Merge branch 'main' of git://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf-next

Florian Westphal says:

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

1. nf_tables 'brouting' support, from Sriram Yagnaraman.

2. Update bridge netfilter and ovs conntrack helpers to handle
   IPv6 Jumbo packets properly, i.e. fetch the packet length
   from hop-by-hop extension header, from Xin Long.

   This comes with a test BIG TCP test case, added to
   tools/testing/selftests/net/.

3. Fix spelling and indentation in conntrack, from Jeremy Sowden.

* 'main' of git://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf-next:
  netfilter: nat: fix indentation of function arguments
  netfilter: conntrack: fix typo
  selftests: add a selftest for big tcp
  netfilter: use nf_ip6_check_hbh_len in nf_ct_skb_network_trim
  netfilter: move br_nf_check_hbh_len to utils
  netfilter: bridge: move pskb_trim_rcsum out of br_nf_check_hbh_len
  netfilter: bridge: check len before accessing more nh data
  netfilter: bridge: call pskb_may_pull in br_nf_check_hbh_len
  netfilter: bridge: introduce broute meta statement
====================

Link: https://lore.kernel.org/r/20230308193033.13965-1-fw@strlen.deSigned-off-by: Jakub Kicinski <kuba@kernel.org>

include/linux/netfilter_ipv6.h

@@ -197,6 +197,8 @@ static inline int nf_cookie_v6_check(const struct ipv6hdr *iph,
 __sum16 nf_ip6_checksum(struct sk_buff *skb, unsigned int hook,
 			unsigned int dataoff, u_int8_t protocol);
 
+int nf_ip6_check_hbh_len(struct sk_buff *skb, u32 *plen);
+
 int ipv6_netfilter_init(void);
 void ipv6_netfilter_fini(void);
 

include/uapi/linux/netfilter/nf_tables.h

@@ -931,6 +931,7 @@ enum nft_exthdr_attributes {
  * @NFT_META_TIME_HOUR: hour of day (in seconds)
  * @NFT_META_SDIF: slave device interface index
  * @NFT_META_SDIFNAME: slave device interface name
+ * @NFT_META_BRI_BROUTE: packet br_netfilter_broute bit
  */
 enum nft_meta_keys {
 	NFT_META_LEN,
@@ -969,6 +970,7 @@ enum nft_meta_keys {
 	NFT_META_TIME_HOUR,
 	NFT_META_SDIF,
 	NFT_META_SDIFNAME,
+	NFT_META_BRI_BROUTE,
 	__NFT_META_IIFTYPE,
 };
 

net/bridge/br_netfilter_ipv6.c

@@ -40,62 +40,6 @@
 #include <linux/sysctl.h>
 #endif
 
-/* We only check the length. A bridge shouldn't do any hop-by-hop stuff
- * anyway
- */
-static int br_nf_check_hbh_len(struct sk_buff *skb)
-{
-	unsigned char *raw = (u8 *)(ipv6_hdr(skb) + 1);
-	u32 pkt_len;
-	const unsigned char *nh = skb_network_header(skb);
-	int off = raw - nh;
-	int len = (raw[1] + 1) << 3;
-
-	if ((raw + len) - skb->data > skb_headlen(skb))
-		goto bad;
-
-	off += 2;
-	len -= 2;
-
-	while (len > 0) {
-		int optlen = nh[off + 1] + 2;
-
-		switch (nh[off]) {
-		case IPV6_TLV_PAD1:
-			optlen = 1;
-			break;
-
-		case IPV6_TLV_PADN:
-			break;
-
-		case IPV6_TLV_JUMBO:
-			if (nh[off + 1] != 4 || (off & 3) != 2)
-				goto bad;
-			pkt_len = ntohl(*(__be32 *)(nh + off + 2));
-			if (pkt_len <= IPV6_MAXPLEN ||
-			    ipv6_hdr(skb)->payload_len)
-				goto bad;
-			if (pkt_len > skb->len - sizeof(struct ipv6hdr))
-				goto bad;
-			if (pskb_trim_rcsum(skb,
-					    pkt_len + sizeof(struct ipv6hdr)))
-				goto bad;
-			nh = skb_network_header(skb);
-			break;
-		default:
-			if (optlen > len)
-				goto bad;
-			break;
-		}
-		off += optlen;
-		len -= optlen;
-	}
-	if (len == 0)
-		return 0;
-bad:
-	return -1;
-}
-
 int br_validate_ipv6(struct net *net, struct sk_buff *skb)
 {
 	const struct ipv6hdr *hdr;
@@ -115,22 +59,19 @@ int br_validate_ipv6(struct net *net, struct sk_buff *skb)
 		goto inhdr_error;
 
 	pkt_len = ntohs(hdr->payload_len);
+	if (hdr->nexthdr == NEXTHDR_HOP && nf_ip6_check_hbh_len(skb, &pkt_len))
+		goto drop;
 
-	if (pkt_len || hdr->nexthdr != NEXTHDR_HOP) {
-		if (pkt_len + ip6h_len > skb->len) {
-			__IP6_INC_STATS(net, idev,
-					IPSTATS_MIB_INTRUNCATEDPKTS);
-			goto drop;
-		}
-		if (pskb_trim_rcsum(skb, pkt_len + ip6h_len)) {
-			__IP6_INC_STATS(net, idev,
-					IPSTATS_MIB_INDISCARDS);
-			goto drop;
-		}
-		hdr = ipv6_hdr(skb);
+	if (pkt_len + ip6h_len > skb->len) {
+		__IP6_INC_STATS(net, idev,
+				IPSTATS_MIB_INTRUNCATEDPKTS);
+		goto drop;
 	}
-	if (hdr->nexthdr == NEXTHDR_HOP && br_nf_check_hbh_len(skb))
+	if (pskb_trim_rcsum(skb, pkt_len + ip6h_len)) {
+		__IP6_INC_STATS(net, idev,
+				IPSTATS_MIB_INDISCARDS);
 		goto drop;
+	}
 
 	memset(IP6CB(skb), 0, sizeof(struct inet6_skb_parm));
 	/* No IP options in IPv6 header; however it should be

net/bridge/netfilter/nft_meta_bridge.c

@@ -8,6 +8,9 @@
 #include <net/netfilter/nf_tables.h>
 #include <net/netfilter/nft_meta.h>
 #include <linux/if_bridge.h>
+#include <uapi/linux/netfilter_bridge.h> /* NF_BR_PRE_ROUTING */
+
+#include "../br_private.h"
 
 static const struct net_device *
 nft_meta_get_bridge(const struct net_device *dev)
@@ -102,6 +105,50 @@ static const struct nft_expr_ops nft_meta_bridge_get_ops = {
 	.reduce		= nft_meta_get_reduce,
 };
 
+static void nft_meta_bridge_set_eval(const struct nft_expr *expr,
+				     struct nft_regs *regs,
+				     const struct nft_pktinfo *pkt)
+{
+	const struct nft_meta *meta = nft_expr_priv(expr);
+	u32 *sreg = &regs->data[meta->sreg];
+	struct sk_buff *skb = pkt->skb;
+	u8 value8;
+
+	switch (meta->key) {
+	case NFT_META_BRI_BROUTE:
+		value8 = nft_reg_load8(sreg);
+		BR_INPUT_SKB_CB(skb)->br_netfilter_broute = !!value8;
+		break;
+	default:
+		nft_meta_set_eval(expr, regs, pkt);
+	}
+}
+
+static int nft_meta_bridge_set_init(const struct nft_ctx *ctx,
+				    const struct nft_expr *expr,
+				    const struct nlattr * const tb[])
+{
+	struct nft_meta *priv = nft_expr_priv(expr);
+	unsigned int len;
+	int err;
+
+	priv->key = ntohl(nla_get_be32(tb[NFTA_META_KEY]));
+	switch (priv->key) {
+	case NFT_META_BRI_BROUTE:
+		len = sizeof(u8);
+		break;
+	default:
+		return nft_meta_set_init(ctx, expr, tb);
+	}
+
+	priv->len = len;
+	err = nft_parse_register_load(tb[NFTA_META_SREG], &priv->sreg, len);
+	if (err < 0)
+		return err;
+
+	return 0;
+}
+
 static bool nft_meta_bridge_set_reduce(struct nft_regs_track *track,
 				       const struct nft_expr *expr)
 {
@@ -120,15 +167,33 @@ static bool nft_meta_bridge_set_reduce(struct nft_regs_track *track,
 	return false;
 }
 
+static int nft_meta_bridge_set_validate(const struct nft_ctx *ctx,
+					const struct nft_expr *expr,
+					const struct nft_data **data)
+{
+	struct nft_meta *priv = nft_expr_priv(expr);
+	unsigned int hooks;
+
+	switch (priv->key) {
+	case NFT_META_BRI_BROUTE:
+		hooks = 1 << NF_BR_PRE_ROUTING;
+		break;
+	default:
+		return nft_meta_set_validate(ctx, expr, data);
+	}
+
+	return nft_chain_validate_hooks(ctx->chain, hooks);
+}
+
 static const struct nft_expr_ops nft_meta_bridge_set_ops = {
 	.type		= &nft_meta_bridge_type,
 	.size		= NFT_EXPR_SIZE(sizeof(struct nft_meta)),
-	.eval		= nft_meta_set_eval,
-	.init		= nft_meta_set_init,
+	.eval		= nft_meta_bridge_set_eval,
+	.init		= nft_meta_bridge_set_init,
 	.destroy	= nft_meta_set_destroy,
 	.dump		= nft_meta_set_dump,
 	.reduce		= nft_meta_bridge_set_reduce,
-	.validate	= nft_meta_set_validate,
+	.validate	= nft_meta_bridge_set_validate,
 };
 
 static const struct nft_expr_ops *

net/netfilter/nf_conntrack_core.c

@@ -1294,7 +1294,7 @@ __nf_conntrack_confirm(struct sk_buff *skb)
 }
 EXPORT_SYMBOL_GPL(__nf_conntrack_confirm);
 
-/* Returns true if a connection correspondings to the tuple (required
+/* Returns true if a connection corresponds to the tuple (required
    for NAT). */
 int
 nf_conntrack_tuple_taken(const struct nf_conntrack_tuple *tuple,

net/netfilter/nf_conntrack_ovs.c

@@ -6,6 +6,7 @@
 #include <net/netfilter/ipv6/nf_defrag_ipv6.h>
 #include <net/ipv6_frag.h>
 #include <net/ip.h>
+#include <linux/netfilter_ipv6.h>
 
 /* 'skb' should already be pulled to nh_ofs. */
 int nf_ct_helper(struct sk_buff *skb, struct nf_conn *ct,
@@ -120,8 +121,14 @@ int nf_ct_skb_network_trim(struct sk_buff *skb, int family)
 		len = skb_ip_totlen(skb);
 		break;
 	case NFPROTO_IPV6:
-		len = sizeof(struct ipv6hdr)
-			+ ntohs(ipv6_hdr(skb)->payload_len);
+		len = ntohs(ipv6_hdr(skb)->payload_len);
+		if (ipv6_hdr(skb)->nexthdr == NEXTHDR_HOP) {
+			int err = nf_ip6_check_hbh_len(skb, &len);
+
+			if (err)
+				return err;
+		}
+		len += sizeof(struct ipv6hdr);
 		break;
 	default:
 		len = skb->len;

net/netfilter/nf_nat_core.c

@@ -549,8 +549,8 @@ get_unique_tuple(struct nf_conntrack_tuple *tuple,
 		if (range->flags & NF_NAT_RANGE_PROTO_SPECIFIED) {
 			if (!(range->flags & NF_NAT_RANGE_PROTO_OFFSET) &&
 			    l4proto_in_range(tuple, maniptype,
-			          &range->min_proto,
-			          &range->max_proto) &&
+					     &range->min_proto,
+					     &range->max_proto) &&
 			    (range->min_proto.all == range->max_proto.all ||
 			     !nf_nat_used_tuple(tuple, ct)))
 				return;

net/netfilter/utils.c

@@ -215,3 +215,55 @@ int nf_reroute(struct sk_buff *skb, struct nf_queue_entry *entry)
 	}
 	return ret;
 }
+
+/* Only get and check the lengths, not do any hop-by-hop stuff. */
+int nf_ip6_check_hbh_len(struct sk_buff *skb, u32 *plen)
+{
+	int len, off = sizeof(struct ipv6hdr);
+	unsigned char *nh;
+
+	if (!pskb_may_pull(skb, off + 8))
+		return -ENOMEM;
+	nh = (unsigned char *)(ipv6_hdr(skb) + 1);
+	len = (nh[1] + 1) << 3;
+
+	if (!pskb_may_pull(skb, off + len))
+		return -ENOMEM;
+	nh = skb_network_header(skb);
+
+	off += 2;
+	len -= 2;
+	while (len > 0) {
+		int optlen;
+
+		if (nh[off] == IPV6_TLV_PAD1) {
+			off++;
+			len--;
+			continue;
+		}
+		if (len < 2)
+			return -EBADMSG;
+		optlen = nh[off + 1] + 2;
+		if (optlen > len)
+			return -EBADMSG;
+
+		if (nh[off] == IPV6_TLV_JUMBO) {
+			u32 pkt_len;
+
+			if (nh[off + 1] != 4 || (off & 3) != 2)
+				return -EBADMSG;
+			pkt_len = ntohl(*(__be32 *)(nh + off + 2));
+			if (pkt_len <= IPV6_MAXPLEN ||
+			    ipv6_hdr(skb)->payload_len)
+				return -EBADMSG;
+			if (pkt_len > skb->len - sizeof(struct ipv6hdr))
+				return -EBADMSG;
+			*plen = pkt_len;
+		}
+		off += optlen;
+		len -= optlen;
+	}
+
+	return len ? -EBADMSG : 0;
+}
+EXPORT_SYMBOL_GPL(nf_ip6_check_hbh_len);

tools/testing/selftests/net/Makefile

@@ -48,6 +48,7 @@ TEST_PROGS += l2_tos_ttl_inherit.sh
 TEST_PROGS += bind_bhash.sh
 TEST_PROGS += ip_local_port_range.sh
 TEST_PROGS += rps_default_mask.sh
+TEST_PROGS += big_tcp.sh
 TEST_PROGS_EXTENDED := in_netns.sh setup_loopback.sh setup_veth.sh
 TEST_PROGS_EXTENDED += toeplitz_client.sh toeplitz.sh
 TEST_GEN_FILES =  socket nettest

tools/testing/selftests/net/big_tcp.sh

+#!/bin/bash
+# SPDX-License-Identifier: GPL-2.0
+#
+# Testing For IPv4 and IPv6 BIG TCP.
+# TOPO: CLIENT_NS (link0)<--->(link1) ROUTER_NS (link2)<--->(link3) SERVER_NS
+
+CLIENT_NS=$(mktemp -u client-XXXXXXXX)
+CLIENT_IP4="198.51.100.1"
+CLIENT_IP6="2001:db8:1::1"
+
+SERVER_NS=$(mktemp -u server-XXXXXXXX)
+SERVER_IP4="203.0.113.1"
+SERVER_IP6="2001:db8:2::1"
+
+ROUTER_NS=$(mktemp -u router-XXXXXXXX)
+SERVER_GW4="203.0.113.2"
+CLIENT_GW4="198.51.100.2"
+SERVER_GW6="2001:db8:2::2"
+CLIENT_GW6="2001:db8:1::2"
+
+MAX_SIZE=128000
+CHK_SIZE=65535
+
+# Kselftest framework requirement - SKIP code is 4.
+ksft_skip=4
+
+setup() {
+	ip netns add $CLIENT_NS
+	ip netns add $SERVER_NS
+	ip netns add $ROUTER_NS
+	ip -net $ROUTER_NS link add link1 type veth peer name link0 netns $CLIENT_NS
+	ip -net $ROUTER_NS link add link2 type veth peer name link3 netns $SERVER_NS
+
+	ip -net $CLIENT_NS link set link0 up
+	ip -net $CLIENT_NS link set link0 mtu 1442
+	ip -net $CLIENT_NS addr add $CLIENT_IP4/24 dev link0
+	ip -net $CLIENT_NS addr add $CLIENT_IP6/64 dev link0 nodad
+	ip -net $CLIENT_NS route add $SERVER_IP4 dev link0 via $CLIENT_GW4
+	ip -net $CLIENT_NS route add $SERVER_IP6 dev link0 via $CLIENT_GW6
+	ip -net $CLIENT_NS link set dev link0 \
+		gro_ipv4_max_size $MAX_SIZE gso_ipv4_max_size $MAX_SIZE
+	ip -net $CLIENT_NS link set dev link0 \
+		gro_max_size $MAX_SIZE gso_max_size $MAX_SIZE
+	ip net exec $CLIENT_NS sysctl -wq net.ipv4.tcp_window_scaling=10
+
+	ip -net $ROUTER_NS link set link1 up
+	ip -net $ROUTER_NS link set link2 up
+	ip -net $ROUTER_NS addr add $CLIENT_GW4/24 dev link1
+	ip -net $ROUTER_NS addr add $CLIENT_GW6/64 dev link1 nodad
+	ip -net $ROUTER_NS addr add $SERVER_GW4/24 dev link2
+	ip -net $ROUTER_NS addr add $SERVER_GW6/64 dev link2 nodad
+	ip -net $ROUTER_NS link set dev link1 \
+		gro_ipv4_max_size $MAX_SIZE gso_ipv4_max_size $MAX_SIZE
+	ip -net $ROUTER_NS link set dev link2 \
+		gro_ipv4_max_size $MAX_SIZE gso_ipv4_max_size $MAX_SIZE
+	ip -net $ROUTER_NS link set dev link1 \
+		gro_max_size $MAX_SIZE gso_max_size $MAX_SIZE
+	ip -net $ROUTER_NS link set dev link2 \
+		gro_max_size $MAX_SIZE gso_max_size $MAX_SIZE
+	# test for nf_ct_skb_network_trim in nf_conntrack_ovs used by TC ct action.
+	ip net exec $ROUTER_NS tc qdisc add dev link1 ingress
+	ip net exec $ROUTER_NS tc filter add dev link1 ingress \
+		proto ip flower ip_proto tcp action ct
+	ip net exec $ROUTER_NS tc filter add dev link1 ingress \
+		proto ipv6 flower ip_proto tcp action ct
+	ip net exec $ROUTER_NS sysctl -wq net.ipv4.ip_forward=1
+	ip net exec $ROUTER_NS sysctl -wq net.ipv6.conf.all.forwarding=1
+
+	ip -net $SERVER_NS link set link3 up
+	ip -net $SERVER_NS addr add $SERVER_IP4/24 dev link3
+	ip -net $SERVER_NS addr add $SERVER_IP6/64 dev link3 nodad
+	ip -net $SERVER_NS route add $CLIENT_IP4 dev link3 via $SERVER_GW4
+	ip -net $SERVER_NS route add $CLIENT_IP6 dev link3 via $SERVER_GW6
+	ip -net $SERVER_NS link set dev link3 \
+		gro_ipv4_max_size $MAX_SIZE gso_ipv4_max_size $MAX_SIZE
+	ip -net $SERVER_NS link set dev link3 \
+		gro_max_size $MAX_SIZE gso_max_size $MAX_SIZE
+	ip net exec $SERVER_NS sysctl -wq net.ipv4.tcp_window_scaling=10
+	ip net exec $SERVER_NS netserver 2>&1 >/dev/null
+}
+
+cleanup() {
+	ip net exec $SERVER_NS pkill netserver
+	ip -net $ROUTER_NS link del link1
+	ip -net $ROUTER_NS link del link2
+	ip netns del "$CLIENT_NS"
+	ip netns del "$SERVER_NS"
+	ip netns del "$ROUTER_NS"
+}
+
+start_counter() {
+	local ipt="iptables"
+	local iface=$1
+	local netns=$2
+
+	[ "$NF" = "6" ] && ipt="ip6tables"
+	ip net exec $netns $ipt -t raw -A PREROUTING -i $iface \
+		-m length ! --length 0:$CHK_SIZE -j ACCEPT
+}
+
+check_counter() {
+	local ipt="iptables"
+	local iface=$1
+	local netns=$2
+
+	[ "$NF" = "6" ] && ipt="ip6tables"
+	test `ip net exec $netns $ipt -t raw -L -v |grep $iface | awk '{print $1}'` != "0"
+}
+
+stop_counter() {
+	local ipt="iptables"
+	local iface=$1
+	local netns=$2
+
+	[ "$NF" = "6" ] && ipt="ip6tables"
+	ip net exec $netns $ipt -t raw -D PREROUTING -i $iface \
+		-m length ! --length 0:$CHK_SIZE -j ACCEPT
+}
+
+do_netperf() {
+	local serip=$SERVER_IP4
+	local netns=$1
+
+	[ "$NF" = "6" ] && serip=$SERVER_IP6
+	ip net exec $netns netperf -$NF -t TCP_STREAM -H $serip 2>&1 >/dev/null
+}
+
+do_test() {
+	local cli_tso=$1
+	local gw_gro=$2
+	local gw_tso=$3
+	local ser_gro=$4
+	local ret="PASS"
+
+	ip net exec $CLIENT_NS ethtool -K link0 tso $cli_tso
+	ip net exec $ROUTER_NS ethtool -K link1 gro $gw_gro
+	ip net exec $ROUTER_NS ethtool -K link2 tso $gw_tso
+	ip net exec $SERVER_NS ethtool -K link3 gro $ser_gro
+
+	start_counter link1 $ROUTER_NS
+	start_counter link3 $SERVER_NS
+	do_netperf $CLIENT_NS
+
+	if check_counter link1 $ROUTER_NS; then
+		check_counter link3 $SERVER_NS || ret="FAIL_on_link3"
+	else
+		ret="FAIL_on_link1"
+	fi
+
+	stop_counter link1 $ROUTER_NS
+	stop_counter link3 $SERVER_NS
+	printf "%-9s %-8s %-8s %-8s: [%s]\n" \
+		$cli_tso $gw_gro $gw_tso $ser_gro $ret
+	test $ret = "PASS"
+}
+
+testup() {
+	echo "CLI GSO | GW GRO | GW GSO | SER GRO" && \
+	do_test "on"  "on"  "on"  "on"  && \
+	do_test "on"  "off" "on"  "off" && \
+	do_test "off" "on"  "on"  "on"  && \
+	do_test "on"  "on"  "off" "on"  && \
+	do_test "off" "on"  "off" "on"
+}
+
+if ! netperf -V &> /dev/null; then
+	echo "SKIP: Could not run test without netperf tool"
+	exit $ksft_skip
+fi
+
+if ! ip link help 2>&1 | grep gso_ipv4_max_size &> /dev/null; then
+	echo "SKIP: Could not run test without gso/gro_ipv4_max_size supported in ip-link"
+	exit $ksft_skip
+fi
+
+trap cleanup EXIT
+setup && echo "Testing for BIG TCP:" && \
+NF=4 testup && echo "***v4 Tests Done***" && \
+NF=6 testup && echo "***v6 Tests Done***"
+exit $?