[PATCH mptcp-next 0/2] TLS disconnect support

[PATCH mptcp-next 0/2] TLS disconnect support

[Geliang Tang]

From: Geliang Tang <tanggeliang@kylinos.cn>

To verify the correctness of MPTCP KTLS implementation, I added TLS test
support to MPTCP selftests mptcp_connect.c. The test results showed that
the disconnect test group failed with error code EOPNOTSUPP, so I added
this series to support TLS disconnect.

Patch 1 implements tls_disconnect() to support disconnect operation for
TLS sockets via connect(AF_UNSPEC).

Patch 2 updates selftests to verify the full disconnect/reconnect cycle,
including data transfer verification before and after disconnect.

Geliang Tang (2):
  tls: implement tls_disconnect
  selftests: tls: update disconnect tests

 include/net/tls.h                 |  12 +++-
 net/tls/tls_main.c                | 112 ++++++++++++++++++++++++++----
 tools/testing/selftests/net/tls.c |  57 ++++++++++++++-
 3 files changed, 163 insertions(+), 18 deletions(-)

-- 
2.55.0

[PATCH mptcp-next 1/2] tls: implement tls_disconnect

[Geliang Tang]

From: Geliang Tang <tanggeliang@kylinos.cn>

Implement disconnect support for TLS sockets via connect(AF_UNSPEC).
The handler invokes the underlying disconnect() first, then on success
cancels TX work (for TLS_SW), calls tls_sk_proto_free() to restore
original protocol operations and clear TLS state, cleans up RX context
and frees the TLS context. RX context uses kfree(), TLS context uses
kfree_rcu() for RCU safety.

tls_sk_proto_free() restores sk_prot, sk_socket->ops, sk_write_space,
and sk_destruct to their original values, clears icsk_ulp_data and
icsk_ulp_ops when free_ctx is true, and frees the TX sub-context. The
RX context and main TLS context are freed separately by tls_disconnect().

Add sk_proto_ops field to tls_context to save the original sk_socket->ops.
Unlike close() where the socket is destroyed, disconnect leaves the
socket alive for potential reconnection. sk_prot and sk_socket->ops are
restored inside tls_sk_proto_free() to prevent NULL pointer dereference
from concurrent poll/read/splice operations that dispatch through
sk_socket->ops.

After releasing the socket lock to cancel TX work, re-check the context
pointer to handle concurrent close(). Clear icsk_ulp_ops when freeing the
context to prevent tcp_set_ulp from failing with -EEXIST on reconnection.

Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
---
 include/net/tls.h  |  12 +++--
 net/tls/tls_main.c | 112 +++++++++++++++++++++++++++++++++++++++------
 2 files changed, 108 insertions(+), 16 deletions(-)

diff --git a/include/net/tls.h b/include/net/tls.h
index e57bef58851e..25ad76df97cb 100644
--- a/include/net/tls.h
+++ b/include/net/tls.h
@@ -193,6 +193,10 @@ enum tls_context_flags {
 	 * tls_dev_del call in tls_device_down if it happens simultaneously.
 	 */
 	TLS_RX_DEV_CLOSED = 2,
+	/* tls_disconnect() is in progress, prevent concurrent close() from
+	 * freeing resources while disconnect is using them.
+	 */
+	TLS_DISCONNECTING = 3,
 };
 
 struct cipher_context {
@@ -259,6 +263,7 @@ struct tls_context {
 
 	/* cache cold stuff */
 	struct proto *sk_proto;
+	const struct proto_ops *sk_proto_ops;
 	struct sock *sk;
 
 	void (*sk_destruct)(struct sock *sk);
@@ -374,10 +379,11 @@ static inline struct tls_context *tls_get_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,
-	 * TLS data path doesn't need rcu_dereference().
+	/* Use rcu_access_pointer() to safely read icsk_ulp_data,
+	 * as it can be cleared concurrently by tls_disconnect()
+	 * or tls_sk_proto_close().
 	 */
-	return (__force void *)icsk->icsk_ulp_data;
+	return (__force void *)rcu_access_pointer(icsk->icsk_ulp_data);
 }
 
 static inline struct tls_sw_context_rx *tls_sw_ctx_rx(
diff --git a/net/tls/tls_main.c b/net/tls/tls_main.c
index 8bdb78718a8d..d00e56af628a 100644
--- a/net/tls/tls_main.c
+++ b/net/tls/tls_main.c
@@ -358,32 +358,71 @@ static void tls_sk_proto_cleanup(struct sock *sk,
 	}
 }
 
-static void tls_sk_proto_close(struct sock *sk, long timeout)
+static void tls_sk_proto_free(struct sock *sk,
+			      struct tls_context *ctx,
+			      bool free_ctx)
 {
 	struct inet_connection_sock *icsk = inet_csk(sk);
-	struct tls_context *ctx = tls_get_ctx(sk);
 	long timeo = sock_sndtimeo(sk, 0);
-	bool free_ctx;
-
-	if (ctx->tx_conf == TLS_SW)
-		tls_sw_cancel_work_tx(ctx);
-
-	lock_sock(sk);
-	free_ctx = ctx->tx_conf != TLS_HW && ctx->rx_conf != TLS_HW;
 
 	if (ctx->tx_conf != TLS_BASE || ctx->rx_conf != TLS_BASE)
 		tls_sk_proto_cleanup(sk, ctx, timeo);
 
 	write_lock_bh(&sk->sk_callback_lock);
-	if (free_ctx)
-		rcu_assign_pointer(icsk->icsk_ulp_data, NULL);
 	WRITE_ONCE(sk->sk_prot, ctx->sk_proto);
+	if (sk->sk_socket)
+		WRITE_ONCE(sk->sk_socket->ops, ctx->sk_proto_ops);
 	if (sk->sk_write_space == tls_write_space)
 		sk->sk_write_space = ctx->sk_write_space;
+	if (sk->sk_destruct == tls_device_sk_destruct)
+		sk->sk_destruct = ctx->sk_destruct;
+	if (free_ctx) {
+		rcu_assign_pointer(icsk->icsk_ulp_data, NULL);
+		if (icsk->icsk_ulp_ops) {
+			/* Balance try_module_get in __tcp_ulp_find_autoload */
+			module_put(icsk->icsk_ulp_ops->owner);
+			icsk->icsk_ulp_ops = NULL;
+		}
+	}
 	write_unlock_bh(&sk->sk_callback_lock);
-	release_sock(sk);
 	if (ctx->tx_conf == TLS_SW)
 		tls_sw_free_ctx_tx(ctx);
+}
+
+static void tls_sk_proto_close(struct sock *sk, long timeout)
+{
+	struct tls_context *ctx;
+	bool free_ctx;
+
+	ctx = tls_get_ctx(sk);
+	if (!ctx) {
+		sk->sk_prot->close(sk, timeout);
+		return;
+	}
+
+	if (ctx->tx_conf == TLS_SW)
+		tls_sw_cancel_work_tx(ctx);
+
+	lock_sock(sk);
+	/* Re-read ctx under lock, as it may have been cleared by disconnect */
+	ctx = tls_get_ctx(sk);
+	if (!ctx) {
+		release_sock(sk);
+		sk->sk_prot->close(sk, timeout);
+		return;
+	}
+
+	/* Check if disconnect is in progress after acquiring lock */
+	if (test_bit(TLS_DISCONNECTING, &ctx->flags)) {
+		release_sock(sk);
+		ctx->sk_proto->close(sk, timeout);
+		return;
+	}
+
+	free_ctx = ctx->tx_conf != TLS_HW && ctx->rx_conf != TLS_HW;
+	tls_sk_proto_free(sk, ctx, free_ctx);
+	release_sock(sk);
+
 	if (ctx->rx_conf == TLS_SW || ctx->rx_conf == TLS_HW)
 		tls_sw_strparser_done(ctx);
 	if (ctx->rx_conf == TLS_SW)
@@ -904,7 +943,52 @@ static int tls_setsockopt(struct sock *sk, int level, int optname,
 
 static int tls_disconnect(struct sock *sk, int flags)
 {
-	return -EOPNOTSUPP;
+	struct tls_context *ctx;
+	bool free_ctx;
+	int rc;
+
+	ctx = tls_get_ctx(sk);
+	if (!ctx)
+		return 0;
+
+	if (!ctx->sk_proto->disconnect)
+		return -EOPNOTSUPP;
+
+	set_bit(TLS_DISCONNECTING, &ctx->flags);
+
+	rc = ctx->sk_proto->disconnect(sk, flags);
+	if (rc) {
+		clear_bit(TLS_DISCONNECTING, &ctx->flags);
+		return rc;
+	}
+
+	if (ctx->tx_conf == TLS_SW) {
+		struct tls_sw_context_tx *sw_ctx_tx = tls_sw_ctx_tx(ctx);
+
+		set_bit(BIT_TX_CLOSING, &sw_ctx_tx->tx_bitmask);
+		set_bit(BIT_TX_SCHEDULED, &sw_ctx_tx->tx_bitmask);
+
+		release_sock(sk);
+		disable_delayed_work_sync(&sw_ctx_tx->tx_work.work);
+		lock_sock(sk);
+	}
+
+	free_ctx = ctx->tx_conf != TLS_HW && ctx->rx_conf != TLS_HW;
+	tls_sk_proto_free(sk, ctx, free_ctx);
+
+	if (ctx->rx_conf == TLS_SW) {
+		struct tls_sw_context_rx *sw_ctx = tls_sw_ctx_rx(ctx);
+
+		release_sock(sk);
+		tls_strp_done(&sw_ctx->strp);
+		tls_sw_free_ctx_rx(ctx);
+		lock_sock(sk);
+	}
+
+	if (free_ctx)
+		tls_ctx_free(sk, ctx);
+
+	return 0;
 }
 
 struct tls_context *tls_ctx_create(struct sock *sk)
@@ -918,6 +1002,8 @@ struct tls_context *tls_ctx_create(struct sock *sk)
 
 	mutex_init(&ctx->tx_lock);
 	ctx->sk_proto = READ_ONCE(sk->sk_prot);
+	if (sk->sk_socket)
+		ctx->sk_proto_ops = READ_ONCE(sk->sk_socket->ops);
 	ctx->sk = sk;
 	/* Release semantic of rcu_assign_pointer() ensures that
 	 * ctx->sk_proto is visible before changing sk->sk_prot in
-- 
2.55.0

Re: [PATCH mptcp-next 1/2] tls: implement tls_disconnect

[sashiko-bot]

Thank you for your contribution! Sashiko AI review found 5 potential issue(s) to consider:
- [High] net/tls: double free of tls_context on concurrent connect(AF_UNSPEC)
- [High] net/tls: memory leak of TLS_HW contexts due to unconditional sk_destruct overwrite
- [High] net/tls: NULL pointer dereference in lockless socket ops on concurrent disconnect
- [High] tls: UAF in recvmsg/sendmsg due to concurrent tls_disconnect
- [High] tls: memory leak of HW offload context due to unconditional sk_destruct restore
--

commit c85d98242c0f64ea4b94701ff52f11b27b61f2a3
Author: Geliang Tang <geliang@kernel.org>

tls: implement tls_disconnect

Implement disconnect support for TLS sockets via connect(AF_UNSPEC).
The handler invokes the underlying disconnect() first, then on success
cancels TX work (for TLS_SW), calls tls_sk_proto_free() to restore
original protocol operations and clear TLS state, cleans up RX context
and frees the TLS context.

> diff --git a/net/tls/tls_main.c b/net/tls/tls_main.c
> index 8bdb78718a8d4..d00e56af628a0 100644
> --- a/net/tls/tls_main.c
> +++ b/net/tls/tls_main.c

[ ... ]

> -static void tls_sk_proto_close(struct sock *sk, long timeout)
> +static void tls_sk_proto_free(struct sock *sk,
> +			      struct tls_context *ctx,
> +			      bool free_ctx)
>  {
>  	struct inet_connection_sock *icsk = inet_csk(sk);
> -	struct tls_context *ctx = tls_get_ctx(sk);
>  	long timeo = sock_sndtimeo(sk, 0);
> -	bool free_ctx;
> -
> -	if (ctx->tx_conf == TLS_SW)
> -		tls_sw_cancel_work_tx(ctx);
> -
> -	lock_sock(sk);
> -	free_ctx = ctx->tx_conf != TLS_HW && ctx->rx_conf != TLS_HW;
>  
>  	if (ctx->tx_conf != TLS_BASE || ctx->rx_conf != TLS_BASE)
>  		tls_sk_proto_cleanup(sk, ctx, timeo);
>  
>  	write_lock_bh(&sk->sk_callback_lock);
> -	if (free_ctx)
> -		rcu_assign_pointer(icsk->icsk_ulp_data, NULL);
>  	WRITE_ONCE(sk->sk_prot, ctx->sk_proto);
> +	if (sk->sk_socket)
> +		WRITE_ONCE(sk->sk_socket->ops, ctx->sk_proto_ops);
>  	if (sk->sk_write_space == tls_write_space)
>  		sk->sk_write_space = ctx->sk_write_space;
> +	if (sk->sk_destruct == tls_device_sk_destruct)
> +		sk->sk_destruct = ctx->sk_destruct;

[Severity: High]
For hardware-offloaded TLS sockets (TLS_HW), free_ctx evaluates to false. This
means tls_ctx_free() is intentionally skipped so the context can be cleaned up
asynchronously later.

However, this code unconditionally unhooks tls_device_sk_destruct and restores
the original socket destructor. Does this completely bypass the hardware
offload cleanup and permanently leak the tls_context and associated NIC
hardware state when the socket is ultimately destroyed?

> +	if (free_ctx) {
> +		rcu_assign_pointer(icsk->icsk_ulp_data, NULL);

[Severity: High]
When tls_disconnect() is called, it can run concurrently with lockless
operations like setsockopt() or poll() on the same file descriptor.

If tls_disconnect() sets icsk_ulp_data to NULL here while tls_setsockopt() or
tls_sk_poll() are fetching it locklessly via tls_get_ctx(), won't those
functions dereference a NULL pointer since they don't validate the pointer
before dereferencing ctx?

> +		if (icsk->icsk_ulp_ops) {
> +			/* Balance try_module_get in __tcp_ulp_find_autoload */
> +			module_put(icsk->icsk_ulp_ops->owner);
> +			icsk->icsk_ulp_ops = NULL;
> +		}
> +	}
>  	write_unlock_bh(&sk->sk_callback_lock);
> -	release_sock(sk);
>  	if (ctx->tx_conf == TLS_SW)
>  		tls_sw_free_ctx_tx(ctx);
> +}

[ ... ]

> @@ -904,7 +943,52 @@ static int tls_setsockopt(struct sock *sk, int level, int optname,
>  
>  static int tls_disconnect(struct sock *sk, int flags)
>  {
> -	return -EOPNOTSUPP;
> +	struct tls_context *ctx;
> +	bool free_ctx;
> +	int rc;
> +
> +	ctx = tls_get_ctx(sk);
> +	if (!ctx)
> +		return 0;
> +
> +	if (!ctx->sk_proto->disconnect)
> +		return -EOPNOTSUPP;
> +
> +	set_bit(TLS_DISCONNECTING, &ctx->flags);

[Severity: High]
Two threads can concurrently call connect() with AF_UNSPEC on the same socket
file descriptor.

If thread 1 enters tls_disconnect(), sets the TLS_DISCONNECTING bit, and
temporarily drops the socket lock to wait for disable_delayed_work_sync(),
thread 2 can acquire the lock and enter tls_disconnect() as well.

Since set_bit() is used instead of test_and_set_bit() to abort, thread 2 will
resolve the same ctx pointer and proceed. Will this result in both threads
sequentially executing tls_sk_proto_free() and tls_ctx_free(), causing a
double free of the contexts?

> +
> +	rc = ctx->sk_proto->disconnect(sk, flags);
> +	if (rc) {
> +		clear_bit(TLS_DISCONNECTING, &ctx->flags);
> +		return rc;
> +	}

[ ... ]

> +	if (ctx->rx_conf == TLS_SW) {
> +		struct tls_sw_context_rx *sw_ctx = tls_sw_ctx_rx(ctx);
> +
> +		release_sock(sk);
> +		tls_strp_done(&sw_ctx->strp);
> +		tls_sw_free_ctx_rx(ctx);
> +		lock_sock(sk);
> +	}
> +
> +	if (free_ctx)
> +		tls_ctx_free(sk, ctx);

[Severity: High]
Threads calling tls_sw_recvmsg() or tls_sw_sendmsg() cache pointers to the
RX/TX contexts and then temporarily drop the socket lock to wait for data or
memory via sk_wait_event() or sk_stream_wait_memory().

If a concurrent disconnect triggers tls_disconnect(), it acquires the
socket lock and synchronously frees the RX and TX contexts here.

When the waiting recvmsg or sendmsg thread wakes up, won't it re-acquire the
socket lock and use its locally cached context pointer, leading to a
use-after-free read or write?

> +
> +	return 0;
>  }

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/cover.1791967771.git.tanggeliang@kylinos.cn?part=1

Re: [PATCH mptcp-next 1/2] tls: implement tls_disconnect

[kernel test robot]

Hi Geliang,

kernel test robot noticed the following build errors:

[auto build test ERROR on mptcp/export]
[also build test ERROR on mptcp/export-net shuah-kselftest/next shuah-kselftest/fixes linus/master v7.3-rc7 next-20261012]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch#_base_tree_information]

url:    https://github.com/intel-lab-lkp/linux/commits/Geliang-Tang/tls-implement-tls_disconnect/20261014-165935
base:   https://github.com/multipath-tcp/mptcp_net-next.git export
patch link:    https://lore.kernel.org/r/6ffac9e36dab9b4f5fdb6e55a744b35f0d51ffb4.1791967771.git.tanggeliang%40kylinos.cn
patch subject: [PATCH mptcp-next 1/2] tls: implement tls_disconnect
config: m68k-defconfig (https://download.01.org/0day-ci/archive/20261015/202610150856.tk1Ryxlc-lkp@intel.com/config)
compiler: m68k-linux-gcc (GCC) 16.1.0
reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20261015/202610150856.tk1Ryxlc-lkp@intel.com/reproduce)

If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@intel.com>
| Closes: https://lore.kernel.org/oe-kbuild-all/202610150856.tk1Ryxlc-lkp@intel.com/

All errors (new ones prefixed by >>):

   net/tls/tls_main.c: In function 'tls_sk_proto_free':
>> net/tls/tls_main.c:377:32: error: 'tls_device_sk_destruct' undeclared (first use in this function); did you mean 'tls_device_sendmsg'?
     377 |         if (sk->sk_destruct == tls_device_sk_destruct)
         |                                ^~~~~~~~~~~~~~~~~~~~~~
         |                                tls_device_sendmsg
   net/tls/tls_main.c:377:32: note: each undeclared identifier is reported only once for each function it appears in


vim +377 net/tls/tls_main.c

   360	
   361	static void tls_sk_proto_free(struct sock *sk,
   362				      struct tls_context *ctx,
   363				      bool free_ctx)
   364	{
   365		struct inet_connection_sock *icsk = inet_csk(sk);
   366		long timeo = sock_sndtimeo(sk, 0);
   367	
   368		if (ctx->tx_conf != TLS_BASE || ctx->rx_conf != TLS_BASE)
   369			tls_sk_proto_cleanup(sk, ctx, timeo);
   370	
   371		write_lock_bh(&sk->sk_callback_lock);
   372		WRITE_ONCE(sk->sk_prot, ctx->sk_proto);
   373		if (sk->sk_socket)
   374			WRITE_ONCE(sk->sk_socket->ops, ctx->sk_proto_ops);
   375		if (sk->sk_write_space == tls_write_space)
   376			sk->sk_write_space = ctx->sk_write_space;
 > 377		if (sk->sk_destruct == tls_device_sk_destruct)
   378			sk->sk_destruct = ctx->sk_destruct;
   379		if (free_ctx) {
   380			rcu_assign_pointer(icsk->icsk_ulp_data, NULL);
   381			if (icsk->icsk_ulp_ops) {
   382				/* Balance try_module_get in __tcp_ulp_find_autoload */
   383				module_put(icsk->icsk_ulp_ops->owner);
   384				icsk->icsk_ulp_ops = NULL;
   385			}
   386		}
   387		write_unlock_bh(&sk->sk_callback_lock);
   388		if (ctx->tx_conf == TLS_SW)
   389			tls_sw_free_ctx_tx(ctx);
   390	}
   391	

--
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki

[PATCH mptcp-next 2/2] selftests: tls: update disconnect tests

[Geliang Tang]

From: Geliang Tang <tanggeliang@kylinos.cn>

Update the TLS disconnect test to verify the full disconnect/reconnect
cycle.

The test first establishes a TLS connection and verifies data transfer
works. It then disconnects both ends of the connection using
connect(AF_UNSPEC) and verifies that send/recv operations fail with
appropriate error codes (EPIPE/ECONNRESET for send, ENOTCONN for recv).
The client socket is then reconnected to a new server, TLS is
re-established, and data transfer is verified to work again.

This ensures TLS sockets can be properly disconnected and reconnected
without state corruption.

Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
---
 tools/testing/selftests/net/tls.c | 57 +++++++++++++++++++++++++++++--
 1 file changed, 55 insertions(+), 2 deletions(-)

diff --git a/tools/testing/selftests/net/tls.c b/tools/testing/selftests/net/tls.c
index 9d3cd4fff062..ada2f43d05d3 100644
--- a/tools/testing/selftests/net/tls.c
+++ b/tools/testing/selftests/net/tls.c
@@ -1985,8 +1985,10 @@ TEST_F(tls_basic, disconnect)
 	int send_len = strlen(test_str) + 1;
 	struct tls_crypto_info_keys key;
 	struct sockaddr_in addr;
+	socklen_t len;
 	char buf[20];
 	int ret;
+	int sfd;
 
 	if (self->notls)
 		return;
@@ -2005,14 +2007,65 @@ TEST_F(tls_basic, disconnect)
 	ret = setsockopt(self->cfd, SOL_TLS, TLS_RX, &key, key.len);
 	ASSERT_EQ(ret, 0);
 
+	bzero(buf, sizeof(buf));
+	EXPECT_EQ(recv(self->cfd, buf, send_len, 0), send_len);
+	EXPECT_EQ(memcmp(buf, test_str, send_len), 0);
+
 	addr.sin_family = AF_UNSPEC;
 	addr.sin_addr.s_addr = htonl(INADDR_ANY);
 	addr.sin_port = 0;
 	ret = connect(self->cfd, &addr, sizeof(addr));
-	EXPECT_EQ(ret, -1);
-	EXPECT_EQ(errno, EOPNOTSUPP);
+	EXPECT_EQ(ret, 0);
+
+	/* Verify send/recv fail after disconnect. */
+	EXPECT_EQ(send(self->cfd, buf, sizeof(buf),
+		       MSG_DONTWAIT | MSG_NOSIGNAL), -1);
+	EXPECT_TRUE(errno == EPIPE || errno == ECONNRESET);
+	EXPECT_EQ(recv(self->cfd, buf, sizeof(buf), MSG_DONTWAIT), -1);
+	EXPECT_EQ(errno, ENOTCONN);
+
+	/* Disconnect fd as well. */
+	ret = connect(self->fd, &addr, sizeof(addr));
+	ASSERT_EQ(ret, 0);
+	EXPECT_EQ(send(self->fd, buf, sizeof(buf),
+		       MSG_DONTWAIT | MSG_NOSIGNAL), -1);
+	EXPECT_TRUE(errno == EPIPE || errno == ECONNRESET);
+	EXPECT_EQ(recv(self->fd, buf, sizeof(buf), MSG_DONTWAIT), -1);
+	EXPECT_EQ(errno, ENOTCONN);
+
+	/* Reconnect fd to a new server and re-setup TLS. */
+	addr.sin_family = AF_INET;
+	addr.sin_port = 0;
+	sfd = socket(AF_INET, SOCK_STREAM, 0);
+	EXPECT_GE(sfd, 0);
+	if (sfd < 0)
+		return;
+
+	/* sfd not closed on ASSERT failure, kernel cleans up on test abort */
+	ASSERT_EQ(bind(sfd, &addr, sizeof(addr)), 0);
+	ASSERT_EQ(listen(sfd, 10), 0);
+	len = sizeof(addr);
+	ASSERT_EQ(getsockname(sfd, &addr, &len), 0);
+	ASSERT_EQ(connect(self->fd, (struct sockaddr *)&addr, sizeof(addr)), 0);
 
+	ret = accept(sfd, &addr, &len);
+	close(sfd);
+	ASSERT_GE(ret, 0);
+	close(self->cfd);
+	self->cfd = ret;
+
+	ASSERT_EQ(setsockopt(self->fd, IPPROTO_TCP, TCP_ULP, "tls",
+			     sizeof("tls")), 0);
+	ASSERT_EQ(setsockopt(self->cfd, IPPROTO_TCP, TCP_ULP, "tls",
+			     sizeof("tls")), 0);
+	ASSERT_EQ(setsockopt(self->fd, SOL_TLS, TLS_TX, &key, key.len), 0);
+	ASSERT_EQ(setsockopt(self->cfd, SOL_TLS, TLS_RX, &key, key.len), 0);
+
+	/* Verify data transfer works after reconnect. */
+	EXPECT_EQ(send(self->fd, test_str, send_len, 0), send_len);
+	bzero(buf, sizeof(buf));
 	EXPECT_EQ(recv(self->cfd, buf, send_len, 0), send_len);
+	EXPECT_EQ(memcmp(buf, test_str, send_len), 0);
 }
 
 TEST_F(tls, rekey)
-- 
2.55.0

Re: [PATCH mptcp-next 0/2] TLS disconnect support

[MPTCP CI]

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): Success! ✅
- KVM Validation: normal (only selftest_mptcp_join): Success! ✅
- KVM Validation: debug (except selftest_mptcp_join): Success! ✅
- KVM Validation: debug (only selftest_mptcp_join): Success! ✅
- KVM Validation: btf-normal (only bpftest_all): Success! ✅
- KVM Validation: btf-debug (only bpftest_all): Success! ✅
- Perf: Success! ✅
- Task: https://github.com/multipath-tcp/mptcp_net-next/actions/runs/38541308008

Initiator: Patchew Applier
Commits: https://github.com/multipath-tcp/mptcp_net-next/commits/cab552cd4a4d
Patchwork: https://patchwork.kernel.org/project/mptcp/list/?series=1185404


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)

Re: [PATCH mptcp-next 0/2] TLS disconnect support

[Matthieu Baerts]

Hi Geliang,

On 14/10/2026 10:59, Geliang Tang wrote:
> From: Geliang Tang <tanggeliang@kylinos.cn>
> 
> To verify the correctness of MPTCP KTLS implementation, I added TLS test
> support to MPTCP selftests mptcp_connect.c. The test results showed that
> the disconnect test group failed with error code EOPNOTSUPP, so I added
> this series to support TLS disconnect.

I guess that's more a question for the KTLS maintainers, but my feeling
is that the disconnect is not supported on purpose because that's
probably not used in production, and source of many errors: see all the
bugs found around that for TCP and MPTCP (and others).

In fact, I just did a git blame, and found this: 5071a1e606b3 ("net:
tls: explicitly disallow disconnect").

Cheers,
Matt
-- 
Sponsored by the NGI0 Core fund.