On Tue, 13 Jun 2006, Rainer Becker wrote: > Alan Stern schrieb: > > On Fri, 2 Jun 2006, Rainer Becker wrote: > > > >>> I'm getting a little confused. Is "the problem" that you have so many > >>> messages in the log and it takes a long time, or is "the problem" that > >>> the > >>> port cannot be enabled? This patch won't allow the port to be enabled. > >> This repeater uses a 4 port hub chip, but only port 4 is connected to the > >> outside. So the other ports should be disabled. > >> > >> The USB bus is slowed down by these retries and therefor the patch is > >> necessary. > > > > Why are the other ports enabled in the first place? Linux won't try to > > enable a port unless the port indicates that a device is connected to it. > > > > Try setting CONFIG_USB_DEBUG in the kernel configuration and see what the > > kernel log shows. > > > > Alan Stern > > > I've been on holiday for a week. Therefore I couldn't answer immediately. > > Here is the kernel log with CONFIG_USB_DEBUG set. I've attached all the > messages > I got. they may be repeated. > > Two of this repeater cables (hubs) are connected in series.
What does the log show when your patch is installed? It looks like the real problem here is a hardware bug in the repeater cables' hubs. They keep indicating connect-change and reset-change events for ports that aren't connected. This is a violation of the USB specification. Alan Stern _______________________________________________ linux-usb-devel@lists.sourceforge.net To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
