[PATCH 0/6] Netfilter/IPVS updates for net-next

[PATCH 0/6] Netfilter/IPVS updates for net-next

[Pablo Neira Ayuso]

Hi,

The following patchset contains Netfilter/IPVS updates for net-next:

1) Inspect the reply packets coming from DR/TUN and refresh connection
   state and timeout, from longguang yue and Julian Anastasov.

2) Series to add support for the inet ingress chain type in nf_tables.

Please, pull these changes from:

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

Thank you.

----------------------------------------------------------------

The following changes since commit bc081a693a56061f68f736c5d596134ee3c87689:

  Merge branch 'Offload-tc-vlan-mangle-to-mscc_ocelot-switch' (2020-10-11 11:19:25 -0700)

are available in the Git repository at:

  git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf-next.git HEAD

for you to fetch changes up to 793d5d61242695142460ce74f124281e0681fbc7:

  netfilter: flowtable: reduce calls to pskb_may_pull() (2020-10-12 01:58:10 +0200)

----------------------------------------------------------------
Pablo Neira Ayuso (5):
      netfilter: add nf_static_key_{inc,dec}
      netfilter: add nf_ingress_hook() helper function
      netfilter: add inet ingress support
      netfilter: nf_tables: add inet ingress support
      netfilter: flowtable: reduce calls to pskb_may_pull()

longguang.yue (1):
      ipvs: inspect reply packets from DR/TUN real servers

 include/net/netfilter/nf_tables.h      |   6 ++
 include/net/netfilter/nf_tables_ipv4.h |  33 +++++++++
 include/net/netfilter/nf_tables_ipv6.h |  46 ++++++++++++
 include/uapi/linux/netfilter.h         |   1 +
 net/netfilter/core.c                   | 129 ++++++++++++++++++++++++++-------
 net/netfilter/ipvs/ip_vs_conn.c        |  18 ++++-
 net/netfilter/ipvs/ip_vs_core.c        |  19 ++---
 net/netfilter/nf_flow_table_core.c     |  12 +--
 net/netfilter/nf_flow_table_ip.c       |  45 +++++++-----
 net/netfilter/nf_tables_api.c          |  14 ++--
 net/netfilter/nft_chain_filter.c       |  35 ++++++++-
 11 files changed, 282 insertions(+), 76 deletions(-)

[PATCH 1/6] ipvs: inspect reply packets from DR/TUN real servers

[Pablo Neira Ayuso]

From: "longguang.yue" <bigclouds@163.com>

Just like for MASQ, inspect the reply packets coming from DR/TUN
real servers and alter the connection's state and timeout
according to the protocol.

It's ipvs's duty to do traffic statistic if packets get hit,
no matter what mode it is.

Signed-off-by: longguang.yue <bigclouds@163.com>
Signed-off-by: Julian Anastasov <ja@ssi.bg>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
 net/netfilter/ipvs/ip_vs_conn.c | 18 +++++++++++++++---
 net/netfilter/ipvs/ip_vs_core.c | 19 +++++++------------
 2 files changed, 22 insertions(+), 15 deletions(-)

diff --git a/net/netfilter/ipvs/ip_vs_conn.c b/net/netfilter/ipvs/ip_vs_conn.c
index a90b8eac16ac..c100c6b112c8 100644
--- a/net/netfilter/ipvs/ip_vs_conn.c
+++ b/net/netfilter/ipvs/ip_vs_conn.c
@@ -402,6 +402,8 @@ struct ip_vs_conn *ip_vs_conn_out_get(const struct ip_vs_conn_param *p)
 {
 	unsigned int hash;
 	struct ip_vs_conn *cp, *ret=NULL;
+	const union nf_inet_addr *saddr;
+	__be16 sport;
 
 	/*
 	 *	Check for "full" addressed entries
@@ -411,10 +413,20 @@ struct ip_vs_conn *ip_vs_conn_out_get(const struct ip_vs_conn_param *p)
 	rcu_read_lock();
 
 	hlist_for_each_entry_rcu(cp, &ip_vs_conn_tab[hash], c_list) {
-		if (p->vport == cp->cport && p->cport == cp->dport &&
-		    cp->af == p->af &&
+		if (p->vport != cp->cport)
+			continue;
+
+		if (IP_VS_FWD_METHOD(cp) != IP_VS_CONN_F_MASQ) {
+			sport = cp->vport;
+			saddr = &cp->vaddr;
+		} else {
+			sport = cp->dport;
+			saddr = &cp->daddr;
+		}
+
+		if (p->cport == sport && cp->af == p->af &&
 		    ip_vs_addr_equal(p->af, p->vaddr, &cp->caddr) &&
-		    ip_vs_addr_equal(p->af, p->caddr, &cp->daddr) &&
+		    ip_vs_addr_equal(p->af, p->caddr, saddr) &&
 		    p->protocol == cp->protocol &&
 		    cp->ipvs == p->ipvs) {
 			if (!__ip_vs_conn_get(cp))
diff --git a/net/netfilter/ipvs/ip_vs_core.c b/net/netfilter/ipvs/ip_vs_core.c
index e3668a6e54e4..cc3c275934f4 100644
--- a/net/netfilter/ipvs/ip_vs_core.c
+++ b/net/netfilter/ipvs/ip_vs_core.c
@@ -875,7 +875,7 @@ static int handle_response_icmp(int af, struct sk_buff *skb,
 	unsigned int verdict = NF_DROP;
 
 	if (IP_VS_FWD_METHOD(cp) != IP_VS_CONN_F_MASQ)
-		goto ignore_cp;
+		goto after_nat;
 
 	/* Ensure the checksum is correct */
 	if (!skb_csum_unnecessary(skb) && ip_vs_checksum_complete(skb, ihl)) {
@@ -901,6 +901,7 @@ static int handle_response_icmp(int af, struct sk_buff *skb,
 	if (ip_vs_route_me_harder(cp->ipvs, af, skb, hooknum))
 		goto out;
 
+after_nat:
 	/* do the statistics and put it back */
 	ip_vs_out_stats(cp, skb);
 
@@ -909,8 +910,6 @@ static int handle_response_icmp(int af, struct sk_buff *skb,
 		ip_vs_notrack(skb);
 	else
 		ip_vs_update_conntrack(skb, cp, 0);
-
-ignore_cp:
 	verdict = NF_ACCEPT;
 
 out:
@@ -1276,6 +1275,9 @@ handle_response(int af, struct sk_buff *skb, struct ip_vs_proto_data *pd,
 {
 	struct ip_vs_protocol *pp = pd->pp;
 
+	if (IP_VS_FWD_METHOD(cp) != IP_VS_CONN_F_MASQ)
+		goto after_nat;
+
 	IP_VS_DBG_PKT(11, af, pp, skb, iph->off, "Outgoing packet");
 
 	if (skb_ensure_writable(skb, iph->len))
@@ -1316,6 +1318,7 @@ handle_response(int af, struct sk_buff *skb, struct ip_vs_proto_data *pd,
 
 	IP_VS_DBG_PKT(10, af, pp, skb, iph->off, "After SNAT");
 
+after_nat:
 	ip_vs_out_stats(cp, skb);
 	ip_vs_set_state(cp, IP_VS_DIR_OUTPUT, skb, pd);
 	skb->ipvs_property = 1;
@@ -1412,11 +1415,8 @@ ip_vs_out(struct netns_ipvs *ipvs, unsigned int hooknum, struct sk_buff *skb, in
 	cp = INDIRECT_CALL_1(pp->conn_out_get, ip_vs_conn_out_get_proto,
 			     ipvs, af, skb, &iph);
 
-	if (likely(cp)) {
-		if (IP_VS_FWD_METHOD(cp) != IP_VS_CONN_F_MASQ)
-			goto ignore_cp;
+	if (likely(cp))
 		return handle_response(af, skb, pd, cp, &iph, hooknum);
-	}
 
 	/* Check for real-server-started requests */
 	if (atomic_read(&ipvs->conn_out_counter)) {
@@ -1475,14 +1475,9 @@ ip_vs_out(struct netns_ipvs *ipvs, unsigned int hooknum, struct sk_buff *skb, in
 		}
 	}
 
-out:
 	IP_VS_DBG_PKT(12, af, pp, skb, iph.off,
 		      "ip_vs_out: packet continues traversal as normal");
 	return NF_ACCEPT;
-
-ignore_cp:
-	__ip_vs_conn_put(cp);
-	goto out;
 }
 
 /*
-- 
2.20.1

[PATCH 2/6] netfilter: add nf_static_key_{inc,dec}

[Pablo Neira Ayuso]

Add helper functions increment and decrement the hook static keys.

Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
 net/netfilter/core.c | 23 +++++++++++++++++------
 1 file changed, 17 insertions(+), 6 deletions(-)

diff --git a/net/netfilter/core.c b/net/netfilter/core.c
index 3ac7c8c1548d..b9ec8ecf7e30 100644
--- a/net/netfilter/core.c
+++ b/net/netfilter/core.c
@@ -311,6 +311,20 @@ nf_hook_entry_head(struct net *net, int pf, unsigned int hooknum,
 	return NULL;
 }
 
+static void nf_static_key_inc(const struct nf_hook_ops *reg, int pf)
+{
+#ifdef CONFIG_JUMP_LABEL
+       static_key_slow_inc(&nf_hooks_needed[pf][reg->hooknum]);
+#endif
+}
+
+static void nf_static_key_dec(const struct nf_hook_ops *reg, int pf)
+{
+#ifdef CONFIG_JUMP_LABEL
+       static_key_slow_dec(&nf_hooks_needed[pf][reg->hooknum]);
+#endif
+}
+
 static int __nf_register_net_hook(struct net *net, int pf,
 				  const struct nf_hook_ops *reg)
 {
@@ -348,9 +362,8 @@ static int __nf_register_net_hook(struct net *net, int pf,
 	if (pf == NFPROTO_NETDEV && reg->hooknum == NF_NETDEV_INGRESS)
 		net_inc_ingress_queue();
 #endif
-#ifdef CONFIG_JUMP_LABEL
-	static_key_slow_inc(&nf_hooks_needed[pf][reg->hooknum]);
-#endif
+	nf_static_key_inc(reg, pf);
+
 	BUG_ON(p == new_hooks);
 	nf_hook_entries_free(p);
 	return 0;
@@ -406,9 +419,7 @@ static void __nf_unregister_net_hook(struct net *net, int pf,
 		if (pf == NFPROTO_NETDEV && reg->hooknum == NF_NETDEV_INGRESS)
 			net_dec_ingress_queue();
 #endif
-#ifdef CONFIG_JUMP_LABEL
-		static_key_slow_dec(&nf_hooks_needed[pf][reg->hooknum]);
-#endif
+		nf_static_key_dec(reg, pf);
 	} else {
 		WARN_ONCE(1, "hook not found, pf %d num %d", pf, reg->hooknum);
 	}
-- 
2.20.1

[PATCH 3/6] netfilter: add nf_ingress_hook() helper function

[Pablo Neira Ayuso]

Add helper function to check if this is an ingress hook.

Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
 net/netfilter/core.c | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)

diff --git a/net/netfilter/core.c b/net/netfilter/core.c
index b9ec8ecf7e30..c82f779a587e 100644
--- a/net/netfilter/core.c
+++ b/net/netfilter/core.c
@@ -311,6 +311,11 @@ nf_hook_entry_head(struct net *net, int pf, unsigned int hooknum,
 	return NULL;
 }
 
+static inline bool nf_ingress_hook(const struct nf_hook_ops *reg, int pf)
+{
+	return pf == NFPROTO_NETDEV && reg->hooknum == NF_NETDEV_INGRESS;
+}
+
 static void nf_static_key_inc(const struct nf_hook_ops *reg, int pf)
 {
 #ifdef CONFIG_JUMP_LABEL
@@ -359,7 +364,7 @@ static int __nf_register_net_hook(struct net *net, int pf,
 
 	hooks_validate(new_hooks);
 #ifdef CONFIG_NETFILTER_INGRESS
-	if (pf == NFPROTO_NETDEV && reg->hooknum == NF_NETDEV_INGRESS)
+	if (nf_ingress_hook(reg, pf))
 		net_inc_ingress_queue();
 #endif
 	nf_static_key_inc(reg, pf);
@@ -416,7 +421,7 @@ static void __nf_unregister_net_hook(struct net *net, int pf,
 
 	if (nf_remove_net_hook(p, reg)) {
 #ifdef CONFIG_NETFILTER_INGRESS
-		if (pf == NFPROTO_NETDEV && reg->hooknum == NF_NETDEV_INGRESS)
+		if (nf_ingress_hook(reg, pf))
 			net_dec_ingress_queue();
 #endif
 		nf_static_key_dec(reg, pf);
-- 
2.20.1

[PATCH 4/6] netfilter: add inet ingress support

[Pablo Neira Ayuso]

This patch adds the NF_INET_INGRESS pseudohook for the NFPROTO_INET
family. This is a mapping this new hook to the existing NFPROTO_NETDEV
and NF_NETDEV_INGRESS hook. The hook does not guarantee that packets are
inet only, users must filter out non-ip traffic explicitly.

This infrastructure makes it easier to support this new hook in nf_tables.

Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
 include/uapi/linux/netfilter.h |   1 +
 net/netfilter/core.c           | 103 ++++++++++++++++++++++++++-------
 2 files changed, 83 insertions(+), 21 deletions(-)

diff --git a/include/uapi/linux/netfilter.h b/include/uapi/linux/netfilter.h
index ca9e63d6e0e4..6a6179af0d7c 100644
--- a/include/uapi/linux/netfilter.h
+++ b/include/uapi/linux/netfilter.h
@@ -45,6 +45,7 @@ enum nf_inet_hooks {
 	NF_INET_FORWARD,
 	NF_INET_LOCAL_OUT,
 	NF_INET_POST_ROUTING,
+	NF_INET_INGRESS,
 	NF_INET_NUMHOOKS
 };
 
diff --git a/net/netfilter/core.c b/net/netfilter/core.c
index c82f779a587e..63d032191e62 100644
--- a/net/netfilter/core.c
+++ b/net/netfilter/core.c
@@ -281,6 +281,16 @@ nf_hook_entry_head(struct net *net, int pf, unsigned int hooknum,
 		if (WARN_ON_ONCE(ARRAY_SIZE(net->nf.hooks_bridge) <= hooknum))
 			return NULL;
 		return net->nf.hooks_bridge + hooknum;
+#endif
+#ifdef CONFIG_NETFILTER_INGRESS
+	case NFPROTO_INET:
+		if (WARN_ON_ONCE(hooknum != NF_INET_INGRESS))
+			return NULL;
+		if (!dev || dev_net(dev) != net) {
+			WARN_ON_ONCE(1);
+			return NULL;
+		}
+		return &dev->nf_hooks_ingress;
 #endif
 	case NFPROTO_IPV4:
 		if (WARN_ON_ONCE(ARRAY_SIZE(net->nf.hooks_ipv4) <= hooknum))
@@ -311,22 +321,56 @@ nf_hook_entry_head(struct net *net, int pf, unsigned int hooknum,
 	return NULL;
 }
 
+static int nf_ingress_check(struct net *net, const struct nf_hook_ops *reg,
+			    int hooknum)
+{
+#ifndef CONFIG_NETFILTER_INGRESS
+	if (reg->hooknum == hooknum)
+		return -EOPNOTSUPP;
+#endif
+	if (reg->hooknum != hooknum ||
+	    !reg->dev || dev_net(reg->dev) != net)
+		return -EINVAL;
+
+	return 0;
+}
+
 static inline bool nf_ingress_hook(const struct nf_hook_ops *reg, int pf)
 {
-	return pf == NFPROTO_NETDEV && reg->hooknum == NF_NETDEV_INGRESS;
+	if ((pf == NFPROTO_NETDEV && reg->hooknum == NF_NETDEV_INGRESS) ||
+	    (pf == NFPROTO_INET && reg->hooknum == NF_INET_INGRESS))
+		return true;
+
+	return false;
 }
 
 static void nf_static_key_inc(const struct nf_hook_ops *reg, int pf)
 {
 #ifdef CONFIG_JUMP_LABEL
-       static_key_slow_inc(&nf_hooks_needed[pf][reg->hooknum]);
+	int hooknum;
+
+	if (pf == NFPROTO_INET && reg->hooknum == NF_INET_INGRESS) {
+		pf = NFPROTO_NETDEV;
+		hooknum = NF_NETDEV_INGRESS;
+	} else {
+		hooknum = reg->hooknum;
+	}
+	static_key_slow_inc(&nf_hooks_needed[pf][hooknum]);
 #endif
 }
 
 static void nf_static_key_dec(const struct nf_hook_ops *reg, int pf)
 {
 #ifdef CONFIG_JUMP_LABEL
-       static_key_slow_dec(&nf_hooks_needed[pf][reg->hooknum]);
+	int hooknum;
+
+	if (pf == NFPROTO_INET && reg->hooknum == NF_INET_INGRESS) {
+		pf = NFPROTO_NETDEV;
+		hooknum = NF_NETDEV_INGRESS;
+	} else {
+		hooknum = reg->hooknum;
+	}
+	static_key_slow_dec(&nf_hooks_needed[pf][hooknum]);
 #endif
 }
 
@@ -335,15 +379,22 @@ static int __nf_register_net_hook(struct net *net, int pf,
 {
 	struct nf_hook_entries *p, *new_hooks;
 	struct nf_hook_entries __rcu **pp;
+	int err;
 
-	if (pf == NFPROTO_NETDEV) {
-#ifndef CONFIG_NETFILTER_INGRESS
-		if (reg->hooknum == NF_NETDEV_INGRESS)
-			return -EOPNOTSUPP;
-#endif
-		if (reg->hooknum != NF_NETDEV_INGRESS ||
-		    !reg->dev || dev_net(reg->dev) != net)
-			return -EINVAL;
+	switch (pf) {
+	case NFPROTO_NETDEV:
+		err = nf_ingress_check(net, reg, NF_NETDEV_INGRESS);
+		if (err < 0)
+			return err;
+		break;
+	case NFPROTO_INET:
+		if (reg->hooknum != NF_INET_INGRESS)
+			break;
+
+		err = nf_ingress_check(net, reg, NF_INET_INGRESS);
+		if (err < 0)
+			return err;
+		break;
 	}
 
 	pp = nf_hook_entry_head(net, pf, reg->hooknum, reg->dev);
@@ -441,8 +492,12 @@ static void __nf_unregister_net_hook(struct net *net, int pf,
 void nf_unregister_net_hook(struct net *net, const struct nf_hook_ops *reg)
 {
 	if (reg->pf == NFPROTO_INET) {
-		__nf_unregister_net_hook(net, NFPROTO_IPV4, reg);
-		__nf_unregister_net_hook(net, NFPROTO_IPV6, reg);
+		if (reg->hooknum == NF_INET_INGRESS) {
+			__nf_unregister_net_hook(net, NFPROTO_INET, reg);
+		} else {
+			__nf_unregister_net_hook(net, NFPROTO_IPV4, reg);
+			__nf_unregister_net_hook(net, NFPROTO_IPV6, reg);
+		}
 	} else {
 		__nf_unregister_net_hook(net, reg->pf, reg);
 	}
@@ -467,14 +522,20 @@ int nf_register_net_hook(struct net *net, const struct nf_hook_ops *reg)
 	int err;
 
 	if (reg->pf == NFPROTO_INET) {
-		err = __nf_register_net_hook(net, NFPROTO_IPV4, reg);
-		if (err < 0)
-			return err;
-
-		err = __nf_register_net_hook(net, NFPROTO_IPV6, reg);
-		if (err < 0) {
-			__nf_unregister_net_hook(net, NFPROTO_IPV4, reg);
-			return err;
+		if (reg->hooknum == NF_INET_INGRESS) {
+			err = __nf_register_net_hook(net, NFPROTO_INET, reg);
+			if (err < 0)
+				return err;
+		} else {
+			err = __nf_register_net_hook(net, NFPROTO_IPV4, reg);
+			if (err < 0)
+				return err;
+
+			err = __nf_register_net_hook(net, NFPROTO_IPV6, reg);
+			if (err < 0) {
+				__nf_unregister_net_hook(net, NFPROTO_IPV4, reg);
+				return err;
+			}
 		}
 	} else {
 		err = __nf_register_net_hook(net, reg->pf, reg);
-- 
2.20.1

[PATCH 5/6] netfilter: nf_tables: add inet ingress support

[Pablo Neira Ayuso]

This patch adds a new ingress hook for the inet family. The inet ingress
hook emulates the IP receive path code, therefore, unclean packets are
drop before walking over the ruleset in this basechain.

This patch also introduces the nft_base_chain_netdev() helper function
to check if this hook is bound to one or more devices (through the hook
list infrastructure). This check allows to perform the same handling for
the inet ingress as it would be a netdev ingress chain from the control
plane.

Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
 include/net/netfilter/nf_tables.h      |  6 ++++
 include/net/netfilter/nf_tables_ipv4.h | 33 ++++++++++++++++++
 include/net/netfilter/nf_tables_ipv6.h | 46 ++++++++++++++++++++++++++
 net/netfilter/nf_tables_api.c          | 14 ++++----
 net/netfilter/nft_chain_filter.c       | 35 +++++++++++++++++++-
 5 files changed, 126 insertions(+), 8 deletions(-)

diff --git a/include/net/netfilter/nf_tables.h b/include/net/netfilter/nf_tables.h
index 0bd2a081ae39..3965ce18226f 100644
--- a/include/net/netfilter/nf_tables.h
+++ b/include/net/netfilter/nf_tables.h
@@ -1081,6 +1081,12 @@ struct nft_table {
 	u8				*udata;
 };
 
+static inline bool nft_base_chain_netdev(int family, u32 hooknum)
+{
+	return family == NFPROTO_NETDEV ||
+	       (family == NFPROTO_INET && hooknum == NF_INET_INGRESS);
+}
+
 void nft_register_chain_type(const struct nft_chain_type *);
 void nft_unregister_chain_type(const struct nft_chain_type *);
 
diff --git a/include/net/netfilter/nf_tables_ipv4.h b/include/net/netfilter/nf_tables_ipv4.h
index ed7b511f0a59..1f7bea39ad1b 100644
--- a/include/net/netfilter/nf_tables_ipv4.h
+++ b/include/net/netfilter/nf_tables_ipv4.h
@@ -53,4 +53,37 @@ static inline void nft_set_pktinfo_ipv4_validate(struct nft_pktinfo *pkt,
 		nft_set_pktinfo_unspec(pkt, skb);
 }
 
+static inline int nft_set_pktinfo_ipv4_ingress(struct nft_pktinfo *pkt,
+					       struct sk_buff *skb)
+{
+	struct iphdr *iph;
+	u32 len, thoff;
+
+	if (!pskb_may_pull(skb, sizeof(*iph)))
+		return -1;
+
+	iph = ip_hdr(skb);
+	if (iph->ihl < 5 || iph->version != 4)
+		goto inhdr_error;
+
+	len = ntohs(iph->tot_len);
+	thoff = iph->ihl * 4;
+	if (skb->len < len) {
+		__IP_INC_STATS(nft_net(pkt), IPSTATS_MIB_INTRUNCATEDPKTS);
+		return -1;
+	} else if (len < thoff) {
+		goto inhdr_error;
+	}
+
+	pkt->tprot_set = true;
+	pkt->tprot = iph->protocol;
+	pkt->xt.thoff = thoff;
+	pkt->xt.fragoff = ntohs(iph->frag_off) & IP_OFFSET;
+
+	return 0;
+
+inhdr_error:
+	__IP_INC_STATS(nft_net(pkt), IPSTATS_MIB_INHDRERRORS);
+	return -1;
+}
 #endif
diff --git a/include/net/netfilter/nf_tables_ipv6.h b/include/net/netfilter/nf_tables_ipv6.h
index d0f1c537b017..867de29f3f7a 100644
--- a/include/net/netfilter/nf_tables_ipv6.h
+++ b/include/net/netfilter/nf_tables_ipv6.h
@@ -70,4 +70,50 @@ static inline void nft_set_pktinfo_ipv6_validate(struct nft_pktinfo *pkt,
 		nft_set_pktinfo_unspec(pkt, skb);
 }
 
+static inline int nft_set_pktinfo_ipv6_ingress(struct nft_pktinfo *pkt,
+					       struct sk_buff *skb)
+{
+#if IS_ENABLED(CONFIG_IPV6)
+	unsigned int flags = IP6_FH_F_AUTH;
+	unsigned short frag_off;
+	unsigned int thoff = 0;
+	struct inet6_dev *idev;
+	struct ipv6hdr *ip6h;
+	int protohdr;
+	u32 pkt_len;
+
+	if (!pskb_may_pull(skb, sizeof(*ip6h)))
+		return -1;
+
+	ip6h = ipv6_hdr(skb);
+	if (ip6h->version != 6)
+		goto inhdr_error;
+
+	pkt_len = ntohs(ip6h->payload_len);
+	if (pkt_len + sizeof(*ip6h) > skb->len) {
+		idev = __in6_dev_get(nft_in(pkt));
+		__IP6_INC_STATS(nft_net(pkt), idev, IPSTATS_MIB_INTRUNCATEDPKTS);
+		return -1;
+	}
+
+	protohdr = ipv6_find_hdr(pkt->skb, &thoff, -1, &frag_off, &flags);
+	if (protohdr < 0)
+		goto inhdr_error;
+
+	pkt->tprot_set = true;
+	pkt->tprot = protohdr;
+	pkt->xt.thoff = thoff;
+	pkt->xt.fragoff = frag_off;
+
+	return 0;
+
+inhdr_error:
+	idev = __in6_dev_get(nft_in(pkt));
+	__IP6_INC_STATS(nft_net(pkt), idev, IPSTATS_MIB_INHDRERRORS);
+	return -1;
+#else
+	return -1;
+#endif
+}
+
 #endif
diff --git a/net/netfilter/nf_tables_api.c b/net/netfilter/nf_tables_api.c
index ae2c04d411b1..f22ad21d0230 100644
--- a/net/netfilter/nf_tables_api.c
+++ b/net/netfilter/nf_tables_api.c
@@ -206,7 +206,7 @@ static int nf_tables_register_hook(struct net *net,
 	if (basechain->type->ops_register)
 		return basechain->type->ops_register(net, ops);
 
-	if (table->family == NFPROTO_NETDEV)
+	if (nft_base_chain_netdev(table->family, basechain->ops.hooknum))
 		return nft_netdev_register_hooks(net, &basechain->hook_list);
 
 	return nf_register_net_hook(net, &basechain->ops);
@@ -228,7 +228,7 @@ static void nf_tables_unregister_hook(struct net *net,
 	if (basechain->type->ops_unregister)
 		return basechain->type->ops_unregister(net, ops);
 
-	if (table->family == NFPROTO_NETDEV)
+	if (nft_base_chain_netdev(table->family, basechain->ops.hooknum))
 		nft_netdev_unregister_hooks(net, &basechain->hook_list);
 	else
 		nf_unregister_net_hook(net, &basechain->ops);
@@ -1381,7 +1381,7 @@ static int nft_dump_basechain_hook(struct sk_buff *skb, int family,
 	if (nla_put_be32(skb, NFTA_HOOK_PRIORITY, htonl(ops->priority)))
 		goto nla_put_failure;
 
-	if (family == NFPROTO_NETDEV) {
+	if (nft_base_chain_netdev(family, ops->hooknum)) {
 		nest_devs = nla_nest_start_noflag(skb, NFTA_HOOK_DEVS);
 		list_for_each_entry(hook, &basechain->hook_list, list) {
 			if (!first)
@@ -1685,7 +1685,7 @@ void nf_tables_chain_destroy(struct nft_ctx *ctx)
 	if (nft_is_base_chain(chain)) {
 		struct nft_base_chain *basechain = nft_base_chain(chain);
 
-		if (ctx->family == NFPROTO_NETDEV) {
+		if (nft_base_chain_netdev(ctx->family, basechain->ops.hooknum)) {
 			list_for_each_entry_safe(hook, next,
 						 &basechain->hook_list, list) {
 				list_del_rcu(&hook->list);
@@ -1877,7 +1877,7 @@ static int nft_chain_parse_hook(struct net *net,
 	hook->type = type;
 
 	INIT_LIST_HEAD(&hook->list);
-	if (family == NFPROTO_NETDEV) {
+	if (nft_base_chain_netdev(family, hook->num)) {
 		err = nft_chain_parse_netdev(net, ha, &hook->list);
 		if (err < 0) {
 			module_put(type->owner);
@@ -1944,7 +1944,7 @@ static int nft_basechain_init(struct nft_base_chain *basechain, u8 family,
 	INIT_LIST_HEAD(&basechain->hook_list);
 	chain = &basechain->chain;
 
-	if (family == NFPROTO_NETDEV) {
+	if (nft_base_chain_netdev(family, hook->num)) {
 		list_splice_init(&hook->list, &basechain->hook_list);
 		list_for_each_entry(h, &basechain->hook_list, list)
 			nft_basechain_hook_init(&h->ops, family, hook, chain);
@@ -2168,7 +2168,7 @@ static int nf_tables_updchain(struct nft_ctx *ctx, u8 genmask, u8 policy,
 			return -EEXIST;
 		}
 
-		if (ctx->family == NFPROTO_NETDEV) {
+		if (nft_base_chain_netdev(ctx->family, hook.num)) {
 			if (!nft_hook_list_equal(&basechain->hook_list,
 						 &hook.list)) {
 				nft_chain_release_hook(&hook);
diff --git a/net/netfilter/nft_chain_filter.c b/net/netfilter/nft_chain_filter.c
index c78d01bc02e9..ff8528ad3dc6 100644
--- a/net/netfilter/nft_chain_filter.c
+++ b/net/netfilter/nft_chain_filter.c
@@ -161,16 +161,49 @@ static unsigned int nft_do_chain_inet(void *priv, struct sk_buff *skb,
 	return nft_do_chain(&pkt, priv);
 }
 
+static unsigned int nft_do_chain_inet_ingress(void *priv, struct sk_buff *skb,
+					      const struct nf_hook_state *state)
+{
+	struct nf_hook_state ingress_state = *state;
+	struct nft_pktinfo pkt;
+
+	switch (skb->protocol) {
+	case htons(ETH_P_IP):
+		/* Original hook is NFPROTO_NETDEV and NF_NETDEV_INGRESS. */
+		ingress_state.pf = NFPROTO_IPV4;
+		ingress_state.hook = NF_INET_INGRESS;
+		nft_set_pktinfo(&pkt, skb, &ingress_state);
+
+		if (nft_set_pktinfo_ipv4_ingress(&pkt, skb) < 0)
+			return NF_DROP;
+		break;
+	case htons(ETH_P_IPV6):
+		ingress_state.pf = NFPROTO_IPV6;
+		ingress_state.hook = NF_INET_INGRESS;
+		nft_set_pktinfo(&pkt, skb, &ingress_state);
+
+		if (nft_set_pktinfo_ipv6_ingress(&pkt, skb) < 0)
+			return NF_DROP;
+		break;
+	default:
+		return NF_ACCEPT;
+	}
+
+	return nft_do_chain(&pkt, priv);
+}
+
 static const struct nft_chain_type nft_chain_filter_inet = {
 	.name		= "filter",
 	.type		= NFT_CHAIN_T_DEFAULT,
 	.family		= NFPROTO_INET,
-	.hook_mask	= (1 << NF_INET_LOCAL_IN) |
+	.hook_mask	= (1 << NF_INET_INGRESS) |
+			  (1 << NF_INET_LOCAL_IN) |
 			  (1 << NF_INET_LOCAL_OUT) |
 			  (1 << NF_INET_FORWARD) |
 			  (1 << NF_INET_PRE_ROUTING) |
 			  (1 << NF_INET_POST_ROUTING),
 	.hooks		= {
+		[NF_INET_INGRESS]	= nft_do_chain_inet_ingress,
 		[NF_INET_LOCAL_IN]	= nft_do_chain_inet,
 		[NF_INET_LOCAL_OUT]	= nft_do_chain_inet,
 		[NF_INET_FORWARD]	= nft_do_chain_inet,
-- 
2.20.1

[PATCH 6/6] netfilter: flowtable: reduce calls to pskb_may_pull()

[Pablo Neira Ayuso]

Make two unfront calls to pskb_may_pull() to linearize the network and
transport header.

Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
 net/netfilter/nf_flow_table_core.c | 12 +++-----
 net/netfilter/nf_flow_table_ip.c   | 45 +++++++++++++++++-------------
 2 files changed, 30 insertions(+), 27 deletions(-)

diff --git a/net/netfilter/nf_flow_table_core.c b/net/netfilter/nf_flow_table_core.c
index 4f7a567c536e..513f78db3cb2 100644
--- a/net/netfilter/nf_flow_table_core.c
+++ b/net/netfilter/nf_flow_table_core.c
@@ -395,8 +395,7 @@ static int nf_flow_nat_port_tcp(struct sk_buff *skb, unsigned int thoff,
 {
 	struct tcphdr *tcph;
 
-	if (!pskb_may_pull(skb, thoff + sizeof(*tcph)) ||
-	    skb_try_make_writable(skb, thoff + sizeof(*tcph)))
+	if (skb_try_make_writable(skb, thoff + sizeof(*tcph)))
 		return -1;
 
 	tcph = (void *)(skb_network_header(skb) + thoff);
@@ -410,8 +409,7 @@ static int nf_flow_nat_port_udp(struct sk_buff *skb, unsigned int thoff,
 {
 	struct udphdr *udph;
 
-	if (!pskb_may_pull(skb, thoff + sizeof(*udph)) ||
-	    skb_try_make_writable(skb, thoff + sizeof(*udph)))
+	if (skb_try_make_writable(skb, thoff + sizeof(*udph)))
 		return -1;
 
 	udph = (void *)(skb_network_header(skb) + thoff);
@@ -449,8 +447,7 @@ int nf_flow_snat_port(const struct flow_offload *flow,
 	struct flow_ports *hdr;
 	__be16 port, new_port;
 
-	if (!pskb_may_pull(skb, thoff + sizeof(*hdr)) ||
-	    skb_try_make_writable(skb, thoff + sizeof(*hdr)))
+	if (skb_try_make_writable(skb, thoff + sizeof(*hdr)))
 		return -1;
 
 	hdr = (void *)(skb_network_header(skb) + thoff);
@@ -481,8 +478,7 @@ int nf_flow_dnat_port(const struct flow_offload *flow,
 	struct flow_ports *hdr;
 	__be16 port, new_port;
 
-	if (!pskb_may_pull(skb, thoff + sizeof(*hdr)) ||
-	    skb_try_make_writable(skb, thoff + sizeof(*hdr)))
+	if (skb_try_make_writable(skb, thoff + sizeof(*hdr)))
 		return -1;
 
 	hdr = (void *)(skb_network_header(skb) + thoff);
diff --git a/net/netfilter/nf_flow_table_ip.c b/net/netfilter/nf_flow_table_ip.c
index a3bca758b849..a698dbe28ef5 100644
--- a/net/netfilter/nf_flow_table_ip.c
+++ b/net/netfilter/nf_flow_table_ip.c
@@ -25,9 +25,6 @@ static int nf_flow_state_check(struct flow_offload *flow, int proto,
 	if (proto != IPPROTO_TCP)
 		return 0;
 
-	if (!pskb_may_pull(skb, thoff + sizeof(*tcph)))
-		return -1;
-
 	tcph = (void *)(skb_network_header(skb) + thoff);
 	if (unlikely(tcph->fin || tcph->rst)) {
 		flow_offload_teardown(flow);
@@ -42,8 +39,7 @@ static int nf_flow_nat_ip_tcp(struct sk_buff *skb, unsigned int thoff,
 {
 	struct tcphdr *tcph;
 
-	if (!pskb_may_pull(skb, thoff + sizeof(*tcph)) ||
-	    skb_try_make_writable(skb, thoff + sizeof(*tcph)))
+	if (skb_try_make_writable(skb, thoff + sizeof(*tcph)))
 		return -1;
 
 	tcph = (void *)(skb_network_header(skb) + thoff);
@@ -57,8 +53,7 @@ static int nf_flow_nat_ip_udp(struct sk_buff *skb, unsigned int thoff,
 {
 	struct udphdr *udph;
 
-	if (!pskb_may_pull(skb, thoff + sizeof(*udph)) ||
-	    skb_try_make_writable(skb, thoff + sizeof(*udph)))
+	if (skb_try_make_writable(skb, thoff + sizeof(*udph)))
 		return -1;
 
 	udph = (void *)(skb_network_header(skb) + thoff);
@@ -167,8 +162,8 @@ static bool ip_has_options(unsigned int thoff)
 static int nf_flow_tuple_ip(struct sk_buff *skb, const struct net_device *dev,
 			    struct flow_offload_tuple *tuple)
 {
+	unsigned int thoff, hdrsize;
 	struct flow_ports *ports;
-	unsigned int thoff;
 	struct iphdr *iph;
 
 	if (!pskb_may_pull(skb, sizeof(*iph)))
@@ -181,15 +176,22 @@ static int nf_flow_tuple_ip(struct sk_buff *skb, const struct net_device *dev,
 	    unlikely(ip_has_options(thoff)))
 		return -1;
 
-	if (iph->protocol != IPPROTO_TCP &&
-	    iph->protocol != IPPROTO_UDP)
+	switch (iph->protocol) {
+	case IPPROTO_TCP:
+		hdrsize = sizeof(struct tcphdr);
+		break;
+	case IPPROTO_UDP:
+		hdrsize = sizeof(struct udphdr);
+		break;
+	default:
 		return -1;
+	}
 
 	if (iph->ttl <= 1)
 		return -1;
 
 	thoff = iph->ihl * 4;
-	if (!pskb_may_pull(skb, thoff + sizeof(*ports)))
+	if (!pskb_may_pull(skb, thoff + hdrsize))
 		return -1;
 
 	iph = ip_hdr(skb);
@@ -315,8 +317,7 @@ static int nf_flow_nat_ipv6_tcp(struct sk_buff *skb, unsigned int thoff,
 {
 	struct tcphdr *tcph;
 
-	if (!pskb_may_pull(skb, thoff + sizeof(*tcph)) ||
-	    skb_try_make_writable(skb, thoff + sizeof(*tcph)))
+	if (skb_try_make_writable(skb, thoff + sizeof(*tcph)))
 		return -1;
 
 	tcph = (void *)(skb_network_header(skb) + thoff);
@@ -332,8 +333,7 @@ static int nf_flow_nat_ipv6_udp(struct sk_buff *skb, unsigned int thoff,
 {
 	struct udphdr *udph;
 
-	if (!pskb_may_pull(skb, thoff + sizeof(*udph)) ||
-	    skb_try_make_writable(skb, thoff + sizeof(*udph)))
+	if (skb_try_make_writable(skb, thoff + sizeof(*udph)))
 		return -1;
 
 	udph = (void *)(skb_network_header(skb) + thoff);
@@ -439,24 +439,31 @@ static int nf_flow_nat_ipv6(const struct flow_offload *flow,
 static int nf_flow_tuple_ipv6(struct sk_buff *skb, const struct net_device *dev,
 			      struct flow_offload_tuple *tuple)
 {
+	unsigned int thoff, hdrsize;
 	struct flow_ports *ports;
 	struct ipv6hdr *ip6h;
-	unsigned int thoff;
 
 	if (!pskb_may_pull(skb, sizeof(*ip6h)))
 		return -1;
 
 	ip6h = ipv6_hdr(skb);
 
-	if (ip6h->nexthdr != IPPROTO_TCP &&
-	    ip6h->nexthdr != IPPROTO_UDP)
+	switch (ip6h->nexthdr) {
+	case IPPROTO_TCP:
+		hdrsize = sizeof(struct tcphdr);
+		break;
+	case IPPROTO_UDP:
+		hdrsize = sizeof(struct udphdr);
+		break;
+	default:
 		return -1;
+	}
 
 	if (ip6h->hop_limit <= 1)
 		return -1;
 
 	thoff = sizeof(*ip6h);
-	if (!pskb_may_pull(skb, thoff + sizeof(*ports)))
+	if (!pskb_may_pull(skb, thoff + hdrsize))
 		return -1;
 
 	ip6h = ipv6_hdr(skb);
-- 
2.20.1

Re: [PATCH 0/6] Netfilter/IPVS updates for net-next

[Jakub Kicinski]

On Mon, 12 Oct 2020 03:38:13 +0200 Pablo Neira Ayuso wrote:
> Hi,
> 
> The following patchset contains Netfilter/IPVS updates for net-next:
> 
> 1) Inspect the reply packets coming from DR/TUN and refresh connection
>    state and timeout, from longguang yue and Julian Anastasov.
> 
> 2) Series to add support for the inet ingress chain type in nf_tables.

Pulled, thanks!