bluetooth: eliminate the potential race condition when removing the HCI controller

There is a possible race condition vulnerability between issuing a HCI
command and removing the cont.  Specifically, functions hci_req_sync()
and hci_dev_do_close() can race each other like below:

thread-A in hci_req_sync()      |   thread-B in hci_dev_do_close()
                                |   hci_req_sync_lock(hdev);
test_bit(HCI_UP, &hdev->flags); |
...                             |   test_and_clear_bit(HCI_UP, &hdev->flags)
hci_req_sync_lock(hdev);        |
                                |
In this commit we alter the sequence in function hci_req_sync(). Hence,
the thread-A cannot issue th.
Signed-off-by: Lin Ma <linma@zju.edu.cn>
Cc: Marcel Holtmann <marcel@holtmann.org>
Fixes: 7c6a329e ("[Bluetooth] Fix regression from using default link policy")
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

net/bluetooth/hci_request.c

@@ -272,12 +272,16 @@ int hci_req_sync(struct hci_dev *hdev, int (*req)(struct hci_request *req,
 {
 	int ret;
 
-	if (!test_bit(HCI_UP, &hdev->flags))
-		return -ENETDOWN;
-
 	/* Serialize all requests */
 	hci_req_sync_lock(hdev);
-	ret = __hci_req_sync(hdev, req, opt, timeout, hci_status);
+	/* check the state after obtaing the lock to protect the HCI_UP
+	 * against any races from hci_dev_do_close when the controller
+	 * gets removed.
+	 */
+	if (test_bit(HCI_UP, &hdev->flags))
+		ret = __hci_req_sync(hdev, req, opt, timeout, hci_status);
+	else
+		ret = -ENETDOWN;
 	hci_req_sync_unlock(hdev);
 
 	return ret;