On Tuesday 18 July 2006 4:48 am, [EMAIL PROTECTED] wrote: > >So the real fix would involve finding out how the "eject HC" (CardBus?) > >case is interacting wrongly with the other cleanup code, and then making > >that behave properly. > > Thanks, Dave. > I will try to find the code and expect to report it if I could find.
An anti-oops band-aid would be OK too, if updated as I described earlier. > By the way "eject HC" is that HC is pci board inserted in the machine > and it is ejected by stopping to provide electric power to the board. At that point, if the CPU issues a read to an OHCI register what will get reported? One option would be a PCI bus error; you didn't show any of that happening in the limited trace info you sent, and I'm not quite sure how those would be handled in Linux. Another option is to do what CardBus does: return all-ones on the reads. That option is expected in the code in several places. If your test setup produces a third error mode, that might be trouble. If that's a PCI hotplug kind of setup, I'd be kind of surprised that the driver doesn't get cleanly shut down. - Dave ------------------------------------------------------------------------- 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
