On Mon, 25 Apr 2011, Sarah Sharp wrote: > Random question time. :) > > Do USB device drivers have to cancel all URBs that are pending for a > device before calling usb_reset_device? Will the drivers have to wait > for the URBs to complete before resetting the device? There is no requirement that drivers cancel all URBs before a reset occurs. Indeed, some drivers can't because they have no way to know when another driver bound to another interface on the same device wants to do a reset, i.e., they have no pre_reset callback. Of course, an URB that remains active during a reset is quite likely to fail. > The specific random question I got asked by the chipset folks is: > > "Before issuing a Port Reset, is there any condition that the driver > should remove the TD associated with the device?" > > Not quite sure what to make of that, since I don't know EHCI internals. I'm not quite sure what to make of it either. TDs aren't associated with devices, or at least, not in a one-to-one relationship. Anyway, the answer appears to be "No". > The second question relates to the CLEAR_TT_BUFFER request, and it only > applies to an Intel EHCI controller with a built-in rate matching hub > (RMH). The chipset folks want to know if a CLEAR_TT_BUFFER request > might be used before a port suspend or a port reset. Yes, it might. In fact, as far as I can tell, a Clear-TT-Buffer request might still be in progress when a suspend or reset is issued for one of the ports associated with that TT. > It looks like the EHCI driver automatically clears the TT buffer, except > when the host controller has a Transaction Translator built into the > root hub, or we need to clear a TT buffer on an external HS hub. You got that reversed. TT buffers on non-root high-speed hubs do get cleared. In theory a root hub's built-in TT (if there is one) should have its buffers cleared as well, but there doesn't appear to be a standard way of doing it so we don't even try. > I > assume that means we won't clear the TT buffer for the EHCI roothub, > since this platform has a rate matching hub? On the platform you're talking about, the root hub does not have a TT. On the other hand, we _will_ clear the TT buffers in the rate-matching hub when they need it. I trust these answers will dismay your chipset people... :-) 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
