summaryrefslogtreecommitdiff
path: root/src/journal/coredump.c
blob: 442c969778a5ddbd861a5a13b04c2b8ef0f03e1f (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
/*-*- Mode: C; c-basic-offset: 8; indent-tabs-mode: nil -*-*/

/***
  This file is part of systemd.

  Copyright 2012 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 <errno.h>
#include <unistd.h>
#include <stdio.h>
#include <sys/prctl.h>
#include <sys/types.h>
#include <sys/xattr.h>

#include <systemd/sd-journal.h>
#include <systemd/sd-login.h>

#include "log.h"
#include "util.h"
#include "macro.h"
#include "mkdir.h"
#include "special.h"
#include "cgroup-util.h"
#include "journald-native.h"
#include "conf-parser.h"
#include "copy.h"
#include "stacktrace.h"

#ifdef HAVE_ACL
#include <sys/acl.h>
#include "acl-util.h"
#endif

/* The maximum size up to which we process coredumps */
#define PROCESS_SIZE_MAX ((off_t) (2LLU*1024LLU*1024LLU*1024LLU))

/* The maximum size up to which we leave the coredump around on
 * disk */
#define EXTERNAL_SIZE_MAX PROCESS_SIZE_MAX

/* The maximum size up to which we store the coredump in the
 * journal */
#define JOURNAL_SIZE_MAX ((size_t) (767LU*1024LU*1024LU))

/* Make sure to not make this larger than the maximum journal entry
 * size. See ENTRY_SIZE_MAX in journald-native.c. */
assert_cc(JOURNAL_SIZE_MAX <= ENTRY_SIZE_MAX);

enum {
        ARG_PID = 1,
        ARG_UID,
        ARG_GID,
        ARG_SIGNAL,
        ARG_TIMESTAMP,
        ARG_COMM,
        _ARG_MAX
};

typedef enum CoredumpStorage {
        COREDUMP_STORAGE_NONE,
        COREDUMP_STORAGE_EXTERNAL,
        COREDUMP_STORAGE_JOURNAL,
        COREDUMP_STORAGE_BOTH,
        _COREDUMP_STORAGE_MAX,
        _COREDUMP_STORAGE_INVALID = -1
} CoredumpStorage;

static CoredumpStorage arg_storage = COREDUMP_STORAGE_EXTERNAL;
static off_t arg_process_size_max = PROCESS_SIZE_MAX;
static off_t arg_external_size_max = EXTERNAL_SIZE_MAX;
static size_t arg_journal_size_max = JOURNAL_SIZE_MAX;

static const char* const coredump_storage_table[_COREDUMP_STORAGE_MAX] = {
        [COREDUMP_STORAGE_NONE] = "none",
        [COREDUMP_STORAGE_EXTERNAL] = "external",
        [COREDUMP_STORAGE_JOURNAL] = "journal",
        [COREDUMP_STORAGE_BOTH] = "both",
};

DEFINE_PRIVATE_STRING_TABLE_LOOKUP(coredump_storage, CoredumpStorage);
static DEFINE_CONFIG_PARSE_ENUM(config_parse_coredump_storage, coredump_storage, CoredumpStorage, "Failed to parse storage setting");

static int parse_config(void) {

        static const ConfigTableItem items[] = {
                { "Coredump", "ProcessSizeMax",  config_parse_iec_off,           0, &arg_process_size_max  },
                { "Coredump", "ExternalSizeMax", config_parse_iec_off,           0, &arg_external_size_max },
                { "Coredump", "JournalSizeMax",  config_parse_iec_size,          0, &arg_journal_size_max  },
                { "Coredump", "Storage",         config_parse_coredump_storage,  0, &arg_storage           },
                {}
        };

        return config_parse(
                        NULL,
                        "/etc/systemd/coredump.conf",
                        NULL,
                        "Coredump\0",
                        config_item_table_lookup,
                        (void*) items,
                        false,
                        false,
                        NULL);
}

static int fix_acl(int fd, uid_t uid) {

#ifdef HAVE_ACL
        _cleanup_(acl_freep) acl_t acl = NULL;
        acl_entry_t entry;
        acl_permset_t permset;

        if (uid <= SYSTEM_UID_MAX)
                return 0;

        /* Make sure normal users can read (but not write or delete)
         * their own coredumps */

        acl = acl_get_fd(fd);
        if (!acl) {
                log_error("Failed to get ACL: %m");
                return -errno;
        }

        if (acl_create_entry(&acl, &entry) < 0 ||
            acl_set_tag_type(entry, ACL_USER) < 0 ||
            acl_set_qualifier(entry, &uid) < 0) {
                log_error("Failed to patch ACL: %m");
                return -errno;
        }

        if (acl_get_permset(entry, &permset) < 0 ||
            acl_add_perm(permset, ACL_READ) < 0 ||
            calc_acl_mask_if_needed(&acl) < 0) {
                log_warning("Failed to patch ACL: %m");
                return -errno;
        }

        if (acl_set_fd(fd, acl) < 0) {
                log_error("Failed to apply ACL: %m");
                return -errno;
        }
#endif

        return 0;
}

static int fix_xattr(int fd, char *argv[]) {

        static const char * const xattrs[_ARG_MAX] = {
                [ARG_PID] = "user.coredump.pid",
                [ARG_UID] = "user.coredump.uid",
                [ARG_GID] = "user.coredump.gid",
                [ARG_SIGNAL] = "user.coredump.signal",
                [ARG_TIMESTAMP] = "user.coredump.timestamp",
                [ARG_COMM] = "user.coredump.comm",
        };

        int r = 0;
        unsigned i;

        /* Attach some metadate to coredumps via extended
         * attributes. Just because we can. */

        for (i = 0; i < _ARG_MAX; i++) {
                if (isempty(argv[i]))
                        continue;

                if (fsetxattr(fd, xattrs[i], argv[i], strlen(argv[i]), XATTR_CREATE) < 0)
                        r = -errno;
        }

        return r;
}

#define filename_escape(s) xescape((s), "./")

static int save_external_coredump(char **argv, uid_t uid, char **ret_filename, int *ret_fd, off_t *ret_size) {
        _cleanup_free_ char *p = NULL, *t = NULL, *c = NULL, *fn = NULL, *tmp = NULL;
        _cleanup_close_ int fd = -1;
        sd_id128_t boot;
        struct stat st;
        int r;

        assert(argv);
        assert(ret_filename);
        assert(ret_fd);
        assert(ret_size);

        c = filename_escape(argv[ARG_COMM]);
        if (!c)
                return log_oom();

        p = filename_escape(argv[ARG_PID]);
        if (!p)
                return log_oom();

        t = filename_escape(argv[ARG_TIMESTAMP]);
        if (!t)
                return log_oom();

        r = sd_id128_get_boot(&boot);
        if (r < 0) {
                log_error("Failed to determine boot ID: %s", strerror(-r));
                return r;
        }

        r = asprintf(&fn,
                     "/var/lib/systemd/coredump/core.%s." SD_ID128_FORMAT_STR ".%s.%s000000",
                     c,
                     SD_ID128_FORMAT_VAL(boot),
                     p,
                     t);
        if (r < 0)
                return log_oom();

        tmp = tempfn_random(fn);
        if (!tmp)
                return log_oom();

        mkdir_p_label("/var/lib/systemd/coredump", 0755);

        fd = open(tmp, O_CREAT|O_EXCL|O_RDWR|O_CLOEXEC|O_NOCTTY|O_NOFOLLOW, 0640);
        if (fd < 0) {
                log_error("Failed to create coredump file: %m");
                return -errno;
        }

        r = copy_bytes(STDIN_FILENO, fd);
        if (r < 0) {
                log_error("Failed to dump coredump to file: %s", strerror(-r));
                goto fail;
        }

        /* Ignore errors on these */
        fchmod(fd, 0640);
        fix_acl(fd, uid);
        fix_xattr(fd, argv);

        if (fsync(fd) < 0) {
                log_error("Failed to sync coredump: %m");
                r = -errno;
                goto fail;
        }

        if (fstat(fd, &st) < 0) {
                log_error("Failed to fstat coredump: %m");
                r = -errno;
                goto fail;
        }

        if (rename(tmp, fn) < 0) {
                log_error("Failed to rename coredump: %m");
                r = -errno;
                goto fail;
        }

        *ret_filename = fn;
        *ret_fd = fd;
        *ret_size = st.st_size;

        fn = NULL;
        fd = -1;

        return 0;

fail:
        unlink_noerrno(tmp);
        return r;
}

static int allocate_journal_field(int fd, size_t size, char **ret, size_t *ret_size) {
        _cleanup_free_ char *field = NULL;
        ssize_t n;

        assert(fd >= 0);
        assert(ret);
        assert(ret_size);

        if (lseek(fd, 0, SEEK_SET) == (off_t) -1) {
                log_warning("Failed to seek: %m");
                return -errno;
        }

        field = malloc(9 + size);
        if (!field) {
                log_warning("Failed to allocate memory fore coredump, coredump will not be stored.");
                return -ENOMEM;
        }

        memcpy(field, "COREDUMP=", 9);

        n = read(fd, field + 9, size);
        if (n < 0) {
                log_error("Failed to read core data: %s", strerror(-n));
                return (int) n;
        }
        if ((size_t) n < size) {
                log_error("Core data too short.");
                return -EIO;
        }

        *ret = field;
        *ret_size = size + 9;

        field = NULL;

        return 0;
}

static int maybe_remove_external_coredump(const char *filename, off_t size) {

        if (!filename)
                return 0;

        if (IN_SET(arg_storage, COREDUMP_STORAGE_EXTERNAL, COREDUMP_STORAGE_BOTH) &&
            size <= arg_external_size_max)
                return 0;

        if (unlink(filename) < 0) {
                log_error("Failed to unlink %s: %m", filename);
                return -errno;
        }

        return 0;
}

int main(int argc, char* argv[]) {

        _cleanup_free_ char *core_pid = NULL, *core_uid = NULL, *core_gid = NULL, *core_signal = NULL,
                *core_timestamp = NULL, *core_comm = NULL, *core_exe = NULL, *core_unit = NULL,
                *core_session = NULL, *core_message = NULL, *core_cmdline = NULL, *coredump_data = NULL,
                *coredump_filename = NULL, *core_slice = NULL, *core_cgroup = NULL, *core_owner_uid = NULL,
                *exe = NULL;

        _cleanup_close_ int coredump_fd = -1;

        struct iovec iovec[17];
        off_t coredump_size;
        int r, j = 0;
        uid_t uid, owner_uid;
        gid_t gid;
        pid_t pid;
        char *t;

        /* Make sure we never enter a loop */
        prctl(PR_SET_DUMPABLE, 0);

        /* First, log to a safe place, since we don't know what
         * crashed and it might be journald which we'd rather not log
         * to then. */
        log_set_target(LOG_TARGET_KMSG);
        log_open();

        if (argc != _ARG_MAX) {
                log_error("Invalid number of arguments passed from kernel.");
                r = -EINVAL;
                goto finish;
        }

        /* Ignore all parse errors */
        parse_config();
        log_debug("Selected storage '%s'.", coredump_storage_to_string(arg_storage));

        r = parse_uid(argv[ARG_UID], &uid);
        if (r < 0) {
                log_error("Failed to parse UID.");
                goto finish;
        }

        r = parse_pid(argv[ARG_PID], &pid);
        if (r < 0) {
                log_error("Failed to parse PID.");
                goto finish;
        }

        r = parse_gid(argv[ARG_GID], &gid);
        if (r < 0) {
                log_error("Failed to parse GID.");
                goto finish;
        }

        if (cg_pid_get_unit(pid, &t) >= 0) {

                if (streq(t, SPECIAL_JOURNALD_SERVICE)) {

                        /* If we are journald, we cut things short,
                         * don't write to the journal, but still
                         * create a coredump. */

                        if (arg_storage != COREDUMP_STORAGE_NONE)
                                arg_storage = COREDUMP_STORAGE_EXTERNAL;

                        r = save_external_coredump(argv, uid, &coredump_filename, &coredump_fd, &coredump_size);
                        if (r < 0)
                                goto finish;

                        r = maybe_remove_external_coredump(coredump_filename, coredump_size);
                        if (r < 0)
                                goto finish;

                        log_info("Detected coredump of the journal daemon itself, diverted to %s.", coredump_filename);
                        goto finish;
                }

                core_unit = strappend("COREDUMP_UNIT=", t);
        } else if (cg_pid_get_user_unit(pid, &t) >= 0)
                core_unit = strappend("COREDUMP_USER_UNIT=", t);

        if (core_unit)
                IOVEC_SET_STRING(iovec[j++], core_unit);

        /* OK, now we know it's not the journal, hence we can make use
         * of it now. */
        log_set_target(LOG_TARGET_JOURNAL_OR_KMSG);
        log_open();

        core_pid = strappend("COREDUMP_PID=", argv[ARG_PID]);
        if (core_pid)
                IOVEC_SET_STRING(iovec[j++], core_pid);

        core_uid = strappend("COREDUMP_UID=", argv[ARG_UID]);
        if (core_uid)
                IOVEC_SET_STRING(iovec[j++], core_uid);

        core_gid = strappend("COREDUMP_GID=", argv[ARG_GID]);
        if (core_gid)
                IOVEC_SET_STRING(iovec[j++], core_gid);

        core_signal = strappend("COREDUMP_SIGNAL=", argv[ARG_SIGNAL]);
        if (core_signal)
                IOVEC_SET_STRING(iovec[j++], core_signal);

        core_comm = strappend("COREDUMP_COMM=", argv[ARG_COMM]);
        if (core_comm)
                IOVEC_SET_STRING(iovec[j++], core_comm);

        if (sd_pid_get_session(pid, &t) >= 0) {
                core_session = strappend("COREDUMP_SESSION=", t);
                free(t);

                if (core_session)
                        IOVEC_SET_STRING(iovec[j++], core_session);
        }

        if (sd_pid_get_owner_uid(pid, &owner_uid) >= 0) {
                asprintf(&core_owner_uid, "COREDUMP_OWNER_UID=" UID_FMT, owner_uid);

                if (core_owner_uid)
                        IOVEC_SET_STRING(iovec[j++], core_owner_uid);
        }

        if (sd_pid_get_slice(pid, &t) >= 0) {
                core_slice = strappend("COREDUMP_SLICE=", t);
                free(t);

                if (core_slice)
                        IOVEC_SET_STRING(iovec[j++], core_slice);
        }

        if (get_process_exe(pid, &exe) >= 0) {
                core_exe = strappend("COREDUMP_EXE=", exe);
                if (core_exe)
                        IOVEC_SET_STRING(iovec[j++], core_exe);
        }

        if (get_process_cmdline(pid, 0, false, &t) >= 0) {
                core_cmdline = strappend("COREDUMP_CMDLINE=", t);
                free(t);

                if (core_cmdline)
                        IOVEC_SET_STRING(iovec[j++], core_cmdline);
        }

        if (cg_pid_get_path_shifted(pid, NULL, &t) >= 0) {
                core_cgroup = strappend("COREDUMP_CGROUP=", t);
                free(t);

                if (core_cgroup)
                        IOVEC_SET_STRING(iovec[j++], core_cgroup);
        }

        core_timestamp = strjoin("COREDUMP_TIMESTAMP=", argv[ARG_TIMESTAMP], "000000", NULL);
        if (core_timestamp)
                IOVEC_SET_STRING(iovec[j++], core_timestamp);

        IOVEC_SET_STRING(iovec[j++], "MESSAGE_ID=fc2e22bc6ee647b6b90729ab34a250b1");
        IOVEC_SET_STRING(iovec[j++], "PRIORITY=2");

        /* Always stream the coredump to disk, if that's possible */
        r = save_external_coredump(argv, uid, &coredump_filename, &coredump_fd, &coredump_size);
        if (r < 0)
                goto finish;

        /* If we don't want to keep the coredump on disk, remove it
         * now, as later on we will lack the privileges for
         * it. However, we keep the fd to it, so that we can still
         * process it and log it. */
        r = maybe_remove_external_coredump(coredump_filename, coredump_size);
        if (r < 0)
                goto finish;

        /* Now, let's drop privileges to become the user who owns the
         * segfaulted process and allocate the coredump memory under
         * his uid. This also ensures that the credentials journald
         * will see are the ones of the coredumping user, thus making
         * sure the user himself gets access to the core dump. */
        if (setresgid(gid, gid, gid) < 0 ||
            setresuid(uid, uid, uid) < 0) {
                log_error("Failed to drop privileges: %m");
                r = -errno;
                goto finish;
        }

#ifdef HAVE_ELFUTILS
        /* Try to get a strack trace if we can */
        if (coredump_size <= arg_process_size_max) {
                _cleanup_free_ char *stacktrace = NULL;

                r = coredump_make_stack_trace(coredump_fd, exe, &stacktrace);
                if (r >= 0)
                        core_message = strjoin("MESSAGE=Process ", argv[ARG_PID], " (", argv[ARG_COMM], ") of user ", argv[ARG_UID], " dumped core.\n\n", stacktrace, NULL);
                else
                        log_warning("Failed to generate stack trace: %s", strerror(-r));
        }

        if (!core_message)
#endif
        core_message = strjoin("MESSAGE=Process ", argv[ARG_PID], " (", argv[ARG_COMM], ") of user ", argv[ARG_UID], " dumped core.", NULL);
        if (core_message)
                IOVEC_SET_STRING(iovec[j++], core_message);

        /* Optionally store the entire coredump in the journal */
        if (IN_SET(arg_storage, COREDUMP_STORAGE_JOURNAL, COREDUMP_STORAGE_BOTH) &&
            coredump_size <= (off_t) arg_journal_size_max) {
                size_t sz;

                /* Store the coredump itself in the journal */

                r = allocate_journal_field(coredump_fd, (size_t) coredump_size, &coredump_data, &sz);
                if (r >= 0) {
                        iovec[j].iov_base = coredump_data;
                        iovec[j].iov_len = sz;
                        j++;
                }
        }

        r = sd_journal_sendv(iovec, j);
        if (r < 0)
                log_error("Failed to log coredump: %s", strerror(-r));

finish:
        return r < 0 ? EXIT_FAILURE : EXIT_SUCCESS;
}