2.6.6-mm1 Oops with dummy network device (sysfs related?)

2.6.6-mm1 Oops with dummy network device (sysfs related?)

[Richard A Nelson]

: divert: allocating divert_blk for dummy0
: Unable to handle kernel NULL pointer dereference at virtual address 00000000
:  printing eip:
: c016dcad
: *pde = 00000000
:        ___      ______
:       0--,|    /OOOOOO\
:      {_o  /  /OO plop OO\
:        \__\_/OO oh dear OOO\s
:           \OOOOOOOOOOOOOOOO/
:            __XXX__   __XXX__
: Oops: 0000 [#1]
: PREEMPT
: CPU:    0
: EIP:    0060:[d_move+109/576]    Not tainted VLI
: EFLAGS: 00210246   (2.6.6-mm1)
: EIP is at d_move+0x6d/0x240
: eax: 00000000   ebx: c25549a4   ecx: c2554a0c   edx: 00000000
: esi: c313d908   edi: c155f928   ebp: c496aebc   esp: c496aea8
: ds: 007b   es: 007b   ss: 0068
: Process ip (pid: 5604, threadinfo=c496a000 task=c7cc54c0)
: Stack: 00000006 c0149c27 c480cca0 c313d908 c155f928 c496aedc c018e251 00200286
:        c496af44 c480ce50 c480ce50 c480cca0 c50d5113 c496aeec c01d18b1 c480ce48
:        c480cca0 c496aefc c0231ca8 c480cca0 c480cdcc c496af24 c028f56b c496af24
: Call Trace:
:  [show_stack+122/144] show_stack+0x7a/0x90
:  [show_registers+324/432] show_registers+0x144/0x1b0
:  [die+153/272] die+0x99/0x110
:  [do_page_fault+485/1327] do_page_fault+0x1e5/0x52f
:  [error_code+45/56] error_code+0x2d/0x38
:  [sysfs_rename_dir+193/224] sysfs_rename_dir+0xc1/0xe0
:  [kobject_rename+33/64] kobject_rename+0x21/0x40
:  [class_device_rename+56/80] class_device_rename+0x38/0x50
:  [dev_change_name+315/448] dev_change_name+0x13b/0x1c0
:  [dev_ioctl+339/704] dev_ioctl+0x153/0x2c0
:  [inet_ioctl+154/176] inet_ioctl+0x9a/0xb0
:  [sock_ioctl+239/656] sock_ioctl+0xef/0x290
:  [sys_ioctl+261/608] sys_ioctl+0x105/0x260
:  [sysenter_past_esp+82/113] sysenter_past_esp+0x52/0x71
:
: Code: 00 8b 46 10 39 43 10 74 37 8b 43 68 8d 4b 68 8b 51 04 85 c0 89 02 74 03 89 50 04 c7 41 04 00 02 20 00 8b
46 10 89 43 10 8b 56 10 <8b> 02 89 51 04 89 43 68 85 c0 74 03 89 48 04 89 0a 83 63 04 ef
:  <6>note: ip[5604] exited with preempt_count 4
: bad: scheduling while atomic!
:  [dump_stack+23/32] dump_stack+0x17/0x20
:  [schedule+1190/1200] schedule+0x4a6/0x4b0
:  [shmem_file_write+523/624] shmem_file_write+0x20b/0x270
:  [do_acct_process+780/816] do_acct_process+0x30c/0x330
:  [acct_process+66/138] acct_process+0x42/0x8a
:  [do_exit+128/1008] do_exit+0x80/0x3f0
:  [die+260/272] die+0x104/0x110
:  [do_page_fault+485/1327] do_page_fault+0x1e5/0x52f
:  [error_code+45/56] error_code+0x2d/0x38
:  [sysfs_rename_dir+193/224] sysfs_rename_dir+0xc1/0xe0
:  [kobject_rename+33/64] kobject_rename+0x21/0x40
:  [class_device_rename+56/80] class_device_rename+0x38/0x50
:  [dev_change_name+315/448] dev_change_name+0x13b/0x1c0
:  [dev_ioctl+339/704] dev_ioctl+0x153/0x2c0
:  [inet_ioctl+154/176] inet_ioctl+0x9a/0xb0
:  [sock_ioctl+239/656] sock_ioctl+0xef/0x290
:  [sys_ioctl+261/608] sys_ioctl+0x105/0x260
:  [sysenter_past_esp+82/113] sysenter_past_esp+0x52/0x71
:

-- 
Rick Nelson
<Crow_> hmm, is there a --now-dammit option for exim?

Re: 2.6.6-mm1 Oops with dummy network device (sysfs related?)

[Stephen Hemminger]

It would be easier to know what is wrong, if you said what you
did that started the problem.  Looks like ifrename or something
like that.


On Mon, 10 May 2004 16:59:54 -0400 (EDT)
Richard A Nelson <cowboy@debian.org> wrote:

> : divert: allocating divert_blk for dummy0
> : Unable to handle kernel NULL pointer dereference at virtual address 00000000
> :  printing eip:
> : c016dcad
> : *pde = 00000000
> :        ___      ______
> :       0--,|    /OOOOOO\
> :      {_o  /  /OO plop OO\
> :        \__\_/OO oh dear OOO\s
> :           \OOOOOOOOOOOOOOOO/
> :            __XXX__   __XXX__
> : Oops: 0000 [#1]
> : PREEMPT
> : CPU:    0
> : EIP:    0060:[d_move+109/576]    Not tainted VLI
> : EFLAGS: 00210246   (2.6.6-mm1)
> : EIP is at d_move+0x6d/0x240
> : eax: 00000000   ebx: c25549a4   ecx: c2554a0c   edx: 00000000
> : esi: c313d908   edi: c155f928   ebp: c496aebc   esp: c496aea8
> : ds: 007b   es: 007b   ss: 0068
> : Process ip (pid: 5604, threadinfo=c496a000 task=c7cc54c0)
> : Stack: 00000006 c0149c27 c480cca0 c313d908 c155f928 c496aedc c018e251 00200286
> :        c496af44 c480ce50 c480ce50 c480cca0 c50d5113 c496aeec c01d18b1 c480ce48
> :        c480cca0 c496aefc c0231ca8 c480cca0 c480cdcc c496af24 c028f56b c496af24
> : Call Trace:
> :  [show_stack+122/144] show_stack+0x7a/0x90
> :  [show_registers+324/432] show_registers+0x144/0x1b0
> :  [die+153/272] die+0x99/0x110
> :  [do_page_fault+485/1327] do_page_fault+0x1e5/0x52f
> :  [error_code+45/56] error_code+0x2d/0x38
> :  [sysfs_rename_dir+193/224] sysfs_rename_dir+0xc1/0xe0
> :  [kobject_rename+33/64] kobject_rename+0x21/0x40
> :  [class_device_rename+56/80] class_device_rename+0x38/0x50
> :  [dev_change_name+315/448] dev_change_name+0x13b/0x1c0
> :  [dev_ioctl+339/704] dev_ioctl+0x153/0x2c0
> :  [inet_ioctl+154/176] inet_ioctl+0x9a/0xb0
> :  [sock_ioctl+239/656] sock_ioctl+0xef/0x290
> :  [sys_ioctl+261/608] sys_ioctl+0x105/0x260
> :  [sysenter_past_esp+82/113] sysenter_past_esp+0x52/0x71
> :
> : Code: 00 8b 46 10 39 43 10 74 37 8b 43 68 8d 4b 68 8b 51 04 85 c0 89 02 74 03 89 50 04 c7 41 04 00 02 20 00 8b
> 46 10 89 43 10 8b 56 10 <8b> 02 89 51 04 89 43 68 85 c0 74 03 89 48 04 89 0a 83 63 04 ef
> :  <6>note: ip[5604] exited with preempt_count 4
> : bad: scheduling while atomic!
> :  [dump_stack+23/32] dump_stack+0x17/0x20
> :  [schedule+1190/1200] schedule+0x4a6/0x4b0
> :  [shmem_file_write+523/624] shmem_file_write+0x20b/0x270
> :  [do_acct_process+780/816] do_acct_process+0x30c/0x330
> :  [acct_process+66/138] acct_process+0x42/0x8a
> :  [do_exit+128/1008] do_exit+0x80/0x3f0
> :  [die+260/272] die+0x104/0x110
> :  [do_page_fault+485/1327] do_page_fault+0x1e5/0x52f
> :  [error_code+45/56] error_code+0x2d/0x38
> :  [sysfs_rename_dir+193/224] sysfs_rename_dir+0xc1/0xe0
> :  [kobject_rename+33/64] kobject_rename+0x21/0x40
> :  [class_device_rename+56/80] class_device_rename+0x38/0x50
> :  [dev_change_name+315/448] dev_change_name+0x13b/0x1c0
> :  [dev_ioctl+339/704] dev_ioctl+0x153/0x2c0
> :  [inet_ioctl+154/176] inet_ioctl+0x9a/0xb0
> :  [sock_ioctl+239/656] sock_ioctl+0xef/0x290
> :  [sys_ioctl+261/608] sys_ioctl+0x105/0x260
> :  [sysenter_past_esp+82/113] sysenter_past_esp+0x52/0x71
> :
> 
> -- 
> Rick Nelson
> <Crow_> hmm, is there a --now-dammit option for exim?
> -
> 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/


-- 
Stephen Hemminger 		mailto:shemminger@osdl.org
Open Source Development Lab	http://developer.osdl.org/shemminger

Re: 2.6.6-mm1 Oops with dummy network device (sysfs related?)

[Richard A Nelson]

On Mon, 10 May 2004, Stephen Hemminger wrote:

> It would be easier to know what is wrong, if you said what you
> did that started the problem.  Looks like ifrename or something
> like that.

hrm, been trying to track that down (several oopses later...)

the modprobe dummy worked ok (and is seen in the log)
the next command is 'ip link set name ipsec0 dev dummy0'
and I've yet to get passed that

-- 
Rick Nelson
People disagree with me.  I just ignore them.
	-- Linus Torvalds, regarding the use of C++ for the Linux kernel

Re: 2.6.6-mm1 Oops with dummy network device (sysfs related?)

[Maneesh Soni]

On Mon, May 10, 2004 at 11:35:00PM +0000, Richard A Nelson wrote:
> On Mon, 10 May 2004, Stephen Hemminger wrote:
> 
> > It would be easier to know what is wrong, if you said what you
> > did that started the problem.  Looks like ifrename or something
> > like that.
> 
> hrm, been trying to track that down (several oopses later...)
> 
> the modprobe dummy worked ok (and is seen in the log)
> the next command is 'ip link set name ipsec0 dev dummy0'
> and I've yet to get passed that

ok it is sysfs related and because of the backing store patches from me.

And here is the fix. Hope this solves your problem. 


Thanks
Maneesh

o Fix sysfs_rename_dir(). The sysfs_lookup() does not hash
  negative dentries so just hash it before calling d_move

 fs/sysfs/dir.c |    1 +
 1 files changed, 1 insertion(+)

diff -puN fs/sysfs/dir.c~sysfs-backing-store-sysfs_rename_dir-fix fs/sysfs/dir.c
--- linux-2.6.6-mm1/fs/sysfs/dir.c~sysfs-backing-store-sysfs_rename_dir-fix	2004-10-06 15:21:37.000000000 +0530
+++ linux-2.6.6-mm1-maneesh/fs/sysfs/dir.c	2004-10-06 15:22:03.000000000 +0530
@@ -314,6 +314,7 @@ void sysfs_rename_dir(struct kobject * k
 	new_dentry = sysfs_get_dentry(parent, new_name);
 	if (!IS_ERR(new_dentry)) {
 		if (!new_dentry->d_inode) {
+			d_add(new_dentry, NULL);
 			d_move(kobj->dentry, new_dentry);
 			kobject_set_name(kobj,new_name);
 		}

_


-- 
Maneesh Soni
Linux Technology Center, 
IBM Software Lab, Bangalore, India
email: maneesh@in.ibm.com
Phone: 91-80-25044999 Fax: 91-80-25268553
T/L : 9243696

Re: 2.6.6-mm1 Oops with dummy network device (sysfs related?)

[Valdis.Kletnieks]

[-- Attachment #1: Type: text/plain, Size: 668 bytes --]

On Wed, 06 Oct 2004 15:26:02 +0530, Maneesh Soni said:

> o Fix sysfs_rename_dir(). The sysfs_lookup() does not hash
>   negative dentries so just hash it before calling d_move
> 
>  fs/sysfs/dir.c |    1 +
>  1 files changed, 1 insertion(+)
> 
> diff -puN fs/sysfs/dir.c~sysfs-backing-store-sysfs_rename_dir-fix fs/sysfs/dir.c
> --- linux-2.6.6-mm1/fs/sysfs/dir.c~sysfs-backing-store-sysfs_rename_dir-fix	2004-10-06 15:21:37.000000000 +0530
> +++ linux-2.6.6-mm1-maneesh/fs/sysfs/dir.c	2004-10-06 15:22:03.000000000 +0530

Confirming that this patch also resolves my 'nameif' issue...  System booted fine
with the sysfs-backing-store patches and this one on top...



[-- Attachment #2: Type: application/pgp-signature, Size: 226 bytes --]

Re: 2.6.6-mm1 Oops with dummy network device (sysfs related?)

[Valdis.Kletnieks]

[-- Attachment #1: Type: text/plain, Size: 1912 bytes --]

On Mon, 10 May 2004 16:59:54 EDT, Richard A Nelson <cowboy@debian.org>  said:
> : divert: allocating divert_blk for dummy0
> : Unable to handle kernel NULL pointer dereference at virtual address 0000000
0
> :  printing eip:
> : c016dcad
> : *pde = 00000000
> :        ___      ______
> :       0--,|    /OOOOOO\
> :      {_o  /  /OO plop OO\
> :        \__\_/OO oh dear OOO\s
> :           \OOOOOOOOOOOOOOOO/
> :            __XXX__   __XXX__
> : Oops: 0000 [#1]
> : PREEMPT
> : CPU:    0
> : EIP:    0060:[d_move+109/576]    Not tainted VLI
> : EFLAGS: 00210246   (2.6.6-mm1)
> : EIP is at d_move+0x6d/0x240
> : eax: 00000000   ebx: c25549a4   ecx: c2554a0c   edx: 00000000
> : esi: c313d908   edi: c155f928   ebp: c496aebc   esp: c496aea8
> : ds: 007b   es: 007b   ss: 0068
> : Process ip (pid: 5604, threadinfo=c496a000 task=c7cc54c0)
> : Stack: 00000006 c0149c27 c480cca0 c313d908 c155f928 c496aedc c018e251 00200
286
> :        c496af44 c480ce50 c480ce50 c480cca0 c50d5113 c496aeec c01d18b1 c480c
e48
> :        c480cca0 c496aefc c0231ca8 c480cca0 c480cdcc c496af24 c028f56b c496a
f24
> : Call Trace:
> :  [show_stack+122/144] show_stack+0x7a/0x90
> :  [show_registers+324/432] show_registers+0x144/0x1b0
> :  [die+153/272] die+0x99/0x110
> :  [do_page_fault+485/1327] do_page_fault+0x1e5/0x52f
> :  [error_code+45/56] error_code+0x2d/0x38
> :  [sysfs_rename_dir+193/224] sysfs_rename_dir+0xc1/0xe0
> :  [kobject_rename+33/64] kobject_rename+0x21/0x40
> :  [class_device_rename+56/80] class_device_rename+0x38/0x50
> :  [dev_change_name+315/448] dev_change_name+0x13b/0x1c0
> :  [dev_ioctl+339/704] dev_ioctl+0x153/0x2c0
> :  [inet_ioctl+154/176] inet_ioctl+0x9a/0xb0
> :  [sock_ioctl+239/656] sock_ioctl+0xef/0x290
> :  [sys_ioctl+261/608] sys_ioctl+0x105/0x260
> :  [sysenter_past_esp+82/113] sysenter_past_esp+0x52/0x71

Am seeing this as  well, while running 'nameif' to rename real Ethernet interfaces.

[-- Attachment #2: Type: application/pgp-signature, Size: 226 bytes --]

Re: 2.6.6-mm1 Oops with dummy network device (sysfs related?)

[Valdis.Kletnieks]

[-- Attachment #1: Type: text/plain, Size: 1315 bytes --]

On Mon, 10 May 2004 21:06:50 EDT, Valdis.Kletnieks@vt.edu said:

> > : Unable to handle kernel NULL pointer dereference at virtual address 0000000

> > :  [show_stack+122/144] show_stack+0x7a/0x90
> > :  [show_registers+324/432] show_registers+0x144/0x1b0
> > :  [die+153/272] die+0x99/0x110
> > :  [do_page_fault+485/1327] do_page_fault+0x1e5/0x52f
> > :  [error_code+45/56] error_code+0x2d/0x38
> > :  [sysfs_rename_dir+193/224] sysfs_rename_dir+0xc1/0xe0
> > :  [kobject_rename+33/64] kobject_rename+0x21/0x40
> > :  [class_device_rename+56/80] class_device_rename+0x38/0x50
> > :  [dev_change_name+315/448] dev_change_name+0x13b/0x1c0
> > :  [dev_ioctl+339/704] dev_ioctl+0x153/0x2c0
> > :  [inet_ioctl+154/176] inet_ioctl+0x9a/0xb0
> > :  [sock_ioctl+239/656] sock_ioctl+0xef/0x290
> > :  [sys_ioctl+261/608] sys_ioctl+0x105/0x260
> > :  [sysenter_past_esp+82/113] sysenter_past_esp+0x52/0x71
> 
> Am seeing this as  well, while running 'nameif' to rename real Ethernet interfaces.

In case nobody shot this bug overnight... More info

The problem is in one of the sysfs-backing-store patches.  I backed out the 6
sysfs-leaves-* patches and my system boots OK. (My kernel *does* still have
fix-sysfs-symlinks on).  It won't be till late afternoon before I have time to
play binary-search on the 6 remaining..


[-- Attachment #2: Type: application/pgp-signature, Size: 226 bytes --]