Comment 4 for bug 889226

On Mon, Nov 14, 2011 at 10:43:17PM -0000, Bryce Harrington wrote:
> Hi Steve,
>
> In the Xorg.1.log file I'm seeing:
>
> [ 99307.560] (II) config/udev: Adding input device Logitech USB-PS/2 Optical Mouse (/dev/input/mouse1)
> [ 99307.560] (II) No input driver/identifier specified (ignoring)
> [ 99307.560] (II) config/udev: Adding input device USB KVM (/dev/input/event10)
> [ 99307.560] (**) USB KVM: Applying InputClass "evdev keyboard catchall"
> [ 99307.560] (II) Using input driver 'evdev' for 'USB KVM'
> [ 99307.560] (II) Loading /usr/lib/xorg/modules/input/evdev_drv.so
> [ 99307.560] (**) USB KVM: always reports core events
> [ 99307.560] (**) USB KVM: Device: "/dev/input/event10"
> [ 99307.561] (--) USB KVM: Found keys
> [ 99307.561] (II) USB KVM: Configuring as keyboard
> [ 99307.561] (**) Option "config_info" "udev:/sys/devices/pci0000:00/0000:00:1a.0/usb1/1-1/1-1.2/1-1.2.4/1-1.2.4:1.0/input/input18/event10"
> [ 99307.561] (II) XINPUT: Adding extended input device "USB KVM" (type: KEYBOARD)
>
> and
>
> [ 99307.632] (II) config/udev: Adding input device USB KVM (/dev/input/event11)
> [ 99307.632] (**) USB KVM: Applying InputClass "evdev pointer catchall"
> [ 99307.632] (**) USB KVM: Applying InputClass "evdev keyboard catchall"
> [ 99307.632] (II) Using input driver 'evdev' for 'USB KVM'
> [ 99307.632] (II) Loading /usr/lib/xorg/modules/input/evdev_drv.so
> [ 99307.633] (**) USB KVM: always reports core events
> [ 99307.633] (**) USB KVM: Device: "/dev/input/event11"
> [ 99307.633] (--) USB KVM: Found 9 mouse buttons
> [ 99307.633] (--) USB KVM: Found scroll wheel(s)
> [ 99307.633] (--) USB KVM: Found relative axes
> [ 99307.633] (--) USB KVM: Found x and y relative axes
> [ 99307.633] (--) USB KVM: Found absolute axes
> [ 99307.714] (II) evdev-grail: failed to open grail, no gesture support
> [ 99307.714] (--) USB KVM: Found keys
> [ 99307.714] (II) USB KVM: Configuring as mouse
> [ 99307.714] (II) USB KVM: Configuring as keyboard
> [ 99307.714] (II) USB KVM: Adding scrollwheel support
> [ 99307.714] (**) USB KVM: YAxisMapping: buttons 4 and 5
> [ 99307.714] (**) USB KVM: EmulateWheelButton: 4, EmulateWheelInertia: 10, EmulateWheelTimeout: 200
> [ 99307.714] (**) Option "config_info" "udev:/sys/devices/pci0000:00/0000:00:1a.0/usb1/1-1/1-1.2/1-1.2.4/1-1.2.4:1.1/input/input19/event11"
>
> A HID and keyboard appear to be getting mapped to event13 and event14.
> A PS/2 USB mouse is on event9.

Yes. event13 and event14 are incorrect mappings; as mentioned in the
report, the actual devices are event9 and event10. The current lsinput
results are:

/dev/input/event9
   bustype : BUS_USB
   vendor : 0x45e
   product : 0xdb
   version : 273
   name : "Microsoft Natural® Ergonomic Ke"
   phys : "usb-0000:00:1a.0-1.2.2/input1"
   uniq : ""
   bits ev : EV_SYN EV_KEY EV_REL EV_ABS EV_MSC EV_REP

/dev/input/event10
   bustype : BUS_USB
   vendor : 0x46d
   product : 0xc03e
   version : 272
   name : "Logitech USB-PS/2 Optical Mouse"
   phys : "usb-0000:00:1a.0-1.2.3/input0"
   uniq : ""
   bits ev : EV_SYN EV_KEY EV_REL EV_MSC

(this may not correspond to what was shown in the log at the time since the
USB hub has been unplugged and replugged a few times since then - but they
have consistently had indices 9 and 10 here.)

> What are the brand/model of mouse being plugged into the KVM? Is the
> keyboard plugged into the KVM a Microsoft Natural Ergonomic Keyboard
> 4000? Also, what's the brand/model of the USB KVM itself?

Yes, that's the correct keyboard model. The KVM is an off-brand unit that
reports itself as a Cypress TetraHub in lsusb - but that shouldn't be
relevant, the KVM works fine and the mouse and keyboard endpoints are being
correctly exposed. It's only X that's failing to pick up the devices.

The mouse is, as above, a Logitech optical mouse.

> I'm interpreting from your wording that this is a regression and the
> laptop with the KVM worked properly in Ubuntu at some point in the past?
> What has changed on your system since that time? kernel 3.0 vs. 3.1
> maybe?

It worked reliably under oneiric. I'm now running precise, which is where
I've seen the failure. I have no reason to believe it's tied to kernel
version at all - note that the failure to pick up the devices occurred
during an X session in which the keyboard and mouse previously *had* been
picked up just fine. And since filing this bug report, a subsequent
unplug/replug has resulted in the keyboard being picked up reliably again,
but the mouse is still not being seen by X (but is being seen by the
kernel). So it's not even consistent with this version of the X packages.

I've left this X session running for debugging in spite of the inconvenience
of not having my external mouse, because I fully expect that restarting will
cause the bug to be unreproducible again.

--
Steve Langasek Give me a lever long enough and a Free OS
Debian Developer to set it on, and I can move the world.
Ubuntu Developer http://www.debian.org/
<email address hidden> <email address hidden>