netfilter: allow exp not to be removed in nf_ct_find_expectation

Currently nf_conntrack_in() calling nf_ct_find_expectation() will
remove the exp from the hash table. However, in some scenario, we
expect the exp not to be removed when the created ct will not be
confirmed, like in OVS and TC conntrack in the following patches.

This patch allows exp not to be removed by setting IPS_CONFIRMED
in the status of the tmpl.

Signed-off-by: Xin Long <lucien.xin@gmail.com>
Acked-by: Aaron Conole <aconole@redhat.com>
Acked-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>

diff --git a/include/net/netfilter/nf_conntrack_expect.h b/include/net/netfilter/nf_conntrack_expect.h
index cf0d81b..165e7a0 100644 (file)
--- a/include/net/netfilter/nf_conntrack_expect.h
+++ b/include/net/netfilter/nf_conntrack_expect.h
@@ -100,7 +100,7 @@ nf_ct_expect_find_get(struct net *net,
 struct nf_conntrack_expect *
 nf_ct_find_expectation(struct net *net,
                       const struct nf_conntrack_zone *zone,
-                      const struct nf_conntrack_tuple *tuple);
+                      const struct nf_conntrack_tuple *tuple, bool unlink);
 
 void nf_ct_unlink_expect_report(struct nf_conntrack_expect *exp,
                                u32 portid, int report);

diff --git a/net/netfilter/nf_conntrack_core.c b/net/netfilter/nf_conntrack_core.c
index 9923931..9f6f2e6 100644 (file)
--- a/net/netfilter/nf_conntrack_core.c
+++ b/net/netfilter/nf_conntrack_core.c
@@ -1756,7 +1756,7 @@ init_conntrack(struct net *net, struct nf_conn *tmpl,
        cnet = nf_ct_pernet(net);
        if (cnet->expect_count) {
                spin_lock_bh(&nf_conntrack_expect_lock);
-               exp = nf_ct_find_expectation(net, zone, tuple);
+               exp = nf_ct_find_expectation(net, zone, tuple, !tmpl || nf_ct_is_confirmed(tmpl));
                if (exp) {
                        /* Welcome, Mr. Bond.  We've been expecting you... */
                        __set_bit(IPS_EXPECTED_BIT, &ct->status);

diff --git a/net/netfilter/nf_conntrack_expect.c b/net/netfilter/nf_conntrack_expect.c
index 96948e9..81ca348 100644 (file)
--- a/net/netfilter/nf_conntrack_expect.c
+++ b/net/netfilter/nf_conntrack_expect.c
@@ -171,7 +171,7 @@ EXPORT_SYMBOL_GPL(nf_ct_expect_find_get);
 struct nf_conntrack_expect *
 nf_ct_find_expectation(struct net *net,
                       const struct nf_conntrack_zone *zone,
-                      const struct nf_conntrack_tuple *tuple)
+                      const struct nf_conntrack_tuple *tuple, bool unlink)
 {
        struct nf_conntrack_net *cnet = nf_ct_pernet(net);
        struct nf_conntrack_expect *i, *exp = NULL;
@@ -211,7 +211,7 @@ nf_ct_find_expectation(struct net *net,
                     !refcount_inc_not_zero(&exp->master->ct_general.use)))
                return NULL;
 
-       if (exp->flags & NF_CT_EXPECT_PERMANENT) {
+       if (exp->flags & NF_CT_EXPECT_PERMANENT || !unlink) {
                refcount_inc(&exp->use);
                return exp;
        } else if (del_timer(&exp->timeout)) {

diff --git a/net/netfilter/nft_ct.c b/net/netfilter/nft_ct.c
index 38958e0..e87fd43 100644 (file)
--- a/net/netfilter/nft_ct.c
+++ b/net/netfilter/nft_ct.c
@@ -262,6 +262,7 @@ static void nft_ct_set_zone_eval(const struct nft_expr *expr,
                        regs->verdict.code = NF_DROP;
                        return;
                }
+               __set_bit(IPS_CONFIRMED_BIT, &ct->status);
        }
 
        nf_ct_set(skb, ct, IP_CT_NEW);
@@ -368,6 +369,7 @@ static bool nft_ct_tmpl_alloc_pcpu(void)
                        return false;
                }
 
+               __set_bit(IPS_CONFIRMED_BIT, &tmp->status);
                per_cpu(nft_ct_pcpu_template, cpu) = tmp;
        }