Re: 2.6.17-mm6

[Andrew Morton <akpm@osdl.org>]

Reuben Farrelly <reuben-lkml@reub.net> wrote:
>
> 
> > 
> > The core slab data structures were wrecked.  For kmalloc(), no less. 
> > Something secretly destroyed your kernel, and it could be anything.  Nice.
> 
> Having now turned on slab debugging, is it possibly related to this message 
> which appeared in my log when I booting up earlier?
> 
> Jul  8 02:49:39 tornado kernel: EXT3-fs: mounted filesystem with ordered data mode.
> Jul  8 02:49:39 tornado kernel: Adding 497972k swap on /dev/sdc9.  Priority:-1 
> extents:1 across:497972k
> Jul  8 02:49:40 tornado kernel: ip_tables: (C) 2000-2006 Netfilter Core Team
> Jul  8 02:49:40 tornado kernel: Netfilter messages via NETLINK v0.30.
> Jul  8 02:49:40 tornado kernel: ip_conntrack version 2.4 (4060 buckets, 32480 
> max) - 288 bytes per conntrack
> Jul  8 02:49:40 tornado kernel: Slab corruption: start=ffff81003efd7000, len=4096
> Jul  8 02:49:40 tornado kernel: 170: ff ff ff ff 00 00 00 00 6b 6b 6b 6b 6b 6b 6b 6b
> Jul  8 02:49:40 tornado kernel: e1000: eth0: e1000_watchdog: NIC Link is Up 1000 
> Mbps Full Duplex
> Jul  8 02:49:40 tornado kernel: GRE over IPv4 tunneling driver
> 

Yikes!  Until we fix that there's no point in looking at anything else.

CONFIG_DEBUG_PAGEALLOC would nail this bug in a flash, but x86_64 doesn't
implement the damn thing :(

So if this is repeatable it would be of some value if you can work out what
causes it - start by disabling netfilter.

But to fix it for real we'll probably need to twiddle thumbs until an x86
person can hit it.

-
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/