Merge tag 'batman-adv-for-davem' of git://git.open-mesh.org/linux-merge

Included changes:
- minimal fixes to the packet layout to avoid the __packed attribute when not
  needed
- new packet type called UNICAST_4ADDR: in this packet it is possible to find
  both source and destination node (in the classic UNICAST header only the
  destination field exists).
- a new feature: Distributed ARP Table (D.A.T.). It aims to reduce ARP lookups
  latency by means of a simil-DHT approach.

Documentation/networking/batman-adv.txt

@@ -203,7 +203,8 @@ abled  during run time. Following log_levels are defined:
 2 - Enable messages related to route added / changed / deleted
 4 - Enable messages related to translation table operations
 8 - Enable messages related to bridge loop avoidance
-15 - enable all messages
+16 - Enable messaged related to DAT, ARP snooping and parsing
+31 - Enable all messages
 
 The debug output can be changed at runtime  using  the  file
 /sys/class/net/bat0/mesh/log_level. e.g.

net/batman-adv/Kconfig

@@ -25,6 +25,16 @@ config BATMAN_ADV_BLA
 	  more than one mesh node in the same LAN, you can safely remove
 	  this feature and save some space.
 
+config BATMAN_ADV_DAT
+	bool "Distributed ARP Table"
+	depends on BATMAN_ADV && INET
+	default n
+	help
+	  This option enables DAT (Distributed ARP Table), a DHT based
+	  mechanism that increases ARP reliability on sparse wireless
+	  mesh networks. If you think that your network does not need
+	  this option you can safely remove it and save some space.
+
 config BATMAN_ADV_DEBUG
 	bool "B.A.T.M.A.N. debugging"
 	depends on BATMAN_ADV

net/batman-adv/Makefile

@@ -23,6 +23,7 @@ batman-adv-y += bat_iv_ogm.o
 batman-adv-y += bitarray.o
 batman-adv-$(CONFIG_BATMAN_ADV_BLA) += bridge_loop_avoidance.o
 batman-adv-y += debugfs.o
+batman-adv-$(CONFIG_BATMAN_ADV_DAT) += distributed-arp-table.o
 batman-adv-y += gateway_client.o
 batman-adv-y += gateway_common.o
 batman-adv-y += hard-interface.o

net/batman-adv/bat_iv_ogm.c

@@ -411,9 +411,11 @@ static void batadv_iv_ogm_aggregate_new(const unsigned char *packet_buff,
 
 	if ((atomic_read(&bat_priv->aggregated_ogms)) &&
 	    (packet_len < BATADV_MAX_AGGREGATION_BYTES))
-		skb_size = BATADV_MAX_AGGREGATION_BYTES + ETH_HLEN;
+		skb_size = BATADV_MAX_AGGREGATION_BYTES;
 	else
-		skb_size = packet_len + ETH_HLEN;
+		skb_size = packet_len;
+
+	skb_size += ETH_HLEN + NET_IP_ALIGN;
 
 	forw_packet_aggr->skb = dev_alloc_skb(skb_size);
 	if (!forw_packet_aggr->skb) {
@@ -422,7 +424,7 @@ static void batadv_iv_ogm_aggregate_new(const unsigned char *packet_buff,
 		kfree(forw_packet_aggr);
 		goto out;
 	}
-	skb_reserve(forw_packet_aggr->skb, ETH_HLEN);
+	skb_reserve(forw_packet_aggr->skb, ETH_HLEN + NET_IP_ALIGN);
 
 	INIT_HLIST_NODE(&forw_packet_aggr->list);
 

net/batman-adv/debugfs.c

@@ -31,6 +31,7 @@
 #include "vis.h"
 #include "icmp_socket.h"
 #include "bridge_loop_avoidance.h"
+#include "distributed-arp-table.h"
 
 static struct dentry *batadv_debugfs;
 
@@ -280,6 +281,19 @@ static int batadv_bla_backbone_table_open(struct inode *inode,
 
 #endif
 
+#ifdef CONFIG_BATMAN_ADV_DAT
+/**
+ * batadv_dat_cache_open - Prepare file handler for reads from dat_chache
+ * @inode: inode which was opened
+ * @file: file handle to be initialized
+ */
+static int batadv_dat_cache_open(struct inode *inode, struct file *file)
+{
+	struct net_device *net_dev = (struct net_device *)inode->i_private;
+	return single_open(file, batadv_dat_cache_seq_print_text, net_dev);
+}
+#endif
+
 static int batadv_transtable_local_open(struct inode *inode, struct file *file)
 {
 	struct net_device *net_dev = (struct net_device *)inode->i_private;
@@ -319,6 +333,9 @@ static BATADV_DEBUGINFO(bla_claim_table, S_IRUGO, batadv_bla_claim_table_open);
 static BATADV_DEBUGINFO(bla_backbone_table, S_IRUGO,
 			batadv_bla_backbone_table_open);
 #endif
+#ifdef CONFIG_BATMAN_ADV_DAT
+static BATADV_DEBUGINFO(dat_cache, S_IRUGO, batadv_dat_cache_open);
+#endif
 static BATADV_DEBUGINFO(transtable_local, S_IRUGO,
 			batadv_transtable_local_open);
 static BATADV_DEBUGINFO(vis_data, S_IRUGO, batadv_vis_data_open);
@@ -330,6 +347,9 @@ static struct batadv_debuginfo *batadv_mesh_debuginfos[] = {
 #ifdef CONFIG_BATMAN_ADV_BLA
 	&batadv_debuginfo_bla_claim_table,
 	&batadv_debuginfo_bla_backbone_table,
+#endif
+#ifdef CONFIG_BATMAN_ADV_DAT
+	&batadv_debuginfo_dat_cache,
 #endif
 	&batadv_debuginfo_transtable_local,
 	&batadv_debuginfo_vis_data,

net/batman-adv/distributed-arp-table.c

+/* Copyright (C) 2011-2012 B.A.T.M.A.N. contributors:
+ *
+ * Antonio Quartulli
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of version 2 of the GNU General Public
+ * License as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
+ * 02110-1301, USA
+ */
+
+#include <linux/if_ether.h>
+#include <linux/if_arp.h>
+#include <net/arp.h>
+
+#include "main.h"
+#include "hash.h"
+#include "distributed-arp-table.h"
+#include "hard-interface.h"
+#include "originator.h"
+#include "send.h"
+#include "types.h"
+#include "translation-table.h"
+#include "unicast.h"
+
+static void batadv_dat_purge(struct work_struct *work);
+
+/**
+ * batadv_dat_start_timer - initialise the DAT periodic worker
+ * @bat_priv: the bat priv with all the soft interface information
+ */
+static void batadv_dat_start_timer(struct batadv_priv *bat_priv)
+{
+	INIT_DELAYED_WORK(&bat_priv->dat.work, batadv_dat_purge);
+	queue_delayed_work(batadv_event_workqueue, &bat_priv->dat.work,
+			   msecs_to_jiffies(10000));
+}
+
+/**
+ * batadv_dat_entry_free_ref - decrements the dat_entry refcounter and possibly
+ * free it
+ * @dat_entry: the oentry to free
+ */
+static void batadv_dat_entry_free_ref(struct batadv_dat_entry *dat_entry)
+{
+	if (atomic_dec_and_test(&dat_entry->refcount))
+		kfree_rcu(dat_entry, rcu);
+}
+
+/**
+ * batadv_dat_to_purge - checks whether a dat_entry has to be purged or not
+ * @dat_entry: the entry to check
+ *
+ * Returns true if the entry has to be purged now, false otherwise
+ */
+static bool batadv_dat_to_purge(struct batadv_dat_entry *dat_entry)
+{
+	return batadv_has_timed_out(dat_entry->last_update,
+				    BATADV_DAT_ENTRY_TIMEOUT);
+}
+
+/**
+ * __batadv_dat_purge - delete entries from the DAT local storage
+ * @bat_priv: the bat priv with all the soft interface information
+ * @to_purge: function in charge to decide whether an entry has to be purged or
+ *	      not. This function takes the dat_entry as argument and has to
+ *	      returns a boolean value: true is the entry has to be deleted,
+ *	      false otherwise
+ *
+ * Loops over each entry in the DAT local storage and delete it if and only if
+ * the to_purge function passed as argument returns true
+ */
+static void __batadv_dat_purge(struct batadv_priv *bat_priv,
+			       bool (*to_purge)(struct batadv_dat_entry *))
+{
+	spinlock_t *list_lock; /* protects write access to the hash lists */
+	struct batadv_dat_entry *dat_entry;
+	struct hlist_node *node, *node_tmp;
+	struct hlist_head *head;
+	uint32_t i;
+
+	if (!bat_priv->dat.hash)
+		return;
+
+	for (i = 0; i < bat_priv->dat.hash->size; i++) {
+		head = &bat_priv->dat.hash->table[i];
+		list_lock = &bat_priv->dat.hash->list_locks[i];
+
+		spin_lock_bh(list_lock);
+		hlist_for_each_entry_safe(dat_entry, node, node_tmp, head,
+					  hash_entry) {
+			/* if an helper function has been passed as parameter,
+			 * ask it if the entry has to be purged or not
+			 */
+			if (to_purge && !to_purge(dat_entry))
+				continue;
+
+			hlist_del_rcu(node);
+			batadv_dat_entry_free_ref(dat_entry);
+		}
+		spin_unlock_bh(list_lock);
+	}
+}
+
+/**
+ * batadv_dat_purge - periodic task that deletes old entries from the local DAT
+ * hash table
+ * @work: kernel work struct
+ */
+static void batadv_dat_purge(struct work_struct *work)
+{
+	struct delayed_work *delayed_work;
+	struct batadv_priv_dat *priv_dat;
+	struct batadv_priv *bat_priv;
+
+	delayed_work = container_of(work, struct delayed_work, work);
+	priv_dat = container_of(delayed_work, struct batadv_priv_dat, work);
+	bat_priv = container_of(priv_dat, struct batadv_priv, dat);
+
+	__batadv_dat_purge(bat_priv, batadv_dat_to_purge);
+	batadv_dat_start_timer(bat_priv);
+}
+
+/**
+ * batadv_compare_dat - comparing function used in the local DAT hash table
+ * @node: node in the local table
+ * @data2: second object to compare the node to
+ *
+ * Returns 1 if the two entry are the same, 0 otherwise
+ */
+static int batadv_compare_dat(const struct hlist_node *node, const void *data2)
+{
+	const void *data1 = container_of(node, struct batadv_dat_entry,
+					 hash_entry);
+
+	return (memcmp(data1, data2, sizeof(__be32)) == 0 ? 1 : 0);
+}
+
+/**
+ * batadv_arp_hw_src - extract the hw_src field from an ARP packet
+ * @skb: ARP packet
+ * @hdr_size: size of the possible header before the ARP packet
+ *
+ * Returns the value of the hw_src field in the ARP packet
+ */
+static uint8_t *batadv_arp_hw_src(struct sk_buff *skb, int hdr_size)
+{
+	uint8_t *addr;
+
+	addr = (uint8_t *)(skb->data + hdr_size);
+	addr += ETH_HLEN + sizeof(struct arphdr);
+
+	return addr;
+}
+
+/**
+ * batadv_arp_ip_src - extract the ip_src field from an ARP packet
+ * @skb: ARP packet
+ * @hdr_size: size of the possible header before the ARP packet
+ *
+ * Returns the value of the ip_src field in the ARP packet
+ */
+static __be32 batadv_arp_ip_src(struct sk_buff *skb, int hdr_size)
+{
+	return *(__be32 *)(batadv_arp_hw_src(skb, hdr_size) + ETH_ALEN);
+}
+
+/**
+ * batadv_arp_hw_dst - extract the hw_dst field from an ARP packet
+ * @skb: ARP packet
+ * @hdr_size: size of the possible header before the ARP packet
+ *
+ * Returns the value of the hw_dst field in the ARP packet
+ */
+static uint8_t *batadv_arp_hw_dst(struct sk_buff *skb, int hdr_size)
+{
+	return batadv_arp_hw_src(skb, hdr_size) + ETH_ALEN + 4;
+}
+
+/**
+ * batadv_arp_ip_dst - extract the ip_dst field from an ARP packet
+ * @skb: ARP packet
+ * @hdr_size: size of the possible header before the ARP packet
+ *
+ * Returns the value of the ip_dst field in the ARP packet
+ */
+static __be32 batadv_arp_ip_dst(struct sk_buff *skb, int hdr_size)
+{
+	return *(__be32 *)(batadv_arp_hw_src(skb, hdr_size) + ETH_ALEN * 2 + 4);
+}
+
+/**
+ * batadv_hash_dat - compute the hash value for an IP address
+ * @data: data to hash
+ * @size: size of the hash table
+ *
+ * Returns the selected index in the hash table for the given data
+ */
+static uint32_t batadv_hash_dat(const void *data, uint32_t size)
+{
+	const unsigned char *key = data;
+	uint32_t hash = 0;
+	size_t i;
+
+	for (i = 0; i < 4; i++) {
+		hash += key[i];
+		hash += (hash << 10);
+		hash ^= (hash >> 6);
+	}
+
+	hash += (hash << 3);
+	hash ^= (hash >> 11);
+	hash += (hash << 15);
+
+	return hash % size;
+}
+
+/**
+ * batadv_dat_entry_hash_find - looks for a given dat_entry in the local hash
+ * table
+ * @bat_priv: the bat priv with all the soft interface information
+ * @ip: search key
+ *
+ * Returns the dat_entry if found, NULL otherwise
+ */
+static struct batadv_dat_entry *
+batadv_dat_entry_hash_find(struct batadv_priv *bat_priv, __be32 ip)
+{
+	struct hlist_head *head;
+	struct hlist_node *node;
+	struct batadv_dat_entry *dat_entry, *dat_entry_tmp = NULL;
+	struct batadv_hashtable *hash = bat_priv->dat.hash;
+	uint32_t index;
+
+	if (!hash)
+		return NULL;
+
+	index = batadv_hash_dat(&ip, hash->size);
+	head = &hash->table[index];
+
+	rcu_read_lock();
+	hlist_for_each_entry_rcu(dat_entry, node, head, hash_entry) {
+		if (dat_entry->ip != ip)
+			continue;
+
+		if (!atomic_inc_not_zero(&dat_entry->refcount))
+			continue;
+
+		dat_entry_tmp = dat_entry;
+		break;
+	}
+	rcu_read_unlock();
+
+	return dat_entry_tmp;
+}
+
+/**
+ * batadv_dat_entry_add - add a new dat entry or update it if already exists
+ * @bat_priv: the bat priv with all the soft interface information
+ * @ip: ipv4 to add/edit
+ * @mac_addr: mac address to assign to the given ipv4
+ */
+static void batadv_dat_entry_add(struct batadv_priv *bat_priv, __be32 ip,
+				 uint8_t *mac_addr)
+{
+	struct batadv_dat_entry *dat_entry;
+	int hash_added;
+
+	dat_entry = batadv_dat_entry_hash_find(bat_priv, ip);
+	/* if this entry is already known, just update it */
+	if (dat_entry) {
+		if (!batadv_compare_eth(dat_entry->mac_addr, mac_addr))
+			memcpy(dat_entry->mac_addr, mac_addr, ETH_ALEN);
+		dat_entry->last_update = jiffies;
+		batadv_dbg(BATADV_DBG_DAT, bat_priv,
+			   "Entry updated: %pI4 %pM\n", &dat_entry->ip,
+			   dat_entry->mac_addr);
+		goto out;
+	}
+
+	dat_entry = kmalloc(sizeof(*dat_entry), GFP_ATOMIC);
+	if (!dat_entry)
+		goto out;
+
+	dat_entry->ip = ip;
+	memcpy(dat_entry->mac_addr, mac_addr, ETH_ALEN);
+	dat_entry->last_update = jiffies;
+	atomic_set(&dat_entry->refcount, 2);
+
+	hash_added = batadv_hash_add(bat_priv->dat.hash, batadv_compare_dat,
+				     batadv_hash_dat, &dat_entry->ip,
+				     &dat_entry->hash_entry);
+
+	if (unlikely(hash_added != 0)) {
+		/* remove the reference for the hash */
+		batadv_dat_entry_free_ref(dat_entry);
+		goto out;
+	}
+
+	batadv_dbg(BATADV_DBG_DAT, bat_priv, "New entry added: %pI4 %pM\n",
+		   &dat_entry->ip, dat_entry->mac_addr);
+
+out:
+	if (dat_entry)
+		batadv_dat_entry_free_ref(dat_entry);
+}
+
+#ifdef CONFIG_BATMAN_ADV_DEBUG
+
+/**
+ * batadv_dbg_arp - print a debug message containing all the ARP packet details
+ * @bat_priv: the bat priv with all the soft interface information
+ * @skb: ARP packet
+ * @type: ARP type
+ * @hdr_size: size of the possible header before the ARP packet
+ * @msg: message to print together with the debugging information
+ */
+static void batadv_dbg_arp(struct batadv_priv *bat_priv, struct sk_buff *skb,
+			   uint16_t type, int hdr_size, char *msg)
+{
+	struct batadv_unicast_4addr_packet *unicast_4addr_packet;
+	struct batadv_bcast_packet *bcast_pkt;
+	uint8_t *orig_addr;
+	__be32 ip_src, ip_dst;
+
+	if (msg)
+		batadv_dbg(BATADV_DBG_DAT, bat_priv, "%s\n", msg);
+
+	ip_src = batadv_arp_ip_src(skb, hdr_size);
+	ip_dst = batadv_arp_ip_dst(skb, hdr_size);
+	batadv_dbg(BATADV_DBG_DAT, bat_priv,
+		   "ARP MSG = [src: %pM-%pI4 dst: %pM-%pI4]\n",
+		   batadv_arp_hw_src(skb, hdr_size), &ip_src,
+		   batadv_arp_hw_dst(skb, hdr_size), &ip_dst);
+
+	if (hdr_size == 0)
+		return;
+
+	/* if the ARP packet is encapsulated in a batman packet, let's print
+	 * some debug messages
+	 */
+	unicast_4addr_packet = (struct batadv_unicast_4addr_packet *)skb->data;
+
+	switch (unicast_4addr_packet->u.header.packet_type) {
+	case BATADV_UNICAST:
+		batadv_dbg(BATADV_DBG_DAT, bat_priv,
+			   "* encapsulated within a UNICAST packet\n");
+		break;
+	case BATADV_UNICAST_4ADDR:
+		batadv_dbg(BATADV_DBG_DAT, bat_priv,
+			   "* encapsulated within a UNICAST_4ADDR packet (src: %pM)\n",
+			   unicast_4addr_packet->src);
+		switch (unicast_4addr_packet->subtype) {
+		case BATADV_P_DAT_DHT_PUT:
+			batadv_dbg(BATADV_DBG_DAT, bat_priv, "* type: DAT_DHT_PUT\n");
+			break;
+		case BATADV_P_DAT_DHT_GET:
+			batadv_dbg(BATADV_DBG_DAT, bat_priv, "* type: DAT_DHT_GET\n");
+			break;
+		case BATADV_P_DAT_CACHE_REPLY:
+			batadv_dbg(BATADV_DBG_DAT, bat_priv,
+				   "* type: DAT_CACHE_REPLY\n");
+			break;
+		case BATADV_P_DATA:
+			batadv_dbg(BATADV_DBG_DAT, bat_priv, "* type: DATA\n");
+			break;
+		default:
+			batadv_dbg(BATADV_DBG_DAT, bat_priv, "* type: Unknown (%u)!\n",
+				   unicast_4addr_packet->u.header.packet_type);
+		}
+		break;
+	case BATADV_BCAST:
+		bcast_pkt = (struct batadv_bcast_packet *)unicast_4addr_packet;
+		orig_addr = bcast_pkt->orig;
+		batadv_dbg(BATADV_DBG_DAT, bat_priv,
+			   "* encapsulated within a BCAST packet (src: %pM)\n",
+			   orig_addr);
+		break;
+	default:
+		batadv_dbg(BATADV_DBG_DAT, bat_priv,
+			   "* encapsulated within an unknown packet type (0x%x)\n",
+			   unicast_4addr_packet->u.header.packet_type);
+	}
+}
+
+#else
+
+static void batadv_dbg_arp(struct batadv_priv *bat_priv, struct sk_buff *skb,
+			   uint16_t type, int hdr_size, char *msg)
+{
+}
+
+#endif /* CONFIG_BATMAN_ADV_DEBUG */
+
+/**
+ * batadv_is_orig_node_eligible - check whether a node can be a DHT candidate
+ * @res: the array with the already selected candidates
+ * @select: number of already selected candidates
+ * @tmp_max: address of the currently evaluated node
+ * @max: current round max address
+ * @last_max: address of the last selected candidate
+ * @candidate: orig_node under evaluation
+ * @max_orig_node: last selected candidate
+ *
+ * Returns true if the node has been elected as next candidate or false othrwise
+ */
+static bool batadv_is_orig_node_eligible(struct batadv_dat_candidate *res,
+					 int select, batadv_dat_addr_t tmp_max,
+					 batadv_dat_addr_t max,
+					 batadv_dat_addr_t last_max,
+					 struct batadv_orig_node *candidate,
+					 struct batadv_orig_node *max_orig_node)
+{
+	bool ret = false;
+	int j;
+
+	/* Check if this node has already been selected... */
+	for (j = 0; j < select; j++)
+		if (res[j].orig_node == candidate)
+			break;
+	/* ..and possibly skip it */
+	if (j < select)
+		goto out;
+	/* sanity check: has it already been selected? This should not happen */
+	if (tmp_max > last_max)
+		goto out;
+	/* check if during this iteration an originator with a closer dht
+	 * address has already been found
+	 */
+	if (tmp_max < max)
+		goto out;
+	/* this is an hash collision with the temporary selected node. Choose
+	 * the one with the lowest address
+	 */
+	if ((tmp_max == max) &&
+	    (batadv_compare_eth(candidate->orig, max_orig_node->orig) > 0))
+		goto out;
+
+	ret = true;
+out:
+	return ret;
+}
+
+/**
+ * batadv_choose_next_candidate - select the next DHT candidate
+ * @bat_priv: the bat priv with all the soft interface information
+ * @cands: candidates array
+ * @select: number of candidates already present in the array
+ * @ip_key: key to look up in the DHT
+ * @last_max: pointer where the address of the selected candidate will be saved
+ */
+static void batadv_choose_next_candidate(struct batadv_priv *bat_priv,
+					 struct batadv_dat_candidate *cands,
+					 int select, batadv_dat_addr_t ip_key,
+					 batadv_dat_addr_t *last_max)
+{
+	batadv_dat_addr_t max = 0, tmp_max = 0;
+	struct batadv_orig_node *orig_node, *max_orig_node = NULL;
+	struct batadv_hashtable *hash = bat_priv->orig_hash;
+	struct hlist_node *node;
+	struct hlist_head *head;
+	int i;
+
+	/* if no node is eligible as candidate, leave the candidate type as
+	 * NOT_FOUND
+	 */
+	cands[select].type = BATADV_DAT_CANDIDATE_NOT_FOUND;
+
+	/* iterate over the originator list and find the node with closest
+	 * dat_address which has not been selected yet
+	 */
+	for (i = 0; i < hash->size; i++) {
+		head = &hash->table[i];
+
+		rcu_read_lock();
+		hlist_for_each_entry_rcu(orig_node, node, head, hash_entry) {
+			/* the dht space is a ring and addresses are unsigned */
+			tmp_max = BATADV_DAT_ADDR_MAX - orig_node->dat_addr +
+				  ip_key;
+
+			if (!batadv_is_orig_node_eligible(cands, select,
+							  tmp_max, max,
+							  *last_max, orig_node,
+							  max_orig_node))
+				continue;
+
+			if (!atomic_inc_not_zero(&orig_node->refcount))
+				continue;
+
+			max = tmp_max;
+			if (max_orig_node)
+				batadv_orig_node_free_ref(max_orig_node);
+			max_orig_node = orig_node;
+		}
+		rcu_read_unlock();
+	}
+	if (max_orig_node) {
+		cands[select].type = BATADV_DAT_CANDIDATE_ORIG;
+		cands[select].orig_node = max_orig_node;
+		batadv_dbg(BATADV_DBG_DAT, bat_priv,
+			   "dat_select_candidates() %d: selected %pM addr=%u dist=%u\n",
+			   select, max_orig_node->orig, max_orig_node->dat_addr,
+			   max);
+	}
+	*last_max = max;
+}
+
+/**
+ * batadv_dat_select_candidates - selects the nodes which the DHT message has to
+ * be sent to
+ * @bat_priv: the bat priv with all the soft interface information
+ * @ip_dst: ipv4 to look up in the DHT
+ *
+ * An originator O is selected if and only if its DHT_ID value is one of three
+ * closest values (from the LEFT, with wrap around if needed) then the hash
+ * value of the key. ip_dst is the key.
+ *
+ * Returns the candidate array of size BATADV_DAT_CANDIDATE_NUM
+ */
+static struct batadv_dat_candidate *
+batadv_dat_select_candidates(struct batadv_priv *bat_priv, __be32 ip_dst)
+{
+	int select;
+	batadv_dat_addr_t last_max = BATADV_DAT_ADDR_MAX, ip_key;
+	struct batadv_dat_candidate *res;
+
+	if (!bat_priv->orig_hash)
+		return NULL;
+
+	res = kmalloc(BATADV_DAT_CANDIDATES_NUM * sizeof(*res), GFP_ATOMIC);
+	if (!res)
+		return NULL;
+
+	ip_key = (batadv_dat_addr_t)batadv_hash_dat(&ip_dst,
+						    BATADV_DAT_ADDR_MAX);
+
+	batadv_dbg(BATADV_DBG_DAT, bat_priv,
+		   "dat_select_candidates(): IP=%pI4 hash(IP)=%u\n", &ip_dst,
+		   ip_key);
+
+	for (select = 0; select < BATADV_DAT_CANDIDATES_NUM; select++)
+		batadv_choose_next_candidate(bat_priv, res, select, ip_key,
+					     &last_max);
+
+	return res;
+}
+
+/**
+ * batadv_dat_send_data - send a payload to the selected candidates
+ * @bat_priv: the bat priv with all the soft interface information
+ * @skb: payload to send
+ * @ip: the DHT key
+ * @packet_subtype: unicast4addr packet subtype to use
+ *
+ * In this function the skb is copied by means of pskb_copy() and is sent as
+ * unicast packet to each of the selected candidates
+ *
+ * Returns true if the packet is sent to at least one candidate, false otherwise
+ */
+static bool batadv_dat_send_data(struct batadv_priv *bat_priv,
+				 struct sk_buff *skb, __be32 ip,
+				 int packet_subtype)
+{
+	int i;
+	bool ret = false;
+	int send_status;
+	struct batadv_neigh_node *neigh_node = NULL;
+	struct sk_buff *tmp_skb;
+	struct batadv_dat_candidate *cand;
+
+	cand = batadv_dat_select_candidates(bat_priv, ip);
+	if (!cand)
+		goto out;
+
+	batadv_dbg(BATADV_DBG_DAT, bat_priv, "DHT_SEND for %pI4\n", &ip);
+
+	for (i = 0; i < BATADV_DAT_CANDIDATES_NUM; i++) {
+		if (cand[i].type == BATADV_DAT_CANDIDATE_NOT_FOUND)
+			continue;
+
+		neigh_node = batadv_orig_node_get_router(cand[i].orig_node);
+		if (!neigh_node)
+			goto free_orig;
+
+		tmp_skb = pskb_copy(skb, GFP_ATOMIC);
+		if (!batadv_unicast_4addr_prepare_skb(bat_priv, tmp_skb,
+						      cand[i].orig_node,
+						      packet_subtype)) {
+			kfree_skb(tmp_skb);
+			goto free_neigh;
+		}
+
+		send_status = batadv_send_skb_packet(tmp_skb,
+						     neigh_node->if_incoming,
+						     neigh_node->addr);
+		if (send_status == NET_XMIT_SUCCESS) {
+			/* count the sent packet */
+			switch (packet_subtype) {
+			case BATADV_P_DAT_DHT_GET:
+				batadv_inc_counter(bat_priv,
+						   BATADV_CNT_DAT_GET_TX);
+				break;
+			case BATADV_P_DAT_DHT_PUT:
+				batadv_inc_counter(bat_priv,
+						   BATADV_CNT_DAT_PUT_TX);
+				break;
+			}
+
+			/* packet sent to a candidate: return true */
+			ret = true;
+		}
+free_neigh:
+		batadv_neigh_node_free_ref(neigh_node);
+free_orig:
+		batadv_orig_node_free_ref(cand[i].orig_node);
+	}
+
+out:
+	kfree(cand);
+	return ret;
+}
+
+/**
+ * batadv_dat_hash_free - free the local DAT hash table
+ * @bat_priv: the bat priv with all the soft interface information
+ */
+static void batadv_dat_hash_free(struct batadv_priv *bat_priv)
+{
+	if (!bat_priv->dat.hash)
+		return;
+
+	__batadv_dat_purge(bat_priv, NULL);
+
+	batadv_hash_destroy(bat_priv->dat.hash);
+
+	bat_priv->dat.hash = NULL;
+}
+
+/**
+ * batadv_dat_init - initialise the DAT internals
+ * @bat_priv: the bat priv with all the soft interface information
+ */
+int batadv_dat_init(struct batadv_priv *bat_priv)
+{
+	if (bat_priv->dat.hash)
+		return 0;
+
+	bat_priv->dat.hash = batadv_hash_new(1024);
+
+	if (!bat_priv->dat.hash)
+		return -ENOMEM;
+
+	batadv_dat_start_timer(bat_priv);
+
+	return 0;
+}
+
+/**
+ * batadv_dat_free - free the DAT internals
+ * @bat_priv: the bat priv with all the soft interface information
+ */
+void batadv_dat_free(struct batadv_priv *bat_priv)
+{
+	cancel_delayed_work_sync(&bat_priv->dat.work);
+
+	batadv_dat_hash_free(bat_priv);
+}
+
+/**
+ * batadv_dat_cache_seq_print_text - print the local DAT hash table
+ * @seq: seq file to print on
+ * @offset: not used
+ */
+int batadv_dat_cache_seq_print_text(struct seq_file *seq, void *offset)
+{
+	struct net_device *net_dev = (struct net_device *)seq->private;
+	struct batadv_priv *bat_priv = netdev_priv(net_dev);
+	struct batadv_hashtable *hash = bat_priv->dat.hash;
+	struct batadv_dat_entry *dat_entry;
+	struct batadv_hard_iface *primary_if;
+	struct hlist_node *node;
+	struct hlist_head *head;
+	unsigned long last_seen_jiffies;
+	int last_seen_msecs, last_seen_secs, last_seen_mins;
+	uint32_t i;
+
+	primary_if = batadv_seq_print_text_primary_if_get(seq);
+	if (!primary_if)
+		goto out;
+
+	seq_printf(seq, "Distributed ARP Table (%s):\n", net_dev->name);
+	seq_printf(seq, "          %-7s          %-13s %5s\n", "IPv4", "MAC",
+		   "last-seen");
+
+	for (i = 0; i < hash->size; i++) {
+		head = &hash->table[i];
+
+		rcu_read_lock();
+		hlist_for_each_entry_rcu(dat_entry, node, head, hash_entry) {
+			last_seen_jiffies = jiffies - dat_entry->last_update;
+			last_seen_msecs = jiffies_to_msecs(last_seen_jiffies);
+			last_seen_mins = last_seen_msecs / 60000;
+			last_seen_msecs = last_seen_msecs % 60000;
+			last_seen_secs = last_seen_msecs / 1000;
+
+			seq_printf(seq, " * %15pI4 %14pM %6i:%02i\n",
+				   &dat_entry->ip, dat_entry->mac_addr,
+				   last_seen_mins, last_seen_secs);
+		}
+		rcu_read_unlock();
+	}
+
+out:
+	if (primary_if)
+		batadv_hardif_free_ref(primary_if);
+	return 0;
+}
+
+/**
+ * batadv_arp_get_type - parse an ARP packet and gets the type
+ * @bat_priv: the bat priv with all the soft interface information
+ * @skb: packet to analyse
+ * @hdr_size: size of the possible header before the ARP packet in the skb
+ *
+ * Returns the ARP type if the skb contains a valid ARP packet, 0 otherwise
+ */
+static uint16_t batadv_arp_get_type(struct batadv_priv *bat_priv,
+				    struct sk_buff *skb, int hdr_size)
+{
+	struct arphdr *arphdr;
+	struct ethhdr *ethhdr;
+	__be32 ip_src, ip_dst;
+	uint16_t type = 0;
+
+	/* pull the ethernet header */
+	if (unlikely(!pskb_may_pull(skb, hdr_size + ETH_HLEN)))
+		goto out;
+
+	ethhdr = (struct ethhdr *)(skb->data + hdr_size);
+
+	if (ethhdr->h_proto != htons(ETH_P_ARP))
+		goto out;
+
+	/* pull the ARP payload */
+	if (unlikely(!pskb_may_pull(skb, hdr_size + ETH_HLEN +
+				    arp_hdr_len(skb->dev))))
+		goto out;
+
+	arphdr = (struct arphdr *)(skb->data + hdr_size + ETH_HLEN);
+
+	/* Check whether the ARP packet carries a valid
+	 * IP information
+	 */
+	if (arphdr->ar_hrd != htons(ARPHRD_ETHER))
+		goto out;
+
+	if (arphdr->ar_pro != htons(ETH_P_IP))
+		goto out;
+
+	if (arphdr->ar_hln != ETH_ALEN)
+		goto out;
+
+	if (arphdr->ar_pln != 4)
+		goto out;
+
+	/* Check for bad reply/request. If the ARP message is not sane, DAT
+	 * will simply ignore it
+	 */
+	ip_src = batadv_arp_ip_src(skb, hdr_size);
+	ip_dst = batadv_arp_ip_dst(skb, hdr_size);
+	if (ipv4_is_loopback(ip_src) || ipv4_is_multicast(ip_src) ||
+	    ipv4_is_loopback(ip_dst) || ipv4_is_multicast(ip_dst))
+		goto out;
+
+	type = ntohs(arphdr->ar_op);
+out:
+	return type;
+}
+
+/**
+ * batadv_dat_snoop_outgoing_arp_request - snoop the ARP request and try to
+ * answer using DAT
+ * @bat_priv: the bat priv with all the soft interface information
+ * @skb: packet to check
+ *
+ * Returns true if the message has been sent to the dht candidates, false
+ * otherwise. In case of true the message has to be enqueued to permit the
+ * fallback
+ */
+bool batadv_dat_snoop_outgoing_arp_request(struct batadv_priv *bat_priv,
+					   struct sk_buff *skb)
+{
+	uint16_t type = 0;
+	__be32 ip_dst, ip_src;
+	uint8_t *hw_src;
+	bool ret = false;
+	struct batadv_dat_entry *dat_entry = NULL;
+	struct sk_buff *skb_new;
+	struct batadv_hard_iface *primary_if = NULL;
+
+	if (!atomic_read(&bat_priv->distributed_arp_table))
+		goto out;
+
+	type = batadv_arp_get_type(bat_priv, skb, 0);
+	/* If the node gets an ARP_REQUEST it has to send a DHT_GET unicast
+	 * message to the selected DHT candidates
+	 */
+	if (type != ARPOP_REQUEST)
+		goto out;
+
+	batadv_dbg_arp(bat_priv, skb, type, 0, "Parsing outgoing ARP REQUEST");
+
+	ip_src = batadv_arp_ip_src(skb, 0);
+	hw_src = batadv_arp_hw_src(skb, 0);
+	ip_dst = batadv_arp_ip_dst(skb, 0);
+
+	batadv_dat_entry_add(bat_priv, ip_src, hw_src);
+
+	dat_entry = batadv_dat_entry_hash_find(bat_priv, ip_dst);
+	if (dat_entry) {
+		primary_if = batadv_primary_if_get_selected(bat_priv);
+		if (!primary_if)
+			goto out;
+
+		skb_new = arp_create(ARPOP_REPLY, ETH_P_ARP, ip_src,
+				     primary_if->soft_iface, ip_dst, hw_src,
+				     dat_entry->mac_addr, hw_src);
+		if (!skb_new)
+			goto out;
+
+		skb_reset_mac_header(skb_new);
+		skb_new->protocol = eth_type_trans(skb_new,
+						   primary_if->soft_iface);
+		bat_priv->stats.rx_packets++;
+		bat_priv->stats.rx_bytes += skb->len + ETH_HLEN;
+		primary_if->soft_iface->last_rx = jiffies;
+
+		netif_rx(skb_new);
+		batadv_dbg(BATADV_DBG_DAT, bat_priv, "ARP request replied locally\n");
+		ret = true;
+	} else {
+		/* Send the request on the DHT */
+		ret = batadv_dat_send_data(bat_priv, skb, ip_dst,
+					   BATADV_P_DAT_DHT_GET);
+	}
+out:
+	if (dat_entry)
+		batadv_dat_entry_free_ref(dat_entry);
+	if (primary_if)
+		batadv_hardif_free_ref(primary_if);
+	return ret;
+}
+
+/**
+ * batadv_dat_snoop_incoming_arp_request - snoop the ARP request and try to
+ * answer using the local DAT storage
+ * @bat_priv: the bat priv with all the soft interface information
+ * @skb: packet to check
+ * @hdr_size: size of the encapsulation header
+ *
+ * Returns true if the request has been answered, false otherwise
+ */
+bool batadv_dat_snoop_incoming_arp_request(struct batadv_priv *bat_priv,
+					   struct sk_buff *skb, int hdr_size)
+{
+	uint16_t type;
+	__be32 ip_src, ip_dst;
+	uint8_t *hw_src;
+	struct sk_buff *skb_new;
+	struct batadv_hard_iface *primary_if = NULL;
+	struct batadv_dat_entry *dat_entry = NULL;
+	bool ret = false;
+	int err;
+
+	if (!atomic_read(&bat_priv->distributed_arp_table))
+		goto out;
+
+	type = batadv_arp_get_type(bat_priv, skb, hdr_size);
+	if (type != ARPOP_REQUEST)
+		goto out;
+
+	hw_src = batadv_arp_hw_src(skb, hdr_size);
+	ip_src = batadv_arp_ip_src(skb, hdr_size);
+	ip_dst = batadv_arp_ip_dst(skb, hdr_size);
+
+	batadv_dbg_arp(bat_priv, skb, type, hdr_size,
+		       "Parsing incoming ARP REQUEST");
+
+	batadv_dat_entry_add(bat_priv, ip_src, hw_src);
+
+	dat_entry = batadv_dat_entry_hash_find(bat_priv, ip_dst);
+	if (!dat_entry)
+		goto out;
+
+	primary_if = batadv_primary_if_get_selected(bat_priv);
+	if (!primary_if)
+		goto out;
+
+	skb_new = arp_create(ARPOP_REPLY, ETH_P_ARP, ip_src,
+			     primary_if->soft_iface, ip_dst, hw_src,
+			     dat_entry->mac_addr, hw_src);
+
+	if (!skb_new)
+		goto out;
+
+	/* to preserve backwards compatibility, here the node has to answer
+	 * using the same packet type it received for the request. This is due
+	 * to that if a node is not using the 4addr packet format it may not
+	 * support it.
+	 */
+	if (hdr_size == sizeof(struct batadv_unicast_4addr_packet))
+		err = batadv_unicast_4addr_send_skb(bat_priv, skb_new,
+						    BATADV_P_DAT_CACHE_REPLY);
+	else
+		err = batadv_unicast_send_skb(bat_priv, skb_new);
+
+	if (!err) {
+		batadv_inc_counter(bat_priv, BATADV_CNT_DAT_CACHED_REPLY_TX);
+		ret = true;
+	}
+out:
+	if (dat_entry)
+		batadv_dat_entry_free_ref(dat_entry);
+	if (primary_if)
+		batadv_hardif_free_ref(primary_if);
+	if (ret)
+		kfree_skb(skb);
+	return ret;
+}
+
+/**
+ * batadv_dat_snoop_outgoing_arp_reply - snoop the ARP reply and fill the DHT
+ * @bat_priv: the bat priv with all the soft interface information
+ * @skb: packet to check
+ */
+void batadv_dat_snoop_outgoing_arp_reply(struct batadv_priv *bat_priv,
+					 struct sk_buff *skb)
+{
+	uint16_t type;
+	__be32 ip_src, ip_dst;
+	uint8_t *hw_src, *hw_dst;
+
+	if (!atomic_read(&bat_priv->distributed_arp_table))
+		return;
+
+	type = batadv_arp_get_type(bat_priv, skb, 0);
+	if (type != ARPOP_REPLY)
+		return;
+
+	batadv_dbg_arp(bat_priv, skb, type, 0, "Parsing outgoing ARP REPLY");
+
+	hw_src = batadv_arp_hw_src(skb, 0);
+	ip_src = batadv_arp_ip_src(skb, 0);
+	hw_dst = batadv_arp_hw_dst(skb, 0);
+	ip_dst = batadv_arp_ip_dst(skb, 0);
+
+	batadv_dat_entry_add(bat_priv, ip_src, hw_src);
+	batadv_dat_entry_add(bat_priv, ip_dst, hw_dst);
+
+	/* Send the ARP reply to the candidates for both the IP addresses that
+	 * the node got within the ARP reply
+	 */
+	batadv_dat_send_data(bat_priv, skb, ip_src, BATADV_P_DAT_DHT_PUT);
+	batadv_dat_send_data(bat_priv, skb, ip_dst, BATADV_P_DAT_DHT_PUT);
+}
+/**
+ * batadv_dat_snoop_incoming_arp_reply - snoop the ARP reply and fill the local
+ * DAT storage only
+ * @bat_priv: the bat priv with all the soft interface information
+ * @skb: packet to check
+ * @hdr_size: siaze of the encapsulation header
+ */
+bool batadv_dat_snoop_incoming_arp_reply(struct batadv_priv *bat_priv,
+					 struct sk_buff *skb, int hdr_size)
+{
+	uint16_t type;
+	__be32 ip_src, ip_dst;
+	uint8_t *hw_src, *hw_dst;
+	bool ret = false;
+
+	if (!atomic_read(&bat_priv->distributed_arp_table))
+		goto out;
+
+	type = batadv_arp_get_type(bat_priv, skb, hdr_size);
+	if (type != ARPOP_REPLY)
+		goto out;
+
+	batadv_dbg_arp(bat_priv, skb, type, hdr_size,
+		       "Parsing incoming ARP REPLY");
+
+	hw_src = batadv_arp_hw_src(skb, hdr_size);
+	ip_src = batadv_arp_ip_src(skb, hdr_size);
+	hw_dst = batadv_arp_hw_dst(skb, hdr_size);
+	ip_dst = batadv_arp_ip_dst(skb, hdr_size);
+
+	/* Update our internal cache with both the IP addresses the node got
+	 * within the ARP reply
+	 */
+	batadv_dat_entry_add(bat_priv, ip_src, hw_src);
+	batadv_dat_entry_add(bat_priv, ip_dst, hw_dst);
+
+	/* if this REPLY is directed to a client of mine, let's deliver the
+	 * packet to the interface
+	 */
+	ret = !batadv_is_my_client(bat_priv, hw_dst);
+out:
+	/* if ret == false -> packet has to be delivered to the interface */
+	return ret;
+}
+
+/**
+ * batadv_dat_drop_broadcast_packet - check if an ARP request has to be dropped
+ * (because the node has already got the reply via DAT) or not
+ * @bat_priv: the bat priv with all the soft interface information
+ * @forw_packet: the broadcast packet
+ *
+ * Returns true if the node can drop the packet, false otherwise
+ */
+bool batadv_dat_drop_broadcast_packet(struct batadv_priv *bat_priv,
+				      struct batadv_forw_packet *forw_packet)
+{
+	uint16_t type;
+	__be32 ip_dst;
+	struct batadv_dat_entry *dat_entry = NULL;
+	bool ret = false;
+	const size_t bcast_len = sizeof(struct batadv_bcast_packet);
+
+	if (!atomic_read(&bat_priv->distributed_arp_table))
+		goto out;
+
+	/* If this packet is an ARP_REQUEST and the node already has the
+	 * information that it is going to ask, then the packet can be dropped
+	 */
+	if (forw_packet->num_packets)
+		goto out;
+
+	type = batadv_arp_get_type(bat_priv, forw_packet->skb, bcast_len);
+	if (type != ARPOP_REQUEST)
+		goto out;
+
+	ip_dst = batadv_arp_ip_dst(forw_packet->skb, bcast_len);
+	dat_entry = batadv_dat_entry_hash_find(bat_priv, ip_dst);
+	/* check if the node already got this entry */
+	if (!dat_entry) {
+		batadv_dbg(BATADV_DBG_DAT, bat_priv,
+			   "ARP Request for %pI4: fallback\n", &ip_dst);
+		goto out;
+	}
+
+	batadv_dbg(BATADV_DBG_DAT, bat_priv,
+		   "ARP Request for %pI4: fallback prevented\n", &ip_dst);
+	ret = true;
+
+out:
+	if (dat_entry)
+		batadv_dat_entry_free_ref(dat_entry);
+	return ret;
+}

net/batman-adv/distributed-arp-table.h

+/* Copyright (C) 2011-2012 B.A.T.M.A.N. contributors:
+ *
+ * Antonio Quartulli
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of version 2 of the GNU General Public
+ * License as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
+ * 02110-1301, USA
+ */
+
+#ifndef _NET_BATMAN_ADV_ARP_H_
+#define _NET_BATMAN_ADV_ARP_H_
+
+#ifdef CONFIG_BATMAN_ADV_DAT
+
+#include "types.h"
+#include "originator.h"
+
+#include <linux/if_arp.h>
+
+#define BATADV_DAT_ADDR_MAX ((batadv_dat_addr_t)~(batadv_dat_addr_t)0)
+
+bool batadv_dat_snoop_outgoing_arp_request(struct batadv_priv *bat_priv,
+					   struct sk_buff *skb);
+bool batadv_dat_snoop_incoming_arp_request(struct batadv_priv *bat_priv,
+					   struct sk_buff *skb, int hdr_size);
+void batadv_dat_snoop_outgoing_arp_reply(struct batadv_priv *bat_priv,
+					 struct sk_buff *skb);
+bool batadv_dat_snoop_incoming_arp_reply(struct batadv_priv *bat_priv,
+					 struct sk_buff *skb, int hdr_size);
+bool batadv_dat_drop_broadcast_packet(struct batadv_priv *bat_priv,
+				      struct batadv_forw_packet *forw_packet);
+
+/**
+ * batadv_dat_init_orig_node_addr - assign a DAT address to the orig_node
+ * @orig_node: the node to assign the DAT address to
+ */
+static inline void
+batadv_dat_init_orig_node_addr(struct batadv_orig_node *orig_node)
+{
+	uint32_t addr;
+
+	addr = batadv_choose_orig(orig_node->orig, BATADV_DAT_ADDR_MAX);
+	orig_node->dat_addr = (batadv_dat_addr_t)addr;
+}
+
+/**
+ * batadv_dat_init_own_addr - assign a DAT address to the node itself
+ * @bat_priv: the bat priv with all the soft interface information
+ * @primary_if: a pointer to the primary interface
+ */
+static inline void
+batadv_dat_init_own_addr(struct batadv_priv *bat_priv,
+			 struct batadv_hard_iface *primary_if)
+{
+	uint32_t addr;
+
+	addr = batadv_choose_orig(primary_if->net_dev->dev_addr,
+				  BATADV_DAT_ADDR_MAX);
+
+	bat_priv->dat.addr = (batadv_dat_addr_t)addr;
+}
+
+int batadv_dat_init(struct batadv_priv *bat_priv);
+void batadv_dat_free(struct batadv_priv *bat_priv);
+int batadv_dat_cache_seq_print_text(struct seq_file *seq, void *offset);
+
+/**
+ * batadv_dat_inc_counter - increment the correct DAT packet counter
+ * @bat_priv: the bat priv with all the soft interface information
+ * @subtype: the 4addr subtype of the packet to be counted
+ *
+ * Updates the ethtool statistics for the received packet if it is a DAT subtype
+ */
+static inline void batadv_dat_inc_counter(struct batadv_priv *bat_priv,
+					  uint8_t subtype)
+{
+	switch (subtype) {
+	case BATADV_P_DAT_DHT_GET:
+		batadv_inc_counter(bat_priv,
+				   BATADV_CNT_DAT_GET_RX);
+		break;
+	case BATADV_P_DAT_DHT_PUT:
+		batadv_inc_counter(bat_priv,
+				   BATADV_CNT_DAT_PUT_RX);
+		break;
+	}
+}
+
+#else
+
+static inline bool
+batadv_dat_snoop_outgoing_arp_request(struct batadv_priv *bat_priv,
+				      struct sk_buff *skb)
+{
+	return false;
+}
+
+static inline bool
+batadv_dat_snoop_incoming_arp_request(struct batadv_priv *bat_priv,
+				      struct sk_buff *skb, int hdr_size)
+{
+	return false;
+}
+
+static inline bool
+batadv_dat_snoop_outgoing_arp_reply(struct batadv_priv *bat_priv,
+				    struct sk_buff *skb)
+{
+	return false;
+}
+
+static inline bool
+batadv_dat_snoop_incoming_arp_reply(struct batadv_priv *bat_priv,
+				    struct sk_buff *skb, int hdr_size)
+{
+	return false;
+}
+
+static inline bool
+batadv_dat_drop_broadcast_packet(struct batadv_priv *bat_priv,
+				 struct batadv_forw_packet *forw_packet)
+{
+	return false;
+}
+
+static inline void
+batadv_dat_init_orig_node_addr(struct batadv_orig_node *orig_node)
+{
+}
+
+static inline void batadv_dat_init_own_addr(struct batadv_priv *bat_priv,
+					    struct batadv_hard_iface *iface)
+{
+}
+
+static inline void batadv_arp_change_timeout(struct net_device *soft_iface,
+					     const char *name)
+{
+}
+
+static inline int batadv_dat_init(struct batadv_priv *bat_priv)
+{
+	return 0;
+}
+
+static inline void batadv_dat_free(struct batadv_priv *bat_priv)
+{
+}
+
+static inline void batadv_dat_inc_counter(struct batadv_priv *bat_priv,
+					  uint8_t subtype)
+{
+}
+
+#endif /* CONFIG_BATMAN_ADV_DAT */
+
+#endif /* _NET_BATMAN_ADV_ARP_H_ */

net/batman-adv/hard-interface.c

@@ -18,6 +18,7 @@
  */
 
 #include "main.h"
+#include "distributed-arp-table.h"
 #include "hard-interface.h"
 #include "soft-interface.h"
 #include "send.h"
@@ -109,6 +110,8 @@ static void batadv_primary_if_update_addr(struct batadv_priv *bat_priv,
 	if (!primary_if)
 		goto out;
 
+	batadv_dat_init_own_addr(bat_priv, primary_if);
+
 	skb = bat_priv->vis.my_info->skb_packet;
 	vis_packet = (struct batadv_vis_packet *)skb->data;
 	memcpy(vis_packet->vis_orig, primary_if->net_dev->dev_addr, ETH_ALEN);

net/batman-adv/icmp_socket.c

@@ -177,13 +177,13 @@ static ssize_t batadv_socket_write(struct file *file, const char __user *buff,
 	if (len >= sizeof(struct batadv_icmp_packet_rr))
 		packet_len = sizeof(struct batadv_icmp_packet_rr);
 
-	skb = dev_alloc_skb(packet_len + ETH_HLEN);
+	skb = dev_alloc_skb(packet_len + ETH_HLEN + NET_IP_ALIGN);
 	if (!skb) {
 		len = -ENOMEM;
 		goto out;
 	}
 
-	skb_reserve(skb, ETH_HLEN);
+	skb_reserve(skb, ETH_HLEN + NET_IP_ALIGN);
 	icmp_packet = (struct batadv_icmp_packet_rr *)skb_put(skb, packet_len);
 
 	if (copy_from_user(icmp_packet, buff, packet_len)) {

net/batman-adv/main.c

@@ -29,6 +29,7 @@
 #include "hard-interface.h"
 #include "gateway_client.h"
 #include "bridge_loop_avoidance.h"
+#include "distributed-arp-table.h"
 #include "vis.h"
 #include "hash.h"
 #include "bat_algo.h"
@@ -128,6 +129,10 @@ int batadv_mesh_init(struct net_device *soft_iface)
 	if (ret < 0)
 		goto err;
 
+	ret = batadv_dat_init(bat_priv);
+	if (ret < 0)
+		goto err;
+
 	atomic_set(&bat_priv->gw.reselect, 0);
 	atomic_set(&bat_priv->mesh_state, BATADV_MESH_ACTIVE);
 
@@ -155,6 +160,8 @@ void batadv_mesh_free(struct net_device *soft_iface)
 
 	batadv_bla_free(bat_priv);
 
+	batadv_dat_free(bat_priv);
+
 	free_percpu(bat_priv->bat_counters);
 
 	atomic_set(&bat_priv->mesh_state, BATADV_MESH_INACTIVE);
@@ -300,6 +307,8 @@ static void batadv_recv_handler_init(void)
 
 	/* batman icmp packet */
 	batadv_rx_handler[BATADV_ICMP] = batadv_recv_icmp_packet;
+	/* unicast with 4 addresses packet */
+	batadv_rx_handler[BATADV_UNICAST_4ADDR] = batadv_recv_unicast_packet;
 	/* unicast packet */
 	batadv_rx_handler[BATADV_UNICAST] = batadv_recv_unicast_packet;
 	/* fragmented unicast packet */

net/batman-adv/main.h

@@ -44,6 +44,7 @@
 #define BATADV_TT_LOCAL_TIMEOUT 3600000 /* in milliseconds */
 #define BATADV_TT_CLIENT_ROAM_TIMEOUT 600000 /* in milliseconds */
 #define BATADV_TT_CLIENT_TEMP_TIMEOUT 600000 /* in milliseconds */
+#define BATADV_DAT_ENTRY_TIMEOUT (5*60000) /* 5 mins in milliseconds */
 /* sliding packet range of received originator messages in sequence numbers
  * (should be a multiple of our word size)
  */
@@ -73,6 +74,11 @@
 
 #define BATADV_LOG_BUF_LEN 8192	  /* has to be a power of 2 */
 
+/* msecs after which an ARP_REQUEST is sent in broadcast as fallback */
+#define ARP_REQ_DELAY 250
+/* numbers of originator to contact for any PUT/GET DHT operation */
+#define BATADV_DAT_CANDIDATES_NUM 3
+
 #define BATADV_VIS_INTERVAL 5000	/* 5 seconds */
 
 /* how much worse secondary interfaces may be to be considered as bonding
@@ -117,6 +123,9 @@ enum batadv_uev_type {
 
 #define BATADV_GW_THRESHOLD	50
 
+#define BATADV_DAT_CANDIDATE_NOT_FOUND	0
+#define BATADV_DAT_CANDIDATE_ORIG	1
+
 /* Debug Messages */
 #ifdef pr_fmt
 #undef pr_fmt
@@ -171,6 +180,7 @@ int batadv_algo_seq_print_text(struct seq_file *seq, void *offset);
  * @BATADV_DBG_ROUTES: route added / changed / deleted
  * @BATADV_DBG_TT: translation table messages
  * @BATADV_DBG_BLA: bridge loop avoidance messages
+ * @BATADV_DBG_DAT: ARP snooping and DAT related messages
  * @BATADV_DBG_ALL: the union of all the above log levels
  */
 enum batadv_dbg_level {
@@ -178,7 +188,8 @@ enum batadv_dbg_level {
 	BATADV_DBG_ROUTES = BIT(1),
 	BATADV_DBG_TT	  = BIT(2),
 	BATADV_DBG_BLA    = BIT(3),
-	BATADV_DBG_ALL    = 15,
+	BATADV_DBG_DAT    = BIT(4),
+	BATADV_DBG_ALL    = 31,
 };
 
 #ifdef CONFIG_BATMAN_ADV_DEBUG

net/batman-adv/originator.c

@@ -18,6 +18,7 @@
  */
 
 #include "main.h"
+#include "distributed-arp-table.h"
 #include "originator.h"
 #include "hash.h"
 #include "translation-table.h"
@@ -223,6 +224,7 @@ struct batadv_orig_node *batadv_get_orig_node(struct batadv_priv *bat_priv,
 	orig_node->tt_poss_change = false;
 	orig_node->bat_priv = bat_priv;
 	memcpy(orig_node->orig, addr, ETH_ALEN);
+	batadv_dat_init_orig_node_addr(orig_node);
 	orig_node->router = NULL;
 	orig_node->tt_crc = 0;
 	atomic_set(&orig_node->last_ttvn, 0);

net/batman-adv/packet.h

@@ -31,6 +31,21 @@ enum batadv_packettype {
 	BATADV_UNICAST_FRAG	= 0x06,
 	BATADV_TT_QUERY		= 0x07,
 	BATADV_ROAM_ADV		= 0x08,
+	BATADV_UNICAST_4ADDR	= 0x09,
+};
+
+/**
+ * enum batadv_subtype - packet subtype for unicast4addr
+ * @BATADV_P_DATA: user payload
+ * @BATADV_P_DAT_DHT_GET: DHT request message
+ * @BATADV_P_DAT_DHT_PUT: DHT store message
+ * @BATADV_P_DAT_CACHE_REPLY: ARP reply generated by DAT
+ */
+enum batadv_subtype {
+	BATADV_P_DATA			= 0x01,
+	BATADV_P_DAT_DHT_GET		= 0x02,
+	BATADV_P_DAT_DHT_PUT		= 0x03,
+	BATADV_P_DAT_CACHE_REPLY	= 0x04,
 };
 
 /* this file is included by batctl which needs these defines */
@@ -106,13 +121,16 @@ struct batadv_bla_claim_dst {
 	uint8_t magic[3];	/* FF:43:05 */
 	uint8_t type;		/* bla_claimframe */
 	__be16 group;		/* group id */
-} __packed;
+};
 
 struct batadv_header {
 	uint8_t  packet_type;
 	uint8_t  version;  /* batman version field */
 	uint8_t  ttl;
-} __packed;
+	/* the parent struct has to add a byte after the header to make
+	 * everything 4 bytes aligned again
+	 */
+};
 
 struct batadv_ogm_packet {
 	struct batadv_header header;
@@ -137,7 +155,7 @@ struct batadv_icmp_packet {
 	__be16   seqno;
 	uint8_t  uid;
 	uint8_t  reserved;
-} __packed;
+};
 
 #define BATADV_RR_LEN 16
 
@@ -153,13 +171,32 @@ struct batadv_icmp_packet_rr {
 	uint8_t  uid;
 	uint8_t  rr_cur;
 	uint8_t  rr[BATADV_RR_LEN][ETH_ALEN];
-} __packed;
+};
 
 struct batadv_unicast_packet {
 	struct batadv_header header;
 	uint8_t  ttvn; /* destination translation table version number */
 	uint8_t  dest[ETH_ALEN];
-} __packed;
+	/* "4 bytes boundary + 2 bytes" long to make the payload after the
+	 * following ethernet header again 4 bytes boundary aligned
+	 */
+};
+
+/**
+ * struct batadv_unicast_4addr_packet - extended unicast packet
+ * @u: common unicast packet header
+ * @src: address of the source
+ * @subtype: packet subtype
+ */
+struct batadv_unicast_4addr_packet {
+	struct batadv_unicast_packet u;
+	uint8_t src[ETH_ALEN];
+	uint8_t subtype;
+	uint8_t reserved;
+	/* "4 bytes boundary + 2 bytes" long to make the payload after the
+	 * following ethernet header again 4 bytes boundary aligned
+	 */
+};
 
 struct batadv_unicast_frag_packet {
 	struct batadv_header header;
@@ -176,6 +213,9 @@ struct batadv_bcast_packet {
 	uint8_t  reserved;
 	__be32   seqno;
 	uint8_t  orig[ETH_ALEN];
+	/* "4 bytes boundary + 2 bytes" long to make the payload after the
+	 * following ethernet header again 4 bytes boundary aligned
+	 */
 } __packed;
 
 struct batadv_vis_packet {
@@ -187,7 +227,7 @@ struct batadv_vis_packet {
 	uint8_t  vis_orig[ETH_ALEN];	/* originator reporting its neighbors */
 	uint8_t  target_orig[ETH_ALEN]; /* who should receive this packet */
 	uint8_t  sender_orig[ETH_ALEN]; /* who sent or forwarded this packet */
-} __packed;
+};
 
 struct batadv_tt_query_packet {
 	struct batadv_header header;

net/batman-adv/routing.c

@@ -28,6 +28,7 @@
 #include "vis.h"
 #include "unicast.h"
 #include "bridge_loop_avoidance.h"
+#include "distributed-arp-table.h"
 
 static int batadv_route_unicast_packet(struct sk_buff *skb,
 				       struct batadv_hard_iface *recv_if);
@@ -984,7 +985,19 @@ int batadv_recv_unicast_packet(struct sk_buff *skb,
 {
 	struct batadv_priv *bat_priv = netdev_priv(recv_if->soft_iface);
 	struct batadv_unicast_packet *unicast_packet;
+	struct batadv_unicast_4addr_packet *unicast_4addr_packet;
+	uint8_t *orig_addr;
+	struct batadv_orig_node *orig_node = NULL;
 	int hdr_size = sizeof(*unicast_packet);
+	bool is4addr;
+
+	unicast_packet = (struct batadv_unicast_packet *)skb->data;
+	unicast_4addr_packet = (struct batadv_unicast_4addr_packet *)skb->data;
+
+	is4addr = unicast_packet->header.packet_type == BATADV_UNICAST_4ADDR;
+	/* the caller function should have already pulled 2 bytes */
+	if (is4addr)
+		hdr_size = sizeof(*unicast_4addr_packet);
 
 	if (batadv_check_unicast_packet(skb, hdr_size) < 0)
 		return NET_RX_DROP;
@@ -992,12 +1005,28 @@ int batadv_recv_unicast_packet(struct sk_buff *skb,
 	if (!batadv_check_unicast_ttvn(bat_priv, skb))
 		return NET_RX_DROP;
 
-	unicast_packet = (struct batadv_unicast_packet *)skb->data;
-
 	/* packet for me */
 	if (batadv_is_my_mac(unicast_packet->dest)) {
+		if (is4addr) {
+			batadv_dat_inc_counter(bat_priv,
+					       unicast_4addr_packet->subtype);
+			orig_addr = unicast_4addr_packet->src;
+			orig_node = batadv_orig_hash_find(bat_priv, orig_addr);
+		}
+
+		if (batadv_dat_snoop_incoming_arp_request(bat_priv, skb,
+							  hdr_size))
+			goto rx_success;
+		if (batadv_dat_snoop_incoming_arp_reply(bat_priv, skb,
+							hdr_size))
+			goto rx_success;
+
 		batadv_interface_rx(recv_if->soft_iface, skb, recv_if, hdr_size,
-				    NULL);
+				    orig_node);
+
+rx_success:
+		if (orig_node)
+			batadv_orig_node_free_ref(orig_node);
 
 		return NET_RX_SUCCESS;
 	}
@@ -1034,8 +1063,17 @@ int batadv_recv_ucast_frag_packet(struct sk_buff *skb,
 		if (!new_skb)
 			return NET_RX_SUCCESS;
 
+		if (batadv_dat_snoop_incoming_arp_request(bat_priv, new_skb,
+							  hdr_size))
+			goto rx_success;
+		if (batadv_dat_snoop_incoming_arp_reply(bat_priv, new_skb,
+							hdr_size))
+			goto rx_success;
+
 		batadv_interface_rx(recv_if->soft_iface, new_skb, recv_if,
 				    sizeof(struct batadv_unicast_packet), NULL);
+
+rx_success:
 		return NET_RX_SUCCESS;
 	}
 
@@ -1127,9 +1165,16 @@ int batadv_recv_bcast_packet(struct sk_buff *skb,
 	if (batadv_bla_is_backbone_gw(skb, orig_node, hdr_size))
 		goto out;
 
+	if (batadv_dat_snoop_incoming_arp_request(bat_priv, skb, hdr_size))
+		goto rx_success;
+	if (batadv_dat_snoop_incoming_arp_reply(bat_priv, skb, hdr_size))
+		goto rx_success;
+
 	/* broadcast for me */
 	batadv_interface_rx(recv_if->soft_iface, skb, recv_if, hdr_size,
 			    orig_node);
+
+rx_success:
 	ret = NET_RX_SUCCESS;
 	goto out;
 

net/batman-adv/send.c

@@ -18,6 +18,7 @@
  */
 
 #include "main.h"
+#include "distributed-arp-table.h"
 #include "send.h"
 #include "routing.h"
 #include "translation-table.h"
@@ -209,6 +210,9 @@ static void batadv_send_outstanding_bcast_packet(struct work_struct *work)
 	if (atomic_read(&bat_priv->mesh_state) == BATADV_MESH_DEACTIVATING)
 		goto out;
 
+	if (batadv_dat_drop_broadcast_packet(bat_priv, forw_packet))
+		goto out;
+
 	/* rebroadcast packet */
 	rcu_read_lock();
 	list_for_each_entry_rcu(hard_iface, &batadv_hardif_list, list) {

net/batman-adv/soft-interface.c

@@ -20,6 +20,7 @@
 #include "main.h"
 #include "soft-interface.h"
 #include "hard-interface.h"
+#include "distributed-arp-table.h"
 #include "routing.h"
 #include "send.h"
 #include "debugfs.h"
@@ -155,6 +156,7 @@ static int batadv_interface_tx(struct sk_buff *skb,
 	short vid __maybe_unused = -1;
 	bool do_bcast = false;
 	uint32_t seqno;
+	unsigned long brd_delay = 1;
 
 	if (atomic_read(&bat_priv->mesh_state) != BATADV_MESH_ACTIVE)
 		goto dropped;
@@ -224,6 +226,13 @@ static int batadv_interface_tx(struct sk_buff *skb,
 		if (!primary_if)
 			goto dropped;
 
+		/* in case of ARP request, we do not immediately broadcasti the
+		 * packet, instead we first wait for DAT to try to retrieve the
+		 * correct ARP entry
+		 */
+		if (batadv_dat_snoop_outgoing_arp_request(bat_priv, skb))
+			brd_delay = msecs_to_jiffies(ARP_REQ_DELAY);
+
 		if (batadv_skb_head_push(skb, sizeof(*bcast_packet)) < 0)
 			goto dropped;
 
@@ -245,7 +254,7 @@ static int batadv_interface_tx(struct sk_buff *skb,
 		seqno = atomic_inc_return(&bat_priv->bcast_seqno);
 		bcast_packet->seqno = htonl(seqno);
 
-		batadv_add_bcast_packet_to_list(bat_priv, skb, 1);
+		batadv_add_bcast_packet_to_list(bat_priv, skb, brd_delay);
 
 		/* a copy is stored in the bcast list, therefore removing
 		 * the original skb.
@@ -260,7 +269,12 @@ static int batadv_interface_tx(struct sk_buff *skb,
 				goto dropped;
 		}
 
-		ret = batadv_unicast_send_skb(skb, bat_priv);
+		if (batadv_dat_snoop_outgoing_arp_request(bat_priv, skb))
+			goto dropped;
+
+		batadv_dat_snoop_outgoing_arp_reply(bat_priv, skb);
+
+		ret = batadv_unicast_send_skb(bat_priv, skb);
 		if (ret != 0)
 			goto dropped_freed;
 	}
@@ -466,6 +480,9 @@ struct net_device *batadv_softif_create(const char *name)
 	atomic_set(&bat_priv->aggregated_ogms, 1);
 	atomic_set(&bat_priv->bonding, 0);
 	atomic_set(&bat_priv->bridge_loop_avoidance, 0);
+#ifdef CONFIG_BATMAN_ADV_DAT
+	atomic_set(&bat_priv->distributed_arp_table, 1);
+#endif
 	atomic_set(&bat_priv->ap_isolation, 0);
 	atomic_set(&bat_priv->vis_mode, BATADV_VIS_TYPE_CLIENT_UPDATE);
 	atomic_set(&bat_priv->gw_mode, BATADV_GW_MODE_OFF);
@@ -608,6 +625,13 @@ static const struct {
 	{ "tt_response_rx" },
 	{ "tt_roam_adv_tx" },
 	{ "tt_roam_adv_rx" },
+#ifdef CONFIG_BATMAN_ADV_DAT
+	{ "dat_get_tx" },
+	{ "dat_get_rx" },
+	{ "dat_put_tx" },
+	{ "dat_put_rx" },
+	{ "dat_cached_reply_tx" },
+#endif
 };
 
 static void batadv_get_strings(struct net_device *dev, uint32_t stringset,

net/batman-adv/sysfs.c

@@ -20,6 +20,7 @@
 #include "main.h"
 #include "sysfs.h"
 #include "translation-table.h"
+#include "distributed-arp-table.h"
 #include "originator.h"
 #include "hard-interface.h"
 #include "gateway_common.h"
@@ -420,6 +421,9 @@ BATADV_ATTR_SIF_BOOL(bonding, S_IRUGO | S_IWUSR, NULL);
 #ifdef CONFIG_BATMAN_ADV_BLA
 BATADV_ATTR_SIF_BOOL(bridge_loop_avoidance, S_IRUGO | S_IWUSR, NULL);
 #endif
+#ifdef CONFIG_BATMAN_ADV_DAT
+BATADV_ATTR_SIF_BOOL(distributed_arp_table, S_IRUGO | S_IWUSR, NULL);
+#endif
 BATADV_ATTR_SIF_BOOL(fragmentation, S_IRUGO | S_IWUSR, batadv_update_min_mtu);
 BATADV_ATTR_SIF_BOOL(ap_isolation, S_IRUGO | S_IWUSR, NULL);
 static BATADV_ATTR(vis_mode, S_IRUGO | S_IWUSR, batadv_show_vis_mode,
@@ -444,6 +448,9 @@ static struct batadv_attribute *batadv_mesh_attrs[] = {
 	&batadv_attr_bonding,
 #ifdef CONFIG_BATMAN_ADV_BLA
 	&batadv_attr_bridge_loop_avoidance,
+#endif
+#ifdef CONFIG_BATMAN_ADV_DAT
+	&batadv_attr_distributed_arp_table,
 #endif
 	&batadv_attr_fragmentation,
 	&batadv_attr_ap_isolation,

net/batman-adv/translation-table.c

@@ -1472,11 +1472,11 @@ batadv_tt_response_fill_table(uint16_t tt_len, uint8_t ttvn,
 	tt_tot = tt_len / sizeof(struct batadv_tt_change);
 
 	len = tt_query_size + tt_len;
-	skb = dev_alloc_skb(len + ETH_HLEN);
+	skb = dev_alloc_skb(len + ETH_HLEN + NET_IP_ALIGN);
 	if (!skb)
 		goto out;
 
-	skb_reserve(skb, ETH_HLEN);
+	skb_reserve(skb, ETH_HLEN + NET_IP_ALIGN);
 	tt_response = (struct batadv_tt_query_packet *)skb_put(skb, len);
 	tt_response->ttvn = ttvn;
 
@@ -1538,11 +1538,11 @@ static int batadv_send_tt_request(struct batadv_priv *bat_priv,
 	if (!tt_req_node)
 		goto out;
 
-	skb = dev_alloc_skb(sizeof(*tt_request) + ETH_HLEN);
+	skb = dev_alloc_skb(sizeof(*tt_request) + ETH_HLEN + NET_IP_ALIGN);
 	if (!skb)
 		goto out;
 
-	skb_reserve(skb, ETH_HLEN);
+	skb_reserve(skb, ETH_HLEN + NET_IP_ALIGN);
 
 	tt_req_len = sizeof(*tt_request);
 	tt_request = (struct batadv_tt_query_packet *)skb_put(skb, tt_req_len);
@@ -1653,11 +1653,11 @@ batadv_send_other_tt_response(struct batadv_priv *bat_priv,
 		tt_tot = tt_len / sizeof(struct batadv_tt_change);
 
 		len = sizeof(*tt_response) + tt_len;
-		skb = dev_alloc_skb(len + ETH_HLEN);
+		skb = dev_alloc_skb(len + ETH_HLEN + NET_IP_ALIGN);
 		if (!skb)
 			goto unlock;
 
-		skb_reserve(skb, ETH_HLEN);
+		skb_reserve(skb, ETH_HLEN + NET_IP_ALIGN);
 		packet_pos = skb_put(skb, len);
 		tt_response = (struct batadv_tt_query_packet *)packet_pos;
 		tt_response->ttvn = req_ttvn;
@@ -1780,11 +1780,11 @@ batadv_send_my_tt_response(struct batadv_priv *bat_priv,
 		tt_tot = tt_len / sizeof(struct batadv_tt_change);
 
 		len = sizeof(*tt_response) + tt_len;
-		skb = dev_alloc_skb(len + ETH_HLEN);
+		skb = dev_alloc_skb(len + ETH_HLEN + NET_IP_ALIGN);
 		if (!skb)
 			goto unlock;
 
-		skb_reserve(skb, ETH_HLEN);
+		skb_reserve(skb, ETH_HLEN + NET_IP_ALIGN);
 		packet_pos = skb_put(skb, len);
 		tt_response = (struct batadv_tt_query_packet *)packet_pos;
 		tt_response->ttvn = req_ttvn;
@@ -2118,11 +2118,11 @@ static void batadv_send_roam_adv(struct batadv_priv *bat_priv, uint8_t *client,
 	if (!batadv_tt_check_roam_count(bat_priv, client))
 		goto out;
 
-	skb = dev_alloc_skb(sizeof(*roam_adv_packet) + ETH_HLEN);
+	skb = dev_alloc_skb(sizeof(*roam_adv_packet) + ETH_HLEN + NET_IP_ALIGN);
 	if (!skb)
 		goto out;
 
-	skb_reserve(skb, ETH_HLEN);
+	skb_reserve(skb, ETH_HLEN + NET_IP_ALIGN);
 
 	roam_adv_packet = (struct batadv_roam_adv_packet *)skb_put(skb, len);
 

net/batman-adv/types.h

@@ -28,6 +28,17 @@
 	(ETH_HLEN + max(sizeof(struct batadv_unicast_packet), \
 			sizeof(struct batadv_bcast_packet)))
 
+#ifdef CONFIG_BATMAN_ADV_DAT
+
+/* batadv_dat_addr_t is the type used for all DHT addresses. If it is changed,
+ * BATADV_DAT_ADDR_MAX is changed as well.
+ *
+ * *Please be careful: batadv_dat_addr_t must be UNSIGNED*
+ */
+#define batadv_dat_addr_t uint16_t
+
+#endif /* CONFIG_BATMAN_ADV_DAT */
+
 /**
  * struct batadv_hard_iface_bat_iv - per hard interface B.A.T.M.A.N. IV data
  * @ogm_buff: buffer holding the OGM packet
@@ -73,6 +84,9 @@ struct batadv_orig_node {
 	uint8_t orig[ETH_ALEN];
 	uint8_t primary_addr[ETH_ALEN];
 	struct batadv_neigh_node __rcu *router; /* rcu protected pointer */
+#ifdef CONFIG_BATMAN_ADV_DAT
+	batadv_dat_addr_t dat_addr;
+#endif
 	unsigned long *bcast_own;
 	uint8_t *bcast_own_sum;
 	unsigned long last_seen;
@@ -172,6 +186,13 @@ enum batadv_counters {
 	BATADV_CNT_TT_RESPONSE_RX,
 	BATADV_CNT_TT_ROAM_ADV_TX,
 	BATADV_CNT_TT_ROAM_ADV_RX,
+#ifdef CONFIG_BATMAN_ADV_DAT
+	BATADV_CNT_DAT_GET_TX,
+	BATADV_CNT_DAT_GET_RX,
+	BATADV_CNT_DAT_PUT_TX,
+	BATADV_CNT_DAT_PUT_RX,
+	BATADV_CNT_DAT_CACHED_REPLY_TX,
+#endif
 	BATADV_CNT_NUM,
 };
 
@@ -238,6 +259,20 @@ struct batadv_priv_vis {
 	struct batadv_vis_info *my_info;
 };
 
+/**
+ * struct batadv_priv_dat - per mesh interface DAT private data
+ * @addr: node DAT address
+ * @hash: hashtable representing the local ARP cache
+ * @work: work queue callback item for cache purging
+ */
+#ifdef CONFIG_BATMAN_ADV_DAT
+struct batadv_priv_dat {
+	batadv_dat_addr_t addr;
+	struct batadv_hashtable *hash;
+	struct delayed_work work;
+};
+#endif
+
 struct batadv_priv {
 	atomic_t mesh_state;
 	struct net_device_stats stats;
@@ -247,6 +282,9 @@ struct batadv_priv {
 	atomic_t fragmentation;		/* boolean */
 	atomic_t ap_isolation;		/* boolean */
 	atomic_t bridge_loop_avoidance;	/* boolean */
+#ifdef CONFIG_BATMAN_ADV_DAT
+	atomic_t distributed_arp_table;	/* boolean */
+#endif
 	atomic_t vis_mode;		/* VIS_TYPE_* */
 	atomic_t gw_mode;		/* GW_MODE_* */
 	atomic_t gw_sel_class;		/* uint */
@@ -275,6 +313,9 @@ struct batadv_priv {
 	struct batadv_priv_gw gw;
 	struct batadv_priv_tt tt;
 	struct batadv_priv_vis vis;
+#ifdef CONFIG_BATMAN_ADV_DAT
+	struct batadv_priv_dat dat;
+#endif
 };
 
 struct batadv_socket_client {
@@ -447,4 +488,36 @@ struct batadv_algo_ops {
 	void (*bat_ogm_emit)(struct batadv_forw_packet *forw_packet);
 };
 
+/**
+ * struct batadv_dat_entry - it is a single entry of batman-adv ARP backend. It
+ * is used to stored ARP entries needed for the global DAT cache
+ * @ip: the IPv4 corresponding to this DAT/ARP entry
+ * @mac_addr: the MAC address associated to the stored IPv4
+ * @last_update: time in jiffies when this entry was refreshed last time
+ * @hash_entry: hlist node for batadv_priv_dat::hash
+ * @refcount: number of contexts the object is used
+ * @rcu: struct used for freeing in an RCU-safe manner
+ */
+struct batadv_dat_entry {
+	__be32 ip;
+	uint8_t mac_addr[ETH_ALEN];
+	unsigned long last_update;
+	struct hlist_node hash_entry;
+	atomic_t refcount;
+	struct rcu_head rcu;
+};
+
+/**
+ * struct batadv_dat_candidate - candidate destination for DAT operations
+ * @type: the type of the selected candidate. It can one of the following:
+ *	  - BATADV_DAT_CANDIDATE_NOT_FOUND
+ *	  - BATADV_DAT_CANDIDATE_ORIG
+ * @orig_node: if type is BATADV_DAT_CANDIDATE_ORIG this field points to the
+ *	       corresponding originator node structure
+ */
+struct batadv_dat_candidate {
+	int type;
+	struct batadv_orig_node *orig_node;
+};
+
 #endif /* _NET_BATMAN_ADV_TYPES_H_ */

net/batman-adv/unicast.c

@@ -291,7 +291,111 @@ int batadv_frag_send_skb(struct sk_buff *skb, struct batadv_priv *bat_priv,
 	return ret;
 }
 
-int batadv_unicast_send_skb(struct sk_buff *skb, struct batadv_priv *bat_priv)
+/**
+ * batadv_unicast_push_and_fill_skb - extends the buffer and initializes the
+ * common fields for unicast packets
+ * @skb: packet
+ * @hdr_size: amount of bytes to push at the beginning of the skb
+ * @orig_node: the destination node
+ *
+ * Returns false if the buffer extension was not possible or true otherwise
+ */
+static bool batadv_unicast_push_and_fill_skb(struct sk_buff *skb, int hdr_size,
+					     struct batadv_orig_node *orig_node)
+{
+	struct batadv_unicast_packet *unicast_packet;
+	uint8_t ttvn = (uint8_t)atomic_read(&orig_node->last_ttvn);
+
+	if (batadv_skb_head_push(skb, hdr_size) < 0)
+		return false;
+
+	unicast_packet = (struct batadv_unicast_packet *)skb->data;
+	unicast_packet->header.version = BATADV_COMPAT_VERSION;
+	/* batman packet type: unicast */
+	unicast_packet->header.packet_type = BATADV_UNICAST;
+	/* set unicast ttl */
+	unicast_packet->header.ttl = BATADV_TTL;
+	/* copy the destination for faster routing */
+	memcpy(unicast_packet->dest, orig_node->orig, ETH_ALEN);
+	/* set the destination tt version number */
+	unicast_packet->ttvn = ttvn;
+
+	return true;
+}
+
+/**
+ * batadv_unicast_prepare_skb - encapsulate an skb with a unicast header
+ * @skb: the skb containing the payload to encapsulate
+ * @orig_node: the destination node
+ *
+ * Returns false if the payload could not be encapsulated or true otherwise
+ */
+static bool batadv_unicast_prepare_skb(struct sk_buff *skb,
+				       struct batadv_orig_node *orig_node)
+{
+	size_t uni_size = sizeof(struct batadv_unicast_packet);
+	return batadv_unicast_push_and_fill_skb(skb, uni_size, orig_node);
+}
+
+/**
+ * batadv_unicast_4addr_prepare_skb - encapsulate an skb with a unicast4addr
+ * header
+ * @bat_priv: the bat priv with all the soft interface information
+ * @skb: the skb containing the payload to encapsulate
+ * @orig_node: the destination node
+ * @packet_subtype: the batman 4addr packet subtype to use
+ *
+ * Returns false if the payload could not be encapsulated or true otherwise
+ */
+bool batadv_unicast_4addr_prepare_skb(struct batadv_priv *bat_priv,
+				      struct sk_buff *skb,
+				      struct batadv_orig_node *orig,
+				      int packet_subtype)
+{
+	struct batadv_hard_iface *primary_if;
+	struct batadv_unicast_4addr_packet *unicast_4addr_packet;
+	bool ret = false;
+
+	primary_if = batadv_primary_if_get_selected(bat_priv);
+	if (!primary_if)
+		goto out;
+
+	/* pull the header space and fill the unicast_packet substructure.
+	 * We can do that because the first member of the unicast_4addr_packet
+	 * is of type struct unicast_packet
+	 */
+	if (!batadv_unicast_push_and_fill_skb(skb,
+					      sizeof(*unicast_4addr_packet),
+					      orig))
+		goto out;
+
+	unicast_4addr_packet = (struct batadv_unicast_4addr_packet *)skb->data;
+	unicast_4addr_packet->u.header.packet_type = BATADV_UNICAST_4ADDR;
+	memcpy(unicast_4addr_packet->src, primary_if->net_dev->dev_addr,
+	       ETH_ALEN);
+	unicast_4addr_packet->subtype = packet_subtype;
+	unicast_4addr_packet->reserved = 0;
+
+	ret = true;
+out:
+	if (primary_if)
+		batadv_hardif_free_ref(primary_if);
+	return ret;
+}
+
+/**
+ * batadv_unicast_generic_send_skb - send an skb as unicast
+ * @bat_priv: the bat priv with all the soft interface information
+ * @skb: payload to send
+ * @packet_type: the batman unicast packet type to use
+ * @packet_subtype: the batman packet subtype. It is ignored if packet_type is
+ *		    not BATADV_UNICAT_4ADDR
+ *
+ * Returns 1 in case of error or 0 otherwise
+ */
+int batadv_unicast_generic_send_skb(struct batadv_priv *bat_priv,
+				    struct sk_buff *skb, int packet_type,
+				    int packet_subtype)
 {
 	struct ethhdr *ethhdr = (struct ethhdr *)skb->data;
 	struct batadv_unicast_packet *unicast_packet;
@@ -324,21 +428,23 @@ int batadv_unicast_send_skb(struct sk_buff *skb, struct batadv_priv *bat_priv)
 	if (!neigh_node)
 		goto out;
 
-	if (batadv_skb_head_push(skb, sizeof(*unicast_packet)) < 0)
+	switch (packet_type) {
+	case BATADV_UNICAST:
+		batadv_unicast_prepare_skb(skb, orig_node);
+		break;
+	case BATADV_UNICAST_4ADDR:
+		batadv_unicast_4addr_prepare_skb(bat_priv, skb, orig_node,
+						 packet_subtype);
+		break;
+	default:
+		/* this function supports UNICAST and UNICAST_4ADDR only. It
+		 * should never be invoked with any other packet type
+		 */
 		goto out;
+	}
 
 	unicast_packet = (struct batadv_unicast_packet *)skb->data;
 
-	unicast_packet->header.version = BATADV_COMPAT_VERSION;
-	/* batman packet type: unicast */
-	unicast_packet->header.packet_type = BATADV_UNICAST;
-	/* set unicast ttl */
-	unicast_packet->header.ttl = BATADV_TTL;
-	/* copy the destination for faster routing */
-	memcpy(unicast_packet->dest, orig_node->orig, ETH_ALEN);
-	/* set the destination tt version number */
-	unicast_packet->ttvn = (uint8_t)atomic_read(&orig_node->last_ttvn);
-
 	/* inform the destination node that we are still missing a correct route
 	 * for this client. The destination will receive this packet and will
 	 * try to reroute it because the ttvn contained in the header is less
@@ -348,7 +454,9 @@ int batadv_unicast_send_skb(struct sk_buff *skb, struct batadv_priv *bat_priv)
 		unicast_packet->ttvn = unicast_packet->ttvn - 1;
 
 	dev_mtu = neigh_node->if_incoming->net_dev->mtu;
-	if (atomic_read(&bat_priv->fragmentation) &&
+	/* fragmentation mechanism only works for UNICAST (now) */
+	if (packet_type == BATADV_UNICAST &&
+	    atomic_read(&bat_priv->fragmentation) &&
 	    data_len + sizeof(*unicast_packet) > dev_mtu) {
 		/* send frag skb decreases ttl */
 		unicast_packet->header.ttl++;
@@ -360,7 +468,6 @@ int batadv_unicast_send_skb(struct sk_buff *skb, struct batadv_priv *bat_priv)
 
 	batadv_send_skb_packet(skb, neigh_node->if_incoming, neigh_node->addr);
 	ret = 0;
-	goto out;
 
 out:
 	if (neigh_node)

net/batman-adv/unicast.h

@@ -29,10 +29,44 @@ int batadv_frag_reassemble_skb(struct sk_buff *skb,
 			       struct batadv_priv *bat_priv,
 			       struct sk_buff **new_skb);
 void batadv_frag_list_free(struct list_head *head);
-int batadv_unicast_send_skb(struct sk_buff *skb, struct batadv_priv *bat_priv);
 int batadv_frag_send_skb(struct sk_buff *skb, struct batadv_priv *bat_priv,
 			 struct batadv_hard_iface *hard_iface,
 			 const uint8_t dstaddr[]);
+bool batadv_unicast_4addr_prepare_skb(struct batadv_priv *bat_priv,
+				      struct sk_buff *skb,
+				      struct batadv_orig_node *orig_node,
+				      int packet_subtype);
+int batadv_unicast_generic_send_skb(struct batadv_priv *bat_priv,
+				    struct sk_buff *skb, int packet_type,
+				    int packet_subtype);
+
+
+/**
+ * batadv_unicast_send_skb - send the skb encapsulated in a unicast packet
+ * @bat_priv: the bat priv with all the soft interface information
+ * @skb: the payload to send
+ */
+static inline int batadv_unicast_send_skb(struct batadv_priv *bat_priv,
+					  struct sk_buff *skb)
+{
+	return batadv_unicast_generic_send_skb(bat_priv, skb, BATADV_UNICAST,
+					       0);
+}
+
+/**
+ * batadv_unicast_send_skb - send the skb encapsulated in a unicast4addr packet
+ * @bat_priv: the bat priv with all the soft interface information
+ * @skb: the payload to send
+ * @packet_subtype: the batman 4addr packet subtype to use
+ */
+static inline int batadv_unicast_4addr_send_skb(struct batadv_priv *bat_priv,
+						struct sk_buff *skb,
+						int packet_subtype)
+{
+	return batadv_unicast_generic_send_skb(bat_priv, skb,
+					       BATADV_UNICAST_4ADDR,
+					       packet_subtype);
+}
 
 static inline int batadv_frag_can_reassemble(const struct sk_buff *skb, int mtu)
 {

net/batman-adv/vis.c

@@ -396,12 +396,12 @@ batadv_add_packet(struct batadv_priv *bat_priv,
 		return NULL;
 
 	len = sizeof(*packet) + vis_info_len;
-	info->skb_packet = dev_alloc_skb(len + ETH_HLEN);
+	info->skb_packet = dev_alloc_skb(len + ETH_HLEN + NET_IP_ALIGN);
 	if (!info->skb_packet) {
 		kfree(info);
 		return NULL;
 	}
-	skb_reserve(info->skb_packet, ETH_HLEN);
+	skb_reserve(info->skb_packet, ETH_HLEN + NET_IP_ALIGN);
 	packet = (struct batadv_vis_packet *)skb_put(info->skb_packet, len);
 
 	kref_init(&info->refcount);
@@ -873,12 +873,13 @@ int batadv_vis_init(struct batadv_priv *bat_priv)
 	if (!bat_priv->vis.my_info)
 		goto err;
 
-	len = sizeof(*packet) + BATADV_MAX_VIS_PACKET_SIZE + ETH_HLEN;
+	len = sizeof(*packet) + BATADV_MAX_VIS_PACKET_SIZE;
+	len += ETH_HLEN + NET_IP_ALIGN;
 	bat_priv->vis.my_info->skb_packet = dev_alloc_skb(len);
 	if (!bat_priv->vis.my_info->skb_packet)
 		goto free_info;
 
-	skb_reserve(bat_priv->vis.my_info->skb_packet, ETH_HLEN);
+	skb_reserve(bat_priv->vis.my_info->skb_packet, ETH_HLEN + NET_IP_ALIGN);
 	tmp_skb = bat_priv->vis.my_info->skb_packet;
 	packet = (struct batadv_vis_packet *)skb_put(tmp_skb, sizeof(*packet));