[Xenomai-help] Kernel Oops with latest SVN Xenomai

[Xenomai-help] Kernel Oops with latest SVN Xenomai

[M. Koehrer]

Hi all,

using the latest SVN version of Xenomai and the integrated patches on a 2.6.19 kernel
I get a kernel Oops with my application and the system freezes.
Here is the oops:
------------[ cut here ]------------
kernel BUG at fs/buffer.c:1233!
invalid opcode: 0000 [#1]
CPU:    0
EIP:    0060:[__find_get_block+36/350]    Not tainted VLI
EFLAGS: 00013046   (2.6.19.1 #3)
EIP is at __find_get_block+0x24/0x15e
eax: 00000000   ebx: dae1fcf4   ecx: 00000000   edx: 0007035e
esi: 0007035e   edi: 00000000   ebp: c146fac0   esp: dae1fb5c
ds: 007b   es: 007b   ss: 0068
Process model (pid: 1674, ti=dae1e000 task=df81f560 task.ti=dae1e000)
Stack: dae1fb68 c0124a84 dae1fbe8 00000000 c015b39a dae1fbe8 c015b3bb 00000000
       dae434c0 dae1fcf4 0007035e 00000000 00000000 c016be11 c146fac0 0007035e
       00000000 00001000 dae1fe50 dae1fe54 dae1fe58 00000200 00000000 00000000
Call Trace:
 [remove_wait_queue+21/93] remove_wait_queue+0x15/0x5d
 [free_poll_entry+14/22] free_poll_entry+0xe/0x16
 [poll_freewait+25/78] poll_freewait+0x19/0x4e
 [__getblk+43/628] __getblk+0x2b/0x274
 [search_by_key+130/3465] search_by_key+0x82/0xd89
 [default_wake_function+0/21] default_wake_function+0x0/0x15
 [__getblk+43/628] __getblk+0x2b/0x274
 [ll_rw_block+130/147] ll_rw_block+0x82/0x93
 [search_by_key+420/3465] search_by_key+0x1a4/0xd89
 [reiserfs_update_sd_size+160/722] reiserfs_update_sd_size+0xa0/0x2d2
 [journal_begin+185/250] journal_begin+0xb9/0xfa
 [__d_lookup+157/209] __d_lookup+0x9d/0xd1
 [reiserfs_dirty_inode+100/132] reiserfs_dirty_inode+0x64/0x84
 [__mark_inode_dirty+48/338] __mark_inode_dirty+0x30/0x152
 [touch_atime+115/173] touch_atime+0x73/0xad
 [__link_path_walk+1892/2828] __link_path_walk+0x764/0xb0c
 [mntput_no_expire+28/92] mntput_no_expire+0x1c/0x5c
 [link_path_walk+198/208] link_path_walk+0xc6/0xd0
 [link_path_walk+83/208] link_path_walk+0x53/0xd0
 [__handle_mm_fault+1686/1810] __handle_mm_fault+0x696/0x712
 [<e0a03096>] schedule_event+0x0/0x13c [xeno_nucleus]
 [do_path_lookup+354/444] do_path_lookup+0x162/0x1bc
 [getname+105/201] getname+0x69/0xc9
 [__user_walk_fd+48/66] __user_walk_fd+0x30/0x42
 [vfs_stat_fd+30/82] vfs_stat_fd+0x1e/0x52
 [__handle_mm_fault+1686/1810] __handle_mm_fault+0x696/0x712
 [<e0a03096>] schedule_event+0x0/0x13c [xeno_nucleus]
 [vfs_stat+31/35] vfs_stat+0x1f/0x23
 [sys_stat64+24/44] sys_stat64+0x18/0x2c
 [do_gettimeofday+40/192] do_gettimeofday+0x28/0xc0
 [sys_time+37/53] sys_time+0x25/0x35
 [syscall_call+7/11] syscall_call+0x7/0xb
 =======================
Code: 0c 83 c4 0c 5b 5e c3 55 57 56 53 83 ec 24 8b 6c 24 38 8b 74 24 3c 8b 7c 24 40 e8 6d a8 fc ff 83 f8 01 19 c0 31 c9 f6 c4 02 75 08 <0f> 0b d1 04 6a 6a 36 c0 8b 1c 8d 60 aa 46 c0 85 db 74 47 39 6b
EIP: [__find_get_block+36/350] __find_get_block+0x24/0x15e SS:ESP 0068:dae1fb5c
------------[ cut here ]------------


Unfortunately I have no idea, what causes this oops.
Is there anybody that has some hints on this oops?

Thanks in advance.

Regards

Mathias

-- 
Mathias Koehrer
mathias_koehrer@domain.hid


Viel oder wenig? Schnell oder langsam? Unbegrenzt surfen + telefonieren
ohne Zeit- und Volumenbegrenzung? DAS TOP ANGEBOT JETZT bei Arcor: günstig
und schnell mit DSL - das All-Inclusive-Paket für clevere Doppel-Sparer,
nur  44,85 €  inkl. DSL- und ISDN-Grundgebühr!
http://www.arcor.de/rd/emf-dsl-2

Re: [Xenomai-help] Kernel Oops with latest SVN Xenomai

[Philippe Gerum]

On Wed, 2007-01-03 at 16:29 +0100, M. Koehrer wrote:
> Hi all,
> 
> using the latest SVN version of Xenomai and the integrated patches on a 2.6.19 kernel

Which means? 1.6-03 + the tasklist_lock fix?

> I get a kernel Oops with my application and the system freezes.
> Here is the oops:
> ------------[ cut here ]------------
> kernel BUG at fs/buffer.c:1233!

> [__user_walk_fd+48/66] __user_walk_fd+0x30/0x42
> [vfs_stat_fd+30/82] vfs_stat_fd+0x1e/0x52
> [__handle_mm_fault+1686/1810] __handle_mm_fault+0x696/0x712
> [<e0a03096>] schedule_event+0x0/0x13c [xeno_nucleus]

[...]

At first sight, the backtrace is close to meaningless, and in any case,
really weird. Either the kernel badly misses unwinding information, or
you have a real build problem pending. Try enabling CONFIG_DEBUG_KERNEL
if not already done to get full backtracing support, and don't forget to
reinstall all modules if any.

> Unfortunately I have no idea, what causes this oops.
> Is there anybody that has some hints on this oops?

It's always more efficient to compile all Xenomai support statically
into the kernel instead of using modules when debugging: virtual address
mapping to symbolic info is much easier.

-- 
Philippe.

Re: [Xenomai-help] Kernel Oops with latest SVN Xenomai

[Gilles Chanteperdrix]

M. Koehrer wrote:
> Hi all,
> 
> using the latest SVN version of Xenomai and the integrated patches on a 2.6.19 kernel
> I get a kernel Oops with my application and the system freezes.

Could you send your .config to the list ?

-- 
                                                 Gilles Chanteperdrix

Re: Re: [Xenomai-help] Kernel Oops with latest SVN Xenomai

[M. Koehrer]

[-- Attachment #1.1: Type: text/plain, Size: 729 bytes --]

Hi Gilles, 

enclosed is the .config file.

Regards

Mathias
> > using the latest SVN version of Xenomai and the integrated patches on a
> 2.6.19 kernel
> > I get a kernel Oops with my application and the system freezes.
> 
> Could you send your .config to the list ?
> 
> -- 
>                                                  Gilles Chanteperdrix
> 


-- 
Mathias Koehrer
mathias_koehrer@domain.hid


Viel oder wenig? Schnell oder langsam? Unbegrenzt surfen + telefonieren
ohne Zeit- und Volumenbegrenzung? DAS TOP ANGEBOT JETZT bei Arcor: günstig
und schnell mit DSL - das All-Inclusive-Paket für clevere Doppel-Sparer,
nur  44,85 €  inkl. DSL- und ISDN-Grundgebühr!
http://www.arcor.de/rd/emf-dsl-2

[-- Attachment #2: config.gz --]
[-- Type: application/octetstream, Size: 8478 bytes --]

Re: [Xenomai-help] Kernel Oops with latest SVN Xenomai

[Gilles Chanteperdrix]

M. Koehrer wrote:
> Hi Gilles, 
> 
> enclosed is the .config file.
> 
> Regards

There seem to be nothing suspicious in your configuration. Maybe you
could try disabling CONFIG_X86_PM_TIMER. Also, the error you have seem
to happen in the filesystem code, so maybe you should try running fsck
on your filesystems.

-- 
                                                 Gilles Chanteperdrix