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

Pablo Neira Ayuso says:

====================
Netfilter/IPVS updates for net-next

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

1) Add new run_estimation toggle to IPVS to stop the estimation_timer
   logic, from Dust Li.

2) Relax superfluous dynset check on NFT_SET_TIMEOUT.

3) Add egress hook, from Lukas Wunner.

4) Nowadays, almost all hook functions in x_table land just call the hook
   evaluation loop. Remove remaining hook wrappers from iptables and IPVS.
   From Florian Westphal.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

10 files changed, 175 insertions, 70 deletions

diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index f9cd6fea213f..3ec42495a43a 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -1861,6 +1861,7 @@ enum netdev_ml_priv_type {
  *	@xps_maps:	XXX: need comments on this one
  *	@miniq_egress:		clsact qdisc specific data for
  *				egress processing
+ *	@nf_hooks_egress:	netfilter hooks executed for egress packets
  *	@qdisc_hash:		qdisc hash table
  *	@watchdog_timeo:	Represents the timeout that is used by
  *				the watchdog (see dev_watchdog())
@@ -2160,6 +2161,9 @@ struct net_device {
 #ifdef CONFIG_NET_CLS_ACT
 	struct mini_Qdisc __rcu	*miniq_egress;
 #endif
+#ifdef CONFIG_NETFILTER_EGRESS
+	struct nf_hook_entries __rcu *nf_hooks_egress;
+#endif
 
 #ifdef CONFIG_NET_SCHED
 	DECLARE_HASHTABLE	(qdisc_hash, 4);
diff --git a/include/linux/netfilter_arp/arp_tables.h b/include/linux/netfilter_arp/arp_tables.h
index 4f9a4b3c5892..a40aaf645fa4 100644
--- a/include/linux/netfilter_arp/arp_tables.h
+++ b/include/linux/netfilter_arp/arp_tables.h
@@ -54,9 +54,8 @@ int arpt_register_table(struct net *net, const struct xt_table *table,
 			const struct nf_hook_ops *ops);
 void arpt_unregister_table(struct net *net, const char *name);
 void arpt_unregister_table_pre_exit(struct net *net, const char *name);
-extern unsigned int arpt_do_table(struct sk_buff *skb,
-				  const struct nf_hook_state *state,
-				  struct xt_table *table);
+extern unsigned int arpt_do_table(void *priv, struct sk_buff *skb,
+				  const struct nf_hook_state *state);
 
 #ifdef CONFIG_NETFILTER_XTABLES_COMPAT
 #include <net/compat.h>
diff --git a/include/linux/netfilter_bridge/ebtables.h b/include/linux/netfilter_bridge/ebtables.h
index 10a01978bc0d..a13296d6c7ce 100644
--- a/include/linux/netfilter_bridge/ebtables.h
+++ b/include/linux/netfilter_bridge/ebtables.h
@@ -112,9 +112,8 @@ extern int ebt_register_table(struct net *net,
 			      const struct nf_hook_ops *ops);
 extern void ebt_unregister_table(struct net *net, const char *tablename);
 void ebt_unregister_table_pre_exit(struct net *net, const char *tablename);
-extern unsigned int ebt_do_table(struct sk_buff *skb,
-				 const struct nf_hook_state *state,
-				 struct ebt_table *table);
+extern unsigned int ebt_do_table(void *priv, struct sk_buff *skb,
+				 const struct nf_hook_state *state);
 
 /* True if the hook mask denotes that the rule is in a base chain,
  * used in the check() functions */
diff --git a/include/linux/netfilter_ingress.h b/include/linux/netfilter_ingress.h
deleted file mode 100644
index a13774be2eb5..000000000000
--- a/include/linux/netfilter_ingress.h
+++ /dev/null
@@ -1,58 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0 */
-#ifndef _NETFILTER_INGRESS_H_
-#define _NETFILTER_INGRESS_H_
-
-#include <linux/netfilter.h>
-#include <linux/netdevice.h>
-
-#ifdef CONFIG_NETFILTER_INGRESS
-static inline bool nf_hook_ingress_active(const struct sk_buff *skb)
-{
-#ifdef CONFIG_JUMP_LABEL
-	if (!static_key_false(&nf_hooks_needed[NFPROTO_NETDEV][NF_NETDEV_INGRESS]))
-		return false;
-#endif
-	return rcu_access_pointer(skb->dev->nf_hooks_ingress);
-}
-
-/* caller must hold rcu_read_lock */
-static inline int nf_hook_ingress(struct sk_buff *skb)
-{
-	struct nf_hook_entries *e = rcu_dereference(skb->dev->nf_hooks_ingress);
-	struct nf_hook_state state;
-	int ret;
-
-	/* Must recheck the ingress hook head, in the event it became NULL
-	 * after the check in nf_hook_ingress_active evaluated to true.
-	 */
-	if (unlikely(!e))
-		return 0;
-
-	nf_hook_state_init(&state, NF_NETDEV_INGRESS,
-			   NFPROTO_NETDEV, skb->dev, NULL, NULL,
-			   dev_net(skb->dev), NULL);
-	ret = nf_hook_slow(skb, &state, e, 0);
-	if (ret == 0)
-		return -1;
-
-	return ret;
-}
-
-static inline void nf_hook_ingress_init(struct net_device *dev)
-{
-	RCU_INIT_POINTER(dev->nf_hooks_ingress, NULL);
-}
-#else /* CONFIG_NETFILTER_INGRESS */
-static inline int nf_hook_ingress_active(struct sk_buff *skb)
-{
-	return 0;
-}
-
-static inline int nf_hook_ingress(struct sk_buff *skb)
-{
-	return 0;
-}
-
-static inline void nf_hook_ingress_init(struct net_device *dev) {}
-#endif /* CONFIG_NETFILTER_INGRESS */
-#endif /* _NETFILTER_INGRESS_H_ */
diff --git a/include/linux/netfilter_ipv4/ip_tables.h b/include/linux/netfilter_ipv4/ip_tables.h
index 8d09bfe850dc..132b0e4a6d4d 100644
--- a/include/linux/netfilter_ipv4/ip_tables.h
+++ b/include/linux/netfilter_ipv4/ip_tables.h
@@ -63,9 +63,9 @@ struct ipt_error {
 }
 
 extern void *ipt_alloc_initial_table(const struct xt_table *);
-extern unsigned int ipt_do_table(struct sk_buff *skb,
-				 const struct nf_hook_state *state,
-				 struct xt_table *table);
+extern unsigned int ipt_do_table(void *priv,
+				 struct sk_buff *skb,
+				 const struct nf_hook_state *state);
 
 #ifdef CONFIG_NETFILTER_XTABLES_COMPAT
 #include <net/compat.h>
diff --git a/include/linux/netfilter_ipv6/ip6_tables.h b/include/linux/netfilter_ipv6/ip6_tables.h
index 79e73fd7d965..8b8885a73c76 100644
--- a/include/linux/netfilter_ipv6/ip6_tables.h
+++ b/include/linux/netfilter_ipv6/ip6_tables.h
@@ -29,9 +29,8 @@ int ip6t_register_table(struct net *net, const struct xt_table *table,
 			const struct nf_hook_ops *ops);
 void ip6t_unregister_table_pre_exit(struct net *net, const char *name);
 void ip6t_unregister_table_exit(struct net *net, const char *name);
-extern unsigned int ip6t_do_table(struct sk_buff *skb,
-				  const struct nf_hook_state *state,
-				  struct xt_table *table);
+extern unsigned int ip6t_do_table(void *priv, struct sk_buff *skb,
+				  const struct nf_hook_state *state);
 
 #ifdef CONFIG_NETFILTER_XTABLES_COMPAT
 #include <net/compat.h>
diff --git a/include/linux/netfilter_netdev.h b/include/linux/netfilter_netdev.h
new file mode 100644
index 000000000000..b71b57a83bb4
--- /dev/null
+++ b/include/linux/netfilter_netdev.h
@@ -0,0 +1,146 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef _NETFILTER_NETDEV_H_
+#define _NETFILTER_NETDEV_H_
+
+#include <linux/netfilter.h>
+#include <linux/netdevice.h>
+
+#ifdef CONFIG_NETFILTER_INGRESS
+static inline bool nf_hook_ingress_active(const struct sk_buff *skb)
+{
+#ifdef CONFIG_JUMP_LABEL
+	if (!static_key_false(&nf_hooks_needed[NFPROTO_NETDEV][NF_NETDEV_INGRESS]))
+		return false;
+#endif
+	return rcu_access_pointer(skb->dev->nf_hooks_ingress);
+}
+
+/* caller must hold rcu_read_lock */
+static inline int nf_hook_ingress(struct sk_buff *skb)
+{
+	struct nf_hook_entries *e = rcu_dereference(skb->dev->nf_hooks_ingress);
+	struct nf_hook_state state;
+	int ret;
+
+	/* Must recheck the ingress hook head, in the event it became NULL
+	 * after the check in nf_hook_ingress_active evaluated to true.
+	 */
+	if (unlikely(!e))
+		return 0;
+
+	nf_hook_state_init(&state, NF_NETDEV_INGRESS,
+			   NFPROTO_NETDEV, skb->dev, NULL, NULL,
+			   dev_net(skb->dev), NULL);
+	ret = nf_hook_slow(skb, &state, e, 0);
+	if (ret == 0)
+		return -1;
+
+	return ret;
+}
+
+#else /* CONFIG_NETFILTER_INGRESS */
+static inline int nf_hook_ingress_active(struct sk_buff *skb)
+{
+	return 0;
+}
+
+static inline int nf_hook_ingress(struct sk_buff *skb)
+{
+	return 0;
+}
+#endif /* CONFIG_NETFILTER_INGRESS */
+
+#ifdef CONFIG_NETFILTER_EGRESS
+static inline bool nf_hook_egress_active(void)
+{
+#ifdef CONFIG_JUMP_LABEL
+	if (!static_key_false(&nf_hooks_needed[NFPROTO_NETDEV][NF_NETDEV_EGRESS]))
+		return false;
+#endif
+	return true;
+}
+
+/**
+ * nf_hook_egress - classify packets before transmission
+ * @skb: packet to be classified
+ * @rc: result code which shall be returned by __dev_queue_xmit() on failure
+ * @dev: netdev whose egress hooks shall be applied to @skb
+ *
+ * Returns @skb on success or %NULL if the packet was consumed or filtered.
+ * Caller must hold rcu_read_lock.
+ *
+ * On ingress, packets are classified first by tc, then by netfilter.
+ * On egress, the order is reversed for symmetry.  Conceptually, tc and
+ * netfilter can be thought of as layers, with netfilter layered above tc:
+ * When tc redirects a packet to another interface, netfilter is not applied
+ * because the packet is on the tc layer.
+ *
+ * The nf_skip_egress flag controls whether netfilter is applied on egress.
+ * It is updated by __netif_receive_skb_core() and __dev_queue_xmit() when the
+ * packet passes through tc and netfilter.  Because __dev_queue_xmit() may be
+ * called recursively by tunnel drivers such as vxlan, the flag is reverted to
+ * false after sch_handle_egress().  This ensures that netfilter is applied
+ * both on the overlay and underlying network.
+ */
+static inline struct sk_buff *nf_hook_egress(struct sk_buff *skb, int *rc,
+					     struct net_device *dev)
+{
+	struct nf_hook_entries *e;
+	struct nf_hook_state state;
+	int ret;
+
+#ifdef CONFIG_NETFILTER_SKIP_EGRESS
+	if (skb->nf_skip_egress)
+		return skb;
+#endif
+
+	e = rcu_dereference(dev->nf_hooks_egress);
+	if (!e)
+		return skb;
+
+	nf_hook_state_init(&state, NF_NETDEV_EGRESS,
+			   NFPROTO_NETDEV, dev, NULL, NULL,
+			   dev_net(dev), NULL);
+	ret = nf_hook_slow(skb, &state, e, 0);
+
+	if (ret == 1) {
+		return skb;
+	} else if (ret < 0) {
+		*rc = NET_XMIT_DROP;
+		return NULL;
+	} else { /* ret == 0 */
+		*rc = NET_XMIT_SUCCESS;
+		return NULL;
+	}
+}
+#else /* CONFIG_NETFILTER_EGRESS */
+static inline bool nf_hook_egress_active(void)
+{
+	return false;
+}
+
+static inline struct sk_buff *nf_hook_egress(struct sk_buff *skb, int *rc,
+					     struct net_device *dev)
+{
+	return skb;
+}
+#endif /* CONFIG_NETFILTER_EGRESS */
+
+static inline void nf_skip_egress(struct sk_buff *skb, bool skip)
+{
+#ifdef CONFIG_NETFILTER_SKIP_EGRESS
+	skb->nf_skip_egress = skip;
+#endif
+}
+
+static inline void nf_hook_netdev_init(struct net_device *dev)
+{
+#ifdef CONFIG_NETFILTER_INGRESS
+	RCU_INIT_POINTER(dev->nf_hooks_ingress, NULL);
+#endif
+#ifdef CONFIG_NETFILTER_EGRESS
+	RCU_INIT_POINTER(dev->nf_hooks_egress, NULL);
+#endif
+}
+
+#endif /* _NETFILTER_NETDEV_H_ */
diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index 841e2f0f5240..cb96f1e6460c 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -652,6 +652,7 @@ typedef unsigned char *sk_buff_data_t;
  *	@tc_at_ingress: used within tc_classify to distinguish in/egress
  *	@redirected: packet was redirected by packet classifier
  *	@from_ingress: packet was redirected from the ingress path
+ *	@nf_skip_egress: packet shall skip nf egress - see netfilter_netdev.h
  *	@peeked: this packet has been seen already, so stats have been
  *		done for it, don't do them again
  *	@nf_trace: netfilter packet trace flag
@@ -868,6 +869,9 @@ struct sk_buff {
 #ifdef CONFIG_NET_REDIRECT
 	__u8			from_ingress:1;
 #endif
+#ifdef CONFIG_NETFILTER_SKIP_EGRESS
+	__u8			nf_skip_egress:1;
+#endif
 #ifdef CONFIG_TLS_DEVICE
 	__u8			decrypted:1;
 #endif
diff --git a/include/net/ip_vs.h b/include/net/ip_vs.h
index 7cb5a1aace40..ff1804a0c469 100644
--- a/include/net/ip_vs.h
+++ b/include/net/ip_vs.h
@@ -931,6 +931,7 @@ struct netns_ipvs {
 	int			sysctl_conn_reuse_mode;
 	int			sysctl_schedule_icmp;
 	int			sysctl_ignore_tunneled;
+	int			sysctl_run_estimation;
 
 	/* ip_vs_lblc */
 	int			sysctl_lblc_expiration;
@@ -1071,6 +1072,11 @@ static inline int sysctl_cache_bypass(struct netns_ipvs *ipvs)
 	return ipvs->sysctl_cache_bypass;
 }
 
+static inline int sysctl_run_estimation(struct netns_ipvs *ipvs)
+{
+	return ipvs->sysctl_run_estimation;
+}
+
 #else
 
 static inline int sysctl_sync_threshold(struct netns_ipvs *ipvs)
@@ -1163,6 +1169,11 @@ static inline int sysctl_cache_bypass(struct netns_ipvs *ipvs)
 	return 0;
 }
 
+static inline int sysctl_run_estimation(struct netns_ipvs *ipvs)
+{
+	return 1;
+}
+
 #endif
 
 /* IPVS core functions
diff --git a/include/uapi/linux/netfilter.h b/include/uapi/linux/netfilter.h
index ef9a44286e23..53411ccc69db 100644
--- a/include/uapi/linux/netfilter.h
+++ b/include/uapi/linux/netfilter.h
@@ -51,6 +51,7 @@ enum nf_inet_hooks {
 
 enum nf_dev_hooks {
 	NF_NETDEV_INGRESS,
+	NF_NETDEV_EGRESS,
 	NF_NETDEV_NUMHOOKS
 };