From 934993137199ffb56fef50664f87e71cdb3471b0 Mon Sep 17 00:00:00 2001 From: Allan Stephens Date: Tue, 25 Oct 2011 15:14:46 -0400 Subject: tipc: Ensure broadcast link re-acquires node after link failure Fix a bug that can prevent TIPC from sending broadcast messages to a node if contact with the node is lost and then regained. The problem occurs if the broadcast link first clears the flag indicating the node is part of the link's distribution set (when it loses contact with the node), and later fails to restore the flag (when contact is regained); restoration fails if contact with the node is regained by implicit unicast link activation triggered by the arrival of a data message, rather than explicitly by the arrival of a link activation message. The broadcast link now uses separate fields to track whether a node is theoretically capable of receiving broadcast messages versus whether it is actually part of the link's distribution set. The former member is updated by the receipt of link protocol messages, which can occur at any time; the latter member is updated only when contact with the node is gained or lost. This change also permits the simplification of several conditional expressions since the broadcast link's "supported" field can now only be set if there are working links to the associated node. Signed-off-by: Allan Stephens Signed-off-by: Paul Gortmaker --- net/tipc/node.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'net/tipc/node.c') diff --git a/net/tipc/node.c b/net/tipc/node.c index 6b226faad89f..9196f943b835 100644 --- a/net/tipc/node.c +++ b/net/tipc/node.c @@ -306,8 +306,9 @@ static void node_established_contact(struct tipc_node *n_ptr) /* Syncronize broadcast acks */ n_ptr->bclink.acked = tipc_bclink_get_last_sent(); - if (n_ptr->bclink.supported) { + if (n_ptr->bclink.supportable) { tipc_bclink_add_node(n_ptr->addr); + n_ptr->bclink.supported = 1; if (n_ptr->addr < tipc_own_addr) tipc_own_tag++; } -- cgit v1.2.3 From 7a54d4a99dcbbfdf1d4550faa19b615091137953 Mon Sep 17 00:00:00 2001 From: Allan Stephens Date: Thu, 27 Oct 2011 14:17:53 -0400 Subject: tipc: Major redesign of broadcast link ACK/NACK algorithms Completely redesigns broadcast link ACK and NACK mechanisms to prevent spurious retransmit requests in dual LAN networks, and to prevent the broadcast link from stalling due to the failure of a receiving node to acknowledge receiving a broadcast message or request its retransmission. Note: These changes only impact the timing of when ACK and NACK messages are sent, and not the basic broadcast link protocol itself, so inter- operability with nodes using the "classic" algorithms is maintained. The revised algorithms are as follows: 1) An explicit ACK message is still sent after receiving 16 in-sequence messages, and implicit ACK information continues to be carried in other unicast link message headers (including link state messages). However, the timing of explicit ACKs is now based on the receiving node's absolute network address rather than its relative network address to ensure that the failure of another node does not delay the ACK beyond its 16 message target. 2) A NACK message is now typically sent only when a message gap persists for two consecutive incoming link state messages; this ensures that a suspected gap is not confirmed until both LANs in a dual LAN network have had an opportunity to deliver the message, thereby preventing spurious NACKs. A NACK message can also be generated by the arrival of a single link state message, if the deferred queue is so big that the current message gap cannot be the result of "normal" mis-ordering due to the use of dual LANs (or one LAN using a bonded interface). Since link state messages typically arrive at different nodes at different times the problem of multiple nodes issuing identical NACKs simultaneously is inherently avoided. 3) Nodes continue to "peek" at NACK messages sent by other nodes. If another node requests retransmission of a message gap suspected (but not yet confirmed) by the peeking node, the peeking node forgets about the gap and does not generate a duplicate retransmit request. (If the peeking node subsequently fails to receive the lost message, later link state messages will cause it to rediscover and confirm the gap and send another NACK.) 4) Message gap "equality" is now determined by the start of the gap only. This is sufficient to deal with the most common cases of message loss, and eliminates the need for complex end of gap computations. 5) A peeking node no longer tries to determine whether it should send a complementary NACK, since the most common cases of message loss don't require it to be sent. Consequently, the node no longer examines the "broadcast tag" field of a NACK message when peeking. Signed-off-by: Allan Stephens Signed-off-by: Paul Gortmaker --- net/tipc/node.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'net/tipc/node.c') diff --git a/net/tipc/node.c b/net/tipc/node.c index 9196f943b835..6d8bdfd95cd6 100644 --- a/net/tipc/node.c +++ b/net/tipc/node.c @@ -339,12 +339,12 @@ static void node_lost_contact(struct tipc_node *n_ptr) /* Flush broadcast link info associated with lost node */ if (n_ptr->bclink.supported) { - n_ptr->bclink.gap_after = n_ptr->bclink.gap_to = 0; while (n_ptr->bclink.deferred_head) { struct sk_buff *buf = n_ptr->bclink.deferred_head; n_ptr->bclink.deferred_head = buf->next; buf_discard(buf); } + n_ptr->bclink.deferred_size = 0; if (n_ptr->bclink.defragm) { buf_discard(n_ptr->bclink.defragm); @@ -450,7 +450,7 @@ struct sk_buff *tipc_node_get_links(const void *req_tlv_area, int req_tlv_space) read_lock_bh(&tipc_net_lock); - /* Get space for all unicast links + multicast link */ + /* Get space for all unicast links + broadcast link */ payload_size = TLV_SPACE(sizeof(link_info)) * (atomic_read(&tipc_num_links) + 1); -- cgit v1.2.3 From 1ec2bb08407b377e5954b3f9479c2bf67fc925a9 Mon Sep 17 00:00:00 2001 From: Allan Stephens Date: Thu, 27 Oct 2011 15:03:24 -0400 Subject: tipc: Remove obsolete broadcast tag capability Eliminates support for the broadcast tag field, which is no longer used by broadcast link NACK messages. Signed-off-by: Allan Stephens Signed-off-by: Paul Gortmaker --- net/tipc/node.c | 7 +------ 1 file changed, 1 insertion(+), 6 deletions(-) (limited to 'net/tipc/node.c') diff --git a/net/tipc/node.c b/net/tipc/node.c index 6d8bdfd95cd6..7bc45e135fb4 100644 --- a/net/tipc/node.c +++ b/net/tipc/node.c @@ -49,9 +49,8 @@ LIST_HEAD(tipc_node_list); static u32 tipc_num_nodes; static atomic_t tipc_num_links = ATOMIC_INIT(0); -u32 tipc_own_tag; -/** +/* * tipc_node_find - locate specified node object, if it exists */ @@ -309,8 +308,6 @@ static void node_established_contact(struct tipc_node *n_ptr) if (n_ptr->bclink.supportable) { tipc_bclink_add_node(n_ptr->addr); n_ptr->bclink.supported = 1; - if (n_ptr->addr < tipc_own_addr) - tipc_own_tag++; } } @@ -353,8 +350,6 @@ static void node_lost_contact(struct tipc_node *n_ptr) tipc_bclink_remove_node(n_ptr->addr); tipc_bclink_acknowledge(n_ptr, INVALID_LINK_SEQ); - if (n_ptr->addr < tipc_own_addr) - tipc_own_tag--; n_ptr->bclink.supported = 0; } -- cgit v1.2.3 From fc0eea691a06ba8516795fb7a198239fb9db1cfc Mon Sep 17 00:00:00 2001 From: Allan Stephens Date: Fri, 28 Oct 2011 16:26:41 -0400 Subject: tipc: Introduce node signature field in neighbor discovery message Adds support for the new "node signature" in neighbor discovery messages, which is a 16 bit identifier chosen randomly when TIPC is initialized. This field makes it possible for nodes receiving a neighbor discovery message to detect if multiple neighboring nodes are using the same network address (i.e. ), even when the messages are arriving on different interfaces. This first phase of node signature support creates the signature, incorporates it into outgoing neighbor discovery messages, and tracks the signature used by valid neighbors. An upcoming patch builds on this foundation to implement the improved duplicate neighbor detection checking. Signed-off-by: Allan Stephens Signed-off-by: Paul Gortmaker --- net/tipc/node.c | 1 + 1 file changed, 1 insertion(+) (limited to 'net/tipc/node.c') diff --git a/net/tipc/node.c b/net/tipc/node.c index 7bc45e135fb4..153425997cce 100644 --- a/net/tipc/node.c +++ b/net/tipc/node.c @@ -112,6 +112,7 @@ struct tipc_node *tipc_node_create(u32 addr) } list_add_tail(&n_ptr->list, &temp_node->list); n_ptr->block_setup = WAIT_PEER_DOWN; + n_ptr->signature = INVALID_NODE_SIG; tipc_num_nodes++; -- cgit v1.2.3 From a635b46bd884efc1fc98819cb5a200da255d575c Mon Sep 17 00:00:00 2001 From: Allan Stephens Date: Fri, 4 Nov 2011 11:54:43 -0400 Subject: tipc: Hide internal details of node table implementation Relocates information about the size of TIPC's node table index and its associated hash function, since only node subsystem routines need to have access to this information. Note that these changes are essentially cosmetic in nature, and have no impact on the actual operation of TIPC. Signed-off-by: Allan Stephens Signed-off-by: Paul Gortmaker --- net/tipc/node.c | 13 +++++++++++++ 1 file changed, 13 insertions(+) (limited to 'net/tipc/node.c') diff --git a/net/tipc/node.c b/net/tipc/node.c index 153425997cce..1790f503f57b 100644 --- a/net/tipc/node.c +++ b/net/tipc/node.c @@ -39,6 +39,8 @@ #include "node.h" #include "name_distr.h" +#define NODE_HTABLE_SIZE 512 + static void node_lost_contact(struct tipc_node *n_ptr); static void node_established_contact(struct tipc_node *n_ptr); @@ -50,6 +52,17 @@ static u32 tipc_num_nodes; static atomic_t tipc_num_links = ATOMIC_INIT(0); +/* + * A trivial power-of-two bitmask technique is used for speed, since this + * operation is done for every incoming TIPC packet. The number of hash table + * entries has been chosen so that no hash chain exceeds 8 nodes and will + * usually be much smaller (typically only a single node). + */ +static inline unsigned int tipc_hashfn(u32 addr) +{ + return addr & (NODE_HTABLE_SIZE - 1); +} + /* * tipc_node_find - locate specified node object, if it exists */ -- cgit v1.2.3 From 5f6d9123f1c7ef7297b0da1620988fe16c738e75 Mon Sep 17 00:00:00 2001 From: Allan Stephens Date: Fri, 4 Nov 2011 13:24:29 -0400 Subject: tipc: Eliminate trivial buffer manipulation helper routines Gets rid of two inlined routines that simply call existing sk_buff manipulation routines, since there is no longer any extra processing done by the helper routines. Note that these changes are essentially cosmetic in nature, and have no impact on the actual operation of TIPC. Signed-off-by: Allan Stephens Signed-off-by: Paul Gortmaker --- net/tipc/node.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'net/tipc/node.c') diff --git a/net/tipc/node.c b/net/tipc/node.c index 1790f503f57b..24c42f7568ac 100644 --- a/net/tipc/node.c +++ b/net/tipc/node.c @@ -353,12 +353,12 @@ static void node_lost_contact(struct tipc_node *n_ptr) while (n_ptr->bclink.deferred_head) { struct sk_buff *buf = n_ptr->bclink.deferred_head; n_ptr->bclink.deferred_head = buf->next; - buf_discard(buf); + kfree_skb(buf); } n_ptr->bclink.deferred_size = 0; if (n_ptr->bclink.defragm) { - buf_discard(n_ptr->bclink.defragm); + kfree_skb(n_ptr->bclink.defragm); n_ptr->bclink.defragm = NULL; } -- cgit v1.2.3 From 1cc35df847f7dba4ba06cb65bc41713df5d41404 Mon Sep 17 00:00:00 2001 From: Allan Stephens Date: Fri, 4 Nov 2011 14:54:06 -0400 Subject: tipc: Remove obsolete comments about routing table updates Eliminates a block of comments that describe how routing table updates are to be handled. These comments no longer apply following the removal of TIPC's prototype multi-cluster support. Note that these changes are essentially cosmetic in nature, and have no impact on the actual operation of TIPC. Signed-off-by: Allan Stephens Signed-off-by: Paul Gortmaker --- net/tipc/node.c | 46 ---------------------------------------------- 1 file changed, 46 deletions(-) (limited to 'net/tipc/node.c') diff --git a/net/tipc/node.c b/net/tipc/node.c index 24c42f7568ac..d7f8b7be3d8f 100644 --- a/net/tipc/node.c +++ b/net/tipc/node.c @@ -266,52 +266,6 @@ void tipc_node_detach_link(struct tipc_node *n_ptr, struct tipc_link *l_ptr) n_ptr->link_cnt--; } -/* - * Routing table management - five cases to handle: - * - * 1: A link towards a zone/cluster external node comes up. - * => Send a multicast message updating routing tables of all - * system nodes within own cluster that the new destination - * can be reached via this node. - * (node.establishedContact()=>cluster.multicastNewRoute()) - * - * 2: A link towards a slave node comes up. - * => Send a multicast message updating routing tables of all - * system nodes within own cluster that the new destination - * can be reached via this node. - * (node.establishedContact()=>cluster.multicastNewRoute()) - * => Send a message to the slave node about existence - * of all system nodes within cluster: - * (node.establishedContact()=>cluster.sendLocalRoutes()) - * - * 3: A new cluster local system node becomes available. - * => Send message(s) to this particular node containing - * information about all cluster external and slave - * nodes which can be reached via this node. - * (node.establishedContact()==>network.sendExternalRoutes()) - * (node.establishedContact()==>network.sendSlaveRoutes()) - * => Send messages to all directly connected slave nodes - * containing information about the existence of the new node - * (node.establishedContact()=>cluster.multicastNewRoute()) - * - * 4: The link towards a zone/cluster external node or slave - * node goes down. - * => Send a multcast message updating routing tables of all - * nodes within cluster that the new destination can not any - * longer be reached via this node. - * (node.lostAllLinks()=>cluster.bcastLostRoute()) - * - * 5: A cluster local system node becomes unavailable. - * => Remove all references to this node from the local - * routing tables. Note: This is a completely node - * local operation. - * (node.lostAllLinks()=>network.removeAsRouter()) - * => Send messages to all directly connected slave nodes - * containing information about loss of the node - * (node.establishedContact()=>cluster.multicastLostRoute()) - * - */ - static void node_established_contact(struct tipc_node *n_ptr) { tipc_k_signal((Handler)tipc_named_node_up, n_ptr->addr); -- cgit v1.2.3 From 75aba9af2410ae8fc70600d9dcda0651f20e091e Mon Sep 17 00:00:00 2001 From: Allan Stephens Date: Fri, 4 Nov 2011 15:00:02 -0400 Subject: tipc: Minor optimization to broadcast link synchronization logic Optimizes processing done when contact with a neighboring node is established to avoid recording the current state of outgoing broadcast messages if the neighboring node isn't a valid broadcast link destination, since this state information isn't needed for such nodes. Signed-off-by: Allan Stephens Signed-off-by: Paul Gortmaker --- net/tipc/node.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) (limited to 'net/tipc/node.c') diff --git a/net/tipc/node.c b/net/tipc/node.c index d7f8b7be3d8f..12ddc6581572 100644 --- a/net/tipc/node.c +++ b/net/tipc/node.c @@ -270,10 +270,8 @@ static void node_established_contact(struct tipc_node *n_ptr) { tipc_k_signal((Handler)tipc_named_node_up, n_ptr->addr); - /* Syncronize broadcast acks */ - n_ptr->bclink.acked = tipc_bclink_get_last_sent(); - if (n_ptr->bclink.supportable) { + n_ptr->bclink.acked = tipc_bclink_get_last_sent(); tipc_bclink_add_node(n_ptr->addr); n_ptr->bclink.supported = 1; } -- cgit v1.2.3 From b58343f9ea75f02ef48b984767511c6b3ba76eaf Mon Sep 17 00:00:00 2001 From: Allan Stephens Date: Tue, 8 Nov 2011 13:48:28 -0500 Subject: tipc: Eliminate support for tipc_mode global variable Removes all references to the global variable that records whether TIPC is running in "single node" mode or "network" mode, since this information can be easily deduced from the global variable that records TIPC's network address. (i.e. a non-zero network address means that TIPC is running in network mode.) The changes made update most existing mode-based checks to use the network address global variable. A few checks that are no longer needed are removed entirely, along with any associated code lying on non-executable control paths. Signed-off-by: Allan Stephens Signed-off-by: Paul Gortmaker --- net/tipc/node.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'net/tipc/node.c') diff --git a/net/tipc/node.c b/net/tipc/node.c index 12ddc6581572..a34cabc2c43a 100644 --- a/net/tipc/node.c +++ b/net/tipc/node.c @@ -406,7 +406,7 @@ struct sk_buff *tipc_node_get_links(const void *req_tlv_area, int req_tlv_space) return tipc_cfg_reply_error_string(TIPC_CFG_INVALID_VALUE " (network address)"); - if (tipc_mode != TIPC_NET_MODE) + if (!tipc_own_addr) return tipc_cfg_reply_none(); read_lock_bh(&tipc_net_lock); -- cgit v1.2.3