forked from systemd/systemd
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathbtrfs-util.c
2168 lines (1649 loc) · 74 KB
/
btrfs-util.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: LGPL-2.1-or-later */
#include <errno.h>
#include <fcntl.h>
#include <inttypes.h>
#include <linux/btrfs_tree.h>
#include <linux/fs.h>
#include <linux/loop.h>
#include <linux/magic.h>
#include <stddef.h>
#include <stdio.h>
#include <stdlib.h>
#include <sys/file.h>
#include <sys/ioctl.h>
#include <sys/sysmacros.h>
#include <unistd.h>
#include "alloc-util.h"
#include "blockdev-util.h"
#include "btrfs-util.h"
#include "chase.h"
#include "chattr-util.h"
#include "copy.h"
#include "fd-util.h"
#include "fileio.h"
#include "fs-util.h"
#include "io-util.h"
#include "macro.h"
#include "path-util.h"
#include "rm-rf.h"
#include "smack-util.h"
#include "sparse-endian.h"
#include "stat-util.h"
#include "string-util.h"
#include "time-util.h"
/* WARNING: Be careful with file system ioctls! When we get an fd, we
* need to make sure it either refers to only a regular file or
* directory, or that it is located on btrfs, before invoking any
* btrfs ioctls. The ioctl numbers are reused by some device drivers
* (such as DRM), and hence might have bad effects when invoked on
* device nodes (that reference drivers) rather than fds to normal
* files or directories. */
int btrfs_is_subvol_at(int dir_fd, const char *path) {
struct stat st;
assert(dir_fd >= 0 || dir_fd == AT_FDCWD);
/* On btrfs subvolumes always have the inode 256 */
if (fstatat(dir_fd, strempty(path), &st, isempty(path) ? AT_EMPTY_PATH : 0) < 0)
return -errno;
if (!btrfs_might_be_subvol(&st))
return 0;
return is_fs_type_at(dir_fd, path, BTRFS_SUPER_MAGIC);
}
int btrfs_subvol_set_read_only_at(int dir_fd, const char *path, bool b) {
_cleanup_close_ int fd = -EBADF;
uint64_t flags, nflags;
struct stat st;
assert(dir_fd >= 0 || dir_fd == AT_FDCWD);
fd = xopenat(dir_fd, path, O_RDONLY|O_NOCTTY|O_CLOEXEC|O_DIRECTORY, /* xopen_flags = */ 0, /* mode = */ 0);
if (fd < 0)
return fd;
if (fstat(fd, &st) < 0)
return -errno;
if (!btrfs_might_be_subvol(&st))
return -EINVAL;
if (ioctl(fd, BTRFS_IOC_SUBVOL_GETFLAGS, &flags) < 0)
return -errno;
nflags = UPDATE_FLAG(flags, BTRFS_SUBVOL_RDONLY, b);
if (flags == nflags)
return 0;
return RET_NERRNO(ioctl(fd, BTRFS_IOC_SUBVOL_SETFLAGS, &nflags));
}
int btrfs_subvol_get_read_only_fd(int fd) {
uint64_t flags;
struct stat st;
assert(fd >= 0);
if (fstat(fd, &st) < 0)
return -errno;
if (!btrfs_might_be_subvol(&st))
return -EINVAL;
if (ioctl(fd, BTRFS_IOC_SUBVOL_GETFLAGS, &flags) < 0)
return -errno;
return !!(flags & BTRFS_SUBVOL_RDONLY);
}
int btrfs_get_block_device_at(int dir_fd, const char *path, dev_t *ret) {
struct btrfs_ioctl_fs_info_args fsi = {};
_cleanup_close_ int fd = -EBADF;
uint64_t id;
int r;
assert(dir_fd >= 0 || dir_fd == AT_FDCWD);
assert(path);
assert(ret);
fd = xopenat(dir_fd, path, O_RDONLY|O_CLOEXEC|O_NONBLOCK|O_NOCTTY, /* xopen_flags = */ 0, /* mode = */ 0);
if (fd < 0)
return fd;
r = fd_is_fs_type(fd, BTRFS_SUPER_MAGIC);
if (r < 0)
return r;
if (r == 0)
return -ENOTTY;
if (ioctl(fd, BTRFS_IOC_FS_INFO, &fsi) < 0)
return -errno;
/* We won't do this for btrfs RAID */
if (fsi.num_devices != 1) {
*ret = 0;
return 0;
}
for (id = 1; id <= fsi.max_id; id++) {
struct btrfs_ioctl_dev_info_args di = {
.devid = id,
};
struct stat st;
if (ioctl(fd, BTRFS_IOC_DEV_INFO, &di) < 0) {
if (errno == ENODEV)
continue;
return -errno;
}
/* For the root fs — when no initrd is involved — btrfs returns /dev/root on any kernels from
* the past few years. That sucks, as we have no API to determine the actual root then. let's
* return an recognizable error for this case, so that the caller can maybe print a nice
* message about this.
*
* https://bugzilla.kernel.org/show_bug.cgi?id=89721 */
if (path_equal((char*) di.path, "/dev/root"))
return -EUCLEAN;
if (stat((char*) di.path, &st) < 0)
return -errno;
if (!S_ISBLK(st.st_mode))
return -ENOTBLK;
if (major(st.st_rdev) == 0)
return -ENODEV;
*ret = st.st_rdev;
return 1;
}
return -ENODEV;
}
int btrfs_subvol_get_id_fd(int fd, uint64_t *ret) {
struct btrfs_ioctl_ino_lookup_args args = {
.objectid = BTRFS_FIRST_FREE_OBJECTID
};
int r;
assert(fd >= 0);
assert(ret);
r = fd_is_fs_type(fd, BTRFS_SUPER_MAGIC);
if (r < 0)
return r;
if (r == 0)
return -ENOTTY;
if (ioctl(fd, BTRFS_IOC_INO_LOOKUP, &args) < 0)
return -errno;
*ret = args.treeid;
return 0;
}
int btrfs_subvol_get_id(int fd, const char *subvol, uint64_t *ret) {
_cleanup_close_ int subvol_fd = -EBADF;
assert(fd >= 0);
assert(ret);
subvol_fd = openat(fd, subvol, O_RDONLY|O_CLOEXEC|O_NOCTTY|O_NOFOLLOW);
if (subvol_fd < 0)
return -errno;
return btrfs_subvol_get_id_fd(subvol_fd, ret);
}
static bool btrfs_ioctl_search_args_inc(struct btrfs_ioctl_search_args *args) {
assert(args);
/* the objectid, type, offset together make up the btrfs key,
* which is considered a single 136byte integer when
* comparing. This call increases the counter by one, dealing
* with the overflow between the overflows */
if (args->key.min_offset < UINT64_MAX) {
args->key.min_offset++;
return true;
}
if (args->key.min_type < UINT8_MAX) {
args->key.min_type++;
args->key.min_offset = 0;
return true;
}
if (args->key.min_objectid < UINT64_MAX) {
args->key.min_objectid++;
args->key.min_offset = 0;
args->key.min_type = 0;
return true;
}
return 0;
}
static void btrfs_ioctl_search_args_set(struct btrfs_ioctl_search_args *args, const struct btrfs_ioctl_search_header *h) {
assert(args);
assert(h);
args->key.min_objectid = h->objectid;
args->key.min_type = h->type;
args->key.min_offset = h->offset;
}
static int btrfs_ioctl_search_args_compare(const struct btrfs_ioctl_search_args *args) {
int r;
assert(args);
/* Compare min and max */
r = CMP(args->key.min_objectid, args->key.max_objectid);
if (r != 0)
return r;
r = CMP(args->key.min_type, args->key.max_type);
if (r != 0)
return r;
return CMP(args->key.min_offset, args->key.max_offset);
}
#define FOREACH_BTRFS_IOCTL_SEARCH_HEADER(i, sh, args) \
for ((i) = 0, \
(sh) = (const struct btrfs_ioctl_search_header*) (args).buf; \
(i) < (args).key.nr_items; \
(i)++, \
(sh) = (const struct btrfs_ioctl_search_header*) ((uint8_t*) (sh) + sizeof(struct btrfs_ioctl_search_header) + (sh)->len))
#define BTRFS_IOCTL_SEARCH_HEADER_BODY(sh) \
((void*) ((uint8_t*) sh + sizeof(struct btrfs_ioctl_search_header)))
int btrfs_subvol_get_info_fd(int fd, uint64_t subvol_id, BtrfsSubvolInfo *ret) {
struct btrfs_ioctl_search_args args = {
/* Tree of tree roots */
.key.tree_id = BTRFS_ROOT_TREE_OBJECTID,
/* Look precisely for the subvolume items */
.key.min_type = BTRFS_ROOT_ITEM_KEY,
.key.max_type = BTRFS_ROOT_ITEM_KEY,
.key.min_offset = 0,
.key.max_offset = UINT64_MAX,
/* No restrictions on the other components */
.key.min_transid = 0,
.key.max_transid = UINT64_MAX,
};
bool found = false;
int r;
assert(fd >= 0);
assert(ret);
if (subvol_id == 0) {
r = btrfs_subvol_get_id_fd(fd, &subvol_id);
if (r < 0)
return r;
} else {
r = fd_is_fs_type(fd, BTRFS_SUPER_MAGIC);
if (r < 0)
return r;
if (r == 0)
return -ENOTTY;
}
args.key.min_objectid = args.key.max_objectid = subvol_id;
while (btrfs_ioctl_search_args_compare(&args) <= 0) {
const struct btrfs_ioctl_search_header *sh;
unsigned i;
args.key.nr_items = 256;
if (ioctl(fd, BTRFS_IOC_TREE_SEARCH, &args) < 0)
return -errno;
if (args.key.nr_items <= 0)
break;
FOREACH_BTRFS_IOCTL_SEARCH_HEADER(i, sh, args) {
const struct btrfs_root_item *ri;
/* Make sure we start the next search at least from this entry */
btrfs_ioctl_search_args_set(&args, sh);
if (sh->objectid != subvol_id)
continue;
if (sh->type != BTRFS_ROOT_ITEM_KEY)
continue;
/* Older versions of the struct lacked the otime setting */
if (sh->len < offsetof(struct btrfs_root_item, otime) + sizeof(struct btrfs_timespec))
continue;
ri = BTRFS_IOCTL_SEARCH_HEADER_BODY(sh);
ret->otime = (usec_t) le64toh(ri->otime.sec) * USEC_PER_SEC +
(usec_t) le32toh(ri->otime.nsec) / NSEC_PER_USEC;
ret->subvol_id = subvol_id;
ret->read_only = le64toh(ri->flags) & BTRFS_ROOT_SUBVOL_RDONLY;
assert_cc(sizeof(ri->uuid) == sizeof(ret->uuid));
memcpy(&ret->uuid, ri->uuid, sizeof(ret->uuid));
memcpy(&ret->parent_uuid, ri->parent_uuid, sizeof(ret->parent_uuid));
found = true;
goto finish;
}
/* Increase search key by one, to read the next item, if we can. */
if (!btrfs_ioctl_search_args_inc(&args))
break;
}
finish:
return found ? 0 : -ENODATA;
}
int btrfs_qgroup_get_quota_fd(int fd, uint64_t qgroupid, BtrfsQuotaInfo *ret) {
struct btrfs_ioctl_search_args args = {
/* Tree of quota items */
.key.tree_id = BTRFS_QUOTA_TREE_OBJECTID,
/* The object ID is always 0 */
.key.min_objectid = 0,
.key.max_objectid = 0,
/* Look precisely for the quota items */
.key.min_type = BTRFS_QGROUP_STATUS_KEY,
.key.max_type = BTRFS_QGROUP_LIMIT_KEY,
/* No restrictions on the other components */
.key.min_transid = 0,
.key.max_transid = UINT64_MAX,
};
bool found_info = false, found_limit = false;
int r;
assert(fd >= 0);
assert(ret);
if (qgroupid == 0) {
r = btrfs_subvol_get_id_fd(fd, &qgroupid);
if (r < 0)
return r;
} else {
r = fd_is_fs_type(fd, BTRFS_SUPER_MAGIC);
if (r < 0)
return r;
if (r == 0)
return -ENOTTY;
}
args.key.min_offset = args.key.max_offset = qgroupid;
while (btrfs_ioctl_search_args_compare(&args) <= 0) {
const struct btrfs_ioctl_search_header *sh;
unsigned i;
args.key.nr_items = 256;
if (ioctl(fd, BTRFS_IOC_TREE_SEARCH, &args) < 0) {
if (errno == ENOENT) /* quota tree is missing: quota disabled */
break;
return -errno;
}
if (args.key.nr_items <= 0)
break;
FOREACH_BTRFS_IOCTL_SEARCH_HEADER(i, sh, args) {
/* Make sure we start the next search at least from this entry */
btrfs_ioctl_search_args_set(&args, sh);
if (sh->objectid != 0)
continue;
if (sh->offset != qgroupid)
continue;
if (sh->type == BTRFS_QGROUP_INFO_KEY) {
const struct btrfs_qgroup_info_item *qii = BTRFS_IOCTL_SEARCH_HEADER_BODY(sh);
ret->referenced = le64toh(qii->rfer);
ret->exclusive = le64toh(qii->excl);
found_info = true;
} else if (sh->type == BTRFS_QGROUP_LIMIT_KEY) {
const struct btrfs_qgroup_limit_item *qli = BTRFS_IOCTL_SEARCH_HEADER_BODY(sh);
if (le64toh(qli->flags) & BTRFS_QGROUP_LIMIT_MAX_RFER)
ret->referenced_max = le64toh(qli->max_rfer);
else
ret->referenced_max = UINT64_MAX;
if (le64toh(qli->flags) & BTRFS_QGROUP_LIMIT_MAX_EXCL)
ret->exclusive_max = le64toh(qli->max_excl);
else
ret->exclusive_max = UINT64_MAX;
found_limit = true;
}
if (found_info && found_limit)
goto finish;
}
/* Increase search key by one, to read the next item, if we can. */
if (!btrfs_ioctl_search_args_inc(&args))
break;
}
finish:
if (!found_limit && !found_info)
return -ENODATA;
if (!found_info) {
ret->referenced = UINT64_MAX;
ret->exclusive = UINT64_MAX;
}
if (!found_limit) {
ret->referenced_max = UINT64_MAX;
ret->exclusive_max = UINT64_MAX;
}
return 0;
}
int btrfs_qgroup_get_quota(const char *path, uint64_t qgroupid, BtrfsQuotaInfo *ret) {
_cleanup_close_ int fd = -EBADF;
fd = open(path, O_RDONLY|O_CLOEXEC|O_NOCTTY|O_NOFOLLOW);
if (fd < 0)
return -errno;
return btrfs_qgroup_get_quota_fd(fd, qgroupid, ret);
}
int btrfs_subvol_find_subtree_qgroup(int fd, uint64_t subvol_id, uint64_t *ret) {
uint64_t level, lowest = UINT64_MAX, lowest_qgroupid = 0;
_cleanup_free_ uint64_t *qgroups = NULL;
int r, n;
assert(fd >= 0);
assert(ret);
/* This finds the "subtree" qgroup for a specific
* subvolume. This only works for subvolumes that have been
* prepared with btrfs_subvol_auto_qgroup_fd() with
* insert_intermediary_qgroup=true (or equivalent). For others
* it will return the leaf qgroup instead. The two cases may
* be distinguished via the return value, which is 1 in case
* an appropriate "subtree" qgroup was found, and 0
* otherwise. */
if (subvol_id == 0) {
r = btrfs_subvol_get_id_fd(fd, &subvol_id);
if (r < 0)
return r;
}
r = btrfs_qgroupid_split(subvol_id, &level, NULL);
if (r < 0)
return r;
if (level != 0) /* Input must be a leaf qgroup */
return -EINVAL;
n = btrfs_qgroup_find_parents(fd, subvol_id, &qgroups);
if (n < 0)
return n;
for (int i = 0; i < n; i++) {
uint64_t id;
r = btrfs_qgroupid_split(qgroups[i], &level, &id);
if (r < 0)
return r;
if (id != subvol_id)
continue;
if (lowest == UINT64_MAX || level < lowest) {
lowest_qgroupid = qgroups[i];
lowest = level;
}
}
if (lowest == UINT64_MAX) {
/* No suitable higher-level qgroup found, let's return
* the leaf qgroup instead, and indicate that with the
* return value. */
*ret = subvol_id;
return 0;
}
*ret = lowest_qgroupid;
return 1;
}
int btrfs_subvol_get_subtree_quota_fd(int fd, uint64_t subvol_id, BtrfsQuotaInfo *ret) {
uint64_t qgroupid;
int r;
assert(fd >= 0);
assert(ret);
/* This determines the quota data of the qgroup with the
* lowest level, that shares the id part with the specified
* subvolume. This is useful for determining the quota data
* for entire subvolume subtrees, as long as the subtrees have
* been set up with btrfs_qgroup_subvol_auto_fd() or in a
* compatible way */
r = btrfs_subvol_find_subtree_qgroup(fd, subvol_id, &qgroupid);
if (r < 0)
return r;
return btrfs_qgroup_get_quota_fd(fd, qgroupid, ret);
}
int btrfs_subvol_get_subtree_quota(const char *path, uint64_t subvol_id, BtrfsQuotaInfo *ret) {
_cleanup_close_ int fd = -EBADF;
fd = open(path, O_RDONLY|O_CLOEXEC|O_NOCTTY|O_NOFOLLOW);
if (fd < 0)
return -errno;
return btrfs_subvol_get_subtree_quota_fd(fd, subvol_id, ret);
}
int btrfs_defrag_fd(int fd) {
int r;
assert(fd >= 0);
r = fd_verify_regular(fd);
if (r < 0)
return r;
return RET_NERRNO(ioctl(fd, BTRFS_IOC_DEFRAG, NULL));
}
int btrfs_defrag(const char *p) {
_cleanup_close_ int fd = -EBADF;
fd = open(p, O_RDWR|O_CLOEXEC|O_NOCTTY|O_NOFOLLOW);
if (fd < 0)
return -errno;
return btrfs_defrag_fd(fd);
}
int btrfs_quota_enable_fd(int fd, bool b) {
struct btrfs_ioctl_quota_ctl_args args = {
.cmd = b ? BTRFS_QUOTA_CTL_ENABLE : BTRFS_QUOTA_CTL_DISABLE,
};
int r;
assert(fd >= 0);
r = fd_is_fs_type(fd, BTRFS_SUPER_MAGIC);
if (r < 0)
return r;
if (r == 0)
return -ENOTTY;
return RET_NERRNO(ioctl(fd, BTRFS_IOC_QUOTA_CTL, &args));
}
int btrfs_quota_enable(const char *path, bool b) {
_cleanup_close_ int fd = -EBADF;
fd = open(path, O_RDONLY|O_CLOEXEC|O_NOCTTY|O_NOFOLLOW);
if (fd < 0)
return -errno;
return btrfs_quota_enable_fd(fd, b);
}
int btrfs_qgroup_set_limit_fd(int fd, uint64_t qgroupid, uint64_t referenced_max) {
struct btrfs_ioctl_qgroup_limit_args args = {
.lim.max_rfer = referenced_max,
.lim.flags = BTRFS_QGROUP_LIMIT_MAX_RFER,
};
int r;
assert(fd >= 0);
if (qgroupid == 0) {
r = btrfs_subvol_get_id_fd(fd, &qgroupid);
if (r < 0)
return r;
} else {
r = fd_is_fs_type(fd, BTRFS_SUPER_MAGIC);
if (r < 0)
return r;
if (r == 0)
return -ENOTTY;
}
args.qgroupid = qgroupid;
for (unsigned c = 0;; c++) {
if (ioctl(fd, BTRFS_IOC_QGROUP_LIMIT, &args) < 0) {
if (errno == EBUSY && c < 10) {
(void) btrfs_quota_scan_wait(fd);
continue;
}
return -errno;
}
break;
}
return 0;
}
int btrfs_qgroup_set_limit(const char *path, uint64_t qgroupid, uint64_t referenced_max) {
_cleanup_close_ int fd = -EBADF;
fd = open(path, O_RDONLY|O_CLOEXEC|O_NOCTTY|O_NOFOLLOW);
if (fd < 0)
return -errno;
return btrfs_qgroup_set_limit_fd(fd, qgroupid, referenced_max);
}
int btrfs_subvol_set_subtree_quota_limit_fd(int fd, uint64_t subvol_id, uint64_t referenced_max) {
uint64_t qgroupid;
int r;
assert(fd >= 0);
r = btrfs_subvol_find_subtree_qgroup(fd, subvol_id, &qgroupid);
if (r < 0)
return r;
return btrfs_qgroup_set_limit_fd(fd, qgroupid, referenced_max);
}
int btrfs_subvol_set_subtree_quota_limit(const char *path, uint64_t subvol_id, uint64_t referenced_max) {
_cleanup_close_ int fd = -EBADF;
fd = open(path, O_RDONLY|O_CLOEXEC|O_NOCTTY|O_NOFOLLOW);
if (fd < 0)
return -errno;
return btrfs_subvol_set_subtree_quota_limit_fd(fd, subvol_id, referenced_max);
}
int btrfs_qgroupid_make(uint64_t level, uint64_t id, uint64_t *ret) {
assert(ret);
if (level >= (UINT64_C(1) << (64 - BTRFS_QGROUP_LEVEL_SHIFT)))
return -EINVAL;
if (id >= (UINT64_C(1) << BTRFS_QGROUP_LEVEL_SHIFT))
return -EINVAL;
*ret = (level << BTRFS_QGROUP_LEVEL_SHIFT) | id;
return 0;
}
int btrfs_qgroupid_split(uint64_t qgroupid, uint64_t *level, uint64_t *id) {
assert(level || id);
if (level)
*level = qgroupid >> BTRFS_QGROUP_LEVEL_SHIFT;
if (id)
*id = qgroupid & ((UINT64_C(1) << BTRFS_QGROUP_LEVEL_SHIFT) - 1);
return 0;
}
static int qgroup_create_or_destroy(int fd, bool b, uint64_t qgroupid) {
struct btrfs_ioctl_qgroup_create_args args = {
.create = b,
.qgroupid = qgroupid,
};
int r;
r = fd_is_fs_type(fd, BTRFS_SUPER_MAGIC);
if (r < 0)
return r;
if (r == 0)
return -ENOTTY;
for (unsigned c = 0;; c++) {
if (ioctl(fd, BTRFS_IOC_QGROUP_CREATE, &args) < 0) {
/* On old kernels if quota is not enabled, we get EINVAL. On newer kernels we get
* ENOTCONN. Let's always convert this to ENOTCONN to make this recognizable
* everywhere the same way. */
if (IN_SET(errno, EINVAL, ENOTCONN))
return -ENOTCONN;
if (errno == EBUSY && c < 10) {
(void) btrfs_quota_scan_wait(fd);
continue;
}
return -errno;
}
break;
}
return 0;
}
int btrfs_qgroup_create(int fd, uint64_t qgroupid) {
return qgroup_create_or_destroy(fd, true, qgroupid);
}
int btrfs_qgroup_destroy(int fd, uint64_t qgroupid) {
return qgroup_create_or_destroy(fd, false, qgroupid);
}
int btrfs_qgroup_destroy_recursive(int fd, uint64_t qgroupid) {
_cleanup_free_ uint64_t *qgroups = NULL;
uint64_t subvol_id;
int n, r;
/* Destroys the specified qgroup, but unassigns it from all
* its parents first. Also, it recursively destroys all
* qgroups it is assigned to that have the same id part of the
* qgroupid as the specified group. */
r = btrfs_qgroupid_split(qgroupid, NULL, &subvol_id);
if (r < 0)
return r;
n = btrfs_qgroup_find_parents(fd, qgroupid, &qgroups);
if (n < 0)
return n;
for (int i = 0; i < n; i++) {
uint64_t id;
r = btrfs_qgroupid_split(qgroups[i], NULL, &id);
if (r < 0)
return r;
r = btrfs_qgroup_unassign(fd, qgroupid, qgroups[i]);
if (r < 0)
return r;
if (id != subvol_id)
continue;
/* The parent qgroupid shares the same id part with
* us? If so, destroy it too. */
(void) btrfs_qgroup_destroy_recursive(fd, qgroups[i]);
}
return btrfs_qgroup_destroy(fd, qgroupid);
}
int btrfs_quota_scan_start(int fd) {
struct btrfs_ioctl_quota_rescan_args args = {};
assert(fd >= 0);
return RET_NERRNO(ioctl(fd, BTRFS_IOC_QUOTA_RESCAN, &args));
}
int btrfs_quota_scan_wait(int fd) {
assert(fd >= 0);
return RET_NERRNO(ioctl(fd, BTRFS_IOC_QUOTA_RESCAN_WAIT));
}
int btrfs_quota_scan_ongoing(int fd) {
struct btrfs_ioctl_quota_rescan_args args = {};
assert(fd >= 0);
if (ioctl(fd, BTRFS_IOC_QUOTA_RESCAN_STATUS, &args) < 0)
return -errno;
return !!args.flags;
}
static int qgroup_assign_or_unassign(int fd, bool b, uint64_t child, uint64_t parent) {
struct btrfs_ioctl_qgroup_assign_args args = {
.assign = b,
.src = child,
.dst = parent,
};
int r;
r = fd_is_fs_type(fd, BTRFS_SUPER_MAGIC);
if (r < 0)
return r;
if (r == 0)
return -ENOTTY;
for (unsigned c = 0;; c++) {
r = ioctl(fd, BTRFS_IOC_QGROUP_ASSIGN, &args);
if (r < 0) {
if (errno == EBUSY && c < 10) {
(void) btrfs_quota_scan_wait(fd);
continue;
}
return -errno;
}
if (r == 0)
return 0;
/* If the return value is > 0, we need to request a rescan */
(void) btrfs_quota_scan_start(fd);
return 1;
}
}
int btrfs_qgroup_assign(int fd, uint64_t child, uint64_t parent) {
return qgroup_assign_or_unassign(fd, true, child, parent);
}
int btrfs_qgroup_unassign(int fd, uint64_t child, uint64_t parent) {
return qgroup_assign_or_unassign(fd, false, child, parent);
}
static int subvol_remove_children(int fd, const char *subvolume, uint64_t subvol_id, BtrfsRemoveFlags flags) {
struct btrfs_ioctl_search_args args = {
.key.tree_id = BTRFS_ROOT_TREE_OBJECTID,
.key.min_objectid = BTRFS_FIRST_FREE_OBJECTID,
.key.max_objectid = BTRFS_LAST_FREE_OBJECTID,
.key.min_type = BTRFS_ROOT_BACKREF_KEY,
.key.max_type = BTRFS_ROOT_BACKREF_KEY,
.key.min_transid = 0,
.key.max_transid = UINT64_MAX,
};
struct btrfs_ioctl_vol_args vol_args = {};
_cleanup_close_ int subvol_fd = -EBADF;
struct stat st;
bool made_writable = false;
int r;
assert(fd >= 0);
assert(subvolume);
if (fstat(fd, &st) < 0)
return -errno;
if (!S_ISDIR(st.st_mode))
return -EINVAL;
subvol_fd = openat(fd, subvolume, O_RDONLY|O_NOCTTY|O_CLOEXEC|O_DIRECTORY|O_NOFOLLOW);
if (subvol_fd < 0)
return -errno;
/* Let's check if this is actually a subvolume. Note that this is mostly redundant, as BTRFS_IOC_SNAP_DESTROY
* would fail anyway if it is not. However, it's a good thing to check this ahead of time so that we can return
* ENOTTY unconditionally in this case. This is different from the ioctl() which will return EPERM/EACCES if we
* don't have the privileges to remove subvolumes, regardless if the specified directory is actually a
* subvolume or not. In order to make it easy for callers to cover the "this is not a btrfs subvolume" case
* let's prefer ENOTTY over EPERM/EACCES though. */
r = btrfs_is_subvol_fd(subvol_fd);
if (r < 0)
return r;
if (r == 0) /* Not a btrfs subvolume */
return -ENOTTY;
if (subvol_id == 0) {
r = btrfs_subvol_get_id_fd(subvol_fd, &subvol_id);
if (r < 0)
return r;
}
/* First, try to remove the subvolume. If it happens to be
* already empty, this will just work. */
strncpy(vol_args.name, subvolume, sizeof(vol_args.name)-1);
if (ioctl(fd, BTRFS_IOC_SNAP_DESTROY, &vol_args) >= 0) {
(void) btrfs_qgroup_destroy_recursive(fd, subvol_id); /* for the leaf subvolumes, the qgroup id is identical to the subvol id */
return 0;
}
if (!(flags & BTRFS_REMOVE_RECURSIVE) || errno != ENOTEMPTY)
return -errno;
/* OK, the subvolume is not empty, let's look for child
* subvolumes, and remove them, first */
args.key.min_offset = args.key.max_offset = subvol_id;
while (btrfs_ioctl_search_args_compare(&args) <= 0) {
const struct btrfs_ioctl_search_header *sh;
unsigned i;
args.key.nr_items = 256;
if (ioctl(fd, BTRFS_IOC_TREE_SEARCH, &args) < 0)
return -errno;
if (args.key.nr_items <= 0)
break;
FOREACH_BTRFS_IOCTL_SEARCH_HEADER(i, sh, args) {
_cleanup_free_ char *p = NULL;
const struct btrfs_root_ref *ref;
btrfs_ioctl_search_args_set(&args, sh);
if (sh->type != BTRFS_ROOT_BACKREF_KEY)
continue;
if (sh->offset != subvol_id)
continue;
ref = BTRFS_IOCTL_SEARCH_HEADER_BODY(sh);
p = strndup((char*) ref + sizeof(struct btrfs_root_ref), le64toh(ref->name_len));
if (!p)
return -ENOMEM;
struct btrfs_ioctl_ino_lookup_args ino_args = {
.treeid = subvol_id,
.objectid = htole64(ref->dirid),
};
if (ioctl(fd, BTRFS_IOC_INO_LOOKUP, &ino_args) < 0)
return -errno;
if (!made_writable) {
r = btrfs_subvol_set_read_only_fd(subvol_fd, false);
if (r < 0)
return r;
made_writable = true;
}
if (isempty(ino_args.name))
/* Subvolume is in the top-level
* directory of the subvolume. */
r = subvol_remove_children(subvol_fd, p, sh->objectid, flags);
else {
_cleanup_close_ int child_fd = -EBADF;
/* Subvolume is somewhere further down,