Add the rt linux 4.1.3-rt3 as base
[kvmfornfv.git] / kernel / drivers / staging / lustre / lustre / lov / lov_lock.c
1 /*
2  * GPL HEADER START
3  *
4  * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
5  *
6  * This program is free software; you can redistribute it and/or modify
7  * it under the terms of the GNU General Public License version 2 only,
8  * as published by the Free Software Foundation.
9  *
10  * This program is distributed in the hope that it will be useful, but
11  * WITHOUT ANY WARRANTY; without even the implied warranty of
12  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
13  * General Public License version 2 for more details (a copy is included
14  * in the LICENSE file that accompanied this code).
15  *
16  * You should have received a copy of the GNU General Public License
17  * version 2 along with this program; If not, see
18  * http://www.sun.com/software/products/lustre/docs/GPLv2.pdf
19  *
20  * Please contact Sun Microsystems, Inc., 4150 Network Circle, Santa Clara,
21  * CA 95054 USA or visit www.sun.com if you need additional information or
22  * have any questions.
23  *
24  * GPL HEADER END
25  */
26 /*
27  * Copyright (c) 2008, 2010, Oracle and/or its affiliates. All rights reserved.
28  * Use is subject to license terms.
29  *
30  * Copyright (c) 2011, 2012, Intel Corporation.
31  */
32 /*
33  * This file is part of Lustre, http://www.lustre.org/
34  * Lustre is a trademark of Sun Microsystems, Inc.
35  *
36  * Implementation of cl_lock for LOV layer.
37  *
38  *   Author: Nikita Danilov <nikita.danilov@sun.com>
39  */
40
41 #define DEBUG_SUBSYSTEM S_LOV
42
43 #include "lov_cl_internal.h"
44
45 /** \addtogroup lov
46  *  @{
47  */
48
49 static struct cl_lock_closure *lov_closure_get(const struct lu_env *env,
50                                                struct cl_lock *parent);
51
52 static int lov_lock_unuse(const struct lu_env *env,
53                           const struct cl_lock_slice *slice);
54 /*****************************************************************************
55  *
56  * Lov lock operations.
57  *
58  */
59
60 static struct lov_sublock_env *lov_sublock_env_get(const struct lu_env *env,
61                                                    struct cl_lock *parent,
62                                                    struct lov_lock_sub *lls)
63 {
64         struct lov_sublock_env *subenv;
65         struct lov_io     *lio    = lov_env_io(env);
66         struct cl_io       *io     = lio->lis_cl.cis_io;
67         struct lov_io_sub      *sub;
68
69         subenv = &lov_env_session(env)->ls_subenv;
70
71         /*
72          * FIXME: We tend to use the subio's env & io to call the sublock
73          * lock operations because osc lock sometimes stores some control
74          * variables in thread's IO information(Now only lockless information).
75          * However, if the lock's host(object) is different from the object
76          * for current IO, we have no way to get the subenv and subio because
77          * they are not initialized at all. As a temp fix, in this case,
78          * we still borrow the parent's env to call sublock operations.
79          */
80         if (!io || !cl_object_same(io->ci_obj, parent->cll_descr.cld_obj)) {
81                 subenv->lse_env = env;
82                 subenv->lse_io  = io;
83                 subenv->lse_sub = NULL;
84         } else {
85                 sub = lov_sub_get(env, lio, lls->sub_stripe);
86                 if (!IS_ERR(sub)) {
87                         subenv->lse_env = sub->sub_env;
88                         subenv->lse_io  = sub->sub_io;
89                         subenv->lse_sub = sub;
90                 } else {
91                         subenv = (void *)sub;
92                 }
93         }
94         return subenv;
95 }
96
97 static void lov_sublock_env_put(struct lov_sublock_env *subenv)
98 {
99         if (subenv && subenv->lse_sub)
100                 lov_sub_put(subenv->lse_sub);
101 }
102
103 static void lov_sublock_adopt(const struct lu_env *env, struct lov_lock *lck,
104                               struct cl_lock *sublock, int idx,
105                               struct lov_lock_link *link)
106 {
107         struct lovsub_lock *lsl;
108         struct cl_lock     *parent = lck->lls_cl.cls_lock;
109         int              rc;
110
111         LASSERT(cl_lock_is_mutexed(parent));
112         LASSERT(cl_lock_is_mutexed(sublock));
113
114         lsl = cl2sub_lock(sublock);
115         /*
116          * check that sub-lock doesn't have lock link to this top-lock.
117          */
118         LASSERT(lov_lock_link_find(env, lck, lsl) == NULL);
119         LASSERT(idx < lck->lls_nr);
120
121         lck->lls_sub[idx].sub_lock = lsl;
122         lck->lls_nr_filled++;
123         LASSERT(lck->lls_nr_filled <= lck->lls_nr);
124         list_add_tail(&link->lll_list, &lsl->lss_parents);
125         link->lll_idx = idx;
126         link->lll_super = lck;
127         cl_lock_get(parent);
128         lu_ref_add(&parent->cll_reference, "lov-child", sublock);
129         lck->lls_sub[idx].sub_flags |= LSF_HELD;
130         cl_lock_user_add(env, sublock);
131
132         rc = lov_sublock_modify(env, lck, lsl, &sublock->cll_descr, idx);
133         LASSERT(rc == 0); /* there is no way this can fail, currently */
134 }
135
136 static struct cl_lock *lov_sublock_alloc(const struct lu_env *env,
137                                          const struct cl_io *io,
138                                          struct lov_lock *lck,
139                                          int idx, struct lov_lock_link **out)
140 {
141         struct cl_lock       *sublock;
142         struct cl_lock       *parent;
143         struct lov_lock_link *link;
144
145         LASSERT(idx < lck->lls_nr);
146
147         OBD_SLAB_ALLOC_PTR_GFP(link, lov_lock_link_kmem, GFP_NOFS);
148         if (link != NULL) {
149                 struct lov_sublock_env *subenv;
150                 struct lov_lock_sub  *lls;
151                 struct cl_lock_descr *descr;
152
153                 parent = lck->lls_cl.cls_lock;
154                 lls    = &lck->lls_sub[idx];
155                 descr  = &lls->sub_got;
156
157                 subenv = lov_sublock_env_get(env, parent, lls);
158                 if (!IS_ERR(subenv)) {
159                         /* CAVEAT: Don't try to add a field in lov_lock_sub
160                          * to remember the subio. This is because lock is able
161                          * to be cached, but this is not true for IO. This
162                          * further means a sublock might be referenced in
163                          * different io context. -jay */
164
165                         sublock = cl_lock_hold(subenv->lse_env, subenv->lse_io,
166                                                descr, "lov-parent", parent);
167                         lov_sublock_env_put(subenv);
168                 } else {
169                         /* error occurs. */
170                         sublock = (void *)subenv;
171                 }
172
173                 if (!IS_ERR(sublock))
174                         *out = link;
175                 else
176                         OBD_SLAB_FREE_PTR(link, lov_lock_link_kmem);
177         } else
178                 sublock = ERR_PTR(-ENOMEM);
179         return sublock;
180 }
181
182 static void lov_sublock_unlock(const struct lu_env *env,
183                                struct lovsub_lock *lsl,
184                                struct cl_lock_closure *closure,
185                                struct lov_sublock_env *subenv)
186 {
187         lov_sublock_env_put(subenv);
188         lsl->lss_active = NULL;
189         cl_lock_disclosure(env, closure);
190 }
191
192 static int lov_sublock_lock(const struct lu_env *env,
193                             struct lov_lock *lck,
194                             struct lov_lock_sub *lls,
195                             struct cl_lock_closure *closure,
196                             struct lov_sublock_env **lsep)
197 {
198         struct lovsub_lock *sublock;
199         struct cl_lock     *child;
200         int              result = 0;
201
202         LASSERT(list_empty(&closure->clc_list));
203
204         sublock = lls->sub_lock;
205         child = sublock->lss_cl.cls_lock;
206         result = cl_lock_closure_build(env, child, closure);
207         if (result == 0) {
208                 struct cl_lock *parent = closure->clc_origin;
209
210                 LASSERT(cl_lock_is_mutexed(child));
211                 sublock->lss_active = parent;
212
213                 if (unlikely((child->cll_state == CLS_FREEING) ||
214                              (child->cll_flags & CLF_CANCELLED))) {
215                         struct lov_lock_link *link;
216                         /*
217                          * we could race with lock deletion which temporarily
218                          * put the lock in freeing state, bug 19080.
219                          */
220                         LASSERT(!(lls->sub_flags & LSF_HELD));
221
222                         link = lov_lock_link_find(env, lck, sublock);
223                         LASSERT(link != NULL);
224                         lov_lock_unlink(env, link, sublock);
225                         lov_sublock_unlock(env, sublock, closure, NULL);
226                         lck->lls_cancel_race = 1;
227                         result = CLO_REPEAT;
228                 } else if (lsep) {
229                         struct lov_sublock_env *subenv;
230                         subenv = lov_sublock_env_get(env, parent, lls);
231                         if (IS_ERR(subenv)) {
232                                 lov_sublock_unlock(env, sublock,
233                                                    closure, NULL);
234                                 result = PTR_ERR(subenv);
235                         } else {
236                                 *lsep = subenv;
237                         }
238                 }
239         }
240         return result;
241 }
242
243 /**
244  * Updates the result of a top-lock operation from a result of sub-lock
245  * sub-operations. Top-operations like lov_lock_{enqueue,use,unuse}() iterate
246  * over sub-locks and lov_subresult() is used to calculate return value of a
247  * top-operation. To this end, possible return values of sub-operations are
248  * ordered as
249  *
250  *     - 0                success
251  *     - CLO_WAIT          wait for event
252  *     - CLO_REPEAT      repeat top-operation
253  *     - -ne            fundamental error
254  *
255  * Top-level return code can only go down through this list. CLO_REPEAT
256  * overwrites CLO_WAIT, because lock mutex was released and sleeping condition
257  * has to be rechecked by the upper layer.
258  */
259 static int lov_subresult(int result, int rc)
260 {
261         int result_rank;
262         int rc_rank;
263
264         LASSERTF(result <= 0 || result == CLO_REPEAT || result == CLO_WAIT,
265                  "result = %d", result);
266         LASSERTF(rc <= 0 || rc == CLO_REPEAT || rc == CLO_WAIT,
267                  "rc = %d\n", rc);
268         CLASSERT(CLO_WAIT < CLO_REPEAT);
269
270         /* calculate ranks in the ordering above */
271         result_rank = result < 0 ? 1 + CLO_REPEAT : result;
272         rc_rank = rc < 0 ? 1 + CLO_REPEAT : rc;
273
274         if (result_rank < rc_rank)
275                 result = rc;
276         return result;
277 }
278
279 /**
280  * Creates sub-locks for a given lov_lock for the first time.
281  *
282  * Goes through all sub-objects of top-object, and creates sub-locks on every
283  * sub-object intersecting with top-lock extent. This is complicated by the
284  * fact that top-lock (that is being created) can be accessed concurrently
285  * through already created sub-locks (possibly shared with other top-locks).
286  */
287 static int lov_lock_sub_init(const struct lu_env *env,
288                              struct lov_lock *lck, const struct cl_io *io)
289 {
290         int result = 0;
291         int i;
292         int nr;
293         u64 start;
294         u64 end;
295         u64 file_start;
296         u64 file_end;
297
298         struct lov_object       *loo    = cl2lov(lck->lls_cl.cls_obj);
299         struct lov_layout_raid0 *r0     = lov_r0(loo);
300         struct cl_lock    *parent = lck->lls_cl.cls_lock;
301
302         lck->lls_orig = parent->cll_descr;
303         file_start = cl_offset(lov2cl(loo), parent->cll_descr.cld_start);
304         file_end   = cl_offset(lov2cl(loo), parent->cll_descr.cld_end + 1) - 1;
305
306         for (i = 0, nr = 0; i < r0->lo_nr; i++) {
307                 /*
308                  * XXX for wide striping smarter algorithm is desirable,
309                  * breaking out of the loop, early.
310                  */
311                 if (likely(r0->lo_sub[i] != NULL) &&
312                     lov_stripe_intersects(loo->lo_lsm, i,
313                                           file_start, file_end, &start, &end))
314                         nr++;
315         }
316         LASSERT(nr > 0);
317         OBD_ALLOC_LARGE(lck->lls_sub, nr * sizeof(lck->lls_sub[0]));
318         if (lck->lls_sub == NULL)
319                 return -ENOMEM;
320
321         lck->lls_nr = nr;
322         /*
323          * First, fill in sub-lock descriptions in
324          * lck->lls_sub[].sub_descr. They are used by lov_sublock_alloc()
325          * (called below in this function, and by lov_lock_enqueue()) to
326          * create sub-locks. At this moment, no other thread can access
327          * top-lock.
328          */
329         for (i = 0, nr = 0; i < r0->lo_nr; ++i) {
330                 if (likely(r0->lo_sub[i] != NULL) &&
331                     lov_stripe_intersects(loo->lo_lsm, i,
332                                           file_start, file_end, &start, &end)) {
333                         struct cl_lock_descr *descr;
334
335                         descr = &lck->lls_sub[nr].sub_descr;
336
337                         LASSERT(descr->cld_obj == NULL);
338                         descr->cld_obj   = lovsub2cl(r0->lo_sub[i]);
339                         descr->cld_start = cl_index(descr->cld_obj, start);
340                         descr->cld_end   = cl_index(descr->cld_obj, end);
341                         descr->cld_mode  = parent->cll_descr.cld_mode;
342                         descr->cld_gid   = parent->cll_descr.cld_gid;
343                         descr->cld_enq_flags   = parent->cll_descr.cld_enq_flags;
344                         /* XXX has no effect */
345                         lck->lls_sub[nr].sub_got = *descr;
346                         lck->lls_sub[nr].sub_stripe = i;
347                         nr++;
348                 }
349         }
350         LASSERT(nr == lck->lls_nr);
351
352         /*
353          * Some sub-locks can be missing at this point. This is not a problem,
354          * because enqueue will create them anyway. Main duty of this function
355          * is to fill in sub-lock descriptions in a race free manner.
356          */
357         return result;
358 }
359
360 static int lov_sublock_release(const struct lu_env *env, struct lov_lock *lck,
361                                int i, int deluser, int rc)
362 {
363         struct cl_lock *parent = lck->lls_cl.cls_lock;
364
365         LASSERT(cl_lock_is_mutexed(parent));
366
367         if (lck->lls_sub[i].sub_flags & LSF_HELD) {
368                 struct cl_lock    *sublock;
369                 int dying;
370
371                 LASSERT(lck->lls_sub[i].sub_lock != NULL);
372                 sublock = lck->lls_sub[i].sub_lock->lss_cl.cls_lock;
373                 LASSERT(cl_lock_is_mutexed(sublock));
374
375                 lck->lls_sub[i].sub_flags &= ~LSF_HELD;
376                 if (deluser)
377                         cl_lock_user_del(env, sublock);
378                 /*
379                  * If the last hold is released, and cancellation is pending
380                  * for a sub-lock, release parent mutex, to avoid keeping it
381                  * while sub-lock is being paged out.
382                  */
383                 dying = (sublock->cll_descr.cld_mode == CLM_PHANTOM ||
384                          sublock->cll_descr.cld_mode == CLM_GROUP ||
385                          (sublock->cll_flags & (CLF_CANCELPEND|CLF_DOOMED))) &&
386                         sublock->cll_holds == 1;
387                 if (dying)
388                         cl_lock_mutex_put(env, parent);
389                 cl_lock_unhold(env, sublock, "lov-parent", parent);
390                 if (dying) {
391                         cl_lock_mutex_get(env, parent);
392                         rc = lov_subresult(rc, CLO_REPEAT);
393                 }
394                 /*
395                  * From now on lck->lls_sub[i].sub_lock is a "weak" pointer,
396                  * not backed by a reference on a
397                  * sub-lock. lovsub_lock_delete() will clear
398                  * lck->lls_sub[i].sub_lock under semaphores, just before
399                  * sub-lock is destroyed.
400                  */
401         }
402         return rc;
403 }
404
405 static void lov_sublock_hold(const struct lu_env *env, struct lov_lock *lck,
406                              int i)
407 {
408         struct cl_lock *parent = lck->lls_cl.cls_lock;
409
410         LASSERT(cl_lock_is_mutexed(parent));
411
412         if (!(lck->lls_sub[i].sub_flags & LSF_HELD)) {
413                 struct cl_lock *sublock;
414
415                 LASSERT(lck->lls_sub[i].sub_lock != NULL);
416                 sublock = lck->lls_sub[i].sub_lock->lss_cl.cls_lock;
417                 LASSERT(cl_lock_is_mutexed(sublock));
418                 LASSERT(sublock->cll_state != CLS_FREEING);
419
420                 lck->lls_sub[i].sub_flags |= LSF_HELD;
421
422                 cl_lock_get_trust(sublock);
423                 cl_lock_hold_add(env, sublock, "lov-parent", parent);
424                 cl_lock_user_add(env, sublock);
425                 cl_lock_put(env, sublock);
426         }
427 }
428
429 static void lov_lock_fini(const struct lu_env *env,
430                           struct cl_lock_slice *slice)
431 {
432         struct lov_lock *lck;
433         int i;
434
435         lck = cl2lov_lock(slice);
436         LASSERT(lck->lls_nr_filled == 0);
437         if (lck->lls_sub != NULL) {
438                 for (i = 0; i < lck->lls_nr; ++i)
439                         /*
440                          * No sub-locks exists at this point, as sub-lock has
441                          * a reference on its parent.
442                          */
443                         LASSERT(lck->lls_sub[i].sub_lock == NULL);
444                 OBD_FREE_LARGE(lck->lls_sub,
445                                lck->lls_nr * sizeof(lck->lls_sub[0]));
446         }
447         OBD_SLAB_FREE_PTR(lck, lov_lock_kmem);
448 }
449
450 static int lov_lock_enqueue_wait(const struct lu_env *env,
451                                  struct lov_lock *lck,
452                                  struct cl_lock *sublock)
453 {
454         struct cl_lock *lock = lck->lls_cl.cls_lock;
455         int          result;
456
457         LASSERT(cl_lock_is_mutexed(lock));
458
459         cl_lock_mutex_put(env, lock);
460         result = cl_lock_enqueue_wait(env, sublock, 0);
461         cl_lock_mutex_get(env, lock);
462         return result ?: CLO_REPEAT;
463 }
464
465 /**
466  * Tries to advance a state machine of a given sub-lock toward enqueuing of
467  * the top-lock.
468  *
469  * \retval 0 if state-transition can proceed
470  * \retval -ve otherwise.
471  */
472 static int lov_lock_enqueue_one(const struct lu_env *env, struct lov_lock *lck,
473                                 struct cl_lock *sublock,
474                                 struct cl_io *io, __u32 enqflags, int last)
475 {
476         int result;
477
478         /* first, try to enqueue a sub-lock ... */
479         result = cl_enqueue_try(env, sublock, io, enqflags);
480         if ((sublock->cll_state == CLS_ENQUEUED) && !(enqflags & CEF_AGL)) {
481                 /* if it is enqueued, try to `wait' on it---maybe it's already
482                  * granted */
483                 result = cl_wait_try(env, sublock);
484                 if (result == CLO_REENQUEUED)
485                         result = CLO_WAIT;
486         }
487         /*
488          * If CEF_ASYNC flag is set, then all sub-locks can be enqueued in
489          * parallel, otherwise---enqueue has to wait until sub-lock is granted
490          * before proceeding to the next one.
491          */
492         if ((result == CLO_WAIT) && (sublock->cll_state <= CLS_HELD) &&
493             (enqflags & CEF_ASYNC) && (!last || (enqflags & CEF_AGL)))
494                 result = 0;
495         return result;
496 }
497
498 /**
499  * Helper function for lov_lock_enqueue() that creates missing sub-lock.
500  */
501 static int lov_sublock_fill(const struct lu_env *env, struct cl_lock *parent,
502                             struct cl_io *io, struct lov_lock *lck, int idx)
503 {
504         struct lov_lock_link *link = NULL;
505         struct cl_lock       *sublock;
506         int                result;
507
508         LASSERT(parent->cll_depth == 1);
509         cl_lock_mutex_put(env, parent);
510         sublock = lov_sublock_alloc(env, io, lck, idx, &link);
511         if (!IS_ERR(sublock))
512                 cl_lock_mutex_get(env, sublock);
513         cl_lock_mutex_get(env, parent);
514
515         if (!IS_ERR(sublock)) {
516                 cl_lock_get_trust(sublock);
517                 if (parent->cll_state == CLS_QUEUING &&
518                     lck->lls_sub[idx].sub_lock == NULL) {
519                         lov_sublock_adopt(env, lck, sublock, idx, link);
520                 } else {
521                         OBD_SLAB_FREE_PTR(link, lov_lock_link_kmem);
522                         /* other thread allocated sub-lock, or enqueue is no
523                          * longer going on */
524                         cl_lock_mutex_put(env, parent);
525                         cl_lock_unhold(env, sublock, "lov-parent", parent);
526                         cl_lock_mutex_get(env, parent);
527                 }
528                 cl_lock_mutex_put(env, sublock);
529                 cl_lock_put(env, sublock);
530                 result = CLO_REPEAT;
531         } else
532                 result = PTR_ERR(sublock);
533         return result;
534 }
535
536 /**
537  * Implementation of cl_lock_operations::clo_enqueue() for lov layer. This
538  * function is rather subtle, as it enqueues top-lock (i.e., advances top-lock
539  * state machine from CLS_QUEUING to CLS_ENQUEUED states) by juggling sub-lock
540  * state machines in the face of sub-locks sharing (by multiple top-locks),
541  * and concurrent sub-lock cancellations.
542  */
543 static int lov_lock_enqueue(const struct lu_env *env,
544                             const struct cl_lock_slice *slice,
545                             struct cl_io *io, __u32 enqflags)
546 {
547         struct cl_lock   *lock    = slice->cls_lock;
548         struct lov_lock *lck     = cl2lov_lock(slice);
549         struct cl_lock_closure *closure = lov_closure_get(env, lock);
550         int i;
551         int result;
552         enum cl_lock_state minstate;
553
554         for (result = 0, minstate = CLS_FREEING, i = 0; i < lck->lls_nr; ++i) {
555                 int rc;
556                 struct lovsub_lock     *sub;
557                 struct lov_lock_sub    *lls;
558                 struct cl_lock   *sublock;
559                 struct lov_sublock_env *subenv;
560
561                 if (lock->cll_state != CLS_QUEUING) {
562                         /*
563                          * Lock might have left QUEUING state if previous
564                          * iteration released its mutex. Stop enqueing in this
565                          * case and let the upper layer to decide what to do.
566                          */
567                         LASSERT(i > 0 && result != 0);
568                         break;
569                 }
570
571                 lls = &lck->lls_sub[i];
572                 sub = lls->sub_lock;
573                 /*
574                  * Sub-lock might have been canceled, while top-lock was
575                  * cached.
576                  */
577                 if (sub == NULL) {
578                         result = lov_sublock_fill(env, lock, io, lck, i);
579                         /* lov_sublock_fill() released @lock mutex,
580                          * restart. */
581                         break;
582                 }
583                 sublock = sub->lss_cl.cls_lock;
584                 rc = lov_sublock_lock(env, lck, lls, closure, &subenv);
585                 if (rc == 0) {
586                         lov_sublock_hold(env, lck, i);
587                         rc = lov_lock_enqueue_one(subenv->lse_env, lck, sublock,
588                                                   subenv->lse_io, enqflags,
589                                                   i == lck->lls_nr - 1);
590                         minstate = min(minstate, sublock->cll_state);
591                         if (rc == CLO_WAIT) {
592                                 switch (sublock->cll_state) {
593                                 case CLS_QUEUING:
594                                         /* take recursive mutex, the lock is
595                                          * released in lov_lock_enqueue_wait.
596                                          */
597                                         cl_lock_mutex_get(env, sublock);
598                                         lov_sublock_unlock(env, sub, closure,
599                                                            subenv);
600                                         rc = lov_lock_enqueue_wait(env, lck,
601                                                                    sublock);
602                                         break;
603                                 case CLS_CACHED:
604                                         cl_lock_get(sublock);
605                                         /* take recursive mutex of sublock */
606                                         cl_lock_mutex_get(env, sublock);
607                                         /* need to release all locks in closure
608                                          * otherwise it may deadlock. LU-2683.*/
609                                         lov_sublock_unlock(env, sub, closure,
610                                                            subenv);
611                                         /* sublock and parent are held. */
612                                         rc = lov_sublock_release(env, lck, i,
613                                                                  1, rc);
614                                         cl_lock_mutex_put(env, sublock);
615                                         cl_lock_put(env, sublock);
616                                         break;
617                                 default:
618                                         lov_sublock_unlock(env, sub, closure,
619                                                            subenv);
620                                         break;
621                                 }
622                         } else {
623                                 LASSERT(sublock->cll_conflict == NULL);
624                                 lov_sublock_unlock(env, sub, closure, subenv);
625                         }
626                 }
627                 result = lov_subresult(result, rc);
628                 if (result != 0)
629                         break;
630         }
631         cl_lock_closure_fini(closure);
632         return result ?: minstate >= CLS_ENQUEUED ? 0 : CLO_WAIT;
633 }
634
635 static int lov_lock_unuse(const struct lu_env *env,
636                           const struct cl_lock_slice *slice)
637 {
638         struct lov_lock *lck     = cl2lov_lock(slice);
639         struct cl_lock_closure *closure = lov_closure_get(env, slice->cls_lock);
640         int i;
641         int result;
642
643         for (result = 0, i = 0; i < lck->lls_nr; ++i) {
644                 int rc;
645                 struct lovsub_lock     *sub;
646                 struct cl_lock   *sublock;
647                 struct lov_lock_sub    *lls;
648                 struct lov_sublock_env *subenv;
649
650                 /* top-lock state cannot change concurrently, because single
651                  * thread (one that released the last hold) carries unlocking
652                  * to the completion. */
653                 LASSERT(slice->cls_lock->cll_state == CLS_INTRANSIT);
654                 lls = &lck->lls_sub[i];
655                 sub = lls->sub_lock;
656                 if (sub == NULL)
657                         continue;
658
659                 sublock = sub->lss_cl.cls_lock;
660                 rc = lov_sublock_lock(env, lck, lls, closure, &subenv);
661                 if (rc == 0) {
662                         if (lls->sub_flags & LSF_HELD) {
663                                 LASSERT(sublock->cll_state == CLS_HELD ||
664                                         sublock->cll_state == CLS_ENQUEUED);
665                                 rc = cl_unuse_try(subenv->lse_env, sublock);
666                                 rc = lov_sublock_release(env, lck, i, 0, rc);
667                         }
668                         lov_sublock_unlock(env, sub, closure, subenv);
669                 }
670                 result = lov_subresult(result, rc);
671         }
672
673         if (result == 0 && lck->lls_cancel_race) {
674                 lck->lls_cancel_race = 0;
675                 result = -ESTALE;
676         }
677         cl_lock_closure_fini(closure);
678         return result;
679 }
680
681
682 static void lov_lock_cancel(const struct lu_env *env,
683                            const struct cl_lock_slice *slice)
684 {
685         struct lov_lock *lck     = cl2lov_lock(slice);
686         struct cl_lock_closure *closure = lov_closure_get(env, slice->cls_lock);
687         int i;
688         int result;
689
690         for (result = 0, i = 0; i < lck->lls_nr; ++i) {
691                 int rc;
692                 struct lovsub_lock     *sub;
693                 struct cl_lock   *sublock;
694                 struct lov_lock_sub    *lls;
695                 struct lov_sublock_env *subenv;
696
697                 /* top-lock state cannot change concurrently, because single
698                  * thread (one that released the last hold) carries unlocking
699                  * to the completion. */
700                 lls = &lck->lls_sub[i];
701                 sub = lls->sub_lock;
702                 if (sub == NULL)
703                         continue;
704
705                 sublock = sub->lss_cl.cls_lock;
706                 rc = lov_sublock_lock(env, lck, lls, closure, &subenv);
707                 if (rc == 0) {
708                         if (!(lls->sub_flags & LSF_HELD)) {
709                                 lov_sublock_unlock(env, sub, closure, subenv);
710                                 continue;
711                         }
712
713                         switch (sublock->cll_state) {
714                         case CLS_HELD:
715                                 rc = cl_unuse_try(subenv->lse_env, sublock);
716                                 lov_sublock_release(env, lck, i, 0, 0);
717                                 break;
718                         default:
719                                 lov_sublock_release(env, lck, i, 1, 0);
720                                 break;
721                         }
722                         lov_sublock_unlock(env, sub, closure, subenv);
723                 }
724
725                 if (rc == CLO_REPEAT) {
726                         --i;
727                         continue;
728                 }
729
730                 result = lov_subresult(result, rc);
731         }
732
733         if (result)
734                 CL_LOCK_DEBUG(D_ERROR, env, slice->cls_lock,
735                               "lov_lock_cancel fails with %d.\n", result);
736
737         cl_lock_closure_fini(closure);
738 }
739
740 static int lov_lock_wait(const struct lu_env *env,
741                          const struct cl_lock_slice *slice)
742 {
743         struct lov_lock *lck     = cl2lov_lock(slice);
744         struct cl_lock_closure *closure = lov_closure_get(env, slice->cls_lock);
745         enum cl_lock_state      minstate;
746         int                  reenqueued;
747         int                  result;
748         int                  i;
749
750 again:
751         for (result = 0, minstate = CLS_FREEING, i = 0, reenqueued = 0;
752              i < lck->lls_nr; ++i) {
753                 int rc;
754                 struct lovsub_lock     *sub;
755                 struct cl_lock   *sublock;
756                 struct lov_lock_sub    *lls;
757                 struct lov_sublock_env *subenv;
758
759                 lls = &lck->lls_sub[i];
760                 sub = lls->sub_lock;
761                 LASSERT(sub != NULL);
762                 sublock = sub->lss_cl.cls_lock;
763                 rc = lov_sublock_lock(env, lck, lls, closure, &subenv);
764                 if (rc == 0) {
765                         LASSERT(sublock->cll_state >= CLS_ENQUEUED);
766                         if (sublock->cll_state < CLS_HELD)
767                                 rc = cl_wait_try(env, sublock);
768
769                         minstate = min(minstate, sublock->cll_state);
770                         lov_sublock_unlock(env, sub, closure, subenv);
771                 }
772                 if (rc == CLO_REENQUEUED) {
773                         reenqueued++;
774                         rc = 0;
775                 }
776                 result = lov_subresult(result, rc);
777                 if (result != 0)
778                         break;
779         }
780         /* Each sublock only can be reenqueued once, so will not loop for
781          * ever. */
782         if (result == 0 && reenqueued != 0)
783                 goto again;
784         cl_lock_closure_fini(closure);
785         return result ?: minstate >= CLS_HELD ? 0 : CLO_WAIT;
786 }
787
788 static int lov_lock_use(const struct lu_env *env,
789                         const struct cl_lock_slice *slice)
790 {
791         struct lov_lock *lck     = cl2lov_lock(slice);
792         struct cl_lock_closure *closure = lov_closure_get(env, slice->cls_lock);
793         int                  result;
794         int                  i;
795
796         LASSERT(slice->cls_lock->cll_state == CLS_INTRANSIT);
797
798         for (result = 0, i = 0; i < lck->lls_nr; ++i) {
799                 int rc;
800                 struct lovsub_lock     *sub;
801                 struct cl_lock   *sublock;
802                 struct lov_lock_sub    *lls;
803                 struct lov_sublock_env *subenv;
804
805                 LASSERT(slice->cls_lock->cll_state == CLS_INTRANSIT);
806
807                 lls = &lck->lls_sub[i];
808                 sub = lls->sub_lock;
809                 if (sub == NULL) {
810                         /*
811                          * Sub-lock might have been canceled, while top-lock was
812                          * cached.
813                          */
814                         result = -ESTALE;
815                         break;
816                 }
817
818                 sublock = sub->lss_cl.cls_lock;
819                 rc = lov_sublock_lock(env, lck, lls, closure, &subenv);
820                 if (rc == 0) {
821                         LASSERT(sublock->cll_state != CLS_FREEING);
822                         lov_sublock_hold(env, lck, i);
823                         if (sublock->cll_state == CLS_CACHED) {
824                                 rc = cl_use_try(subenv->lse_env, sublock, 0);
825                                 if (rc != 0)
826                                         rc = lov_sublock_release(env, lck,
827                                                                  i, 1, rc);
828                         } else if (sublock->cll_state == CLS_NEW) {
829                                 /* Sub-lock might have been canceled, while
830                                  * top-lock was cached. */
831                                 result = -ESTALE;
832                                 lov_sublock_release(env, lck, i, 1, result);
833                         }
834                         lov_sublock_unlock(env, sub, closure, subenv);
835                 }
836                 result = lov_subresult(result, rc);
837                 if (result != 0)
838                         break;
839         }
840
841         if (lck->lls_cancel_race) {
842                 /*
843                  * If there is unlocking happened at the same time, then
844                  * sublock_lock state should be FREEING, and lov_sublock_lock
845                  * should return CLO_REPEAT. In this case, it should return
846                  * ESTALE, and up layer should reset the lock state to be NEW.
847                  */
848                 lck->lls_cancel_race = 0;
849                 LASSERT(result != 0);
850                 result = -ESTALE;
851         }
852         cl_lock_closure_fini(closure);
853         return result;
854 }
855
856 #if 0
857 static int lock_lock_multi_match()
858 {
859         struct cl_lock    *lock    = slice->cls_lock;
860         struct cl_lock_descr    *subneed = &lov_env_info(env)->lti_ldescr;
861         struct lov_object       *loo     = cl2lov(lov->lls_cl.cls_obj);
862         struct lov_layout_raid0 *r0      = lov_r0(loo);
863         struct lov_lock_sub     *sub;
864         struct cl_object        *subobj;
865         u64  fstart;
866         u64  fend;
867         u64  start;
868         u64  end;
869         int i;
870
871         fstart = cl_offset(need->cld_obj, need->cld_start);
872         fend   = cl_offset(need->cld_obj, need->cld_end + 1) - 1;
873         subneed->cld_mode = need->cld_mode;
874         cl_lock_mutex_get(env, lock);
875         for (i = 0; i < lov->lls_nr; ++i) {
876                 sub = &lov->lls_sub[i];
877                 if (sub->sub_lock == NULL)
878                         continue;
879                 subobj = sub->sub_descr.cld_obj;
880                 if (!lov_stripe_intersects(loo->lo_lsm, sub->sub_stripe,
881                                            fstart, fend, &start, &end))
882                         continue;
883                 subneed->cld_start = cl_index(subobj, start);
884                 subneed->cld_end   = cl_index(subobj, end);
885                 subneed->cld_obj   = subobj;
886                 if (!cl_lock_ext_match(&sub->sub_got, subneed)) {
887                         result = 0;
888                         break;
889                 }
890         }
891         cl_lock_mutex_put(env, lock);
892 }
893 #endif
894
895 /**
896  * Check if the extent region \a descr is covered by \a child against the
897  * specific \a stripe.
898  */
899 static int lov_lock_stripe_is_matching(const struct lu_env *env,
900                                        struct lov_object *lov, int stripe,
901                                        const struct cl_lock_descr *child,
902                                        const struct cl_lock_descr *descr)
903 {
904         struct lov_stripe_md *lsm = lov->lo_lsm;
905         u64 start;
906         u64 end;
907         int result;
908
909         if (lov_r0(lov)->lo_nr == 1)
910                 return cl_lock_ext_match(child, descr);
911
912         /*
913          * For a multi-stripes object:
914          * - make sure the descr only covers child's stripe, and
915          * - check if extent is matching.
916          */
917         start = cl_offset(&lov->lo_cl, descr->cld_start);
918         end   = cl_offset(&lov->lo_cl, descr->cld_end + 1) - 1;
919         result = 0;
920         /* glimpse should work on the object with LOV EA hole. */
921         if (end - start <= lsm->lsm_stripe_size) {
922                 int idx;
923
924                 idx = lov_stripe_number(lsm, start);
925                 if (idx == stripe ||
926                     unlikely(lov_r0(lov)->lo_sub[idx] == NULL)) {
927                         idx = lov_stripe_number(lsm, end);
928                         if (idx == stripe ||
929                             unlikely(lov_r0(lov)->lo_sub[idx] == NULL))
930                                 result = 1;
931                 }
932         }
933
934         if (result != 0) {
935                 struct cl_lock_descr *subd = &lov_env_info(env)->lti_ldescr;
936                 u64 sub_start;
937                 u64 sub_end;
938
939                 subd->cld_obj  = NULL;   /* don't need sub object at all */
940                 subd->cld_mode = descr->cld_mode;
941                 subd->cld_gid  = descr->cld_gid;
942                 result = lov_stripe_intersects(lsm, stripe, start, end,
943                                                &sub_start, &sub_end);
944                 LASSERT(result);
945                 subd->cld_start = cl_index(child->cld_obj, sub_start);
946                 subd->cld_end   = cl_index(child->cld_obj, sub_end);
947                 result = cl_lock_ext_match(child, subd);
948         }
949         return result;
950 }
951
952 /**
953  * An implementation of cl_lock_operations::clo_fits_into() method.
954  *
955  * Checks whether a lock (given by \a slice) is suitable for \a
956  * io. Multi-stripe locks can be used only for "quick" io, like truncate, or
957  * O_APPEND write.
958  *
959  * \see ccc_lock_fits_into().
960  */
961 static int lov_lock_fits_into(const struct lu_env *env,
962                               const struct cl_lock_slice *slice,
963                               const struct cl_lock_descr *need,
964                               const struct cl_io *io)
965 {
966         struct lov_lock   *lov = cl2lov_lock(slice);
967         struct lov_object *obj = cl2lov(slice->cls_obj);
968         int result;
969
970         LASSERT(cl_object_same(need->cld_obj, slice->cls_obj));
971         LASSERT(lov->lls_nr > 0);
972
973         /* for top lock, it's necessary to match enq flags otherwise it will
974          * run into problem if a sublock is missing and reenqueue. */
975         if (need->cld_enq_flags != lov->lls_orig.cld_enq_flags)
976                 return 0;
977
978         if (need->cld_mode == CLM_GROUP)
979                 /*
980                  * always allow to match group lock.
981                  */
982                 result = cl_lock_ext_match(&lov->lls_orig, need);
983         else if (lov->lls_nr == 1) {
984                 struct cl_lock_descr *got = &lov->lls_sub[0].sub_got;
985                 result = lov_lock_stripe_is_matching(env,
986                                                      cl2lov(slice->cls_obj),
987                                                      lov->lls_sub[0].sub_stripe,
988                                                      got, need);
989         } else if (io->ci_type != CIT_SETATTR && io->ci_type != CIT_MISC &&
990                    !cl_io_is_append(io) && need->cld_mode != CLM_PHANTOM)
991                 /*
992                  * Multi-stripe locks are only suitable for `quick' IO and for
993                  * glimpse.
994                  */
995                 result = 0;
996         else
997                 /*
998                  * Most general case: multi-stripe existing lock, and
999                  * (potentially) multi-stripe @need lock. Check that @need is
1000                  * covered by @lov's sub-locks.
1001                  *
1002                  * For now, ignore lock expansions made by the server, and
1003                  * match against original lock extent.
1004                  */
1005                 result = cl_lock_ext_match(&lov->lls_orig, need);
1006         CDEBUG(D_DLMTRACE, DDESCR"/"DDESCR" %d %d/%d: %d\n",
1007                PDESCR(&lov->lls_orig), PDESCR(&lov->lls_sub[0].sub_got),
1008                lov->lls_sub[0].sub_stripe, lov->lls_nr, lov_r0(obj)->lo_nr,
1009                result);
1010         return result;
1011 }
1012
1013 void lov_lock_unlink(const struct lu_env *env,
1014                      struct lov_lock_link *link, struct lovsub_lock *sub)
1015 {
1016         struct lov_lock *lck    = link->lll_super;
1017         struct cl_lock  *parent = lck->lls_cl.cls_lock;
1018
1019         LASSERT(cl_lock_is_mutexed(parent));
1020         LASSERT(cl_lock_is_mutexed(sub->lss_cl.cls_lock));
1021
1022         list_del_init(&link->lll_list);
1023         LASSERT(lck->lls_sub[link->lll_idx].sub_lock == sub);
1024         /* yank this sub-lock from parent's array */
1025         lck->lls_sub[link->lll_idx].sub_lock = NULL;
1026         LASSERT(lck->lls_nr_filled > 0);
1027         lck->lls_nr_filled--;
1028         lu_ref_del(&parent->cll_reference, "lov-child", sub->lss_cl.cls_lock);
1029         cl_lock_put(env, parent);
1030         OBD_SLAB_FREE_PTR(link, lov_lock_link_kmem);
1031 }
1032
1033 struct lov_lock_link *lov_lock_link_find(const struct lu_env *env,
1034                                          struct lov_lock *lck,
1035                                          struct lovsub_lock *sub)
1036 {
1037         struct lov_lock_link *scan;
1038
1039         LASSERT(cl_lock_is_mutexed(sub->lss_cl.cls_lock));
1040
1041         list_for_each_entry(scan, &sub->lss_parents, lll_list) {
1042                 if (scan->lll_super == lck)
1043                         return scan;
1044         }
1045         return NULL;
1046 }
1047
1048 /**
1049  * An implementation of cl_lock_operations::clo_delete() method. This is
1050  * invoked for "top-to-bottom" delete, when lock destruction starts from the
1051  * top-lock, e.g., as a result of inode destruction.
1052  *
1053  * Unlinks top-lock from all its sub-locks. Sub-locks are not deleted there:
1054  * this is done separately elsewhere:
1055  *
1056  *     - for inode destruction, lov_object_delete() calls cl_object_kill() for
1057  *       each sub-object, purging its locks;
1058  *
1059  *     - in other cases (e.g., a fatal error with a top-lock) sub-locks are
1060  *       left in the cache.
1061  */
1062 static void lov_lock_delete(const struct lu_env *env,
1063                             const struct cl_lock_slice *slice)
1064 {
1065         struct lov_lock *lck     = cl2lov_lock(slice);
1066         struct cl_lock_closure *closure = lov_closure_get(env, slice->cls_lock);
1067         struct lov_lock_link   *link;
1068         int                  rc;
1069         int                  i;
1070
1071         LASSERT(slice->cls_lock->cll_state == CLS_FREEING);
1072
1073         for (i = 0; i < lck->lls_nr; ++i) {
1074                 struct lov_lock_sub *lls = &lck->lls_sub[i];
1075                 struct lovsub_lock  *lsl = lls->sub_lock;
1076
1077                 if (lsl == NULL) /* already removed */
1078                         continue;
1079
1080                 rc = lov_sublock_lock(env, lck, lls, closure, NULL);
1081                 if (rc == CLO_REPEAT) {
1082                         --i;
1083                         continue;
1084                 }
1085
1086                 LASSERT(rc == 0);
1087                 LASSERT(lsl->lss_cl.cls_lock->cll_state < CLS_FREEING);
1088
1089                 if (lls->sub_flags & LSF_HELD)
1090                         lov_sublock_release(env, lck, i, 1, 0);
1091
1092                 link = lov_lock_link_find(env, lck, lsl);
1093                 LASSERT(link != NULL);
1094                 lov_lock_unlink(env, link, lsl);
1095                 LASSERT(lck->lls_sub[i].sub_lock == NULL);
1096
1097                 lov_sublock_unlock(env, lsl, closure, NULL);
1098         }
1099
1100         cl_lock_closure_fini(closure);
1101 }
1102
1103 static int lov_lock_print(const struct lu_env *env, void *cookie,
1104                           lu_printer_t p, const struct cl_lock_slice *slice)
1105 {
1106         struct lov_lock *lck = cl2lov_lock(slice);
1107         int           i;
1108
1109         (*p)(env, cookie, "%d\n", lck->lls_nr);
1110         for (i = 0; i < lck->lls_nr; ++i) {
1111                 struct lov_lock_sub *sub;
1112
1113                 sub = &lck->lls_sub[i];
1114                 (*p)(env, cookie, "    %d %x: ", i, sub->sub_flags);
1115                 if (sub->sub_lock != NULL)
1116                         cl_lock_print(env, cookie, p,
1117                                       sub->sub_lock->lss_cl.cls_lock);
1118                 else
1119                         (*p)(env, cookie, "---\n");
1120         }
1121         return 0;
1122 }
1123
1124 static const struct cl_lock_operations lov_lock_ops = {
1125         .clo_fini      = lov_lock_fini,
1126         .clo_enqueue   = lov_lock_enqueue,
1127         .clo_wait      = lov_lock_wait,
1128         .clo_use       = lov_lock_use,
1129         .clo_unuse     = lov_lock_unuse,
1130         .clo_cancel    = lov_lock_cancel,
1131         .clo_fits_into = lov_lock_fits_into,
1132         .clo_delete    = lov_lock_delete,
1133         .clo_print     = lov_lock_print
1134 };
1135
1136 int lov_lock_init_raid0(const struct lu_env *env, struct cl_object *obj,
1137                         struct cl_lock *lock, const struct cl_io *io)
1138 {
1139         struct lov_lock *lck;
1140         int result;
1141
1142         OBD_SLAB_ALLOC_PTR_GFP(lck, lov_lock_kmem, GFP_NOFS);
1143         if (lck != NULL) {
1144                 cl_lock_slice_add(lock, &lck->lls_cl, obj, &lov_lock_ops);
1145                 result = lov_lock_sub_init(env, lck, io);
1146         } else
1147                 result = -ENOMEM;
1148         return result;
1149 }
1150
1151 static void lov_empty_lock_fini(const struct lu_env *env,
1152                                 struct cl_lock_slice *slice)
1153 {
1154         struct lov_lock *lck = cl2lov_lock(slice);
1155         OBD_SLAB_FREE_PTR(lck, lov_lock_kmem);
1156 }
1157
1158 static int lov_empty_lock_print(const struct lu_env *env, void *cookie,
1159                         lu_printer_t p, const struct cl_lock_slice *slice)
1160 {
1161         (*p)(env, cookie, "empty\n");
1162         return 0;
1163 }
1164
1165 /* XXX: more methods will be added later. */
1166 static const struct cl_lock_operations lov_empty_lock_ops = {
1167         .clo_fini  = lov_empty_lock_fini,
1168         .clo_print = lov_empty_lock_print
1169 };
1170
1171 int lov_lock_init_empty(const struct lu_env *env, struct cl_object *obj,
1172                 struct cl_lock *lock, const struct cl_io *io)
1173 {
1174         struct lov_lock *lck;
1175         int result = -ENOMEM;
1176
1177         OBD_SLAB_ALLOC_PTR_GFP(lck, lov_lock_kmem, GFP_NOFS);
1178         if (lck != NULL) {
1179                 cl_lock_slice_add(lock, &lck->lls_cl, obj, &lov_empty_lock_ops);
1180                 lck->lls_orig = lock->cll_descr;
1181                 result = 0;
1182         }
1183         return result;
1184 }
1185
1186 static struct cl_lock_closure *lov_closure_get(const struct lu_env *env,
1187                                                struct cl_lock *parent)
1188 {
1189         struct cl_lock_closure *closure;
1190
1191         closure = &lov_env_info(env)->lti_closure;
1192         LASSERT(list_empty(&closure->clc_list));
1193         cl_lock_closure_init(env, closure, parent, 1);
1194         return closure;
1195 }
1196
1197
1198 /** @} lov */