Skip to contentSkip to conversation

MPTCP KTLS support

RFC Superseded Geliang Tang 4 participants 16 messages Dec 12, 2025 – Dec 18, 2025

Subsystems (4)

v4 of 25 Superseded Latest is v25 →

lore
Share & tools
lore

Geliang Tang

author reviewer
#0 lore raw

v4 cover MPTCP KTLS support

From: Geliang Tang <[email protected]>

v4:
- split "tls: add MPTCP protocol support" into smaller, more focused patches.
- a new mptcp_inq helper has been implemented instead of directly using mptcp_inq_hint to fix the issue mentioned in [1].
- add sk_is_msk helper.
- the 'expect' parameter will no longer be added to sock_test_tcpulp. Instead, SOCK_TEST_TCPULP items causing the tests failure will be directly removed.
- remove the "TCP KTLS" tests, keeping only the MPTCP-related ones.

[1]
https://patchwork.kernel.org/project/mptcp/patch/ce74452f4c095a1761ef493b767b4bd9f9c14359.1764333805.git.tanggeliang@kylinos.cn/

v3:
- mptcp_read_sock() and mptcp_poll() are not exported, as mptcp_sockopt
test does not use read_sock/poll interfaces. They will be exported when
new tests are added in the future.
- call mptcp_inq_hint in tls_device_rx_resync_new_rec(), tls_device_core_ctrl_rx_resync() and tls_read_flush_backlog() too.
- update selftests.
- Link: https://patchwork.kernel.org/project/mptcp/cover/[email protected]/

v2:
- fix disconnect.
- update selftests.

This series adds KTLS support for MPTCP. Since the ULP of msk is not being used, ULP KTLS can be directly configured onto msk without affecting its communication.

Geliang Tang (10):
mptcp: add sk_is_msk helper
tls: switch to MPTCP_SKB_CB
tls: switch to mptcp_inq
tls: switch to mptcp_sendmsg_locked
tls: switch to mptcp_recv_skb
tls: switch to mptcp_read_done
mptcp: update ULP getsockopt
mptcp: enable TLS setsockopt
selftests: mptcp: connect: update sock_test_tcpulp
selftests: mptcp: sockopt: implement MPTCP KTLS tests

 include/net/mptcp.h                           | 48 ++++++++++++
 net/mptcp/protocol.c                          | 66 ++++++++++++++--
 net/mptcp/protocol.h                          | 10 ---
 net/mptcp/sockopt.c                           | 30 +++++++-
 net/tls/tls_main.c                            |  4 +-
 net/tls/tls_strp.c                            | 24 ++++--
 net/tls/tls_sw.c                              |  8 +-
 tools/testing/selftests/net/mptcp/config      |  1 +
 .../selftests/net/mptcp/mptcp_connect.c       | 20 +++--
 .../selftests/net/mptcp/mptcp_sockopt.c       | 76 ++++++++++++++++++-
 .../selftests/net/mptcp/mptcp_sockopt.sh      | 36 +++++++++
 11 files changed, 289 insertions(+), 34 deletions(-)
Signature · 1 line
2.51.0
author reviewer
#1 lore raw

RFC v4 1/10 mptcp: add sk_is_msk helper

From: Geliang Tang <[email protected]>

This patch introduces a sk_is_msk() helper modeled after sk_is_tcp() to determine whether the socket is an MPTCP one. Unlike sk_is_mptcp(), which accepts a subflow socket as its parameter, this new helper specifically accepts an MPTCP socket parameter.

---
 include/net/mptcp.h | 12 ++++++++++++
 1 file changed, 12 insertions(+)
include/net/mptcp.h +12 -0
index 4cf59e83c1c5..82660374859a 100644
@@ -150,6 +150,13 @@ static inline bool rsk_drop_req(const struct request_sock *req)
 	return tcp_rsk(req)->is_mptcp && tcp_rsk(req)->drop_req;
 }
 
+static inline bool sk_is_msk(const struct sock *sk)
+{
+	return sk_is_inet(sk) &&
+	       sk->sk_type == SOCK_STREAM &&
+	       sk->sk_protocol == IPPROTO_MPTCP;
+}
+
 void mptcp_space(const struct sock *ssk, int *space, int *full_space);
 bool mptcp_syn_options(struct sock *sk, const struct sk_buff *skb,
 		       unsigned int *size, struct mptcp_out_options *opts);
@@ -258,6 +265,11 @@ static inline bool rsk_drop_req(const struct request_sock *req)
 	return false;
 }
 
+static inline bool sk_is_msk(const struct sock *sk)
+{
+	return false;
+}
+
 static inline bool mptcp_syn_options(struct sock *sk, const struct sk_buff *skb,
 				     unsigned int *size,
 				     struct mptcp_out_options *opts)
Signature · 1 line
2.51.0
author reviewer
#2 lore raw

RFC v4 2/10 tls: switch to MPTCP_SKB_CB

In reply to Geliang Tang

From: Geliang Tang <[email protected]>

To extend MPTCP support based on TCP TLS, corresponding MPTCP-specific macro MPTCP_SKB_CB() has been exported into net/mptcp.h.

TLS implementation switches between the respective TCP and MPTCP macros based on the detected protocol.

---
 include/net/mptcp.h  | 10 ++++++++++
 net/mptcp/protocol.h | 10 ----------
 net/tls/tls_strp.c   |  8 ++++++--
 net/tls/tls_sw.c     |  4 +++-
 4 files changed, 19 insertions(+), 13 deletions(-)
include/net/mptcp.h +10 -0
index 82660374859a..050c2e820fc4 100644
@@ -132,6 +132,16 @@ struct mptcp_pm_ops {
 	void (*release)(struct mptcp_sock *msk);
 } ____cacheline_aligned_in_smp;
 
+struct mptcp_skb_cb {
+	u64 map_seq;
+	u64 end_seq;
+	u32 offset;
+	u8  has_rxtstamp;
+	u8  cant_coalesce;
+};
+
+#define MPTCP_SKB_CB(__skb)	((struct mptcp_skb_cb *)&((__skb)->cb[0]))
+
 #ifdef CONFIG_MPTCP
 void mptcp_init(void);
 
net/mptcp/protocol.h +0 -10
index cd5266099993..75f2fd05ece4 100644
@@ -125,16 +125,6 @@
 #define MPTCP_SYNC_STATE	6
 #define MPTCP_SYNC_SNDBUF	7
 
-struct mptcp_skb_cb {
-	u64 map_seq;
-	u64 end_seq;
-	u32 offset;
-	u8  has_rxtstamp;
-	u8  cant_coalesce;
-};
-
-#define MPTCP_SKB_CB(__skb)	((struct mptcp_skb_cb *)&((__skb)->cb[0]))
-
 static inline bool before64(__u64 seq1, __u64 seq2)
 {
 	return (__s64)(seq1 - seq2) < 0;
net/tls/tls_strp.c +6 -2
index 98e12f0ff57e..cdbcb07a52f8 100644
@@ -439,7 +439,9 @@ static bool tls_strp_check_queue_ok(struct tls_strparser *strp)
 
 	first = skb_shinfo(strp->anchor)->frag_list;
 	skb = first;
-	seq = TCP_SKB_CB(first)->seq;
+	seq = sk_is_msk(strp->sk) ?
+	      MPTCP_SKB_CB(first)->map_seq :
+	      TCP_SKB_CB(first)->seq;
 
 	/* Make sure there's no duplicate data in the queue,
 	 * and the decrypted status matches.
@@ -449,7 +451,9 @@ static bool tls_strp_check_queue_ok(struct tls_strparser *strp)
 		len -= skb->len;
 		skb = skb->next;
 
-		if (TCP_SKB_CB(skb)->seq != seq)
+		if ((sk_is_msk(strp->sk) ?
+		     MPTCP_SKB_CB(skb)->map_seq :
+		     TCP_SKB_CB(skb)->seq) != seq)
 			return false;
 		if (skb_cmp_decrypted(first, skb))
 			return false;
net/tls/tls_sw.c +3 -1
index 9937d4c810f2..c35c25abc40c 100644
@@ -2489,7 +2489,9 @@ int tls_rx_msg_size(struct tls_strparser *strp, struct sk_buff *skb)
 	}
 
 	tls_device_rx_resync_new_rec(strp->sk, data_len + TLS_HEADER_SIZE,
-				     TCP_SKB_CB(skb)->seq + strp->stm.offset);
+				     (sk_is_msk(strp->sk) ?
+				      MPTCP_SKB_CB(skb)->map_seq :
+				      TCP_SKB_CB(skb)->seq) + strp->stm.offset);
 	return data_len + TLS_HEADER_SIZE;
 
 read_failure:
Signature · 1 line
2.51.0
author reviewer
#3 lore raw

RFC v4 3/10 tls: switch to mptcp_inq

In reply to Geliang Tang

From: Geliang Tang <[email protected]>

To extend MPTCP support based on TCP TLS, corresponding MPTCP-specific helper mptcp_inq() has been exported into net/mptcp.h.

Note that there is a slight difference between mptcp_inq() and mptcp_inq_hint(), it does not return 1 when the socket is closed or shut down; instead, it returns 0. Otherwise, it would break the condition "inq < 1" in tls_strp_read_sock().

TLS implementation switches between the respective TCP and MPTCP helpers based on the detected protocol.

---
 include/net/mptcp.h  |  7 +++++++
 net/mptcp/protocol.c | 12 +++++++++++-
 net/tls/tls_strp.c   |  8 ++++++--
 net/tls/tls_sw.c     |  4 +++-
 4 files changed, 27 insertions(+), 4 deletions(-)
include/net/mptcp.h +7 -0
index 050c2e820fc4..4c8e483fcc4b 100644
@@ -254,6 +254,8 @@ static inline __be32 mptcp_reset_option(const struct sk_buff *skb)
 }
 
 void mptcp_active_detect_blackhole(struct sock *sk, bool expired);
+
+unsigned int mptcp_inq(const struct sock *sk);
 #else
 
 static inline void mptcp_init(void)
@@ -345,6 +347,11 @@ static inline struct request_sock *mptcp_subflow_reqsk_alloc(const struct reques
 static inline __be32 mptcp_reset_option(const struct sk_buff *skb)  { return htonl(0u); }
 
 static inline void mptcp_active_detect_blackhole(struct sock *sk, bool expired) { }
+
+static inline unsigned int mptcp_inq(const struct sock *sk)
+{
+	return 0;
+}
 #endif /* CONFIG_MPTCP */
 
 #if IS_ENABLED(CONFIG_MPTCP_IPV6)
net/mptcp/protocol.c +11 -1
index 334fdb10fdf3..82a3b3c70ad1 100644
@@ -2242,7 +2242,7 @@ static bool mptcp_move_skbs(struct sock *sk)
 	return enqueued;
 }
 
-static unsigned int mptcp_inq_hint(const struct sock *sk)
+unsigned int mptcp_inq(const struct sock *sk)
 {
 	const struct mptcp_sock *msk = mptcp_sk(sk);
 	const struct sk_buff *skb;
@@ -2257,6 +2257,16 @@ static unsigned int mptcp_inq_hint(const struct sock *sk)
 		return (unsigned int)hint_val;
 	}
 
+	return 0;
+}
+
+static unsigned int mptcp_inq_hint(const struct sock *sk)
+{
+	unsigned int inq = mptcp_inq(sk);
+
+	if (inq)
+		return inq;
+
 	if (sk->sk_state == TCP_CLOSE || (sk->sk_shutdown & RCV_SHUTDOWN))
 		return 1;
 
net/tls/tls_strp.c +6 -2
index cdbcb07a52f8..0639130b8d5e 100644
@@ -494,7 +494,9 @@ bool tls_strp_msg_load(struct tls_strparser *strp, bool force_refresh)
 	DEBUG_NET_WARN_ON_ONCE(!strp->stm.full_len);
 
 	if (!strp->copy_mode && force_refresh) {
-		if (unlikely(tcp_inq(strp->sk) < strp->stm.full_len)) {
+		if (unlikely((sk_is_msk(strp->sk) ?
+			      mptcp_inq(strp->sk) :
+			      tcp_inq(strp->sk)) < strp->stm.full_len)) {
 			WRITE_ONCE(strp->msg_ready, 0);
 			memset(&strp->stm, 0, sizeof(strp->stm));
 			return false;
@@ -517,7 +519,9 @@ static int tls_strp_read_sock(struct tls_strparser *strp)
 {
 	int sz, inq;
 
-	inq = tcp_inq(strp->sk);
+	inq = sk_is_msk(strp->sk) ?
+	      mptcp_inq(strp->sk) :
+	      tcp_inq(strp->sk);
 	if (inq < 1)
 		return 0;
 
net/tls/tls_sw.c +3 -1
index c35c25abc40c..30ca332afe41 100644
@@ -1958,7 +1958,9 @@ tls_read_flush_backlog(struct sock *sk, struct tls_prot_info *prot,
 		return false;
 
 	max_rec = prot->overhead_size - prot->tail_size + TLS_MAX_PAYLOAD_SIZE;
-	if (done - *flushed_at < SZ_128K && tcp_inq(sk) > max_rec)
+	if (done - *flushed_at < SZ_128K && (sk_is_msk(sk) ?
+					     mptcp_inq(sk) :
+					     tcp_inq(sk)) > max_rec)
 		return false;
 
 	*flushed_at = done;
Signature · 1 line
2.51.0
author reviewer
#4 lore raw

RFC v4 4/10 tls: switch to mptcp_sendmsg_locked

In reply to Geliang Tang

From: Geliang Tang <[email protected]>

To extend MPTCP support based on TCP TLS, corresponding MPTCP-specific helper mptcp_sendmsg_locked() has been exported into net/mptcp.h.

TLS implementation switches between the respective TCP and MPTCP helpers based on the detected protocol.

---
 include/net/mptcp.h  |  8 ++++++++
 net/mptcp/protocol.c | 16 ++++++++++++----
 net/tls/tls_main.c   |  4 +++-
 3 files changed, 23 insertions(+), 5 deletions(-)
include/net/mptcp.h +8 -0
index 4c8e483fcc4b..cdb726865486 100644
@@ -256,6 +256,8 @@ static inline __be32 mptcp_reset_option(const struct sk_buff *skb)
 void mptcp_active_detect_blackhole(struct sock *sk, bool expired);
 
 unsigned int mptcp_inq(const struct sock *sk);
+
+int mptcp_sendmsg_locked(struct sock *sk, struct msghdr *msg, size_t len);
 #else
 
 static inline void mptcp_init(void)
@@ -352,6 +354,12 @@ static inline unsigned int mptcp_inq(const struct sock *sk)
 {
 	return 0;
 }
+
+static inline int mptcp_sendmsg_locked(struct sock *sk, struct msghdr *msg,
+				       size_t len)
+{
+	return 0;
+}
 #endif /* CONFIG_MPTCP */
 
 #if IS_ENABLED(CONFIG_MPTCP_IPV6)
net/mptcp/protocol.c +12 -4
index 82a3b3c70ad1..feac59a92b32 100644
@@ -1884,7 +1884,7 @@ static void mptcp_rps_record_subflows(const struct mptcp_sock *msk)
 	}
 }
 
-static int mptcp_sendmsg(struct sock *sk, struct msghdr *msg, size_t len)
+int mptcp_sendmsg_locked(struct sock *sk, struct msghdr *msg, size_t len)
 {
 	struct mptcp_sock *msk = mptcp_sk(sk);
 	struct page_frag *pfrag;
@@ -1895,8 +1895,6 @@ static int mptcp_sendmsg(struct sock *sk, struct msghdr *msg, size_t len)
 	/* silently ignore everything else */
 	msg->msg_flags &= MSG_MORE | MSG_DONTWAIT | MSG_NOSIGNAL | MSG_FASTOPEN;
 
-	lock_sock(sk);
-
 	mptcp_rps_record_subflows(msk);
 
 	if (unlikely(inet_test_bit(DEFER_CONNECT, sk) ||
@@ -2004,7 +2002,6 @@ static int mptcp_sendmsg(struct sock *sk, struct msghdr *msg, size_t len)
 		__mptcp_push_pending(sk, msg->msg_flags);
 
 out:
-	release_sock(sk);
 	return copied;
 
 do_error:
@@ -2015,6 +2012,17 @@ static int mptcp_sendmsg(struct sock *sk, struct msghdr *msg, size_t len)
 	goto out;
 }
 
+static int mptcp_sendmsg(struct sock *sk, struct msghdr *msg, size_t len)
+{
+	int ret;
+
+	lock_sock(sk);
+	ret = mptcp_sendmsg_locked(sk, msg, len);
+	release_sock(sk);
+
+	return ret;
+}
+
 static void mptcp_rcv_space_adjust(struct mptcp_sock *msk, int copied);
 
 static void mptcp_eat_recv_skb(struct sock *sk, struct sk_buff *skb)
net/tls/tls_main.c +3 -1
index 56ce0bc8317b..18c767b7388d 100644
@@ -194,7 +194,9 @@ int tls_push_sg(struct sock *sk,
 		bvec_set_page(&bvec, p, size, offset);
 		iov_iter_bvec(&msg.msg_iter, ITER_SOURCE, &bvec, 1, size);
 
-		ret = tcp_sendmsg_locked(sk, &msg, size);
+		ret = sk_is_msk(sk) ?
+		      mptcp_sendmsg_locked(sk, &msg, size) :
+		      tcp_sendmsg_locked(sk, &msg, size);
 
 		if (ret != size) {
 			if (ret > 0) {
Signature · 1 line
2.51.0
author reviewer
#5 lore raw

RFC v4 5/10 tls: switch to mptcp_recv_skb

In reply to Geliang Tang

From: Geliang Tang <[email protected]>

To extend MPTCP support based on TCP TLS, corresponding MPTCP-specific helper mptcp_recv_skb() has been exported into net/mptcp.h.

TLS implementation switches between the respective TCP and MPTCP helpers based on the detected protocol.

---
 include/net/mptcp.h  | 7 +++++++
 net/mptcp/protocol.c | 2 +-
 net/tls/tls_strp.c   | 4 +++-
 3 files changed, 11 insertions(+), 2 deletions(-)
include/net/mptcp.h +7 -0
index cdb726865486..3fe60818cada 100644
@@ -258,6 +258,8 @@ void mptcp_active_detect_blackhole(struct sock *sk, bool expired);
 unsigned int mptcp_inq(const struct sock *sk);
 
 int mptcp_sendmsg_locked(struct sock *sk, struct msghdr *msg, size_t len);
+
+struct sk_buff *mptcp_recv_skb(struct sock *sk, u32 *off);
 #else
 
 static inline void mptcp_init(void)
@@ -360,6 +362,11 @@ static inline int mptcp_sendmsg_locked(struct sock *sk, struct msghdr *msg,
 {
 	return 0;
 }
+
+static inline struct sk_buff *mptcp_recv_skb(struct sock *sk, u32 *off)
+{
+	return NULL;
+}
 #endif /* CONFIG_MPTCP */
 
 #if IS_ENABLED(CONFIG_MPTCP_IPV6)
net/mptcp/protocol.c +1 -1
index feac59a92b32..92141e87ed25 100644
@@ -4340,7 +4340,7 @@ static __poll_t mptcp_poll(struct file *file, struct socket *sock,
 	return mask;
 }
 
-static struct sk_buff *mptcp_recv_skb(struct sock *sk, u32 *off)
+struct sk_buff *mptcp_recv_skb(struct sock *sk, u32 *off)
 {
 	struct mptcp_sock *msk = mptcp_sk(sk);
 	struct sk_buff *skb;
net/tls/tls_strp.c +3 -1
index 0639130b8d5e..3bddb871d90b 100644
@@ -468,7 +468,9 @@ static void tls_strp_load_anchor_with_queue(struct tls_strparser *strp, int len)
 	struct sk_buff *first;
 	u32 offset;
 
-	first = tcp_recv_skb(strp->sk, tp->copied_seq, &offset);
+	first = sk_is_msk(strp->sk) ?
+		mptcp_recv_skb(strp->sk, &offset) :
+		tcp_recv_skb(strp->sk, tp->copied_seq, &offset);
 	if (WARN_ON_ONCE(!first))
 		return;
 
Signature · 1 line
2.51.0
author reviewer
#6 lore raw

RFC v4 6/10 tls: switch to mptcp_read_done

In reply to Geliang Tang

From: Geliang Tang <[email protected]>

To extend MPTCP support based on TCP TLS, corresponding MPTCP-specific helper mptcp_read_done() has been implemented.

TLS implementation switches between the respective TCP and MPTCP helpers based on the detected protocol.

---
 include/net/mptcp.h  |  4 ++++
 net/mptcp/protocol.c | 36 ++++++++++++++++++++++++++++++++++++
 net/tls/tls_strp.c   |  4 ++++
 3 files changed, 44 insertions(+)
include/net/mptcp.h +4 -0
index 3fe60818cada..7346eb9b19bd 100644
@@ -260,6 +260,8 @@ unsigned int mptcp_inq(const struct sock *sk);
 int mptcp_sendmsg_locked(struct sock *sk, struct msghdr *msg, size_t len);
 
 struct sk_buff *mptcp_recv_skb(struct sock *sk, u32 *off);
+
+void mptcp_read_done(struct sock *sk, size_t len);
 #else
 
 static inline void mptcp_init(void)
@@ -367,6 +369,8 @@ static inline struct sk_buff *mptcp_recv_skb(struct sock *sk, u32 *off)
 {
 	return NULL;
 }
+
+static inline void mptcp_read_done(struct sock *sk, size_t len) { }
 #endif /* CONFIG_MPTCP */
 
 #if IS_ENABLED(CONFIG_MPTCP_IPV6)
net/mptcp/protocol.c +36 -0
index 92141e87ed25..205492bc6345 100644
@@ -4532,6 +4532,42 @@ static ssize_t mptcp_splice_read(struct socket *sock, loff_t *ppos,
 	return ret;
 }
 
+void mptcp_read_done(struct sock *sk, size_t len)
+{
+	struct mptcp_sock *msk = mptcp_sk(sk);
+	struct sk_buff *skb;
+	size_t left;
+	u32 offset;
+
+	msk_owned_by_me(msk);
+
+	if (sk->sk_state == TCP_LISTEN)
+		return;
+
+	left = len;
+	while (left && (skb = mptcp_recv_skb(sk, &offset)) != NULL) {
+		int used;
+
+		used = min_t(size_t, skb->len - offset, left);
+		msk->bytes_consumed += used;
+		MPTCP_SKB_CB(skb)->offset += used;
+		MPTCP_SKB_CB(skb)->map_seq += used;
+		left -= used;
+
+		if (skb->len > offset + used)
+			break;
+
+		mptcp_eat_recv_skb(sk, skb);
+	}
+
+	mptcp_rcv_space_adjust(msk, len - left);
+
+	/* Clean up data we have read: This will do ACK frames. */
+	if (left != len)
+		mptcp_cleanup_rbuf(msk, len - left);
+}
+EXPORT_SYMBOL(mptcp_read_done);
+
 static const struct proto_ops mptcp_stream_ops = {
 	.family		   = PF_INET,
 	.owner		   = THIS_MODULE,
net/tls/tls_strp.c +4 -0
index 3bddb871d90b..81ff6be22162 100644
@@ -132,6 +132,8 @@ int tls_strp_msg_cow(struct tls_sw_context_rx *ctx)
 	tls_strp_anchor_free(strp);
 	strp->anchor = skb;
 
+	sk_is_msk(strp->sk) ?
+	mptcp_read_done(strp->sk, strp->stm.full_len) :
 	tcp_read_done(strp->sk, strp->stm.full_len);
 	strp->copy_mode = 1;
 
@@ -596,6 +598,8 @@ void tls_strp_msg_done(struct tls_strparser *strp)
 	WARN_ON(!strp->stm.full_len);
 
 	if (likely(!strp->copy_mode))
+		sk_is_msk(strp->sk) ?
+		mptcp_read_done(strp->sk, strp->stm.full_len) :
 		tcp_read_done(strp->sk, strp->stm.full_len);
 	else
 		tls_strp_flush_anchor_copy(strp);
Signature · 1 line
2.51.0
author reviewer
#7 lore raw

RFC v4 7/10 mptcp: update ULP getsockopt

In reply to Geliang Tang

From: Geliang Tang <[email protected]>

TCP_ULP was obtained by calling mptcp_getsockopt_first_sf_only() to get the ULP of the first subflow. Now that the mechanism has changed, a new helper needs to be implemented to get the ULP of the msk.

---
 net/mptcp/sockopt.c | 12 ++++++++++++
 1 file changed, 12 insertions(+)
net/mptcp/sockopt.c +12 -0
index de90a2897d2d..f3db4f2e8f81 100644
@@ -1393,6 +1393,17 @@ static int mptcp_put_int_option(struct mptcp_sock *msk, char __user *optval,
 	return 0;
 }
 
+static int mptcp_getsockopt_msk(struct sock *sk, int level, int optname,
+				char __user *optval, int __user *optlen)
+{
+	int ret;
+
+	lock_sock(sk);
+	ret = tcp_getsockopt(sk, level, optname, optval, optlen);
+	release_sock(sk);
+	return ret;
+}
+
 static int mptcp_getsockopt_sol_tcp(struct mptcp_sock *msk, int optname,
 				    char __user *optval, int __user *optlen)
 {
@@ -1400,6 +1411,7 @@ static int mptcp_getsockopt_sol_tcp(struct mptcp_sock *msk, int optname,
 
 	switch (optname) {
 	case TCP_ULP:
+		return mptcp_getsockopt_msk(sk, SOL_TCP, optname, optval, optlen);
 	case TCP_CONGESTION:
 	case TCP_INFO:
 	case TCP_CC_INFO:
Signature · 1 line
2.51.0
author reviewer
#8 lore raw

RFC v4 8/10 mptcp: enable TLS setsockopt

In reply to Geliang Tang

From: Geliang Tang <[email protected]>

This patch adds MPTCP TLS setsockopt support. It allows setting the TCP_ULP option to 'tls' exclusively, and enables configuration of the TLS_TX and TLS_RX options at the SOL_TLS level.

This option cannot be set when the socket is in CLOSE or LISTEN state.

---
 net/mptcp/sockopt.c | 18 +++++++++++++++++-
 1 file changed, 17 insertions(+), 1 deletion(-)
net/mptcp/sockopt.c +17 -1
index f3db4f2e8f81..52ff75702404 100644
@@ -12,6 +12,7 @@
 #include <net/protocol.h>
 #include <net/tcp.h>
 #include <net/mptcp.h>
+#include <net/tls.h>
 #include "protocol.h"
 
 #define MIN_INFO_OPTLEN_SIZE		16
@@ -567,6 +568,7 @@ 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;
 		}
 
@@ -576,6 +578,13 @@ 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;
 }
 
@@ -819,11 +828,18 @@ static int mptcp_setsockopt_sol_tcp(struct mptcp_sock *msk, int optname,
 				    sockptr_t optval, unsigned int optlen)
 {
 	struct sock *sk = (void *)msk;
+	char ulp[4] = "";
 	int ret, val;
 
 	switch (optname) {
 	case TCP_ULP:
-		return -EOPNOTSUPP;
+		if (copy_from_user(ulp, optval.user, 4))
+			return -EFAULT;
+		if (strcmp(ulp, "tls\0"))
+			return -EOPNOTSUPP;
+		if ((1 << sk->sk_state) & (TCPF_CLOSE | TCPF_LISTEN))
+			return -EINVAL;
+		return tcp_setsockopt(sk, SOL_TCP, optname, optval, optlen);
 	case TCP_CONGESTION:
 		return mptcp_setsockopt_sol_tcp_congestion(msk, optval, optlen);
 	case TCP_DEFER_ACCEPT:
Signature · 1 line
2.51.0
#9 lore raw

Re: [RFC mptcp-next v4 08/10] mptcp: enable TLS setsockopt

Hi, Geliang:

Quoted 56 lines · Geliang Tang

From: Geliang Tang <[email protected]>

This patch adds MPTCP TLS setsockopt support. It allows setting the TCP_ULP option to 'tls' exclusively, and enables configuration of the TLS_TX and TLS_RX options at the SOL_TLS level.

This option cannot be set when the socket is in CLOSE or LISTEN state.

---
 net/mptcp/sockopt.c | 18 +++++++++++++++++-
 1 file changed, 17 insertions(+), 1 deletion(-)
net/mptcp/sockopt.c +10 -0
index f3db4f2e8f81..52ff75702404 100644
@@ -12,6 +12,7 @@
 #include <net/protocol.h>
 #include <net/tcp.h>
 #include <net/mptcp.h>
+#include <net/tls.h>
 #include "protocol.h"
 
 #define MIN_INFO_OPTLEN_SIZE		16
@@ -567,6 +568,7 @@ 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;
 		}
 
@@ -576,6 +578,13 @@ 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;
 }
 
@@ -819,11 +828,18 @@ static int mptcp_setsockopt_sol_tcp(struct mptcp_sock *msk, int optname,
 				    sockptr_t optval, unsigned int optlen)
 {
 	struct sock *sk = (void *)msk;
+	char ulp[4] = "";
 	int ret, val;

net/mptcp/sockopt.c static int mptcp_setsockopt_sol_tcp(struct mptcp_sock *msk, int optname, · line 841

Quoted 10 lines
net/mptcp/sockopt.c +6 -1
 
 	switch (optname) {
 	case TCP_ULP:
-		return -EOPNOTSUPP;
+		if (copy_from_user(ulp, optval.user, 4))
+			return -EFAULT;
+		if (strcmp(ulp, "tls\0"))
+			return -EOPNOTSUPP;
+		if ((1 << sk->sk_state) & (TCPF_CLOSE | TCPF_LISTEN))
+			return -EINVAL;

Here should return -ENOTCONN

I'm running the tls selftest(tools/testing/selftest/net/tls.c), and the '-EINVAL' will cause an error in 'non_established' test, it checks the errno should be 'ENOTCONN'.

If we don't return here is also OK, because the 'tcp_setsockopt' can return too, but I think a state validation at the MPTCP layer is necessary, and 'ENOTCONN' is more accurate for 'TCPF_CLOSE | TCPF_LISTEN'.

WDYT

Thanks
Gang

Quoted 8 lines
net/mptcp/sockopt.c +1 -0
+		return tcp_setsockopt(sk, SOL_TCP, optname, optval, optlen);
 	case TCP_CONGESTION:
 		return mptcp_setsockopt_sol_tcp_congestion(msk, optval, optlen);
 	case TCP_DEFER_ACCEPT:
Signature · 1 line
2.51.0
author reviewer
#10 lore raw

Re: [RFC mptcp-next v4 08/10] mptcp: enable TLS setsockopt

Quoted 69 lines · GangYan

Hi, Geliang:

Quoted 68 lines · Geliang Tang

From: Geliang Tang <[email protected]>

This patch adds MPTCP TLS setsockopt support. It allows setting the TCP_ULP
option to 'tls' exclusively, and enables configuration of the TLS_TX and
TLS_RX options at the SOL_TLS level.

This option cannot be set when the socket is in CLOSE or LISTEN state.

---
 net/mptcp/sockopt.c | 18 +++++++++++++++++-

 1 file changed, 17 insertions(+), 1 deletion(-)
net/mptcp/sockopt.c +1 -0
index f3db4f2e8f81..52ff75702404 100644
@@ -12,6 +12,7 @@
  #include <net/protocol.h>
  #include <net/tcp.h>
  #include <net/mptcp.h>
+#include <net/tls.h>
  #include "protocol.h"


 #define MIN_INFO_OPTLEN_SIZE 16

net/mptcp/sockopt.c +0 -0
@@ -567,6 +568,7 @@ static bool mptcp_supported_sockopt(int level,
 int optname)

  case TCP_FASTOPEN_CONNECT:
  case TCP_FASTOPEN_KEY:
  case TCP_FASTOPEN_NO_COOKIE:

net/mptcp/sockopt.c +1 -0
+		case TCP_ULP:

  return true;
  }

net/mptcp/sockopt.c +0 -0
@@ -576,6 +578,13 @@ static bool mptcp_supported_sockopt(int level,
 int optname)
  		 * TCP_REPAIR_WINDOW are not supported, better

avoid this mess
  */
  }

net/mptcp/sockopt.c +7 -0
+	if (level == SOL_TLS) {
+		switch (optname) {
+		case TLS_TX:
+		case TLS_RX:
+			return true;
+		}
+	}

  return false;
 }

net/mptcp/sockopt.c +0 -0
@@ -819,11 +828,18 @@ static int mptcp_setsockopt_sol_tcp(struct
 mptcp_sock *msk, int optname,

      sockptr_t optval, unsigned int optlen)
 {
  struct sock *sk = (void *)msk;

net/mptcp/sockopt.c +1 -0
+	char ulp[4] = "";

  int ret, val;

  switch (optname) {
  case TCP_ULP:

net/mptcp/sockopt.c +0 -1
-		return -EOPNOTSUPP;
Quoted 8 lines

net/mptcp/sockopt.c static int mptcp_setsockopt_sol_tcp(struct mptcp_sock *msk, int optname, · line 841

Quoted 7 lines
net/mptcp/sockopt.c +5 -0
+		if (copy_from_user(ulp, optval.user, 4))
+			return -EFAULT;
+		if (strcmp(ulp, "tls\0"))
+			return -EOPNOTSUPP;
+		if ((1 << sk->sk_state) & (TCPF_CLOSE |

TCPF_LISTEN))

net/mptcp/sockopt.c +1 -0
+			return -EINVAL;

Here should return -ENOTCONN

Yes, you're right. I'll update it in v5.

Thanks,
-Geliang

Quoted 27 lines

I'm running the tls selftest(tools/testing/selftest/net/tls.c), and the
'-EINVAL' will cause an error in 'non_established' test, it checks the
errno should be 'ENOTCONN'.

If we don't return here is also OK, because the 'tcp_setsockopt' can return too, but I think a state validation at the
MPTCP layer is necessary, and 'ENOTCONN' is more accurate for 'TCPF_CLOSE | TCPF_LISTEN'.

WDYT

Thanks
Gang

Quoted 10 lines
net/mptcp/sockopt.c +1 -0
+		return tcp_setsockopt(sk, SOL_TCP, optname,

optval, optlen);
  case TCP_CONGESTION:
  return mptcp_setsockopt_sol_tcp_congestion(msk, optval, optlen);
  case TCP_DEFER_ACCEPT:

Signature · 1 line
2.51.0
author reviewer
#11 lore raw

RFC v4 9/10 selftests: mptcp: connect: update sock_test_tcpulp

In reply to Geliang Tang

From: Geliang Tang <[email protected]>

Update sock_test_tcpulp() to ensure the mptcp_connect.c tests are not broken.

fallback() helper has been implemented to verify fallback scenarios, ensuring that MPTCP fallback-to-TCP tests continue to pass.

---
 .../selftests/net/mptcp/mptcp_connect.c       | 20 ++++++++++++-------
 1 file changed, 13 insertions(+), 7 deletions(-)
tools/testing/selftests/net/mptcp/mptcp_connect.c +13 -7
index b82df82e0594..e5c394ea4e11 100644
@@ -266,6 +266,18 @@ static void set_mptfo(int fd, int pf)
 		perror("TCP_FASTOPEN");
 }
 
+static int fallback(int fd)
+{
+	int is_mptcp = 0;
+	socklen_t optlen;
+
+	optlen = sizeof(is_mptcp);
+	if (getsockopt(fd, IPPROTO_TCP, TCP_IS_MPTCP, &is_mptcp, &optlen) == -1)
+		perror("TCP_IS_MPTCP");
+
+	return !is_mptcp;
+}
+
 static int do_ulp_so(int sock, const char *name)
 {
 	return setsockopt(sock, IPPROTO_TCP, TCP_ULP, name, strlen(name));
@@ -282,7 +294,7 @@ static void sock_test_tcpulp(int sock, int proto, unsigned int line)
 		X("getsockopt");
 
 	if (buflen > 0) {
-		if (strcmp(buf, "mptcp") != 0)
+		if (strcmp(buf, fallback(sock) ? "mptcp" : "tls") != 0)
 			xerror("unexpected ULP '%s' for proto %d at line %u", buf, proto, line);
 		ret = do_ulp_so(sock, "tls");
 		if (ret == 0)
@@ -424,8 +436,6 @@ static int sock_connect_mptcp(const char * const remoteaddr,
 	}
 
 	freeaddrinfo(addr);
-	if (sock != -1)
-		SOCK_TEST_TCPULP(sock, proto);
 	return sock;
 }
 
@@ -1197,8 +1207,6 @@ int main_loop_s(int listensock)
 				xerror("can't open %s: %d", cfg_input, errno);
 		}
 
-		SOCK_TEST_TCPULP(remotesock, 0);
-
 		memset(&winfo, 0, sizeof(winfo));
 		err = copyfd_io(fd, remotesock, 1, true, &winfo);
 	} else {
@@ -1371,8 +1379,6 @@ int main_loop(void)
 again:
 	check_getpeername_connect(fd);
 
-	SOCK_TEST_TCPULP(fd, cfg_sock_proto);
-
 	if (cfg_rcvbuf)
 		set_rcvbuf(fd, cfg_rcvbuf);
 	if (cfg_sndbuf)
Signature · 1 line
2.51.0
author reviewer
#12 lore raw

RFC v4 10/10 selftests: mptcp: sockopt: implement MPTCP KTLS tests

In reply to Geliang Tang

From: Geliang Tang <[email protected]>

This patch adds Kernel TLS (KTLS) testing infrastructure to MPTCP sockopt selftest, introducing a new '-c' option to enable TLS tests. It includes a helper for configuring TLS socket options and implements MPTCP-specific KTLS test cases for both IPv4 and IPv6, along with the necessary header includes and config updates.

TLS_OVERHEAD_SIZE macro is defined to account for the overhead in sent and received data length.

---
 tools/testing/selftests/net/mptcp/config      |  1 +
 .../selftests/net/mptcp/mptcp_sockopt.c       | 76 ++++++++++++++++++-
 .../selftests/net/mptcp/mptcp_sockopt.sh      | 36 +++++++++
 3 files changed, 111 insertions(+), 2 deletions(-)
tools/testing/selftests/net/mptcp/config +1 -0
index 59051ee2a986..18bd29ac5b24 100644
@@ -34,3 +34,4 @@ CONFIG_NFT_SOCKET=m
 CONFIG_NFT_TPROXY=m
 CONFIG_SYN_COOKIES=y
 CONFIG_VETH=y
+CONFIG_TLS=y
tools/testing/selftests/net/mptcp/mptcp_sockopt.c +74 -2

138 lines not on this page · raw message

tools/testing/selftests/net/mptcp/mptcp_sockopt.sh +36 -0
index ab8bce06b262..23b483241f01 100755
@@ -351,6 +351,41 @@ do_tcpinq_tests()
 	return $?
 }
 
+do_tls_test()
+{
+	print_title "KTLS $*" | head -c 53
+	ip netns exec "$ns_sbox" ./mptcp_sockopt "$@"
+	local lret=$?
+	if [ $lret -ne 0 ];then
+		ret=$lret
+		mptcp_lib_pr_fail
+		mptcp_lib_result_fail "KTLS: $*"
+		return $lret
+	fi
+
+	mptcp_lib_pr_ok
+	mptcp_lib_result_pass "KTLS: $*"
+	return $lret
+}
+
+do_tls_tests()
+{
+	local lret=0
+
+	mptcp_lib_print_info "sockopt KTLS"
+
+	local args
+	for args in "-c" "-6 -c"; do
+		do_tls_test $args
+		lret=$?
+		if [ $lret -ne 0 ] ; then
+			return $lret
+		fi
+	done
+
+	return $lret
+}
+
 sin=$(mktemp)
 sout=$(mktemp)
 cin=$(mktemp)
@@ -366,6 +401,7 @@ run_tests $ns1 $ns2 dead:beef:1::1
 
 do_mptcp_sockopt_tests
 do_tcpinq_tests
+do_tls_tests
 
 mptcp_lib_result_print_all_tap
 exit $ret
Signature · 1 line
2.51.0
#13 lore raw
In reply to Geliang Tang

Hi Geliang,

Thank you for your modifications, that's great!

Our CI did some validations and here is its report:

- KVM Validation: normal (except selftest_mptcp_join): Unstable: 1 failed test(s): selftest_simult_flows 🔴
- KVM Validation: normal (only selftest_mptcp_join): Success! ✅
- KVM Validation: debug (except selftest_mptcp_join): Unstable: 1 failed test(s): packetdrill_add_addr 🔴
- KVM Validation: debug (only selftest_mptcp_join): Critical: 2 Call Trace(s) - Critical: Unexpected stop of the VM ❌
- KVM Validation: btf-normal (only bpftest_all): Success! ✅
- KVM Validation: btf-debug (only bpftest_all): Success! ✅
- Task: https://github.com/multipath-tcp/mptcp_net-next/actions/runs/20154532531

Initiator: Patchew Applier
Commits: https://github.com/multipath-tcp/mptcp_net-next/commits/397a7ac03cb4
Patchwork: https://patchwork.kernel.org/project/mptcp/list/?series=1032489

If there are some issues, you can reproduce them using the same environment as the one used by the CI thanks to a docker image, e.g.:

    $ cd [kernel source code]
    $ docker run -v "${PWD}:${PWD}:rw" -w "${PWD}" --privileged --rm -it \
        --pull always mptcp/mptcp-upstream-virtme-docker:latest \
        auto-normal

For more details:

https://github.com/multipath-tcp/mptcp-upstream-virtme-docker

Please note that despite all the efforts that have been already done to have a stable tests suite when executed on a public CI like here, it is possible some reported issues are not due to your modifications. Still, do not hesitate to help us improve that ;-)

Cheers,
MPTCP GH Action bot
Bot operated by Matthieu Baerts (NGI0 Core)

#14 lore raw
In reply to Geliang Tang

Hi Geliang,

Thank you for your modifications, that's great!

Our CI did some validations and here is its report:

- KVM Validation: normal (except selftest_mptcp_join): Unstable: 1 failed test(s): selftest_simult_flows 🔴
- KVM Validation: normal (only selftest_mptcp_join): Success! ✅
- KVM Validation: debug (except selftest_mptcp_join): Unstable: 1 failed test(s): packetdrill_add_addr 🔴
- KVM Validation: debug (only selftest_mptcp_join): Success! ✅
- KVM Validation: btf-normal (only bpftest_all): Success! ✅
- KVM Validation: btf-debug (only bpftest_all): Success! ✅
- Task: https://github.com/multipath-tcp/mptcp_net-next/actions/runs/20154532531

Initiator: Patchew Applier
Commits: https://github.com/multipath-tcp/mptcp_net-next/commits/397a7ac03cb4
Patchwork: https://patchwork.kernel.org/project/mptcp/list/?series=1032489

If there are some issues, you can reproduce them using the same environment as the one used by the CI thanks to a docker image, e.g.:

    $ cd [kernel source code]
    $ docker run -v "${PWD}:${PWD}:rw" -w "${PWD}" --privileged --rm -it \
        --pull always mptcp/mptcp-upstream-virtme-docker:latest \
        auto-normal

For more details:

https://github.com/multipath-tcp/mptcp-upstream-virtme-docker

Please note that despite all the efforts that have been already done to have a stable tests suite when executed on a public CI like here, it is possible some reported issues are not due to your modifications. Still, do not hesitate to help us improve that ;-)

Cheers,
MPTCP GH Action bot
Bot operated by Matthieu Baerts (NGI0 Core)

maintainer
#15 lore raw
In reply to Geliang Tang

Hi Geliang,

Quoted 12 lines · Geliang Tang

From: Geliang Tang <[email protected]>

v4:
- split "tls: add MPTCP protocol support" into smaller, more focused patches.
- a new mptcp_inq helper has been implemented instead of directly using mptcp_inq_hint to fix the issue mentioned in [1].
- add sk_is_msk helper.
- the 'expect' parameter will no longer be added to sock_test_tcpulp. Instead, SOCK_TEST_TCPULP items causing the tests failure will be directly removed.
- remove the "TCP KTLS" tests, keeping only the MPTCP-related ones.

Thank you for the new version.

We briefly looked at the series with Mat yesterday, and we think that this series would be better accepted by the KTLS maintainers if function pointers are used instead of all the sk_is_msk() you added in many places. In other words, adding a new layer, where "struct tls_strparser" (or another one?) would have new fields with function pointer to call tcp_inq() or mptcp_inq(), etc. e.g. strp->ops->tcp_inq(strp->sk).

MPTCP would then only be checked once at the initialisation, not before each call.

Notes: If adding an indirection is an issue for the KTLS maintainers, INDIRECT_CALL_*(...) macros can be used. Also, it might be good to send the RFC to KTLS people, but probably best to wait for the read-sock series to be in net-next? But if you are not sure about the new suggested way, I guess it is always possible to email KTLS maintainers, and ask for their feedback.

Cheers,
Matt

Signature · 1 line
Sponsored by the NGI0 Core fund.