Re: [Libusb-devel] Strange control-out requests in usbmon / libusb 0.1.12

[Tim R. <ti...@pr...>]

Sarah Messer wrote:
> I've had some trouble with my scopes under libusb 0.1.12, as mentioned a few times in the past.  Learned to read usbmon files and started looking at control messages (so that I could write some & emulate as-much-as-possible what was happening when I used the vendor-provided software on a Windows box).  After writing some code, I tried running usbmon on my Linux machine to see how close I was to the Tektronix (vendor) sequence of USB commands...  The file starts recording before the device is powered on, so all setup & attachment info should be here...  I haven't run any other programs, so this should _only_ be the setup & attachment process.
>
> One of the first lines in the usbmon output file is:
>
> ffff810019363b80 3144200568 S Co:2:001:0 s 23 01 0010 0002 0000 0
>
> This is apparently a Control-Out packet on bus 2, dev 1, endpoint 0 (all fine), that usbmon successfully captured (that's the "s").  The confusing part is the "23 01", which I parse (via USB 2.0 spec section 9.3.1, p. 248 and USBTMC 1.0 spec section ) to be requesting a Class-specific Host-to-device, recipient=Other, INITIATE_ABORT_BULK_OUT request.
>   

I believe, if you look further, you'll find that your device is device 
9, not device 1.  I'm guessing that what you have posted there is 
actually a command being sent to the USB hub that holds your device, 
managing the establishment of communcations.  When the device number 
changes to 2:009:0, THAT'S where you should start looking.

-- 
Tim Roberts, ti...@pr...
Providenza & Boekelheide, Inc.