Question "DRIVER_POWER_STATE_FAILURE" BSOD ?

[HenryPottier]

Hi people,

After assembling my new PC, I've had four blue screens of death (two different sessions, twice at the beginning of each session, no longer after). The first two times, the PC crashed when I launched GPUZ (PC semi froze and then after 20 seconds BSODed), second session was during a random internet session (GPUz has been uninstalled and reinstalled, it now works).


I collected the dmp files and tried to look into it, by looking how some people here did, but I'm a bit out of my depth, I have to admit. I can't find a device that would be the responsible and beyond that... big ? for me.

If any charitable soul could help me, that would be great.

PC Specs
Motherboard: MSI B850 Tomahawk
CPU: Ryzen 7 9800X3D
GPU: Asus 5070 Ti Prime
RAM: 64GB DDR5 [2 x 32GB] in the correct slots
3 x NVMe SSD and 1 x SATA SSD

Only one NVME is new and I cloned Windows on it from my old rig with Macrium Reflect.
The other drives come from the old PC.

Nothing is overclocked. I launched two demanding games (ultra modded Skyrim and Doom dark ages), seemed perfectly fluid. Temps seem normal.


Dmp files for the 4 crashes in one zip :
https://drive.google.com/file/d/1ZlUTywiFI7wKNo-V-GZ53ZmcDEQktA10/view?usp=drive_link

Thanks in advance.

 

[ubuysa]

Two of those dumps are 0x9F bugchecks and both caused by the Crucial MX500 SSD failing to complete a power transition in time. Here, for example, is the IRP that manages the power transition...

0: kd> !irp ffff800bbc1759c0
Irp is active with 8 stacks 4 is current (= 0xffff800bbc175b68)
 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
 [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 ffff800bb7f8e050 00000000 00000000-00000000    pending
           \Driver\storahci
            Args: 00000000 00000001 00000001 00000000
 [IRP_MJ_POWER(16), IRP_MN_SET_POWER(2)]
            0 e1 ffff800bbc04ada0 00000000 fffff80044f68510-ffff800bbc155730 Success Error Cancel pending
           \Driver\EhStorClass    CLASSPNP!ClasspPowerUpCompletion
            Args: 00000000 00000001 00000001 00000000
 [IRP_MJ_POWER(16), IRP_MN_SET_POWER(2)]
            0 e1 ffff800bbc1551f0 00000000 fffff80043e3ec60-00000000 Success Error Cancel pending
           \Driver\disk    partmgr!PmPowerCompletion
            Args: 00000000 00000001 00000001 00000000
 [IRP_MJ_POWER(16), IRP_MN_SET_POWER(2)]
            0 e1 ffff800bbbf888d0 00000000 fffff8003f7a6b30-ffff800bbc175228 Success Error Cancel pending
           \Driver\partmgr    nt!PopRequestCompletion
            Args: 00000000 00000001 00000001 00000000
 [N/A(0), N/A(0)]
            0  0 00000000 00000000 00000000-ffff800bbc175228    

            Args: 00000000 00000000 00000000 00000000

At the bottom there are the drivers that are holding up the IRP and preventing it from completing. You can see that they are all disk related. The dumps also give us the address of the device object, from which we can access the device node. This describes the actual device...

0: kd> !devnode ffff800bb7fa9c20
DevNode 0xffff800bb7fa9c20 for PDO 0xffff800bb7f8e050
  Parent 0xffff800bb7f35a60   Sibling 0000000000   Child 0xffff800bbbff68a0   
  InstancePath is "SCSI\Disk&Ven_&Prod_CT500MX500SSD1\7&2a9b7bbb&0&010000"
  ServiceName is "disk"
  State = DeviceNodeStarted (0x30a) @ 2023 May 26 08:32:08.995
  Previous State = DeviceNodeEnumerateCompletion (0x30f) @ 2023 May 26 08:32:08.995
  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 (0x00002180)  SilentInstall, RawDeviceOK, 
                                WakeFromD3

The InstancePath entry contains the hardware description: SCSI\Disk&Ven_&Prod_CT500MX500SSD1\7&2a9b7bbb&0&010000 and that is your Crucial MX500. You can see at the bottom above that it is able to wake from the D3 state, this is one of the low power states and the BSOD resulted because it failed to wake from that state into the fully running state (D0).

You could try changing the hard disk timeout time in your power option plan and setting it to 0. This stops the hard disk from powering down and AFAIK it also stops SSDs entering low power states.

 

[johnbl]

your drive came out in 2017, you might want to start with downloading a tool like crystaldiskinfo.exe to read the smart data from the drive and check the drives health and firmware version.

note:
-you can update the firmware for this drive if are just hitting a bug.
-your motherboard bios is 23 versions behind in updates

- I would also make sure there is plenty of free drive space on your drive. As the drive wears out the available space will shrink. if the drive runs out of space the TRIM command will not get enough time to work and the drive will start to time out.
(check the drive health and available disk space)

Note: if you update the bios, you will have to update the cpu chipset drivers also. (the bios update restart can take up to 15 minutes)

removing/deleting temp files and unwanted programs can free up space and extend the life of a drive. You can boot into bios and leave the system idle for a hour or so to give the drive firmware time to do its cleanup.

firmware update to drive might fix problem of not being able to wake. but you might also have to update bios and chipset drivers. You might also be able to just tell windows not to put your drive to sleep. IE go to windows device manager, right click to bring up properties and look for the power management tab and check the box to tell windows not to put the device to sleep.
(option might be on the sata port or drive)

 

[HenryPottier]

Thanks people ! I had someone look into it. Well above my paygrade.