Commit ff823c79a5c33194c2e5594f7c4686ea3547910c ("usb: move children
to struct usb_port") forgot to consider the hub_disconnect sequence,
which releases ports before quiescing the hub, which will lead to a
use-after-free, since hub_quiesce() will try to disconnect ports'
children, which are already deallocated. Simple modprobe dummy_hcd &&
rmmod dummy_hcd will illustrate the problem.

This patch moves deallocation of hub's ports after hub_quiesce() call
in hub_disconnect().

Cc: Lan Tianyu <tianyu....@intel.com>
Signed-off-by: Alexander Shishkin <alexander.shish...@linux.intel.com>
---
 drivers/usb/core/hub.c |    6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/usb/core/hub.c b/drivers/usb/core/hub.c
index aa45e43..6dc41c6 100644
--- a/drivers/usb/core/hub.c
+++ b/drivers/usb/core/hub.c
@@ -1584,9 +1584,6 @@ static void hub_disconnect(struct usb_interface *intf)
        struct usb_device *hdev = interface_to_usbdev(intf);
        int i;
 
-       for (i = 0; i < hdev->maxchild; i++)
-               usb_hub_remove_port_device(hub, i + 1);
-
        /* Take the hub off the event list and don't let it be added again */
        spin_lock_irq(&hub_event_lock);
        if (!list_empty(&hub->event_list)) {
@@ -1601,6 +1598,9 @@ static void hub_disconnect(struct usb_interface *intf)
        hub_quiesce(hub, HUB_DISCONNECT);
 
        usb_set_intfdata (intf, NULL);
+
+       for (i = 0; i < hdev->maxchild; i++)
+               usb_hub_remove_port_device(hub, i + 1);
        hub->hdev->maxchild = 0;
 
        if (hub->hdev->speed == USB_SPEED_HIGH)
-- 
1.7.10.4

--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Reply via email to