Question Frequent crashes and freezes

[RealMiel]

Hello everyone,

Since a couple of weeks my pc has been crashing almost daily.
When I am playing a game the screen freezes and the sound goes on for a little bit. After that I get a bluescreen. I can’t find out what the issue is.

A couple of months ago I also had a issue that kept crashing my pc. But after getting a new PSU the crashes where gone.
I already tested my memory, reset drivers and updated drivers.
I know my PSU is strong enough for my pc but is it possible that it is faulty again?

The link below has a folder with all the information I have. Component temps, a minidump file and my event logs.
https://drive.google.com/drive/folders/1DqflY3E31SctkK2h2QveTxuT0qThndDs

The minidump says something about ntkrnlmp.exe and nt!KeAccumulateTicks+1d5510

PC Specs are:
CPU: AMD Ryzen 5 3600
GPU: NVIDIA GeForce GTX 1660 Ti
Motherboard: MSI b450 Tomahawk
RAM: Corsair CMK16GX4M2B3000C15 DDR4 x2
PSU: Corsair TX-M Series TX650M V2
HDD: Toshiba P300 1TB,
HDD: Seagate BarraCuda 2TB
SSD: Hyperx Savage 480GB

Apologies if I missed anything. If you need me to add anything to this post just ask!
I hope someone can help me. Thank you!

 

[rgd1101]

full system spec? include brand and model of the psu

 

[RealMiel]

full system spec? include brand and model of the psu

CPU: AMD Ryzen 5 3600
GPU: NVIDIA GeForce GTX 1660 Ti
Motherboard: MSI b450 Tomahawk
RAM: Corsair CMK16GX4M2B3000C15 DDR4 x2
PSU: Corsair TX-M Series TX650M V2
HDD: Toshiba P300 1TB,
HDD: Seagate BarraCuda 2TB
SSD: Hyperx Savage 480GB

 

[johnbl]

graphic driver stopped or slowly processed deferred procedure calls
the backup got too long and a bugcheck was called
DPC_TIMEOUT_TYPE: DPC_QUEUE_EXECUTION_TIMEOUT_EXCEEDED
I do not see a direct cause of the problem:
maybe your android hypervisor emulator? gvm.sys / something with your vm?
over heated graphics card? PCI bus not running at 100MHz ?

notes:
cpu micrcode patches not installed: (microsoft amd cpu micrcode dll not loaded)
(mcupdate.dll was loaded)

0: kd> !sysinfo cpumicrocode
Initial Microcode Version: 00000000:08701021
Cached Microcode Version: 00000000:00000000
Processor Family: 17
Processor Model: 71
Processor Stepping: 00
Note: Cached Microcode Version is not detected!


ScpVBus.sys Sun May 5 14:31:26 2013 (causes random kernel corruption)
Hamdrv.sys Mon Mar 30 06:28:42 2015 (a lot of times left over after uninstall)
NTIOLib_X64. (overclock driver)

Oculus drivers installed
Oculus_ViGEmBus.sys Wed Aug 2 16:42:02 2017
OCULUSUD.sys Mon Apr 22 18:29:45 2019
oculusvad.sys Thu Feb 11 13:03:49 2021

nvhda64v.sys Mon Jan 3 02:46:34 2022
nvlddmkm.sys Wed May 4 13:32:29 2022 <-- driver with problem
nvvad64v.sys Mon Sep 13 00:32:00 2021 (NVIDIA Virtual Audio Driver )
nvvhci.sys Fri Jan 10 10:31:07 2020 (nvidia Virtual USB Host Controller driver )


BIOS Version 1.H7
BIOS Starting Address Segment f000
BIOS Release Date 01/18/2022
Manufacturer Micro-Star International Co., Ltd
Product B450 TOMAHAWK (MS-7C02)


Processor Version AMD Ryzen 5 3600 6-Core Processor
Processor Voltage 8bh - 1.1V
External Clock 100MHz
Max Speed 4200MHz
Current Speed 3600MHz



-----------
cpu 0 had 640 stacks waiting for the gpu to process.
most of the stacks looked like this:
[Stack 625] Elapsed time since previous stack: 46 mSec; since reference stack: 9765 mSec
Exiting function: fffff80693defea9 nvlddmkm+0x8FEA9
# RetAddr Call Site
05 fffff80693e001f3 nt!KzLowerIrql
06 fffff80693e01a51 nvlddmkm+0xA01F3
07 fffff80693df006a nvlddmkm+0xA1A51
08 fffff806800a181e nvlddmkm+0x9006A
09 fffff806800a0b04 nt!KiExecuteAllDpcs+0x30E
0a fffff806801fb8ce nt!KiRetireDpcList+0x1F4
0b ---------------- nt!KiIdleLoop+0x9E

DPC Watchdog capture size: 641 stacks.
Number of unique stacks: 339.

--------------

 

[RealMiel]

graphic driver stopped or slowly processed deferred procedure calls
the backup got too long and a bugcheck was called
DPC_TIMEOUT_TYPE: DPC_QUEUE_EXECUTION_TIMEOUT_EXCEEDED
I do not see a direct cause of the problem:
maybe your android hypervisor emulator? gvm.sys / something with your vm?
over heated graphics card? PCI bus not running at 100MHz ?

notes:
cpu micrcode patches not installed: (microsoft amd cpu micrcode dll not loaded)
(mcupdate.dll was loaded)

0: kd> !sysinfo cpumicrocode
Initial Microcode Version: 00000000:08701021
Cached Microcode Version: 00000000:00000000
Processor Family: 17
Processor Model: 71
Processor Stepping: 00
Note: Cached Microcode Version is not detected!


ScpVBus.sys Sun May 5 14:31:26 2013 (causes random kernel corruption)
Hamdrv.sys Mon Mar 30 06:28:42 2015 (a lot of times left over after uninstall)
NTIOLib_X64. (overclock driver)

Oculus drivers installed
Oculus_ViGEmBus.sys Wed Aug 2 16:42:02 2017
OCULUSUD.sys Mon Apr 22 18:29:45 2019
oculusvad.sys Thu Feb 11 13:03:49 2021

nvhda64v.sys Mon Jan 3 02:46:34 2022
nvlddmkm.sys Wed May 4 13:32:29 2022 <-- driver with problem
nvvad64v.sys Mon Sep 13 00:32:00 2021 (NVIDIA Virtual Audio Driver )
nvvhci.sys Fri Jan 10 10:31:07 2020 (nvidia Virtual USB Host Controller driver )


BIOS Version 1.H7
BIOS Starting Address Segment f000
BIOS Release Date 01/18/2022
Manufacturer Micro-Star International Co., Ltd
Product B450 TOMAHAWK (MS-7C02)


Processor Version AMD Ryzen 5 3600 6-Core Processor
Processor Voltage 8bh - 1.1V
External Clock 100MHz
Max Speed 4200MHz
Current Speed 3600MHz



-----------
cpu 0 had 640 stacks waiting for the gpu to process.
most of the stacks looked like this:
[Stack 625] Elapsed time since previous stack: 46 mSec; since reference stack: 9765 mSec
Exiting function: fffff80693defea9 nvlddmkm+0x8FEA9
# RetAddr Call Site
05 fffff80693e001f3 nt!KzLowerIrql
06 fffff80693e01a51 nvlddmkm+0xA01F3
07 fffff80693df006a nvlddmkm+0xA1A51
08 fffff806800a181e nvlddmkm+0x9006A
09 fffff806800a0b04 nt!KiExecuteAllDpcs+0x30E
0a fffff806801fb8ce nt!KiRetireDpcList+0x1F4
0b ---------------- nt!KiIdleLoop+0x9E

DPC Watchdog capture size: 641 stacks.
Number of unique stacks: 339.

--------------

Do you think this issue is hardware related? And do you know a good way to test if it is?
I still have a new PSU laying around so i can try replacing my current PSU with that one.

 

[johnbl]

Do you think this issue is hardware related? And do you know a good way to test if it is?
I still have a new PSU laying around so i can try replacing my current PSU with that one.

I would start by removing the android emulator and see if the problem goes away. If i remember correctly your system was up for over 11 hrs before the crash. the problem could be slow and continuous build up of the calls. or some bug occurred and all the stacks built up quickly.
only way to know is to do a memory dump on a working system and check to see if there is a queue of stacks to process. to do this you google how to force a memory dump with a keyboard, make the registry change and force a memory dump on a working system, load the dump into the debugger.

you could also roll back the graphic driver if something changed.
most likely going to be something related to the vm android emulator since we are not seeing a lot of people with this problem (yet)

if you are playing a game in the vm android emulator, you might change the graphics mode to a common setting.
ie full screen 1080. often the defaults work the best since they get more testing

if you change the memory dump to kernel, and get a dump I might be able to figure out the owner of the calls.

bunch of security updates for amd cpu, keep looking for bios updates since microsoft will start pushing out updated files for amd and these might break a emulator or crash your system. some vendors already put out the updates a week ago.

not likely to be a power supply issue. most likely a driver issue. (or driver is being starved of gpu time and getting behind on its work to the point a bugcheck is called)

 

[RealMiel]

I would start by removing the android emulator and see if the problem goes away. If i remember correctly your system was up for over 11 hrs before the crash. the problem could be slow and continuous build up of the calls. or some bug occurred and all the stacks built up quickly.
only way to know is to do a memory dump on a working system and check to see if there is a queue of stacks to process. to do this you google how to force a memory dump with a keyboard, make the registry change and force a memory dump on a working system, load the dump into the debugger.

you could also roll back the graphic driver if something changed.
most likely going to be something related to the vm android emulator since we are not seeing a lot of people with this problem (yet)

if you are playing a game in the vm android emulator, you might change the graphics mode to a common setting.
ie full screen 1080. often the defaults work the best since they get more testing

if you change the memory dump to kernel, and get a dump I might be able to figure out the owner of the calls.

bunch of security updates for amd cpu, keep looking for bios updates since microsoft will start pushing out updated files for amd and these might break a emulator or crash your system. some vendors already put out the updates a week ago.

not likely to be a power supply issue. most likely a driver issue. (or driver is being starved of cpu time and getting behind on its work to the point a bugcheck is called)

I created a kernel dump and you can download it on the link below.
https://drive.google.com/file/d/1MWWujrY_8PCIRAZcwITnhHIw95m8BLPO/view?usp=sharing

I'm gonna try out some of the things you said in the last post.

Thanks for the help btw

 

[johnbl]

could just be some of the vm software does not understand the sleep functions windows now has. I would disable sleep so windows does not sleep devices that VM is using and does not know how to wake up. (usb devices)
------------
from looking at the dump, there is some issue with sleep functions causing problems with some devices:
you will want to tell your system not to put your usb devices to sleep. windows control panel, device manager, right mouse click on usb hubs and see if there is a power management tab, see if you can tell windows not to all windows to put the device to sleep to save power.
I would look at all of the usb device tree and the hid device tree. (or put the whole system in to high performance mode)

guessing Shinko Shoji Co

device is the one that put the 䯀 as the owner of a bunch of the DPC
when I translate the character to english it comes out as
"Yikes" maybe google translate does not know the word?

I would just set the system to not allow any usb hub to sleep reboot and see if it helps.
-------------
seems to be issues in the dump.
system up for 51 minutes
system process using a lot of virtual memory.
large amount of info paged to disk. I would turn of system virtual memory, reboot and turn it back on to dump old pagefile.sys

kind of a unexpected file name
[Image]ImageLoad of 䯀

will look in dump logs to see what i can find.
-----------
debugger thinks your
nt kernel has been modified
USBXHCI.sys modified
HIDCLASS.sys driver modified

lots of stuff paged out to disk so it can not be looked at.

i would remove as many suspect drivers as you can.
I would also run microsoft process explorer and take a look at what the system process is doing and why it is using so much virtual memory.

or just start removing software you don't need and reboot and see run resource monitor and see if the system starts eating up a bunch of virtual memory.

try and make another kernel dump after doing driver cleanup (remove suspect drivers)

something not right with usb subsytem. not sure if it is related to vm. will see what the usb driver logs indicate.

usb device vendor id 198c device id 0003
vendor = c-scape

several of your usb devices are waiting for a wake up signal
this includes a network driver rt640x64.sys
also some device
it has a log entry:
Reporting no runtime wake capability for device (b:28 d:0 f:4) because the BIOS left Express in Compatibility mode

vendor 1022 device id 43c7
listed as from
Shinko Shoji Co., Ltd