Re: [RFC PATCH net-next 00/11] netns: don't switch namespace while creating kernel sockets

[Ying Xue <ying.xue@windriver.com>]

On 05/08/2015 12:14 AM, Eric W. Biederman wrote:
> I agree that commit 23fe18669e7f ("[NETNS]: Fix race between put_net()
> and netlink_kernel_create()."  was a hack.
>

Thanks for the agreement :)

> However it is not appropriate to call get_net on a network namespace
> whose count might be zero.

I will explain why it's still safe for us in another mail even if we do this.

  I believe all of your patches rely on that
> currently.

Yes, you are right.

  Instead we need to build something like sk_release_kernel
> that does not increase the network namespace reference count

Please refer to above comments.

 if you are
> going to avoid changing the network namespace on a socket (a worthy
> goal).
> 

Thanks to the conformation for the effort!

> The following change shows how it is possible to always know that your
> network namespace has a non-zero reference count in the network
> namespace initialization methods.  My implementation of
> lock_network_namespaces is problematic in that it does not sleep
> while network namespaces are unregistering.  But it is enough to show
> how the locking and reference counting can be fixed.
> 

If my understanding for your proposal is right, register_pernet_subsys() will
return a failed error code to its caller if it's found that there is a net whose
refcount is zero from net_namespace_list in lock_network_namespaces(), which
means the per namespace operation is not registered at all in this situation.
But in practice we should not reject the registration even if there is a dead
net linked in the global net_namespace_list.

Regards,
Ying

> Eric
> 
> 
> diff --git a/net/core/net_namespace.c b/net/core/net_namespace.c
> index a3abb719221f..81c53ccc5764 100644
> --- a/net/core/net_namespace.c
> +++ b/net/core/net_namespace.c
> @@ -822,6 +822,49 @@ static void unregister_pernet_operations(struct pernet_operations *ops)
>  		ida_remove(&net_generic_ids, *ops->id);
>  }
>  
> +static void unlock_network_namespaces(void)
> +{
> +	/* Drop the reference count to every network namespace
> +	 * and then release the net_mutex.
> +	 */
> +	struct net *net;
> +
> +	for_each_net(net)
> +		put_net(net);
> +
> +	mutex_unlock(&net_mutex);
> +}
> +
> +static void lock_network_namespaces(void)
> +{
> +	/* Take the mutex lock ensuring no new network namespaces
> +	 * and take a reference on all existing network namespaces
> +	 * allowing network namespace initialization code to take
> +	 * further references
> +	 */
> +	for (;;) {
> +		struct net *net, *stop;
> +
> +		mutex_lock(&net_mutex);
> +		for_each_net(net) {
> +			if (!maybe_get_net(net))
> +				goto undo;
> +		}
> +		return;
> +undo:
> +		/* Remember the network namespace whose reference
> +		 * count was not acquired. */
> +		stop = net;
> +		for_each_net(net) {
> +			if (net_eq(net, stop))
> +				goto undone;
> +			put_net(net);
> +		}
> +undone:
> +		mutex_unlock(&net_mutex);
> +	}
> +}
> +
>  /**
>   *      register_pernet_subsys - register a network namespace subsystem
>   *	@ops:  pernet operations structure for the subsystem
> @@ -844,9 +887,9 @@ static void unregister_pernet_operations(struct pernet_operations *ops)
>  int register_pernet_subsys(struct pernet_operations *ops)
>  {
>  	int error;
> -	mutex_lock(&net_mutex);
> +	lock_network_namespaces();
>  	error =  register_pernet_operations(first_device, ops);
> -	mutex_unlock(&net_mutex);
> +	unlock_network_namespaces();
>  	return error;
>  }
>  EXPORT_SYMBOL_GPL(register_pernet_subsys);
> @@ -890,11 +933,11 @@ EXPORT_SYMBOL_GPL(unregister_pernet_subsys);
>  int register_pernet_device(struct pernet_operations *ops)
>  {
>  	int error;
> -	mutex_lock(&net_mutex);
> +	lock_network_namespaces();
>  	error = register_pernet_operations(&pernet_list, ops);
>  	if (!error && (first_device == &pernet_list))
>  		first_device = &ops->list;
> -	mutex_unlock(&net_mutex);
> +	unlock_network_namespaces();
>  	return error;
>  }
>  EXPORT_SYMBOL_GPL(register_pernet_device);