ia64/linux-2.6.18-xen.hg

view fs/ocfs2/dlmglue.c @ 524:7f8b544237bf

netfront: Allow netfront in domain 0.

This is useful if your physical network device is in a utility domain.

Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
author Keir Fraser <keir.fraser@citrix.com>
date Tue Apr 15 15:18:58 2008 +0100 (2008-04-15)
parents 831230e53067
children
line source
1 /* -*- mode: c; c-basic-offset: 8; -*-
2 * vim: noexpandtab sw=8 ts=8 sts=0:
3 *
4 * dlmglue.c
5 *
6 * Code which implements an OCFS2 specific interface to our DLM.
7 *
8 * Copyright (C) 2003, 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 */
26 #include <linux/types.h>
27 #include <linux/slab.h>
28 #include <linux/highmem.h>
29 #include <linux/mm.h>
30 #include <linux/smp_lock.h>
31 #include <linux/crc32.h>
32 #include <linux/kthread.h>
33 #include <linux/pagemap.h>
34 #include <linux/debugfs.h>
35 #include <linux/seq_file.h>
37 #include <cluster/heartbeat.h>
38 #include <cluster/nodemanager.h>
39 #include <cluster/tcp.h>
41 #include <dlm/dlmapi.h>
43 #define MLOG_MASK_PREFIX ML_DLM_GLUE
44 #include <cluster/masklog.h>
46 #include "ocfs2.h"
48 #include "alloc.h"
49 #include "dlmglue.h"
50 #include "extent_map.h"
51 #include "heartbeat.h"
52 #include "inode.h"
53 #include "journal.h"
54 #include "slot_map.h"
55 #include "super.h"
56 #include "uptodate.h"
57 #include "vote.h"
59 #include "buffer_head_io.h"
61 struct ocfs2_mask_waiter {
62 struct list_head mw_item;
63 int mw_status;
64 struct completion mw_complete;
65 unsigned long mw_mask;
66 unsigned long mw_goal;
67 };
69 static void ocfs2_inode_ast_func(void *opaque);
70 static void ocfs2_inode_bast_func(void *opaque,
71 int level);
72 static void ocfs2_super_ast_func(void *opaque);
73 static void ocfs2_super_bast_func(void *opaque,
74 int level);
75 static void ocfs2_rename_ast_func(void *opaque);
76 static void ocfs2_rename_bast_func(void *opaque,
77 int level);
79 /* so far, all locks have gotten along with the same unlock ast */
80 static void ocfs2_unlock_ast_func(void *opaque,
81 enum dlm_status status);
82 static int ocfs2_do_unblock_meta(struct inode *inode,
83 int *requeue);
84 static int ocfs2_unblock_meta(struct ocfs2_lock_res *lockres,
85 int *requeue);
86 static int ocfs2_unblock_data(struct ocfs2_lock_res *lockres,
87 int *requeue);
88 static int ocfs2_unblock_inode_lock(struct ocfs2_lock_res *lockres,
89 int *requeue);
90 static int ocfs2_unblock_osb_lock(struct ocfs2_lock_res *lockres,
91 int *requeue);
92 typedef void (ocfs2_convert_worker_t)(struct ocfs2_lock_res *, int);
93 static int ocfs2_generic_unblock_lock(struct ocfs2_super *osb,
94 struct ocfs2_lock_res *lockres,
95 int *requeue,
96 ocfs2_convert_worker_t *worker);
98 struct ocfs2_lock_res_ops {
99 void (*ast)(void *);
100 void (*bast)(void *, int);
101 void (*unlock_ast)(void *, enum dlm_status);
102 int (*unblock)(struct ocfs2_lock_res *, int *);
103 };
105 static struct ocfs2_lock_res_ops ocfs2_inode_rw_lops = {
106 .ast = ocfs2_inode_ast_func,
107 .bast = ocfs2_inode_bast_func,
108 .unlock_ast = ocfs2_unlock_ast_func,
109 .unblock = ocfs2_unblock_inode_lock,
110 };
112 static struct ocfs2_lock_res_ops ocfs2_inode_meta_lops = {
113 .ast = ocfs2_inode_ast_func,
114 .bast = ocfs2_inode_bast_func,
115 .unlock_ast = ocfs2_unlock_ast_func,
116 .unblock = ocfs2_unblock_meta,
117 };
119 static void ocfs2_data_convert_worker(struct ocfs2_lock_res *lockres,
120 int blocking);
122 static struct ocfs2_lock_res_ops ocfs2_inode_data_lops = {
123 .ast = ocfs2_inode_ast_func,
124 .bast = ocfs2_inode_bast_func,
125 .unlock_ast = ocfs2_unlock_ast_func,
126 .unblock = ocfs2_unblock_data,
127 };
129 static struct ocfs2_lock_res_ops ocfs2_super_lops = {
130 .ast = ocfs2_super_ast_func,
131 .bast = ocfs2_super_bast_func,
132 .unlock_ast = ocfs2_unlock_ast_func,
133 .unblock = ocfs2_unblock_osb_lock,
134 };
136 static struct ocfs2_lock_res_ops ocfs2_rename_lops = {
137 .ast = ocfs2_rename_ast_func,
138 .bast = ocfs2_rename_bast_func,
139 .unlock_ast = ocfs2_unlock_ast_func,
140 .unblock = ocfs2_unblock_osb_lock,
141 };
143 static inline int ocfs2_is_inode_lock(struct ocfs2_lock_res *lockres)
144 {
145 return lockres->l_type == OCFS2_LOCK_TYPE_META ||
146 lockres->l_type == OCFS2_LOCK_TYPE_DATA ||
147 lockres->l_type == OCFS2_LOCK_TYPE_RW;
148 }
150 static inline int ocfs2_is_super_lock(struct ocfs2_lock_res *lockres)
151 {
152 return lockres->l_type == OCFS2_LOCK_TYPE_SUPER;
153 }
155 static inline int ocfs2_is_rename_lock(struct ocfs2_lock_res *lockres)
156 {
157 return lockres->l_type == OCFS2_LOCK_TYPE_RENAME;
158 }
160 static inline struct ocfs2_super *ocfs2_lock_res_super(struct ocfs2_lock_res *lockres)
161 {
162 BUG_ON(!ocfs2_is_super_lock(lockres)
163 && !ocfs2_is_rename_lock(lockres));
165 return (struct ocfs2_super *) lockres->l_priv;
166 }
168 static inline struct inode *ocfs2_lock_res_inode(struct ocfs2_lock_res *lockres)
169 {
170 BUG_ON(!ocfs2_is_inode_lock(lockres));
172 return (struct inode *) lockres->l_priv;
173 }
175 static int ocfs2_lock_create(struct ocfs2_super *osb,
176 struct ocfs2_lock_res *lockres,
177 int level,
178 int dlm_flags);
179 static inline int ocfs2_may_continue_on_blocked_lock(struct ocfs2_lock_res *lockres,
180 int wanted);
181 static void ocfs2_cluster_unlock(struct ocfs2_super *osb,
182 struct ocfs2_lock_res *lockres,
183 int level);
184 static inline void ocfs2_generic_handle_downconvert_action(struct ocfs2_lock_res *lockres);
185 static inline void ocfs2_generic_handle_convert_action(struct ocfs2_lock_res *lockres);
186 static inline void ocfs2_generic_handle_attach_action(struct ocfs2_lock_res *lockres);
187 static int ocfs2_generic_handle_bast(struct ocfs2_lock_res *lockres, int level);
188 static void ocfs2_schedule_blocked_lock(struct ocfs2_super *osb,
189 struct ocfs2_lock_res *lockres);
190 static inline void ocfs2_recover_from_dlm_error(struct ocfs2_lock_res *lockres,
191 int convert);
192 #define ocfs2_log_dlm_error(_func, _stat, _lockres) do { \
193 mlog(ML_ERROR, "Dlm error \"%s\" while calling %s on " \
194 "resource %s: %s\n", dlm_errname(_stat), _func, \
195 _lockres->l_name, dlm_errmsg(_stat)); \
196 } while (0)
197 static void ocfs2_vote_on_unlock(struct ocfs2_super *osb,
198 struct ocfs2_lock_res *lockres);
199 static int ocfs2_meta_lock_update(struct inode *inode,
200 struct buffer_head **bh);
201 static void ocfs2_drop_osb_locks(struct ocfs2_super *osb);
202 static inline int ocfs2_highest_compat_lock_level(int level);
203 static inline int ocfs2_can_downconvert_meta_lock(struct inode *inode,
204 struct ocfs2_lock_res *lockres,
205 int new_level);
207 static char *ocfs2_lock_type_strings[] = {
208 [OCFS2_LOCK_TYPE_META] = "Meta",
209 [OCFS2_LOCK_TYPE_DATA] = "Data",
210 [OCFS2_LOCK_TYPE_SUPER] = "Super",
211 [OCFS2_LOCK_TYPE_RENAME] = "Rename",
212 /* Need to differntiate from [R]ename.. serializing writes is the
213 * important job it does, anyway. */
214 [OCFS2_LOCK_TYPE_RW] = "Write/Read",
215 };
217 static char *ocfs2_lock_type_string(enum ocfs2_lock_type type)
218 {
219 mlog_bug_on_msg(type >= OCFS2_NUM_LOCK_TYPES, "%d\n", type);
220 return ocfs2_lock_type_strings[type];
221 }
223 static void ocfs2_build_lock_name(enum ocfs2_lock_type type,
224 u64 blkno,
225 u32 generation,
226 char *name)
227 {
228 int len;
230 mlog_entry_void();
232 BUG_ON(type >= OCFS2_NUM_LOCK_TYPES);
234 len = snprintf(name, OCFS2_LOCK_ID_MAX_LEN, "%c%s%016llx%08x",
235 ocfs2_lock_type_char(type), OCFS2_LOCK_ID_PAD,
236 (long long)blkno, generation);
238 BUG_ON(len != (OCFS2_LOCK_ID_MAX_LEN - 1));
240 mlog(0, "built lock resource with name: %s\n", name);
242 mlog_exit_void();
243 }
245 static DEFINE_SPINLOCK(ocfs2_dlm_tracking_lock);
247 static void ocfs2_add_lockres_tracking(struct ocfs2_lock_res *res,
248 struct ocfs2_dlm_debug *dlm_debug)
249 {
250 mlog(0, "Add tracking for lockres %s\n", res->l_name);
252 spin_lock(&ocfs2_dlm_tracking_lock);
253 list_add(&res->l_debug_list, &dlm_debug->d_lockres_tracking);
254 spin_unlock(&ocfs2_dlm_tracking_lock);
255 }
257 static void ocfs2_remove_lockres_tracking(struct ocfs2_lock_res *res)
258 {
259 spin_lock(&ocfs2_dlm_tracking_lock);
260 if (!list_empty(&res->l_debug_list))
261 list_del_init(&res->l_debug_list);
262 spin_unlock(&ocfs2_dlm_tracking_lock);
263 }
265 static void ocfs2_lock_res_init_common(struct ocfs2_super *osb,
266 struct ocfs2_lock_res *res,
267 enum ocfs2_lock_type type,
268 u64 blkno,
269 u32 generation,
270 struct ocfs2_lock_res_ops *ops,
271 void *priv)
272 {
273 ocfs2_build_lock_name(type, blkno, generation, res->l_name);
275 res->l_type = type;
276 res->l_ops = ops;
277 res->l_priv = priv;
279 res->l_level = LKM_IVMODE;
280 res->l_requested = LKM_IVMODE;
281 res->l_blocking = LKM_IVMODE;
282 res->l_action = OCFS2_AST_INVALID;
283 res->l_unlock_action = OCFS2_UNLOCK_INVALID;
285 res->l_flags = OCFS2_LOCK_INITIALIZED;
287 ocfs2_add_lockres_tracking(res, osb->osb_dlm_debug);
288 }
290 void ocfs2_lock_res_init_once(struct ocfs2_lock_res *res)
291 {
292 /* This also clears out the lock status block */
293 memset(res, 0, sizeof(struct ocfs2_lock_res));
294 spin_lock_init(&res->l_lock);
295 init_waitqueue_head(&res->l_event);
296 INIT_LIST_HEAD(&res->l_blocked_list);
297 INIT_LIST_HEAD(&res->l_mask_waiters);
298 }
300 void ocfs2_inode_lock_res_init(struct ocfs2_lock_res *res,
301 enum ocfs2_lock_type type,
302 struct inode *inode)
303 {
304 struct ocfs2_lock_res_ops *ops;
306 switch(type) {
307 case OCFS2_LOCK_TYPE_RW:
308 ops = &ocfs2_inode_rw_lops;
309 break;
310 case OCFS2_LOCK_TYPE_META:
311 ops = &ocfs2_inode_meta_lops;
312 break;
313 case OCFS2_LOCK_TYPE_DATA:
314 ops = &ocfs2_inode_data_lops;
315 break;
316 default:
317 mlog_bug_on_msg(1, "type: %d\n", type);
318 ops = NULL; /* thanks, gcc */
319 break;
320 };
322 ocfs2_lock_res_init_common(OCFS2_SB(inode->i_sb), res, type,
323 OCFS2_I(inode)->ip_blkno,
324 inode->i_generation, ops, inode);
325 }
327 static void ocfs2_super_lock_res_init(struct ocfs2_lock_res *res,
328 struct ocfs2_super *osb)
329 {
330 /* Superblock lockres doesn't come from a slab so we call init
331 * once on it manually. */
332 ocfs2_lock_res_init_once(res);
333 ocfs2_lock_res_init_common(osb, res, OCFS2_LOCK_TYPE_SUPER,
334 OCFS2_SUPER_BLOCK_BLKNO, 0,
335 &ocfs2_super_lops, osb);
336 }
338 static void ocfs2_rename_lock_res_init(struct ocfs2_lock_res *res,
339 struct ocfs2_super *osb)
340 {
341 /* Rename lockres doesn't come from a slab so we call init
342 * once on it manually. */
343 ocfs2_lock_res_init_once(res);
344 ocfs2_lock_res_init_common(osb, res, OCFS2_LOCK_TYPE_RENAME, 0, 0,
345 &ocfs2_rename_lops, osb);
346 }
348 void ocfs2_lock_res_free(struct ocfs2_lock_res *res)
349 {
350 mlog_entry_void();
352 if (!(res->l_flags & OCFS2_LOCK_INITIALIZED))
353 return;
355 ocfs2_remove_lockres_tracking(res);
357 mlog_bug_on_msg(!list_empty(&res->l_blocked_list),
358 "Lockres %s is on the blocked list\n",
359 res->l_name);
360 mlog_bug_on_msg(!list_empty(&res->l_mask_waiters),
361 "Lockres %s has mask waiters pending\n",
362 res->l_name);
363 mlog_bug_on_msg(spin_is_locked(&res->l_lock),
364 "Lockres %s is locked\n",
365 res->l_name);
366 mlog_bug_on_msg(res->l_ro_holders,
367 "Lockres %s has %u ro holders\n",
368 res->l_name, res->l_ro_holders);
369 mlog_bug_on_msg(res->l_ex_holders,
370 "Lockres %s has %u ex holders\n",
371 res->l_name, res->l_ex_holders);
373 /* Need to clear out the lock status block for the dlm */
374 memset(&res->l_lksb, 0, sizeof(res->l_lksb));
376 res->l_flags = 0UL;
377 mlog_exit_void();
378 }
380 static inline void ocfs2_inc_holders(struct ocfs2_lock_res *lockres,
381 int level)
382 {
383 mlog_entry_void();
385 BUG_ON(!lockres);
387 switch(level) {
388 case LKM_EXMODE:
389 lockres->l_ex_holders++;
390 break;
391 case LKM_PRMODE:
392 lockres->l_ro_holders++;
393 break;
394 default:
395 BUG();
396 }
398 mlog_exit_void();
399 }
401 static inline void ocfs2_dec_holders(struct ocfs2_lock_res *lockres,
402 int level)
403 {
404 mlog_entry_void();
406 BUG_ON(!lockres);
408 switch(level) {
409 case LKM_EXMODE:
410 BUG_ON(!lockres->l_ex_holders);
411 lockres->l_ex_holders--;
412 break;
413 case LKM_PRMODE:
414 BUG_ON(!lockres->l_ro_holders);
415 lockres->l_ro_holders--;
416 break;
417 default:
418 BUG();
419 }
420 mlog_exit_void();
421 }
423 /* WARNING: This function lives in a world where the only three lock
424 * levels are EX, PR, and NL. It *will* have to be adjusted when more
425 * lock types are added. */
426 static inline int ocfs2_highest_compat_lock_level(int level)
427 {
428 int new_level = LKM_EXMODE;
430 if (level == LKM_EXMODE)
431 new_level = LKM_NLMODE;
432 else if (level == LKM_PRMODE)
433 new_level = LKM_PRMODE;
434 return new_level;
435 }
437 static void lockres_set_flags(struct ocfs2_lock_res *lockres,
438 unsigned long newflags)
439 {
440 struct list_head *pos, *tmp;
441 struct ocfs2_mask_waiter *mw;
443 assert_spin_locked(&lockres->l_lock);
445 lockres->l_flags = newflags;
447 list_for_each_safe(pos, tmp, &lockres->l_mask_waiters) {
448 mw = list_entry(pos, struct ocfs2_mask_waiter, mw_item);
449 if ((lockres->l_flags & mw->mw_mask) != mw->mw_goal)
450 continue;
452 list_del_init(&mw->mw_item);
453 mw->mw_status = 0;
454 complete(&mw->mw_complete);
455 }
456 }
457 static void lockres_or_flags(struct ocfs2_lock_res *lockres, unsigned long or)
458 {
459 lockres_set_flags(lockres, lockres->l_flags | or);
460 }
461 static void lockres_clear_flags(struct ocfs2_lock_res *lockres,
462 unsigned long clear)
463 {
464 lockres_set_flags(lockres, lockres->l_flags & ~clear);
465 }
467 static inline void ocfs2_generic_handle_downconvert_action(struct ocfs2_lock_res *lockres)
468 {
469 mlog_entry_void();
471 BUG_ON(!(lockres->l_flags & OCFS2_LOCK_BUSY));
472 BUG_ON(!(lockres->l_flags & OCFS2_LOCK_ATTACHED));
473 BUG_ON(!(lockres->l_flags & OCFS2_LOCK_BLOCKED));
474 BUG_ON(lockres->l_blocking <= LKM_NLMODE);
476 lockres->l_level = lockres->l_requested;
477 if (lockres->l_level <=
478 ocfs2_highest_compat_lock_level(lockres->l_blocking)) {
479 lockres->l_blocking = LKM_NLMODE;
480 lockres_clear_flags(lockres, OCFS2_LOCK_BLOCKED);
481 }
482 lockres_clear_flags(lockres, OCFS2_LOCK_BUSY);
484 mlog_exit_void();
485 }
487 static inline void ocfs2_generic_handle_convert_action(struct ocfs2_lock_res *lockres)
488 {
489 mlog_entry_void();
491 BUG_ON(!(lockres->l_flags & OCFS2_LOCK_BUSY));
492 BUG_ON(!(lockres->l_flags & OCFS2_LOCK_ATTACHED));
494 /* Convert from RO to EX doesn't really need anything as our
495 * information is already up to data. Convert from NL to
496 * *anything* however should mark ourselves as needing an
497 * update */
498 if (lockres->l_level == LKM_NLMODE)
499 lockres_or_flags(lockres, OCFS2_LOCK_NEEDS_REFRESH);
501 lockres->l_level = lockres->l_requested;
502 lockres_clear_flags(lockres, OCFS2_LOCK_BUSY);
504 mlog_exit_void();
505 }
507 static inline void ocfs2_generic_handle_attach_action(struct ocfs2_lock_res *lockres)
508 {
509 mlog_entry_void();
511 BUG_ON((!lockres->l_flags & OCFS2_LOCK_BUSY));
512 BUG_ON(lockres->l_flags & OCFS2_LOCK_ATTACHED);
514 if (lockres->l_requested > LKM_NLMODE &&
515 !(lockres->l_flags & OCFS2_LOCK_LOCAL))
516 lockres_or_flags(lockres, OCFS2_LOCK_NEEDS_REFRESH);
518 lockres->l_level = lockres->l_requested;
519 lockres_or_flags(lockres, OCFS2_LOCK_ATTACHED);
520 lockres_clear_flags(lockres, OCFS2_LOCK_BUSY);
522 mlog_exit_void();
523 }
525 static void ocfs2_inode_ast_func(void *opaque)
526 {
527 struct ocfs2_lock_res *lockres = opaque;
528 struct inode *inode;
529 struct dlm_lockstatus *lksb;
530 unsigned long flags;
532 mlog_entry_void();
534 inode = ocfs2_lock_res_inode(lockres);
536 mlog(0, "AST fired for inode %llu, l_action = %u, type = %s\n",
537 (unsigned long long)OCFS2_I(inode)->ip_blkno, lockres->l_action,
538 ocfs2_lock_type_string(lockres->l_type));
540 BUG_ON(!ocfs2_is_inode_lock(lockres));
542 spin_lock_irqsave(&lockres->l_lock, flags);
544 lksb = &(lockres->l_lksb);
545 if (lksb->status != DLM_NORMAL) {
546 mlog(ML_ERROR, "ocfs2_inode_ast_func: lksb status value of %u "
547 "on inode %llu\n", lksb->status,
548 (unsigned long long)OCFS2_I(inode)->ip_blkno);
549 spin_unlock_irqrestore(&lockres->l_lock, flags);
550 mlog_exit_void();
551 return;
552 }
554 switch(lockres->l_action) {
555 case OCFS2_AST_ATTACH:
556 ocfs2_generic_handle_attach_action(lockres);
557 lockres_clear_flags(lockres, OCFS2_LOCK_LOCAL);
558 break;
559 case OCFS2_AST_CONVERT:
560 ocfs2_generic_handle_convert_action(lockres);
561 break;
562 case OCFS2_AST_DOWNCONVERT:
563 ocfs2_generic_handle_downconvert_action(lockres);
564 break;
565 default:
566 mlog(ML_ERROR, "lockres %s: ast fired with invalid action: %u "
567 "lockres flags = 0x%lx, unlock action: %u\n",
568 lockres->l_name, lockres->l_action, lockres->l_flags,
569 lockres->l_unlock_action);
571 BUG();
572 }
574 /* data and rw locking ignores refresh flag for now. */
575 if (lockres->l_type != OCFS2_LOCK_TYPE_META)
576 lockres_clear_flags(lockres, OCFS2_LOCK_NEEDS_REFRESH);
578 /* set it to something invalid so if we get called again we
579 * can catch it. */
580 lockres->l_action = OCFS2_AST_INVALID;
581 spin_unlock_irqrestore(&lockres->l_lock, flags);
582 wake_up(&lockres->l_event);
584 mlog_exit_void();
585 }
587 static int ocfs2_generic_handle_bast(struct ocfs2_lock_res *lockres,
588 int level)
589 {
590 int needs_downconvert = 0;
591 mlog_entry_void();
593 assert_spin_locked(&lockres->l_lock);
595 lockres_or_flags(lockres, OCFS2_LOCK_BLOCKED);
597 if (level > lockres->l_blocking) {
598 /* only schedule a downconvert if we haven't already scheduled
599 * one that goes low enough to satisfy the level we're
600 * blocking. this also catches the case where we get
601 * duplicate BASTs */
602 if (ocfs2_highest_compat_lock_level(level) <
603 ocfs2_highest_compat_lock_level(lockres->l_blocking))
604 needs_downconvert = 1;
606 lockres->l_blocking = level;
607 }
609 mlog_exit(needs_downconvert);
610 return needs_downconvert;
611 }
613 static void ocfs2_generic_bast_func(struct ocfs2_super *osb,
614 struct ocfs2_lock_res *lockres,
615 int level)
616 {
617 int needs_downconvert;
618 unsigned long flags;
620 mlog_entry_void();
622 BUG_ON(level <= LKM_NLMODE);
624 spin_lock_irqsave(&lockres->l_lock, flags);
625 needs_downconvert = ocfs2_generic_handle_bast(lockres, level);
626 if (needs_downconvert)
627 ocfs2_schedule_blocked_lock(osb, lockres);
628 spin_unlock_irqrestore(&lockres->l_lock, flags);
630 ocfs2_kick_vote_thread(osb);
632 wake_up(&lockres->l_event);
633 mlog_exit_void();
634 }
636 static void ocfs2_inode_bast_func(void *opaque, int level)
637 {
638 struct ocfs2_lock_res *lockres = opaque;
639 struct inode *inode;
640 struct ocfs2_super *osb;
642 mlog_entry_void();
644 BUG_ON(!ocfs2_is_inode_lock(lockres));
646 inode = ocfs2_lock_res_inode(lockres);
647 osb = OCFS2_SB(inode->i_sb);
649 mlog(0, "BAST fired for inode %llu, blocking %d, level %d type %s\n",
650 (unsigned long long)OCFS2_I(inode)->ip_blkno, level,
651 lockres->l_level, ocfs2_lock_type_string(lockres->l_type));
653 ocfs2_generic_bast_func(osb, lockres, level);
655 mlog_exit_void();
656 }
658 static void ocfs2_generic_ast_func(struct ocfs2_lock_res *lockres,
659 int ignore_refresh)
660 {
661 struct dlm_lockstatus *lksb = &lockres->l_lksb;
662 unsigned long flags;
664 spin_lock_irqsave(&lockres->l_lock, flags);
666 if (lksb->status != DLM_NORMAL) {
667 mlog(ML_ERROR, "lockres %s: lksb status value of %u!\n",
668 lockres->l_name, lksb->status);
669 spin_unlock_irqrestore(&lockres->l_lock, flags);
670 return;
671 }
673 switch(lockres->l_action) {
674 case OCFS2_AST_ATTACH:
675 ocfs2_generic_handle_attach_action(lockres);
676 break;
677 case OCFS2_AST_CONVERT:
678 ocfs2_generic_handle_convert_action(lockres);
679 break;
680 case OCFS2_AST_DOWNCONVERT:
681 ocfs2_generic_handle_downconvert_action(lockres);
682 break;
683 default:
684 BUG();
685 }
687 if (ignore_refresh)
688 lockres_clear_flags(lockres, OCFS2_LOCK_NEEDS_REFRESH);
690 /* set it to something invalid so if we get called again we
691 * can catch it. */
692 lockres->l_action = OCFS2_AST_INVALID;
693 spin_unlock_irqrestore(&lockres->l_lock, flags);
695 wake_up(&lockres->l_event);
696 }
698 static void ocfs2_super_ast_func(void *opaque)
699 {
700 struct ocfs2_lock_res *lockres = opaque;
702 mlog_entry_void();
703 mlog(0, "Superblock AST fired\n");
705 BUG_ON(!ocfs2_is_super_lock(lockres));
706 ocfs2_generic_ast_func(lockres, 0);
708 mlog_exit_void();
709 }
711 static void ocfs2_super_bast_func(void *opaque,
712 int level)
713 {
714 struct ocfs2_lock_res *lockres = opaque;
715 struct ocfs2_super *osb;
717 mlog_entry_void();
718 mlog(0, "Superblock BAST fired\n");
720 BUG_ON(!ocfs2_is_super_lock(lockres));
721 osb = ocfs2_lock_res_super(lockres);
722 ocfs2_generic_bast_func(osb, lockres, level);
724 mlog_exit_void();
725 }
727 static void ocfs2_rename_ast_func(void *opaque)
728 {
729 struct ocfs2_lock_res *lockres = opaque;
731 mlog_entry_void();
733 mlog(0, "Rename AST fired\n");
735 BUG_ON(!ocfs2_is_rename_lock(lockres));
737 ocfs2_generic_ast_func(lockres, 1);
739 mlog_exit_void();
740 }
742 static void ocfs2_rename_bast_func(void *opaque,
743 int level)
744 {
745 struct ocfs2_lock_res *lockres = opaque;
746 struct ocfs2_super *osb;
748 mlog_entry_void();
750 mlog(0, "Rename BAST fired\n");
752 BUG_ON(!ocfs2_is_rename_lock(lockres));
754 osb = ocfs2_lock_res_super(lockres);
755 ocfs2_generic_bast_func(osb, lockres, level);
757 mlog_exit_void();
758 }
760 static inline void ocfs2_recover_from_dlm_error(struct ocfs2_lock_res *lockres,
761 int convert)
762 {
763 unsigned long flags;
765 mlog_entry_void();
766 spin_lock_irqsave(&lockres->l_lock, flags);
767 lockres_clear_flags(lockres, OCFS2_LOCK_BUSY);
768 if (convert)
769 lockres->l_action = OCFS2_AST_INVALID;
770 else
771 lockres->l_unlock_action = OCFS2_UNLOCK_INVALID;
772 spin_unlock_irqrestore(&lockres->l_lock, flags);
774 wake_up(&lockres->l_event);
775 mlog_exit_void();
776 }
778 /* Note: If we detect another process working on the lock (i.e.,
779 * OCFS2_LOCK_BUSY), we'll bail out returning 0. It's up to the caller
780 * to do the right thing in that case.
781 */
782 static int ocfs2_lock_create(struct ocfs2_super *osb,
783 struct ocfs2_lock_res *lockres,
784 int level,
785 int dlm_flags)
786 {
787 int ret = 0;
788 enum dlm_status status;
789 unsigned long flags;
791 mlog_entry_void();
793 mlog(0, "lock %s, level = %d, flags = %d\n", lockres->l_name, level,
794 dlm_flags);
796 spin_lock_irqsave(&lockres->l_lock, flags);
797 if ((lockres->l_flags & OCFS2_LOCK_ATTACHED) ||
798 (lockres->l_flags & OCFS2_LOCK_BUSY)) {
799 spin_unlock_irqrestore(&lockres->l_lock, flags);
800 goto bail;
801 }
803 lockres->l_action = OCFS2_AST_ATTACH;
804 lockres->l_requested = level;
805 lockres_or_flags(lockres, OCFS2_LOCK_BUSY);
806 spin_unlock_irqrestore(&lockres->l_lock, flags);
808 status = dlmlock(osb->dlm,
809 level,
810 &lockres->l_lksb,
811 dlm_flags,
812 lockres->l_name,
813 lockres->l_ops->ast,
814 lockres,
815 lockres->l_ops->bast);
816 if (status != DLM_NORMAL) {
817 ocfs2_log_dlm_error("dlmlock", status, lockres);
818 ret = -EINVAL;
819 ocfs2_recover_from_dlm_error(lockres, 1);
820 }
822 mlog(0, "lock %s, successfull return from dlmlock\n", lockres->l_name);
824 bail:
825 mlog_exit(ret);
826 return ret;
827 }
829 static inline int ocfs2_check_wait_flag(struct ocfs2_lock_res *lockres,
830 int flag)
831 {
832 unsigned long flags;
833 int ret;
835 spin_lock_irqsave(&lockres->l_lock, flags);
836 ret = lockres->l_flags & flag;
837 spin_unlock_irqrestore(&lockres->l_lock, flags);
839 return ret;
840 }
842 static inline void ocfs2_wait_on_busy_lock(struct ocfs2_lock_res *lockres)
844 {
845 wait_event(lockres->l_event,
846 !ocfs2_check_wait_flag(lockres, OCFS2_LOCK_BUSY));
847 }
849 static inline void ocfs2_wait_on_refreshing_lock(struct ocfs2_lock_res *lockres)
851 {
852 wait_event(lockres->l_event,
853 !ocfs2_check_wait_flag(lockres, OCFS2_LOCK_REFRESHING));
854 }
856 /* predict what lock level we'll be dropping down to on behalf
857 * of another node, and return true if the currently wanted
858 * level will be compatible with it. */
859 static inline int ocfs2_may_continue_on_blocked_lock(struct ocfs2_lock_res *lockres,
860 int wanted)
861 {
862 BUG_ON(!(lockres->l_flags & OCFS2_LOCK_BLOCKED));
864 return wanted <= ocfs2_highest_compat_lock_level(lockres->l_blocking);
865 }
867 static void ocfs2_init_mask_waiter(struct ocfs2_mask_waiter *mw)
868 {
869 INIT_LIST_HEAD(&mw->mw_item);
870 init_completion(&mw->mw_complete);
871 }
873 static int ocfs2_wait_for_mask(struct ocfs2_mask_waiter *mw)
874 {
875 wait_for_completion(&mw->mw_complete);
876 /* Re-arm the completion in case we want to wait on it again */
877 INIT_COMPLETION(mw->mw_complete);
878 return mw->mw_status;
879 }
881 static void lockres_add_mask_waiter(struct ocfs2_lock_res *lockres,
882 struct ocfs2_mask_waiter *mw,
883 unsigned long mask,
884 unsigned long goal)
885 {
886 BUG_ON(!list_empty(&mw->mw_item));
888 assert_spin_locked(&lockres->l_lock);
890 list_add_tail(&mw->mw_item, &lockres->l_mask_waiters);
891 mw->mw_mask = mask;
892 mw->mw_goal = goal;
893 }
895 /* returns 0 if the mw that was removed was already satisfied, -EBUSY
896 * if the mask still hadn't reached its goal */
897 static int lockres_remove_mask_waiter(struct ocfs2_lock_res *lockres,
898 struct ocfs2_mask_waiter *mw)
899 {
900 unsigned long flags;
901 int ret = 0;
903 spin_lock_irqsave(&lockres->l_lock, flags);
904 if (!list_empty(&mw->mw_item)) {
905 if ((lockres->l_flags & mw->mw_mask) != mw->mw_goal)
906 ret = -EBUSY;
908 list_del_init(&mw->mw_item);
909 init_completion(&mw->mw_complete);
910 }
911 spin_unlock_irqrestore(&lockres->l_lock, flags);
913 return ret;
915 }
917 static int ocfs2_cluster_lock(struct ocfs2_super *osb,
918 struct ocfs2_lock_res *lockres,
919 int level,
920 int lkm_flags,
921 int arg_flags)
922 {
923 struct ocfs2_mask_waiter mw;
924 enum dlm_status status;
925 int wait, catch_signals = !(osb->s_mount_opt & OCFS2_MOUNT_NOINTR);
926 int ret = 0; /* gcc doesn't realize wait = 1 guarantees ret is set */
927 unsigned long flags;
929 mlog_entry_void();
931 ocfs2_init_mask_waiter(&mw);
933 again:
934 wait = 0;
936 if (catch_signals && signal_pending(current)) {
937 ret = -ERESTARTSYS;
938 goto out;
939 }
941 spin_lock_irqsave(&lockres->l_lock, flags);
943 mlog_bug_on_msg(lockres->l_flags & OCFS2_LOCK_FREEING,
944 "Cluster lock called on freeing lockres %s! flags "
945 "0x%lx\n", lockres->l_name, lockres->l_flags);
947 /* We only compare against the currently granted level
948 * here. If the lock is blocked waiting on a downconvert,
949 * we'll get caught below. */
950 if (lockres->l_flags & OCFS2_LOCK_BUSY &&
951 level > lockres->l_level) {
952 /* is someone sitting in dlm_lock? If so, wait on
953 * them. */
954 lockres_add_mask_waiter(lockres, &mw, OCFS2_LOCK_BUSY, 0);
955 wait = 1;
956 goto unlock;
957 }
959 if (!(lockres->l_flags & OCFS2_LOCK_ATTACHED)) {
960 /* lock has not been created yet. */
961 spin_unlock_irqrestore(&lockres->l_lock, flags);
963 ret = ocfs2_lock_create(osb, lockres, LKM_NLMODE, 0);
964 if (ret < 0) {
965 mlog_errno(ret);
966 goto out;
967 }
968 goto again;
969 }
971 if (lockres->l_flags & OCFS2_LOCK_BLOCKED &&
972 !ocfs2_may_continue_on_blocked_lock(lockres, level)) {
973 /* is the lock is currently blocked on behalf of
974 * another node */
975 lockres_add_mask_waiter(lockres, &mw, OCFS2_LOCK_BLOCKED, 0);
976 wait = 1;
977 goto unlock;
978 }
980 if (level > lockres->l_level) {
981 if (lockres->l_action != OCFS2_AST_INVALID)
982 mlog(ML_ERROR, "lockres %s has action %u pending\n",
983 lockres->l_name, lockres->l_action);
985 lockres->l_action = OCFS2_AST_CONVERT;
986 lockres->l_requested = level;
987 lockres_or_flags(lockres, OCFS2_LOCK_BUSY);
988 spin_unlock_irqrestore(&lockres->l_lock, flags);
990 BUG_ON(level == LKM_IVMODE);
991 BUG_ON(level == LKM_NLMODE);
993 mlog(0, "lock %s, convert from %d to level = %d\n",
994 lockres->l_name, lockres->l_level, level);
996 /* call dlm_lock to upgrade lock now */
997 status = dlmlock(osb->dlm,
998 level,
999 &lockres->l_lksb,
1000 lkm_flags|LKM_CONVERT|LKM_VALBLK,
1001 lockres->l_name,
1002 lockres->l_ops->ast,
1003 lockres,
1004 lockres->l_ops->bast);
1005 if (status != DLM_NORMAL) {
1006 if ((lkm_flags & LKM_NOQUEUE) &&
1007 (status == DLM_NOTQUEUED))
1008 ret = -EAGAIN;
1009 else {
1010 ocfs2_log_dlm_error("dlmlock", status,
1011 lockres);
1012 ret = -EINVAL;
1014 ocfs2_recover_from_dlm_error(lockres, 1);
1015 goto out;
1018 mlog(0, "lock %s, successfull return from dlmlock\n",
1019 lockres->l_name);
1021 /* At this point we've gone inside the dlm and need to
1022 * complete our work regardless. */
1023 catch_signals = 0;
1025 /* wait for busy to clear and carry on */
1026 goto again;
1029 /* Ok, if we get here then we're good to go. */
1030 ocfs2_inc_holders(lockres, level);
1032 ret = 0;
1033 unlock:
1034 spin_unlock_irqrestore(&lockres->l_lock, flags);
1035 out:
1036 /*
1037 * This is helping work around a lock inversion between the page lock
1038 * and dlm locks. One path holds the page lock while calling aops
1039 * which block acquiring dlm locks. The voting thread holds dlm
1040 * locks while acquiring page locks while down converting data locks.
1041 * This block is helping an aop path notice the inversion and back
1042 * off to unlock its page lock before trying the dlm lock again.
1043 */
1044 if (wait && arg_flags & OCFS2_LOCK_NONBLOCK &&
1045 mw.mw_mask & (OCFS2_LOCK_BUSY|OCFS2_LOCK_BLOCKED)) {
1046 wait = 0;
1047 if (lockres_remove_mask_waiter(lockres, &mw))
1048 ret = -EAGAIN;
1049 else
1050 goto again;
1052 if (wait) {
1053 ret = ocfs2_wait_for_mask(&mw);
1054 if (ret == 0)
1055 goto again;
1056 mlog_errno(ret);
1059 mlog_exit(ret);
1060 return ret;
1063 static void ocfs2_cluster_unlock(struct ocfs2_super *osb,
1064 struct ocfs2_lock_res *lockres,
1065 int level)
1067 unsigned long flags;
1069 mlog_entry_void();
1070 spin_lock_irqsave(&lockres->l_lock, flags);
1071 ocfs2_dec_holders(lockres, level);
1072 ocfs2_vote_on_unlock(osb, lockres);
1073 spin_unlock_irqrestore(&lockres->l_lock, flags);
1074 mlog_exit_void();
1077 static int ocfs2_create_new_inode_lock(struct inode *inode,
1078 struct ocfs2_lock_res *lockres)
1080 struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
1081 unsigned long flags;
1083 spin_lock_irqsave(&lockres->l_lock, flags);
1084 BUG_ON(lockres->l_flags & OCFS2_LOCK_ATTACHED);
1085 lockres_or_flags(lockres, OCFS2_LOCK_LOCAL);
1086 spin_unlock_irqrestore(&lockres->l_lock, flags);
1088 return ocfs2_lock_create(osb, lockres, LKM_EXMODE, LKM_LOCAL);
1091 /* Grants us an EX lock on the data and metadata resources, skipping
1092 * the normal cluster directory lookup. Use this ONLY on newly created
1093 * inodes which other nodes can't possibly see, and which haven't been
1094 * hashed in the inode hash yet. This can give us a good performance
1095 * increase as it'll skip the network broadcast normally associated
1096 * with creating a new lock resource. */
1097 int ocfs2_create_new_inode_locks(struct inode *inode)
1099 int ret;
1101 BUG_ON(!inode);
1102 BUG_ON(!ocfs2_inode_is_new(inode));
1104 mlog_entry_void();
1106 mlog(0, "Inode %llu\n", (unsigned long long)OCFS2_I(inode)->ip_blkno);
1108 /* NOTE: That we don't increment any of the holder counts, nor
1109 * do we add anything to a journal handle. Since this is
1110 * supposed to be a new inode which the cluster doesn't know
1111 * about yet, there is no need to. As far as the LVB handling
1112 * is concerned, this is basically like acquiring an EX lock
1113 * on a resource which has an invalid one -- we'll set it
1114 * valid when we release the EX. */
1116 ret = ocfs2_create_new_inode_lock(inode,
1117 &OCFS2_I(inode)->ip_rw_lockres);
1118 if (ret) {
1119 mlog_errno(ret);
1120 goto bail;
1123 ret = ocfs2_create_new_inode_lock(inode,
1124 &OCFS2_I(inode)->ip_meta_lockres);
1125 if (ret) {
1126 mlog_errno(ret);
1127 goto bail;
1130 ret = ocfs2_create_new_inode_lock(inode,
1131 &OCFS2_I(inode)->ip_data_lockres);
1132 if (ret) {
1133 mlog_errno(ret);
1134 goto bail;
1137 bail:
1138 mlog_exit(ret);
1139 return ret;
1142 int ocfs2_rw_lock(struct inode *inode, int write)
1144 int status, level;
1145 struct ocfs2_lock_res *lockres;
1147 BUG_ON(!inode);
1149 mlog_entry_void();
1151 mlog(0, "inode %llu take %s RW lock\n",
1152 (unsigned long long)OCFS2_I(inode)->ip_blkno,
1153 write ? "EXMODE" : "PRMODE");
1155 lockres = &OCFS2_I(inode)->ip_rw_lockres;
1157 level = write ? LKM_EXMODE : LKM_PRMODE;
1159 status = ocfs2_cluster_lock(OCFS2_SB(inode->i_sb), lockres, level, 0,
1160 0);
1161 if (status < 0)
1162 mlog_errno(status);
1164 mlog_exit(status);
1165 return status;
1168 void ocfs2_rw_unlock(struct inode *inode, int write)
1170 int level = write ? LKM_EXMODE : LKM_PRMODE;
1171 struct ocfs2_lock_res *lockres = &OCFS2_I(inode)->ip_rw_lockres;
1173 mlog_entry_void();
1175 mlog(0, "inode %llu drop %s RW lock\n",
1176 (unsigned long long)OCFS2_I(inode)->ip_blkno,
1177 write ? "EXMODE" : "PRMODE");
1179 ocfs2_cluster_unlock(OCFS2_SB(inode->i_sb), lockres, level);
1181 mlog_exit_void();
1184 int ocfs2_data_lock_full(struct inode *inode,
1185 int write,
1186 int arg_flags)
1188 int status = 0, level;
1189 struct ocfs2_lock_res *lockres;
1191 BUG_ON(!inode);
1193 mlog_entry_void();
1195 mlog(0, "inode %llu take %s DATA lock\n",
1196 (unsigned long long)OCFS2_I(inode)->ip_blkno,
1197 write ? "EXMODE" : "PRMODE");
1199 /* We'll allow faking a readonly data lock for
1200 * rodevices. */
1201 if (ocfs2_is_hard_readonly(OCFS2_SB(inode->i_sb))) {
1202 if (write) {
1203 status = -EROFS;
1204 mlog_errno(status);
1206 goto out;
1209 lockres = &OCFS2_I(inode)->ip_data_lockres;
1211 level = write ? LKM_EXMODE : LKM_PRMODE;
1213 status = ocfs2_cluster_lock(OCFS2_SB(inode->i_sb), lockres, level,
1214 0, arg_flags);
1215 if (status < 0 && status != -EAGAIN)
1216 mlog_errno(status);
1218 out:
1219 mlog_exit(status);
1220 return status;
1223 /* see ocfs2_meta_lock_with_page() */
1224 int ocfs2_data_lock_with_page(struct inode *inode,
1225 int write,
1226 struct page *page)
1228 int ret;
1230 ret = ocfs2_data_lock_full(inode, write, OCFS2_LOCK_NONBLOCK);
1231 if (ret == -EAGAIN) {
1232 unlock_page(page);
1233 if (ocfs2_data_lock(inode, write) == 0)
1234 ocfs2_data_unlock(inode, write);
1235 ret = AOP_TRUNCATED_PAGE;
1238 return ret;
1241 static void ocfs2_vote_on_unlock(struct ocfs2_super *osb,
1242 struct ocfs2_lock_res *lockres)
1244 int kick = 0;
1246 mlog_entry_void();
1248 /* If we know that another node is waiting on our lock, kick
1249 * the vote thread * pre-emptively when we reach a release
1250 * condition. */
1251 if (lockres->l_flags & OCFS2_LOCK_BLOCKED) {
1252 switch(lockres->l_blocking) {
1253 case LKM_EXMODE:
1254 if (!lockres->l_ex_holders && !lockres->l_ro_holders)
1255 kick = 1;
1256 break;
1257 case LKM_PRMODE:
1258 if (!lockres->l_ex_holders)
1259 kick = 1;
1260 break;
1261 default:
1262 BUG();
1266 if (kick)
1267 ocfs2_kick_vote_thread(osb);
1269 mlog_exit_void();
1272 void ocfs2_data_unlock(struct inode *inode,
1273 int write)
1275 int level = write ? LKM_EXMODE : LKM_PRMODE;
1276 struct ocfs2_lock_res *lockres = &OCFS2_I(inode)->ip_data_lockres;
1278 mlog_entry_void();
1280 mlog(0, "inode %llu drop %s DATA lock\n",
1281 (unsigned long long)OCFS2_I(inode)->ip_blkno,
1282 write ? "EXMODE" : "PRMODE");
1284 if (!ocfs2_is_hard_readonly(OCFS2_SB(inode->i_sb)))
1285 ocfs2_cluster_unlock(OCFS2_SB(inode->i_sb), lockres, level);
1287 mlog_exit_void();
1290 #define OCFS2_SEC_BITS 34
1291 #define OCFS2_SEC_SHIFT (64 - 34)
1292 #define OCFS2_NSEC_MASK ((1ULL << OCFS2_SEC_SHIFT) - 1)
1294 /* LVB only has room for 64 bits of time here so we pack it for
1295 * now. */
1296 static u64 ocfs2_pack_timespec(struct timespec *spec)
1298 u64 res;
1299 u64 sec = spec->tv_sec;
1300 u32 nsec = spec->tv_nsec;
1302 res = (sec << OCFS2_SEC_SHIFT) | (nsec & OCFS2_NSEC_MASK);
1304 return res;
1307 /* Call this with the lockres locked. I am reasonably sure we don't
1308 * need ip_lock in this function as anyone who would be changing those
1309 * values is supposed to be blocked in ocfs2_meta_lock right now. */
1310 static void __ocfs2_stuff_meta_lvb(struct inode *inode)
1312 struct ocfs2_inode_info *oi = OCFS2_I(inode);
1313 struct ocfs2_lock_res *lockres = &oi->ip_meta_lockres;
1314 struct ocfs2_meta_lvb *lvb;
1316 mlog_entry_void();
1318 lvb = (struct ocfs2_meta_lvb *) lockres->l_lksb.lvb;
1320 lvb->lvb_version = cpu_to_be32(OCFS2_LVB_VERSION);
1321 lvb->lvb_isize = cpu_to_be64(i_size_read(inode));
1322 lvb->lvb_iclusters = cpu_to_be32(oi->ip_clusters);
1323 lvb->lvb_iuid = cpu_to_be32(inode->i_uid);
1324 lvb->lvb_igid = cpu_to_be32(inode->i_gid);
1325 lvb->lvb_imode = cpu_to_be16(inode->i_mode);
1326 lvb->lvb_inlink = cpu_to_be16(inode->i_nlink);
1327 lvb->lvb_iatime_packed =
1328 cpu_to_be64(ocfs2_pack_timespec(&inode->i_atime));
1329 lvb->lvb_ictime_packed =
1330 cpu_to_be64(ocfs2_pack_timespec(&inode->i_ctime));
1331 lvb->lvb_imtime_packed =
1332 cpu_to_be64(ocfs2_pack_timespec(&inode->i_mtime));
1334 mlog_meta_lvb(0, lockres);
1336 mlog_exit_void();
1339 static void ocfs2_unpack_timespec(struct timespec *spec,
1340 u64 packed_time)
1342 spec->tv_sec = packed_time >> OCFS2_SEC_SHIFT;
1343 spec->tv_nsec = packed_time & OCFS2_NSEC_MASK;
1346 static void ocfs2_refresh_inode_from_lvb(struct inode *inode)
1348 struct ocfs2_inode_info *oi = OCFS2_I(inode);
1349 struct ocfs2_lock_res *lockres = &oi->ip_meta_lockres;
1350 struct ocfs2_meta_lvb *lvb;
1352 mlog_entry_void();
1354 mlog_meta_lvb(0, lockres);
1356 lvb = (struct ocfs2_meta_lvb *) lockres->l_lksb.lvb;
1358 /* We're safe here without the lockres lock... */
1359 spin_lock(&oi->ip_lock);
1360 oi->ip_clusters = be32_to_cpu(lvb->lvb_iclusters);
1361 i_size_write(inode, be64_to_cpu(lvb->lvb_isize));
1363 /* fast-symlinks are a special case */
1364 if (S_ISLNK(inode->i_mode) && !oi->ip_clusters)
1365 inode->i_blocks = 0;
1366 else
1367 inode->i_blocks =
1368 ocfs2_align_bytes_to_sectors(i_size_read(inode));
1370 inode->i_uid = be32_to_cpu(lvb->lvb_iuid);
1371 inode->i_gid = be32_to_cpu(lvb->lvb_igid);
1372 inode->i_mode = be16_to_cpu(lvb->lvb_imode);
1373 inode->i_nlink = be16_to_cpu(lvb->lvb_inlink);
1374 ocfs2_unpack_timespec(&inode->i_atime,
1375 be64_to_cpu(lvb->lvb_iatime_packed));
1376 ocfs2_unpack_timespec(&inode->i_mtime,
1377 be64_to_cpu(lvb->lvb_imtime_packed));
1378 ocfs2_unpack_timespec(&inode->i_ctime,
1379 be64_to_cpu(lvb->lvb_ictime_packed));
1380 spin_unlock(&oi->ip_lock);
1382 mlog_exit_void();
1385 static inline int ocfs2_meta_lvb_is_trustable(struct ocfs2_lock_res *lockres)
1387 struct ocfs2_meta_lvb *lvb = (struct ocfs2_meta_lvb *) lockres->l_lksb.lvb;
1389 if (be32_to_cpu(lvb->lvb_version) == OCFS2_LVB_VERSION)
1390 return 1;
1391 return 0;
1394 /* Determine whether a lock resource needs to be refreshed, and
1395 * arbitrate who gets to refresh it.
1397 * 0 means no refresh needed.
1399 * > 0 means you need to refresh this and you MUST call
1400 * ocfs2_complete_lock_res_refresh afterwards. */
1401 static int ocfs2_should_refresh_lock_res(struct ocfs2_lock_res *lockres)
1403 unsigned long flags;
1404 int status = 0;
1406 mlog_entry_void();
1408 refresh_check:
1409 spin_lock_irqsave(&lockres->l_lock, flags);
1410 if (!(lockres->l_flags & OCFS2_LOCK_NEEDS_REFRESH)) {
1411 spin_unlock_irqrestore(&lockres->l_lock, flags);
1412 goto bail;
1415 if (lockres->l_flags & OCFS2_LOCK_REFRESHING) {
1416 spin_unlock_irqrestore(&lockres->l_lock, flags);
1418 ocfs2_wait_on_refreshing_lock(lockres);
1419 goto refresh_check;
1422 /* Ok, I'll be the one to refresh this lock. */
1423 lockres_or_flags(lockres, OCFS2_LOCK_REFRESHING);
1424 spin_unlock_irqrestore(&lockres->l_lock, flags);
1426 status = 1;
1427 bail:
1428 mlog_exit(status);
1429 return status;
1432 /* If status is non zero, I'll mark it as not being in refresh
1433 * anymroe, but i won't clear the needs refresh flag. */
1434 static inline void ocfs2_complete_lock_res_refresh(struct ocfs2_lock_res *lockres,
1435 int status)
1437 unsigned long flags;
1438 mlog_entry_void();
1440 spin_lock_irqsave(&lockres->l_lock, flags);
1441 lockres_clear_flags(lockres, OCFS2_LOCK_REFRESHING);
1442 if (!status)
1443 lockres_clear_flags(lockres, OCFS2_LOCK_NEEDS_REFRESH);
1444 spin_unlock_irqrestore(&lockres->l_lock, flags);
1446 wake_up(&lockres->l_event);
1448 mlog_exit_void();
1451 /* may or may not return a bh if it went to disk. */
1452 static int ocfs2_meta_lock_update(struct inode *inode,
1453 struct buffer_head **bh)
1455 int status = 0;
1456 struct ocfs2_inode_info *oi = OCFS2_I(inode);
1457 struct ocfs2_lock_res *lockres;
1458 struct ocfs2_dinode *fe;
1460 mlog_entry_void();
1462 spin_lock(&oi->ip_lock);
1463 if (oi->ip_flags & OCFS2_INODE_DELETED) {
1464 mlog(0, "Orphaned inode %llu was deleted while we "
1465 "were waiting on a lock. ip_flags = 0x%x\n",
1466 (unsigned long long)oi->ip_blkno, oi->ip_flags);
1467 spin_unlock(&oi->ip_lock);
1468 status = -ENOENT;
1469 goto bail;
1471 spin_unlock(&oi->ip_lock);
1473 lockres = &oi->ip_meta_lockres;
1475 if (!ocfs2_should_refresh_lock_res(lockres))
1476 goto bail;
1478 /* This will discard any caching information we might have had
1479 * for the inode metadata. */
1480 ocfs2_metadata_cache_purge(inode);
1482 /* will do nothing for inode types that don't use the extent
1483 * map (directories, bitmap files, etc) */
1484 ocfs2_extent_map_trunc(inode, 0);
1486 if (ocfs2_meta_lvb_is_trustable(lockres)) {
1487 mlog(0, "Trusting LVB on inode %llu\n",
1488 (unsigned long long)oi->ip_blkno);
1489 ocfs2_refresh_inode_from_lvb(inode);
1490 } else {
1491 /* Boo, we have to go to disk. */
1492 /* read bh, cast, ocfs2_refresh_inode */
1493 status = ocfs2_read_block(OCFS2_SB(inode->i_sb), oi->ip_blkno,
1494 bh, OCFS2_BH_CACHED, inode);
1495 if (status < 0) {
1496 mlog_errno(status);
1497 goto bail_refresh;
1499 fe = (struct ocfs2_dinode *) (*bh)->b_data;
1501 /* This is a good chance to make sure we're not
1502 * locking an invalid object.
1504 * We bug on a stale inode here because we checked
1505 * above whether it was wiped from disk. The wiping
1506 * node provides a guarantee that we receive that
1507 * message and can mark the inode before dropping any
1508 * locks associated with it. */
1509 if (!OCFS2_IS_VALID_DINODE(fe)) {
1510 OCFS2_RO_ON_INVALID_DINODE(inode->i_sb, fe);
1511 status = -EIO;
1512 goto bail_refresh;
1514 mlog_bug_on_msg(inode->i_generation !=
1515 le32_to_cpu(fe->i_generation),
1516 "Invalid dinode %llu disk generation: %u "
1517 "inode->i_generation: %u\n",
1518 (unsigned long long)oi->ip_blkno,
1519 le32_to_cpu(fe->i_generation),
1520 inode->i_generation);
1521 mlog_bug_on_msg(le64_to_cpu(fe->i_dtime) ||
1522 !(fe->i_flags & cpu_to_le32(OCFS2_VALID_FL)),
1523 "Stale dinode %llu dtime: %llu flags: 0x%x\n",
1524 (unsigned long long)oi->ip_blkno,
1525 (unsigned long long)le64_to_cpu(fe->i_dtime),
1526 le32_to_cpu(fe->i_flags));
1528 ocfs2_refresh_inode(inode, fe);
1531 status = 0;
1532 bail_refresh:
1533 ocfs2_complete_lock_res_refresh(lockres, status);
1534 bail:
1535 mlog_exit(status);
1536 return status;
1539 static int ocfs2_assign_bh(struct inode *inode,
1540 struct buffer_head **ret_bh,
1541 struct buffer_head *passed_bh)
1543 int status;
1545 if (passed_bh) {
1546 /* Ok, the update went to disk for us, use the
1547 * returned bh. */
1548 *ret_bh = passed_bh;
1549 get_bh(*ret_bh);
1551 return 0;
1554 status = ocfs2_read_block(OCFS2_SB(inode->i_sb),
1555 OCFS2_I(inode)->ip_blkno,
1556 ret_bh,
1557 OCFS2_BH_CACHED,
1558 inode);
1559 if (status < 0)
1560 mlog_errno(status);
1562 return status;
1565 /*
1566 * returns < 0 error if the callback will never be called, otherwise
1567 * the result of the lock will be communicated via the callback.
1568 */
1569 int ocfs2_meta_lock_full(struct inode *inode,
1570 struct ocfs2_journal_handle *handle,
1571 struct buffer_head **ret_bh,
1572 int ex,
1573 int arg_flags)
1575 int status, level, dlm_flags, acquired;
1576 struct ocfs2_lock_res *lockres;
1577 struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
1578 struct buffer_head *local_bh = NULL;
1580 BUG_ON(!inode);
1582 mlog_entry_void();
1584 mlog(0, "inode %llu, take %s META lock\n",
1585 (unsigned long long)OCFS2_I(inode)->ip_blkno,
1586 ex ? "EXMODE" : "PRMODE");
1588 status = 0;
1589 acquired = 0;
1590 /* We'll allow faking a readonly metadata lock for
1591 * rodevices. */
1592 if (ocfs2_is_hard_readonly(osb)) {
1593 if (ex)
1594 status = -EROFS;
1595 goto bail;
1598 if (!(arg_flags & OCFS2_META_LOCK_RECOVERY))
1599 wait_event(osb->recovery_event,
1600 ocfs2_node_map_is_empty(osb, &osb->recovery_map));
1602 acquired = 0;
1603 lockres = &OCFS2_I(inode)->ip_meta_lockres;
1604 level = ex ? LKM_EXMODE : LKM_PRMODE;
1605 dlm_flags = 0;
1606 if (arg_flags & OCFS2_META_LOCK_NOQUEUE)
1607 dlm_flags |= LKM_NOQUEUE;
1609 status = ocfs2_cluster_lock(osb, lockres, level, dlm_flags, arg_flags);
1610 if (status < 0) {
1611 if (status != -EAGAIN && status != -EIOCBRETRY)
1612 mlog_errno(status);
1613 goto bail;
1616 /* Notify the error cleanup path to drop the cluster lock. */
1617 acquired = 1;
1619 /* We wait twice because a node may have died while we were in
1620 * the lower dlm layers. The second time though, we've
1621 * committed to owning this lock so we don't allow signals to
1622 * abort the operation. */
1623 if (!(arg_flags & OCFS2_META_LOCK_RECOVERY))
1624 wait_event(osb->recovery_event,
1625 ocfs2_node_map_is_empty(osb, &osb->recovery_map));
1627 /* This is fun. The caller may want a bh back, or it may
1628 * not. ocfs2_meta_lock_update definitely wants one in, but
1629 * may or may not read one, depending on what's in the
1630 * LVB. The result of all of this is that we've *only* gone to
1631 * disk if we have to, so the complexity is worthwhile. */
1632 status = ocfs2_meta_lock_update(inode, &local_bh);
1633 if (status < 0) {
1634 if (status != -ENOENT)
1635 mlog_errno(status);
1636 goto bail;
1639 if (ret_bh) {
1640 status = ocfs2_assign_bh(inode, ret_bh, local_bh);
1641 if (status < 0) {
1642 mlog_errno(status);
1643 goto bail;
1647 if (handle) {
1648 status = ocfs2_handle_add_lock(handle, inode);
1649 if (status < 0)
1650 mlog_errno(status);
1653 bail:
1654 if (status < 0) {
1655 if (ret_bh && (*ret_bh)) {
1656 brelse(*ret_bh);
1657 *ret_bh = NULL;
1659 if (acquired)
1660 ocfs2_meta_unlock(inode, ex);
1663 if (local_bh)
1664 brelse(local_bh);
1666 mlog_exit(status);
1667 return status;
1670 /*
1671 * This is working around a lock inversion between tasks acquiring DLM locks
1672 * while holding a page lock and the vote thread which blocks dlm lock acquiry
1673 * while acquiring page locks.
1675 * ** These _with_page variantes are only intended to be called from aop
1676 * methods that hold page locks and return a very specific *positive* error
1677 * code that aop methods pass up to the VFS -- test for errors with != 0. **
1679 * The DLM is called such that it returns -EAGAIN if it would have blocked
1680 * waiting for the vote thread. In that case we unlock our page so the vote
1681 * thread can make progress. Once we've done this we have to return
1682 * AOP_TRUNCATED_PAGE so the aop method that called us can bubble that back up
1683 * into the VFS who will then immediately retry the aop call.
1685 * We do a blocking lock and immediate unlock before returning, though, so that
1686 * the lock has a great chance of being cached on this node by the time the VFS
1687 * calls back to retry the aop. This has a potential to livelock as nodes
1688 * ping locks back and forth, but that's a risk we're willing to take to avoid
1689 * the lock inversion simply.
1690 */
1691 int ocfs2_meta_lock_with_page(struct inode *inode,
1692 struct ocfs2_journal_handle *handle,
1693 struct buffer_head **ret_bh,
1694 int ex,
1695 struct page *page)
1697 int ret;
1699 ret = ocfs2_meta_lock_full(inode, handle, ret_bh, ex,
1700 OCFS2_LOCK_NONBLOCK);
1701 if (ret == -EAGAIN) {
1702 unlock_page(page);
1703 if (ocfs2_meta_lock(inode, handle, ret_bh, ex) == 0)
1704 ocfs2_meta_unlock(inode, ex);
1705 ret = AOP_TRUNCATED_PAGE;
1708 return ret;
1711 void ocfs2_meta_unlock(struct inode *inode,
1712 int ex)
1714 int level = ex ? LKM_EXMODE : LKM_PRMODE;
1715 struct ocfs2_lock_res *lockres = &OCFS2_I(inode)->ip_meta_lockres;
1717 mlog_entry_void();
1719 mlog(0, "inode %llu drop %s META lock\n",
1720 (unsigned long long)OCFS2_I(inode)->ip_blkno,
1721 ex ? "EXMODE" : "PRMODE");
1723 if (!ocfs2_is_hard_readonly(OCFS2_SB(inode->i_sb)))
1724 ocfs2_cluster_unlock(OCFS2_SB(inode->i_sb), lockres, level);
1726 mlog_exit_void();
1729 int ocfs2_super_lock(struct ocfs2_super *osb,
1730 int ex)
1732 int status;
1733 int level = ex ? LKM_EXMODE : LKM_PRMODE;
1734 struct ocfs2_lock_res *lockres = &osb->osb_super_lockres;
1735 struct buffer_head *bh;
1736 struct ocfs2_slot_info *si = osb->slot_info;
1738 mlog_entry_void();
1740 if (ocfs2_is_hard_readonly(osb))
1741 return -EROFS;
1743 status = ocfs2_cluster_lock(osb, lockres, level, 0, 0);
1744 if (status < 0) {
1745 mlog_errno(status);
1746 goto bail;
1749 /* The super block lock path is really in the best position to
1750 * know when resources covered by the lock need to be
1751 * refreshed, so we do it here. Of course, making sense of
1752 * everything is up to the caller :) */
1753 status = ocfs2_should_refresh_lock_res(lockres);
1754 if (status < 0) {
1755 mlog_errno(status);
1756 goto bail;
1758 if (status) {
1759 bh = si->si_bh;
1760 status = ocfs2_read_block(osb, bh->b_blocknr, &bh, 0,
1761 si->si_inode);
1762 if (status == 0)
1763 ocfs2_update_slot_info(si);
1765 ocfs2_complete_lock_res_refresh(lockres, status);
1767 if (status < 0)
1768 mlog_errno(status);
1770 bail:
1771 mlog_exit(status);
1772 return status;
1775 void ocfs2_super_unlock(struct ocfs2_super *osb,
1776 int ex)
1778 int level = ex ? LKM_EXMODE : LKM_PRMODE;
1779 struct ocfs2_lock_res *lockres = &osb->osb_super_lockres;
1781 ocfs2_cluster_unlock(osb, lockres, level);
1784 int ocfs2_rename_lock(struct ocfs2_super *osb)
1786 int status;
1787 struct ocfs2_lock_res *lockres = &osb->osb_rename_lockres;
1789 if (ocfs2_is_hard_readonly(osb))
1790 return -EROFS;
1792 status = ocfs2_cluster_lock(osb, lockres, LKM_EXMODE, 0, 0);
1793 if (status < 0)
1794 mlog_errno(status);
1796 return status;
1799 void ocfs2_rename_unlock(struct ocfs2_super *osb)
1801 struct ocfs2_lock_res *lockres = &osb->osb_rename_lockres;
1803 ocfs2_cluster_unlock(osb, lockres, LKM_EXMODE);
1806 /* Reference counting of the dlm debug structure. We want this because
1807 * open references on the debug inodes can live on after a mount, so
1808 * we can't rely on the ocfs2_super to always exist. */
1809 static void ocfs2_dlm_debug_free(struct kref *kref)
1811 struct ocfs2_dlm_debug *dlm_debug;
1813 dlm_debug = container_of(kref, struct ocfs2_dlm_debug, d_refcnt);
1815 kfree(dlm_debug);
1818 void ocfs2_put_dlm_debug(struct ocfs2_dlm_debug *dlm_debug)
1820 if (dlm_debug)
1821 kref_put(&dlm_debug->d_refcnt, ocfs2_dlm_debug_free);
1824 static void ocfs2_get_dlm_debug(struct ocfs2_dlm_debug *debug)
1826 kref_get(&debug->d_refcnt);
1829 struct ocfs2_dlm_debug *ocfs2_new_dlm_debug(void)
1831 struct ocfs2_dlm_debug *dlm_debug;
1833 dlm_debug = kmalloc(sizeof(struct ocfs2_dlm_debug), GFP_KERNEL);
1834 if (!dlm_debug) {
1835 mlog_errno(-ENOMEM);
1836 goto out;
1839 kref_init(&dlm_debug->d_refcnt);
1840 INIT_LIST_HEAD(&dlm_debug->d_lockres_tracking);
1841 dlm_debug->d_locking_state = NULL;
1842 out:
1843 return dlm_debug;
1846 /* Access to this is arbitrated for us via seq_file->sem. */
1847 struct ocfs2_dlm_seq_priv {
1848 struct ocfs2_dlm_debug *p_dlm_debug;
1849 struct ocfs2_lock_res p_iter_res;
1850 struct ocfs2_lock_res p_tmp_res;
1851 };
1853 static struct ocfs2_lock_res *ocfs2_dlm_next_res(struct ocfs2_lock_res *start,
1854 struct ocfs2_dlm_seq_priv *priv)
1856 struct ocfs2_lock_res *iter, *ret = NULL;
1857 struct ocfs2_dlm_debug *dlm_debug = priv->p_dlm_debug;
1859 assert_spin_locked(&ocfs2_dlm_tracking_lock);
1861 list_for_each_entry(iter, &start->l_debug_list, l_debug_list) {
1862 /* discover the head of the list */
1863 if (&iter->l_debug_list == &dlm_debug->d_lockres_tracking) {
1864 mlog(0, "End of list found, %p\n", ret);
1865 break;
1868 /* We track our "dummy" iteration lockres' by a NULL
1869 * l_ops field. */
1870 if (iter->l_ops != NULL) {
1871 ret = iter;
1872 break;
1876 return ret;
1879 static void *ocfs2_dlm_seq_start(struct seq_file *m, loff_t *pos)
1881 struct ocfs2_dlm_seq_priv *priv = m->private;
1882 struct ocfs2_lock_res *iter;
1884 spin_lock(&ocfs2_dlm_tracking_lock);
1885 iter = ocfs2_dlm_next_res(&priv->p_iter_res, priv);
1886 if (iter) {
1887 /* Since lockres' have the lifetime of their container
1888 * (which can be inodes, ocfs2_supers, etc) we want to
1889 * copy this out to a temporary lockres while still
1890 * under the spinlock. Obviously after this we can't
1891 * trust any pointers on the copy returned, but that's
1892 * ok as the information we want isn't typically held
1893 * in them. */
1894 priv->p_tmp_res = *iter;
1895 iter = &priv->p_tmp_res;
1897 spin_unlock(&ocfs2_dlm_tracking_lock);
1899 return iter;
1902 static void ocfs2_dlm_seq_stop(struct seq_file *m, void *v)
1906 static void *ocfs2_dlm_seq_next(struct seq_file *m, void *v, loff_t *pos)
1908 struct ocfs2_dlm_seq_priv *priv = m->private;
1909 struct ocfs2_lock_res *iter = v;
1910 struct ocfs2_lock_res *dummy = &priv->p_iter_res;
1912 spin_lock(&ocfs2_dlm_tracking_lock);
1913 iter = ocfs2_dlm_next_res(iter, priv);
1914 list_del_init(&dummy->l_debug_list);
1915 if (iter) {
1916 list_add(&dummy->l_debug_list, &iter->l_debug_list);
1917 priv->p_tmp_res = *iter;
1918 iter = &priv->p_tmp_res;
1920 spin_unlock(&ocfs2_dlm_tracking_lock);
1922 return iter;
1925 /* So that debugfs.ocfs2 can determine which format is being used */
1926 #define OCFS2_DLM_DEBUG_STR_VERSION 1
1927 static int ocfs2_dlm_seq_show(struct seq_file *m, void *v)
1929 int i;
1930 char *lvb;
1931 struct ocfs2_lock_res *lockres = v;
1933 if (!lockres)
1934 return -EINVAL;
1936 seq_printf(m, "0x%x\t"
1937 "%.*s\t"
1938 "%d\t"
1939 "0x%lx\t"
1940 "0x%x\t"
1941 "0x%x\t"
1942 "%u\t"
1943 "%u\t"
1944 "%d\t"
1945 "%d\t",
1946 OCFS2_DLM_DEBUG_STR_VERSION,
1947 OCFS2_LOCK_ID_MAX_LEN, lockres->l_name,
1948 lockres->l_level,
1949 lockres->l_flags,
1950 lockres->l_action,
1951 lockres->l_unlock_action,
1952 lockres->l_ro_holders,
1953 lockres->l_ex_holders,
1954 lockres->l_requested,
1955 lockres->l_blocking);
1957 /* Dump the raw LVB */
1958 lvb = lockres->l_lksb.lvb;
1959 for(i = 0; i < DLM_LVB_LEN; i++)
1960 seq_printf(m, "0x%x\t", lvb[i]);
1962 /* End the line */
1963 seq_printf(m, "\n");
1964 return 0;
1967 static struct seq_operations ocfs2_dlm_seq_ops = {
1968 .start = ocfs2_dlm_seq_start,
1969 .stop = ocfs2_dlm_seq_stop,
1970 .next = ocfs2_dlm_seq_next,
1971 .show = ocfs2_dlm_seq_show,
1972 };
1974 static int ocfs2_dlm_debug_release(struct inode *inode, struct file *file)
1976 struct seq_file *seq = (struct seq_file *) file->private_data;
1977 struct ocfs2_dlm_seq_priv *priv = seq->private;
1978 struct ocfs2_lock_res *res = &priv->p_iter_res;
1980 ocfs2_remove_lockres_tracking(res);
1981 ocfs2_put_dlm_debug(priv->p_dlm_debug);
1982 return seq_release_private(inode, file);
1985 static int ocfs2_dlm_debug_open(struct inode *inode, struct file *file)
1987 int ret;
1988 struct ocfs2_dlm_seq_priv *priv;
1989 struct seq_file *seq;
1990 struct ocfs2_super *osb;
1992 priv = kzalloc(sizeof(struct ocfs2_dlm_seq_priv), GFP_KERNEL);
1993 if (!priv) {
1994 ret = -ENOMEM;
1995 mlog_errno(ret);
1996 goto out;
1998 osb = (struct ocfs2_super *) inode->u.generic_ip;
1999 ocfs2_get_dlm_debug(osb->osb_dlm_debug);
2000 priv->p_dlm_debug = osb->osb_dlm_debug;
2001 INIT_LIST_HEAD(&priv->p_iter_res.l_debug_list);
2003 ret = seq_open(file, &ocfs2_dlm_seq_ops);
2004 if (ret) {
2005 kfree(priv);
2006 mlog_errno(ret);
2007 goto out;
2010 seq = (struct seq_file *) file->private_data;
2011 seq->private = priv;
2013 ocfs2_add_lockres_tracking(&priv->p_iter_res,
2014 priv->p_dlm_debug);
2016 out:
2017 return ret;
2020 static const struct file_operations ocfs2_dlm_debug_fops = {
2021 .open = ocfs2_dlm_debug_open,
2022 .release = ocfs2_dlm_debug_release,
2023 .read = seq_read,
2024 .llseek = seq_lseek,
2025 };
2027 static int ocfs2_dlm_init_debug(struct ocfs2_super *osb)
2029 int ret = 0;
2030 struct ocfs2_dlm_debug *dlm_debug = osb->osb_dlm_debug;
2032 dlm_debug->d_locking_state = debugfs_create_file("locking_state",
2033 S_IFREG|S_IRUSR,
2034 osb->osb_debug_root,
2035 osb,
2036 &ocfs2_dlm_debug_fops);
2037 if (!dlm_debug->d_locking_state) {
2038 ret = -EINVAL;
2039 mlog(ML_ERROR,
2040 "Unable to create locking state debugfs file.\n");
2041 goto out;
2044 ocfs2_get_dlm_debug(dlm_debug);
2045 out:
2046 return ret;
2049 static void ocfs2_dlm_shutdown_debug(struct ocfs2_super *osb)
2051 struct ocfs2_dlm_debug *dlm_debug = osb->osb_dlm_debug;
2053 if (dlm_debug) {
2054 debugfs_remove(dlm_debug->d_locking_state);
2055 ocfs2_put_dlm_debug(dlm_debug);
2059 int ocfs2_dlm_init(struct ocfs2_super *osb)
2061 int status;
2062 u32 dlm_key;
2063 struct dlm_ctxt *dlm;
2065 mlog_entry_void();
2067 status = ocfs2_dlm_init_debug(osb);
2068 if (status < 0) {
2069 mlog_errno(status);
2070 goto bail;
2073 /* launch vote thread */
2074 osb->vote_task = kthread_run(ocfs2_vote_thread, osb, "ocfs2vote");
2075 if (IS_ERR(osb->vote_task)) {
2076 status = PTR_ERR(osb->vote_task);
2077 osb->vote_task = NULL;
2078 mlog_errno(status);
2079 goto bail;
2082 /* used by the dlm code to make message headers unique, each
2083 * node in this domain must agree on this. */
2084 dlm_key = crc32_le(0, osb->uuid_str, strlen(osb->uuid_str));
2086 /* for now, uuid == domain */
2087 dlm = dlm_register_domain(osb->uuid_str, dlm_key);
2088 if (IS_ERR(dlm)) {
2089 status = PTR_ERR(dlm);
2090 mlog_errno(status);
2091 goto bail;
2094 ocfs2_super_lock_res_init(&osb->osb_super_lockres, osb);
2095 ocfs2_rename_lock_res_init(&osb->osb_rename_lockres, osb);
2097 dlm_register_eviction_cb(dlm, &osb->osb_eviction_cb);
2099 osb->dlm = dlm;
2101 status = 0;
2102 bail:
2103 if (status < 0) {
2104 ocfs2_dlm_shutdown_debug(osb);
2105 if (osb->vote_task)
2106 kthread_stop(osb->vote_task);
2109 mlog_exit(status);
2110 return status;
2113 void ocfs2_dlm_shutdown(struct ocfs2_super *osb)
2115 mlog_entry_void();
2117 dlm_unregister_eviction_cb(&osb->osb_eviction_cb);
2119 ocfs2_drop_osb_locks(osb);
2121 if (osb->vote_task) {
2122 kthread_stop(osb->vote_task);
2123 osb->vote_task = NULL;
2126 ocfs2_lock_res_free(&osb->osb_super_lockres);
2127 ocfs2_lock_res_free(&osb->osb_rename_lockres);
2129 dlm_unregister_domain(osb->dlm);
2130 osb->dlm = NULL;
2132 ocfs2_dlm_shutdown_debug(osb);
2134 mlog_exit_void();
2137 static void ocfs2_unlock_ast_func(void *opaque, enum dlm_status status)
2139 struct ocfs2_lock_res *lockres = opaque;
2140 unsigned long flags;
2142 mlog_entry_void();
2144 mlog(0, "UNLOCK AST called on lock %s, action = %d\n", lockres->l_name,
2145 lockres->l_unlock_action);
2147 spin_lock_irqsave(&lockres->l_lock, flags);
2148 /* We tried to cancel a convert request, but it was already
2149 * granted. All we want to do here is clear our unlock
2150 * state. The wake_up call done at the bottom is redundant
2151 * (ocfs2_prepare_cancel_convert doesn't sleep on this) but doesn't
2152 * hurt anything anyway */
2153 if (status == DLM_CANCELGRANT &&
2154 lockres->l_unlock_action == OCFS2_UNLOCK_CANCEL_CONVERT) {
2155 mlog(0, "Got cancelgrant for %s\n", lockres->l_name);
2157 /* We don't clear the busy flag in this case as it
2158 * should have been cleared by the ast which the dlm
2159 * has called. */
2160 goto complete_unlock;
2163 if (status != DLM_NORMAL) {
2164 mlog(ML_ERROR, "Dlm passes status %d for lock %s, "
2165 "unlock_action %d\n", status, lockres->l_name,
2166 lockres->l_unlock_action);
2167 spin_unlock_irqrestore(&lockres->l_lock, flags);
2168 return;
2171 switch(lockres->l_unlock_action) {
2172 case OCFS2_UNLOCK_CANCEL_CONVERT:
2173 mlog(0, "Cancel convert success for %s\n", lockres->l_name);
2174 lockres->l_action = OCFS2_AST_INVALID;
2175 break;
2176 case OCFS2_UNLOCK_DROP_LOCK:
2177 lockres->l_level = LKM_IVMODE;
2178 break;
2179 default:
2180 BUG();
2183 lockres_clear_flags(lockres, OCFS2_LOCK_BUSY);
2184 complete_unlock:
2185 lockres->l_unlock_action = OCFS2_UNLOCK_INVALID;
2186 spin_unlock_irqrestore(&lockres->l_lock, flags);
2188 wake_up(&lockres->l_event);
2190 mlog_exit_void();
2193 typedef void (ocfs2_pre_drop_cb_t)(struct ocfs2_lock_res *, void *);
2195 struct drop_lock_cb {
2196 ocfs2_pre_drop_cb_t *drop_func;
2197 void *drop_data;
2198 };
2200 static int ocfs2_drop_lock(struct ocfs2_super *osb,
2201 struct ocfs2_lock_res *lockres,
2202 struct drop_lock_cb *dcb)
2204 enum dlm_status status;
2205 unsigned long flags;
2207 /* We didn't get anywhere near actually using this lockres. */
2208 if (!(lockres->l_flags & OCFS2_LOCK_INITIALIZED))
2209 goto out;
2211 spin_lock_irqsave(&lockres->l_lock, flags);
2213 mlog_bug_on_msg(!(lockres->l_flags & OCFS2_LOCK_FREEING),
2214 "lockres %s, flags 0x%lx\n",
2215 lockres->l_name, lockres->l_flags);
2217 while (lockres->l_flags & OCFS2_LOCK_BUSY) {
2218 mlog(0, "waiting on busy lock \"%s\": flags = %lx, action = "
2219 "%u, unlock_action = %u\n",
2220 lockres->l_name, lockres->l_flags, lockres->l_action,
2221 lockres->l_unlock_action);
2223 spin_unlock_irqrestore(&lockres->l_lock, flags);
2225 /* XXX: Today we just wait on any busy
2226 * locks... Perhaps we need to cancel converts in the
2227 * future? */
2228 ocfs2_wait_on_busy_lock(lockres);
2230 spin_lock_irqsave(&lockres->l_lock, flags);
2233 if (dcb)
2234 dcb->drop_func(lockres, dcb->drop_data);
2236 if (lockres->l_flags & OCFS2_LOCK_BUSY)
2237 mlog(ML_ERROR, "destroying busy lock: \"%s\"\n",
2238 lockres->l_name);
2239 if (lockres->l_flags & OCFS2_LOCK_BLOCKED)
2240 mlog(0, "destroying blocked lock: \"%s\"\n", lockres->l_name);
2242 if (!(lockres->l_flags & OCFS2_LOCK_ATTACHED)) {
2243 spin_unlock_irqrestore(&lockres->l_lock, flags);
2244 goto out;
2247 lockres_clear_flags(lockres, OCFS2_LOCK_ATTACHED);
2249 /* make sure we never get here while waiting for an ast to
2250 * fire. */
2251 BUG_ON(lockres->l_action != OCFS2_AST_INVALID);
2253 /* is this necessary? */
2254 lockres_or_flags(lockres, OCFS2_LOCK_BUSY);
2255 lockres->l_unlock_action = OCFS2_UNLOCK_DROP_LOCK;
2256 spin_unlock_irqrestore(&lockres->l_lock, flags);
2258 mlog(0, "lock %s\n", lockres->l_name);
2260 status = dlmunlock(osb->dlm, &lockres->l_lksb, LKM_VALBLK,
2261 lockres->l_ops->unlock_ast, lockres);
2262 if (status != DLM_NORMAL) {
2263 ocfs2_log_dlm_error("dlmunlock", status, lockres);
2264 mlog(ML_ERROR, "lockres flags: %lu\n", lockres->l_flags);
2265 dlm_print_one_lock(lockres->l_lksb.lockid);
2266 BUG();
2268 mlog(0, "lock %s, successfull return from dlmunlock\n",
2269 lockres->l_name);
2271 ocfs2_wait_on_busy_lock(lockres);
2272 out:
2273 mlog_exit(0);
2274 return 0;
2277 /* Mark the lockres as being dropped. It will no longer be
2278 * queued if blocking, but we still may have to wait on it
2279 * being dequeued from the vote thread before we can consider
2280 * it safe to drop.
2282 * You can *not* attempt to call cluster_lock on this lockres anymore. */
2283 void ocfs2_mark_lockres_freeing(struct ocfs2_lock_res *lockres)
2285 int status;
2286 struct ocfs2_mask_waiter mw;
2287 unsigned long flags;
2289 ocfs2_init_mask_waiter(&mw);
2291 spin_lock_irqsave(&lockres->l_lock, flags);
2292 lockres->l_flags |= OCFS2_LOCK_FREEING;
2293 while (lockres->l_flags & OCFS2_LOCK_QUEUED) {
2294 lockres_add_mask_waiter(lockres, &mw, OCFS2_LOCK_QUEUED, 0);
2295 spin_unlock_irqrestore(&lockres->l_lock, flags);
2297 mlog(0, "Waiting on lockres %s\n", lockres->l_name);
2299 status = ocfs2_wait_for_mask(&mw);
2300 if (status)
2301 mlog_errno(status);
2303 spin_lock_irqsave(&lockres->l_lock, flags);
2305 spin_unlock_irqrestore(&lockres->l_lock, flags);
2308 static void ocfs2_drop_osb_locks(struct ocfs2_super *osb)
2310 int status;
2312 mlog_entry_void();
2314 ocfs2_mark_lockres_freeing(&osb->osb_super_lockres);
2316 status = ocfs2_drop_lock(osb, &osb->osb_super_lockres, NULL);
2317 if (status < 0)
2318 mlog_errno(status);
2320 ocfs2_mark_lockres_freeing(&osb->osb_rename_lockres);
2322 status = ocfs2_drop_lock(osb, &osb->osb_rename_lockres, NULL);
2323 if (status < 0)
2324 mlog_errno(status);
2326 mlog_exit(status);
2329 static void ocfs2_meta_pre_drop(struct ocfs2_lock_res *lockres, void *data)
2331 struct inode *inode = data;
2333 /* the metadata lock requires a bit more work as we have an
2334 * LVB to worry about. */
2335 if (lockres->l_flags & OCFS2_LOCK_ATTACHED &&
2336 lockres->l_level == LKM_EXMODE &&
2337 !(lockres->l_flags & OCFS2_LOCK_NEEDS_REFRESH))
2338 __ocfs2_stuff_meta_lvb(inode);
2341 int ocfs2_drop_inode_locks(struct inode *inode)
2343 int status, err;
2344 struct drop_lock_cb meta_dcb = { ocfs2_meta_pre_drop, inode, };
2346 mlog_entry_void();
2348 /* No need to call ocfs2_mark_lockres_freeing here -
2349 * ocfs2_clear_inode has done it for us. */
2351 err = ocfs2_drop_lock(OCFS2_SB(inode->i_sb),
2352 &OCFS2_I(inode)->ip_data_lockres,
2353 NULL);
2354 if (err < 0)
2355 mlog_errno(err);
2357 status = err;
2359 err = ocfs2_drop_lock(OCFS2_SB(inode->i_sb),
2360 &OCFS2_I(inode)->ip_meta_lockres,
2361 &meta_dcb);
2362 if (err < 0)
2363 mlog_errno(err);
2364 if (err < 0 && !status)
2365 status = err;
2367 err = ocfs2_drop_lock(OCFS2_SB(inode->i_sb),
2368 &OCFS2_I(inode)->ip_rw_lockres,
2369 NULL);
2370 if (err < 0)
2371 mlog_errno(err);
2372 if (err < 0 && !status)
2373 status = err;
2375 mlog_exit(status);
2376 return status;
2379 static void ocfs2_prepare_downconvert(struct ocfs2_lock_res *lockres,
2380 int new_level)
2382 assert_spin_locked(&lockres->l_lock);
2384 BUG_ON(lockres->l_blocking <= LKM_NLMODE);
2386 if (lockres->l_level <= new_level) {
2387 mlog(ML_ERROR, "lockres->l_level (%u) <= new_level (%u)\n",
2388 lockres->l_level, new_level);
2389 BUG();
2392 mlog(0, "lock %s, new_level = %d, l_blocking = %d\n",
2393 lockres->l_name, new_level, lockres->l_blocking);
2395 lockres->l_action = OCFS2_AST_DOWNCONVERT;
2396 lockres->l_requested = new_level;
2397 lockres_or_flags(lockres, OCFS2_LOCK_BUSY);
2400 static int ocfs2_downconvert_lock(struct ocfs2_super *osb,
2401 struct ocfs2_lock_res *lockres,
2402 int new_level,
2403 int lvb)
2405 int ret, dlm_flags = LKM_CONVERT;
2406 enum dlm_status status;
2408 mlog_entry_void();
2410 if (lvb)
2411 dlm_flags |= LKM_VALBLK;
2413 status = dlmlock(osb->dlm,
2414 new_level,
2415 &lockres->l_lksb,
2416 dlm_flags,
2417 lockres->l_name,
2418 lockres->l_ops->ast,
2419 lockres,
2420 lockres->l_ops->bast);
2421 if (status != DLM_NORMAL) {
2422 ocfs2_log_dlm_error("dlmlock", status, lockres);
2423 ret = -EINVAL;
2424 ocfs2_recover_from_dlm_error(lockres, 1);
2425 goto bail;
2428 ret = 0;
2429 bail:
2430 mlog_exit(ret);
2431 return ret;
2434 /* returns 1 when the caller should unlock and call dlmunlock */
2435 static int ocfs2_prepare_cancel_convert(struct ocfs2_super *osb,
2436 struct ocfs2_lock_res *lockres)
2438 assert_spin_locked(&lockres->l_lock);
2440 mlog_entry_void();
2441 mlog(0, "lock %s\n", lockres->l_name);
2443 if (lockres->l_unlock_action == OCFS2_UNLOCK_CANCEL_CONVERT) {
2444 /* If we're already trying to cancel a lock conversion
2445 * then just drop the spinlock and allow the caller to
2446 * requeue this lock. */
2448 mlog(0, "Lockres %s, skip convert\n", lockres->l_name);
2449 return 0;
2452 /* were we in a convert when we got the bast fire? */
2453 BUG_ON(lockres->l_action != OCFS2_AST_CONVERT &&
2454 lockres->l_action != OCFS2_AST_DOWNCONVERT);
2455 /* set things up for the unlockast to know to just
2456 * clear out the ast_action and unset busy, etc. */
2457 lockres->l_unlock_action = OCFS2_UNLOCK_CANCEL_CONVERT;
2459 mlog_bug_on_msg(!(lockres->l_flags & OCFS2_LOCK_BUSY),
2460 "lock %s, invalid flags: 0x%lx\n",
2461 lockres->l_name, lockres->l_flags);
2463 return 1;
2466 static int ocfs2_cancel_convert(struct ocfs2_super *osb,
2467 struct ocfs2_lock_res *lockres)
2469 int ret;
2470 enum dlm_status status;
2472 mlog_entry_void();
2473 mlog(0, "lock %s\n", lockres->l_name);
2475 ret = 0;
2476 status = dlmunlock(osb->dlm,
2477 &lockres->l_lksb,
2478 LKM_CANCEL,
2479 lockres->l_ops->unlock_ast,
2480 lockres);
2481 if (status != DLM_NORMAL) {
2482 ocfs2_log_dlm_error("dlmunlock", status, lockres);
2483 ret = -EINVAL;
2484 ocfs2_recover_from_dlm_error(lockres, 0);
2487 mlog(0, "lock %s return from dlmunlock\n", lockres->l_name);
2489 mlog_exit(ret);
2490 return ret;
2493 static inline int ocfs2_can_downconvert_meta_lock(struct inode *inode,
2494 struct ocfs2_lock_res *lockres,
2495 int new_level)
2497 int ret;
2499 mlog_entry_void();
2501 BUG_ON(new_level != LKM_NLMODE && new_level != LKM_PRMODE);
2503 if (lockres->l_flags & OCFS2_LOCK_REFRESHING) {
2504 ret = 0;
2505 mlog(0, "lockres %s currently being refreshed -- backing "
2506 "off!\n", lockres->l_name);
2507 } else if (new_level == LKM_PRMODE)
2508 ret = !lockres->l_ex_holders &&
2509 ocfs2_inode_fully_checkpointed(inode);
2510 else /* Must be NLMODE we're converting to. */
2511 ret = !lockres->l_ro_holders && !lockres->l_ex_holders &&
2512 ocfs2_inode_fully_checkpointed(inode);
2514 mlog_exit(ret);
2515 return ret;
2518 static int ocfs2_do_unblock_meta(struct inode *inode,
2519 int *requeue)
2521 int new_level;
2522 int set_lvb = 0;
2523 int ret = 0;
2524 struct ocfs2_lock_res *lockres = &OCFS2_I(inode)->ip_meta_lockres;
2525 unsigned long flags;
2527 struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
2529 mlog_entry_void();
2531 spin_lock_irqsave(&lockres->l_lock, flags);
2533 BUG_ON(!(lockres->l_flags & OCFS2_LOCK_BLOCKED));
2535 mlog(0, "l_level=%d, l_blocking=%d\n", lockres->l_level,
2536 lockres->l_blocking);
2538 BUG_ON(lockres->l_level != LKM_EXMODE &&
2539 lockres->l_level != LKM_PRMODE);
2541 if (lockres->l_flags & OCFS2_LOCK_BUSY) {
2542 *requeue = 1;
2543 ret = ocfs2_prepare_cancel_convert(osb, lockres);
2544 spin_unlock_irqrestore(&lockres->l_lock, flags);
2545 if (ret) {
2546 ret = ocfs2_cancel_convert(osb, lockres);
2547 if (ret < 0)
2548 mlog_errno(ret);
2550 goto leave;
2553 new_level = ocfs2_highest_compat_lock_level(lockres->l_blocking);
2555 mlog(0, "l_level=%d, l_blocking=%d, new_level=%d\n",
2556 lockres->l_level, lockres->l_blocking, new_level);
2558 if (ocfs2_can_downconvert_meta_lock(inode, lockres, new_level)) {
2559 if (lockres->l_level == LKM_EXMODE)
2560 set_lvb = 1;
2562 /* If the lock hasn't been refreshed yet (rare), then
2563 * our memory inode values are old and we skip
2564 * stuffing the lvb. There's no need to actually clear
2565 * out the lvb here as it's value is still valid. */
2566 if (!(lockres->l_flags & OCFS2_LOCK_NEEDS_REFRESH)) {
2567 if (set_lvb)
2568 __ocfs2_stuff_meta_lvb(inode);
2569 } else
2570 mlog(0, "lockres %s: downconverting stale lock!\n",
2571 lockres->l_name);
2573 mlog(0, "calling ocfs2_downconvert_lock with l_level=%d, "
2574 "l_blocking=%d, new_level=%d\n",
2575 lockres->l_level, lockres->l_blocking, new_level);
2577 ocfs2_prepare_downconvert(lockres, new_level);
2578 spin_unlock_irqrestore(&lockres->l_lock, flags);
2579 ret = ocfs2_downconvert_lock(osb, lockres, new_level, set_lvb);
2580 goto leave;
2582 if (!ocfs2_inode_fully_checkpointed(inode))
2583 ocfs2_start_checkpoint(osb);
2585 *requeue = 1;
2586 spin_unlock_irqrestore(&lockres->l_lock, flags);
2587 ret = 0;
2588 leave:
2589 mlog_exit(ret);
2590 return ret;
2593 static int ocfs2_generic_unblock_lock(struct ocfs2_super *osb,
2594 struct ocfs2_lock_res *lockres,
2595 int *requeue,
2596 ocfs2_convert_worker_t *worker)
2598 unsigned long flags;
2599 int blocking;
2600 int new_level;
2601 int ret = 0;
2603 mlog_entry_void();
2605 spin_lock_irqsave(&lockres->l_lock, flags);
2607 BUG_ON(!(lockres->l_flags & OCFS2_LOCK_BLOCKED));
2609 recheck:
2610 if (lockres->l_flags & OCFS2_LOCK_BUSY) {
2611 *requeue = 1;
2612 ret = ocfs2_prepare_cancel_convert(osb, lockres);
2613 spin_unlock_irqrestore(&lockres->l_lock, flags);
2614 if (ret) {
2615 ret = ocfs2_cancel_convert(osb, lockres);
2616 if (ret < 0)
2617 mlog_errno(ret);
2619 goto leave;
2622 /* if we're blocking an exclusive and we have *any* holders,
2623 * then requeue. */
2624 if ((lockres->l_blocking == LKM_EXMODE)
2625 && (lockres->l_ex_holders || lockres->l_ro_holders)) {
2626 spin_unlock_irqrestore(&lockres->l_lock, flags);
2627 *requeue = 1;
2628 ret = 0;
2629 goto leave;
2632 /* If it's a PR we're blocking, then only
2633 * requeue if we've got any EX holders */
2634 if (lockres->l_blocking == LKM_PRMODE &&
2635 lockres->l_ex_holders) {
2636 spin_unlock_irqrestore(&lockres->l_lock, flags);
2637 *requeue = 1;
2638 ret = 0;
2639 goto leave;
2642 /* If we get here, then we know that there are no more
2643 * incompatible holders (and anyone asking for an incompatible
2644 * lock is blocked). We can now downconvert the lock */
2645 if (!worker)
2646 goto downconvert;
2648 /* Some lockres types want to do a bit of work before
2649 * downconverting a lock. Allow that here. The worker function
2650 * may sleep, so we save off a copy of what we're blocking as
2651 * it may change while we're not holding the spin lock. */
2652 blocking = lockres->l_blocking;
2653 spin_unlock_irqrestore(&lockres->l_lock, flags);
2655 worker(lockres, blocking);
2657 spin_lock_irqsave(&lockres->l_lock, flags);
2658 if (blocking != lockres->l_blocking) {
2659 /* If this changed underneath us, then we can't drop
2660 * it just yet. */
2661 goto recheck;
2664 downconvert:
2665 *requeue = 0;
2666 new_level = ocfs2_highest_compat_lock_level(lockres->l_blocking);
2668 ocfs2_prepare_downconvert(lockres, new_level);
2669 spin_unlock_irqrestore(&lockres->l_lock, flags);
2670 ret = ocfs2_downconvert_lock(osb, lockres, new_level, 0);
2671 leave:
2672 mlog_exit(ret);
2673 return ret;
2676 static void ocfs2_data_convert_worker(struct ocfs2_lock_res *lockres,
2677 int blocking)
2679 struct inode *inode;
2680 struct address_space *mapping;
2682 mlog_entry_void();
2684 inode = ocfs2_lock_res_inode(lockres);
2685 mapping = inode->i_mapping;
2687 if (filemap_fdatawrite(mapping)) {
2688 mlog(ML_ERROR, "Could not sync inode %llu for downconvert!",
2689 (unsigned long long)OCFS2_I(inode)->ip_blkno);
2691 sync_mapping_buffers(mapping);
2692 if (blocking == LKM_EXMODE) {
2693 truncate_inode_pages(mapping, 0);
2694 unmap_mapping_range(mapping, 0, 0, 0);
2695 } else {
2696 /* We only need to wait on the I/O if we're not also
2697 * truncating pages because truncate_inode_pages waits
2698 * for us above. We don't truncate pages if we're
2699 * blocking anything < EXMODE because we want to keep
2700 * them around in that case. */
2701 filemap_fdatawait(mapping);
2704 mlog_exit_void();
2707 int ocfs2_unblock_data(struct ocfs2_lock_res *lockres,
2708 int *requeue)
2710 int status;
2711 struct inode *inode;
2712 struct ocfs2_super *osb;
2714 mlog_entry_void();
2716 inode = ocfs2_lock_res_inode(lockres);
2717 osb = OCFS2_SB(inode->i_sb);
2719 mlog(0, "unblock inode %llu\n",
2720 (unsigned long long)OCFS2_I(inode)->ip_blkno);
2722 status = ocfs2_generic_unblock_lock(osb,
2723 lockres,
2724 requeue,
2725 ocfs2_data_convert_worker);
2726 if (status < 0)
2727 mlog_errno(status);
2729 mlog(0, "inode %llu, requeue = %d\n",
2730 (unsigned long long)OCFS2_I(inode)->ip_blkno, *requeue);
2732 mlog_exit(status);
2733 return status;
2736 static int ocfs2_unblock_inode_lock(struct ocfs2_lock_res *lockres,
2737 int *requeue)
2739 int status;
2740 struct inode *inode;
2742 mlog_entry_void();
2744 mlog(0, "Unblock lockres %s\n", lockres->l_name);
2746 inode = ocfs2_lock_res_inode(lockres);
2748 status = ocfs2_generic_unblock_lock(OCFS2_SB(inode->i_sb),
2749 lockres,
2750 requeue,
2751 NULL);
2752 if (status < 0)
2753 mlog_errno(status);
2755 mlog_exit(status);
2756 return status;
2760 int ocfs2_unblock_meta(struct ocfs2_lock_res *lockres,
2761 int *requeue)
2763 int status;
2764 struct inode *inode;
2766 mlog_entry_void();
2768 inode = ocfs2_lock_res_inode(lockres);
2770 mlog(0, "unblock inode %llu\n",
2771 (unsigned long long)OCFS2_I(inode)->ip_blkno);
2773 status = ocfs2_do_unblock_meta(inode, requeue);
2774 if (status < 0)
2775 mlog_errno(status);
2777 mlog(0, "inode %llu, requeue = %d\n",
2778 (unsigned long long)OCFS2_I(inode)->ip_blkno, *requeue);
2780 mlog_exit(status);
2781 return status;
2784 /* Generic unblock function for any lockres whose private data is an
2785 * ocfs2_super pointer. */
2786 static int ocfs2_unblock_osb_lock(struct ocfs2_lock_res *lockres,
2787 int *requeue)
2789 int status;
2790 struct ocfs2_super *osb;
2792 mlog_entry_void();
2794 mlog(0, "Unblock lockres %s\n", lockres->l_name);
2796 osb = ocfs2_lock_res_super(lockres);
2798 status = ocfs2_generic_unblock_lock(osb,
2799 lockres,
2800 requeue,
2801 NULL);
2802 if (status < 0)
2803 mlog_errno(status);
2805 mlog_exit(status);
2806 return status;
2809 void ocfs2_process_blocked_lock(struct ocfs2_super *osb,
2810 struct ocfs2_lock_res *lockres)
2812 int status;
2813 int requeue = 0;
2814 unsigned long flags;
2816 /* Our reference to the lockres in this function can be
2817 * considered valid until we remove the OCFS2_LOCK_QUEUED
2818 * flag. */
2820 mlog_entry_void();
2822 BUG_ON(!lockres);
2823 BUG_ON(!lockres->l_ops);
2824 BUG_ON(!lockres->l_ops->unblock);
2826 mlog(0, "lockres %s blocked.\n", lockres->l_name);
2828 /* Detect whether a lock has been marked as going away while
2829 * the vote thread was processing other things. A lock can
2830 * still be marked with OCFS2_LOCK_FREEING after this check,
2831 * but short circuiting here will still save us some
2832 * performance. */
2833 spin_lock_irqsave(&lockres->l_lock, flags);
2834 if (lockres->l_flags & OCFS2_LOCK_FREEING)
2835 goto unqueue;
2836 spin_unlock_irqrestore(&lockres->l_lock, flags);
2838 status = lockres->l_ops->unblock(lockres, &requeue);
2839 if (status < 0)
2840 mlog_errno(status);
2842 spin_lock_irqsave(&lockres->l_lock, flags);
2843 unqueue:
2844 if (lockres->l_flags & OCFS2_LOCK_FREEING || !requeue) {
2845 lockres_clear_flags(lockres, OCFS2_LOCK_QUEUED);
2846 } else
2847 ocfs2_schedule_blocked_lock(osb, lockres);
2849 mlog(0, "lockres %s, requeue = %s.\n", lockres->l_name,
2850 requeue ? "yes" : "no");
2851 spin_unlock_irqrestore(&lockres->l_lock, flags);
2853 mlog_exit_void();
2856 static void ocfs2_schedule_blocked_lock(struct ocfs2_super *osb,
2857 struct ocfs2_lock_res *lockres)
2859 mlog_entry_void();
2861 assert_spin_locked(&lockres->l_lock);
2863 if (lockres->l_flags & OCFS2_LOCK_FREEING) {
2864 /* Do not schedule a lock for downconvert when it's on
2865 * the way to destruction - any nodes wanting access
2866 * to the resource will get it soon. */
2867 mlog(0, "Lockres %s won't be scheduled: flags 0x%lx\n",
2868 lockres->l_name, lockres->l_flags);
2869 return;
2872 lockres_or_flags(lockres, OCFS2_LOCK_QUEUED);
2874 spin_lock(&osb->vote_task_lock);
2875 if (list_empty(&lockres->l_blocked_list)) {
2876 list_add_tail(&lockres->l_blocked_list,
2877 &osb->blocked_lock_list);
2878 osb->blocked_lock_count++;
2880 spin_unlock(&osb->vote_task_lock);
2882 mlog_exit_void();
2885 /* This aids in debugging situations where a bad LVB might be involved. */
2886 void ocfs2_dump_meta_lvb_info(u64 level,
2887 const char *function,
2888 unsigned int line,
2889 struct ocfs2_lock_res *lockres)
2891 struct ocfs2_meta_lvb *lvb = (struct ocfs2_meta_lvb *) lockres->l_lksb.lvb;
2893 mlog(level, "LVB information for %s (called from %s:%u):\n",
2894 lockres->l_name, function, line);
2895 mlog(level, "version: %u, clusters: %u\n",
2896 be32_to_cpu(lvb->lvb_version), be32_to_cpu(lvb->lvb_iclusters));
2897 mlog(level, "size: %llu, uid %u, gid %u, mode 0x%x\n",
2898 (unsigned long long)be64_to_cpu(lvb->lvb_isize),
2899 be32_to_cpu(lvb->lvb_iuid), be32_to_cpu(lvb->lvb_igid),
2900 be16_to_cpu(lvb->lvb_imode));
2901 mlog(level, "nlink %u, atime_packed 0x%llx, ctime_packed 0x%llx, "
2902 "mtime_packed 0x%llx\n", be16_to_cpu(lvb->lvb_inlink),
2903 (long long)be64_to_cpu(lvb->lvb_iatime_packed),
2904 (long long)be64_to_cpu(lvb->lvb_ictime_packed),
2905 (long long)be64_to_cpu(lvb->lvb_imtime_packed));