Re: [RFC PATCH ghak90 (was ghak32) V3 07/10] audit: add support for containerid to network namespaces

[Richard Guy Briggs <rgb@redhat.com>]

On 2018-07-24 16:33, Paul Moore wrote:
> On Tue, Jul 24, 2018 at 10:06 AM Richard Guy Briggs <rgb@redhat.com> wrote:
> > On 2018-07-20 18:14, Paul Moore wrote:
> > > On Wed, Jun 6, 2018 at 1:03 PM Richard Guy Briggs <rgb@redhat.com> wrote:
> > > > Audit events could happen in a network namespace outside of a task
> > > > context due to packets received from the net that trigger an auditing
> > > > rule prior to being associated with a running task.  The network
> > > > namespace could in use by multiple containers by association to the
> > > > tasks in that network namespace.  We still want a way to attribute
> > > > these events to any potential containers.  Keep a list per network
> > > > namespace to track these audit container identifiiers.
> > > >
> > > > Add/increment the audit container identifier on:
> > > > - initial setting of the audit container identifier via /proc
> > > > - clone/fork call that inherits an audit container identifier
> > > > - unshare call that inherits an audit container identifier
> > > > - setns call that inherits an audit container identifier
> > > > Delete/decrement the audit container identifier on:
> > > > - an inherited audit container identifier dropped when child set
> > > > - process exit
> > > > - unshare call that drops a net namespace
> > > > - setns call that drops a net namespace
> > > >
> > > > See: https://github.com/linux-audit/audit-kernel/issues/92
> > > > See: https://github.com/linux-audit/audit-testsuite/issues/64
> > > > See: https://github.com/linux-audit/audit-kernel/wiki/RFE-Audit-Container-ID
> > > > Signed-off-by: Richard Guy Briggs <rgb@redhat.com>
> > > > ---
> > > >  include/linux/audit.h | 23 ++++++++++++++++
> > > >  kernel/audit.c        | 72 +++++++++++++++++++++++++++++++++++++++++++++++++++
> > > >  kernel/auditsc.c      |  5 ++++
> > > >  kernel/nsproxy.c      |  4 +++
> > > >  4 files changed, 104 insertions(+)
> 
> ...
> 
> > > > +                       }
> > > > +       if (!found)
> > > > +               return;
> > > > +       list_del(&cont->list);
> > > > +       if (refcount_dec_and_test(&cont->refcount))
> > > > +               kfree(cont);
> > >
> > > Don't you want to dec_and_test first and only remove it from the list
> > > if there are no other references?
> >
> > I don't think so.  Let me try to describe it in prose to see if I
> > understood this properly and see if this makes more sense: I want to
> > remove this audit_contid list member from this net's audit_contid list
> > and decrement unconditionally this member's refcount so it knows there
> > is one less thing pointing at it and when there is no longer anything
> > pointing at it, free it.
> 
> Yep, sorry, my mistake, I was thinking the other way around (netns
> going away) ... which actually, this patchset doesn't handle that does
> it (I don't see any new code in audit_net_exit())?  Is is in a later
> patch?  If so, it really should be in the same patch as this code to
> prevent bisect nasties.

I don't think any code is needed in audit_net_exit() other than a WARN()
that the contid list is empty, since if a netns is going away, all its
tasks would have exited (removing its contid from the list) or moved to
a new nstns (moving its contid from this list to another netns).
At worst, this is a memory leak and not a bad pointer dereference.

> paul moore

- RGB

--
Richard Guy Briggs <rgb@redhat.com>
Sr. S/W Engineer, Kernel Security, Base Operating Systems
Remote, Ottawa, Red Hat Canada
IRC: rgb, SunRaycer
Voice: +1.647.777.2635, Internal: (81) 32635