2006-01-03 02:04:38 +08:00
|
|
|
/*
|
|
|
|
* net/tipc/node.c: TIPC node management routines
|
2007-02-09 22:25:21 +08:00
|
|
|
*
|
2006-01-12 02:14:19 +08:00
|
|
|
* Copyright (c) 2000-2006, Ericsson AB
|
2011-01-08 02:00:11 +08:00
|
|
|
* Copyright (c) 2005-2006, 2010-2011, Wind River Systems
|
2006-01-03 02:04:38 +08:00
|
|
|
* All rights reserved.
|
|
|
|
*
|
2006-01-11 20:30:43 +08:00
|
|
|
* Redistribution and use in source and binary forms, with or without
|
2006-01-03 02:04:38 +08:00
|
|
|
* modification, are permitted provided that the following conditions are met:
|
|
|
|
*
|
2006-01-11 20:30:43 +08:00
|
|
|
* 1. Redistributions of source code must retain the above copyright
|
|
|
|
* notice, this list of conditions and the following disclaimer.
|
|
|
|
* 2. Redistributions in binary form must reproduce the above copyright
|
|
|
|
* notice, this list of conditions and the following disclaimer in the
|
|
|
|
* documentation and/or other materials provided with the distribution.
|
|
|
|
* 3. Neither the names of the copyright holders nor the names of its
|
|
|
|
* contributors may be used to endorse or promote products derived from
|
|
|
|
* this software without specific prior written permission.
|
2006-01-03 02:04:38 +08:00
|
|
|
*
|
2006-01-11 20:30:43 +08:00
|
|
|
* Alternatively, this software may be distributed under the terms of the
|
|
|
|
* GNU General Public License ("GPL") version 2 as published by the Free
|
|
|
|
* Software Foundation.
|
|
|
|
*
|
|
|
|
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
|
|
|
|
* AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
|
|
|
|
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
|
|
|
|
* ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
|
|
|
|
* LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
|
|
|
|
* CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
|
|
|
|
* SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
|
|
|
|
* INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
|
|
|
|
* CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
|
|
|
|
* ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
|
2006-01-03 02:04:38 +08:00
|
|
|
* POSSIBILITY OF SUCH DAMAGE.
|
|
|
|
*/
|
|
|
|
|
|
|
|
#include "core.h"
|
|
|
|
#include "config.h"
|
|
|
|
#include "node.h"
|
|
|
|
#include "name_distr.h"
|
|
|
|
|
2011-11-04 23:54:43 +08:00
|
|
|
#define NODE_HTABLE_SIZE 512
|
|
|
|
|
2008-09-03 14:38:32 +08:00
|
|
|
static void node_lost_contact(struct tipc_node *n_ptr);
|
|
|
|
static void node_established_contact(struct tipc_node *n_ptr);
|
2006-01-03 02:04:38 +08:00
|
|
|
|
2008-05-22 05:53:00 +08:00
|
|
|
static DEFINE_SPINLOCK(node_create_lock);
|
|
|
|
|
2011-02-26 07:42:52 +08:00
|
|
|
static struct hlist_head node_htable[NODE_HTABLE_SIZE];
|
|
|
|
LIST_HEAD(tipc_node_list);
|
|
|
|
static u32 tipc_num_nodes;
|
2011-02-26 08:11:25 +08:00
|
|
|
|
|
|
|
static atomic_t tipc_num_links = ATOMIC_INIT(0);
|
2006-01-03 02:04:38 +08:00
|
|
|
|
2011-11-04 23:54:43 +08:00
|
|
|
/*
|
|
|
|
* 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).
|
|
|
|
*/
|
2012-04-23 12:49:13 +08:00
|
|
|
static unsigned int tipc_hashfn(u32 addr)
|
2011-11-04 23:54:43 +08:00
|
|
|
{
|
|
|
|
return addr & (NODE_HTABLE_SIZE - 1);
|
|
|
|
}
|
|
|
|
|
2011-10-28 03:03:24 +08:00
|
|
|
/*
|
2011-02-26 07:42:52 +08:00
|
|
|
* tipc_node_find - locate specified node object, if it exists
|
|
|
|
*/
|
|
|
|
struct tipc_node *tipc_node_find(u32 addr)
|
|
|
|
{
|
|
|
|
struct tipc_node *node;
|
|
|
|
struct hlist_node *pos;
|
|
|
|
|
2012-04-18 06:02:01 +08:00
|
|
|
if (unlikely(!in_own_cluster_exact(addr)))
|
2011-02-26 07:42:52 +08:00
|
|
|
return NULL;
|
|
|
|
|
|
|
|
hlist_for_each_entry(node, pos, &node_htable[tipc_hashfn(addr)], hash) {
|
|
|
|
if (node->addr == addr)
|
|
|
|
return node;
|
|
|
|
}
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
2008-05-22 05:53:00 +08:00
|
|
|
/**
|
|
|
|
* tipc_node_create - create neighboring node
|
|
|
|
*
|
|
|
|
* Currently, this routine is called by neighbor discovery code, which holds
|
|
|
|
* net_lock for reading only. We must take node_create_lock to ensure a node
|
|
|
|
* isn't created twice if two different bearers discover the node at the same
|
|
|
|
* time. (It would be preferable to switch to holding net_lock in write mode,
|
|
|
|
* but this is a non-trivial change.)
|
|
|
|
*/
|
2008-09-03 14:38:32 +08:00
|
|
|
struct tipc_node *tipc_node_create(u32 addr)
|
2006-01-03 02:04:38 +08:00
|
|
|
{
|
2011-02-26 07:42:52 +08:00
|
|
|
struct tipc_node *n_ptr, *temp_node;
|
2006-01-03 02:04:38 +08:00
|
|
|
|
2008-05-22 05:53:00 +08:00
|
|
|
spin_lock_bh(&node_create_lock);
|
|
|
|
|
2011-01-01 02:59:23 +08:00
|
|
|
n_ptr = tipc_node_find(addr);
|
|
|
|
if (n_ptr) {
|
|
|
|
spin_unlock_bh(&node_create_lock);
|
|
|
|
return n_ptr;
|
2008-05-22 05:53:00 +08:00
|
|
|
}
|
|
|
|
|
2011-01-01 02:59:23 +08:00
|
|
|
n_ptr = kzalloc(sizeof(*n_ptr), GFP_ATOMIC);
|
2006-06-26 14:52:17 +08:00
|
|
|
if (!n_ptr) {
|
2008-05-22 05:53:00 +08:00
|
|
|
spin_unlock_bh(&node_create_lock);
|
2012-06-29 12:16:37 +08:00
|
|
|
pr_warn("Node creation failed, no memory\n");
|
2006-06-26 14:52:17 +08:00
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
n_ptr->addr = addr;
|
2011-01-01 02:59:18 +08:00
|
|
|
spin_lock_init(&n_ptr->lock);
|
2011-02-26 07:42:52 +08:00
|
|
|
INIT_HLIST_NODE(&n_ptr->hash);
|
|
|
|
INIT_LIST_HEAD(&n_ptr->list);
|
2006-06-26 14:52:17 +08:00
|
|
|
INIT_LIST_HEAD(&n_ptr->nsub);
|
2011-01-01 02:59:19 +08:00
|
|
|
|
2011-02-26 07:42:52 +08:00
|
|
|
hlist_add_head(&n_ptr->hash, &node_htable[tipc_hashfn(addr)]);
|
|
|
|
|
|
|
|
list_for_each_entry(temp_node, &tipc_node_list, list) {
|
|
|
|
if (n_ptr->addr < temp_node->addr)
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
list_add_tail(&n_ptr->list, &temp_node->list);
|
tipc: Ensure both nodes recognize loss of contact between them
Enhances TIPC to ensure that a node that loses contact with a
neighboring node does not allow contact to be re-established until
it sees that its peer has also recognized the loss of contact.
Previously, nodes that were connected by two or more links could
encounter a situation in which node A would lose contact with node B
on all of its links, purge its name table of names published by B,
and then fail to repopulate those names once contact with B was restored.
This would happen because B was able to re-establish one or more links
so quickly that it never reached a point where it had no links to A --
meaning that B never saw a loss of contact with A, and consequently
didn't re-publish its names to A.
This problem is now prevented by enhancing the cleanup done by TIPC
following a loss of contact with a neighboring node to ensure that
node A ignores all messages sent by B until it receives a LINK_PROTOCOL
message that indicates B has lost contact with A, thereby preventing
the (re)establishment of links between the nodes. The loss of contact
is recognized when a RESET or ACTIVATE message is received that has
a "redundant link exists" field of 0, indicating that B's sending link
endpoint is in a reset state and that B has no other working links.
Additionally, TIPC now suppresses the sending of (most) link protocol
messages to a neighboring node while it is cleaning up after an earlier
loss of contact with that node. This stops the peer node from prematurely
activating its link endpoint, which would prevent TIPC from later
activating its own end. TIPC still allows outgoing RESET messages to
occur during cleanup, to avoid problems if its own node recognizes
the loss of contact first and tries to notify the peer of the situation.
Finally, TIPC now recognizes an impending loss of contact with a peer node
as soon as it receives a RESET message on a working link that is the
peer's only link to the node, and ensures that the link protocol
suppression mentioned above goes into effect right away -- that is,
even before its own link endpoints have failed. This is necessary to
ensure correct operation when there are redundant links between the nodes,
since otherwise TIPC would send an ACTIVATE message upon receiving a RESET
on its first link and only begin suppressing when a RESET on its second
link was received, instead of initiating suppression with the first RESET
message as it needs to.
Note: The reworked cleanup code also eliminates a check that prevented
a link endpoint's discovery object from responding to incoming messages
while stale name table entries are being purged. This check is now
unnecessary and would have slowed down re-establishment of communication
between the nodes in some situations.
Signed-off-by: Allan Stephens <allan.stephens@windriver.com>
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
2011-05-27 23:00:51 +08:00
|
|
|
n_ptr->block_setup = WAIT_PEER_DOWN;
|
2011-10-29 04:26:41 +08:00
|
|
|
n_ptr->signature = INVALID_NODE_SIG;
|
2011-02-26 07:42:52 +08:00
|
|
|
|
|
|
|
tipc_num_nodes++;
|
2006-06-26 14:52:17 +08:00
|
|
|
|
2008-05-22 05:53:00 +08:00
|
|
|
spin_unlock_bh(&node_create_lock);
|
2006-01-03 02:04:38 +08:00
|
|
|
return n_ptr;
|
|
|
|
}
|
|
|
|
|
2008-09-03 14:38:32 +08:00
|
|
|
void tipc_node_delete(struct tipc_node *n_ptr)
|
2006-01-03 02:04:38 +08:00
|
|
|
{
|
2011-02-26 07:42:52 +08:00
|
|
|
list_del(&n_ptr->list);
|
|
|
|
hlist_del(&n_ptr->hash);
|
2006-01-03 02:04:38 +08:00
|
|
|
kfree(n_ptr);
|
2011-01-01 02:59:19 +08:00
|
|
|
|
2011-02-26 07:42:52 +08:00
|
|
|
tipc_num_nodes--;
|
2006-01-03 02:04:38 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
2006-01-18 07:38:21 +08:00
|
|
|
* tipc_node_link_up - handle addition of link
|
2007-02-09 22:25:21 +08:00
|
|
|
*
|
2006-01-03 02:04:38 +08:00
|
|
|
* Link becomes active (alone or shared) or standby, depending on its priority.
|
|
|
|
*/
|
2011-12-30 09:58:42 +08:00
|
|
|
void tipc_node_link_up(struct tipc_node *n_ptr, struct tipc_link *l_ptr)
|
2006-01-03 02:04:38 +08:00
|
|
|
{
|
2011-12-30 09:58:42 +08:00
|
|
|
struct tipc_link **active = &n_ptr->active_links[0];
|
2006-01-03 02:04:38 +08:00
|
|
|
|
2006-06-26 14:52:50 +08:00
|
|
|
n_ptr->working_links++;
|
|
|
|
|
2012-06-29 12:16:37 +08:00
|
|
|
pr_info("Established link <%s> on network plane %c\n",
|
|
|
|
l_ptr->name, l_ptr->b_ptr->net_plane);
|
2007-02-09 22:25:21 +08:00
|
|
|
|
2006-01-03 02:04:38 +08:00
|
|
|
if (!active[0]) {
|
|
|
|
active[0] = active[1] = l_ptr;
|
|
|
|
node_established_contact(n_ptr);
|
|
|
|
return;
|
|
|
|
}
|
2007-02-09 22:25:21 +08:00
|
|
|
if (l_ptr->priority < active[0]->priority) {
|
2012-06-29 12:16:37 +08:00
|
|
|
pr_info("New link <%s> becomes standby\n", l_ptr->name);
|
2006-01-03 02:04:38 +08:00
|
|
|
return;
|
|
|
|
}
|
2006-01-18 07:38:21 +08:00
|
|
|
tipc_link_send_duplicate(active[0], l_ptr);
|
2007-02-09 22:25:21 +08:00
|
|
|
if (l_ptr->priority == active[0]->priority) {
|
2006-01-03 02:04:38 +08:00
|
|
|
active[0] = l_ptr;
|
|
|
|
return;
|
|
|
|
}
|
2012-06-29 12:16:37 +08:00
|
|
|
pr_info("Old link <%s> becomes standby\n", active[0]->name);
|
2006-06-26 14:52:17 +08:00
|
|
|
if (active[1] != active[0])
|
2012-06-29 12:16:37 +08:00
|
|
|
pr_info("Old link <%s> becomes standby\n", active[1]->name);
|
2006-01-03 02:04:38 +08:00
|
|
|
active[0] = active[1] = l_ptr;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* node_select_active_links - select active link
|
|
|
|
*/
|
2008-09-03 14:38:32 +08:00
|
|
|
static void node_select_active_links(struct tipc_node *n_ptr)
|
2006-01-03 02:04:38 +08:00
|
|
|
{
|
2011-12-30 09:58:42 +08:00
|
|
|
struct tipc_link **active = &n_ptr->active_links[0];
|
2006-01-03 02:04:38 +08:00
|
|
|
u32 i;
|
|
|
|
u32 highest_prio = 0;
|
|
|
|
|
2007-02-09 22:25:21 +08:00
|
|
|
active[0] = active[1] = NULL;
|
2006-01-03 02:04:38 +08:00
|
|
|
|
|
|
|
for (i = 0; i < MAX_BEARERS; i++) {
|
2011-12-30 09:58:42 +08:00
|
|
|
struct tipc_link *l_ptr = n_ptr->links[i];
|
2006-01-03 02:04:38 +08:00
|
|
|
|
2006-01-18 07:38:21 +08:00
|
|
|
if (!l_ptr || !tipc_link_is_up(l_ptr) ||
|
2006-01-03 02:04:38 +08:00
|
|
|
(l_ptr->priority < highest_prio))
|
|
|
|
continue;
|
|
|
|
|
|
|
|
if (l_ptr->priority > highest_prio) {
|
2007-02-09 22:25:21 +08:00
|
|
|
highest_prio = l_ptr->priority;
|
2006-01-03 02:04:38 +08:00
|
|
|
active[0] = active[1] = l_ptr;
|
|
|
|
} else {
|
|
|
|
active[1] = l_ptr;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
2006-01-18 07:38:21 +08:00
|
|
|
* tipc_node_link_down - handle loss of link
|
2006-01-03 02:04:38 +08:00
|
|
|
*/
|
2011-12-30 09:58:42 +08:00
|
|
|
void tipc_node_link_down(struct tipc_node *n_ptr, struct tipc_link *l_ptr)
|
2006-01-03 02:04:38 +08:00
|
|
|
{
|
2011-12-30 09:58:42 +08:00
|
|
|
struct tipc_link **active;
|
2006-01-03 02:04:38 +08:00
|
|
|
|
2006-06-26 14:52:50 +08:00
|
|
|
n_ptr->working_links--;
|
|
|
|
|
2006-01-18 07:38:21 +08:00
|
|
|
if (!tipc_link_is_active(l_ptr)) {
|
2012-06-29 12:16:37 +08:00
|
|
|
pr_info("Lost standby link <%s> on network plane %c\n",
|
|
|
|
l_ptr->name, l_ptr->b_ptr->net_plane);
|
2006-01-03 02:04:38 +08:00
|
|
|
return;
|
|
|
|
}
|
2012-06-29 12:16:37 +08:00
|
|
|
pr_info("Lost link <%s> on network plane %c\n",
|
2006-01-03 02:04:38 +08:00
|
|
|
l_ptr->name, l_ptr->b_ptr->net_plane);
|
|
|
|
|
|
|
|
active = &n_ptr->active_links[0];
|
|
|
|
if (active[0] == l_ptr)
|
|
|
|
active[0] = active[1];
|
|
|
|
if (active[1] == l_ptr)
|
|
|
|
active[1] = active[0];
|
|
|
|
if (active[0] == l_ptr)
|
|
|
|
node_select_active_links(n_ptr);
|
2007-02-09 22:25:21 +08:00
|
|
|
if (tipc_node_is_up(n_ptr))
|
2006-01-18 07:38:21 +08:00
|
|
|
tipc_link_changeover(l_ptr);
|
2007-02-09 22:25:21 +08:00
|
|
|
else
|
2006-01-03 02:04:38 +08:00
|
|
|
node_lost_contact(n_ptr);
|
|
|
|
}
|
|
|
|
|
2011-02-28 23:36:21 +08:00
|
|
|
int tipc_node_active_links(struct tipc_node *n_ptr)
|
2006-01-03 02:04:38 +08:00
|
|
|
{
|
2010-08-17 19:00:12 +08:00
|
|
|
return n_ptr->active_links[0] != NULL;
|
2006-01-03 02:04:38 +08:00
|
|
|
}
|
|
|
|
|
2011-02-28 23:36:21 +08:00
|
|
|
int tipc_node_redundant_links(struct tipc_node *n_ptr)
|
2006-01-03 02:04:38 +08:00
|
|
|
{
|
2010-09-23 04:43:57 +08:00
|
|
|
return n_ptr->working_links > 1;
|
2006-01-03 02:04:38 +08:00
|
|
|
}
|
|
|
|
|
2008-09-03 14:38:32 +08:00
|
|
|
int tipc_node_is_up(struct tipc_node *n_ptr)
|
2006-01-03 02:04:38 +08:00
|
|
|
{
|
2011-02-28 23:36:21 +08:00
|
|
|
return tipc_node_active_links(n_ptr);
|
2006-01-03 02:04:38 +08:00
|
|
|
}
|
|
|
|
|
2011-12-30 09:58:42 +08:00
|
|
|
void tipc_node_attach_link(struct tipc_node *n_ptr, struct tipc_link *l_ptr)
|
2006-01-03 02:04:38 +08:00
|
|
|
{
|
2011-03-01 00:32:27 +08:00
|
|
|
n_ptr->links[l_ptr->b_ptr->identity] = l_ptr;
|
|
|
|
atomic_inc(&tipc_num_links);
|
|
|
|
n_ptr->link_cnt++;
|
2006-01-03 02:04:38 +08:00
|
|
|
}
|
|
|
|
|
2011-12-30 09:58:42 +08:00
|
|
|
void tipc_node_detach_link(struct tipc_node *n_ptr, struct tipc_link *l_ptr)
|
2006-01-03 02:04:38 +08:00
|
|
|
{
|
2006-03-21 14:36:47 +08:00
|
|
|
n_ptr->links[l_ptr->b_ptr->identity] = NULL;
|
2011-02-25 23:01:58 +08:00
|
|
|
atomic_dec(&tipc_num_links);
|
2006-01-03 02:04:38 +08:00
|
|
|
n_ptr->link_cnt--;
|
|
|
|
}
|
|
|
|
|
2008-09-03 14:38:32 +08:00
|
|
|
static void node_established_contact(struct tipc_node *n_ptr)
|
2006-01-03 02:04:38 +08:00
|
|
|
{
|
2011-01-01 02:59:18 +08:00
|
|
|
tipc_k_signal((Handler)tipc_named_node_up, n_ptr->addr);
|
2006-01-03 02:04:38 +08:00
|
|
|
|
2012-11-16 13:51:29 +08:00
|
|
|
n_ptr->bclink.acked = tipc_bclink_get_last_sent();
|
|
|
|
tipc_bclink_add_node(n_ptr->addr);
|
|
|
|
n_ptr->bclink.supported = 1;
|
2006-01-03 02:04:38 +08:00
|
|
|
}
|
|
|
|
|
tipc: Ensure both nodes recognize loss of contact between them
Enhances TIPC to ensure that a node that loses contact with a
neighboring node does not allow contact to be re-established until
it sees that its peer has also recognized the loss of contact.
Previously, nodes that were connected by two or more links could
encounter a situation in which node A would lose contact with node B
on all of its links, purge its name table of names published by B,
and then fail to repopulate those names once contact with B was restored.
This would happen because B was able to re-establish one or more links
so quickly that it never reached a point where it had no links to A --
meaning that B never saw a loss of contact with A, and consequently
didn't re-publish its names to A.
This problem is now prevented by enhancing the cleanup done by TIPC
following a loss of contact with a neighboring node to ensure that
node A ignores all messages sent by B until it receives a LINK_PROTOCOL
message that indicates B has lost contact with A, thereby preventing
the (re)establishment of links between the nodes. The loss of contact
is recognized when a RESET or ACTIVATE message is received that has
a "redundant link exists" field of 0, indicating that B's sending link
endpoint is in a reset state and that B has no other working links.
Additionally, TIPC now suppresses the sending of (most) link protocol
messages to a neighboring node while it is cleaning up after an earlier
loss of contact with that node. This stops the peer node from prematurely
activating its link endpoint, which would prevent TIPC from later
activating its own end. TIPC still allows outgoing RESET messages to
occur during cleanup, to avoid problems if its own node recognizes
the loss of contact first and tries to notify the peer of the situation.
Finally, TIPC now recognizes an impending loss of contact with a peer node
as soon as it receives a RESET message on a working link that is the
peer's only link to the node, and ensures that the link protocol
suppression mentioned above goes into effect right away -- that is,
even before its own link endpoints have failed. This is necessary to
ensure correct operation when there are redundant links between the nodes,
since otherwise TIPC would send an ACTIVATE message upon receiving a RESET
on its first link and only begin suppressing when a RESET on its second
link was received, instead of initiating suppression with the first RESET
message as it needs to.
Note: The reworked cleanup code also eliminates a check that prevented
a link endpoint's discovery object from responding to incoming messages
while stale name table entries are being purged. This check is now
unnecessary and would have slowed down re-establishment of communication
between the nodes in some situations.
Signed-off-by: Allan Stephens <allan.stephens@windriver.com>
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
2011-05-27 23:00:51 +08:00
|
|
|
static void node_name_purge_complete(unsigned long node_addr)
|
2010-08-17 19:00:16 +08:00
|
|
|
{
|
|
|
|
struct tipc_node *n_ptr;
|
|
|
|
|
|
|
|
read_lock_bh(&tipc_net_lock);
|
|
|
|
n_ptr = tipc_node_find(node_addr);
|
|
|
|
if (n_ptr) {
|
|
|
|
tipc_node_lock(n_ptr);
|
tipc: Ensure both nodes recognize loss of contact between them
Enhances TIPC to ensure that a node that loses contact with a
neighboring node does not allow contact to be re-established until
it sees that its peer has also recognized the loss of contact.
Previously, nodes that were connected by two or more links could
encounter a situation in which node A would lose contact with node B
on all of its links, purge its name table of names published by B,
and then fail to repopulate those names once contact with B was restored.
This would happen because B was able to re-establish one or more links
so quickly that it never reached a point where it had no links to A --
meaning that B never saw a loss of contact with A, and consequently
didn't re-publish its names to A.
This problem is now prevented by enhancing the cleanup done by TIPC
following a loss of contact with a neighboring node to ensure that
node A ignores all messages sent by B until it receives a LINK_PROTOCOL
message that indicates B has lost contact with A, thereby preventing
the (re)establishment of links between the nodes. The loss of contact
is recognized when a RESET or ACTIVATE message is received that has
a "redundant link exists" field of 0, indicating that B's sending link
endpoint is in a reset state and that B has no other working links.
Additionally, TIPC now suppresses the sending of (most) link protocol
messages to a neighboring node while it is cleaning up after an earlier
loss of contact with that node. This stops the peer node from prematurely
activating its link endpoint, which would prevent TIPC from later
activating its own end. TIPC still allows outgoing RESET messages to
occur during cleanup, to avoid problems if its own node recognizes
the loss of contact first and tries to notify the peer of the situation.
Finally, TIPC now recognizes an impending loss of contact with a peer node
as soon as it receives a RESET message on a working link that is the
peer's only link to the node, and ensures that the link protocol
suppression mentioned above goes into effect right away -- that is,
even before its own link endpoints have failed. This is necessary to
ensure correct operation when there are redundant links between the nodes,
since otherwise TIPC would send an ACTIVATE message upon receiving a RESET
on its first link and only begin suppressing when a RESET on its second
link was received, instead of initiating suppression with the first RESET
message as it needs to.
Note: The reworked cleanup code also eliminates a check that prevented
a link endpoint's discovery object from responding to incoming messages
while stale name table entries are being purged. This check is now
unnecessary and would have slowed down re-establishment of communication
between the nodes in some situations.
Signed-off-by: Allan Stephens <allan.stephens@windriver.com>
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
2011-05-27 23:00:51 +08:00
|
|
|
n_ptr->block_setup &= ~WAIT_NAMES_GONE;
|
2010-08-17 19:00:16 +08:00
|
|
|
tipc_node_unlock(n_ptr);
|
|
|
|
}
|
|
|
|
read_unlock_bh(&tipc_net_lock);
|
|
|
|
}
|
|
|
|
|
2008-09-03 14:38:32 +08:00
|
|
|
static void node_lost_contact(struct tipc_node *n_ptr)
|
2006-01-03 02:04:38 +08:00
|
|
|
{
|
|
|
|
char addr_string[16];
|
|
|
|
u32 i;
|
|
|
|
|
2012-06-29 12:16:37 +08:00
|
|
|
pr_info("Lost contact with %s\n",
|
|
|
|
tipc_addr_string_fill(addr_string, n_ptr->addr));
|
2011-04-07 23:58:08 +08:00
|
|
|
|
|
|
|
/* Flush broadcast link info associated with lost node */
|
2011-01-01 02:59:18 +08:00
|
|
|
if (n_ptr->bclink.supported) {
|
2011-04-07 23:58:08 +08:00
|
|
|
while (n_ptr->bclink.deferred_head) {
|
|
|
|
struct sk_buff *buf = n_ptr->bclink.deferred_head;
|
|
|
|
n_ptr->bclink.deferred_head = buf->next;
|
2011-11-05 01:24:29 +08:00
|
|
|
kfree_skb(buf);
|
2011-04-07 23:58:08 +08:00
|
|
|
}
|
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 <allan.stephens@windriver.com>
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
2011-10-28 02:17:53 +08:00
|
|
|
n_ptr->bclink.deferred_size = 0;
|
2011-04-07 23:58:08 +08:00
|
|
|
|
|
|
|
if (n_ptr->bclink.defragm) {
|
2011-11-05 01:24:29 +08:00
|
|
|
kfree_skb(n_ptr->bclink.defragm);
|
2011-04-07 23:58:08 +08:00
|
|
|
n_ptr->bclink.defragm = NULL;
|
|
|
|
}
|
|
|
|
|
2011-10-24 23:18:12 +08:00
|
|
|
tipc_bclink_remove_node(n_ptr->addr);
|
2011-10-25 03:26:24 +08:00
|
|
|
tipc_bclink_acknowledge(n_ptr, INVALID_LINK_SEQ);
|
2006-01-03 02:04:38 +08:00
|
|
|
|
2011-04-07 23:58:08 +08:00
|
|
|
n_ptr->bclink.supported = 0;
|
|
|
|
}
|
2006-01-03 02:04:38 +08:00
|
|
|
|
|
|
|
/* Abort link changeover */
|
|
|
|
for (i = 0; i < MAX_BEARERS; i++) {
|
2011-12-30 09:58:42 +08:00
|
|
|
struct tipc_link *l_ptr = n_ptr->links[i];
|
2007-02-09 22:25:21 +08:00
|
|
|
if (!l_ptr)
|
2006-01-03 02:04:38 +08:00
|
|
|
continue;
|
|
|
|
l_ptr->reset_checkpoint = l_ptr->next_in_no;
|
|
|
|
l_ptr->exp_msg_count = 0;
|
2006-01-18 07:38:21 +08:00
|
|
|
tipc_link_reset_fragments(l_ptr);
|
2006-01-03 02:04:38 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
/* Notify subscribers */
|
2011-02-24 03:13:41 +08:00
|
|
|
tipc_nodesub_notify(n_ptr);
|
2010-08-17 19:00:16 +08:00
|
|
|
|
tipc: Ensure both nodes recognize loss of contact between them
Enhances TIPC to ensure that a node that loses contact with a
neighboring node does not allow contact to be re-established until
it sees that its peer has also recognized the loss of contact.
Previously, nodes that were connected by two or more links could
encounter a situation in which node A would lose contact with node B
on all of its links, purge its name table of names published by B,
and then fail to repopulate those names once contact with B was restored.
This would happen because B was able to re-establish one or more links
so quickly that it never reached a point where it had no links to A --
meaning that B never saw a loss of contact with A, and consequently
didn't re-publish its names to A.
This problem is now prevented by enhancing the cleanup done by TIPC
following a loss of contact with a neighboring node to ensure that
node A ignores all messages sent by B until it receives a LINK_PROTOCOL
message that indicates B has lost contact with A, thereby preventing
the (re)establishment of links between the nodes. The loss of contact
is recognized when a RESET or ACTIVATE message is received that has
a "redundant link exists" field of 0, indicating that B's sending link
endpoint is in a reset state and that B has no other working links.
Additionally, TIPC now suppresses the sending of (most) link protocol
messages to a neighboring node while it is cleaning up after an earlier
loss of contact with that node. This stops the peer node from prematurely
activating its link endpoint, which would prevent TIPC from later
activating its own end. TIPC still allows outgoing RESET messages to
occur during cleanup, to avoid problems if its own node recognizes
the loss of contact first and tries to notify the peer of the situation.
Finally, TIPC now recognizes an impending loss of contact with a peer node
as soon as it receives a RESET message on a working link that is the
peer's only link to the node, and ensures that the link protocol
suppression mentioned above goes into effect right away -- that is,
even before its own link endpoints have failed. This is necessary to
ensure correct operation when there are redundant links between the nodes,
since otherwise TIPC would send an ACTIVATE message upon receiving a RESET
on its first link and only begin suppressing when a RESET on its second
link was received, instead of initiating suppression with the first RESET
message as it needs to.
Note: The reworked cleanup code also eliminates a check that prevented
a link endpoint's discovery object from responding to incoming messages
while stale name table entries are being purged. This check is now
unnecessary and would have slowed down re-establishment of communication
between the nodes in some situations.
Signed-off-by: Allan Stephens <allan.stephens@windriver.com>
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
2011-05-27 23:00:51 +08:00
|
|
|
/* Prevent re-contact with node until cleanup is done */
|
|
|
|
n_ptr->block_setup = WAIT_PEER_DOWN | WAIT_NAMES_GONE;
|
|
|
|
tipc_k_signal((Handler)node_name_purge_complete, n_ptr->addr);
|
2006-01-03 02:04:38 +08:00
|
|
|
}
|
|
|
|
|
2006-01-18 07:38:21 +08:00
|
|
|
struct sk_buff *tipc_node_get_nodes(const void *req_tlv_area, int req_tlv_space)
|
2006-01-03 02:04:38 +08:00
|
|
|
{
|
|
|
|
u32 domain;
|
|
|
|
struct sk_buff *buf;
|
2008-09-03 14:38:32 +08:00
|
|
|
struct tipc_node *n_ptr;
|
2007-02-09 22:25:21 +08:00
|
|
|
struct tipc_node_info node_info;
|
2006-06-30 03:33:20 +08:00
|
|
|
u32 payload_size;
|
2006-01-03 02:04:38 +08:00
|
|
|
|
|
|
|
if (!TLV_CHECK(req_tlv_area, req_tlv_space, TIPC_TLV_NET_ADDR))
|
2006-01-18 07:38:21 +08:00
|
|
|
return tipc_cfg_reply_error_string(TIPC_CFG_TLV_ERROR);
|
2006-01-03 02:04:38 +08:00
|
|
|
|
2006-11-08 16:19:09 +08:00
|
|
|
domain = ntohl(*(__be32 *)TLV_DATA(req_tlv_area));
|
2006-01-18 07:38:21 +08:00
|
|
|
if (!tipc_addr_domain_valid(domain))
|
|
|
|
return tipc_cfg_reply_error_string(TIPC_CFG_INVALID_VALUE
|
|
|
|
" (network address)");
|
2006-01-03 02:04:38 +08:00
|
|
|
|
2008-07-15 13:44:58 +08:00
|
|
|
read_lock_bh(&tipc_net_lock);
|
2011-02-26 07:42:52 +08:00
|
|
|
if (!tipc_num_nodes) {
|
2008-07-15 13:44:58 +08:00
|
|
|
read_unlock_bh(&tipc_net_lock);
|
2007-02-09 22:25:21 +08:00
|
|
|
return tipc_cfg_reply_none();
|
2008-07-15 13:44:58 +08:00
|
|
|
}
|
2006-01-03 02:04:38 +08:00
|
|
|
|
2011-01-01 02:59:17 +08:00
|
|
|
/* For now, get space for all other nodes */
|
2011-02-26 07:42:52 +08:00
|
|
|
payload_size = TLV_SPACE(sizeof(node_info)) * tipc_num_nodes;
|
2008-07-15 13:44:58 +08:00
|
|
|
if (payload_size > 32768u) {
|
|
|
|
read_unlock_bh(&tipc_net_lock);
|
2006-06-30 03:33:20 +08:00
|
|
|
return tipc_cfg_reply_error_string(TIPC_CFG_NOT_SUPPORTED
|
|
|
|
" (too many nodes)");
|
2008-07-15 13:44:58 +08:00
|
|
|
}
|
2006-06-30 03:33:20 +08:00
|
|
|
buf = tipc_cfg_reply_alloc(payload_size);
|
2008-07-15 13:44:58 +08:00
|
|
|
if (!buf) {
|
|
|
|
read_unlock_bh(&tipc_net_lock);
|
2006-01-03 02:04:38 +08:00
|
|
|
return NULL;
|
2008-07-15 13:44:58 +08:00
|
|
|
}
|
2006-01-03 02:04:38 +08:00
|
|
|
|
|
|
|
/* Add TLVs for all nodes in scope */
|
2011-02-26 07:42:52 +08:00
|
|
|
list_for_each_entry(n_ptr, &tipc_node_list, list) {
|
|
|
|
if (!tipc_in_scope(domain, n_ptr->addr))
|
2006-01-03 02:04:38 +08:00
|
|
|
continue;
|
2007-02-09 22:25:21 +08:00
|
|
|
node_info.addr = htonl(n_ptr->addr);
|
|
|
|
node_info.up = htonl(tipc_node_is_up(n_ptr));
|
|
|
|
tipc_cfg_append_tlv(buf, TIPC_TLV_NODE_INFO,
|
2006-01-18 07:38:21 +08:00
|
|
|
&node_info, sizeof(node_info));
|
2006-01-03 02:04:38 +08:00
|
|
|
}
|
|
|
|
|
2008-07-15 13:44:58 +08:00
|
|
|
read_unlock_bh(&tipc_net_lock);
|
2006-01-03 02:04:38 +08:00
|
|
|
return buf;
|
|
|
|
}
|
|
|
|
|
2006-01-18 07:38:21 +08:00
|
|
|
struct sk_buff *tipc_node_get_links(const void *req_tlv_area, int req_tlv_space)
|
2006-01-03 02:04:38 +08:00
|
|
|
{
|
|
|
|
u32 domain;
|
|
|
|
struct sk_buff *buf;
|
2008-09-03 14:38:32 +08:00
|
|
|
struct tipc_node *n_ptr;
|
2007-02-09 22:25:21 +08:00
|
|
|
struct tipc_link_info link_info;
|
2006-06-30 03:33:20 +08:00
|
|
|
u32 payload_size;
|
2006-01-03 02:04:38 +08:00
|
|
|
|
|
|
|
if (!TLV_CHECK(req_tlv_area, req_tlv_space, TIPC_TLV_NET_ADDR))
|
2006-01-18 07:38:21 +08:00
|
|
|
return tipc_cfg_reply_error_string(TIPC_CFG_TLV_ERROR);
|
2006-01-03 02:04:38 +08:00
|
|
|
|
2006-11-08 16:19:09 +08:00
|
|
|
domain = ntohl(*(__be32 *)TLV_DATA(req_tlv_area));
|
2006-01-18 07:38:21 +08:00
|
|
|
if (!tipc_addr_domain_valid(domain))
|
|
|
|
return tipc_cfg_reply_error_string(TIPC_CFG_INVALID_VALUE
|
|
|
|
" (network address)");
|
2006-01-03 02:04:38 +08:00
|
|
|
|
2011-11-09 02:48:28 +08:00
|
|
|
if (!tipc_own_addr)
|
2007-02-09 22:25:21 +08:00
|
|
|
return tipc_cfg_reply_none();
|
|
|
|
|
2008-07-15 13:44:58 +08:00
|
|
|
read_lock_bh(&tipc_net_lock);
|
|
|
|
|
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 <allan.stephens@windriver.com>
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
2011-10-28 02:17:53 +08:00
|
|
|
/* Get space for all unicast links + broadcast link */
|
2011-02-25 02:20:20 +08:00
|
|
|
payload_size = TLV_SPACE(sizeof(link_info)) *
|
2011-02-25 23:01:58 +08:00
|
|
|
(atomic_read(&tipc_num_links) + 1);
|
2008-07-15 13:44:58 +08:00
|
|
|
if (payload_size > 32768u) {
|
|
|
|
read_unlock_bh(&tipc_net_lock);
|
2006-06-30 03:33:20 +08:00
|
|
|
return tipc_cfg_reply_error_string(TIPC_CFG_NOT_SUPPORTED
|
|
|
|
" (too many links)");
|
2008-07-15 13:44:58 +08:00
|
|
|
}
|
2006-06-30 03:33:20 +08:00
|
|
|
buf = tipc_cfg_reply_alloc(payload_size);
|
2008-07-15 13:44:58 +08:00
|
|
|
if (!buf) {
|
|
|
|
read_unlock_bh(&tipc_net_lock);
|
2006-01-03 02:04:38 +08:00
|
|
|
return NULL;
|
2008-07-15 13:44:58 +08:00
|
|
|
}
|
2006-01-03 02:04:38 +08:00
|
|
|
|
|
|
|
/* Add TLV for broadcast link */
|
2011-02-24 00:44:49 +08:00
|
|
|
link_info.dest = htonl(tipc_cluster_mask(tipc_own_addr));
|
2007-02-09 22:25:21 +08:00
|
|
|
link_info.up = htonl(1);
|
2009-03-19 10:11:29 +08:00
|
|
|
strlcpy(link_info.str, tipc_bclink_name, TIPC_MAX_LINK_NAME);
|
2006-01-18 07:38:21 +08:00
|
|
|
tipc_cfg_append_tlv(buf, TIPC_TLV_LINK_INFO, &link_info, sizeof(link_info));
|
2006-01-03 02:04:38 +08:00
|
|
|
|
|
|
|
/* Add TLVs for any other links in scope */
|
2011-02-26 07:42:52 +08:00
|
|
|
list_for_each_entry(n_ptr, &tipc_node_list, list) {
|
2007-02-09 22:25:21 +08:00
|
|
|
u32 i;
|
2006-01-03 02:04:38 +08:00
|
|
|
|
2011-02-26 07:42:52 +08:00
|
|
|
if (!tipc_in_scope(domain, n_ptr->addr))
|
2006-01-03 02:04:38 +08:00
|
|
|
continue;
|
2008-07-15 13:44:58 +08:00
|
|
|
tipc_node_lock(n_ptr);
|
2007-02-09 22:25:21 +08:00
|
|
|
for (i = 0; i < MAX_BEARERS; i++) {
|
|
|
|
if (!n_ptr->links[i])
|
|
|
|
continue;
|
|
|
|
link_info.dest = htonl(n_ptr->addr);
|
|
|
|
link_info.up = htonl(tipc_link_is_up(n_ptr->links[i]));
|
|
|
|
strcpy(link_info.str, n_ptr->links[i]->name);
|
|
|
|
tipc_cfg_append_tlv(buf, TIPC_TLV_LINK_INFO,
|
2006-01-18 07:38:21 +08:00
|
|
|
&link_info, sizeof(link_info));
|
2007-02-09 22:25:21 +08:00
|
|
|
}
|
2008-07-15 13:44:58 +08:00
|
|
|
tipc_node_unlock(n_ptr);
|
2006-01-03 02:04:38 +08:00
|
|
|
}
|
|
|
|
|
2008-07-15 13:44:58 +08:00
|
|
|
read_unlock_bh(&tipc_net_lock);
|
2006-01-03 02:04:38 +08:00
|
|
|
return buf;
|
|
|
|
}
|