On Mon, 23 Mar 2009, Dave Mielke wrote: > Your statement regarding an even vs an odd number of packets being sent to the > device prompted me to try something. It indeed does seem to matter whether the > previous use of the device (before usbfs close/reopen) sent an odd vs an even > number of bulk writes to the device. Is that really what you meant, though, or > did you mean, rather, that a bulk write is really a two-packet oeration (like > preparation stuff followed by actual data)? If the latter, I assume the device > is counting packets incorrectly or something. A bulk write can contain an arbitrary number of data packets. The actual number depends on the length of the data and the maxpacket size of the bulk-out endpoint. Each data packet other than the last one will have size equal to the maxpacket size. The size of the last data packet might be equal to maxpacket or it might be smaller -- it might even be 0. I don't know what the maxpacket size is for your device, but let's say it is 64 (which is typical for full-speed devices). If the length of your Then the data packets sent bulk write is: will be: 0 1 packet of len 0 63 1 packet of len 63 64 1 packet of len 64 65 1 packet of len 64 and 1 packet of len 1 127 1 packet of len 64 and 1 packet of len 63 128 2 packets of len 64 129 2 packets of len 64 and 1 packet of len 1 You get the picture. Using this approach you can count up the total number of data packets sent. 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
