Re: ip_conntrack related slab error (Re: Fw: Re: 2.6.16-rc1-mm3)

[Harald Welte <laforge@netfilter.org>]

[-- Attachment #1: Type: text/plain, Size: 1978 bytes --]

On Tue, Jan 31, 2006 at 11:05:43PM +1300, Reuben Farrelly wrote:

> >>slab error in kmem_cache_destroy(): cache `ip_conntrack': Can't free all objects
> >>  [<b010412b>] show_trace+0xd/0xf
> >>  [<b01041cc>] dump_stack+0x17/0x19
> >>  [<b0155d04>] kmem_cache_destroy+0x9b/0x1a9
> >>  [<f0ebf701>] ip_conntrack_cleanup+0x5d/0x10e [ip_conntrack]
> >>  [<f0ebe31e>] init_or_cleanup+0x1f8/0x283 [ip_conntrack]
> >>  [<f0ec2c4e>] fini+0xa/0x66 [ip_conntrack]
> >>  [<b0136d06>] sys_delete_module+0x161/0x1fb
> >>  [<b0102b3f>] sysenter_past_esp+0x54/0x75
> >>Removing netfilter NETLINK layer.
> >>[root@tornado log]#
> >>I was just reading IMAP mail at the time, ie same as I'd been doing for an hour or two beforehand and not 
> >>altering config of the box in any way.  I was able to log on via console but lost all network connectivity and 
> >>had to reboot :(
> >The codepath you see in that backtrace is only hit during load or
> >removal of the 'ip_conntrack' module.  While this certainly still should
> >not oops, your description of 'not doing anything but IMAP reading' is
> >certainly not true.  
> 
> With the greatest of respect (which I do have for you Harald), I don't
> think being essentially called a liar is very fair.

I didn't want to imply that you are lying, but merely point out that
something was going on on your system that you didn't be aware of.

Sorry if that was to be misunderstood.

Without knowing what actually happened at the time you encountered the
bug, it is hard for me to try and reproduce / understand it.

This is not a lame excuse.

-- 
- Harald Welte <laforge@netfilter.org>                 http://netfilter.org/
============================================================================
  "Fragmentation is like classful addressing -- an interesting early
   architectural error that shows how much experimentation was going
   on while IP was being designed."                    -- Paul Vixie

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]