Bug: Please fix mouse pointer / mouse cursor with 5k displays issue in Sierra 10.12
  1 / 3    
This bug was introduced in Sierra 10.12 and is NOT present in 10.11 (El Capitan). Despite of several Sierra AND Nvidia web drivers updates, this issue hasn't been fixed. Bug behaves as follows: - Normally, when you point your mouse over a link, it changes into a hand, if you want to resize a window, the cursor changes into a double arrow, and so forth. - The 5k resolution is a virtual 2-screen-setup, seamlessly split in the middle. - Since 10.12, the mouse pointer gets often stuck into either a hand, a double arrow, a cross or whatever. It doesn't change back to normal, unless when you hover over the other half of the screen. In the other half the mouse pointer appears normal again. But move the cursor back to the first half, where the bug occured, and the mouse is still stuck in a different symbol. - the bug appears completely random. It can appear after 10 seconds or 2 hours. I just held off updating to Sierra because of this, because working with a broken mouse pointer is really confusing. Please address this bug Nvidia!
This bug was introduced in Sierra 10.12 and is NOT present in 10.11 (El Capitan). Despite of several Sierra AND Nvidia web drivers updates, this issue hasn't been fixed.

Bug behaves as follows:
- Normally, when you point your mouse over a link, it changes into a hand, if you want to resize a window, the cursor changes into a double arrow, and so forth.
- The 5k resolution is a virtual 2-screen-setup, seamlessly split in the middle.
- Since 10.12, the mouse pointer gets often stuck into either a hand, a double arrow, a cross or whatever. It doesn't change back to normal, unless when you hover over the other half of the screen. In the other half the mouse pointer appears normal again. But move the cursor back to the first half, where the bug occured, and the mouse is still stuck in a different symbol.
- the bug appears completely random. It can appear after 10 seconds or 2 hours.

I just held off updating to Sierra because of this, because working with a broken mouse pointer is really confusing. Please address this bug Nvidia!

#1
Posted 04/07/2017 11:47 AM   
Can you tell me what monitor make and model you are using with your Mac? What graphics card are you using as well? Last but not least, are you using any other monitors with your Mac or just the 5K display?
Can you tell me what monitor make and model you are using with your Mac? What graphics card are you using as well? Last but not least, are you using any other monitors with your Mac or just the 5K display?

We can't fix your bug if we can't reproduce it. Please provide as much details as possible including specs, steps to reproduce, graphics settings used, etc..

For GeForce driver issues, please leave feedback in driver feedback form: Driver Feedback

For SHIELD Tablet issues/requests, please leave feedback in the SHIELD Tablet feedback form: SHIELD Tablet Feedback

For SHIELD Android TV issues/requests, please leave feedback in the SHIELD feedback form: SHIELD Feedback

I am also available on Twitter:
www.twitter.com/manuelguzman

#2
Posted 04/10/2017 09:24 PM   
Monitor is 5K Dell UP2715K. Tested with Sierra 10.12.4, GTX 970 and GTX 1070 with the recent Nvidia web drivers (with pascal support) The bug occures only when in 5K mode connected with 2 DP cables and was introduced with Sierra 10.12. The bug does NOT appear with 10.11 El Capitan, however the pascal drivers are only available for Sierra 10.12 :( I made a video which shows this bug and the mouse pointer changing exactly in the middle of the 5K screen: https://streamable.com/smuwk
Monitor is 5K Dell UP2715K.

Tested with Sierra 10.12.4, GTX 970 and GTX 1070 with the recent Nvidia web drivers (with pascal support)
The bug occures only when in 5K mode connected with 2 DP cables and was introduced with Sierra 10.12. The bug does NOT appear with 10.11 El Capitan, however the pascal drivers are only available for Sierra 10.12 :(

I made a video which shows this bug and the mouse pointer changing exactly in the middle of the 5K screen:
https://streamable.com/smuwk

#3
Posted 04/11/2017 02:42 PM   
I have the same issue with a EVGA GeForce GTX TITAN X 12GB SC GAMING card and the same Dell UP2715K (EDID vendor ID 10AC, product 40B6). I wonder if the problem occurs on other dual cable 5K displays, or MST displays that require two streams to create a 4K display. The Apple support document "Using 4K displays, 5K displays, and Ultra HD TVs with your Mac" lists some of those displays https://support.apple.com/en-ca/HT206587
I have the same issue with a EVGA GeForce GTX TITAN X 12GB SC GAMING card and the same Dell UP2715K (EDID vendor ID 10AC, product 40B6).

I wonder if the problem occurs on other dual cable 5K displays, or MST displays that require two streams to create a 4K display. The Apple support document "Using 4K displays, 5K displays, and Ultra HD TVs with your Mac" lists some of those displays

https://support.apple.com/en-ca/HT206587

#4
Posted 04/12/2017 11:42 AM   
Also with EVGA GTX 980 I am having the same problem concerning cursor misbehavior. I am connected to a Dell UP2715K screen as my only display. (5K, 5120*2880 panel)
Also with EVGA GTX 980 I am having the same problem concerning cursor misbehavior.
I am connected to a Dell UP2715K screen as my only display. (5K, 5120*2880 panel)

#5
Posted 04/12/2017 01:27 PM   
Thank you. I will go ahead and file a bug for this.
Thank you. I will go ahead and file a bug for this.

We can't fix your bug if we can't reproduce it. Please provide as much details as possible including specs, steps to reproduce, graphics settings used, etc..

For GeForce driver issues, please leave feedback in driver feedback form: Driver Feedback

For SHIELD Tablet issues/requests, please leave feedback in the SHIELD Tablet feedback form: SHIELD Tablet Feedback

For SHIELD Android TV issues/requests, please leave feedback in the SHIELD feedback form: SHIELD Feedback

I am also available on Twitter:
www.twitter.com/manuelguzman

#6
Posted 04/12/2017 01:57 PM   
Thank you! On a side note: funnily enough, when you record the screen during this mouse pointer bug, the recorded video doesn't show the bug. In the video, the mouse pointer appears normal on the entire screen. (PS: The video posted above which shows the bug was actually recorded by a phone filming the monitor)
Thank you!
On a side note: funnily enough, when you record the screen during this mouse pointer bug, the recorded video doesn't show the bug. In the video, the mouse pointer appears normal on the entire screen.

(PS: The video posted above which shows the bug was actually recorded by a phone filming the monitor)

#7
Posted 04/12/2017 02:29 PM   
[quote="Kendurax"]On a side note: funnily enough, when you record the screen during this mouse pointer bug, the recorded video doesn't show the bug. In the video, the mouse pointer appears normal on the entire screen.[/quote]The OS uses a hardware cursor, where the pixels of the cursor are stored in a different part of vram and are not copied to the vram of the frame buffer. This problem would go away if there was an option to use a software cursor. However, a software cursor does not draw as fast, and may appear to lag behind your mouse movements. If your screen recording software has an option to record the cursor, then it will draw the cursor itself (like a software cursor). [quote="joevt"]I wonder if the problem occurs on other dual cable 5K displays, or MST displays that require two streams to create a 4K display. The Apple support document "Using 4K displays, 5K displays, and Ultra HD TVs with your Mac" lists some of those displays https://support.apple.com/en-ca/HT206587[/quote] I found a post in Apple's discussion forums about this issue at https://discussions.apple.com/thread/7679637 which mentions similar problems for 4K MST displays. I did not check the specs of all the MacBook Pro computers involved to see if they used AMD or Nvidia or Intel GPU's. If a MacBook Pro has the choice between integrated (Intel) and discrete (Nvidia or AMD) graphics, it uses discrete graphics for external displays.
Kendurax said:On a side note: funnily enough, when you record the screen during this mouse pointer bug, the recorded video doesn't show the bug. In the video, the mouse pointer appears normal on the entire screen.
The OS uses a hardware cursor, where the pixels of the cursor are stored in a different part of vram and are not copied to the vram of the frame buffer. This problem would go away if there was an option to use a software cursor. However, a software cursor does not draw as fast, and may appear to lag behind your mouse movements. If your screen recording software has an option to record the cursor, then it will draw the cursor itself (like a software cursor).

joevt said:I wonder if the problem occurs on other dual cable 5K displays, or MST displays that require two streams to create a 4K display. The Apple support document "Using 4K displays, 5K displays, and Ultra HD TVs with your Mac" lists some of those displays

https://support.apple.com/en-ca/HT206587
I found a post in Apple's discussion forums about this issue at https://discussions.apple.com/thread/7679637 which mentions similar problems for 4K MST displays. I did not check the specs of all the MacBook Pro computers involved to see if they used AMD or Nvidia or Intel GPU's. If a MacBook Pro has the choice between integrated (Intel) and discrete (Nvidia or AMD) graphics, it uses discrete graphics for external displays.

#8
Posted 04/12/2017 03:14 PM   
The models mentioned in the Apple forums all have a Nvidia GPU. Yesterday, I connected my Macbook Pro (2015) with AMD GPU to the Dell 5K via MST and the mouse pointer bug doesn't happen (tested for 3 hours and various reboots). It all points to a Apple X Nvidia issue with MST connections.
The models mentioned in the Apple forums all have a Nvidia GPU.
Yesterday, I connected my Macbook Pro (2015) with AMD GPU to the Dell 5K via MST and the mouse pointer bug doesn't happen (tested for 3 hours and various reboots).

It all points to a Apple X Nvidia issue with MST connections.

#9
Posted 04/12/2017 03:55 PM   
Here with the same Dell UP2715K and GTX 1060, same bug. Please fix it, it would be awesome! :D
Here with the same Dell UP2715K and GTX 1060, same bug. Please fix it, it would be awesome! :D

#10
Posted 04/15/2017 06:12 PM   
I'm getting this cursor issue running a GTX 980ti on a early 2009 Classic Mac Pro tower. I'm running Mac OS Seirra 10.12.4 and have the latest web drivers 378.05.05.05f01. This bug is kinda a deal breaker for me as I need precision with the mouse for my work. Hoping nVidia can get a fix for this out quick.
I'm getting this cursor issue running a GTX 980ti on a early 2009 Classic Mac Pro tower. I'm running Mac OS Seirra 10.12.4 and have the latest web drivers 378.05.05.05f01.

This bug is kinda a deal breaker for me as I need precision with the mouse for my work. Hoping nVidia can get a fix for this out quick.

Mac Pro (Early 2009)
CPU: 2 x 3.46 GHz 6-Core Intel Xeon
RAM: 64 GB 1333 MHz DDR3
Video: NVIDIA GeForce GTX 980 TI
Monitor: Dell UP2715K
macOS High Sierra 10.13.1
--
Web Driver 378.10.10.10.20.109
CUDA Driver 9.0.222

#11
Posted 04/18/2017 10:43 AM   
Same cursor issue using a GTX 1080 in a Mac Pro 5,1 and Dell 5K UP2715K (latest 378.05.05.05f01 drivers)
Same cursor issue using a GTX 1080 in a Mac Pro 5,1 and Dell 5K UP2715K (latest 378.05.05.05f01 drivers)

Mac Pro (Mid 2010)
CPU: 2 x 3.46 GHz 6-Core Intel Xeon
RAM: 128 GB 1333 MHz DDR3 ECC
Video: EVGA GTX 1080
Monitor: Dell UP2715K
macOS Sierra 10.13.2

#12
Posted 04/18/2017 01:45 PM   
Can anyone confirm the latest drivers (378.05.05.05f02) fixes this issue?
Can anyone confirm the latest drivers (378.05.05.05f02) fixes this issue?

Mac Pro (Early 2009)
CPU: 2 x 3.46 GHz 6-Core Intel Xeon
RAM: 64 GB 1333 MHz DDR3
Video: NVIDIA GeForce GTX 980 TI
Monitor: Dell UP2715K
macOS High Sierra 10.13.1
--
Web Driver 378.10.10.10.20.109
CUDA Driver 9.0.222

#13
Posted 04/21/2017 07:44 AM   
[quote="Stez"]Can anyone confirm the latest drivers (378.05.05.05f02) fixes this issue?[/quote] I don't know if they fix this issue. But I don't think they do.
Stez said:Can anyone confirm the latest drivers (378.05.05.05f02) fixes this issue?


I don't know if they fix this issue. But I don't think they do.

#14
Posted 04/21/2017 04:32 PM   
The bug is still present with 378.05.05.05f02
The bug is still present with 378.05.05.05f02

#15
Posted 04/21/2017 11:23 PM   
  1 / 3    
Scroll To Top