netdev - [PATCH AUTOSEL 4.14 19/33] net: add {READ|WRITE}_ONCE() annotations on ->rskq_accept

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [day] [month] [year] [list]

Message-Id: <20191026132110.4026-19-sashal@kernel.org>
Date:   Sat, 26 Oct 2019 09:20:56 -0400
From:   Sasha Levin <sashal@...nel.org>
To:     linux-kernel@...r.kernel.org, stable@...r.kernel.org
Cc:     Eric Dumazet <edumazet@...gle.com>,
        Jakub Kicinski <jakub.kicinski@...ronome.com>,
        Sasha Levin <sashal@...nel.org>, netdev@...r.kernel.org
Subject: [PATCH AUTOSEL 4.14 19/33] net: add {READ|WRITE}_ONCE() annotations on ->rskq_accept_head

From: Eric Dumazet <edumazet@...gle.com>

[ Upstream commit 60b173ca3d1cd1782bd0096dc17298ec242f6fb1 ]

reqsk_queue_empty() is called from inet_csk_listen_poll() while
other cpus might write ->rskq_accept_head value.

Use {READ|WRITE}_ONCE() to avoid compiler tricks
and potential KCSAN splats.

Fixes: fff1f3001cc5 ("tcp: add a spinlock to protect struct request_sock_queue")
Signed-off-by: Eric Dumazet <edumazet@...gle.com>
Signed-off-by: Jakub Kicinski <jakub.kicinski@...ronome.com>
Signed-off-by: Sasha Levin <sashal@...nel.org>
---
 drivers/xen/pvcalls-back.c      | 2 +-
 include/net/request_sock.h      | 4 ++--
 net/ipv4/inet_connection_sock.c | 2 +-
 3 files changed, 4 insertions(+), 4 deletions(-)

diff --git a/drivers/xen/pvcalls-back.c b/drivers/xen/pvcalls-back.c
index abd6dbc29ac28..58be15c27b6d6 100644
--- a/drivers/xen/pvcalls-back.c
+++ b/drivers/xen/pvcalls-back.c
@@ -792,7 +792,7 @@ static int pvcalls_back_poll(struct xenbus_device *dev,
 	mappass->reqcopy = *req;
 	icsk = inet_csk(mappass->sock->sk);
 	queue = &icsk->icsk_accept_queue;
-	data = queue->rskq_accept_head != NULL;
+	data = READ_ONCE(queue->rskq_accept_head) != NULL;
 	if (data) {
 		mappass->reqcopy.cmd = 0;
 		ret = 0;
diff --git a/include/net/request_sock.h b/include/net/request_sock.h
index 23e22054aa60d..04aa2c7d35c4e 100644
--- a/include/net/request_sock.h
+++ b/include/net/request_sock.h
@@ -181,7 +181,7 @@ void reqsk_fastopen_remove(struct sock *sk, struct request_sock *req,
 
 static inline bool reqsk_queue_empty(const struct request_sock_queue *queue)
 {
-	return queue->rskq_accept_head == NULL;
+	return READ_ONCE(queue->rskq_accept_head) == NULL;
 }
 
 static inline struct request_sock *reqsk_queue_remove(struct request_sock_queue *queue,
@@ -193,7 +193,7 @@ static inline struct request_sock *reqsk_queue_remove(struct request_sock_queue
 	req = queue->rskq_accept_head;
 	if (req) {
 		sk_acceptq_removed(parent);
-		queue->rskq_accept_head = req->dl_next;
+		WRITE_ONCE(queue->rskq_accept_head, req->dl_next);
 		if (queue->rskq_accept_head == NULL)
 			queue->rskq_accept_tail = NULL;
 	}
diff --git a/net/ipv4/inet_connection_sock.c b/net/ipv4/inet_connection_sock.c
index 9d6b172caf6cf..8d48f68e2a772 100644
--- a/net/ipv4/inet_connection_sock.c
+++ b/net/ipv4/inet_connection_sock.c
@@ -936,7 +936,7 @@ struct sock *inet_csk_reqsk_queue_add(struct sock *sk,
 		req->sk = child;
 		req->dl_next = NULL;
 		if (queue->rskq_accept_head == NULL)
-			queue->rskq_accept_head = req;
+			WRITE_ONCE(queue->rskq_accept_head, req);
 		else
 			queue->rskq_accept_tail->dl_next = req;
 		queue->rskq_accept_tail = req;
-- 
2.20.1