With this patch a multicast packet is not always simply flooded anymore, the behaviour for the following cases is changed to reduce unnecessary overhead:
If all nodes within the horizon of a certain node have signalized multicast listener announcement capability then an IPv6 multicast packet with a destination of IPv6 link-local scope (excluding ff02::1) coming from the upstream of this node...
* ...is dropped if there is no according multicast listener in the translation table, * ...is forwarded via unicast if there is a single node with interested multicast listeners * ...and otherwise still gets flooded.
Signed-off-by: Linus Lüssing linus.luessing@web.de --- multicast.c | 134 ++++++++++++++++++++++++++++++++++++++++++++++++++ multicast.h | 24 +++++++++ send.c | 10 ++-- send.h | 5 ++ soft-interface.c | 20 +++++++- sysfs-class-net-mesh | 9 ++++ sysfs.c | 6 +++ translation-table.c | 92 ++++++++++++++++++++++++++-------- translation-table.h | 2 + types.h | 7 +++ 10 files changed, 283 insertions(+), 26 deletions(-)
diff --git a/multicast.c b/multicast.c index fe9e1e3..66a4fe8 100644 --- a/multicast.c +++ b/multicast.c @@ -17,6 +17,7 @@ #include "originator.h" #include "hard-interface.h" #include "translation-table.h" +#include "multicast.h"
/** * batadv_mcast_mla_softif_get - get softif multicast listeners @@ -271,6 +272,139 @@ out: }
/** + * batadv_mcast_forw_mode_check_ipv6 - check for optimized forwarding potential + * @bat_priv: the bat priv with all the soft interface information + * @skb: the IPv6 packet to check + * + * Check whether the given IPv6 packet has the potential to + * be forwarded with a mode more optimal than classic flooding. + * + * If so then return 0. Otherwise -EINVAL is returned or -ENOMEM if we are + * out of memory. + */ +static int batadv_mcast_forw_mode_check_ipv6(struct batadv_priv *bat_priv, + struct sk_buff *skb) +{ + struct ipv6hdr *ip6hdr; + + /* We might fail due to out-of-memory -> drop it */ + if (!pskb_may_pull(skb, sizeof(struct ethhdr) + sizeof(*ip6hdr))) + return -ENOMEM; + + ip6hdr = ipv6_hdr(skb); + + /* TODO: Implement Multicast Router Discovery (RFC4286), + * then allow scope > link local, too + */ + if (IPV6_ADDR_MC_SCOPE(&ip6hdr->daddr) != + IPV6_ADDR_SCOPE_LINKLOCAL) + return -EINVAL; + + /* link-local-all-nodes multicast listeners behind a bridge are + * not snoopable (see RFC4541, section 3, paragraph 3) + */ + if (ipv6_addr_is_ll_all_nodes(&ip6hdr->daddr)) + return -EINVAL; + + return 0; +} + +/** + * batadv_mcast_forw_mode_check - check for optimized forwarding potential + * @bat_priv: the bat priv with all the soft interface information + * @skb: the multicast frame to check + * + * Check whether the given multicast ethernet frame has the potential to + * be forwarded with a mode more optimal than classic flooding. + * + * If so then return 0. Otherwise -EINVAL is returned or -ENOMEM if we are + * out of memory. + */ +static int batadv_mcast_forw_mode_check(struct batadv_priv *bat_priv, + struct sk_buff *skb) +{ + struct ethhdr *ethhdr = eth_hdr(skb); + + if (!atomic_read(&bat_priv->multicast_mode)) + return -EINVAL; + + if (atomic_read(&bat_priv->mcast.num_disabled)) + return -EINVAL; + + switch (ntohs(ethhdr->h_proto)) { + case ETH_P_IPV6: + return batadv_mcast_forw_mode_check_ipv6(bat_priv, skb); + default: + return -EINVAL; + } +} + +/** + * batadv_mcast_tt_count - count the number of matching TT entries + * @bat_priv: the bat priv with all the soft interface information + * @ethhdr: ethernet header of a packet + * @orig_node: an originator to be set to forward the skb to + * + * Set an orig_node matching the given ethhdr source address and increase its + * refcount or keep it unchanged if there is none. Finally return the total + * number of matching orig_nodes. + */ +static int batadv_mcast_tt_count(struct batadv_priv *bat_priv, + struct ethhdr *ethhdr, + struct batadv_orig_node **orig_node) +{ + int ret = batadv_tt_global_hash_count(bat_priv, ethhdr->h_dest, + BATADV_NO_FLAGS); + + if (!ret) + goto out; + + *orig_node = batadv_transtable_search(bat_priv, ethhdr->h_source, + ethhdr->h_dest, + BATADV_NO_FLAGS); + if (!*orig_node) + ret = 0; + +out: + return ret; +} + +/** + * batadv_mcast_forw_mode - check on how to forward a multicast packet + * @bat_priv: the bat priv with all the soft interface information + * @skb: The multicast packet to check + * @mcast_single_orig: an originator to be set to forward the skb to + * + * Return the forwarding mode as enum batadv_forw_mode and in case of + * BATADV_FORW_SINGLE set the mcast_single_orig to the single originator + * the skb should be forwarded to. + */ +enum batadv_forw_mode +batadv_mcast_forw_mode(struct batadv_priv *bat_priv, struct sk_buff *skb, + struct batadv_orig_node **mcast_single_orig) +{ + struct ethhdr *ethhdr = eth_hdr(skb); + int ret; + + ret = batadv_mcast_forw_mode_check(bat_priv, skb); + if (ret == -ENOMEM) + return BATADV_FORW_NONE; + else if (ret < 0) + return BATADV_FORW_ALL; + + ret = batadv_mcast_tt_count(bat_priv, ethhdr, mcast_single_orig); + + switch (ret) { + case 0: + return BATADV_FORW_NONE; + case 1: + return BATADV_FORW_SINGLE; + default: + return BATADV_FORW_ALL; + } +} + +/** * batadv_mcast_tvlv_ogm_handler_v1 - process incoming multicast tvlv container * @bat_priv: the bat priv with all the soft interface information * @orig: the orig_node of the ogm diff --git a/multicast.h b/multicast.h index d736335..61e631b 100644 --- a/multicast.h +++ b/multicast.h @@ -15,10 +15,28 @@ #ifndef _NET_BATMAN_ADV_MULTICAST_H_ #define _NET_BATMAN_ADV_MULTICAST_H_
+/** + * batadv_forw_mode - the way a packet should be forwarded as + * @BATADV_FORW_ALL: forward the packet to all nodes + * (currently via classic flooding) + * @BATADV_FORW_SINGLE: forward the packet to a single node + * (currently via the BATMAN unicast routing protocol) + * @BATADV_FORW_NONE: don't forward, drop it + */ +enum batadv_forw_mode { + BATADV_FORW_ALL, + BATADV_FORW_SINGLE, + BATADV_FORW_NONE, +}; + #ifdef CONFIG_BATMAN_ADV_MCAST
void batadv_mcast_mla_update(struct batadv_priv *bat_priv);
+enum batadv_forw_mode +batadv_mcast_forw_mode(struct batadv_priv *bat_priv, struct sk_buff *skb, + struct batadv_orig_node **mcast_single_orig); + void batadv_mcast_init(struct batadv_priv *bat_priv);
void batadv_mcast_free(struct batadv_priv *bat_priv); @@ -32,6 +50,12 @@ static inline void batadv_mcast_mla_update(struct batadv_priv *bat_priv) return; }
+static inline enum batadv_forw_mode +batadv_mcast_forw_mode(struct batadv_priv *bat_priv, struct sk_buff *skb) +{ + return BATADV_FORW_ALL; +} + static inline int batadv_mcast_init(struct batadv_priv *bat_priv) { return 0; diff --git a/send.c b/send.c index 31a8c19..ef4c93e 100644 --- a/send.c +++ b/send.c @@ -248,11 +248,11 @@ out: * * Returns NET_XMIT_DROP in case of error or NET_XMIT_SUCCESS otherwise. */ -static int batadv_send_skb_unicast(struct batadv_priv *bat_priv, - struct sk_buff *skb, int packet_type, - int packet_subtype, - struct batadv_orig_node *orig_node, - unsigned short vid) +int batadv_send_skb_unicast(struct batadv_priv *bat_priv, + struct sk_buff *skb, int packet_type, + int packet_subtype, + struct batadv_orig_node *orig_node, + unsigned short vid) { struct batadv_unicast_packet *unicast_packet; struct ethhdr *ethhdr; diff --git a/send.h b/send.h index aaddaa9..887b63f 100644 --- a/send.h +++ b/send.h @@ -36,6 +36,11 @@ bool batadv_send_skb_prepare_unicast_4addr(struct batadv_priv *bat_priv, struct sk_buff *skb, struct batadv_orig_node *orig_node, int packet_subtype); +int batadv_send_skb_unicast(struct batadv_priv *bat_priv, + struct sk_buff *skb, int packet_type, + int packet_subtype, + struct batadv_orig_node *orig_node, + unsigned short vid); int batadv_send_skb_via_tt_generic(struct batadv_priv *bat_priv, struct sk_buff *skb, int packet_type, int packet_subtype, uint8_t *dst_hint, diff --git a/soft-interface.c b/soft-interface.c index ef9d0aa..6b1eb75 100644 --- a/soft-interface.c +++ b/soft-interface.c @@ -32,6 +32,7 @@ #include <linux/ethtool.h> #include <linux/etherdevice.h> #include <linux/if_vlan.h> +#include "multicast.h" #include "bridge_loop_avoidance.h" #include "network-coding.h"
@@ -170,6 +171,8 @@ static int batadv_interface_tx(struct sk_buff *skb, unsigned short vid; uint32_t seqno; int gw_mode; + enum batadv_forw_mode forw_mode; + struct batadv_orig_node *mcast_single_orig = NULL;
if (atomic_read(&bat_priv->mesh_state) != BATADV_MESH_ACTIVE) goto dropped; @@ -247,9 +250,19 @@ static int batadv_interface_tx(struct sk_buff *skb, * directed to a DHCP server */ goto dropped; - }
send: + if (do_bcast && !is_broadcast_ether_addr(ethhdr->h_dest)) { + forw_mode = batadv_mcast_forw_mode(bat_priv, skb, + &mcast_single_orig); + if (forw_mode == BATADV_FORW_NONE) + goto dropped; + + if (forw_mode == BATADV_FORW_SINGLE) + do_bcast = false; + } + } + batadv_skb_set_priority(skb, 0);
/* ethernet packet should be broadcasted */ @@ -301,6 +314,10 @@ send: if (ret) goto dropped; ret = batadv_send_skb_via_gw(bat_priv, skb, vid); + } else if (mcast_single_orig) { + ret = batadv_send_skb_unicast(bat_priv, skb, + BATADV_UNICAST, 0, + mcast_single_orig, vid); } else { if (batadv_dat_snoop_outgoing_arp_request(bat_priv, skb)) @@ -694,6 +711,7 @@ static int batadv_softif_init_late(struct net_device *dev) #endif #ifdef CONFIG_BATMAN_ADV_MCAST bat_priv->mcast.flags = BATADV_NO_FLAGS; + atomic_set(&bat_priv->multicast_mode, 1); atomic_set(&bat_priv->mcast.num_disabled, 0); #endif atomic_set(&bat_priv->gw_mode, BATADV_GW_MODE_OFF); diff --git a/sysfs-class-net-mesh b/sysfs-class-net-mesh index 4793d3d..a8e918c 100644 --- a/sysfs-class-net-mesh +++ b/sysfs-class-net-mesh @@ -76,6 +76,15 @@ Description: is used to classify clients as "isolated" by the Extended Isolation feature.
+What: /sys/class/net/<mesh_iface>/mesh/multicast_mode +Date: June 2013 +Contact: Linus Lüssing linus.luessing@web.de +Description: + Indicates whether multicast optimizations are enabled + or disabled. If set to zero then all nodes in the + mesh are going to use classic flooding for any + multicast packet with no optimizations. + What: /sys/class/net/<mesh_iface>/mesh/network_coding Date: Nov 2012 Contact: Martin Hundeboll martin@hundeboll.net diff --git a/sysfs.c b/sysfs.c index e456bf6..1ebb0d9 100644 --- a/sysfs.c +++ b/sysfs.c @@ -539,6 +539,9 @@ BATADV_ATTR_SIF_UINT(gw_sel_class, S_IRUGO | S_IWUSR, 1, BATADV_TQ_MAX_VALUE, batadv_post_gw_reselect); static BATADV_ATTR(gw_bandwidth, S_IRUGO | S_IWUSR, batadv_show_gw_bwidth, batadv_store_gw_bwidth); +#ifdef CONFIG_BATMAN_ADV_MCAST +BATADV_ATTR_SIF_BOOL(multicast_mode, S_IRUGO | S_IWUSR, NULL); +#endif #ifdef CONFIG_BATMAN_ADV_DEBUG BATADV_ATTR_SIF_UINT(log_level, S_IRUGO | S_IWUSR, 0, BATADV_DBG_ALL, NULL); #endif @@ -558,6 +561,9 @@ static struct batadv_attribute *batadv_mesh_attrs[] = { #ifdef CONFIG_BATMAN_ADV_DAT &batadv_attr_distributed_arp_table, #endif +#ifdef CONFIG_BATMAN_ADV_MCAST + &batadv_attr_multicast_mode, +#endif &batadv_attr_fragmentation, &batadv_attr_routing_algo, &batadv_attr_gw_mode, diff --git a/translation-table.c b/translation-table.c index 6b5268d..1d858d5 100644 --- a/translation-table.c +++ b/translation-table.c @@ -193,6 +193,32 @@ batadv_tt_global_entry_free_ref(struct batadv_tt_global_entry *tt_global_entry) } }
+/** + * batadv_tt_global_hash_count - count the number of orig entries + * @hash: hash table containing the tt entries + * @addr: the mac address of the client to count entries for + * @vid: VLAN identifier + * + * Return the number of originators advertising the given address/data + * (excluding ourself). + */ +int batadv_tt_global_hash_count(struct batadv_priv *bat_priv, + const uint8_t *addr, unsigned short vid) +{ + struct batadv_tt_global_entry *tt_global_entry; + int count = 0; + + tt_global_entry = batadv_tt_global_hash_find(bat_priv, addr, vid); + if (!tt_global_entry) + goto out; + + count = atomic_read(&tt_global_entry->orig_list_count); + batadv_tt_global_entry_free_ref(tt_global_entry); + +out: + return count; +} + static void batadv_tt_orig_list_entry_free_rcu(struct rcu_head *rcu) { struct batadv_tt_orig_list_entry *orig_entry; @@ -1225,6 +1251,8 @@ batadv_tt_global_orig_entry_add(struct batadv_tt_global_entry *tt_global, hlist_add_head_rcu(&orig_entry->list, &tt_global->orig_list); spin_unlock_bh(&tt_global->list_lock); + atomic_inc(&tt_global->orig_list_count); + out: if (orig_entry) batadv_tt_orig_list_entry_free_ref(orig_entry); @@ -1298,6 +1326,7 @@ static bool batadv_tt_global_add(struct batadv_priv *bat_priv, common->added_at = jiffies;
INIT_HLIST_HEAD(&tt_global_entry->orig_list); + atomic_set(&tt_global_entry->orig_list_count, 0); spin_lock_init(&tt_global_entry->list_lock);
hash_added = batadv_hash_add(bat_priv->tt.global_hash, @@ -1563,6 +1592,25 @@ out: return 0; }
+/** + * batadv_tt_global_del_orig_entry - remove and free an orig_entry + * @tt_global_entry: the global entry to remove the orig_entry from + * @orig_entry: the orig entry to remove and free + * + * Remove an orig_entry from its list in the given tt_global_entry and + * free this orig_entry afterwards. + */ +static void +batadv_tt_global_del_orig_entry(struct batadv_tt_global_entry *tt_global_entry, + struct batadv_tt_orig_list_entry *orig_entry) +{ + batadv_tt_global_size_dec(orig_entry->orig_node, + tt_global_entry->common.vid); + atomic_dec(&tt_global_entry->orig_list_count); + hlist_del_rcu(&orig_entry->list); + batadv_tt_orig_list_entry_free_ref(orig_entry); +} + /* deletes the orig list of a tt_global_entry */ static void batadv_tt_global_del_orig_list(struct batadv_tt_global_entry *tt_global_entry) @@ -1573,20 +1621,26 @@ batadv_tt_global_del_orig_list(struct batadv_tt_global_entry *tt_global_entry)
spin_lock_bh(&tt_global_entry->list_lock); head = &tt_global_entry->orig_list; - hlist_for_each_entry_safe(orig_entry, safe, head, list) { - hlist_del_rcu(&orig_entry->list); - batadv_tt_global_size_dec(orig_entry->orig_node, - tt_global_entry->common.vid); - batadv_tt_orig_list_entry_free_ref(orig_entry); - } + hlist_for_each_entry_safe(orig_entry, safe, head, list) + batadv_tt_global_del_orig_entry(tt_global_entry, orig_entry); spin_unlock_bh(&tt_global_entry->list_lock); }
+/** + * batadv_tt_global_del_orig_node - remove orig_node from a global tt entry + * @bat_priv: the bat priv with all the soft interface information + * @tt_global_entry: the global entry to remove the orig_node from + * @orig_node: the originator announcing the client + * @message: message to append to the log on deletion + * + * Remove the given orig_node and its according orig_entry from the given + * global tt entry. + */ static void -batadv_tt_global_del_orig_entry(struct batadv_priv *bat_priv, - struct batadv_tt_global_entry *tt_global_entry, - struct batadv_orig_node *orig_node, - const char *message) +batadv_tt_global_del_orig_node(struct batadv_priv *bat_priv, + struct batadv_tt_global_entry *tt_global_entry, + struct batadv_orig_node *orig_node, + const char *message) { struct hlist_head *head; struct hlist_node *safe; @@ -1603,10 +1657,8 @@ batadv_tt_global_del_orig_entry(struct batadv_priv *bat_priv, orig_node->orig, tt_global_entry->common.addr, BATADV_PRINT_VID(vid), message); - hlist_del_rcu(&orig_entry->list); - batadv_tt_global_size_dec(orig_node, - tt_global_entry->common.vid); - batadv_tt_orig_list_entry_free_ref(orig_entry); + batadv_tt_global_del_orig_entry(tt_global_entry, + orig_entry); } } spin_unlock_bh(&tt_global_entry->list_lock); @@ -1648,8 +1700,8 @@ batadv_tt_global_del_roaming(struct batadv_priv *bat_priv, /* there is another entry, we can simply delete this * one and can still use the other one. */ - batadv_tt_global_del_orig_entry(bat_priv, tt_global_entry, - orig_node, message); + batadv_tt_global_del_orig_node(bat_priv, tt_global_entry, + orig_node, message); }
/** @@ -1675,8 +1727,8 @@ static void batadv_tt_global_del(struct batadv_priv *bat_priv, goto out;
if (!roaming) { - batadv_tt_global_del_orig_entry(bat_priv, tt_global_entry, - orig_node, message); + batadv_tt_global_del_orig_node(bat_priv, tt_global_entry, + orig_node, message);
if (hlist_empty(&tt_global_entry->orig_list)) batadv_tt_global_free(bat_priv, tt_global_entry, @@ -1759,8 +1811,8 @@ void batadv_tt_global_del_orig(struct batadv_priv *bat_priv, struct batadv_tt_global_entry, common);
- batadv_tt_global_del_orig_entry(bat_priv, tt_global, - orig_node, message); + batadv_tt_global_del_orig_node(bat_priv, tt_global, + orig_node, message);
if (hlist_empty(&tt_global->orig_list)) { vid = tt_global->common.vid; diff --git a/translation-table.h b/translation-table.h index 20a1d78..ad84d7b 100644 --- a/translation-table.h +++ b/translation-table.h @@ -29,6 +29,8 @@ int batadv_tt_global_seq_print_text(struct seq_file *seq, void *offset); void batadv_tt_global_del_orig(struct batadv_priv *bat_priv, struct batadv_orig_node *orig_node, int32_t match_vid, const char *message); +int batadv_tt_global_hash_count(struct batadv_priv *bat_priv, + const uint8_t *addr, unsigned short vid); struct batadv_orig_node *batadv_transtable_search(struct batadv_priv *bat_priv, const uint8_t *src, const uint8_t *addr, diff --git a/types.h b/types.h index 96ee0d2..c28fc4a 100644 --- a/types.h +++ b/types.h @@ -696,6 +696,8 @@ struct batadv_softif_vlan { * enabled * @distributed_arp_table: bool indicating whether distributed ARP table is * enabled + * @multicast_mode: Enable or disable multicast optimizations on this node's + * sender/originating side * @gw_mode: gateway operation: off, client or server (see batadv_gw_modes) * @gw_sel_class: gateway selection class (applies if gw_mode client) * @orig_interval: OGM broadcast interval in milliseconds @@ -746,6 +748,9 @@ struct batadv_priv { #ifdef CONFIG_BATMAN_ADV_DAT atomic_t distributed_arp_table; #endif +#ifdef CONFIG_BATMAN_ADV_MCAST + atomic_t multicast_mode; +#endif atomic_t gw_mode; atomic_t gw_sel_class; atomic_t orig_interval; @@ -909,12 +914,14 @@ struct batadv_tt_local_entry { * struct batadv_tt_global_entry - translation table global entry data * @common: general translation table data * @orig_list: list of orig nodes announcing this non-mesh client + * @orig_list_count: number of items in the orig_list * @list_lock: lock protecting orig_list * @roam_at: time at which TT_GLOBAL_ROAM was set */ struct batadv_tt_global_entry { struct batadv_tt_common_entry common; struct hlist_head orig_list; + atomic_t orig_list_count; spinlock_t list_lock; /* protects orig_list */ unsigned long roam_at; };