On Mon, 12 Mar 2007 15:31:48 -0400, Jack Tanner <ihok@xxxxxxxxxxx> wrote: > Is it evidence enough that sometimes it works and sometimes it doesn't > (on the same FC6 box, with no reboots in between), and that it always > works on two other Windows boxes? Close, but not there yet. It would be nice if you tried it on other systems with Linux, because -71 is a pretty good indicator of something broken physically. It means that low level bus tokens get lost, or a frame corruption is detected by CRCs. But sometimes devices cause it by becoming temporarily deaf if they process some command or other. The workaround is to identify the conditions under which it happens and find a way to avoid poking the device for a few milliseconds. Again, that is only if we rule out poor signal quality on your motherboard. > Er, running a logger on Windows I can do. I snooped on the web, and > there are lots of Snoopies. Is this the right one? > http://benoit.papillault.free.fr/usbsnoop/ I don't remember the exact URL, but probably. I usually start with that Swiss page at linux-usb.org and follow links. Most people use something called Snoopy Pro. There's a free as in beer verision. > The formulate the patch part is going to be difficult. I don't suppose I > could get away with just sending you some log files? :) The problem is, it's time-consuming, which is why I tried to off-load all the work on you. If I take on it, it's going to take like forever. I think you better take it to linux-usb-devel@xxxxxxxxxxxxxxxxxxxxx at this time. -- Pete -- fedora-devel-list mailing list fedora-devel-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/fedora-devel-list
