Bluetooth: timeout handling in new connect procedure

Currently, when trying to connect to already paired device that just
rotated its RPA MAC address, old address would be used and connection
would fail. In order to fix that, kernel must scan and receive
advertisement with fresh RPA before connecting.

This patch makes sure that when new procedure is in use, and we're stuck
in scan phase because no advertisement was received and timeout happened,
or app decided to close socket, scan whitelist gets properly cleaned up.

Signed-off-by: Jakub Pawlowski <jpawlowski@google.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>

diff --git a/net/bluetooth/hci_conn.c b/net/bluetooth/hci_conn.c
index 85c6aa5d5bbc0332817ee84f16802d1e93019c69..b4548c739a6475446d643bd5b01ab8627ef1f08e 100644 (file)
--- a/net/bluetooth/hci_conn.c
+++ b/net/bluetooth/hci_conn.c
@@ -382,8 +382,12 @@ static void hci_conn_timeout(struct work_struct *work)
                if (conn->out) {
                        if (conn->type == ACL_LINK)
                                hci_acl_create_connection_cancel(conn);
-                       else if (conn->type == LE_LINK)
-                               hci_le_create_connection_cancel(conn);
+                       else if (conn->type == LE_LINK) {
+                               if (test_bit(HCI_CONN_SCANNING, &conn->flags))
+                                       hci_connect_le_scan_remove(conn);
+                               else
+                                       hci_le_create_connection_cancel(conn);
+                       }
                } else if (conn->type == SCO_LINK || conn->type == ESCO_LINK) {
                        hci_reject_sco(conn);
                }