1 /******************************************************************************
2 *******************************************************************************
4 ** Copyright (C) Sistina Software, Inc. 1997-2003 All rights reserved.
5 ** Copyright (C) 2004-2009 Red Hat, Inc. All rights reserved.
7 ** This copyrighted material is made available to anyone wishing to use,
8 ** modify, copy, or redistribute it subject to the terms and conditions
9 ** of the GNU General Public License v.2.
11 *******************************************************************************
12 ******************************************************************************/
17 * This is the "low-level" comms layer.
19 * It is responsible for sending/receiving messages
20 * from other nodes in the cluster.
22 * Cluster nodes are referred to by their nodeids. nodeids are
23 * simply 32 bit numbers to the locking module - if they need to
24 * be expanded for the cluster infrastructure then that is its
25 * responsibility. It is this layer's
26 * responsibility to resolve these into IP address or
27 * whatever it needs for inter-node communication.
29 * The comms level is two kernel threads that deal mainly with
30 * the receiving of messages from other nodes and passing them
31 * up to the mid-level comms layer (which understands the
32 * message format) for execution by the locking core, and
33 * a send thread which does all the setting up of connections
34 * to remote nodes and the sending of data. Threads are not allowed
35 * to send their own data because it may cause them to wait in times
36 * of high load. Also, this way, the sending thread can collect together
37 * messages bound for one node and send them in one block.
39 * lowcomms will choose to use either TCP or SCTP as its transport layer
40 * depending on the configuration variable 'protocol'. This should be set
41 * to 0 (default) for TCP or 1 for SCTP. It should be configured using a
42 * cluster-wide mechanism as it must be the same on all nodes of the cluster
43 * for the DLM to function.
47 #include <asm/ioctls.h>
50 #include <linux/pagemap.h>
51 #include <linux/file.h>
52 #include <linux/mutex.h>
53 #include <linux/sctp.h>
54 #include <linux/slab.h>
55 #include <net/sctp/sctp.h>
56 #include <net/sctp/user.h>
59 #include "dlm_internal.h"
64 #define NEEDED_RMEM (4*1024*1024)
65 #define CONN_HASH_SIZE 32
67 /* Number of messages to send before rescheduling */
68 #define MAX_SEND_MSG_COUNT 25
76 static void cbuf_add(struct cbuf *cb, int n)
81 static int cbuf_data(struct cbuf *cb)
83 return ((cb->base + cb->len) & cb->mask);
86 static void cbuf_init(struct cbuf *cb, int size)
88 cb->base = cb->len = 0;
92 static void cbuf_eat(struct cbuf *cb, int n)
99 static bool cbuf_empty(struct cbuf *cb)
105 struct socket *sock; /* NULL if not connected */
106 uint32_t nodeid; /* So we know who we are in the list */
107 struct mutex sock_mutex;
109 #define CF_READ_PENDING 1
110 #define CF_WRITE_PENDING 2
111 #define CF_CONNECT_PENDING 3
112 #define CF_INIT_PENDING 4
113 #define CF_IS_OTHERCON 5
115 #define CF_APP_LIMITED 7
116 struct list_head writequeue; /* List of outgoing writequeue_entries */
117 spinlock_t writequeue_lock;
118 int (*rx_action) (struct connection *); /* What to do when active */
119 void (*connect_action) (struct connection *); /* What to do to connect */
120 struct page *rx_page;
123 #define MAX_CONNECT_RETRIES 3
125 struct hlist_node list;
126 struct connection *othercon;
127 struct work_struct rwork; /* Receive workqueue */
128 struct work_struct swork; /* Send workqueue */
130 #define sock2con(x) ((struct connection *)(x)->sk_user_data)
132 /* An entry waiting to be sent */
133 struct writequeue_entry {
134 struct list_head list;
140 struct connection *con;
143 static struct sockaddr_storage *dlm_local_addr[DLM_MAX_ADDR_COUNT];
144 static int dlm_local_count;
145 static int dlm_allow_conn;
148 static struct workqueue_struct *recv_workqueue;
149 static struct workqueue_struct *send_workqueue;
151 static struct hlist_head connection_hash[CONN_HASH_SIZE];
152 static DEFINE_MUTEX(connections_lock);
153 static struct kmem_cache *con_cache;
155 static void process_recv_sockets(struct work_struct *work);
156 static void process_send_sockets(struct work_struct *work);
159 /* This is deliberately very simple because most clusters have simple
160 sequential nodeids, so we should be able to go straight to a connection
161 struct in the array */
162 static inline int nodeid_hash(int nodeid)
164 return nodeid & (CONN_HASH_SIZE-1);
167 static struct connection *__find_con(int nodeid)
170 struct hlist_node *h;
171 struct connection *con;
173 r = nodeid_hash(nodeid);
175 hlist_for_each_entry(con, h, &connection_hash[r], list) {
176 if (con->nodeid == nodeid)
183 * If 'allocation' is zero then we don't attempt to create a new
184 * connection structure for this node.
186 static struct connection *__nodeid2con(int nodeid, gfp_t alloc)
188 struct connection *con = NULL;
191 con = __find_con(nodeid);
195 con = kmem_cache_zalloc(con_cache, alloc);
199 r = nodeid_hash(nodeid);
200 hlist_add_head(&con->list, &connection_hash[r]);
202 con->nodeid = nodeid;
203 mutex_init(&con->sock_mutex);
204 INIT_LIST_HEAD(&con->writequeue);
205 spin_lock_init(&con->writequeue_lock);
206 INIT_WORK(&con->swork, process_send_sockets);
207 INIT_WORK(&con->rwork, process_recv_sockets);
209 /* Setup action pointers for child sockets */
211 struct connection *zerocon = __find_con(0);
213 con->connect_action = zerocon->connect_action;
215 con->rx_action = zerocon->rx_action;
221 /* Loop round all connections */
222 static void foreach_conn(void (*conn_func)(struct connection *c))
225 struct hlist_node *h, *n;
226 struct connection *con;
228 for (i = 0; i < CONN_HASH_SIZE; i++) {
229 hlist_for_each_entry_safe(con, h, n, &connection_hash[i], list){
235 static struct connection *nodeid2con(int nodeid, gfp_t allocation)
237 struct connection *con;
239 mutex_lock(&connections_lock);
240 con = __nodeid2con(nodeid, allocation);
241 mutex_unlock(&connections_lock);
246 /* This is a bit drastic, but only called when things go wrong */
247 static struct connection *assoc2con(int assoc_id)
250 struct hlist_node *h;
251 struct connection *con;
253 mutex_lock(&connections_lock);
255 for (i = 0 ; i < CONN_HASH_SIZE; i++) {
256 hlist_for_each_entry(con, h, &connection_hash[i], list) {
257 if (con->sctp_assoc == assoc_id) {
258 mutex_unlock(&connections_lock);
263 mutex_unlock(&connections_lock);
267 static int nodeid_to_addr(int nodeid, struct sockaddr *retaddr)
269 struct sockaddr_storage addr;
272 if (!dlm_local_count)
275 error = dlm_nodeid_to_addr(nodeid, &addr);
279 if (dlm_local_addr[0]->ss_family == AF_INET) {
280 struct sockaddr_in *in4 = (struct sockaddr_in *) &addr;
281 struct sockaddr_in *ret4 = (struct sockaddr_in *) retaddr;
282 ret4->sin_addr.s_addr = in4->sin_addr.s_addr;
284 struct sockaddr_in6 *in6 = (struct sockaddr_in6 *) &addr;
285 struct sockaddr_in6 *ret6 = (struct sockaddr_in6 *) retaddr;
286 ret6->sin6_addr = in6->sin6_addr;
292 /* Data available on socket or listen socket received a connect */
293 static void lowcomms_data_ready(struct sock *sk, int count_unused)
295 struct connection *con = sock2con(sk);
296 if (con && !test_and_set_bit(CF_READ_PENDING, &con->flags))
297 queue_work(recv_workqueue, &con->rwork);
300 static void lowcomms_write_space(struct sock *sk)
302 struct connection *con = sock2con(sk);
307 clear_bit(SOCK_NOSPACE, &con->sock->flags);
309 if (test_and_clear_bit(CF_APP_LIMITED, &con->flags)) {
310 con->sock->sk->sk_write_pending--;
311 clear_bit(SOCK_ASYNC_NOSPACE, &con->sock->flags);
314 if (!test_and_set_bit(CF_WRITE_PENDING, &con->flags))
315 queue_work(send_workqueue, &con->swork);
318 static inline void lowcomms_connect_sock(struct connection *con)
320 if (test_bit(CF_CLOSE, &con->flags))
322 if (!test_and_set_bit(CF_CONNECT_PENDING, &con->flags))
323 queue_work(send_workqueue, &con->swork);
326 static void lowcomms_state_change(struct sock *sk)
328 if (sk->sk_state == TCP_ESTABLISHED)
329 lowcomms_write_space(sk);
332 int dlm_lowcomms_connect_node(int nodeid)
334 struct connection *con;
336 /* with sctp there's no connecting without sending */
337 if (dlm_config.ci_protocol != 0)
340 if (nodeid == dlm_our_nodeid())
343 con = nodeid2con(nodeid, GFP_NOFS);
346 lowcomms_connect_sock(con);
350 /* Make a socket active */
351 static int add_sock(struct socket *sock, struct connection *con)
355 /* Install a data_ready callback */
356 con->sock->sk->sk_data_ready = lowcomms_data_ready;
357 con->sock->sk->sk_write_space = lowcomms_write_space;
358 con->sock->sk->sk_state_change = lowcomms_state_change;
359 con->sock->sk->sk_user_data = con;
360 con->sock->sk->sk_allocation = GFP_NOFS;
364 /* Add the port number to an IPv6 or 4 sockaddr and return the address
366 static void make_sockaddr(struct sockaddr_storage *saddr, uint16_t port,
369 saddr->ss_family = dlm_local_addr[0]->ss_family;
370 if (saddr->ss_family == AF_INET) {
371 struct sockaddr_in *in4_addr = (struct sockaddr_in *)saddr;
372 in4_addr->sin_port = cpu_to_be16(port);
373 *addr_len = sizeof(struct sockaddr_in);
374 memset(&in4_addr->sin_zero, 0, sizeof(in4_addr->sin_zero));
376 struct sockaddr_in6 *in6_addr = (struct sockaddr_in6 *)saddr;
377 in6_addr->sin6_port = cpu_to_be16(port);
378 *addr_len = sizeof(struct sockaddr_in6);
380 memset((char *)saddr + *addr_len, 0, sizeof(struct sockaddr_storage) - *addr_len);
383 /* Close a remote connection and tidy up */
384 static void close_connection(struct connection *con, bool and_other)
386 mutex_lock(&con->sock_mutex);
389 sock_release(con->sock);
392 if (con->othercon && and_other) {
393 /* Will only re-enter once. */
394 close_connection(con->othercon, false);
397 __free_page(con->rx_page);
402 mutex_unlock(&con->sock_mutex);
405 /* We only send shutdown messages to nodes that are not part of the cluster */
406 static void sctp_send_shutdown(sctp_assoc_t associd)
408 static char outcmsg[CMSG_SPACE(sizeof(struct sctp_sndrcvinfo))];
409 struct msghdr outmessage;
410 struct cmsghdr *cmsg;
411 struct sctp_sndrcvinfo *sinfo;
413 struct connection *con;
415 con = nodeid2con(0,0);
418 outmessage.msg_name = NULL;
419 outmessage.msg_namelen = 0;
420 outmessage.msg_control = outcmsg;
421 outmessage.msg_controllen = sizeof(outcmsg);
422 outmessage.msg_flags = MSG_EOR;
424 cmsg = CMSG_FIRSTHDR(&outmessage);
425 cmsg->cmsg_level = IPPROTO_SCTP;
426 cmsg->cmsg_type = SCTP_SNDRCV;
427 cmsg->cmsg_len = CMSG_LEN(sizeof(struct sctp_sndrcvinfo));
428 outmessage.msg_controllen = cmsg->cmsg_len;
429 sinfo = CMSG_DATA(cmsg);
430 memset(sinfo, 0x00, sizeof(struct sctp_sndrcvinfo));
432 sinfo->sinfo_flags |= MSG_EOF;
433 sinfo->sinfo_assoc_id = associd;
435 ret = kernel_sendmsg(con->sock, &outmessage, NULL, 0, 0);
438 log_print("send EOF to node failed: %d", ret);
441 static void sctp_init_failed_foreach(struct connection *con)
444 if (test_and_clear_bit(CF_CONNECT_PENDING, &con->flags)) {
445 if (!test_and_set_bit(CF_WRITE_PENDING, &con->flags))
446 queue_work(send_workqueue, &con->swork);
450 /* INIT failed but we don't know which node...
451 restart INIT on all pending nodes */
452 static void sctp_init_failed(void)
454 mutex_lock(&connections_lock);
456 foreach_conn(sctp_init_failed_foreach);
458 mutex_unlock(&connections_lock);
461 /* Something happened to an association */
462 static void process_sctp_notification(struct connection *con,
463 struct msghdr *msg, char *buf)
465 union sctp_notification *sn = (union sctp_notification *)buf;
467 if (sn->sn_header.sn_type == SCTP_ASSOC_CHANGE) {
468 switch (sn->sn_assoc_change.sac_state) {
473 /* Check that the new node is in the lockspace */
474 struct sctp_prim prim;
478 struct connection *new_con;
481 * We get this before any data for an association.
482 * We verify that the node is in the cluster and
483 * then peel off a socket for it.
485 if ((int)sn->sn_assoc_change.sac_assoc_id <= 0) {
486 log_print("COMM_UP for invalid assoc ID %d",
487 (int)sn->sn_assoc_change.sac_assoc_id);
491 memset(&prim, 0, sizeof(struct sctp_prim));
492 prim_len = sizeof(struct sctp_prim);
493 prim.ssp_assoc_id = sn->sn_assoc_change.sac_assoc_id;
495 ret = kernel_getsockopt(con->sock,
501 log_print("getsockopt/sctp_primary_addr on "
502 "new assoc %d failed : %d",
503 (int)sn->sn_assoc_change.sac_assoc_id,
506 /* Retry INIT later */
507 new_con = assoc2con(sn->sn_assoc_change.sac_assoc_id);
509 clear_bit(CF_CONNECT_PENDING, &con->flags);
512 make_sockaddr(&prim.ssp_addr, 0, &addr_len);
513 if (dlm_addr_to_nodeid(&prim.ssp_addr, &nodeid)) {
514 unsigned char *b=(unsigned char *)&prim.ssp_addr;
515 log_print("reject connect from unknown addr");
516 print_hex_dump_bytes("ss: ", DUMP_PREFIX_NONE,
517 b, sizeof(struct sockaddr_storage));
518 sctp_send_shutdown(prim.ssp_assoc_id);
522 new_con = nodeid2con(nodeid, GFP_NOFS);
526 /* Peel off a new sock */
527 sctp_lock_sock(con->sock->sk);
528 ret = sctp_do_peeloff(con->sock->sk,
529 sn->sn_assoc_change.sac_assoc_id,
531 sctp_release_sock(con->sock->sk);
533 log_print("Can't peel off a socket for "
534 "connection %d to node %d: err=%d",
535 (int)sn->sn_assoc_change.sac_assoc_id,
539 add_sock(new_con->sock, new_con);
541 log_print("connecting to %d sctp association %d",
542 nodeid, (int)sn->sn_assoc_change.sac_assoc_id);
544 /* Send any pending writes */
545 clear_bit(CF_CONNECT_PENDING, &new_con->flags);
546 clear_bit(CF_INIT_PENDING, &con->flags);
547 if (!test_and_set_bit(CF_WRITE_PENDING, &new_con->flags)) {
548 queue_work(send_workqueue, &new_con->swork);
550 if (!test_and_set_bit(CF_READ_PENDING, &new_con->flags))
551 queue_work(recv_workqueue, &new_con->rwork);
556 case SCTP_SHUTDOWN_COMP:
558 con = assoc2con(sn->sn_assoc_change.sac_assoc_id);
565 /* We don't know which INIT failed, so clear the PENDING flags
566 * on them all. if assoc_id is zero then it will then try
569 case SCTP_CANT_STR_ASSOC:
571 log_print("Can't start SCTP association - retrying");
577 log_print("unexpected SCTP assoc change id=%d state=%d",
578 (int)sn->sn_assoc_change.sac_assoc_id,
579 sn->sn_assoc_change.sac_state);
584 /* Data received from remote end */
585 static int receive_from_sock(struct connection *con)
588 struct msghdr msg = {};
592 int call_again_soon = 0;
594 char incmsg[CMSG_SPACE(sizeof(struct sctp_sndrcvinfo))];
596 mutex_lock(&con->sock_mutex);
598 if (con->sock == NULL) {
603 if (con->rx_page == NULL) {
605 * This doesn't need to be atomic, but I think it should
606 * improve performance if it is.
608 con->rx_page = alloc_page(GFP_ATOMIC);
609 if (con->rx_page == NULL)
611 cbuf_init(&con->cb, PAGE_CACHE_SIZE);
614 /* Only SCTP needs these really */
615 memset(&incmsg, 0, sizeof(incmsg));
616 msg.msg_control = incmsg;
617 msg.msg_controllen = sizeof(incmsg);
620 * iov[0] is the bit of the circular buffer between the current end
621 * point (cb.base + cb.len) and the end of the buffer.
623 iov[0].iov_len = con->cb.base - cbuf_data(&con->cb);
624 iov[0].iov_base = page_address(con->rx_page) + cbuf_data(&con->cb);
629 * iov[1] is the bit of the circular buffer between the start of the
630 * buffer and the start of the currently used section (cb.base)
632 if (cbuf_data(&con->cb) >= con->cb.base) {
633 iov[0].iov_len = PAGE_CACHE_SIZE - cbuf_data(&con->cb);
634 iov[1].iov_len = con->cb.base;
635 iov[1].iov_base = page_address(con->rx_page);
638 len = iov[0].iov_len + iov[1].iov_len;
640 r = ret = kernel_recvmsg(con->sock, &msg, iov, nvec, len,
641 MSG_DONTWAIT | MSG_NOSIGNAL);
645 /* Process SCTP notifications */
646 if (msg.msg_flags & MSG_NOTIFICATION) {
647 msg.msg_control = incmsg;
648 msg.msg_controllen = sizeof(incmsg);
650 process_sctp_notification(con, &msg,
651 page_address(con->rx_page) + con->cb.base);
652 mutex_unlock(&con->sock_mutex);
655 BUG_ON(con->nodeid == 0);
659 cbuf_add(&con->cb, ret);
660 ret = dlm_process_incoming_buffer(con->nodeid,
661 page_address(con->rx_page),
662 con->cb.base, con->cb.len,
664 if (ret == -EBADMSG) {
665 log_print("lowcomms: addr=%p, base=%u, len=%u, "
666 "iov_len=%u, iov_base[0]=%p, read=%d",
667 page_address(con->rx_page), con->cb.base, con->cb.len,
668 len, iov[0].iov_base, r);
672 cbuf_eat(&con->cb, ret);
674 if (cbuf_empty(&con->cb) && !call_again_soon) {
675 __free_page(con->rx_page);
681 mutex_unlock(&con->sock_mutex);
685 if (!test_and_set_bit(CF_READ_PENDING, &con->flags))
686 queue_work(recv_workqueue, &con->rwork);
687 mutex_unlock(&con->sock_mutex);
691 mutex_unlock(&con->sock_mutex);
692 if (ret != -EAGAIN) {
693 close_connection(con, false);
694 /* Reconnect when there is something to send */
696 /* Don't return success if we really got EOF */
703 /* Listening socket is busy, accept a connection */
704 static int tcp_accept_from_sock(struct connection *con)
707 struct sockaddr_storage peeraddr;
708 struct socket *newsock;
711 struct connection *newcon;
712 struct connection *addcon;
714 mutex_lock(&connections_lock);
715 if (!dlm_allow_conn) {
716 mutex_unlock(&connections_lock);
719 mutex_unlock(&connections_lock);
721 memset(&peeraddr, 0, sizeof(peeraddr));
722 result = sock_create_kern(dlm_local_addr[0]->ss_family, SOCK_STREAM,
723 IPPROTO_TCP, &newsock);
727 mutex_lock_nested(&con->sock_mutex, 0);
730 if (con->sock == NULL)
733 newsock->type = con->sock->type;
734 newsock->ops = con->sock->ops;
736 result = con->sock->ops->accept(con->sock, newsock, O_NONBLOCK);
740 /* Get the connected socket's peer */
741 memset(&peeraddr, 0, sizeof(peeraddr));
742 if (newsock->ops->getname(newsock, (struct sockaddr *)&peeraddr,
744 result = -ECONNABORTED;
748 /* Get the new node's NODEID */
749 make_sockaddr(&peeraddr, 0, &len);
750 if (dlm_addr_to_nodeid(&peeraddr, &nodeid)) {
751 unsigned char *b=(unsigned char *)&peeraddr;
752 log_print("connect from non cluster node");
753 print_hex_dump_bytes("ss: ", DUMP_PREFIX_NONE,
754 b, sizeof(struct sockaddr_storage));
755 sock_release(newsock);
756 mutex_unlock(&con->sock_mutex);
760 log_print("got connection from %d", nodeid);
762 /* Check to see if we already have a connection to this node. This
763 * could happen if the two nodes initiate a connection at roughly
764 * the same time and the connections cross on the wire.
765 * In this case we store the incoming one in "othercon"
767 newcon = nodeid2con(nodeid, GFP_NOFS);
772 mutex_lock_nested(&newcon->sock_mutex, 1);
774 struct connection *othercon = newcon->othercon;
777 othercon = kmem_cache_zalloc(con_cache, GFP_NOFS);
779 log_print("failed to allocate incoming socket");
780 mutex_unlock(&newcon->sock_mutex);
784 othercon->nodeid = nodeid;
785 othercon->rx_action = receive_from_sock;
786 mutex_init(&othercon->sock_mutex);
787 INIT_WORK(&othercon->swork, process_send_sockets);
788 INIT_WORK(&othercon->rwork, process_recv_sockets);
789 set_bit(CF_IS_OTHERCON, &othercon->flags);
791 if (!othercon->sock) {
792 newcon->othercon = othercon;
793 othercon->sock = newsock;
794 newsock->sk->sk_user_data = othercon;
795 add_sock(newsock, othercon);
799 printk("Extra connection from node %d attempted\n", nodeid);
801 mutex_unlock(&newcon->sock_mutex);
806 newsock->sk->sk_user_data = newcon;
807 newcon->rx_action = receive_from_sock;
808 add_sock(newsock, newcon);
812 mutex_unlock(&newcon->sock_mutex);
815 * Add it to the active queue in case we got data
816 * between processing the accept adding the socket
817 * to the read_sockets list
819 if (!test_and_set_bit(CF_READ_PENDING, &addcon->flags))
820 queue_work(recv_workqueue, &addcon->rwork);
821 mutex_unlock(&con->sock_mutex);
826 mutex_unlock(&con->sock_mutex);
827 sock_release(newsock);
829 if (result != -EAGAIN)
830 log_print("error accepting connection from node: %d", result);
834 static void free_entry(struct writequeue_entry *e)
836 __free_page(e->page);
840 /* Initiate an SCTP association.
841 This is a special case of send_to_sock() in that we don't yet have a
842 peeled-off socket for this association, so we use the listening socket
843 and add the primary IP address of the remote node.
845 static void sctp_init_assoc(struct connection *con)
847 struct sockaddr_storage rem_addr;
848 char outcmsg[CMSG_SPACE(sizeof(struct sctp_sndrcvinfo))];
849 struct msghdr outmessage;
850 struct cmsghdr *cmsg;
851 struct sctp_sndrcvinfo *sinfo;
852 struct connection *base_con;
853 struct writequeue_entry *e;
859 if (test_and_set_bit(CF_INIT_PENDING, &con->flags))
862 if (con->retries++ > MAX_CONNECT_RETRIES)
865 if (nodeid_to_addr(con->nodeid, (struct sockaddr *)&rem_addr)) {
866 log_print("no address for nodeid %d", con->nodeid);
869 base_con = nodeid2con(0, 0);
870 BUG_ON(base_con == NULL);
872 make_sockaddr(&rem_addr, dlm_config.ci_tcp_port, &addrlen);
874 outmessage.msg_name = &rem_addr;
875 outmessage.msg_namelen = addrlen;
876 outmessage.msg_control = outcmsg;
877 outmessage.msg_controllen = sizeof(outcmsg);
878 outmessage.msg_flags = MSG_EOR;
880 spin_lock(&con->writequeue_lock);
882 if (list_empty(&con->writequeue)) {
883 spin_unlock(&con->writequeue_lock);
884 log_print("writequeue empty for nodeid %d", con->nodeid);
888 e = list_first_entry(&con->writequeue, struct writequeue_entry, list);
891 spin_unlock(&con->writequeue_lock);
893 /* Send the first block off the write queue */
894 iov[0].iov_base = page_address(e->page)+offset;
895 iov[0].iov_len = len;
897 cmsg = CMSG_FIRSTHDR(&outmessage);
898 cmsg->cmsg_level = IPPROTO_SCTP;
899 cmsg->cmsg_type = SCTP_SNDRCV;
900 cmsg->cmsg_len = CMSG_LEN(sizeof(struct sctp_sndrcvinfo));
901 sinfo = CMSG_DATA(cmsg);
902 memset(sinfo, 0x00, sizeof(struct sctp_sndrcvinfo));
903 sinfo->sinfo_ppid = cpu_to_le32(dlm_our_nodeid());
904 outmessage.msg_controllen = cmsg->cmsg_len;
906 ret = kernel_sendmsg(base_con->sock, &outmessage, iov, 1, len);
908 log_print("Send first packet to node %d failed: %d",
911 /* Try again later */
912 clear_bit(CF_CONNECT_PENDING, &con->flags);
913 clear_bit(CF_INIT_PENDING, &con->flags);
916 spin_lock(&con->writequeue_lock);
920 if (e->len == 0 && e->users == 0) {
924 spin_unlock(&con->writequeue_lock);
928 /* Connect a new socket to its peer */
929 static void tcp_connect_to_sock(struct connection *con)
931 int result = -EHOSTUNREACH;
932 struct sockaddr_storage saddr, src_addr;
934 struct socket *sock = NULL;
937 if (con->nodeid == 0) {
938 log_print("attempt to connect sock 0 foiled");
942 mutex_lock(&con->sock_mutex);
943 if (con->retries++ > MAX_CONNECT_RETRIES)
946 /* Some odd races can cause double-connects, ignore them */
952 /* Create a socket to communicate with */
953 result = sock_create_kern(dlm_local_addr[0]->ss_family, SOCK_STREAM,
958 memset(&saddr, 0, sizeof(saddr));
959 if (dlm_nodeid_to_addr(con->nodeid, &saddr))
962 sock->sk->sk_user_data = con;
963 con->rx_action = receive_from_sock;
964 con->connect_action = tcp_connect_to_sock;
967 /* Bind to our cluster-known address connecting to avoid
969 memcpy(&src_addr, dlm_local_addr[0], sizeof(src_addr));
970 make_sockaddr(&src_addr, 0, &addr_len);
971 result = sock->ops->bind(sock, (struct sockaddr *) &src_addr,
974 log_print("could not bind for connect: %d", result);
975 /* This *may* not indicate a critical error */
978 make_sockaddr(&saddr, dlm_config.ci_tcp_port, &addr_len);
980 log_print("connecting to %d", con->nodeid);
982 /* Turn off Nagle's algorithm */
983 kernel_setsockopt(sock, SOL_TCP, TCP_NODELAY, (char *)&one,
987 sock->ops->connect(sock, (struct sockaddr *)&saddr, addr_len,
989 if (result == -EINPROGRESS)
996 sock_release(con->sock);
1002 * Some errors are fatal and this list might need adjusting. For other
1003 * errors we try again until the max number of retries is reached.
1005 if (result != -EHOSTUNREACH && result != -ENETUNREACH &&
1006 result != -ENETDOWN && result != -EINVAL
1007 && result != -EPROTONOSUPPORT) {
1008 lowcomms_connect_sock(con);
1012 mutex_unlock(&con->sock_mutex);
1016 static struct socket *tcp_create_listen_sock(struct connection *con,
1017 struct sockaddr_storage *saddr)
1019 struct socket *sock = NULL;
1024 if (dlm_local_addr[0]->ss_family == AF_INET)
1025 addr_len = sizeof(struct sockaddr_in);
1027 addr_len = sizeof(struct sockaddr_in6);
1029 /* Create a socket to communicate with */
1030 result = sock_create_kern(dlm_local_addr[0]->ss_family, SOCK_STREAM,
1031 IPPROTO_TCP, &sock);
1033 log_print("Can't create listening comms socket");
1037 /* Turn off Nagle's algorithm */
1038 kernel_setsockopt(sock, SOL_TCP, TCP_NODELAY, (char *)&one,
1041 result = kernel_setsockopt(sock, SOL_SOCKET, SO_REUSEADDR,
1042 (char *)&one, sizeof(one));
1045 log_print("Failed to set SO_REUSEADDR on socket: %d", result);
1047 sock->sk->sk_user_data = con;
1048 con->rx_action = tcp_accept_from_sock;
1049 con->connect_action = tcp_connect_to_sock;
1052 /* Bind to our port */
1053 make_sockaddr(saddr, dlm_config.ci_tcp_port, &addr_len);
1054 result = sock->ops->bind(sock, (struct sockaddr *) saddr, addr_len);
1056 log_print("Can't bind to port %d", dlm_config.ci_tcp_port);
1062 result = kernel_setsockopt(sock, SOL_SOCKET, SO_KEEPALIVE,
1063 (char *)&one, sizeof(one));
1065 log_print("Set keepalive failed: %d", result);
1068 result = sock->ops->listen(sock, 5);
1070 log_print("Can't listen on port %d", dlm_config.ci_tcp_port);
1080 /* Get local addresses */
1081 static void init_local(void)
1083 struct sockaddr_storage sas, *addr;
1086 dlm_local_count = 0;
1087 for (i = 0; i < DLM_MAX_ADDR_COUNT; i++) {
1088 if (dlm_our_addr(&sas, i))
1091 addr = kmalloc(sizeof(*addr), GFP_NOFS);
1094 memcpy(addr, &sas, sizeof(*addr));
1095 dlm_local_addr[dlm_local_count++] = addr;
1099 /* Bind to an IP address. SCTP allows multiple address so it can do
1101 static int add_sctp_bind_addr(struct connection *sctp_con,
1102 struct sockaddr_storage *addr,
1103 int addr_len, int num)
1108 result = kernel_bind(sctp_con->sock,
1109 (struct sockaddr *) addr,
1112 result = kernel_setsockopt(sctp_con->sock, SOL_SCTP,
1113 SCTP_SOCKOPT_BINDX_ADD,
1114 (char *)addr, addr_len);
1117 log_print("Can't bind to port %d addr number %d",
1118 dlm_config.ci_tcp_port, num);
1123 /* Initialise SCTP socket and bind to all interfaces */
1124 static int sctp_listen_for_all(void)
1126 struct socket *sock = NULL;
1127 struct sockaddr_storage localaddr;
1128 struct sctp_event_subscribe subscribe;
1129 int result = -EINVAL, num = 1, i, addr_len;
1130 struct connection *con = nodeid2con(0, GFP_NOFS);
1131 int bufsize = NEEDED_RMEM;
1136 log_print("Using SCTP for communications");
1138 result = sock_create_kern(dlm_local_addr[0]->ss_family, SOCK_SEQPACKET,
1139 IPPROTO_SCTP, &sock);
1141 log_print("Can't create comms socket, check SCTP is loaded");
1145 /* Listen for events */
1146 memset(&subscribe, 0, sizeof(subscribe));
1147 subscribe.sctp_data_io_event = 1;
1148 subscribe.sctp_association_event = 1;
1149 subscribe.sctp_send_failure_event = 1;
1150 subscribe.sctp_shutdown_event = 1;
1151 subscribe.sctp_partial_delivery_event = 1;
1153 result = kernel_setsockopt(sock, SOL_SOCKET, SO_RCVBUFFORCE,
1154 (char *)&bufsize, sizeof(bufsize));
1156 log_print("Error increasing buffer space on socket %d", result);
1158 result = kernel_setsockopt(sock, SOL_SCTP, SCTP_EVENTS,
1159 (char *)&subscribe, sizeof(subscribe));
1161 log_print("Failed to set SCTP_EVENTS on socket: result=%d",
1163 goto create_delsock;
1166 /* Init con struct */
1167 sock->sk->sk_user_data = con;
1169 con->sock->sk->sk_data_ready = lowcomms_data_ready;
1170 con->rx_action = receive_from_sock;
1171 con->connect_action = sctp_init_assoc;
1173 /* Bind to all interfaces. */
1174 for (i = 0; i < dlm_local_count; i++) {
1175 memcpy(&localaddr, dlm_local_addr[i], sizeof(localaddr));
1176 make_sockaddr(&localaddr, dlm_config.ci_tcp_port, &addr_len);
1178 result = add_sctp_bind_addr(con, &localaddr, addr_len, num);
1180 goto create_delsock;
1184 result = sock->ops->listen(sock, 5);
1186 log_print("Can't set socket listening");
1187 goto create_delsock;
1199 static int tcp_listen_for_all(void)
1201 struct socket *sock = NULL;
1202 struct connection *con = nodeid2con(0, GFP_NOFS);
1203 int result = -EINVAL;
1208 /* We don't support multi-homed hosts */
1209 if (dlm_local_addr[1] != NULL) {
1210 log_print("TCP protocol can't handle multi-homed hosts, "
1215 log_print("Using TCP for communications");
1217 sock = tcp_create_listen_sock(con, dlm_local_addr[0]);
1219 add_sock(sock, con);
1223 result = -EADDRINUSE;
1231 static struct writequeue_entry *new_writequeue_entry(struct connection *con,
1234 struct writequeue_entry *entry;
1236 entry = kmalloc(sizeof(struct writequeue_entry), allocation);
1240 entry->page = alloc_page(allocation);
1255 void *dlm_lowcomms_get_buffer(int nodeid, int len, gfp_t allocation, char **ppc)
1257 struct connection *con;
1258 struct writequeue_entry *e;
1262 con = nodeid2con(nodeid, allocation);
1266 spin_lock(&con->writequeue_lock);
1267 e = list_entry(con->writequeue.prev, struct writequeue_entry, list);
1268 if ((&e->list == &con->writequeue) ||
1269 (PAGE_CACHE_SIZE - e->end < len)) {
1276 spin_unlock(&con->writequeue_lock);
1280 *ppc = page_address(e->page) + offset;
1284 e = new_writequeue_entry(con, allocation);
1286 spin_lock(&con->writequeue_lock);
1290 list_add_tail(&e->list, &con->writequeue);
1291 spin_unlock(&con->writequeue_lock);
1297 void dlm_lowcomms_commit_buffer(void *mh)
1299 struct writequeue_entry *e = (struct writequeue_entry *)mh;
1300 struct connection *con = e->con;
1303 spin_lock(&con->writequeue_lock);
1307 e->len = e->end - e->offset;
1308 spin_unlock(&con->writequeue_lock);
1310 if (!test_and_set_bit(CF_WRITE_PENDING, &con->flags)) {
1311 queue_work(send_workqueue, &con->swork);
1316 spin_unlock(&con->writequeue_lock);
1320 /* Send a message */
1321 static void send_to_sock(struct connection *con)
1324 const int msg_flags = MSG_DONTWAIT | MSG_NOSIGNAL;
1325 struct writequeue_entry *e;
1329 mutex_lock(&con->sock_mutex);
1330 if (con->sock == NULL)
1333 spin_lock(&con->writequeue_lock);
1335 e = list_entry(con->writequeue.next, struct writequeue_entry,
1337 if ((struct list_head *) e == &con->writequeue)
1342 BUG_ON(len == 0 && e->users == 0);
1343 spin_unlock(&con->writequeue_lock);
1347 ret = kernel_sendpage(con->sock, e->page, offset, len,
1349 if (ret == -EAGAIN || ret == 0) {
1350 if (ret == -EAGAIN &&
1351 test_bit(SOCK_ASYNC_NOSPACE, &con->sock->flags) &&
1352 !test_and_set_bit(CF_APP_LIMITED, &con->flags)) {
1353 /* Notify TCP that we're limited by the
1354 * application window size.
1356 set_bit(SOCK_NOSPACE, &con->sock->flags);
1357 con->sock->sk->sk_write_pending++;
1366 /* Don't starve people filling buffers */
1367 if (++count >= MAX_SEND_MSG_COUNT) {
1372 spin_lock(&con->writequeue_lock);
1376 if (e->len == 0 && e->users == 0) {
1382 spin_unlock(&con->writequeue_lock);
1384 mutex_unlock(&con->sock_mutex);
1388 mutex_unlock(&con->sock_mutex);
1389 close_connection(con, false);
1390 lowcomms_connect_sock(con);
1394 mutex_unlock(&con->sock_mutex);
1395 if (!test_bit(CF_INIT_PENDING, &con->flags))
1396 lowcomms_connect_sock(con);
1400 static void clean_one_writequeue(struct connection *con)
1402 struct writequeue_entry *e, *safe;
1404 spin_lock(&con->writequeue_lock);
1405 list_for_each_entry_safe(e, safe, &con->writequeue, list) {
1409 spin_unlock(&con->writequeue_lock);
1412 /* Called from recovery when it knows that a node has
1414 int dlm_lowcomms_close(int nodeid)
1416 struct connection *con;
1418 log_print("closing connection to node %d", nodeid);
1419 con = nodeid2con(nodeid, 0);
1421 clear_bit(CF_CONNECT_PENDING, &con->flags);
1422 clear_bit(CF_WRITE_PENDING, &con->flags);
1423 set_bit(CF_CLOSE, &con->flags);
1424 if (cancel_work_sync(&con->swork))
1425 log_print("canceled swork for node %d", nodeid);
1426 if (cancel_work_sync(&con->rwork))
1427 log_print("canceled rwork for node %d", nodeid);
1428 clean_one_writequeue(con);
1429 close_connection(con, true);
1434 /* Receive workqueue function */
1435 static void process_recv_sockets(struct work_struct *work)
1437 struct connection *con = container_of(work, struct connection, rwork);
1440 clear_bit(CF_READ_PENDING, &con->flags);
1442 err = con->rx_action(con);
1446 /* Send workqueue function */
1447 static void process_send_sockets(struct work_struct *work)
1449 struct connection *con = container_of(work, struct connection, swork);
1451 if (test_and_clear_bit(CF_CONNECT_PENDING, &con->flags)) {
1452 con->connect_action(con);
1453 set_bit(CF_WRITE_PENDING, &con->flags);
1455 if (test_and_clear_bit(CF_WRITE_PENDING, &con->flags))
1460 /* Discard all entries on the write queues */
1461 static void clean_writequeues(void)
1463 foreach_conn(clean_one_writequeue);
1466 static void work_stop(void)
1468 destroy_workqueue(recv_workqueue);
1469 destroy_workqueue(send_workqueue);
1472 static int work_start(void)
1474 recv_workqueue = alloc_workqueue("dlm_recv",
1475 WQ_UNBOUND | WQ_MEM_RECLAIM, 1);
1476 if (!recv_workqueue) {
1477 log_print("can't start dlm_recv");
1481 send_workqueue = alloc_workqueue("dlm_send",
1482 WQ_UNBOUND | WQ_MEM_RECLAIM, 1);
1483 if (!send_workqueue) {
1484 log_print("can't start dlm_send");
1485 destroy_workqueue(recv_workqueue);
1492 static void stop_conn(struct connection *con)
1495 if (con->sock && con->sock->sk)
1496 con->sock->sk->sk_user_data = NULL;
1499 static void free_conn(struct connection *con)
1501 close_connection(con, true);
1503 kmem_cache_free(con_cache, con->othercon);
1504 hlist_del(&con->list);
1505 kmem_cache_free(con_cache, con);
1508 void dlm_lowcomms_stop(void)
1510 /* Set all the flags to prevent any
1513 mutex_lock(&connections_lock);
1515 foreach_conn(stop_conn);
1516 mutex_unlock(&connections_lock);
1520 mutex_lock(&connections_lock);
1521 clean_writequeues();
1523 foreach_conn(free_conn);
1525 mutex_unlock(&connections_lock);
1526 kmem_cache_destroy(con_cache);
1529 int dlm_lowcomms_start(void)
1531 int error = -EINVAL;
1532 struct connection *con;
1535 for (i = 0; i < CONN_HASH_SIZE; i++)
1536 INIT_HLIST_HEAD(&connection_hash[i]);
1539 if (!dlm_local_count) {
1541 log_print("no local IP address has been set");
1546 con_cache = kmem_cache_create("dlm_conn", sizeof(struct connection),
1547 __alignof__(struct connection), 0,
1552 error = work_start();
1558 /* Start listening */
1559 if (dlm_config.ci_protocol == 0)
1560 error = tcp_listen_for_all();
1562 error = sctp_listen_for_all();
1570 con = nodeid2con(0,0);
1572 close_connection(con, false);
1573 kmem_cache_free(con_cache, con);
1576 kmem_cache_destroy(con_cache);