On Wed, 17 Jun 2015, Stefan Koch wrote:
> The kernel supports the device authorization because of wireless USB.
> These is usable for wired USB devices, too.
> These new interface authorization allows to enable or disable
> individual interfaces instead a whole device.
>
> To avoid side effects the driver probing process
> for the interface and it's siblings is triggered after each authorization.
No, don't do this. Allow the user to trigger probing, don't do it
automatically.
> Signed-off-by: Stefan Koch <skoch@xxxxxxx>
> ---
> drivers/usb/core/hub.c | 77 +++++++++++++++++++++++++++++++++++++++++++++++++-
hub.c deals with devices, not interfaces. The new routines belong in
message.c, along with all the other routines that manage interfaces.
> drivers/usb/core/usb.h | 2 ++
> 2 files changed, 78 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/usb/core/hub.c b/drivers/usb/core/hub.c
> index 3b71516..6837281 100644
> --- a/drivers/usb/core/hub.c
> +++ b/drivers/usb/core/hub.c
> @@ -2528,6 +2528,82 @@ out_unauthorized:
> return 0;
> }
>
> +/*
> + * usb_deauthorize_interface - deauthorize an USB interface
> + *
> + * @intf: USB interface structure
> + *
> + * Returns: 0 at success, <0 at failure
> + */
> +int usb_deauthorize_interface(struct usb_interface *intf)
> +{
> + struct device *dev = &intf->dev;
> +
> + if (!dev || !dev->parent)
> + return -ENODEV;
This can never happen. Get rid of it.
> +
> + device_lock(dev->parent);
> +
> + if (intf->authorized) {
> + intf->unregistering = 1;
Don't set intf->unregistering. You aren't unregistering the interface;
you are de-authorizing it. They are different things.
> + intf->authorized = 0;
> + usb_forced_unbind_intf(intf);
> + }
> +
> + device_unlock(dev->parent);
> +
> + return 0;
> +}
Alan Stern
--
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
