RHEL 5.8 Guest mouse left clicks when moved

[Shaftstall]

I have searched the internet high and low, and I am at a loss. I have a RHEL 5.8 Guest running on a RHEL 6.3 Host using the most recent version of VBox 4.1.20 with the newest extension, and I cannot figure out what is causing this. Does anyone have any insight into what could be causing the mouse to sporadically left click when moving? I have checked the hardware just to be sure, also the issue is not present on the Host or any other Guests that are on a different OS. I have rebuilt the VM from scratch and still the problem persists. I am at my wits end.

[BCD3]

This sounds similar to the problem I had with xorg-x11-servers (see 'activities screen glitches in Fed. 17 after 4.1.20 install'). Instead of left click, when moving the mouse the mouse would occasionally 'snapto' the hot spot in the upper left corner of Gnome3 causing the screen to switch. The cuprit was the xorg-x11-servers in Fedora (xorg-x11-server-common, xorg-x11-server-Xorg, xorg-x11-server-Xephyr) all version 1.12.3-1.fc17. The problem had been there for a while, but when I upgraded to 4.1.20 the problem got much worse.
I downgraded the xorg-x11-servers above to 1.12.0-2.fc17 and the problem was resolved.

Again I'm not sure if this is the same problem but it seems similar.

[yug]

I am having mouse pointer problems as well and it's not just fedora. In both Fedora 17 and Arch Linux -- while moving the mouse the pointer will jump suddenly to some extreme off-screen location and then return. This is very evident, (and extremely annoying), in graphics applications such as gimp or blender. I switched my mouse to another logitech wireless rule it out as defective and the problem persists.
I believe there may be a bug in 4.1.20 drivers.

[yug]

I jumped the gun with my assumption. The problem appears to be in xorg-server later than 1.12.2. Reverting just xorg-server made the problem go away.

[Shaftstall]

Appreciate the responses. After getting desperate I went ahead and started from scratch. The new VMs don't seem to have the issue, and I went back through my notes and realized that the affected VMs were all clones of the original faulty VM.