[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <20180330.110012.2226179718404929635.davem@davemloft.net>
Date: Fri, 30 Mar 2018 11:00:12 -0400 (EDT)
From: David Miller <davem@...emloft.net>
To: ktkhai@...tuozzo.com
Cc: steffen.klassert@...unet.com, herbert@...dor.apana.org.au,
pablo@...filter.org, kadlec@...ckhole.kfki.hu, fw@...len.de,
daniel@...earbox.net, jakub.kicinski@...ronome.com, ast@...nel.org,
brouer@...hat.com, linux@...musvillemoes.dk,
john.fastabend@...il.com, dsahern@...il.com,
netdev@...r.kernel.org, netfilter-devel@...r.kernel.org,
coreteam@...filter.org
Subject: Re: [PATCH net-next 0/3] Close race between {un,
}register_netdevice_notifier and pernet_operations
From: Kirill Tkhai <ktkhai@...tuozzo.com>
Date: Thu, 29 Mar 2018 17:03:15 +0300
> the problem is {,un}register_netdevice_notifier() do not take
> pernet_ops_rwsem, and they don't see network namespaces, being
> initialized in setup_net() and cleanup_net(), since at this
> time net is not hashed to net_namespace_list.
>
> This may lead to imbalance, when a notifier is called at time of
> setup_net()/net is alive, but it's not called at time of cleanup_net(),
> for the devices, hashed to the net, and vise versa. See (3/3) for
> the scheme of imbalance.
>
> This patchset fixes the problem by acquiring pernet_ops_rwsem
> at the time of {,un}register_netdevice_notifier() (3/3).
> (1-2/3) are preparations in xfrm and netfilter subsystems.
>
> The problem was introduced a long ago, but backporting won't be easy,
> since every previous kernel version may have changes in netdevice
> notifiers, and they all need review and testing. Otherwise, there
> may be more pernet_operations, which register or unregister
> netdevice notifiers, and that leads to deadlock (which is was fixed
> in 1-2/3). This patchset is for net-next.
I am applying this series and skipping the rwsem revert.
Thanks Kirill.
Powered by blists - more mailing lists