Received-SPF: pass (google.com: domain of linux-kernel+bounces-30065-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:45e3:2400::1 as permitted sender) client-ip=2604:1380:45e3:2400::1;
From: Hillf Danton <hdanton@sina.com>
To: shaozhengchao <shaozhengchao@huawei.com>
Cc: linux-kernel@vger.kernel.org,
	netdev@vger.kernel.org,
	Davidlohr Bueso <dave@stgolabs.net>,
	Manfred Spraul <manfred@colorfullife.com>,
	jack@suse.cz
Subject: Re: [PATCH v2] ipc/mqueue: fix potential sleeping issue in mqueue_flush_file
Date: Thu, 18 Jan 2024 19:46:31 +0800
Message-Id: <20240118114631.1490-1-hdanton@sina.com>
In-Reply-To: <fee3ec1c-5af6-aad2-c0d0-843de59494a7@huawei.com>
References: <20231220021208.2634523-1-shaozhengchao@huawei.com>
Precedence: bulk
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit

On 2023/12/20 10:12, Zhengchao Shao wrote:
> I analyze the potential sleeping issue of the following processes:
> Thread A                                Thread B
> ...                                     netlink_create  //ref = 1
> do_mq_notify                            ...
>    sock = netlink_getsockbyfilp          ...     //ref = 2
>    info->notify_sock = sock;             ...
> ...                                     netlink_sendmsg
> ...                                       skb = netlink_alloc_large_skb  //skb->head is vmalloced
> ...                                       netlink_unicast
> ...                                         sk = netlink_getsockbyportid //ref = 3
> ...                                         netlink_sendskb
> ...                                           __netlink_sendskb
> ...                                             skb_queue_tail //put skb to sk_receive_queue
> ...                                         sock_put //ref = 2
> ...                                     ...
> ...                                     netlink_release
> ...                                       deferred_put_nlk_sk //ref = 1
> mqueue_flush_file
>    spin_lock
>    remove_notification
>      netlink_sendskb
>        sock_put  //ref = 0
>          sk_free
>            ...
>            __sk_destruct
>              netlink_sock_destruct
>                skb_queue_purge  //get skb from sk_receive_queue
>                  ...
>                  __skb_queue_purge_reason
>                    kfree_skb_reason
>                      __kfree_skb
>                      ...
>                      skb_release_all
>                        skb_release_head_state
>                          netlink_skb_destructor
>                            vfree(skb->head)  //sleeping while holding spinlock
> 
> In netlink_sendmsg, if the memory pointed to by skb->head is allocated by
> vmalloc, and is put to sk_receive_queue queue, also the skb is not freed.
> When the mqueue executes flush, the sleeping bug will occur. Use mutex
> lock instead of spin lock in mqueue_flush_file.

It makes no sense to replace spinlock with mutex just for putting sock.

Only for thoughts.

--- x/ipc/mqueue.c
+++ y/ipc/mqueue.c
@@ -663,12 +663,17 @@ static ssize_t mqueue_read_file(struct f
 static int mqueue_flush_file(struct file *filp, fl_owner_t id)
 {
 	struct mqueue_inode_info *info = MQUEUE_I(file_inode(filp));
+	struct sock *sk = NULL;
 
 	spin_lock(&info->lock);
-	if (task_tgid(current) == info->notify_owner)
+	if (task_tgid(current) == info->notify_owner) {
+		sk = info->notify_sock;
+		sock_hold(sk);
 		remove_notification(info);
-
+	}
 	spin_unlock(&info->lock);
+	if (sk)
+		sock_put(sk);
 	return 0;
 }