VirtualBox

Opened 8 years ago

Closed 4 years ago

#15684 closed defect (obsolete)

VM craches after a while

Reported by: Panos Kavalagios Owned by:
Component: VM control Version: VirtualBox 5.1.2
Keywords: Cc:
Guest type: Windows Host type: Linux

Description

The VM crashes after a while. The log is attached.

Attachments (5)

VBox.log (72.2 KB ) - added by Panos Kavalagios 8 years ago.
VBox.2.log (86.2 KB ) - added by Panos Kavalagios 8 years ago.
virtualbox-core-dump.txt (4.7 KB ) - added by Panos Kavalagios 8 years ago.
VBox.3.log (129.9 KB ) - added by aanderson 8 years ago.
apport-unpack-report.zip (58.2 KB ) - added by aanderson 8 years ago.

Download all attachments as: .zip

Change History (22)

by Panos Kavalagios, 8 years ago

Attachment: VBox.log added

by Panos Kavalagios, 8 years ago

Attachment: VBox.2.log added

comment:1 by Panos Kavalagios, 8 years ago

It happened again and emitted the following line in the dmesg:

[11697.614264] EMT-3[2158]: segfault at 10 ip 00007fd0c21901a1 sp 00007fd0f13d8bb0 error 4 in VBoxDD.so[7fd0c207e000+283000]

comment:2 by Socratis, 8 years ago

The log doesn't tell me too much, except the fact that your host has 4 cores and you assign all of them to the guest. Lower them to a max of 3 and see if it changes the situation.

For future reference, it's best if crashes and other issues get first triaged in the forums (https://forums.virtualbox.org/). 95% of the issues are resolved over there, which keeps the developers focusing on the bug fixes and enhancements.

comment:3 by Panos Kavalagios, 8 years ago

According to my understanding, a crash in an application is considered an application bug that should be logged to a bug tracking system. The maximum number of CPUs were assigned since VirtualBox 3 and the problem is observed only in the latest version. It does not affect previous versions.

If the log of the application is not enough, you may indicate additional information that can be retrieved from the host or guest system. In case of a segmentation fault, the core file produced could provide more information with gdb.

Anyway, I apologise for wasting your time. You may close the ticket and I will ask in the forums as suggested.

comment:4 by Socratis, 8 years ago

I'm not sure that you read my post as I intended it. *I* can't find anything wrong in your logs; others might. That is exactly why I suggested to first try the forums, there are a lot more eyes there compared to the bugtracker. I don't think you're wasting anybody's time, I was saying you might have a better chance of resolving it.

Plus I can't close a bug. I'm a simple user, just like you ;)

comment:5 by Panos Kavalagios, 8 years ago

Sorry, I thought you were VirtualBox support/developer. OK, we can wait for Frank then.

comment:6 by gojul, 8 years ago

Same here actually. The issue appeared with VirtualBox 5.1.2.

Host : Debian Jessie AMD64

comment:7 by Don Hughes, 8 years ago

See Ticket#15677

comment:8 by Frank Mehnert, 8 years ago

The VirtualBox process crashes so this is definitely a VirtualBox bug. Can you provide a core dump?

comment:9 by Panos Kavalagios, 8 years ago

You may find attached as virtualbox-core-dump.txt the requested information. It is the result of the "coredumpctl info 18364". Let me know if it is enough or you need the actual core file.

by Panos Kavalagios, 8 years ago

Attachment: virtualbox-core-dump.txt added

comment:10 by aeichner, 8 years ago

Please provide the actual core dump.

comment:11 by Panos Kavalagios, 8 years ago

The core file itself does not exist, since it is set as following:

# grep -i storage /etc/systemd/coredump.conf
Storage=none

I am changing the value of storage to external, so that it can be kept in the /var/lib/systemd/coredump disk location.

comment:12 by Panos Kavalagios, 8 years ago

It happened again, but no core file was generated. Nothing. The VM was closed unexpectingly leaving no trace. From the logs:

Aug 26 10:39:01 bb229 kernel: EMT-3[2801]: segfault at 10 ip 00007f55a4e4b859 sp
 00007f55a7ffebb0 error 4 in VBoxDD.so[7f55a4d38000+286000]
Aug 26 10:39:02 bb229 systemd: Created slice system-systemd\x2dcoredump.slice.
Aug 26 10:39:02 bb229 systemd: Started Process Core Dump (PID 4884/UID 0).
Aug 26 10:39:02 bb229 systemd-coredump: Coredump file descriptor missing.

coredumpctl does not report the crash and neiher /var/lib/systemd/coredump has any vbox core file.

by aanderson, 8 years ago

Attachment: VBox.3.log added

by aanderson, 8 years ago

Attachment: apport-unpack-report.zip added

comment:13 by aanderson, 8 years ago

Host:

uname -a
Linux aanderson-dt 4.4.0-34-generic #53-Ubuntu SMP Wed Jul 27 16:06:39 UTC 2016 x86_64 x86_64 x86_64 GNU/Linux
lsb_release -r
Release:	16.04

installed from package: virtualbox-5.1_5.1.4-110228-Ubuntu-xenial_amd64.deb

sudo apt list -a virtualbox

Listing... Done
virtualbox/xenial-updates 5.0.24-dfsg-0ubuntu1.16.04.1 amd64
virtualbox/xenial 5.0.18-dfsg-2build1 amd64

sudo apt list -a virtualbox-5.1
Listing... Done
virtualbox-5.1/now 5.1.4-110228~Ubuntu~xenial amd64 [installed,local]

sudo dpkg --get-selections | grep -v deinstall|grep virtual
unity-scope-virtualbox				install
virtualbox-5.1					install


Guest: Windows 10 Pro Version 1511 OS Build 10586.545 Processor AMD FX(tm) 8350 Additions: 5.1.4r110228

I am encountering the same VBoxDD.so. I was running Unbuntu 15.10 and VirtualBox 5.0.26 without issue but since I upgraded to Ubuntu 16.04 and VirtualBox 5.1.4 my Windows 10 x64 guest as been crashing multiple times a day. I attached the VBox.log file and an apport-unpack of the core dump. The full core dump is 83MB and I don't know if I am able to upload it. Here is what appears in dmesg:

[217970.488838] vboxdrv: ffffffffc085d020 VMMR0.r0
[217970.868595] vboxdrv: ffffffffc0595020 VBoxDDR0.r0
[217971.034533] vboxdrv: ffffffffc0006020 VBoxEhciR0.r0
[219694.365710] vboxdrv: ffffffffc085d020 VMMR0.r0
[219702.015544] vboxdrv: ffffffffc0595020 VBoxDDR0.r0
[219702.201068] vboxdrv: ffffffffc0006020 VBoxEhciR0.r0
[246959.997660] EMT-3[11821]: segfault at 10 ip 00007f79916059db sp 00007f79bd3a5bb0 error 4 in VBoxDD.so[7f79914ee000+28d000]
Last edited 8 years ago by aanderson (previous) (diff)

comment:14 by aeichner, 8 years ago

Can you try to disable audio for the VM and see if the crash disappears?

comment:15 by aanderson, 8 years ago

I disabled the audio for the VM and it has been running several days without aborting.

comment:16 by pentagonik, 8 years ago

@aanderson Please try the latest 5.1 build from the Testbuilds page.

Scroll down to the Latest test builds / 5.1.x builds section and download the build appropriate for your host platform.

Could you please supply the verbose VBox.log file as shown here: Investigating audio problems

Thank you!

comment:17 by aeichner, 4 years ago

Resolution: obsolete
Status: newclosed
Note: See TracTickets for help on using tickets.

© 2023 Oracle
ContactPrivacy policyTerms of Use