SLUB cifs_request: kmem_cache_destroy called for cache that still has objects.

SLUB cifs_request: kmem_cache_destroy called for cache that still has objects.

[Jeff Layton]

I saw this today when unplugging cifs.ko in a stock -rc7 kernel. I
never mounted any cifs mounts, so this is entirely due to module
initialization as far as I can tell:

[  464.776843] SLUB cifs_request: kmem_cache_destroy called for cache that still has objects.
[  464.779071] Pid: 4738, comm: rmmod Not tainted 3.4.0-0.rc7.git0.1.fc18.x86_64 #1
[  464.780243] Call Trace:
[  464.780433]  [<ffffffff8116b53a>] kmem_cache_destroy+0x2ba/0x360
[  464.780870]  [<ffffffffa00d6469>] cifs_destroy_request_bufs+0x21/0x3b [cifs]
[  464.781397]  [<ffffffffa00d66a6>] exit_cifs+0x30/0x98a [cifs]
[  464.781944]  [<ffffffff810b4a3e>] sys_delete_module+0x16e/0x2d0
[  464.782482]  [<ffffffff8117c186>] ? filp_close+0x66/0xa0
[  464.782903]  [<ffffffff81603769>] system_call_fastpath+0x16/0x1b

...any ideas?
-- 
Jeff Layton <jlayton-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>

Re: SLUB cifs_request: kmem_cache_destroy called for cache that still has objects.

[Steve French]

I don't see that on my rc7 system and I have slub debugging and
various other debugging options on in my rc7 config.  Let me know if
you find a way to reproduce it or want me to try with your config.

On Tue, May 15, 2012 at 7:46 PM, Jeff Layton <jlayton-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org> wrote:
> I saw this today when unplugging cifs.ko in a stock -rc7 kernel. I
> never mounted any cifs mounts, so this is entirely due to module
> initialization as far as I can tell:
>
> [  464.776843] SLUB cifs_request: kmem_cache_destroy called for cache that still has objects.
> [  464.779071] Pid: 4738, comm: rmmod Not tainted 3.4.0-0.rc7.git0.1.fc18.x86_64 #1
> [  464.780243] Call Trace:
> [  464.780433]  [<ffffffff8116b53a>] kmem_cache_destroy+0x2ba/0x360
> [  464.780870]  [<ffffffffa00d6469>] cifs_destroy_request_bufs+0x21/0x3b [cifs]
> [  464.781397]  [<ffffffffa00d66a6>] exit_cifs+0x30/0x98a [cifs]
> [  464.781944]  [<ffffffff810b4a3e>] sys_delete_module+0x16e/0x2d0
> [  464.782482]  [<ffffffff8117c186>] ? filp_close+0x66/0xa0
> [  464.782903]  [<ffffffff81603769>] system_call_fastpath+0x16/0x1b
>
> ...any ideas?
> --
> Jeff Layton <jlayton-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-cifs" in
> the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html



-- 
Thanks,

Steve

Re: SLUB cifs_request: kmem_cache_destroy called for cache that still has objects.

[Suresh Jayaraman]

On 05/16/2012 09:30 AM, Steve French wrote:
> I don't see that on my rc7 system and I have slub debugging and
> various other debugging options on in my rc7 config.  Let me know if
> you find a way to reproduce it or want me to try with your config.
> 
> On Tue, May 15, 2012 at 7:46 PM, Jeff Layton <jlayton-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org> wrote:
>> I saw this today when unplugging cifs.ko in a stock -rc7 kernel. I
>> never mounted any cifs mounts, so this is entirely due to module
>> initialization as far as I can tell:
>>
>> [  464.776843] SLUB cifs_request: kmem_cache_destroy called for cache that still has objects.
>> [  464.779071] Pid: 4738, comm: rmmod Not tainted 3.4.0-0.rc7.git0.1.fc18.x86_64 #1
>> [  464.780243] Call Trace:
>> [  464.780433]  [<ffffffff8116b53a>] kmem_cache_destroy+0x2ba/0x360
>> [  464.780870]  [<ffffffffa00d6469>] cifs_destroy_request_bufs+0x21/0x3b [cifs]
>> [  464.781397]  [<ffffffffa00d66a6>] exit_cifs+0x30/0x98a [cifs]
>> [  464.781944]  [<ffffffff810b4a3e>] sys_delete_module+0x16e/0x2d0
>> [  464.782482]  [<ffffffff8117c186>] ? filp_close+0x66/0xa0
>> [  464.782903]  [<ffffffff81603769>] system_call_fastpath+0x16/0x1b
>>
>> ...any ideas?

Sounds to me like a slub problem affecting random modules that could be
fixed by this patch

  http://comments.gmane.org/gmane.linux.kernel.mm/78106

If you are able to reliably reproduce, it is worth giving the patch a
shot. But, I could be totally wrong too...


Suresh

Re: SLUB cifs_request: kmem_cache_destroy called for cache that still has objects.

[Jeff Layton]

On Wed, 16 May 2012 11:41:39 +0530
Suresh Jayaraman <sjayaraman-IBi9RG/b67k@public.gmane.org> wrote:

> On 05/16/2012 09:30 AM, Steve French wrote:
> > I don't see that on my rc7 system and I have slub debugging and
> > various other debugging options on in my rc7 config.  Let me know if
> > you find a way to reproduce it or want me to try with your config.
> > 
> > On Tue, May 15, 2012 at 7:46 PM, Jeff Layton <jlayton-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org> wrote:
> >> I saw this today when unplugging cifs.ko in a stock -rc7 kernel. I
> >> never mounted any cifs mounts, so this is entirely due to module
> >> initialization as far as I can tell:
> >>
> >> [  464.776843] SLUB cifs_request: kmem_cache_destroy called for cache that still has objects.
> >> [  464.779071] Pid: 4738, comm: rmmod Not tainted 3.4.0-0.rc7.git0.1.fc18.x86_64 #1
> >> [  464.780243] Call Trace:
> >> [  464.780433]  [<ffffffff8116b53a>] kmem_cache_destroy+0x2ba/0x360
> >> [  464.780870]  [<ffffffffa00d6469>] cifs_destroy_request_bufs+0x21/0x3b [cifs]
> >> [  464.781397]  [<ffffffffa00d66a6>] exit_cifs+0x30/0x98a [cifs]
> >> [  464.781944]  [<ffffffff810b4a3e>] sys_delete_module+0x16e/0x2d0
> >> [  464.782482]  [<ffffffff8117c186>] ? filp_close+0x66/0xa0
> >> [  464.782903]  [<ffffffff81603769>] system_call_fastpath+0x16/0x1b
> >>
> >> ...any ideas?
> 
> Sounds to me like a slub problem affecting random modules that could be
> fixed by this patch
> 
>   http://comments.gmane.org/gmane.linux.kernel.mm/78106
> 
> If you are able to reliably reproduce, it is worth giving the patch a
> shot. But, I could be totally wrong too...
> 

Thanks for the pointer. That seems to have fixed it. I'll see if I can
reply to the thread and give them a "Tested-by" as well...

Thanks!
-- 
Jeff Layton <jlayton-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>