Merge branch 'mptcp-userspace-pm-create-sockets-for-the-right-family'

Matthieu Baerts says:

====================
mptcp: userspace pm: create sockets for the right family

Before these patches, the Userspace Path Manager would allow the
creation of subflows with wrong families: taking the one of the MPTCP
socket instead of the provided ones and resulting in the creation of
subflows with likely not the right source and/or destination IPs. It
would also allow the creation of subflows between different families or
not respecting v4/v6-only socket attributes.

Patch 1 lets the userspace PM select the proper family to avoid creating
subflows with the wrong source and/or destination addresses because the
family is not the expected one.

Patch 2 makes sure the userspace PM doesn't allow the userspace to
create subflows for a family that is not allowed.

Patch 3 validates scenarios with a mix of v4 and v6 subflows for the
same MPTCP connection.

These patches fix issues introduced in v5.19 when the userspace path
manager has been introduced.
====================

Link: https://lore.kernel.org/r/20230112-upstream-net-20230112-netlink-v4-v6-v1-0-6a8363a221d2@tessares.net
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

6 files changed, 90 insertions, 6 deletions

diff --git a/net/mptcp/pm.c b/net/mptcp/pm.c
index 45e2a48397b9..70f0ced3ca86 100644
--- a/net/mptcp/pm.c
+++ b/net/mptcp/pm.c
@@ -420,6 +420,31 @@ void mptcp_pm_subflow_chk_stale(const struct mptcp_sock *msk, struct sock *ssk)
 	}
 }
 
+/* if sk is ipv4 or ipv6_only allows only same-family local and remote addresses,
+ * otherwise allow any matching local/remote pair
+ */
+bool mptcp_pm_addr_families_match(const struct sock *sk,
+				  const struct mptcp_addr_info *loc,
+				  const struct mptcp_addr_info *rem)
+{
+	bool mptcp_is_v4 = sk->sk_family == AF_INET;
+
+#if IS_ENABLED(CONFIG_MPTCP_IPV6)
+	bool loc_is_v4 = loc->family == AF_INET || ipv6_addr_v4mapped(&loc->addr6);
+	bool rem_is_v4 = rem->family == AF_INET || ipv6_addr_v4mapped(&rem->addr6);
+
+	if (mptcp_is_v4)
+		return loc_is_v4 && rem_is_v4;
+
+	if (ipv6_only_sock(sk))
+		return !loc_is_v4 && !rem_is_v4;
+
+	return loc_is_v4 == rem_is_v4;
+#else
+	return mptcp_is_v4 && loc->family == AF_INET && rem->family == AF_INET;
+#endif
+}
+
 void mptcp_pm_data_reset(struct mptcp_sock *msk)
 {
 	u8 pm_type = mptcp_get_pm_type(sock_net((struct sock *)msk));
diff --git a/net/mptcp/pm_userspace.c b/net/mptcp/pm_userspace.c
index 65dcc55a8ad8..ea6ad9da7493 100644
--- a/net/mptcp/pm_userspace.c
+++ b/net/mptcp/pm_userspace.c
@@ -294,6 +294,13 @@ int mptcp_nl_cmd_sf_create(struct sk_buff *skb, struct genl_info *info)
 	}
 
 	sk = (struct sock *)msk;
+
+	if (!mptcp_pm_addr_families_match(sk, &addr_l, &addr_r)) {
+		GENL_SET_ERR_MSG(info, "families mismatch");
+		err = -EINVAL;
+		goto create_err;
+	}
+
 	lock_sock(sk);
 
 	err = __mptcp_subflow_connect(sk, &addr_l, &addr_r);
diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c
index b7ad030dfe89..8cd6cc67c2c5 100644
--- a/net/mptcp/protocol.c
+++ b/net/mptcp/protocol.c
@@ -98,7 +98,7 @@ static int __mptcp_socket_create(struct mptcp_sock *msk)
 	struct socket *ssock;
 	int err;
 
-	err = mptcp_subflow_create_socket(sk, &ssock);
+	err = mptcp_subflow_create_socket(sk, sk->sk_family, &ssock);
 	if (err)
 		return err;
 
diff --git a/net/mptcp/protocol.h b/net/mptcp/protocol.h
index a0d1658ce59e..601469249da8 100644
--- a/net/mptcp/protocol.h
+++ b/net/mptcp/protocol.h
@@ -641,7 +641,8 @@ bool mptcp_addresses_equal(const struct mptcp_addr_info *a,
 /* called with sk socket lock held */
 int __mptcp_subflow_connect(struct sock *sk, const struct mptcp_addr_info *loc,
 			    const struct mptcp_addr_info *remote);
-int mptcp_subflow_create_socket(struct sock *sk, struct socket **new_sock);
+int mptcp_subflow_create_socket(struct sock *sk, unsigned short family,
+				struct socket **new_sock);
 void mptcp_info2sockaddr(const struct mptcp_addr_info *info,
 			 struct sockaddr_storage *addr,
 			 unsigned short family);
@@ -776,6 +777,9 @@ int mptcp_pm_parse_addr(struct nlattr *attr, struct genl_info *info,
 int mptcp_pm_parse_entry(struct nlattr *attr, struct genl_info *info,
 			 bool require_family,
 			 struct mptcp_pm_addr_entry *entry);
+bool mptcp_pm_addr_families_match(const struct sock *sk,
+				  const struct mptcp_addr_info *loc,
+				  const struct mptcp_addr_info *rem);
 void mptcp_pm_subflow_chk_stale(const struct mptcp_sock *msk, struct sock *ssk);
 void mptcp_pm_nl_subflow_chk_stale(const struct mptcp_sock *msk, struct sock *ssk);
 void mptcp_pm_new_connection(struct mptcp_sock *msk, const struct sock *ssk, int server_side);
diff --git a/net/mptcp/subflow.c b/net/mptcp/subflow.c
index bd387d4b5a38..ec54413fb31f 100644
--- a/net/mptcp/subflow.c
+++ b/net/mptcp/subflow.c
@@ -1547,7 +1547,7 @@ int __mptcp_subflow_connect(struct sock *sk, const struct mptcp_addr_info *loc,
 	if (!mptcp_is_fully_established(sk))
 		goto err_out;
 
-	err = mptcp_subflow_create_socket(sk, &sf);
+	err = mptcp_subflow_create_socket(sk, loc->family, &sf);
 	if (err)
 		goto err_out;
 
@@ -1660,7 +1660,9 @@ static void mptcp_subflow_ops_undo_override(struct sock *ssk)
 #endif
 		ssk->sk_prot = &tcp_prot;
 }
-int mptcp_subflow_create_socket(struct sock *sk, struct socket **new_sock)
+
+int mptcp_subflow_create_socket(struct sock *sk, unsigned short family,
+				struct socket **new_sock)
 {
 	struct mptcp_subflow_context *subflow;
 	struct net *net = sock_net(sk);
@@ -1673,8 +1675,7 @@ int mptcp_subflow_create_socket(struct sock *sk, struct socket **new_sock)
 	if (unlikely(!sk->sk_socket))
 		return -EINVAL;
 
-	err = sock_create_kern(net, sk->sk_family, SOCK_STREAM, IPPROTO_TCP,
-			       &sf);
+	err = sock_create_kern(net, family, SOCK_STREAM, IPPROTO_TCP, &sf);
 	if (err)
 		return err;
 
diff --git a/tools/testing/selftests/net/mptcp/userspace_pm.sh b/tools/testing/selftests/net/mptcp/userspace_pm.sh
index a29deb9fa024..ab2d581f28a1 100755
--- a/tools/testing/selftests/net/mptcp/userspace_pm.sh
+++ b/tools/testing/selftests/net/mptcp/userspace_pm.sh
@@ -752,6 +752,52 @@ test_subflows()
 	   "$server4_token" > /dev/null 2>&1
 }
 
+test_subflows_v4_v6_mix()
+{
+	# Attempt to add a listener at 10.0.2.1:<subflow-port>
+	ip netns exec "$ns1" ./pm_nl_ctl listen 10.0.2.1\
+	   $app6_port > /dev/null 2>&1 &
+	local listener_pid=$!
+
+	# ADD_ADDR4 from server to client machine reusing the subflow port on
+	# the established v6 connection
+	:>"$client_evts"
+	ip netns exec "$ns1" ./pm_nl_ctl ann 10.0.2.1 token "$server6_token" id\
+	   $server_addr_id dev ns1eth2 > /dev/null 2>&1
+	stdbuf -o0 -e0 printf "ADD_ADDR4 id:%d 10.0.2.1 (ns1) => ns2, reuse port\t\t" $server_addr_id
+	sleep 0.5
+	verify_announce_event "$client_evts" "$ANNOUNCED" "$client6_token" "10.0.2.1"\
+			      "$server_addr_id" "$app6_port"
+
+	# CREATE_SUBFLOW from client to server machine
+	:>"$client_evts"
+	ip netns exec "$ns2" ./pm_nl_ctl csf lip 10.0.2.2 lid 23 rip 10.0.2.1 rport\
+	   $app6_port token "$client6_token" > /dev/null 2>&1
+	sleep 0.5
+	verify_subflow_events "$client_evts" "$SUB_ESTABLISHED" "$client6_token"\
+			      "$AF_INET" "10.0.2.2" "10.0.2.1" "$app6_port" "23"\
+			      "$server_addr_id" "ns2" "ns1"
+
+	# Delete the listener from the server ns, if one was created
+	kill_wait $listener_pid
+
+	sport=$(sed --unbuffered -n 's/.*\(sport:\)\([[:digit:]]*\).*$/\2/p;q' "$client_evts")
+
+	# DESTROY_SUBFLOW from client to server machine
+	:>"$client_evts"
+	ip netns exec "$ns2" ./pm_nl_ctl dsf lip 10.0.2.2 lport "$sport" rip 10.0.2.1 rport\
+	   $app6_port token "$client6_token" > /dev/null 2>&1
+	sleep 0.5
+	verify_subflow_events "$client_evts" "$SUB_CLOSED" "$client6_token" \
+			      "$AF_INET" "10.0.2.2" "10.0.2.1" "$app6_port" "23"\
+			      "$server_addr_id" "ns2" "ns1"
+
+	# RM_ADDR from server to client machine
+	ip netns exec "$ns1" ./pm_nl_ctl rem id $server_addr_id token\
+	   "$server6_token" > /dev/null 2>&1
+	sleep 0.5
+}
+
 test_prio()
 {
 	local count
@@ -861,6 +907,7 @@ make_connection "v6"
 test_announce
 test_remove
 test_subflows
+test_subflows_v4_v6_mix
 test_prio
 test_listener