On Sun, 29 May 2005, Arlet Ottens wrote: > Hello, > > While trying to write a linux driver for my USB device on kernel 2.4.30, I > noticed that doing an usb_unlink_urb on a bulk IN endpoint, and then > submitting a new urb later, causes the next packet of data on that > endpoint to be lost. I see the packet on my USB bus analyzer, but I don't > get the completion callback in the driver. > > It appears to be caused by the fact that Linux has performed the data > toggle on that endpoint even though there was never any data received. > When the packet comes in, it decides that it has the wrong toggle bit, and > discards it. > > I can work around this by replacing the usb_unlink_urb() by the following > sequence: > > usb_unlink_urb() > usb_submit_urb() > usb_unlink_urb() > > That works, but it's not very elegant.
You didn't say which driver you were using. Presumably it's the "alternate" (or "JE") uhci driver. The corresponding problem with the 2.6 uhci-hcd driver has been known and was fixed last year. See http://marc.theaimsgroup.com/?l=linux-usb-devel&m=109545628228560&w=2 If you're interested you could back-port the fix to 2.4. Alan Stern ------------------------------------------------------- This SF.Net email is sponsored by Yahoo. Introducing Yahoo! Search Developer Network - Create apps using Yahoo! Search APIs Find out how you can build Yahoo! directly into your own Applications - visit http://developer.yahoo.net/?fr=offad-ysdn-ostg-q22005 _______________________________________________ linux-usb-devel@lists.sourceforge.net To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
