USB HCD: avoid duplicate local_irq_disable()

Arnd Bergmann wrote:

usb_hcd_flush_endpoint() has a retry loop that starts with a spin_lock_irq(),
but only gives up the spinlock, not the irq_disable before jumping to the
rescan label.

Alan Stern:

I agree with your sentiment, but it would be better to solve this
problem without using local_irq_disable().

Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Cc: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>