On Wed, 26 Jan 2011, Micah Elizabeth Scott wrote:
> > In my opinion, the best way to deal with this is simply to reject
> > high-speed hubs that don't have any TTs. Presumably there aren't many
> > devices like that floating around. Do you agree?
> >
> > The following patch implements this policy. It should eliminate your
> > bug.
>
> Perhaps this is selfish, but I'd strongly advocate against this patch
> since it will break all of our virtual hubs here at VMware :)
Come to think of it, the patch will also break the root hubs we create
for EHCI controllers.
> We emulate USB hubs in order to dynamically extend the number of ports
> available to a virtual machine. We never create hubs with TTs, partly
> so we can avoid the performance overhead and implementation complexity
> of them, and partly because we really have no need for them. (We can
> always have both a high-speed and a full/low-speed hub available, and
> plug devices into the correct one.)
The alternative is to reject low- and full-speed devices when they are
plugged into a high-speed hub with no TTs. How does this look?
Alan Stern
Index: usb-2.6/drivers/usb/core/hub.c
===================================================================
--- usb-2.6.orig/drivers/usb/core/hub.c
+++ usb-2.6/drivers/usb/core/hub.c
@@ -2732,6 +2732,11 @@ hub_port_init (struct usb_hub *hub, stru
udev->ttport = hdev->ttport;
} else if (udev->speed != USB_SPEED_HIGH
&& hdev->speed == USB_SPEED_HIGH) {
+ if (!hub->tt.hub) {
+ dev_err(&udev->dev, "parent hub has no TT\n");
+ retval = -EINVAL;
+ goto fail;
+ }
udev->tt = &hub->tt;
udev->ttport = port1;
}
--
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
