[B.A.T.M.A.N.] [PATCHv6 0/7] batman-adv: Distributed ARP Table

[B.A.T.M.A.N.] [PATCHv6 0/7] batman-adv: Distributed ARP Table

[Antonio Quartulli]

Hello people,

again, this code is related to the GSOC2011 project: DAT: Distributed ARP Table
for B.A.T.M.A.N.-Advanced. For more details, please have a look at the related
wikipage[1]. A more detailed, formal and technical wikipage will be written
soon.

****
This is the **sixth** version of this patchset. The whole code has been
reviewed
after some discussions on irc and here in the ml in the previous threads.
Several bugs have been spotted and fixed. Thanks Marek, Sven, Simon, Andrew
and
all the others for the comments/feedbacks.

Major changes regards the UNICAST_4ADDR patch which has been moved at the
beginning of the patchset, so that the rest of the code is directly based
onto
it. The same patch has also been hardly modified in order to become cleaner and
reuse the already available code as much as possible, thus reducing the
replicated one.

* v6 Additions:
- handling code for UNICAST_4ADDR has been deeply modified
- README.external has been updated with the new compile option.
- initialisation code for dat has been enclosed into two functions in order to
  avoid having #ifdefs around for the batman-adv code
- some minor fixes
- debugging output improved
****

[1] http://www.open-mesh.org/wiki/batman-adv/GSOC2011_DAT

[B.A.T.M.A.N.] [PATCHv6 1/7] batman-adv: add UNICAST_4ADDR packet type

[Antonio Quartulli]

The current unicast packet type does not contain the orig source address. This
patches add a new unicast packet (called UNICAST_4ADDR) which provides two new
fields: the originator source address and the subtype (the type of the data
contained in the packet payload). The former is useful to identify the node
which injected the packet into the network and the latter is useful to avoid
creating new unicast packet types in the future: a macro defining a new subtype
will be enough.

Signed-off-by: Antonio Quartulli <ordex@autistici.org>
---
 hard-interface.c |    1 +
 packet.h         |   27 ++++++++++----
 routing.c        |    8 +++-
 unicast.c        |  100 +++++++++++++++++++++++++++++++++++++++++++----------
 unicast.h        |   17 +++++++++-
 5 files changed, 123 insertions(+), 30 deletions(-)

diff --git a/hard-interface.c b/hard-interface.c
index dfa948b..f18459c 100644
--- a/hard-interface.c
+++ b/hard-interface.c
@@ -627,6 +627,7 @@ static int batman_skb_recv(struct sk_buff *skb, struct net_device *dev,
 
 		/* unicast packet */
 	case BAT_UNICAST:
+	case BAT_UNICAST_4ADDR:
 		ret = recv_unicast_packet(skb, hard_iface);
 		break;
 
diff --git a/packet.h b/packet.h
index 7971a69..f704c51 100644
--- a/packet.h
+++ b/packet.h
@@ -25,14 +25,19 @@
 #define ETH_P_BATMAN  0x4305	/* unofficial/not registered Ethertype */
 
 enum bat_packettype {
-	BAT_IV_OGM	 = 0x01,
-	BAT_ICMP	 = 0x02,
-	BAT_UNICAST	 = 0x03,
-	BAT_BCAST	 = 0x04,
-	BAT_VIS		 = 0x05,
-	BAT_UNICAST_FRAG = 0x06,
-	BAT_TT_QUERY	 = 0x07,
-	BAT_ROAM_ADV	 = 0x08
+	BAT_IV_OGM		= 0x01,
+	BAT_ICMP		= 0x02,
+	BAT_UNICAST		= 0x03,
+	BAT_BCAST		= 0x04,
+	BAT_VIS			= 0x05,
+	BAT_UNICAST_FRAG	= 0x06,
+	BAT_TT_QUERY		= 0x07,
+	BAT_ROAM_ADV		= 0x08,
+	BAT_UNICAST_4ADDR	= 0x09
+};
+
+enum bat_subtype {
+	BAT_P_DATA		= 0x01
 };
 
 /* this file is included by batctl which needs these defines */
@@ -158,6 +163,12 @@ struct unicast_packet {
 	uint8_t  dest[ETH_ALEN];
 } __packed;
 
+struct unicast_4addr_packet {
+	struct unicast_packet u;
+	uint8_t src[ETH_ALEN];
+	uint8_t subtype;
+} __packed;
+
 struct unicast_frag_packet {
 	struct batman_header header;
 	uint8_t  ttvn; /* destination translation table version number */
diff --git a/routing.c b/routing.c
index a055386..f87b2ae 100644
--- a/routing.c
+++ b/routing.c
@@ -960,14 +960,18 @@ int recv_unicast_packet(struct sk_buff *skb, struct hard_iface *recv_if)
 	struct unicast_packet *unicast_packet;
 	int hdr_size = sizeof(*unicast_packet);
 
+	unicast_packet = (struct unicast_packet *)skb->data;
+
+	/* the caller function should have already pulled 2 bytes */
+	if (unicast_packet->header.packet_type == BAT_UNICAST_4ADDR)
+		hdr_size = sizeof(struct unicast_4addr_packet);
+
 	if (check_unicast_packet(skb, hdr_size) < 0)
 		return NET_RX_DROP;
 
 	if (!check_unicast_ttvn(bat_priv, skb))
 		return NET_RX_DROP;
 
-	unicast_packet = (struct unicast_packet *)skb->data;
-
 	/* packet for me */
 	if (is_my_mac(unicast_packet->dest)) {
 		interface_rx(recv_if->soft_iface, skb, recv_if, hdr_size);
diff --git a/unicast.c b/unicast.c
index 0897dfa..18e5cfd 100644
--- a/unicast.c
+++ b/unicast.c
@@ -283,13 +283,77 @@ out:
 	return ret;
 }
 
-int unicast_send_skb(struct sk_buff *skb, struct bat_priv *bat_priv)
+static bool pull_and_fill_unicast(struct sk_buff *skb, int hdr_size,
+				  struct orig_node *orig_node)
 {
-	struct ethhdr *ethhdr = (struct ethhdr *)skb->data;
 	struct unicast_packet *unicast_packet;
+
+	if (my_skb_head_push(skb, hdr_size) < 0)
+		return false;
+
+	unicast_packet = (struct unicast_packet *)skb->data;
+	unicast_packet->header.version = COMPAT_VERSION;
+	/* batman packet type: unicast */
+	unicast_packet->header.packet_type = BAT_UNICAST;
+	/* set unicast ttl */
+	unicast_packet->header.ttl = 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);
+
+	return true;
+}
+
+static bool prepare_unicast_packet(struct sk_buff *skb,
+				   struct orig_node *orig_node)
+{
+	return pull_and_fill_unicast(skb, sizeof(struct unicast_packet),
+				     orig_node);
+}
+
+static bool prepare_unicast_4addr_packet(struct bat_priv *bat_priv,
+					 struct sk_buff *skb,
+					 struct orig_node *orig_node,
+					 int packet_subtype)
+{
+	struct hard_iface *primary_if;
+	struct unicast_4addr_packet *unicast_4addr_packet;
+	bool ret = false;
+
+	primary_if = 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 (!pull_and_fill_unicast(skb, sizeof(*unicast_4addr_packet),
+				   orig_node))
+		goto out;
+
+	unicast_4addr_packet = (struct unicast_4addr_packet *)skb->data;
+	unicast_4addr_packet->u.header.packet_type = BAT_UNICAST_4ADDR;
+	memcpy(unicast_4addr_packet->src, primary_if->net_dev->dev_addr,
+	       ETH_ALEN);
+	unicast_4addr_packet->subtype = packet_subtype;
+
+	ret = true;
+out:
+	if (primary_if)
+		hardif_free_ref(primary_if);
+	return ret;
+}
+
+int unicast_generic_send_skb(struct sk_buff *skb, struct bat_priv *bat_priv,
+			     int packet_type, int packet_subtype)
+{
+	struct ethhdr *ethhdr = (struct ethhdr *)skb->data;
 	struct orig_node *orig_node;
 	struct neigh_node *neigh_node;
 	int data_len = skb->len;
+	struct unicast_packet *unicast_packet;
 	int ret = 1;
 
 	/* get routing information */
@@ -303,7 +367,6 @@ int unicast_send_skb(struct sk_buff *skb, struct bat_priv *bat_priv)
 	 * returns NULL in case of AP isolation */
 	orig_node = transtable_search(bat_priv, ethhdr->h_source,
 				      ethhdr->h_dest);
-
 find_router:
 	/**
 	 * find_router():
@@ -311,27 +374,27 @@ find_router:
 	 *  - increases neigh_nodes refcount if found.
 	 */
 	neigh_node = find_router(bat_priv, orig_node, NULL);
-
 	if (!neigh_node)
 		goto out;
 
-	if (my_skb_head_push(skb, sizeof(*unicast_packet)) < 0)
+	switch (packet_type) {
+	case BAT_UNICAST:
+		prepare_unicast_packet(skb, orig_node);
+		break;
+	case BAT_UNICAST_4ADDR:
+		prepare_unicast_4addr_packet(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 unicast_packet *)skb->data;
-
-	unicast_packet->header.version = COMPAT_VERSION;
-	/* batman packet type: unicast */
-	unicast_packet->header.packet_type = BAT_UNICAST;
-	/* set unicast ttl */
-	unicast_packet->header.ttl = 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);
-
-	if (atomic_read(&bat_priv->fragmentation) &&
+	/* fragmentation mechanism only works for UNICAST (now) */
+	if (packet_type == BAT_UNICAST &&
+	    atomic_read(&bat_priv->fragmentation) &&
 	    data_len + sizeof(*unicast_packet) >
 				neigh_node->if_incoming->net_dev->mtu) {
 		/* send frag skb decreases ttl */
@@ -343,7 +406,6 @@ find_router:
 
 	send_skb_packet(skb, neigh_node->if_incoming, neigh_node->addr);
 	ret = 0;
-	goto out;
 
 out:
 	if (neigh_node)
diff --git a/unicast.h b/unicast.h
index a9faf6b..ae9775b 100644
--- a/unicast.h
+++ b/unicast.h
@@ -30,9 +30,24 @@
 int frag_reassemble_skb(struct sk_buff *skb, struct bat_priv *bat_priv,
 			struct sk_buff **new_skb);
 void frag_list_free(struct list_head *head);
-int unicast_send_skb(struct sk_buff *skb, struct bat_priv *bat_priv);
 int frag_send_skb(struct sk_buff *skb, struct bat_priv *bat_priv,
 		  struct hard_iface *hard_iface, const uint8_t dstaddr[]);
+int unicast_generic_send_skb(struct sk_buff *skb, struct bat_priv *bat_priv,
+			     int packet_type, int packet_subtype);
+
+static inline int unicast_send_skb(struct sk_buff *skb,
+				   struct bat_priv *bat_priv)
+{
+	return unicast_generic_send_skb(skb, bat_priv, BAT_UNICAST, 0);
+}
+
+static inline int unicast_4addr_send_skb(struct sk_buff *skb,
+					 struct bat_priv *bat_priv,
+					 int packet_subtype)
+{
+	return unicast_generic_send_skb(skb, bat_priv, BAT_UNICAST_4ADDR,
+					packet_subtype);
+}
 
 static inline int frag_can_reassemble(const struct sk_buff *skb, int mtu)
 {
-- 
1.7.3.4

[B.A.T.M.A.N.] [PATCHv6 2/7] batman-adv: add a new log level for DAT/ARP debugging

[Antonio Quartulli]

A new log level has been added to concentrate messages regarding DAT: ARP
snooping, requests, response and DHT related messages.
The new log level is named DBG_ARP

Signed-off-by: Antonio Quartulli <ordex@autistici.org>
---
 README      |    3 ++-
 bat_sysfs.c |    2 +-
 main.h      |    3 ++-
 3 files changed, 5 insertions(+), 3 deletions(-)

diff --git a/README b/README
index 0c7a3c8..4ee2f54 100644
--- a/README
+++ b/README
@@ -201,7 +201,8 @@ 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.
diff --git a/bat_sysfs.c b/bat_sysfs.c
index 3adb183..abf7300 100644
--- a/bat_sysfs.c
+++ b/bat_sysfs.c
@@ -401,7 +401,7 @@ BAT_ATTR_UINT(gw_sel_class, S_IRUGO | S_IWUSR, 1, TQ_MAX_VALUE,
 static BAT_ATTR(gw_bandwidth, S_IRUGO | S_IWUSR, show_gw_bwidth,
 		store_gw_bwidth);
 #ifdef CONFIG_BATMAN_ADV_DEBUG
-BAT_ATTR_UINT(log_level, S_IRUGO | S_IWUSR, 0, 15, NULL);
+BAT_ATTR_UINT(log_level, S_IRUGO | S_IWUSR, 0, 31, NULL);
 #endif
 
 static struct bat_attribute *mesh_attrs[] = {
diff --git a/main.h b/main.h
index e669eea..3060f94 100644
--- a/main.h
+++ b/main.h
@@ -126,7 +126,8 @@ enum dbg_level {
 	DBG_ROUTES = 1 << 1, /* route added / changed / deleted */
 	DBG_TT	   = 1 << 2, /* translation table operations */
 	DBG_BLA    = 1 << 3, /* bridge loop avoidance */
-	DBG_ALL    = 15
+	DBG_ARP    = 1 << 4, /* snooped arp messages / dht operations */
+	DBG_ALL    = 31
 };
 
 
-- 
1.7.3.4

[B.A.T.M.A.N.] [PATCHv6 3/7] batman-adv: Distributed ARP Table - create DHT helper functions

[Antonio Quartulli]

Add all the relevant functions in order to manage a Distributed Hash Table over
the B.A.T.M.A.N.-adv network. It will later be used to store several ARP entries
and implement DAT (Distributed ARP Table)

Signed-off-by: Antonio Quartulli <ordex@autistici.org>
---
 Makefile.kbuild         |    1 +
 distributed-arp-table.c |  214 +++++++++++++++++++++++++++++++++++++++++++++++
 distributed-arp-table.h |   65 ++++++++++++++
 hard-interface.c        |    3 +
 main.h                  |    7 ++
 originator.c            |    2 +
 types.h                 |   15 ++++
 unicast.c               |    8 +-
 unicast.h               |    4 +
 9 files changed, 315 insertions(+), 4 deletions(-)
 create mode 100644 distributed-arp-table.c
 create mode 100644 distributed-arp-table.h

diff --git a/Makefile.kbuild b/Makefile.kbuild
index 6d5c194..84955b3 100644
--- a/Makefile.kbuild
+++ b/Makefile.kbuild
@@ -24,6 +24,7 @@ batman-adv-y += bat_iv_ogm.o
 batman-adv-y += bat_sysfs.o
 batman-adv-y += bitarray.o
 batman-adv-$(CONFIG_BATMAN_ADV_BLA) += bridge_loop_avoidance.o
+batman-adv-y += distributed-arp-table.o
 batman-adv-y += gateway_client.o
 batman-adv-y += gateway_common.o
 batman-adv-y += hard-interface.o
diff --git a/distributed-arp-table.c b/distributed-arp-table.c
new file mode 100644
index 0000000..f36b85c
--- /dev/null
+++ b/distributed-arp-table.c
@@ -0,0 +1,214 @@
+/*
+ * Copyright (C) 2011 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 "main.h"
+#include "distributed-arp-table.h"
+#include "hard-interface.h"
+#include "originator.h"
+#include "send.h"
+#include "types.h"
+#include "unicast.h"
+
+static bool is_orig_node_eligible(struct dht_candidate *res, int select,
+				  dat_addr_t tmp_max, dat_addr_t max,
+				  dat_addr_t last_max,
+				  struct orig_node *candidate,
+				  struct orig_node *max_orig_node)
+{
+	bool ret = false;
+	int j;
+
+	/* Check if we have already selected this neighbour... */
+	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 we have already found an originator
+	 * with a closer dht address */
+	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 && memcmp(candidate->orig, max_orig_node->orig,
+				     ETH_ALEN) > 0)
+		goto out;
+
+	ret = true;
+out:
+	return ret;
+}
+
+
+/* Given a key, selects the candidates which the DHT message has to be sent to.
+ * 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.
+ *
+ * return an array of size DHT_CANDIDATES_NUM */
+static struct dht_candidate *dht_select_candidates(struct bat_priv *bat_priv,
+						   uint32_t ip_dst)
+{
+	struct hashtable_t *hash = bat_priv->orig_hash;
+	struct hlist_node *node;
+	struct hlist_head *head;
+	struct orig_node *orig_node, *max_orig_node = NULL;
+	int select, i;
+	dat_addr_t last_max = DAT_ADDR_MAX, max, tmp_max, ip_key;
+	struct dht_candidate *res;
+	bool chosen_me = false;
+
+	if (!hash)
+		return NULL;
+
+	res = kmalloc(DHT_CANDIDATES_NUM * sizeof(*res), GFP_ATOMIC);
+	if (!res)
+		return NULL;
+
+	ip_key = (dat_addr_t)hash_ipv4(&ip_dst, DAT_ADDR_MAX);
+
+	bat_dbg(DBG_ARP, bat_priv, "dht_select_candidates(): IP=%pI4 "
+		"hash(IP)=%u\n", &ip_dst, ip_key);
+
+	for (select = 0; select < DHT_CANDIDATES_NUM; select++) {
+		max = 0;
+		max_orig_node = NULL;
+		/* Since the local node is not in the originator list, let's
+		 * first compute its distance from the key and assume it's the
+		 * maximum, then compare it to the other originators.
+		 * If the local node has already been chosen as candidate, jump
+		 * this step */
+		if (!chosen_me) {
+			/* if true, wrap around the key space */
+			if (bat_priv->dht_addr > ip_key)
+				max = DAT_ADDR_MAX - bat_priv->dht_addr +
+					ip_key;
+			else
+				max = ip_key - bat_priv->dht_addr;
+			res[select].type = DHT_CANDIDATE_ME;
+		} else
+			res[select].type = DHT_CANDIDATE_NULL;
+		/* iterate over the originator list and find the node with
+		 * closest dht_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, so we have to check
+				 * if we need to wrap around to compute the
+				 * distance from the key */
+				if (orig_node->dht_addr > ip_key)
+					tmp_max = DAT_ADDR_MAX -
+						  orig_node->dht_addr + ip_key;
+				else
+					tmp_max = ip_key - orig_node->dht_addr;
+				if (!is_orig_node_eligible(res, 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)
+					orig_node_free_ref(max_orig_node);
+				max_orig_node = orig_node;
+			}
+			rcu_read_unlock();
+		}
+		last_max = max;
+		if (max_orig_node) {
+			res[select].type = DHT_CANDIDATE_ORIG;
+			res[select].orig_node = max_orig_node;
+			bat_dbg(DBG_ARP, bat_priv, "dht_select_candidates() %d: "
+				"selected %pM addr=%u dist=%u\n", select,
+				max_orig_node->orig, max_orig_node->dht_addr,
+				max);
+		} else if (res[select].type == DHT_CANDIDATE_ME) {
+			chosen_me = true;
+			bat_dbg(DBG_ARP, bat_priv, "dht_select_candidates() %d: "
+				"selected ME addr=%u dist=%u\n", select,
+				bat_priv->dht_addr, max);
+		}
+
+		max_orig_node = NULL;
+	}
+
+	return res;
+}
+
+/*
+ * Sends the skb payload passed as argument to the candidates selected for
+ * 'ip'. The skb is copied by means of pskb_copy() and is sent as unicast packet
+ * to each of the selected candidate. */
+static bool dht_send_data(struct bat_priv *bat_priv, struct sk_buff *skb,
+			  uint32_t ip, int packet_subtype)
+{
+	int i;
+	bool ret = false;
+	struct neigh_node *neigh_node = NULL;
+	struct sk_buff *tmp_skb;
+	struct dht_candidate *cand = dht_select_candidates(bat_priv, ip);
+
+	if (!cand)
+		goto out;
+
+	bat_dbg(DBG_ARP, bat_priv, "DHT_SEND for %pI4\n", &ip);
+
+	for (i = 0; i < DHT_CANDIDATES_NUM; i++) {
+		if (cand[i].type == DHT_CANDIDATE_ME ||
+		    cand[i].type == DHT_CANDIDATE_NULL)
+			continue;
+
+		neigh_node = orig_node_get_router(cand[i].orig_node);
+		if (!neigh_node)
+			goto free_orig;
+
+		tmp_skb = pskb_copy(skb, GFP_ATOMIC);
+		if (prepare_unicast_4addr_packet(bat_priv, tmp_skb,
+						 cand[i].orig_node,
+						 packet_subtype))
+			send_skb_packet(tmp_skb, neigh_node->if_incoming,
+					neigh_node->addr);
+		else
+			kfree_skb(tmp_skb);
+		/* set ret to true only if we send at least one request */
+		ret = true;
+		neigh_node_free_ref(neigh_node);
+free_orig:
+		orig_node_free_ref(cand[i].orig_node);
+	}
+
+out:
+	kfree(cand);
+	return ret;
+}
diff --git a/distributed-arp-table.h b/distributed-arp-table.h
new file mode 100644
index 0000000..89d52c8
--- /dev/null
+++ b/distributed-arp-table.h
@@ -0,0 +1,65 @@
+/*
+ * Copyright (C) 2011 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_
+
+#include "types.h"
+#include "originator.h"
+
+#define DAT_ADDR_MAX biggest_unsigned_int(dat_addr_t)
+
+/* hash function to choose an entry in a hash table of given size */
+/* hash algorithm from http://en.wikipedia.org/wiki/Hash_table */
+static inline uint32_t hash_ipv4(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;
+}
+
+static inline void dat_init_orig_node_dht_addr(struct orig_node *orig_node)
+{
+	orig_node->dht_addr = (dat_addr_t)choose_orig(orig_node->orig,
+						      DAT_ADDR_MAX);
+}
+
+static inline void dat_init_own_dht_addr(struct bat_priv *bat_priv,
+					 struct hard_iface *primary_if)
+{
+	bat_priv->dht_addr = (dat_addr_t)
+				choose_orig(primary_if->net_dev->dev_addr,
+					    DAT_ADDR_MAX);
+}
+
+#endif /* _NET_BATMAN_ADV_ARP_H_ */
diff --git a/hard-interface.c b/hard-interface.c
index f18459c..77e3163 100644
--- a/hard-interface.c
+++ b/hard-interface.c
@@ -20,6 +20,7 @@
  */
 
 #include "main.h"
+#include "distributed-arp-table.h"
 #include "hard-interface.h"
 #include "soft-interface.h"
 #include "send.h"
@@ -118,6 +119,8 @@ static void primary_if_update_addr(struct bat_priv *bat_priv,
 	if (!primary_if)
 		goto out;
 
+	dat_init_own_dht_addr(bat_priv, primary_if);
+
 	vis_packet = (struct vis_packet *)
 				bat_priv->my_vis_info->skb_packet->data;
 	memcpy(vis_packet->vis_orig, primary_if->net_dev->dev_addr, ETH_ALEN);
diff --git a/main.h b/main.h
index 3060f94..86a3d4a 100644
--- a/main.h
+++ b/main.h
@@ -67,6 +67,9 @@
 
 #define NUM_WORDS BITS_TO_LONGS(TQ_LOCAL_WINDOW_SIZE)
 
+/* numbers of originator to contact for any PUT/GET DHT operation */
+#define DHT_CANDIDATES_NUM 3
+
 #define LOG_BUF_LEN 8192	  /* has to be a power of 2 */
 
 #define VIS_INTERVAL 5000	/* 5 seconds */
@@ -111,6 +114,10 @@ enum uev_type {
 
 #define GW_THRESHOLD	50
 
+#define DHT_CANDIDATE_NULL 0
+#define DHT_CANDIDATE_ME   1
+#define DHT_CANDIDATE_ORIG 2
+
 /*
  * Debug Messages
  */
diff --git a/originator.c b/originator.c
index 2db4b53..1ee6e82 100644
--- a/originator.c
+++ b/originator.c
@@ -20,6 +20,7 @@
  */
 
 #include "main.h"
+#include "distributed-arp-table.h"
 #include "originator.h"
 #include "hash.h"
 #include "translation-table.h"
@@ -224,6 +225,7 @@ struct orig_node *get_orig_node(struct bat_priv *bat_priv, const uint8_t *addr)
 	orig_node->tt_poss_change = false;
 	orig_node->bat_priv = bat_priv;
 	memcpy(orig_node->orig, addr, ETH_ALEN);
+	dat_init_orig_node_dht_addr(orig_node);
 	orig_node->router = NULL;
 	orig_node->tt_crc = 0;
 	atomic_set(&orig_node->last_ttvn, 0);
diff --git a/types.h b/types.h
index 2f4848b..210ee3e 100644
--- a/types.h
+++ b/types.h
@@ -27,6 +27,14 @@
 #include "packet.h"
 #include "bitarray.h"
 
+/*
+ * dat_addr_t is the type used for all DHT addresses. If it is changed,
+ * DAT_ADDR_MAX is changed as well.
+ *
+ * *Please be careful: dat_addr_t must be UNSIGNED*
+ */
+#define dat_addr_t uint16_t
+
 #define BAT_HEADER_LEN (ETH_HLEN + \
 	((sizeof(struct unicast_packet) > sizeof(struct bcast_packet) ? \
 	 sizeof(struct unicast_packet) : \
@@ -67,6 +75,7 @@ struct hard_iface {
 struct orig_node {
 	uint8_t orig[ETH_ALEN];
 	uint8_t primary_addr[ETH_ALEN];
+	dat_addr_t dht_addr;
 	struct neigh_node __rcu *router; /* rcu protected pointer */
 	unsigned long *bcast_own;
 	uint8_t *bcast_own_sum;
@@ -221,6 +230,7 @@ struct bat_priv {
 	struct gw_node __rcu *curr_gw;  /* rcu protected pointer */
 	atomic_t gw_reselect;
 	struct hard_iface __rcu *primary_if;  /* rcu protected pointer */
+	dat_addr_t dht_addr;
 	struct vis_info *my_vis_info;
 	struct bat_algo_ops *bat_algo_ops;
 };
@@ -395,4 +405,9 @@ struct bat_algo_ops {
 				struct sk_buff *skb);
 };
 
+struct dht_candidate {
+	int type;
+	struct orig_node *orig_node;
+};
+
 #endif /* _NET_BATMAN_ADV_TYPES_H_ */
diff --git a/unicast.c b/unicast.c
index 18e5cfd..e363a48 100644
--- a/unicast.c
+++ b/unicast.c
@@ -313,10 +313,10 @@ static bool prepare_unicast_packet(struct sk_buff *skb,
 				     orig_node);
 }
 
-static bool prepare_unicast_4addr_packet(struct bat_priv *bat_priv,
-					 struct sk_buff *skb,
-					 struct orig_node *orig_node,
-					 int packet_subtype)
+bool prepare_unicast_4addr_packet(struct bat_priv *bat_priv,
+				  struct sk_buff *skb,
+				  struct orig_node *orig_node,
+				  int packet_subtype)
 {
 	struct hard_iface *primary_if;
 	struct unicast_4addr_packet *unicast_4addr_packet;
diff --git a/unicast.h b/unicast.h
index ae9775b..e15aa62 100644
--- a/unicast.h
+++ b/unicast.h
@@ -32,6 +32,10 @@ int frag_reassemble_skb(struct sk_buff *skb, struct bat_priv *bat_priv,
 void frag_list_free(struct list_head *head);
 int frag_send_skb(struct sk_buff *skb, struct bat_priv *bat_priv,
 		  struct hard_iface *hard_iface, const uint8_t dstaddr[]);
+bool prepare_unicast_4addr_packet(struct bat_priv *bat_priv,
+				  struct sk_buff *skb,
+				  struct orig_node *orig_node,
+				  int packet_subtype);
 int unicast_generic_send_skb(struct sk_buff *skb, struct bat_priv *bat_priv,
 			     int packet_type, int packet_subtype);
 
-- 
1.7.3.4

[B.A.T.M.A.N.] [PATCHv6 4/7] batman-adv: Distributed ARP Table - add ARP parsing functions

[Antonio Quartulli]

ARP messages are now parsed to make it possible to trigger special actions
depending on their types (snooping).

Signed-off-by: Antonio Quartulli <ordex@autistici.org>
---
 distributed-arp-table.c |  117 ++++++++++++++++++++++++++++++++++++++++++++++-
 distributed-arp-table.h |   10 ++++
 packet.h                |    5 ++-
 3 files changed, 130 insertions(+), 2 deletions(-)

diff --git a/distributed-arp-table.c b/distributed-arp-table.c
index f36b85c..84e4630 100644
--- a/distributed-arp-table.c
+++ b/distributed-arp-table.c
@@ -30,6 +30,70 @@
 #include "types.h"
 #include "unicast.h"
 
+#ifdef CONFIG_BATMAN_ADV_DEBUG
+
+static void bat_dbg_arp(struct bat_priv *bat_priv, struct sk_buff *skb,
+			uint16_t type, int hdr_size, char *msg) {
+	struct unicast_4addr_packet *unicast_4addr_packet;
+
+	if (msg)
+		bat_dbg(DBG_ARP, bat_priv, "%s\n", msg);
+
+	bat_dbg(DBG_ARP, bat_priv, "ARP MSG = [src: %pM-%pI4 dst: %pM-%pI4]\n",
+		ARP_HW_SRC(skb, hdr_size), &ARP_IP_SRC(skb, hdr_size),
+		ARP_HW_DST(skb, hdr_size), &ARP_IP_DST(skb, hdr_size));
+
+	if (hdr_size == 0)
+		return;
+
+	/* if the AP packet is encapsulated in a batman packet, let's print some
+	 * debug messages */
+	unicast_4addr_packet = (struct unicast_4addr_packet *)skb->data;
+
+	switch (unicast_4addr_packet->u.header.packet_type) {
+	case BAT_UNICAST:
+		bat_dbg(DBG_ARP, bat_priv, "* encapsulated within a UNICAST "
+			"packet\n");
+		break;
+	case BAT_UNICAST_4ADDR:
+		bat_dbg(DBG_ARP, bat_priv, "* encapsulated within a "
+			"UNICAST_4ADDR packet (src: %pM)\n",
+			unicast_4addr_packet->src);
+		switch (unicast_4addr_packet->subtype) {
+		case BAT_P_DAT_DHT_PUT:
+			bat_dbg(DBG_ARP, bat_priv, "* type: DAT_DHT_PUT\n");
+			break;
+		case BAT_P_DAT_DHT_GET:
+			bat_dbg(DBG_ARP, bat_priv, "* type: DAT_DHT_GET\n");
+			break;
+		case BAT_P_DAT_CACHE_REPLY:
+			bat_dbg(DBG_ARP, bat_priv, "* type: DAT_CACHE_REPLY\n");
+			break;
+		case BAT_P_DATA:
+			bat_dbg(DBG_ARP, bat_priv, "* type: DATA\n");
+			break;
+		default:
+			bat_dbg(DBG_ARP, bat_priv, "* type: Unknown!\n");
+		}
+		break;
+	case BAT_BCAST:
+		bat_dbg(DBG_ARP, bat_priv, "* encapsulated within a BCAST "
+			"packet (src: %pM)\n",
+			((struct bcast_packet *)unicast_4addr_packet)->orig);
+		break;
+	default:
+		bat_dbg(DBG_ARP, bat_priv, "* encapsulated within an unknown "
+			"packet type (0x%x)\n",
+			unicast_4addr_packet->u.header.packet_type);
+	}
+}
+
+#else
+
+#define bat_dbg_arp(...)
+
+#endif /* CONFIG_BATMAN_ADV_DEBUG */
+
 static bool is_orig_node_eligible(struct dht_candidate *res, int select,
 				  dat_addr_t tmp_max, dat_addr_t max,
 				  dat_addr_t last_max,
@@ -64,7 +128,6 @@ out:
 	return ret;
 }
 
-
 /* Given a key, selects the candidates which the DHT message has to be sent to.
  * 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
@@ -212,3 +275,55 @@ out:
 	kfree(cand);
 	return ret;
 }
+
+/* Returns arphdr->ar_op if the skb contains a valid ARP packet, otherwise
+ * returns 0 */
+static uint16_t arp_get_type(struct bat_priv *bat_priv, struct sk_buff *skb,
+			     int hdr_size)
+{
+	struct arphdr *arphdr;
+	struct ethhdr *ethhdr;
+	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 */
+	if (ipv4_is_loopback(ARP_IP_SRC(skb, hdr_size)) ||
+	    ipv4_is_multicast(ARP_IP_SRC(skb, hdr_size)) ||
+	    ipv4_is_loopback(ARP_IP_DST(skb, hdr_size)) ||
+	    ipv4_is_multicast(ARP_IP_DST(skb, hdr_size)))
+		goto out;
+
+	type = ntohs(arphdr->ar_op);
+out:
+	return type;
+}
diff --git a/distributed-arp-table.h b/distributed-arp-table.h
index 89d52c8..b5bf2d1 100644
--- a/distributed-arp-table.h
+++ b/distributed-arp-table.h
@@ -25,8 +25,18 @@
 #include "types.h"
 #include "originator.h"
 
+#include <linux/if_arp.h>
+
 #define DAT_ADDR_MAX biggest_unsigned_int(dat_addr_t)
 
+#define ARP_HW_SRC(skb, hdr_size) ((uint8_t *)(skb->data + hdr_size) + \
+				   ETH_HLEN + sizeof(struct arphdr))
+#define ARP_IP_SRC(skb, hdr_size) (*(uint32_t *)(ARP_HW_SRC(skb, hdr_size) + \
+				   ETH_ALEN))
+#define ARP_HW_DST(skb, hdr_size) (ARP_HW_SRC(skb, hdr_size) + ETH_ALEN + 4)
+#define ARP_IP_DST(skb, hdr_size) (*(uint32_t *)(ARP_HW_SRC(skb, hdr_size) + \
+				   ETH_ALEN * 2 + 4))
+
 /* hash function to choose an entry in a hash table of given size */
 /* hash algorithm from http://en.wikipedia.org/wiki/Hash_table */
 static inline uint32_t hash_ipv4(const void *data, uint32_t size)
diff --git a/packet.h b/packet.h
index f704c51..02b0c87 100644
--- a/packet.h
+++ b/packet.h
@@ -37,7 +37,10 @@ enum bat_packettype {
 };
 
 enum bat_subtype {
-	BAT_P_DATA		= 0x01
+	BAT_P_DATA		= 0x01,
+	BAT_P_DAT_DHT_GET	= 0x02,
+	BAT_P_DAT_DHT_PUT	= 0x03,
+	BAT_P_DAT_CACHE_REPLY	= 0x04
 };
 
 /* this file is included by batctl which needs these defines */
-- 
1.7.3.4

[B.A.T.M.A.N.] [PATCHv6 5/7] batman-adv: Distributed ARP Table - add snooping functions for ARP messages

[Antonio Quartulli]

In case of an ARP message going in or out the soft_iface, it is intercepted and
a special action is performed. In particular the DHT helper functions previously
implemented are used to store all the ARP entries belonging to the network in
order to provide a fast and unicast lookup instead of the classic broadcast
flooding mechanism.
Each node stores the entries it is responsible for (following the DHT rules) in
its soft_iface ARP table. This makes it possible to reuse the kernel data
structures and functions for ARP management.

Signed-off-by: Antonio Quartulli <ordex@autistici.org>
---
 distributed-arp-table.c |  246 +++++++++++++++++++++++++++++++++++++++++++++++
 distributed-arp-table.h |   11 ++
 main.h                  |    2 +
 send.c                  |    6 +
 soft-interface.c        |   18 ++++-
 5 files changed, 282 insertions(+), 1 deletions(-)

diff --git a/distributed-arp-table.c b/distributed-arp-table.c
index 84e4630..3693879 100644
--- a/distributed-arp-table.c
+++ b/distributed-arp-table.c
@@ -21,6 +21,8 @@
 
 #include <linux/if_ether.h>
 #include <linux/if_arp.h>
+/* needed to use arp_tbl */
+#include <net/arp.h>
 
 #include "main.h"
 #include "distributed-arp-table.h"
@@ -28,6 +30,7 @@
 #include "originator.h"
 #include "send.h"
 #include "types.h"
+#include "translation-table.h"
 #include "unicast.h"
 
 #ifdef CONFIG_BATMAN_ADV_DEBUG
@@ -276,6 +279,31 @@ out:
 	return ret;
 }
 
+/* Update the neighbour entry corresponding to the IP passed as parameter with
+ * the hw address hw. If the neighbour entry doesn't exists, then it will be
+ * created */
+static void arp_neigh_update(struct bat_priv *bat_priv, uint32_t ip,
+			     uint8_t *hw)
+{
+	struct neighbour *n = NULL;
+	struct hard_iface *primary_if = primary_if_get_selected(bat_priv);
+	if (!primary_if)
+		goto out;
+
+	n = __neigh_lookup(&arp_tbl, &ip, primary_if->soft_iface, 1);
+	if (!n)
+		goto out;
+
+	bat_dbg(DBG_ARP, bat_priv, "Updating neighbour: %pI4 - %pM\n", &ip, hw);
+
+	neigh_update(n, hw, NUD_REACHABLE, NEIGH_UPDATE_F_OVERRIDE);
+out:
+	if (n && !IS_ERR(n))
+		neigh_release(n);
+	if (primary_if)
+		hardif_free_ref(primary_if);
+}
+
 /* Returns arphdr->ar_op if the skb contains a valid ARP packet, otherwise
  * returns 0 */
 static uint16_t arp_get_type(struct bat_priv *bat_priv, struct sk_buff *skb,
@@ -327,3 +355,221 @@ static uint16_t arp_get_type(struct bat_priv *bat_priv, struct sk_buff *skb,
 out:
 	return type;
 }
+
+/* return 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 dat_snoop_outgoing_arp_request(struct bat_priv *bat_priv,
+				    struct sk_buff *skb)
+{
+	uint16_t type = 0;
+	uint32_t ip_dst, ip_src;
+	uint8_t *hw_src;
+	bool ret = false;
+	struct neighbour *n = NULL;
+	struct hard_iface *primary_if = NULL;
+	struct sk_buff *skb_new;
+
+	type = arp_get_type(bat_priv, skb, 0);
+	/* If we get an ARP_REQUEST we have to send the unicast message to the
+	 * selected DHT candidates */
+	if (type != ARPOP_REQUEST)
+		goto out;
+
+	bat_dbg_arp(bat_priv, skb, type, 0, "Parsing outgoing ARP REQUEST");
+
+	ip_src = ARP_IP_SRC(skb, 0);
+	hw_src = ARP_HW_SRC(skb, 0);
+	ip_dst = ARP_IP_DST(skb, 0);
+
+	primary_if = primary_if_get_selected(bat_priv);
+	if (!primary_if)
+		goto out;
+
+	arp_neigh_update(bat_priv, ip_src, hw_src);
+
+	n = neigh_lookup(&arp_tbl, &ip_dst, primary_if->soft_iface);
+	/* check if it is a valid neigh entry */
+	if (n && (n->nud_state & NUD_CONNECTED)) {
+		skb_new = arp_create(ARPOP_REPLY, ETH_P_ARP, ip_src,
+				     primary_if->soft_iface, ip_dst, hw_src,
+				     n->ha, 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);
+		bat_dbg(DBG_ARP, bat_priv, "ARP request replied locally\n");
+	} else
+		/* Send the request on the DHT */
+		ret = dht_send_data(bat_priv, skb, ip_dst, BAT_P_DAT_DHT_GET);
+out:
+	if (n)
+		neigh_release(n);
+	if (primary_if)
+		hardif_free_ref(primary_if);
+	return ret;
+}
+
+/* This function is meant to be invoked for an ARP request which is coming into
+ * the bat0 interfaces from the mesh network. It will check for the needed data
+ * into the local table. If found, an ARP reply is sent immediately, otherwise
+ * the caller has to deliver the ARP request to the upper layer */
+bool dat_snoop_incoming_arp_request(struct bat_priv *bat_priv,
+				    struct sk_buff *skb, int hdr_size)
+{
+	uint16_t type;
+	uint32_t ip_src, ip_dst;
+	uint8_t *hw_src;
+	struct hard_iface *primary_if = NULL;
+	struct sk_buff *skb_new;
+	struct neighbour *n = NULL;
+	bool ret = false;
+
+	type = arp_get_type(bat_priv, skb, hdr_size);
+	if (type != ARPOP_REQUEST)
+		goto out;
+
+	hw_src = ARP_HW_SRC(skb, hdr_size);
+	ip_src = ARP_IP_SRC(skb, hdr_size);
+	ip_dst = ARP_IP_DST(skb, hdr_size);
+
+	bat_dbg_arp(bat_priv, skb, type, hdr_size,
+		    "Parsing incoming ARP REQUEST");
+
+	primary_if = primary_if_get_selected(bat_priv);
+	if (!primary_if)
+		goto out;
+
+	arp_neigh_update(bat_priv, ip_src, hw_src);
+
+	n = neigh_lookup(&arp_tbl, &ip_dst, primary_if->soft_iface);
+	/* check if it is a valid neigh entry */
+	if (!n || !(n->nud_state & NUD_CONNECTED))
+		goto out;
+
+	skb_new = arp_create(ARPOP_REPLY, ETH_P_ARP, ip_src,
+			     primary_if->soft_iface, ip_dst, hw_src, n->ha,
+			     hw_src);
+
+	if (!skb_new)
+		goto out;
+
+	unicast_4addr_send_skb(skb_new, bat_priv, BAT_P_DAT_CACHE_REPLY);
+
+	ret = true;
+out:
+	if (n)
+		neigh_release(n);
+	if (primary_if)
+		hardif_free_ref(primary_if);
+	if (ret)
+		kfree_skb(skb);
+	return ret;
+}
+
+/* This function is meant to be invoked on an ARP reply packet going into the
+ * soft interface. The related neighbour entry has to be updated and the DHT has
+ * to be populated as well */
+bool dat_snoop_outgoing_arp_reply(struct bat_priv *bat_priv,
+				  struct sk_buff *skb)
+{
+	uint16_t type;
+	uint32_t ip_src, ip_dst;
+	uint8_t *hw_src, *hw_dst;
+	bool ret = false;
+
+	type = arp_get_type(bat_priv, skb, 0);
+	if (type != ARPOP_REPLY)
+		goto out;
+
+	bat_dbg_arp(bat_priv, skb, type, 0, "Parsing outgoing ARP REPLY");
+
+	hw_src = ARP_HW_SRC(skb, 0);
+	ip_src = ARP_IP_SRC(skb, 0);
+	hw_dst = ARP_HW_DST(skb, 0);
+	ip_dst = ARP_IP_DST(skb, 0);
+
+	arp_neigh_update(bat_priv, ip_src, hw_src);
+	arp_neigh_update(bat_priv, ip_dst, hw_dst);
+
+	/* Send the ARP reply to the candidates for both the IP addresses we
+	 * fetched from the ARP reply */
+	dht_send_data(bat_priv, skb, ip_src, BAT_P_DAT_DHT_PUT);
+	dht_send_data(bat_priv, skb, ip_dst, BAT_P_DAT_DHT_PUT);
+	ret = true;
+out:
+	return ret;
+}
+
+/* This function has to be invoked on an ARP reply coming into the soft
+ * interface from the mesh network. The local table has to be updated */
+bool dat_snoop_incoming_arp_reply(struct bat_priv *bat_priv,
+				  struct sk_buff *skb, int hdr_size)
+{
+	uint16_t type;
+	uint32_t ip_src, ip_dst;
+	uint8_t *hw_src, *hw_dst;
+	bool ret = false;
+
+	type = arp_get_type(bat_priv, skb, hdr_size);
+	if (type != ARPOP_REPLY)
+		goto out;
+
+	bat_dbg_arp(bat_priv, skb, type, hdr_size,
+		    "Parsing incoming ARP REPLY");
+
+	hw_src = ARP_HW_SRC(skb, hdr_size);
+	ip_src = ARP_IP_SRC(skb, hdr_size);
+	hw_dst = ARP_HW_DST(skb, hdr_size);
+	ip_dst = ARP_IP_DST(skb, hdr_size);
+
+	/* Update our internal cache with both the IP addresses we fetched from
+	 * the ARP reply */
+	arp_neigh_update(bat_priv, ip_src, hw_src);
+	arp_neigh_update(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 = !is_my_client(bat_priv, hw_dst);
+out:
+	/* if ret == false packet has to be delivered to the interface */
+	return ret;
+}
+
+bool arp_drop_broadcast_packet(struct bat_priv *bat_priv,
+			       struct forw_packet *forw_packet)
+{
+	struct neighbour *n;
+
+	/* If this packet is an ARP_REQUEST and we already have the information
+	 * that it is going to ask, we can drop the packet */
+	if (!forw_packet->num_packets &&
+			(arp_get_type(bat_priv, forw_packet->skb,
+				      sizeof(struct bcast_packet)) ==
+							ARPOP_REQUEST)) {
+		n = neigh_lookup(&arp_tbl,
+				 &ARP_IP_DST(forw_packet->skb,
+					     sizeof(struct bcast_packet)),
+				 forw_packet->if_incoming->soft_iface);
+		/* check if we already know this neigh */
+		if (n && (n->nud_state & NUD_CONNECTED)) {
+			bat_dbg(DBG_ARP, bat_priv, "ARP Request for %pI4: "
+				"fallback prevented\n",
+				&ARP_IP_DST(forw_packet->skb,
+					    sizeof(struct bcast_packet)));
+			return true;
+		}
+
+		bat_dbg(DBG_ARP, bat_priv, "ARP Request for %pI4: fallback\n",
+			&ARP_IP_DST(forw_packet->skb,
+				    sizeof(struct bcast_packet)));
+	}
+	return false;
+}
diff --git a/distributed-arp-table.h b/distributed-arp-table.h
index b5bf2d1..3b90b73 100644
--- a/distributed-arp-table.h
+++ b/distributed-arp-table.h
@@ -37,6 +37,17 @@
 #define ARP_IP_DST(skb, hdr_size) (*(uint32_t *)(ARP_HW_SRC(skb, hdr_size) + \
 				   ETH_ALEN * 2 + 4))
 
+bool dat_snoop_outgoing_arp_request(struct bat_priv *bat_priv,
+				    struct sk_buff *skb);
+bool dat_snoop_incoming_arp_request(struct bat_priv *bat_priv,
+				    struct sk_buff *skb, int hdr_size);
+bool dat_snoop_outgoing_arp_reply(struct bat_priv *bat_priv,
+				  struct sk_buff *skb);
+bool dat_snoop_incoming_arp_reply(struct bat_priv *bat_priv,
+				  struct sk_buff *skb, int hdr_size);
+bool arp_drop_broadcast_packet(struct bat_priv *bat_priv,
+			       struct forw_packet *forw_packet);
+
 /* hash function to choose an entry in a hash table of given size */
 /* hash algorithm from http://en.wikipedia.org/wiki/Hash_table */
 static inline uint32_t hash_ipv4(const void *data, uint32_t size)
diff --git a/main.h b/main.h
index 86a3d4a..7384abb 100644
--- a/main.h
+++ b/main.h
@@ -67,6 +67,8 @@
 
 #define NUM_WORDS BITS_TO_LONGS(TQ_LOCAL_WINDOW_SIZE)
 
+/* 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 DHT_CANDIDATES_NUM 3
 
diff --git a/send.c b/send.c
index 1d5353f..8bd31bb 100644
--- a/send.c
+++ b/send.c
@@ -20,6 +20,7 @@
  */
 
 #include "main.h"
+#include "distributed-arp-table.h"
 #include "send.h"
 #include "routing.h"
 #include "translation-table.h"
@@ -29,6 +30,8 @@
 #include "gateway_common.h"
 #include "originator.h"
 
+#include <net/arp.h>
+
 static void send_outstanding_bcast_packet(struct work_struct *work);
 
 /* send out an already prepared packet to the given address via the
@@ -274,6 +277,9 @@ static void send_outstanding_bcast_packet(struct work_struct *work)
 	if (atomic_read(&bat_priv->mesh_state) == MESH_DEACTIVATING)
 		goto out;
 
+	if (arp_drop_broadcast_packet(bat_priv, forw_packet))
+		goto out;
+
 	/* rebroadcast packet */
 	rcu_read_lock();
 	list_for_each_entry_rcu(hard_iface, &hardif_list, list) {
diff --git a/soft-interface.c b/soft-interface.c
index 3297c40..f27ea7e 100644
--- a/soft-interface.c
+++ b/soft-interface.c
@@ -22,6 +22,7 @@
 #include "main.h"
 #include "soft-interface.h"
 #include "hard-interface.h"
+#include "distributed-arp-table.h"
 #include "routing.h"
 #include "send.h"
 #include "bat_debugfs.h"
@@ -136,6 +137,7 @@ static int interface_tx(struct sk_buff *skb, struct net_device *soft_iface)
 	int data_len = skb->len, ret;
 	short vid __maybe_unused = -1;
 	bool do_bcast = false;
+	unsigned long brd_delay = 1;
 
 	if (atomic_read(&bat_priv->mesh_state) != MESH_ACTIVE)
 		goto dropped;
@@ -196,6 +198,9 @@ static int interface_tx(struct sk_buff *skb, struct net_device *soft_iface)
 		if (!primary_if)
 			goto dropped;
 
+		if (dat_snoop_outgoing_arp_request(bat_priv, skb))
+			brd_delay = msecs_to_jiffies(ARP_REQ_DELAY);
+
 		if (my_skb_head_push(skb, sizeof(*bcast_packet)) < 0)
 			goto dropped;
 
@@ -215,7 +220,7 @@ static int interface_tx(struct sk_buff *skb, struct net_device *soft_iface)
 		bcast_packet->seqno =
 			htonl(atomic_inc_return(&bat_priv->bcast_seqno));
 
-		add_bcast_packet_to_list(bat_priv, skb, 1);
+		add_bcast_packet_to_list(bat_priv, skb, brd_delay);
 
 		/* a copy is stored in the bcast list, therefore removing
 		 * the original skb. */
@@ -229,6 +234,8 @@ static int interface_tx(struct sk_buff *skb, struct net_device *soft_iface)
 				goto dropped;
 		}
 
+		dat_snoop_outgoing_arp_reply(bat_priv, skb);
+
 		ret = unicast_send_skb(skb, bat_priv);
 		if (ret != 0)
 			goto dropped_freed;
@@ -253,6 +260,7 @@ void interface_rx(struct net_device *soft_iface,
 		  int hdr_size)
 {
 	struct bat_priv *bat_priv = netdev_priv(soft_iface);
+	struct unicast_4addr_packet *unicast_4addr_packet;
 	struct ethhdr *ethhdr;
 	struct vlan_ethhdr *vhdr;
 	short vid __maybe_unused = -1;
@@ -261,6 +269,14 @@ void interface_rx(struct net_device *soft_iface,
 	if (!pskb_may_pull(skb, hdr_size))
 		goto dropped;
 
+	unicast_4addr_packet = (struct unicast_4addr_packet *)skb->data;
+
+	if (dat_snoop_incoming_arp_request(bat_priv, skb, hdr_size))
+		goto out;
+
+	if (dat_snoop_incoming_arp_reply(bat_priv, skb, hdr_size))
+		goto out;
+
 	skb_pull_rcsum(skb, hdr_size);
 	skb_reset_mac_header(skb);
 
-- 
1.7.3.4

[B.A.T.M.A.N.] [PATCHv6 6/7] batman-adv: Distributed ARP Table - increase default soft_iface ARP table timeout

[Antonio Quartulli]

The default timeout value for ARP entries belonging to any soft_iface
ARP table has been incremented by a factor 4. This is necessary because the DHT
will store several network entries in the soft_iface ARP table.

Signed-off-by: Antonio Quartulli <ordex@autistici.org>
---
 distributed-arp-table.c |   19 +++++++++++++++++++
 distributed-arp-table.h |    1 +
 main.h                  |    3 +++
 soft-interface.c        |    2 ++
 4 files changed, 25 insertions(+), 0 deletions(-)

diff --git a/distributed-arp-table.c b/distributed-arp-table.c
index 3693879..57b0c43 100644
--- a/distributed-arp-table.c
+++ b/distributed-arp-table.c
@@ -23,6 +23,7 @@
 #include <linux/if_arp.h>
 /* needed to use arp_tbl */
 #include <net/arp.h>
+#include <linux/inetdevice.h>
 
 #include "main.h"
 #include "distributed-arp-table.h"
@@ -573,3 +574,21 @@ bool arp_drop_broadcast_packet(struct bat_priv *bat_priv,
 	}
 	return false;
 }
+
+void arp_change_timeout(struct net_device *soft_iface, const char *name)
+{
+	struct in_device *in_dev = in_dev_get(soft_iface);
+	if (!in_dev) {
+		pr_err("Unable to set ARP parameters for the batman interface "
+		       "'%s'\n", name);
+		return;
+	}
+
+	/* Introduce a delay in the ARP state-machine transactions. Entries
+	 * will be kept in the ARP table for the default time multiplied by 4 */
+	in_dev->arp_parms->base_reachable_time *= ARP_TIMEOUT_FACTOR;
+	in_dev->arp_parms->gc_staletime *= ARP_TIMEOUT_FACTOR;
+	in_dev->arp_parms->reachable_time *= ARP_TIMEOUT_FACTOR;
+
+	in_dev_put(in_dev);
+}
diff --git a/distributed-arp-table.h b/distributed-arp-table.h
index 3b90b73..bf4390d 100644
--- a/distributed-arp-table.h
+++ b/distributed-arp-table.h
@@ -47,6 +47,7 @@ bool dat_snoop_incoming_arp_reply(struct bat_priv *bat_priv,
 				  struct sk_buff *skb, int hdr_size);
 bool arp_drop_broadcast_packet(struct bat_priv *bat_priv,
 			       struct forw_packet *forw_packet);
+void arp_change_timeout(struct net_device *soft_iface, const char *name);
 
 /* hash function to choose an entry in a hash table of given size */
 /* hash algorithm from http://en.wikipedia.org/wiki/Hash_table */
diff --git a/main.h b/main.h
index 7384abb..9c8f4c1 100644
--- a/main.h
+++ b/main.h
@@ -71,6 +71,9 @@
 #define ARP_REQ_DELAY 250
 /* numbers of originator to contact for any PUT/GET DHT operation */
 #define DHT_CANDIDATES_NUM 3
+/* Factor which default ARP timeout values of the soft_iface table are
+ * multiplied by */
+#define ARP_TIMEOUT_FACTOR 4
 
 #define LOG_BUF_LEN 8192	  /* has to be a power of 2 */
 
diff --git a/soft-interface.c b/soft-interface.c
index f27ea7e..766d073 100644
--- a/soft-interface.c
+++ b/soft-interface.c
@@ -382,6 +382,8 @@ struct net_device *softif_create(const char *name)
 		goto free_soft_iface;
 	}
 
+	arp_change_timeout(soft_iface, name);
+
 	bat_priv = netdev_priv(soft_iface);
 
 	atomic_set(&bat_priv->aggregated_ogms, 1);
-- 
1.7.3.4

[B.A.T.M.A.N.] [PATCHv6 7/7] batman-adv: Distributed ARP Table - add compile option

[Antonio Quartulli]

This patch makes it possible to decide whether to include DAT within the
batman-adv binary or not.
It is extremely useful when the user wants to reduce the size of the resulting
module by cutting off any not needed feature.

Signed-off-by: Antonio Quartulli <ordex@autistici.org>
---
 Makefile                |    2 ++
 Makefile.kbuild         |    2 +-
 README.external         |    1 +
 distributed-arp-table.h |   15 +++++++++++++++
 gen-compat-autoconf.sh  |    1 +
 send.c                  |    2 --
 types.h                 |    8 ++++++++
 7 files changed, 28 insertions(+), 3 deletions(-)

diff --git a/Makefile b/Makefile
index 08f8c39..ac84fba 100644
--- a/Makefile
+++ b/Makefile
@@ -23,6 +23,8 @@
 export CONFIG_BATMAN_ADV_DEBUG=n
 # B.A.T.M.A.N. bridge loop avoidance:
 export CONFIG_BATMAN_ADV_BLA=y
+# B.A.T.M.A.N. distributed ARP table:
+export CONFIG_BATMAN_ADV_DAT=y
 
 PWD:=$(shell pwd)
 KERNELPATH ?= /lib/modules/$(shell uname -r)/build
diff --git a/Makefile.kbuild b/Makefile.kbuild
index 84955b3..ad002cd 100644
--- a/Makefile.kbuild
+++ b/Makefile.kbuild
@@ -24,7 +24,7 @@ batman-adv-y += bat_iv_ogm.o
 batman-adv-y += bat_sysfs.o
 batman-adv-y += bitarray.o
 batman-adv-$(CONFIG_BATMAN_ADV_BLA) += bridge_loop_avoidance.o
-batman-adv-y += distributed-arp-table.o
+batman-adv-$(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
diff --git a/README.external b/README.external
index 1eff0f9..95a7eb4 100644
--- a/README.external
+++ b/README.external
@@ -37,6 +37,7 @@ module).  Available  options  and  their    possible   values are
 
  * CONFIG_BATMAN_ADV_DEBUG=[y|n*] (B.A.T.M.A.N. debugging)
  * CONFIG_BATMAN_ADV_BLA=[y*|n] (B.A.T.M.A.N. bridge loop avoidance)
+ * CONFIG_BATMAN_ADV_DAT=[y*|n] (B.A.T.M.A.N. Distributed ARP Table)
 
 e.g., debugging can be enabled by
 
diff --git a/distributed-arp-table.h b/distributed-arp-table.h
index bf4390d..2e37e3f 100644
--- a/distributed-arp-table.h
+++ b/distributed-arp-table.h
@@ -22,6 +22,8 @@
 #ifndef _NET_BATMAN_ADV_ARP_H_
 #define _NET_BATMAN_ADV_ARP_H_
 
+#ifdef CONFIG_BATMAN_ADV_DAT
+
 #include "types.h"
 #include "originator.h"
 
@@ -84,4 +86,17 @@ static inline void dat_init_own_dht_addr(struct bat_priv *bat_priv,
 					    DAT_ADDR_MAX);
 }
 
+#else
+
+#define dat_snoop_outgoing_arp_request(...) (0)
+#define dat_snoop_incoming_arp_request(...) (0)
+#define dat_snoop_outgoing_arp_reply(...)
+#define dat_snoop_incoming_arp_reply(...) (0)
+#define arp_drop_broadcast_packet(...) (0)
+#define arp_change_timeout(...)
+#define dat_init_orig_node_dht_addr(...)
+#define dat_init_own_dht_addr(...)
+
+#endif /* CONFIG_BATMAN_ADV_DAT */
+
 #endif /* _NET_BATMAN_ADV_ARP_H_ */
diff --git a/gen-compat-autoconf.sh b/gen-compat-autoconf.sh
index 7cf621b..33de95d 100755
--- a/gen-compat-autoconf.sh
+++ b/gen-compat-autoconf.sh
@@ -38,6 +38,7 @@ gen_config() {
 # write config variables
 gen_config 'CONFIG_BATMAN_ADV_DEBUG' ${CONFIG_BATMAN_ADV_DEBUG:="n"} >> "${TMP}"
 gen_config 'CONFIG_BATMAN_ADV_BLA' ${CONFIG_BATMAN_ADV_BLA:="y"} >> "${TMP}"
+gen_config 'CONFIG_BATMAN_ADV_DAT' ${CONFIG_BATMAN_ADV_DAT:="n"} >> "${TMP}"
 
 # only regenerate compat-autoconf.h when config was changed
 diff "${TMP}" "${TARGET}" > /dev/null 2>&1 || cp "${TMP}" "${TARGET}"
diff --git a/send.c b/send.c
index 8bd31bb..65c13b2 100644
--- a/send.c
+++ b/send.c
@@ -30,8 +30,6 @@
 #include "gateway_common.h"
 #include "originator.h"
 
-#include <net/arp.h>
-
 static void send_outstanding_bcast_packet(struct work_struct *work);
 
 /* send out an already prepared packet to the given address via the
diff --git a/types.h b/types.h
index 210ee3e..6a3cc88 100644
--- a/types.h
+++ b/types.h
@@ -27,6 +27,8 @@
 #include "packet.h"
 #include "bitarray.h"
 
+#ifdef CONFIG_BATMAN_ADV_DAT
+
 /*
  * dat_addr_t is the type used for all DHT addresses. If it is changed,
  * DAT_ADDR_MAX is changed as well.
@@ -35,6 +37,8 @@
  */
 #define dat_addr_t uint16_t
 
+#endif /* CONFIG_BATMAN_ADV_DAT */
+
 #define BAT_HEADER_LEN (ETH_HLEN + \
 	((sizeof(struct unicast_packet) > sizeof(struct bcast_packet) ? \
 	 sizeof(struct unicast_packet) : \
@@ -75,7 +79,9 @@ struct hard_iface {
 struct orig_node {
 	uint8_t orig[ETH_ALEN];
 	uint8_t primary_addr[ETH_ALEN];
+#ifdef CONFIG_BATMAN_ADV_DAT
 	dat_addr_t dht_addr;
+#endif
 	struct neigh_node __rcu *router; /* rcu protected pointer */
 	unsigned long *bcast_own;
 	uint8_t *bcast_own_sum;
@@ -230,7 +236,9 @@ struct bat_priv {
 	struct gw_node __rcu *curr_gw;  /* rcu protected pointer */
 	atomic_t gw_reselect;
 	struct hard_iface __rcu *primary_if;  /* rcu protected pointer */
+#ifdef CONFIG_BATMAN_ADV_DAT
 	dat_addr_t dht_addr;
+#endif
 	struct vis_info *my_vis_info;
 	struct bat_algo_ops *bat_algo_ops;
 };
-- 
1.7.3.4

Re: [B.A.T.M.A.N.] [PATCHv6 1/7] batman-adv: add UNICAST_4ADDR packet type

[Antonio Quartulli]

[-- Attachment #1: Type: text/plain, Size: 939 bytes --]

On Sat, Feb 18, 2012 at 05:38:47 +0100, Antonio Quartulli wrote:
> The current unicast packet type does not contain the orig source address. This
> patches add a new unicast packet (called UNICAST_4ADDR) which provides two new
> fields: the originator source address and the subtype (the type of the data
> contained in the packet payload). The former is useful to identify the node
> which injected the packet into the network and the latter is useful to avoid
> creating new unicast packet types in the future: a macro defining a new subtype
> will be enough.

Hello people,

I was wondering...what about adding a "payload_size" field to the new packet?

The handling code would probably be generic enough to handle any future change
leading to a more stable compat_version (instead of increasing its value
everyday).


Cheers,

-- 
Antonio Quartulli

..each of us alone is worth nothing..
Ernesto "Che" Guevara

[-- Attachment #2: Type: application/pgp-signature, Size: 490 bytes --]

Re: [B.A.T.M.A.N.] [PATCHv6 1/7] batman-adv: add UNICAST_4ADDR packet type

[Marek Lindner]

On Sunday, February 19, 2012 22:54:19 Antonio Quartulli wrote:
> I was wondering...what about adding a "payload_size" field to the new
> packet?
> 
> The handling code would probably be generic enough to handle any future
> change leading to a more stable compat_version (instead of increasing its
> value everyday).

What will be the purpose of the payload field ?

Regards,
Marek

Re: [B.A.T.M.A.N.] [PATCHv6 1/7] batman-adv: add UNICAST_4ADDR packet type

[Antonio Quartulli]

[-- Attachment #1: Type: text/plain, Size: 713 bytes --]

On Mon, Feb 20, 2012 at 01:43:35PM +0800, Marek Lindner wrote:
> On Sunday, February 19, 2012 22:54:19 Antonio Quartulli wrote:
> > I was wondering...what about adding a "payload_size" field to the new
> > packet?
> > 
> > The handling code would probably be generic enough to handle any future
> > change leading to a more stable compat_version (instead of increasing its
> > value everyday).
> 
> What will be the purpose of the payload field ?

After discussing this on IRC we came up with not introducing this field.
Actually we don't need it so it's not worth bloating the packet format now.

Cheers!


-- 
Antonio Quartulli

..each of us alone is worth nothing..
Ernesto "Che" Guevara

[-- Attachment #2: Type: application/pgp-signature, Size: 490 bytes --]