[PATCH v6.1.y 4/4] ipv4: start using dst_dev_rcu()

[Miguel Gazquez (Schneider Electric)]

From: Eric Dumazet <edumazet@xxxxxxxxxx>

[ Upstream commit 6ad8de3cefdb6ffa6708b21c567df0dbf82c43a8 ]

Change icmpv4_xrlim_allow(), ip_defrag() to prevent possible UAF.

Change ipmr_prepare_xmit(), ipmr_queue_fwd_xmit(), ip_mr_output(),
ipv4_neigh_lookup() to use lockdep enabled dst_dev_rcu().

Fixes: 4a6ce2b6f2ec ("net: introduce a new function dst_dev_put()")
Signed-off-by: Eric Dumazet <edumazet@xxxxxxxxxx>
Reviewed-by: David Ahern <dsahern@xxxxxxxxxx>
Link: https://patch.msgid.link/20250828195823.3958522-9-edumazet@xxxxxxxxxx
Signed-off-by: Jakub Kicinski <kuba@xxxxxxxxxx>
[ minor modifications to fix conflict, added rcu_read_lock and unlock
to ip_defrag function ]
Signed-off-by: Miguel Gazquez (Schneider Electric) <miguel.gazquez@xxxxxxxxxxx>
---
 net/ipv4/icmp.c        | 6 +++---
 net/ipv4/ip_fragment.c | 9 +++++++--
 net/ipv4/ipmr.c        | 2 +-
 net/ipv4/route.c       | 4 ++--
 4 files changed, 13 insertions(+), 8 deletions(-)

diff --git a/net/ipv4/icmp.c b/net/ipv4/icmp.c
index 7763a98ddc96..e87325b80e0c 100644
--- a/net/ipv4/icmp.c
+++ b/net/ipv4/icmp.c
@@ -320,17 +320,17 @@ static bool icmpv4_xrlim_allow(struct net *net, struct rtable *rt,
 		return true;
 
 	/* No rate limit on loopback */
-	dev = dst_dev(dst);
+	rcu_read_lock();
+	dev = dst_dev_rcu(dst);
 	if (dev && (dev->flags & IFF_LOOPBACK))
 		goto out;
 
-	rcu_read_lock();
 	peer = inet_getpeer_v4(net->ipv4.peers, fl4->daddr,
 			       l3mdev_master_ifindex_rcu(dev));
 	rc = inet_peer_xrlim_allow(peer,
 				   READ_ONCE(net->ipv4.sysctl_icmp_ratelimit));
-	rcu_read_unlock();
 out:
+	rcu_read_unlock();
 	if (!rc)
 		__ICMP_INC_STATS(net, ICMP_MIB_RATELIMITHOST);
 	else
diff --git a/net/ipv4/ip_fragment.c b/net/ipv4/ip_fragment.c
index a70ede523297..43b1265f5ac1 100644
--- a/net/ipv4/ip_fragment.c
+++ b/net/ipv4/ip_fragment.c
@@ -481,13 +481,16 @@ static int ip_frag_reasm(struct ipq *qp, struct sk_buff *skb,
 /* Process an incoming IP datagram fragment. */
 int ip_defrag(struct net *net, struct sk_buff *skb, u32 user)
 {
-	struct net_device *dev = skb->dev ? : skb_dst_dev(skb);
-	int vif = l3mdev_master_ifindex_rcu(dev);
+	struct net_device *dev;
 	struct ipq *qp;
+	int vif;
 
 	__IP_INC_STATS(net, IPSTATS_MIB_REASMREQDS);
 
 	/* Lookup (or create) queue header */
+	rcu_read_lock();
+	dev = skb->dev ? : skb_dst_dev_rcu(skb);
+	vif = l3mdev_master_ifindex_rcu(dev);
 	qp = ip_find(net, ip_hdr(skb), user, vif);
 	if (qp) {
 		int ret;
@@ -497,9 +500,11 @@ int ip_defrag(struct net *net, struct sk_buff *skb, u32 user)
 		ret = ip_frag_queue(qp, skb);
 
 		spin_unlock(&qp->q.lock);
+		rcu_read_unlock();
 		ipq_put(qp);
 		return ret;
 	}
+	rcu_read_unlock();
 
 	__IP_INC_STATS(net, IPSTATS_MIB_REASMFAILS);
 	kfree_skb(skb);
diff --git a/net/ipv4/ipmr.c b/net/ipv4/ipmr.c
index f64651c0dea6..8d4f556fced2 100644
--- a/net/ipv4/ipmr.c
+++ b/net/ipv4/ipmr.c
@@ -1879,7 +1879,7 @@ static void ipmr_queue_xmit(struct net *net, struct mr_table *mrt,
 			goto out_free;
 	}
 
-	dev = rt->dst.dev;
+	dev = dst_dev_rcu(&rt->dst);
 
 	if (skb->len+encap > dst_mtu(&rt->dst) && (ntohs(iph->frag_off) & IP_DF)) {
 		/* Do not fragment multicasts. Alas, IPv4 does not
diff --git a/net/ipv4/route.c b/net/ipv4/route.c
index e3cce307e0e9..5528a5543df5 100644
--- a/net/ipv4/route.c
+++ b/net/ipv4/route.c
@@ -416,11 +416,11 @@ static struct neighbour *ipv4_neigh_lookup(const struct dst_entry *dst,
 					   const void *daddr)
 {
 	const struct rtable *rt = container_of(dst, struct rtable, dst);
-	struct net_device *dev = dst_dev(dst);
+	struct net_device *dev;
 	struct neighbour *n;
 
 	rcu_read_lock();
-
+	dev = dst_dev_rcu(dst);
 	if (likely(rt->rt_gw_family == AF_INET)) {
 		n = ip_neigh_gw4(dev, rt->rt_gw4);
 	} else if (rt->rt_gw_family == AF_INET6) {

-- 
2.55.0