BSOD with new laptop

[SOG261]

Hey guys.

Been using the new Lenovo Legion 7i laptop but every now and again it crashes with a blue screen or does a semi crash where I can hear audio but can't do nothing else.

I'm not exactly sure why, sometimes the bluescreen references a "VIDEO_MEMORY_MANAGEMENT_INTERNAL" error. The app "who crashed" says the fault of the error is the igdkmdn64.sys driver, but I'm not too sure what to do with that info.

I've attached the synative file collection folder.

I am overclocking/undervolting, overclocking with the 7i can be done natively in the bios with mild oc (+150 core, +300 mem), and undervolting settings (-60mV core/cache, -12.7mV integrated gpu). Other people with the exact same config have been doing similar or more demanding overclocks/undervolts, so I doubt it's related to this (unless I'm just really unlucky with the silicon).

 

[x BlueRobot]

Please restore your BIOS settings to their factory defaults while we troubleshoot the issue.

12: kd> lmvm igdkmdn64
Browse full module list
start             end                 module name
fffff801`34570000 fffff801`3618b000   igdkmdn64 T (no symbols)           
    Loaded symbol image file: igdkmdn64.sys
    Image path: \SystemRoot\System32\DriverStore\FileRepository\iigd_dch.inf_amd64_fde4b3d3e5046b52\igdkmdn64.sys
    Image name: igdkmdn64.sys
    Browse all global symbols  functions  data
    Timestamp:        Fri Jul 23 18:52:19 2021 (60FB01D3)
    CheckSum:         01C25949
    ImageSize:        01C1B000
    Translations:     0000.04b0 0000.04e4 0409.04b0 0409.04e4
    Information from resource tables:

Please update (if available) your Intel graphics card driver from the Intel support page - Intel® Driver & Support Assistant

 

[SOG261]

Please restore your BIOS settings to their factory defaults while we troubleshoot the issue.

Thanks for your quick reply.

All I changed was the OC settings & flip to boot. Should I disable those?

I'll update my intel gpu, but usually I run with the nvidia gpu on at all times (mux switch) so would that still be a problem?

 

[x BlueRobot]

It should be fine, the Intel driver was the one causing the problem, I presume that it may not immediately use the nVidia card.

All I changed was the OC settings & flip to boot. Should I disable those?

Yes, set them back to how they were beforehand, at least for troubleshooting purposes. If the system seems to be stable after updating the driver, then you can enable them again if you wish.

 

[SOG261]

Just crashed again, this was after updaing the intel drivers (it was only the wifi/bluetooth drivers) and also disabling any bios optimisations. It seems that there could be a fault with the nvidia gpu but it's really strange because this is a brand new laptop so I don't understand how that could be.

One thing to note is that I did update the nvidia gpu's oem driver to the more updated GFE drivers, but I'm not sure if that would cause such an issue.

The error code is nearly always "video memory internal management" "0x10E". I'll upload my memory.dump file from the most recent crash in a new reply.

 

[x BlueRobot]

It seems that there could be a fault with the nvidia gpu but it's really strange because this is a brand new laptop so I don't understand how that could be.

Only because something is new doesn't mean it doesn't have any defects. Please upload the new dump file(s) and I'll have a look.

 

[SOG261]

Only because something is new doesn't mean it doesn't have any defects. Please upload the new dump file(s) and I'll have a look.

MEMORY.DMP

Appreciate you looking. I didn't mean that just because it's new it shouldn't have any defects, said it more as context.

 

[zbook]

Please uninstall and reinstall the Nvidia GPU driver.

Uninstall using DDU:
Display Driver Uninstaller Download version 18.0.4.3 (or newer version if available)
Wagnardsoft

Reinstall using the Lenovo website:
Nvidia VGA Driver for Windows 10 (64-bit) - Legion 5, Legion 5 Pro, Legion 7 - Lenovo Support UY


Some of the dump files for the past few months were typical of overclocking or malfunctioning hardware.

9C
124
101


New laptops without overclocks should not have BSOD.

Had you considered Return Merchandise Authorization (RMA)?


Was the latest BSOD before or after resetting the BIOS?

 

[SOG261]

Please uninstall and reinstall the Nvidia GPU driver.

Uninstall using DDU:
Display Driver Uninstaller Download version 18.0.4.3 (or newer version if available)
Wagnardsoft

Reinstall using the Lenovo website:
Nvidia VGA Driver for Windows 10 (64-bit) - Legion 5, Legion 5 Pro, Legion 7 - Lenovo Support UY


Some of the dump files for the past few months were typical of overclocking or malfunctioning hardware.

9C
124
101


New laptops without overclocks should not have BSOD.

Had you considered Return Merchandise Authorization (RMA)?


Was the latest BSOD before or after resetting the BIOS?

I reinstalled my Nvidia GPU driver using DDU yesterday after the BSOD, this time not installing gfe. Will see if I have any more blue screens. If I do then I'll RMA.

The overclocking is enabled as a "safe" OC in the bios. Any OC gets reset to 0 whenever the laptop enters quiet mode, which is what I have on sometimes and it still crashed yesterday through that, so I don't think it's the OCs.

There is a new intel gpu release (30.0.100.9955) but I'm not sure if I should update.
Same with the bios, new update recently but I haven't updated.

 

[x BlueRobot]

I haven't been able to find much information about Stop 0x10E since most of the Direct X kernel and WDDM is undocumented. The MSDN documentation mentions that the bugcheck is commonly caused by graphics card drivers.

VIDEO_MEMORY_MANAGEMENT_INTERNAL (10e)
The video memory manager encountered a condition that it can't recover from. By crashing,
the video memory manager is attempting to get enough information into the minidump such that
somebody can pinpoint what lead to this condition.
Arguments:
Arg1: 0000000000000017, Unexpected system command failure.
Arg2: ffffe50e41f9c000
Arg3: 0000000000000000
Arg4: ffffffffc00002b6

The error code is always related to a device being removed and I assume that it's referring to the adapter being released as shown in the call stack.

12: kd> !error c00002b6
Error code: (NTSTATUS) 0xc00002b6 (3221226166) - The device has been removed.

12: kd> knL
 # Child-SP          RetAddr           Call Site
00 fffff901`a73a72f8 fffff806`92493ad0 nt!KeBugCheckEx
01 fffff901`a73a7300 fffff806`9381443f watchdog!WdLogEvent5_WdCriticalError+0xe0
02 fffff901`a73a7340 fffff806`93782949 dxgmms2!VIDMM_GLOBAL::RestoreFromPurge+0xbb << Exception is caught here
03 fffff901`a73a73d0 fffff806`922548e2 dxgmms2!VidMmRestoreFromPurge+0x9
04 fffff901`a73a7400 fffff806`921ff44a dxgkrnl!ADAPTER_RENDER::RestoreFromPurgeSegments+0x36
05 fffff901`a73a7430 fffff806`921ff379 dxgkrnl!DXGADAPTER::ReleaseCoreSync+0xa6
06 fffff901`a73a7460 fffff806`92276a17 dxgkrnl!DxgkReleaseAdapterCoreSync+0x4d << Crash here?
07 fffff901`a73a74b0 fffff806`7a555855 dxgkrnl!DpiPowerArbiterThread+0xf7
08 fffff901`a73a7510 fffff806`7a5fe808 nt!PspSystemThreadStartup+0x55
09 fffff901`a73a7560 00000000`00000000 nt!KiStartSystemThread+0x28

I know, the second parameter refers to the video memory manager global state object, which is again, unfortunately undocumented.

12: kd> !pool ffffe50e41f9c000
Pool page ffffe50e41f9c000 region is Nonpaged pool
*ffffe50e41f9c000 : large page allocation, tag is Vi15, size is 0xaec0 bytes
        Pooltag Vi15 : Video memory manager global state, Binary : dxgmms2.sys

08 fffff901a73a7510 fffff8067a5fe808 nt!PspSystemThreadStartup+55 
    Parameter[0] = ffffe50e3ade1040 << Our current thread
    Parameter[1] = ffffe50e3ae49030 << nVidia graphics card device object
    Parameter[2] = (unknown)       
    Parameter[3] = (unknown)

12: kd> !devobj ffffe50e3ae49030
Device object (ffffe50e3ae49030) is for:
  \Driver\nvlddmkm DriverObject ffffe50e3ab84e10
Current Irp 00000000 RefCount 0 Type 00000023 Flags 00002004
SecurityDescriptor ffffb405d1bc2d60 DevExt ffffe50e3ae49180 DevObjExt ffffe50e3ae4a8b8 
ExtensionFlags (0000000000)  
Characteristics (0x00000100)  FILE_DEVICE_SECURE_OPEN
AttachedTo (Lower) ffffe50e34581e00 \Driver\ACPI
Device queue is not busy.

Hmm, I'm wondering if it is the nVidia graphics card driver at fault here? Let's see if you have any more crashes since you've reinstalled the driver.

 

[SOG261]

That's a pretty indepth reply, thanks for taking so much time investigating. I'm sure it's like going on a wild goose chase at times...

I've not experienced any crashing so far since updating, I'm wondering if it's a mix of advanced optimus, multiplexer, intel/nvidia all somehow tangling together resulting in the crashes.

We'll see anyhow. Thanks again for all the help.

 

[x BlueRobot]

That's good news! Hopefully you won't experience any more crashes.