:: Re: [DNG] Virtualbox?
トップ ページ
このメッセージを削除
このメッセージに返信
著者: Olaf Meeuwissen
日付:  
To: fsmithred
CC: dng
題目: Re: [DNG] Virtualbox?
Hi,

fsmithred writes:

> On 10/25/2017 08:05 AM, Olaf Meeuwissen wrote:
>> Using the virtualbox package from either jessie (4.3.36-dfsg-1+deb8u1)
>> or jessie-backports (5.1.8-dfsg-6~bpo8+2) with a stock linux-image-amd64
>> (3.16+63) from jessie works fine. That is, it installs without errors
>> and virtualbox images start fine (based on my provisioning Debian jessie
>> images and ssh-ing into them).
>>
>> Using either virtualbox package with the stock linux-image-amd64 from
>> jessie-backports (4.9+80~bpo8+1) however does *not* work. Installation
>> fails with a compile error of the vboxdrv kernel module. It looks like
>> the kernel API has changed (unsurprisingly when going from a major
>> version of 3 to 4). Neither virtualbox package seems to deal with this,
>> not even the one from backports.
>>
>> Next move is checking whether things work in ascii because my laptop at
>> the office doesn't really work with a 3.16 kernel.
>
> Good luck with that. There is no virtualbox package in ascii/stretch. You
> can get the packages from virtualbox.org, and those will work in ascii.
>
> OK, that's weird - there's one in ascii-backports but not in
> stretch-backports. I don't know what that's about. A few minutes of
> googling tells me what I already know - it's not in stretch. Maybe someone
> else knows the right words to search to find out why that is.


There is a virtualbox package in stretch-backports (and buster and sid).
See https://packages.debian.org/search?keywords=virtualbox

This is the package that is "merged" into ascii-backports. It installs
without any problem against the ascii vanilla kernel (linux-image-amd64
in my case).

I've been using it at the office and it seems to work (depending on your
definition of "work"). I've used in combination with a Vagrant box that
provides a debian/jessie64 based Docker capable Jenkins slave agent.
The only issue I have at the moment is that the box' clock runs behind
at the pace of 20 seconds to the minute! That's despite the fact that
both the host OS and the Vagrant box have ntp/openntpd installed. The
host OS' clock stays in sync. Both use the same set of NTP servers and
can communicate with them.

As to why virtualbox is not in stretch, that's detailed in BTS#794466.
Long story short, it's because the Debian and Oracle policies for CVEs
are at odds. Debian does not allow upstream version bumps for stable
but Oracle only releases their security *and other* fixes via version
bumps and it appears to be too difficult the separate the other fixes
from the security ones. The stable-backports, testing and unstable
suites don't have this version restriction so can use the upstream as
is. From what I read in the BTS, I'd expect the same to happen for
buster, i.e. virtualbox getting dropped from buster when that enters
freeze and to resurface in buster-backports after a while.

Hope this helps,
--
Olaf Meeuwissen, LPIC-2            FSF Associate Member since 2004-01-27
 GnuPG key: F84A2DD9/B3C0 2F47 EA19 64F4 9F13  F43E B8A4 A88A F84A 2DD9
 Support Free Software                        https://my.fsf.org/donate
 Join the Free Software Foundation              https://my.fsf.org/join