Random 0x9F BSOD - even after replacing NVMe drive

cmdwedge · Mar 16, 2023

Hi all, thanks for any assistance you might have for me!

I've just built my first new PC in years and have reused a small number of parts from my old build. Unfortunately, a BSOD has followed me to the new PC - I have a feeling about which piece of hardware is causing the issue but it would be great to have that verified.

Issue is a BSOD (stop code 0x9f) mainly when gaming but sometimes when web browsing. My old PC (i7-6700k system) started doing this about 18 months ago. I haven't used my PC since then, and have only now saved up to buy a new PC to replace it.

System is not a prebuilt. It's a desktop PC. Windows 11 64-bit, activated and running all of the available updates from Windows Update. This is the first OS I have installed on this PC. The installation is 3 days old.

PC is mostly new parts - new case, PSU, motherboard, RAM, CPU and CPU cooler. I have reused the NVME SSD and two mechanical HDDs from my old system - though I will note that the mechanical HDDs were not present in my old PC, I had them mounted into an external enclosure as the system was ITX. The new PC is ATX and I have mounted the two drives into the case directly. I have reused the Gigabyte GTX 1080 from my old system as well - I have a gut feeling that the card is failing.

CPU - AMD Ryzen 9 7950X
RAM - Kingston Fury Beast DDR5-5200 64GB kit. Slots A2 and B2 (per motherboard manual)
Video card - Gigabyte GTX 1080 G2 Gaming
Motherboard - ASRock B650E PG Riptide WiFi
Power Supply - Deepcool PQ1000M - 1000W PSU

I am unsure about the status of Driver Verifier
I am only running Windows Defender, it's up to date
No proxy, VPN, ipfilters etc.
I am not using any disk image tools.
No overclocking whatsoever. RAM is running at XMP speeds (5200 40-40-40 1.25V).

Speccy - http://speccy.piriform.com/results/XiU5jZKwezH4JO3RWkpjYqW

Thanks so much for your help guys! I really appreciate it.

Phil

ubuysa · Mar 17, 2023

I'd suspect the NVMe SSD initially. All three dumps have DRIVER_POWER_STATE_FAILURE (0x9F) stop codes, two of them failed because a power transition took too long, the third because an IRP was held for too long.

All three have the same failure bucket id - 0x9F_4_storahci_IMAGE_pci.sys, which suggests a stoarge device using the PCIe bus (which must mean the NVMe SSD).

The third dump (with the blocked IRP) allows us to see the driver blocking the IRP...

Code:

11: kd> !irp ffffb08425ddb8e0
Irp is active with 6 stacks 3 is current (= 0xffffb08425ddba40)
 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 ffffb08424018050 00000000 fffff802558f8510-ffffb08426016590 Success Error Cancel pending
           \Driver\storahci    CLASSPNP!ClasspPowerUpCompletion
            Args: 00000000 00000001 00000001 00000000
 [IRP_MJ_POWER(16), IRP_MN_SET_POWER(2)]
            0 e1 ffffb08426016050 00000000 fffff8025476ec60-00000000 Success Error Cancel pending
           \Driver\disk    partmgr!PmPowerCompletion
            Args: 00000000 00000001 00000001 00000000
 [IRP_MJ_POWER(16), IRP_MN_SET_POWER(2)]
            0 e1 ffffb08425fc0940 00000000 fffff80250ba6ad0-ffffb08425fda248 Success Error Cancel pending
           \Driver\partmgr    nt!PopRequestCompletion
            Args: 00000000 00000001 00000001 00000000
 [N/A(0), N/A(0)]
            0  0 00000000 00000000 00000000-ffffb08425fda248

            Args: 00000000 00000000 00000000 00000000

The bottom driver, partmgr.sys, is the Windows partition manager driver. It's not the root cause because it's a Microsoft driver, but the issue is clearly with a storage drive.

We thus have a storage drive problem and PCIe problem, and that points I believe at the NVMe SSD.

cmdwedge · Mar 17, 2023

Wow! And of all things - NVME drives have plummeted in price and it's going to be the cheapest thing to replace. Whew!

That really does explain something for me. I ran MemtestX86 for 24 hours once, no errors. I then (because I suspected the GPU) ran Heaven benchmark on loop for 6 hours - and no crash. Of course.. once Heaven is loaded, it doesn't really need to touch the NVME drive!

You're an absolutely genius, thank you so much. Donation on the way.

P

cmdwedge · Apr 8, 2023

Previous thread: [SOLVED] - BSOD when gaming or web browsing

Speccy output: http://speccy.piriform.com/results/wjMpB6udMk6T0rjZgFn0FQA

Since the previous thread, I replaced the faulty Samsung 970 Evo NVMe with a WD Black SN850X 2TB NVMe, into the PCIE 5.0 NVMe slot. I sent the 970 Evo back to Samsung as it was under warranty, and they replaced it with a 970 Evo Plus which was new. I added the 970 Evo Plus to the PCIE 3.0 slot on the motherboard as D drive. Both NVMe drives updated to latest firmware via their respective applications (WD Dashboard, Samsung Magician).

System is not a prebuilt. It's a desktop PC. Windows 11 64-bit, activated and running all of the available updates from Windows Update. This is the first OS I have installed on this PC. The installation is about 6 days old.

CPU - AMD Ryzen 9 7950X
RAM - Kingston Fury Beast DDR5-5200 64GB kit. Slots A2 and B2 (per motherboard manual)
Video card - Gigabyte GTX 1080 G2 Gaming
Motherboard - ASRock B650E PG Riptide WiFi
Power Supply - Deepcool PQ1000M - 1000W PSU
Primary NVMe - WD Black SN850X 2TB
Secondary NVMe - Samsung 970 Evo Plus

The only component which is from my old PC is the GTX 1080, since I have now replaced the NVMe drive.

I am unsure about the status of Driver Verifier
I am only running Windows Defender, it's up to date
No proxy, VPN, ipfilters etc.
I am not using any disk image tools.
No overclocking whatsoever. RAM is running at XMP speeds (5200 40-40-40 1.25V).

Again, thanks for the amazing help you provide so generously.

SQx · Apr 8, 2023

Hello cmdwedge,

I am SQx and I will be handling your log to help you find the cause of BSOD.

Please give me some time to look it over and I will get back to you as soon as possible.

SQx · Apr 10, 2023

Hello cmdwedge,

Welcome to Sysnative.

I processed your dumps, and it looks like a hard drive is causing the problem.

Code:

7: kd> !irp ffffbf04bcd798e0
Irp is active with 6 stacks 3 is current (= 0xffffbf04bcd79a40)
 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 ffffbf04b3f62050 00000000 fffff80350638510-ffffbf04bce72590 Success Error Cancel pending
           \Driver\storahci    CLASSPNP!ClasspPowerUpCompletion
            Args: 00000000 00000001 00000001 00000000
 [IRP_MJ_POWER(16), IRP_MN_SET_POWER(2)]
            0 e1 ffffbf04bce72050 00000000 fffff8034f52ec60-00000000 Success Error Cancel pending
           \Driver\disk    partmgr!PmPowerCompletion
            Args: 00000000 00000001 00000001 00000000
 [IRP_MJ_POWER(16), IRP_MN_SET_POWER(2)]
            0 e1 ffffbf04bccbb9a0 00000000 fffff8034bba6ac0-ffffbf04bce732d8 Success Error Cancel pending
           \Driver\partmgr    nt!PopRequestCompletion
            Args: 00000000 00000001 00000001 00000000
 [N/A(0), N/A(0)]
            0  0 00000000 00000000 00000000-ffffbf04bce732d8 

            Args: 00000000 00000000 00000000 00000000

Learn more:Bug Check 0x9F: DRIVER_POWER_STATE_FAILURE
A device object has been blocking an IRP for too long a time.

There is an assumption that the indicated problem is somehow related to the AsRsok motherboard. Since it can be reproduced by everyone who uses a similar model: ASRock B650E

As a workaround, try to config to keep the hard disks awake all the time in the Power Option (powercfg.cpl).

Please let me know if this solved the specified problem.

cmdwedge · Apr 10, 2023

Thanks so much matw, especially on an Easter weekend. How frustrating! Looks like a firmware issue that will need attention by Asrock. In the meantime I’ll change the setting you e recommended and see how it goes.

Appreciate it!

SQx · Apr 13, 2023

Hello cmdwedge,

Please let me know if you have any changes, is the issue still reproducible?

xilolee · Apr 18, 2023

The problematic hard drive seems to be: "SCSI\Disk&Ven_WDC&Prod_WD4003FZEX-00Z4S\7&35f8e2bf&0&000000".
"WDC WD4003FZEX-00Z4SA0", from msinfo32 and speccy.
Western Digital Hard Disk Desktop Performance - black, 3.5inches, 4TB, 7200RPM, SATA 3.0 (serial ATA-600, 6 Gb/s), first released at the end of 2013 (?).

It could be a good idea to check this disk: chkdsk e: /scan /perf

cmdwedge · Apr 23, 2023

xilolee said:
The problematic hard drive seems to be: "SCSI\Disk&Ven_WDC&Prod_WD4003FZEX-00Z4S\7&35f8e2bf&0&000000".
"WDC WD4003FZEX-00Z4SA0", from msinfo32 and speccy.
Western Digital Hard Disk Desktop Performance - black, 3.5inches, 4TB, 7200RPM, SATA 3.0 (serial ATA-600, 6 Gb/s), first released at the end of 2013 (?).

It could be a good idea to check this disk: chkdsk e: /scan /perf

Thanks @xilolee

Results from chkdsk:
PS C:\Windows\system32> chkdsk e: /scan /perf
The type of the file system is NTFS.
Volume label is HDD Fast.

Stage 1: Examining basic file system structure ...
1458688 file records processed.
File verification completed.
Phase duration (File record verification): 9.22 seconds.
75 large file records processed.
Phase duration (Orphan file record recovery): 0.43 milliseconds.
0 bad file records processed.
Phase duration (Bad file record checking): 0.22 milliseconds.

Stage 2: Examining file name linkage ...
26522 reparse records processed.
1509312 index entries processed.
Index verification completed.
Phase duration (Index verification): 26.25 seconds.
0 unindexed files scanned.
Phase duration (Orphan reconnection): 221.98 milliseconds.
0 unindexed files recovered to lost and found.
Phase duration (Orphan recovery to lost and found): 0.23 milliseconds.
26522 reparse records processed.
Phase duration (Reparse point and Object ID verification): 65.73 milliseconds.

Stage 3: Examining security descriptors ...
Security descriptor verification completed.
Phase duration (Security descriptor verification): 21.90 milliseconds.
25313 data files processed.
Phase duration (Data attribute verification): 0.26 milliseconds.
CHKDSK is verifying Usn Journal...
37443392 USN bytes processed.
Usn Journal verification completed.
Phase duration (USN journal verification): 340.10 milliseconds.

Windows has scanned the file system and found no problems.
No further action is required.

3815441 MB total disk space.
1910255208 KB in 281652 files.
92600 KB in 25314 indexes.
0 KB in bad sectors.
1682284 KB in use by the system.
65536 KB occupied by the log file.
1994982512 KB available on disk.

4096 bytes in each allocation unit.
976753151 total allocation units on disk.
498745628 allocation units available on disk.
Total duration: 36.13 seconds (36134 ms).

cmdwedge · Apr 23, 2023

SQx said:
Hello cmdwedge,

Please let me know if you have any changes, is the issue still reproducible?

Hi mate, since enabling the 'never power down' setting suggested, I have had no further crashes. Ideally I'd like to get to the point where it can be returned to default settings but it'll get expensive to keep replacing SSDs/HDs to narrow it down. The first thread I posted suggested it was a faulty NVMe drive, now this one suggests a fault with a hard drive. It's all very confusing!

Thanks again for all of your help.

xilolee · Apr 23, 2023

Given that the change suggested by SQx seems to work after 13 days, I would be glad if I were you. ;-)

You could also try western digital dashboard (Supported Products: sanDisk, WD_BLACK, WD HDD & SSD (Black, Blue, Green, Red, Purple, Gold)):
https://wddashboarddownloads.wdc.com/wdDashboard/DashboardSetup.exe

Or to remove that hdd altogether and set the previous settings, as a test.

Random 0x9F BSOD - even after replacing NVMe drive

cmdwedge

Contributor

Attachments

ubuysa

Sysnative Staff
BSOD Kernel Dump Senior Analyst
Contributor

cmdwedge

Contributor

cmdwedge

Contributor

Attachments

SQx

BSOD Kernel Dump Analyst

SQx

BSOD Kernel Dump Analyst

cmdwedge

Contributor

SQx

BSOD Kernel Dump Analyst

xilolee

Moderator

cmdwedge

Contributor

cmdwedge

Contributor

xilolee

Moderator

Random 0x9F BSOD - even after replacing NVMe drive

Contributor

Attachments

Sysnative Staff BSOD Kernel Dump Senior AnalystContributor

Contributor

Contributor

Attachments

BSOD Kernel Dump Analyst

BSOD Kernel Dump Analyst

Contributor

BSOD Kernel Dump Analyst

Moderator

Contributor

Contributor

Moderator

Sysnative Staff
BSOD Kernel Dump Senior Analyst
Contributor