blob: e335541727f9bad9130859678b21693f7e25e067 [file] [log] [blame]
Mark Fashehccd979b2005-12-15 14:31:24 -08001/* -*- 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 */
25
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>
36
37#include <cluster/heartbeat.h>
38#include <cluster/nodemanager.h>
39#include <cluster/tcp.h>
40
41#include <dlm/dlmapi.h>
42
43#define MLOG_MASK_PREFIX ML_DLM_GLUE
44#include <cluster/masklog.h>
45
46#include "ocfs2.h"
47
48#include "alloc.h"
Mark Fashehd680efe2006-09-08 14:14:34 -070049#include "dcache.h"
Mark Fashehccd979b2005-12-15 14:31:24 -080050#include "dlmglue.h"
51#include "extent_map.h"
Tiger Yang7f1a37e2006-11-15 15:48:42 +080052#include "file.h"
Mark Fashehccd979b2005-12-15 14:31:24 -080053#include "heartbeat.h"
54#include "inode.h"
55#include "journal.h"
56#include "slot_map.h"
57#include "super.h"
58#include "uptodate.h"
59#include "vote.h"
60
61#include "buffer_head_io.h"
62
63struct ocfs2_mask_waiter {
64 struct list_head mw_item;
65 int mw_status;
66 struct completion mw_complete;
67 unsigned long mw_mask;
68 unsigned long mw_goal;
69};
70
Mark Fasheh54a7e752006-09-12 21:49:13 -070071static struct ocfs2_super *ocfs2_get_dentry_osb(struct ocfs2_lock_res *lockres);
72static struct ocfs2_super *ocfs2_get_inode_osb(struct ocfs2_lock_res *lockres);
Mark Fashehccd979b2005-12-15 14:31:24 -080073
Mark Fashehd680efe2006-09-08 14:14:34 -070074/*
Mark Fashehcc567d82006-09-13 21:52:21 -070075 * Return value from ->downconvert_worker functions.
Mark Fashehd680efe2006-09-08 14:14:34 -070076 *
Mark Fashehb5e500e2006-09-13 22:01:16 -070077 * These control the precise actions of ocfs2_unblock_lock()
Mark Fashehd680efe2006-09-08 14:14:34 -070078 * and ocfs2_process_blocked_lock()
79 *
80 */
81enum ocfs2_unblock_action {
82 UNBLOCK_CONTINUE = 0, /* Continue downconvert */
83 UNBLOCK_CONTINUE_POST = 1, /* Continue downconvert, fire
84 * ->post_unlock callback */
85 UNBLOCK_STOP_POST = 2, /* Do not downconvert, fire
86 * ->post_unlock() callback. */
87};
88
89struct ocfs2_unblock_ctl {
90 int requeue;
91 enum ocfs2_unblock_action unblock_action;
92};
93
Mark Fasheh810d5ae2006-09-13 21:39:52 -070094static int ocfs2_check_meta_downconvert(struct ocfs2_lock_res *lockres,
95 int new_level);
96static void ocfs2_set_meta_lvb(struct ocfs2_lock_res *lockres);
97
Mark Fashehcc567d82006-09-13 21:52:21 -070098static int ocfs2_data_convert_worker(struct ocfs2_lock_res *lockres,
99 int blocking);
100
Mark Fashehcc567d82006-09-13 21:52:21 -0700101static int ocfs2_dentry_convert_worker(struct ocfs2_lock_res *lockres,
102 int blocking);
Mark Fashehd680efe2006-09-08 14:14:34 -0700103
104static void ocfs2_dentry_post_unlock(struct ocfs2_super *osb,
105 struct ocfs2_lock_res *lockres);
Mark Fashehccd979b2005-12-15 14:31:24 -0800106
Mark Fashehf625c972006-09-12 21:24:53 -0700107/*
108 * OCFS2 Lock Resource Operations
109 *
110 * These fine tune the behavior of the generic dlmglue locking infrastructure.
Mark Fasheh0d5dc6c2006-09-14 14:44:51 -0700111 *
112 * The most basic of lock types can point ->l_priv to their respective
113 * struct ocfs2_super and allow the default actions to manage things.
114 *
115 * Right now, each lock type also needs to implement an init function,
116 * and trivial lock/unlock wrappers. ocfs2_simple_drop_lockres()
117 * should be called when the lock is no longer needed (i.e., object
118 * destruction time).
Mark Fashehf625c972006-09-12 21:24:53 -0700119 */
Mark Fashehccd979b2005-12-15 14:31:24 -0800120struct ocfs2_lock_res_ops {
Mark Fasheh54a7e752006-09-12 21:49:13 -0700121 /*
122 * Translate an ocfs2_lock_res * into an ocfs2_super *. Define
123 * this callback if ->l_priv is not an ocfs2_super pointer
124 */
125 struct ocfs2_super * (*get_osb)(struct ocfs2_lock_res *);
Mark Fashehb5e500e2006-09-13 22:01:16 -0700126
Mark Fasheh0d5dc6c2006-09-14 14:44:51 -0700127 /*
128 * Optionally called in the downconvert (or "vote") thread
129 * after a successful downconvert. The lockres will not be
130 * referenced after this callback is called, so it is safe to
131 * free memory, etc.
132 *
133 * The exact semantics of when this is called are controlled
134 * by ->downconvert_worker()
135 */
Mark Fashehd680efe2006-09-08 14:14:34 -0700136 void (*post_unlock)(struct ocfs2_super *, struct ocfs2_lock_res *);
Mark Fashehf625c972006-09-12 21:24:53 -0700137
138 /*
Mark Fasheh16d5b952006-09-13 21:10:12 -0700139 * Allow a lock type to add checks to determine whether it is
140 * safe to downconvert a lock. Return 0 to re-queue the
141 * downconvert at a later time, nonzero to continue.
142 *
143 * For most locks, the default checks that there are no
144 * incompatible holders are sufficient.
145 *
146 * Called with the lockres spinlock held.
147 */
148 int (*check_downconvert)(struct ocfs2_lock_res *, int);
149
150 /*
Mark Fasheh5ef0d4e2006-09-13 21:21:52 -0700151 * Allows a lock type to populate the lock value block. This
152 * is called on downconvert, and when we drop a lock.
153 *
154 * Locks that want to use this should set LOCK_TYPE_USES_LVB
155 * in the flags field.
156 *
157 * Called with the lockres spinlock held.
158 */
159 void (*set_lvb)(struct ocfs2_lock_res *);
160
161 /*
Mark Fashehcc567d82006-09-13 21:52:21 -0700162 * Called from the downconvert thread when it is determined
163 * that a lock will be downconverted. This is called without
164 * any locks held so the function can do work that might
165 * schedule (syncing out data, etc).
166 *
167 * This should return any one of the ocfs2_unblock_action
168 * values, depending on what it wants the thread to do.
169 */
170 int (*downconvert_worker)(struct ocfs2_lock_res *, int);
171
172 /*
Mark Fashehf625c972006-09-12 21:24:53 -0700173 * LOCK_TYPE_* flags which describe the specific requirements
174 * of a lock type. Descriptions of each individual flag follow.
175 */
176 int flags;
Mark Fashehccd979b2005-12-15 14:31:24 -0800177};
178
Mark Fashehf625c972006-09-12 21:24:53 -0700179/*
180 * Some locks want to "refresh" potentially stale data when a
181 * meaningful (PRMODE or EXMODE) lock level is first obtained. If this
182 * flag is set, the OCFS2_LOCK_NEEDS_REFRESH flag will be set on the
183 * individual lockres l_flags member from the ast function. It is
184 * expected that the locking wrapper will clear the
185 * OCFS2_LOCK_NEEDS_REFRESH flag when done.
186 */
187#define LOCK_TYPE_REQUIRES_REFRESH 0x1
188
Mark Fashehb80fc012006-09-12 22:08:14 -0700189/*
Mark Fasheh5ef0d4e2006-09-13 21:21:52 -0700190 * Indicate that a lock type makes use of the lock value block. The
191 * ->set_lvb lock type callback must be defined.
Mark Fashehb80fc012006-09-12 22:08:14 -0700192 */
193#define LOCK_TYPE_USES_LVB 0x2
194
Mark Fashehccd979b2005-12-15 14:31:24 -0800195static struct ocfs2_lock_res_ops ocfs2_inode_rw_lops = {
Mark Fasheh54a7e752006-09-12 21:49:13 -0700196 .get_osb = ocfs2_get_inode_osb,
Mark Fashehf625c972006-09-12 21:24:53 -0700197 .flags = 0,
Mark Fashehccd979b2005-12-15 14:31:24 -0800198};
199
200static struct ocfs2_lock_res_ops ocfs2_inode_meta_lops = {
Mark Fasheh54a7e752006-09-12 21:49:13 -0700201 .get_osb = ocfs2_get_inode_osb,
Mark Fasheh810d5ae2006-09-13 21:39:52 -0700202 .check_downconvert = ocfs2_check_meta_downconvert,
203 .set_lvb = ocfs2_set_meta_lvb,
Mark Fashehb80fc012006-09-12 22:08:14 -0700204 .flags = LOCK_TYPE_REQUIRES_REFRESH|LOCK_TYPE_USES_LVB,
Mark Fashehccd979b2005-12-15 14:31:24 -0800205};
206
Mark Fashehccd979b2005-12-15 14:31:24 -0800207static struct ocfs2_lock_res_ops ocfs2_inode_data_lops = {
Mark Fasheh54a7e752006-09-12 21:49:13 -0700208 .get_osb = ocfs2_get_inode_osb,
Mark Fashehcc567d82006-09-13 21:52:21 -0700209 .downconvert_worker = ocfs2_data_convert_worker,
Mark Fashehf625c972006-09-12 21:24:53 -0700210 .flags = 0,
Mark Fashehccd979b2005-12-15 14:31:24 -0800211};
212
213static struct ocfs2_lock_res_ops ocfs2_super_lops = {
Mark Fashehf625c972006-09-12 21:24:53 -0700214 .flags = LOCK_TYPE_REQUIRES_REFRESH,
Mark Fashehccd979b2005-12-15 14:31:24 -0800215};
216
217static struct ocfs2_lock_res_ops ocfs2_rename_lops = {
Mark Fashehf625c972006-09-12 21:24:53 -0700218 .flags = 0,
Mark Fashehccd979b2005-12-15 14:31:24 -0800219};
220
Mark Fashehd680efe2006-09-08 14:14:34 -0700221static struct ocfs2_lock_res_ops ocfs2_dentry_lops = {
Mark Fasheh54a7e752006-09-12 21:49:13 -0700222 .get_osb = ocfs2_get_dentry_osb,
Mark Fashehd680efe2006-09-08 14:14:34 -0700223 .post_unlock = ocfs2_dentry_post_unlock,
Mark Fashehcc567d82006-09-13 21:52:21 -0700224 .downconvert_worker = ocfs2_dentry_convert_worker,
Mark Fashehf625c972006-09-12 21:24:53 -0700225 .flags = 0,
Mark Fashehd680efe2006-09-08 14:14:34 -0700226};
227
Mark Fashehccd979b2005-12-15 14:31:24 -0800228static inline int ocfs2_is_inode_lock(struct ocfs2_lock_res *lockres)
229{
230 return lockres->l_type == OCFS2_LOCK_TYPE_META ||
231 lockres->l_type == OCFS2_LOCK_TYPE_DATA ||
232 lockres->l_type == OCFS2_LOCK_TYPE_RW;
233}
234
Mark Fashehccd979b2005-12-15 14:31:24 -0800235static inline struct inode *ocfs2_lock_res_inode(struct ocfs2_lock_res *lockres)
236{
237 BUG_ON(!ocfs2_is_inode_lock(lockres));
238
239 return (struct inode *) lockres->l_priv;
240}
241
Mark Fashehd680efe2006-09-08 14:14:34 -0700242static inline struct ocfs2_dentry_lock *ocfs2_lock_res_dl(struct ocfs2_lock_res *lockres)
243{
244 BUG_ON(lockres->l_type != OCFS2_LOCK_TYPE_DENTRY);
245
246 return (struct ocfs2_dentry_lock *)lockres->l_priv;
247}
248
Mark Fasheh54a7e752006-09-12 21:49:13 -0700249static inline struct ocfs2_super *ocfs2_get_lockres_osb(struct ocfs2_lock_res *lockres)
250{
251 if (lockres->l_ops->get_osb)
252 return lockres->l_ops->get_osb(lockres);
253
254 return (struct ocfs2_super *)lockres->l_priv;
255}
256
Mark Fashehccd979b2005-12-15 14:31:24 -0800257static int ocfs2_lock_create(struct ocfs2_super *osb,
258 struct ocfs2_lock_res *lockres,
259 int level,
260 int dlm_flags);
261static inline int ocfs2_may_continue_on_blocked_lock(struct ocfs2_lock_res *lockres,
262 int wanted);
263static void ocfs2_cluster_unlock(struct ocfs2_super *osb,
264 struct ocfs2_lock_res *lockres,
265 int level);
266static inline void ocfs2_generic_handle_downconvert_action(struct ocfs2_lock_res *lockres);
267static inline void ocfs2_generic_handle_convert_action(struct ocfs2_lock_res *lockres);
268static inline void ocfs2_generic_handle_attach_action(struct ocfs2_lock_res *lockres);
269static int ocfs2_generic_handle_bast(struct ocfs2_lock_res *lockres, int level);
270static void ocfs2_schedule_blocked_lock(struct ocfs2_super *osb,
271 struct ocfs2_lock_res *lockres);
272static inline void ocfs2_recover_from_dlm_error(struct ocfs2_lock_res *lockres,
273 int convert);
274#define ocfs2_log_dlm_error(_func, _stat, _lockres) do { \
275 mlog(ML_ERROR, "Dlm error \"%s\" while calling %s on " \
276 "resource %s: %s\n", dlm_errname(_stat), _func, \
277 _lockres->l_name, dlm_errmsg(_stat)); \
278} while (0)
279static void ocfs2_vote_on_unlock(struct ocfs2_super *osb,
280 struct ocfs2_lock_res *lockres);
281static int ocfs2_meta_lock_update(struct inode *inode,
282 struct buffer_head **bh);
283static void ocfs2_drop_osb_locks(struct ocfs2_super *osb);
284static inline int ocfs2_highest_compat_lock_level(int level);
Mark Fashehccd979b2005-12-15 14:31:24 -0800285
Mark Fashehccd979b2005-12-15 14:31:24 -0800286static void ocfs2_build_lock_name(enum ocfs2_lock_type type,
287 u64 blkno,
288 u32 generation,
289 char *name)
290{
291 int len;
292
293 mlog_entry_void();
294
295 BUG_ON(type >= OCFS2_NUM_LOCK_TYPES);
296
Mark Fashehb06970532006-03-03 10:24:33 -0800297 len = snprintf(name, OCFS2_LOCK_ID_MAX_LEN, "%c%s%016llx%08x",
298 ocfs2_lock_type_char(type), OCFS2_LOCK_ID_PAD,
299 (long long)blkno, generation);
Mark Fashehccd979b2005-12-15 14:31:24 -0800300
301 BUG_ON(len != (OCFS2_LOCK_ID_MAX_LEN - 1));
302
303 mlog(0, "built lock resource with name: %s\n", name);
304
305 mlog_exit_void();
306}
307
Ingo Molnar34af9462006-06-27 02:53:55 -0700308static DEFINE_SPINLOCK(ocfs2_dlm_tracking_lock);
Mark Fashehccd979b2005-12-15 14:31:24 -0800309
310static void ocfs2_add_lockres_tracking(struct ocfs2_lock_res *res,
311 struct ocfs2_dlm_debug *dlm_debug)
312{
313 mlog(0, "Add tracking for lockres %s\n", res->l_name);
314
315 spin_lock(&ocfs2_dlm_tracking_lock);
316 list_add(&res->l_debug_list, &dlm_debug->d_lockres_tracking);
317 spin_unlock(&ocfs2_dlm_tracking_lock);
318}
319
320static void ocfs2_remove_lockres_tracking(struct ocfs2_lock_res *res)
321{
322 spin_lock(&ocfs2_dlm_tracking_lock);
323 if (!list_empty(&res->l_debug_list))
324 list_del_init(&res->l_debug_list);
325 spin_unlock(&ocfs2_dlm_tracking_lock);
326}
327
328static void ocfs2_lock_res_init_common(struct ocfs2_super *osb,
329 struct ocfs2_lock_res *res,
330 enum ocfs2_lock_type type,
Mark Fashehccd979b2005-12-15 14:31:24 -0800331 struct ocfs2_lock_res_ops *ops,
332 void *priv)
333{
Mark Fashehccd979b2005-12-15 14:31:24 -0800334 res->l_type = type;
335 res->l_ops = ops;
336 res->l_priv = priv;
337
338 res->l_level = LKM_IVMODE;
339 res->l_requested = LKM_IVMODE;
340 res->l_blocking = LKM_IVMODE;
341 res->l_action = OCFS2_AST_INVALID;
342 res->l_unlock_action = OCFS2_UNLOCK_INVALID;
343
344 res->l_flags = OCFS2_LOCK_INITIALIZED;
345
346 ocfs2_add_lockres_tracking(res, osb->osb_dlm_debug);
347}
348
349void ocfs2_lock_res_init_once(struct ocfs2_lock_res *res)
350{
351 /* This also clears out the lock status block */
352 memset(res, 0, sizeof(struct ocfs2_lock_res));
353 spin_lock_init(&res->l_lock);
354 init_waitqueue_head(&res->l_event);
355 INIT_LIST_HEAD(&res->l_blocked_list);
356 INIT_LIST_HEAD(&res->l_mask_waiters);
357}
358
359void ocfs2_inode_lock_res_init(struct ocfs2_lock_res *res,
360 enum ocfs2_lock_type type,
Mark Fasheh24c19ef2006-09-22 17:28:19 -0700361 unsigned int generation,
Mark Fashehccd979b2005-12-15 14:31:24 -0800362 struct inode *inode)
363{
364 struct ocfs2_lock_res_ops *ops;
365
366 switch(type) {
367 case OCFS2_LOCK_TYPE_RW:
368 ops = &ocfs2_inode_rw_lops;
369 break;
370 case OCFS2_LOCK_TYPE_META:
371 ops = &ocfs2_inode_meta_lops;
372 break;
373 case OCFS2_LOCK_TYPE_DATA:
374 ops = &ocfs2_inode_data_lops;
375 break;
376 default:
377 mlog_bug_on_msg(1, "type: %d\n", type);
378 ops = NULL; /* thanks, gcc */
379 break;
380 };
381
Mark Fashehd680efe2006-09-08 14:14:34 -0700382 ocfs2_build_lock_name(type, OCFS2_I(inode)->ip_blkno,
Mark Fasheh24c19ef2006-09-22 17:28:19 -0700383 generation, res->l_name);
Mark Fashehd680efe2006-09-08 14:14:34 -0700384 ocfs2_lock_res_init_common(OCFS2_SB(inode->i_sb), res, type, ops, inode);
385}
386
Mark Fasheh54a7e752006-09-12 21:49:13 -0700387static struct ocfs2_super *ocfs2_get_inode_osb(struct ocfs2_lock_res *lockres)
388{
389 struct inode *inode = ocfs2_lock_res_inode(lockres);
390
391 return OCFS2_SB(inode->i_sb);
392}
393
Mark Fashehd680efe2006-09-08 14:14:34 -0700394static __u64 ocfs2_get_dentry_lock_ino(struct ocfs2_lock_res *lockres)
395{
396 __be64 inode_blkno_be;
397
398 memcpy(&inode_blkno_be, &lockres->l_name[OCFS2_DENTRY_LOCK_INO_START],
399 sizeof(__be64));
400
401 return be64_to_cpu(inode_blkno_be);
402}
403
Mark Fasheh54a7e752006-09-12 21:49:13 -0700404static struct ocfs2_super *ocfs2_get_dentry_osb(struct ocfs2_lock_res *lockres)
405{
406 struct ocfs2_dentry_lock *dl = lockres->l_priv;
407
408 return OCFS2_SB(dl->dl_inode->i_sb);
409}
410
Mark Fashehd680efe2006-09-08 14:14:34 -0700411void ocfs2_dentry_lock_res_init(struct ocfs2_dentry_lock *dl,
412 u64 parent, struct inode *inode)
413{
414 int len;
415 u64 inode_blkno = OCFS2_I(inode)->ip_blkno;
416 __be64 inode_blkno_be = cpu_to_be64(inode_blkno);
417 struct ocfs2_lock_res *lockres = &dl->dl_lockres;
418
419 ocfs2_lock_res_init_once(lockres);
420
421 /*
422 * Unfortunately, the standard lock naming scheme won't work
423 * here because we have two 16 byte values to use. Instead,
424 * we'll stuff the inode number as a binary value. We still
425 * want error prints to show something without garbling the
426 * display, so drop a null byte in there before the inode
427 * number. A future version of OCFS2 will likely use all
428 * binary lock names. The stringified names have been a
429 * tremendous aid in debugging, but now that the debugfs
430 * interface exists, we can mangle things there if need be.
431 *
432 * NOTE: We also drop the standard "pad" value (the total lock
433 * name size stays the same though - the last part is all
434 * zeros due to the memset in ocfs2_lock_res_init_once()
435 */
436 len = snprintf(lockres->l_name, OCFS2_DENTRY_LOCK_INO_START,
437 "%c%016llx",
438 ocfs2_lock_type_char(OCFS2_LOCK_TYPE_DENTRY),
439 (long long)parent);
440
441 BUG_ON(len != (OCFS2_DENTRY_LOCK_INO_START - 1));
442
443 memcpy(&lockres->l_name[OCFS2_DENTRY_LOCK_INO_START], &inode_blkno_be,
444 sizeof(__be64));
445
446 ocfs2_lock_res_init_common(OCFS2_SB(inode->i_sb), lockres,
447 OCFS2_LOCK_TYPE_DENTRY, &ocfs2_dentry_lops,
448 dl);
Mark Fashehccd979b2005-12-15 14:31:24 -0800449}
450
451static void ocfs2_super_lock_res_init(struct ocfs2_lock_res *res,
452 struct ocfs2_super *osb)
453{
454 /* Superblock lockres doesn't come from a slab so we call init
455 * once on it manually. */
456 ocfs2_lock_res_init_once(res);
Mark Fashehd680efe2006-09-08 14:14:34 -0700457 ocfs2_build_lock_name(OCFS2_LOCK_TYPE_SUPER, OCFS2_SUPER_BLOCK_BLKNO,
458 0, res->l_name);
Mark Fashehccd979b2005-12-15 14:31:24 -0800459 ocfs2_lock_res_init_common(osb, res, OCFS2_LOCK_TYPE_SUPER,
Mark Fashehccd979b2005-12-15 14:31:24 -0800460 &ocfs2_super_lops, osb);
461}
462
463static void ocfs2_rename_lock_res_init(struct ocfs2_lock_res *res,
464 struct ocfs2_super *osb)
465{
466 /* Rename lockres doesn't come from a slab so we call init
467 * once on it manually. */
468 ocfs2_lock_res_init_once(res);
Mark Fashehd680efe2006-09-08 14:14:34 -0700469 ocfs2_build_lock_name(OCFS2_LOCK_TYPE_RENAME, 0, 0, res->l_name);
470 ocfs2_lock_res_init_common(osb, res, OCFS2_LOCK_TYPE_RENAME,
Mark Fashehccd979b2005-12-15 14:31:24 -0800471 &ocfs2_rename_lops, osb);
472}
473
474void ocfs2_lock_res_free(struct ocfs2_lock_res *res)
475{
476 mlog_entry_void();
477
478 if (!(res->l_flags & OCFS2_LOCK_INITIALIZED))
479 return;
480
481 ocfs2_remove_lockres_tracking(res);
482
483 mlog_bug_on_msg(!list_empty(&res->l_blocked_list),
484 "Lockres %s is on the blocked list\n",
485 res->l_name);
486 mlog_bug_on_msg(!list_empty(&res->l_mask_waiters),
487 "Lockres %s has mask waiters pending\n",
488 res->l_name);
489 mlog_bug_on_msg(spin_is_locked(&res->l_lock),
490 "Lockres %s is locked\n",
491 res->l_name);
492 mlog_bug_on_msg(res->l_ro_holders,
493 "Lockres %s has %u ro holders\n",
494 res->l_name, res->l_ro_holders);
495 mlog_bug_on_msg(res->l_ex_holders,
496 "Lockres %s has %u ex holders\n",
497 res->l_name, res->l_ex_holders);
498
499 /* Need to clear out the lock status block for the dlm */
500 memset(&res->l_lksb, 0, sizeof(res->l_lksb));
501
502 res->l_flags = 0UL;
503 mlog_exit_void();
504}
505
506static inline void ocfs2_inc_holders(struct ocfs2_lock_res *lockres,
507 int level)
508{
509 mlog_entry_void();
510
511 BUG_ON(!lockres);
512
513 switch(level) {
514 case LKM_EXMODE:
515 lockres->l_ex_holders++;
516 break;
517 case LKM_PRMODE:
518 lockres->l_ro_holders++;
519 break;
520 default:
521 BUG();
522 }
523
524 mlog_exit_void();
525}
526
527static inline void ocfs2_dec_holders(struct ocfs2_lock_res *lockres,
528 int level)
529{
530 mlog_entry_void();
531
532 BUG_ON(!lockres);
533
534 switch(level) {
535 case LKM_EXMODE:
536 BUG_ON(!lockres->l_ex_holders);
537 lockres->l_ex_holders--;
538 break;
539 case LKM_PRMODE:
540 BUG_ON(!lockres->l_ro_holders);
541 lockres->l_ro_holders--;
542 break;
543 default:
544 BUG();
545 }
546 mlog_exit_void();
547}
548
549/* WARNING: This function lives in a world where the only three lock
550 * levels are EX, PR, and NL. It *will* have to be adjusted when more
551 * lock types are added. */
552static inline int ocfs2_highest_compat_lock_level(int level)
553{
554 int new_level = LKM_EXMODE;
555
556 if (level == LKM_EXMODE)
557 new_level = LKM_NLMODE;
558 else if (level == LKM_PRMODE)
559 new_level = LKM_PRMODE;
560 return new_level;
561}
562
563static void lockres_set_flags(struct ocfs2_lock_res *lockres,
564 unsigned long newflags)
565{
566 struct list_head *pos, *tmp;
567 struct ocfs2_mask_waiter *mw;
568
569 assert_spin_locked(&lockres->l_lock);
570
571 lockres->l_flags = newflags;
572
573 list_for_each_safe(pos, tmp, &lockres->l_mask_waiters) {
574 mw = list_entry(pos, struct ocfs2_mask_waiter, mw_item);
575 if ((lockres->l_flags & mw->mw_mask) != mw->mw_goal)
576 continue;
577
578 list_del_init(&mw->mw_item);
579 mw->mw_status = 0;
580 complete(&mw->mw_complete);
581 }
582}
583static void lockres_or_flags(struct ocfs2_lock_res *lockres, unsigned long or)
584{
585 lockres_set_flags(lockres, lockres->l_flags | or);
586}
587static void lockres_clear_flags(struct ocfs2_lock_res *lockres,
588 unsigned long clear)
589{
590 lockres_set_flags(lockres, lockres->l_flags & ~clear);
591}
592
593static inline void ocfs2_generic_handle_downconvert_action(struct ocfs2_lock_res *lockres)
594{
595 mlog_entry_void();
596
597 BUG_ON(!(lockres->l_flags & OCFS2_LOCK_BUSY));
598 BUG_ON(!(lockres->l_flags & OCFS2_LOCK_ATTACHED));
599 BUG_ON(!(lockres->l_flags & OCFS2_LOCK_BLOCKED));
600 BUG_ON(lockres->l_blocking <= LKM_NLMODE);
601
602 lockres->l_level = lockres->l_requested;
603 if (lockres->l_level <=
604 ocfs2_highest_compat_lock_level(lockres->l_blocking)) {
605 lockres->l_blocking = LKM_NLMODE;
606 lockres_clear_flags(lockres, OCFS2_LOCK_BLOCKED);
607 }
608 lockres_clear_flags(lockres, OCFS2_LOCK_BUSY);
609
610 mlog_exit_void();
611}
612
613static inline void ocfs2_generic_handle_convert_action(struct ocfs2_lock_res *lockres)
614{
615 mlog_entry_void();
616
617 BUG_ON(!(lockres->l_flags & OCFS2_LOCK_BUSY));
618 BUG_ON(!(lockres->l_flags & OCFS2_LOCK_ATTACHED));
619
620 /* Convert from RO to EX doesn't really need anything as our
621 * information is already up to data. Convert from NL to
622 * *anything* however should mark ourselves as needing an
623 * update */
Mark Fashehf625c972006-09-12 21:24:53 -0700624 if (lockres->l_level == LKM_NLMODE &&
625 lockres->l_ops->flags & LOCK_TYPE_REQUIRES_REFRESH)
Mark Fashehccd979b2005-12-15 14:31:24 -0800626 lockres_or_flags(lockres, OCFS2_LOCK_NEEDS_REFRESH);
627
628 lockres->l_level = lockres->l_requested;
629 lockres_clear_flags(lockres, OCFS2_LOCK_BUSY);
630
631 mlog_exit_void();
632}
633
634static inline void ocfs2_generic_handle_attach_action(struct ocfs2_lock_res *lockres)
635{
636 mlog_entry_void();
637
638 BUG_ON((!lockres->l_flags & OCFS2_LOCK_BUSY));
639 BUG_ON(lockres->l_flags & OCFS2_LOCK_ATTACHED);
640
641 if (lockres->l_requested > LKM_NLMODE &&
Mark Fashehf625c972006-09-12 21:24:53 -0700642 !(lockres->l_flags & OCFS2_LOCK_LOCAL) &&
643 lockres->l_ops->flags & LOCK_TYPE_REQUIRES_REFRESH)
Mark Fashehccd979b2005-12-15 14:31:24 -0800644 lockres_or_flags(lockres, OCFS2_LOCK_NEEDS_REFRESH);
645
646 lockres->l_level = lockres->l_requested;
647 lockres_or_flags(lockres, OCFS2_LOCK_ATTACHED);
648 lockres_clear_flags(lockres, OCFS2_LOCK_BUSY);
649
650 mlog_exit_void();
651}
652
Mark Fashehccd979b2005-12-15 14:31:24 -0800653static int ocfs2_generic_handle_bast(struct ocfs2_lock_res *lockres,
654 int level)
655{
656 int needs_downconvert = 0;
657 mlog_entry_void();
658
659 assert_spin_locked(&lockres->l_lock);
660
661 lockres_or_flags(lockres, OCFS2_LOCK_BLOCKED);
662
663 if (level > lockres->l_blocking) {
664 /* only schedule a downconvert if we haven't already scheduled
665 * one that goes low enough to satisfy the level we're
666 * blocking. this also catches the case where we get
667 * duplicate BASTs */
668 if (ocfs2_highest_compat_lock_level(level) <
669 ocfs2_highest_compat_lock_level(lockres->l_blocking))
670 needs_downconvert = 1;
671
672 lockres->l_blocking = level;
673 }
674
675 mlog_exit(needs_downconvert);
676 return needs_downconvert;
677}
678
Mark Fashehaa2623a2006-09-12 21:58:23 -0700679static void ocfs2_blocking_ast(void *opaque, int level)
Mark Fashehccd979b2005-12-15 14:31:24 -0800680{
Mark Fashehaa2623a2006-09-12 21:58:23 -0700681 struct ocfs2_lock_res *lockres = opaque;
682 struct ocfs2_super *osb = ocfs2_get_lockres_osb(lockres);
Mark Fashehccd979b2005-12-15 14:31:24 -0800683 int needs_downconvert;
684 unsigned long flags;
685
Mark Fashehccd979b2005-12-15 14:31:24 -0800686 BUG_ON(level <= LKM_NLMODE);
687
Mark Fashehaa2623a2006-09-12 21:58:23 -0700688 mlog(0, "BAST fired for lockres %s, blocking %d, level %d type %s\n",
689 lockres->l_name, level, lockres->l_level,
690 ocfs2_lock_type_string(lockres->l_type));
691
Mark Fashehccd979b2005-12-15 14:31:24 -0800692 spin_lock_irqsave(&lockres->l_lock, flags);
693 needs_downconvert = ocfs2_generic_handle_bast(lockres, level);
694 if (needs_downconvert)
695 ocfs2_schedule_blocked_lock(osb, lockres);
696 spin_unlock_irqrestore(&lockres->l_lock, flags);
697
Mark Fashehd680efe2006-09-08 14:14:34 -0700698 wake_up(&lockres->l_event);
699
Mark Fashehccd979b2005-12-15 14:31:24 -0800700 ocfs2_kick_vote_thread(osb);
Mark Fashehccd979b2005-12-15 14:31:24 -0800701}
702
Mark Fashehe92d57d2006-09-12 21:34:35 -0700703static void ocfs2_locking_ast(void *opaque)
Mark Fashehccd979b2005-12-15 14:31:24 -0800704{
Mark Fashehe92d57d2006-09-12 21:34:35 -0700705 struct ocfs2_lock_res *lockres = opaque;
Mark Fashehccd979b2005-12-15 14:31:24 -0800706 struct dlm_lockstatus *lksb = &lockres->l_lksb;
707 unsigned long flags;
708
709 spin_lock_irqsave(&lockres->l_lock, flags);
710
711 if (lksb->status != DLM_NORMAL) {
712 mlog(ML_ERROR, "lockres %s: lksb status value of %u!\n",
713 lockres->l_name, lksb->status);
714 spin_unlock_irqrestore(&lockres->l_lock, flags);
715 return;
716 }
717
718 switch(lockres->l_action) {
719 case OCFS2_AST_ATTACH:
720 ocfs2_generic_handle_attach_action(lockres);
Mark Fashehe92d57d2006-09-12 21:34:35 -0700721 lockres_clear_flags(lockres, OCFS2_LOCK_LOCAL);
Mark Fashehccd979b2005-12-15 14:31:24 -0800722 break;
723 case OCFS2_AST_CONVERT:
724 ocfs2_generic_handle_convert_action(lockres);
725 break;
726 case OCFS2_AST_DOWNCONVERT:
727 ocfs2_generic_handle_downconvert_action(lockres);
728 break;
729 default:
Mark Fashehe92d57d2006-09-12 21:34:35 -0700730 mlog(ML_ERROR, "lockres %s: ast fired with invalid action: %u "
731 "lockres flags = 0x%lx, unlock action: %u\n",
732 lockres->l_name, lockres->l_action, lockres->l_flags,
733 lockres->l_unlock_action);
Mark Fashehccd979b2005-12-15 14:31:24 -0800734 BUG();
735 }
736
Mark Fashehccd979b2005-12-15 14:31:24 -0800737 /* set it to something invalid so if we get called again we
738 * can catch it. */
739 lockres->l_action = OCFS2_AST_INVALID;
Mark Fashehccd979b2005-12-15 14:31:24 -0800740
741 wake_up(&lockres->l_event);
Mark Fashehd680efe2006-09-08 14:14:34 -0700742 spin_unlock_irqrestore(&lockres->l_lock, flags);
Mark Fashehccd979b2005-12-15 14:31:24 -0800743}
744
Mark Fashehccd979b2005-12-15 14:31:24 -0800745static inline void ocfs2_recover_from_dlm_error(struct ocfs2_lock_res *lockres,
746 int convert)
747{
748 unsigned long flags;
749
750 mlog_entry_void();
751 spin_lock_irqsave(&lockres->l_lock, flags);
752 lockres_clear_flags(lockres, OCFS2_LOCK_BUSY);
753 if (convert)
754 lockres->l_action = OCFS2_AST_INVALID;
755 else
756 lockres->l_unlock_action = OCFS2_UNLOCK_INVALID;
757 spin_unlock_irqrestore(&lockres->l_lock, flags);
758
759 wake_up(&lockres->l_event);
760 mlog_exit_void();
761}
762
763/* Note: If we detect another process working on the lock (i.e.,
764 * OCFS2_LOCK_BUSY), we'll bail out returning 0. It's up to the caller
765 * to do the right thing in that case.
766 */
767static int ocfs2_lock_create(struct ocfs2_super *osb,
768 struct ocfs2_lock_res *lockres,
769 int level,
770 int dlm_flags)
771{
772 int ret = 0;
Sunil Mushranc271c5c2006-12-05 17:56:35 -0800773 enum dlm_status status = DLM_NORMAL;
Mark Fashehccd979b2005-12-15 14:31:24 -0800774 unsigned long flags;
775
776 mlog_entry_void();
777
778 mlog(0, "lock %s, level = %d, flags = %d\n", lockres->l_name, level,
779 dlm_flags);
780
781 spin_lock_irqsave(&lockres->l_lock, flags);
782 if ((lockres->l_flags & OCFS2_LOCK_ATTACHED) ||
783 (lockres->l_flags & OCFS2_LOCK_BUSY)) {
784 spin_unlock_irqrestore(&lockres->l_lock, flags);
785 goto bail;
786 }
787
788 lockres->l_action = OCFS2_AST_ATTACH;
789 lockres->l_requested = level;
790 lockres_or_flags(lockres, OCFS2_LOCK_BUSY);
791 spin_unlock_irqrestore(&lockres->l_lock, flags);
792
793 status = dlmlock(osb->dlm,
794 level,
795 &lockres->l_lksb,
796 dlm_flags,
797 lockres->l_name,
Mark Fashehf0681062006-09-08 11:40:10 -0700798 OCFS2_LOCK_ID_MAX_LEN - 1,
Mark Fashehe92d57d2006-09-12 21:34:35 -0700799 ocfs2_locking_ast,
Mark Fashehccd979b2005-12-15 14:31:24 -0800800 lockres,
Mark Fashehaa2623a2006-09-12 21:58:23 -0700801 ocfs2_blocking_ast);
Mark Fashehccd979b2005-12-15 14:31:24 -0800802 if (status != DLM_NORMAL) {
803 ocfs2_log_dlm_error("dlmlock", status, lockres);
804 ret = -EINVAL;
805 ocfs2_recover_from_dlm_error(lockres, 1);
806 }
807
808 mlog(0, "lock %s, successfull return from dlmlock\n", lockres->l_name);
809
810bail:
811 mlog_exit(ret);
812 return ret;
813}
814
815static inline int ocfs2_check_wait_flag(struct ocfs2_lock_res *lockres,
816 int flag)
817{
818 unsigned long flags;
819 int ret;
820
821 spin_lock_irqsave(&lockres->l_lock, flags);
822 ret = lockres->l_flags & flag;
823 spin_unlock_irqrestore(&lockres->l_lock, flags);
824
825 return ret;
826}
827
828static inline void ocfs2_wait_on_busy_lock(struct ocfs2_lock_res *lockres)
829
830{
831 wait_event(lockres->l_event,
832 !ocfs2_check_wait_flag(lockres, OCFS2_LOCK_BUSY));
833}
834
835static inline void ocfs2_wait_on_refreshing_lock(struct ocfs2_lock_res *lockres)
836
837{
838 wait_event(lockres->l_event,
839 !ocfs2_check_wait_flag(lockres, OCFS2_LOCK_REFRESHING));
840}
841
842/* predict what lock level we'll be dropping down to on behalf
843 * of another node, and return true if the currently wanted
844 * level will be compatible with it. */
845static inline int ocfs2_may_continue_on_blocked_lock(struct ocfs2_lock_res *lockres,
846 int wanted)
847{
848 BUG_ON(!(lockres->l_flags & OCFS2_LOCK_BLOCKED));
849
850 return wanted <= ocfs2_highest_compat_lock_level(lockres->l_blocking);
851}
852
853static void ocfs2_init_mask_waiter(struct ocfs2_mask_waiter *mw)
854{
855 INIT_LIST_HEAD(&mw->mw_item);
856 init_completion(&mw->mw_complete);
857}
858
859static int ocfs2_wait_for_mask(struct ocfs2_mask_waiter *mw)
860{
861 wait_for_completion(&mw->mw_complete);
862 /* Re-arm the completion in case we want to wait on it again */
863 INIT_COMPLETION(mw->mw_complete);
864 return mw->mw_status;
865}
866
867static void lockres_add_mask_waiter(struct ocfs2_lock_res *lockres,
868 struct ocfs2_mask_waiter *mw,
869 unsigned long mask,
870 unsigned long goal)
871{
872 BUG_ON(!list_empty(&mw->mw_item));
873
874 assert_spin_locked(&lockres->l_lock);
875
876 list_add_tail(&mw->mw_item, &lockres->l_mask_waiters);
877 mw->mw_mask = mask;
878 mw->mw_goal = goal;
879}
880
881/* returns 0 if the mw that was removed was already satisfied, -EBUSY
882 * if the mask still hadn't reached its goal */
883static int lockres_remove_mask_waiter(struct ocfs2_lock_res *lockres,
884 struct ocfs2_mask_waiter *mw)
885{
886 unsigned long flags;
887 int ret = 0;
888
889 spin_lock_irqsave(&lockres->l_lock, flags);
890 if (!list_empty(&mw->mw_item)) {
891 if ((lockres->l_flags & mw->mw_mask) != mw->mw_goal)
892 ret = -EBUSY;
893
894 list_del_init(&mw->mw_item);
895 init_completion(&mw->mw_complete);
896 }
897 spin_unlock_irqrestore(&lockres->l_lock, flags);
898
899 return ret;
900
901}
902
903static int ocfs2_cluster_lock(struct ocfs2_super *osb,
904 struct ocfs2_lock_res *lockres,
905 int level,
906 int lkm_flags,
907 int arg_flags)
908{
909 struct ocfs2_mask_waiter mw;
910 enum dlm_status status;
911 int wait, catch_signals = !(osb->s_mount_opt & OCFS2_MOUNT_NOINTR);
912 int ret = 0; /* gcc doesn't realize wait = 1 guarantees ret is set */
913 unsigned long flags;
914
915 mlog_entry_void();
916
917 ocfs2_init_mask_waiter(&mw);
918
Mark Fashehb80fc012006-09-12 22:08:14 -0700919 if (lockres->l_ops->flags & LOCK_TYPE_USES_LVB)
920 lkm_flags |= LKM_VALBLK;
921
Mark Fashehccd979b2005-12-15 14:31:24 -0800922again:
923 wait = 0;
924
925 if (catch_signals && signal_pending(current)) {
926 ret = -ERESTARTSYS;
927 goto out;
928 }
929
930 spin_lock_irqsave(&lockres->l_lock, flags);
931
932 mlog_bug_on_msg(lockres->l_flags & OCFS2_LOCK_FREEING,
933 "Cluster lock called on freeing lockres %s! flags "
934 "0x%lx\n", lockres->l_name, lockres->l_flags);
935
936 /* We only compare against the currently granted level
937 * here. If the lock is blocked waiting on a downconvert,
938 * we'll get caught below. */
939 if (lockres->l_flags & OCFS2_LOCK_BUSY &&
940 level > lockres->l_level) {
941 /* is someone sitting in dlm_lock? If so, wait on
942 * them. */
943 lockres_add_mask_waiter(lockres, &mw, OCFS2_LOCK_BUSY, 0);
944 wait = 1;
945 goto unlock;
946 }
947
948 if (!(lockres->l_flags & OCFS2_LOCK_ATTACHED)) {
949 /* lock has not been created yet. */
950 spin_unlock_irqrestore(&lockres->l_lock, flags);
951
952 ret = ocfs2_lock_create(osb, lockres, LKM_NLMODE, 0);
953 if (ret < 0) {
954 mlog_errno(ret);
955 goto out;
956 }
957 goto again;
958 }
959
960 if (lockres->l_flags & OCFS2_LOCK_BLOCKED &&
961 !ocfs2_may_continue_on_blocked_lock(lockres, level)) {
962 /* is the lock is currently blocked on behalf of
963 * another node */
964 lockres_add_mask_waiter(lockres, &mw, OCFS2_LOCK_BLOCKED, 0);
965 wait = 1;
966 goto unlock;
967 }
968
969 if (level > lockres->l_level) {
970 if (lockres->l_action != OCFS2_AST_INVALID)
971 mlog(ML_ERROR, "lockres %s has action %u pending\n",
972 lockres->l_name, lockres->l_action);
973
974 lockres->l_action = OCFS2_AST_CONVERT;
975 lockres->l_requested = level;
976 lockres_or_flags(lockres, OCFS2_LOCK_BUSY);
977 spin_unlock_irqrestore(&lockres->l_lock, flags);
978
979 BUG_ON(level == LKM_IVMODE);
980 BUG_ON(level == LKM_NLMODE);
981
982 mlog(0, "lock %s, convert from %d to level = %d\n",
983 lockres->l_name, lockres->l_level, level);
984
985 /* call dlm_lock to upgrade lock now */
986 status = dlmlock(osb->dlm,
987 level,
988 &lockres->l_lksb,
Mark Fashehb80fc012006-09-12 22:08:14 -0700989 lkm_flags|LKM_CONVERT,
Mark Fashehccd979b2005-12-15 14:31:24 -0800990 lockres->l_name,
Mark Fashehf0681062006-09-08 11:40:10 -0700991 OCFS2_LOCK_ID_MAX_LEN - 1,
Mark Fashehe92d57d2006-09-12 21:34:35 -0700992 ocfs2_locking_ast,
Mark Fashehccd979b2005-12-15 14:31:24 -0800993 lockres,
Mark Fashehaa2623a2006-09-12 21:58:23 -0700994 ocfs2_blocking_ast);
Mark Fashehccd979b2005-12-15 14:31:24 -0800995 if (status != DLM_NORMAL) {
996 if ((lkm_flags & LKM_NOQUEUE) &&
997 (status == DLM_NOTQUEUED))
998 ret = -EAGAIN;
999 else {
1000 ocfs2_log_dlm_error("dlmlock", status,
1001 lockres);
1002 ret = -EINVAL;
1003 }
1004 ocfs2_recover_from_dlm_error(lockres, 1);
1005 goto out;
1006 }
1007
1008 mlog(0, "lock %s, successfull return from dlmlock\n",
1009 lockres->l_name);
1010
1011 /* At this point we've gone inside the dlm and need to
1012 * complete our work regardless. */
1013 catch_signals = 0;
1014
1015 /* wait for busy to clear and carry on */
1016 goto again;
1017 }
1018
1019 /* Ok, if we get here then we're good to go. */
1020 ocfs2_inc_holders(lockres, level);
1021
1022 ret = 0;
1023unlock:
1024 spin_unlock_irqrestore(&lockres->l_lock, flags);
1025out:
1026 /*
1027 * This is helping work around a lock inversion between the page lock
1028 * and dlm locks. One path holds the page lock while calling aops
1029 * which block acquiring dlm locks. The voting thread holds dlm
1030 * locks while acquiring page locks while down converting data locks.
1031 * This block is helping an aop path notice the inversion and back
1032 * off to unlock its page lock before trying the dlm lock again.
1033 */
1034 if (wait && arg_flags & OCFS2_LOCK_NONBLOCK &&
1035 mw.mw_mask & (OCFS2_LOCK_BUSY|OCFS2_LOCK_BLOCKED)) {
1036 wait = 0;
1037 if (lockres_remove_mask_waiter(lockres, &mw))
1038 ret = -EAGAIN;
1039 else
1040 goto again;
1041 }
1042 if (wait) {
1043 ret = ocfs2_wait_for_mask(&mw);
1044 if (ret == 0)
1045 goto again;
1046 mlog_errno(ret);
1047 }
1048
1049 mlog_exit(ret);
1050 return ret;
1051}
1052
1053static void ocfs2_cluster_unlock(struct ocfs2_super *osb,
1054 struct ocfs2_lock_res *lockres,
1055 int level)
1056{
1057 unsigned long flags;
1058
1059 mlog_entry_void();
1060 spin_lock_irqsave(&lockres->l_lock, flags);
1061 ocfs2_dec_holders(lockres, level);
1062 ocfs2_vote_on_unlock(osb, lockres);
1063 spin_unlock_irqrestore(&lockres->l_lock, flags);
1064 mlog_exit_void();
1065}
1066
Adrian Bunkda661162006-11-20 03:24:28 +01001067static int ocfs2_create_new_lock(struct ocfs2_super *osb,
1068 struct ocfs2_lock_res *lockres,
1069 int ex,
1070 int local)
Mark Fashehccd979b2005-12-15 14:31:24 -08001071{
Mark Fashehd680efe2006-09-08 14:14:34 -07001072 int level = ex ? LKM_EXMODE : LKM_PRMODE;
Mark Fashehccd979b2005-12-15 14:31:24 -08001073 unsigned long flags;
Mark Fasheh24c19ef2006-09-22 17:28:19 -07001074 int lkm_flags = local ? LKM_LOCAL : 0;
Mark Fashehccd979b2005-12-15 14:31:24 -08001075
1076 spin_lock_irqsave(&lockres->l_lock, flags);
1077 BUG_ON(lockres->l_flags & OCFS2_LOCK_ATTACHED);
1078 lockres_or_flags(lockres, OCFS2_LOCK_LOCAL);
1079 spin_unlock_irqrestore(&lockres->l_lock, flags);
1080
Mark Fasheh24c19ef2006-09-22 17:28:19 -07001081 return ocfs2_lock_create(osb, lockres, level, lkm_flags);
Mark Fashehccd979b2005-12-15 14:31:24 -08001082}
1083
1084/* Grants us an EX lock on the data and metadata resources, skipping
1085 * the normal cluster directory lookup. Use this ONLY on newly created
1086 * inodes which other nodes can't possibly see, and which haven't been
1087 * hashed in the inode hash yet. This can give us a good performance
1088 * increase as it'll skip the network broadcast normally associated
1089 * with creating a new lock resource. */
1090int ocfs2_create_new_inode_locks(struct inode *inode)
1091{
1092 int ret;
Mark Fashehd680efe2006-09-08 14:14:34 -07001093 struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
Mark Fashehccd979b2005-12-15 14:31:24 -08001094
1095 BUG_ON(!inode);
1096 BUG_ON(!ocfs2_inode_is_new(inode));
1097
1098 mlog_entry_void();
1099
Mark Fashehb06970532006-03-03 10:24:33 -08001100 mlog(0, "Inode %llu\n", (unsigned long long)OCFS2_I(inode)->ip_blkno);
Mark Fashehccd979b2005-12-15 14:31:24 -08001101
1102 /* NOTE: That we don't increment any of the holder counts, nor
1103 * do we add anything to a journal handle. Since this is
1104 * supposed to be a new inode which the cluster doesn't know
1105 * about yet, there is no need to. As far as the LVB handling
1106 * is concerned, this is basically like acquiring an EX lock
1107 * on a resource which has an invalid one -- we'll set it
1108 * valid when we release the EX. */
1109
Mark Fasheh24c19ef2006-09-22 17:28:19 -07001110 ret = ocfs2_create_new_lock(osb, &OCFS2_I(inode)->ip_rw_lockres, 1, 1);
Mark Fashehccd979b2005-12-15 14:31:24 -08001111 if (ret) {
1112 mlog_errno(ret);
1113 goto bail;
1114 }
1115
Mark Fasheh24c19ef2006-09-22 17:28:19 -07001116 /*
1117 * We don't want to use LKM_LOCAL on a meta data lock as they
1118 * don't use a generation in their lock names.
1119 */
1120 ret = ocfs2_create_new_lock(osb, &OCFS2_I(inode)->ip_meta_lockres, 1, 0);
Mark Fashehccd979b2005-12-15 14:31:24 -08001121 if (ret) {
1122 mlog_errno(ret);
1123 goto bail;
1124 }
1125
Mark Fasheh24c19ef2006-09-22 17:28:19 -07001126 ret = ocfs2_create_new_lock(osb, &OCFS2_I(inode)->ip_data_lockres, 1, 1);
Mark Fashehccd979b2005-12-15 14:31:24 -08001127 if (ret) {
1128 mlog_errno(ret);
1129 goto bail;
1130 }
1131
1132bail:
1133 mlog_exit(ret);
1134 return ret;
1135}
1136
1137int ocfs2_rw_lock(struct inode *inode, int write)
1138{
1139 int status, level;
1140 struct ocfs2_lock_res *lockres;
Sunil Mushranc271c5c2006-12-05 17:56:35 -08001141 struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
Mark Fashehccd979b2005-12-15 14:31:24 -08001142
1143 BUG_ON(!inode);
1144
1145 mlog_entry_void();
1146
Mark Fashehb06970532006-03-03 10:24:33 -08001147 mlog(0, "inode %llu take %s RW lock\n",
1148 (unsigned long long)OCFS2_I(inode)->ip_blkno,
Mark Fashehccd979b2005-12-15 14:31:24 -08001149 write ? "EXMODE" : "PRMODE");
1150
Sunil Mushranc271c5c2006-12-05 17:56:35 -08001151 if (ocfs2_mount_local(osb))
1152 return 0;
1153
Mark Fashehccd979b2005-12-15 14:31:24 -08001154 lockres = &OCFS2_I(inode)->ip_rw_lockres;
1155
1156 level = write ? LKM_EXMODE : LKM_PRMODE;
1157
1158 status = ocfs2_cluster_lock(OCFS2_SB(inode->i_sb), lockres, level, 0,
1159 0);
1160 if (status < 0)
1161 mlog_errno(status);
1162
1163 mlog_exit(status);
1164 return status;
1165}
1166
1167void ocfs2_rw_unlock(struct inode *inode, int write)
1168{
1169 int level = write ? LKM_EXMODE : LKM_PRMODE;
1170 struct ocfs2_lock_res *lockres = &OCFS2_I(inode)->ip_rw_lockres;
Sunil Mushranc271c5c2006-12-05 17:56:35 -08001171 struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
Mark Fashehccd979b2005-12-15 14:31:24 -08001172
1173 mlog_entry_void();
1174
Mark Fashehb06970532006-03-03 10:24:33 -08001175 mlog(0, "inode %llu drop %s RW lock\n",
1176 (unsigned long long)OCFS2_I(inode)->ip_blkno,
Mark Fashehccd979b2005-12-15 14:31:24 -08001177 write ? "EXMODE" : "PRMODE");
1178
Sunil Mushranc271c5c2006-12-05 17:56:35 -08001179 if (!ocfs2_mount_local(osb))
1180 ocfs2_cluster_unlock(OCFS2_SB(inode->i_sb), lockres, level);
Mark Fashehccd979b2005-12-15 14:31:24 -08001181
1182 mlog_exit_void();
1183}
1184
1185int ocfs2_data_lock_full(struct inode *inode,
1186 int write,
1187 int arg_flags)
1188{
1189 int status = 0, level;
1190 struct ocfs2_lock_res *lockres;
Sunil Mushranc271c5c2006-12-05 17:56:35 -08001191 struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
Mark Fashehccd979b2005-12-15 14:31:24 -08001192
1193 BUG_ON(!inode);
1194
1195 mlog_entry_void();
1196
Mark Fashehb06970532006-03-03 10:24:33 -08001197 mlog(0, "inode %llu take %s DATA lock\n",
1198 (unsigned long long)OCFS2_I(inode)->ip_blkno,
Mark Fashehccd979b2005-12-15 14:31:24 -08001199 write ? "EXMODE" : "PRMODE");
1200
1201 /* We'll allow faking a readonly data lock for
1202 * rodevices. */
1203 if (ocfs2_is_hard_readonly(OCFS2_SB(inode->i_sb))) {
1204 if (write) {
1205 status = -EROFS;
1206 mlog_errno(status);
1207 }
1208 goto out;
1209 }
1210
Sunil Mushranc271c5c2006-12-05 17:56:35 -08001211 if (ocfs2_mount_local(osb))
1212 goto out;
1213
Mark Fashehccd979b2005-12-15 14:31:24 -08001214 lockres = &OCFS2_I(inode)->ip_data_lockres;
1215
1216 level = write ? LKM_EXMODE : LKM_PRMODE;
1217
1218 status = ocfs2_cluster_lock(OCFS2_SB(inode->i_sb), lockres, level,
1219 0, arg_flags);
1220 if (status < 0 && status != -EAGAIN)
1221 mlog_errno(status);
1222
1223out:
1224 mlog_exit(status);
1225 return status;
1226}
1227
1228/* see ocfs2_meta_lock_with_page() */
1229int ocfs2_data_lock_with_page(struct inode *inode,
1230 int write,
1231 struct page *page)
1232{
1233 int ret;
1234
1235 ret = ocfs2_data_lock_full(inode, write, OCFS2_LOCK_NONBLOCK);
1236 if (ret == -EAGAIN) {
1237 unlock_page(page);
1238 if (ocfs2_data_lock(inode, write) == 0)
1239 ocfs2_data_unlock(inode, write);
1240 ret = AOP_TRUNCATED_PAGE;
1241 }
1242
1243 return ret;
1244}
1245
1246static void ocfs2_vote_on_unlock(struct ocfs2_super *osb,
1247 struct ocfs2_lock_res *lockres)
1248{
1249 int kick = 0;
1250
1251 mlog_entry_void();
1252
1253 /* If we know that another node is waiting on our lock, kick
1254 * the vote thread * pre-emptively when we reach a release
1255 * condition. */
1256 if (lockres->l_flags & OCFS2_LOCK_BLOCKED) {
1257 switch(lockres->l_blocking) {
1258 case LKM_EXMODE:
1259 if (!lockres->l_ex_holders && !lockres->l_ro_holders)
1260 kick = 1;
1261 break;
1262 case LKM_PRMODE:
1263 if (!lockres->l_ex_holders)
1264 kick = 1;
1265 break;
1266 default:
1267 BUG();
1268 }
1269 }
1270
1271 if (kick)
1272 ocfs2_kick_vote_thread(osb);
1273
1274 mlog_exit_void();
1275}
1276
1277void ocfs2_data_unlock(struct inode *inode,
1278 int write)
1279{
1280 int level = write ? LKM_EXMODE : LKM_PRMODE;
1281 struct ocfs2_lock_res *lockres = &OCFS2_I(inode)->ip_data_lockres;
Sunil Mushranc271c5c2006-12-05 17:56:35 -08001282 struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
Mark Fashehccd979b2005-12-15 14:31:24 -08001283
1284 mlog_entry_void();
1285
Mark Fashehb06970532006-03-03 10:24:33 -08001286 mlog(0, "inode %llu drop %s DATA lock\n",
1287 (unsigned long long)OCFS2_I(inode)->ip_blkno,
Mark Fashehccd979b2005-12-15 14:31:24 -08001288 write ? "EXMODE" : "PRMODE");
1289
Sunil Mushranc271c5c2006-12-05 17:56:35 -08001290 if (!ocfs2_is_hard_readonly(OCFS2_SB(inode->i_sb)) &&
1291 !ocfs2_mount_local(osb))
Mark Fashehccd979b2005-12-15 14:31:24 -08001292 ocfs2_cluster_unlock(OCFS2_SB(inode->i_sb), lockres, level);
1293
1294 mlog_exit_void();
1295}
1296
1297#define OCFS2_SEC_BITS 34
1298#define OCFS2_SEC_SHIFT (64 - 34)
1299#define OCFS2_NSEC_MASK ((1ULL << OCFS2_SEC_SHIFT) - 1)
1300
1301/* LVB only has room for 64 bits of time here so we pack it for
1302 * now. */
1303static u64 ocfs2_pack_timespec(struct timespec *spec)
1304{
1305 u64 res;
1306 u64 sec = spec->tv_sec;
1307 u32 nsec = spec->tv_nsec;
1308
1309 res = (sec << OCFS2_SEC_SHIFT) | (nsec & OCFS2_NSEC_MASK);
1310
1311 return res;
1312}
1313
1314/* Call this with the lockres locked. I am reasonably sure we don't
1315 * need ip_lock in this function as anyone who would be changing those
1316 * values is supposed to be blocked in ocfs2_meta_lock right now. */
1317static void __ocfs2_stuff_meta_lvb(struct inode *inode)
1318{
1319 struct ocfs2_inode_info *oi = OCFS2_I(inode);
1320 struct ocfs2_lock_res *lockres = &oi->ip_meta_lockres;
1321 struct ocfs2_meta_lvb *lvb;
1322
1323 mlog_entry_void();
1324
1325 lvb = (struct ocfs2_meta_lvb *) lockres->l_lksb.lvb;
1326
Mark Fasheh24c19ef2006-09-22 17:28:19 -07001327 /*
1328 * Invalidate the LVB of a deleted inode - this way other
1329 * nodes are forced to go to disk and discover the new inode
1330 * status.
1331 */
1332 if (oi->ip_flags & OCFS2_INODE_DELETED) {
1333 lvb->lvb_version = 0;
1334 goto out;
1335 }
1336
Mark Fasheh4d3b83f2006-09-12 15:22:18 -07001337 lvb->lvb_version = OCFS2_LVB_VERSION;
Mark Fashehccd979b2005-12-15 14:31:24 -08001338 lvb->lvb_isize = cpu_to_be64(i_size_read(inode));
1339 lvb->lvb_iclusters = cpu_to_be32(oi->ip_clusters);
1340 lvb->lvb_iuid = cpu_to_be32(inode->i_uid);
1341 lvb->lvb_igid = cpu_to_be32(inode->i_gid);
1342 lvb->lvb_imode = cpu_to_be16(inode->i_mode);
1343 lvb->lvb_inlink = cpu_to_be16(inode->i_nlink);
1344 lvb->lvb_iatime_packed =
1345 cpu_to_be64(ocfs2_pack_timespec(&inode->i_atime));
1346 lvb->lvb_ictime_packed =
1347 cpu_to_be64(ocfs2_pack_timespec(&inode->i_ctime));
1348 lvb->lvb_imtime_packed =
1349 cpu_to_be64(ocfs2_pack_timespec(&inode->i_mtime));
Herbert Poetzlca4d1472006-07-03 17:27:12 -07001350 lvb->lvb_iattr = cpu_to_be32(oi->ip_attr);
Mark Fashehf9e2d822006-09-12 15:35:49 -07001351 lvb->lvb_igeneration = cpu_to_be32(inode->i_generation);
Mark Fashehccd979b2005-12-15 14:31:24 -08001352
Mark Fasheh24c19ef2006-09-22 17:28:19 -07001353out:
Mark Fashehccd979b2005-12-15 14:31:24 -08001354 mlog_meta_lvb(0, lockres);
1355
1356 mlog_exit_void();
1357}
1358
1359static void ocfs2_unpack_timespec(struct timespec *spec,
1360 u64 packed_time)
1361{
1362 spec->tv_sec = packed_time >> OCFS2_SEC_SHIFT;
1363 spec->tv_nsec = packed_time & OCFS2_NSEC_MASK;
1364}
1365
1366static void ocfs2_refresh_inode_from_lvb(struct inode *inode)
1367{
1368 struct ocfs2_inode_info *oi = OCFS2_I(inode);
1369 struct ocfs2_lock_res *lockres = &oi->ip_meta_lockres;
1370 struct ocfs2_meta_lvb *lvb;
1371
1372 mlog_entry_void();
1373
1374 mlog_meta_lvb(0, lockres);
1375
1376 lvb = (struct ocfs2_meta_lvb *) lockres->l_lksb.lvb;
1377
1378 /* We're safe here without the lockres lock... */
1379 spin_lock(&oi->ip_lock);
1380 oi->ip_clusters = be32_to_cpu(lvb->lvb_iclusters);
1381 i_size_write(inode, be64_to_cpu(lvb->lvb_isize));
1382
Herbert Poetzlca4d1472006-07-03 17:27:12 -07001383 oi->ip_attr = be32_to_cpu(lvb->lvb_iattr);
1384 ocfs2_set_inode_flags(inode);
1385
Mark Fashehccd979b2005-12-15 14:31:24 -08001386 /* fast-symlinks are a special case */
1387 if (S_ISLNK(inode->i_mode) && !oi->ip_clusters)
1388 inode->i_blocks = 0;
1389 else
1390 inode->i_blocks =
1391 ocfs2_align_bytes_to_sectors(i_size_read(inode));
1392
1393 inode->i_uid = be32_to_cpu(lvb->lvb_iuid);
1394 inode->i_gid = be32_to_cpu(lvb->lvb_igid);
1395 inode->i_mode = be16_to_cpu(lvb->lvb_imode);
1396 inode->i_nlink = be16_to_cpu(lvb->lvb_inlink);
1397 ocfs2_unpack_timespec(&inode->i_atime,
1398 be64_to_cpu(lvb->lvb_iatime_packed));
1399 ocfs2_unpack_timespec(&inode->i_mtime,
1400 be64_to_cpu(lvb->lvb_imtime_packed));
1401 ocfs2_unpack_timespec(&inode->i_ctime,
1402 be64_to_cpu(lvb->lvb_ictime_packed));
1403 spin_unlock(&oi->ip_lock);
1404
1405 mlog_exit_void();
1406}
1407
Mark Fashehf9e2d822006-09-12 15:35:49 -07001408static inline int ocfs2_meta_lvb_is_trustable(struct inode *inode,
1409 struct ocfs2_lock_res *lockres)
Mark Fashehccd979b2005-12-15 14:31:24 -08001410{
1411 struct ocfs2_meta_lvb *lvb = (struct ocfs2_meta_lvb *) lockres->l_lksb.lvb;
1412
Mark Fashehf9e2d822006-09-12 15:35:49 -07001413 if (lvb->lvb_version == OCFS2_LVB_VERSION
1414 && be32_to_cpu(lvb->lvb_igeneration) == inode->i_generation)
Mark Fashehccd979b2005-12-15 14:31:24 -08001415 return 1;
1416 return 0;
1417}
1418
1419/* Determine whether a lock resource needs to be refreshed, and
1420 * arbitrate who gets to refresh it.
1421 *
1422 * 0 means no refresh needed.
1423 *
1424 * > 0 means you need to refresh this and you MUST call
1425 * ocfs2_complete_lock_res_refresh afterwards. */
1426static int ocfs2_should_refresh_lock_res(struct ocfs2_lock_res *lockres)
1427{
1428 unsigned long flags;
1429 int status = 0;
1430
1431 mlog_entry_void();
1432
1433refresh_check:
1434 spin_lock_irqsave(&lockres->l_lock, flags);
1435 if (!(lockres->l_flags & OCFS2_LOCK_NEEDS_REFRESH)) {
1436 spin_unlock_irqrestore(&lockres->l_lock, flags);
1437 goto bail;
1438 }
1439
1440 if (lockres->l_flags & OCFS2_LOCK_REFRESHING) {
1441 spin_unlock_irqrestore(&lockres->l_lock, flags);
1442
1443 ocfs2_wait_on_refreshing_lock(lockres);
1444 goto refresh_check;
1445 }
1446
1447 /* Ok, I'll be the one to refresh this lock. */
1448 lockres_or_flags(lockres, OCFS2_LOCK_REFRESHING);
1449 spin_unlock_irqrestore(&lockres->l_lock, flags);
1450
1451 status = 1;
1452bail:
1453 mlog_exit(status);
1454 return status;
1455}
1456
1457/* If status is non zero, I'll mark it as not being in refresh
1458 * anymroe, but i won't clear the needs refresh flag. */
1459static inline void ocfs2_complete_lock_res_refresh(struct ocfs2_lock_res *lockres,
1460 int status)
1461{
1462 unsigned long flags;
1463 mlog_entry_void();
1464
1465 spin_lock_irqsave(&lockres->l_lock, flags);
1466 lockres_clear_flags(lockres, OCFS2_LOCK_REFRESHING);
1467 if (!status)
1468 lockres_clear_flags(lockres, OCFS2_LOCK_NEEDS_REFRESH);
1469 spin_unlock_irqrestore(&lockres->l_lock, flags);
1470
1471 wake_up(&lockres->l_event);
1472
1473 mlog_exit_void();
1474}
1475
1476/* may or may not return a bh if it went to disk. */
1477static int ocfs2_meta_lock_update(struct inode *inode,
1478 struct buffer_head **bh)
1479{
1480 int status = 0;
1481 struct ocfs2_inode_info *oi = OCFS2_I(inode);
Sunil Mushranc271c5c2006-12-05 17:56:35 -08001482 struct ocfs2_lock_res *lockres = NULL;
Mark Fashehccd979b2005-12-15 14:31:24 -08001483 struct ocfs2_dinode *fe;
Sunil Mushranc271c5c2006-12-05 17:56:35 -08001484 struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
Mark Fashehccd979b2005-12-15 14:31:24 -08001485
1486 mlog_entry_void();
1487
1488 spin_lock(&oi->ip_lock);
1489 if (oi->ip_flags & OCFS2_INODE_DELETED) {
Mark Fashehb06970532006-03-03 10:24:33 -08001490 mlog(0, "Orphaned inode %llu was deleted while we "
Mark Fashehccd979b2005-12-15 14:31:24 -08001491 "were waiting on a lock. ip_flags = 0x%x\n",
Mark Fashehb06970532006-03-03 10:24:33 -08001492 (unsigned long long)oi->ip_blkno, oi->ip_flags);
Mark Fashehccd979b2005-12-15 14:31:24 -08001493 spin_unlock(&oi->ip_lock);
1494 status = -ENOENT;
1495 goto bail;
1496 }
1497 spin_unlock(&oi->ip_lock);
1498
Sunil Mushranc271c5c2006-12-05 17:56:35 -08001499 if (!ocfs2_mount_local(osb)) {
1500 lockres = &oi->ip_meta_lockres;
Mark Fashehccd979b2005-12-15 14:31:24 -08001501
Sunil Mushranc271c5c2006-12-05 17:56:35 -08001502 if (!ocfs2_should_refresh_lock_res(lockres))
1503 goto bail;
1504 }
Mark Fashehccd979b2005-12-15 14:31:24 -08001505
1506 /* This will discard any caching information we might have had
1507 * for the inode metadata. */
1508 ocfs2_metadata_cache_purge(inode);
1509
1510 /* will do nothing for inode types that don't use the extent
1511 * map (directories, bitmap files, etc) */
1512 ocfs2_extent_map_trunc(inode, 0);
1513
Sunil Mushranc271c5c2006-12-05 17:56:35 -08001514 if (lockres && ocfs2_meta_lvb_is_trustable(inode, lockres)) {
Mark Fashehb06970532006-03-03 10:24:33 -08001515 mlog(0, "Trusting LVB on inode %llu\n",
1516 (unsigned long long)oi->ip_blkno);
Mark Fashehccd979b2005-12-15 14:31:24 -08001517 ocfs2_refresh_inode_from_lvb(inode);
1518 } else {
1519 /* Boo, we have to go to disk. */
1520 /* read bh, cast, ocfs2_refresh_inode */
1521 status = ocfs2_read_block(OCFS2_SB(inode->i_sb), oi->ip_blkno,
1522 bh, OCFS2_BH_CACHED, inode);
1523 if (status < 0) {
1524 mlog_errno(status);
1525 goto bail_refresh;
1526 }
1527 fe = (struct ocfs2_dinode *) (*bh)->b_data;
1528
1529 /* This is a good chance to make sure we're not
1530 * locking an invalid object.
1531 *
1532 * We bug on a stale inode here because we checked
1533 * above whether it was wiped from disk. The wiping
1534 * node provides a guarantee that we receive that
1535 * message and can mark the inode before dropping any
1536 * locks associated with it. */
1537 if (!OCFS2_IS_VALID_DINODE(fe)) {
1538 OCFS2_RO_ON_INVALID_DINODE(inode->i_sb, fe);
1539 status = -EIO;
1540 goto bail_refresh;
1541 }
1542 mlog_bug_on_msg(inode->i_generation !=
1543 le32_to_cpu(fe->i_generation),
Mark Fashehb06970532006-03-03 10:24:33 -08001544 "Invalid dinode %llu disk generation: %u "
Mark Fashehccd979b2005-12-15 14:31:24 -08001545 "inode->i_generation: %u\n",
Mark Fashehb06970532006-03-03 10:24:33 -08001546 (unsigned long long)oi->ip_blkno,
1547 le32_to_cpu(fe->i_generation),
Mark Fashehccd979b2005-12-15 14:31:24 -08001548 inode->i_generation);
1549 mlog_bug_on_msg(le64_to_cpu(fe->i_dtime) ||
1550 !(fe->i_flags & cpu_to_le32(OCFS2_VALID_FL)),
Mark Fashehb06970532006-03-03 10:24:33 -08001551 "Stale dinode %llu dtime: %llu flags: 0x%x\n",
1552 (unsigned long long)oi->ip_blkno,
1553 (unsigned long long)le64_to_cpu(fe->i_dtime),
Mark Fashehccd979b2005-12-15 14:31:24 -08001554 le32_to_cpu(fe->i_flags));
1555
1556 ocfs2_refresh_inode(inode, fe);
1557 }
1558
1559 status = 0;
1560bail_refresh:
Sunil Mushranc271c5c2006-12-05 17:56:35 -08001561 if (lockres)
1562 ocfs2_complete_lock_res_refresh(lockres, status);
Mark Fashehccd979b2005-12-15 14:31:24 -08001563bail:
1564 mlog_exit(status);
1565 return status;
1566}
1567
1568static int ocfs2_assign_bh(struct inode *inode,
1569 struct buffer_head **ret_bh,
1570 struct buffer_head *passed_bh)
1571{
1572 int status;
1573
1574 if (passed_bh) {
1575 /* Ok, the update went to disk for us, use the
1576 * returned bh. */
1577 *ret_bh = passed_bh;
1578 get_bh(*ret_bh);
1579
1580 return 0;
1581 }
1582
1583 status = ocfs2_read_block(OCFS2_SB(inode->i_sb),
1584 OCFS2_I(inode)->ip_blkno,
1585 ret_bh,
1586 OCFS2_BH_CACHED,
1587 inode);
1588 if (status < 0)
1589 mlog_errno(status);
1590
1591 return status;
1592}
1593
1594/*
1595 * returns < 0 error if the callback will never be called, otherwise
1596 * the result of the lock will be communicated via the callback.
1597 */
1598int ocfs2_meta_lock_full(struct inode *inode,
Mark Fashehccd979b2005-12-15 14:31:24 -08001599 struct buffer_head **ret_bh,
1600 int ex,
1601 int arg_flags)
1602{
1603 int status, level, dlm_flags, acquired;
Sunil Mushranc271c5c2006-12-05 17:56:35 -08001604 struct ocfs2_lock_res *lockres = NULL;
Mark Fashehccd979b2005-12-15 14:31:24 -08001605 struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
1606 struct buffer_head *local_bh = NULL;
1607
1608 BUG_ON(!inode);
1609
1610 mlog_entry_void();
1611
Mark Fashehb06970532006-03-03 10:24:33 -08001612 mlog(0, "inode %llu, take %s META lock\n",
1613 (unsigned long long)OCFS2_I(inode)->ip_blkno,
Mark Fashehccd979b2005-12-15 14:31:24 -08001614 ex ? "EXMODE" : "PRMODE");
1615
1616 status = 0;
1617 acquired = 0;
1618 /* We'll allow faking a readonly metadata lock for
1619 * rodevices. */
1620 if (ocfs2_is_hard_readonly(osb)) {
1621 if (ex)
1622 status = -EROFS;
1623 goto bail;
1624 }
1625
Sunil Mushranc271c5c2006-12-05 17:56:35 -08001626 if (ocfs2_mount_local(osb))
1627 goto local;
1628
Mark Fashehccd979b2005-12-15 14:31:24 -08001629 if (!(arg_flags & OCFS2_META_LOCK_RECOVERY))
1630 wait_event(osb->recovery_event,
1631 ocfs2_node_map_is_empty(osb, &osb->recovery_map));
1632
1633 acquired = 0;
1634 lockres = &OCFS2_I(inode)->ip_meta_lockres;
1635 level = ex ? LKM_EXMODE : LKM_PRMODE;
1636 dlm_flags = 0;
1637 if (arg_flags & OCFS2_META_LOCK_NOQUEUE)
1638 dlm_flags |= LKM_NOQUEUE;
1639
1640 status = ocfs2_cluster_lock(osb, lockres, level, dlm_flags, arg_flags);
1641 if (status < 0) {
1642 if (status != -EAGAIN && status != -EIOCBRETRY)
1643 mlog_errno(status);
1644 goto bail;
1645 }
1646
1647 /* Notify the error cleanup path to drop the cluster lock. */
1648 acquired = 1;
1649
1650 /* We wait twice because a node may have died while we were in
1651 * the lower dlm layers. The second time though, we've
1652 * committed to owning this lock so we don't allow signals to
1653 * abort the operation. */
1654 if (!(arg_flags & OCFS2_META_LOCK_RECOVERY))
1655 wait_event(osb->recovery_event,
1656 ocfs2_node_map_is_empty(osb, &osb->recovery_map));
1657
Sunil Mushranc271c5c2006-12-05 17:56:35 -08001658local:
Mark Fasheh24c19ef2006-09-22 17:28:19 -07001659 /*
1660 * We only see this flag if we're being called from
1661 * ocfs2_read_locked_inode(). It means we're locking an inode
1662 * which hasn't been populated yet, so clear the refresh flag
1663 * and let the caller handle it.
1664 */
1665 if (inode->i_state & I_NEW) {
1666 status = 0;
Sunil Mushranc271c5c2006-12-05 17:56:35 -08001667 if (lockres)
1668 ocfs2_complete_lock_res_refresh(lockres, 0);
Mark Fasheh24c19ef2006-09-22 17:28:19 -07001669 goto bail;
1670 }
1671
Mark Fashehccd979b2005-12-15 14:31:24 -08001672 /* This is fun. The caller may want a bh back, or it may
1673 * not. ocfs2_meta_lock_update definitely wants one in, but
1674 * may or may not read one, depending on what's in the
1675 * LVB. The result of all of this is that we've *only* gone to
1676 * disk if we have to, so the complexity is worthwhile. */
1677 status = ocfs2_meta_lock_update(inode, &local_bh);
1678 if (status < 0) {
1679 if (status != -ENOENT)
1680 mlog_errno(status);
1681 goto bail;
1682 }
1683
1684 if (ret_bh) {
1685 status = ocfs2_assign_bh(inode, ret_bh, local_bh);
1686 if (status < 0) {
1687 mlog_errno(status);
1688 goto bail;
1689 }
1690 }
1691
Mark Fashehccd979b2005-12-15 14:31:24 -08001692bail:
1693 if (status < 0) {
1694 if (ret_bh && (*ret_bh)) {
1695 brelse(*ret_bh);
1696 *ret_bh = NULL;
1697 }
1698 if (acquired)
1699 ocfs2_meta_unlock(inode, ex);
1700 }
1701
1702 if (local_bh)
1703 brelse(local_bh);
1704
1705 mlog_exit(status);
1706 return status;
1707}
1708
1709/*
1710 * This is working around a lock inversion between tasks acquiring DLM locks
1711 * while holding a page lock and the vote thread which blocks dlm lock acquiry
1712 * while acquiring page locks.
1713 *
1714 * ** These _with_page variantes are only intended to be called from aop
1715 * methods that hold page locks and return a very specific *positive* error
1716 * code that aop methods pass up to the VFS -- test for errors with != 0. **
1717 *
1718 * The DLM is called such that it returns -EAGAIN if it would have blocked
1719 * waiting for the vote thread. In that case we unlock our page so the vote
1720 * thread can make progress. Once we've done this we have to return
1721 * AOP_TRUNCATED_PAGE so the aop method that called us can bubble that back up
1722 * into the VFS who will then immediately retry the aop call.
1723 *
1724 * We do a blocking lock and immediate unlock before returning, though, so that
1725 * the lock has a great chance of being cached on this node by the time the VFS
1726 * calls back to retry the aop. This has a potential to livelock as nodes
1727 * ping locks back and forth, but that's a risk we're willing to take to avoid
1728 * the lock inversion simply.
1729 */
1730int ocfs2_meta_lock_with_page(struct inode *inode,
Mark Fashehccd979b2005-12-15 14:31:24 -08001731 struct buffer_head **ret_bh,
1732 int ex,
1733 struct page *page)
1734{
1735 int ret;
1736
Mark Fasheh4bcec182006-10-09 16:02:40 -07001737 ret = ocfs2_meta_lock_full(inode, ret_bh, ex, OCFS2_LOCK_NONBLOCK);
Mark Fashehccd979b2005-12-15 14:31:24 -08001738 if (ret == -EAGAIN) {
1739 unlock_page(page);
Mark Fasheh4bcec182006-10-09 16:02:40 -07001740 if (ocfs2_meta_lock(inode, ret_bh, ex) == 0)
Mark Fashehccd979b2005-12-15 14:31:24 -08001741 ocfs2_meta_unlock(inode, ex);
1742 ret = AOP_TRUNCATED_PAGE;
1743 }
1744
1745 return ret;
1746}
1747
Tiger Yang7f1a37e2006-11-15 15:48:42 +08001748int ocfs2_meta_lock_atime(struct inode *inode,
1749 struct vfsmount *vfsmnt,
1750 int *level)
1751{
1752 int ret;
1753
1754 mlog_entry_void();
1755 ret = ocfs2_meta_lock(inode, NULL, 0);
1756 if (ret < 0) {
1757 mlog_errno(ret);
1758 return ret;
1759 }
1760
1761 /*
1762 * If we should update atime, we will get EX lock,
1763 * otherwise we just get PR lock.
1764 */
1765 if (ocfs2_should_update_atime(inode, vfsmnt)) {
1766 struct buffer_head *bh = NULL;
1767
1768 ocfs2_meta_unlock(inode, 0);
1769 ret = ocfs2_meta_lock(inode, &bh, 1);
1770 if (ret < 0) {
1771 mlog_errno(ret);
1772 return ret;
1773 }
1774 *level = 1;
1775 if (ocfs2_should_update_atime(inode, vfsmnt))
1776 ocfs2_update_inode_atime(inode, bh);
1777 if (bh)
1778 brelse(bh);
1779 } else
1780 *level = 0;
1781
1782 mlog_exit(ret);
1783 return ret;
1784}
1785
Mark Fashehccd979b2005-12-15 14:31:24 -08001786void ocfs2_meta_unlock(struct inode *inode,
1787 int ex)
1788{
1789 int level = ex ? LKM_EXMODE : LKM_PRMODE;
1790 struct ocfs2_lock_res *lockres = &OCFS2_I(inode)->ip_meta_lockres;
Sunil Mushranc271c5c2006-12-05 17:56:35 -08001791 struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
Mark Fashehccd979b2005-12-15 14:31:24 -08001792
1793 mlog_entry_void();
1794
Mark Fashehb06970532006-03-03 10:24:33 -08001795 mlog(0, "inode %llu drop %s META lock\n",
1796 (unsigned long long)OCFS2_I(inode)->ip_blkno,
Mark Fashehccd979b2005-12-15 14:31:24 -08001797 ex ? "EXMODE" : "PRMODE");
1798
Sunil Mushranc271c5c2006-12-05 17:56:35 -08001799 if (!ocfs2_is_hard_readonly(OCFS2_SB(inode->i_sb)) &&
1800 !ocfs2_mount_local(osb))
Mark Fashehccd979b2005-12-15 14:31:24 -08001801 ocfs2_cluster_unlock(OCFS2_SB(inode->i_sb), lockres, level);
1802
1803 mlog_exit_void();
1804}
1805
1806int ocfs2_super_lock(struct ocfs2_super *osb,
1807 int ex)
1808{
Sunil Mushranc271c5c2006-12-05 17:56:35 -08001809 int status = 0;
Mark Fashehccd979b2005-12-15 14:31:24 -08001810 int level = ex ? LKM_EXMODE : LKM_PRMODE;
1811 struct ocfs2_lock_res *lockres = &osb->osb_super_lockres;
1812 struct buffer_head *bh;
1813 struct ocfs2_slot_info *si = osb->slot_info;
1814
1815 mlog_entry_void();
1816
1817 if (ocfs2_is_hard_readonly(osb))
1818 return -EROFS;
1819
Sunil Mushranc271c5c2006-12-05 17:56:35 -08001820 if (ocfs2_mount_local(osb))
1821 goto bail;
1822
Mark Fashehccd979b2005-12-15 14:31:24 -08001823 status = ocfs2_cluster_lock(osb, lockres, level, 0, 0);
1824 if (status < 0) {
1825 mlog_errno(status);
1826 goto bail;
1827 }
1828
1829 /* The super block lock path is really in the best position to
1830 * know when resources covered by the lock need to be
1831 * refreshed, so we do it here. Of course, making sense of
1832 * everything is up to the caller :) */
1833 status = ocfs2_should_refresh_lock_res(lockres);
1834 if (status < 0) {
1835 mlog_errno(status);
1836 goto bail;
1837 }
1838 if (status) {
1839 bh = si->si_bh;
1840 status = ocfs2_read_block(osb, bh->b_blocknr, &bh, 0,
1841 si->si_inode);
1842 if (status == 0)
1843 ocfs2_update_slot_info(si);
1844
1845 ocfs2_complete_lock_res_refresh(lockres, status);
1846
1847 if (status < 0)
1848 mlog_errno(status);
1849 }
1850bail:
1851 mlog_exit(status);
1852 return status;
1853}
1854
1855void ocfs2_super_unlock(struct ocfs2_super *osb,
1856 int ex)
1857{
1858 int level = ex ? LKM_EXMODE : LKM_PRMODE;
1859 struct ocfs2_lock_res *lockres = &osb->osb_super_lockres;
1860
Sunil Mushranc271c5c2006-12-05 17:56:35 -08001861 if (!ocfs2_mount_local(osb))
1862 ocfs2_cluster_unlock(osb, lockres, level);
Mark Fashehccd979b2005-12-15 14:31:24 -08001863}
1864
1865int ocfs2_rename_lock(struct ocfs2_super *osb)
1866{
1867 int status;
1868 struct ocfs2_lock_res *lockres = &osb->osb_rename_lockres;
1869
1870 if (ocfs2_is_hard_readonly(osb))
1871 return -EROFS;
1872
Sunil Mushranc271c5c2006-12-05 17:56:35 -08001873 if (ocfs2_mount_local(osb))
1874 return 0;
1875
Mark Fashehccd979b2005-12-15 14:31:24 -08001876 status = ocfs2_cluster_lock(osb, lockres, LKM_EXMODE, 0, 0);
1877 if (status < 0)
1878 mlog_errno(status);
1879
1880 return status;
1881}
1882
1883void ocfs2_rename_unlock(struct ocfs2_super *osb)
1884{
1885 struct ocfs2_lock_res *lockres = &osb->osb_rename_lockres;
1886
Sunil Mushranc271c5c2006-12-05 17:56:35 -08001887 if (!ocfs2_mount_local(osb))
1888 ocfs2_cluster_unlock(osb, lockres, LKM_EXMODE);
Mark Fashehccd979b2005-12-15 14:31:24 -08001889}
1890
Mark Fashehd680efe2006-09-08 14:14:34 -07001891int ocfs2_dentry_lock(struct dentry *dentry, int ex)
1892{
1893 int ret;
1894 int level = ex ? LKM_EXMODE : LKM_PRMODE;
1895 struct ocfs2_dentry_lock *dl = dentry->d_fsdata;
1896 struct ocfs2_super *osb = OCFS2_SB(dentry->d_sb);
1897
1898 BUG_ON(!dl);
1899
1900 if (ocfs2_is_hard_readonly(osb))
1901 return -EROFS;
1902
Sunil Mushranc271c5c2006-12-05 17:56:35 -08001903 if (ocfs2_mount_local(osb))
1904 return 0;
1905
Mark Fashehd680efe2006-09-08 14:14:34 -07001906 ret = ocfs2_cluster_lock(osb, &dl->dl_lockres, level, 0, 0);
1907 if (ret < 0)
1908 mlog_errno(ret);
1909
1910 return ret;
1911}
1912
1913void ocfs2_dentry_unlock(struct dentry *dentry, int ex)
1914{
1915 int level = ex ? LKM_EXMODE : LKM_PRMODE;
1916 struct ocfs2_dentry_lock *dl = dentry->d_fsdata;
1917 struct ocfs2_super *osb = OCFS2_SB(dentry->d_sb);
1918
Sunil Mushranc271c5c2006-12-05 17:56:35 -08001919 if (!ocfs2_mount_local(osb))
1920 ocfs2_cluster_unlock(osb, &dl->dl_lockres, level);
Mark Fashehd680efe2006-09-08 14:14:34 -07001921}
1922
Mark Fashehccd979b2005-12-15 14:31:24 -08001923/* Reference counting of the dlm debug structure. We want this because
1924 * open references on the debug inodes can live on after a mount, so
1925 * we can't rely on the ocfs2_super to always exist. */
1926static void ocfs2_dlm_debug_free(struct kref *kref)
1927{
1928 struct ocfs2_dlm_debug *dlm_debug;
1929
1930 dlm_debug = container_of(kref, struct ocfs2_dlm_debug, d_refcnt);
1931
1932 kfree(dlm_debug);
1933}
1934
1935void ocfs2_put_dlm_debug(struct ocfs2_dlm_debug *dlm_debug)
1936{
1937 if (dlm_debug)
1938 kref_put(&dlm_debug->d_refcnt, ocfs2_dlm_debug_free);
1939}
1940
1941static void ocfs2_get_dlm_debug(struct ocfs2_dlm_debug *debug)
1942{
1943 kref_get(&debug->d_refcnt);
1944}
1945
1946struct ocfs2_dlm_debug *ocfs2_new_dlm_debug(void)
1947{
1948 struct ocfs2_dlm_debug *dlm_debug;
1949
1950 dlm_debug = kmalloc(sizeof(struct ocfs2_dlm_debug), GFP_KERNEL);
1951 if (!dlm_debug) {
1952 mlog_errno(-ENOMEM);
1953 goto out;
1954 }
1955
1956 kref_init(&dlm_debug->d_refcnt);
1957 INIT_LIST_HEAD(&dlm_debug->d_lockres_tracking);
1958 dlm_debug->d_locking_state = NULL;
1959out:
1960 return dlm_debug;
1961}
1962
1963/* Access to this is arbitrated for us via seq_file->sem. */
1964struct ocfs2_dlm_seq_priv {
1965 struct ocfs2_dlm_debug *p_dlm_debug;
1966 struct ocfs2_lock_res p_iter_res;
1967 struct ocfs2_lock_res p_tmp_res;
1968};
1969
1970static struct ocfs2_lock_res *ocfs2_dlm_next_res(struct ocfs2_lock_res *start,
1971 struct ocfs2_dlm_seq_priv *priv)
1972{
1973 struct ocfs2_lock_res *iter, *ret = NULL;
1974 struct ocfs2_dlm_debug *dlm_debug = priv->p_dlm_debug;
1975
1976 assert_spin_locked(&ocfs2_dlm_tracking_lock);
1977
1978 list_for_each_entry(iter, &start->l_debug_list, l_debug_list) {
1979 /* discover the head of the list */
1980 if (&iter->l_debug_list == &dlm_debug->d_lockres_tracking) {
1981 mlog(0, "End of list found, %p\n", ret);
1982 break;
1983 }
1984
1985 /* We track our "dummy" iteration lockres' by a NULL
1986 * l_ops field. */
1987 if (iter->l_ops != NULL) {
1988 ret = iter;
1989 break;
1990 }
1991 }
1992
1993 return ret;
1994}
1995
1996static void *ocfs2_dlm_seq_start(struct seq_file *m, loff_t *pos)
1997{
1998 struct ocfs2_dlm_seq_priv *priv = m->private;
1999 struct ocfs2_lock_res *iter;
2000
2001 spin_lock(&ocfs2_dlm_tracking_lock);
2002 iter = ocfs2_dlm_next_res(&priv->p_iter_res, priv);
2003 if (iter) {
2004 /* Since lockres' have the lifetime of their container
2005 * (which can be inodes, ocfs2_supers, etc) we want to
2006 * copy this out to a temporary lockres while still
2007 * under the spinlock. Obviously after this we can't
2008 * trust any pointers on the copy returned, but that's
2009 * ok as the information we want isn't typically held
2010 * in them. */
2011 priv->p_tmp_res = *iter;
2012 iter = &priv->p_tmp_res;
2013 }
2014 spin_unlock(&ocfs2_dlm_tracking_lock);
2015
2016 return iter;
2017}
2018
2019static void ocfs2_dlm_seq_stop(struct seq_file *m, void *v)
2020{
2021}
2022
2023static void *ocfs2_dlm_seq_next(struct seq_file *m, void *v, loff_t *pos)
2024{
2025 struct ocfs2_dlm_seq_priv *priv = m->private;
2026 struct ocfs2_lock_res *iter = v;
2027 struct ocfs2_lock_res *dummy = &priv->p_iter_res;
2028
2029 spin_lock(&ocfs2_dlm_tracking_lock);
2030 iter = ocfs2_dlm_next_res(iter, priv);
2031 list_del_init(&dummy->l_debug_list);
2032 if (iter) {
2033 list_add(&dummy->l_debug_list, &iter->l_debug_list);
2034 priv->p_tmp_res = *iter;
2035 iter = &priv->p_tmp_res;
2036 }
2037 spin_unlock(&ocfs2_dlm_tracking_lock);
2038
2039 return iter;
2040}
2041
2042/* So that debugfs.ocfs2 can determine which format is being used */
2043#define OCFS2_DLM_DEBUG_STR_VERSION 1
2044static int ocfs2_dlm_seq_show(struct seq_file *m, void *v)
2045{
2046 int i;
2047 char *lvb;
2048 struct ocfs2_lock_res *lockres = v;
2049
2050 if (!lockres)
2051 return -EINVAL;
2052
Mark Fashehd680efe2006-09-08 14:14:34 -07002053 seq_printf(m, "0x%x\t", OCFS2_DLM_DEBUG_STR_VERSION);
2054
2055 if (lockres->l_type == OCFS2_LOCK_TYPE_DENTRY)
2056 seq_printf(m, "%.*s%08x\t", OCFS2_DENTRY_LOCK_INO_START - 1,
2057 lockres->l_name,
2058 (unsigned int)ocfs2_get_dentry_lock_ino(lockres));
2059 else
2060 seq_printf(m, "%.*s\t", OCFS2_LOCK_ID_MAX_LEN, lockres->l_name);
2061
2062 seq_printf(m, "%d\t"
Mark Fashehccd979b2005-12-15 14:31:24 -08002063 "0x%lx\t"
2064 "0x%x\t"
2065 "0x%x\t"
2066 "%u\t"
2067 "%u\t"
2068 "%d\t"
2069 "%d\t",
Mark Fashehccd979b2005-12-15 14:31:24 -08002070 lockres->l_level,
2071 lockres->l_flags,
2072 lockres->l_action,
2073 lockres->l_unlock_action,
2074 lockres->l_ro_holders,
2075 lockres->l_ex_holders,
2076 lockres->l_requested,
2077 lockres->l_blocking);
2078
2079 /* Dump the raw LVB */
2080 lvb = lockres->l_lksb.lvb;
2081 for(i = 0; i < DLM_LVB_LEN; i++)
2082 seq_printf(m, "0x%x\t", lvb[i]);
2083
2084 /* End the line */
2085 seq_printf(m, "\n");
2086 return 0;
2087}
2088
2089static struct seq_operations ocfs2_dlm_seq_ops = {
2090 .start = ocfs2_dlm_seq_start,
2091 .stop = ocfs2_dlm_seq_stop,
2092 .next = ocfs2_dlm_seq_next,
2093 .show = ocfs2_dlm_seq_show,
2094};
2095
2096static int ocfs2_dlm_debug_release(struct inode *inode, struct file *file)
2097{
2098 struct seq_file *seq = (struct seq_file *) file->private_data;
2099 struct ocfs2_dlm_seq_priv *priv = seq->private;
2100 struct ocfs2_lock_res *res = &priv->p_iter_res;
2101
2102 ocfs2_remove_lockres_tracking(res);
2103 ocfs2_put_dlm_debug(priv->p_dlm_debug);
2104 return seq_release_private(inode, file);
2105}
2106
2107static int ocfs2_dlm_debug_open(struct inode *inode, struct file *file)
2108{
2109 int ret;
2110 struct ocfs2_dlm_seq_priv *priv;
2111 struct seq_file *seq;
2112 struct ocfs2_super *osb;
2113
2114 priv = kzalloc(sizeof(struct ocfs2_dlm_seq_priv), GFP_KERNEL);
2115 if (!priv) {
2116 ret = -ENOMEM;
2117 mlog_errno(ret);
2118 goto out;
2119 }
Theodore Ts'o8e18e292006-09-27 01:50:46 -07002120 osb = inode->i_private;
Mark Fashehccd979b2005-12-15 14:31:24 -08002121 ocfs2_get_dlm_debug(osb->osb_dlm_debug);
2122 priv->p_dlm_debug = osb->osb_dlm_debug;
2123 INIT_LIST_HEAD(&priv->p_iter_res.l_debug_list);
2124
2125 ret = seq_open(file, &ocfs2_dlm_seq_ops);
2126 if (ret) {
2127 kfree(priv);
2128 mlog_errno(ret);
2129 goto out;
2130 }
2131
2132 seq = (struct seq_file *) file->private_data;
2133 seq->private = priv;
2134
2135 ocfs2_add_lockres_tracking(&priv->p_iter_res,
2136 priv->p_dlm_debug);
2137
2138out:
2139 return ret;
2140}
2141
Arjan van de Ven4b6f5d22006-03-28 01:56:42 -08002142static const struct file_operations ocfs2_dlm_debug_fops = {
Mark Fashehccd979b2005-12-15 14:31:24 -08002143 .open = ocfs2_dlm_debug_open,
2144 .release = ocfs2_dlm_debug_release,
2145 .read = seq_read,
2146 .llseek = seq_lseek,
2147};
2148
2149static int ocfs2_dlm_init_debug(struct ocfs2_super *osb)
2150{
2151 int ret = 0;
2152 struct ocfs2_dlm_debug *dlm_debug = osb->osb_dlm_debug;
2153
2154 dlm_debug->d_locking_state = debugfs_create_file("locking_state",
2155 S_IFREG|S_IRUSR,
2156 osb->osb_debug_root,
2157 osb,
2158 &ocfs2_dlm_debug_fops);
2159 if (!dlm_debug->d_locking_state) {
2160 ret = -EINVAL;
2161 mlog(ML_ERROR,
2162 "Unable to create locking state debugfs file.\n");
2163 goto out;
2164 }
2165
2166 ocfs2_get_dlm_debug(dlm_debug);
2167out:
2168 return ret;
2169}
2170
2171static void ocfs2_dlm_shutdown_debug(struct ocfs2_super *osb)
2172{
2173 struct ocfs2_dlm_debug *dlm_debug = osb->osb_dlm_debug;
2174
2175 if (dlm_debug) {
2176 debugfs_remove(dlm_debug->d_locking_state);
2177 ocfs2_put_dlm_debug(dlm_debug);
2178 }
2179}
2180
2181int ocfs2_dlm_init(struct ocfs2_super *osb)
2182{
Sunil Mushranc271c5c2006-12-05 17:56:35 -08002183 int status = 0;
Mark Fashehccd979b2005-12-15 14:31:24 -08002184 u32 dlm_key;
Sunil Mushranc271c5c2006-12-05 17:56:35 -08002185 struct dlm_ctxt *dlm = NULL;
Mark Fashehccd979b2005-12-15 14:31:24 -08002186
2187 mlog_entry_void();
2188
Sunil Mushranc271c5c2006-12-05 17:56:35 -08002189 if (ocfs2_mount_local(osb))
2190 goto local;
2191
Mark Fashehccd979b2005-12-15 14:31:24 -08002192 status = ocfs2_dlm_init_debug(osb);
2193 if (status < 0) {
2194 mlog_errno(status);
2195 goto bail;
2196 }
2197
2198 /* launch vote thread */
Mark Fasheh78427042006-05-04 12:03:26 -07002199 osb->vote_task = kthread_run(ocfs2_vote_thread, osb, "ocfs2vote");
Mark Fashehccd979b2005-12-15 14:31:24 -08002200 if (IS_ERR(osb->vote_task)) {
2201 status = PTR_ERR(osb->vote_task);
2202 osb->vote_task = NULL;
2203 mlog_errno(status);
2204 goto bail;
2205 }
2206
2207 /* used by the dlm code to make message headers unique, each
2208 * node in this domain must agree on this. */
2209 dlm_key = crc32_le(0, osb->uuid_str, strlen(osb->uuid_str));
2210
2211 /* for now, uuid == domain */
2212 dlm = dlm_register_domain(osb->uuid_str, dlm_key);
2213 if (IS_ERR(dlm)) {
2214 status = PTR_ERR(dlm);
2215 mlog_errno(status);
2216 goto bail;
2217 }
2218
Sunil Mushranc271c5c2006-12-05 17:56:35 -08002219 dlm_register_eviction_cb(dlm, &osb->osb_eviction_cb);
2220
2221local:
Mark Fashehccd979b2005-12-15 14:31:24 -08002222 ocfs2_super_lock_res_init(&osb->osb_super_lockres, osb);
2223 ocfs2_rename_lock_res_init(&osb->osb_rename_lockres, osb);
2224
Mark Fashehccd979b2005-12-15 14:31:24 -08002225 osb->dlm = dlm;
2226
2227 status = 0;
2228bail:
2229 if (status < 0) {
2230 ocfs2_dlm_shutdown_debug(osb);
2231 if (osb->vote_task)
2232 kthread_stop(osb->vote_task);
2233 }
2234
2235 mlog_exit(status);
2236 return status;
2237}
2238
2239void ocfs2_dlm_shutdown(struct ocfs2_super *osb)
2240{
2241 mlog_entry_void();
2242
2243 dlm_unregister_eviction_cb(&osb->osb_eviction_cb);
2244
2245 ocfs2_drop_osb_locks(osb);
2246
2247 if (osb->vote_task) {
2248 kthread_stop(osb->vote_task);
2249 osb->vote_task = NULL;
2250 }
2251
2252 ocfs2_lock_res_free(&osb->osb_super_lockres);
2253 ocfs2_lock_res_free(&osb->osb_rename_lockres);
2254
2255 dlm_unregister_domain(osb->dlm);
2256 osb->dlm = NULL;
2257
2258 ocfs2_dlm_shutdown_debug(osb);
2259
2260 mlog_exit_void();
2261}
2262
Mark Fasheh2a45f2d2006-09-12 21:36:58 -07002263static void ocfs2_unlock_ast(void *opaque, enum dlm_status status)
Mark Fashehccd979b2005-12-15 14:31:24 -08002264{
2265 struct ocfs2_lock_res *lockres = opaque;
2266 unsigned long flags;
2267
2268 mlog_entry_void();
2269
2270 mlog(0, "UNLOCK AST called on lock %s, action = %d\n", lockres->l_name,
2271 lockres->l_unlock_action);
2272
2273 spin_lock_irqsave(&lockres->l_lock, flags);
2274 /* We tried to cancel a convert request, but it was already
2275 * granted. All we want to do here is clear our unlock
2276 * state. The wake_up call done at the bottom is redundant
2277 * (ocfs2_prepare_cancel_convert doesn't sleep on this) but doesn't
2278 * hurt anything anyway */
2279 if (status == DLM_CANCELGRANT &&
2280 lockres->l_unlock_action == OCFS2_UNLOCK_CANCEL_CONVERT) {
2281 mlog(0, "Got cancelgrant for %s\n", lockres->l_name);
2282
2283 /* We don't clear the busy flag in this case as it
2284 * should have been cleared by the ast which the dlm
2285 * has called. */
2286 goto complete_unlock;
2287 }
2288
2289 if (status != DLM_NORMAL) {
2290 mlog(ML_ERROR, "Dlm passes status %d for lock %s, "
2291 "unlock_action %d\n", status, lockres->l_name,
2292 lockres->l_unlock_action);
2293 spin_unlock_irqrestore(&lockres->l_lock, flags);
2294 return;
2295 }
2296
2297 switch(lockres->l_unlock_action) {
2298 case OCFS2_UNLOCK_CANCEL_CONVERT:
2299 mlog(0, "Cancel convert success for %s\n", lockres->l_name);
2300 lockres->l_action = OCFS2_AST_INVALID;
2301 break;
2302 case OCFS2_UNLOCK_DROP_LOCK:
2303 lockres->l_level = LKM_IVMODE;
2304 break;
2305 default:
2306 BUG();
2307 }
2308
2309 lockres_clear_flags(lockres, OCFS2_LOCK_BUSY);
2310complete_unlock:
2311 lockres->l_unlock_action = OCFS2_UNLOCK_INVALID;
2312 spin_unlock_irqrestore(&lockres->l_lock, flags);
2313
2314 wake_up(&lockres->l_event);
2315
2316 mlog_exit_void();
2317}
2318
Mark Fashehccd979b2005-12-15 14:31:24 -08002319static int ocfs2_drop_lock(struct ocfs2_super *osb,
Mark Fasheh0d5dc6c2006-09-14 14:44:51 -07002320 struct ocfs2_lock_res *lockres)
Mark Fashehccd979b2005-12-15 14:31:24 -08002321{
2322 enum dlm_status status;
2323 unsigned long flags;
Mark Fashehb80fc012006-09-12 22:08:14 -07002324 int lkm_flags = 0;
Mark Fashehccd979b2005-12-15 14:31:24 -08002325
2326 /* We didn't get anywhere near actually using this lockres. */
2327 if (!(lockres->l_flags & OCFS2_LOCK_INITIALIZED))
2328 goto out;
2329
Mark Fashehb80fc012006-09-12 22:08:14 -07002330 if (lockres->l_ops->flags & LOCK_TYPE_USES_LVB)
2331 lkm_flags |= LKM_VALBLK;
2332
Mark Fashehccd979b2005-12-15 14:31:24 -08002333 spin_lock_irqsave(&lockres->l_lock, flags);
2334
2335 mlog_bug_on_msg(!(lockres->l_flags & OCFS2_LOCK_FREEING),
2336 "lockres %s, flags 0x%lx\n",
2337 lockres->l_name, lockres->l_flags);
2338
2339 while (lockres->l_flags & OCFS2_LOCK_BUSY) {
2340 mlog(0, "waiting on busy lock \"%s\": flags = %lx, action = "
2341 "%u, unlock_action = %u\n",
2342 lockres->l_name, lockres->l_flags, lockres->l_action,
2343 lockres->l_unlock_action);
2344
2345 spin_unlock_irqrestore(&lockres->l_lock, flags);
2346
2347 /* XXX: Today we just wait on any busy
2348 * locks... Perhaps we need to cancel converts in the
2349 * future? */
2350 ocfs2_wait_on_busy_lock(lockres);
2351
2352 spin_lock_irqsave(&lockres->l_lock, flags);
2353 }
2354
Mark Fasheh0d5dc6c2006-09-14 14:44:51 -07002355 if (lockres->l_ops->flags & LOCK_TYPE_USES_LVB) {
2356 if (lockres->l_flags & OCFS2_LOCK_ATTACHED &&
2357 lockres->l_level == LKM_EXMODE &&
2358 !(lockres->l_flags & OCFS2_LOCK_NEEDS_REFRESH))
2359 lockres->l_ops->set_lvb(lockres);
2360 }
Mark Fashehccd979b2005-12-15 14:31:24 -08002361
2362 if (lockres->l_flags & OCFS2_LOCK_BUSY)
2363 mlog(ML_ERROR, "destroying busy lock: \"%s\"\n",
2364 lockres->l_name);
2365 if (lockres->l_flags & OCFS2_LOCK_BLOCKED)
2366 mlog(0, "destroying blocked lock: \"%s\"\n", lockres->l_name);
2367
2368 if (!(lockres->l_flags & OCFS2_LOCK_ATTACHED)) {
2369 spin_unlock_irqrestore(&lockres->l_lock, flags);
2370 goto out;
2371 }
2372
2373 lockres_clear_flags(lockres, OCFS2_LOCK_ATTACHED);
2374
2375 /* make sure we never get here while waiting for an ast to
2376 * fire. */
2377 BUG_ON(lockres->l_action != OCFS2_AST_INVALID);
2378
2379 /* is this necessary? */
2380 lockres_or_flags(lockres, OCFS2_LOCK_BUSY);
2381 lockres->l_unlock_action = OCFS2_UNLOCK_DROP_LOCK;
2382 spin_unlock_irqrestore(&lockres->l_lock, flags);
2383
2384 mlog(0, "lock %s\n", lockres->l_name);
2385
Mark Fashehb80fc012006-09-12 22:08:14 -07002386 status = dlmunlock(osb->dlm, &lockres->l_lksb, lkm_flags,
Mark Fasheh2a45f2d2006-09-12 21:36:58 -07002387 ocfs2_unlock_ast, lockres);
Mark Fashehccd979b2005-12-15 14:31:24 -08002388 if (status != DLM_NORMAL) {
2389 ocfs2_log_dlm_error("dlmunlock", status, lockres);
2390 mlog(ML_ERROR, "lockres flags: %lu\n", lockres->l_flags);
2391 dlm_print_one_lock(lockres->l_lksb.lockid);
2392 BUG();
2393 }
2394 mlog(0, "lock %s, successfull return from dlmunlock\n",
2395 lockres->l_name);
2396
2397 ocfs2_wait_on_busy_lock(lockres);
2398out:
2399 mlog_exit(0);
2400 return 0;
2401}
2402
2403/* Mark the lockres as being dropped. It will no longer be
2404 * queued if blocking, but we still may have to wait on it
2405 * being dequeued from the vote thread before we can consider
2406 * it safe to drop.
2407 *
2408 * You can *not* attempt to call cluster_lock on this lockres anymore. */
2409void ocfs2_mark_lockres_freeing(struct ocfs2_lock_res *lockres)
2410{
2411 int status;
2412 struct ocfs2_mask_waiter mw;
2413 unsigned long flags;
2414
2415 ocfs2_init_mask_waiter(&mw);
2416
2417 spin_lock_irqsave(&lockres->l_lock, flags);
2418 lockres->l_flags |= OCFS2_LOCK_FREEING;
2419 while (lockres->l_flags & OCFS2_LOCK_QUEUED) {
2420 lockres_add_mask_waiter(lockres, &mw, OCFS2_LOCK_QUEUED, 0);
2421 spin_unlock_irqrestore(&lockres->l_lock, flags);
2422
2423 mlog(0, "Waiting on lockres %s\n", lockres->l_name);
2424
2425 status = ocfs2_wait_for_mask(&mw);
2426 if (status)
2427 mlog_errno(status);
2428
2429 spin_lock_irqsave(&lockres->l_lock, flags);
2430 }
2431 spin_unlock_irqrestore(&lockres->l_lock, flags);
2432}
2433
Mark Fashehd680efe2006-09-08 14:14:34 -07002434void ocfs2_simple_drop_lockres(struct ocfs2_super *osb,
2435 struct ocfs2_lock_res *lockres)
2436{
2437 int ret;
2438
2439 ocfs2_mark_lockres_freeing(lockres);
Mark Fasheh0d5dc6c2006-09-14 14:44:51 -07002440 ret = ocfs2_drop_lock(osb, lockres);
Mark Fashehd680efe2006-09-08 14:14:34 -07002441 if (ret)
2442 mlog_errno(ret);
2443}
2444
Mark Fashehccd979b2005-12-15 14:31:24 -08002445static void ocfs2_drop_osb_locks(struct ocfs2_super *osb)
2446{
Mark Fashehd680efe2006-09-08 14:14:34 -07002447 ocfs2_simple_drop_lockres(osb, &osb->osb_super_lockres);
2448 ocfs2_simple_drop_lockres(osb, &osb->osb_rename_lockres);
Mark Fashehccd979b2005-12-15 14:31:24 -08002449}
2450
Mark Fashehccd979b2005-12-15 14:31:24 -08002451int ocfs2_drop_inode_locks(struct inode *inode)
2452{
2453 int status, err;
Mark Fashehccd979b2005-12-15 14:31:24 -08002454
2455 mlog_entry_void();
2456
2457 /* No need to call ocfs2_mark_lockres_freeing here -
2458 * ocfs2_clear_inode has done it for us. */
2459
2460 err = ocfs2_drop_lock(OCFS2_SB(inode->i_sb),
Mark Fasheh0d5dc6c2006-09-14 14:44:51 -07002461 &OCFS2_I(inode)->ip_data_lockres);
Mark Fashehccd979b2005-12-15 14:31:24 -08002462 if (err < 0)
2463 mlog_errno(err);
2464
2465 status = err;
2466
2467 err = ocfs2_drop_lock(OCFS2_SB(inode->i_sb),
Mark Fasheh0d5dc6c2006-09-14 14:44:51 -07002468 &OCFS2_I(inode)->ip_meta_lockres);
Mark Fashehccd979b2005-12-15 14:31:24 -08002469 if (err < 0)
2470 mlog_errno(err);
2471 if (err < 0 && !status)
2472 status = err;
2473
2474 err = ocfs2_drop_lock(OCFS2_SB(inode->i_sb),
Mark Fasheh0d5dc6c2006-09-14 14:44:51 -07002475 &OCFS2_I(inode)->ip_rw_lockres);
Mark Fashehccd979b2005-12-15 14:31:24 -08002476 if (err < 0)
2477 mlog_errno(err);
2478 if (err < 0 && !status)
2479 status = err;
2480
2481 mlog_exit(status);
2482 return status;
2483}
2484
2485static void ocfs2_prepare_downconvert(struct ocfs2_lock_res *lockres,
2486 int new_level)
2487{
2488 assert_spin_locked(&lockres->l_lock);
2489
2490 BUG_ON(lockres->l_blocking <= LKM_NLMODE);
2491
2492 if (lockres->l_level <= new_level) {
2493 mlog(ML_ERROR, "lockres->l_level (%u) <= new_level (%u)\n",
2494 lockres->l_level, new_level);
2495 BUG();
2496 }
2497
2498 mlog(0, "lock %s, new_level = %d, l_blocking = %d\n",
2499 lockres->l_name, new_level, lockres->l_blocking);
2500
2501 lockres->l_action = OCFS2_AST_DOWNCONVERT;
2502 lockres->l_requested = new_level;
2503 lockres_or_flags(lockres, OCFS2_LOCK_BUSY);
2504}
2505
2506static int ocfs2_downconvert_lock(struct ocfs2_super *osb,
2507 struct ocfs2_lock_res *lockres,
2508 int new_level,
2509 int lvb)
2510{
2511 int ret, dlm_flags = LKM_CONVERT;
2512 enum dlm_status status;
2513
2514 mlog_entry_void();
2515
2516 if (lvb)
2517 dlm_flags |= LKM_VALBLK;
2518
2519 status = dlmlock(osb->dlm,
2520 new_level,
2521 &lockres->l_lksb,
2522 dlm_flags,
2523 lockres->l_name,
Mark Fashehf0681062006-09-08 11:40:10 -07002524 OCFS2_LOCK_ID_MAX_LEN - 1,
Mark Fashehe92d57d2006-09-12 21:34:35 -07002525 ocfs2_locking_ast,
Mark Fashehccd979b2005-12-15 14:31:24 -08002526 lockres,
Mark Fashehaa2623a2006-09-12 21:58:23 -07002527 ocfs2_blocking_ast);
Mark Fashehccd979b2005-12-15 14:31:24 -08002528 if (status != DLM_NORMAL) {
2529 ocfs2_log_dlm_error("dlmlock", status, lockres);
2530 ret = -EINVAL;
2531 ocfs2_recover_from_dlm_error(lockres, 1);
2532 goto bail;
2533 }
2534
2535 ret = 0;
2536bail:
2537 mlog_exit(ret);
2538 return ret;
2539}
2540
2541/* returns 1 when the caller should unlock and call dlmunlock */
2542static int ocfs2_prepare_cancel_convert(struct ocfs2_super *osb,
2543 struct ocfs2_lock_res *lockres)
2544{
2545 assert_spin_locked(&lockres->l_lock);
2546
2547 mlog_entry_void();
2548 mlog(0, "lock %s\n", lockres->l_name);
2549
2550 if (lockres->l_unlock_action == OCFS2_UNLOCK_CANCEL_CONVERT) {
2551 /* If we're already trying to cancel a lock conversion
2552 * then just drop the spinlock and allow the caller to
2553 * requeue this lock. */
2554
2555 mlog(0, "Lockres %s, skip convert\n", lockres->l_name);
2556 return 0;
2557 }
2558
2559 /* were we in a convert when we got the bast fire? */
2560 BUG_ON(lockres->l_action != OCFS2_AST_CONVERT &&
2561 lockres->l_action != OCFS2_AST_DOWNCONVERT);
2562 /* set things up for the unlockast to know to just
2563 * clear out the ast_action and unset busy, etc. */
2564 lockres->l_unlock_action = OCFS2_UNLOCK_CANCEL_CONVERT;
2565
2566 mlog_bug_on_msg(!(lockres->l_flags & OCFS2_LOCK_BUSY),
2567 "lock %s, invalid flags: 0x%lx\n",
2568 lockres->l_name, lockres->l_flags);
2569
2570 return 1;
2571}
2572
2573static int ocfs2_cancel_convert(struct ocfs2_super *osb,
2574 struct ocfs2_lock_res *lockres)
2575{
2576 int ret;
2577 enum dlm_status status;
2578
2579 mlog_entry_void();
2580 mlog(0, "lock %s\n", lockres->l_name);
2581
2582 ret = 0;
2583 status = dlmunlock(osb->dlm,
2584 &lockres->l_lksb,
2585 LKM_CANCEL,
Mark Fasheh2a45f2d2006-09-12 21:36:58 -07002586 ocfs2_unlock_ast,
Mark Fashehccd979b2005-12-15 14:31:24 -08002587 lockres);
2588 if (status != DLM_NORMAL) {
2589 ocfs2_log_dlm_error("dlmunlock", status, lockres);
2590 ret = -EINVAL;
2591 ocfs2_recover_from_dlm_error(lockres, 0);
2592 }
2593
2594 mlog(0, "lock %s return from dlmunlock\n", lockres->l_name);
2595
2596 mlog_exit(ret);
2597 return ret;
2598}
2599
Mark Fashehb5e500e2006-09-13 22:01:16 -07002600static int ocfs2_unblock_lock(struct ocfs2_super *osb,
2601 struct ocfs2_lock_res *lockres,
2602 struct ocfs2_unblock_ctl *ctl)
Mark Fashehccd979b2005-12-15 14:31:24 -08002603{
2604 unsigned long flags;
2605 int blocking;
2606 int new_level;
2607 int ret = 0;
Mark Fasheh5ef0d4e2006-09-13 21:21:52 -07002608 int set_lvb = 0;
Mark Fashehccd979b2005-12-15 14:31:24 -08002609
2610 mlog_entry_void();
2611
2612 spin_lock_irqsave(&lockres->l_lock, flags);
2613
2614 BUG_ON(!(lockres->l_flags & OCFS2_LOCK_BLOCKED));
2615
2616recheck:
2617 if (lockres->l_flags & OCFS2_LOCK_BUSY) {
Mark Fashehd680efe2006-09-08 14:14:34 -07002618 ctl->requeue = 1;
Mark Fashehccd979b2005-12-15 14:31:24 -08002619 ret = ocfs2_prepare_cancel_convert(osb, lockres);
2620 spin_unlock_irqrestore(&lockres->l_lock, flags);
2621 if (ret) {
2622 ret = ocfs2_cancel_convert(osb, lockres);
2623 if (ret < 0)
2624 mlog_errno(ret);
2625 }
2626 goto leave;
2627 }
2628
2629 /* if we're blocking an exclusive and we have *any* holders,
2630 * then requeue. */
2631 if ((lockres->l_blocking == LKM_EXMODE)
Mark Fashehf7fbfdd2006-09-13 21:02:29 -07002632 && (lockres->l_ex_holders || lockres->l_ro_holders))
2633 goto leave_requeue;
Mark Fashehccd979b2005-12-15 14:31:24 -08002634
2635 /* If it's a PR we're blocking, then only
2636 * requeue if we've got any EX holders */
2637 if (lockres->l_blocking == LKM_PRMODE &&
Mark Fashehf7fbfdd2006-09-13 21:02:29 -07002638 lockres->l_ex_holders)
2639 goto leave_requeue;
2640
2641 /*
2642 * Can we get a lock in this state if the holder counts are
2643 * zero? The meta data unblock code used to check this.
2644 */
2645 if ((lockres->l_ops->flags & LOCK_TYPE_REQUIRES_REFRESH)
2646 && (lockres->l_flags & OCFS2_LOCK_REFRESHING))
2647 goto leave_requeue;
Mark Fashehccd979b2005-12-15 14:31:24 -08002648
Mark Fasheh16d5b952006-09-13 21:10:12 -07002649 new_level = ocfs2_highest_compat_lock_level(lockres->l_blocking);
2650
2651 if (lockres->l_ops->check_downconvert
2652 && !lockres->l_ops->check_downconvert(lockres, new_level))
2653 goto leave_requeue;
2654
Mark Fashehccd979b2005-12-15 14:31:24 -08002655 /* If we get here, then we know that there are no more
2656 * incompatible holders (and anyone asking for an incompatible
2657 * lock is blocked). We can now downconvert the lock */
Mark Fashehcc567d82006-09-13 21:52:21 -07002658 if (!lockres->l_ops->downconvert_worker)
Mark Fashehccd979b2005-12-15 14:31:24 -08002659 goto downconvert;
2660
2661 /* Some lockres types want to do a bit of work before
2662 * downconverting a lock. Allow that here. The worker function
2663 * may sleep, so we save off a copy of what we're blocking as
2664 * it may change while we're not holding the spin lock. */
2665 blocking = lockres->l_blocking;
2666 spin_unlock_irqrestore(&lockres->l_lock, flags);
2667
Mark Fashehcc567d82006-09-13 21:52:21 -07002668 ctl->unblock_action = lockres->l_ops->downconvert_worker(lockres, blocking);
Mark Fashehd680efe2006-09-08 14:14:34 -07002669
2670 if (ctl->unblock_action == UNBLOCK_STOP_POST)
2671 goto leave;
Mark Fashehccd979b2005-12-15 14:31:24 -08002672
2673 spin_lock_irqsave(&lockres->l_lock, flags);
2674 if (blocking != lockres->l_blocking) {
2675 /* If this changed underneath us, then we can't drop
2676 * it just yet. */
2677 goto recheck;
2678 }
2679
2680downconvert:
Mark Fashehd680efe2006-09-08 14:14:34 -07002681 ctl->requeue = 0;
Mark Fashehccd979b2005-12-15 14:31:24 -08002682
Mark Fasheh5ef0d4e2006-09-13 21:21:52 -07002683 if (lockres->l_ops->flags & LOCK_TYPE_USES_LVB) {
2684 if (lockres->l_level == LKM_EXMODE)
2685 set_lvb = 1;
2686
2687 /*
2688 * We only set the lvb if the lock has been fully
2689 * refreshed - otherwise we risk setting stale
2690 * data. Otherwise, there's no need to actually clear
2691 * out the lvb here as it's value is still valid.
2692 */
2693 if (set_lvb && !(lockres->l_flags & OCFS2_LOCK_NEEDS_REFRESH))
2694 lockres->l_ops->set_lvb(lockres);
2695 }
2696
Mark Fashehccd979b2005-12-15 14:31:24 -08002697 ocfs2_prepare_downconvert(lockres, new_level);
2698 spin_unlock_irqrestore(&lockres->l_lock, flags);
Mark Fasheh5ef0d4e2006-09-13 21:21:52 -07002699 ret = ocfs2_downconvert_lock(osb, lockres, new_level, set_lvb);
Mark Fashehccd979b2005-12-15 14:31:24 -08002700leave:
2701 mlog_exit(ret);
2702 return ret;
Mark Fashehf7fbfdd2006-09-13 21:02:29 -07002703
2704leave_requeue:
2705 spin_unlock_irqrestore(&lockres->l_lock, flags);
2706 ctl->requeue = 1;
2707
2708 mlog_exit(0);
2709 return 0;
Mark Fashehccd979b2005-12-15 14:31:24 -08002710}
2711
Mark Fashehd680efe2006-09-08 14:14:34 -07002712static int ocfs2_data_convert_worker(struct ocfs2_lock_res *lockres,
2713 int blocking)
Mark Fashehccd979b2005-12-15 14:31:24 -08002714{
2715 struct inode *inode;
2716 struct address_space *mapping;
2717
Mark Fashehccd979b2005-12-15 14:31:24 -08002718 inode = ocfs2_lock_res_inode(lockres);
2719 mapping = inode->i_mapping;
2720
Mark Fasheh7f4a2a92006-12-11 11:06:36 -08002721 /*
2722 * We need this before the filemap_fdatawrite() so that it can
2723 * transfer the dirty bit from the PTE to the
2724 * page. Unfortunately this means that even for EX->PR
2725 * downconverts, we'll lose our mappings and have to build
2726 * them up again.
2727 */
2728 unmap_mapping_range(mapping, 0, 0, 0);
2729
Mark Fashehccd979b2005-12-15 14:31:24 -08002730 if (filemap_fdatawrite(mapping)) {
Mark Fashehb06970532006-03-03 10:24:33 -08002731 mlog(ML_ERROR, "Could not sync inode %llu for downconvert!",
2732 (unsigned long long)OCFS2_I(inode)->ip_blkno);
Mark Fashehccd979b2005-12-15 14:31:24 -08002733 }
2734 sync_mapping_buffers(mapping);
2735 if (blocking == LKM_EXMODE) {
2736 truncate_inode_pages(mapping, 0);
Mark Fashehccd979b2005-12-15 14:31:24 -08002737 } else {
2738 /* We only need to wait on the I/O if we're not also
2739 * truncating pages because truncate_inode_pages waits
2740 * for us above. We don't truncate pages if we're
2741 * blocking anything < EXMODE because we want to keep
2742 * them around in that case. */
2743 filemap_fdatawait(mapping);
2744 }
2745
Mark Fashehd680efe2006-09-08 14:14:34 -07002746 return UNBLOCK_CONTINUE;
Mark Fashehccd979b2005-12-15 14:31:24 -08002747}
2748
Mark Fasheh810d5ae2006-09-13 21:39:52 -07002749static int ocfs2_check_meta_downconvert(struct ocfs2_lock_res *lockres,
2750 int new_level)
2751{
2752 struct inode *inode = ocfs2_lock_res_inode(lockres);
2753 int checkpointed = ocfs2_inode_fully_checkpointed(inode);
2754
2755 BUG_ON(new_level != LKM_NLMODE && new_level != LKM_PRMODE);
2756 BUG_ON(lockres->l_level != LKM_EXMODE && !checkpointed);
2757
2758 if (checkpointed)
2759 return 1;
2760
2761 ocfs2_start_checkpoint(OCFS2_SB(inode->i_sb));
2762 return 0;
2763}
2764
2765static void ocfs2_set_meta_lvb(struct ocfs2_lock_res *lockres)
2766{
2767 struct inode *inode = ocfs2_lock_res_inode(lockres);
2768
2769 __ocfs2_stuff_meta_lvb(inode);
2770}
2771
Mark Fashehd680efe2006-09-08 14:14:34 -07002772/*
2773 * Does the final reference drop on our dentry lock. Right now this
2774 * happens in the vote thread, but we could choose to simplify the
2775 * dlmglue API and push these off to the ocfs2_wq in the future.
2776 */
2777static void ocfs2_dentry_post_unlock(struct ocfs2_super *osb,
2778 struct ocfs2_lock_res *lockres)
2779{
2780 struct ocfs2_dentry_lock *dl = ocfs2_lock_res_dl(lockres);
2781 ocfs2_dentry_lock_put(osb, dl);
2782}
2783
2784/*
2785 * d_delete() matching dentries before the lock downconvert.
2786 *
2787 * At this point, any process waiting to destroy the
2788 * dentry_lock due to last ref count is stopped by the
2789 * OCFS2_LOCK_QUEUED flag.
2790 *
2791 * We have two potential problems
2792 *
2793 * 1) If we do the last reference drop on our dentry_lock (via dput)
2794 * we'll wind up in ocfs2_release_dentry_lock(), waiting on
2795 * the downconvert to finish. Instead we take an elevated
2796 * reference and push the drop until after we've completed our
2797 * unblock processing.
2798 *
2799 * 2) There might be another process with a final reference,
2800 * waiting on us to finish processing. If this is the case, we
2801 * detect it and exit out - there's no more dentries anyway.
2802 */
2803static int ocfs2_dentry_convert_worker(struct ocfs2_lock_res *lockres,
2804 int blocking)
2805{
2806 struct ocfs2_dentry_lock *dl = ocfs2_lock_res_dl(lockres);
2807 struct ocfs2_inode_info *oi = OCFS2_I(dl->dl_inode);
2808 struct dentry *dentry;
2809 unsigned long flags;
2810 int extra_ref = 0;
2811
2812 /*
2813 * This node is blocking another node from getting a read
2814 * lock. This happens when we've renamed within a
2815 * directory. We've forced the other nodes to d_delete(), but
2816 * we never actually dropped our lock because it's still
2817 * valid. The downconvert code will retain a PR for this node,
2818 * so there's no further work to do.
2819 */
2820 if (blocking == LKM_PRMODE)
2821 return UNBLOCK_CONTINUE;
2822
2823 /*
2824 * Mark this inode as potentially orphaned. The code in
2825 * ocfs2_delete_inode() will figure out whether it actually
2826 * needs to be freed or not.
2827 */
2828 spin_lock(&oi->ip_lock);
2829 oi->ip_flags |= OCFS2_INODE_MAYBE_ORPHANED;
2830 spin_unlock(&oi->ip_lock);
2831
2832 /*
2833 * Yuck. We need to make sure however that the check of
2834 * OCFS2_LOCK_FREEING and the extra reference are atomic with
2835 * respect to a reference decrement or the setting of that
2836 * flag.
2837 */
2838 spin_lock_irqsave(&lockres->l_lock, flags);
2839 spin_lock(&dentry_attach_lock);
2840 if (!(lockres->l_flags & OCFS2_LOCK_FREEING)
2841 && dl->dl_count) {
2842 dl->dl_count++;
2843 extra_ref = 1;
2844 }
2845 spin_unlock(&dentry_attach_lock);
2846 spin_unlock_irqrestore(&lockres->l_lock, flags);
2847
2848 mlog(0, "extra_ref = %d\n", extra_ref);
2849
2850 /*
2851 * We have a process waiting on us in ocfs2_dentry_iput(),
2852 * which means we can't have any more outstanding
2853 * aliases. There's no need to do any more work.
2854 */
2855 if (!extra_ref)
2856 return UNBLOCK_CONTINUE;
2857
2858 spin_lock(&dentry_attach_lock);
2859 while (1) {
2860 dentry = ocfs2_find_local_alias(dl->dl_inode,
2861 dl->dl_parent_blkno, 1);
2862 if (!dentry)
2863 break;
2864 spin_unlock(&dentry_attach_lock);
2865
2866 mlog(0, "d_delete(%.*s);\n", dentry->d_name.len,
2867 dentry->d_name.name);
2868
2869 /*
2870 * The following dcache calls may do an
2871 * iput(). Normally we don't want that from the
2872 * downconverting thread, but in this case it's ok
2873 * because the requesting node already has an
2874 * exclusive lock on the inode, so it can't be queued
2875 * for a downconvert.
2876 */
2877 d_delete(dentry);
2878 dput(dentry);
2879
2880 spin_lock(&dentry_attach_lock);
2881 }
2882 spin_unlock(&dentry_attach_lock);
2883
2884 /*
2885 * If we are the last holder of this dentry lock, there is no
2886 * reason to downconvert so skip straight to the unlock.
2887 */
2888 if (dl->dl_count == 1)
2889 return UNBLOCK_STOP_POST;
2890
2891 return UNBLOCK_CONTINUE_POST;
2892}
2893
Mark Fashehccd979b2005-12-15 14:31:24 -08002894void ocfs2_process_blocked_lock(struct ocfs2_super *osb,
2895 struct ocfs2_lock_res *lockres)
2896{
2897 int status;
Mark Fashehd680efe2006-09-08 14:14:34 -07002898 struct ocfs2_unblock_ctl ctl = {0, 0,};
Mark Fashehccd979b2005-12-15 14:31:24 -08002899 unsigned long flags;
2900
2901 /* Our reference to the lockres in this function can be
2902 * considered valid until we remove the OCFS2_LOCK_QUEUED
2903 * flag. */
2904
2905 mlog_entry_void();
2906
2907 BUG_ON(!lockres);
2908 BUG_ON(!lockres->l_ops);
Mark Fashehccd979b2005-12-15 14:31:24 -08002909
2910 mlog(0, "lockres %s blocked.\n", lockres->l_name);
2911
2912 /* Detect whether a lock has been marked as going away while
2913 * the vote thread was processing other things. A lock can
2914 * still be marked with OCFS2_LOCK_FREEING after this check,
2915 * but short circuiting here will still save us some
2916 * performance. */
2917 spin_lock_irqsave(&lockres->l_lock, flags);
2918 if (lockres->l_flags & OCFS2_LOCK_FREEING)
2919 goto unqueue;
2920 spin_unlock_irqrestore(&lockres->l_lock, flags);
2921
Mark Fashehb5e500e2006-09-13 22:01:16 -07002922 status = ocfs2_unblock_lock(osb, lockres, &ctl);
Mark Fashehccd979b2005-12-15 14:31:24 -08002923 if (status < 0)
2924 mlog_errno(status);
2925
2926 spin_lock_irqsave(&lockres->l_lock, flags);
2927unqueue:
Mark Fashehd680efe2006-09-08 14:14:34 -07002928 if (lockres->l_flags & OCFS2_LOCK_FREEING || !ctl.requeue) {
Mark Fashehccd979b2005-12-15 14:31:24 -08002929 lockres_clear_flags(lockres, OCFS2_LOCK_QUEUED);
2930 } else
2931 ocfs2_schedule_blocked_lock(osb, lockres);
2932
2933 mlog(0, "lockres %s, requeue = %s.\n", lockres->l_name,
Mark Fashehd680efe2006-09-08 14:14:34 -07002934 ctl.requeue ? "yes" : "no");
Mark Fashehccd979b2005-12-15 14:31:24 -08002935 spin_unlock_irqrestore(&lockres->l_lock, flags);
2936
Mark Fashehd680efe2006-09-08 14:14:34 -07002937 if (ctl.unblock_action != UNBLOCK_CONTINUE
2938 && lockres->l_ops->post_unlock)
2939 lockres->l_ops->post_unlock(osb, lockres);
2940
Mark Fashehccd979b2005-12-15 14:31:24 -08002941 mlog_exit_void();
2942}
2943
2944static void ocfs2_schedule_blocked_lock(struct ocfs2_super *osb,
2945 struct ocfs2_lock_res *lockres)
2946{
2947 mlog_entry_void();
2948
2949 assert_spin_locked(&lockres->l_lock);
2950
2951 if (lockres->l_flags & OCFS2_LOCK_FREEING) {
2952 /* Do not schedule a lock for downconvert when it's on
2953 * the way to destruction - any nodes wanting access
2954 * to the resource will get it soon. */
2955 mlog(0, "Lockres %s won't be scheduled: flags 0x%lx\n",
2956 lockres->l_name, lockres->l_flags);
2957 return;
2958 }
2959
2960 lockres_or_flags(lockres, OCFS2_LOCK_QUEUED);
2961
2962 spin_lock(&osb->vote_task_lock);
2963 if (list_empty(&lockres->l_blocked_list)) {
2964 list_add_tail(&lockres->l_blocked_list,
2965 &osb->blocked_lock_list);
2966 osb->blocked_lock_count++;
2967 }
2968 spin_unlock(&osb->vote_task_lock);
2969
2970 mlog_exit_void();
2971}
2972
2973/* This aids in debugging situations where a bad LVB might be involved. */
2974void ocfs2_dump_meta_lvb_info(u64 level,
2975 const char *function,
2976 unsigned int line,
2977 struct ocfs2_lock_res *lockres)
2978{
2979 struct ocfs2_meta_lvb *lvb = (struct ocfs2_meta_lvb *) lockres->l_lksb.lvb;
2980
2981 mlog(level, "LVB information for %s (called from %s:%u):\n",
2982 lockres->l_name, function, line);
Mark Fashehf9e2d822006-09-12 15:35:49 -07002983 mlog(level, "version: %u, clusters: %u, generation: 0x%x\n",
2984 lvb->lvb_version, be32_to_cpu(lvb->lvb_iclusters),
2985 be32_to_cpu(lvb->lvb_igeneration));
Mark Fashehb06970532006-03-03 10:24:33 -08002986 mlog(level, "size: %llu, uid %u, gid %u, mode 0x%x\n",
2987 (unsigned long long)be64_to_cpu(lvb->lvb_isize),
2988 be32_to_cpu(lvb->lvb_iuid), be32_to_cpu(lvb->lvb_igid),
2989 be16_to_cpu(lvb->lvb_imode));
2990 mlog(level, "nlink %u, atime_packed 0x%llx, ctime_packed 0x%llx, "
Herbert Poetzlca4d1472006-07-03 17:27:12 -07002991 "mtime_packed 0x%llx iattr 0x%x\n", be16_to_cpu(lvb->lvb_inlink),
Mark Fashehb06970532006-03-03 10:24:33 -08002992 (long long)be64_to_cpu(lvb->lvb_iatime_packed),
2993 (long long)be64_to_cpu(lvb->lvb_ictime_packed),
Herbert Poetzlca4d1472006-07-03 17:27:12 -07002994 (long long)be64_to_cpu(lvb->lvb_imtime_packed),
2995 be32_to_cpu(lvb->lvb_iattr));
Mark Fashehccd979b2005-12-15 14:31:24 -08002996}