[SRU][J][PATCH 1/1] netfilter: allow exp not to be removed in nf_ct_find_expectation

[Tim Whisonant]

From: Xin Long <lucien.xin at gmail.com>

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 at gmail.com>
Acked-by: Aaron Conole <aconole at redhat.com>
Acked-by: Florian Westphal <fw at strlen.de>
Signed-off-by: Paolo Abeni <pabeni at redhat.com>
(backported from commit 4914109a8e1e494c6aa9852f9e84ec77a5fc643f)
[tswhison: adjusted context due to missing commit
0bcfbafbcd3 ("netfilter: conntrack: avoid unconditional local_bh_disable")]
CVE-2023-52927
Signed-off-by: Tim Whisonant <tim.whisonant at canonical.com>
---
 include/net/netfilter/nf_conntrack_expect.h | 2 +-
 net/netfilter/nf_conntrack_core.c           | 2 +-
 net/netfilter/nf_conntrack_expect.c         | 4 ++--
 net/netfilter/nft_ct.c                      | 2 ++
 4 files changed, 6 insertions(+), 4 deletions(-)

diff --git a/include/net/netfilter/nf_conntrack_expect.h b/include/net/netfilter/nf_conntrack_expect.h
index 0855b60fba175..f642a87ea330a 100644
--- 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 00a97743507de..e4a266cb3e5b5 100644
--- a/net/netfilter/nf_conntrack_core.c
+++ b/net/netfilter/nf_conntrack_core.c
@@ -1720,7 +1720,7 @@ init_conntrack(struct net *net, struct nf_conn *tmpl,
 	cnet = nf_ct_pernet(net);
 	if (cnet->expect_count) {
 		spin_lock(&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) {
 			pr_debug("expectation arrives ct=%p exp=%p\n",
 				 ct, exp);
diff --git a/net/netfilter/nf_conntrack_expect.c b/net/netfilter/nf_conntrack_expect.c
index 6d056ebba57c6..43bbfdfc40934 100644
--- 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 69214993b5a2c..5b5f4c8d59de1 100644
--- a/net/netfilter/nft_ct.c
+++ b/net/netfilter/nft_ct.c
@@ -271,6 +271,7 @@ 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);
@@ -377,6 +378,7 @@ 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;
 	}
 
-- 
2.43.0