On 64 bit Ubuntu 14.04, I have two virtual machines installed with Oracle
VirtualBox Version 5.0.20 r106931 that I have used for quite some time.
Several days ago, I attempted to load a virtual machine by pointing to an iso file on my hard drive based on instructions in a magazine.

That, by the way, was unsuccessful, and I deleted the dregs of this (at least I believe I did anyway). Since then, however, each time I reboot, a system error dialog box pops up once I log in to the desktop. When I attempt to run the VirtualBox Manager, I get the following error message:

==============================================================================
VirtualBox - Error in suplibOsInit
------------------------------------------------------------------------------
Kernel driver not installed (rc=-1908)
The VirtualBox Linux kernel driver (vboxdrv) is either not loaded or there is
a permission problem with /dev/vboxdrv. Please reinstall the kernel module
by executing
'sbin/rcvboxdrv setup'
as root ... et cetera, and ending with
where: suplibOsInit what: 3 VERR_VM_DRIVER_NOT_INSTALLED (rc=-1908) -
The support driver is not installed. On linux, open returned ENOENT.
==============================================================================

Being an obedient and trusting sort, I run 'sudo sbin/rcvboxdrv setup' in a
terminal, enter my password and, after thinking about things for a while, the
prompt returns and all is well with my virtual machines.

I've looked at everything I can think of, and can find no "permission problems"
so I'm at a loss as to what to look for now.

Does anyone know what I'm missing? Thanks for any pointers.

(NOTE: I don't believe this is related to Steven G's posting on 26 May
2016, since I don't recall having a kernel update since the last time I used VirtualBox, but who knows? Just for the record, I'm using kernel 3.19.0-59-generic according to 'uname -r'.)

Also, I should mention that lsmod gives the following lines after running the rcvboxdrv setup:
Module Size Used by
vboxpci 24576 0
vboxnetadp 28672 0
vboxnetflt 28672 0
vboxdrv 450560 3 vboxnetadp,vboxnetflt,vboxpci

... just in case.

Do you have a recent backup of 'VirtualBox VMs' from before the changes?

I do have backups of all the VirtualBox VMs, but I'm trying to understand what I did to "break" VirtualBox. The existing VMs (I didn't restore the backups) work fine after I execute the [sudo sbin/rcvboxdrv setup]. As far as I know the kernel module this command adds is attached to my host OS, not any of the virtual OSs (if that's not true I'm very confused).

the command: lsmod | grep vbox gives the following after running the setup:
vboxpci 24576 0
vboxnetadp 28672 0
vboxnetflt 28672 0
vboxdrv 450560 3 vboxnetadp,vboxnetflt,vboxpci

... nothing shows up when I first boot, so I don't think the problems are with the VMs themselves.

Apologies CVAlkan, I thought that is what you were after, after having already posted my premature mention of backups. I don't know that you "broke" anything in VBox by removing the iso that wouldn't launch. Acc'd to Virtualbox it is supposed to be possible to boot an iso and create a vdi on bare metal. I have also at times had to reload the modules (modprobe in Arch) after removing a vm from VirtualBox to get the others to work. Perhaps you can try the install again, then run debugging when you uninstall it to see what happens. There may be information in one of the VirtualBox logs already as to what happened why vboxdrv had to be reloaded.

One thing I have done when getting a permissions problem like this is to open thunar (or whatever your file manager is) with "sudo thunar". Then navigate to 'sbin/rcvboxdrv setup' and make sure the permissions are set so anyone can "read/write" and that ought to take care of it.

Thanks much. I think my problem turned out to be that I needed to shut down all the VMs for the kernel change to take effect permanently. I'm not sure I understand why that is, but once I shut down the one machine I had left "powered on," and then re-booted, things went back to normal.