CFT: vdfuse port (was: Re: Mounting VirtualBox vdi files [SEC=UNCLASSIFIED])

[Brandon Gooch]

On Mon, Sep 13, 2010 at 4:14 PM, Juergen Lock <nox at jelal.kn-bremen.de> wrote:
> On Sat, Sep 11, 2010 at 01:06:24PM -0600, Warren Block wrote:
>> On Sat, 11 Sep 2010, Juergen Lock wrote:
>>
>> > On Wed, Sep 08, 2010 at 07:44:37PM -0600, Warren Block wrote:
>> >> On Thu, 2 Sep 2010, Juergen Lock wrote:
>> >>
>> >>> ...there is a much simpler fusefs tool in debian called vdfuse that I
>> >>> now finally made a port of.  I don't know how stable this is on
>> >>> FreeBSD or if there still may be bugs, so please give this a good test
>> >>> and post your results here.  Here's a simple example:
>> >>>
>> >>>   # vdfuse -r -f ~nox/.VirtualBox/HardDisks/win7-64.vdi /mnt
>> >>>   # mdconfig -a -f /mnt/Partition2 -o readonly
>> >>>   md0
>> >>>   # mount -o ro -t ntfs /dev/md0 /mnt2
>> >>>   # ls -l /mnt2
>> >>>   ...
>> >>>   # umount /mnt2
>> >>>   # mdconfig -d -u 0
>> >>>   # umount /mnt
>> >>
>> >> It works for a .vdi of a UFS drive.  The BSD partitions show up on the
>> >> md0 device (md0a, etc).  Just a trivial test, but looks promising!
>> >
>> > Yep, those kind of tests worked for me too, I just wasn't sure if it
>> > also survives `heavy use', i.e. reading/writing/using files much on
>> > the mounted fs.es...
>>
>> Dumping the partitions of that FreeBSD system worked just now.  Not what
>> I'd call a thorough test, but it did read everything in the filesystem
>> without problems.
>
> Ok that sounds promising...  Someone want to do more tests or should I
> just commit the port? :)

During my first (and only) testing session, I wasn't able to mount the
md(4) device after I created it. Further, after a few attempts each
time, vdfuse would segfault.

Image was form Windows XP 32-bit (1 Partition, NTFS).

I'm in the process of rebuilding my system ATM (kernel, world, and
ports), but I will test further when I'm back up again...

-Brandon