virtualbox.org

HELP, I GIVE UP!
For the last couple of weeks (after installing ver 6.1.3), I've been wrestling with a Win10 Pro 32-bit Guest running an old vb6 sp5 (duh, vb=visual basic) program.
Program has run flawlessly for nigh 20 years, now machine goes into full reboot mode.
No idea which log file would contain the required abend info, so |I'm including all.
Thanks in advance for any & all help!

Please provide a VM log file. Make sure the VM is fully shut down, then right click it in the manager UI. Select "Show Log" and save "VBox.log" (no other file) to a zip file. Attach the zip here.

Didn't I already send you "vbox.log" from my 'logs' folder (unzipped)?
Well, sending as instructed....
And THANKS!

I don't see any sign of an error being registered by the VM. The VM starts, the GAs load after 25s, at 16 minutes I see a reboot followed quickly by a shutdown. No obvious incidents. The heartbeat register does not detect a hung VM either.

If you have a software crash inside the guest OS then what makes you believe VirtualBox has something to do with it?

The VM OS BSODed:

00:15:54.889322 GIM: HyperV: Guest indicates a fatal condition! P0=0x50 P1=0xffffffe8 P2=0x2 P3=0x84a72ee1 P4=0x2
00:15:54.889370 GIMHv: BugCheck 50 {ffffffe8, 2, 84a72ee1, 2}

Host, EP, & GAs are all on 6.1.30.

RetiredGuy wrote:after installing ver 6.1.3

Must be something new to the users, there's another poster who insists they're running 6.1.3, also, when it's probably 6.1.30.

There is a disparity in the VM template, see the VM's General settings, Basic tab, Version: dropdown, which does affect some settings deeper in the VM's needs:

00:00:03.598890 Guest OS type: 'Windows7'
00:00:41.662478 VMMDev: Guest Log: 20:06:26.152972 main OS Product: Windows 10
00:00:41.662511 VMMDev: Guest Log: 20:06:26.152972 main OS Release: 10.0.19044

I'm not certain why this is happening. If you revert to the Virtualbox, EP, and GAs you had before, does the VM run well again?

Good catch. I looked for "Hyper" and "Hyper-V", I wasn't aware I should be searching for "HyperV"!

Bugcheck 0x50 apparantly means "page fault in non-paged area", so a memory error of some kind. Previous reports seem to have been related to superfluous third party AV installed on the host, so perhaps something outside is messing with the VM.

THIS IS IN REPLY TO mpack @ 20:35
MANY new posts to review

>a reboot at 16 mins
It is NOT me rebooting

>the VM shuts down after 15 mins
Why???

>all without incident
Again, why???

>The heartbeat register does not detect a hung VM either
DEFINITELY requires a better cardiologist

>If you have a software crash inside the guest OS
We're talking an environment that goes into meltdown/shutdown

>what makes you believe VirtualBox has something to do with it?
I'd say that being able to run successfully until 6.1.3 install would be a good indicator

Be that as it may (may also be an MS update?), got any suggestions?
Is there an "extreme" debug log I Can switch on?

Thank you for your promptness.

You could look for additional info in the Windows Event Viewer inside the Windows 7 guest OS.

First, my PROFUSE apologies for dropping contact. Was offline, hopefully you'll pardon me...
Restoring the late st good copy, and I intend to test by setting acceleration to 'none' instead of hyper-v (the environment within which the error occurred)
I run a VPN on the host (surfshark) ans zonealarm (firewall+av) on host+guest. I figure I'll shut down the vpn;but the zonealarm, exit host and guest or just guest?
I'd love to get past this as I'd like to install the new 6.1.32 version.
Will hopefully keep you informed.

By the by, 'system' & .log report the OS as Win-7 when that is the OS at time of creation; Win-10 is actually the current OS .

RetiredGuy wrote:my PROFUSE apologies for dropping contact. Was offline, hopefully you'll pardon me...

No problem! Only two days away is very common on the forum. We've seen months between posts...

RetiredGuy wrote:I intend to test by setting acceleration to 'none' instead of hyper-v (the environment within which the error occurred)

I wouldn't do this until you ask first, and make sure you use the Virtualbox names of the settings you want to change. This is called "thrashing" on the forum, and it rarely makes things better; far more often it breaks more than it fixes.

RetiredGuy wrote:setting acceleration to 'none' instead of hyper-v

What is this "acceleration" setting?

RetiredGuy wrote:(the environment within which the error occurred)

What environment are you referring to?

1. Choose vm with LEFT-click in ui
2. Settings-->system
3. Choose 'Acceleration' tab
4. Change 'Paravirtualization Interface' to 'none' from 'hyper-v'

environment=hyper5 (see 'paging' errmsg in log)

Fortunately restored but waited on your input
So what's next?

Thanks for the clarifications.

'Paravirtualization Interface' is a comm channel for virtualization-aware OS's to talk to whatever hypervisor is running them, so they can run more efficiently. Setting this to Hyper-V does not turn on the host OS's Hyper-V. Windows VMs should have 'Paravirtualization Interface' set to Hyper-V as by default.

RetiredGuy wrote:see 'paging' errmsg in log

That would be the BSOD message reported by the VM OS. This message is part of the 'Paravirtualization Interface' and is a good thing, for it tells us what happened in the VM crash, though you must study the VM OS's data per the usual Microsoft channels and web-searches to find out what was the cause of the BSOD.

Please reset 'Paravirtualization Interface' to Hyper-V and try the VM again. If it crashes again, please set the VM's General settings, Basic tab, Version: dropdown to Windows 10 and try once more. If it crashes again, please post another log.

Okay, I'm good.
Was unsuccessful applying your change to the General tab
Equally unsuccessful setting 'nested paging' off.

In an act of frustration with repeated restores, I did as follows:

. Updated to 6.1.132
. Exited Zonealarm i the guest
. Turned off vpn on the host AND STOPPED ITS BACKGROUND SERVICE(S)

A different forum member suggested that AV stuff caused 'page faults'
Sorry, not sure which one did it as I committed the cardinal sin of testing with multiple and concurrent changes