B.A.T.M.A.N August 2013

b.a.t.m.a.n@lists.open-mesh.org

12 participants
45 discussions

[B.A.T.M.A.N.] [PATCH 1/2] batman-adv: Fix broadcast+ogm queue purging race condition
by Linus Lüssing 22 Jul '16

22 Jul '16

So far on purging broadcast and ogm queues we temporarily give up the spin lock of these queues to be able to cancel any scheduled forwarding work. However this is unsafe and can lead to a general protection error in batadv_purge_outstanding_packets(). With this patch we split the queue purging into two steps: First removing forward packets from those queues and signaling the cancelation. Secondly, we are actively canceling any scheduled forwarding, wait for any running forwarding to finish and only free a forw_packet afterwards. Signed-off-by: Linus Lüssing <linus.luessing(a)web.de> --- Fixes issue #168 send.c | 117 ++++++++++++++++++++++++++++++++++++++------------------------- types.h | 1 + 2 files changed, 71 insertions(+), 47 deletions(-) diff --git a/send.c b/send.c index 0a0bb45..f93476b 100644 --- a/send.c +++ b/send.c @@ -245,6 +245,10 @@ static void batadv_send_outstanding_bcast_packet(struct work_struct *work) bat_priv = netdev_priv(soft_iface); spin_lock_bh(&bat_priv->forw_bcast_list_lock); + if (hlist_unhashed(&forw_packet->list)) { + spin_unlock_bh(&bat_priv->forw_bcast_list_lock); + return; + } hlist_del(&forw_packet->list); spin_unlock_bh(&bat_priv->forw_bcast_list_lock); @@ -293,6 +297,10 @@ void batadv_send_outstanding_bat_ogm_packet(struct work_struct *work) delayed_work); bat_priv = netdev_priv(forw_packet->if_incoming->soft_iface); spin_lock_bh(&bat_priv->forw_bat_list_lock); + if (hlist_unhashed(&forw_packet->list)) { + spin_unlock_bh(&bat_priv->forw_bat_list_lock); + return; + } hlist_del(&forw_packet->list); spin_unlock_bh(&bat_priv->forw_bat_list_lock); @@ -316,13 +324,68 @@ out: batadv_forw_packet_free(forw_packet); } +/** + * batadv_cancel_packets - Cancels a list of forward packets + * @forw_list: The to be canceled forward packets + * @canceled_list: The backup list. + * + * This canceles any scheduled forwarding packet tasks in the provided + * forw_list. The packets are being moved from the forw_list to the + * canceled_list afterwards to unhash the forward packet list pointer, + * allowing any already running task to notice the cancelation. + */ +static void batadv_cancel_packets(struct hlist_head *forw_list, + struct hlist_head *canceled_list, + const struct batadv_hard_iface *hard_iface) +{ + struct batadv_forw_packet *forw_packet; + struct hlist_node *tmp_node, *safe_tmp_node; + + hlist_for_each_entry_safe(forw_packet, tmp_node, safe_tmp_node, + forw_list, list) { + /* if purge_outstanding_packets() was called with an argument + * we delete only packets belonging to the given interface + */ + if ((hard_iface) && + (forw_packet->if_incoming != hard_iface)) + continue; + + hlist_del_init(&forw_packet->list); + hlist_add_head(&forw_packet->canceled_list, canceled_list); + } +} + +/** + * batadv_canceled_packets_free - Frees canceled forward packets + * @head: A list of to be freed forw_packets + * + * This function canceles the scheduling of any packet in the provided list, + * waits for any possibly running packet forwarding thread to finish and + * finally, safely frees this forward packet. + * + * This function might sleep. + */ +static void batadv_canceled_packets_free(struct hlist_head *head) +{ + struct batadv_forw_packet *forw_packet; + struct hlist_node *tmp_node, *safe_tmp_node; + + hlist_for_each_entry_safe(forw_packet, tmp_node, safe_tmp_node, head, + canceled_list) { + cancel_delayed_work_sync(&forw_packet->delayed_work); + + hlist_del(&forw_packet->canceled_list); + batadv_forw_packet_free(forw_packet); + } +} + void batadv_purge_outstanding_packets(struct batadv_priv *bat_priv, const struct batadv_hard_iface *hard_iface) { - struct batadv_forw_packet *forw_packet; - struct hlist_node *tmp_node, *safe_tmp_node; - bool pending; + struct hlist_head head; + + INIT_HLIST_HEAD(&head); if (hard_iface) batadv_dbg(BATADV_DBG_BATMAN, bat_priv, @@ -334,53 +397,13 @@ batadv_purge_outstanding_packets(struct batadv_priv *bat_priv, /* free bcast list */ spin_lock_bh(&bat_priv->forw_bcast_list_lock); - hlist_for_each_entry_safe(forw_packet, tmp_node, safe_tmp_node, - &bat_priv->forw_bcast_list, list) { - /* if purge_outstanding_packets() was called with an argument - * we delete only packets belonging to the given interface - */ - if ((hard_iface) && - (forw_packet->if_incoming != hard_iface)) - continue; - - spin_unlock_bh(&bat_priv->forw_bcast_list_lock); - - /* batadv_send_outstanding_bcast_packet() will lock the list to - * delete the item from the list - */ - pending = cancel_delayed_work_sync(&forw_packet->delayed_work); - spin_lock_bh(&bat_priv->forw_bcast_list_lock); - - if (pending) { - hlist_del(&forw_packet->list); - batadv_forw_packet_free(forw_packet); - } - } + batadv_cancel_packets(&bat_priv->forw_bcast_list, &head, hard_iface); spin_unlock_bh(&bat_priv->forw_bcast_list_lock); /* free batman packet list */ spin_lock_bh(&bat_priv->forw_bat_list_lock); - hlist_for_each_entry_safe(forw_packet, tmp_node, safe_tmp_node, - &bat_priv->forw_bat_list, list) { - /* if purge_outstanding_packets() was called with an argument - * we delete only packets belonging to the given interface - */ - if ((hard_iface) && - (forw_packet->if_incoming != hard_iface)) - continue; - - spin_unlock_bh(&bat_priv->forw_bat_list_lock); - - /* send_outstanding_bat_packet() will lock the list to - * delete the item from the list - */ - pending = cancel_delayed_work_sync(&forw_packet->delayed_work); - spin_lock_bh(&bat_priv->forw_bat_list_lock); - - if (pending) { - hlist_del(&forw_packet->list); - batadv_forw_packet_free(forw_packet); - } - } + batadv_cancel_packets(&bat_priv->forw_bat_list, &head, hard_iface); spin_unlock_bh(&bat_priv->forw_bat_list_lock); + + batadv_canceled_packets_free(&head); } diff --git a/types.h b/types.h index aba8364..f62a35f 100644 --- a/types.h +++ b/types.h @@ -853,6 +853,7 @@ struct batadv_skb_cb { */ struct batadv_forw_packet { struct hlist_node list; + struct hlist_node canceled_list; unsigned long send_time; uint8_t own; struct sk_buff *skb; -- 1.7.10.4

4 12

[B.A.T.M.A.N.] [PATCH 1/2] batman-adv: Fix broadcast queue limit on a removed interface
by Linus Lüssing 09 Apr '16

09 Apr '16

When removing a single interface while a broadcast or ogm packet is still pending then we will free the forward packet without releasing the queue slots again. This patch is supposed to fix this issue. Signed-off-by: Linus Lüssing <linus.luessing(a)web.de> --- send.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/send.c b/send.c index ed7072a..2d539d6 100644 --- a/send.c +++ b/send.c @@ -356,6 +356,9 @@ batadv_purge_outstanding_packets(struct batadv_priv *bat_priv, if (pending) { hlist_del(&forw_packet->list); + if (!forw_packet->own) + atomic_inc(&bat_priv->bcast_queue_left); + batadv_forw_packet_free(forw_packet); } } @@ -382,6 +385,9 @@ batadv_purge_outstanding_packets(struct batadv_priv *bat_priv, if (pending) { hlist_del(&forw_packet->list); + if (!forw_packet->own) + atomic_inc(&bat_priv->batman_queue_left); + batadv_forw_packet_free(forw_packet); } } -- 1.7.10.4

2 5

[B.A.T.M.A.N.] [RFC 1/6] batman-adv: Generalize DAT in order to support any type of data, not only IPv4
by mihail.costea2005＠gmail.com 20 Mar '16

20 Mar '16

From: Mihail Costea <mihail.costea90(a)gmail.com> Mades DAT support more types by making its data a void*, adding type field to dat_entry and adding data_type to necessary functions. This change is needed in order to make DAT support any type of data, like IPv6 too. Adds generic function for transforming DAT data to string. The function is used in order to avoid defining different debug messages for different DAT data types. For example, if we had IPv6 as a DAT data, then "%pI4" should be "%pI6c", but all the other text of the debug message would be the same. Also everything is memorized in a struct in order to avoid further switch cases for all types. Signed-off-by: Mihail Costea <mihail.costea90(a)gmail.com> Signed-off-by: Stefan Popa <Stefan.A.Popa(a)intel.com> Reviewed-by: Stefan Popa <Stefan.A.Popa(a)intel.com> --- distributed-arp-table.c | 197 +++++++++++++++++++++++++++++++++++------------ distributed-arp-table.h | 1 + types.h | 24 +++++- 3 files changed, 169 insertions(+), 53 deletions(-) diff --git a/distributed-arp-table.c b/distributed-arp-table.c index f2543c2..90565d0 100644 --- a/distributed-arp-table.c +++ b/distributed-arp-table.c @@ -31,9 +31,32 @@ #include "types.h" #include "translation-table.h" +static struct batadv_dat_type_info batadv_dat_types_info[] = { + { + .size = sizeof(__be32), + .str_fmt = "%pI4", + }, +}; + static void batadv_dat_purge(struct work_struct *work); /** + * batadv_dat_data_to_str: transforms DAT data to string + * @data: the DAT data + * @type: type of data + * @buf: the buf where the data string is stored + * @buf_len: buf length + * + * Returns buf. + */ +static char *batadv_dat_data_to_str(void *data, uint8_t type, + char *buf, size_t buf_len) +{ + snprintf(buf, buf_len, batadv_dat_types_info[type].str_fmt, data); +return buf; +} + +/** * batadv_dat_start_timer - initialise the DAT periodic worker * @bat_priv: the bat priv with all the soft interface information */ @@ -45,6 +68,19 @@ static void batadv_dat_start_timer(struct batadv_priv *bat_priv) } /** + * batadv_dat_entry_free_ref_rcu - free a dat entry using its rcu + * @rcu: the dat entry rcu + */ +static void batadv_dat_entry_free_ref_rcu(struct rcu_head *rcu) +{ + struct batadv_dat_entry *dat_entry; + + dat_entry = container_of(rcu, struct batadv_dat_entry, rcu); + kfree(dat_entry->data); + kfree(dat_entry); +} + +/** * batadv_dat_entry_free_ref - decrement the dat_entry refcounter and possibly * free it * @dat_entry: the entry to free @@ -52,7 +88,7 @@ static void batadv_dat_start_timer(struct batadv_priv *bat_priv) 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); + call_rcu(&dat_entry->rcu, batadv_dat_entry_free_ref_rcu); } /** @@ -136,12 +172,21 @@ static void batadv_dat_purge(struct work_struct *work) * * Returns 1 if the two entries are the same, 0 otherwise. */ -static int batadv_compare_dat(const struct hlist_node *node, const void *data2) +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); + struct batadv_dat_entry *dat_entry1 = + container_of(node, struct batadv_dat_entry, + hash_entry); + struct batadv_dat_entry *dat_entry2 = + container_of(data2, + struct batadv_dat_entry, data); + size_t data_size = batadv_dat_types_info[dat_entry1->type].size; - return (memcmp(data1, data2, sizeof(__be32)) == 0 ? 1 : 0); + if (dat_entry1->type != dat_entry2->type) + return 0; + + return (memcmp(dat_entry1->data, dat_entry2->data, + data_size) == 0 ? 1 : 0); } /** @@ -198,8 +243,9 @@ static __be32 batadv_arp_ip_dst(struct sk_buff *skb, int hdr_size) } /** - * batadv_hash_dat - compute the hash value for an IP address + * batadv_hash_dat - compute the hash value for a DAT data * @data: data to hash + * @data_type: type of data * @size: size of the hash table * * Returns the selected index in the hash table for the given data. @@ -209,7 +255,8 @@ static uint32_t batadv_hash_dat(const void *data, uint32_t size) uint32_t hash = 0; const struct batadv_dat_entry *dat = data; - hash = batadv_hash_bytes(hash, &dat->ip, sizeof(dat->ip)); + hash = batadv_hash_bytes(hash, dat->data, + batadv_dat_types_info[dat->type].size); hash = batadv_hash_bytes(hash, &dat->vid, sizeof(dat->vid)); hash += (hash << 3); @@ -223,32 +270,40 @@ static uint32_t batadv_hash_dat(const void *data, uint32_t size) * batadv_dat_entry_hash_find - look for a given dat_entry in the local hash * table * @bat_priv: the bat priv with all the soft interface information - * @ip: search key + * @data: search key + * @data_type: type of data * @vid: VLAN identifier * * 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, - unsigned short vid) +batadv_dat_entry_hash_find(struct batadv_priv *bat_priv, void *data, + uint8_t data_type, unsigned short vid) { struct hlist_head *head; struct batadv_dat_entry to_find, *dat_entry, *dat_entry_tmp = NULL; struct batadv_hashtable *hash = bat_priv->dat.hash; - uint32_t index; + uint32_t index, data_size = batadv_dat_types_info[data_type].size; if (!hash) return NULL; - to_find.ip = ip; + to_find.data = kmalloc(data_size, GFP_ATOMIC); + if (!to_find.data) + return NULL; + memcpy(to_find.data, data, data_size); + to_find.type = data_type; to_find.vid = vid; index = batadv_hash_dat(&to_find, hash->size); head = &hash->table[index]; + kfree(to_find.data); rcu_read_lock(); hlist_for_each_entry_rcu(dat_entry, head, hash_entry) { - if (dat_entry->ip != ip) + if (dat_entry->type != data_type) + continue; + if (memcmp(dat_entry->data, data, data_size)) continue; if (!atomic_inc_not_zero(&dat_entry->refcount)) @@ -265,25 +320,30 @@ batadv_dat_entry_hash_find(struct batadv_priv *bat_priv, __be32 ip, /** * 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 + * @data: the data to add/edit + * @data_type: type of the data added to DAT + * @mac_addr: mac address to assign to the given data * @vid: VLAN identifier */ -static void batadv_dat_entry_add(struct batadv_priv *bat_priv, __be32 ip, - uint8_t *mac_addr, unsigned short vid) +static void batadv_dat_entry_add(struct batadv_priv *bat_priv, void *data, + uint8_t data_type, uint8_t *mac_addr, + unsigned short vid) { struct batadv_dat_entry *dat_entry; int hash_added; + char dbg_data[BATADV_DAT_DATA_MAX_LEN]; + size_t data_size = batadv_dat_types_info[data_type].size; - dat_entry = batadv_dat_entry_hash_find(bat_priv, ip, vid); + dat_entry = batadv_dat_entry_hash_find(bat_priv, data, data_type, vid); /* 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 (vid: %u)\n", - &dat_entry->ip, dat_entry->mac_addr, vid); + batadv_dbg(BATADV_DBG_DAT, bat_priv, "Entry updated: %s %pM (vid: %u)\n", + batadv_dat_data_to_str(dat_entry->data, data_type, + dbg_data, sizeof(dbg_data)), + dat_entry->mac_addr, vid); goto out; } @@ -291,7 +351,12 @@ static void batadv_dat_entry_add(struct batadv_priv *bat_priv, __be32 ip, if (!dat_entry) goto out; - dat_entry->ip = ip; + dat_entry->data = kmalloc(data_size, GFP_ATOMIC); + if (!dat_entry->data) + goto out; + memcpy(dat_entry->data, data, data_size); + + dat_entry->type = data_type; dat_entry->vid = vid; memcpy(dat_entry->mac_addr, mac_addr, ETH_ALEN); dat_entry->last_update = jiffies; @@ -307,8 +372,10 @@ static void batadv_dat_entry_add(struct batadv_priv *bat_priv, __be32 ip, goto out; } - batadv_dbg(BATADV_DBG_DAT, bat_priv, "New entry added: %pI4 %pM (vid: %u)\n", - &dat_entry->ip, dat_entry->mac_addr, vid); + batadv_dbg(BATADV_DBG_DAT, bat_priv, "New entry added: %s %pM (vid: %u)\n", + batadv_dat_data_to_str(dat_entry->data, data_type, + dbg_data, sizeof(dbg_data)), + dat_entry->mac_addr, vid); out: if (dat_entry) @@ -520,7 +587,8 @@ static void batadv_choose_next_candidate(struct batadv_priv *bat_priv, * batadv_dat_select_candidates - select 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 + * @data: data to look up in the DHT + * @data_type: type of data * * 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 @@ -529,11 +597,15 @@ static void batadv_choose_next_candidate(struct batadv_priv *bat_priv, * 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) +batadv_dat_select_candidates(struct batadv_priv *bat_priv, void *data, + uint8_t data_type) { int select; - batadv_dat_addr_t last_max = BATADV_DAT_ADDR_MAX, ip_key; + batadv_dat_addr_t last_max = BATADV_DAT_ADDR_MAX, data_key; struct batadv_dat_candidate *res; + struct batadv_dat_entry to_find; + char dbg_data[BATADV_DAT_DATA_MAX_LEN]; + size_t data_size = batadv_dat_types_info[data_type].size; if (!bat_priv->orig_hash) return NULL; @@ -542,15 +614,23 @@ batadv_dat_select_candidates(struct batadv_priv *bat_priv, __be32 ip_dst) if (!res) return NULL; - ip_key = (batadv_dat_addr_t)batadv_hash_dat(&ip_dst, - BATADV_DAT_ADDR_MAX); + to_find.data = kmalloc(data_size, GFP_ATOMIC); + if (!to_find.data) + return NULL; + memcpy(to_find.data, data, data_size); + to_find.type = data_type; + data_key = (batadv_dat_addr_t)batadv_hash_dat(&to_find, + BATADV_DAT_ADDR_MAX); + kfree(to_find.data); batadv_dbg(BATADV_DBG_DAT, bat_priv, - "dat_select_candidates(): IP=%pI4 hash(IP)=%u\n", &ip_dst, - ip_key); + "dat_select_candidates(): DATA=%s hash(DATA)=%u\n", + batadv_dat_data_to_str(data, data_type, dbg_data, + sizeof(dbg_data)), + data_key); for (select = 0; select < BATADV_DAT_CANDIDATES_NUM; select++) - batadv_choose_next_candidate(bat_priv, res, select, ip_key, + batadv_choose_next_candidate(bat_priv, res, select, data_key, &last_max); return res; @@ -560,7 +640,8 @@ batadv_dat_select_candidates(struct batadv_priv *bat_priv, __be32 ip_dst) * 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 + * @data: the DHT key + * @data_type: type of data * @packet_subtype: unicast4addr packet subtype to use * * This function copies the skb with pskb_copy() and is sent as unicast packet @@ -570,8 +651,8 @@ batadv_dat_select_candidates(struct batadv_priv *bat_priv, __be32 ip_dst) * otherwise. */ static bool batadv_dat_send_data(struct batadv_priv *bat_priv, - struct sk_buff *skb, __be32 ip, - int packet_subtype) + struct sk_buff *skb, void *data, + uint8_t data_type, int packet_subtype) { int i; bool ret = false; @@ -579,12 +660,15 @@ static bool batadv_dat_send_data(struct batadv_priv *bat_priv, struct batadv_neigh_node *neigh_node = NULL; struct sk_buff *tmp_skb; struct batadv_dat_candidate *cand; + char dbg_data[BATADV_DAT_DATA_MAX_LEN]; - cand = batadv_dat_select_candidates(bat_priv, ip); + cand = batadv_dat_select_candidates(bat_priv, data, data_type); if (!cand) goto out; - batadv_dbg(BATADV_DBG_DAT, bat_priv, "DHT_SEND for %pI4\n", &ip); + batadv_dbg(BATADV_DBG_DAT, bat_priv, "DHT_SEND for %s\n", + batadv_dat_data_to_str(data, data_type, dbg_data, + sizeof(dbg_data))); for (i = 0; i < BATADV_DAT_CANDIDATES_NUM; i++) { if (cand[i].type == BATADV_DAT_CANDIDATE_NOT_FOUND) @@ -754,6 +838,7 @@ int batadv_dat_cache_seq_print_text(struct seq_file *seq, void *offset) unsigned long last_seen_jiffies; int last_seen_msecs, last_seen_secs, last_seen_mins; uint32_t i; + char dbg_data[BATADV_DAT_DATA_MAX_LEN]; primary_if = batadv_seq_print_text_primary_if_get(seq); if (!primary_if) @@ -774,8 +859,12 @@ int batadv_dat_cache_seq_print_text(struct seq_file *seq, void *offset) 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, + seq_printf(seq, " * %15s %14pM %6i:%02i\n", + batadv_dat_data_to_str(dat_entry->data, + dat_entry->type, + dbg_data, + sizeof(dbg_data)), + dat_entry->mac_addr, last_seen_mins, last_seen_secs); } rcu_read_unlock(); @@ -926,9 +1015,10 @@ bool batadv_dat_snoop_outgoing_arp_request(struct batadv_priv *bat_priv, 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, vid); + batadv_dat_entry_add(bat_priv, &ip_src, BATADV_DAT_IPV4, hw_src, vid); - dat_entry = batadv_dat_entry_hash_find(bat_priv, ip_dst, vid); + dat_entry = batadv_dat_entry_hash_find(bat_priv, &ip_dst, + BATADV_DAT_IPV4, vid); if (dat_entry) { /* If the ARP request is destined for a local client the local * client will answer itself. DAT would only generate a @@ -962,7 +1052,8 @@ bool batadv_dat_snoop_outgoing_arp_request(struct batadv_priv *bat_priv, ret = true; } else { /* Send the request to the DHT */ - ret = batadv_dat_send_data(bat_priv, skb, ip_dst, + ret = batadv_dat_send_data(bat_priv, skb, &ip_dst, + BATADV_DAT_IPV4, BATADV_P_DAT_DHT_GET); } out: @@ -1008,9 +1099,10 @@ bool batadv_dat_snoop_incoming_arp_request(struct batadv_priv *bat_priv, batadv_dbg_arp(bat_priv, skb, type, hdr_size, "Parsing incoming ARP REQUEST"); - batadv_dat_entry_add(bat_priv, ip_src, hw_src, vid); + batadv_dat_entry_add(bat_priv, &ip_src, BATADV_DAT_IPV4, hw_src, vid); - dat_entry = batadv_dat_entry_hash_find(bat_priv, ip_dst, vid); + dat_entry = batadv_dat_entry_hash_find(bat_priv, &ip_dst, + BATADV_DAT_IPV4, vid); if (!dat_entry) goto out; @@ -1074,14 +1166,16 @@ void batadv_dat_snoop_outgoing_arp_reply(struct batadv_priv *bat_priv, hw_dst = batadv_arp_hw_dst(skb, hdr_size); ip_dst = batadv_arp_ip_dst(skb, hdr_size); - batadv_dat_entry_add(bat_priv, ip_src, hw_src, vid); - batadv_dat_entry_add(bat_priv, ip_dst, hw_dst, vid); + batadv_dat_entry_add(bat_priv, &ip_src, BATADV_DAT_IPV4, hw_src, vid); + batadv_dat_entry_add(bat_priv, &ip_dst, BATADV_DAT_IPV4, hw_dst, vid); /* Send the ARP reply to the candidates for both the IP addresses that * the node obtained from 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_send_data(bat_priv, skb, &ip_src, BATADV_DAT_IPV4, + BATADV_P_DAT_DHT_PUT); + batadv_dat_send_data(bat_priv, skb, &ip_dst, BATADV_DAT_IPV4, + BATADV_P_DAT_DHT_PUT); } /** * batadv_dat_snoop_incoming_arp_reply - snoop the ARP reply and fill the local @@ -1119,8 +1213,8 @@ bool batadv_dat_snoop_incoming_arp_reply(struct batadv_priv *bat_priv, /* 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, vid); - batadv_dat_entry_add(bat_priv, ip_dst, hw_dst, vid); + batadv_dat_entry_add(bat_priv, &ip_src, BATADV_DAT_IPV4, hw_src, vid); + batadv_dat_entry_add(bat_priv, &ip_dst, BATADV_DAT_IPV4, hw_dst, vid); /* if this REPLY is directed to a client of mine, let's deliver the * packet to the interface @@ -1167,7 +1261,8 @@ bool batadv_dat_drop_broadcast_packet(struct batadv_priv *bat_priv, goto out; ip_dst = batadv_arp_ip_dst(forw_packet->skb, hdr_size); - dat_entry = batadv_dat_entry_hash_find(bat_priv, ip_dst, vid); + dat_entry = batadv_dat_entry_hash_find(bat_priv, &ip_dst, + BATADV_DAT_IPV4, vid); /* check if the node already got this entry */ if (!dat_entry) { batadv_dbg(BATADV_DBG_DAT, bat_priv, diff --git a/distributed-arp-table.h b/distributed-arp-table.h index 60d853b..557bab9 100644 --- a/distributed-arp-table.h +++ b/distributed-arp-table.h @@ -28,6 +28,7 @@ #include <linux/if_arp.h> #define BATADV_DAT_ADDR_MAX ((batadv_dat_addr_t)~(batadv_dat_addr_t)0) +#define BATADV_DAT_DATA_MAX_LEN 16 void batadv_dat_status_update(struct net_device *net_dev); bool batadv_dat_snoop_outgoing_arp_request(struct batadv_priv *bat_priv, diff --git a/types.h b/types.h index 20a1bef..69c187e 100644 --- a/types.h +++ b/types.h @@ -931,7 +931,8 @@ struct batadv_algo_ops { /** * 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 + * @data: the data corresponding to this DAT entry + * @type: the type corresponding to this DAT entry * @mac_addr: the MAC address associated to the stored IPv4 * @vid: the vlan ID associated to this entry * @last_update: time in jiffies when this entry was refreshed last time @@ -940,7 +941,8 @@ struct batadv_algo_ops { * @rcu: struct used for freeing in an RCU-safe manner */ struct batadv_dat_entry { - __be32 ip; + void *data; + uint8_t type; uint8_t mac_addr[ETH_ALEN]; unsigned short vid; unsigned long last_update; @@ -950,6 +952,24 @@ struct batadv_dat_entry { }; /** + * batadv_dat_types - types used in batadv_dat_entry for IP + * @BATADV_DAT_IPv4: IPv4 address type + */ +enum batadv_dat_types { + BATADV_DAT_IPV4 = 0, +}; + +/** + * batadv_dat_type_info - info needed for a DAT type data + * @size: the size of the type data + * @str_fmt: string format used by the data + */ +struct batadv_dat_type_info { + size_t size; + char *str_fmt; +}; + +/** * 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 -- 1.7.10.4

8 32

[B.A.T.M.A.N.] Classes and purposed classes (problem to solve)
by cmsv 16 Nov '15

16 Nov '15

Scenario: I have recently moved 13 nodes previously using wds to batman-adv. One problem that i am encountering is the gateway selection and its actually defectiveness. This problem has be seen when the client nodes are set on class 20. The node clients in fact select the gateways that advertise better link quality usually above 200 but at the same time some clients end up connecting to a gateway that is either too slow or unusable either due to routing or other factors such as distance and even clear visibility. In one case the node connects to something that defies logic. I thought about ways to work around this and came up with an idea that may help which means the creation of 2 new classes. - Preferred gateway fall-back - Preferred fixed gateway For Preferred gateway fall-back: consider the gateway's advertised quality towards the gateway and stick with the selection until the gateway disappears returning to it soon as it returns regardless of any other data. - Preferred fixed gateway Manual gateway selection by the node owner or admin. (wds type) If these classes cannot be created; maybe a plugin of some sort could be. could be called "gordon". I understand that this idea may sound useless the same way blocking originators may be for someone but in some scenarios just like blocking originators it look to me very valid. My current problem is that the link quality seems to go up and down like a roller coaster in a specific order which causes the clients to switch constantly sometimes many times per hour. While this does not seem to be an issue for most client nodes; for 2 of them it means no internet access at all. The solution for now seems to be having these 2 nodes forced to use 1 specific gateway. I am quite open to any ideas that may help resolve this issue that has been casing some "uninformed people" to become headaches. -- Redes wireless http://wirelesspt.net

3 2

[B.A.T.M.A.N.] TVLV infrastructure (v2)
by Marek Lindner 10 Jun '15

10 Jun '15

Hi, at the battlemesh we had plenty of time to discuss and test the tvlv patchset. Therefore, a number of changes/fixes found their way into v2 touching the tvlv API, network coding, unicast sending and gateway announcements. This looks like the final version which is going to be merged unless objections are raised soon-ish. Cheers, Marek

2 15

[B.A.T.M.A.N.] [PATCH] batman-adv: send GW_DEL event in case of soft-iface destruction
by Antonio Quartulli 24 Sep '13

24 Sep '13

From: Antonio Quartulli <antonio(a)open-mesh.com> In case of soft_iface destruction send a GW DEL event to userspace so that applications which are listening for GW events are informed about the lost of connectivity and can react accordingly. Signed-off-by: Antonio Quartulli <antonio(a)open-mesh.com> --- hard-interface.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-) diff --git a/hard-interface.c b/hard-interface.c index c5f871f..8a42745 100644 --- a/hard-interface.c +++ b/hard-interface.c @@ -28,6 +28,7 @@ #include "originator.h" #include "hash.h" #include "bridge_loop_avoidance.h" +#include "gateway_client.h" #include <linux/if_arp.h> #include <linux/if_ether.h> @@ -524,8 +525,12 @@ void batadv_hardif_disable_interface(struct batadv_hard_iface *hard_iface, dev_put(hard_iface->soft_iface); /* nobody uses this interface anymore */ - if (!bat_priv->num_ifaces && autodel == BATADV_IF_CLEANUP_AUTO) - batadv_softif_destroy_sysfs(hard_iface->soft_iface); + if (!bat_priv->num_ifaces) { + batadv_gw_check_client_stop(bat_priv); + + if (autodel == BATADV_IF_CLEANUP_AUTO) + batadv_softif_destroy_sysfs(hard_iface->soft_iface); + } netdev_upper_dev_unlink(hard_iface->net_dev, hard_iface->soft_iface); hard_iface->soft_iface = NULL; -- 1.8.1.5

2 1

[B.A.T.M.A.N.] [PATCH] batctl: add sys framework for VLAN support
by Marco Dalla Torre 04 Sep '13

04 Sep '13

Version 2 - code style fixes If no directory entry corresponding to the user-selected device is found at the standard location for non VLAN interfaces (/sys/class/net/${base_device}), 'batctl' now looks into directory: /sys/devices/virtual/net/${base_device}/mesh/vlan${vid} Where: -${base_device}: the batman device on top of which the VLAN is sitting -${device}: the device interface for the VLAN, -${vid}: the identifier assigned to the VLAN. Information on VLAN devices (base device, vid) necessary to construct the directory path is acquired by parsing /proc/net/vlan/${device}. If the user-selected command is not supported by the VLAN, an appropriate error is shown. --- functions.c | 3 +++ sys.c | 51 +++++++++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 54 insertions(+) diff --git a/functions.c b/functions.c index cc05a48..ed010ea 100644 --- a/functions.c +++ b/functions.c @@ -135,6 +135,9 @@ static void file_open_problem_dbg(char *dir, char *fname, char *full_path) fprintf(stderr, "Error - the folder '/sys/' was not found on the system\n"); fprintf(stderr, "Please make sure that the sys filesystem is properly mounted\n"); return; + } else if (strstr(dir, "/sys/devices/virtual/")) { + fprintf(stderr, "The selected feature '%s' is not supported for vlans\n", fname); + return; } } diff --git a/sys.c b/sys.c index b1d7ea8..942ead0 100644 --- a/sys.c +++ b/sys.c @@ -371,6 +371,37 @@ static void settings_usage(int setting) fprintf(stderr, " \t -h print this help\n"); } +int get_basedev_vid(char *mesh_iface, char **base_dev, unsigned short *vid) +{ + char *vdev; + char line[100]; + const char path[] = "/proc/net/vlan/"; + int size = sizeof(path) + sizeof(mesh_iface); + FILE *fp = NULL; + + char *fpath = malloc(size); + strcpy(fpath, path); + /* prepare path file path: /proc/net/vlan/$mesh_iface*/ + strcat(fpath, mesh_iface); + + fp = fopen(fpath, "r"); + if (fp == NULL) + return 1; + + if (fscanf(fp, "%ms %*s %hu %*s %*d %*s %*d", &vdev, vid) == 0) + return 0; + + while (fgets(line, sizeof(line), fp) != NULL) { + if (sscanf(line, "Device: %ms", base_dev) == 1) + break; + } + // handle base device not found case + if (*base_dev == NULL) + return 0; + + return 1; +} + int handle_sys_setting(char *mesh_iface, int setting, int argc, char **argv) { int optchar, res = EXIT_FAILURE; @@ -392,6 +423,26 @@ int handle_sys_setting(char *mesh_iface, int setting, int argc, char **argv) snprintf(path_buff, PATH_BUFF_LEN, SYS_BATIF_PATH_FMT, mesh_iface); path_buff[PATH_BUFF_LEN - 1] = '\0'; + if (access(path_buff, F_OK) != 0) { + if (errno == ENOENT) { + // path does not exist, no lan interface: check vlan + unsigned short vid = 0; + char *base_dev = NULL; + if (get_basedev_vid(mesh_iface, &base_dev, &vid) == 1) { + free(path_buff); + char sys_vlan_path[] = "/sys/devices/virtual/net/%s/mesh/vlan%d/"; + int size = sizeof(sys_vlan_path) + sizeof(base_dev); + path_buff = malloc(size); + sprintf(path_buff, sys_vlan_path, base_dev, vid); + } + } + if (errno == ENOTDIR) { + // not a directory, something wrong here + fprintf(stderr, "Error - expected directory at '%s'\n", + path_buff); + return EXIT_FAILURE; + } + } if (argc == 1) { res = read_file(path_buff, (char *)batctl_settings[setting].sysfs_name, NO_FLAGS, 0, 0, 0); -- 1.8.3.2

3 7

[B.A.T.M.A.N.] [PATCH] batman-adv: limit local translation table max size
by Marek Lindner 01 Sep '13

01 Sep '13

The local translation table size is limited by what can be transferred from one node to another via a full table request. The number of entries fitting into a full table request depend on whether the fragmentation is enabled or not. Therefore this patch introduces a max table size check and refuses to add more local clients when that size is reached. Moreover, if the max full table packet size changes (MTU change or fragmentation is disabled) the local table is downsized instantaneously. Signed-off-by: Marek Lindner <lindner_marek(a)yahoo.de> --- hard-interface.c | 41 +++++++++----- soft-interface.c | 12 ++-- translation-table.c | 154 ++++++++++++++++++++++++++++++++++++++++++++------- translation-table.h | 3 +- types.h | 3 + 5 files changed, 172 insertions(+), 41 deletions(-) diff --git a/hard-interface.c b/hard-interface.c index c5f871f..c60d3ed 100644 --- a/hard-interface.c +++ b/hard-interface.c @@ -266,16 +266,9 @@ static void batadv_check_known_mac_addr(const struct net_device *net_dev) int batadv_hardif_min_mtu(struct net_device *soft_iface) { - const struct batadv_priv *bat_priv = netdev_priv(soft_iface); + struct batadv_priv *bat_priv = netdev_priv(soft_iface); const struct batadv_hard_iface *hard_iface; - /* allow big frames if all devices are capable to do so - * (have MTU > 1500 + batadv_max_header_len()) - */ int min_mtu = ETH_DATA_LEN; - int max_header_len = batadv_max_header_len(); - - if (atomic_read(&bat_priv->fragmentation)) - goto out; rcu_read_lock(); list_for_each_entry_rcu(hard_iface, &batadv_hardif_list, list) { @@ -286,22 +279,40 @@ int batadv_hardif_min_mtu(struct net_device *soft_iface) if (hard_iface->soft_iface != soft_iface) continue; - min_mtu = min_t(int, hard_iface->net_dev->mtu - max_header_len, - min_mtu); + min_mtu = min_t(int, hard_iface->net_dev->mtu, min_mtu); } rcu_read_unlock(); + + atomic_set(&bat_priv->packet_size_max, min_mtu); + + if (atomic_read(&bat_priv->fragmentation) == 0) + goto out; + + /* with fragmentation enabled the maximum size of internally generated + * packets such as translation table exchanges or tvlv containers, etc + * has to be calculated + */ + min_mtu = min_t(int, min_mtu, BATADV_FRAG_MAX_FRAG_SIZE); + min_mtu -= sizeof(struct batadv_frag_packet); + min_mtu *= BATADV_FRAG_MAX_FRAGMENTS; + atomic_set(&bat_priv->packet_size_max, min_mtu); + + /* with fragmentation enabled we can fragment external packets easily */ + min_mtu = min_t(int, min_mtu, ETH_DATA_LEN); + out: - return min_mtu; + return min_mtu - batadv_max_header_len(); } /* adjusts the MTU if a new interface with a smaller MTU appeared. */ void batadv_update_min_mtu(struct net_device *soft_iface) { - int min_mtu; + soft_iface->mtu = batadv_hardif_min_mtu(soft_iface); - min_mtu = batadv_hardif_min_mtu(soft_iface); - if (soft_iface->mtu != min_mtu) - soft_iface->mtu = min_mtu; + /* Check if the local translate table should be cleaned up to match a + * new (and smaller) MTU. + */ + batadv_tt_local_resize_to_mtu(soft_iface); } static void diff --git a/soft-interface.c b/soft-interface.c index bafc811..a1f00e8 100644 --- a/soft-interface.c +++ b/soft-interface.c @@ -166,7 +166,7 @@ static int batadv_interface_tx(struct sk_buff *skb, unsigned int header_len = 0; int data_len = skb->len, ret; unsigned long brd_delay = 1; - bool do_bcast = false; + bool do_bcast = false, client_added; unsigned short vid; uint32_t seqno; @@ -196,9 +196,12 @@ static int batadv_interface_tx(struct sk_buff *skb, ethhdr = (struct ethhdr *)skb->data; /* Register the client MAC in the transtable */ - if (!is_multicast_ether_addr(ethhdr->h_source)) - batadv_tt_local_add(soft_iface, ethhdr->h_source, vid, - skb->skb_iif); + if (!is_multicast_ether_addr(ethhdr->h_source)) { + client_added = batadv_tt_local_add(soft_iface, ethhdr->h_source, + vid, skb->skb_iif); + if (!client_added) + goto dropped; + } /* don't accept stp packets. STP does not help in meshes. * better use the bridge loop avoidance ... @@ -674,6 +677,7 @@ static int batadv_softif_init_late(struct net_device *dev) atomic_set(&bat_priv->log_level, 0); #endif atomic_set(&bat_priv->fragmentation, 1); + atomic_set(&bat_priv->packet_size_max, ETH_DATA_LEN); atomic_set(&bat_priv->bcast_queue_left, BATADV_BCAST_QUEUE_LEN); atomic_set(&bat_priv->batman_queue_left, BATADV_BATMAN_QUEUE_LEN); diff --git a/translation-table.c b/translation-table.c index 2b4d9ed..6a69510 100644 --- a/translation-table.c +++ b/translation-table.c @@ -401,6 +401,35 @@ static uint16_t batadv_tt_entries(uint16_t tt_len) return tt_len / batadv_tt_len(1); } +/** + * batadv_tt_local_table_transmit_size - calculates the local translation table + * size when transmitted over the air + * @bat_priv: the bat priv with all the soft interface information + * + * Returns local translation table size in bytes. + */ +static int batadv_tt_local_table_transmit_size(struct batadv_priv *bat_priv) +{ + struct batadv_softif_vlan *vlan; + uint16_t num_vlan = 0, tt_local_entries = 0; + int hdr_size; + + rcu_read_lock(); + hlist_for_each_entry_rcu(vlan, &bat_priv->softif_vlan_list, list) { + num_vlan++; + tt_local_entries += atomic_read(&vlan->tt.num_entries); + } + rcu_read_unlock(); + + /* header size of tvlv encapsulated tt response payload */ + hdr_size = sizeof(struct batadv_unicast_tvlv_packet); + hdr_size += sizeof(struct batadv_tvlv_hdr); + hdr_size += sizeof(struct batadv_tvlv_tt_data); + hdr_size += num_vlan * sizeof(struct batadv_tvlv_tt_vlan_data); + + return hdr_size + batadv_tt_len(tt_local_entries); +} + static int batadv_tt_local_init(struct batadv_priv *bat_priv) { if (bat_priv->tt.local_hash) @@ -439,8 +468,10 @@ static void batadv_tt_global_free(struct batadv_priv *bat_priv, * @vid: VLAN identifier * @ifindex: index of the interface where the client is connected to (useful to * identify wireless clients) + * + * Returns true if the client was successfully added, false otherwise. */ -void batadv_tt_local_add(struct net_device *soft_iface, const uint8_t *addr, +bool batadv_tt_local_add(struct net_device *soft_iface, const uint8_t *addr, unsigned short vid, int ifindex) { struct batadv_priv *bat_priv = netdev_priv(soft_iface); @@ -448,8 +479,8 @@ void batadv_tt_local_add(struct net_device *soft_iface, const uint8_t *addr, struct batadv_tt_global_entry *tt_global; struct hlist_head *head; struct batadv_tt_orig_list_entry *orig_entry; - int hash_added; - bool roamed_back = false; + int hash_added, table_size, packet_size_max; + bool ret = false, roamed_back = false; tt_local = batadv_tt_local_hash_find(bat_priv, addr, vid); tt_global = batadv_tt_global_hash_find(bat_priv, addr, vid); @@ -484,6 +515,17 @@ void batadv_tt_local_add(struct net_device *soft_iface, const uint8_t *addr, goto check_roaming; } + /* Ignore the client if we cannot send it in a full table response. */ + table_size = batadv_tt_local_table_transmit_size(bat_priv); + table_size += batadv_tt_len(1); + packet_size_max = atomic_read(&bat_priv->packet_size_max); + if (table_size > packet_size_max) { + net_ratelimited_function(batadv_info, soft_iface, + "Local translation table size (%i) exceeds maximum packet size (%i); Ignoring new local tt entry: %pM\n", + table_size, packet_size_max, addr); + goto out; + } + tt_local = kmalloc(sizeof(*tt_local), GFP_ATOMIC); if (!tt_local) goto out; @@ -550,11 +592,14 @@ check_roaming: } } + ret = true; + out: if (tt_local) batadv_tt_local_entry_free_ref(tt_local); if (tt_global) batadv_tt_global_entry_free_ref(tt_global); + return ret; } /** @@ -717,12 +762,6 @@ static void batadv_tt_tvlv_container_update(struct batadv_priv *bat_priv) tt_diff_entries_num = atomic_read(&bat_priv->tt.local_changes); tt_diff_len = batadv_tt_len(tt_diff_entries_num); - /* if we have too many changes for one packet don't send any - * and wait for the tt table request which will be fragmented - */ - if (tt_diff_len > bat_priv->soft_iface->mtu) - tt_diff_len = 0; - tvlv_len = batadv_tt_prepare_tvlv_local_data(bat_priv, &tt_data, &tt_change, &tt_diff_len); if (!tvlv_len) @@ -926,8 +965,16 @@ out: return curr_flags; } +/** + * batadv_tt_local_purge_list - purge inactive tt local entries + * @bat_priv: the bat priv with all the soft interface information + * @head: pointer to the list containing the local tt entries + * @timeout: parameter deciding whether a given tt local entry is considered + * inactive or not + */ static void batadv_tt_local_purge_list(struct batadv_priv *bat_priv, - struct hlist_head *head) + struct hlist_head *head, + int timeout) { struct batadv_tt_local_entry *tt_local_entry; struct batadv_tt_common_entry *tt_common_entry; @@ -945,8 +992,7 @@ static void batadv_tt_local_purge_list(struct batadv_priv *bat_priv, if (tt_local_entry->common.flags & BATADV_TT_CLIENT_PENDING) continue; - if (!batadv_has_timed_out(tt_local_entry->last_seen, - BATADV_TT_LOCAL_TIMEOUT)) + if (!batadv_has_timed_out(tt_local_entry->last_seen, timeout)) continue; batadv_tt_local_set_pending(bat_priv, tt_local_entry, @@ -954,7 +1000,14 @@ static void batadv_tt_local_purge_list(struct batadv_priv *bat_priv, } } -static void batadv_tt_local_purge(struct batadv_priv *bat_priv) +/** + * batadv_tt_local_purge - purge inactive tt local entries + * @bat_priv: the bat priv with all the soft interface information + * @timeout: parameter deciding whether a given tt local entry is considered + * inactive or not + */ +static void batadv_tt_local_purge(struct batadv_priv *bat_priv, + int timeout) { struct batadv_hashtable *hash = bat_priv->tt.local_hash; struct hlist_head *head; @@ -966,7 +1019,7 @@ static void batadv_tt_local_purge(struct batadv_priv *bat_priv) list_lock = &hash->list_locks[i]; spin_lock_bh(list_lock); - batadv_tt_local_purge_list(bat_priv, head); + batadv_tt_local_purge_list(bat_priv, head, timeout); spin_unlock_bh(list_lock); } } @@ -2370,6 +2423,15 @@ static bool batadv_send_other_tt_response(struct batadv_priv *bat_priv, req_dst_orig_node); } + /* Don't send the response, if larger than fragmented packet. */ + tt_len = sizeof(struct batadv_unicast_tvlv_packet) + tvlv_len; + if (tt_len > atomic_read(&bat_priv->packet_size_max)) { + net_ratelimited_function(batadv_info, bat_priv->soft_iface, + "Ignoring TT_REQUEST from %pM; Response size exceeds max packet size.\n", + res_dst_orig_node->orig); + goto out; + } + tvlv_tt_data->flags = BATADV_TT_RESPONSE; tvlv_tt_data->ttvn = req_ttvn; @@ -2846,7 +2908,7 @@ static void batadv_tt_purge(struct work_struct *work) priv_tt = container_of(delayed_work, struct batadv_priv_tt, work); bat_priv = container_of(priv_tt, struct batadv_priv, tt); - batadv_tt_local_purge(bat_priv); + batadv_tt_local_purge(bat_priv, BATADV_TT_LOCAL_TIMEOUT); batadv_tt_global_purge(bat_priv); batadv_tt_req_purge(bat_priv); batadv_tt_roam_purge(bat_priv); @@ -2959,18 +3021,18 @@ static void batadv_tt_local_purge_pending_clients(struct batadv_priv *bat_priv) } /** - * batadv_tt_local_commit_changes - commit all pending local tt changes which + * batadv_tt_local_commit_changes_nolock - commit all pending local tt changes which * have been queued in the time since the last commit * @bat_priv: the bat priv with all the soft interface information + * + * Caller must hold tt->commit_lock. */ -void batadv_tt_local_commit_changes(struct batadv_priv *bat_priv) +void batadv_tt_local_commit_changes_nolock(struct batadv_priv *bat_priv) { - spin_lock_bh(&bat_priv->tt.commit_lock); - if (atomic_read(&bat_priv->tt.local_changes) < 1) { if (!batadv_atomic_dec_not_zero(&bat_priv->tt.ogm_append_cnt)) batadv_tt_tvlv_container_update(bat_priv); - goto out; + return; } batadv_tt_local_set_flags(bat_priv, BATADV_TT_CLIENT_NEW, false, true); @@ -2987,8 +3049,17 @@ void batadv_tt_local_commit_changes(struct batadv_priv *bat_priv) /* reset the sending counter */ atomic_set(&bat_priv->tt.ogm_append_cnt, BATADV_TT_OGM_APPEND_MAX); batadv_tt_tvlv_container_update(bat_priv); +} -out: +/** + * batadv_tt_local_commit_changes - commit all pending local tt changes which + * have been queued in the time since the last commit + * @bat_priv: the bat priv with all the soft interface information + */ +void batadv_tt_local_commit_changes(struct batadv_priv *bat_priv) +{ + spin_lock_bh(&bat_priv->tt.commit_lock); + batadv_tt_local_commit_changes_nolock(bat_priv); spin_unlock_bh(&bat_priv->tt.commit_lock); } @@ -3184,6 +3255,47 @@ out: } /** + * batadv_tt_local_resize_to_mtu - resize the local translation table fit the + * maximum packet size that can be transported through the mesh + * @soft_iface: netdev struct of the mesh interface + * + * Remove entries older than 'timeout' and half timeout if more entries need + * to be removed. + */ +void batadv_tt_local_resize_to_mtu(struct net_device *soft_iface) +{ + struct batadv_priv *bat_priv = netdev_priv(soft_iface); + int packet_size_max = atomic_read(&bat_priv->packet_size_max); + int table_size, timeout = BATADV_TT_LOCAL_TIMEOUT / 2; + bool reduced = false; + + spin_lock_bh(&bat_priv->tt.commit_lock); + + while (true) { + table_size = batadv_tt_local_table_transmit_size(bat_priv); + if (packet_size_max >= table_size) + break; + + batadv_tt_local_purge(bat_priv, timeout); + batadv_tt_local_purge_pending_clients(bat_priv); + + timeout /= 2; + reduced = true; + net_ratelimited_function(batadv_info, soft_iface, + "Forced to purge local tt entries to fit new maximum fragment MTU (%i)\n", + packet_size_max); + } + + /* commit this changes immediately, to avoid synchronization problem + * with the TTVN + */ + if (reduced) + batadv_tt_local_commit_changes_nolock(bat_priv); + + spin_unlock_bh(&bat_priv->tt.commit_lock); +} + +/** * batadv_tt_tvlv_ogm_handler_v1 - process incoming tt tvlv container * @bat_priv: the bat priv with all the soft interface information * @orig: the orig_node of the ogm diff --git a/translation-table.h b/translation-table.h index dc6db4e..026b1ff 100644 --- a/translation-table.h +++ b/translation-table.h @@ -21,7 +21,7 @@ #define _NET_BATMAN_ADV_TRANSLATION_TABLE_H_ int batadv_tt_init(struct batadv_priv *bat_priv); -void batadv_tt_local_add(struct net_device *soft_iface, const uint8_t *addr, +bool batadv_tt_local_add(struct net_device *soft_iface, const uint8_t *addr, unsigned short vid, int ifindex); uint16_t batadv_tt_local_remove(struct batadv_priv *bat_priv, const uint8_t *addr, unsigned short vid, @@ -45,6 +45,7 @@ bool batadv_tt_global_client_is_roaming(struct batadv_priv *bat_priv, uint8_t *addr, unsigned short vid); bool batadv_tt_local_client_is_roaming(struct batadv_priv *bat_priv, uint8_t *addr, unsigned short vid); +void batadv_tt_local_resize_to_mtu(struct net_device *soft_iface); bool batadv_tt_add_temporary_global_entry(struct batadv_priv *bat_priv, struct batadv_orig_node *orig_node, const unsigned char *addr, diff --git a/types.h b/types.h index 0f938c2..d395bb8 100644 --- a/types.h +++ b/types.h @@ -595,6 +595,8 @@ struct batadv_softif_vlan { * @aggregated_ogms: bool indicating whether OGM aggregation is enabled * @bonding: bool indicating whether traffic bonding is enabled * @fragmentation: bool indicating whether traffic fragmentation is enabled + * @packet_size_max: max packet size that can be transmitted via + * multiple fragmented skbs or a single frame if fragmentation is disabled * @frag_seqno: incremental counter to identify chains of egress fragments * @bridge_loop_avoidance: bool indicating whether bridge loop avoidance is * enabled @@ -641,6 +643,7 @@ struct batadv_priv { atomic_t aggregated_ogms; atomic_t bonding; atomic_t fragmentation; + atomic_t packet_size_max; atomic_t frag_seqno; #ifdef CONFIG_BATMAN_ADV_BLA atomic_t bridge_loop_avoidance; -- 1.7.10.4

2 5

[B.A.T.M.A.N.] [PATCHv6 0/9] Improving the routing protocol abstraction
by Antonio Quartulli 31 Aug '13

31 Aug '13

Hello list, This is the *fifth* version of the "improved routing abstraction" patchset. Thanks for the feedback so far! (Message below is the same as v3) This patchset is introducing a new set of routing API functions meant to improve the routing protocol abstraction. Also the orig_node and the neigh_node structure have been heavily modified in order to split their generic members from those which are metric related. (the rest of the message is what you already had in the RFC cover letter). This changes have been written while developing batman V. The latter helped me in understanding what batman iv and v have in common and what not. The main problem was the metric: the two protocols use different metric domains and different semantics. Therefore all the functions handling/printing the metric needed to be generalised and rearranged to let the protocols decide what to do. Another issue was the way routing protocols handle the orig and neigh node structures. Also these two have been changed and some small APIs have been provided as well. Moreover, after Simon's RFC about the new multi-interface optimisation, we saw the need for a better abstraction so that mechanisms like that could easily be re-used by new algorithms (like batman v) with little effort. ****** API functions changed starting from v4 ****** + neigh comparison related: - bat_neigh_cmp - bat_neigh_is_equiv_or_better + orig_node related: - bat_orig_print: print the originator table - bat_orig_free - bat_orig_add_if - bat_orig_del_if Changes from v1: - removed bat_metric_compare() API. This is not really useful since we are assuming all the metrics are in the same domain (uint32_t) so a simple comparison is more than enough. Changes from v2: - initialise best_metric in the neighbor purging routine - don't optimise ntohs. Will be done in a separate patch Changes from v3: - bat_metric_is_equiv_or_better is now called bat_neigh_is_equiv_or_better and now directly takes two neigh_node as parameters. This allows to do not export the metric at all: all the metric magic is handled inside the function. - bat_metric_compare has been re-introduced again with the bat_neigh_cmp name. As for bat_neigh_is_equiv_or_better it also takes two neigh_node as parameters and the metric is handled inside only. - as result of the two changes above bat_metric_get has been removed. The metric is always handled inside the routing protocol code and there is no need to export it to the rest of the module. This simplified the code and reduced code duplication. (Thanks Marek for the hint!) Changes from v4: - fixed typ0 in 9/9 Changes from v5: - add neigh_node to the neighbor list only after having completed the whole initialisation phase (in 1/9) Cheers, Antonio Quartulli (9): batman-adv: make struct batadv_neigh_node algorithm agnostic batman-adv: make struct batadv_orig_node algorithm agnostic batman-adv: add bat_orig_print API function batman-adv: add bat_neigh_cmp API function batman-adv: add bat_neigh_is_equiv_or_better API function batman-adv: adapt bonding to use the new API functions batman-adv: adapt the neighbor purging routine to use the new API functions batman-adv: provide orig_node routing API batman-adv: adapt the TT component to use the new API functions bat_iv_ogm.c | 394 ++++++++++++++++++++++++++++++++++++++++++++-------- gateway_client.c | 16 +-- main.c | 4 +- main.h | 6 + network-coding.c | 8 +- originator.c | 246 ++++++++------------------------ originator.h | 5 +- routing.c | 39 ++++-- routing.h | 3 +- translation-table.c | 35 +++-- types.h | 95 +++++++++---- 11 files changed, 546 insertions(+), 305 deletions(-) -- 1.8.1.5

1 9

[B.A.T.M.A.N.] [PATCHv5 0/9] Improving the routing protocol abstraction
by Antonio Quartulli 31 Aug '13

31 Aug '13

Hello list, This is the *fifth* version of the "improved routing abstraction" patchset. Thanks for the feedback so far! (Message below is the same as v3) This patchset is introducing a new set of routing API functions meant to improve the routing protocol abstraction. Also the orig_node and the neigh_node structure have been heavily modified in order to split their generic members from those which are metric related. (the rest of the message is what you already had in the RFC cover letter). This changes have been written while developing batman V. The latter helped me in understanding what batman iv and v have in common and what not. The main problem was the metric: the two protocols use different metric domains and different semantics. Therefore all the functions handling/printing the metric needed to be generalised and rearranged to let the protocols decide what to do. Another issue was the way routing protocols handle the orig and neigh node structures. Also these two have been changed and some small APIs have been provided as well. Moreover, after Simon's RFC about the new multi-interface optimisation, we saw the need for a better abstraction so that mechanisms like that could easily be re-used by new algorithms (like batman v) with little effort. ****** API functions changed starting from v4 ****** + neigh comparison related: - bat_neigh_cmp - bat_neigh_is_equiv_or_better + orig_node related: - bat_orig_print: print the originator table - bat_orig_free - bat_orig_add_if - bat_orig_del_if Changes from v1: - removed bat_metric_compare() API. This is not really useful since we are assuming all the metrics are in the same domain (uint32_t) so a simple comparison is more than enough. Changes from v2: - initialise best_metric in the neighbor purging routine - don't optimise ntohs. Will be done in a separate patch Changes from v3: - bat_metric_is_equiv_or_better is now called bat_neigh_is_equiv_or_better and now directly takes two neigh_node as parameters. This allows to do not export the metric at all: all the metric magic is handled inside the function. - bat_metric_compare has been re-introduced again with the bat_neigh_cmp name. As for bat_neigh_is_equiv_or_better it also takes two neigh_node as parameters and the metric is handled inside only. - as result of the two changes above bat_metric_get has been removed. The metric is always handled inside the routing protocol code and there is no need to export it to the rest of the module. This simplified the code and reduced code duplication. (Thanks Marek for the hint!) Changes from v4: - fixed typ0 in 9/9 Cheers, Antonio Quartulli (9): batman-adv: make struct batadv_neigh_node algorithm agnostic batman-adv: make struct batadv_orig_node algorithm agnostic batman-adv: add bat_orig_print API function batman-adv: add bat_neigh_cmp API function batman-adv: add bat_neigh_is_equiv_or_better API function batman-adv: adapt bonding to use the new API functions batman-adv: adapt the neighbor purging routine to use the new API functions batman-adv: provide orig_node routing API batman-adv: adapt the TT component to use the new API functions bat_iv_ogm.c | 398 ++++++++++++++++++++++++++++++++++++++++++++-------- gateway_client.c | 16 +-- main.c | 4 +- main.h | 6 + network-coding.c | 8 +- originator.c | 248 ++++++++------------------------ originator.h | 5 +- routing.c | 42 ++++-- routing.h | 3 +- translation-table.c | 35 +++-- types.h | 91 ++++++++---- 11 files changed, 546 insertions(+), 310 deletions(-) -- 1.8.1.5

2 14

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

B.A.T.M.A.N August 2013