Hi
I have a problem with my Microsoft Natural USB Keyboard.
Since I moved from kernel 2.4.18-rc2-ac1 to 2.4.19-pre3-ac6 the keybord
isn't recogniced as a HID device anymore and I just get an error message,
when I reconnect it. The usb driver finds the integrated hub but not the
keyboard itself.
I have attached the error messages when reconnecting, the kernel config,
the output from lsusb (usbtools) and the diffs to the previous
configuration and the corresponding lsusb.
I looked at the code, but there weren't any changes to the hid*.c files,
but many updates to _hub.[ch]_ and _uhci.[ch]_.
Some loaded modules (no network and gameport included - Not tainted):
mousedev 3904 1
hid 19232 0 (unused)
input 3232 0 [mousedev hid evdev joydev sidewinder]
uhci 24488 0 (unused)
usbcore 56800 1 [hid uhci]
Thanks
Jan-Marek
On Mon, Mar 25, 2002 at 06:57:59PM +0100, Jan-Marek Glogowski wrote:
> Hi
>
> I have a problem with my Microsoft Natural USB Keyboard.
>
> Since I moved from kernel 2.4.18-rc2-ac1 to 2.4.19-pre3-ac6 the keybord
> isn't recogniced as a HID device anymore and I just get an error message,
> when I reconnect it. The usb driver finds the integrated hub but not the
> keyboard itself.
Can you try the patches at:
http://marc.theaimsgroup.com/?l=linux-usb-devel&m=101684196109355
and also:
http://marc.theaimsgroup.com/?l=linux-usb-devel&m=101684207509482
And let us know if they help you out?
If not, try renaming 'usbmodules' on your box to something else (like
'usbmodules.orig' and seeing if that solves your problem?
thanks,
greg k-h
Hi Greg
[schnipp]
> Can you try the patches at:
> http://marc.theaimsgroup.com/?l=linux-usb-devel&m=101684196109355
> and also:
> http://marc.theaimsgroup.com/?l=linux-usb-devel&m=101684207509482
>
> And let us know if they help you out?
[schnapp]
Applied both patches - the keyboard is detected again, but I still have
some errors in the lsusb-output (see attachment).
Anyway the keyboard works.
Thanks
Jan-Marek
On Mon, Mar 25, 2002 at 08:07:21PM +0100, Jan-Marek Glogowski wrote:
> Hi Greg
>
> [schnipp]
> > Can you try the patches at:
> > http://marc.theaimsgroup.com/?l=linux-usb-devel&m=101684196109355
> > and also:
> > http://marc.theaimsgroup.com/?l=linux-usb-devel&m=101684207509482
> >
> > And let us know if they help you out?
> [schnapp]
>
> Applied both patches - the keyboard is detected again, but I still have
> some errors in the lsusb-output (see attachment).
Sounds like a device that is lying about it's strings. If the device
works, I wouldn't worry about it :)
If it bothers you, take it up with the lsusb author.
thank for testing those patches.
greg k-h
Greg KH wrote:
> On Mon, Mar 25, 2002 at 08:07:21PM +0100, Jan-Marek Glogowski wrote:
> > Hi Greg
> >
> > [schnipp]
> > > Can you try the patches at:
> > > http://marc.theaimsgroup.com/?l=linux-usb-devel&m=101684196109355
> > > and also:
> > > http://marc.theaimsgroup.com/?l=linux-usb-devel&m=101684207509482
> > >
> > > And let us know if they help you out?
> > [schnapp]
> >
> > Applied both patches - the keyboard is detected again, but I still have
> > some errors in the lsusb-output (see attachment).
>
> Sounds like a device that is lying about it's strings. If the device
> works, I wouldn't worry about it :)
Greg, bad guessing. This is not the device's fault but the linux usb
drivers are buggy.
The messages:
bInterfaceClass cannot get string descriptor 1, error = Broken
pipe(32)
cannot get string descriptor 2, error = Broken pipe(32)
go away after "rmmod hid" (or whatever driver is using the device).
This is a long standing bug.
Regards, Gunther
On Wed, Mar 27, 2002 at 08:00:18PM +0100, Gunther Mayer wrote:
>
> Greg, bad guessing. This is not the device's fault but the linux usb
> drivers are buggy.
>
> The messages:
> bInterfaceClass cannot get string descriptor 1, error = Broken pipe(32)
> cannot get string descriptor 2, error = Broken pipe(32)
>
> go away after "rmmod hid" (or whatever driver is using the device).
>
> This is a long standing bug.
Bug in lsusb or in the kernel drivers?
Can you send what /proc/bus/usb/devices looks like with the device
plugged in, _and_ the driver bound to the device? Doing that reads the
strings from the device, just like lsusb should be doing. If that is
successful, I think the problem would be in lsusb.
thanks,
greg k-h