SHM objects cannot be isolated in jails, any evolution in future FreeBSD versions?

[James Gritton]

On 2016-03-17 05:54, Simon wrote:
> Le 2016-03-15 09:34, Miroslav Lachman a écrit :
>> Mark Felder wrote on 03/14/2016 22:07:
>>> 
>>> 
>>> On Sat, Mar 12, 2016, at 11:42, James Gritton wrote:
>>>> On 2016-03-12 04:05, Simon wrote:
>>>>> The shm_open()(2) function changed since FreeBSD 7.0: the SHM 
>>>>> objects
>>>>> path are now uncorrelated from the physical file system to become 
>>>>> just
>>>>> abstract objects. Probably due to this, the jail system do not 
>>>>> provide
>>>>> any form of filtering regarding shared memory created using this
>>>>> function. Therefore:
>>>>> 
>>>>> - Anyone can create unauthorized communication channels between 
>>>>> jails,
>>>>> - Users with enough privileges in any jail can access and modify 
>>>>> any
>>>>> SHM objects system-wide, ie. shared memory objects created in any
>>>>> other jail and in the host system.
>>>>> 
>>>>> I've seen a few claims that SHM objects were being handled 
>>>>> differently
>>>>> whether they were created inside or outside a jail. However, I 
>>>>> tested
>>>>> on FreeBSD 10.1 and 9.3 but found no evidence of this: both version
>>>>> were affected by the same issue.
>>>>> 
>>>>> A reference of such claim:
>>>>> https://lists.freebsd.org/pipermail/freebsd-ports-bugs/2015-July/312665.html
>>>>> 
>>>>> My initial post on FreeBSD forum discussing the issue with more
>>>>> details: https://forums.freebsd.org/threads/55468/
>>>>> 
>>>>> Currently, there does not seem to be any way to prevent this.
>>>>> 
>>>>> I'm therefore wondering if there are any concrete plans to change 
>>>>> this
>>>>> situation in future FreeBSD versions? Be able to block the 
>>>>> currently
>>>>> free inter-jail SHM-based communication seems a minimum, however 
>>>>> such
>>>>> setting would also most likely prevent SHM-based application to 
>>>>> work.
>>>>> 
>>>>> Using file based SHM objects in jails seemed a good ideas but it 
>>>>> does
>>>>> not seem implemented this way, I don't know why. Is this planned, 
>>>>> or
>>>>> are there any greater plans ongoing also involving IPC's similar
>>>>> issue?
>>>> 
>>>> There are no concrete plans I'm aware of, but it's definitely a 
>>>> thing
>>>> that should be done.  How about filing a bug report for it?  You've
>>>> already got a good write-up of the situation.
>>>> 
>>> 
>>> Both this and SYSV IPC jail support[1] are badly needed.
>>> 
>>> [1] https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=48471
>> 
>> Yes, it is very sad that original patch was not commited, nor
>> commented or improved by core developers for long 13 years. I am not
>> 100% sure but I thing there was some patch from PJD for SysV IPC too.
>> There were EclipseBSD with resource limits in times of FreeBSD 3.4 and
>> there is FreeVPS for 6.x with virtualized IPC...
>> 
>> So I really hope SysV IPC aware jails will become reality soon.
>> 
>> Miroslav Lachman
> 
> Hi everyone,
> 
> Odd thing, I've seen that the very first exchanges which opened this
> mailing list back in 2007 precisely discussed IPC isolation in Jail
> and some work already done in the Jail2 project part of the now
> abandoned FreeVPS project. At that time IPC virtualization was
> qualified as an easy job:
> 
>> As say about SYSV IPC stuff you say about only virtualization? or
>> also about limits? "virtualization" is easy, but for limits - need 
>> more
>> work
> (https://lists.freebsd.org/pipermail/freebsd-jail/2007-May/000004.html)
> 
> We have now come full circle :).
> 
> As per the SHM objects issue, I've now filled a new bug #208082:
> https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=208082
> 
> I explain in the bug description why it may be different than the
> already existing bug #48471 covering SysV IPC.
> 
> Le 2016-03-17 01:10, Dewayne Geraghty a écrit :
>> PS We don't want/need the complexity (or performance hit) associated
>> with v* additions when a well thought out (simple) jail does the task
>> very nicely :)
> 
> I agree, the main advantage of jails and other lightweight containers
> is precisely their lightness.
> 
> Regards,
> Simon.

I've put a diff on the bug report (Bug 208082), for the shm objects, and 
also for ksem and mqueue which have the same problems.  Any review is 
welcome :-).

SYSV IPC is a separate issue.  I'm following up with bz about my memory 
of hearing there's something vimage-related there, and if there isn't I 
can jump into that one as well (I actually have some work already done 
with it, so it just needs a little more).

- Jamie