Computer rebooted from a bug check 0x0000003b.

Hi and thanks for reaching out. My name is William. I'm a Microsoft Windows Certified Professional and Systems Administrator. I'll be happy to help you out today.

0x0000003b is otherwise known as a SYSTEM_SERVICE_EXCEPTION, i.e. an essential Windows service (or even a 3rd party service) cause a fault in the kernel, halting the system.

When your system crashes it will create a dump file(s) in C:\Windows\Minidumps. I can analyze those files for the potential cause of the system halting. Can you upload the .dmp file(s) to OneDrive, Google Drive, Dropbox or some other file sharing service and share the link here?

Yes, there are two dmp files and they are both from today. The second one has happened since making the original post. Here is the link to a google drive folder containing them both: https://drive.google.com/drive/folders/16Zs5o2zNz6a3yheGlRr0JNd6OFdYjhlv?usp=sharing

The stack trace is not remarkable, no signs of 3rd party drivers. There is some access checking being done on object prior to bugcheck.

11: kd> k
# Child-SP RetAddr Call Site
00 ffffba0c`afb70c38 fffff802`52406469 nt!KeBugCheckEx
01 ffffba0c`afb70c40 fffff802`524058bc nt!KiBugCheckDispatch+0x69
02 ffffba0c`afb70d80 fffff802`523fd402 nt!KiSystemServiceHandler+0x7c
03 ffffba0c`afb70dc0 fffff802`52252d97 nt!RtlpExecuteHandlerForException+0x12
04 ffffba0c`afb70df0 fffff802`522519a6 nt!RtlDispatchException+0x297
05 ffffba0c`afb71510 fffff802`524065ac nt!KiDispatchException+0x186
06 ffffba0c`afb71bd0 fffff802`52402743 nt!KiExceptionDispatch+0x12c
07 ffffba0c`afb71db0 fffff802`522e8cd8 nt!KiPageFault+0x443
08 ffffba0c`afb71f40 fffff802`522e73d4 nt!SepAccessCheck+0x6c8
09 ffffba0c`afb720b0 fffff802`522e5c07 nt!SeAccessCheckWithHintWithAdminlessChecks+0x4f4
0a ffffba0c`afb72290 fffff802`526f0319 nt!SeAccessCheck+0x67
0b ffffba0c`afb72300 fffff802`526afbe9 nt!ObpCreateHandle+0x7c9
0c ffffba0c`afb724f0 fffff802`526ad0c3 nt!ObOpenObjectByPointer+0x1b9
0d ffffba0c`afb72770 fffff802`526acea3 nt!PsOpenThread+0x213
0e ffffba0c`afb72ac0 fffff802`52405eb8 nt!NtOpenThread+0x23
0f ffffba0c`afb72b00 000002a1`f8b0aa34 nt!KiSystemServiceCopyEnd+0x28
10 000000b3`ffffc1f8 00000000`00000000 0x000002a1`f8b0aa34

The object is likely a driver. Which driver, I cant specifically say, but I can infer from the thread dump it is Nvidia display driver, nvlddmkm.sys:

...
ffffba0c`afb70ba8 fffff802`5af5dbffUnable to load image \SystemRoot\System32\DriverStore\FileRepository\nv_dispi.inf_amd64_edf184f24a37bacd\nvlddmkm.sys, Win32 error 0n2
*** WARNING: Unable to verify timestamp for nvlddmkm.sys
nvlddmkm+0x1cdbff
ffffba0c`afb70bb0 00000000`00000001
ffffba0c`afb70bb8 00000000`00000000
ffffba0c`afb70bc0 ffffba0c`afb71d08
ffffba0c`afb70bc8 ffffba0c`afb70e30
ffffba0c`afb70bd0 fffff802`52000000 nt!VrpRegistryString <PERF> (nt+0x0)
ffffba0c`afb70bd8 ffffba0c`afb713e0
ffffba0c`afb70be0 fffff802`52405eb8 nt!KiSystemServiceCopyEnd+0x28
ffffba0c`afb70be8 fffff802`52405840 nt!KiSystemServiceHandler
ffffba0c`afb70bf0 ffffba0c`afb71330
ffffba0c`afb70bf8 fffff802`523f4647 nt!KeBugCheckEx+0x107
ffffba0c`afb70c00 00000000`00000001
ffffba0c`afb70c08 fffff802`5ae9c522 nvlddmkm+0x10c522
ffffba0c`afb70c10 fffff802`520e3868 nt!setjmpexused <PERF> (nt+0xe3868)
ffffba0c`afb70c18 00000000`00000000
ffffba0c`afb70c20 00000000`00000000
ffffba0c`afb70c28 00000000`00000000
ffffba0c`afb70c30 00000000`00040282
ffffba0c`afb70c38 fffff802`52406469 nt!KiBugCheckDispatch+0x69
ffffba0c`afb70c40 00000000`0000003b
ffffba0c`afb70c48 00000000`c0000005
ffffba0c`afb70c50 fffff802`522e8cd8 nt!SepAccessCheck+0x6c8
ffffba0c`afb70c58 ffffba0c`afb71540
...

Without another dump to compare tis against, I would recommend rolling back to an older build for the GPU since you are at the current build.

However, if you can reproduce this fairly consistently, then you should enable driver verifer as it may assist with isolating the cause at next crash. To enable driver verifier, see this guide: https://www.sysnative.com/forums/threads/driver...

Note, you do not want to leave driver verifier running for extended times. Because of its invasive nature hooking 3rd party drivers, it can introduce performance issues into the system or crash it out-right on boot if issues are detected with boot drivers.

When you recommend "rolling back to an older build for the GPU" do you mean an older driver version?

Also if you can confidently say that it is Nvidia display driver, does this mean I am not at risk of my BSoDs causing a harddrive malfunction and I loosing all my data? I am currently creating a clone of my disk, but I am nervous about losing all my data.

So both dumps indicate hardware as the cause, but is rarely the case; the WinDbg heuristics engine just does not have enough to go on. Without a conclusive stack trace, I cannot tell you for sure what is causing bugcheck, I can only infer.

I looked at the other dump just now and this is similar to the first in that system halts when doing some access check routines, but in this case this happened while parsing a registry key. Usually, when I see registry key prior to crash, I suspect 3rd party antivirus reg filters. I did not see in the 1st dump, but in this one I see an unloaded module for mcupdate.dll, a McAfee antivirus module. You wouldn't happen to be running McAfee?

No, I am not currently running any 3rd party antivirus software. I use Windows Defender to prevent viruses and threats. I have noticed that most of the time when it crashed I have Opera GX web browser open....could that be the culprit?

In both crashes I did see that opera was running in the active thread, but normally user-mode processes do not spill over into the kernel and halt the system (although it is known to happen, and specifically where SYSTEM_SERVICE_EXCEPTION (3b) are thrown, which is the case in these two dumps. To see if that is the cause requires enabling user-mode crash dumps and then seeing if their is a corresponding dump with the next bugcheck.

To enable user-mode dumps, see this Microsoft article: https://docs.microsoft.com/en-us/windows/win32/...

Alright, I think I have set up user-mode crash dumps (see image). If this happens again should I just reply to this discussion? In the meantime is there something you recommend me to do? 

You can examine the Event Viewer > Windows Logs > Application & System for errors or warnings that coincide with the bugcheck. Some of the events may offer clues as to the state of system right before the event. You can upload those as well and I can take a look, just right-click Applications and System logs respectively > Save all events as...

Alright I will do that this afternoon, I am currently in a power outage. Now that I am thinking about it, all of my PC problems started after a power outage last week.