On Mon, 4 Mar 2019 at 17:43, Martin Lau <ka...@fb.com> wrote:
>
> On Mon, Mar 04, 2019 at 10:33:46AM +0100, Daniel Borkmann wrote:
> > On 03/02/2019 09:21 PM, Martin Lau wrote:
> > > On Sat, Mar 02, 2019 at 10:03:03AM -0800, Alexei Starovoitov wrote:
> > >> On Sat, Mar 02, 2019 at 08:10:10AM -0800, Martin KaFai Lau wrote:
> > >>> Lorenz Bauer [thanks!] reported that a ptr returned by bpf_tcp_sock(sk)
> > >>> can still be accessed after bpf_sk_release(sk).
> > >>> Both bpf_tcp_sock() and bpf_sk_fullsock() have the same issue.
> > >>> This patch addresses them together.
> > >>>
> > >>> A simple reproducer looks like this:
> > >>>
> > >>> sk = bpf_sk_lookup_tcp();
> > >>> /* if (!sk) ... */
> > >>> tp = bpf_tcp_sock(sk);
> > >>> /* if (!tp) ... */
> > >>> bpf_sk_release(sk);
> > >>> snd_cwnd = tp->snd_cwnd; /* oops! The verifier does not complain. */
> > >>>
> > >>> The problem is the verifier did not scrub the register's states of
> > >>> the tcp_sock ptr (tp) after bpf_sk_release(sk).
> > >>>
> > >>> [ Note that when calling bpf_tcp_sock(sk), the sk is not always
> > >>> refcount-acquired. e.g. bpf_tcp_sock(skb->sk). The verifier works
> > >>> fine for this case. ]
> > >>>
> > >>> Currently, the verifier does not track if a helper's return ptr (in
> > >>> REG_0)
> > >>> is "carry"-ing one of its argument's refcount status. To carry this
> > >>> info,
> > >>> the reg1->id needs to be stored in reg0. The reg0->id has already
> > >>> been used for NULL checking purpose. Hence, a new "refcount_id"
> > >>> is needed in "struct bpf_reg_state".
> > >>>
> > >>> With refcount_id, when bpf_sk_release(sk) is called, the verifier can
> > >>> scrub
> > >>> all reg states which has a refcount_id match. It is done with the
> > >>> changes
> > >>> in release_reg_references().
> > >>>
> > >>> When acquiring and releasing a refcount, the reg->id is still used.
> > >>> Hence, we cannot do "bpf_sk_release(tp)" in the above reproducer
> > >>> example.
> > >>
> > >> I think the choice of returning listener full sock from req sock
> > >> in sk_to_full_sk() was a wrong one.
> > >> It seems better to make semantics of bpf_tcp_sock() and
> > >> bpf_sk_fullsock() as
> > >> always type cast or null.
> > >> And have a separate helper for req socket that returns
> > >> inet_reqsk(sk)->rsk_listener.
> > >>
> > >> Then it will be ok to call bpf_sk_release(tp) when tp came from
> > >> bpf_sk_lookup_tcp.
> > >> The verifier will know that it's the case because its ID will be in
> > >> acquired_refs.
> > >>
> > >> The additional refcount_id won't be necessary.
> > >> bpf_sk_fullsock() and bpf_tcp_sock() will not call sk_to_full_sk
> > >> and the verifier will be copying reg1->id into reg0->id.
> > >>
> > >> In release_reference() the verifier will do
> > >> if (regs[i].id == id)
> > >> mark_reg_unknown(env, regs, i);
> > >> for all socket types.
> > >>
> > >> release_reference_state() will stay as-is.
> > >>
> > >> imo such logic will be easier to follow.
> > >>
> > >> This implicit sk_to_full_sk() makes the whole thing much harder for the
> > >> verifier
> > >> and for the bpf program writers.
> > >>
> > >> The new bpf_get_listener_sock(sk) doesn't have to copy ID from reg1 to
> > >> reg0
> > >> since req socket will not be returned from bpf_sk_lookup_tcp and its ID
> > >> will not be stored in acuired_refs.
> > >>
> > >> Does it make sense ?
> > > I like this idea. Many thanks for thinking it through!
> > >
> > > Allowing bpf_sk_release(tp), no need to call bpf_sk_release() on ptr
> > > returned from bpf_get_listener_sock(sk) and keep one reg->id.
> > >
> > > I think it should work. I will rework the patches.
> >
> > Agree, makes sense, that seems much better fix.
> While I was working on this change, based on the code, one issue I saw is:
>
> if the bpf prog does this:
>
> sk = bpf_sk_lookup_tcp();
> /* if (!sk) ... */
> fullsock = bpf_sk_fullsock(sk);
> if (!fullsock) {
> bpf_sk_release(sk); /* Fail. sk_reg->id not found in ref state */
> return 0;
> }
>
> The bpf_sk_release(sk) failed because the reference state has already
> been released by "release_reference_state(state, fullsock_reg->id)" during
> "if (!fullsock) /* handled by mark_ptr_or_null_regs(is_null == true) */"
> Logically, I think bpf_sk_release(sk) should not fail regardless of
> bpf_sk_fullsock() doing sk_to_full_sk() or not.
>
> bpf_sk_fullsock() could disallow PTR_TO_SOCKET or PTR_TO_TCP_SOCK but that
> would be weird.
>
> I think we still need two id. May be rename the refcount_id proposed in
> this patch to ref_obj_id which is the original refcounted object id.
>
> If the sk_to_full_sk() is removed from bpf_sk_fullsock() and bpf_tcp_sock(),
> these two helpers become a simple cast (i.e. either return the same pointer
> or NULL). Then bpf_sk_release(fullsock) and bpf_sk_release(tp) could work:
>
> - When is_null == true, release_reference_state(state, reg->id) is called.
If I understand correctly, this works because we never
acquire_reference() for tp/ fullsock,
making this a no-op?
> - During bpf_sk_release(fullsock), release_reference(env, reg->ref_obj_id)
> is called so that sk, fullsock and tp with the same ref_obj_id will
> be mark_reg_unknown().
To clarify, the following states are possible:
* id == 0, ref_obj_id == 0: not a pointer / reference
* id != 0, reg_obj_id == 0: a reference which didn't have
acquire_reference() called
* id != 0, reg_obj_id != 0: a reference which had acquire_reference() called
* id == 0, reg_obj_id: illegal
Sounds reasonable to me.
--
Lorenz Bauer | Systems Engineer
25 Lavington St., London SE1 0NZ
www.cloudflare.com
