summaryrefslogtreecommitdiffstats
path: root/src/journal/journald-native.c
blob: 315ec0b51307f994a326f19bd58a2db0a5284285 (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
/* SPDX-License-Identifier: LGPL-2.1-or-later */

#include <stddef.h>
#include <sys/epoll.h>
#include <sys/mman.h>
#include <sys/statvfs.h>
#include <unistd.h>

#include "alloc-util.h"
#include "fd-util.h"
#include "fs-util.h"
#include "iovec-util.h"
#include "journal-importer.h"
#include "journal-internal.h"
#include "journal-util.h"
#include "journald-client.h"
#include "journald-console.h"
#include "journald-kmsg.h"
#include "journald-native.h"
#include "journald-server.h"
#include "journald-syslog.h"
#include "journald-wall.h"
#include "memfd-util.h"
#include "memory-util.h"
#include "missing_fcntl.h"
#include "parse-util.h"
#include "path-util.h"
#include "process-util.h"
#include "selinux-util.h"
#include "socket-util.h"
#include "string-util.h"
#include "strv.h"
#include "unaligned.h"

static bool allow_object_pid(const struct ucred *ucred) {
        return ucred && ucred->uid == 0;
}

static void server_process_entry_meta(
                const char *p, size_t l,
                const struct ucred *ucred,
                int *priority,
                char **identifier,
                char **message,
                pid_t *object_pid) {

        /* We need to determine the priority of this entry for the rate limiting logic */

        if (l == 10 &&
            startswith(p, "PRIORITY=") &&
            p[9] >= '0' && p[9] <= '9')
                *priority = (*priority & LOG_FACMASK) | (p[9] - '0');

        else if (l == 17 &&
                 startswith(p, "SYSLOG_FACILITY=") &&
                 p[16] >= '0' && p[16] <= '9')
                *priority = (*priority & LOG_PRIMASK) | ((p[16] - '0') << 3);

        else if (l == 18 &&
                 startswith(p, "SYSLOG_FACILITY=") &&
                 p[16] >= '0' && p[16] <= '9' &&
                 p[17] >= '0' && p[17] <= '9')
                *priority = (*priority & LOG_PRIMASK) | (((p[16] - '0')*10 + (p[17] - '0')) << 3);

        else if (l >= 19 &&
                 startswith(p, "SYSLOG_IDENTIFIER=")) {
                char *t;

                t = memdup_suffix0(p + 18, l - 18);
                if (t)
                        free_and_replace(*identifier, t);

        } else if (l >= 8 &&
                   startswith(p, "MESSAGE=")) {
                char *t;

                t = memdup_suffix0(p + 8, l - 8);
                if (t)
                        free_and_replace(*message, t);

        } else if (l > STRLEN("OBJECT_PID=") &&
                   l < STRLEN("OBJECT_PID=")  + DECIMAL_STR_MAX(pid_t) &&
                   startswith(p, "OBJECT_PID=") &&
                   allow_object_pid(ucred)) {
                char buf[DECIMAL_STR_MAX(pid_t)];
                memcpy(buf, p + STRLEN("OBJECT_PID="),
                       l - STRLEN("OBJECT_PID="));
                buf[l-STRLEN("OBJECT_PID=")] = '\0';

                (void) parse_pid(buf, object_pid);
        }
}

static int server_process_entry(
                Server *s,
                const void *buffer, size_t *remaining,
                ClientContext *context,
                const struct ucred *ucred,
                const struct timeval *tv,
                const char *label, size_t label_len) {

        /* Process a single entry from a native message. Returns 0 if nothing special happened and the message
         * processing should continue, and a negative or positive value otherwise.
         *
         * Note that *remaining is altered on both success and failure. */

        size_t n = 0, j, tn = SIZE_MAX, entry_size = 0;
        char *identifier = NULL, *message = NULL;
        struct iovec *iovec = NULL;
        int priority = LOG_INFO;
        pid_t object_pid = 0;
        const char *p;
        int r = 1;

        p = buffer;

        while (*remaining > 0) {
                const char *e, *q;

                e = memchr(p, '\n', *remaining);

                if (!e) {
                        /* Trailing noise, let's ignore it, and flush what we collected */
                        log_debug("Received message with trailing noise, ignoring.");
                        break; /* finish processing of the message */
                }

                if (e == p) {
                        /* Entry separator */
                        *remaining -= 1;
                        break;
                }

                if (IN_SET(*p, '.', '#')) {
                        /* Ignore control commands for now, and comments too. */
                        *remaining -= (e - p) + 1;
                        p = e + 1;
                        continue;
                }

                /* A property follows */
                if (n > ENTRY_FIELD_COUNT_MAX) {
                        log_debug("Received an entry that has more than " STRINGIFY(ENTRY_FIELD_COUNT_MAX) " fields, ignoring entry.");
                        goto finish;
                }

                /* n existing properties, 1 new, +1 for _TRANSPORT */
                if (!GREEDY_REALLOC(iovec,
                                    n + 2 +
                                    N_IOVEC_META_FIELDS + N_IOVEC_OBJECT_FIELDS +
                                    client_context_extra_fields_n_iovec(context))) {
                        r = log_oom();
                        goto finish;
                }

                q = memchr(p, '=', e - p);
                if (q) {
                        if (journal_field_valid(p, q - p, false)) {
                                size_t l;

                                l = e - p;
                                if (l > DATA_SIZE_MAX) {
                                        log_debug("Received text block of %zu bytes is too large, ignoring entry.", l);
                                        goto finish;
                                }

                                if (entry_size + l + n + 1 > ENTRY_SIZE_MAX) { /* data + separators + trailer */
                                        log_debug("Entry is too big (%zu bytes after processing %zu entries), ignoring entry.",
                                                  entry_size + l, n + 1);
                                        goto finish;
                                }

                                /* If the field name starts with an underscore, skip the variable, since that indicates
                                 * a trusted field */
                                iovec[n++] = IOVEC_MAKE((char*) p, l);
                                entry_size += l;

                                server_process_entry_meta(p, l, ucred,
                                                          &priority,
                                                          &identifier,
                                                          &message,
                                                          &object_pid);
                        }

                        *remaining -= (e - p) + 1;
                        p = e + 1;
                        continue;
                } else {
                        uint64_t l, total;
                        char *k;

                        if (*remaining < e - p + 1 + sizeof(uint64_t) + 1) {
                                log_debug("Failed to parse message, ignoring.");
                                break;
                        }

                        l = unaligned_read_le64(e + 1);
                        if (l > DATA_SIZE_MAX) {
                                log_debug("Received binary data block of %"PRIu64" bytes is too large, ignoring entry.", l);
                                goto finish;
                        }

                        total = (e - p) + 1 + l;
                        if (entry_size + total + n + 1 > ENTRY_SIZE_MAX) { /* data + separators + trailer */
                                log_debug("Entry is too big (%"PRIu64"bytes after processing %zu fields), ignoring.",
                                          entry_size + total, n + 1);
                                goto finish;
                        }

                        if ((uint64_t) *remaining < e - p + 1 + sizeof(uint64_t) + l + 1 ||
                            e[1+sizeof(uint64_t)+l] != '\n') {
                                log_debug("Failed to parse message, ignoring.");
                                break;
                        }

                        k = malloc(total);
                        if (!k) {
                                log_oom();
                                break;
                        }

                        memcpy(k, p, e - p);
                        k[e - p] = '=';
                        memcpy(k + (e - p) + 1, e + 1 + sizeof(uint64_t), l);

                        if (journal_field_valid(p, e - p, false)) {
                                iovec[n] = IOVEC_MAKE(k, (e - p) + 1 + l);
                                entry_size += iovec[n].iov_len;
                                n++;

                                server_process_entry_meta(k, (e - p) + 1 + l, ucred,
                                                          &priority,
                                                          &identifier,
                                                          &message,
                                                          &object_pid);
                        } else
                                free(k);

                        *remaining -= (e - p) + 1 + sizeof(uint64_t) + l + 1;
                        p = e + 1 + sizeof(uint64_t) + l + 1;
                }
        }

        if (n <= 0)
                goto finish;

        tn = n++;
        iovec[tn] = IOVEC_MAKE_STRING("_TRANSPORT=journal");
        entry_size += STRLEN("_TRANSPORT=journal");

        if (entry_size + n + 1 > ENTRY_SIZE_MAX) { /* data + separators + trailer */
                log_debug("Entry is too big with %zu properties and %zu bytes, ignoring.", n, entry_size);
                goto finish;
        }

        r = 0; /* Success, we read the message. */

        if (!client_context_test_priority(context, priority))
                goto finish;

        if (message) {
                /* Ensure message is not NULL, otherwise strlen(message) would crash. This check needs to
                 * be here until server_process_entry() is able to process messages containing \0 characters,
                 * as we would have access to the actual size of message. */
                r = client_context_check_keep_log(context, message, strlen(message));
                if (r <= 0)
                        goto finish;

                if (s->forward_to_syslog)
                        server_forward_syslog(s, syslog_fixup_facility(priority), identifier, message, ucred, tv);

                if (s->forward_to_kmsg)
                        server_forward_kmsg(s, priority, identifier, message, ucred);

                if (s->forward_to_console)
                        server_forward_console(s, priority, identifier, message, ucred);

                if (s->forward_to_wall)
                        server_forward_wall(s, priority, identifier, message, ucred);
        }

        server_dispatch_message(s, iovec, n, MALLOC_ELEMENTSOF(iovec), context, tv, priority, object_pid);

finish:
        for (j = 0; j < n; j++)  {
                if (j == tn)
                        continue;

                if (iovec[j].iov_base < buffer ||
                    (const char*) iovec[j].iov_base >= p + *remaining)
                        free(iovec[j].iov_base);
        }

        free(iovec);
        free(identifier);
        free(message);

        return r;
}

void server_process_native_message(
                Server *s,
                const char *buffer, size_t buffer_size,
                const struct ucred *ucred,
                const struct timeval *tv,
                const char *label, size_t label_len) {

        size_t remaining = buffer_size;
        ClientContext *context = NULL;
        int r;

        assert(s);
        assert(buffer || buffer_size == 0);

        if (ucred && pid_is_valid(ucred->pid)) {
                r = client_context_get(s, ucred->pid, ucred, label, label_len, NULL, &context);
                if (r < 0)
                        log_ratelimit_warning_errno(r, JOURNAL_LOG_RATELIMIT,
                                                    "Failed to retrieve credentials for PID " PID_FMT ", ignoring: %m",
                                                    ucred->pid);
        }

        do {
                r = server_process_entry(s,
                                         (const uint8_t*) buffer + (buffer_size - remaining), &remaining,
                                         context, ucred, tv, label, label_len);
        } while (r == 0);
}

void server_process_native_file(
                Server *s,
                int fd,
                const struct ucred *ucred,
                const struct timeval *tv,
                const char *label, size_t label_len) {

        struct stat st;
        bool sealed;
        int r;

        /* Data is in the passed fd, probably it didn't fit in a datagram. */

        assert(s);
        assert(fd >= 0);

        if (fstat(fd, &st) < 0) {
                log_ratelimit_error_errno(errno, JOURNAL_LOG_RATELIMIT, "Failed to stat passed file, ignoring: %m");
                return;
        }

        r = stat_verify_regular(&st);
        if (r < 0) {
                log_ratelimit_error_errno(r, JOURNAL_LOG_RATELIMIT, "File passed is not regular, ignoring: %m");
                return;
        }

        if (st.st_size <= 0)
                return;

        int flags = fcntl(fd, F_GETFL);
        if (flags < 0) {
                log_ratelimit_error_errno(errno, JOURNAL_LOG_RATELIMIT, "Failed to get flags of passed file, ignoring: %m");
                return;
        }

        if ((flags & ~(O_ACCMODE|RAW_O_LARGEFILE)) != 0) {
                log_ratelimit_error(JOURNAL_LOG_RATELIMIT, "Unexpected flags of passed memory fd, ignoring message: %m");
                return;
        }

        /* If it's a memfd, check if it is sealed. If so, we can just mmap it and use it, and do not need to
         * copy the data out. */
        sealed = memfd_get_sealed(fd) > 0;

        if (!sealed && (!ucred || ucred->uid != 0)) {
                _cleanup_free_ char *k = NULL;
                const char *e;

                /* If this is not a sealed memfd, and the peer is unknown or
                 * unprivileged, then verify the path. */

                r = fd_get_path(fd, &k);
                if (r < 0) {
                        log_ratelimit_error_errno(r, JOURNAL_LOG_RATELIMIT,
                                                  "readlink(/proc/self/fd/%i) failed: %m", fd);
                        return;
                }

                e = PATH_STARTSWITH_SET(k, "/dev/shm/", "/tmp/", "/var/tmp/");
                if (!e) {
                        log_ratelimit_error(JOURNAL_LOG_RATELIMIT,
                                            "Received file outside of allowed directories. Refusing.");
                        return;
                }

                if (!filename_is_valid(e)) {
                        log_ratelimit_error(JOURNAL_LOG_RATELIMIT,
                                            "Received file in subdirectory of allowed directories. Refusing.");
                        return;
                }
        }

        /* When !sealed, set a lower memory limit. We have to read the file, effectively doubling memory
         * use. */
        if (st.st_size > ENTRY_SIZE_MAX / (sealed ? 1 : 2)) {
                log_ratelimit_error(JOURNAL_LOG_RATELIMIT,
                                    "File passed too large (%"PRIu64" bytes). Ignoring.",
                                    (uint64_t) st.st_size);
                return;
        }

        if (sealed) {
                void *p;
                size_t ps;

                /* The file is sealed, we can just map it and use it. */

                ps = PAGE_ALIGN(st.st_size);
                assert(ps < SIZE_MAX);
                p = mmap(NULL, ps, PROT_READ, MAP_PRIVATE, fd, 0);
                if (p == MAP_FAILED) {
                        log_ratelimit_error_errno(errno, JOURNAL_LOG_RATELIMIT,
                                                  "Failed to map memfd, ignoring: %m");
                        return;
                }

                server_process_native_message(s, p, st.st_size, ucred, tv, label, label_len);
                assert_se(munmap(p, ps) >= 0);
        } else {
                _cleanup_free_ void *p = NULL;
                struct statvfs vfs;
                ssize_t n;

                if (fstatvfs(fd, &vfs) < 0) {
                        log_ratelimit_error_errno(errno, JOURNAL_LOG_RATELIMIT,
                                                  "Failed to stat file system of passed file, not processing it: %m");
                        return;
                }

                /* Refuse operating on file systems that have
                 * mandatory locking enabled, see:
                 *
                 * https://github.com/systemd/systemd/issues/1822
                 */
                if (vfs.f_flag & ST_MANDLOCK) {
                        log_ratelimit_error(JOURNAL_LOG_RATELIMIT,
                                            "Received file descriptor from file system with mandatory locking enabled, not processing it.");
                        return;
                }

                /* Make the fd non-blocking. On regular files this has
                 * the effect of bypassing mandatory locking. Of
                 * course, this should normally not be necessary given
                 * the check above, but let's better be safe than
                 * sorry, after all NFS is pretty confusing regarding
                 * file system flags, and we better don't trust it,
                 * and so is SMB. */
                r = fd_nonblock(fd, true);
                if (r < 0) {
                        log_ratelimit_error_errno(r, JOURNAL_LOG_RATELIMIT,
                                                  "Failed to make fd non-blocking, not processing it: %m");
                        return;
                }

                /* The file is not sealed, we can't map the file here, since
                 * clients might then truncate it and trigger a SIGBUS for
                 * us. So let's stupidly read it. */

                p = malloc(st.st_size);
                if (!p) {
                        log_oom();
                        return;
                }

                n = pread(fd, p, st.st_size, 0);
                if (n < 0)
                        log_ratelimit_error_errno(errno, JOURNAL_LOG_RATELIMIT,
                                                  "Failed to read file, ignoring: %m");
                else if (n > 0)
                        server_process_native_message(s, p, n, ucred, tv, label, label_len);
        }
}

int server_open_native_socket(Server *s, const char *native_socket) {
        int r;

        assert(s);
        assert(native_socket);

        if (s->native_fd < 0) {
                union sockaddr_union sa;
                size_t sa_len;

                r = sockaddr_un_set_path(&sa.un, native_socket);
                if (r < 0)
                        return log_error_errno(r, "Unable to use namespace path %s for AF_UNIX socket: %m", native_socket);
                sa_len = r;

                s->native_fd = socket(AF_UNIX, SOCK_DGRAM|SOCK_CLOEXEC|SOCK_NONBLOCK, 0);
                if (s->native_fd < 0)
                        return log_error_errno(errno, "socket() failed: %m");

                (void) sockaddr_un_unlink(&sa.un);

                r = bind(s->native_fd, &sa.sa, sa_len);
                if (r < 0)
                        return log_error_errno(errno, "bind(%s) failed: %m", sa.un.sun_path);

                (void) chmod(sa.un.sun_path, 0666);
        } else
                (void) fd_nonblock(s->native_fd, true);

        r = setsockopt_int(s->native_fd, SOL_SOCKET, SO_PASSCRED, true);
        if (r < 0)
                return log_error_errno(r, "SO_PASSCRED failed: %m");

        if (mac_selinux_use()) {
                r = setsockopt_int(s->native_fd, SOL_SOCKET, SO_PASSSEC, true);
                if (r < 0)
                        log_warning_errno(r, "SO_PASSSEC failed: %m");
        }

        r = setsockopt_int(s->native_fd, SOL_SOCKET, SO_TIMESTAMP, true);
        if (r < 0)
                return log_error_errno(r, "SO_TIMESTAMP failed: %m");

        r = sd_event_add_io(s->event, &s->native_event_source, s->native_fd, EPOLLIN, server_process_datagram, s);
        if (r < 0)
                return log_error_errno(r, "Failed to add native server fd to event loop: %m");

        r = sd_event_source_set_priority(s->native_event_source, SD_EVENT_PRIORITY_NORMAL+5);
        if (r < 0)
                return log_error_errno(r, "Failed to adjust native event source priority: %m");

        return 0;
}