X-Git-Url: https://gerrit.opnfv.org/gerrit/gitweb?p=kvmfornfv.git;a=blobdiff_plain;f=kernel%2Fnet%2Ftipc%2Flink.c;h=91aea071ab27fee550e3c88ecc4097adac399d83;hp=43a515dc97b0d4a2c7fca5bc9dc2951175201fff;hb=e09b41010ba33a20a87472ee821fa407a5b8da36;hpb=f93b97fd65072de626c074dbe099a1fff05ce060 diff --git a/kernel/net/tipc/link.c b/kernel/net/tipc/link.c index 43a515dc9..91aea071a 100644 --- a/kernel/net/tipc/link.c +++ b/kernel/net/tipc/link.c @@ -48,9 +48,9 @@ /* * Error message prefixes */ -static const char *link_co_err = "Link changeover error, "; +static const char *link_co_err = "Link tunneling error, "; static const char *link_rst_msg = "Resetting link "; -static const char *link_unk_evt = "Unknown link event "; +static const char tipc_bclink_name[] = "broadcast-link"; static const struct nla_policy tipc_nl_link_policy[TIPC_NLA_LINK_MAX + 1] = { [TIPC_NLA_LINK_UNSPEC] = { .type = NLA_UNSPEC }, @@ -76,257 +76,530 @@ static const struct nla_policy tipc_nl_prop_policy[TIPC_NLA_PROP_MAX + 1] = { [TIPC_NLA_PROP_WIN] = { .type = NLA_U32 } }; +/* Send states for broadcast NACKs + */ +enum { + BC_NACK_SND_CONDITIONAL, + BC_NACK_SND_UNCONDITIONAL, + BC_NACK_SND_SUPPRESS, +}; + +/* + * Interval between NACKs when packets arrive out of order + */ +#define TIPC_NACK_INTV (TIPC_MIN_LINK_WIN * 2) /* * Out-of-range value for link session numbers */ -#define INVALID_SESSION 0x10000 +#define WILDCARD_SESSION 0x10000 -/* - * Link state events: +/* Link FSM states: */ -#define STARTING_EVT 856384768 /* link processing trigger */ -#define TRAFFIC_MSG_EVT 560815u /* rx'd ??? */ -#define TIMEOUT_EVT 560817u /* link timer expired */ +enum { + LINK_ESTABLISHED = 0xe, + LINK_ESTABLISHING = 0xe << 4, + LINK_RESET = 0x1 << 8, + LINK_RESETTING = 0x2 << 12, + LINK_PEER_RESET = 0xd << 16, + LINK_FAILINGOVER = 0xf << 20, + LINK_SYNCHING = 0xc << 24 +}; -/* - * State value stored in 'failover_pkts' +/* Link FSM state checking routines */ -#define FIRST_FAILOVER 0xffffu - -static void link_handle_out_of_seq_msg(struct tipc_link *link, - struct sk_buff *skb); -static void tipc_link_proto_rcv(struct tipc_link *link, - struct sk_buff *skb); -static void link_set_supervision_props(struct tipc_link *l_ptr, u32 tol); -static void link_state_event(struct tipc_link *l_ptr, u32 event); +static int link_is_up(struct tipc_link *l) +{ + return l->state & (LINK_ESTABLISHED | LINK_SYNCHING); +} + +static int tipc_link_proto_rcv(struct tipc_link *l, struct sk_buff *skb, + struct sk_buff_head *xmitq); +static void tipc_link_build_proto_msg(struct tipc_link *l, int mtyp, bool probe, + u16 rcvgap, int tolerance, int priority, + struct sk_buff_head *xmitq); static void link_reset_statistics(struct tipc_link *l_ptr); static void link_print(struct tipc_link *l_ptr, const char *str); -static void tipc_link_sync_xmit(struct tipc_link *l); -static void tipc_link_sync_rcv(struct tipc_node *n, struct sk_buff *buf); -static void tipc_link_input(struct tipc_link *l, struct sk_buff *skb); -static bool tipc_data_input(struct tipc_link *l, struct sk_buff *skb); -static bool tipc_link_failover_rcv(struct tipc_link *l, struct sk_buff **skb); +static void tipc_link_build_nack_msg(struct tipc_link *l, + struct sk_buff_head *xmitq); +static void tipc_link_build_bc_init_msg(struct tipc_link *l, + struct sk_buff_head *xmitq); +static bool tipc_link_release_pkts(struct tipc_link *l, u16 to); + /* - * Simple link routines + * Simple non-static link routines (i.e. referenced outside this file) */ -static unsigned int align(unsigned int i) +bool tipc_link_is_up(struct tipc_link *l) { - return (i + 3) & ~3u; + return link_is_up(l); } -static void tipc_link_release(struct kref *kref) +bool tipc_link_peer_is_down(struct tipc_link *l) { - kfree(container_of(kref, struct tipc_link, ref)); + return l->state == LINK_PEER_RESET; } -static void tipc_link_get(struct tipc_link *l_ptr) +bool tipc_link_is_reset(struct tipc_link *l) { - kref_get(&l_ptr->ref); + return l->state & (LINK_RESET | LINK_FAILINGOVER | LINK_ESTABLISHING); } -static void tipc_link_put(struct tipc_link *l_ptr) +bool tipc_link_is_establishing(struct tipc_link *l) { - kref_put(&l_ptr->ref, tipc_link_release); + return l->state == LINK_ESTABLISHING; } -static struct tipc_link *tipc_parallel_link(struct tipc_link *l) +bool tipc_link_is_synching(struct tipc_link *l) { - if (l->owner->active_links[0] != l) - return l->owner->active_links[0]; - return l->owner->active_links[1]; + return l->state == LINK_SYNCHING; } -/* - * Simple non-static link routines (i.e. referenced outside this file) - */ -int tipc_link_is_up(struct tipc_link *l_ptr) +bool tipc_link_is_failingover(struct tipc_link *l) { - if (!l_ptr) - return 0; - return link_working_working(l_ptr) || link_working_unknown(l_ptr); + return l->state == LINK_FAILINGOVER; } -int tipc_link_is_active(struct tipc_link *l_ptr) +bool tipc_link_is_blocked(struct tipc_link *l) { - return (l_ptr->owner->active_links[0] == l_ptr) || - (l_ptr->owner->active_links[1] == l_ptr); + return l->state & (LINK_RESETTING | LINK_PEER_RESET | LINK_FAILINGOVER); } -/** - * link_timeout - handle expiration of link timer - * @l_ptr: pointer to link - */ -static void link_timeout(unsigned long data) +static bool link_is_bc_sndlink(struct tipc_link *l) { - struct tipc_link *l_ptr = (struct tipc_link *)data; - struct sk_buff *skb; + return !l->bc_sndlink; +} + +static bool link_is_bc_rcvlink(struct tipc_link *l) +{ + return ((l->bc_rcvlink == l) && !link_is_bc_sndlink(l)); +} + +int tipc_link_is_active(struct tipc_link *l) +{ + return l->active; +} - tipc_node_lock(l_ptr->owner); +void tipc_link_set_active(struct tipc_link *l, bool active) +{ + l->active = active; +} - /* update counters used in statistical profiling of send traffic */ - l_ptr->stats.accu_queue_sz += skb_queue_len(&l_ptr->transmq); - l_ptr->stats.queue_sz_counts++; +void tipc_link_add_bc_peer(struct tipc_link *snd_l, + struct tipc_link *uc_l, + struct sk_buff_head *xmitq) +{ + struct tipc_link *rcv_l = uc_l->bc_rcvlink; - skb = skb_peek(&l_ptr->transmq); - if (skb) { - struct tipc_msg *msg = buf_msg(skb); - u32 length = msg_size(msg); + snd_l->ackers++; + rcv_l->acked = snd_l->snd_nxt - 1; + snd_l->state = LINK_ESTABLISHED; + tipc_link_build_bc_init_msg(uc_l, xmitq); +} - if ((msg_user(msg) == MSG_FRAGMENTER) && - (msg_type(msg) == FIRST_FRAGMENT)) { - length = msg_size(msg_get_wrapped(msg)); - } - if (length) { - l_ptr->stats.msg_lengths_total += length; - l_ptr->stats.msg_length_counts++; - if (length <= 64) - l_ptr->stats.msg_length_profile[0]++; - else if (length <= 256) - l_ptr->stats.msg_length_profile[1]++; - else if (length <= 1024) - l_ptr->stats.msg_length_profile[2]++; - else if (length <= 4096) - l_ptr->stats.msg_length_profile[3]++; - else if (length <= 16384) - l_ptr->stats.msg_length_profile[4]++; - else if (length <= 32768) - l_ptr->stats.msg_length_profile[5]++; - else - l_ptr->stats.msg_length_profile[6]++; - } +void tipc_link_remove_bc_peer(struct tipc_link *snd_l, + struct tipc_link *rcv_l, + struct sk_buff_head *xmitq) +{ + u16 ack = snd_l->snd_nxt - 1; + + snd_l->ackers--; + tipc_link_bc_ack_rcv(rcv_l, ack, xmitq); + tipc_link_reset(rcv_l); + rcv_l->state = LINK_RESET; + if (!snd_l->ackers) { + tipc_link_reset(snd_l); + snd_l->state = LINK_RESET; + __skb_queue_purge(xmitq); } +} - /* do all other link processing performed on a periodic basis */ - link_state_event(l_ptr, TIMEOUT_EVT); +int tipc_link_bc_peers(struct tipc_link *l) +{ + return l->ackers; +} - if (skb_queue_len(&l_ptr->backlogq)) - tipc_link_push_packets(l_ptr); +void tipc_link_set_mtu(struct tipc_link *l, int mtu) +{ + l->mtu = mtu; +} - tipc_node_unlock(l_ptr->owner); - tipc_link_put(l_ptr); +int tipc_link_mtu(struct tipc_link *l) +{ + return l->mtu; } -static void link_set_timer(struct tipc_link *link, unsigned long time) +static u32 link_own_addr(struct tipc_link *l) { - if (!mod_timer(&link->timer, jiffies + time)) - tipc_link_get(link); + return msg_prevnode(l->pmsg); } /** * tipc_link_create - create a new link - * @n_ptr: pointer to associated node - * @b_ptr: pointer to associated bearer - * @media_addr: media address to use when sending messages over link + * @n: pointer to associated node + * @if_name: associated interface name + * @bearer_id: id (index) of associated bearer + * @tolerance: link tolerance to be used by link + * @net_plane: network plane (A,B,c..) this link belongs to + * @mtu: mtu to be advertised by link + * @priority: priority to be used by link + * @window: send window to be used by link + * @session: session to be used by link + * @ownnode: identity of own node + * @peer: node id of peer node + * @peer_caps: bitmap describing peer node capabilities + * @bc_sndlink: the namespace global link used for broadcast sending + * @bc_rcvlink: the peer specific link used for broadcast reception + * @inputq: queue to put messages ready for delivery + * @namedq: queue to put binding table update messages ready for delivery + * @link: return value, pointer to put the created link * - * Returns pointer to link. + * Returns true if link was created, otherwise false */ -struct tipc_link *tipc_link_create(struct tipc_node *n_ptr, - struct tipc_bearer *b_ptr, - const struct tipc_media_addr *media_addr) +bool tipc_link_create(struct net *net, char *if_name, int bearer_id, + int tolerance, char net_plane, u32 mtu, int priority, + int window, u32 session, u32 ownnode, u32 peer, + u16 peer_caps, + struct tipc_link *bc_sndlink, + struct tipc_link *bc_rcvlink, + struct sk_buff_head *inputq, + struct sk_buff_head *namedq, + struct tipc_link **link) { - struct tipc_net *tn = net_generic(n_ptr->net, tipc_net_id); - struct tipc_link *l_ptr; - struct tipc_msg *msg; - char *if_name; - char addr_string[16]; - u32 peer = n_ptr->addr; - - if (n_ptr->link_cnt >= MAX_BEARERS) { - tipc_addr_string_fill(addr_string, n_ptr->addr); - pr_err("Attempt to establish %uth link to %s. Max %u allowed.\n", - n_ptr->link_cnt, addr_string, MAX_BEARERS); - return NULL; - } + struct tipc_link *l; + struct tipc_msg *hdr; - if (n_ptr->links[b_ptr->identity]) { - tipc_addr_string_fill(addr_string, n_ptr->addr); - pr_err("Attempt to establish second link on <%s> to %s\n", - b_ptr->name, addr_string); - return NULL; - } - - l_ptr = kzalloc(sizeof(*l_ptr), GFP_ATOMIC); - if (!l_ptr) { - pr_warn("Link creation failed, no memory\n"); - return NULL; - } - kref_init(&l_ptr->ref); - l_ptr->addr = peer; - if_name = strchr(b_ptr->name, ':') + 1; - sprintf(l_ptr->name, "%u.%u.%u:%s-%u.%u.%u:unknown", - tipc_zone(tn->own_addr), tipc_cluster(tn->own_addr), - tipc_node(tn->own_addr), - if_name, - tipc_zone(peer), tipc_cluster(peer), tipc_node(peer)); - /* note: peer i/f name is updated by reset/activate message */ - memcpy(&l_ptr->media_addr, media_addr, sizeof(*media_addr)); - l_ptr->owner = n_ptr; - l_ptr->checkpoint = 1; - l_ptr->peer_session = INVALID_SESSION; - l_ptr->bearer_id = b_ptr->identity; - link_set_supervision_props(l_ptr, b_ptr->tolerance); - l_ptr->state = RESET_UNKNOWN; - - l_ptr->pmsg = (struct tipc_msg *)&l_ptr->proto_msg; - msg = l_ptr->pmsg; - tipc_msg_init(tn->own_addr, msg, LINK_PROTOCOL, RESET_MSG, INT_H_SIZE, - l_ptr->addr); - msg_set_size(msg, sizeof(l_ptr->proto_msg)); - msg_set_session(msg, (tn->random & 0xffff)); - msg_set_bearer_id(msg, b_ptr->identity); - strcpy((char *)msg_data(msg), if_name); - l_ptr->net_plane = b_ptr->net_plane; - l_ptr->advertised_mtu = b_ptr->mtu; - l_ptr->mtu = l_ptr->advertised_mtu; - l_ptr->priority = b_ptr->priority; - tipc_link_set_queue_limits(l_ptr, b_ptr->window); - l_ptr->next_out_no = 1; - __skb_queue_head_init(&l_ptr->transmq); - __skb_queue_head_init(&l_ptr->backlogq); - __skb_queue_head_init(&l_ptr->deferdq); - skb_queue_head_init(&l_ptr->wakeupq); - skb_queue_head_init(&l_ptr->inputq); - skb_queue_head_init(&l_ptr->namedq); - link_reset_statistics(l_ptr); - tipc_node_attach_link(n_ptr, l_ptr); - setup_timer(&l_ptr->timer, link_timeout, (unsigned long)l_ptr); - link_state_event(l_ptr, STARTING_EVT); - - return l_ptr; + l = kzalloc(sizeof(*l), GFP_ATOMIC); + if (!l) + return false; + *link = l; + l->pmsg = (struct tipc_msg *)&l->proto_msg; + hdr = l->pmsg; + tipc_msg_init(ownnode, hdr, LINK_PROTOCOL, RESET_MSG, INT_H_SIZE, peer); + msg_set_size(hdr, sizeof(l->proto_msg)); + msg_set_session(hdr, session); + msg_set_bearer_id(hdr, l->bearer_id); + + /* Note: peer i/f name is completed by reset/activate message */ + sprintf(l->name, "%u.%u.%u:%s-%u.%u.%u:unknown", + tipc_zone(ownnode), tipc_cluster(ownnode), tipc_node(ownnode), + if_name, tipc_zone(peer), tipc_cluster(peer), tipc_node(peer)); + strcpy((char *)msg_data(hdr), if_name); + + l->addr = peer; + l->peer_caps = peer_caps; + l->net = net; + l->peer_session = WILDCARD_SESSION; + l->bearer_id = bearer_id; + l->tolerance = tolerance; + l->net_plane = net_plane; + l->advertised_mtu = mtu; + l->mtu = mtu; + l->priority = priority; + tipc_link_set_queue_limits(l, window); + l->ackers = 1; + l->bc_sndlink = bc_sndlink; + l->bc_rcvlink = bc_rcvlink; + l->inputq = inputq; + l->namedq = namedq; + l->state = LINK_RESETTING; + __skb_queue_head_init(&l->transmq); + __skb_queue_head_init(&l->backlogq); + __skb_queue_head_init(&l->deferdq); + skb_queue_head_init(&l->wakeupq); + skb_queue_head_init(l->inputq); + return true; } /** - * tipc_link_delete - Delete a link - * @l: link to be deleted + * tipc_link_bc_create - create new link to be used for broadcast + * @n: pointer to associated node + * @mtu: mtu to be used + * @window: send window to be used + * @inputq: queue to put messages ready for delivery + * @namedq: queue to put binding table update messages ready for delivery + * @link: return value, pointer to put the created link + * + * Returns true if link was created, otherwise false */ -void tipc_link_delete(struct tipc_link *l) +bool tipc_link_bc_create(struct net *net, u32 ownnode, u32 peer, + int mtu, int window, u16 peer_caps, + struct sk_buff_head *inputq, + struct sk_buff_head *namedq, + struct tipc_link *bc_sndlink, + struct tipc_link **link) { + struct tipc_link *l; + + if (!tipc_link_create(net, "", MAX_BEARERS, 0, 'Z', mtu, 0, window, + 0, ownnode, peer, peer_caps, bc_sndlink, + NULL, inputq, namedq, link)) + return false; + + l = *link; + strcpy(l->name, tipc_bclink_name); tipc_link_reset(l); - if (del_timer(&l->timer)) - tipc_link_put(l); - l->flags |= LINK_STOPPED; - /* Delete link now, or when timer is finished: */ - tipc_link_reset_fragments(l); - tipc_node_detach_link(l->owner, l); - tipc_link_put(l); + l->state = LINK_RESET; + l->ackers = 0; + l->bc_rcvlink = l; + + /* Broadcast send link is always up */ + if (link_is_bc_sndlink(l)) + l->state = LINK_ESTABLISHED; + + return true; } -void tipc_link_delete_list(struct net *net, unsigned int bearer_id, - bool shutting_down) +/** + * tipc_link_fsm_evt - link finite state machine + * @l: pointer to link + * @evt: state machine event to be processed + */ +int tipc_link_fsm_evt(struct tipc_link *l, int evt) { - struct tipc_net *tn = net_generic(net, tipc_net_id); - struct tipc_link *link; - struct tipc_node *node; + int rc = 0; - rcu_read_lock(); - list_for_each_entry_rcu(node, &tn->node_list, list) { - tipc_node_lock(node); - link = node->links[bearer_id]; - if (link) - tipc_link_delete(link); - tipc_node_unlock(node); + switch (l->state) { + case LINK_RESETTING: + switch (evt) { + case LINK_PEER_RESET_EVT: + l->state = LINK_PEER_RESET; + break; + case LINK_RESET_EVT: + l->state = LINK_RESET; + break; + case LINK_FAILURE_EVT: + case LINK_FAILOVER_BEGIN_EVT: + case LINK_ESTABLISH_EVT: + case LINK_FAILOVER_END_EVT: + case LINK_SYNCH_BEGIN_EVT: + case LINK_SYNCH_END_EVT: + default: + goto illegal_evt; + } + break; + case LINK_RESET: + switch (evt) { + case LINK_PEER_RESET_EVT: + l->state = LINK_ESTABLISHING; + break; + case LINK_FAILOVER_BEGIN_EVT: + l->state = LINK_FAILINGOVER; + case LINK_FAILURE_EVT: + case LINK_RESET_EVT: + case LINK_ESTABLISH_EVT: + case LINK_FAILOVER_END_EVT: + break; + case LINK_SYNCH_BEGIN_EVT: + case LINK_SYNCH_END_EVT: + default: + goto illegal_evt; + } + break; + case LINK_PEER_RESET: + switch (evt) { + case LINK_RESET_EVT: + l->state = LINK_ESTABLISHING; + break; + case LINK_PEER_RESET_EVT: + case LINK_ESTABLISH_EVT: + case LINK_FAILURE_EVT: + break; + case LINK_SYNCH_BEGIN_EVT: + case LINK_SYNCH_END_EVT: + case LINK_FAILOVER_BEGIN_EVT: + case LINK_FAILOVER_END_EVT: + default: + goto illegal_evt; + } + break; + case LINK_FAILINGOVER: + switch (evt) { + case LINK_FAILOVER_END_EVT: + l->state = LINK_RESET; + break; + case LINK_PEER_RESET_EVT: + case LINK_RESET_EVT: + case LINK_ESTABLISH_EVT: + case LINK_FAILURE_EVT: + break; + case LINK_FAILOVER_BEGIN_EVT: + case LINK_SYNCH_BEGIN_EVT: + case LINK_SYNCH_END_EVT: + default: + goto illegal_evt; + } + break; + case LINK_ESTABLISHING: + switch (evt) { + case LINK_ESTABLISH_EVT: + l->state = LINK_ESTABLISHED; + break; + case LINK_FAILOVER_BEGIN_EVT: + l->state = LINK_FAILINGOVER; + break; + case LINK_RESET_EVT: + l->state = LINK_RESET; + break; + case LINK_FAILURE_EVT: + case LINK_PEER_RESET_EVT: + case LINK_SYNCH_BEGIN_EVT: + case LINK_FAILOVER_END_EVT: + break; + case LINK_SYNCH_END_EVT: + default: + goto illegal_evt; + } + break; + case LINK_ESTABLISHED: + switch (evt) { + case LINK_PEER_RESET_EVT: + l->state = LINK_PEER_RESET; + rc |= TIPC_LINK_DOWN_EVT; + break; + case LINK_FAILURE_EVT: + l->state = LINK_RESETTING; + rc |= TIPC_LINK_DOWN_EVT; + break; + case LINK_RESET_EVT: + l->state = LINK_RESET; + break; + case LINK_ESTABLISH_EVT: + case LINK_SYNCH_END_EVT: + break; + case LINK_SYNCH_BEGIN_EVT: + l->state = LINK_SYNCHING; + break; + case LINK_FAILOVER_BEGIN_EVT: + case LINK_FAILOVER_END_EVT: + default: + goto illegal_evt; + } + break; + case LINK_SYNCHING: + switch (evt) { + case LINK_PEER_RESET_EVT: + l->state = LINK_PEER_RESET; + rc |= TIPC_LINK_DOWN_EVT; + break; + case LINK_FAILURE_EVT: + l->state = LINK_RESETTING; + rc |= TIPC_LINK_DOWN_EVT; + break; + case LINK_RESET_EVT: + l->state = LINK_RESET; + break; + case LINK_ESTABLISH_EVT: + case LINK_SYNCH_BEGIN_EVT: + break; + case LINK_SYNCH_END_EVT: + l->state = LINK_ESTABLISHED; + break; + case LINK_FAILOVER_BEGIN_EVT: + case LINK_FAILOVER_END_EVT: + default: + goto illegal_evt; + } + break; + default: + pr_err("Unknown FSM state %x in %s\n", l->state, l->name); } - rcu_read_unlock(); + return rc; +illegal_evt: + pr_err("Illegal FSM event %x in state %x on link %s\n", + evt, l->state, l->name); + return rc; +} + +/* link_profile_stats - update statistical profiling of traffic + */ +static void link_profile_stats(struct tipc_link *l) +{ + struct sk_buff *skb; + struct tipc_msg *msg; + int length; + + /* Update counters used in statistical profiling of send traffic */ + l->stats.accu_queue_sz += skb_queue_len(&l->transmq); + l->stats.queue_sz_counts++; + + skb = skb_peek(&l->transmq); + if (!skb) + return; + msg = buf_msg(skb); + length = msg_size(msg); + + if (msg_user(msg) == MSG_FRAGMENTER) { + if (msg_type(msg) != FIRST_FRAGMENT) + return; + length = msg_size(msg_get_wrapped(msg)); + } + l->stats.msg_lengths_total += length; + l->stats.msg_length_counts++; + if (length <= 64) + l->stats.msg_length_profile[0]++; + else if (length <= 256) + l->stats.msg_length_profile[1]++; + else if (length <= 1024) + l->stats.msg_length_profile[2]++; + else if (length <= 4096) + l->stats.msg_length_profile[3]++; + else if (length <= 16384) + l->stats.msg_length_profile[4]++; + else if (length <= 32768) + l->stats.msg_length_profile[5]++; + else + l->stats.msg_length_profile[6]++; +} + +/* tipc_link_timeout - perform periodic task as instructed from node timeout + */ +/* tipc_link_timeout - perform periodic task as instructed from node timeout + */ +int tipc_link_timeout(struct tipc_link *l, struct sk_buff_head *xmitq) +{ + int rc = 0; + int mtyp = STATE_MSG; + bool xmit = false; + bool prb = false; + u16 bc_snt = l->bc_sndlink->snd_nxt - 1; + u16 bc_acked = l->bc_rcvlink->acked; + bool bc_up = link_is_up(l->bc_rcvlink); + + link_profile_stats(l); + + switch (l->state) { + case LINK_ESTABLISHED: + case LINK_SYNCHING: + if (!l->silent_intv_cnt) { + if (bc_up && (bc_acked != bc_snt)) + xmit = true; + } else if (l->silent_intv_cnt <= l->abort_limit) { + xmit = true; + prb = true; + } else { + rc |= tipc_link_fsm_evt(l, LINK_FAILURE_EVT); + } + l->silent_intv_cnt++; + break; + case LINK_RESET: + xmit = true; + mtyp = RESET_MSG; + break; + case LINK_ESTABLISHING: + xmit = true; + mtyp = ACTIVATE_MSG; + break; + case LINK_PEER_RESET: + case LINK_RESETTING: + case LINK_FAILINGOVER: + break; + default: + break; + } + + if (xmit) + tipc_link_build_proto_msg(l, mtyp, prb, 0, 0, 0, xmitq); + + return rc; } /** @@ -334,7 +607,7 @@ void tipc_link_delete_list(struct net *net, unsigned int bearer_id, * @link: congested link * @list: message that was attempted sent * Create pseudo msg to send back to user when congestion abates - * Only consumes message if there is an error + * Does not consume buffer list */ static int link_schedule_user(struct tipc_link *link, struct sk_buff_head *list) { @@ -347,8 +620,7 @@ static int link_schedule_user(struct tipc_link *link, struct sk_buff_head *list) /* This really cannot happen... */ if (unlikely(imp > TIPC_CRITICAL_IMPORTANCE)) { pr_warn("%s<%s>, send queue full", link_rst_msg, link->name); - tipc_link_reset(link); - goto err; + return -ENOBUFS; } /* Non-blocking sender: */ if (TIPC_SKB_CB(skb_peek(list))->wakeup_pending) @@ -358,15 +630,12 @@ static int link_schedule_user(struct tipc_link *link, struct sk_buff_head *list) skb = tipc_msg_create(SOCK_WAKEUP, 0, INT_H_SIZE, 0, addr, addr, oport, 0, 0); if (!skb) - goto err; + return -ENOBUFS; TIPC_SKB_CB(skb)->chain_sz = skb_queue_len(list); TIPC_SKB_CB(skb)->chain_imp = imp; skb_queue_tail(&link->wakeupq, skb); link->stats.link_congs++; return -ELINKCONG; -err: - __skb_queue_purge(list); - return -ENOBUFS; } /** @@ -388,782 +657,202 @@ void link_prepare_wakeup(struct tipc_link *l) if ((pnd[imp] + l->backlog[imp].len) >= lim) break; skb_unlink(skb, &l->wakeupq); - skb_queue_tail(&l->inputq, skb); - l->owner->inputq = &l->inputq; - l->owner->action_flags |= TIPC_MSG_EVT; + skb_queue_tail(l->inputq, skb); } } -/** - * tipc_link_reset_fragments - purge link's inbound message fragments queue - * @l_ptr: pointer to link - */ -void tipc_link_reset_fragments(struct tipc_link *l_ptr) +void tipc_link_reset(struct tipc_link *l) { - kfree_skb(l_ptr->reasm_buf); - l_ptr->reasm_buf = NULL; -} + /* Link is down, accept any session */ + l->peer_session = WILDCARD_SESSION; -static void tipc_link_purge_backlog(struct tipc_link *l) -{ + /* If peer is up, it only accepts an incremented session number */ + msg_set_session(l->pmsg, msg_session(l->pmsg) + 1); + + /* Prepare for renewed mtu size negotiation */ + l->mtu = l->advertised_mtu; + + /* Clean up all queues and counters: */ + __skb_queue_purge(&l->transmq); + __skb_queue_purge(&l->deferdq); + skb_queue_splice_init(&l->wakeupq, l->inputq); __skb_queue_purge(&l->backlogq); l->backlog[TIPC_LOW_IMPORTANCE].len = 0; l->backlog[TIPC_MEDIUM_IMPORTANCE].len = 0; l->backlog[TIPC_HIGH_IMPORTANCE].len = 0; l->backlog[TIPC_CRITICAL_IMPORTANCE].len = 0; l->backlog[TIPC_SYSTEM_IMPORTANCE].len = 0; + kfree_skb(l->reasm_buf); + kfree_skb(l->failover_reasm_skb); + l->reasm_buf = NULL; + l->failover_reasm_skb = NULL; + l->rcv_unacked = 0; + l->snd_nxt = 1; + l->rcv_nxt = 1; + l->acked = 0; + l->silent_intv_cnt = 0; + l->stats.recv_info = 0; + l->stale_count = 0; + l->bc_peer_is_up = false; + link_reset_statistics(l); } /** - * tipc_link_purge_queues - purge all pkt queues associated with link - * @l_ptr: pointer to link - */ -void tipc_link_purge_queues(struct tipc_link *l_ptr) -{ - __skb_queue_purge(&l_ptr->deferdq); - __skb_queue_purge(&l_ptr->transmq); - tipc_link_purge_backlog(l_ptr); - tipc_link_reset_fragments(l_ptr); -} - -void tipc_link_reset(struct tipc_link *l_ptr) -{ - u32 prev_state = l_ptr->state; - int was_active_link = tipc_link_is_active(l_ptr); - struct tipc_node *owner = l_ptr->owner; - struct tipc_link *pl = tipc_parallel_link(l_ptr); - - msg_set_session(l_ptr->pmsg, ((msg_session(l_ptr->pmsg) + 1) & 0xffff)); - - /* Link is down, accept any session */ - l_ptr->peer_session = INVALID_SESSION; - - /* Prepare for renewed mtu size negotiation */ - l_ptr->mtu = l_ptr->advertised_mtu; - - l_ptr->state = RESET_UNKNOWN; - - if ((prev_state == RESET_UNKNOWN) || (prev_state == RESET_RESET)) - return; - - tipc_node_link_down(l_ptr->owner, l_ptr); - tipc_bearer_remove_dest(owner->net, l_ptr->bearer_id, l_ptr->addr); - - if (was_active_link && tipc_node_is_up(l_ptr->owner) && (pl != l_ptr)) { - l_ptr->flags |= LINK_FAILINGOVER; - l_ptr->failover_checkpt = l_ptr->next_in_no; - pl->failover_pkts = FIRST_FAILOVER; - pl->failover_checkpt = l_ptr->next_in_no; - pl->failover_skb = l_ptr->reasm_buf; - } else { - kfree_skb(l_ptr->reasm_buf); - } - /* Clean up all queues, except inputq: */ - __skb_queue_purge(&l_ptr->transmq); - __skb_queue_purge(&l_ptr->deferdq); - if (!owner->inputq) - owner->inputq = &l_ptr->inputq; - skb_queue_splice_init(&l_ptr->wakeupq, owner->inputq); - if (!skb_queue_empty(owner->inputq)) - owner->action_flags |= TIPC_MSG_EVT; - tipc_link_purge_backlog(l_ptr); - l_ptr->reasm_buf = NULL; - l_ptr->rcv_unacked = 0; - l_ptr->checkpoint = 1; - l_ptr->next_out_no = 1; - l_ptr->fsm_msg_cnt = 0; - l_ptr->stale_count = 0; - link_reset_statistics(l_ptr); -} - -void tipc_link_reset_list(struct net *net, unsigned int bearer_id) -{ - struct tipc_net *tn = net_generic(net, tipc_net_id); - struct tipc_link *l_ptr; - struct tipc_node *n_ptr; - - rcu_read_lock(); - list_for_each_entry_rcu(n_ptr, &tn->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 *link) -{ - struct tipc_node *node = link->owner; - - link->next_in_no = 1; - link->stats.recv_info = 1; - tipc_node_link_up(node, link); - tipc_bearer_add_dest(node->net, link->bearer_id, link->addr); -} - -/** - * link_state_event - link finite state machine - * @l_ptr: pointer to link - * @event: state machine event to process - */ -static void link_state_event(struct tipc_link *l_ptr, unsigned int event) -{ - struct tipc_link *other; - unsigned long cont_intv = l_ptr->cont_intv; - - if (l_ptr->flags & LINK_STOPPED) - return; - - if (!(l_ptr->flags & LINK_STARTED) && (event != STARTING_EVT)) - return; /* Not yet. */ - - if (l_ptr->flags & LINK_FAILINGOVER) { - if (event == TIMEOUT_EVT) - link_set_timer(l_ptr, cont_intv); - return; - } - - switch (l_ptr->state) { - case WORKING_WORKING: - switch (event) { - case TRAFFIC_MSG_EVT: - case ACTIVATE_MSG: - break; - case TIMEOUT_EVT: - if (l_ptr->next_in_no != l_ptr->checkpoint) { - l_ptr->checkpoint = l_ptr->next_in_no; - if (tipc_bclink_acks_missing(l_ptr->owner)) { - tipc_link_proto_xmit(l_ptr, STATE_MSG, - 0, 0, 0, 0); - l_ptr->fsm_msg_cnt++; - } - link_set_timer(l_ptr, cont_intv); - break; - } - l_ptr->state = WORKING_UNKNOWN; - l_ptr->fsm_msg_cnt = 0; - tipc_link_proto_xmit(l_ptr, STATE_MSG, 1, 0, 0, 0); - l_ptr->fsm_msg_cnt++; - link_set_timer(l_ptr, cont_intv / 4); - break; - case RESET_MSG: - pr_debug("%s<%s>, requested by peer\n", - link_rst_msg, l_ptr->name); - tipc_link_reset(l_ptr); - l_ptr->state = RESET_RESET; - l_ptr->fsm_msg_cnt = 0; - tipc_link_proto_xmit(l_ptr, ACTIVATE_MSG, - 0, 0, 0, 0); - l_ptr->fsm_msg_cnt++; - link_set_timer(l_ptr, cont_intv); - break; - default: - pr_debug("%s%u in WW state\n", link_unk_evt, event); - } - break; - case WORKING_UNKNOWN: - switch (event) { - case TRAFFIC_MSG_EVT: - case ACTIVATE_MSG: - l_ptr->state = WORKING_WORKING; - l_ptr->fsm_msg_cnt = 0; - link_set_timer(l_ptr, cont_intv); - break; - case RESET_MSG: - pr_debug("%s<%s>, requested by peer while probing\n", - link_rst_msg, l_ptr->name); - tipc_link_reset(l_ptr); - l_ptr->state = RESET_RESET; - l_ptr->fsm_msg_cnt = 0; - tipc_link_proto_xmit(l_ptr, ACTIVATE_MSG, - 0, 0, 0, 0); - l_ptr->fsm_msg_cnt++; - link_set_timer(l_ptr, cont_intv); - break; - case TIMEOUT_EVT: - if (l_ptr->next_in_no != l_ptr->checkpoint) { - l_ptr->state = WORKING_WORKING; - l_ptr->fsm_msg_cnt = 0; - l_ptr->checkpoint = l_ptr->next_in_no; - if (tipc_bclink_acks_missing(l_ptr->owner)) { - tipc_link_proto_xmit(l_ptr, STATE_MSG, - 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_proto_xmit(l_ptr, STATE_MSG, - 1, 0, 0, 0); - l_ptr->fsm_msg_cnt++; - link_set_timer(l_ptr, cont_intv / 4); - } else { /* Link has failed */ - pr_debug("%s<%s>, peer not responding\n", - link_rst_msg, l_ptr->name); - tipc_link_reset(l_ptr); - l_ptr->state = RESET_UNKNOWN; - l_ptr->fsm_msg_cnt = 0; - tipc_link_proto_xmit(l_ptr, RESET_MSG, - 0, 0, 0, 0); - l_ptr->fsm_msg_cnt++; - link_set_timer(l_ptr, cont_intv); - } - break; - default: - pr_err("%s%u in WU state\n", link_unk_evt, event); - } - break; - case RESET_UNKNOWN: - switch (event) { - case TRAFFIC_MSG_EVT: - break; - case ACTIVATE_MSG: - other = l_ptr->owner->active_links[0]; - if (other && link_working_unknown(other)) - break; - l_ptr->state = WORKING_WORKING; - l_ptr->fsm_msg_cnt = 0; - link_activate(l_ptr); - tipc_link_proto_xmit(l_ptr, STATE_MSG, 1, 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_proto_xmit(l_ptr, ACTIVATE_MSG, - 1, 0, 0, 0); - l_ptr->fsm_msg_cnt++; - link_set_timer(l_ptr, cont_intv); - break; - case STARTING_EVT: - l_ptr->flags |= LINK_STARTED; - l_ptr->fsm_msg_cnt++; - link_set_timer(l_ptr, cont_intv); - break; - case TIMEOUT_EVT: - tipc_link_proto_xmit(l_ptr, RESET_MSG, 0, 0, 0, 0); - l_ptr->fsm_msg_cnt++; - link_set_timer(l_ptr, cont_intv); - break; - default: - pr_err("%s%u in RU state\n", link_unk_evt, event); - } - break; - case RESET_RESET: - switch (event) { - case TRAFFIC_MSG_EVT: - case ACTIVATE_MSG: - other = l_ptr->owner->active_links[0]; - if (other && link_working_unknown(other)) - break; - l_ptr->state = WORKING_WORKING; - l_ptr->fsm_msg_cnt = 0; - link_activate(l_ptr); - tipc_link_proto_xmit(l_ptr, STATE_MSG, 1, 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_proto_xmit(l_ptr, ACTIVATE_MSG, - 0, 0, 0, 0); - l_ptr->fsm_msg_cnt++; - link_set_timer(l_ptr, cont_intv); - break; - default: - pr_err("%s%u in RR state\n", link_unk_evt, event); - } - break; - default: - pr_err("Unknown link state %u/%u\n", l_ptr->state, event); - } -} - -/** - * __tipc_link_xmit(): same as tipc_link_xmit, but destlink is known & locked + * tipc_link_xmit(): enqueue buffer list according to queue situation * @link: link to use * @list: chain of buffers containing message + * @xmitq: returned list of packets to be sent by caller * * Consumes the buffer chain, except when returning -ELINKCONG, * since the caller then may want to make more send attempts. * Returns 0 if success, or errno: -ELINKCONG, -EMSGSIZE or -ENOBUFS * Messages at TIPC_SYSTEM_IMPORTANCE are always accepted */ -int __tipc_link_xmit(struct net *net, struct tipc_link *link, - struct sk_buff_head *list) +int tipc_link_xmit(struct tipc_link *l, struct sk_buff_head *list, + struct sk_buff_head *xmitq) { - struct tipc_msg *msg = buf_msg(skb_peek(list)); - unsigned int maxwin = link->window; - unsigned int imp = msg_importance(msg); - uint mtu = link->mtu; - uint ack = mod(link->next_in_no - 1); - uint seqno = link->next_out_no; - uint bc_last_in = link->owner->bclink.last_in; - struct tipc_media_addr *addr = &link->media_addr; - struct sk_buff_head *transmq = &link->transmq; - struct sk_buff_head *backlogq = &link->backlogq; - struct sk_buff *skb, *tmp; - - /* Match backlog limit against msg importance: */ - if (unlikely(link->backlog[imp].len >= link->backlog[imp].limit)) - return link_schedule_user(link, list); - - if (unlikely(msg_size(msg) > mtu)) { - __skb_queue_purge(list); - return -EMSGSIZE; + struct tipc_msg *hdr = buf_msg(skb_peek(list)); + unsigned int maxwin = l->window; + unsigned int i, imp = msg_importance(hdr); + unsigned int mtu = l->mtu; + u16 ack = l->rcv_nxt - 1; + u16 seqno = l->snd_nxt; + u16 bc_ack = l->bc_rcvlink->rcv_nxt - 1; + struct sk_buff_head *transmq = &l->transmq; + struct sk_buff_head *backlogq = &l->backlogq; + struct sk_buff *skb, *_skb, *bskb; + + /* Match msg importance against this and all higher backlog limits: */ + for (i = imp; i <= TIPC_SYSTEM_IMPORTANCE; i++) { + if (unlikely(l->backlog[i].len >= l->backlog[i].limit)) + return link_schedule_user(l, list); } + if (unlikely(msg_size(hdr) > mtu)) + return -EMSGSIZE; + /* Prepare each packet for sending, and add to relevant queue: */ - skb_queue_walk_safe(list, skb, tmp) { - __skb_unlink(skb, list); - msg = buf_msg(skb); - msg_set_seqno(msg, seqno); - msg_set_ack(msg, ack); - msg_set_bcast_ack(msg, bc_last_in); + while (skb_queue_len(list)) { + skb = skb_peek(list); + hdr = buf_msg(skb); + msg_set_seqno(hdr, seqno); + msg_set_ack(hdr, ack); + msg_set_bcast_ack(hdr, bc_ack); if (likely(skb_queue_len(transmq) < maxwin)) { + _skb = skb_clone(skb, GFP_ATOMIC); + if (!_skb) + return -ENOBUFS; + __skb_dequeue(list); __skb_queue_tail(transmq, skb); - tipc_bearer_send(net, link->bearer_id, skb, addr); - link->rcv_unacked = 0; + __skb_queue_tail(xmitq, _skb); + TIPC_SKB_CB(skb)->ackers = l->ackers; + l->rcv_unacked = 0; seqno++; continue; } - if (tipc_msg_bundle(skb_peek_tail(backlogq), skb, mtu)) { - link->stats.sent_bundled++; + if (tipc_msg_bundle(skb_peek_tail(backlogq), hdr, mtu)) { + kfree_skb(__skb_dequeue(list)); + l->stats.sent_bundled++; continue; } - if (tipc_msg_make_bundle(&skb, mtu, link->addr)) { - link->stats.sent_bundled++; - link->stats.sent_bundles++; - imp = msg_importance(buf_msg(skb)); + if (tipc_msg_make_bundle(&bskb, hdr, mtu, l->addr)) { + kfree_skb(__skb_dequeue(list)); + __skb_queue_tail(backlogq, bskb); + l->backlog[msg_importance(buf_msg(bskb))].len++; + l->stats.sent_bundled++; + l->stats.sent_bundles++; + continue; } - __skb_queue_tail(backlogq, skb); - link->backlog[imp].len++; - seqno++; + l->backlog[imp].len += skb_queue_len(list); + skb_queue_splice_tail_init(list, backlogq); } - link->next_out_no = seqno; - return 0; -} - -static void skb2list(struct sk_buff *skb, struct sk_buff_head *list) -{ - skb_queue_head_init(list); - __skb_queue_tail(list, skb); -} - -static int __tipc_link_xmit_skb(struct tipc_link *link, struct sk_buff *skb) -{ - struct sk_buff_head head; - - skb2list(skb, &head); - return __tipc_link_xmit(link->owner->net, link, &head); -} - -/* tipc_link_xmit_skb(): send single buffer to destination - * Buffers sent via this functon are generally TIPC_SYSTEM_IMPORTANCE - * messages, which will not be rejected - * The only exception is datagram messages rerouted after secondary - * lookup, which are rare and safe to dispose of anyway. - * TODO: Return real return value, and let callers use - * tipc_wait_for_sendpkt() where applicable - */ -int tipc_link_xmit_skb(struct net *net, struct sk_buff *skb, u32 dnode, - u32 selector) -{ - struct sk_buff_head head; - int rc; - - skb2list(skb, &head); - rc = tipc_link_xmit(net, &head, dnode, selector); - if (rc == -ELINKCONG) - kfree_skb(skb); + l->snd_nxt = seqno; return 0; } -/** - * tipc_link_xmit() is the general link level function for message sending - * @net: the applicable net namespace - * @list: chain of buffers containing message - * @dsz: amount of user data to be sent - * @dnode: address of destination node - * @selector: a number used for deterministic link selection - * Consumes the buffer chain, except when returning -ELINKCONG - * Returns 0 if success, otherwise errno: -ELINKCONG,-EHOSTUNREACH,-EMSGSIZE - */ -int tipc_link_xmit(struct net *net, struct sk_buff_head *list, u32 dnode, - u32 selector) -{ - struct tipc_link *link = NULL; - struct tipc_node *node; - int rc = -EHOSTUNREACH; - - node = tipc_node_find(net, dnode); - if (node) { - tipc_node_lock(node); - link = node->active_links[selector & 1]; - if (link) - rc = __tipc_link_xmit(net, link, list); - tipc_node_unlock(node); - tipc_node_put(node); - } - if (link) - return rc; - - if (likely(in_own_node(net, dnode))) { - tipc_sk_rcv(net, list); - return 0; - } - - __skb_queue_purge(list); - return rc; -} - -/* - * 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 *link) -{ - struct sk_buff *skb; - struct tipc_msg *msg; - - skb = tipc_buf_acquire(INT_H_SIZE); - if (!skb) - return; - - msg = buf_msg(skb); - tipc_msg_init(link_own_addr(link), msg, BCAST_PROTOCOL, STATE_MSG, - INT_H_SIZE, link->addr); - msg_set_last_bcast(msg, link->owner->bclink.acked); - __tipc_link_xmit_skb(link, skb); -} - -/* - * tipc_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_push_packets - push unsent packets to bearer - * - * Push out the unsent messages of a link where congestion - * has abated. Node is locked. - * - * Called with node locked - */ -void tipc_link_push_packets(struct tipc_link *link) +void tipc_link_advance_backlog(struct tipc_link *l, struct sk_buff_head *xmitq) { - struct sk_buff *skb; - struct tipc_msg *msg; - unsigned int ack = mod(link->next_in_no - 1); - - while (skb_queue_len(&link->transmq) < link->window) { - skb = __skb_dequeue(&link->backlogq); + struct sk_buff *skb, *_skb; + struct tipc_msg *hdr; + u16 seqno = l->snd_nxt; + u16 ack = l->rcv_nxt - 1; + u16 bc_ack = l->bc_rcvlink->rcv_nxt - 1; + + while (skb_queue_len(&l->transmq) < l->window) { + skb = skb_peek(&l->backlogq); if (!skb) - break; - msg = buf_msg(skb); - link->backlog[msg_importance(msg)].len--; - msg_set_ack(msg, ack); - msg_set_bcast_ack(msg, link->owner->bclink.last_in); - link->rcv_unacked = 0; - __skb_queue_tail(&link->transmq, skb); - tipc_bearer_send(link->owner->net, link->bearer_id, - skb, &link->media_addr); - } -} - -void tipc_link_reset_all(struct tipc_node *node) -{ - char addr_string[16]; - u32 i; - - tipc_node_lock(node); - - pr_warn("Resetting all links to %s\n", - tipc_addr_string_fill(addr_string, node->addr)); - - for (i = 0; i < MAX_BEARERS; i++) { - if (node->links[i]) { - link_print(node->links[i], "Resetting link\n"); - tipc_link_reset(node->links[i]); - } - } - - tipc_node_unlock(node); -} - -static void link_retransmit_failure(struct tipc_link *l_ptr, - struct sk_buff *buf) -{ - struct tipc_msg *msg = buf_msg(buf); - struct net *net = l_ptr->owner->net; - - pr_warn("Retransmission failure on link <%s>\n", l_ptr->name); - - if (l_ptr->addr) { - /* Handle failure on standard link */ - link_print(l_ptr, "Resetting link\n"); - tipc_link_reset(l_ptr); - - } else { - /* Handle failure on broadcast link */ - struct tipc_node *n_ptr; - char addr_string[16]; - - pr_info("Msg seq number: %u, ", msg_seqno(msg)); - pr_cont("Outstanding acks: %lu\n", - (unsigned long) TIPC_SKB_CB(buf)->handle); - - n_ptr = tipc_bclink_retransmit_to(net); - - tipc_addr_string_fill(addr_string, n_ptr->addr); - pr_info("Broadcast link info for %s\n", addr_string); - pr_info("Reception permitted: %d, Acked: %u\n", - n_ptr->bclink.recv_permitted, - n_ptr->bclink.acked); - pr_info("Last in: %u, Oos state: %u, Last sent: %u\n", - n_ptr->bclink.last_in, - n_ptr->bclink.oos_state, - n_ptr->bclink.last_sent); - - n_ptr->action_flags |= TIPC_BCAST_RESET; - l_ptr->stale_count = 0; - } -} - -void tipc_link_retransmit(struct tipc_link *l_ptr, struct sk_buff *skb, - u32 retransmits) -{ - struct tipc_msg *msg; - - if (!skb) - return; - - msg = buf_msg(skb); - - /* Detect repeated retransmit failures */ - if (l_ptr->last_retransmitted == msg_seqno(msg)) { - if (++l_ptr->stale_count > 100) { - link_retransmit_failure(l_ptr, skb); - return; - } - } else { - l_ptr->last_retransmitted = msg_seqno(msg); - l_ptr->stale_count = 1; - } - - skb_queue_walk_from(&l_ptr->transmq, skb) { - if (!retransmits) - break; - msg = buf_msg(skb); - msg_set_ack(msg, mod(l_ptr->next_in_no - 1)); - msg_set_bcast_ack(msg, l_ptr->owner->bclink.last_in); - tipc_bearer_send(l_ptr->owner->net, l_ptr->bearer_id, skb, - &l_ptr->media_addr); - retransmits--; - l_ptr->stats.retransmitted++; - } -} - -/* link_synch(): check if all packets arrived before the synch - * point have been consumed - * Returns true if the parallel links are synched, otherwise false - */ -static bool link_synch(struct tipc_link *l) -{ - unsigned int post_synch; - struct tipc_link *pl; - - pl = tipc_parallel_link(l); - if (pl == l) - goto synched; - - /* Was last pre-synch packet added to input queue ? */ - if (less_eq(pl->next_in_no, l->synch_point)) - return false; - - /* Is it still in the input queue ? */ - post_synch = mod(pl->next_in_no - l->synch_point) - 1; - if (skb_queue_len(&pl->inputq) > post_synch) - return false; -synched: - l->flags &= ~LINK_SYNCHING; - return true; -} - -static void link_retrieve_defq(struct tipc_link *link, - struct sk_buff_head *list) -{ - u32 seq_no; - - if (skb_queue_empty(&link->deferdq)) - return; - - seq_no = buf_seqno(skb_peek(&link->deferdq)); - if (seq_no == mod(link->next_in_no)) - skb_queue_splice_tail_init(&link->deferdq, list); -} - -/** - * tipc_rcv - process TIPC packets/messages arriving from off-node - * @net: the applicable net namespace - * @skb: TIPC packet - * @b_ptr: pointer to bearer message arrived on - * - * Invoked with no locks held. Bearer pointer must point to a valid bearer - * structure (i.e. cannot be NULL), but bearer can be inactive. - */ -void tipc_rcv(struct net *net, struct sk_buff *skb, struct tipc_bearer *b_ptr) -{ - struct tipc_net *tn = net_generic(net, tipc_net_id); - struct sk_buff_head head; - struct tipc_node *n_ptr; - struct tipc_link *l_ptr; - struct sk_buff *skb1, *tmp; - struct tipc_msg *msg; - u32 seq_no; - u32 ackd; - u32 released; - - skb2list(skb, &head); - - while ((skb = __skb_dequeue(&head))) { - /* Ensure message is well-formed */ - if (unlikely(!tipc_msg_validate(skb))) - goto discard; - - /* Handle arrival of a non-unicast link message */ - msg = buf_msg(skb); - if (unlikely(msg_non_seq(msg))) { - if (msg_user(msg) == LINK_CONFIG) - tipc_disc_rcv(net, skb, b_ptr); - else - tipc_bclink_rcv(net, skb); - continue; - } - - /* Discard unicast link messages destined for another node */ - if (unlikely(!msg_short(msg) && - (msg_destnode(msg) != tn->own_addr))) - goto discard; - - /* Locate neighboring node that sent message */ - n_ptr = tipc_node_find(net, msg_prevnode(msg)); - if (unlikely(!n_ptr)) - goto discard; - - tipc_node_lock(n_ptr); - /* Locate unicast link endpoint that should handle message */ - l_ptr = n_ptr->links[b_ptr->identity]; - if (unlikely(!l_ptr)) - goto unlock; - - /* Verify that communication with node is currently allowed */ - if ((n_ptr->action_flags & TIPC_WAIT_PEER_LINKS_DOWN) && - msg_user(msg) == LINK_PROTOCOL && - (msg_type(msg) == RESET_MSG || - msg_type(msg) == ACTIVATE_MSG) && - !msg_redundant_link(msg)) - n_ptr->action_flags &= ~TIPC_WAIT_PEER_LINKS_DOWN; - - if (tipc_node_blocked(n_ptr)) - goto unlock; - - /* Validate message sequence number info */ - seq_no = msg_seqno(msg); - ackd = msg_ack(msg); - - /* Release acked messages */ - if (unlikely(n_ptr->bclink.acked != msg_bcast_ack(msg))) - tipc_bclink_acknowledge(n_ptr, msg_bcast_ack(msg)); - - released = 0; - skb_queue_walk_safe(&l_ptr->transmq, skb1, tmp) { - if (more(buf_seqno(skb1), ackd)) - break; - __skb_unlink(skb1, &l_ptr->transmq); - kfree_skb(skb1); - released = 1; - } + break; + _skb = skb_clone(skb, GFP_ATOMIC); + if (!_skb) + break; + __skb_dequeue(&l->backlogq); + hdr = buf_msg(skb); + l->backlog[msg_importance(hdr)].len--; + __skb_queue_tail(&l->transmq, skb); + __skb_queue_tail(xmitq, _skb); + TIPC_SKB_CB(skb)->ackers = l->ackers; + msg_set_seqno(hdr, seqno); + msg_set_ack(hdr, ack); + msg_set_bcast_ack(hdr, bc_ack); + l->rcv_unacked = 0; + seqno++; + } + l->snd_nxt = seqno; +} - /* Try sending any messages link endpoint has pending */ - if (unlikely(skb_queue_len(&l_ptr->backlogq))) - tipc_link_push_packets(l_ptr); +static void link_retransmit_failure(struct tipc_link *l, struct sk_buff *skb) +{ + struct tipc_msg *hdr = buf_msg(skb); + + pr_warn("Retransmission failure on link <%s>\n", l->name); + link_print(l, "Resetting link "); + pr_info("Failed msg: usr %u, typ %u, len %u, err %u\n", + msg_user(hdr), msg_type(hdr), msg_size(hdr), msg_errcode(hdr)); + pr_info("sqno %u, prev: %x, src: %x\n", + msg_seqno(hdr), msg_prevnode(hdr), msg_orignode(hdr)); +} - if (released && !skb_queue_empty(&l_ptr->wakeupq)) - link_prepare_wakeup(l_ptr); +int tipc_link_retrans(struct tipc_link *l, u16 from, u16 to, + struct sk_buff_head *xmitq) +{ + struct sk_buff *_skb, *skb = skb_peek(&l->transmq); + struct tipc_msg *hdr; + u16 ack = l->rcv_nxt - 1; + u16 bc_ack = l->bc_rcvlink->rcv_nxt - 1; - /* Process the incoming packet */ - if (unlikely(!link_working_working(l_ptr))) { - if (msg_user(msg) == LINK_PROTOCOL) { - tipc_link_proto_rcv(l_ptr, skb); - link_retrieve_defq(l_ptr, &head); - skb = NULL; - goto unlock; - } + if (!skb) + return 0; - /* Traffic message. Conditionally activate link */ - link_state_event(l_ptr, TRAFFIC_MSG_EVT); + /* Detect repeated retransmit failures on same packet */ + if (likely(l->last_retransm != buf_seqno(skb))) { + l->last_retransm = buf_seqno(skb); + l->stale_count = 1; + } else if (++l->stale_count > 100) { + link_retransmit_failure(l, skb); + return tipc_link_fsm_evt(l, LINK_FAILURE_EVT); + } - if (link_working_working(l_ptr)) { - /* Re-insert buffer in front of queue */ - __skb_queue_head(&head, skb); - skb = NULL; - goto unlock; - } - goto unlock; - } + /* Move forward to where retransmission should start */ + skb_queue_walk(&l->transmq, skb) { + if (!less(buf_seqno(skb), from)) + break; + } - /* 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, skb); - link_retrieve_defq(l_ptr, &head); - skb = NULL; - goto unlock; - } - /* Synchronize with parallel link if applicable */ - if (unlikely((l_ptr->flags & LINK_SYNCHING) && !msg_dup(msg))) { - if (!link_synch(l_ptr)) - goto unlock; - } - l_ptr->next_in_no++; - if (unlikely(!skb_queue_empty(&l_ptr->deferdq))) - link_retrieve_defq(l_ptr, &head); - if (unlikely(++l_ptr->rcv_unacked >= TIPC_MIN_LINK_WIN)) { - l_ptr->stats.sent_acks++; - tipc_link_proto_xmit(l_ptr, STATE_MSG, 0, 0, 0, 0); - } - tipc_link_input(l_ptr, skb); - skb = NULL; -unlock: - tipc_node_unlock(n_ptr); - tipc_node_put(n_ptr); -discard: - if (unlikely(skb)) - kfree_skb(skb); + skb_queue_walk_from(&l->transmq, skb) { + if (more(buf_seqno(skb), to)) + break; + hdr = buf_msg(skb); + _skb = __pskb_copy(skb, MIN_H_SIZE, GFP_ATOMIC); + if (!_skb) + return 0; + hdr = buf_msg(_skb); + msg_set_ack(hdr, ack); + msg_set_bcast_ack(hdr, bc_ack); + _skb->priority = TC_PRIO_CONTROL; + __skb_queue_tail(xmitq, _skb); + l->stats.retransmitted++; } + return 0; } /* tipc_data_input - deliver data and name distr msgs to upper layer @@ -1171,29 +860,20 @@ discard: * Consumes buffer if message is of right type * Node lock must be held */ -static bool tipc_data_input(struct tipc_link *link, struct sk_buff *skb) +static bool tipc_data_input(struct tipc_link *l, struct sk_buff *skb, + struct sk_buff_head *inputq) { - struct tipc_node *node = link->owner; - struct tipc_msg *msg = buf_msg(skb); - u32 dport = msg_destport(msg); - - switch (msg_user(msg)) { + switch (msg_user(buf_msg(skb))) { case TIPC_LOW_IMPORTANCE: case TIPC_MEDIUM_IMPORTANCE: case TIPC_HIGH_IMPORTANCE: case TIPC_CRITICAL_IMPORTANCE: case CONN_MANAGER: - if (tipc_skb_queue_tail(&link->inputq, skb, dport)) { - node->inputq = &link->inputq; - node->action_flags |= TIPC_MSG_EVT; - } + skb_queue_tail(inputq, skb); return true; case NAME_DISTRIBUTOR: - node->bclink.recv_permitted = true; - node->namedq = &link->namedq; - skb_queue_tail(&link->namedq, skb); - if (skb_queue_len(&link->namedq) == 1) - node->action_flags |= TIPC_NAMED_MSG_EVT; + l->bc_rcvlink->state = LINK_ESTABLISHED; + skb_queue_tail(l->namedq, skb); return true; case MSG_BUNDLER: case TUNNEL_PROTOCOL: @@ -1210,540 +890,629 @@ static bool tipc_data_input(struct tipc_link *link, struct sk_buff *skb) /* tipc_link_input - process packet that has passed link protocol check * * Consumes buffer - * Node lock must be held */ -static void tipc_link_input(struct tipc_link *link, struct sk_buff *skb) +static int tipc_link_input(struct tipc_link *l, struct sk_buff *skb, + struct sk_buff_head *inputq) { - struct tipc_node *node = link->owner; - struct tipc_msg *msg = buf_msg(skb); + struct tipc_msg *hdr = buf_msg(skb); + struct sk_buff **reasm_skb = &l->reasm_buf; struct sk_buff *iskb; + struct sk_buff_head tmpq; + int usr = msg_user(hdr); + int rc = 0; int pos = 0; + int ipos = 0; - if (likely(tipc_data_input(link, skb))) - return; - - switch (msg_user(msg)) { - case TUNNEL_PROTOCOL: - if (msg_dup(msg)) { - link->flags |= LINK_SYNCHING; - link->synch_point = msg_seqno(msg_get_wrapped(msg)); - kfree_skb(skb); - break; + if (unlikely(usr == TUNNEL_PROTOCOL)) { + if (msg_type(hdr) == SYNCH_MSG) { + __skb_queue_purge(&l->deferdq); + goto drop; } - if (!tipc_link_failover_rcv(link, &skb)) - break; - if (msg_user(buf_msg(skb)) != MSG_BUNDLER) { - tipc_data_input(link, skb); - break; - } - case MSG_BUNDLER: - link->stats.recv_bundles++; - link->stats.recv_bundled += msg_msgcnt(msg); + if (!tipc_msg_extract(skb, &iskb, &ipos)) + return rc; + kfree_skb(skb); + skb = iskb; + hdr = buf_msg(skb); + if (less(msg_seqno(hdr), l->drop_point)) + goto drop; + if (tipc_data_input(l, skb, inputq)) + return rc; + usr = msg_user(hdr); + reasm_skb = &l->failover_reasm_skb; + } + if (usr == MSG_BUNDLER) { + skb_queue_head_init(&tmpq); + l->stats.recv_bundles++; + l->stats.recv_bundled += msg_msgcnt(hdr); while (tipc_msg_extract(skb, &iskb, &pos)) - tipc_data_input(link, iskb); - break; - case MSG_FRAGMENTER: - link->stats.recv_fragments++; - if (tipc_buf_append(&link->reasm_buf, &skb)) { - link->stats.recv_fragmented++; - tipc_data_input(link, skb); - } else if (!link->reasm_buf) { - tipc_link_reset(link); + tipc_data_input(l, iskb, &tmpq); + tipc_skb_queue_splice_tail(&tmpq, inputq); + return 0; + } else if (usr == MSG_FRAGMENTER) { + l->stats.recv_fragments++; + if (tipc_buf_append(reasm_skb, &skb)) { + l->stats.recv_fragmented++; + tipc_data_input(l, skb, inputq); + } else if (!*reasm_skb && !link_is_bc_rcvlink(l)) { + pr_warn_ratelimited("Unable to build fragment list\n"); + return tipc_link_fsm_evt(l, LINK_FAILURE_EVT); } - break; - case BCAST_PROTOCOL: - tipc_link_sync_rcv(node, skb); - break; - default: - break; - }; + return 0; + } else if (usr == BCAST_PROTOCOL) { + tipc_bcast_lock(l->net); + tipc_link_bc_init_rcv(l->bc_rcvlink, hdr); + tipc_bcast_unlock(l->net); + } +drop: + kfree_skb(skb); + return 0; } -/** - * tipc_link_defer_pkt - Add out-of-sequence message to deferred reception queue - * - * Returns increase in queue length (i.e. 0 or 1) - */ -u32 tipc_link_defer_pkt(struct sk_buff_head *list, struct sk_buff *skb) +static bool tipc_link_release_pkts(struct tipc_link *l, u16 acked) { - struct sk_buff *skb1; - u32 seq_no = buf_seqno(skb); + bool released = false; + struct sk_buff *skb, *tmp; - /* Empty queue ? */ - if (skb_queue_empty(list)) { - __skb_queue_tail(list, skb); - return 1; + skb_queue_walk_safe(&l->transmq, skb, tmp) { + if (more(buf_seqno(skb), acked)) + break; + __skb_unlink(skb, &l->transmq); + kfree_skb(skb); + released = true; } + return released; +} + +/* tipc_link_build_ack_msg: prepare link acknowledge message for transmission + * + * Note that sending of broadcast ack is coordinated among nodes, to reduce + * risk of ack storms towards the sender + */ +int tipc_link_build_ack_msg(struct tipc_link *l, struct sk_buff_head *xmitq) +{ + if (!l) + return 0; - /* Last ? */ - if (less(buf_seqno(skb_peek_tail(list)), seq_no)) { - __skb_queue_tail(list, skb); - return 1; + /* Broadcast ACK must be sent via a unicast link => defer to caller */ + if (link_is_bc_rcvlink(l)) { + if (((l->rcv_nxt ^ link_own_addr(l)) & 0xf) != 0xf) + return 0; + l->rcv_unacked = 0; + return TIPC_LINK_SND_BC_ACK; } - /* Locate insertion point in queue, then insert; discard if duplicate */ - skb_queue_walk(list, skb1) { - u32 curr_seqno = buf_seqno(skb1); + /* Unicast ACK */ + l->rcv_unacked = 0; + l->stats.sent_acks++; + tipc_link_build_proto_msg(l, STATE_MSG, 0, 0, 0, 0, xmitq); + return 0; +} - if (seq_no == curr_seqno) { - kfree_skb(skb); - return 0; - } +/* tipc_link_build_reset_msg: prepare link RESET or ACTIVATE message + */ +void tipc_link_build_reset_msg(struct tipc_link *l, struct sk_buff_head *xmitq) +{ + int mtyp = RESET_MSG; - if (less(seq_no, curr_seqno)) - break; - } + if (l->state == LINK_ESTABLISHING) + mtyp = ACTIVATE_MSG; - __skb_queue_before(list, skb1, skb); - return 1; + tipc_link_build_proto_msg(l, mtyp, 0, 0, 0, 0, xmitq); } -/* - * link_handle_out_of_seq_msg - handle arrival of out-of-sequence packet +/* tipc_link_build_nack_msg: prepare link nack message for transmission */ -static void link_handle_out_of_seq_msg(struct tipc_link *l_ptr, - struct sk_buff *buf) +static void tipc_link_build_nack_msg(struct tipc_link *l, + struct sk_buff_head *xmitq) { - u32 seq_no = buf_seqno(buf); + u32 def_cnt = ++l->stats.deferred_recv; - if (likely(msg_user(buf_msg(buf)) == LINK_PROTOCOL)) { - tipc_link_proto_rcv(l_ptr, buf); + if (link_is_bc_rcvlink(l)) return; - } - /* Record OOS packet arrival (force mismatch on next timeout) */ - l_ptr->checkpoint--; + if ((skb_queue_len(&l->deferdq) == 1) || !(def_cnt % TIPC_NACK_INTV)) + tipc_link_build_proto_msg(l, STATE_MSG, 0, 0, 0, 0, xmitq); +} - /* - * Discard packet if a duplicate; otherwise add it to deferred queue - * and notify peer of gap as per protocol specification - */ - if (less(seq_no, mod(l_ptr->next_in_no))) { - l_ptr->stats.duplicates++; - kfree_skb(buf); - return; - } +/* tipc_link_rcv - process TIPC packets/messages arriving from off-node + * @l: the link that should handle the message + * @skb: TIPC packet + * @xmitq: queue to place packets to be sent after this call + */ +int tipc_link_rcv(struct tipc_link *l, struct sk_buff *skb, + struct sk_buff_head *xmitq) +{ + struct sk_buff_head *defq = &l->deferdq; + struct tipc_msg *hdr; + u16 seqno, rcv_nxt, win_lim; + int rc = 0; + + do { + hdr = buf_msg(skb); + seqno = msg_seqno(hdr); + rcv_nxt = l->rcv_nxt; + win_lim = rcv_nxt + TIPC_MAX_LINK_WIN; + + /* Verify and update link state */ + if (unlikely(msg_user(hdr) == LINK_PROTOCOL)) + return tipc_link_proto_rcv(l, skb, xmitq); + + if (unlikely(!link_is_up(l))) { + if (l->state == LINK_ESTABLISHING) + rc = TIPC_LINK_UP_EVT; + goto drop; + } - if (tipc_link_defer_pkt(&l_ptr->deferdq, buf)) { - l_ptr->stats.deferred_recv++; - if ((skb_queue_len(&l_ptr->deferdq) % TIPC_MIN_LINK_WIN) == 1) - tipc_link_proto_xmit(l_ptr, STATE_MSG, 0, 0, 0, 0); - } else { - l_ptr->stats.duplicates++; - } + /* Don't send probe at next timeout expiration */ + l->silent_intv_cnt = 0; + + /* Drop if outside receive window */ + if (unlikely(less(seqno, rcv_nxt) || more(seqno, win_lim))) { + l->stats.duplicates++; + goto drop; + } + + /* Forward queues and wake up waiting users */ + if (likely(tipc_link_release_pkts(l, msg_ack(hdr)))) { + tipc_link_advance_backlog(l, xmitq); + if (unlikely(!skb_queue_empty(&l->wakeupq))) + link_prepare_wakeup(l); + } + + /* Defer delivery if sequence gap */ + if (unlikely(seqno != rcv_nxt)) { + __tipc_skb_queue_sorted(defq, seqno, skb); + tipc_link_build_nack_msg(l, xmitq); + break; + } + + /* Deliver packet */ + l->rcv_nxt++; + l->stats.recv_info++; + if (!tipc_data_input(l, skb, l->inputq)) + rc |= tipc_link_input(l, skb, l->inputq); + if (unlikely(++l->rcv_unacked >= TIPC_MIN_LINK_WIN)) + rc |= tipc_link_build_ack_msg(l, xmitq); + if (unlikely(rc & ~TIPC_LINK_SND_BC_ACK)) + break; + } while ((skb = __skb_dequeue(defq))); + + return rc; +drop: + kfree_skb(skb); + return rc; } /* * Send protocol message to the other endpoint. */ -void tipc_link_proto_xmit(struct tipc_link *l_ptr, u32 msg_typ, int probe_msg, +void tipc_link_proto_xmit(struct tipc_link *l, u32 msg_typ, int probe_msg, u32 gap, u32 tolerance, u32 priority) { - struct sk_buff *buf = NULL; - struct tipc_msg *msg = l_ptr->pmsg; - u32 msg_size = sizeof(l_ptr->proto_msg); - int r_flag; - - /* Don't send protocol message during link failover */ - if (l_ptr->flags & LINK_FAILINGOVER) - return; + struct sk_buff *skb = NULL; + struct sk_buff_head xmitq; - /* Abort non-RESET send if communication with node is prohibited */ - if ((tipc_node_blocked(l_ptr->owner)) && (msg_typ != RESET_MSG)) + __skb_queue_head_init(&xmitq); + tipc_link_build_proto_msg(l, msg_typ, probe_msg, gap, + tolerance, priority, &xmitq); + skb = __skb_dequeue(&xmitq); + if (!skb) return; + tipc_bearer_xmit_skb(l->net, l->bearer_id, skb, l->media_addr); + l->rcv_unacked = 0; +} - /* Create protocol message with "out-of-sequence" sequence number */ - msg_set_type(msg, msg_typ); - 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(l_ptr->owner->net)); +static void tipc_link_build_proto_msg(struct tipc_link *l, int mtyp, bool probe, + u16 rcvgap, int tolerance, int priority, + struct sk_buff_head *xmitq) +{ + struct sk_buff *skb = NULL; + struct tipc_msg *hdr = l->pmsg; + bool node_up = link_is_up(l->bc_rcvlink); - if (msg_typ == STATE_MSG) { - u32 next_sent = mod(l_ptr->next_out_no); + /* Don't send protocol message during reset or link failover */ + if (tipc_link_is_blocked(l)) + return; - if (!tipc_link_is_up(l_ptr)) + msg_set_type(hdr, mtyp); + msg_set_net_plane(hdr, l->net_plane); + msg_set_next_sent(hdr, l->snd_nxt); + msg_set_ack(hdr, l->rcv_nxt - 1); + msg_set_bcast_ack(hdr, l->bc_rcvlink->rcv_nxt - 1); + msg_set_last_bcast(hdr, l->bc_sndlink->snd_nxt - 1); + msg_set_link_tolerance(hdr, tolerance); + msg_set_linkprio(hdr, priority); + msg_set_redundant_link(hdr, node_up); + msg_set_seq_gap(hdr, 0); + + /* Compatibility: created msg must not be in sequence with pkt flow */ + msg_set_seqno(hdr, l->snd_nxt + U16_MAX / 2); + + if (mtyp == STATE_MSG) { + if (!tipc_link_is_up(l)) return; - if (skb_queue_len(&l_ptr->backlogq)) - next_sent = buf_seqno(skb_peek(&l_ptr->backlogq)); - msg_set_next_sent(msg, next_sent); - if (!skb_queue_empty(&l_ptr->deferdq)) { - u32 rec = buf_seqno(skb_peek(&l_ptr->deferdq)); - gap = mod(rec - mod(l_ptr->next_in_no)); + + /* Override rcvgap if there are packets in deferred queue */ + if (!skb_queue_empty(&l->deferdq)) + rcvgap = buf_seqno(skb_peek(&l->deferdq)) - l->rcv_nxt; + if (rcvgap) { + msg_set_seq_gap(hdr, rcvgap); + l->stats.sent_nacks++; } - msg_set_seq_gap(msg, gap); - if (gap) - l_ptr->stats.sent_nacks++; - msg_set_link_tolerance(msg, tolerance); - msg_set_linkprio(msg, priority); - msg_set_max_pkt(msg, l_ptr->mtu); - msg_set_ack(msg, mod(l_ptr->next_in_no - 1)); - msg_set_probe(msg, probe_msg != 0); - if (probe_msg) - l_ptr->stats.sent_probes++; - l_ptr->stats.sent_states++; - } else { /* RESET_MSG or ACTIVATE_MSG */ - msg_set_ack(msg, mod(l_ptr->failover_checkpt - 1)); - msg_set_seq_gap(msg, 0); - msg_set_next_sent(msg, 1); - msg_set_probe(msg, 0); - msg_set_link_tolerance(msg, l_ptr->tolerance); - msg_set_linkprio(msg, l_ptr->priority); - msg_set_max_pkt(msg, l_ptr->advertised_mtu); + msg_set_probe(hdr, probe); + if (probe) + l->stats.sent_probes++; + l->stats.sent_states++; + l->rcv_unacked = 0; + } else { + /* RESET_MSG or ACTIVATE_MSG */ + msg_set_max_pkt(hdr, l->advertised_mtu); + msg_set_ack(hdr, l->rcv_nxt - 1); + msg_set_next_sent(hdr, 1); } + skb = tipc_buf_acquire(msg_size(hdr)); + if (!skb) + return; + skb_copy_to_linear_data(skb, hdr, msg_size(hdr)); + skb->priority = TC_PRIO_CONTROL; + __skb_queue_tail(xmitq, skb); +} + +/* tipc_link_tnl_prepare(): prepare and return a list of tunnel packets + * with contents of the link's transmit and backlog queues. + */ +void tipc_link_tnl_prepare(struct tipc_link *l, struct tipc_link *tnl, + int mtyp, struct sk_buff_head *xmitq) +{ + struct sk_buff *skb, *tnlskb; + struct tipc_msg *hdr, tnlhdr; + struct sk_buff_head *queue = &l->transmq; + struct sk_buff_head tmpxq, tnlq; + u16 pktlen, pktcnt, seqno = l->snd_nxt; - r_flag = (l_ptr->owner->working_links > tipc_link_is_up(l_ptr)); - msg_set_redundant_link(msg, r_flag); - msg_set_linkprio(msg, l_ptr->priority); - msg_set_size(msg, msg_size); + if (!tnl) + return; - msg_set_seqno(msg, mod(l_ptr->next_out_no + (0xffff/2))); + skb_queue_head_init(&tnlq); + skb_queue_head_init(&tmpxq); - buf = tipc_buf_acquire(msg_size); - if (!buf) + /* At least one packet required for safe algorithm => add dummy */ + skb = tipc_msg_create(TIPC_LOW_IMPORTANCE, TIPC_DIRECT_MSG, + BASIC_H_SIZE, 0, l->addr, link_own_addr(l), + 0, 0, TIPC_ERR_NO_PORT); + if (!skb) { + pr_warn("%sunable to create tunnel packet\n", link_co_err); return; + } + skb_queue_tail(&tnlq, skb); + tipc_link_xmit(l, &tnlq, &tmpxq); + __skb_queue_purge(&tmpxq); + + /* Initialize reusable tunnel packet header */ + tipc_msg_init(link_own_addr(l), &tnlhdr, TUNNEL_PROTOCOL, + mtyp, INT_H_SIZE, l->addr); + pktcnt = skb_queue_len(&l->transmq) + skb_queue_len(&l->backlogq); + msg_set_msgcnt(&tnlhdr, pktcnt); + msg_set_bearer_id(&tnlhdr, l->peer_bearer_id); +tnl: + /* Wrap each packet into a tunnel packet */ + skb_queue_walk(queue, skb) { + hdr = buf_msg(skb); + if (queue == &l->backlogq) + msg_set_seqno(hdr, seqno++); + pktlen = msg_size(hdr); + msg_set_size(&tnlhdr, pktlen + INT_H_SIZE); + tnlskb = tipc_buf_acquire(pktlen + INT_H_SIZE); + if (!tnlskb) { + pr_warn("%sunable to send packet\n", link_co_err); + return; + } + skb_copy_to_linear_data(tnlskb, &tnlhdr, INT_H_SIZE); + skb_copy_to_linear_data_offset(tnlskb, INT_H_SIZE, hdr, pktlen); + __skb_queue_tail(&tnlq, tnlskb); + } + if (queue != &l->backlogq) { + queue = &l->backlogq; + goto tnl; + } + + tipc_link_xmit(tnl, &tnlq, xmitq); - skb_copy_to_linear_data(buf, msg, sizeof(l_ptr->proto_msg)); - buf->priority = TC_PRIO_CONTROL; - tipc_bearer_send(l_ptr->owner->net, l_ptr->bearer_id, buf, - &l_ptr->media_addr); - l_ptr->rcv_unacked = 0; - kfree_skb(buf); + if (mtyp == FAILOVER_MSG) { + tnl->drop_point = l->rcv_nxt; + tnl->failover_reasm_skb = l->reasm_buf; + l->reasm_buf = NULL; + } } -/* - * Receive protocol message : +/* tipc_link_proto_rcv(): receive link level protocol message : * Note that network plane id propagates through the network, and may - * change at any time. The node with lowest address rules + * change at any time. The node with lowest numerical id determines + * network plane */ -static void tipc_link_proto_rcv(struct tipc_link *l_ptr, - struct sk_buff *buf) +static int tipc_link_proto_rcv(struct tipc_link *l, struct sk_buff *skb, + struct sk_buff_head *xmitq) { - u32 rec_gap = 0; - u32 msg_tol; - struct tipc_msg *msg = buf_msg(buf); + struct tipc_msg *hdr = buf_msg(skb); + u16 rcvgap = 0; + u16 ack = msg_ack(hdr); + u16 gap = msg_seq_gap(hdr); + u16 peers_snd_nxt = msg_next_sent(hdr); + u16 peers_tol = msg_link_tolerance(hdr); + u16 peers_prio = msg_linkprio(hdr); + u16 rcv_nxt = l->rcv_nxt; + int mtyp = msg_type(hdr); + char *if_name; + int rc = 0; - if (l_ptr->flags & LINK_FAILINGOVER) + if (tipc_link_is_blocked(l) || !xmitq) goto exit; - if (l_ptr->net_plane != msg_net_plane(msg)) - if (link_own_addr(l_ptr) > msg_prevnode(msg)) - l_ptr->net_plane = msg_net_plane(msg); - - switch (msg_type(msg)) { + if (link_own_addr(l) > msg_prevnode(hdr)) + l->net_plane = msg_net_plane(hdr); + switch (mtyp) { case RESET_MSG: - if (!link_working_unknown(l_ptr) && - (l_ptr->peer_session != INVALID_SESSION)) { - if (less_eq(msg_session(msg), l_ptr->peer_session)) - break; /* duplicate or old reset: ignore */ - } - - if (!msg_redundant_link(msg) && (link_working_working(l_ptr) || - link_working_unknown(l_ptr))) { - /* - * peer has lost contact -- don't allow peer's links - * to reactivate before we recognize loss & clean up - */ - l_ptr->owner->action_flags |= TIPC_WAIT_OWN_LINKS_DOWN; - } - - link_state_event(l_ptr, RESET_MSG); + /* Ignore duplicate RESET with old session number */ + if ((less_eq(msg_session(hdr), l->peer_session)) && + (l->peer_session != WILDCARD_SESSION)) + break; /* fall thru' */ + case ACTIVATE_MSG: - /* Update link settings according other endpoint's values */ - strcpy((strrchr(l_ptr->name, ':') + 1), (char *)msg_data(msg)); - msg_tol = msg_link_tolerance(msg); - if (msg_tol > l_ptr->tolerance) - link_set_supervision_props(l_ptr, msg_tol); + /* Complete own link name with peer's interface name */ + if_name = strrchr(l->name, ':') + 1; + if (sizeof(l->name) - (if_name - l->name) <= TIPC_MAX_IF_NAME) + break; + if (msg_data_sz(hdr) < TIPC_MAX_IF_NAME) + break; + strncpy(if_name, msg_data(hdr), TIPC_MAX_IF_NAME); - if (msg_linkprio(msg) > l_ptr->priority) - l_ptr->priority = msg_linkprio(msg); + /* Update own tolerance if peer indicates a non-zero value */ + if (in_range(peers_tol, TIPC_MIN_LINK_TOL, TIPC_MAX_LINK_TOL)) + l->tolerance = peers_tol; - if (l_ptr->mtu > msg_max_pkt(msg)) - l_ptr->mtu = msg_max_pkt(msg); + /* Update own priority if peer's priority is higher */ + if (in_range(peers_prio, l->priority + 1, TIPC_MAX_LINK_PRI)) + l->priority = peers_prio; - /* 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; - } + /* ACTIVATE_MSG serves as PEER_RESET if link is already down */ + if ((mtyp == RESET_MSG) || !link_is_up(l)) + rc = tipc_link_fsm_evt(l, LINK_PEER_RESET_EVT); - l_ptr->peer_session = msg_session(msg); - l_ptr->peer_bearer_id = msg_bearer_id(msg); + /* ACTIVATE_MSG takes up link if it was already locally reset */ + if ((mtyp == ACTIVATE_MSG) && (l->state == LINK_ESTABLISHING)) + rc = TIPC_LINK_UP_EVT; - if (msg_type(msg) == ACTIVATE_MSG) - link_state_event(l_ptr, ACTIVATE_MSG); + l->peer_session = msg_session(hdr); + l->peer_bearer_id = msg_bearer_id(hdr); + if (l->mtu > msg_max_pkt(hdr)) + l->mtu = msg_max_pkt(hdr); break; + case STATE_MSG: - msg_tol = msg_link_tolerance(msg); - if (msg_tol) - link_set_supervision_props(l_ptr, msg_tol); - - if (msg_linkprio(msg) && - (msg_linkprio(msg) != l_ptr->priority)) { - pr_debug("%s<%s>, priority change %u->%u\n", - link_rst_msg, l_ptr->name, - l_ptr->priority, msg_linkprio(msg)); - l_ptr->priority = msg_linkprio(msg); - tipc_link_reset(l_ptr); /* Enforce change to take effect */ - break; - } + /* Update own tolerance if peer indicates a non-zero value */ + if (in_range(peers_tol, TIPC_MIN_LINK_TOL, TIPC_MAX_LINK_TOL)) + l->tolerance = peers_tol; - /* Record reception; force mismatch at next timeout: */ - l_ptr->checkpoint--; + l->silent_intv_cnt = 0; + l->stats.recv_states++; + if (msg_probe(hdr)) + l->stats.recv_probes++; - link_state_event(l_ptr, TRAFFIC_MSG_EVT); - l_ptr->stats.recv_states++; - if (link_reset_unknown(l_ptr)) + if (!link_is_up(l)) { + if (l->state == LINK_ESTABLISHING) + rc = TIPC_LINK_UP_EVT; break; - - if (less_eq(mod(l_ptr->next_in_no), msg_next_sent(msg))) { - rec_gap = mod(msg_next_sent(msg) - - mod(l_ptr->next_in_no)); } - if (msg_probe(msg)) - l_ptr->stats.recv_probes++; - - /* Protocol message before retransmits, reduce loss risk */ - 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_proto_xmit(l_ptr, STATE_MSG, 0, - rec_gap, 0, 0); + /* Send NACK if peer has sent pkts we haven't received yet */ + if (more(peers_snd_nxt, rcv_nxt) && !tipc_link_is_synching(l)) + rcvgap = peers_snd_nxt - l->rcv_nxt; + if (rcvgap || (msg_probe(hdr))) + tipc_link_build_proto_msg(l, STATE_MSG, 0, rcvgap, + 0, 0, xmitq); + tipc_link_release_pkts(l, ack); + + /* If NACK, retransmit will now start at right position */ + if (gap) { + rc = tipc_link_retrans(l, ack + 1, ack + gap, xmitq); + l->stats.recv_nacks++; } - if (msg_seq_gap(msg)) { - l_ptr->stats.recv_nacks++; - tipc_link_retransmit(l_ptr, skb_peek(&l_ptr->transmq), - msg_seq_gap(msg)); - } - break; + + tipc_link_advance_backlog(l, xmitq); + if (unlikely(!skb_queue_empty(&l->wakeupq))) + link_prepare_wakeup(l); } exit: - kfree_skb(buf); + kfree_skb(skb); + return rc; } - -/* tipc_link_tunnel_xmit(): Tunnel one packet via a link belonging to - * a different bearer. Owner node is locked. +/* tipc_link_build_bc_proto_msg() - create broadcast protocol message */ -static void tipc_link_tunnel_xmit(struct tipc_link *l_ptr, - struct tipc_msg *tunnel_hdr, - struct tipc_msg *msg, - u32 selector) +static bool tipc_link_build_bc_proto_msg(struct tipc_link *l, bool bcast, + u16 peers_snd_nxt, + struct sk_buff_head *xmitq) { - struct tipc_link *tunnel; struct sk_buff *skb; - u32 length = msg_size(msg); + struct tipc_msg *hdr; + struct sk_buff *dfrd_skb = skb_peek(&l->deferdq); + u16 ack = l->rcv_nxt - 1; + u16 gap_to = peers_snd_nxt - 1; - tunnel = l_ptr->owner->active_links[selector & 1]; - if (!tipc_link_is_up(tunnel)) { - pr_warn("%stunnel link no longer available\n", link_co_err); - return; - } - msg_set_size(tunnel_hdr, length + INT_H_SIZE); - skb = tipc_buf_acquire(length + INT_H_SIZE); - if (!skb) { - pr_warn("%sunable to send tunnel msg\n", link_co_err); - return; - } - skb_copy_to_linear_data(skb, tunnel_hdr, INT_H_SIZE); - skb_copy_to_linear_data_offset(skb, INT_H_SIZE, msg, length); - __tipc_link_xmit_skb(tunnel, skb); + skb = tipc_msg_create(BCAST_PROTOCOL, STATE_MSG, INT_H_SIZE, + 0, l->addr, link_own_addr(l), 0, 0, 0); + if (!skb) + return false; + hdr = buf_msg(skb); + msg_set_last_bcast(hdr, l->bc_sndlink->snd_nxt - 1); + msg_set_bcast_ack(hdr, ack); + msg_set_bcgap_after(hdr, ack); + if (dfrd_skb) + gap_to = buf_seqno(dfrd_skb) - 1; + msg_set_bcgap_to(hdr, gap_to); + msg_set_non_seq(hdr, bcast); + __skb_queue_tail(xmitq, skb); + return true; } +/* tipc_link_build_bc_init_msg() - synchronize broadcast link endpoints. + * + * Give a newly added peer node the sequence number where it should + * start receiving and acking broadcast packets. + */ +static void tipc_link_build_bc_init_msg(struct tipc_link *l, + struct sk_buff_head *xmitq) +{ + struct sk_buff_head list; + + __skb_queue_head_init(&list); + if (!tipc_link_build_bc_proto_msg(l->bc_rcvlink, false, 0, &list)) + return; + tipc_link_xmit(l, &list, xmitq); +} -/* 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. +/* tipc_link_bc_init_rcv - receive initial broadcast synch data from peer */ -void tipc_link_failover_send_queue(struct tipc_link *l_ptr) +void tipc_link_bc_init_rcv(struct tipc_link *l, struct tipc_msg *hdr) { - int msgcount; - struct tipc_link *tunnel = l_ptr->owner->active_links[0]; - struct tipc_msg tunnel_hdr; - struct sk_buff *skb; - int split_bundles; + int mtyp = msg_type(hdr); + u16 peers_snd_nxt = msg_bc_snd_nxt(hdr); - if (!tunnel) + if (link_is_up(l)) return; - tipc_msg_init(link_own_addr(l_ptr), &tunnel_hdr, TUNNEL_PROTOCOL, - FAILOVER_MSG, INT_H_SIZE, l_ptr->addr); - skb_queue_splice_tail_init(&l_ptr->backlogq, &l_ptr->transmq); - tipc_link_purge_backlog(l_ptr); - msgcount = skb_queue_len(&l_ptr->transmq); - msg_set_bearer_id(&tunnel_hdr, l_ptr->peer_bearer_id); - msg_set_msgcnt(&tunnel_hdr, msgcount); - - if (skb_queue_empty(&l_ptr->transmq)) { - skb = tipc_buf_acquire(INT_H_SIZE); - if (skb) { - skb_copy_to_linear_data(skb, &tunnel_hdr, INT_H_SIZE); - msg_set_size(&tunnel_hdr, INT_H_SIZE); - __tipc_link_xmit_skb(tunnel, skb); - } else { - pr_warn("%sunable to send changeover msg\n", - link_co_err); - } + if (msg_user(hdr) == BCAST_PROTOCOL) { + l->rcv_nxt = peers_snd_nxt; + l->state = LINK_ESTABLISHED; return; } - split_bundles = (l_ptr->owner->active_links[0] != - l_ptr->owner->active_links[1]); - - skb_queue_walk(&l_ptr->transmq, skb) { - struct tipc_msg *msg = buf_msg(skb); + if (l->peer_caps & TIPC_BCAST_SYNCH) + return; - if ((msg_user(msg) == MSG_BUNDLER) && split_bundles) { - struct tipc_msg *m = msg_get_wrapped(msg); - unchar *pos = (unchar *)m; + if (msg_peer_node_is_up(hdr)) + return; - msgcount = msg_msgcnt(msg); - while (msgcount--) { - msg_set_seqno(m, msg_seqno(msg)); - tipc_link_tunnel_xmit(l_ptr, &tunnel_hdr, m, - msg_link_selector(m)); - pos += align(msg_size(m)); - m = (struct tipc_msg *)pos; - } - } else { - tipc_link_tunnel_xmit(l_ptr, &tunnel_hdr, msg, - msg_link_selector(msg)); - } - } + /* Compatibility: accept older, less safe initial synch data */ + if ((mtyp == RESET_MSG) || (mtyp == ACTIVATE_MSG)) + l->rcv_nxt = peers_snd_nxt; } -/* tipc_link_dup_queue_xmit(): A second link has become active. Tunnel a - * duplicate of the first link's send queue via the new link. This way, we - * are guaranteed that currently queued packets from a socket are delivered - * before future traffic from the same socket, even if this is using the - * new link. The last arriving copy of each duplicate packet is dropped at - * the receiving end by the regular protocol check, so packet cardinality - * and sequence order is preserved per sender/receiver socket pair. - * Owner node is locked. +/* tipc_link_bc_sync_rcv - update rcv link according to peer's send state */ -void tipc_link_dup_queue_xmit(struct tipc_link *link, - struct tipc_link *tnl) +void tipc_link_bc_sync_rcv(struct tipc_link *l, struct tipc_msg *hdr, + struct sk_buff_head *xmitq) { - struct sk_buff *skb; - struct tipc_msg tnl_hdr; - struct sk_buff_head *queue = &link->transmq; - int mcnt; + u16 peers_snd_nxt = msg_bc_snd_nxt(hdr); - tipc_msg_init(link_own_addr(link), &tnl_hdr, TUNNEL_PROTOCOL, - SYNCH_MSG, INT_H_SIZE, link->addr); - mcnt = skb_queue_len(&link->transmq) + skb_queue_len(&link->backlogq); - msg_set_msgcnt(&tnl_hdr, mcnt); - msg_set_bearer_id(&tnl_hdr, link->peer_bearer_id); + if (!link_is_up(l)) + return; -tunnel_queue: - skb_queue_walk(queue, skb) { - struct sk_buff *outskb; - struct tipc_msg *msg = buf_msg(skb); - u32 len = msg_size(msg); - - msg_set_ack(msg, mod(link->next_in_no - 1)); - msg_set_bcast_ack(msg, link->owner->bclink.last_in); - msg_set_size(&tnl_hdr, len + INT_H_SIZE); - outskb = tipc_buf_acquire(len + INT_H_SIZE); - if (outskb == NULL) { - pr_warn("%sunable to send duplicate msg\n", - link_co_err); - return; - } - skb_copy_to_linear_data(outskb, &tnl_hdr, INT_H_SIZE); - skb_copy_to_linear_data_offset(outskb, INT_H_SIZE, - skb->data, len); - __tipc_link_xmit_skb(tnl, outskb); - if (!tipc_link_is_up(link)) - return; - } - if (queue == &link->backlogq) + if (!msg_peer_node_is_up(hdr)) return; - queue = &link->backlogq; - goto tunnel_queue; -} -/* tipc_link_failover_rcv(): Receive a tunnelled FAILOVER_MSG packet - * Owner node is locked. - */ -static bool tipc_link_failover_rcv(struct tipc_link *link, - struct sk_buff **skb) -{ - struct tipc_msg *msg = buf_msg(*skb); - struct sk_buff *iskb = NULL; - struct tipc_link *pl = NULL; - int bearer_id = msg_bearer_id(msg); - int pos = 0; + l->bc_peer_is_up = true; - if (msg_type(msg) != FAILOVER_MSG) { - pr_warn("%sunknown tunnel pkt received\n", link_co_err); - goto exit; + /* Ignore if peers_snd_nxt goes beyond receive window */ + if (more(peers_snd_nxt, l->rcv_nxt + l->window)) + return; + + if (!more(peers_snd_nxt, l->rcv_nxt)) { + l->nack_state = BC_NACK_SND_CONDITIONAL; + return; } - if (bearer_id >= MAX_BEARERS) - goto exit; - if (bearer_id == link->bearer_id) - goto exit; + /* Don't NACK if one was recently sent or peeked */ + if (l->nack_state == BC_NACK_SND_SUPPRESS) { + l->nack_state = BC_NACK_SND_UNCONDITIONAL; + return; + } - pl = link->owner->links[bearer_id]; - if (pl && tipc_link_is_up(pl)) - tipc_link_reset(pl); + /* Conditionally delay NACK sending until next synch rcv */ + if (l->nack_state == BC_NACK_SND_CONDITIONAL) { + l->nack_state = BC_NACK_SND_UNCONDITIONAL; + if ((peers_snd_nxt - l->rcv_nxt) < TIPC_MIN_LINK_WIN) + return; + } - if (link->failover_pkts == FIRST_FAILOVER) - link->failover_pkts = msg_msgcnt(msg); + /* Send NACK now but suppress next one */ + tipc_link_build_bc_proto_msg(l, true, peers_snd_nxt, xmitq); + l->nack_state = BC_NACK_SND_SUPPRESS; +} - /* Should we expect an inner packet? */ - if (!link->failover_pkts) - goto exit; +void tipc_link_bc_ack_rcv(struct tipc_link *l, u16 acked, + struct sk_buff_head *xmitq) +{ + struct sk_buff *skb, *tmp; + struct tipc_link *snd_l = l->bc_sndlink; - if (!tipc_msg_extract(*skb, &iskb, &pos)) { - pr_warn("%sno inner failover pkt\n", link_co_err); - *skb = NULL; - goto exit; - } - link->failover_pkts--; - *skb = NULL; + if (!link_is_up(l) || !l->bc_peer_is_up) + return; - /* Was this packet already delivered? */ - if (less(buf_seqno(iskb), link->failover_checkpt)) { - kfree_skb(iskb); - iskb = NULL; - goto exit; + if (!more(acked, l->acked)) + return; + + /* Skip over packets peer has already acked */ + skb_queue_walk(&snd_l->transmq, skb) { + if (more(buf_seqno(skb), l->acked)) + break; } - if (msg_user(buf_msg(iskb)) == MSG_FRAGMENTER) { - link->stats.recv_fragments++; - tipc_buf_append(&link->failover_skb, &iskb); + + /* Update/release the packets peer is acking now */ + skb_queue_walk_from_safe(&snd_l->transmq, skb, tmp) { + if (more(buf_seqno(skb), acked)) + break; + if (!--TIPC_SKB_CB(skb)->ackers) { + __skb_unlink(skb, &snd_l->transmq); + kfree_skb(skb); + } } -exit: - if (!link->failover_pkts && pl) - pl->flags &= ~LINK_FAILINGOVER; - kfree_skb(*skb); - *skb = iskb; - return *skb; + l->acked = acked; + tipc_link_advance_backlog(snd_l, xmitq); + if (unlikely(!skb_queue_empty(&snd_l->wakeupq))) + link_prepare_wakeup(snd_l); } -static void link_set_supervision_props(struct tipc_link *l_ptr, u32 tol) +/* tipc_link_bc_nack_rcv(): receive broadcast nack message + */ +int tipc_link_bc_nack_rcv(struct tipc_link *l, struct sk_buff *skb, + struct sk_buff_head *xmitq) { - unsigned long intv = ((tol / 4) > 500) ? 500 : tol / 4; + struct tipc_msg *hdr = buf_msg(skb); + u32 dnode = msg_destnode(hdr); + int mtyp = msg_type(hdr); + u16 acked = msg_bcast_ack(hdr); + u16 from = acked + 1; + u16 to = msg_bcgap_to(hdr); + u16 peers_snd_nxt = to + 1; + int rc = 0; + + kfree_skb(skb); + + if (!tipc_link_is_up(l) || !l->bc_peer_is_up) + return 0; - if ((tol < TIPC_MIN_LINK_TOL) || (tol > TIPC_MAX_LINK_TOL)) - return; + if (mtyp != STATE_MSG) + return 0; + + if (dnode == link_own_addr(l)) { + tipc_link_bc_ack_rcv(l, acked, xmitq); + rc = tipc_link_retrans(l->bc_sndlink, from, to, xmitq); + l->stats.recv_nacks++; + return rc; + } + + /* Msg for other node => suppress own NACK at next sync if applicable */ + if (more(peers_snd_nxt, l->rcv_nxt) && !less(l->rcv_nxt, from)) + l->nack_state = BC_NACK_SND_SUPPRESS; - l_ptr->tolerance = tol; - l_ptr->cont_intv = msecs_to_jiffies(intv); - l_ptr->abort_limit = tol / (jiffies_to_msecs(l_ptr->cont_intv) / 4); + return 0; } void tipc_link_set_queue_limits(struct tipc_link *l, u32 win) @@ -1780,7 +1549,7 @@ static struct tipc_node *tipc_link_find_owner(struct net *net, list_for_each_entry_rcu(n_ptr, &tn->node_list, list) { tipc_node_lock(n_ptr); for (i = 0; i < MAX_BEARERS; i++) { - l_ptr = n_ptr->links[i]; + l_ptr = n_ptr->links[i].link; if (l_ptr && !strcmp(l_ptr->name, link_name)) { *bearer_id = i; found_node = n_ptr; @@ -1803,31 +1572,20 @@ static struct tipc_node *tipc_link_find_owner(struct net *net, static void link_reset_statistics(struct tipc_link *l_ptr) { memset(&l_ptr->stats, 0, sizeof(l_ptr->stats)); - l_ptr->stats.sent_info = l_ptr->next_out_no; - l_ptr->stats.recv_info = l_ptr->next_in_no; + l_ptr->stats.sent_info = l_ptr->snd_nxt; + l_ptr->stats.recv_info = l_ptr->rcv_nxt; } -static void link_print(struct tipc_link *l_ptr, const char *str) +static void link_print(struct tipc_link *l, const char *str) { - struct tipc_net *tn = net_generic(l_ptr->owner->net, tipc_net_id); - struct tipc_bearer *b_ptr; - - rcu_read_lock(); - b_ptr = rcu_dereference_rtnl(tn->bearer_list[l_ptr->bearer_id]); - if (b_ptr) - pr_info("%s Link %x<%s>:", str, l_ptr->addr, b_ptr->name); - rcu_read_unlock(); - - if (link_working_unknown(l_ptr)) - pr_cont(":WU\n"); - else if (link_reset_reset(l_ptr)) - pr_cont(":RR\n"); - else if (link_reset_unknown(l_ptr)) - pr_cont(":RU\n"); - else if (link_working_working(l_ptr)) - pr_cont(":WW\n"); - else - pr_cont("\n"); + struct sk_buff *hskb = skb_peek(&l->transmq); + u16 head = hskb ? msg_seqno(buf_msg(hskb)) : l->snd_nxt - 1; + u16 tail = l->snd_nxt - 1; + + pr_info("%s Link <%s> state %x\n", str, l->name, l->state); + pr_info("XMTQ: %u [%u-%u], BKLGQ: %u, SNDNX: %u, RCVNX: %u\n", + skb_queue_len(&l->transmq), head, tail, + skb_queue_len(&l->backlogq), l->snd_nxt, l->rcv_nxt); } /* Parse and validate nested (link) properties valid for media, bearer and link @@ -1893,13 +1651,16 @@ int tipc_nl_link_set(struct sk_buff *skb, struct genl_info *info) name = nla_data(attrs[TIPC_NLA_LINK_NAME]); + if (strcmp(name, tipc_bclink_name) == 0) + return tipc_nl_bc_link_set(net, attrs); + node = tipc_link_find_owner(net, name, &bearer_id); if (!node) return -EINVAL; tipc_node_lock(node); - link = node->links[bearer_id]; + link = node->links[bearer_id].link; if (!link) { res = -EINVAL; goto out; @@ -1919,7 +1680,7 @@ int tipc_nl_link_set(struct sk_buff *skb, struct genl_info *info) u32 tol; tol = nla_get_u32(props[TIPC_NLA_PROP_TOL]); - link_set_supervision_props(link, tol); + link->tolerance = tol; tipc_link_proto_xmit(link, STATE_MSG, 0, 0, tol, 0); } if (props[TIPC_NLA_PROP_PRIO]) { @@ -2034,15 +1795,15 @@ static int __tipc_nl_add_link(struct net *net, struct tipc_nl_msg *msg, goto attr_msg_full; if (nla_put_u32(msg->skb, TIPC_NLA_LINK_MTU, link->mtu)) goto attr_msg_full; - if (nla_put_u32(msg->skb, TIPC_NLA_LINK_RX, link->next_in_no)) + if (nla_put_u32(msg->skb, TIPC_NLA_LINK_RX, link->rcv_nxt)) goto attr_msg_full; - if (nla_put_u32(msg->skb, TIPC_NLA_LINK_TX, link->next_out_no)) + if (nla_put_u32(msg->skb, TIPC_NLA_LINK_TX, link->snd_nxt)) goto attr_msg_full; if (tipc_link_is_up(link)) if (nla_put_flag(msg->skb, TIPC_NLA_LINK_UP)) goto attr_msg_full; - if (tipc_link_is_active(link)) + if (link->active) if (nla_put_flag(msg->skb, TIPC_NLA_LINK_ACTIVE)) goto attr_msg_full; @@ -2089,10 +1850,11 @@ static int __tipc_nl_add_node_links(struct net *net, struct tipc_nl_msg *msg, for (i = *prev_link; i < MAX_BEARERS; i++) { *prev_link = i; - if (!node->links[i]) + if (!node->links[i].link) continue; - err = __tipc_nl_add_link(net, msg, node->links[i], NLM_F_MULTI); + err = __tipc_nl_add_link(net, msg, + node->links[i].link, NLM_F_MULTI); if (err) return err; } @@ -2175,50 +1937,53 @@ out: int tipc_nl_link_get(struct sk_buff *skb, struct genl_info *info) { struct net *net = genl_info_net(info); - struct sk_buff *ans_skb; struct tipc_nl_msg msg; - struct tipc_link *link; - struct tipc_node *node; char *name; - int bearer_id; int err; + msg.portid = info->snd_portid; + msg.seq = info->snd_seq; + if (!info->attrs[TIPC_NLA_LINK_NAME]) return -EINVAL; - name = nla_data(info->attrs[TIPC_NLA_LINK_NAME]); - node = tipc_link_find_owner(net, name, &bearer_id); - if (!node) - return -EINVAL; - ans_skb = nlmsg_new(NLMSG_GOODSIZE, GFP_KERNEL); - if (!ans_skb) + msg.skb = nlmsg_new(NLMSG_GOODSIZE, GFP_KERNEL); + if (!msg.skb) return -ENOMEM; - msg.skb = ans_skb; - msg.portid = info->snd_portid; - msg.seq = info->snd_seq; - - tipc_node_lock(node); - link = node->links[bearer_id]; - if (!link) { - err = -EINVAL; - goto err_out; - } - - err = __tipc_nl_add_link(net, &msg, link, 0); - if (err) - goto err_out; + if (strcmp(name, tipc_bclink_name) == 0) { + err = tipc_nl_add_bc_link(net, &msg); + if (err) { + nlmsg_free(msg.skb); + return err; + } + } else { + int bearer_id; + struct tipc_node *node; + struct tipc_link *link; - tipc_node_unlock(node); + node = tipc_link_find_owner(net, name, &bearer_id); + if (!node) + return -EINVAL; - return genlmsg_reply(ans_skb, info); + tipc_node_lock(node); + link = node->links[bearer_id].link; + if (!link) { + tipc_node_unlock(node); + nlmsg_free(msg.skb); + return -EINVAL; + } -err_out: - tipc_node_unlock(node); - nlmsg_free(ans_skb); + err = __tipc_nl_add_link(net, &msg, link, 0); + tipc_node_unlock(node); + if (err) { + nlmsg_free(msg.skb); + return err; + } + } - return err; + return genlmsg_reply(msg.skb, info); } int tipc_nl_link_reset_stats(struct sk_buff *skb, struct genl_info *info) @@ -2258,7 +2023,7 @@ int tipc_nl_link_reset_stats(struct sk_buff *skb, struct genl_info *info) tipc_node_lock(node); - link = node->links[bearer_id]; + link = node->links[bearer_id].link; if (!link) { tipc_node_unlock(node); return -EINVAL;