15/05/14 20:35, sajolida@??? wrote:
> I've been working on fixing the build of the VirtualBox guest modules in
> Tails 1.1.
>
> ticket: 5972
> branch: bugfix/fix_virtualbox_dkms_build
>
> Summary:
> - Good news: The build is fixed, way less hackish, and that was easy.
> - Bad news: The modules don't work :(
>
> When I say the modules don't work I mean that `VBoxClient --display`,
> and `VBoxClient --clipboard` don't activate the resizable display and
> shared clipboard. But the shared folders work. So that seems like a
> error with X11, not with the modules themselves.
I've tested this branch, and it actually seems to work as good as it
possibly can at the moment:
*
https://www.virtualbox.org/ticket/11037
*
https://www.virtualbox.org/ticket/8336
(Unfortunately these upstream tickets are only referenced in the
resolved tickets #5835 and #5456. I hope you didn't waste too much time
on this... :S)
In other words, everything will work perfectly with the 32-bit kernel,
but not necessarily with the 64-bit kernel since that's not supported.
Only missing out on the display management (resize host window => change
guest's resolution) and clipboard sharing in the latter case isn't so
bad, when we get back file sharing and a greater selection of
resolutions. Heck, it's probably more secure that way (and host -> guest
clipboard sharing is a pretty obvious leak)! :)
> Even more weird, I tried to start this working VM with that last ISO
> image and the resizing worked! Then I tried to create another fresh VM
> from scratch with no disk and started it from that ISO image and the
> resizing didn't work. I tried to compare the config file for both VMs
> and didn't find anything relevant.
Probably the "working VM" was 32-bit, and the "fresh VM" was 64-bit.
> Anyway, I'm absolutely not sure I'll be able to fix that before the
> freeze. I'd rather concentrate on the documentation upgrades. So if any
> VirtualBox user want to have a look...
I haven't done a code review yet, but it looks very promising to me so
expect this to be merged in time for 1.1. Thanks for working on this!
A potential follow-up is how to document the shortcomings. Known issue,
perhaps?
Cheers!
