Alan Stern wrote: > Finally, one of the things we discussed was allowing drivers to specify > urb->start_frame for new streams (to get precise synchronization of two > devices, for instance). For the case that I mentioned, setting start_frame would not be necessary; the only requirement is that the streams do not drift apart because of gaps. (The audio driver always uses ASAP.) > But with xHCI, the hardware does all the scheduling as soon as an > altsetting is installed. As far as I understand the spec, this is the intended (or at least allowed) behaviour for all HCs. And it would be a benefit for drivers if a successful set_interface guaranteed that submitting isoc URBs will not fail due to missing bandwidth. > How can you tell it, later on when the driver submits the first URB, > what the start frame should be? When a stream use an interval greater than one, the scheduling accuracy cannot be guaranteed to be better than that. So allow the HCD to round start_frame to whatever fits the actual schedule. Best regards, Clemens -- 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