Hi Kris > I had a brief ponder over this. > A couple of years ago I modified & expanded Atmel's original MSD project > (for SAM7 which is ARM7) to compile under GCC, but since it's been so > long I'm somewhat fuzzy. I do recall the class USBGenericRequest. > I can't readily access my project files right now to verify my > conjecture but by looking at your log, I think your device might be > using the USBGenericRequest Class to query/report the button press. > > IIRC USBGenericRequest can specify which pipe will be used for the data > transfer, its length/direction, recipient etc. I'm not sure if I understood completely. I found the code of Atmel's USB stack and it looks to me that USBGenericRequest is just Atmel's implementation of creating USB messages. Linux calls them USB Request Blocks (URB). I don't know what the USB standard calls them, I don't have access it. > I do recall that Snoopy didn't log nor display string descriptors et al, > but I thought it did display what's going on in the control pipe (ie. Shouldn't string descriptors show up in lsusb -v? > EP0). If so, see if such requests are used. Is this traffic rate also > observed when you're mounted to a Linux host ? (I dunno if Snoopy runs I only logged USB traffic on the windows box so far. usbsnoop/SnoopyPro are windows only, but I read about ways to monitor USB under linux using debugfs/usbmon. I'll try that next week and post the results. > under Linux, must check that). Else it could be that WIN's standard > driver somehow was told that it should issue some poll on the control > pipe, I'd have thought using GET_STATUS in USBGenericRequest... That would be great. In linux there is a usb_get_status() drivers/usb/core/message.c which would make very simple to poll the button. Unfortunately, as you'll see in my recent post to Daniel, no commands are sent on the control pipe after initialisation. > Also, it's not clear whether the button poll is when that little WIN > button app is installed, or whether usbstor.sys itself is doing that ? Good point, I haven't monitored the device before running the app. I'll try to do that on a clean box and see if it makes a difference. Thanks, Chris -- echo mailto: NOSPAM !#$.'<*>'|sed 's. ..'|tr "<*> !#:2" org@fr33z3 -- To unsubscribe from this list: send an email with "unsubscribe kernelnewbies" to ecartis@xxxxxxxxxxxx Please read the FAQ at http://kernelnewbies.org/FAQ
