Question 5-10 BSODs per hour - ntorskrnl.exe+3f71b0

[forsurexd]

Hello guys!

currently losing my mind while trying to figure out what the <Mod Edit> went wrong with my computer. You guys are pretty much my last resort.
After this there is only one option left and thats buying a new pc. I think i tried every single idea that i found somewhere in the internet but nothing really helped. Honestly i might have made things worse.

I have an absolute insane amount of BSODs to the point where my PC is literally unusable. But before i get into what i tried and what i think could be potential issues let me start with my system configuration:

CPU: Ryzen 5 3600
GPU: Sapphire AMD 5700 XT Nitro+
RAM: 4x 8GB Corsair Vengance RGB Pro
SSD: Kingston SA2000M8500G 500GB
PSU: ill update that during the day as i have to open my case for that but its a silver Corsair one / voltages seem fine tho / gonna have a friend over at the weekend who will check with multimeter
Motherboard: MSI B450M PRO VDH Max
OS: now Win10 updated to the maximum - was on Win11 for the most part of this tho


So a few weeks (?) might honestly be months ago ive started to experience more and more bluescreens with pretty much always the same error code.

DPC Watchdog Violation or IRQL not less or equal i think its these two 90% of the time. Since ive downgraded back to Windows 10 a few hours ago the error codes have become a little bit more versatile.

All of the BSODs are caused by ntorskrnl.exe+3f71b0.

Ive browsed all of the known forums and holy there is a lot of people with almost identical issues and almost all of them dont really have any idea what the problem is.

Let me list a few of the things i tried to fix these issues:

• ive updated the BIOS to the most recent version
• same for my GPU drivers which i always un/installed using DDU in safe mode ( using AMD WQHL drivers at the moment )
• ive tried to fix my issues by enabling / disabling XMP
• run memtest several times without any errors
• malwarebytes did not report any issues
• all drivers are updated
• ram was reseated and pc has been cleaned properly
• voltages of CPU seem to be in a good state
• did run cmd SFC /scannow and chdksk/r

So what else is there to do? Im totally hopeless at this point. Im having like 5-10 BSODs per hour. Would love to change any broken hardware but i dont even know what it could be?


Added minidumps: https://drive.google.com/drive/folders/1htJ-lwY_B-jBCStglCtFfA1xgWiA8syp


A few observations and thoughts after dealing with that <Mod Edit> for months now:

• it happens ALOT while using any browser ( tried chromium and non chromium based, same <Mod Edit> )
• sometimes while using the browser the tab crashes ( also in all browsers ) this feels like the error was not "big" enough to make it a BSOD (?)
• on my old windows 11 built driver verifier said valorant and faceit anticheat where the drivers that were causing problems
• the PC did indeed bluescreen without having any non windows drivers installed completly fresh installed windows crashed in the first 2 minutes with a BSOD
• tried all browsers without extensions = still the same issues
• could it be a wireless Logitech Mouse? I dont know why i even thought about it but the start of my BSOD journey somewhat allignes with the timeframe where i got the mouse ( Logitech GHub not installed currently )

If you guys need any more information any screenshots or more minidumps just let me know just while writing this there were at least 3 new BSODs.

Thank you

PS: this PC ran without ANY issues at all for more than 12 months before the BSODs started happening.

 

[Colif]

on my old windows 11 built driver verifier said valorant and faceit anticheat where the drivers that were causing problems

Vanguard does think it owns the PC


Conversion of dumps

report - Click run as fiddle to see report

Spoiler

File: 091422-6328-01.dmp (Sep 14 2022 - 15:54:01)
BugCheck: [DRIVER_OVERRAN_STACK_BUFFER (F7)]
Probably caused by: ntkrnlmp.exe (Process: System)
Uptime: 0 Day(s), 0 Hour(s), 02 Min(s), and 46 Sec(s)

File: 091422-6312-01.dmp (Sep 14 2022 - 15:57:33)
BugCheck: [IRQL_NOT_LESS_OR_EQUAL (A)]
Probably caused by: ntkrnlmp.exe (Process: System)
Uptime: 0 Day(s), 0 Hour(s), 02 Min(s), and 38 Sec(s)

File: 091422-5828-01.dmp (Sep 14 2022 - 16:38:57)
BugCheck: [CLOCK_WATCHDOG_TIMEOUT (101)]
Probably caused by: Unknown_Image (Process: System)
Uptime: 0 Day(s), 0 Hour(s), 11 Min(s), and 59 Sec(s)

File: 091422-5687-01.dmp (Sep 14 2022 - 15:50:51)
BugCheck: [INVALID_PROCESS_DETACH_ATTEMPT (6)]
Probably caused by: ntkrnlmp.exe (Process: opera.exe)
Uptime: 0 Day(s), 0 Hour(s), 03 Min(s), and 36 Sec(s)

File: 091422-5656-01.dmp (Sep 14 2022 - 15:46:49)
BugCheck: [APC_INDEX_MISMATCH (1)]
Probably caused by: ntkrnlmp.exe (Process: opera.exe)
Uptime: 0 Day(s), 0 Hour(s), 22 Min(s), and 41 Sec(s)

Processes named are victims

Can't say I seen Invalid Process Detach Attempt before (there are over 300 codes, there are a lot of irregular ones), MIcrosofts website says its an infrequent error. I have to agree.
I would update chipset drivers, that might help with the errors - https://www.amd.com/en/support/chipsets/amd-socket-am4/b450

One test you didn't run
Try running this overnight - https://www.guru3d.com/files-details/prime95-download.html
Prime 95 how to Guide: http://www.playtool.com/pages/prime95/prime95.html

 

[Nine Layer Nige]

Two things to try, which you did not mention.

1. Try only 2 sticks of ram.
2. Get a new SSD and try a fresh install with the OS only, nothing else installed to see if its stable.

 

[forsurexd]

Alright gonna head to bed now and gonna update in 8-10 hours how the Prime 95 test went.

Chipset drivers are updated.

Also definitely trying to only use 2 sticks of ram later!

Dont have a second SSD here at the moment, can I use a HDD? Shouldnt be a problem right?

 

[Colif]

wish i had noticed you using 2 sets of ram

8192MB	3200MHz	Unknown	CMW16GX4M2C3200C16
8192MB	3200MHz	Unknown	CMW16GX4M2C3200C16
8192MB	3200MHz	Unknown	CMW16GX4M2C3200C16
8192MB	3200MHz	Unknown	CMW16GX4M2C3200C16

https://www.corsair.com/ww/en/Categ.../Vengeance-PRO-RGB-Black/p/CMW16GX4M2C3200C16

thats all it takes

Just run Prime with 1 set in

 

[forsurexd]

Could that be the issue? Ive upgraded from 16GB RAM to 32GB RAM in December last year.


Ive made sure to buy the same exact RAM again.

Remember when i said me buying a Wireless Mouse roughly alligns with the start of my BSOD issues? The RAM Upgrade was in the same amazon order

 

[Colif]

Sticks in sets only tested to work with other sticks in that set. Adding extra ram to one set that isn't tested can lead to errors. Adding 2 sticks can make it worse. Even if they exactly the same sticks.
There is a reason ram is sold in sets.

If you want 32gb on a Ryzen CPU, safest option is buy 1 set

wish it wasn't the case, I have 2 x 16gb of the same ram as you do, I wouldn't mind 4 sticks but can't afford the cost really. And can't add 2 more sticks without getting same errors as we have same CPU as well.

 

[forsurexd]

Wow so many browsed threads and not a single mention of that.
Prime 95 didnt crash once. All workers were still at work. No errors.

So I guess that was my problem and you guys fixed it.

What am I doing with my other 2 Sticks of RAM now? Ive read a few times while browsing different forums before bed that there is ways to maybe still make it work?
Should I go for better timings and lower MHz? Or do u think that this overall will not really improve my system and especially system stability?


Anyways, thank you guys so much. I really appreciate the time you guys put into this

 

[forsurexd]

Well its apparently not fixed. Its way more stable but for whatever reason its still crashing.

Ive updated the google drive link i posted earlier with two new minidumps: https://drive.google.com/drive/folders/1htJ-lwY_B-jBCStglCtFfA1xgWiA8syp

The ones in the folder "minidump new" are the new ones. Maybe someone could take a look at it.

 

[Colif]

sure the sticks you removed were from a set?
2 dumps sure beats last person I did who had 50
Conversion of dumps

report - Click run as fiddle to see report

Spoiler

File: 091422-5546-01.dmp (Sep 15 2022 - 04:56:28)
BugCheck: [SYSTEM_SERVICE_EXCEPTION (3B)]
Probably caused by: memory_corruption (Process: opera.exe)
Uptime: 0 Day(s), 0 Hour(s), 03 Min(s), and 47 Sec(s)

File: 091422-5281-01.dmp (Sep 15 2022 - 05:08:08)
BugCheck: [SYSTEM_THREAD_EXCEPTION_NOT_HANDLED_M (1000007E)]
Probably caused by: memory_corruption (Process: conhost.exe)
Uptime: 0 Day(s), 0 Hour(s), 07 Min(s), and 48 Sec(s)

try updating gpu drivers

conhost.exe is command prompt.

 

[forsurexd]

Im fairly certain yeah...
Have one pair with serial numbers ending with 948 and 949 in front of me right now. The other one had 50 and 51 so i assume the serial numbers are always +1 on the paired sticks?

Ill update my gpu drivers again sure, Do u think i should go for the PRO drivers, WHQL drivers or just the most recent version?

 

[Colif]

Have one pair with serial numbers ending with 948 and 949 in front of me right now. The other one had 50 and 51 so i assume the serial numbers are always +1 on the paired sticks?

i would think so. It would make sense to me.

I didn't know AMD had so many choices for drivers?
Pro appear to be for work related things & graphics whereas adrenaline are for gaming. So which do you do more of?
I not sure what 3rd are unless they are the ones installed by windows update.

 

[forsurexd]

Afaik WHQL is the recommended driver, the "safest" one.
Pro is as you said more for content creation and the most recent one is kinda just the driver updated for new games.
Im just gonna go with WHQL one again i think. Doing that now and reporting back after.

This whole thing is so confusing to me tho, Like how is my pc so stable while gaming, while running very CPU and RAM intensive programs but as soon as i open the browser everything goes downhill.

 

[Colif]

keep giving me dump files, maybe not one at a time, but if you get any.
Is it still crashing as often?

PSU: ill update that during the day as i have to open my case for that but its a silver Corsair one / voltages seem fine tho / gonna have a friend over at the weekend who will check with multimeter

It could be this. Friend might show more. Nice to know voltages at idle.

Some PSU have problems with how Ryzen CPU act at idle. What normally happens is if a process needs to be run, the CPU will wake a core, boost to speed it needs to run process and rushes to idle again. The constant up/down in voltage can be a problem for some PSU. They don't have problems with load though as its consistent.

its a possibility. Just throwing it out there.

I know you ran SFC but try this as well:

right click start button

choose powershell (admin)

copy/paste this command into window:

Repair-WindowsImage -Online -RestoreHealth

and press enter

Then type SFC /scannow

and press enter

Restart PC if SFC fixes any files as some fixes require a restart to be implemented

First command repairs the files SFC uses to clean files, and SFC fixes system files

SFC = System File Checker. First command runs DISM - https://docs.microsoft.com/en-us/windows-hardware/manufacture/desktop/what-is-dism?view=windows-11

tempted to see if other set of ram makes any difference.

possibly newer BIOS, you appear to be on a beta (way MSI do their bios number system, it can be hard to tell)
https://www.msi.com/Motherboard/B450M-PRO-VDH-MAX/support

ram is slots A2 & B2? sometimes using wrong slots can cause errors, just checking - https://download.msi.com/archive/mnu_exe/mb/M7A38v8.1.pdf

I have to go to sleep soon.

 

[forsurexd]

Yeah its still crashing. Here is one new dumpfile: https://drive.google.com/drive/folders/1htJ-lwY_B-jBCStglCtFfA1xgWiA8syp

I did run the two commands u sent me and there were no issues found.

Im currently using A1 B1 but I was using A2 B2 till like an hour ago.

Just wanted to make sure that the issues are not RAM slot related.

I know that A1 B1 performance can be worse i just wanted to try it to make sure.

New drivers are installed, did use DDU.

I have one more dumpfile with a new errorcode but im unable to upload it at the moment. Google drive tells me its not able to read the file. Trying to fix that rn.

 

[Colif]

I expect you get more errors using wrong slots. Some boards are wired in such a way they only supposed to use the last 2 slots if other ram is already in 1st 2.

Meant to be bad weather here next few hours so I don't know when i be back. Will convert dumps when I can.

You have had some weird errors so far
Prime 95 ran and found no errors? So shouldn't be CPU
ram passed memtest?
PSU being tested on weekend - at least you have a friend with a multimeter
try running crystaldiskinfo (blue icons) - https://crystalmark.info/en/ - no need to do anything, it will read what ssd tells it.

 

[forsurexd]

Crystal Disk Info Screenshots:

https://drive.google.com/drive/folders/1htJ-lwY_B-jBCStglCtFfA1xgWiA8syp?usp=sharing

+ quite a few new dumpfiles: https://1drv.ms/u/s!Ag-U7on1ZMw9iUcdwrpIvnLPyfVy?e=edDKbW

Prime 95 ran and found no errors, correct! Same for Memtest.

Currently debating to try out my old PSU. Did not change it because it had any issues, just bought a new one for my new PC a few months ago. So old one is still working and could just try that one out.
Just no idea what it could be, do u think its worth trying out different GPU Slot? Im using the one closest to the CPU but could try the other one...
Will also revert back to the right RAM slot.

 

[Colif]

11 isn't as bad as 50
11 beats downloading one 3gb dump file to get 1 dump result.
funny thing is 1 dump in a 3gb file is less work than 11 dumps were for the program

Conversion of dumps

report - Click run as fiddle to see report

Spoiler

File: 091522-8500-01.dmp (Sep 15 2022 - 11:34:43)
BugCheck: [SYSTEM_THREAD_EXCEPTION_NOT_HANDLED_M (1000007E)]
Probably caused by: HIDCLASS.SYS (Process: System)
Uptime: 0 Day(s), 0 Hour(s), 00 Min(s), and 54 Sec(s)

File: 091522-8093-01.dmp (Sep 15 2022 - 11:36:22)
BugCheck: [SYSTEM_THREAD_EXCEPTION_NOT_HANDLED_M (1000007E)]
Probably caused by: HIDCLASS.SYS (Process: System)
Uptime: 0 Day(s), 0 Hour(s), 01 Min(s), and 04 Sec(s)

File: 091522-7593-01.dmp (Sep 15 2022 - 11:30:23)
BugCheck: [SYSTEM_THREAD_EXCEPTION_NOT_HANDLED_M (1000007E)]
Probably caused by: HIDCLASS.SYS (Process: System)
Uptime: 0 Day(s), 0 Hour(s), 00 Min(s), and 49 Sec(s)

File: 091522-5531-01.dmp (Sep 15 2022 - 12:21:28)
BugCheck: [APC_INDEX_MISMATCH (1)]
Probably caused by: ntkrnlmp.exe (Process: opera.exe)
Uptime: 0 Day(s), 0 Hour(s), 01 Min(s), and 51 Sec(s)

File: 091522-5328-01.dmp (Sep 15 2022 - 11:04:28)
BugCheck: [DPC_WATCHDOG_VIOLATION (133)]
Probably caused by: ntkrnlmp.exe (Process: opera.exe)
Uptime: 0 Day(s), 0 Hour(s), 10 Min(s), and 56 Sec(s)

File: 091522-5140-01.dmp (Sep 15 2022 - 12:11:36)
BugCheck: [DPC_WATCHDOG_VIOLATION (133)]
Probably caused by: ntkrnlmp.exe (Process: iCUE.exe)
Uptime: 0 Day(s), 0 Hour(s), 05 Min(s), and 31 Sec(s)

File: 091522-4968-01.dmp (Sep 15 2022 - 12:05:39)
BugCheck: [DPC_WATCHDOG_VIOLATION (133)]
Probably caused by: ntkrnlmp.exe (Process: System)
Uptime: 0 Day(s), 0 Hour(s), 13 Min(s), and 00 Sec(s)

File: 091522-4796-01.dmp (Sep 15 2022 - 11:52:12)
BugCheck: [SYSTEM_SERVICE_EXCEPTION (3B)]
Probably caused by: memory_corruption (Process: lghub_agent.ex)
Uptime: 0 Day(s), 0 Hour(s), 13 Min(s), and 04 Sec(s)

File: 091522-4546-01.dmp (Sep 15 2022 - 10:28:10)
BugCheck: [DPC_WATCHDOG_VIOLATION (133)]
Probably caused by: ks.sys (Process: System)
Uptime: 0 Day(s), 0 Hour(s), 14 Min(s), and 58 Sec(s)

File: 091422-5531-01.dmp (Sep 15 2022 - 07:40:56)
BugCheck: [DPC_WATCHDOG_VIOLATION (133)]
*** WARNING: Unable to verify timestamp for rt640x64.sys
Probably caused by: afd.sys (Process: System)
Uptime: 0 Day(s), 0 Hour(s), 21 Min(s), and 35 Sec(s)

File: 091422-5390-01.dmp (Sep 15 2022 - 07:05:06)
BugCheck: [DRIVER_PORTION_MUST_BE_NONPAGED (D3)]
Probably caused by: memory_corruption (Process: System)
Uptime: 0 Day(s), 0 Hour(s), 09 Min(s), and 05 Sec(s)

top 3 blame HID class - its used to communicate with USB devices, looks like a driver not responding was cause of error
Crash 9 mentions USBAudio so is that headphones or DAC?

one blames rt640x64.sys which is

Jun 15 2022

rt640x64.sys

Realtek NICDRV 8169 PCIe GBE Family Controller driver https://www.realtek.com/en/

The continuing feature - BSOD I never seen before - DRIVER_PORTION_MUST_BE_NONPAGED - new one to me.

Honestly, I would clean install win 11 again as if only problems you had were Valorant, thats easy to deal with compared to all these BSOD.

 

[forsurexd]

Nah the problems are everywhere, it was just driver verifier that screamed it would be the valorant anticheat. Im pretty sure that was just a false alarm tho.

Im gonna reinstall win10 one last time tonight and if that doesnt help i just have to buy a new one. I just have no clue what it could be and how to fix it. Spent the whole last four days trying to fix it but yeah didnt really help. Very frustrating.

 

[Colif]

The weird errors like Invalid process detach attempt & DRIVER_PORTION_MUST_BE_NONPAGED just make me think its ram as most of the time I see weird errors I don't normally see, its ram.

though the last one might be because of this:

Will also revert back to the right RAM slot.

Is ram is right slots now?

I would ask on here and see what they suggest: https://www.tenforums.com/bsod-crashes-debugging/

Otherwise I would pay someone to test hardware and see if they can identify it before I bought a new PC.

 

[forsurexd]

Yeah ram is in right slots again. Also tested my old sticks for a time and my two sticks bought in december. Doesnt really change anything. The crashes are happening on both.

Im just sadly out of Ideas. Gonna try to check the PSU with a friend on the weekend but after that yeah no more stuff to try kinda...

Thank you very much for all your help tho!

 

[Colif]

if you ask on tenforums, they should be able to zero in on cause.
Can you give me a link to your thread there as I watch it - I can't answer there though.
I am curious what answer is.

 

[forsurexd]

Hey Mate,
i have new RAM, new PSU and new SSD coming in today just to rule out more hardware parts as broken. If the problem still persits afterwards im gonna let you know and open a thread there.

 

[forsurexd]

Yeah sadly no good news tried every new part on its own to rule out broken hardware but the bluescreens just wont go away.
So its not the RAM, not the SSD and not the PSU. Have RMA available for my 5700XT till mid of October so thats probably gonna be next.
Maybe a friend has a old CPU that i can try before that to rule that part out aswell.

 

[Colif]

. If the problem still persits afterwards im gonna let you know and open a thread there.

I would ask on there before spending any more money.

I get this feeling they will blame the CPU and at this point, they might be right.
I only say that as most times I send people there they say its the CPU.

I don't like sending you elsewhere but I also don't like wasting peoples time/money