On Thu, 3 Dec 2009, Marian wrote: > Thank you, Alan! > > You were right: My investigations of the usb snoop log bring up that there are > only control messages used. Even though you do not know the vendor´s USB > protocol (nor me), I hope to get some answers how the host can access the > responses of the device and how I can implement that with legancy libusb. > > Here is a draft interpretation of a sample section which represents the first > of many communication sequences: > > URB7: > - host sends an APL specific request packet (91 01 00 01 93) to the clock > (that´s for sure) > - device seems to respond with a setup packet? No, you're getting confused by the snooping software. For some strange reason it prints the setup packet contents in the "coming back" part of the report instead of the "going down" part. Since URB 7 is a control-OUT transfer, the device doesn't send any data back. > URB 8: > - host then starts an asynchronous read (am I right?) It's a control-IN (as opposed to an interrupt-IN). Yes, control transfers are asynchronous. > - device answers with the first APL-specific response packet (fore sure) > URB 9: > - host polls length of incoming data (am I right?) I don't know. The host is asking for _something_, but it might not be the incoming data length. > - device responses with 00 0b, wich seems to be the recent uplink packet length > URB 10: > - host sends an APL specific ACK packet (a1 01 00 01 a3) > - device seems to respond with a setup packet? Same as before. 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
