On Sun, 29 Nov 2009, Steve Calfee wrote: > Hi Alan, > > You are right about the correct solution. But in a windows dominated > world where windows drivers can work around gadget bugs this is > difficult. Windows does not dominate the world to the extent that devices can blatantly violate the spec and still obtain the approval of the USBIF. > I believe the way windows does this is the driver will send > the control packet to the gadget to set the transfer size, then change > the host drivers copy of wMaxpacketsize for the isoc endpoint to this > new value. At that point things are back in spec. And normal > scheduling should work. Perhaps so. > However, I am not sure that Linux allows drivers to change > wMaxPacketSize? How far do we go to support buggy devices? There's nothing to stop it. On the other hand, it isn't guaranteed to work. Especially if you're dealing with an xHCI controller. 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
