
Jan 28 19:05:19 mtl1 kernel:  [<ffffffff8027516e>] unmap vmas+0x3f9/0x758
Jan 28 19:05:19 mtl1 kernel:  [<ffffffff8027ad26>] exit mmap+0x76/0x19c
Jan 28 19:05:19 mtl1 kernel:  [<ffffffff8022fe34>] mmput+0x2c/0x9d
Jan 28 19:05:19 mtl1 kernel:  [<ffffffff802356e4>] do exit+0x222/0x947
Jan 28 19:05:19 mtl1 kernel:  [<ffffffff80235e88>] sys exit group+0x0/0xe
Jan 28 19:05:19 mtl1 kernel:  [<ffffffff80209d9e>] system call+0x7e/0x83
Jan 28 19:05:19 mtl1 kernel: 
Jan 28 19:05:19 mtl1 kernel: 
Jan 28 19:05:19 mtl1 kernel: Code: 0f 0b eb fe 8b 77 18 41 58 5b 5d 83 e6 01 f7 de 83
c6 04 e9 
Jan 28 19:05:19 mtl1 kernel: RIP  [<ffffffff8027d383>] page remove rmap+0xed/0x109
Jan 28 19:05:19 mtl1 kernel:  RSP <ffff8101191cfdb8>
Jan 28 19:05:19 mtl1 kernel: Fixing recursive fault but reboot is needed!







-  
Cédric Veilleux
Solutions Internet Neopeak inc.
http://www.neopeak.com/
514-906-5847   #252
1-866-760-1350 #252



["application/x-pkcs7-signature" not shown]