Regression: mouse support stopped working

[Colin Watson]

Binary package hint: kvm

As of a few days ago, I stopped having working mouse support in kvm while booting Hardy daily desktop CD builds. I first saw this while booting either -15 or -16 (I think -15) on a -12 kernel, but am still seeing this while booting an entirely up-to-date Hardy image (-16) on an entirely up-to-date Hardy system (also -16).

It's possible that this was prompted by this change:

xorg (1:7.3+10ubuntu8) hardy; urgency=low

  * dexconf: Add CorePointer to the default mouse section, otherwise
    the server defaults are used. (LP: #197337)

 -- Timo Aaltonen <email address hidden> Thu, 10 Apr 2008 14:54:29 +0300

See the attached Xorg.0.log, where vmmouse seems to detect a virtual machine and then bail out. It's the weekend, though, and I'm just guessing.

[Timo Aaltonen (tjaalton)]

damn vmmouse.. if it really needs the device -option, we lose

[Timo Aaltonen (tjaalton)]

Hmm, actually it's pretty easy.. make dexconf add the device option if it's run on a virtual machine. Fixed in git.

[Launchpad Janitor (janitor)]

This bug was fixed in the package xorg - 1:7.3+10ubuntu9

---------------
xorg (1:7.3+10ubuntu9) hardy; urgency=low

  * control: Add an epoch to the "x11-common Replaces xrgb" version.
    (LP: #212672)
  * dexconf, failsafeDexconf: Revert back to using "mouse" by default, and
    use vmmouse_detect to determine if the system is a virtual machine and
    if yes, use vmmouse and set the device path. (LP: #216667, #216276)
  * control: xserver-xorg needs to Depend on mdetect instead of Recommends.

 -- Timo Aaltonen <email address hidden> Mon, 14 Apr 2008 16:35:12 +0300