* Oops during mkfs, SMP Opteron, 2.6.8.1 / -mm4
@ 2004-08-23 18:12 Christopher Swingley
2004-08-23 22:54 ` Christopher Swingley
0 siblings, 1 reply; 3+ messages in thread
From: Christopher Swingley @ 2004-08-23 18:12 UTC (permalink / raw)
To: linux-kernel
Greetings!
I'm getting a kernel oops (manually typed in at the bottom) when I try
to make an ext3 (or ext2) filesystem on a large (999 GB) RAID array. I
get the same oops whether the disks are set up as a hardware RAID (3Ware
7000 series) or software RAID-5. The oops is 100% repeatable, and
always happens at the same place during mkfs.
The system is a dual Opteron system with an MSI mainboard. The drives
are SATA drives plugged into a 3Ware 7000 series RAID controller.
I get the oops with both 2.6.8.1 and 2.6.8.1-mm4.
If more information would be helpful, just let me know!
Here's the manually typed in oops:
Unable to handle kernel paging request at 0000000000200e6f RIP:
<ffffffff80157081>{kmem_getpages+129}
PML4 1fdcd9067 PGD 1fe351067 PMD 0
Oops: 0000 [1] SMP
CPU 1
Modules linked in: 8139too mii crc32 tg3
Pid: 1719, comm: mkfs.ext3 Not tainted 2.6.8.1
RIP: 0010:[<ffffffff80157081>] <ffffffff80157081>{kmem_getpages+129}
RSP: 0018:00000101fe38bb88 EFLAGS: 00010213
RAX: ffffffff7fffffff RXB: 00000101fffc9680 RCX: 00000000001fffff
RDX: 0000010000011400 RSI: 0000010000011740 RDI: 0000010000011c00
RBP: 00000101fffc9680 R08: 000001016bc1e000 R09: 000001016c331310
R10: 00000101fffc96a8 R11: 00000101fffc96b8 R12: 0000000000000050
R13: 0000000000000000 R14: 0000000000000001 R15: 0000000000000010
FS: 0000002a95bcf380(0000) GS:ffffffff8050f100(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b
CR2: 0000000000200e6f CR3: 00000000dfeca000 CR4: 00000000000006e0
Process mkfs.ext3 (pid: 1719, threadinfo 00000101fe38a000, task 00000100099229f0)
Stack: 000001016c3f4000 0000000000000000 0000000000000050 ffffffff80157eb0
00000050fe38bda1 00000101fffc9680 00000101fffc9698 0000000000000050
0000010005f92658 0000000000000001
Call Trace:<ffffffff80157eb0>{cache_grow+176} <ffffffff801580cf>{cache_alloc_refill+383}
<ffffffff80158396>{kmem_cache_alloc+54} <ffffffff80175161>{alloc_buffer_head+17}
<ffffffff8017290a>{create_buffers+42} <ffffffff801732a6>{create_empty_buffers+22}
<ffffffff8017370f>{__block_prepare_write+175} <ffffffff80177b10>{blkdev_get_block+0}
<ffffffff801538e2>{__alloc_pages+818} <ffffffff8017419a>{block_prepare_write+26}
<ffffffff80151593>{generic_file_aio_write_nolock+1315}
<ffffffff80263617>{write_chan+311} <ffffffff8026370c>{write_chan+556}
<ffffffff80151a37>{generic_file_write_nolock+103} <ffffffff8012f860>{default_wake_function+0}
<ffffffff802634e0>{write_chan+0} <ffffffff80178b5a>{blkdev_file_write+26}
<ffffffff80170497>{vfs_write+199} <ffffffff801705c3>{sys_write+83}
<ffffffff8010ed4a>{system_call+126}
Code: 48 8b 91 70 0e 00 00 76 07 b8 00 00 00 80 eb 0a 48 b8 00 00
RIP <ffffffff80157081>{kmem_getpages+129} RSP <00000101fe38bb88>
CR2: 0000000000200e6f
Thanks!
Chris
--
Christopher S. Swingley email: cswingle@iarc.uaf.edu (work)
Intl. Arctic Research Center cswingle@gmail.com (personal)
University of Alaska Fairbanks www.frontier.iarc.uaf.edu/~cswingle/
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: Oops during mkfs, SMP Opteron, 2.6.8.1 / -mm4
2004-08-23 18:12 Oops during mkfs, SMP Opteron, 2.6.8.1 / -mm4 Christopher Swingley
@ 2004-08-23 22:54 ` Christopher Swingley
2004-08-23 23:13 ` Lee Revell
0 siblings, 1 reply; 3+ messages in thread
From: Christopher Swingley @ 2004-08-23 22:54 UTC (permalink / raw)
To: linux-kernel
Hello again,
* Christopher Swingley <cswingle@iarc.uaf.edu> [2004-Aug-23 10:12 AKDT]:
> Unable to handle kernel paging request at 0000000000200e6f RIP:
> <ffffffff80157081>{kmem_getpages+129}
> PML4 1fdcd9067 PGD 1fe351067 PMD 0
> Oops: 0000 [1] SMP
> CPU 1
> Modules linked in: 8139too mii crc32 tg3
> Pid: 1719, comm: mkfs.ext3 Not tainted 2.6.8.1
I know it's bad form to reply to one's own posting, but I suspect that
more information is probably useful here.
Since posting this, I've tried a single processor version of the 2.6.8.1
kernel (even though it's a dual Opteron) and I didn't get the oops.
I've also tried an SMP kernel without CONFIG_K8_NUMA, and the oops also
disappears.
So the issue appears to be related to the K8_NUMA stuff, and I'll
continue on with a non-NUMA SMP 2.6.8.1 and see if the system stays up
and running.
Thanks,
Chris
--
Christopher S. Swingley email: cswingle@iarc.uaf.edu (work)
Intl. Arctic Research Center cswingle@gmail.com (personal)
University of Alaska Fairbanks www.frontier.iarc.uaf.edu/~cswingle/
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: Oops during mkfs, SMP Opteron, 2.6.8.1 / -mm4
2004-08-23 22:54 ` Christopher Swingley
@ 2004-08-23 23:13 ` Lee Revell
0 siblings, 0 replies; 3+ messages in thread
From: Lee Revell @ 2004-08-23 23:13 UTC (permalink / raw)
To: Christopher Swingley; +Cc: linux-kernel
On Mon, 2004-08-23 at 18:54, Christopher Swingley wrote:
> I know it's bad form to reply to one's own posting, but I suspect that
> more information is probably useful here.
>
Not on LKML it's not. If you report a problem, then discover more
information about it, this is the recommended way. What is considered
bad form is starting a new thread in this situation.
Lee
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2004-08-23 23:17 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2004-08-23 18:12 Oops during mkfs, SMP Opteron, 2.6.8.1 / -mm4 Christopher Swingley
2004-08-23 22:54 ` Christopher Swingley
2004-08-23 23:13 ` Lee Revell
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox