Red Hat Bugzilla – Bug 493618
The Xen virtual keyboard/mouse combination device confuses the X server
Last modified: 2010-04-08 12:21:34 EDT
Created attachment 337793 [details]
Patch to split the combo xen virtual keyboard/pointer device into two virtual devices
Description of problem:
While its possible to configure the X server to use the Xen virtual keyboard
and mouse combination device (see other bugs referenced in the next comment) it
is fraught with problems because the X server only half-believes that you can
have a device that is both a keyboard and a mouse. In particular you must have
exactly one CorePointer and one CoreKeyboard and, although this is not stated
explicitly, they cannot be the same device. It gets worse, though, because
the Xen virtual keyboard (which is also a pointer) must have exactly the same
Xkb configuration as the normal keyboard because which Xkb settings are on a
per-device basis, this isn't reflected through utilities like setxkbmap.
There's more detail in the bugs I'm about to enter and refer to later.
Later kernels, for example, Fedora 9's kernel, split the virtual keyboard and
pointer into two separate devices. This makes configuration much easier and
doesn't run into logical problems with the X server configuration.
Having said that, this is a "nice to have" because we can configure the Xen
pointer and keyboard combo device in a way that is transparent to most users.
Version-Release number of selected component (if
applicable):kernel-xenU-2.6.18-128 (and other, earlier kernels).
Steps to Reproduce:
1. Attempt to configure the Xen virtual pointer so that you get proper mouse
tracking in a Xen guest VNC console.
Without fixing xorg-x11-input-evdev, xkeyboard-config and rhpl and rhpxl to understand more of this combo device you can't do this. Well, you can, so long as you don't mind losing either the keyboard or the mouse or sometimes both.
With the xorg-x11-drv-evdev this should be a piece of cake, but it's not because the evdev driver wants to be both a keyboard and a mouse and this confuses X.
I've attached a patch which does the trick. But see also the rhpl, rhpxl, xkeyboard-config and xorg-x11-drv-evdev patches.
Related bugs: bug 493618, bug 493623, bug 493627, bug 493630, bug 493634 and bug 493642
The bug as reported is about the difficulties to manually configure X
in a certain way, to get proper pointer tracking. It is certainly
valid. However, the real problem users care about is that pointer
tracking works very poorly (mouse hits invisible wall).
Upstram, we solved the problem by fixing numerous issues in various
places to make it work out of the box, from Fedora 11 on.
The patches proposed by the reporter solve the problem by making
manual configuration work. They are unrelated to the upstream
Like the upstream solution, they're fairly invasive: they touch the
kernel and several user space packages. This is a significant risk.
Moreover, splitting the device is an ABI change. The question is
whether there are benefits to justify that.
We already put much less invasive changes into 5.4 to improve pointer
tracking (bug 492866). They fix the "mouse hits invisible wall"
problem. They still require a pointer grab, which virt-viewer makes
So, the benefit of the proposed patches over what we've got already is
to enable manual configuration so that the mouse grab can be avoided.
While that's certainly nice, it doesn't justify the risk, in my
If you disagree, please reopen the bug.
Some technical background:
There are 3 pieces in the stack: VNC client, QEMU VNC server and the
guest's X. The old, misbehaving setup has VNC client sending absolute
coordinates, which the VNC server sends through xenkbd to X. X in
RHEL-5 doesn't have any auto-configuration of input devices, so it
just opens /dev/input/mice, and thus the guest kernel converts from
absolute to relative. Because the VNC client has no knowledge of this
conversion, you get into situation where you hit an invisible wall in
the client when it thinks it has got to the virtual desktop boundary.
Now, if you had absolute coordinates being passed and used all the way
to X then it would trivially work, but this requires X configuration.
Recent X's do that automatically. RHEL-5's X simply isn't capable of
that. The patches proposed by the reporter add manual configuration
steps across a wide range of RPMs.
The fix we did for RHEL-5.4 is to make xenkbd *not* use absolute
coordinates, so the VNC server now sends relative coordinates. This
removes the broken absolute -> relative conversion in the guest
kernel. On its own this isn't sufficient, because you then get broken
absolute to relative conversion in the VNC server instead. So we also
added the VNC relative mouse extension to QEMU and GTK-VNC. Now the
VNC client is in charge of doing the absolute -> relative conversion.
This is good, as this is the only point in the stack capable of doing
the conversion correctly, to avoid hitting an invisible wall. The
only restriction is that you must grab the host mouse pointer and hide
it, so you only see the guest drawn pointer.
By the way, RHEL-6 pointer tracking will be fine out of the box using absolute coordinates, just like upstream.