On Wed, Dec 05, 2012 at 01:50:46PM -0500, Alan Stern wrote: > On Tue, 4 Dec 2012, Sarah Sharp wrote: > > > The USB core hub thread (khubd) is designed with external USB hubs in > > mind. It expects that if a port status change bit is set, the hub will > > continue to send a notification through the hub status data transfer. > > Basically, it expects hub notifications to be level-triggered. > > > > The xHCI host controller is designed to be edge-triggered. When all > > port status change bits are clear, and a new change bit is set, the xHC > > will generate a Port Status Change Event. If another change bit is set > > in the same port status register before the first bit is cleared, it > > will not send another event. > > This explanation is okay, but in general I often find the terms > "edge-triggered" and "level-triggered" to be rather confusing in this > sort of context. > > The real issue here is what input will turn on the IRQ signal (or cause > an event to be raised). With xHCI, that input is the "or" of all the > status-change bits in a port register -- a new IRQ won't be generated > until the "or" changes from 0 to 1. With EHCI, on the other hand, a 0 > -> 1 change in any of the bits will cause a new IRQ. > > Both controllers use edge-triggered IRQs. The difference lies in what > edges they are monitoring. Yes, that's a better expression of the idea. In fact, the xHCI spec's diagram of the port status change triggers includes all change bits feeding to an OR gate. :) > > + /* > > + * xHCI portsc bits are level-triggered. An event is sent on the first > > + * change bit, but won't be sent if another change bit is still set. > > + * Poll to avoid losing change bits. > > + */ > > The first two lines of this comment don't express the idea very well. > I suggest something more like: > > /* > * xHCI port-status-change events occur when the "or" of all the > * status-change bits in the portsc register changes from 0 to 1. > * New status changes won't cause an event if any other change > * bits are still set. When an event occurs, switch over to > * polling to avoid losing status changes. > */ Sure, that sounds fine. Sarah Sharp -- 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
