netlink: Fix dump skb leak/double free
authorHerbert Xu <herbert@gondor.apana.org.au>
Mon, 16 May 2016 09:28:16 +0000 (17:28 +0800)
committerSasha Levin <sasha.levin@oracle.com>
Tue, 12 Jul 2016 12:48:21 +0000 (08:48 -0400)
[ Upstream commit 92964c79b357efd980812c4de5c1fd2ec8bb5520 ]

When we free cb->skb after a dump, we do it after releasing the
lock.  This means that a new dump could have started in the time
being and we'll end up freeing their skb instead of ours.

This patch saves the skb and module before we unlock so we free
the right memory.

Fixes: 16b304f3404f ("netlink: Eliminate kmalloc in netlink dump operation.")
Reported-by: Baozeng Ding <sploving1@gmail.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Acked-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
net/netlink/af_netlink.c

index fe106b50053e8f2f0220c3fb52c7d8e9ca860a44..15fc0938e1c4308c4ed48ce69aed302a2d813fba 100644 (file)
@@ -2630,6 +2630,7 @@ static int netlink_dump(struct sock *sk)
        struct netlink_callback *cb;
        struct sk_buff *skb = NULL;
        struct nlmsghdr *nlh;
+       struct module *module;
        int len, err = -ENOBUFS;
        int alloc_size;
 
@@ -2699,9 +2700,11 @@ static int netlink_dump(struct sock *sk)
                cb->done(cb);
 
        nlk->cb_running = false;
+       module = cb->module;
+       skb = cb->skb;
        mutex_unlock(nlk->cb_mutex);
-       module_put(cb->module);
-       consume_skb(cb->skb);
+       module_put(module);
+       consume_skb(skb);
        return 0;
 
 errout_skb: