Question Getting BSOD in heavy games ?

shaheer_2

Honorable
Aug 15, 2016
18
0
10,510
BSOD ( MEMORY_MANAGEMENT ntoskrnl.exe )ERROR PIC getting after playing 5 mins in heavy games like RDR2 or cyberpunk and didnt happen in small games like dead by daylight i played for 2 hours no error. i was getting NO BSOD when i had RTX 2080ti but i purchased this RTX 3070 and now getting errors.
I installed drivers using DDU in safe mode and also installed old 3070 driver but no help

Specs:
i5-11400f Temp never crosses 60c
B560 gaming HD ( Latest bios )
32gb xpg D50 ( brand new rams)
RTX 3070 ( i also tried another 3070 getting same error ) Temp never crosses 70c
PSU: xpg pylon 650w

I used commands like sfc /scannow and DISM.exe /Online /Cleanup-Image /Restorehealth but didnt help.
 
BSOD ( MEMORY_MANAGEMENT ntoskrnl.exe )ERROR PIC getting after playing 5 mins in heavy games like RDR2 or cyberpunk and didnt happen in small games like dead by daylight i played for 2 hours no error. i was getting NO BSOD when i had RTX 2080ti but i purchased this and getting errors. i installed drivers using DDU in safe mode and also installed old 3070 driver but no help
Specs: i5 11400f Temp never crosses 60c
B560 gaming HD ( Latest bios )
32gb xpg D50 ( brand new rams)
RTX 3070 ( i also tried another 3070 getting same error ) Temp never crosses 70c
psu: xpg pylon 650w
i used commands like sfc /scannow and DISM.exe /Online /Cleanup-Image /Restorehealth but didnt help.
Your psu is a little iffy to run a 3070.

Shop for a 750w unit with at least a 7 year warranty.
 
Looking at the BlueScreenView output the BSOD on 23rd March is the key one, mainly because it's a Driver Verifier Detected error. The Razor driver RZDev_025d.sys was the one trapped by Driver Verifier so that's the one you should probably focus on.

FWIW ntoskrnl.sys is quite often flagged by both BlueScreenView and WinDbg analyze -v, but it's never the real problem. Ntoskrnl.sys is the Windows kernel, it will call a third-party driver, which will do its thing and then return to the kernel. It's when the kernel has control back and checks the parameters passed by the third-party driver that an error (by the third-party driver) is typically detected, so the resulting BSOD points at ntoskrnl.sys....

Without actual dumps it's not possible to say any more.
 
Looking at the BlueScreenView output the BSOD on 23rd March is the key one, mainly because it's a Driver Verifier Detected error. The Razor driver RZDev_025d.sys was the one trapped by Driver Verifier so that's the one you should probably focus on.

FWIW ntoskrnl.sys is quite often flagged by both BlueScreenView and WinDbg analyze -v, but it's never the real problem. Ntoskrnl.sys is the Windows kernel, it will call a third-party driver, which will do its thing and then return to the kernel. It's when the kernel has control back and checks the parameters passed by the third-party driver that an error (by the third-party driver) is typically detected, so the resulting BSOD points at ntoskrnl.sys....

Without actual dumps it's not possible to say any more.
Here is three dump files https://drive.google.com/drive/folders/1wOBwIhJrF0GskVGijzeoDdxHzzQ3Q1G4?usp=sharing
 
The 032623-8187-01.dmp is clear-cut, it's a DRIVER_VERIFIER_DETECTED_VIOLATION showing a code integrity issue, meaning that a driver tried to access the wrong page type (executable instead of non-executable). This is clearly a driver error.

The call stack shows a call to Rz_Dev.sys...
Code:
1: kd> knL
# Child-SP          RetAddr               Call Site
00 fffff40a`908fe578 fffff803`5cd866e3     nt!KeBugCheckEx
01 fffff40a`908fe580 fffff803`5c74072b     nt!VerifierBugCheckIfAppropriate+0xdf
02 fffff40a`908fe5c0 fffff803`5cd7db98     nt!VfReportIssueWithOptions+0x103
03 fffff40a`908fe610 fffff803`5c73f212     nt!VfCheckPoolType+0x90
04 fffff40a`908fe650 fffff803`5d818615     nt!VfCheckNxPoolType+0x12
05 (Inline Function) --------`--------     Wdf01000!FxVerifierCheckNxPoolType+0x27
06 fffff40a`908fe680 fffff804`cd8224cf     Wdf01000!imp_WdfDeviceAllocAndQueryProperty+0x95
07 fffff40a`908fe6e0 ffff9b02`03e8c140     RzDev_025d+0x24cf
08 fffff40a`908fe6e8 00000000`00000000     0xffff9b02`03e8c140
Argument 2 of the bugcheck (0xFFFFF804CD8224CF) is the address in the driver where the failure occurred, identifying the module name confirms that the problem happened in RzDev_025d.sys...
Code:
1: kd> lmDva 0xfffff804`cd8224cf
Browse full module list
start             end                 module name
fffff804`cd820000 fffff804`cd830000   RzDev_025d T (no symbols)         
    Loaded symbol image file: RzDev_025d.sys
    Image path: \SystemRoot\System32\drivers\RzDev_025d.sys
    Image name: RzDev_025d.sys
    Browse all global symbols  functions  data
    Timestamp:        Tue Aug 18 07:00:27 2020 (5F3B525B)
    CheckSum:         0001245D
    ImageSize:        00010000
    Translations:     0000.04b0 0000.04e4 0409.04b0 0409.04e4
    Information from resource tables:
There may be a more recent update for this Raxer (mouse?) driver? This is certainly a flaky driver and it does need either updating or removing/replacing.

The other two dumps are very similar, they are both MEMORY_MANAGEMENT bugchecks with an exception code indicating a page table problem. Argument 2 in each dump provides the virtual address where the bugcheck happened, by displaying the virtual to real address mapping for that address we see that there is indeed a problem...
Code:
8: kd> !vtop 0 0000000055b9c32a
Amd64VtoP: Virt 0000000055b9c32a, pagedir 000000024239a000
Amd64VtoP: PML4E 000000024239a000
Amd64VtoP: PML4E read error 0x8000FFFF
Virtual address 55b9c32a translation fails, error 0x8000FFFF.

This could be a RAM problem as you suspected, but it could also be a rogue driver. Both these dumps were performing graphics operations, the DirectX kernel driver dxgkrnl.sys is called in both. Whilst there is no evidence of the Razer driver RzDev_025d.sys being called in these dumps I don't believe in coincidences and I think it reasonable to lay these two BSODs at the door of RzDev_025d.sys.