mptcp: tls support

ulp array mptcp tls fix Signed-off-by: Geliang Tang <[email protected]>
geliangtang · May 30, 2024 · 098253a · 098253a
1 parent 0e7c36c
commit 098253a
Show file tree

Hide file tree

Showing 23 changed files with 300 additions and 53 deletions.
diff --git a/include/net/espintcp.h b/include/net/espintcp.h
@@ -35,6 +35,6 @@ static inline struct espintcp_ctx *espintcp_getctx(const struct sock *sk)
 	const struct inet_connection_sock *icsk = inet_csk(sk);
 
 	/* RCU is only needed for diag */
-	return (__force void *)icsk->icsk_ulp_data;
+	return (__force void *)icsk->icsk_ulp_data[ULP_INDEX_DEFAULT];
 }
 #endif
diff --git a/include/net/inet_connection_sock.h b/include/net/inet_connection_sock.h
@@ -28,6 +28,12 @@ struct inet_bind_bucket;
 struct inet_bind2_bucket;
 struct tcp_congestion_ops;
 
+enum ulp_index {
+	ULP_INDEX_DEFAULT,
+	ULP_INDEX_MPTCP,
+	ULP_INDEX_MAX,
+};
+
 /*
  * Pointers to address related TCP functions
  * (i.e. things that depend on the address family)
@@ -94,8 +100,8 @@ struct inet_connection_sock {
 	__u32			  icsk_pmtu_cookie;
 	const struct tcp_congestion_ops *icsk_ca_ops;
 	const struct inet_connection_sock_af_ops *icsk_af_ops;
-	const struct tcp_ulp_ops  *icsk_ulp_ops;
-	void __rcu		  *icsk_ulp_data;
+	const struct tcp_ulp_ops  *icsk_ulp_ops[ULP_INDEX_MAX];
+	void __rcu		  *icsk_ulp_data[ULP_INDEX_MAX];
 	void (*icsk_clean_acked)(struct sock *sk, u32 acked_seq);
 	unsigned int		  (*icsk_sync_mss)(struct sock *sk, u32 pmtu);
 	__u8			  icsk_ca_state:5,
@@ -352,7 +358,8 @@ static inline void inet_csk_inc_pingpong_cnt(struct sock *sk)
 
 static inline bool inet_csk_has_ulp(const struct sock *sk)
 {
-	return inet_test_bit(IS_ICSK, sk) && !!inet_csk(sk)->icsk_ulp_ops;
+	return inet_test_bit(IS_ICSK, sk) && (!!inet_csk(sk)->icsk_ulp_ops[ULP_INDEX_DEFAULT] ||
+					      !!inet_csk(sk)->icsk_ulp_ops[ULP_INDEX_MPTCP]);
 }
 
 static inline void inet_init_csk_locks(struct sock *sk)

diff --git a/include/net/tcp.h b/include/net/tcp.h
@@ -2543,6 +2543,7 @@ enum hrtimer_restart tcp_pace_kick(struct hrtimer *timer);
 
 struct tcp_ulp_ops {
 	struct list_head	list;
+	int			id;
 
 	/* initialize ulp */
 	int (*init)(struct sock *sk);

diff --git a/include/net/tls.h b/include/net/tls.h
@@ -367,7 +367,7 @@ static inline struct tls_context *tls_get_ctx(const struct sock *sk)
 	/* Use RCU on icsk_ulp_data only for sock diag code,
 	 * TLS data path doesn't need rcu_dereference().
 	 */
-	return (__force void *)icsk->icsk_ulp_data;
+	return (__force void *)icsk->icsk_ulp_data[ULP_INDEX_DEFAULT];
 }
 
 static inline struct tls_sw_context_rx *tls_sw_ctx_rx(

diff --git a/net/ipv4/inet_connection_sock.c b/net/ipv4/inet_connection_sock.c
@@ -1148,11 +1148,14 @@ static void inet_clone_ulp(const struct request_sock *req, struct sock *newsk,
 			   const gfp_t priority)
 {
 	struct inet_connection_sock *icsk = inet_csk(newsk);
+	const struct tcp_ulp_ops *ulp_ops;
+	int i;
 
-	if (!icsk->icsk_ulp_ops)
-		return;
-
-	icsk->icsk_ulp_ops->clone(req, newsk, priority);
+	for (i = 0; i < ULP_INDEX_MAX; i++) {
+		ulp_ops = icsk->icsk_ulp_ops[i];
+		if (ulp_ops && ulp_ops->clone)
+			ulp_ops->clone(req, newsk, priority);
+	}
 }
 
 /**
@@ -1251,9 +1254,14 @@ EXPORT_SYMBOL(inet_csk_prepare_forced_close);
 static int inet_ulp_can_listen(const struct sock *sk)
 {
 	const struct inet_connection_sock *icsk = inet_csk(sk);
+	const struct tcp_ulp_ops *ulp_ops;
+	int i;
 
-	if (icsk->icsk_ulp_ops && !icsk->icsk_ulp_ops->clone)
-		return -EINVAL;
+	for (i = 0; i < ULP_INDEX_MAX; i++) {
+		ulp_ops = icsk->icsk_ulp_ops[i];
+		if (ulp_ops && !ulp_ops->clone)
+			return -EINVAL;
+	}
 
 	return 0;
 }

diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
@@ -4133,15 +4133,15 @@ int do_tcp_getsockopt(struct sock *sk, int level,
 		if (copy_from_sockptr(&len, optlen, sizeof(int)))
 			return -EFAULT;
 		len = min_t(unsigned int, len, TCP_ULP_NAME_MAX);
-		if (!icsk->icsk_ulp_ops) {
+		if (!icsk->icsk_ulp_ops[ULP_INDEX_DEFAULT]) {
 			len = 0;
 			if (copy_to_sockptr(optlen, &len, sizeof(int)))
 				return -EFAULT;
 			return 0;
 		}
 		if (copy_to_sockptr(optlen, &len, sizeof(int)))
 			return -EFAULT;
-		if (copy_to_sockptr(optval, icsk->icsk_ulp_ops->name, len))
+		if (copy_to_sockptr(optval, icsk->icsk_ulp_ops[ULP_INDEX_DEFAULT]->name, len))
 			return -EFAULT;
 		return 0;
 

diff --git a/net/ipv4/tcp_diag.c b/net/ipv4/tcp_diag.c
@@ -132,7 +132,7 @@ static int tcp_diag_get_aux(struct sock *sk, bool net_admin,
 	if (net_admin) {
 		const struct tcp_ulp_ops *ulp_ops;
 
-		ulp_ops = icsk->icsk_ulp_ops;
+		ulp_ops = icsk->icsk_ulp_ops[ULP_INDEX_DEFAULT];
 		if (ulp_ops)
 			err = tcp_diag_put_ulp(skb, sk, ulp_ops);
 		if (err)
@@ -167,7 +167,7 @@ static size_t tcp_diag_get_aux_size(struct sock *sk, bool net_admin)
 	if (net_admin && sk_fullsock(sk)) {
 		const struct tcp_ulp_ops *ulp_ops;
 
-		ulp_ops = icsk->icsk_ulp_ops;
+		ulp_ops = icsk->icsk_ulp_ops[ULP_INDEX_DEFAULT];
 		if (ulp_ops) {
 			size += nla_total_size(0) +
 				nla_total_size(TCP_ULP_NAME_MAX);

diff --git a/net/ipv4/tcp_ulp.c b/net/ipv4/tcp_ulp.c
@@ -104,27 +104,36 @@ void tcp_update_ulp(struct sock *sk, struct proto *proto,
 		    void (*write_space)(struct sock *sk))
 {
 	struct inet_connection_sock *icsk = inet_csk(sk);
+	const struct tcp_ulp_ops *ulp_ops;
+	int i;
 
-	if (icsk->icsk_ulp_ops->update)
-		icsk->icsk_ulp_ops->update(sk, proto, write_space);
+	for (i = 0; i < ULP_INDEX_MAX; i++) {
+		ulp_ops = icsk->icsk_ulp_ops[i];
+		if (ulp_ops->update)
+			ulp_ops->update(sk, proto, write_space);
+	}
 }
 
 void tcp_cleanup_ulp(struct sock *sk)
 {
 	struct inet_connection_sock *icsk = inet_csk(sk);
+	const struct tcp_ulp_ops *ulp_ops;
+	int i;
 
 	/* No sock_owned_by_me() check here as at the time the
 	 * stack calls this function, the socket is dead and
 	 * about to be destroyed.
 	 */
-	if (!icsk->icsk_ulp_ops)
-		return;
-
-	if (icsk->icsk_ulp_ops->release)
-		icsk->icsk_ulp_ops->release(sk);
-	module_put(icsk->icsk_ulp_ops->owner);
-
-	icsk->icsk_ulp_ops = NULL;
+	for (i = 0; i < ULP_INDEX_MAX; i++) {
+		ulp_ops = icsk->icsk_ulp_ops[i];
+		if (ulp_ops) {
+			//pr_info("%s ulp_ops->name=%s\n", __func__, ulp_ops->name);
+			if (ulp_ops->release)
+				ulp_ops->release(sk);
+			module_put(ulp_ops->owner);
+			ulp_ops = NULL;
+		}
+	}
 }
 
 static int __tcp_set_ulp(struct sock *sk, const struct tcp_ulp_ops *ulp_ops)
@@ -133,7 +142,7 @@ static int __tcp_set_ulp(struct sock *sk, const struct tcp_ulp_ops *ulp_ops)
 	int err;
 
 	err = -EEXIST;
-	if (icsk->icsk_ulp_ops)
+	if (icsk->icsk_ulp_ops[ulp_ops->id])
 		goto out_err;
 
 	if (sk->sk_socket)
@@ -147,7 +156,7 @@ static int __tcp_set_ulp(struct sock *sk, const struct tcp_ulp_ops *ulp_ops)
 	if (err)
 		goto out_err;
 
-	icsk->icsk_ulp_ops = ulp_ops;
+	icsk->icsk_ulp_ops[ulp_ops->id] = ulp_ops;
 	return 0;
 out_err:
 	module_put(ulp_ops->owner);

diff --git a/net/mptcp/diag.c b/net/mptcp/diag.c
@@ -29,7 +29,7 @@ static int subflow_get_info(struct sock *sk, struct sk_buff *skb)
 
 	slow = lock_sock_fast(sk);
 	rcu_read_lock();
-	sf = rcu_dereference(inet_csk(sk)->icsk_ulp_data);
+	sf = rcu_dereference(inet_csk(sk)->icsk_ulp_data[ULP_INDEX_MPTCP]);
 	if (!sf) {
 		err = 0;
 		goto nla_failure;

diff --git a/net/mptcp/mptcp_diag.c b/net/mptcp/mptcp_diag.c
@@ -103,7 +103,7 @@ static void mptcp_diag_dump_listeners(struct sk_buff *skb, struct netlink_callba
 			if (num < diag_ctx->l_num)
 				goto next_listen;
 
-			if (!ctx || strcmp(inet_csk(sk)->icsk_ulp_ops->name, "mptcp"))
+			if (!ctx || strcmp(inet_csk(sk)->icsk_ulp_ops[ULP_INDEX_MPTCP]->name, "mptcp"))
 				goto next_listen;
 
 			sk = ctx->conn;

diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c
@@ -2462,7 +2462,7 @@ static void __mptcp_close_ssk(struct sock *sk, struct sock *ssk,
 	 * the ssk has been already destroyed, we just need to release the
 	 * reference owned by msk;
 	 */
-	if (!inet_csk(ssk)->icsk_ulp_ops) {
+	if (!inet_csk(ssk)->icsk_ulp_ops[ULP_INDEX_MPTCP]) {
 		WARN_ON_ONCE(!sock_flag(ssk, SOCK_DEAD));
 		kfree_rcu(subflow, rcu);
 	} else {

diff --git a/net/mptcp/protocol.h b/net/mptcp/protocol.h
@@ -14,6 +14,7 @@
 #include <uapi/linux/mptcp.h>
 #include <net/genetlink.h>
 #include <net/rstreason.h>
+#include <net/tls.h>
 
 #define MPTCP_SUPPORTED_VERSION	1
 
@@ -499,6 +500,7 @@ DECLARE_PER_CPU(struct mptcp_delegated_action, mptcp_delegated_actions);
 struct mptcp_subflow_context {
 	struct	list_head node;/* conn_list of subflows */
 
+	struct tls_context tls;
 	struct_group(reset,
 
 	unsigned long avg_pacing_rate; /* protected by msk socket lock */
@@ -584,7 +586,7 @@ mptcp_subflow_ctx(const struct sock *sk)
 	const struct inet_connection_sock *icsk = inet_csk(sk);
 
 	/* Use RCU on icsk_ulp_data only for sock diag code */
-	return (__force struct mptcp_subflow_context *)icsk->icsk_ulp_data;
+	return (__force struct mptcp_subflow_context *)icsk->icsk_ulp_data[ULP_INDEX_MPTCP];
 }
 
 static inline struct sock *

diff --git a/net/mptcp/sockopt.c b/net/mptcp/sockopt.c
@@ -567,6 +567,7 @@ static bool mptcp_supported_sockopt(int level, int optname)
 		case TCP_FASTOPEN_CONNECT:
 		case TCP_FASTOPEN_KEY:
 		case TCP_FASTOPEN_NO_COOKIE:
+		case TCP_ULP:
 			return true;
 		}
 
@@ -576,9 +577,54 @@ static bool mptcp_supported_sockopt(int level, int optname)
 		 * TCP_REPAIR_WINDOW are not supported, better avoid this mess
 		 */
 	}
+	if (level == SOL_TLS) {
+		switch (optname) {
+		case TLS_TX:
+		case TLS_RX:
+			return true;
+		}
+	}
 	return false;
 }
 
+static int mptcp_setsockopt_sol_tcp_ulp(struct mptcp_sock *msk, sockptr_t optval,
+					unsigned int optlen)
+{
+	struct mptcp_subflow_context *subflow;
+	struct sock *sk = (struct sock *)msk;
+	char name[TCP_ULP_NAME_MAX];
+	int ret;
+
+	if (optlen < 1)
+		return -EINVAL;
+
+	ret = strncpy_from_sockptr(name, optval,
+				   min_t(long, TCP_ULP_NAME_MAX - 1, optlen));
+	if (ret < 0)
+		return -EFAULT;
+	name[ret] = 0;
+
+	ret = 0;
+	lock_sock(sk);
+	sockopt_seq_inc(msk);
+	mptcp_for_each_subflow(msk, subflow) {
+		struct sock *ssk = mptcp_subflow_tcp_sock(subflow);
+		int err;
+
+		lock_sock(ssk);
+		err = tcp_set_ulp(ssk, name);
+		if (err < 0 && ret == 0) {
+			pr_info("%s err=%d\n", __func__, err);
+			ret = err;
+		}
+		subflow->setsockopt_seq = msk->setsockopt_seq;
+		release_sock(ssk);
+	}
+
+	release_sock(sk);
+	return ret;
+}
+
 static int mptcp_setsockopt_sol_tcp_congestion(struct mptcp_sock *msk, sockptr_t optval,
 					       unsigned int optlen)
 {
@@ -806,7 +852,7 @@ static int mptcp_setsockopt_sol_tcp(struct mptcp_sock *msk, int optname,
 
 	switch (optname) {
 	case TCP_ULP:
-		return -EOPNOTSUPP;
+		return mptcp_setsockopt_sol_tcp_ulp(msk, optval, optlen);
 	case TCP_CONGESTION:
 		return mptcp_setsockopt_sol_tcp_congestion(msk, optval, optlen);
 	case TCP_DEFER_ACCEPT:
@@ -867,6 +913,8 @@ static int mptcp_setsockopt_sol_tcp(struct mptcp_sock *msk, int optname,
 	return ret;
 }
 
+int tls_setsockopt(struct sock *sk, int level, int optname,
+		   sockptr_t optval, unsigned int optlen);
 int mptcp_setsockopt(struct sock *sk, int level, int optname,
 		     sockptr_t optval, unsigned int optlen)
 {
@@ -902,6 +950,9 @@ int mptcp_setsockopt(struct sock *sk, int level, int optname,
 	if (level == SOL_TCP)
 		return mptcp_setsockopt_sol_tcp(msk, optname, optval, optlen);
 
+	if (level == SOL_TLS)
+		return tls_setsockopt(msk->first, level, optname, optval, optlen);
+
 	return -EOPNOTSUPP;
 }
 

diff --git a/net/mptcp/subflow.c b/net/mptcp/subflow.c
@@ -747,8 +747,8 @@ static void subflow_ulp_fallback(struct sock *sk,
 	struct inet_connection_sock *icsk = inet_csk(sk);
 
 	mptcp_subflow_tcp_fallback(sk, old_ctx);
-	icsk->icsk_ulp_ops = NULL;
-	rcu_assign_pointer(icsk->icsk_ulp_data, NULL);
+	icsk->icsk_ulp_ops[ULP_INDEX_MPTCP] = NULL;
+	rcu_assign_pointer(icsk->icsk_ulp_data[ULP_INDEX_MPTCP], NULL);
 	tcp_sk(sk)->is_mptcp = 0;
 
 	mptcp_subflow_ops_undo_override(sk);
@@ -762,7 +762,7 @@ void mptcp_subflow_drop_ctx(struct sock *ssk)
 		return;
 
 	list_del(&mptcp_subflow_ctx(ssk)->node);
-	if (inet_csk(ssk)->icsk_ulp_ops) {
+	if (inet_csk(ssk)->icsk_ulp_ops[ULP_INDEX_MPTCP]) {
 		subflow_ulp_fallback(ssk, ctx);
 		if (ctx->conn)
 			sock_put(ctx->conn);
@@ -1755,7 +1755,7 @@ static struct mptcp_subflow_context *subflow_create_ctx(struct sock *sk,
 	if (!ctx)
 		return NULL;
 
-	rcu_assign_pointer(icsk->icsk_ulp_data, ctx);
+	rcu_assign_pointer(icsk->icsk_ulp_data[ULP_INDEX_MPTCP], ctx);
 	INIT_LIST_HEAD(&ctx->node);
 	INIT_LIST_HEAD(&ctx->delegated_node);
 
@@ -2044,6 +2044,7 @@ static int tcp_abort_override(struct sock *ssk, int err)
 }
 
 static struct tcp_ulp_ops subflow_ulp_ops __read_mostly = {
+	.id		= ULP_INDEX_MPTCP,
 	.name		= "mptcp",
 	.owner		= THIS_MODULE,
 	.init		= subflow_ulp_init,

diff --git a/net/mptcp/token_test.c b/net/mptcp/token_test.c
@@ -76,7 +76,7 @@ static void mptcp_token_test_msk_basic(struct kunit *test)
 	struct mptcp_sock *null_msk = NULL;
 	struct sock *sk;
 
-	rcu_assign_pointer(icsk->icsk_ulp_data, ctx);
+	rcu_assign_pointer(icsk->icsk_ulp_data[ULP_INDEX_MPTCP], ctx);
 	ctx->conn = (struct sock *)msk;
 	sk = (struct sock *)msk;
 

diff --git a/net/smc/af_smc.c b/net/smc/af_smc.c
@@ -3439,7 +3439,7 @@ static void smc_ulp_clone(const struct request_sock *req, struct sock *newsk,
 	struct inet_connection_sock *icsk = inet_csk(newsk);
 
 	/* don't inherit ulp ops to child when listen */
-	icsk->icsk_ulp_ops = NULL;
+	icsk->icsk_ulp_ops[ULP_INDEX_DEFAULT] = NULL;
 }
 
 static struct tcp_ulp_ops smc_ulp_ops __read_mostly = {