DWM fault in Windows VMs

[DaveC1]

I've been using Workstation Pro without issue for more than a year.  I began having an issue recently when starting to use it on a new host machine.  Windows 10 VMs repeatedly get this exception when interacting (as usual) with their desktop:

Faulting application name: dwm.exe, version: 10.0.19041.746, time stamp: 0x6be51595
Faulting module name: KERNELBASE.dll, version: 10.0.19041.1949, time stamp: 0xcb12e58e
Exception code: 0x8898009b

One difference I notice between old host and new host is that 'Workstation Pro' asked me to enable the 'Windows Hypervisor Platform' feature upon installation on the new host. The old host never had that feature enabled.

Old host: Windows 11 21H2 64-bit (Build 22000.978), Intel(R) Core(TM) i7-8700K CPU @ 3.70GHz, 64 GB RAM
New host: Windows 11 21H2 64-bit (Build 22000.978), 12th Gen Intel(R) Core(TM) i7-12700, 64 GB RAM 

I've seen advice online which suggests editing the VM's Windows boot option, setting a flag called 'UsePlatformClock' = 'yes'.  I have done this on one test VM and it seems to have alleviated the issue. However, the description of this flag is a bit vague, so I'd like to know:

1. WHY do I now have to enable this flag when I never had to do so in the past?  Please let me know if more specific host hardware info is needed to help answer?
2. Could enabling that flag cause impact to the performance of a Windows 10 VM?
3. Do I have any other option(s) to alleviate the issue BESIDES this suggestion?

Thanks for your time,

DaveC

Attachment(s)

vmware-0.log   217 KB 1 version

[james3126]

Hello.

Adding to this thread, as I'm also experiencing these issues on Win10 VMs at work. Repeated crashes of the dwm.exe application, causing my screen to black-out for 2-5 seconds at a time.

This often happens back-to-back 3 times before it'll be okay for a short while. However, sometimes this repeats over and over so many times that Windows sends me back to the login screen almost as if the VM has soft rebooted.

Host and guest are running Windows 10 Pro, 64-bit (Build 19044.2006) 10.0.19044
And the guests are running VMWare Tools Version 12.1.0, build-20219665

Host specs: Intel i7-11800H @ 2.30Ghz - 8c/16t, 32GB DDR4 2667MHz memory, Nvidia RTX 3050Ti Laptop GPU.
Guest 1 specs: 6 cores, 10GB memory, tried with and without 3D acceleration.
Guest 2 specs: 4 cores, 4GB memory, currently with 3D acceleration.

I have also come across this 'UsePlatformClock' option, but it seemed to cause slowdown on my VM? Any help would be greatly appreciated.

Kind regards,
James

[DaveC1]

 

"I have also come across this 'UsePlatformClock' option, but it seemed to cause slowdown on my VM? Any help would be greatly appreciated."

Yeah, I can't provide solid evidence as such (yet), but I do feel as if the two Windows 10 VM guests where I used that option do not perform as well as they did when the option was at its default value of 'no/false'.

[javcz]

Same problem here on two notebooks. I noted you mention NVIDIA ... in my case both notebooks have also NVIDIA card. It seems to be related to display system (well ... it's dwm affected in the end), I can use the guest Win system by letting it run in background and accessing it via RDP. It's first an emergency workaround and also a lead this might the host graphic card driver problem. Tested both with and without 3D acceleration.

[DaveC1]

For what it's worth, I've tested a different workaround as follows...

1. On Windows 10 guest VM(s) I reverted the 'UsePlatformClock' BCD flag back to its default value ("no")

2. Shutdown all guest O/S

3. On the Windows 11 host system I removed the 'Windows Hypervisor Platform' feature

4. Restarted Windows 11 host system and then restarted the Windows 10 guest VM(s)

Since then, I've not seen the DWM fault on any Windows 10 guest VM in the past 48+ hours.  As a side note, I have two Linux VMs on this same host, but have not had any issue with them during all of this (as best I can determine).

I'm not at all confident this solution is any better than the BCD option.  I suspect I'm not supposed to take EITHER of these actions, but I can't live with DWM crashing every 15-30 seconds either.  Later this week I intend to re-enable that feature to see if the issue recurs.

-DaveC

 

[HubVirt]

Hi,

Same problem here since a few days, occurring many times per hour. I noticed dwm was failing each time I had a sound notification in the VM.

Removed the virtual sound card 2 days ago solved the issue. Did not encountered any flickering screen anymore...

Not a 100% satisfying workaround, but at least i can work !

[danial]

I've been having this problem too. When I search for something in Notepad++ and it fails to find it, the screen turns black for a second. I never associated it with a sound problem since I always mute the VM. I disabled the sound in the VM and the problem seems to have gone away. If Notepad++ finds a match then it doesn't play a sound and the screen doesn't black out.

The driver is "High Definition Audio Device" version 10.0.19041.3636 and the date of it is "10/19/2023". I've been having the problem since before October 2023 though.

[xenio]

I list trust in vmware now. Can you recommend any other virtualization technology? Is Hyper V now better? Is there anything I would miss?

[tomscase]

I'm seeing the same across multiple VMs. Including Linux OS's locking up completely so the Windows boot options wont fix those.

Re: Sudden issues across multiple VMs - flickerin ... - VMware Technology Network VMTN

I'll give it a try in the windows vms though

[james3126]

Interesting. I'll spin up one of my Linux VMs and see if I experience the same crashing behaviour there. I notice so far we're all using a host device running Win 10/11.

On 4th Oct 2022 I had update KB5017262 install. Giving that KB number a quick search, I see a few mentions of applications crashing. I'll try uninstalling this from the host, and monitor for further occurrences of the issue.

[james3126]

I've had a flip-flop of successful moments and failure moments over the past month.

The longest period of no-crashes was after removing the Windows Hypervisor Platform feature. However in the past few days the issue has unfortunately returned yet again.

From looking at other posts, it seems doubtful that a fix will be coming after this long with no news - Unless I've missed something.

[SmartCookie]

Try disabling Virtual Machine Platform too if you have it enabled, that fixed it for me again (disabling WHP only fixed it for 1 reboot, strangely enough).

[Geofence]

I've been having this problem for about 1-2 years. Any way to fix it? Have disabled drivers, messed with Vmware settings, reinstalling, etc. I even tried upgrading to Workstation 17 but no permanent fix. After about 10-15min of usage I get screen flicker black and lag until the VM slowly becomes unusable. PLEASE HELP!!! Where is support??

[xenio]

Same still problem here. I have the VMs crashing since more than a year now. I chose vmware initially because it was rock solid. Now I doubt since there is not update on this. I'm running vmware 17.5 on Lenovo P15v Gen 3 with Win 10 22H2 with Window 11 23H2. 

[VMware Community Member]

Same... Been a while now.