2006-01-03 02:04:38 +08:00
|
|
|
/*
|
|
|
|
* net/tipc/name_table.c: TIPC name table code
|
2007-02-09 22:25:21 +08:00
|
|
|
*
|
2015-02-05 21:36:43 +08:00
|
|
|
* Copyright (c) 2000-2006, 2014-2015, Ericsson AB
|
2014-12-02 15:00:24 +08:00
|
|
|
* Copyright (c) 2004-2008, 2010-2014, Wind River Systems
|
2006-01-03 02:04:38 +08:00
|
|
|
* All rights reserved.
|
|
|
|
*
|
2006-01-11 20:30:43 +08:00
|
|
|
* Redistribution and use in source and binary forms, with or without
|
2006-01-03 02:04:38 +08:00
|
|
|
* modification, are permitted provided that the following conditions are met:
|
|
|
|
*
|
2006-01-11 20:30:43 +08:00
|
|
|
* 1. Redistributions of source code must retain the above copyright
|
|
|
|
* notice, this list of conditions and the following disclaimer.
|
|
|
|
* 2. Redistributions in binary form must reproduce the above copyright
|
|
|
|
* notice, this list of conditions and the following disclaimer in the
|
|
|
|
* documentation and/or other materials provided with the distribution.
|
|
|
|
* 3. Neither the names of the copyright holders nor the names of its
|
|
|
|
* contributors may be used to endorse or promote products derived from
|
|
|
|
* this software without specific prior written permission.
|
2006-01-03 02:04:38 +08:00
|
|
|
*
|
2006-01-11 20:30:43 +08:00
|
|
|
* Alternatively, this software may be distributed under the terms of the
|
|
|
|
* GNU General Public License ("GPL") version 2 as published by the Free
|
|
|
|
* Software Foundation.
|
|
|
|
*
|
|
|
|
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
|
|
|
|
* AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
|
|
|
|
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
|
|
|
|
* ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
|
|
|
|
* LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
|
|
|
|
* CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
|
|
|
|
* SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
|
|
|
|
* INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
|
|
|
|
* CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
|
|
|
|
* ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
|
2006-01-03 02:04:38 +08:00
|
|
|
* POSSIBILITY OF SUCH DAMAGE.
|
|
|
|
*/
|
|
|
|
|
2015-01-09 15:27:09 +08:00
|
|
|
#include <net/sock.h>
|
2006-01-03 02:04:38 +08:00
|
|
|
#include "core.h"
|
2015-02-09 16:50:18 +08:00
|
|
|
#include "netlink.h"
|
2006-01-03 02:04:38 +08:00
|
|
|
#include "name_table.h"
|
|
|
|
#include "name_distr.h"
|
|
|
|
#include "subscr.h"
|
2015-01-09 15:27:07 +08:00
|
|
|
#include "bcast.h"
|
2015-02-09 16:50:18 +08:00
|
|
|
#include "addr.h"
|
2015-11-20 03:30:42 +08:00
|
|
|
#include "node.h"
|
tipc: introduce communication groups
As a preparation for introducing flow control for multicast and datagram
messaging we need a more strictly defined framework than we have now. A
socket must be able keep track of exactly how many and which other
sockets it is allowed to communicate with at any moment, and keep the
necessary state for those.
We therefore introduce a new concept we have named Communication Group.
Sockets can join a group via a new setsockopt() call TIPC_GROUP_JOIN.
The call takes four parameters: 'type' serves as group identifier,
'instance' serves as an logical member identifier, and 'scope' indicates
the visibility of the group (node/cluster/zone). Finally, 'flags' makes
it possible to set certain properties for the member. For now, there is
only one flag, indicating if the creator of the socket wants to receive
a copy of broadcast or multicast messages it is sending via the socket,
and if wants to be eligible as destination for its own anycasts.
A group is closed, i.e., sockets which have not joined a group will
not be able to send messages to or receive messages from members of
the group, and vice versa.
Any member of a group can send multicast ('group broadcast') messages
to all group members, optionally including itself, using the primitive
send(). The messages are received via the recvmsg() primitive. A socket
can only be member of one group at a time.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Acked-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-13 17:04:23 +08:00
|
|
|
#include "group.h"
|
2015-02-09 16:50:18 +08:00
|
|
|
#include <net/genetlink.h>
|
2006-01-03 02:04:38 +08:00
|
|
|
|
2012-08-16 20:09:11 +08:00
|
|
|
#define TIPC_NAMETBL_SIZE 1024 /* must be a power of 2 */
|
2006-01-03 02:04:38 +08:00
|
|
|
|
|
|
|
/**
|
2011-05-30 21:44:38 +08:00
|
|
|
* struct name_info - name sequence publication info
|
2008-07-15 13:45:33 +08:00
|
|
|
* @node_list: circular list of publications made by own node
|
|
|
|
* @cluster_list: circular list of publications made by own cluster
|
|
|
|
* @zone_list: circular list of publications made by own zone
|
|
|
|
* @node_list_size: number of entries in "node_list"
|
|
|
|
* @cluster_list_size: number of entries in "cluster_list"
|
|
|
|
* @zone_list_size: number of entries in "zone_list"
|
|
|
|
*
|
|
|
|
* Note: The zone list always contains at least one entry, since all
|
|
|
|
* publications of the associated name sequence belong to it.
|
|
|
|
* (The cluster and node lists may be empty.)
|
2006-01-03 02:04:38 +08:00
|
|
|
*/
|
2011-05-30 21:44:38 +08:00
|
|
|
struct name_info {
|
2011-05-30 22:48:48 +08:00
|
|
|
struct list_head node_list;
|
|
|
|
struct list_head cluster_list;
|
|
|
|
struct list_head zone_list;
|
2008-07-15 13:45:33 +08:00
|
|
|
u32 node_list_size;
|
|
|
|
u32 cluster_list_size;
|
|
|
|
u32 zone_list_size;
|
2006-01-03 02:04:38 +08:00
|
|
|
};
|
|
|
|
|
2011-05-30 21:44:38 +08:00
|
|
|
/**
|
|
|
|
* struct sub_seq - container for all published instances of a name sequence
|
|
|
|
* @lower: name sequence lower bound
|
|
|
|
* @upper: name sequence upper bound
|
|
|
|
* @info: pointer to name sequence publication info
|
|
|
|
*/
|
|
|
|
struct sub_seq {
|
|
|
|
u32 lower;
|
|
|
|
u32 upper;
|
|
|
|
struct name_info *info;
|
|
|
|
};
|
|
|
|
|
2007-02-09 22:25:21 +08:00
|
|
|
/**
|
2006-01-03 02:04:38 +08:00
|
|
|
* struct name_seq - container for all published instances of a name type
|
|
|
|
* @type: 32 bit 'type' value for name sequence
|
|
|
|
* @sseq: pointer to dynamically-sized array of sub-sequences of this 'type';
|
|
|
|
* sub-sequences are sorted in ascending order
|
|
|
|
* @alloc: number of sub-sequences currently in array
|
2006-06-26 14:51:37 +08:00
|
|
|
* @first_free: array index of first unused sub-sequence entry
|
2006-01-03 02:04:38 +08:00
|
|
|
* @ns_list: links to adjacent name sequences in hash chain
|
|
|
|
* @subscriptions: list of subscriptions for this 'type'
|
2008-06-05 08:38:22 +08:00
|
|
|
* @lock: spinlock controlling access to publication lists of all sub-sequences
|
2014-12-02 15:00:30 +08:00
|
|
|
* @rcu: RCU callback head used for deferred freeing
|
2006-01-03 02:04:38 +08:00
|
|
|
*/
|
|
|
|
struct name_seq {
|
|
|
|
u32 type;
|
|
|
|
struct sub_seq *sseqs;
|
|
|
|
u32 alloc;
|
|
|
|
u32 first_free;
|
|
|
|
struct hlist_node ns_list;
|
|
|
|
struct list_head subscriptions;
|
|
|
|
spinlock_t lock;
|
2014-12-02 15:00:30 +08:00
|
|
|
struct rcu_head rcu;
|
2006-01-03 02:04:38 +08:00
|
|
|
};
|
|
|
|
|
2006-03-21 14:37:04 +08:00
|
|
|
static int hash(int x)
|
2006-01-03 02:04:38 +08:00
|
|
|
{
|
2012-08-16 20:09:11 +08:00
|
|
|
return x & (TIPC_NAMETBL_SIZE - 1);
|
2006-01-03 02:04:38 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* publ_create - create a publication structure
|
|
|
|
*/
|
2007-02-09 22:25:21 +08:00
|
|
|
static struct publication *publ_create(u32 type, u32 lower, u32 upper,
|
|
|
|
u32 scope, u32 node, u32 port_ref,
|
2006-01-03 02:04:38 +08:00
|
|
|
u32 key)
|
|
|
|
{
|
2006-07-22 05:51:30 +08:00
|
|
|
struct publication *publ = kzalloc(sizeof(*publ), GFP_ATOMIC);
|
2006-01-03 02:04:38 +08:00
|
|
|
if (publ == NULL) {
|
2012-06-29 12:16:37 +08:00
|
|
|
pr_warn("Publication creation failure, no memory\n");
|
2006-03-21 14:36:47 +08:00
|
|
|
return NULL;
|
2006-01-03 02:04:38 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
publ->type = type;
|
|
|
|
publ->lower = lower;
|
|
|
|
publ->upper = upper;
|
|
|
|
publ->scope = scope;
|
|
|
|
publ->node = node;
|
|
|
|
publ->ref = port_ref;
|
|
|
|
publ->key = key;
|
|
|
|
INIT_LIST_HEAD(&publ->pport_list);
|
|
|
|
return publ;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
2006-01-18 07:38:21 +08:00
|
|
|
* tipc_subseq_alloc - allocate a specified number of sub-sequence structures
|
2006-01-03 02:04:38 +08:00
|
|
|
*/
|
2006-03-21 14:37:52 +08:00
|
|
|
static struct sub_seq *tipc_subseq_alloc(u32 cnt)
|
2006-01-03 02:04:38 +08:00
|
|
|
{
|
2013-12-12 09:36:39 +08:00
|
|
|
return kcalloc(cnt, sizeof(struct sub_seq), GFP_ATOMIC);
|
2006-01-03 02:04:38 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
2006-01-18 07:38:21 +08:00
|
|
|
* tipc_nameseq_create - create a name sequence structure for the specified 'type'
|
2007-02-09 22:25:21 +08:00
|
|
|
*
|
2006-01-03 02:04:38 +08:00
|
|
|
* Allocates a single sub-sequence structure and sets it to all 0's.
|
|
|
|
*/
|
2006-03-21 14:37:52 +08:00
|
|
|
static struct name_seq *tipc_nameseq_create(u32 type, struct hlist_head *seq_head)
|
2006-01-03 02:04:38 +08:00
|
|
|
{
|
2006-07-22 05:51:30 +08:00
|
|
|
struct name_seq *nseq = kzalloc(sizeof(*nseq), GFP_ATOMIC);
|
2006-01-18 07:38:21 +08:00
|
|
|
struct sub_seq *sseq = tipc_subseq_alloc(1);
|
2006-01-03 02:04:38 +08:00
|
|
|
|
|
|
|
if (!nseq || !sseq) {
|
2012-06-29 12:16:37 +08:00
|
|
|
pr_warn("Name sequence creation failed, no memory\n");
|
2006-01-03 02:04:38 +08:00
|
|
|
kfree(nseq);
|
|
|
|
kfree(sseq);
|
2006-03-21 14:36:47 +08:00
|
|
|
return NULL;
|
2006-01-03 02:04:38 +08:00
|
|
|
}
|
|
|
|
|
2006-06-27 17:53:55 +08:00
|
|
|
spin_lock_init(&nseq->lock);
|
2006-01-03 02:04:38 +08:00
|
|
|
nseq->type = type;
|
|
|
|
nseq->sseqs = sseq;
|
|
|
|
nseq->alloc = 1;
|
|
|
|
INIT_HLIST_NODE(&nseq->ns_list);
|
|
|
|
INIT_LIST_HEAD(&nseq->subscriptions);
|
2014-12-02 15:00:30 +08:00
|
|
|
hlist_add_head_rcu(&nseq->ns_list, seq_head);
|
2006-01-03 02:04:38 +08:00
|
|
|
return nseq;
|
|
|
|
}
|
|
|
|
|
2012-07-10 18:55:09 +08:00
|
|
|
/**
|
2006-01-03 02:04:38 +08:00
|
|
|
* nameseq_find_subseq - find sub-sequence (if any) matching a name instance
|
2007-02-09 22:25:21 +08:00
|
|
|
*
|
2006-01-03 02:04:38 +08:00
|
|
|
* Very time-critical, so binary searches through sub-sequence array.
|
|
|
|
*/
|
2006-03-21 14:37:04 +08:00
|
|
|
static struct sub_seq *nameseq_find_subseq(struct name_seq *nseq,
|
|
|
|
u32 instance)
|
2006-01-03 02:04:38 +08:00
|
|
|
{
|
|
|
|
struct sub_seq *sseqs = nseq->sseqs;
|
|
|
|
int low = 0;
|
|
|
|
int high = nseq->first_free - 1;
|
|
|
|
int mid;
|
|
|
|
|
|
|
|
while (low <= high) {
|
|
|
|
mid = (low + high) / 2;
|
|
|
|
if (instance < sseqs[mid].lower)
|
|
|
|
high = mid - 1;
|
|
|
|
else if (instance > sseqs[mid].upper)
|
|
|
|
low = mid + 1;
|
|
|
|
else
|
|
|
|
return &sseqs[mid];
|
|
|
|
}
|
2006-03-21 14:36:47 +08:00
|
|
|
return NULL;
|
2006-01-03 02:04:38 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* nameseq_locate_subseq - determine position of name instance in sub-sequence
|
2007-02-09 22:25:21 +08:00
|
|
|
*
|
2006-01-03 02:04:38 +08:00
|
|
|
* Returns index in sub-sequence array of the entry that contains the specified
|
|
|
|
* instance value; if no entry contains that value, returns the position
|
|
|
|
* where a new entry for it would be inserted in the array.
|
|
|
|
*
|
|
|
|
* Note: Similar to binary search code for locating a sub-sequence.
|
|
|
|
*/
|
|
|
|
static u32 nameseq_locate_subseq(struct name_seq *nseq, u32 instance)
|
|
|
|
{
|
|
|
|
struct sub_seq *sseqs = nseq->sseqs;
|
|
|
|
int low = 0;
|
|
|
|
int high = nseq->first_free - 1;
|
|
|
|
int mid;
|
|
|
|
|
|
|
|
while (low <= high) {
|
|
|
|
mid = (low + high) / 2;
|
|
|
|
if (instance < sseqs[mid].lower)
|
|
|
|
high = mid - 1;
|
|
|
|
else if (instance > sseqs[mid].upper)
|
|
|
|
low = mid + 1;
|
|
|
|
else
|
|
|
|
return mid;
|
|
|
|
}
|
|
|
|
return low;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
2012-05-01 03:29:02 +08:00
|
|
|
* tipc_nameseq_insert_publ
|
2006-01-03 02:04:38 +08:00
|
|
|
*/
|
2015-01-09 15:27:10 +08:00
|
|
|
static struct publication *tipc_nameseq_insert_publ(struct net *net,
|
|
|
|
struct name_seq *nseq,
|
|
|
|
u32 type, u32 lower,
|
|
|
|
u32 upper, u32 scope,
|
|
|
|
u32 node, u32 port, u32 key)
|
2006-01-03 02:04:38 +08:00
|
|
|
{
|
2011-12-30 09:43:44 +08:00
|
|
|
struct tipc_subscription *s;
|
|
|
|
struct tipc_subscription *st;
|
2006-01-03 02:04:38 +08:00
|
|
|
struct publication *publ;
|
|
|
|
struct sub_seq *sseq;
|
2011-05-30 21:44:38 +08:00
|
|
|
struct name_info *info;
|
2006-01-03 02:04:38 +08:00
|
|
|
int created_subseq = 0;
|
|
|
|
|
|
|
|
sseq = nameseq_find_subseq(nseq, lower);
|
|
|
|
if (sseq) {
|
|
|
|
|
|
|
|
/* Lower end overlaps existing entry => need an exact match */
|
|
|
|
if ((sseq->lower != lower) || (sseq->upper != upper)) {
|
2006-03-21 14:36:47 +08:00
|
|
|
return NULL;
|
2006-01-03 02:04:38 +08:00
|
|
|
}
|
2011-05-30 21:44:38 +08:00
|
|
|
|
|
|
|
info = sseq->info;
|
2011-11-03 23:12:01 +08:00
|
|
|
|
|
|
|
/* Check if an identical publication already exists */
|
|
|
|
list_for_each_entry(publ, &info->zone_list, zone_list) {
|
|
|
|
if ((publ->ref == port) && (publ->key == key) &&
|
|
|
|
(!publ->node || (publ->node == node)))
|
|
|
|
return NULL;
|
|
|
|
}
|
2006-01-03 02:04:38 +08:00
|
|
|
} else {
|
|
|
|
u32 inspos;
|
|
|
|
struct sub_seq *freesseq;
|
|
|
|
|
|
|
|
/* Find where lower end should be inserted */
|
|
|
|
inspos = nameseq_locate_subseq(nseq, lower);
|
|
|
|
|
|
|
|
/* Fail if upper end overlaps into an existing entry */
|
|
|
|
if ((inspos < nseq->first_free) &&
|
|
|
|
(upper >= nseq->sseqs[inspos].lower)) {
|
2006-03-21 14:36:47 +08:00
|
|
|
return NULL;
|
2006-01-03 02:04:38 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
/* Ensure there is space for new sub-sequence */
|
|
|
|
if (nseq->first_free == nseq->alloc) {
|
2006-06-26 14:37:24 +08:00
|
|
|
struct sub_seq *sseqs = tipc_subseq_alloc(nseq->alloc * 2);
|
|
|
|
|
|
|
|
if (!sseqs) {
|
2012-06-29 12:16:37 +08:00
|
|
|
pr_warn("Cannot publish {%u,%u,%u}, no memory\n",
|
|
|
|
type, lower, upper);
|
2006-03-21 14:36:47 +08:00
|
|
|
return NULL;
|
2006-01-03 02:04:38 +08:00
|
|
|
}
|
2006-06-26 14:37:24 +08:00
|
|
|
memcpy(sseqs, nseq->sseqs,
|
|
|
|
nseq->alloc * sizeof(struct sub_seq));
|
|
|
|
kfree(nseq->sseqs);
|
|
|
|
nseq->sseqs = sseqs;
|
|
|
|
nseq->alloc *= 2;
|
2006-01-03 02:04:38 +08:00
|
|
|
}
|
|
|
|
|
2011-05-30 21:44:38 +08:00
|
|
|
info = kzalloc(sizeof(*info), GFP_ATOMIC);
|
|
|
|
if (!info) {
|
2012-06-29 12:16:37 +08:00
|
|
|
pr_warn("Cannot publish {%u,%u,%u}, no memory\n",
|
|
|
|
type, lower, upper);
|
2011-05-30 21:44:38 +08:00
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
2011-05-30 22:48:48 +08:00
|
|
|
INIT_LIST_HEAD(&info->node_list);
|
|
|
|
INIT_LIST_HEAD(&info->cluster_list);
|
|
|
|
INIT_LIST_HEAD(&info->zone_list);
|
|
|
|
|
2006-01-03 02:04:38 +08:00
|
|
|
/* Insert new sub-sequence */
|
|
|
|
sseq = &nseq->sseqs[inspos];
|
|
|
|
freesseq = &nseq->sseqs[nseq->first_free];
|
2011-01-01 02:59:32 +08:00
|
|
|
memmove(sseq + 1, sseq, (freesseq - sseq) * sizeof(*sseq));
|
|
|
|
memset(sseq, 0, sizeof(*sseq));
|
2006-01-03 02:04:38 +08:00
|
|
|
nseq->first_free++;
|
|
|
|
sseq->lower = lower;
|
|
|
|
sseq->upper = upper;
|
2011-05-30 21:44:38 +08:00
|
|
|
sseq->info = info;
|
2006-01-03 02:04:38 +08:00
|
|
|
created_subseq = 1;
|
|
|
|
}
|
|
|
|
|
2012-05-01 03:29:02 +08:00
|
|
|
/* Insert a publication */
|
2006-01-03 02:04:38 +08:00
|
|
|
publ = publ_create(type, lower, upper, scope, node, port, key);
|
|
|
|
if (!publ)
|
2006-03-21 14:36:47 +08:00
|
|
|
return NULL;
|
2006-01-03 02:04:38 +08:00
|
|
|
|
2011-05-30 22:48:48 +08:00
|
|
|
list_add(&publ->zone_list, &info->zone_list);
|
2011-05-30 21:44:38 +08:00
|
|
|
info->zone_list_size++;
|
2006-01-03 02:04:38 +08:00
|
|
|
|
2015-01-09 15:27:10 +08:00
|
|
|
if (in_own_cluster(net, node)) {
|
2011-05-30 22:48:48 +08:00
|
|
|
list_add(&publ->cluster_list, &info->cluster_list);
|
2011-05-30 21:44:38 +08:00
|
|
|
info->cluster_list_size++;
|
2006-01-03 02:04:38 +08:00
|
|
|
}
|
|
|
|
|
2015-01-09 15:27:10 +08:00
|
|
|
if (in_own_node(net, node)) {
|
2011-05-30 22:48:48 +08:00
|
|
|
list_add(&publ->node_list, &info->node_list);
|
2011-05-30 21:44:38 +08:00
|
|
|
info->node_list_size++;
|
2006-01-03 02:04:38 +08:00
|
|
|
}
|
|
|
|
|
2012-05-01 03:29:02 +08:00
|
|
|
/* Any subscriptions waiting for notification? */
|
2006-01-03 02:04:38 +08:00
|
|
|
list_for_each_entry_safe(s, st, &nseq->subscriptions, nameseq_list) {
|
2018-02-15 17:40:48 +08:00
|
|
|
tipc_sub_report_overlap(s, publ->lower, publ->upper,
|
|
|
|
TIPC_PUBLISHED, publ->ref,
|
|
|
|
publ->node, publ->scope,
|
|
|
|
created_subseq);
|
2006-01-03 02:04:38 +08:00
|
|
|
}
|
|
|
|
return publ;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
2012-05-01 03:29:02 +08:00
|
|
|
* tipc_nameseq_remove_publ
|
2007-02-09 22:25:21 +08:00
|
|
|
*
|
2006-06-26 14:51:37 +08:00
|
|
|
* NOTE: There may be cases where TIPC is asked to remove a publication
|
|
|
|
* that is not in the name table. For example, if another node issues a
|
|
|
|
* publication for a name sequence that overlaps an existing name sequence
|
|
|
|
* the publication will not be recorded, which means the publication won't
|
|
|
|
* be found when the name sequence is later withdrawn by that node.
|
|
|
|
* A failed withdraw request simply returns a failure indication and lets the
|
|
|
|
* caller issue any error or warning messages associated with such a problem.
|
2006-01-03 02:04:38 +08:00
|
|
|
*/
|
2015-01-09 15:27:10 +08:00
|
|
|
static struct publication *tipc_nameseq_remove_publ(struct net *net,
|
|
|
|
struct name_seq *nseq,
|
|
|
|
u32 inst, u32 node,
|
|
|
|
u32 ref, u32 key)
|
2006-01-03 02:04:38 +08:00
|
|
|
{
|
|
|
|
struct publication *publ;
|
|
|
|
struct sub_seq *sseq = nameseq_find_subseq(nseq, inst);
|
2011-05-30 21:44:38 +08:00
|
|
|
struct name_info *info;
|
2006-01-03 02:04:38 +08:00
|
|
|
struct sub_seq *free;
|
2011-12-30 09:43:44 +08:00
|
|
|
struct tipc_subscription *s, *st;
|
2006-01-03 02:04:38 +08:00
|
|
|
int removed_subseq = 0;
|
|
|
|
|
2006-06-26 14:51:37 +08:00
|
|
|
if (!sseq)
|
2006-03-21 14:36:47 +08:00
|
|
|
return NULL;
|
2006-06-26 14:51:37 +08:00
|
|
|
|
2011-05-30 21:44:38 +08:00
|
|
|
info = sseq->info;
|
|
|
|
|
2011-05-30 22:48:48 +08:00
|
|
|
/* Locate publication, if it exists */
|
|
|
|
list_for_each_entry(publ, &info->zone_list, zone_list) {
|
|
|
|
if ((publ->key == key) && (publ->ref == ref) &&
|
|
|
|
(!publ->node || (publ->node == node)))
|
|
|
|
goto found;
|
|
|
|
}
|
|
|
|
return NULL;
|
2007-02-09 22:25:21 +08:00
|
|
|
|
2011-05-30 22:48:48 +08:00
|
|
|
found:
|
|
|
|
/* Remove publication from zone scope list */
|
|
|
|
list_del(&publ->zone_list);
|
2011-05-30 21:44:38 +08:00
|
|
|
info->zone_list_size--;
|
2006-01-03 02:04:38 +08:00
|
|
|
|
2006-06-26 14:51:37 +08:00
|
|
|
/* Remove publication from cluster scope list, if present */
|
2015-01-09 15:27:10 +08:00
|
|
|
if (in_own_cluster(net, node)) {
|
2011-05-30 22:48:48 +08:00
|
|
|
list_del(&publ->cluster_list);
|
2011-05-30 21:44:38 +08:00
|
|
|
info->cluster_list_size--;
|
2006-01-03 02:04:38 +08:00
|
|
|
}
|
2006-06-26 14:51:37 +08:00
|
|
|
|
|
|
|
/* Remove publication from node scope list, if present */
|
2015-01-09 15:27:10 +08:00
|
|
|
if (in_own_node(net, node)) {
|
2011-05-30 22:48:48 +08:00
|
|
|
list_del(&publ->node_list);
|
2011-05-30 21:44:38 +08:00
|
|
|
info->node_list_size--;
|
2006-01-03 02:04:38 +08:00
|
|
|
}
|
|
|
|
|
2006-06-26 14:51:37 +08:00
|
|
|
/* Contract subseq list if no more publications for that subseq */
|
2011-05-30 22:48:48 +08:00
|
|
|
if (list_empty(&info->zone_list)) {
|
2011-05-30 21:44:38 +08:00
|
|
|
kfree(info);
|
2006-01-03 02:04:38 +08:00
|
|
|
free = &nseq->sseqs[nseq->first_free--];
|
2011-01-01 02:59:32 +08:00
|
|
|
memmove(sseq, sseq + 1, (free - (sseq + 1)) * sizeof(*sseq));
|
2006-01-03 02:04:38 +08:00
|
|
|
removed_subseq = 1;
|
|
|
|
}
|
|
|
|
|
2006-06-26 14:51:37 +08:00
|
|
|
/* Notify any waiting subscriptions */
|
2006-01-03 02:04:38 +08:00
|
|
|
list_for_each_entry_safe(s, st, &nseq->subscriptions, nameseq_list) {
|
2018-02-15 17:40:48 +08:00
|
|
|
tipc_sub_report_overlap(s, publ->lower, publ->upper,
|
|
|
|
TIPC_WITHDRAWN, publ->ref, publ->node,
|
|
|
|
publ->scope, removed_subseq);
|
2006-01-03 02:04:38 +08:00
|
|
|
}
|
2006-06-26 14:51:37 +08:00
|
|
|
|
2006-01-03 02:04:38 +08:00
|
|
|
return publ;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
tipc: add option to suppress PUBLISH events for pre-existing publications
Currently, when a user is subscribing for binding table publications,
he will receive a PUBLISH event for all already existing matching items
in the binding table.
However, a group socket making a subscriptions doesn't need this initial
status update from the binding table, because it has already scanned it
during the join operation. Worse, the multiplicatory effect of issuing
mutual events for dozens or hundreds group members within a short time
frame put a heavy load on the topology server, with the end result that
scale out operations on a big group tend to take much longer than needed.
We now add a new filter option, TIPC_SUB_NO_STATUS, for topology server
subscriptions, so that this initial avalanche of events is suppressed.
This change, along with the previous commit, significantly improves the
range and speed of group scale out operations.
We keep the new option internal for the tipc driver, at least for now.
Acked-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2018-01-09 04:03:29 +08:00
|
|
|
* tipc_nameseq_subscribe - attach a subscription, and optionally
|
|
|
|
* issue the prescribed number of events if there is any sub-
|
2006-01-03 02:04:38 +08:00
|
|
|
* sequence overlapping with the requested sequence
|
|
|
|
*/
|
2011-12-30 09:43:44 +08:00
|
|
|
static void tipc_nameseq_subscribe(struct name_seq *nseq,
|
2018-02-15 17:40:46 +08:00
|
|
|
struct tipc_subscription *sub)
|
2006-01-03 02:04:38 +08:00
|
|
|
{
|
|
|
|
struct sub_seq *sseq = nseq->sseqs;
|
2016-02-02 17:52:10 +08:00
|
|
|
struct tipc_name_seq ns;
|
2018-02-15 17:40:46 +08:00
|
|
|
struct tipc_subscr *s = &sub->evt.s;
|
|
|
|
bool no_status;
|
2016-02-02 17:52:10 +08:00
|
|
|
|
2018-02-15 17:40:46 +08:00
|
|
|
ns.type = tipc_sub_read(s, seq.type);
|
|
|
|
ns.lower = tipc_sub_read(s, seq.lower);
|
|
|
|
ns.upper = tipc_sub_read(s, seq.upper);
|
|
|
|
no_status = tipc_sub_read(s, filter) & TIPC_SUB_NO_STATUS;
|
2006-01-03 02:04:38 +08:00
|
|
|
|
2018-02-15 17:40:48 +08:00
|
|
|
tipc_sub_get(sub);
|
2018-02-15 17:40:46 +08:00
|
|
|
list_add(&sub->nameseq_list, &nseq->subscriptions);
|
2006-01-03 02:04:38 +08:00
|
|
|
|
2018-02-15 17:40:46 +08:00
|
|
|
if (no_status || !sseq)
|
2006-01-03 02:04:38 +08:00
|
|
|
return;
|
|
|
|
|
|
|
|
while (sseq != &nseq->sseqs[nseq->first_free]) {
|
2018-02-15 17:40:48 +08:00
|
|
|
if (tipc_sub_check_overlap(&ns, sseq->lower, sseq->upper)) {
|
2011-05-30 22:48:48 +08:00
|
|
|
struct publication *crs;
|
|
|
|
struct name_info *info = sseq->info;
|
2006-01-03 02:04:38 +08:00
|
|
|
int must_report = 1;
|
|
|
|
|
2011-05-30 22:48:48 +08:00
|
|
|
list_for_each_entry(crs, &info->zone_list, zone_list) {
|
2018-02-15 17:40:48 +08:00
|
|
|
tipc_sub_report_overlap(sub, sseq->lower,
|
|
|
|
sseq->upper,
|
|
|
|
TIPC_PUBLISHED,
|
|
|
|
crs->ref, crs->node,
|
|
|
|
crs->scope,
|
|
|
|
must_report);
|
2006-01-03 02:04:38 +08:00
|
|
|
must_report = 0;
|
2011-05-30 22:48:48 +08:00
|
|
|
}
|
2006-01-03 02:04:38 +08:00
|
|
|
}
|
|
|
|
sseq++;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2015-01-09 15:27:09 +08:00
|
|
|
static struct name_seq *nametbl_find_seq(struct net *net, u32 type)
|
2006-01-03 02:04:38 +08:00
|
|
|
{
|
2015-01-09 15:27:09 +08:00
|
|
|
struct tipc_net *tn = net_generic(net, tipc_net_id);
|
2006-01-03 02:04:38 +08:00
|
|
|
struct hlist_head *seq_head;
|
|
|
|
struct name_seq *ns;
|
|
|
|
|
2015-01-09 15:27:09 +08:00
|
|
|
seq_head = &tn->nametbl->seq_hlist[hash(type)];
|
2014-12-02 15:00:30 +08:00
|
|
|
hlist_for_each_entry_rcu(ns, seq_head, ns_list) {
|
2011-01-01 02:59:25 +08:00
|
|
|
if (ns->type == type)
|
2006-01-03 02:04:38 +08:00
|
|
|
return ns;
|
|
|
|
}
|
|
|
|
|
2006-03-21 14:36:47 +08:00
|
|
|
return NULL;
|
2006-01-03 02:04:38 +08:00
|
|
|
};
|
|
|
|
|
2015-01-09 15:27:09 +08:00
|
|
|
struct publication *tipc_nametbl_insert_publ(struct net *net, u32 type,
|
|
|
|
u32 lower, u32 upper, u32 scope,
|
|
|
|
u32 node, u32 port, u32 key)
|
2006-01-03 02:04:38 +08:00
|
|
|
{
|
2015-01-09 15:27:09 +08:00
|
|
|
struct tipc_net *tn = net_generic(net, tipc_net_id);
|
2014-12-02 15:00:26 +08:00
|
|
|
struct publication *publ;
|
2015-01-09 15:27:09 +08:00
|
|
|
struct name_seq *seq = nametbl_find_seq(net, type);
|
2014-12-02 15:00:24 +08:00
|
|
|
int index = hash(type);
|
2006-01-03 02:04:38 +08:00
|
|
|
|
2018-03-15 23:48:51 +08:00
|
|
|
if (scope > TIPC_NODE_SCOPE || lower > upper) {
|
2012-06-29 12:16:37 +08:00
|
|
|
pr_debug("Failed to publish illegal {%u,%u,%u} with scope %u\n",
|
|
|
|
type, lower, upper, scope);
|
2006-03-21 14:36:47 +08:00
|
|
|
return NULL;
|
2006-01-03 02:04:38 +08:00
|
|
|
}
|
|
|
|
|
2011-01-01 02:59:25 +08:00
|
|
|
if (!seq)
|
2015-01-09 15:27:09 +08:00
|
|
|
seq = tipc_nameseq_create(type, &tn->nametbl->seq_hlist[index]);
|
2006-01-03 02:04:38 +08:00
|
|
|
if (!seq)
|
2006-03-21 14:36:47 +08:00
|
|
|
return NULL;
|
2006-01-03 02:04:38 +08:00
|
|
|
|
2014-12-02 15:00:26 +08:00
|
|
|
spin_lock_bh(&seq->lock);
|
2015-01-09 15:27:10 +08:00
|
|
|
publ = tipc_nameseq_insert_publ(net, seq, type, lower, upper,
|
2006-01-18 07:38:21 +08:00
|
|
|
scope, node, port, key);
|
2014-12-02 15:00:26 +08:00
|
|
|
spin_unlock_bh(&seq->lock);
|
|
|
|
return publ;
|
2006-01-03 02:04:38 +08:00
|
|
|
}
|
|
|
|
|
2015-01-09 15:27:09 +08:00
|
|
|
struct publication *tipc_nametbl_remove_publ(struct net *net, u32 type,
|
|
|
|
u32 lower, u32 node, u32 ref,
|
|
|
|
u32 key)
|
2006-01-03 02:04:38 +08:00
|
|
|
{
|
|
|
|
struct publication *publ;
|
2015-01-09 15:27:09 +08:00
|
|
|
struct name_seq *seq = nametbl_find_seq(net, type);
|
2006-01-03 02:04:38 +08:00
|
|
|
|
|
|
|
if (!seq)
|
2006-03-21 14:36:47 +08:00
|
|
|
return NULL;
|
2006-01-03 02:04:38 +08:00
|
|
|
|
2014-12-02 15:00:26 +08:00
|
|
|
spin_lock_bh(&seq->lock);
|
2015-01-09 15:27:10 +08:00
|
|
|
publ = tipc_nameseq_remove_publ(net, seq, lower, node, ref, key);
|
2014-12-02 15:00:26 +08:00
|
|
|
if (!seq->first_free && list_empty(&seq->subscriptions)) {
|
2014-12-02 15:00:30 +08:00
|
|
|
hlist_del_init_rcu(&seq->ns_list);
|
2014-12-02 15:00:26 +08:00
|
|
|
kfree(seq->sseqs);
|
2014-12-02 15:00:30 +08:00
|
|
|
spin_unlock_bh(&seq->lock);
|
|
|
|
kfree_rcu(seq, rcu);
|
2014-12-02 15:00:26 +08:00
|
|
|
return publ;
|
|
|
|
}
|
|
|
|
spin_unlock_bh(&seq->lock);
|
2006-01-03 02:04:38 +08:00
|
|
|
return publ;
|
|
|
|
}
|
|
|
|
|
2012-07-10 18:55:09 +08:00
|
|
|
/**
|
2011-11-08 06:00:54 +08:00
|
|
|
* tipc_nametbl_translate - perform name translation
|
2006-01-03 02:04:38 +08:00
|
|
|
*
|
2011-11-08 06:00:54 +08:00
|
|
|
* On entry, 'destnode' is the search domain used during translation.
|
|
|
|
*
|
|
|
|
* On exit:
|
|
|
|
* - if name translation is deferred to another node/cluster/zone,
|
|
|
|
* leaves 'destnode' unchanged (will be non-zero) and returns 0
|
|
|
|
* - if name translation is attempted and succeeds, sets 'destnode'
|
|
|
|
* to publishing node and returns port reference (will be non-zero)
|
|
|
|
* - if name translation is attempted and fails, sets 'destnode' to 0
|
|
|
|
* and returns 0
|
2006-01-03 02:04:38 +08:00
|
|
|
*/
|
2015-01-09 15:27:09 +08:00
|
|
|
u32 tipc_nametbl_translate(struct net *net, u32 type, u32 instance,
|
|
|
|
u32 *destnode)
|
2006-01-03 02:04:38 +08:00
|
|
|
{
|
2015-01-09 15:27:10 +08:00
|
|
|
struct tipc_net *tn = net_generic(net, tipc_net_id);
|
2006-01-03 02:04:38 +08:00
|
|
|
struct sub_seq *sseq;
|
2011-05-30 21:44:38 +08:00
|
|
|
struct name_info *info;
|
2011-05-30 22:48:48 +08:00
|
|
|
struct publication *publ;
|
2006-01-03 02:04:38 +08:00
|
|
|
struct name_seq *seq;
|
2011-05-30 22:48:48 +08:00
|
|
|
u32 ref = 0;
|
2011-11-08 06:00:54 +08:00
|
|
|
u32 node = 0;
|
2006-01-03 02:04:38 +08:00
|
|
|
|
2015-01-09 15:27:10 +08:00
|
|
|
if (!tipc_in_scope(*destnode, tn->own_addr))
|
2006-01-03 02:04:38 +08:00
|
|
|
return 0;
|
|
|
|
|
2014-12-02 15:00:30 +08:00
|
|
|
rcu_read_lock();
|
2015-01-09 15:27:09 +08:00
|
|
|
seq = nametbl_find_seq(net, type);
|
2006-01-03 02:04:38 +08:00
|
|
|
if (unlikely(!seq))
|
|
|
|
goto not_found;
|
2014-12-02 15:00:26 +08:00
|
|
|
spin_lock_bh(&seq->lock);
|
2006-01-03 02:04:38 +08:00
|
|
|
sseq = nameseq_find_subseq(seq, instance);
|
|
|
|
if (unlikely(!sseq))
|
2014-12-02 15:00:26 +08:00
|
|
|
goto no_match;
|
2011-05-30 21:44:38 +08:00
|
|
|
info = sseq->info;
|
2006-01-03 02:04:38 +08:00
|
|
|
|
2012-05-01 03:29:02 +08:00
|
|
|
/* Closest-First Algorithm */
|
2006-01-03 02:04:38 +08:00
|
|
|
if (likely(!*destnode)) {
|
2011-05-30 22:48:48 +08:00
|
|
|
if (!list_empty(&info->node_list)) {
|
|
|
|
publ = list_first_entry(&info->node_list,
|
|
|
|
struct publication,
|
|
|
|
node_list);
|
|
|
|
list_move_tail(&publ->node_list,
|
|
|
|
&info->node_list);
|
|
|
|
} else if (!list_empty(&info->cluster_list)) {
|
|
|
|
publ = list_first_entry(&info->cluster_list,
|
|
|
|
struct publication,
|
|
|
|
cluster_list);
|
|
|
|
list_move_tail(&publ->cluster_list,
|
|
|
|
&info->cluster_list);
|
2011-05-30 23:27:50 +08:00
|
|
|
} else {
|
2011-05-30 22:48:48 +08:00
|
|
|
publ = list_first_entry(&info->zone_list,
|
|
|
|
struct publication,
|
|
|
|
zone_list);
|
|
|
|
list_move_tail(&publ->zone_list,
|
|
|
|
&info->zone_list);
|
2011-05-30 23:27:50 +08:00
|
|
|
}
|
2006-01-03 02:04:38 +08:00
|
|
|
}
|
|
|
|
|
2012-05-01 03:29:02 +08:00
|
|
|
/* Round-Robin Algorithm */
|
2015-01-09 15:27:10 +08:00
|
|
|
else if (*destnode == tn->own_addr) {
|
2011-05-30 22:48:48 +08:00
|
|
|
if (list_empty(&info->node_list))
|
|
|
|
goto no_match;
|
|
|
|
publ = list_first_entry(&info->node_list, struct publication,
|
|
|
|
node_list);
|
|
|
|
list_move_tail(&publ->node_list, &info->node_list);
|
2015-01-09 15:27:10 +08:00
|
|
|
} else if (in_own_cluster_exact(net, *destnode)) {
|
2011-05-30 22:48:48 +08:00
|
|
|
if (list_empty(&info->cluster_list))
|
|
|
|
goto no_match;
|
|
|
|
publ = list_first_entry(&info->cluster_list, struct publication,
|
|
|
|
cluster_list);
|
|
|
|
list_move_tail(&publ->cluster_list, &info->cluster_list);
|
2006-01-03 02:04:38 +08:00
|
|
|
} else {
|
2011-05-30 22:48:48 +08:00
|
|
|
publ = list_first_entry(&info->zone_list, struct publication,
|
|
|
|
zone_list);
|
|
|
|
list_move_tail(&publ->zone_list, &info->zone_list);
|
2006-01-03 02:04:38 +08:00
|
|
|
}
|
2011-05-30 22:48:48 +08:00
|
|
|
|
|
|
|
ref = publ->ref;
|
2011-11-08 06:00:54 +08:00
|
|
|
node = publ->node;
|
2011-05-30 22:48:48 +08:00
|
|
|
no_match:
|
2006-01-03 02:04:38 +08:00
|
|
|
spin_unlock_bh(&seq->lock);
|
|
|
|
not_found:
|
2014-12-02 15:00:30 +08:00
|
|
|
rcu_read_unlock();
|
2011-11-08 06:00:54 +08:00
|
|
|
*destnode = node;
|
2011-05-30 22:48:48 +08:00
|
|
|
return ref;
|
2006-01-03 02:04:38 +08:00
|
|
|
}
|
|
|
|
|
2018-01-09 04:03:30 +08:00
|
|
|
bool tipc_nametbl_lookup(struct net *net, u32 type, u32 instance, u32 scope,
|
2017-10-13 17:04:28 +08:00
|
|
|
struct list_head *dsts, int *dstcnt, u32 exclude,
|
|
|
|
bool all)
|
|
|
|
{
|
|
|
|
u32 self = tipc_own_addr(net);
|
|
|
|
struct publication *publ;
|
|
|
|
struct name_info *info;
|
|
|
|
struct name_seq *seq;
|
|
|
|
struct sub_seq *sseq;
|
|
|
|
|
|
|
|
*dstcnt = 0;
|
|
|
|
rcu_read_lock();
|
|
|
|
seq = nametbl_find_seq(net, type);
|
|
|
|
if (unlikely(!seq))
|
|
|
|
goto exit;
|
|
|
|
spin_lock_bh(&seq->lock);
|
|
|
|
sseq = nameseq_find_subseq(seq, instance);
|
|
|
|
if (likely(sseq)) {
|
|
|
|
info = sseq->info;
|
|
|
|
list_for_each_entry(publ, &info->zone_list, zone_list) {
|
2018-01-09 04:03:30 +08:00
|
|
|
if (publ->scope != scope)
|
2017-10-13 17:04:28 +08:00
|
|
|
continue;
|
|
|
|
if (publ->ref == exclude && publ->node == self)
|
|
|
|
continue;
|
|
|
|
tipc_dest_push(dsts, publ->node, publ->ref);
|
|
|
|
(*dstcnt)++;
|
|
|
|
if (all)
|
|
|
|
continue;
|
|
|
|
list_move_tail(&publ->zone_list, &info->zone_list);
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
spin_unlock_bh(&seq->lock);
|
|
|
|
exit:
|
|
|
|
rcu_read_unlock();
|
|
|
|
return !list_empty(dsts);
|
|
|
|
}
|
|
|
|
|
2018-01-09 04:03:30 +08:00
|
|
|
int tipc_nametbl_mc_lookup(struct net *net, u32 type, u32 lower, u32 upper,
|
|
|
|
u32 scope, bool exact, struct list_head *dports)
|
2006-01-03 02:04:38 +08:00
|
|
|
{
|
|
|
|
struct sub_seq *sseq_stop;
|
2011-05-30 21:44:38 +08:00
|
|
|
struct name_info *info;
|
2018-01-09 04:03:30 +08:00
|
|
|
struct publication *p;
|
|
|
|
struct name_seq *seq;
|
|
|
|
struct sub_seq *sseq;
|
2006-01-03 02:04:38 +08:00
|
|
|
int res = 0;
|
|
|
|
|
2014-12-02 15:00:30 +08:00
|
|
|
rcu_read_lock();
|
2015-01-09 15:27:09 +08:00
|
|
|
seq = nametbl_find_seq(net, type);
|
2006-01-03 02:04:38 +08:00
|
|
|
if (!seq)
|
|
|
|
goto exit;
|
|
|
|
|
|
|
|
spin_lock_bh(&seq->lock);
|
|
|
|
sseq = seq->sseqs + nameseq_locate_subseq(seq, lower);
|
|
|
|
sseq_stop = seq->sseqs + seq->first_free;
|
|
|
|
for (; sseq != sseq_stop; sseq++) {
|
|
|
|
if (sseq->lower > upper)
|
|
|
|
break;
|
2011-05-30 21:44:38 +08:00
|
|
|
info = sseq->info;
|
2018-01-09 04:03:30 +08:00
|
|
|
list_for_each_entry(p, &info->node_list, node_list) {
|
|
|
|
if (p->scope == scope || (!exact && p->scope < scope))
|
|
|
|
tipc_dest_push(dports, 0, p->ref);
|
2008-07-15 13:45:33 +08:00
|
|
|
}
|
|
|
|
|
2011-05-30 21:44:38 +08:00
|
|
|
if (info->cluster_list_size != info->node_list_size)
|
2008-07-15 13:45:33 +08:00
|
|
|
res = 1;
|
2006-01-03 02:04:38 +08:00
|
|
|
}
|
|
|
|
spin_unlock_bh(&seq->lock);
|
|
|
|
exit:
|
2014-12-02 15:00:30 +08:00
|
|
|
rcu_read_unlock();
|
2006-01-03 02:04:38 +08:00
|
|
|
return res;
|
|
|
|
}
|
|
|
|
|
2017-01-19 02:50:51 +08:00
|
|
|
/* tipc_nametbl_lookup_dst_nodes - find broadcast destination nodes
|
|
|
|
* - Creates list of nodes that overlap the given multicast address
|
|
|
|
* - Determines if any node local ports overlap
|
|
|
|
*/
|
|
|
|
void tipc_nametbl_lookup_dst_nodes(struct net *net, u32 type, u32 lower,
|
2018-01-13 03:56:50 +08:00
|
|
|
u32 upper, struct tipc_nlist *nodes)
|
2017-01-19 02:50:51 +08:00
|
|
|
{
|
|
|
|
struct sub_seq *sseq, *stop;
|
|
|
|
struct publication *publ;
|
|
|
|
struct name_info *info;
|
|
|
|
struct name_seq *seq;
|
|
|
|
|
|
|
|
rcu_read_lock();
|
|
|
|
seq = nametbl_find_seq(net, type);
|
|
|
|
if (!seq)
|
|
|
|
goto exit;
|
|
|
|
|
|
|
|
spin_lock_bh(&seq->lock);
|
|
|
|
sseq = seq->sseqs + nameseq_locate_subseq(seq, lower);
|
|
|
|
stop = seq->sseqs + seq->first_free;
|
2017-10-25 22:19:52 +08:00
|
|
|
for (; sseq != stop && sseq->lower <= upper; sseq++) {
|
2017-01-19 02:50:51 +08:00
|
|
|
info = sseq->info;
|
|
|
|
list_for_each_entry(publ, &info->zone_list, zone_list) {
|
2018-01-13 03:56:50 +08:00
|
|
|
tipc_nlist_add(nodes, publ->node);
|
2017-01-19 02:50:51 +08:00
|
|
|
}
|
|
|
|
}
|
|
|
|
spin_unlock_bh(&seq->lock);
|
|
|
|
exit:
|
|
|
|
rcu_read_unlock();
|
|
|
|
}
|
|
|
|
|
tipc: introduce communication groups
As a preparation for introducing flow control for multicast and datagram
messaging we need a more strictly defined framework than we have now. A
socket must be able keep track of exactly how many and which other
sockets it is allowed to communicate with at any moment, and keep the
necessary state for those.
We therefore introduce a new concept we have named Communication Group.
Sockets can join a group via a new setsockopt() call TIPC_GROUP_JOIN.
The call takes four parameters: 'type' serves as group identifier,
'instance' serves as an logical member identifier, and 'scope' indicates
the visibility of the group (node/cluster/zone). Finally, 'flags' makes
it possible to set certain properties for the member. For now, there is
only one flag, indicating if the creator of the socket wants to receive
a copy of broadcast or multicast messages it is sending via the socket,
and if wants to be eligible as destination for its own anycasts.
A group is closed, i.e., sockets which have not joined a group will
not be able to send messages to or receive messages from members of
the group, and vice versa.
Any member of a group can send multicast ('group broadcast') messages
to all group members, optionally including itself, using the primitive
send(). The messages are received via the recvmsg() primitive. A socket
can only be member of one group at a time.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Acked-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-13 17:04:23 +08:00
|
|
|
/* tipc_nametbl_build_group - build list of communication group members
|
|
|
|
*/
|
|
|
|
void tipc_nametbl_build_group(struct net *net, struct tipc_group *grp,
|
2018-01-09 04:03:30 +08:00
|
|
|
u32 type, u32 scope)
|
tipc: introduce communication groups
As a preparation for introducing flow control for multicast and datagram
messaging we need a more strictly defined framework than we have now. A
socket must be able keep track of exactly how many and which other
sockets it is allowed to communicate with at any moment, and keep the
necessary state for those.
We therefore introduce a new concept we have named Communication Group.
Sockets can join a group via a new setsockopt() call TIPC_GROUP_JOIN.
The call takes four parameters: 'type' serves as group identifier,
'instance' serves as an logical member identifier, and 'scope' indicates
the visibility of the group (node/cluster/zone). Finally, 'flags' makes
it possible to set certain properties for the member. For now, there is
only one flag, indicating if the creator of the socket wants to receive
a copy of broadcast or multicast messages it is sending via the socket,
and if wants to be eligible as destination for its own anycasts.
A group is closed, i.e., sockets which have not joined a group will
not be able to send messages to or receive messages from members of
the group, and vice versa.
Any member of a group can send multicast ('group broadcast') messages
to all group members, optionally including itself, using the primitive
send(). The messages are received via the recvmsg() primitive. A socket
can only be member of one group at a time.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Acked-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-13 17:04:23 +08:00
|
|
|
{
|
|
|
|
struct sub_seq *sseq, *stop;
|
|
|
|
struct name_info *info;
|
|
|
|
struct publication *p;
|
|
|
|
struct name_seq *seq;
|
|
|
|
|
|
|
|
rcu_read_lock();
|
|
|
|
seq = nametbl_find_seq(net, type);
|
|
|
|
if (!seq)
|
|
|
|
goto exit;
|
|
|
|
|
|
|
|
spin_lock_bh(&seq->lock);
|
|
|
|
sseq = seq->sseqs;
|
|
|
|
stop = seq->sseqs + seq->first_free;
|
|
|
|
for (; sseq != stop; sseq++) {
|
|
|
|
info = sseq->info;
|
|
|
|
list_for_each_entry(p, &info->zone_list, zone_list) {
|
2018-01-09 04:03:30 +08:00
|
|
|
if (p->scope != scope)
|
tipc: introduce communication groups
As a preparation for introducing flow control for multicast and datagram
messaging we need a more strictly defined framework than we have now. A
socket must be able keep track of exactly how many and which other
sockets it is allowed to communicate with at any moment, and keep the
necessary state for those.
We therefore introduce a new concept we have named Communication Group.
Sockets can join a group via a new setsockopt() call TIPC_GROUP_JOIN.
The call takes four parameters: 'type' serves as group identifier,
'instance' serves as an logical member identifier, and 'scope' indicates
the visibility of the group (node/cluster/zone). Finally, 'flags' makes
it possible to set certain properties for the member. For now, there is
only one flag, indicating if the creator of the socket wants to receive
a copy of broadcast or multicast messages it is sending via the socket,
and if wants to be eligible as destination for its own anycasts.
A group is closed, i.e., sockets which have not joined a group will
not be able to send messages to or receive messages from members of
the group, and vice versa.
Any member of a group can send multicast ('group broadcast') messages
to all group members, optionally including itself, using the primitive
send(). The messages are received via the recvmsg() primitive. A socket
can only be member of one group at a time.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Acked-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-13 17:04:23 +08:00
|
|
|
continue;
|
2018-01-09 04:03:28 +08:00
|
|
|
tipc_group_add_member(grp, p->node, p->ref, p->lower);
|
tipc: introduce communication groups
As a preparation for introducing flow control for multicast and datagram
messaging we need a more strictly defined framework than we have now. A
socket must be able keep track of exactly how many and which other
sockets it is allowed to communicate with at any moment, and keep the
necessary state for those.
We therefore introduce a new concept we have named Communication Group.
Sockets can join a group via a new setsockopt() call TIPC_GROUP_JOIN.
The call takes four parameters: 'type' serves as group identifier,
'instance' serves as an logical member identifier, and 'scope' indicates
the visibility of the group (node/cluster/zone). Finally, 'flags' makes
it possible to set certain properties for the member. For now, there is
only one flag, indicating if the creator of the socket wants to receive
a copy of broadcast or multicast messages it is sending via the socket,
and if wants to be eligible as destination for its own anycasts.
A group is closed, i.e., sockets which have not joined a group will
not be able to send messages to or receive messages from members of
the group, and vice versa.
Any member of a group can send multicast ('group broadcast') messages
to all group members, optionally including itself, using the primitive
send(). The messages are received via the recvmsg() primitive. A socket
can only be member of one group at a time.
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Acked-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-13 17:04:23 +08:00
|
|
|
}
|
|
|
|
}
|
|
|
|
spin_unlock_bh(&seq->lock);
|
|
|
|
exit:
|
|
|
|
rcu_read_unlock();
|
|
|
|
}
|
|
|
|
|
2011-11-03 03:49:40 +08:00
|
|
|
/*
|
2006-01-18 07:38:21 +08:00
|
|
|
* tipc_nametbl_publish - add name publication to network name tables
|
2006-01-03 02:04:38 +08:00
|
|
|
*/
|
2015-01-09 15:27:05 +08:00
|
|
|
struct publication *tipc_nametbl_publish(struct net *net, u32 type, u32 lower,
|
|
|
|
u32 upper, u32 scope, u32 port_ref,
|
|
|
|
u32 key)
|
2006-01-03 02:04:38 +08:00
|
|
|
{
|
|
|
|
struct publication *publ;
|
2014-04-28 18:00:10 +08:00
|
|
|
struct sk_buff *buf = NULL;
|
2015-01-09 15:27:09 +08:00
|
|
|
struct tipc_net *tn = net_generic(net, tipc_net_id);
|
2006-01-03 02:04:38 +08:00
|
|
|
|
2015-01-09 15:27:09 +08:00
|
|
|
spin_lock_bh(&tn->nametbl_lock);
|
|
|
|
if (tn->nametbl->local_publ_count >= TIPC_MAX_PUBLICATIONS) {
|
2012-06-29 12:16:37 +08:00
|
|
|
pr_warn("Publication failed, local publication limit reached (%u)\n",
|
2012-08-16 20:09:14 +08:00
|
|
|
TIPC_MAX_PUBLICATIONS);
|
2015-01-09 15:27:09 +08:00
|
|
|
spin_unlock_bh(&tn->nametbl_lock);
|
2006-03-21 14:36:47 +08:00
|
|
|
return NULL;
|
2006-01-03 02:04:38 +08:00
|
|
|
}
|
|
|
|
|
2015-01-09 15:27:09 +08:00
|
|
|
publ = tipc_nametbl_insert_publ(net, type, lower, upper, scope,
|
2015-01-09 15:27:10 +08:00
|
|
|
tn->own_addr, port_ref, key);
|
2011-11-10 03:22:52 +08:00
|
|
|
if (likely(publ)) {
|
2015-01-09 15:27:09 +08:00
|
|
|
tn->nametbl->local_publ_count++;
|
|
|
|
buf = tipc_named_publish(net, publ);
|
tipc: add name distributor resiliency queue
TIPC name table updates are distributed asynchronously in a cluster,
entailing a risk of certain race conditions. E.g., if two nodes
simultaneously issue conflicting (overlapping) publications, this may
not be detected until both publications have reached a third node, in
which case one of the publications will be silently dropped on that
node. Hence, we end up with an inconsistent name table.
In most cases this conflict is just a temporary race, e.g., one
node is issuing a publication under the assumption that a previous,
conflicting, publication has already been withdrawn by the other node.
However, because of the (rtt related) distributed update delay, this
may not yet hold true on all nodes. The symptom of this failure is a
syslog message: "tipc: Cannot publish {%u,%u,%u}, overlap error".
In this commit we add a resiliency queue at the receiving end of
the name table distributor. When insertion of an arriving publication
fails, we retain it in this queue for a short amount of time, assuming
that another update will arrive very soon and clear the conflict. If so
happens, we insert the publication, otherwise we drop it.
The (configurable) retention value defaults to 2000 ms. Knowing from
experience that the situation described above is extremely rare, there
is no risk that the queue will accumulate any large number of items.
Signed-off-by: Erik Hugne <erik.hugne@ericsson.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Acked-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-08-28 15:08:47 +08:00
|
|
|
/* Any pending external events? */
|
2015-01-09 15:27:05 +08:00
|
|
|
tipc_named_process_backlog(net);
|
2011-11-10 03:22:52 +08:00
|
|
|
}
|
2015-01-09 15:27:09 +08:00
|
|
|
spin_unlock_bh(&tn->nametbl_lock);
|
2014-04-28 18:00:10 +08:00
|
|
|
|
|
|
|
if (buf)
|
2015-11-20 03:30:42 +08:00
|
|
|
tipc_node_broadcast(net, buf);
|
2006-01-03 02:04:38 +08:00
|
|
|
return publ;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
2006-01-18 07:38:21 +08:00
|
|
|
* tipc_nametbl_withdraw - withdraw name publication from network name tables
|
2006-01-03 02:04:38 +08:00
|
|
|
*/
|
2015-01-09 15:27:05 +08:00
|
|
|
int tipc_nametbl_withdraw(struct net *net, u32 type, u32 lower, u32 ref,
|
|
|
|
u32 key)
|
2006-01-03 02:04:38 +08:00
|
|
|
{
|
|
|
|
struct publication *publ;
|
2014-12-02 15:00:28 +08:00
|
|
|
struct sk_buff *skb = NULL;
|
2015-01-09 15:27:09 +08:00
|
|
|
struct tipc_net *tn = net_generic(net, tipc_net_id);
|
2006-01-03 02:04:38 +08:00
|
|
|
|
2015-01-09 15:27:09 +08:00
|
|
|
spin_lock_bh(&tn->nametbl_lock);
|
2015-01-09 15:27:10 +08:00
|
|
|
publ = tipc_nametbl_remove_publ(net, type, lower, tn->own_addr,
|
2015-01-09 15:27:09 +08:00
|
|
|
ref, key);
|
2006-06-26 14:51:37 +08:00
|
|
|
if (likely(publ)) {
|
2015-01-09 15:27:09 +08:00
|
|
|
tn->nametbl->local_publ_count--;
|
2015-01-09 15:27:10 +08:00
|
|
|
skb = tipc_named_withdraw(net, publ);
|
tipc: add name distributor resiliency queue
TIPC name table updates are distributed asynchronously in a cluster,
entailing a risk of certain race conditions. E.g., if two nodes
simultaneously issue conflicting (overlapping) publications, this may
not be detected until both publications have reached a third node, in
which case one of the publications will be silently dropped on that
node. Hence, we end up with an inconsistent name table.
In most cases this conflict is just a temporary race, e.g., one
node is issuing a publication under the assumption that a previous,
conflicting, publication has already been withdrawn by the other node.
However, because of the (rtt related) distributed update delay, this
may not yet hold true on all nodes. The symptom of this failure is a
syslog message: "tipc: Cannot publish {%u,%u,%u}, overlap error".
In this commit we add a resiliency queue at the receiving end of
the name table distributor. When insertion of an arriving publication
fails, we retain it in this queue for a short amount of time, assuming
that another update will arrive very soon and clear the conflict. If so
happens, we insert the publication, otherwise we drop it.
The (configurable) retention value defaults to 2000 ms. Knowing from
experience that the situation described above is extremely rare, there
is no risk that the queue will accumulate any large number of items.
Signed-off-by: Erik Hugne <erik.hugne@ericsson.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Acked-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-08-28 15:08:47 +08:00
|
|
|
/* Any pending external events? */
|
2015-01-09 15:27:05 +08:00
|
|
|
tipc_named_process_backlog(net);
|
2006-01-03 02:04:38 +08:00
|
|
|
list_del_init(&publ->pport_list);
|
2014-12-02 15:00:30 +08:00
|
|
|
kfree_rcu(publ, rcu);
|
2014-12-02 15:00:28 +08:00
|
|
|
} else {
|
|
|
|
pr_err("Unable to remove local publication\n"
|
|
|
|
"(type=%u, lower=%u, ref=%u, key=%u)\n",
|
|
|
|
type, lower, ref, key);
|
|
|
|
}
|
2015-01-09 15:27:09 +08:00
|
|
|
spin_unlock_bh(&tn->nametbl_lock);
|
2014-04-28 18:00:10 +08:00
|
|
|
|
2014-12-02 15:00:28 +08:00
|
|
|
if (skb) {
|
2015-11-20 03:30:42 +08:00
|
|
|
tipc_node_broadcast(net, skb);
|
2006-01-03 02:04:38 +08:00
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
2006-01-18 07:38:21 +08:00
|
|
|
* tipc_nametbl_subscribe - add a subscription object to the name table
|
2006-01-03 02:04:38 +08:00
|
|
|
*/
|
2018-02-15 17:40:46 +08:00
|
|
|
void tipc_nametbl_subscribe(struct tipc_subscription *sub)
|
2006-01-03 02:04:38 +08:00
|
|
|
{
|
2018-02-15 17:40:49 +08:00
|
|
|
struct tipc_net *tn = tipc_net(sub->net);
|
2018-02-15 17:40:46 +08:00
|
|
|
struct tipc_subscr *s = &sub->evt.s;
|
|
|
|
u32 type = tipc_sub_read(s, seq.type);
|
2014-12-02 15:00:24 +08:00
|
|
|
int index = hash(type);
|
2006-01-03 02:04:38 +08:00
|
|
|
struct name_seq *seq;
|
2016-02-02 17:52:10 +08:00
|
|
|
struct tipc_name_seq ns;
|
2006-01-03 02:04:38 +08:00
|
|
|
|
2015-01-09 15:27:09 +08:00
|
|
|
spin_lock_bh(&tn->nametbl_lock);
|
2018-02-15 17:40:49 +08:00
|
|
|
seq = nametbl_find_seq(sub->net, type);
|
2011-01-01 02:59:35 +08:00
|
|
|
if (!seq)
|
2015-01-09 15:27:09 +08:00
|
|
|
seq = tipc_nameseq_create(type, &tn->nametbl->seq_hlist[index]);
|
2011-01-01 02:59:32 +08:00
|
|
|
if (seq) {
|
2007-02-09 22:25:21 +08:00
|
|
|
spin_lock_bh(&seq->lock);
|
2018-02-15 17:40:46 +08:00
|
|
|
tipc_nameseq_subscribe(seq, sub);
|
2007-02-09 22:25:21 +08:00
|
|
|
spin_unlock_bh(&seq->lock);
|
|
|
|
} else {
|
2018-02-15 17:40:46 +08:00
|
|
|
ns.type = tipc_sub_read(s, seq.type);
|
|
|
|
ns.lower = tipc_sub_read(s, seq.lower);
|
|
|
|
ns.upper = tipc_sub_read(s, seq.upper);
|
2012-06-29 12:16:37 +08:00
|
|
|
pr_warn("Failed to create subscription for {%u,%u,%u}\n",
|
2016-02-02 17:52:10 +08:00
|
|
|
ns.type, ns.lower, ns.upper);
|
2007-02-09 22:25:21 +08:00
|
|
|
}
|
2015-01-09 15:27:09 +08:00
|
|
|
spin_unlock_bh(&tn->nametbl_lock);
|
2006-01-03 02:04:38 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
2006-01-18 07:38:21 +08:00
|
|
|
* tipc_nametbl_unsubscribe - remove a subscription object from name table
|
2006-01-03 02:04:38 +08:00
|
|
|
*/
|
2018-02-15 17:40:46 +08:00
|
|
|
void tipc_nametbl_unsubscribe(struct tipc_subscription *sub)
|
2006-01-03 02:04:38 +08:00
|
|
|
{
|
2018-02-15 17:40:46 +08:00
|
|
|
struct tipc_subscr *s = &sub->evt.s;
|
2018-02-15 17:40:49 +08:00
|
|
|
struct tipc_net *tn = tipc_net(sub->net);
|
2006-01-03 02:04:38 +08:00
|
|
|
struct name_seq *seq;
|
2018-02-15 17:40:46 +08:00
|
|
|
u32 type = tipc_sub_read(s, seq.type);
|
2006-01-03 02:04:38 +08:00
|
|
|
|
2015-01-09 15:27:09 +08:00
|
|
|
spin_lock_bh(&tn->nametbl_lock);
|
2018-02-15 17:40:49 +08:00
|
|
|
seq = nametbl_find_seq(sub->net, type);
|
2011-01-01 02:59:32 +08:00
|
|
|
if (seq != NULL) {
|
2007-02-09 22:25:21 +08:00
|
|
|
spin_lock_bh(&seq->lock);
|
2018-02-15 17:40:46 +08:00
|
|
|
list_del_init(&sub->nameseq_list);
|
2018-02-15 17:40:48 +08:00
|
|
|
tipc_sub_put(sub);
|
2014-12-02 15:00:26 +08:00
|
|
|
if (!seq->first_free && list_empty(&seq->subscriptions)) {
|
2014-12-02 15:00:30 +08:00
|
|
|
hlist_del_init_rcu(&seq->ns_list);
|
2014-12-02 15:00:26 +08:00
|
|
|
kfree(seq->sseqs);
|
2014-12-02 15:00:30 +08:00
|
|
|
spin_unlock_bh(&seq->lock);
|
|
|
|
kfree_rcu(seq, rcu);
|
2014-12-02 15:00:26 +08:00
|
|
|
} else {
|
|
|
|
spin_unlock_bh(&seq->lock);
|
|
|
|
}
|
2007-02-09 22:25:21 +08:00
|
|
|
}
|
2015-01-09 15:27:09 +08:00
|
|
|
spin_unlock_bh(&tn->nametbl_lock);
|
2006-01-03 02:04:38 +08:00
|
|
|
}
|
|
|
|
|
2015-01-09 15:27:09 +08:00
|
|
|
int tipc_nametbl_init(struct net *net)
|
2006-01-03 02:04:38 +08:00
|
|
|
{
|
2015-01-09 15:27:09 +08:00
|
|
|
struct tipc_net *tn = net_generic(net, tipc_net_id);
|
|
|
|
struct name_table *tipc_nametbl;
|
2014-12-02 15:00:24 +08:00
|
|
|
int i;
|
|
|
|
|
|
|
|
tipc_nametbl = kzalloc(sizeof(*tipc_nametbl), GFP_ATOMIC);
|
|
|
|
if (!tipc_nametbl)
|
2006-01-03 02:04:38 +08:00
|
|
|
return -ENOMEM;
|
|
|
|
|
2014-12-02 15:00:24 +08:00
|
|
|
for (i = 0; i < TIPC_NAMETBL_SIZE; i++)
|
|
|
|
INIT_HLIST_HEAD(&tipc_nametbl->seq_hlist[i]);
|
|
|
|
|
|
|
|
INIT_LIST_HEAD(&tipc_nametbl->publ_list[TIPC_ZONE_SCOPE]);
|
|
|
|
INIT_LIST_HEAD(&tipc_nametbl->publ_list[TIPC_CLUSTER_SCOPE]);
|
|
|
|
INIT_LIST_HEAD(&tipc_nametbl->publ_list[TIPC_NODE_SCOPE]);
|
2015-01-09 15:27:09 +08:00
|
|
|
tn->nametbl = tipc_nametbl;
|
|
|
|
spin_lock_init(&tn->nametbl_lock);
|
2006-01-03 02:04:38 +08:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2014-03-06 21:40:20 +08:00
|
|
|
/**
|
|
|
|
* tipc_purge_publications - remove all publications for a given type
|
|
|
|
*
|
|
|
|
* tipc_nametbl_lock must be held when calling this function
|
|
|
|
*/
|
2015-01-09 15:27:09 +08:00
|
|
|
static void tipc_purge_publications(struct net *net, struct name_seq *seq)
|
2014-03-06 21:40:20 +08:00
|
|
|
{
|
|
|
|
struct publication *publ, *safe;
|
|
|
|
struct sub_seq *sseq;
|
|
|
|
struct name_info *info;
|
|
|
|
|
2014-12-02 15:00:26 +08:00
|
|
|
spin_lock_bh(&seq->lock);
|
2014-03-06 21:40:20 +08:00
|
|
|
sseq = seq->sseqs;
|
|
|
|
info = sseq->info;
|
|
|
|
list_for_each_entry_safe(publ, safe, &info->zone_list, zone_list) {
|
tipc: fix a potential deadlock when nametable is purged
[ 28.531768] =============================================
[ 28.532322] [ INFO: possible recursive locking detected ]
[ 28.532322] 3.19.0+ #194 Not tainted
[ 28.532322] ---------------------------------------------
[ 28.532322] insmod/583 is trying to acquire lock:
[ 28.532322] (&(&nseq->lock)->rlock){+.....}, at: [<ffffffffa000d219>] tipc_nametbl_remove_publ+0x49/0x2e0 [tipc]
[ 28.532322]
[ 28.532322] but task is already holding lock:
[ 28.532322] (&(&nseq->lock)->rlock){+.....}, at: [<ffffffffa000e0dc>] tipc_nametbl_stop+0xfc/0x1f0 [tipc]
[ 28.532322]
[ 28.532322] other info that might help us debug this:
[ 28.532322] Possible unsafe locking scenario:
[ 28.532322]
[ 28.532322] CPU0
[ 28.532322] ----
[ 28.532322] lock(&(&nseq->lock)->rlock);
[ 28.532322] lock(&(&nseq->lock)->rlock);
[ 28.532322]
[ 28.532322] *** DEADLOCK ***
[ 28.532322]
[ 28.532322] May be due to missing lock nesting notation
[ 28.532322]
[ 28.532322] 3 locks held by insmod/583:
[ 28.532322] #0: (net_mutex){+.+.+.}, at: [<ffffffff8163e30f>] register_pernet_subsys+0x1f/0x50
[ 28.532322] #1: (&(&tn->nametbl_lock)->rlock){+.....}, at: [<ffffffffa000e091>] tipc_nametbl_stop+0xb1/0x1f0 [tipc]
[ 28.532322] #2: (&(&nseq->lock)->rlock){+.....}, at: [<ffffffffa000e0dc>] tipc_nametbl_stop+0xfc/0x1f0 [tipc]
[ 28.532322]
[ 28.532322] stack backtrace:
[ 28.532322] CPU: 1 PID: 583 Comm: insmod Not tainted 3.19.0+ #194
[ 28.532322] Hardware name: Bochs Bochs, BIOS Bochs 01/01/2007
[ 28.532322] ffffffff82394460 ffff8800144cb928 ffffffff81792f3e 0000000000000007
[ 28.532322] ffffffff82394460 ffff8800144cba28 ffffffff810a8080 ffff8800144cb998
[ 28.532322] ffffffff810a4df3 ffff880013e9cb10 ffffffff82b0d330 ffff880013e9cb38
[ 28.532322] Call Trace:
[ 28.532322] [<ffffffff81792f3e>] dump_stack+0x4c/0x65
[ 28.532322] [<ffffffff810a8080>] __lock_acquire+0x740/0x1ca0
[ 28.532322] [<ffffffff810a4df3>] ? __bfs+0x23/0x270
[ 28.532322] [<ffffffff810a7506>] ? check_irq_usage+0x96/0xe0
[ 28.532322] [<ffffffff810a8a73>] ? __lock_acquire+0x1133/0x1ca0
[ 28.532322] [<ffffffffa000d219>] ? tipc_nametbl_remove_publ+0x49/0x2e0 [tipc]
[ 28.532322] [<ffffffff810a9c0c>] lock_acquire+0x9c/0x140
[ 28.532322] [<ffffffffa000d219>] ? tipc_nametbl_remove_publ+0x49/0x2e0 [tipc]
[ 28.532322] [<ffffffff8179c41f>] _raw_spin_lock_bh+0x3f/0x50
[ 28.532322] [<ffffffffa000d219>] ? tipc_nametbl_remove_publ+0x49/0x2e0 [tipc]
[ 28.532322] [<ffffffffa000d219>] tipc_nametbl_remove_publ+0x49/0x2e0 [tipc]
[ 28.532322] [<ffffffffa000e11e>] tipc_nametbl_stop+0x13e/0x1f0 [tipc]
[ 28.532322] [<ffffffffa000dfe5>] ? tipc_nametbl_stop+0x5/0x1f0 [tipc]
[ 28.532322] [<ffffffffa0004bab>] tipc_init_net+0x13b/0x150 [tipc]
[ 28.532322] [<ffffffffa0004a75>] ? tipc_init_net+0x5/0x150 [tipc]
[ 28.532322] [<ffffffff8163dece>] ops_init+0x4e/0x150
[ 28.532322] [<ffffffff810aa66d>] ? trace_hardirqs_on+0xd/0x10
[ 28.532322] [<ffffffff8163e1d3>] register_pernet_operations+0xf3/0x190
[ 28.532322] [<ffffffff8163e31e>] register_pernet_subsys+0x2e/0x50
[ 28.532322] [<ffffffffa002406a>] tipc_init+0x6a/0x1000 [tipc]
[ 28.532322] [<ffffffffa0024000>] ? 0xffffffffa0024000
[ 28.532322] [<ffffffff810002d9>] do_one_initcall+0x89/0x1c0
[ 28.532322] [<ffffffff811b7cb0>] ? kmem_cache_alloc_trace+0x50/0x1b0
[ 28.532322] [<ffffffff810e725b>] ? do_init_module+0x2b/0x200
[ 28.532322] [<ffffffff810e7294>] do_init_module+0x64/0x200
[ 28.532322] [<ffffffff810e9353>] load_module+0x12f3/0x18e0
[ 28.532322] [<ffffffff810e5890>] ? show_initstate+0x50/0x50
[ 28.532322] [<ffffffff810e9a19>] SyS_init_module+0xd9/0x110
[ 28.532322] [<ffffffff8179f3b3>] sysenter_dispatch+0x7/0x1f
Before tipc_purge_publications() calls tipc_nametbl_remove_publ() to
remove a publication with a name sequence, the name sequence's lock
is held. However, when tipc_nametbl_remove_publ() calling
tipc_nameseq_remove_publ() to remove the publication, it first tries
to query name sequence instance with the publication, and then holds
the lock of the found name sequence. But as the lock may be already
taken in tipc_purge_publications(), deadlock happens like above
scenario demonstrated. As tipc_nameseq_remove_publ() doesn't grab name
sequence's lock, the deadlock can be avoided if it's directly invoked
by tipc_purge_publications().
Fixes: 97ede29e80ee ("tipc: convert name table read-write lock to RCU")
Signed-off-by: Ying Xue <ying.xue@windriver.com>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2015-03-18 09:32:58 +08:00
|
|
|
tipc_nameseq_remove_publ(net, seq, publ->lower, publ->node,
|
|
|
|
publ->ref, publ->key);
|
2014-12-02 15:00:30 +08:00
|
|
|
kfree_rcu(publ, rcu);
|
2014-03-06 21:40:20 +08:00
|
|
|
}
|
2014-12-02 15:00:30 +08:00
|
|
|
hlist_del_init_rcu(&seq->ns_list);
|
|
|
|
kfree(seq->sseqs);
|
2014-12-09 15:17:56 +08:00
|
|
|
spin_unlock_bh(&seq->lock);
|
2014-12-02 15:00:26 +08:00
|
|
|
|
2014-12-02 15:00:30 +08:00
|
|
|
kfree_rcu(seq, rcu);
|
2014-03-06 21:40:20 +08:00
|
|
|
}
|
|
|
|
|
2015-01-09 15:27:09 +08:00
|
|
|
void tipc_nametbl_stop(struct net *net)
|
2006-01-03 02:04:38 +08:00
|
|
|
{
|
|
|
|
u32 i;
|
2014-03-06 21:40:20 +08:00
|
|
|
struct name_seq *seq;
|
|
|
|
struct hlist_head *seq_head;
|
2015-01-09 15:27:09 +08:00
|
|
|
struct tipc_net *tn = net_generic(net, tipc_net_id);
|
|
|
|
struct name_table *tipc_nametbl = tn->nametbl;
|
2006-01-03 02:04:38 +08:00
|
|
|
|
2014-03-06 21:40:20 +08:00
|
|
|
/* Verify name table is empty and purge any lingering
|
|
|
|
* publications, then release the name table
|
|
|
|
*/
|
2015-01-09 15:27:09 +08:00
|
|
|
spin_lock_bh(&tn->nametbl_lock);
|
2012-08-16 20:09:11 +08:00
|
|
|
for (i = 0; i < TIPC_NAMETBL_SIZE; i++) {
|
2014-12-02 15:00:24 +08:00
|
|
|
if (hlist_empty(&tipc_nametbl->seq_hlist[i]))
|
2012-07-12 05:35:01 +08:00
|
|
|
continue;
|
2014-12-02 15:00:24 +08:00
|
|
|
seq_head = &tipc_nametbl->seq_hlist[i];
|
2014-12-02 15:00:30 +08:00
|
|
|
hlist_for_each_entry_rcu(seq, seq_head, ns_list) {
|
2015-01-09 15:27:09 +08:00
|
|
|
tipc_purge_publications(net, seq);
|
2014-03-06 21:40:20 +08:00
|
|
|
}
|
2006-01-03 02:04:38 +08:00
|
|
|
}
|
2015-01-09 15:27:09 +08:00
|
|
|
spin_unlock_bh(&tn->nametbl_lock);
|
2014-12-02 15:00:24 +08:00
|
|
|
|
2014-12-02 15:00:30 +08:00
|
|
|
synchronize_net();
|
2014-12-02 15:00:24 +08:00
|
|
|
kfree(tipc_nametbl);
|
|
|
|
|
2006-01-03 02:04:38 +08:00
|
|
|
}
|
2014-11-20 17:29:20 +08:00
|
|
|
|
2014-11-24 18:10:29 +08:00
|
|
|
static int __tipc_nl_add_nametable_publ(struct tipc_nl_msg *msg,
|
|
|
|
struct name_seq *seq,
|
|
|
|
struct sub_seq *sseq, u32 *last_publ)
|
2014-11-20 17:29:20 +08:00
|
|
|
{
|
|
|
|
void *hdr;
|
|
|
|
struct nlattr *attrs;
|
|
|
|
struct nlattr *publ;
|
|
|
|
struct publication *p;
|
|
|
|
|
|
|
|
if (*last_publ) {
|
|
|
|
list_for_each_entry(p, &sseq->info->zone_list, zone_list)
|
|
|
|
if (p->key == *last_publ)
|
|
|
|
break;
|
|
|
|
if (p->key != *last_publ)
|
|
|
|
return -EPIPE;
|
|
|
|
} else {
|
|
|
|
p = list_first_entry(&sseq->info->zone_list, struct publication,
|
|
|
|
zone_list);
|
|
|
|
}
|
|
|
|
|
|
|
|
list_for_each_entry_from(p, &sseq->info->zone_list, zone_list) {
|
|
|
|
*last_publ = p->key;
|
|
|
|
|
|
|
|
hdr = genlmsg_put(msg->skb, msg->portid, msg->seq,
|
2015-02-09 16:50:03 +08:00
|
|
|
&tipc_genl_family, NLM_F_MULTI,
|
2014-11-20 17:29:20 +08:00
|
|
|
TIPC_NL_NAME_TABLE_GET);
|
|
|
|
if (!hdr)
|
|
|
|
return -EMSGSIZE;
|
|
|
|
|
|
|
|
attrs = nla_nest_start(msg->skb, TIPC_NLA_NAME_TABLE);
|
|
|
|
if (!attrs)
|
|
|
|
goto msg_full;
|
|
|
|
|
|
|
|
publ = nla_nest_start(msg->skb, TIPC_NLA_NAME_TABLE_PUBL);
|
|
|
|
if (!publ)
|
|
|
|
goto attr_msg_full;
|
|
|
|
|
|
|
|
if (nla_put_u32(msg->skb, TIPC_NLA_PUBL_TYPE, seq->type))
|
|
|
|
goto publ_msg_full;
|
|
|
|
if (nla_put_u32(msg->skb, TIPC_NLA_PUBL_LOWER, sseq->lower))
|
|
|
|
goto publ_msg_full;
|
|
|
|
if (nla_put_u32(msg->skb, TIPC_NLA_PUBL_UPPER, sseq->upper))
|
|
|
|
goto publ_msg_full;
|
|
|
|
if (nla_put_u32(msg->skb, TIPC_NLA_PUBL_SCOPE, p->scope))
|
|
|
|
goto publ_msg_full;
|
|
|
|
if (nla_put_u32(msg->skb, TIPC_NLA_PUBL_NODE, p->node))
|
|
|
|
goto publ_msg_full;
|
|
|
|
if (nla_put_u32(msg->skb, TIPC_NLA_PUBL_REF, p->ref))
|
|
|
|
goto publ_msg_full;
|
|
|
|
if (nla_put_u32(msg->skb, TIPC_NLA_PUBL_KEY, p->key))
|
|
|
|
goto publ_msg_full;
|
|
|
|
|
|
|
|
nla_nest_end(msg->skb, publ);
|
|
|
|
nla_nest_end(msg->skb, attrs);
|
|
|
|
genlmsg_end(msg->skb, hdr);
|
|
|
|
}
|
|
|
|
*last_publ = 0;
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
publ_msg_full:
|
|
|
|
nla_nest_cancel(msg->skb, publ);
|
|
|
|
attr_msg_full:
|
|
|
|
nla_nest_cancel(msg->skb, attrs);
|
|
|
|
msg_full:
|
|
|
|
genlmsg_cancel(msg->skb, hdr);
|
|
|
|
|
|
|
|
return -EMSGSIZE;
|
|
|
|
}
|
|
|
|
|
2014-11-24 18:10:29 +08:00
|
|
|
static int __tipc_nl_subseq_list(struct tipc_nl_msg *msg, struct name_seq *seq,
|
|
|
|
u32 *last_lower, u32 *last_publ)
|
2014-11-20 17:29:20 +08:00
|
|
|
{
|
|
|
|
struct sub_seq *sseq;
|
|
|
|
struct sub_seq *sseq_start;
|
|
|
|
int err;
|
|
|
|
|
|
|
|
if (*last_lower) {
|
|
|
|
sseq_start = nameseq_find_subseq(seq, *last_lower);
|
|
|
|
if (!sseq_start)
|
|
|
|
return -EPIPE;
|
|
|
|
} else {
|
|
|
|
sseq_start = seq->sseqs;
|
|
|
|
}
|
|
|
|
|
|
|
|
for (sseq = sseq_start; sseq != &seq->sseqs[seq->first_free]; sseq++) {
|
|
|
|
err = __tipc_nl_add_nametable_publ(msg, seq, sseq, last_publ);
|
|
|
|
if (err) {
|
|
|
|
*last_lower = sseq->lower;
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
*last_lower = 0;
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2015-01-09 15:27:09 +08:00
|
|
|
static int tipc_nl_seq_list(struct net *net, struct tipc_nl_msg *msg,
|
|
|
|
u32 *last_type, u32 *last_lower, u32 *last_publ)
|
2014-11-20 17:29:20 +08:00
|
|
|
{
|
2015-01-09 15:27:09 +08:00
|
|
|
struct tipc_net *tn = net_generic(net, tipc_net_id);
|
2014-11-20 17:29:20 +08:00
|
|
|
struct hlist_head *seq_head;
|
2014-12-02 15:00:30 +08:00
|
|
|
struct name_seq *seq = NULL;
|
2014-11-20 17:29:20 +08:00
|
|
|
int err;
|
|
|
|
int i;
|
|
|
|
|
|
|
|
if (*last_type)
|
|
|
|
i = hash(*last_type);
|
|
|
|
else
|
|
|
|
i = 0;
|
|
|
|
|
|
|
|
for (; i < TIPC_NAMETBL_SIZE; i++) {
|
2015-01-09 15:27:09 +08:00
|
|
|
seq_head = &tn->nametbl->seq_hlist[i];
|
2014-11-20 17:29:20 +08:00
|
|
|
|
|
|
|
if (*last_type) {
|
2015-01-09 15:27:09 +08:00
|
|
|
seq = nametbl_find_seq(net, *last_type);
|
2014-11-20 17:29:20 +08:00
|
|
|
if (!seq)
|
|
|
|
return -EPIPE;
|
|
|
|
} else {
|
2014-12-02 15:00:30 +08:00
|
|
|
hlist_for_each_entry_rcu(seq, seq_head, ns_list)
|
|
|
|
break;
|
2014-11-20 17:29:20 +08:00
|
|
|
if (!seq)
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
2014-12-02 15:00:30 +08:00
|
|
|
hlist_for_each_entry_from_rcu(seq, ns_list) {
|
2014-11-20 17:29:20 +08:00
|
|
|
spin_lock_bh(&seq->lock);
|
|
|
|
err = __tipc_nl_subseq_list(msg, seq, last_lower,
|
|
|
|
last_publ);
|
|
|
|
|
|
|
|
if (err) {
|
|
|
|
*last_type = seq->type;
|
|
|
|
spin_unlock_bh(&seq->lock);
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
spin_unlock_bh(&seq->lock);
|
|
|
|
}
|
|
|
|
*last_type = 0;
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
int tipc_nl_name_table_dump(struct sk_buff *skb, struct netlink_callback *cb)
|
|
|
|
{
|
|
|
|
int err;
|
|
|
|
int done = cb->args[3];
|
|
|
|
u32 last_type = cb->args[0];
|
|
|
|
u32 last_lower = cb->args[1];
|
|
|
|
u32 last_publ = cb->args[2];
|
2015-01-09 15:27:09 +08:00
|
|
|
struct net *net = sock_net(skb->sk);
|
2014-11-20 17:29:20 +08:00
|
|
|
struct tipc_nl_msg msg;
|
|
|
|
|
|
|
|
if (done)
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
msg.skb = skb;
|
|
|
|
msg.portid = NETLINK_CB(cb->skb).portid;
|
|
|
|
msg.seq = cb->nlh->nlmsg_seq;
|
|
|
|
|
2014-12-02 15:00:30 +08:00
|
|
|
rcu_read_lock();
|
2015-01-09 15:27:09 +08:00
|
|
|
err = tipc_nl_seq_list(net, &msg, &last_type, &last_lower, &last_publ);
|
2014-11-20 17:29:20 +08:00
|
|
|
if (!err) {
|
|
|
|
done = 1;
|
|
|
|
} else if (err != -EMSGSIZE) {
|
|
|
|
/* 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 NLMSG_DONE message having the NLM_F_DUMP_INTR flag set if
|
|
|
|
* we got an error.
|
|
|
|
*/
|
|
|
|
cb->prev_seq = 1;
|
|
|
|
}
|
2014-12-02 15:00:30 +08:00
|
|
|
rcu_read_unlock();
|
2014-11-20 17:29:20 +08:00
|
|
|
|
|
|
|
cb->args[0] = last_type;
|
|
|
|
cb->args[1] = last_lower;
|
|
|
|
cb->args[2] = last_publ;
|
|
|
|
cb->args[3] = done;
|
|
|
|
|
|
|
|
return skb->len;
|
|
|
|
}
|
2015-02-05 21:36:43 +08:00
|
|
|
|
2017-10-13 17:04:22 +08:00
|
|
|
struct tipc_dest *tipc_dest_find(struct list_head *l, u32 node, u32 port)
|
2015-02-05 21:36:43 +08:00
|
|
|
{
|
2017-10-13 17:04:22 +08:00
|
|
|
u64 value = (u64)node << 32 | port;
|
|
|
|
struct tipc_dest *dst;
|
2015-02-05 21:36:43 +08:00
|
|
|
|
2017-10-13 17:04:22 +08:00
|
|
|
list_for_each_entry(dst, l, list) {
|
|
|
|
if (dst->value != value)
|
|
|
|
continue;
|
|
|
|
return dst;
|
2015-02-05 21:36:43 +08:00
|
|
|
}
|
2017-10-13 17:04:22 +08:00
|
|
|
return NULL;
|
2017-01-03 23:55:10 +08:00
|
|
|
}
|
|
|
|
|
2017-10-13 17:04:22 +08:00
|
|
|
bool tipc_dest_push(struct list_head *l, u32 node, u32 port)
|
2017-01-03 23:55:10 +08:00
|
|
|
{
|
2017-10-13 17:04:22 +08:00
|
|
|
u64 value = (u64)node << 32 | port;
|
|
|
|
struct tipc_dest *dst;
|
2017-01-03 23:55:10 +08:00
|
|
|
|
2017-10-13 17:04:22 +08:00
|
|
|
if (tipc_dest_find(l, node, port))
|
2017-01-03 23:55:10 +08:00
|
|
|
return false;
|
|
|
|
|
2017-10-13 17:04:22 +08:00
|
|
|
dst = kmalloc(sizeof(*dst), GFP_ATOMIC);
|
|
|
|
if (unlikely(!dst))
|
|
|
|
return false;
|
|
|
|
dst->value = value;
|
|
|
|
list_add(&dst->list, l);
|
2017-01-03 23:55:10 +08:00
|
|
|
return true;
|
|
|
|
}
|
|
|
|
|
2017-10-13 17:04:22 +08:00
|
|
|
bool tipc_dest_pop(struct list_head *l, u32 *node, u32 *port)
|
2017-01-03 23:55:10 +08:00
|
|
|
{
|
2017-10-13 17:04:22 +08:00
|
|
|
struct tipc_dest *dst;
|
2017-01-03 23:55:10 +08:00
|
|
|
|
|
|
|
if (list_empty(l))
|
2017-10-13 17:04:22 +08:00
|
|
|
return false;
|
|
|
|
dst = list_first_entry(l, typeof(*dst), list);
|
|
|
|
if (port)
|
|
|
|
*port = dst->port;
|
|
|
|
if (node)
|
|
|
|
*node = dst->node;
|
|
|
|
list_del(&dst->list);
|
|
|
|
kfree(dst);
|
|
|
|
return true;
|
2017-01-03 23:55:10 +08:00
|
|
|
}
|
|
|
|
|
2017-10-13 17:04:22 +08:00
|
|
|
bool tipc_dest_del(struct list_head *l, u32 node, u32 port)
|
2017-01-03 23:55:10 +08:00
|
|
|
{
|
2017-10-13 17:04:22 +08:00
|
|
|
struct tipc_dest *dst;
|
2017-01-03 23:55:10 +08:00
|
|
|
|
2017-10-13 17:04:22 +08:00
|
|
|
dst = tipc_dest_find(l, node, port);
|
|
|
|
if (!dst)
|
|
|
|
return false;
|
|
|
|
list_del(&dst->list);
|
|
|
|
kfree(dst);
|
|
|
|
return true;
|
2017-01-03 23:55:10 +08:00
|
|
|
}
|
|
|
|
|
2017-10-13 17:04:22 +08:00
|
|
|
void tipc_dest_list_purge(struct list_head *l)
|
2017-01-03 23:55:10 +08:00
|
|
|
{
|
2017-10-13 17:04:22 +08:00
|
|
|
struct tipc_dest *dst, *tmp;
|
2017-01-03 23:55:10 +08:00
|
|
|
|
2017-10-13 17:04:22 +08:00
|
|
|
list_for_each_entry_safe(dst, tmp, l, list) {
|
|
|
|
list_del(&dst->list);
|
|
|
|
kfree(dst);
|
2015-02-05 21:36:43 +08:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2017-10-13 17:04:22 +08:00
|
|
|
int tipc_dest_list_len(struct list_head *l)
|
2015-02-05 21:36:43 +08:00
|
|
|
{
|
2017-10-13 17:04:22 +08:00
|
|
|
struct tipc_dest *dst;
|
2017-01-03 23:55:10 +08:00
|
|
|
int i = 0;
|
2015-02-05 21:36:43 +08:00
|
|
|
|
2017-10-13 17:04:22 +08:00
|
|
|
list_for_each_entry(dst, l, list) {
|
2017-01-03 23:55:10 +08:00
|
|
|
i++;
|
2015-02-05 21:36:43 +08:00
|
|
|
}
|
2017-01-03 23:55:10 +08:00
|
|
|
return i;
|
2015-02-05 21:36:43 +08:00
|
|
|
}
|