mouse jumps to top left corner when using media keys

[Reed Lipman]

Binary package hint: rhythmbox

When I press one of my keyboard's media keys, the mouse jumps to the upper left corner of the screen when I start to move it after pressing the button. This happens with the play, pause, stop, previous track, next track, mute, and eject buttons. The mouse doesn't jump if I use the non-media keyboard shortcuts or click on the buttons.

My keyboard layout is Logitech Cordless Desktop LX-300 as reported by System->Preferences->Keyboard. This has only started happening since lucid.

Keybinds for these keys (from System->Preferences->Keyboard Shortcuts)
Volume mute: XF86AudioMute
Play (or play/pause): XF86AudioPlay
Stop playback: XF86AudioStop
Previous Track: XF86AudioPrev
Next Track: XF86AudioNext
Eject: XF86Eject

ProblemType: Bug
Architecture: amd64
CheckboxSubmission: f5e32f44c6e48eb43576e12312917e86
CheckboxSystem: 4ed15c40009aa6f7770f606350a390a2
Date: Thu Jan 28 11:19:59 2010
DistroRelease: Ubuntu 10.04
EcryptfsInUse: Yes
ExecutablePath: /usr/bin/rhythmbox
NonfreeKernelModules: nvidia
Package: rhythmbox 0.12.6-1ubuntu6
ProcEnviron:
 PATH=(custom, user)
 LANG=en_US.UTF-8
 SHELL=/bin/bash
ProcVersionSignature: Ubuntu 2.6.32-11.15-generic
SourcePackage: rhythmbox
Uname: Linux 2.6.32-11-generic x86_64

[Pedro Villavicencio (pedro)]

not confirming it works fine here, could you try the same on a new user on your system? Thanks in advance.

[Reed Lipman (rmlipman)]

I made a new user and left the settings at the default. Still happening for me.

[ccc1 (cllccl-deactivatedaccount)]

hi,

how to reproduce:
1) press an alphanumeric key
2) press a multimedia key (in my case any of the keys that control the volume)
3) move the mouse
result: mouse cursor jumps immediately to the left upper corner of the screen

i have an external lenovo usb keyboard with trackpoint. xinput output:
Lite-On Technology Corp. ThinkPad USB Keyboard with TrackPoint id=7 [slave keyboard (3)]
this bug doesn't occur with my internal keyboard, which has a trackpoint and multimedia keys as well.

other people with usb keyboard mouse combos observed this behavior too:
http://bbs.archlinux.org/viewtopic.php?id=86344
https://bugs.launchpad.net/ubuntu/+source/rhythmbox/+bug/513852

[ccc1 (cllccl-deactivatedaccount)]

two things i forgot to add:
OS:Linux cl-laptop 2.6.32-ARCH x86_64
and they keyboard is configured via hal.

[Peter Hutterer (peter-hutterer)]

already fixed, thanks for the bug report though.

commit 6f265d55a61f9be323583b8acacae783be72bda9
Refs: xorg-server-1.7.99.2-40-g6f265d5
Author: Peter Hutterer <email address hidden>
AuthorDate: Tue Jan 5 13:15:25 2010 +1000
Commit: Keith Packard <email address hidden>
CommitDate: Tue Jan 5 14:01:51 2010 -0800

    dix: don't update the slave coordinates from the VCK.

    A keyboard event from a device with both valuators and keys will be posted
    through the VCK. In this case, do not update the slave device coordinates
    from the VCK - they're always 0/0. Leave them as-is, for the next pointer
    event will continue where it left.

[José P Valdés (sevmpe-deactivatedaccount)]

I think it might be already fixed upstream (in xorg): https://bugs.freedesktop.org/show_bug.cgi?id=26640

[Reed Lipman (rmlipman)]

Actually affects xorg according to the upstream report, moving accordingly

[Bryce Harrington (bryce)]

commit 6f265d55a61f9be323583b8acacae783be72bda9
Refs: xorg-server-1.7.99.2-40-g6f265d5
Author: Peter Hutterer <email address hidden>
AuthorDate: Tue Jan 5 13:15:25 2010 +1000
Commit: Keith Packard <email address hidden>
CommitDate: Tue Jan 5 14:01:51 2010 -0800

    dix: don't update the slave coordinates from the VCK.

    A keyboard event from a device with both valuators and keys will be posted
    through the VCK. In this case, do not update the slave device coordinates
    from the VCK - they're always 0/0. Leave them as-is, for the next pointer
    event will continue where it left.

[Reed Lipman (rmlipman)]

This has gone away for me, but I don't remember when that happened. Was this change pulled in already?

[Bryce Harrington (bryce)]

Yes, this fix got backported and is included in the 1.7 xserver carried in lucid.