GSoC Idea: per-process filesystem namespaces for FreeBSD

[Mark Saad]

On Tue, Mar 13, 2018 at 8:16 PM, Theron Tarigo <theron.tarigo at gmail.com> wrote:
> On 03/13/18 18:23, Bakul Shah wrote:
>>
>> Plan9 has no root (superuser) or setuid.  You can mangle
>> anything in your namespace but it affects only *your* own
>> process and its future descendants.
>>
>> The following paper on Plan9 authentication in Linux may be
>> worth reading:
>>
>> https://static.googleusercontent.com/media/research.google.com/en//pubs/archive/34433.pdf
>>
>> While I have wanted per-process namespace in BSD for a long
>> time, I agree with Konstantin this is a non-trivial project.
>> Even if the design was fully fleshed out, implementing it
>> would likely take longer than 12 weeks.
>
> Although it would limit the usefulness of it, ignoring any and all file suid
> bits for any process with a non-empty mount table should in theory prevent
> exploitation of setuid.  Allowing safe setuid in combination with ("trusted"
> ?) namespaces would be something to add support for much later if someone
> decides it would be useful.
>
> By focusing on a narrowed case, that of allowing an unprivileged process to
> alter its view into the vfs in a way which is only preserved through
> execve() in specific safe circumstances, I hoped to avoid the insurmountable
> complexity of implementing the feature in the full generality that is
> available on Plan9.
>
> On 03/13/18 18:31, Mark Saad wrote:
>>
>>  A kind of related task; FreeBSD could benefit from : Fixing  and
>> improving unionfs / nullfs. There are some weird issues with the current
>> unionfs and while it works in many cases there are some edge cases where the
>> comments are something like “ FreeBSD needs a proper stacking vfs ...”   the
>> examples I can think of ; imagine you have a jail , chroot or even a Pxe
>> booted system where you want a a read only null mount from the hosts /bin to
>> the targets /bin . Now expand that to most of the base system and the mount
>> tmpfs’s for /tep /var/log etc.  most of that works but try to unmount it in
>> the wrong order or thrash a unionfs with lots of writes ,on top of a tmpfs
>> and things break .
>> So to be clear the project would be to better document the various uses of
>> unionfs and nullfs that work , for the ones that do not diving into the
>> stacking vfs and seeing if it could be implemented and if it would help .
>>
> Using nullfs / unionfs in combination with chroot could be made functionally
> equivalent to per-process namespace, but would have the very same security
> problems as already discussed (as any chroot have) so configuring such
> environments would be available only to superuser.

I was not thinking of you cobbling nullfs and unionfs into a FreeBSD
version of plan9's bind
I was thinking more about how nullfs , and unionfs are influenced by
plan9 and how they could benefit
from someone with some plan9 experience. Not to say that any of the
prior work was done poorly or
incorrectly because of a lack of plan9 experience . But that your
enthusiasm would be better focused
on a part of base that could use some polish and maybe some rethinking .

>
> So it appears that the most significant obstacle to achieving at least an
> approximation of the behavior of user-controlled per-process namespace is
> managing setuid safely.

One idea here could be to use capsicum to enforce this requirement but
that would be a new approach from what I understand.
https://www.freebsd.org/cgi/man.cgi?capsicum(4)

>
> One thought I had is to do all of this purely in user-space by creating an
> extension to libc which allows appropriately linked programs to find files
> according to some set of prefixes defining the namespace, defined through an
> environment variable, but have all system interactions function in the
> normal ways.  Would this be sufficiently general to be of any use?  If this
> approach does pose any security threats, it indicates a hole is already
> present in the system! (MacOS once had a problem with allowing privileged
> programs to be launched under a modified library path...)
>
> On 03/13/18 20:00, Mark Saad wrote:
>>
>> However I still think the unionfs / nullfs work I mentioned before would
>> be a good project related to the plan9 idea in some ways .
>
> Is there a way I could go about fixing up unionfs while also making
> significant progress towards eventual support for true per-process
> namespace?
>
> Theron

I think there is a lot of opportunity here to pick apart a subsystem
that shares some commonality with plan9 . I think that improving
unionfs / nullfs to have less edge cases would be a great project.
You would need to set a smaller goal for managing the expiation
of time you will have for GSoC .

Some small tasks that could be a starting point for working on this
project.  There are a few projects that depend on creating a
unionfs of a tmpfs and a readonly disk image.   One example would be
booting off a CD-Rom and using a unionfs to make a ram disk give the
read-only medium the appearance of have a read / write. Second example
is the Docker / Container setup where a read-only loop-mounted
disk image is unioned to a ram disk in a similar way to the cdrom was
before . Third option is where a Jail is created from a template
directory
that would mount versions of a binary or directory into the jailed
environment for various reason.


-- 
mark saad | nonesuch at longcount.org