2011-11-10 13:30:53

by Andrew Watts

[permalink] [raw]
Subject: [OOPS]: Kernel 3.1 (ext3?)

Hi.

I had the following kernel panic today on 3.1 (machine was compiling code
unattended). It would appear to be a bug/regression introduced sometime
between 2.6.39.4 and 3.1.

Help would be much appreciated.

~ Andy

PS cross-posting to linux-ext4@ given the stack trace.

================


EIP: 0060:[<c114cd83>] EFLAGS: 00210082 CPU: 0
EIP is at init_once+0x13/0x50
EAX: f5406068 EBX: f5406000 ECX: c1962c63 EDX: c1781109
ESI: f5406000 EDI: f5406000 EBP: f0d2dc6c ESP: f0d2dc68
DS: 007b ES: 007b FS: 0000 GS: 00e0 SS: 0068
Process ar (pid 17281, ti=f0d2c000 task=f3060c50 tast.ti=f0d2c000)
Stack:
f5c5aa00 f0d2dc7c c10d953b f5c5aa00 f6fa50c0 f0d2dca0 c10d97d6 00000000
f5406000 f5406000 f5406000 00000000 00000000 f5c5aa00 f0d2dd34 c10dadf1
f5803200 f1f5b2a0 f58b3030 f0d2dcd4 00000000 f0d2dccc 00000000 c114cd22
Call Trace:
[<c10d953b>] setup_object+0x4b/0x60
[<c10d97d6>] new_slab+0x126/0x1d0
[<c10dadf1>] T.1022+0x131/0x340
[<c114cd22>] ? ext3_alloc_inode+0x12/0x60
[<c1141547>] ? ext3_getblk+0xd7/0x1d0
[<c1051827>] ? wake_up_bit+0x57/0x60
[<c114cd22>] ? ext3_alloc_inode+0x12/0x60
[<c11091a0>] ? unlock_buffer+0x10/0x20
[<c10db10a>] kmem_cache_alloc+0x10a/0x120
[<c114cd22>] ? ext3_alloc_inode+0x12/0x60
[<c114634e>] ? ext3_find_entry+0x3fe/0x5d0
[<c114cd22>] ext3_alloc_inode+0x12/0x60
[<c10f7ccc>] alloc_inode+0x1c/0x80
[<c10f7d38>] new_inode_pseudo+0x8/0x30
[<c10f7d72>] new_inode+0x12/0x40
[<c113da8c>] ext3_new_inode+0x4c/0x900
[<c1161db2>] ? journal_start+0x62/0xd0
[<c1161de8>] ? journal_start+0x90/0xd0
[<c114dac9>] ? ext3_journal_start_sb+0x29/0x50
[<c114588c>] ext3_create+0x7c/0xe0
[<c10edda3>] vfs_create+0x93/0xb0
[<c10ee7fc>] do_last+0x38c/0x760
[<c10ef6da>] path_openat+0x9a/0x340
[<c10efa60>] do_filp_open+0x30/0x80
[<c10db095>] ? kmem_cache_alloc+0x95/0x120
[<c10ed088>] ? getname_flags+0x28/0x110
[<c10f98d2>] ? alloc_fd+0x62/0xe0
[<c10ed11b>] ? getname_flags+0xbb/0x110
[<c10e367d>] do_sys_open+0xed/0x1e0
[<c10e46c5>] ? vfs_read+0xf5/0x160
[<c10e37d9>] sys_open+0x29/0x40
[<c159f713>] sysenter_do_call+0x12/0x22
Code: 00 00 00 00 00 89 d0 5d c3 eb 0d 90 90 90 90 90 90 90 90 90
90 90 90 90 55 b9 63 2c 96 c1 89 e5 ba 09 11 78 c1 53 89 c3 8d 40
68 43 68 89 43 6c 8d 43 5c e8 cf 55 09 00 8d 43 7c b9 63 2c 96
EIP: [<c114cd83>] init_once+0x13/0x50 SS:ESP 0068:f0d2dc68
CR2: 00000000f5406068



2011-11-14 19:54:14

by Jan Kara

[permalink] [raw]
Subject: Re: [OOPS]: Kernel 3.1 (ext3?)

Hi,

On Thu 10-11-11 08:29:37, Andrew Watts wrote:
> I had the following kernel panic today on 3.1 (machine was compiling code
> unattended). It would appear to be a bug/regression introduced sometime
> between 2.6.39.4 and 3.1.
Hmm, the report is missing a line (top one) saying why the kernel
actually crashed. Can you add that?

Also it seems you are using SLUB allocator, right? This seems like a
problem there so adding some CCs.

Honza

> ================
>
>
> EIP: 0060:[<c114cd83>] EFLAGS: 00210082 CPU: 0
> EIP is at init_once+0x13/0x50
> EAX: f5406068 EBX: f5406000 ECX: c1962c63 EDX: c1781109
> ESI: f5406000 EDI: f5406000 EBP: f0d2dc6c ESP: f0d2dc68
> DS: 007b ES: 007b FS: 0000 GS: 00e0 SS: 0068
> Process ar (pid 17281, ti=f0d2c000 task=f3060c50 tast.ti=f0d2c000)
> Stack:
> f5c5aa00 f0d2dc7c c10d953b f5c5aa00 f6fa50c0 f0d2dca0 c10d97d6 00000000
> f5406000 f5406000 f5406000 00000000 00000000 f5c5aa00 f0d2dd34 c10dadf1
> f5803200 f1f5b2a0 f58b3030 f0d2dcd4 00000000 f0d2dccc 00000000 c114cd22
> Call Trace:
> [<c10d953b>] setup_object+0x4b/0x60
> [<c10d97d6>] new_slab+0x126/0x1d0
> [<c10dadf1>] T.1022+0x131/0x340
> [<c114cd22>] ? ext3_alloc_inode+0x12/0x60
> [<c1141547>] ? ext3_getblk+0xd7/0x1d0
> [<c1051827>] ? wake_up_bit+0x57/0x60
> [<c114cd22>] ? ext3_alloc_inode+0x12/0x60
> [<c11091a0>] ? unlock_buffer+0x10/0x20
> [<c10db10a>] kmem_cache_alloc+0x10a/0x120
> [<c114cd22>] ? ext3_alloc_inode+0x12/0x60
> [<c114634e>] ? ext3_find_entry+0x3fe/0x5d0
> [<c114cd22>] ext3_alloc_inode+0x12/0x60
> [<c10f7ccc>] alloc_inode+0x1c/0x80
> [<c10f7d38>] new_inode_pseudo+0x8/0x30
> [<c10f7d72>] new_inode+0x12/0x40
> [<c113da8c>] ext3_new_inode+0x4c/0x900
> [<c1161db2>] ? journal_start+0x62/0xd0
> [<c1161de8>] ? journal_start+0x90/0xd0
> [<c114dac9>] ? ext3_journal_start_sb+0x29/0x50
> [<c114588c>] ext3_create+0x7c/0xe0
> [<c10edda3>] vfs_create+0x93/0xb0
> [<c10ee7fc>] do_last+0x38c/0x760
> [<c10ef6da>] path_openat+0x9a/0x340
> [<c10efa60>] do_filp_open+0x30/0x80
> [<c10db095>] ? kmem_cache_alloc+0x95/0x120
> [<c10ed088>] ? getname_flags+0x28/0x110
> [<c10f98d2>] ? alloc_fd+0x62/0xe0
> [<c10ed11b>] ? getname_flags+0xbb/0x110
> [<c10e367d>] do_sys_open+0xed/0x1e0
> [<c10e46c5>] ? vfs_read+0xf5/0x160
> [<c10e37d9>] sys_open+0x29/0x40
> [<c159f713>] sysenter_do_call+0x12/0x22
> Code: 00 00 00 00 00 89 d0 5d c3 eb 0d 90 90 90 90 90 90 90 90 90
> 90 90 90 90 55 b9 63 2c 96 c1 89 e5 ba 09 11 78 c1 53 89 c3 8d 40
> 68 43 68 89 43 6c 8d 43 5c e8 cf 55 09 00 8d 43 7c b9 63 2c 96
> EIP: [<c114cd83>] init_once+0x13/0x50 SS:ESP 0068:f0d2dc68
> CR2: 00000000f5406068
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
--
Jan Kara <[email protected]>
SUSE Labs, CR

2011-11-14 20:55:20

by Andrew Watts

[permalink] [raw]
Subject: Re: [OOPS]: Kernel 3.1 (ext3?)

On Mon, Nov 14, 2011 at 08:53:52PM +0100, Jan Kara wrote:
> Hmm, the report is missing a line (top one) saying why the kernel
> actually crashed. Can you add that?
>
> Also it seems you are using SLUB allocator, right? This seems like a
> problem there so adding some CCs.

Hi Jan.

Unfortunately, the oops trace is a transcription of all that was left on
my screen. I hope that can provide enough clues into what is going on.

Yes, I use the SLUB allocator.

Thank you for your reply and for adding revelant cc's.

~ Andy


Subject: Re: [OOPS]: Kernel 3.1 (ext3?)

On Mon, 14 Nov 2011, Jan Kara wrote:

> On Thu 10-11-11 08:29:37, Andrew Watts wrote:
> > I had the following kernel panic today on 3.1 (machine was compiling code
> > unattended). It would appear to be a bug/regression introduced sometime
> > between 2.6.39.4 and 3.1.
> Hmm, the report is missing a line (top one) saying why the kernel
> actually crashed. Can you add that?
>
> Also it seems you are using SLUB allocator, right? This seems like a
> problem there so adding some CCs.

Likely some data corruption. Enable slub debugging by passing

slub_debug

on the kernel commandline please to get some information as to where and
when this happens.

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to [email protected]. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"[email protected]"> [email protected] </a>

2011-11-22 11:42:49

by Andrew Watts

[permalink] [raw]
Subject: Re: [OOPS]: Kernel 3.1 (ext3?)

On Tue, Nov 15, 2011 at 10:05:25AM -0600, Christoph Lameter wrote:
> On Mon, 14 Nov 2011, Jan Kara wrote:
>
> > On Thu 10-11-11 08:29:37, Andrew Watts wrote:
> > > I had the following kernel panic today on 3.1 (machine was compiling code
> > > unattended). It would appear to be a bug/regression introduced sometime
> > > between 2.6.39.4 and 3.1.
> > Hmm, the report is missing a line (top one) saying why the kernel
> > actually crashed. Can you add that?
> >
> > Also it seems you are using SLUB allocator, right? This seems like a
> > problem there so adding some CCs.
>
> Likely some data corruption. Enable slub debugging by passing
>
> slub_debug
>
> on the kernel commandline please to get some information as to where and
> when this happens.
>

Hi Christoph.

Thank you for your reply. I'll enable slub debugging and post anything of
interest.

~ Andy

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to [email protected]. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"[email protected]"> [email protected] </a>