Windows 7 Guest Randomly Aborts

[dietfig]

This has been going on for some time now. I'm attaching 3 logs that I saved after the machine aborted. The first two are from VirtualBox VM 4.2.10 r84104 and I have looked through them and didn't notice anything that seemed unusual. The latest abort was under VirtualBox VM 4.2.12 r84980 and the last line in the log indicated an OpenGL assertion failure: "00:16:34.760450 OpenGL Warning: Assertion failed: !cr_server.curClient || !cr_server.curClient->currentMural || cr_server.curClient->currentMural == mural, file /home/vbox/vbox-4.2.12/src/VBox/HostServices/SharedOpenGL/crserverlib/server_window.c, line 395".

When I was running 4.2.10 the abort only seemed to happen if the VM was idling and I was off doing something else. Today, with 4.2.12, it happened while I was actively using it. That's the only pattern I've been able to notice.

The host system is Debian 7.0.

Code: Select all

$ uname -a
Linux debianthinkpad 3.2.0-4-amd64 #1 SMP Debian 3.2.41-2 x86_64 GNU/Linux

Anyone have any suggestions on what to do to further troubleshoot the problem?

[Perryg]

I would post the last log (4.2.12) here viewtopic.php?f=1&t=55091 since this is close to what others are seeing.

The abort is another issue. From the logs it appears to be the host side that aborts, since there is no indication of the guest aborting or shutting down. Have you looked at the VBoxSVC.log, and the hosts syslog to see if you are getting segfaults?

[dietfig]

I would post the last log (4.2.12) here <...scrubbed>

I'll do that.

The abort is another issue. From the logs it appears to be the host side that aborts, since there is no indication of the guest aborting or shutting down. Have you looked at the VBoxSVC.log, and the hosts syslog to see if you are getting segfaults?

Reading back through the syslog I did find a segfault that occurred at roughly the time that the VM crashed but I'm not positive that the two are related. I couldn't find anything interesting in VBoxSVC.log but I'm attaching it all the same.

Code: Select all

Apr 15 10:25:03 debianthinkpad kernel: [99210.317670] EMT-0[4782]: segfault at 7
f3e7444fcf0 ip 00007f3e70fd507d sp 00007f3e4b931a20 error 4 in libc-2.13.so[7f3e
70f5e000+180000

[Perryg]

I did find a segfault that occurred at roughly the time that the VM crashed but I'm not positive that the two are related

Actually that explains why nothing is being recorded by VirtualBox. So it is host side.

The tricky part is trying to figure out what triggered the segfault. If you Google "error 4 in libc-2.13.so" you will see what I mean.

Do you hibernate the laptop?

[dietfig]

Do you hibernate the laptop?

Yes, all the time, although never with the VM running.

The tricky part is trying to figure out what triggered the segfault. If you Google "error 4 in libc-2.13.so" you will see what I mean.

Would following the instructions on the wiki for obtaining a core dump get something more useful than the syslog error message?

[Perryg]

Sure that would help but if you are not running the debug build you would not have the necessary files to make any sense of it.
You might want to post a ticket at bugtracker and see if they can tell you what they need.
They will tell you where to send the core dump.

[dietfig]

Thanks, I appreciate your help! I'll wait for the abort to happen again to confirm that it really is a segmentation fault and then open a ticket on the bug tracker.