summaryrefslogtreecommitdiff
path: root/fs/fscache/page.c
blob: b38b13d2a5559330d43d7448b65142c2711204dc (plain)
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
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
/* Cache page management and data I/O routines
 *
 * Copyright (C) 2004-2008 Red Hat, Inc. All Rights Reserved.
 * Written by David Howells (dhowells@redhat.com)
 *
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of the GNU General Public License
 * as published by the Free Software Foundation; either version
 * 2 of the License, or (at your option) any later version.
 */

#define FSCACHE_DEBUG_LEVEL PAGE
#include <linux/module.h>
#include <linux/fscache-cache.h>
#include <linux/buffer_head.h>
#include <linux/pagevec.h>
#include <linux/slab.h>
#include "internal.h"

/*
 * check to see if a page is being written to the cache
 */
bool __fscache_check_page_write(struct fscache_cookie *cookie, struct page *page)
{
	void *val;

	rcu_read_lock();
	val = radix_tree_lookup(&cookie->stores, page->index);
	rcu_read_unlock();

	return val != NULL;
}
EXPORT_SYMBOL(__fscache_check_page_write);

/*
 * wait for a page to finish being written to the cache
 */
void __fscache_wait_on_page_write(struct fscache_cookie *cookie, struct page *page)
{
	wait_queue_head_t *wq = bit_waitqueue(&cookie->flags, 0);

	wait_event(*wq, !__fscache_check_page_write(cookie, page));
}
EXPORT_SYMBOL(__fscache_wait_on_page_write);

/*
 * decide whether a page can be released, possibly by cancelling a store to it
 * - we're allowed to sleep if __GFP_WAIT is flagged
 */
bool __fscache_maybe_release_page(struct fscache_cookie *cookie,
				  struct page *page,
				  gfp_t gfp)
{
	struct page *xpage;
	void *val;

	_enter("%p,%p,%x", cookie, page, gfp);

	rcu_read_lock();
	val = radix_tree_lookup(&cookie->stores, page->index);
	if (!val) {
		rcu_read_unlock();
		fscache_stat(&fscache_n_store_vmscan_not_storing);
		__fscache_uncache_page(cookie, page);
		return true;
	}

	/* see if the page is actually undergoing storage - if so we can't get
	 * rid of it till the cache has finished with it */
	if (radix_tree_tag_get(&cookie->stores, page->index,
			       FSCACHE_COOKIE_STORING_TAG)) {
		rcu_read_unlock();
		goto page_busy;
	}

	/* the page is pending storage, so we attempt to cancel the store and
	 * discard the store request so that the page can be reclaimed */
	spin_lock(&cookie->stores_lock);
	rcu_read_unlock();

	if (radix_tree_tag_get(&cookie->stores, page->index,
			       FSCACHE_COOKIE_STORING_TAG)) {
		/* the page started to undergo storage whilst we were looking,
		 * so now we can only wait or return */
		spin_unlock(&cookie->stores_lock);
		goto page_busy;
	}

	xpage = radix_tree_delete(&cookie->stores, page->index);
	spin_unlock(&cookie->stores_lock);

	if (xpage) {
		fscache_stat(&fscache_n_store_vmscan_cancelled);
		fscache_stat(&fscache_n_store_radix_deletes);
		ASSERTCMP(xpage, ==, page);
	} else {
		fscache_stat(&fscache_n_store_vmscan_gone);
	}

	wake_up_bit(&cookie->flags, 0);
	if (xpage)
		page_cache_release(xpage);
	__fscache_uncache_page(cookie, page);
	return true;

page_busy:
	/* we might want to wait here, but that could deadlock the allocator as
	 * the work threads writing to the cache may all end up sleeping
	 * on memory allocation */
	fscache_stat(&fscache_n_store_vmscan_busy);
	return false;
}
EXPORT_SYMBOL(__fscache_maybe_release_page);

/*
 * note that a page has finished being written to the cache
 */
static void fscache_end_page_write(struct fscache_object *object,
				   struct page *page)
{
	struct fscache_cookie *cookie;
	struct page *xpage = NULL;

	spin_lock(&object->lock);
	cookie = object->cookie;
	if (cookie) {
		/* delete the page from the tree if it is now no longer
		 * pending */
		spin_lock(&cookie->stores_lock);
		radix_tree_tag_clear(&cookie->stores, page->index,
				     FSCACHE_COOKIE_STORING_TAG);
		if (!radix_tree_tag_get(&cookie->stores, page->index,
					FSCACHE_COOKIE_PENDING_TAG)) {
			fscache_stat(&fscache_n_store_radix_deletes);
			xpage = radix_tree_delete(&cookie->stores, page->index);
		}
		spin_unlock(&cookie->stores_lock);
		wake_up_bit(&cookie->flags, 0);
	}
	spin_unlock(&object->lock);
	if (xpage)
		page_cache_release(xpage);
}

/*
 * actually apply the changed attributes to a cache object
 */
static void fscache_attr_changed_op(struct fscache_operation *op)
{
	struct fscache_object *object = op->object;
	int ret;

	_enter("{OBJ%x OP%x}", object->debug_id, op->debug_id);

	fscache_stat(&fscache_n_attr_changed_calls);

	if (fscache_object_is_active(object)) {
		fscache_stat(&fscache_n_cop_attr_changed);
		ret = object->cache->ops->attr_changed(object);
		fscache_stat_d(&fscache_n_cop_attr_changed);
		if (ret < 0)
			fscache_abort_object(object);
	}

	fscache_op_complete(op);
	_leave("");
}

/*
 * notification that the attributes on an object have changed
 */
int __fscache_attr_changed(struct fscache_cookie *cookie)
{
	struct fscache_operation *op;
	struct fscache_object *object;

	_enter("%p", cookie);

	ASSERTCMP(cookie->def->type, !=, FSCACHE_COOKIE_TYPE_INDEX);

	fscache_stat(&fscache_n_attr_changed);

	op = kzalloc(sizeof(*op), GFP_KERNEL);
	if (!op) {
		fscache_stat(&fscache_n_attr_changed_nomem);
		_leave(" = -ENOMEM");
		return -ENOMEM;
	}

	fscache_operation_init(op, fscache_attr_changed_op, NULL);
	op->flags = FSCACHE_OP_ASYNC | (1 << FSCACHE_OP_EXCLUSIVE);

	spin_lock(&cookie->lock);

	if (hlist_empty(&cookie->backing_objects))
		goto nobufs;
	object = hlist_entry(cookie->backing_objects.first,
			     struct fscache_object, cookie_link);

	if (fscache_submit_exclusive_op(object, op) < 0)
		goto nobufs;
	spin_unlock(&cookie->lock);
	fscache_stat(&fscache_n_attr_changed_ok);
	fscache_put_operation(op);
	_leave(" = 0");
	return 0;

nobufs:
	spin_unlock(&cookie->lock);
	kfree(op);
	fscache_stat(&fscache_n_attr_changed_nobufs);
	_leave(" = %d", -ENOBUFS);
	return -ENOBUFS;
}
EXPORT_SYMBOL(__fscache_attr_changed);

/*
 * release a retrieval op reference
 */
static void fscache_release_retrieval_op(struct fscache_operation *_op)
{
	struct fscache_retrieval *op =
		container_of(_op, struct fscache_retrieval, op);

	_enter("{OP%x}", op->op.debug_id);

	ASSERTCMP(op->n_pages, ==, 0);

	fscache_hist(fscache_retrieval_histogram, op->start_time);
	if (op->context)
		fscache_put_context(op->op.object->cookie, op->context);

	_leave("");
}

/*
 * allocate a retrieval op
 */
static struct fscache_retrieval *fscache_alloc_retrieval(
	struct address_space *mapping,
	fscache_rw_complete_t end_io_func,
	void *context)
{
	struct fscache_retrieval *op;

	/* allocate a retrieval operation and attempt to submit it */
	op = kzalloc(sizeof(*op), GFP_NOIO);
	if (!op) {
		fscache_stat(&fscache_n_retrievals_nomem);
		return NULL;
	}

	fscache_operation_init(&op->op, NULL, fscache_release_retrieval_op);
	op->op.flags	= FSCACHE_OP_MYTHREAD | (1 << FSCACHE_OP_WAITING);
	op->mapping	= mapping;
	op->end_io_func	= end_io_func;
	op->context	= context;
	op->start_time	= jiffies;
	INIT_LIST_HEAD(&op->to_do);
	return op;
}

/*
 * wait for a deferred lookup to complete
 */
static int fscache_wait_for_deferred_lookup(struct fscache_cookie *cookie)
{
	unsigned long jif;

	_enter("");

	if (!test_bit(FSCACHE_COOKIE_LOOKING_UP, &cookie->flags)) {
		_leave(" = 0 [imm]");
		return 0;
	}

	fscache_stat(&fscache_n_retrievals_wait);

	jif = jiffies;
	if (wait_on_bit(&cookie->flags, FSCACHE_COOKIE_LOOKING_UP,
			fscache_wait_bit_interruptible,
			TASK_INTERRUPTIBLE) != 0) {
		fscache_stat(&fscache_n_retrievals_intr);
		_leave(" = -ERESTARTSYS");
		return -ERESTARTSYS;
	}

	ASSERT(!test_bit(FSCACHE_COOKIE_LOOKING_UP, &cookie->flags));

	smp_rmb();
	fscache_hist(fscache_retrieval_delay_histogram, jif);
	_leave(" = 0 [dly]");
	return 0;
}

/*
 * wait for an object to become active (or dead)
 */
static int fscache_wait_for_retrieval_activation(struct fscache_object *object,
						 struct fscache_retrieval *op,
						 atomic_t *stat_op_waits,
						 atomic_t *stat_object_dead)
{
	int ret;

	if (!test_bit(FSCACHE_OP_WAITING, &op->op.flags))
		goto check_if_dead;

	_debug(">>> WT");
	fscache_stat(stat_op_waits);
	if (wait_on_bit(&op->op.flags, FSCACHE_OP_WAITING,
			fscache_wait_bit_interruptible,
			TASK_INTERRUPTIBLE) < 0) {
		ret = fscache_cancel_op(&op->op);
		if (ret == 0)
			return -ERESTARTSYS;

		/* it's been removed from the pending queue by another party,
		 * so we should get to run shortly */
		wait_on_bit(&op->op.flags, FSCACHE_OP_WAITING,
			    fscache_wait_bit, TASK_UNINTERRUPTIBLE);
	}
	_debug("<<< GO");

check_if_dead:
	if (op->op.state == FSCACHE_OP_ST_CANCELLED) {
		fscache_stat(stat_object_dead);
		_leave(" = -ENOBUFS [cancelled]");
		return -ENOBUFS;
	}
	if (unlikely(fscache_object_is_dead(object))) {
		fscache_stat(stat_object_dead);
		return -ENOBUFS;
	}
	return 0;
}

/*
 * read a page from the cache or allocate a block in which to store it
 * - we return:
 *   -ENOMEM	- out of memory, nothing done
 *   -ERESTARTSYS - interrupted
 *   -ENOBUFS	- no backing object available in which to cache the block
 *   -ENODATA	- no data available in the backing object for this block
 *   0		- dispatched a read - it'll call end_io_func() when finished
 */
int __fscache_read_or_alloc_page(struct fscache_cookie *cookie,
				 struct page *page,
				 fscache_rw_complete_t end_io_func,
				 void *context,
				 gfp_t gfp)
{
	struct fscache_retrieval *op;
	struct fscache_object *object;
	int ret;

	_enter("%p,%p,,,", cookie, page);

	fscache_stat(&fscache_n_retrievals);

	if (hlist_empty(&cookie->backing_objects))
		goto nobufs;

	ASSERTCMP(cookie->def->type, !=, FSCACHE_COOKIE_TYPE_INDEX);
	ASSERTCMP(page, !=, NULL);

	if (fscache_wait_for_deferred_lookup(cookie) < 0)
		return -ERESTARTSYS;

	op = fscache_alloc_retrieval(page->mapping, end_io_func, context);
	if (!op) {
		_leave(" = -ENOMEM");
		return -ENOMEM;
	}
	op->n_pages = 1;

	spin_lock(&cookie->lock);

	if (hlist_empty(&cookie->backing_objects))
		goto nobufs_unlock;
	object = hlist_entry(cookie->backing_objects.first,
			     struct fscache_object, cookie_link);

	ASSERTCMP(object->state, >, FSCACHE_OBJECT_LOOKING_UP);

	atomic_inc(&object->n_reads);
	__set_bit(FSCACHE_OP_DEC_READ_CNT, &op->op.flags);

	if (fscache_submit_op(object, &op->op) < 0)
		goto nobufs_unlock_dec;
	spin_unlock(&cookie->lock);

	fscache_stat(&fscache_n_retrieval_ops);

	/* pin the netfs read context in case we need to do the actual netfs
	 * read because we've encountered a cache read failure */
	fscache_get_context(object->cookie, op->context);

	/* we wait for the operation to become active, and then process it
	 * *here*, in this thread, and not in the thread pool */
	ret = fscache_wait_for_retrieval_activation(
		object, op,
		__fscache_stat(&fscache_n_retrieval_op_waits),
		__fscache_stat(&fscache_n_retrievals_object_dead));
	if (ret < 0)
		goto error;

	/* ask the cache to honour the operation */
	if (test_bit(FSCACHE_COOKIE_NO_DATA_YET, &object->cookie->flags)) {
		fscache_stat(&fscache_n_cop_allocate_page);
		ret = object->cache->ops->allocate_page(op, page, gfp);
		fscache_stat_d(&fscache_n_cop_allocate_page);
		if (ret == 0)
			ret = -ENODATA;
	} else {
		fscache_stat(&fscache_n_cop_read_or_alloc_page);
		ret = object->cache->ops->read_or_alloc_page(op, page, gfp);
		fscache_stat_d(&fscache_n_cop_read_or_alloc_page);
	}

error:
	if (ret == -ENOMEM)
		fscache_stat(&fscache_n_retrievals_nomem);
	else if (ret == -ERESTARTSYS)
		fscache_stat(&fscache_n_retrievals_intr);
	else if (ret == -ENODATA)
		fscache_stat(&fscache_n_retrievals_nodata);
	else if (ret < 0)
		fscache_stat(&fscache_n_retrievals_nobufs);
	else
		fscache_stat(&fscache_n_retrievals_ok);

	fscache_put_retrieval(op);
	_leave(" = %d", ret);
	return ret;

nobufs_unlock_dec:
	atomic_dec(&object->n_reads);
nobufs_unlock:
	spin_unlock(&cookie->lock);
	kfree(op);
nobufs:
	fscache_stat(&fscache_n_retrievals_nobufs);
	_leave(" = -ENOBUFS");
	return -ENOBUFS;
}
EXPORT_SYMBOL(__fscache_read_or_alloc_page);

/*
 * read a list of page from the cache or allocate a block in which to store
 * them
 * - we return:
 *   -ENOMEM	- out of memory, some pages may be being read
 *   -ERESTARTSYS - interrupted, some pages may be being read
 *   -ENOBUFS	- no backing object or space available in which to cache any
 *                pages not being read
 *   -ENODATA	- no data available in the backing object for some or all of
 *                the pages
 *   0		- dispatched a read on all pages
 *
 * end_io_func() will be called for each page read from the cache as it is
 * finishes being read
 *
 * any pages for which a read is dispatched will be removed from pages and
 * nr_pages
 */
int __fscache_read_or_alloc_pages(struct fscache_cookie *cookie,
				  struct address_space *mapping,
				  struct list_head *pages,
				  unsigned *nr_pages,
				  fscache_rw_complete_t end_io_func,
				  void *context,
				  gfp_t gfp)
{
	struct fscache_retrieval *op;
	struct fscache_object *object;
	int ret;

	_enter("%p,,%d,,,", cookie, *nr_pages);

	fscache_stat(&fscache_n_retrievals);

	if (hlist_empty(&cookie->backing_objects))
		goto nobufs;

	ASSERTCMP(cookie->def->type, !=, FSCACHE_COOKIE_TYPE_INDEX);
	ASSERTCMP(*nr_pages, >, 0);
	ASSERT(!list_empty(pages));

	if (fscache_wait_for_deferred_lookup(cookie) < 0)
		return -ERESTARTSYS;

	op = fscache_alloc_retrieval(mapping, end_io_func, context);
	if (!op)
		return -ENOMEM;
	op->n_pages = *nr_pages;

	spin_lock(&cookie->lock);

	if (hlist_empty(&cookie->backing_objects))
		goto nobufs_unlock;
	object = hlist_entry(cookie->backing_objects.first,
			     struct fscache_object, cookie_link);

	atomic_inc(&object->n_reads);
	__set_bit(FSCACHE_OP_DEC_READ_CNT, &op->op.flags);

	if (fscache_submit_op(object, &op->op) < 0)
		goto nobufs_unlock_dec;
	spin_unlock(&cookie->lock);

	fscache_stat(&fscache_n_retrieval_ops);

	/* pin the netfs read context in case we need to do the actual netfs
	 * read because we've encountered a cache read failure */
	fscache_get_context(object->cookie, op->context);

	/* we wait for the operation to become active, and then process it
	 * *here*, in this thread, and not in the thread pool */
	ret = fscache_wait_for_retrieval_activation(
		object, op,
		__fscache_stat(&fscache_n_retrieval_op_waits),
		__fscache_stat(&fscache_n_retrievals_object_dead));
	if (ret < 0)
		goto error;

	/* ask the cache to honour the operation */
	if (test_bit(FSCACHE_COOKIE_NO_DATA_YET, &object->cookie->flags)) {
		fscache_stat(&fscache_n_cop_allocate_pages);
		ret = object->cache->ops->allocate_pages(
			op, pages, nr_pages, gfp);
		fscache_stat_d(&fscache_n_cop_allocate_pages);
	} else {
		fscache_stat(&fscache_n_cop_read_or_alloc_pages);
		ret = object->cache->ops->read_or_alloc_pages(
			op, pages, nr_pages, gfp);
		fscache_stat_d(&fscache_n_cop_read_or_alloc_pages);
	}

error:
	if (ret == -ENOMEM)
		fscache_stat(&fscache_n_retrievals_nomem);
	else if (ret == -ERESTARTSYS)
		fscache_stat(&fscache_n_retrievals_intr);
	else if (ret == -ENODATA)
		fscache_stat(&fscache_n_retrievals_nodata);
	else if (ret < 0)
		fscache_stat(&fscache_n_retrievals_nobufs);
	else
		fscache_stat(&fscache_n_retrievals_ok);

	fscache_put_retrieval(op);
	_leave(" = %d", ret);
	return ret;

nobufs_unlock_dec:
	atomic_dec(&object->n_reads);
nobufs_unlock:
	spin_unlock(&cookie->lock);
	kfree(op);
nobufs:
	fscache_stat(&fscache_n_retrievals_nobufs);
	_leave(" = -ENOBUFS");
	return -ENOBUFS;
}
EXPORT_SYMBOL(__fscache_read_or_alloc_pages);

/*
 * allocate a block in the cache on which to store a page
 * - we return:
 *   -ENOMEM	- out of memory, nothing done
 *   -ERESTARTSYS - interrupted
 *   -ENOBUFS	- no backing object available in which to cache the block
 *   0		- block allocated
 */
int __fscache_alloc_page(struct fscache_cookie *cookie,
			 struct page *page,
			 gfp_t gfp)
{
	struct fscache_retrieval *op;
	struct fscache_object *object;
	int ret;

	_enter("%p,%p,,,", cookie, page);

	fscache_stat(&fscache_n_allocs);

	if (hlist_empty(&cookie->backing_objects))
		goto nobufs;

	ASSERTCMP(cookie->def->type, !=, FSCACHE_COOKIE_TYPE_INDEX);
	ASSERTCMP(page, !=, NULL);

	if (fscache_wait_for_deferred_lookup(cookie) < 0)
		return -ERESTARTSYS;

	op = fscache_alloc_retrieval(page->mapping, NULL, NULL);
	if (!op)
		return -ENOMEM;
	op->n_pages = 1;

	spin_lock(&cookie->lock);

	if (hlist_empty(&cookie->backing_objects))
		goto nobufs_unlock;
	object = hlist_entry(cookie->backing_objects.first,
			     struct fscache_object, cookie_link);

	if (fscache_submit_op(object, &op->op) < 0)
		goto nobufs_unlock;
	spin_unlock(&cookie->lock);

	fscache_stat(&fscache_n_alloc_ops);

	ret = fscache_wait_for_retrieval_activation(
		object, op,
		__fscache_stat(&fscache_n_alloc_op_waits),
		__fscache_stat(&fscache_n_allocs_object_dead));
	if (ret < 0)
		goto error;

	/* ask the cache to honour the operation */
	fscache_stat(&fscache_n_cop_allocate_page);
	ret = object->cache->ops->allocate_page(op, page, gfp);
	fscache_stat_d(&fscache_n_cop_allocate_page);

error:
	if (ret == -ERESTARTSYS)
		fscache_stat(&fscache_n_allocs_intr);
	else if (ret < 0)
		fscache_stat(&fscache_n_allocs_nobufs);
	else
		fscache_stat(&fscache_n_allocs_ok);

	fscache_put_retrieval(op);
	_leave(" = %d", ret);
	return ret;

nobufs_unlock:
	spin_unlock(&cookie->lock);
	kfree(op);
nobufs:
	fscache_stat(&fscache_n_allocs_nobufs);
	_leave(" = -ENOBUFS");
	return -ENOBUFS;
}
EXPORT_SYMBOL(__fscache_alloc_page);

/*
 * release a write op reference
 */
static void fscache_release_write_op(struct fscache_operation *_op)
{
	_enter("{OP%x}", _op->debug_id);
}

/*
 * perform the background storage of a page into the cache
 */
static void fscache_write_op(struct fscache_operation *_op)
{
	struct fscache_storage *op =
		container_of(_op, struct fscache_storage, op);
	struct fscache_object *object = op->op.object;
	struct fscache_cookie *cookie;
	struct page *page;
	unsigned n;
	void *results[1];
	int ret;

	_enter("{OP%x,%d}", op->op.debug_id, atomic_read(&op->op.usage));

	spin_lock(&object->lock);
	cookie = object->cookie;

	if (!fscache_object_is_active(object) || !cookie) {
		spin_unlock(&object->lock);
		_leave("");
		return;
	}

	spin_lock(&cookie->stores_lock);

	fscache_stat(&fscache_n_store_calls);

	/* find a page to store */
	page = NULL;
	n = radix_tree_gang_lookup_tag(&cookie->stores, results, 0, 1,
				       FSCACHE_COOKIE_PENDING_TAG);
	if (n != 1)
		goto superseded;
	page = results[0];
	_debug("gang %d [%lx]", n, page->index);
	if (page->index > op->store_limit) {
		fscache_stat(&fscache_n_store_pages_over_limit);
		goto superseded;
	}

	radix_tree_tag_set(&cookie->stores, page->index,
			   FSCACHE_COOKIE_STORING_TAG);
	radix_tree_tag_clear(&cookie->stores, page->index,
			     FSCACHE_COOKIE_PENDING_TAG);

	spin_unlock(&cookie->stores_lock);
	spin_unlock(&object->lock);

	fscache_stat(&fscache_n_store_pages);
	fscache_stat(&fscache_n_cop_write_page);
	ret = object->cache->ops->write_page(op, page);
	fscache_stat_d(&fscache_n_cop_write_page);
	fscache_end_page_write(object, page);
	if (ret < 0) {
		fscache_abort_object(object);
		fscache_op_complete(&op->op);
	} else {
		fscache_enqueue_operation(&op->op);
	}

	_leave("");
	return;

superseded:
	/* this writer is going away and there aren't any more things to
	 * write */
	_debug("cease");
	spin_unlock(&cookie->stores_lock);
	clear_bit(FSCACHE_OBJECT_PENDING_WRITE, &object->flags);
	spin_unlock(&object->lock);
	fscache_op_complete(&op->op);
	_leave("");
}

/*
 * request a page be stored in the cache
 * - returns:
 *   -ENOMEM	- out of memory, nothing done
 *   -ENOBUFS	- no backing object available in which to cache the page
 *   0		- dispatched a write - it'll call end_io_func() when finished
 *
 * if the cookie still has a backing object at this point, that object can be
 * in one of a few states with respect to storage processing:
 *
 *  (1) negative lookup, object not yet created (FSCACHE_COOKIE_CREATING is
 *      set)
 *
 *	(a) no writes yet (set FSCACHE_COOKIE_PENDING_FILL and queue deferred
 *	    fill op)
 *
 *	(b) writes deferred till post-creation (mark page for writing and
 *	    return immediately)
 *
 *  (2) negative lookup, object created, initial fill being made from netfs
 *      (FSCACHE_COOKIE_INITIAL_FILL is set)
 *
 *	(a) fill point not yet reached this page (mark page for writing and
 *          return)
 *
 *	(b) fill point passed this page (queue op to store this page)
 *
 *  (3) object extant (queue op to store this page)
 *
 * any other state is invalid
 */
int __fscache_write_page(struct fscache_cookie *cookie,
			 struct page *page,
			 gfp_t gfp)
{
	struct fscache_storage *op;
	struct fscache_object *object;
	int ret;

	_enter("%p,%x,", cookie, (u32) page->flags);

	ASSERTCMP(cookie->def->type, !=, FSCACHE_COOKIE_TYPE_INDEX);
	ASSERT(PageFsCache(page));

	fscache_stat(&fscache_n_stores);

	op = kzalloc(sizeof(*op), GFP_NOIO | __GFP_NOMEMALLOC | __GFP_NORETRY);
	if (!op)
		goto nomem;

	fscache_operation_init(&op->op, fscache_write_op,
			       fscache_release_write_op);
	op->op.flags = FSCACHE_OP_ASYNC | (1 << FSCACHE_OP_WAITING);

	ret = radix_tree_preload(gfp & ~__GFP_HIGHMEM);
	if (ret < 0)
		goto nomem_free;

	ret = -ENOBUFS;
	spin_lock(&cookie->lock);

	if (hlist_empty(&cookie->backing_objects))
		goto nobufs;
	object = hlist_entry(cookie->backing_objects.first,
			     struct fscache_object, cookie_link);
	if (test_bit(FSCACHE_IOERROR, &object->cache->flags))
		goto nobufs;

	/* add the page to the pending-storage radix tree on the backing
	 * object */
	spin_lock(&object->lock);
	spin_lock(&cookie->stores_lock);

	_debug("store limit %llx", (unsigned long long) object->store_limit);

	ret = radix_tree_insert(&cookie->stores, page->index, page);
	if (ret < 0) {
		if (ret == -EEXIST)
			goto already_queued;
		_debug("insert failed %d", ret);
		goto nobufs_unlock_obj;
	}

	radix_tree_tag_set(&cookie->stores, page->index,
			   FSCACHE_COOKIE_PENDING_TAG);
	page_cache_get(page);

	/* we only want one writer at a time, but we do need to queue new
	 * writers after exclusive ops */
	if (test_and_set_bit(FSCACHE_OBJECT_PENDING_WRITE, &object->flags))
		goto already_pending;

	spin_unlock(&cookie->stores_lock);
	spin_unlock(&object->lock);

	op->op.debug_id	= atomic_inc_return(&fscache_op_debug_id);
	op->store_limit = object->store_limit;

	if (fscache_submit_op(object, &op->op) < 0)
		goto submit_failed;

	spin_unlock(&cookie->lock);
	radix_tree_preload_end();
	fscache_stat(&fscache_n_store_ops);
	fscache_stat(&fscache_n_stores_ok);

	/* the work queue now carries its own ref on the object */
	fscache_put_operation(&op->op);
	_leave(" = 0");
	return 0;

already_queued:
	fscache_stat(&fscache_n_stores_again);
already_pending:
	spin_unlock(&cookie->stores_lock);
	spin_unlock(&object->lock);
	spin_unlock(&cookie->lock);
	radix_tree_preload_end();
	kfree(op);
	fscache_stat(&fscache_n_stores_ok);
	_leave(" = 0");
	return 0;

submit_failed:
	spin_lock(&cookie->stores_lock);
	radix_tree_delete(&cookie->stores, page->index);
	spin_unlock(&cookie->stores_lock);
	page_cache_release(page);
	ret = -ENOBUFS;
	goto nobufs;

nobufs_unlock_obj:
	spin_unlock(&cookie->stores_lock);
	spin_unlock(&object->lock);
nobufs:
	spin_unlock(&cookie->lock);
	radix_tree_preload_end();
	kfree(op);
	fscache_stat(&fscache_n_stores_nobufs);
	_leave(" = -ENOBUFS");
	return -ENOBUFS;

nomem_free:
	kfree(op);
nomem:
	fscache_stat(&fscache_n_stores_oom);
	_leave(" = -ENOMEM");
	return -ENOMEM;
}
EXPORT_SYMBOL(__fscache_write_page);

/*
 * remove a page from the cache
 */
void __fscache_uncache_page(struct fscache_cookie *cookie, struct page *page)
{
	struct fscache_object *object;

	_enter(",%p", page);

	ASSERTCMP(cookie->def->type, !=, FSCACHE_COOKIE_TYPE_INDEX);
	ASSERTCMP(page, !=, NULL);

	fscache_stat(&fscache_n_uncaches);

	/* cache withdrawal may beat us to it */
	if (!PageFsCache(page))
		goto done;

	/* get the object */
	spin_lock(&cookie->lock);

	if (hlist_empty(&cookie->backing_objects)) {
		ClearPageFsCache(page);
		goto done_unlock;
	}

	object = hlist_entry(cookie->backing_objects.first,
			     struct fscache_object, cookie_link);

	/* there might now be stuff on disk we could read */
	clear_bit(FSCACHE_COOKIE_NO_DATA_YET, &cookie->flags);

	/* only invoke the cache backend if we managed to mark the page
	 * uncached here; this deals with synchronisation vs withdrawal */
	if (TestClearPageFsCache(page) &&
	    object->cache->ops->uncache_page) {
		/* the cache backend releases the cookie lock */
		fscache_stat(&fscache_n_cop_uncache_page);
		object->cache->ops->uncache_page(object, page);
		fscache_stat_d(&fscache_n_cop_uncache_page);
		goto done;
	}

done_unlock:
	spin_unlock(&cookie->lock);
done:
	_leave("");
}
EXPORT_SYMBOL(__fscache_uncache_page);

/**
 * fscache_mark_page_cached - Mark a page as being cached
 * @op: The retrieval op pages are being marked for
 * @page: The page to be marked
 *
 * Mark a netfs page as being cached.  After this is called, the netfs
 * must call fscache_uncache_page() to remove the mark.
 */
void fscache_mark_page_cached(struct fscache_retrieval *op, struct page *page)
{
	struct fscache_cookie *cookie = op->op.object->cookie;

#ifdef CONFIG_FSCACHE_STATS
	atomic_inc(&fscache_n_marks);
#endif

	_debug("- mark %p{%lx}", page, page->index);
	if (TestSetPageFsCache(page)) {
		static bool once_only;
		if (!once_only) {
			once_only = true;
			printk(KERN_WARNING "FS-Cache:"
			       " Cookie type %s marked page %lx"
			       " multiple times\n",
			       cookie->def->name, page->index);
		}
	}

	if (cookie->def->mark_page_cached)
		cookie->def->mark_page_cached(cookie->netfs_data,
					      op->mapping, page);
}
EXPORT_SYMBOL(fscache_mark_page_cached);

/**
 * fscache_mark_pages_cached - Mark pages as being cached
 * @op: The retrieval op pages are being marked for
 * @pagevec: The pages to be marked
 *
 * Mark a bunch of netfs pages as being cached.  After this is called,
 * the netfs must call fscache_uncache_page() to remove the mark.
 */
void fscache_mark_pages_cached(struct fscache_retrieval *op,
			       struct pagevec *pagevec)
{
	unsigned long loop;

	for (loop = 0; loop < pagevec->nr; loop++)
		fscache_mark_page_cached(op, pagevec->pages[loop]);

	pagevec_reinit(pagevec);
}
EXPORT_SYMBOL(fscache_mark_pages_cached);

/*
 * Uncache all the pages in an inode that are marked PG_fscache, assuming them
 * to be associated with the given cookie.
 */
void __fscache_uncache_all_inode_pages(struct fscache_cookie *cookie,
				       struct inode *inode)
{
	struct address_space *mapping = inode->i_mapping;
	struct pagevec pvec;
	pgoff_t next;
	int i;

	_enter("%p,%p", cookie, inode);

	if (!mapping || mapping->nrpages == 0) {
		_leave(" [no pages]");
		return;
	}

	pagevec_init(&pvec, 0);
	next = 0;
	do {
		if (!pagevec_lookup(&pvec, mapping, next, PAGEVEC_SIZE))
			break;
		for (i = 0; i < pagevec_count(&pvec); i++) {
			struct page *page = pvec.pages[i];
			next = page->index;
			if (PageFsCache(page)) {
				__fscache_wait_on_page_write(cookie, page);
				__fscache_uncache_page(cookie, page);
			}
		}
		pagevec_release(&pvec);
		cond_resched();
	} while (++next);

	_leave("");
}
EXPORT_SYMBOL(__fscache_uncache_all_inode_pages);