linux-kernel - [PATCH net,v2] team: Fix ABBA deadlock caused by race in team_del

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

Message-Id: <20240706041329.96637-1-aha310510@gmail.com>
Date: Sat,  6 Jul 2024 13:13:29 +0900
From: Jeongjun Park <aha310510@...il.com>
To: jiri@...nulli.us
Cc: syzbot+705c61d60b091ef42c04@...kaller.appspotmail.com,
	davem@...emloft.net,
	edumazet@...gle.com,
	kuba@...nel.org,
	linux-kernel@...r.kernel.org,
	netdev@...r.kernel.org,
	pabeni@...hat.com,
	syzkaller-bugs@...glegroups.com,
	Jeongjun Park <aha310510@...il.com>
Subject: [PATCH net,v2] team: Fix ABBA deadlock caused by race in team_del_slave

       CPU0                    CPU1
       ----                    ----
  lock(&rdev->wiphy.mtx);
                               lock(team->team_lock_key#4);
                               lock(&rdev->wiphy.mtx);
  lock(team->team_lock_key#4);

Deadlock occurs due to the above scenario. Therefore, you can prevent
deadlock by briefly releasing the lock before calling dev_open() in
team_port_add() and locking it again after it returns.

Reported-and-tested-by: syzbot+705c61d60b091ef42c04@...kaller.appspotmail.com
Fixes: 3d249d4ca7d0 ("net: introduce ethernet teaming device")
Signed-off-by: Jeongjun Park <aha310510@...il.com>
---
 drivers/net/team/team_core.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/net/team/team_core.c b/drivers/net/team/team_core.c
index ab1935a4aa2c..245566a1875d 100644
--- a/drivers/net/team/team_core.c
+++ b/drivers/net/team/team_core.c
@@ -1213,7 +1213,9 @@ static int team_port_add(struct team *team, struct net_device *port_dev,
 		goto err_port_enter;
 	}
 
+	mutex_unlock(&team->lock);
 	err = dev_open(port_dev, extack);
+	mutex_lock(&team->lock);
 	if (err) {
 		netdev_dbg(dev, "Device %s opening failed\n",
 			   portname);
--