aboutsummaryrefslogtreecommitdiffstats
path: root/fs/lockd/svclock.c
blob: 83ee34203bd7748ffa138a93c21ea964b140a1b8 (plain) (blame)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
/*
 * linux/fs/lockd/svclock.c
 *
 * Handling of server-side locks, mostly of the blocked variety.
 * This is the ugliest part of lockd because we tread on very thin ice.
 * GRANT and CANCEL calls may get stuck, meet in mid-flight, etc.
 * IMNSHO introducing the grant callback into the NLM protocol was one
 * of the worst ideas Sun ever had. Except maybe for the idea of doing
 * NFS file locking at all.
 *
 * I'm trying hard to avoid race conditions by protecting most accesses
 * to a file's list of blocked locks through a semaphore. The global
 * list of blocked locks is not protected in this fashion however.
 * Therefore, some functions (such as the RPC callback for the async grant
 * call) move blocked locks towards the head of the list *while some other
 * process might be traversing it*. This should not be a problem in
 * practice, because this will only cause functions traversing the list
 * to visit some blocks twice.
 *
 * Copyright (C) 1996, Olaf Kirch <okir@monad.swb.de>
 */

#include <linux/types.h>
#include <linux/errno.h>
#include <linux/kernel.h>
#include <linux/sched.h>
#include <linux/smp_lock.h>
#include <linux/sunrpc/clnt.h>
#include <linux/sunrpc/svc.h>
#include <linux/lockd/nlm.h>
#include <linux/lockd/lockd.h>
#include <linux/kthread.h>

#define NLMDBG_FACILITY		NLMDBG_SVCLOCK

#ifdef CONFIG_LOCKD_V4
#define nlm_deadlock	nlm4_deadlock
#else
#define nlm_deadlock	nlm_lck_denied
#endif

static void nlmsvc_release_block(struct nlm_block *block);
static void	nlmsvc_insert_block(struct nlm_block *block, unsigned long);
static void	nlmsvc_remove_block(struct nlm_block *block);

static int nlmsvc_setgrantargs(struct nlm_rqst *call, struct nlm_lock *lock);
static void nlmsvc_freegrantargs(struct nlm_rqst *call);
static const struct rpc_call_ops nlmsvc_grant_ops;

/*
 * The list of blocked locks to retry
 */
static LIST_HEAD(nlm_blocked);

/*
 * Insert a blocked lock into the global list
 */
static void
nlmsvc_insert_block(struct nlm_block *block, unsigned long when)
{
	struct nlm_block *b;
	struct list_head *pos;

	dprintk("lockd: nlmsvc_insert_block(%p, %ld)\n", block, when);
	if (list_empty(&block->b_list)) {
		kref_get(&block->b_count);
	} else {
		list_del_init(&block->b_list);
	}

	pos = &nlm_blocked;
	if (when != NLM_NEVER) {
		if ((when += jiffies) == NLM_NEVER)
			when ++;
		list_for_each(pos, &nlm_blocked) {
			b = list_entry(pos, struct nlm_block, b_list);
			if (time_after(b->b_when,when) || b->b_when == NLM_NEVER)
				break;
		}
		/* On normal exit from the loop, pos == &nlm_blocked,
		 * so we will be adding to the end of the list - good
		 */
	}

	list_add_tail(&block->b_list, pos);
	block->b_when = when;
}

/*
 * Remove a block from the global list
 */
static inline void
nlmsvc_remove_block(struct nlm_block *block)
{
	if (!list_empty(&block->b_list)) {
		list_del_init(&block->b_list);
		nlmsvc_release_block(block);
	}
}

/*
 * Find a block for a given lock
 */
static struct nlm_block *
nlmsvc_lookup_block(struct nlm_file *file, struct nlm_lock *lock)
{
	struct nlm_block	*block;
	struct file_lock	*fl;

	dprintk("lockd: nlmsvc_lookup_block f=%p pd=%d %Ld-%Ld ty=%d\n",
				file, lock->fl.fl_pid,
				(long long)lock->fl.fl_start,
				(long long)lock->fl.fl_end, lock->fl.fl_type);
	list_for_each_entry(block, &nlm_blocked, b_list) {
		fl = &block->b_call->a_args.lock.fl;
		dprintk("lockd: check f=%p pd=%d %Ld-%Ld ty=%d cookie=%s\n",
				block->b_file, fl->fl_pid,
				(long long)fl->fl_start,
				(long long)fl->fl_end, fl->fl_type,
				nlmdbg_cookie2a(&block->b_call->a_args.cookie));
		if (block->b_file == file && nlm_compare_locks(fl, &lock->fl)) {
			kref_get(&block->b_count);
			return block;
		}
	}

	return NULL;
}

static inline int nlm_cookie_match(struct nlm_cookie *a, struct nlm_cookie *b)
{
	if (a->len != b->len)
		return 0;
	if (memcmp(a->data, b->data, a->len))
		return 0;
	return 1;
}

/*
 * Find a block with a given NLM cookie.
 */
static inline struct nlm_block *
nlmsvc_find_block(struct nlm_cookie *cookie)
{
	struct nlm_block *block;

	list_for_each_entry(block, &nlm_blocked, b_list) {
		if (nlm_cookie_match(&block->b_call->a_args.cookie,cookie))
			goto found;
	}

	return NULL;

found:
	dprintk("nlmsvc_find_block(%s): block=%p\n", nlmdbg_cookie2a(cookie), block);
	kref_get(&block->b_count);
	return block;
}

/*
 * Create a block and initialize it.
 *
 * Note: we explicitly set the cookie of the grant reply to that of
 * the blocked lock request. The spec explicitly mentions that the client
 * should _not_ rely on the callback containing the same cookie as the
 * request, but (as I found out later) that's because some implementations
 * do just this. Never mind the standards comittees, they support our
 * logging industries.
 *
 * 10 years later: I hope we can safely ignore these old and broken
 * clients by now. Let's fix this so we can uniquely identify an incoming
 * GRANTED_RES message by cookie, without having to rely on the client's IP
 * address. --okir
 */
static struct nlm_block *
nlmsvc_create_block(struct svc_rqst *rqstp, struct nlm_host *host,
		    struct nlm_file *file, struct nlm_lock *lock,
		    struct nlm_cookie *cookie)
{
	struct nlm_block	*block;
	struct nlm_rqst		*call = NULL;

	nlm_get_host(host);
	call = nlm_alloc_call(host);
	if (call == NULL)
		return NULL;

	/* Allocate memory for block, and initialize arguments */
	block = kzalloc(sizeof(*block), GFP_KERNEL);
	if (block == NULL)
		goto failed;
	kref_init(&block->b_count);
	INIT_LIST_HEAD(&block->b_list);
	INIT_LIST_HEAD(&block->b_flist);

	if (!nlmsvc_setgrantargs(call, lock))
		goto failed_free;

	/* Set notifier function for VFS, and init args */
	call->a_args.lock.fl.fl_flags |= FL_SLEEP;
	call->a_args.lock.fl.fl_lmops = &nlmsvc_lock_operations;
	nlmclnt_next_cookie(&call->a_args.cookie);

	dprintk("lockd: created block %p...\n", block);

	/* Create and initialize the block */
	block->b_daemon = rqstp->rq_server;
	block->b_host   = host;
	block->b_file   = file;
	block->b_fl = NULL;
	file->f_count++;

	/* Add to file's list of blocks */
	list_add(&block->b_flist, &file->f_blocks);

	/* Set up RPC arguments for callback */
	block->b_call = call;
	call->a_flags   = RPC_TASK_ASYNC;
	call->a_block = block;

	return block;

failed_free:
	kfree(block);
failed:
	nlm_release_call(call);
	return NULL;
}

/*
 * Delete a block.
 * It is the caller's responsibility to check whether the file
 * can be closed hereafter.
 */
static int nlmsvc_unlink_block(struct nlm_block *block)
{
	int status;
	dprintk("lockd: unlinking block %p...\n", block);

	/* Remove block from list */
	status = posix_unblock_lock(block->b_file->f_file, &block->b_call->a_args.lock.fl);
	nlmsvc_remove_block(block);
	return status;
}

static void nlmsvc_free_block(struct kref *kref)
{
	struct nlm_block *block = container_of(kref, struct nlm_block, b_count);
	struct nlm_file		*file = block->b_file;

	dprintk("lockd: freeing block %p...\n", block);

	/* Remove block from file's list of blocks */
	mutex_lock(&file->f_mutex);
	list_del_init(&block->b_flist);
	mutex_unlock(&file->f_mutex);

	nlmsvc_freegrantargs(block->b_call);
	nlm_release_call(block->b_call);
	nlm_release_file(block->b_file);
	kfree(block->b_fl);
	kfree(block);
}

static void nlmsvc_release_block(struct nlm_block *block)
{
	if (block != NULL)
		kref_put(&block->b_count, nlmsvc_free_block);
}

/*
 * Loop over all blocks and delete blocks held by
 * a matching host.
 */
void nlmsvc_traverse_blocks(struct nlm_host *host,
			struct nlm_file *file,
			nlm_host_match_fn_t match)
{
	struct nlm_block *block, *next;

restart:
	mutex_lock(&file->f_mutex);
	list_for_each_entry_safe(block, next, &file->f_blocks, b_flist) {
		if (!match(block->b_host, host))
			continue;
		/* Do not destroy blocks that are not on
		 * the global retry list - why? */
		if (list_empty(&block->b_list))
			continue;
		kref_get(&block->b_count);
		mutex_unlock(&file->f_mutex);
		nlmsvc_unlink_block(block);
		nlmsvc_release_block(block);
		goto restart;
	}
	mutex_unlock(&file->f_mutex);
}

/*
 * Initialize arguments for GRANTED call. The nlm_rqst structure
 * has been cleared already.
 */
static int nlmsvc_setgrantargs(struct nlm_rqst *call, struct nlm_lock *lock)
{
	locks_copy_lock(&call->a_args.lock.fl, &lock->fl);
	memcpy(&call->a_args.lock.fh, &lock->fh, sizeof(call->a_args.lock.fh));
	call->a_args.lock.caller = utsname()->nodename;
	call->a_args.lock.oh.len = lock->oh.len;

	/* set default data area */
	call->a_args.lock.oh.data = call->a_owner;
	call->a_args.lock.svid = lock->fl.fl_pid;

	if (lock->oh.len > NLMCLNT_OHSIZE) {
		void *data = kmalloc(lock->oh.len, GFP_KERNEL);
		if (!data)
			return 0;
		call->a_args.lock.oh.data = (u8 *) data;
	}

	memcpy(call->a_args.lock.oh.data, lock->oh.data, lock->oh.len);
	return 1;
}

static void nlmsvc_freegrantargs(struct nlm_rqst *call)
{
	if (call->a_args.lock.oh.data != call->a_owner)
		kfree(call->a_args.lock.oh.data);
}

/*
 * Deferred lock request handling for non-blocking lock
 */
static __be32
nlmsvc_defer_lock_rqst(struct svc_rqst *rqstp, struct nlm_block *block)
{
	__be32 status = nlm_lck_denied_nolocks;

	block->b_flags |= B_QUEUED;

	nlmsvc_insert_block(block, NLM_TIMEOUT);

	block->b_cache_req = &rqstp->rq_chandle;
	if (rqstp->rq_chandle.defer) {
		block->b_deferred_req =
			rqstp->rq_chandle.defer(block->b_cache_req);
		if (block->b_deferred_req != NULL)
			status = nlm_drop_reply;
	}
	dprintk("lockd: nlmsvc_defer_lock_rqst block %p flags %d status %d\n",
		block, block->b_flags, ntohl(status));

	return status;
}

/*
 * Attempt to establish a lock, and if it can't be granted, block it
 * if required.
 */
__be32
nlmsvc_lock(struct svc_rqst *rqstp, struct nlm_file *file,
	    struct nlm_host *host, struct nlm_lock *lock, int wait,
	    struct nlm_cookie *cookie, int reclaim)
{
	struct nlm_block	*block = NULL;
	int			error;
	__be32			ret;

	dprintk("lockd: nlmsvc_lock(%s/%ld, ty=%d, pi=%d, %Ld-%Ld, bl=%d)\n",
				file->f_file->f_path.dentry->d_inode->i_sb->s_id,
				file->f_file->f_path.dentry->d_inode->i_ino,
				lock->fl.fl_type, lock->fl.fl_pid,
				(long long)lock->fl.fl_start,
				(long long)lock->fl.fl_end,
				wait);

	/* Lock file against concurrent access */
	mutex_lock(&file->f_mutex);
	/* Get existing block (in case client is busy-waiting)
	 * or create new block
	 */
	block = nlmsvc_lookup_block(file, lock);
	if (block == NULL) {
		block = nlmsvc_create_block(rqstp, host, file, lock, cookie);
		ret = nlm_lck_denied_nolocks;
		if (block == NULL)
			goto out;
		lock = &block->b_call->a_args.lock;
	} else
		lock->fl.fl_flags &= ~FL_SLEEP;

	if (block->b_flags & B_QUEUED) {
		dprintk("lockd: nlmsvc_lock deferred block %p flags %d\n",
							block, block->b_flags);
		if (block->b_granted) {
			nlmsvc_unlink_block(block);
			ret = nlm_granted;
			goto out;
		}
		if (block->b_flags & B_TIMED_OUT) {
			nlmsvc_unlink_block(block);
			ret = nlm_lck_denied;
			goto out;
		}
		ret = nlm_drop_reply;
		goto out;
	}

	if (locks_in_grace() && !reclaim) {
		ret = nlm_lck_denied_grace_period;
		goto out;
	}
	if (reclaim && !locks_in_grace()) {
		ret = nlm_lck_denied_grace_period;
		goto out;
	}

	if (!wait)
		lock->fl.fl_flags &= ~FL_SLEEP;
	error = vfs_lock_file(file->f_file, F_SETLK, &lock->fl, NULL);
	lock->fl.fl_flags &= ~FL_SLEEP;

	dprintk("lockd: vfs_lock_file returned %d\n", error);
	switch (error) {
		case 0:
			ret = nlm_granted;
			goto out;
		case -EAGAIN:
			/*
			 * If this is a blocking request for an
			 * already pending lock request then we need
			 * to put it back on lockd's block list
			 */
			if (wait)
				break;
			ret = nlm_lck_denied;
			goto out;
		case FILE_LOCK_DEFERRED:
			if (wait)
				break;
			/* Filesystem lock operation is in progress
			   Add it to the queue waiting for callback */
			ret = nlmsvc_defer_lock_rqst(rqstp, block);
			goto out;
		case -EDEADLK:
			ret = nlm_deadlock;
			goto out;
		default:			/* includes ENOLCK */
			ret = nlm_lck_denied_nolocks;
			goto out;
	}

	ret = nlm_lck_blocked;

	/* Append to list of blocked */
	nlmsvc_insert_block(block, NLM_NEVER);
out:
	mutex_unlock(&file->f_mutex);
	nlmsvc_release_block(block);
	dprintk("lockd: nlmsvc_lock returned %u\n", ret);
	return ret;
}

/*
 * Test for presence of a conflicting lock.
 */
__be32
nlmsvc_testlock(struct svc_rqst *rqstp, struct nlm_file *file,
		struct nlm_host *host, struct nlm_lock *lock,
		struct nlm_lock *conflock, struct nlm_cookie *cookie)
{
	struct nlm_block 	*block = NULL;
	int			error;
	__be32			ret;

	dprintk("lockd: nlmsvc_testlock(%s/%ld, ty=%d, %Ld-%Ld)\n",
				file->f_file->f_path.dentry->d_inode->i_sb->s_id,
				file->f_file->f_path.dentry->d_inode->i_ino,
				lock->fl.fl_type,
				(long long)lock->fl.fl_start,
				(long long)lock->fl.fl_end);

	/* Get existing block (in case client is busy-waiting) */
	block = nlmsvc_lookup_block(file, lock);

	if (block == NULL) {
		struct file_lock *conf = kzalloc(sizeof(*conf), GFP_KERNEL);

		if (conf == NULL)
			return nlm_granted;
		block = nlmsvc_create_block(rqstp, host, file, lock, cookie);
		if (block == NULL) {
			kfree(conf);
			return nlm_granted;
		}
		block->b_fl = conf;
	}
	if (block->b_flags & B_QUEUED) {
		dprintk("lockd: nlmsvc_testlock deferred block %p flags %d fl %p\n",
			block, block->b_flags, block->b_fl);
		if (block->b_flags & B_TIMED_OUT) {
			nlmsvc_unlink_block(block);
			ret = nlm_lck_denied;
			goto out;
		}
		if (block->b_flags & B_GOT_CALLBACK) {
			nlmsvc_unlink_block(block);
			if (block->b_fl != NULL
					&& block->b_fl->fl_type != F_UNLCK) {
				lock->fl = *block->b_fl;
				goto conf_lock;
			} else {
				ret = nlm_granted;
				goto out;
			}
		}
		ret = nlm_drop_reply;
		goto out;
	}

	if (locks_in_grace()) {
		ret = nlm_lck_denied_grace_period;
		goto out;
	}
	error = vfs_test_lock(file->f_file, &lock->fl);
	if (error == FILE_LOCK_DEFERRED) {
		ret = nlmsvc_defer_lock_rqst(rqstp, block);
		goto out;
	}
	if (error) {
		ret = nlm_lck_denied_nolocks;
		goto out;
	}
	if (lock->fl.fl_type == F_UNLCK) {
		ret = nlm_granted;
		goto out;
	}

conf_lock:
	dprintk("lockd: conflicting lock(ty=%d, %Ld-%Ld)\n",
		lock->fl.fl_type, (long long)lock->fl.fl_start,
		(long long)lock->fl.fl_end);
	conflock->caller = "somehost";	/* FIXME */
	conflock->len = strlen(conflock->caller);
	conflock->oh.len = 0;		/* don't return OH info */
	conflock->svid = lock->fl.fl_pid;
	conflock->fl.fl_type = lock->fl.fl_type;
	conflock->fl.fl_start = lock->fl.fl_start;
	conflock->fl.fl_end = lock->fl.fl_end;
	ret = nlm_lck_denied;
out:
	if (block)
		nlmsvc_release_block(block);
	return ret;
}

/*
 * Remove a lock.
 * This implies a CANCEL call: We send a GRANT_MSG, the client replies
 * with a GRANT_RES call which gets lost, and calls UNLOCK immediately
 * afterwards. In this case the block will still be there, and hence
 * must be removed.
 */
__be32
nlmsvc_unlock(struct nlm_file *file, struct nlm_lock *lock)
{
	int	error;

	dprintk("lockd: nlmsvc_unlock(%s/%ld, pi=%d, %Ld-%Ld)\n",
				file->f_file->f_path.dentry->d_inode->i_sb->s_id,
				file->f_file->f_path.dentry->d_inode->i_ino,
				lock->fl.fl_pid,
				(long long)lock->fl.fl_start,
				(long long)lock->fl.fl_end);

	/* First, cancel any lock that might be there */
	nlmsvc_cancel_blocked(file, lock);

	lock->fl.fl_type = F_UNLCK;
	error = vfs_lock_file(file->f_file, F_SETLK, &lock->fl, NULL);

	return (error < 0)? nlm_lck_denied_nolocks : nlm_granted;
}

/*
 * Cancel a previously blocked request.
 *
 * A cancel request always overrides any grant that may currently
 * be in progress.
 * The calling procedure must check whether the file can be closed.
 */
__be32
nlmsvc_cancel_blocked(struct nlm_file *file, struct nlm_lock *lock)
{
	struct nlm_block	*block;
	int status = 0;

	dprintk("lockd: nlmsvc_cancel(%s/%ld, pi=%d, %Ld-%Ld)\n",
				file->f_file->f_path.dentry->d_inode->i_sb->s_id,
				file->f_file->f_path.dentry->d_inode->i_ino,
				lock->fl.fl_pid,
				(long long)lock->fl.fl_start,
				(long long)lock->fl.fl_end);

	if (locks_in_grace())
		return nlm_lck_denied_grace_period;

	mutex_lock(&file->f_mutex);
	block = nlmsvc_lookup_block(file, lock);
	mutex_unlock(&file->f_mutex);
	if (block != NULL) {
		vfs_cancel_lock(block->b_file->f_file,
				&block->b_call->a_args.lock.fl);
		status = nlmsvc_unlink_block(block);
		nlmsvc_release_block(block);
	}
	return status ? nlm_lck_denied : nlm_granted;
}

/*
 * This is a callback from the filesystem for VFS file lock requests.
 * It will be used if fl_grant is defined and the filesystem can not
 * respond to the request immediately.
 * For GETLK request it will copy the reply to the nlm_block.
 * For SETLK or SETLKW request it will get the local posix lock.
 * In all cases it will move the block to the head of nlm_blocked q where
 * nlmsvc_retry_blocked() can send back a reply for SETLKW or revisit the
 * deferred rpc for GETLK and SETLK.
 */
static void
nlmsvc_update_deferred_block(struct nlm_block *block, struct file_lock *conf,
			     int result)
{
	block->b_flags |= B_GOT_CALLBACK;
	if (result == 0)
		block->b_granted = 1;
	else
		block->b_flags |= B_TIMED_OUT;
	if (conf) {
		if (block->b_fl)
			__locks_copy_lock(block->b_fl, conf);
	}
}

static int nlmsvc_grant_deferred(struct file_lock *fl, struct file_lock *conf,
					int result)
{
	struct nlm_block *block;
	int rc = -ENOENT;

	lock_kernel();
	list_for_each_entry(block, &nlm_blocked, b_list) {
		if (nlm_compare_locks(&block->b_call->a_args.lock.fl, fl)) {
			dprintk("lockd: nlmsvc_notify_blocked block %p flags %d\n",
							block, block->b_flags);
			if (block->b_flags & B_QUEUED) {
				if (block->b_flags & B_TIMED_OUT) {
					rc = -ENOLCK;
					break;
				}
				nlmsvc_update_deferred_block(block, conf, result);
			} else if (result == 0)
				block->b_granted = 1;

			nlmsvc_insert_block(block, 0);
			svc_wake_up(block->b_daemon);
			rc = 0;
			break;
		}
	}
	unlock_kernel();
	if (rc == -ENOENT)
		printk(KERN_WARNING "lockd: grant for unknown block\n");
	return rc;
}

/*
 * Unblock a blocked lock request. This is a callback invoked from the
 * VFS layer when a lock on which we blocked is removed.
 *
 * This function doesn't grant the blocked lock instantly, but rather moves
 * the block to the head of nlm_blocked where it can be picked up by lockd.
 */
static void
nlmsvc_notify_blocked(struct file_lock *fl)
{
	struct nlm_block	*block;

	dprintk("lockd: VFS unblock notification for block %p\n", fl);
	list_for_each_entry(block, &nlm_blocked, b_list) {
		if (nlm_compare_locks(&block->b_call->a_args.lock.fl, fl)) {
			nlmsvc_insert_block(block, 0);
			svc_wake_up(block->b_daemon);
			return;
		}
	}

	printk(KERN_WARNING "lockd: notification for unknown block!\n");
}

static int nlmsvc_same_owner(struct file_lock *fl1, struct file_lock *fl2)
{
	return fl1->fl_owner == fl2->fl_owner && fl1->fl_pid == fl2->fl_pid;
}

struct lock_manager_operations nlmsvc_lock_operations = {
	.fl_compare_owner = nlmsvc_same_owner,
	.fl_notify = nlmsvc_notify_blocked,
	.fl_grant = nlmsvc_grant_deferred,
};

/*
 * Try to claim a lock that was previously blocked.
 *
 * Note that we use both the RPC_GRANTED_MSG call _and_ an async
 * RPC thread when notifying the client. This seems like overkill...
 * Here's why:
 *  -	we don't want to use a synchronous RPC thread, otherwise
 *	we might find ourselves hanging on a dead portmapper.
 *  -	Some lockd implementations (e.g. HP) don't react to
 *	RPC_GRANTED calls; they seem to insist on RPC_GRANTED_MSG calls.
 */
static void
nlmsvc_grant_blocked(struct nlm_block *block)
{
	struct nlm_file		*file = block->b_file;
	struct nlm_lock		*lock = &block->b_call->a_args.lock;
	int			error;

	dprintk("lockd: grant blocked lock %p\n", block);

	kref_get(&block->b_count);

	/* Unlink block request from list */
	nlmsvc_unlink_block(block);

	/* If b_granted is true this means we've been here before.
	 * Just retry the grant callback, possibly refreshing the RPC
	 * binding */
	if (block->b_granted) {
		nlm_rebind_host(block->b_host);
		goto callback;
	}

	/* Try the lock operation again */
	lock->fl.fl_flags |= FL_SLEEP;
	error = vfs_lock_file(file->f_file, F_SETLK, &lock->fl, NULL);
	lock->fl.fl_flags &= ~FL_SLEEP;

	switch (error) {
	case 0:
		break;
	case FILE_LOCK_DEFERRED:
		dprintk("lockd: lock still blocked error %d\n", error);
		nlmsvc_insert_block(block, NLM_NEVER);
		nlmsvc_release_block(block);
		return;
	default:
		printk(KERN_WARNING "lockd: unexpected error %d in %s!\n",
				-error, __func__);
		nlmsvc_insert_block(block, 10 * HZ);
		nlmsvc_release_block(block);
		return;
	}

callback:
	/* Lock was granted by VFS. */
	dprintk("lockd: GRANTing blocked lock.\n");
	block->b_granted = 1;

	/* keep block on the list, but don't reattempt until the RPC
	 * completes or the submission fails
	 */
	nlmsvc_insert_block(block, NLM_NEVER);

	/* Call the client -- use a soft RPC task since nlmsvc_retry_blocked
	 * will queue up a new one if this one times out
	 */
	error = nlm_async_call(block->b_call, NLMPROC_GRANTED_MSG,
				&nlmsvc_grant_ops);

	/* RPC submission failed, wait a bit and retry */
	if (error < 0)
		nlmsvc_insert_block(block, 10 * HZ);
}

/*
 * This is the callback from the RPC layer when the NLM_GRANTED_MSG
 * RPC call has succeeded or timed out.
 * Like all RPC callbacks, it is invoked by the rpciod process, so it
 * better not sleep. Therefore, we put the blocked lock on the nlm_blocked
 * chain once more in order to have it removed by lockd itself (which can
 * then sleep on the file semaphore without disrupting e.g. the nfs client).
 */
static void nlmsvc_grant_callback(struct rpc_task *task, void *data)
{
	struct nlm_rqst		*call = data;
	struct nlm_block	*block = call->a_block;
	unsigned long		timeout;

	dprintk("lockd: GRANT_MSG RPC callback\n");

	lock_kernel();
	/* if the block is not on a list at this point then it has
	 * been invalidated. Don't try to requeue it.
	 *
	 * FIXME: it's possible that the block is removed from the list
	 * after this check but before the nlmsvc_insert_block. In that
	 * case it will be added back. Perhaps we need better locking
	 * for nlm_blocked?
	 */
	if (list_empty(&block->b_list))
		goto out;

	/* Technically, we should down the file semaphore here. Since we
	 * move the block towards the head of the queue only, no harm
	 * can be done, though. */
	if (task->tk_status < 0) {
		/* RPC error: Re-insert for retransmission */
		timeout = 10 * HZ;
	} else {
		/* Call was successful, now wait for client callback */
		timeout = 60 * HZ;
	}
	nlmsvc_insert_block(block, timeout);
	svc_wake_up(block->b_daemon);
out:
	unlock_kernel();
}

static void nlmsvc_grant_release(void *data)
{
	struct nlm_rqst		*call = data;

	lock_kernel();
	nlmsvc_release_block(call->a_block);
	unlock_kernel();
}

static const struct rpc_call_ops nlmsvc_grant_ops = {
	.rpc_call_done = nlmsvc_grant_callback,
	.rpc_release = nlmsvc_grant_release,
};

/*
 * We received a GRANT_RES callback. Try to find the corresponding
 * block.
 */
void
nlmsvc_grant_reply(struct nlm_cookie *cookie, __be32 status)
{
	struct nlm_block	*block;

	dprintk("grant_reply: looking for cookie %x, s=%d \n",
		*(unsigned int *)(cookie->data), status);
	if (!(block = nlmsvc_find_block(cookie)))
		return;

	if (block) {
		if (status == nlm_lck_denied_grace_period) {
			/* Try again in a couple of seconds */
			nlmsvc_insert_block(block, 10 * HZ);
		} else {
			/* Lock is now held by client, or has been rejected.
			 * In both cases, the block should be removed. */
			nlmsvc_unlink_block(block);
		}
	}
	nlmsvc_release_block(block);
}

/* Helper function to handle retry of a deferred block.
 * If it is a blocking lock, call grant_blocked.
 * For a non-blocking lock or test lock, revisit the request.
 */
static void
retry_deferred_block(struct nlm_block *block)
{
	if (!(block->b_flags & B_GOT_CALLBACK))
		block->b_flags |= B_TIMED_OUT;
	nlmsvc_insert_block(block, NLM_TIMEOUT);
	dprintk("revisit block %p flags %d\n",	block, block->b_flags);
	if (block->b_deferred_req) {
		block->b_deferred_req->revisit(block->b_deferred_req, 0);
		block->b_deferred_req = NULL;
	}
}

/*
 * Retry all blocked locks that have been notified. This is where lockd
 * picks up locks that can be granted, or grant notifications that must
 * be retransmitted.
 */
unsigned long
nlmsvc_retry_blocked(void)
{
	unsigned long	timeout = MAX_SCHEDULE_TIMEOUT;
	struct nlm_block *block;

	while (!list_empty(&nlm_blocked) && !kthread_should_stop()) {
		block = list_entry(nlm_blocked.next, struct nlm_block, b_list);

		if (block->b_when == NLM_NEVER)
			break;
		if (time_after(block->b_when, jiffies)) {
			timeout = block->b_when - jiffies;
			break;
		}

		dprintk("nlmsvc_retry_blocked(%p, when=%ld)\n",
			block, block->b_when);
		if (block->b_flags & B_QUEUED) {
			dprintk("nlmsvc_retry_blocked delete block (%p, granted=%d, flags=%d)\n",
				block, block->b_granted, block->b_flags);
			retry_deferred_block(block);
		} else
			nlmsvc_grant_blocked(block);
	}

	return timeout;
}