On Thursday 13 January 2005 3:07 am, Declan Moriarty wrote: > It has been suggested I report this here. I am not on the list, so > please cc to me, but can supply any information necessary. The > particular implementation has 3 pairs of ports, of which only one has > actual usb ports attached. > > I have a box here with a Via kt400 chipset.
So that means a VT8235 southbridge, I guess. > Loading this module produces > hub 1-0: 1.0: Overcurrent change on port 1 > hub 1-0: 1.0: Overcurrent change on port 2 > hub 1-0: 1.0: Overcurrent change on port 2 > > And the last message gets repeated many times a second, pops up on any > console you turn to, to the extent that all consoles are unusable. That's a new failure mode. Is this with 2.6.11-rc1? And are any USB devices hooked up to those ports? The "uhci_hcd" driver should be running too. Something changed to make some EHCI controllers act wierdly; some boards will report overcurrent, others seem to trigger a chip bug (where the controller wrongly issues a "fatal error" interrupt). But I don't recall any reports of the overcurrent messages spewing like that, and none of the relevant changes (in the EHCI driver or in usbcore) are obviously related to those issues... > The way, in fact to load it is as follows > rmmod ehci_hcd (produces an error) > modprobe ehci_hcd > > Then all you have to do is hit 'up' arrow twice and return :-D > > I believe this module is for usb-2.0 transfers. The error implies > overcurrent and changes in that value while not being specific enough. It's an issue in the high speed root hub, yes. The hardware is reporting overcurrent events, which then clear themselves. Then they seem to thoughtfully re-appear. If your version of "lsusb" will show the hub descriptors (Debian's does, as does "usbutils" cvs at linux-usb.sourceforge.net) please send that info along ... else send the kernel messages you'll get from initializing that controller with CONFIG_USB_DEBUG enabled. If your root hub supports per-port power switching, then the kernel hub driver could do at least something about the overcurrent. If not, I can't think of anything the software can do. > Whether overcurrent is true or not (I doubt it), it should not > happen. Overcurrent is certainly not happening unless something is > switched high and low simultaneously :-/. > > The uhci_hcd module runs them without problems. Did somebody make the > mistake of believing a Via datasheet ;-)? VIA and documentation have thus far been more or less mutually exclusive; nobody has any docs to (dis)believe. However, so far the worst problem I recall hearing about with VT8235 EHCI has been that it loses IRQs. - Dave ------------------------------------------------------- The SF.Net email is sponsored by: Beat the post-holiday blues Get a FREE limited edition SourceForge.net t-shirt from ThinkGeek. It's fun and FREE -- well, almost....http://www.thinkgeek.com/sfshirt _______________________________________________ linux-usb-devel@lists.sourceforge.net To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
