Question BSOD caused by ntoskrnl.exe + 4b7ce0 ?

[6_ERALDI]

hi all,

I've been getting numerous BSODs ever since I built this PC, and although numerous, they are really sporadic and rare (e.g. once every two or three days, after long hours of work).
Up to now, I've installed Windows like 3 or 4 times, with the same results. I haven't tried booting in Safe Mode, because as I said, these errors happen only after a long time of the system being on, so I figured no use trying Safe Mode.

PC Specs
Mobo: MSI MAG B550 Tomahawk
CPU: Ryzen 5 5600
GPU: MSI RX 570 8GB OC
RAM: G Skill something [2 x 8GB] 3600 Mhz (Currently running at ~2100 Mhz)
PSU: EVGA 430w
Boot Drive: 1TB Samsung 990 Pro NVMe SSD with heatsink
Other drives: 480GB SSD, 500GB HDD, 2TB HDD

No XMP, no overclocking whatsoever. Drivers are updated to the latest (AMD chipset, Realtek audio/network, AMD Adrenalin, etc.) and Windows is updated to the latest (just yesterday it got another update).

Dump files
1. https://www.mediafire.com/file/iskhnqghexovqka/031025-9187-01.dmp/file
2. https://www.mediafire.com/file/z6jeh6xpjn4dmfs/031225-10562-01.dmp/file
(suggest me another uploading method if this doesn't work)

Please help

 

[jdobrien]

I wish I could help you, but I started having this error this week. Same setup I assume, Windows 11 on AMD Ryzen etc. Same NVMe as well. As a matter of fact, as I was replying to this I got a BSOD.

I got tired of debugging so I opted on a clean installation of Windows 11. Unfortunately, it nearly immediately started with the BSODs and locking up, screen going black.

At first I thought it was related to all the virtualization I was doing since I use it as a development machine.

I have a hard time believing that we are having hardware related issues at the same time.

 

[evertonmdz]

INTERRUPT_EXCEPTION_NOT_HANDLED, and IRQL_NOT_LESS_OR_EQUAL, here.
Here I am also having this ntoskrnl.exe+4b7ce0 on the nvidia rtx 3080 ti gpu, I noticed that it only happens when you are in the game and press the windows button or alt tab, it has been like this for a few weeks now, but not always, I saw that another person is also having this ntoskrnl.exe+4b7ce0 - https://forums.tomshardware.com/threads/blue-screen-help.3875423/
however I am suspicious that it has to do with hypervsior, it mentions this a lot in the minidump, I will try to disable "memory integrity" core isolation. to see if it solves it, I will try to install the most stable nvidia driver from December too! 566.36

For those who know what I'm talking about, I'll leave the minidump too,
https://drive.google.com/drive/folders/1cmV84cTKNi06EwxFFlmpYjU7k8llUjy8?usp=drive_link

I noticed that it also happens after suddenly the HDR within the game becomes whitish, as if it had been disabled, which only works again after pressing alt tab and returning to the game again.

 

[jdobrien]

INTERRUPT_EXCEPTION_NOT_HANDLED, and IRQL_NOT_LESS_OR_EQUAL, here.
Here I am also having this ntoskrnl.exe+4b7ce0 on the nvidia rtx 3080 ti gpu, I noticed that it only happens when you are in the game and press the windows button or alt tab, it has been like this for a few weeks now, but not always, I saw that another person is also having this ntoskrnl.exe+4b7ce0 - https://forums.tomshardware.com/threads/blue-screen-help.3875423/
however I am suspicious that it has to do with hypervsior, it mentions this a lot in the minidump, I will try to disable "memory integrity" core isolation. to see if it solves it, I will try to install the most stable nvidia driver from December too! 566.36

For those who know what I'm talking about, I'll leave the minidump too,
https://drive.google.com/drive/folders/1cmV84cTKNi06EwxFFlmpYjU7k8llUjy8?usp=drive_link

I noticed that it also happens after suddenly the HDR within the game becomes whitish, as if it had been disabled, which only works again after pressing alt tab and returning to the game again.

I have a clean windows installation, no games nothing. BSOD with the same error.

 

[Dark Lord of Tech]

Hard to diagnose when the PSU is very poor quality.

 

[6_ERALDI]

Hard to diagnose when the PSU is very poor quality.

Assuming you're referring to my 430W 80+ white...

I would agree, but:
1. I've used this PSU with no problems with a somewhat similar build (in terms of power), with a R3 3100 and Gigabyte A320M (so, practically, I've recently bought a CPU, MB and NVMe drive, listed in the OP, and now I'm getting BSODs)
2. I don't get any errors during demanding tasks: I could be gaming with no probs for 4 hours, and after 2-3 hours of Netflix I'd get a BSOD...
Thus, before I go out to get a new PSU (with very little knowledge in regards to value/availability in my region these days) I would really like to exhaust every SW *probable* solutions.

however I am suspicious that it has to do with hypervsior, it mentions this a lot in the minidump, I will try to disable "memory integrity" core isolation. to see if it solves it,

I will do this too. Also, I've seen in the Event viewer a few occurrences of this:
The l1vhlwf service failed to start due to the following error:
A hypervisor feature is not available to the user.
Is it relevant?

: )

 

[evertonmdz]

Assuming you're referring to my 430W 80+ white...

I would agree, but:
1. I've used this PSU with no problems with a somewhat similar build (in terms of power), with a R3 3100 and Gigabyte A320M (so, practically, I've recently bought a CPU, MB and NVMe drive, listed in the OP, and now I'm getting BSODs)
2. I don't get any errors during demanding tasks: I could be gaming with no probs for 4 hours, and after 2-3 hours of Netflix I'd get a BSOD...
Thus, before I go out to get a new PSU (with very little knowledge in regards to value/availability in my region these days) I would really like to exhaust every SW *probable* solutions.

I will do this too. Also, I've seen in the Event viewer a few occurrences of this:
The l1vhlwf service failed to start due to the following error:
A hypervisor feature is not available to the user.
Is it relevant?

: )

Here it apparently solved, I didn't have it once today, I did both things mentioned, I'll try to enable the hypervisor again and continue the tests, because although it seems to be it, the latest nvidia drivers since December are Mega problematic, so it could be that too.

 

[Stranni15k]

Hello everybody! I also started experiencing this problem, it started quite recently, perhaps it was affected by the latest Windows 11 update, I used to use only Windows 10, but recently decided to start using 11, so I think the problem is global, I also work as a developer and virtualization is enabled because I have to deploy virtual machines, before up to this point, I have never experienced problems with the blue screen.
Specifications of my PC:
ASUS TUF GAMING 490 Plus Wi-fi
i9-10850k
RTX 3080 ASUS ROG
1000W power supply unit with Platinum certificate
NVME from Samsung 970 EVO PLUS

 

[6_ERALDI]

so, chances that this windows build is unstable: 📈📈📈
I will need to return to this in a couple of days, as my BSOD aren't really that frequent.

In the mean time, if anyone can take a look at those dmp files and let us know if they see anything, they're welcome to reply

 

[Stranni15k]

so, chances that this windows build is unstable: 📈📈📈
I will need to return to this in a couple of days, as my BSOD aren't really that frequent.

In the mean time, if anyone can take a look at those dmp files and let us know if they see anything, they're welcome to reply

This usually happens in my game, but I don't think it happened because of a hardware malfunction. Do I understand correctly that everyone here has Windows 11 installed?

 

[Stranni15k]

I had recovery points enabled on my PC, the earliest was on March 6, I wanted to try to roll back to it, but I got an error and an endless death screen when turned on, I managed to fix it by removing components from the system recovery menu, I think there is still a problem with an update that breaks the system core

 

[cubemodxd]

So i saw that something oc my ryzen 7 5800x to 1.55 volts what is max of my mainboard

Also try lowering voltage with ryzen.master

In case anyone of u have a fix let me know so this problem is a windows thing becauwe windows fking restarts the pc for me i had to read the bluescreen error out

 

[6_ERALDI]

So i saw that something oc my ryzen 7 5800x to 1.55 volts what is max of my mainboard

Also try lowering voltage with ryzen.master

In case anyone of u have a fix let me know so this problem is a windows thing becauwe windows fking restarts the pc for me i had to read the bluescreen error out

As far as I know, the general opinion (in this thread) is that the issue could be related to virtualization done by Windows Security -> Device Security -> Core isolation -> Memory Integrity.
Therefore, me myself have disabled this feature. I'm no expert whatsoever, I'm just trying every possible trick, so your mileage may vary

So far so good for me, but as I've previously said, I would need a few more days to confirm, as my BSOD weren't as frequent as the other fellows here.

Also try lowering voltage with ryzen.master

Again, no expert here, but I haven't heard good things about RyzenMaster that you are mentioning, so I suggest you do your own research about it.

 

[evertonmdz]

Here it apparently solved, I didn't have it once today, I did both things mentioned, I'll try to enable the hypervisor again and continue the tests, because although it seems to be it, the latest nvidia drivers since December are Mega problematic, so it could be that too.

Here I solved the BSOD by switching to the nvidia 566.12 drivers, but instead of giving BSOD now, it is simply closing the game (although less often), and generating the error "Persistent nvlddmkm Event id 153".

I saw that it may be related to Precision Boost Overdrive, decrease it from 25 to 20 (the undervolt) to see if it stabilizes. I found this thread about this event ID 153. https://www.reddit.com/r/PcBuildHel...istent_nvlddmkm_event_id_15313_errors_on_new/

 

[evertonmdz]

GeForce Game Ready Driver 572.83

Fixed General Bugs

• [GeForce RTX 5080/5090] Graphics cards may not run at full speeds on system reboot when overclocked [5088034]
• [GeForce RTX 50 series] GeForce RTX 50 series GPUs crashes with black screen [5120886]
• Some NVIDIA Control Panel / NVIDIA App settings changes immediately get reset or give error [5160516]
• PC may bugcheck IRQL NOT LESS OR EQUAL 0xa during gameplay with HDR enabled [5091576]
• [VRED 2026] Optix compile error with R570 branch drivers [5122360]
• [Derivative TouchDesigner ] Application stability issues [4606316]

 

[Squeezit]

I'm getting BSOD on a brand new 5090 GPU.

Nvidia's driver update did not resolve.
I never had the issue on my 40 series GPU though.

The cause seems to be this:

 

[johnbl]

looks like both bugcheck are due to csrss.exe terminating

I would suspect ovpn-dco.sys just because of the description:

OpenVPN Data Channel Offload (DCO) is a cutting-edge Linux kernel module designed to revolutionize the performance of VPN servers and clients. By implementing the data channel of the OpenVPN protocol directly within the Linux kernel, DCO enables high-speed data processing and efficient resource utilization.

(just a guess)
nothing should modify csrss.exe, it checks itself and will bugcheck the system if modified.
in both cases the system was running a long time before the bugcheck.
System Uptime: 10 days 2:21:00.213
System Uptime: 0 days 16:46:57.557

you might uninstall the software or try to update it. you might run verifier.exe flags to try to get the system to crash faster.

the client server runtime service does check itself for modifications and will crash pretty fast if modified.

problem is that it can be a bug in the driver, bios, chipset or in the cpu. just hard to tell. first step would be to update everything and see if you still bugcheck.


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

bugcheck was a critical process died
IMAGE_NAME: csrss.exe

this is the windows client server runtime service, it is a target of malware, if it detects corruption it will terminate its own process.

this is a new driver I have not seen before:

ovpn-dco.sys
I would be looking at this:
OpenVPN Data Channel Offload (DCO)
----------------
BiosVersion = A.H0
BiosReleaseDate = 07/23/2024
SystemManufacturer = Micro-Star International Co., Ltd.
SystemProductName = MS-7C91
Identifier = REG_SZ AMD64 Family 25 Model 33 Stepping 2
ProcessorNameString = REG_SZ AMD Ryzen 5 5600 6-Core Processor
Update Status = REG_DWORD 6
6: kd> !sysinfo cpumicrocode
~MHz = REG_DWORD 3500
Initial Microcode Version: 00000000:0a20120e
Cached Microcode Version: 00000000:00000000
Processor Family: 19
Processor Model: 21
Processor Stepping: 02
Note: Cached Microcode Version is not detected!

https://www.msi.com/Motherboard/MAG-B550-TOMAHAWK/support

beta bios (not installed) has a fix for sinkclose vulnerability
and AMD agesa updates

there are also some amd chipset updates that could help:
(dated 2025-01-22)
(8 months newer than what you have installed)

motherboard audio driver was updated today.
2025-05-06 people have been having strange audio problems lately