From Bugzilla Helper: User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.2b) Gecko/20021018 Description of problem: The black 'X' which is the mouse pointer before a window manager loads is drawn in the middle of the screen when the window manager loads. I'm using the CVS RPMs, version 4.2.99.1-0.20021018.5. When metacity loads, I am given the red transparent mouse pointer, and the black 'X' is drawn, always on top, in the center of the screen. I had thought it was permanent, but it has disappeared once. Version-Release number of selected component (if applicable): How reproducible: Always Steps to Reproduce: 1. login 2. start x Actual Results: black X is drawn on top of all windows in center of screen when metacity is loaded. Expected Results: black X should become red mouse pointer when metacity loads. Additional info:
Created attachment 82184 [details] X log
Created attachment 82185 [details] X Config file
Created attachment 82186 [details] lspci -vn output
Please try the latest CVS RPM packages from November 1st. Does the latest package resolve this problem?
*** Bug 77213 has been marked as a duplicate of this bug. ***
Please test the latest release of my CVS RPM's guys from: ftp://people.redhat.com/mharris/testing/live-grenades - the 20021101 release Does the problem persist, or does it go away with this build?
Created attachment 83364 [details] XF86Config for 20021101 packages
Problem still exists. Does not occur when logged in as root.
Hmm, that is uberweird. Notifying XFree86 people of this issue, as I have no Intel i8x0 hardware for troubleshooting. I'll update the report with any changes, and periodically ask you to try new packages probably. thanks again
Ok, David Dawes has acknowledged this bug in the i810 driver and that it affects i830 and i845 hardware not disabling the hardware cursor in some circumstances. He plans on fixing it sometime in the next week or so. I'll monitor CVS commits for when the changes go in, and update the bug report when i've got new packages for testing. Thanks guys, your beta testing is appreciated.
Still present in version from 2002-11-01.
David has fixed this in CVS, and I'm building a new CVS snapshot right now, which will appear in the usual place with datestamp 20021105 within 24 hours. Please test this and confirm it fixes the problem. I'm setting bug state to MODIFIED. After testing, please set bug state to either RAWHIDE if the problem is resolved, or to ASSIGNED, if the problem is still present, and include any new information. If you encounter any new problems/bugs, please open new bug reports. Thanks guys.
2002-11-05 fixed it for me.
20021105 works for me too. :D
Update: There is still a bug in the i8x0 cursor code, and David just fixed it. I'll be rolling new rpm's either today or tomorrow with this fix in it, so if you do encounter a problem, wait for the next snapshot to test before reporting it. Thanks guys.