-
Notifications
You must be signed in to change notification settings - Fork 0
/
minijail0_cli.c
1107 lines (1039 loc) · 30.8 KB
/
minijail0_cli.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
/* Copyright 2018 The Chromium OS Authors. All rights reserved.
* Use of this source code is governed by a BSD-style license that can be
* found in the LICENSE file.
*/
#include <dlfcn.h>
#include <errno.h>
#include <getopt.h>
#include <inttypes.h>
#include <stdbool.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <sys/capability.h>
#include <sys/mount.h>
#include <sys/types.h>
#include <unistd.h>
#include <linux/filter.h>
#include "libminijail.h"
#include "libsyscalls.h"
#include "elfparse.h"
#include "minijail0_cli.h"
#include "system.h"
#include "util.h"
#define IDMAP_LEN 32U
#define DEFAULT_TMP_SIZE (64 * 1024 * 1024)
/*
* A malloc() that aborts on failure. We only implement this in the CLI as
* the library should return ENOMEM errors when allocations fail.
*/
static void *xmalloc(size_t size)
{
void *ret = malloc(size);
if (!ret) {
perror("malloc() failed");
exit(1);
}
return ret;
}
static char *xstrdup(const char *s)
{
char *ret = strdup(s);
if (!ret) {
perror("strdup() failed");
exit(1);
}
return ret;
}
static void set_user(struct minijail *j, const char *arg, uid_t *out_uid,
gid_t *out_gid)
{
char *end = NULL;
int uid = strtod(arg, &end);
if (!*end && *arg) {
*out_uid = uid;
minijail_change_uid(j, uid);
return;
}
int ret = lookup_user(arg, out_uid, out_gid);
if (ret) {
fprintf(stderr, "Bad user '%s': %s\n", arg, strerror(-ret));
exit(1);
}
ret = minijail_change_user(j, arg);
if (ret) {
fprintf(stderr, "minijail_change_user('%s') failed: %s\n", arg,
strerror(-ret));
exit(1);
}
}
static void set_group(struct minijail *j, const char *arg, gid_t *out_gid)
{
char *end = NULL;
int gid = strtod(arg, &end);
if (!*end && *arg) {
*out_gid = gid;
minijail_change_gid(j, gid);
return;
}
int ret = lookup_group(arg, out_gid);
if (ret) {
fprintf(stderr, "Bad group '%s': %s\n", arg, strerror(-ret));
exit(1);
}
minijail_change_gid(j, *out_gid);
}
/*
* Helper function used by --add-suppl-group (possibly more than once),
* to build the supplementary gids array.
*/
static void suppl_group_add(size_t *suppl_gids_count, gid_t **suppl_gids,
char *arg) {
char *end = NULL;
int groupid = strtod(arg, &end);
gid_t gid;
int ret;
if (!*end && *arg) {
/* A gid number has been specified, proceed. */
gid = groupid;
} else if ((ret = lookup_group(arg, &gid))) {
/*
* A group name has been specified,
* but doesn't exist: we bail out.
*/
fprintf(stderr, "Bad group '%s': %s\n", arg, strerror(-ret));
exit(1);
}
/*
* From here, gid is guaranteed to be set and valid,
* we add it to our supplementary gids array.
*/
*suppl_gids = realloc(*suppl_gids,
sizeof(gid_t) * ++(*suppl_gids_count));
if (!suppl_gids) {
fprintf(stderr, "failed to allocate memory.\n");
exit(1);
}
(*suppl_gids)[*suppl_gids_count - 1] = gid;
}
static void skip_securebits(struct minijail *j, const char *arg)
{
uint64_t securebits_skip_mask;
char *end = NULL;
securebits_skip_mask = strtoull(arg, &end, 16);
if (*end) {
fprintf(stderr, "Invalid securebit mask: '%s'\n", arg);
exit(1);
}
minijail_skip_setting_securebits(j, securebits_skip_mask);
}
static void use_caps(struct minijail *j, const char *arg)
{
uint64_t caps = 0;
cap_t parsed_caps = cap_from_text(arg);
if (parsed_caps != NULL) {
unsigned int i;
const uint64_t one = 1;
cap_flag_value_t cap_value;
unsigned int last_valid_cap = get_last_valid_cap();
for (i = 0; i <= last_valid_cap; ++i) {
if (cap_get_flag(parsed_caps, i, CAP_EFFECTIVE,
&cap_value)) {
if (errno == EINVAL) {
/*
* Some versions of libcap reject any
* capabilities they were not compiled
* with by returning EINVAL.
*/
continue;
}
fprintf(stderr,
"Could not get the value of "
"the %d-th capability: %m\n",
i);
exit(1);
}
if (cap_value == CAP_SET)
caps |= (one << i);
}
cap_free(parsed_caps);
} else {
char *end = NULL;
caps = strtoull(arg, &end, 16);
if (*end) {
fprintf(stderr, "Invalid cap set: '%s'\n", arg);
exit(1);
}
}
minijail_use_caps(j, caps);
}
static void add_binding(struct minijail *j, char *arg)
{
char *src = tokenize(&arg, ",");
char *dest = tokenize(&arg, ",");
char *flags = tokenize(&arg, ",");
if (!src || src[0] == '\0' || arg != NULL) {
fprintf(stderr, "Bad binding: %s %s\n", src, dest);
exit(1);
}
if (dest == NULL || dest[0] == '\0')
dest = src;
int writable;
if (flags == NULL || flags[0] == '\0' || !strcmp(flags, "0"))
writable = 0;
else if (!strcmp(flags, "1"))
writable = 1;
else {
fprintf(stderr, "Bad value for <writable>: %s\n", flags);
exit(1);
}
if (minijail_bind(j, src, dest, writable)) {
fprintf(stderr, "minijail_bind failed.\n");
exit(1);
}
}
static void add_rlimit(struct minijail *j, char *arg)
{
char *type = tokenize(&arg, ",");
char *cur = tokenize(&arg, ",");
char *max = tokenize(&arg, ",");
char *end;
if (!type || type[0] == '\0' || !cur || cur[0] == '\0' ||
!max || max[0] == '\0' || arg != NULL) {
fprintf(stderr, "Bad rlimit '%s'.\n", arg);
exit(1);
}
rlim_t cur_rlim;
rlim_t max_rlim;
if (!strcmp(cur, "unlimited")) {
cur_rlim = RLIM_INFINITY;
} else {
end = NULL;
cur_rlim = strtoul(cur, &end, 0);
if (*end) {
fprintf(stderr, "Bad soft limit: '%s'.\n", cur);
exit(1);
}
}
if (!strcmp(max, "unlimited")) {
max_rlim = RLIM_INFINITY;
} else {
end = NULL;
max_rlim = strtoul(max, &end, 0);
if (*end) {
fprintf(stderr, "Bad hard limit: '%s'.\n", max);
exit(1);
}
}
end = NULL;
int resource = parse_single_constant(type, &end);
if (type == end) {
fprintf(stderr, "Bad rlimit: '%s'.\n", type);
exit(1);
}
if (minijail_rlimit(j, resource, cur_rlim, max_rlim)) {
fprintf(stderr, "minijail_rlimit '%s,%s,%s' failed.\n", type,
cur, max);
exit(1);
}
}
static void add_mount(struct minijail *j, char *arg)
{
char *src = tokenize(&arg, ",");
char *dest = tokenize(&arg, ",");
char *type = tokenize(&arg, ",");
char *flags = tokenize(&arg, ",");
char *data = tokenize(&arg, ",");
char *end;
if (!src || src[0] == '\0' || !dest || dest[0] == '\0' ||
!type || type[0] == '\0') {
fprintf(stderr, "Bad mount: %s %s %s\n", src, dest, type);
exit(1);
}
/*
* Fun edge case: the data option itself is comma delimited. If there
* were no more options, then arg would be set to NULL. But if we had
* more pending, it'll be pointing to the next token. Back up and undo
* the null byte so it'll be merged back.
* An example:
* none,/tmp,tmpfs,0xe,mode=0755,uid=10,gid=10
* The tokenize calls above will turn this memory into:
* none\0/tmp\0tmpfs\00xe\0mode=0755\0uid=10,gid=10
* With data pointing at mode=0755 and arg pointing at uid=10,gid=10.
*/
if (arg != NULL)
arg[-1] = ',';
unsigned long mountflags;
if (flags == NULL || flags[0] == '\0') {
mountflags = 0;
} else {
end = NULL;
mountflags = parse_constant(flags, &end);
if (flags == end) {
fprintf(stderr, "Bad mount flags: %s\n", flags);
exit(1);
}
}
if (minijail_mount_with_data(j, src, dest, type,
mountflags, data)) {
fprintf(stderr, "minijail_mount failed.\n");
exit(1);
}
}
static char *build_idmap(id_t id, id_t lowerid)
{
int ret;
char *idmap = xmalloc(IDMAP_LEN);
ret = snprintf(idmap, IDMAP_LEN, "%d %d 1", id, lowerid);
if (ret < 0 || (size_t)ret >= IDMAP_LEN) {
free(idmap);
fprintf(stderr, "Could not build id map.\n");
exit(1);
}
return idmap;
}
static int has_cap_setgid(void)
{
cap_t caps;
cap_flag_value_t cap_value;
if (!CAP_IS_SUPPORTED(CAP_SETGID))
return 0;
caps = cap_get_proc();
if (!caps) {
fprintf(stderr, "Could not get process' capabilities: %m\n");
exit(1);
}
if (cap_get_flag(caps, CAP_SETGID, CAP_EFFECTIVE, &cap_value)) {
fprintf(stderr, "Could not get the value of CAP_SETGID: %m\n");
exit(1);
}
if (cap_free(caps)) {
fprintf(stderr, "Could not free capabilities: %m\n");
exit(1);
}
return cap_value == CAP_SET;
}
static void set_ugid_mapping(struct minijail *j, int set_uidmap, uid_t uid,
char *uidmap, int set_gidmap, gid_t gid,
char *gidmap)
{
if (set_uidmap) {
minijail_namespace_user(j);
minijail_namespace_pids(j);
if (!uidmap) {
/*
* If no map is passed, map the current uid to the
* chosen uid in the target namespace (or root, if none
* was chosen).
*/
uidmap = build_idmap(uid, getuid());
}
if (0 != minijail_uidmap(j, uidmap)) {
fprintf(stderr, "Could not set uid map.\n");
exit(1);
}
free(uidmap);
}
if (set_gidmap) {
minijail_namespace_user(j);
minijail_namespace_pids(j);
if (!gidmap) {
/*
* If no map is passed, map the current gid to the
* chosen gid in the target namespace.
*/
gidmap = build_idmap(gid, getgid());
}
if (!has_cap_setgid()) {
/*
* This means that we are not running as root,
* so we also have to disable setgroups(2) to
* be able to set the gid map.
* See
* http://man7.org/linux/man-pages/man7/user_namespaces.7.html
*/
minijail_namespace_user_disable_setgroups(j);
}
if (0 != minijail_gidmap(j, gidmap)) {
fprintf(stderr, "Could not set gid map.\n");
exit(1);
}
free(gidmap);
}
}
static void use_chroot(struct minijail *j, const char *path, int *chroot,
int pivot_root)
{
if (pivot_root) {
fprintf(stderr, "Could not set chroot because "
"'-P' was specified.\n");
exit(1);
}
if (minijail_enter_chroot(j, path)) {
fprintf(stderr, "Could not set chroot.\n");
exit(1);
}
*chroot = 1;
}
static void use_pivot_root(struct minijail *j, const char *path,
int *pivot_root, int chroot)
{
if (chroot) {
fprintf(stderr, "Could not set pivot_root because "
"'-C' was specified.\n");
exit(1);
}
if (minijail_enter_pivot_root(j, path)) {
fprintf(stderr, "Could not set pivot_root.\n");
exit(1);
}
minijail_namespace_vfs(j);
*pivot_root = 1;
}
static void use_profile(struct minijail *j, const char *profile,
int *pivot_root, int chroot, size_t *tmp_size)
{
/* Note: New profiles should be added in minijail0_cli_unittest.cc. */
if (!strcmp(profile, "minimalistic-mountns") ||
!strcmp(profile, "minimalistic-mountns-nodev")) {
minijail_namespace_vfs(j);
if (minijail_bind(j, "/", "/", 0)) {
fprintf(stderr, "minijail_bind(/) failed.\n");
exit(1);
}
if (minijail_bind(j, "/proc", "/proc", 0)) {
fprintf(stderr, "minijail_bind(/proc) failed.\n");
exit(1);
}
if (!strcmp(profile, "minimalistic-mountns")) {
if (minijail_bind(j, "/dev/log", "/dev/log", 0)) {
fprintf(stderr, "minijail_bind(/dev/log) failed.\n");
exit(1);
}
minijail_mount_dev(j);
}
if (!*tmp_size) {
/* Avoid clobbering |tmp_size| if it was already set. */
*tmp_size = DEFAULT_TMP_SIZE;
}
minijail_remount_proc_readonly(j);
use_pivot_root(j, DEFAULT_PIVOT_ROOT, pivot_root, chroot);
} else {
fprintf(stderr, "Unrecognized profile name '%s'\n", profile);
exit(1);
}
}
static void set_remount_mode(struct minijail *j, const char *mode)
{
unsigned long msmode;
if (!strcmp(mode, "shared"))
msmode = MS_SHARED;
else if (!strcmp(mode, "private"))
msmode = MS_PRIVATE;
else if (!strcmp(mode, "slave"))
msmode = MS_SLAVE;
else if (!strcmp(mode, "unbindable"))
msmode = MS_UNBINDABLE;
else {
fprintf(stderr, "Unknown remount mode: '%s'\n", mode);
exit(1);
}
minijail_remount_mode(j, msmode);
}
static void read_seccomp_filter(const char *filter_path,
struct sock_fprog *filter)
{
FILE *f = fopen(filter_path, "re");
if (!f) {
fprintf(stderr, "failed to open %s: %m", filter_path);
exit(1);
}
off_t filter_size = 0;
if (fseeko(f, 0, SEEK_END) == -1 || (filter_size = ftello(f)) == -1) {
fclose(f);
fprintf(stderr, "failed to get file size of %s: %m",
filter_path);
exit(1);
}
if (filter_size % sizeof(struct sock_filter) != 0) {
fclose(f);
fprintf(stderr,
"filter size (%" PRId64
") of %s is not a multiple of %zu: %m",
filter_size, filter_path, sizeof(struct sock_filter));
exit(1);
}
rewind(f);
filter->len = filter_size / sizeof(struct sock_filter);
filter->filter = xmalloc(filter_size);
if (fread(filter->filter, sizeof(struct sock_filter), filter->len, f) !=
filter->len) {
fclose(f);
fprintf(stderr, "failed read %s: %m", filter_path);
exit(1);
}
fclose(f);
}
static void usage(const char *progn)
{
size_t i;
/* clang-format off */
printf("Usage: %s [-dGhHiIKlLnNprRstUvyYz]\n"
" [-a <table>]\n"
" [-b <src>[,[dest][,<writeable>]]] [-k <src>,<dest>,<type>[,<flags>[,<data>]]]\n"
" [-c <caps>] [-C <dir>] [-P <dir>] [-e[file]] [-f <file>] [-g <group>]\n"
" [-m[<uid> <loweruid> <count>]*] [-M[<gid> <lowergid> <count>]*] [--profile <name>]\n"
" [-R <type,cur,max>] [-S <file>] [-t[size]] [-T <type>] [-u <user>] [-V <file>]\n"
" <program> [args...]\n"
" -a <table>: Use alternate syscall table <table>.\n"
" -b <...>: Bind <src> to <dest> in chroot.\n"
" Multiple instances allowed.\n"
" -B <mask>: Skip setting securebits in <mask> when restricting capabilities (-c).\n"
" By default, SECURE_NOROOT, SECURE_NO_SETUID_FIXUP, and \n"
" SECURE_KEEP_CAPS (together with their respective locks) are set.\n"
" There are eight securebits in total.\n"
" -k <...>: Mount <src> at <dest> in chroot.\n"
" <flags> and <data> can be specified as in mount(2).\n"
" Multiple instances allowed.\n"
" -c <caps>: Restrict caps to <caps>.\n"
" -C <dir>: chroot(2) to <dir>.\n"
" Not compatible with -P.\n"
" -P <dir>: pivot_root(2) to <dir> (implies -v).\n"
" Not compatible with -C.\n"
" --mount-dev, Create a new /dev with a minimal set of device nodes (implies -v).\n"
" -d: See the minijail0(1) man page for the exact set.\n"
" -e[file]: Enter new network namespace, or existing one if |file| is provided.\n"
" -f <file>: Write the pid of the jailed process to <file>.\n"
" -g <group>: Change gid to <group>.\n"
" -G: Inherit supplementary groups from new uid.\n"
" Not compatible with -y or --add-suppl-group.\n"
" -y: Keep original uid's supplementary groups.\n"
" Not compatible with -G or --add-suppl-group.\n"
" --add-suppl-group <g>:Add <g> to the proccess' supplementary groups,\n"
" can be specified multiple times to add several groups.\n"
" Not compatible with -y or -G.\n"
" -h: Help (this message).\n"
" -H: Seccomp filter help message.\n"
" -i: Exit immediately after fork(2). The jailed process will run\n"
" in the background.\n"
" -I: Run <program> as init (pid 1) inside a new pid namespace (implies -p).\n"
" -K: Do not change share mode of any existing mounts.\n"
" -K<mode>: Mark all existing mounts as <mode> instead of MS_PRIVATE.\n"
" -l: Enter new IPC namespace.\n"
" -L: Report blocked syscalls when using seccomp filter.\n"
" If the kernel does not support SECCOMP_RET_LOG,\n"
" forces the following syscalls to be allowed:\n"
" ", progn);
/* clang-format on */
for (i = 0; i < log_syscalls_len; i++)
printf("%s ", log_syscalls[i]);
/* clang-format off */
printf("\n"
" -m[map]: Set the uid map of a user namespace (implies -pU).\n"
" Same arguments as newuidmap(1), multiple mappings should be separated by ',' (comma).\n"
" With no mapping, map the current uid to root inside the user namespace.\n"
" Not compatible with -b without the 'writable' option.\n"
" -M[map]: Set the gid map of a user namespace (implies -pU).\n"
" Same arguments as newgidmap(1), multiple mappings should be separated by ',' (comma).\n"
" With no mapping, map the current gid to root inside the user namespace.\n"
" Not compatible with -b without the 'writable' option.\n"
" -n: Set no_new_privs.\n"
" -N: Enter a new cgroup namespace.\n"
" -p: Enter new pid namespace (implies -vr).\n"
" -r: Remount /proc read-only (implies -v).\n"
" -R: Set rlimits, can be specified multiple times.\n"
" -s: Use seccomp mode 1 (not the same as -S).\n"
" -S <file>: Set seccomp filter using <file>.\n"
" E.g., '-S /usr/share/filters/<prog>.$(uname -m)'.\n"
" Requires -n when not running as root.\n"
" -t[size]: Mount tmpfs at /tmp (implies -v).\n"
" Optional argument specifies size (default \"64M\").\n"
" -T <type>: Assume <program> is a <type> ELF binary; <type> can be 'static' or 'dynamic'.\n"
" This will avoid accessing <program> binary before execve(2).\n"
" Type 'static' will avoid preload hooking.\n"
" -u <user>: Change uid to <user>.\n"
" -U: Enter new user namespace (implies -p).\n"
" -v: Enter new mount namespace.\n"
" -V <file>: Enter specified mount namespace.\n"
" -w: Create and join a new anonymous session keyring.\n"
" -Y: Synchronize seccomp filters across thread group.\n"
" -z: Don't forward signals to jailed process.\n"
" --ambient: Raise ambient capabilities. Requires -c.\n"
" --uts[=name]: Enter a new UTS namespace (and set hostname).\n"
" --logging=<s>:Use <s> as the logging system.\n"
" <s> must be 'auto' (default), 'syslog', or 'stderr'.\n"
" --profile <p>:Configure minijail0 to run with the <p> sandboxing profile,\n"
" which is a convenient way to express multiple flags\n"
" that are typically used together.\n"
" See the minijail0(1) man page for the full list.\n"
" --preload-library=<f>:Overrides the path to \"" PRELOADPATH "\".\n"
" This is only really useful for local testing.\n"
" --seccomp-bpf-binary=<f>:Set a pre-compiled seccomp filter using <f>.\n"
" E.g., '-S /usr/share/filters/<prog>.$(uname -m).bpf'.\n"
" Requires -n when not running as root.\n"
" The user is responsible for ensuring that the binary\n"
" was compiled for the correct architecture / kernel version.\n"
" --allow-speculative-execution:Allow speculative execution and disable\n"
" mitigations for speculative execution attacks.\n");
/* clang-format on */
}
static void seccomp_filter_usage(const char *progn)
{
const struct syscall_entry *entry = syscall_table;
printf("Usage: %s -S <policy.file> <program> [args...]\n\n"
"System call names supported:\n",
progn);
for (; entry->name && entry->nr >= 0; ++entry)
printf(" %s [%d]\n", entry->name, entry->nr);
printf("\nSee minijail0(5) for example policies.\n");
}
int parse_args(struct minijail *j, int argc, char *const argv[],
int *exit_immediately, ElfType *elftype,
const char **preload_path)
{
int opt;
int use_seccomp_filter = 0, use_seccomp_filter_binary = 0;
int forward = 1;
int binding = 0;
int chroot = 0, pivot_root = 0;
int mount_ns = 0, change_remount = 0;
const char *remount_mode = NULL;
int inherit_suppl_gids = 0, keep_suppl_gids = 0;
int caps = 0, ambient_caps = 0;
int seccomp = -1;
bool use_uid = false, use_gid = false;
uid_t uid = 0;
gid_t gid = 0;
gid_t *suppl_gids = NULL;
size_t suppl_gids_count = 0;
char *uidmap = NULL, *gidmap = NULL;
int set_uidmap = 0, set_gidmap = 0;
size_t tmp_size = 0;
const char *filter_path = NULL;
int log_to_stderr = -1;
const char *optstring =
"+u:g:sS:c:C:P:b:B:V:f:m::M::k:a:e::R:T:vrGhHinNplLt::IUK::wyYzd";
/* clang-format off */
const struct option long_options[] = {
{"help", no_argument, 0, 'h'},
{"mount-dev", no_argument, 0, 'd'},
{"ambient", no_argument, 0, 128},
{"uts", optional_argument, 0, 129},
{"logging", required_argument, 0, 130},
{"profile", required_argument, 0, 131},
{"preload-library", required_argument, 0, 132},
{"seccomp-bpf-binary", required_argument, 0, 133},
{"add-suppl-group", required_argument, 0, 134},
{"allow-speculative-execution", no_argument, 0, 135},
{0, 0, 0, 0},
};
/* clang-format on */
while ((opt = getopt_long(argc, argv, optstring, long_options, NULL)) !=
-1) {
switch (opt) {
case 'u':
if (use_uid) {
fprintf(stderr,
"-u provided multiple times.\n");
exit(1);
}
use_uid = true;
set_user(j, optarg, &uid, &gid);
break;
case 'g':
if (use_gid) {
fprintf(stderr,
"-g provided multiple times.\n");
exit(1);
}
use_gid = true;
set_group(j, optarg, &gid);
break;
case 'n':
minijail_no_new_privs(j);
break;
case 's':
if (seccomp != -1 && seccomp != 1) {
fprintf(stderr,
"Do not use -s, -S, or "
"--seccomp-bpf-binary together.\n");
exit(1);
}
seccomp = 1;
minijail_use_seccomp(j);
break;
case 'S':
if (seccomp != -1 && seccomp != 2) {
fprintf(stderr,
"Do not use -s, -S, or "
"--seccomp-bpf-binary together.\n");
exit(1);
}
seccomp = 2;
minijail_use_seccomp_filter(j);
filter_path = optarg;
use_seccomp_filter = 1;
break;
case 'l':
minijail_namespace_ipc(j);
break;
case 'L':
minijail_log_seccomp_filter_failures(j);
break;
case 'b':
add_binding(j, optarg);
binding = 1;
break;
case 'B':
skip_securebits(j, optarg);
break;
case 'c':
caps = 1;
use_caps(j, optarg);
break;
case 'C':
use_chroot(j, optarg, &chroot, pivot_root);
break;
case 'k':
add_mount(j, optarg);
break;
case 'K':
remount_mode = optarg;
change_remount = 1;
break;
case 'P':
use_pivot_root(j, optarg, &pivot_root, chroot);
break;
case 'f':
if (0 != minijail_write_pid_file(j, optarg)) {
fprintf(stderr,
"Could not prepare pid file path.\n");
exit(1);
}
break;
case 't':
minijail_namespace_vfs(j);
if (!tmp_size) {
/*
* Avoid clobbering |tmp_size| if it was already
* set.
*/
tmp_size = DEFAULT_TMP_SIZE;
}
if (optarg != NULL &&
0 != parse_size(&tmp_size, optarg)) {
fprintf(stderr, "Invalid /tmp tmpfs size.\n");
exit(1);
}
break;
case 'v':
minijail_namespace_vfs(j);
/*
* Set the default mount propagation in the command-line
* tool to MS_SLAVE.
*
* When executing the sandboxed program in a new mount
* namespace the Minijail library will by default
* remount all mounts with the MS_PRIVATE flag. While
* this is an appropriate, safe default for the library,
* MS_PRIVATE can be problematic: unmount events will
* not propagate into mountpoints marked as MS_PRIVATE.
* This means that if a mount is unmounted in the root
* mount namespace, it will not be unmounted in the
* non-root mount namespace.
* This in turn can be problematic because activity in
* the non-root mount namespace can now directly
* influence the root mount namespace (e.g. preventing
* re-mounts of said mount), which would be a privilege
* inversion.
*
* Setting the default in the command-line to MS_SLAVE
* will still prevent mounts from leaking out of the
* non-root mount namespace but avoid these
* privilege-inversion issues.
* For cases where mounts should not flow *into* the
* namespace either, the user can pass -Kprivate.
* Note that mounts are marked as MS_PRIVATE by default
* by the kernel, so unless the init process (like
* systemd) or something else marks them as shared, this
* won't do anything.
*/
minijail_remount_mode(j, MS_SLAVE);
mount_ns = 1;
break;
case 'V':
minijail_namespace_enter_vfs(j, optarg);
break;
case 'r':
minijail_remount_proc_readonly(j);
break;
case 'G':
if (keep_suppl_gids) {
fprintf(stderr,
"-y and -G are not compatible.\n");
exit(1);
}
minijail_inherit_usergroups(j);
inherit_suppl_gids = 1;
break;
case 'y':
if (inherit_suppl_gids) {
fprintf(stderr,
"-y and -G are not compatible.\n");
exit(1);
}
minijail_keep_supplementary_gids(j);
keep_suppl_gids = 1;
break;
case 'N':
minijail_namespace_cgroups(j);
break;
case 'p':
minijail_namespace_pids(j);
break;
case 'e':
if (optarg)
minijail_namespace_enter_net(j, optarg);
else
minijail_namespace_net(j);
break;
case 'i':
*exit_immediately = 1;
break;
case 'H':
seccomp_filter_usage(argv[0]);
exit(0);
case 'I':
minijail_namespace_pids(j);
minijail_run_as_init(j);
break;
case 'U':
minijail_namespace_user(j);
minijail_namespace_pids(j);
break;
case 'm':
set_uidmap = 1;
if (uidmap) {
free(uidmap);
uidmap = NULL;
}
if (optarg)
uidmap = xstrdup(optarg);
break;
case 'M':
set_gidmap = 1;
if (gidmap) {
free(gidmap);
gidmap = NULL;
}
if (optarg)
gidmap = xstrdup(optarg);
break;
case 'a':
if (0 != minijail_use_alt_syscall(j, optarg)) {
fprintf(stderr,
"Could not set alt-syscall table.\n");
exit(1);
}
break;
case 'R':
add_rlimit(j, optarg);
break;
case 'T':
if (!strcmp(optarg, "static"))
*elftype = ELFSTATIC;
else if (!strcmp(optarg, "dynamic"))
*elftype = ELFDYNAMIC;
else {
fprintf(stderr, "ELF type must be 'static' or "
"'dynamic'.\n");
exit(1);
}
break;
case 'w':
minijail_new_session_keyring(j);
break;
case 'Y':
minijail_set_seccomp_filter_tsync(j);
break;
case 'z':
forward = 0;
break;
case 'd':
minijail_namespace_vfs(j);
minijail_mount_dev(j);
break;
/* Long options. */
case 128: /* Ambient caps. */
ambient_caps = 1;
minijail_set_ambient_caps(j);
break;
case 129: /* UTS/hostname namespace. */
minijail_namespace_uts(j);
if (optarg)
minijail_namespace_set_hostname(j, optarg);
break;
case 130: /* Logging. */
if (!strcmp(optarg, "auto")) {
log_to_stderr = -1;
} else if (!strcmp(optarg, "syslog")) {
log_to_stderr = 0;
} else if (!strcmp(optarg, "stderr")) {
log_to_stderr = 1;
} else {
fprintf(stderr, "--logger must be 'syslog' or "
"'stderr'.\n");
exit(1);
}
break;
case 131: /* Profile */
use_profile(j, optarg, &pivot_root, chroot, &tmp_size);
break;
case 132: /* PRELOADPATH */
*preload_path = optarg;
break;
case 133: /* seccomp-bpf binary. */
if (seccomp != -1 && seccomp != 3) {
fprintf(stderr,
"Do not use -s, -S, or "
"--seccomp-bpf-binary together.\n");
exit(1);
}
seccomp = 3;
minijail_use_seccomp_filter(j);
filter_path = optarg;
use_seccomp_filter_binary = 1;
break;
case 134:
suppl_group_add(&suppl_gids_count, &suppl_gids,
optarg);
break;
case 135:
minijail_set_seccomp_filter_allow_speculation(j);
break;
default:
usage(argv[0]);
exit(opt == 'h' ? 0 : 1);
}
}
if (log_to_stderr == -1) {
/* Autodetect default logging output. */
log_to_stderr = isatty(STDIN_FILENO) ? 1 : 0;
}
if (log_to_stderr) {
init_logging(LOG_TO_FD, STDERR_FILENO, LOG_INFO);
/*
* When logging to stderr, ensure the FD survives the jailing.
*/
if (0 !=
minijail_preserve_fd(j, STDERR_FILENO, STDERR_FILENO)) {
fprintf(stderr, "Could not preserve stderr.\n");
exit(1);
}
}
/* Set up uid/gid mapping. */
if (set_uidmap || set_gidmap) {
set_ugid_mapping(j, set_uidmap, uid, uidmap, set_gidmap, gid,
gidmap);
}
/* Can only set ambient caps when using regular caps. */
if (ambient_caps && !caps) {
fprintf(stderr, "Can't set ambient capabilities (--ambient) "
"without actually using capabilities (-c).\n");
exit(1);
}