DPC Watchdog Error: 0x133_VRF_ISR_nvlddmkm!unknown_function

W

Werest2

Member
Joined
Dec 7, 2024
Posts
20
Hello,

I built a PC in May of this year and haven't had any issues up until late October. I've gotten BSODs primarily while playing Overwatch 2 from anywhere between a few minutes of play to several days. The first instance occurred when playing Kovaak's (iirc) and it's happened once while rendering a project in Adobe AE.

The crashes always occur as my screen freezing up and the audio continuing. Sometimes I'll keep hearing people talking in voice chat, sometimes it will just play sounds on loop. I can move my cursor for about 10 seconds before it disappears. Initially it would go to BSOD after a minute or two with an error message before shutting down. Then after doing some tinkering it no longer goes to blue screen but just cuts all display out and eventually shuts down after making a dump file. Today I had my first crash after enabling a full memory dump and it just didn't shut off after an hour so I shut it off manually.

After booting up without waiting for more than an hour it will only display out via my CPU's integrated graphics and I have to restart from there in order for my GPU to be recognized again.

I updated NVIDIA drivers to no avail. I downloaded AMD chipset drivers. Not getting any helpful information out of my dump files in WinDBG or WhoCrashed. Ran sfc scannow and fixed some corrupted files but that did nothing. Ran Windows Memory Diagnostic and no problems there.

I used DDU to uninstall and reinstall NVIDIA drivers (not sure how well I did this; apologies I'm just a film major and not a computer person). I downloaded MSI Afterburner. I disabled all non-microsoft services in msconfig besides the three NVIDIA drivers and steam and also disabled low-level IO in Afterburner. One of these things might have done something because I was able to play more than 3 games of Overwatch with no crash and then continued to play for about 12 hours over the span of a week with no issue except that I heard myself speak in my Logitech headset whenever it was enabled (presumably because I had disabled LGHUB in msconfig).

Then I got tired of hearing myself speak and accidentally reenabled all the drivers I disabled when trying to just turn on LGHUB and Razer Synapse. The crashes returned and still persist even after going back to my current status of just the 3 NVIDIA services and Steam Client Service. Tried using Driver Verifier via a youtube tut but couldn't find any culprits. I turned on full system memory dump files and stopped playing for a while.

I tried underclocking with MSI today for the first time and set up a core voltage curve that puts me down about 100 mV at ~2800 MHz in hopes that less strain would somehow prevent bug check 133. Crash within 30 minutes of playing. I've never posted on a forum before but at this point I'm at a complete loss of what to do, so thanks for reading my novella.

SYSTEM SPECS

http://speccy.piriform.com/results/TVgS4JV3FYJKJB8QsY7JewB

ProcessorAMD Ryzen 7 7800X3D 8-core Processor, 4201 Mhz, 8 Cores 16 Logical Processors
MotherboardASUS Prime B650-Plus
GPUAsus TUF GAMING OC GeForce RTX 4070 12 GB
RAMG.Skill Flare X5 64 GB (2 x 32 GB) DDR5-6000 CL30 Memory (Slots 2 & 4)
PSUCorsair RM850e 850 W 80+ Gold Modular (used in prior PC, about a year old)
Boot DriveCrucial P3 Plus 2 TB M.2-2280 PCIe 4.0 X4 NVME SSD
OSWindows 10 Home (installed via thumbdrive created by my prior machine)
 

Attachments

Update: crashed again same dpc watchdog violation but this time when i tried to alt tab out of osu. I got a BSOD instead of straight to black and there was a graphical glitch which was similar to something i saw the other day playing Overwatch but with no crash
 

Attachments

  • IMG_2757.webp
    IMG_2757.webp
    1.3 MB · Views: 3
Hello, and welcome to the forum!

The 0x133 bugcheck with an argument 1 value of 0x1 is a special case, it happens when a DPC (or ISR) runs for longer than 2 seconds. All but one of your dumps are of this type. To analyse that we need to be able to see all processors, so we need a kernel dump. Please upload the file %SystemRoot%\Memory.dmp to the cloud with a link to it here - that file contains the kernel dump.

The other dump is a 0x10E VIDEO_MEMORY_MANAGEMENT_INTERNAL bugcheck, which is related to the use of VRAM in the graphics card. The argument 1 value of 0x33 in this dump indicates that 'the device is being destroyed but there are still allocations resident on it'. I'm not 100% certain what that means, but I suspect it means that a graphics process is ending and and error occurs as the VRAM is being cleaned up afterwards. These are the DirectX kernel function calls immediately preceding the bugcheck (you read them from the bottom up)...
Code: 
dxgmms2!VIDMM_PROCESS_ADAPTER_INFO::~VIDMM_PROCESS_ADAPTER_INFO+0x2362d
dxgmms2!VIDMM_PROCESS_ADAPTER_INFO::`scalar deleting destructor'+0xe
dxgmms2!VIDMM_PROCESS::~VIDMM_PROCESS+0x29421
dxgmms2!VidMmTerminateProcess+0x12
dxgkrnl!VidMmTerminateProcessX+0x65
It's interesting that disabling services in MSI Afterburner stopped the BSODs and they returned when you accidentally re-enabled them. Can you try disabling the same services again (record which services you disable) and see whether the BSODs stop again?

In addition, I can see that you've had a few live kernel events, these are errors from which Windows is able to recover. The bugchecks are 0x1B0 (VIDEO_MINIPORT_FAILED_LIVEDUMP) with an exception code (0xC000009A) that indicates that 'insufficient resources were available', which is probably related to the 0x10E BSOD above. There are also a couple of 0x144 (BUGCODE_USB3_DRIVER) with an exception code (0x3003) that indicates that a USB3 device failed enumeration. These are probably unrelated, though it's worth checking your USB3 attached devices and drivers.

We may ask you to enable Driver Verifier again at some point, but we will give you specific instructions on how to do that.

Please change the Complete Memory Dump setting back to Automatic Memory Dump. With a Complete Memory Dump setting your dump files will be the size of installed RAM and none of us will want to download a file that large! They don't provide us with any new information in any case.

Some things that I notice from the Sysnative upload...
  • Several of the dumps show network activity as well as graphics activity. The issue could potentially be networking related.
  • I also see that you're using a Class B IP4 address. That's most unusual for a home PC, is this a business computer?
  • Several dumps also show a WSD scanner and printer driver unloading immediately before the BSOD. These are Web Services for Devices drivers typically used with network attached scanners and printers. I have no idea whether this is significant but it is worth noting.
  • All but one of the dumps show BlueStack drivers unloading immediately before the BSOD. It would be useful if it were possible to temporarily uninstall BlueStacks, or at least prevent any of the drivers being loaded. I've seen BlueStacks causing problems before.
  • Several of the dumps show the processor coming out of the idle (low-power) C-State. Some CPUs are flaky transitioning from low-power to high-power (see below).
  • The RAM that you are using is not on the QVL of the motherboard. That doesn't mean that it won't work, but it's always a concern when you are getting BSODs. That's just something to keep in mind if we don't find any other issues.
The potential C-State issue can be tested for (and mitigated) by disabling all C-States for all processors in the BIOS. This stops the processors entering a low-power state when they become idle and so the potential power transition issue is avoided when they come out of idle to do work. The only downsides to this are that the CPU will generate a tad more heat in the idle state, but a decent cooler should handle this (what CPU cooler do you have?). It will also consume a tad more power when the processors are idle, but it's probably not measurable.

As well as uploading the kernel dump then, try disabling C-States. Uninstall BlueStacks (temporarily) if you can, or at least completely deactivate it. Also check your networking drivers for updates, especially those for the printer/scanner. Oh, and change the dump file type!
 
Last edited:
Hi Ubuysa, thanks for the response!

I've just changed the dump file back to automatic - I'll upload a memory.dmp that's not 64gb the next time I crash.

After my crash last night playing Osu my headset somehow returned to a state of feedback and I was able to play Overwatch for about 4 hours with no crash and have not crashed yet since then. Before playing I downloaded Adrenaline and set my msconfig to as follows:

1733700246713.webp1733700274263.webp

And this is the only setting I changed in MSI from on to off - it's still at this setting.

1733700715246.webp

As for my IP4 address I live on campus so I'm connected to Student-Wireless with some $20 WiFi adapter plugged into one of my USB3 ports.

I don't think I've touched any printing networking drivers but I do have a printer that I connect to via a different WiFi so I'll try that.

I'll uninstall Bluestacks for now 😔

I have a Thermalright Peerless Assassin and don't think I have an issue with overheating so I'll disable C-states.

And as for my RAM, I didn't know what a QVL was until 10 minutes ago and simply trusted PCpartpicker (whoops).

Thanks again, I'll do all this and play some Overwatch and hopefully I won't have a dump file for you to look at in the near future!
 
Also, in terms of USB3 Peripherals I have:

- Razer Black Widow Keyboard
- Endgame Gear OP1 RGB mouse
- Logitech G635 headset
- Aforementioned Realtek RTL8811AU Wireless LAN 802.11ac USB 2.0 Network Adapter
( at least that's what it displays as in Device manager )

As well as multiple external drives which I edit film projects on ranging from LaCie to WD to Samsung, although crashes occur even when none of these are present (none currently are).
 
Crashed today. 20 minutes of looking at a frozen screen was not enough to make a dump file apparently. Happened while I was AFK for 3 minutes with a twitch stream open after playing Overwatch for several hours with no issues. Switched from "Automatic Memory Dump" to "Kernel Memory Dump" since it seems like my PC was trying to create another full memory dump.
 
It's worth running the Sysnative data collector app again because it picks up a lot more than just dump files.

The difference between 'Automatic memory dump' and 'Kernel memory dump' is minimal. Both generate a kernel dump, however the 'automatic memory dump' option gives Windows more leeway in increasing the size of the pagefile when the dump is written (the dump is written in the first instance to the paging file). You generally only get a kernel (or automatic) memory dump when you have had a BSOD. If it's just a frozen screen you may not get a kernel/automatic dump.

Please navigate to the folder C:\Windows\LiveKernelReports, in there you may find other sub-folders, please check in each of these sub-folders for dump files and upload all that you find to a cloud service with a link to them here. This folder (and its sub-folders) contain live dumps that were taken when a problem occurred from which Windows was able to recover, so any dumps in here will be useful.
 
The dumps since Sunday are all the same, they're all failing on the ISR (interrupt service routine) of the graphics driver (nvlddmkm.sys)...
Code: 
FAILURE_BUCKET_ID:  0x133_ISR_nvlddmkm!unknown_function

It's probably not related to BlueStacks then, but if you can leave it uninstalled until the issue is solved.

I am concerned about those Afterburner changes you made. I'd really like you to restore anything you've changed with Afterburner and restore everything to default values. I'd prefer Afterburner to be uninstalled too.

The live kernel dumps also point at the graphics card/driver. You've had 13 live kernel events since July, all of them graphics card/driver related. The bugchecks vary from a TDR timeout attempting to recover a graphics hang, through a video engine timeout detected (because the graphics card/driver failed to respond), and more recently they are graphics miniport failures to start the device. There is very clearly a graphics card and/or graphics driver problem there and it wouldn't surprise me to find that Afterburner is behind it all.

I think you need to get everything back to stock/default settings, especially where the graphics card/driver are concerned. Another worry is that since the live kernel events seem to be increasing in severity it may be that you graphics card is failing?
 
I've just uninstalled Afterburner but I don't think it's the cause because I installed it long after the crashes began occurring.

I have the Dec 5 game-ready driver from Nvidia and the only non-stock change currently active that I'm aware of is the disabling of Adobe Update Services in System Configuration.

I'd like to hope my GPU isn't failing but IDK. It's very variable in the frequency of crashes (3+ a day last week to now 1 within the last 4 days) and the one crash I did have was when it was basically idle but again I have no clue. I bought it used in great condition so maybe it only had 6 months of life left somehow.
 
Ran Furmark 2 for a couple minutes and nothing happened. Avg GPU temp at 97% utilization was between 60-70C and hotspot up to ~92C.

Ran the P1080 benchmark and had the usual freeze occur within about a second, this time with an actual BSOD.

Here's an updated collector file and a link to my MEMORY.dmp.

Werest2_MEMORY.zip
 

Attachments

Just double clicked to fullscreen Steam and my PC crashed straight to blackscreen. Opened WinDBG to check the dump file and crashed to blue screen. Disconnected my wi-fi adapter and ran P1080 benchmark on startup and it completed with no issue. Ran it again and crashed at 11%. During both tests I accidentally had my main display plugged into my integrated graphics rather than gpu, not sure what exactly this entails.
 
Thanks, i’ll give it a read if i’m able to boot my pc again. Was going to try and move my gpu to a different pcie slot but found that the latch on pciex16_1 (my gpu) was in the open position.
 
That AMD CPU has an iGPU, so remove the RTX 4070 and plug just one monitor the motherboard (iGPU) port. Then run Furmark again on that iGPU and see whether things are stable.
 
Ran furmark with my igpu yesterday and it was stable at like 10fps but I'll run it as specified once I have time to disassemble my pc again (busy today). Once I take out my GPU I'm going to try reinstalling it in the other PCIE slot. Also, I disabled EXPO in bios and furmark still caused bluescreens.
 
I think it's important to verify that it's fully stable without the graphics card, then you'll know exactly where the problem lies.
 
Seller offered to refund me for the card. Given everything you've seen do you think I should take them up on their offer or keep troubleshooting?
 

Has Sysnative Forums helped you? Please consider donating to help us support the site!

Back
Top