Corrupt Mouse Cursor

C

cswake

Weaksauce
Joined
Jun 17, 2002
Messages
89
I'm posting here to get some help with a problem that I think is driver related. (Posted to AMD forums but the mods deleted it for some reason.) I've found various other people who have experienced this, and it seems to only occur on Radeons:

http://forums.amd.com/game/messageview.cfm?catid=279&threadid=137745&enterthread=y
http://devforums.amd.com/game/messageview.cfm?catid=279&threadid=110876&highlight_key=y
http://www.overclock.net/ati/752203-anyone-else-have-corrupt-mouse-cursors.html
http://social.technet.microsoft.com/Forums/en/w7itproui/thread/820764ce-97f6-49e6-9735-014531664e0d

The specific problem is that the mouse cursor will become corrupted and can only be resolved by restarting Windows. Screenshots will show a cursor that is not corrupt, so I have to take a picture of it using a digicam. Here's an example of how it will sometimes look (the corruption will vary):

2hqrp6g.jpg


It occurs randomly, both on the desktop and in games. There is no other corruption other than the cursor. I'm not sure how to try to narrow down what could be causing it, but here are a few that I think I've eliminated:

  • Not overclocked
  • Temperatures look good
  • Not system stability (Prime,OCCT, and Furmark)
  • Not an underpowered power supply
  • Not the card itself (replicated it on a 5770 and a 6850)
  • Not dual-monitors (only have a single screen)
  • Not Catalyst 10.10 (I've tried 10.8 and 10.9)
Relevant system specs:
  • Seasonic X750
  • i7-950
  • Asus Rampage III Gene (latest bios)
  • Powercolor Radeon 6850
  • Catalyst 10.10e
Thanks for any suggestions...
 
Dunno the cause, but this happened to me last night. It's only happening on my main monitor though, not my 2 side screens. I've been busy enough with other stuff I haven't tried a reboot yet. The fact that it's happening on one monitor driven by the card, but not the second screams software glitch to me.
 
Yeah, I have this same issue. There are a number of ways you can temporarily fix it without restarting, but none of them are a good fix and no one seems to know the cause of the issue. Its a pretty well know problem so I have to believe that AMD is looking into it.
 
I have a 5870 and get the corrupt mouse cursor once in a while playing starcraft2. It's a known issue but i don't understand why it hasn't been fixed yet.
 
Yep, I'm lookin at that very same mouse pointer right now. Not sure what causes it or how to fix it.
 
OP, have you tried disabling the shadow behind the cursor, or changing it from the default white to a different color/size like large/black? Try this after it is corrupted.
 
Ive found that if I have a multi monitor setup going on and I alt tab out of a game it gets corrupted. Probably the biggest turn off for me from an AMD setup.
 
Driver problem, maybe? I have a 5870 and I haven't upgraded my drivers in ages. Maybe some within the last few months or so that was introduced?
 
cswake said:
I'm posting here to get some help with a problem that I think is driver related. (Posted to AMD forums but the mods deleted it for some reason.) I've found various other people who have experienced this, and it seems to only occur on Radeons:

http://forums.amd.com/game/messageview.cfm?catid=279&threadid=137745&enterthread=y
http://devforums.amd.com/game/messageview.cfm?catid=279&threadid=110876&highlight_key=y
http://www.overclock.net/ati/752203-anyone-else-have-corrupt-mouse-cursors.html
http://social.technet.microsoft.com/Forums/en/w7itproui/thread/820764ce-97f6-49e6-9735-014531664e0d

The specific problem is that the mouse cursor will become corrupted and can only be resolved by restarting Windows. Screenshots will show a cursor that is not corrupt, so I have to take a picture of it using a digicam. Here's an example of how it will sometimes look (the corruption will vary):

2hqrp6g.jpg


It occurs randomly, both on the desktop and in games. There is no other corruption other than the cursor. I'm not sure how to try to narrow down what could be causing it, but here are a few that I think I've eliminated:

  • Not overclocked
  • Temperatures look good
  • Not system stability (Prime,OCCT, and Furmark)
  • Not an underpowered power supply
  • Not the card itself (replicated it on a 5770 and a 6850)
  • Not dual-monitors (only have a single screen)
  • Not Catalyst 10.10 (I've tried 10.8 and 10.9)
Relevant system specs:
  • Seasonic X750
  • i7-950
  • Asus Rampage III Gene (latest bios)
  • Powercolor Radeon 6850
  • Catalyst 10.10e
Thanks for any suggestions...
Click to expand...

Its a driver issue, it really happens bad if using afterburner. You can also unplug the mouse for 30 seconds and plug it back in that also fixes it. AMD knows about this bug but will they fix it any time soon, we will see..
 
Yeah seems to be a driver issue... It started happening to me when I upgraded to the 10.10 drivers, Seems to happen randomly, haven't found a pattern when it happens yet.


Y.
 
EliteComputerBuilds said:
Its a driver issue, it really happens bad if using afterburner. You can also unplug the mouse for 30 seconds and plug it back in that also fixes it. AMD knows about this bug but will they fix it any time soon, we will see..
Click to expand...
Didn't work for me despite a 90 secondish unplug 2 minutes ago.
 
If its similar to this one, try turning on mouse trails and limit the trail. :) Seems to be a MB issue from what I have read earlier. Never seen it in person though on any driver set. AMD is aware of this, as you can see from the thread.
 
It is definitely an issue AMD needs to get on. But a reboot is not usually needed to fix it. It almost always hits me at random and pretty much exclusively on my center monitor. Changing res, activating an eyefinty profile, hot plugging the mouse, or sometimes, even just dragging an open folder from one screen to the next fixes it. I normally just hot key from standard extended to Eyefinity then back. That fixes it pretty quick and easy for me, though I sorely want this little annoyance fixed on a permanent basis.
 
I'm (sort of) glad that it isn't just isolated to me. For those of you who do want to try, I've compiled a list of possible workarounds:
  • Enabling/disabiling shadows for mouse
  • Enabling mouse trails (some games' cursor won't have an in-game cursor, ie. Team Fortress 2)
  • Disable Catalyst AI (have to do this through ATI Tray Tools)
And here's a possible list of ways to "fix" the cursor without rebooting:
  • Unplugging mouse
  • Open magnifier and zoom to 100%
  • Activating an eyefinity profile (for those with multiple screens)
  • Changing resolution
  • Put Windows into sleep mode
Right now I'm running with Catalyst AI disabled and haven't had it happen for a day. Is there some way we can confirm that AMD is aware of the issue and actually working to resolve it?

Let me also add, for the sake of those who might be able to resolve this, that someone in the Microsoft thread I had originally posted what he thought was the actual problem:
This leads me to believe that it is a bounds-checking issue in the DWM. First, because the problem only occurs on a display in which the cursor will be partially drawn and partially offscreen, where the other display would be displaying the cursor if the cursor could be visible on two displays simultaneously. That is, if you move your mouse left to right, the cursor will gradually disappear beginning on the right, and then when the top left pixel is no longer on the left display, the entire cursor will appear on the right display. The problem will only occur on the left display, wherein the cursor, if drawn as an object on the screen, would appear on both displays, but due to the way the display path works, is only drawn on one display. So in a vertical arrangement, the bug only occurs on the top display, in a horizontal arrangement, the bug only occurs on the left display. Second, I believe this is a DWM issue because the problem persists through driver updates but not sleep and wake, at which point I believe the DWM clears itself.
Click to expand...
 
Last edited:
CSWake's list is excellent, although I'm sad to say none of the options work for me. Thanks for compiling them all, though.

For the record, this is caused by pretty much all RTS games, I get it in C&C3, the original StarCraft as well as SC2, plus World in Conflict, the list goes on. Any game of that genre that uses a mouse pointer seems to be susceptible. RPG games which also have a pointer seem unaffected, oddly enough. Same goes for casual games, no issues. What's so special about RTSs?

My determination is to find a way to cure it without a restart, as I loathe restarting my PC under any circumstances, once every couple of weeks when Windows Update forces the issue is my absolute limit. Plus, it bugs me that I should have to just to cure such a basic function as my mouse pointer!

The only "fix" that works on mine is the pointer trails option, but that just makes my mouse pointer flickery and will annoy me no end. Changing resolution, screwing with the Catalyst, using the magnifier, shadows, none of them fix it. And my new PC has decided it doesn't like Sleep any more, it used to work but now when I sleep it won't come back up.

As for Hibernate, it's barely quicker than a restart (my BIOS takes about 30 seconds to go through all the various controllers), so it's really not worth the bother.

This has been going on now across two PCs for over FOUR years. I remember playing Company of Heroes in 2006, using XP on my old HP machine with a Radeon X1950 Pro, and being annoyed by it happening all the time. Now with W7 x64 on a Quad Core with an HD5850, the problem still remains. It's crazy that it's still an issue after so long.

You'd think there'd be something - a service you can stop and start (I tried DWM, it doesn't work) that makes Windows "rebuild" the mouse pointer. Surely there's something? What does Sleep/Hibernate do that's so special? If we could identify what process during wakeup resets the mouse, we could create a fix, and to hell with ATI's indifference.

Ah well. Time for yet another unwanted restart. Bah!
 
It'd be nice, but I'm not holding my breath. Driver updates across two different cards on two different operating systems spanning 4 years haven't cured it, so I'm not expecting anything different from this one.
 
This issue has been around for a very long time . It hasn't been fixed by now and I don't think it ever will.
 
Sadly, I think you're right. It's insane, really, something so fundamental, and in theory so incredibly easy to resolve. Hell, a mouse pointer reset utility would do it, but Google shows nothing relevant. Ah well. So much for technological progress.
 
cswake said:
I'm (sort of) glad that it isn't just isolated to me. For those of you who do want to try, I've compiled a list of possible workarounds:
  • Enabling/disabiling shadows for mouse
  • Enabling mouse trails (some games' cursor won't have an in-game cursor, ie. Team Fortress 2)
  • Disable Catalyst AI (have to do this through ATI Tray Tools)
And here's a possible list of ways to "fix" the cursor without rebooting:
  • Unplugging mouse
  • Open magnifier and zoom to 100%
  • Activating an eyefinity profile (for those with multiple screens)
  • Changing resolution
  • Put Windows into sleep mode
Right now I'm running with Catalyst AI disabled and haven't had it happen for a day. Is there some way we can confirm that AMD is aware of the issue and actually working to resolve it?

Let me also add, for the sake of those who might be able to resolve this, that someone in the Microsoft thread I had originally posted what he thought was the actual problem:
This leads me to believe that it is a bounds-checking issue in the DWM. First, because the problem only occurs on a display in which the cursor will be partially drawn and partially offscreen, where the other display would be displaying the cursor if the cursor could be visible on two displays simultaneously. That is, if you move your mouse left to right, the cursor will gradually disappear beginning on the right, and then when the top left pixel is no longer on the left display, the entire cursor will appear on the right display. The problem will only occur on the left display, wherein the cursor, if drawn as an object on the screen, would appear on both displays, but due to the way the display path works, is only drawn on one display. So in a vertical arrangement, the bug only occurs on the top display, in a horizontal arrangement, the bug only occurs on the left display. Second, I believe this is a DWM issue because the problem persists through driver updates but not sleep and wake, at which point I believe the DWM clears itself.
Last edited by cswake; 11-14-2010 at 12:08 PM..
Click to expand...
Click to expand...

I believe that very well could be the actual problem- I have noticed that it only happened to me when I was playing SC2 when I would scroll to the right, which is the side where my second monitor is located. Never happened any other time. Another note though- has not yet happened with my new 6850s, but did happen occasionally with my 5750s.
 
Just want to give an update for anyone experiencing this problem. In my case, disabling Catalyst AI doesn't prevent the cursor from getting corrupted, but it does happen less frequently AND it eventually goes away without having to reboot.
 
Thanks for the update, I hate having to turn off features just to cure problems, so it's nice to know at least that this doesn't help.

I can offer some further insight into this: If you're lucky, you can reverse it the moment it happens. You just have to be quick to notice it. It's caused by moving the cursor off the right hand edge of the screen, and I've found that if, as soon as you notice, you immediately move the mouse back to the right edge, it often sets the cursor back again. If you leave it too long, you end up stuck with it corrupted, but several times in recent gaming sessions I've managed to catch it and reset it.

Other information, I thought for a while that resolution made a difference, but whether running at 1024x768 or 1280x1024 it still happens.

I managed an entire session on C&C3 the other day, played for several hours without a glitch. The next session it did it about ten minutes in, so it's entirely random and circumstantial, some combination of moving the mouse and drawing the screen causes it, but I've no idea what.
 
Also have this problem; knew about the magnifier fix will try some of the others.

Yeah this is insane. :T Such a basic thing.
 
I am a loyal Radeon user, having owned more than a dozen cards. I was a loyal Matrox fan previously, for those who remember that era. I am not looking forward to switching to nVidia.

In my case I see corrupt cursors on a single monitor Windows 7 Media Center. The magnifier and trailing cursor fixes do not work in media center. I do not use ATI Tray Tools. I have a pre eyefinity card. There is no work around for me.

Before I decide to fight or switch. I would like to gauge how big of an issue this is for this group. Is anyone willing to do any of the following...

  • Determine what formal bug reporting and other feedback systems AMD has.
  • Document and round up users with similar issues on other forums.
  • Find and contact product managers, evangelists and similar types at AMD.
  • Find and contact AMD OEMs.
  • Prioritize and contact members of the press.
  • Prioritize and contact industry analysts.
  • Provide objective feedback on why an effort to get this bug fixed could or couldn't be launched from this forum.
 
If your mouse cursor is corrupt, you should consider making it take some ethics classes. *rimshot*
 
You must log in or register to reply here.
Back
Top