Question Yet another BSOD post - Driver_System_Power_Failure apparently caused by ntoskrnl.exe

[StumpedManiac]

I've been trying to self-diagnose and self-fix this problem for ages, but not even checking the other similar posts here has helped me figure out what the problem is so I am posting myself.

Periodically, every few hours or so when I'm using my computer, I'll get a BSOD with the stop code Driver_Power_System_Failure. I've made sure all my drivers were updated, and using tools like BlueScreenViewer have claimed that it was ntoskrnl.exe to be the culprit, which doesn't tell me much. Here are the dumps: https://drive.google.com/drive/folders/1jECE_GpeR4VeE9aXD5dd85iqGe4QmFts?usp=drive_link

And here are my PC specs:
CPU: Ryzen 7950x
GPU: RTX 3070
Storage: WD Blue SN550 500gig (boot drive, I have a feeling it is the culprit)
WD Black SN750 1tb (game drive)
Samsung 990 Pro 2tb (footage drive)
OS: Windows 11, upgraded from 10 (it's not a fresh install, which could also be the issue)
Ram: 64 gig kit (32x2) of G.Skill TridentZ RGB

I'm truly at a loss, as this only started a few days ago. If anyone can analyze the dumps and let me know what the problem is, I'd love to know! Thank you!

 

[Deleted member 2962262]

Whilst waiting for someone to comment on the dumps, can I suggest you follow through on your instinct: back up your data, unplug the HDDs do a fresh install to the SSD, reconnect the HDDs and see how you go.

 

[Colif]

nice, so completely reinstall windows just to see if you don't get the driver... a little excessive.

BSV always blames ntoskrnl as it was what crashed, its not the cause though... just the victim

It is going to be a driver, we just have to figure out which.

report - mostly for me


File: 122223-16031-01.dmp (Dec 23 2023 - 07:32:46)
BugCheck: [DRIVER_POWER_STATE_FAILURE (9F)]
Probably caused by: memory_corruption (Process: System)
Uptime: 0 Day(s), 19 Hour(s), 02 Min(s), and 01 Sec(s)

File: 122123-18156-01.dmp (Dec 22 2023 - 05:33:13)
BugCheck: [DRIVER_POWER_STATE_FAILURE (9F)]
Probably caused by: memory_corruption (Process: System)
Uptime: 2 Day(s), 8 Hour(s), 16 Min(s), and 55 Sec(s)

File: 122123-16437-01.dmp (Dec 22 2023 - 08:58:58)
BugCheck: [DRIVER_POWER_STATE_FAILURE (9F)]
Probably caused by: memory_corruption (Process: System)
Uptime: 0 Day(s), 3 Hour(s), 24 Min(s), and 27 Sec(s)

File: 122123-16000-01.dmp (Dec 22 2023 - 12:24:51)
BugCheck: [DRIVER_POWER_STATE_FAILURE (9F)]
Probably caused by: memory_corruption (Process: System)
Uptime: 0 Day(s), 3 Hour(s), 05 Min(s), and 30 Sec(s)

File: 121923-18531-01.dmp (Dec 19 2023 - 21:14:59)
BugCheck: [DRIVER_POWER_STATE_FAILURE (9F)]
Probably caused by: memory_corruption (Process: System)
Uptime: 0 Day(s), 6 Hour(s), 08 Min(s), and 05 Sec(s)

it looks to be storage drivers

try updating chipset drivers - https://www.amd.com/en/support/chipsets/amd-socket-am5/b650
you don't need raid mi

Did you update BIOS recently as there is one from this month. I can't tell if you on latest as the report grabs the wrong info for MSI motherboards. There is a beta bios from beginning of month, could be what you have. Not ideal to be on beta bios.

 

[Ralston18]

Question:

PSU: make, model, wattage, age, condition (original to build, new, refurbished, used)?

History of heavy use for gaming, video editing, or even bit-mining?

 

[Deleted member 2962262]

nice, so completely reinstall windows just to see if you don't get the driver

I misread the boot drive as being a HDD, my bad.

 

[StumpedManiac]

Question:

PSU: make, model, wattage, age, condition (original to build, new, refurbished, used)?

History of heavy use for gaming, video editing, or even bit-mining?

The PSU is a fresh out of the box Corsair RM1000x, been using it for 6 months now. I mainly do gaming and video editing, no mining.

Honestly one of my theories is that the power draw is getting messed up somewhere, as the problem started happening after I plugged a space heater into a nearby outlet. It's entirely possible it's a power issue

 

[StumpedManiac]

Did you update BIOS recently as there is one from this month. I can't tell if you on latest as the report grabs the wrong info for MSI motherboards. There is a beta bios from beginning of month, could be what you have. Not ideal to be on beta bios.

Haha I forgot to include my motherboard in the parts list, glad you discerned it was a b650. The thing is, the first thing I did after the first BSOD was update the chipset drivers and it kept happening. Maybe I downloaded the beta bios like you said by accident? Regardless, I went ahead and downloaded the latest drivers again, this time from the link you shared. Maybe that solved it, thank you!

 

[Ralston18]

If the problems continue then there are three things that can be immediately done:

1 ) Disconnect the space heater to determine if the problems stop. If the heater is necessary then try to power if from another physical circuit that uses a different breaker.

2) Look in Reliability History/Monitor for any error codes, warnings, and even informational events that started appearing after the space heater was originally plugged in. Check the timeline.,

3) Run "dism" and "sfc /scannow". Either one or both tools may find and fix corrupted files due to power spikes/surges, etc.

FYI:

https://www.windowscentral.com/how-use-dism-command-line-utility-repair-windows-10-image

https://www.lifewire.com/how-to-use-sfc-scannow-to-repair-windows-system-files-2626161


https://www.windowscentral.com/how-use-dism-command-line-utility-repair-windows-10-image

 

[StumpedManiac]

If the problems continue then there are three things that can be immediately done:

1 ) Disconnect the space heater to determine if the problems stop. If the heater is necessary then try to power if from another physical circuit that uses a different breaker.

2) Look in Reliability History/Monitor for any error codes, warnings, and even informational events that started appearing after the space heater was originally plugged in. Check the timeline.,

3) Run "dism" and "sfc /scannow". Either one or both tools may find and fix corrupted files due to power spikes/surges, etc.

FYI:

https://www.windowscentral.com/how-use-dism-command-line-utility-repair-windows-10-image

https://www.lifewire.com/how-to-use-sfc-scannow-to-repair-windows-system-files-2626161


https://www.windowscentral.com/how-use-dism-command-line-utility-repair-windows-10-image

The heater has been unplugged (though I have since BSOD'd twice since then), however, I when i did the /sfc scannow, it said it found 3 corrupted drivers that it has now fixed (see pic). I don't know what these drivers do, so do you think that them being corrupted was the issue, or a symptom following the BSOD crashes?

EDIT: I forgot to mention this but this is actually a huge breakthrough, looking through the reliability monitor on each day of my crashes shows something similar to this (see pic). The common thread is "ASUS Com Service", I don't even use ASUS products (though my Windows 10 was loaded on an ASUS Motherboard before I upgraded to 11 on an MSI board. Is this all just leftover junk data causing conflictions or something?)

 

[Hotrod2go]

The heater has been unplugged (though I have since BSOD'd twice since then), however, I when i did the /sfc scannow, it said it found 3 corrupted drivers that it has now fixed (see pic). I don't know what these drivers do, so do you think that them being corrupted was the issue, or a symptom following the BSOD crashes?

EDIT: I forgot to mention this but this is actually a huge breakthrough, looking through the reliability monitor on each day of my crashes shows something similar to this (see pic). The common thread is "ASUS Com Service", I don't even use ASUS products (though my Windows 10 was loaded on an ASUS Motherboard before I upgraded to 11 on an MSI board. Is this all just leftover junk data causing conflictions or something?)

Those drivers that windows fixed for you are from the blue tooth device on your motherboard.
Also, going by the extra info you provided - I would most certainly reinstall the OS because going from an Asus board to another manufacturer like MSI, & upgrading the OS from 10 to 11 would most definitely be begging for a clean install (because like you say - left over "junk") as opposed to the incremental install you have done going from 10 - 11.
Download the latest win 11 package from MS & do a clean install, when windows prompts you in the first stages, tell it to completely format your nominated storage device for C drive. Be sure to run the bios at defaults & then continue what you typically like to do on the system after all the software has been setup, then see if these BSOD continue.

 

[Ralston18]

It all may come down to doing a clean install.

There are still other things that can be done before going forward with a clean install.

Just in case:

https://forums.tomshardware.com/threads/windows-11-clean-install-tutorial.3831442/

= = = =

What Bluetooth devices are being used?

My suggestion would be to at least disable (if not uninstall) those bluetooth drivers.

And check that they are not being reinstalled or launched via Task Manager > Startup, Task Scheduler, or some other utility/app.

Run "dism" and "sfc /scannow" again. Determine if the BSOD's stop and that any installed Bluetooth devices are working via generic drivers.

If not, then manually download, reinstall, and reconfigure the device Bluetooth drivers via the applicable manufacturer's website. No third party tools or installers.

Also be sure that the source website is truly the applicable manufacturer's website. Just because "Manufactuer X's" name appears in the URL does not mean that the website is actually the manufacturer.

 

[StumpedManiac]

It all may come down to doing a clean install.

Those drivers that windows fixed for you are from the blue tooth device on your motherboard.
Also, going by the extra info you provided - I would most certainly reinstall the OS because going from an Asus board to another manufacturer like MSI, & upgrading the OS from 10 to 11 would most definitely be begging for a clean install (because like you say - left over "junk") as opposed to the incremental install you have done going from 10 - 11.
Download the latest win 11 package from MS & do a clean install, when windows prompts you in the first stages, tell it to completely format your nominated storage device for C drive. Be sure to run the bios at defaults & then continue what you typically like to do on the system after all the software has been setup, then see if these BSOD continue.

So yeah, I ended up doing a complete total reinstall of Windows- freshly downloaded off of the MS website onto a thumbdrive. However, the BSOD's have still not been fixed. Everything is as default and vanilla as possible (besides reinstalling the latest drivers), yet it still crashes. I'm at a loss here. The thing is, this whole time, it never has crashed while I was using it. It has only ever crashed whenever I leave it alone for a while, so that's my last lead on diagnosing this before resorting to simply replacing the SSD.

Here's a dump post-refresh, I only have one this time https://drive.google.com/file/d/1XCZDCn1j7fFJHwgBYWYjtpWMai3yMNje/view?usp=sharing

EDIT: I have one more lead. If this turns out to be the culprit, I will both laugh and cry. I checked reliability montitor, and the only thing that has been consistent throughout all of this has been "Microsoft GameInput" having a "successful application installation" right before the crash.

GameInput has been consistent across every single crash. I didn't think any of it because how could a game controller input cause a crash? But you never know. If this *is* the problem, how would I go about fixing it?

 

[Ralston18]

Click the "View technical details....." for Critical Events.

Make note of any corresponding error codes, warnings, or informational events: post accordingly.

 

[StumpedManiac]

Click the "View technical details....." for Critical Events.

Make note of any corresponding error codes, warnings, or informational events: post accordingly.

This is all I could find, under "Windows stopped working". The other one just says "Windows shut down unexpectedly", which... duh. Edit: As for the informational events, GameInput is still the only informational event with any relevant data, aka the time of the crash being the same as the time as GameInput had a "successful application installation". I seriously can't see that as a coincidence that all 6ish of my crashes have had GameInput update immediately before them.

Edit 2: Oh, and there's "WindowsPackageManagerServer.exe Stopped Working", but these happen seemingly at random. Idk if they have anything to do with the crashes, but they seem important nonetheless so I'm putting an example of it here.

 

[Hotrod2go]

OP, I notice you've got 2x32GB kit of DDR5 ram & even though its running at default settings in bios as you claim when reinstalling windows, is this set of RAM on the support list for your motherboard?

Even though gameinput package installer just gets installed & then it crashes everytime after that, its got me thinking this may be tipping your system over the edge if the RAM or/& the IMC don't get enough AND steady supply of voltage due to minor but critical power spikes. 64GB kits of DDR5 ram are not as mature with bios settings as the more commonly used 32GB kits or even the now 48GB kits that are becoming more popular are on DDR5 based platforms. Of course, in time 64GB kits will mature with bios settings but they are not a common thing yet.

 

[StumpedManiac]

OP, I notice you've got 2x32GB kit of DDR5 ram & even though its running at default settings in bios as you claim when reinstalling windows, is this set of RAM on the support list for your motherboard?

Even though gameinput package installer just gets installed & then it crashes everytime after that, its got me thinking this may be tipping your system over the edge if the RAM or/& the IMC don't get enough AND steady supply of voltage due to minor but critical power spikes. 64GB kits of DDR5 ram are not as mature with bios settings as the more commonly used 32GB kits or even the now 48GB kits that are becoming more popular are on DDR5 based platforms. Of course, in time 64GB kits will mature with bios settings but they are not a common thing yet.

This is likely as anything else. Though I couldn't for the life of me find the boards supported RAM, beyond it saying it could handle 64 gigs. But for now, I'm about to go out of town for a few days and won't be able to test the RAM or make any other changes until I'm back home.

Besides that though, an update. I haven't had any crashes for a whole 24 hours! I even purposely left the computer on overnight to see if I'd come back to another crash and it didn't. The only things I changed were doing a storage drive check using a CHKDSK /f /r /x command and setting the "GameInput" service to automatic instead of manual (I think I read that this would fix it somewhere). I actually think that one of those things fixed it, but I'll have to wait until I'm back home to fully make sure. But hey, no crashes! This is definitely an improvement. I just hope the previous crashes didn't cause any damage.

 

[StumpedManiac]

The only things I changed were doing a storage drive check using a CHKDSK /f /r /x command and setting the "GameInput" service to automatic instead of manual (I think I read that this would fix it somewhere). I actually think that one of those things fixed it, but I'll have to wait until I'm back home to fully make sure. But hey, no crashes! This is definitely an improvement. I just hope the previous crashes didn't cause any damage.

Welp, now that I'm home, I can safely say that one of those two things mostly likely fixed it. I ran the computer for another full 24 hours since I've been home and I haven't had a single crash, whereas before I'd have multiple. Now I'm using it for work, no issues. Maybe I just jinxed it, but either way I'm VERY relieved. If it starts happening again, I'll likely try replacing the ram or just asking here again. Thank you to everyone who has helped so far!

 

[ubuysa]

Apologies for not responding earlier, I've been having issues with my left eye and my ophthalmologist ordered me not to use the PC for at least 10 days(!) and then there were the holidays....

The issue seems sorted but I can confirm from the dumps that the issue was (is) with the 2TB Seagate Barracuda HDD. All the dumps show that a device failed to power up (from idle low power state to running normal power state). The dump also shows that the problem device was the Seagate HDD. Here's the key data from one of the dumps...

4: kd> !devnode ffffe403919a88a0
DevNode 0xffffe403919a88a0 for PDO 0xffffe403919c0050
  Parent 0xffffe40391933a20   Sibling 0000000000   Child 0000000000 
  InstancePath is "SCSI\Disk&Ven_&Prod_ST2000DM008-2FR1\7&3611b6d6&0&010000"
  ServiceName is "disk"
  State = DeviceNodeStarted (0x30a)
  Previous State = DeviceNodeEnumerateCompletion (0x30f)
  StateHistory[15] = DeviceNodeEnumerateCompletion (0x30f)
  StateHistory[14] = DeviceNodeEnumeratePending (0x30e)
  StateHistory[13] = DeviceNodeStarted (0x30a)
  StateHistory[12] = DeviceNodeEnumerateCompletion (0x30f)
  StateHistory[11] = DeviceNodeEnumeratePending (0x30e)
  StateHistory[10] = DeviceNodeStarted (0x30a)
  StateHistory[09] = DeviceNodeEnumerateCompletion (0x30f)
  StateHistory[08] = DeviceNodeEnumeratePending (0x30e)
  StateHistory[07] = DeviceNodeStarted (0x30a)
  StateHistory[06] = DeviceNodeStartPostWork (0x309)
  StateHistory[05] = DeviceNodeStartCompletion (0x308)
  StateHistory[04] = DeviceNodeStartPending (0x307)
  StateHistory[03] = DeviceNodeResourcesAssigned (0x306)
  StateHistory[02] = DeviceNodeDriversAdded (0x305)
  StateHistory[01] = DeviceNodeInitialized (0x304)
  StateHistory[00] = DeviceNodeUninitialized (0x301)
  StateHistory[19] = Unknown State (0x0)
  StateHistory[18] = Unknown State (0x0)
  StateHistory[17] = Unknown State (0x0)
  StateHistory[16] = Unknown State (0x0)
  Flags (0x24000130)  DNF_ENUMERATED, DNF_IDS_QUERIED,
                      DNF_NO_RESOURCE_REQUIRED, DNF_NO_LOWER_DEVICE_FILTERS,
                      DNF_NO_UPPER_DEVICE_FILTERS
  CapabilityFlags (0x00402180)  SilentInstall, RawDeviceOK,
                                WakeFromD3
                                Unknown flags 0x00400000

The InstancePath value of SCSI\Disk&Ven_&Prod_ST2000DM008-2FR1 identifies the Seagate drive as the problem. Notice at the bottom there that it's supposed to be able to wake from D3 state (idle low power) but the BSOD shows that it didn't.

Displaying the IRP managing the power transition also clearly shows that it's a disk drive problem...

4: kd> !irp ffffe4039bd388a0
Irp is active with 6 stacks 3 is current (= 0xffffe4039bd38a00)
 No Mdl: No System Buffer: Thread 00000000:  Irp stack trace.
     cmd  flg cl Device   File     Completion-Context
 [N/A(0), N/A(0)]
            0  0 00000000 00000000 00000000-00000000  

            Args: 00000000 00000000 00000000 00000000
 [N/A(0), N/A(0)]
            0  0 00000000 00000000 00000000-00000000  

            Args: 00000000 00000000 00000000 00000000
>[IRP_MJ_POWER(16), IRP_MN_SET_POWER(2)]
            0 e1 ffffe403919c0050 00000000 fffff80115498510-ffffe4039c0ac590 Success Error Cancel pending
           \Driver\storahci    CLASSPNP!ClasspPowerUpCompletion
            Args: 00000000 00000001 00000001 00000000
 [IRP_MJ_POWER(16), IRP_MN_SET_POWER(2)]
            0 e1 ffffe4039c0ac050 00000000 fffff8011435ec60-00000000 Success Error Cancel pending
           \Driver\disk    partmgr!PmPowerCompletion
            Args: 00000000 00000001 00000001 00000000
 [IRP_MJ_POWER(16), IRP_MN_SET_POWER(2)]
            0 e1 ffffe4039beee9a0 00000000 fffff8011112cf80-ffffe4039c046248 Success Error Cancel pending
           \Driver\partmgr    nt!PopRequestCompletion
            Args: 00000000 00000001 00000001 00000000
 [N/A(0), N/A(0)]
            0  0 00000000 00000000 00000000-ffffe4039c046248  

            Args: 00000000 00000000 00000000 00000000

You can see three drivers on the IRP queue - all power related. The bottom one ( \Driver\partmgr nt!PopRequestCompletion) is the one that's stuck waiting for the drive to power up.

The chkdsk may have found and fixed a bad sector or a glitch in the filesystem, but this may also be a drive controller issue. If this were mine I would ensure that all data on that drive was backed up at all times, and if this BSOD happens again I would replace the drive.

There is a workaround for this issue too that you may want to try. In your power profile, change the 'Turn off hard disk' setting to 0 minutes. This stops the HDDs from powering down, so this powering up problem should not happen again.

 

[StumpedManiac]

Apologies for not responding earlier, I've been having issues with my left eye and my ophthalmologist ordered me not to use the PC for at least 10 days(!) and then there were the holidays....

The issue seems sorted but I can confirm from the dumps that the issue was (is) with the 2TB Seagate Barracuda HDD. All the dumps show that a device failed to power up (from idle low power state to running normal power state). The dump also shows that the problem device was the Seagate HDD. Here's the key data from one of the dumps...
~~
The InstancePath value of SCSI\Disk&Ven_&Prod_ST2000DM008-2FR1 identifies the Seagate drive as the problem. Notice at the bottom there that it's supposed to be able to wake from D3 state (idle low power) but the BSOD shows that it didn't.

Displaying the IRP managing the power transition also clearly shows that it's a disk drive problem...
~~~
You can see three drivers on the IRP queue - all power related. The bottom one ( \Driver\partmgr nt!PopRequestCompletion) is the one that's stuck waiting for the drive to power up.

The chkdsk may have found and fixed a bad sector or a glitch in the filesystem, but this may also be a drive controller issue. If this were mine I would ensure that all data on that drive was backed up at all times, and if this BSOD happens again I would replace the drive.

There is a workaround for this issue too that you may want to try. In your power profile, change the 'Turn off hard disk' setting to 0 minutes. This stops the HDDs from powering down, so this powering up problem should not happen again.

Oh, sorry about your eyes! Hope you've recovered well. Also... you've gotta be kidding me! I was so focused on my M.2s that I neglected to mention the backup HDD I also had because I never use it. A huge oversight on my part, I feel like this could've been solved a lot faster had I mentioned it! But also, WOW! Im amazed that you actually found a definitive cause! That means not only was GameInput not the issue, but it means that it was the chkdsk that saved me! It also means that the permanent fix should be as easy as grabbing a new HDD rather than an M.2 or even RAM, which is the best outcome honestly.

Thank you so much! I'm glad to have a definitive answer about why that kept happening. The drive is pretty old so maybe I might as well replace it regardless- maybe with a SATA SSD, as a treat!

 

[ubuysa]

Thank you so much! I'm glad to have a definitive answer about why that kept happening. The drive is pretty old so maybe I might as well replace it regardless- maybe with a SATA SSD, as a treat!

You're more than welcome. If the drive is old I would replace it. If it's a drive used only for backup I would replace it with an HDD rather than an SSD. You don't need blistering read/write speeds for backup and restore and when an SSD fails it fails catastrophically. When an HDD fails as long as the drive spins up you can usually get the data off it.