Red Hat Bugzilla – Bug 454948
cursor movement is constrained
Last modified: 2008-07-11 04:35:19 EDT
+++ This bug was initially created as a clone of Bug #238825 +++ Description of problem: When using virtual graphical consoles after the pointer is grabbed often the cursor is constrained away from one edges of the screen, which is annoying. Steps to Reproduce: 1. start virtual machine from virt-manager 2. click to stay in window 3. move cursor around Actual results: Usually some part (edge) of the screen can't be reached without moving first to the opposite edge. Expected results: Pointer to be directly movable to any part of screen. Additional info: This does not seem to happen if one runs qemu directly for instance. -- Additional comment from bryan.christ@hp.com on 2007-06-29 11:39 EST -- I have this problem too. I've noticed that by moving the mouse really fast, it is possible to "break through" the artificial barrier. -- Additional comment from ville.skytta@iki.fi on 2007-07-01 10:44 EST -- Same effects observed here, and the trick in comment #1 also works around it for me (by moving really fast to the opposite direction from the constrained edge). The workaround isn't enough to make the machines actually usable though, it has to be done all the time. Tested only with a QEMU VM, F-7 x86_64. Plain QEMU works fine. -- Additional comment from petersen@redhat.com on 2007-07-01 22:02 EST -- I haven't tried "moving really fast" but for me moving to the opposite edge and then returning clears the constraint for me. Yes it is very annoying. -- Additional comment from bugzilla@redhat.com on 2007-07-24 22:10 EST -- change QA contact -- Additional comment from berrange@redhat.com on 2007-09-27 17:11 EST -- This should be much improved in current (F8) rawhide version, since we have re-written the VNC widget used for the console. Please let me know if it still gets 'stuck' on invisible walls with virt-manager 0.5.1 or later. -- Additional comment from petersen@redhat.com on 2007-10-15 20:58 EST -- Thanks - looks good to me now. :) -- Additional comment from bryan.christ@hp.com on 2008-07-10 15:01 EST -- This problem has resurfaced with F9 and latest stable updates via yum. -- Additional comment from berrange@redhat.com on 2008-07-10 17:24 EST -- Do *NOT* re-open old bugs. Please file a new bug and if relevant add a comment refering to the original bug. -- Additional comment from petersen@redhat.com on 2008-07-10 19:08 EST -- Yes thanks for bringing this up - it has been back for too long. :-/
It is particularly painful when using a virtual console. It took me a long time to understand that the problem is the invisible real X pointer underneath which moves much quicker than the virtual pointer and hence reaches the edges of the console very quickly before the visible cursor has time to move. This became clear when watching the real and virtual pointer when they are separated.
*** This bug has been marked as a duplicate of 454938 ***