On Sat, 29 Jan 2005, Chris Clayton wrote: > Thanks for the reply Alan. > > On Saturday 29 Jan 2005 15:42, Alan Stern wrote:
> > A timeout like that is more likely to result from the _device_ being old > > and slow than from the _host_ being old and slow. > > I see what you mean, but why then would the device I am experimenting with > (and others) work fine with a 2.4 kernel and Windows ME and, when attached > through a cardbus USB2 adapter, with the same 2.6.10 kernel (and, I assume, > the same ohci-hcd driver)? In any case, both the hub and the wireless adapter > are only about 24 months old. The devices also work with my (much newer) > desktop system, which is also running a 2.6.10 kernel. I didn't mean to imply that this was the source of your problem; I only intended to point out that your reasoning about relative speeds was illogical. > > You can try using the patch in this message: > > Thanks, but it made no difference. Whilst rebuilding, I switched verbose usb > debugging on (again) and have attached a couple of additional diagnostic > shots from dmesg. The first is from attaching the (unpowered) 4-port hub > immediately after a reboot. The second also follows a reboot but the device > attached is a Belkin F5D6050 wireless adapter (the drivers are built out of > kernel and are the ones from http://at76c503a.berlios.de/index.html. they > work fine with 2.4 and when the adapter is attached via cardbus/USB2) > > You can also try using the "old_scheme_first=y" module parameter for > > Yes, I saw that tip on the linux-usb home page and gave it a try even though > it was advice specific to 2.6.10. I have the prblem with other 2.6 kernels > too - I've now tried 2.6.5, 2.6.6 and 2.6.9 too. I've also tried acpi=off, > acpi=noirq, pci=routeirq and what I think are valid combinations thereof but > to no avail. Yes, your logs seem to indicate that interrupts are getting through all right. If that weren't so you'd be getting error messages about "Device not accepting address" instead. > The additional diags I mention above are attached. They reinforce my impression that something funny is happening with your OHCI hardware and driver. Alan Stern ------------------------------------------------------- This SF.Net email is sponsored by: IntelliVIEW -- Interactive Reporting Tool for open source databases. Create drag-&-drop reports. Save time by over 75%! Publish reports on the web. Export to DOC, XLS, RTF, etc. Download a FREE copy at http://www.intelliview.com/go/osdn_nl _______________________________________________ linux-usb-devel@lists.sourceforge.net To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
