forked from torvalds/linux
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathspace-info.c
1852 lines (1658 loc) · 57.7 KB
/
space-info.c
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
// SPDX-License-Identifier: GPL-2.0
#include "misc.h"
#include "ctree.h"
#include "space-info.h"
#include "sysfs.h"
#include "volumes.h"
#include "free-space-cache.h"
#include "ordered-data.h"
#include "transaction.h"
#include "block-group.h"
#include "zoned.h"
#include "fs.h"
#include "accessors.h"
#include "extent-tree.h"
/*
* HOW DOES SPACE RESERVATION WORK
*
* If you want to know about delalloc specifically, there is a separate comment
* for that with the delalloc code. This comment is about how the whole system
* works generally.
*
* BASIC CONCEPTS
*
* 1) space_info. This is the ultimate arbiter of how much space we can use.
* There's a description of the bytes_ fields with the struct declaration,
* refer to that for specifics on each field. Suffice it to say that for
* reservations we care about total_bytes - SUM(space_info->bytes_) when
* determining if there is space to make an allocation. There is a space_info
* for METADATA, SYSTEM, and DATA areas.
*
* 2) block_rsv's. These are basically buckets for every different type of
* metadata reservation we have. You can see the comment in the block_rsv
* code on the rules for each type, but generally block_rsv->reserved is how
* much space is accounted for in space_info->bytes_may_use.
*
* 3) btrfs_calc*_size. These are the worst case calculations we used based
* on the number of items we will want to modify. We have one for changing
* items, and one for inserting new items. Generally we use these helpers to
* determine the size of the block reserves, and then use the actual bytes
* values to adjust the space_info counters.
*
* MAKING RESERVATIONS, THE NORMAL CASE
*
* We call into either btrfs_reserve_data_bytes() or
* btrfs_reserve_metadata_bytes(), depending on which we're looking for, with
* num_bytes we want to reserve.
*
* ->reserve
* space_info->bytes_may_reserve += num_bytes
*
* ->extent allocation
* Call btrfs_add_reserved_bytes() which does
* space_info->bytes_may_reserve -= num_bytes
* space_info->bytes_reserved += extent_bytes
*
* ->insert reference
* Call btrfs_update_block_group() which does
* space_info->bytes_reserved -= extent_bytes
* space_info->bytes_used += extent_bytes
*
* MAKING RESERVATIONS, FLUSHING NORMALLY (non-priority)
*
* Assume we are unable to simply make the reservation because we do not have
* enough space
*
* -> __reserve_bytes
* create a reserve_ticket with ->bytes set to our reservation, add it to
* the tail of space_info->tickets, kick async flush thread
*
* ->handle_reserve_ticket
* wait on ticket->wait for ->bytes to be reduced to 0, or ->error to be set
* on the ticket.
*
* -> btrfs_async_reclaim_metadata_space/btrfs_async_reclaim_data_space
* Flushes various things attempting to free up space.
*
* -> btrfs_try_granting_tickets()
* This is called by anything that either subtracts space from
* space_info->bytes_may_use, ->bytes_pinned, etc, or adds to the
* space_info->total_bytes. This loops through the ->priority_tickets and
* then the ->tickets list checking to see if the reservation can be
* completed. If it can the space is added to space_info->bytes_may_use and
* the ticket is woken up.
*
* -> ticket wakeup
* Check if ->bytes == 0, if it does we got our reservation and we can carry
* on, if not return the appropriate error (ENOSPC, but can be EINTR if we
* were interrupted.)
*
* MAKING RESERVATIONS, FLUSHING HIGH PRIORITY
*
* Same as the above, except we add ourselves to the
* space_info->priority_tickets, and we do not use ticket->wait, we simply
* call flush_space() ourselves for the states that are safe for us to call
* without deadlocking and hope for the best.
*
* THE FLUSHING STATES
*
* Generally speaking we will have two cases for each state, a "nice" state
* and a "ALL THE THINGS" state. In btrfs we delay a lot of work in order to
* reduce the locking over head on the various trees, and even to keep from
* doing any work at all in the case of delayed refs. Each of these delayed
* things however hold reservations, and so letting them run allows us to
* reclaim space so we can make new reservations.
*
* FLUSH_DELAYED_ITEMS
* Every inode has a delayed item to update the inode. Take a simple write
* for example, we would update the inode item at write time to update the
* mtime, and then again at finish_ordered_io() time in order to update the
* isize or bytes. We keep these delayed items to coalesce these operations
* into a single operation done on demand. These are an easy way to reclaim
* metadata space.
*
* FLUSH_DELALLOC
* Look at the delalloc comment to get an idea of how much space is reserved
* for delayed allocation. We can reclaim some of this space simply by
* running delalloc, but usually we need to wait for ordered extents to
* reclaim the bulk of this space.
*
* FLUSH_DELAYED_REFS
* We have a block reserve for the outstanding delayed refs space, and every
* delayed ref operation holds a reservation. Running these is a quick way
* to reclaim space, but we want to hold this until the end because COW can
* churn a lot and we can avoid making some extent tree modifications if we
* are able to delay for as long as possible.
*
* ALLOC_CHUNK
* We will skip this the first time through space reservation, because of
* overcommit and we don't want to have a lot of useless metadata space when
* our worst case reservations will likely never come true.
*
* RUN_DELAYED_IPUTS
* If we're freeing inodes we're likely freeing checksums, file extent
* items, and extent tree items. Loads of space could be freed up by these
* operations, however they won't be usable until the transaction commits.
*
* COMMIT_TRANS
* This will commit the transaction. Historically we had a lot of logic
* surrounding whether or not we'd commit the transaction, but this waits born
* out of a pre-tickets era where we could end up committing the transaction
* thousands of times in a row without making progress. Now thanks to our
* ticketing system we know if we're not making progress and can error
* everybody out after a few commits rather than burning the disk hoping for
* a different answer.
*
* OVERCOMMIT
*
* Because we hold so many reservations for metadata we will allow you to
* reserve more space than is currently free in the currently allocate
* metadata space. This only happens with metadata, data does not allow
* overcommitting.
*
* You can see the current logic for when we allow overcommit in
* btrfs_can_overcommit(), but it only applies to unallocated space. If there
* is no unallocated space to be had, all reservations are kept within the
* free space in the allocated metadata chunks.
*
* Because of overcommitting, you generally want to use the
* btrfs_can_overcommit() logic for metadata allocations, as it does the right
* thing with or without extra unallocated space.
*/
u64 __pure btrfs_space_info_used(struct btrfs_space_info *s_info,
bool may_use_included)
{
ASSERT(s_info);
return s_info->bytes_used + s_info->bytes_reserved +
s_info->bytes_pinned + s_info->bytes_readonly +
s_info->bytes_zone_unusable +
(may_use_included ? s_info->bytes_may_use : 0);
}
/*
* after adding space to the filesystem, we need to clear the full flags
* on all the space infos.
*/
void btrfs_clear_space_info_full(struct btrfs_fs_info *info)
{
struct list_head *head = &info->space_info;
struct btrfs_space_info *found;
list_for_each_entry(found, head, list)
found->full = 0;
}
/*
* Block groups with more than this value (percents) of unusable space will be
* scheduled for background reclaim.
*/
#define BTRFS_DEFAULT_ZONED_RECLAIM_THRESH (75)
/*
* Calculate chunk size depending on volume type (regular or zoned).
*/
static u64 calc_chunk_size(const struct btrfs_fs_info *fs_info, u64 flags)
{
if (btrfs_is_zoned(fs_info))
return fs_info->zone_size;
ASSERT(flags & BTRFS_BLOCK_GROUP_TYPE_MASK);
if (flags & BTRFS_BLOCK_GROUP_DATA)
return BTRFS_MAX_DATA_CHUNK_SIZE;
else if (flags & BTRFS_BLOCK_GROUP_SYSTEM)
return SZ_32M;
/* Handle BTRFS_BLOCK_GROUP_METADATA */
if (fs_info->fs_devices->total_rw_bytes > 50ULL * SZ_1G)
return SZ_1G;
return SZ_256M;
}
/*
* Update default chunk size.
*/
void btrfs_update_space_info_chunk_size(struct btrfs_space_info *space_info,
u64 chunk_size)
{
WRITE_ONCE(space_info->chunk_size, chunk_size);
}
static int create_space_info(struct btrfs_fs_info *info, u64 flags)
{
struct btrfs_space_info *space_info;
int i;
int ret;
space_info = kzalloc(sizeof(*space_info), GFP_NOFS);
if (!space_info)
return -ENOMEM;
for (i = 0; i < BTRFS_NR_RAID_TYPES; i++)
INIT_LIST_HEAD(&space_info->block_groups[i]);
init_rwsem(&space_info->groups_sem);
spin_lock_init(&space_info->lock);
space_info->flags = flags & BTRFS_BLOCK_GROUP_TYPE_MASK;
space_info->force_alloc = CHUNK_ALLOC_NO_FORCE;
INIT_LIST_HEAD(&space_info->ro_bgs);
INIT_LIST_HEAD(&space_info->tickets);
INIT_LIST_HEAD(&space_info->priority_tickets);
space_info->clamp = 1;
btrfs_update_space_info_chunk_size(space_info, calc_chunk_size(info, flags));
if (btrfs_is_zoned(info))
space_info->bg_reclaim_threshold = BTRFS_DEFAULT_ZONED_RECLAIM_THRESH;
ret = btrfs_sysfs_add_space_info_type(info, space_info);
if (ret)
return ret;
list_add(&space_info->list, &info->space_info);
if (flags & BTRFS_BLOCK_GROUP_DATA)
info->data_sinfo = space_info;
return ret;
}
int btrfs_init_space_info(struct btrfs_fs_info *fs_info)
{
struct btrfs_super_block *disk_super;
u64 features;
u64 flags;
int mixed = 0;
int ret;
disk_super = fs_info->super_copy;
if (!btrfs_super_root(disk_super))
return -EINVAL;
features = btrfs_super_incompat_flags(disk_super);
if (features & BTRFS_FEATURE_INCOMPAT_MIXED_GROUPS)
mixed = 1;
flags = BTRFS_BLOCK_GROUP_SYSTEM;
ret = create_space_info(fs_info, flags);
if (ret)
goto out;
if (mixed) {
flags = BTRFS_BLOCK_GROUP_METADATA | BTRFS_BLOCK_GROUP_DATA;
ret = create_space_info(fs_info, flags);
} else {
flags = BTRFS_BLOCK_GROUP_METADATA;
ret = create_space_info(fs_info, flags);
if (ret)
goto out;
flags = BTRFS_BLOCK_GROUP_DATA;
ret = create_space_info(fs_info, flags);
}
out:
return ret;
}
void btrfs_add_bg_to_space_info(struct btrfs_fs_info *info,
struct btrfs_block_group *block_group)
{
struct btrfs_space_info *found;
int factor, index;
factor = btrfs_bg_type_to_factor(block_group->flags);
found = btrfs_find_space_info(info, block_group->flags);
ASSERT(found);
spin_lock(&found->lock);
found->total_bytes += block_group->length;
found->disk_total += block_group->length * factor;
found->bytes_used += block_group->used;
found->disk_used += block_group->used * factor;
found->bytes_readonly += block_group->bytes_super;
found->bytes_zone_unusable += block_group->zone_unusable;
if (block_group->length > 0)
found->full = 0;
btrfs_try_granting_tickets(info, found);
spin_unlock(&found->lock);
block_group->space_info = found;
index = btrfs_bg_flags_to_raid_index(block_group->flags);
down_write(&found->groups_sem);
list_add_tail(&block_group->list, &found->block_groups[index]);
up_write(&found->groups_sem);
}
struct btrfs_space_info *btrfs_find_space_info(struct btrfs_fs_info *info,
u64 flags)
{
struct list_head *head = &info->space_info;
struct btrfs_space_info *found;
flags &= BTRFS_BLOCK_GROUP_TYPE_MASK;
list_for_each_entry(found, head, list) {
if (found->flags & flags)
return found;
}
return NULL;
}
static u64 calc_available_free_space(struct btrfs_fs_info *fs_info,
struct btrfs_space_info *space_info,
enum btrfs_reserve_flush_enum flush)
{
u64 profile;
u64 avail;
int factor;
if (space_info->flags & BTRFS_BLOCK_GROUP_SYSTEM)
profile = btrfs_system_alloc_profile(fs_info);
else
profile = btrfs_metadata_alloc_profile(fs_info);
avail = atomic64_read(&fs_info->free_chunk_space);
/*
* If we have dup, raid1 or raid10 then only half of the free
* space is actually usable. For raid56, the space info used
* doesn't include the parity drive, so we don't have to
* change the math
*/
factor = btrfs_bg_type_to_factor(profile);
avail = div_u64(avail, factor);
/*
* If we aren't flushing all things, let us overcommit up to
* 1/2th of the space. If we can flush, don't let us overcommit
* too much, let it overcommit up to 1/8 of the space.
*/
if (flush == BTRFS_RESERVE_FLUSH_ALL)
avail >>= 3;
else
avail >>= 1;
return avail;
}
int btrfs_can_overcommit(struct btrfs_fs_info *fs_info,
struct btrfs_space_info *space_info, u64 bytes,
enum btrfs_reserve_flush_enum flush)
{
u64 avail;
u64 used;
/* Don't overcommit when in mixed mode */
if (space_info->flags & BTRFS_BLOCK_GROUP_DATA)
return 0;
used = btrfs_space_info_used(space_info, true);
if (test_bit(BTRFS_FS_ACTIVE_ZONE_TRACKING, &fs_info->flags) &&
(space_info->flags & BTRFS_BLOCK_GROUP_METADATA))
avail = 0;
else
avail = calc_available_free_space(fs_info, space_info, flush);
if (used + bytes < space_info->total_bytes + avail)
return 1;
return 0;
}
static void remove_ticket(struct btrfs_space_info *space_info,
struct reserve_ticket *ticket)
{
if (!list_empty(&ticket->list)) {
list_del_init(&ticket->list);
ASSERT(space_info->reclaim_size >= ticket->bytes);
space_info->reclaim_size -= ticket->bytes;
}
}
/*
* This is for space we already have accounted in space_info->bytes_may_use, so
* basically when we're returning space from block_rsv's.
*/
void btrfs_try_granting_tickets(struct btrfs_fs_info *fs_info,
struct btrfs_space_info *space_info)
{
struct list_head *head;
enum btrfs_reserve_flush_enum flush = BTRFS_RESERVE_NO_FLUSH;
lockdep_assert_held(&space_info->lock);
head = &space_info->priority_tickets;
again:
while (!list_empty(head)) {
struct reserve_ticket *ticket;
u64 used = btrfs_space_info_used(space_info, true);
ticket = list_first_entry(head, struct reserve_ticket, list);
/* Check and see if our ticket can be satisfied now. */
if ((used + ticket->bytes <= space_info->total_bytes) ||
btrfs_can_overcommit(fs_info, space_info, ticket->bytes,
flush)) {
btrfs_space_info_update_bytes_may_use(fs_info,
space_info,
ticket->bytes);
remove_ticket(space_info, ticket);
ticket->bytes = 0;
space_info->tickets_id++;
wake_up(&ticket->wait);
} else {
break;
}
}
if (head == &space_info->priority_tickets) {
head = &space_info->tickets;
flush = BTRFS_RESERVE_FLUSH_ALL;
goto again;
}
}
#define DUMP_BLOCK_RSV(fs_info, rsv_name) \
do { \
struct btrfs_block_rsv *__rsv = &(fs_info)->rsv_name; \
spin_lock(&__rsv->lock); \
btrfs_info(fs_info, #rsv_name ": size %llu reserved %llu", \
__rsv->size, __rsv->reserved); \
spin_unlock(&__rsv->lock); \
} while (0)
static const char *space_info_flag_to_str(const struct btrfs_space_info *space_info)
{
switch (space_info->flags) {
case BTRFS_BLOCK_GROUP_SYSTEM:
return "SYSTEM";
case BTRFS_BLOCK_GROUP_METADATA | BTRFS_BLOCK_GROUP_DATA:
return "DATA+METADATA";
case BTRFS_BLOCK_GROUP_DATA:
return "DATA";
case BTRFS_BLOCK_GROUP_METADATA:
return "METADATA";
default:
return "UNKNOWN";
}
}
static void dump_global_block_rsv(struct btrfs_fs_info *fs_info)
{
DUMP_BLOCK_RSV(fs_info, global_block_rsv);
DUMP_BLOCK_RSV(fs_info, trans_block_rsv);
DUMP_BLOCK_RSV(fs_info, chunk_block_rsv);
DUMP_BLOCK_RSV(fs_info, delayed_block_rsv);
DUMP_BLOCK_RSV(fs_info, delayed_refs_rsv);
}
static void __btrfs_dump_space_info(struct btrfs_fs_info *fs_info,
struct btrfs_space_info *info)
{
const char *flag_str = space_info_flag_to_str(info);
lockdep_assert_held(&info->lock);
/* The free space could be negative in case of overcommit */
btrfs_info(fs_info, "space_info %s has %lld free, is %sfull",
flag_str,
(s64)(info->total_bytes - btrfs_space_info_used(info, true)),
info->full ? "" : "not ");
btrfs_info(fs_info,
"space_info total=%llu, used=%llu, pinned=%llu, reserved=%llu, may_use=%llu, readonly=%llu zone_unusable=%llu",
info->total_bytes, info->bytes_used, info->bytes_pinned,
info->bytes_reserved, info->bytes_may_use,
info->bytes_readonly, info->bytes_zone_unusable);
}
void btrfs_dump_space_info(struct btrfs_fs_info *fs_info,
struct btrfs_space_info *info, u64 bytes,
int dump_block_groups)
{
struct btrfs_block_group *cache;
int index = 0;
spin_lock(&info->lock);
__btrfs_dump_space_info(fs_info, info);
dump_global_block_rsv(fs_info);
spin_unlock(&info->lock);
if (!dump_block_groups)
return;
down_read(&info->groups_sem);
again:
list_for_each_entry(cache, &info->block_groups[index], list) {
spin_lock(&cache->lock);
btrfs_info(fs_info,
"block group %llu has %llu bytes, %llu used %llu pinned %llu reserved %llu zone_unusable %s",
cache->start, cache->length, cache->used, cache->pinned,
cache->reserved, cache->zone_unusable,
cache->ro ? "[readonly]" : "");
spin_unlock(&cache->lock);
btrfs_dump_free_space(cache, bytes);
}
if (++index < BTRFS_NR_RAID_TYPES)
goto again;
up_read(&info->groups_sem);
}
static inline u64 calc_reclaim_items_nr(const struct btrfs_fs_info *fs_info,
u64 to_reclaim)
{
u64 bytes;
u64 nr;
bytes = btrfs_calc_insert_metadata_size(fs_info, 1);
nr = div64_u64(to_reclaim, bytes);
if (!nr)
nr = 1;
return nr;
}
static inline u64 calc_delayed_refs_nr(const struct btrfs_fs_info *fs_info,
u64 to_reclaim)
{
const u64 bytes = btrfs_calc_delayed_ref_bytes(fs_info, 1);
u64 nr;
nr = div64_u64(to_reclaim, bytes);
if (!nr)
nr = 1;
return nr;
}
#define EXTENT_SIZE_PER_ITEM SZ_256K
/*
* shrink metadata reservation for delalloc
*/
static void shrink_delalloc(struct btrfs_fs_info *fs_info,
struct btrfs_space_info *space_info,
u64 to_reclaim, bool wait_ordered,
bool for_preempt)
{
struct btrfs_trans_handle *trans;
u64 delalloc_bytes;
u64 ordered_bytes;
u64 items;
long time_left;
int loops;
delalloc_bytes = percpu_counter_sum_positive(&fs_info->delalloc_bytes);
ordered_bytes = percpu_counter_sum_positive(&fs_info->ordered_bytes);
if (delalloc_bytes == 0 && ordered_bytes == 0)
return;
/* Calc the number of the pages we need flush for space reservation */
if (to_reclaim == U64_MAX) {
items = U64_MAX;
} else {
/*
* to_reclaim is set to however much metadata we need to
* reclaim, but reclaiming that much data doesn't really track
* exactly. What we really want to do is reclaim full inode's
* worth of reservations, however that's not available to us
* here. We will take a fraction of the delalloc bytes for our
* flushing loops and hope for the best. Delalloc will expand
* the amount we write to cover an entire dirty extent, which
* will reclaim the metadata reservation for that range. If
* it's not enough subsequent flush stages will be more
* aggressive.
*/
to_reclaim = max(to_reclaim, delalloc_bytes >> 3);
items = calc_reclaim_items_nr(fs_info, to_reclaim) * 2;
}
trans = current->journal_info;
/*
* If we are doing more ordered than delalloc we need to just wait on
* ordered extents, otherwise we'll waste time trying to flush delalloc
* that likely won't give us the space back we need.
*/
if (ordered_bytes > delalloc_bytes && !for_preempt)
wait_ordered = true;
loops = 0;
while ((delalloc_bytes || ordered_bytes) && loops < 3) {
u64 temp = min(delalloc_bytes, to_reclaim) >> PAGE_SHIFT;
long nr_pages = min_t(u64, temp, LONG_MAX);
int async_pages;
btrfs_start_delalloc_roots(fs_info, nr_pages, true);
/*
* We need to make sure any outstanding async pages are now
* processed before we continue. This is because things like
* sync_inode() try to be smart and skip writing if the inode is
* marked clean. We don't use filemap_fwrite for flushing
* because we want to control how many pages we write out at a
* time, thus this is the only safe way to make sure we've
* waited for outstanding compressed workers to have started
* their jobs and thus have ordered extents set up properly.
*
* This exists because we do not want to wait for each
* individual inode to finish its async work, we simply want to
* start the IO on everybody, and then come back here and wait
* for all of the async work to catch up. Once we're done with
* that we know we'll have ordered extents for everything and we
* can decide if we wait for that or not.
*
* If we choose to replace this in the future, make absolutely
* sure that the proper waiting is being done in the async case,
* as there have been bugs in that area before.
*/
async_pages = atomic_read(&fs_info->async_delalloc_pages);
if (!async_pages)
goto skip_async;
/*
* We don't want to wait forever, if we wrote less pages in this
* loop than we have outstanding, only wait for that number of
* pages, otherwise we can wait for all async pages to finish
* before continuing.
*/
if (async_pages > nr_pages)
async_pages -= nr_pages;
else
async_pages = 0;
wait_event(fs_info->async_submit_wait,
atomic_read(&fs_info->async_delalloc_pages) <=
async_pages);
skip_async:
loops++;
if (wait_ordered && !trans) {
btrfs_wait_ordered_roots(fs_info, items, 0, (u64)-1);
} else {
time_left = schedule_timeout_killable(1);
if (time_left)
break;
}
/*
* If we are for preemption we just want a one-shot of delalloc
* flushing so we can stop flushing if we decide we don't need
* to anymore.
*/
if (for_preempt)
break;
spin_lock(&space_info->lock);
if (list_empty(&space_info->tickets) &&
list_empty(&space_info->priority_tickets)) {
spin_unlock(&space_info->lock);
break;
}
spin_unlock(&space_info->lock);
delalloc_bytes = percpu_counter_sum_positive(
&fs_info->delalloc_bytes);
ordered_bytes = percpu_counter_sum_positive(
&fs_info->ordered_bytes);
}
}
/*
* Try to flush some data based on policy set by @state. This is only advisory
* and may fail for various reasons. The caller is supposed to examine the
* state of @space_info to detect the outcome.
*/
static void flush_space(struct btrfs_fs_info *fs_info,
struct btrfs_space_info *space_info, u64 num_bytes,
enum btrfs_flush_state state, bool for_preempt)
{
struct btrfs_root *root = fs_info->tree_root;
struct btrfs_trans_handle *trans;
int nr;
int ret = 0;
switch (state) {
case FLUSH_DELAYED_ITEMS_NR:
case FLUSH_DELAYED_ITEMS:
if (state == FLUSH_DELAYED_ITEMS_NR)
nr = calc_reclaim_items_nr(fs_info, num_bytes) * 2;
else
nr = -1;
trans = btrfs_join_transaction(root);
if (IS_ERR(trans)) {
ret = PTR_ERR(trans);
break;
}
ret = btrfs_run_delayed_items_nr(trans, nr);
btrfs_end_transaction(trans);
break;
case FLUSH_DELALLOC:
case FLUSH_DELALLOC_WAIT:
case FLUSH_DELALLOC_FULL:
if (state == FLUSH_DELALLOC_FULL)
num_bytes = U64_MAX;
shrink_delalloc(fs_info, space_info, num_bytes,
state != FLUSH_DELALLOC, for_preempt);
break;
case FLUSH_DELAYED_REFS_NR:
case FLUSH_DELAYED_REFS:
trans = btrfs_join_transaction(root);
if (IS_ERR(trans)) {
ret = PTR_ERR(trans);
break;
}
if (state == FLUSH_DELAYED_REFS_NR)
nr = calc_delayed_refs_nr(fs_info, num_bytes);
else
nr = 0;
btrfs_run_delayed_refs(trans, nr);
btrfs_end_transaction(trans);
break;
case ALLOC_CHUNK:
case ALLOC_CHUNK_FORCE:
/*
* For metadata space on zoned filesystem, reaching here means we
* don't have enough space left in active_total_bytes. Try to
* activate a block group first, because we may have inactive
* block group already allocated.
*/
ret = btrfs_zoned_activate_one_bg(fs_info, space_info, false);
if (ret < 0)
break;
else if (ret == 1)
break;
trans = btrfs_join_transaction(root);
if (IS_ERR(trans)) {
ret = PTR_ERR(trans);
break;
}
ret = btrfs_chunk_alloc(trans,
btrfs_get_alloc_profile(fs_info, space_info->flags),
(state == ALLOC_CHUNK) ? CHUNK_ALLOC_NO_FORCE :
CHUNK_ALLOC_FORCE);
btrfs_end_transaction(trans);
/*
* For metadata space on zoned filesystem, allocating a new chunk
* is not enough. We still need to activate the block * group.
* Active the newly allocated block group by (maybe) finishing
* a block group.
*/
if (ret == 1) {
ret = btrfs_zoned_activate_one_bg(fs_info, space_info, true);
/*
* Revert to the original ret regardless we could finish
* one block group or not.
*/
if (ret >= 0)
ret = 1;
}
if (ret > 0 || ret == -ENOSPC)
ret = 0;
break;
case RUN_DELAYED_IPUTS:
/*
* If we have pending delayed iputs then we could free up a
* bunch of pinned space, so make sure we run the iputs before
* we do our pinned bytes check below.
*/
btrfs_run_delayed_iputs(fs_info);
btrfs_wait_on_delayed_iputs(fs_info);
break;
case COMMIT_TRANS:
ASSERT(current->journal_info == NULL);
trans = btrfs_join_transaction(root);
if (IS_ERR(trans)) {
ret = PTR_ERR(trans);
break;
}
ret = btrfs_commit_transaction(trans);
break;
default:
ret = -ENOSPC;
break;
}
trace_btrfs_flush_space(fs_info, space_info->flags, num_bytes, state,
ret, for_preempt);
return;
}
static inline u64
btrfs_calc_reclaim_metadata_size(struct btrfs_fs_info *fs_info,
struct btrfs_space_info *space_info)
{
u64 used;
u64 avail;
u64 to_reclaim = space_info->reclaim_size;
lockdep_assert_held(&space_info->lock);
avail = calc_available_free_space(fs_info, space_info,
BTRFS_RESERVE_FLUSH_ALL);
used = btrfs_space_info_used(space_info, true);
/*
* We may be flushing because suddenly we have less space than we had
* before, and now we're well over-committed based on our current free
* space. If that's the case add in our overage so we make sure to put
* appropriate pressure on the flushing state machine.
*/
if (space_info->total_bytes + avail < used)
to_reclaim += used - (space_info->total_bytes + avail);
return to_reclaim;
}
static bool need_preemptive_reclaim(struct btrfs_fs_info *fs_info,
struct btrfs_space_info *space_info)
{
u64 global_rsv_size = fs_info->global_block_rsv.reserved;
u64 ordered, delalloc;
u64 thresh;
u64 used;
thresh = mult_perc(space_info->total_bytes, 90);
lockdep_assert_held(&space_info->lock);
/* If we're just plain full then async reclaim just slows us down. */
if ((space_info->bytes_used + space_info->bytes_reserved +
global_rsv_size) >= thresh)
return false;
used = space_info->bytes_may_use + space_info->bytes_pinned;
/* The total flushable belongs to the global rsv, don't flush. */
if (global_rsv_size >= used)
return false;
/*
* 128MiB is 1/4 of the maximum global rsv size. If we have less than
* that devoted to other reservations then there's no sense in flushing,
* we don't have a lot of things that need flushing.
*/
if (used - global_rsv_size <= SZ_128M)
return false;
/*
* We have tickets queued, bail so we don't compete with the async
* flushers.
*/
if (space_info->reclaim_size)
return false;
/*
* If we have over half of the free space occupied by reservations or
* pinned then we want to start flushing.
*
* We do not do the traditional thing here, which is to say
*
* if (used >= ((total_bytes + avail) / 2))
* return 1;
*
* because this doesn't quite work how we want. If we had more than 50%
* of the space_info used by bytes_used and we had 0 available we'd just
* constantly run the background flusher. Instead we want it to kick in
* if our reclaimable space exceeds our clamped free space.
*
* Our clamping range is 2^1 -> 2^8. Practically speaking that means
* the following:
*
* Amount of RAM Minimum threshold Maximum threshold
*
* 256GiB 1GiB 128GiB
* 128GiB 512MiB 64GiB
* 64GiB 256MiB 32GiB
* 32GiB 128MiB 16GiB
* 16GiB 64MiB 8GiB
*
* These are the range our thresholds will fall in, corresponding to how
* much delalloc we need for the background flusher to kick in.
*/
thresh = calc_available_free_space(fs_info, space_info,
BTRFS_RESERVE_FLUSH_ALL);
used = space_info->bytes_used + space_info->bytes_reserved +
space_info->bytes_readonly + global_rsv_size;
if (used < space_info->total_bytes)
thresh += space_info->total_bytes - used;
thresh >>= space_info->clamp;
used = space_info->bytes_pinned;
/*
* If we have more ordered bytes than delalloc bytes then we're either
* doing a lot of DIO, or we simply don't have a lot of delalloc waiting
* around. Preemptive flushing is only useful in that it can free up
* space before tickets need to wait for things to finish. In the case
* of ordered extents, preemptively waiting on ordered extents gets us
* nothing, if our reservations are tied up in ordered extents we'll
* simply have to slow down writers by forcing them to wait on ordered
* extents.
*
* In the case that ordered is larger than delalloc, only include the
* block reserves that we would actually be able to directly reclaim
* from. In this case if we're heavy on metadata operations this will
* clearly be heavy enough to warrant preemptive flushing. In the case
* of heavy DIO or ordered reservations, preemptive flushing will just
* waste time and cause us to slow down.
*
* We want to make sure we truly are maxed out on ordered however, so
* cut ordered in half, and if it's still higher than delalloc then we
* can keep flushing. This is to avoid the case where we start
* flushing, and now delalloc == ordered and we stop preemptively
* flushing when we could still have several gigs of delalloc to flush.
*/
ordered = percpu_counter_read_positive(&fs_info->ordered_bytes) >> 1;
delalloc = percpu_counter_read_positive(&fs_info->delalloc_bytes);
if (ordered >= delalloc)
used += fs_info->delayed_refs_rsv.reserved +
fs_info->delayed_block_rsv.reserved;
else
used += space_info->bytes_may_use - global_rsv_size;
return (used >= thresh && !btrfs_fs_closing(fs_info) &&
!test_bit(BTRFS_FS_STATE_REMOUNTING, &fs_info->fs_state));
}
static bool steal_from_global_rsv(struct btrfs_fs_info *fs_info,
struct btrfs_space_info *space_info,
struct reserve_ticket *ticket)
{
struct btrfs_block_rsv *global_rsv = &fs_info->global_block_rsv;
u64 min_bytes;
if (!ticket->steal)
return false;
if (global_rsv->space_info != space_info)
return false;
spin_lock(&global_rsv->lock);
min_bytes = mult_perc(global_rsv->size, 10);
if (global_rsv->reserved < min_bytes + ticket->bytes) {
spin_unlock(&global_rsv->lock);
return false;
}
global_rsv->reserved -= ticket->bytes;
remove_ticket(space_info, ticket);
ticket->bytes = 0;
wake_up(&ticket->wait);
space_info->tickets_id++;
if (global_rsv->reserved < global_rsv->size)
global_rsv->full = 0;
spin_unlock(&global_rsv->lock);
return true;
}
/*
* maybe_fail_all_tickets - we've exhausted our flushing, start failing tickets
* @fs_info - fs_info for this fs
* @space_info - the space info we were flushing
*
* We call this when we've exhausted our flushing ability and haven't made
* progress in satisfying tickets. The reservation code handles tickets in
* order, so if there is a large ticket first and then smaller ones we could
* very well satisfy the smaller tickets. This will attempt to wake up any
* tickets in the list to catch this case.
*
* This function returns true if it was able to make progress by clearing out