blob: 609998de533e894d9cb98f3ff7ef3e05b41d424a [file] [log] [blame]
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001/******************************************************************************
2*******************************************************************************
3**
4** Copyright (C) Sistina Software, Inc. 1997-2003 All rights reserved.
Christine Caulfield5e9ccc32009-01-28 12:57:40 -06005** Copyright (C) 2004-2009 Red Hat, Inc. All rights reserved.
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01006**
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.
10**
11*******************************************************************************
12******************************************************************************/
13
14/*
15 * lowcomms.c
16 *
17 * This is the "low-level" comms layer.
18 *
19 * It is responsible for sending/receiving messages
20 * from other nodes in the cluster.
21 *
22 * Cluster nodes are referred to by their nodeids. nodeids are
23 * simply 32 bit numbers to the locking module - if they need to
Joe Perches2cf12c02009-01-22 13:26:47 -080024 * be expanded for the cluster infrastructure then that is its
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +010025 * responsibility. It is this layer's
26 * responsibility to resolve these into IP address or
27 * whatever it needs for inter-node communication.
28 *
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.
38 *
Joe Perches2cf12c02009-01-22 13:26:47 -080039 * lowcomms will choose to use either TCP or SCTP as its transport layer
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +010040 * depending on the configuration variable 'protocol'. This should be set
Joe Perches2cf12c02009-01-22 13:26:47 -080041 * to 0 (default) for TCP or 1 for SCTP. It should be configured using a
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +010042 * cluster-wide mechanism as it must be the same on all nodes of the cluster
43 * for the DLM to function.
44 *
45 */
46
47#include <asm/ioctls.h>
48#include <net/sock.h>
49#include <net/tcp.h>
50#include <linux/pagemap.h>
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +010051#include <linux/file.h>
Matthias Kaehlcke7a936ce2008-05-12 10:04:51 -050052#include <linux/mutex.h>
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +010053#include <linux/sctp.h>
Tejun Heo5a0e3ad2010-03-24 17:04:11 +090054#include <linux/slab.h>
Benjamin Poirier2f2d76c2012-03-08 05:55:59 +000055#include <net/sctp/sctp.h>
Joe Perches44ad5322009-01-22 13:24:49 -080056#include <net/ipv6.h>
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +010057
58#include "dlm_internal.h"
59#include "lowcomms.h"
60#include "midcomms.h"
61#include "config.h"
62
63#define NEEDED_RMEM (4*1024*1024)
Christine Caulfield5e9ccc32009-01-28 12:57:40 -060064#define CONN_HASH_SIZE 32
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +010065
Bob Petersonf92c8dd2010-11-12 11:15:20 -060066/* Number of messages to send before rescheduling */
67#define MAX_SEND_MSG_COUNT 25
68
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +010069struct cbuf {
70 unsigned int base;
71 unsigned int len;
72 unsigned int mask;
73};
74
75static void cbuf_add(struct cbuf *cb, int n)
76{
77 cb->len += n;
78}
79
80static int cbuf_data(struct cbuf *cb)
81{
82 return ((cb->base + cb->len) & cb->mask);
83}
84
85static void cbuf_init(struct cbuf *cb, int size)
86{
87 cb->base = cb->len = 0;
88 cb->mask = size-1;
89}
90
91static void cbuf_eat(struct cbuf *cb, int n)
92{
93 cb->len -= n;
94 cb->base += n;
95 cb->base &= cb->mask;
96}
97
98static bool cbuf_empty(struct cbuf *cb)
99{
100 return cb->len == 0;
101}
102
103struct connection {
104 struct socket *sock; /* NULL if not connected */
105 uint32_t nodeid; /* So we know who we are in the list */
106 struct mutex sock_mutex;
107 unsigned long flags;
108#define CF_READ_PENDING 1
109#define CF_WRITE_PENDING 2
110#define CF_CONNECT_PENDING 3
111#define CF_INIT_PENDING 4
112#define CF_IS_OTHERCON 5
Lars Marowsky-Bree063c4c92009-08-11 16:18:23 -0500113#define CF_CLOSE 6
David Millerb36930d2010-11-10 21:56:39 -0800114#define CF_APP_LIMITED 7
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100115 struct list_head writequeue; /* List of outgoing writequeue_entries */
116 spinlock_t writequeue_lock;
117 int (*rx_action) (struct connection *); /* What to do when active */
118 void (*connect_action) (struct connection *); /* What to do to connect */
119 struct page *rx_page;
120 struct cbuf cb;
121 int retries;
122#define MAX_CONNECT_RETRIES 3
Christine Caulfield5e9ccc32009-01-28 12:57:40 -0600123 struct hlist_node list;
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100124 struct connection *othercon;
125 struct work_struct rwork; /* Receive workqueue */
126 struct work_struct swork; /* Send workqueue */
Bob Petersonb81171c2016-02-05 14:39:02 -0500127 void (*orig_error_report)(struct sock *);
128 void (*orig_data_ready)(struct sock *);
129 void (*orig_state_change)(struct sock *);
130 void (*orig_write_space)(struct sock *);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100131};
132#define sock2con(x) ((struct connection *)(x)->sk_user_data)
133
134/* An entry waiting to be sent */
135struct writequeue_entry {
136 struct list_head list;
137 struct page *page;
138 int offset;
139 int len;
140 int end;
141 int users;
142 struct connection *con;
143};
144
David Teigland36b71a82012-07-26 12:44:30 -0500145struct dlm_node_addr {
146 struct list_head list;
147 int nodeid;
148 int addr_count;
Mike Christie98e1b602013-06-14 04:56:12 -0500149 int curr_addr_index;
David Teigland36b71a82012-07-26 12:44:30 -0500150 struct sockaddr_storage *addr[DLM_MAX_ADDR_COUNT];
151};
152
153static LIST_HEAD(dlm_node_addrs);
154static DEFINE_SPINLOCK(dlm_node_addrs_spin);
155
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100156static struct sockaddr_storage *dlm_local_addr[DLM_MAX_ADDR_COUNT];
157static int dlm_local_count;
David Teigland513ef592012-03-30 11:46:08 -0500158static int dlm_allow_conn;
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100159
160/* Work queues */
161static struct workqueue_struct *recv_workqueue;
162static struct workqueue_struct *send_workqueue;
163
Christine Caulfield5e9ccc32009-01-28 12:57:40 -0600164static struct hlist_head connection_hash[CONN_HASH_SIZE];
Matthias Kaehlcke7a936ce2008-05-12 10:04:51 -0500165static DEFINE_MUTEX(connections_lock);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100166static struct kmem_cache *con_cache;
167
168static void process_recv_sockets(struct work_struct *work);
169static void process_send_sockets(struct work_struct *work);
170
Christine Caulfield5e9ccc32009-01-28 12:57:40 -0600171
172/* This is deliberately very simple because most clusters have simple
173 sequential nodeids, so we should be able to go straight to a connection
174 struct in the array */
175static inline int nodeid_hash(int nodeid)
176{
177 return nodeid & (CONN_HASH_SIZE-1);
178}
179
180static struct connection *__find_con(int nodeid)
181{
182 int r;
Christine Caulfield5e9ccc32009-01-28 12:57:40 -0600183 struct connection *con;
184
185 r = nodeid_hash(nodeid);
186
Sasha Levinb67bfe02013-02-27 17:06:00 -0800187 hlist_for_each_entry(con, &connection_hash[r], list) {
Christine Caulfield5e9ccc32009-01-28 12:57:40 -0600188 if (con->nodeid == nodeid)
189 return con;
190 }
191 return NULL;
192}
193
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100194/*
195 * If 'allocation' is zero then we don't attempt to create a new
196 * connection structure for this node.
197 */
198static struct connection *__nodeid2con(int nodeid, gfp_t alloc)
199{
200 struct connection *con = NULL;
201 int r;
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100202
Christine Caulfield5e9ccc32009-01-28 12:57:40 -0600203 con = __find_con(nodeid);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100204 if (con || !alloc)
205 return con;
206
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100207 con = kmem_cache_zalloc(con_cache, alloc);
208 if (!con)
209 return NULL;
210
Christine Caulfield5e9ccc32009-01-28 12:57:40 -0600211 r = nodeid_hash(nodeid);
212 hlist_add_head(&con->list, &connection_hash[r]);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100213
214 con->nodeid = nodeid;
215 mutex_init(&con->sock_mutex);
216 INIT_LIST_HEAD(&con->writequeue);
217 spin_lock_init(&con->writequeue_lock);
218 INIT_WORK(&con->swork, process_send_sockets);
219 INIT_WORK(&con->rwork, process_recv_sockets);
220
221 /* Setup action pointers for child sockets */
222 if (con->nodeid) {
Christine Caulfield5e9ccc32009-01-28 12:57:40 -0600223 struct connection *zerocon = __find_con(0);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100224
225 con->connect_action = zerocon->connect_action;
226 if (!con->rx_action)
227 con->rx_action = zerocon->rx_action;
228 }
229
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100230 return con;
231}
232
Christine Caulfield5e9ccc32009-01-28 12:57:40 -0600233/* Loop round all connections */
234static void foreach_conn(void (*conn_func)(struct connection *c))
235{
236 int i;
Sasha Levinb67bfe02013-02-27 17:06:00 -0800237 struct hlist_node *n;
Christine Caulfield5e9ccc32009-01-28 12:57:40 -0600238 struct connection *con;
239
240 for (i = 0; i < CONN_HASH_SIZE; i++) {
Sasha Levinb67bfe02013-02-27 17:06:00 -0800241 hlist_for_each_entry_safe(con, n, &connection_hash[i], list)
Christine Caulfield5e9ccc32009-01-28 12:57:40 -0600242 conn_func(con);
Christine Caulfield5e9ccc32009-01-28 12:57:40 -0600243 }
244}
245
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100246static struct connection *nodeid2con(int nodeid, gfp_t allocation)
247{
248 struct connection *con;
249
Matthias Kaehlcke7a936ce2008-05-12 10:04:51 -0500250 mutex_lock(&connections_lock);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100251 con = __nodeid2con(nodeid, allocation);
Matthias Kaehlcke7a936ce2008-05-12 10:04:51 -0500252 mutex_unlock(&connections_lock);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100253
254 return con;
255}
256
David Teigland36b71a82012-07-26 12:44:30 -0500257static struct dlm_node_addr *find_node_addr(int nodeid)
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100258{
David Teigland36b71a82012-07-26 12:44:30 -0500259 struct dlm_node_addr *na;
260
261 list_for_each_entry(na, &dlm_node_addrs, list) {
262 if (na->nodeid == nodeid)
263 return na;
264 }
265 return NULL;
266}
267
268static int addr_compare(struct sockaddr_storage *x, struct sockaddr_storage *y)
269{
270 switch (x->ss_family) {
271 case AF_INET: {
272 struct sockaddr_in *sinx = (struct sockaddr_in *)x;
273 struct sockaddr_in *siny = (struct sockaddr_in *)y;
274 if (sinx->sin_addr.s_addr != siny->sin_addr.s_addr)
275 return 0;
276 if (sinx->sin_port != siny->sin_port)
277 return 0;
278 break;
279 }
280 case AF_INET6: {
281 struct sockaddr_in6 *sinx = (struct sockaddr_in6 *)x;
282 struct sockaddr_in6 *siny = (struct sockaddr_in6 *)y;
283 if (!ipv6_addr_equal(&sinx->sin6_addr, &siny->sin6_addr))
284 return 0;
285 if (sinx->sin6_port != siny->sin6_port)
286 return 0;
287 break;
288 }
289 default:
290 return 0;
291 }
292 return 1;
293}
294
295static int nodeid_to_addr(int nodeid, struct sockaddr_storage *sas_out,
Mike Christie98e1b602013-06-14 04:56:12 -0500296 struct sockaddr *sa_out, bool try_new_addr)
David Teigland36b71a82012-07-26 12:44:30 -0500297{
298 struct sockaddr_storage sas;
299 struct dlm_node_addr *na;
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100300
301 if (!dlm_local_count)
302 return -1;
303
David Teigland36b71a82012-07-26 12:44:30 -0500304 spin_lock(&dlm_node_addrs_spin);
305 na = find_node_addr(nodeid);
Mike Christie98e1b602013-06-14 04:56:12 -0500306 if (na && na->addr_count) {
Marcelo Ricardo Leitneree44b4b2015-08-11 19:22:23 -0300307 memcpy(&sas, na->addr[na->curr_addr_index],
308 sizeof(struct sockaddr_storage));
309
Mike Christie98e1b602013-06-14 04:56:12 -0500310 if (try_new_addr) {
311 na->curr_addr_index++;
312 if (na->curr_addr_index == na->addr_count)
313 na->curr_addr_index = 0;
314 }
Mike Christie98e1b602013-06-14 04:56:12 -0500315 }
David Teigland36b71a82012-07-26 12:44:30 -0500316 spin_unlock(&dlm_node_addrs_spin);
317
318 if (!na)
319 return -EEXIST;
320
321 if (!na->addr_count)
322 return -ENOENT;
323
324 if (sas_out)
325 memcpy(sas_out, &sas, sizeof(struct sockaddr_storage));
326
327 if (!sa_out)
328 return 0;
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100329
330 if (dlm_local_addr[0]->ss_family == AF_INET) {
David Teigland36b71a82012-07-26 12:44:30 -0500331 struct sockaddr_in *in4 = (struct sockaddr_in *) &sas;
332 struct sockaddr_in *ret4 = (struct sockaddr_in *) sa_out;
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100333 ret4->sin_addr.s_addr = in4->sin_addr.s_addr;
334 } else {
David Teigland36b71a82012-07-26 12:44:30 -0500335 struct sockaddr_in6 *in6 = (struct sockaddr_in6 *) &sas;
336 struct sockaddr_in6 *ret6 = (struct sockaddr_in6 *) sa_out;
Alexey Dobriyan4e3fd7a2011-11-21 03:39:03 +0000337 ret6->sin6_addr = in6->sin6_addr;
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100338 }
339
340 return 0;
341}
342
David Teigland36b71a82012-07-26 12:44:30 -0500343static int addr_to_nodeid(struct sockaddr_storage *addr, int *nodeid)
344{
345 struct dlm_node_addr *na;
346 int rv = -EEXIST;
Mike Christie98e1b602013-06-14 04:56:12 -0500347 int addr_i;
David Teigland36b71a82012-07-26 12:44:30 -0500348
349 spin_lock(&dlm_node_addrs_spin);
350 list_for_each_entry(na, &dlm_node_addrs, list) {
351 if (!na->addr_count)
352 continue;
353
Mike Christie98e1b602013-06-14 04:56:12 -0500354 for (addr_i = 0; addr_i < na->addr_count; addr_i++) {
355 if (addr_compare(na->addr[addr_i], addr)) {
356 *nodeid = na->nodeid;
357 rv = 0;
358 goto unlock;
359 }
360 }
David Teigland36b71a82012-07-26 12:44:30 -0500361 }
Mike Christie98e1b602013-06-14 04:56:12 -0500362unlock:
David Teigland36b71a82012-07-26 12:44:30 -0500363 spin_unlock(&dlm_node_addrs_spin);
364 return rv;
365}
366
367int dlm_lowcomms_addr(int nodeid, struct sockaddr_storage *addr, int len)
368{
369 struct sockaddr_storage *new_addr;
370 struct dlm_node_addr *new_node, *na;
371
372 new_node = kzalloc(sizeof(struct dlm_node_addr), GFP_NOFS);
373 if (!new_node)
374 return -ENOMEM;
375
376 new_addr = kzalloc(sizeof(struct sockaddr_storage), GFP_NOFS);
377 if (!new_addr) {
378 kfree(new_node);
379 return -ENOMEM;
380 }
381
382 memcpy(new_addr, addr, len);
383
384 spin_lock(&dlm_node_addrs_spin);
385 na = find_node_addr(nodeid);
386 if (!na) {
387 new_node->nodeid = nodeid;
388 new_node->addr[0] = new_addr;
389 new_node->addr_count = 1;
390 list_add(&new_node->list, &dlm_node_addrs);
391 spin_unlock(&dlm_node_addrs_spin);
392 return 0;
393 }
394
395 if (na->addr_count >= DLM_MAX_ADDR_COUNT) {
396 spin_unlock(&dlm_node_addrs_spin);
397 kfree(new_addr);
398 kfree(new_node);
399 return -ENOSPC;
400 }
401
402 na->addr[na->addr_count++] = new_addr;
403 spin_unlock(&dlm_node_addrs_spin);
404 kfree(new_node);
405 return 0;
406}
407
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100408/* Data available on socket or listen socket received a connect */
David S. Miller676d2362014-04-11 16:15:36 -0400409static void lowcomms_data_ready(struct sock *sk)
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100410{
411 struct connection *con = sock2con(sk);
Patrick Caulfieldafb853f2007-06-01 10:07:26 -0500412 if (con && !test_and_set_bit(CF_READ_PENDING, &con->flags))
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100413 queue_work(recv_workqueue, &con->rwork);
414}
415
416static void lowcomms_write_space(struct sock *sk)
417{
418 struct connection *con = sock2con(sk);
419
David Millerb36930d2010-11-10 21:56:39 -0800420 if (!con)
421 return;
422
423 clear_bit(SOCK_NOSPACE, &con->sock->flags);
424
425 if (test_and_clear_bit(CF_APP_LIMITED, &con->flags)) {
426 con->sock->sk->sk_write_pending--;
Eric Dumazet9cd3e072015-11-29 20:03:10 -0800427 clear_bit(SOCKWQ_ASYNC_NOSPACE, &con->sock->flags);
David Millerb36930d2010-11-10 21:56:39 -0800428 }
429
430 if (!test_and_set_bit(CF_WRITE_PENDING, &con->flags))
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100431 queue_work(send_workqueue, &con->swork);
432}
433
434static inline void lowcomms_connect_sock(struct connection *con)
435{
Lars Marowsky-Bree063c4c92009-08-11 16:18:23 -0500436 if (test_bit(CF_CLOSE, &con->flags))
437 return;
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100438 if (!test_and_set_bit(CF_CONNECT_PENDING, &con->flags))
439 queue_work(send_workqueue, &con->swork);
440}
441
442static void lowcomms_state_change(struct sock *sk)
443{
Marcelo Ricardo Leitneree44b4b2015-08-11 19:22:23 -0300444 /* SCTP layer is not calling sk_data_ready when the connection
445 * is done, so we catch the signal through here. Also, it
446 * doesn't switch socket state when entering shutdown, so we
447 * skip the write in that case.
448 */
449 if (sk->sk_shutdown) {
450 if (sk->sk_shutdown == RCV_SHUTDOWN)
451 lowcomms_data_ready(sk);
452 } else if (sk->sk_state == TCP_ESTABLISHED) {
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100453 lowcomms_write_space(sk);
Marcelo Ricardo Leitneree44b4b2015-08-11 19:22:23 -0300454 }
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100455}
456
Christine Caulfield391fbdc2009-05-07 10:54:16 -0500457int dlm_lowcomms_connect_node(int nodeid)
458{
459 struct connection *con;
460
461 if (nodeid == dlm_our_nodeid())
462 return 0;
463
464 con = nodeid2con(nodeid, GFP_NOFS);
465 if (!con)
466 return -ENOMEM;
467 lowcomms_connect_sock(con);
468 return 0;
469}
470
Bob Petersonb3a5bbf2015-08-27 09:34:47 -0500471static void lowcomms_error_report(struct sock *sk)
472{
Bob Petersonb81171c2016-02-05 14:39:02 -0500473 struct connection *con;
Bob Petersonb3a5bbf2015-08-27 09:34:47 -0500474 struct sockaddr_storage saddr;
Bob Peterson1a318332016-01-18 12:29:15 -0500475 int buflen;
Bob Petersonb81171c2016-02-05 14:39:02 -0500476 void (*orig_report)(struct sock *) = NULL;
Bob Petersonb3a5bbf2015-08-27 09:34:47 -0500477
Bob Petersonb81171c2016-02-05 14:39:02 -0500478 read_lock_bh(&sk->sk_callback_lock);
479 con = sock2con(sk);
480 if (con == NULL)
481 goto out;
482
483 orig_report = con->orig_error_report;
Bob Peterson1a318332016-01-18 12:29:15 -0500484 if (con->sock == NULL ||
485 kernel_getpeername(con->sock, (struct sockaddr *)&saddr, &buflen)) {
Bob Petersonb3a5bbf2015-08-27 09:34:47 -0500486 printk_ratelimited(KERN_ERR "dlm: node %d: socket error "
487 "sending to node %d, port %d, "
488 "sk_err=%d/%d\n", dlm_our_nodeid(),
489 con->nodeid, dlm_config.ci_tcp_port,
490 sk->sk_err, sk->sk_err_soft);
Bob Petersonb3a5bbf2015-08-27 09:34:47 -0500491 } else if (saddr.ss_family == AF_INET) {
492 struct sockaddr_in *sin4 = (struct sockaddr_in *)&saddr;
493
494 printk_ratelimited(KERN_ERR "dlm: node %d: socket error "
495 "sending to node %d at %pI4, port %d, "
496 "sk_err=%d/%d\n", dlm_our_nodeid(),
497 con->nodeid, &sin4->sin_addr.s_addr,
498 dlm_config.ci_tcp_port, sk->sk_err,
499 sk->sk_err_soft);
500 } else {
501 struct sockaddr_in6 *sin6 = (struct sockaddr_in6 *)&saddr;
502
503 printk_ratelimited(KERN_ERR "dlm: node %d: socket error "
504 "sending to node %d at %u.%u.%u.%u, "
505 "port %d, sk_err=%d/%d\n", dlm_our_nodeid(),
506 con->nodeid, sin6->sin6_addr.s6_addr32[0],
507 sin6->sin6_addr.s6_addr32[1],
508 sin6->sin6_addr.s6_addr32[2],
509 sin6->sin6_addr.s6_addr32[3],
510 dlm_config.ci_tcp_port, sk->sk_err,
511 sk->sk_err_soft);
512 }
Bob Petersonb81171c2016-02-05 14:39:02 -0500513out:
514 read_unlock_bh(&sk->sk_callback_lock);
515 if (orig_report)
516 orig_report(sk);
517}
518
519/* Note: sk_callback_lock must be locked before calling this function. */
520static void save_callbacks(struct connection *con, struct sock *sk)
521{
522 lock_sock(sk);
523 con->orig_data_ready = sk->sk_data_ready;
524 con->orig_state_change = sk->sk_state_change;
525 con->orig_write_space = sk->sk_write_space;
526 con->orig_error_report = sk->sk_error_report;
527 release_sock(sk);
528}
529
530static void restore_callbacks(struct connection *con, struct sock *sk)
531{
532 write_lock_bh(&sk->sk_callback_lock);
533 lock_sock(sk);
534 sk->sk_user_data = NULL;
535 sk->sk_data_ready = con->orig_data_ready;
536 sk->sk_state_change = con->orig_state_change;
537 sk->sk_write_space = con->orig_write_space;
538 sk->sk_error_report = con->orig_error_report;
539 release_sock(sk);
540 write_unlock_bh(&sk->sk_callback_lock);
Bob Petersonb3a5bbf2015-08-27 09:34:47 -0500541}
542
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100543/* Make a socket active */
Ying Xue4dd40f02012-08-10 14:58:42 +0800544static void add_sock(struct socket *sock, struct connection *con)
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100545{
Bob Petersonb81171c2016-02-05 14:39:02 -0500546 struct sock *sk = sock->sk;
547
548 write_lock_bh(&sk->sk_callback_lock);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100549 con->sock = sock;
550
Bob Petersonb81171c2016-02-05 14:39:02 -0500551 sk->sk_user_data = con;
552 if (!test_bit(CF_IS_OTHERCON, &con->flags))
553 save_callbacks(con, sk);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100554 /* Install a data_ready callback */
Bob Petersonb81171c2016-02-05 14:39:02 -0500555 sk->sk_data_ready = lowcomms_data_ready;
556 sk->sk_write_space = lowcomms_write_space;
557 sk->sk_state_change = lowcomms_state_change;
558 sk->sk_allocation = GFP_NOFS;
559 sk->sk_error_report = lowcomms_error_report;
560 write_unlock_bh(&sk->sk_callback_lock);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100561}
562
563/* Add the port number to an IPv6 or 4 sockaddr and return the address
564 length */
565static void make_sockaddr(struct sockaddr_storage *saddr, uint16_t port,
566 int *addr_len)
567{
568 saddr->ss_family = dlm_local_addr[0]->ss_family;
569 if (saddr->ss_family == AF_INET) {
570 struct sockaddr_in *in4_addr = (struct sockaddr_in *)saddr;
571 in4_addr->sin_port = cpu_to_be16(port);
572 *addr_len = sizeof(struct sockaddr_in);
573 memset(&in4_addr->sin_zero, 0, sizeof(in4_addr->sin_zero));
574 } else {
575 struct sockaddr_in6 *in6_addr = (struct sockaddr_in6 *)saddr;
576 in6_addr->sin6_port = cpu_to_be16(port);
577 *addr_len = sizeof(struct sockaddr_in6);
578 }
Patrick Caulfield01c8cab2007-07-17 16:53:15 +0100579 memset((char *)saddr + *addr_len, 0, sizeof(struct sockaddr_storage) - *addr_len);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100580}
581
582/* Close a remote connection and tidy up */
Marcelo Ricardo Leitner0d737a82015-08-11 19:22:21 -0300583static void close_connection(struct connection *con, bool and_other,
584 bool tx, bool rx)
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100585{
Marcelo Ricardo Leitner0d737a82015-08-11 19:22:21 -0300586 clear_bit(CF_CONNECT_PENDING, &con->flags);
587 clear_bit(CF_WRITE_PENDING, &con->flags);
588 if (tx && cancel_work_sync(&con->swork))
589 log_print("canceled swork for node %d", con->nodeid);
590 if (rx && cancel_work_sync(&con->rwork))
591 log_print("canceled rwork for node %d", con->nodeid);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100592
Marcelo Ricardo Leitner0d737a82015-08-11 19:22:21 -0300593 mutex_lock(&con->sock_mutex);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100594 if (con->sock) {
Bob Petersonb81171c2016-02-05 14:39:02 -0500595 if (!test_bit(CF_IS_OTHERCON, &con->flags))
596 restore_callbacks(con, con->sock->sk);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100597 sock_release(con->sock);
598 con->sock = NULL;
599 }
600 if (con->othercon && and_other) {
601 /* Will only re-enter once. */
Marcelo Ricardo Leitner0d737a82015-08-11 19:22:21 -0300602 close_connection(con->othercon, false, true, true);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100603 }
604 if (con->rx_page) {
605 __free_page(con->rx_page);
606 con->rx_page = NULL;
607 }
Patrick Caulfield9e5f2822007-08-02 14:58:14 +0100608
Patrick Caulfield61d96be2007-08-20 15:13:38 +0100609 con->retries = 0;
610 mutex_unlock(&con->sock_mutex);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100611}
612
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100613/* Data received from remote end */
614static int receive_from_sock(struct connection *con)
615{
616 int ret = 0;
617 struct msghdr msg = {};
618 struct kvec iov[2];
619 unsigned len;
620 int r;
621 int call_again_soon = 0;
622 int nvec;
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100623
624 mutex_lock(&con->sock_mutex);
625
626 if (con->sock == NULL) {
627 ret = -EAGAIN;
628 goto out_close;
629 }
Marcelo Ricardo Leitneracee4e52015-08-11 19:22:24 -0300630 if (con->nodeid == 0) {
631 ret = -EINVAL;
632 goto out_close;
633 }
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100634
635 if (con->rx_page == NULL) {
636 /*
637 * This doesn't need to be atomic, but I think it should
638 * improve performance if it is.
639 */
640 con->rx_page = alloc_page(GFP_ATOMIC);
641 if (con->rx_page == NULL)
642 goto out_resched;
Kirill A. Shutemov09cbfea2016-04-01 15:29:47 +0300643 cbuf_init(&con->cb, PAGE_SIZE);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100644 }
645
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100646 /*
647 * iov[0] is the bit of the circular buffer between the current end
648 * point (cb.base + cb.len) and the end of the buffer.
649 */
650 iov[0].iov_len = con->cb.base - cbuf_data(&con->cb);
651 iov[0].iov_base = page_address(con->rx_page) + cbuf_data(&con->cb);
652 iov[1].iov_len = 0;
653 nvec = 1;
654
655 /*
656 * iov[1] is the bit of the circular buffer between the start of the
657 * buffer and the start of the currently used section (cb.base)
658 */
659 if (cbuf_data(&con->cb) >= con->cb.base) {
Kirill A. Shutemov09cbfea2016-04-01 15:29:47 +0300660 iov[0].iov_len = PAGE_SIZE - cbuf_data(&con->cb);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100661 iov[1].iov_len = con->cb.base;
662 iov[1].iov_base = page_address(con->rx_page);
663 nvec = 2;
664 }
665 len = iov[0].iov_len + iov[1].iov_len;
666
667 r = ret = kernel_recvmsg(con->sock, &msg, iov, nvec, len,
668 MSG_DONTWAIT | MSG_NOSIGNAL);
669 if (ret <= 0)
670 goto out_close;
Marcelo Ricardo Leitneree44b4b2015-08-11 19:22:23 -0300671 else if (ret == len)
672 call_again_soon = 1;
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100673
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100674 cbuf_add(&con->cb, ret);
675 ret = dlm_process_incoming_buffer(con->nodeid,
676 page_address(con->rx_page),
677 con->cb.base, con->cb.len,
Kirill A. Shutemov09cbfea2016-04-01 15:29:47 +0300678 PAGE_SIZE);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100679 if (ret == -EBADMSG) {
Marcelo Ricardo Leitneree44b4b2015-08-11 19:22:23 -0300680 log_print("lowcomms: addr=%p, base=%u, len=%u, read=%d",
681 page_address(con->rx_page), con->cb.base,
682 con->cb.len, r);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100683 }
684 if (ret < 0)
685 goto out_close;
686 cbuf_eat(&con->cb, ret);
687
688 if (cbuf_empty(&con->cb) && !call_again_soon) {
689 __free_page(con->rx_page);
690 con->rx_page = NULL;
691 }
692
693 if (call_again_soon)
694 goto out_resched;
695 mutex_unlock(&con->sock_mutex);
696 return 0;
697
698out_resched:
699 if (!test_and_set_bit(CF_READ_PENDING, &con->flags))
700 queue_work(recv_workqueue, &con->rwork);
701 mutex_unlock(&con->sock_mutex);
702 return -EAGAIN;
703
704out_close:
705 mutex_unlock(&con->sock_mutex);
Patrick Caulfield9e5f2822007-08-02 14:58:14 +0100706 if (ret != -EAGAIN) {
Marcelo Ricardo Leitner0d737a82015-08-11 19:22:21 -0300707 close_connection(con, false, true, false);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100708 /* Reconnect when there is something to send */
709 }
710 /* Don't return success if we really got EOF */
711 if (ret == 0)
712 ret = -EAGAIN;
713
714 return ret;
715}
716
717/* Listening socket is busy, accept a connection */
718static int tcp_accept_from_sock(struct connection *con)
719{
720 int result;
721 struct sockaddr_storage peeraddr;
722 struct socket *newsock;
723 int len;
724 int nodeid;
725 struct connection *newcon;
726 struct connection *addcon;
727
David Teigland513ef592012-03-30 11:46:08 -0500728 mutex_lock(&connections_lock);
729 if (!dlm_allow_conn) {
730 mutex_unlock(&connections_lock);
731 return -1;
732 }
733 mutex_unlock(&connections_lock);
734
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100735 memset(&peeraddr, 0, sizeof(peeraddr));
Eric W. Biedermaneeb1bd52015-05-08 21:08:05 -0500736 result = sock_create_kern(&init_net, dlm_local_addr[0]->ss_family,
737 SOCK_STREAM, IPPROTO_TCP, &newsock);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100738 if (result < 0)
739 return -ENOMEM;
740
741 mutex_lock_nested(&con->sock_mutex, 0);
742
743 result = -ENOTCONN;
744 if (con->sock == NULL)
745 goto accept_err;
746
747 newsock->type = con->sock->type;
748 newsock->ops = con->sock->ops;
749
750 result = con->sock->ops->accept(con->sock, newsock, O_NONBLOCK);
751 if (result < 0)
752 goto accept_err;
753
754 /* Get the connected socket's peer */
755 memset(&peeraddr, 0, sizeof(peeraddr));
756 if (newsock->ops->getname(newsock, (struct sockaddr *)&peeraddr,
757 &len, 2)) {
758 result = -ECONNABORTED;
759 goto accept_err;
760 }
761
762 /* Get the new node's NODEID */
763 make_sockaddr(&peeraddr, 0, &len);
David Teigland36b71a82012-07-26 12:44:30 -0500764 if (addr_to_nodeid(&peeraddr, &nodeid)) {
Masatake YAMATObcaadf52011-07-04 12:25:51 +0900765 unsigned char *b=(unsigned char *)&peeraddr;
David Teigland617e82e2007-04-26 13:46:49 -0500766 log_print("connect from non cluster node");
Masatake YAMATObcaadf52011-07-04 12:25:51 +0900767 print_hex_dump_bytes("ss: ", DUMP_PREFIX_NONE,
768 b, sizeof(struct sockaddr_storage));
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100769 sock_release(newsock);
770 mutex_unlock(&con->sock_mutex);
771 return -1;
772 }
773
774 log_print("got connection from %d", nodeid);
775
776 /* Check to see if we already have a connection to this node. This
777 * could happen if the two nodes initiate a connection at roughly
778 * the same time and the connections cross on the wire.
779 * In this case we store the incoming one in "othercon"
780 */
David Teigland748285c2009-05-15 10:50:57 -0500781 newcon = nodeid2con(nodeid, GFP_NOFS);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100782 if (!newcon) {
783 result = -ENOMEM;
784 goto accept_err;
785 }
786 mutex_lock_nested(&newcon->sock_mutex, 1);
787 if (newcon->sock) {
788 struct connection *othercon = newcon->othercon;
789
790 if (!othercon) {
David Teigland748285c2009-05-15 10:50:57 -0500791 othercon = kmem_cache_zalloc(con_cache, GFP_NOFS);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100792 if (!othercon) {
David Teigland617e82e2007-04-26 13:46:49 -0500793 log_print("failed to allocate incoming socket");
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100794 mutex_unlock(&newcon->sock_mutex);
795 result = -ENOMEM;
796 goto accept_err;
797 }
798 othercon->nodeid = nodeid;
799 othercon->rx_action = receive_from_sock;
800 mutex_init(&othercon->sock_mutex);
801 INIT_WORK(&othercon->swork, process_send_sockets);
802 INIT_WORK(&othercon->rwork, process_recv_sockets);
803 set_bit(CF_IS_OTHERCON, &othercon->flags);
Patrick Caulfield61d96be2007-08-20 15:13:38 +0100804 }
805 if (!othercon->sock) {
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100806 newcon->othercon = othercon;
Patrick Caulfield97d84832007-06-27 11:36:23 +0100807 othercon->sock = newsock;
808 newsock->sk->sk_user_data = othercon;
809 add_sock(newsock, othercon);
810 addcon = othercon;
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100811 }
Patrick Caulfield97d84832007-06-27 11:36:23 +0100812 else {
813 printk("Extra connection from node %d attempted\n", nodeid);
814 result = -EAGAIN;
akpm@linux-foundation.orgf4fadb23c2007-06-27 14:43:37 -0700815 mutex_unlock(&newcon->sock_mutex);
Patrick Caulfield97d84832007-06-27 11:36:23 +0100816 goto accept_err;
817 }
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100818 }
819 else {
820 newsock->sk->sk_user_data = newcon;
821 newcon->rx_action = receive_from_sock;
822 add_sock(newsock, newcon);
823 addcon = newcon;
824 }
825
826 mutex_unlock(&newcon->sock_mutex);
827
828 /*
829 * Add it to the active queue in case we got data
Lucas De Marchi25985ed2011-03-30 22:57:33 -0300830 * between processing the accept adding the socket
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100831 * to the read_sockets list
832 */
833 if (!test_and_set_bit(CF_READ_PENDING, &addcon->flags))
834 queue_work(recv_workqueue, &addcon->rwork);
835 mutex_unlock(&con->sock_mutex);
836
837 return 0;
838
839accept_err:
840 mutex_unlock(&con->sock_mutex);
841 sock_release(newsock);
842
843 if (result != -EAGAIN)
David Teigland617e82e2007-04-26 13:46:49 -0500844 log_print("error accepting connection from node: %d", result);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100845 return result;
846}
847
kbuild test robot18df8a82015-08-18 04:50:36 +0800848static int sctp_accept_from_sock(struct connection *con)
Marcelo Ricardo Leitneree44b4b2015-08-11 19:22:23 -0300849{
850 /* Check that the new node is in the lockspace */
851 struct sctp_prim prim;
852 int nodeid;
853 int prim_len, ret;
854 int addr_len;
855 struct connection *newcon;
856 struct connection *addcon;
857 struct socket *newsock;
858
859 mutex_lock(&connections_lock);
860 if (!dlm_allow_conn) {
861 mutex_unlock(&connections_lock);
862 return -1;
863 }
864 mutex_unlock(&connections_lock);
865
866 mutex_lock_nested(&con->sock_mutex, 0);
867
868 ret = kernel_accept(con->sock, &newsock, O_NONBLOCK);
869 if (ret < 0)
870 goto accept_err;
871
872 memset(&prim, 0, sizeof(struct sctp_prim));
873 prim_len = sizeof(struct sctp_prim);
874
875 ret = kernel_getsockopt(newsock, IPPROTO_SCTP, SCTP_PRIMARY_ADDR,
876 (char *)&prim, &prim_len);
877 if (ret < 0) {
878 log_print("getsockopt/sctp_primary_addr failed: %d", ret);
879 goto accept_err;
880 }
881
882 make_sockaddr(&prim.ssp_addr, 0, &addr_len);
883 if (addr_to_nodeid(&prim.ssp_addr, &nodeid)) {
884 unsigned char *b = (unsigned char *)&prim.ssp_addr;
885
886 log_print("reject connect from unknown addr");
887 print_hex_dump_bytes("ss: ", DUMP_PREFIX_NONE,
888 b, sizeof(struct sockaddr_storage));
889 goto accept_err;
890 }
891
892 newcon = nodeid2con(nodeid, GFP_NOFS);
893 if (!newcon) {
894 ret = -ENOMEM;
895 goto accept_err;
896 }
897
898 mutex_lock_nested(&newcon->sock_mutex, 1);
899
900 if (newcon->sock) {
901 struct connection *othercon = newcon->othercon;
902
903 if (!othercon) {
904 othercon = kmem_cache_zalloc(con_cache, GFP_NOFS);
905 if (!othercon) {
906 log_print("failed to allocate incoming socket");
907 mutex_unlock(&newcon->sock_mutex);
908 ret = -ENOMEM;
909 goto accept_err;
910 }
911 othercon->nodeid = nodeid;
912 othercon->rx_action = receive_from_sock;
913 mutex_init(&othercon->sock_mutex);
914 INIT_WORK(&othercon->swork, process_send_sockets);
915 INIT_WORK(&othercon->rwork, process_recv_sockets);
916 set_bit(CF_IS_OTHERCON, &othercon->flags);
917 }
918 if (!othercon->sock) {
919 newcon->othercon = othercon;
920 othercon->sock = newsock;
921 newsock->sk->sk_user_data = othercon;
922 add_sock(newsock, othercon);
923 addcon = othercon;
924 } else {
925 printk("Extra connection from node %d attempted\n", nodeid);
926 ret = -EAGAIN;
927 mutex_unlock(&newcon->sock_mutex);
928 goto accept_err;
929 }
930 } else {
931 newsock->sk->sk_user_data = newcon;
932 newcon->rx_action = receive_from_sock;
933 add_sock(newsock, newcon);
934 addcon = newcon;
935 }
936
937 log_print("connected to %d", nodeid);
938
939 mutex_unlock(&newcon->sock_mutex);
940
941 /*
942 * Add it to the active queue in case we got data
943 * between processing the accept adding the socket
944 * to the read_sockets list
945 */
946 if (!test_and_set_bit(CF_READ_PENDING, &addcon->flags))
947 queue_work(recv_workqueue, &addcon->rwork);
948 mutex_unlock(&con->sock_mutex);
949
950 return 0;
951
952accept_err:
953 mutex_unlock(&con->sock_mutex);
954 if (newsock)
955 sock_release(newsock);
956 if (ret != -EAGAIN)
957 log_print("error accepting connection from node: %d", ret);
958
959 return ret;
960}
961
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +0100962static void free_entry(struct writequeue_entry *e)
963{
964 __free_page(e->page);
965 kfree(e);
966}
967
Mike Christie5d689872013-06-14 04:56:13 -0500968/*
969 * writequeue_entry_complete - try to delete and free write queue entry
970 * @e: write queue entry to try to delete
971 * @completed: bytes completed
972 *
973 * writequeue_lock must be held.
974 */
975static void writequeue_entry_complete(struct writequeue_entry *e, int completed)
976{
977 e->offset += completed;
978 e->len -= completed;
979
980 if (e->len == 0 && e->users == 0) {
981 list_del(&e->list);
982 free_entry(e);
983 }
984}
985
Marcelo Ricardo Leitneree44b4b2015-08-11 19:22:23 -0300986/*
987 * sctp_bind_addrs - bind a SCTP socket to all our addresses
988 */
989static int sctp_bind_addrs(struct connection *con, uint16_t port)
990{
991 struct sockaddr_storage localaddr;
992 int i, addr_len, result = 0;
993
994 for (i = 0; i < dlm_local_count; i++) {
995 memcpy(&localaddr, dlm_local_addr[i], sizeof(localaddr));
996 make_sockaddr(&localaddr, port, &addr_len);
997
998 if (!i)
999 result = kernel_bind(con->sock,
1000 (struct sockaddr *)&localaddr,
1001 addr_len);
1002 else
1003 result = kernel_setsockopt(con->sock, SOL_SCTP,
1004 SCTP_SOCKOPT_BINDX_ADD,
1005 (char *)&localaddr, addr_len);
1006
1007 if (result < 0) {
1008 log_print("Can't bind to %d addr number %d, %d.\n",
1009 port, i + 1, result);
1010 break;
1011 }
1012 }
1013 return result;
1014}
1015
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001016/* Initiate an SCTP association.
1017 This is a special case of send_to_sock() in that we don't yet have a
1018 peeled-off socket for this association, so we use the listening socket
1019 and add the primary IP address of the remote node.
1020 */
Marcelo Ricardo Leitneree44b4b2015-08-11 19:22:23 -03001021static void sctp_connect_to_sock(struct connection *con)
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001022{
Marcelo Ricardo Leitneree44b4b2015-08-11 19:22:23 -03001023 struct sockaddr_storage daddr;
1024 int one = 1;
1025 int result;
1026 int addr_len;
1027 struct socket *sock;
1028
1029 if (con->nodeid == 0) {
1030 log_print("attempt to connect sock 0 foiled");
1031 return;
1032 }
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001033
Mike Christie5d689872013-06-14 04:56:13 -05001034 mutex_lock(&con->sock_mutex);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001035
Marcelo Ricardo Leitneree44b4b2015-08-11 19:22:23 -03001036 /* Some odd races can cause double-connects, ignore them */
1037 if (con->retries++ > MAX_CONNECT_RETRIES)
1038 goto out;
1039
1040 if (con->sock) {
1041 log_print("node %d already connected.", con->nodeid);
1042 goto out;
1043 }
1044
1045 memset(&daddr, 0, sizeof(daddr));
1046 result = nodeid_to_addr(con->nodeid, &daddr, NULL, true);
1047 if (result < 0) {
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001048 log_print("no address for nodeid %d", con->nodeid);
Marcelo Ricardo Leitneree44b4b2015-08-11 19:22:23 -03001049 goto out;
David Teigland04bedd72009-09-18 14:31:47 -05001050 }
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001051
Marcelo Ricardo Leitneree44b4b2015-08-11 19:22:23 -03001052 /* Create a socket to communicate with */
1053 result = sock_create_kern(&init_net, dlm_local_addr[0]->ss_family,
1054 SOCK_STREAM, IPPROTO_SCTP, &sock);
1055 if (result < 0)
1056 goto socket_err;
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001057
Marcelo Ricardo Leitneree44b4b2015-08-11 19:22:23 -03001058 sock->sk->sk_user_data = con;
1059 con->rx_action = receive_from_sock;
1060 con->connect_action = sctp_connect_to_sock;
1061 add_sock(sock, con);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001062
Marcelo Ricardo Leitneree44b4b2015-08-11 19:22:23 -03001063 /* Bind to all addresses. */
1064 if (sctp_bind_addrs(con, 0))
1065 goto bind_err;
Mike Christie98e1b602013-06-14 04:56:12 -05001066
Marcelo Ricardo Leitneree44b4b2015-08-11 19:22:23 -03001067 make_sockaddr(&daddr, dlm_config.ci_tcp_port, &addr_len);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001068
Marcelo Ricardo Leitneree44b4b2015-08-11 19:22:23 -03001069 log_print("connecting to %d", con->nodeid);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001070
Marcelo Ricardo Leitneree44b4b2015-08-11 19:22:23 -03001071 /* Turn off Nagle's algorithm */
1072 kernel_setsockopt(sock, SOL_TCP, TCP_NODELAY, (char *)&one,
1073 sizeof(one));
1074
1075 result = sock->ops->connect(sock, (struct sockaddr *)&daddr, addr_len,
1076 O_NONBLOCK);
1077 if (result == -EINPROGRESS)
1078 result = 0;
1079 if (result == 0)
1080 goto out;
1081
1082
1083bind_err:
1084 con->sock = NULL;
1085 sock_release(sock);
1086
1087socket_err:
1088 /*
1089 * Some errors are fatal and this list might need adjusting. For other
1090 * errors we try again until the max number of retries is reached.
1091 */
1092 if (result != -EHOSTUNREACH &&
1093 result != -ENETUNREACH &&
1094 result != -ENETDOWN &&
1095 result != -EINVAL &&
1096 result != -EPROTONOSUPPORT) {
1097 log_print("connect %d try %d error %d", con->nodeid,
1098 con->retries, result);
1099 mutex_unlock(&con->sock_mutex);
1100 msleep(1000);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001101 clear_bit(CF_CONNECT_PENDING, &con->flags);
Marcelo Ricardo Leitneree44b4b2015-08-11 19:22:23 -03001102 lowcomms_connect_sock(con);
1103 return;
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001104 }
Mike Christie5d689872013-06-14 04:56:13 -05001105
Marcelo Ricardo Leitneree44b4b2015-08-11 19:22:23 -03001106out:
Mike Christie5d689872013-06-14 04:56:13 -05001107 mutex_unlock(&con->sock_mutex);
Marcelo Ricardo Leitner00dcffa2015-08-11 19:22:25 -03001108 set_bit(CF_WRITE_PENDING, &con->flags);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001109}
1110
1111/* Connect a new socket to its peer */
1112static void tcp_connect_to_sock(struct connection *con)
1113{
Lon Hohberger6bd8fed2007-10-25 18:51:54 -04001114 struct sockaddr_storage saddr, src_addr;
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001115 int addr_len;
Casey Dahlina89d63a2009-07-14 12:17:51 -05001116 struct socket *sock = NULL;
David Teiglandcb2d45d2010-11-12 11:12:55 -06001117 int one = 1;
David Teigland36b71a82012-07-26 12:44:30 -05001118 int result;
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001119
1120 if (con->nodeid == 0) {
1121 log_print("attempt to connect sock 0 foiled");
1122 return;
1123 }
1124
1125 mutex_lock(&con->sock_mutex);
1126 if (con->retries++ > MAX_CONNECT_RETRIES)
1127 goto out;
1128
1129 /* Some odd races can cause double-connects, ignore them */
David Teigland36b71a82012-07-26 12:44:30 -05001130 if (con->sock)
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001131 goto out;
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001132
1133 /* Create a socket to communicate with */
Eric W. Biedermaneeb1bd52015-05-08 21:08:05 -05001134 result = sock_create_kern(&init_net, dlm_local_addr[0]->ss_family,
1135 SOCK_STREAM, IPPROTO_TCP, &sock);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001136 if (result < 0)
1137 goto out_err;
1138
1139 memset(&saddr, 0, sizeof(saddr));
Mike Christie98e1b602013-06-14 04:56:12 -05001140 result = nodeid_to_addr(con->nodeid, &saddr, NULL, false);
David Teigland36b71a82012-07-26 12:44:30 -05001141 if (result < 0) {
1142 log_print("no address for nodeid %d", con->nodeid);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001143 goto out_err;
David Teigland36b71a82012-07-26 12:44:30 -05001144 }
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001145
1146 sock->sk->sk_user_data = con;
1147 con->rx_action = receive_from_sock;
1148 con->connect_action = tcp_connect_to_sock;
1149 add_sock(sock, con);
1150
Lon Hohberger6bd8fed2007-10-25 18:51:54 -04001151 /* Bind to our cluster-known address connecting to avoid
1152 routing problems */
1153 memcpy(&src_addr, dlm_local_addr[0], sizeof(src_addr));
1154 make_sockaddr(&src_addr, 0, &addr_len);
1155 result = sock->ops->bind(sock, (struct sockaddr *) &src_addr,
1156 addr_len);
1157 if (result < 0) {
1158 log_print("could not bind for connect: %d", result);
1159 /* This *may* not indicate a critical error */
1160 }
1161
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001162 make_sockaddr(&saddr, dlm_config.ci_tcp_port, &addr_len);
1163
1164 log_print("connecting to %d", con->nodeid);
David Teiglandcb2d45d2010-11-12 11:12:55 -06001165
1166 /* Turn off Nagle's algorithm */
1167 kernel_setsockopt(sock, SOL_TCP, TCP_NODELAY, (char *)&one,
1168 sizeof(one));
1169
David Teigland36b71a82012-07-26 12:44:30 -05001170 result = sock->ops->connect(sock, (struct sockaddr *)&saddr, addr_len,
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001171 O_NONBLOCK);
1172 if (result == -EINPROGRESS)
1173 result = 0;
1174 if (result == 0)
1175 goto out;
1176
1177out_err:
1178 if (con->sock) {
1179 sock_release(con->sock);
1180 con->sock = NULL;
Casey Dahlina89d63a2009-07-14 12:17:51 -05001181 } else if (sock) {
1182 sock_release(sock);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001183 }
1184 /*
1185 * Some errors are fatal and this list might need adjusting. For other
1186 * errors we try again until the max number of retries is reached.
1187 */
David Teigland36b71a82012-07-26 12:44:30 -05001188 if (result != -EHOSTUNREACH &&
1189 result != -ENETUNREACH &&
1190 result != -ENETDOWN &&
1191 result != -EINVAL &&
1192 result != -EPROTONOSUPPORT) {
1193 log_print("connect %d try %d error %d", con->nodeid,
1194 con->retries, result);
1195 mutex_unlock(&con->sock_mutex);
1196 msleep(1000);
Marcelo Ricardo Leitner356344c2015-08-11 19:22:22 -03001197 clear_bit(CF_CONNECT_PENDING, &con->flags);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001198 lowcomms_connect_sock(con);
David Teigland36b71a82012-07-26 12:44:30 -05001199 return;
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001200 }
1201out:
1202 mutex_unlock(&con->sock_mutex);
Marcelo Ricardo Leitner00dcffa2015-08-11 19:22:25 -03001203 set_bit(CF_WRITE_PENDING, &con->flags);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001204 return;
1205}
1206
1207static struct socket *tcp_create_listen_sock(struct connection *con,
1208 struct sockaddr_storage *saddr)
1209{
1210 struct socket *sock = NULL;
1211 int result = 0;
1212 int one = 1;
1213 int addr_len;
1214
1215 if (dlm_local_addr[0]->ss_family == AF_INET)
1216 addr_len = sizeof(struct sockaddr_in);
1217 else
1218 addr_len = sizeof(struct sockaddr_in6);
1219
1220 /* Create a socket to communicate with */
Eric W. Biedermaneeb1bd52015-05-08 21:08:05 -05001221 result = sock_create_kern(&init_net, dlm_local_addr[0]->ss_family,
1222 SOCK_STREAM, IPPROTO_TCP, &sock);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001223 if (result < 0) {
David Teigland617e82e2007-04-26 13:46:49 -05001224 log_print("Can't create listening comms socket");
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001225 goto create_out;
1226 }
1227
David Teiglandcb2d45d2010-11-12 11:12:55 -06001228 /* Turn off Nagle's algorithm */
1229 kernel_setsockopt(sock, SOL_TCP, TCP_NODELAY, (char *)&one,
1230 sizeof(one));
1231
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001232 result = kernel_setsockopt(sock, SOL_SOCKET, SO_REUSEADDR,
1233 (char *)&one, sizeof(one));
1234
1235 if (result < 0) {
David Teigland617e82e2007-04-26 13:46:49 -05001236 log_print("Failed to set SO_REUSEADDR on socket: %d", result);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001237 }
Bob Petersonb81171c2016-02-05 14:39:02 -05001238 sock->sk->sk_user_data = con;
1239
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001240 con->rx_action = tcp_accept_from_sock;
1241 con->connect_action = tcp_connect_to_sock;
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001242
1243 /* Bind to our port */
1244 make_sockaddr(saddr, dlm_config.ci_tcp_port, &addr_len);
1245 result = sock->ops->bind(sock, (struct sockaddr *) saddr, addr_len);
1246 if (result < 0) {
David Teigland617e82e2007-04-26 13:46:49 -05001247 log_print("Can't bind to port %d", dlm_config.ci_tcp_port);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001248 sock_release(sock);
1249 sock = NULL;
1250 con->sock = NULL;
1251 goto create_out;
1252 }
1253 result = kernel_setsockopt(sock, SOL_SOCKET, SO_KEEPALIVE,
1254 (char *)&one, sizeof(one));
1255 if (result < 0) {
David Teigland617e82e2007-04-26 13:46:49 -05001256 log_print("Set keepalive failed: %d", result);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001257 }
1258
1259 result = sock->ops->listen(sock, 5);
1260 if (result < 0) {
David Teigland617e82e2007-04-26 13:46:49 -05001261 log_print("Can't listen on port %d", dlm_config.ci_tcp_port);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001262 sock_release(sock);
1263 sock = NULL;
1264 goto create_out;
1265 }
1266
1267create_out:
1268 return sock;
1269}
1270
1271/* Get local addresses */
1272static void init_local(void)
1273{
1274 struct sockaddr_storage sas, *addr;
1275 int i;
1276
Patrick Caulfield30d3a232007-04-23 16:26:21 +01001277 dlm_local_count = 0;
David Teigland1b189b82012-03-21 09:18:34 -05001278 for (i = 0; i < DLM_MAX_ADDR_COUNT; i++) {
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001279 if (dlm_our_addr(&sas, i))
1280 break;
1281
Amitoj Kaur Chawla5c93f562016-06-23 10:22:01 +05301282 addr = kmemdup(&sas, sizeof(*addr), GFP_NOFS);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001283 if (!addr)
1284 break;
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001285 dlm_local_addr[dlm_local_count++] = addr;
1286 }
1287}
1288
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001289/* Initialise SCTP socket and bind to all interfaces */
1290static int sctp_listen_for_all(void)
1291{
1292 struct socket *sock = NULL;
Marcelo Ricardo Leitneree44b4b2015-08-11 19:22:23 -03001293 int result = -EINVAL;
David Teigland573c24c2009-11-30 16:34:43 -06001294 struct connection *con = nodeid2con(0, GFP_NOFS);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001295 int bufsize = NEEDED_RMEM;
Mike Christie86e92ad2013-06-14 04:56:14 -05001296 int one = 1;
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001297
1298 if (!con)
1299 return -ENOMEM;
1300
1301 log_print("Using SCTP for communications");
1302
Eric W. Biedermaneeb1bd52015-05-08 21:08:05 -05001303 result = sock_create_kern(&init_net, dlm_local_addr[0]->ss_family,
Marcelo Ricardo Leitneree44b4b2015-08-11 19:22:23 -03001304 SOCK_STREAM, IPPROTO_SCTP, &sock);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001305 if (result < 0) {
1306 log_print("Can't create comms socket, check SCTP is loaded");
1307 goto out;
1308 }
1309
David S. Millerdf61c952007-11-06 23:48:57 -08001310 result = kernel_setsockopt(sock, SOL_SOCKET, SO_RCVBUFFORCE,
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001311 (char *)&bufsize, sizeof(bufsize));
1312 if (result)
David Teigland617e82e2007-04-26 13:46:49 -05001313 log_print("Error increasing buffer space on socket %d", result);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001314
Mike Christie86e92ad2013-06-14 04:56:14 -05001315 result = kernel_setsockopt(sock, SOL_SCTP, SCTP_NODELAY, (char *)&one,
1316 sizeof(one));
1317 if (result < 0)
1318 log_print("Could not set SCTP NODELAY error %d\n", result);
1319
Bob Petersonb81171c2016-02-05 14:39:02 -05001320 write_lock_bh(&sock->sk->sk_callback_lock);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001321 /* Init con struct */
1322 sock->sk->sk_user_data = con;
1323 con->sock = sock;
1324 con->sock->sk->sk_data_ready = lowcomms_data_ready;
Marcelo Ricardo Leitneree44b4b2015-08-11 19:22:23 -03001325 con->rx_action = sctp_accept_from_sock;
1326 con->connect_action = sctp_connect_to_sock;
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001327
Bob Petersonb81171c2016-02-05 14:39:02 -05001328 write_unlock_bh(&sock->sk->sk_callback_lock);
1329
Marcelo Ricardo Leitneree44b4b2015-08-11 19:22:23 -03001330 /* Bind to all addresses. */
1331 if (sctp_bind_addrs(con, dlm_config.ci_tcp_port))
1332 goto create_delsock;
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001333
1334 result = sock->ops->listen(sock, 5);
1335 if (result < 0) {
1336 log_print("Can't set socket listening");
1337 goto create_delsock;
1338 }
1339
1340 return 0;
1341
1342create_delsock:
1343 sock_release(sock);
1344 con->sock = NULL;
1345out:
1346 return result;
1347}
1348
1349static int tcp_listen_for_all(void)
1350{
1351 struct socket *sock = NULL;
David Teigland573c24c2009-11-30 16:34:43 -06001352 struct connection *con = nodeid2con(0, GFP_NOFS);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001353 int result = -EINVAL;
1354
1355 if (!con)
1356 return -ENOMEM;
1357
1358 /* We don't support multi-homed hosts */
1359 if (dlm_local_addr[1] != NULL) {
David Teigland617e82e2007-04-26 13:46:49 -05001360 log_print("TCP protocol can't handle multi-homed hosts, "
1361 "try SCTP");
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001362 return -EINVAL;
1363 }
1364
1365 log_print("Using TCP for communications");
1366
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001367 sock = tcp_create_listen_sock(con, dlm_local_addr[0]);
1368 if (sock) {
1369 add_sock(sock, con);
1370 result = 0;
1371 }
1372 else {
1373 result = -EADDRINUSE;
1374 }
1375
1376 return result;
1377}
1378
1379
1380
1381static struct writequeue_entry *new_writequeue_entry(struct connection *con,
1382 gfp_t allocation)
1383{
1384 struct writequeue_entry *entry;
1385
1386 entry = kmalloc(sizeof(struct writequeue_entry), allocation);
1387 if (!entry)
1388 return NULL;
1389
1390 entry->page = alloc_page(allocation);
1391 if (!entry->page) {
1392 kfree(entry);
1393 return NULL;
1394 }
1395
1396 entry->offset = 0;
1397 entry->len = 0;
1398 entry->end = 0;
1399 entry->users = 0;
1400 entry->con = con;
1401
1402 return entry;
1403}
1404
David Teigland617e82e2007-04-26 13:46:49 -05001405void *dlm_lowcomms_get_buffer(int nodeid, int len, gfp_t allocation, char **ppc)
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001406{
1407 struct connection *con;
1408 struct writequeue_entry *e;
1409 int offset = 0;
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001410
1411 con = nodeid2con(nodeid, allocation);
1412 if (!con)
1413 return NULL;
1414
1415 spin_lock(&con->writequeue_lock);
1416 e = list_entry(con->writequeue.prev, struct writequeue_entry, list);
1417 if ((&e->list == &con->writequeue) ||
Kirill A. Shutemov09cbfea2016-04-01 15:29:47 +03001418 (PAGE_SIZE - e->end < len)) {
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001419 e = NULL;
1420 } else {
1421 offset = e->end;
1422 e->end += len;
Wei Yongjuneeee2b52012-10-18 22:57:19 +08001423 e->users++;
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001424 }
1425 spin_unlock(&con->writequeue_lock);
1426
1427 if (e) {
1428 got_one:
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001429 *ppc = page_address(e->page) + offset;
1430 return e;
1431 }
1432
1433 e = new_writequeue_entry(con, allocation);
1434 if (e) {
1435 spin_lock(&con->writequeue_lock);
1436 offset = e->end;
1437 e->end += len;
Wei Yongjuneeee2b52012-10-18 22:57:19 +08001438 e->users++;
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001439 list_add_tail(&e->list, &con->writequeue);
1440 spin_unlock(&con->writequeue_lock);
1441 goto got_one;
1442 }
1443 return NULL;
1444}
1445
1446void dlm_lowcomms_commit_buffer(void *mh)
1447{
1448 struct writequeue_entry *e = (struct writequeue_entry *)mh;
1449 struct connection *con = e->con;
1450 int users;
1451
1452 spin_lock(&con->writequeue_lock);
1453 users = --e->users;
1454 if (users)
1455 goto out;
1456 e->len = e->end - e->offset;
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001457 spin_unlock(&con->writequeue_lock);
1458
1459 if (!test_and_set_bit(CF_WRITE_PENDING, &con->flags)) {
1460 queue_work(send_workqueue, &con->swork);
1461 }
1462 return;
1463
1464out:
1465 spin_unlock(&con->writequeue_lock);
1466 return;
1467}
1468
1469/* Send a message */
1470static void send_to_sock(struct connection *con)
1471{
1472 int ret = 0;
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001473 const int msg_flags = MSG_DONTWAIT | MSG_NOSIGNAL;
1474 struct writequeue_entry *e;
1475 int len, offset;
Bob Petersonf92c8dd2010-11-12 11:15:20 -06001476 int count = 0;
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001477
1478 mutex_lock(&con->sock_mutex);
1479 if (con->sock == NULL)
1480 goto out_connect;
1481
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001482 spin_lock(&con->writequeue_lock);
1483 for (;;) {
1484 e = list_entry(con->writequeue.next, struct writequeue_entry,
1485 list);
1486 if ((struct list_head *) e == &con->writequeue)
1487 break;
1488
1489 len = e->len;
1490 offset = e->offset;
1491 BUG_ON(len == 0 && e->users == 0);
1492 spin_unlock(&con->writequeue_lock);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001493
1494 ret = 0;
1495 if (len) {
Paolo Bonzini1329e3f2009-08-24 13:18:04 -05001496 ret = kernel_sendpage(con->sock, e->page, offset, len,
1497 msg_flags);
Patrick Caulfieldd66f8272007-09-14 08:49:21 +01001498 if (ret == -EAGAIN || ret == 0) {
David Millerb36930d2010-11-10 21:56:39 -08001499 if (ret == -EAGAIN &&
Eric Dumazet9cd3e072015-11-29 20:03:10 -08001500 test_bit(SOCKWQ_ASYNC_NOSPACE, &con->sock->flags) &&
David Millerb36930d2010-11-10 21:56:39 -08001501 !test_and_set_bit(CF_APP_LIMITED, &con->flags)) {
1502 /* Notify TCP that we're limited by the
1503 * application window size.
1504 */
1505 set_bit(SOCK_NOSPACE, &con->sock->flags);
1506 con->sock->sk->sk_write_pending++;
1507 }
Patrick Caulfieldd66f8272007-09-14 08:49:21 +01001508 cond_resched();
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001509 goto out;
Ying Xue9c5bef52012-08-13 14:29:55 +08001510 } else if (ret < 0)
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001511 goto send_error;
Patrick Caulfieldd66f8272007-09-14 08:49:21 +01001512 }
Bob Petersonf92c8dd2010-11-12 11:15:20 -06001513
1514 /* Don't starve people filling buffers */
1515 if (++count >= MAX_SEND_MSG_COUNT) {
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001516 cond_resched();
Bob Petersonf92c8dd2010-11-12 11:15:20 -06001517 count = 0;
1518 }
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001519
1520 spin_lock(&con->writequeue_lock);
Mike Christie5d689872013-06-14 04:56:13 -05001521 writequeue_entry_complete(e, ret);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001522 }
1523 spin_unlock(&con->writequeue_lock);
1524out:
1525 mutex_unlock(&con->sock_mutex);
1526 return;
1527
1528send_error:
1529 mutex_unlock(&con->sock_mutex);
Marcelo Ricardo Leitner0d737a82015-08-11 19:22:21 -03001530 close_connection(con, false, false, true);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001531 lowcomms_connect_sock(con);
1532 return;
1533
1534out_connect:
1535 mutex_unlock(&con->sock_mutex);
Marcelo Ricardo Leitneree44b4b2015-08-11 19:22:23 -03001536 lowcomms_connect_sock(con);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001537}
1538
1539static void clean_one_writequeue(struct connection *con)
1540{
Christine Caulfield5e9ccc32009-01-28 12:57:40 -06001541 struct writequeue_entry *e, *safe;
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001542
1543 spin_lock(&con->writequeue_lock);
Christine Caulfield5e9ccc32009-01-28 12:57:40 -06001544 list_for_each_entry_safe(e, safe, &con->writequeue, list) {
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001545 list_del(&e->list);
1546 free_entry(e);
1547 }
1548 spin_unlock(&con->writequeue_lock);
1549}
1550
1551/* Called from recovery when it knows that a node has
1552 left the cluster */
1553int dlm_lowcomms_close(int nodeid)
1554{
1555 struct connection *con;
David Teigland36b71a82012-07-26 12:44:30 -05001556 struct dlm_node_addr *na;
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001557
1558 log_print("closing connection to node %d", nodeid);
1559 con = nodeid2con(nodeid, 0);
1560 if (con) {
Lars Marowsky-Bree063c4c92009-08-11 16:18:23 -05001561 set_bit(CF_CLOSE, &con->flags);
Marcelo Ricardo Leitner0d737a82015-08-11 19:22:21 -03001562 close_connection(con, true, true, true);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001563 clean_one_writequeue(con);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001564 }
David Teigland36b71a82012-07-26 12:44:30 -05001565
1566 spin_lock(&dlm_node_addrs_spin);
1567 na = find_node_addr(nodeid);
1568 if (na) {
1569 list_del(&na->list);
1570 while (na->addr_count--)
1571 kfree(na->addr[na->addr_count]);
1572 kfree(na);
1573 }
1574 spin_unlock(&dlm_node_addrs_spin);
1575
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001576 return 0;
1577}
1578
1579/* Receive workqueue function */
1580static void process_recv_sockets(struct work_struct *work)
1581{
1582 struct connection *con = container_of(work, struct connection, rwork);
1583 int err;
1584
1585 clear_bit(CF_READ_PENDING, &con->flags);
1586 do {
1587 err = con->rx_action(con);
1588 } while (!err);
1589}
1590
1591/* Send workqueue function */
1592static void process_send_sockets(struct work_struct *work)
1593{
1594 struct connection *con = container_of(work, struct connection, swork);
1595
Marcelo Ricardo Leitner00dcffa2015-08-11 19:22:25 -03001596 if (test_and_clear_bit(CF_CONNECT_PENDING, &con->flags))
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001597 con->connect_action(con);
Lars Marowsky-Bree063c4c92009-08-11 16:18:23 -05001598 if (test_and_clear_bit(CF_WRITE_PENDING, &con->flags))
1599 send_to_sock(con);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001600}
1601
1602
1603/* Discard all entries on the write queues */
1604static void clean_writequeues(void)
1605{
Christine Caulfield5e9ccc32009-01-28 12:57:40 -06001606 foreach_conn(clean_one_writequeue);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001607}
1608
1609static void work_stop(void)
1610{
1611 destroy_workqueue(recv_workqueue);
1612 destroy_workqueue(send_workqueue);
1613}
1614
1615static int work_start(void)
1616{
David Teiglande43f0552011-03-10 13:22:34 -06001617 recv_workqueue = alloc_workqueue("dlm_recv",
1618 WQ_UNBOUND | WQ_MEM_RECLAIM, 1);
Namhyung Kimb9d41052010-12-13 13:42:24 -06001619 if (!recv_workqueue) {
1620 log_print("can't start dlm_recv");
1621 return -ENOMEM;
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001622 }
1623
David Teiglande43f0552011-03-10 13:22:34 -06001624 send_workqueue = alloc_workqueue("dlm_send",
1625 WQ_UNBOUND | WQ_MEM_RECLAIM, 1);
Namhyung Kimb9d41052010-12-13 13:42:24 -06001626 if (!send_workqueue) {
1627 log_print("can't start dlm_send");
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001628 destroy_workqueue(recv_workqueue);
Namhyung Kimb9d41052010-12-13 13:42:24 -06001629 return -ENOMEM;
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001630 }
1631
1632 return 0;
1633}
1634
Christine Caulfield5e9ccc32009-01-28 12:57:40 -06001635static void stop_conn(struct connection *con)
1636{
1637 con->flags |= 0x0F;
Christine Caulfield391fbdc2009-05-07 10:54:16 -05001638 if (con->sock && con->sock->sk)
Christine Caulfield5e9ccc32009-01-28 12:57:40 -06001639 con->sock->sk->sk_user_data = NULL;
1640}
1641
1642static void free_conn(struct connection *con)
1643{
Marcelo Ricardo Leitner0d737a82015-08-11 19:22:21 -03001644 close_connection(con, true, true, true);
Christine Caulfield5e9ccc32009-01-28 12:57:40 -06001645 if (con->othercon)
1646 kmem_cache_free(con_cache, con->othercon);
1647 hlist_del(&con->list);
1648 kmem_cache_free(con_cache, con);
1649}
1650
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001651void dlm_lowcomms_stop(void)
1652{
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001653 /* Set all the flags to prevent any
1654 socket activity.
1655 */
Matthias Kaehlcke7a936ce2008-05-12 10:04:51 -05001656 mutex_lock(&connections_lock);
David Teigland513ef592012-03-30 11:46:08 -05001657 dlm_allow_conn = 0;
Christine Caulfield5e9ccc32009-01-28 12:57:40 -06001658 foreach_conn(stop_conn);
Marcelo Ricardo Leitner3a8db792016-10-08 10:14:37 -03001659 clean_writequeues();
1660 foreach_conn(free_conn);
Matthias Kaehlcke7a936ce2008-05-12 10:04:51 -05001661 mutex_unlock(&connections_lock);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001662
1663 work_stop();
1664
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001665 kmem_cache_destroy(con_cache);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001666}
1667
1668int dlm_lowcomms_start(void)
1669{
1670 int error = -EINVAL;
1671 struct connection *con;
Christine Caulfield5e9ccc32009-01-28 12:57:40 -06001672 int i;
1673
1674 for (i = 0; i < CONN_HASH_SIZE; i++)
1675 INIT_HLIST_HEAD(&connection_hash[i]);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001676
1677 init_local();
1678 if (!dlm_local_count) {
David Teigland617e82e2007-04-26 13:46:49 -05001679 error = -ENOTCONN;
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001680 log_print("no local IP address has been set");
David Teigland513ef592012-03-30 11:46:08 -05001681 goto fail;
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001682 }
1683
1684 error = -ENOMEM;
1685 con_cache = kmem_cache_create("dlm_conn", sizeof(struct connection),
1686 __alignof__(struct connection), 0,
Paul Mundt20c2df82007-07-20 10:11:58 +09001687 NULL);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001688 if (!con_cache)
David Teigland513ef592012-03-30 11:46:08 -05001689 goto fail;
1690
1691 error = work_start();
1692 if (error)
1693 goto fail_destroy;
1694
1695 dlm_allow_conn = 1;
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001696
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001697 /* Start listening */
1698 if (dlm_config.ci_protocol == 0)
1699 error = tcp_listen_for_all();
1700 else
1701 error = sctp_listen_for_all();
1702 if (error)
1703 goto fail_unlisten;
1704
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001705 return 0;
1706
1707fail_unlisten:
David Teigland513ef592012-03-30 11:46:08 -05001708 dlm_allow_conn = 0;
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001709 con = nodeid2con(0,0);
1710 if (con) {
Marcelo Ricardo Leitner0d737a82015-08-11 19:22:21 -03001711 close_connection(con, false, true, true);
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001712 kmem_cache_free(con_cache, con);
1713 }
David Teigland513ef592012-03-30 11:46:08 -05001714fail_destroy:
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001715 kmem_cache_destroy(con_cache);
David Teigland513ef592012-03-30 11:46:08 -05001716fail:
Patrick Caulfield6ed7257b2007-04-17 15:39:57 +01001717 return error;
1718}
David Teigland36b71a82012-07-26 12:44:30 -05001719
1720void dlm_lowcomms_exit(void)
1721{
1722 struct dlm_node_addr *na, *safe;
1723
1724 spin_lock(&dlm_node_addrs_spin);
1725 list_for_each_entry_safe(na, safe, &dlm_node_addrs, list) {
1726 list_del(&na->list);
1727 while (na->addr_count--)
1728 kfree(na->addr[na->addr_count]);
1729 kfree(na);
1730 }
1731 spin_unlock(&dlm_node_addrs_spin);
1732}