Question PC randomly getting BSODs, possibly due to Hypervisor ?

[USAFRet]

About a year.

A year?
To reinstall your applications, and get setting to where they should be?

An entire year?


What is your current backup procedure?

 

[ZivDero]

The last time I ran Prime95, my computer instantly got a BSOD right after I ran it.

Yeah, while I'm not a super tech expert, Prime95 failing (and this fast and spectacularly, for me it was just 1 worker that failed) would probably indicate that your hardware is very unstable.

I would probably first run memtest to make sure it isn't your RAM, and then probably blame the CPU. But someone more experienced than me should chime in before the CPU is condemned 😀

P. S. When you get into a boot loop with driver verifier, it should eventually let you into recovery mode wherefrom you can boot into safe mode, and from there disable verifier.

 

[ConorDuey2000]

A year?
To reinstall your applications, and get setting to where they should be?

An entire year?


What is your current backup procedure?

My apps and settings are on my C drive. All of my files and documents are on another drive that I don't have to reformat.

 

[USAFRet]

My apps and settings are on my C drive. All of my files and documents are on another drive that I don't have to reformat.

And that is a good construct.
My systems are the same.
OS and application on one drive...other data on other drives.

But that does not address what your current backup procedure is.


For me, if my C drive died....an hour after I slotted in a replacement, done.

 

[ConorDuey2000]

And that is a good construct.
My systems are the same.
OS and application on one drive...other data on other drives.

But that does not address what your current backup procedure is.


For me, if my C drive died....an hour after I slotted in a replacement, done.

I heard that backing up my C drive in a case like this would be pointless.

 

[USAFRet]

I heard that backing up my C drive in a case like this would be pointless.

Backups are what you do before the BadThing happens.

To prevent all the crying, trying to do it after.


In the current situation, have you ruled out actual hardware issues?
Install the OS on a new drive, just as a test.
(I know this has been said before)

 

[Colif]

Yeah, while I'm not a super tech expert, Prime95 failing (and this fast and spectacularly, for me it was just 1 worker that failed) would probably indicate that your hardware is very unstable.

I would probably first run memtest to make sure it isn't your RAM, and then probably blame the CPU. But someone more experienced than me should chime in before the CPU is condemned

He has had problem for about 3 years, in that time every single person who understands BSOD on this forum has tried to solve it. I did for a few years, then Ubuysa took over... others have poked heads in along way.

I am fairly sure in that time I would have suggested memtest, its one of the obvious tests for BSOD. Reason I can see for it not being ram is, memory errors are normally a lot more random and generally point at drivers, not the hypervisor.

I think Others would have been able to tell if it was ram or not from the dump files.

Mobo: ASUS TUF GAMING X570-PLUS (WI-FI)
Mobo BIOS Version: American Megatrends Inc. 5021
CPU: AMD Ryzen 7 3800X 8-Core CPU
GPU: Nvidia GeForce RTX 2060
OS: Windows 11 Pro 24H2

Prime95 crashing could mean a number of things, could be PSU... he says he has an EVGA but I don't know if he ever looked at label to see which one. (mentioned in last thread)
could be ram not getting enough power.
could be CPU
could even be heat as it depends on what cooler he has, Prime will push CPU to max temp pretty quick.

It creates a report in the same folder it runs from, but it may not have recorded anything if it crashes right away.

Its not always the hypervisor but its been a reoccurring BSOD, it seems to come and go. BSOD generally don't fix themselves.

In a previous thread I did suggest it could be the CPU, but the testing a new install is one way to see if its not just a combo of old drivers causing the errors.

 

[ConorDuey2000]

I got another BSOD that's worth posting about. WinDbg said that it was a SYSTEM_SERVICE_EXCEPTION BSOD caused by the driver, dxgmms2.sys, and the program, Discord.exe. It seems like a pretty weird reason to get a BSOD and it looks like it's a graphics issue. Also, here is the kernel dump and here are my most recent minidumps.

 

[Colif]

Buy the sata cable and install windows on the ssd.

If you still get BSOD on the ssd, we will look at those. But we have spent 4 years trying to fix one install. I highly doubt your problem is just a graphics one.

Spoiler

File: 012225-81578-01.dmp (Jan 23 2025 - 11:46:00)
BugCheck: [SYSTEM_SERVICE_EXCEPTION (3B)]
Probably caused by: memory_corruption (Process: Discord.exe)
Uptime: 1 Day(s), 1 Hour(s), 32 Min(s), and 35 Sec(s)

File: 012125-78937-01.dmp (Jan 21 2025 - 19:09:18)
BugCheck: [IRQL_NOT_LESS_OR_EQUAL (A)]
Probably caused by: memory_corruption (Process: System)
Uptime: 0 Day(s), 16 Hour(s), 51 Min(s), and 59 Sec(s)

File: 012125-76890-01.dmp (Jan 22 2025 - 08:06:54)
BugCheck: [IRQL_NOT_LESS_OR_EQUAL (A)]
Probably caused by: memory_corruption (Process: System)
Uptime: 0 Day(s), 12 Hour(s), 32 Min(s), and 48 Sec(s)

File: 012125-16218-01.dmp (Jan 22 2025 - 10:12:40)
BugCheck: [IRQL_NOT_LESS_OR_EQUAL (A)]
Probably caused by: memory_corruption (Process: qbittorrent.ex)
Uptime: 0 Day(s), 2 Hour(s), 04 Min(s), and 14 Sec(s)

File: 012025-63890-01.dmp (Jan 21 2025 - 02:11:17)
BugCheck: [IRQL_NOT_LESS_OR_EQUAL (A)]
Probably caused by: memory_corruption (Process: System)
Uptime: 0 Day(s), 1 Hour(s), 07 Min(s), and 24 Sec(s)

top one did mention directX and video scheduling, but IRQ errors were as obscure as ever.

90% of time a BSOD mentions DirectX it is GPU related, I will give you that. But its not the only BSOD you have got. Its just the latest

But one fast answer is to run DDU in safe mode, uninstall current GPU drivers, boot back into normal mode and install latest Nvidia drivers

How To - How to perform a CLEAN install of your video card drivers.

Updated 9-19-19 How to do a CLEAN install of your graphics card drivers using the Wagnard tools Display Driver Uninstaller A major issue on many systems is the presence of multiple graphics card drivers or drivers that have been incompletely uninstalled followed by the installation of...

forums.tomshardware.com

Until you clean install on the Sata drive, I don't think Ubuysa will help as he told you to clean install in the last thread. If you going to ignore peoples suggestions after they helped you for ages, the problem isn't going to be fixed. And they will just stop trying

 

[ConorDuey2000]

Buy the sata cable and install windows on the ssd.

If you still get BSOD on the ssd, we will look at those. But we have spent 4 years trying to fix one install. I highly doubt your problem is just a graphics one.

I can try to find or reuse a SATA cable that I have. Once I install Windows onto the spare SSD and I leave it on overnight, I'll tell you if I got a BSOD or not. For now, I'm gonna focus on reinstalling DDU through the method that you suggested.

Until you clean install on the Sata drive, I don't think Ubuysa will help as he told you to clean install in the last thread. If you going to ignore peoples suggestions after they helped you for ages, the problem isn't going to be fixed. And they will just stop trying

I mostly didn't ignore those people and I was really happy to take their advice. I only decided not to take their advice on reformatting my computer or buying a new computer or expensive computer parts.

 

[Colif]

I only decided not to take their advice on reformatting my computer

Reinstalling windows can clean it out and forces you to get all new drivers. Since many of yours are old, and have been pointed at as possible causes, it would clear those out and let us see if the problem is windows or not.

Reinstalling windows is only a pain if you don't have copies of all the files you can't afford to lose, and really as probably pointed out above, you always need backups anyway. If you know you have to do it, you can make sure to have copies of anything you want to keep and not download again.

How To - Windows 11 clean install tutorial

Windows 11 Clean install tutorial This tutorial is intended to help you, step by step, to perform a clean install of Windows 11, version 23H2 or newer but can be equally applied to earlier versions which may have some basic differences in the screens you see but will largely remain similar in...

forums.tomshardware.com

Its really easy. It isn't a huge task that needs outside help, and its relatively painless if you know you have copies of everything already.

So if the sata drive doesn't have any errors, you should wipe it after the testing period (are bsod that common that only one night is enough?) and put it in PC as 2nd drive, and put anything you don't want to redownload again, onto the ssd.

A long time ago I had a run of bad hdd and kept needing to reinstall windows. This taught me a lesson, have more than one drive in PC and put everything you don't want to remake again on the 2nd drive.
That speeds up the reinstall process as you already have all your info and just point the Windows library folders at the location on other drive.

With cloud servers and Larger USB drives, the need to have two drives is diminishing for me. But I still like having more than one drive. Even if there really isn't anything essential on the 2nd drive now.

You can get pretty big USB flash drives now. I have a 256gb one myself.
When I made this PC I made sure all the files I needed were saved on Onedrive.
So all I needed to do was logon and it copied them all across.

 

[ConorDuey2000]

Reinstalling windows can clean it out and forces you to get all new drivers. Since many of yours are old, and have been pointed at as possible causes, it would clear those out and let us see if the problem is windows or not.

Reinstalling windows is only a pain if you don't have copies of all the files you can't afford to lose, and really as probably pointed out above, you always need backups anyway. If you know you have to do it, you can make sure to have copies of anything you want to keep and not download again.

How To - Windows 11 clean install tutorial

Windows 11 Clean install tutorial This tutorial is intended to help you, step by step, to perform a clean install of Windows 11, version 23H2 or newer but can be equally applied to earlier versions which may have some basic differences in the screens you see but will largely remain similar in...

forums.tomshardware.com

Its really easy. It isn't a huge task that needs outside help, and its relatively painless if you know you have copies of everything already.

So if the sata drive doesn't have any errors, you should wipe it after the testing period (are bsod that common that only one night is enough?) and put it in PC as 2nd drive, and put anything you don't want to redownload again, onto the ssd.

A long time ago I had a run of bad hdd and kept needing to reinstall windows. This taught me a lesson, have more than one drive in PC and put everything you don't want to remake again on the 2nd drive.
That speeds up the reinstall process as you already have all your info and just point the Windows library folders at the location on other drive.

With cloud servers and Larger USB drives, the need to have two drives is diminishing for me. But I still like having more than one drive. Even if there really isn't anything essential on the 2nd drive now.

You can get pretty big USB flash drives now. I have a 256gb one myself.
When I made this PC I made sure all the files I needed were saved on Onedrive.
So all I needed to do was logon and it copied them all across.

I already have a second drive. One 10TB internal HDD, one 5TB external SSD, and one 6TB external SSD. That's where all of my miscellaneous files are. My C drive contains all of the programs and settings that I have on my computer which I'm unable to back up regardless if I have enough storage on my other hard drives or not. If I were to reinstall Windows, I'll lose all of my programs and settings and it'll take forever for me to reinstall all of my programs and reconfigure all of my settings.

 

[ZivDero]

He has had problem for about 3 years, in that time every single person who understands BSOD on this forum has tried to solve it. I did for a few years, then Ubuysa took over... others have poked heads in along way.

I am fairly sure in that time I would have suggested memtest, its one of the obvious tests for BSOD. Reason I can see for it not being ram is, memory errors are normally a lot more random and generally point at drivers, not the hypervisor.

I think Others would have been able to tell if it was ram or not from the dump files.

Mobo: ASUS TUF GAMING X570-PLUS (WI-FI)
Mobo BIOS Version: American Megatrends Inc. 5021
CPU: AMD Ryzen 7 3800X 8-Core CPU
GPU: Nvidia GeForce RTX 2060
OS: Windows 11 Pro 24H2

Prime95 crashing could mean a number of things, could be PSU... he says he has an EVGA but I don't know if he ever looked at label to see which one. (mentioned in last thread)
could be ram not getting enough power.
could be CPU
could even be heat as it depends on what cooler he has, Prime will push CPU to max temp pretty quick.

It creates a report in the same folder it runs from, but it may not have recorded anything if it crashes right away.

Its not always the hypervisor but its been a reoccurring BSOD, it seems to come and go. BSOD generally don't fix themselves.

In a previous thread I did suggest it could be the CPU, but the testing a new install is one way to see if its not just a combo of old drivers causing the errors.

The reason I am inclined to blame the CPU is, in fact, the randomness of the BSODs. If the problem is that the CPU is failing and as a result is occasionally doing wronful operations (like what I got in my Prime95 test, one of the workers consistently being stopped due to an error), it would explain why Windows keeps detecting kernel-level errors with seemingly random reasons and either pointing at the kernel itself as the source, or at just at random drivers and applications. Because the error itself would be the CPU wrongfully changing some state, which then trips the bugcheck.

When I myself had this issue, I also had IRQL, QMODE, HYPERVISOR, etc. as the errors, and all of them pointed just to ntos kernel (sometimes list a random application, too). And running verifier just kept blaming all the drivers, one after the other as I uninstalled them. Swapping out the CPU has solved the issue.

 

[Colif]

When I myself had this issue, I also had IRQL, QMODE, HYPERVISOR, etc. as the errors, and all of them pointed just to ntos kernel (sometimes list a random application, too). And running verifier just kept blaming all the drivers, one after the other as I uninstalled them. Swapping out the CPU has solved the issue.

ever moving errors also can be explained by bad memory. But as the memory controller is on the CPU, I can see why it would be to blame as well.

@ConorDuey2000
Did you ever run memtest?
I see it mentioned in threads but was it one of the things you never did? In a previous one Ubuysa was saying they looked like memory errors. That was last post in thread so no idea if you just ignored it again.

its created on a USB drive, you boot from it and it checks ram. It doesn't do anything else.

Try running memtest86 on each of your ram sticks, one stick at a time, up to 4 passes. Only error count you want is 0, any higher could be cause of the BSOD. Remove/replace ram sticks with errors. Memtest is created as a bootable USB so that you don’t need windows to run it

Hardware errors won't fix themselves.

 

[ConorDuey2000]

ever moving errors also can be explained by bad memory. But as the memory controller is on the CPU, I can see why it would be to blame as well.

@ConorDuey2000
Did you ever run memtest?
I see it mentioned in threads but was it one of the things you never did? In a previous one Ubuysa was saying they looked like memory errors. That was last post in thread so no idea if you just ignored it again.

its created on a USB drive, you boot from it and it checks ram. It doesn't do anything else.

Try running memtest86 on each of your ram sticks, one stick at a time, up to 4 passes. Only error count you want is 0, any higher could be cause of the BSOD. Remove/replace ram sticks with errors. Memtest is created as a bootable USB so that you don’t need windows to run it

Hardware errors won't fix themselves.

If my computer gets a BSOD just before I go to sleep, I'll run memtest86 on it and leave it on overnight. The reason why I haven't ran it, yet, is because my RAM isn't old. I bought it about a year ago.

 

[Colif]

That probably means its not ram, since you had BSOD before, but it doesn't hurt to check. Age of ram doesn't really have any meaning, ram can be bad from start.
Yours are too consistent I think for them to be ram.

I already have a second drive. One 10TB internal HDD, one 5TB external SSD, and one 6TB external SSD. That's where all of my miscellaneous files are. My C drive contains all of the programs and settings that I have on my computer which I'm unable to back up regardless if I have enough storage on my other hard drives or not. If I were to reinstall Windows, I'll lose all of my programs and settings and it'll take forever for me to reinstall all of my programs and reconfigure all of my settings.

you could clone the drive onto another one so you have a copy of your install. 2nd drive needs to be bigger than first. The function just isn't built into windows.
https://www.easeus.com/partition-master/how-to-clone-a-hard-drive-on-windows-11.html

But I wouldn't clone an install that gets BSOD. Not if you don't get them on the ssd.

Next PC you will have to reinstall them anyway as you can't take an install from one PC to another. Most of the time if you try you will get errors. The hardware is normally too different. You can take data but generally not applications.

 

[ConorDuey2000]

Something bizarre just happened. I was using my computer as normal and suddenly, the computer completely froze and I had to hold down my computer's power button to turn it off. When I turned it back on, I noticed that I didn't get any memory dumps. Event Viewer showed absolutely nothing and neither did Task Scheduler. I'm pretty sure that my computer is self aware.

I'd like to install Windows on my blank SATA SSD and leave it on overnight, but first, I'd like to find some other fixes. By the way, I'm curious. Can I create a bootable Windows 11 USB and leave it running overnight instead of installing Windows 11 onto my SATA SSD?

 

[ConorDuey2000]

I have an update. Two days ago, I downloaded Display Driver Uninstaller and followed the appropriate tutorial on how to use it and reinstall my display driver. Unfortunately, this didn't stop my computer from having those BSODs. Two days after I reinstalled my display driver, I got an IRQL_NOT_LESS_OR_EQUAL BSOD while I was using my computer normally. Although reinstalling my display driver might've decreased the amount of BSODs I've gotten, I still get them. Here is the kernel dump and here is the minidump.

 

[Colif]

I'd like to install Windows on my blank SATA SSD and leave it on overnight, but first, I'd like to find some other fixes.

Just humour us and install it on the ssd.

Like what? We been trying to fix this install for Four years now. Pretty sure, apart from the reinstall, we have tried everything.

Have you left PC on in safe mode overnight? If you BSOD in safe mode, its not going to be drivers.

 

[ConorDuey2000]

Just humour us and install it on the ssd.

Like what? We been trying to fix this install for Four years now. Pretty sure, apart from the reinstall, we have tried everything.

Have you left PC on in safe mode overnight? If you BSOD in safe mode, its not going to be drivers.

I could try leaving my computer on in safe mode overnight. In the meantime, people can try finding other possible ways to stop the BSODs. They can try looking through my BSOD dumps and the ZIP files that Sysnative app generates.

 

[Colif]

How did safe mode go? Do you BSOD every day or just some?

I haven't had to convert one in a few years so some of the sites I use don't work like they used to


File: 013125-276406-01.dmp (Feb 1 2025 - 16:45:04)
BugCheck: [IRQL_NOT_LESS_OR_EQUAL (A)]
Probably caused by: memory_corruption (Process: System)
Uptime: 2 Day(s), 3 Hour(s), 44 Min(s), and 10 Sec(s)


IRQ errors were never my favorites anyway. Rarely ever point at a driver name.

I can't see much in them anymore, its why I let Ubuysa and Johnbl answer any.

They probably both told you to reinstall windows already. Ignoring advice just makes the people offering it IGNORE you. So... if you want to actually fix this and not be back in a few months again asking US to waste our time, you will follow advice.

No one here has to help you, we do it because we want to but we can only go so far.
You don't want to get PC repaired by a shop
And you are reluctant to install windows (which is really easy btw)

If you don't help as well, we can't.

 

[ConorDuey2000]

I have an update. I put my computer in safe mode and left it on safe mode overnight. When I woke up the next morning, I saw that my computer got a BSOD overnight while it was in safe mode. It was just another IQRL_NOT_LESS_OR_EQUAL BSOD caused by the process "System". This should mean that my BSOD problem isn't a driver problem. Here is the kernel dump and here is the minidump.

 

[ubuysa]

The dump from that Safe Mode BSOD suggests that the failure happened during the Windows processor power management (PPM) module....

8: kd> .frame /r 3
03 ffff8085`18509750 fffff805`81821d72     nt!PpmUpdatePerformanceFeedback+0x3b4
rax=0000000000000011 rbx=ffff8b80ec12ca10 rcx=0000000000000000
rdx=0000000000000082 rsi=ffff8b80ec124180 rdi=ffff8b80ec124100
rip=fffff80581620af4 rsp=ffff808518509750 rbp=0000000000000000
 r8=0000000000000000  r9=0000000000000000 r10=0000000000000001
r11=0000000000000000 r12=0000000000000000 r13=0000000000000000
r14=0000000000000000 r15=0000000000000000
iopl=0         nv up di ng nz na pe nc
cs=0010  ss=0018  ds=002b  es=002b  fs=0053  gs=002b             efl=00040082
nt!PpmUpdatePerformanceFeedback+0x3b4:
fffff805`81620af4 40386a20        cmp     byte ptr [rdx+20h],bpl ds:002b:00000000`000000a2=??

You can see there that the CMP (compare) instruction failed because the RDX register (used as a memory pointer) was clearly invalid. The ?? indicates that the memory referenced (via RDX) was invalid.

This could well be bad RAM so running Memtes86 (as advised already) is important...

1. Download Memtest86 (free), use the imageUSB.exe tool extracted from the download to make a bootable USB drive containing Memtest86 (1GB is plenty big enough). Do this on a different PC if you can, because you can't fully trust yours at the moment.
2. Then boot that USB drive on your PC, Memtest86 will start running as soon as it boots.
3. If no errors have been found after the four iterations of the 13 different tests that the free version does, then restart Memtest86 and do another four iterations. Even a single bit error is a failure.

However, I have seen this type of failure before, and mostly with AMD CPUs. It appears that some CPUs become tardy in transitioning from the low-power idle state to the high-power running state.

The test for this, and the mitigation against it, is to disable C-States for all processors in the BIOS. This prevents the processors from entering the low-power state when they become idle. The power transition problem then disappears. The only negative aspect of this is that the CPU runs a tad hotter, but not much, and the idle power drain is a tad higher, but again, not by much.

If the RAM test finds errors then that is the problem, but if Memtst86 runs clean (for two consecutive runs) then try disabling C-States.

 

[ConorDuey2000]

The dump from that Safe Mode BSOD suggests that the failure happened during the Windows processor power management (PPM) module....

8: kd> .frame /r 3
03 ffff8085`18509750 fffff805`81821d72     nt!PpmUpdatePerformanceFeedback+0x3b4
rax=0000000000000011 rbx=ffff8b80ec12ca10 rcx=0000000000000000
rdx=0000000000000082 rsi=ffff8b80ec124180 rdi=ffff8b80ec124100
rip=fffff80581620af4 rsp=ffff808518509750 rbp=0000000000000000
 r8=0000000000000000  r9=0000000000000000 r10=0000000000000001
r11=0000000000000000 r12=0000000000000000 r13=0000000000000000
r14=0000000000000000 r15=0000000000000000
iopl=0         nv up di ng nz na pe nc
cs=0010  ss=0018  ds=002b  es=002b  fs=0053  gs=002b             efl=00040082
nt!PpmUpdatePerformanceFeedback+0x3b4:
fffff805`81620af4 40386a20        cmp     byte ptr [rdx+20h],bpl ds:002b:00000000`000000a2=??

You can see there that the CMP (compare) instruction failed because the RDX register (used as a memory pointer) was clearly invalid. The ?? indicates that the memory referenced (via RDX) was invalid.

This could well be bad RAM so running Memtes86 (as advised already) is important...

1. Download Memtest86 (free), use the imageUSB.exe tool extracted from the download to make a bootable USB drive containing Memtest86 (1GB is plenty big enough). Do this on a different PC if you can, because you can't fully trust yours at the moment.
2. Then boot that USB drive on your PC, Memtest86 will start running as soon as it boots.
3. If no errors have been found after the four iterations of the 13 different tests that the free version does, then restart Memtest86 and do another four iterations. Even a single bit error is a failure.

However, I have seen this type of failure before, and mostly with AMD CPUs. It appears that some CPUs become tardy in transitioning from the low-power idle state to the high-power running state.

The test for this, and the mitigation against it, is to disable C-States for all processors in the BIOS. This prevents the processors from entering the low-power state when they become idle. The power transition problem then disappears. The only negative aspect of this is that the CPU runs a tad hotter, but not much, and the idle power drain is a tad higher, but again, not by much.

If the RAM test finds errors then that is the problem, but if Memtst86 runs clean (for two consecutive runs) then try disabling C-States.

I don't think that my RAM would have any problems since it's only a little over a year old, but I'll run Memtest86, anyway, if it's possible to leave it running overnight as I sleep. Anyway, I disabled "Global C-state Control" in the BIOS. If I still continue to get BSODs, I'll set it back to "Auto" and run Memtest86 overnight.

 

[ConorDuey2000]

I don't think that my RAM would have any problems since it's only a little over a year old, but I'll run Memtest86, anyway, if it's possible to leave it running overnight as I sleep. Anyway, I disabled "Global C-state Control" in the BIOS. If I still continue to get BSODs, I'll set it back to "Auto" and run Memtest86 overnight.

It turns out that disabling "Global C-state Control" did nothing, so I switched it back to "Auto". I have an "ASUS TUF GAMING X570-PLUS (WI-FI)" motherboard. What setting(s) should I change to disable C-states?