projects / linux-2.6-block.git / blame
? search:
summary | shortlog | log | commit | commitdiff | tree
blob | blame (incremental) | history | HEAD
ocfs2/dlm: Track number of mles
[linux-2.6-block.git] / fs / ocfs2 / dlm / dlmmaster.c
CommitLineData
6714d8e8
KH		 1/* -*- mode: c; c-basic-offset: 8; -*-
2 * vim: noexpandtab sw=8 ts=8 sts=0:
3 *
4 * dlmmod.c
5 *
6 * standalone DLM module
7 *
8 * Copyright (C) 2004 Oracle. All rights reserved.
9 *
10 * This program is free software; you can redistribute it and/or
11 * modify it under the terms of the GNU General Public
12 * License as published by the Free Software Foundation; either
13 * version 2 of the License, or (at your option) any later version.
14 *
15 * This program is distributed in the hope that it will be useful,
16 * but WITHOUT ANY WARRANTY; without even the implied warranty of
17 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
18 * General Public License for more details.
19 *
20 * You should have received a copy of the GNU General Public
21 * License along with this program; if not, write to the
22 * Free Software Foundation, Inc., 59 Temple Place - Suite 330,
23 * Boston, MA 021110-1307, USA.
24 *
25 */
26
27
28#include <linux/module.h>
29#include <linux/fs.h>
30#include <linux/types.h>
31#include <linux/slab.h>
32#include <linux/highmem.h>
33#include <linux/utsname.h>
34#include <linux/init.h>
35#include <linux/sysctl.h>
36#include <linux/random.h>
37#include <linux/blkdev.h>
38#include <linux/socket.h>
39#include <linux/inet.h>
40#include <linux/spinlock.h>
41#include <linux/delay.h>
42
43
44#include "cluster/heartbeat.h"
45#include "cluster/nodemanager.h"
46#include "cluster/tcp.h"
47
48#include "dlmapi.h"
49#include "dlmcommon.h"
82353b59 50#include "dlmdomain.h"
e5a0334c 51#include "dlmdebug.h"
6714d8e8
KH		 52
53#define MLOG_MASK_PREFIX (ML_DLM|ML_DLM_MASTER)
54#include "cluster/masklog.h"
55
6714d8e8
KH		 56static void dlm_mle_node_down(struct dlm_ctxt *dlm,
57 struct dlm_master_list_entry *mle,
58 struct o2nm_node *node,
59 int idx);
60static void dlm_mle_node_up(struct dlm_ctxt *dlm,
61 struct dlm_master_list_entry *mle,
62 struct o2nm_node *node,
63 int idx);
64
65static void dlm_assert_master_worker(struct dlm_work_item *item, void *data);
ba2bf218
KH		 66static int dlm_do_assert_master(struct dlm_ctxt *dlm,
67 struct dlm_lock_resource *res,
68 void *nodemap, u32 flags);
f3f85464 69static void dlm_deref_lockres_worker(struct dlm_work_item *item, void *data);
6714d8e8 70
f77a9a78 71static inline void __dlm_mle_name(struct dlm_master_list_entry *mle,
2ed6c750
SM		 72 unsigned char **name, unsigned int *namelen,
73 unsigned int *namehash)
f77a9a78
SM		 74{
75 BUG_ON(mle->type != DLM_MLE_BLOCK &&
76 mle->type != DLM_MLE_MASTER &&
77 mle->type != DLM_MLE_MIGRATION);
78
79 if (mle->type != DLM_MLE_MASTER) {
80 *name = mle->u.mlename.name;
81 *namelen = mle->u.mlename.len;
2ed6c750
SM		 82 if (namehash)
83 *namehash = mle->u.mlename.hash;
f77a9a78
SM		 84 } else {
85 *name = (unsigned char *)mle->u.mleres->lockname.name;
86 *namelen = mle->u.mleres->lockname.len;
2ed6c750
SM		 87 if (namehash)
88 *namehash = mle->u.mleres->lockname.hash;
f77a9a78
SM		 89 }
90}
91
6714d8e8
KH		 92static inline int dlm_mle_equal(struct dlm_ctxt *dlm,
93 struct dlm_master_list_entry *mle,
94 const char *name,
95 unsigned int namelen)
96{
f77a9a78
SM		 97 unsigned char *mlename;
98 unsigned int mlelen;
6714d8e8
KH		 99
100 if (dlm != mle->dlm)
101 return 0;
102
2ed6c750 103 __dlm_mle_name(mle, &mlename, &mlelen, NULL);
f77a9a78
SM		 104
105 if (namelen != mlelen || memcmp(name, mlename, namelen) != 0)
106 return 0;
107
6714d8e8
KH		 108 return 1;
109}
110
724bdca9
SM		 111static struct kmem_cache *dlm_lockres_cache = NULL;
112static struct kmem_cache *dlm_lockname_cache = NULL;
e18b890b 113static struct kmem_cache *dlm_mle_cache = NULL;
6714d8e8 114
6714d8e8
KH		 115static void dlm_mle_release(struct kref *kref);
116static void dlm_init_mle(struct dlm_master_list_entry *mle,
117 enum dlm_mle_type type,
118 struct dlm_ctxt *dlm,
119 struct dlm_lock_resource *res,
120 const char *name,
121 unsigned int namelen);
122static void dlm_put_mle(struct dlm_master_list_entry *mle);
123static void __dlm_put_mle(struct dlm_master_list_entry *mle);
124static int dlm_find_mle(struct dlm_ctxt *dlm,
125 struct dlm_master_list_entry **mle,
126 char *name, unsigned int namelen);
127
ba2bf218
KH		 128static int dlm_do_master_request(struct dlm_lock_resource *res,
129 struct dlm_master_list_entry *mle, int to);
6714d8e8
KH		 130
131
132static int dlm_wait_for_lock_mastery(struct dlm_ctxt *dlm,
133 struct dlm_lock_resource *res,
134 struct dlm_master_list_entry *mle,
135 int *blocked);
136static int dlm_restart_lock_mastery(struct dlm_ctxt *dlm,
137 struct dlm_lock_resource *res,
138 struct dlm_master_list_entry *mle,
139 int blocked);
140static int dlm_add_migration_mle(struct dlm_ctxt *dlm,
141 struct dlm_lock_resource *res,
142 struct dlm_master_list_entry *mle,
143 struct dlm_master_list_entry **oldmle,
144 const char *name, unsigned int namelen,
145 u8 new_master, u8 master);
146
147static u8 dlm_pick_migration_target(struct dlm_ctxt *dlm,
148 struct dlm_lock_resource *res);
149static void dlm_remove_nonlocal_locks(struct dlm_ctxt *dlm,
150 struct dlm_lock_resource *res);
151static int dlm_mark_lockres_migrating(struct dlm_ctxt *dlm,
152 struct dlm_lock_resource *res,
153 u8 target);
c03872f5
KH		 154static int dlm_pre_master_reco_lockres(struct dlm_ctxt *dlm,
155 struct dlm_lock_resource *res);
6714d8e8
KH		 156
157
158int dlm_is_host_down(int errno)
159{
160 switch (errno) {
161 case -EBADF:
162 case -ECONNREFUSED:
163 case -ENOTCONN:
164 case -ECONNRESET:
165 case -EPIPE:
166 case -EHOSTDOWN:
167 case -EHOSTUNREACH:
168 case -ETIMEDOUT:
169 case -ECONNABORTED:
170 case -ENETDOWN:
171 case -ENETUNREACH:
172 case -ENETRESET:
173 case -ESHUTDOWN:
174 case -ENOPROTOOPT:
175 case -EINVAL: /* if returned from our tcp code,
176 this means there is no socket */
177 return 1;
178 }
179 return 0;
180}
181
182
183/*
184 * MASTER LIST FUNCTIONS
185 */
186
187
188/*
189 * regarding master list entries and heartbeat callbacks:
190 *
191 * in order to avoid sleeping and allocation that occurs in
192 * heartbeat, master list entries are simply attached to the
193 * dlm's established heartbeat callbacks. the mle is attached
194 * when it is created, and since the dlm->spinlock is held at
195 * that time, any heartbeat event will be properly discovered
196 * by the mle. the mle needs to be detached from the
197 * dlm->mle_hb_events list as soon as heartbeat events are no
198 * longer useful to the mle, and before the mle is freed.
199 *
200 * as a general rule, heartbeat events are no longer needed by
201 * the mle once an "answer" regarding the lock master has been
202 * received.
203 */
204static inline void __dlm_mle_attach_hb_events(struct dlm_ctxt *dlm,
205 struct dlm_master_list_entry *mle)
206{
207 assert_spin_locked(&dlm->spinlock);
208
209 list_add_tail(&mle->hb_events, &dlm->mle_hb_events);
210}
211
212
213static inline void __dlm_mle_detach_hb_events(struct dlm_ctxt *dlm,
214 struct dlm_master_list_entry *mle)
215{
216 if (!list_empty(&mle->hb_events))
217 list_del_init(&mle->hb_events);
218}
219
220
221static inline void dlm_mle_detach_hb_events(struct dlm_ctxt *dlm,
222 struct dlm_master_list_entry *mle)
223{
224 spin_lock(&dlm->spinlock);
225 __dlm_mle_detach_hb_events(dlm, mle);
226 spin_unlock(&dlm->spinlock);
227}
228
a2bf0477
KH		 229static void dlm_get_mle_inuse(struct dlm_master_list_entry *mle)
230{
231 struct dlm_ctxt *dlm;
232 dlm = mle->dlm;
233
234 assert_spin_locked(&dlm->spinlock);
235 assert_spin_locked(&dlm->master_lock);
236 mle->inuse++;
237 kref_get(&mle->mle_refs);
238}
239
240static void dlm_put_mle_inuse(struct dlm_master_list_entry *mle)
241{
242 struct dlm_ctxt *dlm;
243 dlm = mle->dlm;
244
245 spin_lock(&dlm->spinlock);
246 spin_lock(&dlm->master_lock);
247 mle->inuse--;
248 __dlm_put_mle(mle);
249 spin_unlock(&dlm->master_lock);
250 spin_unlock(&dlm->spinlock);
251
252}
253
6714d8e8
KH		 254/* remove from list and free */
255static void __dlm_put_mle(struct dlm_master_list_entry *mle)
256{
257 struct dlm_ctxt *dlm;
258 dlm = mle->dlm;
259
260 assert_spin_locked(&dlm->spinlock);
261 assert_spin_locked(&dlm->master_lock);
aa852354
KH		 262 if (!atomic_read(&mle->mle_refs.refcount)) {
263 /* this may or may not crash, but who cares.
264 * it's a BUG. */
265 mlog(ML_ERROR, "bad mle: %p\n", mle);
266 dlm_print_one_mle(mle);
267 BUG();
268 } else
269 kref_put(&mle->mle_refs, dlm_mle_release);
6714d8e8
KH		 270}
271
272
273/* must not have any spinlocks coming in */
274static void dlm_put_mle(struct dlm_master_list_entry *mle)
275{
276 struct dlm_ctxt *dlm;
277 dlm = mle->dlm;
278
279 spin_lock(&dlm->spinlock);
280 spin_lock(&dlm->master_lock);
281 __dlm_put_mle(mle);
282 spin_unlock(&dlm->master_lock);
283 spin_unlock(&dlm->spinlock);
284}
285
286static inline void dlm_get_mle(struct dlm_master_list_entry *mle)
287{
288 kref_get(&mle->mle_refs);
289}
290
291static void dlm_init_mle(struct dlm_master_list_entry *mle,
292 enum dlm_mle_type type,
293 struct dlm_ctxt *dlm,
294 struct dlm_lock_resource *res,
295 const char *name,
296 unsigned int namelen)
297{
298 assert_spin_locked(&dlm->spinlock);
299
300 mle->dlm = dlm;
301 mle->type = type;
2ed6c750 302 INIT_HLIST_NODE(&mle->master_hash_node);
6714d8e8
KH		 303 INIT_LIST_HEAD(&mle->hb_events);
304 memset(mle->maybe_map, 0, sizeof(mle->maybe_map));
305 spin_lock_init(&mle->spinlock);
306 init_waitqueue_head(&mle->wq);
307 atomic_set(&mle->woken, 0);
308 kref_init(&mle->mle_refs);
309 memset(mle->response_map, 0, sizeof(mle->response_map));
310 mle->master = O2NM_MAX_NODES;
311 mle->new_master = O2NM_MAX_NODES;
a2bf0477 312 mle->inuse = 0;
6714d8e8 313
f77a9a78
SM		 314 BUG_ON(mle->type != DLM_MLE_BLOCK &&
315 mle->type != DLM_MLE_MASTER &&
316 mle->type != DLM_MLE_MIGRATION);
317
6714d8e8
KH		 318 if (mle->type == DLM_MLE_MASTER) {
319 BUG_ON(!res);
f77a9a78
SM		 320 mle->u.mleres = res;
321 } else {
6714d8e8 322 BUG_ON(!name);
f77a9a78
SM		 323 memcpy(mle->u.mlename.name, name, namelen);
324 mle->u.mlename.len = namelen;
2ed6c750 325 mle->u.mlename.hash = dlm_lockid_hash(name, namelen);
6714d8e8
KH		 326 }
327
2041d8fd
SM		 328 atomic_inc(&dlm->mle_tot_count[mle->type]);
329 atomic_inc(&dlm->mle_cur_count[mle->type]);
330
6714d8e8
KH		 331 /* copy off the node_map and register hb callbacks on our copy */
332 memcpy(mle->node_map, dlm->domain_map, sizeof(mle->node_map));
333 memcpy(mle->vote_map, dlm->domain_map, sizeof(mle->vote_map));
334 clear_bit(dlm->node_num, mle->vote_map);
335 clear_bit(dlm->node_num, mle->node_map);
336
337 /* attach the mle to the domain node up/down events */
338 __dlm_mle_attach_hb_events(dlm, mle);
339}
340
1c084577
SM		 341void __dlm_unlink_mle(struct dlm_ctxt *dlm, struct dlm_master_list_entry *mle)
342{
343 assert_spin_locked(&dlm->spinlock);
344 assert_spin_locked(&dlm->master_lock);
345
2ed6c750
SM		 346 if (!hlist_unhashed(&mle->master_hash_node))
347 hlist_del_init(&mle->master_hash_node);
1c084577
SM		 348}
349
350void __dlm_insert_mle(struct dlm_ctxt *dlm, struct dlm_master_list_entry *mle)
351{
2ed6c750
SM		 352 struct hlist_head *bucket;
353 unsigned char *mname;
354 unsigned int mlen, hash;
355
1c084577
SM		 356 assert_spin_locked(&dlm->master_lock);
357
2ed6c750
SM		 358 __dlm_mle_name(mle, &mname, &mlen, &hash);
359 bucket = dlm_master_hash(dlm, hash);
360 hlist_add_head(&mle->master_hash_node, bucket);
1c084577 361}
6714d8e8
KH		 362
363/* returns 1 if found, 0 if not */
364static int dlm_find_mle(struct dlm_ctxt *dlm,
365 struct dlm_master_list_entry **mle,
366 char *name, unsigned int namelen)
367{
368 struct dlm_master_list_entry *tmpmle;
2ed6c750
SM		 369 struct hlist_head *bucket;
370 struct hlist_node *list;
371 unsigned int hash;
6714d8e8
KH		 372
373 assert_spin_locked(&dlm->master_lock);
374
2ed6c750
SM		 375 hash = dlm_lockid_hash(name, namelen);
376 bucket = dlm_master_hash(dlm, hash);
377 hlist_for_each(list, bucket) {
378 tmpmle = hlist_entry(list, struct dlm_master_list_entry,
379 master_hash_node);
6714d8e8
KH		 380 if (!dlm_mle_equal(dlm, tmpmle, name, namelen))
381 continue;
382 dlm_get_mle(tmpmle);
383 *mle = tmpmle;
384 return 1;
385 }
386 return 0;
387}
388
389void dlm_hb_event_notify_attached(struct dlm_ctxt *dlm, int idx, int node_up)
390{
391 struct dlm_master_list_entry *mle;
6714d8e8
KH		 392
393 assert_spin_locked(&dlm->spinlock);
394
800deef3 395 list_for_each_entry(mle, &dlm->mle_hb_events, hb_events) {
6714d8e8
KH		 396 if (node_up)
397 dlm_mle_node_up(dlm, mle, NULL, idx);
398 else
399 dlm_mle_node_down(dlm, mle, NULL, idx);
400 }
401}
402
403static void dlm_mle_node_down(struct dlm_ctxt *dlm,
404 struct dlm_master_list_entry *mle,
405 struct o2nm_node *node, int idx)
406{
407 spin_lock(&mle->spinlock);
408
409 if (!test_bit(idx, mle->node_map))
410 mlog(0, "node %u already removed from nodemap!\n", idx);
411 else
412 clear_bit(idx, mle->node_map);
413
414 spin_unlock(&mle->spinlock);
415}
416
417static void dlm_mle_node_up(struct dlm_ctxt *dlm,
418 struct dlm_master_list_entry *mle,
419 struct o2nm_node *node, int idx)
420{
421 spin_lock(&mle->spinlock);
422
423 if (test_bit(idx, mle->node_map))
424 mlog(0, "node %u already in node map!\n", idx);
425 else
426 set_bit(idx, mle->node_map);
427
428 spin_unlock(&mle->spinlock);
429}
430
431
432int dlm_init_mle_cache(void)
433{
12eb0035 434 dlm_mle_cache = kmem_cache_create("o2dlm_mle",
6714d8e8
KH		 435 sizeof(struct dlm_master_list_entry),
436 0, SLAB_HWCACHE_ALIGN,
20c2df83 437 NULL);
6714d8e8
KH		 438 if (dlm_mle_cache == NULL)
439 return -ENOMEM;
440 return 0;
441}
442
443void dlm_destroy_mle_cache(void)
444{
445 if (dlm_mle_cache)
446 kmem_cache_destroy(dlm_mle_cache);
447}
448
449static void dlm_mle_release(struct kref *kref)
450{
451 struct dlm_master_list_entry *mle;
452 struct dlm_ctxt *dlm;
2ed6c750
SM		 453 unsigned char *mname;
454 unsigned int mlen;
6714d8e8
KH		 455
456 mlog_entry_void();
457
458 mle = container_of(kref, struct dlm_master_list_entry, mle_refs);
459 dlm = mle->dlm;
460
6714d8e8
KH		 461 assert_spin_locked(&dlm->spinlock);
462 assert_spin_locked(&dlm->master_lock);
463
2ed6c750
SM		 464 __dlm_mle_name(mle, &mname, &mlen, NULL);
465 mlog(0, "Releasing mle for %.*s, type %d\n", mlen, mname, mle->type);
466
6714d8e8 467 /* remove from list if not already */
1c084577 468 __dlm_unlink_mle(dlm, mle);
6714d8e8
KH		 469
470 /* detach the mle from the domain node up/down events */
471 __dlm_mle_detach_hb_events(dlm, mle);
472
2041d8fd
SM		 473 atomic_dec(&dlm->mle_cur_count[mle->type]);
474
6714d8e8
KH		 475 /* NOTE: kfree under spinlock here.
476 * if this is bad, we can move this to a freelist. */
477 kmem_cache_free(dlm_mle_cache, mle);
478}
479
480
481/*
482 * LOCK RESOURCE FUNCTIONS
483 */
484
724bdca9
SM		 485int dlm_init_master_caches(void)
486{
487 dlm_lockres_cache = kmem_cache_create("o2dlm_lockres",
488 sizeof(struct dlm_lock_resource),
489 0, SLAB_HWCACHE_ALIGN, NULL);
490 if (!dlm_lockres_cache)
491 goto bail;
492
493 dlm_lockname_cache = kmem_cache_create("o2dlm_lockname",
494 DLM_LOCKID_NAME_MAX, 0,
495 SLAB_HWCACHE_ALIGN, NULL);
496 if (!dlm_lockname_cache)
497 goto bail;
498
499 return 0;
500bail:
501 dlm_destroy_master_caches();
502 return -ENOMEM;
503}
504
505void dlm_destroy_master_caches(void)
506{
507 if (dlm_lockname_cache)
508 kmem_cache_destroy(dlm_lockname_cache);
509
510 if (dlm_lockres_cache)
511 kmem_cache_destroy(dlm_lockres_cache);
512}
513
6714d8e8
KH		 514static void dlm_set_lockres_owner(struct dlm_ctxt *dlm,
515 struct dlm_lock_resource *res,
516 u8 owner)
517{
518 assert_spin_locked(&res->spinlock);
519
520 mlog_entry("%.*s, %u\n", res->lockname.len, res->lockname.name, owner);
521
522 if (owner == dlm->node_num)
523 atomic_inc(&dlm->local_resources);
524 else if (owner == DLM_LOCK_RES_OWNER_UNKNOWN)
525 atomic_inc(&dlm->unknown_resources);
526 else
527 atomic_inc(&dlm->remote_resources);
528
529 res->owner = owner;
530}
531
532void dlm_change_lockres_owner(struct dlm_ctxt *dlm,
533 struct dlm_lock_resource *res, u8 owner)
534{
535 assert_spin_locked(&res->spinlock);
536
537 if (owner == res->owner)
538 return;
539
540 if (res->owner == dlm->node_num)
541 atomic_dec(&dlm->local_resources);
542 else if (res->owner == DLM_LOCK_RES_OWNER_UNKNOWN)
543 atomic_dec(&dlm->unknown_resources);
544 else
545 atomic_dec(&dlm->remote_resources);
546
547 dlm_set_lockres_owner(dlm, res, owner);
548}
549
550
551static void dlm_lockres_release(struct kref *kref)
552{
553 struct dlm_lock_resource *res;
b0d4f817 554 struct dlm_ctxt *dlm;
6714d8e8
KH		 555
556 res = container_of(kref, struct dlm_lock_resource, refs);
b0d4f817 557 dlm = res->dlm;
6714d8e8
KH		 558
559 /* This should not happen -- all lockres' have a name
560 * associated with them at init time. */
561 BUG_ON(!res->lockname.name);
562
563 mlog(0, "destroying lockres %.*s\n", res->lockname.len,
564 res->lockname.name);
565
b0d4f817 566 spin_lock(&dlm->track_lock);
29576f8b
SM		 567 if (!list_empty(&res->tracking))
568 list_del_init(&res->tracking);
569 else {
570 mlog(ML_ERROR, "Resource %.*s not on the Tracking list\n",
571 res->lockname.len, res->lockname.name);
572 dlm_print_one_lock_resource(res);
573 }
b0d4f817
SM		 574 spin_unlock(&dlm->track_lock);
575
576 dlm_put(dlm);
29576f8b 577
a7f90d83
KH		 578 if (!hlist_unhashed(&res->hash_node) ||
579 !list_empty(&res->granted) ||
580 !list_empty(&res->converting) ||
581 !list_empty(&res->blocked) ||
582 !list_empty(&res->dirty) ||
583 !list_empty(&res->recovering) ||
584 !list_empty(&res->purge)) {
585 mlog(ML_ERROR,
586 "Going to BUG for resource %.*s."
587 " We're on a list! [%c%c%c%c%c%c%c]\n",
588 res->lockname.len, res->lockname.name,
589 !hlist_unhashed(&res->hash_node) ? 'H' : ' ',
590 !list_empty(&res->granted) ? 'G' : ' ',
591 !list_empty(&res->converting) ? 'C' : ' ',
592 !list_empty(&res->blocked) ? 'B' : ' ',
593 !list_empty(&res->dirty) ? 'D' : ' ',
594 !list_empty(&res->recovering) ? 'R' : ' ',
595 !list_empty(&res->purge) ? 'P' : ' ');
596
597 dlm_print_one_lock_resource(res);
598 }
599
6714d8e8
KH		 600 /* By the time we're ready to blow this guy away, we shouldn't
601 * be on any lists. */
81f2094a 602 BUG_ON(!hlist_unhashed(&res->hash_node));
6714d8e8
KH		 603 BUG_ON(!list_empty(&res->granted));
604 BUG_ON(!list_empty(&res->converting));
605 BUG_ON(!list_empty(&res->blocked));
606 BUG_ON(!list_empty(&res->dirty));
607 BUG_ON(!list_empty(&res->recovering));
608 BUG_ON(!list_empty(&res->purge));
609
724bdca9 610 kmem_cache_free(dlm_lockname_cache, (void *)res->lockname.name);
6714d8e8 611
724bdca9 612 kmem_cache_free(dlm_lockres_cache, res);
6714d8e8
KH		 613}
614
6714d8e8
KH		 615void dlm_lockres_put(struct dlm_lock_resource *res)
616{
617 kref_put(&res->refs, dlm_lockres_release);
618}
619
620static void dlm_init_lockres(struct dlm_ctxt *dlm,
621 struct dlm_lock_resource *res,
622 const char *name, unsigned int namelen)
623{
624 char *qname;
625
626 /* If we memset here, we lose our reference to the kmalloc'd
627 * res->lockname.name, so be sure to init every field
628 * correctly! */
629
630 qname = (char *) res->lockname.name;
631 memcpy(qname, name, namelen);
632
633 res->lockname.len = namelen;
a3d33291 634 res->lockname.hash = dlm_lockid_hash(name, namelen);
6714d8e8
KH		 635
636 init_waitqueue_head(&res->wq);
637 spin_lock_init(&res->spinlock);
81f2094a 638 INIT_HLIST_NODE(&res->hash_node);
6714d8e8
KH		 639 INIT_LIST_HEAD(&res->granted);
640 INIT_LIST_HEAD(&res->converting);
641 INIT_LIST_HEAD(&res->blocked);
642 INIT_LIST_HEAD(&res->dirty);
643 INIT_LIST_HEAD(&res->recovering);
644 INIT_LIST_HEAD(&res->purge);
29576f8b 645 INIT_LIST_HEAD(&res->tracking);
6714d8e8
KH		 646 atomic_set(&res->asts_reserved, 0);
647 res->migration_pending = 0;
ba2bf218 648 res->inflight_locks = 0;
6714d8e8 649
b0d4f817
SM		 650 /* put in dlm_lockres_release */
651 dlm_grab(dlm);
652 res->dlm = dlm;
653
6714d8e8
KH		 654 kref_init(&res->refs);
655
656 /* just for consistency */
657 spin_lock(&res->spinlock);
658 dlm_set_lockres_owner(dlm, res, DLM_LOCK_RES_OWNER_UNKNOWN);
659 spin_unlock(&res->spinlock);
660
661 res->state = DLM_LOCK_RES_IN_PROGRESS;
662
663 res->last_used = 0;
664
18c6ac38 665 spin_lock(&dlm->spinlock);
29576f8b 666 list_add_tail(&res->tracking, &dlm->tracking_list);
18c6ac38 667 spin_unlock(&dlm->spinlock);
29576f8b 668
6714d8e8 669 memset(res->lvb, 0, DLM_LVB_LEN);
ba2bf218 670 memset(res->refmap, 0, sizeof(res->refmap));
6714d8e8
KH		 671}
672
673struct dlm_lock_resource *dlm_new_lockres(struct dlm_ctxt *dlm,
674 const char *name,
675 unsigned int namelen)
676{
724bdca9 677 struct dlm_lock_resource *res = NULL;
6714d8e8 678
724bdca9
SM		 679 res = (struct dlm_lock_resource *)
680 kmem_cache_zalloc(dlm_lockres_cache, GFP_NOFS);
6714d8e8 681 if (!res)
724bdca9 682 goto error;
6714d8e8 683
724bdca9
SM		 684 res->lockname.name = (char *)
685 kmem_cache_zalloc(dlm_lockname_cache, GFP_NOFS);
686 if (!res->lockname.name)
687 goto error;
6714d8e8
KH		 688
689 dlm_init_lockres(dlm, res, name, namelen);
690 return res;
724bdca9
SM		 691
692error:
693 if (res && res->lockname.name)
694 kmem_cache_free(dlm_lockname_cache, (void *)res->lockname.name);
695
696 if (res)
697 kmem_cache_free(dlm_lockres_cache, res);
698 return NULL;
6714d8e8
KH		 699}
700
ba2bf218
KH		 701void __dlm_lockres_grab_inflight_ref(struct dlm_ctxt *dlm,
702 struct dlm_lock_resource *res,
703 int new_lockres,
704 const char *file,
705 int line)
706{
707 if (!new_lockres)
708 assert_spin_locked(&res->spinlock);
709
710 if (!test_bit(dlm->node_num, res->refmap)) {
711 BUG_ON(res->inflight_locks != 0);
712 dlm_lockres_set_refmap_bit(dlm->node_num, res);
713 }
714 res->inflight_locks++;
715 mlog(0, "%s:%.*s: inflight++: now %u\n",
716 dlm->name, res->lockname.len, res->lockname.name,
717 res->inflight_locks);
718}
719
720void __dlm_lockres_drop_inflight_ref(struct dlm_ctxt *dlm,
721 struct dlm_lock_resource *res,
722 const char *file,
723 int line)
724{
725 assert_spin_locked(&res->spinlock);
726
727 BUG_ON(res->inflight_locks == 0);
728 res->inflight_locks--;
729 mlog(0, "%s:%.*s: inflight--: now %u\n",
730 dlm->name, res->lockname.len, res->lockname.name,
731 res->inflight_locks);
732 if (res->inflight_locks == 0)
733 dlm_lockres_clear_refmap_bit(dlm->node_num, res);
734 wake_up(&res->wq);
735}
736
6714d8e8
KH		 737/*
738 * lookup a lock resource by name.
739 * may already exist in the hashtable.
740 * lockid is null terminated
741 *
742 * if not, allocate enough for the lockres and for
743 * the temporary structure used in doing the mastering.
744 *
745 * also, do a lookup in the dlm->master_list to see
746 * if another node has begun mastering the same lock.
747 * if so, there should be a block entry in there
748 * for this name, and we should *not* attempt to master
749 * the lock here. need to wait around for that node
750 * to assert_master (or die).
751 *
752 */
753struct dlm_lock_resource * dlm_get_lock_resource(struct dlm_ctxt *dlm,
754 const char *lockid,
3384f3df 755 int namelen,
6714d8e8
KH		 756 int flags)
757{
758 struct dlm_lock_resource *tmpres=NULL, *res=NULL;
759 struct dlm_master_list_entry *mle = NULL;
760 struct dlm_master_list_entry *alloc_mle = NULL;
761 int blocked = 0;
762 int ret, nodenum;
763 struct dlm_node_iter iter;
3384f3df 764 unsigned int hash;
6714d8e8 765 int tries = 0;
c03872f5 766 int bit, wait_on_recovery = 0;
ba2bf218 767 int drop_inflight_if_nonlocal = 0;
6714d8e8
KH		 768
769 BUG_ON(!lockid);
770
a3d33291 771 hash = dlm_lockid_hash(lockid, namelen);
6714d8e8
KH		 772
773 mlog(0, "get lockres %s (len %d)\n", lockid, namelen);
774
775lookup:
776 spin_lock(&dlm->spinlock);
ba2bf218 777 tmpres = __dlm_lookup_lockres_full(dlm, lockid, namelen, hash);
6714d8e8 778 if (tmpres) {
ba2bf218
KH		 779 int dropping_ref = 0;
780
7b791d68
SM		 781 spin_unlock(&dlm->spinlock);
782
ba2bf218 783 spin_lock(&tmpres->spinlock);
7b791d68
SM		 784 /* We wait for the other thread that is mastering the resource */
785 if (tmpres->owner == DLM_LOCK_RES_OWNER_UNKNOWN) {
786 __dlm_wait_on_lockres(tmpres);
787 BUG_ON(tmpres->owner == DLM_LOCK_RES_OWNER_UNKNOWN);
788 }
789
ba2bf218
KH		 790 if (tmpres->owner == dlm->node_num) {
791 BUG_ON(tmpres->state & DLM_LOCK_RES_DROPPING_REF);
792 dlm_lockres_grab_inflight_ref(dlm, tmpres);
793 } else if (tmpres->state & DLM_LOCK_RES_DROPPING_REF)
794 dropping_ref = 1;
795 spin_unlock(&tmpres->spinlock);
ba2bf218
KH		 796
797 /* wait until done messaging the master, drop our ref to allow
798 * the lockres to be purged, start over. */
799 if (dropping_ref) {
800 spin_lock(&tmpres->spinlock);
801 __dlm_wait_on_lockres_flags(tmpres, DLM_LOCK_RES_DROPPING_REF);
802 spin_unlock(&tmpres->spinlock);
803 dlm_lockres_put(tmpres);
804 tmpres = NULL;
805 goto lookup;
806 }
807
6714d8e8
KH		 808 mlog(0, "found in hash!\n");
809 if (res)
810 dlm_lockres_put(res);
811 res = tmpres;
812 goto leave;
813 }
814
815 if (!res) {
816 spin_unlock(&dlm->spinlock);
817 mlog(0, "allocating a new resource\n");
818 /* nothing found and we need to allocate one. */
819 alloc_mle = (struct dlm_master_list_entry *)
ad8100e0 820 kmem_cache_alloc(dlm_mle_cache, GFP_NOFS);
6714d8e8
KH		 821 if (!alloc_mle)
822 goto leave;
823 res = dlm_new_lockres(dlm, lockid, namelen);
824 if (!res)
825 goto leave;
826 goto lookup;
827 }
828
829 mlog(0, "no lockres found, allocated our own: %p\n", res);
830
831 if (flags & LKM_LOCAL) {
832 /* caller knows it's safe to assume it's not mastered elsewhere
833 * DONE! return right away */
834 spin_lock(&res->spinlock);
835 dlm_change_lockres_owner(dlm, res, dlm->node_num);
836 __dlm_insert_lockres(dlm, res);
ba2bf218 837 dlm_lockres_grab_inflight_ref(dlm, res);
6714d8e8
KH		 838 spin_unlock(&res->spinlock);
839 spin_unlock(&dlm->spinlock);
840 /* lockres still marked IN_PROGRESS */
841 goto wake_waiters;
842 }
843
844 /* check master list to see if another node has started mastering it */
845 spin_lock(&dlm->master_lock);
846
847 /* if we found a block, wait for lock to be mastered by another node */
848 blocked = dlm_find_mle(dlm, &mle, (char *)lockid, namelen);
849 if (blocked) {
ba2bf218 850 int mig;
6714d8e8
KH		 851 if (mle->type == DLM_MLE_MASTER) {
852 mlog(ML_ERROR, "master entry for nonexistent lock!\n");
853 BUG();
ba2bf218
KH		 854 }
855 mig = (mle->type == DLM_MLE_MIGRATION);
856 /* if there is a migration in progress, let the migration
857 * finish before continuing. we can wait for the absence
858 * of the MIGRATION mle: either the migrate finished or
859 * one of the nodes died and the mle was cleaned up.
860 * if there is a BLOCK here, but it already has a master
861 * set, we are too late. the master does not have a ref
862 * for us in the refmap. detach the mle and drop it.
863 * either way, go back to the top and start over. */
864 if (mig || mle->master != O2NM_MAX_NODES) {
865 BUG_ON(mig && mle->master == dlm->node_num);
866 /* we arrived too late. the master does not
867 * have a ref for us. retry. */
868 mlog(0, "%s:%.*s: late on %s\n",
869 dlm->name, namelen, lockid,
870 mig ? "MIGRATION" : "BLOCK");
6714d8e8 871 spin_unlock(&dlm->master_lock);
6714d8e8
KH		 872 spin_unlock(&dlm->spinlock);
873
874 /* master is known, detach */
ba2bf218
KH		 875 if (!mig)
876 dlm_mle_detach_hb_events(dlm, mle);
6714d8e8
KH		 877 dlm_put_mle(mle);
878 mle = NULL;
ba2bf218
KH		 879 /* this is lame, but we cant wait on either
880 * the mle or lockres waitqueue here */
881 if (mig)
882 msleep(100);
883 goto lookup;
6714d8e8
KH		 884 }
885 } else {
886 /* go ahead and try to master lock on this node */
887 mle = alloc_mle;
888 /* make sure this does not get freed below */
889 alloc_mle = NULL;
890 dlm_init_mle(mle, DLM_MLE_MASTER, dlm, res, NULL, 0);
891 set_bit(dlm->node_num, mle->maybe_map);
1c084577 892 __dlm_insert_mle(dlm, mle);
c03872f5
KH		 893
894 /* still holding the dlm spinlock, check the recovery map
895 * to see if there are any nodes that still need to be
896 * considered. these will not appear in the mle nodemap
897 * but they might own this lockres. wait on them. */
898 bit = find_next_bit(dlm->recovery_map, O2NM_MAX_NODES, 0);
899 if (bit < O2NM_MAX_NODES) {
2759236f 900 mlog(ML_NOTICE, "%s:%.*s: at least one node (%d) to "
c03872f5
KH		 901 "recover before lock mastery can begin\n",
902 dlm->name, namelen, (char *)lockid, bit);
903 wait_on_recovery = 1;
904 }
6714d8e8
KH		 905 }
906
907 /* at this point there is either a DLM_MLE_BLOCK or a
908 * DLM_MLE_MASTER on the master list, so it's safe to add the
909 * lockres to the hashtable. anyone who finds the lock will
910 * still have to wait on the IN_PROGRESS. */
911
912 /* finally add the lockres to its hash bucket */
913 __dlm_insert_lockres(dlm, res);
ba2bf218
KH		 914 /* since this lockres is new it doesnt not require the spinlock */
915 dlm_lockres_grab_inflight_ref_new(dlm, res);
916
917 /* if this node does not become the master make sure to drop
918 * this inflight reference below */
919 drop_inflight_if_nonlocal = 1;
920
6714d8e8
KH		 921 /* get an extra ref on the mle in case this is a BLOCK
922 * if so, the creator of the BLOCK may try to put the last
923 * ref at this time in the assert master handler, so we
924 * need an extra one to keep from a bad ptr deref. */
a2bf0477 925 dlm_get_mle_inuse(mle);
6714d8e8
KH		 926 spin_unlock(&dlm->master_lock);
927 spin_unlock(&dlm->spinlock);
928
e7e69eb3 929redo_request:
c03872f5
KH		 930 while (wait_on_recovery) {
931 /* any cluster changes that occurred after dropping the
932 * dlm spinlock would be detectable be a change on the mle,
933 * so we only need to clear out the recovery map once. */
934 if (dlm_is_recovery_lock(lockid, namelen)) {
935 mlog(ML_NOTICE, "%s: recovery map is not empty, but "
936 "must master $RECOVERY lock now\n", dlm->name);
937 if (!dlm_pre_master_reco_lockres(dlm, res))
938 wait_on_recovery = 0;
939 else {
940 mlog(0, "%s: waiting 500ms for heartbeat state "
941 "change\n", dlm->name);
942 msleep(500);
943 }
944 continue;
945 }
946
947 dlm_kick_recovery_thread(dlm);
aa087b84 948 msleep(1000);
c03872f5
KH		 949 dlm_wait_for_recovery(dlm);
950
951 spin_lock(&dlm->spinlock);
952 bit = find_next_bit(dlm->recovery_map, O2NM_MAX_NODES, 0);
953 if (bit < O2NM_MAX_NODES) {
2759236f 954 mlog(ML_NOTICE, "%s:%.*s: at least one node (%d) to "
c03872f5
KH		 955 "recover before lock mastery can begin\n",
956 dlm->name, namelen, (char *)lockid, bit);
957 wait_on_recovery = 1;
958 } else
959 wait_on_recovery = 0;
960 spin_unlock(&dlm->spinlock);
b7084ab5
KH		 961
962 if (wait_on_recovery)
963 dlm_wait_for_node_recovery(dlm, bit, 10000);
c03872f5
KH		 964 }
965
6714d8e8
KH		 966 /* must wait for lock to be mastered elsewhere */
967 if (blocked)
968 goto wait;
969
6714d8e8
KH		 970 ret = -EINVAL;
971 dlm_node_iter_init(mle->vote_map, &iter);
972 while ((nodenum = dlm_node_iter_next(&iter)) >= 0) {
ba2bf218 973 ret = dlm_do_master_request(res, mle, nodenum);
6714d8e8
KH		 974 if (ret < 0)
975 mlog_errno(ret);
976 if (mle->master != O2NM_MAX_NODES) {
977 /* found a master ! */
9c6510a5
KH		 978 if (mle->master <= nodenum)
979 break;
980 /* if our master request has not reached the master
981 * yet, keep going until it does. this is how the
982 * master will know that asserts are needed back to
983 * the lower nodes. */
984 mlog(0, "%s:%.*s: requests only up to %u but master "
985 "is %u, keep going\n", dlm->name, namelen,
986 lockid, nodenum, mle->master);
6714d8e8
KH		 987 }
988 }
989
990wait:
991 /* keep going until the response map includes all nodes */
992 ret = dlm_wait_for_lock_mastery(dlm, res, mle, &blocked);
993 if (ret < 0) {
e7e69eb3 994 wait_on_recovery = 1;
6714d8e8
KH		 995 mlog(0, "%s:%.*s: node map changed, redo the "
996 "master request now, blocked=%d\n",
997 dlm->name, res->lockname.len,
998 res->lockname.name, blocked);
999 if (++tries > 20) {
1000 mlog(ML_ERROR, "%s:%.*s: spinning on "
1001 "dlm_wait_for_lock_mastery, blocked=%d\n",
1002 dlm->name, res->lockname.len,
1003 res->lockname.name, blocked);
1004 dlm_print_one_lock_resource(res);
8a9343fa 1005 dlm_print_one_mle(mle);
6714d8e8
KH		 1006 tries = 0;
1007 }
1008 goto redo_request;
1009 }
1010
1011 mlog(0, "lockres mastered by %u\n", res->owner);
1012 /* make sure we never continue without this */
1013 BUG_ON(res->owner == O2NM_MAX_NODES);
1014
1015 /* master is known, detach if not already detached */
1016 dlm_mle_detach_hb_events(dlm, mle);
1017 dlm_put_mle(mle);
1018 /* put the extra ref */
a2bf0477 1019 dlm_put_mle_inuse(mle);
6714d8e8
KH		 1020
1021wake_waiters:
1022 spin_lock(&res->spinlock);
ba2bf218
KH		 1023 if (res->owner != dlm->node_num && drop_inflight_if_nonlocal)
1024 dlm_lockres_drop_inflight_ref(dlm, res);
6714d8e8
KH		 1025 res->state &= ~DLM_LOCK_RES_IN_PROGRESS;
1026 spin_unlock(&res->spinlock);
1027 wake_up(&res->wq);
1028
1029leave:
1030 /* need to free the unused mle */
1031 if (alloc_mle)
1032 kmem_cache_free(dlm_mle_cache, alloc_mle);
1033
1034 return res;
1035}
1036
1037
1038#define DLM_MASTERY_TIMEOUT_MS 5000
1039
1040static int dlm_wait_for_lock_mastery(struct dlm_ctxt *dlm,
1041 struct dlm_lock_resource *res,
1042 struct dlm_master_list_entry *mle,
1043 int *blocked)
1044{
1045 u8 m;
1046 int ret, bit;
1047 int map_changed, voting_done;
1048 int assert, sleep;
1049
1050recheck:
1051 ret = 0;
1052 assert = 0;
1053
1054 /* check if another node has already become the owner */
1055 spin_lock(&res->spinlock);
1056 if (res->owner != DLM_LOCK_RES_OWNER_UNKNOWN) {
9c6510a5
KH		 1057 mlog(0, "%s:%.*s: owner is suddenly %u\n", dlm->name,
1058 res->lockname.len, res->lockname.name, res->owner);
6714d8e8 1059 spin_unlock(&res->spinlock);
9c6510a5
KH		 1060 /* this will cause the master to re-assert across
1061 * the whole cluster, freeing up mles */
588e0090 1062 if (res->owner != dlm->node_num) {
ba2bf218 1063 ret = dlm_do_master_request(res, mle, res->owner);
588e0090
KH		 1064 if (ret < 0) {
1065 /* give recovery a chance to run */
1066 mlog(ML_ERROR, "link to %u went down?: %d\n", res->owner, ret);
1067 msleep(500);
1068 goto recheck;
1069 }
9c6510a5
KH		 1070 }
1071 ret = 0;
6714d8e8
KH		 1072 goto leave;
1073 }
1074 spin_unlock(&res->spinlock);
1075
1076 spin_lock(&mle->spinlock);
1077 m = mle->master;
1078 map_changed = (memcmp(mle->vote_map, mle->node_map,
1079 sizeof(mle->vote_map)) != 0);
1080 voting_done = (memcmp(mle->vote_map, mle->response_map,
1081 sizeof(mle->vote_map)) == 0);
1082
1083 /* restart if we hit any errors */
1084 if (map_changed) {
1085 int b;
1086 mlog(0, "%s: %.*s: node map changed, restarting\n",
1087 dlm->name, res->lockname.len, res->lockname.name);
1088 ret = dlm_restart_lock_mastery(dlm, res, mle, *blocked);
1089 b = (mle->type == DLM_MLE_BLOCK);
1090 if ((*blocked && !b) || (!*blocked && b)) {
1091 mlog(0, "%s:%.*s: status change: old=%d new=%d\n",
1092 dlm->name, res->lockname.len, res->lockname.name,
1093 *blocked, b);
1094 *blocked = b;
1095 }
1096 spin_unlock(&mle->spinlock);
1097 if (ret < 0) {
1098 mlog_errno(ret);
1099 goto leave;
1100 }
1101 mlog(0, "%s:%.*s: restart lock mastery succeeded, "
1102 "rechecking now\n", dlm->name, res->lockname.len,
1103 res->lockname.name);
1104 goto recheck;
aa852354
KH		 1105 } else {
1106 if (!voting_done) {
1107 mlog(0, "map not changed and voting not done "
1108 "for %s:%.*s\n", dlm->name, res->lockname.len,
1109 res->lockname.name);
1110 }
6714d8e8
KH		 1111 }
1112
1113 if (m != O2NM_MAX_NODES) {
1114 /* another node has done an assert!
1115 * all done! */
1116 sleep = 0;
1117 } else {
1118 sleep = 1;
1119 /* have all nodes responded? */
1120 if (voting_done && !*blocked) {
1121 bit = find_next_bit(mle->maybe_map, O2NM_MAX_NODES, 0);
1122 if (dlm->node_num <= bit) {
1123 /* my node number is lowest.
1124 * now tell other nodes that I am
1125 * mastering this. */
1126 mle->master = dlm->node_num;
ba2bf218
KH		 1127 /* ref was grabbed in get_lock_resource
1128 * will be dropped in dlmlock_master */
6714d8e8
KH		 1129 assert = 1;
1130 sleep = 0;
1131 }
1132 /* if voting is done, but we have not received
1133 * an assert master yet, we must sleep */
1134 }
1135 }
1136
1137 spin_unlock(&mle->spinlock);
1138
1139 /* sleep if we haven't finished voting yet */
1140 if (sleep) {
1141 unsigned long timeo = msecs_to_jiffies(DLM_MASTERY_TIMEOUT_MS);
1142
1143 /*
1144 if (atomic_read(&mle->mle_refs.refcount) < 2)
1145 mlog(ML_ERROR, "mle (%p) refs=%d, name=%.*s\n", mle,
1146 atomic_read(&mle->mle_refs.refcount),
1147 res->lockname.len, res->lockname.name);
1148 */
1149 atomic_set(&mle->woken, 0);
1150 (void)wait_event_timeout(mle->wq,
1151 (atomic_read(&mle->woken) == 1),
1152 timeo);
1153 if (res->owner == O2NM_MAX_NODES) {
ba2bf218
KH		 1154 mlog(0, "%s:%.*s: waiting again\n", dlm->name,
1155 res->lockname.len, res->lockname.name);
6714d8e8
KH		 1156 goto recheck;
1157 }
1158 mlog(0, "done waiting, master is %u\n", res->owner);
1159 ret = 0;
1160 goto leave;
1161 }
1162
1163 ret = 0; /* done */
1164 if (assert) {
1165 m = dlm->node_num;
1166 mlog(0, "about to master %.*s here, this=%u\n",
1167 res->lockname.len, res->lockname.name, m);
ba2bf218 1168 ret = dlm_do_assert_master(dlm, res, mle->vote_map, 0);
6714d8e8
KH		 1169 if (ret) {
1170 /* This is a failure in the network path,
1171 * not in the response to the assert_master
1172 * (any nonzero response is a BUG on this node).
1173 * Most likely a socket just got disconnected
1174 * due to node death. */
1175 mlog_errno(ret);
1176 }
1177 /* no longer need to restart lock mastery.
1178 * all living nodes have been contacted. */
1179 ret = 0;
1180 }
1181
1182 /* set the lockres owner */
1183 spin_lock(&res->spinlock);
ba2bf218
KH		 1184 /* mastery reference obtained either during
1185 * assert_master_handler or in get_lock_resource */
6714d8e8
KH		 1186 dlm_change_lockres_owner(dlm, res, m);
1187 spin_unlock(&res->spinlock);
1188
1189leave:
1190 return ret;
1191}
1192
1193struct dlm_bitmap_diff_iter
1194{
1195 int curnode;
1196 unsigned long *orig_bm;
1197 unsigned long *cur_bm;
1198 unsigned long diff_bm[BITS_TO_LONGS(O2NM_MAX_NODES)];
1199};
1200
1201enum dlm_node_state_change
1202{
1203 NODE_DOWN = -1,
1204 NODE_NO_CHANGE = 0,
1205 NODE_UP
1206};
1207
1208static void dlm_bitmap_diff_iter_init(struct dlm_bitmap_diff_iter *iter,
1209 unsigned long *orig_bm,
1210 unsigned long *cur_bm)
1211{
1212 unsigned long p1, p2;
1213 int i;
1214
1215 iter->curnode = -1;
1216 iter->orig_bm = orig_bm;
1217 iter->cur_bm = cur_bm;
1218
1219 for (i = 0; i < BITS_TO_LONGS(O2NM_MAX_NODES); i++) {
1220 p1 = *(iter->orig_bm + i);
1221 p2 = *(iter->cur_bm + i);
1222 iter->diff_bm[i] = (p1 & ~p2) | (p2 & ~p1);
1223 }
1224}
1225
1226static int dlm_bitmap_diff_iter_next(struct dlm_bitmap_diff_iter *iter,
1227 enum dlm_node_state_change *state)
1228{
1229 int bit;
1230
1231 if (iter->curnode >= O2NM_MAX_NODES)
1232 return -ENOENT;
1233
1234 bit = find_next_bit(iter->diff_bm, O2NM_MAX_NODES,
1235 iter->curnode+1);
1236 if (bit >= O2NM_MAX_NODES) {
1237 iter->curnode = O2NM_MAX_NODES;
1238 return -ENOENT;
1239 }
1240
1241 /* if it was there in the original then this node died */
1242 if (test_bit(bit, iter->orig_bm))
1243 *state = NODE_DOWN;
1244 else
1245 *state = NODE_UP;
1246
1247 iter->curnode = bit;
1248 return bit;
1249}
1250
1251
1252static int dlm_restart_lock_mastery(struct dlm_ctxt *dlm,
1253 struct dlm_lock_resource *res,
1254 struct dlm_master_list_entry *mle,
1255 int blocked)
1256{
1257 struct dlm_bitmap_diff_iter bdi;
1258 enum dlm_node_state_change sc;
1259 int node;
1260 int ret = 0;
1261
1262 mlog(0, "something happened such that the "
1263 "master process may need to be restarted!\n");
1264
1265 assert_spin_locked(&mle->spinlock);
1266
1267 dlm_bitmap_diff_iter_init(&bdi, mle->vote_map, mle->node_map);
1268 node = dlm_bitmap_diff_iter_next(&bdi, &sc);
1269 while (node >= 0) {
1270 if (sc == NODE_UP) {
e2faea4c
KH		 1271 /* a node came up. clear any old vote from
1272 * the response map and set it in the vote map
1273 * then restart the mastery. */
1274 mlog(ML_NOTICE, "node %d up while restarting\n", node);
6714d8e8
KH		 1275
1276 /* redo the master request, but only for the new node */
1277 mlog(0, "sending request to new node\n");
1278 clear_bit(node, mle->response_map);
1279 set_bit(node, mle->vote_map);
1280 } else {
1281 mlog(ML_ERROR, "node down! %d\n", node);
6714d8e8
KH		 1282 if (blocked) {
1283 int lowest = find_next_bit(mle->maybe_map,
1284 O2NM_MAX_NODES, 0);
1285
1286 /* act like it was never there */
1287 clear_bit(node, mle->maybe_map);
1288
e7e69eb3
KH		 1289 if (node == lowest) {
1290 mlog(0, "expected master %u died"
1291 " while this node was blocked "
1292 "waiting on it!\n", node);
1293 lowest = find_next_bit(mle->maybe_map,
1294 O2NM_MAX_NODES,
1295 lowest+1);
1296 if (lowest < O2NM_MAX_NODES) {
1297 mlog(0, "%s:%.*s:still "
1298 "blocked. waiting on %u "
1299 "now\n", dlm->name,
1300 res->lockname.len,
1301 res->lockname.name,
1302 lowest);
1303 } else {
1304 /* mle is an MLE_BLOCK, but
1305 * there is now nothing left to
1306 * block on. we need to return
1307 * all the way back out and try
1308 * again with an MLE_MASTER.
1309 * dlm_do_local_recovery_cleanup
1310 * has already run, so the mle
1311 * refcount is ok */
1312 mlog(0, "%s:%.*s: no "
1313 "longer blocking. try to "
1314 "master this here\n",
1315 dlm->name,
1316 res->lockname.len,
1317 res->lockname.name);
1318 mle->type = DLM_MLE_MASTER;
f77a9a78 1319 mle->u.mleres = res;
e7e69eb3 1320 }
6714d8e8 1321 }
6714d8e8
KH		 1322 }
1323
e7e69eb3
KH		 1324 /* now blank out everything, as if we had never
1325 * contacted anyone */
1326 memset(mle->maybe_map, 0, sizeof(mle->maybe_map));
1327 memset(mle->response_map, 0, sizeof(mle->response_map));
1328 /* reset the vote_map to the current node_map */
1329 memcpy(mle->vote_map, mle->node_map,
1330 sizeof(mle->node_map));
1331 /* put myself into the maybe map */
1332 if (mle->type != DLM_MLE_BLOCK)
1333 set_bit(dlm->node_num, mle->maybe_map);
6714d8e8
KH		 1334 }
1335 ret = -EAGAIN;
6714d8e8
KH		 1336 node = dlm_bitmap_diff_iter_next(&bdi, &sc);
1337 }
1338 return ret;
1339}
1340
1341
1342/*
1343 * DLM_MASTER_REQUEST_MSG
1344 *
1345 * returns: 0 on success,
1346 * -errno on a network error
1347 *
1348 * on error, the caller should assume the target node is "dead"
1349 *
1350 */
1351
ba2bf218
KH		 1352static int dlm_do_master_request(struct dlm_lock_resource *res,
1353 struct dlm_master_list_entry *mle, int to)
6714d8e8
KH		 1354{
1355 struct dlm_ctxt *dlm = mle->dlm;
1356 struct dlm_master_request request;
1357 int ret, response=0, resend;
f77a9a78
SM		 1358 unsigned char *mlename;
1359 unsigned int mlenamelen;
6714d8e8
KH		 1360
1361 memset(&request, 0, sizeof(request));
1362 request.node_idx = dlm->node_num;
1363
1364 BUG_ON(mle->type == DLM_MLE_MIGRATION);
1365
2ed6c750 1366 __dlm_mle_name(mle, &mlename, &mlenamelen, NULL);
f77a9a78
SM		 1367
1368 request.namelen = (u8)mlenamelen;
1369 memcpy(request.name, mlename, request.namelen);
6714d8e8
KH		 1370
1371again:
1372 ret = o2net_send_message(DLM_MASTER_REQUEST_MSG, dlm->key, &request,
1373 sizeof(request), to, &response);
1374 if (ret < 0) {
1375 if (ret == -ESRCH) {
1376 /* should never happen */
1377 mlog(ML_ERROR, "TCP stack not ready!\n");
1378 BUG();
1379 } else if (ret == -EINVAL) {
1380 mlog(ML_ERROR, "bad args passed to o2net!\n");
1381 BUG();
1382 } else if (ret == -ENOMEM) {
1383 mlog(ML_ERROR, "out of memory while trying to send "
1384 "network message! retrying\n");
1385 /* this is totally crude */
1386 msleep(50);
1387 goto again;
1388 } else if (!dlm_is_host_down(ret)) {
1389 /* not a network error. bad. */
1390 mlog_errno(ret);
1391 mlog(ML_ERROR, "unhandled error!");
1392 BUG();
1393 }
1394 /* all other errors should be network errors,
1395 * and likely indicate node death */
1396 mlog(ML_ERROR, "link to %d went down!\n", to);
1397 goto out;
1398 }
1399
1400 ret = 0;
1401 resend = 0;
1402 spin_lock(&mle->spinlock);
1403 switch (response) {
1404 case DLM_MASTER_RESP_YES:
1405 set_bit(to, mle->response_map);
1406 mlog(0, "node %u is the master, response=YES\n", to);
ba2bf218
KH		 1407 mlog(0, "%s:%.*s: master node %u now knows I have a "
1408 "reference\n", dlm->name, res->lockname.len,
1409 res->lockname.name, to);
6714d8e8
KH		 1410 mle->master = to;
1411 break;
1412 case DLM_MASTER_RESP_NO:
1413 mlog(0, "node %u not master, response=NO\n", to);
1414 set_bit(to, mle->response_map);
1415 break;
1416 case DLM_MASTER_RESP_MAYBE:
1417 mlog(0, "node %u not master, response=MAYBE\n", to);
1418 set_bit(to, mle->response_map);
1419 set_bit(to, mle->maybe_map);
1420 break;
1421 case DLM_MASTER_RESP_ERROR:
1422 mlog(0, "node %u hit an error, resending\n", to);
1423 resend = 1;
1424 response = 0;
1425 break;
1426 default:
1427 mlog(ML_ERROR, "bad response! %u\n", response);
1428 BUG();
1429 }
1430 spin_unlock(&mle->spinlock);
1431 if (resend) {
1432 /* this is also totally crude */
1433 msleep(50);
1434 goto again;
1435 }
1436
1437out:
1438 return ret;
1439}
1440
1441/*
1442 * locks that can be taken here:
1443 * dlm->spinlock
1444 * res->spinlock
1445 * mle->spinlock
1446 * dlm->master_list
1447 *
1448 * if possible, TRIM THIS DOWN!!!
1449 */
d74c9803
KH		 1450int dlm_master_request_handler(struct o2net_msg *msg, u32 len, void *data,
1451 void **ret_data)
6714d8e8
KH		 1452{
1453 u8 response = DLM_MASTER_RESP_MAYBE;
1454 struct dlm_ctxt *dlm = data;
9c6510a5 1455 struct dlm_lock_resource *res = NULL;
6714d8e8
KH		 1456 struct dlm_master_request *request = (struct dlm_master_request *) msg->buf;
1457 struct dlm_master_list_entry *mle = NULL, *tmpmle = NULL;
1458 char *name;
a3d33291 1459 unsigned int namelen, hash;
6714d8e8
KH		 1460 int found, ret;
1461 int set_maybe;
9c6510a5 1462 int dispatch_assert = 0;
6714d8e8
KH		 1463
1464 if (!dlm_grab(dlm))
1465 return DLM_MASTER_RESP_NO;
1466
1467 if (!dlm_domain_fully_joined(dlm)) {
1468 response = DLM_MASTER_RESP_NO;
1469 goto send_response;
1470 }
1471
1472 name = request->name;
1473 namelen = request->namelen;
a3d33291 1474 hash = dlm_lockid_hash(name, namelen);
6714d8e8
KH		 1475
1476 if (namelen > DLM_LOCKID_NAME_MAX) {
1477 response = DLM_IVBUFLEN;
1478 goto send_response;
1479 }
1480
1481way_up_top:
1482 spin_lock(&dlm->spinlock);
a3d33291 1483 res = __dlm_lookup_lockres(dlm, name, namelen, hash);
6714d8e8
KH		 1484 if (res) {
1485 spin_unlock(&dlm->spinlock);
1486
1487 /* take care of the easy cases up front */
1488 spin_lock(&res->spinlock);
1cd04dbe
KH		 1489 if (res->state & (DLM_LOCK_RES_RECOVERING|
1490 DLM_LOCK_RES_MIGRATING)) {
6714d8e8
KH		 1491 spin_unlock(&res->spinlock);
1492 mlog(0, "returning DLM_MASTER_RESP_ERROR since res is "
1cd04dbe 1493 "being recovered/migrated\n");
6714d8e8
KH		 1494 response = DLM_MASTER_RESP_ERROR;
1495 if (mle)
1496 kmem_cache_free(dlm_mle_cache, mle);
1497 goto send_response;
1498 }
1499
1500 if (res->owner == dlm->node_num) {
ba2bf218
KH		 1501 mlog(0, "%s:%.*s: setting bit %u in refmap\n",
1502 dlm->name, namelen, name, request->node_idx);
1503 dlm_lockres_set_refmap_bit(request->node_idx, res);
6714d8e8 1504 spin_unlock(&res->spinlock);
6714d8e8
KH		 1505 response = DLM_MASTER_RESP_YES;
1506 if (mle)
1507 kmem_cache_free(dlm_mle_cache, mle);
1508
1509 /* this node is the owner.
1510 * there is some extra work that needs to
1511 * happen now. the requesting node has
1512 * caused all nodes up to this one to
1513 * create mles. this node now needs to
1514 * go back and clean those up. */
9c6510a5 1515 dispatch_assert = 1;
6714d8e8
KH		 1516 goto send_response;
1517 } else if (res->owner != DLM_LOCK_RES_OWNER_UNKNOWN) {
1518 spin_unlock(&res->spinlock);
1519 // mlog(0, "node %u is the master\n", res->owner);
1520 response = DLM_MASTER_RESP_NO;
1521 if (mle)
1522 kmem_cache_free(dlm_mle_cache, mle);
1523 goto send_response;
1524 }
1525
1526 /* ok, there is no owner. either this node is
1527 * being blocked, or it is actively trying to
1528 * master this lock. */
1529 if (!(res->state & DLM_LOCK_RES_IN_PROGRESS)) {
1530 mlog(ML_ERROR, "lock with no owner should be "
1531 "in-progress!\n");
1532 BUG();
1533 }
1534
1535 // mlog(0, "lockres is in progress...\n");
1536 spin_lock(&dlm->master_lock);
1537 found = dlm_find_mle(dlm, &tmpmle, name, namelen);
1538 if (!found) {
1539 mlog(ML_ERROR, "no mle found for this lock!\n");
1540 BUG();
1541 }
1542 set_maybe = 1;
1543 spin_lock(&tmpmle->spinlock);
1544 if (tmpmle->type == DLM_MLE_BLOCK) {
1545 // mlog(0, "this node is waiting for "
1546 // "lockres to be mastered\n");
1547 response = DLM_MASTER_RESP_NO;
1548 } else if (tmpmle->type == DLM_MLE_MIGRATION) {
1549 mlog(0, "node %u is master, but trying to migrate to "
1550 "node %u.\n", tmpmle->master, tmpmle->new_master);
1551 if (tmpmle->master == dlm->node_num) {
6714d8e8
KH		 1552 mlog(ML_ERROR, "no owner on lockres, but this "
1553 "node is trying to migrate it to %u?!\n",
1554 tmpmle->new_master);
1555 BUG();
1556 } else {
1557 /* the real master can respond on its own */
1558 response = DLM_MASTER_RESP_NO;
1559 }
1560 } else if (tmpmle->master != DLM_LOCK_RES_OWNER_UNKNOWN) {
1561 set_maybe = 0;
9c6510a5 1562 if (tmpmle->master == dlm->node_num) {
6714d8e8 1563 response = DLM_MASTER_RESP_YES;
9c6510a5
KH		 1564 /* this node will be the owner.
1565 * go back and clean the mles on any
1566 * other nodes */
1567 dispatch_assert = 1;
ba2bf218
KH		 1568 dlm_lockres_set_refmap_bit(request->node_idx, res);
1569 mlog(0, "%s:%.*s: setting bit %u in refmap\n",
1570 dlm->name, namelen, name,
1571 request->node_idx);
9c6510a5 1572 } else
6714d8e8
KH		 1573 response = DLM_MASTER_RESP_NO;
1574 } else {
1575 // mlog(0, "this node is attempting to "
1576 // "master lockres\n");
1577 response = DLM_MASTER_RESP_MAYBE;
1578 }
1579 if (set_maybe)
1580 set_bit(request->node_idx, tmpmle->maybe_map);
1581 spin_unlock(&tmpmle->spinlock);
1582
1583 spin_unlock(&dlm->master_lock);
1584 spin_unlock(&res->spinlock);
1585
1586 /* keep the mle attached to heartbeat events */
1587 dlm_put_mle(tmpmle);
1588 if (mle)
1589 kmem_cache_free(dlm_mle_cache, mle);
1590 goto send_response;
1591 }
1592
1593 /*
1594 * lockres doesn't exist on this node
1595 * if there is an MLE_BLOCK, return NO
1596 * if there is an MLE_MASTER, return MAYBE
1597 * otherwise, add an MLE_BLOCK, return NO
1598 */
1599 spin_lock(&dlm->master_lock);
1600 found = dlm_find_mle(dlm, &tmpmle, name, namelen);
1601 if (!found) {
1602 /* this lockid has never been seen on this node yet */
1603 // mlog(0, "no mle found\n");
1604 if (!mle) {
1605 spin_unlock(&dlm->master_lock);
1606 spin_unlock(&dlm->spinlock);
1607
1608 mle = (struct dlm_master_list_entry *)
ad8100e0 1609 kmem_cache_alloc(dlm_mle_cache, GFP_NOFS);
6714d8e8 1610 if (!mle) {
6714d8e8 1611 response = DLM_MASTER_RESP_ERROR;
9c6510a5 1612 mlog_errno(-ENOMEM);
6714d8e8
KH		 1613 goto send_response;
1614 }
6714d8e8
KH		 1615 goto way_up_top;
1616 }
1617
1618 // mlog(0, "this is second time thru, already allocated, "
1619 // "add the block.\n");
41b8c8a1 1620 dlm_init_mle(mle, DLM_MLE_BLOCK, dlm, NULL, name, namelen);
6714d8e8 1621 set_bit(request->node_idx, mle->maybe_map);
1c084577 1622 __dlm_insert_mle(dlm, mle);
6714d8e8
KH		 1623 response = DLM_MASTER_RESP_NO;
1624 } else {
1625 // mlog(0, "mle was found\n");
1626 set_maybe = 1;
1627 spin_lock(&tmpmle->spinlock);
9c6510a5
KH		 1628 if (tmpmle->master == dlm->node_num) {
1629 mlog(ML_ERROR, "no lockres, but an mle with this node as master!\n");
1630 BUG();
1631 }
6714d8e8
KH		 1632 if (tmpmle->type == DLM_MLE_BLOCK)
1633 response = DLM_MASTER_RESP_NO;
1634 else if (tmpmle->type == DLM_MLE_MIGRATION) {
1635 mlog(0, "migration mle was found (%u->%u)\n",
1636 tmpmle->master, tmpmle->new_master);
6714d8e8
KH		 1637 /* real master can respond on its own */
1638 response = DLM_MASTER_RESP_NO;
9c6510a5
KH		 1639 } else
1640 response = DLM_MASTER_RESP_MAYBE;
6714d8e8
KH		 1641 if (set_maybe)
1642 set_bit(request->node_idx, tmpmle->maybe_map);
1643 spin_unlock(&tmpmle->spinlock);
1644 }
1645 spin_unlock(&dlm->master_lock);
1646 spin_unlock(&dlm->spinlock);
1647
1648 if (found) {
1649 /* keep the mle attached to heartbeat events */
1650 dlm_put_mle(tmpmle);
1651 }
1652send_response:
b31cfc02
SM		 1653 /*
1654 * __dlm_lookup_lockres() grabbed a reference to this lockres.
1655 * The reference is released by dlm_assert_master_worker() under
1656 * the call to dlm_dispatch_assert_master(). If
1657 * dlm_assert_master_worker() isn't called, we drop it here.
1658 */
9c6510a5
KH		 1659 if (dispatch_assert) {
1660 if (response != DLM_MASTER_RESP_YES)
1661 mlog(ML_ERROR, "invalid response %d\n", response);
1662 if (!res) {
1663 mlog(ML_ERROR, "bad lockres while trying to assert!\n");
1664 BUG();
1665 }
1666 mlog(0, "%u is the owner of %.*s, cleaning everyone else\n",
1667 dlm->node_num, res->lockname.len, res->lockname.name);
1668 ret = dlm_dispatch_assert_master(dlm, res, 0, request->node_idx,
1669 DLM_ASSERT_MASTER_MLE_CLEANUP);
1670 if (ret < 0) {
1671 mlog(ML_ERROR, "failed to dispatch assert master work\n");
1672 response = DLM_MASTER_RESP_ERROR;
b31cfc02 1673 dlm_lockres_put(res);
9c6510a5 1674 }
b31cfc02
SM		 1675 } else {
1676 if (res)
1677 dlm_lockres_put(res);
9c6510a5
KH		 1678 }
1679
6714d8e8
KH		 1680 dlm_put(dlm);
1681 return response;
1682}
1683
1684/*
1685 * DLM_ASSERT_MASTER_MSG
1686 */
1687
1688
1689/*
1690 * NOTE: this can be used for debugging
1691 * can periodically run all locks owned by this node
1692 * and re-assert across the cluster...
1693 */
05488bbe
AB		 1694static int dlm_do_assert_master(struct dlm_ctxt *dlm,
1695 struct dlm_lock_resource *res,
1696 void *nodemap, u32 flags)
6714d8e8
KH		 1697{
1698 struct dlm_assert_master assert;
1699 int to, tmpret;
1700 struct dlm_node_iter iter;
1701 int ret = 0;
9c6510a5 1702 int reassert;
ba2bf218
KH		 1703 const char *lockname = res->lockname.name;
1704 unsigned int namelen = res->lockname.len;
6714d8e8
KH		 1705
1706 BUG_ON(namelen > O2NM_MAX_NAME_LEN);
f3f85464
SM		 1707
1708 spin_lock(&res->spinlock);
1709 res->state |= DLM_LOCK_RES_SETREF_INPROG;
1710 spin_unlock(&res->spinlock);
1711
9c6510a5
KH		 1712again:
1713 reassert = 0;
6714d8e8
KH		 1714
1715 /* note that if this nodemap is empty, it returns 0 */
1716 dlm_node_iter_init(nodemap, &iter);
1717 while ((to = dlm_node_iter_next(&iter)) >= 0) {
1718 int r = 0;
a9ee4c8a
KH		 1719 struct dlm_master_list_entry *mle = NULL;
1720
6714d8e8
KH		 1721 mlog(0, "sending assert master to %d (%.*s)\n", to,
1722 namelen, lockname);
1723 memset(&assert, 0, sizeof(assert));
1724 assert.node_idx = dlm->node_num;
1725 assert.namelen = namelen;
1726 memcpy(assert.name, lockname, namelen);
1727 assert.flags = cpu_to_be32(flags);
1728
1729 tmpret = o2net_send_message(DLM_ASSERT_MASTER_MSG, dlm->key,
1730 &assert, sizeof(assert), to, &r);
1731 if (tmpret < 0) {
3b3b84a8 1732 mlog(0, "assert_master returned %d!\n", tmpret);
6714d8e8 1733 if (!dlm_is_host_down(tmpret)) {
3b3b84a8 1734 mlog(ML_ERROR, "unhandled error=%d!\n", tmpret);
6714d8e8
KH		 1735 BUG();
1736 }
1737 /* a node died. finish out the rest of the nodes. */
3b3b84a8 1738 mlog(0, "link to %d went down!\n", to);
6714d8e8
KH		 1739 /* any nonzero status return will do */
1740 ret = tmpret;
ba2bf218 1741 r = 0;
6714d8e8
KH		 1742 } else if (r < 0) {
1743 /* ok, something horribly messed. kill thyself. */
1744 mlog(ML_ERROR,"during assert master of %.*s to %u, "
1745 "got %d.\n", namelen, lockname, to, r);
a9ee4c8a
KH		 1746 spin_lock(&dlm->spinlock);
1747 spin_lock(&dlm->master_lock);
1748 if (dlm_find_mle(dlm, &mle, (char *)lockname,
1749 namelen)) {
1750 dlm_print_one_mle(mle);
1751 __dlm_put_mle(mle);
1752 }
1753 spin_unlock(&dlm->master_lock);
1754 spin_unlock(&dlm->spinlock);
6714d8e8 1755 BUG();
ba2bf218
KH		 1756 }
1757
1758 if (r & DLM_ASSERT_RESPONSE_REASSERT &&
1759 !(r & DLM_ASSERT_RESPONSE_MASTERY_REF)) {
1760 mlog(ML_ERROR, "%.*s: very strange, "
1761 "master MLE but no lockres on %u\n",
1762 namelen, lockname, to);
1763 }
1764
1765 if (r & DLM_ASSERT_RESPONSE_REASSERT) {
9c6510a5
KH		 1766 mlog(0, "%.*s: node %u create mles on other "
1767 "nodes and requests a re-assert\n",
1768 namelen, lockname, to);
1769 reassert = 1;
6714d8e8 1770 }
ba2bf218
KH		 1771 if (r & DLM_ASSERT_RESPONSE_MASTERY_REF) {
1772 mlog(0, "%.*s: node %u has a reference to this "
1773 "lockres, set the bit in the refmap\n",
1774 namelen, lockname, to);
1775 spin_lock(&res->spinlock);
1776 dlm_lockres_set_refmap_bit(to, res);
1777 spin_unlock(&res->spinlock);
1778 }
6714d8e8
KH		 1779 }
1780
9c6510a5
KH		 1781 if (reassert)
1782 goto again;
1783
f3f85464
SM		 1784 spin_lock(&res->spinlock);
1785 res->state &= ~DLM_LOCK_RES_SETREF_INPROG;
1786 spin_unlock(&res->spinlock);
1787 wake_up(&res->wq);
1788
6714d8e8
KH		 1789 return ret;
1790}
1791
1792/*
1793 * locks that can be taken here:
1794 * dlm->spinlock
1795 * res->spinlock
1796 * mle->spinlock
1797 * dlm->master_list
1798 *
1799 * if possible, TRIM THIS DOWN!!!
1800 */
d74c9803
KH		 1801int dlm_assert_master_handler(struct o2net_msg *msg, u32 len, void *data,
1802 void **ret_data)
6714d8e8
KH		 1803{
1804 struct dlm_ctxt *dlm = data;
1805 struct dlm_master_list_entry *mle = NULL;
1806 struct dlm_assert_master *assert = (struct dlm_assert_master *)msg->buf;
1807 struct dlm_lock_resource *res = NULL;
1808 char *name;
a3d33291 1809 unsigned int namelen, hash;
6714d8e8 1810 u32 flags;
ba2bf218 1811 int master_request = 0, have_lockres_ref = 0;
9c6510a5 1812 int ret = 0;
6714d8e8
KH		 1813
1814 if (!dlm_grab(dlm))
1815 return 0;
1816
1817 name = assert->name;
1818 namelen = assert->namelen;
a3d33291 1819 hash = dlm_lockid_hash(name, namelen);
6714d8e8
KH		 1820 flags = be32_to_cpu(assert->flags);
1821
1822 if (namelen > DLM_LOCKID_NAME_MAX) {
1823 mlog(ML_ERROR, "Invalid name length!");
1824 goto done;
1825 }
1826
1827 spin_lock(&dlm->spinlock);
1828
1829 if (flags)
1830 mlog(0, "assert_master with flags: %u\n", flags);
1831
1832 /* find the MLE */
1833 spin_lock(&dlm->master_lock);
1834 if (!dlm_find_mle(dlm, &mle, name, namelen)) {
1835 /* not an error, could be master just re-asserting */
1836 mlog(0, "just got an assert_master from %u, but no "
1837 "MLE for it! (%.*s)\n", assert->node_idx,
1838 namelen, name);
1839 } else {
1840 int bit = find_next_bit (mle->maybe_map, O2NM_MAX_NODES, 0);
1841 if (bit >= O2NM_MAX_NODES) {
1842 /* not necessarily an error, though less likely.
1843 * could be master just re-asserting. */
aa852354 1844 mlog(0, "no bits set in the maybe_map, but %u "
6714d8e8
KH		 1845 "is asserting! (%.*s)\n", assert->node_idx,
1846 namelen, name);
1847 } else if (bit != assert->node_idx) {
1848 if (flags & DLM_ASSERT_MASTER_MLE_CLEANUP) {
1849 mlog(0, "master %u was found, %u should "
1850 "back off\n", assert->node_idx, bit);
1851 } else {
1852 /* with the fix for bug 569, a higher node
1853 * number winning the mastery will respond
1854 * YES to mastery requests, but this node
1855 * had no way of knowing. let it pass. */
aa852354 1856 mlog(0, "%u is the lowest node, "
6714d8e8
KH		 1857 "%u is asserting. (%.*s) %u must "
1858 "have begun after %u won.\n", bit,
1859 assert->node_idx, namelen, name, bit,
1860 assert->node_idx);
1861 }
1862 }
2d1a868c
KH		 1863 if (mle->type == DLM_MLE_MIGRATION) {
1864 if (flags & DLM_ASSERT_MASTER_MLE_CLEANUP) {
1865 mlog(0, "%s:%.*s: got cleanup assert"
1866 " from %u for migration\n",
1867 dlm->name, namelen, name,
1868 assert->node_idx);
1869 } else if (!(flags & DLM_ASSERT_MASTER_FINISH_MIGRATION)) {
1870 mlog(0, "%s:%.*s: got unrelated assert"
1871 " from %u for migration, ignoring\n",
1872 dlm->name, namelen, name,
1873 assert->node_idx);
1874 __dlm_put_mle(mle);
1875 spin_unlock(&dlm->master_lock);
1876 spin_unlock(&dlm->spinlock);
1877 goto done;
1878 }
1879 }
6714d8e8
KH		 1880 }
1881 spin_unlock(&dlm->master_lock);
1882
1883 /* ok everything checks out with the MLE
1884 * now check to see if there is a lockres */
a3d33291 1885 res = __dlm_lookup_lockres(dlm, name, namelen, hash);
6714d8e8
KH		 1886 if (res) {
1887 spin_lock(&res->spinlock);
1888 if (res->state & DLM_LOCK_RES_RECOVERING) {
1889 mlog(ML_ERROR, "%u asserting but %.*s is "
1890 "RECOVERING!\n", assert->node_idx, namelen, name);
1891 goto kill;
1892 }
1893 if (!mle) {
dc2ed195
KH		 1894 if (res->owner != DLM_LOCK_RES_OWNER_UNKNOWN &&
1895 res->owner != assert->node_idx) {
53ecd25e
SM		 1896 mlog(ML_ERROR, "DIE! Mastery assert from %u, "
1897 "but current owner is %u! (%.*s)\n",
1898 assert->node_idx, res->owner, namelen,
1899 name);
1900 __dlm_print_one_lock_resource(res);
1901 BUG();
6714d8e8
KH		 1902 }
1903 } else if (mle->type != DLM_MLE_MIGRATION) {
1904 if (res->owner != DLM_LOCK_RES_OWNER_UNKNOWN) {
1905 /* owner is just re-asserting */
1906 if (res->owner == assert->node_idx) {
1907 mlog(0, "owner %u re-asserting on "
1908 "lock %.*s\n", assert->node_idx,
1909 namelen, name);
1910 goto ok;
1911 }
1912 mlog(ML_ERROR, "got assert_master from "
1913 "node %u, but %u is the owner! "
1914 "(%.*s)\n", assert->node_idx,
1915 res->owner, namelen, name);
1916 goto kill;
1917 }
1918 if (!(res->state & DLM_LOCK_RES_IN_PROGRESS)) {
1919 mlog(ML_ERROR, "got assert from %u, but lock "
1920 "with no owner should be "
1921 "in-progress! (%.*s)\n",
1922 assert->node_idx,
1923 namelen, name);
1924 goto kill;
1925 }
1926 } else /* mle->type == DLM_MLE_MIGRATION */ {
1927 /* should only be getting an assert from new master */
1928 if (assert->node_idx != mle->new_master) {
1929 mlog(ML_ERROR, "got assert from %u, but "
1930 "new master is %u, and old master "
1931 "was %u (%.*s)\n",
1932 assert->node_idx, mle->new_master,
1933 mle->master, namelen, name);
1934 goto kill;
1935 }
1936
1937 }
1938ok:
1939 spin_unlock(&res->spinlock);
1940 }
1941 spin_unlock(&dlm->spinlock);
1942
1943 // mlog(0, "woo! got an assert_master from node %u!\n",
1944 // assert->node_idx);
1945 if (mle) {
9c6510a5
KH		 1946 int extra_ref = 0;
1947 int nn = -1;
a2bf0477 1948 int rr, err = 0;
6714d8e8
KH		 1949
1950 spin_lock(&mle->spinlock);
9c6510a5
KH		 1951 if (mle->type == DLM_MLE_BLOCK || mle->type == DLM_MLE_MIGRATION)
1952 extra_ref = 1;
1953 else {
1954 /* MASTER mle: if any bits set in the response map
1955 * then the calling node needs to re-assert to clear
1956 * up nodes that this node contacted */
1957 while ((nn = find_next_bit (mle->response_map, O2NM_MAX_NODES,
1958 nn+1)) < O2NM_MAX_NODES) {
1959 if (nn != dlm->node_num && nn != assert->node_idx)
1960 master_request = 1;
1961 }
1962 }
6714d8e8
KH		 1963 mle->master = assert->node_idx;
1964 atomic_set(&mle->woken, 1);
1965 wake_up(&mle->wq);
1966 spin_unlock(&mle->spinlock);
1967
a2bf0477 1968 if (res) {
a6fa3640 1969 int wake = 0;
6714d8e8 1970 spin_lock(&res->spinlock);
a2bf0477
KH		 1971 if (mle->type == DLM_MLE_MIGRATION) {
1972 mlog(0, "finishing off migration of lockres %.*s, "
1973 "from %u to %u\n",
1974 res->lockname.len, res->lockname.name,
1975 dlm->node_num, mle->new_master);
1976 res->state &= ~DLM_LOCK_RES_MIGRATING;
a6fa3640 1977 wake = 1;
a2bf0477
KH		 1978 dlm_change_lockres_owner(dlm, res, mle->new_master);
1979 BUG_ON(res->state & DLM_LOCK_RES_DIRTY);
1980 } else {
1981 dlm_change_lockres_owner(dlm, res, mle->master);
1982 }
6714d8e8 1983 spin_unlock(&res->spinlock);
ba2bf218 1984 have_lockres_ref = 1;
a6fa3640
KH		 1985 if (wake)
1986 wake_up(&res->wq);
6714d8e8 1987 }
a2bf0477
KH		 1988
1989 /* master is known, detach if not already detached.
1990 * ensures that only one assert_master call will happen
1991 * on this mle. */
1992 spin_lock(&dlm->spinlock);
1993 spin_lock(&dlm->master_lock);
1994
1995 rr = atomic_read(&mle->mle_refs.refcount);
1996 if (mle->inuse > 0) {
1997 if (extra_ref && rr < 3)
1998 err = 1;
1999 else if (!extra_ref && rr < 2)
2000 err = 1;
2001 } else {
2002 if (extra_ref && rr < 2)
2003 err = 1;
2004 else if (!extra_ref && rr < 1)
2005 err = 1;
2006 }
2007 if (err) {
2008 mlog(ML_ERROR, "%s:%.*s: got assert master from %u "
2009 "that will mess up this node, refs=%d, extra=%d, "
2010 "inuse=%d\n", dlm->name, namelen, name,
2011 assert->node_idx, rr, extra_ref, mle->inuse);
2012 dlm_print_one_mle(mle);
2013 }
1c084577 2014 __dlm_unlink_mle(dlm, mle);
a2bf0477
KH		 2015 __dlm_mle_detach_hb_events(dlm, mle);
2016 __dlm_put_mle(mle);
6714d8e8
KH		 2017 if (extra_ref) {
2018 /* the assert master message now balances the extra
2019 * ref given by the master / migration request message.
2020 * if this is the last put, it will be removed
2021 * from the list. */
a2bf0477
KH		 2022 __dlm_put_mle(mle);
2023 }
2024 spin_unlock(&dlm->master_lock);
2025 spin_unlock(&dlm->spinlock);
2026 } else if (res) {
2027 if (res->owner != assert->node_idx) {
2028 mlog(0, "assert_master from %u, but current "
2029 "owner is %u (%.*s), no mle\n", assert->node_idx,
2030 res->owner, namelen, name);
6714d8e8
KH		 2031 }
2032 }
2033
2034done:
9c6510a5 2035 ret = 0;
3b8118cf
KH		 2036 if (res) {
2037 spin_lock(&res->spinlock);
2038 res->state |= DLM_LOCK_RES_SETREF_INPROG;
2039 spin_unlock(&res->spinlock);
2040 *ret_data = (void *)res;
2041 }
6714d8e8 2042 dlm_put(dlm);
9c6510a5
KH		 2043 if (master_request) {
2044 mlog(0, "need to tell master to reassert\n");
ba2bf218
KH		 2045 /* positive. negative would shoot down the node. */
2046 ret |= DLM_ASSERT_RESPONSE_REASSERT;
2047 if (!have_lockres_ref) {
2048 mlog(ML_ERROR, "strange, got assert from %u, MASTER "
2049 "mle present here for %s:%.*s, but no lockres!\n",
2050 assert->node_idx, dlm->name, namelen, name);
2051 }
2052 }
2053 if (have_lockres_ref) {
2054 /* let the master know we have a reference to the lockres */
2055 ret |= DLM_ASSERT_RESPONSE_MASTERY_REF;
2056 mlog(0, "%s:%.*s: got assert from %u, need a ref\n",
2057 dlm->name, namelen, name, assert->node_idx);
9c6510a5
KH		 2058 }
2059 return ret;
6714d8e8
KH		 2060
2061kill:
2062 /* kill the caller! */
a9ee4c8a
KH		 2063 mlog(ML_ERROR, "Bad message received from another node. Dumping state "
2064 "and killing the other node now! This node is OK and can continue.\n");
2065 __dlm_print_one_lock_resource(res);
6714d8e8
KH		 2066 spin_unlock(&res->spinlock);
2067 spin_unlock(&dlm->spinlock);
3b8118cf 2068 *ret_data = (void *)res;
6714d8e8
KH		 2069 dlm_put(dlm);
2070 return -EINVAL;
2071}
2072
3b8118cf
KH		 2073void dlm_assert_master_post_handler(int status, void *data, void *ret_data)
2074{
2075 struct dlm_lock_resource *res = (struct dlm_lock_resource *)ret_data;
2076
2077 if (ret_data) {
2078 spin_lock(&res->spinlock);
2079 res->state &= ~DLM_LOCK_RES_SETREF_INPROG;
2080 spin_unlock(&res->spinlock);
2081 wake_up(&res->wq);
2082 dlm_lockres_put(res);
2083 }
2084 return;
2085}
2086
6714d8e8
KH		 2087int dlm_dispatch_assert_master(struct dlm_ctxt *dlm,
2088 struct dlm_lock_resource *res,
2089 int ignore_higher, u8 request_from, u32 flags)
2090{
2091 struct dlm_work_item *item;
cd861280 2092 item = kzalloc(sizeof(*item), GFP_NOFS);
6714d8e8
KH		 2093 if (!item)
2094 return -ENOMEM;
2095
2096
2097 /* queue up work for dlm_assert_master_worker */
2098 dlm_grab(dlm); /* get an extra ref for the work item */
2099 dlm_init_work_item(dlm, item, dlm_assert_master_worker, NULL);
2100 item->u.am.lockres = res; /* already have a ref */
2101 /* can optionally ignore node numbers higher than this node */
2102 item->u.am.ignore_higher = ignore_higher;
2103 item->u.am.request_from = request_from;
2104 item->u.am.flags = flags;
2105
9c6510a5
KH		 2106 if (ignore_higher)
2107 mlog(0, "IGNORE HIGHER: %.*s\n", res->lockname.len,
2108 res->lockname.name);
2109
6714d8e8
KH		 2110 spin_lock(&dlm->work_lock);
2111 list_add_tail(&item->list, &dlm->work_list);
2112 spin_unlock(&dlm->work_lock);
2113
3156d267 2114 queue_work(dlm->dlm_worker, &dlm->dispatched_work);
6714d8e8
KH		 2115 return 0;
2116}
2117
2118static void dlm_assert_master_worker(struct dlm_work_item *item, void *data)
2119{
2120 struct dlm_ctxt *dlm = data;
2121 int ret = 0;
2122 struct dlm_lock_resource *res;
2123 unsigned long nodemap[BITS_TO_LONGS(O2NM_MAX_NODES)];
2124 int ignore_higher;
2125 int bit;
2126 u8 request_from;
2127 u32 flags;
2128
2129 dlm = item->dlm;
2130 res = item->u.am.lockres;
2131 ignore_higher = item->u.am.ignore_higher;
2132 request_from = item->u.am.request_from;
2133 flags = item->u.am.flags;
2134
2135 spin_lock(&dlm->spinlock);
2136 memcpy(nodemap, dlm->domain_map, sizeof(nodemap));
2137 spin_unlock(&dlm->spinlock);
2138
2139 clear_bit(dlm->node_num, nodemap);
2140 if (ignore_higher) {
2141 /* if is this just to clear up mles for nodes below
2142 * this node, do not send the message to the original
2143 * caller or any node number higher than this */
2144 clear_bit(request_from, nodemap);
2145 bit = dlm->node_num;
2146 while (1) {
2147 bit = find_next_bit(nodemap, O2NM_MAX_NODES,
2148 bit+1);
2149 if (bit >= O2NM_MAX_NODES)
2150 break;
2151 clear_bit(bit, nodemap);
2152 }
2153 }
2154
36407488
KH		 2155 /*
2156 * If we're migrating this lock to someone else, we are no
2157 * longer allowed to assert out own mastery. OTOH, we need to
2158 * prevent migration from starting while we're still asserting
2159 * our dominance. The reserved ast delays migration.
2160 */
2161 spin_lock(&res->spinlock);
2162 if (res->state & DLM_LOCK_RES_MIGRATING) {
2163 mlog(0, "Someone asked us to assert mastery, but we're "
2164 "in the middle of migration. Skipping assert, "
2165 "the new master will handle that.\n");
2166 spin_unlock(&res->spinlock);
2167 goto put;
2168 } else
2169 __dlm_lockres_reserve_ast(res);
2170 spin_unlock(&res->spinlock);
2171
6714d8e8
KH		 2172 /* this call now finishes out the nodemap
2173 * even if one or more nodes die */
2174 mlog(0, "worker about to master %.*s here, this=%u\n",
2175 res->lockname.len, res->lockname.name, dlm->node_num);
ba2bf218 2176 ret = dlm_do_assert_master(dlm, res, nodemap, flags);
6714d8e8
KH		 2177 if (ret < 0) {
2178 /* no need to restart, we are done */
3b3b84a8
KH		 2179 if (!dlm_is_host_down(ret))
2180 mlog_errno(ret);
6714d8e8
KH		 2181 }
2182
36407488
KH		 2183 /* Ok, we've asserted ourselves. Let's let migration start. */
2184 dlm_lockres_release_ast(dlm, res);
2185
2186put:
6714d8e8
KH		 2187 dlm_lockres_put(res);
2188
2189 mlog(0, "finished with dlm_assert_master_worker\n");
2190}
2191
c03872f5
KH		 2192/* SPECIAL CASE for the $RECOVERY lock used by the recovery thread.
2193 * We cannot wait for node recovery to complete to begin mastering this
2194 * lockres because this lockres is used to kick off recovery! ;-)
2195 * So, do a pre-check on all living nodes to see if any of those nodes
2196 * think that $RECOVERY is currently mastered by a dead node. If so,
2197 * we wait a short time to allow that node to get notified by its own
2198 * heartbeat stack, then check again. All $RECOVERY lock resources
2199 * mastered by dead nodes are purged when the hearbeat callback is
2200 * fired, so we can know for sure that it is safe to continue once
2201 * the node returns a live node or no node. */
2202static int dlm_pre_master_reco_lockres(struct dlm_ctxt *dlm,
2203 struct dlm_lock_resource *res)
2204{
2205 struct dlm_node_iter iter;
2206 int nodenum;
2207 int ret = 0;
2208 u8 master = DLM_LOCK_RES_OWNER_UNKNOWN;
2209
2210 spin_lock(&dlm->spinlock);
2211 dlm_node_iter_init(dlm->domain_map, &iter);
2212 spin_unlock(&dlm->spinlock);
2213
2214 while ((nodenum = dlm_node_iter_next(&iter)) >= 0) {
2215 /* do not send to self */
2216 if (nodenum == dlm->node_num)
2217 continue;
2218 ret = dlm_do_master_requery(dlm, res, nodenum, &master);
2219 if (ret < 0) {
2220 mlog_errno(ret);
2221 if (!dlm_is_host_down(ret))
2222 BUG();
2223 /* host is down, so answer for that node would be
2224 * DLM_LOCK_RES_OWNER_UNKNOWN. continue. */
f42a100b 2225 ret = 0;
c03872f5
KH		 2226 }
2227
2228 if (master != DLM_LOCK_RES_OWNER_UNKNOWN) {
2229 /* check to see if this master is in the recovery map */
2230 spin_lock(&dlm->spinlock);
2231 if (test_bit(master, dlm->recovery_map)) {
2232 mlog(ML_NOTICE, "%s: node %u has not seen "
2233 "node %u go down yet, and thinks the "
2234 "dead node is mastering the recovery "
2235 "lock. must wait.\n", dlm->name,
2236 nodenum, master);
2237 ret = -EAGAIN;
2238 }
2239 spin_unlock(&dlm->spinlock);
2240 mlog(0, "%s: reco lock master is %u\n", dlm->name,
2241 master);
2242 break;
2243 }
2244 }
2245 return ret;
2246}
2247
ba2bf218
KH		 2248/*
2249 * DLM_DEREF_LOCKRES_MSG
2250 */
2251
2252int dlm_drop_lockres_ref(struct dlm_ctxt *dlm, struct dlm_lock_resource *res)
2253{
2254 struct dlm_deref_lockres deref;
2255 int ret = 0, r;
2256 const char *lockname;
2257 unsigned int namelen;
2258
2259 lockname = res->lockname.name;
2260 namelen = res->lockname.len;
2261 BUG_ON(namelen > O2NM_MAX_NAME_LEN);
2262
2263 mlog(0, "%s:%.*s: sending deref to %d\n",
2264 dlm->name, namelen, lockname, res->owner);
2265 memset(&deref, 0, sizeof(deref));
2266 deref.node_idx = dlm->node_num;
2267 deref.namelen = namelen;
2268 memcpy(deref.name, lockname, namelen);
2269
2270 ret = o2net_send_message(DLM_DEREF_LOCKRES_MSG, dlm->key,
2271 &deref, sizeof(deref), res->owner, &r);
2272 if (ret < 0)
2273 mlog_errno(ret);
2274 else if (r < 0) {
2275 /* BAD. other node says I did not have a ref. */
2276 mlog(ML_ERROR,"while dropping ref on %s:%.*s "
2277 "(master=%u) got %d.\n", dlm->name, namelen,
2278 lockname, res->owner, r);
2279 dlm_print_one_lock_resource(res);
2280 BUG();
2281 }
2282 return ret;
2283}
2284
d74c9803
KH		 2285int dlm_deref_lockres_handler(struct o2net_msg *msg, u32 len, void *data,
2286 void **ret_data)
ba2bf218
KH		 2287{
2288 struct dlm_ctxt *dlm = data;
2289 struct dlm_deref_lockres *deref = (struct dlm_deref_lockres *)msg->buf;
2290 struct dlm_lock_resource *res = NULL;
2291 char *name;
2292 unsigned int namelen;
2293 int ret = -EINVAL;
2294 u8 node;
2295 unsigned int hash;
f3f85464
SM		 2296 struct dlm_work_item *item;
2297 int cleared = 0;
2298 int dispatch = 0;
ba2bf218
KH		 2299
2300 if (!dlm_grab(dlm))
2301 return 0;
2302
2303 name = deref->name;
2304 namelen = deref->namelen;
2305 node = deref->node_idx;
2306
2307 if (namelen > DLM_LOCKID_NAME_MAX) {
2308 mlog(ML_ERROR, "Invalid name length!");
2309 goto done;
2310 }
2311 if (deref->node_idx >= O2NM_MAX_NODES) {
2312 mlog(ML_ERROR, "Invalid node number: %u\n", node);
2313 goto done;
2314 }
2315
2316 hash = dlm_lockid_hash(name, namelen);
2317
2318 spin_lock(&dlm->spinlock);
2319 res = __dlm_lookup_lockres_full(dlm, name, namelen, hash);
2320 if (!res) {
2321 spin_unlock(&dlm->spinlock);
2322 mlog(ML_ERROR, "%s:%.*s: bad lockres name\n",
2323 dlm->name, namelen, name);
2324 goto done;
2325 }
2326 spin_unlock(&dlm->spinlock);
2327
2328 spin_lock(&res->spinlock);
f3f85464
SM		 2329 if (res->state & DLM_LOCK_RES_SETREF_INPROG)
2330 dispatch = 1;
2331 else {
2332 BUG_ON(res->state & DLM_LOCK_RES_DROPPING_REF);
2333 if (test_bit(node, res->refmap)) {
2334 dlm_lockres_clear_refmap_bit(node, res);
2335 cleared = 1;
2336 }
ba2bf218
KH		 2337 }
2338 spin_unlock(&res->spinlock);
2339
f3f85464
SM		 2340 if (!dispatch) {
2341 if (cleared)
2342 dlm_lockres_calc_usage(dlm, res);
2343 else {
2344 mlog(ML_ERROR, "%s:%.*s: node %u trying to drop ref "
2345 "but it is already dropped!\n", dlm->name,
2346 res->lockname.len, res->lockname.name, node);
2af37ce8 2347 dlm_print_one_lock_resource(res);
f3f85464
SM		 2348 }
2349 ret = 0;
2350 goto done;
2351 }
2352
2353 item = kzalloc(sizeof(*item), GFP_NOFS);
2354 if (!item) {
2355 ret = -ENOMEM;
2356 mlog_errno(ret);
2357 goto done;
2358 }
2359
2360 dlm_init_work_item(dlm, item, dlm_deref_lockres_worker, NULL);
2361 item->u.dl.deref_res = res;
2362 item->u.dl.deref_node = node;
2363
2364 spin_lock(&dlm->work_lock);
2365 list_add_tail(&item->list, &dlm->work_list);
2366 spin_unlock(&dlm->work_lock);
2367
2368 queue_work(dlm->dlm_worker, &dlm->dispatched_work);
2369 return 0;
2370
ba2bf218
KH		 2371done:
2372 if (res)
2373 dlm_lockres_put(res);
2374 dlm_put(dlm);
f3f85464 2375
ba2bf218
KH		 2376 return ret;
2377}
2378
f3f85464
SM		 2379static void dlm_deref_lockres_worker(struct dlm_work_item *item, void *data)
2380{
2381 struct dlm_ctxt *dlm;
2382 struct dlm_lock_resource *res;
2383 u8 node;
2384 u8 cleared = 0;
2385
2386 dlm = item->dlm;
2387 res = item->u.dl.deref_res;
2388 node = item->u.dl.deref_node;
2389
2390 spin_lock(&res->spinlock);
2391 BUG_ON(res->state & DLM_LOCK_RES_DROPPING_REF);
2392 if (test_bit(node, res->refmap)) {
2393 __dlm_wait_on_lockres_flags(res, DLM_LOCK_RES_SETREF_INPROG);
2394 dlm_lockres_clear_refmap_bit(node, res);
2395 cleared = 1;
2396 }
2397 spin_unlock(&res->spinlock);
2398
2399 if (cleared) {
2400 mlog(0, "%s:%.*s node %u ref dropped in dispatch\n",
2401 dlm->name, res->lockname.len, res->lockname.name, node);
2402 dlm_lockres_calc_usage(dlm, res);
2403 } else {
2404 mlog(ML_ERROR, "%s:%.*s: node %u trying to drop ref "
2405 "but it is already dropped!\n", dlm->name,
2406 res->lockname.len, res->lockname.name, node);
2af37ce8 2407 dlm_print_one_lock_resource(res);
f3f85464
SM		 2408 }
2409
2410 dlm_lockres_put(res);
2411}
2412
2f5bf1f2
SM		 2413/* Checks whether the lockres can be migrated. Returns 0 if yes, < 0
2414 * if not. If 0, numlocks is set to the number of locks in the lockres.
2415 */
2416static int dlm_is_lockres_migrateable(struct dlm_ctxt *dlm,
2417 struct dlm_lock_resource *res,
2418 int *numlocks)
2419{
2420 int ret;
2421 int i;
2422 int count = 0;
800deef3 2423 struct list_head *queue;
2f5bf1f2
SM		 2424 struct dlm_lock *lock;
2425
2426 assert_spin_locked(&res->spinlock);
2427
2428 ret = -EINVAL;
2429 if (res->owner == DLM_LOCK_RES_OWNER_UNKNOWN) {
2430 mlog(0, "cannot migrate lockres with unknown owner!\n");
2431 goto leave;
2432 }
2433
2434 if (res->owner != dlm->node_num) {
2435 mlog(0, "cannot migrate lockres this node doesn't own!\n");
2436 goto leave;
2437 }
2438
2439 ret = 0;
2440 queue = &res->granted;
2441 for (i = 0; i < 3; i++) {
800deef3 2442 list_for_each_entry(lock, queue, list) {
2f5bf1f2
SM		 2443 ++count;
2444 if (lock->ml.node == dlm->node_num) {
2445 mlog(0, "found a lock owned by this node still "
2446 "on the %s queue! will not migrate this "
2447 "lockres\n", (i == 0 ? "granted" :
2448 (i == 1 ? "converting" :
2449 "blocked")));
2450 ret = -ENOTEMPTY;
2451 goto leave;
2452 }
2453 }
2454 queue++;
2455 }
2456
2457 *numlocks = count;
2458 mlog(0, "migrateable lockres having %d locks\n", *numlocks);
2459
2460leave:
2461 return ret;
2462}
6714d8e8
KH		 2463
2464/*
2465 * DLM_MIGRATE_LOCKRES
2466 */
2467
2468
faf0ec9f
AB		 2469static int dlm_migrate_lockres(struct dlm_ctxt *dlm,
2470 struct dlm_lock_resource *res,
2471 u8 target)
6714d8e8
KH		 2472{
2473 struct dlm_master_list_entry *mle = NULL;
2474 struct dlm_master_list_entry *oldmle = NULL;
2475 struct dlm_migratable_lockres *mres = NULL;
2f5bf1f2 2476 int ret = 0;
6714d8e8
KH		 2477 const char *name;
2478 unsigned int namelen;
2479 int mle_added = 0;
2f5bf1f2
SM		 2480 int numlocks;
2481 int wake = 0;
6714d8e8
KH		 2482
2483 if (!dlm_grab(dlm))
2484 return -EINVAL;
2485
2486 name = res->lockname.name;
2487 namelen = res->lockname.len;
2488
2489 mlog(0, "migrating %.*s to %u\n", namelen, name, target);
2490
2491 /*
2492 * ensure this lockres is a proper candidate for migration
2493 */
2494 spin_lock(&res->spinlock);
2f5bf1f2
SM		 2495 ret = dlm_is_lockres_migrateable(dlm, res, &numlocks);
2496 if (ret < 0) {
6714d8e8
KH		 2497 spin_unlock(&res->spinlock);
2498 goto leave;
2499 }
6714d8e8
KH		 2500 spin_unlock(&res->spinlock);
2501
2502 /* no work to do */
2f5bf1f2 2503 if (numlocks == 0) {
6714d8e8 2504 mlog(0, "no locks were found on this lockres! done!\n");
6714d8e8
KH		 2505 goto leave;
2506 }
2507
2508 /*
2509 * preallocate up front
2510 * if this fails, abort
2511 */
2512
2513 ret = -ENOMEM;
ad8100e0 2514 mres = (struct dlm_migratable_lockres *) __get_free_page(GFP_NOFS);
6714d8e8
KH		 2515 if (!mres) {
2516 mlog_errno(ret);
2517 goto leave;
2518 }
2519
2520 mle = (struct dlm_master_list_entry *) kmem_cache_alloc(dlm_mle_cache,
ad8100e0 2521 GFP_NOFS);
6714d8e8
KH		 2522 if (!mle) {
2523 mlog_errno(ret);
2524 goto leave;
2525 }
2526 ret = 0;
2527
2528 /*
2529 * find a node to migrate the lockres to
2530 */
2531
2532 mlog(0, "picking a migration node\n");
2533 spin_lock(&dlm->spinlock);
2534 /* pick a new node */
2535 if (!test_bit(target, dlm->domain_map) ||
2536 target >= O2NM_MAX_NODES) {
2537 target = dlm_pick_migration_target(dlm, res);
2538 }
2539 mlog(0, "node %u chosen for migration\n", target);
2540
2541 if (target >= O2NM_MAX_NODES ||
2542 !test_bit(target, dlm->domain_map)) {
2543 /* target chosen is not alive */
2544 ret = -EINVAL;
2545 }
2546
2547 if (ret) {
2548 spin_unlock(&dlm->spinlock);
2549 goto fail;
2550 }
2551
2552 mlog(0, "continuing with target = %u\n", target);
2553
2554 /*
2555 * clear any existing master requests and
2556 * add the migration mle to the list
2557 */
2558 spin_lock(&dlm->master_lock);
2559 ret = dlm_add_migration_mle(dlm, res, mle, &oldmle, name,
2560 namelen, target, dlm->node_num);
2561 spin_unlock(&dlm->master_lock);
2562 spin_unlock(&dlm->spinlock);
2563
2564 if (ret == -EEXIST) {
2565 mlog(0, "another process is already migrating it\n");
2566 goto fail;
2567 }
2568 mle_added = 1;
2569
2570 /*
2571 * set the MIGRATING flag and flush asts
2572 * if we fail after this we need to re-dirty the lockres
2573 */
2574 if (dlm_mark_lockres_migrating(dlm, res, target) < 0) {
2575 mlog(ML_ERROR, "tried to migrate %.*s to %u, but "
2576 "the target went down.\n", res->lockname.len,
2577 res->lockname.name, target);
2578 spin_lock(&res->spinlock);
2579 res->state &= ~DLM_LOCK_RES_MIGRATING;
a6fa3640 2580 wake = 1;
6714d8e8
KH		 2581 spin_unlock(&res->spinlock);
2582 ret = -EINVAL;
2583 }
2584
2585fail:
2586 if (oldmle) {
2587 /* master is known, detach if not already detached */
2588 dlm_mle_detach_hb_events(dlm, oldmle);
2589 dlm_put_mle(oldmle);
2590 }
2591
2592 if (ret < 0) {
2593 if (mle_added) {
2594 dlm_mle_detach_hb_events(dlm, mle);
2595 dlm_put_mle(mle);
2596 } else if (mle) {
2597 kmem_cache_free(dlm_mle_cache, mle);
2598 }
2599 goto leave;
2600 }
2601
2602 /*
2603 * at this point, we have a migration target, an mle
2604 * in the master list, and the MIGRATING flag set on
2605 * the lockres
2606 */
2607
1cd04dbe
KH		 2608 /* now that remote nodes are spinning on the MIGRATING flag,
2609 * ensure that all assert_master work is flushed. */
2610 flush_workqueue(dlm->dlm_worker);
6714d8e8
KH		 2611
2612 /* get an extra reference on the mle.
2613 * otherwise the assert_master from the new
2614 * master will destroy this.
2615 * also, make sure that all callers of dlm_get_mle
2616 * take both dlm->spinlock and dlm->master_lock */
2617 spin_lock(&dlm->spinlock);
2618 spin_lock(&dlm->master_lock);
a2bf0477 2619 dlm_get_mle_inuse(mle);
6714d8e8
KH		 2620 spin_unlock(&dlm->master_lock);
2621 spin_unlock(&dlm->spinlock);
2622
2623 /* notify new node and send all lock state */
2624 /* call send_one_lockres with migration flag.
2625 * this serves as notice to the target node that a
2626 * migration is starting. */
2627 ret = dlm_send_one_lockres(dlm, res, mres, target,
2628 DLM_MRES_MIGRATION);
2629
2630 if (ret < 0) {
2631 mlog(0, "migration to node %u failed with %d\n",
2632 target, ret);
2633 /* migration failed, detach and clean up mle */
2634 dlm_mle_detach_hb_events(dlm, mle);
2635 dlm_put_mle(mle);
a2bf0477
KH		 2636 dlm_put_mle_inuse(mle);
2637 spin_lock(&res->spinlock);
2638 res->state &= ~DLM_LOCK_RES_MIGRATING;
a6fa3640 2639 wake = 1;
a2bf0477 2640 spin_unlock(&res->spinlock);
6714d8e8
KH		 2641 goto leave;
2642 }
2643
2644 /* at this point, the target sends a message to all nodes,
2645 * (using dlm_do_migrate_request). this node is skipped since
2646 * we had to put an mle in the list to begin the process. this
2647 * node now waits for target to do an assert master. this node
2648 * will be the last one notified, ensuring that the migration
2649 * is complete everywhere. if the target dies while this is
2650 * going on, some nodes could potentially see the target as the
2651 * master, so it is important that my recovery finds the migration
2652 * mle and sets the master to UNKNONWN. */
2653
2654
2655 /* wait for new node to assert master */
2656 while (1) {
2657 ret = wait_event_interruptible_timeout(mle->wq,
2658 (atomic_read(&mle->woken) == 1),
2659 msecs_to_jiffies(5000));
2660
2661 if (ret >= 0) {
2662 if (atomic_read(&mle->woken) == 1 ||
2663 res->owner == target)
2664 break;
2665
1cd04dbe
KH		 2666 mlog(0, "%s:%.*s: timed out during migration\n",
2667 dlm->name, res->lockname.len, res->lockname.name);
e2faea4c
KH		 2668 /* avoid hang during shutdown when migrating lockres
2669 * to a node which also goes down */
2670 if (dlm_is_node_dead(dlm, target)) {
aa852354
KH		 2671 mlog(0, "%s:%.*s: expected migration "
2672 "target %u is no longer up, restarting\n",
e2faea4c
KH		 2673 dlm->name, res->lockname.len,
2674 res->lockname.name, target);
1cd04dbe
KH		 2675 ret = -EINVAL;
2676 /* migration failed, detach and clean up mle */
2677 dlm_mle_detach_hb_events(dlm, mle);
2678 dlm_put_mle(mle);
2679 dlm_put_mle_inuse(mle);
2680 spin_lock(&res->spinlock);
2681 res->state &= ~DLM_LOCK_RES_MIGRATING;
a6fa3640 2682 wake = 1;
1cd04dbe
KH		 2683 spin_unlock(&res->spinlock);
2684 goto leave;
e2faea4c 2685 }
1cd04dbe
KH		 2686 } else
2687 mlog(0, "%s:%.*s: caught signal during migration\n",
2688 dlm->name, res->lockname.len, res->lockname.name);
6714d8e8
KH		 2689 }
2690
2691 /* all done, set the owner, clear the flag */
2692 spin_lock(&res->spinlock);
2693 dlm_set_lockres_owner(dlm, res, target);
2694 res->state &= ~DLM_LOCK_RES_MIGRATING;
2695 dlm_remove_nonlocal_locks(dlm, res);
2696 spin_unlock(&res->spinlock);
2697 wake_up(&res->wq);
2698
2699 /* master is known, detach if not already detached */
2700 dlm_mle_detach_hb_events(dlm, mle);
a2bf0477 2701 dlm_put_mle_inuse(mle);
6714d8e8
KH		 2702 ret = 0;
2703
2704 dlm_lockres_calc_usage(dlm, res);
2705
2706leave:
2707 /* re-dirty the lockres if we failed */
2708 if (ret < 0)
2709 dlm_kick_thread(dlm, res);
2710
a6fa3640
KH		 2711 /* wake up waiters if the MIGRATING flag got set
2712 * but migration failed */
2713 if (wake)
2714 wake_up(&res->wq);
2715
6714d8e8
KH		 2716 /* TODO: cleanup */
2717 if (mres)
2718 free_page((unsigned long)mres);
2719
2720 dlm_put(dlm);
2721
2722 mlog(0, "returning %d\n", ret);
2723 return ret;
2724}
6714d8e8 2725
ba2bf218
KH		 2726#define DLM_MIGRATION_RETRY_MS 100
2727
2728/* Should be called only after beginning the domain leave process.
2729 * There should not be any remaining locks on nonlocal lock resources,
2730 * and there should be no local locks left on locally mastered resources.
2731 *
2732 * Called with the dlm spinlock held, may drop it to do migration, but
2733 * will re-acquire before exit.
2734 *
2735 * Returns: 1 if dlm->spinlock was dropped/retaken, 0 if never dropped */
2736int dlm_empty_lockres(struct dlm_ctxt *dlm, struct dlm_lock_resource *res)
2737{
2738 int ret;
2739 int lock_dropped = 0;
2f5bf1f2 2740 int numlocks;
ba2bf218 2741
b36c3f84 2742 spin_lock(&res->spinlock);
ba2bf218
KH		 2743 if (res->owner != dlm->node_num) {
2744 if (!__dlm_lockres_unused(res)) {
2745 mlog(ML_ERROR, "%s:%.*s: this node is not master, "
2746 "trying to free this but locks remain\n",
2747 dlm->name, res->lockname.len, res->lockname.name);
2748 }
b36c3f84 2749 spin_unlock(&res->spinlock);
ba2bf218
KH		 2750 goto leave;
2751 }
2f5bf1f2
SM		 2752
2753 /* No need to migrate a lockres having no locks */
2754 ret = dlm_is_lockres_migrateable(dlm, res, &numlocks);
2755 if (ret >= 0 && numlocks == 0) {
2756 spin_unlock(&res->spinlock);
2757 goto leave;
2758 }
b36c3f84 2759 spin_unlock(&res->spinlock);
ba2bf218
KH		 2760
2761 /* Wheee! Migrate lockres here! Will sleep so drop spinlock. */
2762 spin_unlock(&dlm->spinlock);
2763 lock_dropped = 1;
2764 while (1) {
2765 ret = dlm_migrate_lockres(dlm, res, O2NM_MAX_NODES);
2766 if (ret >= 0)
2767 break;
2768 if (ret == -ENOTEMPTY) {
2769 mlog(ML_ERROR, "lockres %.*s still has local locks!\n",
2770 res->lockname.len, res->lockname.name);
2771 BUG();
2772 }
2773
2774 mlog(0, "lockres %.*s: migrate failed, "
2775 "retrying\n", res->lockname.len,
2776 res->lockname.name);
2777 msleep(DLM_MIGRATION_RETRY_MS);
2778 }
2779 spin_lock(&dlm->spinlock);
2780leave:
2781 return lock_dropped;
2782}
2783
6714d8e8
KH		 2784int dlm_lock_basts_flushed(struct dlm_ctxt *dlm, struct dlm_lock *lock)
2785{
2786 int ret;
2787 spin_lock(&dlm->ast_lock);
2788 spin_lock(&lock->spinlock);
2789 ret = (list_empty(&lock->bast_list) && !lock->bast_pending);
2790 spin_unlock(&lock->spinlock);
2791 spin_unlock(&dlm->ast_lock);
2792 return ret;
2793}
2794
2795static int dlm_migration_can_proceed(struct dlm_ctxt *dlm,
2796 struct dlm_lock_resource *res,
2797 u8 mig_target)
2798{
2799 int can_proceed;
2800 spin_lock(&res->spinlock);
2801 can_proceed = !!(res->state & DLM_LOCK_RES_MIGRATING);
2802 spin_unlock(&res->spinlock);
2803
2804 /* target has died, so make the caller break out of the
2805 * wait_event, but caller must recheck the domain_map */
2806 spin_lock(&dlm->spinlock);
2807 if (!test_bit(mig_target, dlm->domain_map))
2808 can_proceed = 1;
2809 spin_unlock(&dlm->spinlock);
2810 return can_proceed;
2811}
2812
faf0ec9f
AB		 2813static int dlm_lockres_is_dirty(struct dlm_ctxt *dlm,
2814 struct dlm_lock_resource *res)
6714d8e8
KH		 2815{
2816 int ret;
2817 spin_lock(&res->spinlock);
2818 ret = !!(res->state & DLM_LOCK_RES_DIRTY);
2819 spin_unlock(&res->spinlock);
2820 return ret;
2821}
2822
2823
2824static int dlm_mark_lockres_migrating(struct dlm_ctxt *dlm,
2825 struct dlm_lock_resource *res,
2826 u8 target)
2827{
2828 int ret = 0;
2829
2830 mlog(0, "dlm_mark_lockres_migrating: %.*s, from %u to %u\n",
2831 res->lockname.len, res->lockname.name, dlm->node_num,
2832 target);
2833 /* need to set MIGRATING flag on lockres. this is done by
2834 * ensuring that all asts have been flushed for this lockres. */
2835 spin_lock(&res->spinlock);
2836 BUG_ON(res->migration_pending);
2837 res->migration_pending = 1;
2838 /* strategy is to reserve an extra ast then release
2839 * it below, letting the release do all of the work */
2840 __dlm_lockres_reserve_ast(res);
2841 spin_unlock(&res->spinlock);
2842
ddc09c8d 2843 /* now flush all the pending asts */
6714d8e8 2844 dlm_kick_thread(dlm, res);
ddc09c8d
KH		 2845 /* before waiting on DIRTY, block processes which may
2846 * try to dirty the lockres before MIGRATING is set */
2847 spin_lock(&res->spinlock);
2848 BUG_ON(res->state & DLM_LOCK_RES_BLOCK_DIRTY);
2849 res->state |= DLM_LOCK_RES_BLOCK_DIRTY;
2850 spin_unlock(&res->spinlock);
2851 /* now wait on any pending asts and the DIRTY state */
6714d8e8
KH		 2852 wait_event(dlm->ast_wq, !dlm_lockres_is_dirty(dlm, res));
2853 dlm_lockres_release_ast(dlm, res);
2854
2855 mlog(0, "about to wait on migration_wq, dirty=%s\n",
2856 res->state & DLM_LOCK_RES_DIRTY ? "yes" : "no");
2857 /* if the extra ref we just put was the final one, this
2858 * will pass thru immediately. otherwise, we need to wait
2859 * for the last ast to finish. */
2860again:
2861 ret = wait_event_interruptible_timeout(dlm->migration_wq,
2862 dlm_migration_can_proceed(dlm, res, target),
2863 msecs_to_jiffies(1000));
2864 if (ret < 0) {
2865 mlog(0, "woken again: migrating? %s, dead? %s\n",
2866 res->state & DLM_LOCK_RES_MIGRATING ? "yes":"no",
2867 test_bit(target, dlm->domain_map) ? "no":"yes");
2868 } else {
2869 mlog(0, "all is well: migrating? %s, dead? %s\n",
2870 res->state & DLM_LOCK_RES_MIGRATING ? "yes":"no",
2871 test_bit(target, dlm->domain_map) ? "no":"yes");
2872 }
2873 if (!dlm_migration_can_proceed(dlm, res, target)) {
2874 mlog(0, "trying again...\n");
2875 goto again;
2876 }
ddc09c8d
KH		 2877 /* now that we are sure the MIGRATING state is there, drop
2878 * the unneded state which blocked threads trying to DIRTY */
2879 spin_lock(&res->spinlock);
2880 BUG_ON(!(res->state & DLM_LOCK_RES_BLOCK_DIRTY));
2881 BUG_ON(!(res->state & DLM_LOCK_RES_MIGRATING));
2882 res->state &= ~DLM_LOCK_RES_BLOCK_DIRTY;
2883 spin_unlock(&res->spinlock);
6714d8e8
KH		 2884
2885 /* did the target go down or die? */
2886 spin_lock(&dlm->spinlock);
2887 if (!test_bit(target, dlm->domain_map)) {
2888 mlog(ML_ERROR, "aha. migration target %u just went down\n",
2889 target);
2890 ret = -EHOSTDOWN;
2891 }
2892 spin_unlock(&dlm->spinlock);
2893
2894 /*
2895 * at this point:
2896 *
2897 * o the DLM_LOCK_RES_MIGRATING flag is set
2898 * o there are no pending asts on this lockres
2899 * o all processes trying to reserve an ast on this
2900 * lockres must wait for the MIGRATING flag to clear
2901 */
2902 return ret;
2903}
2904
2905/* last step in the migration process.
2906 * original master calls this to free all of the dlm_lock
2907 * structures that used to be for other nodes. */
2908static void dlm_remove_nonlocal_locks(struct dlm_ctxt *dlm,
2909 struct dlm_lock_resource *res)
2910{
6714d8e8 2911 struct list_head *queue = &res->granted;
ba2bf218 2912 int i, bit;
800deef3 2913 struct dlm_lock *lock, *next;
6714d8e8
KH		 2914
2915 assert_spin_locked(&res->spinlock);
2916
2917 BUG_ON(res->owner == dlm->node_num);
2918
2919 for (i=0; i<3; i++) {
800deef3 2920 list_for_each_entry_safe(lock, next, queue, list) {
6714d8e8
KH		 2921 if (lock->ml.node != dlm->node_num) {
2922 mlog(0, "putting lock for node %u\n",
2923 lock->ml.node);
2924 /* be extra careful */
2925 BUG_ON(!list_empty(&lock->ast_list));
2926 BUG_ON(!list_empty(&lock->bast_list));
2927 BUG_ON(lock->ast_pending);
2928 BUG_ON(lock->bast_pending);
ba2bf218 2929 dlm_lockres_clear_refmap_bit(lock->ml.node, res);
6714d8e8
KH		 2930 list_del_init(&lock->list);
2931 dlm_lock_put(lock);
2c5c54ac
SM		 2932 /* In a normal unlock, we would have added a
2933 * DLM_UNLOCK_FREE_LOCK action. Force it. */
2934 dlm_lock_put(lock);
6714d8e8
KH		 2935 }
2936 }
2937 queue++;
2938 }
ba2bf218
KH		 2939 bit = 0;
2940 while (1) {
2941 bit = find_next_bit(res->refmap, O2NM_MAX_NODES, bit);
2942 if (bit >= O2NM_MAX_NODES)
2943 break;
2944 /* do not clear the local node reference, if there is a
2945 * process holding this, let it drop the ref itself */
2946 if (bit != dlm->node_num) {
2947 mlog(0, "%s:%.*s: node %u had a ref to this "
2948 "migrating lockres, clearing\n", dlm->name,
2949 res->lockname.len, res->lockname.name, bit);
2950 dlm_lockres_clear_refmap_bit(bit, res);
2951 }
2952 bit++;
2953 }
6714d8e8
KH		 2954}
2955
2956/* for now this is not too intelligent. we will
2957 * need stats to make this do the right thing.
2958 * this just finds the first lock on one of the
2959 * queues and uses that node as the target. */
2960static u8 dlm_pick_migration_target(struct dlm_ctxt *dlm,
2961 struct dlm_lock_resource *res)
2962{
2963 int i;
2964 struct list_head *queue = &res->granted;
6714d8e8
KH		 2965 struct dlm_lock *lock;
2966 int nodenum;
2967
2968 assert_spin_locked(&dlm->spinlock);
2969
2970 spin_lock(&res->spinlock);
2971 for (i=0; i<3; i++) {
800deef3 2972 list_for_each_entry(lock, queue, list) {
6714d8e8
KH		 2973 /* up to the caller to make sure this node
2974 * is alive */
6714d8e8
KH		 2975 if (lock->ml.node != dlm->node_num) {
2976 spin_unlock(&res->spinlock);
2977 return lock->ml.node;
2978 }
2979 }
2980 queue++;
2981 }
2982 spin_unlock(&res->spinlock);
2983 mlog(0, "have not found a suitable target yet! checking domain map\n");
2984
2985 /* ok now we're getting desperate. pick anyone alive. */
2986 nodenum = -1;
2987 while (1) {
2988 nodenum = find_next_bit(dlm->domain_map,
2989 O2NM_MAX_NODES, nodenum+1);
2990 mlog(0, "found %d in domain map\n", nodenum);
2991 if (nodenum >= O2NM_MAX_NODES)
2992 break;
2993 if (nodenum != dlm->node_num) {
2994 mlog(0, "picking %d\n", nodenum);
2995 return nodenum;
2996 }
2997 }
2998
2999 mlog(0, "giving up. no master to migrate to\n");
3000 return DLM_LOCK_RES_OWNER_UNKNOWN;
3001}
3002
3003
3004
3005/* this is called by the new master once all lockres
3006 * data has been received */
3007static int dlm_do_migrate_request(struct dlm_ctxt *dlm,
3008 struct dlm_lock_resource *res,
3009 u8 master, u8 new_master,
3010 struct dlm_node_iter *iter)
3011{
3012 struct dlm_migrate_request migrate;
2b832564 3013 int ret, skip, status = 0;
6714d8e8
KH		 3014 int nodenum;
3015
3016 memset(&migrate, 0, sizeof(migrate));
3017 migrate.namelen = res->lockname.len;
3018 memcpy(migrate.name, res->lockname.name, migrate.namelen);
3019 migrate.new_master = new_master;
3020 migrate.master = master;
3021
3022 ret = 0;
3023
3024 /* send message to all nodes, except the master and myself */
3025 while ((nodenum = dlm_node_iter_next(iter)) >= 0) {
3026 if (nodenum == master ||
3027 nodenum == new_master)
3028 continue;
3029
2b832564
SM		 3030 /* We could race exit domain. If exited, skip. */
3031 spin_lock(&dlm->spinlock);
3032 skip = (!test_bit(nodenum, dlm->domain_map));
3033 spin_unlock(&dlm->spinlock);
3034 if (skip) {
3035 clear_bit(nodenum, iter->node_map);
3036 continue;
3037 }
3038
6714d8e8
KH		 3039 ret = o2net_send_message(DLM_MIGRATE_REQUEST_MSG, dlm->key,
3040 &migrate, sizeof(migrate), nodenum,
3041 &status);
2b832564
SM		 3042 if (ret < 0) {
3043 mlog(0, "migrate_request returned %d!\n", ret);
3044 if (!dlm_is_host_down(ret)) {
3045 mlog(ML_ERROR, "unhandled error=%d!\n", ret);
3046 BUG();
3047 }
3048 clear_bit(nodenum, iter->node_map);
3049 ret = 0;
3050 } else if (status < 0) {
6714d8e8
KH		 3051 mlog(0, "migrate request (node %u) returned %d!\n",
3052 nodenum, status);
3053 ret = status;
ba2bf218
KH		 3054 } else if (status == DLM_MIGRATE_RESPONSE_MASTERY_REF) {
3055 /* during the migration request we short-circuited
3056 * the mastery of the lockres. make sure we have
3057 * a mastery ref for nodenum */
3058 mlog(0, "%s:%.*s: need ref for node %u\n",
3059 dlm->name, res->lockname.len, res->lockname.name,
3060 nodenum);
3061 spin_lock(&res->spinlock);
3062 dlm_lockres_set_refmap_bit(nodenum, res);
3063 spin_unlock(&res->spinlock);
6714d8e8
KH		 3064 }
3065 }
3066
3067 if (ret < 0)
3068 mlog_errno(ret);
3069
3070 mlog(0, "returning ret=%d\n", ret);
3071 return ret;
3072}
3073
3074
3075/* if there is an existing mle for this lockres, we now know who the master is.
3076 * (the one who sent us *this* message) we can clear it up right away.
3077 * since the process that put the mle on the list still has a reference to it,
3078 * we can unhash it now, set the master and wake the process. as a result,
3079 * we will have no mle in the list to start with. now we can add an mle for
3080 * the migration and this should be the only one found for those scanning the
3081 * list. */
d74c9803
KH		 3082int dlm_migrate_request_handler(struct o2net_msg *msg, u32 len, void *data,
3083 void **ret_data)
6714d8e8
KH		 3084{
3085 struct dlm_ctxt *dlm = data;
3086 struct dlm_lock_resource *res = NULL;
3087 struct dlm_migrate_request *migrate = (struct dlm_migrate_request *) msg->buf;
3088 struct dlm_master_list_entry *mle = NULL, *oldmle = NULL;
3089 const char *name;
a3d33291 3090 unsigned int namelen, hash;
6714d8e8
KH		 3091 int ret = 0;
3092
3093 if (!dlm_grab(dlm))
3094 return -EINVAL;
3095
3096 name = migrate->name;
3097 namelen = migrate->namelen;
a3d33291 3098 hash = dlm_lockid_hash(name, namelen);
6714d8e8
KH		 3099
3100 /* preallocate.. if this fails, abort */
3101 mle = (struct dlm_master_list_entry *) kmem_cache_alloc(dlm_mle_cache,
ad8100e0 3102 GFP_NOFS);
6714d8e8
KH		 3103
3104 if (!mle) {
3105 ret = -ENOMEM;
3106 goto leave;
3107 }
3108
3109 /* check for pre-existing lock */
3110 spin_lock(&dlm->spinlock);
a3d33291 3111 res = __dlm_lookup_lockres(dlm, name, namelen, hash);
6714d8e8
KH		 3112 spin_lock(&dlm->master_lock);
3113
3114 if (res) {
3115 spin_lock(&res->spinlock);
3116 if (res->state & DLM_LOCK_RES_RECOVERING) {
3117 /* if all is working ok, this can only mean that we got
3118 * a migrate request from a node that we now see as
3119 * dead. what can we do here? drop it to the floor? */
3120 spin_unlock(&res->spinlock);
3121 mlog(ML_ERROR, "Got a migrate request, but the "
3122 "lockres is marked as recovering!");
3123 kmem_cache_free(dlm_mle_cache, mle);
3124 ret = -EINVAL; /* need a better solution */
3125 goto unlock;
3126 }
3127 res->state |= DLM_LOCK_RES_MIGRATING;
3128 spin_unlock(&res->spinlock);
3129 }
3130
3131 /* ignore status. only nonzero status would BUG. */
3132 ret = dlm_add_migration_mle(dlm, res, mle, &oldmle,
3133 name, namelen,
3134 migrate->new_master,
3135 migrate->master);
3136
3137unlock:
3138 spin_unlock(&dlm->master_lock);
3139 spin_unlock(&dlm->spinlock);
3140
3141 if (oldmle) {
3142 /* master is known, detach if not already detached */
3143 dlm_mle_detach_hb_events(dlm, oldmle);
3144 dlm_put_mle(oldmle);
3145 }
3146
3147 if (res)
3148 dlm_lockres_put(res);
3149leave:
3150 dlm_put(dlm);
3151 return ret;
3152}
3153
3154/* must be holding dlm->spinlock and dlm->master_lock
3155 * when adding a migration mle, we can clear any other mles
3156 * in the master list because we know with certainty that
3157 * the master is "master". so we remove any old mle from
3158 * the list after setting it's master field, and then add
3159 * the new migration mle. this way we can hold with the rule
3160 * of having only one mle for a given lock name at all times. */
3161static int dlm_add_migration_mle(struct dlm_ctxt *dlm,
3162 struct dlm_lock_resource *res,
3163 struct dlm_master_list_entry *mle,
3164 struct dlm_master_list_entry **oldmle,
3165 const char *name, unsigned int namelen,
3166 u8 new_master, u8 master)
3167{
3168 int found;
3169 int ret = 0;
3170
3171 *oldmle = NULL;
3172
3173 mlog_entry_void();
3174
3175 assert_spin_locked(&dlm->spinlock);
3176 assert_spin_locked(&dlm->master_lock);
3177
3178 /* caller is responsible for any ref taken here on oldmle */
3179 found = dlm_find_mle(dlm, oldmle, (char *)name, namelen);
3180 if (found) {
3181 struct dlm_master_list_entry *tmp = *oldmle;
3182 spin_lock(&tmp->spinlock);
3183 if (tmp->type == DLM_MLE_MIGRATION) {
3184 if (master == dlm->node_num) {
3185 /* ah another process raced me to it */
3186 mlog(0, "tried to migrate %.*s, but some "
3187 "process beat me to it\n",
3188 namelen, name);
3189 ret = -EEXIST;
3190 } else {
3191 /* bad. 2 NODES are trying to migrate! */
3192 mlog(ML_ERROR, "migration error mle: "
3193 "master=%u new_master=%u // request: "
3194 "master=%u new_master=%u // "
3195 "lockres=%.*s\n",
3196 tmp->master, tmp->new_master,
3197 master, new_master,
3198 namelen, name);
3199 BUG();
3200 }
3201 } else {
3202 /* this is essentially what assert_master does */
3203 tmp->master = master;
3204 atomic_set(&tmp->woken, 1);
3205 wake_up(&tmp->wq);
1c084577
SM		 3206 /* remove it so that only one mle will be found */
3207 __dlm_unlink_mle(dlm, tmp);
ba2bf218
KH		 3208 __dlm_mle_detach_hb_events(dlm, tmp);
3209 ret = DLM_MIGRATE_RESPONSE_MASTERY_REF;
3210 mlog(0, "%s:%.*s: master=%u, newmaster=%u, "
3211 "telling master to get ref for cleared out mle "
3212 "during migration\n", dlm->name, namelen, name,
3213 master, new_master);
6714d8e8
KH		 3214 }
3215 spin_unlock(&tmp->spinlock);
3216 }
3217
3218 /* now add a migration mle to the tail of the list */
3219 dlm_init_mle(mle, DLM_MLE_MIGRATION, dlm, res, name, namelen);
3220 mle->new_master = new_master;
ba2bf218
KH		 3221 /* the new master will be sending an assert master for this.
3222 * at that point we will get the refmap reference */
6714d8e8
KH		 3223 mle->master = master;
3224 /* do this for consistency with other mle types */
3225 set_bit(new_master, mle->maybe_map);
1c084577 3226 __dlm_insert_mle(dlm, mle);
6714d8e8
KH		 3227
3228 return ret;
3229}
3230
c2cd4a44
SM		 3231/*
3232 * Sets the owner of the lockres, associated to the mle, to UNKNOWN
3233 */
3234static struct dlm_lock_resource *dlm_reset_mleres_owner(struct dlm_ctxt *dlm,
3235 struct dlm_master_list_entry *mle)
3236{
3237 struct dlm_lock_resource *res;
3238 unsigned int hash;
3239
3240 /* Find the lockres associated to the mle and set its owner to UNK */
3241 hash = dlm_lockid_hash(mle->u.mlename.name, mle->u.mlename.len);
3242 res = __dlm_lookup_lockres(dlm, mle->u.mlename.name, mle->u.mlename.len,
3243 hash);
3244 if (res) {
3245 spin_unlock(&dlm->master_lock);
3246
3247 /* move lockres onto recovery list */
3248 spin_lock(&res->spinlock);
3249 dlm_set_lockres_owner(dlm, res, DLM_LOCK_RES_OWNER_UNKNOWN);
3250 dlm_move_lockres_to_recovery_list(dlm, res);
3251 spin_unlock(&res->spinlock);
3252 dlm_lockres_put(res);
3253
3254 /* about to get rid of mle, detach from heartbeat */
3255 __dlm_mle_detach_hb_events(dlm, mle);
3256
3257 /* dump the mle */
3258 spin_lock(&dlm->master_lock);
3259 __dlm_put_mle(mle);
3260 spin_unlock(&dlm->master_lock);
3261 }
3262
3263 return res;
3264}
3265
3266static void dlm_clean_migration_mle(struct dlm_ctxt *dlm,
3267 struct dlm_master_list_entry *mle)
3268{
3269 __dlm_mle_detach_hb_events(dlm, mle);
3270
3271 spin_lock(&mle->spinlock);
3272 __dlm_unlink_mle(dlm, mle);
3273 atomic_set(&mle->woken, 1);
3274 spin_unlock(&mle->spinlock);
3275
3276 wake_up(&mle->wq);
3277}
3278
3279static void dlm_clean_block_mle(struct dlm_ctxt *dlm,
3280 struct dlm_master_list_entry *mle, u8 dead_node)
3281{
3282 int bit;
3283
3284 BUG_ON(mle->type != DLM_MLE_BLOCK);
3285
3286 spin_lock(&mle->spinlock);
3287 bit = find_next_bit(mle->maybe_map, O2NM_MAX_NODES, 0);
3288 if (bit != dead_node) {
3289 mlog(0, "mle found, but dead node %u would not have been "
3290 "master\n", dead_node);
3291 spin_unlock(&mle->spinlock);
3292 } else {
3293 /* Must drop the refcount by one since the assert_master will
3294 * never arrive. This may result in the mle being unlinked and
3295 * freed, but there may still be a process waiting in the
3296 * dlmlock path which is fine. */
3297 mlog(0, "node %u was expected master\n", dead_node);
3298 atomic_set(&mle->woken, 1);
3299 spin_unlock(&mle->spinlock);
3300 wake_up(&mle->wq);
3301
3302 /* Do not need events any longer, so detach from heartbeat */
3303 __dlm_mle_detach_hb_events(dlm, mle);
3304 __dlm_put_mle(mle);
3305 }
3306}
6714d8e8
KH		 3307
3308void dlm_clean_master_list(struct dlm_ctxt *dlm, u8 dead_node)
3309{
2ed6c750 3310 struct dlm_master_list_entry *mle;
6714d8e8 3311 struct dlm_lock_resource *res;
2ed6c750
SM		 3312 struct hlist_head *bucket;
3313 struct hlist_node *list;
3314 unsigned int i;
6714d8e8
KH		 3315
3316 mlog_entry("dlm=%s, dead node=%u\n", dlm->name, dead_node);
3317top:
3318 assert_spin_locked(&dlm->spinlock);
3319
3320 /* clean the master list */
3321 spin_lock(&dlm->master_lock);
2ed6c750
SM		 3322 for (i = 0; i < DLM_HASH_BUCKETS; i++) {
3323 bucket = dlm_master_hash(dlm, i);
3324 hlist_for_each(list, bucket) {
3325 mle = hlist_entry(list, struct dlm_master_list_entry,
3326 master_hash_node);
3327
67ae1f06
SM		 3328 BUG_ON(mle->type != DLM_MLE_BLOCK &&
3329 mle->type != DLM_MLE_MASTER &&
3330 mle->type != DLM_MLE_MIGRATION);
3331
3332 /* MASTER mles are initiated locally. The waiting
3333 * process will notice the node map change shortly.
3334 * Let that happen as normal. */
3335 if (mle->type == DLM_MLE_MASTER)
3336 continue;
3337
3338 /* BLOCK mles are initiated by other nodes. Need to
3339 * clean up if the dead node would have been the
3340 * master. */
3341 if (mle->type == DLM_MLE_BLOCK) {
3342 dlm_clean_block_mle(dlm, mle, dead_node);
3343 continue;
3344 }
6714d8e8 3345
67ae1f06
SM		 3346 /* Everything else is a MIGRATION mle */
3347
3348 /* The rule for MIGRATION mles is that the master
3349 * becomes UNKNOWN if *either* the original or the new
3350 * master dies. All UNKNOWN lockres' are sent to
3351 * whichever node becomes the recovery master. The new
3352 * master is responsible for determining if there is
3353 * still a master for this lockres, or if he needs to
3354 * take over mastery. Either way, this node should
3355 * expect another message to resolve this. */
3356
3357 if (mle->master != dead_node &&
3358 mle->new_master != dead_node)
3359 continue;
3360
3361 /* If we have reached this point, this mle needs to be
3362 * removed from the list and freed. */
3363 dlm_clean_migration_mle(dlm, mle);
3364
3365 mlog(0, "%s: node %u died during migration from "
3366 "%u to %u!\n", dlm->name, dead_node, mle->master,
3367 mle->new_master);
3368
3369 /* If we find a lockres associated with the mle, we've
3370 * hit this rare case that messes up our lock ordering.
3371 * If so, we need to drop the master lock so that we can
3372 * take the lockres lock, meaning that we will have to
3373 * restart from the head of list. */
3374 res = dlm_reset_mleres_owner(dlm, mle);
3375 if (res)
3376 /* restart */
3377 goto top;
3378
3379 /* This may be the last reference */
3380 __dlm_put_mle(mle);
6714d8e8 3381 }
2ed6c750 3382 }
6714d8e8
KH		 3383 spin_unlock(&dlm->master_lock);
3384}
3385
6714d8e8
KH		 3386int dlm_finish_migration(struct dlm_ctxt *dlm, struct dlm_lock_resource *res,
3387 u8 old_master)
3388{
3389 struct dlm_node_iter iter;
3390 int ret = 0;
3391
3392 spin_lock(&dlm->spinlock);
3393 dlm_node_iter_init(dlm->domain_map, &iter);
3394 clear_bit(old_master, iter.node_map);
3395 clear_bit(dlm->node_num, iter.node_map);
3396 spin_unlock(&dlm->spinlock);
3397
ba2bf218
KH		 3398 /* ownership of the lockres is changing. account for the
3399 * mastery reference here since old_master will briefly have
3400 * a reference after the migration completes */
3401 spin_lock(&res->spinlock);
3402 dlm_lockres_set_refmap_bit(old_master, res);
3403 spin_unlock(&res->spinlock);
3404
6714d8e8
KH		 3405 mlog(0, "now time to do a migrate request to other nodes\n");
3406 ret = dlm_do_migrate_request(dlm, res, old_master,
3407 dlm->node_num, &iter);
3408 if (ret < 0) {
3409 mlog_errno(ret);
3410 goto leave;
3411 }
3412
3413 mlog(0, "doing assert master of %.*s to all except the original node\n",
3414 res->lockname.len, res->lockname.name);
3415 /* this call now finishes out the nodemap
3416 * even if one or more nodes die */
ba2bf218 3417 ret = dlm_do_assert_master(dlm, res, iter.node_map,
6714d8e8
KH		 3418 DLM_ASSERT_MASTER_FINISH_MIGRATION);
3419 if (ret < 0) {
3420 /* no longer need to retry. all living nodes contacted. */
3421 mlog_errno(ret);
3422 ret = 0;
3423 }
3424
3425 memset(iter.node_map, 0, sizeof(iter.node_map));
3426 set_bit(old_master, iter.node_map);
3427 mlog(0, "doing assert master of %.*s back to %u\n",
3428 res->lockname.len, res->lockname.name, old_master);
ba2bf218 3429 ret = dlm_do_assert_master(dlm, res, iter.node_map,
6714d8e8
KH		 3430 DLM_ASSERT_MASTER_FINISH_MIGRATION);
3431 if (ret < 0) {
3432 mlog(0, "assert master to original master failed "
3433 "with %d.\n", ret);
3434 /* the only nonzero status here would be because of
3435 * a dead original node. we're done. */
3436 ret = 0;
3437 }
3438
3439 /* all done, set the owner, clear the flag */
3440 spin_lock(&res->spinlock);
3441 dlm_set_lockres_owner(dlm, res, dlm->node_num);
3442 res->state &= ~DLM_LOCK_RES_MIGRATING;
3443 spin_unlock(&res->spinlock);
3444 /* re-dirty it on the new master */
3445 dlm_kick_thread(dlm, res);
3446 wake_up(&res->wq);
3447leave:
3448 return ret;
3449}
3450
3451/*
3452 * LOCKRES AST REFCOUNT
3453 * this is integral to migration
3454 */
3455
3456/* for future intent to call an ast, reserve one ahead of time.
3457 * this should be called only after waiting on the lockres
3458 * with dlm_wait_on_lockres, and while still holding the
3459 * spinlock after the call. */
3460void __dlm_lockres_reserve_ast(struct dlm_lock_resource *res)
3461{
3462 assert_spin_locked(&res->spinlock);
3463 if (res->state & DLM_LOCK_RES_MIGRATING) {
3464 __dlm_print_one_lock_resource(res);
3465 }
3466 BUG_ON(res->state & DLM_LOCK_RES_MIGRATING);
3467
3468 atomic_inc(&res->asts_reserved);
3469}
3470
3471/*
3472 * used to drop the reserved ast, either because it went unused,
3473 * or because the ast/bast was actually called.
3474 *
3475 * also, if there is a pending migration on this lockres,
3476 * and this was the last pending ast on the lockres,
3477 * atomically set the MIGRATING flag before we drop the lock.
3478 * this is how we ensure that migration can proceed with no
3479 * asts in progress. note that it is ok if the state of the
3480 * queues is such that a lock should be granted in the future
3481 * or that a bast should be fired, because the new master will
3482 * shuffle the lists on this lockres as soon as it is migrated.
3483 */
3484void dlm_lockres_release_ast(struct dlm_ctxt *dlm,
3485 struct dlm_lock_resource *res)
3486{
3487 if (!atomic_dec_and_lock(&res->asts_reserved, &res->spinlock))
3488 return;
3489
3490 if (!res->migration_pending) {
3491 spin_unlock(&res->spinlock);
3492 return;
3493 }
3494
3495 BUG_ON(res->state & DLM_LOCK_RES_MIGRATING);
3496 res->migration_pending = 0;
3497 res->state |= DLM_LOCK_RES_MIGRATING;
3498 spin_unlock(&res->spinlock);
3499 wake_up(&res->wq);
3500 wake_up(&dlm->migration_wq);
3501}
Linux 6.x block layer and io_uring tree(s)