summaryrefslogtreecommitdiff
path: root/net/tipc
diff options
context:
space:
mode:
Diffstat (limited to 'net/tipc')
-rw-r--r--net/tipc/Kconfig22
-rw-r--r--net/tipc/Makefile5
-rw-r--r--net/tipc/bcast.c57
-rw-r--r--net/tipc/bcast.h3
-rw-r--r--net/tipc/bearer.c147
-rw-r--r--net/tipc/bearer.h40
-rw-r--r--net/tipc/config.c165
-rw-r--r--net/tipc/core.c61
-rw-r--r--net/tipc/core.h35
-rw-r--r--net/tipc/discover.c11
-rw-r--r--net/tipc/eth_media.c85
-rw-r--r--net/tipc/handler.c3
-rw-r--r--net/tipc/ib_media.c372
-rw-r--r--net/tipc/link.c335
-rw-r--r--net/tipc/link.h4
-rw-r--r--net/tipc/msg.c19
-rw-r--r--net/tipc/msg.h8
-rw-r--r--net/tipc/name_distr.c2
-rw-r--r--net/tipc/name_table.c34
-rw-r--r--net/tipc/name_table.h11
-rw-r--r--net/tipc/net.c3
-rw-r--r--net/tipc/net.h2
-rw-r--r--net/tipc/netlink.c8
-rw-r--r--net/tipc/node.c18
-rw-r--r--net/tipc/node.h6
-rw-r--r--net/tipc/node_subscr.c2
-rw-r--r--net/tipc/port.c350
-rw-r--r--net/tipc/port.h91
-rw-r--r--net/tipc/server.c605
-rw-r--r--net/tipc/server.h94
-rw-r--r--net/tipc/socket.c624
-rw-r--r--net/tipc/subscr.c350
-rw-r--r--net/tipc/subscr.h21
-rw-r--r--net/tipc/sysctl.c64
34 files changed, 2107 insertions, 1550 deletions
diff --git a/net/tipc/Kconfig b/net/tipc/Kconfig
index 585460180ffb..c890848f9d56 100644
--- a/net/tipc/Kconfig
+++ b/net/tipc/Kconfig
@@ -3,8 +3,8 @@
#
menuconfig TIPC
- tristate "The TIPC Protocol (EXPERIMENTAL)"
- depends on INET && EXPERIMENTAL
+ tristate "The TIPC Protocol"
+ depends on INET
---help---
The Transparent Inter Process Communication (TIPC) protocol is
specially designed for intra cluster communication. This protocol
@@ -20,18 +20,9 @@ menuconfig TIPC
If in doubt, say N.
-if TIPC
-
-config TIPC_ADVANCED
- bool "Advanced TIPC configuration"
- default n
- help
- Saying Y here will open some advanced configuration for TIPC.
- Most users do not need to bother; if unsure, just say N.
-
config TIPC_PORTS
int "Maximum number of ports in a node"
- depends on TIPC_ADVANCED
+ depends on TIPC
range 127 65535
default "8191"
help
@@ -41,4 +32,9 @@ config TIPC_PORTS
Setting this to a smaller value saves some memory,
setting it to higher allows for more ports.
-endif # TIPC
+config TIPC_MEDIA_IB
+ bool "InfiniBand media type support"
+ depends on TIPC && INFINIBAND_IPOIB
+ help
+ Saying Y here will enable support for running TIPC on
+ IP-over-InfiniBand devices.
diff --git a/net/tipc/Makefile b/net/tipc/Makefile
index 6cd55d671d3a..b282f7130d2b 100644
--- a/net/tipc/Makefile
+++ b/net/tipc/Makefile
@@ -8,4 +8,7 @@ tipc-y += addr.o bcast.o bearer.o config.o \
core.o handler.o link.o discover.o msg.o \
name_distr.o subscr.o name_table.o net.o \
netlink.o node.o node_subscr.o port.o ref.o \
- socket.o log.o eth_media.o
+ socket.o log.o eth_media.o server.o
+
+tipc-$(CONFIG_TIPC_MEDIA_IB) += ib_media.o
+tipc-$(CONFIG_SYSCTL) += sysctl.o
diff --git a/net/tipc/bcast.c b/net/tipc/bcast.c
index e4e6d8cd47e6..716de1ac6cb5 100644
--- a/net/tipc/bcast.c
+++ b/net/tipc/bcast.c
@@ -347,7 +347,7 @@ static void bclink_peek_nack(struct tipc_msg *msg)
tipc_node_lock(n_ptr);
- if (n_ptr->bclink.supported &&
+ if (n_ptr->bclink.recv_permitted &&
(n_ptr->bclink.last_in != n_ptr->bclink.last_sent) &&
(n_ptr->bclink.last_in == msg_bcgap_after(msg)))
n_ptr->bclink.oos_state = 2;
@@ -429,7 +429,7 @@ void tipc_bclink_recv_pkt(struct sk_buff *buf)
goto exit;
tipc_node_lock(node);
- if (unlikely(!node->bclink.supported))
+ if (unlikely(!node->bclink.recv_permitted))
goto unlock;
/* Handle broadcast protocol message */
@@ -564,7 +564,7 @@ exit:
u32 tipc_bclink_acks_missing(struct tipc_node *n_ptr)
{
- return (n_ptr->bclink.supported &&
+ return (n_ptr->bclink.recv_permitted &&
(tipc_bclink_get_last_sent() != n_ptr->bclink.acked));
}
@@ -578,14 +578,12 @@ u32 tipc_bclink_acks_missing(struct tipc_node *n_ptr)
* Returns 0 (packet sent successfully) under all circumstances,
* since the broadcast link's pseudo-bearer never blocks
*/
-static int tipc_bcbearer_send(struct sk_buff *buf,
- struct tipc_bearer *unused1,
+static int tipc_bcbearer_send(struct sk_buff *buf, struct tipc_bearer *unused1,
struct tipc_media_addr *unused2)
{
int bp_index;
- /*
- * Prepare broadcast link message for reliable transmission,
+ /* Prepare broadcast link message for reliable transmission,
* if first time trying to send it;
* preparation is skipped for broadcast link protocol messages
* since they are sent in an unreliable manner and don't need it
@@ -611,32 +609,43 @@ static int tipc_bcbearer_send(struct sk_buff *buf,
for (bp_index = 0; bp_index < MAX_BEARERS; bp_index++) {
struct tipc_bearer *p = bcbearer->bpairs[bp_index].primary;
struct tipc_bearer *s = bcbearer->bpairs[bp_index].secondary;
+ struct tipc_bearer *b = p;
+ struct sk_buff *tbuf;
if (!p)
- break; /* no more bearers to try */
+ break; /* No more bearers to try */
+
+ if (tipc_bearer_blocked(p)) {
+ if (!s || tipc_bearer_blocked(s))
+ continue; /* Can't use either bearer */
+ b = s;
+ }
- tipc_nmap_diff(&bcbearer->remains, &p->nodes, &bcbearer->remains_new);
+ tipc_nmap_diff(&bcbearer->remains, &b->nodes,
+ &bcbearer->remains_new);
if (bcbearer->remains_new.count == bcbearer->remains.count)
- continue; /* bearer pair doesn't add anything */
-
- if (p->blocked ||
- p->media->send_msg(buf, p, &p->media->bcast_addr)) {
- /* unable to send on primary bearer */
- if (!s || s->blocked ||
- s->media->send_msg(buf, s,
- &s->media->bcast_addr)) {
- /* unable to send on either bearer */
- continue;
- }
+ continue; /* Nothing added by bearer pair */
+
+ if (bp_index == 0) {
+ /* Use original buffer for first bearer */
+ tipc_bearer_send(b, buf, &b->bcast_addr);
+ } else {
+ /* Avoid concurrent buffer access */
+ tbuf = pskb_copy(buf, GFP_ATOMIC);
+ if (!tbuf)
+ break;
+ tipc_bearer_send(b, tbuf, &b->bcast_addr);
+ kfree_skb(tbuf); /* Bearer keeps a clone */
}
+ /* Swap bearers for next packet */
if (s) {
bcbearer->bpairs[bp_index].primary = s;
bcbearer->bpairs[bp_index].secondary = p;
}
if (bcbearer->remains_new.count == 0)
- break; /* all targets reached */
+ break; /* All targets reached */
bcbearer->remains = bcbearer->remains_new;
}
@@ -731,8 +740,8 @@ int tipc_bclink_stats(char *buf, const u32 buf_size)
" TX naks:%u acks:%u dups:%u\n",
s->sent_nacks, s->sent_acks, s->retransmitted);
ret += tipc_snprintf(buf + ret, buf_size - ret,
- " Congestion bearer:%u link:%u Send queue max:%u avg:%u\n",
- s->bearer_congs, s->link_congs, s->max_queue_sz,
+ " Congestion link:%u Send queue max:%u avg:%u\n",
+ s->link_congs, s->max_queue_sz,
s->queue_sz_counts ?
(s->accu_queue_sz / s->queue_sz_counts) : 0);
@@ -766,7 +775,6 @@ int tipc_bclink_set_queue_limits(u32 limit)
void tipc_bclink_init(void)
{
- INIT_LIST_HEAD(&bcbearer->bearer.cong_links);
bcbearer->bearer.media = &bcbearer->media;
bcbearer->media.send_msg = tipc_bcbearer_send;
sprintf(bcbearer->media.name, "tipc-broadcast");
@@ -777,6 +785,7 @@ void tipc_bclink_init(void)
bcl->owner = &bclink->node;
bcl->max_pkt = MAX_PKT_DEFAULT_MCAST;
tipc_link_set_queue_limits(bcl, BCLINK_WIN_DEFAULT);
+ spin_lock_init(&bcbearer->bearer.lock);
bcl->b_ptr = &bcbearer->bearer;
bcl->state = WORKING_WORKING;
strlcpy(bcl->name, tipc_bclink_name, TIPC_MAX_LINK_NAME);
diff --git a/net/tipc/bcast.h b/net/tipc/bcast.h
index a93306557e00..6ee587b469fd 100644
--- a/net/tipc/bcast.h
+++ b/net/tipc/bcast.h
@@ -75,7 +75,8 @@ void tipc_nmap_remove(struct tipc_node_map *nm_ptr, u32 node);
/**
* tipc_nmap_equal - test for equality of node maps
*/
-static inline int tipc_nmap_equal(struct tipc_node_map *nm_a, struct tipc_node_map *nm_b)
+static inline int tipc_nmap_equal(struct tipc_node_map *nm_a,
+ struct tipc_node_map *nm_b)
{
return !memcmp(nm_a, nm_b, sizeof(*nm_a));
}
diff --git a/net/tipc/bearer.c b/net/tipc/bearer.c
index 09e71241265d..609c30c80816 100644
--- a/net/tipc/bearer.c
+++ b/net/tipc/bearer.c
@@ -39,7 +39,7 @@
#include "bearer.h"
#include "discover.h"
-#define MAX_ADDR_STR 32
+#define MAX_ADDR_STR 60
static struct tipc_media *media_list[MAX_MEDIA];
static u32 media_count;
@@ -49,21 +49,6 @@ struct tipc_bearer tipc_bearers[MAX_BEARERS];
static void bearer_disable(struct tipc_bearer *b_ptr);
/**
- * media_name_valid - validate media name
- *
- * Returns 1 if media name is valid, otherwise 0.
- */
-static int media_name_valid(const char *name)
-{
- u32 len;
-
- len = strlen(name);
- if ((len + 1) > TIPC_MAX_MEDIA_NAME)
- return 0;
- return strspn(name, tipc_alphabet) == len;
-}
-
-/**
* tipc_media_find - locates specified media object by name
*/
struct tipc_media *tipc_media_find(const char *name)
@@ -102,10 +87,7 @@ int tipc_register_media(struct tipc_media *m_ptr)
write_lock_bh(&tipc_net_lock);
- if (!media_name_valid(m_ptr->name))
- goto exit;
- if ((m_ptr->bcast_addr.media_id != m_ptr->type_id) ||
- !m_ptr->bcast_addr.broadcast)
+ if ((strlen(m_ptr->name) + 1) > TIPC_MAX_MEDIA_NAME)
goto exit;
if (m_ptr->priority > TIPC_MAX_LINK_PRI)
goto exit;
@@ -206,9 +188,7 @@ static int bearer_name_validate(const char *name,
/* validate component parts of bearer name */
if ((media_len <= 1) || (media_len > TIPC_MAX_MEDIA_NAME) ||
- (if_len <= 1) || (if_len > TIPC_MAX_IF_NAME) ||
- (strspn(media_name, tipc_alphabet) != (media_len - 1)) ||
- (strspn(if_name, tipc_alphabet) != (if_len - 1)))
+ (if_len <= 1) || (if_len > TIPC_MAX_IF_NAME))
return 0;
/* return bearer name components, if necessary */
@@ -296,116 +276,31 @@ void tipc_bearer_remove_dest(struct tipc_bearer *b_ptr, u32 dest)
}
/*
- * bearer_push(): Resolve bearer congestion. Force the waiting
- * links to push out their unsent packets, one packet per link
- * per iteration, until all packets are gone or congestion reoccurs.
- * 'tipc_net_lock' is read_locked when this function is called
- * bearer.lock must be taken before calling
- * Returns binary true(1) ore false(0)
- */
-static int bearer_push(struct tipc_bearer *b_ptr)
-{
- u32 res = 0;
- struct tipc_link *ln, *tln;
-
- if (b_ptr->blocked)
- return 0;
-
- while (!list_empty(&b_ptr->cong_links) && (res != PUSH_FAILED)) {
- list_for_each_entry_safe(ln, tln, &b_ptr->cong_links, link_list) {
- res = tipc_link_push_packet(ln);
- if (res == PUSH_FAILED)
- break;
- if (res == PUSH_FINISHED)
- list_move_tail(&ln->link_list, &b_ptr->links);
- }
- }
- return list_empty(&b_ptr->cong_links);
-}
-
-void tipc_bearer_lock_push(struct tipc_bearer *b_ptr)
-{
- spin_lock_bh(&b_ptr->lock);
- bearer_push(b_ptr);
- spin_unlock_bh(&b_ptr->lock);
-}
-
-
-/*
- * Interrupt enabling new requests after bearer congestion or blocking:
+ * Interrupt enabling new requests after bearer blocking:
* See bearer_send().
*/
-void tipc_continue(struct tipc_bearer *b_ptr)
-{
- spin_lock_bh(&b_ptr->lock);
- if (!list_empty(&b_ptr->cong_links))
- tipc_k_signal((Handler)tipc_bearer_lock_push, (unsigned long)b_ptr);
- b_ptr->blocked = 0;
- spin_unlock_bh(&b_ptr->lock);
-}
-
-/*
- * Schedule link for sending of messages after the bearer
- * has been deblocked by 'continue()'. This method is called
- * when somebody tries to send a message via this link while
- * the bearer is congested. 'tipc_net_lock' is in read_lock here
- * bearer.lock is busy
- */
-static void tipc_bearer_schedule_unlocked(struct tipc_bearer *b_ptr,
- struct tipc_link *l_ptr)
+void tipc_continue(struct tipc_bearer *b)
{
- list_move_tail(&l_ptr->link_list, &b_ptr->cong_links);
+ spin_lock_bh(&b->lock);
+ b->blocked = 0;
+ spin_unlock_bh(&b->lock);
}
/*
- * Schedule link for sending of messages after the bearer
- * has been deblocked by 'continue()'. This method is called
- * when somebody tries to send a message via this link while
- * the bearer is congested. 'tipc_net_lock' is in read_lock here,
- * bearer.lock is free
+ * tipc_bearer_blocked - determines if bearer is currently blocked
*/
-void tipc_bearer_schedule(struct tipc_bearer *b_ptr, struct tipc_link *l_ptr)
+int tipc_bearer_blocked(struct tipc_bearer *b)
{
- spin_lock_bh(&b_ptr->lock);
- tipc_bearer_schedule_unlocked(b_ptr, l_ptr);
- spin_unlock_bh(&b_ptr->lock);
-}
-
+ int res;
-/*
- * tipc_bearer_resolve_congestion(): Check if there is bearer congestion,
- * and if there is, try to resolve it before returning.
- * 'tipc_net_lock' is read_locked when this function is called
- */
-int tipc_bearer_resolve_congestion(struct tipc_bearer *b_ptr,
- struct tipc_link *l_ptr)
-{
- int res = 1;
+ spin_lock_bh(&b->lock);
+ res = b->blocked;
+ spin_unlock_bh(&b->lock);
- if (list_empty(&b_ptr->cong_links))
- return 1;
- spin_lock_bh(&b_ptr->lock);
- if (!bearer_push(b_ptr)) {
- tipc_bearer_schedule_unlocked(b_ptr, l_ptr);
- res = 0;
- }
- spin_unlock_bh(&b_ptr->lock);
return res;
}
/**
- * tipc_bearer_congested - determines if bearer is currently congested
- */
-int tipc_bearer_congested(struct tipc_bearer *b_ptr, struct tipc_link *l_ptr)
-{
- if (unlikely(b_ptr->blocked))
- return 1;
- if (likely(list_empty(&b_ptr->cong_links)))
- return 0;
- return !tipc_bearer_resolve_congestion(b_ptr, l_ptr);
-}
-
-/**
* tipc_enable_bearer - enable bearer with the given name
*/
int tipc_enable_bearer(const char *name, u32 disc_domain, u32 priority)
@@ -506,11 +401,10 @@ restart:
b_ptr->net_plane = bearer_id + 'A';
b_ptr->active = 1;
b_ptr->priority = priority;
- INIT_LIST_HEAD(&b_ptr->cong_links);
INIT_LIST_HEAD(&b_ptr->links);
spin_lock_init(&b_ptr->lock);
- res = tipc_disc_create(b_ptr, &m_ptr->bcast_addr, disc_domain);
+ res = tipc_disc_create(b_ptr, &b_ptr->bcast_addr, disc_domain);
if (res) {
bearer_disable(b_ptr);
pr_warn("Bearer <%s> rejected, discovery object creation failed\n",
@@ -545,7 +439,6 @@ int tipc_block_bearer(const char *name)
pr_info("Blocking bearer <%s>\n", name);
spin_lock_bh(&b_ptr->lock);
b_ptr->blocked = 1;
- list_splice_init(&b_ptr->cong_links, &b_ptr->links);
list_for_each_entry_safe(l_ptr, temp_l_ptr, &b_ptr->links, link_list) {
struct tipc_node *n_ptr = l_ptr->owner;
@@ -567,18 +460,22 @@ static void bearer_disable(struct tipc_bearer *b_ptr)
{
struct tipc_link *l_ptr;
struct tipc_link *temp_l_ptr;
+ struct tipc_link_req *temp_req;
pr_info("Disabling bearer <%s>\n", b_ptr->name);
spin_lock_bh(&b_ptr->lock);
b_ptr->blocked = 1;
b_ptr->media->disable_bearer(b_ptr);
- list_splice_init(&b_ptr->cong_links, &b_ptr->links);
list_for_each_entry_safe(l_ptr, temp_l_ptr, &b_ptr->links, link_list) {
tipc_link_delete(l_ptr);
}
- if (b_ptr->link_req)
- tipc_disc_delete(b_ptr->link_req);
+ temp_req = b_ptr->link_req;
+ b_ptr->link_req = NULL;
spin_unlock_bh(&b_ptr->lock);
+
+ if (temp_req)
+ tipc_disc_delete(temp_req);
+
memset(b_ptr, 0, sizeof(struct tipc_bearer));
}
diff --git a/net/tipc/bearer.h b/net/tipc/bearer.h
index dd4c2abf08e7..09c869adcfcf 100644
--- a/net/tipc/bearer.h
+++ b/net/tipc/bearer.h
@@ -56,6 +56,7 @@
* Identifiers of supported TIPC media types
*/
#define TIPC_MEDIA_TYPE_ETH 1
+#define TIPC_MEDIA_TYPE_IB 2
/**
* struct tipc_media_addr - destination address used by TIPC bearers
@@ -77,7 +78,6 @@ struct tipc_bearer;
* @enable_bearer: routine which enables a bearer
* @disable_bearer: routine which disables a bearer
* @addr2str: routine which converts media address to string
- * @str2addr: routine which converts media address from string
* @addr2msg: routine which converts media address to protocol message area
* @msg2addr: routine which converts media address from protocol message area
* @bcast_addr: media address used in broadcasting
@@ -94,10 +94,9 @@ struct tipc_media {
int (*enable_bearer)(struct tipc_bearer *b_ptr);
void (*disable_bearer)(struct tipc_bearer *b_ptr);
int (*addr2str)(struct tipc_media_addr *a, char *str_buf, int str_size);
- int (*str2addr)(struct tipc_media_addr *a, char *str_buf);
int (*addr2msg)(struct tipc_media_addr *a, char *msg_area);
- int (*msg2addr)(struct tipc_media_addr *a, char *msg_area);
- struct tipc_media_addr bcast_addr;
+ int (*msg2addr)(const struct tipc_bearer *b_ptr,
+ struct tipc_media_addr *a, char *msg_area);
u32 priority;
u32 tolerance;
u32 window;
@@ -120,7 +119,6 @@ struct tipc_media {
* @identity: array index of this bearer within TIPC bearer array
* @link_req: ptr to (optional) structure making periodic link setup requests
* @links: list of non-congested links associated with bearer
- * @cong_links: list of congested links associated with bearer
* @active: non-zero if bearer structure is represents a bearer
* @net_plane: network plane ('A' through 'H') currently associated with bearer
* @nodes: indicates which nodes in cluster can be reached through bearer
@@ -137,13 +135,13 @@ struct tipc_bearer {
char name[TIPC_MAX_BEARER_NAME];
spinlock_t lock;
struct tipc_media *media;
+ struct tipc_media_addr bcast_addr;
u32 priority;
u32 window;
u32 tolerance;
u32 identity;
struct tipc_link_req *link_req;
struct list_head links;
- struct list_head cong_links;
int active;
char net_plane;
struct tipc_node_map nodes;
@@ -177,6 +175,14 @@ int tipc_disable_bearer(const char *name);
int tipc_eth_media_start(void);
void tipc_eth_media_stop(void);
+#ifdef CONFIG_TIPC_MEDIA_IB
+int tipc_ib_media_start(void);
+void tipc_ib_media_stop(void);
+#else
+static inline int tipc_ib_media_start(void) { return 0; }
+static inline void tipc_ib_media_stop(void) { return; }
+#endif
+
int tipc_media_set_priority(const char *name, u32 new_value);
int tipc_media_set_window(const char *name, u32 new_value);
void tipc_media_addr_printf(char *buf, int len, struct tipc_media_addr *a);
@@ -185,39 +191,23 @@ struct sk_buff *tipc_media_get_names(void);
struct sk_buff *tipc_bearer_get_names(void);
void tipc_bearer_add_dest(struct tipc_bearer *b_ptr, u32 dest);
void tipc_bearer_remove_dest(struct tipc_bearer *b_ptr, u32 dest);
-void tipc_bearer_schedule(struct tipc_bearer *b_ptr, struct tipc_link *l_ptr);
struct tipc_bearer *tipc_bearer_find(const char *name);
struct tipc_bearer *tipc_bearer_find_interface(const char *if_name);
struct tipc_media *tipc_media_find(const char *name);
-int tipc_bearer_resolve_congestion(struct tipc_bearer *b_ptr,
- struct tipc_link *l_ptr);
-int tipc_bearer_congested(struct tipc_bearer *b_ptr, struct tipc_link *l_ptr);
+int tipc_bearer_blocked(struct tipc_bearer *b_ptr);
void tipc_bearer_stop(void);
-void tipc_bearer_lock_push(struct tipc_bearer *b_ptr);
-
/**
* tipc_bearer_send- sends buffer to destination over bearer
*
- * Returns true (1) if successful, or false (0) if unable to send
- *
* IMPORTANT:
* The media send routine must not alter the buffer being passed in
* as it may be needed for later retransmission!
- *
- * If the media send routine returns a non-zero value (indicating that
- * it was unable to send the buffer), it must:
- * 1) mark the bearer as blocked,
- * 2) call tipc_continue() once the bearer is able to send again.
- * Media types that are unable to meet these two critera must ensure their
- * send routine always returns success -- even if the buffer was not sent --
- * and let TIPC's link code deal with the undelivered message.
*/
-static inline int tipc_bearer_send(struct tipc_bearer *b_ptr,
- struct sk_buff *buf,
+static inline void tipc_bearer_send(struct tipc_bearer *b, struct sk_buff *buf,
struct tipc_media_addr *dest)
{
- return !b_ptr->media->send_msg(buf, b_ptr, dest);
+ b->media->send_msg(buf, b, dest);
}
#endif /* _TIPC_BEARER_H */
diff --git a/net/tipc/config.c b/net/tipc/config.c
index a056a3852f71..c301a9a592d8 100644
--- a/net/tipc/config.c
+++ b/net/tipc/config.c
@@ -2,7 +2,7 @@
* net/tipc/config.c: TIPC configuration management code
*
* Copyright (c) 2002-2006, Ericsson AB
- * Copyright (c) 2004-2007, 2010-2011, Wind River Systems
+ * Copyright (c) 2004-2007, 2010-2013, Wind River Systems
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
@@ -38,12 +38,12 @@
#include "port.h"
#include "name_table.h"
#include "config.h"
+#include "server.h"
#define REPLY_TRUNCATED "<truncated>\n"
-static u32 config_port_ref;
-
-static DEFINE_SPINLOCK(config_lock);
+static DEFINE_MUTEX(config_mutex);
+static struct tipc_server cfgsrv;
static const void *req_tlv_area; /* request message TLV area */
static int req_tlv_space; /* request message TLV area size */
@@ -181,18 +181,7 @@ static struct sk_buff *cfg_set_own_addr(void)
if (tipc_own_addr)
return tipc_cfg_reply_error_string(TIPC_CFG_NOT_SUPPORTED
" (cannot change node address once assigned)");
-
- /*
- * Must temporarily release configuration spinlock while switching into
- * networking mode as it calls tipc_eth_media_start(), which may sleep.
- * Releasing the lock is harmless as other locally-issued configuration
- * commands won't occur until this one completes, and remotely-issued
- * configuration commands can't be received until a local configuration
- * command to enable the first bearer is received and processed.
- */
- spin_unlock_bh(&config_lock);
tipc_core_start_net(addr);
- spin_lock_bh(&config_lock);
return tipc_cfg_reply_none();
}
@@ -208,36 +197,6 @@ static struct sk_buff *cfg_set_remote_mng(void)
return tipc_cfg_reply_none();
}
-static struct sk_buff *cfg_set_max_publications(void)
-{
- u32 value;
-
- if (!TLV_CHECK(req_tlv_area, req_tlv_space, TIPC_TLV_UNSIGNED))
- return tipc_cfg_reply_error_string(TIPC_CFG_TLV_ERROR);
-
- value = ntohl(*(__be32 *)TLV_DATA(req_tlv_area));
- if (value < 1 || value > 65535)
- return tipc_cfg_reply_error_string(TIPC_CFG_INVALID_VALUE
- " (max publications must be 1-65535)");
- tipc_max_publications = value;
- return tipc_cfg_reply_none();
-}
-
-static struct sk_buff *cfg_set_max_subscriptions(void)
-{
- u32 value;
-
- if (!TLV_CHECK(req_tlv_area, req_tlv_space, TIPC_TLV_UNSIGNED))
- return tipc_cfg_reply_error_string(TIPC_CFG_TLV_ERROR);
-
- value = ntohl(*(__be32 *)TLV_DATA(req_tlv_area));
- if (value < 1 || value > 65535)
- return tipc_cfg_reply_error_string(TIPC_CFG_INVALID_VALUE
- " (max subscriptions must be 1-65535");
- tipc_max_subscriptions = value;
- return tipc_cfg_reply_none();
-}
-
static struct sk_buff *cfg_set_max_ports(void)
{
u32 value;
@@ -278,7 +237,7 @@ struct sk_buff *tipc_cfg_do_cmd(u32 orig_node, u16 cmd, const void *request_area
{
struct sk_buff *rep_tlv_buf;
- spin_lock_bh(&config_lock);
+ mutex_lock(&config_mutex);
/* Save request and reply details in a well-known location */
req_tlv_area = request_area;
@@ -357,12 +316,6 @@ struct sk_buff *tipc_cfg_do_cmd(u32 orig_node, u16 cmd, const void *request_area
case TIPC_CMD_SET_MAX_PORTS:
rep_tlv_buf = cfg_set_max_ports();
break;
- case TIPC_CMD_SET_MAX_PUBL:
- rep_tlv_buf = cfg_set_max_publications();
- break;
- case TIPC_CMD_SET_MAX_SUBSCR:
- rep_tlv_buf = cfg_set_max_subscriptions();
- break;
case TIPC_CMD_SET_NETID:
rep_tlv_buf = cfg_set_netid();
break;
@@ -372,12 +325,6 @@ struct sk_buff *tipc_cfg_do_cmd(u32 orig_node, u16 cmd, const void *request_area
case TIPC_CMD_GET_MAX_PORTS:
rep_tlv_buf = tipc_cfg_reply_unsigned(tipc_max_ports);
break;
- case TIPC_CMD_GET_MAX_PUBL:
- rep_tlv_buf = tipc_cfg_reply_unsigned(tipc_max_publications);
- break;
- case TIPC_CMD_GET_MAX_SUBSCR:
- rep_tlv_buf = tipc_cfg_reply_unsigned(tipc_max_subscriptions);
- break;
case TIPC_CMD_GET_NETID:
rep_tlv_buf = tipc_cfg_reply_unsigned(tipc_net_id);
break;
@@ -393,6 +340,10 @@ struct sk_buff *tipc_cfg_do_cmd(u32 orig_node, u16 cmd, const void *request_area
case TIPC_CMD_GET_MAX_CLUSTERS:
case TIPC_CMD_SET_MAX_NODES:
case TIPC_CMD_GET_MAX_NODES:
+ case TIPC_CMD_SET_MAX_SUBSCR:
+ case TIPC_CMD_GET_MAX_SUBSCR:
+ case TIPC_CMD_SET_MAX_PUBL:
+ case TIPC_CMD_GET_MAX_PUBL:
case TIPC_CMD_SET_LOG_SIZE:
case TIPC_CMD_DUMP_LOG:
rep_tlv_buf = tipc_cfg_reply_error_string(TIPC_CFG_NOT_SUPPORTED
@@ -415,37 +366,31 @@ struct sk_buff *tipc_cfg_do_cmd(u32 orig_node, u16 cmd, const void *request_area
/* Return reply buffer */
exit:
- spin_unlock_bh(&config_lock);
+ mutex_unlock(&config_mutex);
return rep_tlv_buf;
}
-static void cfg_named_msg_event(void *userdata,
- u32 port_ref,
- struct sk_buff **buf,
- const unchar *msg,
- u32 size,
- u32 importance,
- struct tipc_portid const *orig,
- struct tipc_name_seq const *dest)
+static void cfg_conn_msg_event(int conid, struct sockaddr_tipc *addr,
+ void *usr_data, void *buf, size_t len)
{
struct tipc_cfg_msg_hdr *req_hdr;
struct tipc_cfg_msg_hdr *rep_hdr;
struct sk_buff *rep_buf;
+ int ret;
/* Validate configuration message header (ignore invalid message) */
- req_hdr = (struct tipc_cfg_msg_hdr *)msg;
- if ((size < sizeof(*req_hdr)) ||
- (size != TCM_ALIGN(ntohl(req_hdr->tcm_len))) ||
+ req_hdr = (struct tipc_cfg_msg_hdr *)buf;
+ if ((len < sizeof(*req_hdr)) ||
+ (len != TCM_ALIGN(ntohl(req_hdr->tcm_len))) ||
(ntohs(req_hdr->tcm_flags) != TCM_F_REQUEST)) {
pr_warn("Invalid configuration message discarded\n");
return;
}
/* Generate reply for request (if can't, return request) */
- rep_buf = tipc_cfg_do_cmd(orig->node,
- ntohs(req_hdr->tcm_type),
- msg + sizeof(*req_hdr),
- size - sizeof(*req_hdr),
+ rep_buf = tipc_cfg_do_cmd(addr->addr.id.node, ntohs(req_hdr->tcm_type),
+ buf + sizeof(*req_hdr),
+ len - sizeof(*req_hdr),
BUF_HEADROOM + MAX_H_SIZE + sizeof(*rep_hdr));
if (rep_buf) {
skb_push(rep_buf, sizeof(*rep_hdr));
@@ -453,57 +398,51 @@ static void cfg_named_msg_event(void *userdata,
memcpy(rep_hdr, req_hdr, sizeof(*rep_hdr));
rep_hdr->tcm_len = htonl(rep_buf->len);
rep_hdr->tcm_flags &= htons(~TCM_F_REQUEST);
- } else {
- rep_buf = *buf;
- *buf = NULL;
- }
- /* NEED TO ADD CODE TO HANDLE FAILED SEND (SUCH AS CONGESTION) */
- tipc_send_buf2port(port_ref, orig, rep_buf, rep_buf->len);
+ ret = tipc_conn_sendmsg(&cfgsrv, conid, addr, rep_buf->data,
+ rep_buf->len);
+ if (ret < 0)
+ pr_err("Sending cfg reply message failed, no memory\n");
+
+ kfree_skb(rep_buf);
+ }
}
+static struct sockaddr_tipc cfgsrv_addr __read_mostly = {
+ .family = AF_TIPC,
+ .addrtype = TIPC_ADDR_NAMESEQ,
+ .addr.nameseq.type = TIPC_CFG_SRV,
+ .addr.nameseq.lower = 0,
+ .addr.nameseq.upper = 0,
+ .scope = TIPC_ZONE_SCOPE
+};
+
+static struct tipc_server cfgsrv __read_mostly = {
+ .saddr = &cfgsrv_addr,
+ .imp = TIPC_CRITICAL_IMPORTANCE,
+ .type = SOCK_RDM,
+ .max_rcvbuf_size = 64 * 1024,
+ .name = "cfg_server",
+ .tipc_conn_recvmsg = cfg_conn_msg_event,
+ .tipc_conn_new = NULL,
+ .tipc_conn_shutdown = NULL
+};
+
int tipc_cfg_init(void)
{
- struct tipc_name_seq seq;
- int res;
-
- res = tipc_createport(NULL, TIPC_CRITICAL_IMPORTANCE,
- NULL, NULL, NULL,
- NULL, cfg_named_msg_event, NULL,
- NULL, &config_port_ref);
- if (res)
- goto failed;
-
- seq.type = TIPC_CFG_SRV;
- seq.lower = seq.upper = tipc_own_addr;
- res = tipc_publish(config_port_ref, TIPC_ZONE_SCOPE, &seq);
- if (res)
- goto failed;
-
- return 0;
-
-failed:
- pr_err("Unable to create configuration service\n");
- return res;
+ return tipc_server_start(&cfgsrv);
}
void tipc_cfg_reinit(void)
{
- struct tipc_name_seq seq;
- int res;
-
- seq.type = TIPC_CFG_SRV;
- seq.lower = seq.upper = 0;
- tipc_withdraw(config_port_ref, TIPC_ZONE_SCOPE, &seq);
+ tipc_server_stop(&cfgsrv);
- seq.lower = seq.upper = tipc_own_addr;
- res = tipc_publish(config_port_ref, TIPC_ZONE_SCOPE, &seq);
- if (res)
- pr_err("Unable to reinitialize configuration service\n");
+ cfgsrv_addr.addr.nameseq.lower = tipc_own_addr;
+ cfgsrv_addr.addr.nameseq.upper = tipc_own_addr;
+ tipc_server_start(&cfgsrv);
}
void tipc_cfg_stop(void)
{
- tipc_deleteport(config_port_ref);
- config_port_ref = 0;
+ tipc_server_stop(&cfgsrv);
}
diff --git a/net/tipc/core.c b/net/tipc/core.c
index 6586eac6a50e..fd4eeeaa972a 100644
--- a/net/tipc/core.c
+++ b/net/tipc/core.c
@@ -2,7 +2,7 @@
* net/tipc/core.c: TIPC module code
*
* Copyright (c) 2003-2006, Ericsson AB
- * Copyright (c) 2005-2006, 2010-2011, Wind River Systems
+ * Copyright (c) 2005-2006, 2010-2013, Wind River Systems
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
@@ -39,28 +39,19 @@
#include "name_table.h"
#include "subscr.h"
#include "config.h"
+#include "port.h"
#include <linux/module.h>
-#ifndef CONFIG_TIPC_PORTS
-#define CONFIG_TIPC_PORTS 8191
-#endif
-
-
/* global variables used by multiple sub-systems within TIPC */
-int tipc_random;
-
-const char tipc_alphabet[] =
- "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789_.";
+int tipc_random __read_mostly;
/* configurable TIPC parameters */
-u32 tipc_own_addr;
-int tipc_max_ports;
-int tipc_max_subscriptions;
-int tipc_max_publications;
-int tipc_net_id;
-int tipc_remote_management;
-
+u32 tipc_own_addr __read_mostly;
+int tipc_max_ports __read_mostly;
+int tipc_net_id __read_mostly;
+int tipc_remote_management __read_mostly;
+int sysctl_tipc_rmem[3] __read_mostly; /* min/default/max */
/**
* tipc_buf_acquire - creates a TIPC message buffer
@@ -92,6 +83,7 @@ static void tipc_core_stop_net(void)
{
tipc_net_stop();
tipc_eth_media_stop();
+ tipc_ib_media_stop();
}
/**
@@ -101,11 +93,17 @@ int tipc_core_start_net(unsigned long addr)
{
int res;
- res = tipc_net_start(addr);
- if (!res)
- res = tipc_eth_media_start();
- if (res)
- tipc_core_stop_net();
+ tipc_net_start(addr);
+ res = tipc_eth_media_start();
+ if (res < 0)
+ goto err;
+ res = tipc_ib_media_start();
+ if (res < 0)
+ goto err;
+ return res;
+
+err:
+ tipc_core_stop_net();
return res;
}
@@ -121,6 +119,7 @@ static void tipc_core_stop(void)
tipc_nametbl_stop();
tipc_ref_table_stop();
tipc_socket_stop();
+ tipc_unregister_sysctl();
}
/**
@@ -138,20 +137,21 @@ static int tipc_core_start(void)
if (!res)
res = tipc_nametbl_init();
if (!res)
- res = tipc_subscr_start();
- if (!res)
- res = tipc_cfg_init();
- if (!res)
res = tipc_netlink_start();
if (!res)
res = tipc_socket_init();
+ if (!res)
+ res = tipc_register_sysctl();
+ if (!res)
+ res = tipc_subscr_start();
+ if (!res)
+ res = tipc_cfg_init();
if (res)
tipc_core_stop();
return res;
}
-
static int __init tipc_init(void)
{
int res;
@@ -160,11 +160,14 @@ static int __init tipc_init(void)
tipc_own_addr = 0;
tipc_remote_management = 1;
- tipc_max_publications = 10000;
- tipc_max_subscriptions = 2000;
tipc_max_ports = CONFIG_TIPC_PORTS;
tipc_net_id = 4711;
+ sysctl_tipc_rmem[0] = CONN_OVERLOAD_LIMIT >> 4 << TIPC_LOW_IMPORTANCE;
+ sysctl_tipc_rmem[1] = CONN_OVERLOAD_LIMIT >> 4 <<
+ TIPC_CRITICAL_IMPORTANCE;
+ sysctl_tipc_rmem[2] = CONN_OVERLOAD_LIMIT;
+
res = tipc_core_start();
if (res)
pr_err("Unable to start in single node mode\n");
diff --git a/net/tipc/core.h b/net/tipc/core.h
index fd42e106c185..be72f8cebc53 100644
--- a/net/tipc/core.h
+++ b/net/tipc/core.h
@@ -1,8 +1,8 @@
/*
* net/tipc/core.h: Include file for TIPC global declarations
*
- * Copyright (c) 2005-2006, Ericsson AB
- * Copyright (c) 2005-2007, 2010-2011, Wind River Systems
+ * Copyright (c) 2005-2006, 2013 Ericsson AB
+ * Copyright (c) 2005-2007, 2010-2013, Wind River Systems
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
@@ -60,7 +60,9 @@
#define TIPC_MOD_VER "2.0.0"
-#define ULTRA_STRING_MAX_LEN 32768
+#define ULTRA_STRING_MAX_LEN 32768
+#define TIPC_MAX_SUBSCRIPTIONS 65535
+#define TIPC_MAX_PUBLICATIONS 65535
struct tipc_msg; /* msg.h */
@@ -74,19 +76,16 @@ int tipc_snprintf(char *buf, int len, const char *fmt, ...);
/*
* Global configuration variables
*/
-extern u32 tipc_own_addr;
-extern int tipc_max_ports;
-extern int tipc_max_subscriptions;
-extern int tipc_max_publications;
-extern int tipc_net_id;
-extern int tipc_remote_management;
+extern u32 tipc_own_addr __read_mostly;
+extern int tipc_max_ports __read_mostly;
+extern int tipc_net_id __read_mostly;
+extern int tipc_remote_management __read_mostly;
+extern int sysctl_tipc_rmem[3] __read_mostly;
/*
* Other global variables
*/
-extern int tipc_random;
-extern const char tipc_alphabet[];
-
+extern int tipc_random __read_mostly;
/*
* Routines available to privileged subsystems
@@ -98,6 +97,18 @@ extern int tipc_netlink_start(void);
extern void tipc_netlink_stop(void);
extern int tipc_socket_init(void);
extern void tipc_socket_stop(void);
+extern int tipc_sock_create_local(int type, struct socket **res);
+extern void tipc_sock_release_local(struct socket *sock);
+extern int tipc_sock_accept_local(struct socket *sock,
+ struct socket **newsock, int flags);
+
+#ifdef CONFIG_SYSCTL
+extern int tipc_register_sysctl(void);
+extern void tipc_unregister_sysctl(void);
+#else
+#define tipc_register_sysctl() 0
+#define tipc_unregister_sysctl()
+#endif
/*
* TIPC timer and signal code
diff --git a/net/tipc/discover.c b/net/tipc/discover.c
index 50eaa403eb6e..ecc758c6eacf 100644
--- a/net/tipc/discover.c
+++ b/net/tipc/discover.c
@@ -70,8 +70,7 @@ struct tipc_link_req {
* @dest_domain: network domain of node(s) which should respond to message
* @b_ptr: ptr to bearer issuing message
*/
-static struct sk_buff *tipc_disc_init_msg(u32 type,
- u32 dest_domain,
+static struct sk_buff *tipc_disc_init_msg(u32 type, u32 dest_domain,
struct tipc_bearer *b_ptr)
{
struct sk_buff *buf = tipc_buf_acquire(INT_H_SIZE);
@@ -129,7 +128,7 @@ void tipc_disc_recv_msg(struct sk_buff *buf, struct tipc_bearer *b_ptr)
int link_fully_up;
media_addr.broadcast = 1;
- b_ptr->media->msg2addr(&media_addr, msg_media_addr(msg));
+ b_ptr->media->msg2addr(b_ptr, &media_addr, msg_media_addr(msg));
kfree_skb(buf);
/* Ensure message from node is valid and communication is permitted */
@@ -243,7 +242,7 @@ void tipc_disc_recv_msg(struct sk_buff *buf, struct tipc_bearer *b_ptr)
if ((type == DSC_REQ_MSG) && !link_fully_up && !b_ptr->blocked) {
rbuf = tipc_disc_init_msg(DSC_RESP_MSG, orig, b_ptr);
if (rbuf) {
- b_ptr->media->send_msg(rbuf, b_ptr, &media_addr);
+ tipc_bearer_send(b_ptr, rbuf, &media_addr);
kfree_skb(rbuf);
}
}
@@ -346,8 +345,8 @@ exit:
*
* Returns 0 if successful, otherwise -errno.
*/
-int tipc_disc_create(struct tipc_bearer *b_ptr,
- struct tipc_media_addr *dest, u32 dest_domain)
+int tipc_disc_create(struct tipc_bearer *b_ptr, struct tipc_media_addr *dest,
+ u32 dest_domain)
{
struct tipc_link_req *req;
diff --git a/net/tipc/eth_media.c b/net/tipc/eth_media.c
index 90ac9bfa7abb..40ea40cf6204 100644
--- a/net/tipc/eth_media.c
+++ b/net/tipc/eth_media.c
@@ -46,19 +46,30 @@
* @bearer: ptr to associated "generic" bearer structure
* @dev: ptr to associated Ethernet network device
* @tipc_packet_type: used in binding TIPC to Ethernet driver
+ * @setup: work item used when enabling bearer
* @cleanup: work item used when disabling bearer
*/
struct eth_bearer {
struct tipc_bearer *bearer;
struct net_device *dev;
struct packet_type tipc_packet_type;
+ struct work_struct setup;
struct work_struct cleanup;
};
static struct tipc_media eth_media_info;
static struct eth_bearer eth_bearers[MAX_ETH_BEARERS];
static int eth_started;
-static struct notifier_block notifier;
+
+static int recv_notification(struct notifier_block *nb, unsigned long evt,
+ void *dv);
+/*
+ * Network device notifier info
+ */
+static struct notifier_block notifier = {
+ .notifier_call = recv_notification,
+ .priority = 0
+};
/**
* eth_media_addr_set - initialize Ethernet media address structure
@@ -66,12 +77,13 @@ static struct notifier_block notifier;
* Media-dependent "value" field stores MAC address in first 6 bytes
* and zeroes out the remaining bytes.
*/
-static void eth_media_addr_set(struct tipc_media_addr *a, char *mac)
+static void eth_media_addr_set(const struct tipc_bearer *tb_ptr,
+ struct tipc_media_addr *a, char *mac)
{
memcpy(a->value, mac, ETH_ALEN);
memset(a->value + ETH_ALEN, 0, sizeof(a->value) - ETH_ALEN);
a->media_id = TIPC_MEDIA_TYPE_ETH;
- a->broadcast = !memcmp(mac, eth_media_info.bcast_addr.value, ETH_ALEN);
+ a->broadcast = !memcmp(mac, tb_ptr->bcast_addr.value, ETH_ALEN);
}
/**
@@ -99,6 +111,7 @@ static int send_msg(struct sk_buff *buf, struct tipc_bearer *tb_ptr,
skb_reset_network_header(clone);
clone->dev = dev;
+ clone->protocol = htons(ETH_P_TIPC);
dev_hard_header(clone, dev, ETH_P_TIPC, dest->value,
dev->dev_addr, clone->len);
dev_queue_xmit(clone);
@@ -134,12 +147,22 @@ static int recv_msg(struct sk_buff *buf, struct net_device *dev,
}
/**
+ * setup_bearer - setup association between Ethernet bearer and interface
+ */
+static void setup_bearer(struct work_struct *work)
+{
+ struct eth_bearer *eb_ptr =
+ container_of(work, struct eth_bearer, setup);
+
+ dev_add_pack(&eb_ptr->tipc_packet_type);
+}
+
+/**
* enable_bearer - attach TIPC bearer to an Ethernet interface
*/
static int enable_bearer(struct tipc_bearer *tb_ptr)
{
- struct net_device *dev = NULL;
- struct net_device *pdev = NULL;
+ struct net_device *dev;
struct eth_bearer *eb_ptr = &eth_bearers[0];
struct eth_bearer *stop = &eth_bearers[MAX_ETH_BEARERS];
char *driver_name = strchr((const char *)tb_ptr->name, ':') + 1;
@@ -154,15 +177,7 @@ static int enable_bearer(struct tipc_bearer *tb_ptr)
}
/* Find device with specified name */
- read_lock(&dev_base_lock);
- for_each_netdev(&init_net, pdev) {
- if (!strncmp(pdev->name, driver_name, IFNAMSIZ)) {
- dev = pdev;
- dev_hold(dev);
- break;
- }
- }
- read_unlock(&dev_base_lock);
+ dev = dev_get_by_name(&init_net, driver_name);
if (!dev)
return -ENODEV;
@@ -173,14 +188,19 @@ static int enable_bearer(struct tipc_bearer *tb_ptr)
eb_ptr->tipc_packet_type.func = recv_msg;
eb_ptr->tipc_packet_type.af_packet_priv = eb_ptr;
INIT_LIST_HEAD(&(eb_ptr->tipc_packet_type.list));
- dev_add_pack(&eb_ptr->tipc_packet_type);
+ INIT_WORK(&eb_ptr->setup, setup_bearer);
+ schedule_work(&eb_ptr->setup);
/* Associate TIPC bearer with Ethernet bearer */
eb_ptr->bearer = tb_ptr;
tb_ptr->usr_handle = (void *)eb_ptr;
+ memset(tb_ptr->bcast_addr.value, 0, sizeof(tb_ptr->bcast_addr.value));
+ memcpy(tb_ptr->bcast_addr.value, dev->broadcast, ETH_ALEN);
+ tb_ptr->bcast_addr.media_id = TIPC_MEDIA_TYPE_ETH;
+ tb_ptr->bcast_addr.broadcast = 1;
tb_ptr->mtu = dev->mtu;
tb_ptr->blocked = 0;
- eth_media_addr_set(&tb_ptr->addr, (char *)dev->dev_addr);
+ eth_media_addr_set(tb_ptr, &tb_ptr->addr, (char *)dev->dev_addr);
return 0;
}
@@ -222,9 +242,9 @@ static void disable_bearer(struct tipc_bearer *tb_ptr)
* specified device.
*/
static int recv_notification(struct notifier_block *nb, unsigned long evt,
- void *dv)
+ void *ptr)
{
- struct net_device *dev = (struct net_device *)dv;
+ struct net_device *dev = netdev_notifier_info_to_dev(ptr);
struct eth_bearer *eb_ptr = &eth_bearers[0];
struct eth_bearer *stop = &eth_bearers[MAX_ETH_BEARERS];
@@ -279,25 +299,6 @@ static int eth_addr2str(struct tipc_media_addr *a, char *str_buf, int str_size)
}
/**
- * eth_str2addr - convert string to Ethernet address
- */
-static int eth_str2addr(struct tipc_media_addr *a, char *str_buf)
-{
- char mac[ETH_ALEN];
- int r;
-
- r = sscanf(str_buf, "%02x:%02x:%02x:%02x:%02x:%02x",
- (u32 *)&mac[0], (u32 *)&mac[1], (u32 *)&mac[2],
- (u32 *)&mac[3], (u32 *)&mac[4], (u32 *)&mac[5]);
-
- if (r != ETH_ALEN)
- return 1;
-
- eth_media_addr_set(a, mac);
- return 0;
-}
-
-/**
* eth_str2addr - convert Ethernet address format to message header format
*/
static int eth_addr2msg(struct tipc_media_addr *a, char *msg_area)
@@ -311,12 +312,13 @@ static int eth_addr2msg(struct tipc_media_addr *a, char *msg_area)
/**
* eth_str2addr - convert message header address format to Ethernet format
*/
-static int eth_msg2addr(struct tipc_media_addr *a, char *msg_area)
+static int eth_msg2addr(const struct tipc_bearer *tb_ptr,
+ struct tipc_media_addr *a, char *msg_area)
{
if (msg_area[TIPC_MEDIA_TYPE_OFFSET] != TIPC_MEDIA_TYPE_ETH)
return 1;
- eth_media_addr_set(a, msg_area + ETH_ADDR_OFFSET);
+ eth_media_addr_set(tb_ptr, a, msg_area + ETH_ADDR_OFFSET);
return 0;
}
@@ -328,11 +330,8 @@ static struct tipc_media eth_media_info = {
.enable_bearer = enable_bearer,
.disable_bearer = disable_bearer,
.addr2str = eth_addr2str,
- .str2addr = eth_str2addr,
.addr2msg = eth_addr2msg,
.msg2addr = eth_msg2addr,
- .bcast_addr = { { 0xff, 0xff, 0xff, 0xff, 0xff, 0xff },
- TIPC_MEDIA_TYPE_ETH, 1 },
.priority = TIPC_DEF_LINK_PRI,
.tolerance = TIPC_DEF_LINK_TOL,
.window = TIPC_DEF_LINK_WIN,
@@ -357,8 +356,6 @@ int tipc_eth_media_start(void)
if (res)
return res;
- notifier.notifier_call = &recv_notification;
- notifier.priority = 0;
res = register_netdevice_notifier(&notifier);
if (!res)
eth_started = 1;
diff --git a/net/tipc/handler.c b/net/tipc/handler.c
index 7a52d3922f3c..b36f0fcd9bdf 100644
--- a/net/tipc/handler.c
+++ b/net/tipc/handler.c
@@ -45,7 +45,7 @@ struct queue_item {
static struct kmem_cache *tipc_queue_item_cache;
static struct list_head signal_queue_head;
static DEFINE_SPINLOCK(qitem_lock);
-static int handler_enabled;
+static int handler_enabled __read_mostly;
static void process_signal_queue(unsigned long dummy);
@@ -116,7 +116,6 @@ void tipc_handler_stop(void)
return;
handler_enabled = 0;
- tasklet_disable(&tipc_tasklet);
tasklet_kill(&tipc_tasklet);
spin_lock_bh(&qitem_lock);
diff --git a/net/tipc/ib_media.c b/net/tipc/ib_media.c
new file mode 100644
index 000000000000..9934a32bfa87
--- /dev/null
+++ b/net/tipc/ib_media.c
@@ -0,0 +1,372 @@
+/*
+ * net/tipc/ib_media.c: Infiniband bearer support for TIPC
+ *
+ * Copyright (c) 2013 Patrick McHardy <kaber@trash.net>
+ *
+ * Based on eth_media.c, which carries the following copyright notice:
+ *
+ * Copyright (c) 2001-2007, Ericsson AB
+ * Copyright (c) 2005-2008, 2011, 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 <linux/if_infiniband.h>
+#include "core.h"
+#include "bearer.h"
+
+#define MAX_IB_BEARERS MAX_BEARERS
+
+/**
+ * struct ib_bearer - Infiniband bearer data structure
+ * @bearer: ptr to associated "generic" bearer structure
+ * @dev: ptr to associated Infiniband network device
+ * @tipc_packet_type: used in binding TIPC to Infiniband driver
+ * @cleanup: work item used when disabling bearer
+ */
+
+struct ib_bearer {
+ struct tipc_bearer *bearer;
+ struct net_device *dev;
+ struct packet_type tipc_packet_type;
+ struct work_struct setup;
+ struct work_struct cleanup;
+};
+
+static struct tipc_media ib_media_info;
+static struct ib_bearer ib_bearers[MAX_IB_BEARERS];
+static int ib_started;
+
+/**
+ * ib_media_addr_set - initialize Infiniband media address structure
+ *
+ * Media-dependent "value" field stores MAC address in first 6 bytes
+ * and zeroes out the remaining bytes.
+ */
+static void ib_media_addr_set(const struct tipc_bearer *tb_ptr,
+ struct tipc_media_addr *a, char *mac)
+{
+ BUILD_BUG_ON(sizeof(a->value) < INFINIBAND_ALEN);
+ memcpy(a->value, mac, INFINIBAND_ALEN);
+ a->media_id = TIPC_MEDIA_TYPE_IB;
+ a->broadcast = !memcmp(mac, tb_ptr->bcast_addr.value, INFINIBAND_ALEN);
+}
+
+/**
+ * send_msg - send a TIPC message out over an InfiniBand interface
+ */
+static int send_msg(struct sk_buff *buf, struct tipc_bearer *tb_ptr,
+ struct tipc_media_addr *dest)
+{
+ struct sk_buff *clone;
+ struct net_device *dev;
+ int delta;
+
+ clone = skb_clone(buf, GFP_ATOMIC);
+ if (!clone)
+ return 0;
+
+ dev = ((struct ib_bearer *)(tb_ptr->usr_handle))->dev;
+ delta = dev->hard_header_len - skb_headroom(buf);
+
+ if ((delta > 0) &&
+ pskb_expand_head(clone, SKB_DATA_ALIGN(delta), 0, GFP_ATOMIC)) {
+ kfree_skb(clone);
+ return 0;
+ }
+
+ skb_reset_network_header(clone);
+ clone->dev = dev;
+ clone->protocol = htons(ETH_P_TIPC);
+ dev_hard_header(clone, dev, ETH_P_TIPC, dest->value,
+ dev->dev_addr, clone->len);
+ dev_queue_xmit(clone);
+ return 0;
+}
+
+/**
+ * recv_msg - handle incoming TIPC message from an InfiniBand interface
+ *
+ * Accept only packets explicitly sent to this node, or broadcast packets;
+ * ignores packets sent using InfiniBand multicast, and traffic sent to other
+ * nodes (which can happen if interface is running in promiscuous mode).
+ */
+static int recv_msg(struct sk_buff *buf, struct net_device *dev,
+ struct packet_type *pt, struct net_device *orig_dev)
+{
+ struct ib_bearer *ib_ptr = (struct ib_bearer *)pt->af_packet_priv;
+
+ if (!net_eq(dev_net(dev), &init_net)) {
+ kfree_skb(buf);
+ return 0;
+ }
+
+ if (likely(ib_ptr->bearer)) {
+ if (likely(buf->pkt_type <= PACKET_BROADCAST)) {
+ buf->next = NULL;
+ tipc_recv_msg(buf, ib_ptr->bearer);
+ return 0;
+ }
+ }
+ kfree_skb(buf);
+ return 0;
+}
+
+/**
+ * setup_bearer - setup association between InfiniBand bearer and interface
+ */
+static void setup_bearer(struct work_struct *work)
+{
+ struct ib_bearer *ib_ptr =
+ container_of(work, struct ib_bearer, setup);
+
+ dev_add_pack(&ib_ptr->tipc_packet_type);
+}
+
+/**
+ * enable_bearer - attach TIPC bearer to an InfiniBand interface
+ */
+static int enable_bearer(struct tipc_bearer *tb_ptr)
+{
+ struct net_device *dev;
+ struct ib_bearer *ib_ptr = &ib_bearers[0];
+ struct ib_bearer *stop = &ib_bearers[MAX_IB_BEARERS];
+ char *driver_name = strchr((const char *)tb_ptr->name, ':') + 1;
+ int pending_dev = 0;
+
+ /* Find unused InfiniBand bearer structure */
+ while (ib_ptr->dev) {
+ if (!ib_ptr->bearer)
+ pending_dev++;
+ if (++ib_ptr == stop)
+ return pending_dev ? -EAGAIN : -EDQUOT;
+ }
+
+ /* Find device with specified name */
+ dev = dev_get_by_name(&init_net, driver_name);
+ if (!dev)
+ return -ENODEV;
+
+ /* Create InfiniBand bearer for device */
+ ib_ptr->dev = dev;
+ ib_ptr->tipc_packet_type.type = htons(ETH_P_TIPC);
+ ib_ptr->tipc_packet_type.dev = dev;
+ ib_ptr->tipc_packet_type.func = recv_msg;
+ ib_ptr->tipc_packet_type.af_packet_priv = ib_ptr;
+ INIT_LIST_HEAD(&(ib_ptr->tipc_packet_type.list));
+ INIT_WORK(&ib_ptr->setup, setup_bearer);
+ schedule_work(&ib_ptr->setup);
+
+ /* Associate TIPC bearer with InfiniBand bearer */
+ ib_ptr->bearer = tb_ptr;
+ tb_ptr->usr_handle = (void *)ib_ptr;
+ memset(tb_ptr->bcast_addr.value, 0, sizeof(tb_ptr->bcast_addr.value));
+ memcpy(tb_ptr->bcast_addr.value, dev->broadcast, INFINIBAND_ALEN);
+ tb_ptr->bcast_addr.media_id = TIPC_MEDIA_TYPE_IB;
+ tb_ptr->bcast_addr.broadcast = 1;
+ tb_ptr->mtu = dev->mtu;
+ tb_ptr->blocked = 0;
+ ib_media_addr_set(tb_ptr, &tb_ptr->addr, (char *)dev->dev_addr);
+ return 0;
+}
+
+/**
+ * cleanup_bearer - break association between InfiniBand bearer and interface
+ *
+ * This routine must be invoked from a work queue because it can sleep.
+ */
+static void cleanup_bearer(struct work_struct *work)
+{
+ struct ib_bearer *ib_ptr =
+ container_of(work, struct ib_bearer, cleanup);
+
+ dev_remove_pack(&ib_ptr->tipc_packet_type);
+ dev_put(ib_ptr->dev);
+ ib_ptr->dev = NULL;
+}
+
+/**
+ * disable_bearer - detach TIPC bearer from an InfiniBand interface
+ *
+ * Mark InfiniBand bearer as inactive so that incoming buffers are thrown away,
+ * then get worker thread to complete bearer cleanup. (Can't do cleanup
+ * here because cleanup code needs to sleep and caller holds spinlocks.)
+ */
+static void disable_bearer(struct tipc_bearer *tb_ptr)
+{
+ struct ib_bearer *ib_ptr = (struct ib_bearer *)tb_ptr->usr_handle;
+
+ ib_ptr->bearer = NULL;
+ INIT_WORK(&ib_ptr->cleanup, cleanup_bearer);
+ schedule_work(&ib_ptr->cleanup);
+}
+
+/**
+ * recv_notification - handle device updates from OS
+ *
+ * Change the state of the InfiniBand bearer (if any) associated with the
+ * specified device.
+ */
+static int recv_notification(struct notifier_block *nb, unsigned long evt,
+ void *ptr)
+{
+ struct net_device *dev = netdev_notifier_info_to_dev(ptr);
+ struct ib_bearer *ib_ptr = &ib_bearers[0];
+ struct ib_bearer *stop = &ib_bearers[MAX_IB_BEARERS];
+
+ if (!net_eq(dev_net(dev), &init_net))
+ return NOTIFY_DONE;
+
+ while ((ib_ptr->dev != dev)) {
+ if (++ib_ptr == stop)
+ return NOTIFY_DONE; /* couldn't find device */
+ }
+ if (!ib_ptr->bearer)
+ return NOTIFY_DONE; /* bearer had been disabled */
+
+ ib_ptr->bearer->mtu = dev->mtu;
+
+ switch (evt) {
+ case NETDEV_CHANGE:
+ if (netif_carrier_ok(dev))
+ tipc_continue(ib_ptr->bearer);
+ else
+ tipc_block_bearer(ib_ptr->bearer->name);
+ break;
+ case NETDEV_UP:
+ tipc_continue(ib_ptr->bearer);
+ break;
+ case NETDEV_DOWN:
+ tipc_block_bearer(ib_ptr->bearer->name);
+ break;
+ case NETDEV_CHANGEMTU:
+ case NETDEV_CHANGEADDR:
+ tipc_block_bearer(ib_ptr->bearer->name);
+ tipc_continue(ib_ptr->bearer);
+ break;
+ case NETDEV_UNREGISTER:
+ case NETDEV_CHANGENAME:
+ tipc_disable_bearer(ib_ptr->bearer->name);
+ break;
+ }
+ return NOTIFY_OK;
+}
+
+static struct notifier_block notifier = {
+ .notifier_call = recv_notification,
+ .priority = 0,
+};
+
+/**
+ * ib_addr2str - convert InfiniBand address to string
+ */
+static int ib_addr2str(struct tipc_media_addr *a, char *str_buf, int str_size)
+{
+ if (str_size < 60) /* 60 = 19 * strlen("xx:") + strlen("xx\0") */
+ return 1;
+
+ sprintf(str_buf, "%20phC", a->value);
+
+ return 0;
+}
+
+/**
+ * ib_addr2msg - convert InfiniBand address format to message header format
+ */
+static int ib_addr2msg(struct tipc_media_addr *a, char *msg_area)
+{
+ memset(msg_area, 0, TIPC_MEDIA_ADDR_SIZE);
+ msg_area[TIPC_MEDIA_TYPE_OFFSET] = TIPC_MEDIA_TYPE_IB;
+ memcpy(msg_area, a->value, INFINIBAND_ALEN);
+ return 0;
+}
+
+/**
+ * ib_msg2addr - convert message header address format to InfiniBand format
+ */
+static int ib_msg2addr(const struct tipc_bearer *tb_ptr,
+ struct tipc_media_addr *a, char *msg_area)
+{
+ ib_media_addr_set(tb_ptr, a, msg_area);
+ return 0;
+}
+
+/*
+ * InfiniBand media registration info
+ */
+static struct tipc_media ib_media_info = {
+ .send_msg = send_msg,
+ .enable_bearer = enable_bearer,
+ .disable_bearer = disable_bearer,
+ .addr2str = ib_addr2str,
+ .addr2msg = ib_addr2msg,
+ .msg2addr = ib_msg2addr,
+ .priority = TIPC_DEF_LINK_PRI,
+ .tolerance = TIPC_DEF_LINK_TOL,
+ .window = TIPC_DEF_LINK_WIN,
+ .type_id = TIPC_MEDIA_TYPE_IB,
+ .name = "ib"
+};
+
+/**
+ * tipc_ib_media_start - activate InfiniBand bearer support
+ *
+ * Register InfiniBand media type with TIPC bearer code. Also register
+ * with OS for notifications about device state changes.
+ */
+int tipc_ib_media_start(void)
+{
+ int res;
+
+ if (ib_started)
+ return -EINVAL;
+
+ res = tipc_register_media(&ib_media_info);
+ if (res)
+ return res;
+
+ res = register_netdevice_notifier(&notifier);
+ if (!res)
+ ib_started = 1;
+ return res;
+}
+
+/**
+ * tipc_ib_media_stop - deactivate InfiniBand bearer support
+ */
+void tipc_ib_media_stop(void)
+{
+ if (!ib_started)
+ return;
+
+ flush_scheduled_work();
+ unregister_netdevice_notifier(&notifier);
+ ib_started = 0;
+}
diff --git a/net/tipc/link.c b/net/tipc/link.c
index 1c1e6151875e..0cc3d9015c5d 100644
--- a/net/tipc/link.c
+++ b/net/tipc/link.c
@@ -1,8 +1,8 @@
/*
* net/tipc/link.c: TIPC link code
*
- * Copyright (c) 1996-2007, Ericsson AB
- * Copyright (c) 2004-2007, 2010-2011, Wind River Systems
+ * Copyright (c) 1996-2007, 2012, 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
@@ -41,6 +41,8 @@
#include "discover.h"
#include "config.h"
+#include <linux/pkt_sched.h>
+
/*
* Error message prefixes
*/
@@ -97,12 +99,13 @@ static int link_send_sections_long(struct tipc_port *sender,
struct iovec const *msg_sect,
u32 num_sect, unsigned int total_len,
u32 destnode);
-static void link_check_defragm_bufs(struct tipc_link *l_ptr);
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);
static void link_start(struct tipc_link *l_ptr);
static int link_send_long_buf(struct tipc_link *l_ptr, struct sk_buff *buf);
+static void tipc_link_send_sync(struct tipc_link *l);
+static void tipc_link_recv_sync(struct tipc_node *n, struct sk_buff *buf);
/*
* Simple link routines
@@ -210,9 +213,7 @@ static int link_name_validate(const char *name,
(z_local > 255) || (c_local > 4095) || (n_local > 4095) ||
(z_peer > 255) || (c_peer > 4095) || (n_peer > 4095) ||
(if_local_len <= 1) || (if_local_len > TIPC_MAX_IF_NAME) ||
- (if_peer_len <= 1) || (if_peer_len > TIPC_MAX_IF_NAME) ||
- (strspn(if_local, tipc_alphabet) != (if_local_len - 1)) ||
- (strspn(if_peer, tipc_alphabet) != (if_peer_len - 1)))
+ (if_peer_len <= 1) || (if_peer_len > TIPC_MAX_IF_NAME))
return 0;
/* return link name components, if necessary */
@@ -271,7 +272,6 @@ static void link_timeout(struct tipc_link *l_ptr)
}
/* do all other link processing performed on a periodic basis */
- link_check_defragm_bufs(l_ptr);
link_state_event(l_ptr, TIMEOUT_EVT);
@@ -714,6 +714,8 @@ static void link_state_event(struct tipc_link *l_ptr, unsigned int event)
link_activate(l_ptr);
tipc_link_send_proto_msg(l_ptr, STATE_MSG, 1, 0, 0, 0, 0);
l_ptr->fsm_msg_cnt++;
+ if (l_ptr->owner->working_links == 1)
+ tipc_link_send_sync(l_ptr);
link_set_timer(l_ptr, cont_intv);
break;
case RESET_MSG:
@@ -747,6 +749,8 @@ static void link_state_event(struct tipc_link *l_ptr, unsigned int event)
link_activate(l_ptr);
tipc_link_send_proto_msg(l_ptr, STATE_MSG, 1, 0, 0, 0, 0);
l_ptr->fsm_msg_cnt++;
+ if (l_ptr->owner->working_links == 1)
+ tipc_link_send_sync(l_ptr);
link_set_timer(l_ptr, cont_intv);
break;
case RESET_MSG:
@@ -769,8 +773,7 @@ static void link_state_event(struct tipc_link *l_ptr, unsigned int event)
* link_bundle_buf(): Append contents of a buffer to
* the tail of an existing one.
*/
-static int link_bundle_buf(struct tipc_link *l_ptr,
- struct sk_buff *bundler,
+static int link_bundle_buf(struct tipc_link *l_ptr, struct sk_buff *bundler,
struct sk_buff *buf)
{
struct tipc_msg *bundler_msg = buf_msg(bundler);
@@ -874,17 +877,12 @@ int tipc_link_send_buf(struct tipc_link *l_ptr, struct sk_buff *buf)
return link_send_long_buf(l_ptr, buf);
/* Packet can be queued or sent. */
- if (likely(!tipc_bearer_congested(l_ptr->b_ptr, l_ptr) &&
+ if (likely(!tipc_bearer_blocked(l_ptr->b_ptr) &&
!link_congested(l_ptr))) {
link_add_to_outqueue(l_ptr, buf, msg);
- if (likely(tipc_bearer_send(l_ptr->b_ptr, buf, &l_ptr->media_addr))) {
- l_ptr->unacked_window = 0;
- } else {
- tipc_bearer_schedule(l_ptr->b_ptr, l_ptr);
- l_ptr->stats.bearer_congs++;
- l_ptr->next_out = buf;
- }
+ tipc_bearer_send(l_ptr->b_ptr, buf, &l_ptr->media_addr);
+ l_ptr->unacked_window = 0;
return dsz;
}
/* Congestion: can message be bundled ? */
@@ -893,10 +891,8 @@ int tipc_link_send_buf(struct tipc_link *l_ptr, struct sk_buff *buf)
/* Try adding message to an existing bundle */
if (l_ptr->next_out &&
- link_bundle_buf(l_ptr, l_ptr->last_out, buf)) {
- tipc_bearer_resolve_congestion(l_ptr->b_ptr, l_ptr);
+ link_bundle_buf(l_ptr, l_ptr->last_out, buf))
return dsz;
- }
/* Try creating a new bundle */
if (size <= max_packet * 2 / 3) {
@@ -919,7 +915,6 @@ int tipc_link_send_buf(struct tipc_link *l_ptr, struct sk_buff *buf)
if (!l_ptr->next_out)
l_ptr->next_out = buf;
link_add_to_outqueue(l_ptr, buf, msg);
- tipc_bearer_resolve_congestion(l_ptr->b_ptr, l_ptr);
return dsz;
}
@@ -951,7 +946,48 @@ int tipc_link_send(struct sk_buff *buf, u32 dest, u32 selector)
return res;
}
-/**
+/*
+ * tipc_link_send_sync - synchronize broadcast link endpoints.
+ *
+ * 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_send_sync(struct tipc_link *l)
+{
+ struct sk_buff *buf;
+ struct tipc_msg *msg;
+
+ buf = tipc_buf_acquire(INT_H_SIZE);
+ if (!buf)
+ return;
+
+ msg = buf_msg(buf);
+ tipc_msg_init(msg, BCAST_PROTOCOL, STATE_MSG, INT_H_SIZE, l->addr);
+ msg_set_last_bcast(msg, l->owner->bclink.acked);
+ link_add_chain_to_outqueue(l, buf, 0);
+ tipc_link_push_queue(l);
+}
+
+/*
+ * tipc_link_recv_sync - synchronize broadcast link endpoints.
+ * 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
+ */
+static void tipc_link_recv_sync(struct tipc_node *n, struct sk_buff *buf)
+{
+ 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);
+}
+
+/*
* tipc_link_send_names - send name table entries to new neighbor
*
* Send routine for bulk delivery of name table messages when contact
@@ -1008,16 +1044,11 @@ static int link_send_buf_fast(struct tipc_link *l_ptr, struct sk_buff *buf,
if (likely(!link_congested(l_ptr))) {
if (likely(msg_size(msg) <= l_ptr->max_pkt)) {
- if (likely(list_empty(&l_ptr->b_ptr->cong_links))) {
+ if (likely(!tipc_bearer_blocked(l_ptr->b_ptr))) {
link_add_to_outqueue(l_ptr, buf, msg);
- if (likely(tipc_bearer_send(l_ptr->b_ptr, buf,
- &l_ptr->media_addr))) {
- l_ptr->unacked_window = 0;
- return res;
- }
- tipc_bearer_schedule(l_ptr->b_ptr, l_ptr);
- l_ptr->stats.bearer_congs++;
- l_ptr->next_out = buf;
+ tipc_bearer_send(l_ptr->b_ptr, buf,
+ &l_ptr->media_addr);
+ l_ptr->unacked_window = 0;
return res;
}
} else
@@ -1027,40 +1058,6 @@ static int link_send_buf_fast(struct tipc_link *l_ptr, struct sk_buff *buf,
}
/*
- * tipc_send_buf_fast: Entry for data messages where the
- * destination node is known and the header is complete,
- * inclusive total message length.
- * Returns user data length.
- */
-int tipc_send_buf_fast(struct sk_buff *buf, u32 destnode)
-{
- struct tipc_link *l_ptr;
- struct tipc_node *n_ptr;
- int res;
- u32 selector = msg_origport(buf_msg(buf)) & 1;
- u32 dummy;
-
- read_lock_bh(&tipc_net_lock);
- n_ptr = tipc_node_find(destnode);
- if (likely(n_ptr)) {
- tipc_node_lock(n_ptr);
- l_ptr = n_ptr->active_links[selector];
- if (likely(l_ptr)) {
- res = link_send_buf_fast(l_ptr, buf, &dummy);
- tipc_node_unlock(n_ptr);
- read_unlock_bh(&tipc_net_lock);
- return res;
- }
- tipc_node_unlock(n_ptr);
- }
- read_unlock_bh(&tipc_net_lock);
- res = msg_data_sz(buf_msg(buf));
- tipc_reject_msg(buf, TIPC_ERR_NO_NODE);
- return res;
-}
-
-
-/*
* tipc_link_send_sections_fast: Entry for messages where the
* destination processor is known and the header is complete,
* except for total message length.
@@ -1068,8 +1065,7 @@ int tipc_send_buf_fast(struct sk_buff *buf, u32 destnode)
*/
int tipc_link_send_sections_fast(struct tipc_port *sender,
struct iovec const *msg_sect,
- const u32 num_sect,
- unsigned int total_len,
+ const u32 num_sect, unsigned int total_len,
u32 destaddr)
{
struct tipc_msg *hdr = &sender->phdr;
@@ -1085,7 +1081,10 @@ again:
* (Must not hold any locks while building message.)
*/
res = tipc_msg_build(hdr, msg_sect, num_sect, total_len,
- sender->max_pkt, !sender->user_port, &buf);
+ sender->max_pkt, &buf);
+ /* Exit if build request was invalid */
+ if (unlikely(res < 0))
+ return res;
read_lock_bh(&tipc_net_lock);
node = tipc_node_find(destaddr);
@@ -1102,13 +1101,9 @@ exit:
return res;
}
- /* Exit if build request was invalid */
- if (unlikely(res < 0))
- goto exit;
-
/* Exit if link (or bearer) is congested */
if (link_congested(l_ptr) ||
- !list_empty(&l_ptr->b_ptr->cong_links)) {
+ tipc_bearer_blocked(l_ptr->b_ptr)) {
res = link_schedule_port(l_ptr,
sender->ref, res);
goto exit;
@@ -1159,8 +1154,7 @@ exit:
*/
static int link_send_sections_long(struct tipc_port *sender,
struct iovec const *msg_sect,
- u32 num_sect,
- unsigned int total_len,
+ u32 num_sect, unsigned int total_len,
u32 destaddr)
{
struct tipc_link *l_ptr;
@@ -1174,6 +1168,7 @@ static int link_send_sections_long(struct tipc_port *sender,
const unchar *sect_crs;
int curr_sect;
u32 fragm_no;
+ int res = 0;
again:
fragm_no = 1;
@@ -1220,18 +1215,15 @@ again:
else
sz = fragm_rest;
- if (likely(!sender->user_port)) {
- if (copy_from_user(buf->data + fragm_crs, sect_crs, sz)) {
+ if (copy_from_user(buf->data + fragm_crs, sect_crs, sz)) {
+ res = -EFAULT;
error:
- for (; buf_chain; buf_chain = buf) {
- buf = buf_chain->next;
- kfree_skb(buf_chain);
- }
- return -EFAULT;
+ for (; buf_chain; buf_chain = buf) {
+ buf = buf_chain->next;
+ kfree_skb(buf_chain);
}
- } else
- skb_copy_to_linear_data_offset(buf, fragm_crs,
- sect_crs, sz);
+ return res;
+ }
sect_crs += sz;
sect_rest -= sz;
fragm_crs += sz;
@@ -1251,8 +1243,10 @@ error:
msg_set_fragm_no(&fragm_hdr, ++fragm_no);
prev = buf;
buf = tipc_buf_acquire(fragm_sz + INT_H_SIZE);
- if (!buf)
+ if (!buf) {
+ res = -ENOMEM;
goto error;
+ }
buf->next = NULL;
prev->next = buf;
@@ -1331,15 +1325,11 @@ u32 tipc_link_push_packet(struct tipc_link *l_ptr)
if (r_q_size && buf) {
msg_set_ack(buf_msg(buf), mod(l_ptr->next_in_no - 1));
msg_set_bcast_ack(buf_msg(buf), l_ptr->owner->bclink.last_in);
- if (tipc_bearer_send(l_ptr->b_ptr, buf, &l_ptr->media_addr)) {
- l_ptr->retransm_queue_head = mod(++r_q_head);
- l_ptr->retransm_queue_size = --r_q_size;
- l_ptr->stats.retransmitted++;
- return 0;
- } else {
- l_ptr->stats.bearer_congs++;
- return PUSH_FAILED;
- }
+ tipc_bearer_send(l_ptr->b_ptr, buf, &l_ptr->media_addr);
+ l_ptr->retransm_queue_head = mod(++r_q_head);
+ l_ptr->retransm_queue_size = --r_q_size;
+ l_ptr->stats.retransmitted++;
+ return 0;
}
/* Send deferred protocol message, if any: */
@@ -1347,15 +1337,11 @@ u32 tipc_link_push_packet(struct tipc_link *l_ptr)
if (buf) {
msg_set_ack(buf_msg(buf), mod(l_ptr->next_in_no - 1));
msg_set_bcast_ack(buf_msg(buf), l_ptr->owner->bclink.last_in);
- if (tipc_bearer_send(l_ptr->b_ptr, buf, &l_ptr->media_addr)) {
- l_ptr->unacked_window = 0;
- kfree_skb(buf);
- l_ptr->proto_msg_queue = NULL;
- return 0;
- } else {
- l_ptr->stats.bearer_congs++;
- return PUSH_FAILED;
- }
+ tipc_bearer_send(l_ptr->b_ptr, buf, &l_ptr->media_addr);
+ l_ptr->unacked_window = 0;
+ kfree_skb(buf);
+ l_ptr->proto_msg_queue = NULL;
+ return 0;
}
/* Send one deferred data message, if send window not full: */
@@ -1368,18 +1354,14 @@ u32 tipc_link_push_packet(struct tipc_link *l_ptr)
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 (tipc_bearer_send(l_ptr->b_ptr, buf, &l_ptr->media_addr)) {
- if (msg_user(msg) == MSG_BUNDLER)
- msg_set_type(msg, CLOSED_MSG);
- l_ptr->next_out = buf->next;
- return 0;
- } else {
- l_ptr->stats.bearer_congs++;
- return PUSH_FAILED;
- }
+ tipc_bearer_send(l_ptr->b_ptr, buf, &l_ptr->media_addr);
+ if (msg_user(msg) == MSG_BUNDLER)
+ msg_set_type(msg, CLOSED_MSG);
+ l_ptr->next_out = buf->next;
+ return 0;
}
}
- return PUSH_FINISHED;
+ return 1;
}
/*
@@ -1390,15 +1372,12 @@ void tipc_link_push_queue(struct tipc_link *l_ptr)
{
u32 res;
- if (tipc_bearer_congested(l_ptr->b_ptr, l_ptr))
+ if (tipc_bearer_blocked(l_ptr->b_ptr))
return;
do {
res = tipc_link_push_packet(l_ptr);
} while (!res);
-
- if (res == PUSH_FAILED)
- tipc_bearer_schedule(l_ptr->b_ptr, l_ptr);
}
static void link_reset_all(unsigned long addr)
@@ -1431,7 +1410,7 @@ static void link_reset_all(unsigned long addr)
}
static void link_retransmit_failure(struct tipc_link *l_ptr,
- struct sk_buff *buf)
+ struct sk_buff *buf)
{
struct tipc_msg *msg = buf_msg(buf);
@@ -1456,9 +1435,8 @@ static void link_retransmit_failure(struct tipc_link *l_ptr,
tipc_addr_string_fill(addr_string, n_ptr->addr);
pr_info("Broadcast link info for %s\n", addr_string);
- pr_info("Supportable: %d, Supported: %d, Acked: %u\n",
- n_ptr->bclink.supportable,
- n_ptr->bclink.supported,
+ 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,
@@ -1483,7 +1461,7 @@ void tipc_link_retransmit(struct tipc_link *l_ptr, struct sk_buff *buf,
msg = buf_msg(buf);
- if (tipc_bearer_congested(l_ptr->b_ptr, l_ptr)) {
+ if (tipc_bearer_blocked(l_ptr->b_ptr)) {
if (l_ptr->retransm_queue_size == 0) {
l_ptr->retransm_queue_head = msg_seqno(msg);
l_ptr->retransm_queue_size = retransmits;
@@ -1493,7 +1471,7 @@ void tipc_link_retransmit(struct tipc_link *l_ptr, struct sk_buff *buf,
}
return;
} else {
- /* Detect repeated retransmit failures on uncongested bearer */
+ /* Detect repeated retransmit failures on unblocked bearer */
if (l_ptr->last_retransmitted == msg_seqno(msg)) {
if (++l_ptr->stale_count > 100) {
link_retransmit_failure(l_ptr, buf);
@@ -1509,17 +1487,10 @@ void tipc_link_retransmit(struct tipc_link *l_ptr, struct sk_buff *buf,
msg = buf_msg(buf);
msg_set_ack(msg, mod(l_ptr->next_in_no - 1));
msg_set_bcast_ack(msg, l_ptr->owner->bclink.last_in);
- if (tipc_bearer_send(l_ptr->b_ptr, buf, &l_ptr->media_addr)) {
- buf = buf->next;
- retransmits--;
- l_ptr->stats.retransmitted++;
- } else {
- tipc_bearer_schedule(l_ptr->b_ptr, l_ptr);
- l_ptr->stats.bearer_congs++;
- l_ptr->retransm_queue_head = buf_seqno(buf);
- l_ptr->retransm_queue_size = retransmits;
- return;
- }
+ tipc_bearer_send(l_ptr->b_ptr, buf, &l_ptr->media_addr);
+ buf = buf->next;
+ retransmits--;
+ l_ptr->stats.retransmitted++;
}
l_ptr->retransm_queue_head = l_ptr->retransm_queue_size = 0;
@@ -1678,7 +1649,7 @@ void tipc_recv_msg(struct sk_buff *head, struct tipc_bearer *b_ptr)
ackd = msg_ack(msg);
/* Release acked messages */
- if (n_ptr->bclink.supported)
+ if (n_ptr->bclink.recv_permitted)
tipc_bclink_acknowledge(n_ptr, msg_bcast_ack(msg));
crs = l_ptr->first_out;
@@ -1729,9 +1700,14 @@ deliver:
tipc_link_recv_bundle(buf);
continue;
case NAME_DISTRIBUTOR:
+ n_ptr->bclink.recv_permitted = true;
tipc_node_unlock(n_ptr);
tipc_named_recv(buf);
continue;
+ case BCAST_PROTOCOL:
+ tipc_link_recv_sync(n_ptr, buf);
+ tipc_node_unlock(n_ptr);
+ continue;
case CONN_MANAGER:
tipc_node_unlock(n_ptr);
tipc_port_recv_proto_msg(buf);
@@ -1774,16 +1750,19 @@ deliver:
continue;
}
+ /* Link is not in state WORKING_WORKING */
if (msg_user(msg) == LINK_PROTOCOL) {
link_recv_proto_msg(l_ptr, buf);
head = link_insert_deferred_queue(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 in front of queue */
+ /* Re-insert buffer in front of queue */
buf->next = head;
head = buf;
tipc_node_unlock(n_ptr);
@@ -1886,8 +1865,8 @@ static void link_handle_out_of_seq_msg(struct tipc_link *l_ptr,
* Send protocol message to the other endpoint.
*/
void tipc_link_send_proto_msg(struct tipc_link *l_ptr, u32 msg_typ,
- int probe_msg, u32 gap, u32 tolerance,
- u32 priority, u32 ack_mtu)
+ int probe_msg, u32 gap, u32 tolerance,
+ u32 priority, u32 ack_mtu)
{
struct sk_buff *buf = NULL;
struct tipc_msg *msg = l_ptr->pmsg;
@@ -1973,22 +1952,15 @@ void tipc_link_send_proto_msg(struct tipc_link *l_ptr, u32 msg_typ,
return;
skb_copy_to_linear_data(buf, msg, sizeof(l_ptr->proto_msg));
+ buf->priority = TC_PRIO_CONTROL;
- /* Defer message if bearer is already congested */
- if (tipc_bearer_congested(l_ptr->b_ptr, l_ptr)) {
+ /* Defer message if bearer is already blocked */
+ if (tipc_bearer_blocked(l_ptr->b_ptr)) {
l_ptr->proto_msg_queue = buf;
return;
}
- /* Defer message if attempting to send results in bearer congestion */
- if (!tipc_bearer_send(l_ptr->b_ptr, buf, &l_ptr->media_addr)) {
- tipc_bearer_schedule(l_ptr->b_ptr, l_ptr);
- l_ptr->proto_msg_queue = buf;
- l_ptr->stats.bearer_congs++;
- return;
- }
-
- /* Discard message if it was sent successfully */
+ tipc_bearer_send(l_ptr->b_ptr, buf, &l_ptr->media_addr);
l_ptr->unacked_window = 0;
kfree_skb(buf);
}
@@ -2059,7 +2031,6 @@ static void link_recv_proto_msg(struct tipc_link *l_ptr, struct sk_buff *buf)
} else {
l_ptr->max_pkt = l_ptr->max_pkt_target;
}
- l_ptr->owner->bclink.supportable = (max_pkt_info != 0);
/* Synchronize broadcast link info, if not done previously */
if (!tipc_node_is_up(l_ptr->owner)) {
@@ -2114,7 +2085,7 @@ static void link_recv_proto_msg(struct tipc_link *l_ptr, struct sk_buff *buf)
}
/* Protocol message before retransmits, reduce loss risk */
- if (l_ptr->owner->bclink.supported)
+ if (l_ptr->owner->bclink.recv_permitted)
tipc_bclink_update_link_state(l_ptr->owner,
msg_last_bcast(msg));
@@ -2139,8 +2110,7 @@ exit:
* another bearer. Owner node is locked.
*/
static void tipc_link_tunnel(struct tipc_link *l_ptr,
- struct tipc_msg *tunnel_hdr,
- struct tipc_msg *msg,
+ struct tipc_msg *tunnel_hdr, struct tipc_msg *msg,
u32 selector)
{
struct tipc_link *tunnel;
@@ -2300,8 +2270,11 @@ static int link_recv_changeover_msg(struct tipc_link **l_ptr,
struct tipc_msg *tunnel_msg = buf_msg(tunnel_buf);
u32 msg_typ = msg_type(tunnel_msg);
u32 msg_count = msg_msgcnt(tunnel_msg);
+ u32 bearer_id = msg_bearer_id(tunnel_msg);
- dest_link = (*l_ptr)->owner->links[msg_bearer_id(tunnel_msg)];
+ if (bearer_id >= MAX_BEARERS)
+ goto exit;
+ dest_link = (*l_ptr)->owner->links[bearer_id];
if (!dest_link)
goto exit;
if (dest_link == *l_ptr) {
@@ -2489,16 +2462,6 @@ static void set_expected_frags(struct sk_buff *buf, u32 exp)
msg_set_bcast_ack(buf_msg(buf), exp);
}
-static u32 get_timer_cnt(struct sk_buff *buf)
-{
- return msg_reroute_cnt(buf_msg(buf));
-}
-
-static void incr_timer_cnt(struct sk_buff *buf)
-{
- msg_incr_reroute_cnt(buf_msg(buf));
-}
-
/*
* tipc_link_recv_fragment(): Called with node lock on. Returns
* the reassembled buffer if message is complete.
@@ -2525,14 +2488,16 @@ int tipc_link_recv_fragment(struct sk_buff **pending, struct sk_buff **fb,
struct tipc_msg *imsg = (struct tipc_msg *)msg_data(fragm);
u32 msg_sz = msg_size(imsg);
u32 fragm_sz = msg_data_sz(fragm);
- u32 exp_fragm_cnt = msg_sz/fragm_sz + !!(msg_sz % fragm_sz);
+ u32 exp_fragm_cnt;
u32 max = TIPC_MAX_USER_MSG_SIZE + NAMED_H_SIZE;
+
if (msg_type(imsg) == TIPC_MCAST_MSG)
max = TIPC_MAX_USER_MSG_SIZE + MCAST_H_SIZE;
- if (msg_size(imsg) > max) {
+ if (fragm_sz == 0 || msg_size(imsg) > max) {
kfree_skb(fbuf);
return 0;
}
+ exp_fragm_cnt = msg_sz / fragm_sz + !!(msg_sz % fragm_sz);
pbuf = tipc_buf_acquire(msg_size(imsg));
if (pbuf != NULL) {
pbuf->next = *pending;
@@ -2577,38 +2542,6 @@ int tipc_link_recv_fragment(struct sk_buff **pending, struct sk_buff **fb,
return 0;
}
-/**
- * link_check_defragm_bufs - flush stale incoming message fragments
- * @l_ptr: pointer to link
- */
-static void link_check_defragm_bufs(struct tipc_link *l_ptr)
-{
- struct sk_buff *prev = NULL;
- struct sk_buff *next = NULL;
- struct sk_buff *buf = l_ptr->defragm_buf;
-
- if (!buf)
- return;
- if (!link_working_working(l_ptr))
- return;
- while (buf) {
- u32 cnt = get_timer_cnt(buf);
-
- next = buf->next;
- if (cnt < 4) {
- incr_timer_cnt(buf);
- prev = buf;
- } else {
- if (prev)
- prev->next = buf->next;
- else
- l_ptr->defragm_buf = buf->next;
- kfree_skb(buf);
- }
- buf = next;
- }
-}
-
static void link_set_supervision_props(struct tipc_link *l_ptr, u32 tolerance)
{
if ((tolerance < TIPC_MIN_LINK_TOL) || (tolerance > TIPC_MAX_LINK_TOL))
@@ -2939,8 +2872,8 @@ static int tipc_link_stats(const char *name, char *buf, const u32 buf_size)
s->sent_nacks, s->sent_acks, s->retransmitted);
ret += tipc_snprintf(buf + ret, buf_size - ret,
- " Congestion bearer:%u link:%u Send queue"
- " max:%u avg:%u\n", s->bearer_congs, s->link_congs,
+ " Congestion link:%u Send queue"
+ " max:%u avg:%u\n", s->link_congs,
s->max_queue_sz, s->queue_sz_counts ?
(s->accu_queue_sz / s->queue_sz_counts) : 0);
diff --git a/net/tipc/link.h b/net/tipc/link.h
index 6e921121be06..c048ed1cbd76 100644
--- a/net/tipc/link.h
+++ b/net/tipc/link.h
@@ -40,9 +40,6 @@
#include "msg.h"
#include "node.h"
-#define PUSH_FAILED 1
-#define PUSH_FINISHED 2
-
/*
* Out-of-range value for link sequence numbers
*/
@@ -82,7 +79,6 @@ struct tipc_stats {
u32 recv_fragmented;
u32 recv_fragments;
u32 link_congs; /* # port sends blocked by congestion */
- u32 bearer_congs;
u32 deferred_recv;
u32 duplicates;
u32 max_queue_sz; /* send queue size high water mark */
diff --git a/net/tipc/msg.c b/net/tipc/msg.c
index f2db8a87d9c5..ced60e2fc4f7 100644
--- a/net/tipc/msg.c
+++ b/net/tipc/msg.c
@@ -51,8 +51,8 @@ u32 tipc_msg_tot_importance(struct tipc_msg *m)
}
-void tipc_msg_init(struct tipc_msg *m, u32 user, u32 type,
- u32 hsize, u32 destnode)
+void tipc_msg_init(struct tipc_msg *m, u32 user, u32 type, u32 hsize,
+ u32 destnode)
{
memset(m, 0, hsize);
msg_set_version(m);
@@ -73,8 +73,8 @@ void tipc_msg_init(struct tipc_msg *m, u32 user, u32 type,
* Returns message data size or errno
*/
int tipc_msg_build(struct tipc_msg *hdr, struct iovec const *msg_sect,
- u32 num_sect, unsigned int total_len,
- int max_size, int usrmem, struct sk_buff **buf)
+ u32 num_sect, unsigned int total_len, int max_size,
+ struct sk_buff **buf)
{
int dsz, sz, hsz, pos, res, cnt;
@@ -92,14 +92,9 @@ int tipc_msg_build(struct tipc_msg *hdr, struct iovec const *msg_sect,
return -ENOMEM;
skb_copy_to_linear_data(*buf, hdr, hsz);
for (res = 1, cnt = 0; res && (cnt < num_sect); cnt++) {
- if (likely(usrmem))
- res = !copy_from_user((*buf)->data + pos,
- msg_sect[cnt].iov_base,
- msg_sect[cnt].iov_len);
- else
- skb_copy_to_linear_data_offset(*buf, pos,
- msg_sect[cnt].iov_base,
- msg_sect[cnt].iov_len);
+ skb_copy_to_linear_data_offset(*buf, pos,
+ msg_sect[cnt].iov_base,
+ msg_sect[cnt].iov_len);
pos += msg_sect[cnt].iov_len;
}
if (likely(res))
diff --git a/net/tipc/msg.h b/net/tipc/msg.h
index ba2a72beea68..5e4ccf5c27df 100644
--- a/net/tipc/msg.h
+++ b/net/tipc/msg.h
@@ -719,9 +719,9 @@ static inline void msg_set_link_tolerance(struct tipc_msg *m, u32 n)
}
u32 tipc_msg_tot_importance(struct tipc_msg *m);
-void tipc_msg_init(struct tipc_msg *m, u32 user, u32 type,
- u32 hsize, u32 destnode);
+void tipc_msg_init(struct tipc_msg *m, u32 user, u32 type, u32 hsize,
+ u32 destnode);
int tipc_msg_build(struct tipc_msg *hdr, struct iovec const *msg_sect,
- u32 num_sect, unsigned int total_len,
- int max_size, int usrmem, struct sk_buff **buf);
+ u32 num_sect, unsigned int total_len, int max_size,
+ struct sk_buff **buf);
#endif
diff --git a/net/tipc/name_distr.c b/net/tipc/name_distr.c
index 55d3928dfd67..e0d08055754e 100644
--- a/net/tipc/name_distr.c
+++ b/net/tipc/name_distr.c
@@ -262,7 +262,7 @@ void tipc_named_node_up(unsigned long nodearg)
named_distribute(&message_list, node, &publ_zone, max_item_buf);
read_unlock_bh(&tipc_nametbl_lock);
- tipc_link_send_names(&message_list, (u32)node);
+ tipc_link_send_names(&message_list, node);
}
/**
diff --git a/net/tipc/name_table.c b/net/tipc/name_table.c
index 360c478b0b53..09dcd54b04e1 100644
--- a/net/tipc/name_table.c
+++ b/net/tipc/name_table.c
@@ -41,7 +41,7 @@
#include "subscr.h"
#include "port.h"
-static int tipc_nametbl_size = 1024; /* must be a power of 2 */
+#define TIPC_NAMETBL_SIZE 1024 /* must be a power of 2 */
/**
* struct name_info - name sequence publication info
@@ -114,7 +114,7 @@ DEFINE_RWLOCK(tipc_nametbl_lock);
static int hash(int x)
{
- return x & (tipc_nametbl_size - 1);
+ return x & (TIPC_NAMETBL_SIZE - 1);
}
/**
@@ -440,7 +440,7 @@ found:
* sequence overlapping with the requested sequence
*/
static void tipc_nameseq_subscribe(struct name_seq *nseq,
- struct tipc_subscription *s)
+ struct tipc_subscription *s)
{
struct sub_seq *sseq = nseq->sseqs;
@@ -473,11 +473,10 @@ static void tipc_nameseq_subscribe(struct name_seq *nseq,
static struct name_seq *nametbl_find_seq(u32 type)
{
struct hlist_head *seq_head;
- struct hlist_node *seq_node;
struct name_seq *ns;
seq_head = &table.types[hash(type)];
- hlist_for_each_entry(ns, seq_node, seq_head, ns_list) {
+ hlist_for_each_entry(ns, seq_head, ns_list) {
if (ns->type == type)
return ns;
}
@@ -663,13 +662,13 @@ exit:
* tipc_nametbl_publish - add name publication to network name tables
*/
struct publication *tipc_nametbl_publish(u32 type, u32 lower, u32 upper,
- u32 scope, u32 port_ref, u32 key)
+ u32 scope, u32 port_ref, u32 key)
{
struct publication *publ;
- if (table.local_publ_count >= tipc_max_publications) {
+ if (table.local_publ_count >= TIPC_MAX_PUBLICATIONS) {
pr_warn("Publication failed, local publication limit reached (%u)\n",
- tipc_max_publications);
+ TIPC_MAX_PUBLICATIONS);
return NULL;
}
@@ -754,7 +753,7 @@ void tipc_nametbl_unsubscribe(struct tipc_subscription *s)
* subseq_list - print specified sub-sequence contents into the given buffer
*/
static int subseq_list(struct sub_seq *sseq, char *buf, int len, u32 depth,
- u32 index)
+ u32 index)
{
char portIdStr[27];
const char *scope_str[] = {"", " zone", " cluster", " node"};
@@ -783,7 +782,7 @@ static int subseq_list(struct sub_seq *sseq, char *buf, int len, u32 depth,
if (!list_is_last(&publ->zone_list, &info->zone_list))
ret += tipc_snprintf(buf + ret, len - ret,
"\n%33s", " ");
- };
+ }
ret += tipc_snprintf(buf + ret, len - ret, "\n");
return ret;
@@ -793,7 +792,7 @@ static int subseq_list(struct sub_seq *sseq, char *buf, int len, u32 depth,
* nameseq_list - print specified name sequence contents into the given buffer
*/
static int nameseq_list(struct name_seq *seq, char *buf, int len, u32 depth,
- u32 type, u32 lowbound, u32 upbound, u32 index)
+ u32 type, u32 lowbound, u32 upbound, u32 index)
{
struct sub_seq *sseq;
char typearea[11];
@@ -850,10 +849,9 @@ static int nametbl_header(char *buf, int len, u32 depth)
* nametbl_list - print specified name table contents into the given buffer
*/
static int nametbl_list(char *buf, int len, u32 depth_info,
- u32 type, u32 lowbound, u32 upbound)
+ u32 type, u32 lowbound, u32 upbound)
{
struct hlist_head *seq_head;
- struct hlist_node *seq_node;
struct name_seq *seq;
int all_types;
int ret = 0;
@@ -871,9 +869,9 @@ static int nametbl_list(char *buf, int len, u32 depth_info,
ret += nametbl_header(buf, len, depth);
lowbound = 0;
upbound = ~0;
- for (i = 0; i < tipc_nametbl_size; i++) {
+ for (i = 0; i < TIPC_NAMETBL_SIZE; i++) {
seq_head = &table.types[i];
- hlist_for_each_entry(seq, seq_node, seq_head, ns_list) {
+ hlist_for_each_entry(seq, seq_head, ns_list) {
ret += nameseq_list(seq, buf + ret, len - ret,
depth, seq->type,
lowbound, upbound, i);
@@ -889,7 +887,7 @@ static int nametbl_list(char *buf, int len, u32 depth_info,
ret += nametbl_header(buf + ret, len - ret, depth);
i = hash(type);
seq_head = &table.types[i];
- hlist_for_each_entry(seq, seq_node, seq_head, ns_list) {
+ hlist_for_each_entry(seq, seq_head, ns_list) {
if (seq->type == type) {
ret += nameseq_list(seq, buf + ret, len - ret,
depth, type,
@@ -935,7 +933,7 @@ struct sk_buff *tipc_nametbl_get(const void *req_tlv_area, int req_tlv_space)
int tipc_nametbl_init(void)
{
- table.types = kcalloc(tipc_nametbl_size, sizeof(struct hlist_head),
+ table.types = kcalloc(TIPC_NAMETBL_SIZE, sizeof(struct hlist_head),
GFP_ATOMIC);
if (!table.types)
return -ENOMEM;
@@ -953,7 +951,7 @@ void tipc_nametbl_stop(void)
/* Verify name table is empty, then release it */
write_lock_bh(&tipc_nametbl_lock);
- for (i = 0; i < tipc_nametbl_size; i++) {
+ for (i = 0; i < TIPC_NAMETBL_SIZE; i++) {
if (hlist_empty(&table.types[i]))
continue;
pr_err("nametbl_stop(): orphaned hash chain detected\n");
diff --git a/net/tipc/name_table.h b/net/tipc/name_table.h
index 71cb4dc712df..f02f48b9a216 100644
--- a/net/tipc/name_table.h
+++ b/net/tipc/name_table.h
@@ -87,14 +87,15 @@ extern rwlock_t tipc_nametbl_lock;
struct sk_buff *tipc_nametbl_get(const void *req_tlv_area, int req_tlv_space);
u32 tipc_nametbl_translate(u32 type, u32 instance, u32 *node);
int tipc_nametbl_mc_translate(u32 type, u32 lower, u32 upper, u32 limit,
- struct tipc_port_list *dports);
+ struct tipc_port_list *dports);
struct publication *tipc_nametbl_publish(u32 type, u32 lower, u32 upper,
- u32 scope, u32 port_ref, u32 key);
+ u32 scope, u32 port_ref, u32 key);
int tipc_nametbl_withdraw(u32 type, u32 lower, u32 ref, u32 key);
struct publication *tipc_nametbl_insert_publ(u32 type, u32 lower, u32 upper,
- u32 scope, u32 node, u32 ref, u32 key);
-struct publication *tipc_nametbl_remove_publ(u32 type, u32 lower,
- u32 node, u32 ref, u32 key);
+ u32 scope, u32 node, u32 ref,
+ u32 key);
+struct publication *tipc_nametbl_remove_publ(u32 type, u32 lower, u32 node,
+ u32 ref, u32 key);
void tipc_nametbl_subscribe(struct tipc_subscription *s);
void tipc_nametbl_unsubscribe(struct tipc_subscription *s);
int tipc_nametbl_init(void);
diff --git a/net/tipc/net.c b/net/tipc/net.c
index 5b5cea259caf..7d305ecc09c2 100644
--- a/net/tipc/net.c
+++ b/net/tipc/net.c
@@ -171,7 +171,7 @@ void tipc_net_route_msg(struct sk_buff *buf)
tipc_link_send(buf, dnode, msg_link_selector(msg));
}
-int tipc_net_start(u32 addr)
+void tipc_net_start(u32 addr)
{
char addr_string[16];
@@ -187,7 +187,6 @@ int tipc_net_start(u32 addr)
pr_info("Started in network mode\n");
pr_info("Own node address %s, network identity %u\n",
tipc_addr_string_fill(addr_string, tipc_own_addr), tipc_net_id);
- return 0;
}
void tipc_net_stop(void)
diff --git a/net/tipc/net.h b/net/tipc/net.h
index 9eb4b9e220eb..079daadb3f72 100644
--- a/net/tipc/net.h
+++ b/net/tipc/net.h
@@ -41,7 +41,7 @@ extern rwlock_t tipc_net_lock;
void tipc_net_route_msg(struct sk_buff *buf);
-int tipc_net_start(u32 addr);
+void tipc_net_start(u32 addr);
void tipc_net_stop(void);
#endif
diff --git a/net/tipc/netlink.c b/net/tipc/netlink.c
index 47a839df27dc..8bcd4985d0fb 100644
--- a/net/tipc/netlink.c
+++ b/net/tipc/netlink.c
@@ -44,7 +44,7 @@ static int handle_cmd(struct sk_buff *skb, struct genl_info *info)
struct nlmsghdr *rep_nlh;
struct nlmsghdr *req_nlh = info->nlhdr;
struct tipc_genlmsghdr *req_userhdr = info->userhdr;
- int hdr_space = NLMSG_SPACE(GENL_HDRLEN + TIPC_GENL_HDRLEN);
+ int hdr_space = nlmsg_total_size(GENL_HDRLEN + TIPC_GENL_HDRLEN);
u16 cmd;
if ((req_userhdr->cmd & 0xC000) && (!capable(CAP_NET_ADMIN)))
@@ -53,8 +53,8 @@ static int handle_cmd(struct sk_buff *skb, struct genl_info *info)
cmd = req_userhdr->cmd;
rep_buf = tipc_cfg_do_cmd(req_userhdr->dest, cmd,
- NLMSG_DATA(req_nlh) + GENL_HDRLEN + TIPC_GENL_HDRLEN,
- NLMSG_PAYLOAD(req_nlh, GENL_HDRLEN + TIPC_GENL_HDRLEN),
+ nlmsg_data(req_nlh) + GENL_HDRLEN + TIPC_GENL_HDRLEN,
+ nlmsg_attrlen(req_nlh, GENL_HDRLEN + TIPC_GENL_HDRLEN),
hdr_space);
if (rep_buf) {
@@ -62,7 +62,7 @@ static int handle_cmd(struct sk_buff *skb, struct genl_info *info)
rep_nlh = nlmsg_hdr(rep_buf);
memcpy(rep_nlh, req_nlh, hdr_space);
rep_nlh->nlmsg_len = rep_buf->len;
- genlmsg_unicast(&init_net, rep_buf, NETLINK_CB(skb).pid);
+ genlmsg_unicast(&init_net, rep_buf, NETLINK_CB(skb).portid);
}
return 0;
diff --git a/net/tipc/node.c b/net/tipc/node.c
index d21db204e25a..6e6c434872e8 100644
--- a/net/tipc/node.c
+++ b/net/tipc/node.c
@@ -1,7 +1,7 @@
/*
* net/tipc/node.c: TIPC node management routines
*
- * Copyright (c) 2000-2006, Ericsson AB
+ * Copyright (c) 2000-2006, 2012 Ericsson AB
* Copyright (c) 2005-2006, 2010-2011, Wind River Systems
* All rights reserved.
*
@@ -69,12 +69,11 @@ static unsigned int tipc_hashfn(u32 addr)
struct tipc_node *tipc_node_find(u32 addr)
{
struct tipc_node *node;
- struct hlist_node *pos;
if (unlikely(!in_own_cluster_exact(addr)))
return NULL;
- hlist_for_each_entry(node, pos, &node_htable[tipc_hashfn(addr)], hash) {
+ hlist_for_each_entry(node, &node_htable[tipc_hashfn(addr)], hash) {
if (node->addr == addr)
return node;
}
@@ -263,12 +262,9 @@ void tipc_node_detach_link(struct tipc_node *n_ptr, struct tipc_link *l_ptr)
static void node_established_contact(struct tipc_node *n_ptr)
{
tipc_k_signal((Handler)tipc_named_node_up, n_ptr->addr);
-
- if (n_ptr->bclink.supportable) {
- n_ptr->bclink.acked = tipc_bclink_get_last_sent();
- tipc_bclink_add_node(n_ptr->addr);
- n_ptr->bclink.supported = 1;
- }
+ n_ptr->bclink.oos_state = 0;
+ n_ptr->bclink.acked = tipc_bclink_get_last_sent();
+ tipc_bclink_add_node(n_ptr->addr);
}
static void node_name_purge_complete(unsigned long node_addr)
@@ -294,7 +290,7 @@ static void node_lost_contact(struct tipc_node *n_ptr)
tipc_addr_string_fill(addr_string, n_ptr->addr));
/* Flush broadcast link info associated with lost node */
- if (n_ptr->bclink.supported) {
+ if (n_ptr->bclink.recv_permitted) {
while (n_ptr->bclink.deferred_head) {
struct sk_buff *buf = n_ptr->bclink.deferred_head;
n_ptr->bclink.deferred_head = buf->next;
@@ -310,7 +306,7 @@ static void node_lost_contact(struct tipc_node *n_ptr)
tipc_bclink_remove_node(n_ptr->addr);
tipc_bclink_acknowledge(n_ptr, INVALID_LINK_SEQ);
- n_ptr->bclink.supported = 0;
+ n_ptr->bclink.recv_permitted = false;
}
/* Abort link changeover */
diff --git a/net/tipc/node.h b/net/tipc/node.h
index cfcaf4d6e480..3c189b35b102 100644
--- a/net/tipc/node.h
+++ b/net/tipc/node.h
@@ -67,8 +67,6 @@
* @permit_changeover: non-zero if node has redundant links to this system
* @signature: node instance identifier
* @bclink: broadcast-related info
- * @supportable: non-zero if node supports TIPC b'cast link capability
- * @supported: non-zero if node supports TIPC b'cast capability
* @acked: sequence # of last outbound b'cast message acknowledged by node
* @last_in: sequence # of last in-sequence b'cast message received from node
* @last_sent: sequence # of last b'cast message sent by node
@@ -77,6 +75,7 @@
* @deferred_head: oldest OOS b'cast message received from node
* @deferred_tail: newest OOS b'cast message received from node
* @defragm: list of partially reassembled b'cast message fragments from node
+ * @recv_permitted: true if node is allowed to receive b'cast messages
*/
struct tipc_node {
u32 addr;
@@ -92,8 +91,6 @@ struct tipc_node {
int permit_changeover;
u32 signature;
struct {
- u8 supportable;
- u8 supported;
u32 acked;
u32 last_in;
u32 last_sent;
@@ -102,6 +99,7 @@ struct tipc_node {
struct sk_buff *deferred_head;
struct sk_buff *deferred_tail;
struct sk_buff *defragm;
+ bool recv_permitted;
} bclink;
};
diff --git a/net/tipc/node_subscr.c b/net/tipc/node_subscr.c
index 5e34b015da45..8a7384c04add 100644
--- a/net/tipc/node_subscr.c
+++ b/net/tipc/node_subscr.c
@@ -42,7 +42,7 @@
* tipc_nodesub_subscribe - create "node down" subscription for specified node
*/
void tipc_nodesub_subscribe(struct tipc_node_subscr *node_sub, u32 addr,
- void *usr_handle, net_ev_handler handle_down)
+ void *usr_handle, net_ev_handler handle_down)
{
if (in_own_node(addr)) {
node_sub->node = NULL;
diff --git a/net/tipc/port.c b/net/tipc/port.c
index 07c42fba672b..b3ed2fcab4fb 100644
--- a/net/tipc/port.c
+++ b/net/tipc/port.c
@@ -2,7 +2,7 @@
* net/tipc/port.c: TIPC port code
*
* Copyright (c) 1992-2007, Ericsson AB
- * Copyright (c) 2004-2008, 2010-2011, Wind River Systems
+ * Copyright (c) 2004-2008, 2010-2013, Wind River Systems
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
@@ -46,11 +46,7 @@
#define MAX_REJECT_SIZE 1024
-static struct sk_buff *msg_queue_head;
-static struct sk_buff *msg_queue_tail;
-
DEFINE_SPINLOCK(tipc_port_list_lock);
-static DEFINE_SPINLOCK(queue_lock);
static LIST_HEAD(ports);
static void port_handle_node_down(unsigned long ref);
@@ -119,7 +115,7 @@ int tipc_multicast(u32 ref, struct tipc_name_seq const *seq,
msg_set_nameupper(hdr, seq->upper);
msg_set_hdr_sz(hdr, MCAST_H_SIZE);
res = tipc_msg_build(hdr, msg_sect, num_sect, total_len, MAX_MSG_SIZE,
- !oport->user_port, &buf);
+ &buf);
if (unlikely(!buf))
return res;
@@ -206,14 +202,15 @@ exit:
}
/**
- * tipc_createport_raw - create a generic TIPC port
+ * tipc_createport - create a generic TIPC port
*
* Returns pointer to (locked) TIPC port, or NULL if unable to create it
*/
-struct tipc_port *tipc_createport_raw(void *usr_handle,
- u32 (*dispatcher)(struct tipc_port *, struct sk_buff *),
- void (*wakeup)(struct tipc_port *),
- const u32 importance)
+struct tipc_port *tipc_createport(struct sock *sk,
+ u32 (*dispatcher)(struct tipc_port *,
+ struct sk_buff *),
+ void (*wakeup)(struct tipc_port *),
+ const u32 importance)
{
struct tipc_port *p_ptr;
struct tipc_msg *msg;
@@ -231,14 +228,13 @@ struct tipc_port *tipc_createport_raw(void *usr_handle,
return NULL;
}
- p_ptr->usr_handle = usr_handle;
+ p_ptr->sk = sk;
p_ptr->max_pkt = MAX_PKT_DEFAULT;
p_ptr->ref = ref;
INIT_LIST_HEAD(&p_ptr->wait_list);
INIT_LIST_HEAD(&p_ptr->subscription.nodesub_list);
p_ptr->dispatcher = dispatcher;
p_ptr->wakeup = wakeup;
- p_ptr->user_port = NULL;
k_init_timer(&p_ptr->timer, (Handler)port_timeout, ref);
INIT_LIST_HEAD(&p_ptr->publications);
INIT_LIST_HEAD(&p_ptr->port_list);
@@ -275,7 +271,6 @@ int tipc_deleteport(u32 ref)
buf = port_build_peer_abort_msg(p_ptr, TIPC_ERR_NO_PORT);
tipc_nodesub_unsubscribe(&p_ptr->subscription);
}
- kfree(p_ptr->user_port);
spin_lock_bh(&tipc_port_list_lock);
list_del(&p_ptr->port_list);
@@ -448,7 +443,7 @@ int tipc_port_reject_sections(struct tipc_port *p_ptr, struct tipc_msg *hdr,
int res;
res = tipc_msg_build(hdr, msg_sect, num_sect, total_len, MAX_MSG_SIZE,
- !p_ptr->user_port, &buf);
+ &buf);
if (!buf)
return res;
@@ -668,215 +663,6 @@ void tipc_port_reinit(void)
spin_unlock_bh(&tipc_port_list_lock);
}
-
-/*
- * port_dispatcher_sigh(): Signal handler for messages destinated
- * to the tipc_port interface.
- */
-static void port_dispatcher_sigh(void *dummy)
-{
- struct sk_buff *buf;
-
- spin_lock_bh(&queue_lock);
- buf = msg_queue_head;
- msg_queue_head = NULL;
- spin_unlock_bh(&queue_lock);
-
- while (buf) {
- struct tipc_port *p_ptr;
- struct user_port *up_ptr;
- struct tipc_portid orig;
- struct tipc_name_seq dseq;
- void *usr_handle;
- int connected;
- int peer_invalid;
- int published;
- u32 message_type;
-
- struct sk_buff *next = buf->next;
- struct tipc_msg *msg = buf_msg(buf);
- u32 dref = msg_destport(msg);
-
- message_type = msg_type(msg);
- if (message_type > TIPC_DIRECT_MSG)
- goto reject; /* Unsupported message type */
-
- p_ptr = tipc_port_lock(dref);
- if (!p_ptr)
- goto reject; /* Port deleted while msg in queue */
-
- orig.ref = msg_origport(msg);
- orig.node = msg_orignode(msg);
- up_ptr = p_ptr->user_port;
- usr_handle = up_ptr->usr_handle;
- connected = p_ptr->connected;
- peer_invalid = connected && !tipc_port_peer_msg(p_ptr, msg);
- published = p_ptr->published;
-
- if (unlikely(msg_errcode(msg)))
- goto err;
-
- switch (message_type) {
-
- case TIPC_CONN_MSG:{
- tipc_conn_msg_event cb = up_ptr->conn_msg_cb;
- u32 dsz;
-
- tipc_port_unlock(p_ptr);
- if (unlikely(!cb))
- goto reject;
- if (unlikely(!connected)) {
- if (tipc_connect2port(dref, &orig))
- goto reject;
- } else if (peer_invalid)
- goto reject;
- dsz = msg_data_sz(msg);
- if (unlikely(dsz &&
- (++p_ptr->conn_unacked >=
- TIPC_FLOW_CONTROL_WIN)))
- tipc_acknowledge(dref,
- p_ptr->conn_unacked);
- skb_pull(buf, msg_hdr_sz(msg));
- cb(usr_handle, dref, &buf, msg_data(msg), dsz);
- break;
- }
- case TIPC_DIRECT_MSG:{
- tipc_msg_event cb = up_ptr->msg_cb;
-
- tipc_port_unlock(p_ptr);
- if (unlikely(!cb || connected))
- goto reject;
- skb_pull(buf, msg_hdr_sz(msg));
- cb(usr_handle, dref, &buf, msg_data(msg),
- msg_data_sz(msg), msg_importance(msg),
- &orig);
- break;
- }
- case TIPC_MCAST_MSG:
- case TIPC_NAMED_MSG:{
- tipc_named_msg_event cb = up_ptr->named_msg_cb;
-
- tipc_port_unlock(p_ptr);
- if (unlikely(!cb || connected || !published))
- goto reject;
- dseq.type = msg_nametype(msg);
- dseq.lower = msg_nameinst(msg);
- dseq.upper = (message_type == TIPC_NAMED_MSG)
- ? dseq.lower : msg_nameupper(msg);
- skb_pull(buf, msg_hdr_sz(msg));
- cb(usr_handle, dref, &buf, msg_data(msg),
- msg_data_sz(msg), msg_importance(msg),
- &orig, &dseq);
- break;
- }
- }
- if (buf)
- kfree_skb(buf);
- buf = next;
- continue;
-err:
- switch (message_type) {
-
- case TIPC_CONN_MSG:{
- tipc_conn_shutdown_event cb =
- up_ptr->conn_err_cb;
-
- tipc_port_unlock(p_ptr);
- if (!cb || !connected || peer_invalid)
- break;
- tipc_disconnect(dref);
- skb_pull(buf, msg_hdr_sz(msg));
- cb(usr_handle, dref, &buf, msg_data(msg),
- msg_data_sz(msg), msg_errcode(msg));
- break;
- }
- case TIPC_DIRECT_MSG:{
- tipc_msg_err_event cb = up_ptr->err_cb;
-
- tipc_port_unlock(p_ptr);
- if (!cb || connected)
- break;
- skb_pull(buf, msg_hdr_sz(msg));
- cb(usr_handle, dref, &buf, msg_data(msg),
- msg_data_sz(msg), msg_errcode(msg), &orig);
- break;
- }
- case TIPC_MCAST_MSG:
- case TIPC_NAMED_MSG:{
- tipc_named_msg_err_event cb =
- up_ptr->named_err_cb;
-
- tipc_port_unlock(p_ptr);
- if (!cb || connected)
- break;
- dseq.type = msg_nametype(msg);
- dseq.lower = msg_nameinst(msg);
- dseq.upper = (message_type == TIPC_NAMED_MSG)
- ? dseq.lower : msg_nameupper(msg);
- skb_pull(buf, msg_hdr_sz(msg));
- cb(usr_handle, dref, &buf, msg_data(msg),
- msg_data_sz(msg), msg_errcode(msg), &dseq);
- break;
- }
- }
- if (buf)
- kfree_skb(buf);
- buf = next;
- continue;
-reject:
- tipc_reject_msg(buf, TIPC_ERR_NO_PORT);
- buf = next;
- }
-}
-
-/*
- * port_dispatcher(): Dispatcher for messages destinated
- * to the tipc_port interface. Called with port locked.
- */
-static u32 port_dispatcher(struct tipc_port *dummy, struct sk_buff *buf)
-{
- buf->next = NULL;
- spin_lock_bh(&queue_lock);
- if (msg_queue_head) {
- msg_queue_tail->next = buf;
- msg_queue_tail = buf;
- } else {
- msg_queue_tail = msg_queue_head = buf;
- tipc_k_signal((Handler)port_dispatcher_sigh, 0);
- }
- spin_unlock_bh(&queue_lock);
- return 0;
-}
-
-/*
- * Wake up port after congestion: Called with port locked
- */
-static void port_wakeup_sh(unsigned long ref)
-{
- struct tipc_port *p_ptr;
- struct user_port *up_ptr;
- tipc_continue_event cb = NULL;
- void *uh = NULL;
-
- p_ptr = tipc_port_lock(ref);
- if (p_ptr) {
- up_ptr = p_ptr->user_port;
- if (up_ptr) {
- cb = up_ptr->continue_event_cb;
- uh = up_ptr->usr_handle;
- }
- tipc_port_unlock(p_ptr);
- }
- if (cb)
- cb(uh, ref);
-}
-
-
-static void port_wakeup(struct tipc_port *p_ptr)
-{
- tipc_k_signal((Handler)port_wakeup_sh, p_ptr->ref);
-}
-
void tipc_acknowledge(u32 ref, u32 ack)
{
struct tipc_port *p_ptr;
@@ -893,50 +679,6 @@ void tipc_acknowledge(u32 ref, u32 ack)
tipc_net_route_msg(buf);
}
-/*
- * tipc_createport(): user level call.
- */
-int tipc_createport(void *usr_handle,
- unsigned int importance,
- tipc_msg_err_event error_cb,
- tipc_named_msg_err_event named_error_cb,
- tipc_conn_shutdown_event conn_error_cb,
- tipc_msg_event msg_cb,
- tipc_named_msg_event named_msg_cb,
- tipc_conn_msg_event conn_msg_cb,
- tipc_continue_event continue_event_cb, /* May be zero */
- u32 *portref)
-{
- struct user_port *up_ptr;
- struct tipc_port *p_ptr;
-
- up_ptr = kmalloc(sizeof(*up_ptr), GFP_ATOMIC);
- if (!up_ptr) {
- pr_warn("Port creation failed, no memory\n");
- return -ENOMEM;
- }
- p_ptr = tipc_createport_raw(NULL, port_dispatcher, port_wakeup,
- importance);
- if (!p_ptr) {
- kfree(up_ptr);
- return -ENOMEM;
- }
-
- p_ptr->user_port = up_ptr;
- up_ptr->usr_handle = usr_handle;
- up_ptr->ref = p_ptr->ref;
- up_ptr->err_cb = error_cb;
- up_ptr->named_err_cb = named_error_cb;
- up_ptr->conn_err_cb = conn_error_cb;
- up_ptr->msg_cb = msg_cb;
- up_ptr->named_msg_cb = named_msg_cb;
- up_ptr->conn_msg_cb = conn_msg_cb;
- up_ptr->continue_event_cb = continue_event_cb;
- *portref = p_ptr->ref;
- tipc_port_unlock(p_ptr);
- return 0;
-}
-
int tipc_portimportance(u32 ref, unsigned int *importance)
{
struct tipc_port *p_ptr;
@@ -1036,15 +778,30 @@ int tipc_withdraw(u32 ref, unsigned int scope, struct tipc_name_seq const *seq)
return res;
}
-int tipc_connect2port(u32 ref, struct tipc_portid const *peer)
+int tipc_connect(u32 ref, struct tipc_portid const *peer)
{
struct tipc_port *p_ptr;
- struct tipc_msg *msg;
- int res = -EINVAL;
+ int res;
p_ptr = tipc_port_lock(ref);
if (!p_ptr)
return -EINVAL;
+ res = __tipc_connect(ref, p_ptr, peer);
+ tipc_port_unlock(p_ptr);
+ return res;
+}
+
+/*
+ * __tipc_connect - connect to a remote peer
+ *
+ * Port must be locked.
+ */
+int __tipc_connect(u32 ref, struct tipc_port *p_ptr,
+ struct tipc_portid const *peer)
+{
+ struct tipc_msg *msg;
+ int res = -EINVAL;
+
if (p_ptr->published || p_ptr->connected)
goto exit;
if (!peer->ref)
@@ -1067,17 +824,16 @@ int tipc_connect2port(u32 ref, struct tipc_portid const *peer)
(net_ev_handler)port_handle_node_down);
res = 0;
exit:
- tipc_port_unlock(p_ptr);
p_ptr->max_pkt = tipc_link_get_max_pkt(peer->node, ref);
return res;
}
-/**
- * tipc_disconnect_port - disconnect port from peer
+/*
+ * __tipc_disconnect - disconnect port from peer
*
* Port must be locked.
*/
-int tipc_disconnect_port(struct tipc_port *tp_ptr)
+int __tipc_disconnect(struct tipc_port *tp_ptr)
{
int res;
@@ -1104,7 +860,7 @@ int tipc_disconnect(u32 ref)
p_ptr = tipc_port_lock(ref);
if (!p_ptr)
return -EINVAL;
- res = tipc_disconnect_port(p_ptr);
+ res = __tipc_disconnect(p_ptr);
tipc_port_unlock(p_ptr);
return res;
}
@@ -1170,7 +926,7 @@ static int tipc_port_recv_sections(struct tipc_port *sender, unsigned int num_se
int res;
res = tipc_msg_build(&sender->phdr, msg_sect, num_sect, total_len,
- MAX_MSG_SIZE, !sender->user_port, &buf);
+ MAX_MSG_SIZE, &buf);
if (likely(buf))
tipc_port_recv_msg(buf);
return res;
@@ -1308,43 +1064,3 @@ int tipc_send2port(u32 ref, struct tipc_portid const *dest,
}
return -ELINKCONG;
}
-
-/**
- * tipc_send_buf2port - send message buffer to port identity
- */
-int tipc_send_buf2port(u32 ref, struct tipc_portid const *dest,
- struct sk_buff *buf, unsigned int dsz)
-{
- struct tipc_port *p_ptr;
- struct tipc_msg *msg;
- int res;
-
- p_ptr = (struct tipc_port *)tipc_ref_deref(ref);
- if (!p_ptr || p_ptr->connected)
- return -EINVAL;
-
- msg = &p_ptr->phdr;
- msg_set_type(msg, TIPC_DIRECT_MSG);
- msg_set_destnode(msg, dest->node);
- msg_set_destport(msg, dest->ref);
- msg_set_hdr_sz(msg, BASIC_H_SIZE);
- msg_set_size(msg, BASIC_H_SIZE + dsz);
- if (skb_cow(buf, BASIC_H_SIZE))
- return -ENOMEM;
-
- skb_push(buf, BASIC_H_SIZE);
- skb_copy_to_linear_data(buf, msg, BASIC_H_SIZE);
-
- if (in_own_node(dest->node))
- res = tipc_port_recv_msg(buf);
- else
- res = tipc_send_buf_fast(buf, dest->node);
- if (likely(res != -ELINKCONG)) {
- if (res > 0)
- p_ptr->sent++;
- return res;
- }
- if (port_unreliable(p_ptr))
- return dsz;
- return -ELINKCONG;
-}
diff --git a/net/tipc/port.h b/net/tipc/port.h
index 4660e3065790..5a7026b9c345 100644
--- a/net/tipc/port.h
+++ b/net/tipc/port.h
@@ -2,7 +2,7 @@
* net/tipc/port.h: Include file for TIPC port code
*
* Copyright (c) 1994-2007, Ericsson AB
- * Copyright (c) 2004-2007, 2010-2011, Wind River Systems
+ * Copyright (c) 2004-2007, 2010-2013, Wind River Systems
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
@@ -43,60 +43,12 @@
#include "node_subscr.h"
#define TIPC_FLOW_CONTROL_WIN 512
-
-typedef void (*tipc_msg_err_event) (void *usr_handle, u32 portref,
- struct sk_buff **buf, unsigned char const *data,
- unsigned int size, int reason,
- struct tipc_portid const *attmpt_destid);
-
-typedef void (*tipc_named_msg_err_event) (void *usr_handle, u32 portref,
- struct sk_buff **buf, unsigned char const *data,
- unsigned int size, int reason,
- struct tipc_name_seq const *attmpt_dest);
-
-typedef void (*tipc_conn_shutdown_event) (void *usr_handle, u32 portref,
- struct sk_buff **buf, unsigned char const *data,
- unsigned int size, int reason);
-
-typedef void (*tipc_msg_event) (void *usr_handle, u32 portref,
- struct sk_buff **buf, unsigned char const *data,
- unsigned int size, unsigned int importance,
- struct tipc_portid const *origin);
-
-typedef void (*tipc_named_msg_event) (void *usr_handle, u32 portref,
- struct sk_buff **buf, unsigned char const *data,
- unsigned int size, unsigned int importance,
- struct tipc_portid const *orig,
- struct tipc_name_seq const *dest);
-
-typedef void (*tipc_conn_msg_event) (void *usr_handle, u32 portref,
- struct sk_buff **buf, unsigned char const *data,
- unsigned int size);
-
-typedef void (*tipc_continue_event) (void *usr_handle, u32 portref);
-
-/**
- * struct user_port - TIPC user port (used with native API)
- * @usr_handle: user-specified field
- * @ref: object reference to associated TIPC port
- *
- * <various callback routines>
- */
-struct user_port {
- void *usr_handle;
- u32 ref;
- tipc_msg_err_event err_cb;
- tipc_named_msg_err_event named_err_cb;
- tipc_conn_shutdown_event conn_err_cb;
- tipc_msg_event msg_cb;
- tipc_named_msg_event named_msg_cb;
- tipc_conn_msg_event conn_msg_cb;
- tipc_continue_event continue_event_cb;
-};
+#define CONN_OVERLOAD_LIMIT ((TIPC_FLOW_CONTROL_WIN * 2 + 1) * \
+ SKB_TRUESIZE(TIPC_MAX_USER_MSG_SIZE))
/**
* struct tipc_port - TIPC port structure
- * @usr_handle: pointer to additional user-defined information about port
+ * @sk: pointer to socket handle
* @lock: pointer to spinlock for controlling access to port
* @connected: non-zero if port is currently connected to a peer port
* @conn_type: TIPC type used when connection was established
@@ -110,7 +62,6 @@ struct user_port {
* @port_list: adjacent ports in TIPC's global list of ports
* @dispatcher: ptr to routine which handles received messages
* @wakeup: ptr to routine to call when port is no longer congested
- * @user_port: ptr to user port associated with port (if any)
* @wait_list: adjacent ports in list of ports waiting on link congestion
* @waiting_pkts:
* @sent: # of non-empty messages sent by port
@@ -123,7 +74,7 @@ struct user_port {
* @subscription: "node down" subscription used to terminate failed connections
*/
struct tipc_port {
- void *usr_handle;
+ struct sock *sk;
spinlock_t *lock;
int connected;
u32 conn_type;
@@ -137,7 +88,6 @@ struct tipc_port {
struct list_head port_list;
u32 (*dispatcher)(struct tipc_port *, struct sk_buff *);
void (*wakeup)(struct tipc_port *);
- struct user_port *user_port;
struct list_head wait_list;
u32 waiting_pkts;
u32 sent;
@@ -156,24 +106,16 @@ struct tipc_port_list;
/*
* TIPC port manipulation routines
*/
-struct tipc_port *tipc_createport_raw(void *usr_handle,
- u32 (*dispatcher)(struct tipc_port *, struct sk_buff *),
- void (*wakeup)(struct tipc_port *), const u32 importance);
+struct tipc_port *tipc_createport(struct sock *sk,
+ u32 (*dispatcher)(struct tipc_port *,
+ struct sk_buff *),
+ void (*wakeup)(struct tipc_port *),
+ const u32 importance);
int tipc_reject_msg(struct sk_buff *buf, u32 err);
-int tipc_send_buf_fast(struct sk_buff *buf, u32 destnode);
-
void tipc_acknowledge(u32 port_ref, u32 ack);
-int tipc_createport(void *usr_handle,
- unsigned int importance, tipc_msg_err_event error_cb,
- tipc_named_msg_err_event named_error_cb,
- tipc_conn_shutdown_event conn_error_cb, tipc_msg_event msg_cb,
- tipc_named_msg_event named_msg_cb,
- tipc_conn_msg_event conn_msg_cb,
- tipc_continue_event continue_event_cb, u32 *portref);
-
int tipc_deleteport(u32 portref);
int tipc_portimportance(u32 portref, unsigned int *importance);
@@ -186,11 +128,11 @@ int tipc_portunreturnable(u32 portref, unsigned int *isunreturnable);
int tipc_set_portunreturnable(u32 portref, unsigned int isunreturnable);
int tipc_publish(u32 portref, unsigned int scope,
- struct tipc_name_seq const *name_seq);
+ struct tipc_name_seq const *name_seq);
int tipc_withdraw(u32 portref, unsigned int scope,
- struct tipc_name_seq const *name_seq);
+ struct tipc_name_seq const *name_seq);
-int tipc_connect2port(u32 portref, struct tipc_portid const *port);
+int tipc_connect(u32 portref, struct tipc_portid const *port);
int tipc_disconnect(u32 portref);
@@ -200,7 +142,9 @@ int tipc_shutdown(u32 ref);
/*
* The following routines require that the port be locked on entry
*/
-int tipc_disconnect_port(struct tipc_port *tp_ptr);
+int __tipc_disconnect(struct tipc_port *tp_ptr);
+int __tipc_connect(u32 ref, struct tipc_port *p_ptr,
+ struct tipc_portid const *peer);
int tipc_port_peer_msg(struct tipc_port *p_ptr, struct tipc_msg *msg);
/*
@@ -218,9 +162,6 @@ int tipc_send2port(u32 portref, struct tipc_portid const *dest,
unsigned int num_sect, struct iovec const *msg_sect,
unsigned int total_len);
-int tipc_send_buf2port(u32 portref, struct tipc_portid const *dest,
- struct sk_buff *buf, unsigned int dsz);
-
int tipc_multicast(u32 portref, struct tipc_name_seq const *seq,
unsigned int section_count, struct iovec const *msg,
unsigned int total_len);
diff --git a/net/tipc/server.c b/net/tipc/server.c
new file mode 100644
index 000000000000..fd3fa57a410e
--- /dev/null
+++ b/net/tipc/server.c
@@ -0,0 +1,605 @@
+/*
+ * net/tipc/server.c: TIPC server infrastructure
+ *
+ * Copyright (c) 2012-2013, 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 "server.h"
+#include "core.h"
+#include <net/sock.h>
+
+/* Number of messages to send before rescheduling */
+#define MAX_SEND_MSG_COUNT 25
+#define MAX_RECV_MSG_COUNT 25
+#define CF_CONNECTED 1
+
+#define sock2con(x) ((struct tipc_conn *)(x)->sk_user_data)
+
+/**
+ * struct tipc_conn - TIPC connection structure
+ * @kref: reference counter to connection object
+ * @conid: connection identifier
+ * @sock: socket handler associated with connection
+ * @flags: indicates connection state
+ * @server: pointer to connected server
+ * @rwork: receive work item
+ * @usr_data: user-specified field
+ * @rx_action: what to do when connection socket is active
+ * @outqueue: pointer to first outbound message in queue
+ * @outqueue_lock: controll access to the outqueue
+ * @outqueue: list of connection objects for its server
+ * @swork: send work item
+ */
+struct tipc_conn {
+ struct kref kref;
+ int conid;
+ struct socket *sock;
+ unsigned long flags;
+ struct tipc_server *server;
+ struct work_struct rwork;
+ int (*rx_action) (struct tipc_conn *con);
+ void *usr_data;
+ struct list_head outqueue;
+ spinlock_t outqueue_lock;
+ struct work_struct swork;
+};
+
+/* An entry waiting to be sent */
+struct outqueue_entry {
+ struct list_head list;
+ struct kvec iov;
+ struct sockaddr_tipc dest;
+};
+
+static void tipc_recv_work(struct work_struct *work);
+static void tipc_send_work(struct work_struct *work);
+static void tipc_clean_outqueues(struct tipc_conn *con);
+
+static void tipc_conn_kref_release(struct kref *kref)
+{
+ struct tipc_conn *con = container_of(kref, struct tipc_conn, kref);
+ struct tipc_server *s = con->server;
+
+ if (con->sock) {
+ tipc_sock_release_local(con->sock);
+ con->sock = NULL;
+ }
+
+ tipc_clean_outqueues(con);
+
+ if (con->conid)
+ s->tipc_conn_shutdown(con->conid, con->usr_data);
+
+ kfree(con);
+}
+
+static void conn_put(struct tipc_conn *con)
+{
+ kref_put(&con->kref, tipc_conn_kref_release);
+}
+
+static void conn_get(struct tipc_conn *con)
+{
+ kref_get(&con->kref);
+}
+
+static struct tipc_conn *tipc_conn_lookup(struct tipc_server *s, int conid)
+{
+ struct tipc_conn *con;
+
+ spin_lock_bh(&s->idr_lock);
+ con = idr_find(&s->conn_idr, conid);
+ if (con)
+ conn_get(con);
+ spin_unlock_bh(&s->idr_lock);
+ return con;
+}
+
+static void sock_data_ready(struct sock *sk, int unused)
+{
+ struct tipc_conn *con;
+
+ read_lock(&sk->sk_callback_lock);
+ con = sock2con(sk);
+ if (con && test_bit(CF_CONNECTED, &con->flags)) {
+ conn_get(con);
+ if (!queue_work(con->server->rcv_wq, &con->rwork))
+ conn_put(con);
+ }
+ read_unlock(&sk->sk_callback_lock);
+}
+
+static void sock_write_space(struct sock *sk)
+{
+ struct tipc_conn *con;
+
+ read_lock(&sk->sk_callback_lock);
+ con = sock2con(sk);
+ if (con && test_bit(CF_CONNECTED, &con->flags)) {
+ conn_get(con);
+ if (!queue_work(con->server->send_wq, &con->swork))
+ conn_put(con);
+ }
+ read_unlock(&sk->sk_callback_lock);
+}
+
+static void tipc_register_callbacks(struct socket *sock, struct tipc_conn *con)
+{
+ struct sock *sk = sock->sk;
+
+ write_lock_bh(&sk->sk_callback_lock);
+
+ sk->sk_data_ready = sock_data_ready;
+ sk->sk_write_space = sock_write_space;
+ sk->sk_user_data = con;
+
+ con->sock = sock;
+
+ write_unlock_bh(&sk->sk_callback_lock);
+}
+
+static void tipc_unregister_callbacks(struct tipc_conn *con)
+{
+ struct sock *sk = con->sock->sk;
+
+ write_lock_bh(&sk->sk_callback_lock);
+ sk->sk_user_data = NULL;
+ write_unlock_bh(&sk->sk_callback_lock);
+}
+
+static void tipc_close_conn(struct tipc_conn *con)
+{
+ struct tipc_server *s = con->server;
+
+ if (test_and_clear_bit(CF_CONNECTED, &con->flags)) {
+ spin_lock_bh(&s->idr_lock);
+ idr_remove(&s->conn_idr, con->conid);
+ s->idr_in_use--;
+ spin_unlock_bh(&s->idr_lock);
+
+ tipc_unregister_callbacks(con);
+
+ /* We shouldn't flush pending works as we may be in the
+ * thread. In fact the races with pending rx/tx work structs
+ * are harmless for us here as we have already deleted this
+ * connection from server connection list and set
+ * sk->sk_user_data to 0 before releasing connection object.
+ */
+ kernel_sock_shutdown(con->sock, SHUT_RDWR);
+
+ conn_put(con);
+ }
+}
+
+static struct tipc_conn *tipc_alloc_conn(struct tipc_server *s)
+{
+ struct tipc_conn *con;
+ int ret;
+
+ con = kzalloc(sizeof(struct tipc_conn), GFP_ATOMIC);
+ if (!con)
+ return ERR_PTR(-ENOMEM);
+
+ kref_init(&con->kref);
+ INIT_LIST_HEAD(&con->outqueue);
+ spin_lock_init(&con->outqueue_lock);
+ INIT_WORK(&con->swork, tipc_send_work);
+ INIT_WORK(&con->rwork, tipc_recv_work);
+
+ spin_lock_bh(&s->idr_lock);
+ ret = idr_alloc(&s->conn_idr, con, 0, 0, GFP_ATOMIC);
+ if (ret < 0) {
+ kfree(con);
+ spin_unlock_bh(&s->idr_lock);
+ return ERR_PTR(-ENOMEM);
+ }
+ con->conid = ret;
+ s->idr_in_use++;
+ spin_unlock_bh(&s->idr_lock);
+
+ set_bit(CF_CONNECTED, &con->flags);
+ con->server = s;
+
+ return con;
+}
+
+static int tipc_receive_from_sock(struct tipc_conn *con)
+{
+ struct msghdr msg = {};
+ struct tipc_server *s = con->server;
+ struct sockaddr_tipc addr;
+ struct kvec iov;
+ void *buf;
+ int ret;
+
+ buf = kmem_cache_alloc(s->rcvbuf_cache, GFP_ATOMIC);
+ if (!buf) {
+ ret = -ENOMEM;
+ goto out_close;
+ }
+
+ iov.iov_base = buf;
+ iov.iov_len = s->max_rcvbuf_size;
+ msg.msg_name = &addr;
+ ret = kernel_recvmsg(con->sock, &msg, &iov, 1, iov.iov_len,
+ MSG_DONTWAIT);
+ if (ret <= 0) {
+ kmem_cache_free(s->rcvbuf_cache, buf);
+ goto out_close;
+ }
+
+ s->tipc_conn_recvmsg(con->conid, &addr, con->usr_data, buf, ret);
+
+ kmem_cache_free(s->rcvbuf_cache, buf);
+
+ return 0;
+
+out_close:
+ if (ret != -EWOULDBLOCK)
+ tipc_close_conn(con);
+ else if (ret == 0)
+ /* Don't return success if we really got EOF */
+ ret = -EAGAIN;
+
+ return ret;
+}
+
+static int tipc_accept_from_sock(struct tipc_conn *con)
+{
+ struct tipc_server *s = con->server;
+ struct socket *sock = con->sock;
+ struct socket *newsock;
+ struct tipc_conn *newcon;
+ int ret;
+
+ ret = tipc_sock_accept_local(sock, &newsock, O_NONBLOCK);
+ if (ret < 0)
+ return ret;
+
+ newcon = tipc_alloc_conn(con->server);
+ if (IS_ERR(newcon)) {
+ ret = PTR_ERR(newcon);
+ sock_release(newsock);
+ return ret;
+ }
+
+ newcon->rx_action = tipc_receive_from_sock;
+ tipc_register_callbacks(newsock, newcon);
+
+ /* Notify that new connection is incoming */
+ newcon->usr_data = s->tipc_conn_new(newcon->conid);
+
+ /* Wake up receive process in case of 'SYN+' message */
+ newsock->sk->sk_data_ready(newsock->sk, 0);
+ return ret;
+}
+
+static struct socket *tipc_create_listen_sock(struct tipc_conn *con)
+{
+ struct tipc_server *s = con->server;
+ struct socket *sock = NULL;
+ int ret;
+
+ ret = tipc_sock_create_local(s->type, &sock);
+ if (ret < 0)
+ return NULL;
+ ret = kernel_setsockopt(sock, SOL_TIPC, TIPC_IMPORTANCE,
+ (char *)&s->imp, sizeof(s->imp));
+ if (ret < 0)
+ goto create_err;
+ ret = kernel_bind(sock, (struct sockaddr *)s->saddr, sizeof(*s->saddr));
+ if (ret < 0)
+ goto create_err;
+
+ switch (s->type) {
+ case SOCK_STREAM:
+ case SOCK_SEQPACKET:
+ con->rx_action = tipc_accept_from_sock;
+
+ ret = kernel_listen(sock, 0);
+ if (ret < 0)
+ goto create_err;
+ break;
+ case SOCK_DGRAM:
+ case SOCK_RDM:
+ con->rx_action = tipc_receive_from_sock;
+ break;
+ default:
+ pr_err("Unknown socket type %d\n", s->type);
+ goto create_err;
+ }
+ return sock;
+
+create_err:
+ sock_release(sock);
+ con->sock = NULL;
+ return NULL;
+}
+
+static int tipc_open_listening_sock(struct tipc_server *s)
+{
+ struct socket *sock;
+ struct tipc_conn *con;
+
+ con = tipc_alloc_conn(s);
+ if (IS_ERR(con))
+ return PTR_ERR(con);
+
+ sock = tipc_create_listen_sock(con);
+ if (!sock) {
+ idr_remove(&s->conn_idr, con->conid);
+ s->idr_in_use--;
+ kfree(con);
+ return -EINVAL;
+ }
+
+ tipc_register_callbacks(sock, con);
+ return 0;
+}
+
+static struct outqueue_entry *tipc_alloc_entry(void *data, int len)
+{
+ struct outqueue_entry *entry;
+ void *buf;
+
+ entry = kmalloc(sizeof(struct outqueue_entry), GFP_ATOMIC);
+ if (!entry)
+ return NULL;
+
+ buf = kmalloc(len, GFP_ATOMIC);
+ if (!buf) {
+ kfree(entry);
+ return NULL;
+ }
+
+ memcpy(buf, data, len);
+ entry->iov.iov_base = buf;
+ entry->iov.iov_len = len;
+
+ return entry;
+}
+
+static void tipc_free_entry(struct outqueue_entry *e)
+{
+ kfree(e->iov.iov_base);
+ kfree(e);
+}
+
+static void tipc_clean_outqueues(struct tipc_conn *con)
+{
+ struct outqueue_entry *e, *safe;
+
+ spin_lock_bh(&con->outqueue_lock);
+ list_for_each_entry_safe(e, safe, &con->outqueue, list) {
+ list_del(&e->list);
+ tipc_free_entry(e);
+ }
+ spin_unlock_bh(&con->outqueue_lock);
+}
+
+int tipc_conn_sendmsg(struct tipc_server *s, int conid,
+ struct sockaddr_tipc *addr, void *data, size_t len)
+{
+ struct outqueue_entry *e;
+ struct tipc_conn *con;
+
+ con = tipc_conn_lookup(s, conid);
+ if (!con)
+ return -EINVAL;
+
+ e = tipc_alloc_entry(data, len);
+ if (!e) {
+ conn_put(con);
+ return -ENOMEM;
+ }
+
+ if (addr)
+ memcpy(&e->dest, addr, sizeof(struct sockaddr_tipc));
+
+ spin_lock_bh(&con->outqueue_lock);
+ list_add_tail(&e->list, &con->outqueue);
+ spin_unlock_bh(&con->outqueue_lock);
+
+ if (test_bit(CF_CONNECTED, &con->flags))
+ if (!queue_work(s->send_wq, &con->swork))
+ conn_put(con);
+
+ return 0;
+}
+
+void tipc_conn_terminate(struct tipc_server *s, int conid)
+{
+ struct tipc_conn *con;
+
+ con = tipc_conn_lookup(s, conid);
+ if (con) {
+ tipc_close_conn(con);
+ conn_put(con);
+ }
+}
+
+static void tipc_send_to_sock(struct tipc_conn *con)
+{
+ int count = 0;
+ struct tipc_server *s = con->server;
+ struct outqueue_entry *e;
+ struct msghdr msg;
+ int ret;
+
+ spin_lock_bh(&con->outqueue_lock);
+ while (1) {
+ e = list_entry(con->outqueue.next, struct outqueue_entry,
+ list);
+ if ((struct list_head *) e == &con->outqueue)
+ break;
+ spin_unlock_bh(&con->outqueue_lock);
+
+ memset(&msg, 0, sizeof(msg));
+ msg.msg_flags = MSG_DONTWAIT;
+
+ if (s->type == SOCK_DGRAM || s->type == SOCK_RDM) {
+ msg.msg_name = &e->dest;
+ msg.msg_namelen = sizeof(struct sockaddr_tipc);
+ }
+ ret = kernel_sendmsg(con->sock, &msg, &e->iov, 1,
+ e->iov.iov_len);
+ if (ret == -EWOULDBLOCK || ret == 0) {
+ cond_resched();
+ goto out;
+ } else if (ret < 0) {
+ goto send_err;
+ }
+
+ /* Don't starve users filling buffers */
+ if (++count >= MAX_SEND_MSG_COUNT) {
+ cond_resched();
+ count = 0;
+ }
+
+ spin_lock_bh(&con->outqueue_lock);
+ list_del(&e->list);
+ tipc_free_entry(e);
+ }
+ spin_unlock_bh(&con->outqueue_lock);
+out:
+ return;
+
+send_err:
+ tipc_close_conn(con);
+}
+
+static void tipc_recv_work(struct work_struct *work)
+{
+ struct tipc_conn *con = container_of(work, struct tipc_conn, rwork);
+ int count = 0;
+
+ while (test_bit(CF_CONNECTED, &con->flags)) {
+ if (con->rx_action(con))
+ break;
+
+ /* Don't flood Rx machine */
+ if (++count >= MAX_RECV_MSG_COUNT) {
+ cond_resched();
+ count = 0;
+ }
+ }
+ conn_put(con);
+}
+
+static void tipc_send_work(struct work_struct *work)
+{
+ struct tipc_conn *con = container_of(work, struct tipc_conn, swork);
+
+ if (test_bit(CF_CONNECTED, &con->flags))
+ tipc_send_to_sock(con);
+
+ conn_put(con);
+}
+
+static void tipc_work_stop(struct tipc_server *s)
+{
+ destroy_workqueue(s->rcv_wq);
+ destroy_workqueue(s->send_wq);
+}
+
+static int tipc_work_start(struct tipc_server *s)
+{
+ s->rcv_wq = alloc_workqueue("tipc_rcv", WQ_UNBOUND, 1);
+ if (!s->rcv_wq) {
+ pr_err("can't start tipc receive workqueue\n");
+ return -ENOMEM;
+ }
+
+ s->send_wq = alloc_workqueue("tipc_send", WQ_UNBOUND, 1);
+ if (!s->send_wq) {
+ pr_err("can't start tipc send workqueue\n");
+ destroy_workqueue(s->rcv_wq);
+ return -ENOMEM;
+ }
+
+ return 0;
+}
+
+int tipc_server_start(struct tipc_server *s)
+{
+ int ret;
+
+ spin_lock_init(&s->idr_lock);
+ idr_init(&s->conn_idr);
+ s->idr_in_use = 0;
+
+ s->rcvbuf_cache = kmem_cache_create(s->name, s->max_rcvbuf_size,
+ 0, SLAB_HWCACHE_ALIGN, NULL);
+ if (!s->rcvbuf_cache)
+ return -ENOMEM;
+
+ ret = tipc_work_start(s);
+ if (ret < 0) {
+ kmem_cache_destroy(s->rcvbuf_cache);
+ return ret;
+ }
+ ret = tipc_open_listening_sock(s);
+ if (ret < 0) {
+ tipc_work_stop(s);
+ kmem_cache_destroy(s->rcvbuf_cache);
+ return ret;
+ }
+ s->enabled = 1;
+ return ret;
+}
+
+void tipc_server_stop(struct tipc_server *s)
+{
+ struct tipc_conn *con;
+ int total = 0;
+ int id;
+
+ if (!s->enabled)
+ return;
+
+ s->enabled = 0;
+ spin_lock_bh(&s->idr_lock);
+ for (id = 0; total < s->idr_in_use; id++) {
+ con = idr_find(&s->conn_idr, id);
+ if (con) {
+ total++;
+ spin_unlock_bh(&s->idr_lock);
+ tipc_close_conn(con);
+ spin_lock_bh(&s->idr_lock);
+ }
+ }
+ spin_unlock_bh(&s->idr_lock);
+
+ tipc_work_stop(s);
+ kmem_cache_destroy(s->rcvbuf_cache);
+ idr_destroy(&s->conn_idr);
+}
diff --git a/net/tipc/server.h b/net/tipc/server.h
new file mode 100644
index 000000000000..98b23f20bc0f
--- /dev/null
+++ b/net/tipc/server.h
@@ -0,0 +1,94 @@
+/*
+ * net/tipc/server.h: Include file for TIPC server code
+ *
+ * Copyright (c) 2012-2013, 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.
+ */
+
+#ifndef _TIPC_SERVER_H
+#define _TIPC_SERVER_H
+
+#include "core.h"
+
+#define TIPC_SERVER_NAME_LEN 32
+
+/**
+ * struct tipc_server - TIPC server structure
+ * @conn_idr: identifier set of connection
+ * @idr_lock: protect the connection identifier set
+ * @idr_in_use: amount of allocated identifier entry
+ * @rcvbuf_cache: memory cache of server receive buffer
+ * @rcv_wq: receive workqueue
+ * @send_wq: send workqueue
+ * @max_rcvbuf_size: maximum permitted receive message length
+ * @tipc_conn_new: callback will be called when new connection is incoming
+ * @tipc_conn_shutdown: callback will be called when connection is shut down
+ * @tipc_conn_recvmsg: callback will be called when message arrives
+ * @saddr: TIPC server address
+ * @name: server name
+ * @imp: message importance
+ * @type: socket type
+ * @enabled: identify whether server is launched or not
+ */
+struct tipc_server {
+ struct idr conn_idr;
+ spinlock_t idr_lock;
+ int idr_in_use;
+ struct kmem_cache *rcvbuf_cache;
+ struct workqueue_struct *rcv_wq;
+ struct workqueue_struct *send_wq;
+ int max_rcvbuf_size;
+ void *(*tipc_conn_new) (int conid);
+ void (*tipc_conn_shutdown) (int conid, void *usr_data);
+ void (*tipc_conn_recvmsg) (int conid, struct sockaddr_tipc *addr,
+ void *usr_data, void *buf, size_t len);
+ struct sockaddr_tipc *saddr;
+ const char name[TIPC_SERVER_NAME_LEN];
+ int imp;
+ int type;
+ int enabled;
+};
+
+int tipc_conn_sendmsg(struct tipc_server *s, int conid,
+ struct sockaddr_tipc *addr, void *data, size_t len);
+
+/**
+ * tipc_conn_terminate - terminate connection with server
+ *
+ * Note: Must call it in process context since it might sleep
+ */
+void tipc_conn_terminate(struct tipc_server *s, int conid);
+
+int tipc_server_start(struct tipc_server *s);
+
+void tipc_server_stop(struct tipc_server *s);
+
+#endif
diff --git a/net/tipc/socket.c b/net/tipc/socket.c
index 09dc5b97e079..6cc7ddd2fb7c 100644
--- a/net/tipc/socket.c
+++ b/net/tipc/socket.c
@@ -1,8 +1,8 @@
/*
* net/tipc/socket.c: TIPC socket API
*
- * Copyright (c) 2001-2007, Ericsson AB
- * Copyright (c) 2004-2008, 2010-2011, Wind River Systems
+ * Copyright (c) 2001-2007, 2012 Ericsson AB
+ * Copyright (c) 2004-2008, 2010-2013, Wind River Systems
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
@@ -43,7 +43,6 @@
#define SS_LISTENING -1 /* socket is listening */
#define SS_READY -2 /* socket is connectionless */
-#define OVERLOAD_LIMIT_BASE 5000
#define CONN_TIMEOUT_DEFAULT 8000 /* default connect timeout = 8s */
struct tipc_sock {
@@ -62,17 +61,20 @@ struct tipc_sock {
static int backlog_rcv(struct sock *sk, struct sk_buff *skb);
static u32 dispatch(struct tipc_port *tport, struct sk_buff *buf);
static void wakeupdispatch(struct tipc_port *tport);
+static void tipc_data_ready(struct sock *sk, int len);
+static void tipc_write_space(struct sock *sk);
+static int release(struct socket *sock);
+static int accept(struct socket *sock, struct socket *new_sock, int flags);
static const struct proto_ops packet_ops;
static const struct proto_ops stream_ops;
static const struct proto_ops msg_ops;
static struct proto tipc_proto;
+static struct proto tipc_proto_kern;
static int sockets_enabled;
-static atomic_t tipc_queue_size = ATOMIC_INIT(0);
-
/*
* Revised TIPC socket locking policy:
*
@@ -126,22 +128,6 @@ static atomic_t tipc_queue_size = ATOMIC_INIT(0);
static void advance_rx_queue(struct sock *sk)
{
kfree_skb(__skb_dequeue(&sk->sk_receive_queue));
- atomic_dec(&tipc_queue_size);
-}
-
-/**
- * discard_rx_queue - discard all buffers in socket receive queue
- *
- * Caller must hold socket lock
- */
-static void discard_rx_queue(struct sock *sk)
-{
- struct sk_buff *buf;
-
- while ((buf = __skb_dequeue(&sk->sk_receive_queue))) {
- atomic_dec(&tipc_queue_size);
- kfree_skb(buf);
- }
}
/**
@@ -153,14 +139,12 @@ static void reject_rx_queue(struct sock *sk)
{
struct sk_buff *buf;
- while ((buf = __skb_dequeue(&sk->sk_receive_queue))) {
+ while ((buf = __skb_dequeue(&sk->sk_receive_queue)))
tipc_reject_msg(buf, TIPC_ERR_NO_PORT);
- atomic_dec(&tipc_queue_size);
- }
}
/**
- * tipc_create - create a TIPC socket
+ * tipc_sk_create - create a TIPC socket
* @net: network namespace (must be default network)
* @sock: pre-allocated socket structure
* @protocol: protocol indicator (must be 0)
@@ -171,8 +155,8 @@ static void reject_rx_queue(struct sock *sk)
*
* Returns 0 on success, errno otherwise
*/
-static int tipc_create(struct net *net, struct socket *sock, int protocol,
- int kern)
+static int tipc_sk_create(struct net *net, struct socket *sock, int protocol,
+ int kern)
{
const struct proto_ops *ops;
socket_state state;
@@ -202,13 +186,17 @@ static int tipc_create(struct net *net, struct socket *sock, int protocol,
}
/* Allocate socket's protocol area */
- sk = sk_alloc(net, AF_TIPC, GFP_KERNEL, &tipc_proto);
+ if (!kern)
+ sk = sk_alloc(net, AF_TIPC, GFP_KERNEL, &tipc_proto);
+ else
+ sk = sk_alloc(net, AF_TIPC, GFP_KERNEL, &tipc_proto_kern);
+
if (sk == NULL)
return -ENOMEM;
/* Allocate TIPC port for socket to use */
- tp_ptr = tipc_createport_raw(sk, &dispatch, &wakeupdispatch,
- TIPC_LOW_IMPORTANCE);
+ tp_ptr = tipc_createport(sk, &dispatch, &wakeupdispatch,
+ TIPC_LOW_IMPORTANCE);
if (unlikely(!tp_ptr)) {
sk_free(sk);
return -ENOMEM;
@@ -220,6 +208,9 @@ static int tipc_create(struct net *net, struct socket *sock, int protocol,
sock_init_data(sock, sk);
sk->sk_backlog_rcv = backlog_rcv;
+ sk->sk_rcvbuf = sysctl_tipc_rmem[1];
+ sk->sk_data_ready = tipc_data_ready;
+ sk->sk_write_space = tipc_write_space;
tipc_sk(sk)->p = tp_ptr;
tipc_sk(sk)->conn_timeout = CONN_TIMEOUT_DEFAULT;
@@ -235,6 +226,78 @@ static int tipc_create(struct net *net, struct socket *sock, int protocol,
}
/**
+ * tipc_sock_create_local - create TIPC socket from inside TIPC module
+ * @type: socket type - SOCK_RDM or SOCK_SEQPACKET
+ *
+ * We cannot use sock_creat_kern here because it bumps module user count.
+ * Since socket owner and creator is the same module we must make sure
+ * that module count remains zero for module local sockets, otherwise
+ * we cannot do rmmod.
+ *
+ * Returns 0 on success, errno otherwise
+ */
+int tipc_sock_create_local(int type, struct socket **res)
+{
+ int rc;
+ struct sock *sk;
+
+ rc = sock_create_lite(AF_TIPC, type, 0, res);
+ if (rc < 0) {
+ pr_err("Failed to create kernel socket\n");
+ return rc;
+ }
+ tipc_sk_create(&init_net, *res, 0, 1);
+
+ sk = (*res)->sk;
+
+ return 0;
+}
+
+/**
+ * tipc_sock_release_local - release socket created by tipc_sock_create_local
+ * @sock: the socket to be released.
+ *
+ * Module reference count is not incremented when such sockets are created,
+ * so we must keep it from being decremented when they are released.
+ */
+void tipc_sock_release_local(struct socket *sock)
+{
+ release(sock);
+ sock->ops = NULL;
+ sock_release(sock);
+}
+
+/**
+ * tipc_sock_accept_local - accept a connection on a socket created
+ * with tipc_sock_create_local. Use this function to avoid that
+ * module reference count is inadvertently incremented.
+ *
+ * @sock: the accepting socket
+ * @newsock: reference to the new socket to be created
+ * @flags: socket flags
+ */
+
+int tipc_sock_accept_local(struct socket *sock, struct socket **newsock,
+ int flags)
+{
+ struct sock *sk = sock->sk;
+ int ret;
+
+ ret = sock_create_lite(sk->sk_family, sk->sk_type,
+ sk->sk_protocol, newsock);
+ if (ret < 0)
+ return ret;
+
+ ret = accept(sock, *newsock, flags);
+ if (ret < 0) {
+ sock_release(*newsock);
+ return ret;
+ }
+ (*newsock)->ops = sock->ops;
+ return ret;
+}
+
+/**
* release - destroy a TIPC socket
* @sock: socket to destroy
*
@@ -275,7 +338,6 @@ static int release(struct socket *sock)
buf = __skb_dequeue(&sk->sk_receive_queue);
if (buf == NULL)
break;
- atomic_dec(&tipc_queue_size);
if (TIPC_SKB_CB(buf)->handle != 0)
kfree_skb(buf);
else {
@@ -295,7 +357,7 @@ static int release(struct socket *sock)
res = tipc_deleteport(tport->ref);
/* Discard any remaining (connection-based) messages in receive queue */
- discard_rx_queue(sk);
+ __skb_queue_purge(&sk->sk_receive_queue);
/* Reject any messages that accumulated in backlog queue */
sock->state = SS_DISCONNECTING;
@@ -340,7 +402,9 @@ static int bind(struct socket *sock, struct sockaddr *uaddr, int uaddr_len)
else if (addr->addrtype != TIPC_ADDR_NAMESEQ)
return -EAFNOSUPPORT;
- if (addr->addr.nameseq.type < TIPC_RESERVED_TYPES)
+ if ((addr->addr.nameseq.type < TIPC_RESERVED_TYPES) &&
+ (addr->addr.nameseq.type != TIPC_TOP_SRV) &&
+ (addr->addr.nameseq.type != TIPC_CFG_SRV))
return -EACCES;
return (addr->scope > 0) ?
@@ -407,7 +471,7 @@ static int get_name(struct socket *sock, struct sockaddr *uaddr,
* socket state flags set
* ------------ ---------
* unconnected no read flags
- * no write flags
+ * POLLOUT if port is not congested
*
* connecting POLLIN/POLLRDNORM if ACK/NACK in rx queue
* no write flags
@@ -434,9 +498,13 @@ static unsigned int poll(struct file *file, struct socket *sock,
struct sock *sk = sock->sk;
u32 mask = 0;
- poll_wait(file, sk_sleep(sk), wait);
+ sock_poll_wait(file, sk_sleep(sk), wait);
switch ((int)sock->state) {
+ case SS_UNCONNECTED:
+ if (!tipc_sk_port(sk)->congested)
+ mask |= POLLOUT;
+ break;
case SS_READY:
case SS_CONNECTED:
if (!tipc_sk_port(sk)->congested)
@@ -515,8 +583,7 @@ static int send_msg(struct kiocb *iocb, struct socket *sock,
if (unlikely((m->msg_namelen < sizeof(*dest)) ||
(dest->family != AF_TIPC)))
return -EINVAL;
- if ((total_len > TIPC_MAX_USER_MSG_SIZE) ||
- (m->msg_iovlen > (unsigned int)INT_MAX))
+ if (total_len > TIPC_MAX_USER_MSG_SIZE)
return -EMSGSIZE;
if (iocb)
@@ -532,8 +599,7 @@ static int send_msg(struct kiocb *iocb, struct socket *sock,
res = -EISCONN;
goto exit;
}
- if ((tport->published) ||
- ((sock->type == SOCK_STREAM) && (total_len != 0))) {
+ if (tport->published) {
res = -EOPNOTSUPP;
goto exit;
}
@@ -624,8 +690,7 @@ static int send_packet(struct kiocb *iocb, struct socket *sock,
if (unlikely(dest))
return send_msg(iocb, sock, m, total_len);
- if ((total_len > TIPC_MAX_USER_MSG_SIZE) ||
- (m->msg_iovlen > (unsigned int)INT_MAX))
+ if (total_len > TIPC_MAX_USER_MSG_SIZE)
return -EMSGSIZE;
if (iocb)
@@ -710,8 +775,7 @@ static int send_stream(struct kiocb *iocb, struct socket *sock,
goto exit;
}
- if ((total_len > (unsigned int)INT_MAX) ||
- (m->msg_iovlen > (unsigned int)INT_MAX)) {
+ if (total_len > (unsigned int)INT_MAX) {
res = -EMSGSIZE;
goto exit;
}
@@ -774,16 +838,19 @@ exit:
static int auto_connect(struct socket *sock, struct tipc_msg *msg)
{
struct tipc_sock *tsock = tipc_sk(sock->sk);
-
- if (msg_errcode(msg)) {
- sock->state = SS_DISCONNECTING;
- return -ECONNREFUSED;
- }
+ struct tipc_port *p_ptr;
tsock->peer_name.ref = msg_origport(msg);
tsock->peer_name.node = msg_orignode(msg);
- tipc_connect2port(tsock->p->ref, &tsock->peer_name);
- tipc_set_portimportance(tsock->p->ref, msg_importance(msg));
+ p_ptr = tipc_port_deref(tsock->p->ref);
+ if (!p_ptr)
+ return -EINVAL;
+
+ __tipc_connect(tsock->p->ref, p_ptr, &tsock->peer_name);
+
+ if (msg_importance(msg) > TIPC_CRITICAL_IMPORTANCE)
+ return -EINVAL;
+ msg_set_importance(&p_ptr->phdr, (u32)msg_importance(msg));
sock->state = SS_CONNECTED;
return 0;
}
@@ -802,6 +869,7 @@ static void set_orig_addr(struct msghdr *m, struct tipc_msg *msg)
if (addr) {
addr->family = AF_TIPC;
addr->addrtype = TIPC_ADDR_ID;
+ memset(&addr->addr, 0, sizeof(addr->addr));
addr->addr.id.ref = msg_origport(msg);
addr->addr.id.node = msg_orignode(msg);
addr->addr.name.domain = 0; /* could leave uninitialized */
@@ -821,7 +889,7 @@ static void set_orig_addr(struct msghdr *m, struct tipc_msg *msg)
* Returns 0 if successful, otherwise errno
*/
static int anc_data_recv(struct msghdr *m, struct tipc_msg *msg,
- struct tipc_port *tport)
+ struct tipc_port *tport)
{
u32 anc_data[3];
u32 err;
@@ -916,6 +984,9 @@ static int recv_msg(struct kiocb *iocb, struct socket *sock,
goto exit;
}
+ /* will be updated in set_orig_addr() if needed */
+ m->msg_namelen = 0;
+
timeout = sock_rcvtimeo(sk, flags & MSG_DONTWAIT);
restart:
@@ -942,13 +1013,6 @@ restart:
sz = msg_data_sz(msg);
err = msg_errcode(msg);
- /* Complete connection setup for an implied connect */
- if (unlikely(sock->state == SS_CONNECTING)) {
- res = auto_connect(sock, msg);
- if (res)
- goto exit;
- }
-
/* Discard an empty non-errored message & try again */
if ((!sz) && (!err)) {
advance_rx_queue(sk);
@@ -1026,12 +1090,14 @@ static int recv_stream(struct kiocb *iocb, struct socket *sock,
lock_sock(sk);
- if (unlikely((sock->state == SS_UNCONNECTED) ||
- (sock->state == SS_CONNECTING))) {
+ if (unlikely((sock->state == SS_UNCONNECTED))) {
res = -ENOTCONN;
goto exit;
}
+ /* will be updated in set_orig_addr() if needed */
+ m->msg_namelen = 0;
+
target = sock_rcvlowat(sk, flags & MSG_WAITALL, buf_len);
timeout = sock_rcvtimeo(sk, flags & MSG_DONTWAIT);
@@ -1125,31 +1191,144 @@ exit:
}
/**
- * rx_queue_full - determine if receive queue can accept another message
- * @msg: message to be added to queue
- * @queue_size: current size of queue
- * @base: nominal maximum size of queue
+ * tipc_write_space - wake up thread if port congestion is released
+ * @sk: socket
+ */
+static void tipc_write_space(struct sock *sk)
+{
+ struct socket_wq *wq;
+
+ rcu_read_lock();
+ wq = rcu_dereference(sk->sk_wq);
+ if (wq_has_sleeper(wq))
+ wake_up_interruptible_sync_poll(&wq->wait, POLLOUT |
+ POLLWRNORM | POLLWRBAND);
+ rcu_read_unlock();
+}
+
+/**
+ * tipc_data_ready - wake up threads to indicate messages have been received
+ * @sk: socket
+ * @len: the length of messages
+ */
+static void tipc_data_ready(struct sock *sk, int len)
+{
+ struct socket_wq *wq;
+
+ rcu_read_lock();
+ wq = rcu_dereference(sk->sk_wq);
+ if (wq_has_sleeper(wq))
+ wake_up_interruptible_sync_poll(&wq->wait, POLLIN |
+ POLLRDNORM | POLLRDBAND);
+ rcu_read_unlock();
+}
+
+/**
+ * filter_connect - Handle all incoming messages for a connection-based socket
+ * @tsock: TIPC socket
+ * @msg: message
*
- * Returns 1 if queue is unable to accept message, 0 otherwise
+ * Returns TIPC error status code and socket error status code
+ * once it encounters some errors
*/
-static int rx_queue_full(struct tipc_msg *msg, u32 queue_size, u32 base)
+static u32 filter_connect(struct tipc_sock *tsock, struct sk_buff **buf)
{
- u32 threshold;
- u32 imp = msg_importance(msg);
-
- if (imp == TIPC_LOW_IMPORTANCE)
- threshold = base;
- else if (imp == TIPC_MEDIUM_IMPORTANCE)
- threshold = base * 2;
- else if (imp == TIPC_HIGH_IMPORTANCE)
- threshold = base * 100;
- else
- return 0;
+ struct socket *sock = tsock->sk.sk_socket;
+ struct tipc_msg *msg = buf_msg(*buf);
+ struct sock *sk = &tsock->sk;
+ u32 retval = TIPC_ERR_NO_PORT;
+ int res;
- if (msg_connected(msg))
- threshold *= 4;
+ if (msg_mcast(msg))
+ return retval;
+
+ switch ((int)sock->state) {
+ case SS_CONNECTED:
+ /* Accept only connection-based messages sent by peer */
+ if (msg_connected(msg) && tipc_port_peer_msg(tsock->p, msg)) {
+ if (unlikely(msg_errcode(msg))) {
+ sock->state = SS_DISCONNECTING;
+ __tipc_disconnect(tsock->p);
+ }
+ retval = TIPC_OK;
+ }
+ break;
+ case SS_CONNECTING:
+ /* Accept only ACK or NACK message */
+ if (unlikely(msg_errcode(msg))) {
+ sock->state = SS_DISCONNECTING;
+ sk->sk_err = ECONNREFUSED;
+ retval = TIPC_OK;
+ break;
+ }
- return queue_size >= threshold;
+ if (unlikely(!msg_connected(msg)))
+ break;
+
+ res = auto_connect(sock, msg);
+ if (res) {
+ sock->state = SS_DISCONNECTING;
+ sk->sk_err = -res;
+ retval = TIPC_OK;
+ break;
+ }
+
+ /* If an incoming message is an 'ACK-', it should be
+ * discarded here because it doesn't contain useful
+ * data. In addition, we should try to wake up
+ * connect() routine if sleeping.
+ */
+ if (msg_data_sz(msg) == 0) {
+ kfree_skb(*buf);
+ *buf = NULL;
+ if (waitqueue_active(sk_sleep(sk)))
+ wake_up_interruptible(sk_sleep(sk));
+ }
+ retval = TIPC_OK;
+ break;
+ case SS_LISTENING:
+ case SS_UNCONNECTED:
+ /* Accept only SYN message */
+ if (!msg_connected(msg) && !(msg_errcode(msg)))
+ retval = TIPC_OK;
+ break;
+ case SS_DISCONNECTING:
+ break;
+ default:
+ pr_err("Unknown socket state %u\n", sock->state);
+ }
+ return retval;
+}
+
+/**
+ * rcvbuf_limit - get proper overload limit of socket receive queue
+ * @sk: socket
+ * @buf: message
+ *
+ * For all connection oriented messages, irrespective of importance,
+ * the default overload value (i.e. 67MB) is set as limit.
+ *
+ * For all connectionless messages, by default new queue limits are
+ * as belows:
+ *
+ * TIPC_LOW_IMPORTANCE (4 MB)
+ * TIPC_MEDIUM_IMPORTANCE (8 MB)
+ * TIPC_HIGH_IMPORTANCE (16 MB)
+ * TIPC_CRITICAL_IMPORTANCE (32 MB)
+ *
+ * Returns overload limit according to corresponding message importance
+ */
+static unsigned int rcvbuf_limit(struct sock *sk, struct sk_buff *buf)
+{
+ struct tipc_msg *msg = buf_msg(buf);
+ unsigned int limit;
+
+ if (msg_connected(msg))
+ limit = sysctl_tipc_rmem[2];
+ else
+ limit = sk->sk_rcvbuf >> TIPC_CRITICAL_IMPORTANCE <<
+ msg_importance(msg);
+ return limit;
}
/**
@@ -1168,7 +1347,8 @@ static u32 filter_rcv(struct sock *sk, struct sk_buff *buf)
{
struct socket *sock = sk->sk_socket;
struct tipc_msg *msg = buf_msg(buf);
- u32 recv_q_len;
+ unsigned int limit = rcvbuf_limit(sk, buf);
+ u32 res = TIPC_OK;
/* Reject message if it is wrong sort of message for socket */
if (msg_type(msg) > TIPC_DIRECT_MSG)
@@ -1178,51 +1358,21 @@ static u32 filter_rcv(struct sock *sk, struct sk_buff *buf)
if (msg_connected(msg))
return TIPC_ERR_NO_PORT;
} else {
- if (msg_mcast(msg))
- return TIPC_ERR_NO_PORT;
- if (sock->state == SS_CONNECTED) {
- if (!msg_connected(msg) ||
- !tipc_port_peer_msg(tipc_sk_port(sk), msg))
- return TIPC_ERR_NO_PORT;
- } else if (sock->state == SS_CONNECTING) {
- if (!msg_connected(msg) && (msg_errcode(msg) == 0))
- return TIPC_ERR_NO_PORT;
- } else if (sock->state == SS_LISTENING) {
- if (msg_connected(msg) || msg_errcode(msg))
- return TIPC_ERR_NO_PORT;
- } else if (sock->state == SS_DISCONNECTING) {
- return TIPC_ERR_NO_PORT;
- } else /* (sock->state == SS_UNCONNECTED) */ {
- if (msg_connected(msg) || msg_errcode(msg))
- return TIPC_ERR_NO_PORT;
- }
+ res = filter_connect(tipc_sk(sk), &buf);
+ if (res != TIPC_OK || buf == NULL)
+ return res;
}
/* Reject message if there isn't room to queue it */
- recv_q_len = (u32)atomic_read(&tipc_queue_size);
- if (unlikely(recv_q_len >= OVERLOAD_LIMIT_BASE)) {
- if (rx_queue_full(msg, recv_q_len, OVERLOAD_LIMIT_BASE))
- return TIPC_ERR_OVERLOAD;
- }
- recv_q_len = skb_queue_len(&sk->sk_receive_queue);
- if (unlikely(recv_q_len >= (OVERLOAD_LIMIT_BASE / 2))) {
- if (rx_queue_full(msg, recv_q_len, OVERLOAD_LIMIT_BASE / 2))
- return TIPC_ERR_OVERLOAD;
- }
+ if (sk_rmem_alloc_get(sk) + buf->truesize >= limit)
+ return TIPC_ERR_OVERLOAD;
- /* Enqueue message (finally!) */
+ /* Enqueue message */
TIPC_SKB_CB(buf)->handle = 0;
- atomic_inc(&tipc_queue_size);
__skb_queue_tail(&sk->sk_receive_queue, buf);
+ skb_set_owner_r(buf, sk);
- /* Initiate connection termination for an incoming 'FIN' */
- if (unlikely(msg_errcode(msg) && (sock->state == SS_CONNECTED))) {
- sock->state = SS_DISCONNECTING;
- tipc_disconnect_port(tipc_sk_port(sk));
- }
-
- if (waitqueue_active(sk_sleep(sk)))
- wake_up_interruptible(sk_sleep(sk));
+ sk->sk_data_ready(sk, 0);
return TIPC_OK;
}
@@ -1256,7 +1406,7 @@ static int backlog_rcv(struct sock *sk, struct sk_buff *buf)
*/
static u32 dispatch(struct tipc_port *tport, struct sk_buff *buf)
{
- struct sock *sk = (struct sock *)tport->usr_handle;
+ struct sock *sk = tport->sk;
u32 res;
/*
@@ -1269,7 +1419,7 @@ static u32 dispatch(struct tipc_port *tport, struct sk_buff *buf)
if (!sock_owned_by_user(sk)) {
res = filter_rcv(sk, buf);
} else {
- if (sk_add_backlog(sk, buf, sk->sk_rcvbuf))
+ if (sk_add_backlog(sk, buf, rcvbuf_limit(sk, buf)))
res = TIPC_ERR_OVERLOAD;
else
res = TIPC_OK;
@@ -1287,10 +1437,9 @@ static u32 dispatch(struct tipc_port *tport, struct sk_buff *buf)
*/
static void wakeupdispatch(struct tipc_port *tport)
{
- struct sock *sk = (struct sock *)tport->usr_handle;
+ struct sock *sk = tport->sk;
- if (waitqueue_active(sk_sleep(sk)))
- wake_up_interruptible(sk_sleep(sk));
+ sk->sk_write_space(sk);
}
/**
@@ -1308,8 +1457,6 @@ static int connect(struct socket *sock, struct sockaddr *dest, int destlen,
struct sock *sk = sock->sk;
struct sockaddr_tipc *dst = (struct sockaddr_tipc *)dest;
struct msghdr m = {NULL,};
- struct sk_buff *buf;
- struct tipc_msg *msg;
unsigned int timeout;
int res;
@@ -1321,26 +1468,6 @@ static int connect(struct socket *sock, struct sockaddr *dest, int destlen,
goto exit;
}
- /* For now, TIPC does not support the non-blocking form of connect() */
- if (flags & O_NONBLOCK) {
- res = -EOPNOTSUPP;
- goto exit;
- }
-
- /* Issue Posix-compliant error code if socket is in the wrong state */
- if (sock->state == SS_LISTENING) {
- res = -EOPNOTSUPP;
- goto exit;
- }
- if (sock->state == SS_CONNECTING) {
- res = -EALREADY;
- goto exit;
- }
- if (sock->state != SS_UNCONNECTED) {
- res = -EISCONN;
- goto exit;
- }
-
/*
* Reject connection attempt using multicast address
*
@@ -1352,49 +1479,66 @@ static int connect(struct socket *sock, struct sockaddr *dest, int destlen,
goto exit;
}
- /* Reject any messages already in receive queue (very unlikely) */
- reject_rx_queue(sk);
+ timeout = (flags & O_NONBLOCK) ? 0 : tipc_sk(sk)->conn_timeout;
+
+ switch (sock->state) {
+ case SS_UNCONNECTED:
+ /* Send a 'SYN-' to destination */
+ m.msg_name = dest;
+ m.msg_namelen = destlen;
+
+ /* If connect is in non-blocking case, set MSG_DONTWAIT to
+ * indicate send_msg() is never blocked.
+ */
+ if (!timeout)
+ m.msg_flags = MSG_DONTWAIT;
+
+ res = send_msg(NULL, sock, &m, 0);
+ if ((res < 0) && (res != -EWOULDBLOCK))
+ goto exit;
- /* Send a 'SYN-' to destination */
- m.msg_name = dest;
- m.msg_namelen = destlen;
- res = send_msg(NULL, sock, &m, 0);
- if (res < 0)
+ /* Just entered SS_CONNECTING state; the only
+ * difference is that return value in non-blocking
+ * case is EINPROGRESS, rather than EALREADY.
+ */
+ res = -EINPROGRESS;
+ break;
+ case SS_CONNECTING:
+ res = -EALREADY;
+ break;
+ case SS_CONNECTED:
+ res = -EISCONN;
+ break;
+ default:
+ res = -EINVAL;
goto exit;
+ }
- /* Wait until an 'ACK' or 'RST' arrives, or a timeout occurs */
- timeout = tipc_sk(sk)->conn_timeout;
- release_sock(sk);
- res = wait_event_interruptible_timeout(*sk_sleep(sk),
- (!skb_queue_empty(&sk->sk_receive_queue) ||
- (sock->state != SS_CONNECTING)),
- timeout ? (long)msecs_to_jiffies(timeout)
- : MAX_SCHEDULE_TIMEOUT);
- lock_sock(sk);
+ if (sock->state == SS_CONNECTING) {
+ if (!timeout)
+ goto exit;
- if (res > 0) {
- buf = skb_peek(&sk->sk_receive_queue);
- if (buf != NULL) {
- msg = buf_msg(buf);
- res = auto_connect(sock, msg);
- if (!res) {
- if (!msg_data_sz(msg))
- advance_rx_queue(sk);
- }
- } else {
- if (sock->state == SS_CONNECTED)
- res = -EISCONN;
+ /* Wait until an 'ACK' or 'RST' arrives, or a timeout occurs */
+ release_sock(sk);
+ res = wait_event_interruptible_timeout(*sk_sleep(sk),
+ sock->state != SS_CONNECTING,
+ timeout ? (long)msecs_to_jiffies(timeout)
+ : MAX_SCHEDULE_TIMEOUT);
+ lock_sock(sk);
+ if (res <= 0) {
+ if (res == 0)
+ res = -ETIMEDOUT;
else
- res = -ECONNREFUSED;
+ ; /* leave "res" unchanged */
+ goto exit;
}
- } else {
- if (res == 0)
- res = -ETIMEDOUT;
- else
- ; /* leave "res" unchanged */
- sock->state = SS_DISCONNECTING;
}
+ if (unlikely(sock->state == SS_DISCONNECTING))
+ res = sock_error(sk);
+ else
+ res = 0;
+
exit:
release_sock(sk);
return res;
@@ -1435,8 +1579,13 @@ static int listen(struct socket *sock, int len)
*/
static int accept(struct socket *sock, struct socket *new_sock, int flags)
{
- struct sock *sk = sock->sk;
+ struct sock *new_sk, *sk = sock->sk;
struct sk_buff *buf;
+ struct tipc_sock *new_tsock;
+ struct tipc_port *new_tport;
+ struct tipc_msg *msg;
+ u32 new_ref;
+
int res;
lock_sock(sk);
@@ -1461,49 +1610,53 @@ static int accept(struct socket *sock, struct socket *new_sock, int flags)
buf = skb_peek(&sk->sk_receive_queue);
- res = tipc_create(sock_net(sock->sk), new_sock, 0, 0);
- if (!res) {
- struct sock *new_sk = new_sock->sk;
- struct tipc_sock *new_tsock = tipc_sk(new_sk);
- struct tipc_port *new_tport = new_tsock->p;
- u32 new_ref = new_tport->ref;
- struct tipc_msg *msg = buf_msg(buf);
+ res = tipc_sk_create(sock_net(sock->sk), new_sock, 0, 1);
+ if (res)
+ goto exit;
- lock_sock(new_sk);
+ new_sk = new_sock->sk;
+ new_tsock = tipc_sk(new_sk);
+ new_tport = new_tsock->p;
+ new_ref = new_tport->ref;
+ msg = buf_msg(buf);
- /*
- * Reject any stray messages received by new socket
- * before the socket lock was taken (very, very unlikely)
- */
- reject_rx_queue(new_sk);
-
- /* Connect new socket to it's peer */
- new_tsock->peer_name.ref = msg_origport(msg);
- new_tsock->peer_name.node = msg_orignode(msg);
- tipc_connect2port(new_ref, &new_tsock->peer_name);
- new_sock->state = SS_CONNECTED;
-
- tipc_set_portimportance(new_ref, msg_importance(msg));
- if (msg_named(msg)) {
- new_tport->conn_type = msg_nametype(msg);
- new_tport->conn_instance = msg_nameinst(msg);
- }
+ /* we lock on new_sk; but lockdep sees the lock on sk */
+ lock_sock_nested(new_sk, SINGLE_DEPTH_NESTING);
- /*
- * Respond to 'SYN-' by discarding it & returning 'ACK'-.
- * Respond to 'SYN+' by queuing it on new socket.
- */
- if (!msg_data_sz(msg)) {
- struct msghdr m = {NULL,};
+ /*
+ * Reject any stray messages received by new socket
+ * before the socket lock was taken (very, very unlikely)
+ */
+ reject_rx_queue(new_sk);
+
+ /* Connect new socket to it's peer */
+ new_tsock->peer_name.ref = msg_origport(msg);
+ new_tsock->peer_name.node = msg_orignode(msg);
+ tipc_connect(new_ref, &new_tsock->peer_name);
+ new_sock->state = SS_CONNECTED;
+
+ tipc_set_portimportance(new_ref, msg_importance(msg));
+ if (msg_named(msg)) {
+ new_tport->conn_type = msg_nametype(msg);
+ new_tport->conn_instance = msg_nameinst(msg);
+ }
- advance_rx_queue(sk);
- send_packet(NULL, new_sock, &m, 0);
- } else {
- __skb_dequeue(&sk->sk_receive_queue);
- __skb_queue_head(&new_sk->sk_receive_queue, buf);
- }
- release_sock(new_sk);
+ /*
+ * Respond to 'SYN-' by discarding it & returning 'ACK'-.
+ * Respond to 'SYN+' by queuing it on new socket.
+ */
+ if (!msg_data_sz(msg)) {
+ struct msghdr m = {NULL,};
+
+ advance_rx_queue(sk);
+ send_packet(NULL, new_sock, &m, 0);
+ } else {
+ __skb_dequeue(&sk->sk_receive_queue);
+ __skb_queue_head(&new_sk->sk_receive_queue, buf);
+ skb_set_owner_r(buf, new_sk);
}
+ release_sock(new_sk);
+
exit:
release_sock(sk);
return res;
@@ -1538,7 +1691,6 @@ restart:
/* Disconnect and send a 'FIN+' or 'FIN-' message to peer */
buf = __skb_dequeue(&sk->sk_receive_queue);
if (buf) {
- atomic_dec(&tipc_queue_size);
if (TIPC_SKB_CB(buf)->handle != 0) {
kfree_skb(buf);
goto restart;
@@ -1555,10 +1707,11 @@ restart:
case SS_DISCONNECTING:
- /* Discard any unreceived messages; wake up sleeping tasks */
- discard_rx_queue(sk);
- if (waitqueue_active(sk_sleep(sk)))
- wake_up_interruptible(sk_sleep(sk));
+ /* Discard any unreceived messages */
+ __skb_queue_purge(&sk->sk_receive_queue);
+
+ /* Wake up anyone sleeping in poll */
+ sk->sk_state_change(sk);
res = 0;
break;
@@ -1583,8 +1736,8 @@ restart:
*
* Returns 0 on success, errno otherwise
*/
-static int setsockopt(struct socket *sock,
- int lvl, int opt, char __user *ov, unsigned int ol)
+static int setsockopt(struct socket *sock, int lvl, int opt, char __user *ov,
+ unsigned int ol)
{
struct sock *sk = sock->sk;
struct tipc_port *tport = tipc_sk_port(sk);
@@ -1642,8 +1795,8 @@ static int setsockopt(struct socket *sock,
*
* Returns 0 on success, errno otherwise
*/
-static int getsockopt(struct socket *sock,
- int lvl, int opt, char __user *ov, int __user *ol)
+static int getsockopt(struct socket *sock, int lvl, int opt, char __user *ov,
+ int __user *ol)
{
struct sock *sk = sock->sk;
struct tipc_port *tport = tipc_sk_port(sk);
@@ -1676,7 +1829,7 @@ static int getsockopt(struct socket *sock,
/* no need to set "res", since already 0 at this point */
break;
case TIPC_NODE_RECVQ_DEPTH:
- value = (u32)atomic_read(&tipc_queue_size);
+ value = 0; /* was tipc_queue_size, now obsolete */
break;
case TIPC_SOCK_RECVQ_DEPTH:
value = skb_queue_len(&sk->sk_receive_queue);
@@ -1767,13 +1920,20 @@ static const struct proto_ops stream_ops = {
static const struct net_proto_family tipc_family_ops = {
.owner = THIS_MODULE,
.family = AF_TIPC,
- .create = tipc_create
+ .create = tipc_sk_create
};
static struct proto tipc_proto = {
.name = "TIPC",
.owner = THIS_MODULE,
- .obj_size = sizeof(struct tipc_sock)
+ .obj_size = sizeof(struct tipc_sock),
+ .sysctl_rmem = sysctl_tipc_rmem
+};
+
+static struct proto tipc_proto_kern = {
+ .name = "TIPC",
+ .obj_size = sizeof(struct tipc_sock),
+ .sysctl_rmem = sysctl_tipc_rmem
};
/**
diff --git a/net/tipc/subscr.c b/net/tipc/subscr.c
index 5ed5965eb0be..d38bb45d82e9 100644
--- a/net/tipc/subscr.c
+++ b/net/tipc/subscr.c
@@ -2,7 +2,7 @@
* net/tipc/subscr.c: TIPC network topology service
*
* Copyright (c) 2000-2006, Ericsson AB
- * Copyright (c) 2005-2007, 2010-2011, Wind River Systems
+ * Copyright (c) 2005-2007, 2010-2013, Wind River Systems
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
@@ -41,33 +41,42 @@
/**
* struct tipc_subscriber - TIPC network topology subscriber
- * @port_ref: object reference to server port connecting to subscriber
- * @lock: pointer to spinlock controlling access to subscriber's server port
- * @subscriber_list: adjacent subscribers in top. server's list of subscribers
+ * @conid: connection identifier to server connecting to subscriber
+ * @lock: controll access to subscriber
* @subscription_list: list of subscription objects for this subscriber
*/
struct tipc_subscriber {
- u32 port_ref;
- spinlock_t *lock;
- struct list_head subscriber_list;
+ int conid;
+ spinlock_t lock;
struct list_head subscription_list;
};
-/**
- * struct top_srv - TIPC network topology subscription service
- * @setup_port: reference to TIPC port that handles subscription requests
- * @subscription_count: number of active subscriptions (not subscribers!)
- * @subscriber_list: list of ports subscribing to service
- * @lock: spinlock govering access to subscriber list
- */
-struct top_srv {
- u32 setup_port;
- atomic_t subscription_count;
- struct list_head subscriber_list;
- spinlock_t lock;
+static void subscr_conn_msg_event(int conid, struct sockaddr_tipc *addr,
+ void *usr_data, void *buf, size_t len);
+static void *subscr_named_msg_event(int conid);
+static void subscr_conn_shutdown_event(int conid, void *usr_data);
+
+static atomic_t subscription_count = ATOMIC_INIT(0);
+
+static struct sockaddr_tipc topsrv_addr __read_mostly = {
+ .family = AF_TIPC,
+ .addrtype = TIPC_ADDR_NAMESEQ,
+ .addr.nameseq.type = TIPC_TOP_SRV,
+ .addr.nameseq.lower = TIPC_TOP_SRV,
+ .addr.nameseq.upper = TIPC_TOP_SRV,
+ .scope = TIPC_NODE_SCOPE
};
-static struct top_srv topsrv;
+static struct tipc_server topsrv __read_mostly = {
+ .saddr = &topsrv_addr,
+ .imp = TIPC_CRITICAL_IMPORTANCE,
+ .type = SOCK_SEQPACKET,
+ .max_rcvbuf_size = sizeof(struct tipc_subscr),
+ .name = "topology_server",
+ .tipc_conn_recvmsg = subscr_conn_msg_event,
+ .tipc_conn_new = subscr_named_msg_event,
+ .tipc_conn_shutdown = subscr_conn_shutdown_event,
+};
/**
* htohl - convert value to endianness used by destination
@@ -81,20 +90,13 @@ static u32 htohl(u32 in, int swap)
return swap ? swab32(in) : in;
}
-/**
- * subscr_send_event - send a message containing a tipc_event to the subscriber
- *
- * Note: Must not hold subscriber's server port lock, since tipc_send() will
- * try to take the lock if the message is rejected and returned!
- */
-static void subscr_send_event(struct tipc_subscription *sub,
- u32 found_lower,
- u32 found_upper,
- u32 event,
- u32 port_ref,
+static void subscr_send_event(struct tipc_subscription *sub, u32 found_lower,
+ u32 found_upper, u32 event, u32 port_ref,
u32 node)
{
- struct iovec msg_sect;
+ struct tipc_subscriber *subscriber = sub->subscriber;
+ struct kvec msg_sect;
+ int ret;
msg_sect.iov_base = (void *)&sub->evt;
msg_sect.iov_len = sizeof(struct tipc_event);
@@ -104,7 +106,10 @@ static void subscr_send_event(struct tipc_subscription *sub,
sub->evt.found_upper = htohl(found_upper, sub->swap);
sub->evt.port.ref = htohl(port_ref, sub->swap);
sub->evt.port.node = htohl(node, sub->swap);
- tipc_send(sub->server_ref, 1, &msg_sect, msg_sect.iov_len);
+ ret = tipc_conn_sendmsg(&topsrv, subscriber->conid, NULL,
+ msg_sect.iov_base, msg_sect.iov_len);
+ if (ret < 0)
+ pr_err("Sending subscription event failed, no memory\n");
}
/**
@@ -112,10 +117,8 @@ static void subscr_send_event(struct tipc_subscription *sub,
*
* Returns 1 if there is overlap, otherwise 0.
*/
-int tipc_subscr_overlap(struct tipc_subscription *sub,
- u32 found_lower,
+int tipc_subscr_overlap(struct tipc_subscription *sub, u32 found_lower,
u32 found_upper)
-
{
if (found_lower < sub->seq.lower)
found_lower = sub->seq.lower;
@@ -131,13 +134,9 @@ int tipc_subscr_overlap(struct tipc_subscription *sub,
*
* Protected by nameseq.lock in name_table.c
*/
-void tipc_subscr_report_overlap(struct tipc_subscription *sub,
- u32 found_lower,
- u32 found_upper,
- u32 event,
- u32 port_ref,
- u32 node,
- int must)
+void tipc_subscr_report_overlap(struct tipc_subscription *sub, u32 found_lower,
+ u32 found_upper, u32 event, u32 port_ref,
+ u32 node, int must)
{
if (!tipc_subscr_overlap(sub, found_lower, found_upper))
return;
@@ -147,21 +146,24 @@ void tipc_subscr_report_overlap(struct tipc_subscription *sub,
subscr_send_event(sub, found_lower, found_upper, event, port_ref, node);
}
-/**
- * subscr_timeout - subscription timeout has occurred
- */
static void subscr_timeout(struct tipc_subscription *sub)
{
- struct tipc_port *server_port;
+ struct tipc_subscriber *subscriber = sub->subscriber;
+
+ /* The spin lock per subscriber is used to protect its members */
+ spin_lock_bh(&subscriber->lock);
- /* Validate server port reference (in case subscriber is terminating) */
- server_port = tipc_port_lock(sub->server_ref);
- if (server_port == NULL)
+ /* Validate if the connection related to the subscriber is
+ * closed (in case subscriber is terminating)
+ */
+ if (subscriber->conid == 0) {
+ spin_unlock_bh(&subscriber->lock);
return;
+ }
/* Validate timeout (in case subscription is being cancelled) */
if (sub->timeout == TIPC_WAIT_FOREVER) {
- tipc_port_unlock(server_port);
+ spin_unlock_bh(&subscriber->lock);
return;
}
@@ -171,8 +173,7 @@ static void subscr_timeout(struct tipc_subscription *sub)
/* Unlink subscription from subscriber */
list_del(&sub->subscription_list);
- /* Release subscriber's server port */
- tipc_port_unlock(server_port);
+ spin_unlock_bh(&subscriber->lock);
/* Notify subscriber of timeout */
subscr_send_event(sub, sub->evt.s.seq.lower, sub->evt.s.seq.upper,
@@ -181,64 +182,54 @@ static void subscr_timeout(struct tipc_subscription *sub)
/* Now destroy subscription */
k_term_timer(&sub->timer);
kfree(sub);
- atomic_dec(&topsrv.subscription_count);
+ atomic_dec(&subscription_count);
}
/**
* subscr_del - delete a subscription within a subscription list
*
- * Called with subscriber port locked.
+ * Called with subscriber lock held.
*/
static void subscr_del(struct tipc_subscription *sub)
{
tipc_nametbl_unsubscribe(sub);
list_del(&sub->subscription_list);
kfree(sub);
- atomic_dec(&topsrv.subscription_count);
+ atomic_dec(&subscription_count);
}
/**
* subscr_terminate - terminate communication with a subscriber
*
- * Called with subscriber port locked. Routine must temporarily release lock
- * to enable subscription timeout routine(s) to finish without deadlocking;
- * the lock is then reclaimed to allow caller to release it upon return.
- * (This should work even in the unlikely event some other thread creates
- * a new object reference in the interim that uses this lock; this routine will
- * simply wait for it to be released, then claim it.)
+ * Note: Must call it in process context since it might sleep.
*/
static void subscr_terminate(struct tipc_subscriber *subscriber)
{
- u32 port_ref;
+ tipc_conn_terminate(&topsrv, subscriber->conid);
+}
+
+static void subscr_release(struct tipc_subscriber *subscriber)
+{
struct tipc_subscription *sub;
struct tipc_subscription *sub_temp;
- /* Invalidate subscriber reference */
- port_ref = subscriber->port_ref;
- subscriber->port_ref = 0;
- spin_unlock_bh(subscriber->lock);
+ spin_lock_bh(&subscriber->lock);
- /* Sever connection to subscriber */
- tipc_shutdown(port_ref);
- tipc_deleteport(port_ref);
+ /* Invalidate subscriber reference */
+ subscriber->conid = 0;
/* Destroy any existing subscriptions for subscriber */
list_for_each_entry_safe(sub, sub_temp, &subscriber->subscription_list,
subscription_list) {
if (sub->timeout != TIPC_WAIT_FOREVER) {
+ spin_unlock_bh(&subscriber->lock);
k_cancel_timer(&sub->timer);
k_term_timer(&sub->timer);
+ spin_lock_bh(&subscriber->lock);
}
subscr_del(sub);
}
-
- /* Remove subscriber from topology server's subscriber list */
- spin_lock_bh(&topsrv.lock);
- list_del(&subscriber->subscriber_list);
- spin_unlock_bh(&topsrv.lock);
-
- /* Reclaim subscriber lock */
- spin_lock_bh(subscriber->lock);
+ spin_unlock_bh(&subscriber->lock);
/* Now destroy subscriber */
kfree(subscriber);
@@ -247,7 +238,7 @@ static void subscr_terminate(struct tipc_subscriber *subscriber)
/**
* subscr_cancel - handle subscription cancellation request
*
- * Called with subscriber port locked. Routine must temporarily release lock
+ * Called with subscriber lock held. Routine must temporarily release lock
* to enable the subscription timeout routine to finish without deadlocking;
* the lock is then reclaimed to allow caller to release it upon return.
*
@@ -274,10 +265,10 @@ static void subscr_cancel(struct tipc_subscr *s,
/* Cancel subscription timer (if used), then delete subscription */
if (sub->timeout != TIPC_WAIT_FOREVER) {
sub->timeout = TIPC_WAIT_FOREVER;
- spin_unlock_bh(subscriber->lock);
+ spin_unlock_bh(&subscriber->lock);
k_cancel_timer(&sub->timer);
k_term_timer(&sub->timer);
- spin_lock_bh(subscriber->lock);
+ spin_lock_bh(&subscriber->lock);
}
subscr_del(sub);
}
@@ -285,7 +276,7 @@ static void subscr_cancel(struct tipc_subscr *s,
/**
* subscr_subscribe - create subscription for subscriber
*
- * Called with subscriber port locked.
+ * Called with subscriber lock held.
*/
static struct tipc_subscription *subscr_subscribe(struct tipc_subscr *s,
struct tipc_subscriber *subscriber)
@@ -304,9 +295,9 @@ static struct tipc_subscription *subscr_subscribe(struct tipc_subscr *s,
}
/* Refuse subscription if global limit exceeded */
- if (atomic_read(&topsrv.subscription_count) >= tipc_max_subscriptions) {
+ if (atomic_read(&subscription_count) >= TIPC_MAX_SUBSCRIPTIONS) {
pr_warn("Subscription rejected, limit reached (%u)\n",
- tipc_max_subscriptions);
+ TIPC_MAX_SUBSCRIPTIONS);
subscr_terminate(subscriber);
return NULL;
}
@@ -335,10 +326,10 @@ static struct tipc_subscription *subscr_subscribe(struct tipc_subscr *s,
}
INIT_LIST_HEAD(&sub->nameseq_list);
list_add(&sub->subscription_list, &subscriber->subscription_list);
- sub->server_ref = subscriber->port_ref;
+ sub->subscriber = subscriber;
sub->swap = swap;
memcpy(&sub->evt.s, s, sizeof(struct tipc_subscr));
- atomic_inc(&topsrv.subscription_count);
+ atomic_inc(&subscription_count);
if (sub->timeout != TIPC_WAIT_FOREVER) {
k_init_timer(&sub->timer,
(Handler)subscr_timeout, (unsigned long)sub);
@@ -348,196 +339,51 @@ static struct tipc_subscription *subscr_subscribe(struct tipc_subscr *s,
return sub;
}
-/**
- * subscr_conn_shutdown_event - handle termination request from subscriber
- *
- * Called with subscriber's server port unlocked.
- */
-static void subscr_conn_shutdown_event(void *usr_handle,
- u32 port_ref,
- struct sk_buff **buf,
- unsigned char const *data,
- unsigned int size,
- int reason)
+/* Handle one termination request for the subscriber */
+static void subscr_conn_shutdown_event(int conid, void *usr_data)
{
- struct tipc_subscriber *subscriber = usr_handle;
- spinlock_t *subscriber_lock;
-
- if (tipc_port_lock(port_ref) == NULL)
- return;
-
- subscriber_lock = subscriber->lock;
- subscr_terminate(subscriber);
- spin_unlock_bh(subscriber_lock);
+ subscr_release((struct tipc_subscriber *)usr_data);
}
-/**
- * subscr_conn_msg_event - handle new subscription request from subscriber
- *
- * Called with subscriber's server port unlocked.
- */
-static void subscr_conn_msg_event(void *usr_handle,
- u32 port_ref,
- struct sk_buff **buf,
- const unchar *data,
- u32 size)
+/* Handle one request to create a new subscription for the subscriber */
+static void subscr_conn_msg_event(int conid, struct sockaddr_tipc *addr,
+ void *usr_data, void *buf, size_t len)
{
- struct tipc_subscriber *subscriber = usr_handle;
- spinlock_t *subscriber_lock;
+ struct tipc_subscriber *subscriber = usr_data;
struct tipc_subscription *sub;
- /*
- * Lock subscriber's server port (& make a local copy of lock pointer,
- * in case subscriber is deleted while processing subscription request)
- */
- if (tipc_port_lock(port_ref) == NULL)
- return;
-
- subscriber_lock = subscriber->lock;
-
- if (size != sizeof(struct tipc_subscr)) {
- subscr_terminate(subscriber);
- spin_unlock_bh(subscriber_lock);
- } else {
- sub = subscr_subscribe((struct tipc_subscr *)data, subscriber);
- spin_unlock_bh(subscriber_lock);
- if (sub != NULL) {
-
- /*
- * We must release the server port lock before adding a
- * subscription to the name table since TIPC needs to be
- * able to (re)acquire the port lock if an event message
- * issued by the subscription process is rejected and
- * returned. The subscription cannot be deleted while
- * it is being added to the name table because:
- * a) the single-threading of the native API port code
- * ensures the subscription cannot be cancelled and
- * the subscriber connection cannot be broken, and
- * b) the name table lock ensures the subscription
- * timeout code cannot delete the subscription,
- * so the subscription object is still protected.
- */
- tipc_nametbl_subscribe(sub);
- }
- }
+ spin_lock_bh(&subscriber->lock);
+ sub = subscr_subscribe((struct tipc_subscr *)buf, subscriber);
+ if (sub)
+ tipc_nametbl_subscribe(sub);
+ spin_unlock_bh(&subscriber->lock);
}
-/**
- * subscr_named_msg_event - handle request to establish a new subscriber
- */
-static void subscr_named_msg_event(void *usr_handle,
- u32 port_ref,
- struct sk_buff **buf,
- const unchar *data,
- u32 size,
- u32 importance,
- struct tipc_portid const *orig,
- struct tipc_name_seq const *dest)
+
+/* Handle one request to establish a new subscriber */
+static void *subscr_named_msg_event(int conid)
{
struct tipc_subscriber *subscriber;
- u32 server_port_ref;
/* Create subscriber object */
subscriber = kzalloc(sizeof(struct tipc_subscriber), GFP_ATOMIC);
if (subscriber == NULL) {
pr_warn("Subscriber rejected, no memory\n");
- return;
+ return NULL;
}
INIT_LIST_HEAD(&subscriber->subscription_list);
- INIT_LIST_HEAD(&subscriber->subscriber_list);
-
- /* Create server port & establish connection to subscriber */
- tipc_createport(subscriber,
- importance,
- NULL,
- NULL,
- subscr_conn_shutdown_event,
- NULL,
- NULL,
- subscr_conn_msg_event,
- NULL,
- &subscriber->port_ref);
- if (subscriber->port_ref == 0) {
- pr_warn("Subscriber rejected, unable to create port\n");
- kfree(subscriber);
- return;
- }
- tipc_connect2port(subscriber->port_ref, orig);
-
- /* Lock server port (& save lock address for future use) */
- subscriber->lock = tipc_port_lock(subscriber->port_ref)->lock;
-
- /* Add subscriber to topology server's subscriber list */
- spin_lock_bh(&topsrv.lock);
- list_add(&subscriber->subscriber_list, &topsrv.subscriber_list);
- spin_unlock_bh(&topsrv.lock);
-
- /* Unlock server port */
- server_port_ref = subscriber->port_ref;
- spin_unlock_bh(subscriber->lock);
-
- /* Send an ACK- to complete connection handshaking */
- tipc_send(server_port_ref, 0, NULL, 0);
+ subscriber->conid = conid;
+ spin_lock_init(&subscriber->lock);
- /* Handle optional subscription request */
- if (size != 0) {
- subscr_conn_msg_event(subscriber, server_port_ref,
- buf, data, size);
- }
+ return (void *)subscriber;
}
int tipc_subscr_start(void)
{
- struct tipc_name_seq seq = {TIPC_TOP_SRV, TIPC_TOP_SRV, TIPC_TOP_SRV};
- int res;
-
- spin_lock_init(&topsrv.lock);
- INIT_LIST_HEAD(&topsrv.subscriber_list);
-
- res = tipc_createport(NULL,
- TIPC_CRITICAL_IMPORTANCE,
- NULL,
- NULL,
- NULL,
- NULL,
- subscr_named_msg_event,
- NULL,
- NULL,
- &topsrv.setup_port);
- if (res)
- goto failed;
-
- res = tipc_publish(topsrv.setup_port, TIPC_NODE_SCOPE, &seq);
- if (res) {
- tipc_deleteport(topsrv.setup_port);
- topsrv.setup_port = 0;
- goto failed;
- }
-
- return 0;
-
-failed:
- pr_err("Failed to create subscription service\n");
- return res;
+ return tipc_server_start(&topsrv);
}
void tipc_subscr_stop(void)
{
- struct tipc_subscriber *subscriber;
- struct tipc_subscriber *subscriber_temp;
- spinlock_t *subscriber_lock;
-
- if (topsrv.setup_port) {
- tipc_deleteport(topsrv.setup_port);
- topsrv.setup_port = 0;
-
- list_for_each_entry_safe(subscriber, subscriber_temp,
- &topsrv.subscriber_list,
- subscriber_list) {
- subscriber_lock = subscriber->lock;
- spin_lock_bh(subscriber_lock);
- subscr_terminate(subscriber);
- spin_unlock_bh(subscriber_lock);
- }
- }
+ tipc_server_stop(&topsrv);
}
diff --git a/net/tipc/subscr.h b/net/tipc/subscr.h
index 218d2e07f0cc..393e417bee3f 100644
--- a/net/tipc/subscr.h
+++ b/net/tipc/subscr.h
@@ -2,7 +2,7 @@
* net/tipc/subscr.h: Include file for TIPC network topology service
*
* Copyright (c) 2003-2006, Ericsson AB
- * Copyright (c) 2005-2007, Wind River Systems
+ * Copyright (c) 2005-2007, 2012-2013, Wind River Systems
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
@@ -37,10 +37,14 @@
#ifndef _TIPC_SUBSCR_H
#define _TIPC_SUBSCR_H
+#include "server.h"
+
struct tipc_subscription;
+struct tipc_subscriber;
/**
* struct tipc_subscription - TIPC network topology subscription object
+ * @subscriber: pointer to its subscriber
* @seq: name sequence associated with subscription
* @timeout: duration of subscription (in ms)
* @filter: event filtering to be done for subscription
@@ -52,28 +56,23 @@ struct tipc_subscription;
* @evt: template for events generated by subscription
*/
struct tipc_subscription {
+ struct tipc_subscriber *subscriber;
struct tipc_name_seq seq;
u32 timeout;
u32 filter;
struct timer_list timer;
struct list_head nameseq_list;
struct list_head subscription_list;
- u32 server_ref;
int swap;
struct tipc_event evt;
};
-int tipc_subscr_overlap(struct tipc_subscription *sub,
- u32 found_lower,
+int tipc_subscr_overlap(struct tipc_subscription *sub, u32 found_lower,
u32 found_upper);
-void tipc_subscr_report_overlap(struct tipc_subscription *sub,
- u32 found_lower,
- u32 found_upper,
- u32 event,
- u32 port_ref,
- u32 node,
- int must_report);
+void tipc_subscr_report_overlap(struct tipc_subscription *sub, u32 found_lower,
+ u32 found_upper, u32 event, u32 port_ref,
+ u32 node, int must);
int tipc_subscr_start(void);
diff --git a/net/tipc/sysctl.c b/net/tipc/sysctl.c
new file mode 100644
index 000000000000..f3fef93325a8
--- /dev/null
+++ b/net/tipc/sysctl.c
@@ -0,0 +1,64 @@
+/*
+ * net/tipc/sysctl.c: sysctl interface to TIPC subsystem
+ *
+ * Copyright (c) 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 <linux/sysctl.h>
+
+static struct ctl_table_header *tipc_ctl_hdr;
+
+static struct ctl_table tipc_table[] = {
+ {
+ .procname = "tipc_rmem",
+ .data = &sysctl_tipc_rmem,
+ .maxlen = sizeof(sysctl_tipc_rmem),
+ .mode = 0644,
+ .proc_handler = proc_dointvec,
+ },
+ {}
+};
+
+int tipc_register_sysctl(void)
+{
+ tipc_ctl_hdr = register_net_sysctl(&init_net, "net/tipc", tipc_table);
+ if (tipc_ctl_hdr == NULL)
+ return -ENOMEM;
+ return 0;
+}
+
+void tipc_unregister_sysctl(void)
+{
+ unregister_net_sysctl_table(tipc_ctl_hdr);
+}