Re: 2.6.12-mm1 (kernel BUG at fs/open.c:935!)
From: Alexey Dobriyan (adobriyan_at_gmail.com)
Date: 06/20/05
- Previous message: Dave Jones: "Re: [SCSI] allow sleeping in ->eh_device_reset_handler()"
- In reply to: Dipankar Sarma: "Re: 2.6.12-mm1 (kernel BUG at fs/open.c:935!)"
- Next in thread: Bernhard Rosenkraenzer: "Re: 2.6.12-mm1 (kernel BUG at fs/open.c:935!)"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
To: dipankar@in.ibm.com Date: Tue, 21 Jun 2005 01:39:55 +0400
On Tuesday 21 June 2005 00:21, Dipankar Sarma wrote:
> On Mon, Jun 20, 2005 at 03:04:02PM +0200, jan malstrom wrote:
> > Jun 20 14:38:07 hades kernel: kernel BUG at fs/open.c:935!
> > Jun 20 14:38:07 hades kernel: invalid operand: 0000 [#1]
> > Jun 20 14:38:07 hades kernel: PREEMPT
> > Jun 20 14:38:07 hades kernel: Modules linked in: ipw2100 i2c_i801
> > Jun 20 14:38:07 hades kernel: CPU: 0
> > Jun 20 14:38:07 hades kernel: EIP: 0060:[fd_install+309/400] Not
> > tainted VLI
>
> Can you try the following patch and let me know if it fixes any
> of your problems ?
> If expand_fdtable() sees that someone else expanded the fdtable
> while it dropped the lock, it can return 0 which in turn
> can be returned by expand_files() even though there has
> been an expansion of the fdtable since expand_files()
> was originally called. This could lead to locate_fd()
> not repeating the fd search and returning a bogus fd.
> --- linux-2.6.12-mm1-test/fs/file.c~fix-expand-files
> +++ linux-2.6.12-mm1-test-dipankar/fs/file.c
Doesn't fix for me.
2.6.12-mm1-935
============================================================================
kernel BUG at fs/open.c:935!
invalid operand: 0000 [#1]
PREEMPT
Modules linked in: floppy evdev ide_cd cdrom snd_intel8x0 snd_ac97_codec
CPU: 0
EIP: 0060:[<c01517ad>] Not tainted VLI
EFLAGS: 00210282 (2.6.12-mm1)
EIP is at fd_install+0x7d/0x90
eax: c1573280 ebx: deee0280 ecx: 00000001 edx: de597840
esi: de45c000 edi: 00000080 ebp: c1573a80 esp: de45cf68
ds: 007b es: 007b ss: 0068
Process kded (pid: 6947, threadinfo=de45c000 task=de6d10a0)
Stack: 00000080 deee0280 de597840 de45c000 c0163de2 c1573a80 00000080 ffffffea
0000000c 0000000c c016424a 00000000 c1573a80 fffffff7 c0164400 c1573a80
0000000c 00000080 b6c58ff4 de45c000 c0102d05 0000000c 00000000 00000080
Call Trace:
[<c0163de2>] dupfd+0x62/0xa0
[<c016424a>] do_fcntl+0xba/0x190
[<c0164400>] sys_fcntl64+0x80/0x90
[<c0102d05>] syscall_call+0x7/0xb
Code: 13 8b 1c 24 8b 74 24 04 8b 7c 24 08 8b 6c 24 0c 83 c4 10 c3 8b 1c 24 8b 74 24 04 8b 7c 24 08 8b 6c 24 0c 83 c4 10 e9 23 06 19 00 <0f> 0b a7 03 1c 6b 2f c0 eb b5 89 f6 8d bc 27 00 00 00 00 83 ec
<6>note: kded[6947] exited with preempt_count 1
...
2.6.12-mm1
============================================================================
kernel BUG at fs/open.c:935!
invalid operand: 0000 [#1]
PREEMPT
Modules linked in: floppy evdev ide_cd cdrom snd_intel8x0 snd_ac97_codec
CPU: 0
EIP: 0060:[<c01517ad>] Not tainted VLI
EFLAGS: 00210282 (2.6.12-mm1)
EIP is at fd_install+0x7d/0x90
eax: de5a9980 ebx: def38280 ecx: 00000001 edx: de534840
esi: de2f9000 edi: 00000080 ebp: de6c2280 esp: de2f9f68
ds: 007b es: 007b ss: 0068
Process kded (pid: 6947, threadinfo=de2f9000 task=de927510)
Stack: 00000080 def38280 de534840 de2f9000 c0163de2 de6c2280 00000080 ffffffea
0000000c 0000000c c016424a 00000000 de6c2280 fffffff7 c0164400 de6c2280
0000000c 00000080 b6c68ff4 de2f9000 c0102d05 0000000c 00000000 00000080
Call Trace:
[<c0163de2>] dupfd+0x62/0xa0
[<c016424a>] do_fcntl+0xba/0x190
[<c0164400>] sys_fcntl64+0x80/0x90
[<c0102d05>] syscall_call+0x7/0xb
Code: 13 8b 1c 24 8b 74 24 04 8b 7c 24 08 8b 6c 24 0c 83 c4 10 c3 8b 1c 24 8b 74 24 04 8b 7c 24 08 8b 6c 24 0c 83 c4 10 e9 43 06 19 00 <0f> 0b a7 03 3c 6b 2f c0 eb b5 89 f6 8d bc 27 00 00 00 00 83 ec
<6>note: kded[6947] exited with preempt_count 1
scheduling while atomic: kded/0x10000001/6947
[<c02e1de2>] schedule+0x672/0x680
[<c014254e>] zap_pte_range+0xde/0x1b0
[<c01426ab>] unmap_page_range+0x8b/0xb0
[<c02e280a>] cond_resched+0x2a/0x50
[<c0142890>] unmap_vmas+0x1c0/0x220
[<c014707a>] exit_mmap+0x7a/0x160
[<c0112c01>] mmput+0x41/0x110
[<c011769d>] do_exit+0xcd/0x4e0
[<c01036db>] die+0x16b/0x170
[<c0103a40>] do_invalid_op+0x0/0xc0
[<c0103adf>] do_invalid_op+0x9f/0xc0
[<c01517ad>] fd_install+0x7d/0x90
[<c0123f7d>] in_group_p+0x3d/0xa0
[<c01b1cfe>] __reiserfs_permission+0x23e/0x290
[<c01b1d50>] reiserfs_permission+0x0/0x20
[<c01b1d5f>] reiserfs_permission+0xf/0x20
[<c015f9eb>] permission+0x8b/0xa0
[<c0161797>] may_open+0x47/0x1e0
[<c015301b>] get_empty_filp+0x5b/0xd0
[<c0102f1f>] error_code+0x4f/0x54
[<c01517ad>] fd_install+0x7d/0x90
[<c0163de2>] dupfd+0x62/0xa0
[<c016424a>] do_fcntl+0xba/0x190
[<c0164400>] sys_fcntl64+0x80/0x90
[<c0102d05>] syscall_call+0x7/0xb
...
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
- Previous message: Dave Jones: "Re: [SCSI] allow sleeping in ->eh_device_reset_handler()"
- In reply to: Dipankar Sarma: "Re: 2.6.12-mm1 (kernel BUG at fs/open.c:935!)"
- Next in thread: Bernhard Rosenkraenzer: "Re: 2.6.12-mm1 (kernel BUG at fs/open.c:935!)"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Relevant Pages
|
|
