KMODE EXCEPTION NOT HANDLED BSOD after reinstalling windows 10

Hi

I have an old desktop that needed to be refreshed one last time before I save up for a new one. I attempted a clean install of Windows 10 and have been suffering the Kmode exception BSOD ever since. This can occur during installation, boot-up, after 5 seconds or half an hour but it always happens with no apparent trigger. As there is no additional information to the error code and because it occurs during installation I presume it is a hardware issue? I have tried most of the common fault finding operations for this error detailed below.

The system specs are an ASUS P8P67 Pro, (Bios 0402), with an i5-2500K, 16Gb of Ram and a GTX 970, 'running' Windows 10 Home 64-bit (10.0, Build 19045)


Things I've tried already:

Disabled Fast Startup

Installed all the drivers suggested by Snappy Driver installer

Run the memory diagnostic tool

Removed and swapped the memory slots over

Unplugged all cables and re-assembled

Installed windows on various hard drives (SSD and HDD)

I'm at a bit of a loss what to try next, I was thinking of trying to install Windows 7 next as this machine was upgraded from this and I don't recall doing a fresh install of Windows 10 before.

As well as the log files I have included minidump files from when the machine had all the drivers updated and also minidump files from a clean install with verifier running. It would not start when verifier was running with all the drivers updated.

One Drive

Any help or direction you can give me so that I can nurse this machine through 1 more year would be greatly appreciated.

If you're saying (as I think you are) that you've been getting BSODs during Windows installation then that's 100% a hardware issue.

All the dumps you uploaded are almost identical. All of them fail at the same place, in the nt!PpmPerfSelectProcessorState function call...

Code:

FAILURE_BUCKET_ID:  0x1E_c0000094_VRF_STACKPTR_ERROR_nt!PpmPerfSelectProcessorState

I can see from this that you have Driver Verifier enabled - was that deliberate?

You can see that the problem is a stackpointer error, that means that somehow a stack pointer has been corrupted. The exception code here is an 0xC0000094, which is an attempt to divide by zero - possibly because of the stack pointer error.

The other important information we can get from this is the name of the kernel function responsible; nt!ppm. The ppm function is the processor power management function, and in all of the dumps the processor starts in the idle loop (nt!KiIdleLoop) which means that the processor has no work to do. In the idle state processor power management put the processor in a lower power state to conserve power and reduce heat.

We then see the nt!KiExecuteAllDpcs function call, to begin processing the DPC queue (DPCs are the back-end of device interrupt processing). That causes the processor to begin executing again, and so processor power management needs to bring it out of the lower power state into the normal power state - and that's where the problem starts.

If you look at this call stack, which is similar in all dumps, you can see what's happened...

Code:

0: kd> knL
 # Child-SP          RetAddr               Call Site
00 fffff806`42689158 fffff806`3fd0c1ae     nt!KeBugCheckEx
01 fffff806`42689160 fffff806`3fc19392     nt!HvlpVtlCallExceptionHandler+0x22
02 fffff806`426891a0 fffff806`3faf88c7     nt!RtlpExecuteHandlerForException+0x12
03 fffff806`426891d0 fffff806`3fafa896     nt!RtlDispatchException+0x297
04 fffff806`426898f0 fffff806`3fc0fdb2     nt!KiDispatchException+0x186
05 fffff806`42689fb0 fffff806`3fc0fd80     nt!KxExceptionDispatchOnExceptionStack+0x12
06 fffff806`42674238 fffff806`3fc22e65     nt!KiExceptionDispatchOnExceptionStackContinue
07 fffff806`42674240 fffff806`3fc1b204     nt!KiExceptionDispatch+0x125
08 fffff806`42674420 fffff806`3fb02a86     nt!KiDivideErrorFault+0x304
09 fffff806`426745b0 fffff806`3fb028ae     nt!PpmPerfSelectProcessorState+0x1a6
0a fffff806`426746d0 fffff806`3fa9b810     nt!PpmPerfSelectProcessorStates+0x5e
0b fffff806`42674700 fffff806`3fa9b9c6     nt!PpmCheckRun+0x40
0c fffff806`42674770 fffff806`3fb1eeee     nt!PpmCheckStart+0x106
0d fffff806`42674810 fffff806`3fa9e57e     nt!PpmCheckPeriodicStart+0x3e
0e fffff806`42674860 fffff806`3fa9d864     nt!KiExecuteAllDpcs+0x30e
0f fffff806`426749d0 fffff806`3fc1359e     nt!KiRetireDpcList+0x1f4
10 fffff806`42674c60 00000000`00000000     nt!KiIdleLoop+0x9e

You read these call stacks from the bottom up. After the nt!KiExecuteAllDpcs function call you can see a series of nt!ppm calls as the processor power manager raises the power state of the processor to normal. Immediately after the nt!PpmPerfSelectProcessorState function call we see the problem - the nt!KiDivideErrorFault.

This all looks to be a CPU issue, it's having a problem changing power states, and that's not that uncommon - I've seen it several times before. If your BIOS allows you to disable the processor C-States (they're the different power states) then disable them. Otherwise try setting the power profile to the Windows default High Performance profile. In that profile the C-States are disabled. Another way to try disabling C-States is to set the Processor Power Management settings in whichever power profile you use, so that the maximum processor state and the minimum processor state are both 99%.

Hi thank you for your response.

I was not sure about the options in the bios and could not see any options in my bios that related to disabling the c-states of the CPU. Options listed were CPU C1E, CPU C3 Report, CPU C6 Report. Old forum posts I found suggested that these were c-states so I tried with all these disabled with no luck.

I switched to the high performance profile and the BSOD still occurred. I then tried standard power management at 99% and high performance at 99% and the BSOD was still present.

Hi Zbook

Just a quick reply to your pointers

1) I will swap the installation over to a larger hard drive, this has been done before but I ended up with this one when I collected the minidumps.

2)Will do

3)I tried to upgrade the BIOS to latest version, still over 10 years old, and got an error message. I will repeat this so that I can note down the message.

4)WDV is off it was only on for the second group of minidump files.

5)I left the computer on in safe mode with networking last night and it is still on 9 hours later, I will leave it on today.

6) I stopped updating the drivers after each re install of windows, I will update the graphics card driver though if there are errors relating to this in the minidump files.

Many thanks

Please continue in safe mode continuously for 24 to 48 hours and for any BSOD post a new V2 share link into the newest post.

The longer the continuous safe mode the better the testing environment.



There were multiple missed BIOS upgrades.

Upgrade to one of the other more recent versions as needed.



For any new BSOD with greater than 30 GB free space open administrative command prompt and copy and paste:

PowerShell $D=(New-Object -ComObject Wscript.Shell).SpecialFolders(4);Compress-Archive "%WinDir%\MEMORY.DMP" "$D\Memory_dmp"

Post a share link with only the compressed memory dump into this thread.

Code:

Event[854]:
  Log Name: System
  Source: Microsoft-Windows-WER-SystemErrorReporting
  Date: 2023-07-21T21:15:13.4450000Z
  Event ID: 1018
  Task: N/A
  Level: Information
  Opcode: N/A
  Keyword: Classic
  User: N/A
  User Name: N/A
  Computer: DESKTOP-OGEIDLE
  Description: 
The dump file at location: C:\Windows\MEMORY.DMP was deleted because the disk volume had less than 25 GB free space.

So I´ve installed to a larger SSD, updated all the drivers and tried to flash all the different updates to the BIOS from Asus. I was doing this through the tool in the BIOS and got the error ´"image integrity check failed".

I´ve run tune up plus and the file is now in my sharepoint folder.

I´ve put it into safe mode and will report back after 24hrs

- - - Updated - - -

Ok so its been between 24 to 48 hours running in safe mode, no crashes, so I presume you don´t need a new V2 report or minidump?

Monitor the computer using reliability monitor > for any new BOD post a new V2 share link into the newest post

Reliability Monitor is the Best Windows Troubleshooting Tool You Aren't Using

It hasn´t done this before... I was running reliability monitor when it crashed and the bsod stalled at 100%, not sure if that´s related. It does this every time now. Should I Post the log from this or do I need to keep running it?

The V2

Did the latest BSOD occur in normal boot or safe mode?


Open administrative command prompt and copy and paste:

PowerShell $D=(New-Object -ComObject Wscript.Shell).SpecialFolders(4);Compress-Archive "%WinDir%\MEMORY.DMP" "$D\Memory_dmp"

Post a separate share link into this thread with only the compressed memory dump.