Windows 10: Constant BSODs under medium load while gaming

Hi TenForums, I'm posting an issue here that I've already fruitlessly posted on the microsoft forums because I'm all out of ideas.... The text from that post is copied below:

Recently I've started having BSODs occur while running games, and more infrequently at random times when the PC is not really under any load. These have been consistently DPC_WATCHDOG_VIOLATION, IRQL_LESS_THAN_OR_EQUAL, BAD_OBJECT_HEADER, and a few other less frequent stop codes. I'm now able to pretty consistently recreate the crash by running a game of Overwatch, usually with 3 games I'll BSOD with one of these stop codes.
I've tried a series of debugging methods so far:

1. Memtest86, found no errors, also did the RAM DIMM slot shuffle, again finding no issues...
2. Replaced the RAM anyway, No change.
3. Reinstalled Windows on a series of different physical hard drives and one SSD, no change.
4. Replaced the motherboard entierly (after my machine started throwing these same errors while trying to install windows from a bootable USB made with the media creation tool. No longer convinced the old Mobo was bad), allowed me to install and run windows without immediate BSOD, but original problem not solved
5. Swapped out the GPU & updated GPU drivers, no effect
6. Swapped PSU, in case old psu went bad, no effect
7. Updated BIOS on both Mobos, no effect
8. Ran the Intel Processor Diagnostic Tool, CPU Passes, including load tests
9. Implemented the suggestions for googling "DPC_WATCHDOG_VIOLATION windows 10" [rolled SATA AHCI driver to default, disabled fast boot in bios/windows], no effect
10. Tried to some success to recreate the crash by just loading the CPU via CPU-z, while I did get it to BSOD once I haven't been able to repeat it.
11. Updated all the drivers available on my motherboards manufacturer page, no effect
12. Ran System Tests in Ubuntu finding no issues, though I'll admit these were not really comprehensive.

While the CPU does pass the Intel tool, I've not had a chance to replace it with a known working unit as I don't currently have access to one of the correct package for my motherboard.
Pretty much at the end of my wits here as far as what to do next...

System info:
Operating System: Windows 10 Pro 64-bit (10.0, Build 15063) (15063.rs2_release.170317-1834)
Language: English (Regional Setting: English)
System Manufacturer: MSI
System Model: MS-7A64
BIOS: 1.20
Processor: Intel(R) Core(TM) i7-6700K CPU @ 4.00GHz (8 CPUs), ~4.0GHz
Memory: 16384MB RAM
Available OS Memory: 16344MB RAM
Page File: 8034MB used, 11252MB available
Windows Dir: C:\Windows
DirectX Version: DirectX 12
Link to some of the latest crash dumps/minidumps:
CrashDumps.zip

And here is a CPU-Z report of my current system:
DESKTOP-PMEM6GD.txt


I'll add to that post:

Since Posting I've now run the prime95 stress test with no issues for several hours.
SFC has verified the integrity of my file system...
I've now individually recreated the BSOD while testing every single peripheral I have.
As mentioned in the above post I dont have access to a spare LGA 1151 CPU to really, truly, eliminate it as the cause, but no tests I've run point to it as the cause.
Voltage levels, both idle and under load, displayed via HWMonitor are nominal...
I've down graded my NVidia drivers to those from March 1, recreated the BSOD, and then upgraded to the latest. No effect...

I've also made a reddit thread on the matter and was asked to provide model info. I'll include it below:
Model Info:
CPU
Intel Core i7 6700K @ 4.00GHz
RAM
16.0GB HyperX FURY (2 x 8GB) DDR4 2133
Motherboard
MSI B250 GAMING PRO CARBON (MS-7A64) BIOS: 1.20
Graphics
4095MB NVIDIA GeForce GTX 1070 (EVGA)
Storage
KINGSTON SV300S37A120G (SSD) (os)
Seagate ST2000DX002-2DV164 (SATA)
SAMSUNG HD103SJ (SATA)
Hitachi HDP725050GLA360 (SATA)
Optical Drives
ATAPI iHAS224 A
PSU
Corsair 750w CMPSU-750TX-C
Peripherals
Logitech G502
Razor Blackwidow Ultimate
All other disconnected for now

Sorry in advance if I've broken any posting guidelines, only discovered this forum tonight.

:)

 

Constant BSODs under medium load while gaming.

If that's the case we suggest that you do a rollback of the previous build of your Windows 10 and let see if that helps.

You may refer to the link below on how you rollback to the previous build.

Recovery options in Windows 10

Hope to hear from you soon.

 

Constant BSODs under medium load while gaming.

Just a follow up question, did the issue started after updating your computer to the Creators Update?

 

Microsoft (R) Windows Debugger Version 10.0.15063.137 AMD64 is malfunctioning with GetContextState failed, 0x8007001E

Two of the dumps: Probably caused by : memory_corruption


Eventlog:

Code: Event[3553]: Log Name: System Source: Service Control Manager Date: 2017-06-02T21:05:08.432 Event ID: 7026 Task: N/A Level: Information Opcode: N/A Keyword: Classic User: N/A User Name: N/A Computer: DESKTOP-PMEM6GD Description: The following boot-start or system-start driver(s) did not load: dam[/quote]
Code: Event[3499]: Log Name: System Source: Service Control Manager Date: 2017-06-02T20:44:55.369 Event ID: 7026 Task: N/A Level: Information Opcode: N/A Keyword: Classic User: N/A User Name: N/A Computer: DESKTOP-PMEM6GD Description: The following boot-start or system-start driver(s) did not load: dam[/quote] Code: Event[3393]: Log Name: System Source: Service Control Manager Date: 2017-06-02T20:05:55.276 Event ID: 7026 Task: N/A Level: Information Opcode: N/A Keyword: Classic User: N/A User Name: N/A Computer: DESKTOP-PMEM6GD Description: The following boot-start or system-start driver(s) did not load: dam[/quote]
There were multiple crashes and this event has preceded the crashes:

Code: Event[1130]: Log Name: System Source: Microsoft-Windows-FilterManager Date: 2017-05-30T21:39:42.180 Event ID: 6 Task: N/A Level: Information Opcode: Info Keyword: N/A User: S-1-5-18 User Name: NT AUTHORITY\SYSTEM Computer: DESKTOP-PMEM6GD Description: File System Filter 'npsvctrig' (10.0, ?2097?-?07?-?25T06:18:05.000000000Z) has successfully loaded and registered with Filter Manager.[/quote]




Code: Event[1610]: Log Name: System Source: Microsoft-Windows-Kernel-PnP Date: 2017-05-31T00:12:43.633 Event ID: 225 Task: N/A Level: Warning Opcode: Info Keyword: N/A User: S-1-5-18 User Name: NT AUTHORITY\SYSTEM Computer: DESKTOP-PMEM6GD Description: The application \Device\HarddiskVolume4\Program Files (x86)\Google\Update\1.3.33.5\GoogleCrashHandler.exe with process id 9356 stopped the removal or ejection for the device PCI\VEN_8086&DEV_A282&SUBSYS_7A641462&REV_00\3&11583659&0&B8.[/quote] Code: Event[217]: Log Name: System Source: Microsoft-Windows-Kernel-PnP Date: 2017-05-31T04:57:06.383 Event ID: 219 Task: N/A Level: Warning Opcode: Info Keyword: N/A User: S-1-5-18 User Name: NT AUTHORITY\SYSTEM Computer: DESKTOP-PMEM6GD Description: The driver \Driver\WudfRd failed to load for the device SWD\WPDBUSENUM\_??_USBSTOR#Disk&Ven_SanDisk&Prod_Cruzer_Glide&Rev_1.27#20042605700307F0A611&0#{53f56307-b6bf-11d0-94f2-00a0c91efb8b}.[/quote]
Code: Event[328]: Log Name: System Source: Microsoft-Windows-Kernel-Boot Date: 2017-05-30T15:18:11.920 Event ID: 29 Task: N/A Level: Error Opcode: Info Keyword: N/A User: S-1-5-18 User Name: NT AUTHORITY\SYSTEM Computer: DESKTOP-PMEM6GD Description: Windows failed fast startup with error status 0xC00000D4.[/quote]
Code: Event[451]: Log Name: System Source: Microsoft-Windows-WindowsUpdateClient Date: 2017-05-30T15:23:06.797 Event ID: 20 Task: Windows Update Agent Level: Error Opcode: Installation Keyword: Installation,Failure User: S-1-5-18 User Name: NT AUTHORITY\SYSTEM Computer: DESKTOP-PMEM6GD Description: Installation Failure: Windows failed to install the following update with error 0x80246007: VCLibs 12.[/quote]
Code: Event[604]: Log Name: System Source: Service Control Manager Date: 2017-05-30T15:39:27.353 Event ID: 7000 Task: N/A Level: Error Opcode: N/A Keyword: Classic User: N/A User Name: N/A Computer: DESKTOP-PMEM6GD Description: The MSICDSetup service failed to start due to the following error: This driver has been blocked from loading[/quote]
Code: Event[1221]: Log Name: System Source: Microsoft-Windows-Kernel-Boot Date: 2017-05-30T22:06:38.810 Event ID: 29 Task: N/A Level: Error Opcode: Info Keyword: N/A User: S-1-5-18 User Name: NT AUTHORITY\SYSTEM Computer: DESKTOP-PMEM6GD Description: Windows failed fast startup with error status 0xC00000D4.[/quote] Code: Event[1678]: Log Name: System Source: User32 Date: 2017-05-31T01:52:39.359 Event ID: 1073 Task: N/A Level: Warning Opcode: N/A Keyword: Classic User: S-1-5-21-1390793526-556943757-3711158197-1001 User Name: DESKTOP-PMEM6GD\svikh Computer: DESKTOP-PMEM6GD Description: The attempt by user DESKTOP-PMEM6GD\svikh to restart/shutdown computer DESKTOP-PMEM6GD failed[/quote] Code: Event[1813]: Log Name: System Source: Microsoft-Windows-Kernel-Boot Date: 2017-05-31T10:00:25.786 Event ID: 29 Task: N/A Level: Error Opcode: Info Keyword: N/A User: S-1-5-18 User Name: NT AUTHORITY\SYSTEM Computer: DESKTOP-PMEM6GD Description: Windows failed fast startup with error status 0xC00000D4.[/quote] Code: Event[2095]: Log Name: System Source: Microsoft-Windows-WindowsUpdateClient Date: 2017-05-31T12:01:19.535 Event ID: 20 Task: Windows Update Agent Level: Error Opcode: Installation Keyword: Installation,Failure User: S-1-5-18 User Name: NT AUTHORITY\SYSTEM Computer: DESKTOP-PMEM6GD Description: Installation Failure: Windows failed to install the following update with error 0x80246007: Microsoft Sticky Notes.[/quote] Code: Event[2154]: Log Name: System Source: Microsoft-Windows-WindowsUpdateClient Date: 2017-05-31T12:02:10.509 Event ID: 20 Task: Windows Update Agent Level: Error Opcode: Installation Keyword: Installation,Failure User: S-1-5-18 User Name: NT AUTHORITY\SYSTEM Computer: DESKTOP-PMEM6GD Description: Installation Failure: Windows failed to install the following update with error 0x80073D02: Microsoft Sticky Notes.[/quote]
Code: Event[2318]: Log Name: System Source: Service Control Manager Date: 2017-05-31T12:34:10.454 Event ID: 7000 Task: N/A Level: Error Opcode: N/A Keyword: Classic User: N/A User Name: N/A Computer: DESKTOP-PMEM6GD Description: The MSICDSetup service failed to start due to the following error: This driver has been blocked from loading[/quote]
Code: Event[2324]: Log Name: System Source: Service Control Manager Date: 2017-05-31T12:35:40.470 Event ID: 7000 Task: N/A Level: Error Opcode: N/A Keyword: Classic User: N/A User Name: N/A Computer: DESKTOP-PMEM6GD Description: The MSICDSetup service failed to start due to the following error: This driver has been blocked from loading[/quote]
Code: Event[2373]: Log Name: System Source: Microsoft-Windows-Kernel-PnP Date: 2017-05-31T12:44:24.900 Event ID: 219 Task: N/A Level: Warning Opcode: Info Keyword: N/A User: S-1-5-18 User Name: NT AUTHORITY\SYSTEM Computer: DESKTOP-PMEM6GD Description: The driver \Driver\WudfRd failed to load for the device SWD\WPDBUSENUM\_??_USBSTOR#Disk&Ven_Mass&Prod_Storage_Device&Rev_1.00#121220130416&0#{53f56307-b6bf-11d0-94f2-00a0c91efb8b}.[/quote]
Code: Event[2806]: Log Name: System Source: Microsoft-Windows-Directory-Services-SAM Date: 2017-05-31T21:53:59.900 Event ID: 12291 Task: N/A Level: Error Opcode: Info Keyword: N/A User: S-1-5-18 User Name: NT AUTHORITY\SYSTEM Computer: DESKTOP-PMEM6GD Description: SAM failed to start the TCP/IP or SPX/IPX listening thread[/quote]



These are some of the notable or recurrent finding in the event log related to failure:
The MSICDSetup service failed to start due to the following error: This driver has been blocked from loading
The driver \Driver\WudfRd failed to load for the device SWD\WPDBUSENUM\_??_USBSTOR#Disk&Ven_Mass&Prod_Storage_Device&Rev_1.00#121220130416&0#{53f56307-b6bf-11d0-94f2-00a0c91efb8b}.
The following boot-start or system-start driver(s) did not load: dam
Windows failed fast startup with error status 0xC00000D4.
Installation Failure: Windows failed to install the following update with error 0x80246007: VCLibs 12.
Installation Failure: Windows failed to install the following update with error 0x80246007: Microsoft Sticky Notes.
Installation Failure: Windows failed to install the following update with error 0x80073D02: Microsoft Sticky Notes.


dam.sys DAM Kernel Driver Windows Update

WUDFRd.sys Windows Driver Foundation - User-mode Driver Framework Reflector Windows Update

npsvctrig.sys Named pipe service t Windows Update

For the bsod what information did you see from the whocrashed analysis?
Are you doing any over clocking?
Are you using software or hardware RAID?



In administrative command prompt type or copy and paste these commands:

1) winver (in the pop up view the windows version and build. the latest build is 15063.332) There were some failed windows updates in the event log. If there are currently any failed windows updates post the KB# and failure code.

2) sfc /scannow
3) dism /online /cleanup-image /restorehealth
4) chkdsk /scan

5) There were multiple windows failed fast start up.
Turn off fast startup:
Turn On or Off Fast Startup in Windows 10 Windows 10 Performance Maintenance Tutorials

6) In your post you indicated that you had ran memtest86.
With two bsod related to memory corruption please test the memory using memtest86+
Version 5.01
Memtest86+ - Advanced Memory Diagnostic Tool
Make at least 8 runs. Sometimes it takes runs in the teens to find ram or dim related problems. Run memtest86+ overnight.


7) After the above plan to run windows driver verifier.
Are you comfortable navigating through the windows recovery menu or windows advanced troubleshooting menu to safe mode with command prompt?

Open administrative command prompt and type or copy and paste this command:
shutdown /r /o /f /t 00
Navigate to safe mode with command prompt.
Type: shutdown /r

When you have completed memtest86+ with 8 or more runs and it is a pass this link has the settings for windows driver verifier:
Driver Verifier-- tracking down a mis-behaving driver. - Microsoft Community

Open system control panel > advanced system settings > startup and recovery settings > un-check automatically restart.

For windows driver verifier to start type verifier in the left lower corner search and then follow the link above.
To turn off windows driver verifier you will need to navigate through the windows advanced troubleshooting menu to startup options then safe mode with command prompt and type verifier /reset

When back on the desktop you should have already installed both whocrashed and bluescreenview. The minidump files are immediately analyzed and the driver/applications are uninstalled.
Resplendence Software - WhoCrashed, automatic crash dump analyzer
Blue screen of death (STOP error) information in dump files.
Windows driver verifier is restarted and the next bsod is analyzed, etc.
Typically windows driver verifier will post a stop code with driver verifier detected violation. Sometimes it will display the misbehaving driver with *.sys. Most often the minidump files need to be analyzed with whocrashed and bluescreenview. Sometimes the analysis is different so make sure you also post the minidump files into the forum for analysis.


Please report each:
1) windows version and build, and any failed update KB# and error code
2) scannow
3) restorehealth
4) chkdsk /scan (you can use this administrative command prompt command to generate a desktop text file:

Read Chkdsk Log in Event Viewer in Windows 10 Windows 10 Performance Maintenance Tutorials

get-winevent -FilterHashTable @{logname="Application"; id="1001"}| ?{$_.providername –match "wininit"} | fl timecreated, message | out-file Desktop\CHKDSKResults.txt


5) memtest86+ version 5.01 with at least 8 runs

 

Hello Haroo *Smile

If I may also add a suggestion:
When updating graphics drivers I recommend a complete uninstall of the existing drivers and all NVidia software before reinstalling just the driver and PhysX components. These are the instructions you will see repeated a lot on this forum:

Please update the NVidia driver. The best way to do this is to uninstall everything of Nvidia using Display Driver Uninstaller and install new drivers from your Graphics card manufacturer or from the NVidia website. When you do this be sure the "clean install" box is checked and only install the Graphics driver and the PhysX driver.

I'm not sure from your write up if this is the method you used.

Also like @zbook I must stress that the memtest86+ result is only valid if you have tested for a minimum of 8 passes, sometimes errors have only been spotted after 10 or more passes.

 

@zbook: Working thru your list now, I did run 8 (and then 4) passes of memtest86 previously, on both pairs of ram (the Ripjaws and HyperX), Ill run a -teen number of passes tonight in 86+. Following that I'll run the suggested Driver Verification process, and will report both of these back.

Ok, as for the list:

• Version 1703 - Build 15063.332, I'm not sure how to check for failed KB updates. I'm not seeing an failed in the update log in the update pane of the settings app.
• SCANNOW returned no integrity violations.
• The DISM completed successfully.
• CHKDSK also found now issues as far as I can tell. I've attached the report to my OP, though I had to manually copy it from the event viewer since the Get-Winevent command was just generating a blank file (and the event ID in my event viewer is not 1001, but 26226, though changing that part of the command still generated a blank file....) Let me know if the copied report is insufficient.

I've also attached a WhoCrashed output. I've been running it after each BSOD to try and spot if its a hardware issue (which I'm way more comfortable solving...) or some software issue. I'd say, across my multiple windows installs (since last Friday), WhoCrashed has consistently pointed to software driver bugs, though I've been totally unable to fix these. The NVida Kernal driver nvlddmkm.sys (nvlddmkm+0xD9E17), Windows Kernal ntoskrnl.exe (nt+0x16C310), and dxgmms2.sys (0xFFFFF80D1FD05BC7) have been the most commonly identified sources, though HAL.dll, has also occasionally made an appearance.


No Overclocking, nor any RAID configuration. Fast Boot is disabled both in Power Options, and in my motherboard's BIOS settings.

@philc43: I totally forgot about DDU! I'll roll back the drivers using that method and report back.

Thank's for the replies!

 

So @philc43, did a clean install of the latest NVidia Drivers with DDU and the 'clean install' option. Still resulted in BSOD, but that is what I was expecting given that installing the drivers on a fresh windows install also BSOD'd. Though I'll make sure to follow this method for future driver updates.
@zbook: I was unable to get memtest86+ onto a bootable USB, I tried a variety of image writers (RUFUS, UUI, & imageUSB), I also tried the USB Key installer, as well as trying to burn the bootable binary. Internet searches on the issue did not yield a solution. Not wanting to let the night go to waste I re-imaged memtest86 and ran it for 13 passes. Found 2 errors! This is on the new sticks of RAM, I'll run the same passes tonight on the old I guess. I'll attach the HTML output of the test to my OP.

Given that there are errors I'll first try and figure out if I own any non-faulty RAM before diving into the driver verification.

Not sure how likely it is that I had RAM go bad and then ordered new, but faulty, RAM, but I guess it's not impossible...

 

The next step is to determine if it is the RAM module or the slot on the motherboard that is faulty. You can do this by re-running the single stick in different slots and repeating for the other stick to determine which RAM module or slot is faulty. By a process of elimination you will find out. I just hope you don't have to wait for 13 passes each time!

 

OK as Philc43 posted you will need to run the RAM one at a time to find the faulty RAM and exclude faulty dim.

 

So smallish update:

I've been running many memtests and, much to my frustration, have been unable to repeat those errors... Even after 20 passes on each stick and dimm slot....

Furthermore the BSODs are back, even on my assumed 'good' ram....

So I've been trying the verifier route, to some result?

Following what zbook recommended I've been running the verifier following the included guide, and basically being unable to boot into windows without a "DRIVER_VERIFIER_DETECTED_VIOLATION" BSOD. I have a restore point up that I keep reverting too and I'm uninstalling the flagged driver (flagged in BSODviewer, whocrashed, and winDBG), but it seems that every driver flagged for verification is causing the BSOD within seconds of boot (before I hit login). I'm uninstalling in batches and not checking certain drivers in verifier setup, so its not a matter of me forgetting that the restore point reversion reinstalled the drivers.

So I wanted to clarify some instructions: The guide in-closed in Zbook's post says not to run the verifier in a safe mode setting since some of the offending drivers might not launch in safemode, though Zbook seem's to indicate that I should be running in safe mode. I see the point made in the guide, maybe I missed zbook's point?

Some of the drivers flagged were drivers not associated with any installed software, such as ndisfrl.sys (Intel (R) Technology Access), so I assume they were automatically installed by windows update?

I'll keep plugging away at checking which drivers are tripping the verifier, but given that I've been unable to reach the login screen at all with verifier without hitting a fault I'm not sure I've set verifier correctly (or my system is in really bad shape, which is odd given this is a fairly fresh install).

Any advice from here?

Ninja edit: Decided to test if I was doing something totally wrong with Verifier and set it up to to verify only 1 random driver (Hal.dll), and I boot into windows just fine, so guess not all drivers get auto flagged, though I'm sure this just has to do with when they are loaded....

Edit2: Managed to exclude the right drivers to get to login (not verifying any intel, logitech, or razor drivers. As well as one from FINTEK which is apparently associated with the MSI gaming app. Machine is running sluggishly which I'm guessing is expected. I've already gotten 2 BSODs with these settings, one BAD_OBJECT_HEADER originating from ntoskrnl.exe, and a DPC_WATCHDOG_VIOLATION that didnt finish making a dump for some reason (I saw the BSOD stop code on the actual BSOD, but no dump files reference that particular BSOD). winDBG pulls up inconclusive results for the BAD_OBJECT_HEADER BSOD, so im only able to go off whocrashed and bluescreenview for that one.

 

When looking for misbehaving drivers using windows driver verifier you should be using a normal boot. You should not be testing while in safe mode as in safe mode the non-Microsoft drivers are off.
In order to move from a windows driver verifier bsod to the desktop you must turn off windows driver verifier. Without turning it off it will always bsod. In order to turn off windows driver verifier you need to use the windows advanced troubleshooting menu, choose startup options, pick choice #6 safe mode with command prompt, then enter verifier /reset.
Once you enter the command verifier /reset you then should be able to return to the desktop. Then run whocrashed and bluescreen view to identify the misbehaving driver and uninstall it.
Then repeat the use of windows driver verifier. Find another bsod, turn off windows driver verifier, return to the desktop, run who crashed and bluescreenview, uninstall the next driver etc. It's a rinse and repeat.
At some point you will need to install the most up to date driver and retest them using windows driver verifier.
You can keep repeating the process as long as you turn on windows driver verifier with the command verifier and turn it off with the command verifier /reset.

Every non-Microsoft driver should be tested.
When you have new bsod you can update the zip file or create a new one so we can process the bsod information.

 

So an update: It was the CPU. Finally caved and ordered a new one, assuming I'd just end up returning it, but it ended up being the solution. At this point I'm running all 4 sticks of ram with no further issues. Guess the CPU testing tools I was using weren't hitting the right functionality? Thanks for all the help!

 

What lead you to the CPU and how did you confirm it as the problem?

 

This is interesting and a salutary lesson. It is not the first time that I have seen CPU tests pass, including Intel's own CPU test, only for it to be later confirmed that the CPU is faulty. I'm pleased your perseverance paid off. *Smile