Updated nVidia drivers, screens blacking out

I"m not getting a BSOD, but the system is recording all sorts of errors and warnings, so hopefully this is the right place...

Windows 10 Pro, EVGA Geforce 970 GTX. Winver SAYS it’s running 1803.

I GENERALLY try to keep the nvidia drivers current, but that may have been a mistake...

Yesterday morning I installed the “current” version of nvidia drivers – 398.82 (as of 10 minutes ago, this is still the current driver shown at EVGA for the 970 GTX.
Some time after that I started getting a screen black-out. Both monitors turn black, no response to keyboard or mouse, disk activity continues. Eventually, I hit the reset and crash the system. At some point AFTER that it happens again. May be minutes, may be hours.

The errors yesterday were similar to the ones below (these are happening today) – this is from the application log
Event 1000, Application Error
Faulting application name: dwm.exe, version: 10.0.17134.1, time stamp: 0xf5178e97
Faulting module name: unknown, version: 0.0.0.0, time stamp: 0x00000000
Exception code: 0xe0464645

Followed by Event 1001, Windows Error reporting
Fault bucket , type 0
Event Name: APPCRASH
Response: Not available
Cab Id: 0

Then by Event 0, Dwminit
The Desktop Window Manager process has exited. (Process exit code: 0x000000ff, Restart count: 1, Primary display device ID: NVIDIA GeForce GTX 970)

SO, I re-installed nvidia driver set 397.31, which had been on the machine previously. Problem CONTINUED…

Later last night, I found a note elsewhere that doing a reinstall MAY not clean things out and recommended using the Control Panel to completely uninstall the nVidia drivers. So, I went into Control panel and uninstalled all the 397.31 drivers I could find. Rebooted. System APPEARED to be OK. Running.

I’m not sure about performance, but at least it APPEARED to stay alive from 11:15 pm last night to 3:34 pm today (during which period there was no one using the system so I'm not sure that has any meaning), when everything went black again in the middle of a backup.
Between 3:23:47 and 3:23:52 the event log recorded SEVEN occurrences of the three errors above. As near as I can tell, they’re all the same, with dwm failing or crashing or whatever it does, over and over.

In the SYSTEM log, starting at 3:23:47 I received TWO occurrences of
Warning Event 4101, Display - Display driver nvlddmkm stopped responding and has successfully recovered.
At the same 3:23:47 timestamp

THEN, between 3:23:47 and 3:23:51 I received FOURTEEN occurrences of
Warning Event 263 Win32k - A pointer device has no information about the monitor it is attached to.

Then at 3:23:52 a Warning Event 4115, Display - Display driver failed to start; using Microsoft Basic Display Driver instead. Check Windows Update for a newer display driver.

Followed immediately by 3 more Warning Event 263 Win32k - A pointer device has no information about the monitor it is attached to.


At that point I hit the reset button and crashed the system.

This system is a normal, 30” primary monitor and 24” secondary.
It is used primarily for image processing.
It is NOT being used for games.
It is not a touch screen.

There have, prior to the latest nvidia drive update, NEVER been any errors or warnings of these types recorded as far as I know.

I’ve looked at all the errors on the Internet, but unfortunately the majority or reporting appears to happen in Micorsoft forums, and I’ve yet to see ANY ACTUAL SOLUTIONS for any question asked there. There’s an incessant demand for more information, eventually ending with something like “please let us know if your troubleshooting results in a solution. So, asking anything over there seems pointless.

The errors I’m seeing APPEAR to point to the GTX 970, and it’s worked flawlessly since installation. NOTHING new is happening on the system. No new work, no new hardware, no new software – other than the nVidia driver update (not counting all the updates Microsoft does).

Having gone back one version (of those I’ve installed) of nVidia drivers, then removing ALL the nVidia drivers, WHAT do I do now to get this thing to STOP locking up the screens?

I’ve got three older sets of nVidia drivers – do I start installing THOSE, hoping the one of them will work?
Is the problem being caused BY the nVidia drivers or is there a problem with the Micorsoft Windows driver(s)? Is there a way to UPDATE the Microsoft display driver or revert to some older driver?

What else should I be doing to get this fixed?

Hi GracieA. To my knowledge Dwm.exe is a Windows process to run certain hardware accelerated features on the desktop. Something may've been corrupted in the update method so the display driver (nvlddmkm - TDR) is crashing at random intervals, or at specific tasks.

Try this
1.) Download and install DDU (Display Driver Uninstaller) Official Display Driver Uninstaller DDU Download
2.) Run DDU. Select NVIDIA Software and drivers | Clean and restart
3.) Download the latest compatible Nvidia driver for your machine: https://www.nvidia.co.uk/Download/dr...x/136721/en-uk
4.) Install the driver and select "Custom" and "Clean" install.

Following this step should assure you're on a completely clean installation of your graphics card driver. If the driver install is corrupted this should clear it up. If not you may look at other pssibilites for this issue.

Hi,
Too much sludge in 1803 to go by error reporting
Although best drivers I've run into are 391.01 & 397.64
All time favorite was 385.69

Not sure why MS is in love with probably the worst driver series of which there were 6-388. releases none were worth beans most borked just by using monitoring utilities

There was an windows update ended for me with kb....19 that did cause black screens/ graphic fixes may also end with 18 saw in the news section Brink posts so many items I can't even begin to dig for it :)

Thanks for the replies. I'll try DDU. Things here are getting a little desperate since the system is now essentially unusable.

I found a topic over at nVidia discussions that claimed this problem was known with the GTX 970 since 1803 and that people were going back to 378.? (it was late last night). I did that. Made NO difference.

I'm getting mentally ready to reformat the partition and do a complete Windows re-install, which will be a complete nightmare and take approximately 4 days before I'm even remotely productive again.

OK, MAYBE fixed... System has been up for over an hour without puking... I ran DDU then installed the set of drivers that the nVidia forum SAYS is the newest version that DOESN'T cause the black screen...

I'll believe it when I wiggle the mouse in the morning and it wakes up........

Some versions of the drivers may cause problems for some users, but I know of no drivers that cause problems for all users.

I've never experienced your problem. I have two systems with nVidia graphics cards. One card is a GTX 970, the other is a 900 series Titan X (basically a 980ti).

I suppose that I'm unique, but I doubt that.

Well, it was nice while it lasted for the couple hours last evening...

Last night:
10:34:25 to 10:36:19 there were 36 msiinstaller entries, so apparently MS did some kind of update.
10:36:21 Event 1021 Perflib error
Windows cannot open the 32-bit extensible counter DLL Outlook in a 64-bit environment. Contact the file vendor to obtain a 64-bit version. Alternatively, you can open the 32-bit extensible counter DLL by using the 32-bit version of Performance Monitor. To use this tool, open the Windows folder, open the Syswow64 folder, and then start Perfmon.exe.

Event 1017 Perflib error
Disabled performance counter data collection from the "Outlook" service because the performance counter library for that service has generated one or more errors. The errors that forced this action have been written to the application event log. Correct the errors before enabling the performance counters for this service.

Event 2003 warning Perflib
The configuration information of the performance library "C:\Windows\System32\perfts.dll" for the "TermService" service does not match the trusted performance library information stored in the registry. The functions in this library will not be treated as trusted.

11:32:36 Event 916 - got the usual slew of ESENT information entries (looks to be about 30 of 'em in a row - usually a few HUNDRED a day)

11:32:45 - a couple Event 1001 Windows Error Reporting

And a Warning Event 6004 Winlogin - The winlogon notification subscriber <TrustedInstaller> failed a critical notification event.

Then the next slew of 20 or 30 ESENT information messages

Then for the next couple minutes a couple HUNDRED Event 1004 Security SPP starting services... A LOT of services.

And on and on - lots more ESENT information entries and so on.

I PRESUME ALL THIS STUFF IS SOMEHOW RELATED TO THE UPDATE? At least I HOPE so.

-------------------------------------------------------------------------

BUT, at 12:07:50 Got the Event 1000, Application Error
Faulting application name: dwm.exe, version: 10.0.17134.1, time stamp: 0xf5178e97

Event 1001, Windows Error Reporting APPCRASH

Event 0, Dwnminit - The Desktop Window Manager process has exited. (Process exit code: 0x000000ff, Restart count: 1, Primary display device ID: NVIDIA GeForce GTX 970)

This was all repeated 9 times...

System stayed up and continued writing to the event logs, but, of course, had to crash and restart it at 4:30 a.m. when I tried to wake the system.

--------------------------------------------

The SYSTEM log shows the herd of Windws Updates, then says the system crashed at 11:32 last evening

At 11:32 it logged an error that the system rebooted from a bugcheck.

At 12:06 it rebooted from another bugcheck.

Then at 12:07 a list of a dozen or so Event 263 A pointer device has no information about the monitor it is attached to.
I usually see this error when dwm.exe is having whatever fit it's having.....

And then just the "normal" several hundred information messages and
several Event 34 BTHUSB warnings (The local adapter does not support an important Low Energy controller state to support peripheral mode. The minimum required supported state mask is 0x491f7fffff, got 0x1fffffff. Low Energy peripheral role functionality will not be available.)
and several Event 10016 ERROR DistributedCOM entries that everyone says to just ignore.

I've tried the things I've found on the Internet for getting rid of the ESENT, BTHUSB and DistributedCOM stuff, but nothing has worked...

The Event 263 makes no sense to me 'cause it's not a touch screen and The Touch Keyboard and Handwriting Panel Service is disabled.

SO - Did the MS update of whatever it was installing CAUSE this thing to go back to having dwm fail, or is all this just coincidence?

I'll let it go today and see if it starts frequently giving me the black screen.

NO GOOD. Reran DDU and it's still dumping to the black screen. A LOT. System is essentially unusable at this point.

1) Boot into safe mode:
Boot into Safe Mode on Windows 10

2) Run the BETA log collector and post a zip into this thread:
https://www.tenforums.com/attachment...-v2-beta15.zip
Use the txt and images in this thread in case there are problems running the beta log collector: (post #5)
DM Log tool problem

3) Check to see whether you can remove the Nvidia GPU and use default motherboard graphics