Blame view
fs/dlm/lowcomms.c
36.2 KB
fdda387f7
|
1 2 3 4 |
/****************************************************************************** ******************************************************************************* ** ** Copyright (C) Sistina Software, Inc. 1997-2003 All rights reserved. |
a34fbc636
|
5 |
** Copyright (C) 2004-2007 Red Hat, Inc. All rights reserved. |
fdda387f7
|
6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 |
** ** This copyrighted material is made available to anyone wishing to use, ** modify, copy, or redistribute it subject to the terms and conditions ** of the GNU General Public License v.2. ** ******************************************************************************* ******************************************************************************/ /* * lowcomms.c * * This is the "low-level" comms layer. * * It is responsible for sending/receiving messages * from other nodes in the cluster. * * Cluster nodes are referred to by their nodeids. nodeids are * simply 32 bit numbers to the locking module - if they need to * be expanded for the cluster infrastructure then that is it's * responsibility. It is this layer's * responsibility to resolve these into IP address or * whatever it needs for inter-node communication. * * The comms level is two kernel threads that deal mainly with * the receiving of messages from other nodes and passing them * up to the mid-level comms layer (which understands the * message format) for execution by the locking core, and * a send thread which does all the setting up of connections * to remote nodes and the sending of data. Threads are not allowed * to send their own data because it may cause them to wait in times * of high load. Also, this way, the sending thread can collect together * messages bound for one node and send them in one block. * |
6ed7257b4
|
39 40 41 42 43 |
* lowcomms will choose to use wither TCP or SCTP as its transport layer * depending on the configuration variable 'protocol'. This should be set * to 0 (default) for TCP or 1 for SCTP. It shouldbe configured using a * cluster-wide mechanism as it must be the same on all nodes of the cluster * for the DLM to function. |
fdda387f7
|
44 45 |
* */ |
fdda387f7
|
46 47 48 49 |
#include <asm/ioctls.h> #include <net/sock.h> #include <net/tcp.h> #include <linux/pagemap.h> |
6ed7257b4
|
50 51 |
#include <linux/idr.h> #include <linux/file.h> |
7a936ce71
|
52 |
#include <linux/mutex.h> |
6ed7257b4
|
53 54 |
#include <linux/sctp.h> #include <net/sctp/user.h> |
fdda387f7
|
55 56 57 58 59 |
#include "dlm_internal.h" #include "lowcomms.h" #include "midcomms.h" #include "config.h" |
6ed7257b4
|
60 |
#define NEEDED_RMEM (4*1024*1024) |
fdda387f7
|
61 |
struct cbuf { |
ac33d0710
|
62 63 64 |
unsigned int base; unsigned int len; unsigned int mask; |
fdda387f7
|
65 |
}; |
ac33d0710
|
66 67 68 69 |
static void cbuf_add(struct cbuf *cb, int n) { cb->len += n; } |
fdda387f7
|
70 |
|
ac33d0710
|
71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 |
static int cbuf_data(struct cbuf *cb) { return ((cb->base + cb->len) & cb->mask); } static void cbuf_init(struct cbuf *cb, int size) { cb->base = cb->len = 0; cb->mask = size-1; } static void cbuf_eat(struct cbuf *cb, int n) { cb->len -= n; cb->base += n; cb->base &= cb->mask; } static bool cbuf_empty(struct cbuf *cb) { return cb->len == 0; } |
fdda387f7
|
93 |
|
fdda387f7
|
94 95 96 |
struct connection { struct socket *sock; /* NULL if not connected */ uint32_t nodeid; /* So we know who we are in the list */ |
f1f1c1ccf
|
97 |
struct mutex sock_mutex; |
6ed7257b4
|
98 |
unsigned long flags; |
fdda387f7
|
99 100 101 |
#define CF_READ_PENDING 1 #define CF_WRITE_PENDING 2 #define CF_CONNECT_PENDING 3 |
6ed7257b4
|
102 103 |
#define CF_INIT_PENDING 4 #define CF_IS_OTHERCON 5 |
ac33d0710
|
104 |
struct list_head writequeue; /* List of outgoing writequeue_entries */ |
fdda387f7
|
105 106 |
spinlock_t writequeue_lock; int (*rx_action) (struct connection *); /* What to do when active */ |
6ed7257b4
|
107 |
void (*connect_action) (struct connection *); /* What to do to connect */ |
fdda387f7
|
108 109 110 |
struct page *rx_page; struct cbuf cb; int retries; |
fdda387f7
|
111 |
#define MAX_CONNECT_RETRIES 3 |
6ed7257b4
|
112 |
int sctp_assoc; |
fdda387f7
|
113 |
struct connection *othercon; |
1d6e8131c
|
114 115 |
struct work_struct rwork; /* Receive workqueue */ struct work_struct swork; /* Send workqueue */ |
fdda387f7
|
116 117 118 119 120 121 122 123 124 125 126 127 128 |
}; #define sock2con(x) ((struct connection *)(x)->sk_user_data) /* An entry waiting to be sent */ struct writequeue_entry { struct list_head list; struct page *page; int offset; int len; int end; int users; struct connection *con; }; |
6ed7257b4
|
129 130 |
static struct sockaddr_storage *dlm_local_addr[DLM_MAX_ADDR_COUNT]; static int dlm_local_count; |
fdda387f7
|
131 |
|
1d6e8131c
|
132 133 134 |
/* Work queues */ static struct workqueue_struct *recv_workqueue; static struct workqueue_struct *send_workqueue; |
fdda387f7
|
135 |
|
6ed7257b4
|
136 |
static DEFINE_IDR(connections_idr); |
7a936ce71
|
137 |
static DEFINE_MUTEX(connections_lock); |
6ed7257b4
|
138 |
static int max_nodeid; |
c80e7c83d
|
139 |
static struct kmem_cache *con_cache; |
fdda387f7
|
140 |
|
1d6e8131c
|
141 142 |
static void process_recv_sockets(struct work_struct *work); static void process_send_sockets(struct work_struct *work); |
fdda387f7
|
143 |
|
6ed7257b4
|
144 145 146 147 148 |
/* * If 'allocation' is zero then we don't attempt to create a new * connection structure for this node. */ static struct connection *__nodeid2con(int nodeid, gfp_t alloc) |
fdda387f7
|
149 150 |
{ struct connection *con = NULL; |
6ed7257b4
|
151 152 |
int r; int n; |
fdda387f7
|
153 |
|
6ed7257b4
|
154 155 156 |
con = idr_find(&connections_idr, nodeid); if (con || !alloc) return con; |
fdda387f7
|
157 |
|
6ed7257b4
|
158 159 160 161 162 163 164 |
r = idr_pre_get(&connections_idr, alloc); if (!r) return NULL; con = kmem_cache_zalloc(con_cache, alloc); if (!con) return NULL; |
fdda387f7
|
165 |
|
6ed7257b4
|
166 167 168 169 170 |
r = idr_get_new_above(&connections_idr, con, nodeid, &n); if (r) { kmem_cache_free(con_cache, con); return NULL; } |
fdda387f7
|
171 |
|
6ed7257b4
|
172 173 174 175 |
if (n != nodeid) { idr_remove(&connections_idr, n); kmem_cache_free(con_cache, con); return NULL; |
fdda387f7
|
176 |
} |
6ed7257b4
|
177 178 179 180 181 182 |
con->nodeid = nodeid; mutex_init(&con->sock_mutex); INIT_LIST_HEAD(&con->writequeue); spin_lock_init(&con->writequeue_lock); INIT_WORK(&con->swork, process_send_sockets); INIT_WORK(&con->rwork, process_recv_sockets); |
fdda387f7
|
183 |
|
6ed7257b4
|
184 185 186 |
/* Setup action pointers for child sockets */ if (con->nodeid) { struct connection *zerocon = idr_find(&connections_idr, 0); |
fdda387f7
|
187 |
|
6ed7257b4
|
188 189 190 |
con->connect_action = zerocon->connect_action; if (!con->rx_action) con->rx_action = zerocon->rx_action; |
fdda387f7
|
191 |
} |
6ed7257b4
|
192 193 194 195 196 197 198 199 200 |
if (nodeid > max_nodeid) max_nodeid = nodeid; return con; } static struct connection *nodeid2con(int nodeid, gfp_t allocation) { struct connection *con; |
7a936ce71
|
201 |
mutex_lock(&connections_lock); |
6ed7257b4
|
202 |
con = __nodeid2con(nodeid, allocation); |
7a936ce71
|
203 |
mutex_unlock(&connections_lock); |
6ed7257b4
|
204 |
|
fdda387f7
|
205 206 |
return con; } |
6ed7257b4
|
207 208 209 210 211 |
/* This is a bit drastic, but only called when things go wrong */ static struct connection *assoc2con(int assoc_id) { int i; struct connection *con; |
7a936ce71
|
212 |
mutex_lock(&connections_lock); |
30d3a2373
|
213 |
for (i=0; i<=max_nodeid; i++) { |
6ed7257b4
|
214 215 |
con = __nodeid2con(i, 0); if (con && con->sctp_assoc == assoc_id) { |
7a936ce71
|
216 |
mutex_unlock(&connections_lock); |
6ed7257b4
|
217 218 219 |
return con; } } |
7a936ce71
|
220 |
mutex_unlock(&connections_lock); |
6ed7257b4
|
221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 |
return NULL; } static int nodeid_to_addr(int nodeid, struct sockaddr *retaddr) { struct sockaddr_storage addr; int error; if (!dlm_local_count) return -1; error = dlm_nodeid_to_addr(nodeid, &addr); if (error) return error; if (dlm_local_addr[0]->ss_family == AF_INET) { struct sockaddr_in *in4 = (struct sockaddr_in *) &addr; struct sockaddr_in *ret4 = (struct sockaddr_in *) retaddr; ret4->sin_addr.s_addr = in4->sin_addr.s_addr; } else { struct sockaddr_in6 *in6 = (struct sockaddr_in6 *) &addr; struct sockaddr_in6 *ret6 = (struct sockaddr_in6 *) retaddr; memcpy(&ret6->sin6_addr, &in6->sin6_addr, sizeof(in6->sin6_addr)); } return 0; } |
fdda387f7
|
249 250 251 252 |
/* Data available on socket or listen socket received a connect */ static void lowcomms_data_ready(struct sock *sk, int count_unused) { struct connection *con = sock2con(sk); |
afb853fb4
|
253 |
if (con && !test_and_set_bit(CF_READ_PENDING, &con->flags)) |
1d6e8131c
|
254 |
queue_work(recv_workqueue, &con->rwork); |
fdda387f7
|
255 256 257 258 259 |
} static void lowcomms_write_space(struct sock *sk) { struct connection *con = sock2con(sk); |
afb853fb4
|
260 |
if (con && !test_and_set_bit(CF_WRITE_PENDING, &con->flags)) |
1d6e8131c
|
261 |
queue_work(send_workqueue, &con->swork); |
fdda387f7
|
262 263 264 265 |
} static inline void lowcomms_connect_sock(struct connection *con) { |
1d6e8131c
|
266 267 |
if (!test_and_set_bit(CF_CONNECT_PENDING, &con->flags)) queue_work(send_workqueue, &con->swork); |
fdda387f7
|
268 269 270 271 |
} static void lowcomms_state_change(struct sock *sk) { |
ac33d0710
|
272 |
if (sk->sk_state == TCP_ESTABLISHED) |
fdda387f7
|
273 |
lowcomms_write_space(sk); |
fdda387f7
|
274 275 276 277 278 279 280 281 282 283 284 |
} /* Make a socket active */ static int add_sock(struct socket *sock, struct connection *con) { con->sock = sock; /* Install a data_ready callback */ con->sock->sk->sk_data_ready = lowcomms_data_ready; con->sock->sk->sk_write_space = lowcomms_write_space; con->sock->sk->sk_state_change = lowcomms_state_change; |
6ed7257b4
|
285 |
con->sock->sk->sk_user_data = con; |
fdda387f7
|
286 287 |
return 0; } |
6ed7257b4
|
288 |
/* Add the port number to an IPv6 or 4 sockaddr and return the address |
fdda387f7
|
289 290 291 292 |
length */ static void make_sockaddr(struct sockaddr_storage *saddr, uint16_t port, int *addr_len) { |
6ed7257b4
|
293 |
saddr->ss_family = dlm_local_addr[0]->ss_family; |
ac33d0710
|
294 |
if (saddr->ss_family == AF_INET) { |
fdda387f7
|
295 296 297 |
struct sockaddr_in *in4_addr = (struct sockaddr_in *)saddr; in4_addr->sin_port = cpu_to_be16(port); *addr_len = sizeof(struct sockaddr_in); |
6ed7257b4
|
298 |
memset(&in4_addr->sin_zero, 0, sizeof(in4_addr->sin_zero)); |
ac33d0710
|
299 |
} else { |
fdda387f7
|
300 301 302 303 |
struct sockaddr_in6 *in6_addr = (struct sockaddr_in6 *)saddr; in6_addr->sin6_port = cpu_to_be16(port); *addr_len = sizeof(struct sockaddr_in6); } |
01c8cab25
|
304 |
memset((char *)saddr + *addr_len, 0, sizeof(struct sockaddr_storage) - *addr_len); |
fdda387f7
|
305 306 307 |
} /* Close a remote connection and tidy up */ |
ac33d0710
|
308 |
static void close_connection(struct connection *con, bool and_other) |
fdda387f7
|
309 |
{ |
f1f1c1ccf
|
310 |
mutex_lock(&con->sock_mutex); |
fdda387f7
|
311 312 313 314 315 316 |
if (con->sock) { sock_release(con->sock); con->sock = NULL; } if (con->othercon && and_other) { |
ac33d0710
|
317 318 |
/* Will only re-enter once. */ close_connection(con->othercon, false); |
fdda387f7
|
319 320 321 322 323 |
} if (con->rx_page) { __free_page(con->rx_page); con->rx_page = NULL; } |
9e5f2825a
|
324 |
|
61d96be0f
|
325 326 |
con->retries = 0; mutex_unlock(&con->sock_mutex); |
fdda387f7
|
327 |
} |
6ed7257b4
|
328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 |
/* We only send shutdown messages to nodes that are not part of the cluster */ static void sctp_send_shutdown(sctp_assoc_t associd) { static char outcmsg[CMSG_SPACE(sizeof(struct sctp_sndrcvinfo))]; struct msghdr outmessage; struct cmsghdr *cmsg; struct sctp_sndrcvinfo *sinfo; int ret; struct connection *con; con = nodeid2con(0,0); BUG_ON(con == NULL); outmessage.msg_name = NULL; outmessage.msg_namelen = 0; outmessage.msg_control = outcmsg; outmessage.msg_controllen = sizeof(outcmsg); outmessage.msg_flags = MSG_EOR; cmsg = CMSG_FIRSTHDR(&outmessage); cmsg->cmsg_level = IPPROTO_SCTP; cmsg->cmsg_type = SCTP_SNDRCV; cmsg->cmsg_len = CMSG_LEN(sizeof(struct sctp_sndrcvinfo)); outmessage.msg_controllen = cmsg->cmsg_len; sinfo = CMSG_DATA(cmsg); memset(sinfo, 0x00, sizeof(struct sctp_sndrcvinfo)); sinfo->sinfo_flags |= MSG_EOF; sinfo->sinfo_assoc_id = associd; ret = kernel_sendmsg(con->sock, &outmessage, NULL, 0, 0); if (ret != 0) log_print("send EOF to node failed: %d", ret); } /* INIT failed but we don't know which node... restart INIT on all pending nodes */ static void sctp_init_failed(void) { int i; struct connection *con; |
7a936ce71
|
370 |
mutex_lock(&connections_lock); |
6ed7257b4
|
371 372 373 374 375 376 377 378 379 380 381 |
for (i=1; i<=max_nodeid; i++) { con = __nodeid2con(i, 0); if (!con) continue; con->sctp_assoc = 0; if (test_and_clear_bit(CF_CONNECT_PENDING, &con->flags)) { if (!test_and_set_bit(CF_WRITE_PENDING, &con->flags)) { queue_work(send_workqueue, &con->swork); } } } |
7a936ce71
|
382 |
mutex_unlock(&connections_lock); |
6ed7257b4
|
383 384 385 |
} /* Something happened to an association */ |
617e82e10
|
386 387 |
static void process_sctp_notification(struct connection *con, struct msghdr *msg, char *buf) |
6ed7257b4
|
388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 |
{ union sctp_notification *sn = (union sctp_notification *)buf; if (sn->sn_header.sn_type == SCTP_ASSOC_CHANGE) { switch (sn->sn_assoc_change.sac_state) { case SCTP_COMM_UP: case SCTP_RESTART: { /* Check that the new node is in the lockspace */ struct sctp_prim prim; int nodeid; int prim_len, ret; int addr_len; struct connection *new_con; struct file *file; sctp_peeloff_arg_t parg; int parglen = sizeof(parg); /* * We get this before any data for an association. * We verify that the node is in the cluster and * then peel off a socket for it. */ if ((int)sn->sn_assoc_change.sac_assoc_id <= 0) { log_print("COMM_UP for invalid assoc ID %d", |
617e82e10
|
414 |
(int)sn->sn_assoc_change.sac_assoc_id); |
6ed7257b4
|
415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 |
sctp_init_failed(); return; } memset(&prim, 0, sizeof(struct sctp_prim)); prim_len = sizeof(struct sctp_prim); prim.ssp_assoc_id = sn->sn_assoc_change.sac_assoc_id; ret = kernel_getsockopt(con->sock, IPPROTO_SCTP, SCTP_PRIMARY_ADDR, (char*)&prim, &prim_len); if (ret < 0) { log_print("getsockopt/sctp_primary_addr on " "new assoc %d failed : %d", (int)sn->sn_assoc_change.sac_assoc_id, ret); /* Retry INIT later */ new_con = assoc2con(sn->sn_assoc_change.sac_assoc_id); if (new_con) clear_bit(CF_CONNECT_PENDING, &con->flags); return; } make_sockaddr(&prim.ssp_addr, 0, &addr_len); if (dlm_addr_to_nodeid(&prim.ssp_addr, &nodeid)) { int i; unsigned char *b=(unsigned char *)&prim.ssp_addr; log_print("reject connect from unknown addr"); for (i=0; i<sizeof(struct sockaddr_storage);i++) printk("%02x ", b[i]); printk(" "); sctp_send_shutdown(prim.ssp_assoc_id); return; } new_con = nodeid2con(nodeid, GFP_KERNEL); if (!new_con) return; /* Peel off a new sock */ parg.associd = sn->sn_assoc_change.sac_assoc_id; |
617e82e10
|
458 459 |
ret = kernel_getsockopt(con->sock, IPPROTO_SCTP, SCTP_SOCKOPT_PEELOFF, |
6ed7257b4
|
460 |
(void *)&parg, &parglen); |
30d3a2373
|
461 |
if (ret) { |
617e82e10
|
462 463 464 |
log_print("Can't peel off a socket for " "connection %d to node %d: err=%d ", |
6ed7257b4
|
465 |
parg.associd, nodeid, ret); |
6ed7257b4
|
466 |
} |
6ed7257b4
|
467 468 469 470 471 472 473 |
file = fget(parg.sd); new_con->sock = SOCKET_I(file->f_dentry->d_inode); add_sock(new_con->sock, new_con); fput(file); put_unused_fd(parg.sd); log_print("got new/restarted association %d nodeid %d", |
617e82e10
|
474 |
(int)sn->sn_assoc_change.sac_assoc_id, nodeid); |
6ed7257b4
|
475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 |
/* Send any pending writes */ clear_bit(CF_CONNECT_PENDING, &new_con->flags); clear_bit(CF_INIT_PENDING, &con->flags); if (!test_and_set_bit(CF_WRITE_PENDING, &new_con->flags)) { queue_work(send_workqueue, &new_con->swork); } if (!test_and_set_bit(CF_READ_PENDING, &new_con->flags)) queue_work(recv_workqueue, &new_con->rwork); } break; case SCTP_COMM_LOST: case SCTP_SHUTDOWN_COMP: { con = assoc2con(sn->sn_assoc_change.sac_assoc_id); if (con) { con->sctp_assoc = 0; } } break; /* We don't know which INIT failed, so clear the PENDING flags * on them all. if assoc_id is zero then it will then try * again */ case SCTP_CANT_STR_ASSOC: { log_print("Can't start SCTP association - retrying"); sctp_init_failed(); } break; default: log_print("unexpected SCTP assoc change id=%d state=%d", (int)sn->sn_assoc_change.sac_assoc_id, sn->sn_assoc_change.sac_state); } } } |
fdda387f7
|
515 516 517 518 |
/* Data received from remote end */ static int receive_from_sock(struct connection *con) { int ret = 0; |
58addbffd
|
519 520 |
struct msghdr msg = {}; struct kvec iov[2]; |
fdda387f7
|
521 522 523 |
unsigned len; int r; int call_again_soon = 0; |
58addbffd
|
524 |
int nvec; |
6ed7257b4
|
525 |
char incmsg[CMSG_SPACE(sizeof(struct sctp_sndrcvinfo))]; |
fdda387f7
|
526 |
|
f1f1c1ccf
|
527 |
mutex_lock(&con->sock_mutex); |
fdda387f7
|
528 |
|
a34fbc636
|
529 530 531 532 |
if (con->sock == NULL) { ret = -EAGAIN; goto out_close; } |
fdda387f7
|
533 534 535 536 537 538 539 540 |
if (con->rx_page == NULL) { /* * This doesn't need to be atomic, but I think it should * improve performance if it is. */ con->rx_page = alloc_page(GFP_ATOMIC); if (con->rx_page == NULL) goto out_resched; |
ac33d0710
|
541 |
cbuf_init(&con->cb, PAGE_CACHE_SIZE); |
fdda387f7
|
542 |
} |
6ed7257b4
|
543 544 545 546 |
/* Only SCTP needs these really */ memset(&incmsg, 0, sizeof(incmsg)); msg.msg_control = incmsg; msg.msg_controllen = sizeof(incmsg); |
fdda387f7
|
547 548 549 550 |
/* * iov[0] is the bit of the circular buffer between the current end * point (cb.base + cb.len) and the end of the buffer. */ |
ac33d0710
|
551 552 |
iov[0].iov_len = con->cb.base - cbuf_data(&con->cb); iov[0].iov_base = page_address(con->rx_page) + cbuf_data(&con->cb); |
89adc934f
|
553 |
iov[1].iov_len = 0; |
58addbffd
|
554 |
nvec = 1; |
fdda387f7
|
555 556 557 558 559 |
/* * iov[1] is the bit of the circular buffer between the start of the * buffer and the start of the currently used section (cb.base) */ |
ac33d0710
|
560 561 |
if (cbuf_data(&con->cb) >= con->cb.base) { iov[0].iov_len = PAGE_CACHE_SIZE - cbuf_data(&con->cb); |
fdda387f7
|
562 563 |
iov[1].iov_len = con->cb.base; iov[1].iov_base = page_address(con->rx_page); |
58addbffd
|
564 |
nvec = 2; |
fdda387f7
|
565 566 |
} len = iov[0].iov_len + iov[1].iov_len; |
58addbffd
|
567 |
r = ret = kernel_recvmsg(con->sock, &msg, iov, nvec, len, |
fdda387f7
|
568 |
MSG_DONTWAIT | MSG_NOSIGNAL); |
fdda387f7
|
569 570 |
if (ret <= 0) goto out_close; |
bd44e2b00
|
571 |
|
6ed7257b4
|
572 573 |
/* Process SCTP notifications */ if (msg.msg_flags & MSG_NOTIFICATION) { |
6ed7257b4
|
574 575 576 577 |
msg.msg_control = incmsg; msg.msg_controllen = sizeof(incmsg); process_sctp_notification(con, &msg, |
617e82e10
|
578 |
page_address(con->rx_page) + con->cb.base); |
6ed7257b4
|
579 580 581 582 |
mutex_unlock(&con->sock_mutex); return 0; } BUG_ON(con->nodeid == 0); |
fdda387f7
|
583 584 |
if (ret == len) call_again_soon = 1; |
ac33d0710
|
585 |
cbuf_add(&con->cb, ret); |
fdda387f7
|
586 587 588 589 590 |
ret = dlm_process_incoming_buffer(con->nodeid, page_address(con->rx_page), con->cb.base, con->cb.len, PAGE_CACHE_SIZE); if (ret == -EBADMSG) { |
617e82e10
|
591 592 593 594 |
log_print("lowcomms: addr=%p, base=%u, len=%u, " "iov_len=%u, iov_base[0]=%p, read=%d", page_address(con->rx_page), con->cb.base, con->cb.len, len, iov[0].iov_base, r); |
fdda387f7
|
595 596 597 |
} if (ret < 0) goto out_close; |
ac33d0710
|
598 |
cbuf_eat(&con->cb, ret); |
fdda387f7
|
599 |
|
ac33d0710
|
600 |
if (cbuf_empty(&con->cb) && !call_again_soon) { |
fdda387f7
|
601 602 603 |
__free_page(con->rx_page); con->rx_page = NULL; } |
fdda387f7
|
604 605 |
if (call_again_soon) goto out_resched; |
f1f1c1ccf
|
606 |
mutex_unlock(&con->sock_mutex); |
ac33d0710
|
607 |
return 0; |
fdda387f7
|
608 |
|
ac33d0710
|
609 |
out_resched: |
1d6e8131c
|
610 611 |
if (!test_and_set_bit(CF_READ_PENDING, &con->flags)) queue_work(recv_workqueue, &con->rwork); |
f1f1c1ccf
|
612 |
mutex_unlock(&con->sock_mutex); |
bd44e2b00
|
613 |
return -EAGAIN; |
fdda387f7
|
614 |
|
ac33d0710
|
615 |
out_close: |
f1f1c1ccf
|
616 |
mutex_unlock(&con->sock_mutex); |
9e5f2825a
|
617 |
if (ret != -EAGAIN) { |
ac33d0710
|
618 |
close_connection(con, false); |
fdda387f7
|
619 620 |
/* Reconnect when there is something to send */ } |
a34fbc636
|
621 622 623 |
/* Don't return success if we really got EOF */ if (ret == 0) ret = -EAGAIN; |
fdda387f7
|
624 |
|
fdda387f7
|
625 626 627 628 |
return ret; } /* Listening socket is busy, accept a connection */ |
6ed7257b4
|
629 |
static int tcp_accept_from_sock(struct connection *con) |
fdda387f7
|
630 631 632 633 634 635 636 |
{ int result; struct sockaddr_storage peeraddr; struct socket *newsock; int len; int nodeid; struct connection *newcon; |
bd44e2b00
|
637 |
struct connection *addcon; |
fdda387f7
|
638 639 |
memset(&peeraddr, 0, sizeof(peeraddr)); |
6ed7257b4
|
640 |
result = sock_create_kern(dlm_local_addr[0]->ss_family, SOCK_STREAM, |
ac33d0710
|
641 |
IPPROTO_TCP, &newsock); |
fdda387f7
|
642 643 |
if (result < 0) return -ENOMEM; |
f1f1c1ccf
|
644 |
mutex_lock_nested(&con->sock_mutex, 0); |
fdda387f7
|
645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 |
result = -ENOTCONN; if (con->sock == NULL) goto accept_err; newsock->type = con->sock->type; newsock->ops = con->sock->ops; result = con->sock->ops->accept(con->sock, newsock, O_NONBLOCK); if (result < 0) goto accept_err; /* Get the connected socket's peer */ memset(&peeraddr, 0, sizeof(peeraddr)); if (newsock->ops->getname(newsock, (struct sockaddr *)&peeraddr, &len, 2)) { result = -ECONNABORTED; goto accept_err; } /* Get the new node's NODEID */ make_sockaddr(&peeraddr, 0, &len); if (dlm_addr_to_nodeid(&peeraddr, &nodeid)) { |
617e82e10
|
668 |
log_print("connect from non cluster node"); |
fdda387f7
|
669 |
sock_release(newsock); |
f1f1c1ccf
|
670 |
mutex_unlock(&con->sock_mutex); |
fdda387f7
|
671 672 673 674 675 676 677 678 |
return -1; } log_print("got connection from %d", nodeid); /* Check to see if we already have a connection to this node. This * could happen if the two nodes initiate a connection at roughly * the same time and the connections cross on the wire. |
fdda387f7
|
679 680 681 682 683 684 685 |
* In this case we store the incoming one in "othercon" */ newcon = nodeid2con(nodeid, GFP_KERNEL); if (!newcon) { result = -ENOMEM; goto accept_err; } |
f1f1c1ccf
|
686 |
mutex_lock_nested(&newcon->sock_mutex, 1); |
fdda387f7
|
687 |
if (newcon->sock) { |
ac33d0710
|
688 |
struct connection *othercon = newcon->othercon; |
fdda387f7
|
689 690 |
if (!othercon) { |
ac33d0710
|
691 |
othercon = kmem_cache_zalloc(con_cache, GFP_KERNEL); |
fdda387f7
|
692 |
if (!othercon) { |
617e82e10
|
693 |
log_print("failed to allocate incoming socket"); |
f1f1c1ccf
|
694 |
mutex_unlock(&newcon->sock_mutex); |
fdda387f7
|
695 696 697 |
result = -ENOMEM; goto accept_err; } |
fdda387f7
|
698 699 |
othercon->nodeid = nodeid; othercon->rx_action = receive_from_sock; |
f1f1c1ccf
|
700 |
mutex_init(&othercon->sock_mutex); |
1d6e8131c
|
701 702 |
INIT_WORK(&othercon->swork, process_send_sockets); INIT_WORK(&othercon->rwork, process_recv_sockets); |
fdda387f7
|
703 |
set_bit(CF_IS_OTHERCON, &othercon->flags); |
61d96be0f
|
704 705 |
} if (!othercon->sock) { |
fdda387f7
|
706 |
newcon->othercon = othercon; |
97d848365
|
707 708 709 710 711 712 713 714 715 |
othercon->sock = newsock; newsock->sk->sk_user_data = othercon; add_sock(newsock, othercon); addcon = othercon; } else { printk("Extra connection from node %d attempted ", nodeid); result = -EAGAIN; |
f4fadb23c
|
716 |
mutex_unlock(&newcon->sock_mutex); |
97d848365
|
717 |
goto accept_err; |
fdda387f7
|
718 |
} |
fdda387f7
|
719 720 721 722 723 |
} else { newsock->sk->sk_user_data = newcon; newcon->rx_action = receive_from_sock; add_sock(newsock, newcon); |
bd44e2b00
|
724 |
addcon = newcon; |
fdda387f7
|
725 |
} |
f1f1c1ccf
|
726 |
mutex_unlock(&newcon->sock_mutex); |
fdda387f7
|
727 728 729 730 731 732 |
/* * Add it to the active queue in case we got data * beween processing the accept adding the socket * to the read_sockets list */ |
bd44e2b00
|
733 734 |
if (!test_and_set_bit(CF_READ_PENDING, &addcon->flags)) queue_work(recv_workqueue, &addcon->rwork); |
f1f1c1ccf
|
735 |
mutex_unlock(&con->sock_mutex); |
fdda387f7
|
736 737 |
return 0; |
ac33d0710
|
738 |
accept_err: |
f1f1c1ccf
|
739 |
mutex_unlock(&con->sock_mutex); |
fdda387f7
|
740 741 742 |
sock_release(newsock); if (result != -EAGAIN) |
617e82e10
|
743 |
log_print("error accepting connection from node: %d", result); |
fdda387f7
|
744 745 |
return result; } |
6ed7257b4
|
746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 |
static void free_entry(struct writequeue_entry *e) { __free_page(e->page); kfree(e); } /* Initiate an SCTP association. This is a special case of send_to_sock() in that we don't yet have a peeled-off socket for this association, so we use the listening socket and add the primary IP address of the remote node. */ static void sctp_init_assoc(struct connection *con) { struct sockaddr_storage rem_addr; char outcmsg[CMSG_SPACE(sizeof(struct sctp_sndrcvinfo))]; struct msghdr outmessage; struct cmsghdr *cmsg; struct sctp_sndrcvinfo *sinfo; struct connection *base_con; struct writequeue_entry *e; int len, offset; int ret; int addrlen; struct kvec iov[1]; if (test_and_set_bit(CF_INIT_PENDING, &con->flags)) return; if (con->retries++ > MAX_CONNECT_RETRIES) return; log_print("Initiating association with node %d", con->nodeid); if (nodeid_to_addr(con->nodeid, (struct sockaddr *)&rem_addr)) { log_print("no address for nodeid %d", con->nodeid); return; } base_con = nodeid2con(0, 0); BUG_ON(base_con == NULL); make_sockaddr(&rem_addr, dlm_config.ci_tcp_port, &addrlen); outmessage.msg_name = &rem_addr; outmessage.msg_namelen = addrlen; outmessage.msg_control = outcmsg; outmessage.msg_controllen = sizeof(outcmsg); outmessage.msg_flags = MSG_EOR; spin_lock(&con->writequeue_lock); e = list_entry(con->writequeue.next, struct writequeue_entry, list); BUG_ON((struct list_head *) e == &con->writequeue); len = e->len; offset = e->offset; spin_unlock(&con->writequeue_lock); kmap(e->page); /* Send the first block off the write queue */ iov[0].iov_base = page_address(e->page)+offset; iov[0].iov_len = len; cmsg = CMSG_FIRSTHDR(&outmessage); cmsg->cmsg_level = IPPROTO_SCTP; cmsg->cmsg_type = SCTP_SNDRCV; cmsg->cmsg_len = CMSG_LEN(sizeof(struct sctp_sndrcvinfo)); sinfo = CMSG_DATA(cmsg); memset(sinfo, 0x00, sizeof(struct sctp_sndrcvinfo)); sinfo->sinfo_ppid = cpu_to_le32(dlm_our_nodeid()); outmessage.msg_controllen = cmsg->cmsg_len; ret = kernel_sendmsg(base_con->sock, &outmessage, iov, 1, len); if (ret < 0) { |
617e82e10
|
820 821 |
log_print("Send first packet to node %d failed: %d", con->nodeid, ret); |
6ed7257b4
|
822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 |
/* Try again later */ clear_bit(CF_CONNECT_PENDING, &con->flags); clear_bit(CF_INIT_PENDING, &con->flags); } else { spin_lock(&con->writequeue_lock); e->offset += ret; e->len -= ret; if (e->len == 0 && e->users == 0) { list_del(&e->list); kunmap(e->page); free_entry(e); } spin_unlock(&con->writequeue_lock); } } |
fdda387f7
|
840 |
/* Connect a new socket to its peer */ |
6ed7257b4
|
841 |
static void tcp_connect_to_sock(struct connection *con) |
fdda387f7
|
842 843 |
{ int result = -EHOSTUNREACH; |
6bd8fedaa
|
844 |
struct sockaddr_storage saddr, src_addr; |
fdda387f7
|
845 846 847 848 849 |
int addr_len; struct socket *sock; if (con->nodeid == 0) { log_print("attempt to connect sock 0 foiled"); |
ac33d0710
|
850 |
return; |
fdda387f7
|
851 |
} |
f1f1c1ccf
|
852 |
mutex_lock(&con->sock_mutex); |
fdda387f7
|
853 854 855 856 857 858 859 860 861 862 |
if (con->retries++ > MAX_CONNECT_RETRIES) goto out; /* Some odd races can cause double-connects, ignore them */ if (con->sock) { result = 0; goto out; } /* Create a socket to communicate with */ |
6ed7257b4
|
863 |
result = sock_create_kern(dlm_local_addr[0]->ss_family, SOCK_STREAM, |
ac33d0710
|
864 |
IPPROTO_TCP, &sock); |
fdda387f7
|
865 866 867 868 869 |
if (result < 0) goto out_err; memset(&saddr, 0, sizeof(saddr)); if (dlm_nodeid_to_addr(con->nodeid, &saddr)) |
ac33d0710
|
870 |
goto out_err; |
fdda387f7
|
871 872 873 |
sock->sk->sk_user_data = con; con->rx_action = receive_from_sock; |
6ed7257b4
|
874 875 |
con->connect_action = tcp_connect_to_sock; add_sock(sock, con); |
fdda387f7
|
876 |
|
6bd8fedaa
|
877 878 879 880 881 882 883 884 885 886 |
/* Bind to our cluster-known address connecting to avoid routing problems */ memcpy(&src_addr, dlm_local_addr[0], sizeof(src_addr)); make_sockaddr(&src_addr, 0, &addr_len); result = sock->ops->bind(sock, (struct sockaddr *) &src_addr, addr_len); if (result < 0) { log_print("could not bind for connect: %d", result); /* This *may* not indicate a critical error */ } |
68c817a1c
|
887 |
make_sockaddr(&saddr, dlm_config.ci_tcp_port, &addr_len); |
fdda387f7
|
888 |
|
fdda387f7
|
889 890 891 |
log_print("connecting to %d", con->nodeid); result = sock->ops->connect(sock, (struct sockaddr *)&saddr, addr_len, |
ac33d0710
|
892 |
O_NONBLOCK); |
fdda387f7
|
893 894 |
if (result == -EINPROGRESS) result = 0; |
ac33d0710
|
895 896 |
if (result == 0) goto out; |
fdda387f7
|
897 |
|
ac33d0710
|
898 |
out_err: |
fdda387f7
|
899 900 901 902 903 904 905 906 907 |
if (con->sock) { sock_release(con->sock); con->sock = NULL; } /* * Some errors are fatal and this list might need adjusting. For other * errors we try again until the max number of retries is reached. */ if (result != -EHOSTUNREACH && result != -ENETUNREACH && |
0035a4b14
|
908 |
result != -ENETDOWN && result != -EINVAL |
fdda387f7
|
909 910 911 912 |
&& result != -EPROTONOSUPPORT) { lowcomms_connect_sock(con); result = 0; } |
ac33d0710
|
913 |
out: |
f1f1c1ccf
|
914 |
mutex_unlock(&con->sock_mutex); |
ac33d0710
|
915 |
return; |
fdda387f7
|
916 |
} |
6ed7257b4
|
917 918 |
static struct socket *tcp_create_listen_sock(struct connection *con, struct sockaddr_storage *saddr) |
fdda387f7
|
919 |
{ |
ac33d0710
|
920 |
struct socket *sock = NULL; |
fdda387f7
|
921 922 923 |
int result = 0; int one = 1; int addr_len; |
6ed7257b4
|
924 |
if (dlm_local_addr[0]->ss_family == AF_INET) |
fdda387f7
|
925 926 927 928 929 |
addr_len = sizeof(struct sockaddr_in); else addr_len = sizeof(struct sockaddr_in6); /* Create a socket to communicate with */ |
617e82e10
|
930 931 |
result = sock_create_kern(dlm_local_addr[0]->ss_family, SOCK_STREAM, IPPROTO_TCP, &sock); |
fdda387f7
|
932 |
if (result < 0) { |
617e82e10
|
933 |
log_print("Can't create listening comms socket"); |
fdda387f7
|
934 935 |
goto create_out; } |
6ed7257b4
|
936 937 |
result = kernel_setsockopt(sock, SOL_SOCKET, SO_REUSEADDR, (char *)&one, sizeof(one)); |
fdda387f7
|
938 |
if (result < 0) { |
617e82e10
|
939 |
log_print("Failed to set SO_REUSEADDR on socket: %d", result); |
fdda387f7
|
940 941 |
} sock->sk->sk_user_data = con; |
6ed7257b4
|
942 943 |
con->rx_action = tcp_accept_from_sock; con->connect_action = tcp_connect_to_sock; |
fdda387f7
|
944 945 946 |
con->sock = sock; /* Bind to our port */ |
68c817a1c
|
947 |
make_sockaddr(saddr, dlm_config.ci_tcp_port, &addr_len); |
fdda387f7
|
948 949 |
result = sock->ops->bind(sock, (struct sockaddr *) saddr, addr_len); if (result < 0) { |
617e82e10
|
950 |
log_print("Can't bind to port %d", dlm_config.ci_tcp_port); |
fdda387f7
|
951 952 953 954 955 |
sock_release(sock); sock = NULL; con->sock = NULL; goto create_out; } |
6ed7257b4
|
956 |
result = kernel_setsockopt(sock, SOL_SOCKET, SO_KEEPALIVE, |
ac33d0710
|
957 |
(char *)&one, sizeof(one)); |
fdda387f7
|
958 |
if (result < 0) { |
617e82e10
|
959 |
log_print("Set keepalive failed: %d", result); |
fdda387f7
|
960 961 962 963 |
} result = sock->ops->listen(sock, 5); if (result < 0) { |
617e82e10
|
964 |
log_print("Can't listen on port %d", dlm_config.ci_tcp_port); |
fdda387f7
|
965 966 967 968 |
sock_release(sock); sock = NULL; goto create_out; } |
ac33d0710
|
969 |
create_out: |
fdda387f7
|
970 971 |
return sock; } |
6ed7257b4
|
972 973 974 975 976 |
/* Get local addresses */ static void init_local(void) { struct sockaddr_storage sas, *addr; int i; |
30d3a2373
|
977 |
dlm_local_count = 0; |
6ed7257b4
|
978 979 980 981 982 983 984 985 986 987 988 |
for (i = 0; i < DLM_MAX_ADDR_COUNT - 1; i++) { if (dlm_our_addr(&sas, i)) break; addr = kmalloc(sizeof(*addr), GFP_KERNEL); if (!addr) break; memcpy(addr, &sas, sizeof(*addr)); dlm_local_addr[dlm_local_count++] = addr; } } |
617e82e10
|
989 990 991 992 993 |
/* Bind to an IP address. SCTP allows multiple address so it can do multi-homing */ static int add_sctp_bind_addr(struct connection *sctp_con, struct sockaddr_storage *addr, int addr_len, int num) |
6ed7257b4
|
994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 |
{ int result = 0; if (num == 1) result = kernel_bind(sctp_con->sock, (struct sockaddr *) addr, addr_len); else result = kernel_setsockopt(sctp_con->sock, SOL_SCTP, SCTP_SOCKOPT_BINDX_ADD, (char *)addr, addr_len); if (result < 0) log_print("Can't bind to port %d addr number %d", dlm_config.ci_tcp_port, num); return result; } |
fdda387f7
|
1012 |
|
6ed7257b4
|
1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 |
/* Initialise SCTP socket and bind to all interfaces */ static int sctp_listen_for_all(void) { struct socket *sock = NULL; struct sockaddr_storage localaddr; struct sctp_event_subscribe subscribe; int result = -EINVAL, num = 1, i, addr_len; struct connection *con = nodeid2con(0, GFP_KERNEL); int bufsize = NEEDED_RMEM; if (!con) return -ENOMEM; log_print("Using SCTP for communications"); result = sock_create_kern(dlm_local_addr[0]->ss_family, SOCK_SEQPACKET, IPPROTO_SCTP, &sock); if (result < 0) { log_print("Can't create comms socket, check SCTP is loaded"); goto out; } /* Listen for events */ memset(&subscribe, 0, sizeof(subscribe)); subscribe.sctp_data_io_event = 1; subscribe.sctp_association_event = 1; subscribe.sctp_send_failure_event = 1; subscribe.sctp_shutdown_event = 1; subscribe.sctp_partial_delivery_event = 1; |
df61c9526
|
1042 |
result = kernel_setsockopt(sock, SOL_SOCKET, SO_RCVBUFFORCE, |
6ed7257b4
|
1043 1044 |
(char *)&bufsize, sizeof(bufsize)); if (result) |
617e82e10
|
1045 |
log_print("Error increasing buffer space on socket %d", result); |
6ed7257b4
|
1046 1047 |
result = kernel_setsockopt(sock, SOL_SCTP, SCTP_EVENTS, |
617e82e10
|
1048 |
(char *)&subscribe, sizeof(subscribe)); |
6ed7257b4
|
1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 |
if (result < 0) { log_print("Failed to set SCTP_EVENTS on socket: result=%d", result); goto create_delsock; } /* Init con struct */ sock->sk->sk_user_data = con; con->sock = sock; con->sock->sk->sk_data_ready = lowcomms_data_ready; con->rx_action = receive_from_sock; con->connect_action = sctp_init_assoc; /* Bind to all interfaces. */ for (i = 0; i < dlm_local_count; i++) { memcpy(&localaddr, dlm_local_addr[i], sizeof(localaddr)); make_sockaddr(&localaddr, dlm_config.ci_tcp_port, &addr_len); result = add_sctp_bind_addr(con, &localaddr, addr_len, num); if (result) goto create_delsock; ++num; } result = sock->ops->listen(sock, 5); if (result < 0) { log_print("Can't set socket listening"); goto create_delsock; } return 0; create_delsock: sock_release(sock); con->sock = NULL; out: return result; } static int tcp_listen_for_all(void) |
fdda387f7
|
1089 1090 1091 1092 |
{ struct socket *sock = NULL; struct connection *con = nodeid2con(0, GFP_KERNEL); int result = -EINVAL; |
6ed7257b4
|
1093 1094 |
if (!con) return -ENOMEM; |
fdda387f7
|
1095 |
/* We don't support multi-homed hosts */ |
6ed7257b4
|
1096 |
if (dlm_local_addr[1] != NULL) { |
617e82e10
|
1097 1098 |
log_print("TCP protocol can't handle multi-homed hosts, " "try SCTP"); |
6ed7257b4
|
1099 1100 1101 1102 |
return -EINVAL; } log_print("Using TCP for communications"); |
6ed7257b4
|
1103 |
sock = tcp_create_listen_sock(con, dlm_local_addr[0]); |
fdda387f7
|
1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 |
if (sock) { add_sock(sock, con); result = 0; } else { result = -EADDRINUSE; } return result; } static struct writequeue_entry *new_writequeue_entry(struct connection *con, gfp_t allocation) { struct writequeue_entry *entry; entry = kmalloc(sizeof(struct writequeue_entry), allocation); if (!entry) return NULL; entry->page = alloc_page(allocation); if (!entry->page) { kfree(entry); return NULL; } entry->offset = 0; entry->len = 0; entry->end = 0; entry->users = 0; entry->con = con; return entry; } |
617e82e10
|
1140 |
void *dlm_lowcomms_get_buffer(int nodeid, int len, gfp_t allocation, char **ppc) |
fdda387f7
|
1141 1142 1143 1144 1145 |
{ struct connection *con; struct writequeue_entry *e; int offset = 0; int users = 0; |
fdda387f7
|
1146 1147 1148 |
con = nodeid2con(nodeid, allocation); if (!con) return NULL; |
4edde74ee
|
1149 |
spin_lock(&con->writequeue_lock); |
fdda387f7
|
1150 |
e = list_entry(con->writequeue.prev, struct writequeue_entry, list); |
ac33d0710
|
1151 |
if ((&e->list == &con->writequeue) || |
fdda387f7
|
1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 |
(PAGE_CACHE_SIZE - e->end < len)) { e = NULL; } else { offset = e->end; e->end += len; users = e->users++; } spin_unlock(&con->writequeue_lock); if (e) { |
ac33d0710
|
1162 |
got_one: |
fdda387f7
|
1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 |
if (users == 0) kmap(e->page); *ppc = page_address(e->page) + offset; return e; } e = new_writequeue_entry(con, allocation); if (e) { spin_lock(&con->writequeue_lock); offset = e->end; e->end += len; users = e->users++; list_add_tail(&e->list, &con->writequeue); spin_unlock(&con->writequeue_lock); goto got_one; } return NULL; } void dlm_lowcomms_commit_buffer(void *mh) { struct writequeue_entry *e = (struct writequeue_entry *)mh; struct connection *con = e->con; int users; |
4edde74ee
|
1187 |
spin_lock(&con->writequeue_lock); |
fdda387f7
|
1188 1189 1190 1191 1192 1193 |
users = --e->users; if (users) goto out; e->len = e->end - e->offset; kunmap(e->page); spin_unlock(&con->writequeue_lock); |
1d6e8131c
|
1194 1195 |
if (!test_and_set_bit(CF_WRITE_PENDING, &con->flags)) { queue_work(send_workqueue, &con->swork); |
fdda387f7
|
1196 1197 |
} return; |
ac33d0710
|
1198 |
out: |
fdda387f7
|
1199 1200 1201 |
spin_unlock(&con->writequeue_lock); return; } |
fdda387f7
|
1202 |
/* Send a message */ |
ac33d0710
|
1203 |
static void send_to_sock(struct connection *con) |
fdda387f7
|
1204 1205 1206 1207 1208 1209 |
{ int ret = 0; ssize_t(*sendpage) (struct socket *, struct page *, int, size_t, int); const int msg_flags = MSG_DONTWAIT | MSG_NOSIGNAL; struct writequeue_entry *e; int len, offset; |
f1f1c1ccf
|
1210 |
mutex_lock(&con->sock_mutex); |
fdda387f7
|
1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 |
if (con->sock == NULL) goto out_connect; sendpage = con->sock->ops->sendpage; spin_lock(&con->writequeue_lock); for (;;) { e = list_entry(con->writequeue.next, struct writequeue_entry, list); if ((struct list_head *) e == &con->writequeue) break; len = e->len; offset = e->offset; BUG_ON(len == 0 && e->users == 0); spin_unlock(&con->writequeue_lock); |
1d6e8131c
|
1227 |
kmap(e->page); |
fdda387f7
|
1228 1229 1230 1231 1232 |
ret = 0; if (len) { ret = sendpage(con->sock, e->page, offset, len, msg_flags); |
d66f8277f
|
1233 1234 |
if (ret == -EAGAIN || ret == 0) { cond_resched(); |
fdda387f7
|
1235 |
goto out; |
d66f8277f
|
1236 |
} |
fdda387f7
|
1237 1238 |
if (ret <= 0) goto send_error; |
d66f8277f
|
1239 |
} |
fdda387f7
|
1240 |
/* Don't starve people filling buffers */ |
ac33d0710
|
1241 |
cond_resched(); |
fdda387f7
|
1242 1243 1244 1245 1246 1247 1248 |
spin_lock(&con->writequeue_lock); e->offset += ret; e->len -= ret; if (e->len == 0 && e->users == 0) { list_del(&e->list); |
ac33d0710
|
1249 |
kunmap(e->page); |
fdda387f7
|
1250 1251 1252 1253 1254 |
free_entry(e); continue; } } spin_unlock(&con->writequeue_lock); |
ac33d0710
|
1255 |
out: |
f1f1c1ccf
|
1256 |
mutex_unlock(&con->sock_mutex); |
ac33d0710
|
1257 |
return; |
fdda387f7
|
1258 |
|
ac33d0710
|
1259 |
send_error: |
f1f1c1ccf
|
1260 |
mutex_unlock(&con->sock_mutex); |
ac33d0710
|
1261 |
close_connection(con, false); |
fdda387f7
|
1262 |
lowcomms_connect_sock(con); |
ac33d0710
|
1263 |
return; |
fdda387f7
|
1264 |
|
ac33d0710
|
1265 |
out_connect: |
f1f1c1ccf
|
1266 |
mutex_unlock(&con->sock_mutex); |
6ed7257b4
|
1267 1268 |
if (!test_bit(CF_INIT_PENDING, &con->flags)) lowcomms_connect_sock(con); |
ac33d0710
|
1269 |
return; |
fdda387f7
|
1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 |
} static void clean_one_writequeue(struct connection *con) { struct list_head *list; struct list_head *temp; spin_lock(&con->writequeue_lock); list_for_each_safe(list, temp, &con->writequeue) { struct writequeue_entry *e = list_entry(list, struct writequeue_entry, list); list_del(&e->list); free_entry(e); } spin_unlock(&con->writequeue_lock); } /* Called from recovery when it knows that a node has left the cluster */ int dlm_lowcomms_close(int nodeid) { struct connection *con; |
fdda387f7
|
1292 1293 1294 1295 |
log_print("closing connection to node %d", nodeid); con = nodeid2con(nodeid, 0); if (con) { clean_one_writequeue(con); |
ac33d0710
|
1296 |
close_connection(con, true); |
fdda387f7
|
1297 1298 |
} return 0; |
fdda387f7
|
1299 |
} |
6ed7257b4
|
1300 |
/* Receive workqueue function */ |
1d6e8131c
|
1301 |
static void process_recv_sockets(struct work_struct *work) |
fdda387f7
|
1302 |
{ |
1d6e8131c
|
1303 1304 |
struct connection *con = container_of(work, struct connection, rwork); int err; |
fdda387f7
|
1305 |
|
1d6e8131c
|
1306 1307 1308 1309 |
clear_bit(CF_READ_PENDING, &con->flags); do { err = con->rx_action(con); } while (!err); |
fdda387f7
|
1310 |
} |
6ed7257b4
|
1311 |
/* Send workqueue function */ |
1d6e8131c
|
1312 |
static void process_send_sockets(struct work_struct *work) |
fdda387f7
|
1313 |
{ |
1d6e8131c
|
1314 |
struct connection *con = container_of(work, struct connection, swork); |
fdda387f7
|
1315 |
|
1d6e8131c
|
1316 |
if (test_and_clear_bit(CF_CONNECT_PENDING, &con->flags)) { |
6ed7257b4
|
1317 |
con->connect_action(con); |
fdda387f7
|
1318 |
} |
bd44e2b00
|
1319 1320 |
clear_bit(CF_WRITE_PENDING, &con->flags); send_to_sock(con); |
fdda387f7
|
1321 1322 1323 1324 1325 1326 1327 |
} /* Discard all entries on the write queues */ static void clean_writequeues(void) { int nodeid; |
30d3a2373
|
1328 1329 |
for (nodeid = 1; nodeid <= max_nodeid; nodeid++) { struct connection *con = __nodeid2con(nodeid, 0); |
fdda387f7
|
1330 1331 1332 1333 1334 |
if (con) clean_one_writequeue(con); } } |
1d6e8131c
|
1335 |
static void work_stop(void) |
fdda387f7
|
1336 |
{ |
1d6e8131c
|
1337 1338 |
destroy_workqueue(recv_workqueue); destroy_workqueue(send_workqueue); |
fdda387f7
|
1339 |
} |
1d6e8131c
|
1340 |
static int work_start(void) |
fdda387f7
|
1341 |
{ |
fdda387f7
|
1342 |
int error; |
1d6e8131c
|
1343 1344 |
recv_workqueue = create_workqueue("dlm_recv"); error = IS_ERR(recv_workqueue); |
ac33d0710
|
1345 |
if (error) { |
1d6e8131c
|
1346 |
log_print("can't start dlm_recv %d", error); |
fdda387f7
|
1347 1348 |
return error; } |
fdda387f7
|
1349 |
|
1d6e8131c
|
1350 1351 |
send_workqueue = create_singlethread_workqueue("dlm_send"); error = IS_ERR(send_workqueue); |
ac33d0710
|
1352 |
if (error) { |
1d6e8131c
|
1353 1354 |
log_print("can't start dlm_send %d", error); destroy_workqueue(recv_workqueue); |
fdda387f7
|
1355 1356 |
return error; } |
fdda387f7
|
1357 1358 1359 |
return 0; } |
fdda387f7
|
1360 1361 1362 |
void dlm_lowcomms_stop(void) { int i; |
6ed7257b4
|
1363 |
struct connection *con; |
fdda387f7
|
1364 |
|
ac33d0710
|
1365 |
/* Set all the flags to prevent any |
fdda387f7
|
1366 1367 |
socket activity. */ |
7a936ce71
|
1368 |
mutex_lock(&connections_lock); |
30d3a2373
|
1369 |
for (i = 0; i <= max_nodeid; i++) { |
6ed7257b4
|
1370 |
con = __nodeid2con(i, 0); |
afb853fb4
|
1371 |
if (con) { |
9e5f2825a
|
1372 |
con->flags |= 0x0F; |
afb853fb4
|
1373 1374 1375 |
if (con->sock) con->sock->sk->sk_user_data = NULL; } |
fdda387f7
|
1376 |
} |
7a936ce71
|
1377 |
mutex_unlock(&connections_lock); |
ac33d0710
|
1378 |
|
1d6e8131c
|
1379 |
work_stop(); |
6ed7257b4
|
1380 |
|
7a936ce71
|
1381 |
mutex_lock(&connections_lock); |
fdda387f7
|
1382 |
clean_writequeues(); |
30d3a2373
|
1383 |
for (i = 0; i <= max_nodeid; i++) { |
2439fe507
|
1384 |
con = __nodeid2con(i, 0); |
6ed7257b4
|
1385 1386 |
if (con) { close_connection(con, true); |
39bd4177d
|
1387 1388 |
if (con->othercon) kmem_cache_free(con_cache, con->othercon); |
6ed7257b4
|
1389 |
kmem_cache_free(con_cache, con); |
fdda387f7
|
1390 1391 |
} } |
30d3a2373
|
1392 |
max_nodeid = 0; |
7a936ce71
|
1393 |
mutex_unlock(&connections_lock); |
fdda387f7
|
1394 |
kmem_cache_destroy(con_cache); |
30d3a2373
|
1395 |
idr_init(&connections_idr); |
fdda387f7
|
1396 |
} |
fdda387f7
|
1397 1398 |
int dlm_lowcomms_start(void) { |
6ed7257b4
|
1399 1400 |
int error = -EINVAL; struct connection *con; |
fdda387f7
|
1401 |
|
6ed7257b4
|
1402 1403 |
init_local(); if (!dlm_local_count) { |
617e82e10
|
1404 |
error = -ENOTCONN; |
fdda387f7
|
1405 |
log_print("no local IP address has been set"); |
6ed7257b4
|
1406 |
goto out; |
fdda387f7
|
1407 |
} |
6ed7257b4
|
1408 |
error = -ENOMEM; |
fdda387f7
|
1409 |
con_cache = kmem_cache_create("dlm_conn", sizeof(struct connection), |
ac33d0710
|
1410 |
__alignof__(struct connection), 0, |
20c2df83d
|
1411 |
NULL); |
fdda387f7
|
1412 |
if (!con_cache) |
6ed7257b4
|
1413 |
goto out; |
fdda387f7
|
1414 |
|
fdda387f7
|
1415 |
/* Start listening */ |
6ed7257b4
|
1416 1417 1418 1419 |
if (dlm_config.ci_protocol == 0) error = tcp_listen_for_all(); else error = sctp_listen_for_all(); |
fdda387f7
|
1420 1421 |
if (error) goto fail_unlisten; |
1d6e8131c
|
1422 |
error = work_start(); |
fdda387f7
|
1423 1424 |
if (error) goto fail_unlisten; |
fdda387f7
|
1425 |
return 0; |
ac33d0710
|
1426 |
fail_unlisten: |
6ed7257b4
|
1427 1428 1429 1430 1431 |
con = nodeid2con(0,0); if (con) { close_connection(con, false); kmem_cache_free(con_cache, con); } |
fdda387f7
|
1432 |
kmem_cache_destroy(con_cache); |
ac33d0710
|
1433 |
out: |
fdda387f7
|
1434 1435 |
return error; } |