FS#45791 - [virtualbox] start VBoxClient in Virtualbox guest only

Attached to Project: Community Packages
Opened by Christian Hesse (eworm) - Monday, 27 July 2015, 07:52 GMT
Last edited by Sébastien Luttringer (seblu) - Sunday, 13 March 2016, 12:36 GMT
Task Type Feature Request
Category Packages
Status Closed
Assigned To Sébastien Luttringer (seblu)
Architecture All
Severity Low
Priority Normal
Reported Version
Due in Version Undecided
Due Date Undecided
Percent Complete 100%
Votes 0
Private No

Details

Description:
Installing virtualbox-guest-utils on a non-Virtualbox system results in an error message popping up via libnotify:

VBoxCLient: the VirtualBox kernel service is not running. Exiting.

This is really annoying...
As I use this on live media built with archiso uninstalling the package is not an option. The system should as well on Virtualbox and non-Virtualbox machines.

The simplest solution is to use systemd-detect-virt to detect the virtualization and start within Virtualbox guest only. A proposed patch is attached.

(Currently this will fail (or not start) with virtualbox 5.0.0 and guest virtualization set to 'kvm'. However everything is in place for system 223.)

Additional info:
virtualbox-guest-utils 5.0.0-1
   0001-start-within-vbox-guest-... (2.1 KiB)
This task depends upon

Closed by  Sébastien Luttringer (seblu)
Sunday, 13 March 2016, 12:36 GMT
Reason for closing:  Upstream
Additional comments about closing:  https://www.virtualbox.org/ticket/14474
Comment by Doug Newgard (Scimmia) - Monday, 27 July 2015, 14:20 GMT
How is this getting run on install? Or is the problem the xdg autostart?
Comment by Christian Hesse (eworm) - Monday, 27 July 2015, 15:23 GMT
The package installation is no problem. The (modified) file is installed as /usr/bin/VBoxClient-all and autostarted via xdg autostart (/etc/xdg/autostart/vboxclient.desktop) at user login.
Comment by Sébastien Luttringer (seblu) - Tuesday, 18 August 2015, 21:36 GMT
Well, your patch make sense. Did you propose it upstream?
Comment by Christian Hesse (eworm) - Wednesday, 19 August 2015, 06:20 GMT
No, I did not. In Arch we can expect to have systemd (and systemd-detect-virt) around. Upstream would have to care about some more cases when systemd-detect-virt is not available... So not sure what exactly to propose upstream.
Comment by Sébastien Luttringer (seblu) - Wednesday, 19 August 2015, 09:48 GMT
I don't get it. There is more and more distribution using systemd, testing the presence of systemd-detect-virt to make your patch generic enough to be submitted is easy.

As you also know, in Arch we try to not feature patch upstream code source. This is why I'm a bit reluctant; pushing your feature suggestion upstream is welcome.
Comment by Christian Hesse (eworm) - Wednesday, 19 August 2015, 10:45 GMT
Done. :D
Let's wait for a response...

https://www.virtualbox.org/ticket/14474
Comment by Sébastien Luttringer (seblu) - Saturday, 12 March 2016, 18:01 GMT
Christian, what do we do with this? Reading the last comment in the upstream BR, an equivalent patch seems to be prepared.
Should I close here and let the works done upstream? Or you prefer we do differently?
Comment by Christian Hesse (eworm) - Saturday, 12 March 2016, 19:09 GMT
I am fine with closing here. Let's wait for upstream to get a proper fix.
I need this for myself, but I can handle it just fine for now. ;)

Loading...