Re: [PATCH bpf-next v2 1/2] bpf/flow_dissector: add mode to enforce global BPF flow dissector

[Stanislav Fomichev <sdf@fomichev.me>]

On 10/05, Andrii Nakryiko wrote:
> On Fri, Oct 4, 2019 at 8:58 AM Stanislav Fomichev <sdf@google.com> wrote:
> >
> > Always use init_net flow dissector BPF program if it's attached and fall
> > back to the per-net namespace one. Also, deny installing new programs if
> > there is already one attached to the root namespace.
> > Users can still detach their BPF programs, but can't attach any
> > new ones (-EEXIST).
> >
> > Cc: Petar Penkov <ppenkov@google.com>
> > Signed-off-by: Stanislav Fomichev <sdf@google.com>
> > ---
> 
> Looks good, but see my note below. Regardless:
> 
> Acked-by: Andrii Nakryiko <andriin@fb.com>
> 
> >  Documentation/bpf/prog_flow_dissector.rst |  3 ++
> >  net/core/flow_dissector.c                 | 42 ++++++++++++++++++++---
> >  2 files changed, 41 insertions(+), 4 deletions(-)
> >
> > diff --git a/Documentation/bpf/prog_flow_dissector.rst b/Documentation/bpf/prog_flow_dissector.rst
> > index a78bf036cadd..4d86780ab0f1 100644
> > --- a/Documentation/bpf/prog_flow_dissector.rst
> > +++ b/Documentation/bpf/prog_flow_dissector.rst
> > @@ -142,3 +142,6 @@ BPF flow dissector doesn't support exporting all the metadata that in-kernel
> >  C-based implementation can export. Notable example is single VLAN (802.1Q)
> >  and double VLAN (802.1AD) tags. Please refer to the ``struct bpf_flow_keys``
> >  for a set of information that's currently can be exported from the BPF context.
> > +
> > +When BPF flow dissector is attached to the root network namespace (machine-wide
> > +policy), users can't override it in their child network namespaces.
> > diff --git a/net/core/flow_dissector.c b/net/core/flow_dissector.c
> > index 7c09d87d3269..9821e730fc70 100644
> > --- a/net/core/flow_dissector.c
> > +++ b/net/core/flow_dissector.c
> > @@ -114,19 +114,50 @@ int skb_flow_dissector_bpf_prog_attach(const union bpf_attr *attr,
> >  {
> >         struct bpf_prog *attached;
> >         struct net *net;
> > +       int ret = 0;
> >
> >         net = current->nsproxy->net_ns;
> >         mutex_lock(&flow_dissector_mutex);
> > +
> > +       if (net == &init_net) {
> > +               /* BPF flow dissector in the root namespace overrides
> > +                * any per-net-namespace one. When attaching to root,
> > +                * make sure we don't have any BPF program attached
> > +                * to the non-root namespaces.
> > +                */
> > +               struct net *ns;
> > +
> > +               for_each_net(ns) {
> > +                       if (net == &init_net)
> > +                               continue;
> 
> You don't need this condition, if something is attached to init_net,
> you will return -EEXIST anyway. Or is this a performance optimization?
Ah, I agree, will remove an respin.

> > +
> > +                       if (rcu_access_pointer(ns->flow_dissector_prog)) {
> > +                               ret = -EEXIST;
> > +                               goto out;
> > +                       }
> > +               }
> > +       } else {
> > +               /* Make sure root flow dissector is not attached
> > +                * when attaching to the non-root namespace.
> > +                */
> > +
> 
> nit: extra empty line
Thx, will fix.

> > +               if (rcu_access_pointer(init_net.flow_dissector_prog)) {
> > +                       ret = -EEXIST;
> > +                       goto out;
> > +               }
> > +       }
> > +
> >         attached = rcu_dereference_protected(net->flow_dissector_prog,
> >                                              lockdep_is_held(&flow_dissector_mutex));
> >         if (attached) {
> >                 /* Only one BPF program can be attached at a time */
> > -               mutex_unlock(&flow_dissector_mutex);
> > -               return -EEXIST;
> > +               ret = -EEXIST;
> > +               goto out;
> >         }
> >         rcu_assign_pointer(net->flow_dissector_prog, prog);
> > +out:
> >         mutex_unlock(&flow_dissector_mutex);
> > -       return 0;
> > +       return ret;
> >  }
> >
> >  int skb_flow_dissector_bpf_prog_detach(const union bpf_attr *attr)
> > @@ -910,7 +941,10 @@ bool __skb_flow_dissect(const struct net *net,
> >         WARN_ON_ONCE(!net);
> >         if (net) {
> >                 rcu_read_lock();
> > -               attached = rcu_dereference(net->flow_dissector_prog);
> > +               attached = rcu_dereference(init_net.flow_dissector_prog);
> > +
> > +               if (!attached)
> > +                       attached = rcu_dereference(net->flow_dissector_prog);
> >
> >                 if (attached) {
> >                         struct bpf_flow_keys flow_keys;
> > --
> > 2.23.0.581.g78d2f28ef7-goog
> >