]>
Commit | Line | Data |
---|---|---|
1 | /* SPDX-License-Identifier: LGPL-2.1+ */ | |
2 | ||
3 | #if HAVE_SELINUX | |
4 | #include <selinux/selinux.h> | |
5 | #endif | |
6 | #include <sys/ioctl.h> | |
7 | #include <sys/mman.h> | |
8 | #include <sys/signalfd.h> | |
9 | #include <sys/statvfs.h> | |
10 | #include <linux/sockios.h> | |
11 | ||
12 | #include "sd-daemon.h" | |
13 | #include "sd-journal.h" | |
14 | #include "sd-messages.h" | |
15 | ||
16 | #include "acl-util.h" | |
17 | #include "alloc-util.h" | |
18 | #include "audit-util.h" | |
19 | #include "cgroup-util.h" | |
20 | #include "conf-parser.h" | |
21 | #include "dirent-util.h" | |
22 | #include "extract-word.h" | |
23 | #include "fd-util.h" | |
24 | #include "fileio.h" | |
25 | #include "format-util.h" | |
26 | #include "fs-util.h" | |
27 | #include "hashmap.h" | |
28 | #include "hostname-util.h" | |
29 | #include "id128-util.h" | |
30 | #include "io-util.h" | |
31 | #include "journal-authenticate.h" | |
32 | #include "journal-file.h" | |
33 | #include "journal-internal.h" | |
34 | #include "journal-vacuum.h" | |
35 | #include "journald-audit.h" | |
36 | #include "journald-context.h" | |
37 | #include "journald-kmsg.h" | |
38 | #include "journald-native.h" | |
39 | #include "journald-rate-limit.h" | |
40 | #include "journald-server.h" | |
41 | #include "journald-stream.h" | |
42 | #include "journald-syslog.h" | |
43 | #include "log.h" | |
44 | #include "missing.h" | |
45 | #include "mkdir.h" | |
46 | #include "parse-util.h" | |
47 | #include "proc-cmdline.h" | |
48 | #include "process-util.h" | |
49 | #include "rm-rf.h" | |
50 | #include "selinux-util.h" | |
51 | #include "signal-util.h" | |
52 | #include "socket-util.h" | |
53 | #include "stdio-util.h" | |
54 | #include "string-table.h" | |
55 | #include "string-util.h" | |
56 | #include "syslog-util.h" | |
57 | #include "user-util.h" | |
58 | ||
59 | #define USER_JOURNALS_MAX 1024 | |
60 | ||
61 | #define DEFAULT_SYNC_INTERVAL_USEC (5*USEC_PER_MINUTE) | |
62 | #define DEFAULT_RATE_LIMIT_INTERVAL (30*USEC_PER_SEC) | |
63 | #define DEFAULT_RATE_LIMIT_BURST 10000 | |
64 | #define DEFAULT_MAX_FILE_USEC USEC_PER_MONTH | |
65 | ||
66 | #define RECHECK_SPACE_USEC (30*USEC_PER_SEC) | |
67 | ||
68 | #define NOTIFY_SNDBUF_SIZE (8*1024*1024) | |
69 | ||
70 | /* The period to insert between posting changes for coalescing */ | |
71 | #define POST_CHANGE_TIMER_INTERVAL_USEC (250*USEC_PER_MSEC) | |
72 | ||
73 | /* Pick a good default that is likely to fit into AF_UNIX and AF_INET SOCK_DGRAM datagrams, and even leaves some room | |
74 | * for a bit of additional metadata. */ | |
75 | #define DEFAULT_LINE_MAX (48*1024) | |
76 | ||
77 | #define DEFERRED_CLOSES_MAX (4096) | |
78 | ||
79 | static int determine_path_usage(Server *s, const char *path, uint64_t *ret_used, uint64_t *ret_free) { | |
80 | _cleanup_closedir_ DIR *d = NULL; | |
81 | struct dirent *de; | |
82 | struct statvfs ss; | |
83 | ||
84 | assert(ret_used); | |
85 | assert(ret_free); | |
86 | ||
87 | d = opendir(path); | |
88 | if (!d) | |
89 | return log_full_errno(errno == ENOENT ? LOG_DEBUG : LOG_ERR, | |
90 | errno, "Failed to open %s: %m", path); | |
91 | ||
92 | if (fstatvfs(dirfd(d), &ss) < 0) | |
93 | return log_error_errno(errno, "Failed to fstatvfs(%s): %m", path); | |
94 | ||
95 | *ret_free = ss.f_bsize * ss.f_bavail; | |
96 | *ret_used = 0; | |
97 | FOREACH_DIRENT_ALL(de, d, break) { | |
98 | struct stat st; | |
99 | ||
100 | if (!endswith(de->d_name, ".journal") && | |
101 | !endswith(de->d_name, ".journal~")) | |
102 | continue; | |
103 | ||
104 | if (fstatat(dirfd(d), de->d_name, &st, AT_SYMLINK_NOFOLLOW) < 0) { | |
105 | log_debug_errno(errno, "Failed to stat %s/%s, ignoring: %m", path, de->d_name); | |
106 | continue; | |
107 | } | |
108 | ||
109 | if (!S_ISREG(st.st_mode)) | |
110 | continue; | |
111 | ||
112 | *ret_used += (uint64_t) st.st_blocks * 512UL; | |
113 | } | |
114 | ||
115 | return 0; | |
116 | } | |
117 | ||
118 | static void cache_space_invalidate(JournalStorageSpace *space) { | |
119 | zero(*space); | |
120 | } | |
121 | ||
122 | static int cache_space_refresh(Server *s, JournalStorage *storage) { | |
123 | JournalStorageSpace *space; | |
124 | JournalMetrics *metrics; | |
125 | uint64_t vfs_used, vfs_avail, avail; | |
126 | usec_t ts; | |
127 | int r; | |
128 | ||
129 | assert(s); | |
130 | ||
131 | metrics = &storage->metrics; | |
132 | space = &storage->space; | |
133 | ||
134 | ts = now(CLOCK_MONOTONIC); | |
135 | ||
136 | if (space->timestamp != 0 && space->timestamp + RECHECK_SPACE_USEC > ts) | |
137 | return 0; | |
138 | ||
139 | r = determine_path_usage(s, storage->path, &vfs_used, &vfs_avail); | |
140 | if (r < 0) | |
141 | return r; | |
142 | ||
143 | space->vfs_used = vfs_used; | |
144 | space->vfs_available = vfs_avail; | |
145 | ||
146 | avail = LESS_BY(vfs_avail, metrics->keep_free); | |
147 | ||
148 | space->limit = MIN(MAX(vfs_used + avail, metrics->min_use), metrics->max_use); | |
149 | space->available = LESS_BY(space->limit, vfs_used); | |
150 | space->timestamp = ts; | |
151 | return 1; | |
152 | } | |
153 | ||
154 | static void patch_min_use(JournalStorage *storage) { | |
155 | assert(storage); | |
156 | ||
157 | /* Let's bump the min_use limit to the current usage on disk. We do | |
158 | * this when starting up and first opening the journal files. This way | |
159 | * sudden spikes in disk usage will not cause journald to vacuum files | |
160 | * without bounds. Note that this means that only a restart of journald | |
161 | * will make it reset this value. */ | |
162 | ||
163 | storage->metrics.min_use = MAX(storage->metrics.min_use, storage->space.vfs_used); | |
164 | } | |
165 | ||
166 | static int determine_space(Server *s, uint64_t *available, uint64_t *limit) { | |
167 | JournalStorage *js; | |
168 | int r; | |
169 | ||
170 | assert(s); | |
171 | ||
172 | js = s->system_journal ? &s->system_storage : &s->runtime_storage; | |
173 | ||
174 | r = cache_space_refresh(s, js); | |
175 | if (r >= 0) { | |
176 | if (available) | |
177 | *available = js->space.available; | |
178 | if (limit) | |
179 | *limit = js->space.limit; | |
180 | } | |
181 | return r; | |
182 | } | |
183 | ||
184 | void server_space_usage_message(Server *s, JournalStorage *storage) { | |
185 | char fb1[FORMAT_BYTES_MAX], fb2[FORMAT_BYTES_MAX], fb3[FORMAT_BYTES_MAX], | |
186 | fb4[FORMAT_BYTES_MAX], fb5[FORMAT_BYTES_MAX], fb6[FORMAT_BYTES_MAX]; | |
187 | JournalMetrics *metrics; | |
188 | ||
189 | assert(s); | |
190 | ||
191 | if (!storage) | |
192 | storage = s->system_journal ? &s->system_storage : &s->runtime_storage; | |
193 | ||
194 | if (cache_space_refresh(s, storage) < 0) | |
195 | return; | |
196 | ||
197 | metrics = &storage->metrics; | |
198 | format_bytes(fb1, sizeof(fb1), storage->space.vfs_used); | |
199 | format_bytes(fb2, sizeof(fb2), metrics->max_use); | |
200 | format_bytes(fb3, sizeof(fb3), metrics->keep_free); | |
201 | format_bytes(fb4, sizeof(fb4), storage->space.vfs_available); | |
202 | format_bytes(fb5, sizeof(fb5), storage->space.limit); | |
203 | format_bytes(fb6, sizeof(fb6), storage->space.available); | |
204 | ||
205 | server_driver_message(s, 0, | |
206 | "MESSAGE_ID=" SD_MESSAGE_JOURNAL_USAGE_STR, | |
207 | LOG_MESSAGE("%s (%s) is %s, max %s, %s free.", | |
208 | storage->name, storage->path, fb1, fb5, fb6), | |
209 | "JOURNAL_NAME=%s", storage->name, | |
210 | "JOURNAL_PATH=%s", storage->path, | |
211 | "CURRENT_USE=%"PRIu64, storage->space.vfs_used, | |
212 | "CURRENT_USE_PRETTY=%s", fb1, | |
213 | "MAX_USE=%"PRIu64, metrics->max_use, | |
214 | "MAX_USE_PRETTY=%s", fb2, | |
215 | "DISK_KEEP_FREE=%"PRIu64, metrics->keep_free, | |
216 | "DISK_KEEP_FREE_PRETTY=%s", fb3, | |
217 | "DISK_AVAILABLE=%"PRIu64, storage->space.vfs_available, | |
218 | "DISK_AVAILABLE_PRETTY=%s", fb4, | |
219 | "LIMIT=%"PRIu64, storage->space.limit, | |
220 | "LIMIT_PRETTY=%s", fb5, | |
221 | "AVAILABLE=%"PRIu64, storage->space.available, | |
222 | "AVAILABLE_PRETTY=%s", fb6, | |
223 | NULL); | |
224 | } | |
225 | ||
226 | static bool uid_for_system_journal(uid_t uid) { | |
227 | ||
228 | /* Returns true if the specified UID shall get its data stored in the system journal*/ | |
229 | ||
230 | return uid_is_system(uid) || uid_is_dynamic(uid) || uid == UID_NOBODY; | |
231 | } | |
232 | ||
233 | static void server_add_acls(JournalFile *f, uid_t uid) { | |
234 | #if HAVE_ACL | |
235 | int r; | |
236 | #endif | |
237 | assert(f); | |
238 | ||
239 | #if HAVE_ACL | |
240 | if (uid_for_system_journal(uid)) | |
241 | return; | |
242 | ||
243 | r = add_acls_for_user(f->fd, uid); | |
244 | if (r < 0) | |
245 | log_warning_errno(r, "Failed to set ACL on %s, ignoring: %m", f->path); | |
246 | #endif | |
247 | } | |
248 | ||
249 | static int open_journal( | |
250 | Server *s, | |
251 | bool reliably, | |
252 | const char *fname, | |
253 | int flags, | |
254 | bool seal, | |
255 | JournalMetrics *metrics, | |
256 | JournalFile **ret) { | |
257 | ||
258 | _cleanup_(journal_file_closep) JournalFile *f = NULL; | |
259 | int r; | |
260 | ||
261 | assert(s); | |
262 | assert(fname); | |
263 | assert(ret); | |
264 | ||
265 | if (reliably) | |
266 | r = journal_file_open_reliably(fname, flags, 0640, s->compress.enabled, s->compress.threshold_bytes, | |
267 | seal, metrics, s->mmap, s->deferred_closes, NULL, &f); | |
268 | else | |
269 | r = journal_file_open(-1, fname, flags, 0640, s->compress.enabled, s->compress.threshold_bytes, seal, | |
270 | metrics, s->mmap, s->deferred_closes, NULL, &f); | |
271 | ||
272 | if (r < 0) | |
273 | return r; | |
274 | ||
275 | r = journal_file_enable_post_change_timer(f, s->event, POST_CHANGE_TIMER_INTERVAL_USEC); | |
276 | if (r < 0) | |
277 | return r; | |
278 | ||
279 | *ret = TAKE_PTR(f); | |
280 | return r; | |
281 | } | |
282 | ||
283 | static bool flushed_flag_is_set(void) { | |
284 | return access("/run/systemd/journal/flushed", F_OK) >= 0; | |
285 | } | |
286 | ||
287 | static int system_journal_open(Server *s, bool flush_requested, bool relinquish_requested) { | |
288 | const char *fn; | |
289 | int r = 0; | |
290 | ||
291 | if (!s->system_journal && | |
292 | IN_SET(s->storage, STORAGE_PERSISTENT, STORAGE_AUTO) && | |
293 | (flush_requested || flushed_flag_is_set()) && | |
294 | !relinquish_requested) { | |
295 | ||
296 | /* If in auto mode: first try to create the machine | |
297 | * path, but not the prefix. | |
298 | * | |
299 | * If in persistent mode: create /var/log/journal and | |
300 | * the machine path */ | |
301 | ||
302 | if (s->storage == STORAGE_PERSISTENT) | |
303 | (void) mkdir_p("/var/log/journal/", 0755); | |
304 | ||
305 | (void) mkdir(s->system_storage.path, 0755); | |
306 | ||
307 | fn = strjoina(s->system_storage.path, "/system.journal"); | |
308 | r = open_journal(s, true, fn, O_RDWR|O_CREAT, s->seal, &s->system_storage.metrics, &s->system_journal); | |
309 | if (r >= 0) { | |
310 | server_add_acls(s->system_journal, 0); | |
311 | (void) cache_space_refresh(s, &s->system_storage); | |
312 | patch_min_use(&s->system_storage); | |
313 | } else { | |
314 | if (!IN_SET(r, -ENOENT, -EROFS)) | |
315 | log_warning_errno(r, "Failed to open system journal: %m"); | |
316 | ||
317 | r = 0; | |
318 | } | |
319 | ||
320 | /* If the runtime journal is open, and we're post-flush, we're | |
321 | * recovering from a failed system journal rotate (ENOSPC) | |
322 | * for which the runtime journal was reopened. | |
323 | * | |
324 | * Perform an implicit flush to var, leaving the runtime | |
325 | * journal closed, now that the system journal is back. | |
326 | */ | |
327 | if (!flush_requested) | |
328 | (void) server_flush_to_var(s, true); | |
329 | } | |
330 | ||
331 | if (!s->runtime_journal && | |
332 | (s->storage != STORAGE_NONE)) { | |
333 | ||
334 | fn = strjoina(s->runtime_storage.path, "/system.journal"); | |
335 | ||
336 | if (s->system_journal && !relinquish_requested) { | |
337 | ||
338 | /* Try to open the runtime journal, but only | |
339 | * if it already exists, so that we can flush | |
340 | * it into the system journal */ | |
341 | ||
342 | r = open_journal(s, false, fn, O_RDWR, false, &s->runtime_storage.metrics, &s->runtime_journal); | |
343 | if (r < 0) { | |
344 | if (r != -ENOENT) | |
345 | log_warning_errno(r, "Failed to open runtime journal: %m"); | |
346 | ||
347 | r = 0; | |
348 | } | |
349 | ||
350 | } else { | |
351 | ||
352 | /* OK, we really need the runtime journal, so create | |
353 | * it if necessary. */ | |
354 | ||
355 | (void) mkdir("/run/log", 0755); | |
356 | (void) mkdir("/run/log/journal", 0755); | |
357 | (void) mkdir_parents(fn, 0750); | |
358 | ||
359 | r = open_journal(s, true, fn, O_RDWR|O_CREAT, false, &s->runtime_storage.metrics, &s->runtime_journal); | |
360 | if (r < 0) | |
361 | return log_error_errno(r, "Failed to open runtime journal: %m"); | |
362 | } | |
363 | ||
364 | if (s->runtime_journal) { | |
365 | server_add_acls(s->runtime_journal, 0); | |
366 | (void) cache_space_refresh(s, &s->runtime_storage); | |
367 | patch_min_use(&s->runtime_storage); | |
368 | } | |
369 | } | |
370 | ||
371 | return r; | |
372 | } | |
373 | ||
374 | static JournalFile* find_journal(Server *s, uid_t uid) { | |
375 | _cleanup_free_ char *p = NULL; | |
376 | int r; | |
377 | JournalFile *f; | |
378 | sd_id128_t machine; | |
379 | ||
380 | assert(s); | |
381 | ||
382 | /* A rotate that fails to create the new journal (ENOSPC) leaves the | |
383 | * rotated journal as NULL. Unless we revisit opening, even after | |
384 | * space is made available we'll continue to return NULL indefinitely. | |
385 | * | |
386 | * system_journal_open() is a noop if the journals are already open, so | |
387 | * we can just call it here to recover from failed rotates (or anything | |
388 | * else that's left the journals as NULL). | |
389 | * | |
390 | * Fixes https://github.com/systemd/systemd/issues/3968 */ | |
391 | (void) system_journal_open(s, false, false); | |
392 | ||
393 | /* We split up user logs only on /var, not on /run. If the | |
394 | * runtime file is open, we write to it exclusively, in order | |
395 | * to guarantee proper order as soon as we flush /run to | |
396 | * /var and close the runtime file. */ | |
397 | ||
398 | if (s->runtime_journal) | |
399 | return s->runtime_journal; | |
400 | ||
401 | if (uid_for_system_journal(uid)) | |
402 | return s->system_journal; | |
403 | ||
404 | f = ordered_hashmap_get(s->user_journals, UID_TO_PTR(uid)); | |
405 | if (f) | |
406 | return f; | |
407 | ||
408 | r = sd_id128_get_machine(&machine); | |
409 | if (r < 0) { | |
410 | log_debug_errno(r, "Failed to determine machine ID, using system log: %m"); | |
411 | return s->system_journal; | |
412 | } | |
413 | ||
414 | if (asprintf(&p, "/var/log/journal/" SD_ID128_FORMAT_STR "/user-"UID_FMT".journal", | |
415 | SD_ID128_FORMAT_VAL(machine), uid) < 0) { | |
416 | log_oom(); | |
417 | return s->system_journal; | |
418 | } | |
419 | ||
420 | while (ordered_hashmap_size(s->user_journals) >= USER_JOURNALS_MAX) { | |
421 | /* Too many open? Then let's close one */ | |
422 | f = ordered_hashmap_steal_first(s->user_journals); | |
423 | assert(f); | |
424 | (void) journal_file_close(f); | |
425 | } | |
426 | ||
427 | r = open_journal(s, true, p, O_RDWR|O_CREAT, s->seal, &s->system_storage.metrics, &f); | |
428 | if (r < 0) | |
429 | return s->system_journal; | |
430 | ||
431 | server_add_acls(f, uid); | |
432 | ||
433 | r = ordered_hashmap_put(s->user_journals, UID_TO_PTR(uid), f); | |
434 | if (r < 0) { | |
435 | (void) journal_file_close(f); | |
436 | return s->system_journal; | |
437 | } | |
438 | ||
439 | return f; | |
440 | } | |
441 | ||
442 | static int do_rotate( | |
443 | Server *s, | |
444 | JournalFile **f, | |
445 | const char* name, | |
446 | bool seal, | |
447 | uint32_t uid) { | |
448 | ||
449 | int r; | |
450 | assert(s); | |
451 | ||
452 | if (!*f) | |
453 | return -EINVAL; | |
454 | ||
455 | r = journal_file_rotate(f, s->compress.enabled, s->compress.threshold_bytes, seal, s->deferred_closes); | |
456 | if (r < 0) { | |
457 | if (*f) | |
458 | return log_error_errno(r, "Failed to rotate %s: %m", (*f)->path); | |
459 | else | |
460 | return log_error_errno(r, "Failed to create new %s journal: %m", name); | |
461 | } | |
462 | ||
463 | server_add_acls(*f, uid); | |
464 | ||
465 | return r; | |
466 | } | |
467 | ||
468 | static void server_process_deferred_closes(Server *s) { | |
469 | JournalFile *f; | |
470 | Iterator i; | |
471 | ||
472 | /* Perform any deferred closes which aren't still offlining. */ | |
473 | SET_FOREACH(f, s->deferred_closes, i) { | |
474 | if (journal_file_is_offlining(f)) | |
475 | continue; | |
476 | ||
477 | (void) set_remove(s->deferred_closes, f); | |
478 | (void) journal_file_close(f); | |
479 | } | |
480 | } | |
481 | ||
482 | static void server_vacuum_deferred_closes(Server *s) { | |
483 | assert(s); | |
484 | ||
485 | /* Make some room in the deferred closes list, so that it doesn't grow without bounds */ | |
486 | if (set_size(s->deferred_closes) < DEFERRED_CLOSES_MAX) | |
487 | return; | |
488 | ||
489 | /* Let's first remove all journal files that might already have completed closing */ | |
490 | server_process_deferred_closes(s); | |
491 | ||
492 | /* And now, let's close some more until we reach the limit again. */ | |
493 | while (set_size(s->deferred_closes) >= DEFERRED_CLOSES_MAX) { | |
494 | JournalFile *f; | |
495 | ||
496 | assert_se(f = set_steal_first(s->deferred_closes)); | |
497 | journal_file_close(f); | |
498 | } | |
499 | } | |
500 | ||
501 | static int open_user_journal_directory(Server *s, DIR **ret_dir, char **ret_path) { | |
502 | _cleanup_closedir_ DIR *dir = NULL; | |
503 | _cleanup_free_ char *path = NULL; | |
504 | sd_id128_t machine; | |
505 | int r; | |
506 | ||
507 | assert(s); | |
508 | ||
509 | r = sd_id128_get_machine(&machine); | |
510 | if (r < 0) | |
511 | return log_error_errno(r, "Failed to determine machine ID, ignoring: %m"); | |
512 | ||
513 | if (asprintf(&path, "/var/log/journal/" SD_ID128_FORMAT_STR "/", SD_ID128_FORMAT_VAL(machine)) < 0) | |
514 | return log_oom(); | |
515 | ||
516 | dir = opendir(path); | |
517 | if (!dir) | |
518 | return log_error_errno(errno, "Failed to open user journal directory '%s': %m", path); | |
519 | ||
520 | if (ret_dir) | |
521 | *ret_dir = TAKE_PTR(dir); | |
522 | if (ret_path) | |
523 | *ret_path = TAKE_PTR(path); | |
524 | ||
525 | return 0; | |
526 | } | |
527 | ||
528 | void server_rotate(Server *s) { | |
529 | _cleanup_free_ char *path = NULL; | |
530 | _cleanup_closedir_ DIR *d = NULL; | |
531 | JournalFile *f; | |
532 | Iterator i; | |
533 | void *k; | |
534 | int r; | |
535 | ||
536 | log_debug("Rotating..."); | |
537 | ||
538 | /* First, rotate the system journal (either in its runtime flavour or in its runtime flavour) */ | |
539 | (void) do_rotate(s, &s->runtime_journal, "runtime", false, 0); | |
540 | (void) do_rotate(s, &s->system_journal, "system", s->seal, 0); | |
541 | ||
542 | /* Then, rotate all user journals we have open (keeping them open) */ | |
543 | ORDERED_HASHMAP_FOREACH_KEY(f, k, s->user_journals, i) { | |
544 | r = do_rotate(s, &f, "user", s->seal, PTR_TO_UID(k)); | |
545 | if (r >= 0) | |
546 | ordered_hashmap_replace(s->user_journals, k, f); | |
547 | else if (!f) | |
548 | /* Old file has been closed and deallocated */ | |
549 | ordered_hashmap_remove(s->user_journals, k); | |
550 | } | |
551 | ||
552 | /* Finally, also rotate all user journals we currently do not have open. (But do so only if we actually have | |
553 | * access to /var, i.e. are not in the log-to-runtime-journal mode). */ | |
554 | if (!s->runtime_journal && | |
555 | open_user_journal_directory(s, &d, &path) >= 0) { | |
556 | ||
557 | struct dirent *de; | |
558 | ||
559 | FOREACH_DIRENT(de, d, log_warning_errno(errno, "Failed to enumerate %s, ignoring: %m", path)) { | |
560 | _cleanup_free_ char *u = NULL, *full = NULL; | |
561 | _cleanup_close_ int fd = -1; | |
562 | const char *a, *b; | |
563 | uid_t uid; | |
564 | ||
565 | a = startswith(de->d_name, "user-"); | |
566 | if (!a) | |
567 | continue; | |
568 | b = endswith(de->d_name, ".journal"); | |
569 | if (!b) | |
570 | continue; | |
571 | ||
572 | u = strndup(a, b-a); | |
573 | if (!u) { | |
574 | log_oom(); | |
575 | break; | |
576 | } | |
577 | ||
578 | r = parse_uid(u, &uid); | |
579 | if (r < 0) { | |
580 | log_debug_errno(r, "Failed to parse UID from file name '%s', ignoring: %m", de->d_name); | |
581 | continue; | |
582 | } | |
583 | ||
584 | /* Already rotated in the above loop? i.e. is it an open user journal? */ | |
585 | if (ordered_hashmap_contains(s->user_journals, UID_TO_PTR(uid))) | |
586 | continue; | |
587 | ||
588 | full = strjoin(path, de->d_name); | |
589 | if (!full) { | |
590 | log_oom(); | |
591 | break; | |
592 | } | |
593 | ||
594 | fd = openat(dirfd(d), de->d_name, O_RDWR|O_CLOEXEC|O_NOCTTY|O_NOFOLLOW|O_NONBLOCK); | |
595 | if (fd < 0) { | |
596 | log_full_errno(IN_SET(errno, ELOOP, ENOENT) ? LOG_DEBUG : LOG_WARNING, errno, | |
597 | "Failed to open journal file '%s' for rotation: %m", full); | |
598 | continue; | |
599 | } | |
600 | ||
601 | /* Make some room in the set of deferred close()s */ | |
602 | server_vacuum_deferred_closes(s); | |
603 | ||
604 | /* Open the file briefly, so that we can archive it */ | |
605 | r = journal_file_open(fd, | |
606 | full, | |
607 | O_RDWR, | |
608 | 0640, | |
609 | s->compress.enabled, | |
610 | s->compress.threshold_bytes, | |
611 | s->seal, | |
612 | &s->system_storage.metrics, | |
613 | s->mmap, | |
614 | s->deferred_closes, | |
615 | NULL, | |
616 | &f); | |
617 | if (r < 0) { | |
618 | log_warning_errno(r, "Failed to read journal file %s for rotation, trying to move it out of the way: %m", full); | |
619 | ||
620 | r = journal_file_dispose(dirfd(d), de->d_name); | |
621 | if (r < 0) | |
622 | log_warning_errno(r, "Failed to move %s out of the way, ignoring: %m", full); | |
623 | else | |
624 | log_debug("Successfully moved %s out of the way.", full); | |
625 | ||
626 | continue; | |
627 | } | |
628 | ||
629 | TAKE_FD(fd); /* Donated to journal_file_open() */ | |
630 | ||
631 | r = journal_file_archive(f); | |
632 | if (r < 0) | |
633 | log_debug_errno(r, "Failed to archive journal file '%s', ignoring: %m", full); | |
634 | ||
635 | f = journal_initiate_close(f, s->deferred_closes); | |
636 | } | |
637 | } | |
638 | ||
639 | server_process_deferred_closes(s); | |
640 | } | |
641 | ||
642 | void server_sync(Server *s) { | |
643 | JournalFile *f; | |
644 | Iterator i; | |
645 | int r; | |
646 | ||
647 | if (s->system_journal) { | |
648 | r = journal_file_set_offline(s->system_journal, false); | |
649 | if (r < 0) | |
650 | log_warning_errno(r, "Failed to sync system journal, ignoring: %m"); | |
651 | } | |
652 | ||
653 | ORDERED_HASHMAP_FOREACH(f, s->user_journals, i) { | |
654 | r = journal_file_set_offline(f, false); | |
655 | if (r < 0) | |
656 | log_warning_errno(r, "Failed to sync user journal, ignoring: %m"); | |
657 | } | |
658 | ||
659 | if (s->sync_event_source) { | |
660 | r = sd_event_source_set_enabled(s->sync_event_source, SD_EVENT_OFF); | |
661 | if (r < 0) | |
662 | log_error_errno(r, "Failed to disable sync timer source: %m"); | |
663 | } | |
664 | ||
665 | s->sync_scheduled = false; | |
666 | } | |
667 | ||
668 | static void do_vacuum(Server *s, JournalStorage *storage, bool verbose) { | |
669 | ||
670 | int r; | |
671 | ||
672 | assert(s); | |
673 | assert(storage); | |
674 | ||
675 | (void) cache_space_refresh(s, storage); | |
676 | ||
677 | if (verbose) | |
678 | server_space_usage_message(s, storage); | |
679 | ||
680 | r = journal_directory_vacuum(storage->path, storage->space.limit, | |
681 | storage->metrics.n_max_files, s->max_retention_usec, | |
682 | &s->oldest_file_usec, verbose); | |
683 | if (r < 0 && r != -ENOENT) | |
684 | log_warning_errno(r, "Failed to vacuum %s, ignoring: %m", storage->path); | |
685 | ||
686 | cache_space_invalidate(&storage->space); | |
687 | } | |
688 | ||
689 | int server_vacuum(Server *s, bool verbose) { | |
690 | assert(s); | |
691 | ||
692 | log_debug("Vacuuming..."); | |
693 | ||
694 | s->oldest_file_usec = 0; | |
695 | ||
696 | if (s->system_journal) | |
697 | do_vacuum(s, &s->system_storage, verbose); | |
698 | if (s->runtime_journal) | |
699 | do_vacuum(s, &s->runtime_storage, verbose); | |
700 | ||
701 | return 0; | |
702 | } | |
703 | ||
704 | static void server_cache_machine_id(Server *s) { | |
705 | sd_id128_t id; | |
706 | int r; | |
707 | ||
708 | assert(s); | |
709 | ||
710 | r = sd_id128_get_machine(&id); | |
711 | if (r < 0) | |
712 | return; | |
713 | ||
714 | sd_id128_to_string(id, stpcpy(s->machine_id_field, "_MACHINE_ID=")); | |
715 | } | |
716 | ||
717 | static void server_cache_boot_id(Server *s) { | |
718 | sd_id128_t id; | |
719 | int r; | |
720 | ||
721 | assert(s); | |
722 | ||
723 | r = sd_id128_get_boot(&id); | |
724 | if (r < 0) | |
725 | return; | |
726 | ||
727 | sd_id128_to_string(id, stpcpy(s->boot_id_field, "_BOOT_ID=")); | |
728 | } | |
729 | ||
730 | static void server_cache_hostname(Server *s) { | |
731 | _cleanup_free_ char *t = NULL; | |
732 | char *x; | |
733 | ||
734 | assert(s); | |
735 | ||
736 | t = gethostname_malloc(); | |
737 | if (!t) | |
738 | return; | |
739 | ||
740 | x = strappend("_HOSTNAME=", t); | |
741 | if (!x) | |
742 | return; | |
743 | ||
744 | free(s->hostname_field); | |
745 | s->hostname_field = x; | |
746 | } | |
747 | ||
748 | static bool shall_try_append_again(JournalFile *f, int r) { | |
749 | switch(r) { | |
750 | ||
751 | case -E2BIG: /* Hit configured limit */ | |
752 | case -EFBIG: /* Hit fs limit */ | |
753 | case -EDQUOT: /* Quota limit hit */ | |
754 | case -ENOSPC: /* Disk full */ | |
755 | log_debug("%s: Allocation limit reached, rotating.", f->path); | |
756 | return true; | |
757 | ||
758 | case -EIO: /* I/O error of some kind (mmap) */ | |
759 | log_warning("%s: IO error, rotating.", f->path); | |
760 | return true; | |
761 | ||
762 | case -EHOSTDOWN: /* Other machine */ | |
763 | log_info("%s: Journal file from other machine, rotating.", f->path); | |
764 | return true; | |
765 | ||
766 | case -EBUSY: /* Unclean shutdown */ | |
767 | log_info("%s: Unclean shutdown, rotating.", f->path); | |
768 | return true; | |
769 | ||
770 | case -EPROTONOSUPPORT: /* Unsupported feature */ | |
771 | log_info("%s: Unsupported feature, rotating.", f->path); | |
772 | return true; | |
773 | ||
774 | case -EBADMSG: /* Corrupted */ | |
775 | case -ENODATA: /* Truncated */ | |
776 | case -ESHUTDOWN: /* Already archived */ | |
777 | log_warning("%s: Journal file corrupted, rotating.", f->path); | |
778 | return true; | |
779 | ||
780 | case -EIDRM: /* Journal file has been deleted */ | |
781 | log_warning("%s: Journal file has been deleted, rotating.", f->path); | |
782 | return true; | |
783 | ||
784 | case -ETXTBSY: /* Journal file is from the future */ | |
785 | log_warning("%s: Journal file is from the future, rotating.", f->path); | |
786 | return true; | |
787 | ||
788 | case -EAFNOSUPPORT: | |
789 | log_warning("%s: underlying file system does not support memory mapping or another required file system feature.", f->path); | |
790 | return false; | |
791 | ||
792 | default: | |
793 | return false; | |
794 | } | |
795 | } | |
796 | ||
797 | static void write_to_journal(Server *s, uid_t uid, struct iovec *iovec, size_t n, int priority) { | |
798 | bool vacuumed = false, rotate = false; | |
799 | struct dual_timestamp ts; | |
800 | JournalFile *f; | |
801 | int r; | |
802 | ||
803 | assert(s); | |
804 | assert(iovec); | |
805 | assert(n > 0); | |
806 | ||
807 | /* Get the closest, linearized time we have for this log event from the event loop. (Note that we do not use | |
808 | * the source time, and not even the time the event was originally seen, but instead simply the time we started | |
809 | * processing it, as we want strictly linear ordering in what we write out.) */ | |
810 | assert_se(sd_event_now(s->event, CLOCK_REALTIME, &ts.realtime) >= 0); | |
811 | assert_se(sd_event_now(s->event, CLOCK_MONOTONIC, &ts.monotonic) >= 0); | |
812 | ||
813 | if (ts.realtime < s->last_realtime_clock) { | |
814 | /* When the time jumps backwards, let's immediately rotate. Of course, this should not happen during | |
815 | * regular operation. However, when it does happen, then we should make sure that we start fresh files | |
816 | * to ensure that the entries in the journal files are strictly ordered by time, in order to ensure | |
817 | * bisection works correctly. */ | |
818 | ||
819 | log_debug("Time jumped backwards, rotating."); | |
820 | rotate = true; | |
821 | } else { | |
822 | ||
823 | f = find_journal(s, uid); | |
824 | if (!f) | |
825 | return; | |
826 | ||
827 | if (journal_file_rotate_suggested(f, s->max_file_usec)) { | |
828 | log_debug("%s: Journal header limits reached or header out-of-date, rotating.", f->path); | |
829 | rotate = true; | |
830 | } | |
831 | } | |
832 | ||
833 | if (rotate) { | |
834 | server_rotate(s); | |
835 | server_vacuum(s, false); | |
836 | vacuumed = true; | |
837 | ||
838 | f = find_journal(s, uid); | |
839 | if (!f) | |
840 | return; | |
841 | } | |
842 | ||
843 | s->last_realtime_clock = ts.realtime; | |
844 | ||
845 | r = journal_file_append_entry(f, &ts, NULL, iovec, n, &s->seqnum, NULL, NULL); | |
846 | if (r >= 0) { | |
847 | server_schedule_sync(s, priority); | |
848 | return; | |
849 | } | |
850 | ||
851 | if (vacuumed || !shall_try_append_again(f, r)) { | |
852 | log_error_errno(r, "Failed to write entry (%zu items, %zu bytes), ignoring: %m", n, IOVEC_TOTAL_SIZE(iovec, n)); | |
853 | return; | |
854 | } | |
855 | ||
856 | server_rotate(s); | |
857 | server_vacuum(s, false); | |
858 | ||
859 | f = find_journal(s, uid); | |
860 | if (!f) | |
861 | return; | |
862 | ||
863 | log_debug("Retrying write."); | |
864 | r = journal_file_append_entry(f, &ts, NULL, iovec, n, &s->seqnum, NULL, NULL); | |
865 | if (r < 0) | |
866 | log_error_errno(r, "Failed to write entry (%zu items, %zu bytes) despite vacuuming, ignoring: %m", n, IOVEC_TOTAL_SIZE(iovec, n)); | |
867 | else | |
868 | server_schedule_sync(s, priority); | |
869 | } | |
870 | ||
871 | #define IOVEC_ADD_NUMERIC_FIELD(iovec, n, value, type, isset, format, field) \ | |
872 | if (isset(value)) { \ | |
873 | char *k; \ | |
874 | k = newa(char, STRLEN(field "=") + DECIMAL_STR_MAX(type) + 1); \ | |
875 | sprintf(k, field "=" format, value); \ | |
876 | iovec[n++] = IOVEC_MAKE_STRING(k); \ | |
877 | } | |
878 | ||
879 | #define IOVEC_ADD_STRING_FIELD(iovec, n, value, field) \ | |
880 | if (!isempty(value)) { \ | |
881 | char *k; \ | |
882 | k = strjoina(field "=", value); \ | |
883 | iovec[n++] = IOVEC_MAKE_STRING(k); \ | |
884 | } | |
885 | ||
886 | #define IOVEC_ADD_ID128_FIELD(iovec, n, value, field) \ | |
887 | if (!sd_id128_is_null(value)) { \ | |
888 | char *k; \ | |
889 | k = newa(char, STRLEN(field "=") + SD_ID128_STRING_MAX); \ | |
890 | sd_id128_to_string(value, stpcpy(k, field "=")); \ | |
891 | iovec[n++] = IOVEC_MAKE_STRING(k); \ | |
892 | } | |
893 | ||
894 | #define IOVEC_ADD_SIZED_FIELD(iovec, n, value, value_size, field) \ | |
895 | if (value_size > 0) { \ | |
896 | char *k; \ | |
897 | k = newa(char, STRLEN(field "=") + value_size + 1); \ | |
898 | *((char*) mempcpy(stpcpy(k, field "="), value, value_size)) = 0; \ | |
899 | iovec[n++] = IOVEC_MAKE_STRING(k); \ | |
900 | } \ | |
901 | ||
902 | static void dispatch_message_real( | |
903 | Server *s, | |
904 | struct iovec *iovec, size_t n, size_t m, | |
905 | const ClientContext *c, | |
906 | const struct timeval *tv, | |
907 | int priority, | |
908 | pid_t object_pid) { | |
909 | ||
910 | char source_time[sizeof("_SOURCE_REALTIME_TIMESTAMP=") + DECIMAL_STR_MAX(usec_t)]; | |
911 | _cleanup_free_ char *cmdline1 = NULL, *cmdline2 = NULL; | |
912 | uid_t journal_uid; | |
913 | ClientContext *o; | |
914 | ||
915 | assert(s); | |
916 | assert(iovec); | |
917 | assert(n > 0); | |
918 | assert(n + | |
919 | N_IOVEC_META_FIELDS + | |
920 | (pid_is_valid(object_pid) ? N_IOVEC_OBJECT_FIELDS : 0) + | |
921 | client_context_extra_fields_n_iovec(c) <= m); | |
922 | ||
923 | if (c) { | |
924 | IOVEC_ADD_NUMERIC_FIELD(iovec, n, c->pid, pid_t, pid_is_valid, PID_FMT, "_PID"); | |
925 | IOVEC_ADD_NUMERIC_FIELD(iovec, n, c->uid, uid_t, uid_is_valid, UID_FMT, "_UID"); | |
926 | IOVEC_ADD_NUMERIC_FIELD(iovec, n, c->gid, gid_t, gid_is_valid, GID_FMT, "_GID"); | |
927 | ||
928 | IOVEC_ADD_STRING_FIELD(iovec, n, c->comm, "_COMM"); /* At most TASK_COMM_LENGTH (16 bytes) */ | |
929 | IOVEC_ADD_STRING_FIELD(iovec, n, c->exe, "_EXE"); /* A path, so at most PATH_MAX (4096 bytes) */ | |
930 | ||
931 | if (c->cmdline) | |
932 | /* At most _SC_ARG_MAX (2MB usually), which is too much to put on stack. | |
933 | * Let's use a heap allocation for this one. */ | |
934 | cmdline1 = set_iovec_string_field(iovec, &n, "_CMDLINE=", c->cmdline); | |
935 | ||
936 | IOVEC_ADD_STRING_FIELD(iovec, n, c->capeff, "_CAP_EFFECTIVE"); /* Read from /proc/.../status */ | |
937 | IOVEC_ADD_SIZED_FIELD(iovec, n, c->label, c->label_size, "_SELINUX_CONTEXT"); | |
938 | IOVEC_ADD_NUMERIC_FIELD(iovec, n, c->auditid, uint32_t, audit_session_is_valid, "%" PRIu32, "_AUDIT_SESSION"); | |
939 | IOVEC_ADD_NUMERIC_FIELD(iovec, n, c->loginuid, uid_t, uid_is_valid, UID_FMT, "_AUDIT_LOGINUID"); | |
940 | ||
941 | IOVEC_ADD_STRING_FIELD(iovec, n, c->cgroup, "_SYSTEMD_CGROUP"); /* A path */ | |
942 | IOVEC_ADD_STRING_FIELD(iovec, n, c->session, "_SYSTEMD_SESSION"); | |
943 | IOVEC_ADD_NUMERIC_FIELD(iovec, n, c->owner_uid, uid_t, uid_is_valid, UID_FMT, "_SYSTEMD_OWNER_UID"); | |
944 | IOVEC_ADD_STRING_FIELD(iovec, n, c->unit, "_SYSTEMD_UNIT"); /* Unit names are bounded by UNIT_NAME_MAX */ | |
945 | IOVEC_ADD_STRING_FIELD(iovec, n, c->user_unit, "_SYSTEMD_USER_UNIT"); | |
946 | IOVEC_ADD_STRING_FIELD(iovec, n, c->slice, "_SYSTEMD_SLICE"); | |
947 | IOVEC_ADD_STRING_FIELD(iovec, n, c->user_slice, "_SYSTEMD_USER_SLICE"); | |
948 | ||
949 | IOVEC_ADD_ID128_FIELD(iovec, n, c->invocation_id, "_SYSTEMD_INVOCATION_ID"); | |
950 | ||
951 | if (c->extra_fields_n_iovec > 0) { | |
952 | memcpy(iovec + n, c->extra_fields_iovec, c->extra_fields_n_iovec * sizeof(struct iovec)); | |
953 | n += c->extra_fields_n_iovec; | |
954 | } | |
955 | } | |
956 | ||
957 | assert(n <= m); | |
958 | ||
959 | if (pid_is_valid(object_pid) && client_context_get(s, object_pid, NULL, NULL, 0, NULL, &o) >= 0) { | |
960 | ||
961 | IOVEC_ADD_NUMERIC_FIELD(iovec, n, o->pid, pid_t, pid_is_valid, PID_FMT, "OBJECT_PID"); | |
962 | IOVEC_ADD_NUMERIC_FIELD(iovec, n, o->uid, uid_t, uid_is_valid, UID_FMT, "OBJECT_UID"); | |
963 | IOVEC_ADD_NUMERIC_FIELD(iovec, n, o->gid, gid_t, gid_is_valid, GID_FMT, "OBJECT_GID"); | |
964 | ||
965 | /* See above for size limits, only ->cmdline may be large, so use a heap allocation for it. */ | |
966 | IOVEC_ADD_STRING_FIELD(iovec, n, o->comm, "OBJECT_COMM"); | |
967 | IOVEC_ADD_STRING_FIELD(iovec, n, o->exe, "OBJECT_EXE"); | |
968 | if (o->cmdline) | |
969 | cmdline2 = set_iovec_string_field(iovec, &n, "OBJECT_CMDLINE=", o->cmdline); | |
970 | ||
971 | IOVEC_ADD_STRING_FIELD(iovec, n, o->capeff, "OBJECT_CAP_EFFECTIVE"); | |
972 | IOVEC_ADD_SIZED_FIELD(iovec, n, o->label, o->label_size, "OBJECT_SELINUX_CONTEXT"); | |
973 | IOVEC_ADD_NUMERIC_FIELD(iovec, n, o->auditid, uint32_t, audit_session_is_valid, "%" PRIu32, "OBJECT_AUDIT_SESSION"); | |
974 | IOVEC_ADD_NUMERIC_FIELD(iovec, n, o->loginuid, uid_t, uid_is_valid, UID_FMT, "OBJECT_AUDIT_LOGINUID"); | |
975 | ||
976 | IOVEC_ADD_STRING_FIELD(iovec, n, o->cgroup, "OBJECT_SYSTEMD_CGROUP"); | |
977 | IOVEC_ADD_STRING_FIELD(iovec, n, o->session, "OBJECT_SYSTEMD_SESSION"); | |
978 | IOVEC_ADD_NUMERIC_FIELD(iovec, n, o->owner_uid, uid_t, uid_is_valid, UID_FMT, "OBJECT_SYSTEMD_OWNER_UID"); | |
979 | IOVEC_ADD_STRING_FIELD(iovec, n, o->unit, "OBJECT_SYSTEMD_UNIT"); | |
980 | IOVEC_ADD_STRING_FIELD(iovec, n, o->user_unit, "OBJECT_SYSTEMD_USER_UNIT"); | |
981 | IOVEC_ADD_STRING_FIELD(iovec, n, o->slice, "OBJECT_SYSTEMD_SLICE"); | |
982 | IOVEC_ADD_STRING_FIELD(iovec, n, o->user_slice, "OBJECT_SYSTEMD_USER_SLICE"); | |
983 | ||
984 | IOVEC_ADD_ID128_FIELD(iovec, n, o->invocation_id, "OBJECT_SYSTEMD_INVOCATION_ID="); | |
985 | } | |
986 | ||
987 | assert(n <= m); | |
988 | ||
989 | if (tv) { | |
990 | sprintf(source_time, "_SOURCE_REALTIME_TIMESTAMP=" USEC_FMT, timeval_load(tv)); | |
991 | iovec[n++] = IOVEC_MAKE_STRING(source_time); | |
992 | } | |
993 | ||
994 | /* Note that strictly speaking storing the boot id here is | |
995 | * redundant since the entry includes this in-line | |
996 | * anyway. However, we need this indexed, too. */ | |
997 | if (!isempty(s->boot_id_field)) | |
998 | iovec[n++] = IOVEC_MAKE_STRING(s->boot_id_field); | |
999 | ||
1000 | if (!isempty(s->machine_id_field)) | |
1001 | iovec[n++] = IOVEC_MAKE_STRING(s->machine_id_field); | |
1002 | ||
1003 | if (!isempty(s->hostname_field)) | |
1004 | iovec[n++] = IOVEC_MAKE_STRING(s->hostname_field); | |
1005 | ||
1006 | assert(n <= m); | |
1007 | ||
1008 | if (s->split_mode == SPLIT_UID && c && uid_is_valid(c->uid)) | |
1009 | /* Split up strictly by (non-root) UID */ | |
1010 | journal_uid = c->uid; | |
1011 | else if (s->split_mode == SPLIT_LOGIN && c && c->uid > 0 && uid_is_valid(c->owner_uid)) | |
1012 | /* Split up by login UIDs. We do this only if the | |
1013 | * realuid is not root, in order not to accidentally | |
1014 | * leak privileged information to the user that is | |
1015 | * logged by a privileged process that is part of an | |
1016 | * unprivileged session. */ | |
1017 | journal_uid = c->owner_uid; | |
1018 | else | |
1019 | journal_uid = 0; | |
1020 | ||
1021 | write_to_journal(s, journal_uid, iovec, n, priority); | |
1022 | } | |
1023 | ||
1024 | void server_driver_message(Server *s, pid_t object_pid, const char *message_id, const char *format, ...) { | |
1025 | ||
1026 | struct iovec *iovec; | |
1027 | size_t n = 0, k, m; | |
1028 | va_list ap; | |
1029 | int r; | |
1030 | ||
1031 | assert(s); | |
1032 | assert(format); | |
1033 | ||
1034 | m = N_IOVEC_META_FIELDS + 5 + N_IOVEC_PAYLOAD_FIELDS + client_context_extra_fields_n_iovec(s->my_context) + N_IOVEC_OBJECT_FIELDS; | |
1035 | iovec = newa(struct iovec, m); | |
1036 | ||
1037 | assert_cc(3 == LOG_FAC(LOG_DAEMON)); | |
1038 | iovec[n++] = IOVEC_MAKE_STRING("SYSLOG_FACILITY=3"); | |
1039 | iovec[n++] = IOVEC_MAKE_STRING("SYSLOG_IDENTIFIER=systemd-journald"); | |
1040 | ||
1041 | iovec[n++] = IOVEC_MAKE_STRING("_TRANSPORT=driver"); | |
1042 | assert_cc(6 == LOG_INFO); | |
1043 | iovec[n++] = IOVEC_MAKE_STRING("PRIORITY=6"); | |
1044 | ||
1045 | if (message_id) | |
1046 | iovec[n++] = IOVEC_MAKE_STRING(message_id); | |
1047 | k = n; | |
1048 | ||
1049 | va_start(ap, format); | |
1050 | r = log_format_iovec(iovec, m, &n, false, 0, format, ap); | |
1051 | /* Error handling below */ | |
1052 | va_end(ap); | |
1053 | ||
1054 | if (r >= 0) | |
1055 | dispatch_message_real(s, iovec, n, m, s->my_context, NULL, LOG_INFO, object_pid); | |
1056 | ||
1057 | while (k < n) | |
1058 | free(iovec[k++].iov_base); | |
1059 | ||
1060 | if (r < 0) { | |
1061 | /* We failed to format the message. Emit a warning instead. */ | |
1062 | char buf[LINE_MAX]; | |
1063 | ||
1064 | xsprintf(buf, "MESSAGE=Entry printing failed: %s", strerror(-r)); | |
1065 | ||
1066 | n = 3; | |
1067 | iovec[n++] = IOVEC_MAKE_STRING("PRIORITY=4"); | |
1068 | iovec[n++] = IOVEC_MAKE_STRING(buf); | |
1069 | dispatch_message_real(s, iovec, n, m, s->my_context, NULL, LOG_INFO, object_pid); | |
1070 | } | |
1071 | } | |
1072 | ||
1073 | void server_dispatch_message( | |
1074 | Server *s, | |
1075 | struct iovec *iovec, size_t n, size_t m, | |
1076 | ClientContext *c, | |
1077 | const struct timeval *tv, | |
1078 | int priority, | |
1079 | pid_t object_pid) { | |
1080 | ||
1081 | uint64_t available = 0; | |
1082 | int rl; | |
1083 | ||
1084 | assert(s); | |
1085 | assert(iovec || n == 0); | |
1086 | ||
1087 | if (n == 0) | |
1088 | return; | |
1089 | ||
1090 | if (LOG_PRI(priority) > s->max_level_store) | |
1091 | return; | |
1092 | ||
1093 | /* Stop early in case the information will not be stored | |
1094 | * in a journal. */ | |
1095 | if (s->storage == STORAGE_NONE) | |
1096 | return; | |
1097 | ||
1098 | if (c && c->unit) { | |
1099 | (void) determine_space(s, &available, NULL); | |
1100 | ||
1101 | rl = journal_rate_limit_test(s->rate_limit, c->unit, c->log_rate_limit_interval, c->log_rate_limit_burst, priority & LOG_PRIMASK, available); | |
1102 | if (rl == 0) | |
1103 | return; | |
1104 | ||
1105 | /* Write a suppression message if we suppressed something */ | |
1106 | if (rl > 1) | |
1107 | server_driver_message(s, c->pid, | |
1108 | "MESSAGE_ID=" SD_MESSAGE_JOURNAL_DROPPED_STR, | |
1109 | LOG_MESSAGE("Suppressed %i messages from %s", rl - 1, c->unit), | |
1110 | "N_DROPPED=%i", rl - 1, | |
1111 | NULL); | |
1112 | } | |
1113 | ||
1114 | dispatch_message_real(s, iovec, n, m, c, tv, priority, object_pid); | |
1115 | } | |
1116 | ||
1117 | int server_flush_to_var(Server *s, bool require_flag_file) { | |
1118 | sd_id128_t machine; | |
1119 | sd_journal *j = NULL; | |
1120 | char ts[FORMAT_TIMESPAN_MAX]; | |
1121 | usec_t start; | |
1122 | unsigned n = 0; | |
1123 | int r, k; | |
1124 | ||
1125 | assert(s); | |
1126 | ||
1127 | if (!IN_SET(s->storage, STORAGE_AUTO, STORAGE_PERSISTENT)) | |
1128 | return 0; | |
1129 | ||
1130 | if (!s->runtime_journal) | |
1131 | return 0; | |
1132 | ||
1133 | if (require_flag_file && !flushed_flag_is_set()) | |
1134 | return 0; | |
1135 | ||
1136 | (void) system_journal_open(s, true, false); | |
1137 | ||
1138 | if (!s->system_journal) | |
1139 | return 0; | |
1140 | ||
1141 | log_debug("Flushing to /var..."); | |
1142 | ||
1143 | start = now(CLOCK_MONOTONIC); | |
1144 | ||
1145 | r = sd_id128_get_machine(&machine); | |
1146 | if (r < 0) | |
1147 | return r; | |
1148 | ||
1149 | r = sd_journal_open(&j, SD_JOURNAL_RUNTIME_ONLY); | |
1150 | if (r < 0) | |
1151 | return log_error_errno(r, "Failed to read runtime journal: %m"); | |
1152 | ||
1153 | sd_journal_set_data_threshold(j, 0); | |
1154 | ||
1155 | SD_JOURNAL_FOREACH(j) { | |
1156 | Object *o = NULL; | |
1157 | JournalFile *f; | |
1158 | ||
1159 | f = j->current_file; | |
1160 | assert(f && f->current_offset > 0); | |
1161 | ||
1162 | n++; | |
1163 | ||
1164 | r = journal_file_move_to_object(f, OBJECT_ENTRY, f->current_offset, &o); | |
1165 | if (r < 0) { | |
1166 | log_error_errno(r, "Can't read entry: %m"); | |
1167 | goto finish; | |
1168 | } | |
1169 | ||
1170 | r = journal_file_copy_entry(f, s->system_journal, o, f->current_offset); | |
1171 | if (r >= 0) | |
1172 | continue; | |
1173 | ||
1174 | if (!shall_try_append_again(s->system_journal, r)) { | |
1175 | log_error_errno(r, "Can't write entry: %m"); | |
1176 | goto finish; | |
1177 | } | |
1178 | ||
1179 | server_rotate(s); | |
1180 | server_vacuum(s, false); | |
1181 | ||
1182 | if (!s->system_journal) { | |
1183 | log_notice("Didn't flush runtime journal since rotation of system journal wasn't successful."); | |
1184 | r = -EIO; | |
1185 | goto finish; | |
1186 | } | |
1187 | ||
1188 | log_debug("Retrying write."); | |
1189 | r = journal_file_copy_entry(f, s->system_journal, o, f->current_offset); | |
1190 | if (r < 0) { | |
1191 | log_error_errno(r, "Can't write entry: %m"); | |
1192 | goto finish; | |
1193 | } | |
1194 | } | |
1195 | ||
1196 | r = 0; | |
1197 | ||
1198 | finish: | |
1199 | if (s->system_journal) | |
1200 | journal_file_post_change(s->system_journal); | |
1201 | ||
1202 | s->runtime_journal = journal_file_close(s->runtime_journal); | |
1203 | ||
1204 | if (r >= 0) | |
1205 | (void) rm_rf("/run/log/journal", REMOVE_ROOT); | |
1206 | ||
1207 | sd_journal_close(j); | |
1208 | ||
1209 | server_driver_message(s, 0, NULL, | |
1210 | LOG_MESSAGE("Time spent on flushing to /var is %s for %u entries.", | |
1211 | format_timespan(ts, sizeof(ts), now(CLOCK_MONOTONIC) - start, 0), | |
1212 | n), | |
1213 | NULL); | |
1214 | ||
1215 | k = touch("/run/systemd/journal/flushed"); | |
1216 | if (k < 0) | |
1217 | log_warning_errno(k, "Failed to touch /run/systemd/journal/flushed, ignoring: %m"); | |
1218 | ||
1219 | return r; | |
1220 | } | |
1221 | ||
1222 | static int server_relinquish_var(Server *s) { | |
1223 | assert(s); | |
1224 | ||
1225 | if (s->storage == STORAGE_NONE) | |
1226 | return 0; | |
1227 | ||
1228 | if (s->runtime_journal && !s->system_journal) | |
1229 | return 0; | |
1230 | ||
1231 | log_debug("Relinquishing /var..."); | |
1232 | ||
1233 | (void) system_journal_open(s, false, true); | |
1234 | ||
1235 | s->system_journal = journal_file_close(s->system_journal); | |
1236 | ordered_hashmap_clear_with_destructor(s->user_journals, journal_file_close); | |
1237 | set_clear_with_destructor(s->deferred_closes, journal_file_close); | |
1238 | ||
1239 | if (unlink("/run/systemd/journal/flushed") < 0 && errno != ENOENT) | |
1240 | log_warning_errno(errno, "Failed to unlink /run/systemd/journal/flushed, ignoring: %m"); | |
1241 | ||
1242 | return 0; | |
1243 | } | |
1244 | ||
1245 | int server_process_datagram(sd_event_source *es, int fd, uint32_t revents, void *userdata) { | |
1246 | Server *s = userdata; | |
1247 | struct ucred *ucred = NULL; | |
1248 | struct timeval *tv = NULL; | |
1249 | struct cmsghdr *cmsg; | |
1250 | char *label = NULL; | |
1251 | size_t label_len = 0, m; | |
1252 | struct iovec iovec; | |
1253 | ssize_t n; | |
1254 | int *fds = NULL, v = 0; | |
1255 | size_t n_fds = 0; | |
1256 | ||
1257 | union { | |
1258 | struct cmsghdr cmsghdr; | |
1259 | ||
1260 | /* We use NAME_MAX space for the SELinux label | |
1261 | * here. The kernel currently enforces no | |
1262 | * limit, but according to suggestions from | |
1263 | * the SELinux people this will change and it | |
1264 | * will probably be identical to NAME_MAX. For | |
1265 | * now we use that, but this should be updated | |
1266 | * one day when the final limit is known. */ | |
1267 | uint8_t buf[CMSG_SPACE(sizeof(struct ucred)) + | |
1268 | CMSG_SPACE(sizeof(struct timeval)) + | |
1269 | CMSG_SPACE(sizeof(int)) + /* fd */ | |
1270 | CMSG_SPACE(NAME_MAX)]; /* selinux label */ | |
1271 | } control = {}; | |
1272 | ||
1273 | union sockaddr_union sa = {}; | |
1274 | ||
1275 | struct msghdr msghdr = { | |
1276 | .msg_iov = &iovec, | |
1277 | .msg_iovlen = 1, | |
1278 | .msg_control = &control, | |
1279 | .msg_controllen = sizeof(control), | |
1280 | .msg_name = &sa, | |
1281 | .msg_namelen = sizeof(sa), | |
1282 | }; | |
1283 | ||
1284 | assert(s); | |
1285 | assert(fd == s->native_fd || fd == s->syslog_fd || fd == s->audit_fd); | |
1286 | ||
1287 | if (revents != EPOLLIN) | |
1288 | return log_error_errno(SYNTHETIC_ERRNO(EIO), | |
1289 | "Got invalid event from epoll for datagram fd: %" PRIx32, | |
1290 | revents); | |
1291 | ||
1292 | /* Try to get the right size, if we can. (Not all sockets support SIOCINQ, hence we just try, but don't rely on | |
1293 | * it.) */ | |
1294 | (void) ioctl(fd, SIOCINQ, &v); | |
1295 | ||
1296 | /* Fix it up, if it is too small. We use the same fixed value as auditd here. Awful! */ | |
1297 | m = PAGE_ALIGN(MAX3((size_t) v + 1, | |
1298 | (size_t) LINE_MAX, | |
1299 | ALIGN(sizeof(struct nlmsghdr)) + ALIGN((size_t) MAX_AUDIT_MESSAGE_LENGTH)) + 1); | |
1300 | ||
1301 | if (!GREEDY_REALLOC(s->buffer, s->buffer_size, m)) | |
1302 | return log_oom(); | |
1303 | ||
1304 | iovec = IOVEC_MAKE(s->buffer, s->buffer_size - 1); /* Leave room for trailing NUL we add later */ | |
1305 | ||
1306 | n = recvmsg(fd, &msghdr, MSG_DONTWAIT|MSG_CMSG_CLOEXEC); | |
1307 | if (n < 0) { | |
1308 | if (IN_SET(errno, EINTR, EAGAIN)) | |
1309 | return 0; | |
1310 | ||
1311 | return log_error_errno(errno, "recvmsg() failed: %m"); | |
1312 | } | |
1313 | ||
1314 | CMSG_FOREACH(cmsg, &msghdr) | |
1315 | if (cmsg->cmsg_level == SOL_SOCKET && | |
1316 | cmsg->cmsg_type == SCM_CREDENTIALS && | |
1317 | cmsg->cmsg_len == CMSG_LEN(sizeof(struct ucred))) | |
1318 | ucred = (struct ucred*) CMSG_DATA(cmsg); | |
1319 | else if (cmsg->cmsg_level == SOL_SOCKET && | |
1320 | cmsg->cmsg_type == SCM_SECURITY) { | |
1321 | label = (char*) CMSG_DATA(cmsg); | |
1322 | label_len = cmsg->cmsg_len - CMSG_LEN(0); | |
1323 | } else if (cmsg->cmsg_level == SOL_SOCKET && | |
1324 | cmsg->cmsg_type == SO_TIMESTAMP && | |
1325 | cmsg->cmsg_len == CMSG_LEN(sizeof(struct timeval))) | |
1326 | tv = (struct timeval*) CMSG_DATA(cmsg); | |
1327 | else if (cmsg->cmsg_level == SOL_SOCKET && | |
1328 | cmsg->cmsg_type == SCM_RIGHTS) { | |
1329 | fds = (int*) CMSG_DATA(cmsg); | |
1330 | n_fds = (cmsg->cmsg_len - CMSG_LEN(0)) / sizeof(int); | |
1331 | } | |
1332 | ||
1333 | /* And a trailing NUL, just in case */ | |
1334 | s->buffer[n] = 0; | |
1335 | ||
1336 | if (fd == s->syslog_fd) { | |
1337 | if (n > 0 && n_fds == 0) | |
1338 | server_process_syslog_message(s, s->buffer, n, ucred, tv, label, label_len); | |
1339 | else if (n_fds > 0) | |
1340 | log_warning("Got file descriptors via syslog socket. Ignoring."); | |
1341 | ||
1342 | } else if (fd == s->native_fd) { | |
1343 | if (n > 0 && n_fds == 0) | |
1344 | server_process_native_message(s, s->buffer, n, ucred, tv, label, label_len); | |
1345 | else if (n == 0 && n_fds == 1) | |
1346 | server_process_native_file(s, fds[0], ucred, tv, label, label_len); | |
1347 | else if (n_fds > 0) | |
1348 | log_warning("Got too many file descriptors via native socket. Ignoring."); | |
1349 | ||
1350 | } else { | |
1351 | assert(fd == s->audit_fd); | |
1352 | ||
1353 | if (n > 0 && n_fds == 0) | |
1354 | server_process_audit_message(s, s->buffer, n, ucred, &sa, msghdr.msg_namelen); | |
1355 | else if (n_fds > 0) | |
1356 | log_warning("Got file descriptors via audit socket. Ignoring."); | |
1357 | } | |
1358 | ||
1359 | close_many(fds, n_fds); | |
1360 | return 0; | |
1361 | } | |
1362 | ||
1363 | static void server_full_flush(Server *s) { | |
1364 | assert(s); | |
1365 | ||
1366 | (void) server_flush_to_var(s, false); | |
1367 | server_sync(s); | |
1368 | server_vacuum(s, false); | |
1369 | ||
1370 | server_space_usage_message(s, NULL); | |
1371 | } | |
1372 | ||
1373 | static int dispatch_sigusr1(sd_event_source *es, const struct signalfd_siginfo *si, void *userdata) { | |
1374 | Server *s = userdata; | |
1375 | ||
1376 | assert(s); | |
1377 | ||
1378 | log_info("Received SIGUSR1 signal from PID " PID_FMT ", as request to flush runtime journal.", si->ssi_pid); | |
1379 | server_full_flush(s); | |
1380 | ||
1381 | return 0; | |
1382 | } | |
1383 | ||
1384 | static void server_full_rotate(Server *s) { | |
1385 | int r; | |
1386 | ||
1387 | assert(s); | |
1388 | ||
1389 | server_rotate(s); | |
1390 | server_vacuum(s, true); | |
1391 | ||
1392 | if (s->system_journal) | |
1393 | patch_min_use(&s->system_storage); | |
1394 | if (s->runtime_journal) | |
1395 | patch_min_use(&s->runtime_storage); | |
1396 | ||
1397 | /* Let clients know when the most recent rotation happened. */ | |
1398 | r = write_timestamp_file_atomic("/run/systemd/journal/rotated", now(CLOCK_MONOTONIC)); | |
1399 | if (r < 0) | |
1400 | log_warning_errno(r, "Failed to write /run/systemd/journal/rotated, ignoring: %m"); | |
1401 | } | |
1402 | ||
1403 | static int dispatch_sigusr2(sd_event_source *es, const struct signalfd_siginfo *si, void *userdata) { | |
1404 | Server *s = userdata; | |
1405 | ||
1406 | assert(s); | |
1407 | ||
1408 | log_info("Received SIGUSR2 signal from PID " PID_FMT ", as request to rotate journal.", si->ssi_pid); | |
1409 | server_full_rotate(s); | |
1410 | ||
1411 | return 0; | |
1412 | } | |
1413 | ||
1414 | static int dispatch_sigterm(sd_event_source *es, const struct signalfd_siginfo *si, void *userdata) { | |
1415 | Server *s = userdata; | |
1416 | ||
1417 | assert(s); | |
1418 | ||
1419 | log_received_signal(LOG_INFO, si); | |
1420 | ||
1421 | sd_event_exit(s->event, 0); | |
1422 | return 0; | |
1423 | } | |
1424 | ||
1425 | static void server_full_sync(Server *s) { | |
1426 | int r; | |
1427 | ||
1428 | assert(s); | |
1429 | ||
1430 | server_sync(s); | |
1431 | ||
1432 | /* Let clients know when the most recent sync happened. */ | |
1433 | r = write_timestamp_file_atomic("/run/systemd/journal/synced", now(CLOCK_MONOTONIC)); | |
1434 | if (r < 0) | |
1435 | log_warning_errno(r, "Failed to write /run/systemd/journal/synced, ignoring: %m"); | |
1436 | ||
1437 | return; | |
1438 | } | |
1439 | ||
1440 | static int dispatch_sigrtmin1(sd_event_source *es, const struct signalfd_siginfo *si, void *userdata) { | |
1441 | Server *s = userdata; | |
1442 | ||
1443 | assert(s); | |
1444 | ||
1445 | log_debug("Received SIGRTMIN1 signal from PID " PID_FMT ", as request to sync.", si->ssi_pid ); | |
1446 | server_full_sync(s); | |
1447 | ||
1448 | return 0; | |
1449 | } | |
1450 | ||
1451 | static int setup_signals(Server *s) { | |
1452 | int r; | |
1453 | ||
1454 | assert(s); | |
1455 | ||
1456 | assert_se(sigprocmask_many(SIG_SETMASK, NULL, SIGINT, SIGTERM, SIGUSR1, SIGUSR2, SIGRTMIN+1, -1) >= 0); | |
1457 | ||
1458 | r = sd_event_add_signal(s->event, &s->sigusr1_event_source, SIGUSR1, dispatch_sigusr1, s); | |
1459 | if (r < 0) | |
1460 | return r; | |
1461 | ||
1462 | r = sd_event_add_signal(s->event, &s->sigusr2_event_source, SIGUSR2, dispatch_sigusr2, s); | |
1463 | if (r < 0) | |
1464 | return r; | |
1465 | ||
1466 | r = sd_event_add_signal(s->event, &s->sigterm_event_source, SIGTERM, dispatch_sigterm, s); | |
1467 | if (r < 0) | |
1468 | return r; | |
1469 | ||
1470 | /* Let's process SIGTERM late, so that we flush all queued messages to disk before we exit */ | |
1471 | r = sd_event_source_set_priority(s->sigterm_event_source, SD_EVENT_PRIORITY_NORMAL+20); | |
1472 | if (r < 0) | |
1473 | return r; | |
1474 | ||
1475 | /* When journald is invoked on the terminal (when debugging), it's useful if C-c is handled | |
1476 | * equivalent to SIGTERM. */ | |
1477 | r = sd_event_add_signal(s->event, &s->sigint_event_source, SIGINT, dispatch_sigterm, s); | |
1478 | if (r < 0) | |
1479 | return r; | |
1480 | ||
1481 | r = sd_event_source_set_priority(s->sigint_event_source, SD_EVENT_PRIORITY_NORMAL+20); | |
1482 | if (r < 0) | |
1483 | return r; | |
1484 | ||
1485 | /* SIGRTMIN+1 causes an immediate sync. We process this very late, so that everything else queued at | |
1486 | * this point is really written to disk. Clients can watch /run/systemd/journal/synced with inotify | |
1487 | * until its mtime changes to see when a sync happened. */ | |
1488 | r = sd_event_add_signal(s->event, &s->sigrtmin1_event_source, SIGRTMIN+1, dispatch_sigrtmin1, s); | |
1489 | if (r < 0) | |
1490 | return r; | |
1491 | ||
1492 | r = sd_event_source_set_priority(s->sigrtmin1_event_source, SD_EVENT_PRIORITY_NORMAL+15); | |
1493 | if (r < 0) | |
1494 | return r; | |
1495 | ||
1496 | return 0; | |
1497 | } | |
1498 | ||
1499 | static int parse_proc_cmdline_item(const char *key, const char *value, void *data) { | |
1500 | Server *s = data; | |
1501 | int r; | |
1502 | ||
1503 | assert(s); | |
1504 | ||
1505 | if (proc_cmdline_key_streq(key, "systemd.journald.forward_to_syslog")) { | |
1506 | ||
1507 | r = value ? parse_boolean(value) : true; | |
1508 | if (r < 0) | |
1509 | log_warning("Failed to parse forward to syslog switch \"%s\". Ignoring.", value); | |
1510 | else | |
1511 | s->forward_to_syslog = r; | |
1512 | ||
1513 | } else if (proc_cmdline_key_streq(key, "systemd.journald.forward_to_kmsg")) { | |
1514 | ||
1515 | r = value ? parse_boolean(value) : true; | |
1516 | if (r < 0) | |
1517 | log_warning("Failed to parse forward to kmsg switch \"%s\". Ignoring.", value); | |
1518 | else | |
1519 | s->forward_to_kmsg = r; | |
1520 | ||
1521 | } else if (proc_cmdline_key_streq(key, "systemd.journald.forward_to_console")) { | |
1522 | ||
1523 | r = value ? parse_boolean(value) : true; | |
1524 | if (r < 0) | |
1525 | log_warning("Failed to parse forward to console switch \"%s\". Ignoring.", value); | |
1526 | else | |
1527 | s->forward_to_console = r; | |
1528 | ||
1529 | } else if (proc_cmdline_key_streq(key, "systemd.journald.forward_to_wall")) { | |
1530 | ||
1531 | r = value ? parse_boolean(value) : true; | |
1532 | if (r < 0) | |
1533 | log_warning("Failed to parse forward to wall switch \"%s\". Ignoring.", value); | |
1534 | else | |
1535 | s->forward_to_wall = r; | |
1536 | ||
1537 | } else if (proc_cmdline_key_streq(key, "systemd.journald.max_level_console")) { | |
1538 | ||
1539 | if (proc_cmdline_value_missing(key, value)) | |
1540 | return 0; | |
1541 | ||
1542 | r = log_level_from_string(value); | |
1543 | if (r < 0) | |
1544 | log_warning("Failed to parse max level console value \"%s\". Ignoring.", value); | |
1545 | else | |
1546 | s->max_level_console = r; | |
1547 | ||
1548 | } else if (proc_cmdline_key_streq(key, "systemd.journald.max_level_store")) { | |
1549 | ||
1550 | if (proc_cmdline_value_missing(key, value)) | |
1551 | return 0; | |
1552 | ||
1553 | r = log_level_from_string(value); | |
1554 | if (r < 0) | |
1555 | log_warning("Failed to parse max level store value \"%s\". Ignoring.", value); | |
1556 | else | |
1557 | s->max_level_store = r; | |
1558 | ||
1559 | } else if (proc_cmdline_key_streq(key, "systemd.journald.max_level_syslog")) { | |
1560 | ||
1561 | if (proc_cmdline_value_missing(key, value)) | |
1562 | return 0; | |
1563 | ||
1564 | r = log_level_from_string(value); | |
1565 | if (r < 0) | |
1566 | log_warning("Failed to parse max level syslog value \"%s\". Ignoring.", value); | |
1567 | else | |
1568 | s->max_level_syslog = r; | |
1569 | ||
1570 | } else if (proc_cmdline_key_streq(key, "systemd.journald.max_level_kmsg")) { | |
1571 | ||
1572 | if (proc_cmdline_value_missing(key, value)) | |
1573 | return 0; | |
1574 | ||
1575 | r = log_level_from_string(value); | |
1576 | if (r < 0) | |
1577 | log_warning("Failed to parse max level kmsg value \"%s\". Ignoring.", value); | |
1578 | else | |
1579 | s->max_level_kmsg = r; | |
1580 | ||
1581 | } else if (proc_cmdline_key_streq(key, "systemd.journald.max_level_wall")) { | |
1582 | ||
1583 | if (proc_cmdline_value_missing(key, value)) | |
1584 | return 0; | |
1585 | ||
1586 | r = log_level_from_string(value); | |
1587 | if (r < 0) | |
1588 | log_warning("Failed to parse max level wall value \"%s\". Ignoring.", value); | |
1589 | else | |
1590 | s->max_level_wall = r; | |
1591 | ||
1592 | } else if (startswith(key, "systemd.journald")) | |
1593 | log_warning("Unknown journald kernel command line option \"%s\". Ignoring.", key); | |
1594 | ||
1595 | /* do not warn about state here, since probably systemd already did */ | |
1596 | return 0; | |
1597 | } | |
1598 | ||
1599 | static int server_parse_config_file(Server *s) { | |
1600 | assert(s); | |
1601 | ||
1602 | return config_parse_many_nulstr(PKGSYSCONFDIR "/journald.conf", | |
1603 | CONF_PATHS_NULSTR("systemd/journald.conf.d"), | |
1604 | "Journal\0", | |
1605 | config_item_perf_lookup, journald_gperf_lookup, | |
1606 | CONFIG_PARSE_WARN, s); | |
1607 | } | |
1608 | ||
1609 | static int server_dispatch_sync(sd_event_source *es, usec_t t, void *userdata) { | |
1610 | Server *s = userdata; | |
1611 | ||
1612 | assert(s); | |
1613 | ||
1614 | server_sync(s); | |
1615 | return 0; | |
1616 | } | |
1617 | ||
1618 | int server_schedule_sync(Server *s, int priority) { | |
1619 | int r; | |
1620 | ||
1621 | assert(s); | |
1622 | ||
1623 | if (priority <= LOG_CRIT) { | |
1624 | /* Immediately sync to disk when this is of priority CRIT, ALERT, EMERG */ | |
1625 | server_sync(s); | |
1626 | return 0; | |
1627 | } | |
1628 | ||
1629 | if (s->sync_scheduled) | |
1630 | return 0; | |
1631 | ||
1632 | if (s->sync_interval_usec > 0) { | |
1633 | usec_t when; | |
1634 | ||
1635 | r = sd_event_now(s->event, CLOCK_MONOTONIC, &when); | |
1636 | if (r < 0) | |
1637 | return r; | |
1638 | ||
1639 | when += s->sync_interval_usec; | |
1640 | ||
1641 | if (!s->sync_event_source) { | |
1642 | r = sd_event_add_time( | |
1643 | s->event, | |
1644 | &s->sync_event_source, | |
1645 | CLOCK_MONOTONIC, | |
1646 | when, 0, | |
1647 | server_dispatch_sync, s); | |
1648 | if (r < 0) | |
1649 | return r; | |
1650 | ||
1651 | r = sd_event_source_set_priority(s->sync_event_source, SD_EVENT_PRIORITY_IMPORTANT); | |
1652 | } else { | |
1653 | r = sd_event_source_set_time(s->sync_event_source, when); | |
1654 | if (r < 0) | |
1655 | return r; | |
1656 | ||
1657 | r = sd_event_source_set_enabled(s->sync_event_source, SD_EVENT_ONESHOT); | |
1658 | } | |
1659 | if (r < 0) | |
1660 | return r; | |
1661 | ||
1662 | s->sync_scheduled = true; | |
1663 | } | |
1664 | ||
1665 | return 0; | |
1666 | } | |
1667 | ||
1668 | static int dispatch_hostname_change(sd_event_source *es, int fd, uint32_t revents, void *userdata) { | |
1669 | Server *s = userdata; | |
1670 | ||
1671 | assert(s); | |
1672 | ||
1673 | server_cache_hostname(s); | |
1674 | return 0; | |
1675 | } | |
1676 | ||
1677 | static int server_open_hostname(Server *s) { | |
1678 | int r; | |
1679 | ||
1680 | assert(s); | |
1681 | ||
1682 | s->hostname_fd = open("/proc/sys/kernel/hostname", | |
1683 | O_RDONLY|O_CLOEXEC|O_NONBLOCK|O_NOCTTY); | |
1684 | if (s->hostname_fd < 0) | |
1685 | return log_error_errno(errno, "Failed to open /proc/sys/kernel/hostname: %m"); | |
1686 | ||
1687 | r = sd_event_add_io(s->event, &s->hostname_event_source, s->hostname_fd, 0, dispatch_hostname_change, s); | |
1688 | if (r < 0) { | |
1689 | /* kernels prior to 3.2 don't support polling this file. Ignore | |
1690 | * the failure. */ | |
1691 | if (r == -EPERM) { | |
1692 | log_warning_errno(r, "Failed to register hostname fd in event loop, ignoring: %m"); | |
1693 | s->hostname_fd = safe_close(s->hostname_fd); | |
1694 | return 0; | |
1695 | } | |
1696 | ||
1697 | return log_error_errno(r, "Failed to register hostname fd in event loop: %m"); | |
1698 | } | |
1699 | ||
1700 | r = sd_event_source_set_priority(s->hostname_event_source, SD_EVENT_PRIORITY_IMPORTANT-10); | |
1701 | if (r < 0) | |
1702 | return log_error_errno(r, "Failed to adjust priority of host name event source: %m"); | |
1703 | ||
1704 | return 0; | |
1705 | } | |
1706 | ||
1707 | static int dispatch_notify_event(sd_event_source *es, int fd, uint32_t revents, void *userdata) { | |
1708 | Server *s = userdata; | |
1709 | int r; | |
1710 | ||
1711 | assert(s); | |
1712 | assert(s->notify_event_source == es); | |
1713 | assert(s->notify_fd == fd); | |
1714 | ||
1715 | /* The $NOTIFY_SOCKET is writable again, now send exactly one | |
1716 | * message on it. Either it's the watchdog event, the initial | |
1717 | * READY=1 event or an stdout stream event. If there's nothing | |
1718 | * to write anymore, turn our event source off. The next time | |
1719 | * there's something to send it will be turned on again. */ | |
1720 | ||
1721 | if (!s->sent_notify_ready) { | |
1722 | static const char p[] = | |
1723 | "READY=1\n" | |
1724 | "STATUS=Processing requests..."; | |
1725 | ssize_t l; | |
1726 | ||
1727 | l = send(s->notify_fd, p, strlen(p), MSG_DONTWAIT); | |
1728 | if (l < 0) { | |
1729 | if (errno == EAGAIN) | |
1730 | return 0; | |
1731 | ||
1732 | return log_error_errno(errno, "Failed to send READY=1 notification message: %m"); | |
1733 | } | |
1734 | ||
1735 | s->sent_notify_ready = true; | |
1736 | log_debug("Sent READY=1 notification."); | |
1737 | ||
1738 | } else if (s->send_watchdog) { | |
1739 | ||
1740 | static const char p[] = | |
1741 | "WATCHDOG=1"; | |
1742 | ||
1743 | ssize_t l; | |
1744 | ||
1745 | l = send(s->notify_fd, p, strlen(p), MSG_DONTWAIT); | |
1746 | if (l < 0) { | |
1747 | if (errno == EAGAIN) | |
1748 | return 0; | |
1749 | ||
1750 | return log_error_errno(errno, "Failed to send WATCHDOG=1 notification message: %m"); | |
1751 | } | |
1752 | ||
1753 | s->send_watchdog = false; | |
1754 | log_debug("Sent WATCHDOG=1 notification."); | |
1755 | ||
1756 | } else if (s->stdout_streams_notify_queue) | |
1757 | /* Dispatch one stream notification event */ | |
1758 | stdout_stream_send_notify(s->stdout_streams_notify_queue); | |
1759 | ||
1760 | /* Leave us enabled if there's still more to do. */ | |
1761 | if (s->send_watchdog || s->stdout_streams_notify_queue) | |
1762 | return 0; | |
1763 | ||
1764 | /* There was nothing to do anymore, let's turn ourselves off. */ | |
1765 | r = sd_event_source_set_enabled(es, SD_EVENT_OFF); | |
1766 | if (r < 0) | |
1767 | return log_error_errno(r, "Failed to turn off notify event source: %m"); | |
1768 | ||
1769 | return 0; | |
1770 | } | |
1771 | ||
1772 | static int dispatch_watchdog(sd_event_source *es, uint64_t usec, void *userdata) { | |
1773 | Server *s = userdata; | |
1774 | int r; | |
1775 | ||
1776 | assert(s); | |
1777 | ||
1778 | s->send_watchdog = true; | |
1779 | ||
1780 | r = sd_event_source_set_enabled(s->notify_event_source, SD_EVENT_ON); | |
1781 | if (r < 0) | |
1782 | log_warning_errno(r, "Failed to turn on notify event source: %m"); | |
1783 | ||
1784 | r = sd_event_source_set_time(s->watchdog_event_source, usec + s->watchdog_usec / 2); | |
1785 | if (r < 0) | |
1786 | return log_error_errno(r, "Failed to restart watchdog event source: %m"); | |
1787 | ||
1788 | r = sd_event_source_set_enabled(s->watchdog_event_source, SD_EVENT_ON); | |
1789 | if (r < 0) | |
1790 | return log_error_errno(r, "Failed to enable watchdog event source: %m"); | |
1791 | ||
1792 | return 0; | |
1793 | } | |
1794 | ||
1795 | static int server_connect_notify(Server *s) { | |
1796 | union sockaddr_union sa = {}; | |
1797 | const char *e; | |
1798 | int r, salen; | |
1799 | ||
1800 | assert(s); | |
1801 | assert(s->notify_fd < 0); | |
1802 | assert(!s->notify_event_source); | |
1803 | ||
1804 | /* | |
1805 | * So here's the problem: we'd like to send notification messages to PID 1, but we cannot do that via | |
1806 | * sd_notify(), since that's synchronous, and we might end up blocking on it. Specifically: given | |
1807 | * that PID 1 might block on dbus-daemon during IPC, and dbus-daemon is logging to us, and might | |
1808 | * hence block on us, we might end up in a deadlock if we block on sending PID 1 notification | |
1809 | * messages — by generating a full blocking circle. To avoid this, let's create a non-blocking | |
1810 | * socket, and connect it to the notification socket, and then wait for POLLOUT before we send | |
1811 | * anything. This should efficiently avoid any deadlocks, as we'll never block on PID 1, hence PID 1 | |
1812 | * can safely block on dbus-daemon which can safely block on us again. | |
1813 | * | |
1814 | * Don't think that this issue is real? It is, see: https://github.com/systemd/systemd/issues/1505 | |
1815 | */ | |
1816 | ||
1817 | e = getenv("NOTIFY_SOCKET"); | |
1818 | if (!e) | |
1819 | return 0; | |
1820 | ||
1821 | salen = sockaddr_un_set_path(&sa.un, e); | |
1822 | if (salen < 0) | |
1823 | return log_error_errno(salen, "NOTIFY_SOCKET set to invalid value '%s': %m", e); | |
1824 | ||
1825 | s->notify_fd = socket(AF_UNIX, SOCK_DGRAM|SOCK_CLOEXEC|SOCK_NONBLOCK, 0); | |
1826 | if (s->notify_fd < 0) | |
1827 | return log_error_errno(errno, "Failed to create notify socket: %m"); | |
1828 | ||
1829 | (void) fd_inc_sndbuf(s->notify_fd, NOTIFY_SNDBUF_SIZE); | |
1830 | ||
1831 | r = connect(s->notify_fd, &sa.sa, salen); | |
1832 | if (r < 0) | |
1833 | return log_error_errno(errno, "Failed to connect to notify socket: %m"); | |
1834 | ||
1835 | r = sd_event_add_io(s->event, &s->notify_event_source, s->notify_fd, EPOLLOUT, dispatch_notify_event, s); | |
1836 | if (r < 0) | |
1837 | return log_error_errno(r, "Failed to watch notification socket: %m"); | |
1838 | ||
1839 | if (sd_watchdog_enabled(false, &s->watchdog_usec) > 0) { | |
1840 | s->send_watchdog = true; | |
1841 | ||
1842 | r = sd_event_add_time(s->event, &s->watchdog_event_source, CLOCK_MONOTONIC, now(CLOCK_MONOTONIC) + s->watchdog_usec/2, s->watchdog_usec/4, dispatch_watchdog, s); | |
1843 | if (r < 0) | |
1844 | return log_error_errno(r, "Failed to add watchdog time event: %m"); | |
1845 | } | |
1846 | ||
1847 | /* This should fire pretty soon, which we'll use to send the READY=1 event. */ | |
1848 | ||
1849 | return 0; | |
1850 | } | |
1851 | ||
1852 | static int synchronize_second_half(sd_event_source *event_source, void *userdata) { | |
1853 | Varlink *link = userdata; | |
1854 | Server *s; | |
1855 | int r; | |
1856 | ||
1857 | assert(link); | |
1858 | assert_se(s = varlink_get_userdata(link)); | |
1859 | ||
1860 | /* This is the "second half" of the Synchronize() varlink method. This function is called as deferred | |
1861 | * event source at a low priority to ensure the synchronization completes after all queued log | |
1862 | * messages are processed. */ | |
1863 | server_full_sync(s); | |
1864 | ||
1865 | /* Let's get rid of the event source now, by marking it as non-floating again. It then has no ref | |
1866 | * anymore and is immediately destroyed after we return from this function, i.e. from this event | |
1867 | * source handler at the end. */ | |
1868 | r = sd_event_source_set_floating(event_source, false); | |
1869 | if (r < 0) | |
1870 | return log_error_errno(r, "Failed to mark event source as non-floating: %m"); | |
1871 | ||
1872 | return varlink_reply(link, NULL); | |
1873 | } | |
1874 | ||
1875 | static void synchronize_destroy(void *userdata) { | |
1876 | varlink_unref(userdata); | |
1877 | } | |
1878 | ||
1879 | static int vl_method_synchronize(Varlink *link, JsonVariant *parameters, VarlinkMethodFlags flags, void *userdata) { | |
1880 | _cleanup_(sd_event_source_unrefp) sd_event_source *event_source = NULL; | |
1881 | Server *s = userdata; | |
1882 | int r; | |
1883 | ||
1884 | assert(link); | |
1885 | assert(s); | |
1886 | ||
1887 | if (json_variant_elements(parameters) > 0) | |
1888 | return varlink_error_invalid_parameter(link, parameters); | |
1889 | ||
1890 | log_info("Received client request to rotate journal."); | |
1891 | ||
1892 | /* We don't do the main work now, but instead enqueue a deferred event loop job which will do | |
1893 | * it. That job is scheduled at low priority, so that we return from this method call only after all | |
1894 | * queued but not processed log messages are written to disk, so that this method call returning can | |
1895 | * be used as nice synchronization point. */ | |
1896 | r = sd_event_add_defer(s->event, &event_source, synchronize_second_half, link); | |
1897 | if (r < 0) | |
1898 | return log_error_errno(r, "Failed to allocate defer event source: %m"); | |
1899 | ||
1900 | r = sd_event_source_set_destroy_callback(event_source, synchronize_destroy); | |
1901 | if (r < 0) | |
1902 | return log_error_errno(r, "Failed to set event source destroy callback: %m"); | |
1903 | ||
1904 | varlink_ref(link); /* The varlink object is now left to the destroy callack to unref */ | |
1905 | ||
1906 | r = sd_event_source_set_priority(event_source, SD_EVENT_PRIORITY_NORMAL+15); | |
1907 | if (r < 0) | |
1908 | return log_error_errno(r, "Failed to set defer event source priority: %m"); | |
1909 | ||
1910 | /* Give up ownership of this event source. It will now be destroyed along with event loop itself, | |
1911 | * unless it destroys itself earlier. */ | |
1912 | r = sd_event_source_set_floating(event_source, true); | |
1913 | if (r < 0) | |
1914 | return log_error_errno(r, "Failed to mark event source as floating: %m"); | |
1915 | ||
1916 | (void) sd_event_source_set_description(event_source, "deferred-sync"); | |
1917 | ||
1918 | return 0; | |
1919 | } | |
1920 | ||
1921 | static int vl_method_rotate(Varlink *link, JsonVariant *parameters, VarlinkMethodFlags flags, void *userdata) { | |
1922 | Server *s = userdata; | |
1923 | ||
1924 | assert(link); | |
1925 | assert(s); | |
1926 | ||
1927 | if (json_variant_elements(parameters) > 0) | |
1928 | return varlink_error_invalid_parameter(link, parameters); | |
1929 | ||
1930 | log_info("Received client request to rotate journal."); | |
1931 | server_full_rotate(s); | |
1932 | ||
1933 | return varlink_reply(link, NULL); | |
1934 | } | |
1935 | ||
1936 | static int vl_method_flush_to_var(Varlink *link, JsonVariant *parameters, VarlinkMethodFlags flags, void *userdata) { | |
1937 | Server *s = userdata; | |
1938 | ||
1939 | assert(link); | |
1940 | assert(s); | |
1941 | ||
1942 | if (json_variant_elements(parameters) > 0) | |
1943 | return varlink_error_invalid_parameter(link, parameters); | |
1944 | ||
1945 | log_info("Received client request to flush runtime journal."); | |
1946 | server_full_flush(s); | |
1947 | ||
1948 | return varlink_reply(link, NULL); | |
1949 | } | |
1950 | ||
1951 | static int vl_method_relinquish_var(Varlink *link, JsonVariant *parameters, VarlinkMethodFlags flags, void *userdata) { | |
1952 | Server *s = userdata; | |
1953 | ||
1954 | assert(link); | |
1955 | assert(s); | |
1956 | ||
1957 | if (json_variant_elements(parameters) > 0) | |
1958 | return varlink_error_invalid_parameter(link, parameters); | |
1959 | ||
1960 | log_info("Received client request to relinquish /var access."); | |
1961 | server_relinquish_var(s); | |
1962 | ||
1963 | return varlink_reply(link, NULL); | |
1964 | } | |
1965 | ||
1966 | static int server_open_varlink(Server *s) { | |
1967 | int r; | |
1968 | ||
1969 | assert(s); | |
1970 | ||
1971 | r = varlink_server_new(&s->varlink_server, VARLINK_SERVER_ROOT_ONLY); | |
1972 | if (r < 0) | |
1973 | return r; | |
1974 | ||
1975 | varlink_server_set_userdata(s->varlink_server, s); | |
1976 | ||
1977 | r = varlink_server_bind_method_many( | |
1978 | s->varlink_server, | |
1979 | "io.systemd.Journal.Synchronize", vl_method_synchronize, | |
1980 | "io.systemd.Journal.Rotate", vl_method_rotate, | |
1981 | "io.systemd.Journal.FlushToVar", vl_method_flush_to_var, | |
1982 | "io.systemd.Journal.RelinquishVar", vl_method_relinquish_var); | |
1983 | if (r < 0) | |
1984 | return r; | |
1985 | ||
1986 | r = varlink_server_listen_address(s->varlink_server, "/run/systemd/journal/io.systemd.journal", 0600); | |
1987 | if (r < 0) | |
1988 | return r; | |
1989 | ||
1990 | r = varlink_server_attach_event(s->varlink_server, s->event, SD_EVENT_PRIORITY_NORMAL); | |
1991 | if (r < 0) | |
1992 | return r; | |
1993 | ||
1994 | return 0; | |
1995 | } | |
1996 | ||
1997 | int server_init(Server *s) { | |
1998 | _cleanup_fdset_free_ FDSet *fds = NULL; | |
1999 | int n, r, fd; | |
2000 | bool no_sockets; | |
2001 | ||
2002 | assert(s); | |
2003 | ||
2004 | *s = (Server) { | |
2005 | .syslog_fd = -1, | |
2006 | .native_fd = -1, | |
2007 | .stdout_fd = -1, | |
2008 | .dev_kmsg_fd = -1, | |
2009 | .audit_fd = -1, | |
2010 | .hostname_fd = -1, | |
2011 | .notify_fd = -1, | |
2012 | ||
2013 | .compress.enabled = true, | |
2014 | .compress.threshold_bytes = (uint64_t) -1, | |
2015 | .seal = true, | |
2016 | .read_kmsg = true, | |
2017 | ||
2018 | .watchdog_usec = USEC_INFINITY, | |
2019 | ||
2020 | .sync_interval_usec = DEFAULT_SYNC_INTERVAL_USEC, | |
2021 | .sync_scheduled = false, | |
2022 | ||
2023 | .rate_limit_interval = DEFAULT_RATE_LIMIT_INTERVAL, | |
2024 | .rate_limit_burst = DEFAULT_RATE_LIMIT_BURST, | |
2025 | ||
2026 | .forward_to_wall = true, | |
2027 | ||
2028 | .max_file_usec = DEFAULT_MAX_FILE_USEC, | |
2029 | ||
2030 | .max_level_store = LOG_DEBUG, | |
2031 | .max_level_syslog = LOG_DEBUG, | |
2032 | .max_level_kmsg = LOG_NOTICE, | |
2033 | .max_level_console = LOG_INFO, | |
2034 | .max_level_wall = LOG_EMERG, | |
2035 | ||
2036 | .line_max = DEFAULT_LINE_MAX, | |
2037 | ||
2038 | .runtime_storage.name = "Runtime Journal", | |
2039 | .system_storage.name = "System Journal", | |
2040 | }; | |
2041 | ||
2042 | journal_reset_metrics(&s->system_storage.metrics); | |
2043 | journal_reset_metrics(&s->runtime_storage.metrics); | |
2044 | ||
2045 | server_parse_config_file(s); | |
2046 | ||
2047 | r = proc_cmdline_parse(parse_proc_cmdline_item, s, PROC_CMDLINE_STRIP_RD_PREFIX); | |
2048 | if (r < 0) | |
2049 | log_warning_errno(r, "Failed to parse kernel command line, ignoring: %m"); | |
2050 | ||
2051 | if (!!s->rate_limit_interval ^ !!s->rate_limit_burst) { | |
2052 | log_debug("Setting both rate limit interval and burst from "USEC_FMT",%u to 0,0", | |
2053 | s->rate_limit_interval, s->rate_limit_burst); | |
2054 | s->rate_limit_interval = s->rate_limit_burst = 0; | |
2055 | } | |
2056 | ||
2057 | (void) mkdir_p("/run/systemd/journal", 0755); | |
2058 | ||
2059 | s->user_journals = ordered_hashmap_new(NULL); | |
2060 | if (!s->user_journals) | |
2061 | return log_oom(); | |
2062 | ||
2063 | s->mmap = mmap_cache_new(); | |
2064 | if (!s->mmap) | |
2065 | return log_oom(); | |
2066 | ||
2067 | s->deferred_closes = set_new(NULL); | |
2068 | if (!s->deferred_closes) | |
2069 | return log_oom(); | |
2070 | ||
2071 | r = sd_event_default(&s->event); | |
2072 | if (r < 0) | |
2073 | return log_error_errno(r, "Failed to create event loop: %m"); | |
2074 | ||
2075 | n = sd_listen_fds(true); | |
2076 | if (n < 0) | |
2077 | return log_error_errno(n, "Failed to read listening file descriptors from environment: %m"); | |
2078 | ||
2079 | for (fd = SD_LISTEN_FDS_START; fd < SD_LISTEN_FDS_START + n; fd++) { | |
2080 | ||
2081 | if (sd_is_socket_unix(fd, SOCK_DGRAM, -1, "/run/systemd/journal/socket", 0) > 0) { | |
2082 | ||
2083 | if (s->native_fd >= 0) | |
2084 | return log_error_errno(SYNTHETIC_ERRNO(EINVAL), | |
2085 | "Too many native sockets passed."); | |
2086 | ||
2087 | s->native_fd = fd; | |
2088 | ||
2089 | } else if (sd_is_socket_unix(fd, SOCK_STREAM, 1, "/run/systemd/journal/stdout", 0) > 0) { | |
2090 | ||
2091 | if (s->stdout_fd >= 0) | |
2092 | return log_error_errno(SYNTHETIC_ERRNO(EINVAL), | |
2093 | "Too many stdout sockets passed."); | |
2094 | ||
2095 | s->stdout_fd = fd; | |
2096 | ||
2097 | } else if (sd_is_socket_unix(fd, SOCK_DGRAM, -1, "/dev/log", 0) > 0 || | |
2098 | sd_is_socket_unix(fd, SOCK_DGRAM, -1, "/run/systemd/journal/dev-log", 0) > 0) { | |
2099 | ||
2100 | if (s->syslog_fd >= 0) | |
2101 | return log_error_errno(SYNTHETIC_ERRNO(EINVAL), | |
2102 | "Too many /dev/log sockets passed."); | |
2103 | ||
2104 | s->syslog_fd = fd; | |
2105 | ||
2106 | } else if (sd_is_socket(fd, AF_NETLINK, SOCK_RAW, -1) > 0) { | |
2107 | ||
2108 | if (s->audit_fd >= 0) | |
2109 | return log_error_errno(SYNTHETIC_ERRNO(EINVAL), | |
2110 | "Too many audit sockets passed."); | |
2111 | ||
2112 | s->audit_fd = fd; | |
2113 | ||
2114 | } else { | |
2115 | ||
2116 | if (!fds) { | |
2117 | fds = fdset_new(); | |
2118 | if (!fds) | |
2119 | return log_oom(); | |
2120 | } | |
2121 | ||
2122 | r = fdset_put(fds, fd); | |
2123 | if (r < 0) | |
2124 | return log_oom(); | |
2125 | } | |
2126 | } | |
2127 | ||
2128 | /* Try to restore streams, but don't bother if this fails */ | |
2129 | (void) server_restore_streams(s, fds); | |
2130 | ||
2131 | if (fdset_size(fds) > 0) { | |
2132 | log_warning("%u unknown file descriptors passed, closing.", fdset_size(fds)); | |
2133 | fds = fdset_free(fds); | |
2134 | } | |
2135 | ||
2136 | no_sockets = s->native_fd < 0 && s->stdout_fd < 0 && s->syslog_fd < 0 && s->audit_fd < 0; | |
2137 | ||
2138 | /* always open stdout, syslog, native, and kmsg sockets */ | |
2139 | ||
2140 | /* systemd-journald.socket: /run/systemd/journal/stdout */ | |
2141 | r = server_open_stdout_socket(s); | |
2142 | if (r < 0) | |
2143 | return r; | |
2144 | ||
2145 | /* systemd-journald-dev-log.socket: /run/systemd/journal/dev-log */ | |
2146 | r = server_open_syslog_socket(s); | |
2147 | if (r < 0) | |
2148 | return r; | |
2149 | ||
2150 | /* systemd-journald.socket: /run/systemd/journal/socket */ | |
2151 | r = server_open_native_socket(s); | |
2152 | if (r < 0) | |
2153 | return r; | |
2154 | ||
2155 | /* /dev/kmsg */ | |
2156 | r = server_open_dev_kmsg(s); | |
2157 | if (r < 0) | |
2158 | return r; | |
2159 | ||
2160 | /* Unless we got *some* sockets and not audit, open audit socket */ | |
2161 | if (s->audit_fd >= 0 || no_sockets) { | |
2162 | r = server_open_audit(s); | |
2163 | if (r < 0) | |
2164 | return r; | |
2165 | } | |
2166 | ||
2167 | r = server_open_varlink(s); | |
2168 | if (r < 0) | |
2169 | return r; | |
2170 | ||
2171 | r = server_open_kernel_seqnum(s); | |
2172 | if (r < 0) | |
2173 | return r; | |
2174 | ||
2175 | r = server_open_hostname(s); | |
2176 | if (r < 0) | |
2177 | return r; | |
2178 | ||
2179 | r = setup_signals(s); | |
2180 | if (r < 0) | |
2181 | return r; | |
2182 | ||
2183 | s->rate_limit = journal_rate_limit_new(); | |
2184 | if (!s->rate_limit) | |
2185 | return -ENOMEM; | |
2186 | ||
2187 | r = cg_get_root_path(&s->cgroup_root); | |
2188 | if (r < 0) | |
2189 | return r; | |
2190 | ||
2191 | server_cache_hostname(s); | |
2192 | server_cache_boot_id(s); | |
2193 | server_cache_machine_id(s); | |
2194 | ||
2195 | s->runtime_storage.path = path_join("/run/log/journal", SERVER_MACHINE_ID(s)); | |
2196 | s->system_storage.path = path_join("/var/log/journal", SERVER_MACHINE_ID(s)); | |
2197 | if (!s->runtime_storage.path || !s->system_storage.path) | |
2198 | return -ENOMEM; | |
2199 | ||
2200 | (void) server_connect_notify(s); | |
2201 | ||
2202 | (void) client_context_acquire_default(s); | |
2203 | ||
2204 | return system_journal_open(s, false, false); | |
2205 | } | |
2206 | ||
2207 | void server_maybe_append_tags(Server *s) { | |
2208 | #if HAVE_GCRYPT | |
2209 | JournalFile *f; | |
2210 | Iterator i; | |
2211 | usec_t n; | |
2212 | ||
2213 | n = now(CLOCK_REALTIME); | |
2214 | ||
2215 | if (s->system_journal) | |
2216 | journal_file_maybe_append_tag(s->system_journal, n); | |
2217 | ||
2218 | ORDERED_HASHMAP_FOREACH(f, s->user_journals, i) | |
2219 | journal_file_maybe_append_tag(f, n); | |
2220 | #endif | |
2221 | } | |
2222 | ||
2223 | void server_done(Server *s) { | |
2224 | assert(s); | |
2225 | ||
2226 | set_free_with_destructor(s->deferred_closes, journal_file_close); | |
2227 | ||
2228 | while (s->stdout_streams) | |
2229 | stdout_stream_free(s->stdout_streams); | |
2230 | ||
2231 | client_context_flush_all(s); | |
2232 | ||
2233 | (void) journal_file_close(s->system_journal); | |
2234 | (void) journal_file_close(s->runtime_journal); | |
2235 | ||
2236 | ordered_hashmap_free_with_destructor(s->user_journals, journal_file_close); | |
2237 | ||
2238 | varlink_server_unref(s->varlink_server); | |
2239 | ||
2240 | sd_event_source_unref(s->syslog_event_source); | |
2241 | sd_event_source_unref(s->native_event_source); | |
2242 | sd_event_source_unref(s->stdout_event_source); | |
2243 | sd_event_source_unref(s->dev_kmsg_event_source); | |
2244 | sd_event_source_unref(s->audit_event_source); | |
2245 | sd_event_source_unref(s->sync_event_source); | |
2246 | sd_event_source_unref(s->sigusr1_event_source); | |
2247 | sd_event_source_unref(s->sigusr2_event_source); | |
2248 | sd_event_source_unref(s->sigterm_event_source); | |
2249 | sd_event_source_unref(s->sigint_event_source); | |
2250 | sd_event_source_unref(s->sigrtmin1_event_source); | |
2251 | sd_event_source_unref(s->hostname_event_source); | |
2252 | sd_event_source_unref(s->notify_event_source); | |
2253 | sd_event_source_unref(s->watchdog_event_source); | |
2254 | sd_event_unref(s->event); | |
2255 | ||
2256 | safe_close(s->syslog_fd); | |
2257 | safe_close(s->native_fd); | |
2258 | safe_close(s->stdout_fd); | |
2259 | safe_close(s->dev_kmsg_fd); | |
2260 | safe_close(s->audit_fd); | |
2261 | safe_close(s->hostname_fd); | |
2262 | safe_close(s->notify_fd); | |
2263 | ||
2264 | if (s->rate_limit) | |
2265 | journal_rate_limit_free(s->rate_limit); | |
2266 | ||
2267 | if (s->kernel_seqnum) | |
2268 | munmap(s->kernel_seqnum, sizeof(uint64_t)); | |
2269 | ||
2270 | free(s->buffer); | |
2271 | free(s->tty_path); | |
2272 | free(s->cgroup_root); | |
2273 | free(s->hostname_field); | |
2274 | free(s->runtime_storage.path); | |
2275 | free(s->system_storage.path); | |
2276 | ||
2277 | mmap_cache_unref(s->mmap); | |
2278 | } | |
2279 | ||
2280 | static const char* const storage_table[_STORAGE_MAX] = { | |
2281 | [STORAGE_AUTO] = "auto", | |
2282 | [STORAGE_VOLATILE] = "volatile", | |
2283 | [STORAGE_PERSISTENT] = "persistent", | |
2284 | [STORAGE_NONE] = "none" | |
2285 | }; | |
2286 | ||
2287 | DEFINE_STRING_TABLE_LOOKUP(storage, Storage); | |
2288 | DEFINE_CONFIG_PARSE_ENUM(config_parse_storage, storage, Storage, "Failed to parse storage setting"); | |
2289 | ||
2290 | static const char* const split_mode_table[_SPLIT_MAX] = { | |
2291 | [SPLIT_LOGIN] = "login", | |
2292 | [SPLIT_UID] = "uid", | |
2293 | [SPLIT_NONE] = "none", | |
2294 | }; | |
2295 | ||
2296 | DEFINE_STRING_TABLE_LOOKUP(split_mode, SplitMode); | |
2297 | DEFINE_CONFIG_PARSE_ENUM(config_parse_split_mode, split_mode, SplitMode, "Failed to parse split mode setting"); | |
2298 | ||
2299 | int config_parse_line_max( | |
2300 | const char* unit, | |
2301 | const char *filename, | |
2302 | unsigned line, | |
2303 | const char *section, | |
2304 | unsigned section_line, | |
2305 | const char *lvalue, | |
2306 | int ltype, | |
2307 | const char *rvalue, | |
2308 | void *data, | |
2309 | void *userdata) { | |
2310 | ||
2311 | size_t *sz = data; | |
2312 | int r; | |
2313 | ||
2314 | assert(filename); | |
2315 | assert(lvalue); | |
2316 | assert(rvalue); | |
2317 | assert(data); | |
2318 | ||
2319 | if (isempty(rvalue)) | |
2320 | /* Empty assignment means default */ | |
2321 | *sz = DEFAULT_LINE_MAX; | |
2322 | else { | |
2323 | uint64_t v; | |
2324 | ||
2325 | r = parse_size(rvalue, 1024, &v); | |
2326 | if (r < 0) { | |
2327 | log_syntax(unit, LOG_ERR, filename, line, r, "Failed to parse LineMax= value, ignoring: %s", rvalue); | |
2328 | return 0; | |
2329 | } | |
2330 | ||
2331 | if (v < 79) { | |
2332 | /* Why specify 79 here as minimum line length? Simply, because the most common traditional | |
2333 | * terminal size is 80ch, and it might make sense to break one character before the natural | |
2334 | * line break would occur on that. */ | |
2335 | log_syntax(unit, LOG_WARNING, filename, line, 0, "LineMax= too small, clamping to 79: %s", rvalue); | |
2336 | *sz = 79; | |
2337 | } else if (v > (uint64_t) (SSIZE_MAX-1)) { | |
2338 | /* So, why specify SSIZE_MAX-1 here? Because that's one below the largest size value read() | |
2339 | * can return, and we need one extra byte for the trailing NUL byte. Of course IRL such large | |
2340 | * memory allocations will fail anyway, hence this limit is mostly theoretical anyway, as we'll | |
2341 | * fail much earlier anyway. */ | |
2342 | log_syntax(unit, LOG_WARNING, filename, line, 0, "LineMax= too large, clamping to %" PRIu64 ": %s", (uint64_t) (SSIZE_MAX-1), rvalue); | |
2343 | *sz = SSIZE_MAX-1; | |
2344 | } else | |
2345 | *sz = (size_t) v; | |
2346 | } | |
2347 | ||
2348 | return 0; | |
2349 | } | |
2350 | ||
2351 | int config_parse_compress( | |
2352 | const char* unit, | |
2353 | const char *filename, | |
2354 | unsigned line, | |
2355 | const char *section, | |
2356 | unsigned section_line, | |
2357 | const char *lvalue, | |
2358 | int ltype, | |
2359 | const char *rvalue, | |
2360 | void *data, | |
2361 | void *userdata) { | |
2362 | ||
2363 | JournalCompressOptions* compress = data; | |
2364 | int r; | |
2365 | ||
2366 | if (isempty(rvalue)) { | |
2367 | compress->enabled = true; | |
2368 | compress->threshold_bytes = (uint64_t) -1; | |
2369 | } else if (streq(rvalue, "1")) { | |
2370 | log_syntax(unit, LOG_WARNING, filename, line, 0, | |
2371 | "Compress= ambiguously specified as 1, enabling compression with default threshold"); | |
2372 | compress->enabled = true; | |
2373 | } else if (streq(rvalue, "0")) { | |
2374 | log_syntax(unit, LOG_WARNING, filename, line, 0, | |
2375 | "Compress= ambiguously specified as 0, disabling compression"); | |
2376 | compress->enabled = false; | |
2377 | } else { | |
2378 | r = parse_boolean(rvalue); | |
2379 | if (r < 0) { | |
2380 | r = parse_size(rvalue, 1024, &compress->threshold_bytes); | |
2381 | if (r < 0) | |
2382 | log_syntax(unit, LOG_ERR, filename, line, r, | |
2383 | "Failed to parse Compress= value, ignoring: %s", rvalue); | |
2384 | else | |
2385 | compress->enabled = true; | |
2386 | } else | |
2387 | compress->enabled = r; | |
2388 | } | |
2389 | ||
2390 | return 0; | |
2391 | } |