> > I wonder, what is wrong in reporting mounts in other namespaces that
quoted text
> > either receive and send propagation to mounts in our namespace?
> 
> A plenty.  E.g. if foo trusts control over /var/blah to bar, it's not
> obvious that foo has any business knowing if bar gets it from somebody
> else in turn.  And I'm not sure that bar has any business knowing that
> foo has the damn thing attached in five places instead of just one,
> let alone _where_ it has been attached.
> 
> If you get down to it, the thing is about delegating control over part
> of namespace to somebody, without letting them control, see, etc. the
> rest of it.  So I'd rather be very conservative about extra information
> we allow to piggyback on that.  I don't know... perhaps with stable peer
> group IDs it would be OK to show peer group ID by (our) vfsmount + peer
> group ID of master + peer group ID of nearest dominating group that has
> intersection with our namespace.  Then we don't leak information (AFAICS),
> get full propagation information between our vfsmounts and cooperating
> tasks in different namespaces can figure the things out as much as possible
> without leaking 3rd-party information to either.

This sounds fine.

I'll have a look at implementing a stable peer group ID (it doesn't
need a separate object, I realized that now).

Miklos
-
To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html