Merge git://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf-next

Florian Westphal says:

====================
Netfilter updates for net-next

following patch set includes netfilter updates for your *net-next* tree.

1. Replace pr_debug use with nf_log infra for debugging in sctp
   conntrack.
2. Remove pr_debug calls, they are either useless or we have better
   options in place.
3. Avoid repeated load of ct->status in some spots.
   Some bit-flags cannot change during the lifeetime of
   a connection, so no need to re-fetch those.
4. Avoid uneeded nesting of rcu_read_lock during tuple lookup.
5. Remove the CLUSTERIP target.  Marked as obsolete for years,
   and we still have WARN splats wrt. races of the out-of-band
   /proc interface installed by this target.
6. Add static key to nf_tables to avoid the retpoline mitigation
   if/else if cascade provided the cpu doesn't need the retpoline thunk.
7. add nf_tables objref calls to the retpoline mitigation workaround.
8. Split parts of nft_ct.c that do not need symbols exported by
   the conntrack modules and place them in nf_tables directly.
   This allows to avoid indirect call for 'ct status' checks.
9. Add 'destroy' commands to nf_tables.  They are identical
   to the existing 'delete' commands, but do not indicate
   an error if the referenced object (set, chain, rule...)
   did not exist, from Fernando.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

include/net/netfilter/nf_tables_core.h

@@ -61,6 +61,16 @@ struct nft_immediate_expr {
 extern const struct nft_expr_ops nft_cmp_fast_ops;
 extern const struct nft_expr_ops nft_cmp16_fast_ops;
 
+struct nft_ct {
+	enum nft_ct_keys	key:8;
+	enum ip_conntrack_dir	dir:8;
+	u8			len;
+	union {
+		u8		dreg;
+		u8		sreg;
+	};
+};
+
 struct nft_payload {
 	enum nft_payload_bases	base:8;
 	u8			offset;
@@ -140,6 +150,8 @@ void nft_rt_get_eval(const struct nft_expr *expr,
 		     struct nft_regs *regs, const struct nft_pktinfo *pkt);
 void nft_counter_eval(const struct nft_expr *expr, struct nft_regs *regs,
                       const struct nft_pktinfo *pkt);
+void nft_ct_get_fast_eval(const struct nft_expr *expr,
+			  struct nft_regs *regs, const struct nft_pktinfo *pkt);
 
 enum {
 	NFT_PAYLOAD_CTX_INNER_TUN	= (1 << 0),
@@ -164,4 +176,8 @@ void nft_payload_inner_eval(const struct nft_expr *expr, struct nft_regs *regs,
 			    const struct nft_pktinfo *pkt,
 			    struct nft_inner_tun_ctx *ctx);
 
+void nft_objref_eval(const struct nft_expr *expr, struct nft_regs *regs,
+		     const struct nft_pktinfo *pkt);
+void nft_objref_map_eval(const struct nft_expr *expr, struct nft_regs *regs,
+			 const struct nft_pktinfo *pkt);
 #endif /* _NET_NF_TABLES_CORE_H */

include/uapi/linux/netfilter/nf_tables.h

@@ -98,6 +98,13 @@ enum nft_verdicts {
  * @NFT_MSG_GETFLOWTABLE: get flow table (enum nft_flowtable_attributes)
  * @NFT_MSG_DELFLOWTABLE: delete flow table (enum nft_flowtable_attributes)
  * @NFT_MSG_GETRULE_RESET: get rules and reset stateful expressions (enum nft_obj_attributes)
+ * @NFT_MSG_DESTROYTABLE: destroy a table (enum nft_table_attributes)
+ * @NFT_MSG_DESTROYCHAIN: destroy a chain (enum nft_chain_attributes)
+ * @NFT_MSG_DESTROYRULE: destroy a rule (enum nft_rule_attributes)
+ * @NFT_MSG_DESTROYSET: destroy a set (enum nft_set_attributes)
+ * @NFT_MSG_DESTROYSETELEM: destroy a set element (enum nft_set_elem_attributes)
+ * @NFT_MSG_DESTROYOBJ: destroy a stateful object (enum nft_object_attributes)
+ * @NFT_MSG_DESTROYFLOWTABLE: destroy flow table (enum nft_flowtable_attributes)
  */
 enum nf_tables_msg_types {
 	NFT_MSG_NEWTABLE,
@@ -126,6 +133,13 @@ enum nf_tables_msg_types {
 	NFT_MSG_GETFLOWTABLE,
 	NFT_MSG_DELFLOWTABLE,
 	NFT_MSG_GETRULE_RESET,
+	NFT_MSG_DESTROYTABLE,
+	NFT_MSG_DESTROYCHAIN,
+	NFT_MSG_DESTROYRULE,
+	NFT_MSG_DESTROYSET,
+	NFT_MSG_DESTROYSETELEM,
+	NFT_MSG_DESTROYOBJ,
+	NFT_MSG_DESTROYFLOWTABLE,
 	NFT_MSG_MAX,
 };
 

net/ipv4/netfilter/Kconfig

@@ -259,20 +259,6 @@ config IP_NF_MANGLE
 
 	  To compile it as a module, choose M here.  If unsure, say N.
 
-config IP_NF_TARGET_CLUSTERIP
-	tristate "CLUSTERIP target support"
-	depends on IP_NF_MANGLE
-	depends on NF_CONNTRACK
-	depends on NETFILTER_ADVANCED
-	select NF_CONNTRACK_MARK
-	select NETFILTER_FAMILY_ARP
-	help
-	  The CLUSTERIP target allows you to build load-balancing clusters of
-	  network servers without having a dedicated load-balancing
-	  router/server/switch.
-
-	  To compile it as a module, choose M here.  If unsure, say N.
-
 config IP_NF_TARGET_ECN
 	tristate "ECN target support"
 	depends on IP_NF_MANGLE

net/ipv4/netfilter/Makefile

@@ -39,7 +39,6 @@ obj-$(CONFIG_IP_NF_MATCH_AH) += ipt_ah.o
 obj-$(CONFIG_IP_NF_MATCH_RPFILTER) += ipt_rpfilter.o
 
 # targets
-obj-$(CONFIG_IP_NF_TARGET_CLUSTERIP) += ipt_CLUSTERIP.o
 obj-$(CONFIG_IP_NF_TARGET_ECN) += ipt_ECN.o
 obj-$(CONFIG_IP_NF_TARGET_REJECT) += ipt_REJECT.o
 obj-$(CONFIG_IP_NF_TARGET_SYNPROXY) += ipt_SYNPROXY.o

net/netfilter/Makefile

@@ -98,6 +98,12 @@ nf_tables-objs += nft_set_pipapo_avx2.o
 endif
 endif
 
+ifdef CONFIG_NFT_CT
+ifdef CONFIG_RETPOLINE
+nf_tables-objs += nft_ct_fast.o
+endif
+endif
+
 obj-$(CONFIG_NF_TABLES)		+= nf_tables.o
 obj-$(CONFIG_NFT_COMPAT)	+= nft_compat.o
 obj-$(CONFIG_NFT_CONNLIMIT)	+= nft_connlimit.o

net/netfilter/nf_conntrack_core.c

@@ -514,7 +514,6 @@ EXPORT_SYMBOL_GPL(nf_ct_get_id);
 static void
 clean_from_lists(struct nf_conn *ct)
 {
-	pr_debug("clean_from_lists(%p)\n", ct);
 	hlist_nulls_del_rcu(&ct->tuplehash[IP_CT_DIR_ORIGINAL].hnnode);
 	hlist_nulls_del_rcu(&ct->tuplehash[IP_CT_DIR_REPLY].hnnode);
 
@@ -582,7 +581,6 @@ void nf_ct_destroy(struct nf_conntrack *nfct)
 {
 	struct nf_conn *ct = (struct nf_conn *)nfct;
 
-	pr_debug("%s(%p)\n", __func__, ct);
 	WARN_ON(refcount_read(&nfct->use) != 0);
 
 	if (unlikely(nf_ct_is_template(ct))) {
@@ -603,7 +601,6 @@ void nf_ct_destroy(struct nf_conntrack *nfct)
 	if (ct->master)
 		nf_ct_put(ct->master);
 
-	pr_debug("%s: returning ct=%p to slab\n", __func__, ct);
 	nf_conntrack_free(ct);
 }
 EXPORT_SYMBOL(nf_ct_destroy);
@@ -786,8 +783,6 @@ __nf_conntrack_find_get(struct net *net, const struct nf_conntrack_zone *zone,
 	struct nf_conntrack_tuple_hash *h;
 	struct nf_conn *ct;
 
-	rcu_read_lock();
-
 	h = ____nf_conntrack_find(net, zone, tuple, hash);
 	if (h) {
 		/* We have a candidate that matches the tuple we're interested
@@ -799,7 +794,7 @@ __nf_conntrack_find_get(struct net *net, const struct nf_conntrack_zone *zone,
 			smp_acquire__after_ctrl_dep();
 
 			if (likely(nf_ct_key_equal(h, tuple, zone, net)))
-				goto found;
+				return h;
 
 			/* TYPESAFE_BY_RCU recycled the candidate */
 			nf_ct_put(ct);
@@ -807,8 +802,6 @@ __nf_conntrack_find_get(struct net *net, const struct nf_conntrack_zone *zone,
 
 		h = NULL;
 	}
-found:
-	rcu_read_unlock();
 
 	return h;
 }
@@ -820,16 +813,21 @@ nf_conntrack_find_get(struct net *net, const struct nf_conntrack_zone *zone,
 	unsigned int rid, zone_id = nf_ct_zone_id(zone, IP_CT_DIR_ORIGINAL);
 	struct nf_conntrack_tuple_hash *thash;
 
+	rcu_read_lock();
+
 	thash = __nf_conntrack_find_get(net, zone, tuple,
 					hash_conntrack_raw(tuple, zone_id, net));
 
 	if (thash)
-		return thash;
+		goto out_unlock;
 
 	rid = nf_ct_zone_id(zone, IP_CT_DIR_REPLY);
 	if (rid != zone_id)
-		return __nf_conntrack_find_get(net, zone, tuple,
-					       hash_conntrack_raw(tuple, rid, net));
+		thash = __nf_conntrack_find_get(net, zone, tuple,
+						hash_conntrack_raw(tuple, rid, net));
+
+out_unlock:
+	rcu_read_unlock();
 	return thash;
 }
 EXPORT_SYMBOL_GPL(nf_conntrack_find_get);
@@ -1210,7 +1208,6 @@ __nf_conntrack_confirm(struct sk_buff *skb)
 		goto dying;
 	}
 
-	pr_debug("Confirming conntrack %p\n", ct);
 	/* We have to check the DYING flag after unlink to prevent
 	 * a race against nf_ct_get_next_corpse() possibly called from
 	 * user context, else we insert an already 'dead' hash, blocking
@@ -1721,10 +1718,8 @@ init_conntrack(struct net *net, struct nf_conn *tmpl,
 	struct nf_conntrack_zone tmp;
 	struct nf_conntrack_net *cnet;
 
-	if (!nf_ct_invert_tuple(&repl_tuple, tuple)) {
-		pr_debug("Can't invert tuple.\n");
+	if (!nf_ct_invert_tuple(&repl_tuple, tuple))
 		return NULL;
-	}
 
 	zone = nf_ct_zone_tmpl(tmpl, skb, &tmp);
 	ct = __nf_conntrack_alloc(net, zone, tuple, &repl_tuple, GFP_ATOMIC,
@@ -1764,8 +1759,6 @@ init_conntrack(struct net *net, struct nf_conn *tmpl,
 		spin_lock_bh(&nf_conntrack_expect_lock);
 		exp = nf_ct_find_expectation(net, zone, tuple);
 		if (exp) {
-			pr_debug("expectation arrives ct=%p exp=%p\n",
-				 ct, exp);
 			/* Welcome, Mr. Bond.  We've been expecting you... */
 			__set_bit(IPS_EXPECTED_BIT, &ct->status);
 			/* exp->master safe, refcnt bumped in nf_ct_find_expectation */
@@ -1829,10 +1822,8 @@ resolve_normal_ct(struct nf_conn *tmpl,
 
 	if (!nf_ct_get_tuple(skb, skb_network_offset(skb),
 			     dataoff, state->pf, protonum, state->net,
-			     &tuple)) {
-		pr_debug("Can't get tuple\n");
+			     &tuple))
 		return 0;
-	}
 
 	/* look for tuple match */
 	zone = nf_ct_zone_tmpl(tmpl, skb, &tmp);
@@ -1864,17 +1855,15 @@ resolve_normal_ct(struct nf_conn *tmpl,
 	if (NF_CT_DIRECTION(h) == IP_CT_DIR_REPLY) {
 		ctinfo = IP_CT_ESTABLISHED_REPLY;
 	} else {
+		unsigned long status = READ_ONCE(ct->status);
+
 		/* Once we've had two way comms, always ESTABLISHED. */
-		if (test_bit(IPS_SEEN_REPLY_BIT, &ct->status)) {
-			pr_debug("normal packet for %p\n", ct);
+		if (likely(status & IPS_SEEN_REPLY))
 			ctinfo = IP_CT_ESTABLISHED;
-		} else if (test_bit(IPS_EXPECTED_BIT, &ct->status)) {
-			pr_debug("related packet for %p\n", ct);
+		else if (status & IPS_EXPECTED)
 			ctinfo = IP_CT_RELATED;
-		} else {
-			pr_debug("new packet for %p\n", ct);
+		else
 			ctinfo = IP_CT_NEW;
-		}
 	}
 	nf_ct_set(skb, ct, ctinfo);
 	return 0;
@@ -1988,7 +1977,6 @@ nf_conntrack_in(struct sk_buff *skb, const struct nf_hook_state *state)
 	/* rcu_read_lock()ed by nf_hook_thresh */
 	dataoff = get_l4proto(skb, skb_network_offset(skb), state->pf, &protonum);
 	if (dataoff <= 0) {
-		pr_debug("not prepared to track yet or error occurred\n");
 		NF_CT_STAT_INC_ATOMIC(state->net, invalid);
 		ret = NF_ACCEPT;
 		goto out;
@@ -2027,7 +2015,6 @@ nf_conntrack_in(struct sk_buff *skb, const struct nf_hook_state *state)
 	if (ret <= 0) {
 		/* Invalid: inverse of the return code tells
 		 * the netfilter core what to do */
-		pr_debug("nf_conntrack_in: Can't track with proto module\n");
 		nf_ct_put(ct);
 		skb->_nfct = 0;
 		/* Special case: TCP tracker reports an attempt to reopen a
@@ -2066,7 +2053,6 @@ void nf_conntrack_alter_reply(struct nf_conn *ct,
 	/* Should be unconfirmed, so not in hash table yet */
 	WARN_ON(nf_ct_is_confirmed(ct));
 
-	pr_debug("Altering reply tuple of %p to ", ct);
 	nf_ct_dump_tuple(newreply);
 
 	ct->tuplehash[IP_CT_DIR_REPLY].tuple = *newreply;

net/netfilter/nf_conntrack_proto.c

@@ -284,16 +284,11 @@ getorigdst(struct sock *sk, int optval, void __user *user, int *len)
 
 	/* We only do TCP and SCTP at the moment: is there a better way? */
 	if (tuple.dst.protonum != IPPROTO_TCP &&
-	    tuple.dst.protonum != IPPROTO_SCTP) {
-		pr_debug("SO_ORIGINAL_DST: Not a TCP/SCTP socket\n");
+	    tuple.dst.protonum != IPPROTO_SCTP)
 		return -ENOPROTOOPT;
-	}
 
-	if ((unsigned int)*len < sizeof(struct sockaddr_in)) {
-		pr_debug("SO_ORIGINAL_DST: len %d not %zu\n",
-			 *len, sizeof(struct sockaddr_in));
+	if ((unsigned int)*len < sizeof(struct sockaddr_in))
 		return -EINVAL;
-	}
 
 	h = nf_conntrack_find_get(sock_net(sk), &nf_ct_zone_dflt, &tuple);
 	if (h) {
@@ -307,17 +302,12 @@ getorigdst(struct sock *sk, int optval, void __user *user, int *len)
 			.tuple.dst.u3.ip;
 		memset(sin.sin_zero, 0, sizeof(sin.sin_zero));
 
-		pr_debug("SO_ORIGINAL_DST: %pI4 %u\n",
-			 &sin.sin_addr.s_addr, ntohs(sin.sin_port));
 		nf_ct_put(ct);
 		if (copy_to_user(user, &sin, sizeof(sin)) != 0)
 			return -EFAULT;
 		else
 			return 0;
 	}
-	pr_debug("SO_ORIGINAL_DST: Can't find %pI4/%u-%pI4/%u.\n",
-		 &tuple.src.u3.ip, ntohs(tuple.src.u.tcp.port),
-		 &tuple.dst.u3.ip, ntohs(tuple.dst.u.tcp.port));
 	return -ENOENT;
 }
 
@@ -360,12 +350,8 @@ ipv6_getorigdst(struct sock *sk, int optval, void __user *user, int *len)
 		return -EINVAL;
 
 	h = nf_conntrack_find_get(sock_net(sk), &nf_ct_zone_dflt, &tuple);
-	if (!h) {
-		pr_debug("IP6T_SO_ORIGINAL_DST: Can't find %pI6c/%u-%pI6c/%u.\n",
-			 &tuple.src.u3.ip6, ntohs(tuple.src.u.tcp.port),
-			 &tuple.dst.u3.ip6, ntohs(tuple.dst.u.tcp.port));
+	if (!h)
 		return -ENOENT;
-	}
 
 	ct = nf_ct_tuplehash_to_ctrack(h);
 

net/netfilter/nf_conntrack_proto_sctp.c

@@ -168,7 +168,8 @@ for ((offset) = (dataoff) + sizeof(struct sctphdr), (count) = 0;	\
 static int do_basic_checks(struct nf_conn *ct,
 			   const struct sk_buff *skb,
 			   unsigned int dataoff,
-			   unsigned long *map)
+			   unsigned long *map,
+			   const struct nf_hook_state *state)
 {
 	u_int32_t offset, count;
 	struct sctp_chunkhdr _sch, *sch;
@@ -177,8 +178,6 @@ static int do_basic_checks(struct nf_conn *ct,
 	flag = 0;
 
 	for_each_sctp_chunk (skb, sch, _sch, offset, dataoff, count) {
-		pr_debug("Chunk Num: %d  Type: %d\n", count, sch->type);
-
 		if (sch->type == SCTP_CID_INIT ||
 		    sch->type == SCTP_CID_INIT_ACK ||
 		    sch->type == SCTP_CID_SHUTDOWN_COMPLETE)
@@ -193,7 +192,9 @@ static int do_basic_checks(struct nf_conn *ct,
 		      sch->type == SCTP_CID_COOKIE_ECHO ||
 		      flag) &&
 		     count != 0) || !sch->length) {
-			pr_debug("Basic checks failed\n");
+			nf_ct_l4proto_log_invalid(skb, ct, state,
+						  "%s failed. chunk num %d, type %d, len %d flag %d\n",
+						  __func__, count, sch->type, sch->length, flag);
 			return 1;
 		}
 
@@ -201,7 +202,6 @@ static int do_basic_checks(struct nf_conn *ct,
 			set_bit(sch->type, map);
 	}
 
-	pr_debug("Basic checks passed\n");
 	return count == 0;
 }
 
@@ -211,69 +211,51 @@ static int sctp_new_state(enum ip_conntrack_dir dir,
 {
 	int i;
 
-	pr_debug("Chunk type: %d\n", chunk_type);
-
 	switch (chunk_type) {
 	case SCTP_CID_INIT:
-		pr_debug("SCTP_CID_INIT\n");
 		i = 0;
 		break;
 	case SCTP_CID_INIT_ACK:
-		pr_debug("SCTP_CID_INIT_ACK\n");
 		i = 1;
 		break;
 	case SCTP_CID_ABORT:
-		pr_debug("SCTP_CID_ABORT\n");
 		i = 2;
 		break;
 	case SCTP_CID_SHUTDOWN:
-		pr_debug("SCTP_CID_SHUTDOWN\n");
 		i = 3;
 		break;
 	case SCTP_CID_SHUTDOWN_ACK:
-		pr_debug("SCTP_CID_SHUTDOWN_ACK\n");
 		i = 4;
 		break;
 	case SCTP_CID_ERROR:
-		pr_debug("SCTP_CID_ERROR\n");
 		i = 5;
 		break;
 	case SCTP_CID_COOKIE_ECHO:
-		pr_debug("SCTP_CID_COOKIE_ECHO\n");
 		i = 6;
 		break;
 	case SCTP_CID_COOKIE_ACK:
-		pr_debug("SCTP_CID_COOKIE_ACK\n");
 		i = 7;
 		break;
 	case SCTP_CID_SHUTDOWN_COMPLETE:
-		pr_debug("SCTP_CID_SHUTDOWN_COMPLETE\n");
 		i = 8;
 		break;
 	case SCTP_CID_HEARTBEAT:
-		pr_debug("SCTP_CID_HEARTBEAT");
 		i = 9;
 		break;
 	case SCTP_CID_HEARTBEAT_ACK:
-		pr_debug("SCTP_CID_HEARTBEAT_ACK");
 		i = 10;
 		break;
 	case SCTP_CID_DATA:
 	case SCTP_CID_SACK:
-		pr_debug("SCTP_CID_DATA/SACK");
 		i = 11;
 		break;
 	default:
 		/* Other chunks like DATA or SACK do not change the state */
-		pr_debug("Unknown chunk type, Will stay in %s\n",
-			 sctp_conntrack_names[cur_state]);
+		pr_debug("Unknown chunk type %d, Will stay in %s\n",
+			 chunk_type, sctp_conntrack_names[cur_state]);
 		return cur_state;
 	}
 
-	pr_debug("dir: %d   cur_state: %s  chunk_type: %d  new_state: %s\n",
-		 dir, sctp_conntrack_names[cur_state], chunk_type,
-		 sctp_conntrack_names[sctp_conntracks[dir][i][cur_state]]);
-
 	return sctp_conntracks[dir][i][cur_state];
 }
 
@@ -392,7 +374,7 @@ int nf_conntrack_sctp_packet(struct nf_conn *ct,
 	if (sh == NULL)
 		goto out;
 
-	if (do_basic_checks(ct, skb, dataoff, map) != 0)
+	if (do_basic_checks(ct, skb, dataoff, map, state) != 0)
 		goto out;
 
 	if (!nf_ct_is_confirmed(ct)) {
@@ -414,7 +396,9 @@ int nf_conntrack_sctp_packet(struct nf_conn *ct,
 		    !test_bit(SCTP_CID_HEARTBEAT, map) &&
 		    !test_bit(SCTP_CID_HEARTBEAT_ACK, map) &&
 		    sh->vtag != ct->proto.sctp.vtag[dir]) {
-			pr_debug("Verification tag check failed\n");
+			nf_ct_l4proto_log_invalid(skb, ct, state,
+						  "verification tag check failed %x vs %x for dir %d",
+						  sh->vtag, ct->proto.sctp.vtag[dir], dir);
 			goto out;
 		}
 	}
@@ -488,9 +472,10 @@ int nf_conntrack_sctp_packet(struct nf_conn *ct,
 
 		/* Invalid */
 		if (new_state == SCTP_CONNTRACK_MAX) {
-			pr_debug("nf_conntrack_sctp: Invalid dir=%i ctype=%u "
-				 "conntrack=%u\n",
-				 dir, sch->type, old_state);
+			nf_ct_l4proto_log_invalid(skb, ct, state,
+						  "Invalid, old_state %d, dir %d, type %d",
+						  old_state, dir, sch->type);
+
 			goto out_unlock;
 		}
 
@@ -536,7 +521,6 @@ int nf_conntrack_sctp_packet(struct nf_conn *ct,
 	if (old_state == SCTP_CONNTRACK_COOKIE_ECHOED &&
 	    dir == IP_CT_DIR_REPLY &&
 	    new_state == SCTP_CONNTRACK_ESTABLISHED) {
-		pr_debug("Setting assured bit\n");
 		set_bit(IPS_ASSURED_BIT, &ct->status);
 		nf_conntrack_event_cache(IPCT_ASSURED, ct);
 	}

net/netfilter/nf_conntrack_proto_tcp.c

@@ -930,7 +930,6 @@ int nf_conntrack_tcp_packet(struct nf_conn *ct,
 {
 	struct net *net = nf_ct_net(ct);
 	struct nf_tcp_net *tn = nf_tcp_pernet(net);
-	struct nf_conntrack_tuple *tuple;
 	enum tcp_conntrack new_state, old_state;
 	unsigned int index, *timeouts;
 	enum nf_ct_tcp_action res;
@@ -954,7 +953,6 @@ int nf_conntrack_tcp_packet(struct nf_conn *ct,
 	dir = CTINFO2DIR(ctinfo);
 	index = get_conntrack_index(th);
 	new_state = tcp_conntracks[dir][index][old_state];
-	tuple = &ct->tuplehash[dir].tuple;
 
 	switch (new_state) {
 	case TCP_CONNTRACK_SYN_SENT:
@@ -1217,13 +1215,6 @@ int nf_conntrack_tcp_packet(struct nf_conn *ct,
 	ct->proto.tcp.last_index = index;
 	ct->proto.tcp.last_dir = dir;
 
-	pr_debug("tcp_conntracks: ");
-	nf_ct_dump_tuple(tuple);
-	pr_debug("syn=%i ack=%i fin=%i rst=%i old=%i new=%i\n",
-		 (th->syn ? 1 : 0), (th->ack ? 1 : 0),
-		 (th->fin ? 1 : 0), (th->rst ? 1 : 0),
-		 old_state, new_state);
-
 	ct->proto.tcp.state = new_state;
 	if (old_state != new_state
 	    && new_state == TCP_CONNTRACK_FIN_WAIT)

net/netfilter/nf_conntrack_proto_udp.c

@@ -88,6 +88,7 @@ int nf_conntrack_udp_packet(struct nf_conn *ct,
 			    const struct nf_hook_state *state)
 {
 	unsigned int *timeouts;
+	unsigned long status;
 
 	if (udp_error(skb, dataoff, state))
 		return -NF_ACCEPT;
@@ -96,26 +97,27 @@ int nf_conntrack_udp_packet(struct nf_conn *ct,
 	if (!timeouts)
 		timeouts = udp_get_timeouts(nf_ct_net(ct));
 
-	if (!nf_ct_is_confirmed(ct))
+	status = READ_ONCE(ct->status);
+	if ((status & IPS_CONFIRMED) == 0)
 		ct->proto.udp.stream_ts = 2 * HZ + jiffies;
 
 	/* If we've seen traffic both ways, this is some kind of UDP
 	 * stream. Set Assured.
 	 */
-	if (test_bit(IPS_SEEN_REPLY_BIT, &ct->status)) {
+	if (status & IPS_SEEN_REPLY_BIT) {
 		unsigned long extra = timeouts[UDP_CT_UNREPLIED];
 		bool stream = false;
 
 		/* Still active after two seconds? Extend timeout. */
 		if (time_after(jiffies, ct->proto.udp.stream_ts)) {
 			extra = timeouts[UDP_CT_REPLIED];
-			stream = true;
+			stream = (status & IPS_ASSURED) == 0;
 		}
 
 		nf_ct_refresh_acct(ct, ctinfo, skb, extra);
 
 		/* never set ASSURED for IPS_NAT_CLASH, they time out soon */
-		if (unlikely((ct->status & IPS_NAT_CLASH)))
+		if (unlikely((status & IPS_NAT_CLASH)))
 			return NF_ACCEPT;
 
 		/* Also, more likely to be important, and not a probe */

net/netfilter/nf_tables_core.c

@@ -21,6 +21,26 @@
 #include <net/netfilter/nf_log.h>
 #include <net/netfilter/nft_meta.h>
 
+#if defined(CONFIG_RETPOLINE) && defined(CONFIG_X86)
+
+static struct static_key_false nf_tables_skip_direct_calls;
+
+static bool nf_skip_indirect_calls(void)
+{
+	return static_branch_likely(&nf_tables_skip_direct_calls);
+}
+
+static void __init nf_skip_indirect_calls_enable(void)
+{
+	if (!cpu_feature_enabled(X86_FEATURE_RETPOLINE))
+		static_branch_enable(&nf_tables_skip_direct_calls);
+}
+#else
+static inline bool nf_skip_indirect_calls(void) { return false; }
+
+static inline void nf_skip_indirect_calls_enable(void) { }
+#endif
+
 static noinline void __nft_trace_packet(struct nft_traceinfo *info,
 					const struct nft_chain *chain,
 					enum nft_trace_types type)
@@ -193,7 +213,12 @@ static void expr_call_ops_eval(const struct nft_expr *expr,
 			       struct nft_pktinfo *pkt)
 {
 #ifdef CONFIG_RETPOLINE
-	unsigned long e = (unsigned long)expr->ops->eval;
+	unsigned long e;
+
+	if (nf_skip_indirect_calls())
+		goto indirect_call;
+
+	e = (unsigned long)expr->ops->eval;
 #define X(e, fun) \
 	do { if ((e) == (unsigned long)(fun)) \
 		return fun(expr, regs, pkt); } while (0)
@@ -203,13 +228,19 @@ static void expr_call_ops_eval(const struct nft_expr *expr,
 	X(e, nft_counter_eval);
 	X(e, nft_meta_get_eval);
 	X(e, nft_lookup_eval);
+#if IS_ENABLED(CONFIG_NFT_CT)
+	X(e, nft_ct_get_fast_eval);
+#endif
 	X(e, nft_range_eval);
 	X(e, nft_immediate_eval);
 	X(e, nft_byteorder_eval);
 	X(e, nft_dynset_eval);
 	X(e, nft_rt_get_eval);
 	X(e, nft_bitwise_eval);
+	X(e, nft_objref_eval);
+	X(e, nft_objref_map_eval);
 #undef  X
+indirect_call:
 #endif /* CONFIG_RETPOLINE */
 	expr->ops->eval(expr, regs, pkt);
 }
@@ -369,6 +400,8 @@ int __init nf_tables_core_module_init(void)
 			goto err;
 	}
 
+	nf_skip_indirect_calls_enable();
+
 	return 0;
 
 err:

net/netfilter/nft_ct.c

@@ -12,7 +12,7 @@
 #include <linux/netlink.h>
 #include <linux/netfilter.h>
 #include <linux/netfilter/nf_tables.h>
-#include <net/netfilter/nf_tables.h>
+#include <net/netfilter/nf_tables_core.h>
 #include <net/netfilter/nf_conntrack.h>
 #include <net/netfilter/nf_conntrack_acct.h>
 #include <net/netfilter/nf_conntrack_tuple.h>
@@ -23,16 +23,6 @@
 #include <net/netfilter/nf_conntrack_l4proto.h>
 #include <net/netfilter/nf_conntrack_expect.h>
 
-struct nft_ct {
-	enum nft_ct_keys	key:8;
-	enum ip_conntrack_dir	dir:8;
-	u8			len;
-	union {
-		u8		dreg;
-		u8		sreg;
-	};
-};
-
 struct nft_ct_helper_obj  {
 	struct nf_conntrack_helper *helper4;
 	struct nf_conntrack_helper *helper6;
@@ -759,6 +749,18 @@ static bool nft_ct_set_reduce(struct nft_regs_track *track,
 	return false;
 }
 
+#ifdef CONFIG_RETPOLINE
+static const struct nft_expr_ops nft_ct_get_fast_ops = {
+	.type		= &nft_ct_type,
+	.size		= NFT_EXPR_SIZE(sizeof(struct nft_ct)),
+	.eval		= nft_ct_get_fast_eval,
+	.init		= nft_ct_get_init,
+	.destroy	= nft_ct_get_destroy,
+	.dump		= nft_ct_get_dump,
+	.reduce		= nft_ct_set_reduce,
+};
+#endif
+
 static const struct nft_expr_ops nft_ct_set_ops = {
 	.type		= &nft_ct_type,
 	.size		= NFT_EXPR_SIZE(sizeof(struct nft_ct)),
@@ -791,8 +793,21 @@ nft_ct_select_ops(const struct nft_ctx *ctx,
 	if (tb[NFTA_CT_DREG] && tb[NFTA_CT_SREG])
 		return ERR_PTR(-EINVAL);
 
-	if (tb[NFTA_CT_DREG])
+	if (tb[NFTA_CT_DREG]) {
+#ifdef CONFIG_RETPOLINE
+		u32 k = ntohl(nla_get_be32(tb[NFTA_CT_KEY]));
+
+		switch (k) {
+		case NFT_CT_STATE:
+		case NFT_CT_DIRECTION:
+		case NFT_CT_STATUS:
+		case NFT_CT_MARK:
+		case NFT_CT_SECMARK:
+			return &nft_ct_get_fast_ops;
+		}
+#endif
 		return &nft_ct_get_ops;
+	}
 
 	if (tb[NFTA_CT_SREG]) {
 #ifdef CONFIG_NF_CONNTRACK_ZONES

net/netfilter/nft_ct_fast.c

+// SPDX-License-Identifier: GPL-2.0-only
+#if IS_ENABLED(CONFIG_NFT_CT)
+#include <linux/netfilter/nf_tables.h>
+#include <net/netfilter/nf_tables_core.h>
+#include <net/netfilter/nf_conntrack.h>
+
+void nft_ct_get_fast_eval(const struct nft_expr *expr,
+			  struct nft_regs *regs,
+			  const struct nft_pktinfo *pkt)
+{
+	const struct nft_ct *priv = nft_expr_priv(expr);
+	u32 *dest = &regs->data[priv->dreg];
+	enum ip_conntrack_info ctinfo;
+	const struct nf_conn *ct;
+	unsigned int state;
+
+	ct = nf_ct_get(pkt->skb, &ctinfo);
+	if (!ct) {
+		regs->verdict.code = NFT_BREAK;
+		return;
+	}
+
+	switch (priv->key) {
+	case NFT_CT_STATE:
+		if (ct)
+			state = NF_CT_STATE_BIT(ctinfo);
+		else if (ctinfo == IP_CT_UNTRACKED)
+			state = NF_CT_STATE_UNTRACKED_BIT;
+		else
+			state = NF_CT_STATE_INVALID_BIT;
+		*dest = state;
+		return;
+	case NFT_CT_DIRECTION:
+		nft_reg_store8(dest, CTINFO2DIR(ctinfo));
+		return;
+	case NFT_CT_STATUS:
+		*dest = ct->status;
+		return;
+#ifdef CONFIG_NF_CONNTRACK_MARK
+	case NFT_CT_MARK:
+		*dest = ct->mark;
+		return;
+#endif
+#ifdef CONFIG_NF_CONNTRACK_SECMARK
+	case NFT_CT_SECMARK:
+		*dest = ct->secmark;
+		return;
+#endif
+	default:
+		WARN_ON_ONCE(1);
+		regs->verdict.code = NFT_BREAK;
+		break;
+	}
+}
+EXPORT_SYMBOL_GPL(nft_ct_get_fast_eval);
+#endif

net/netfilter/nft_objref.c

@@ -13,9 +13,9 @@
 
 #define nft_objref_priv(expr)	*((struct nft_object **)nft_expr_priv(expr))
 
-static void nft_objref_eval(const struct nft_expr *expr,
-			    struct nft_regs *regs,
-			    const struct nft_pktinfo *pkt)
+void nft_objref_eval(const struct nft_expr *expr,
+		     struct nft_regs *regs,
+		     const struct nft_pktinfo *pkt)
 {
 	struct nft_object *obj = nft_objref_priv(expr);
 
@@ -100,9 +100,9 @@ struct nft_objref_map {
 	struct nft_set_binding	binding;
 };
 
-static void nft_objref_map_eval(const struct nft_expr *expr,
-				struct nft_regs *regs,
-				const struct nft_pktinfo *pkt)
+void nft_objref_map_eval(const struct nft_expr *expr,
+			 struct nft_regs *regs,
+			 const struct nft_pktinfo *pkt)
 {
 	struct nft_objref_map *priv = nft_expr_priv(expr);
 	const struct nft_set *set = priv->set;