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
|
/***
This file is part of systemd.
Copyright 2013 Lennart Poettering
systemd is free software; you can redistribute it and/or modify it
under the terms of the GNU Lesser General Public License as published by
the Free Software Foundation; either version 2.1 of the License, or
(at your option) any later version.
systemd is distributed in the hope that it will be useful, but
WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
Lesser General Public License for more details.
You should have received a copy of the GNU Lesser General Public License
along with systemd; If not, see <http://www.gnu.org/licenses/>.
***/
#include <dirent.h>
#include <errno.h>
#include <fcntl.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <sys/file.h>
#include <sys/stat.h>
#include <unistd.h>
#include <linux/fs.h>
#include "alloc-util.h"
#include "btrfs-util.h"
#include "chattr-util.h"
#include "copy.h"
#include "dirent-util.h"
#include "env-util.h"
#include "fd-util.h"
#include "fs-util.h"
#include "hashmap.h"
#include "lockfile-util.h"
#include "log.h"
#include "machine-image.h"
#include "macro.h"
#include "mkdir.h"
#include "path-util.h"
#include "rm-rf.h"
#include "string-table.h"
#include "string-util.h"
#include "strv.h"
#include "time-util.h"
#include "utf8.h"
#include "util.h"
#include "xattr-util.h"
static const char image_search_path[] =
"/var/lib/machines\0"
"/var/lib/container\0" /* legacy */
"/usr/local/lib/machines\0"
"/usr/lib/machines\0";
Image *image_unref(Image *i) {
if (!i)
return NULL;
free(i->name);
free(i->path);
return mfree(i);
}
static char **image_settings_path(Image *image) {
_cleanup_strv_free_ char **l = NULL;
char **ret;
const char *fn, *s;
unsigned i = 0;
assert(image);
l = new0(char*, 4);
if (!l)
return NULL;
fn = strjoina(image->name, ".nspawn");
FOREACH_STRING(s, "/etc/systemd/nspawn/", "/run/systemd/nspawn/") {
l[i] = strappend(s, fn);
if (!l[i])
return NULL;
i++;
}
l[i] = file_in_same_dir(image->path, fn);
if (!l[i])
return NULL;
ret = l;
l = NULL;
return ret;
}
static int image_new(
ImageType t,
const char *pretty,
const char *path,
const char *filename,
bool read_only,
usec_t crtime,
usec_t mtime,
Image **ret) {
_cleanup_(image_unrefp) Image *i = NULL;
assert(t >= 0);
assert(t < _IMAGE_TYPE_MAX);
assert(pretty);
assert(filename);
assert(ret);
i = new0(Image, 1);
if (!i)
return -ENOMEM;
i->type = t;
i->read_only = read_only;
i->crtime = crtime;
i->mtime = mtime;
i->usage = i->usage_exclusive = (uint64_t) -1;
i->limit = i->limit_exclusive = (uint64_t) -1;
i->name = strdup(pretty);
if (!i->name)
return -ENOMEM;
if (path)
i->path = strjoin(path, "/", filename);
else
i->path = strdup(filename);
if (!i->path)
return -ENOMEM;
path_kill_slashes(i->path);
*ret = i;
i = NULL;
return 0;
}
static int image_make(
const char *pretty,
int dfd,
const char *path,
const char *filename,
Image **ret) {
struct stat st;
bool read_only;
int r;
assert(filename);
/* We explicitly *do* follow symlinks here, since we want to
* allow symlinking trees into /var/lib/machines/, and treat
* them normally. */
if (fstatat(dfd, filename, &st, 0) < 0)
return -errno;
read_only =
(path && path_startswith(path, "/usr")) ||
(faccessat(dfd, filename, W_OK, AT_EACCESS) < 0 && errno == EROFS);
if (S_ISDIR(st.st_mode)) {
_cleanup_close_ int fd = -1;
unsigned file_attr = 0;
if (!ret)
return 1;
if (!pretty)
pretty = filename;
fd = openat(dfd, filename, O_CLOEXEC|O_NOCTTY|O_DIRECTORY);
if (fd < 0)
return -errno;
/* btrfs subvolumes have inode 256 */
if (st.st_ino == 256) {
r = btrfs_is_filesystem(fd);
if (r < 0)
return r;
if (r) {
BtrfsSubvolInfo info;
/* It's a btrfs subvolume */
r = btrfs_subvol_get_info_fd(fd, 0, &info);
if (r < 0)
return r;
r = image_new(IMAGE_SUBVOLUME,
pretty,
path,
filename,
info.read_only || read_only,
info.otime,
0,
ret);
if (r < 0)
return r;
if (btrfs_quota_scan_ongoing(fd) == 0) {
BtrfsQuotaInfo quota;
r = btrfs_subvol_get_subtree_quota_fd(fd, 0, "a);
if (r >= 0) {
(*ret)->usage = quota.referenced;
(*ret)->usage_exclusive = quota.exclusive;
(*ret)->limit = quota.referenced_max;
(*ret)->limit_exclusive = quota.exclusive_max;
}
}
return 1;
}
}
/* If the IMMUTABLE bit is set, we consider the
* directory read-only. Since the ioctl is not
* supported everywhere we ignore failures. */
(void) read_attr_fd(fd, &file_attr);
/* It's just a normal directory. */
r = image_new(IMAGE_DIRECTORY,
pretty,
path,
filename,
read_only || (file_attr & FS_IMMUTABLE_FL),
0,
0,
ret);
if (r < 0)
return r;
return 1;
} else if (S_ISREG(st.st_mode) && endswith(filename, ".raw")) {
usec_t crtime = 0;
/* It's a RAW disk image */
if (!ret)
return 1;
fd_getcrtime_at(dfd, filename, &crtime, 0);
if (!pretty)
pretty = strndupa(filename, strlen(filename) - 4);
r = image_new(IMAGE_RAW,
pretty,
path,
filename,
!(st.st_mode & 0222) || read_only,
crtime,
timespec_load(&st.st_mtim),
ret);
if (r < 0)
return r;
(*ret)->usage = (*ret)->usage_exclusive = st.st_blocks * 512;
(*ret)->limit = (*ret)->limit_exclusive = st.st_size;
return 1;
}
return 0;
}
int image_find(const char *name, Image **ret) {
const char *path;
int r;
assert(name);
/* There are no images with invalid names */
if (!image_name_is_valid(name))
return 0;
NULSTR_FOREACH(path, image_search_path) {
_cleanup_closedir_ DIR *d = NULL;
d = opendir(path);
if (!d) {
if (errno == ENOENT)
continue;
return -errno;
}
r = image_make(NULL, dirfd(d), path, name, ret);
if (r == 0 || r == -ENOENT) {
_cleanup_free_ char *raw = NULL;
raw = strappend(name, ".raw");
if (!raw)
return -ENOMEM;
r = image_make(NULL, dirfd(d), path, raw, ret);
if (r == 0 || r == -ENOENT)
continue;
}
if (r < 0)
return r;
return 1;
}
if (streq(name, ".host"))
return image_make(".host", AT_FDCWD, NULL, "/", ret);
return 0;
};
int image_discover(Hashmap *h) {
const char *path;
int r;
assert(h);
NULSTR_FOREACH(path, image_search_path) {
_cleanup_closedir_ DIR *d = NULL;
struct dirent *de;
d = opendir(path);
if (!d) {
if (errno == ENOENT)
continue;
return -errno;
}
FOREACH_DIRENT_ALL(de, d, return -errno) {
_cleanup_(image_unrefp) Image *image = NULL;
if (!image_name_is_valid(de->d_name))
continue;
if (hashmap_contains(h, de->d_name))
continue;
r = image_make(NULL, dirfd(d), path, de->d_name, &image);
if (r == 0 || r == -ENOENT)
continue;
if (r < 0)
return r;
r = hashmap_put(h, image->name, image);
if (r < 0)
return r;
image = NULL;
}
}
if (!hashmap_contains(h, ".host")) {
_cleanup_(image_unrefp) Image *image = NULL;
r = image_make(".host", AT_FDCWD, NULL, "/", &image);
if (r < 0)
return r;
r = hashmap_put(h, image->name, image);
if (r < 0)
return r;
image = NULL;
}
return 0;
}
void image_hashmap_free(Hashmap *map) {
Image *i;
while ((i = hashmap_steal_first(map)))
image_unref(i);
hashmap_free(map);
}
int image_remove(Image *i) {
_cleanup_release_lock_file_ LockFile global_lock = LOCK_FILE_INIT, local_lock = LOCK_FILE_INIT;
_cleanup_strv_free_ char **settings = NULL;
char **j;
int r;
assert(i);
if (IMAGE_IS_VENDOR(i) || IMAGE_IS_HOST(i))
return -EROFS;
settings = image_settings_path(i);
if (!settings)
return -ENOMEM;
/* Make sure we don't interfere with a running nspawn */
r = image_path_lock(i->path, LOCK_EX|LOCK_NB, &global_lock, &local_lock);
if (r < 0)
return r;
switch (i->type) {
case IMAGE_SUBVOLUME:
r = btrfs_subvol_remove(i->path, BTRFS_REMOVE_RECURSIVE|BTRFS_REMOVE_QUOTA);
if (r < 0)
return r;
break;
case IMAGE_DIRECTORY:
/* Allow deletion of read-only directories */
(void) chattr_path(i->path, 0, FS_IMMUTABLE_FL);
r = rm_rf(i->path, REMOVE_ROOT|REMOVE_PHYSICAL|REMOVE_SUBVOLUME);
if (r < 0)
return r;
break;
case IMAGE_RAW:
if (unlink(i->path) < 0)
return -errno;
break;
default:
return -EOPNOTSUPP;
}
STRV_FOREACH(j, settings) {
if (unlink(*j) < 0 && errno != ENOENT)
log_debug_errno(errno, "Failed to unlink %s, ignoring: %m", *j);
}
return 0;
}
static int rename_settings_file(const char *path, const char *new_name) {
_cleanup_free_ char *rs = NULL;
const char *fn;
fn = strjoina(new_name, ".nspawn");
rs = file_in_same_dir(path, fn);
if (!rs)
return -ENOMEM;
return rename_noreplace(AT_FDCWD, path, AT_FDCWD, rs);
}
int image_rename(Image *i, const char *new_name) {
_cleanup_release_lock_file_ LockFile global_lock = LOCK_FILE_INIT, local_lock = LOCK_FILE_INIT, name_lock = LOCK_FILE_INIT;
_cleanup_free_ char *new_path = NULL, *nn = NULL;
_cleanup_strv_free_ char **settings = NULL;
unsigned file_attr = 0;
char **j;
int r;
assert(i);
if (!image_name_is_valid(new_name))
return -EINVAL;
if (IMAGE_IS_VENDOR(i) || IMAGE_IS_HOST(i))
return -EROFS;
settings = image_settings_path(i);
if (!settings)
return -ENOMEM;
/* Make sure we don't interfere with a running nspawn */
r = image_path_lock(i->path, LOCK_EX|LOCK_NB, &global_lock, &local_lock);
if (r < 0)
return r;
/* Make sure nobody takes the new name, between the time we
* checked it is currently unused in all search paths, and the
* time we take possession of it */
r = image_name_lock(new_name, LOCK_EX|LOCK_NB, &name_lock);
if (r < 0)
return r;
r = image_find(new_name, NULL);
if (r < 0)
return r;
if (r > 0)
return -EEXIST;
switch (i->type) {
case IMAGE_DIRECTORY:
/* Turn of the immutable bit while we rename the image, so that we can rename it */
(void) read_attr_path(i->path, &file_attr);
if (file_attr & FS_IMMUTABLE_FL)
(void) chattr_path(i->path, 0, FS_IMMUTABLE_FL);
/* fall through */
case IMAGE_SUBVOLUME:
new_path = file_in_same_dir(i->path, new_name);
break;
case IMAGE_RAW: {
const char *fn;
fn = strjoina(new_name, ".raw");
new_path = file_in_same_dir(i->path, fn);
break;
}
default:
return -EOPNOTSUPP;
}
if (!new_path)
return -ENOMEM;
nn = strdup(new_name);
if (!nn)
return -ENOMEM;
r = rename_noreplace(AT_FDCWD, i->path, AT_FDCWD, new_path);
if (r < 0)
return r;
/* Restore the immutable bit, if it was set before */
if (file_attr & FS_IMMUTABLE_FL)
(void) chattr_path(new_path, FS_IMMUTABLE_FL, FS_IMMUTABLE_FL);
free(i->path);
i->path = new_path;
new_path = NULL;
free(i->name);
i->name = nn;
nn = NULL;
STRV_FOREACH(j, settings) {
r = rename_settings_file(*j, new_name);
if (r < 0 && r != -ENOENT)
log_debug_errno(r, "Failed to rename settings file %s, ignoring: %m", *j);
}
return 0;
}
static int clone_settings_file(const char *path, const char *new_name) {
_cleanup_free_ char *rs = NULL;
const char *fn;
fn = strjoina(new_name, ".nspawn");
rs = file_in_same_dir(path, fn);
if (!rs)
return -ENOMEM;
return copy_file_atomic(path, rs, 0664, false, 0);
}
int image_clone(Image *i, const char *new_name, bool read_only) {
_cleanup_release_lock_file_ LockFile name_lock = LOCK_FILE_INIT;
_cleanup_strv_free_ char **settings = NULL;
const char *new_path;
char **j;
int r;
assert(i);
if (!image_name_is_valid(new_name))
return -EINVAL;
settings = image_settings_path(i);
if (!settings)
return -ENOMEM;
/* Make sure nobody takes the new name, between the time we
* checked it is currently unused in all search paths, and the
* time we take possession of it */
r = image_name_lock(new_name, LOCK_EX|LOCK_NB, &name_lock);
if (r < 0)
return r;
r = image_find(new_name, NULL);
if (r < 0)
return r;
if (r > 0)
return -EEXIST;
switch (i->type) {
case IMAGE_SUBVOLUME:
case IMAGE_DIRECTORY:
/* If we can we'll always try to create a new btrfs subvolume here, even if the source is a plain
* directory.*/
new_path = strjoina("/var/lib/machines/", new_name);
r = btrfs_subvol_snapshot(i->path, new_path,
(read_only ? BTRFS_SNAPSHOT_READ_ONLY : 0) |
BTRFS_SNAPSHOT_FALLBACK_COPY |
BTRFS_SNAPSHOT_FALLBACK_DIRECTORY |
BTRFS_SNAPSHOT_FALLBACK_IMMUTABLE |
BTRFS_SNAPSHOT_RECURSIVE |
BTRFS_SNAPSHOT_QUOTA);
if (r >= 0)
/* Enable "subtree" quotas for the copy, if we didn't copy any quota from the source. */
(void) btrfs_subvol_auto_qgroup(new_path, 0, true);
break;
case IMAGE_RAW:
new_path = strjoina("/var/lib/machines/", new_name, ".raw");
r = copy_file_atomic(i->path, new_path, read_only ? 0444 : 0644, false, FS_NOCOW_FL);
break;
default:
return -EOPNOTSUPP;
}
if (r < 0)
return r;
STRV_FOREACH(j, settings) {
r = clone_settings_file(*j, new_name);
if (r < 0 && r != -ENOENT)
log_debug_errno(r, "Failed to clone settings %s, ignoring: %m", *j);
}
return 0;
}
int image_read_only(Image *i, bool b) {
_cleanup_release_lock_file_ LockFile global_lock = LOCK_FILE_INIT, local_lock = LOCK_FILE_INIT;
int r;
assert(i);
if (IMAGE_IS_VENDOR(i) || IMAGE_IS_HOST(i))
return -EROFS;
/* Make sure we don't interfere with a running nspawn */
r = image_path_lock(i->path, LOCK_EX|LOCK_NB, &global_lock, &local_lock);
if (r < 0)
return r;
switch (i->type) {
case IMAGE_SUBVOLUME:
/* Note that we set the flag only on the top-level
* subvolume of the image. */
r = btrfs_subvol_set_read_only(i->path, b);
if (r < 0)
return r;
break;
case IMAGE_DIRECTORY:
/* For simple directory trees we cannot use the access
mode of the top-level directory, since it has an
effect on the container itself. However, we can
use the "immutable" flag, to at least make the
top-level directory read-only. It's not as good as
a read-only subvolume, but at least something, and
we can read the value back.*/
r = chattr_path(i->path, b ? FS_IMMUTABLE_FL : 0, FS_IMMUTABLE_FL);
if (r < 0)
return r;
break;
case IMAGE_RAW: {
struct stat st;
if (stat(i->path, &st) < 0)
return -errno;
if (chmod(i->path, (st.st_mode & 0444) | (b ? 0000 : 0200)) < 0)
return -errno;
/* If the images is now read-only, it's a good time to
* defrag it, given that no write patterns will
* fragment it again. */
if (b)
(void) btrfs_defrag(i->path);
break;
}
default:
return -EOPNOTSUPP;
}
return 0;
}
int image_path_lock(const char *path, int operation, LockFile *global, LockFile *local) {
_cleanup_free_ char *p = NULL;
LockFile t = LOCK_FILE_INIT;
struct stat st;
int r;
assert(path);
assert(global);
assert(local);
/* Locks an image path. This actually creates two locks: one
* "local" one, next to the image path itself, which might be
* shared via NFS. And another "global" one, in /run, that
* uses the device/inode number. This has the benefit that we
* can even lock a tree that is a mount point, correctly. */
if (!path_is_absolute(path))
return -EINVAL;
if (getenv_bool("SYSTEMD_NSPAWN_LOCK") == 0) {
*local = *global = (LockFile) LOCK_FILE_INIT;
return 0;
}
if (path_equal(path, "/"))
return -EBUSY;
if (stat(path, &st) >= 0) {
if (asprintf(&p, "/run/systemd/nspawn/locks/inode-%lu:%lu", (unsigned long) st.st_dev, (unsigned long) st.st_ino) < 0)
return -ENOMEM;
}
r = make_lock_file_for(path, operation, &t);
if (r < 0)
return r;
if (p) {
mkdir_p("/run/systemd/nspawn/locks", 0700);
r = make_lock_file(p, operation, global);
if (r < 0) {
release_lock_file(&t);
return r;
}
} else
*global = (LockFile) LOCK_FILE_INIT;
*local = t;
return 0;
}
int image_set_limit(Image *i, uint64_t referenced_max) {
assert(i);
if (IMAGE_IS_VENDOR(i) || IMAGE_IS_HOST(i))
return -EROFS;
if (i->type != IMAGE_SUBVOLUME)
return -EOPNOTSUPP;
/* We set the quota both for the subvolume as well as for the
* subtree. The latter is mostly for historical reasons, since
* we didn't use to have a concept of subtree quota, and hence
* only modified the subvolume quota. */
(void) btrfs_qgroup_set_limit(i->path, 0, referenced_max);
(void) btrfs_subvol_auto_qgroup(i->path, 0, true);
return btrfs_subvol_set_subtree_quota_limit(i->path, 0, referenced_max);
}
int image_name_lock(const char *name, int operation, LockFile *ret) {
const char *p;
assert(name);
assert(ret);
/* Locks an image name, regardless of the precise path used. */
if (!image_name_is_valid(name))
return -EINVAL;
if (getenv_bool("SYSTEMD_NSPAWN_LOCK") == 0) {
*ret = (LockFile) LOCK_FILE_INIT;
return 0;
}
if (streq(name, ".host"))
return -EBUSY;
mkdir_p("/run/systemd/nspawn/locks", 0700);
p = strjoina("/run/systemd/nspawn/locks/name-", name);
return make_lock_file(p, operation, ret);
}
bool image_name_is_valid(const char *s) {
if (!filename_is_valid(s))
return false;
if (string_has_cc(s, NULL))
return false;
if (!utf8_is_valid(s))
return false;
/* Temporary files for atomically creating new files */
if (startswith(s, ".#"))
return false;
return true;
}
static const char* const image_type_table[_IMAGE_TYPE_MAX] = {
[IMAGE_DIRECTORY] = "directory",
[IMAGE_SUBVOLUME] = "subvolume",
[IMAGE_RAW] = "raw",
};
DEFINE_STRING_TABLE_LOOKUP(image_type, ImageType);
|