2
0
mirror of https://github.com/edk2-porting/linux-next.git synced 2024-12-27 22:53:55 +08:00
linux-next/net/tipc/link.c

2792 lines
74 KiB
C
Raw Normal View History

/*
* net/tipc/link.c: TIPC link code
*
* Copyright (c) 1996-2007, 2012-2014, Ericsson AB
* Copyright (c) 2004-2007, 2010-2013, Wind River Systems
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions are met:
*
* 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.
*
* 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
* POSSIBILITY OF SUCH DAMAGE.
*/
#include "core.h"
#include "link.h"
#include "bcast.h"
#include "socket.h"
#include "name_distr.h"
#include "discover.h"
#include "config.h"
#include "netlink.h"
#include <linux/pkt_sched.h>
/*
* Error message prefixes
*/
static const char *link_co_err = "Link changeover error, ";
static const char *link_rst_msg = "Resetting link ";
static const char *link_unk_evt = "Unknown link event ";
static const struct nla_policy tipc_nl_link_policy[TIPC_NLA_LINK_MAX + 1] = {
[TIPC_NLA_LINK_UNSPEC] = { .type = NLA_UNSPEC },
[TIPC_NLA_LINK_NAME] = {
.type = NLA_STRING,
.len = TIPC_MAX_LINK_NAME
},
[TIPC_NLA_LINK_MTU] = { .type = NLA_U32 },
[TIPC_NLA_LINK_BROADCAST] = { .type = NLA_FLAG },
[TIPC_NLA_LINK_UP] = { .type = NLA_FLAG },
[TIPC_NLA_LINK_ACTIVE] = { .type = NLA_FLAG },
[TIPC_NLA_LINK_PROP] = { .type = NLA_NESTED },
[TIPC_NLA_LINK_STATS] = { .type = NLA_NESTED },
[TIPC_NLA_LINK_RX] = { .type = NLA_U32 },
[TIPC_NLA_LINK_TX] = { .type = NLA_U32 }
};
/* Properties valid for media, bearar and link */
static const struct nla_policy tipc_nl_prop_policy[TIPC_NLA_PROP_MAX + 1] = {
[TIPC_NLA_PROP_UNSPEC] = { .type = NLA_UNSPEC },
[TIPC_NLA_PROP_PRIO] = { .type = NLA_U32 },
[TIPC_NLA_PROP_TOL] = { .type = NLA_U32 },
[TIPC_NLA_PROP_WIN] = { .type = NLA_U32 }
};
/*
* Out-of-range value for link session numbers
*/
#define INVALID_SESSION 0x10000
/*
* Link state events:
*/
#define STARTING_EVT 856384768 /* link processing trigger */
#define TRAFFIC_MSG_EVT 560815u /* rx'd ??? */
#define TIMEOUT_EVT 560817u /* link timer expired */
/*
* The following two 'message types' is really just implementation
* data conveniently stored in the message header.
* They must not be considered part of the protocol
*/
#define OPEN_MSG 0
#define CLOSED_MSG 1
/*
* State value stored in 'exp_msg_count'
*/
#define START_CHANGEOVER 100000u
2015-02-05 21:36:36 +08:00
static void link_handle_out_of_seq_msg(struct tipc_link *link,
struct sk_buff *skb);
static void tipc_link_proto_rcv(struct tipc_link *link,
struct sk_buff *skb);
static int tipc_link_tunnel_rcv(struct tipc_node *node,
struct sk_buff **skb);
static void link_set_supervision_props(struct tipc_link *l_ptr, u32 tol);
static void link_state_event(struct tipc_link *l_ptr, u32 event);
static void link_reset_statistics(struct tipc_link *l_ptr);
static void link_print(struct tipc_link *l_ptr, const char *str);
tipc: align tipc function names with common naming practice in the network Rename the following functions, which are shorter and more in line with common naming practice in the network subsystem. tipc_bclink_send_msg->tipc_bclink_xmit tipc_bclink_recv_pkt->tipc_bclink_rcv tipc_disc_recv_msg->tipc_disc_rcv tipc_link_send_proto_msg->tipc_link_proto_xmit link_recv_proto_msg->tipc_link_proto_rcv link_send_sections_long->tipc_link_iovec_long_xmit tipc_link_send_sections_fast->tipc_link_iovec_xmit_fast tipc_link_send_sync->tipc_link_sync_xmit tipc_link_recv_sync->tipc_link_sync_rcv tipc_link_send_buf->__tipc_link_xmit tipc_link_send->tipc_link_xmit tipc_link_send_names->tipc_link_names_xmit tipc_named_recv->tipc_named_rcv tipc_link_recv_bundle->tipc_link_bundle_rcv tipc_link_dup_send_queue->tipc_link_dup_queue_xmit link_send_long_buf->tipc_link_frag_xmit tipc_multicast->tipc_port_mcast_xmit tipc_port_recv_mcast->tipc_port_mcast_rcv tipc_port_reject_sections->tipc_port_iovec_reject tipc_port_recv_proto_msg->tipc_port_proto_rcv tipc_connect->tipc_port_connect __tipc_connect->__tipc_port_connect __tipc_disconnect->__tipc_port_disconnect tipc_disconnect->tipc_port_disconnect tipc_shutdown->tipc_port_shutdown tipc_port_recv_msg->tipc_port_rcv tipc_port_recv_sections->tipc_port_iovec_rcv release->tipc_release accept->tipc_accept bind->tipc_bind get_name->tipc_getname poll->tipc_poll send_msg->tipc_sendmsg send_packet->tipc_send_packet send_stream->tipc_send_stream recv_msg->tipc_recvmsg recv_stream->tipc_recv_stream connect->tipc_connect listen->tipc_listen shutdown->tipc_shutdown setsockopt->tipc_setsockopt getsockopt->tipc_getsockopt Above changes have no impact on current users of the functions. Signed-off-by: Ying Xue <ying.xue@windriver.com> Reviewed-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-18 16:06:46 +08:00
static void tipc_link_sync_xmit(struct tipc_link *l);
static void tipc_link_sync_rcv(struct tipc_node *n, struct sk_buff *buf);
static int tipc_link_input(struct net *net, struct tipc_link *l,
struct sk_buff *buf);
static int tipc_link_prepare_input(struct net *net, struct tipc_link *l,
struct sk_buff **buf);
/*
* Simple link routines
*/
static unsigned int align(unsigned int i)
{
return (i + 3) & ~3u;
}
static void tipc_link_release(struct kref *kref)
{
kfree(container_of(kref, struct tipc_link, ref));
}
static void tipc_link_get(struct tipc_link *l_ptr)
{
kref_get(&l_ptr->ref);
}
static void tipc_link_put(struct tipc_link *l_ptr)
{
kref_put(&l_ptr->ref, tipc_link_release);
}
static void link_init_max_pkt(struct tipc_link *l_ptr)
{
struct tipc_node *node = l_ptr->owner;
struct tipc_net *tn = net_generic(node->net, tipc_net_id);
struct tipc_bearer *b_ptr;
u32 max_pkt;
rcu_read_lock();
b_ptr = rcu_dereference_rtnl(tn->bearer_list[l_ptr->bearer_id]);
if (!b_ptr) {
rcu_read_unlock();
return;
}
max_pkt = (b_ptr->mtu & ~3);
rcu_read_unlock();
if (max_pkt > MAX_MSG_SIZE)
max_pkt = MAX_MSG_SIZE;
l_ptr->max_pkt_target = max_pkt;
if (l_ptr->max_pkt_target < MAX_PKT_DEFAULT)
l_ptr->max_pkt = l_ptr->max_pkt_target;
else
l_ptr->max_pkt = MAX_PKT_DEFAULT;
l_ptr->max_pkt_probes = 0;
}
/*
* Simple non-static link routines (i.e. referenced outside this file)
*/
int tipc_link_is_up(struct tipc_link *l_ptr)
{
if (!l_ptr)
return 0;
return link_working_working(l_ptr) || link_working_unknown(l_ptr);
}
int tipc_link_is_active(struct tipc_link *l_ptr)
{
return (l_ptr->owner->active_links[0] == l_ptr) ||
(l_ptr->owner->active_links[1] == l_ptr);
}
/**
* link_timeout - handle expiration of link timer
* @l_ptr: pointer to link
*/
static void link_timeout(unsigned long data)
{
struct tipc_link *l_ptr = (struct tipc_link *)data;
struct sk_buff *skb;
tipc_node_lock(l_ptr->owner);
/* update counters used in statistical profiling of send traffic */
l_ptr->stats.accu_queue_sz += skb_queue_len(&l_ptr->outqueue);
l_ptr->stats.queue_sz_counts++;
skb = skb_peek(&l_ptr->outqueue);
if (skb) {
struct tipc_msg *msg = buf_msg(skb);
u32 length = msg_size(msg);
if ((msg_user(msg) == MSG_FRAGMENTER) &&
(msg_type(msg) == FIRST_FRAGMENT)) {
length = msg_size(msg_get_wrapped(msg));
}
if (length) {
l_ptr->stats.msg_lengths_total += length;
l_ptr->stats.msg_length_counts++;
if (length <= 64)
l_ptr->stats.msg_length_profile[0]++;
else if (length <= 256)
l_ptr->stats.msg_length_profile[1]++;
else if (length <= 1024)
l_ptr->stats.msg_length_profile[2]++;
else if (length <= 4096)
l_ptr->stats.msg_length_profile[3]++;
else if (length <= 16384)
l_ptr->stats.msg_length_profile[4]++;
else if (length <= 32768)
l_ptr->stats.msg_length_profile[5]++;
else
l_ptr->stats.msg_length_profile[6]++;
}
}
/* do all other link processing performed on a periodic basis */
link_state_event(l_ptr, TIMEOUT_EVT);
if (l_ptr->next_out)
tipc_link_push_packets(l_ptr);
tipc_node_unlock(l_ptr->owner);
tipc_link_put(l_ptr);
}
static void link_set_timer(struct tipc_link *link, unsigned long time)
{
if (!mod_timer(&link->timer, jiffies + time))
tipc_link_get(link);
}
/**
* tipc_link_create - create a new link
* @n_ptr: pointer to associated node
* @b_ptr: pointer to associated bearer
* @media_addr: media address to use when sending messages over link
*
* Returns pointer to link.
*/
struct tipc_link *tipc_link_create(struct tipc_node *n_ptr,
tipc: remove 'links' list from tipc_bearer struct In our ongoing effort to simplify the TIPC locking structure, we see a need to remove the linked list for tipc_links in the bearer. This can be explained as follows. Currently, we have three different ways to access a link, via three different lists/tables: 1: Via a node hash table: Used by the time-critical outgoing/incoming data paths. (e.g. link_send_sections_fast() and tipc_recv_msg() ): grab net_lock(read) find node from node hash table grab node_lock select link grab bearer_lock send_msg() release bearer_lock release node lock release net_lock 2: Via a global linked list for nodes: Used by configuration commands (link_cmd_set_value()) grab net_lock(read) find node and link from global node list (using link name) grab node_lock update link release node lock release net_lock (Same locking order as above. No problem.) 3: Via the bearer's linked link list: Used by notifications from interface (e.g. tipc_disable_bearer() ) grab net_lock(write) grab bearer_lock get link ptr from bearer's link list get node from link grab node_lock delete link release node lock release bearer_lock release net_lock (Different order from above, but works because we grab the outer net_lock in write mode first, excluding all other access.) The first major goal in our simplification effort is to get rid of the "big" net_lock, replacing it with rcu-locks when accessing the node list and node hash array. This will come in a later patch series. But to get there we first need to rewrite access methods ##2 and 3, since removal of net_lock would introduce three major problems: a) In access method #2, we access the link before taking the protecting node_lock. This will not work once net_lock is gone, so we will have to change the access order. We will deal with this in a later commit in this series, "tipc: add node lock protection to link found by link_find_link()". b) When the outer protection from net_lock is gone, taking bearer_lock and node_lock in opposite order of method 1) and 2) will become an obvious deadlock hazard. This is fixed in the commit ("tipc: remove bearer_lock from tipc_bearer struct") later in this series. c) Similar to what is described in problem a), access method #3 starts with using a link pointer that is unprotected by node_lock, in order to via that pointer find the correct node struct and lock it. Before we remove net_lock, this access order must be altered. This is what we do with this commit. We can avoid introducing problem problem c) by even here using the global node list to find the node, before accessing its links. When we loop though the node list we use the own bearer identity as search criteria, thus easily finding the links that are associated to the resetting/disabling bearer. It should be noted that although this method is somewhat slower than the current list traversal, it is in no way time critical. This is only about resetting or deleting links, something that must be considered relatively infrequent events. As a bonus, we can get rid of the mutual pointers between links and bearers. After this commit, pointer dependency go in one direction only: from the link to the bearer. This commit pre-empts introduction of problem c) as described above. Signed-off-by: Ying Xue <ying.xue@windriver.com> Reviewed-by: Paul Gortmaker <paul.gortmaker@windriver.com> Signed-off-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-14 06:29:09 +08:00
struct tipc_bearer *b_ptr,
const struct tipc_media_addr *media_addr)
{
struct tipc_net *tn = net_generic(n_ptr->net, tipc_net_id);
struct tipc_link *l_ptr;
struct tipc_msg *msg;
char *if_name;
char addr_string[16];
u32 peer = n_ptr->addr;
if (n_ptr->link_cnt >= MAX_BEARERS) {
tipc_addr_string_fill(addr_string, n_ptr->addr);
pr_err("Attempt to establish %uth link to %s. Max %u allowed.\n",
n_ptr->link_cnt, addr_string, MAX_BEARERS);
return NULL;
}
if (n_ptr->links[b_ptr->identity]) {
tipc_addr_string_fill(addr_string, n_ptr->addr);
pr_err("Attempt to establish second link on <%s> to %s\n",
b_ptr->name, addr_string);
return NULL;
}
l_ptr = kzalloc(sizeof(*l_ptr), GFP_ATOMIC);
if (!l_ptr) {
pr_warn("Link creation failed, no memory\n");
return NULL;
}
kref_init(&l_ptr->ref);
l_ptr->addr = peer;
if_name = strchr(b_ptr->name, ':') + 1;
sprintf(l_ptr->name, "%u.%u.%u:%s-%u.%u.%u:unknown",
tipc_zone(tn->own_addr), tipc_cluster(tn->own_addr),
tipc_node(tn->own_addr),
if_name,
tipc_zone(peer), tipc_cluster(peer), tipc_node(peer));
/* note: peer i/f name is updated by reset/activate message */
memcpy(&l_ptr->media_addr, media_addr, sizeof(*media_addr));
l_ptr->owner = n_ptr;
l_ptr->checkpoint = 1;
l_ptr->peer_session = INVALID_SESSION;
l_ptr->bearer_id = b_ptr->identity;
link_set_supervision_props(l_ptr, b_ptr->tolerance);
l_ptr->state = RESET_UNKNOWN;
l_ptr->pmsg = (struct tipc_msg *)&l_ptr->proto_msg;
msg = l_ptr->pmsg;
2015-02-05 21:36:36 +08:00
tipc_msg_init(tn->own_addr, msg, LINK_PROTOCOL, RESET_MSG, INT_H_SIZE,
l_ptr->addr);
msg_set_size(msg, sizeof(l_ptr->proto_msg));
msg_set_session(msg, (tn->random & 0xffff));
msg_set_bearer_id(msg, b_ptr->identity);
strcpy((char *)msg_data(msg), if_name);
l_ptr->priority = b_ptr->priority;
tipc_link_set_queue_limits(l_ptr, b_ptr->window);
l_ptr->net_plane = b_ptr->net_plane;
link_init_max_pkt(l_ptr);
l_ptr->next_out_no = 1;
__skb_queue_head_init(&l_ptr->outqueue);
__skb_queue_head_init(&l_ptr->deferred_queue);
skb_queue_head_init(&l_ptr->waiting_sks);
link_reset_statistics(l_ptr);
tipc_node_attach_link(n_ptr, l_ptr);
setup_timer(&l_ptr->timer, link_timeout, (unsigned long)l_ptr);
link_state_event(l_ptr, STARTING_EVT);
return l_ptr;
}
/**
* link_delete - Conditional deletion of link.
* If timer still running, real delete is done when it expires
* @link: link to be deleted
*/
void tipc_link_delete(struct tipc_link *link)
{
tipc_link_reset_fragments(link);
tipc_node_detach_link(link->owner, link);
tipc_link_put(link);
}
void tipc_link_delete_list(struct net *net, unsigned int bearer_id,
bool shutting_down)
{
struct tipc_net *tn = net_generic(net, tipc_net_id);
struct tipc_link *link;
struct tipc_node *node;
rcu_read_lock();
list_for_each_entry_rcu(node, &tn->node_list, list) {
tipc_node_lock(node);
link = node->links[bearer_id];
if (!link) {
tipc_node_unlock(node);
tipc: remove 'links' list from tipc_bearer struct In our ongoing effort to simplify the TIPC locking structure, we see a need to remove the linked list for tipc_links in the bearer. This can be explained as follows. Currently, we have three different ways to access a link, via three different lists/tables: 1: Via a node hash table: Used by the time-critical outgoing/incoming data paths. (e.g. link_send_sections_fast() and tipc_recv_msg() ): grab net_lock(read) find node from node hash table grab node_lock select link grab bearer_lock send_msg() release bearer_lock release node lock release net_lock 2: Via a global linked list for nodes: Used by configuration commands (link_cmd_set_value()) grab net_lock(read) find node and link from global node list (using link name) grab node_lock update link release node lock release net_lock (Same locking order as above. No problem.) 3: Via the bearer's linked link list: Used by notifications from interface (e.g. tipc_disable_bearer() ) grab net_lock(write) grab bearer_lock get link ptr from bearer's link list get node from link grab node_lock delete link release node lock release bearer_lock release net_lock (Different order from above, but works because we grab the outer net_lock in write mode first, excluding all other access.) The first major goal in our simplification effort is to get rid of the "big" net_lock, replacing it with rcu-locks when accessing the node list and node hash array. This will come in a later patch series. But to get there we first need to rewrite access methods ##2 and 3, since removal of net_lock would introduce three major problems: a) In access method #2, we access the link before taking the protecting node_lock. This will not work once net_lock is gone, so we will have to change the access order. We will deal with this in a later commit in this series, "tipc: add node lock protection to link found by link_find_link()". b) When the outer protection from net_lock is gone, taking bearer_lock and node_lock in opposite order of method 1) and 2) will become an obvious deadlock hazard. This is fixed in the commit ("tipc: remove bearer_lock from tipc_bearer struct") later in this series. c) Similar to what is described in problem a), access method #3 starts with using a link pointer that is unprotected by node_lock, in order to via that pointer find the correct node struct and lock it. Before we remove net_lock, this access order must be altered. This is what we do with this commit. We can avoid introducing problem problem c) by even here using the global node list to find the node, before accessing its links. When we loop though the node list we use the own bearer identity as search criteria, thus easily finding the links that are associated to the resetting/disabling bearer. It should be noted that although this method is somewhat slower than the current list traversal, it is in no way time critical. This is only about resetting or deleting links, something that must be considered relatively infrequent events. As a bonus, we can get rid of the mutual pointers between links and bearers. After this commit, pointer dependency go in one direction only: from the link to the bearer. This commit pre-empts introduction of problem c) as described above. Signed-off-by: Ying Xue <ying.xue@windriver.com> Reviewed-by: Paul Gortmaker <paul.gortmaker@windriver.com> Signed-off-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-14 06:29:09 +08:00
continue;
}
tipc_link_reset(link);
if (del_timer(&link->timer))
tipc_link_put(link);
link->flags |= LINK_STOPPED;
/* Delete link now, or when failover is finished: */
if (shutting_down || !tipc_node_is_up(node))
tipc_link_delete(link);
tipc_node_unlock(node);
}
rcu_read_unlock();
}
/**
* link_schedule_user - schedule user for wakeup after congestion
* @link: congested link
* @oport: sending port
* @chain_sz: size of buffer chain that was attempted sent
* @imp: importance of message attempted sent
* Create pseudo msg to send back to user when congestion abates
*/
static bool link_schedule_user(struct tipc_link *link, u32 oport,
uint chain_sz, uint imp)
{
struct sk_buff *buf;
2015-02-05 21:36:36 +08:00
buf = tipc_msg_create(SOCK_WAKEUP, 0, INT_H_SIZE, 0,
link_own_addr(link), link_own_addr(link),
oport, 0, 0);
if (!buf)
return false;
TIPC_SKB_CB(buf)->chain_sz = chain_sz;
TIPC_SKB_CB(buf)->chain_imp = imp;
skb_queue_tail(&link->waiting_sks, buf);
link->stats.link_congs++;
return true;
}
/**
* link_prepare_wakeup - prepare users for wakeup after congestion
* @link: congested link
* Move a number of waiting users, as permitted by available space in
* the send queue, from link wait queue to node wait queue for wakeup
*/
static void link_prepare_wakeup(struct tipc_link *link)
{
uint pend_qsz = skb_queue_len(&link->outqueue);
struct sk_buff *skb, *tmp;
skb_queue_walk_safe(&link->waiting_sks, skb, tmp) {
if (pend_qsz >= link->queue_limit[TIPC_SKB_CB(skb)->chain_imp])
break;
pend_qsz += TIPC_SKB_CB(skb)->chain_sz;
skb_unlink(skb, &link->waiting_sks);
skb_queue_tail(&link->owner->waiting_sks, skb);
}
}
/**
* tipc_link_reset_fragments - purge link's inbound message fragments queue
* @l_ptr: pointer to link
*/
void tipc_link_reset_fragments(struct tipc_link *l_ptr)
{
kfree_skb(l_ptr->reasm_buf);
l_ptr->reasm_buf = NULL;
}
/**
* tipc_link_purge_queues - purge all pkt queues associated with link
* @l_ptr: pointer to link
*/
void tipc_link_purge_queues(struct tipc_link *l_ptr)
{
__skb_queue_purge(&l_ptr->deferred_queue);
__skb_queue_purge(&l_ptr->outqueue);
tipc_link_reset_fragments(l_ptr);
}
void tipc_link_reset(struct tipc_link *l_ptr)
{
u32 prev_state = l_ptr->state;
u32 checkpoint = l_ptr->next_in_no;
int was_active_link = tipc_link_is_active(l_ptr);
struct tipc_node *owner = l_ptr->owner;
msg_set_session(l_ptr->pmsg, ((msg_session(l_ptr->pmsg) + 1) & 0xffff));
/* Link is down, accept any session */
l_ptr->peer_session = INVALID_SESSION;
/* Prepare for max packet size negotiation */
link_init_max_pkt(l_ptr);
l_ptr->state = RESET_UNKNOWN;
if ((prev_state == RESET_UNKNOWN) || (prev_state == RESET_RESET))
return;
tipc_node_link_down(l_ptr->owner, l_ptr);
tipc_bearer_remove_dest(owner->net, l_ptr->bearer_id, l_ptr->addr);
tipc: remove 'has_redundant_link' flag from STATE link protocol messages The flag 'has_redundant_link' is defined only in RESET and ACTIVATE protocol messages. Due to an ambiguity in the protocol specification it is currently also transferred in STATE messages. Its value is used to initialize a link state variable, 'permit_changeover', which is used to inhibit futile link failover attempts when it is known that the peer node has no working links at the moment, although the local node may still think it has one. The fact that 'has_redundant_link' incorrectly is read from STATE messages has the effect that 'permit_changeover' sometimes gets a wrong value, and permanently blocks any links from being re-established. Such failures can only occur in in dual-link systems, and are extremely rare. This bug seems to have always been present in the code. Furthermore, since commit b4b5610223f17790419b03eaa962b0e3ecf930d7 ("tipc: Ensure both nodes recognize loss of contact between them"), the 'permit_changeover' field serves no purpose any more. The task of enforcing 'lost contact' cycles at both peer endpoints is now taken by a new mechanism, using the flags WAIT_NODE_DOWN and WAIT_PEER_DOWN in struct tipc_node to abort unnecessary failover attempts. We therefore remove the 'has_redundant_link' flag from STATE messages, as well as the now redundant 'permit_changeover' variable. Signed-off-by: Jon Maloy <jon.maloy@ericsson.com> Reviewed-by: Ying Xue <ying.xue@windriver.com> Reviewed-by: Paul Gortmaker <paul.gortmaker@windriver.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-01-08 06:02:42 +08:00
if (was_active_link && tipc_node_active_links(l_ptr->owner)) {
l_ptr->reset_checkpoint = checkpoint;
l_ptr->exp_msg_count = START_CHANGEOVER;
}
/* Clean up all queues: */
__skb_queue_purge(&l_ptr->outqueue);
__skb_queue_purge(&l_ptr->deferred_queue);
if (!skb_queue_empty(&l_ptr->waiting_sks)) {
skb_queue_splice_init(&l_ptr->waiting_sks, &owner->waiting_sks);
owner->action_flags |= TIPC_WAKEUP_USERS;
}
l_ptr->next_out = NULL;
l_ptr->unacked_window = 0;
l_ptr->checkpoint = 1;
l_ptr->next_out_no = 1;
l_ptr->fsm_msg_cnt = 0;
l_ptr->stale_count = 0;
link_reset_statistics(l_ptr);
}
void tipc_link_reset_list(struct net *net, unsigned int bearer_id)
{
struct tipc_net *tn = net_generic(net, tipc_net_id);
struct tipc_link *l_ptr;
tipc: remove 'links' list from tipc_bearer struct In our ongoing effort to simplify the TIPC locking structure, we see a need to remove the linked list for tipc_links in the bearer. This can be explained as follows. Currently, we have three different ways to access a link, via three different lists/tables: 1: Via a node hash table: Used by the time-critical outgoing/incoming data paths. (e.g. link_send_sections_fast() and tipc_recv_msg() ): grab net_lock(read) find node from node hash table grab node_lock select link grab bearer_lock send_msg() release bearer_lock release node lock release net_lock 2: Via a global linked list for nodes: Used by configuration commands (link_cmd_set_value()) grab net_lock(read) find node and link from global node list (using link name) grab node_lock update link release node lock release net_lock (Same locking order as above. No problem.) 3: Via the bearer's linked link list: Used by notifications from interface (e.g. tipc_disable_bearer() ) grab net_lock(write) grab bearer_lock get link ptr from bearer's link list get node from link grab node_lock delete link release node lock release bearer_lock release net_lock (Different order from above, but works because we grab the outer net_lock in write mode first, excluding all other access.) The first major goal in our simplification effort is to get rid of the "big" net_lock, replacing it with rcu-locks when accessing the node list and node hash array. This will come in a later patch series. But to get there we first need to rewrite access methods ##2 and 3, since removal of net_lock would introduce three major problems: a) In access method #2, we access the link before taking the protecting node_lock. This will not work once net_lock is gone, so we will have to change the access order. We will deal with this in a later commit in this series, "tipc: add node lock protection to link found by link_find_link()". b) When the outer protection from net_lock is gone, taking bearer_lock and node_lock in opposite order of method 1) and 2) will become an obvious deadlock hazard. This is fixed in the commit ("tipc: remove bearer_lock from tipc_bearer struct") later in this series. c) Similar to what is described in problem a), access method #3 starts with using a link pointer that is unprotected by node_lock, in order to via that pointer find the correct node struct and lock it. Before we remove net_lock, this access order must be altered. This is what we do with this commit. We can avoid introducing problem problem c) by even here using the global node list to find the node, before accessing its links. When we loop though the node list we use the own bearer identity as search criteria, thus easily finding the links that are associated to the resetting/disabling bearer. It should be noted that although this method is somewhat slower than the current list traversal, it is in no way time critical. This is only about resetting or deleting links, something that must be considered relatively infrequent events. As a bonus, we can get rid of the mutual pointers between links and bearers. After this commit, pointer dependency go in one direction only: from the link to the bearer. This commit pre-empts introduction of problem c) as described above. Signed-off-by: Ying Xue <ying.xue@windriver.com> Reviewed-by: Paul Gortmaker <paul.gortmaker@windriver.com> Signed-off-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-14 06:29:09 +08:00
struct tipc_node *n_ptr;
rcu_read_lock();
list_for_each_entry_rcu(n_ptr, &tn->node_list, list) {
tipc_node_lock(n_ptr);
tipc: remove 'links' list from tipc_bearer struct In our ongoing effort to simplify the TIPC locking structure, we see a need to remove the linked list for tipc_links in the bearer. This can be explained as follows. Currently, we have three different ways to access a link, via three different lists/tables: 1: Via a node hash table: Used by the time-critical outgoing/incoming data paths. (e.g. link_send_sections_fast() and tipc_recv_msg() ): grab net_lock(read) find node from node hash table grab node_lock select link grab bearer_lock send_msg() release bearer_lock release node lock release net_lock 2: Via a global linked list for nodes: Used by configuration commands (link_cmd_set_value()) grab net_lock(read) find node and link from global node list (using link name) grab node_lock update link release node lock release net_lock (Same locking order as above. No problem.) 3: Via the bearer's linked link list: Used by notifications from interface (e.g. tipc_disable_bearer() ) grab net_lock(write) grab bearer_lock get link ptr from bearer's link list get node from link grab node_lock delete link release node lock release bearer_lock release net_lock (Different order from above, but works because we grab the outer net_lock in write mode first, excluding all other access.) The first major goal in our simplification effort is to get rid of the "big" net_lock, replacing it with rcu-locks when accessing the node list and node hash array. This will come in a later patch series. But to get there we first need to rewrite access methods ##2 and 3, since removal of net_lock would introduce three major problems: a) In access method #2, we access the link before taking the protecting node_lock. This will not work once net_lock is gone, so we will have to change the access order. We will deal with this in a later commit in this series, "tipc: add node lock protection to link found by link_find_link()". b) When the outer protection from net_lock is gone, taking bearer_lock and node_lock in opposite order of method 1) and 2) will become an obvious deadlock hazard. This is fixed in the commit ("tipc: remove bearer_lock from tipc_bearer struct") later in this series. c) Similar to what is described in problem a), access method #3 starts with using a link pointer that is unprotected by node_lock, in order to via that pointer find the correct node struct and lock it. Before we remove net_lock, this access order must be altered. This is what we do with this commit. We can avoid introducing problem problem c) by even here using the global node list to find the node, before accessing its links. When we loop though the node list we use the own bearer identity as search criteria, thus easily finding the links that are associated to the resetting/disabling bearer. It should be noted that although this method is somewhat slower than the current list traversal, it is in no way time critical. This is only about resetting or deleting links, something that must be considered relatively infrequent events. As a bonus, we can get rid of the mutual pointers between links and bearers. After this commit, pointer dependency go in one direction only: from the link to the bearer. This commit pre-empts introduction of problem c) as described above. Signed-off-by: Ying Xue <ying.xue@windriver.com> Reviewed-by: Paul Gortmaker <paul.gortmaker@windriver.com> Signed-off-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-14 06:29:09 +08:00
l_ptr = n_ptr->links[bearer_id];
if (l_ptr)
tipc_link_reset(l_ptr);
tipc_node_unlock(n_ptr);
}
rcu_read_unlock();
}
static void link_activate(struct tipc_link *link)
{
struct tipc_node *node = link->owner;
link->next_in_no = 1;
link->stats.recv_info = 1;
tipc_node_link_up(node, link);
tipc_bearer_add_dest(node->net, link->bearer_id, link->addr);
}
/**
* link_state_event - link finite state machine
* @l_ptr: pointer to link
* @event: state machine event to process
*/
static void link_state_event(struct tipc_link *l_ptr, unsigned int event)
{
struct tipc_link *other;
unsigned long cont_intv = l_ptr->cont_intv;
if (l_ptr->flags & LINK_STOPPED)
return;
if (!(l_ptr->flags & LINK_STARTED) && (event != STARTING_EVT))
return; /* Not yet. */
/* Check whether changeover is going on */
if (l_ptr->exp_msg_count) {
if (event == TIMEOUT_EVT)
link_set_timer(l_ptr, cont_intv);
return;
}
switch (l_ptr->state) {
case WORKING_WORKING:
switch (event) {
case TRAFFIC_MSG_EVT:
case ACTIVATE_MSG:
break;
case TIMEOUT_EVT:
if (l_ptr->next_in_no != l_ptr->checkpoint) {
l_ptr->checkpoint = l_ptr->next_in_no;
if (tipc_bclink_acks_missing(l_ptr->owner)) {
tipc: align tipc function names with common naming practice in the network Rename the following functions, which are shorter and more in line with common naming practice in the network subsystem. tipc_bclink_send_msg->tipc_bclink_xmit tipc_bclink_recv_pkt->tipc_bclink_rcv tipc_disc_recv_msg->tipc_disc_rcv tipc_link_send_proto_msg->tipc_link_proto_xmit link_recv_proto_msg->tipc_link_proto_rcv link_send_sections_long->tipc_link_iovec_long_xmit tipc_link_send_sections_fast->tipc_link_iovec_xmit_fast tipc_link_send_sync->tipc_link_sync_xmit tipc_link_recv_sync->tipc_link_sync_rcv tipc_link_send_buf->__tipc_link_xmit tipc_link_send->tipc_link_xmit tipc_link_send_names->tipc_link_names_xmit tipc_named_recv->tipc_named_rcv tipc_link_recv_bundle->tipc_link_bundle_rcv tipc_link_dup_send_queue->tipc_link_dup_queue_xmit link_send_long_buf->tipc_link_frag_xmit tipc_multicast->tipc_port_mcast_xmit tipc_port_recv_mcast->tipc_port_mcast_rcv tipc_port_reject_sections->tipc_port_iovec_reject tipc_port_recv_proto_msg->tipc_port_proto_rcv tipc_connect->tipc_port_connect __tipc_connect->__tipc_port_connect __tipc_disconnect->__tipc_port_disconnect tipc_disconnect->tipc_port_disconnect tipc_shutdown->tipc_port_shutdown tipc_port_recv_msg->tipc_port_rcv tipc_port_recv_sections->tipc_port_iovec_rcv release->tipc_release accept->tipc_accept bind->tipc_bind get_name->tipc_getname poll->tipc_poll send_msg->tipc_sendmsg send_packet->tipc_send_packet send_stream->tipc_send_stream recv_msg->tipc_recvmsg recv_stream->tipc_recv_stream connect->tipc_connect listen->tipc_listen shutdown->tipc_shutdown setsockopt->tipc_setsockopt getsockopt->tipc_getsockopt Above changes have no impact on current users of the functions. Signed-off-by: Ying Xue <ying.xue@windriver.com> Reviewed-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-18 16:06:46 +08:00
tipc_link_proto_xmit(l_ptr, STATE_MSG,
0, 0, 0, 0, 0);
l_ptr->fsm_msg_cnt++;
} else if (l_ptr->max_pkt < l_ptr->max_pkt_target) {
tipc: align tipc function names with common naming practice in the network Rename the following functions, which are shorter and more in line with common naming practice in the network subsystem. tipc_bclink_send_msg->tipc_bclink_xmit tipc_bclink_recv_pkt->tipc_bclink_rcv tipc_disc_recv_msg->tipc_disc_rcv tipc_link_send_proto_msg->tipc_link_proto_xmit link_recv_proto_msg->tipc_link_proto_rcv link_send_sections_long->tipc_link_iovec_long_xmit tipc_link_send_sections_fast->tipc_link_iovec_xmit_fast tipc_link_send_sync->tipc_link_sync_xmit tipc_link_recv_sync->tipc_link_sync_rcv tipc_link_send_buf->__tipc_link_xmit tipc_link_send->tipc_link_xmit tipc_link_send_names->tipc_link_names_xmit tipc_named_recv->tipc_named_rcv tipc_link_recv_bundle->tipc_link_bundle_rcv tipc_link_dup_send_queue->tipc_link_dup_queue_xmit link_send_long_buf->tipc_link_frag_xmit tipc_multicast->tipc_port_mcast_xmit tipc_port_recv_mcast->tipc_port_mcast_rcv tipc_port_reject_sections->tipc_port_iovec_reject tipc_port_recv_proto_msg->tipc_port_proto_rcv tipc_connect->tipc_port_connect __tipc_connect->__tipc_port_connect __tipc_disconnect->__tipc_port_disconnect tipc_disconnect->tipc_port_disconnect tipc_shutdown->tipc_port_shutdown tipc_port_recv_msg->tipc_port_rcv tipc_port_recv_sections->tipc_port_iovec_rcv release->tipc_release accept->tipc_accept bind->tipc_bind get_name->tipc_getname poll->tipc_poll send_msg->tipc_sendmsg send_packet->tipc_send_packet send_stream->tipc_send_stream recv_msg->tipc_recvmsg recv_stream->tipc_recv_stream connect->tipc_connect listen->tipc_listen shutdown->tipc_shutdown setsockopt->tipc_setsockopt getsockopt->tipc_getsockopt Above changes have no impact on current users of the functions. Signed-off-by: Ying Xue <ying.xue@windriver.com> Reviewed-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-18 16:06:46 +08:00
tipc_link_proto_xmit(l_ptr, STATE_MSG,
1, 0, 0, 0, 0);
l_ptr->fsm_msg_cnt++;
}
link_set_timer(l_ptr, cont_intv);
break;
}
l_ptr->state = WORKING_UNKNOWN;
l_ptr->fsm_msg_cnt = 0;
tipc: align tipc function names with common naming practice in the network Rename the following functions, which are shorter and more in line with common naming practice in the network subsystem. tipc_bclink_send_msg->tipc_bclink_xmit tipc_bclink_recv_pkt->tipc_bclink_rcv tipc_disc_recv_msg->tipc_disc_rcv tipc_link_send_proto_msg->tipc_link_proto_xmit link_recv_proto_msg->tipc_link_proto_rcv link_send_sections_long->tipc_link_iovec_long_xmit tipc_link_send_sections_fast->tipc_link_iovec_xmit_fast tipc_link_send_sync->tipc_link_sync_xmit tipc_link_recv_sync->tipc_link_sync_rcv tipc_link_send_buf->__tipc_link_xmit tipc_link_send->tipc_link_xmit tipc_link_send_names->tipc_link_names_xmit tipc_named_recv->tipc_named_rcv tipc_link_recv_bundle->tipc_link_bundle_rcv tipc_link_dup_send_queue->tipc_link_dup_queue_xmit link_send_long_buf->tipc_link_frag_xmit tipc_multicast->tipc_port_mcast_xmit tipc_port_recv_mcast->tipc_port_mcast_rcv tipc_port_reject_sections->tipc_port_iovec_reject tipc_port_recv_proto_msg->tipc_port_proto_rcv tipc_connect->tipc_port_connect __tipc_connect->__tipc_port_connect __tipc_disconnect->__tipc_port_disconnect tipc_disconnect->tipc_port_disconnect tipc_shutdown->tipc_port_shutdown tipc_port_recv_msg->tipc_port_rcv tipc_port_recv_sections->tipc_port_iovec_rcv release->tipc_release accept->tipc_accept bind->tipc_bind get_name->tipc_getname poll->tipc_poll send_msg->tipc_sendmsg send_packet->tipc_send_packet send_stream->tipc_send_stream recv_msg->tipc_recvmsg recv_stream->tipc_recv_stream connect->tipc_connect listen->tipc_listen shutdown->tipc_shutdown setsockopt->tipc_setsockopt getsockopt->tipc_getsockopt Above changes have no impact on current users of the functions. Signed-off-by: Ying Xue <ying.xue@windriver.com> Reviewed-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-18 16:06:46 +08:00
tipc_link_proto_xmit(l_ptr, STATE_MSG, 1, 0, 0, 0, 0);
l_ptr->fsm_msg_cnt++;
link_set_timer(l_ptr, cont_intv / 4);
break;
case RESET_MSG:
pr_debug("%s<%s>, requested by peer\n",
link_rst_msg, l_ptr->name);
tipc_link_reset(l_ptr);
l_ptr->state = RESET_RESET;
l_ptr->fsm_msg_cnt = 0;
tipc: align tipc function names with common naming practice in the network Rename the following functions, which are shorter and more in line with common naming practice in the network subsystem. tipc_bclink_send_msg->tipc_bclink_xmit tipc_bclink_recv_pkt->tipc_bclink_rcv tipc_disc_recv_msg->tipc_disc_rcv tipc_link_send_proto_msg->tipc_link_proto_xmit link_recv_proto_msg->tipc_link_proto_rcv link_send_sections_long->tipc_link_iovec_long_xmit tipc_link_send_sections_fast->tipc_link_iovec_xmit_fast tipc_link_send_sync->tipc_link_sync_xmit tipc_link_recv_sync->tipc_link_sync_rcv tipc_link_send_buf->__tipc_link_xmit tipc_link_send->tipc_link_xmit tipc_link_send_names->tipc_link_names_xmit tipc_named_recv->tipc_named_rcv tipc_link_recv_bundle->tipc_link_bundle_rcv tipc_link_dup_send_queue->tipc_link_dup_queue_xmit link_send_long_buf->tipc_link_frag_xmit tipc_multicast->tipc_port_mcast_xmit tipc_port_recv_mcast->tipc_port_mcast_rcv tipc_port_reject_sections->tipc_port_iovec_reject tipc_port_recv_proto_msg->tipc_port_proto_rcv tipc_connect->tipc_port_connect __tipc_connect->__tipc_port_connect __tipc_disconnect->__tipc_port_disconnect tipc_disconnect->tipc_port_disconnect tipc_shutdown->tipc_port_shutdown tipc_port_recv_msg->tipc_port_rcv tipc_port_recv_sections->tipc_port_iovec_rcv release->tipc_release accept->tipc_accept bind->tipc_bind get_name->tipc_getname poll->tipc_poll send_msg->tipc_sendmsg send_packet->tipc_send_packet send_stream->tipc_send_stream recv_msg->tipc_recvmsg recv_stream->tipc_recv_stream connect->tipc_connect listen->tipc_listen shutdown->tipc_shutdown setsockopt->tipc_setsockopt getsockopt->tipc_getsockopt Above changes have no impact on current users of the functions. Signed-off-by: Ying Xue <ying.xue@windriver.com> Reviewed-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-18 16:06:46 +08:00
tipc_link_proto_xmit(l_ptr, ACTIVATE_MSG,
0, 0, 0, 0, 0);
l_ptr->fsm_msg_cnt++;
link_set_timer(l_ptr, cont_intv);
break;
default:
pr_debug("%s%u in WW state\n", link_unk_evt, event);
}
break;
case WORKING_UNKNOWN:
switch (event) {
case TRAFFIC_MSG_EVT:
case ACTIVATE_MSG:
l_ptr->state = WORKING_WORKING;
l_ptr->fsm_msg_cnt = 0;
link_set_timer(l_ptr, cont_intv);
break;
case RESET_MSG:
pr_debug("%s<%s>, requested by peer while probing\n",
link_rst_msg, l_ptr->name);
tipc_link_reset(l_ptr);
l_ptr->state = RESET_RESET;
l_ptr->fsm_msg_cnt = 0;
tipc: align tipc function names with common naming practice in the network Rename the following functions, which are shorter and more in line with common naming practice in the network subsystem. tipc_bclink_send_msg->tipc_bclink_xmit tipc_bclink_recv_pkt->tipc_bclink_rcv tipc_disc_recv_msg->tipc_disc_rcv tipc_link_send_proto_msg->tipc_link_proto_xmit link_recv_proto_msg->tipc_link_proto_rcv link_send_sections_long->tipc_link_iovec_long_xmit tipc_link_send_sections_fast->tipc_link_iovec_xmit_fast tipc_link_send_sync->tipc_link_sync_xmit tipc_link_recv_sync->tipc_link_sync_rcv tipc_link_send_buf->__tipc_link_xmit tipc_link_send->tipc_link_xmit tipc_link_send_names->tipc_link_names_xmit tipc_named_recv->tipc_named_rcv tipc_link_recv_bundle->tipc_link_bundle_rcv tipc_link_dup_send_queue->tipc_link_dup_queue_xmit link_send_long_buf->tipc_link_frag_xmit tipc_multicast->tipc_port_mcast_xmit tipc_port_recv_mcast->tipc_port_mcast_rcv tipc_port_reject_sections->tipc_port_iovec_reject tipc_port_recv_proto_msg->tipc_port_proto_rcv tipc_connect->tipc_port_connect __tipc_connect->__tipc_port_connect __tipc_disconnect->__tipc_port_disconnect tipc_disconnect->tipc_port_disconnect tipc_shutdown->tipc_port_shutdown tipc_port_recv_msg->tipc_port_rcv tipc_port_recv_sections->tipc_port_iovec_rcv release->tipc_release accept->tipc_accept bind->tipc_bind get_name->tipc_getname poll->tipc_poll send_msg->tipc_sendmsg send_packet->tipc_send_packet send_stream->tipc_send_stream recv_msg->tipc_recvmsg recv_stream->tipc_recv_stream connect->tipc_connect listen->tipc_listen shutdown->tipc_shutdown setsockopt->tipc_setsockopt getsockopt->tipc_getsockopt Above changes have no impact on current users of the functions. Signed-off-by: Ying Xue <ying.xue@windriver.com> Reviewed-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-18 16:06:46 +08:00
tipc_link_proto_xmit(l_ptr, ACTIVATE_MSG,
0, 0, 0, 0, 0);
l_ptr->fsm_msg_cnt++;
link_set_timer(l_ptr, cont_intv);
break;
case TIMEOUT_EVT:
if (l_ptr->next_in_no != l_ptr->checkpoint) {
l_ptr->state = WORKING_WORKING;
l_ptr->fsm_msg_cnt = 0;
l_ptr->checkpoint = l_ptr->next_in_no;
if (tipc_bclink_acks_missing(l_ptr->owner)) {
tipc: align tipc function names with common naming practice in the network Rename the following functions, which are shorter and more in line with common naming practice in the network subsystem. tipc_bclink_send_msg->tipc_bclink_xmit tipc_bclink_recv_pkt->tipc_bclink_rcv tipc_disc_recv_msg->tipc_disc_rcv tipc_link_send_proto_msg->tipc_link_proto_xmit link_recv_proto_msg->tipc_link_proto_rcv link_send_sections_long->tipc_link_iovec_long_xmit tipc_link_send_sections_fast->tipc_link_iovec_xmit_fast tipc_link_send_sync->tipc_link_sync_xmit tipc_link_recv_sync->tipc_link_sync_rcv tipc_link_send_buf->__tipc_link_xmit tipc_link_send->tipc_link_xmit tipc_link_send_names->tipc_link_names_xmit tipc_named_recv->tipc_named_rcv tipc_link_recv_bundle->tipc_link_bundle_rcv tipc_link_dup_send_queue->tipc_link_dup_queue_xmit link_send_long_buf->tipc_link_frag_xmit tipc_multicast->tipc_port_mcast_xmit tipc_port_recv_mcast->tipc_port_mcast_rcv tipc_port_reject_sections->tipc_port_iovec_reject tipc_port_recv_proto_msg->tipc_port_proto_rcv tipc_connect->tipc_port_connect __tipc_connect->__tipc_port_connect __tipc_disconnect->__tipc_port_disconnect tipc_disconnect->tipc_port_disconnect tipc_shutdown->tipc_port_shutdown tipc_port_recv_msg->tipc_port_rcv tipc_port_recv_sections->tipc_port_iovec_rcv release->tipc_release accept->tipc_accept bind->tipc_bind get_name->tipc_getname poll->tipc_poll send_msg->tipc_sendmsg send_packet->tipc_send_packet send_stream->tipc_send_stream recv_msg->tipc_recvmsg recv_stream->tipc_recv_stream connect->tipc_connect listen->tipc_listen shutdown->tipc_shutdown setsockopt->tipc_setsockopt getsockopt->tipc_getsockopt Above changes have no impact on current users of the functions. Signed-off-by: Ying Xue <ying.xue@windriver.com> Reviewed-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-18 16:06:46 +08:00
tipc_link_proto_xmit(l_ptr, STATE_MSG,
0, 0, 0, 0, 0);
l_ptr->fsm_msg_cnt++;
}
link_set_timer(l_ptr, cont_intv);
} else if (l_ptr->fsm_msg_cnt < l_ptr->abort_limit) {
tipc: align tipc function names with common naming practice in the network Rename the following functions, which are shorter and more in line with common naming practice in the network subsystem. tipc_bclink_send_msg->tipc_bclink_xmit tipc_bclink_recv_pkt->tipc_bclink_rcv tipc_disc_recv_msg->tipc_disc_rcv tipc_link_send_proto_msg->tipc_link_proto_xmit link_recv_proto_msg->tipc_link_proto_rcv link_send_sections_long->tipc_link_iovec_long_xmit tipc_link_send_sections_fast->tipc_link_iovec_xmit_fast tipc_link_send_sync->tipc_link_sync_xmit tipc_link_recv_sync->tipc_link_sync_rcv tipc_link_send_buf->__tipc_link_xmit tipc_link_send->tipc_link_xmit tipc_link_send_names->tipc_link_names_xmit tipc_named_recv->tipc_named_rcv tipc_link_recv_bundle->tipc_link_bundle_rcv tipc_link_dup_send_queue->tipc_link_dup_queue_xmit link_send_long_buf->tipc_link_frag_xmit tipc_multicast->tipc_port_mcast_xmit tipc_port_recv_mcast->tipc_port_mcast_rcv tipc_port_reject_sections->tipc_port_iovec_reject tipc_port_recv_proto_msg->tipc_port_proto_rcv tipc_connect->tipc_port_connect __tipc_connect->__tipc_port_connect __tipc_disconnect->__tipc_port_disconnect tipc_disconnect->tipc_port_disconnect tipc_shutdown->tipc_port_shutdown tipc_port_recv_msg->tipc_port_rcv tipc_port_recv_sections->tipc_port_iovec_rcv release->tipc_release accept->tipc_accept bind->tipc_bind get_name->tipc_getname poll->tipc_poll send_msg->tipc_sendmsg send_packet->tipc_send_packet send_stream->tipc_send_stream recv_msg->tipc_recvmsg recv_stream->tipc_recv_stream connect->tipc_connect listen->tipc_listen shutdown->tipc_shutdown setsockopt->tipc_setsockopt getsockopt->tipc_getsockopt Above changes have no impact on current users of the functions. Signed-off-by: Ying Xue <ying.xue@windriver.com> Reviewed-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-18 16:06:46 +08:00
tipc_link_proto_xmit(l_ptr, STATE_MSG,
1, 0, 0, 0, 0);
l_ptr->fsm_msg_cnt++;
link_set_timer(l_ptr, cont_intv / 4);
} else { /* Link has failed */
pr_debug("%s<%s>, peer not responding\n",
link_rst_msg, l_ptr->name);
tipc_link_reset(l_ptr);
l_ptr->state = RESET_UNKNOWN;
l_ptr->fsm_msg_cnt = 0;
tipc: align tipc function names with common naming practice in the network Rename the following functions, which are shorter and more in line with common naming practice in the network subsystem. tipc_bclink_send_msg->tipc_bclink_xmit tipc_bclink_recv_pkt->tipc_bclink_rcv tipc_disc_recv_msg->tipc_disc_rcv tipc_link_send_proto_msg->tipc_link_proto_xmit link_recv_proto_msg->tipc_link_proto_rcv link_send_sections_long->tipc_link_iovec_long_xmit tipc_link_send_sections_fast->tipc_link_iovec_xmit_fast tipc_link_send_sync->tipc_link_sync_xmit tipc_link_recv_sync->tipc_link_sync_rcv tipc_link_send_buf->__tipc_link_xmit tipc_link_send->tipc_link_xmit tipc_link_send_names->tipc_link_names_xmit tipc_named_recv->tipc_named_rcv tipc_link_recv_bundle->tipc_link_bundle_rcv tipc_link_dup_send_queue->tipc_link_dup_queue_xmit link_send_long_buf->tipc_link_frag_xmit tipc_multicast->tipc_port_mcast_xmit tipc_port_recv_mcast->tipc_port_mcast_rcv tipc_port_reject_sections->tipc_port_iovec_reject tipc_port_recv_proto_msg->tipc_port_proto_rcv tipc_connect->tipc_port_connect __tipc_connect->__tipc_port_connect __tipc_disconnect->__tipc_port_disconnect tipc_disconnect->tipc_port_disconnect tipc_shutdown->tipc_port_shutdown tipc_port_recv_msg->tipc_port_rcv tipc_port_recv_sections->tipc_port_iovec_rcv release->tipc_release accept->tipc_accept bind->tipc_bind get_name->tipc_getname poll->tipc_poll send_msg->tipc_sendmsg send_packet->tipc_send_packet send_stream->tipc_send_stream recv_msg->tipc_recvmsg recv_stream->tipc_recv_stream connect->tipc_connect listen->tipc_listen shutdown->tipc_shutdown setsockopt->tipc_setsockopt getsockopt->tipc_getsockopt Above changes have no impact on current users of the functions. Signed-off-by: Ying Xue <ying.xue@windriver.com> Reviewed-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-18 16:06:46 +08:00
tipc_link_proto_xmit(l_ptr, RESET_MSG,
0, 0, 0, 0, 0);
l_ptr->fsm_msg_cnt++;
link_set_timer(l_ptr, cont_intv);
}
break;
default:
pr_err("%s%u in WU state\n", link_unk_evt, event);
}
break;
case RESET_UNKNOWN:
switch (event) {
case TRAFFIC_MSG_EVT:
break;
case ACTIVATE_MSG:
other = l_ptr->owner->active_links[0];
if (other && link_working_unknown(other))
break;
l_ptr->state = WORKING_WORKING;
l_ptr->fsm_msg_cnt = 0;
link_activate(l_ptr);
tipc: align tipc function names with common naming practice in the network Rename the following functions, which are shorter and more in line with common naming practice in the network subsystem. tipc_bclink_send_msg->tipc_bclink_xmit tipc_bclink_recv_pkt->tipc_bclink_rcv tipc_disc_recv_msg->tipc_disc_rcv tipc_link_send_proto_msg->tipc_link_proto_xmit link_recv_proto_msg->tipc_link_proto_rcv link_send_sections_long->tipc_link_iovec_long_xmit tipc_link_send_sections_fast->tipc_link_iovec_xmit_fast tipc_link_send_sync->tipc_link_sync_xmit tipc_link_recv_sync->tipc_link_sync_rcv tipc_link_send_buf->__tipc_link_xmit tipc_link_send->tipc_link_xmit tipc_link_send_names->tipc_link_names_xmit tipc_named_recv->tipc_named_rcv tipc_link_recv_bundle->tipc_link_bundle_rcv tipc_link_dup_send_queue->tipc_link_dup_queue_xmit link_send_long_buf->tipc_link_frag_xmit tipc_multicast->tipc_port_mcast_xmit tipc_port_recv_mcast->tipc_port_mcast_rcv tipc_port_reject_sections->tipc_port_iovec_reject tipc_port_recv_proto_msg->tipc_port_proto_rcv tipc_connect->tipc_port_connect __tipc_connect->__tipc_port_connect __tipc_disconnect->__tipc_port_disconnect tipc_disconnect->tipc_port_disconnect tipc_shutdown->tipc_port_shutdown tipc_port_recv_msg->tipc_port_rcv tipc_port_recv_sections->tipc_port_iovec_rcv release->tipc_release accept->tipc_accept bind->tipc_bind get_name->tipc_getname poll->tipc_poll send_msg->tipc_sendmsg send_packet->tipc_send_packet send_stream->tipc_send_stream recv_msg->tipc_recvmsg recv_stream->tipc_recv_stream connect->tipc_connect listen->tipc_listen shutdown->tipc_shutdown setsockopt->tipc_setsockopt getsockopt->tipc_getsockopt Above changes have no impact on current users of the functions. Signed-off-by: Ying Xue <ying.xue@windriver.com> Reviewed-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-18 16:06:46 +08:00
tipc_link_proto_xmit(l_ptr, STATE_MSG, 1, 0, 0, 0, 0);
l_ptr->fsm_msg_cnt++;
tipc: introduce message to synchronize broadcast link Upon establishing a first link between two nodes, there is currently a risk that the two endpoints will disagree on exactly which sequence number reception and acknowleding of broadcast packets should start. The following scenarios may happen: 1: Node A sends an ACTIVATE message to B, telling it to start acking packets from sequence number N. 2: Node A sends out broadcast N, but does not expect an acknowledge from B, since B is not yet in its broadcast receiver's list. 3: Node A receives ACK for N from all nodes except B, and releases packet N. 4: Node B receives the ACTIVATE, activates its link endpoint, and stores the value N as sequence number of first expected packet. 5: Node B sends a NAME_DISTR message to A. 6: Node A receives the NAME_DISTR message, and activates its endpoint. At this moment B is added to A's broadcast receiver's set. Node A also sets sequence number 0 as the first broadcast packet to be received from B. 7: Node A sends broadcast N+1. 8: B receives N+1, determines there is a gap in the sequence, since it is expecting N, and sends a NACK for N back to A. 9: Node A has already released N, so no retransmission is possible. The broadcast link in direction A->B is stale. In addition to, or instead of, 7-9 above, the following may happen: 10: Node B sends broadcast M > 0 to A. 11: Node A receives M, falsely decides there must be a gap, since it is expecting packet 0, and asks for retransmission of packets [0,M-1]. 12: Node B has already released these packets, so the broadcast link is stale in direction B->A. We solve this problem by introducing a new unicast message type, BCAST_PROTOCOL/STATE, to convey the sequence number of the next sent broadcast packet to the other endpoint, at exactly the moment that endpoint is added to the own node's broadcast receivers list, and before any other unicast messages are permitted to be sent. Furthermore, we don't allow any node to start receiving and processing broadcast packets until this new synchronization message has been received. To maintain backwards compatibility, we still open up for broadcast reception if we receive a NAME_DISTR message without any preceding broadcast sync message. In this case, we must assume that the other end has an older code version, and will never send out the new synchronization message. Hence, for mixed old and new nodes, the issue arising in 7-12 of the above may happen with the same probability as before. Signed-off-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: Ying Xue <ying.xue@windriver.com> Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
2012-11-16 13:51:31 +08:00
if (l_ptr->owner->working_links == 1)
tipc: align tipc function names with common naming practice in the network Rename the following functions, which are shorter and more in line with common naming practice in the network subsystem. tipc_bclink_send_msg->tipc_bclink_xmit tipc_bclink_recv_pkt->tipc_bclink_rcv tipc_disc_recv_msg->tipc_disc_rcv tipc_link_send_proto_msg->tipc_link_proto_xmit link_recv_proto_msg->tipc_link_proto_rcv link_send_sections_long->tipc_link_iovec_long_xmit tipc_link_send_sections_fast->tipc_link_iovec_xmit_fast tipc_link_send_sync->tipc_link_sync_xmit tipc_link_recv_sync->tipc_link_sync_rcv tipc_link_send_buf->__tipc_link_xmit tipc_link_send->tipc_link_xmit tipc_link_send_names->tipc_link_names_xmit tipc_named_recv->tipc_named_rcv tipc_link_recv_bundle->tipc_link_bundle_rcv tipc_link_dup_send_queue->tipc_link_dup_queue_xmit link_send_long_buf->tipc_link_frag_xmit tipc_multicast->tipc_port_mcast_xmit tipc_port_recv_mcast->tipc_port_mcast_rcv tipc_port_reject_sections->tipc_port_iovec_reject tipc_port_recv_proto_msg->tipc_port_proto_rcv tipc_connect->tipc_port_connect __tipc_connect->__tipc_port_connect __tipc_disconnect->__tipc_port_disconnect tipc_disconnect->tipc_port_disconnect tipc_shutdown->tipc_port_shutdown tipc_port_recv_msg->tipc_port_rcv tipc_port_recv_sections->tipc_port_iovec_rcv release->tipc_release accept->tipc_accept bind->tipc_bind get_name->tipc_getname poll->tipc_poll send_msg->tipc_sendmsg send_packet->tipc_send_packet send_stream->tipc_send_stream recv_msg->tipc_recvmsg recv_stream->tipc_recv_stream connect->tipc_connect listen->tipc_listen shutdown->tipc_shutdown setsockopt->tipc_setsockopt getsockopt->tipc_getsockopt Above changes have no impact on current users of the functions. Signed-off-by: Ying Xue <ying.xue@windriver.com> Reviewed-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-18 16:06:46 +08:00
tipc_link_sync_xmit(l_ptr);
link_set_timer(l_ptr, cont_intv);
break;
case RESET_MSG:
l_ptr->state = RESET_RESET;
l_ptr->fsm_msg_cnt = 0;
tipc: align tipc function names with common naming practice in the network Rename the following functions, which are shorter and more in line with common naming practice in the network subsystem. tipc_bclink_send_msg->tipc_bclink_xmit tipc_bclink_recv_pkt->tipc_bclink_rcv tipc_disc_recv_msg->tipc_disc_rcv tipc_link_send_proto_msg->tipc_link_proto_xmit link_recv_proto_msg->tipc_link_proto_rcv link_send_sections_long->tipc_link_iovec_long_xmit tipc_link_send_sections_fast->tipc_link_iovec_xmit_fast tipc_link_send_sync->tipc_link_sync_xmit tipc_link_recv_sync->tipc_link_sync_rcv tipc_link_send_buf->__tipc_link_xmit tipc_link_send->tipc_link_xmit tipc_link_send_names->tipc_link_names_xmit tipc_named_recv->tipc_named_rcv tipc_link_recv_bundle->tipc_link_bundle_rcv tipc_link_dup_send_queue->tipc_link_dup_queue_xmit link_send_long_buf->tipc_link_frag_xmit tipc_multicast->tipc_port_mcast_xmit tipc_port_recv_mcast->tipc_port_mcast_rcv tipc_port_reject_sections->tipc_port_iovec_reject tipc_port_recv_proto_msg->tipc_port_proto_rcv tipc_connect->tipc_port_connect __tipc_connect->__tipc_port_connect __tipc_disconnect->__tipc_port_disconnect tipc_disconnect->tipc_port_disconnect tipc_shutdown->tipc_port_shutdown tipc_port_recv_msg->tipc_port_rcv tipc_port_recv_sections->tipc_port_iovec_rcv release->tipc_release accept->tipc_accept bind->tipc_bind get_name->tipc_getname poll->tipc_poll send_msg->tipc_sendmsg send_packet->tipc_send_packet send_stream->tipc_send_stream recv_msg->tipc_recvmsg recv_stream->tipc_recv_stream connect->tipc_connect listen->tipc_listen shutdown->tipc_shutdown setsockopt->tipc_setsockopt getsockopt->tipc_getsockopt Above changes have no impact on current users of the functions. Signed-off-by: Ying Xue <ying.xue@windriver.com> Reviewed-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-18 16:06:46 +08:00
tipc_link_proto_xmit(l_ptr, ACTIVATE_MSG,
1, 0, 0, 0, 0);
l_ptr->fsm_msg_cnt++;
link_set_timer(l_ptr, cont_intv);
break;
case STARTING_EVT:
l_ptr->flags |= LINK_STARTED;
l_ptr->fsm_msg_cnt++;
link_set_timer(l_ptr, cont_intv);
break;
case TIMEOUT_EVT:
tipc: align tipc function names with common naming practice in the network Rename the following functions, which are shorter and more in line with common naming practice in the network subsystem. tipc_bclink_send_msg->tipc_bclink_xmit tipc_bclink_recv_pkt->tipc_bclink_rcv tipc_disc_recv_msg->tipc_disc_rcv tipc_link_send_proto_msg->tipc_link_proto_xmit link_recv_proto_msg->tipc_link_proto_rcv link_send_sections_long->tipc_link_iovec_long_xmit tipc_link_send_sections_fast->tipc_link_iovec_xmit_fast tipc_link_send_sync->tipc_link_sync_xmit tipc_link_recv_sync->tipc_link_sync_rcv tipc_link_send_buf->__tipc_link_xmit tipc_link_send->tipc_link_xmit tipc_link_send_names->tipc_link_names_xmit tipc_named_recv->tipc_named_rcv tipc_link_recv_bundle->tipc_link_bundle_rcv tipc_link_dup_send_queue->tipc_link_dup_queue_xmit link_send_long_buf->tipc_link_frag_xmit tipc_multicast->tipc_port_mcast_xmit tipc_port_recv_mcast->tipc_port_mcast_rcv tipc_port_reject_sections->tipc_port_iovec_reject tipc_port_recv_proto_msg->tipc_port_proto_rcv tipc_connect->tipc_port_connect __tipc_connect->__tipc_port_connect __tipc_disconnect->__tipc_port_disconnect tipc_disconnect->tipc_port_disconnect tipc_shutdown->tipc_port_shutdown tipc_port_recv_msg->tipc_port_rcv tipc_port_recv_sections->tipc_port_iovec_rcv release->tipc_release accept->tipc_accept bind->tipc_bind get_name->tipc_getname poll->tipc_poll send_msg->tipc_sendmsg send_packet->tipc_send_packet send_stream->tipc_send_stream recv_msg->tipc_recvmsg recv_stream->tipc_recv_stream connect->tipc_connect listen->tipc_listen shutdown->tipc_shutdown setsockopt->tipc_setsockopt getsockopt->tipc_getsockopt Above changes have no impact on current users of the functions. Signed-off-by: Ying Xue <ying.xue@windriver.com> Reviewed-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-18 16:06:46 +08:00
tipc_link_proto_xmit(l_ptr, RESET_MSG, 0, 0, 0, 0, 0);
l_ptr->fsm_msg_cnt++;
link_set_timer(l_ptr, cont_intv);
break;
default:
pr_err("%s%u in RU state\n", link_unk_evt, event);
}
break;
case RESET_RESET:
switch (event) {
case TRAFFIC_MSG_EVT:
case ACTIVATE_MSG:
other = l_ptr->owner->active_links[0];
if (other && link_working_unknown(other))
break;
l_ptr->state = WORKING_WORKING;
l_ptr->fsm_msg_cnt = 0;
link_activate(l_ptr);
tipc: align tipc function names with common naming practice in the network Rename the following functions, which are shorter and more in line with common naming practice in the network subsystem. tipc_bclink_send_msg->tipc_bclink_xmit tipc_bclink_recv_pkt->tipc_bclink_rcv tipc_disc_recv_msg->tipc_disc_rcv tipc_link_send_proto_msg->tipc_link_proto_xmit link_recv_proto_msg->tipc_link_proto_rcv link_send_sections_long->tipc_link_iovec_long_xmit tipc_link_send_sections_fast->tipc_link_iovec_xmit_fast tipc_link_send_sync->tipc_link_sync_xmit tipc_link_recv_sync->tipc_link_sync_rcv tipc_link_send_buf->__tipc_link_xmit tipc_link_send->tipc_link_xmit tipc_link_send_names->tipc_link_names_xmit tipc_named_recv->tipc_named_rcv tipc_link_recv_bundle->tipc_link_bundle_rcv tipc_link_dup_send_queue->tipc_link_dup_queue_xmit link_send_long_buf->tipc_link_frag_xmit tipc_multicast->tipc_port_mcast_xmit tipc_port_recv_mcast->tipc_port_mcast_rcv tipc_port_reject_sections->tipc_port_iovec_reject tipc_port_recv_proto_msg->tipc_port_proto_rcv tipc_connect->tipc_port_connect __tipc_connect->__tipc_port_connect __tipc_disconnect->__tipc_port_disconnect tipc_disconnect->tipc_port_disconnect tipc_shutdown->tipc_port_shutdown tipc_port_recv_msg->tipc_port_rcv tipc_port_recv_sections->tipc_port_iovec_rcv release->tipc_release accept->tipc_accept bind->tipc_bind get_name->tipc_getname poll->tipc_poll send_msg->tipc_sendmsg send_packet->tipc_send_packet send_stream->tipc_send_stream recv_msg->tipc_recvmsg recv_stream->tipc_recv_stream connect->tipc_connect listen->tipc_listen shutdown->tipc_shutdown setsockopt->tipc_setsockopt getsockopt->tipc_getsockopt Above changes have no impact on current users of the functions. Signed-off-by: Ying Xue <ying.xue@windriver.com> Reviewed-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-18 16:06:46 +08:00
tipc_link_proto_xmit(l_ptr, STATE_MSG, 1, 0, 0, 0, 0);
l_ptr->fsm_msg_cnt++;
tipc: introduce message to synchronize broadcast link Upon establishing a first link between two nodes, there is currently a risk that the two endpoints will disagree on exactly which sequence number reception and acknowleding of broadcast packets should start. The following scenarios may happen: 1: Node A sends an ACTIVATE message to B, telling it to start acking packets from sequence number N. 2: Node A sends out broadcast N, but does not expect an acknowledge from B, since B is not yet in its broadcast receiver's list. 3: Node A receives ACK for N from all nodes except B, and releases packet N. 4: Node B receives the ACTIVATE, activates its link endpoint, and stores the value N as sequence number of first expected packet. 5: Node B sends a NAME_DISTR message to A. 6: Node A receives the NAME_DISTR message, and activates its endpoint. At this moment B is added to A's broadcast receiver's set. Node A also sets sequence number 0 as the first broadcast packet to be received from B. 7: Node A sends broadcast N+1. 8: B receives N+1, determines there is a gap in the sequence, since it is expecting N, and sends a NACK for N back to A. 9: Node A has already released N, so no retransmission is possible. The broadcast link in direction A->B is stale. In addition to, or instead of, 7-9 above, the following may happen: 10: Node B sends broadcast M > 0 to A. 11: Node A receives M, falsely decides there must be a gap, since it is expecting packet 0, and asks for retransmission of packets [0,M-1]. 12: Node B has already released these packets, so the broadcast link is stale in direction B->A. We solve this problem by introducing a new unicast message type, BCAST_PROTOCOL/STATE, to convey the sequence number of the next sent broadcast packet to the other endpoint, at exactly the moment that endpoint is added to the own node's broadcast receivers list, and before any other unicast messages are permitted to be sent. Furthermore, we don't allow any node to start receiving and processing broadcast packets until this new synchronization message has been received. To maintain backwards compatibility, we still open up for broadcast reception if we receive a NAME_DISTR message without any preceding broadcast sync message. In this case, we must assume that the other end has an older code version, and will never send out the new synchronization message. Hence, for mixed old and new nodes, the issue arising in 7-12 of the above may happen with the same probability as before. Signed-off-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: Ying Xue <ying.xue@windriver.com> Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
2012-11-16 13:51:31 +08:00
if (l_ptr->owner->working_links == 1)
tipc: align tipc function names with common naming practice in the network Rename the following functions, which are shorter and more in line with common naming practice in the network subsystem. tipc_bclink_send_msg->tipc_bclink_xmit tipc_bclink_recv_pkt->tipc_bclink_rcv tipc_disc_recv_msg->tipc_disc_rcv tipc_link_send_proto_msg->tipc_link_proto_xmit link_recv_proto_msg->tipc_link_proto_rcv link_send_sections_long->tipc_link_iovec_long_xmit tipc_link_send_sections_fast->tipc_link_iovec_xmit_fast tipc_link_send_sync->tipc_link_sync_xmit tipc_link_recv_sync->tipc_link_sync_rcv tipc_link_send_buf->__tipc_link_xmit tipc_link_send->tipc_link_xmit tipc_link_send_names->tipc_link_names_xmit tipc_named_recv->tipc_named_rcv tipc_link_recv_bundle->tipc_link_bundle_rcv tipc_link_dup_send_queue->tipc_link_dup_queue_xmit link_send_long_buf->tipc_link_frag_xmit tipc_multicast->tipc_port_mcast_xmit tipc_port_recv_mcast->tipc_port_mcast_rcv tipc_port_reject_sections->tipc_port_iovec_reject tipc_port_recv_proto_msg->tipc_port_proto_rcv tipc_connect->tipc_port_connect __tipc_connect->__tipc_port_connect __tipc_disconnect->__tipc_port_disconnect tipc_disconnect->tipc_port_disconnect tipc_shutdown->tipc_port_shutdown tipc_port_recv_msg->tipc_port_rcv tipc_port_recv_sections->tipc_port_iovec_rcv release->tipc_release accept->tipc_accept bind->tipc_bind get_name->tipc_getname poll->tipc_poll send_msg->tipc_sendmsg send_packet->tipc_send_packet send_stream->tipc_send_stream recv_msg->tipc_recvmsg recv_stream->tipc_recv_stream connect->tipc_connect listen->tipc_listen shutdown->tipc_shutdown setsockopt->tipc_setsockopt getsockopt->tipc_getsockopt Above changes have no impact on current users of the functions. Signed-off-by: Ying Xue <ying.xue@windriver.com> Reviewed-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-18 16:06:46 +08:00
tipc_link_sync_xmit(l_ptr);
link_set_timer(l_ptr, cont_intv);
break;
case RESET_MSG:
break;
case TIMEOUT_EVT:
tipc: align tipc function names with common naming practice in the network Rename the following functions, which are shorter and more in line with common naming practice in the network subsystem. tipc_bclink_send_msg->tipc_bclink_xmit tipc_bclink_recv_pkt->tipc_bclink_rcv tipc_disc_recv_msg->tipc_disc_rcv tipc_link_send_proto_msg->tipc_link_proto_xmit link_recv_proto_msg->tipc_link_proto_rcv link_send_sections_long->tipc_link_iovec_long_xmit tipc_link_send_sections_fast->tipc_link_iovec_xmit_fast tipc_link_send_sync->tipc_link_sync_xmit tipc_link_recv_sync->tipc_link_sync_rcv tipc_link_send_buf->__tipc_link_xmit tipc_link_send->tipc_link_xmit tipc_link_send_names->tipc_link_names_xmit tipc_named_recv->tipc_named_rcv tipc_link_recv_bundle->tipc_link_bundle_rcv tipc_link_dup_send_queue->tipc_link_dup_queue_xmit link_send_long_buf->tipc_link_frag_xmit tipc_multicast->tipc_port_mcast_xmit tipc_port_recv_mcast->tipc_port_mcast_rcv tipc_port_reject_sections->tipc_port_iovec_reject tipc_port_recv_proto_msg->tipc_port_proto_rcv tipc_connect->tipc_port_connect __tipc_connect->__tipc_port_connect __tipc_disconnect->__tipc_port_disconnect tipc_disconnect->tipc_port_disconnect tipc_shutdown->tipc_port_shutdown tipc_port_recv_msg->tipc_port_rcv tipc_port_recv_sections->tipc_port_iovec_rcv release->tipc_release accept->tipc_accept bind->tipc_bind get_name->tipc_getname poll->tipc_poll send_msg->tipc_sendmsg send_packet->tipc_send_packet send_stream->tipc_send_stream recv_msg->tipc_recvmsg recv_stream->tipc_recv_stream connect->tipc_connect listen->tipc_listen shutdown->tipc_shutdown setsockopt->tipc_setsockopt getsockopt->tipc_getsockopt Above changes have no impact on current users of the functions. Signed-off-by: Ying Xue <ying.xue@windriver.com> Reviewed-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-18 16:06:46 +08:00
tipc_link_proto_xmit(l_ptr, ACTIVATE_MSG,
0, 0, 0, 0, 0);
l_ptr->fsm_msg_cnt++;
link_set_timer(l_ptr, cont_intv);
break;
default:
pr_err("%s%u in RR state\n", link_unk_evt, event);
}
break;
default:
pr_err("Unknown link state %u/%u\n", l_ptr->state, event);
}
}
/* tipc_link_cong: determine return value and how to treat the
* sent buffer during link congestion.
* - For plain, errorless user data messages we keep the buffer and
* return -ELINKONG.
* - For all other messages we discard the buffer and return -EHOSTUNREACH
* - For TIPC internal messages we also reset the link
*/
static int tipc_link_cong(struct tipc_link *link, struct sk_buff_head *list)
{
struct sk_buff *skb = skb_peek(list);
struct tipc_msg *msg = buf_msg(skb);
uint imp = tipc_msg_tot_importance(msg);
u32 oport = msg_tot_origport(msg);
if (unlikely(imp > TIPC_CRITICAL_IMPORTANCE)) {
pr_warn("%s<%s>, send queue full", link_rst_msg, link->name);
tipc_link_reset(link);
goto drop;
}
if (unlikely(msg_errcode(msg)))
goto drop;
if (unlikely(msg_reroute_cnt(msg)))
goto drop;
if (TIPC_SKB_CB(skb)->wakeup_pending)
return -ELINKCONG;
if (link_schedule_user(link, oport, skb_queue_len(list), imp))
return -ELINKCONG;
drop:
__skb_queue_purge(list);
return -EHOSTUNREACH;
}
/**
* __tipc_link_xmit(): same as tipc_link_xmit, but destlink is known & locked
* @link: link to use
* @list: chain of buffers containing message
*
* Consumes the buffer chain, except when returning -ELINKCONG
* Returns 0 if success, otherwise errno: -ELINKCONG, -EMSGSIZE (plain socket
* user data messages) or -EHOSTUNREACH (all other messages/senders)
* Only the socket functions tipc_send_stream() and tipc_send_packet() need
* to act on the return value, since they may need to do more send attempts.
*/
int __tipc_link_xmit(struct net *net, struct tipc_link *link,
struct sk_buff_head *list)
{
struct tipc_msg *msg = buf_msg(skb_peek(list));
uint psz = msg_size(msg);
uint sndlim = link->queue_limit[0];
uint imp = tipc_msg_tot_importance(msg);
uint mtu = link->max_pkt;
uint ack = mod(link->next_in_no - 1);
uint seqno = link->next_out_no;
uint bc_last_in = link->owner->bclink.last_in;
struct tipc_media_addr *addr = &link->media_addr;
struct sk_buff_head *outqueue = &link->outqueue;
struct sk_buff *skb, *tmp;
/* Match queue limits against msg importance: */
if (unlikely(skb_queue_len(outqueue) >= link->queue_limit[imp]))
return tipc_link_cong(link, list);
/* Has valid packet limit been used ? */
if (unlikely(psz > mtu)) {
__skb_queue_purge(list);
return -EMSGSIZE;
}
/* Prepare each packet for sending, and add to outqueue: */
skb_queue_walk_safe(list, skb, tmp) {
__skb_unlink(skb, list);
msg = buf_msg(skb);
msg_set_word(msg, 2, ((ack << 16) | mod(seqno)));
msg_set_bcast_ack(msg, bc_last_in);
if (skb_queue_len(outqueue) < sndlim) {
__skb_queue_tail(outqueue, skb);
tipc_bearer_send(net, link->bearer_id,
skb, addr);
link->next_out = NULL;
link->unacked_window = 0;
} else if (tipc_msg_bundle(outqueue, skb, mtu)) {
link->stats.sent_bundled++;
continue;
2015-02-05 21:36:36 +08:00
} else if (tipc_msg_make_bundle(outqueue, skb, mtu,
link->addr)) {
link->stats.sent_bundled++;
link->stats.sent_bundles++;
if (!link->next_out)
link->next_out = skb_peek_tail(outqueue);
} else {
__skb_queue_tail(outqueue, skb);
if (!link->next_out)
link->next_out = skb;
}
seqno++;
}
link->next_out_no = seqno;
return 0;
}
static void skb2list(struct sk_buff *skb, struct sk_buff_head *list)
{
__skb_queue_head_init(list);
__skb_queue_tail(list, skb);
}
static int __tipc_link_xmit_skb(struct tipc_link *link, struct sk_buff *skb)
{
struct sk_buff_head head;
skb2list(skb, &head);
return __tipc_link_xmit(link->owner->net, link, &head);
}
int tipc_link_xmit_skb(struct net *net, struct sk_buff *skb, u32 dnode,
u32 selector)
{
struct sk_buff_head head;
skb2list(skb, &head);
return tipc_link_xmit(net, &head, dnode, selector);
}
/**
* tipc_link_xmit() is the general link level function for message sending
* @net: the applicable net namespace
* @list: chain of buffers containing message
* @dsz: amount of user data to be sent
* @dnode: address of destination node
* @selector: a number used for deterministic link selection
* Consumes the buffer chain, except when returning -ELINKCONG
* Returns 0 if success, otherwise errno: -ELINKCONG,-EHOSTUNREACH,-EMSGSIZE
*/
int tipc_link_xmit(struct net *net, struct sk_buff_head *list, u32 dnode,
u32 selector)
{
struct tipc_link *link = NULL;
struct tipc_node *node;
int rc = -EHOSTUNREACH;
node = tipc_node_find(net, dnode);
if (node) {
tipc_node_lock(node);
link = node->active_links[selector & 1];
if (link)
rc = __tipc_link_xmit(net, link, list);
tipc_node_unlock(node);
}
if (link)
return rc;
if (likely(in_own_node(net, dnode))) {
/* As a node local message chain never contains more than one
* buffer, we just need to dequeue one SKB buffer from the
* head list.
*/
return tipc_sk_rcv(net, __skb_dequeue(list));
}
__skb_queue_purge(list);
return rc;
}
tipc: introduce message to synchronize broadcast link Upon establishing a first link between two nodes, there is currently a risk that the two endpoints will disagree on exactly which sequence number reception and acknowleding of broadcast packets should start. The following scenarios may happen: 1: Node A sends an ACTIVATE message to B, telling it to start acking packets from sequence number N. 2: Node A sends out broadcast N, but does not expect an acknowledge from B, since B is not yet in its broadcast receiver's list. 3: Node A receives ACK for N from all nodes except B, and releases packet N. 4: Node B receives the ACTIVATE, activates its link endpoint, and stores the value N as sequence number of first expected packet. 5: Node B sends a NAME_DISTR message to A. 6: Node A receives the NAME_DISTR message, and activates its endpoint. At this moment B is added to A's broadcast receiver's set. Node A also sets sequence number 0 as the first broadcast packet to be received from B. 7: Node A sends broadcast N+1. 8: B receives N+1, determines there is a gap in the sequence, since it is expecting N, and sends a NACK for N back to A. 9: Node A has already released N, so no retransmission is possible. The broadcast link in direction A->B is stale. In addition to, or instead of, 7-9 above, the following may happen: 10: Node B sends broadcast M > 0 to A. 11: Node A receives M, falsely decides there must be a gap, since it is expecting packet 0, and asks for retransmission of packets [0,M-1]. 12: Node B has already released these packets, so the broadcast link is stale in direction B->A. We solve this problem by introducing a new unicast message type, BCAST_PROTOCOL/STATE, to convey the sequence number of the next sent broadcast packet to the other endpoint, at exactly the moment that endpoint is added to the own node's broadcast receivers list, and before any other unicast messages are permitted to be sent. Furthermore, we don't allow any node to start receiving and processing broadcast packets until this new synchronization message has been received. To maintain backwards compatibility, we still open up for broadcast reception if we receive a NAME_DISTR message without any preceding broadcast sync message. In this case, we must assume that the other end has an older code version, and will never send out the new synchronization message. Hence, for mixed old and new nodes, the issue arising in 7-12 of the above may happen with the same probability as before. Signed-off-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: Ying Xue <ying.xue@windriver.com> Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
2012-11-16 13:51:31 +08:00
/*
tipc: align tipc function names with common naming practice in the network Rename the following functions, which are shorter and more in line with common naming practice in the network subsystem. tipc_bclink_send_msg->tipc_bclink_xmit tipc_bclink_recv_pkt->tipc_bclink_rcv tipc_disc_recv_msg->tipc_disc_rcv tipc_link_send_proto_msg->tipc_link_proto_xmit link_recv_proto_msg->tipc_link_proto_rcv link_send_sections_long->tipc_link_iovec_long_xmit tipc_link_send_sections_fast->tipc_link_iovec_xmit_fast tipc_link_send_sync->tipc_link_sync_xmit tipc_link_recv_sync->tipc_link_sync_rcv tipc_link_send_buf->__tipc_link_xmit tipc_link_send->tipc_link_xmit tipc_link_send_names->tipc_link_names_xmit tipc_named_recv->tipc_named_rcv tipc_link_recv_bundle->tipc_link_bundle_rcv tipc_link_dup_send_queue->tipc_link_dup_queue_xmit link_send_long_buf->tipc_link_frag_xmit tipc_multicast->tipc_port_mcast_xmit tipc_port_recv_mcast->tipc_port_mcast_rcv tipc_port_reject_sections->tipc_port_iovec_reject tipc_port_recv_proto_msg->tipc_port_proto_rcv tipc_connect->tipc_port_connect __tipc_connect->__tipc_port_connect __tipc_disconnect->__tipc_port_disconnect tipc_disconnect->tipc_port_disconnect tipc_shutdown->tipc_port_shutdown tipc_port_recv_msg->tipc_port_rcv tipc_port_recv_sections->tipc_port_iovec_rcv release->tipc_release accept->tipc_accept bind->tipc_bind get_name->tipc_getname poll->tipc_poll send_msg->tipc_sendmsg send_packet->tipc_send_packet send_stream->tipc_send_stream recv_msg->tipc_recvmsg recv_stream->tipc_recv_stream connect->tipc_connect listen->tipc_listen shutdown->tipc_shutdown setsockopt->tipc_setsockopt getsockopt->tipc_getsockopt Above changes have no impact on current users of the functions. Signed-off-by: Ying Xue <ying.xue@windriver.com> Reviewed-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-18 16:06:46 +08:00
* tipc_link_sync_xmit - synchronize broadcast link endpoints.
tipc: introduce message to synchronize broadcast link Upon establishing a first link between two nodes, there is currently a risk that the two endpoints will disagree on exactly which sequence number reception and acknowleding of broadcast packets should start. The following scenarios may happen: 1: Node A sends an ACTIVATE message to B, telling it to start acking packets from sequence number N. 2: Node A sends out broadcast N, but does not expect an acknowledge from B, since B is not yet in its broadcast receiver's list. 3: Node A receives ACK for N from all nodes except B, and releases packet N. 4: Node B receives the ACTIVATE, activates its link endpoint, and stores the value N as sequence number of first expected packet. 5: Node B sends a NAME_DISTR message to A. 6: Node A receives the NAME_DISTR message, and activates its endpoint. At this moment B is added to A's broadcast receiver's set. Node A also sets sequence number 0 as the first broadcast packet to be received from B. 7: Node A sends broadcast N+1. 8: B receives N+1, determines there is a gap in the sequence, since it is expecting N, and sends a NACK for N back to A. 9: Node A has already released N, so no retransmission is possible. The broadcast link in direction A->B is stale. In addition to, or instead of, 7-9 above, the following may happen: 10: Node B sends broadcast M > 0 to A. 11: Node A receives M, falsely decides there must be a gap, since it is expecting packet 0, and asks for retransmission of packets [0,M-1]. 12: Node B has already released these packets, so the broadcast link is stale in direction B->A. We solve this problem by introducing a new unicast message type, BCAST_PROTOCOL/STATE, to convey the sequence number of the next sent broadcast packet to the other endpoint, at exactly the moment that endpoint is added to the own node's broadcast receivers list, and before any other unicast messages are permitted to be sent. Furthermore, we don't allow any node to start receiving and processing broadcast packets until this new synchronization message has been received. To maintain backwards compatibility, we still open up for broadcast reception if we receive a NAME_DISTR message without any preceding broadcast sync message. In this case, we must assume that the other end has an older code version, and will never send out the new synchronization message. Hence, for mixed old and new nodes, the issue arising in 7-12 of the above may happen with the same probability as before. Signed-off-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: Ying Xue <ying.xue@windriver.com> Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
2012-11-16 13:51:31 +08:00
*
* Give a newly added peer node the sequence number where it should
* start receiving and acking broadcast packets.
*
* Called with node locked
*/
static void tipc_link_sync_xmit(struct tipc_link *link)
tipc: introduce message to synchronize broadcast link Upon establishing a first link between two nodes, there is currently a risk that the two endpoints will disagree on exactly which sequence number reception and acknowleding of broadcast packets should start. The following scenarios may happen: 1: Node A sends an ACTIVATE message to B, telling it to start acking packets from sequence number N. 2: Node A sends out broadcast N, but does not expect an acknowledge from B, since B is not yet in its broadcast receiver's list. 3: Node A receives ACK for N from all nodes except B, and releases packet N. 4: Node B receives the ACTIVATE, activates its link endpoint, and stores the value N as sequence number of first expected packet. 5: Node B sends a NAME_DISTR message to A. 6: Node A receives the NAME_DISTR message, and activates its endpoint. At this moment B is added to A's broadcast receiver's set. Node A also sets sequence number 0 as the first broadcast packet to be received from B. 7: Node A sends broadcast N+1. 8: B receives N+1, determines there is a gap in the sequence, since it is expecting N, and sends a NACK for N back to A. 9: Node A has already released N, so no retransmission is possible. The broadcast link in direction A->B is stale. In addition to, or instead of, 7-9 above, the following may happen: 10: Node B sends broadcast M > 0 to A. 11: Node A receives M, falsely decides there must be a gap, since it is expecting packet 0, and asks for retransmission of packets [0,M-1]. 12: Node B has already released these packets, so the broadcast link is stale in direction B->A. We solve this problem by introducing a new unicast message type, BCAST_PROTOCOL/STATE, to convey the sequence number of the next sent broadcast packet to the other endpoint, at exactly the moment that endpoint is added to the own node's broadcast receivers list, and before any other unicast messages are permitted to be sent. Furthermore, we don't allow any node to start receiving and processing broadcast packets until this new synchronization message has been received. To maintain backwards compatibility, we still open up for broadcast reception if we receive a NAME_DISTR message without any preceding broadcast sync message. In this case, we must assume that the other end has an older code version, and will never send out the new synchronization message. Hence, for mixed old and new nodes, the issue arising in 7-12 of the above may happen with the same probability as before. Signed-off-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: Ying Xue <ying.xue@windriver.com> Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
2012-11-16 13:51:31 +08:00
{
struct sk_buff *skb;
tipc: introduce message to synchronize broadcast link Upon establishing a first link between two nodes, there is currently a risk that the two endpoints will disagree on exactly which sequence number reception and acknowleding of broadcast packets should start. The following scenarios may happen: 1: Node A sends an ACTIVATE message to B, telling it to start acking packets from sequence number N. 2: Node A sends out broadcast N, but does not expect an acknowledge from B, since B is not yet in its broadcast receiver's list. 3: Node A receives ACK for N from all nodes except B, and releases packet N. 4: Node B receives the ACTIVATE, activates its link endpoint, and stores the value N as sequence number of first expected packet. 5: Node B sends a NAME_DISTR message to A. 6: Node A receives the NAME_DISTR message, and activates its endpoint. At this moment B is added to A's broadcast receiver's set. Node A also sets sequence number 0 as the first broadcast packet to be received from B. 7: Node A sends broadcast N+1. 8: B receives N+1, determines there is a gap in the sequence, since it is expecting N, and sends a NACK for N back to A. 9: Node A has already released N, so no retransmission is possible. The broadcast link in direction A->B is stale. In addition to, or instead of, 7-9 above, the following may happen: 10: Node B sends broadcast M > 0 to A. 11: Node A receives M, falsely decides there must be a gap, since it is expecting packet 0, and asks for retransmission of packets [0,M-1]. 12: Node B has already released these packets, so the broadcast link is stale in direction B->A. We solve this problem by introducing a new unicast message type, BCAST_PROTOCOL/STATE, to convey the sequence number of the next sent broadcast packet to the other endpoint, at exactly the moment that endpoint is added to the own node's broadcast receivers list, and before any other unicast messages are permitted to be sent. Furthermore, we don't allow any node to start receiving and processing broadcast packets until this new synchronization message has been received. To maintain backwards compatibility, we still open up for broadcast reception if we receive a NAME_DISTR message without any preceding broadcast sync message. In this case, we must assume that the other end has an older code version, and will never send out the new synchronization message. Hence, for mixed old and new nodes, the issue arising in 7-12 of the above may happen with the same probability as before. Signed-off-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: Ying Xue <ying.xue@windriver.com> Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
2012-11-16 13:51:31 +08:00
struct tipc_msg *msg;
skb = tipc_buf_acquire(INT_H_SIZE);
if (!skb)
tipc: introduce message to synchronize broadcast link Upon establishing a first link between two nodes, there is currently a risk that the two endpoints will disagree on exactly which sequence number reception and acknowleding of broadcast packets should start. The following scenarios may happen: 1: Node A sends an ACTIVATE message to B, telling it to start acking packets from sequence number N. 2: Node A sends out broadcast N, but does not expect an acknowledge from B, since B is not yet in its broadcast receiver's list. 3: Node A receives ACK for N from all nodes except B, and releases packet N. 4: Node B receives the ACTIVATE, activates its link endpoint, and stores the value N as sequence number of first expected packet. 5: Node B sends a NAME_DISTR message to A. 6: Node A receives the NAME_DISTR message, and activates its endpoint. At this moment B is added to A's broadcast receiver's set. Node A also sets sequence number 0 as the first broadcast packet to be received from B. 7: Node A sends broadcast N+1. 8: B receives N+1, determines there is a gap in the sequence, since it is expecting N, and sends a NACK for N back to A. 9: Node A has already released N, so no retransmission is possible. The broadcast link in direction A->B is stale. In addition to, or instead of, 7-9 above, the following may happen: 10: Node B sends broadcast M > 0 to A. 11: Node A receives M, falsely decides there must be a gap, since it is expecting packet 0, and asks for retransmission of packets [0,M-1]. 12: Node B has already released these packets, so the broadcast link is stale in direction B->A. We solve this problem by introducing a new unicast message type, BCAST_PROTOCOL/STATE, to convey the sequence number of the next sent broadcast packet to the other endpoint, at exactly the moment that endpoint is added to the own node's broadcast receivers list, and before any other unicast messages are permitted to be sent. Furthermore, we don't allow any node to start receiving and processing broadcast packets until this new synchronization message has been received. To maintain backwards compatibility, we still open up for broadcast reception if we receive a NAME_DISTR message without any preceding broadcast sync message. In this case, we must assume that the other end has an older code version, and will never send out the new synchronization message. Hence, for mixed old and new nodes, the issue arising in 7-12 of the above may happen with the same probability as before. Signed-off-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: Ying Xue <ying.xue@windriver.com> Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
2012-11-16 13:51:31 +08:00
return;
msg = buf_msg(skb);
2015-02-05 21:36:36 +08:00
tipc_msg_init(link_own_addr(link), msg, BCAST_PROTOCOL, STATE_MSG,
INT_H_SIZE, link->addr);
msg_set_last_bcast(msg, link->owner->bclink.acked);
__tipc_link_xmit_skb(link, skb);
tipc: introduce message to synchronize broadcast link Upon establishing a first link between two nodes, there is currently a risk that the two endpoints will disagree on exactly which sequence number reception and acknowleding of broadcast packets should start. The following scenarios may happen: 1: Node A sends an ACTIVATE message to B, telling it to start acking packets from sequence number N. 2: Node A sends out broadcast N, but does not expect an acknowledge from B, since B is not yet in its broadcast receiver's list. 3: Node A receives ACK for N from all nodes except B, and releases packet N. 4: Node B receives the ACTIVATE, activates its link endpoint, and stores the value N as sequence number of first expected packet. 5: Node B sends a NAME_DISTR message to A. 6: Node A receives the NAME_DISTR message, and activates its endpoint. At this moment B is added to A's broadcast receiver's set. Node A also sets sequence number 0 as the first broadcast packet to be received from B. 7: Node A sends broadcast N+1. 8: B receives N+1, determines there is a gap in the sequence, since it is expecting N, and sends a NACK for N back to A. 9: Node A has already released N, so no retransmission is possible. The broadcast link in direction A->B is stale. In addition to, or instead of, 7-9 above, the following may happen: 10: Node B sends broadcast M > 0 to A. 11: Node A receives M, falsely decides there must be a gap, since it is expecting packet 0, and asks for retransmission of packets [0,M-1]. 12: Node B has already released these packets, so the broadcast link is stale in direction B->A. We solve this problem by introducing a new unicast message type, BCAST_PROTOCOL/STATE, to convey the sequence number of the next sent broadcast packet to the other endpoint, at exactly the moment that endpoint is added to the own node's broadcast receivers list, and before any other unicast messages are permitted to be sent. Furthermore, we don't allow any node to start receiving and processing broadcast packets until this new synchronization message has been received. To maintain backwards compatibility, we still open up for broadcast reception if we receive a NAME_DISTR message without any preceding broadcast sync message. In this case, we must assume that the other end has an older code version, and will never send out the new synchronization message. Hence, for mixed old and new nodes, the issue arising in 7-12 of the above may happen with the same probability as before. Signed-off-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: Ying Xue <ying.xue@windriver.com> Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
2012-11-16 13:51:31 +08:00
}
/*
tipc: align tipc function names with common naming practice in the network Rename the following functions, which are shorter and more in line with common naming practice in the network subsystem. tipc_bclink_send_msg->tipc_bclink_xmit tipc_bclink_recv_pkt->tipc_bclink_rcv tipc_disc_recv_msg->tipc_disc_rcv tipc_link_send_proto_msg->tipc_link_proto_xmit link_recv_proto_msg->tipc_link_proto_rcv link_send_sections_long->tipc_link_iovec_long_xmit tipc_link_send_sections_fast->tipc_link_iovec_xmit_fast tipc_link_send_sync->tipc_link_sync_xmit tipc_link_recv_sync->tipc_link_sync_rcv tipc_link_send_buf->__tipc_link_xmit tipc_link_send->tipc_link_xmit tipc_link_send_names->tipc_link_names_xmit tipc_named_recv->tipc_named_rcv tipc_link_recv_bundle->tipc_link_bundle_rcv tipc_link_dup_send_queue->tipc_link_dup_queue_xmit link_send_long_buf->tipc_link_frag_xmit tipc_multicast->tipc_port_mcast_xmit tipc_port_recv_mcast->tipc_port_mcast_rcv tipc_port_reject_sections->tipc_port_iovec_reject tipc_port_recv_proto_msg->tipc_port_proto_rcv tipc_connect->tipc_port_connect __tipc_connect->__tipc_port_connect __tipc_disconnect->__tipc_port_disconnect tipc_disconnect->tipc_port_disconnect tipc_shutdown->tipc_port_shutdown tipc_port_recv_msg->tipc_port_rcv tipc_port_recv_sections->tipc_port_iovec_rcv release->tipc_release accept->tipc_accept bind->tipc_bind get_name->tipc_getname poll->tipc_poll send_msg->tipc_sendmsg send_packet->tipc_send_packet send_stream->tipc_send_stream recv_msg->tipc_recvmsg recv_stream->tipc_recv_stream connect->tipc_connect listen->tipc_listen shutdown->tipc_shutdown setsockopt->tipc_setsockopt getsockopt->tipc_getsockopt Above changes have no impact on current users of the functions. Signed-off-by: Ying Xue <ying.xue@windriver.com> Reviewed-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-18 16:06:46 +08:00
* tipc_link_sync_rcv - synchronize broadcast link endpoints.
tipc: introduce message to synchronize broadcast link Upon establishing a first link between two nodes, there is currently a risk that the two endpoints will disagree on exactly which sequence number reception and acknowleding of broadcast packets should start. The following scenarios may happen: 1: Node A sends an ACTIVATE message to B, telling it to start acking packets from sequence number N. 2: Node A sends out broadcast N, but does not expect an acknowledge from B, since B is not yet in its broadcast receiver's list. 3: Node A receives ACK for N from all nodes except B, and releases packet N. 4: Node B receives the ACTIVATE, activates its link endpoint, and stores the value N as sequence number of first expected packet. 5: Node B sends a NAME_DISTR message to A. 6: Node A receives the NAME_DISTR message, and activates its endpoint. At this moment B is added to A's broadcast receiver's set. Node A also sets sequence number 0 as the first broadcast packet to be received from B. 7: Node A sends broadcast N+1. 8: B receives N+1, determines there is a gap in the sequence, since it is expecting N, and sends a NACK for N back to A. 9: Node A has already released N, so no retransmission is possible. The broadcast link in direction A->B is stale. In addition to, or instead of, 7-9 above, the following may happen: 10: Node B sends broadcast M > 0 to A. 11: Node A receives M, falsely decides there must be a gap, since it is expecting packet 0, and asks for retransmission of packets [0,M-1]. 12: Node B has already released these packets, so the broadcast link is stale in direction B->A. We solve this problem by introducing a new unicast message type, BCAST_PROTOCOL/STATE, to convey the sequence number of the next sent broadcast packet to the other endpoint, at exactly the moment that endpoint is added to the own node's broadcast receivers list, and before any other unicast messages are permitted to be sent. Furthermore, we don't allow any node to start receiving and processing broadcast packets until this new synchronization message has been received. To maintain backwards compatibility, we still open up for broadcast reception if we receive a NAME_DISTR message without any preceding broadcast sync message. In this case, we must assume that the other end has an older code version, and will never send out the new synchronization message. Hence, for mixed old and new nodes, the issue arising in 7-12 of the above may happen with the same probability as before. Signed-off-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: Ying Xue <ying.xue@windriver.com> Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
2012-11-16 13:51:31 +08:00
* Receive the sequence number where we should start receiving and
* acking broadcast packets from a newly added peer node, and open
* up for reception of such packets.
*
* Called with node locked
*/
tipc: align tipc function names with common naming practice in the network Rename the following functions, which are shorter and more in line with common naming practice in the network subsystem. tipc_bclink_send_msg->tipc_bclink_xmit tipc_bclink_recv_pkt->tipc_bclink_rcv tipc_disc_recv_msg->tipc_disc_rcv tipc_link_send_proto_msg->tipc_link_proto_xmit link_recv_proto_msg->tipc_link_proto_rcv link_send_sections_long->tipc_link_iovec_long_xmit tipc_link_send_sections_fast->tipc_link_iovec_xmit_fast tipc_link_send_sync->tipc_link_sync_xmit tipc_link_recv_sync->tipc_link_sync_rcv tipc_link_send_buf->__tipc_link_xmit tipc_link_send->tipc_link_xmit tipc_link_send_names->tipc_link_names_xmit tipc_named_recv->tipc_named_rcv tipc_link_recv_bundle->tipc_link_bundle_rcv tipc_link_dup_send_queue->tipc_link_dup_queue_xmit link_send_long_buf->tipc_link_frag_xmit tipc_multicast->tipc_port_mcast_xmit tipc_port_recv_mcast->tipc_port_mcast_rcv tipc_port_reject_sections->tipc_port_iovec_reject tipc_port_recv_proto_msg->tipc_port_proto_rcv tipc_connect->tipc_port_connect __tipc_connect->__tipc_port_connect __tipc_disconnect->__tipc_port_disconnect tipc_disconnect->tipc_port_disconnect tipc_shutdown->tipc_port_shutdown tipc_port_recv_msg->tipc_port_rcv tipc_port_recv_sections->tipc_port_iovec_rcv release->tipc_release accept->tipc_accept bind->tipc_bind get_name->tipc_getname poll->tipc_poll send_msg->tipc_sendmsg send_packet->tipc_send_packet send_stream->tipc_send_stream recv_msg->tipc_recvmsg recv_stream->tipc_recv_stream connect->tipc_connect listen->tipc_listen shutdown->tipc_shutdown setsockopt->tipc_setsockopt getsockopt->tipc_getsockopt Above changes have no impact on current users of the functions. Signed-off-by: Ying Xue <ying.xue@windriver.com> Reviewed-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-18 16:06:46 +08:00
static void tipc_link_sync_rcv(struct tipc_node *n, struct sk_buff *buf)
tipc: introduce message to synchronize broadcast link Upon establishing a first link between two nodes, there is currently a risk that the two endpoints will disagree on exactly which sequence number reception and acknowleding of broadcast packets should start. The following scenarios may happen: 1: Node A sends an ACTIVATE message to B, telling it to start acking packets from sequence number N. 2: Node A sends out broadcast N, but does not expect an acknowledge from B, since B is not yet in its broadcast receiver's list. 3: Node A receives ACK for N from all nodes except B, and releases packet N. 4: Node B receives the ACTIVATE, activates its link endpoint, and stores the value N as sequence number of first expected packet. 5: Node B sends a NAME_DISTR message to A. 6: Node A receives the NAME_DISTR message, and activates its endpoint. At this moment B is added to A's broadcast receiver's set. Node A also sets sequence number 0 as the first broadcast packet to be received from B. 7: Node A sends broadcast N+1. 8: B receives N+1, determines there is a gap in the sequence, since it is expecting N, and sends a NACK for N back to A. 9: Node A has already released N, so no retransmission is possible. The broadcast link in direction A->B is stale. In addition to, or instead of, 7-9 above, the following may happen: 10: Node B sends broadcast M > 0 to A. 11: Node A receives M, falsely decides there must be a gap, since it is expecting packet 0, and asks for retransmission of packets [0,M-1]. 12: Node B has already released these packets, so the broadcast link is stale in direction B->A. We solve this problem by introducing a new unicast message type, BCAST_PROTOCOL/STATE, to convey the sequence number of the next sent broadcast packet to the other endpoint, at exactly the moment that endpoint is added to the own node's broadcast receivers list, and before any other unicast messages are permitted to be sent. Furthermore, we don't allow any node to start receiving and processing broadcast packets until this new synchronization message has been received. To maintain backwards compatibility, we still open up for broadcast reception if we receive a NAME_DISTR message without any preceding broadcast sync message. In this case, we must assume that the other end has an older code version, and will never send out the new synchronization message. Hence, for mixed old and new nodes, the issue arising in 7-12 of the above may happen with the same probability as before. Signed-off-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: Ying Xue <ying.xue@windriver.com> Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
2012-11-16 13:51:31 +08:00
{
struct tipc_msg *msg = buf_msg(buf);
n->bclink.last_sent = n->bclink.last_in = msg_last_bcast(msg);
n->bclink.recv_permitted = true;
kfree_skb(buf);
}
struct sk_buff *tipc_skb_queue_next(const struct sk_buff_head *list,
const struct sk_buff *skb)
{
if (skb_queue_is_last(list, skb))
return NULL;
return skb->next;
}
/*
* tipc_link_push_packets - push unsent packets to bearer
*
* Push out the unsent messages of a link where congestion
* has abated. Node is locked.
*
* Called with node locked
*/
void tipc_link_push_packets(struct tipc_link *l_ptr)
{
struct sk_buff_head *outqueue = &l_ptr->outqueue;
struct sk_buff *skb = l_ptr->next_out;
struct tipc_msg *msg;
u32 next, first;
skb_queue_walk_from(outqueue, skb) {
msg = buf_msg(skb);
next = msg_seqno(msg);
first = buf_seqno(skb_peek(outqueue));
if (mod(next - first) < l_ptr->queue_limit[0]) {
msg_set_ack(msg, mod(l_ptr->next_in_no - 1));
msg_set_bcast_ack(msg, l_ptr->owner->bclink.last_in);
if (msg_user(msg) == MSG_BUNDLER)
TIPC_SKB_CB(skb)->bundling = false;
tipc_bearer_send(l_ptr->owner->net,
l_ptr->bearer_id, skb,
&l_ptr->media_addr);
l_ptr->next_out = tipc_skb_queue_next(outqueue, skb);
} else {
break;
}
}
}
void tipc_link_reset_all(struct tipc_node *node)
{
char addr_string[16];
u32 i;
tipc_node_lock(node);
pr_warn("Resetting all links to %s\n",
tipc_addr_string_fill(addr_string, node->addr));
for (i = 0; i < MAX_BEARERS; i++) {
if (node->links[i]) {
link_print(node->links[i], "Resetting link\n");
tipc_link_reset(node->links[i]);
}
}
tipc_node_unlock(node);
}
static void link_retransmit_failure(struct tipc_link *l_ptr,
struct sk_buff *buf)
{
struct tipc_msg *msg = buf_msg(buf);
struct net *net = l_ptr->owner->net;
pr_warn("Retransmission failure on link <%s>\n", l_ptr->name);
if (l_ptr->addr) {
/* Handle failure on standard link */
link_print(l_ptr, "Resetting link\n");
tipc_link_reset(l_ptr);
} else {
/* Handle failure on broadcast link */
struct tipc_node *n_ptr;
char addr_string[16];
pr_info("Msg seq number: %u, ", msg_seqno(msg));
pr_cont("Outstanding acks: %lu\n",
(unsigned long) TIPC_SKB_CB(buf)->handle);
n_ptr = tipc_bclink_retransmit_to(net);
tipc_node_lock(n_ptr);
tipc_addr_string_fill(addr_string, n_ptr->addr);
pr_info("Broadcast link info for %s\n", addr_string);
pr_info("Reception permitted: %d, Acked: %u\n",
n_ptr->bclink.recv_permitted,
n_ptr->bclink.acked);
pr_info("Last in: %u, Oos state: %u, Last sent: %u\n",
n_ptr->bclink.last_in,
n_ptr->bclink.oos_state,
n_ptr->bclink.last_sent);
tipc_node_unlock(n_ptr);
tipc_bclink_set_flags(net, TIPC_BCLINK_RESET);
l_ptr->stale_count = 0;
}
}
void tipc_link_retransmit(struct tipc_link *l_ptr, struct sk_buff *skb,
u32 retransmits)
{
struct tipc_msg *msg;
if (!skb)
return;
msg = buf_msg(skb);
tipc: remove interface state mirroring in bearer struct 'tipc_bearer' is a generic representation of the underlying media type, and exists in a one-to-one relationship to each interface TIPC is using. The struct contains a 'blocked' flag that mirrors the operational and execution state of the represented interface, and is updated through notification calls from the latter. The users of tipc_bearer are checking this flag before each attempt to send a packet via the interface. This state mirroring serves no purpose in the current code base. TIPC links will not discover a media failure any faster through this mechanism, and in reality the flag only adds overhead at packet sending and reception. Furthermore, the fact that the flag needs to be protected by a spinlock aggregated into tipc_bearer has turned out to cause a serious and completely unnecessary deadlock problem. CPU0 CPU1 ---- ---- Time 0: bearer_disable() link_timeout() Time 1: spin_lock_bh(&b_ptr->lock) tipc_link_push_queue() Time 2: tipc_link_delete() tipc_bearer_blocked(b_ptr) Time 3: k_cancel_timer(&req->timer) spin_lock_bh(&b_ptr->lock) Time 4: del_timer_sync(&req->timer) I.e., del_timer_sync() on CPU0 never returns, because the timer handler on CPU1 is waiting for the bearer lock. We eliminate the 'blocked' flag from struct tipc_bearer, along with all tests on this flag. This not only resolves the deadlock, but also simplifies and speeds up the data path execution of TIPC. It also fits well into our ongoing effort to make the locking policy simpler and more manageable. An effect of this change is that we can get rid of functions such as tipc_bearer_blocked(), tipc_continue() and tipc_block_bearer(). We replace the latter with a new function, tipc_reset_bearer(), which resets all links associated to the bearer immediately after an interface goes down. A user might notice one slight change in link behaviour after this change. When an interface goes down, (e.g. through a NETDEV_DOWN event) all attached links will be reset immediately, instead of leaving it to each link to detect the failure through a timer-driven mechanism. We consider this an improvement, and see no obvious risks with the new behavior. Signed-off-by: Erik Hugne <erik.hugne@ericsson.com> Reviewed-by: Ying Xue <ying.xue@windriver.com> Reviewed-by: Paul Gortmaker <Paul.Gortmaker@windriver.com> Signed-off-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-06 23:08:00 +08:00
/* Detect repeated retransmit failures */
if (l_ptr->last_retransmitted == msg_seqno(msg)) {
if (++l_ptr->stale_count > 100) {
link_retransmit_failure(l_ptr, skb);
tipc: remove interface state mirroring in bearer struct 'tipc_bearer' is a generic representation of the underlying media type, and exists in a one-to-one relationship to each interface TIPC is using. The struct contains a 'blocked' flag that mirrors the operational and execution state of the represented interface, and is updated through notification calls from the latter. The users of tipc_bearer are checking this flag before each attempt to send a packet via the interface. This state mirroring serves no purpose in the current code base. TIPC links will not discover a media failure any faster through this mechanism, and in reality the flag only adds overhead at packet sending and reception. Furthermore, the fact that the flag needs to be protected by a spinlock aggregated into tipc_bearer has turned out to cause a serious and completely unnecessary deadlock problem. CPU0 CPU1 ---- ---- Time 0: bearer_disable() link_timeout() Time 1: spin_lock_bh(&b_ptr->lock) tipc_link_push_queue() Time 2: tipc_link_delete() tipc_bearer_blocked(b_ptr) Time 3: k_cancel_timer(&req->timer) spin_lock_bh(&b_ptr->lock) Time 4: del_timer_sync(&req->timer) I.e., del_timer_sync() on CPU0 never returns, because the timer handler on CPU1 is waiting for the bearer lock. We eliminate the 'blocked' flag from struct tipc_bearer, along with all tests on this flag. This not only resolves the deadlock, but also simplifies and speeds up the data path execution of TIPC. It also fits well into our ongoing effort to make the locking policy simpler and more manageable. An effect of this change is that we can get rid of functions such as tipc_bearer_blocked(), tipc_continue() and tipc_block_bearer(). We replace the latter with a new function, tipc_reset_bearer(), which resets all links associated to the bearer immediately after an interface goes down. A user might notice one slight change in link behaviour after this change. When an interface goes down, (e.g. through a NETDEV_DOWN event) all attached links will be reset immediately, instead of leaving it to each link to detect the failure through a timer-driven mechanism. We consider this an improvement, and see no obvious risks with the new behavior. Signed-off-by: Erik Hugne <erik.hugne@ericsson.com> Reviewed-by: Ying Xue <ying.xue@windriver.com> Reviewed-by: Paul Gortmaker <Paul.Gortmaker@windriver.com> Signed-off-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-06 23:08:00 +08:00
return;
}
} else {
tipc: remove interface state mirroring in bearer struct 'tipc_bearer' is a generic representation of the underlying media type, and exists in a one-to-one relationship to each interface TIPC is using. The struct contains a 'blocked' flag that mirrors the operational and execution state of the represented interface, and is updated through notification calls from the latter. The users of tipc_bearer are checking this flag before each attempt to send a packet via the interface. This state mirroring serves no purpose in the current code base. TIPC links will not discover a media failure any faster through this mechanism, and in reality the flag only adds overhead at packet sending and reception. Furthermore, the fact that the flag needs to be protected by a spinlock aggregated into tipc_bearer has turned out to cause a serious and completely unnecessary deadlock problem. CPU0 CPU1 ---- ---- Time 0: bearer_disable() link_timeout() Time 1: spin_lock_bh(&b_ptr->lock) tipc_link_push_queue() Time 2: tipc_link_delete() tipc_bearer_blocked(b_ptr) Time 3: k_cancel_timer(&req->timer) spin_lock_bh(&b_ptr->lock) Time 4: del_timer_sync(&req->timer) I.e., del_timer_sync() on CPU0 never returns, because the timer handler on CPU1 is waiting for the bearer lock. We eliminate the 'blocked' flag from struct tipc_bearer, along with all tests on this flag. This not only resolves the deadlock, but also simplifies and speeds up the data path execution of TIPC. It also fits well into our ongoing effort to make the locking policy simpler and more manageable. An effect of this change is that we can get rid of functions such as tipc_bearer_blocked(), tipc_continue() and tipc_block_bearer(). We replace the latter with a new function, tipc_reset_bearer(), which resets all links associated to the bearer immediately after an interface goes down. A user might notice one slight change in link behaviour after this change. When an interface goes down, (e.g. through a NETDEV_DOWN event) all attached links will be reset immediately, instead of leaving it to each link to detect the failure through a timer-driven mechanism. We consider this an improvement, and see no obvious risks with the new behavior. Signed-off-by: Erik Hugne <erik.hugne@ericsson.com> Reviewed-by: Ying Xue <ying.xue@windriver.com> Reviewed-by: Paul Gortmaker <Paul.Gortmaker@windriver.com> Signed-off-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-06 23:08:00 +08:00
l_ptr->last_retransmitted = msg_seqno(msg);
l_ptr->stale_count = 1;
}
skb_queue_walk_from(&l_ptr->outqueue, skb) {
if (!retransmits || skb == l_ptr->next_out)
break;
msg = buf_msg(skb);
msg_set_ack(msg, mod(l_ptr->next_in_no - 1));
msg_set_bcast_ack(msg, l_ptr->owner->bclink.last_in);
tipc_bearer_send(l_ptr->owner->net, l_ptr->bearer_id, skb,
&l_ptr->media_addr);
retransmits--;
l_ptr->stats.retransmitted++;
}
}
static void link_retrieve_defq(struct tipc_link *link,
struct sk_buff_head *list)
{
u32 seq_no;
if (skb_queue_empty(&link->deferred_queue))
return;
seq_no = buf_seqno(skb_peek(&link->deferred_queue));
if (seq_no == mod(link->next_in_no))
skb_queue_splice_tail_init(&link->deferred_queue, list);
}
/**
* link_recv_buf_validate - validate basic format of received message
*
* This routine ensures a TIPC message has an acceptable header, and at least
* as much data as the header indicates it should. The routine also ensures
* that the entire message header is stored in the main fragment of the message
* buffer, to simplify future access to message header fields.
*
* Note: Having extra info present in the message header or data areas is OK.
* TIPC will ignore the excess, under the assumption that it is optional info
* introduced by a later release of the protocol.
*/
static int link_recv_buf_validate(struct sk_buff *buf)
{
static u32 min_data_hdr_size[8] = {
SHORT_H_SIZE, MCAST_H_SIZE, NAMED_H_SIZE, BASIC_H_SIZE,
MAX_H_SIZE, MAX_H_SIZE, MAX_H_SIZE, MAX_H_SIZE
};
struct tipc_msg *msg;
u32 tipc_hdr[2];
u32 size;
u32 hdr_size;
u32 min_hdr_size;
/* If this packet comes from the defer queue, the skb has already
* been validated
*/
if (unlikely(TIPC_SKB_CB(buf)->deferred))
return 1;
if (unlikely(buf->len < MIN_H_SIZE))
return 0;
msg = skb_header_pointer(buf, 0, sizeof(tipc_hdr), tipc_hdr);
if (msg == NULL)
return 0;
if (unlikely(msg_version(msg) != TIPC_VERSION))
return 0;
size = msg_size(msg);
hdr_size = msg_hdr_sz(msg);
min_hdr_size = msg_isdata(msg) ?
min_data_hdr_size[msg_type(msg)] : INT_H_SIZE;
if (unlikely((hdr_size < min_hdr_size) ||
(size < hdr_size) ||
(buf->len < size) ||
(size - hdr_size > TIPC_MAX_USER_MSG_SIZE)))
return 0;
return pskb_may_pull(buf, hdr_size);
}
/**
* tipc_rcv - process TIPC packets/messages arriving from off-node
* @net: the applicable net namespace
* @skb: TIPC packet
* @b_ptr: pointer to bearer message arrived on
*
* Invoked with no locks held. Bearer pointer must point to a valid bearer
* structure (i.e. cannot be NULL), but bearer can be inactive.
*/
void tipc_rcv(struct net *net, struct sk_buff *skb, struct tipc_bearer *b_ptr)
{
struct tipc_net *tn = net_generic(net, tipc_net_id);
struct sk_buff_head head;
struct tipc_node *n_ptr;
struct tipc_link *l_ptr;
struct sk_buff *skb1, *tmp;
struct tipc_msg *msg;
u32 seq_no;
u32 ackd;
u32 released;
skb2list(skb, &head);
while ((skb = __skb_dequeue(&head))) {
/* Ensure message is well-formed */
if (unlikely(!link_recv_buf_validate(skb)))
goto discard;
/* Ensure message data is a single contiguous unit */
if (unlikely(skb_linearize(skb)))
goto discard;
/* Handle arrival of a non-unicast link message */
msg = buf_msg(skb);
if (unlikely(msg_non_seq(msg))) {
if (msg_user(msg) == LINK_CONFIG)
tipc_disc_rcv(net, skb, b_ptr);
else
tipc_bclink_rcv(net, skb);
continue;
}
/* Discard unicast link messages destined for another node */
if (unlikely(!msg_short(msg) &&
(msg_destnode(msg) != tn->own_addr)))
goto discard;
/* Locate neighboring node that sent message */
n_ptr = tipc_node_find(net, msg_prevnode(msg));
if (unlikely(!n_ptr))
goto discard;
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
/* Locate unicast link endpoint that should handle message */
l_ptr = n_ptr->links[b_ptr->identity];
if (unlikely(!l_ptr))
goto unlock_discard;
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
/* Verify that communication with node is currently allowed */
if ((n_ptr->action_flags & TIPC_WAIT_PEER_LINKS_DOWN) &&
msg_user(msg) == LINK_PROTOCOL &&
(msg_type(msg) == RESET_MSG ||
msg_type(msg) == ACTIVATE_MSG) &&
!msg_redundant_link(msg))
n_ptr->action_flags &= ~TIPC_WAIT_PEER_LINKS_DOWN;
if (tipc_node_blocked(n_ptr))
goto unlock_discard;
/* Validate message sequence number info */
seq_no = msg_seqno(msg);
ackd = msg_ack(msg);
/* Release acked messages */
if (n_ptr->bclink.recv_permitted)
tipc_bclink_acknowledge(n_ptr, msg_bcast_ack(msg));
released = 0;
skb_queue_walk_safe(&l_ptr->outqueue, skb1, tmp) {
if (skb1 == l_ptr->next_out ||
more(buf_seqno(skb1), ackd))
break;
__skb_unlink(skb1, &l_ptr->outqueue);
kfree_skb(skb1);
released = 1;
}
/* Try sending any messages link endpoint has pending */
if (unlikely(l_ptr->next_out))
tipc_link_push_packets(l_ptr);
if (released && !skb_queue_empty(&l_ptr->waiting_sks)) {
link_prepare_wakeup(l_ptr);
l_ptr->owner->action_flags |= TIPC_WAKEUP_USERS;
}
/* Process the incoming packet */
if (unlikely(!link_working_working(l_ptr))) {
if (msg_user(msg) == LINK_PROTOCOL) {
2015-02-05 21:36:36 +08:00
tipc_link_proto_rcv(l_ptr, skb);
link_retrieve_defq(l_ptr, &head);
tipc_node_unlock(n_ptr);
continue;
}
/* Traffic message. Conditionally activate link */
link_state_event(l_ptr, TRAFFIC_MSG_EVT);
if (link_working_working(l_ptr)) {
/* Re-insert buffer in front of queue */
__skb_queue_head(&head, skb);
tipc_node_unlock(n_ptr);
continue;
}
goto unlock_discard;
}
/* Link is now in state WORKING_WORKING */
if (unlikely(seq_no != mod(l_ptr->next_in_no))) {
2015-02-05 21:36:36 +08:00
link_handle_out_of_seq_msg(l_ptr, skb);
link_retrieve_defq(l_ptr, &head);
tipc_node_unlock(n_ptr);
continue;
}
l_ptr->next_in_no++;
if (unlikely(!skb_queue_empty(&l_ptr->deferred_queue)))
link_retrieve_defq(l_ptr, &head);
if (unlikely(++l_ptr->unacked_window >= TIPC_MIN_LINK_WIN)) {
l_ptr->stats.sent_acks++;
tipc_link_proto_xmit(l_ptr, STATE_MSG, 0, 0, 0, 0, 0);
}
if (tipc_link_prepare_input(net, l_ptr, &skb)) {
tipc_node_unlock(n_ptr);
continue;
}
tipc_node_unlock(n_ptr);
if (tipc_link_input(net, l_ptr, skb) != 0)
goto discard;
continue;
unlock_discard:
tipc_node_unlock(n_ptr);
discard:
kfree_skb(skb);
}
}
/**
* tipc_link_prepare_input - process TIPC link messages
*
* returns nonzero if the message was consumed
*
* Node lock must be held
*/
static int tipc_link_prepare_input(struct net *net, struct tipc_link *l,
struct sk_buff **buf)
{
struct tipc_node *n;
struct tipc_msg *msg;
int res = -EINVAL;
n = l->owner;
msg = buf_msg(*buf);
switch (msg_user(msg)) {
case CHANGEOVER_PROTOCOL:
2015-02-05 21:36:36 +08:00
if (tipc_link_tunnel_rcv(n, buf))
res = 0;
break;
case MSG_FRAGMENTER:
l->stats.recv_fragments++;
if (tipc_buf_append(&l->reasm_buf, buf)) {
l->stats.recv_fragmented++;
res = 0;
} else if (!l->reasm_buf) {
tipc_link_reset(l);
}
break;
case MSG_BUNDLER:
l->stats.recv_bundles++;
l->stats.recv_bundled += msg_msgcnt(msg);
res = 0;
break;
case NAME_DISTRIBUTOR:
n->bclink.recv_permitted = true;
res = 0;
break;
case BCAST_PROTOCOL:
tipc_link_sync_rcv(n, *buf);
break;
default:
res = 0;
}
return res;
}
/**
* tipc_link_input - Deliver message too higher layers
*/
static int tipc_link_input(struct net *net, struct tipc_link *l,
struct sk_buff *buf)
{
struct tipc_msg *msg = buf_msg(buf);
int res = 0;
switch (msg_user(msg)) {
case TIPC_LOW_IMPORTANCE:
case TIPC_MEDIUM_IMPORTANCE:
case TIPC_HIGH_IMPORTANCE:
case TIPC_CRITICAL_IMPORTANCE:
case CONN_MANAGER:
tipc_sk_rcv(net, buf);
break;
case NAME_DISTRIBUTOR:
tipc_named_rcv(net, buf);
break;
case MSG_BUNDLER:
tipc_link_bundle_rcv(net, buf);
break;
default:
res = -EINVAL;
}
return res;
}
/**
* tipc_link_defer_pkt - Add out-of-sequence message to deferred reception queue
*
* Returns increase in queue length (i.e. 0 or 1)
*/
u32 tipc_link_defer_pkt(struct sk_buff_head *list, struct sk_buff *skb)
{
struct sk_buff *skb1;
u32 seq_no = buf_seqno(skb);
/* Empty queue ? */
if (skb_queue_empty(list)) {
__skb_queue_tail(list, skb);
return 1;
}
/* Last ? */
if (less(buf_seqno(skb_peek_tail(list)), seq_no)) {
__skb_queue_tail(list, skb);
return 1;
}
/* Locate insertion point in queue, then insert; discard if duplicate */
skb_queue_walk(list, skb1) {
u32 curr_seqno = buf_seqno(skb1);
if (seq_no == curr_seqno) {
kfree_skb(skb);
return 0;
}
if (less(seq_no, curr_seqno))
break;
}
__skb_queue_before(list, skb1, skb);
return 1;
}
/*
* link_handle_out_of_seq_msg - handle arrival of out-of-sequence packet
*/
2015-02-05 21:36:36 +08:00
static void link_handle_out_of_seq_msg(struct tipc_link *l_ptr,
struct sk_buff *buf)
{
u32 seq_no = buf_seqno(buf);
if (likely(msg_user(buf_msg(buf)) == LINK_PROTOCOL)) {
2015-02-05 21:36:36 +08:00
tipc_link_proto_rcv(l_ptr, buf);
return;
}
/* Record OOS packet arrival (force mismatch on next timeout) */
l_ptr->checkpoint--;
/*
* Discard packet if a duplicate; otherwise add it to deferred queue
* and notify peer of gap as per protocol specification
*/
if (less(seq_no, mod(l_ptr->next_in_no))) {
l_ptr->stats.duplicates++;
kfree_skb(buf);
return;
}
if (tipc_link_defer_pkt(&l_ptr->deferred_queue, buf)) {
l_ptr->stats.deferred_recv++;
TIPC_SKB_CB(buf)->deferred = true;
if ((skb_queue_len(&l_ptr->deferred_queue) % 16) == 1)
tipc: align tipc function names with common naming practice in the network Rename the following functions, which are shorter and more in line with common naming practice in the network subsystem. tipc_bclink_send_msg->tipc_bclink_xmit tipc_bclink_recv_pkt->tipc_bclink_rcv tipc_disc_recv_msg->tipc_disc_rcv tipc_link_send_proto_msg->tipc_link_proto_xmit link_recv_proto_msg->tipc_link_proto_rcv link_send_sections_long->tipc_link_iovec_long_xmit tipc_link_send_sections_fast->tipc_link_iovec_xmit_fast tipc_link_send_sync->tipc_link_sync_xmit tipc_link_recv_sync->tipc_link_sync_rcv tipc_link_send_buf->__tipc_link_xmit tipc_link_send->tipc_link_xmit tipc_link_send_names->tipc_link_names_xmit tipc_named_recv->tipc_named_rcv tipc_link_recv_bundle->tipc_link_bundle_rcv tipc_link_dup_send_queue->tipc_link_dup_queue_xmit link_send_long_buf->tipc_link_frag_xmit tipc_multicast->tipc_port_mcast_xmit tipc_port_recv_mcast->tipc_port_mcast_rcv tipc_port_reject_sections->tipc_port_iovec_reject tipc_port_recv_proto_msg->tipc_port_proto_rcv tipc_connect->tipc_port_connect __tipc_connect->__tipc_port_connect __tipc_disconnect->__tipc_port_disconnect tipc_disconnect->tipc_port_disconnect tipc_shutdown->tipc_port_shutdown tipc_port_recv_msg->tipc_port_rcv tipc_port_recv_sections->tipc_port_iovec_rcv release->tipc_release accept->tipc_accept bind->tipc_bind get_name->tipc_getname poll->tipc_poll send_msg->tipc_sendmsg send_packet->tipc_send_packet send_stream->tipc_send_stream recv_msg->tipc_recvmsg recv_stream->tipc_recv_stream connect->tipc_connect listen->tipc_listen shutdown->tipc_shutdown setsockopt->tipc_setsockopt getsockopt->tipc_getsockopt Above changes have no impact on current users of the functions. Signed-off-by: Ying Xue <ying.xue@windriver.com> Reviewed-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-18 16:06:46 +08:00
tipc_link_proto_xmit(l_ptr, STATE_MSG, 0, 0, 0, 0, 0);
} else {
l_ptr->stats.duplicates++;
}
}
/*
* Send protocol message to the other endpoint.
*/
tipc: align tipc function names with common naming practice in the network Rename the following functions, which are shorter and more in line with common naming practice in the network subsystem. tipc_bclink_send_msg->tipc_bclink_xmit tipc_bclink_recv_pkt->tipc_bclink_rcv tipc_disc_recv_msg->tipc_disc_rcv tipc_link_send_proto_msg->tipc_link_proto_xmit link_recv_proto_msg->tipc_link_proto_rcv link_send_sections_long->tipc_link_iovec_long_xmit tipc_link_send_sections_fast->tipc_link_iovec_xmit_fast tipc_link_send_sync->tipc_link_sync_xmit tipc_link_recv_sync->tipc_link_sync_rcv tipc_link_send_buf->__tipc_link_xmit tipc_link_send->tipc_link_xmit tipc_link_send_names->tipc_link_names_xmit tipc_named_recv->tipc_named_rcv tipc_link_recv_bundle->tipc_link_bundle_rcv tipc_link_dup_send_queue->tipc_link_dup_queue_xmit link_send_long_buf->tipc_link_frag_xmit tipc_multicast->tipc_port_mcast_xmit tipc_port_recv_mcast->tipc_port_mcast_rcv tipc_port_reject_sections->tipc_port_iovec_reject tipc_port_recv_proto_msg->tipc_port_proto_rcv tipc_connect->tipc_port_connect __tipc_connect->__tipc_port_connect __tipc_disconnect->__tipc_port_disconnect tipc_disconnect->tipc_port_disconnect tipc_shutdown->tipc_port_shutdown tipc_port_recv_msg->tipc_port_rcv tipc_port_recv_sections->tipc_port_iovec_rcv release->tipc_release accept->tipc_accept bind->tipc_bind get_name->tipc_getname poll->tipc_poll send_msg->tipc_sendmsg send_packet->tipc_send_packet send_stream->tipc_send_stream recv_msg->tipc_recvmsg recv_stream->tipc_recv_stream connect->tipc_connect listen->tipc_listen shutdown->tipc_shutdown setsockopt->tipc_setsockopt getsockopt->tipc_getsockopt Above changes have no impact on current users of the functions. Signed-off-by: Ying Xue <ying.xue@windriver.com> Reviewed-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-18 16:06:46 +08:00
void tipc_link_proto_xmit(struct tipc_link *l_ptr, u32 msg_typ, int probe_msg,
u32 gap, u32 tolerance, u32 priority, u32 ack_mtu)
{
struct sk_buff *buf = NULL;
struct tipc_msg *msg = l_ptr->pmsg;
u32 msg_size = sizeof(l_ptr->proto_msg);
int r_flag;
/* Don't send protocol message during link changeover */
if (l_ptr->exp_msg_count)
return;
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
/* Abort non-RESET send if communication with node is prohibited */
if ((tipc_node_blocked(l_ptr->owner)) && (msg_typ != RESET_MSG))
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
return;
/* Create protocol message with "out-of-sequence" sequence number */
msg_set_type(msg, msg_typ);
msg_set_net_plane(msg, l_ptr->net_plane);
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
msg_set_bcast_ack(msg, l_ptr->owner->bclink.last_in);
msg_set_last_bcast(msg, tipc_bclink_get_last_sent(l_ptr->owner->net));
if (msg_typ == STATE_MSG) {
u32 next_sent = mod(l_ptr->next_out_no);
if (!tipc_link_is_up(l_ptr))
return;
if (l_ptr->next_out)
next_sent = buf_seqno(l_ptr->next_out);
msg_set_next_sent(msg, next_sent);
if (!skb_queue_empty(&l_ptr->deferred_queue)) {
u32 rec = buf_seqno(skb_peek(&l_ptr->deferred_queue));
gap = mod(rec - mod(l_ptr->next_in_no));
}
msg_set_seq_gap(msg, gap);
if (gap)
l_ptr->stats.sent_nacks++;
msg_set_link_tolerance(msg, tolerance);
msg_set_linkprio(msg, priority);
msg_set_max_pkt(msg, ack_mtu);
msg_set_ack(msg, mod(l_ptr->next_in_no - 1));
msg_set_probe(msg, probe_msg != 0);
if (probe_msg) {
u32 mtu = l_ptr->max_pkt;
if ((mtu < l_ptr->max_pkt_target) &&
link_working_working(l_ptr) &&
l_ptr->fsm_msg_cnt) {
msg_size = (mtu + (l_ptr->max_pkt_target - mtu)/2 + 2) & ~3;
if (l_ptr->max_pkt_probes == 10) {
l_ptr->max_pkt_target = (msg_size - 4);
l_ptr->max_pkt_probes = 0;
msg_size = (mtu + (l_ptr->max_pkt_target - mtu)/2 + 2) & ~3;
}
l_ptr->max_pkt_probes++;
}
l_ptr->stats.sent_probes++;
}
l_ptr->stats.sent_states++;
} else { /* RESET_MSG or ACTIVATE_MSG */
msg_set_ack(msg, mod(l_ptr->reset_checkpoint - 1));
msg_set_seq_gap(msg, 0);
msg_set_next_sent(msg, 1);
msg_set_probe(msg, 0);
msg_set_link_tolerance(msg, l_ptr->tolerance);
msg_set_linkprio(msg, l_ptr->priority);
msg_set_max_pkt(msg, l_ptr->max_pkt_target);
}
r_flag = (l_ptr->owner->working_links > tipc_link_is_up(l_ptr));
msg_set_redundant_link(msg, r_flag);
msg_set_linkprio(msg, l_ptr->priority);
msg_set_size(msg, msg_size);
msg_set_seqno(msg, mod(l_ptr->next_out_no + (0xffff/2)));
buf = tipc_buf_acquire(msg_size);
if (!buf)
return;
skb_copy_to_linear_data(buf, msg, sizeof(l_ptr->proto_msg));
buf->priority = TC_PRIO_CONTROL;
tipc_bearer_send(l_ptr->owner->net, l_ptr->bearer_id, buf,
&l_ptr->media_addr);
l_ptr->unacked_window = 0;
kfree_skb(buf);
}
/*
* Receive protocol message :
* Note that network plane id propagates through the network, and may
* change at any time. The node with lowest address rules
*/
2015-02-05 21:36:36 +08:00
static void tipc_link_proto_rcv(struct tipc_link *l_ptr,
struct sk_buff *buf)
{
u32 rec_gap = 0;
u32 max_pkt_info;
u32 max_pkt_ack;
u32 msg_tol;
struct tipc_msg *msg = buf_msg(buf);
/* Discard protocol message during link changeover */
if (l_ptr->exp_msg_count)
goto exit;
if (l_ptr->net_plane != msg_net_plane(msg))
2015-02-05 21:36:36 +08:00
if (link_own_addr(l_ptr) > msg_prevnode(msg))
l_ptr->net_plane = msg_net_plane(msg);
switch (msg_type(msg)) {
case RESET_MSG:
if (!link_working_unknown(l_ptr) &&
(l_ptr->peer_session != INVALID_SESSION)) {
if (less_eq(msg_session(msg), l_ptr->peer_session))
break; /* duplicate or old reset: ignore */
}
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
if (!msg_redundant_link(msg) && (link_working_working(l_ptr) ||
link_working_unknown(l_ptr))) {
/*
* peer has lost contact -- don't allow peer's links
* to reactivate before we recognize loss & clean up
*/
l_ptr->owner->action_flags |= TIPC_WAIT_OWN_LINKS_DOWN;
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
}
link_state_event(l_ptr, RESET_MSG);
/* fall thru' */
case ACTIVATE_MSG:
/* Update link settings according other endpoint's values */
strcpy((strrchr(l_ptr->name, ':') + 1), (char *)msg_data(msg));
msg_tol = msg_link_tolerance(msg);
if (msg_tol > l_ptr->tolerance)
link_set_supervision_props(l_ptr, msg_tol);
if (msg_linkprio(msg) > l_ptr->priority)
l_ptr->priority = msg_linkprio(msg);
max_pkt_info = msg_max_pkt(msg);
if (max_pkt_info) {
if (max_pkt_info < l_ptr->max_pkt_target)
l_ptr->max_pkt_target = max_pkt_info;
if (l_ptr->max_pkt > l_ptr->max_pkt_target)
l_ptr->max_pkt = l_ptr->max_pkt_target;
} else {
l_ptr->max_pkt = l_ptr->max_pkt_target;
}
/* Synchronize broadcast link info, if not done previously */
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
if (!tipc_node_is_up(l_ptr->owner)) {
l_ptr->owner->bclink.last_sent =
l_ptr->owner->bclink.last_in =
msg_last_bcast(msg);
l_ptr->owner->bclink.oos_state = 0;
}
l_ptr->peer_session = msg_session(msg);
l_ptr->peer_bearer_id = msg_bearer_id(msg);
if (msg_type(msg) == ACTIVATE_MSG)
link_state_event(l_ptr, ACTIVATE_MSG);
break;
case STATE_MSG:
msg_tol = msg_link_tolerance(msg);
if (msg_tol)
link_set_supervision_props(l_ptr, msg_tol);
if (msg_linkprio(msg) &&
(msg_linkprio(msg) != l_ptr->priority)) {
pr_debug("%s<%s>, priority change %u->%u\n",
link_rst_msg, l_ptr->name,
l_ptr->priority, msg_linkprio(msg));
l_ptr->priority = msg_linkprio(msg);
tipc_link_reset(l_ptr); /* Enforce change to take effect */
break;
}
tipc: don't record link RESET or ACTIVATE messages as traffic In the current code, all incoming LINK_PROTOCOL messages, irrespective of type, nudge the "last message received" checkpoint, informing the link state machine that a message was received from the peer since last supervision timeout event. This inhibits the link from starting probing the peer unnecessarily. However, not only STATE messages are recorded as legitimate incoming traffic this way, but even RESET and ACTIVATE messages, which in reality are there to inform the link that the peer endpoint has been reset. At the same time, some RESET messages may be dropped instead of causing a link reset. This happens when the link endpoint thinks it is fully up and working, and the session number of the RESET is lower than or equal to the current link session. In such cases the RESET is perceived as a delayed remnant from an earlier session, or the current one, and dropped. Now, if a TIPC module is removed and then immediately reinserted, e.g. when using a script, RESET messages may arrive at the peer link endpoint before this one has had time to discover the failure. The RESET may be dropped because of the session number, but only after it has been recorded as a legitimate traffic event. Hence, the receiving link will not start probing, and not discover that the peer endpoint is down, at the same time ignoring the periodic RESET messages coming from that endpoint. We have ended up in a stale state where a failed link cannot be re-established. In this commit, we remedy this by nudging the checkpoint only for received STATE messages, not for RESET or ACTIVATE messages. Signed-off-by: Jon Maloy <jon.maloy@ericsson.com> Reviewed-by: Ying Xue <ying.xue@windriver.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-05-14 17:39:10 +08:00
/* Record reception; force mismatch at next timeout: */
l_ptr->checkpoint--;
link_state_event(l_ptr, TRAFFIC_MSG_EVT);
l_ptr->stats.recv_states++;
if (link_reset_unknown(l_ptr))
break;
if (less_eq(mod(l_ptr->next_in_no), msg_next_sent(msg))) {
rec_gap = mod(msg_next_sent(msg) -
mod(l_ptr->next_in_no));
}
max_pkt_ack = msg_max_pkt(msg);
if (max_pkt_ack > l_ptr->max_pkt) {
l_ptr->max_pkt = max_pkt_ack;
l_ptr->max_pkt_probes = 0;
}
max_pkt_ack = 0;
if (msg_probe(msg)) {
l_ptr->stats.recv_probes++;
if (msg_size(msg) > sizeof(l_ptr->proto_msg))
max_pkt_ack = msg_size(msg);
}
/* Protocol message before retransmits, reduce loss risk */
if (l_ptr->owner->bclink.recv_permitted)
2015-02-05 21:36:36 +08:00
tipc_bclink_update_link_state(l_ptr->owner,
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
msg_last_bcast(msg));
if (rec_gap || (msg_probe(msg))) {
tipc: align tipc function names with common naming practice in the network Rename the following functions, which are shorter and more in line with common naming practice in the network subsystem. tipc_bclink_send_msg->tipc_bclink_xmit tipc_bclink_recv_pkt->tipc_bclink_rcv tipc_disc_recv_msg->tipc_disc_rcv tipc_link_send_proto_msg->tipc_link_proto_xmit link_recv_proto_msg->tipc_link_proto_rcv link_send_sections_long->tipc_link_iovec_long_xmit tipc_link_send_sections_fast->tipc_link_iovec_xmit_fast tipc_link_send_sync->tipc_link_sync_xmit tipc_link_recv_sync->tipc_link_sync_rcv tipc_link_send_buf->__tipc_link_xmit tipc_link_send->tipc_link_xmit tipc_link_send_names->tipc_link_names_xmit tipc_named_recv->tipc_named_rcv tipc_link_recv_bundle->tipc_link_bundle_rcv tipc_link_dup_send_queue->tipc_link_dup_queue_xmit link_send_long_buf->tipc_link_frag_xmit tipc_multicast->tipc_port_mcast_xmit tipc_port_recv_mcast->tipc_port_mcast_rcv tipc_port_reject_sections->tipc_port_iovec_reject tipc_port_recv_proto_msg->tipc_port_proto_rcv tipc_connect->tipc_port_connect __tipc_connect->__tipc_port_connect __tipc_disconnect->__tipc_port_disconnect tipc_disconnect->tipc_port_disconnect tipc_shutdown->tipc_port_shutdown tipc_port_recv_msg->tipc_port_rcv tipc_port_recv_sections->tipc_port_iovec_rcv release->tipc_release accept->tipc_accept bind->tipc_bind get_name->tipc_getname poll->tipc_poll send_msg->tipc_sendmsg send_packet->tipc_send_packet send_stream->tipc_send_stream recv_msg->tipc_recvmsg recv_stream->tipc_recv_stream connect->tipc_connect listen->tipc_listen shutdown->tipc_shutdown setsockopt->tipc_setsockopt getsockopt->tipc_getsockopt Above changes have no impact on current users of the functions. Signed-off-by: Ying Xue <ying.xue@windriver.com> Reviewed-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-18 16:06:46 +08:00
tipc_link_proto_xmit(l_ptr, STATE_MSG, 0, rec_gap, 0,
0, max_pkt_ack);
}
if (msg_seq_gap(msg)) {
l_ptr->stats.recv_nacks++;
tipc_link_retransmit(l_ptr, skb_peek(&l_ptr->outqueue),
msg_seq_gap(msg));
}
break;
}
exit:
kfree_skb(buf);
}
/* tipc_link_tunnel_xmit(): Tunnel one packet via a link belonging to
* a different bearer. Owner node is locked.
*/
static void tipc_link_tunnel_xmit(struct tipc_link *l_ptr,
struct tipc_msg *tunnel_hdr,
struct tipc_msg *msg,
u32 selector)
{
struct tipc_link *tunnel;
struct sk_buff *skb;
u32 length = msg_size(msg);
tunnel = l_ptr->owner->active_links[selector & 1];
if (!tipc_link_is_up(tunnel)) {
pr_warn("%stunnel link no longer available\n", link_co_err);
return;
}
msg_set_size(tunnel_hdr, length + INT_H_SIZE);
skb = tipc_buf_acquire(length + INT_H_SIZE);
if (!skb) {
pr_warn("%sunable to send tunnel msg\n", link_co_err);
return;
}
skb_copy_to_linear_data(skb, tunnel_hdr, INT_H_SIZE);
skb_copy_to_linear_data_offset(skb, INT_H_SIZE, msg, length);
__tipc_link_xmit_skb(tunnel, skb);
}
/* tipc_link_failover_send_queue(): A link has gone down, but a second
* link is still active. We can do failover. Tunnel the failing link's
* whole send queue via the remaining link. This way, we don't lose
* any packets, and sequence order is preserved for subsequent traffic
* sent over the remaining link. Owner node is locked.
*/
void tipc_link_failover_send_queue(struct tipc_link *l_ptr)
{
u32 msgcount = skb_queue_len(&l_ptr->outqueue);
struct tipc_link *tunnel = l_ptr->owner->active_links[0];
struct tipc_msg tunnel_hdr;
struct sk_buff *skb;
int split_bundles;
if (!tunnel)
return;
2015-02-05 21:36:36 +08:00
tipc_msg_init(link_own_addr(l_ptr), &tunnel_hdr, CHANGEOVER_PROTOCOL,
ORIGINAL_MSG, INT_H_SIZE, l_ptr->addr);
msg_set_bearer_id(&tunnel_hdr, l_ptr->peer_bearer_id);
msg_set_msgcnt(&tunnel_hdr, msgcount);
if (skb_queue_empty(&l_ptr->outqueue)) {
skb = tipc_buf_acquire(INT_H_SIZE);
if (skb) {
skb_copy_to_linear_data(skb, &tunnel_hdr, INT_H_SIZE);
msg_set_size(&tunnel_hdr, INT_H_SIZE);
__tipc_link_xmit_skb(tunnel, skb);
} else {
pr_warn("%sunable to send changeover msg\n",
link_co_err);
}
return;
}
split_bundles = (l_ptr->owner->active_links[0] !=
l_ptr->owner->active_links[1]);
skb_queue_walk(&l_ptr->outqueue, skb) {
struct tipc_msg *msg = buf_msg(skb);
if ((msg_user(msg) == MSG_BUNDLER) && split_bundles) {
struct tipc_msg *m = msg_get_wrapped(msg);
unchar *pos = (unchar *)m;
msgcount = msg_msgcnt(msg);
while (msgcount--) {
msg_set_seqno(m, msg_seqno(msg));
tipc_link_tunnel_xmit(l_ptr, &tunnel_hdr, m,
msg_link_selector(m));
pos += align(msg_size(m));
m = (struct tipc_msg *)pos;
}
} else {
tipc_link_tunnel_xmit(l_ptr, &tunnel_hdr, msg,
msg_link_selector(msg));
}
}
}
tipc: align tipc function names with common naming practice in the network Rename the following functions, which are shorter and more in line with common naming practice in the network subsystem. tipc_bclink_send_msg->tipc_bclink_xmit tipc_bclink_recv_pkt->tipc_bclink_rcv tipc_disc_recv_msg->tipc_disc_rcv tipc_link_send_proto_msg->tipc_link_proto_xmit link_recv_proto_msg->tipc_link_proto_rcv link_send_sections_long->tipc_link_iovec_long_xmit tipc_link_send_sections_fast->tipc_link_iovec_xmit_fast tipc_link_send_sync->tipc_link_sync_xmit tipc_link_recv_sync->tipc_link_sync_rcv tipc_link_send_buf->__tipc_link_xmit tipc_link_send->tipc_link_xmit tipc_link_send_names->tipc_link_names_xmit tipc_named_recv->tipc_named_rcv tipc_link_recv_bundle->tipc_link_bundle_rcv tipc_link_dup_send_queue->tipc_link_dup_queue_xmit link_send_long_buf->tipc_link_frag_xmit tipc_multicast->tipc_port_mcast_xmit tipc_port_recv_mcast->tipc_port_mcast_rcv tipc_port_reject_sections->tipc_port_iovec_reject tipc_port_recv_proto_msg->tipc_port_proto_rcv tipc_connect->tipc_port_connect __tipc_connect->__tipc_port_connect __tipc_disconnect->__tipc_port_disconnect tipc_disconnect->tipc_port_disconnect tipc_shutdown->tipc_port_shutdown tipc_port_recv_msg->tipc_port_rcv tipc_port_recv_sections->tipc_port_iovec_rcv release->tipc_release accept->tipc_accept bind->tipc_bind get_name->tipc_getname poll->tipc_poll send_msg->tipc_sendmsg send_packet->tipc_send_packet send_stream->tipc_send_stream recv_msg->tipc_recvmsg recv_stream->tipc_recv_stream connect->tipc_connect listen->tipc_listen shutdown->tipc_shutdown setsockopt->tipc_setsockopt getsockopt->tipc_getsockopt Above changes have no impact on current users of the functions. Signed-off-by: Ying Xue <ying.xue@windriver.com> Reviewed-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-18 16:06:46 +08:00
/* tipc_link_dup_queue_xmit(): A second link has become active. Tunnel a
* duplicate of the first link's send queue via the new link. This way, we
* are guaranteed that currently queued packets from a socket are delivered
* before future traffic from the same socket, even if this is using the
* new link. The last arriving copy of each duplicate packet is dropped at
* the receiving end by the regular protocol check, so packet cardinality
* and sequence order is preserved per sender/receiver socket pair.
* Owner node is locked.
*/
tipc: align tipc function names with common naming practice in the network Rename the following functions, which are shorter and more in line with common naming practice in the network subsystem. tipc_bclink_send_msg->tipc_bclink_xmit tipc_bclink_recv_pkt->tipc_bclink_rcv tipc_disc_recv_msg->tipc_disc_rcv tipc_link_send_proto_msg->tipc_link_proto_xmit link_recv_proto_msg->tipc_link_proto_rcv link_send_sections_long->tipc_link_iovec_long_xmit tipc_link_send_sections_fast->tipc_link_iovec_xmit_fast tipc_link_send_sync->tipc_link_sync_xmit tipc_link_recv_sync->tipc_link_sync_rcv tipc_link_send_buf->__tipc_link_xmit tipc_link_send->tipc_link_xmit tipc_link_send_names->tipc_link_names_xmit tipc_named_recv->tipc_named_rcv tipc_link_recv_bundle->tipc_link_bundle_rcv tipc_link_dup_send_queue->tipc_link_dup_queue_xmit link_send_long_buf->tipc_link_frag_xmit tipc_multicast->tipc_port_mcast_xmit tipc_port_recv_mcast->tipc_port_mcast_rcv tipc_port_reject_sections->tipc_port_iovec_reject tipc_port_recv_proto_msg->tipc_port_proto_rcv tipc_connect->tipc_port_connect __tipc_connect->__tipc_port_connect __tipc_disconnect->__tipc_port_disconnect tipc_disconnect->tipc_port_disconnect tipc_shutdown->tipc_port_shutdown tipc_port_recv_msg->tipc_port_rcv tipc_port_recv_sections->tipc_port_iovec_rcv release->tipc_release accept->tipc_accept bind->tipc_bind get_name->tipc_getname poll->tipc_poll send_msg->tipc_sendmsg send_packet->tipc_send_packet send_stream->tipc_send_stream recv_msg->tipc_recvmsg recv_stream->tipc_recv_stream connect->tipc_connect listen->tipc_listen shutdown->tipc_shutdown setsockopt->tipc_setsockopt getsockopt->tipc_getsockopt Above changes have no impact on current users of the functions. Signed-off-by: Ying Xue <ying.xue@windriver.com> Reviewed-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-18 16:06:46 +08:00
void tipc_link_dup_queue_xmit(struct tipc_link *l_ptr,
struct tipc_link *tunnel)
{
struct sk_buff *skb;
struct tipc_msg tunnel_hdr;
2015-02-05 21:36:36 +08:00
tipc_msg_init(link_own_addr(l_ptr), &tunnel_hdr, CHANGEOVER_PROTOCOL,
DUPLICATE_MSG, INT_H_SIZE, l_ptr->addr);
msg_set_msgcnt(&tunnel_hdr, skb_queue_len(&l_ptr->outqueue));
msg_set_bearer_id(&tunnel_hdr, l_ptr->peer_bearer_id);
skb_queue_walk(&l_ptr->outqueue, skb) {
struct sk_buff *outskb;
struct tipc_msg *msg = buf_msg(skb);
u32 length = msg_size(msg);
if (msg_user(msg) == MSG_BUNDLER)
msg_set_type(msg, CLOSED_MSG);
msg_set_ack(msg, mod(l_ptr->next_in_no - 1)); /* Update */
msg_set_bcast_ack(msg, l_ptr->owner->bclink.last_in);
msg_set_size(&tunnel_hdr, length + INT_H_SIZE);
outskb = tipc_buf_acquire(length + INT_H_SIZE);
if (outskb == NULL) {
pr_warn("%sunable to send duplicate msg\n",
link_co_err);
return;
}
skb_copy_to_linear_data(outskb, &tunnel_hdr, INT_H_SIZE);
skb_copy_to_linear_data_offset(outskb, INT_H_SIZE, skb->data,
length);
__tipc_link_xmit_skb(tunnel, outskb);
if (!tipc_link_is_up(l_ptr))
return;
}
}
/**
* buf_extract - extracts embedded TIPC message from another message
* @skb: encapsulating message buffer
* @from_pos: offset to extract from
*
* Returns a new message buffer containing an embedded message. The
* encapsulating message itself is left unchanged.
*/
static struct sk_buff *buf_extract(struct sk_buff *skb, u32 from_pos)
{
struct tipc_msg *msg = (struct tipc_msg *)(skb->data + from_pos);
u32 size = msg_size(msg);
struct sk_buff *eb;
eb = tipc_buf_acquire(size);
if (eb)
skb_copy_to_linear_data(eb, msg, size);
return eb;
}
/* tipc_link_dup_rcv(): Receive a tunnelled DUPLICATE_MSG packet.
* Owner node is locked.
*/
2015-02-05 21:36:36 +08:00
static void tipc_link_dup_rcv(struct tipc_link *l_ptr,
struct sk_buff *t_buf)
{
struct sk_buff *buf;
if (!tipc_link_is_up(l_ptr))
return;
buf = buf_extract(t_buf, INT_H_SIZE);
if (buf == NULL) {
pr_warn("%sfailed to extract inner dup pkt\n", link_co_err);
return;
}
/* Add buffer to deferred queue, if applicable: */
2015-02-05 21:36:36 +08:00
link_handle_out_of_seq_msg(l_ptr, buf);
}
2014-02-14 06:29:11 +08:00
/* tipc_link_failover_rcv(): Receive a tunnelled ORIGINAL_MSG packet
* Owner node is locked.
*/
static struct sk_buff *tipc_link_failover_rcv(struct tipc_link *l_ptr,
struct sk_buff *t_buf)
{
struct tipc_msg *t_msg = buf_msg(t_buf);
struct sk_buff *buf = NULL;
struct tipc_msg *msg;
if (tipc_link_is_up(l_ptr))
tipc_link_reset(l_ptr);
/* First failover packet? */
if (l_ptr->exp_msg_count == START_CHANGEOVER)
l_ptr->exp_msg_count = msg_msgcnt(t_msg);
/* Should there be an inner packet? */
if (l_ptr->exp_msg_count) {
l_ptr->exp_msg_count--;
buf = buf_extract(t_buf, INT_H_SIZE);
if (buf == NULL) {
pr_warn("%sno inner failover pkt\n", link_co_err);
goto exit;
}
msg = buf_msg(buf);
if (less(msg_seqno(msg), l_ptr->reset_checkpoint)) {
kfree_skb(buf);
buf = NULL;
goto exit;
}
if (msg_user(msg) == MSG_FRAGMENTER) {
l_ptr->stats.recv_fragments++;
tipc_buf_append(&l_ptr->reasm_buf, &buf);
2014-02-14 06:29:11 +08:00
}
}
exit:
if ((!l_ptr->exp_msg_count) && (l_ptr->flags & LINK_STOPPED))
tipc_link_delete(l_ptr);
2014-02-14 06:29:11 +08:00
return buf;
}
/* tipc_link_tunnel_rcv(): Receive a tunnelled packet, sent
* via other link as result of a failover (ORIGINAL_MSG) or
* a new active link (DUPLICATE_MSG). Failover packets are
* returned to the active link for delivery upwards.
* Owner node is locked.
*/
2015-02-05 21:36:36 +08:00
static int tipc_link_tunnel_rcv(struct tipc_node *n_ptr,
struct sk_buff **buf)
{
struct sk_buff *t_buf = *buf;
struct tipc_link *l_ptr;
struct tipc_msg *t_msg = buf_msg(t_buf);
u32 bearer_id = msg_bearer_id(t_msg);
*buf = NULL;
if (bearer_id >= MAX_BEARERS)
goto exit;
l_ptr = n_ptr->links[bearer_id];
if (!l_ptr)
goto exit;
if (msg_type(t_msg) == DUPLICATE_MSG)
2015-02-05 21:36:36 +08:00
tipc_link_dup_rcv(l_ptr, t_buf);
else if (msg_type(t_msg) == ORIGINAL_MSG)
*buf = tipc_link_failover_rcv(l_ptr, t_buf);
else
pr_warn("%sunknown tunnel pkt received\n", link_co_err);
exit:
kfree_skb(t_buf);
return *buf != NULL;
}
/*
* Bundler functionality:
*/
void tipc_link_bundle_rcv(struct net *net, struct sk_buff *buf)
{
u32 msgcount = msg_msgcnt(buf_msg(buf));
u32 pos = INT_H_SIZE;
struct sk_buff *obuf;
struct tipc_msg *omsg;
while (msgcount--) {
obuf = buf_extract(buf, pos);
if (obuf == NULL) {
pr_warn("Link unable to unbundle message(s)\n");
break;
}
omsg = buf_msg(obuf);
pos += align(msg_size(omsg));
if (msg_isdata(omsg)) {
if (unlikely(msg_type(omsg) == TIPC_MCAST_MSG))
tipc_sk_mcast_rcv(net, obuf);
else
tipc_sk_rcv(net, obuf);
} else if (msg_user(omsg) == CONN_MANAGER) {
tipc_sk_rcv(net, obuf);
} else if (msg_user(omsg) == NAME_DISTRIBUTOR) {
tipc_named_rcv(net, obuf);
} else {
pr_warn("Illegal bundled msg: %u\n", msg_user(omsg));
kfree_skb(obuf);
}
}
kfree_skb(buf);
}
static void link_set_supervision_props(struct tipc_link *l_ptr, u32 tol)
{
unsigned long intv = ((tol / 4) > 500) ? 500 : tol / 4;
if ((tol < TIPC_MIN_LINK_TOL) || (tol > TIPC_MAX_LINK_TOL))
return;
l_ptr->tolerance = tol;
l_ptr->cont_intv = msecs_to_jiffies(intv);
l_ptr->abort_limit = tol / (jiffies_to_msecs(l_ptr->cont_intv) / 4);
}
void tipc_link_set_queue_limits(struct tipc_link *l_ptr, u32 window)
{
/* Data messages from this node, inclusive FIRST_FRAGM */
l_ptr->queue_limit[TIPC_LOW_IMPORTANCE] = window;
l_ptr->queue_limit[TIPC_MEDIUM_IMPORTANCE] = (window / 3) * 4;
l_ptr->queue_limit[TIPC_HIGH_IMPORTANCE] = (window / 3) * 5;
l_ptr->queue_limit[TIPC_CRITICAL_IMPORTANCE] = (window / 3) * 6;
/* Transiting data messages,inclusive FIRST_FRAGM */
l_ptr->queue_limit[TIPC_LOW_IMPORTANCE + 4] = 300;
l_ptr->queue_limit[TIPC_MEDIUM_IMPORTANCE + 4] = 600;
l_ptr->queue_limit[TIPC_HIGH_IMPORTANCE + 4] = 900;
l_ptr->queue_limit[TIPC_CRITICAL_IMPORTANCE + 4] = 1200;
l_ptr->queue_limit[CONN_MANAGER] = 1200;
l_ptr->queue_limit[CHANGEOVER_PROTOCOL] = 2500;
l_ptr->queue_limit[NAME_DISTRIBUTOR] = 3000;
/* FRAGMENT and LAST_FRAGMENT packets */
l_ptr->queue_limit[MSG_FRAGMENTER] = 4000;
}
/* tipc_link_find_owner - locate owner node of link by link's name
* @net: the applicable net namespace
* @name: pointer to link name string
* @bearer_id: pointer to index in 'node->links' array where the link was found.
*
* Returns pointer to node owning the link, or 0 if no matching link is found.
*/
static struct tipc_node *tipc_link_find_owner(struct net *net,
const char *link_name,
unsigned int *bearer_id)
{
struct tipc_net *tn = net_generic(net, tipc_net_id);
struct tipc_link *l_ptr;
struct tipc_node *n_ptr;
struct tipc_node *found_node = NULL;
int i;
*bearer_id = 0;
rcu_read_lock();
list_for_each_entry_rcu(n_ptr, &tn->node_list, list) {
tipc_node_lock(n_ptr);
for (i = 0; i < MAX_BEARERS; i++) {
l_ptr = n_ptr->links[i];
if (l_ptr && !strcmp(l_ptr->name, link_name)) {
*bearer_id = i;
found_node = n_ptr;
break;
}
}
tipc_node_unlock(n_ptr);
if (found_node)
break;
}
rcu_read_unlock();
return found_node;
}
/**
* link_value_is_valid -- validate proposed link tolerance/priority/window
*
* @cmd: value type (TIPC_CMD_SET_LINK_*)
* @new_value: the new value
*
* Returns 1 if value is within range, 0 if not.
*/
static int link_value_is_valid(u16 cmd, u32 new_value)
{
switch (cmd) {
case TIPC_CMD_SET_LINK_TOL:
return (new_value >= TIPC_MIN_LINK_TOL) &&
(new_value <= TIPC_MAX_LINK_TOL);
case TIPC_CMD_SET_LINK_PRI:
return (new_value <= TIPC_MAX_LINK_PRI);
case TIPC_CMD_SET_LINK_WINDOW:
return (new_value >= TIPC_MIN_LINK_WIN) &&
(new_value <= TIPC_MAX_LINK_WIN);
}
return 0;
}
/**
* link_cmd_set_value - change priority/tolerance/window for link/bearer/media
* @net: the applicable net namespace
* @name: ptr to link, bearer, or media name
* @new_value: new value of link, bearer, or media setting
* @cmd: which link, bearer, or media attribute to set (TIPC_CMD_SET_LINK_*)
*
tipc: purge tipc_net_lock lock Now tipc routing hierarchy comprises the structures 'node', 'link'and 'bearer'. The whole hierarchy is protected by a big read/write lock, tipc_net_lock, to ensure that nothing is added or removed while code is accessing any of these structures. Obviously the locking policy makes node, link and bearer components closely bound together so that their relationship becomes unnecessarily complex. In the worst case, such locking policy not only has a negative influence on performance, but also it's prone to lead to deadlock occasionally. In order o decouple the complex relationship between bearer and node as well as link, the locking policy is adjusted as follows: - Bearer level RTNL lock is used on update side, and RCU is used on read side. Meanwhile, all bearer instances including broadcast bearer are saved into bearer_list array. - Node and link level All node instances are saved into two tipc_node_list and node_htable lists. The two lists are protected by node_list_lock on write side, and they are guarded with RCU lock on read side. All members in node structure including link instances are protected by node spin lock. - The relationship between bearer and node When link accesses bearer, it first needs to find the bearer with its bearer identity from the bearer_list array. When bearer accesses node, it can iterate the node_htable hash list with the node address to find the corresponding node. In the new locking policy, every component has its private locking solution and the relationship between bearer and node is very simple, that is, they can find each other with node address or bearer identity from node_htable hash list or bearer_list array. Until now above all changes have been done, so tipc_net_lock can be removed safely. Signed-off-by: Ying Xue <ying.xue@windriver.com> Reviewed-by: Jon Maloy <jon.maloy@ericsson.com> Reviewed-by: Erik Hugne <erik.hugne@ericsson.com> Tested-by: Erik Hugne <erik.hugne@ericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-04-21 10:55:48 +08:00
* Caller must hold RTNL lock to ensure link/bearer/media is not deleted.
*
* Returns 0 if value updated and negative value on error.
*/
static int link_cmd_set_value(struct net *net, const char *name, u32 new_value,
u16 cmd)
{
struct tipc_node *node;
struct tipc_link *l_ptr;
struct tipc_bearer *b_ptr;
struct tipc_media *m_ptr;
int bearer_id;
int res = 0;
node = tipc_link_find_owner(net, name, &bearer_id);
if (node) {
tipc_node_lock(node);
l_ptr = node->links[bearer_id];
if (l_ptr) {
switch (cmd) {
case TIPC_CMD_SET_LINK_TOL:
link_set_supervision_props(l_ptr, new_value);
tipc: align tipc function names with common naming practice in the network Rename the following functions, which are shorter and more in line with common naming practice in the network subsystem. tipc_bclink_send_msg->tipc_bclink_xmit tipc_bclink_recv_pkt->tipc_bclink_rcv tipc_disc_recv_msg->tipc_disc_rcv tipc_link_send_proto_msg->tipc_link_proto_xmit link_recv_proto_msg->tipc_link_proto_rcv link_send_sections_long->tipc_link_iovec_long_xmit tipc_link_send_sections_fast->tipc_link_iovec_xmit_fast tipc_link_send_sync->tipc_link_sync_xmit tipc_link_recv_sync->tipc_link_sync_rcv tipc_link_send_buf->__tipc_link_xmit tipc_link_send->tipc_link_xmit tipc_link_send_names->tipc_link_names_xmit tipc_named_recv->tipc_named_rcv tipc_link_recv_bundle->tipc_link_bundle_rcv tipc_link_dup_send_queue->tipc_link_dup_queue_xmit link_send_long_buf->tipc_link_frag_xmit tipc_multicast->tipc_port_mcast_xmit tipc_port_recv_mcast->tipc_port_mcast_rcv tipc_port_reject_sections->tipc_port_iovec_reject tipc_port_recv_proto_msg->tipc_port_proto_rcv tipc_connect->tipc_port_connect __tipc_connect->__tipc_port_connect __tipc_disconnect->__tipc_port_disconnect tipc_disconnect->tipc_port_disconnect tipc_shutdown->tipc_port_shutdown tipc_port_recv_msg->tipc_port_rcv tipc_port_recv_sections->tipc_port_iovec_rcv release->tipc_release accept->tipc_accept bind->tipc_bind get_name->tipc_getname poll->tipc_poll send_msg->tipc_sendmsg send_packet->tipc_send_packet send_stream->tipc_send_stream recv_msg->tipc_recvmsg recv_stream->tipc_recv_stream connect->tipc_connect listen->tipc_listen shutdown->tipc_shutdown setsockopt->tipc_setsockopt getsockopt->tipc_getsockopt Above changes have no impact on current users of the functions. Signed-off-by: Ying Xue <ying.xue@windriver.com> Reviewed-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-18 16:06:46 +08:00
tipc_link_proto_xmit(l_ptr, STATE_MSG, 0, 0,
new_value, 0, 0);
break;
case TIPC_CMD_SET_LINK_PRI:
l_ptr->priority = new_value;
tipc: align tipc function names with common naming practice in the network Rename the following functions, which are shorter and more in line with common naming practice in the network subsystem. tipc_bclink_send_msg->tipc_bclink_xmit tipc_bclink_recv_pkt->tipc_bclink_rcv tipc_disc_recv_msg->tipc_disc_rcv tipc_link_send_proto_msg->tipc_link_proto_xmit link_recv_proto_msg->tipc_link_proto_rcv link_send_sections_long->tipc_link_iovec_long_xmit tipc_link_send_sections_fast->tipc_link_iovec_xmit_fast tipc_link_send_sync->tipc_link_sync_xmit tipc_link_recv_sync->tipc_link_sync_rcv tipc_link_send_buf->__tipc_link_xmit tipc_link_send->tipc_link_xmit tipc_link_send_names->tipc_link_names_xmit tipc_named_recv->tipc_named_rcv tipc_link_recv_bundle->tipc_link_bundle_rcv tipc_link_dup_send_queue->tipc_link_dup_queue_xmit link_send_long_buf->tipc_link_frag_xmit tipc_multicast->tipc_port_mcast_xmit tipc_port_recv_mcast->tipc_port_mcast_rcv tipc_port_reject_sections->tipc_port_iovec_reject tipc_port_recv_proto_msg->tipc_port_proto_rcv tipc_connect->tipc_port_connect __tipc_connect->__tipc_port_connect __tipc_disconnect->__tipc_port_disconnect tipc_disconnect->tipc_port_disconnect tipc_shutdown->tipc_port_shutdown tipc_port_recv_msg->tipc_port_rcv tipc_port_recv_sections->tipc_port_iovec_rcv release->tipc_release accept->tipc_accept bind->tipc_bind get_name->tipc_getname poll->tipc_poll send_msg->tipc_sendmsg send_packet->tipc_send_packet send_stream->tipc_send_stream recv_msg->tipc_recvmsg recv_stream->tipc_recv_stream connect->tipc_connect listen->tipc_listen shutdown->tipc_shutdown setsockopt->tipc_setsockopt getsockopt->tipc_getsockopt Above changes have no impact on current users of the functions. Signed-off-by: Ying Xue <ying.xue@windriver.com> Reviewed-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-18 16:06:46 +08:00
tipc_link_proto_xmit(l_ptr, STATE_MSG, 0, 0,
0, new_value, 0);
break;
case TIPC_CMD_SET_LINK_WINDOW:
tipc_link_set_queue_limits(l_ptr, new_value);
break;
default:
res = -EINVAL;
break;
}
}
tipc_node_unlock(node);
return res;
}
b_ptr = tipc_bearer_find(net, name);
if (b_ptr) {
switch (cmd) {
case TIPC_CMD_SET_LINK_TOL:
b_ptr->tolerance = new_value;
break;
case TIPC_CMD_SET_LINK_PRI:
b_ptr->priority = new_value;
break;
case TIPC_CMD_SET_LINK_WINDOW:
b_ptr->window = new_value;
break;
default:
res = -EINVAL;
break;
}
return res;
}
m_ptr = tipc_media_find(name);
if (!m_ptr)
return -ENODEV;
switch (cmd) {
case TIPC_CMD_SET_LINK_TOL:
m_ptr->tolerance = new_value;
break;
case TIPC_CMD_SET_LINK_PRI:
m_ptr->priority = new_value;
break;
case TIPC_CMD_SET_LINK_WINDOW:
m_ptr->window = new_value;
break;
default:
res = -EINVAL;
break;
}
return res;
}
struct sk_buff *tipc_link_cmd_config(struct net *net, const void *req_tlv_area,
int req_tlv_space, u16 cmd)
{
struct tipc_link_config *args;
u32 new_value;
int res;
if (!TLV_CHECK(req_tlv_area, req_tlv_space, TIPC_TLV_LINK_CONFIG))
return tipc_cfg_reply_error_string(TIPC_CFG_TLV_ERROR);
args = (struct tipc_link_config *)TLV_DATA(req_tlv_area);
new_value = ntohl(args->value);
if (!link_value_is_valid(cmd, new_value))
return tipc_cfg_reply_error_string(
"cannot change, value invalid");
if (!strcmp(args->name, tipc_bclink_name)) {
if ((cmd == TIPC_CMD_SET_LINK_WINDOW) &&
(tipc_bclink_set_queue_limits(net, new_value) == 0))
return tipc_cfg_reply_none();
return tipc_cfg_reply_error_string(TIPC_CFG_NOT_SUPPORTED
" (cannot change setting on broadcast link)");
}
res = link_cmd_set_value(net, args->name, new_value, cmd);
if (res)
return tipc_cfg_reply_error_string("cannot change link setting");
return tipc_cfg_reply_none();
}
/**
* link_reset_statistics - reset link statistics
* @l_ptr: pointer to link
*/
static void link_reset_statistics(struct tipc_link *l_ptr)
{
memset(&l_ptr->stats, 0, sizeof(l_ptr->stats));
l_ptr->stats.sent_info = l_ptr->next_out_no;
l_ptr->stats.recv_info = l_ptr->next_in_no;
}
struct sk_buff *tipc_link_cmd_reset_stats(struct net *net,
const void *req_tlv_area,
int req_tlv_space)
{
char *link_name;
struct tipc_link *l_ptr;
struct tipc_node *node;
unsigned int bearer_id;
if (!TLV_CHECK(req_tlv_area, req_tlv_space, TIPC_TLV_LINK_NAME))
return tipc_cfg_reply_error_string(TIPC_CFG_TLV_ERROR);
link_name = (char *)TLV_DATA(req_tlv_area);
if (!strcmp(link_name, tipc_bclink_name)) {
if (tipc_bclink_reset_stats(net))
return tipc_cfg_reply_error_string("link not found");
return tipc_cfg_reply_none();
}
node = tipc_link_find_owner(net, link_name, &bearer_id);
tipc: purge tipc_net_lock lock Now tipc routing hierarchy comprises the structures 'node', 'link'and 'bearer'. The whole hierarchy is protected by a big read/write lock, tipc_net_lock, to ensure that nothing is added or removed while code is accessing any of these structures. Obviously the locking policy makes node, link and bearer components closely bound together so that their relationship becomes unnecessarily complex. In the worst case, such locking policy not only has a negative influence on performance, but also it's prone to lead to deadlock occasionally. In order o decouple the complex relationship between bearer and node as well as link, the locking policy is adjusted as follows: - Bearer level RTNL lock is used on update side, and RCU is used on read side. Meanwhile, all bearer instances including broadcast bearer are saved into bearer_list array. - Node and link level All node instances are saved into two tipc_node_list and node_htable lists. The two lists are protected by node_list_lock on write side, and they are guarded with RCU lock on read side. All members in node structure including link instances are protected by node spin lock. - The relationship between bearer and node When link accesses bearer, it first needs to find the bearer with its bearer identity from the bearer_list array. When bearer accesses node, it can iterate the node_htable hash list with the node address to find the corresponding node. In the new locking policy, every component has its private locking solution and the relationship between bearer and node is very simple, that is, they can find each other with node address or bearer identity from node_htable hash list or bearer_list array. Until now above all changes have been done, so tipc_net_lock can be removed safely. Signed-off-by: Ying Xue <ying.xue@windriver.com> Reviewed-by: Jon Maloy <jon.maloy@ericsson.com> Reviewed-by: Erik Hugne <erik.hugne@ericsson.com> Tested-by: Erik Hugne <erik.hugne@ericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-04-21 10:55:48 +08:00
if (!node)
return tipc_cfg_reply_error_string("link not found");
tipc: purge tipc_net_lock lock Now tipc routing hierarchy comprises the structures 'node', 'link'and 'bearer'. The whole hierarchy is protected by a big read/write lock, tipc_net_lock, to ensure that nothing is added or removed while code is accessing any of these structures. Obviously the locking policy makes node, link and bearer components closely bound together so that their relationship becomes unnecessarily complex. In the worst case, such locking policy not only has a negative influence on performance, but also it's prone to lead to deadlock occasionally. In order o decouple the complex relationship between bearer and node as well as link, the locking policy is adjusted as follows: - Bearer level RTNL lock is used on update side, and RCU is used on read side. Meanwhile, all bearer instances including broadcast bearer are saved into bearer_list array. - Node and link level All node instances are saved into two tipc_node_list and node_htable lists. The two lists are protected by node_list_lock on write side, and they are guarded with RCU lock on read side. All members in node structure including link instances are protected by node spin lock. - The relationship between bearer and node When link accesses bearer, it first needs to find the bearer with its bearer identity from the bearer_list array. When bearer accesses node, it can iterate the node_htable hash list with the node address to find the corresponding node. In the new locking policy, every component has its private locking solution and the relationship between bearer and node is very simple, that is, they can find each other with node address or bearer identity from node_htable hash list or bearer_list array. Until now above all changes have been done, so tipc_net_lock can be removed safely. Signed-off-by: Ying Xue <ying.xue@windriver.com> Reviewed-by: Jon Maloy <jon.maloy@ericsson.com> Reviewed-by: Erik Hugne <erik.hugne@ericsson.com> Tested-by: Erik Hugne <erik.hugne@ericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-04-21 10:55:48 +08:00
tipc_node_lock(node);
l_ptr = node->links[bearer_id];
if (!l_ptr) {
tipc_node_unlock(node);
return tipc_cfg_reply_error_string("link not found");
}
link_reset_statistics(l_ptr);
tipc_node_unlock(node);
return tipc_cfg_reply_none();
}
/**
* percent - convert count to a percentage of total (rounding up or down)
*/
static u32 percent(u32 count, u32 total)
{
return (count * 100 + (total / 2)) / total;
}
/**
* tipc_link_stats - print link statistics
* @net: the applicable net namespace
* @name: link name
* @buf: print buffer area
* @buf_size: size of print buffer area
*
* Returns length of print buffer data string (or 0 if error)
*/
static int tipc_link_stats(struct net *net, const char *name, char *buf,
const u32 buf_size)
{
tipc: phase out most of the struct print_buf usage The tipc_printf is renamed to tipc_snprintf, as the new name describes more what the function actually does. It is also changed to take a buffer and length parameter and return number of characters written to the buffer. All callers of this function that used to pass a print_buf are updated. Final removal of the struct print_buf itself will be done synchronously with the pending removal of the deprecated logging code that also was using it. Functions that build up a response message with a list of ports, nametable contents etc. are changed to return the number of characters written to the output buffer. This information was previously hidden in a field of the print_buf struct, and the number of chars written was fetched with a call to tipc_printbuf_validate. This function is removed since it is no longer referenced nor needed. A generic max size ULTRA_STRING_MAX_LEN is defined, named in keeping with the existing TIPC_TLV_ULTRA_STRING, and the various definitions in port, link and nametable code that largely duplicated this information are removed. This means that amount of link statistics that can be returned is now increased from 2k to 32k. The buffer overflow check is now done just before the reply message is passed over netlink or TIPC to a remote node and the message indicating a truncated buffer is changed to a less dramatic one (less CAPS), placed at the end of the message. Signed-off-by: Erik Hugne <erik.hugne@ericsson.com> Signed-off-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
2012-06-29 12:50:23 +08:00
struct tipc_link *l;
struct tipc_stats *s;
struct tipc_node *node;
char *status;
u32 profile_total = 0;
unsigned int bearer_id;
tipc: phase out most of the struct print_buf usage The tipc_printf is renamed to tipc_snprintf, as the new name describes more what the function actually does. It is also changed to take a buffer and length parameter and return number of characters written to the buffer. All callers of this function that used to pass a print_buf are updated. Final removal of the struct print_buf itself will be done synchronously with the pending removal of the deprecated logging code that also was using it. Functions that build up a response message with a list of ports, nametable contents etc. are changed to return the number of characters written to the output buffer. This information was previously hidden in a field of the print_buf struct, and the number of chars written was fetched with a call to tipc_printbuf_validate. This function is removed since it is no longer referenced nor needed. A generic max size ULTRA_STRING_MAX_LEN is defined, named in keeping with the existing TIPC_TLV_ULTRA_STRING, and the various definitions in port, link and nametable code that largely duplicated this information are removed. This means that amount of link statistics that can be returned is now increased from 2k to 32k. The buffer overflow check is now done just before the reply message is passed over netlink or TIPC to a remote node and the message indicating a truncated buffer is changed to a less dramatic one (less CAPS), placed at the end of the message. Signed-off-by: Erik Hugne <erik.hugne@ericsson.com> Signed-off-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
2012-06-29 12:50:23 +08:00
int ret;
if (!strcmp(name, tipc_bclink_name))
return tipc_bclink_stats(net, buf, buf_size);
node = tipc_link_find_owner(net, name, &bearer_id);
tipc: purge tipc_net_lock lock Now tipc routing hierarchy comprises the structures 'node', 'link'and 'bearer'. The whole hierarchy is protected by a big read/write lock, tipc_net_lock, to ensure that nothing is added or removed while code is accessing any of these structures. Obviously the locking policy makes node, link and bearer components closely bound together so that their relationship becomes unnecessarily complex. In the worst case, such locking policy not only has a negative influence on performance, but also it's prone to lead to deadlock occasionally. In order o decouple the complex relationship between bearer and node as well as link, the locking policy is adjusted as follows: - Bearer level RTNL lock is used on update side, and RCU is used on read side. Meanwhile, all bearer instances including broadcast bearer are saved into bearer_list array. - Node and link level All node instances are saved into two tipc_node_list and node_htable lists. The two lists are protected by node_list_lock on write side, and they are guarded with RCU lock on read side. All members in node structure including link instances are protected by node spin lock. - The relationship between bearer and node When link accesses bearer, it first needs to find the bearer with its bearer identity from the bearer_list array. When bearer accesses node, it can iterate the node_htable hash list with the node address to find the corresponding node. In the new locking policy, every component has its private locking solution and the relationship between bearer and node is very simple, that is, they can find each other with node address or bearer identity from node_htable hash list or bearer_list array. Until now above all changes have been done, so tipc_net_lock can be removed safely. Signed-off-by: Ying Xue <ying.xue@windriver.com> Reviewed-by: Jon Maloy <jon.maloy@ericsson.com> Reviewed-by: Erik Hugne <erik.hugne@ericsson.com> Tested-by: Erik Hugne <erik.hugne@ericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-04-21 10:55:48 +08:00
if (!node)
return 0;
tipc: purge tipc_net_lock lock Now tipc routing hierarchy comprises the structures 'node', 'link'and 'bearer'. The whole hierarchy is protected by a big read/write lock, tipc_net_lock, to ensure that nothing is added or removed while code is accessing any of these structures. Obviously the locking policy makes node, link and bearer components closely bound together so that their relationship becomes unnecessarily complex. In the worst case, such locking policy not only has a negative influence on performance, but also it's prone to lead to deadlock occasionally. In order o decouple the complex relationship between bearer and node as well as link, the locking policy is adjusted as follows: - Bearer level RTNL lock is used on update side, and RCU is used on read side. Meanwhile, all bearer instances including broadcast bearer are saved into bearer_list array. - Node and link level All node instances are saved into two tipc_node_list and node_htable lists. The two lists are protected by node_list_lock on write side, and they are guarded with RCU lock on read side. All members in node structure including link instances are protected by node spin lock. - The relationship between bearer and node When link accesses bearer, it first needs to find the bearer with its bearer identity from the bearer_list array. When bearer accesses node, it can iterate the node_htable hash list with the node address to find the corresponding node. In the new locking policy, every component has its private locking solution and the relationship between bearer and node is very simple, that is, they can find each other with node address or bearer identity from node_htable hash list or bearer_list array. Until now above all changes have been done, so tipc_net_lock can be removed safely. Signed-off-by: Ying Xue <ying.xue@windriver.com> Reviewed-by: Jon Maloy <jon.maloy@ericsson.com> Reviewed-by: Erik Hugne <erik.hugne@ericsson.com> Tested-by: Erik Hugne <erik.hugne@ericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-04-21 10:55:48 +08:00
tipc_node_lock(node);
l = node->links[bearer_id];
if (!l) {
tipc_node_unlock(node);
return 0;
}
tipc: phase out most of the struct print_buf usage The tipc_printf is renamed to tipc_snprintf, as the new name describes more what the function actually does. It is also changed to take a buffer and length parameter and return number of characters written to the buffer. All callers of this function that used to pass a print_buf are updated. Final removal of the struct print_buf itself will be done synchronously with the pending removal of the deprecated logging code that also was using it. Functions that build up a response message with a list of ports, nametable contents etc. are changed to return the number of characters written to the output buffer. This information was previously hidden in a field of the print_buf struct, and the number of chars written was fetched with a call to tipc_printbuf_validate. This function is removed since it is no longer referenced nor needed. A generic max size ULTRA_STRING_MAX_LEN is defined, named in keeping with the existing TIPC_TLV_ULTRA_STRING, and the various definitions in port, link and nametable code that largely duplicated this information are removed. This means that amount of link statistics that can be returned is now increased from 2k to 32k. The buffer overflow check is now done just before the reply message is passed over netlink or TIPC to a remote node and the message indicating a truncated buffer is changed to a less dramatic one (less CAPS), placed at the end of the message. Signed-off-by: Erik Hugne <erik.hugne@ericsson.com> Signed-off-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
2012-06-29 12:50:23 +08:00
s = &l->stats;
tipc: phase out most of the struct print_buf usage The tipc_printf is renamed to tipc_snprintf, as the new name describes more what the function actually does. It is also changed to take a buffer and length parameter and return number of characters written to the buffer. All callers of this function that used to pass a print_buf are updated. Final removal of the struct print_buf itself will be done synchronously with the pending removal of the deprecated logging code that also was using it. Functions that build up a response message with a list of ports, nametable contents etc. are changed to return the number of characters written to the output buffer. This information was previously hidden in a field of the print_buf struct, and the number of chars written was fetched with a call to tipc_printbuf_validate. This function is removed since it is no longer referenced nor needed. A generic max size ULTRA_STRING_MAX_LEN is defined, named in keeping with the existing TIPC_TLV_ULTRA_STRING, and the various definitions in port, link and nametable code that largely duplicated this information are removed. This means that amount of link statistics that can be returned is now increased from 2k to 32k. The buffer overflow check is now done just before the reply message is passed over netlink or TIPC to a remote node and the message indicating a truncated buffer is changed to a less dramatic one (less CAPS), placed at the end of the message. Signed-off-by: Erik Hugne <erik.hugne@ericsson.com> Signed-off-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
2012-06-29 12:50:23 +08:00
if (tipc_link_is_active(l))
status = "ACTIVE";
tipc: phase out most of the struct print_buf usage The tipc_printf is renamed to tipc_snprintf, as the new name describes more what the function actually does. It is also changed to take a buffer and length parameter and return number of characters written to the buffer. All callers of this function that used to pass a print_buf are updated. Final removal of the struct print_buf itself will be done synchronously with the pending removal of the deprecated logging code that also was using it. Functions that build up a response message with a list of ports, nametable contents etc. are changed to return the number of characters written to the output buffer. This information was previously hidden in a field of the print_buf struct, and the number of chars written was fetched with a call to tipc_printbuf_validate. This function is removed since it is no longer referenced nor needed. A generic max size ULTRA_STRING_MAX_LEN is defined, named in keeping with the existing TIPC_TLV_ULTRA_STRING, and the various definitions in port, link and nametable code that largely duplicated this information are removed. This means that amount of link statistics that can be returned is now increased from 2k to 32k. The buffer overflow check is now done just before the reply message is passed over netlink or TIPC to a remote node and the message indicating a truncated buffer is changed to a less dramatic one (less CAPS), placed at the end of the message. Signed-off-by: Erik Hugne <erik.hugne@ericsson.com> Signed-off-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
2012-06-29 12:50:23 +08:00
else if (tipc_link_is_up(l))
status = "STANDBY";
else
status = "DEFUNCT";
tipc: phase out most of the struct print_buf usage The tipc_printf is renamed to tipc_snprintf, as the new name describes more what the function actually does. It is also changed to take a buffer and length parameter and return number of characters written to the buffer. All callers of this function that used to pass a print_buf are updated. Final removal of the struct print_buf itself will be done synchronously with the pending removal of the deprecated logging code that also was using it. Functions that build up a response message with a list of ports, nametable contents etc. are changed to return the number of characters written to the output buffer. This information was previously hidden in a field of the print_buf struct, and the number of chars written was fetched with a call to tipc_printbuf_validate. This function is removed since it is no longer referenced nor needed. A generic max size ULTRA_STRING_MAX_LEN is defined, named in keeping with the existing TIPC_TLV_ULTRA_STRING, and the various definitions in port, link and nametable code that largely duplicated this information are removed. This means that amount of link statistics that can be returned is now increased from 2k to 32k. The buffer overflow check is now done just before the reply message is passed over netlink or TIPC to a remote node and the message indicating a truncated buffer is changed to a less dramatic one (less CAPS), placed at the end of the message. Signed-off-by: Erik Hugne <erik.hugne@ericsson.com> Signed-off-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
2012-06-29 12:50:23 +08:00
ret = tipc_snprintf(buf, buf_size, "Link <%s>\n"
" %s MTU:%u Priority:%u Tolerance:%u ms"
" Window:%u packets\n",
l->name, status, l->max_pkt, l->priority,
l->tolerance, l->queue_limit[0]);
ret += tipc_snprintf(buf + ret, buf_size - ret,
" RX packets:%u fragments:%u/%u bundles:%u/%u\n",
l->next_in_no - s->recv_info, s->recv_fragments,
s->recv_fragmented, s->recv_bundles,
s->recv_bundled);
ret += tipc_snprintf(buf + ret, buf_size - ret,
" TX packets:%u fragments:%u/%u bundles:%u/%u\n",
l->next_out_no - s->sent_info, s->sent_fragments,
s->sent_fragmented, s->sent_bundles,
s->sent_bundled);
profile_total = s->msg_length_counts;
if (!profile_total)
profile_total = 1;
tipc: phase out most of the struct print_buf usage The tipc_printf is renamed to tipc_snprintf, as the new name describes more what the function actually does. It is also changed to take a buffer and length parameter and return number of characters written to the buffer. All callers of this function that used to pass a print_buf are updated. Final removal of the struct print_buf itself will be done synchronously with the pending removal of the deprecated logging code that also was using it. Functions that build up a response message with a list of ports, nametable contents etc. are changed to return the number of characters written to the output buffer. This information was previously hidden in a field of the print_buf struct, and the number of chars written was fetched with a call to tipc_printbuf_validate. This function is removed since it is no longer referenced nor needed. A generic max size ULTRA_STRING_MAX_LEN is defined, named in keeping with the existing TIPC_TLV_ULTRA_STRING, and the various definitions in port, link and nametable code that largely duplicated this information are removed. This means that amount of link statistics that can be returned is now increased from 2k to 32k. The buffer overflow check is now done just before the reply message is passed over netlink or TIPC to a remote node and the message indicating a truncated buffer is changed to a less dramatic one (less CAPS), placed at the end of the message. Signed-off-by: Erik Hugne <erik.hugne@ericsson.com> Signed-off-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
2012-06-29 12:50:23 +08:00
ret += tipc_snprintf(buf + ret, buf_size - ret,
" TX profile sample:%u packets average:%u octets\n"
" 0-64:%u%% -256:%u%% -1024:%u%% -4096:%u%% "
"-16384:%u%% -32768:%u%% -66000:%u%%\n",
s->msg_length_counts,
s->msg_lengths_total / profile_total,
percent(s->msg_length_profile[0], profile_total),
percent(s->msg_length_profile[1], profile_total),
percent(s->msg_length_profile[2], profile_total),
percent(s->msg_length_profile[3], profile_total),
percent(s->msg_length_profile[4], profile_total),
percent(s->msg_length_profile[5], profile_total),
percent(s->msg_length_profile[6], profile_total));
ret += tipc_snprintf(buf + ret, buf_size - ret,
" RX states:%u probes:%u naks:%u defs:%u"
" dups:%u\n", s->recv_states, s->recv_probes,
s->recv_nacks, s->deferred_recv, s->duplicates);
ret += tipc_snprintf(buf + ret, buf_size - ret,
" TX states:%u probes:%u naks:%u acks:%u"
" dups:%u\n", s->sent_states, s->sent_probes,
s->sent_nacks, s->sent_acks, s->retransmitted);
ret += tipc_snprintf(buf + ret, buf_size - ret,
" Congestion link:%u Send queue"
" max:%u avg:%u\n", s->link_congs,
tipc: phase out most of the struct print_buf usage The tipc_printf is renamed to tipc_snprintf, as the new name describes more what the function actually does. It is also changed to take a buffer and length parameter and return number of characters written to the buffer. All callers of this function that used to pass a print_buf are updated. Final removal of the struct print_buf itself will be done synchronously with the pending removal of the deprecated logging code that also was using it. Functions that build up a response message with a list of ports, nametable contents etc. are changed to return the number of characters written to the output buffer. This information was previously hidden in a field of the print_buf struct, and the number of chars written was fetched with a call to tipc_printbuf_validate. This function is removed since it is no longer referenced nor needed. A generic max size ULTRA_STRING_MAX_LEN is defined, named in keeping with the existing TIPC_TLV_ULTRA_STRING, and the various definitions in port, link and nametable code that largely duplicated this information are removed. This means that amount of link statistics that can be returned is now increased from 2k to 32k. The buffer overflow check is now done just before the reply message is passed over netlink or TIPC to a remote node and the message indicating a truncated buffer is changed to a less dramatic one (less CAPS), placed at the end of the message. Signed-off-by: Erik Hugne <erik.hugne@ericsson.com> Signed-off-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
2012-06-29 12:50:23 +08:00
s->max_queue_sz, s->queue_sz_counts ?
(s->accu_queue_sz / s->queue_sz_counts) : 0);
tipc_node_unlock(node);
tipc: phase out most of the struct print_buf usage The tipc_printf is renamed to tipc_snprintf, as the new name describes more what the function actually does. It is also changed to take a buffer and length parameter and return number of characters written to the buffer. All callers of this function that used to pass a print_buf are updated. Final removal of the struct print_buf itself will be done synchronously with the pending removal of the deprecated logging code that also was using it. Functions that build up a response message with a list of ports, nametable contents etc. are changed to return the number of characters written to the output buffer. This information was previously hidden in a field of the print_buf struct, and the number of chars written was fetched with a call to tipc_printbuf_validate. This function is removed since it is no longer referenced nor needed. A generic max size ULTRA_STRING_MAX_LEN is defined, named in keeping with the existing TIPC_TLV_ULTRA_STRING, and the various definitions in port, link and nametable code that largely duplicated this information are removed. This means that amount of link statistics that can be returned is now increased from 2k to 32k. The buffer overflow check is now done just before the reply message is passed over netlink or TIPC to a remote node and the message indicating a truncated buffer is changed to a less dramatic one (less CAPS), placed at the end of the message. Signed-off-by: Erik Hugne <erik.hugne@ericsson.com> Signed-off-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
2012-06-29 12:50:23 +08:00
return ret;
}
struct sk_buff *tipc_link_cmd_show_stats(struct net *net,
const void *req_tlv_area,
int req_tlv_space)
{
struct sk_buff *buf;
struct tlv_desc *rep_tlv;
int str_len;
tipc: phase out most of the struct print_buf usage The tipc_printf is renamed to tipc_snprintf, as the new name describes more what the function actually does. It is also changed to take a buffer and length parameter and return number of characters written to the buffer. All callers of this function that used to pass a print_buf are updated. Final removal of the struct print_buf itself will be done synchronously with the pending removal of the deprecated logging code that also was using it. Functions that build up a response message with a list of ports, nametable contents etc. are changed to return the number of characters written to the output buffer. This information was previously hidden in a field of the print_buf struct, and the number of chars written was fetched with a call to tipc_printbuf_validate. This function is removed since it is no longer referenced nor needed. A generic max size ULTRA_STRING_MAX_LEN is defined, named in keeping with the existing TIPC_TLV_ULTRA_STRING, and the various definitions in port, link and nametable code that largely duplicated this information are removed. This means that amount of link statistics that can be returned is now increased from 2k to 32k. The buffer overflow check is now done just before the reply message is passed over netlink or TIPC to a remote node and the message indicating a truncated buffer is changed to a less dramatic one (less CAPS), placed at the end of the message. Signed-off-by: Erik Hugne <erik.hugne@ericsson.com> Signed-off-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
2012-06-29 12:50:23 +08:00
int pb_len;
char *pb;
if (!TLV_CHECK(req_tlv_area, req_tlv_space, TIPC_TLV_LINK_NAME))
return tipc_cfg_reply_error_string(TIPC_CFG_TLV_ERROR);
tipc: phase out most of the struct print_buf usage The tipc_printf is renamed to tipc_snprintf, as the new name describes more what the function actually does. It is also changed to take a buffer and length parameter and return number of characters written to the buffer. All callers of this function that used to pass a print_buf are updated. Final removal of the struct print_buf itself will be done synchronously with the pending removal of the deprecated logging code that also was using it. Functions that build up a response message with a list of ports, nametable contents etc. are changed to return the number of characters written to the output buffer. This information was previously hidden in a field of the print_buf struct, and the number of chars written was fetched with a call to tipc_printbuf_validate. This function is removed since it is no longer referenced nor needed. A generic max size ULTRA_STRING_MAX_LEN is defined, named in keeping with the existing TIPC_TLV_ULTRA_STRING, and the various definitions in port, link and nametable code that largely duplicated this information are removed. This means that amount of link statistics that can be returned is now increased from 2k to 32k. The buffer overflow check is now done just before the reply message is passed over netlink or TIPC to a remote node and the message indicating a truncated buffer is changed to a less dramatic one (less CAPS), placed at the end of the message. Signed-off-by: Erik Hugne <erik.hugne@ericsson.com> Signed-off-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
2012-06-29 12:50:23 +08:00
buf = tipc_cfg_reply_alloc(TLV_SPACE(ULTRA_STRING_MAX_LEN));
if (!buf)
return NULL;
rep_tlv = (struct tlv_desc *)buf->data;
tipc: phase out most of the struct print_buf usage The tipc_printf is renamed to tipc_snprintf, as the new name describes more what the function actually does. It is also changed to take a buffer and length parameter and return number of characters written to the buffer. All callers of this function that used to pass a print_buf are updated. Final removal of the struct print_buf itself will be done synchronously with the pending removal of the deprecated logging code that also was using it. Functions that build up a response message with a list of ports, nametable contents etc. are changed to return the number of characters written to the output buffer. This information was previously hidden in a field of the print_buf struct, and the number of chars written was fetched with a call to tipc_printbuf_validate. This function is removed since it is no longer referenced nor needed. A generic max size ULTRA_STRING_MAX_LEN is defined, named in keeping with the existing TIPC_TLV_ULTRA_STRING, and the various definitions in port, link and nametable code that largely duplicated this information are removed. This means that amount of link statistics that can be returned is now increased from 2k to 32k. The buffer overflow check is now done just before the reply message is passed over netlink or TIPC to a remote node and the message indicating a truncated buffer is changed to a less dramatic one (less CAPS), placed at the end of the message. Signed-off-by: Erik Hugne <erik.hugne@ericsson.com> Signed-off-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
2012-06-29 12:50:23 +08:00
pb = TLV_DATA(rep_tlv);
pb_len = ULTRA_STRING_MAX_LEN;
str_len = tipc_link_stats(net, (char *)TLV_DATA(req_tlv_area),
tipc: phase out most of the struct print_buf usage The tipc_printf is renamed to tipc_snprintf, as the new name describes more what the function actually does. It is also changed to take a buffer and length parameter and return number of characters written to the buffer. All callers of this function that used to pass a print_buf are updated. Final removal of the struct print_buf itself will be done synchronously with the pending removal of the deprecated logging code that also was using it. Functions that build up a response message with a list of ports, nametable contents etc. are changed to return the number of characters written to the output buffer. This information was previously hidden in a field of the print_buf struct, and the number of chars written was fetched with a call to tipc_printbuf_validate. This function is removed since it is no longer referenced nor needed. A generic max size ULTRA_STRING_MAX_LEN is defined, named in keeping with the existing TIPC_TLV_ULTRA_STRING, and the various definitions in port, link and nametable code that largely duplicated this information are removed. This means that amount of link statistics that can be returned is now increased from 2k to 32k. The buffer overflow check is now done just before the reply message is passed over netlink or TIPC to a remote node and the message indicating a truncated buffer is changed to a less dramatic one (less CAPS), placed at the end of the message. Signed-off-by: Erik Hugne <erik.hugne@ericsson.com> Signed-off-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
2012-06-29 12:50:23 +08:00
pb, pb_len);
if (!str_len) {
kfree_skb(buf);
return tipc_cfg_reply_error_string("link not found");
}
tipc: phase out most of the struct print_buf usage The tipc_printf is renamed to tipc_snprintf, as the new name describes more what the function actually does. It is also changed to take a buffer and length parameter and return number of characters written to the buffer. All callers of this function that used to pass a print_buf are updated. Final removal of the struct print_buf itself will be done synchronously with the pending removal of the deprecated logging code that also was using it. Functions that build up a response message with a list of ports, nametable contents etc. are changed to return the number of characters written to the output buffer. This information was previously hidden in a field of the print_buf struct, and the number of chars written was fetched with a call to tipc_printbuf_validate. This function is removed since it is no longer referenced nor needed. A generic max size ULTRA_STRING_MAX_LEN is defined, named in keeping with the existing TIPC_TLV_ULTRA_STRING, and the various definitions in port, link and nametable code that largely duplicated this information are removed. This means that amount of link statistics that can be returned is now increased from 2k to 32k. The buffer overflow check is now done just before the reply message is passed over netlink or TIPC to a remote node and the message indicating a truncated buffer is changed to a less dramatic one (less CAPS), placed at the end of the message. Signed-off-by: Erik Hugne <erik.hugne@ericsson.com> Signed-off-by: Jon Maloy <jon.maloy@ericsson.com> Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
2012-06-29 12:50:23 +08:00
str_len += 1; /* for "\0" */
skb_put(buf, TLV_SPACE(str_len));
TLV_SET(rep_tlv, TIPC_TLV_ULTRA_STRING, NULL, str_len);
return buf;
}
static void link_print(struct tipc_link *l_ptr, const char *str)
{
struct tipc_net *tn = net_generic(l_ptr->owner->net, tipc_net_id);
struct tipc_bearer *b_ptr;
rcu_read_lock();
b_ptr = rcu_dereference_rtnl(tn->bearer_list[l_ptr->bearer_id]);
if (b_ptr)
pr_info("%s Link %x<%s>:", str, l_ptr->addr, b_ptr->name);
rcu_read_unlock();
if (link_working_unknown(l_ptr))
pr_cont(":WU\n");
else if (link_reset_reset(l_ptr))
pr_cont(":RR\n");
else if (link_reset_unknown(l_ptr))
pr_cont(":RU\n");
else if (link_working_working(l_ptr))
pr_cont(":WW\n");
else
pr_cont("\n");
}
/* Parse and validate nested (link) properties valid for media, bearer and link
*/
int tipc_nl_parse_link_prop(struct nlattr *prop, struct nlattr *props[])
{
int err;
err = nla_parse_nested(props, TIPC_NLA_PROP_MAX, prop,
tipc_nl_prop_policy);
if (err)
return err;
if (props[TIPC_NLA_PROP_PRIO]) {
u32 prio;
prio = nla_get_u32(props[TIPC_NLA_PROP_PRIO]);
if (prio > TIPC_MAX_LINK_PRI)
return -EINVAL;
}
if (props[TIPC_NLA_PROP_TOL]) {
u32 tol;
tol = nla_get_u32(props[TIPC_NLA_PROP_TOL]);
if ((tol < TIPC_MIN_LINK_TOL) || (tol > TIPC_MAX_LINK_TOL))
return -EINVAL;
}
if (props[TIPC_NLA_PROP_WIN]) {
u32 win;
win = nla_get_u32(props[TIPC_NLA_PROP_WIN]);
if ((win < TIPC_MIN_LINK_WIN) || (win > TIPC_MAX_LINK_WIN))
return -EINVAL;
}
return 0;
}
int tipc_nl_link_set(struct sk_buff *skb, struct genl_info *info)
{
int err;
int res = 0;
int bearer_id;
char *name;
struct tipc_link *link;
struct tipc_node *node;
struct nlattr *attrs[TIPC_NLA_LINK_MAX + 1];
struct net *net = genl_info_net(info);
if (!info->attrs[TIPC_NLA_LINK])
return -EINVAL;
err = nla_parse_nested(attrs, TIPC_NLA_LINK_MAX,
info->attrs[TIPC_NLA_LINK],
tipc_nl_link_policy);
if (err)
return err;
if (!attrs[TIPC_NLA_LINK_NAME])
return -EINVAL;
name = nla_data(attrs[TIPC_NLA_LINK_NAME]);
node = tipc_link_find_owner(net, name, &bearer_id);
if (!node)
return -EINVAL;
tipc_node_lock(node);
link = node->links[bearer_id];
if (!link) {
res = -EINVAL;
goto out;
}
if (attrs[TIPC_NLA_LINK_PROP]) {
struct nlattr *props[TIPC_NLA_PROP_MAX + 1];
err = tipc_nl_parse_link_prop(attrs[TIPC_NLA_LINK_PROP],
props);
if (err) {
res = err;
goto out;
}
if (props[TIPC_NLA_PROP_TOL]) {
u32 tol;
tol = nla_get_u32(props[TIPC_NLA_PROP_TOL]);
link_set_supervision_props(link, tol);
tipc_link_proto_xmit(link, STATE_MSG, 0, 0, tol, 0, 0);
}
if (props[TIPC_NLA_PROP_PRIO]) {
u32 prio;
prio = nla_get_u32(props[TIPC_NLA_PROP_PRIO]);
link->priority = prio;
tipc_link_proto_xmit(link, STATE_MSG, 0, 0, 0, prio, 0);
}
if (props[TIPC_NLA_PROP_WIN]) {
u32 win;
win = nla_get_u32(props[TIPC_NLA_PROP_WIN]);
tipc_link_set_queue_limits(link, win);
}
}
out:
tipc_node_unlock(node);
return res;
}
static int __tipc_nl_add_stats(struct sk_buff *skb, struct tipc_stats *s)
{
int i;
struct nlattr *stats;
struct nla_map {
u32 key;
u32 val;
};
struct nla_map map[] = {
{TIPC_NLA_STATS_RX_INFO, s->recv_info},
{TIPC_NLA_STATS_RX_FRAGMENTS, s->recv_fragments},
{TIPC_NLA_STATS_RX_FRAGMENTED, s->recv_fragmented},
{TIPC_NLA_STATS_RX_BUNDLES, s->recv_bundles},
{TIPC_NLA_STATS_RX_BUNDLED, s->recv_bundled},
{TIPC_NLA_STATS_TX_INFO, s->sent_info},
{TIPC_NLA_STATS_TX_FRAGMENTS, s->sent_fragments},
{TIPC_NLA_STATS_TX_FRAGMENTED, s->sent_fragmented},
{TIPC_NLA_STATS_TX_BUNDLES, s->sent_bundles},
{TIPC_NLA_STATS_TX_BUNDLED, s->sent_bundled},
{TIPC_NLA_STATS_MSG_PROF_TOT, (s->msg_length_counts) ?
s->msg_length_counts : 1},
{TIPC_NLA_STATS_MSG_LEN_CNT, s->msg_length_counts},
{TIPC_NLA_STATS_MSG_LEN_TOT, s->msg_lengths_total},
{TIPC_NLA_STATS_MSG_LEN_P0, s->msg_length_profile[0]},
{TIPC_NLA_STATS_MSG_LEN_P1, s->msg_length_profile[1]},
{TIPC_NLA_STATS_MSG_LEN_P2, s->msg_length_profile[2]},
{TIPC_NLA_STATS_MSG_LEN_P3, s->msg_length_profile[3]},
{TIPC_NLA_STATS_MSG_LEN_P4, s->msg_length_profile[4]},
{TIPC_NLA_STATS_MSG_LEN_P5, s->msg_length_profile[5]},
{TIPC_NLA_STATS_MSG_LEN_P6, s->msg_length_profile[6]},
{TIPC_NLA_STATS_RX_STATES, s->recv_states},
{TIPC_NLA_STATS_RX_PROBES, s->recv_probes},
{TIPC_NLA_STATS_RX_NACKS, s->recv_nacks},
{TIPC_NLA_STATS_RX_DEFERRED, s->deferred_recv},
{TIPC_NLA_STATS_TX_STATES, s->sent_states},
{TIPC_NLA_STATS_TX_PROBES, s->sent_probes},
{TIPC_NLA_STATS_TX_NACKS, s->sent_nacks},
{TIPC_NLA_STATS_TX_ACKS, s->sent_acks},
{TIPC_NLA_STATS_RETRANSMITTED, s->retransmitted},
{TIPC_NLA_STATS_DUPLICATES, s->duplicates},
{TIPC_NLA_STATS_LINK_CONGS, s->link_congs},
{TIPC_NLA_STATS_MAX_QUEUE, s->max_queue_sz},
{TIPC_NLA_STATS_AVG_QUEUE, s->queue_sz_counts ?
(s->accu_queue_sz / s->queue_sz_counts) : 0}
};
stats = nla_nest_start(skb, TIPC_NLA_LINK_STATS);
if (!stats)
return -EMSGSIZE;
for (i = 0; i < ARRAY_SIZE(map); i++)
if (nla_put_u32(skb, map[i].key, map[i].val))
goto msg_full;
nla_nest_end(skb, stats);
return 0;
msg_full:
nla_nest_cancel(skb, stats);
return -EMSGSIZE;
}
/* Caller should hold appropriate locks to protect the link */
static int __tipc_nl_add_link(struct net *net, struct tipc_nl_msg *msg,
struct tipc_link *link)
{
int err;
void *hdr;
struct nlattr *attrs;
struct nlattr *prop;
struct tipc_net *tn = net_generic(net, tipc_net_id);
hdr = genlmsg_put(msg->skb, msg->portid, msg->seq, &tipc_genl_v2_family,
NLM_F_MULTI, TIPC_NL_LINK_GET);
if (!hdr)
return -EMSGSIZE;
attrs = nla_nest_start(msg->skb, TIPC_NLA_LINK);
if (!attrs)
goto msg_full;
if (nla_put_string(msg->skb, TIPC_NLA_LINK_NAME, link->name))
goto attr_msg_full;
if (nla_put_u32(msg->skb, TIPC_NLA_LINK_DEST,
tipc_cluster_mask(tn->own_addr)))
goto attr_msg_full;
if (nla_put_u32(msg->skb, TIPC_NLA_LINK_MTU, link->max_pkt))
goto attr_msg_full;
if (nla_put_u32(msg->skb, TIPC_NLA_LINK_RX, link->next_in_no))
goto attr_msg_full;
if (nla_put_u32(msg->skb, TIPC_NLA_LINK_TX, link->next_out_no))
goto attr_msg_full;
if (tipc_link_is_up(link))
if (nla_put_flag(msg->skb, TIPC_NLA_LINK_UP))
goto attr_msg_full;
if (tipc_link_is_active(link))
if (nla_put_flag(msg->skb, TIPC_NLA_LINK_ACTIVE))
goto attr_msg_full;
prop = nla_nest_start(msg->skb, TIPC_NLA_LINK_PROP);
if (!prop)
goto attr_msg_full;
if (nla_put_u32(msg->skb, TIPC_NLA_PROP_PRIO, link->priority))
goto prop_msg_full;
if (nla_put_u32(msg->skb, TIPC_NLA_PROP_TOL, link->tolerance))
goto prop_msg_full;
if (nla_put_u32(msg->skb, TIPC_NLA_PROP_WIN,
link->queue_limit[TIPC_LOW_IMPORTANCE]))
goto prop_msg_full;
if (nla_put_u32(msg->skb, TIPC_NLA_PROP_PRIO, link->priority))
goto prop_msg_full;
nla_nest_end(msg->skb, prop);
err = __tipc_nl_add_stats(msg->skb, &link->stats);
if (err)
goto attr_msg_full;
nla_nest_end(msg->skb, attrs);
genlmsg_end(msg->skb, hdr);
return 0;
prop_msg_full:
nla_nest_cancel(msg->skb, prop);
attr_msg_full:
nla_nest_cancel(msg->skb, attrs);
msg_full:
genlmsg_cancel(msg->skb, hdr);
return -EMSGSIZE;
}
/* Caller should hold node lock */
static int __tipc_nl_add_node_links(struct net *net, struct tipc_nl_msg *msg,
struct tipc_node *node, u32 *prev_link)
{
u32 i;
int err;
for (i = *prev_link; i < MAX_BEARERS; i++) {
*prev_link = i;
if (!node->links[i])
continue;
err = __tipc_nl_add_link(net, msg, node->links[i]);
if (err)
return err;
}
*prev_link = 0;
return 0;
}
int tipc_nl_link_dump(struct sk_buff *skb, struct netlink_callback *cb)
{
struct net *net = sock_net(skb->sk);
struct tipc_net *tn = net_generic(net, tipc_net_id);
struct tipc_node *node;
struct tipc_nl_msg msg;
u32 prev_node = cb->args[0];
u32 prev_link = cb->args[1];
int done = cb->args[2];
int err;
if (done)
return 0;
msg.skb = skb;
msg.portid = NETLINK_CB(cb->skb).portid;
msg.seq = cb->nlh->nlmsg_seq;
rcu_read_lock();
if (prev_node) {
node = tipc_node_find(net, prev_node);
if (!node) {
/* We never set seq or call nl_dump_check_consistent()
* this means that setting prev_seq here will cause the
* consistence check to fail in the netlink callback
* handler. Resulting in the last NLMSG_DONE message
* having the NLM_F_DUMP_INTR flag set.
*/
cb->prev_seq = 1;
goto out;
}
list_for_each_entry_continue_rcu(node, &tn->node_list,
list) {
tipc_node_lock(node);
err = __tipc_nl_add_node_links(net, &msg, node,
&prev_link);
tipc_node_unlock(node);
if (err)
goto out;
prev_node = node->addr;
}
} else {
err = tipc_nl_add_bc_link(net, &msg);
if (err)
goto out;
list_for_each_entry_rcu(node, &tn->node_list, list) {
tipc_node_lock(node);
err = __tipc_nl_add_node_links(net, &msg, node,
&prev_link);
tipc_node_unlock(node);
if (err)
goto out;
prev_node = node->addr;
}
}
done = 1;
out:
rcu_read_unlock();
cb->args[0] = prev_node;
cb->args[1] = prev_link;
cb->args[2] = done;
return skb->len;
}
int tipc_nl_link_get(struct sk_buff *skb, struct genl_info *info)
{
struct net *net = genl_info_net(info);
struct sk_buff *ans_skb;
struct tipc_nl_msg msg;
struct tipc_link *link;
struct tipc_node *node;
char *name;
int bearer_id;
int err;
if (!info->attrs[TIPC_NLA_LINK_NAME])
return -EINVAL;
name = nla_data(info->attrs[TIPC_NLA_LINK_NAME]);
node = tipc_link_find_owner(net, name, &bearer_id);
if (!node)
return -EINVAL;
ans_skb = nlmsg_new(NLMSG_GOODSIZE, GFP_KERNEL);
if (!ans_skb)
return -ENOMEM;
msg.skb = ans_skb;
msg.portid = info->snd_portid;
msg.seq = info->snd_seq;
tipc_node_lock(node);
link = node->links[bearer_id];
if (!link) {
err = -EINVAL;
goto err_out;
}
err = __tipc_nl_add_link(net, &msg, link);
if (err)
goto err_out;
tipc_node_unlock(node);
return genlmsg_reply(ans_skb, info);
err_out:
tipc_node_unlock(node);
nlmsg_free(ans_skb);
return err;
}
int tipc_nl_link_reset_stats(struct sk_buff *skb, struct genl_info *info)
{
int err;
char *link_name;
unsigned int bearer_id;
struct tipc_link *link;
struct tipc_node *node;
struct nlattr *attrs[TIPC_NLA_LINK_MAX + 1];
struct net *net = genl_info_net(info);
if (!info->attrs[TIPC_NLA_LINK])
return -EINVAL;
err = nla_parse_nested(attrs, TIPC_NLA_LINK_MAX,
info->attrs[TIPC_NLA_LINK],
tipc_nl_link_policy);
if (err)
return err;
if (!attrs[TIPC_NLA_LINK_NAME])
return -EINVAL;
link_name = nla_data(attrs[TIPC_NLA_LINK_NAME]);
if (strcmp(link_name, tipc_bclink_name) == 0) {
err = tipc_bclink_reset_stats(net);
if (err)
return err;
return 0;
}
node = tipc_link_find_owner(net, link_name, &bearer_id);
if (!node)
return -EINVAL;
tipc_node_lock(node);
link = node->links[bearer_id];
if (!link) {
tipc_node_unlock(node);
return -EINVAL;
}
link_reset_statistics(link);
tipc_node_unlock(node);
return 0;
}