Re: [PATCH] USB: fix deadlock in HCD code

[Oliver Neukum]

Am Mittwoch 21 Mai 2008 15:27:50 schrieb Jiri Kosina:
> On Wed, 21 May 2008, Oliver Neukum wrote:
> 
> > > hcd_urb_list_lock is used for synchronization between IRQ and non-IRQ 
> > > contexts, so the non-IRQ context has to disable IRQs in order to prevent 
> > > deadlocking with IRQ context.
> > Which non-irq context is that?
> 
> One example -- assume usb_submit_urb() called from non-IRQ context. Then
> 
> usb_hcd_submit_urb() -> rh_urb_enqueue() -> rh_queue_status() -> 
> usb_hcd_link_urb_to_ep().
> 

This turns out not to be the case. Interrupts are disabled.

static int rh_queue_status (struct usb_hcd *hcd, struct urb *urb)
{
	int		retval;
	unsigned long	flags;
	int		len = 1 + (urb->dev->maxchild / 8);

	spin_lock_irqsave (&hcd_root_hub_lock, flags);
	if (hcd->status_urb || urb->transfer_buffer_length < len) {
		dev_dbg (hcd->self.controller, "not queuing rh status urb\n");
		retval = -EINVAL;
		goto done;
	}

	retval = usb_hcd_link_urb_to_ep(hcd, urb);

I'll investigate.

	Regards
		Oliver
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/