Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
From:   Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To:     linux-kernel@vger.kernel.org
Cc:     Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
        stable@vger.kernel.org, Eric Dumazet <eric.dumazet@gmail.com>,
        Florian Westphal <fw@strlen.de>,
        Pablo Neira Ayuso <pablo@netfilter.org>,
        Sasha Levin <sashal@kernel.org>
Subject: [PATCH 5.16 18/28] netfilter: egress: silence egress hook lockdep splats
Date:   Thu, 17 Mar 2022 13:46:09 +0100
Message-Id: <20220317124527.285482124@linuxfoundation.org>
In-Reply-To: <20220317124526.768423926@linuxfoundation.org>
References: <20220317124526.768423926@linuxfoundation.org>
User-Agent: quilt/0.66
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Precedence: bulk

From: Florian Westphal <fw@strlen.de>

[ Upstream commit 17a8f31bba7bac8cce4bd12bab50697da96e7710 ]

Netfilter assumes its called with rcu_read_lock held, but in egress
hook case it may be called with BH readlock.

This triggers lockdep splat.

In order to avoid to change all rcu_dereference() to
rcu_dereference_check(..., rcu_read_lock_bh_held()), wrap nf_hook_slow
with read lock/unlock pair.

Reported-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 include/linux/netfilter_netdev.h | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/include/linux/netfilter_netdev.h b/include/linux/netfilter_netdev.h
index b4dd96e4dc8d..e6487a691136 100644
--- a/include/linux/netfilter_netdev.h
+++ b/include/linux/netfilter_netdev.h
@@ -101,7 +101,11 @@ static inline struct sk_buff *nf_hook_egress(struct sk_buff *skb, int *rc,
 	nf_hook_state_init(&state, NF_NETDEV_EGRESS,
 			   NFPROTO_NETDEV, dev, NULL, NULL,
 			   dev_net(dev), NULL);
+
+	/* nf assumes rcu_read_lock, not just read_lock_bh */
+	rcu_read_lock();
 	ret = nf_hook_slow(skb, &state, e, 0);
+	rcu_read_unlock();
 
 	if (ret == 1) {
 		return skb;
-- 
2.34.1