Hi,
it seems to me that usb_serial_suspend() is buggy. If a subdriver
vetos a suspension we already have killed all URBs, which we cannot
undo in the generic layer because we don't know what damage we have
done. Comments?
Regards
Oliver
--
commit bb521ef5aafd9745e3a04970e90f4db7cbc20416
Author: Oliver Neukum <oliver@xxxxxxxxxx>
Date: Mon Jul 20 23:15:51 2009 +0200
usb: fix order of calls in usb serial suspend
if a subdriver has an additional suspend method, it must be called
first to allow the subdriver to return -EBUSY, because the second
half cannot be easily undone.
diff --git a/drivers/usb/serial/usb-serial.c b/drivers/usb/serial/usb-serial.c
index c6f69c0..b739516 100644
--- a/drivers/usb/serial/usb-serial.c
+++ b/drivers/usb/serial/usb-serial.c
@@ -1180,15 +1180,19 @@ int usb_serial_suspend(struct usb_interface *intf, pm_message_t message)
serial->suspending = 1;
+ if (serial->type->suspend) {
+ r = serial->type->suspend(serial, message);
+ if (r < 0)
+ goto err_out;
+ }
+
for (i = 0; i < serial->num_ports; ++i) {
port = serial->port[i];
if (port)
kill_traffic(port);
}
- if (serial->type->suspend)
- r = serial->type->suspend(serial, message);
-
+err_out:
return r;
}
EXPORT_SYMBOL(usb_serial_suspend);
--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
