[PATCH] Bluetooth: L2CAP: reject accept queue add unless BT_LISTEN

[PATCH] Bluetooth: L2CAP: reject accept queue add unless BT_LISTEN

[Pauli Virtanen]

New sk should not be added to parent socket accept queue after last
l2cap_sock_cleanup_listen() has run in l2cap_sock_teardown_cb() and
state set to BT_CLOSED, as that can result to UAF on dereferencing the
dangling parent reference.

l2cap_sock_new_connection_cb() may race with parent l2cap_chan teardown,
due to chan->state accessed without consistent locking:

  [Task 1]                           [Task 2]
  l2cap_sock_release(parent)         l2cap_connect
    l2cap_sock_shutdown                pchan = l2cap_global_chan_by_psm
      l2cap_chan_lock(pchan)
      l2cap_chan_close
        l2cap_sock_teardown_cb
          pchan->state = BT_CLOSED
      l2cap_chan_unlock(pchan) ------> l2cap_chan_lock(pchan)
                                       l2cap_new_connection
                                         l2cap_sock_new_connection_cb
      l2cap_chan_lock(pchan) <-------- l2cap_chan_unlock(pchan)
      l2cap_sock_kill(parent)          /* bt_sk(sk)->parent dangling */

Fix by adding check for sk_state == BT_LISTEN after acquiring sk lock in
l2cap_sock_new_connection_cb().  Add lock_sock() around sk_state writes
where missing, to avoid data races.

Although the data races on pchan->state should be fixed too, this
defensive sk_state check probably makes sense in any case.

Fixes: 2ff1a41a912d ("Bluetooth: L2CAP: Fix null-ptr-deref in l2cap_sock_state_change_cb()")
Reported-by: syzbot+9265e754091c2d27ea29@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=9265e754091c2d27ea29
Signed-off-by: Pauli Virtanen <pav@iki.fi>
---

Notes:
    Fixing up chan->state locking is a bigger changeset, and should be done
    separately.
    
#syz test

 net/bluetooth/l2cap_sock.c | 13 +++++++++++++
 1 file changed, 13 insertions(+)

diff --git a/net/bluetooth/l2cap_sock.c b/net/bluetooth/l2cap_sock.c
index 735167f73f31..8c2ac8b911e0 100644
--- a/net/bluetooth/l2cap_sock.c
+++ b/net/bluetooth/l2cap_sock.c
@@ -1568,6 +1568,11 @@ static int l2cap_sock_new_connection_cb(struct l2cap_chan *chan,
 
 	lock_sock(parent);
 
+	if (parent->sk_state != BT_LISTEN) {
+		release_sock(parent);
+		return -EINVAL;
+	}
+
 	/* Check for backlog size */
 	if (sk_acceptq_is_full(parent)) {
 		BT_DBG("backlog full %d", parent->sk_ack_backlog);
@@ -1731,10 +1736,14 @@ static void l2cap_sock_state_change_cb(struct l2cap_chan *chan, int state,
 	if (!sk)
 		return;
 
+	lock_sock(sk);
+
 	sk->sk_state = state;
 
 	if (err)
 		sk->sk_err = err;
+
+	release_sock(sk);
 }
 
 static struct sk_buff *l2cap_sock_alloc_skb_cb(struct l2cap_chan *chan,
@@ -1810,6 +1819,8 @@ static void l2cap_sock_resume_cb(struct l2cap_chan *chan)
 	if (!sk)
 		return;
 
+	lock_sock(sk);
+
 	if (test_and_clear_bit(FLAG_PENDING_SECURITY, &chan->flags)) {
 		sk->sk_state = BT_CONNECTED;
 		chan->state = BT_CONNECTED;
@@ -1817,6 +1828,8 @@ static void l2cap_sock_resume_cb(struct l2cap_chan *chan)
 
 	clear_bit(BT_SK_SUSPEND, &bt_sk(sk)->flags);
 	sk->sk_state_change(sk);
+
+	release_sock(sk);
 }
 
 static void l2cap_sock_set_shutdown_cb(struct l2cap_chan *chan)
-- 
2.55.0

Re: [syzbot] [bluetooth?] KASAN: slab-use-after-free Read in l2cap_sock_ready_cb (2)

[syzbot]

Hello,

syzbot has tested the proposed patch and the reproducer did not trigger any issue:

Reported-by: syzbot+9265e754091c2d27ea29@syzkaller.appspotmail.com
Tested-by: syzbot+9265e754091c2d27ea29@syzkaller.appspotmail.com

Tested on:

commit:         a7c7074b Merge tag 'fbdev-for-7.2-rc7' of git://git.ke..
git tree:       upstream
console output: https://syzkaller.appspot.com/x/log.txt?x=1105bfb9580000
kernel config:  https://syzkaller.appspot.com/x/.config?x=4887c52b0f7d06a5
dashboard link: https://syzkaller.appspot.com/bug?extid=9265e754091c2d27ea29
compiler:       Debian clang version 22.1.8 (++20260613092233+e80beda6e255-1~exp1~20260613092250.77), Debian LLD 22.1.8
patch:          https://syzkaller.appspot.com/x/patch.diff?x=10176132580000

Note: testing is done by a robot and is best-effort only.

Re: [PATCH] Bluetooth: L2CAP: reject accept queue add unless BT_LISTEN

[patchwork-bot+bluetooth]

Hello:

This patch was applied to bluetooth/bluetooth-next.git (master)
by Luiz Augusto von Dentz <luiz.von.dentz@intel.com>:

On Sun,  9 Aug 2026 01:06:05 +0300 you wrote:
> New sk should not be added to parent socket accept queue after last
> l2cap_sock_cleanup_listen() has run in l2cap_sock_teardown_cb() and
> state set to BT_CLOSED, as that can result to UAF on dereferencing the
> dangling parent reference.
> 
> l2cap_sock_new_connection_cb() may race with parent l2cap_chan teardown,
> due to chan->state accessed without consistent locking:
> 
> [...]

Here is the summary with links:
  - Bluetooth: L2CAP: reject accept queue add unless BT_LISTEN
    https://git.kernel.org/bluetooth/bluetooth-next/c/9db7e5fffbae

You are awesome, thank you!
-- 
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html