diff options
Diffstat (limited to 'net/tipc')
44 files changed, 4804 insertions, 6194 deletions
diff --git a/net/tipc/Kconfig b/net/tipc/Kconfig index 2c5954b8593..c890848f9d5 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 @@ -41,29 +32,9 @@ config TIPC_PORTS Setting this to a smaller value saves some memory, setting it to higher allows for more ports. -config TIPC_LOG - int "Size of log buffer" - depends on TIPC_ADVANCED - range 0 32768 - default "0" - help - Size (in bytes) of TIPC's internal log buffer, which records the - occurrence of significant events. Can range from 0 to 32768 bytes; - default is 0. - - There is no need to enable the log buffer unless the node will be - managed remotely via TIPC. - -config TIPC_DEBUG - bool "Enable debugging support" - default n +config TIPC_MEDIA_IB + bool "InfiniBand media type support" + depends on TIPC && INFINIBAND_IPOIB help - Saying Y here enables TIPC debugging capabilities used by developers. - Most users do not need to bother; if unsure, just say N. - - Enabling debugging support causes TIPC to display data about its - internal state when certain abnormal conditions occur. It also - makes it easy for developers to capture additional information of - interest using the dbg() or msg_dbg() macros. - -endif # TIPC + 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 521d24d04ab..a080c66d819 100644 --- a/net/tipc/Makefile +++ b/net/tipc/Makefile @@ -5,9 +5,10 @@ obj-$(CONFIG_TIPC) := tipc.o tipc-y += addr.o bcast.o bearer.o config.o \ - core.o handler.o link.o discover.o msg.o \ + core.o link.o discover.o msg.o \ name_distr.o subscr.o name_table.o net.o \ netlink.o node.o node_subscr.o port.o ref.o \ - socket.o log.o eth_media.o + socket.o log.o eth_media.o server.o -# End of file +tipc-$(CONFIG_TIPC_MEDIA_IB) += ib_media.o +tipc-$(CONFIG_SYSCTL) += sysctl.o diff --git a/net/tipc/addr.c b/net/tipc/addr.c index a6fdab33877..357b74b26f9 100644 --- a/net/tipc/addr.c +++ b/net/tipc/addr.c @@ -45,7 +45,6 @@ * * Returns 1 if domain address is valid, otherwise 0 */ - int tipc_addr_domain_valid(u32 addr) { u32 n = tipc_node(addr); @@ -66,7 +65,6 @@ int tipc_addr_domain_valid(u32 addr) * * Returns 1 if address can be used, otherwise 0 */ - int tipc_addr_node_valid(u32 addr) { return tipc_addr_domain_valid(addr) && tipc_node(addr); @@ -86,7 +84,6 @@ int tipc_in_scope(u32 domain, u32 addr) /** * tipc_addr_scope - convert message lookup domain to a 2-bit scope value */ - int tipc_addr_scope(u32 domain) { if (likely(!domain)) diff --git a/net/tipc/addr.h b/net/tipc/addr.h index e4f35afe320..a74acf9ee80 100644 --- a/net/tipc/addr.h +++ b/net/tipc/addr.h @@ -37,6 +37,8 @@ #ifndef _TIPC_ADDR_H #define _TIPC_ADDR_H +#include "core.h" + #define TIPC_ZONE_MASK 0xff000000u #define TIPC_CLUSTER_MASK 0xfffff000u @@ -50,18 +52,33 @@ static inline u32 tipc_cluster_mask(u32 addr) return addr & TIPC_CLUSTER_MASK; } -static inline int in_own_cluster(u32 addr) +static inline int in_own_cluster_exact(u32 addr) { return !((addr ^ tipc_own_addr) >> 12); } /** + * in_own_node - test for node inclusion; <0.0.0> always matches + */ +static inline int in_own_node(u32 addr) +{ + return (addr == tipc_own_addr) || !addr; +} + +/** + * in_own_cluster - test for cluster inclusion; <0.0.0> always matches + */ +static inline int in_own_cluster(u32 addr) +{ + return in_own_cluster_exact(addr) || !addr; +} + +/** * addr_domain - convert 2-bit scope value to equivalent message lookup domain * * Needed when address of a named message must be looked up a second time * after a network hop. */ - static inline u32 addr_domain(u32 sc) { if (likely(sc == TIPC_NODE_SCOPE)) diff --git a/net/tipc/bcast.c b/net/tipc/bcast.c index 8eb87b11d10..55c6c9d3e1c 100644 --- a/net/tipc/bcast.c +++ b/net/tipc/bcast.c @@ -41,9 +41,9 @@ #include "bcast.h" #include "name_distr.h" -#define MAX_PKT_DEFAULT_MCAST 1500 /* bcast link max packet size (fixed) */ - -#define BCLINK_WIN_DEFAULT 20 /* bcast link window size (default) */ +#define MAX_PKT_DEFAULT_MCAST 1500 /* bcast link max packet size (fixed) */ +#define BCLINK_WIN_DEFAULT 20 /* bcast link window size (default) */ +#define BCBEARER MAX_BEARERS /** * struct tipc_bcbearer_pair - a pair of bearers used by broadcast link @@ -71,9 +71,8 @@ struct tipc_bcbearer_pair { * Note: The fields labelled "temporary" are incorporated into the bearer * to avoid consuming potentially limited stack space through the use of * large local variables within multicast routines. Concurrent access is - * prevented through use of the spinlock "bc_lock". + * prevented through use of the spinlock "bclink_lock". */ - struct tipc_bcbearer { struct tipc_bearer bearer; struct tipc_media media; @@ -85,35 +84,64 @@ struct tipc_bcbearer { /** * struct tipc_bclink - link used for broadcast messages + * @lock: spinlock governing access to structure * @link: (non-standard) broadcast link structure * @node: (non-standard) node structure representing b'cast link's peer node + * @flags: represent bclink states * @bcast_nodes: map of broadcast-capable nodes * @retransmit_to: node that most recently requested a retransmit * * Handles sequence numbering, fragmentation, bundling, etc. */ - struct tipc_bclink { + spinlock_t lock; struct tipc_link link; struct tipc_node node; + unsigned int flags; struct tipc_node_map bcast_nodes; struct tipc_node *retransmit_to; }; -static struct tipc_bcbearer bcast_bearer; -static struct tipc_bclink bcast_link; - -static struct tipc_bcbearer *bcbearer = &bcast_bearer; -static struct tipc_bclink *bclink = &bcast_link; -static struct tipc_link *bcl = &bcast_link.link; - -static DEFINE_SPINLOCK(bc_lock); +static struct tipc_bcbearer *bcbearer; +static struct tipc_bclink *bclink; +static struct tipc_link *bcl; const char tipc_bclink_name[] = "broadcast-link"; static void tipc_nmap_diff(struct tipc_node_map *nm_a, struct tipc_node_map *nm_b, struct tipc_node_map *nm_diff); +static void tipc_nmap_add(struct tipc_node_map *nm_ptr, u32 node); +static void tipc_nmap_remove(struct tipc_node_map *nm_ptr, u32 node); + +static void tipc_bclink_lock(void) +{ + spin_lock_bh(&bclink->lock); +} + +static void tipc_bclink_unlock(void) +{ + struct tipc_node *node = NULL; + + if (likely(!bclink->flags)) { + spin_unlock_bh(&bclink->lock); + return; + } + + if (bclink->flags & TIPC_BCLINK_RESET) { + bclink->flags &= ~TIPC_BCLINK_RESET; + node = tipc_bclink_retransmit_to(); + } + spin_unlock_bh(&bclink->lock); + + if (node) + tipc_link_reset_all(node); +} + +void tipc_bclink_set_flags(unsigned int flags) +{ + bclink->flags |= flags; +} static u32 bcbuf_acks(struct sk_buff *buf) { @@ -132,16 +160,16 @@ static void bcbuf_decr_acks(struct sk_buff *buf) void tipc_bclink_add_node(u32 addr) { - spin_lock_bh(&bc_lock); + tipc_bclink_lock(); tipc_nmap_add(&bclink->bcast_nodes, addr); - spin_unlock_bh(&bc_lock); + tipc_bclink_unlock(); } void tipc_bclink_remove_node(u32 addr) { - spin_lock_bh(&bc_lock); + tipc_bclink_lock(); tipc_nmap_remove(&bclink->bcast_nodes, addr); - spin_unlock_bh(&bc_lock); + tipc_bclink_unlock(); } static void bclink_set_last_sent(void) @@ -157,44 +185,18 @@ u32 tipc_bclink_get_last_sent(void) return bcl->fsm_msg_cnt; } -/** - * bclink_set_gap - set gap according to contents of current deferred pkt queue - * - * Called with 'node' locked, bc_lock unlocked - */ - -static void bclink_set_gap(struct tipc_node *n_ptr) +static void bclink_update_last_sent(struct tipc_node *node, u32 seqno) { - struct sk_buff *buf = n_ptr->bclink.deferred_head; - - n_ptr->bclink.gap_after = n_ptr->bclink.gap_to = - mod(n_ptr->bclink.last_in); - if (unlikely(buf != NULL)) - n_ptr->bclink.gap_to = mod(buf_seqno(buf) - 1); -} - -/** - * bclink_ack_allowed - test if ACK or NACK message can be sent at this moment - * - * This mechanism endeavours to prevent all nodes in network from trying - * to ACK or NACK at the same time. - * - * Note: TIPC uses a different trigger to distribute ACKs than it does to - * distribute NACKs, but tries to use the same spacing (divide by 16). - */ - -static int bclink_ack_allowed(u32 n) -{ - return (n % TIPC_MIN_LINK_WIN) == tipc_own_tag; + node->bclink.last_sent = less_eq(node->bclink.last_sent, seqno) ? + seqno : node->bclink.last_sent; } /** * tipc_bclink_retransmit_to - get most recent node to request retransmission * - * Called with bc_lock locked + * Called with bclink_lock locked */ - struct tipc_node *tipc_bclink_retransmit_to(void) { return bclink->retransmit_to; @@ -205,9 +207,8 @@ struct tipc_node *tipc_bclink_retransmit_to(void) * @after: sequence number of last packet to *not* retransmit * @to: sequence number of last packet to retransmit * - * Called with bc_lock locked + * Called with bclink_lock locked */ - static void bclink_retransmit_pkt(u32 after, u32 to) { struct sk_buff *buf; @@ -223,17 +224,15 @@ static void bclink_retransmit_pkt(u32 after, u32 to) * @n_ptr: node that sent acknowledgement info * @acked: broadcast sequence # that has been acknowledged * - * Node is locked, bc_lock unlocked. + * Node is locked, bclink_lock unlocked. */ - void tipc_bclink_acknowledge(struct tipc_node *n_ptr, u32 acked) { struct sk_buff *crs; struct sk_buff *next; unsigned int released = 0; - spin_lock_bh(&bc_lock); - + tipc_bclink_lock(); /* Bail out if tx queue is empty (no clean up is required) */ crs = bcl->first_out; if (!crs) @@ -281,7 +280,7 @@ void tipc_bclink_acknowledge(struct tipc_node *n_ptr, u32 acked) if (bcbuf_acks(crs) == 0) { bcl->first_out = next; bcl->out_queue_size--; - buf_discard(crs); + kfree_skb(crs); released = 1; } crs = next; @@ -297,182 +296,154 @@ void tipc_bclink_acknowledge(struct tipc_node *n_ptr, u32 acked) if (unlikely(released && !list_empty(&bcl->waiting_ports))) tipc_link_wakeup_ports(bcl, 0); exit: - spin_unlock_bh(&bc_lock); + tipc_bclink_unlock(); } /** - * bclink_send_ack - unicast an ACK msg + * tipc_bclink_update_link_state - update broadcast link state * - * tipc_net_lock and node lock set + * RCU and node lock set */ - -static void bclink_send_ack(struct tipc_node *n_ptr) +void tipc_bclink_update_link_state(struct tipc_node *n_ptr, u32 last_sent) { - struct tipc_link *l_ptr = n_ptr->active_links[n_ptr->addr & 1]; + struct sk_buff *buf; - if (l_ptr != NULL) - tipc_link_send_proto_msg(l_ptr, STATE_MSG, 0, 0, 0, 0, 0); -} + /* Ignore "stale" link state info */ -/** - * bclink_send_nack- broadcast a NACK msg - * - * tipc_net_lock and node lock set - */ + if (less_eq(last_sent, n_ptr->bclink.last_in)) + return; -static void bclink_send_nack(struct tipc_node *n_ptr) -{ - struct sk_buff *buf; - struct tipc_msg *msg; + /* Update link synchronization state; quit if in sync */ + + bclink_update_last_sent(n_ptr, last_sent); + + if (n_ptr->bclink.last_sent == n_ptr->bclink.last_in) + return; - if (!less(n_ptr->bclink.gap_after, n_ptr->bclink.gap_to)) + /* Update out-of-sync state; quit if loss is still unconfirmed */ + + if ((++n_ptr->bclink.oos_state) == 1) { + if (n_ptr->bclink.deferred_size < (TIPC_MIN_LINK_WIN / 2)) + return; + n_ptr->bclink.oos_state++; + } + + /* Don't NACK if one has been recently sent (or seen) */ + + if (n_ptr->bclink.oos_state & 0x1) return; + /* Send NACK */ + buf = tipc_buf_acquire(INT_H_SIZE); if (buf) { - msg = buf_msg(buf); + struct tipc_msg *msg = buf_msg(buf); + tipc_msg_init(msg, BCAST_PROTOCOL, STATE_MSG, - INT_H_SIZE, n_ptr->addr); + INT_H_SIZE, n_ptr->addr); msg_set_non_seq(msg, 1); msg_set_mc_netid(msg, tipc_net_id); - msg_set_bcast_ack(msg, mod(n_ptr->bclink.last_in)); - msg_set_bcgap_after(msg, n_ptr->bclink.gap_after); - msg_set_bcgap_to(msg, n_ptr->bclink.gap_to); - msg_set_bcast_tag(msg, tipc_own_tag); - - tipc_bearer_send(&bcbearer->bearer, buf, NULL); + msg_set_bcast_ack(msg, n_ptr->bclink.last_in); + msg_set_bcgap_after(msg, n_ptr->bclink.last_in); + msg_set_bcgap_to(msg, n_ptr->bclink.deferred_head + ? buf_seqno(n_ptr->bclink.deferred_head) - 1 + : n_ptr->bclink.last_sent); + + tipc_bclink_lock(); + tipc_bearer_send(MAX_BEARERS, buf, NULL); bcl->stats.sent_nacks++; - buf_discard(buf); + tipc_bclink_unlock(); + kfree_skb(buf); - /* - * Ensure we doesn't send another NACK msg to the node - * until 16 more deferred messages arrive from it - * (i.e. helps prevent all nodes from NACK'ing at same time) - */ - - n_ptr->bclink.nack_sync = tipc_own_tag; + n_ptr->bclink.oos_state++; } } /** - * tipc_bclink_check_gap - send a NACK if a sequence gap exists + * bclink_peek_nack - monitor retransmission requests sent by other nodes * - * tipc_net_lock and node lock set + * Delay any upcoming NACK by this node if another node has already + * requested the first message this node is going to ask for. */ - -void tipc_bclink_check_gap(struct tipc_node *n_ptr, u32 last_sent) +static void bclink_peek_nack(struct tipc_msg *msg) { - if (!n_ptr->bclink.supported || - less_eq(last_sent, mod(n_ptr->bclink.last_in))) - return; - - bclink_set_gap(n_ptr); - if (n_ptr->bclink.gap_after == n_ptr->bclink.gap_to) - n_ptr->bclink.gap_to = last_sent; - bclink_send_nack(n_ptr); -} - -/** - * tipc_bclink_peek_nack - process a NACK msg meant for another node - * - * Only tipc_net_lock set. - */ + struct tipc_node *n_ptr = tipc_node_find(msg_destnode(msg)); -static void tipc_bclink_peek_nack(u32 dest, u32 sender_tag, u32 gap_after, u32 gap_to) -{ - struct tipc_node *n_ptr = tipc_node_find(dest); - u32 my_after, my_to; - - if (unlikely(!n_ptr || !tipc_node_is_up(n_ptr))) + if (unlikely(!n_ptr)) return; + tipc_node_lock(n_ptr); - /* - * Modify gap to suppress unnecessary NACKs from this node - */ - my_after = n_ptr->bclink.gap_after; - my_to = n_ptr->bclink.gap_to; - - if (less_eq(gap_after, my_after)) { - if (less(my_after, gap_to) && less(gap_to, my_to)) - n_ptr->bclink.gap_after = gap_to; - else if (less_eq(my_to, gap_to)) - n_ptr->bclink.gap_to = n_ptr->bclink.gap_after; - } else if (less_eq(gap_after, my_to)) { - if (less_eq(my_to, gap_to)) - n_ptr->bclink.gap_to = gap_after; - } else { - /* - * Expand gap if missing bufs not in deferred queue: - */ - struct sk_buff *buf = n_ptr->bclink.deferred_head; - u32 prev = n_ptr->bclink.gap_to; - for (; buf; buf = buf->next) { - u32 seqno = buf_seqno(buf); + 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; - if (mod(seqno - prev) != 1) { - buf = NULL; - break; - } - if (seqno == gap_after) - break; - prev = seqno; - } - if (buf == NULL) - n_ptr->bclink.gap_to = gap_after; - } - /* - * Some nodes may send a complementary NACK now: - */ - if (bclink_ack_allowed(sender_tag + 1)) { - if (n_ptr->bclink.gap_to != n_ptr->bclink.gap_after) { - bclink_send_nack(n_ptr); - bclink_set_gap(n_ptr); - } - } tipc_node_unlock(n_ptr); } -/** - * tipc_bclink_send_msg - broadcast a packet to all nodes in cluster +/* + * tipc_bclink_xmit - broadcast a packet to all nodes in cluster */ - -int tipc_bclink_send_msg(struct sk_buff *buf) +int tipc_bclink_xmit(struct sk_buff *buf) { int res; - spin_lock_bh(&bc_lock); + tipc_bclink_lock(); if (!bclink->bcast_nodes.count) { res = msg_data_sz(buf_msg(buf)); - buf_discard(buf); + kfree_skb(buf); goto exit; } - res = tipc_link_send_buf(bcl, buf); + res = __tipc_link_xmit(bcl, buf); if (likely(res >= 0)) { bclink_set_last_sent(); bcl->stats.queue_sz_counts++; bcl->stats.accu_queue_sz += bcl->out_queue_size; } exit: - spin_unlock_bh(&bc_lock); + tipc_bclink_unlock(); return res; } /** - * tipc_bclink_recv_pkt - receive a broadcast packet, and deliver upwards + * bclink_accept_pkt - accept an incoming, in-sequence broadcast packet * - * tipc_net_lock is read_locked, no other locks set + * Called with both sending node's lock and bclink_lock taken. */ +static void bclink_accept_pkt(struct tipc_node *node, u32 seqno) +{ + bclink_update_last_sent(node, seqno); + node->bclink.last_in = seqno; + node->bclink.oos_state = 0; + bcl->stats.recv_info++; + + /* + * Unicast an ACK periodically, ensuring that + * all nodes in the cluster don't ACK at the same time + */ -void tipc_bclink_recv_pkt(struct sk_buff *buf) + if (((seqno - tipc_own_addr) % TIPC_MIN_LINK_WIN) == 0) { + tipc_link_proto_xmit(node->active_links[node->addr & 1], + STATE_MSG, 0, 0, 0, 0, 0); + bcl->stats.sent_acks++; + } +} + +/** + * tipc_bclink_rcv - receive a broadcast packet, and deliver upwards + * + * RCU is locked, no other locks set + */ +void tipc_bclink_rcv(struct sk_buff *buf) { struct tipc_msg *msg = buf_msg(buf); struct tipc_node *node; u32 next_in; u32 seqno; - struct sk_buff *deferred; + int deferred; /* Screen out unwanted broadcast messages */ @@ -484,111 +455,145 @@ 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 */ + if (unlikely(msg_user(msg) == BCAST_PROTOCOL)) { if (msg_type(msg) != STATE_MSG) goto unlock; if (msg_destnode(msg) == tipc_own_addr) { tipc_bclink_acknowledge(node, msg_bcast_ack(msg)); tipc_node_unlock(node); - spin_lock_bh(&bc_lock); + tipc_bclink_lock(); bcl->stats.recv_nacks++; bclink->retransmit_to = node; bclink_retransmit_pkt(msg_bcgap_after(msg), msg_bcgap_to(msg)); - spin_unlock_bh(&bc_lock); + tipc_bclink_unlock(); } else { tipc_node_unlock(node); - tipc_bclink_peek_nack(msg_destnode(msg), - msg_bcast_tag(msg), - msg_bcgap_after(msg), - msg_bcgap_to(msg)); + bclink_peek_nack(msg); } goto exit; } /* Handle in-sequence broadcast message */ -receive: - next_in = mod(node->bclink.last_in + 1); seqno = msg_seqno(msg); + next_in = mod(node->bclink.last_in + 1); if (likely(seqno == next_in)) { - bcl->stats.recv_info++; - node->bclink.last_in++; - bclink_set_gap(node); - if (unlikely(bclink_ack_allowed(seqno))) { - bclink_send_ack(node); - bcl->stats.sent_acks++; - } +receive: + /* Deliver message to destination */ + if (likely(msg_isdata(msg))) { + tipc_bclink_lock(); + bclink_accept_pkt(node, seqno); + tipc_bclink_unlock(); tipc_node_unlock(node); if (likely(msg_mcast(msg))) - tipc_port_recv_mcast(buf, NULL); + tipc_port_mcast_rcv(buf, NULL); else - buf_discard(buf); + kfree_skb(buf); } else if (msg_user(msg) == MSG_BUNDLER) { + tipc_bclink_lock(); + bclink_accept_pkt(node, seqno); bcl->stats.recv_bundles++; bcl->stats.recv_bundled += msg_msgcnt(msg); + tipc_bclink_unlock(); tipc_node_unlock(node); - tipc_link_recv_bundle(buf); + tipc_link_bundle_rcv(buf); } else if (msg_user(msg) == MSG_FRAGMENTER) { + tipc_buf_append(&node->bclink.reasm_buf, &buf); + if (unlikely(!buf && !node->bclink.reasm_buf)) + goto unlock; + tipc_bclink_lock(); + bclink_accept_pkt(node, seqno); bcl->stats.recv_fragments++; - if (tipc_link_recv_fragment(&node->bclink.defragm, - &buf, &msg)) + if (buf) { bcl->stats.recv_fragmented++; + msg = buf_msg(buf); + tipc_bclink_unlock(); + goto receive; + } + tipc_bclink_unlock(); tipc_node_unlock(node); - tipc_net_route_msg(buf); } else if (msg_user(msg) == NAME_DISTRIBUTOR) { + tipc_bclink_lock(); + bclink_accept_pkt(node, seqno); + tipc_bclink_unlock(); tipc_node_unlock(node); - tipc_named_recv(buf); + tipc_named_rcv(buf); } else { + tipc_bclink_lock(); + bclink_accept_pkt(node, seqno); + tipc_bclink_unlock(); tipc_node_unlock(node); - buf_discard(buf); + kfree_skb(buf); } buf = NULL; + + /* Determine new synchronization state */ + tipc_node_lock(node); - deferred = node->bclink.deferred_head; - if (deferred && (buf_seqno(deferred) == mod(next_in + 1))) { - buf = deferred; - msg = buf_msg(buf); - node->bclink.deferred_head = deferred->next; - goto receive; - } - } else if (less(next_in, seqno)) { - u32 gap_after = node->bclink.gap_after; - u32 gap_to = node->bclink.gap_to; - - if (tipc_link_defer_pkt(&node->bclink.deferred_head, - &node->bclink.deferred_tail, - buf)) { - node->bclink.nack_sync++; - bcl->stats.deferred_recv++; - if (seqno == mod(gap_after + 1)) - node->bclink.gap_after = seqno; - else if (less(gap_after, seqno) && less(seqno, gap_to)) - node->bclink.gap_to = seqno; + if (unlikely(!tipc_node_is_up(node))) + goto unlock; + + if (node->bclink.last_in == node->bclink.last_sent) + goto unlock; + + if (!node->bclink.deferred_head) { + node->bclink.oos_state = 1; + goto unlock; } + + msg = buf_msg(node->bclink.deferred_head); + seqno = msg_seqno(msg); + next_in = mod(next_in + 1); + if (seqno != next_in) + goto unlock; + + /* Take in-sequence message from deferred queue & deliver it */ + + buf = node->bclink.deferred_head; + node->bclink.deferred_head = buf->next; + buf->next = NULL; + node->bclink.deferred_size--; + goto receive; + } + + /* Handle out-of-sequence broadcast message */ + + if (less(next_in, seqno)) { + deferred = tipc_link_defer_pkt(&node->bclink.deferred_head, + &node->bclink.deferred_tail, + buf); + node->bclink.deferred_size += deferred; + bclink_update_last_sent(node, seqno); buf = NULL; - if (bclink_ack_allowed(node->bclink.nack_sync)) { - if (gap_to != gap_after) - bclink_send_nack(node); - bclink_set_gap(node); - } - } else { + } else + deferred = 0; + + tipc_bclink_lock(); + + if (deferred) + bcl->stats.deferred_recv++; + else bcl->stats.duplicates++; - } + + tipc_bclink_unlock(); + unlock: tipc_node_unlock(node); exit: - buf_discard(buf); + kfree_skb(buf); } 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)); } @@ -602,20 +607,16 @@ 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 */ - if (likely(!msg_non_seq(buf_msg(buf)))) { struct tipc_msg *msg; @@ -632,38 +633,42 @@ static int tipc_bcbearer_send(struct sk_buff *buf, } /* Send buffer over bearers until all targets reached */ - bcbearer->remains = bclink->bcast_nodes; 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 */ - 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->identity, buf, &b->bcast_addr); + } else { + /* Avoid concurrent buffer access */ + tbuf = pskb_copy_for_clone(buf, GFP_ATOMIC); + if (!tbuf) + break; + tipc_bearer_send(b->identity, 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; } @@ -674,24 +679,28 @@ static int tipc_bcbearer_send(struct sk_buff *buf, /** * tipc_bcbearer_sort - create sets of bearer pairs used by broadcast bearer */ - -void tipc_bcbearer_sort(void) +void tipc_bcbearer_sort(struct tipc_node_map *nm_ptr, u32 node, bool action) { struct tipc_bcbearer_pair *bp_temp = bcbearer->bpairs_temp; struct tipc_bcbearer_pair *bp_curr; + struct tipc_bearer *b; int b_index; int pri; - spin_lock_bh(&bc_lock); + tipc_bclink_lock(); - /* Group bearers by priority (can assume max of two per priority) */ + if (action) + tipc_nmap_add(nm_ptr, node); + else + tipc_nmap_remove(nm_ptr, node); + /* Group bearers by priority (can assume max of two per priority) */ memset(bp_temp, 0, sizeof(bcbearer->bpairs_temp)); + rcu_read_lock(); for (b_index = 0; b_index < MAX_BEARERS; b_index++) { - struct tipc_bearer *b = &tipc_bearers[b_index]; - - if (!b->active || !b->nodes.count) + b = rcu_dereference_rtnl(bearer_list[b_index]); + if (!b || !b->nodes.count) continue; if (!bp_temp[b->priority].primary) @@ -699,9 +708,9 @@ void tipc_bcbearer_sort(void) else bp_temp[b->priority].secondary = b; } + rcu_read_unlock(); /* Create array of bearer pairs for broadcasting */ - bp_curr = bcbearer->bpairs; memset(bcbearer->bpairs, 0, sizeof(bcbearer->bpairs)); @@ -725,54 +734,49 @@ void tipc_bcbearer_sort(void) bp_curr++; } - spin_unlock_bh(&bc_lock); + tipc_bclink_unlock(); } int tipc_bclink_stats(char *buf, const u32 buf_size) { - struct print_buf pb; + int ret; + struct tipc_stats *s; if (!bcl) return 0; - tipc_printbuf_init(&pb, buf, buf_size); - - spin_lock_bh(&bc_lock); - - tipc_printf(&pb, "Link <%s>\n" - " Window:%u packets\n", - bcl->name, bcl->queue_limit[0]); - tipc_printf(&pb, " RX packets:%u fragments:%u/%u bundles:%u/%u\n", - bcl->stats.recv_info, - bcl->stats.recv_fragments, - bcl->stats.recv_fragmented, - bcl->stats.recv_bundles, - bcl->stats.recv_bundled); - tipc_printf(&pb, " TX packets:%u fragments:%u/%u bundles:%u/%u\n", - bcl->stats.sent_info, - bcl->stats.sent_fragments, - bcl->stats.sent_fragmented, - bcl->stats.sent_bundles, - bcl->stats.sent_bundled); - tipc_printf(&pb, " RX naks:%u defs:%u dups:%u\n", - bcl->stats.recv_nacks, - bcl->stats.deferred_recv, - bcl->stats.duplicates); - tipc_printf(&pb, " TX naks:%u acks:%u dups:%u\n", - bcl->stats.sent_nacks, - bcl->stats.sent_acks, - bcl->stats.retransmitted); - tipc_printf(&pb, " Congestion bearer:%u link:%u Send queue max:%u avg:%u\n", - bcl->stats.bearer_congs, - bcl->stats.link_congs, - bcl->stats.max_queue_sz, - bcl->stats.queue_sz_counts - ? (bcl->stats.accu_queue_sz / bcl->stats.queue_sz_counts) - : 0); - - spin_unlock_bh(&bc_lock); - return tipc_printbuf_validate(&pb); + tipc_bclink_lock(); + + s = &bcl->stats; + + ret = tipc_snprintf(buf, buf_size, "Link <%s>\n" + " Window:%u packets\n", + bcl->name, bcl->queue_limit[0]); + ret += tipc_snprintf(buf + ret, buf_size - ret, + " RX packets:%u fragments:%u/%u bundles:%u/%u\n", + s->recv_info, s->recv_fragments, + s->recv_fragmented, s->recv_bundles, + s->recv_bundled); + ret += tipc_snprintf(buf + ret, buf_size - ret, + " TX packets:%u fragments:%u/%u bundles:%u/%u\n", + s->sent_info, s->sent_fragments, + s->sent_fragmented, s->sent_bundles, + s->sent_bundled); + ret += tipc_snprintf(buf + ret, buf_size - ret, + " RX naks:%u defs:%u dups:%u\n", + s->recv_nacks, s->deferred_recv, s->duplicates); + ret += tipc_snprintf(buf + ret, buf_size - ret, + " 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 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); + + tipc_bclink_unlock(); + return ret; } int tipc_bclink_reset_stats(void) @@ -780,9 +784,9 @@ int tipc_bclink_reset_stats(void) if (!bcl) return -ENOPROTOOPT; - spin_lock_bh(&bc_lock); + tipc_bclink_lock(); memset(&bcl->stats, 0, sizeof(bcl->stats)); - spin_unlock_bh(&bc_lock); + tipc_bclink_unlock(); return 0; } @@ -793,46 +797,59 @@ int tipc_bclink_set_queue_limits(u32 limit) if ((limit < TIPC_MIN_LINK_WIN) || (limit > TIPC_MAX_LINK_WIN)) return -EINVAL; - spin_lock_bh(&bc_lock); + tipc_bclink_lock(); tipc_link_set_queue_limits(bcl, limit); - spin_unlock_bh(&bc_lock); + tipc_bclink_unlock(); return 0; } -void tipc_bclink_init(void) +int tipc_bclink_init(void) { - INIT_LIST_HEAD(&bcbearer->bearer.cong_links); + bcbearer = kzalloc(sizeof(*bcbearer), GFP_ATOMIC); + if (!bcbearer) + return -ENOMEM; + + bclink = kzalloc(sizeof(*bclink), GFP_ATOMIC); + if (!bclink) { + kfree(bcbearer); + return -ENOMEM; + } + + bcl = &bclink->link; bcbearer->bearer.media = &bcbearer->media; bcbearer->media.send_msg = tipc_bcbearer_send; sprintf(bcbearer->media.name, "tipc-broadcast"); + spin_lock_init(&bclink->lock); INIT_LIST_HEAD(&bcl->waiting_ports); bcl->next_out_no = 1; spin_lock_init(&bclink->node.lock); bcl->owner = &bclink->node; bcl->max_pkt = MAX_PKT_DEFAULT_MCAST; tipc_link_set_queue_limits(bcl, BCLINK_WIN_DEFAULT); - bcl->b_ptr = &bcbearer->bearer; + bcl->bearer_id = MAX_BEARERS; + rcu_assign_pointer(bearer_list[MAX_BEARERS], &bcbearer->bearer); bcl->state = WORKING_WORKING; strlcpy(bcl->name, tipc_bclink_name, TIPC_MAX_LINK_NAME); + return 0; } void tipc_bclink_stop(void) { - spin_lock_bh(&bc_lock); - tipc_link_stop(bcl); - spin_unlock_bh(&bc_lock); - - memset(bclink, 0, sizeof(*bclink)); - memset(bcbearer, 0, sizeof(*bcbearer)); + tipc_bclink_lock(); + tipc_link_purge_queues(bcl); + tipc_bclink_unlock(); + + RCU_INIT_POINTER(bearer_list[BCBEARER], NULL); + synchronize_net(); + kfree(bcbearer); + kfree(bclink); } - /** * tipc_nmap_add - add a node to a node map */ - -void tipc_nmap_add(struct tipc_node_map *nm_ptr, u32 node) +static void tipc_nmap_add(struct tipc_node_map *nm_ptr, u32 node) { int n = tipc_node(node); int w = n / WSIZE; @@ -847,8 +864,7 @@ void tipc_nmap_add(struct tipc_node_map *nm_ptr, u32 node) /** * tipc_nmap_remove - remove a node from a node map */ - -void tipc_nmap_remove(struct tipc_node_map *nm_ptr, u32 node) +static void tipc_nmap_remove(struct tipc_node_map *nm_ptr, u32 node) { int n = tipc_node(node); int w = n / WSIZE; @@ -866,7 +882,6 @@ void tipc_nmap_remove(struct tipc_node_map *nm_ptr, u32 node) * @nm_b: input node map B * @nm_diff: output node map A-B (i.e. nodes of A that are not in B) */ - static void tipc_nmap_diff(struct tipc_node_map *nm_a, struct tipc_node_map *nm_b, struct tipc_node_map *nm_diff) @@ -892,7 +907,6 @@ static void tipc_nmap_diff(struct tipc_node_map *nm_a, /** * tipc_port_list_add - add a port to a port list, ensuring no duplicates */ - void tipc_port_list_add(struct tipc_port_list *pl_ptr, u32 port) { struct tipc_port_list *item = pl_ptr; @@ -914,7 +928,7 @@ void tipc_port_list_add(struct tipc_port_list *pl_ptr, u32 port) if (!item->next) { item->next = kmalloc(sizeof(*item), GFP_ATOMIC); if (!item->next) { - warn("Incomplete multicast delivery, no memory\n"); + pr_warn("Incomplete multicast delivery, no memory\n"); return; } item->next->next = NULL; @@ -926,7 +940,6 @@ void tipc_port_list_add(struct tipc_port_list *pl_ptr, u32 port) * tipc_port_list_free - free dynamically created entries in port_list chain * */ - void tipc_port_list_free(struct tipc_port_list *pl_ptr) { struct tipc_port_list *item; @@ -937,4 +950,3 @@ void tipc_port_list_free(struct tipc_port_list *pl_ptr) kfree(item); } } - diff --git a/net/tipc/bcast.h b/net/tipc/bcast.h index b009666c60b..00330c45df3 100644 --- a/net/tipc/bcast.h +++ b/net/tipc/bcast.h @@ -39,13 +39,13 @@ #define MAX_NODES 4096 #define WSIZE 32 +#define TIPC_BCLINK_RESET 1 /** * struct tipc_node_map - set of node identifiers * @count: # of nodes in set * @map: bitmap of node identifiers that are in the set */ - struct tipc_node_map { u32 count; u32 map[MAX_NODES / WSIZE]; @@ -59,7 +59,6 @@ struct tipc_node_map { * @next: pointer to next entry in list * @ports: array of port references */ - struct tipc_port_list { int count; struct tipc_port_list *next; @@ -71,14 +70,11 @@ struct tipc_node; extern const char tipc_bclink_name[]; -void tipc_nmap_add(struct tipc_node_map *nm_ptr, u32 node); -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)); } @@ -86,20 +82,21 @@ static inline int tipc_nmap_equal(struct tipc_node_map *nm_a, struct tipc_node_m void tipc_port_list_add(struct tipc_port_list *pl_ptr, u32 port); void tipc_port_list_free(struct tipc_port_list *pl_ptr); -void tipc_bclink_init(void); +int tipc_bclink_init(void); void tipc_bclink_stop(void); +void tipc_bclink_set_flags(unsigned int flags); void tipc_bclink_add_node(u32 addr); void tipc_bclink_remove_node(u32 addr); struct tipc_node *tipc_bclink_retransmit_to(void); void tipc_bclink_acknowledge(struct tipc_node *n_ptr, u32 acked); -int tipc_bclink_send_msg(struct sk_buff *buf); -void tipc_bclink_recv_pkt(struct sk_buff *buf); +int tipc_bclink_xmit(struct sk_buff *buf); +void tipc_bclink_rcv(struct sk_buff *buf); u32 tipc_bclink_get_last_sent(void); u32 tipc_bclink_acks_missing(struct tipc_node *n_ptr); -void tipc_bclink_check_gap(struct tipc_node *n_ptr, u32 seqno); +void tipc_bclink_update_link_state(struct tipc_node *n_ptr, u32 last_sent); int tipc_bclink_stats(char *stats_buf, const u32 buf_size); int tipc_bclink_reset_stats(void); int tipc_bclink_set_queue_limits(u32 limit); -void tipc_bcbearer_sort(void); +void tipc_bcbearer_sort(struct tipc_node_map *nm_ptr, u32 node, bool action); #endif diff --git a/net/tipc/bearer.c b/net/tipc/bearer.c index 329fb659fae..264474394f9 100644 --- a/net/tipc/bearer.c +++ b/net/tipc/bearer.c @@ -1,8 +1,8 @@ /* * net/tipc/bearer.c: TIPC bearer code * - * Copyright (c) 1996-2006, Ericsson AB - * Copyright (c) 2004-2006, 2010-2011, Wind River Systems + * Copyright (c) 1996-2006, 2013, Ericsson AB + * Copyright (c) 2004-2006, 2010-2013, Wind River Systems * All rights reserved. * * Redistribution and use in source and binary forms, with or without @@ -39,124 +39,74 @@ #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; +static struct tipc_media * const media_info_array[] = { + ð_media_info, +#ifdef CONFIG_TIPC_MEDIA_IB + &ib_media_info, +#endif + NULL +}; -struct tipc_bearer tipc_bearers[MAX_BEARERS]; +struct tipc_bearer __rcu *bearer_list[MAX_BEARERS + 1]; -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; -} +static void bearer_disable(struct tipc_bearer *b_ptr, bool shutting_down); /** * tipc_media_find - locates specified media object by name */ - struct tipc_media *tipc_media_find(const char *name) { u32 i; - for (i = 0; i < media_count; i++) { - if (!strcmp(media_list[i]->name, name)) - return media_list[i]; + for (i = 0; media_info_array[i] != NULL; i++) { + if (!strcmp(media_info_array[i]->name, name)) + break; } - return NULL; + return media_info_array[i]; } /** * media_find_id - locates specified media object by type identifier */ - static struct tipc_media *media_find_id(u8 type) { u32 i; - for (i = 0; i < media_count; i++) { - if (media_list[i]->type_id == type) - return media_list[i]; + for (i = 0; media_info_array[i] != NULL; i++) { + if (media_info_array[i]->type_id == type) + break; } - return NULL; -} - -/** - * tipc_register_media - register a media type - * - * Bearers for this media type must be activated separately at a later stage. - */ - -int tipc_register_media(struct tipc_media *m_ptr) -{ - int res = -EINVAL; - - 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) - goto exit; - if (m_ptr->priority > TIPC_MAX_LINK_PRI) - goto exit; - if ((m_ptr->tolerance < TIPC_MIN_LINK_TOL) || - (m_ptr->tolerance > TIPC_MAX_LINK_TOL)) - goto exit; - if (media_count >= MAX_MEDIA) - goto exit; - if (tipc_media_find(m_ptr->name) || media_find_id(m_ptr->type_id)) - goto exit; - - media_list[media_count] = m_ptr; - media_count++; - res = 0; -exit: - write_unlock_bh(&tipc_net_lock); - if (res) - warn("Media <%s> registration error\n", m_ptr->name); - return res; + return media_info_array[i]; } /** * tipc_media_addr_printf - record media address in print buffer */ - -void tipc_media_addr_printf(struct print_buf *pb, struct tipc_media_addr *a) +void tipc_media_addr_printf(char *buf, int len, struct tipc_media_addr *a) { char addr_str[MAX_ADDR_STR]; struct tipc_media *m_ptr; + int ret; m_ptr = media_find_id(a->media_id); if (m_ptr && !m_ptr->addr2str(a, addr_str, sizeof(addr_str))) - tipc_printf(pb, "%s(%s)", m_ptr->name, addr_str); + ret = tipc_snprintf(buf, len, "%s(%s)", m_ptr->name, addr_str); else { u32 i; - tipc_printf(pb, "UNKNOWN(%u)", a->media_id); + ret = tipc_snprintf(buf, len, "UNKNOWN(%u)", a->media_id); for (i = 0; i < sizeof(a->value); i++) - tipc_printf(pb, "-%02x", a->value[i]); + ret += tipc_snprintf(buf - ret, len + ret, + "-%02x", a->value[i]); } } /** * tipc_media_get_names - record names of registered media in buffer */ - struct sk_buff *tipc_media_get_names(void) { struct sk_buff *buf; @@ -166,24 +116,21 @@ struct sk_buff *tipc_media_get_names(void) if (!buf) return NULL; - read_lock_bh(&tipc_net_lock); - for (i = 0; i < media_count; i++) { + for (i = 0; media_info_array[i] != NULL; i++) { tipc_cfg_append_tlv(buf, TIPC_TLV_MEDIA_NAME, - media_list[i]->name, - strlen(media_list[i]->name) + 1); + media_info_array[i]->name, + strlen(media_info_array[i]->name) + 1); } - read_unlock_bh(&tipc_net_lock); return buf; } /** * bearer_name_validate - validate & (optionally) deconstruct bearer name - * @name - ptr to bearer name string - * @name_parts - ptr to area for bearer name components (or NULL if not needed) + * @name: ptr to bearer name string + * @name_parts: ptr to area for bearer name components (or NULL if not needed) * * Returns 1 if bearer name is valid, otherwise 0. */ - static int bearer_name_validate(const char *name, struct tipc_bearer_names *name_parts) { @@ -194,7 +141,6 @@ static int bearer_name_validate(const char *name, u32 if_len; /* copy bearer name & ensure length is OK */ - name_copy[TIPC_MAX_BEARER_NAME - 1] = 0; /* need above in case non-Posix strncpy() doesn't pad with nulls */ strncpy(name_copy, name, TIPC_MAX_BEARER_NAME); @@ -202,7 +148,6 @@ static int bearer_name_validate(const char *name, return 0; /* ensure all component parts of bearer name are present */ - media_name = name_copy; if_name = strchr(media_name, ':'); if (if_name == NULL) @@ -212,15 +157,11 @@ static int bearer_name_validate(const char *name, if_len = strlen(if_name) + 1; /* 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 */ - if (name_parts) { strcpy(name_parts->media_name, media_name); strcpy(name_parts->if_name, if_name); @@ -231,34 +172,14 @@ static int bearer_name_validate(const char *name, /** * tipc_bearer_find - locates bearer object with matching bearer name */ - struct tipc_bearer *tipc_bearer_find(const char *name) { struct tipc_bearer *b_ptr; u32 i; - for (i = 0, b_ptr = tipc_bearers; i < MAX_BEARERS; i++, b_ptr++) { - if (b_ptr->active && (!strcmp(b_ptr->name, name))) - return b_ptr; - } - return NULL; -} - -/** - * tipc_bearer_find_interface - locates bearer object with matching interface name - */ - -struct tipc_bearer *tipc_bearer_find_interface(const char *if_name) -{ - struct tipc_bearer *b_ptr; - char *b_if_name; - u32 i; - - for (i = 0, b_ptr = tipc_bearers; i < MAX_BEARERS; i++, b_ptr++) { - if (!b_ptr->active) - continue; - b_if_name = strchr(b_ptr->name, ':') + 1; - if (!strcmp(b_if_name, if_name)) + for (i = 0; i < MAX_BEARERS; i++) { + b_ptr = rtnl_dereference(bearer_list[i]); + if (b_ptr && (!strcmp(b_ptr->name, name))) return b_ptr; } return NULL; @@ -267,163 +188,60 @@ struct tipc_bearer *tipc_bearer_find_interface(const char *if_name) /** * tipc_bearer_get_names - record names of bearers in buffer */ - struct sk_buff *tipc_bearer_get_names(void) { struct sk_buff *buf; - struct tipc_bearer *b_ptr; + struct tipc_bearer *b; int i, j; buf = tipc_cfg_reply_alloc(MAX_BEARERS * TLV_SPACE(TIPC_MAX_BEARER_NAME)); if (!buf) return NULL; - read_lock_bh(&tipc_net_lock); - for (i = 0; i < media_count; i++) { + for (i = 0; media_info_array[i] != NULL; i++) { for (j = 0; j < MAX_BEARERS; j++) { - b_ptr = &tipc_bearers[j]; - if (b_ptr->active && (b_ptr->media == media_list[i])) { + b = rtnl_dereference(bearer_list[j]); + if (!b) + continue; + if (b->media == media_info_array[i]) { tipc_cfg_append_tlv(buf, TIPC_TLV_BEARER_NAME, - b_ptr->name, - strlen(b_ptr->name) + 1); + b->name, + strlen(b->name) + 1); } } } - read_unlock_bh(&tipc_net_lock); return buf; } -void tipc_bearer_add_dest(struct tipc_bearer *b_ptr, u32 dest) -{ - tipc_nmap_add(&b_ptr->nodes, dest); - tipc_bcbearer_sort(); - tipc_disc_add_dest(b_ptr->link_req); -} - -void tipc_bearer_remove_dest(struct tipc_bearer *b_ptr, u32 dest) -{ - tipc_nmap_remove(&b_ptr->nodes, dest); - tipc_bcbearer_sort(); - tipc_disc_remove_dest(b_ptr->link_req); -} - -/* - * 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) +void tipc_bearer_add_dest(u32 bearer_id, u32 dest) { - u32 res = 0; - struct tipc_link *ln, *tln; - - if (b_ptr->blocked) - return 0; + struct tipc_bearer *b_ptr; - 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); - } + rcu_read_lock(); + b_ptr = rcu_dereference_rtnl(bearer_list[bearer_id]); + if (b_ptr) { + tipc_bcbearer_sort(&b_ptr->nodes, dest, true); + tipc_disc_add_dest(b_ptr->link_req); } - 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: - * 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) -{ - list_move_tail(&l_ptr->link_list, &b_ptr->cong_links); + rcu_read_unlock(); } -/* - * 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 - */ - -void tipc_bearer_schedule(struct tipc_bearer *b_ptr, struct tipc_link *l_ptr) +void tipc_bearer_remove_dest(u32 bearer_id, u32 dest) { - spin_lock_bh(&b_ptr->lock); - tipc_bearer_schedule_unlocked(b_ptr, l_ptr); - spin_unlock_bh(&b_ptr->lock); -} - - -/* - * 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; + struct tipc_bearer *b_ptr; - 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; + rcu_read_lock(); + b_ptr = rcu_dereference_rtnl(bearer_list[bearer_id]); + if (b_ptr) { + tipc_bcbearer_sort(&b_ptr->nodes, dest, false); + tipc_disc_remove_dest(b_ptr->link_req); } - 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); + rcu_read_unlock(); } /** * tipc_enable_bearer - enable bearer with the given name */ - int tipc_enable_bearer(const char *name, u32 disc_domain, u32 priority) { struct tipc_bearer *b_ptr; @@ -435,13 +253,13 @@ int tipc_enable_bearer(const char *name, u32 disc_domain, u32 priority) u32 i; int res = -EINVAL; - if (tipc_mode != TIPC_NET_MODE) { - warn("Bearer <%s> rejected, not supported in standalone mode\n", - name); + if (!tipc_own_addr) { + pr_warn("Bearer <%s> rejected, not supported in standalone mode\n", + name); return -ENOPROTOOPT; } if (!bearer_name_validate(name, &b_names)) { - warn("Bearer <%s> rejected, illegal name\n", name); + pr_warn("Bearer <%s> rejected, illegal name\n", name); return -EINVAL; } if (tipc_addr_domain_valid(disc_domain) && @@ -449,27 +267,25 @@ int tipc_enable_bearer(const char *name, u32 disc_domain, u32 priority) if (tipc_in_scope(disc_domain, tipc_own_addr)) { disc_domain = tipc_own_addr & TIPC_CLUSTER_MASK; res = 0; /* accept any node in own cluster */ - } else if (in_own_cluster(disc_domain)) + } else if (in_own_cluster_exact(disc_domain)) res = 0; /* accept specified node in own cluster */ } if (res) { - warn("Bearer <%s> rejected, illegal discovery domain\n", name); + pr_warn("Bearer <%s> rejected, illegal discovery domain\n", + name); return -EINVAL; } - if ((priority < TIPC_MIN_LINK_PRI || - priority > TIPC_MAX_LINK_PRI) && + if ((priority > TIPC_MAX_LINK_PRI) && (priority != TIPC_MEDIA_LINK_PRI)) { - warn("Bearer <%s> rejected, illegal priority\n", name); + pr_warn("Bearer <%s> rejected, illegal priority\n", name); return -EINVAL; } - write_lock_bh(&tipc_net_lock); - m_ptr = tipc_media_find(b_names.media_name); if (!m_ptr) { - warn("Bearer <%s> rejected, media <%s> not registered\n", name, - b_names.media_name); - goto exit; + pr_warn("Bearer <%s> rejected, media <%s> not registered\n", + name, b_names.media_name); + return -EINVAL; } if (priority == TIPC_MEDIA_LINK_PRI) @@ -479,123 +295,104 @@ restart: bearer_id = MAX_BEARERS; with_this_prio = 1; for (i = MAX_BEARERS; i-- != 0; ) { - if (!tipc_bearers[i].active) { + b_ptr = rtnl_dereference(bearer_list[i]); + if (!b_ptr) { bearer_id = i; continue; } - if (!strcmp(name, tipc_bearers[i].name)) { - warn("Bearer <%s> rejected, already enabled\n", name); - goto exit; + if (!strcmp(name, b_ptr->name)) { + pr_warn("Bearer <%s> rejected, already enabled\n", + name); + return -EINVAL; } - if ((tipc_bearers[i].priority == priority) && + if ((b_ptr->priority == priority) && (++with_this_prio > 2)) { if (priority-- == 0) { - warn("Bearer <%s> rejected, duplicate priority\n", - name); - goto exit; + pr_warn("Bearer <%s> rejected, duplicate priority\n", + name); + return -EINVAL; } - warn("Bearer <%s> priority adjustment required %u->%u\n", - name, priority + 1, priority); + pr_warn("Bearer <%s> priority adjustment required %u->%u\n", + name, priority + 1, priority); goto restart; } } if (bearer_id >= MAX_BEARERS) { - warn("Bearer <%s> rejected, bearer limit reached (%u)\n", - name, MAX_BEARERS); - goto exit; + pr_warn("Bearer <%s> rejected, bearer limit reached (%u)\n", + name, MAX_BEARERS); + return -EINVAL; } - b_ptr = &tipc_bearers[bearer_id]; + b_ptr = kzalloc(sizeof(*b_ptr), GFP_ATOMIC); + if (!b_ptr) + return -ENOMEM; + strcpy(b_ptr->name, name); - res = m_ptr->enable_bearer(b_ptr); + b_ptr->media = m_ptr; + res = m_ptr->enable_media(b_ptr); if (res) { - warn("Bearer <%s> rejected, enable failure (%d)\n", name, -res); - goto exit; + pr_warn("Bearer <%s> rejected, enable failure (%d)\n", + name, -res); + return -EINVAL; } b_ptr->identity = bearer_id; - b_ptr->media = m_ptr; b_ptr->tolerance = m_ptr->tolerance; b_ptr->window = m_ptr->window; + b_ptr->domain = disc_domain; 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); if (res) { - bearer_disable(b_ptr); - warn("Bearer <%s> rejected, discovery object creation failed\n", - name); - goto exit; + bearer_disable(b_ptr, false); + pr_warn("Bearer <%s> rejected, discovery object creation failed\n", + name); + return -EINVAL; } - info("Enabled bearer <%s>, discovery domain %s, priority %u\n", - name, tipc_addr_string_fill(addr_string, disc_domain), priority); -exit: - write_unlock_bh(&tipc_net_lock); + + rcu_assign_pointer(bearer_list[bearer_id], b_ptr); + + pr_info("Enabled bearer <%s>, discovery domain %s, priority %u\n", + name, + tipc_addr_string_fill(addr_string, disc_domain), priority); return res; } /** - * tipc_block_bearer(): Block the bearer with the given name, - * and reset all its links + * tipc_reset_bearer - Reset all links established over this bearer */ - -int tipc_block_bearer(const char *name) +static int tipc_reset_bearer(struct tipc_bearer *b_ptr) { - struct tipc_bearer *b_ptr = NULL; - struct tipc_link *l_ptr; - struct tipc_link *temp_l_ptr; - - read_lock_bh(&tipc_net_lock); - b_ptr = tipc_bearer_find(name); - if (!b_ptr) { - warn("Attempt to block unknown bearer <%s>\n", name); - read_unlock_bh(&tipc_net_lock); - return -EINVAL; - } - - 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; - - spin_lock_bh(&n_ptr->lock); - tipc_link_reset(l_ptr); - spin_unlock_bh(&n_ptr->lock); - } - spin_unlock_bh(&b_ptr->lock); - read_unlock_bh(&tipc_net_lock); + pr_info("Resetting bearer <%s>\n", b_ptr->name); + tipc_link_reset_list(b_ptr->identity); + tipc_disc_reset(b_ptr); return 0; } /** - * bearer_disable - + * bearer_disable * - * Note: This routine assumes caller holds tipc_net_lock. + * Note: This routine assumes caller holds RTNL lock. */ - -static void bearer_disable(struct tipc_bearer *b_ptr) +static void bearer_disable(struct tipc_bearer *b_ptr, bool shutting_down) { - struct tipc_link *l_ptr; - struct tipc_link *temp_l_ptr; - - 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); - } + u32 i; + + pr_info("Disabling bearer <%s>\n", b_ptr->name); + b_ptr->media->disable_media(b_ptr); + + tipc_link_delete_list(b_ptr->identity, shutting_down); if (b_ptr->link_req) tipc_disc_delete(b_ptr->link_req); - spin_unlock_bh(&b_ptr->lock); - memset(b_ptr, 0, sizeof(struct tipc_bearer)); + + for (i = 0; i < MAX_BEARERS; i++) { + if (b_ptr == rtnl_dereference(bearer_list[i])) { + RCU_INIT_POINTER(bearer_list[i], NULL); + break; + } + } + kfree_rcu(b_ptr, rcu); } int tipc_disable_bearer(const char *name) @@ -603,28 +400,230 @@ int tipc_disable_bearer(const char *name) struct tipc_bearer *b_ptr; int res; - write_lock_bh(&tipc_net_lock); b_ptr = tipc_bearer_find(name); if (b_ptr == NULL) { - warn("Attempt to disable unknown bearer <%s>\n", name); + pr_warn("Attempt to disable unknown bearer <%s>\n", name); res = -EINVAL; } else { - bearer_disable(b_ptr); + bearer_disable(b_ptr, false); res = 0; } - write_unlock_bh(&tipc_net_lock); return res; } +int tipc_enable_l2_media(struct tipc_bearer *b) +{ + struct net_device *dev; + char *driver_name = strchr((const char *)b->name, ':') + 1; + + /* Find device with specified name */ + dev = dev_get_by_name(&init_net, driver_name); + if (!dev) + return -ENODEV; + + /* Associate TIPC bearer with L2 bearer */ + rcu_assign_pointer(b->media_ptr, dev); + memset(&b->bcast_addr, 0, sizeof(b->bcast_addr)); + memcpy(b->bcast_addr.value, dev->broadcast, b->media->hwaddr_len); + b->bcast_addr.media_id = b->media->type_id; + b->bcast_addr.broadcast = 1; + b->mtu = dev->mtu; + b->media->raw2addr(b, &b->addr, (char *)dev->dev_addr); + rcu_assign_pointer(dev->tipc_ptr, b); + return 0; +} +/* tipc_disable_l2_media - detach TIPC bearer from an L2 interface + * + * Mark L2 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.) + */ +void tipc_disable_l2_media(struct tipc_bearer *b) +{ + struct net_device *dev; + + dev = (struct net_device *)rtnl_dereference(b->media_ptr); + RCU_INIT_POINTER(b->media_ptr, NULL); + RCU_INIT_POINTER(dev->tipc_ptr, NULL); + synchronize_net(); + dev_put(dev); +} + +/** + * tipc_l2_send_msg - send a TIPC packet out over an L2 interface + * @buf: the packet to be sent + * @b_ptr: the bearer through which the packet is to be sent + * @dest: peer destination address + */ +int tipc_l2_send_msg(struct sk_buff *buf, struct tipc_bearer *b, + struct tipc_media_addr *dest) +{ + struct sk_buff *clone; + struct net_device *dev; + int delta; + + dev = (struct net_device *)rcu_dereference_rtnl(b->media_ptr); + if (!dev) + return 0; + + clone = skb_clone(buf, GFP_ATOMIC); + if (!clone) + return 0; + + 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; +} + +/* tipc_bearer_send- sends buffer to destination over bearer + * + * IMPORTANT: + * The media send routine must not alter the buffer being passed in + * as it may be needed for later retransmission! + */ +void tipc_bearer_send(u32 bearer_id, struct sk_buff *buf, + struct tipc_media_addr *dest) +{ + struct tipc_bearer *b_ptr; + + rcu_read_lock(); + b_ptr = rcu_dereference_rtnl(bearer_list[bearer_id]); + if (likely(b_ptr)) + b_ptr->media->send_msg(buf, b_ptr, dest); + rcu_read_unlock(); +} + +/** + * tipc_l2_rcv_msg - handle incoming TIPC message from an interface + * @buf: the received packet + * @dev: the net device that the packet was received on + * @pt: the packet_type structure which was used to register this handler + * @orig_dev: the original receive net device in case the device is a bond + * + * Accept only packets explicitly sent to this node, or broadcast packets; + * ignores packets sent using interface multicast, and traffic sent to other + * nodes (which can happen if interface is running in promiscuous mode). + */ +static int tipc_l2_rcv_msg(struct sk_buff *buf, struct net_device *dev, + struct packet_type *pt, struct net_device *orig_dev) +{ + struct tipc_bearer *b_ptr; + + if (!net_eq(dev_net(dev), &init_net)) { + kfree_skb(buf); + return NET_RX_DROP; + } + + rcu_read_lock(); + b_ptr = rcu_dereference_rtnl(dev->tipc_ptr); + if (likely(b_ptr)) { + if (likely(buf->pkt_type <= PACKET_BROADCAST)) { + buf->next = NULL; + tipc_rcv(buf, b_ptr); + rcu_read_unlock(); + return NET_RX_SUCCESS; + } + } + rcu_read_unlock(); + + kfree_skb(buf); + return NET_RX_DROP; +} + +/** + * tipc_l2_device_event - handle device events from network device + * @nb: the context of the notification + * @evt: the type of event + * @ptr: the net device that the event was on + * + * This function is called by the Ethernet driver in case of link + * change event. + */ +static int tipc_l2_device_event(struct notifier_block *nb, unsigned long evt, + void *ptr) +{ + struct tipc_bearer *b_ptr; + struct net_device *dev = netdev_notifier_info_to_dev(ptr); + + if (!net_eq(dev_net(dev), &init_net)) + return NOTIFY_DONE; + + b_ptr = rtnl_dereference(dev->tipc_ptr); + if (!b_ptr) + return NOTIFY_DONE; + + b_ptr->mtu = dev->mtu; + + switch (evt) { + case NETDEV_CHANGE: + if (netif_carrier_ok(dev)) + break; + case NETDEV_DOWN: + case NETDEV_CHANGEMTU: + tipc_reset_bearer(b_ptr); + break; + case NETDEV_CHANGEADDR: + b_ptr->media->raw2addr(b_ptr, &b_ptr->addr, + (char *)dev->dev_addr); + tipc_reset_bearer(b_ptr); + break; + case NETDEV_UNREGISTER: + case NETDEV_CHANGENAME: + bearer_disable(b_ptr, false); + break; + } + return NOTIFY_OK; +} + +static struct packet_type tipc_packet_type __read_mostly = { + .type = htons(ETH_P_TIPC), + .func = tipc_l2_rcv_msg, +}; + +static struct notifier_block notifier = { + .notifier_call = tipc_l2_device_event, + .priority = 0, +}; + +int tipc_bearer_setup(void) +{ + int err; + + err = register_netdevice_notifier(¬ifier); + if (err) + return err; + dev_add_pack(&tipc_packet_type); + return 0; +} + +void tipc_bearer_cleanup(void) +{ + unregister_netdevice_notifier(¬ifier); + dev_remove_pack(&tipc_packet_type); +} void tipc_bearer_stop(void) { + struct tipc_bearer *b_ptr; u32 i; for (i = 0; i < MAX_BEARERS; i++) { - if (tipc_bearers[i].active) - bearer_disable(&tipc_bearers[i]); + b_ptr = rtnl_dereference(bearer_list[i]); + if (b_ptr) { + bearer_disable(b_ptr, true); + bearer_list[i] = NULL; + } } - media_count = 0; } diff --git a/net/tipc/bearer.h b/net/tipc/bearer.h index d3eac56b8c2..78fccc49de2 100644 --- a/net/tipc/bearer.h +++ b/net/tipc/bearer.h @@ -1,7 +1,7 @@ /* * net/tipc/bearer.h: Include file for TIPC bearer code * - * Copyright (c) 1996-2006, Ericsson AB + * Copyright (c) 1996-2006, 2013, Ericsson AB * Copyright (c) 2005, 2010-2011, Wind River Systems * All rights reserved. * @@ -42,29 +42,26 @@ #define MAX_BEARERS 2 #define MAX_MEDIA 2 -/* - * Identifiers associated with TIPC message header media address info - * - * - address info field is 20 bytes long - * - media type identifier located at offset 3 - * - remaining bytes vary according to media type +/* Identifiers associated with TIPC message header media address info + * - address info field is 32 bytes long + * - the field's actual content and length is defined per media + * - remaining unused bytes in the field are set to zero */ - -#define TIPC_MEDIA_ADDR_SIZE 20 +#define TIPC_MEDIA_ADDR_SIZE 32 #define TIPC_MEDIA_TYPE_OFFSET 3 /* * 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 * @value: address info (format defined by media) * @media_id: TIPC media type identifier * @broadcast: non-zero if address is a broadcast address */ - struct tipc_media_addr { u8 value[TIPC_MEDIA_ADDR_SIZE]; u8 media_id; @@ -74,57 +71,60 @@ struct tipc_media_addr { struct tipc_bearer; /** - * struct tipc_media - TIPC media information available to internal users + * struct tipc_media - Media specific info exposed to generic bearer layer * @send_msg: routine which handles buffer transmission - * @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 + * @enable_media: routine which enables a media + * @disable_media: routine which disables a media + * @addr2str: convert media address format to string + * @addr2msg: convert from media addr format to discovery msg addr format + * @msg2addr: convert from discovery msg addr format to media addr format + * @raw2addr: convert from raw addr format to media addr format * @priority: default link (and bearer) priority * @tolerance: default time (in ms) before declaring link failure * @window: default window (in packets) before declaring link congestion * @type_id: TIPC media identifier + * @hwaddr_len: TIPC media address len * @name: media name */ - struct tipc_media { int (*send_msg)(struct sk_buff *buf, struct tipc_bearer *b_ptr, struct tipc_media_addr *dest); - 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 (*enable_media)(struct tipc_bearer *b_ptr); + void (*disable_media)(struct tipc_bearer *b_ptr); + int (*addr2str)(struct tipc_media_addr *addr, + char *strbuf, + int bufsz); + int (*addr2msg)(char *msg, struct tipc_media_addr *addr); + int (*msg2addr)(struct tipc_bearer *b, + struct tipc_media_addr *addr, + char *msg); + int (*raw2addr)(struct tipc_bearer *b, + struct tipc_media_addr *addr, + char *raw); u32 priority; u32 tolerance; u32 window; u32 type_id; + u32 hwaddr_len; char name[TIPC_MAX_MEDIA_NAME]; }; /** - * struct tipc_bearer - TIPC bearer structure - * @usr_handle: pointer to additional media-specific information about bearer + * struct tipc_bearer - Generic TIPC bearer structure + * @media_ptr: pointer to additional media-specific information about bearer * @mtu: max packet size bearer can support - * @blocked: non-zero if bearer is blocked - * @lock: spinlock for controlling access to bearer * @addr: media-specific address associated with bearer * @name: bearer name (format = media:interface) * @media: ptr to media structure associated with bearer + * @bcast_addr: media address used in broadcasting + * @rcu: rcu struct for tipc_bearer * @priority: default link priority for bearer * @window: default window size for bearer * @tolerance: default link tolerance for bearer + * @domain: network domain to which links can be established * @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 * @@ -133,21 +133,19 @@ struct tipc_media { * care of initializing all other fields. */ struct tipc_bearer { - void *usr_handle; /* initalized by media */ + void __rcu *media_ptr; /* initalized by media */ u32 mtu; /* initalized by media */ - int blocked; /* initalized by media */ struct tipc_media_addr addr; /* initalized by media */ char name[TIPC_MAX_BEARER_NAME]; - spinlock_t lock; struct tipc_media *media; + struct tipc_media_addr bcast_addr; + struct rcu_head rcu; u32 priority; u32 window; u32 tolerance; + u32 domain; 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; }; @@ -159,69 +157,43 @@ struct tipc_bearer_names { struct tipc_link; -extern struct tipc_bearer tipc_bearers[]; +extern struct tipc_bearer __rcu *bearer_list[]; /* * TIPC routines available to supported media types */ -int tipc_register_media(struct tipc_media *m_ptr); - -void tipc_recv_msg(struct sk_buff *buf, struct tipc_bearer *tb_ptr); - -int tipc_block_bearer(const char *name); -void tipc_continue(struct tipc_bearer *tb_ptr); +void tipc_rcv(struct sk_buff *buf, struct tipc_bearer *tb_ptr); int tipc_enable_bearer(const char *bearer_name, u32 disc_domain, u32 priority); int tipc_disable_bearer(const char *name); /* * Routines made available to TIPC by supported media types */ -int tipc_eth_media_start(void); -void tipc_eth_media_stop(void); +extern struct tipc_media eth_media_info; + +#ifdef CONFIG_TIPC_MEDIA_IB +extern struct tipc_media ib_media_info; +#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(struct print_buf *pb, struct tipc_media_addr *a); +void tipc_media_addr_printf(char *buf, int len, struct tipc_media_addr *a); struct sk_buff *tipc_media_get_names(void); +int tipc_enable_l2_media(struct tipc_bearer *b); +void tipc_disable_l2_media(struct tipc_bearer *b); +int tipc_l2_send_msg(struct sk_buff *buf, struct tipc_bearer *b, + struct tipc_media_addr *dest); 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); +void tipc_bearer_add_dest(u32 bearer_id, u32 dest); +void tipc_bearer_remove_dest(u32 bearer_id, u32 dest); 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_setup(void); +void tipc_bearer_cleanup(void); 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, - struct tipc_media_addr *dest) -{ - return !b_ptr->media->send_msg(buf, b_ptr, dest); -} +void tipc_bearer_send(u32 bearer_id, struct sk_buff *buf, + struct tipc_media_addr *dest); #endif /* _TIPC_BEARER_H */ diff --git a/net/tipc/config.c b/net/tipc/config.c index 4785bf26cdf..2b42403ad33 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 @@ -38,16 +38,14 @@ #include "port.h" #include "name_table.h" #include "config.h" +#include "server.h" -static u32 config_port_ref; - -static DEFINE_SPINLOCK(config_lock); +#define REPLY_TRUNCATED "<truncated>\n" static const void *req_tlv_area; /* request message TLV area */ static int req_tlv_space; /* request message TLV area size */ static int rep_headroom; /* reply message headroom to use */ - struct sk_buff *tipc_cfg_reply_alloc(int payload_size) { struct sk_buff *buf; @@ -104,13 +102,12 @@ struct sk_buff *tipc_cfg_reply_string_type(u16 tlv_type, char *string) return buf; } -#define MAX_STATS_INFO 2000 - static struct sk_buff *tipc_show_stats(void) { struct sk_buff *buf; struct tlv_desc *rep_tlv; - struct print_buf pb; + char *pb; + int pb_len; int str_len; u32 value; @@ -121,18 +118,16 @@ static struct sk_buff *tipc_show_stats(void) if (value != 0) return tipc_cfg_reply_error_string("unsupported argument"); - buf = tipc_cfg_reply_alloc(TLV_SPACE(MAX_STATS_INFO)); + buf = tipc_cfg_reply_alloc(TLV_SPACE(ULTRA_STRING_MAX_LEN)); if (buf == NULL) return NULL; rep_tlv = (struct tlv_desc *)buf->data; - tipc_printbuf_init(&pb, (char *)TLV_DATA(rep_tlv), MAX_STATS_INFO); + pb = TLV_DATA(rep_tlv); + pb_len = ULTRA_STRING_MAX_LEN; - tipc_printf(&pb, "TIPC version " TIPC_MOD_VER "\n"); - - /* Use additional tipc_printf()'s to return more info ... */ - - str_len = tipc_printbuf_validate(&pb); + str_len = tipc_snprintf(pb, pb_len, "TIPC version " TIPC_MOD_VER "\n"); + str_len += 1; /* for "\0" */ skb_put(buf, TLV_SPACE(str_len)); TLV_SET(rep_tlv, TIPC_TLV_ULTRA_STRING, NULL, str_len); @@ -179,65 +174,13 @@ static struct sk_buff *cfg_set_own_addr(void) if (!tipc_addr_node_valid(addr)) return tipc_cfg_reply_error_string(TIPC_CFG_INVALID_VALUE " (node address)"); - if (tipc_mode == TIPC_NET_MODE) + if (tipc_own_addr) return tipc_cfg_reply_error_string(TIPC_CFG_NOT_SUPPORTED " (cannot change node address once assigned)"); + if (!tipc_net_start(addr)) + return tipc_cfg_reply_none(); - /* - * 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(); -} - -static struct sk_buff *cfg_set_remote_mng(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)); - tipc_remote_management = (value != 0); - 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 != delimit(value, 1, 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 != delimit(value, 1, 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(); + return tipc_cfg_reply_error_string("cannot change to network mode"); } static struct sk_buff *cfg_set_max_ports(void) @@ -249,14 +192,11 @@ static struct sk_buff *cfg_set_max_ports(void) value = ntohl(*(__be32 *)TLV_DATA(req_tlv_area)); if (value == tipc_max_ports) return tipc_cfg_reply_none(); - if (value != delimit(value, 127, 65535)) + if (value < 127 || value > 65535) return tipc_cfg_reply_error_string(TIPC_CFG_INVALID_VALUE " (max ports must be 127-65535)"); - if (tipc_mode != TIPC_NOT_RUNNING) - return tipc_cfg_reply_error_string(TIPC_CFG_NOT_SUPPORTED - " (cannot change max ports while TIPC is active)"); - tipc_max_ports = value; - return tipc_cfg_reply_none(); + return tipc_cfg_reply_error_string(TIPC_CFG_NOT_SUPPORTED + " (cannot change max ports while TIPC is active)"); } static struct sk_buff *cfg_set_netid(void) @@ -268,10 +208,10 @@ static struct sk_buff *cfg_set_netid(void) value = ntohl(*(__be32 *)TLV_DATA(req_tlv_area)); if (value == tipc_net_id) return tipc_cfg_reply_none(); - if (value != delimit(value, 1, 9999)) + if (value < 1 || value > 9999) return tipc_cfg_reply_error_string(TIPC_CFG_INVALID_VALUE " (network id must be 1-9999)"); - if (tipc_mode == TIPC_NET_MODE) + if (tipc_own_addr) return tipc_cfg_reply_error_string(TIPC_CFG_NOT_SUPPORTED " (cannot change network id once TIPC has joined a network)"); tipc_net_id = value; @@ -283,37 +223,23 @@ 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); + rtnl_lock(); /* Save request and reply details in a well-known location */ - req_tlv_area = request_area; req_tlv_space = request_space; rep_headroom = reply_headroom; /* Check command authorization */ - - if (likely(orig_node == tipc_own_addr)) { + if (likely(in_own_node(orig_node))) { /* command is permitted */ - } else if (cmd >= 0x8000) { + } else { rep_tlv_buf = tipc_cfg_reply_error_string(TIPC_CFG_NOT_SUPPORTED " (cannot be done remotely)"); goto exit; - } else if (!tipc_remote_management) { - rep_tlv_buf = tipc_cfg_reply_error_string(TIPC_CFG_NO_REMOTE); - goto exit; - } else if (cmd >= 0x4000) { - u32 domain = 0; - - if ((tipc_nametbl_translate(TIPC_ZM_SRV, 0, &domain) == 0) || - (domain != orig_node)) { - rep_tlv_buf = tipc_cfg_reply_error_string(TIPC_CFG_NOT_ZONE_MSTR); - goto exit; - } } /* Call appropriate processing routine */ - switch (cmd) { case TIPC_CMD_NOOP: rep_tlv_buf = tipc_cfg_reply_none(); @@ -342,12 +268,6 @@ struct sk_buff *tipc_cfg_do_cmd(u32 orig_node, u16 cmd, const void *request_area case TIPC_CMD_SHOW_PORTS: rep_tlv_buf = tipc_port_get_ports(); break; - case TIPC_CMD_SET_LOG_SIZE: - rep_tlv_buf = tipc_log_resize_cmd(req_tlv_area, req_tlv_space); - break; - case TIPC_CMD_DUMP_LOG: - rep_tlv_buf = tipc_log_dump(); - break; case TIPC_CMD_SHOW_STATS: rep_tlv_buf = tipc_show_stats(); break; @@ -365,33 +285,15 @@ struct sk_buff *tipc_cfg_do_cmd(u32 orig_node, u16 cmd, const void *request_area case TIPC_CMD_SET_NODE_ADDR: rep_tlv_buf = cfg_set_own_addr(); break; - case TIPC_CMD_SET_REMOTE_MNG: - rep_tlv_buf = cfg_set_remote_mng(); - break; 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; - case TIPC_CMD_GET_REMOTE_MNG: - rep_tlv_buf = tipc_cfg_reply_unsigned(tipc_remote_management); - break; 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; @@ -407,6 +309,14 @@ 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_SET_REMOTE_MNG: + case TIPC_CMD_GET_REMOTE_MNG: + case TIPC_CMD_DUMP_LOG: rep_tlv_buf = tipc_cfg_reply_error_string(TIPC_CFG_NOT_SUPPORTED " (obsolete command)"); break; @@ -416,86 +326,17 @@ struct sk_buff *tipc_cfg_do_cmd(u32 orig_node, u16 cmd, const void *request_area break; } - /* Return reply buffer */ -exit: - spin_unlock_bh(&config_lock); - 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) -{ - struct tipc_cfg_msg_hdr *req_hdr; - struct tipc_cfg_msg_hdr *rep_hdr; - struct sk_buff *rep_buf; - - /* 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))) || - (ntohs(req_hdr->tcm_flags) != TCM_F_REQUEST)) { - warn("Invalid configuration message discarded\n"); - return; - } + WARN_ON(rep_tlv_buf->len > TLV_SPACE(ULTRA_STRING_MAX_LEN)); - /* 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), - BUF_HEADROOM + MAX_H_SIZE + sizeof(*rep_hdr)); - if (rep_buf) { - skb_push(rep_buf, sizeof(*rep_hdr)); - rep_hdr = (struct tipc_cfg_msg_hdr *)rep_buf->data; - 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; + /* Append an error message if we cannot return all requested data */ + if (rep_tlv_buf->len == TLV_SPACE(ULTRA_STRING_MAX_LEN)) { + if (*(rep_tlv_buf->data + ULTRA_STRING_MAX_LEN) != '\0') + sprintf(rep_tlv_buf->data + rep_tlv_buf->len - + sizeof(REPLY_TRUNCATED) - 1, REPLY_TRUNCATED); } - /* NEED TO ADD CODE TO HANDLE FAILED SEND (SUCH AS CONGESTION) */ - tipc_send_buf2port(port_ref, orig, rep_buf, rep_buf->len); -} - -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_nametbl_publish_rsv(config_port_ref, TIPC_ZONE_SCOPE, &seq); - if (res) - goto failed; - - return 0; - -failed: - err("Unable to create configuration service\n"); - return res; -} - -void tipc_cfg_stop(void) -{ - if (config_port_ref) { - tipc_deleteport(config_port_ref); - config_port_ref = 0; - } + /* Return reply buffer */ +exit: + rtnl_unlock(); + return rep_tlv_buf; } diff --git a/net/tipc/config.h b/net/tipc/config.h index 80da6ebc278..47b1bf18161 100644 --- a/net/tipc/config.h +++ b/net/tipc/config.h @@ -64,8 +64,4 @@ static inline struct sk_buff *tipc_cfg_reply_ultra_string(char *string) struct sk_buff *tipc_cfg_do_cmd(u32 orig_node, u16 cmd, const void *req_tlv_area, int req_tlv_space, int headroom); - -int tipc_cfg_init(void); -void tipc_cfg_stop(void); - #endif diff --git a/net/tipc/core.c b/net/tipc/core.c index 2691cd57b8a..676d18015dd 100644 --- a/net/tipc/core.c +++ b/net/tipc/core.c @@ -1,8 +1,8 @@ /* * net/tipc/core.c: TIPC module code * - * Copyright (c) 2003-2006, Ericsson AB - * Copyright (c) 2005-2006, 2010-2011, Wind River Systems + * Copyright (c) 2003-2006, 2013, Ericsson AB + * Copyright (c) 2005-2006, 2010-2013, Wind River Systems * All rights reserved. * * Redistribution and use in source and binary forms, with or without @@ -34,40 +34,23 @@ * POSSIBILITY OF SUCH DAMAGE. */ -#include <linux/module.h> - #include "core.h" #include "ref.h" #include "name_table.h" #include "subscr.h" #include "config.h" +#include "port.h" - -#ifndef CONFIG_TIPC_PORTS -#define CONFIG_TIPC_PORTS 8191 -#endif - -#ifndef CONFIG_TIPC_LOG -#define CONFIG_TIPC_LOG 0 -#endif +#include <linux/module.h> /* global variables used by multiple sub-systems within TIPC */ - -int tipc_mode = TIPC_NOT_RUNNING; -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 sysctl_tipc_rmem[3] __read_mostly; /* min/default/max */ /** * tipc_buf_acquire - creates a TIPC message buffer @@ -78,7 +61,6 @@ int tipc_remote_management; * NOTE: Headroom is reserved to allow prepending of a data link header. * There may also be unrequested tailroom present at the buffer's end. */ - struct sk_buff *tipc_buf_acquire(u32 size) { struct sk_buff *skb; @@ -94,115 +76,102 @@ struct sk_buff *tipc_buf_acquire(u32 size) } /** - * tipc_core_stop_net - shut down TIPC networking sub-systems - */ - -static void tipc_core_stop_net(void) -{ - tipc_net_stop(); - tipc_eth_media_stop(); -} - -/** - * start_net - start TIPC networking sub-systems - */ - -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(); - return res; -} - -/** * tipc_core_stop - switch TIPC from SINGLE NODE to NOT RUNNING mode */ - static void tipc_core_stop(void) { - if (tipc_mode != TIPC_NODE_MODE) - return; - - tipc_mode = TIPC_NOT_RUNNING; - + tipc_net_stop(); + tipc_bearer_cleanup(); tipc_netlink_stop(); - tipc_handler_stop(); - tipc_cfg_stop(); tipc_subscr_stop(); tipc_nametbl_stop(); tipc_ref_table_stop(); tipc_socket_stop(); - tipc_log_resize(0); + tipc_unregister_sysctl(); } /** * tipc_core_start - switch TIPC from NOT RUNNING to SINGLE NODE mode */ - static int tipc_core_start(void) { - int res; - - if (tipc_mode != TIPC_NOT_RUNNING) - return -ENOPROTOOPT; + int err; get_random_bytes(&tipc_random, sizeof(tipc_random)); - tipc_mode = TIPC_NODE_MODE; - - res = tipc_handler_start(); - if (!res) - res = tipc_ref_table_init(tipc_max_ports, tipc_random); - if (!res) - res = tipc_nametbl_init(); - if (!res) - res = tipc_k_signal((Handler)tipc_subscr_start, 0); - if (!res) - res = tipc_k_signal((Handler)tipc_cfg_init, 0); - if (!res) - res = tipc_netlink_start(); - if (!res) - res = tipc_socket_init(); - if (res) - tipc_core_stop(); - return res; -} + err = tipc_ref_table_init(tipc_max_ports, tipc_random); + if (err) + goto out_reftbl; + + err = tipc_nametbl_init(); + if (err) + goto out_nametbl; + + err = tipc_netlink_start(); + if (err) + goto out_netlink; + err = tipc_socket_init(); + if (err) + goto out_socket; + + err = tipc_register_sysctl(); + if (err) + goto out_sysctl; + + err = tipc_subscr_start(); + if (err) + goto out_subscr; + + err = tipc_bearer_setup(); + if (err) + goto out_bearer; + + return 0; +out_bearer: + tipc_subscr_stop(); +out_subscr: + tipc_unregister_sysctl(); +out_sysctl: + tipc_socket_stop(); +out_socket: + tipc_netlink_stop(); +out_netlink: + tipc_nametbl_stop(); +out_nametbl: + tipc_ref_table_stop(); +out_reftbl: + return err; +} static int __init tipc_init(void) { int res; - if (tipc_log_resize(CONFIG_TIPC_LOG) != 0) - warn("Unable to create log buffer\n"); - - info("Activated (version " TIPC_MOD_VER ")\n"); + pr_info("Activated (version " TIPC_MOD_VER ")\n"); 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] = TIPC_CONN_OVERLOAD_LIMIT >> 4 << + TIPC_LOW_IMPORTANCE; + sysctl_tipc_rmem[1] = TIPC_CONN_OVERLOAD_LIMIT >> 4 << + TIPC_CRITICAL_IMPORTANCE; + sysctl_tipc_rmem[2] = TIPC_CONN_OVERLOAD_LIMIT; + res = tipc_core_start(); if (res) - err("Unable to start in single node mode\n"); + pr_err("Unable to start in single node mode\n"); else - info("Started in single node mode\n"); + pr_info("Started in single node mode\n"); return res; } static void __exit tipc_exit(void) { - tipc_core_stop_net(); tipc_core_stop(); - info("Deactivated\n"); + pr_info("Deactivated\n"); } module_init(tipc_init); diff --git a/net/tipc/core.h b/net/tipc/core.h index 2761af36d14..bb26ed1ee96 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 @@ -37,6 +37,8 @@ #ifndef _TIPC_CORE_H #define _TIPC_CORE_H +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + #include <linux/tipc.h> #include <linux/tipc_config.h> #include <linux/types.h> @@ -45,7 +47,7 @@ #include <linux/mm.h> #include <linux/timer.h> #include <linux/string.h> -#include <asm/uaccess.h> +#include <linux/uaccess.h> #include <linux/interrupt.h> #include <linux/atomic.h> #include <asm/hardirq.h> @@ -54,138 +56,62 @@ #include <linux/list.h> #include <linux/slab.h> #include <linux/vmalloc.h> - +#include <linux/rtnetlink.h> +#include <linux/etherdevice.h> #define TIPC_MOD_VER "2.0.0" -struct tipc_msg; /* msg.h */ -struct print_buf; /* log.h */ - -/* - * TIPC system monitoring code - */ - -/* - * TIPC's print buffer subsystem supports the following print buffers: - * - * TIPC_NULL : null buffer (i.e. print nowhere) - * TIPC_CONS : system console - * TIPC_LOG : TIPC log buffer - * &buf : user-defined buffer (struct print_buf *) - * - * Note: TIPC_LOG is configured to echo its output to the system console; - * user-defined buffers can be configured to do the same thing. - */ -extern struct print_buf *const TIPC_NULL; -extern struct print_buf *const TIPC_CONS; -extern struct print_buf *const TIPC_LOG; - -void tipc_printf(struct print_buf *, const char *fmt, ...); - -/* - * TIPC_OUTPUT is the destination print buffer for system messages. - */ - -#ifndef TIPC_OUTPUT -#define TIPC_OUTPUT TIPC_LOG -#endif - -#define err(fmt, arg...) tipc_printf(TIPC_OUTPUT, \ - KERN_ERR "TIPC: " fmt, ## arg) -#define warn(fmt, arg...) tipc_printf(TIPC_OUTPUT, \ - KERN_WARNING "TIPC: " fmt, ## arg) -#define info(fmt, arg...) tipc_printf(TIPC_OUTPUT, \ - KERN_NOTICE "TIPC: " fmt, ## arg) +#define ULTRA_STRING_MAX_LEN 32768 +#define TIPC_MAX_SUBSCRIPTIONS 65535 +#define TIPC_MAX_PUBLICATIONS 65535 -#ifdef CONFIG_TIPC_DEBUG - -/* - * DBG_OUTPUT is the destination print buffer for debug messages. - */ - -#ifndef DBG_OUTPUT -#define DBG_OUTPUT TIPC_LOG -#endif - -#define dbg(fmt, arg...) tipc_printf(DBG_OUTPUT, KERN_DEBUG fmt, ## arg); - -#define msg_dbg(msg, txt) tipc_msg_dbg(DBG_OUTPUT, msg, txt); - -void tipc_msg_dbg(struct print_buf *, struct tipc_msg *, const char *); - -#else - -#define dbg(fmt, arg...) do {} while (0) -#define msg_dbg(msg, txt) do {} while (0) - -#define tipc_msg_dbg(buf, msg, txt) do {} while (0) - -#endif +struct tipc_msg; /* msg.h */ +int tipc_snprintf(char *buf, int len, const char *fmt, ...); /* * TIPC-specific error codes */ - #define ELINKCONG EAGAIN /* link congestion <=> resource unavailable */ /* - * TIPC operating mode routines - */ -#define TIPC_NOT_RUNNING 0 -#define TIPC_NODE_MODE 1 -#define TIPC_NET_MODE 2 - -/* * 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 sysctl_tipc_rmem[3] __read_mostly; /* * Other global variables */ - -extern int tipc_mode; -extern int tipc_random; -extern const char tipc_alphabet[]; - +extern int tipc_random __read_mostly; /* * Routines available to privileged subsystems */ - -extern int tipc_core_start_net(unsigned long); -extern int tipc_handler_start(void); -extern void tipc_handler_stop(void); -extern int tipc_netlink_start(void); -extern void tipc_netlink_stop(void); -extern int tipc_socket_init(void); -extern void tipc_socket_stop(void); - -static inline int delimit(int val, int min, int max) -{ - if (val > max) - return max; - if (val < min) - return min; - return val; -} - +int tipc_netlink_start(void); +void tipc_netlink_stop(void); +int tipc_socket_init(void); +void tipc_socket_stop(void); +int tipc_sock_create_local(int type, struct socket **res); +void tipc_sock_release_local(struct socket *sock); +int tipc_sock_accept_local(struct socket *sock, struct socket **newsock, + int flags); + +#ifdef CONFIG_SYSCTL +int tipc_register_sysctl(void); +void tipc_unregister_sysctl(void); +#else +#define tipc_register_sysctl() 0 +#define tipc_unregister_sysctl() +#endif /* - * TIPC timer and signal code + * TIPC timer code */ - typedef void (*Handler) (unsigned long); -u32 tipc_k_signal(Handler routine, unsigned long argument); - /** * k_init_timer - initialize a timer * @timer: pointer to timer structure @@ -194,7 +120,6 @@ u32 tipc_k_signal(Handler routine, unsigned long argument); * * Timer must be initialized before use (and terminated when no longer needed). */ - static inline void k_init_timer(struct timer_list *timer, Handler routine, unsigned long argument) { @@ -214,7 +139,6 @@ static inline void k_init_timer(struct timer_list *timer, Handler routine, * then an additional jiffy is added to account for the fact that * the starting time may be in the middle of the current jiffy. */ - static inline void k_start_timer(struct timer_list *timer, unsigned long msec) { mod_timer(timer, jiffies + msecs_to_jiffies(msec) + 1); @@ -230,7 +154,6 @@ static inline void k_start_timer(struct timer_list *timer, unsigned long msec) * WARNING: Must not be called when holding locks required by the timer's * timeout routine, otherwise deadlock can occur on SMP systems! */ - static inline void k_cancel_timer(struct timer_list *timer) { del_timer_sync(timer); @@ -247,12 +170,10 @@ static inline void k_cancel_timer(struct timer_list *timer) * (Do not "enhance" this routine to automatically cancel an active timer, * otherwise deadlock can arise when a timeout routine calls k_term_timer.) */ - static inline void k_term_timer(struct timer_list *timer) { } - /* * TIPC message buffer code * @@ -262,45 +183,21 @@ static inline void k_term_timer(struct timer_list *timer) * Note: Headroom should be a multiple of 4 to ensure the TIPC header fields * are word aligned for quicker access */ - #define BUF_HEADROOM LL_MAX_HEADER struct tipc_skb_cb { void *handle; + bool deferred; + struct sk_buff *tail; }; #define TIPC_SKB_CB(__skb) ((struct tipc_skb_cb *)&((__skb)->cb[0])) - static inline struct tipc_msg *buf_msg(struct sk_buff *skb) { return (struct tipc_msg *)skb->data; } -extern struct sk_buff *tipc_buf_acquire(u32 size); - -/** - * buf_discard - frees a TIPC message buffer - * @skb: message buffer - * - * Frees a message buffer. If passed NULL, just returns. - */ - -static inline void buf_discard(struct sk_buff *skb) -{ - kfree_skb(skb); -} - -/** - * buf_linearize - convert a TIPC message buffer into a single contiguous piece - * @skb: message buffer - * - * Returns 0 on success. - */ - -static inline int buf_linearize(struct sk_buff *skb) -{ - return skb_linearize(skb); -} +struct sk_buff *tipc_buf_acquire(u32 size); #endif diff --git a/net/tipc/discover.c b/net/tipc/discover.c index a00e5f81156..aa722a42ef8 100644 --- a/net/tipc/discover.c +++ b/net/tipc/discover.c @@ -1,7 +1,7 @@ /* * net/tipc/discover.c * - * Copyright (c) 2003-2006, Ericsson AB + * Copyright (c) 2003-2006, 2014, Ericsson AB * Copyright (c) 2005-2006, 2010-2011, Wind River Systems * All rights reserved. * @@ -46,19 +46,21 @@ /** * struct tipc_link_req - information about an ongoing link setup request - * @bearer: bearer issuing requests + * @bearer_id: identity of bearer issuing requests * @dest: destination address for request messages * @domain: network domain to which links can be established * @num_nodes: number of nodes currently discovered (i.e. with an active link) + * @lock: spinlock for controlling access to requests * @buf: request message to be (repeatedly) sent * @timer: timer governing period between requests * @timer_intv: current interval between requests (in ms) */ struct tipc_link_req { - struct tipc_bearer *bearer; + u32 bearer_id; struct tipc_media_addr dest; u32 domain; int num_nodes; + spinlock_t lock; struct sk_buff *buf; struct timer_list timer; unsigned int timer_intv; @@ -67,26 +69,21 @@ struct tipc_link_req { /** * tipc_disc_init_msg - initialize a link setup message * @type: message type (request or response) - * @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, - struct tipc_bearer *b_ptr) +static void tipc_disc_init_msg(struct sk_buff *buf, u32 type, + struct tipc_bearer *b_ptr) { - struct sk_buff *buf = tipc_buf_acquire(INT_H_SIZE); struct tipc_msg *msg; - - if (buf) { - msg = buf_msg(buf); - tipc_msg_init(msg, LINK_CONFIG, type, INT_H_SIZE, dest_domain); - msg_set_non_seq(msg, 1); - msg_set_dest_domain(msg, dest_domain); - msg_set_bc_netid(msg, tipc_net_id); - b_ptr->media->addr2msg(&b_ptr->addr, msg_media_addr(msg)); - } - return buf; + u32 dest_domain = b_ptr->domain; + + msg = buf_msg(buf); + tipc_msg_init(msg, LINK_CONFIG, type, INT_H_SIZE, dest_domain); + msg_set_non_seq(msg, 1); + msg_set_node_sig(msg, tipc_random); + msg_set_dest_domain(msg, dest_domain); + msg_set_bc_netid(msg, tipc_net_id); + b_ptr->media->addr2msg(msg_media_addr(msg), &b_ptr->addr); } /** @@ -95,119 +92,164 @@ static struct sk_buff *tipc_disc_init_msg(u32 type, * @node_addr: duplicated node address * @media_addr: media address advertised by duplicated node */ - static void disc_dupl_alert(struct tipc_bearer *b_ptr, u32 node_addr, struct tipc_media_addr *media_addr) { char node_addr_str[16]; char media_addr_str[64]; - struct print_buf pb; tipc_addr_string_fill(node_addr_str, node_addr); - tipc_printbuf_init(&pb, media_addr_str, sizeof(media_addr_str)); - tipc_media_addr_printf(&pb, media_addr); - tipc_printbuf_validate(&pb); - warn("Duplicate %s using %s seen on <%s>\n", - node_addr_str, media_addr_str, b_ptr->name); + tipc_media_addr_printf(media_addr_str, sizeof(media_addr_str), + media_addr); + pr_warn("Duplicate %s using %s seen on <%s>\n", node_addr_str, + media_addr_str, b_ptr->name); } /** - * tipc_disc_recv_msg - handle incoming link setup message (request or response) + * tipc_disc_rcv - handle incoming discovery message (request or response) * @buf: buffer containing message - * @b_ptr: bearer that message arrived on + * @bearer: bearer that message arrived on */ - -void tipc_disc_recv_msg(struct sk_buff *buf, struct tipc_bearer *b_ptr) +void tipc_disc_rcv(struct sk_buff *buf, struct tipc_bearer *bearer) { - struct tipc_node *n_ptr; + struct tipc_node *node; struct tipc_link *link; - struct tipc_media_addr media_addr, *addr; + struct tipc_media_addr maddr; struct sk_buff *rbuf; struct tipc_msg *msg = buf_msg(buf); - u32 dest = msg_dest_domain(msg); - u32 orig = msg_prevnode(msg); + u32 ddom = msg_dest_domain(msg); + u32 onode = msg_prevnode(msg); u32 net_id = msg_bc_netid(msg); - u32 type = msg_type(msg); - int link_fully_up; - - media_addr.broadcast = 1; - b_ptr->media->msg2addr(&media_addr, msg_media_addr(msg)); - buf_discard(buf); - - /* Validate discovery message from requesting node */ + u32 mtyp = msg_type(msg); + u32 signature = msg_node_sig(msg); + bool addr_match = false; + bool sign_match = false; + bool link_up = false; + bool accept_addr = false; + bool accept_sign = false; + bool respond = false; + + bearer->media->msg2addr(bearer, &maddr, msg_media_addr(msg)); + kfree_skb(buf); + + /* Ensure message from node is valid and communication is permitted */ if (net_id != tipc_net_id) return; - if (media_addr.broadcast) + if (maddr.broadcast) return; - if (!tipc_addr_domain_valid(dest)) + if (!tipc_addr_domain_valid(ddom)) return; - if (!tipc_addr_node_valid(orig)) + if (!tipc_addr_node_valid(onode)) return; - if (orig == tipc_own_addr) { - if (memcmp(&media_addr, &b_ptr->addr, sizeof(media_addr))) - disc_dupl_alert(b_ptr, tipc_own_addr, &media_addr); + + if (in_own_node(onode)) { + if (memcmp(&maddr, &bearer->addr, sizeof(maddr))) + disc_dupl_alert(bearer, tipc_own_addr, &maddr); return; } - if (!tipc_in_scope(dest, tipc_own_addr)) + if (!tipc_in_scope(ddom, tipc_own_addr)) return; - if (!tipc_in_scope(b_ptr->link_req->domain, orig)) + if (!tipc_in_scope(bearer->domain, onode)) return; - /* Locate structure corresponding to requesting node */ - n_ptr = tipc_node_find(orig); - if (!n_ptr) { - n_ptr = tipc_node_create(orig); - if (!n_ptr) - return; - } - tipc_node_lock(n_ptr); - - link = n_ptr->links[b_ptr->identity]; + /* Locate, or if necessary, create, node: */ + node = tipc_node_find(onode); + if (!node) + node = tipc_node_create(onode); + if (!node) + return; - /* Create a link endpoint for this bearer, if necessary */ - if (!link) { - link = tipc_link_create(n_ptr, b_ptr, &media_addr); - if (!link) { - tipc_node_unlock(n_ptr); - return; - } + tipc_node_lock(node); + link = node->links[bearer->identity]; + + /* Prepare to validate requesting node's signature and media address */ + sign_match = (signature == node->signature); + addr_match = link && !memcmp(&link->media_addr, &maddr, sizeof(maddr)); + link_up = link && tipc_link_is_up(link); + + + /* These three flags give us eight permutations: */ + + if (sign_match && addr_match && link_up) { + /* All is fine. Do nothing. */ + } else if (sign_match && addr_match && !link_up) { + /* Respond. The link will come up in due time */ + respond = true; + } else if (sign_match && !addr_match && link_up) { + /* Peer has changed i/f address without rebooting. + * If so, the link will reset soon, and the next + * discovery will be accepted. So we can ignore it. + * It may also be an cloned or malicious peer having + * chosen the same node address and signature as an + * existing one. + * Ignore requests until the link goes down, if ever. + */ + disc_dupl_alert(bearer, onode, &maddr); + } else if (sign_match && !addr_match && !link_up) { + /* Peer link has changed i/f address without rebooting. + * It may also be a cloned or malicious peer; we can't + * distinguish between the two. + * The signature is correct, so we must accept. + */ + accept_addr = true; + respond = true; + } else if (!sign_match && addr_match && link_up) { + /* Peer node rebooted. Two possibilities: + * - Delayed re-discovery; this link endpoint has already + * reset and re-established contact with the peer, before + * receiving a discovery message from that node. + * (The peer happened to receive one from this node first). + * - The peer came back so fast that our side has not + * discovered it yet. Probing from this side will soon + * reset the link, since there can be no working link + * endpoint at the peer end, and the link will re-establish. + * Accept the signature, since it comes from a known peer. + */ + accept_sign = true; + } else if (!sign_match && addr_match && !link_up) { + /* The peer node has rebooted. + * Accept signature, since it is a known peer. + */ + accept_sign = true; + respond = true; + } else if (!sign_match && !addr_match && link_up) { + /* Peer rebooted with new address, or a new/duplicate peer. + * Ignore until the link goes down, if ever. + */ + disc_dupl_alert(bearer, onode, &maddr); + } else if (!sign_match && !addr_match && !link_up) { + /* Peer rebooted with new address, or it is a new peer. + * Accept signature and address. + */ + accept_sign = true; + accept_addr = true; + respond = true; } - /* - * Ensure requesting node's media address is correct - * - * If media address doesn't match and the link is working, reject the - * request (must be from a duplicate node). - * - * If media address doesn't match and the link is not working, accept - * the new media address and reset the link to ensure it starts up - * cleanly. - */ - addr = &link->media_addr; - if (memcmp(addr, &media_addr, sizeof(*addr))) { - if (tipc_link_is_up(link) || (!link->started)) { - disc_dupl_alert(b_ptr, orig, &media_addr); - tipc_node_unlock(n_ptr); - return; + if (accept_sign) + node->signature = signature; + + if (accept_addr) { + if (!link) + link = tipc_link_create(node, bearer, &maddr); + if (link) { + memcpy(&link->media_addr, &maddr, sizeof(maddr)); + tipc_link_reset(link); + } else { + respond = false; } - warn("Resetting link <%s>, peer interface address changed\n", - link->name); - memcpy(addr, &media_addr, sizeof(*addr)); - tipc_link_reset(link); } - /* Accept discovery message & send response, if necessary */ - link_fully_up = link_working_working(link); - - if ((type == DSC_REQ_MSG) && !link_fully_up && !b_ptr->blocked) { - rbuf = tipc_disc_init_msg(DSC_RESP_MSG, orig, b_ptr); + /* Send response, if necessary */ + if (respond && (mtyp == DSC_REQ_MSG)) { + rbuf = tipc_buf_acquire(INT_H_SIZE); if (rbuf) { - b_ptr->media->send_msg(rbuf, b_ptr, &media_addr); - buf_discard(rbuf); + tipc_disc_init_msg(rbuf, DSC_RESP_MSG, bearer); + tipc_bearer_send(bearer->identity, rbuf, &maddr); + kfree_skb(rbuf); } } - - tipc_node_unlock(n_ptr); + tipc_node_unlock(node); } /** @@ -217,7 +259,6 @@ void tipc_disc_recv_msg(struct sk_buff *buf, struct tipc_bearer *b_ptr) * Reinitiates discovery process if discovery object has no associated nodes * and is either not currently searching or is searching at a slow rate */ - static void disc_update(struct tipc_link_req *req) { if (!req->num_nodes) { @@ -233,32 +274,23 @@ static void disc_update(struct tipc_link_req *req) * tipc_disc_add_dest - increment set of discovered nodes * @req: ptr to link request structure */ - void tipc_disc_add_dest(struct tipc_link_req *req) { + spin_lock_bh(&req->lock); req->num_nodes++; + spin_unlock_bh(&req->lock); } /** * tipc_disc_remove_dest - decrement set of discovered nodes * @req: ptr to link request structure */ - void tipc_disc_remove_dest(struct tipc_link_req *req) { + spin_lock_bh(&req->lock); req->num_nodes--; disc_update(req); -} - -/** - * disc_send_msg - send link setup request message - * @req: ptr to link request structure - */ - -static void disc_send_msg(struct tipc_link_req *req) -{ - if (!req->bearer->blocked) - tipc_bearer_send(req->bearer, req->buf, &req->dest); + spin_unlock_bh(&req->lock); } /** @@ -267,15 +299,13 @@ static void disc_send_msg(struct tipc_link_req *req) * * Called whenever a link setup request timer associated with a bearer expires. */ - static void disc_timeout(struct tipc_link_req *req) { int max_delay; - spin_lock_bh(&req->bearer->lock); + spin_lock_bh(&req->lock); /* Stop searching if only desired node has been found */ - if (tipc_node(req->domain) && req->num_nodes) { req->timer_intv = TIPC_LINK_REQ_INACTIVE; goto exit; @@ -288,8 +318,8 @@ static void disc_timeout(struct tipc_link_req *req) * hold at fast polling rate if don't have any associated nodes, * otherwise hold at slow polling rate */ + tipc_bearer_send(req->bearer_id, req->buf, &req->dest); - disc_send_msg(req); req->timer_intv *= 2; if (req->num_nodes) @@ -301,7 +331,7 @@ static void disc_timeout(struct tipc_link_req *req) k_start_timer(&req->timer, req->timer_intv); exit: - spin_unlock_bh(&req->bearer->lock); + spin_unlock_bh(&req->lock); } /** @@ -312,9 +342,7 @@ 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) { struct tipc_link_req *req; @@ -322,21 +350,23 @@ int tipc_disc_create(struct tipc_bearer *b_ptr, if (!req) return -ENOMEM; - req->buf = tipc_disc_init_msg(DSC_REQ_MSG, dest_domain, b_ptr); + req->buf = tipc_buf_acquire(INT_H_SIZE); if (!req->buf) { kfree(req); - return -ENOMSG; + return -ENOMEM; } + tipc_disc_init_msg(req->buf, DSC_REQ_MSG, b_ptr); memcpy(&req->dest, dest, sizeof(*dest)); - req->bearer = b_ptr; - req->domain = dest_domain; + req->bearer_id = b_ptr->identity; + req->domain = b_ptr->domain; req->num_nodes = 0; req->timer_intv = TIPC_LINK_REQ_INIT; + spin_lock_init(&req->lock); k_init_timer(&req->timer, (Handler)disc_timeout, (unsigned long)req); k_start_timer(&req->timer, req->timer_intv); b_ptr->link_req = req; - disc_send_msg(req); + tipc_bearer_send(req->bearer_id, req->buf, &req->dest); return 0; } @@ -344,12 +374,30 @@ int tipc_disc_create(struct tipc_bearer *b_ptr, * tipc_disc_delete - destroy object sending periodic link setup requests * @req: ptr to link request structure */ - void tipc_disc_delete(struct tipc_link_req *req) { k_cancel_timer(&req->timer); k_term_timer(&req->timer); - buf_discard(req->buf); + kfree_skb(req->buf); kfree(req); } +/** + * tipc_disc_reset - reset object to send periodic link setup requests + * @b_ptr: ptr to bearer issuing requests + * @dest_domain: network domain to which links can be established + */ +void tipc_disc_reset(struct tipc_bearer *b_ptr) +{ + struct tipc_link_req *req = b_ptr->link_req; + + spin_lock_bh(&req->lock); + tipc_disc_init_msg(req->buf, DSC_REQ_MSG, b_ptr); + req->bearer_id = b_ptr->identity; + req->domain = b_ptr->domain; + req->num_nodes = 0; + req->timer_intv = TIPC_LINK_REQ_INIT; + k_start_timer(&req->timer, req->timer_intv); + tipc_bearer_send(req->bearer_id, req->buf, &req->dest); + spin_unlock_bh(&req->lock); +} diff --git a/net/tipc/discover.h b/net/tipc/discover.h index 75b67c403aa..515b57392f4 100644 --- a/net/tipc/discover.h +++ b/net/tipc/discover.h @@ -39,11 +39,11 @@ struct tipc_link_req; -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); void tipc_disc_delete(struct tipc_link_req *req); +void tipc_disc_reset(struct tipc_bearer *b_ptr); void tipc_disc_add_dest(struct tipc_link_req *req); void tipc_disc_remove_dest(struct tipc_link_req *req); -void tipc_disc_recv_msg(struct sk_buff *buf, struct tipc_bearer *b_ptr); +void tipc_disc_rcv(struct sk_buff *buf, struct tipc_bearer *b_ptr); #endif diff --git a/net/tipc/eth_media.c b/net/tipc/eth_media.c index 527e3f0e165..5e1426f1751 100644 --- a/net/tipc/eth_media.c +++ b/net/tipc/eth_media.c @@ -1,8 +1,8 @@ /* * net/tipc/eth_media.c: Ethernet bearer support for TIPC * - * Copyright (c) 2001-2007, Ericsson AB - * Copyright (c) 2005-2008, 2011, Wind River Systems + * Copyright (c) 2001-2007, 2013-2014, Ericsson AB + * Copyright (c) 2005-2008, 2011-2013, Wind River Systems * All rights reserved. * * Redistribution and use in source and binary forms, with or without @@ -37,362 +37,65 @@ #include "core.h" #include "bearer.h" -#define MAX_ETH_BEARERS MAX_BEARERS +#define ETH_ADDR_OFFSET 4 /* MAC addr position inside address field */ -#define ETH_ADDR_OFFSET 4 /* message header offset of MAC address */ - -/** - * struct eth_bearer - Ethernet bearer data structure - * @bearer: ptr to associated "generic" bearer structure - * @dev: ptr to associated Ethernet network device - * @tipc_packet_type: used in binding TIPC to Ethernet driver - * @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 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; - -/** - * eth_media_addr_set - initialize Ethernet media address structure - * - * 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) -{ - 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); -} - -/** - * send_msg - send a TIPC message out over an Ethernet 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 eth_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; - 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 Ethernet interface - * - * Accept only packets explicitly sent to this node, or broadcast packets; - * ignores packets sent using Ethernet 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 eth_bearer *eb_ptr = (struct eth_bearer *)pt->af_packet_priv; - - if (!net_eq(dev_net(dev), &init_net)) { - kfree_skb(buf); - return 0; - } - - if (likely(eb_ptr->bearer)) { - if (likely(buf->pkt_type <= PACKET_BROADCAST)) { - buf->next = NULL; - tipc_recv_msg(buf, eb_ptr->bearer); - return 0; - } - } - kfree_skb(buf); - return 0; -} - -/** - * 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 eth_bearer *eb_ptr = ð_bearers[0]; - struct eth_bearer *stop = ð_bearers[MAX_ETH_BEARERS]; - char *driver_name = strchr((const char *)tb_ptr->name, ':') + 1; - int pending_dev = 0; - - /* Find unused Ethernet bearer structure */ - - while (eb_ptr->dev) { - if (!eb_ptr->bearer) - pending_dev++; - if (++eb_ptr == stop) - return pending_dev ? -EAGAIN : -EDQUOT; - } - - /* 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); - if (!dev) - return -ENODEV; - - /* Create Ethernet bearer for device */ - - eb_ptr->dev = dev; - eb_ptr->tipc_packet_type.type = htons(ETH_P_TIPC); - eb_ptr->tipc_packet_type.dev = dev; - 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); - - /* Associate TIPC bearer with Ethernet bearer */ - - eb_ptr->bearer = tb_ptr; - tb_ptr->usr_handle = (void *)eb_ptr; - tb_ptr->mtu = dev->mtu; - tb_ptr->blocked = 0; - eth_media_addr_set(&tb_ptr->addr, (char *)dev->dev_addr); - return 0; -} - -/** - * cleanup_bearer - break association between Ethernet bearer and interface - * - * This routine must be invoked from a work queue because it can sleep. - */ - -static void cleanup_bearer(struct work_struct *work) +/* Convert Ethernet address (media address format) to string */ +static int tipc_eth_addr2str(struct tipc_media_addr *addr, + char *strbuf, int bufsz) { - struct eth_bearer *eb_ptr = - container_of(work, struct eth_bearer, cleanup); - - dev_remove_pack(&eb_ptr->tipc_packet_type); - dev_put(eb_ptr->dev); - eb_ptr->dev = NULL; -} - -/** - * disable_bearer - detach TIPC bearer from an Ethernet interface - * - * Mark Ethernet 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 eth_bearer *eb_ptr = (struct eth_bearer *)tb_ptr->usr_handle; - - eb_ptr->bearer = NULL; - INIT_WORK(&eb_ptr->cleanup, cleanup_bearer); - schedule_work(&eb_ptr->cleanup); -} - -/** - * recv_notification - handle device updates from OS - * - * Change the state of the Ethernet bearer (if any) associated with the - * specified device. - */ - -static int recv_notification(struct notifier_block *nb, unsigned long evt, - void *dv) -{ - struct net_device *dev = (struct net_device *)dv; - struct eth_bearer *eb_ptr = ð_bearers[0]; - struct eth_bearer *stop = ð_bearers[MAX_ETH_BEARERS]; - - if (!net_eq(dev_net(dev), &init_net)) - return NOTIFY_DONE; - - while ((eb_ptr->dev != dev)) { - if (++eb_ptr == stop) - return NOTIFY_DONE; /* couldn't find device */ - } - if (!eb_ptr->bearer) - return NOTIFY_DONE; /* bearer had been disabled */ - - eb_ptr->bearer->mtu = dev->mtu; - - switch (evt) { - case NETDEV_CHANGE: - if (netif_carrier_ok(dev)) - tipc_continue(eb_ptr->bearer); - else - tipc_block_bearer(eb_ptr->bearer->name); - break; - case NETDEV_UP: - tipc_continue(eb_ptr->bearer); - break; - case NETDEV_DOWN: - tipc_block_bearer(eb_ptr->bearer->name); - break; - case NETDEV_CHANGEMTU: - case NETDEV_CHANGEADDR: - tipc_block_bearer(eb_ptr->bearer->name); - tipc_continue(eb_ptr->bearer); - break; - case NETDEV_UNREGISTER: - case NETDEV_CHANGENAME: - tipc_disable_bearer(eb_ptr->bearer->name); - break; - } - return NOTIFY_OK; -} - -/** - * eth_addr2str - convert Ethernet address to string - */ - -static int eth_addr2str(struct tipc_media_addr *a, char *str_buf, int str_size) -{ - if (str_size < 18) /* 18 = strlen("aa:bb:cc:dd:ee:ff\0") */ + if (bufsz < 18) /* 18 = strlen("aa:bb:cc:dd:ee:ff\0") */ return 1; - sprintf(str_buf, "%pM", a->value); + sprintf(strbuf, "%pM", addr->value); return 0; } -/** - * eth_str2addr - convert string to Ethernet address - */ - -static int eth_str2addr(struct tipc_media_addr *a, char *str_buf) +/* Convert from media address format to discovery message addr format */ +static int tipc_eth_addr2msg(char *msg, struct tipc_media_addr *addr) { - 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); + memset(msg, 0, TIPC_MEDIA_ADDR_SIZE); + msg[TIPC_MEDIA_TYPE_OFFSET] = TIPC_MEDIA_TYPE_ETH; + memcpy(msg + ETH_ADDR_OFFSET, addr->value, ETH_ALEN); return 0; } -/** - * eth_str2addr - convert Ethernet address format to message header format - */ - -static int eth_addr2msg(struct tipc_media_addr *a, char *msg_area) +/* Convert raw mac address format to media addr format */ +static int tipc_eth_raw2addr(struct tipc_bearer *b, + struct tipc_media_addr *addr, + char *msg) { - memset(msg_area, 0, TIPC_MEDIA_ADDR_SIZE); - msg_area[TIPC_MEDIA_TYPE_OFFSET] = TIPC_MEDIA_TYPE_ETH; - memcpy(msg_area + ETH_ADDR_OFFSET, a->value, ETH_ALEN); + char bcast_mac[ETH_ALEN] = {0xff, 0xff, 0xff, 0xff, 0xff, 0xff}; + + memset(addr, 0, sizeof(*addr)); + ether_addr_copy(addr->value, msg); + addr->media_id = TIPC_MEDIA_TYPE_ETH; + addr->broadcast = !memcmp(addr->value, bcast_mac, ETH_ALEN); return 0; } -/** - * eth_str2addr - convert message header address format to Ethernet format - */ - -static int eth_msg2addr(struct tipc_media_addr *a, char *msg_area) +/* Convert discovery msg addr format to Ethernet media addr format */ +static int tipc_eth_msg2addr(struct tipc_bearer *b, + struct tipc_media_addr *addr, + char *msg) { - if (msg_area[TIPC_MEDIA_TYPE_OFFSET] != TIPC_MEDIA_TYPE_ETH) - return 1; - - eth_media_addr_set(a, msg_area + ETH_ADDR_OFFSET); - return 0; + /* Skip past preamble: */ + msg += ETH_ADDR_OFFSET; + return tipc_eth_raw2addr(b, addr, msg); } -/* - * Ethernet media registration info - */ - -static struct tipc_media eth_media_info = { - .send_msg = send_msg, - .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 }, +/* Ethernet media registration info */ +struct tipc_media eth_media_info = { + .send_msg = tipc_l2_send_msg, + .enable_media = tipc_enable_l2_media, + .disable_media = tipc_disable_l2_media, + .addr2str = tipc_eth_addr2str, + .addr2msg = tipc_eth_addr2msg, + .msg2addr = tipc_eth_msg2addr, + .raw2addr = tipc_eth_raw2addr, .priority = TIPC_DEF_LINK_PRI, .tolerance = TIPC_DEF_LINK_TOL, .window = TIPC_DEF_LINK_WIN, .type_id = TIPC_MEDIA_TYPE_ETH, + .hwaddr_len = ETH_ALEN, .name = "eth" }; - -/** - * tipc_eth_media_start - activate Ethernet bearer support - * - * Register Ethernet media type with TIPC bearer code. Also register - * with OS for notifications about device state changes. - */ - -int tipc_eth_media_start(void) -{ - int res; - - if (eth_started) - return -EINVAL; - - res = tipc_register_media(ð_media_info); - if (res) - return res; - - notifier.notifier_call = &recv_notification; - notifier.priority = 0; - res = register_netdevice_notifier(¬ifier); - if (!res) - eth_started = 1; - return res; -} - -/** - * tipc_eth_media_stop - deactivate Ethernet bearer support - */ - -void tipc_eth_media_stop(void) -{ - if (!eth_started) - return; - - flush_scheduled_work(); - unregister_netdevice_notifier(¬ifier); - eth_started = 0; -} diff --git a/net/tipc/handler.c b/net/tipc/handler.c deleted file mode 100644 index 274c98e164b..00000000000 --- a/net/tipc/handler.c +++ /dev/null @@ -1,132 +0,0 @@ -/* - * net/tipc/handler.c: TIPC signal handling - * - * Copyright (c) 2000-2006, Ericsson AB - * Copyright (c) 2005, 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" - -struct queue_item { - struct list_head next_signal; - void (*handler) (unsigned long); - unsigned long data; -}; - -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 void process_signal_queue(unsigned long dummy); - -static DECLARE_TASKLET_DISABLED(tipc_tasklet, process_signal_queue, 0); - - -unsigned int tipc_k_signal(Handler routine, unsigned long argument) -{ - struct queue_item *item; - - if (!handler_enabled) { - err("Signal request ignored by handler\n"); - return -ENOPROTOOPT; - } - - spin_lock_bh(&qitem_lock); - item = kmem_cache_alloc(tipc_queue_item_cache, GFP_ATOMIC); - if (!item) { - err("Signal queue out of memory\n"); - spin_unlock_bh(&qitem_lock); - return -ENOMEM; - } - item->handler = routine; - item->data = argument; - list_add_tail(&item->next_signal, &signal_queue_head); - spin_unlock_bh(&qitem_lock); - tasklet_schedule(&tipc_tasklet); - return 0; -} - -static void process_signal_queue(unsigned long dummy) -{ - struct queue_item *__volatile__ item; - struct list_head *l, *n; - - spin_lock_bh(&qitem_lock); - list_for_each_safe(l, n, &signal_queue_head) { - item = list_entry(l, struct queue_item, next_signal); - list_del(&item->next_signal); - spin_unlock_bh(&qitem_lock); - item->handler(item->data); - spin_lock_bh(&qitem_lock); - kmem_cache_free(tipc_queue_item_cache, item); - } - spin_unlock_bh(&qitem_lock); -} - -int tipc_handler_start(void) -{ - tipc_queue_item_cache = - kmem_cache_create("tipc_queue_items", sizeof(struct queue_item), - 0, SLAB_HWCACHE_ALIGN, NULL); - if (!tipc_queue_item_cache) - return -ENOMEM; - - INIT_LIST_HEAD(&signal_queue_head); - tasklet_enable(&tipc_tasklet); - handler_enabled = 1; - return 0; -} - -void tipc_handler_stop(void) -{ - struct list_head *l, *n; - struct queue_item *item; - - if (!handler_enabled) - return; - - handler_enabled = 0; - tasklet_disable(&tipc_tasklet); - tasklet_kill(&tipc_tasklet); - - spin_lock_bh(&qitem_lock); - list_for_each_safe(l, n, &signal_queue_head) { - item = list_entry(l, struct queue_item, next_signal); - list_del(&item->next_signal); - kmem_cache_free(tipc_queue_item_cache, item); - } - spin_unlock_bh(&qitem_lock); - - kmem_cache_destroy(tipc_queue_item_cache); -} - diff --git a/net/tipc/ib_media.c b/net/tipc/ib_media.c new file mode 100644 index 00000000000..8522eef9c13 --- /dev/null +++ b/net/tipc/ib_media.c @@ -0,0 +1,101 @@ +/* + * 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" + +/* convert InfiniBand address (media address format) media address to string */ +static int tipc_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; +} + +/* Convert from media address format to discovery message addr format */ +static int tipc_ib_addr2msg(char *msg, struct tipc_media_addr *addr) +{ + memset(msg, 0, TIPC_MEDIA_ADDR_SIZE); + memcpy(msg, addr->value, INFINIBAND_ALEN); + return 0; +} + +/* Convert raw InfiniBand address format to media addr format */ +static int tipc_ib_raw2addr(struct tipc_bearer *b, + struct tipc_media_addr *addr, + char *msg) +{ + memset(addr, 0, sizeof(*addr)); + memcpy(addr->value, msg, INFINIBAND_ALEN); + addr->media_id = TIPC_MEDIA_TYPE_IB; + addr->broadcast = !memcmp(msg, b->bcast_addr.value, + INFINIBAND_ALEN); + return 0; +} + +/* Convert discovery msg addr format to InfiniBand media addr format */ +static int tipc_ib_msg2addr(struct tipc_bearer *b, + struct tipc_media_addr *addr, + char *msg) +{ + return tipc_ib_raw2addr(b, addr, msg); +} + +/* InfiniBand media registration info */ +struct tipc_media ib_media_info = { + .send_msg = tipc_l2_send_msg, + .enable_media = tipc_enable_l2_media, + .disable_media = tipc_disable_l2_media, + .addr2str = tipc_ib_addr2str, + .addr2msg = tipc_ib_addr2msg, + .msg2addr = tipc_ib_msg2addr, + .raw2addr = tipc_ib_raw2addr, + .priority = TIPC_DEF_LINK_PRI, + .tolerance = TIPC_DEF_LINK_TOL, + .window = TIPC_DEF_LINK_WIN, + .type_id = TIPC_MEDIA_TYPE_IB, + .hwaddr_len = INFINIBAND_ALEN, + .name = "ib" +}; diff --git a/net/tipc/link.c b/net/tipc/link.c index ac1832a66f8..ad2c57f5868 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, 2012-2014, Ericsson AB + * Copyright (c) 2004-2007, 2010-2013, Wind River Systems * All rights reserved. * * Redistribution and use in source and binary forms, with or without @@ -37,21 +37,28 @@ #include "core.h" #include "link.h" #include "port.h" +#include "socket.h" #include "name_distr.h" #include "discover.h" #include "config.h" +#include <linux/pkt_sched.h> /* - * Out-of-range value for link session numbers + * Error message prefixes */ +static const char *link_co_err = "Link changeover error, "; +static const char *link_rst_msg = "Resetting link "; +static const char *link_unk_evt = "Unknown link event "; +/* + * Out-of-range value for link session numbers + */ #define INVALID_SESSION 0x10000 /* * Link state events: */ - #define STARTING_EVT 856384768 /* link processing trigger */ #define TRAFFIC_MSG_EVT 560815u /* rx'd ??? */ #define TIMEOUT_EVT 560817u /* link timer expired */ @@ -67,45 +74,27 @@ /* * State value stored in 'exp_msg_count' */ - #define START_CHANGEOVER 100000u -/** - * struct tipc_link_name - deconstructed link name - * @addr_local: network address of node at this end - * @if_local: name of interface at this end - * @addr_peer: network address of node at far end - * @if_peer: name of interface at far end - */ - -struct tipc_link_name { - u32 addr_local; - char if_local[TIPC_MAX_IF_NAME]; - u32 addr_peer; - char if_peer[TIPC_MAX_IF_NAME]; -}; - static void link_handle_out_of_seq_msg(struct tipc_link *l_ptr, struct sk_buff *buf); -static void link_recv_proto_msg(struct tipc_link *l_ptr, struct sk_buff *buf); -static int link_recv_changeover_msg(struct tipc_link **l_ptr, - struct sk_buff **buf); +static void tipc_link_proto_rcv(struct tipc_link *l_ptr, struct sk_buff *buf); +static int tipc_link_tunnel_rcv(struct tipc_node *n_ptr, + struct sk_buff **buf); static void link_set_supervision_props(struct tipc_link *l_ptr, u32 tolerance); -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 int tipc_link_iovec_long_xmit(struct tipc_port *sender, + struct iovec const *msg_sect, + unsigned int len, u32 destnode); 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 int tipc_link_frag_xmit(struct tipc_link *l_ptr, struct sk_buff *buf); +static void tipc_link_sync_xmit(struct tipc_link *l); +static void tipc_link_sync_rcv(struct tipc_node *n, struct sk_buff *buf); /* * Simple link routines */ - static unsigned int align(unsigned int i) { return (i + 3) & ~3u; @@ -113,9 +102,18 @@ static unsigned int align(unsigned int i) static void link_init_max_pkt(struct tipc_link *l_ptr) { + struct tipc_bearer *b_ptr; u32 max_pkt; - max_pkt = (l_ptr->b_ptr->mtu & ~3); + rcu_read_lock(); + b_ptr = rcu_dereference_rtnl(bearer_list[l_ptr->bearer_id]); + if (!b_ptr) { + rcu_read_unlock(); + return; + } + max_pkt = (b_ptr->mtu & ~3); + rcu_read_unlock(); + if (max_pkt > MAX_MSG_SIZE) max_pkt = MAX_MSG_SIZE; @@ -143,7 +141,6 @@ static u32 link_last_sent(struct tipc_link *l_ptr) /* * Simple non-static link routines (i.e. referenced outside this file) */ - int tipc_link_is_up(struct tipc_link *l_ptr) { if (!l_ptr) @@ -158,94 +155,14 @@ int tipc_link_is_active(struct tipc_link *l_ptr) } /** - * link_name_validate - validate & (optionally) deconstruct tipc_link name - * @name - ptr to link name string - * @name_parts - ptr to area for link name components (or NULL if not needed) - * - * Returns 1 if link name is valid, otherwise 0. - */ - -static int link_name_validate(const char *name, - struct tipc_link_name *name_parts) -{ - char name_copy[TIPC_MAX_LINK_NAME]; - char *addr_local; - char *if_local; - char *addr_peer; - char *if_peer; - char dummy; - u32 z_local, c_local, n_local; - u32 z_peer, c_peer, n_peer; - u32 if_local_len; - u32 if_peer_len; - - /* copy link name & ensure length is OK */ - - name_copy[TIPC_MAX_LINK_NAME - 1] = 0; - /* need above in case non-Posix strncpy() doesn't pad with nulls */ - strncpy(name_copy, name, TIPC_MAX_LINK_NAME); - if (name_copy[TIPC_MAX_LINK_NAME - 1] != 0) - return 0; - - /* ensure all component parts of link name are present */ - - addr_local = name_copy; - if_local = strchr(addr_local, ':'); - if (if_local == NULL) - return 0; - *(if_local++) = 0; - addr_peer = strchr(if_local, '-'); - if (addr_peer == NULL) - return 0; - *(addr_peer++) = 0; - if_local_len = addr_peer - if_local; - if_peer = strchr(addr_peer, ':'); - if (if_peer == NULL) - return 0; - *(if_peer++) = 0; - if_peer_len = strlen(if_peer) + 1; - - /* validate component parts of link name */ - - if ((sscanf(addr_local, "%u.%u.%u%c", - &z_local, &c_local, &n_local, &dummy) != 3) || - (sscanf(addr_peer, "%u.%u.%u%c", - &z_peer, &c_peer, &n_peer, &dummy) != 3) || - (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))) - return 0; - - /* return link name components, if necessary */ - - if (name_parts) { - name_parts->addr_local = tipc_addr(z_local, c_local, n_local); - strcpy(name_parts->if_local, if_local); - name_parts->addr_peer = tipc_addr(z_peer, c_peer, n_peer); - strcpy(name_parts->if_peer, if_peer); - } - return 1; -} - -/** * link_timeout - handle expiration of link timer * @l_ptr: pointer to link - * - * This routine must not grab "tipc_net_lock" to avoid a potential deadlock conflict - * with tipc_link_delete(). (There is no risk that the node will be deleted by - * another thread because tipc_link_delete() always cancels the link timer before - * tipc_node_delete() is called.) */ - static void link_timeout(struct tipc_link *l_ptr) { tipc_node_lock(l_ptr->owner); /* update counters used in statistical profiling of send traffic */ - l_ptr->stats.accu_queue_sz += l_ptr->out_queue_size; l_ptr->stats.queue_sz_counts++; @@ -279,8 +196,6 @@ 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); if (l_ptr->next_out) @@ -302,10 +217,9 @@ static void link_set_timer(struct tipc_link *l_ptr, u32 time) * * Returns pointer to link. */ - struct tipc_link *tipc_link_create(struct tipc_node *n_ptr, - struct tipc_bearer *b_ptr, - const struct tipc_media_addr *media_addr) + struct tipc_bearer *b_ptr, + const struct tipc_media_addr *media_addr) { struct tipc_link *l_ptr; struct tipc_msg *msg; @@ -315,20 +229,20 @@ struct tipc_link *tipc_link_create(struct tipc_node *n_ptr, if (n_ptr->link_cnt >= 2) { tipc_addr_string_fill(addr_string, n_ptr->addr); - err("Attempt to establish third link to %s\n", addr_string); + pr_err("Attempt to establish third link to %s\n", addr_string); return NULL; } if (n_ptr->links[b_ptr->identity]) { tipc_addr_string_fill(addr_string, n_ptr->addr); - err("Attempt to establish second link on <%s> to %s\n", - b_ptr->name, addr_string); + pr_err("Attempt to establish second link on <%s> to %s\n", + b_ptr->name, addr_string); return NULL; } l_ptr = kzalloc(sizeof(*l_ptr), GFP_ATOMIC); if (!l_ptr) { - warn("Link creation failed, no memory\n"); + pr_warn("Link creation failed, no memory\n"); return NULL; } @@ -344,7 +258,7 @@ struct tipc_link *tipc_link_create(struct tipc_node *n_ptr, l_ptr->owner = n_ptr; l_ptr->checkpoint = 1; l_ptr->peer_session = INVALID_SESSION; - l_ptr->b_ptr = b_ptr; + l_ptr->bearer_id = b_ptr->identity; link_set_supervision_props(l_ptr, b_ptr->tolerance); l_ptr->state = RESET_UNKNOWN; @@ -359,6 +273,7 @@ struct tipc_link *tipc_link_create(struct tipc_node *n_ptr, l_ptr->priority = b_ptr->priority; tipc_link_set_queue_limits(l_ptr, b_ptr->window); + l_ptr->net_plane = b_ptr->net_plane; link_init_max_pkt(l_ptr); l_ptr->next_out_no = 1; @@ -368,46 +283,44 @@ struct tipc_link *tipc_link_create(struct tipc_node *n_ptr, tipc_node_attach_link(n_ptr, l_ptr); - k_init_timer(&l_ptr->timer, (Handler)link_timeout, (unsigned long)l_ptr); - list_add_tail(&l_ptr->link_list, &b_ptr->links); - tipc_k_signal((Handler)link_start, (unsigned long)l_ptr); + k_init_timer(&l_ptr->timer, (Handler)link_timeout, + (unsigned long)l_ptr); + + link_state_event(l_ptr, STARTING_EVT); return l_ptr; } -/** - * tipc_link_delete - delete a link - * @l_ptr: pointer to link - * - * Note: 'tipc_net_lock' is write_locked, bearer is locked. - * This routine must not grab the node lock until after link timer cancellation - * to avoid a potential deadlock situation. - */ - -void tipc_link_delete(struct tipc_link *l_ptr) +void tipc_link_delete_list(unsigned int bearer_id, bool shutting_down) { - if (!l_ptr) { - err("Attempt to delete non-existent link\n"); - return; - } - - k_cancel_timer(&l_ptr->timer); + struct tipc_link *l_ptr; + struct tipc_node *n_ptr; - tipc_node_lock(l_ptr->owner); - tipc_link_reset(l_ptr); - tipc_node_detach_link(l_ptr->owner, l_ptr); - tipc_link_stop(l_ptr); - list_del_init(&l_ptr->link_list); - tipc_node_unlock(l_ptr->owner); - k_term_timer(&l_ptr->timer); - kfree(l_ptr); -} + rcu_read_lock(); + list_for_each_entry_rcu(n_ptr, &tipc_node_list, list) { + tipc_node_lock(n_ptr); + l_ptr = n_ptr->links[bearer_id]; + if (l_ptr) { + tipc_link_reset(l_ptr); + if (shutting_down || !tipc_node_is_up(n_ptr)) { + tipc_node_detach_link(l_ptr->owner, l_ptr); + tipc_link_reset_fragments(l_ptr); + tipc_node_unlock(n_ptr); -static void link_start(struct tipc_link *l_ptr) -{ - tipc_node_lock(l_ptr->owner); - link_state_event(l_ptr, STARTING_EVT); - tipc_node_unlock(l_ptr->owner); + /* Nobody else can access this link now: */ + del_timer_sync(&l_ptr->timer); + kfree(l_ptr); + } else { + /* Detach/delete when failover is finished: */ + l_ptr->flags |= LINK_STOPPED; + tipc_node_unlock(n_ptr); + del_timer_sync(&l_ptr->timer); + } + continue; + } + tipc_node_unlock(n_ptr); + } + rcu_read_unlock(); } /** @@ -419,7 +332,6 @@ static void link_start(struct tipc_link *l_ptr) * Schedules port for renewed sending of messages after link congestion * has abated. */ - static int link_schedule_port(struct tipc_link *l_ptr, u32 origport, u32 sz) { struct tipc_port *p_ptr; @@ -427,8 +339,6 @@ static int link_schedule_port(struct tipc_link *l_ptr, u32 origport, u32 sz) spin_lock_bh(&tipc_port_list_lock); p_ptr = tipc_port_lock(origport); if (p_ptr) { - if (!p_ptr->wakeup) - goto exit; if (!list_empty(&p_ptr->wait_list)) goto exit; p_ptr->congested = 1; @@ -463,7 +373,7 @@ void tipc_link_wakeup_ports(struct tipc_link *l_ptr, int all) list_del_init(&p_ptr->wait_list); spin_lock_bh(p_ptr->lock); p_ptr->congested = 0; - p_ptr->wakeup(p_ptr); + tipc_port_wakeup(p_ptr); win -= p_ptr->waiting_pkts; spin_unlock_bh(p_ptr->lock); } @@ -476,17 +386,9 @@ exit: * link_release_outqueue - purge link's outbound message queue * @l_ptr: pointer to link */ - static void link_release_outqueue(struct tipc_link *l_ptr) { - struct sk_buff *buf = l_ptr->first_out; - struct sk_buff *next; - - while (buf) { - next = buf->next; - buf_discard(buf); - buf = next; - } + kfree_skb_list(l_ptr->first_out); l_ptr->first_out = NULL; l_ptr->out_queue_size = 0; } @@ -495,53 +397,27 @@ static void link_release_outqueue(struct tipc_link *l_ptr) * tipc_link_reset_fragments - purge link's inbound message fragments queue * @l_ptr: pointer to link */ - void tipc_link_reset_fragments(struct tipc_link *l_ptr) { - struct sk_buff *buf = l_ptr->defragm_buf; - struct sk_buff *next; - - while (buf) { - next = buf->next; - buf_discard(buf); - buf = next; - } - l_ptr->defragm_buf = NULL; + kfree_skb(l_ptr->reasm_buf); + l_ptr->reasm_buf = NULL; } /** - * tipc_link_stop - purge all inbound and outbound messages associated with link + * tipc_link_purge_queues - purge all pkt queues associated with link * @l_ptr: pointer to link */ - -void tipc_link_stop(struct tipc_link *l_ptr) +void tipc_link_purge_queues(struct tipc_link *l_ptr) { - struct sk_buff *buf; - struct sk_buff *next; - - buf = l_ptr->oldest_deferred_in; - while (buf) { - next = buf->next; - buf_discard(buf); - buf = next; - } - - buf = l_ptr->first_out; - while (buf) { - next = buf->next; - buf_discard(buf); - buf = next; - } - + kfree_skb_list(l_ptr->oldest_deferred_in); + kfree_skb_list(l_ptr->first_out); tipc_link_reset_fragments(l_ptr); - - buf_discard(l_ptr->proto_msg_queue); + kfree_skb(l_ptr->proto_msg_queue); l_ptr->proto_msg_queue = NULL; } void tipc_link_reset(struct tipc_link *l_ptr) { - struct sk_buff *buf; u32 prev_state = l_ptr->state; u32 checkpoint = l_ptr->next_in_no; int was_active_link = tipc_link_is_active(l_ptr); @@ -560,25 +436,18 @@ void tipc_link_reset(struct tipc_link *l_ptr) return; tipc_node_link_down(l_ptr->owner, l_ptr); - tipc_bearer_remove_dest(l_ptr->b_ptr, l_ptr->addr); + tipc_bearer_remove_dest(l_ptr->bearer_id, l_ptr->addr); - if (was_active_link && tipc_node_active_links(l_ptr->owner) && - l_ptr->owner->permit_changeover) { + if (was_active_link && tipc_node_active_links(l_ptr->owner)) { l_ptr->reset_checkpoint = checkpoint; l_ptr->exp_msg_count = START_CHANGEOVER; } /* Clean up all queues: */ - link_release_outqueue(l_ptr); - buf_discard(l_ptr->proto_msg_queue); + kfree_skb(l_ptr->proto_msg_queue); l_ptr->proto_msg_queue = NULL; - buf = l_ptr->oldest_deferred_in; - while (buf) { - struct sk_buff *next = buf->next; - buf_discard(buf); - buf = next; - } + kfree_skb_list(l_ptr->oldest_deferred_in); if (!list_empty(&l_ptr->waiting_ports)) tipc_link_wakeup_ports(l_ptr, 1); @@ -598,12 +467,27 @@ void tipc_link_reset(struct tipc_link *l_ptr) link_reset_statistics(l_ptr); } +void tipc_link_reset_list(unsigned int bearer_id) +{ + struct tipc_link *l_ptr; + struct tipc_node *n_ptr; + + rcu_read_lock(); + list_for_each_entry_rcu(n_ptr, &tipc_node_list, list) { + tipc_node_lock(n_ptr); + l_ptr = n_ptr->links[bearer_id]; + if (l_ptr) + tipc_link_reset(l_ptr); + tipc_node_unlock(n_ptr); + } + rcu_read_unlock(); +} static void link_activate(struct tipc_link *l_ptr) { l_ptr->next_in_no = l_ptr->stats.recv_info = 1; tipc_node_link_up(l_ptr->owner, l_ptr); - tipc_bearer_add_dest(l_ptr->b_ptr, l_ptr->addr); + tipc_bearer_add_dest(l_ptr->bearer_id, l_ptr->addr); } /** @@ -611,19 +495,22 @@ static void link_activate(struct tipc_link *l_ptr) * @l_ptr: pointer to link * @event: state machine event to process */ - -static void link_state_event(struct tipc_link *l_ptr, unsigned event) +static void link_state_event(struct tipc_link *l_ptr, unsigned int event) { struct tipc_link *other; u32 cont_intv = l_ptr->continuity_interval; - if (!l_ptr->started && (event != STARTING_EVT)) + if (l_ptr->flags & LINK_STOPPED) + return; + + if (!(l_ptr->flags & LINK_STARTED) && (event != STARTING_EVT)) return; /* Not yet. */ - if (link_blocked(l_ptr)) { + /* Check whether changeover is going on */ + if (l_ptr->exp_msg_count) { if (event == TIMEOUT_EVT) link_set_timer(l_ptr, cont_intv); - return; /* Changeover going on */ + return; } switch (l_ptr->state) { @@ -636,12 +523,12 @@ static void link_state_event(struct tipc_link *l_ptr, unsigned event) if (l_ptr->next_in_no != l_ptr->checkpoint) { l_ptr->checkpoint = l_ptr->next_in_no; if (tipc_bclink_acks_missing(l_ptr->owner)) { - tipc_link_send_proto_msg(l_ptr, STATE_MSG, - 0, 0, 0, 0, 0); + tipc_link_proto_xmit(l_ptr, STATE_MSG, + 0, 0, 0, 0, 0); l_ptr->fsm_msg_cnt++; } else if (l_ptr->max_pkt < l_ptr->max_pkt_target) { - tipc_link_send_proto_msg(l_ptr, STATE_MSG, - 1, 0, 0, 0, 0); + tipc_link_proto_xmit(l_ptr, STATE_MSG, + 1, 0, 0, 0, 0); l_ptr->fsm_msg_cnt++; } link_set_timer(l_ptr, cont_intv); @@ -649,22 +536,23 @@ static void link_state_event(struct tipc_link *l_ptr, unsigned event) } l_ptr->state = WORKING_UNKNOWN; l_ptr->fsm_msg_cnt = 0; - tipc_link_send_proto_msg(l_ptr, STATE_MSG, 1, 0, 0, 0, 0); + tipc_link_proto_xmit(l_ptr, STATE_MSG, 1, 0, 0, 0, 0); l_ptr->fsm_msg_cnt++; link_set_timer(l_ptr, cont_intv / 4); break; case RESET_MSG: - info("Resetting link <%s>, requested by peer\n", - l_ptr->name); + pr_info("%s<%s>, requested by peer\n", link_rst_msg, + l_ptr->name); tipc_link_reset(l_ptr); l_ptr->state = RESET_RESET; l_ptr->fsm_msg_cnt = 0; - tipc_link_send_proto_msg(l_ptr, ACTIVATE_MSG, 0, 0, 0, 0, 0); + tipc_link_proto_xmit(l_ptr, ACTIVATE_MSG, + 0, 0, 0, 0, 0); l_ptr->fsm_msg_cnt++; link_set_timer(l_ptr, cont_intv); break; default: - err("Unknown link event %u in WW state\n", event); + pr_err("%s%u in WW state\n", link_unk_evt, event); } break; case WORKING_UNKNOWN: @@ -676,12 +564,13 @@ static void link_state_event(struct tipc_link *l_ptr, unsigned event) link_set_timer(l_ptr, cont_intv); break; case RESET_MSG: - info("Resetting link <%s>, requested by peer " - "while probing\n", l_ptr->name); + pr_info("%s<%s>, requested by peer while probing\n", + link_rst_msg, l_ptr->name); tipc_link_reset(l_ptr); l_ptr->state = RESET_RESET; l_ptr->fsm_msg_cnt = 0; - tipc_link_send_proto_msg(l_ptr, ACTIVATE_MSG, 0, 0, 0, 0, 0); + tipc_link_proto_xmit(l_ptr, ACTIVATE_MSG, + 0, 0, 0, 0, 0); l_ptr->fsm_msg_cnt++; link_set_timer(l_ptr, cont_intv); break; @@ -691,30 +580,30 @@ static void link_state_event(struct tipc_link *l_ptr, unsigned event) l_ptr->fsm_msg_cnt = 0; l_ptr->checkpoint = l_ptr->next_in_no; if (tipc_bclink_acks_missing(l_ptr->owner)) { - tipc_link_send_proto_msg(l_ptr, STATE_MSG, - 0, 0, 0, 0, 0); + tipc_link_proto_xmit(l_ptr, STATE_MSG, + 0, 0, 0, 0, 0); l_ptr->fsm_msg_cnt++; } link_set_timer(l_ptr, cont_intv); } else if (l_ptr->fsm_msg_cnt < l_ptr->abort_limit) { - tipc_link_send_proto_msg(l_ptr, STATE_MSG, - 1, 0, 0, 0, 0); + tipc_link_proto_xmit(l_ptr, STATE_MSG, + 1, 0, 0, 0, 0); l_ptr->fsm_msg_cnt++; link_set_timer(l_ptr, cont_intv / 4); } else { /* Link has failed */ - warn("Resetting link <%s>, peer not responding\n", - l_ptr->name); + pr_warn("%s<%s>, peer not responding\n", + link_rst_msg, l_ptr->name); tipc_link_reset(l_ptr); l_ptr->state = RESET_UNKNOWN; l_ptr->fsm_msg_cnt = 0; - tipc_link_send_proto_msg(l_ptr, RESET_MSG, - 0, 0, 0, 0, 0); + tipc_link_proto_xmit(l_ptr, RESET_MSG, + 0, 0, 0, 0, 0); l_ptr->fsm_msg_cnt++; link_set_timer(l_ptr, cont_intv); } break; default: - err("Unknown link event %u in WU state\n", event); + pr_err("%s%u in WU state\n", link_unk_evt, event); } break; case RESET_UNKNOWN: @@ -728,27 +617,30 @@ static void link_state_event(struct tipc_link *l_ptr, unsigned event) l_ptr->state = WORKING_WORKING; l_ptr->fsm_msg_cnt = 0; link_activate(l_ptr); - tipc_link_send_proto_msg(l_ptr, STATE_MSG, 1, 0, 0, 0, 0); + tipc_link_proto_xmit(l_ptr, STATE_MSG, 1, 0, 0, 0, 0); l_ptr->fsm_msg_cnt++; + if (l_ptr->owner->working_links == 1) + tipc_link_sync_xmit(l_ptr); link_set_timer(l_ptr, cont_intv); break; case RESET_MSG: l_ptr->state = RESET_RESET; l_ptr->fsm_msg_cnt = 0; - tipc_link_send_proto_msg(l_ptr, ACTIVATE_MSG, 1, 0, 0, 0, 0); + tipc_link_proto_xmit(l_ptr, ACTIVATE_MSG, + 1, 0, 0, 0, 0); l_ptr->fsm_msg_cnt++; link_set_timer(l_ptr, cont_intv); break; case STARTING_EVT: - l_ptr->started = 1; + l_ptr->flags |= LINK_STARTED; /* fall through */ case TIMEOUT_EVT: - tipc_link_send_proto_msg(l_ptr, RESET_MSG, 0, 0, 0, 0, 0); + tipc_link_proto_xmit(l_ptr, RESET_MSG, 0, 0, 0, 0, 0); l_ptr->fsm_msg_cnt++; link_set_timer(l_ptr, cont_intv); break; default: - err("Unknown link event %u in RU state\n", event); + pr_err("%s%u in RU state\n", link_unk_evt, event); } break; case RESET_RESET: @@ -761,23 +653,26 @@ static void link_state_event(struct tipc_link *l_ptr, unsigned event) l_ptr->state = WORKING_WORKING; l_ptr->fsm_msg_cnt = 0; link_activate(l_ptr); - tipc_link_send_proto_msg(l_ptr, STATE_MSG, 1, 0, 0, 0, 0); + tipc_link_proto_xmit(l_ptr, STATE_MSG, 1, 0, 0, 0, 0); l_ptr->fsm_msg_cnt++; + if (l_ptr->owner->working_links == 1) + tipc_link_sync_xmit(l_ptr); link_set_timer(l_ptr, cont_intv); break; case RESET_MSG: break; case TIMEOUT_EVT: - tipc_link_send_proto_msg(l_ptr, ACTIVATE_MSG, 0, 0, 0, 0, 0); + tipc_link_proto_xmit(l_ptr, ACTIVATE_MSG, + 0, 0, 0, 0, 0); l_ptr->fsm_msg_cnt++; link_set_timer(l_ptr, cont_intv); break; default: - err("Unknown link event %u in RR state\n", event); + pr_err("%s%u in RR state\n", link_unk_evt, event); } break; default: - err("Unknown link state %u/%u\n", l_ptr->state, event); + pr_err("Unknown link state %u/%u\n", l_ptr->state, event); } } @@ -785,9 +680,7 @@ static void link_state_event(struct tipc_link *l_ptr, unsigned 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); @@ -810,7 +703,7 @@ static int link_bundle_buf(struct tipc_link *l_ptr, skb_copy_to_linear_data_offset(bundler, to_pos, buf->data, size); msg_set_size(bundler_msg, to_pos + size); msg_set_msgcnt(bundler_msg, msg_msgcnt(bundler_msg) + 1); - buf_discard(buf); + kfree_skb(buf); l_ptr->stats.sent_bundled++; return 1; } @@ -856,12 +749,11 @@ static void link_add_chain_to_outqueue(struct tipc_link *l_ptr, } /* - * tipc_link_send_buf() is the 'full path' for messages, called from - * inside TIPC when the 'fast path' in tipc_send_buf + * tipc_link_xmit() is the 'full path' for messages, called from + * inside TIPC when the 'fast path' in tipc_send_xmit * has failed, and from link_send() */ - -int tipc_link_send_buf(struct tipc_link *l_ptr, struct sk_buff *buf) +int __tipc_link_xmit(struct tipc_link *l_ptr, struct sk_buff *buf) { struct tipc_msg *msg = buf_msg(buf); u32 size = msg_size(msg); @@ -871,59 +763,44 @@ int tipc_link_send_buf(struct tipc_link *l_ptr, struct sk_buff *buf) u32 queue_limit = l_ptr->queue_limit[imp]; u32 max_packet = l_ptr->max_pkt; - msg_set_prevnode(msg, tipc_own_addr); /* If routed message */ - /* Match msg importance against queue limits: */ - if (unlikely(queue_size >= queue_limit)) { if (imp <= TIPC_CRITICAL_IMPORTANCE) { link_schedule_port(l_ptr, msg_origport(msg), size); - buf_discard(buf); + kfree_skb(buf); return -ELINKCONG; } - buf_discard(buf); + kfree_skb(buf); if (imp > CONN_MANAGER) { - warn("Resetting link <%s>, send queue full", l_ptr->name); + pr_warn("%s<%s>, send queue full", link_rst_msg, + l_ptr->name); tipc_link_reset(l_ptr); } return dsz; } /* Fragmentation needed ? */ - if (size > max_packet) - return link_send_long_buf(l_ptr, buf); - - /* Packet can be queued or sent: */ + return tipc_link_frag_xmit(l_ptr, buf); - if (likely(!tipc_bearer_congested(l_ptr->b_ptr, l_ptr) && - !link_congested(l_ptr))) { + /* Packet can be queued or sent. */ + if (likely(!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->bearer_id, buf, &l_ptr->media_addr); + l_ptr->unacked_window = 0; return dsz; } - /* Congestion: can message be bundled ?: */ - + /* Congestion: can message be bundled ? */ if ((msg_user(msg) != CHANGEOVER_PROTOCOL) && (msg_user(msg) != MSG_FRAGMENTER)) { /* 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) { struct sk_buff *bundler = tipc_buf_acquire(max_packet); struct tipc_msg bundler_hdr; @@ -944,41 +821,78 @@ 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; } /* - * tipc_link_send(): same as tipc_link_send_buf(), but the link to use has - * not been selected yet, and the the owner node is not locked + * tipc_link_xmit(): same as __tipc_link_xmit(), but the link to use + * has not been selected yet, and the the owner node is not locked * Called by TIPC internal users, e.g. the name distributor */ - -int tipc_link_send(struct sk_buff *buf, u32 dest, u32 selector) +int tipc_link_xmit(struct sk_buff *buf, u32 dest, u32 selector) { struct tipc_link *l_ptr; struct tipc_node *n_ptr; int res = -ELINKCONG; - read_lock_bh(&tipc_net_lock); n_ptr = tipc_node_find(dest); if (n_ptr) { tipc_node_lock(n_ptr); l_ptr = n_ptr->active_links[selector & 1]; if (l_ptr) - res = tipc_link_send_buf(l_ptr, buf); + res = __tipc_link_xmit(l_ptr, buf); else - buf_discard(buf); + kfree_skb(buf); tipc_node_unlock(n_ptr); } else { - buf_discard(buf); + kfree_skb(buf); } - read_unlock_bh(&tipc_net_lock); return res; } /* - * tipc_link_send_names - send name table entries to new neighbor + * tipc_link_sync_xmit - 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_sync_xmit(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_sync_rcv - 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_sync_rcv(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_names_xmit - send name table entries to new neighbor * * Send routine for bulk delivery of name table messages when contact * with a new neighbor occurs. No link congestion checking is performed @@ -986,8 +900,7 @@ int tipc_link_send(struct sk_buff *buf, u32 dest, u32 selector) * small enough not to require fragmentation. * Called without any locks held. */ - -void tipc_link_send_names(struct list_head *message_list, u32 dest) +void tipc_link_names_xmit(struct list_head *message_list, u32 dest) { struct tipc_node *n_ptr; struct tipc_link *l_ptr; @@ -997,7 +910,6 @@ void tipc_link_send_names(struct list_head *message_list, u32 dest) if (list_empty(message_list)) return; - read_lock_bh(&tipc_net_lock); n_ptr = tipc_node_find(dest); if (n_ptr) { tipc_node_lock(n_ptr); @@ -1012,94 +924,49 @@ void tipc_link_send_names(struct list_head *message_list, u32 dest) } tipc_node_unlock(n_ptr); } - read_unlock_bh(&tipc_net_lock); /* discard the messages if they couldn't be sent */ - list_for_each_safe(buf, temp_buf, ((struct sk_buff *)message_list)) { list_del((struct list_head *)buf); - buf_discard(buf); + kfree_skb(buf); } } /* - * link_send_buf_fast: Entry for data messages where the + * tipc_link_xmit_fast: Entry for data messages where the * destination link is known and the header is complete, * inclusive total message length. Very time critical. * Link is locked. Returns user data length. */ - -static int link_send_buf_fast(struct tipc_link *l_ptr, struct sk_buff *buf, - u32 *used_max_pkt) +static int tipc_link_xmit_fast(struct tipc_link *l_ptr, struct sk_buff *buf, + u32 *used_max_pkt) { struct tipc_msg *msg = buf_msg(buf); int res = msg_data_sz(msg); 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))) { - 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; - return res; - } - } else - *used_max_pkt = l_ptr->max_pkt; - } - return tipc_link_send_buf(l_ptr, buf); /* All other cases */ -} - -/* - * 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); + link_add_to_outqueue(l_ptr, buf, msg); + tipc_bearer_send(l_ptr->bearer_id, buf, + &l_ptr->media_addr); + l_ptr->unacked_window = 0; return res; } - tipc_node_unlock(n_ptr); + else + *used_max_pkt = l_ptr->max_pkt; } - read_unlock_bh(&tipc_net_lock); - res = msg_data_sz(buf_msg(buf)); - tipc_reject_msg(buf, TIPC_ERR_NO_NODE); - return res; + return __tipc_link_xmit(l_ptr, buf); /* All other cases */ } - /* - * tipc_link_send_sections_fast: Entry for messages where the + * tipc_link_iovec_xmit_fast: Entry for messages where the * destination processor is known and the header is complete, * except for total message length. * Returns user data length or errno. */ -int tipc_link_send_sections_fast(struct tipc_port *sender, - struct iovec const *msg_sect, - const u32 num_sect, - unsigned int total_len, - u32 destaddr) +int tipc_link_iovec_xmit_fast(struct tipc_port *sender, + struct iovec const *msg_sect, + unsigned int len, u32 destaddr) { struct tipc_msg *hdr = &sender->phdr; struct tipc_link *l_ptr; @@ -1113,34 +980,26 @@ again: * Try building message using port's max_pkt hint. * (Must not hold any locks while building message.) */ + res = tipc_msg_build(hdr, msg_sect, len, sender->max_pkt, &buf); + /* Exit if build request was invalid */ + if (unlikely(res < 0)) + return res; - res = tipc_msg_build(hdr, msg_sect, num_sect, total_len, - sender->max_pkt, !sender->user_port, &buf); - - read_lock_bh(&tipc_net_lock); node = tipc_node_find(destaddr); if (likely(node)) { tipc_node_lock(node); l_ptr = node->active_links[selector]; if (likely(l_ptr)) { if (likely(buf)) { - res = link_send_buf_fast(l_ptr, buf, - &sender->max_pkt); + res = tipc_link_xmit_fast(l_ptr, buf, + &sender->max_pkt); exit: tipc_node_unlock(node); - read_unlock_bh(&tipc_net_lock); 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)) { + if (link_congested(l_ptr)) { res = link_schedule_port(l_ptr, sender->ref, res); goto exit; @@ -1150,35 +1009,27 @@ exit: * Message size exceeds max_pkt hint; update hint, * then re-try fast path or fragment the message */ - sender->max_pkt = l_ptr->max_pkt; tipc_node_unlock(node); - read_unlock_bh(&tipc_net_lock); if ((msg_hdr_sz(hdr) + res) <= sender->max_pkt) goto again; - return link_send_sections_long(sender, msg_sect, - num_sect, total_len, - destaddr); + return tipc_link_iovec_long_xmit(sender, msg_sect, + len, destaddr); } tipc_node_unlock(node); } - read_unlock_bh(&tipc_net_lock); /* Couldn't find a link to the destination node */ - - if (buf) - return tipc_reject_msg(buf, TIPC_ERR_NO_NODE); - if (res >= 0) - return tipc_port_reject_sections(sender, hdr, msg_sect, num_sect, - total_len, TIPC_ERR_NO_NODE); - return res; + kfree_skb(buf); + tipc_port_iovec_reject(sender, hdr, msg_sect, len, TIPC_ERR_NO_NODE); + return -ENETUNREACH; } /* - * link_send_sections_long(): Entry for long messages where the + * tipc_link_iovec_long_xmit(): Entry for long messages where the * destination node is known and the header is complete, * inclusive total message length. * Link and bearer congestion status have been checked to be ok, @@ -1191,23 +1042,22 @@ exit: * * Returns user data length or errno. */ -static int link_send_sections_long(struct tipc_port *sender, - struct iovec const *msg_sect, - u32 num_sect, - unsigned int total_len, - u32 destaddr) +static int tipc_link_iovec_long_xmit(struct tipc_port *sender, + struct iovec const *msg_sect, + unsigned int len, u32 destaddr) { struct tipc_link *l_ptr; struct tipc_node *node; struct tipc_msg *hdr = &sender->phdr; - u32 dsz = total_len; + u32 dsz = len; u32 max_pkt, fragm_sz, rest; struct tipc_msg fragm_hdr; struct sk_buff *buf, *buf_chain, *prev; u32 fragm_crs, fragm_rest, hsz, sect_rest; - const unchar *sect_crs; + const unchar __user *sect_crs; int curr_sect; u32 fragm_no; + int res = 0; again: fragm_no = 1; @@ -1222,15 +1072,13 @@ again: sect_crs = NULL; curr_sect = -1; - /* Prepare reusable fragment header: */ - + /* Prepare reusable fragment header */ tipc_msg_init(&fragm_hdr, MSG_FRAGMENTER, FIRST_FRAGMENT, INT_H_SIZE, msg_destnode(hdr)); msg_set_size(&fragm_hdr, max_pkt); msg_set_fragm_no(&fragm_hdr, 1); - /* Prepare header of first fragment: */ - + /* Prepare header of first fragment */ buf_chain = buf = tipc_buf_acquire(max_pkt); if (!buf) return -ENOMEM; @@ -1239,8 +1087,7 @@ again: hsz = msg_hdr_sz(hdr); skb_copy_to_linear_data_offset(buf, INT_H_SIZE, hdr, hsz); - /* Chop up message: */ - + /* Chop up message */ fragm_crs = INT_H_SIZE + hsz; fragm_rest = fragm_sz - hsz; @@ -1249,7 +1096,7 @@ again: if (!sect_rest) { sect_rest = msg_sect[++curr_sect].iov_len; - sect_crs = (const unchar *)msg_sect[curr_sect].iov_base; + sect_crs = msg_sect[curr_sect].iov_base; } if (sect_rest < fragm_rest) @@ -1257,18 +1104,12 @@ 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; - buf_discard(buf_chain); - } - return -EFAULT; - } - } else - skb_copy_to_linear_data_offset(buf, fragm_crs, - sect_crs, sz); + kfree_skb_list(buf_chain); + return res; + } sect_crs += sz; sect_rest -= sz; fragm_crs += sz; @@ -1288,8 +1129,10 @@ 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; @@ -1314,24 +1157,18 @@ error: if (l_ptr->max_pkt < max_pkt) { sender->max_pkt = l_ptr->max_pkt; tipc_node_unlock(node); - for (; buf_chain; buf_chain = buf) { - buf = buf_chain->next; - buf_discard(buf_chain); - } + kfree_skb_list(buf_chain); goto again; } } else { reject: - for (; buf_chain; buf_chain = buf) { - buf = buf_chain->next; - buf_discard(buf_chain); - } - return tipc_port_reject_sections(sender, hdr, msg_sect, num_sect, - total_len, TIPC_ERR_NO_NODE); + kfree_skb_list(buf_chain); + tipc_port_iovec_reject(sender, hdr, msg_sect, len, + TIPC_ERR_NO_NODE); + return -ENETUNREACH; } /* Append chain of fragments to send queue & send them */ - l_ptr->long_msg_seq_no++; link_add_chain_to_outqueue(l_ptr, buf_chain, l_ptr->long_msg_seq_no); l_ptr->stats.sent_fragments += fragm_no; @@ -1344,7 +1181,7 @@ reject: /* * tipc_link_push_packet: Push one unsent packet to the media */ -u32 tipc_link_push_packet(struct tipc_link *l_ptr) +static u32 tipc_link_push_packet(struct tipc_link *l_ptr) { struct sk_buff *buf = l_ptr->first_out; u32 r_q_size = l_ptr->retransm_queue_size; @@ -1352,7 +1189,6 @@ u32 tipc_link_push_packet(struct tipc_link *l_ptr) /* Step to position where retransmission failed, if any, */ /* consider that buffers may have been released in meantime */ - if (r_q_size && buf) { u32 last = lesser(mod(r_q_head + r_q_size), link_last_sent(l_ptr)); @@ -1367,40 +1203,29 @@ u32 tipc_link_push_packet(struct tipc_link *l_ptr) } /* Continue retransmission now, if there is anything: */ - 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->bearer_id, 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: */ - buf = l_ptr->proto_msg_queue; 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; - buf_discard(buf); - l_ptr->proto_msg_queue = NULL; - return 0; - } else { - l_ptr->stats.bearer_congs++; - return PUSH_FAILED; - } + tipc_bearer_send(l_ptr->bearer_id, 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: */ - buf = l_ptr->next_out; if (buf) { struct tipc_msg *msg = buf_msg(buf); @@ -1410,18 +1235,15 @@ 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->bearer_id, 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; } /* @@ -1432,87 +1254,68 @@ void tipc_link_push_queue(struct tipc_link *l_ptr) { u32 res; - if (tipc_bearer_congested(l_ptr->b_ptr, l_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) +void tipc_link_reset_all(struct tipc_node *node) { - struct tipc_node *n_ptr; char addr_string[16]; u32 i; - read_lock_bh(&tipc_net_lock); - n_ptr = tipc_node_find((u32)addr); - if (!n_ptr) { - read_unlock_bh(&tipc_net_lock); - return; /* node no longer exists */ - } - - tipc_node_lock(n_ptr); + tipc_node_lock(node); - warn("Resetting all links to %s\n", - tipc_addr_string_fill(addr_string, n_ptr->addr)); + pr_warn("Resetting all links to %s\n", + tipc_addr_string_fill(addr_string, node->addr)); for (i = 0; i < MAX_BEARERS; i++) { - if (n_ptr->links[i]) { - link_print(n_ptr->links[i], "Resetting link\n"); - tipc_link_reset(n_ptr->links[i]); + if (node->links[i]) { + link_print(node->links[i], "Resetting link\n"); + tipc_link_reset(node->links[i]); } } - tipc_node_unlock(n_ptr); - read_unlock_bh(&tipc_net_lock); + tipc_node_unlock(node); } 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); - warn("Retransmission failure on link <%s>\n", l_ptr->name); + pr_warn("Retransmission failure on link <%s>\n", l_ptr->name); if (l_ptr->addr) { - /* Handle failure on standard link */ - link_print(l_ptr, "Resetting link\n"); tipc_link_reset(l_ptr); } else { - /* Handle failure on broadcast link */ - struct tipc_node *n_ptr; char addr_string[16]; - info("Msg seq number: %u, ", msg_seqno(msg)); - info("Outstanding acks: %lu\n", - (unsigned long) TIPC_SKB_CB(buf)->handle); + pr_info("Msg seq number: %u, ", msg_seqno(msg)); + pr_cont("Outstanding acks: %lu\n", + (unsigned long) TIPC_SKB_CB(buf)->handle); n_ptr = tipc_bclink_retransmit_to(); tipc_node_lock(n_ptr); tipc_addr_string_fill(addr_string, n_ptr->addr); - info("Multicast link info for %s\n", addr_string); - info("Supported: %d, ", n_ptr->bclink.supported); - info("Acked: %u\n", n_ptr->bclink.acked); - info("Last in: %u, ", n_ptr->bclink.last_in); - info("Gap after: %u, ", n_ptr->bclink.gap_after); - info("Gap to: %u\n", n_ptr->bclink.gap_to); - info("Nack sync: %u\n\n", n_ptr->bclink.nack_sync); - - tipc_k_signal((Handler)link_reset_all, (unsigned long)n_ptr->addr); + pr_info("Broadcast link info for %s\n", addr_string); + pr_info("Reception permitted: %d, Acked: %u\n", + n_ptr->bclink.recv_permitted, + n_ptr->bclink.acked); + pr_info("Last in: %u, Oos state: %u, Last sent: %u\n", + n_ptr->bclink.last_in, + n_ptr->bclink.oos_state, + n_ptr->bclink.last_sent); tipc_node_unlock(n_ptr); + tipc_bclink_set_flags(TIPC_BCLINK_RESET); l_ptr->stale_count = 0; } } @@ -1527,44 +1330,25 @@ 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 (l_ptr->retransm_queue_size == 0) { - l_ptr->retransm_queue_head = msg_seqno(msg); - l_ptr->retransm_queue_size = retransmits; - } else { - err("Unexpected retransmit on link %s (qsize=%d)\n", - l_ptr->name, l_ptr->retransm_queue_size); + /* Detect repeated retransmit failures */ + if (l_ptr->last_retransmitted == msg_seqno(msg)) { + if (++l_ptr->stale_count > 100) { + link_retransmit_failure(l_ptr, buf); + return; } - return; } else { - /* Detect repeated retransmit failures on uncongested bearer */ - - if (l_ptr->last_retransmitted == msg_seqno(msg)) { - if (++l_ptr->stale_count > 100) { - link_retransmit_failure(l_ptr, buf); - return; - } - } else { - l_ptr->last_retransmitted = msg_seqno(msg); - l_ptr->stale_count = 1; - } + l_ptr->last_retransmitted = msg_seqno(msg); + l_ptr->stale_count = 1; } while (retransmits && (buf != l_ptr->next_out) && 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->bearer_id, buf, &l_ptr->media_addr); + buf = buf->next; + retransmits--; + l_ptr->stats.retransmitted++; } l_ptr->retransm_queue_head = l_ptr->retransm_queue_size = 0; @@ -1573,7 +1357,6 @@ void tipc_link_retransmit(struct tipc_link *l_ptr, struct sk_buff *buf, /** * link_insert_deferred_queue - insert deferred messages back into receive chain */ - static struct sk_buff *link_insert_deferred_queue(struct tipc_link *l_ptr, struct sk_buff *buf) { @@ -1604,7 +1387,6 @@ static struct sk_buff *link_insert_deferred_queue(struct tipc_link *l_ptr, * TIPC will ignore the excess, under the assumption that it is optional info * introduced by a later release of the protocol. */ - static int link_recv_buf_validate(struct sk_buff *buf) { static u32 min_data_hdr_size[8] = { @@ -1618,6 +1400,12 @@ static int link_recv_buf_validate(struct sk_buff *buf) u32 hdr_size; u32 min_hdr_size; + /* If this packet comes from the defer queue, the skb has already + * been validated + */ + if (unlikely(TIPC_SKB_CB(buf)->deferred)) + return 1; + if (unlikely(buf->len < MIN_H_SIZE)) return 0; @@ -1643,17 +1431,15 @@ static int link_recv_buf_validate(struct sk_buff *buf) } /** - * tipc_recv_msg - process TIPC messages arriving from off-node + * tipc_rcv - process TIPC packets/messages arriving from off-node * @head: pointer to message buffer chain - * @tb_ptr: pointer to bearer message arrived on + * @b_ptr: pointer to bearer message arrived on * * Invoked with no locks held. Bearer pointer must point to a valid bearer * structure (i.e. cannot be NULL), but bearer can be inactive. */ - -void tipc_recv_msg(struct sk_buff *head, struct tipc_bearer *b_ptr) +void tipc_rcv(struct sk_buff *head, struct tipc_bearer *b_ptr) { - read_lock_bh(&tipc_net_lock); while (head) { struct tipc_node *n_ptr; struct tipc_link *l_ptr; @@ -1663,88 +1449,69 @@ void tipc_recv_msg(struct sk_buff *head, struct tipc_bearer *b_ptr) u32 seq_no; u32 ackd; u32 released = 0; - int type; head = head->next; - - /* Ensure bearer is still enabled */ - - if (unlikely(!b_ptr->active)) - goto cont; + buf->next = NULL; /* Ensure message is well-formed */ - if (unlikely(!link_recv_buf_validate(buf))) - goto cont; + goto discard; /* Ensure message data is a single contiguous unit */ - - if (unlikely(buf_linearize(buf))) - goto cont; + if (unlikely(skb_linearize(buf))) + goto discard; /* Handle arrival of a non-unicast link message */ - msg = buf_msg(buf); if (unlikely(msg_non_seq(msg))) { if (msg_user(msg) == LINK_CONFIG) - tipc_disc_recv_msg(buf, b_ptr); + tipc_disc_rcv(buf, b_ptr); else - tipc_bclink_recv_pkt(buf); + tipc_bclink_rcv(buf); continue; } /* Discard unicast link messages destined for another node */ - if (unlikely(!msg_short(msg) && (msg_destnode(msg) != tipc_own_addr))) - goto cont; + goto discard; /* Locate neighboring node that sent message */ - n_ptr = tipc_node_find(msg_prevnode(msg)); if (unlikely(!n_ptr)) - goto cont; + goto discard; tipc_node_lock(n_ptr); /* Locate unicast link endpoint that should handle message */ - l_ptr = n_ptr->links[b_ptr->identity]; - if (unlikely(!l_ptr)) { - tipc_node_unlock(n_ptr); - goto cont; - } + if (unlikely(!l_ptr)) + goto unlock_discard; /* Verify that communication with node is currently allowed */ + if ((n_ptr->action_flags & TIPC_WAIT_PEER_LINKS_DOWN) && + msg_user(msg) == LINK_PROTOCOL && + (msg_type(msg) == RESET_MSG || + msg_type(msg) == ACTIVATE_MSG) && + !msg_redundant_link(msg)) + n_ptr->action_flags &= ~TIPC_WAIT_PEER_LINKS_DOWN; - if ((n_ptr->block_setup & WAIT_PEER_DOWN) && - msg_user(msg) == LINK_PROTOCOL && - (msg_type(msg) == RESET_MSG || - msg_type(msg) == ACTIVATE_MSG) && - !msg_redundant_link(msg)) - n_ptr->block_setup &= ~WAIT_PEER_DOWN; - - if (n_ptr->block_setup) { - tipc_node_unlock(n_ptr); - goto cont; - } + if (tipc_node_blocked(n_ptr)) + goto unlock_discard; /* Validate message sequence number info */ - seq_no = msg_seqno(msg); ackd = msg_ack(msg); /* Release acked messages */ - - if (tipc_node_is_up(n_ptr) && n_ptr->bclink.supported) + if (n_ptr->bclink.recv_permitted) tipc_bclink_acknowledge(n_ptr, msg_bcast_ack(msg)); crs = l_ptr->first_out; while ((crs != l_ptr->next_out) && less_eq(buf_seqno(crs), ackd)) { struct sk_buff *next = crs->next; - - buf_discard(crs); + kfree_skb(crs); crs = next; released++; } @@ -1754,116 +1521,119 @@ void tipc_recv_msg(struct sk_buff *head, struct tipc_bearer *b_ptr) } /* Try sending any messages link endpoint has pending */ - if (unlikely(l_ptr->next_out)) tipc_link_push_queue(l_ptr); + if (unlikely(!list_empty(&l_ptr->waiting_ports))) tipc_link_wakeup_ports(l_ptr, 0); + if (unlikely(++l_ptr->unacked_window >= TIPC_MIN_LINK_WIN)) { l_ptr->stats.sent_acks++; - tipc_link_send_proto_msg(l_ptr, STATE_MSG, 0, 0, 0, 0, 0); + tipc_link_proto_xmit(l_ptr, STATE_MSG, 0, 0, 0, 0, 0); } - /* Now (finally!) process the incoming message */ - -protocol_check: - if (likely(link_working_working(l_ptr))) { - if (likely(seq_no == mod(l_ptr->next_in_no))) { - l_ptr->next_in_no++; - if (unlikely(l_ptr->oldest_deferred_in)) - head = link_insert_deferred_queue(l_ptr, - head); - if (likely(msg_is_dest(msg, tipc_own_addr))) { -deliver: - if (likely(msg_isdata(msg))) { - tipc_node_unlock(n_ptr); - tipc_port_recv_msg(buf); - continue; - } - switch (msg_user(msg)) { - case MSG_BUNDLER: - l_ptr->stats.recv_bundles++; - l_ptr->stats.recv_bundled += - msg_msgcnt(msg); - tipc_node_unlock(n_ptr); - tipc_link_recv_bundle(buf); - continue; - case NAME_DISTRIBUTOR: - tipc_node_unlock(n_ptr); - tipc_named_recv(buf); - continue; - case CONN_MANAGER: - tipc_node_unlock(n_ptr); - tipc_port_recv_proto_msg(buf); - continue; - case MSG_FRAGMENTER: - l_ptr->stats.recv_fragments++; - if (tipc_link_recv_fragment(&l_ptr->defragm_buf, - &buf, &msg)) { - l_ptr->stats.recv_fragmented++; - goto deliver; - } - break; - case CHANGEOVER_PROTOCOL: - type = msg_type(msg); - if (link_recv_changeover_msg(&l_ptr, &buf)) { - msg = buf_msg(buf); - seq_no = msg_seqno(msg); - if (type == ORIGINAL_MSG) - goto deliver; - goto protocol_check; - } - break; - default: - buf_discard(buf); - buf = NULL; - break; - } - } + /* Process the incoming packet */ + if (unlikely(!link_working_working(l_ptr))) { + if (msg_user(msg) == LINK_PROTOCOL) { + tipc_link_proto_rcv(l_ptr, buf); + head = link_insert_deferred_queue(l_ptr, head); tipc_node_unlock(n_ptr); - tipc_net_route_msg(buf); continue; } + + /* Traffic message. Conditionally activate link */ + link_state_event(l_ptr, TRAFFIC_MSG_EVT); + + if (link_working_working(l_ptr)) { + /* Re-insert buffer in front of queue */ + buf->next = head; + head = buf; + tipc_node_unlock(n_ptr); + continue; + } + goto unlock_discard; + } + + /* Link is now in state WORKING_WORKING */ + if (unlikely(seq_no != mod(l_ptr->next_in_no))) { link_handle_out_of_seq_msg(l_ptr, buf); head = link_insert_deferred_queue(l_ptr, head); tipc_node_unlock(n_ptr); continue; } - - if (msg_user(msg) == LINK_PROTOCOL) { - link_recv_proto_msg(l_ptr, buf); + l_ptr->next_in_no++; + if (unlikely(l_ptr->oldest_deferred_in)) head = link_insert_deferred_queue(l_ptr, head); - tipc_node_unlock(n_ptr); - continue; + + /* Deliver packet/message to correct user: */ + if (unlikely(msg_user(msg) == CHANGEOVER_PROTOCOL)) { + if (!tipc_link_tunnel_rcv(n_ptr, &buf)) { + tipc_node_unlock(n_ptr); + continue; + } + msg = buf_msg(buf); + } else if (msg_user(msg) == MSG_FRAGMENTER) { + l_ptr->stats.recv_fragments++; + if (tipc_buf_append(&l_ptr->reasm_buf, &buf)) { + l_ptr->stats.recv_fragmented++; + msg = buf_msg(buf); + } else { + if (!l_ptr->reasm_buf) + tipc_link_reset(l_ptr); + tipc_node_unlock(n_ptr); + continue; + } } - link_state_event(l_ptr, TRAFFIC_MSG_EVT); - if (link_working_working(l_ptr)) { - /* Re-insert in front of queue */ - buf->next = head; - head = buf; + switch (msg_user(msg)) { + case TIPC_LOW_IMPORTANCE: + case TIPC_MEDIUM_IMPORTANCE: + case TIPC_HIGH_IMPORTANCE: + case TIPC_CRITICAL_IMPORTANCE: + tipc_node_unlock(n_ptr); + tipc_sk_rcv(buf); + continue; + case MSG_BUNDLER: + l_ptr->stats.recv_bundles++; + l_ptr->stats.recv_bundled += msg_msgcnt(msg); + tipc_node_unlock(n_ptr); + tipc_link_bundle_rcv(buf); + continue; + case NAME_DISTRIBUTOR: + n_ptr->bclink.recv_permitted = true; + tipc_node_unlock(n_ptr); + tipc_named_rcv(buf); + continue; + case CONN_MANAGER: tipc_node_unlock(n_ptr); + tipc_port_proto_rcv(buf); continue; + case BCAST_PROTOCOL: + tipc_link_sync_rcv(n_ptr, buf); + break; + default: + kfree_skb(buf); + break; } tipc_node_unlock(n_ptr); -cont: - buf_discard(buf); + continue; +unlock_discard: + tipc_node_unlock(n_ptr); +discard: + kfree_skb(buf); } - read_unlock_bh(&tipc_net_lock); } -/* - * link_defer_buf(): Sort a received out-of-sequence packet - * into the deferred reception queue. - * Returns the increase of the queue length,i.e. 0 or 1 +/** + * tipc_link_defer_pkt - Add out-of-sequence message to deferred reception queue + * + * Returns increase in queue length (i.e. 0 or 1) */ - -u32 tipc_link_defer_pkt(struct sk_buff **head, - struct sk_buff **tail, +u32 tipc_link_defer_pkt(struct sk_buff **head, struct sk_buff **tail, struct sk_buff *buf) { - struct sk_buff *prev = NULL; - struct sk_buff *crs = *head; + struct sk_buff *queue_buf; + struct sk_buff **prev; u32 seq_no = buf_seqno(buf); buf->next = NULL; @@ -1881,56 +1651,52 @@ u32 tipc_link_defer_pkt(struct sk_buff **head, return 1; } - /* Scan through queue and sort it in */ - do { - struct tipc_msg *msg = buf_msg(crs); + /* Locate insertion point in queue, then insert; discard if duplicate */ + prev = head; + queue_buf = *head; + for (;;) { + u32 curr_seqno = buf_seqno(queue_buf); - if (less(seq_no, msg_seqno(msg))) { - buf->next = crs; - if (prev) - prev->next = buf; - else - *head = buf; - return 1; + if (seq_no == curr_seqno) { + kfree_skb(buf); + return 0; } - if (seq_no == msg_seqno(msg)) + + if (less(seq_no, curr_seqno)) break; - prev = crs; - crs = crs->next; - } while (crs); - /* Message is a duplicate of an existing message */ + prev = &queue_buf->next; + queue_buf = queue_buf->next; + } - buf_discard(buf); - return 0; + buf->next = queue_buf; + *prev = buf; + return 1; } -/** +/* * link_handle_out_of_seq_msg - handle arrival of out-of-sequence packet */ - static void link_handle_out_of_seq_msg(struct tipc_link *l_ptr, struct sk_buff *buf) { u32 seq_no = buf_seqno(buf); if (likely(msg_user(buf_msg(buf)) == LINK_PROTOCOL)) { - link_recv_proto_msg(l_ptr, buf); + tipc_link_proto_rcv(l_ptr, buf); return; } /* Record OOS packet arrival (force mismatch on next timeout) */ - l_ptr->checkpoint--; /* * Discard packet if a duplicate; otherwise add it to deferred queue * and notify peer of gap as per protocol specification */ - if (less(seq_no, mod(l_ptr->next_in_no))) { l_ptr->stats.duplicates++; - buf_discard(buf); + kfree_skb(buf); return; } @@ -1938,8 +1704,9 @@ static void link_handle_out_of_seq_msg(struct tipc_link *l_ptr, &l_ptr->newest_deferred_in, buf)) { l_ptr->deferred_inqueue_sz++; l_ptr->stats.deferred_recv++; + TIPC_SKB_CB(buf)->deferred = true; if ((l_ptr->deferred_inqueue_sz % 16) == 1) - tipc_link_send_proto_msg(l_ptr, STATE_MSG, 0, 0, 0, 0, 0); + tipc_link_proto_xmit(l_ptr, STATE_MSG, 0, 0, 0, 0, 0); } else l_ptr->stats.duplicates++; } @@ -1947,26 +1714,32 @@ 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) +void tipc_link_proto_xmit(struct tipc_link *l_ptr, u32 msg_typ, int probe_msg, + u32 gap, u32 tolerance, u32 priority, u32 ack_mtu) { struct sk_buff *buf = NULL; struct tipc_msg *msg = l_ptr->pmsg; u32 msg_size = sizeof(l_ptr->proto_msg); int r_flag; - if (link_blocked(l_ptr)) + /* Discard any previous message that was deferred due to congestion */ + if (l_ptr->proto_msg_queue) { + kfree_skb(l_ptr->proto_msg_queue); + l_ptr->proto_msg_queue = NULL; + } + + /* Don't send protocol message during link changeover */ + if (l_ptr->exp_msg_count) return; /* Abort non-RESET send if communication with node is prohibited */ - - if ((l_ptr->owner->block_setup) && (msg_typ != RESET_MSG)) + if ((tipc_node_blocked(l_ptr->owner)) && (msg_typ != RESET_MSG)) return; + /* Create protocol message with "out-of-sequence" sequence number */ msg_set_type(msg, msg_typ); - msg_set_net_plane(msg, l_ptr->b_ptr->net_plane); - msg_set_bcast_ack(msg, mod(l_ptr->owner->bclink.last_in)); + msg_set_net_plane(msg, l_ptr->net_plane); + msg_set_bcast_ack(msg, l_ptr->owner->bclink.last_in); msg_set_last_bcast(msg, tipc_bclink_get_last_sent()); if (msg_typ == STATE_MSG) { @@ -2020,44 +1793,20 @@ void tipc_link_send_proto_msg(struct tipc_link *l_ptr, u32 msg_typ, r_flag = (l_ptr->owner->working_links > tipc_link_is_up(l_ptr)); msg_set_redundant_link(msg, r_flag); msg_set_linkprio(msg, l_ptr->priority); - - /* Ensure sequence number will not fit : */ + msg_set_size(msg, msg_size); msg_set_seqno(msg, mod(l_ptr->next_out_no + (0xffff/2))); - /* Congestion? */ - - if (tipc_bearer_congested(l_ptr->b_ptr, l_ptr)) { - if (!l_ptr->proto_msg_queue) { - l_ptr->proto_msg_queue = - tipc_buf_acquire(sizeof(l_ptr->proto_msg)); - } - buf = l_ptr->proto_msg_queue; - if (!buf) - return; - skb_copy_to_linear_data(buf, msg, sizeof(l_ptr->proto_msg)); - return; - } - - /* Message can be sent */ - buf = tipc_buf_acquire(msg_size); if (!buf) return; skb_copy_to_linear_data(buf, msg, sizeof(l_ptr->proto_msg)); - msg_set_size(buf_msg(buf), msg_size); + buf->priority = TC_PRIO_CONTROL; - if (tipc_bearer_send(l_ptr->b_ptr, buf, &l_ptr->media_addr)) { - l_ptr->unacked_window = 0; - buf_discard(buf); - return; - } - - /* New congestion */ - tipc_bearer_schedule(l_ptr->b_ptr, l_ptr); - l_ptr->proto_msg_queue = buf; - l_ptr->stats.bearer_congs++; + tipc_bearer_send(l_ptr->bearer_id, buf, &l_ptr->media_addr); + l_ptr->unacked_window = 0; + kfree_skb(buf); } /* @@ -2065,8 +1814,7 @@ void tipc_link_send_proto_msg(struct tipc_link *l_ptr, u32 msg_typ, * Note that network plane id propagates through the network, and may * change at any time. The node with lowest address rules */ - -static void link_recv_proto_msg(struct tipc_link *l_ptr, struct sk_buff *buf) +static void tipc_link_proto_rcv(struct tipc_link *l_ptr, struct sk_buff *buf) { u32 rec_gap = 0; u32 max_pkt_info; @@ -2074,18 +1822,13 @@ static void link_recv_proto_msg(struct tipc_link *l_ptr, struct sk_buff *buf) u32 msg_tol; struct tipc_msg *msg = buf_msg(buf); - if (link_blocked(l_ptr)) + /* Discard protocol message during link changeover */ + if (l_ptr->exp_msg_count) goto exit; - /* record unnumbered packet arrival (force mismatch on next timeout) */ - - l_ptr->checkpoint--; - - if (l_ptr->b_ptr->net_plane != msg_net_plane(msg)) + if (l_ptr->net_plane != msg_net_plane(msg)) if (tipc_own_addr > msg_prevnode(msg)) - l_ptr->b_ptr->net_plane = msg_net_plane(msg); - - l_ptr->owner->permit_changeover = msg_redundant_link(msg); + l_ptr->net_plane = msg_net_plane(msg); switch (msg_type(msg)) { @@ -2102,13 +1845,14 @@ static void link_recv_proto_msg(struct tipc_link *l_ptr, struct sk_buff *buf) * peer has lost contact -- don't allow peer's links * to reactivate before we recognize loss & clean up */ - l_ptr->owner->block_setup = WAIT_NODE_DOWN; + l_ptr->owner->action_flags |= TIPC_WAIT_OWN_LINKS_DOWN; } + link_state_event(l_ptr, RESET_MSG); + /* fall thru' */ case ACTIVATE_MSG: /* Update link settings according other endpoint's values */ - strcpy((strrchr(l_ptr->name, ':') + 1), (char *)msg_data(msg)); msg_tol = msg_link_tolerance(msg); @@ -2127,16 +1871,20 @@ 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.supported = (max_pkt_info != 0); - link_state_event(l_ptr, msg_type(msg)); + /* Synchronize broadcast link info, if not done previously */ + if (!tipc_node_is_up(l_ptr->owner)) { + l_ptr->owner->bclink.last_sent = + l_ptr->owner->bclink.last_in = + msg_last_bcast(msg); + l_ptr->owner->bclink.oos_state = 0; + } l_ptr->peer_session = msg_session(msg); l_ptr->peer_bearer_id = msg_bearer_id(msg); - /* Synchronize broadcast sequence numbers */ - if (!tipc_node_redundant_links(l_ptr->owner)) - l_ptr->owner->bclink.last_in = mod(msg_last_bcast(msg)); + if (msg_type(msg) == ACTIVATE_MSG) + link_state_event(l_ptr, ACTIVATE_MSG); break; case STATE_MSG: @@ -2146,12 +1894,17 @@ static void link_recv_proto_msg(struct tipc_link *l_ptr, struct sk_buff *buf) if (msg_linkprio(msg) && (msg_linkprio(msg) != l_ptr->priority)) { - warn("Resetting link <%s>, priority change %u->%u\n", - l_ptr->name, l_ptr->priority, msg_linkprio(msg)); + pr_warn("%s<%s>, priority change %u->%u\n", + link_rst_msg, l_ptr->name, l_ptr->priority, + msg_linkprio(msg)); l_ptr->priority = msg_linkprio(msg); tipc_link_reset(l_ptr); /* Enforce change to take effect */ break; } + + /* Record reception; force mismatch at next timeout: */ + l_ptr->checkpoint--; + link_state_event(l_ptr, TRAFFIC_MSG_EVT); l_ptr->stats.recv_states++; if (link_reset_unknown(l_ptr)) @@ -2176,12 +1929,13 @@ static void link_recv_proto_msg(struct tipc_link *l_ptr, struct sk_buff *buf) } /* Protocol message before retransmits, reduce loss risk */ - - tipc_bclink_check_gap(l_ptr->owner, msg_last_bcast(msg)); + if (l_ptr->owner->bclink.recv_permitted) + tipc_bclink_update_link_state(l_ptr->owner, + msg_last_bcast(msg)); if (rec_gap || (msg_probe(msg))) { - tipc_link_send_proto_msg(l_ptr, STATE_MSG, - 0, rec_gap, 0, 0, max_pkt_ack); + tipc_link_proto_xmit(l_ptr, STATE_MSG, 0, rec_gap, 0, + 0, max_pkt_ack); } if (msg_seq_gap(msg)) { l_ptr->stats.recv_nacks++; @@ -2191,18 +1945,17 @@ static void link_recv_proto_msg(struct tipc_link *l_ptr, struct sk_buff *buf) break; } exit: - buf_discard(buf); + kfree_skb(buf); } -/* - * tipc_link_tunnel(): Send one message via a link belonging to - * another bearer. Owner node is locked. +/* tipc_link_tunnel_xmit(): Tunnel one packet via a link belonging to + * a different bearer. Owner node is locked. */ -static void tipc_link_tunnel(struct tipc_link *l_ptr, - struct tipc_msg *tunnel_hdr, - struct tipc_msg *msg, - u32 selector) +static void tipc_link_tunnel_xmit(struct tipc_link *l_ptr, + struct tipc_msg *tunnel_hdr, + struct tipc_msg *msg, + u32 selector) { struct tipc_link *tunnel; struct sk_buff *buf; @@ -2210,30 +1963,28 @@ static void tipc_link_tunnel(struct tipc_link *l_ptr, tunnel = l_ptr->owner->active_links[selector & 1]; if (!tipc_link_is_up(tunnel)) { - warn("Link changeover error, " - "tunnel link no longer available\n"); + pr_warn("%stunnel link no longer available\n", link_co_err); return; } msg_set_size(tunnel_hdr, length + INT_H_SIZE); buf = tipc_buf_acquire(length + INT_H_SIZE); if (!buf) { - warn("Link changeover error, " - "unable to send tunnel msg\n"); + pr_warn("%sunable to send tunnel msg\n", link_co_err); return; } skb_copy_to_linear_data(buf, tunnel_hdr, INT_H_SIZE); skb_copy_to_linear_data_offset(buf, INT_H_SIZE, msg, length); - tipc_link_send_buf(tunnel, buf); + __tipc_link_xmit(tunnel, buf); } - -/* - * changeover(): Send whole message queue via the remaining link - * Owner node is locked. +/* tipc_link_failover_send_queue(): A link has gone down, but a second + * link is still active. We can do failover. Tunnel the failing link's + * whole send queue via the remaining link. This way, we don't lose + * any packets, and sequence order is preserved for subsequent traffic + * sent over the remaining link. Owner node is locked. */ - -void tipc_link_changeover(struct tipc_link *l_ptr) +void tipc_link_failover_send_queue(struct tipc_link *l_ptr) { u32 msgcount = l_ptr->out_queue_size; struct sk_buff *crs = l_ptr->first_out; @@ -2244,12 +1995,6 @@ void tipc_link_changeover(struct tipc_link *l_ptr) if (!tunnel) return; - if (!l_ptr->owner->permit_changeover) { - warn("Link changeover error, " - "peer did not permit changeover\n"); - return; - } - tipc_msg_init(&tunnel_hdr, CHANGEOVER_PROTOCOL, ORIGINAL_MSG, INT_H_SIZE, l_ptr->addr); msg_set_bearer_id(&tunnel_hdr, l_ptr->peer_bearer_id); @@ -2262,10 +2007,10 @@ void tipc_link_changeover(struct tipc_link *l_ptr) if (buf) { skb_copy_to_linear_data(buf, &tunnel_hdr, INT_H_SIZE); msg_set_size(&tunnel_hdr, INT_H_SIZE); - tipc_link_send_buf(tunnel, buf); + __tipc_link_xmit(tunnel, buf); } else { - warn("Link changeover error, " - "unable to send changeover msg\n"); + pr_warn("%sunable to send changeover msg\n", + link_co_err); } return; } @@ -2283,20 +2028,30 @@ void tipc_link_changeover(struct tipc_link *l_ptr) msgcount = msg_msgcnt(msg); while (msgcount--) { msg_set_seqno(m, msg_seqno(msg)); - tipc_link_tunnel(l_ptr, &tunnel_hdr, m, - msg_link_selector(m)); + tipc_link_tunnel_xmit(l_ptr, &tunnel_hdr, m, + msg_link_selector(m)); pos += align(msg_size(m)); m = (struct tipc_msg *)pos; } } else { - tipc_link_tunnel(l_ptr, &tunnel_hdr, msg, - msg_link_selector(msg)); + tipc_link_tunnel_xmit(l_ptr, &tunnel_hdr, msg, + msg_link_selector(msg)); } crs = crs->next; } } -void tipc_link_send_duplicate(struct tipc_link *l_ptr, struct tipc_link *tunnel) +/* tipc_link_dup_queue_xmit(): A second link has become active. Tunnel a + * duplicate of the first link's send queue via the new link. This way, we + * are guaranteed that currently queued packets from a socket are delivered + * before future traffic from the same socket, even if this is using the + * new link. The last arriving copy of each duplicate packet is dropped at + * the receiving end by the regular protocol check, so packet cardinality + * and sequence order is preserved per sender/receiver socket pair. + * Owner node is locked. + */ +void tipc_link_dup_queue_xmit(struct tipc_link *l_ptr, + struct tipc_link *tunnel) { struct sk_buff *iter; struct tipc_msg tunnel_hdr; @@ -2318,22 +2073,20 @@ void tipc_link_send_duplicate(struct tipc_link *l_ptr, struct tipc_link *tunnel) msg_set_size(&tunnel_hdr, length + INT_H_SIZE); outbuf = tipc_buf_acquire(length + INT_H_SIZE); if (outbuf == NULL) { - warn("Link changeover error, " - "unable to send duplicate msg\n"); + pr_warn("%sunable to send duplicate msg\n", + link_co_err); return; } skb_copy_to_linear_data(outbuf, &tunnel_hdr, INT_H_SIZE); skb_copy_to_linear_data_offset(outbuf, INT_H_SIZE, iter->data, length); - tipc_link_send_buf(tunnel, outbuf); + __tipc_link_xmit(tunnel, outbuf); if (!tipc_link_is_up(l_ptr)) return; iter = iter->next; } } - - /** * buf_extract - extracts embedded TIPC message from another message * @skb: encapsulating message buffer @@ -2342,7 +2095,6 @@ void tipc_link_send_duplicate(struct tipc_link *l_ptr, struct tipc_link *tunnel) * Returns a new message buffer containing an embedded message. The * encapsulating message itself is left unchanged. */ - static struct sk_buff *buf_extract(struct sk_buff *skb, u32 from_pos) { struct tipc_msg *msg = (struct tipc_msg *)(skb->data + from_pos); @@ -2355,88 +2107,112 @@ static struct sk_buff *buf_extract(struct sk_buff *skb, u32 from_pos) return eb; } -/* - * link_recv_changeover_msg(): Receive tunneled packet sent - * via other link. Node is locked. Return extracted buffer. + + +/* tipc_link_dup_rcv(): Receive a tunnelled DUPLICATE_MSG packet. + * Owner node is locked. */ +static void tipc_link_dup_rcv(struct tipc_link *l_ptr, + struct sk_buff *t_buf) +{ + struct sk_buff *buf; + + if (!tipc_link_is_up(l_ptr)) + return; + + buf = buf_extract(t_buf, INT_H_SIZE); + if (buf == NULL) { + pr_warn("%sfailed to extract inner dup pkt\n", link_co_err); + return; + } -static int link_recv_changeover_msg(struct tipc_link **l_ptr, - struct sk_buff **buf) + /* Add buffer to deferred queue, if applicable: */ + link_handle_out_of_seq_msg(l_ptr, buf); +} + +/* tipc_link_failover_rcv(): Receive a tunnelled ORIGINAL_MSG packet + * Owner node is locked. + */ +static struct sk_buff *tipc_link_failover_rcv(struct tipc_link *l_ptr, + struct sk_buff *t_buf) { - struct sk_buff *tunnel_buf = *buf; - struct tipc_link *dest_link; + struct tipc_msg *t_msg = buf_msg(t_buf); + struct sk_buff *buf = NULL; struct tipc_msg *msg; - struct tipc_msg *tunnel_msg = buf_msg(tunnel_buf); - u32 msg_typ = msg_type(tunnel_msg); - u32 msg_count = msg_msgcnt(tunnel_msg); - dest_link = (*l_ptr)->owner->links[msg_bearer_id(tunnel_msg)]; - if (!dest_link) - goto exit; - if (dest_link == *l_ptr) { - err("Unexpected changeover message on link <%s>\n", - (*l_ptr)->name); - goto exit; - } - *l_ptr = dest_link; - msg = msg_get_wrapped(tunnel_msg); + if (tipc_link_is_up(l_ptr)) + tipc_link_reset(l_ptr); - if (msg_typ == DUPLICATE_MSG) { - if (less(msg_seqno(msg), mod(dest_link->next_in_no))) - goto exit; - *buf = buf_extract(tunnel_buf, INT_H_SIZE); - if (*buf == NULL) { - warn("Link changeover error, duplicate msg dropped\n"); + /* First failover packet? */ + if (l_ptr->exp_msg_count == START_CHANGEOVER) + l_ptr->exp_msg_count = msg_msgcnt(t_msg); + + /* Should there be an inner packet? */ + if (l_ptr->exp_msg_count) { + l_ptr->exp_msg_count--; + buf = buf_extract(t_buf, INT_H_SIZE); + if (buf == NULL) { + pr_warn("%sno inner failover pkt\n", link_co_err); goto exit; } - buf_discard(tunnel_buf); - return 1; - } - - /* First original message ?: */ + msg = buf_msg(buf); - if (tipc_link_is_up(dest_link)) { - info("Resetting link <%s>, changeover initiated by peer\n", - dest_link->name); - tipc_link_reset(dest_link); - dest_link->exp_msg_count = msg_count; - if (!msg_count) - goto exit; - } else if (dest_link->exp_msg_count == START_CHANGEOVER) { - dest_link->exp_msg_count = msg_count; - if (!msg_count) + if (less(msg_seqno(msg), l_ptr->reset_checkpoint)) { + kfree_skb(buf); + buf = NULL; goto exit; + } + if (msg_user(msg) == MSG_FRAGMENTER) { + l_ptr->stats.recv_fragments++; + tipc_buf_append(&l_ptr->reasm_buf, &buf); + } + } +exit: + if ((l_ptr->exp_msg_count == 0) && (l_ptr->flags & LINK_STOPPED)) { + tipc_node_detach_link(l_ptr->owner, l_ptr); + kfree(l_ptr); } + return buf; +} - /* Receive original message */ +/* tipc_link_tunnel_rcv(): Receive a tunnelled packet, sent + * via other link as result of a failover (ORIGINAL_MSG) or + * a new active link (DUPLICATE_MSG). Failover packets are + * returned to the active link for delivery upwards. + * Owner node is locked. + */ +static int tipc_link_tunnel_rcv(struct tipc_node *n_ptr, + struct sk_buff **buf) +{ + struct sk_buff *t_buf = *buf; + struct tipc_link *l_ptr; + struct tipc_msg *t_msg = buf_msg(t_buf); + u32 bearer_id = msg_bearer_id(t_msg); - if (dest_link->exp_msg_count == 0) { - warn("Link switchover error, " - "got too many tunnelled messages\n"); + *buf = NULL; + + if (bearer_id >= MAX_BEARERS) goto exit; - } - dest_link->exp_msg_count--; - if (less(msg_seqno(msg), dest_link->reset_checkpoint)) { + + l_ptr = n_ptr->links[bearer_id]; + if (!l_ptr) goto exit; - } else { - *buf = buf_extract(tunnel_buf, INT_H_SIZE); - if (*buf != NULL) { - buf_discard(tunnel_buf); - return 1; - } else { - warn("Link changeover error, original msg dropped\n"); - } - } + + if (msg_type(t_msg) == DUPLICATE_MSG) + tipc_link_dup_rcv(l_ptr, t_buf); + else if (msg_type(t_msg) == ORIGINAL_MSG) + *buf = tipc_link_failover_rcv(l_ptr, t_buf); + else + pr_warn("%sunknown tunnel pkt received\n", link_co_err); exit: - *buf = NULL; - buf_discard(tunnel_buf); - return 0; + kfree_skb(t_buf); + return *buf != NULL; } /* * Bundler functionality: */ -void tipc_link_recv_bundle(struct sk_buff *buf) +void tipc_link_bundle_rcv(struct sk_buff *buf) { u32 msgcount = msg_msgcnt(buf_msg(buf)); u32 pos = INT_H_SIZE; @@ -2445,26 +2221,25 @@ void tipc_link_recv_bundle(struct sk_buff *buf) while (msgcount--) { obuf = buf_extract(buf, pos); if (obuf == NULL) { - warn("Link unable to unbundle message(s)\n"); + pr_warn("Link unable to unbundle message(s)\n"); break; } pos += align(msg_size(buf_msg(obuf))); tipc_net_route_msg(obuf); } - buf_discard(buf); + kfree_skb(buf); } /* * Fragmentation/defragmentation: */ - /* - * link_send_long_buf: Entry for buffers needing fragmentation. + * tipc_link_frag_xmit: Entry for buffers needing fragmentation. * The buffer is complete, inclusive total message length. * Returns user data length. */ -static int link_send_long_buf(struct tipc_link *l_ptr, struct sk_buff *buf) +static int tipc_link_frag_xmit(struct tipc_link *l_ptr, struct sk_buff *buf) { struct sk_buff *buf_chain = NULL; struct sk_buff *buf_chain_tail = (struct sk_buff *)&buf_chain; @@ -2485,12 +2260,10 @@ static int link_send_long_buf(struct tipc_link *l_ptr, struct sk_buff *buf) destaddr = msg_destnode(inmsg); /* Prepare reusable fragment header: */ - tipc_msg_init(&fragm_hdr, MSG_FRAGMENTER, FIRST_FRAGMENT, INT_H_SIZE, destaddr); /* Chop up message: */ - while (rest > 0) { struct sk_buff *fragm; @@ -2500,12 +2273,8 @@ static int link_send_long_buf(struct tipc_link *l_ptr, struct sk_buff *buf) } fragm = tipc_buf_acquire(fragm_sz + INT_H_SIZE); if (fragm == NULL) { - buf_discard(buf); - while (buf_chain) { - buf = buf_chain; - buf_chain = buf_chain->next; - buf_discard(buf); - } + kfree_skb(buf); + kfree_skb_list(buf_chain); return -ENOMEM; } msg_set_size(&fragm_hdr, fragm_sz + INT_H_SIZE); @@ -2521,10 +2290,9 @@ static int link_send_long_buf(struct tipc_link *l_ptr, struct sk_buff *buf) crs += fragm_sz; msg_set_type(&fragm_hdr, FRAGMENT); } - buf_discard(buf); + kfree_skb(buf); /* Append chain of fragments to send queue & send them */ - l_ptr->long_msg_seq_no++; link_add_chain_to_outqueue(l_ptr, buf_chain, l_ptr->long_msg_seq_no); l_ptr->stats.sent_fragments += fragm_no; @@ -2534,162 +2302,6 @@ static int link_send_long_buf(struct tipc_link *l_ptr, struct sk_buff *buf) return dsz; } -/* - * A pending message being re-assembled must store certain values - * to handle subsequent fragments correctly. The following functions - * help storing these values in unused, available fields in the - * pending message. This makes dynamic memory allocation unnecessary. - */ - -static void set_long_msg_seqno(struct sk_buff *buf, u32 seqno) -{ - msg_set_seqno(buf_msg(buf), seqno); -} - -static u32 get_fragm_size(struct sk_buff *buf) -{ - return msg_ack(buf_msg(buf)); -} - -static void set_fragm_size(struct sk_buff *buf, u32 sz) -{ - msg_set_ack(buf_msg(buf), sz); -} - -static u32 get_expected_frags(struct sk_buff *buf) -{ - return msg_bcast_ack(buf_msg(buf)); -} - -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. - */ -int tipc_link_recv_fragment(struct sk_buff **pending, struct sk_buff **fb, - struct tipc_msg **m) -{ - struct sk_buff *prev = NULL; - struct sk_buff *fbuf = *fb; - struct tipc_msg *fragm = buf_msg(fbuf); - struct sk_buff *pbuf = *pending; - u32 long_msg_seq_no = msg_long_msgno(fragm); - - *fb = NULL; - - /* Is there an incomplete message waiting for this fragment? */ - - while (pbuf && ((buf_seqno(pbuf) != long_msg_seq_no) || - (msg_orignode(fragm) != msg_orignode(buf_msg(pbuf))))) { - prev = pbuf; - pbuf = pbuf->next; - } - - if (!pbuf && (msg_type(fragm) == FIRST_FRAGMENT)) { - 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 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) { - buf_discard(fbuf); - return 0; - } - pbuf = tipc_buf_acquire(msg_size(imsg)); - if (pbuf != NULL) { - pbuf->next = *pending; - *pending = pbuf; - skb_copy_to_linear_data(pbuf, imsg, - msg_data_sz(fragm)); - /* Prepare buffer for subsequent fragments. */ - - set_long_msg_seqno(pbuf, long_msg_seq_no); - set_fragm_size(pbuf, fragm_sz); - set_expected_frags(pbuf, exp_fragm_cnt - 1); - } else { - warn("Link unable to reassemble fragmented message\n"); - } - buf_discard(fbuf); - return 0; - } else if (pbuf && (msg_type(fragm) != FIRST_FRAGMENT)) { - u32 dsz = msg_data_sz(fragm); - u32 fsz = get_fragm_size(pbuf); - u32 crs = ((msg_fragm_no(fragm) - 1) * fsz); - u32 exp_frags = get_expected_frags(pbuf) - 1; - skb_copy_to_linear_data_offset(pbuf, crs, - msg_data(fragm), dsz); - buf_discard(fbuf); - - /* Is message complete? */ - - if (exp_frags == 0) { - if (prev) - prev->next = pbuf->next; - else - *pending = pbuf->next; - msg_reset_reroute_cnt(buf_msg(pbuf)); - *fb = pbuf; - *m = buf_msg(pbuf); - return 1; - } - set_expected_frags(pbuf, exp_frags); - return 0; - } - buf_discard(fbuf); - 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; - buf_discard(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)) @@ -2701,7 +2313,6 @@ static void link_set_supervision_props(struct tipc_link *l_ptr, u32 tolerance) l_ptr->abort_limit = tolerance / (l_ptr->continuity_interval / 4); } - void tipc_link_set_queue_limits(struct tipc_link *l_ptr, u32 window) { /* Data messages from this node, inclusive FIRST_FRAGM */ @@ -2721,51 +2332,49 @@ void tipc_link_set_queue_limits(struct tipc_link *l_ptr, u32 window) l_ptr->queue_limit[MSG_FRAGMENTER] = 4000; } -/** - * link_find_link - locate link by name - * @name - ptr to link name string - * @node - ptr to area to be filled with ptr to associated node - * - * Caller must hold 'tipc_net_lock' to ensure node and bearer are not deleted; - * this also prevents link deletion. +/* tipc_link_find_owner - locate owner node of link by link's name + * @name: pointer to link name string + * @bearer_id: pointer to index in 'node->links' array where the link was found. * - * Returns pointer to link (or 0 if invalid link name). + * Returns pointer to node owning the link, or 0 if no matching link is found. */ - -static struct tipc_link *link_find_link(const char *name, - struct tipc_node **node) +static struct tipc_node *tipc_link_find_owner(const char *link_name, + unsigned int *bearer_id) { - struct tipc_link_name link_name_parts; - struct tipc_bearer *b_ptr; struct tipc_link *l_ptr; + struct tipc_node *n_ptr; + struct tipc_node *found_node = 0; + int i; - if (!link_name_validate(name, &link_name_parts)) - return NULL; - - b_ptr = tipc_bearer_find_interface(link_name_parts.if_local); - if (!b_ptr) - return NULL; - - *node = tipc_node_find(link_name_parts.addr_peer); - if (!*node) - return NULL; - - l_ptr = (*node)->links[b_ptr->identity]; - if (!l_ptr || strcmp(l_ptr->name, name)) - return NULL; + *bearer_id = 0; + rcu_read_lock(); + list_for_each_entry_rcu(n_ptr, &tipc_node_list, list) { + tipc_node_lock(n_ptr); + for (i = 0; i < MAX_BEARERS; i++) { + l_ptr = n_ptr->links[i]; + if (l_ptr && !strcmp(l_ptr->name, link_name)) { + *bearer_id = i; + found_node = n_ptr; + break; + } + } + tipc_node_unlock(n_ptr); + if (found_node) + break; + } + rcu_read_unlock(); - return l_ptr; + return found_node; } /** * link_value_is_valid -- validate proposed link tolerance/priority/window * - * @cmd - value type (TIPC_CMD_SET_LINK_*) - * @new_value - the new value + * @cmd: value type (TIPC_CMD_SET_LINK_*) + * @new_value: the new value * * Returns 1 if value is within range, 0 if not. */ - static int link_value_is_valid(u16 cmd, u32 new_value) { switch (cmd) { @@ -2781,49 +2390,52 @@ static int link_value_is_valid(u16 cmd, u32 new_value) return 0; } - /** * link_cmd_set_value - change priority/tolerance/window for link/bearer/media - * @name - ptr to link, bearer, or media name - * @new_value - new value of link, bearer, or media setting - * @cmd - which link, bearer, or media attribute to set (TIPC_CMD_SET_LINK_*) + * @name: ptr to link, bearer, or media name + * @new_value: new value of link, bearer, or media setting + * @cmd: which link, bearer, or media attribute to set (TIPC_CMD_SET_LINK_*) * - * Caller must hold 'tipc_net_lock' to ensure link/bearer/media is not deleted. + * Caller must hold RTNL lock to ensure link/bearer/media is not deleted. * * Returns 0 if value updated and negative value on error. */ - static int link_cmd_set_value(const char *name, u32 new_value, u16 cmd) { struct tipc_node *node; struct tipc_link *l_ptr; struct tipc_bearer *b_ptr; struct tipc_media *m_ptr; + int bearer_id; + int res = 0; - l_ptr = link_find_link(name, &node); - if (l_ptr) { - /* - * acquire node lock for tipc_link_send_proto_msg(). - * see "TIPC locking policy" in net.c. - */ + node = tipc_link_find_owner(name, &bearer_id); + if (node) { tipc_node_lock(node); - switch (cmd) { - case TIPC_CMD_SET_LINK_TOL: - link_set_supervision_props(l_ptr, new_value); - tipc_link_send_proto_msg(l_ptr, - STATE_MSG, 0, 0, new_value, 0, 0); - break; - case TIPC_CMD_SET_LINK_PRI: - l_ptr->priority = new_value; - tipc_link_send_proto_msg(l_ptr, - STATE_MSG, 0, 0, 0, new_value, 0); - break; - case TIPC_CMD_SET_LINK_WINDOW: - tipc_link_set_queue_limits(l_ptr, new_value); - break; + l_ptr = node->links[bearer_id]; + + if (l_ptr) { + switch (cmd) { + case TIPC_CMD_SET_LINK_TOL: + link_set_supervision_props(l_ptr, new_value); + tipc_link_proto_xmit(l_ptr, STATE_MSG, 0, 0, + new_value, 0, 0); + break; + case TIPC_CMD_SET_LINK_PRI: + l_ptr->priority = new_value; + tipc_link_proto_xmit(l_ptr, STATE_MSG, 0, 0, + 0, new_value, 0); + break; + case TIPC_CMD_SET_LINK_WINDOW: + tipc_link_set_queue_limits(l_ptr, new_value); + break; + default: + res = -EINVAL; + break; + } } tipc_node_unlock(node); - return 0; + return res; } b_ptr = tipc_bearer_find(name); @@ -2831,15 +2443,18 @@ static int link_cmd_set_value(const char *name, u32 new_value, u16 cmd) switch (cmd) { case TIPC_CMD_SET_LINK_TOL: b_ptr->tolerance = new_value; - return 0; + break; case TIPC_CMD_SET_LINK_PRI: b_ptr->priority = new_value; - return 0; + break; case TIPC_CMD_SET_LINK_WINDOW: b_ptr->window = new_value; - return 0; + break; + default: + res = -EINVAL; + break; } - return -EINVAL; + return res; } m_ptr = tipc_media_find(name); @@ -2848,15 +2463,18 @@ static int link_cmd_set_value(const char *name, u32 new_value, u16 cmd) switch (cmd) { case TIPC_CMD_SET_LINK_TOL: m_ptr->tolerance = new_value; - return 0; + break; case TIPC_CMD_SET_LINK_PRI: m_ptr->priority = new_value; - return 0; + break; case TIPC_CMD_SET_LINK_WINDOW: m_ptr->window = new_value; - return 0; + break; + default: + res = -EINVAL; + break; } - return -EINVAL; + return res; } struct sk_buff *tipc_link_cmd_config(const void *req_tlv_area, int req_tlv_space, @@ -2884,9 +2502,7 @@ struct sk_buff *tipc_link_cmd_config(const void *req_tlv_area, int req_tlv_space " (cannot change setting on broadcast link)"); } - read_lock_bh(&tipc_net_lock); res = link_cmd_set_value(args->name, new_value, cmd); - read_unlock_bh(&tipc_net_lock); if (res) return tipc_cfg_reply_error_string("cannot change link setting"); @@ -2897,7 +2513,6 @@ struct sk_buff *tipc_link_cmd_config(const void *req_tlv_area, int req_tlv_space * link_reset_statistics - reset link statistics * @l_ptr: pointer to link */ - static void link_reset_statistics(struct tipc_link *l_ptr) { memset(&l_ptr->stats, 0, sizeof(l_ptr->stats)); @@ -2910,6 +2525,7 @@ struct sk_buff *tipc_link_cmd_reset_stats(const void *req_tlv_area, int req_tlv_ char *link_name; struct tipc_link *l_ptr; struct tipc_node *node; + unsigned int bearer_id; if (!TLV_CHECK(req_tlv_area, req_tlv_space, TIPC_TLV_LINK_NAME)) return tipc_cfg_reply_error_string(TIPC_CFG_TLV_ERROR); @@ -2920,25 +2536,24 @@ struct sk_buff *tipc_link_cmd_reset_stats(const void *req_tlv_area, int req_tlv_ return tipc_cfg_reply_error_string("link not found"); return tipc_cfg_reply_none(); } + node = tipc_link_find_owner(link_name, &bearer_id); + if (!node) + return tipc_cfg_reply_error_string("link not found"); - read_lock_bh(&tipc_net_lock); - l_ptr = link_find_link(link_name, &node); + tipc_node_lock(node); + l_ptr = node->links[bearer_id]; if (!l_ptr) { - read_unlock_bh(&tipc_net_lock); + tipc_node_unlock(node); return tipc_cfg_reply_error_string("link not found"); } - - tipc_node_lock(node); link_reset_statistics(l_ptr); tipc_node_unlock(node); - read_unlock_bh(&tipc_net_lock); return tipc_cfg_reply_none(); } /** * percent - convert count to a percentage of total (rounding up or down) */ - static u32 percent(u32 count, u32 total) { return (count * 100 + (total / 2)) / total; @@ -2952,115 +2567,121 @@ static u32 percent(u32 count, u32 total) * * Returns length of print buffer data string (or 0 if error) */ - static int tipc_link_stats(const char *name, char *buf, const u32 buf_size) { - struct print_buf pb; - struct tipc_link *l_ptr; + struct tipc_link *l; + struct tipc_stats *s; struct tipc_node *node; char *status; u32 profile_total = 0; + unsigned int bearer_id; + int ret; if (!strcmp(name, tipc_bclink_name)) return tipc_bclink_stats(buf, buf_size); - tipc_printbuf_init(&pb, buf, buf_size); + node = tipc_link_find_owner(name, &bearer_id); + if (!node) + return 0; - read_lock_bh(&tipc_net_lock); - l_ptr = link_find_link(name, &node); - if (!l_ptr) { - read_unlock_bh(&tipc_net_lock); + tipc_node_lock(node); + + l = node->links[bearer_id]; + if (!l) { + tipc_node_unlock(node); return 0; } - tipc_node_lock(node); - if (tipc_link_is_active(l_ptr)) + s = &l->stats; + + if (tipc_link_is_active(l)) status = "ACTIVE"; - else if (tipc_link_is_up(l_ptr)) + else if (tipc_link_is_up(l)) status = "STANDBY"; else status = "DEFUNCT"; - tipc_printf(&pb, "Link <%s>\n" - " %s MTU:%u Priority:%u Tolerance:%u ms" - " Window:%u packets\n", - l_ptr->name, status, l_ptr->max_pkt, - l_ptr->priority, l_ptr->tolerance, l_ptr->queue_limit[0]); - tipc_printf(&pb, " RX packets:%u fragments:%u/%u bundles:%u/%u\n", - l_ptr->next_in_no - l_ptr->stats.recv_info, - l_ptr->stats.recv_fragments, - l_ptr->stats.recv_fragmented, - l_ptr->stats.recv_bundles, - l_ptr->stats.recv_bundled); - tipc_printf(&pb, " TX packets:%u fragments:%u/%u bundles:%u/%u\n", - l_ptr->next_out_no - l_ptr->stats.sent_info, - l_ptr->stats.sent_fragments, - l_ptr->stats.sent_fragmented, - l_ptr->stats.sent_bundles, - l_ptr->stats.sent_bundled); - profile_total = l_ptr->stats.msg_length_counts; + + ret = tipc_snprintf(buf, buf_size, "Link <%s>\n" + " %s MTU:%u Priority:%u Tolerance:%u ms" + " Window:%u packets\n", + l->name, status, l->max_pkt, l->priority, + l->tolerance, l->queue_limit[0]); + + ret += tipc_snprintf(buf + ret, buf_size - ret, + " RX packets:%u fragments:%u/%u bundles:%u/%u\n", + l->next_in_no - s->recv_info, s->recv_fragments, + s->recv_fragmented, s->recv_bundles, + s->recv_bundled); + + ret += tipc_snprintf(buf + ret, buf_size - ret, + " TX packets:%u fragments:%u/%u bundles:%u/%u\n", + l->next_out_no - s->sent_info, s->sent_fragments, + s->sent_fragmented, s->sent_bundles, + s->sent_bundled); + + profile_total = s->msg_length_counts; if (!profile_total) profile_total = 1; - tipc_printf(&pb, " TX profile sample:%u packets average:%u octets\n" - " 0-64:%u%% -256:%u%% -1024:%u%% -4096:%u%% " - "-16384:%u%% -32768:%u%% -66000:%u%%\n", - l_ptr->stats.msg_length_counts, - l_ptr->stats.msg_lengths_total / profile_total, - percent(l_ptr->stats.msg_length_profile[0], profile_total), - percent(l_ptr->stats.msg_length_profile[1], profile_total), - percent(l_ptr->stats.msg_length_profile[2], profile_total), - percent(l_ptr->stats.msg_length_profile[3], profile_total), - percent(l_ptr->stats.msg_length_profile[4], profile_total), - percent(l_ptr->stats.msg_length_profile[5], profile_total), - percent(l_ptr->stats.msg_length_profile[6], profile_total)); - tipc_printf(&pb, " RX states:%u probes:%u naks:%u defs:%u dups:%u\n", - l_ptr->stats.recv_states, - l_ptr->stats.recv_probes, - l_ptr->stats.recv_nacks, - l_ptr->stats.deferred_recv, - l_ptr->stats.duplicates); - tipc_printf(&pb, " TX states:%u probes:%u naks:%u acks:%u dups:%u\n", - l_ptr->stats.sent_states, - l_ptr->stats.sent_probes, - l_ptr->stats.sent_nacks, - l_ptr->stats.sent_acks, - l_ptr->stats.retransmitted); - tipc_printf(&pb, " Congestion bearer:%u link:%u Send queue max:%u avg:%u\n", - l_ptr->stats.bearer_congs, - l_ptr->stats.link_congs, - l_ptr->stats.max_queue_sz, - l_ptr->stats.queue_sz_counts - ? (l_ptr->stats.accu_queue_sz / l_ptr->stats.queue_sz_counts) - : 0); + + ret += tipc_snprintf(buf + ret, buf_size - ret, + " TX profile sample:%u packets average:%u octets\n" + " 0-64:%u%% -256:%u%% -1024:%u%% -4096:%u%% " + "-16384:%u%% -32768:%u%% -66000:%u%%\n", + s->msg_length_counts, + s->msg_lengths_total / profile_total, + percent(s->msg_length_profile[0], profile_total), + percent(s->msg_length_profile[1], profile_total), + percent(s->msg_length_profile[2], profile_total), + percent(s->msg_length_profile[3], profile_total), + percent(s->msg_length_profile[4], profile_total), + percent(s->msg_length_profile[5], profile_total), + percent(s->msg_length_profile[6], profile_total)); + + ret += tipc_snprintf(buf + ret, buf_size - ret, + " RX states:%u probes:%u naks:%u defs:%u" + " dups:%u\n", s->recv_states, s->recv_probes, + s->recv_nacks, s->deferred_recv, s->duplicates); + + ret += tipc_snprintf(buf + ret, buf_size - ret, + " TX states:%u probes:%u naks:%u acks:%u" + " dups:%u\n", s->sent_states, s->sent_probes, + s->sent_nacks, s->sent_acks, s->retransmitted); + + ret += tipc_snprintf(buf + ret, buf_size - ret, + " Congestion link:%u Send queue" + " max:%u avg:%u\n", s->link_congs, + s->max_queue_sz, s->queue_sz_counts ? + (s->accu_queue_sz / s->queue_sz_counts) : 0); tipc_node_unlock(node); - read_unlock_bh(&tipc_net_lock); - return tipc_printbuf_validate(&pb); + return ret; } -#define MAX_LINK_STATS_INFO 2000 - struct sk_buff *tipc_link_cmd_show_stats(const void *req_tlv_area, int req_tlv_space) { struct sk_buff *buf; struct tlv_desc *rep_tlv; int str_len; + int pb_len; + char *pb; if (!TLV_CHECK(req_tlv_area, req_tlv_space, TIPC_TLV_LINK_NAME)) return tipc_cfg_reply_error_string(TIPC_CFG_TLV_ERROR); - buf = tipc_cfg_reply_alloc(TLV_SPACE(MAX_LINK_STATS_INFO)); + buf = tipc_cfg_reply_alloc(TLV_SPACE(ULTRA_STRING_MAX_LEN)); if (!buf) return NULL; rep_tlv = (struct tlv_desc *)buf->data; - + pb = TLV_DATA(rep_tlv); + pb_len = ULTRA_STRING_MAX_LEN; str_len = tipc_link_stats((char *)TLV_DATA(req_tlv_area), - (char *)TLV_DATA(rep_tlv), MAX_LINK_STATS_INFO); + pb, pb_len); if (!str_len) { - buf_discard(buf); + kfree_skb(buf); return tipc_cfg_reply_error_string("link not found"); } - + str_len += 1; /* for "\0" */ skb_put(buf, TLV_SPACE(str_len)); TLV_SET(rep_tlv, TIPC_TLV_ULTRA_STRING, NULL, str_len); @@ -3074,7 +2695,6 @@ struct sk_buff *tipc_link_cmd_show_stats(const void *req_tlv_area, int req_tlv_s * * If no active link can be found, uses default maximum packet size. */ - u32 tipc_link_get_max_pkt(u32 dest, u32 selector) { struct tipc_node *n_ptr; @@ -3084,7 +2704,6 @@ u32 tipc_link_get_max_pkt(u32 dest, u32 selector) if (dest == tipc_own_addr) return MAX_MSG_SIZE; - read_lock_bh(&tipc_net_lock); n_ptr = tipc_node_find(dest); if (n_ptr) { tipc_node_lock(n_ptr); @@ -3093,69 +2712,27 @@ u32 tipc_link_get_max_pkt(u32 dest, u32 selector) res = l_ptr->max_pkt; tipc_node_unlock(n_ptr); } - read_unlock_bh(&tipc_net_lock); return res; } static void link_print(struct tipc_link *l_ptr, const char *str) { - char print_area[256]; - struct print_buf pb; - struct print_buf *buf = &pb; - - tipc_printbuf_init(buf, print_area, sizeof(print_area)); - - tipc_printf(buf, str); - tipc_printf(buf, "Link %x<%s>:", - l_ptr->addr, l_ptr->b_ptr->name); + struct tipc_bearer *b_ptr; -#ifdef CONFIG_TIPC_DEBUG - if (link_reset_reset(l_ptr) || link_reset_unknown(l_ptr)) - goto print_state; - - tipc_printf(buf, ": NXO(%u):", mod(l_ptr->next_out_no)); - tipc_printf(buf, "NXI(%u):", mod(l_ptr->next_in_no)); - tipc_printf(buf, "SQUE"); - if (l_ptr->first_out) { - tipc_printf(buf, "[%u..", buf_seqno(l_ptr->first_out)); - if (l_ptr->next_out) - tipc_printf(buf, "%u..", buf_seqno(l_ptr->next_out)); - tipc_printf(buf, "%u]", buf_seqno(l_ptr->last_out)); - if ((mod(buf_seqno(l_ptr->last_out) - - buf_seqno(l_ptr->first_out)) - != (l_ptr->out_queue_size - 1)) || - (l_ptr->last_out->next != NULL)) { - tipc_printf(buf, "\nSend queue inconsistency\n"); - tipc_printf(buf, "first_out= %p ", l_ptr->first_out); - tipc_printf(buf, "next_out= %p ", l_ptr->next_out); - tipc_printf(buf, "last_out= %p ", l_ptr->last_out); - } - } else - tipc_printf(buf, "[]"); - tipc_printf(buf, "SQSIZ(%u)", l_ptr->out_queue_size); - if (l_ptr->oldest_deferred_in) { - u32 o = buf_seqno(l_ptr->oldest_deferred_in); - u32 n = buf_seqno(l_ptr->newest_deferred_in); - tipc_printf(buf, ":RQUE[%u..%u]", o, n); - if (l_ptr->deferred_inqueue_sz != mod((n + 1) - o)) { - tipc_printf(buf, ":RQSIZ(%u)", - l_ptr->deferred_inqueue_sz); - } - } -print_state: -#endif + rcu_read_lock(); + b_ptr = rcu_dereference_rtnl(bearer_list[l_ptr->bearer_id]); + if (b_ptr) + pr_info("%s Link %x<%s>:", str, l_ptr->addr, b_ptr->name); + rcu_read_unlock(); if (link_working_unknown(l_ptr)) - tipc_printf(buf, ":WU"); + pr_cont(":WU\n"); else if (link_reset_reset(l_ptr)) - tipc_printf(buf, ":RR"); + pr_cont(":RR\n"); else if (link_reset_unknown(l_ptr)) - tipc_printf(buf, ":RU"); + pr_cont(":RU\n"); else if (link_working_working(l_ptr)) - tipc_printf(buf, ":WW"); - tipc_printf(buf, "\n"); - - tipc_printbuf_validate(buf); - info("%s", print_area); + pr_cont(":WW\n"); + else + pr_cont("\n"); } - diff --git a/net/tipc/link.h b/net/tipc/link.h index 73c18c140e1..200d518b218 100644 --- a/net/tipc/link.h +++ b/net/tipc/link.h @@ -1,7 +1,7 @@ /* * net/tipc/link.h: Include file for TIPC link code * - * Copyright (c) 1995-2006, Ericsson AB + * Copyright (c) 1995-2006, 2013, Ericsson AB * Copyright (c) 2004-2005, 2010-2011, Wind River Systems * All rights reserved. * @@ -37,35 +37,60 @@ #ifndef _TIPC_LINK_H #define _TIPC_LINK_H -#include "log.h" #include "msg.h" #include "node.h" -#define PUSH_FAILED 1 -#define PUSH_FINISHED 2 - -/* - * Out-of-range value for link sequence numbers +/* Out-of-range value for link sequence numbers */ - #define INVALID_LINK_SEQ 0x10000 -/* - * Link states +/* Link working states */ - #define WORKING_WORKING 560810u #define WORKING_UNKNOWN 560811u #define RESET_UNKNOWN 560812u #define RESET_RESET 560813u -/* - * Starting value for maximum packet size negotiation on unicast links - * (unless bearer MTU is less) +/* Link endpoint execution states */ +#define LINK_STARTED 0x0001 +#define LINK_STOPPED 0x0002 +/* Starting value for maximum packet size negotiation on unicast links + * (unless bearer MTU is less) + */ #define MAX_PKT_DEFAULT 1500 +struct tipc_stats { + u32 sent_info; /* used in counting # sent packets */ + u32 recv_info; /* used in counting # recv'd packets */ + u32 sent_states; + u32 recv_states; + u32 sent_probes; + u32 recv_probes; + u32 sent_nacks; + u32 recv_nacks; + u32 sent_acks; + u32 sent_bundled; + u32 sent_bundles; + u32 recv_bundled; + u32 recv_bundles; + u32 retransmitted; + u32 sent_fragmented; + u32 sent_fragments; + u32 recv_fragmented; + u32 recv_fragments; + u32 link_congs; /* # port sends blocked by congestion */ + u32 deferred_recv; + u32 duplicates; + u32 max_queue_sz; /* send queue size high water mark */ + u32 accu_queue_sz; /* used for send queue size profiling */ + u32 queue_sz_counts; /* used for send queue size profiling */ + u32 msg_length_counts; /* used for message length profiling */ + u32 msg_lengths_total; /* used for message length profiling */ + u32 msg_length_profile[7]; /* used for msg. length profiling */ +}; + /** * struct tipc_link - TIPC link data structure * @addr: network address of link's peer node @@ -73,21 +98,20 @@ * @media_addr: media address to use when sending messages over link * @timer: link timer * @owner: pointer to peer node - * @link_list: adjacent links in bearer's list of links - * @started: indicates if link has been started + * @flags: execution state flags for link endpoint instance * @checkpoint: reference point for triggering link continuity checking * @peer_session: link session # being used by peer end of link * @peer_bearer_id: bearer id used by link's peer endpoint - * @b_ptr: pointer to bearer used by link + * @bearer_id: local bearer id used by link * @tolerance: minimum link continuity loss needed to reset link [in ms] * @continuity_interval: link continuity testing interval [in ms] * @abort_limit: # of unacknowledged continuity probes needed to reset link * @state: current state of link FSM - * @blocked: indicates if link has been administratively blocked * @fsm_msg_cnt: # of protocol messages link FSM has sent in current state * @proto_msg: template for control messages generated by link * @pmsg: convenience pointer to "proto_msg" field * @priority: current link priority + * @net_plane: current link network plane ('A' through 'H') * @queue_limit: outbound message queue congestion thresholds (indexed by user) * @exp_msg_count: # of tunnelled messages expected during link changeover * @reset_checkpoint: seq # of last acknowledged message at time of link reset @@ -111,29 +135,26 @@ * @next_out: ptr to first unsent outbound message in queue * @waiting_ports: linked list of ports waiting for link congestion to abate * @long_msg_seq_no: next identifier to use for outbound fragmented messages - * @defragm_buf: list of partially reassembled inbound message fragments + * @reasm_buf: head of partially reassembled inbound message fragments * @stats: collects statistics regarding link activity */ - struct tipc_link { u32 addr; char name[TIPC_MAX_LINK_NAME]; struct tipc_media_addr media_addr; struct timer_list timer; struct tipc_node *owner; - struct list_head link_list; /* Management and link supervision data */ - int started; + unsigned int flags; u32 checkpoint; u32 peer_session; u32 peer_bearer_id; - struct tipc_bearer *b_ptr; + u32 bearer_id; u32 tolerance; u32 continuity_interval; u32 abort_limit; int state; - int blocked; u32 fsm_msg_cnt; struct { unchar hdr[INT_H_SIZE]; @@ -141,6 +162,7 @@ struct tipc_link { } proto_msg; struct tipc_msg *pmsg; u32 priority; + char net_plane; u32 queue_limit[15]; /* queue_limit[0]==window limit */ /* Changeover */ @@ -174,41 +196,12 @@ struct tipc_link { struct sk_buff *next_out; struct list_head waiting_ports; - /* Fragmentation/defragmentation */ + /* Fragmentation/reassembly */ u32 long_msg_seq_no; - struct sk_buff *defragm_buf; + struct sk_buff *reasm_buf; /* Statistics */ - struct { - u32 sent_info; /* used in counting # sent packets */ - u32 recv_info; /* used in counting # recv'd packets */ - u32 sent_states; - u32 recv_states; - u32 sent_probes; - u32 recv_probes; - u32 sent_nacks; - u32 recv_nacks; - u32 sent_acks; - u32 sent_bundled; - u32 sent_bundles; - u32 recv_bundled; - u32 recv_bundles; - u32 retransmitted; - u32 sent_fragmented; - u32 sent_fragments; - 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 */ - u32 accu_queue_sz; /* used for send queue size profiling */ - u32 queue_sz_counts; /* used for send queue size profiling */ - u32 msg_length_counts; /* used for message length profiling */ - u32 msg_lengths_total; /* used for message length profiling */ - u32 msg_length_profile[7]; /* used for msg. length profiling */ - } stats; + struct tipc_stats stats; }; struct tipc_port; @@ -216,37 +209,37 @@ struct tipc_port; struct tipc_link *tipc_link_create(struct tipc_node *n_ptr, struct tipc_bearer *b_ptr, const struct tipc_media_addr *media_addr); -void tipc_link_delete(struct tipc_link *l_ptr); -void tipc_link_changeover(struct tipc_link *l_ptr); -void tipc_link_send_duplicate(struct tipc_link *l_ptr, struct tipc_link *dest); +void tipc_link_delete_list(unsigned int bearer_id, bool shutting_down); +void tipc_link_failover_send_queue(struct tipc_link *l_ptr); +void tipc_link_dup_queue_xmit(struct tipc_link *l_ptr, struct tipc_link *dest); void tipc_link_reset_fragments(struct tipc_link *l_ptr); int tipc_link_is_up(struct tipc_link *l_ptr); int tipc_link_is_active(struct tipc_link *l_ptr); -u32 tipc_link_push_packet(struct tipc_link *l_ptr); -void tipc_link_stop(struct tipc_link *l_ptr); -struct sk_buff *tipc_link_cmd_config(const void *req_tlv_area, int req_tlv_space, u16 cmd); -struct sk_buff *tipc_link_cmd_show_stats(const void *req_tlv_area, int req_tlv_space); -struct sk_buff *tipc_link_cmd_reset_stats(const void *req_tlv_area, int req_tlv_space); +void tipc_link_purge_queues(struct tipc_link *l_ptr); +struct sk_buff *tipc_link_cmd_config(const void *req_tlv_area, + int req_tlv_space, + u16 cmd); +struct sk_buff *tipc_link_cmd_show_stats(const void *req_tlv_area, + int req_tlv_space); +struct sk_buff *tipc_link_cmd_reset_stats(const void *req_tlv_area, + int req_tlv_space); +void tipc_link_reset_all(struct tipc_node *node); void tipc_link_reset(struct tipc_link *l_ptr); -int tipc_link_send(struct sk_buff *buf, u32 dest, u32 selector); -void tipc_link_send_names(struct list_head *message_list, u32 dest); +void tipc_link_reset_list(unsigned int bearer_id); +int tipc_link_xmit(struct sk_buff *buf, u32 dest, u32 selector); +void tipc_link_names_xmit(struct list_head *message_list, u32 dest); +int __tipc_link_xmit(struct tipc_link *l_ptr, struct sk_buff *buf); int tipc_link_send_buf(struct tipc_link *l_ptr, struct sk_buff *buf); u32 tipc_link_get_max_pkt(u32 dest, u32 selector); -int tipc_link_send_sections_fast(struct tipc_port *sender, - struct iovec const *msg_sect, - const u32 num_sect, - unsigned int total_len, - u32 destnode); -void tipc_link_recv_bundle(struct sk_buff *buf); -int tipc_link_recv_fragment(struct sk_buff **pending, - struct sk_buff **fb, - struct tipc_msg **msg); -void tipc_link_send_proto_msg(struct tipc_link *l_ptr, u32 msg_typ, int prob, - u32 gap, u32 tolerance, u32 priority, - u32 acked_mtu); +int tipc_link_iovec_xmit_fast(struct tipc_port *sender, + struct iovec const *msg_sect, + unsigned int len, u32 destnode); +void tipc_link_bundle_rcv(struct sk_buff *buf); +void tipc_link_proto_xmit(struct tipc_link *l_ptr, u32 msg_typ, int prob, + u32 gap, u32 tolerance, u32 priority, u32 acked_mtu); void tipc_link_push_queue(struct tipc_link *l_ptr); u32 tipc_link_defer_pkt(struct sk_buff **head, struct sk_buff **tail, - struct sk_buff *buf); + struct sk_buff *buf); void tipc_link_wakeup_ports(struct tipc_link *l_ptr, int all); void tipc_link_set_queue_limits(struct tipc_link *l_ptr, u32 window); void tipc_link_retransmit(struct tipc_link *l_ptr, @@ -255,7 +248,6 @@ void tipc_link_retransmit(struct tipc_link *l_ptr, /* * Link sequence number manipulation routines (uses modulo 2**16 arithmetic) */ - static inline u32 buf_seqno(struct sk_buff *buf) { return msg_seqno(buf_msg(buf)); @@ -294,7 +286,6 @@ static inline u32 lesser(u32 left, u32 right) /* * Link status checking routines */ - static inline int link_working_working(struct tipc_link *l_ptr) { return l_ptr->state == WORKING_WORKING; @@ -315,11 +306,6 @@ static inline int link_reset_reset(struct tipc_link *l_ptr) return l_ptr->state == RESET_RESET; } -static inline int link_blocked(struct tipc_link *l_ptr) -{ - return l_ptr->exp_msg_count || l_ptr->blocked; -} - static inline int link_congested(struct tipc_link *l_ptr) { return l_ptr->out_queue_size >= l_ptr->queue_limit[0]; diff --git a/net/tipc/log.c b/net/tipc/log.c index 952c39f643e..abef644f27d 100644 --- a/net/tipc/log.c +++ b/net/tipc/log.c @@ -36,316 +36,20 @@ #include "core.h" #include "config.h" -#include "log.h" - -/* - * TIPC pre-defines the following print buffers: - * - * TIPC_NULL : null buffer (i.e. print nowhere) - * TIPC_CONS : system console - * TIPC_LOG : TIPC log buffer - * - * Additional user-defined print buffers are also permitted. - */ - -static struct print_buf null_buf = { NULL, 0, NULL, 0 }; -struct print_buf *const TIPC_NULL = &null_buf; - -static struct print_buf cons_buf = { NULL, 0, NULL, 1 }; -struct print_buf *const TIPC_CONS = &cons_buf; - -static struct print_buf log_buf = { NULL, 0, NULL, 1 }; -struct print_buf *const TIPC_LOG = &log_buf; - -/* - * Locking policy when using print buffers. - * - * 1) tipc_printf() uses 'print_lock' to protect against concurrent access to - * 'print_string' when writing to a print buffer. This also protects against - * concurrent writes to the print buffer being written to. - * - * 2) tipc_log_XXX() leverages the aforementioned use of 'print_lock' to - * protect against all types of concurrent operations on their associated - * print buffer (not just write operations). - * - * Note: All routines of the form tipc_printbuf_XXX() are lock-free, and rely - * on the caller to prevent simultaneous use of the print buffer(s) being - * manipulated. - */ - -static char print_string[TIPC_PB_MAX_STR]; -static DEFINE_SPINLOCK(print_lock); - -static void tipc_printbuf_move(struct print_buf *pb_to, - struct print_buf *pb_from); - -#define FORMAT(PTR, LEN, FMT) \ -{\ - va_list args;\ - va_start(args, FMT);\ - LEN = vsprintf(PTR, FMT, args);\ - va_end(args);\ - *(PTR + LEN) = '\0';\ -} - -/** - * tipc_printbuf_init - initialize print buffer to empty - * @pb: pointer to print buffer structure - * @raw: pointer to character array used by print buffer - * @size: size of character array - * - * Note: If the character array is too small (or absent), the print buffer - * becomes a null device that discards anything written to it. - */ - -void tipc_printbuf_init(struct print_buf *pb, char *raw, u32 size) -{ - pb->buf = raw; - pb->crs = raw; - pb->size = size; - pb->echo = 0; - - if (size < TIPC_PB_MIN_SIZE) { - pb->buf = NULL; - } else if (raw) { - pb->buf[0] = 0; - pb->buf[size - 1] = ~0; - } -} - -/** - * tipc_printbuf_reset - reinitialize print buffer to empty state - * @pb: pointer to print buffer structure - */ - -static void tipc_printbuf_reset(struct print_buf *pb) -{ - if (pb->buf) { - pb->crs = pb->buf; - pb->buf[0] = 0; - pb->buf[pb->size - 1] = ~0; - } -} - -/** - * tipc_printbuf_empty - test if print buffer is in empty state - * @pb: pointer to print buffer structure - * - * Returns non-zero if print buffer is empty. - */ - -static int tipc_printbuf_empty(struct print_buf *pb) -{ - return !pb->buf || (pb->crs == pb->buf); -} - -/** - * tipc_printbuf_validate - check for print buffer overflow - * @pb: pointer to print buffer structure - * - * Verifies that a print buffer has captured all data written to it. - * If data has been lost, linearize buffer and prepend an error message - * - * Returns length of print buffer data string (including trailing NUL) - */ - -int tipc_printbuf_validate(struct print_buf *pb) -{ - char *err = "\n\n*** PRINT BUFFER OVERFLOW ***\n\n"; - char *cp_buf; - struct print_buf cb; - - if (!pb->buf) - return 0; - - if (pb->buf[pb->size - 1] == 0) { - cp_buf = kmalloc(pb->size, GFP_ATOMIC); - if (cp_buf) { - tipc_printbuf_init(&cb, cp_buf, pb->size); - tipc_printbuf_move(&cb, pb); - tipc_printbuf_move(pb, &cb); - kfree(cp_buf); - memcpy(pb->buf, err, strlen(err)); - } else { - tipc_printbuf_reset(pb); - tipc_printf(pb, err); - } - } - return pb->crs - pb->buf + 1; -} - -/** - * tipc_printbuf_move - move print buffer contents to another print buffer - * @pb_to: pointer to destination print buffer structure - * @pb_from: pointer to source print buffer structure - * - * Current contents of destination print buffer (if any) are discarded. - * Source print buffer becomes empty if a successful move occurs. - */ - -static void tipc_printbuf_move(struct print_buf *pb_to, - struct print_buf *pb_from) -{ - int len; - - /* Handle the cases where contents can't be moved */ - - if (!pb_to->buf) - return; - - if (!pb_from->buf) { - tipc_printbuf_reset(pb_to); - return; - } - - if (pb_to->size < pb_from->size) { - strcpy(pb_to->buf, "*** PRINT BUFFER MOVE ERROR ***"); - pb_to->buf[pb_to->size - 1] = ~0; - pb_to->crs = strchr(pb_to->buf, 0); - return; - } - - /* Copy data from char after cursor to end (if used) */ - - len = pb_from->buf + pb_from->size - pb_from->crs - 2; - if ((pb_from->buf[pb_from->size - 1] == 0) && (len > 0)) { - strcpy(pb_to->buf, pb_from->crs + 1); - pb_to->crs = pb_to->buf + len; - } else - pb_to->crs = pb_to->buf; - - /* Copy data from start to cursor (always) */ - - len = pb_from->crs - pb_from->buf; - strcpy(pb_to->crs, pb_from->buf); - pb_to->crs += len; - - tipc_printbuf_reset(pb_from); -} /** - * tipc_printf - append formatted output to print buffer - * @pb: pointer to print buffer + * tipc_snprintf - append formatted output to print buffer + * @buf: pointer to print buffer + * @len: buffer length * @fmt: formatted info to be printed */ - -void tipc_printf(struct print_buf *pb, const char *fmt, ...) +int tipc_snprintf(char *buf, int len, const char *fmt, ...) { - int chars_to_add; - int chars_left; - char save_char; - - spin_lock_bh(&print_lock); - - FORMAT(print_string, chars_to_add, fmt); - if (chars_to_add >= TIPC_PB_MAX_STR) - strcpy(print_string, "*** PRINT BUFFER STRING TOO LONG ***"); - - if (pb->buf) { - chars_left = pb->buf + pb->size - pb->crs - 1; - if (chars_to_add <= chars_left) { - strcpy(pb->crs, print_string); - pb->crs += chars_to_add; - } else if (chars_to_add >= (pb->size - 1)) { - strcpy(pb->buf, print_string + chars_to_add + 1 - - pb->size); - pb->crs = pb->buf + pb->size - 1; - } else { - strcpy(pb->buf, print_string + chars_left); - save_char = print_string[chars_left]; - print_string[chars_left] = 0; - strcpy(pb->crs, print_string); - print_string[chars_left] = save_char; - pb->crs = pb->buf + chars_to_add - chars_left; - } - } - - if (pb->echo) - printk("%s", print_string); - - spin_unlock_bh(&print_lock); -} - -/** - * tipc_log_resize - change the size of the TIPC log buffer - * @log_size: print buffer size to use - */ - -int tipc_log_resize(int log_size) -{ - int res = 0; - - spin_lock_bh(&print_lock); - kfree(TIPC_LOG->buf); - TIPC_LOG->buf = NULL; - if (log_size) { - if (log_size < TIPC_PB_MIN_SIZE) - log_size = TIPC_PB_MIN_SIZE; - res = TIPC_LOG->echo; - tipc_printbuf_init(TIPC_LOG, kmalloc(log_size, GFP_ATOMIC), - log_size); - TIPC_LOG->echo = res; - res = !TIPC_LOG->buf; - } - spin_unlock_bh(&print_lock); - - return res; -} - -/** - * tipc_log_resize_cmd - reconfigure size of TIPC log buffer - */ - -struct sk_buff *tipc_log_resize_cmd(const void *req_tlv_area, int req_tlv_space) -{ - 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 != delimit(value, 0, 32768)) - return tipc_cfg_reply_error_string(TIPC_CFG_INVALID_VALUE - " (log size must be 0-32768)"); - if (tipc_log_resize(value)) - return tipc_cfg_reply_error_string( - "unable to create specified log (log size is now 0)"); - return tipc_cfg_reply_none(); -} - -/** - * tipc_log_dump - capture TIPC log buffer contents in configuration message - */ - -struct sk_buff *tipc_log_dump(void) -{ - struct sk_buff *reply; - - spin_lock_bh(&print_lock); - if (!TIPC_LOG->buf) { - spin_unlock_bh(&print_lock); - reply = tipc_cfg_reply_ultra_string("log not activated\n"); - } else if (tipc_printbuf_empty(TIPC_LOG)) { - spin_unlock_bh(&print_lock); - reply = tipc_cfg_reply_ultra_string("log is empty\n"); - } else { - struct tlv_desc *rep_tlv; - struct print_buf pb; - int str_len; + int i; + va_list args; - str_len = min(TIPC_LOG->size, 32768u); - spin_unlock_bh(&print_lock); - reply = tipc_cfg_reply_alloc(TLV_SPACE(str_len)); - if (reply) { - rep_tlv = (struct tlv_desc *)reply->data; - tipc_printbuf_init(&pb, TLV_DATA(rep_tlv), str_len); - spin_lock_bh(&print_lock); - tipc_printbuf_move(&pb, TIPC_LOG); - spin_unlock_bh(&print_lock); - str_len = strlen(TLV_DATA(rep_tlv)) + 1; - skb_put(reply, TLV_SPACE(str_len)); - TLV_SET(rep_tlv, TIPC_TLV_ULTRA_STRING, NULL, str_len); - } - } - return reply; + va_start(args, fmt); + i = vscnprintf(buf, len, fmt, args); + va_end(args); + return i; } diff --git a/net/tipc/msg.c b/net/tipc/msg.c index 3e4d3e29be6..0a37a472c29 100644 --- a/net/tipc/msg.c +++ b/net/tipc/msg.c @@ -1,7 +1,7 @@ /* * net/tipc/msg.c: TIPC message header routines * - * Copyright (c) 2000-2006, Ericsson AB + * Copyright (c) 2000-2006, 2014, Ericsson AB * Copyright (c) 2005, 2010-2011, Wind River Systems * All rights reserved. * @@ -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); @@ -72,15 +72,14 @@ 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) + unsigned int len, int max_size, struct sk_buff **buf) { - int dsz, sz, hsz, pos, res, cnt; + int dsz, sz, hsz; + unsigned char *to; - dsz = total_len; - pos = hsz = msg_hdr_sz(hdr); + dsz = len; + hsz = msg_hdr_sz(hdr); sz = hsz + dsz; msg_set_size(hdr, sz); if (unlikely(sz > max_size)) { @@ -92,265 +91,69 @@ int tipc_msg_build(struct tipc_msg *hdr, struct iovec const *msg_sect, if (!(*buf)) 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); - pos += msg_sect[cnt].iov_len; + to = (*buf)->data + hsz; + if (len && memcpy_fromiovecend(to, msg_sect, 0, dsz)) { + kfree_skb(*buf); + *buf = NULL; + return -EFAULT; } - if (likely(res)) - return dsz; - - buf_discard(*buf); - *buf = NULL; - return -EFAULT; + return dsz; } -#ifdef CONFIG_TIPC_DEBUG - -void tipc_msg_dbg(struct print_buf *buf, struct tipc_msg *msg, const char *str) +/* tipc_buf_append(): Append a buffer to the fragment list of another buffer + * @*headbuf: in: NULL for first frag, otherwise value returned from prev call + * out: set when successful non-complete reassembly, otherwise NULL + * @*buf: in: the buffer to append. Always defined + * out: head buf after sucessful complete reassembly, otherwise NULL + * Returns 1 when reassembly complete, otherwise 0 + */ +int tipc_buf_append(struct sk_buff **headbuf, struct sk_buff **buf) { - u32 usr = msg_user(msg); - tipc_printf(buf, KERN_DEBUG); - tipc_printf(buf, str); - - switch (usr) { - case MSG_BUNDLER: - tipc_printf(buf, "BNDL::"); - tipc_printf(buf, "MSGS(%u):", msg_msgcnt(msg)); - break; - case BCAST_PROTOCOL: - tipc_printf(buf, "BCASTP::"); - break; - case MSG_FRAGMENTER: - tipc_printf(buf, "FRAGM::"); - switch (msg_type(msg)) { - case FIRST_FRAGMENT: - tipc_printf(buf, "FIRST:"); - break; - case FRAGMENT: - tipc_printf(buf, "BODY:"); - break; - case LAST_FRAGMENT: - tipc_printf(buf, "LAST:"); - break; - default: - tipc_printf(buf, "UNKNOWN:%x", msg_type(msg)); - - } - tipc_printf(buf, "NO(%u/%u):", msg_long_msgno(msg), - msg_fragm_no(msg)); - break; - case TIPC_LOW_IMPORTANCE: - case TIPC_MEDIUM_IMPORTANCE: - case TIPC_HIGH_IMPORTANCE: - case TIPC_CRITICAL_IMPORTANCE: - tipc_printf(buf, "DAT%u:", msg_user(msg)); - if (msg_short(msg)) { - tipc_printf(buf, "CON:"); - break; - } - switch (msg_type(msg)) { - case TIPC_CONN_MSG: - tipc_printf(buf, "CON:"); - break; - case TIPC_MCAST_MSG: - tipc_printf(buf, "MCST:"); - break; - case TIPC_NAMED_MSG: - tipc_printf(buf, "NAM:"); - break; - case TIPC_DIRECT_MSG: - tipc_printf(buf, "DIR:"); - break; - default: - tipc_printf(buf, "UNKNOWN TYPE %u", msg_type(msg)); - } - if (msg_reroute_cnt(msg)) - tipc_printf(buf, "REROUTED(%u):", - msg_reroute_cnt(msg)); - break; - case NAME_DISTRIBUTOR: - tipc_printf(buf, "NMD::"); - switch (msg_type(msg)) { - case PUBLICATION: - tipc_printf(buf, "PUBL(%u):", (msg_size(msg) - msg_hdr_sz(msg)) / 20); /* Items */ - break; - case WITHDRAWAL: - tipc_printf(buf, "WDRW:"); - break; - default: - tipc_printf(buf, "UNKNOWN:%x", msg_type(msg)); - } - if (msg_reroute_cnt(msg)) - tipc_printf(buf, "REROUTED(%u):", - msg_reroute_cnt(msg)); - break; - case CONN_MANAGER: - tipc_printf(buf, "CONN_MNG:"); - switch (msg_type(msg)) { - case CONN_PROBE: - tipc_printf(buf, "PROBE:"); - break; - case CONN_PROBE_REPLY: - tipc_printf(buf, "PROBE_REPLY:"); - break; - case CONN_ACK: - tipc_printf(buf, "CONN_ACK:"); - tipc_printf(buf, "ACK(%u):", msg_msgcnt(msg)); - break; - default: - tipc_printf(buf, "UNKNOWN TYPE:%x", msg_type(msg)); - } - if (msg_reroute_cnt(msg)) - tipc_printf(buf, "REROUTED(%u):", msg_reroute_cnt(msg)); - break; - case LINK_PROTOCOL: - switch (msg_type(msg)) { - case STATE_MSG: - tipc_printf(buf, "STATE:"); - tipc_printf(buf, "%s:", msg_probe(msg) ? "PRB" : ""); - tipc_printf(buf, "NXS(%u):", msg_next_sent(msg)); - tipc_printf(buf, "GAP(%u):", msg_seq_gap(msg)); - tipc_printf(buf, "LSTBC(%u):", msg_last_bcast(msg)); - break; - case RESET_MSG: - tipc_printf(buf, "RESET:"); - if (msg_size(msg) != msg_hdr_sz(msg)) - tipc_printf(buf, "BEAR:%s:", msg_data(msg)); - break; - case ACTIVATE_MSG: - tipc_printf(buf, "ACTIVATE:"); - break; - default: - tipc_printf(buf, "UNKNOWN TYPE:%x", msg_type(msg)); - } - tipc_printf(buf, "PLANE(%c):", msg_net_plane(msg)); - tipc_printf(buf, "SESS(%u):", msg_session(msg)); - break; - case CHANGEOVER_PROTOCOL: - tipc_printf(buf, "TUNL:"); - switch (msg_type(msg)) { - case DUPLICATE_MSG: - tipc_printf(buf, "DUPL:"); - break; - case ORIGINAL_MSG: - tipc_printf(buf, "ORIG:"); - tipc_printf(buf, "EXP(%u)", msg_msgcnt(msg)); - break; - default: - tipc_printf(buf, "UNKNOWN TYPE:%x", msg_type(msg)); - } - break; - case LINK_CONFIG: - tipc_printf(buf, "CFG:"); - switch (msg_type(msg)) { - case DSC_REQ_MSG: - tipc_printf(buf, "DSC_REQ:"); - break; - case DSC_RESP_MSG: - tipc_printf(buf, "DSC_RESP:"); - break; - default: - tipc_printf(buf, "UNKNOWN TYPE:%x:", msg_type(msg)); - break; - } - break; - default: - tipc_printf(buf, "UNKNOWN USER:"); - } - - switch (usr) { - case CONN_MANAGER: - case TIPC_LOW_IMPORTANCE: - case TIPC_MEDIUM_IMPORTANCE: - case TIPC_HIGH_IMPORTANCE: - case TIPC_CRITICAL_IMPORTANCE: - switch (msg_errcode(msg)) { - case TIPC_OK: - break; - case TIPC_ERR_NO_NAME: - tipc_printf(buf, "NO_NAME:"); - break; - case TIPC_ERR_NO_PORT: - tipc_printf(buf, "NO_PORT:"); - break; - case TIPC_ERR_NO_NODE: - tipc_printf(buf, "NO_PROC:"); - break; - case TIPC_ERR_OVERLOAD: - tipc_printf(buf, "OVERLOAD:"); - break; - case TIPC_CONN_SHUTDOWN: - tipc_printf(buf, "SHUTDOWN:"); - break; - default: - tipc_printf(buf, "UNKNOWN ERROR(%x):", - msg_errcode(msg)); - } - default: - break; - } - - tipc_printf(buf, "HZ(%u):", msg_hdr_sz(msg)); - tipc_printf(buf, "SZ(%u):", msg_size(msg)); - tipc_printf(buf, "SQNO(%u):", msg_seqno(msg)); - - if (msg_non_seq(msg)) - tipc_printf(buf, "NOSEQ:"); - else - tipc_printf(buf, "ACK(%u):", msg_ack(msg)); - tipc_printf(buf, "BACK(%u):", msg_bcast_ack(msg)); - tipc_printf(buf, "PRND(%x)", msg_prevnode(msg)); - - if (msg_isdata(msg)) { - if (msg_named(msg)) { - tipc_printf(buf, "NTYP(%u):", msg_nametype(msg)); - tipc_printf(buf, "NINST(%u)", msg_nameinst(msg)); - } - } - - if ((usr != LINK_PROTOCOL) && (usr != LINK_CONFIG) && - (usr != MSG_BUNDLER)) { - if (!msg_short(msg)) { - tipc_printf(buf, ":ORIG(%x:%u):", - msg_orignode(msg), msg_origport(msg)); - tipc_printf(buf, ":DEST(%x:%u):", - msg_destnode(msg), msg_destport(msg)); - } else { - tipc_printf(buf, ":OPRT(%u):", msg_origport(msg)); - tipc_printf(buf, ":DPRT(%u):", msg_destport(msg)); - } - } - if (msg_user(msg) == NAME_DISTRIBUTOR) { - tipc_printf(buf, ":ONOD(%x):", msg_orignode(msg)); - tipc_printf(buf, ":DNOD(%x):", msg_destnode(msg)); + struct sk_buff *head = *headbuf; + struct sk_buff *frag = *buf; + struct sk_buff *tail; + struct tipc_msg *msg = buf_msg(frag); + u32 fragid = msg_type(msg); + bool headstolen; + int delta; + + skb_pull(frag, msg_hdr_sz(msg)); + + if (fragid == FIRST_FRAGMENT) { + if (head || skb_unclone(frag, GFP_ATOMIC)) + goto out_free; + head = *headbuf = frag; + skb_frag_list_init(head); + *buf = NULL; + return 0; } - - if (msg_user(msg) == LINK_CONFIG) { - struct tipc_media_addr orig; - - tipc_printf(buf, ":DDOM(%x):", msg_dest_domain(msg)); - tipc_printf(buf, ":NETID(%u):", msg_bc_netid(msg)); - memcpy(orig.value, msg_media_addr(msg), sizeof(orig.value)); - orig.media_id = 0; - orig.broadcast = 0; - tipc_media_addr_printf(buf, &orig); + if (!head) + goto out_free; + tail = TIPC_SKB_CB(head)->tail; + if (skb_try_coalesce(head, frag, &headstolen, &delta)) { + kfree_skb_partial(frag, headstolen); + } else { + if (!skb_has_frag_list(head)) + skb_shinfo(head)->frag_list = frag; + else + tail->next = frag; + head->truesize += frag->truesize; + head->data_len += frag->len; + head->len += frag->len; + TIPC_SKB_CB(head)->tail = frag; } - if (msg_user(msg) == BCAST_PROTOCOL) { - tipc_printf(buf, "BCNACK:AFTER(%u):", msg_bcgap_after(msg)); - tipc_printf(buf, "TO(%u):", msg_bcgap_to(msg)); + if (fragid == LAST_FRAGMENT) { + *buf = head; + TIPC_SKB_CB(head)->tail = NULL; + *headbuf = NULL; + return 1; } - tipc_printf(buf, "\n"); - if ((usr == CHANGEOVER_PROTOCOL) && (msg_msgcnt(msg))) - tipc_msg_dbg(buf, msg_get_wrapped(msg), " /"); - if ((usr == MSG_FRAGMENTER) && (msg_type(msg) == FIRST_FRAGMENT)) - tipc_msg_dbg(buf, msg_get_wrapped(msg), " /"); + *buf = NULL; + return 0; +out_free: + pr_warn_ratelimited("Unable to build fragment list\n"); + kfree_skb(*buf); + kfree_skb(*headbuf); + *buf = *headbuf = NULL; + return 0; } - -#endif diff --git a/net/tipc/msg.h b/net/tipc/msg.h index 7b0cda16710..503511903d1 100644 --- a/net/tipc/msg.h +++ b/net/tipc/msg.h @@ -1,7 +1,7 @@ /* * net/tipc/msg.h: Include file for TIPC message header routines * - * Copyright (c) 2000-2007, Ericsson AB + * Copyright (c) 2000-2007, 2014, Ericsson AB * Copyright (c) 2005-2008, 2010-2011, Wind River Systems * All rights reserved. * @@ -44,7 +44,6 @@ * * Note: Some items are also used with TIPC internal message headers */ - #define TIPC_VERSION 2 /* @@ -58,7 +57,6 @@ /* * Payload message types */ - #define TIPC_CONN_MSG 0 #define TIPC_MCAST_MSG 1 #define TIPC_NAMED_MSG 2 @@ -67,7 +65,6 @@ /* * Message header sizes */ - #define SHORT_H_SIZE 24 /* In-cluster basic payload message */ #define BASIC_H_SIZE 32 /* Basic payload message */ #define NAMED_H_SIZE 40 /* Named payload message */ @@ -121,7 +118,6 @@ static inline void msg_swap_words(struct tipc_msg *msg, u32 a, u32 b) /* * Word 0 */ - static inline u32 msg_version(struct tipc_msg *m) { return msg_bits(m, 0, 29, 7); @@ -216,7 +212,6 @@ static inline void msg_set_size(struct tipc_msg *m, u32 sz) /* * Word 1 */ - static inline u32 msg_type(struct tipc_msg *m) { return msg_bits(m, 1, 29, 0x7); @@ -291,7 +286,6 @@ static inline void msg_set_bcast_ack(struct tipc_msg *m, u32 n) /* * Word 2 */ - static inline u32 msg_ack(struct tipc_msg *m) { return msg_bits(m, 2, 16, 0xffff); @@ -315,8 +309,6 @@ static inline void msg_set_seqno(struct tipc_msg *m, u32 n) /* * Words 3-10 */ - - static inline u32 msg_prevnode(struct tipc_msg *m) { return msg_word(m, 3); @@ -384,11 +376,6 @@ static inline void msg_set_destnode(struct tipc_msg *m, u32 a) msg_set_word(m, 7, a); } -static inline int msg_is_dest(struct tipc_msg *m, u32 d) -{ - return msg_short(m) || (msg_destnode(m) == d); -} - static inline u32 msg_nametype(struct tipc_msg *m) { return msg_word(m, 8); @@ -439,7 +426,6 @@ static inline struct tipc_msg *msg_get_wrapped(struct tipc_msg *m) return (struct tipc_msg *)msg_data(m); } - /* * Constants and routines used to read and write TIPC internal message headers */ @@ -447,7 +433,6 @@ static inline struct tipc_msg *msg_get_wrapped(struct tipc_msg *m) /* * Internal message users */ - #define BCAST_PROTOCOL 5 #define MSG_BUNDLER 6 #define LINK_PROTOCOL 7 @@ -461,7 +446,6 @@ static inline struct tipc_msg *msg_get_wrapped(struct tipc_msg *m) /* * Connection management protocol message types */ - #define CONN_PROBE 0 #define CONN_PROBE_REPLY 1 #define CONN_ACK 2 @@ -469,14 +453,12 @@ static inline struct tipc_msg *msg_get_wrapped(struct tipc_msg *m) /* * Name distributor message types */ - #define PUBLICATION 0 #define WITHDRAWAL 1 /* * Segmentation message types */ - #define FIRST_FRAGMENT 0 #define FRAGMENT 1 #define LAST_FRAGMENT 2 @@ -484,7 +466,6 @@ static inline struct tipc_msg *msg_get_wrapped(struct tipc_msg *m) /* * Link management protocol message types */ - #define STATE_MSG 0 #define RESET_MSG 1 #define ACTIVATE_MSG 2 @@ -498,7 +479,6 @@ static inline struct tipc_msg *msg_get_wrapped(struct tipc_msg *m) /* * Config protocol message types */ - #define DSC_REQ_MSG 0 #define DSC_RESP_MSG 1 @@ -506,7 +486,6 @@ static inline struct tipc_msg *msg_get_wrapped(struct tipc_msg *m) /* * Word 1 */ - static inline u32 msg_seq_gap(struct tipc_msg *m) { return msg_bits(m, 1, 16, 0x1fff); @@ -517,11 +496,20 @@ static inline void msg_set_seq_gap(struct tipc_msg *m, u32 n) msg_set_bits(m, 1, 16, 0x1fff, n); } +static inline u32 msg_node_sig(struct tipc_msg *m) +{ + return msg_bits(m, 1, 0, 0xffff); +} + +static inline void msg_set_node_sig(struct tipc_msg *m, u32 n) +{ + msg_set_bits(m, 1, 0, 0xffff, n); +} + /* * Word 2 */ - static inline u32 msg_dest_domain(struct tipc_msg *m) { return msg_word(m, 2); @@ -556,7 +544,6 @@ static inline void msg_set_bcgap_to(struct tipc_msg *m, u32 n) /* * Word 4 */ - static inline u32 msg_last_bcast(struct tipc_msg *m) { return msg_bits(m, 4, 16, 0xffff); @@ -567,12 +554,6 @@ static inline void msg_set_last_bcast(struct tipc_msg *m, u32 n) msg_set_bits(m, 4, 16, 0xffff, n); } - -static inline u32 msg_fragm_no(struct tipc_msg *m) -{ - return msg_bits(m, 4, 16, 0xffff); -} - static inline void msg_set_fragm_no(struct tipc_msg *m, u32 n) { msg_set_bits(m, 4, 16, 0xffff, n); @@ -589,12 +570,6 @@ static inline void msg_set_next_sent(struct tipc_msg *m, u32 n) msg_set_bits(m, 4, 0, 0xffff, n); } - -static inline u32 msg_long_msgno(struct tipc_msg *m) -{ - return msg_bits(m, 4, 0, 0xffff); -} - static inline void msg_set_long_msgno(struct tipc_msg *m, u32 n) { msg_set_bits(m, 4, 0, 0xffff, n); @@ -623,7 +598,6 @@ static inline void msg_set_link_selector(struct tipc_msg *m, u32 n) /* * Word 5 */ - static inline u32 msg_session(struct tipc_msg *m) { return msg_bits(m, 5, 16, 0xffff); @@ -692,7 +666,6 @@ static inline char *msg_media_addr(struct tipc_msg *m) /* * Word 9 */ - static inline u32 msg_msgcnt(struct tipc_msg *m) { return msg_bits(m, 9, 16, 0xffff); @@ -734,10 +707,11 @@ 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); + unsigned int len, int max_size, struct sk_buff **buf); + +int tipc_buf_append(struct sk_buff **headbuf, struct sk_buff **buf); #endif diff --git a/net/tipc/name_distr.c b/net/tipc/name_distr.c index 98ebb37f180..8ce730984aa 100644 --- a/net/tipc/name_distr.c +++ b/net/tipc/name_distr.c @@ -38,47 +38,42 @@ #include "link.h" #include "name_distr.h" -#define ITEM_SIZE sizeof(struct distr_item) - /** - * struct distr_item - publication info distributed to other nodes - * @type: name sequence type - * @lower: name sequence lower bound - * @upper: name sequence upper bound - * @ref: publishing port reference - * @key: publication key - * - * ===> All fields are stored in network byte order. <=== - * - * First 3 fields identify (name or) name sequence being published. - * Reference field uniquely identifies port that published name sequence. - * Key field uniquely identifies publication, in the event a port has - * multiple publications of the same name sequence. - * - * Note: There is no field that identifies the publishing node because it is - * the same for all items contained within a publication message. + * struct publ_list - list of publications made by this node + * @list: circular list of publications + * @list_size: number of entries in list */ +struct publ_list { + struct list_head list; + u32 size; +}; -struct distr_item { - __be32 type; - __be32 lower; - __be32 upper; - __be32 ref; - __be32 key; +static struct publ_list publ_zone = { + .list = LIST_HEAD_INIT(publ_zone.list), + .size = 0, }; -/** - * List of externally visible publications by this node -- - * that is, all publications having scope > TIPC_NODE_SCOPE. - */ +static struct publ_list publ_cluster = { + .list = LIST_HEAD_INIT(publ_cluster.list), + .size = 0, +}; + +static struct publ_list publ_node = { + .list = LIST_HEAD_INIT(publ_node.list), + .size = 0, +}; + +static struct publ_list *publ_lists[] = { + NULL, + &publ_zone, /* publ_lists[TIPC_ZONE_SCOPE] */ + &publ_cluster, /* publ_lists[TIPC_CLUSTER_SCOPE] */ + &publ_node /* publ_lists[TIPC_NODE_SCOPE] */ +}; -static LIST_HEAD(publ_root); -static u32 publ_cnt; /** * publ_to_item - add publication info to a publication message */ - static void publ_to_item(struct distr_item *i, struct publication *p) { i->type = htonl(p->type); @@ -91,7 +86,6 @@ static void publ_to_item(struct distr_item *i, struct publication *p) /** * named_prepare_buf - allocate & initialize a publication message */ - static struct sk_buff *named_prepare_buf(u32 type, u32 size, u32 dest) { struct sk_buff *buf = tipc_buf_acquire(INT_H_SIZE + size); @@ -105,118 +99,102 @@ static struct sk_buff *named_prepare_buf(u32 type, u32 size, u32 dest) return buf; } -static void named_cluster_distribute(struct sk_buff *buf) +void named_cluster_distribute(struct sk_buff *buf) { struct sk_buff *buf_copy; struct tipc_node *n_ptr; + struct tipc_link *l_ptr; - list_for_each_entry(n_ptr, &tipc_node_list, list) { - if (tipc_node_active_links(n_ptr)) { + rcu_read_lock(); + list_for_each_entry_rcu(n_ptr, &tipc_node_list, list) { + tipc_node_lock(n_ptr); + l_ptr = n_ptr->active_links[n_ptr->addr & 1]; + if (l_ptr) { buf_copy = skb_copy(buf, GFP_ATOMIC); - if (!buf_copy) + if (!buf_copy) { + tipc_node_unlock(n_ptr); break; + } msg_set_destnode(buf_msg(buf_copy), n_ptr->addr); - tipc_link_send(buf_copy, n_ptr->addr, n_ptr->addr); + __tipc_link_xmit(l_ptr, buf_copy); } + tipc_node_unlock(n_ptr); } + rcu_read_unlock(); - buf_discard(buf); + kfree_skb(buf); } /** * tipc_named_publish - tell other nodes about a new publication by this node */ - -void tipc_named_publish(struct publication *publ) +struct sk_buff *tipc_named_publish(struct publication *publ) { struct sk_buff *buf; struct distr_item *item; - list_add_tail(&publ->local_list, &publ_root); - publ_cnt++; + list_add_tail(&publ->local_list, &publ_lists[publ->scope]->list); + publ_lists[publ->scope]->size++; + + if (publ->scope == TIPC_NODE_SCOPE) + return NULL; buf = named_prepare_buf(PUBLICATION, ITEM_SIZE, 0); if (!buf) { - warn("Publication distribution failure\n"); - return; + pr_warn("Publication distribution failure\n"); + return NULL; } item = (struct distr_item *)msg_data(buf_msg(buf)); publ_to_item(item, publ); - named_cluster_distribute(buf); + return buf; } /** * tipc_named_withdraw - tell other nodes about a withdrawn publication by this node */ - -void tipc_named_withdraw(struct publication *publ) +struct sk_buff *tipc_named_withdraw(struct publication *publ) { struct sk_buff *buf; struct distr_item *item; list_del(&publ->local_list); - publ_cnt--; + publ_lists[publ->scope]->size--; + + if (publ->scope == TIPC_NODE_SCOPE) + return NULL; buf = named_prepare_buf(WITHDRAWAL, ITEM_SIZE, 0); if (!buf) { - warn("Withdrawal distribution failure\n"); - return; + pr_warn("Withdrawal distribution failure\n"); + return NULL; } item = (struct distr_item *)msg_data(buf_msg(buf)); publ_to_item(item, publ); - named_cluster_distribute(buf); + return buf; } -/** - * tipc_named_node_up - tell specified node about all publications by this node +/* + * named_distribute - prepare name info for bulk distribution to another node */ - -void tipc_named_node_up(unsigned long nodearg) +static void named_distribute(struct list_head *message_list, u32 node, + struct publ_list *pls, u32 max_item_buf) { - struct tipc_node *n_ptr; - struct tipc_link *l_ptr; struct publication *publ; - struct distr_item *item = NULL; struct sk_buff *buf = NULL; - struct list_head message_list; - u32 node = (u32)nodearg; + struct distr_item *item = NULL; u32 left = 0; - u32 rest; - u32 max_item_buf = 0; - - /* compute maximum amount of publication data to send per message */ - - read_lock_bh(&tipc_net_lock); - n_ptr = tipc_node_find(node); - if (n_ptr) { - tipc_node_lock(n_ptr); - l_ptr = n_ptr->active_links[0]; - if (l_ptr) - max_item_buf = ((l_ptr->max_pkt - INT_H_SIZE) / - ITEM_SIZE) * ITEM_SIZE; - tipc_node_unlock(n_ptr); - } - read_unlock_bh(&tipc_net_lock); - if (!max_item_buf) - return; - - /* create list of publication messages, then send them as a unit */ - - INIT_LIST_HEAD(&message_list); + u32 rest = pls->size * ITEM_SIZE; - read_lock_bh(&tipc_nametbl_lock); - rest = publ_cnt * ITEM_SIZE; - - list_for_each_entry(publ, &publ_root, local_list) { + list_for_each_entry(publ, &pls->list, local_list) { if (!buf) { left = (rest <= max_item_buf) ? rest : max_item_buf; rest -= left; buf = named_prepare_buf(PUBLICATION, left, node); if (!buf) { - warn("Bulk publication distribution failure\n"); - goto exit; + pr_warn("Bulk publication failure\n"); + return; } item = (struct distr_item *)msg_data(buf_msg(buf)); } @@ -224,14 +202,25 @@ void tipc_named_node_up(unsigned long nodearg) item++; left -= ITEM_SIZE; if (!left) { - list_add_tail((struct list_head *)buf, &message_list); + list_add_tail((struct list_head *)buf, message_list); buf = NULL; } } -exit: +} + +/** + * tipc_named_node_up - tell specified node about all publications by this node + */ +void tipc_named_node_up(u32 max_item_buf, u32 node) +{ + LIST_HEAD(message_list); + + read_lock_bh(&tipc_nametbl_lock); + named_distribute(&message_list, node, &publ_cluster, max_item_buf); + 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_names_xmit(&message_list, node); } /** @@ -239,17 +228,12 @@ exit: * * Invoked for each publication issued by a newly failed node. * Removes publication structure from name table & deletes it. - * In rare cases the link may have come back up again when this - * function is called, and we have two items representing the same - * publication. Nudge this item's key to distinguish it from the other. */ - static void named_purge_publ(struct publication *publ) { struct publication *p; write_lock_bh(&tipc_nametbl_lock); - publ->key += 1222345; p = tipc_nametbl_remove_publ(publ->type, publ->lower, publ->node, publ->ref, publ->key); if (p) @@ -257,19 +241,19 @@ static void named_purge_publ(struct publication *publ) write_unlock_bh(&tipc_nametbl_lock); if (p != publ) { - err("Unable to remove publication from failed node\n" - "(type=%u, lower=%u, node=0x%x, ref=%u, key=%u)\n", - publ->type, publ->lower, publ->node, publ->ref, publ->key); + pr_err("Unable to remove publication from failed node\n" + " (type=%u, lower=%u, node=0x%x, ref=%u, key=%u)\n", + publ->type, publ->lower, publ->node, publ->ref, + publ->key); } kfree(p); } /** - * tipc_named_recv - process name table update message sent by another node + * tipc_named_rcv - process name table update message sent by another node */ - -void tipc_named_recv(struct sk_buff *buf) +void tipc_named_rcv(struct sk_buff *buf) { struct publication *publ; struct tipc_msg *msg = buf_msg(buf); @@ -304,37 +288,38 @@ void tipc_named_recv(struct sk_buff *buf) tipc_nodesub_unsubscribe(&publ->subscr); kfree(publ); } else { - err("Unable to remove publication by node 0x%x\n" - "(type=%u, lower=%u, ref=%u, key=%u)\n", - msg_orignode(msg), - ntohl(item->type), ntohl(item->lower), - ntohl(item->ref), ntohl(item->key)); + pr_err("Unable to remove publication by node 0x%x\n" + " (type=%u, lower=%u, ref=%u, key=%u)\n", + msg_orignode(msg), ntohl(item->type), + ntohl(item->lower), ntohl(item->ref), + ntohl(item->key)); } } else { - warn("Unrecognized name table message received\n"); + pr_warn("Unrecognized name table message received\n"); } item++; } write_unlock_bh(&tipc_nametbl_lock); - buf_discard(buf); + kfree_skb(buf); } /** - * tipc_named_reinit - re-initialize local publication list + * tipc_named_reinit - re-initialize local publications * * This routine is called whenever TIPC networking is enabled. - * All existing publications by this node that have "cluster" or "zone" scope - * are updated to reflect the node's new network address. + * All name table entries published by this node are updated to reflect + * the node's new network address. */ - void tipc_named_reinit(void) { struct publication *publ; + int scope; write_lock_bh(&tipc_nametbl_lock); - list_for_each_entry(publ, &publ_root, local_list) - publ->node = tipc_own_addr; + for (scope = TIPC_ZONE_SCOPE; scope <= TIPC_NODE_SCOPE; scope++) + list_for_each_entry(publ, &publ_lists[scope]->list, local_list) + publ->node = tipc_own_addr; write_unlock_bh(&tipc_nametbl_lock); } diff --git a/net/tipc/name_distr.h b/net/tipc/name_distr.h index 1e41bdd4f25..b2eed4ec152 100644 --- a/net/tipc/name_distr.h +++ b/net/tipc/name_distr.h @@ -39,10 +39,39 @@ #include "name_table.h" -void tipc_named_publish(struct publication *publ); -void tipc_named_withdraw(struct publication *publ); -void tipc_named_node_up(unsigned long node); -void tipc_named_recv(struct sk_buff *buf); +#define ITEM_SIZE sizeof(struct distr_item) + +/** + * struct distr_item - publication info distributed to other nodes + * @type: name sequence type + * @lower: name sequence lower bound + * @upper: name sequence upper bound + * @ref: publishing port reference + * @key: publication key + * + * ===> All fields are stored in network byte order. <=== + * + * First 3 fields identify (name or) name sequence being published. + * Reference field uniquely identifies port that published name sequence. + * Key field uniquely identifies publication, in the event a port has + * multiple publications of the same name sequence. + * + * Note: There is no field that identifies the publishing node because it is + * the same for all items contained within a publication message. + */ +struct distr_item { + __be32 type; + __be32 lower; + __be32 upper; + __be32 ref; + __be32 key; +}; + +struct sk_buff *tipc_named_publish(struct publication *publ); +struct sk_buff *tipc_named_withdraw(struct publication *publ); +void named_cluster_distribute(struct sk_buff *buf); +void tipc_named_node_up(u32 max_item_buf, u32 node); +void tipc_named_rcv(struct sk_buff *buf); void tipc_named_reinit(void); #endif diff --git a/net/tipc/name_table.c b/net/tipc/name_table.c index 89eb5621ebb..9d7d37d9518 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 @@ -56,7 +56,6 @@ static int tipc_nametbl_size = 1024; /* must be a power of 2 */ * publications of the associated name sequence belong to it. * (The cluster and node lists may be empty.) */ - struct name_info { struct list_head node_list; struct list_head cluster_list; @@ -72,7 +71,6 @@ struct name_info { * @upper: name sequence upper bound * @info: pointer to name sequence publication info */ - struct sub_seq { u32 lower; u32 upper; @@ -90,7 +88,6 @@ struct sub_seq { * @subscriptions: list of subscriptions for this 'type' * @lock: spinlock controlling access to publication lists of all sub-sequences */ - struct name_seq { u32 type; struct sub_seq *sseqs; @@ -107,33 +104,29 @@ struct name_seq { * accessed via hashing on 'type'; name sequence lists are *not* sorted * @local_publ_count: number of publications issued by this node */ - struct name_table { struct hlist_head *types; u32 local_publ_count; }; static struct name_table table; -static atomic_t rsv_publ_ok = ATOMIC_INIT(0); DEFINE_RWLOCK(tipc_nametbl_lock); - static int hash(int x) { - return x & (tipc_nametbl_size - 1); + return x & (TIPC_NAMETBL_SIZE - 1); } /** * publ_create - create a publication structure */ - static struct publication *publ_create(u32 type, u32 lower, u32 upper, u32 scope, u32 node, u32 port_ref, u32 key) { struct publication *publ = kzalloc(sizeof(*publ), GFP_ATOMIC); if (publ == NULL) { - warn("Publication creation failure, no memory\n"); + pr_warn("Publication creation failure, no memory\n"); return NULL; } @@ -153,11 +146,9 @@ static struct publication *publ_create(u32 type, u32 lower, u32 upper, /** * tipc_subseq_alloc - allocate a specified number of sub-sequence structures */ - static struct sub_seq *tipc_subseq_alloc(u32 cnt) { - struct sub_seq *sseq = kcalloc(cnt, sizeof(struct sub_seq), GFP_ATOMIC); - return sseq; + return kcalloc(cnt, sizeof(struct sub_seq), GFP_ATOMIC); } /** @@ -165,14 +156,13 @@ static struct sub_seq *tipc_subseq_alloc(u32 cnt) * * Allocates a single sub-sequence structure and sets it to all 0's. */ - static struct name_seq *tipc_nameseq_create(u32 type, struct hlist_head *seq_head) { struct name_seq *nseq = kzalloc(sizeof(*nseq), GFP_ATOMIC); struct sub_seq *sseq = tipc_subseq_alloc(1); if (!nseq || !sseq) { - warn("Name sequence creation failed, no memory\n"); + pr_warn("Name sequence creation failed, no memory\n"); kfree(nseq); kfree(sseq); return NULL; @@ -188,12 +178,23 @@ static struct name_seq *tipc_nameseq_create(u32 type, struct hlist_head *seq_hea return nseq; } +/* + * nameseq_delete_empty - deletes a name sequence structure if now unused + */ +static void nameseq_delete_empty(struct name_seq *seq) +{ + if (!seq->first_free && list_empty(&seq->subscriptions)) { + hlist_del_init(&seq->ns_list); + kfree(seq->sseqs); + kfree(seq); + } +} + /** * nameseq_find_subseq - find sub-sequence (if any) matching a name instance * * Very time-critical, so binary searches through sub-sequence array. */ - static struct sub_seq *nameseq_find_subseq(struct name_seq *nseq, u32 instance) { @@ -223,7 +224,6 @@ static struct sub_seq *nameseq_find_subseq(struct name_seq *nseq, * * Note: Similar to binary search code for locating a sub-sequence. */ - static u32 nameseq_locate_subseq(struct name_seq *nseq, u32 instance) { struct sub_seq *sseqs = nseq->sseqs; @@ -244,9 +244,8 @@ static u32 nameseq_locate_subseq(struct name_seq *nseq, u32 instance) } /** - * tipc_nameseq_insert_publ - + * tipc_nameseq_insert_publ */ - static struct publication *tipc_nameseq_insert_publ(struct name_seq *nseq, u32 type, u32 lower, u32 upper, u32 scope, u32 node, u32 port, u32 key) @@ -262,39 +261,42 @@ static struct publication *tipc_nameseq_insert_publ(struct name_seq *nseq, if (sseq) { /* Lower end overlaps existing entry => need an exact match */ - if ((sseq->lower != lower) || (sseq->upper != upper)) { - warn("Cannot publish {%u,%u,%u}, overlap error\n", - type, lower, upper); + pr_warn("Cannot publish {%u,%u,%u}, overlap error\n", + type, lower, upper); return NULL; } info = sseq->info; + + /* Check if an identical publication already exists */ + list_for_each_entry(publ, &info->zone_list, zone_list) { + if ((publ->ref == port) && (publ->key == key) && + (!publ->node || (publ->node == node))) + return NULL; + } } else { u32 inspos; struct sub_seq *freesseq; /* Find where lower end should be inserted */ - inspos = nameseq_locate_subseq(nseq, lower); /* Fail if upper end overlaps into an existing entry */ - if ((inspos < nseq->first_free) && (upper >= nseq->sseqs[inspos].lower)) { - warn("Cannot publish {%u,%u,%u}, overlap error\n", - type, lower, upper); + pr_warn("Cannot publish {%u,%u,%u}, overlap error\n", + type, lower, upper); return NULL; } /* Ensure there is space for new sub-sequence */ - if (nseq->first_free == nseq->alloc) { struct sub_seq *sseqs = tipc_subseq_alloc(nseq->alloc * 2); if (!sseqs) { - warn("Cannot publish {%u,%u,%u}, no memory\n", - type, lower, upper); + pr_warn("Cannot publish {%u,%u,%u}, no memory\n", + type, lower, upper); return NULL; } memcpy(sseqs, nseq->sseqs, @@ -306,8 +308,8 @@ static struct publication *tipc_nameseq_insert_publ(struct name_seq *nseq, info = kzalloc(sizeof(*info), GFP_ATOMIC); if (!info) { - warn("Cannot publish {%u,%u,%u}, no memory\n", - type, lower, upper); + pr_warn("Cannot publish {%u,%u,%u}, no memory\n", + type, lower, upper); return NULL; } @@ -316,7 +318,6 @@ static struct publication *tipc_nameseq_insert_publ(struct name_seq *nseq, INIT_LIST_HEAD(&info->zone_list); /* Insert new sub-sequence */ - sseq = &nseq->sseqs[inspos]; freesseq = &nseq->sseqs[nseq->first_free]; memmove(sseq + 1, sseq, (freesseq - sseq) * sizeof(*sseq)); @@ -328,8 +329,7 @@ static struct publication *tipc_nameseq_insert_publ(struct name_seq *nseq, created_subseq = 1; } - /* Insert a publication: */ - + /* Insert a publication */ publ = publ_create(type, lower, upper, scope, node, port, key); if (!publ) return NULL; @@ -342,14 +342,12 @@ static struct publication *tipc_nameseq_insert_publ(struct name_seq *nseq, info->cluster_list_size++; } - if (node == tipc_own_addr) { + if (in_own_node(node)) { list_add(&publ->node_list, &info->node_list); info->node_list_size++; } - /* - * Any subscriptions waiting for notification? - */ + /* Any subscriptions waiting for notification? */ list_for_each_entry_safe(s, st, &nseq->subscriptions, nameseq_list) { tipc_subscr_report_overlap(s, publ->lower, @@ -363,7 +361,7 @@ static struct publication *tipc_nameseq_insert_publ(struct name_seq *nseq, } /** - * tipc_nameseq_remove_publ - + * tipc_nameseq_remove_publ * * NOTE: There may be cases where TIPC is asked to remove a publication * that is not in the name table. For example, if another node issues a @@ -373,7 +371,6 @@ static struct publication *tipc_nameseq_insert_publ(struct name_seq *nseq, * A failed withdraw request simply returns a failure indication and lets the * caller issue any error or warning messages associated with such a problem. */ - static struct publication *tipc_nameseq_remove_publ(struct name_seq *nseq, u32 inst, u32 node, u32 ref, u32 key) { @@ -390,7 +387,6 @@ static struct publication *tipc_nameseq_remove_publ(struct name_seq *nseq, u32 i info = sseq->info; /* Locate publication, if it exists */ - list_for_each_entry(publ, &info->zone_list, zone_list) { if ((publ->key == key) && (publ->ref == ref) && (!publ->node || (publ->node == node))) @@ -400,26 +396,22 @@ static struct publication *tipc_nameseq_remove_publ(struct name_seq *nseq, u32 i found: /* Remove publication from zone scope list */ - list_del(&publ->zone_list); info->zone_list_size--; /* Remove publication from cluster scope list, if present */ - if (in_own_cluster(node)) { list_del(&publ->cluster_list); info->cluster_list_size--; } /* Remove publication from node scope list, if present */ - - if (node == tipc_own_addr) { + if (in_own_node(node)) { list_del(&publ->node_list); info->node_list_size--; } /* Contract subseq list if no more publications for that subseq */ - if (list_empty(&info->zone_list)) { kfree(info); free = &nseq->sseqs[nseq->first_free--]; @@ -428,7 +420,6 @@ found: } /* Notify any waiting subscriptions */ - list_for_each_entry_safe(s, st, &nseq->subscriptions, nameseq_list) { tipc_subscr_report_overlap(s, publ->lower, @@ -443,13 +434,12 @@ found: } /** - * tipc_nameseq_subscribe: attach a subscription, and issue + * tipc_nameseq_subscribe - attach a subscription, and issue * the prescribed number of events if there is any sub- * 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; @@ -482,11 +472,10 @@ 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; } @@ -499,9 +488,10 @@ struct publication *tipc_nametbl_insert_publ(u32 type, u32 lower, u32 upper, { struct name_seq *seq = nametbl_find_seq(type); - if (lower > upper) { - warn("Failed to publish illegal {%u,%u,%u}\n", - type, lower, upper); + if ((scope < TIPC_ZONE_SCOPE) || (scope > TIPC_NODE_SCOPE) || + (lower > upper)) { + pr_debug("Failed to publish illegal {%u,%u,%u} with scope %u\n", + type, lower, upper, scope); return NULL; } @@ -524,22 +514,23 @@ struct publication *tipc_nametbl_remove_publ(u32 type, u32 lower, return NULL; publ = tipc_nameseq_remove_publ(seq, lower, node, ref, key); - - if (!seq->first_free && list_empty(&seq->subscriptions)) { - hlist_del_init(&seq->ns_list); - kfree(seq->sseqs); - kfree(seq); - } + nameseq_delete_empty(seq); return publ; } -/* - * tipc_nametbl_translate - translate name to port id +/** + * tipc_nametbl_translate - perform name translation + * + * On entry, 'destnode' is the search domain used during translation. * - * Note: on entry 'destnode' is the search domain used during translation; - * on exit it passes back the node address of the matching port (if any) + * On exit: + * - if name translation is deferred to another node/cluster/zone, + * leaves 'destnode' unchanged (will be non-zero) and returns 0 + * - if name translation is attempted and succeeds, sets 'destnode' + * to publishing node and returns port reference (will be non-zero) + * - if name translation is attempted and fails, sets 'destnode' to 0 + * and returns 0 */ - u32 tipc_nametbl_translate(u32 type, u32 instance, u32 *destnode) { struct sub_seq *sseq; @@ -547,6 +538,7 @@ u32 tipc_nametbl_translate(u32 type, u32 instance, u32 *destnode) struct publication *publ; struct name_seq *seq; u32 ref = 0; + u32 node = 0; if (!tipc_in_scope(*destnode, tipc_own_addr)) return 0; @@ -561,7 +553,7 @@ u32 tipc_nametbl_translate(u32 type, u32 instance, u32 *destnode) spin_lock_bh(&seq->lock); info = sseq->info; - /* Closest-First Algorithm: */ + /* Closest-First Algorithm */ if (likely(!*destnode)) { if (!list_empty(&info->node_list)) { publ = list_first_entry(&info->node_list, @@ -584,14 +576,14 @@ u32 tipc_nametbl_translate(u32 type, u32 instance, u32 *destnode) } } - /* Round-Robin Algorithm: */ + /* Round-Robin Algorithm */ else if (*destnode == tipc_own_addr) { if (list_empty(&info->node_list)) goto no_match; publ = list_first_entry(&info->node_list, struct publication, node_list); list_move_tail(&publ->node_list, &info->node_list); - } else if (in_own_cluster(*destnode)) { + } else if (in_own_cluster_exact(*destnode)) { if (list_empty(&info->cluster_list)) goto no_match; publ = list_first_entry(&info->cluster_list, struct publication, @@ -604,11 +596,12 @@ u32 tipc_nametbl_translate(u32 type, u32 instance, u32 *destnode) } ref = publ->ref; - *destnode = publ->node; + node = publ->node; no_match: spin_unlock_bh(&seq->lock); not_found: read_unlock_bh(&tipc_nametbl_lock); + *destnode = node; return ref; } @@ -624,7 +617,6 @@ not_found: * * Returns non-zero if any off-node ports overlap */ - int tipc_nametbl_mc_translate(u32 type, u32 lower, u32 upper, u32 limit, struct tipc_port_list *dports) { @@ -665,81 +657,66 @@ exit: return res; } -/** - * tipc_nametbl_publish_rsv - publish port name using a reserved name type - */ - -int tipc_nametbl_publish_rsv(u32 ref, unsigned int scope, - struct tipc_name_seq const *seq) -{ - int res; - - atomic_inc(&rsv_publ_ok); - res = tipc_publish(ref, scope, seq); - atomic_dec(&rsv_publ_ok); - return res; -} - -/** +/* * 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; + struct sk_buff *buf = NULL; - if (table.local_publ_count >= tipc_max_publications) { - warn("Publication failed, local publication limit reached (%u)\n", - tipc_max_publications); - return NULL; - } - if ((type < TIPC_RESERVED_TYPES) && !atomic_read(&rsv_publ_ok)) { - warn("Publication failed, reserved name {%u,%u,%u}\n", - type, lower, upper); + if (table.local_publ_count >= TIPC_MAX_PUBLICATIONS) { + pr_warn("Publication failed, local publication limit reached (%u)\n", + TIPC_MAX_PUBLICATIONS); return NULL; } write_lock_bh(&tipc_nametbl_lock); - table.local_publ_count++; publ = tipc_nametbl_insert_publ(type, lower, upper, scope, tipc_own_addr, port_ref, key); - if (publ && (scope != TIPC_NODE_SCOPE)) - tipc_named_publish(publ); + if (likely(publ)) { + table.local_publ_count++; + buf = tipc_named_publish(publ); + } write_unlock_bh(&tipc_nametbl_lock); + + if (buf) + named_cluster_distribute(buf); return publ; } /** * tipc_nametbl_withdraw - withdraw name publication from network name tables */ - int tipc_nametbl_withdraw(u32 type, u32 lower, u32 ref, u32 key) { struct publication *publ; + struct sk_buff *buf; write_lock_bh(&tipc_nametbl_lock); publ = tipc_nametbl_remove_publ(type, lower, tipc_own_addr, ref, key); if (likely(publ)) { table.local_publ_count--; - if (publ->scope != TIPC_NODE_SCOPE) - tipc_named_withdraw(publ); + buf = tipc_named_withdraw(publ); write_unlock_bh(&tipc_nametbl_lock); list_del_init(&publ->pport_list); kfree(publ); + + if (buf) + named_cluster_distribute(buf); return 1; } write_unlock_bh(&tipc_nametbl_lock); - err("Unable to remove local publication\n" - "(type=%u, lower=%u, ref=%u, key=%u)\n", - type, lower, ref, key); + pr_err("Unable to remove local publication\n" + "(type=%u, lower=%u, ref=%u, key=%u)\n", + type, lower, ref, key); return 0; } /** * tipc_nametbl_subscribe - add a subscription object to the name table */ - void tipc_nametbl_subscribe(struct tipc_subscription *s) { u32 type = s->seq.type; @@ -754,8 +731,8 @@ void tipc_nametbl_subscribe(struct tipc_subscription *s) tipc_nameseq_subscribe(seq, s); spin_unlock_bh(&seq->lock); } else { - warn("Failed to create subscription for {%u,%u,%u}\n", - s->seq.type, s->seq.lower, s->seq.upper); + pr_warn("Failed to create subscription for {%u,%u,%u}\n", + s->seq.type, s->seq.lower, s->seq.upper); } write_unlock_bh(&tipc_nametbl_lock); } @@ -763,7 +740,6 @@ void tipc_nametbl_subscribe(struct tipc_subscription *s) /** * tipc_nametbl_unsubscribe - remove a subscription object from name table */ - void tipc_nametbl_unsubscribe(struct tipc_subscription *s) { struct name_seq *seq; @@ -774,33 +750,29 @@ void tipc_nametbl_unsubscribe(struct tipc_subscription *s) spin_lock_bh(&seq->lock); list_del_init(&s->nameseq_list); spin_unlock_bh(&seq->lock); - if ((seq->first_free == 0) && list_empty(&seq->subscriptions)) { - hlist_del_init(&seq->ns_list); - kfree(seq->sseqs); - kfree(seq); - } + nameseq_delete_empty(seq); } write_unlock_bh(&tipc_nametbl_lock); } /** - * subseq_list: print specified sub-sequence contents into the given buffer + * subseq_list - print specified sub-sequence contents into the given buffer */ - -static void subseq_list(struct sub_seq *sseq, struct print_buf *buf, u32 depth, - u32 index) +static int subseq_list(struct sub_seq *sseq, char *buf, int len, u32 depth, + u32 index) { char portIdStr[27]; const char *scope_str[] = {"", " zone", " cluster", " node"}; struct publication *publ; struct name_info *info; + int ret; - tipc_printf(buf, "%-10u %-10u ", sseq->lower, sseq->upper); + ret = tipc_snprintf(buf, len, "%-10u %-10u ", sseq->lower, sseq->upper); if (depth == 2) { - tipc_printf(buf, "\n"); - return; + ret += tipc_snprintf(buf - ret, len + ret, "\n"); + return ret; } info = sseq->info; @@ -809,54 +781,58 @@ static void subseq_list(struct sub_seq *sseq, struct print_buf *buf, u32 depth, sprintf(portIdStr, "<%u.%u.%u:%u>", tipc_zone(publ->node), tipc_cluster(publ->node), tipc_node(publ->node), publ->ref); - tipc_printf(buf, "%-26s ", portIdStr); + ret += tipc_snprintf(buf + ret, len - ret, "%-26s ", portIdStr); if (depth > 3) { - tipc_printf(buf, "%-10u %s", publ->key, - scope_str[publ->scope]); + ret += tipc_snprintf(buf + ret, len - ret, "%-10u %s", + publ->key, scope_str[publ->scope]); } if (!list_is_last(&publ->zone_list, &info->zone_list)) - tipc_printf(buf, "\n%33s", " "); - }; + ret += tipc_snprintf(buf + ret, len - ret, + "\n%33s", " "); + } - tipc_printf(buf, "\n"); + ret += tipc_snprintf(buf + ret, len - ret, "\n"); + return ret; } /** - * nameseq_list: print specified name sequence contents into the given buffer + * nameseq_list - print specified name sequence contents into the given buffer */ - -static void nameseq_list(struct name_seq *seq, struct print_buf *buf, u32 depth, - u32 type, u32 lowbound, u32 upbound, u32 index) +static int nameseq_list(struct name_seq *seq, char *buf, int len, u32 depth, + u32 type, u32 lowbound, u32 upbound, u32 index) { struct sub_seq *sseq; char typearea[11]; + int ret = 0; if (seq->first_free == 0) - return; + return 0; sprintf(typearea, "%-10u", seq->type); if (depth == 1) { - tipc_printf(buf, "%s\n", typearea); - return; + ret += tipc_snprintf(buf, len, "%s\n", typearea); + return ret; } for (sseq = seq->sseqs; sseq != &seq->sseqs[seq->first_free]; sseq++) { if ((lowbound <= sseq->upper) && (upbound >= sseq->lower)) { - tipc_printf(buf, "%s ", typearea); + ret += tipc_snprintf(buf + ret, len - ret, "%s ", + typearea); spin_lock_bh(&seq->lock); - subseq_list(sseq, buf, depth, index); + ret += subseq_list(sseq, buf + ret, len - ret, + depth, index); spin_unlock_bh(&seq->lock); sprintf(typearea, "%10s", " "); } } + return ret; } /** * nametbl_header - print name table header into the given buffer */ - -static void nametbl_header(struct print_buf *buf, u32 depth) +static int nametbl_header(char *buf, int len, u32 depth) { const char *header[] = { "Type ", @@ -866,25 +842,26 @@ static void nametbl_header(struct print_buf *buf, u32 depth) }; int i; + int ret = 0; if (depth > 4) depth = 4; for (i = 0; i < depth; i++) - tipc_printf(buf, header[i]); - tipc_printf(buf, "\n"); + ret += tipc_snprintf(buf + ret, len - ret, header[i]); + ret += tipc_snprintf(buf + ret, len - ret, "\n"); + return ret; } /** * nametbl_list - print specified name table contents into the given buffer */ - -static void nametbl_list(struct print_buf *buf, u32 depth_info, - u32 type, u32 lowbound, u32 upbound) +static int nametbl_list(char *buf, int len, u32 depth_info, + 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; u32 depth; u32 i; @@ -892,65 +869,69 @@ static void nametbl_list(struct print_buf *buf, u32 depth_info, depth = (depth_info & ~TIPC_NTQ_ALLTYPES); if (depth == 0) - return; + return 0; if (all_types) { /* display all entries in name table to specified depth */ - nametbl_header(buf, depth); + 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) { - nameseq_list(seq, buf, depth, seq->type, - lowbound, upbound, i); + hlist_for_each_entry(seq, seq_head, ns_list) { + ret += nameseq_list(seq, buf + ret, len - ret, + depth, seq->type, + lowbound, upbound, i); } } } else { /* display only the sequence that matches the specified type */ if (upbound < lowbound) { - tipc_printf(buf, "invalid name sequence specified\n"); - return; + ret += tipc_snprintf(buf + ret, len - ret, + "invalid name sequence specified\n"); + return ret; } - nametbl_header(buf, depth); + 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) { - nameseq_list(seq, buf, depth, type, - lowbound, upbound, i); + ret += nameseq_list(seq, buf + ret, len - ret, + depth, type, + lowbound, upbound, i); break; } } } + return ret; } -#define MAX_NAME_TBL_QUERY 32768 - struct sk_buff *tipc_nametbl_get(const void *req_tlv_area, int req_tlv_space) { struct sk_buff *buf; struct tipc_name_table_query *argv; struct tlv_desc *rep_tlv; - struct print_buf b; + char *pb; + int pb_len; int str_len; if (!TLV_CHECK(req_tlv_area, req_tlv_space, TIPC_TLV_NAME_TBL_QUERY)) return tipc_cfg_reply_error_string(TIPC_CFG_TLV_ERROR); - buf = tipc_cfg_reply_alloc(TLV_SPACE(MAX_NAME_TBL_QUERY)); + buf = tipc_cfg_reply_alloc(TLV_SPACE(ULTRA_STRING_MAX_LEN)); if (!buf) return NULL; rep_tlv = (struct tlv_desc *)buf->data; - tipc_printbuf_init(&b, TLV_DATA(rep_tlv), MAX_NAME_TBL_QUERY); + pb = TLV_DATA(rep_tlv); + pb_len = ULTRA_STRING_MAX_LEN; argv = (struct tipc_name_table_query *)TLV_DATA(req_tlv_area); read_lock_bh(&tipc_nametbl_lock); - nametbl_list(&b, ntohl(argv->depth), ntohl(argv->type), - ntohl(argv->lowbound), ntohl(argv->upbound)); + str_len = nametbl_list(pb, pb_len, ntohl(argv->depth), + ntohl(argv->type), + ntohl(argv->lowbound), ntohl(argv->upbound)); read_unlock_bh(&tipc_nametbl_lock); - str_len = tipc_printbuf_validate(&b); - + str_len += 1; /* for "\0" */ skb_put(buf, TLV_SPACE(str_len)); TLV_SET(rep_tlv, TIPC_TLV_ULTRA_STRING, NULL, str_len); @@ -959,7 +940,7 @@ 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; @@ -968,22 +949,50 @@ int tipc_nametbl_init(void) return 0; } -void tipc_nametbl_stop(void) +/** + * tipc_purge_publications - remove all publications for a given type + * + * tipc_nametbl_lock must be held when calling this function + */ +static void tipc_purge_publications(struct name_seq *seq) { - u32 i; + struct publication *publ, *safe; + struct sub_seq *sseq; + struct name_info *info; - if (!table.types) + if (!seq->sseqs) { + nameseq_delete_empty(seq); return; + } + sseq = seq->sseqs; + info = sseq->info; + list_for_each_entry_safe(publ, safe, &info->zone_list, zone_list) { + tipc_nametbl_remove_publ(publ->type, publ->lower, publ->node, + publ->ref, publ->key); + kfree(publ); + } +} - /* Verify name table is empty, then release it */ +void tipc_nametbl_stop(void) +{ + u32 i; + struct name_seq *seq; + struct hlist_head *seq_head; + struct hlist_node *safe; + /* Verify name table is empty and purge any lingering + * publications, then release the name table + */ write_lock_bh(&tipc_nametbl_lock); - for (i = 0; i < tipc_nametbl_size; i++) { - if (!hlist_empty(&table.types[i])) - err("tipc_nametbl_stop(): hash chain %u is non-null\n", i); + for (i = 0; i < TIPC_NAMETBL_SIZE; i++) { + if (hlist_empty(&table.types[i])) + continue; + seq_head = &table.types[i]; + hlist_for_each_entry_safe(seq, safe, seq_head, ns_list) { + tipc_purge_publications(seq); + } } kfree(table.types); table.types = NULL; write_unlock_bh(&tipc_nametbl_lock); } - diff --git a/net/tipc/name_table.h b/net/tipc/name_table.h index 8086b42f92a..f02f48b9a21 100644 --- a/net/tipc/name_table.h +++ b/net/tipc/name_table.h @@ -45,10 +45,8 @@ struct tipc_port_list; /* * TIPC name types reserved for internal TIPC use (both current and planned) */ - #define TIPC_ZM_SRV 3 /* zone master service name type */ - /** * struct publication - info about a published (name or) name sequence * @type: name sequence type @@ -67,7 +65,6 @@ struct tipc_port_list; * * Note that the node list, cluster list, and zone list are circular lists. */ - struct publication { u32 type; u32 lower; @@ -90,16 +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); -int tipc_nametbl_publish_rsv(u32 ref, unsigned int scope, - struct tipc_name_seq const *seq); + 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 61afee7e829..f64375e7f99 100644 --- a/net/tipc/net.c +++ b/net/tipc/net.c @@ -39,45 +39,41 @@ #include "name_distr.h" #include "subscr.h" #include "port.h" +#include "socket.h" #include "node.h" #include "config.h" /* * The TIPC locking policy is designed to ensure a very fine locking * granularity, permitting complete parallel access to individual - * port and node/link instances. The code consists of three major + * port and node/link instances. The code consists of four major * locking domains, each protected with their own disjunct set of locks. * - * 1: The routing hierarchy. - * Comprises the structures 'zone', 'cluster', 'node', 'link' - * and 'bearer'. The whole hierarchy is protected by a big - * read/write lock, tipc_net_lock, to enssure that nothing is added - * or removed while code is accessing any of these structures. - * This layer must not be called from the two others while they - * hold any of their own locks. - * Neither must it itself do any upcalls to the other two before - * it has released tipc_net_lock and other protective locks. + * 1: The bearer level. + * RTNL lock is used to serialize the process of configuring bearer + * on update side, and RCU lock is applied on read side to make + * bearer instance valid on both paths of message transmission and + * reception. * - * Within the tipc_net_lock domain there are two sub-domains;'node' and - * 'bearer', where local write operations are permitted, - * provided that those are protected by individual spin_locks - * per instance. Code holding tipc_net_lock(read) and a node spin_lock - * is permitted to poke around in both the node itself and its - * subordinate links. I.e, it can update link counters and queues, - * change link state, send protocol messages, and alter the - * "active_links" array in the node; but it can _not_ remove a link - * or a node from the overall structure. - * Correspondingly, individual bearers may change status within a - * tipc_net_lock(read), protected by an individual spin_lock ber bearer - * instance, but it needs tipc_net_lock(write) to remove/add any bearers. + * 2: The node and link level. + * All node instances are saved into two tipc_node_list and node_htable + * lists. The two lists are protected by node_list_lock on write side, + * and they are guarded with RCU lock on read side. Especially node + * instance is destroyed only when TIPC module is removed, and we can + * confirm that there has no any user who is accessing the node at the + * moment. Therefore, Except for iterating the two lists within RCU + * protection, it's no needed to hold RCU that we access node instance + * in other places. * + * In addition, all members in node structure including link instances + * are protected by node spin lock. * - * 2: The transport level of the protocol. - * This consists of the structures port, (and its user level - * representations, such as user_port and tipc_sock), reference and - * tipc_user (port.c, reg.c, socket.c). + * 3: The transport level of the protocol. + * This consists of the structures port, (and its user level + * representations, such as user_port and tipc_sock), reference and + * tipc_user (port.c, reg.c, socket.c). * - * This layer has four different locks: + * This layer has four different locks: * - The tipc_port spin_lock. This is protecting each port instance * from parallel data access and removal. Since we can not place * this lock in the port itself, it has been placed in the @@ -96,7 +92,7 @@ * There are two such lists; 'port_list', which is used for management, * and 'wait_list', which is used to queue ports during congestion. * - * 3: The name table (name_table.c, name_distr.c, subscription.c) + * 4: The name table (name_table.c, name_distr.c, subscription.c) * - There is one big read/write-lock (tipc_nametbl_lock) protecting the * overall name table structure. Nothing must be added/removed to * this structure without holding write access to it. @@ -108,8 +104,6 @@ * - A local spin_lock protecting the queue of subscriber events. */ -DEFINE_RWLOCK(tipc_net_lock); - static void net_route_named_msg(struct sk_buff *buf) { struct tipc_msg *msg = buf_msg(buf); @@ -117,7 +111,7 @@ static void net_route_named_msg(struct sk_buff *buf) u32 dport; if (!msg_named(msg)) { - buf_discard(buf); + kfree_skb(buf); return; } @@ -146,69 +140,63 @@ void tipc_net_route_msg(struct sk_buff *buf) if (tipc_in_scope(dnode, tipc_own_addr)) { if (msg_isdata(msg)) { if (msg_mcast(msg)) - tipc_port_recv_mcast(buf, NULL); + tipc_port_mcast_rcv(buf, NULL); else if (msg_destport(msg)) - tipc_port_recv_msg(buf); + tipc_sk_rcv(buf); else net_route_named_msg(buf); return; } switch (msg_user(msg)) { case NAME_DISTRIBUTOR: - tipc_named_recv(buf); + tipc_named_rcv(buf); break; case CONN_MANAGER: - tipc_port_recv_proto_msg(buf); + tipc_port_proto_rcv(buf); break; default: - buf_discard(buf); + kfree_skb(buf); } return; } /* Handle message for another node */ skb_trim(buf, msg_size(msg)); - tipc_link_send(buf, dnode, msg_link_selector(msg)); + tipc_link_xmit(buf, dnode, msg_link_selector(msg)); } int tipc_net_start(u32 addr) { char addr_string[16]; - - if (tipc_mode != TIPC_NODE_MODE) - return -ENOPROTOOPT; - - tipc_subscr_stop(); - tipc_cfg_stop(); + int res; tipc_own_addr = addr; - tipc_mode = TIPC_NET_MODE; tipc_named_reinit(); tipc_port_reinit(); + res = tipc_bclink_init(); + if (res) + return res; - tipc_bclink_init(); - - tipc_k_signal((Handler)tipc_subscr_start, 0); - tipc_k_signal((Handler)tipc_cfg_init, 0); + tipc_nametbl_publish(TIPC_CFG_SRV, tipc_own_addr, tipc_own_addr, + TIPC_ZONE_SCOPE, 0, tipc_own_addr); - info("Started in network mode\n"); - info("Own node address %s, network identity %u\n", - tipc_addr_string_fill(addr_string, tipc_own_addr), tipc_net_id); + 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) { - struct tipc_node *node, *t_node; - - if (tipc_mode != TIPC_NET_MODE) + if (!tipc_own_addr) return; - write_lock_bh(&tipc_net_lock); - tipc_mode = TIPC_NODE_MODE; + + tipc_nametbl_withdraw(TIPC_CFG_SRV, tipc_own_addr, 0, tipc_own_addr); + rtnl_lock(); tipc_bearer_stop(); tipc_bclink_stop(); - list_for_each_entry_safe(node, t_node, &tipc_node_list, list) - tipc_node_delete(node); - write_unlock_bh(&tipc_net_lock); - info("Left network mode\n"); + tipc_node_stop(); + rtnl_unlock(); + + pr_info("Left network mode\n"); } diff --git a/net/tipc/net.h b/net/tipc/net.h index 9eb4b9e220e..c6c2b46f7c2 100644 --- a/net/tipc/net.h +++ b/net/tipc/net.h @@ -37,8 +37,6 @@ #ifndef _TIPC_NET_H #define _TIPC_NET_H -extern rwlock_t tipc_net_lock; - void tipc_net_route_msg(struct sk_buff *buf); int tipc_net_start(u32 addr); diff --git a/net/tipc/netlink.c b/net/tipc/netlink.c index 7bda8e3d139..ad844d36534 100644 --- a/net/tipc/netlink.c +++ b/net/tipc/netlink.c @@ -44,17 +44,17 @@ 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))) + if ((req_userhdr->cmd & 0xC000) && (!netlink_capable(skb, CAP_NET_ADMIN))) cmd = TIPC_CMD_NOT_NET_ADMIN; else 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; @@ -76,33 +76,26 @@ static struct genl_family tipc_genl_family = { .maxattr = 0, }; -static struct genl_ops tipc_genl_ops = { - .cmd = TIPC_GENL_CMD, - .doit = handle_cmd, +static struct genl_ops tipc_genl_ops[] = { + { + .cmd = TIPC_GENL_CMD, + .doit = handle_cmd, + }, }; -static int tipc_genl_family_registered; - int tipc_netlink_start(void) { int res; - res = genl_register_family_with_ops(&tipc_genl_family, - &tipc_genl_ops, 1); + res = genl_register_family_with_ops(&tipc_genl_family, tipc_genl_ops); if (res) { - err("Failed to register netlink interface\n"); + pr_err("Failed to register netlink interface\n"); return res; } - - tipc_genl_family_registered = 1; return 0; } void tipc_netlink_stop(void) { - if (!tipc_genl_family_registered) - return; - genl_unregister_family(&tipc_genl_family); - tipc_genl_family_registered = 0; } diff --git a/net/tipc/node.c b/net/tipc/node.c index 6b226faad89..5b44c3041be 100644 --- a/net/tipc/node.c +++ b/net/tipc/node.c @@ -1,8 +1,8 @@ /* * net/tipc/node.c: TIPC node management routines * - * Copyright (c) 2000-2006, Ericsson AB - * Copyright (c) 2005-2006, 2010-2011, Wind River Systems + * Copyright (c) 2000-2006, 2012 Ericsson AB + * Copyright (c) 2005-2006, 2010-2014, Wind River Systems * All rights reserved. * * Redistribution and use in source and binary forms, with or without @@ -39,63 +39,59 @@ #include "node.h" #include "name_distr.h" +#define NODE_HTABLE_SIZE 512 + static void node_lost_contact(struct tipc_node *n_ptr); static void node_established_contact(struct tipc_node *n_ptr); -static DEFINE_SPINLOCK(node_create_lock); - static struct hlist_head node_htable[NODE_HTABLE_SIZE]; LIST_HEAD(tipc_node_list); static u32 tipc_num_nodes; +static u32 tipc_num_links; +static DEFINE_SPINLOCK(node_list_lock); -static atomic_t tipc_num_links = ATOMIC_INIT(0); -u32 tipc_own_tag; +/* + * A trivial power-of-two bitmask technique is used for speed, since this + * operation is done for every incoming TIPC packet. The number of hash table + * entries has been chosen so that no hash chain exceeds 8 nodes and will + * usually be much smaller (typically only a single node). + */ +static unsigned int tipc_hashfn(u32 addr) +{ + return addr & (NODE_HTABLE_SIZE - 1); +} -/** +/* * tipc_node_find - locate specified node object, if it exists */ - struct tipc_node *tipc_node_find(u32 addr) { struct tipc_node *node; - struct hlist_node *pos; - if (unlikely(!in_own_cluster(addr))) + if (unlikely(!in_own_cluster_exact(addr))) return NULL; - hlist_for_each_entry(node, pos, &node_htable[tipc_hashfn(addr)], hash) { - if (node->addr == addr) + rcu_read_lock(); + hlist_for_each_entry_rcu(node, &node_htable[tipc_hashfn(addr)], hash) { + if (node->addr == addr) { + rcu_read_unlock(); return node; + } } + rcu_read_unlock(); return NULL; } -/** - * tipc_node_create - create neighboring node - * - * Currently, this routine is called by neighbor discovery code, which holds - * net_lock for reading only. We must take node_create_lock to ensure a node - * isn't created twice if two different bearers discover the node at the same - * time. (It would be preferable to switch to holding net_lock in write mode, - * but this is a non-trivial change.) - */ - struct tipc_node *tipc_node_create(u32 addr) { struct tipc_node *n_ptr, *temp_node; - spin_lock_bh(&node_create_lock); - - n_ptr = tipc_node_find(addr); - if (n_ptr) { - spin_unlock_bh(&node_create_lock); - return n_ptr; - } + spin_lock_bh(&node_list_lock); n_ptr = kzalloc(sizeof(*n_ptr), GFP_ATOMIC); if (!n_ptr) { - spin_unlock_bh(&node_create_lock); - warn("Node creation failed, no memory\n"); + spin_unlock_bh(&node_list_lock); + pr_warn("Node creation failed, no memory\n"); return NULL; } @@ -105,45 +101,56 @@ struct tipc_node *tipc_node_create(u32 addr) INIT_LIST_HEAD(&n_ptr->list); INIT_LIST_HEAD(&n_ptr->nsub); - hlist_add_head(&n_ptr->hash, &node_htable[tipc_hashfn(addr)]); + hlist_add_head_rcu(&n_ptr->hash, &node_htable[tipc_hashfn(addr)]); - list_for_each_entry(temp_node, &tipc_node_list, list) { + list_for_each_entry_rcu(temp_node, &tipc_node_list, list) { if (n_ptr->addr < temp_node->addr) break; } - list_add_tail(&n_ptr->list, &temp_node->list); - n_ptr->block_setup = WAIT_PEER_DOWN; + list_add_tail_rcu(&n_ptr->list, &temp_node->list); + n_ptr->action_flags = TIPC_WAIT_PEER_LINKS_DOWN; + n_ptr->signature = INVALID_NODE_SIG; tipc_num_nodes++; - spin_unlock_bh(&node_create_lock); + spin_unlock_bh(&node_list_lock); return n_ptr; } -void tipc_node_delete(struct tipc_node *n_ptr) +static void tipc_node_delete(struct tipc_node *n_ptr) { - list_del(&n_ptr->list); - hlist_del(&n_ptr->hash); - kfree(n_ptr); + list_del_rcu(&n_ptr->list); + hlist_del_rcu(&n_ptr->hash); + kfree_rcu(n_ptr, rcu); tipc_num_nodes--; } +void tipc_node_stop(void) +{ + struct tipc_node *node, *t_node; + + spin_lock_bh(&node_list_lock); + list_for_each_entry_safe(node, t_node, &tipc_node_list, list) + tipc_node_delete(node); + spin_unlock_bh(&node_list_lock); +} /** * tipc_node_link_up - handle addition of link * * Link becomes active (alone or shared) or standby, depending on its priority. */ - void tipc_node_link_up(struct tipc_node *n_ptr, struct tipc_link *l_ptr) { struct tipc_link **active = &n_ptr->active_links[0]; + u32 addr = n_ptr->addr; n_ptr->working_links++; - - info("Established link <%s> on network plane %c\n", - l_ptr->name, l_ptr->b_ptr->net_plane); + tipc_nametbl_publish(TIPC_LINK_STATE, addr, addr, TIPC_NODE_SCOPE, + l_ptr->bearer_id, addr); + pr_info("Established link <%s> on network plane %c\n", + l_ptr->name, l_ptr->net_plane); if (!active[0]) { active[0] = active[1] = l_ptr; @@ -151,24 +158,23 @@ void tipc_node_link_up(struct tipc_node *n_ptr, struct tipc_link *l_ptr) return; } if (l_ptr->priority < active[0]->priority) { - info("New link <%s> becomes standby\n", l_ptr->name); + pr_info("New link <%s> becomes standby\n", l_ptr->name); return; } - tipc_link_send_duplicate(active[0], l_ptr); + tipc_link_dup_queue_xmit(active[0], l_ptr); if (l_ptr->priority == active[0]->priority) { active[0] = l_ptr; return; } - info("Old link <%s> becomes standby\n", active[0]->name); + pr_info("Old link <%s> becomes standby\n", active[0]->name); if (active[1] != active[0]) - info("Old link <%s> becomes standby\n", active[1]->name); + pr_info("Old link <%s> becomes standby\n", active[1]->name); active[0] = active[1] = l_ptr; } /** * node_select_active_links - select active link */ - static void node_select_active_links(struct tipc_node *n_ptr) { struct tipc_link **active = &n_ptr->active_links[0]; @@ -196,20 +202,21 @@ static void node_select_active_links(struct tipc_node *n_ptr) /** * tipc_node_link_down - handle loss of link */ - void tipc_node_link_down(struct tipc_node *n_ptr, struct tipc_link *l_ptr) { struct tipc_link **active; + u32 addr = n_ptr->addr; n_ptr->working_links--; + tipc_nametbl_withdraw(TIPC_LINK_STATE, addr, l_ptr->bearer_id, addr); if (!tipc_link_is_active(l_ptr)) { - info("Lost standby link <%s> on network plane %c\n", - l_ptr->name, l_ptr->b_ptr->net_plane); + pr_info("Lost standby link <%s> on network plane %c\n", + l_ptr->name, l_ptr->net_plane); return; } - info("Lost link <%s> on network plane %c\n", - l_ptr->name, l_ptr->b_ptr->net_plane); + pr_info("Lost link <%s> on network plane %c\n", + l_ptr->name, l_ptr->net_plane); active = &n_ptr->active_links[0]; if (active[0] == l_ptr) @@ -219,7 +226,7 @@ void tipc_node_link_down(struct tipc_node *n_ptr, struct tipc_link *l_ptr) if (active[0] == l_ptr) node_select_active_links(n_ptr); if (tipc_node_is_up(n_ptr)) - tipc_link_changeover(l_ptr); + tipc_link_failover_send_queue(l_ptr); else node_lost_contact(n_ptr); } @@ -229,11 +236,6 @@ int tipc_node_active_links(struct tipc_node *n_ptr) return n_ptr->active_links[0] != NULL; } -int tipc_node_redundant_links(struct tipc_node *n_ptr) -{ - return n_ptr->working_links > 1; -} - int tipc_node_is_up(struct tipc_node *n_ptr) { return tipc_node_active_links(n_ptr); @@ -241,90 +243,34 @@ int tipc_node_is_up(struct tipc_node *n_ptr) void tipc_node_attach_link(struct tipc_node *n_ptr, struct tipc_link *l_ptr) { - n_ptr->links[l_ptr->b_ptr->identity] = l_ptr; - atomic_inc(&tipc_num_links); + n_ptr->links[l_ptr->bearer_id] = l_ptr; + spin_lock_bh(&node_list_lock); + tipc_num_links++; + spin_unlock_bh(&node_list_lock); n_ptr->link_cnt++; } void tipc_node_detach_link(struct tipc_node *n_ptr, struct tipc_link *l_ptr) { - n_ptr->links[l_ptr->b_ptr->identity] = NULL; - atomic_dec(&tipc_num_links); - n_ptr->link_cnt--; -} + int i; -/* - * Routing table management - five cases to handle: - * - * 1: A link towards a zone/cluster external node comes up. - * => Send a multicast message updating routing tables of all - * system nodes within own cluster that the new destination - * can be reached via this node. - * (node.establishedContact()=>cluster.multicastNewRoute()) - * - * 2: A link towards a slave node comes up. - * => Send a multicast message updating routing tables of all - * system nodes within own cluster that the new destination - * can be reached via this node. - * (node.establishedContact()=>cluster.multicastNewRoute()) - * => Send a message to the slave node about existence - * of all system nodes within cluster: - * (node.establishedContact()=>cluster.sendLocalRoutes()) - * - * 3: A new cluster local system node becomes available. - * => Send message(s) to this particular node containing - * information about all cluster external and slave - * nodes which can be reached via this node. - * (node.establishedContact()==>network.sendExternalRoutes()) - * (node.establishedContact()==>network.sendSlaveRoutes()) - * => Send messages to all directly connected slave nodes - * containing information about the existence of the new node - * (node.establishedContact()=>cluster.multicastNewRoute()) - * - * 4: The link towards a zone/cluster external node or slave - * node goes down. - * => Send a multcast message updating routing tables of all - * nodes within cluster that the new destination can not any - * longer be reached via this node. - * (node.lostAllLinks()=>cluster.bcastLostRoute()) - * - * 5: A cluster local system node becomes unavailable. - * => Remove all references to this node from the local - * routing tables. Note: This is a completely node - * local operation. - * (node.lostAllLinks()=>network.removeAsRouter()) - * => Send messages to all directly connected slave nodes - * containing information about loss of the node - * (node.establishedContact()=>cluster.multicastLostRoute()) - * - */ - -static void node_established_contact(struct tipc_node *n_ptr) -{ - tipc_k_signal((Handler)tipc_named_node_up, n_ptr->addr); - - /* Syncronize broadcast acks */ - n_ptr->bclink.acked = tipc_bclink_get_last_sent(); - - if (n_ptr->bclink.supported) { - tipc_bclink_add_node(n_ptr->addr); - if (n_ptr->addr < tipc_own_addr) - tipc_own_tag++; + for (i = 0; i < MAX_BEARERS; i++) { + if (l_ptr != n_ptr->links[i]) + continue; + n_ptr->links[i] = NULL; + spin_lock_bh(&node_list_lock); + tipc_num_links--; + spin_unlock_bh(&node_list_lock); + n_ptr->link_cnt--; } } -static void node_name_purge_complete(unsigned long node_addr) +static void node_established_contact(struct tipc_node *n_ptr) { - struct tipc_node *n_ptr; - - read_lock_bh(&tipc_net_lock); - n_ptr = tipc_node_find(node_addr); - if (n_ptr) { - tipc_node_lock(n_ptr); - n_ptr->block_setup &= ~WAIT_NAMES_GONE; - tipc_node_unlock(n_ptr); - } - read_unlock_bh(&tipc_net_lock); + n_ptr->action_flags |= TIPC_NOTIFY_NODE_UP; + 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_lost_contact(struct tipc_node *n_ptr) @@ -332,30 +278,23 @@ static void node_lost_contact(struct tipc_node *n_ptr) char addr_string[16]; u32 i; - info("Lost contact with %s\n", - tipc_addr_string_fill(addr_string, n_ptr->addr)); + pr_info("Lost contact with %s\n", + tipc_addr_string_fill(addr_string, n_ptr->addr)); /* Flush broadcast link info associated with lost node */ + if (n_ptr->bclink.recv_permitted) { + kfree_skb_list(n_ptr->bclink.deferred_head); + n_ptr->bclink.deferred_size = 0; - if (n_ptr->bclink.supported) { - n_ptr->bclink.gap_after = n_ptr->bclink.gap_to = 0; - while (n_ptr->bclink.deferred_head) { - struct sk_buff *buf = n_ptr->bclink.deferred_head; - n_ptr->bclink.deferred_head = buf->next; - buf_discard(buf); - } - - if (n_ptr->bclink.defragm) { - buf_discard(n_ptr->bclink.defragm); - n_ptr->bclink.defragm = NULL; + if (n_ptr->bclink.reasm_buf) { + kfree_skb(n_ptr->bclink.reasm_buf); + n_ptr->bclink.reasm_buf = NULL; } tipc_bclink_remove_node(n_ptr->addr); tipc_bclink_acknowledge(n_ptr, INVALID_LINK_SEQ); - if (n_ptr->addr < tipc_own_addr) - tipc_own_tag--; - n_ptr->bclink.supported = 0; + n_ptr->bclink.recv_permitted = false; } /* Abort link changeover */ @@ -368,13 +307,13 @@ static void node_lost_contact(struct tipc_node *n_ptr) tipc_link_reset_fragments(l_ptr); } - /* Notify subscribers */ - tipc_nodesub_notify(n_ptr); - - /* Prevent re-contact with node until cleanup is done */ + n_ptr->action_flags &= ~TIPC_WAIT_OWN_LINKS_DOWN; - n_ptr->block_setup = WAIT_PEER_DOWN | WAIT_NAMES_GONE; - tipc_k_signal((Handler)node_name_purge_complete, n_ptr->addr); + /* Notify subscribers and prevent re-contact with node until + * cleanup is done. + */ + n_ptr->action_flags |= TIPC_WAIT_PEER_LINKS_DOWN | + TIPC_NOTIFY_NODE_DOWN; } struct sk_buff *tipc_node_get_nodes(const void *req_tlv_area, int req_tlv_space) @@ -393,29 +332,28 @@ struct sk_buff *tipc_node_get_nodes(const void *req_tlv_area, int req_tlv_space) return tipc_cfg_reply_error_string(TIPC_CFG_INVALID_VALUE " (network address)"); - read_lock_bh(&tipc_net_lock); + spin_lock_bh(&node_list_lock); if (!tipc_num_nodes) { - read_unlock_bh(&tipc_net_lock); + spin_unlock_bh(&node_list_lock); return tipc_cfg_reply_none(); } /* For now, get space for all other nodes */ - payload_size = TLV_SPACE(sizeof(node_info)) * tipc_num_nodes; if (payload_size > 32768u) { - read_unlock_bh(&tipc_net_lock); + spin_unlock_bh(&node_list_lock); return tipc_cfg_reply_error_string(TIPC_CFG_NOT_SUPPORTED " (too many nodes)"); } + spin_unlock_bh(&node_list_lock); + buf = tipc_cfg_reply_alloc(payload_size); - if (!buf) { - read_unlock_bh(&tipc_net_lock); + if (!buf) return NULL; - } /* Add TLVs for all nodes in scope */ - - list_for_each_entry(n_ptr, &tipc_node_list, list) { + rcu_read_lock(); + list_for_each_entry_rcu(n_ptr, &tipc_node_list, list) { if (!tipc_in_scope(domain, n_ptr->addr)) continue; node_info.addr = htonl(n_ptr->addr); @@ -423,8 +361,7 @@ struct sk_buff *tipc_node_get_nodes(const void *req_tlv_area, int req_tlv_space) tipc_cfg_append_tlv(buf, TIPC_TLV_NODE_INFO, &node_info, sizeof(node_info)); } - - read_unlock_bh(&tipc_net_lock); + rcu_read_unlock(); return buf; } @@ -444,36 +381,32 @@ struct sk_buff *tipc_node_get_links(const void *req_tlv_area, int req_tlv_space) return tipc_cfg_reply_error_string(TIPC_CFG_INVALID_VALUE " (network address)"); - if (tipc_mode != TIPC_NET_MODE) + if (!tipc_own_addr) return tipc_cfg_reply_none(); - read_lock_bh(&tipc_net_lock); - - /* Get space for all unicast links + multicast link */ - - payload_size = TLV_SPACE(sizeof(link_info)) * - (atomic_read(&tipc_num_links) + 1); + spin_lock_bh(&node_list_lock); + /* Get space for all unicast links + broadcast link */ + payload_size = TLV_SPACE((sizeof(link_info)) * (tipc_num_links + 1)); if (payload_size > 32768u) { - read_unlock_bh(&tipc_net_lock); + spin_unlock_bh(&node_list_lock); return tipc_cfg_reply_error_string(TIPC_CFG_NOT_SUPPORTED " (too many links)"); } + spin_unlock_bh(&node_list_lock); + buf = tipc_cfg_reply_alloc(payload_size); - if (!buf) { - read_unlock_bh(&tipc_net_lock); + if (!buf) return NULL; - } /* Add TLV for broadcast link */ - link_info.dest = htonl(tipc_cluster_mask(tipc_own_addr)); link_info.up = htonl(1); strlcpy(link_info.str, tipc_bclink_name, TIPC_MAX_LINK_NAME); tipc_cfg_append_tlv(buf, TIPC_TLV_LINK_INFO, &link_info, sizeof(link_info)); /* Add TLVs for any other links in scope */ - - list_for_each_entry(n_ptr, &tipc_node_list, list) { + rcu_read_lock(); + list_for_each_entry_rcu(n_ptr, &tipc_node_list, list) { u32 i; if (!tipc_in_scope(domain, n_ptr->addr)) @@ -490,7 +423,66 @@ struct sk_buff *tipc_node_get_links(const void *req_tlv_area, int req_tlv_space) } tipc_node_unlock(n_ptr); } - - read_unlock_bh(&tipc_net_lock); + rcu_read_unlock(); return buf; } + +/** + * tipc_node_get_linkname - get the name of a link + * + * @bearer_id: id of the bearer + * @node: peer node address + * @linkname: link name output buffer + * + * Returns 0 on success + */ +int tipc_node_get_linkname(u32 bearer_id, u32 addr, char *linkname, size_t len) +{ + struct tipc_link *link; + struct tipc_node *node = tipc_node_find(addr); + + if ((bearer_id >= MAX_BEARERS) || !node) + return -EINVAL; + tipc_node_lock(node); + link = node->links[bearer_id]; + if (link) { + strncpy(linkname, link->name, len); + tipc_node_unlock(node); + return 0; + } + tipc_node_unlock(node); + return -EINVAL; +} + +void tipc_node_unlock(struct tipc_node *node) +{ + LIST_HEAD(nsub_list); + struct tipc_link *link; + int pkt_sz = 0; + u32 addr = 0; + + if (likely(!node->action_flags)) { + spin_unlock_bh(&node->lock); + return; + } + + if (node->action_flags & TIPC_NOTIFY_NODE_DOWN) { + list_replace_init(&node->nsub, &nsub_list); + node->action_flags &= ~TIPC_NOTIFY_NODE_DOWN; + } + if (node->action_flags & TIPC_NOTIFY_NODE_UP) { + link = node->active_links[0]; + node->action_flags &= ~TIPC_NOTIFY_NODE_UP; + if (link) { + pkt_sz = ((link->max_pkt - INT_H_SIZE) / ITEM_SIZE) * + ITEM_SIZE; + addr = node->addr; + } + } + spin_unlock_bh(&node->lock); + + if (!list_empty(&nsub_list)) + tipc_nodesub_notify(&nsub_list); + if (pkt_sz) + tipc_named_node_up(pkt_sz, addr); +} diff --git a/net/tipc/node.h b/net/tipc/node.h index 0b1c5f8b699..9087063793f 100644 --- a/net/tipc/node.h +++ b/net/tipc/node.h @@ -2,7 +2,7 @@ * net/tipc/node.h: Include file for TIPC node management routines * * Copyright (c) 2000-2006, Ericsson AB - * Copyright (c) 2005, 2010-2011, Wind River Systems + * Copyright (c) 2005, 2010-2014, Wind River Systems * All rights reserved. * * Redistribution and use in source and binary forms, with or without @@ -42,99 +42,105 @@ #include "net.h" #include "bearer.h" -/* Flags used to block (re)establishment of contact with a neighboring node */ +/* + * Out-of-range value for node signature + */ +#define INVALID_NODE_SIG 0x10000 -#define WAIT_PEER_DOWN 0x0001 /* wait to see that peer's links are down */ -#define WAIT_NAMES_GONE 0x0002 /* wait for peer's publications to be purged */ -#define WAIT_NODE_DOWN 0x0004 /* wait until peer node is declared down */ +/* Flags used to take different actions according to flag type + * TIPC_WAIT_PEER_LINKS_DOWN: wait to see that peer's links are down + * TIPC_WAIT_OWN_LINKS_DOWN: wait until peer node is declared down + * TIPC_NOTIFY_NODE_DOWN: notify node is down + * TIPC_NOTIFY_NODE_UP: notify node is up + */ +enum { + TIPC_WAIT_PEER_LINKS_DOWN = (1 << 1), + TIPC_WAIT_OWN_LINKS_DOWN = (1 << 2), + TIPC_NOTIFY_NODE_DOWN = (1 << 3), + TIPC_NOTIFY_NODE_UP = (1 << 4) +}; + +/** + * struct tipc_node_bclink - TIPC node bclink structure + * @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 + * @oos_state: state tracker for handling OOS b'cast messages + * @deferred_size: number of OOS b'cast messages in deferred queue + * @deferred_head: oldest OOS b'cast message received from node + * @deferred_tail: newest OOS b'cast message received from node + * @reasm_buf: broadcast reassembly queue head from node + * @recv_permitted: true if node is allowed to receive b'cast messages + */ +struct tipc_node_bclink { + u32 acked; + u32 last_in; + u32 last_sent; + u32 oos_state; + u32 deferred_size; + struct sk_buff *deferred_head; + struct sk_buff *deferred_tail; + struct sk_buff *reasm_buf; + bool recv_permitted; +}; /** * struct tipc_node - TIPC node structure * @addr: network address of node * @lock: spinlock governing access to structure * @hash: links to adjacent nodes in unsorted hash chain - * @list: links to adjacent nodes in sorted list of cluster's nodes - * @nsub: list of "node down" subscriptions monitoring node * @active_links: pointers to active links to node * @links: pointers to all links to node + * @action_flags: bit mask of different types of node actions + * @bclink: broadcast-related info + * @list: links to adjacent nodes in sorted list of cluster's nodes * @working_links: number of working links to node (both active and standby) - * @block_setup: bit mask of conditions preventing link establishment to node * @link_cnt: number of links to node - * @permit_changeover: non-zero if node has redundant links to this system - * @bclink: broadcast-related info - * @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 - * @gap_after: sequence # of last message not requiring a NAK request - * @gap_to: sequence # of last message requiring a NAK request - * @nack_sync: counter that determines when NAK requests should be sent - * @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 + * @signature: node instance identifier + * @nsub: list of "node down" subscriptions monitoring node + * @rcu: rcu struct for tipc_node */ - struct tipc_node { u32 addr; spinlock_t lock; struct hlist_node hash; - struct list_head list; - struct list_head nsub; struct tipc_link *active_links[2]; struct tipc_link *links[MAX_BEARERS]; + unsigned int action_flags; + struct tipc_node_bclink bclink; + struct list_head list; int link_cnt; int working_links; - int block_setup; - int permit_changeover; - struct { - int supported; - u32 acked; - u32 last_in; - u32 gap_after; - u32 gap_to; - u32 nack_sync; - struct sk_buff *deferred_head; - struct sk_buff *deferred_tail; - struct sk_buff *defragm; - } bclink; + u32 signature; + struct list_head nsub; + struct rcu_head rcu; }; -#define NODE_HTABLE_SIZE 512 extern struct list_head tipc_node_list; -/* - * A trivial power-of-two bitmask technique is used for speed, since this - * operation is done for every incoming TIPC packet. The number of hash table - * entries has been chosen so that no hash chain exceeds 8 nodes and will - * usually be much smaller (typically only a single node). - */ -static inline unsigned int tipc_hashfn(u32 addr) -{ - return addr & (NODE_HTABLE_SIZE - 1); -} - -extern u32 tipc_own_tag; - struct tipc_node *tipc_node_find(u32 addr); struct tipc_node *tipc_node_create(u32 addr); -void tipc_node_delete(struct tipc_node *n_ptr); +void tipc_node_stop(void); void tipc_node_attach_link(struct tipc_node *n_ptr, struct tipc_link *l_ptr); void tipc_node_detach_link(struct tipc_node *n_ptr, struct tipc_link *l_ptr); void tipc_node_link_down(struct tipc_node *n_ptr, struct tipc_link *l_ptr); void tipc_node_link_up(struct tipc_node *n_ptr, struct tipc_link *l_ptr); int tipc_node_active_links(struct tipc_node *n_ptr); -int tipc_node_redundant_links(struct tipc_node *n_ptr); int tipc_node_is_up(struct tipc_node *n_ptr); struct sk_buff *tipc_node_get_links(const void *req_tlv_area, int req_tlv_space); struct sk_buff *tipc_node_get_nodes(const void *req_tlv_area, int req_tlv_space); +int tipc_node_get_linkname(u32 bearer_id, u32 node, char *linkname, size_t len); +void tipc_node_unlock(struct tipc_node *node); -static inline void tipc_node_lock(struct tipc_node *n_ptr) +static inline void tipc_node_lock(struct tipc_node *node) { - spin_lock_bh(&n_ptr->lock); + spin_lock_bh(&node->lock); } -static inline void tipc_node_unlock(struct tipc_node *n_ptr) +static inline bool tipc_node_blocked(struct tipc_node *node) { - spin_unlock_bh(&n_ptr->lock); + return (node->action_flags & (TIPC_WAIT_PEER_LINKS_DOWN | + TIPC_NOTIFY_NODE_DOWN | TIPC_WAIT_OWN_LINKS_DOWN)); } #endif diff --git a/net/tipc/node_subscr.c b/net/tipc/node_subscr.c index c3c2815ae63..7c59ab1d6ec 100644 --- a/net/tipc/node_subscr.c +++ b/net/tipc/node_subscr.c @@ -41,18 +41,18 @@ /** * 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 (addr == tipc_own_addr) { + if (in_own_node(addr)) { node_sub->node = NULL; return; } node_sub->node = tipc_node_find(addr); if (!node_sub->node) { - warn("Node subscription rejected, unknown node 0x%x\n", addr); + pr_warn("Node subscription rejected, unknown node 0x%x\n", + addr); return; } node_sub->handle_node_down = handle_down; @@ -66,7 +66,6 @@ void tipc_nodesub_subscribe(struct tipc_node_subscr *node_sub, u32 addr, /** * tipc_nodesub_unsubscribe - cancel "node down" subscription (if any) */ - void tipc_nodesub_unsubscribe(struct tipc_node_subscr *node_sub) { if (!node_sub->node) @@ -82,15 +81,13 @@ void tipc_nodesub_unsubscribe(struct tipc_node_subscr *node_sub) * * Note: node is locked by caller */ - -void tipc_nodesub_notify(struct tipc_node *node) +void tipc_nodesub_notify(struct list_head *nsub_list) { - struct tipc_node_subscr *ns; + struct tipc_node_subscr *ns, *safe; - list_for_each_entry(ns, &node->nsub, nodesub_list) { + list_for_each_entry_safe(ns, safe, nsub_list, nodesub_list) { if (ns->handle_node_down) { - tipc_k_signal((Handler)ns->handle_node_down, - (unsigned long)ns->usr_handle); + ns->handle_node_down(ns->usr_handle); ns->handle_node_down = NULL; } } diff --git a/net/tipc/node_subscr.h b/net/tipc/node_subscr.h index 4bc2ca0867a..d91b8cc81e3 100644 --- a/net/tipc/node_subscr.h +++ b/net/tipc/node_subscr.h @@ -48,7 +48,6 @@ typedef void (*net_ev_handler) (void *usr_handle); * @usr_handle: argument to pass to routine when node fails * @nodesub_list: adjacent entries in list of subscriptions for the node */ - struct tipc_node_subscr { struct tipc_node *node; net_ev_handler handle_node_down; @@ -59,6 +58,6 @@ struct tipc_node_subscr { void tipc_nodesub_subscribe(struct tipc_node_subscr *node_sub, u32 addr, void *usr_handle, net_ev_handler handle_down); void tipc_nodesub_unsubscribe(struct tipc_node_subscr *node_sub); -void tipc_nodesub_notify(struct tipc_node *node); +void tipc_nodesub_notify(struct list_head *nsub_list); #endif diff --git a/net/tipc/port.c b/net/tipc/port.c index d91efc69e6f..5fd7acce01e 100644 --- a/net/tipc/port.c +++ b/net/tipc/port.c @@ -1,8 +1,8 @@ /* * net/tipc/port.c: TIPC port code * - * Copyright (c) 1992-2007, Ericsson AB - * Copyright (c) 2004-2008, 2010-2011, Wind River Systems + * Copyright (c) 1992-2007, 2014, 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 @@ -38,6 +38,7 @@ #include "config.h" #include "port.h" #include "name_table.h" +#include "socket.h" /* Connection management: */ #define PROBING_INTERVAL 3600000 /* [ms] => 1 h */ @@ -46,11 +47,7 @@ #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); @@ -58,38 +55,44 @@ static struct sk_buff *port_build_self_abort_msg(struct tipc_port *, u32 err); static struct sk_buff *port_build_peer_abort_msg(struct tipc_port *, u32 err); static void port_timeout(unsigned long ref); - -static u32 port_peernode(struct tipc_port *p_ptr) +/** + * tipc_port_peer_msg - verify message was sent by connected port's peer + * + * Handles cases where the node's network address has changed from + * the default of <0.0.0> to its configured setting. + */ +int tipc_port_peer_msg(struct tipc_port *p_ptr, struct tipc_msg *msg) { - return msg_destnode(&p_ptr->phdr); -} + u32 peernode; + u32 orignode; -static u32 port_peerport(struct tipc_port *p_ptr) -{ - return msg_destport(&p_ptr->phdr); + if (msg_origport(msg) != tipc_port_peerport(p_ptr)) + return 0; + + orignode = msg_orignode(msg); + peernode = tipc_port_peernode(p_ptr); + return (orignode == peernode) || + (!orignode && (peernode == tipc_own_addr)) || + (!peernode && (orignode == tipc_own_addr)); } /** - * tipc_multicast - send a multicast message to local and remote destinations + * tipc_port_mcast_xmit - send a multicast message to local and remote + * destinations */ - -int tipc_multicast(u32 ref, struct tipc_name_seq const *seq, - u32 num_sect, struct iovec const *msg_sect, - unsigned int total_len) +int tipc_port_mcast_xmit(struct tipc_port *oport, + struct tipc_name_seq const *seq, + struct iovec const *msg_sect, + unsigned int len) { struct tipc_msg *hdr; struct sk_buff *buf; struct sk_buff *ibuf = NULL; struct tipc_port_list dports = {0, NULL, }; - struct tipc_port *oport = tipc_port_deref(ref); int ext_targets; int res; - if (unlikely(!oport)) - return -EINVAL; - /* Create multicast message */ - hdr = &oport->phdr; msg_set_type(hdr, TIPC_MCAST_MSG); msg_set_lookup_scope(hdr, TIPC_CLUSTER_SCOPE); @@ -99,37 +102,34 @@ int tipc_multicast(u32 ref, struct tipc_name_seq const *seq, msg_set_namelower(hdr, seq->lower); 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); + res = tipc_msg_build(hdr, msg_sect, len, MAX_MSG_SIZE, &buf); if (unlikely(!buf)) return res; /* Figure out where to send multicast message */ - ext_targets = tipc_nametbl_mc_translate(seq->type, seq->lower, seq->upper, TIPC_NODE_SCOPE, &dports); /* Send message to destinations (duplicate it only if necessary) */ - if (ext_targets) { if (dports.count != 0) { ibuf = skb_copy(buf, GFP_ATOMIC); if (ibuf == NULL) { tipc_port_list_free(&dports); - buf_discard(buf); + kfree_skb(buf); return -ENOMEM; } } - res = tipc_bclink_send_msg(buf); + res = tipc_bclink_xmit(buf); if ((res < 0) && (dports.count != 0)) - buf_discard(ibuf); + kfree_skb(ibuf); } else { ibuf = buf; } if (res >= 0) { if (ibuf) - tipc_port_recv_mcast(ibuf, &dports); + tipc_port_mcast_rcv(ibuf, &dports); } else { tipc_port_list_free(&dports); } @@ -137,12 +137,11 @@ int tipc_multicast(u32 ref, struct tipc_name_seq const *seq, } /** - * tipc_port_recv_mcast - deliver multicast message to all destination ports + * tipc_port_mcast_rcv - deliver multicast message to all destination ports * * If there is no port list, perform a lookup to create one */ - -void tipc_port_recv_mcast(struct sk_buff *buf, struct tipc_port_list *dp) +void tipc_port_mcast_rcv(struct sk_buff *buf, struct tipc_port_list *dp) { struct tipc_msg *msg; struct tipc_port_list dports = {0, NULL, }; @@ -152,7 +151,6 @@ void tipc_port_recv_mcast(struct sk_buff *buf, struct tipc_port_list *dp) msg = buf_msg(buf); /* Create destination port list, if one wasn't supplied */ - if (dp == NULL) { tipc_nametbl_mc_translate(msg_nametype(msg), msg_namelower(msg), @@ -163,12 +161,11 @@ void tipc_port_recv_mcast(struct sk_buff *buf, struct tipc_port_list *dp) } /* Deliver a copy of message to each destination port */ - if (dp->count != 0) { msg_set_destnode(msg, tipc_own_addr); if (dp->count == 1) { msg_set_destport(msg, dp->ports[0]); - tipc_port_recv_msg(buf); + tipc_sk_rcv(buf); tipc_port_list_free(dp); return; } @@ -177,153 +174,86 @@ void tipc_port_recv_mcast(struct sk_buff *buf, struct tipc_port_list *dp) struct sk_buff *b = skb_clone(buf, GFP_ATOMIC); if (b == NULL) { - warn("Unable to deliver multicast message(s)\n"); + pr_warn("Unable to deliver multicast message(s)\n"); goto exit; } if ((index == 0) && (cnt != 0)) item = item->next; msg_set_destport(buf_msg(b), item->ports[index]); - tipc_port_recv_msg(b); + tipc_sk_rcv(b); } } exit: - buf_discard(buf); + kfree_skb(buf); tipc_port_list_free(dp); } -/** - * tipc_createport_raw - create a generic TIPC port + +void tipc_port_wakeup(struct tipc_port *port) +{ + tipc_sock_wakeup(tipc_port_to_sock(port)); +} + +/* tipc_port_init - intiate TIPC port and lock it * - * Returns pointer to (locked) TIPC port, or NULL if unable to create it + * Returns obtained reference if initialization is successful, zero otherwise */ - -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) +u32 tipc_port_init(struct tipc_port *p_ptr, + const unsigned int importance) { - struct tipc_port *p_ptr; struct tipc_msg *msg; u32 ref; - p_ptr = kzalloc(sizeof(*p_ptr), GFP_ATOMIC); - if (!p_ptr) { - warn("Port creation failed, no memory\n"); - return NULL; - } ref = tipc_ref_acquire(p_ptr, &p_ptr->lock); if (!ref) { - warn("Port creation failed, reference table exhausted\n"); - kfree(p_ptr); - return NULL; + pr_warn("Port registration failed, ref. table exhausted\n"); + return 0; } - p_ptr->usr_handle = usr_handle; p_ptr->max_pkt = MAX_PKT_DEFAULT; p_ptr->ref = ref; - msg = &p_ptr->phdr; - tipc_msg_init(msg, importance, TIPC_NAMED_MSG, NAMED_H_SIZE, 0); - msg_set_origport(msg, 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); - spin_lock_bh(&tipc_port_list_lock); INIT_LIST_HEAD(&p_ptr->publications); INIT_LIST_HEAD(&p_ptr->port_list); + + /* + * Must hold port list lock while initializing message header template + * to ensure a change to node's own network address doesn't result + * in template containing out-dated network address information + */ + spin_lock_bh(&tipc_port_list_lock); + msg = &p_ptr->phdr; + tipc_msg_init(msg, importance, TIPC_NAMED_MSG, NAMED_H_SIZE, 0); + msg_set_origport(msg, ref); list_add_tail(&p_ptr->port_list, &ports); spin_unlock_bh(&tipc_port_list_lock); - return p_ptr; + return ref; } -int tipc_deleteport(u32 ref) +void tipc_port_destroy(struct tipc_port *p_ptr) { - struct tipc_port *p_ptr; struct sk_buff *buf = NULL; - tipc_withdraw(ref, 0, NULL); - p_ptr = tipc_port_lock(ref); - if (!p_ptr) - return -EINVAL; + tipc_withdraw(p_ptr, 0, NULL); - tipc_ref_discard(ref); - tipc_port_unlock(p_ptr); + spin_lock_bh(p_ptr->lock); + tipc_ref_discard(p_ptr->ref); + spin_unlock_bh(p_ptr->lock); k_cancel_timer(&p_ptr->timer); if (p_ptr->connected) { 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); list_del(&p_ptr->wait_list); spin_unlock_bh(&tipc_port_list_lock); k_term_timer(&p_ptr->timer); - kfree(p_ptr); tipc_net_route_msg(buf); - return 0; -} - -static int port_unreliable(struct tipc_port *p_ptr) -{ - return msg_src_droppable(&p_ptr->phdr); -} - -int tipc_portunreliable(u32 ref, unsigned int *isunreliable) -{ - struct tipc_port *p_ptr; - - p_ptr = tipc_port_lock(ref); - if (!p_ptr) - return -EINVAL; - *isunreliable = port_unreliable(p_ptr); - tipc_port_unlock(p_ptr); - return 0; -} - -int tipc_set_portunreliable(u32 ref, unsigned int isunreliable) -{ - struct tipc_port *p_ptr; - - p_ptr = tipc_port_lock(ref); - if (!p_ptr) - return -EINVAL; - msg_set_src_droppable(&p_ptr->phdr, (isunreliable != 0)); - tipc_port_unlock(p_ptr); - return 0; -} - -static int port_unreturnable(struct tipc_port *p_ptr) -{ - return msg_dest_droppable(&p_ptr->phdr); -} - -int tipc_portunreturnable(u32 ref, unsigned int *isunrejectable) -{ - struct tipc_port *p_ptr; - - p_ptr = tipc_port_lock(ref); - if (!p_ptr) - return -EINVAL; - *isunrejectable = port_unreturnable(p_ptr); - tipc_port_unlock(p_ptr); - return 0; -} - -int tipc_set_portunreturnable(u32 ref, unsigned int isunrejectable) -{ - struct tipc_port *p_ptr; - - p_ptr = tipc_port_lock(ref); - if (!p_ptr) - return -EINVAL; - msg_set_dest_droppable(&p_ptr->phdr, (isunrejectable != 0)); - tipc_port_unlock(p_ptr); - return 0; } /* @@ -341,8 +271,8 @@ static struct sk_buff *port_build_proto_msg(struct tipc_port *p_ptr, if (buf) { msg = buf_msg(buf); tipc_msg_init(msg, CONN_MANAGER, type, INT_H_SIZE, - port_peernode(p_ptr)); - msg_set_destport(msg, port_peerport(p_ptr)); + tipc_port_peernode(p_ptr)); + msg_set_destport(msg, tipc_port_peerport(p_ptr)); msg_set_origport(msg, p_ptr->ref); msg_set_msgcnt(msg, ack); } @@ -361,7 +291,6 @@ int tipc_reject_msg(struct sk_buff *buf, u32 err) u32 rmsg_sz; /* discard rejected message if it shouldn't be returned to sender */ - if (WARN(!msg_isdata(msg), "attempt to reject message with user=%u", msg_user(msg))) { dump_stack(); @@ -374,7 +303,6 @@ int tipc_reject_msg(struct sk_buff *buf, u32 err) * construct returned message by copying rejected message header and * data (or subset), then updating header fields that need adjusting */ - hdr_sz = msg_hdr_sz(msg); rmsg_sz = hdr_sz + min_t(u32, data_sz, MAX_REJECT_SIZE); @@ -400,38 +328,37 @@ int tipc_reject_msg(struct sk_buff *buf, u32 err) /* send self-abort message when rejecting on a connected port */ if (msg_connected(msg)) { - struct sk_buff *abuf = NULL; struct tipc_port *p_ptr = tipc_port_lock(msg_destport(msg)); if (p_ptr) { + struct sk_buff *abuf = NULL; + if (p_ptr->connected) abuf = port_build_self_abort_msg(p_ptr, err); tipc_port_unlock(p_ptr); + tipc_net_route_msg(abuf); } - tipc_net_route_msg(abuf); } /* send returned message & dispose of rejected message */ - src_node = msg_prevnode(msg); - if (src_node == tipc_own_addr) - tipc_port_recv_msg(rbuf); + if (in_own_node(src_node)) + tipc_sk_rcv(rbuf); else - tipc_link_send(rbuf, src_node, msg_link_selector(rmsg)); + tipc_link_xmit(rbuf, src_node, msg_link_selector(rmsg)); exit: - buf_discard(buf); + kfree_skb(buf); return data_sz; } -int tipc_port_reject_sections(struct tipc_port *p_ptr, struct tipc_msg *hdr, - struct iovec const *msg_sect, u32 num_sect, - unsigned int total_len, int err) +int tipc_port_iovec_reject(struct tipc_port *p_ptr, struct tipc_msg *hdr, + struct iovec const *msg_sect, unsigned int len, + int err) { struct sk_buff *buf; int res; - res = tipc_msg_build(hdr, msg_sect, num_sect, total_len, MAX_MSG_SIZE, - !p_ptr->user_port, &buf); + res = tipc_msg_build(hdr, msg_sect, len, MAX_MSG_SIZE, &buf); if (!buf) return res; @@ -513,30 +440,25 @@ static struct sk_buff *port_build_peer_abort_msg(struct tipc_port *p_ptr, u32 er return buf; } -void tipc_port_recv_proto_msg(struct sk_buff *buf) +void tipc_port_proto_rcv(struct sk_buff *buf) { struct tipc_msg *msg = buf_msg(buf); struct tipc_port *p_ptr; struct sk_buff *r_buf = NULL; - u32 orignode = msg_orignode(msg); - u32 origport = msg_origport(msg); u32 destport = msg_destport(msg); int wakeable; /* Validate connection */ - p_ptr = tipc_port_lock(destport); - if (!p_ptr || !p_ptr->connected || - (port_peernode(p_ptr) != orignode) || - (port_peerport(p_ptr) != origport)) { + if (!p_ptr || !p_ptr->connected || !tipc_port_peer_msg(p_ptr, msg)) { r_buf = tipc_buf_acquire(BASIC_H_SIZE); if (r_buf) { msg = buf_msg(r_buf); tipc_msg_init(msg, TIPC_HIGH_IMPORTANCE, TIPC_CONN_MSG, - BASIC_H_SIZE, orignode); + BASIC_H_SIZE, msg_orignode(msg)); msg_set_errcode(msg, TIPC_ERR_NO_PORT); msg_set_origport(msg, destport); - msg_set_destport(msg, origport); + msg_set_destport(msg, msg_origport(msg)); } if (p_ptr) tipc_port_unlock(p_ptr); @@ -544,16 +466,14 @@ void tipc_port_recv_proto_msg(struct sk_buff *buf) } /* Process protocol message sent by peer */ - switch (msg_type(msg)) { case CONN_ACK: - wakeable = tipc_port_congested(p_ptr) && p_ptr->congested && - p_ptr->wakeup; + wakeable = tipc_port_congested(p_ptr) && p_ptr->congested; p_ptr->acked += msg_msgcnt(msg); if (!tipc_port_congested(p_ptr)) { p_ptr->congested = 0; if (wakeable) - p_ptr->wakeup(p_ptr); + tipc_port_wakeup(p_ptr); } break; case CONN_PROBE: @@ -567,70 +487,76 @@ void tipc_port_recv_proto_msg(struct sk_buff *buf) tipc_port_unlock(p_ptr); exit: tipc_net_route_msg(r_buf); - buf_discard(buf); + kfree_skb(buf); } -static void port_print(struct tipc_port *p_ptr, struct print_buf *buf, int full_id) +static int port_print(struct tipc_port *p_ptr, char *buf, int len, int full_id) { struct publication *publ; + int ret; if (full_id) - tipc_printf(buf, "<%u.%u.%u:%u>:", - tipc_zone(tipc_own_addr), tipc_cluster(tipc_own_addr), - tipc_node(tipc_own_addr), p_ptr->ref); + ret = tipc_snprintf(buf, len, "<%u.%u.%u:%u>:", + tipc_zone(tipc_own_addr), + tipc_cluster(tipc_own_addr), + tipc_node(tipc_own_addr), p_ptr->ref); else - tipc_printf(buf, "%-10u:", p_ptr->ref); + ret = tipc_snprintf(buf, len, "%-10u:", p_ptr->ref); if (p_ptr->connected) { - u32 dport = port_peerport(p_ptr); - u32 destnode = port_peernode(p_ptr); - - tipc_printf(buf, " connected to <%u.%u.%u:%u>", - tipc_zone(destnode), tipc_cluster(destnode), - tipc_node(destnode), dport); + u32 dport = tipc_port_peerport(p_ptr); + u32 destnode = tipc_port_peernode(p_ptr); + + ret += tipc_snprintf(buf + ret, len - ret, + " connected to <%u.%u.%u:%u>", + tipc_zone(destnode), + tipc_cluster(destnode), + tipc_node(destnode), dport); if (p_ptr->conn_type != 0) - tipc_printf(buf, " via {%u,%u}", - p_ptr->conn_type, - p_ptr->conn_instance); + ret += tipc_snprintf(buf + ret, len - ret, + " via {%u,%u}", p_ptr->conn_type, + p_ptr->conn_instance); } else if (p_ptr->published) { - tipc_printf(buf, " bound to"); + ret += tipc_snprintf(buf + ret, len - ret, " bound to"); list_for_each_entry(publ, &p_ptr->publications, pport_list) { if (publ->lower == publ->upper) - tipc_printf(buf, " {%u,%u}", publ->type, - publ->lower); + ret += tipc_snprintf(buf + ret, len - ret, + " {%u,%u}", publ->type, + publ->lower); else - tipc_printf(buf, " {%u,%u,%u}", publ->type, - publ->lower, publ->upper); + ret += tipc_snprintf(buf + ret, len - ret, + " {%u,%u,%u}", publ->type, + publ->lower, publ->upper); } } - tipc_printf(buf, "\n"); + ret += tipc_snprintf(buf + ret, len - ret, "\n"); + return ret; } -#define MAX_PORT_QUERY 32768 - struct sk_buff *tipc_port_get_ports(void) { struct sk_buff *buf; struct tlv_desc *rep_tlv; - struct print_buf pb; + char *pb; + int pb_len; struct tipc_port *p_ptr; - int str_len; + int str_len = 0; - buf = tipc_cfg_reply_alloc(TLV_SPACE(MAX_PORT_QUERY)); + buf = tipc_cfg_reply_alloc(TLV_SPACE(ULTRA_STRING_MAX_LEN)); if (!buf) return NULL; rep_tlv = (struct tlv_desc *)buf->data; + pb = TLV_DATA(rep_tlv); + pb_len = ULTRA_STRING_MAX_LEN; - tipc_printbuf_init(&pb, TLV_DATA(rep_tlv), MAX_PORT_QUERY); spin_lock_bh(&tipc_port_list_lock); list_for_each_entry(p_ptr, &ports, port_list) { spin_lock_bh(p_ptr->lock); - port_print(p_ptr, &pb, 0); + str_len += port_print(p_ptr, pb, pb_len, 0); spin_unlock_bh(p_ptr->lock); } spin_unlock_bh(&tipc_port_list_lock); - str_len = tipc_printbuf_validate(&pb); - + str_len += 1; /* for "\0" */ skb_put(buf, TLV_SPACE(str_len)); TLV_SET(rep_tlv, TIPC_TLV_ULTRA_STRING, NULL, str_len); @@ -645,233 +571,12 @@ void tipc_port_reinit(void) spin_lock_bh(&tipc_port_list_lock); list_for_each_entry(p_ptr, &ports, port_list) { msg = &p_ptr->phdr; - if (msg_orignode(msg) == tipc_own_addr) - break; msg_set_prevnode(msg, tipc_own_addr); msg_set_orignode(msg, tipc_own_addr); } 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 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; - 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 peer_port = port_peerport(p_ptr); - u32 peer_node = port_peernode(p_ptr); - u32 dsz; - - tipc_port_unlock(p_ptr); - if (unlikely(!cb)) - goto reject; - if (unlikely(!connected)) { - if (tipc_connect2port(dref, &orig)) - goto reject; - } else if ((msg_origport(msg) != peer_port) || - (msg_orignode(msg) != peer_node)) - 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) - buf_discard(buf); - buf = next; - continue; -err: - switch (message_type) { - - case TIPC_CONN_MSG:{ - tipc_conn_shutdown_event cb = - up_ptr->conn_err_cb; - u32 peer_port = port_peerport(p_ptr); - u32 peer_node = port_peernode(p_ptr); - - tipc_port_unlock(p_ptr); - if (!cb || !connected) - break; - if ((msg_origport(msg) != peer_port) || - (msg_orignode(msg) != peer_node)) - 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) - buf_discard(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; @@ -888,124 +593,36 @@ 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) { - warn("Port creation failed, no memory\n"); - return -ENOMEM; - } - p_ptr = (struct tipc_port *)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) +int tipc_publish(struct tipc_port *p_ptr, unsigned int scope, + struct tipc_name_seq const *seq) { - struct tipc_port *p_ptr; - - p_ptr = tipc_port_lock(ref); - if (!p_ptr) - return -EINVAL; - *importance = (unsigned int)msg_importance(&p_ptr->phdr); - tipc_port_unlock(p_ptr); - return 0; -} - -int tipc_set_portimportance(u32 ref, unsigned int imp) -{ - struct tipc_port *p_ptr; - - if (imp > TIPC_CRITICAL_IMPORTANCE) - return -EINVAL; - - p_ptr = tipc_port_lock(ref); - if (!p_ptr) - return -EINVAL; - msg_set_importance(&p_ptr->phdr, (u32)imp); - tipc_port_unlock(p_ptr); - return 0; -} - - -int tipc_publish(u32 ref, unsigned int scope, struct tipc_name_seq const *seq) -{ - struct tipc_port *p_ptr; struct publication *publ; u32 key; - int res = -EINVAL; - p_ptr = tipc_port_lock(ref); - if (!p_ptr) + if (p_ptr->connected) return -EINVAL; + key = p_ptr->ref + p_ptr->pub_count + 1; + if (key == p_ptr->ref) + return -EADDRINUSE; - if (p_ptr->connected) - goto exit; - if (seq->lower > seq->upper) - goto exit; - if ((scope < TIPC_ZONE_SCOPE) || (scope > TIPC_NODE_SCOPE)) - goto exit; - key = ref + p_ptr->pub_count + 1; - if (key == ref) { - res = -EADDRINUSE; - goto exit; - } publ = tipc_nametbl_publish(seq->type, seq->lower, seq->upper, scope, p_ptr->ref, key); if (publ) { list_add(&publ->pport_list, &p_ptr->publications); p_ptr->pub_count++; p_ptr->published = 1; - res = 0; + return 0; } -exit: - tipc_port_unlock(p_ptr); - return res; + return -EINVAL; } -int tipc_withdraw(u32 ref, unsigned int scope, struct tipc_name_seq const *seq) +int tipc_withdraw(struct tipc_port *p_ptr, unsigned int scope, + struct tipc_name_seq const *seq) { - struct tipc_port *p_ptr; struct publication *publ; struct publication *tpubl; int res = -EINVAL; - p_ptr = tipc_port_lock(ref); - if (!p_ptr) - return -EINVAL; if (!seq) { list_for_each_entry_safe(publ, tpubl, &p_ptr->publications, pport_list) { @@ -1032,19 +649,33 @@ int tipc_withdraw(u32 ref, unsigned int scope, struct tipc_name_seq const *seq) } if (list_empty(&p_ptr->publications)) p_ptr->published = 0; - tipc_port_unlock(p_ptr); return res; } -int tipc_connect2port(u32 ref, struct tipc_portid const *peer) +int tipc_port_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_port_connect(ref, p_ptr, peer); + tipc_port_unlock(p_ptr); + return res; +} + +/* + * __tipc_port_connect - connect to a remote peer + * + * Port must be locked. + */ +int __tipc_port_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) @@ -1053,8 +684,6 @@ int tipc_connect2port(u32 ref, struct tipc_portid const *peer) msg = &p_ptr->phdr; msg_set_destnode(msg, peer->node); msg_set_destport(msg, peer->ref); - msg_set_orignode(msg, tipc_own_addr); - msg_set_origport(msg, p_ptr->ref); msg_set_type(msg, TIPC_CONN_MSG); msg_set_lookup_scope(msg, 0); msg_set_hdr_sz(msg, SHORT_H_SIZE); @@ -1069,39 +698,32 @@ 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_port_disconnect(struct tipc_port *tp_ptr) { - int res; - if (tp_ptr->connected) { tp_ptr->connected = 0; /* let timer expire on it's own to avoid deadlock! */ - tipc_nodesub_unsubscribe( - &((struct tipc_port *)tp_ptr)->subscription); - res = 0; - } else { - res = -ENOTCONN; + tipc_nodesub_unsubscribe(&tp_ptr->subscription); + return 0; } - return res; + + return -ENOTCONN; } /* - * tipc_disconnect(): Disconnect port form peer. + * tipc_port_disconnect(): Disconnect port form peer. * This is a node local operation. */ - -int tipc_disconnect(u32 ref) +int tipc_port_disconnect(u32 ref) { struct tipc_port *p_ptr; int res; @@ -1109,15 +731,15 @@ int tipc_disconnect(u32 ref) p_ptr = tipc_port_lock(ref); if (!p_ptr) return -EINVAL; - res = tipc_disconnect_port((struct tipc_port *)p_ptr); + res = __tipc_port_disconnect(p_ptr); tipc_port_unlock(p_ptr); return res; } /* - * tipc_shutdown(): Send a SHUTDOWN msg to peer and disconnect + * tipc_port_shutdown(): Send a SHUTDOWN msg to peer and disconnect */ -int tipc_shutdown(u32 ref) +int tipc_port_shutdown(u32 ref) { struct tipc_port *p_ptr; struct sk_buff *buf = NULL; @@ -1129,52 +751,47 @@ int tipc_shutdown(u32 ref) buf = port_build_peer_abort_msg(p_ptr, TIPC_CONN_SHUTDOWN); tipc_port_unlock(p_ptr); tipc_net_route_msg(buf); - return tipc_disconnect(ref); + return tipc_port_disconnect(ref); } /* - * tipc_port_recv_sections(): Concatenate and deliver sectioned - * message for this node. + * tipc_port_iovec_rcv: Concatenate and deliver sectioned + * message for this node. */ - -static int tipc_port_recv_sections(struct tipc_port *sender, unsigned int num_sect, - struct iovec const *msg_sect, - unsigned int total_len) +static int tipc_port_iovec_rcv(struct tipc_port *sender, + struct iovec const *msg_sect, + unsigned int len) { struct sk_buff *buf; int res; - res = tipc_msg_build(&sender->phdr, msg_sect, num_sect, total_len, - MAX_MSG_SIZE, !sender->user_port, &buf); + res = tipc_msg_build(&sender->phdr, msg_sect, len, MAX_MSG_SIZE, &buf); if (likely(buf)) - tipc_port_recv_msg(buf); + tipc_sk_rcv(buf); return res; } /** * tipc_send - send message sections on connection */ - -int tipc_send(u32 ref, unsigned int num_sect, struct iovec const *msg_sect, - unsigned int total_len) +int tipc_send(struct tipc_port *p_ptr, + struct iovec const *msg_sect, + unsigned int len) { - struct tipc_port *p_ptr; u32 destnode; int res; - p_ptr = tipc_port_deref(ref); - if (!p_ptr || !p_ptr->connected) + if (!p_ptr->connected) return -EINVAL; p_ptr->congested = 1; if (!tipc_port_congested(p_ptr)) { - destnode = port_peernode(p_ptr); - if (likely(destnode != tipc_own_addr)) - res = tipc_link_send_sections_fast(p_ptr, msg_sect, num_sect, - total_len, destnode); + destnode = tipc_port_peernode(p_ptr); + if (likely(!in_own_node(destnode))) + res = tipc_link_iovec_xmit_fast(p_ptr, msg_sect, len, + destnode); else - res = tipc_port_recv_sections(p_ptr, num_sect, msg_sect, - total_len); + res = tipc_port_iovec_rcv(p_ptr, msg_sect, len); if (likely(res != -ELINKCONG)) { p_ptr->congested = 0; @@ -1183,9 +800,9 @@ int tipc_send(u32 ref, unsigned int num_sect, struct iovec const *msg_sect, return res; } } - if (port_unreliable(p_ptr)) { + if (tipc_port_unreliable(p_ptr)) { p_ptr->congested = 0; - return total_len; + return len; } return -ELINKCONG; } @@ -1193,25 +810,22 @@ int tipc_send(u32 ref, unsigned int num_sect, struct iovec const *msg_sect, /** * tipc_send2name - send message sections to port name */ - -int tipc_send2name(u32 ref, struct tipc_name const *name, unsigned int domain, - unsigned int num_sect, struct iovec const *msg_sect, - unsigned int total_len) +int tipc_send2name(struct tipc_port *p_ptr, + struct tipc_name const *name, + unsigned int domain, + struct iovec const *msg_sect, + unsigned int len) { - struct tipc_port *p_ptr; struct tipc_msg *msg; u32 destnode = domain; u32 destport; int res; - p_ptr = tipc_port_deref(ref); - if (!p_ptr || p_ptr->connected) + if (p_ptr->connected) return -EINVAL; msg = &p_ptr->phdr; msg_set_type(msg, TIPC_NAMED_MSG); - msg_set_orignode(msg, tipc_own_addr); - msg_set_origport(msg, ref); msg_set_hdr_sz(msg, NAMED_H_SIZE); msg_set_nametype(msg, name->type); msg_set_nameinst(msg, name->instance); @@ -1220,110 +834,65 @@ int tipc_send2name(u32 ref, struct tipc_name const *name, unsigned int domain, msg_set_destnode(msg, destnode); msg_set_destport(msg, destport); - if (likely(destport)) { - if (likely(destnode == tipc_own_addr)) - res = tipc_port_recv_sections(p_ptr, num_sect, - msg_sect, total_len); + if (likely(destport || destnode)) { + if (likely(in_own_node(destnode))) + res = tipc_port_iovec_rcv(p_ptr, msg_sect, len); + else if (tipc_own_addr) + res = tipc_link_iovec_xmit_fast(p_ptr, msg_sect, len, + destnode); else - res = tipc_link_send_sections_fast(p_ptr, msg_sect, - num_sect, total_len, - destnode); + res = tipc_port_iovec_reject(p_ptr, msg, msg_sect, + len, TIPC_ERR_NO_NODE); if (likely(res != -ELINKCONG)) { if (res > 0) p_ptr->sent++; return res; } - if (port_unreliable(p_ptr)) { - return total_len; - } + if (tipc_port_unreliable(p_ptr)) + return len; + return -ELINKCONG; } - return tipc_port_reject_sections(p_ptr, msg, msg_sect, num_sect, - total_len, TIPC_ERR_NO_NAME); + return tipc_port_iovec_reject(p_ptr, msg, msg_sect, len, + TIPC_ERR_NO_NAME); } /** * tipc_send2port - send message sections to port identity */ - -int tipc_send2port(u32 ref, struct tipc_portid const *dest, - unsigned int num_sect, struct iovec const *msg_sect, - unsigned int total_len) +int tipc_send2port(struct tipc_port *p_ptr, + struct tipc_portid const *dest, + struct iovec const *msg_sect, + unsigned int len) { - struct tipc_port *p_ptr; struct tipc_msg *msg; int res; - p_ptr = tipc_port_deref(ref); - if (!p_ptr || p_ptr->connected) + if (p_ptr->connected) return -EINVAL; msg = &p_ptr->phdr; msg_set_type(msg, TIPC_DIRECT_MSG); msg_set_lookup_scope(msg, 0); - msg_set_orignode(msg, tipc_own_addr); - msg_set_origport(msg, ref); msg_set_destnode(msg, dest->node); msg_set_destport(msg, dest->ref); msg_set_hdr_sz(msg, BASIC_H_SIZE); - if (dest->node == tipc_own_addr) - res = tipc_port_recv_sections(p_ptr, num_sect, msg_sect, - total_len); + if (in_own_node(dest->node)) + res = tipc_port_iovec_rcv(p_ptr, msg_sect, len); + else if (tipc_own_addr) + res = tipc_link_iovec_xmit_fast(p_ptr, msg_sect, len, + dest->node); else - res = tipc_link_send_sections_fast(p_ptr, msg_sect, num_sect, - total_len, dest->node); + res = tipc_port_iovec_reject(p_ptr, msg, msg_sect, len, + TIPC_ERR_NO_NODE); if (likely(res != -ELINKCONG)) { if (res > 0) p_ptr->sent++; return res; } - if (port_unreliable(p_ptr)) { - return total_len; - } - return -ELINKCONG; -} + if (tipc_port_unreliable(p_ptr)) + return len; -/** - * 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_orignode(msg, tipc_own_addr); - msg_set_origport(msg, ref); - 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 (dest->node == tipc_own_addr) - 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 f751807e2a9..cf4ca5b1d9a 100644 --- a/net/tipc/port.h +++ b/net/tipc/port.h @@ -1,8 +1,8 @@ /* * 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) 1994-2007, 2014, Ericsson AB + * Copyright (c) 2004-2007, 2010-2013, Wind River Systems * All rights reserved. * * Redistribution and use in source and binary forms, with or without @@ -42,61 +42,13 @@ #include "msg.h" #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 TIPC_CONNACK_INTV 256 +#define TIPC_FLOWCTRL_WIN (TIPC_CONNACK_INTV * 2) +#define TIPC_CONN_OVERLOAD_LIMIT ((TIPC_FLOWCTRL_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 * @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 @@ -108,9 +60,6 @@ struct user_port { * @ref: unique reference to port in TIPC object registry * @phdr: preformatted message header used when sending messages * @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 +72,6 @@ struct user_port { * @subscription: "node down" subscription used to terminate failed connections */ struct tipc_port { - void *usr_handle; spinlock_t *lock; int connected; u32 conn_type; @@ -135,9 +83,6 @@ struct tipc_port { u32 ref; struct tipc_msg phdr; 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; @@ -156,85 +101,75 @@ 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); +u32 tipc_port_init(struct tipc_port *p_ptr, + const unsigned int 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); +void tipc_port_destroy(struct tipc_port *p_ptr); -int tipc_portimportance(u32 portref, unsigned int *importance); -int tipc_set_portimportance(u32 portref, unsigned int importance); +int tipc_publish(struct tipc_port *p_ptr, unsigned int scope, + struct tipc_name_seq const *name_seq); -int tipc_portunreliable(u32 portref, unsigned int *isunreliable); -int tipc_set_portunreliable(u32 portref, unsigned int isunreliable); +int tipc_withdraw(struct tipc_port *p_ptr, unsigned int scope, + struct tipc_name_seq const *name_seq); -int tipc_portunreturnable(u32 portref, unsigned int *isunreturnable); -int tipc_set_portunreturnable(u32 portref, unsigned int isunreturnable); +int tipc_port_connect(u32 portref, struct tipc_portid const *port); -int tipc_publish(u32 portref, unsigned int scope, - struct tipc_name_seq const *name_seq); -int tipc_withdraw(u32 portref, unsigned int scope, - struct tipc_name_seq const *name_seq); +int tipc_port_disconnect(u32 portref); -int tipc_connect2port(u32 portref, struct tipc_portid const *port); - -int tipc_disconnect(u32 portref); - -int tipc_shutdown(u32 ref); +int tipc_port_shutdown(u32 ref); +void tipc_port_wakeup(struct tipc_port *port); /* * The following routines require that the port be locked on entry */ -int tipc_disconnect_port(struct tipc_port *tp_ptr); +int __tipc_port_disconnect(struct tipc_port *tp_ptr); +int __tipc_port_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); /* * TIPC messaging routines */ -int tipc_send(u32 portref, unsigned int num_sect, struct iovec const *msg_sect, - unsigned int total_len); -int tipc_send2name(u32 portref, struct tipc_name const *name, u32 domain, - unsigned int num_sect, struct iovec const *msg_sect, - unsigned int total_len); +int tipc_send(struct tipc_port *port, + struct iovec const *msg_sect, + unsigned int len); + +int tipc_send2name(struct tipc_port *port, + struct tipc_name const *name, + u32 domain, + struct iovec const *msg_sect, + unsigned int len); -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_send2port(struct tipc_port *port, + struct tipc_portid const *dest, + struct iovec const *msg_sect, + unsigned int len); -int tipc_send_buf2port(u32 portref, struct tipc_portid const *dest, - struct sk_buff *buf, unsigned int dsz); +int tipc_port_mcast_xmit(struct tipc_port *port, + struct tipc_name_seq const *seq, + struct iovec const *msg, + unsigned int len); -int tipc_multicast(u32 portref, struct tipc_name_seq const *seq, - unsigned int section_count, struct iovec const *msg, - unsigned int total_len); +int tipc_port_iovec_reject(struct tipc_port *p_ptr, + struct tipc_msg *hdr, + struct iovec const *msg_sect, + unsigned int len, + int err); -int tipc_port_reject_sections(struct tipc_port *p_ptr, struct tipc_msg *hdr, - struct iovec const *msg_sect, u32 num_sect, - unsigned int total_len, int err); struct sk_buff *tipc_port_get_ports(void); -void tipc_port_recv_proto_msg(struct sk_buff *buf); -void tipc_port_recv_mcast(struct sk_buff *buf, struct tipc_port_list *dp); +void tipc_port_proto_rcv(struct sk_buff *buf); +void tipc_port_mcast_rcv(struct sk_buff *buf, struct tipc_port_list *dp); void tipc_port_reinit(void); /** * tipc_port_lock - lock port instance referred to and return its pointer */ - static inline struct tipc_port *tipc_port_lock(u32 ref) { return (struct tipc_port *)tipc_ref_lock(ref); @@ -245,71 +180,58 @@ static inline struct tipc_port *tipc_port_lock(u32 ref) * * Can use pointer instead of tipc_ref_unlock() since port is already locked. */ - static inline void tipc_port_unlock(struct tipc_port *p_ptr) { spin_unlock_bh(p_ptr->lock); } -static inline struct tipc_port *tipc_port_deref(u32 ref) +static inline int tipc_port_congested(struct tipc_port *p_ptr) +{ + return ((p_ptr->sent - p_ptr->acked) >= TIPC_FLOWCTRL_WIN); +} + + +static inline u32 tipc_port_peernode(struct tipc_port *p_ptr) { - return (struct tipc_port *)tipc_ref_deref(ref); + return msg_destnode(&p_ptr->phdr); } -static inline u32 tipc_peer_port(struct tipc_port *p_ptr) +static inline u32 tipc_port_peerport(struct tipc_port *p_ptr) { return msg_destport(&p_ptr->phdr); } -static inline u32 tipc_peer_node(struct tipc_port *p_ptr) +static inline bool tipc_port_unreliable(struct tipc_port *port) { - return msg_destnode(&p_ptr->phdr); + return msg_src_droppable(&port->phdr) != 0; } -static inline int tipc_port_congested(struct tipc_port *p_ptr) +static inline void tipc_port_set_unreliable(struct tipc_port *port, + bool unreliable) { - return (p_ptr->sent - p_ptr->acked) >= (TIPC_FLOW_CONTROL_WIN * 2); + msg_set_src_droppable(&port->phdr, unreliable ? 1 : 0); } -/** - * tipc_port_recv_msg - receive message from lower layer and deliver to port user - */ +static inline bool tipc_port_unreturnable(struct tipc_port *port) +{ + return msg_dest_droppable(&port->phdr) != 0; +} + +static inline void tipc_port_set_unreturnable(struct tipc_port *port, + bool unreturnable) +{ + msg_set_dest_droppable(&port->phdr, unreturnable ? 1 : 0); +} + + +static inline int tipc_port_importance(struct tipc_port *port) +{ + return msg_importance(&port->phdr); +} -static inline int tipc_port_recv_msg(struct sk_buff *buf) +static inline void tipc_port_set_importance(struct tipc_port *port, int imp) { - struct tipc_port *p_ptr; - struct tipc_msg *msg = buf_msg(buf); - u32 destport = msg_destport(msg); - u32 dsz = msg_data_sz(msg); - u32 err; - - /* forward unresolved named message */ - if (unlikely(!destport)) { - tipc_net_route_msg(buf); - return dsz; - } - - /* validate destination & pass to port, otherwise reject message */ - p_ptr = tipc_port_lock(destport); - if (likely(p_ptr)) { - if (likely(p_ptr->connected)) { - if ((unlikely(msg_origport(msg) != tipc_peer_port(p_ptr))) || - (unlikely(msg_orignode(msg) != tipc_peer_node(p_ptr))) || - (unlikely(!msg_connected(msg)))) { - err = TIPC_ERR_NO_PORT; - tipc_port_unlock(p_ptr); - goto reject; - } - } - err = p_ptr->dispatcher(p_ptr, buf); - tipc_port_unlock(p_ptr); - if (likely(!err)) - return dsz; - } else { - err = TIPC_ERR_NO_PORT; - } -reject: - return tipc_reject_msg(buf, err); + msg_set_importance(&port->phdr, (u32)imp); } #endif diff --git a/net/tipc/ref.c b/net/tipc/ref.c index 9e37b7812c3..3d4ecd754ee 100644 --- a/net/tipc/ref.c +++ b/net/tipc/ref.c @@ -43,7 +43,6 @@ * @lock: spinlock controlling access to object * @ref: reference value for object (combines instance & array index info) */ - struct reference { void *object; spinlock_t lock; @@ -60,7 +59,6 @@ struct reference { * @index_mask: bitmask for array index portion of reference values * @start_mask: initial value for instance value portion of reference values */ - struct ref_table { struct reference *entries; u32 capacity; @@ -91,12 +89,11 @@ struct ref_table { static struct ref_table tipc_ref_table; -static DEFINE_RWLOCK(ref_table_lock); +static DEFINE_SPINLOCK(ref_table_lock); /** * tipc_ref_table_init - create reference table for objects */ - int tipc_ref_table_init(u32 requested_size, u32 start) { struct reference *table; @@ -109,7 +106,6 @@ int tipc_ref_table_init(u32 requested_size, u32 start) /* do nothing */ ; /* allocate table & mark all entries as uninitialized */ - table = vzalloc(actual_size * sizeof(struct reference)); if (table == NULL) return -ENOMEM; @@ -128,12 +124,8 @@ int tipc_ref_table_init(u32 requested_size, u32 start) /** * tipc_ref_table_stop - destroy reference table for objects */ - void tipc_ref_table_stop(void) { - if (!tipc_ref_table.entries) - return; - vfree(tipc_ref_table.entries); tipc_ref_table.entries = NULL; } @@ -149,7 +141,6 @@ void tipc_ref_table_stop(void) * register a partially initialized object, without running the risk that * the object will be accessed before initialization is complete. */ - u32 tipc_ref_acquire(void *object, spinlock_t **lock) { u32 index; @@ -159,17 +150,16 @@ u32 tipc_ref_acquire(void *object, spinlock_t **lock) struct reference *entry = NULL; if (!object) { - err("Attempt to acquire reference to non-existent object\n"); + pr_err("Attempt to acquire ref. to non-existent obj\n"); return 0; } if (!tipc_ref_table.entries) { - err("Reference table not found during acquisition attempt\n"); + pr_err("Ref. table not found in acquisition attempt\n"); return 0; } /* take a free entry, if available; otherwise initialize a new entry */ - - write_lock_bh(&ref_table_lock); + spin_lock_bh(&ref_table_lock); if (tipc_ref_table.first_free) { index = tipc_ref_table.first_free; entry = &(tipc_ref_table.entries[index]); @@ -185,7 +175,7 @@ u32 tipc_ref_acquire(void *object, spinlock_t **lock) } else { ref = 0; } - write_unlock_bh(&ref_table_lock); + spin_unlock_bh(&ref_table_lock); /* * Grab the lock so no one else can modify this entry @@ -211,7 +201,6 @@ u32 tipc_ref_acquire(void *object, spinlock_t **lock) * Disallow future references to an object and free up the entry for re-use. * Note: The entry's spin_lock may still be busy after discard */ - void tipc_ref_discard(u32 ref) { struct reference *entry; @@ -219,7 +208,7 @@ void tipc_ref_discard(u32 ref) u32 index_mask; if (!tipc_ref_table.entries) { - err("Reference table not found during discard attempt\n"); + pr_err("Ref. table not found during discard attempt\n"); return; } @@ -227,14 +216,14 @@ void tipc_ref_discard(u32 ref) index = ref & index_mask; entry = &(tipc_ref_table.entries[index]); - write_lock_bh(&ref_table_lock); + spin_lock_bh(&ref_table_lock); if (!entry->object) { - err("Attempt to discard reference to non-existent object\n"); + pr_err("Attempt to discard ref. to non-existent obj\n"); goto exit; } if (entry->ref != ref) { - err("Attempt to discard non-existent reference\n"); + pr_err("Attempt to discard non-existent reference\n"); goto exit; } @@ -242,12 +231,10 @@ void tipc_ref_discard(u32 ref) * mark entry as unused; increment instance part of entry's reference * to invalidate any subsequent references */ - entry->object = NULL; entry->ref = (ref & ~index_mask) + (index_mask + 1); /* append entry to free entry list */ - if (tipc_ref_table.first_free == 0) tipc_ref_table.first_free = index; else @@ -255,13 +242,12 @@ void tipc_ref_discard(u32 ref) tipc_ref_table.last_free = index; exit: - write_unlock_bh(&ref_table_lock); + spin_unlock_bh(&ref_table_lock); } /** * tipc_ref_lock - lock referenced object and return pointer to it */ - void *tipc_ref_lock(u32 ref) { if (likely(tipc_ref_table.entries)) { @@ -278,22 +264,3 @@ void *tipc_ref_lock(u32 ref) } return NULL; } - - -/** - * tipc_ref_deref - return pointer referenced object (without locking it) - */ - -void *tipc_ref_deref(u32 ref) -{ - if (likely(tipc_ref_table.entries)) { - struct reference *entry; - - entry = &tipc_ref_table.entries[ref & - tipc_ref_table.index_mask]; - if (likely(entry->ref == ref)) - return entry->object; - } - return NULL; -} - diff --git a/net/tipc/ref.h b/net/tipc/ref.h index 5bc8e7ab84d..d01aa1df63b 100644 --- a/net/tipc/ref.h +++ b/net/tipc/ref.h @@ -44,6 +44,5 @@ u32 tipc_ref_acquire(void *object, spinlock_t **lock); void tipc_ref_discard(u32 ref); void *tipc_ref_lock(u32 ref); -void *tipc_ref_deref(u32 ref); #endif diff --git a/net/tipc/server.c b/net/tipc/server.c new file mode 100644 index 00000000000..a538a02f869 --- /dev/null +++ b/net/tipc/server.c @@ -0,0 +1,600 @@ +/* + * 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: control 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); + + if (con->sock) { + tipc_sock_release_local(con->sock); + con->sock = NULL; + } + + tipc_clean_outqueues(con); + 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) +{ + 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)) { + if (con->conid) + s->tipc_conn_shutdown(con->conid, con->usr_data); + + 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); + 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); + } else { + 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; + } + return ret; +} + +void tipc_server_stop(struct tipc_server *s) +{ + struct tipc_conn *con; + int total = 0; + int id; + + 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 00000000000..be817b0b547 --- /dev/null +++ b/net/tipc/server.h @@ -0,0 +1,92 @@ +/* + * 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 + */ +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 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 e2f7c5d370b..ef0475568f9 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, 2012-2014, 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 @@ -34,44 +34,29 @@ * POSSIBILITY OF SUCH DAMAGE. */ -#include <linux/export.h> -#include <net/sock.h> - #include "core.h" #include "port.h" +#include "node.h" + +#include <linux/export.h> #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 { - struct sock sk; - struct tipc_port *p; - struct tipc_portid peer_name; - unsigned int conn_timeout; -}; - -#define tipc_sk(sk) ((struct tipc_sock *)(sk)) -#define tipc_sk_port(sk) ((struct tipc_port *)(tipc_sk(sk)->p)) - -#define tipc_rx_ready(sock) (!skb_queue_empty(&sock->sk->sk_receive_queue) || \ - (sock->state == SS_DISCONNECTING)) - -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 int tipc_backlog_rcv(struct sock *sk, struct sk_buff *skb); +static void tipc_data_ready(struct sock *sk); +static void tipc_write_space(struct sock *sk); +static int tipc_release(struct socket *sock); +static int tipc_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 int sockets_enabled; - -static atomic_t tipc_queue_size = ATOMIC_INIT(0); +static struct proto tipc_proto_kern; /* * Revised TIPC socket locking policy: @@ -118,32 +103,16 @@ static atomic_t tipc_queue_size = ATOMIC_INIT(0); * - port reference */ +#include "socket.h" + /** * advance_rx_queue - discard first buffer in socket receive queue * * Caller must hold socket lock */ - static void advance_rx_queue(struct sock *sk) { - buf_discard(__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); - buf_discard(buf); - } + kfree_skb(__skb_dequeue(&sk->sk_receive_queue)); } /** @@ -151,19 +120,16 @@ static void discard_rx_queue(struct sock *sk) * * Caller must hold socket lock */ - 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) @@ -174,17 +140,17 @@ 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; struct sock *sk; - struct tipc_port *tp_ptr; + struct tipc_sock *tsk; + struct tipc_port *port; + u32 ref; /* Validate arguments */ - if (unlikely(protocol != 0)) return -EPROTONOSUPPORT; @@ -207,43 +173,116 @@ static int tipc_create(struct net *net, struct socket *sock, int protocol, } /* Allocate socket's protocol area */ + if (!kern) + sk = sk_alloc(net, AF_TIPC, GFP_KERNEL, &tipc_proto); + else + sk = sk_alloc(net, AF_TIPC, GFP_KERNEL, &tipc_proto_kern); - sk = sk_alloc(net, AF_TIPC, GFP_KERNEL, &tipc_proto); if (sk == NULL) return -ENOMEM; - /* Allocate TIPC port for socket to use */ + tsk = tipc_sk(sk); + port = &tsk->port; - tp_ptr = tipc_createport_raw(sk, &dispatch, &wakeupdispatch, - TIPC_LOW_IMPORTANCE); - if (unlikely(!tp_ptr)) { + ref = tipc_port_init(port, TIPC_LOW_IMPORTANCE); + if (!ref) { + pr_warn("Socket registration failed, ref. table exhausted\n"); sk_free(sk); return -ENOMEM; } /* Finish initializing socket data structures */ - sock->ops = ops; sock->state = state; sock_init_data(sock, sk); - sk->sk_backlog_rcv = backlog_rcv; - tipc_sk(sk)->p = tp_ptr; - tipc_sk(sk)->conn_timeout = CONN_TIMEOUT_DEFAULT; - - spin_unlock_bh(tp_ptr->lock); + sk->sk_backlog_rcv = tipc_backlog_rcv; + sk->sk_rcvbuf = sysctl_tipc_rmem[1]; + sk->sk_data_ready = tipc_data_ready; + sk->sk_write_space = tipc_write_space; + tsk->conn_timeout = CONN_TIMEOUT_DEFAULT; + atomic_set(&tsk->dupl_rcvcnt, 0); + tipc_port_unlock(port); if (sock->state == SS_READY) { - tipc_set_portunreturnable(tp_ptr->ref, 1); + tipc_port_set_unreturnable(port, true); if (sock->type == SOCK_DGRAM) - tipc_set_portunreliable(tp_ptr->ref, 1); + tipc_port_set_unreliable(port, true); + } + return 0; +} + +/** + * 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; + + 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); return 0; } /** - * release - destroy a TIPC socket + * 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) +{ + tipc_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 = tipc_accept(sock, *newsock, flags); + if (ret < 0) { + sock_release(*newsock); + return ret; + } + (*newsock)->ops = sock->ops; + return ret; +} + +/** + * tipc_release - destroy a TIPC socket * @sock: socket to destroy * * This routine cleans up any messages that are still queued on the socket. @@ -258,71 +297,64 @@ static int tipc_create(struct net *net, struct socket *sock, int protocol, * * Returns 0 on success, errno otherwise */ - -static int release(struct socket *sock) +static int tipc_release(struct socket *sock) { struct sock *sk = sock->sk; - struct tipc_port *tport; + struct tipc_sock *tsk; + struct tipc_port *port; struct sk_buff *buf; - int res; /* * Exit if socket isn't fully initialized (occurs when a failed accept() * releases a pre-allocated child socket that was never used) */ - if (sk == NULL) return 0; - tport = tipc_sk_port(sk); + tsk = tipc_sk(sk); + port = &tsk->port; lock_sock(sk); /* * Reject all unreceived messages, except on an active connection * (which disconnects locally & sends a 'FIN+' to peer) */ - while (sock->state != SS_DISCONNECTING) { buf = __skb_dequeue(&sk->sk_receive_queue); if (buf == NULL) break; - atomic_dec(&tipc_queue_size); - if (TIPC_SKB_CB(buf)->handle != 0) - buf_discard(buf); + if (TIPC_SKB_CB(buf)->handle != NULL) + kfree_skb(buf); else { if ((sock->state == SS_CONNECTING) || (sock->state == SS_CONNECTED)) { sock->state = SS_DISCONNECTING; - tipc_disconnect(tport->ref); + tipc_port_disconnect(port->ref); } tipc_reject_msg(buf, TIPC_ERR_NO_PORT); } } - /* - * Delete TIPC port; this ensures no more messages are queued - * (also disconnects an active connection & sends a 'FIN-' to peer) + /* Destroy TIPC port; also disconnects an active connection and + * sends a 'FIN-' to peer. */ - - res = tipc_deleteport(tport->ref); + tipc_port_destroy(port); /* 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; release_sock(sk); sock_put(sk); sock->sk = NULL; - return res; + return 0; } /** - * bind - associate or disassocate TIPC name(s) with a socket + * tipc_bind - associate or disassocate TIPC name(s) with a socket * @sock: socket structure * @uaddr: socket address describing name(s) and desired operation * @uaddr_len: size of socket address data structure @@ -336,32 +368,53 @@ static int release(struct socket *sock) * NOTE: This routine doesn't need to take the socket lock since it doesn't * access any non-constant socket information. */ - -static int bind(struct socket *sock, struct sockaddr *uaddr, int uaddr_len) +static int tipc_bind(struct socket *sock, struct sockaddr *uaddr, + int uaddr_len) { + struct sock *sk = sock->sk; struct sockaddr_tipc *addr = (struct sockaddr_tipc *)uaddr; - u32 portref = tipc_sk_port(sock->sk)->ref; + struct tipc_sock *tsk = tipc_sk(sk); + int res = -EINVAL; - if (unlikely(!uaddr_len)) - return tipc_withdraw(portref, 0, NULL); + lock_sock(sk); + if (unlikely(!uaddr_len)) { + res = tipc_withdraw(&tsk->port, 0, NULL); + goto exit; + } - if (uaddr_len < sizeof(struct sockaddr_tipc)) - return -EINVAL; - if (addr->family != AF_TIPC) - return -EAFNOSUPPORT; + if (uaddr_len < sizeof(struct sockaddr_tipc)) { + res = -EINVAL; + goto exit; + } + if (addr->family != AF_TIPC) { + res = -EAFNOSUPPORT; + goto exit; + } if (addr->addrtype == TIPC_ADDR_NAME) addr->addr.nameseq.upper = addr->addr.nameseq.lower; - else if (addr->addrtype != TIPC_ADDR_NAMESEQ) - return -EAFNOSUPPORT; + else if (addr->addrtype != TIPC_ADDR_NAMESEQ) { + res = -EAFNOSUPPORT; + goto exit; + } - return (addr->scope > 0) ? - tipc_publish(portref, addr->scope, &addr->addr.nameseq) : - tipc_withdraw(portref, -addr->scope, &addr->addr.nameseq); + if ((addr->addr.nameseq.type < TIPC_RESERVED_TYPES) && + (addr->addr.nameseq.type != TIPC_TOP_SRV) && + (addr->addr.nameseq.type != TIPC_CFG_SRV)) { + res = -EACCES; + goto exit; + } + + res = (addr->scope > 0) ? + tipc_publish(&tsk->port, addr->scope, &addr->addr.nameseq) : + tipc_withdraw(&tsk->port, -addr->scope, &addr->addr.nameseq); +exit: + release_sock(sk); + return res; } /** - * get_name - get port ID of socket or peer socket + * tipc_getname - get port ID of socket or peer socket * @sock: socket structure * @uaddr: area for returned socket address * @uaddr_len: area for returned length of socket address @@ -373,22 +426,21 @@ static int bind(struct socket *sock, struct sockaddr *uaddr, int uaddr_len) * accesses socket information that is unchanging (or which changes in * a completely predictable manner). */ - -static int get_name(struct socket *sock, struct sockaddr *uaddr, - int *uaddr_len, int peer) +static int tipc_getname(struct socket *sock, struct sockaddr *uaddr, + int *uaddr_len, int peer) { struct sockaddr_tipc *addr = (struct sockaddr_tipc *)uaddr; - struct tipc_sock *tsock = tipc_sk(sock->sk); + struct tipc_sock *tsk = tipc_sk(sock->sk); memset(addr, 0, sizeof(*addr)); if (peer) { if ((sock->state != SS_CONNECTED) && ((peer != 2) || (sock->state != SS_DISCONNECTING))) return -ENOTCONN; - addr->addr.id.ref = tsock->peer_name.ref; - addr->addr.id.node = tsock->peer_name.node; + addr->addr.id.ref = tipc_port_peerport(&tsk->port); + addr->addr.id.node = tipc_port_peernode(&tsk->port); } else { - addr->addr.id.ref = tsock->p->ref; + addr->addr.id.ref = tsk->port.ref; addr->addr.id.node = tipc_own_addr; } @@ -402,7 +454,7 @@ static int get_name(struct socket *sock, struct sockaddr *uaddr, } /** - * poll - read and possibly block on pollmask + * tipc_poll - read and possibly block on pollmask * @file: file structure associated with the socket * @sock: socket for which to calculate the poll bits * @wait: ??? @@ -420,7 +472,7 @@ 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 @@ -441,19 +493,23 @@ static int get_name(struct socket *sock, struct sockaddr *uaddr, * imply that the operation will succeed, merely that it should be performed * and will not block. */ - -static unsigned int poll(struct file *file, struct socket *sock, - poll_table *wait) +static unsigned int tipc_poll(struct file *file, struct socket *sock, + poll_table *wait) { struct sock *sk = sock->sk; + struct tipc_sock *tsk = tipc_sk(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 (!tsk->port.congested) + mask |= POLLOUT; + break; case SS_READY: case SS_CONNECTED: - if (!tipc_sk_port(sk)->congested) + if (!tsk->port.congested) mask |= POLLOUT; /* fall thru' */ case SS_CONNECTING: @@ -479,7 +535,6 @@ static unsigned int poll(struct file *file, struct socket *sock, * * Returns 0 if permission is granted, otherwise errno */ - static int dest_name_check(struct sockaddr_tipc *dest, struct msghdr *m) { struct tipc_cfg_msg_hdr hdr; @@ -501,8 +556,34 @@ static int dest_name_check(struct sockaddr_tipc *dest, struct msghdr *m) return 0; } +static int tipc_wait_for_sndmsg(struct socket *sock, long *timeo_p) +{ + struct sock *sk = sock->sk; + struct tipc_sock *tsk = tipc_sk(sk); + DEFINE_WAIT(wait); + int done; + + do { + int err = sock_error(sk); + if (err) + return err; + if (sock->state == SS_DISCONNECTING) + return -EPIPE; + if (!*timeo_p) + return -EAGAIN; + if (signal_pending(current)) + return sock_intr_errno(*timeo_p); + + prepare_to_wait(sk_sleep(sk), &wait, TASK_INTERRUPTIBLE); + done = sk_wait_event(sk, timeo_p, !tsk->port.congested); + finish_wait(sk_sleep(sk), &wait); + } while (!done); + return 0; +} + + /** - * send_msg - send message in connectionless manner + * tipc_sendmsg - send message in connectionless manner * @iocb: if NULL, indicates that socket lock is already held * @sock: socket structure * @m: message to send @@ -515,15 +596,15 @@ static int dest_name_check(struct sockaddr_tipc *dest, struct msghdr *m) * * Returns the number of bytes sent on success, or errno otherwise */ - -static int send_msg(struct kiocb *iocb, struct socket *sock, - struct msghdr *m, size_t total_len) +static int tipc_sendmsg(struct kiocb *iocb, struct socket *sock, + struct msghdr *m, size_t total_len) { struct sock *sk = sock->sk; - struct tipc_port *tport = tipc_sk_port(sk); - struct sockaddr_tipc *dest = (struct sockaddr_tipc *)m->msg_name; + struct tipc_sock *tsk = tipc_sk(sk); + struct tipc_port *port = &tsk->port; + DECLARE_SOCKADDR(struct sockaddr_tipc *, dest, m->msg_name); int needs_conn; - long timeout_val; + long timeo; int res = -EINVAL; if (unlikely(!dest)) @@ -531,8 +612,7 @@ 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_MAX)) + if (total_len > TIPC_MAX_USER_MSG_SIZE) return -EMSGSIZE; if (iocb) @@ -548,38 +628,33 @@ 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 (tsk->port.published) { res = -EOPNOTSUPP; goto exit; } if (dest->addrtype == TIPC_ADDR_NAME) { - tport->conn_type = dest->addr.name.name.type; - tport->conn_instance = dest->addr.name.name.instance; + tsk->port.conn_type = dest->addr.name.name.type; + tsk->port.conn_instance = dest->addr.name.name.instance; } /* Abort any pending connection attempts (very unlikely) */ - reject_rx_queue(sk); } - timeout_val = sock_sndtimeo(sk, m->msg_flags & MSG_DONTWAIT); - + timeo = sock_sndtimeo(sk, m->msg_flags & MSG_DONTWAIT); do { if (dest->addrtype == TIPC_ADDR_NAME) { res = dest_name_check(dest, m); if (res) break; - res = tipc_send2name(tport->ref, + res = tipc_send2name(port, &dest->addr.name.name, dest->addr.name.domain, - m->msg_iovlen, m->msg_iov, total_len); } else if (dest->addrtype == TIPC_ADDR_ID) { - res = tipc_send2port(tport->ref, + res = tipc_send2port(port, &dest->addr.id, - m->msg_iovlen, m->msg_iov, total_len); } else if (dest->addrtype == TIPC_ADDR_MCAST) { @@ -590,25 +665,19 @@ static int send_msg(struct kiocb *iocb, struct socket *sock, res = dest_name_check(dest, m); if (res) break; - res = tipc_multicast(tport->ref, - &dest->addr.nameseq, - m->msg_iovlen, - m->msg_iov, - total_len); + res = tipc_port_mcast_xmit(port, + &dest->addr.nameseq, + m->msg_iov, + total_len); } if (likely(res != -ELINKCONG)) { if (needs_conn && (res >= 0)) sock->state = SS_CONNECTING; break; } - if (timeout_val <= 0L) { - res = timeout_val ? timeout_val : -EWOULDBLOCK; + res = tipc_wait_for_sndmsg(sock, &timeo); + if (res) break; - } - release_sock(sk); - timeout_val = wait_event_interruptible_timeout(*sk_sleep(sk), - !tport->congested, timeout_val); - lock_sock(sk); } while (1); exit: @@ -617,8 +686,37 @@ exit: return res; } +static int tipc_wait_for_sndpkt(struct socket *sock, long *timeo_p) +{ + struct sock *sk = sock->sk; + struct tipc_sock *tsk = tipc_sk(sk); + struct tipc_port *port = &tsk->port; + DEFINE_WAIT(wait); + int done; + + do { + int err = sock_error(sk); + if (err) + return err; + if (sock->state == SS_DISCONNECTING) + return -EPIPE; + else if (sock->state != SS_CONNECTED) + return -ENOTCONN; + if (!*timeo_p) + return -EAGAIN; + if (signal_pending(current)) + return sock_intr_errno(*timeo_p); + + prepare_to_wait(sk_sleep(sk), &wait, TASK_INTERRUPTIBLE); + done = sk_wait_event(sk, timeo_p, + (!port->congested || !port->connected)); + finish_wait(sk_sleep(sk), &wait); + } while (!done); + return 0; +} + /** - * send_packet - send a connection-oriented message + * tipc_send_packet - send a connection-oriented message * @iocb: if NULL, indicates that socket lock is already held * @sock: socket structure * @m: message to send @@ -628,60 +726,50 @@ exit: * * Returns the number of bytes sent on success, or errno otherwise */ - -static int send_packet(struct kiocb *iocb, struct socket *sock, - struct msghdr *m, size_t total_len) +static int tipc_send_packet(struct kiocb *iocb, struct socket *sock, + struct msghdr *m, size_t total_len) { struct sock *sk = sock->sk; - struct tipc_port *tport = tipc_sk_port(sk); - struct sockaddr_tipc *dest = (struct sockaddr_tipc *)m->msg_name; - long timeout_val; - int res; + struct tipc_sock *tsk = tipc_sk(sk); + DECLARE_SOCKADDR(struct sockaddr_tipc *, dest, m->msg_name); + int res = -EINVAL; + long timeo; /* Handle implied connection establishment */ - if (unlikely(dest)) - return send_msg(iocb, sock, m, total_len); + return tipc_sendmsg(iocb, sock, m, total_len); - if ((total_len > TIPC_MAX_USER_MSG_SIZE) || - (m->msg_iovlen > (unsigned)INT_MAX)) + if (total_len > TIPC_MAX_USER_MSG_SIZE) return -EMSGSIZE; if (iocb) lock_sock(sk); - timeout_val = sock_sndtimeo(sk, m->msg_flags & MSG_DONTWAIT); + if (unlikely(sock->state != SS_CONNECTED)) { + if (sock->state == SS_DISCONNECTING) + res = -EPIPE; + else + res = -ENOTCONN; + goto exit; + } + timeo = sock_sndtimeo(sk, m->msg_flags & MSG_DONTWAIT); do { - if (unlikely(sock->state != SS_CONNECTED)) { - if (sock->state == SS_DISCONNECTING) - res = -EPIPE; - else - res = -ENOTCONN; - break; - } - - res = tipc_send(tport->ref, m->msg_iovlen, m->msg_iov, - total_len); + res = tipc_send(&tsk->port, m->msg_iov, total_len); if (likely(res != -ELINKCONG)) break; - if (timeout_val <= 0L) { - res = timeout_val ? timeout_val : -EWOULDBLOCK; + res = tipc_wait_for_sndpkt(sock, &timeo); + if (res) break; - } - release_sock(sk); - timeout_val = wait_event_interruptible_timeout(*sk_sleep(sk), - (!tport->congested || !tport->connected), timeout_val); - lock_sock(sk); } while (1); - +exit: if (iocb) release_sock(sk); return res; } /** - * send_stream - send stream-oriented data + * tipc_send_stream - send stream-oriented data * @iocb: (unused) * @sock: socket structure * @m: data to send @@ -692,12 +780,11 @@ static int send_packet(struct kiocb *iocb, struct socket *sock, * Returns the number of bytes sent on success (or partial success), * or errno if no data sent */ - -static int send_stream(struct kiocb *iocb, struct socket *sock, - struct msghdr *m, size_t total_len) +static int tipc_send_stream(struct kiocb *iocb, struct socket *sock, + struct msghdr *m, size_t total_len) { struct sock *sk = sock->sk; - struct tipc_port *tport = tipc_sk_port(sk); + struct tipc_sock *tsk = tipc_sk(sk); struct msghdr my_msg; struct iovec my_iov; struct iovec *curr_iov; @@ -712,18 +799,12 @@ static int send_stream(struct kiocb *iocb, struct socket *sock, lock_sock(sk); /* Handle special cases where there is no connection */ - if (unlikely(sock->state != SS_CONNECTED)) { - if (sock->state == SS_UNCONNECTED) { - res = send_packet(NULL, sock, m, total_len); - goto exit; - } else if (sock->state == SS_DISCONNECTING) { - res = -EPIPE; - goto exit; - } else { - res = -ENOTCONN; - goto exit; - } + if (sock->state == SS_UNCONNECTED) + res = tipc_send_packet(NULL, sock, m, total_len); + else + res = sock->state == SS_DISCONNECTING ? -EPIPE : -ENOTCONN; + goto exit; } if (unlikely(m->msg_name)) { @@ -731,8 +812,7 @@ static int send_stream(struct kiocb *iocb, struct socket *sock, goto exit; } - if ((total_len > (unsigned)INT_MAX) || - (m->msg_iovlen > (unsigned)INT_MAX)) { + if (total_len > (unsigned int)INT_MAX) { res = -EMSGSIZE; goto exit; } @@ -744,7 +824,6 @@ static int send_stream(struct kiocb *iocb, struct socket *sock, * (i.e. one large iovec entry), but could be improved to pass sets * of small iovec entries into send_packet(). */ - curr_iov = m->msg_iov; curr_iovlen = m->msg_iovlen; my_msg.msg_iov = &my_iov; @@ -753,21 +832,22 @@ static int send_stream(struct kiocb *iocb, struct socket *sock, my_msg.msg_name = NULL; bytes_sent = 0; - hdr_size = msg_hdr_sz(&tport->phdr); + hdr_size = msg_hdr_sz(&tsk->port.phdr); while (curr_iovlen--) { curr_start = curr_iov->iov_base; curr_left = curr_iov->iov_len; while (curr_left) { - bytes_to_send = tport->max_pkt - hdr_size; + bytes_to_send = tsk->port.max_pkt - hdr_size; if (bytes_to_send > TIPC_MAX_USER_MSG_SIZE) bytes_to_send = TIPC_MAX_USER_MSG_SIZE; if (curr_left < bytes_to_send) bytes_to_send = curr_left; my_iov.iov_base = curr_start; my_iov.iov_len = bytes_to_send; - res = send_packet(NULL, sock, &my_msg, bytes_to_send); + res = tipc_send_packet(NULL, sock, &my_msg, + bytes_to_send); if (res < 0) { if (bytes_sent) res = bytes_sent; @@ -788,25 +868,25 @@ exit: /** * auto_connect - complete connection setup to a remote port - * @sock: socket structure + * @tsk: tipc socket structure * @msg: peer's response message * * Returns 0 on success, errno otherwise */ - -static int auto_connect(struct socket *sock, struct tipc_msg *msg) +static int auto_connect(struct tipc_sock *tsk, struct tipc_msg *msg) { - struct tipc_sock *tsock = tipc_sk(sock->sk); + struct tipc_port *port = &tsk->port; + struct socket *sock = tsk->sk.sk_socket; + struct tipc_portid peer; - if (msg_errcode(msg)) { - sock->state = SS_DISCONNECTING; - return -ECONNREFUSED; - } + peer.ref = msg_origport(msg); + peer.node = msg_orignode(msg); + + __tipc_port_connect(port->ref, port, &peer); - 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)); + if (msg_importance(msg) > TIPC_CRITICAL_IMPORTANCE) + return -EINVAL; + msg_set_importance(&port->phdr, (u32)msg_importance(msg)); sock->state = SS_CONNECTED; return 0; } @@ -818,14 +898,14 @@ static int auto_connect(struct socket *sock, struct tipc_msg *msg) * * Note: Address is not captured if not requested by receiver. */ - static void set_orig_addr(struct msghdr *m, struct tipc_msg *msg) { - struct sockaddr_tipc *addr = (struct sockaddr_tipc *)m->msg_name; + DECLARE_SOCKADDR(struct sockaddr_tipc *, addr, m->msg_name); 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 */ @@ -844,9 +924,8 @@ 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; @@ -858,7 +937,6 @@ static int anc_data_recv(struct msghdr *m, struct tipc_msg *msg, return 0; /* Optionally capture errored message object(s) */ - err = msg ? msg_errcode(msg) : 0; if (unlikely(err)) { anc_data[0] = err; @@ -875,7 +953,6 @@ static int anc_data_recv(struct msghdr *m, struct tipc_msg *msg, } /* Optionally capture message destination object */ - dest_type = msg ? msg_type(msg) : TIPC_DIRECT_MSG; switch (dest_type) { case TIPC_NAMED_MSG: @@ -908,8 +985,41 @@ static int anc_data_recv(struct msghdr *m, struct tipc_msg *msg, return 0; } +static int tipc_wait_for_rcvmsg(struct socket *sock, long *timeop) +{ + struct sock *sk = sock->sk; + DEFINE_WAIT(wait); + long timeo = *timeop; + int err; + + for (;;) { + prepare_to_wait(sk_sleep(sk), &wait, TASK_INTERRUPTIBLE); + if (timeo && skb_queue_empty(&sk->sk_receive_queue)) { + if (sock->state == SS_DISCONNECTING) { + err = -ENOTCONN; + break; + } + release_sock(sk); + timeo = schedule_timeout(timeo); + lock_sock(sk); + } + err = 0; + if (!skb_queue_empty(&sk->sk_receive_queue)) + break; + err = sock_intr_errno(timeo); + if (signal_pending(current)) + break; + err = -EAGAIN; + if (!timeo) + break; + } + finish_wait(sk_sleep(sk), &wait); + *timeop = timeo; + return err; +} + /** - * recv_msg - receive packet-oriented message + * tipc_recvmsg - receive packet-oriented message * @iocb: (unused) * @m: descriptor for message info * @buf_len: total size of user buffer area @@ -920,21 +1030,20 @@ static int anc_data_recv(struct msghdr *m, struct tipc_msg *msg, * * Returns size of returned message data, errno otherwise */ - -static int recv_msg(struct kiocb *iocb, struct socket *sock, - struct msghdr *m, size_t buf_len, int flags) +static int tipc_recvmsg(struct kiocb *iocb, struct socket *sock, + struct msghdr *m, size_t buf_len, int flags) { struct sock *sk = sock->sk; - struct tipc_port *tport = tipc_sk_port(sk); + struct tipc_sock *tsk = tipc_sk(sk); + struct tipc_port *port = &tsk->port; struct sk_buff *buf; struct tipc_msg *msg; - long timeout; + long timeo; unsigned int sz; u32 err; int res; /* Catch invalid receive requests */ - if (unlikely(!buf_len)) return -EINVAL; @@ -945,61 +1054,35 @@ static int recv_msg(struct kiocb *iocb, struct socket *sock, goto exit; } - timeout = sock_rcvtimeo(sk, flags & MSG_DONTWAIT); + timeo = sock_rcvtimeo(sk, flags & MSG_DONTWAIT); restart: /* Look for a message in receive queue; wait if necessary */ - - while (skb_queue_empty(&sk->sk_receive_queue)) { - if (sock->state == SS_DISCONNECTING) { - res = -ENOTCONN; - goto exit; - } - if (timeout <= 0L) { - res = timeout ? timeout : -EWOULDBLOCK; - goto exit; - } - release_sock(sk); - timeout = wait_event_interruptible_timeout(*sk_sleep(sk), - tipc_rx_ready(sock), - timeout); - lock_sock(sk); - } + res = tipc_wait_for_rcvmsg(sock, &timeo); + if (res) + goto exit; /* Look at first message in receive queue */ - buf = skb_peek(&sk->sk_receive_queue); msg = buf_msg(buf); 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); goto restart; } /* Capture sender's address (optional) */ - set_orig_addr(m, msg); /* Capture ancillary data (optional) */ - - res = anc_data_recv(m, msg, tport); + res = anc_data_recv(m, msg, port); if (res) goto exit; /* Capture message data (if valid) & compute return value (always) */ - if (!err) { if (unlikely(buf_len < sz)) { sz = buf_len; @@ -1019,11 +1102,10 @@ restart: } /* Consume received message (optional) */ - if (likely(!(flags & MSG_PEEK))) { if ((sock->state != SS_READY) && - (++tport->conn_unacked >= TIPC_FLOW_CONTROL_WIN)) - tipc_acknowledge(tport->ref, tport->conn_unacked); + (++port->conn_unacked >= TIPC_CONNACK_INTV)) + tipc_acknowledge(port->ref, port->conn_unacked); advance_rx_queue(sk); } exit: @@ -1032,7 +1114,7 @@ exit: } /** - * recv_stream - receive stream-oriented data + * tipc_recv_stream - receive stream-oriented data * @iocb: (unused) * @m: descriptor for message info * @buf_len: total size of user buffer area @@ -1043,15 +1125,15 @@ exit: * * Returns size of returned message data, errno otherwise */ - -static int recv_stream(struct kiocb *iocb, struct socket *sock, - struct msghdr *m, size_t buf_len, int flags) +static int tipc_recv_stream(struct kiocb *iocb, struct socket *sock, + struct msghdr *m, size_t buf_len, int flags) { struct sock *sk = sock->sk; - struct tipc_port *tport = tipc_sk_port(sk); + struct tipc_sock *tsk = tipc_sk(sk); + struct tipc_port *port = &tsk->port; struct sk_buff *buf; struct tipc_msg *msg; - long timeout; + long timeo; unsigned int sz; int sz_to_copy, target, needed; int sz_copied = 0; @@ -1059,65 +1141,46 @@ static int recv_stream(struct kiocb *iocb, struct socket *sock, int res = 0; /* Catch invalid receive attempts */ - if (unlikely(!buf_len)) return -EINVAL; lock_sock(sk); - if (unlikely((sock->state == SS_UNCONNECTED) || - (sock->state == SS_CONNECTING))) { + if (unlikely(sock->state == SS_UNCONNECTED)) { res = -ENOTCONN; goto exit; } target = sock_rcvlowat(sk, flags & MSG_WAITALL, buf_len); - timeout = sock_rcvtimeo(sk, flags & MSG_DONTWAIT); -restart: + timeo = sock_rcvtimeo(sk, flags & MSG_DONTWAIT); +restart: /* Look for a message in receive queue; wait if necessary */ - - while (skb_queue_empty(&sk->sk_receive_queue)) { - if (sock->state == SS_DISCONNECTING) { - res = -ENOTCONN; - goto exit; - } - if (timeout <= 0L) { - res = timeout ? timeout : -EWOULDBLOCK; - goto exit; - } - release_sock(sk); - timeout = wait_event_interruptible_timeout(*sk_sleep(sk), - tipc_rx_ready(sock), - timeout); - lock_sock(sk); - } + res = tipc_wait_for_rcvmsg(sock, &timeo); + if (res) + goto exit; /* Look at first message in receive queue */ - buf = skb_peek(&sk->sk_receive_queue); msg = buf_msg(buf); sz = msg_data_sz(msg); err = msg_errcode(msg); /* Discard an empty non-errored message & try again */ - if ((!sz) && (!err)) { advance_rx_queue(sk); goto restart; } /* Optionally capture sender's address & ancillary data of first msg */ - if (sz_copied == 0) { set_orig_addr(m, msg); - res = anc_data_recv(m, msg, tport); + res = anc_data_recv(m, msg, port); if (res) goto exit; } /* Capture message data (if valid) & compute return value (always) */ - if (!err) { u32 offset = (u32)(unsigned long)(TIPC_SKB_CB(buf)->handle); @@ -1149,15 +1212,13 @@ restart: } /* Consume received message (optional) */ - if (likely(!(flags & MSG_PEEK))) { - if (unlikely(++tport->conn_unacked >= TIPC_FLOW_CONTROL_WIN)) - tipc_acknowledge(tport->ref, tport->conn_unacked); + if (unlikely(++port->conn_unacked >= TIPC_CONNACK_INTV)) + tipc_acknowledge(port->ref, port->conn_unacked); advance_rx_queue(sk); } /* Loop around if more data is required */ - if ((sz_copied < buf_len) && /* didn't get all requested data */ (!skb_queue_empty(&sk->sk_receive_queue) || (sz_copied < target)) && /* and more is ready or required */ @@ -1171,32 +1232,144 @@ 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) +{ + 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 + * @tsk: 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 u32 filter_connect(struct tipc_sock *tsk, struct sk_buff **buf) +{ + struct sock *sk = &tsk->sk; + struct tipc_port *port = &tsk->port; + struct socket *sock = sk->sk_socket; + struct tipc_msg *msg = buf_msg(*buf); + + u32 retval = TIPC_ERR_NO_PORT; + int res; + + 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(port, msg)) { + if (unlikely(msg_errcode(msg))) { + sock->state = SS_DISCONNECTING; + __tipc_port_disconnect(port); + } + 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; + } + + if (unlikely(!msg_connected(msg))) + break; + + res = auto_connect(tsk, msg); + if (res) { + sock->state = SS_DISCONNECTING; + sk->sk_err = -res; + retval = TIPC_OK; + break; + } -static int rx_queue_full(struct tipc_msg *msg, u32 queue_size, u32 base) + /* 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) { - 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 tipc_msg *msg = buf_msg(buf); if (msg_connected(msg)) - threshold *= 4; + return sysctl_tipc_rmem[2]; - return queue_size >= threshold; + return sk->sk_rcvbuf >> TIPC_CRITICAL_IMPORTANCE << + msg_importance(msg); } /** @@ -1211,77 +1384,42 @@ static int rx_queue_full(struct tipc_msg *msg, u32 queue_size, u32 base) * * Returns TIPC error status code (TIPC_OK if message is not to be rejected) */ - static u32 filter_rcv(struct sock *sk, struct sk_buff *buf) { struct socket *sock = sk->sk_socket; + struct tipc_sock *tsk = tipc_sk(sk); 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 */ - - /* - * WOULD IT BE BETTER TO JUST DISCARD THESE MESSAGES INSTEAD? - * "NO PORT" ISN'T REALLY THE RIGHT ERROR CODE, AND THERE MAY - * BE SECURITY IMPLICATIONS INHERENT IN REJECTING INVALID TRAFFIC - */ + if (msg_type(msg) > TIPC_DIRECT_MSG) + return TIPC_ERR_NO_PORT; if (sock->state == SS_READY) { 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)) - 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(tsk, &buf); + if (res != TIPC_OK || buf == NULL) + return res; } /* Reject message if there isn't room to queue it */ + if (sk_rmem_alloc_get(sk) + buf->truesize >= limit) + return TIPC_ERR_OVERLOAD; - 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; - } - - /* Enqueue message (finally!) */ - - TIPC_SKB_CB(buf)->handle = 0; - atomic_inc(&tipc_queue_size); + /* Enqueue message */ + TIPC_SKB_CB(buf)->handle = NULL; __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); return TIPC_OK; } /** - * backlog_rcv - handle incoming message from backlog queue + * tipc_backlog_rcv - handle incoming message from backlog queue * @sk: socket * @buf: message * @@ -1289,70 +1427,100 @@ static u32 filter_rcv(struct sock *sk, struct sk_buff *buf) * * Returns 0 */ - -static int backlog_rcv(struct sock *sk, struct sk_buff *buf) +static int tipc_backlog_rcv(struct sock *sk, struct sk_buff *buf) { u32 res; + struct tipc_sock *tsk = tipc_sk(sk); + uint truesize = buf->truesize; res = filter_rcv(sk, buf); - if (res) + if (unlikely(res)) tipc_reject_msg(buf, res); + + if (atomic_read(&tsk->dupl_rcvcnt) < TIPC_CONN_OVERLOAD_LIMIT) + atomic_add(truesize, &tsk->dupl_rcvcnt); + return 0; } /** - * dispatch - handle incoming message - * @tport: TIPC port that received message - * @buf: message - * - * Called with port lock already taken. - * - * Returns TIPC error status code (TIPC_OK if message is not to be rejected) + * tipc_sk_rcv - handle incoming message + * @buf: buffer containing arriving message + * Consumes buffer + * Returns 0 if success, or errno: -EHOSTUNREACH */ - -static u32 dispatch(struct tipc_port *tport, struct sk_buff *buf) +int tipc_sk_rcv(struct sk_buff *buf) { - struct sock *sk = (struct sock *)tport->usr_handle; - u32 res; + struct tipc_sock *tsk; + struct tipc_port *port; + struct sock *sk; + u32 dport = msg_destport(buf_msg(buf)); + int err = TIPC_OK; + uint limit; - /* - * Process message if socket is unlocked; otherwise add to backlog queue - * - * This code is based on sk_receive_skb(), but must be distinct from it - * since a TIPC-specific filter/reject mechanism is utilized - */ + /* Forward unresolved named message */ + if (unlikely(!dport)) { + tipc_net_route_msg(buf); + return 0; + } + /* Validate destination */ + port = tipc_port_lock(dport); + if (unlikely(!port)) { + err = TIPC_ERR_NO_PORT; + goto exit; + } + + tsk = tipc_port_to_sock(port); + sk = &tsk->sk; + + /* Queue message */ bh_lock_sock(sk); + if (!sock_owned_by_user(sk)) { - res = filter_rcv(sk, buf); + err = filter_rcv(sk, buf); } else { - if (sk_add_backlog(sk, buf)) - res = TIPC_ERR_OVERLOAD; - else - res = TIPC_OK; + if (sk->sk_backlog.len == 0) + atomic_set(&tsk->dupl_rcvcnt, 0); + limit = rcvbuf_limit(sk, buf) + atomic_read(&tsk->dupl_rcvcnt); + if (sk_add_backlog(sk, buf, limit)) + err = TIPC_ERR_OVERLOAD; } + bh_unlock_sock(sk); + tipc_port_unlock(port); - return res; + if (likely(!err)) + return 0; +exit: + tipc_reject_msg(buf, err); + return -EHOSTUNREACH; } -/** - * wakeupdispatch - wake up port after congestion - * @tport: port to wakeup - * - * Called with port lock already taken. - */ - -static void wakeupdispatch(struct tipc_port *tport) +static int tipc_wait_for_connect(struct socket *sock, long *timeo_p) { - struct sock *sk = (struct sock *)tport->usr_handle; + struct sock *sk = sock->sk; + DEFINE_WAIT(wait); + int done; - if (waitqueue_active(sk_sleep(sk))) - wake_up_interruptible(sk_sleep(sk)); + do { + int err = sock_error(sk); + if (err) + return err; + if (!*timeo_p) + return -ETIMEDOUT; + if (signal_pending(current)) + return sock_intr_errno(*timeo_p); + + prepare_to_wait(sk_sleep(sk), &wait, TASK_INTERRUPTIBLE); + done = sk_wait_event(sk, timeo_p, sock->state != SS_CONNECTING); + finish_wait(sk_sleep(sk), &wait); + } while (!done); + return 0; } /** - * connect - establish a connection to another TIPC port + * tipc_connect - establish a connection to another TIPC port * @sock: socket structure * @dest: socket address for destination port * @destlen: size of socket address data structure @@ -1360,121 +1528,86 @@ static void wakeupdispatch(struct tipc_port *tport) * * Returns 0 on success, errno otherwise */ - -static int connect(struct socket *sock, struct sockaddr *dest, int destlen, - int flags) +static int tipc_connect(struct socket *sock, struct sockaddr *dest, + int destlen, int flags) { 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; + long timeout = (flags & O_NONBLOCK) ? 0 : tipc_sk(sk)->conn_timeout; + socket_state previous; int res; lock_sock(sk); /* For now, TIPC does not allow use of connect() with DGRAM/RDM types */ - if (sock->state == SS_READY) { res = -EOPNOTSUPP; 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 * * Note: send_msg() validates the rest of the address fields, * so there's no need to do it here */ - if (dst->addrtype == TIPC_ADDR_MCAST) { res = -EINVAL; goto exit; } - /* Reject any messages already in receive queue (very unlikely) */ - - reject_rx_queue(sk); - - /* Send a 'SYN-' to destination */ - - m.msg_name = dest; - m.msg_namelen = destlen; - res = send_msg(NULL, sock, &m, 0); - if (res < 0) - goto exit; + previous = sock->state; + switch (sock->state) { + case SS_UNCONNECTED: + /* Send a 'SYN-' to destination */ + m.msg_name = dest; + m.msg_namelen = destlen; - /* Wait until an 'ACK' or 'RST' arrives, or a timeout occurs */ + /* If connect is in non-blocking case, set MSG_DONTWAIT to + * indicate send_msg() is never blocked. + */ + if (!timeout) + m.msg_flags = MSG_DONTWAIT; - 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); + res = tipc_sendmsg(NULL, sock, &m, 0); + if ((res < 0) && (res != -EWOULDBLOCK)) + 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; - else - res = -ECONNREFUSED; - } - } else { - if (res == 0) - res = -ETIMEDOUT; - else - ; /* leave "res" unchanged */ - sock->state = SS_DISCONNECTING; + /* Just entered SS_CONNECTING state; the only + * difference is that return value in non-blocking + * case is EINPROGRESS, rather than EALREADY. + */ + res = -EINPROGRESS; + case SS_CONNECTING: + if (previous == SS_CONNECTING) + res = -EALREADY; + if (!timeout) + goto exit; + timeout = msecs_to_jiffies(timeout); + /* Wait until an 'ACK' or 'RST' arrives, or a timeout occurs */ + res = tipc_wait_for_connect(sock, &timeout); + break; + case SS_CONNECTED: + res = -EISCONN; + break; + default: + res = -EINVAL; + break; } - exit: release_sock(sk); return res; } /** - * listen - allow socket to listen for incoming connections + * tipc_listen - allow socket to listen for incoming connections * @sock: socket structure * @len: (unused) * * Returns 0 on success, errno otherwise */ - -static int listen(struct socket *sock, int len) +static int tipc_listen(struct socket *sock, int len) { struct sock *sk = sock->sk; int res; @@ -1492,19 +1625,59 @@ static int listen(struct socket *sock, int len) return res; } +static int tipc_wait_for_accept(struct socket *sock, long timeo) +{ + struct sock *sk = sock->sk; + DEFINE_WAIT(wait); + int err; + + /* True wake-one mechanism for incoming connections: only + * one process gets woken up, not the 'whole herd'. + * Since we do not 'race & poll' for established sockets + * anymore, the common case will execute the loop only once. + */ + for (;;) { + prepare_to_wait_exclusive(sk_sleep(sk), &wait, + TASK_INTERRUPTIBLE); + if (timeo && skb_queue_empty(&sk->sk_receive_queue)) { + release_sock(sk); + timeo = schedule_timeout(timeo); + lock_sock(sk); + } + err = 0; + if (!skb_queue_empty(&sk->sk_receive_queue)) + break; + err = -EINVAL; + if (sock->state != SS_LISTENING) + break; + err = sock_intr_errno(timeo); + if (signal_pending(current)) + break; + err = -EAGAIN; + if (!timeo) + break; + } + finish_wait(sk_sleep(sk), &wait); + return err; +} + /** - * accept - wait for connection request + * tipc_accept - wait for connection request * @sock: listening socket * @newsock: new socket that is to be connected * @flags: file-related flags associated with socket * * Returns 0 on success, errno otherwise */ - -static int accept(struct socket *sock, struct socket *new_sock, int flags) +static int tipc_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_port *new_port; + struct tipc_msg *msg; + struct tipc_portid peer; + u32 new_ref; + long timeo; int res; lock_sock(sk); @@ -1513,75 +1686,65 @@ static int accept(struct socket *sock, struct socket *new_sock, int flags) res = -EINVAL; goto exit; } - - while (skb_queue_empty(&sk->sk_receive_queue)) { - if (flags & O_NONBLOCK) { - res = -EWOULDBLOCK; - goto exit; - } - release_sock(sk); - res = wait_event_interruptible(*sk_sleep(sk), - (!skb_queue_empty(&sk->sk_receive_queue))); - lock_sock(sk); - if (res) - goto exit; - } + timeo = sock_rcvtimeo(sk, flags & O_NONBLOCK); + res = tipc_wait_for_accept(sock, timeo); + if (res) + goto exit; 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); - - lock_sock(new_sk); - - /* - * 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 */ + res = tipc_sk_create(sock_net(sock->sk), new_sock, 0, 1); + if (res) + goto exit; - 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; + new_sk = new_sock->sk; + new_port = &tipc_sk(new_sk)->port; + new_ref = new_port->ref; + msg = buf_msg(buf); - 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. - */ + /* + * 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 */ + peer.ref = msg_origport(msg); + peer.node = msg_orignode(msg); + tipc_port_connect(new_ref, &peer); + new_sock->state = SS_CONNECTED; + + tipc_port_set_importance(new_port, msg_importance(msg)); + if (msg_named(msg)) { + new_port->conn_type = msg_nametype(msg); + new_port->conn_instance = msg_nameinst(msg); + } - if (!msg_data_sz(msg)) { - struct msghdr m = {NULL,}; + /* + * 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); - } - release_sock(new_sk); + advance_rx_queue(sk); + tipc_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; } /** - * shutdown - shutdown socket connection + * tipc_shutdown - shutdown socket connection * @sock: socket structure * @how: direction to close (must be SHUT_RDWR) * @@ -1589,11 +1752,11 @@ exit: * * Returns 0 on success, errno otherwise */ - -static int shutdown(struct socket *sock, int how) +static int tipc_shutdown(struct socket *sock, int how) { struct sock *sk = sock->sk; - struct tipc_port *tport = tipc_sk_port(sk); + struct tipc_sock *tsk = tipc_sk(sk); + struct tipc_port *port = &tsk->port; struct sk_buff *buf; int res; @@ -1606,19 +1769,18 @@ static int shutdown(struct socket *sock, int how) case SS_CONNECTING: case SS_CONNECTED: - /* Disconnect and send a 'FIN+' or 'FIN-' message to peer */ 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) { - buf_discard(buf); + if (TIPC_SKB_CB(buf)->handle != NULL) { + kfree_skb(buf); goto restart; } - tipc_disconnect(tport->ref); + tipc_port_disconnect(port->ref); tipc_reject_msg(buf, TIPC_CONN_SHUTDOWN); } else { - tipc_shutdown(tport->ref); + tipc_port_shutdown(port->ref); } sock->state = SS_DISCONNECTING; @@ -1627,11 +1789,11 @@ restart: case SS_DISCONNECTING: - /* Discard any unreceived messages; wake up sleeping tasks */ + /* Discard any unreceived messages */ + __skb_queue_purge(&sk->sk_receive_queue); - discard_rx_queue(sk); - if (waitqueue_active(sk_sleep(sk))) - wake_up_interruptible(sk_sleep(sk)); + /* Wake up anyone sleeping in poll */ + sk->sk_state_change(sk); res = 0; break; @@ -1644,7 +1806,7 @@ restart: } /** - * setsockopt - set socket option + * tipc_setsockopt - set socket option * @sock: socket structure * @lvl: option level * @opt: option identifier @@ -1656,12 +1818,12 @@ 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 tipc_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); + struct tipc_sock *tsk = tipc_sk(sk); + struct tipc_port *port = &tsk->port; u32 value; int res; @@ -1679,16 +1841,16 @@ static int setsockopt(struct socket *sock, switch (opt) { case TIPC_IMPORTANCE: - res = tipc_set_portimportance(tport->ref, value); + tipc_port_set_importance(port, value); break; case TIPC_SRC_DROPPABLE: if (sock->type != SOCK_STREAM) - res = tipc_set_portunreliable(tport->ref, value); + tipc_port_set_unreliable(port, value); else res = -ENOPROTOOPT; break; case TIPC_DEST_DROPPABLE: - res = tipc_set_portunreturnable(tport->ref, value); + tipc_port_set_unreturnable(port, value); break; case TIPC_CONN_TIMEOUT: tipc_sk(sk)->conn_timeout = value; @@ -1704,7 +1866,7 @@ static int setsockopt(struct socket *sock, } /** - * getsockopt - get socket option + * tipc_getsockopt - get socket option * @sock: socket structure * @lvl: option level * @opt: option identifier @@ -1716,12 +1878,12 @@ 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 tipc_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); + struct tipc_sock *tsk = tipc_sk(sk); + struct tipc_port *port = &tsk->port; int len; u32 value; int res; @@ -1738,20 +1900,20 @@ static int getsockopt(struct socket *sock, switch (opt) { case TIPC_IMPORTANCE: - res = tipc_portimportance(tport->ref, &value); + value = tipc_port_importance(port); break; case TIPC_SRC_DROPPABLE: - res = tipc_portunreliable(tport->ref, &value); + value = tipc_port_unreliable(port); break; case TIPC_DEST_DROPPABLE: - res = tipc_portunreturnable(tport->ref, &value); + value = tipc_port_unreturnable(port); break; case TIPC_CONN_TIMEOUT: value = tipc_sk(sk)->conn_timeout; /* 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); @@ -1774,27 +1936,47 @@ static int getsockopt(struct socket *sock, return put_user(sizeof(value), ol); } -/** - * Protocol switches for the various types of TIPC sockets - */ +int tipc_ioctl(struct socket *sk, unsigned int cmd, unsigned long arg) +{ + struct tipc_sioc_ln_req lnr; + void __user *argp = (void __user *)arg; + + switch (cmd) { + case SIOCGETLINKNAME: + if (copy_from_user(&lnr, argp, sizeof(lnr))) + return -EFAULT; + if (!tipc_node_get_linkname(lnr.bearer_id, lnr.peer, + lnr.linkname, TIPC_MAX_LINK_NAME)) { + if (copy_to_user(argp, &lnr, sizeof(lnr))) + return -EFAULT; + return 0; + } + return -EADDRNOTAVAIL; + break; + default: + return -ENOIOCTLCMD; + } +} + +/* Protocol switches for the various types of TIPC sockets */ static const struct proto_ops msg_ops = { .owner = THIS_MODULE, .family = AF_TIPC, - .release = release, - .bind = bind, - .connect = connect, + .release = tipc_release, + .bind = tipc_bind, + .connect = tipc_connect, .socketpair = sock_no_socketpair, .accept = sock_no_accept, - .getname = get_name, - .poll = poll, - .ioctl = sock_no_ioctl, + .getname = tipc_getname, + .poll = tipc_poll, + .ioctl = tipc_ioctl, .listen = sock_no_listen, - .shutdown = shutdown, - .setsockopt = setsockopt, - .getsockopt = getsockopt, - .sendmsg = send_msg, - .recvmsg = recv_msg, + .shutdown = tipc_shutdown, + .setsockopt = tipc_setsockopt, + .getsockopt = tipc_getsockopt, + .sendmsg = tipc_sendmsg, + .recvmsg = tipc_recvmsg, .mmap = sock_no_mmap, .sendpage = sock_no_sendpage }; @@ -1802,20 +1984,20 @@ static const struct proto_ops msg_ops = { static const struct proto_ops packet_ops = { .owner = THIS_MODULE, .family = AF_TIPC, - .release = release, - .bind = bind, - .connect = connect, + .release = tipc_release, + .bind = tipc_bind, + .connect = tipc_connect, .socketpair = sock_no_socketpair, - .accept = accept, - .getname = get_name, - .poll = poll, - .ioctl = sock_no_ioctl, - .listen = listen, - .shutdown = shutdown, - .setsockopt = setsockopt, - .getsockopt = getsockopt, - .sendmsg = send_packet, - .recvmsg = recv_msg, + .accept = tipc_accept, + .getname = tipc_getname, + .poll = tipc_poll, + .ioctl = tipc_ioctl, + .listen = tipc_listen, + .shutdown = tipc_shutdown, + .setsockopt = tipc_setsockopt, + .getsockopt = tipc_getsockopt, + .sendmsg = tipc_send_packet, + .recvmsg = tipc_recvmsg, .mmap = sock_no_mmap, .sendpage = sock_no_sendpage }; @@ -1823,20 +2005,20 @@ static const struct proto_ops packet_ops = { static const struct proto_ops stream_ops = { .owner = THIS_MODULE, .family = AF_TIPC, - .release = release, - .bind = bind, - .connect = connect, + .release = tipc_release, + .bind = tipc_bind, + .connect = tipc_connect, .socketpair = sock_no_socketpair, - .accept = accept, - .getname = get_name, - .poll = poll, - .ioctl = sock_no_ioctl, - .listen = listen, - .shutdown = shutdown, - .setsockopt = setsockopt, - .getsockopt = getsockopt, - .sendmsg = send_stream, - .recvmsg = recv_stream, + .accept = tipc_accept, + .getname = tipc_getname, + .poll = tipc_poll, + .ioctl = tipc_ioctl, + .listen = tipc_listen, + .shutdown = tipc_shutdown, + .setsockopt = tipc_setsockopt, + .getsockopt = tipc_getsockopt, + .sendmsg = tipc_send_stream, + .recvmsg = tipc_recv_stream, .mmap = sock_no_mmap, .sendpage = sock_no_sendpage }; @@ -1844,13 +2026,20 @@ 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 }; /** @@ -1864,18 +2053,16 @@ int tipc_socket_init(void) res = proto_register(&tipc_proto, 1); if (res) { - err("Failed to register TIPC protocol type\n"); + pr_err("Failed to register TIPC protocol type\n"); goto out; } res = sock_register(&tipc_family_ops); if (res) { - err("Failed to register TIPC socket type\n"); + pr_err("Failed to register TIPC socket type\n"); proto_unregister(&tipc_proto); goto out; } - - sockets_enabled = 1; out: return res; } @@ -1883,14 +2070,8 @@ int tipc_socket_init(void) /** * tipc_socket_stop - stop TIPC socket interface */ - void tipc_socket_stop(void) { - if (!sockets_enabled) - return; - - sockets_enabled = 0; sock_unregister(tipc_family_ops.family); proto_unregister(&tipc_proto); } - diff --git a/net/tipc/socket.h b/net/tipc/socket.h new file mode 100644 index 00000000000..3afcd2a70b3 --- /dev/null +++ b/net/tipc/socket.h @@ -0,0 +1,74 @@ +/* net/tipc/socket.h: Include file for TIPC socket code + * + * Copyright (c) 2014, Ericsson AB + * 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_SOCK_H +#define _TIPC_SOCK_H + +#include "port.h" +#include <net/sock.h> + +/** + * struct tipc_sock - TIPC socket structure + * @sk: socket - interacts with 'port' and with user via the socket API + * @port: port - interacts with 'sk' and with the rest of the TIPC stack + * @peer_name: the peer of the connection, if any + * @conn_timeout: the time we can wait for an unresponded setup request + * @dupl_rcvcnt: number of bytes counted twice, in both backlog and rcv queue + */ + +struct tipc_sock { + struct sock sk; + struct tipc_port port; + unsigned int conn_timeout; + atomic_t dupl_rcvcnt; +}; + +static inline struct tipc_sock *tipc_sk(const struct sock *sk) +{ + return container_of(sk, struct tipc_sock, sk); +} + +static inline struct tipc_sock *tipc_port_to_sock(const struct tipc_port *port) +{ + return container_of(port, struct tipc_sock, port); +} + +static inline void tipc_sock_wakeup(struct tipc_sock *tsk) +{ + tsk->sk.sk_write_space(&tsk->sk); +} + +int tipc_sk_rcv(struct sk_buff *buf); + +#endif diff --git a/net/tipc/subscr.c b/net/tipc/subscr.c index 8c49566da8f..642437231ad 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 @@ -41,36 +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: control 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 - * @user_ref: TIPC userid of 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 - */ +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); -struct top_srv { - u32 setup_port; - atomic_t subscription_count; - struct list_head subscriber_list; - spinlock_t lock; +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 @@ -79,37 +85,27 @@ static struct top_srv topsrv; * * Returns converted value */ - 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; msg_sect.iov_base = (void *)&sub->evt; msg_sect.iov_len = sizeof(struct tipc_event); - sub->evt.event = htohl(event, sub->swap); sub->evt.found_lower = htohl(found_lower, sub->swap); 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); + tipc_conn_sendmsg(&topsrv, subscriber->conid, NULL, msg_sect.iov_base, + msg_sect.iov_len); } /** @@ -117,11 +113,8 @@ 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; @@ -137,14 +130,9 @@ 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; @@ -154,129 +142,93 @@ 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; - /* Validate server port reference (in case subscriber is terminating) */ - - server_port = tipc_port_lock(sub->server_ref); - if (server_port == NULL) - return; + /* The spin lock per subscriber is used to protect its members */ + spin_lock_bh(&subscriber->lock); /* 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; } /* Unlink subscription from name table */ - tipc_nametbl_unsubscribe(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, TIPC_SUBSCR_TIMEOUT, 0, 0); /* 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); - - /* Sever connection to subscriber */ - - tipc_shutdown(port_ref); - tipc_deleteport(port_ref); + spin_lock_bh(&subscriber->lock); /* 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); } /** * 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. * * Note that fields of 's' use subscriber's endianness! */ - static void subscr_cancel(struct tipc_subscr *s, struct tipc_subscriber *subscriber) { @@ -285,7 +237,6 @@ static void subscr_cancel(struct tipc_subscr *s, int found = 0; /* Find first matching subscription, exit if not found */ - list_for_each_entry_safe(sub, sub_temp, &subscriber->subscription_list, subscription_list) { if (!memcmp(s, &sub->evt.s, sizeof(struct tipc_subscr))) { @@ -297,13 +248,12 @@ static void subscr_cancel(struct tipc_subscr *s, return; /* 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); } @@ -311,47 +261,39 @@ 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) -{ +static int subscr_subscribe(struct tipc_subscr *s, + struct tipc_subscriber *subscriber, + struct tipc_subscription **sub_p) { struct tipc_subscription *sub; int swap; /* Determine subscriber's endianness */ - swap = !(s->filter & (TIPC_SUB_PORTS | TIPC_SUB_SERVICE)); /* Detect & process a subscription cancellation request */ - if (s->filter & htohl(TIPC_SUB_CANCEL, swap)) { s->filter &= ~htohl(TIPC_SUB_CANCEL, swap); subscr_cancel(s, subscriber); - return NULL; + return 0; } /* Refuse subscription if global limit exceeded */ - - if (atomic_read(&topsrv.subscription_count) >= tipc_max_subscriptions) { - warn("Subscription rejected, subscription limit reached (%u)\n", - tipc_max_subscriptions); - subscr_terminate(subscriber); - return NULL; + if (atomic_read(&subscription_count) >= TIPC_MAX_SUBSCRIPTIONS) { + pr_warn("Subscription rejected, limit reached (%u)\n", + TIPC_MAX_SUBSCRIPTIONS); + return -EINVAL; } /* Allocate subscription object */ - sub = kmalloc(sizeof(*sub), GFP_ATOMIC); if (!sub) { - warn("Subscription rejected, no memory\n"); - subscr_terminate(subscriber); - return NULL; + pr_warn("Subscription rejected, no memory\n"); + return -ENOMEM; } /* Initialize subscription object */ - sub->seq.type = htohl(s->seq.type, swap); sub->seq.lower = htohl(s->seq.lower, swap); sub->seq.upper = htohl(s->seq.upper, swap); @@ -360,229 +302,74 @@ static struct tipc_subscription *subscr_subscribe(struct tipc_subscr *s, if ((!(sub->filter & TIPC_SUB_PORTS) == !(sub->filter & TIPC_SUB_SERVICE)) || (sub->seq.lower > sub->seq.upper)) { - warn("Subscription rejected, illegal request\n"); + pr_warn("Subscription rejected, illegal request\n"); kfree(sub); - subscr_terminate(subscriber); - return NULL; + return -EINVAL; } 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); k_start_timer(&sub->timer, sub->timeout); } - - return sub; + *sub_p = sub; + return 0; } -/** - * 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_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; + struct tipc_subscriber *subscriber = usr_data; + struct tipc_subscription *sub = NULL; - if (size != sizeof(struct tipc_subscr)) { + spin_lock_bh(&subscriber->lock); + if (subscr_subscribe((struct tipc_subscr *)buf, subscriber, &sub) < 0) { + spin_unlock_bh(&subscriber->lock); 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); - } + return; } + 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) { - warn("Subscriber rejected, no memory\n"); - return; + pr_warn("Subscriber rejected, no memory\n"); + 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) { - 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); + subscriber->conid = conid; + spin_lock_init(&subscriber->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); - - /* 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; - - memset(&topsrv, 0, sizeof(topsrv)); - 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_nametbl_publish_rsv(topsrv.setup_port, TIPC_NODE_SCOPE, &seq); - if (res) { - tipc_deleteport(topsrv.setup_port); - topsrv.setup_port = 0; - goto failed; - } - - return 0; - -failed: - 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 ef6529c8456..393e417bee3 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 @@ -51,34 +55,27 @@ struct tipc_subscription; * @swap: indicates if subscriber uses opposite endianness in its messages * @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); void tipc_subscr_stop(void); - #endif diff --git a/net/tipc/log.h b/net/tipc/sysctl.c index 2248d96238e..f3fef93325a 100644 --- a/net/tipc/log.h +++ b/net/tipc/sysctl.c @@ -1,8 +1,7 @@ /* - * net/tipc/log.h: Include file for TIPC print buffer routines + * net/tipc/sysctl.c: sysctl interface to TIPC subsystem * - * Copyright (c) 1997-2006, Ericsson AB - * Copyright (c) 2005-2007, Wind River Systems + * Copyright (c) 2013, Wind River Systems * All rights reserved. * * Redistribution and use in source and binary forms, with or without @@ -34,34 +33,32 @@ * POSSIBILITY OF SUCH DAMAGE. */ -#ifndef _TIPC_LOG_H -#define _TIPC_LOG_H +#include "core.h" -/** - * struct print_buf - TIPC print buffer structure - * @buf: pointer to character array containing print buffer contents - * @size: size of character array - * @crs: pointer to first unused space in character array (i.e. final NUL) - * @echo: echo output to system console if non-zero - */ - -struct print_buf { - char *buf; - u32 size; - char *crs; - int echo; -}; +#include <linux/sysctl.h> -#define TIPC_PB_MIN_SIZE 64 /* minimum size for a print buffer's array */ -#define TIPC_PB_MAX_STR 512 /* max printable string (with trailing NUL) */ +static struct ctl_table_header *tipc_ctl_hdr; -void tipc_printbuf_init(struct print_buf *pb, char *buf, u32 size); -int tipc_printbuf_validate(struct print_buf *pb); - -int tipc_log_resize(int log_size); +static struct ctl_table tipc_table[] = { + { + .procname = "tipc_rmem", + .data = &sysctl_tipc_rmem, + .maxlen = sizeof(sysctl_tipc_rmem), + .mode = 0644, + .proc_handler = proc_dointvec, + }, + {} +}; -struct sk_buff *tipc_log_resize_cmd(const void *req_tlv_area, - int req_tlv_space); -struct sk_buff *tipc_log_dump(void); +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; +} -#endif +void tipc_unregister_sysctl(void) +{ + unregister_net_sysctl_table(tipc_ctl_hdr); +} |