Question Computer keeps crashing - - - Kernel Power - Event ID 41 ?

[cablelayer]

Consistantly happening on daily basis - black screen crash to restart.

Minidump here:

https://we.tl/t-KBwXpuuPqm

 

[faalin]

can you give full system specs. make, model, and how old is the PSU. when is the system crashing, while its idle, while loading a game, while gaming?

 

[ubuysa]

You talk about having many crashes/BSODs but you uploaded only one dump. It's always useful to see all the relevant dumps, but instead of just uploading the dumps I'd appreciate it if you would download and run the SysnativeBSODCollectionApp. This gathers all the necessary troubleshooting data (including all dumps) and writes it all to a zip file. Please upload this zip file (SysnativeFileCollectionApp.zip).

The SysnativeBSODCollectionApp does not collect any personally identifying data. It's used by many Windows help forums and is known to be safe. If you want to see what data the app collects there are full details here.

 

[cablelayer]

Thanks guys, sorry for the long wait, im at work and we are actively using this machine (offshore on a ship).

PC runs a Survey Aquistion software called Qinsy - basically it timestamps and stores lot of data streaming to the PC over 2 networks from all sorts of instruments - doesnt seem too taxing on the CPU but i believe the aquisiton side of the software still runs in 32bit.

Computer runs basically 24hrs a day - we get this BSOD about once per week

Some peripherals:
2 ethernet adapters
4 screens (quite shitty quality ones a few have been flickering)
One physical COM port (Recieving PPS time sync from GPS)
USB extension cables x2
Software Dongle
Swapped wireless mouse for wired ones - no improvement

Sometimes we have down-time for a few hours so I was able to run SysnativeBSODCollectionApp here are the results:

https://we.tl/t-LlR84WpQjB

Thanks again!

 

[ubuysa]

Your problem would appear to be an outdated driver for a Moxa NPort COM Port device. The drive in question is nptdrv2.sys and it's clearly involved in four of your five dumps. The version you have installed is ancient, dating from 2016...

4: kd> lmvm nptdrv2
Browse full module list
start             end                 module name
fffff804`47df0000 fffff804`47e04000   nptdrv2  T (no symbols)          
    Loaded symbol image file: nptdrv2.sys
    Image path: nptdrv2.sys
    Image name: nptdrv2.sys
    Browse all global symbols  functions  data
    Timestamp:        Mon Jul 11 15:16:06 2016 (57838E06)
    CheckSum:         00014752
    ImageSize:        00014000
    Translations:     0000.04b0 0000.04e4 0409.04b0 0409.04e4
    Information from resource tables:

I would suggest that you contact Moxa Support and seek an updated driver.

 

[cablelayer]

Good day all and thanks for your for your assistance so far. I followed the advice and contacted Moxa who confirmed likely cause is an outdated driver. I uninstalled the old and installed the windows 10 drivers reccomended by Moxa. However after a couple of days I got another BSOD and reboot. I checked in windbg and got this:

SYMBOL_NAME: ACPI!CmosTranslatePNPIDToEnum+49

MODULE_NAME: ACPI

IMAGE_NAME: ACPI.sys

IMAGE_VERSION: 10.0.19041.4106

STACK_COMMAND: .cxr 0xfffff18dd1667f00 ; kb

BUCKET_ID_FUNC_OFFSET: 49

FAILURE_BUCKET_ID: AV_ACPI!CmosTranslatePNPIDToEnum

OSPLATFORM_TYPE: x64

OSNAME: Windows 10

FAILURE_ID_HASH: {13fa9b99-1049-b0b3-106a-a3b80e048cac}

Followup: MachineOwner

Some quick googling shows again a driver issue, but no mention of Moxa.

Please see full dump using SysnativeBSODCollectionApp here:

https://we.tl/t-MaZM5pYlvB

Any help appreciated.

 

[cablelayer]

Further to my last post, I ran sfc /scannow as administrator. The system confirmed that some errors were fixed. Please find the log file it directed me to here:

https://we.tl/t-RhUIU6yIHZ

Any chance that the scan fixed the error causing the BSOD?

Thanks!

 

[ubuysa]

I just went back over all the uploaded dumps, the Moxa driver dump is an outlier (which I recognised earlier, but one problem at a time!), all of the others are similar in that they are all ACPI failures. Close examination of all these dumps shows that the BSOD happens because an ACPI function failed due to insufficient resources. Here's the problem function call, which we see in all these ACPI dumps...

01 ffffd50bc5268910 fffff8050813c267 ACPI!CmosTranslatePNPIDToEnum+49
    Parameter[0] = ffffc50bfef4fb08
    Parameter[1] = 00000000c000009a
    Parameter[2] = 0000000000000000
    Parameter[3] = ffffc50c09598bf0

The function here appears to be a call to a CMOS (UEFI BIOS) ACPI function to enumerate (translate) a PnP device ID, and so is probably part of device enumeration at boot time or when a PnP device is plugged in, powered on, or transitions from a low power (idle) state to a high power (running) state. None of this is documented in any great detail of course. Parameter 1 in that function call contains an error code 0xC000009A, and if you look that up you'll see it's an 'insufficient resources' error....

Error code: (NTSTATUS) 0xc000009a (3221225626) - Insufficient system resources exist to complete the API.

This explains why the BSOD happened, but sadly those two addresses in parameters 1 and 3 are not included in the minidump, they might enable identification of the device and help us discover what happened. Can you please upload the kernel dump? It's the file C:\Windows\Memory.dmp and it will be large.

I would suggest also that you unplug everything except keyboard, mouse, and monitor and see whether the BSODs persist. It's also worth removing and re-seating all PCIe cards - these are all PnP deviuces managed by ACPI.

In addition, there have been several BIOS updates since the F7 version that you have. Since this seems as though there may be a UEFI BIOS component I think it would be wise to bring your BIOS version up to the current F28 version (2nd Jan 2024).

BTW: The sfc scan only fixed a bad shortcut link, so no, I didn't fix this problem..

 

[cablelayer]

Hi Ubuysa,

Wow. Thanks for the assistance.
There was no memory dump file (or I coldnt find it at least) So after quick googling I found

NotMyFault - Sysinternals ​

Which when I crashed it created the dump. Ive attached the momory.dmp file here (11gb):

https://we.tl/t-3zWKMhhaZ4

Unfortunatley due to the type of work we are doing I cant unplug everything except the essentials at this time. The operation would stop. So short of swtiching out the PC that will be difficult for now - the crash happends only once every few days, but of course always at the worst possible time!

Maybe the 11gb file sheds some light on the cause? I can definately look to reseat the PCI cards. That only takes a few minutes. Ill give that a try.

 

[ubuysa]

Ah, you misunderstand I think. A memory dump (a copy of the contents of RAM) is taken whenever the system BSODs. By examining the resulting dump we can often determine what caused the BSOD. If I looked at the dump you have uploaded it would tell me that notmayfault.sys was the cause - but that's no help at all of course!

Do you run any disk cleaning tools? They do delete dumps. Also, please enter the command sysdm.cpl in the Run command box. In the window that opens click the Advanced tab. In there click the bottom Settings button. Post a screenshot of the display that results.

Since it seems to be a PnP device that's at fault here I would remove and re-seat all PCIe cards and remove and re-plug all USB connected devices. Also make sure that all drivers for your PCIe, USB attached devices, AND your motherboard chipset drivers are up to date.

 

[cablelayer]

Understood on the memory dump. I was just confused at there was no dump where it should be. I did check the startup and recovery settings as you menion above. Its set to write an automatic dump file to %systemroot%\MEMORY.DMP

But I couldnt find the .DMP file on any of the drives - I found all the Minidumps. Thats why I created a .DMP, but as youve explained above I realise now that its useless.

I took apart the PC and reseated the PCI cards, air duster etc. I was also able to remove entirely 2 x Moxa PCI 8 port cards that are not in use. Im hoping that these might have been the cause. Ill update the Motherboard and chipset drivers when convinient.

Ill report back one was or the other.

Again huge thanks for taking you time to assist. Its much appreciated.

 

[ubuysa]

Understood on the memory dump. I was just confused at there was no dump where it should be. I did check the startup and recovery settings as you menion above. Its set to write an automatic dump file to %systemroot%\MEMORY.DMP

But I couldnt find the .DMP file on any of the drives - I found all the Minidumps. Thats why I created a .DMP, but as youve explained above I realise now that its useless.

Curious. Those settings are correct. Never mind, I was only hoping the kernel dumps might allow us to identify the device.

I took apart the PC and reseated the PCI cards, air duster etc. I was also able to remove entirely 2 x Moxa PCI 8 port cards that are not in use. Im hoping that these might have been the cause.

Ahhh. Now that does sound like a possibility. It ties in with the Moxa driver BSOD too and I don't believe in having two different problems at the same time.

Ill update the Motherboard and chipset drivers when convinient.

Ill report back one was or the other.

Perfect. Please do.

 

[cablelayer]

Ok So I has another crash overnight even after removing the Moxa Serial com port PCI cards. Again Minidump said ACPI.sys is the cause. Oddly again no full dump file (?!)

The Symbol name did say: ACPI!CmosTranslatePNPIDToEnum+49 as before.

So this morning ive updated the Bios to the newest from Gigabyte website and flashed using Qflash. It installed some kind of app updater, so I temporarly connected the PC to the internet to let it do its thing and download required drivers.

Ill monitor and report back in a few days.

 

[cablelayer]

Hi Ubuysa,

I have an update. 5 days later and the PC is still running without error/crash. It seems that the Bios update did the trick.

Thanks again for all of your assistance - much appreciated.

Ill continue to monitor and only will report back if it falls over again, but for not I think that the issue is solved.

Thanks again!

 

[cablelayer]

Update. More than a month later the problem is certainly solved.

 

[ubuysa]

Good to hear