On Mon, 2 Jul 2012, Sarah Sharp wrote:
> > > > @@ -4705,6 +4759,8 @@ static int usb_reset_and_verify_device(struct usb_device *udev)
> > > > if (ret) {
> > > > dev_err(&udev->dev, "%s Failed to disable LPM\n.", __func__);
> > > > mutex_unlock(hcd->bandwidth_mutex);
> > > > + /* Re-enable LTM, since the default is disabled after reset. */
> > > > + usb_enable_ltm(udev);
> > >
> > > These enable calls don't belong all over the place in this function.
> > > They should occur in exactly one place, in hub_port_reset(), since they
> > > are needed after each reset.
>
> The Set Feature LTM_ENABLE request is only valid when the device is in
> the Configured state, so I can't push it down into hub_port_reset(). I
> can't even push it down to hub_port_init() because that function doesn't
> set the configuration.
Yes -- I realized it too the day after sending that message. We may
not even have the device descriptor when a reset occurs.
Still, it should be possible to use a single call in a single place.
> I could change hub_port_reset() to disable LPM and LTM before the port
> is reset, if the device is configured. Callers of hub_port_init() will
> need to re-enable LPM and LTM once the device is configured.
Or maybe add something to usb_set_configuration.
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
