On Mon, 2004-01-26 at 17:18, Alan Stern wrote: > Let me see if I understand this. If you plug in your device, run the > test, and unplug it then your driver works okay and the outstanding URB is > unlinked. But then the next time you plug anything in, even the mouse, it > fails to enumerate.
Yes. > Does this failure happen always or only sometimes? Yes (tested about 15 times). > On the other hand, if you plug in the mouse and then unplug it, or plug in > your device but don't run the test before unplugging it, then the next > time you plug something in it enumerates okay. > > Is that right? Exactly. > And if it is, do you think the problem is caused by unplugging while an > URB is still in progress, by some unusual electrical characteristic of > your device (or its cable), or by a combination of both? I can not claim that the device/cable is behaving/build according to the USB spec, therefore it might be its electrical characteristic is part of the problem. But this type of device/cable has been in use for thousands of hours on Windows machines without this specific problem (disconnect during transfer -> enumeration fails). As mentioned earlier, I made a similar test with exactly the same hardware with the 2.4.x usb-uhci module without problems. And also the fact that the 2.6 uhci-hcd did work with the additional interrupts from the soundcard makes me believe that the controller hardware is generally capable to deal with it. This pretty much convinces me that fault is to be searched in the controller/hub driver. Axel ------------------------------------------------------- The SF.Net email is sponsored by EclipseCon 2004 Premiere Conference on Open Tools Development and Integration See the breadth of Eclipse activity. February 3-5 in Anaheim, CA. http://www.eclipsecon.org/osdn _______________________________________________ [EMAIL PROTECTED] To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
