From: Seung Min Shin <guncraft2000@naver.com>
To: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>,
Xin Long <lucien.xin@gmail.com>
Cc: "David S . Miller" <davem@davemloft.net>,
Eric Dumazet <edumazet@google.com>,
Jakub Kicinski <kuba@kernel.org>, Paolo Abeni <pabeni@redhat.com>,
Simon Horman <horms@kernel.org>,
linux-sctp@vger.kernel.org, netdev@vger.kernel.org,
linux-kernel@vger.kernel.org
Subject: [PATCH] sctp: hold transport reference across connect wait
Date: Mon, 28 Sep 2026 15:21:32 +0000 [thread overview]
Message-ID: <20260928152132.3705137-1-guncraft2000@naver.com> (raw)
sctp_sendmsg_to_asoc() selects a transport before waiting for a new
interleaved association to become established. sctp_wait_for_connect()
drops the socket lock while sleeping.
While the lock is dropped, backlog processing can handle a COOKIE ACK and
an ASCONF DEL-IP that removes the selected transport. sctp_assoc_rm_peer()
then marks the transport dead and drops its association-owned reference.
The transport can be freed through RCU before the send path resumes.
After the wait, the stale pointer is assigned to newly created DATA or
I-DATA chunks and later dereferenced by sctp_outq_select_transport().
Hold a reference to the selected transport across
sctp_wait_for_connect(). After the socket lock is reacquired, abort the
send if the transport was marked dead. Callers without a selected
transport pass NULL and retain their existing behavior.
Cc: stable@kernel.org
Fixes: 668c9beb9020 ("sctp: implement assign_number for sctp_stream_interleave")
Assisted-by: LLM
Signed-off-by: Seung Min Shin <guncraft2000@naver.com>
---
Notes:
Tested on 165768bb70265b5c38cf0b73fafd75be235f8b14 with
CONFIG_KASAN=y on x86-64.
Before the patch:
- The attached reproducer triggered the reported transport use-after-free.
After the patch:
- The exact attached reproducer completed 64 consecutive attempts without a
KASAN, refcount, BUG, WARNING, or panic report.
- A diagnostic run returned ESRCH in all 64 attempts and observed the
association in SCTP_ESTABLISHED state after each return.
- A normal stream-interleaving connect/send smoke test passed.
The test config does not enable LOCKDEP, so this was not a LOCKDEP test.
net/sctp/socket.c | 26 +++++++++++++++++++-------
1 file changed, 19 insertions(+), 7 deletions(-)
diff --git a/net/sctp/socket.c b/net/sctp/socket.c
index c7b9e325ec1c..621df93d3450 100644
--- a/net/sctp/socket.c
+++ b/net/sctp/socket.c
@@ -75,7 +75,8 @@ static int sctp_wait_for_sndbuf(struct sctp_association *asoc,
struct sctp_transport *transport,
long *timeo_p, size_t msg_len);
static int sctp_wait_for_packet(struct sock *sk, int *err, long *timeo_p);
-static int sctp_wait_for_connect(struct sctp_association *, long *timeo_p);
+static int sctp_wait_for_connect(struct sctp_association *,
+ struct sctp_transport *, long *timeo_p);
static int sctp_wait_for_accept(struct sock *sk, long timeo);
static void sctp_wait_for_close(struct sock *sk, long timeo);
static void sctp_destruct_sock(struct sock *sk);
@@ -1245,7 +1246,7 @@ static int __sctp_connect(struct sock *sk, struct sockaddr *kaddrs,
*assoc_id = asoc->assoc_id;
timeo = sock_sndtimeo(sk, flags & O_NONBLOCK);
- return sctp_wait_for_connect(asoc, &timeo);
+ return sctp_wait_for_connect(asoc, NULL, &timeo);
out_free:
pr_debug("%s: took out_free path with asoc:%p kaddrs:%p err:%d\n",
@@ -1843,7 +1844,7 @@ static int sctp_sendmsg_to_asoc(struct sctp_association *asoc,
if (asoc->ep->intl_enable) {
timeo = sock_sndtimeo(sk, 0);
- err = sctp_wait_for_connect(asoc, &timeo);
+ err = sctp_wait_for_connect(asoc, transport, &timeo);
if (err) {
err = -ESRCH;
goto err;
@@ -1881,7 +1882,7 @@ static int sctp_sendmsg_to_asoc(struct sctp_association *asoc,
if (unlikely(wait_connect)) {
timeo = sock_sndtimeo(sk, msg->msg_flags & MSG_DONTWAIT);
- sctp_wait_for_connect(asoc, &timeo);
+ sctp_wait_for_connect(asoc, NULL, &timeo);
}
err = msg_len;
@@ -9385,7 +9386,9 @@ static bool sctp_writeable(const struct sock *sk)
/* Wait for an association to go into ESTABLISHED state. If timeout is 0,
* returns immediately with EINPROGRESS.
*/
-static int sctp_wait_for_connect(struct sctp_association *asoc, long *timeo_p)
+static int sctp_wait_for_connect(struct sctp_association *asoc,
+ struct sctp_transport *transport,
+ long *timeo_p)
{
struct sock *sk = asoc->base.sk;
int err = 0;
@@ -9394,12 +9397,16 @@ static int sctp_wait_for_connect(struct sctp_association *asoc, long *timeo_p)
pr_debug("%s: asoc:%p, timeo:%ld\n", __func__, asoc, *timeo_p);
- /* Increment the association's refcnt. */
+ /* Hold references while the socket lock is dropped. */
+ if (transport)
+ sctp_transport_hold(transport);
sctp_association_hold(asoc);
for (;;) {
prepare_to_wait_exclusive(&asoc->wait, &wait,
TASK_INTERRUPTIBLE);
+ if (transport && transport->dead)
+ goto do_transport_dead;
if (!*timeo_p)
goto do_nonblock;
if (sk->sk_shutdown & RCV_SHUTDOWN)
@@ -9428,7 +9435,8 @@ static int sctp_wait_for_connect(struct sctp_association *asoc, long *timeo_p)
out:
finish_wait(&asoc->wait, &wait);
- /* Release the association's refcnt. */
+ if (transport)
+ sctp_transport_put(transport);
sctp_association_put(asoc);
return err;
@@ -9447,6 +9455,10 @@ static int sctp_wait_for_connect(struct sctp_association *asoc, long *timeo_p)
do_nonblock:
err = -EINPROGRESS;
goto out;
+
+do_transport_dead:
+ err = -EAGAIN;
+ goto out;
}
static int sctp_wait_for_accept(struct sock *sk, long timeo)
base-commit: a7bfaba4823e3c165bb2004c74eff7c096672bc7
--
2.34.1
next reply other threads:[~2026-09-28 15:31 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-09-28 15:21 Seung Min Shin [this message]
2026-10-01 18:53 ` [PATCH] sctp: hold transport reference across connect wait Xin Long
2026-10-02 0:38 ` Jakub Kicinski
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20260928152132.3705137-1-guncraft2000@naver.com \
--to=guncraft2000@naver.com \
--cc=davem@davemloft.net \
--cc=edumazet@google.com \
--cc=horms@kernel.org \
--cc=kuba@kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-sctp@vger.kernel.org \
--cc=lucien.xin@gmail.com \
--cc=marcelo.leitner@gmail.com \
--cc=netdev@vger.kernel.org \
--cc=pabeni@redhat.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox