Re: [PATCH net] caif: Fix crash due to uninitialized net name-space.

[Eric Dumazet]

Le mercredi 25 janvier 2012 Ã 21:33 +0100, Sjur BrÃndeland a Ãcrit :
> net_generic() calls BUG_ON() if called with uninitialized
> network name-space. Add check if net is initialized before
> calling net_generic(). This fixes the following oops:
> 
> [  200.752016] kernel BUG at include/net/netns/generic.h:40!
> ...
> [  200.752016]  [<ffffffff825c3cea>] ? get_cfcnfg+0x3a/0x180
> [  200.752016]  [<ffffffff821cf0b0>] ? lockdep_rtnl_is_held+0x10/0x20
> [  200.752016]  [<ffffffff825c41be>] caif_device_notify+0x2e/0x530
> [  200.752016]  [<ffffffff810d61b7>] notifier_call_chain+0x67/0x110
> [  200.752016]  [<ffffffff810d67c1>] raw_notifier_call_chain+0x11/0x20
> [  200.752016]  [<ffffffff821bae82>] call_netdevice_notifiers+0x32/0x60
> [  200.752016]  [<ffffffff821c2b26>] register_netdevice+0x196/0x300
> [  200.752016]  [<ffffffff821c2ca9>] register_netdev+0x19/0x30
> [  200.752016]  [<ffffffff81c1c67a>] loopback_net_init+0x4a/0xa0
> [  200.752016]  [<ffffffff821b5e62>] ops_init+0x42/0x180
> [  200.752016]  [<ffffffff821b600b>] setup_net+0x6b/0x100
> [  200.752016]  [<ffffffff821b6466>] copy_net_ns+0x86/0x110
> [  200.752016]  [<ffffffff810d5789>] create_new_namespaces+0xd9/0x190
> 
> Signed-off-by: Sjur BrÃndeland <sjur.brandeland@xxxxxxxxxxxxxx>
> Tested-by: Sasha Levin <levinsasha928@xxxxxxxxx>
> 
> ---
> 
> Hi Sasha and Dave,
> 
> [Sasha]
> >Works for me.
> Thank you Sasha for reporting this bug and testing my patch,
> I appreciate it.
> 
> [Dave]
> >Please post all networking patches CC:'d
> Sorry, I missed the obvious.
> I'm resending the same patch as yesterday, this time to:netdev
> and with "Tested-by: Sasha". Please apply to net.
> 
> Thanks,
> Sjur
> 
>  net/caif/caif_dev.c |    8 ++++++++
>  1 files changed, 8 insertions(+), 0 deletions(-)
> 
> diff --git a/net/caif/caif_dev.c b/net/caif/caif_dev.c
> index 673728a..6110ade 100644
> --- a/net/caif/caif_dev.c
> +++ b/net/caif/caif_dev.c
> @@ -371,6 +371,14 @@ static int caif_device_notify(struct notifier_block *me, unsigned long what,
>  	struct cflayer *layer, *link_support;
>  	int head_room = 0;
>  	struct caif_device_entry_list *caifdevs;
> +	int len;
> +
> +	rcu_read_lock();
> +	len = rcu_dereference(dev_net(dev)->gen)->len;
> +	rcu_read_unlock();
> +
> +	if (caif_net_id == 0 || caif_net_id > len)
> +		return 0;
>  
>  	cfg = get_cfcnfg(dev_net(dev));
>  	caifdevs = caif_device_list(dev_net(dev));

This looks wrong.

This should not be needed, something is broken elsewhere.



--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/