[<prev] [next>] [thread-next>] [day] [month] [year] [list]
Message-Id: <1fc1fd4512e4709d1fbeb7f008f38b273ee1d798.1590767183.git.pabeni@redhat.com>
Date: Fri, 29 May 2020 17:49:18 +0200
From: Paolo Abeni <pabeni@...hat.com>
To: netdev@...r.kernel.org
Cc: Mat Martineau <mathew.j.martineau@...ux.intel.com>,
"David S. Miller" <davem@...emloft.net>, mptcp@...ts.01.org
Subject: [PATCH net-next] mptcp: fix NULL ptr dereference in MP_JOIN error path
When token lookup on MP_JOIN 3rd ack fails, the server
socket closes with a reset the incoming child. Such socket
has the 'is_mptcp' flag set, but no msk socket associated
- due to the failed lookup.
While crafting the reset packet mptcp_established_options_mp()
will try to dereference the child's master socket, causing
a NULL ptr dereference.
This change addresses the issue with explicit fallback to
TCP in such error path.
Fixes: 729cd6436f35 ("mptcp: cope better with MP_JOIN failure")
Signed-off-by: Paolo Abeni <pabeni@...hat.com>
---
net/mptcp/subflow.c | 20 ++++++++++++++++----
1 file changed, 16 insertions(+), 4 deletions(-)
diff --git a/net/mptcp/subflow.c b/net/mptcp/subflow.c
index f3c06b8af92d..493b98a0825c 100644
--- a/net/mptcp/subflow.c
+++ b/net/mptcp/subflow.c
@@ -413,6 +413,20 @@ static void subflow_ulp_fallback(struct sock *sk,
tcp_sk(sk)->is_mptcp = 0;
}
+static void subflow_drop_ctx(struct sock *ssk)
+{
+ struct mptcp_subflow_context *ctx = mptcp_subflow_ctx(ssk);
+
+ if (!ctx)
+ return;
+
+ subflow_ulp_fallback(ssk, ctx);
+ if (ctx->conn)
+ sock_put(ctx->conn);
+
+ kfree_rcu(ctx, rcu);
+}
+
static struct sock *subflow_syn_recv_sock(const struct sock *sk,
struct sk_buff *skb,
struct request_sock *req,
@@ -485,10 +499,7 @@ static struct sock *subflow_syn_recv_sock(const struct sock *sk,
if (fallback_is_fatal)
goto dispose_child;
- if (ctx) {
- subflow_ulp_fallback(child, ctx);
- kfree_rcu(ctx, rcu);
- }
+ subflow_drop_ctx(child);
goto out;
}
@@ -537,6 +548,7 @@ static struct sock *subflow_syn_recv_sock(const struct sock *sk,
return child;
dispose_child:
+ subflow_drop_ctx(child);
tcp_rsk(req)->drop_req = true;
tcp_send_active_reset(child, GFP_ATOMIC);
inet_csk_prepare_for_destroy_sock(child);
--
2.21.3
Powered by blists - more mailing lists