journald-server.c 51 KB
Newer Older
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
/*-*- Mode: C; c-basic-offset: 8; indent-tabs-mode: nil -*-*/

/***
  This file is part of systemd.

  Copyright 2011 Lennart Poettering

  systemd is free software; you can redistribute it and/or modify it
  under the terms of the GNU Lesser General Public License as published by
  the Free Software Foundation; either version 2.1 of the License, or
  (at your option) any later version.

  systemd is distributed in the hope that it will be useful, but
  WITHOUT ANY WARRANTY; without even the implied warranty of
  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
  Lesser General Public License for more details.

  You should have received a copy of the GNU Lesser General Public License
  along with systemd; If not, see <http://www.gnu.org/licenses/>.
***/

#include <sys/signalfd.h>
#include <sys/ioctl.h>
#include <linux/sockios.h>
#include <sys/statvfs.h>
#include <sys/mman.h>
27
#include <sys/timerfd.h>
28 29 30 31 32 33 34 35 36 37

#include <libudev.h>
#include <systemd/sd-journal.h>
#include <systemd/sd-messages.h>
#include <systemd/sd-daemon.h>

#ifdef HAVE_LOGIND
#include <systemd/sd-login.h>
#endif

38
#include "fileio.h"
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
#include "mkdir.h"
#include "hashmap.h"
#include "journal-file.h"
#include "socket-util.h"
#include "cgroup-util.h"
#include "list.h"
#include "virt.h"
#include "missing.h"
#include "conf-parser.h"
#include "journal-internal.h"
#include "journal-vacuum.h"
#include "journal-authenticate.h"
#include "journald-server.h"
#include "journald-rate-limit.h"
#include "journald-kmsg.h"
#include "journald-syslog.h"
#include "journald-stream.h"
#include "journald-console.h"
#include "journald-native.h"

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

#ifdef HAVE_SELINUX
#include <selinux/selinux.h>
#endif

#define USER_JOURNALS_MAX 1024

71
#define DEFAULT_SYNC_INTERVAL_USEC (5*USEC_PER_MINUTE)
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
#define DEFAULT_RATE_LIMIT_INTERVAL (10*USEC_PER_SEC)
#define DEFAULT_RATE_LIMIT_BURST 200

#define RECHECK_AVAILABLE_SPACE_USEC (30*USEC_PER_SEC)

static const char* const storage_table[] = {
        [STORAGE_AUTO] = "auto",
        [STORAGE_VOLATILE] = "volatile",
        [STORAGE_PERSISTENT] = "persistent",
        [STORAGE_NONE] = "none"
};

DEFINE_STRING_TABLE_LOOKUP(storage, Storage);
DEFINE_CONFIG_PARSE_ENUM(config_parse_storage, storage, Storage, "Failed to parse storage setting");

static const char* const split_mode_table[] = {
        [SPLIT_NONE] = "none",
        [SPLIT_UID] = "uid",
        [SPLIT_LOGIN] = "login"
};

DEFINE_STRING_TABLE_LOOKUP(split_mode, SplitMode);
DEFINE_CONFIG_PARSE_ENUM(config_parse_split_mode, split_mode, SplitMode, "Failed to parse split mode setting");

static uint64_t available_space(Server *s) {
97 98
        char ids[33];
        char _cleanup_free_ *p = NULL;
99 100 101 102 103
        const char *f;
        sd_id128_t machine;
        struct statvfs ss;
        uint64_t sum = 0, avail = 0, ss_avail = 0;
        int r;
104
        DIR _cleanup_closedir_ *d = NULL;
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
        usec_t ts;
        JournalMetrics *m;

        ts = now(CLOCK_MONOTONIC);

        if (s->cached_available_space_timestamp + RECHECK_AVAILABLE_SPACE_USEC > ts)
                return s->cached_available_space;

        r = sd_id128_get_machine(&machine);
        if (r < 0)
                return 0;

        if (s->system_journal) {
                f = "/var/log/journal/";
                m = &s->system_metrics;
        } else {
                f = "/run/log/journal/";
                m = &s->runtime_metrics;
        }

        assert(m);

        p = strappend(f, sd_id128_to_string(machine, ids));
        if (!p)
                return 0;

        d = opendir(p);
        if (!d)
                return 0;

        if (fstatvfs(dirfd(d), &ss) < 0)
136
                return 0;
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

        for (;;) {
                struct stat st;
                struct dirent *de;
                union dirent_storage buf;

                r = readdir_r(d, &buf.de, &de);
                if (r != 0)
                        break;

                if (!de)
                        break;

                if (!endswith(de->d_name, ".journal") &&
                    !endswith(de->d_name, ".journal~"))
                        continue;

                if (fstatat(dirfd(d), de->d_name, &st, AT_SYMLINK_NOFOLLOW) < 0)
                        continue;

                if (!S_ISREG(st.st_mode))
                        continue;

                sum += (uint64_t) st.st_blocks * 512UL;
        }

        avail = sum >= m->max_use ? 0 : m->max_use - sum;

        ss_avail = ss.f_bsize * ss.f_bavail;

        ss_avail = ss_avail < m->keep_free ? 0 : ss_avail - m->keep_free;

        if (ss_avail < avail)
                avail = ss_avail;

        s->cached_available_space = avail;
        s->cached_available_space_timestamp = ts;

        return avail;
}

static void server_read_file_gid(Server *s) {
179
        const char *g = "systemd-journal";
180 181 182 183 184 185 186
        int r;

        assert(s);

        if (s->file_gid_valid)
                return;

187
        r = get_group_creds(&g, &s->file_gid);
188
        if (r < 0)
189
                log_warning("Failed to resolve '%s' group: %s", g, strerror(-r));
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

        /* if we couldn't read the gid, then it will be 0, but that's
         * fine and we shouldn't try to resolve the group again, so
         * let's just pretend it worked right-away. */
        s->file_gid_valid = true;
}

void server_fix_perms(Server *s, JournalFile *f, uid_t uid) {
        int r;
#ifdef HAVE_ACL
        acl_t acl;
        acl_entry_t entry;
        acl_permset_t permset;
#endif

        assert(f);

        server_read_file_gid(s);

        r = fchmod_and_fchown(f->fd, 0640, 0, s->file_gid);
        if (r < 0)
                log_warning("Failed to fix access mode/rights on %s, ignoring: %s", f->path, strerror(-r));

#ifdef HAVE_ACL
        if (uid <= 0)
                return;

        acl = acl_get_fd(f->fd);
        if (!acl) {
                log_warning("Failed to read ACL on %s, ignoring: %m", f->path);
                return;
        }

        r = acl_find_uid(acl, uid, &entry);
        if (r <= 0) {

                if (acl_create_entry(&acl, &entry) < 0 ||
                    acl_set_tag_type(entry, ACL_USER) < 0 ||
                    acl_set_qualifier(entry, &uid) < 0) {
                        log_warning("Failed to patch ACL on %s, ignoring: %m", f->path);
                        goto finish;
                }
        }

        if (acl_get_permset(entry, &permset) < 0 ||
            acl_add_perm(permset, ACL_READ) < 0 ||
            acl_calc_mask(&acl) < 0) {
                log_warning("Failed to patch ACL on %s, ignoring: %m", f->path);
                goto finish;
        }

        if (acl_set_fd(f->fd, acl) < 0)
                log_warning("Failed to set ACL on %s, ignoring: %m", f->path);

finish:
        acl_free(acl);
#endif
}

static JournalFile* find_journal(Server *s, uid_t uid) {
        char *p;
        int r;
        JournalFile *f;
        sd_id128_t machine;

        assert(s);

        /* We split up user logs only on /var, not on /run. If the
         * runtime file is open, we write to it exclusively, in order
         * to guarantee proper order as soon as we flush /run to
         * /var and close the runtime file. */

        if (s->runtime_journal)
                return s->runtime_journal;

        if (uid <= 0)
                return s->system_journal;

        r = sd_id128_get_machine(&machine);
        if (r < 0)
                return s->system_journal;

        f = hashmap_get(s->user_journals, UINT32_TO_PTR(uid));
        if (f)
                return f;

        if (asprintf(&p, "/var/log/journal/" SD_ID128_FORMAT_STR "/user-%lu.journal",
                     SD_ID128_FORMAT_VAL(machine), (unsigned long) uid) < 0)
                return s->system_journal;

        while (hashmap_size(s->user_journals) >= USER_JOURNALS_MAX) {
                /* Too many open? Then let's close one */
                f = hashmap_steal_first(s->user_journals);
                assert(f);
                journal_file_close(f);
        }

        r = journal_file_open_reliably(p, O_RDWR|O_CREAT, 0640, s->compress, s->seal, &s->system_metrics, s->mmap, s->system_journal, &f);
        free(p);

        if (r < 0)
                return s->system_journal;

        server_fix_perms(s, f, uid);

        r = hashmap_put(s->user_journals, UINT32_TO_PTR(uid), f);
        if (r < 0) {
                journal_file_close(f);
                return s->system_journal;
        }

        return f;
}

void server_rotate(Server *s) {
        JournalFile *f;
        void *k;
        Iterator i;
        int r;

        log_debug("Rotating...");

        if (s->runtime_journal) {
                r = journal_file_rotate(&s->runtime_journal, s->compress, false);
                if (r < 0)
                        if (s->runtime_journal)
                                log_error("Failed to rotate %s: %s", s->runtime_journal->path, strerror(-r));
                        else
                                log_error("Failed to create new runtime journal: %s", strerror(-r));
                else
                        server_fix_perms(s, s->runtime_journal, 0);
        }

        if (s->system_journal) {
                r = journal_file_rotate(&s->system_journal, s->compress, s->seal);
                if (r < 0)
                        if (s->system_journal)
                                log_error("Failed to rotate %s: %s", s->system_journal->path, strerror(-r));
                        else
                                log_error("Failed to create new system journal: %s", strerror(-r));

                else
                        server_fix_perms(s, s->system_journal, 0);
        }

        HASHMAP_FOREACH_KEY(f, k, s->user_journals, i) {
                r = journal_file_rotate(&f, s->compress, s->seal);
                if (r < 0)
338
                        if (f)
339 340 341 342 343 344 345 346 347 348
                                log_error("Failed to rotate %s: %s", f->path, strerror(-r));
                        else
                                log_error("Failed to create user journal: %s", strerror(-r));
                else {
                        hashmap_replace(s->user_journals, k, f);
                        server_fix_perms(s, f, PTR_TO_UINT32(k));
                }
        }
}

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
void server_sync(Server *s) {
        JournalFile *f;
        void *k;
        Iterator i;
        int r;

        static const struct itimerspec sync_timer_disable = {};

        if (s->system_journal) {
                r = journal_file_set_offline(s->system_journal);
                if (r < 0)
                        log_error("Failed to sync system journal: %s", strerror(-r));
        }

        HASHMAP_FOREACH_KEY(f, k, s->user_journals, i) {
                r = journal_file_set_offline(f);
                if (r < 0)
                        log_error("Failed to sync user journal: %s", strerror(-r));
        }

        r = timerfd_settime(s->sync_timer_fd, 0, &sync_timer_disable, NULL);
        if (r < 0)
                log_error("Failed to disable max timer: %m");

        s->sync_scheduled = false;
}

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
void server_vacuum(Server *s) {
        char *p;
        char ids[33];
        sd_id128_t machine;
        int r;

        log_debug("Vacuuming...");

        s->oldest_file_usec = 0;

        r = sd_id128_get_machine(&machine);
        if (r < 0) {
                log_error("Failed to get machine ID: %s", strerror(-r));
                return;
        }

        sd_id128_to_string(machine, ids);

        if (s->system_journal) {
                p = strappend("/var/log/journal/", ids);
                if (!p) {
                        log_oom();
                        return;
                }

                r = journal_directory_vacuum(p, s->system_metrics.max_use, s->system_metrics.keep_free, s->max_retention_usec, &s->oldest_file_usec);
                if (r < 0 && r != -ENOENT)
                        log_error("Failed to vacuum %s: %s", p, strerror(-r));
                free(p);
        }

        if (s->runtime_journal) {
                p = strappend("/run/log/journal/", ids);
                if (!p) {
                        log_oom();
                        return;
                }

                r = journal_directory_vacuum(p, s->runtime_metrics.max_use, s->runtime_metrics.keep_free, s->max_retention_usec, &s->oldest_file_usec);
                if (r < 0 && r != -ENOENT)
                        log_error("Failed to vacuum %s: %s", p, strerror(-r));
                free(p);
        }

        s->cached_available_space_timestamp = 0;
}

bool shall_try_append_again(JournalFile *f, int r) {

        /* -E2BIG            Hit configured limit
           -EFBIG            Hit fs limit
           -EDQUOT           Quota limit hit
           -ENOSPC           Disk full
           -EHOSTDOWN        Other machine
           -EBUSY            Unclean shutdown
           -EPROTONOSUPPORT  Unsupported feature
           -EBADMSG          Corrupted
           -ENODATA          Truncated
           -ESHUTDOWN        Already archived */

        if (r == -E2BIG || r == -EFBIG || r == -EDQUOT || r == -ENOSPC)
                log_debug("%s: Allocation limit reached, rotating.", f->path);
        else if (r == -EHOSTDOWN)
                log_info("%s: Journal file from other machine, rotating.", f->path);
        else if (r == -EBUSY)
                log_info("%s: Unclean shutdown, rotating.", f->path);
        else if (r == -EPROTONOSUPPORT)
                log_info("%s: Unsupported feature, rotating.", f->path);
        else if (r == -EBADMSG || r == -ENODATA || r == ESHUTDOWN)
                log_warning("%s: Journal file corrupted, rotating.", f->path);
        else
                return false;

        return true;
}

static void write_to_journal(Server *s, uid_t uid, struct iovec *iovec, unsigned n) {
        JournalFile *f;
        bool vacuumed = false;
        int r;

        assert(s);
        assert(iovec);
        assert(n > 0);

        f = find_journal(s, uid);
        if (!f)
                return;

        if (journal_file_rotate_suggested(f, s->max_file_usec)) {
                log_debug("%s: Journal header limits reached or header out-of-date, rotating.", f->path);
                server_rotate(s);
                server_vacuum(s);
                vacuumed = true;

                f = find_journal(s, uid);
                if (!f)
                        return;
        }

        r = journal_file_append_entry(f, NULL, iovec, n, &s->seqnum, NULL, NULL);
477 478
        if (r >= 0) {
                server_schedule_sync(s);
479
                return;
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

        if (vacuumed || !shall_try_append_again(f, r)) {
                log_error("Failed to write entry, ignoring: %s", strerror(-r));
                return;
        }

        server_rotate(s);
        server_vacuum(s);

        f = find_journal(s, uid);
        if (!f)
                return;

        log_debug("Retrying write.");
        r = journal_file_append_entry(f, NULL, iovec, n, &s->seqnum, NULL, NULL);
        if (r < 0)
                log_error("Failed to write entry, ignoring: %s", strerror(-r));
}

static void dispatch_message_real(
                Server *s,
                struct iovec *iovec, unsigned n, unsigned m,
                struct ucred *ucred,
                struct timeval *tv,
                const char *label, size_t label_len,
                const char *unit_id) {

508 509
        char pid[sizeof("_PID=") + DECIMAL_STR_MAX(ucred->pid)],
                uid[sizeof("_UID=") + DECIMAL_STR_MAX(ucred->uid)],
510
                gid[sizeof("_GID=") + DECIMAL_STR_MAX(ucred->gid)],
511
                source_time[sizeof("_SOURCE_REALTIME_TIMESTAMP=") + DECIMAL_STR_MAX(usec_t)],
512 513
                boot_id[sizeof("_BOOT_ID=") + 32] = "_BOOT_ID=",
                machine_id[sizeof("_MACHINE_ID=") + 32] = "_MACHINE_ID=";
514

515
        char _cleanup_free_ *comm = NULL, *cmdline = NULL, *hostname = NULL,
516 517 518
                *exe = NULL, *cgroup = NULL, *session = NULL,
                *owner_uid = NULL, *unit = NULL, *selinux_context = NULL;

519 520 521 522
#ifdef HAVE_AUDIT
        char _cleanup_free_ *audit_session = NULL, *audit_loginuid = NULL;
#endif

523 524 525
        sd_id128_t id;
        int r;
        char *t;
526 527
        uid_t realuid = 0, owner = 0, journal_uid;
        bool owner_valid = false;
528 529 530 531 532 533 534

        assert(s);
        assert(iovec);
        assert(n > 0);
        assert(n + N_IOVEC_META_FIELDS <= m);

        if (ucred) {
535
#ifdef HAVE_AUDIT
536
                uint32_t audit;
537
                uid_t loginuid;
538
#endif
539 540 541

                realuid = ucred->uid;

542 543 544
                snprintf(pid, sizeof(pid) - 1, "_PID=%lu", (unsigned long) ucred->pid);
                char_array_0(pid);
                IOVEC_SET_STRING(iovec[n++], pid);
545

546 547 548
                snprintf(uid, sizeof(uid) - 1, "_UID=%lu", (unsigned long) ucred->uid);
                char_array_0(uid);
                IOVEC_SET_STRING(iovec[n++], uid);
549

550 551 552
                snprintf(gid, sizeof(gid) - 1, "_GID=%lu", (unsigned long) ucred->gid);
                char_array_0(gid);
                IOVEC_SET_STRING(iovec[n++], gid);
553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571

                r = get_process_comm(ucred->pid, &t);
                if (r >= 0) {
                        comm = strappend("_COMM=", t);
                        free(t);

                        if (comm)
                                IOVEC_SET_STRING(iovec[n++], comm);
                }

                r = get_process_exe(ucred->pid, &t);
                if (r >= 0) {
                        exe = strappend("_EXE=", t);
                        free(t);

                        if (exe)
                                IOVEC_SET_STRING(iovec[n++], exe);
                }

572
                r = get_process_cmdline(ucred->pid, 0, false, &t);
573 574 575 576 577 578 579 580
                if (r >= 0) {
                        cmdline = strappend("_CMDLINE=", t);
                        free(t);

                        if (cmdline)
                                IOVEC_SET_STRING(iovec[n++], cmdline);
                }

581
#ifdef HAVE_AUDIT
582 583 584 585 586 587
                r = audit_session_from_pid(ucred->pid, &audit);
                if (r >= 0)
                        if (asprintf(&audit_session, "_AUDIT_SESSION=%lu", (unsigned long) audit) >= 0)
                                IOVEC_SET_STRING(iovec[n++], audit_session);

                r = audit_loginuid_from_pid(ucred->pid, &loginuid);
588
                if (r >= 0)
589 590
                        if (asprintf(&audit_loginuid, "_AUDIT_LOGINUID=%lu", (unsigned long) loginuid) >= 0)
                                IOVEC_SET_STRING(iovec[n++], audit_loginuid);
591
#endif
592

593 594
                r = cg_pid_get_path(NULL, ucred->pid, &t);
                if (r >= 0) {
595 596 597 598 599 600 601 602
                        cgroup = strappend("_SYSTEMD_CGROUP=", t);
                        free(t);

                        if (cgroup)
                                IOVEC_SET_STRING(iovec[n++], cgroup);
                }

#ifdef HAVE_LOGIND
603 604
                r = cg_pid_get_session(ucred->pid, &t);
                if (r >= 0) {
605 606 607 608 609 610 611
                        session = strappend("_SYSTEMD_SESSION=", t);
                        free(t);

                        if (session)
                                IOVEC_SET_STRING(iovec[n++], session);
                }

612
                if (sd_pid_get_owner_uid(ucred->pid, &owner) >= 0) {
613
                        owner_valid = true;
614 615
                        if (asprintf(&owner_uid, "_SYSTEMD_OWNER_UID=%lu", (unsigned long) owner) >= 0)
                                IOVEC_SET_STRING(iovec[n++], owner_uid);
616
                }
617 618 619 620 621
#endif

                if (cg_pid_get_unit(ucred->pid, &t) >= 0) {
                        unit = strappend("_SYSTEMD_UNIT=", t);
                        free(t);
622 623 624 625 626 627 628 629 630
                } else if (cg_pid_get_user_unit(ucred->pid, &t) >= 0) {
                        unit = strappend("_SYSTEMD_USER_UNIT=", t);
                        free(t);
                } else if (unit_id) {
                        if (session)
                                unit = strappend("_SYSTEMD_USER_UNIT=", unit_id);
                        else
                                unit = strappend("_SYSTEMD_UNIT=", unit_id);
                }
631 632 633 634 635 636 637 638

                if (unit)
                        IOVEC_SET_STRING(iovec[n++], unit);

#ifdef HAVE_SELINUX
                if (label) {
                        selinux_context = malloc(sizeof("_SELINUX_CONTEXT=") + label_len);
                        if (selinux_context) {
639
                                *((char*) mempcpy(stpcpy(selinux_context, "_SELINUX_CONTEXT="), label, label_len)) = 0;
640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655
                                IOVEC_SET_STRING(iovec[n++], selinux_context);
                        }
                } else {
                        security_context_t con;

                        if (getpidcon(ucred->pid, &con) >= 0) {
                                selinux_context = strappend("_SELINUX_CONTEXT=", con);
                                if (selinux_context)
                                        IOVEC_SET_STRING(iovec[n++], selinux_context);
                                freecon(con);
                        }
                }
#endif
        }

        if (tv) {
656 657 658 659
                snprintf(source_time, sizeof(source_time) - 1, "_SOURCE_REALTIME_TIMESTAMP=%llu",
                         (unsigned long long) timeval_load(tv));
                char_array_0(source_time);
                IOVEC_SET_STRING(iovec[n++], source_time);
660 661 662 663 664 665
        }

        /* Note that strictly speaking storing the boot id here is
         * redundant since the entry includes this in-line
         * anyway. However, we need this indexed, too. */
        r = sd_id128_get_boot(&id);
666 667 668 669
        if (r >= 0) {
                sd_id128_to_string(id, boot_id + sizeof("_BOOT_ID=") - 1);
                IOVEC_SET_STRING(iovec[n++], boot_id);
        }
670 671

        r = sd_id128_get_machine(&id);
672
        if (r >= 0) {
673
                sd_id128_to_string(id, machine_id + sizeof("_MACHINE_ID=") - 1);
674 675
                IOVEC_SET_STRING(iovec[n++], machine_id);
        }
676 677 678 679 680 681 682 683 684 685 686

        t = gethostname_malloc();
        if (t) {
                hostname = strappend("_HOSTNAME=", t);
                free(t);
                if (hostname)
                        IOVEC_SET_STRING(iovec[n++], hostname);
        }

        assert(n <= m);

687
        if (s->split_mode == SPLIT_UID && realuid > 0)
688
                /* Split up strictly by any UID */
689
                journal_uid = realuid;
690
        else if (s->split_mode == SPLIT_LOGIN && realuid > 0 && owner_valid && owner > 0)
691 692 693
                /* Split up by login UIDs, this avoids creation of
                 * individual journals for system UIDs.  We do this
                 * only if the realuid is not root, in order not to
694 695 696
                 * accidentally leak privileged information to the
                 * user that is logged by a privileged process that is
                 * part of an unprivileged session.*/
697
                journal_uid = owner;
698 699
        else
                journal_uid = 0;
700 701

        write_to_journal(s, journal_uid, iovec, n);
702 703 704 705 706 707 708 709
}

void server_driver_message(Server *s, sd_id128_t message_id, const char *format, ...) {
        char mid[11 + 32 + 1];
        char buffer[16 + LINE_MAX + 1];
        struct iovec iovec[N_IOVEC_META_FIELDS + 4];
        int n = 0;
        va_list ap;
710
        struct ucred ucred = {};
711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746

        assert(s);
        assert(format);

        IOVEC_SET_STRING(iovec[n++], "PRIORITY=6");
        IOVEC_SET_STRING(iovec[n++], "_TRANSPORT=driver");

        memcpy(buffer, "MESSAGE=", 8);
        va_start(ap, format);
        vsnprintf(buffer + 8, sizeof(buffer) - 8, format, ap);
        va_end(ap);
        char_array_0(buffer);
        IOVEC_SET_STRING(iovec[n++], buffer);

        if (!sd_id128_equal(message_id, SD_ID128_NULL)) {
                snprintf(mid, sizeof(mid), MESSAGE_ID(message_id));
                char_array_0(mid);
                IOVEC_SET_STRING(iovec[n++], mid);
        }

        ucred.pid = getpid();
        ucred.uid = getuid();
        ucred.gid = getgid();

        dispatch_message_real(s, iovec, n, ELEMENTSOF(iovec), &ucred, NULL, NULL, 0, NULL);
}

void server_dispatch_message(
                Server *s,
                struct iovec *iovec, unsigned n, unsigned m,
                struct ucred *ucred,
                struct timeval *tv,
                const char *label, size_t label_len,
                const char *unit_id,
                int priority) {

747
        int rl, r;
748 749
        char _cleanup_free_ *path = NULL;
        char *c;
750 751 752 753 754 755 756 757 758 759 760 761 762

        assert(s);
        assert(iovec || n == 0);

        if (n == 0)
                return;

        if (LOG_PRI(priority) > s->max_level_store)
                return;

        if (!ucred)
                goto finish;

763 764
        r = cg_pid_get_path_shifted(ucred->pid, NULL, &path);
        if (r < 0)
765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782
                goto finish;

        /* example: /user/lennart/3/foobar
         *          /system/dbus.service/foobar
         *
         * So let's cut of everything past the third /, since that is
         * where user directories start */

        c = strchr(path, '/');
        if (c) {
                c = strchr(c+1, '/');
                if (c) {
                        c = strchr(c+1, '/');
                        if (c)
                                *c = 0;
                }
        }

783 784
        rl = journal_rate_limit_test(s->rate_limit, path,
                                     priority & LOG_PRIMASK, available_space(s));
785

786
        if (rl == 0)
787 788 789 790
                return;

        /* Write a suppression message if we suppressed something */
        if (rl > 1)
791 792
                server_driver_message(s, SD_MESSAGE_JOURNAL_DROPPED,
                                      "Suppressed %u messages from %s", rl - 1, path);
793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936

finish:
        dispatch_message_real(s, iovec, n, m, ucred, tv, label, label_len, unit_id);
}


static int system_journal_open(Server *s) {
        int r;
        char *fn;
        sd_id128_t machine;
        char ids[33];

        r = sd_id128_get_machine(&machine);
        if (r < 0)
                return r;

        sd_id128_to_string(machine, ids);

        if (!s->system_journal &&
            (s->storage == STORAGE_PERSISTENT || s->storage == STORAGE_AUTO) &&
            access("/run/systemd/journal/flushed", F_OK) >= 0) {

                /* If in auto mode: first try to create the machine
                 * path, but not the prefix.
                 *
                 * If in persistent mode: create /var/log/journal and
                 * the machine path */

                if (s->storage == STORAGE_PERSISTENT)
                        (void) mkdir("/var/log/journal/", 0755);

                fn = strappend("/var/log/journal/", ids);
                if (!fn)
                        return -ENOMEM;

                (void) mkdir(fn, 0755);
                free(fn);

                fn = strjoin("/var/log/journal/", ids, "/system.journal", NULL);
                if (!fn)
                        return -ENOMEM;

                r = journal_file_open_reliably(fn, O_RDWR|O_CREAT, 0640, s->compress, s->seal, &s->system_metrics, s->mmap, NULL, &s->system_journal);
                free(fn);

                if (r >= 0) {
                        char fb[FORMAT_BYTES_MAX];

                        server_fix_perms(s, s->system_journal, 0);
                        server_driver_message(s, SD_ID128_NULL, "Allowing system journal files to grow to %s.",
                                              format_bytes(fb, sizeof(fb), s->system_metrics.max_use));

                } else if (r < 0) {

                        if (r != -ENOENT && r != -EROFS)
                                log_warning("Failed to open system journal: %s", strerror(-r));

                        r = 0;
                }
        }

        if (!s->runtime_journal &&
            (s->storage != STORAGE_NONE)) {

                fn = strjoin("/run/log/journal/", ids, "/system.journal", NULL);
                if (!fn)
                        return -ENOMEM;

                if (s->system_journal) {

                        /* Try to open the runtime journal, but only
                         * if it already exists, so that we can flush
                         * it into the system journal */

                        r = journal_file_open(fn, O_RDWR, 0640, s->compress, false, &s->runtime_metrics, s->mmap, NULL, &s->runtime_journal);
                        free(fn);

                        if (r < 0) {
                                if (r != -ENOENT)
                                        log_warning("Failed to open runtime journal: %s", strerror(-r));

                                r = 0;
                        }

                } else {

                        /* OK, we really need the runtime journal, so create
                         * it if necessary. */

                        (void) mkdir_parents(fn, 0755);
                        r = journal_file_open_reliably(fn, O_RDWR|O_CREAT, 0640, s->compress, false, &s->runtime_metrics, s->mmap, NULL, &s->runtime_journal);
                        free(fn);

                        if (r < 0) {
                                log_error("Failed to open runtime journal: %s", strerror(-r));
                                return r;
                        }
                }

                if (s->runtime_journal) {
                        char fb[FORMAT_BYTES_MAX];

                        server_fix_perms(s, s->runtime_journal, 0);
                        server_driver_message(s, SD_ID128_NULL, "Allowing runtime journal files to grow to %s.",
                                              format_bytes(fb, sizeof(fb), s->runtime_metrics.max_use));
                }
        }

        return r;
}

int server_flush_to_var(Server *s) {
        int r;
        sd_id128_t machine;
        sd_journal *j = NULL;

        assert(s);

        if (s->storage != STORAGE_AUTO &&
            s->storage != STORAGE_PERSISTENT)
                return 0;

        if (!s->runtime_journal)
                return 0;

        system_journal_open(s);

        if (!s->system_journal)
                return 0;

        log_debug("Flushing to /var...");

        r = sd_id128_get_machine(&machine);
        if (r < 0) {
                log_error("Failed to get machine id: %s", strerror(-r));
                return r;
        }

        r = sd_journal_open(&j, SD_JOURNAL_RUNTIME_ONLY);
        if (r < 0) {
                log_error("Failed to read runtime journal: %s", strerror(-r));
                return r;
        }

937 938
        sd_journal_set_data_threshold(j, 0);

939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980
        SD_JOURNAL_FOREACH(j) {
                Object *o = NULL;
                JournalFile *f;

                f = j->current_file;
                assert(f && f->current_offset > 0);

                r = journal_file_move_to_object(f, OBJECT_ENTRY, f->current_offset, &o);
                if (r < 0) {
                        log_error("Can't read entry: %s", strerror(-r));
                        goto finish;
                }

                r = journal_file_copy_entry(f, s->system_journal, o, f->current_offset, NULL, NULL, NULL);
                if (r >= 0)
                        continue;

                if (!shall_try_append_again(s->system_journal, r)) {
                        log_error("Can't write entry: %s", strerror(-r));
                        goto finish;
                }

                server_rotate(s);
                server_vacuum(s);

                log_debug("Retrying write.");
                r = journal_file_copy_entry(f, s->system_journal, o, f->current_offset, NULL, NULL, NULL);
                if (r < 0) {
                        log_error("Can't write entry: %s", strerror(-r));
                        goto finish;
                }
        }

finish:
        journal_file_post_change(s->system_journal);

        journal_file_close(s->runtime_journal);
        s->runtime_journal = NULL;

        if (r >= 0)
                rm_rf("/run/log/journal", false, true, false);

981
        sd_journal_close(j);
982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013

        return r;
}

int process_event(Server *s, struct epoll_event *ev) {
        assert(s);
        assert(ev);

        if (ev->data.fd == s->signal_fd) {
                struct signalfd_siginfo sfsi;
                ssize_t n;

                if (ev->events != EPOLLIN) {
                        log_error("Got invalid event from epoll.");
                        return -EIO;
                }

                n = read(s->signal_fd, &sfsi, sizeof(sfsi));
                if (n != sizeof(sfsi)) {

                        if (n >= 0)
                                return -EIO;

                        if (errno == EINTR || errno == EAGAIN)
                                return 1;

                        return -errno;
                }

                if (sfsi.ssi_signo == SIGUSR1) {
                        touch("/run/systemd/journal/flushed");
                        server_flush_to_var(s);
1014
                        server_sync(s);
1015 1016 1017 1018 1019 1020 1021 1022 1023
                        return 1;
                }

                if (sfsi.ssi_signo == SIGUSR2) {
                        server_rotate(s);
                        server_vacuum(s);
                        return 1;
                }

1024 1025
                log_info("Received SIG%s", signal_to_string(sfsi.ssi_signo));

1026 1027
                return 0;

1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040
        } else if (ev->data.fd == s->sync_timer_fd) {
                int r;
                uint64_t t;

                log_debug("Got sync request from epoll.");

                r = read(ev->data.fd, (void *)&t, sizeof(t));
                if (r < 0)
                        return 0;

                server_sync(s);
                return 1;

1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249
        } else if (ev->data.fd == s->dev_kmsg_fd) {
                int r;

                if (ev->events != EPOLLIN) {
                        log_error("Got invalid event from epoll.");
                        return -EIO;
                }

                r = server_read_dev_kmsg(s);
                if (r < 0)
                        return r;

                return 1;

        } else if (ev->data.fd == s->native_fd ||
                   ev->data.fd == s->syslog_fd) {

                if (ev->events != EPOLLIN) {
                        log_error("Got invalid event from epoll.");
                        return -EIO;
                }

                for (;;) {
                        struct msghdr msghdr;
                        struct iovec iovec;
                        struct ucred *ucred = NULL;
                        struct timeval *tv = NULL;
                        struct cmsghdr *cmsg;
                        char *label = NULL;
                        size_t label_len = 0;
                        union {
                                struct cmsghdr cmsghdr;

                                /* We use NAME_MAX space for the
                                 * SELinux label here. The kernel
                                 * currently enforces no limit, but
                                 * according to suggestions from the
                                 * SELinux people this will change and
                                 * it will probably be identical to
                                 * NAME_MAX. For now we use that, but
                                 * this should be updated one day when
                                 * the final limit is known.*/
                                uint8_t buf[CMSG_SPACE(sizeof(struct ucred)) +
                                            CMSG_SPACE(sizeof(struct timeval)) +
                                            CMSG_SPACE(sizeof(int)) + /* fd */
                                            CMSG_SPACE(NAME_MAX)]; /* selinux label */
                        } control;
                        ssize_t n;
                        int v;
                        int *fds = NULL;
                        unsigned n_fds = 0;

                        if (ioctl(ev->data.fd, SIOCINQ, &v) < 0) {
                                log_error("SIOCINQ failed: %m");
                                return -errno;
                        }

                        if (s->buffer_size < (size_t) v) {
                                void *b;
                                size_t l;

                                l = MAX(LINE_MAX + (size_t) v, s->buffer_size * 2);
                                b = realloc(s->buffer, l+1);

                                if (!b) {
                                        log_error("Couldn't increase buffer.");
                                        return -ENOMEM;
                                }

                                s->buffer_size = l;
                                s->buffer = b;
                        }

                        zero(iovec);
                        iovec.iov_base = s->buffer;
                        iovec.iov_len = s->buffer_size;

                        zero(control);
                        zero(msghdr);
                        msghdr.msg_iov = &iovec;
                        msghdr.msg_iovlen = 1;
                        msghdr.msg_control = &control;
                        msghdr.msg_controllen = sizeof(control);

                        n = recvmsg(ev->data.fd, &msghdr, MSG_DONTWAIT|MSG_CMSG_CLOEXEC);
                        if (n < 0) {

                                if (errno == EINTR || errno == EAGAIN)
                                        return 1;

                                log_error("recvmsg() failed: %m");
                                return -errno;
                        }

                        for (cmsg = CMSG_FIRSTHDR(&msghdr); cmsg; cmsg = CMSG_NXTHDR(&msghdr, cmsg)) {

                                if (cmsg->cmsg_level == SOL_SOCKET &&
                                    cmsg->cmsg_type == SCM_CREDENTIALS &&
                                    cmsg->cmsg_len == CMSG_LEN(sizeof(struct ucred)))
                                        ucred = (struct ucred*) CMSG_DATA(cmsg);
                                else if (cmsg->cmsg_level == SOL_SOCKET &&
                                         cmsg->cmsg_type == SCM_SECURITY) {
                                        label = (char*) CMSG_DATA(cmsg);
                                        label_len = cmsg->cmsg_len - CMSG_LEN(0);
                                } else if (cmsg->cmsg_level == SOL_SOCKET &&
                                         cmsg->cmsg_type == SO_TIMESTAMP &&
                                         cmsg->cmsg_len == CMSG_LEN(sizeof(struct timeval)))
                                        tv = (struct timeval*) CMSG_DATA(cmsg);
                                else if (cmsg->cmsg_level == SOL_SOCKET &&
                                         cmsg->cmsg_type == SCM_RIGHTS) {
                                        fds = (int*) CMSG_DATA(cmsg);
                                        n_fds = (cmsg->cmsg_len - CMSG_LEN(0)) / sizeof(int);
                                }
                        }

                        if (ev->data.fd == s->syslog_fd) {
                                char *e;

                                if (n > 0 && n_fds == 0) {
                                        e = memchr(s->buffer, '\n', n);
                                        if (e)
                                                *e = 0;
                                        else
                                                s->buffer[n] = 0;

                                        server_process_syslog_message(s, strstrip(s->buffer), ucred, tv, label, label_len);
                                } else if (n_fds > 0)
                                        log_warning("Got file descriptors via syslog socket. Ignoring.");

                        } else {
                                if (n > 0 && n_fds == 0)
                                        server_process_native_message(s, s->buffer, n, ucred, tv, label, label_len);
                                else if (n == 0 && n_fds == 1)
                                        server_process_native_file(s, fds[0], ucred, tv, label, label_len);
                                else if (n_fds > 0)
                                        log_warning("Got too many file descriptors via native socket. Ignoring.");
                        }

                        close_many(fds, n_fds);
                }

                return 1;

        } else if (ev->data.fd == s->stdout_fd) {

                if (ev->events != EPOLLIN) {
                        log_error("Got invalid event from epoll.");
                        return -EIO;
                }

                stdout_stream_new(s);
                return 1;

        } else {
                StdoutStream *stream;

                if ((ev->events|EPOLLIN|EPOLLHUP) != (EPOLLIN|EPOLLHUP)) {
                        log_error("Got invalid event from epoll.");
                        return -EIO;
                }

                /* If it is none of the well-known fds, it must be an
                 * stdout stream fd. Note that this is a bit ugly here
                 * (since we rely that none of the well-known fds
                 * could be interpreted as pointer), but nonetheless
                 * safe, since the well-known fds would never get an
                 * fd > 4096, i.e. beyond the first memory page */

                stream = ev->data.ptr;

                if (stdout_stream_process(stream) <= 0)
                        stdout_stream_free(stream);

                return 1;
        }

        log_error("Unknown event.");
        return 0;
}

static int open_signalfd(Server *s) {
        sigset_t mask;
        struct epoll_event ev;

        assert(s);

        assert_se(sigemptyset(&mask) == 0);
        sigset_add_many(&mask, SIGINT, SIGTERM, SIGUSR1, SIGUSR2, -1);
        assert_se(sigprocmask(SIG_SETMASK, &mask, NULL) == 0);

        s->signal_fd = signalfd(-1, &mask, SFD_NONBLOCK|SFD_CLOEXEC);
        if (s->signal_fd < 0) {
                log_error("signalfd(): %m");
                return -errno;
        }

        zero(ev);
        ev.events = EPOLLIN;
        ev.data.fd = s->signal_fd;

        if (epoll_ctl(s->epoll_fd, EPOLL_CTL_ADD, s->signal_fd, &ev) < 0) {
                log_error("epoll_ctl(): %m");
                return -errno;
        }

        return 0;
}

static int server_parse_proc_cmdline(Server *s) {
1250 1251
        char _cleanup_free_ *line = NULL;
        char *w, *state;
1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264
        int r;
        size_t l;

        if (detect_container(NULL) > 0)
                return 0;

        r = read_one_line_file("/proc/cmdline", &line);
        if (r < 0) {
                log_warning("Failed to read /proc/cmdline, ignoring: %s", strerror(-r));
                return 0;
        }

        FOREACH_WORD_QUOTED(w, l, line, state) {
1265
                char _cleanup_free_ *word;
1266 1267

                word = strndup(w, l);
1268 1269
                if (!word)
                        return -ENOMEM;
1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292

                if (startswith(word, "systemd.journald.forward_to_syslog=")) {
                        r = parse_boolean(word + 35);
                        if (r < 0)
                                log_warning("Failed to parse forward to syslog switch %s. Ignoring.", word + 35);
                        else
                                s->forward_to_syslog = r;
                } else if (startswith(word, "systemd.journald.forward_to_kmsg=")) {
                        r = parse_boolean(word + 33);
                        if (r < 0)
                                log_warning("Failed to parse forward to kmsg switch %s. Ignoring.", word + 33);
                        else
                                s->forward_to_kmsg = r;
                } else if (startswith(word, "systemd.journald.forward_to_console=")) {
                        r = parse_boolean(word + 36);
                        if (r < 0)
                                log_warning("Failed to parse forward to console switch %s. Ignoring.", word + 36);
                        else
                                s->forward_to_console = r;
                } else if (startswith(word, "systemd.journald"))
                        log_warning("Invalid systemd.journald parameter. Ignoring.");
        }

1293
        return 0;
1294 1295 1296
}

static int server_parse_config_file(Server *s) {
1297 1298
        static const char *fn = "/etc/systemd/journald.conf";
        FILE _cleanup_fclose_ *f = NULL;
1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311
        int r;

        assert(s);

        f = fopen(fn, "re");
        if (!f) {
                if (errno == ENOENT)
                        return 0;

                log_warning("Failed to open configuration file %s: %m", fn);
                return -errno;
        }

1312
        r = config_parse(NULL, fn, f, "Journal\0", config_item_perf_lookup,
1313
                         (void*) journald_gperf_lookup, false, s);
1314 1315 1316 1317 1318 1319
        if (r < 0)
                log_warning("Failed to parse configuration file: %s", strerror(-r));

        return r;
}

1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351
static int server_open_sync_timer(Server *s) {
        int r;
        struct epoll_event ev;

        assert(s);

        s->sync_timer_fd = timerfd_create(CLOCK_MONOTONIC, TFD_CLOEXEC);
        if (s->sync_timer_fd < 0)
                return -errno;

        zero(ev);
        ev.events = EPOLLIN;
        ev.data.fd = s->sync_timer_fd;

        r = epoll_ctl(s->epoll_fd, EPOLL_CTL_ADD, s->sync_timer_fd, &ev);
        if (r < 0) {
                log_error("Failed to add idle timer fd to epoll object: %m");
                return -errno;
        }

        return 0;
}

int server_schedule_sync(Server *s) {
        int r;

        assert(s);

        if (s->sync_scheduled)
                return 0;

        if (s->sync_interval_usec) {
1352 1353 1354 1355
                struct itimerspec sync_timer_enable = {
                        .it_value.tv_sec = s->sync_interval_usec / USEC_PER_SEC,
                        .it_value.tv_nsec = s->sync_interval_usec % MSEC_PER_SEC,
                };
1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366

                r = timerfd_settime(s->sync_timer_fd, 0, &sync_timer_enable, NULL);
                if (r < 0)
                        return -errno;
        }

        s->sync_scheduled = true;

        return 0;
}

1367 1368 1369 1370 1371 1372
int server_init(Server *s) {
        int n, r, fd;

        assert(s);

        zero(*s);
1373 1374
        s->sync_timer_fd = s->syslog_fd = s->native_fd = s->stdout_fd =
            s->signal_fd = s->epoll_fd = s->dev_kmsg_fd = -1;
1375 1376 1377
        s->compress = true;
        s->seal = true;

1378 1379 1380
        s->sync_interval_usec = DEFAULT_SYNC_INTERVAL_USEC;
        s->sync_scheduled = false;

1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395
        s->rate_limit_interval = DEFAULT_RATE_LIMIT_INTERVAL;
        s->rate_limit_burst = DEFAULT_RATE_LIMIT_BURST;

        s->forward_to_syslog = true;

        s->max_level_store = LOG_DEBUG;
        s->max_level_syslog = LOG_DEBUG;
        s->max_level_kmsg = LOG_NOTICE;
        s->max_level_console = LOG_INFO;

        memset(&s->system_metrics, 0xFF, sizeof(s->system_metrics));
        memset(&s->runtime_metrics, 0xFF, sizeof(s->runtime_metrics));

        server_parse_config_file(s);
        server_parse_proc_cmdline(s);
1396 1397 1398 1399 1400 1401
        if (!!s->rate_limit_interval ^ !!s->rate_limit_burst) {
                log_debug("Setting both rate limit interval and burst from %llu,%u to 0,0",
                          (long long unsigned) s->rate_limit_interval,
                          s->rate_limit_burst);
                s->rate_limit_interval = s->rate_limit_burst = 0;
        }
1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479

        mkdir_p("/run/systemd/journal", 0755);

        s->user_journals = hashmap_new(trivial_hash_func, trivial_compare_func);
        if (!s->user_journals)
                return log_oom();

        s->mmap = mmap_cache_new();
        if (!s->mmap)
                return log_oom();

        s->epoll_fd = epoll_create1(EPOLL_CLOEXEC);
        if (s->epoll_fd < 0) {
                log_error("Failed to create epoll object: %m");
                return -errno;
        }

        n = sd_listen_fds(true);
        if (n < 0) {
                log_error("Failed to read listening file descriptors from environment: %s", strerror(-n));
                return n;
        }

        for (fd = SD_LISTEN_FDS_START; fd < SD_LISTEN_FDS_START + n; fd++) {

                if (sd_is_socket_unix(fd, SOCK_DGRAM, -1, "/run/systemd/journal/socket", 0) > 0) {

                        if (s->native_fd >= 0) {
                                log_error("Too many native sockets passed.");
                                return -EINVAL;
                        }

                        s->native_fd = fd;

                } else if (sd_is_socket_unix(fd, SOCK_STREAM, 1, "/run/systemd/journal/stdout", 0) > 0) {

                        if (s->stdout_fd >= 0) {
                                log_error("Too many stdout sockets passed.");
                                return -EINVAL;
                        }

                        s->stdout_fd = fd;

                } else if (sd_is_socket_unix(fd, SOCK_DGRAM, -1, "/dev/log", 0) > 0) {

                        if (s->syslog_fd >= 0) {
                                log_error("Too many /dev/log sockets passed.");
                                return -EINVAL;
                        }

                        s->syslog_fd = fd;

                } else {
                        log_error("Unknown socket passed.");
                        return -EINVAL;
                }
        }

        r = server_open_syslog_socket(s);
        if (r < 0)
                return r;

        r = server_open_native_socket(s);
        if (r < 0)
                return r;

        r = server_open_stdout_socket(s);
        if (r < 0)
                return r;

        r = server_open_dev_kmsg(s);
        if (r < 0)
                return r;

        r = server_open_kernel_seqnum(s);
        if (r < 0)
                return r;

1480 1481 1482 1483
        r = server_open_sync_timer(s);
        if (r < 0)
                return r;

1484 1485 1486 1487 1488 1489 1490 1491
        r = open_signalfd(s);
        if (r < 0)
                return r;

        s->udev = udev_new();
        if (!s->udev)
                return -ENOMEM;

1492 1493
        s->rate_limit = journal_rate_limit_new(s->rate_limit_interval,
                                               s->rate_limit_burst);
1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555
        if (!s->rate_limit)
                return -ENOMEM;

        r = system_journal_open(s);
        if (r < 0)
                return r;

        return 0;
}

void server_maybe_append_tags(Server *s) {
#ifdef HAVE_GCRYPT
        JournalFile *f;
        Iterator i;
        usec_t n;

        n = now(CLOCK_REALTIME);

        if (s->system_journal)
                journal_file_maybe_append_tag(s->system_journal, n);

        HASHMAP_FOREACH(f, s->user_journals, i)
                journal_file_maybe_append_tag(f, n);
#endif
}

void server_done(Server *s) {
        JournalFile *f;
        assert(s);

        while (s->stdout_streams)
                stdout_stream_free(s->stdout_streams);

        if (s->system_journal)
                journal_file_close(s->system_journal);

        if (s->runtime_journal)
                journal_file_close(s->runtime_journal);

        while ((f = hashmap_steal_first(s->user_journals)))
                journal_file_close(f);

        hashmap_free(s->user_journals);

        if (s->epoll_fd >= 0)
                close_nointr_nofail(s->epoll_fd);

        if (s->signal_fd >= 0)
                close_nointr_nofail(s->signal_fd);

        if (s->syslog_fd >= 0)
                close_nointr_nofail(s->syslog_fd);

        if (s->native_fd >= 0)
                close_nointr_nofail(s->native_fd);

        if (s->stdout_fd >= 0)
                close_nointr_nofail(s->stdout_fd);

        if (s->dev_kmsg_fd >= 0)
                close_nointr_nofail(s->dev_kmsg_fd);

1556 1557 1558
        if (s->sync_timer_fd >= 0)
                close_nointr_nofail(s->sync_timer_fd);

1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573
        if (s->rate_limit)
                journal_rate_limit_free(s->rate_limit);

        if (s->kernel_seqnum)
                munmap(s->kernel_seqnum, sizeof(uint64_t));

        free(s->buffer);
        free(s->tty_path);

        if (s->mmap)
                mmap_cache_unref(s->mmap);

        if (s->udev)
                udev_unref(s->udev);
}