On Fri, 19 Jan 2007, David Brownell wrote: > On Friday 12 January 2007 2:12 pm, you wrote: > > > > Phil: > > > > I haven't tried testing your patch. But I have tried using the existing > > gadgetfs with net2280 and uhci-hcd. There was no trouble running test #10 > > with the usb program or with g_zero instead of gadgetfs. I can't imagine > > what's going wrong on your system. > > What kernel were you using?
2.6.20-rc4 plus gregkh-all-2.6.20-rc4.patch plus the AIO patch I wrote, on x86 with FC6. > Given that the problem I observed was that kill_fasync() didn't get signals > up to userspace, knowing what kernel you used would help narrow down when > (and where) the kill_fasync() thing happened. > > I observed it to not work on 2.6.20 rc4 and rc4 on an ARM920T ... and I think > that Phil was using x86. If you had tested it on an earlier kernel with x86, > a "git bisect" could help track down when kill_fasync() broke ... It doesn't look like this data point will help your search very much. Note that my test was not particularly thorough; I simply ran test #10 with a count of 2000, and I did it only once. By the way, here's another possibly related matter which I never bothered to bring up before because it didn't seem very important. The usb.c + gadgetfs combination doesn't (or at least, didn't -- maybe you have already fixed it) handle disconnects properly. I didn't save any logs, but my impression was that the program was trying to continue using a stale file descriptor. It's easy to test whether this happens. Run usb.c, run one of the testusb tests, then unplug and replug the cable (I don't think there's any way to simulate this with dummy-hcd, but there ought to be) and then see what shows up in the stderr for usb.c and in the system log. Alan Stern ------------------------------------------------------------------------- Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys - and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV _______________________________________________ linux-usb-devel@lists.sourceforge.net To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
