| /*-*- 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 <errno.h> |
| #include <fcntl.h> |
| #include <stddef.h> |
| #include <unistd.h> |
| #include <sys/inotify.h> |
| #include <sys/poll.h> |
| #include <sys/vfs.h> |
| #include <linux/magic.h> |
| |
| #include "sd-journal.h" |
| #include "journal-def.h" |
| #include "journal-file.h" |
| #include "hashmap.h" |
| #include "list.h" |
| #include "path-util.h" |
| #include "lookup3.h" |
| #include "compress.h" |
| #include "journal-internal.h" |
| #include "missing.h" |
| #include "catalog.h" |
| #include "replace-var.h" |
| |
| #define JOURNAL_FILES_MAX 1024 |
| |
| #define JOURNAL_FILES_RECHECK_USEC (2 * USEC_PER_SEC) |
| |
| #define REPLACE_VAR_MAX 256 |
| |
| #define DEFAULT_DATA_THRESHOLD (64*1024) |
| |
| /* We return an error here only if we didn't manage to |
| memorize the real error. */ |
| static int set_put_error(sd_journal *j, int r) { |
| int k; |
| |
| if (r >= 0) |
| return r; |
| |
| k = set_ensure_allocated(&j->errors, trivial_hash_func, trivial_compare_func); |
| if (k < 0) |
| return k; |
| |
| return set_put(j->errors, INT_TO_PTR(r)); |
| } |
| |
| static void detach_location(sd_journal *j) { |
| Iterator i; |
| JournalFile *f; |
| |
| assert(j); |
| |
| j->current_file = NULL; |
| j->current_field = 0; |
| |
| HASHMAP_FOREACH(f, j->files, i) |
| f->current_offset = 0; |
| } |
| |
| static void reset_location(sd_journal *j) { |
| assert(j); |
| |
| detach_location(j); |
| zero(j->current_location); |
| } |
| |
| static void init_location(Location *l, LocationType type, JournalFile *f, Object *o) { |
| assert(l); |
| assert(type == LOCATION_DISCRETE || type == LOCATION_SEEK); |
| assert(f); |
| assert(o->object.type == OBJECT_ENTRY); |
| |
| l->type = type; |
| l->seqnum = le64toh(o->entry.seqnum); |
| l->seqnum_id = f->header->seqnum_id; |
| l->realtime = le64toh(o->entry.realtime); |
| l->monotonic = le64toh(o->entry.monotonic); |
| l->boot_id = o->entry.boot_id; |
| l->xor_hash = le64toh(o->entry.xor_hash); |
| |
| l->seqnum_set = l->realtime_set = l->monotonic_set = l->xor_hash_set = true; |
| } |
| |
| static void set_location(sd_journal *j, LocationType type, JournalFile *f, Object *o, uint64_t offset) { |
| assert(j); |
| assert(type == LOCATION_DISCRETE || type == LOCATION_SEEK); |
| assert(f); |
| assert(o); |
| |
| init_location(&j->current_location, type, f, o); |
| |
| j->current_file = f; |
| j->current_field = 0; |
| |
| f->current_offset = offset; |
| } |
| |
| static int match_is_valid(const void *data, size_t size) { |
| const char *b, *p; |
| |
| assert(data); |
| |
| if (size < 2) |
| return false; |
| |
| if (startswith(data, "__")) |
| return false; |
| |
| b = data; |
| for (p = b; p < b + size; p++) { |
| |
| if (*p == '=') |
| return p > b; |
| |
| if (*p == '_') |
| continue; |
| |
| if (*p >= 'A' && *p <= 'Z') |
| continue; |
| |
| if (*p >= '0' && *p <= '9') |
| continue; |
| |
| return false; |
| } |
| |
| return false; |
| } |
| |
| static bool same_field(const void *_a, size_t s, const void *_b, size_t t) { |
| const uint8_t *a = _a, *b = _b; |
| size_t j; |
| |
| for (j = 0; j < s && j < t; j++) { |
| |
| if (a[j] != b[j]) |
| return false; |
| |
| if (a[j] == '=') |
| return true; |
| } |
| |
| return true; |
| } |
| |
| static Match *match_new(Match *p, MatchType t) { |
| Match *m; |
| |
| m = new0(Match, 1); |
| if (!m) |
| return NULL; |
| |
| m->type = t; |
| |
| if (p) { |
| m->parent = p; |
| LIST_PREPEND(Match, matches, p->matches, m); |
| } |
| |
| return m; |
| } |
| |
| static void match_free(Match *m) { |
| assert(m); |
| |
| while (m->matches) |
| match_free(m->matches); |
| |
| if (m->parent) |
| LIST_REMOVE(Match, matches, m->parent->matches, m); |
| |
| free(m->data); |
| free(m); |
| } |
| |
| static void match_free_if_empty(Match *m) { |
| assert(m); |
| |
| if (m->matches) |
| return; |
| |
| match_free(m); |
| } |
| |
| _public_ int sd_journal_add_match(sd_journal *j, const void *data, size_t size) { |
| Match *l2, *l3, *add_here = NULL, *m; |
| le64_t le_hash; |
| |
| if (!j) |
| return -EINVAL; |
| |
| if (!data) |
| return -EINVAL; |
| |
| if (size == 0) |
| size = strlen(data); |
| |
| if (!match_is_valid(data, size)) |
| return -EINVAL; |
| |
| /* level 0: OR term |
| * level 1: AND terms |
| * level 2: OR terms |
| * level 3: concrete matches */ |
| |
| if (!j->level0) { |
| j->level0 = match_new(NULL, MATCH_OR_TERM); |
| if (!j->level0) |
| return -ENOMEM; |
| } |
| |
| if (!j->level1) { |
| j->level1 = match_new(j->level0, MATCH_AND_TERM); |
| if (!j->level1) |
| return -ENOMEM; |
| } |
| |
| assert(j->level0->type == MATCH_OR_TERM); |
| assert(j->level1->type == MATCH_AND_TERM); |
| |
| le_hash = htole64(hash64(data, size)); |
| |
| LIST_FOREACH(matches, l2, j->level1->matches) { |
| assert(l2->type == MATCH_OR_TERM); |
| |
| LIST_FOREACH(matches, l3, l2->matches) { |
| assert(l3->type == MATCH_DISCRETE); |
| |
| /* Exactly the same match already? Then ignore |
| * this addition */ |
| if (l3->le_hash == le_hash && |
| l3->size == size && |
| memcmp(l3->data, data, size) == 0) |
| return 0; |
| |
| /* Same field? Then let's add this to this OR term */ |
| if (same_field(data, size, l3->data, l3->size)) { |
| add_here = l2; |
| break; |
| } |
| } |
| |
| if (add_here) |
| break; |
| } |
| |
| if (!add_here) { |
| add_here = match_new(j->level1, MATCH_OR_TERM); |
| if (!add_here) |
| goto fail; |
| } |
| |
| m = match_new(add_here, MATCH_DISCRETE); |
| if (!m) |
| goto fail; |
| |
| m->le_hash = le_hash; |
| m->size = size; |
| m->data = memdup(data, size); |
| if (!m->data) |
| goto fail; |
| |
| detach_location(j); |
| |
| return 0; |
| |
| fail: |
| if (add_here) |
| match_free_if_empty(add_here); |
| |
| if (j->level1) |
| match_free_if_empty(j->level1); |
| |
| if (j->level0) |
| match_free_if_empty(j->level0); |
| |
| return -ENOMEM; |
| } |
| |
| _public_ int sd_journal_add_disjunction(sd_journal *j) { |
| Match *m; |
| |
| assert(j); |
| |
| if (!j->level0) |
| return 0; |
| |
| if (!j->level1) |
| return 0; |
| |
| if (!j->level1->matches) |
| return 0; |
| |
| m = match_new(j->level0, MATCH_AND_TERM); |
| if (!m) |
| return -ENOMEM; |
| |
| j->level1 = m; |
| return 0; |
| } |
| |
| static char *match_make_string(Match *m) { |
| char *p, *r; |
| Match *i; |
| bool enclose = false; |
| |
| if (!m) |
| return strdup(""); |
| |
| if (m->type == MATCH_DISCRETE) |
| return strndup(m->data, m->size); |
| |
| p = NULL; |
| LIST_FOREACH(matches, i, m->matches) { |
| char *t, *k; |
| |
| t = match_make_string(i); |
| if (!t) { |
| free(p); |
| return NULL; |
| } |
| |
| if (p) { |
| k = strjoin(p, m->type == MATCH_OR_TERM ? " OR " : " AND ", t, NULL); |
| free(p); |
| free(t); |
| |
| if (!k) |
| return NULL; |
| |
| p = k; |
| |
| enclose = true; |
| } else { |
| free(p); |
| p = t; |
| } |
| } |
| |
| if (enclose) { |
| r = strjoin("(", p, ")", NULL); |
| free(p); |
| return r; |
| } |
| |
| return p; |
| } |
| |
| char *journal_make_match_string(sd_journal *j) { |
| assert(j); |
| |
| return match_make_string(j->level0); |
| } |
| |
| _public_ void sd_journal_flush_matches(sd_journal *j) { |
| |
| if (!j) |
| return; |
| |
| if (j->level0) |
| match_free(j->level0); |
| |
| j->level0 = j->level1 = NULL; |
| |
| detach_location(j); |
| } |
| |
| static int compare_entry_order(JournalFile *af, Object *_ao, |
| JournalFile *bf, uint64_t bp) { |
| |
| uint64_t a, b; |
| Object *ao, *bo; |
| int r; |
| |
| assert(af); |
| assert(bf); |
| assert(_ao); |
| |
| /* The mmap cache might invalidate the object from the first |
| * file if we look at the one from the second file. Hence |
| * temporarily copy the header of the first one, and look at |
| * that only. */ |
| ao = alloca(offsetof(EntryObject, items)); |
| memcpy(ao, _ao, offsetof(EntryObject, items)); |
| |
| r = journal_file_move_to_object(bf, OBJECT_ENTRY, bp, &bo); |
| if (r < 0) |
| return strcmp(af->path, bf->path); |
| |
| /* We operate on two different files here, hence we can access |
| * two objects at the same time, which we normally can't. |
| * |
| * If contents and timestamps match, these entries are |
| * identical, even if the seqnum does not match */ |
| |
| if (sd_id128_equal(ao->entry.boot_id, bo->entry.boot_id) && |
| ao->entry.monotonic == bo->entry.monotonic && |
| ao->entry.realtime == bo->entry.realtime && |
| ao->entry.xor_hash == bo->entry.xor_hash) |
| return 0; |
| |
| if (sd_id128_equal(af->header->seqnum_id, bf->header->seqnum_id)) { |
| |
| /* If this is from the same seqnum source, compare |
| * seqnums */ |
| a = le64toh(ao->entry.seqnum); |
| b = le64toh(bo->entry.seqnum); |
| |
| if (a < b) |
| return -1; |
| if (a > b) |
| return 1; |
| |
| /* Wow! This is weird, different data but the same |
| * seqnums? Something is borked, but let's make the |
| * best of it and compare by time. */ |
| } |
| |
| if (sd_id128_equal(ao->entry.boot_id, bo->entry.boot_id)) { |
| |
| /* If the boot id matches compare monotonic time */ |
| a = le64toh(ao->entry.monotonic); |
| b = le64toh(bo->entry.monotonic); |
| |
| if (a < b) |
| return -1; |
| if (a > b) |
| return 1; |
| } |
| |
| /* Otherwise compare UTC time */ |
| a = le64toh(ao->entry.realtime); |
| b = le64toh(bo->entry.realtime); |
| |
| if (a < b) |
| return -1; |
| if (a > b) |
| return 1; |
| |
| /* Finally, compare by contents */ |
| a = le64toh(ao->entry.xor_hash); |
| b = le64toh(bo->entry.xor_hash); |
| |
| if (a < b) |
| return -1; |
| if (a > b) |
| return 1; |
| |
| return 0; |
| } |
| |
| static int compare_with_location(JournalFile *af, Object *ao, Location *l) { |
| uint64_t a; |
| |
| assert(af); |
| assert(ao); |
| assert(l); |
| assert(l->type == LOCATION_DISCRETE || l->type == LOCATION_SEEK); |
| |
| if (l->monotonic_set && |
| sd_id128_equal(ao->entry.boot_id, l->boot_id) && |
| l->realtime_set && |
| le64toh(ao->entry.realtime) == l->realtime && |
| l->xor_hash_set && |
| le64toh(ao->entry.xor_hash) == l->xor_hash) |
| return 0; |
| |
| if (l->seqnum_set && |
| sd_id128_equal(af->header->seqnum_id, l->seqnum_id)) { |
| |
| a = le64toh(ao->entry.seqnum); |
| |
| if (a < l->seqnum) |
| return -1; |
| if (a > l->seqnum) |
| return 1; |
| } |
| |
| if (l->monotonic_set && |
| sd_id128_equal(ao->entry.boot_id, l->boot_id)) { |
| |
| a = le64toh(ao->entry.monotonic); |
| |
| if (a < l->monotonic) |
| return -1; |
| if (a > l->monotonic) |
| return 1; |
| } |
| |
| if (l->realtime_set) { |
| |
| a = le64toh(ao->entry.realtime); |
| |
| if (a < l->realtime) |
| return -1; |
| if (a > l->realtime) |
| return 1; |
| } |
| |
| if (l->xor_hash_set) { |
| a = le64toh(ao->entry.xor_hash); |
| |
| if (a < l->xor_hash) |
| return -1; |
| if (a > l->xor_hash) |
| return 1; |
| } |
| |
| return 0; |
| } |
| |
| static int next_for_match( |
| sd_journal *j, |
| Match *m, |
| JournalFile *f, |
| uint64_t after_offset, |
| direction_t direction, |
| Object **ret, |
| uint64_t *offset) { |
| |
| int r; |
| uint64_t np = 0; |
| Object *n; |
| |
| assert(j); |
| assert(m); |
| assert(f); |
| |
| if (m->type == MATCH_DISCRETE) { |
| uint64_t dp; |
| |
| r = journal_file_find_data_object_with_hash(f, m->data, m->size, le64toh(m->le_hash), NULL, &dp); |
| if (r <= 0) |
| return r; |
| |
| return journal_file_move_to_entry_by_offset_for_data(f, dp, after_offset, direction, ret, offset); |
| |
| } else if (m->type == MATCH_OR_TERM) { |
| Match *i; |
| |
| /* Find the earliest match beyond after_offset */ |
| |
| LIST_FOREACH(matches, i, m->matches) { |
| uint64_t cp; |
| |
| r = next_for_match(j, i, f, after_offset, direction, NULL, &cp); |
| if (r < 0) |
| return r; |
| else if (r > 0) { |
| if (np == 0 || (direction == DIRECTION_DOWN ? np > cp : np < cp)) |
| np = cp; |
| } |
| } |
| |
| } else if (m->type == MATCH_AND_TERM) { |
| Match *i; |
| bool continue_looking; |
| |
| /* Always jump to the next matching entry and repeat |
| * this until we fine and offset that matches for all |
| * matches. */ |
| |
| if (!m->matches) |
| return 0; |
| |
| np = 0; |
| do { |
| continue_looking = false; |
| |
| LIST_FOREACH(matches, i, m->matches) { |
| uint64_t cp, limit; |
| |
| if (np == 0) |
| limit = after_offset; |
| else if (direction == DIRECTION_DOWN) |
| limit = MAX(np, after_offset); |
| else |
| limit = MIN(np, after_offset); |
| |
| r = next_for_match(j, i, f, limit, direction, NULL, &cp); |
| if (r <= 0) |
| return r; |
| |
| if ((direction == DIRECTION_DOWN ? cp >= after_offset : cp <= after_offset) && |
| (np == 0 || (direction == DIRECTION_DOWN ? cp > np : cp < np))) { |
| np = cp; |
| continue_looking = true; |
| } |
| } |
| |
| } while (continue_looking); |
| } |
| |
| if (np == 0) |
| return 0; |
| |
| r = journal_file_move_to_object(f, OBJECT_ENTRY, np, &n); |
| if (r < 0) |
| return r; |
| |
| if (ret) |
| *ret = n; |
| if (offset) |
| *offset = np; |
| |
| return 1; |
| } |
| |
| static int find_location_for_match( |
| sd_journal *j, |
| Match *m, |
| JournalFile *f, |
| direction_t direction, |
| Object **ret, |
| uint64_t *offset) { |
| |
| int r; |
| |
| assert(j); |
| assert(m); |
| assert(f); |
| |
| if (m->type == MATCH_DISCRETE) { |
| uint64_t dp; |
| |
| r = journal_file_find_data_object_with_hash(f, m->data, m->size, le64toh(m->le_hash), NULL, &dp); |
| if (r <= 0) |
| return r; |
| |
| /* FIXME: missing: find by monotonic */ |
| |
| if (j->current_location.type == LOCATION_HEAD) |
| return journal_file_next_entry_for_data(f, NULL, 0, dp, DIRECTION_DOWN, ret, offset); |
| if (j->current_location.type == LOCATION_TAIL) |
| return journal_file_next_entry_for_data(f, NULL, 0, dp, DIRECTION_UP, ret, offset); |
| if (j->current_location.seqnum_set && sd_id128_equal(j->current_location.seqnum_id, f->header->seqnum_id)) |
| return journal_file_move_to_entry_by_seqnum_for_data(f, dp, j->current_location.seqnum, direction, ret, offset); |
| if (j->current_location.monotonic_set) { |
| r = journal_file_move_to_entry_by_monotonic_for_data(f, dp, j->current_location.boot_id, j->current_location.monotonic, direction, ret, offset); |
| if (r != -ENOENT) |
| return r; |
| } |
| if (j->current_location.realtime_set) |
| return journal_file_move_to_entry_by_realtime_for_data(f, dp, j->current_location.realtime, direction, ret, offset); |
| |
| return journal_file_next_entry_for_data(f, NULL, 0, dp, direction, ret, offset); |
| |
| } else if (m->type == MATCH_OR_TERM) { |
| uint64_t np = 0; |
| Object *n; |
| Match *i; |
| |
| /* Find the earliest match */ |
| |
| LIST_FOREACH(matches, i, m->matches) { |
| uint64_t cp; |
| |
| r = find_location_for_match(j, i, f, direction, NULL, &cp); |
| if (r < 0) |
| return r; |
| else if (r > 0) { |
| if (np == 0 || (direction == DIRECTION_DOWN ? np > cp : np < cp)) |
| np = cp; |
| } |
| } |
| |
| if (np == 0) |
| return 0; |
| |
| r = journal_file_move_to_object(f, OBJECT_ENTRY, np, &n); |
| if (r < 0) |
| return r; |
| |
| if (ret) |
| *ret = n; |
| if (offset) |
| *offset = np; |
| |
| return 1; |
| |
| } else { |
| Match *i; |
| uint64_t np = 0; |
| |
| assert(m->type == MATCH_AND_TERM); |
| |
| /* First jump to the last match, and then find the |
| * next one where all matches match */ |
| |
| if (!m->matches) |
| return 0; |
| |
| LIST_FOREACH(matches, i, m->matches) { |
| uint64_t cp; |
| |
| r = find_location_for_match(j, i, f, direction, NULL, &cp); |
| if (r <= 0) |
| return r; |
| |
| if (np == 0 || (direction == DIRECTION_DOWN ? np < cp : np > cp)) |
| np = cp; |
| } |
| |
| return next_for_match(j, m, f, np, direction, ret, offset); |
| } |
| } |
| |
| static int find_location_with_matches( |
| sd_journal *j, |
| JournalFile *f, |
| direction_t direction, |
| Object **ret, |
| uint64_t *offset) { |
| |
| int r; |
| |
| assert(j); |
| assert(f); |
| assert(ret); |
| assert(offset); |
| |
| if (!j->level0) { |
| /* No matches is simple */ |
| |
| if (j->current_location.type == LOCATION_HEAD) |
| return journal_file_next_entry(f, NULL, 0, DIRECTION_DOWN, ret, offset); |
| if (j->current_location.type == LOCATION_TAIL) |
| return journal_file_next_entry(f, NULL, 0, DIRECTION_UP, ret, offset); |
| if (j->current_location.seqnum_set && sd_id128_equal(j->current_location.seqnum_id, f->header->seqnum_id)) |
| return journal_file_move_to_entry_by_seqnum(f, j->current_location.seqnum, direction, ret, offset); |
| if (j->current_location.monotonic_set) { |
| r = journal_file_move_to_entry_by_monotonic(f, j->current_location.boot_id, j->current_location.monotonic, direction, ret, offset); |
| if (r != -ENOENT) |
| return r; |
| } |
| if (j->current_location.realtime_set) |
| return journal_file_move_to_entry_by_realtime(f, j->current_location.realtime, direction, ret, offset); |
| |
| return journal_file_next_entry(f, NULL, 0, direction, ret, offset); |
| } else |
| return find_location_for_match(j, j->level0, f, direction, ret, offset); |
| } |
| |
| static int next_with_matches( |
| sd_journal *j, |
| JournalFile *f, |
| direction_t direction, |
| Object **ret, |
| uint64_t *offset) { |
| |
| Object *c; |
| uint64_t cp; |
| |
| assert(j); |
| assert(f); |
| assert(ret); |
| assert(offset); |
| |
| c = *ret; |
| cp = *offset; |
| |
| /* No matches is easy. We simple advance the file |
| * pointer by one. */ |
| if (!j->level0) |
| return journal_file_next_entry(f, c, cp, direction, ret, offset); |
| |
| /* If we have a match then we look for the next matching entry |
| * with an offset at least one step larger */ |
| return next_for_match(j, j->level0, f, direction == DIRECTION_DOWN ? cp+1 : cp-1, direction, ret, offset); |
| } |
| |
| static int next_beyond_location(sd_journal *j, JournalFile *f, direction_t direction, Object **ret, uint64_t *offset) { |
| Object *c; |
| uint64_t cp; |
| int r; |
| |
| assert(j); |
| assert(f); |
| |
| if (f->current_offset > 0) { |
| cp = f->current_offset; |
| |
| r = journal_file_move_to_object(f, OBJECT_ENTRY, cp, &c); |
| if (r < 0) |
| return r; |
| |
| r = next_with_matches(j, f, direction, &c, &cp); |
| if (r <= 0) |
| return r; |
| } else { |
| r = find_location_with_matches(j, f, direction, &c, &cp); |
| if (r <= 0) |
| return r; |
| } |
| |
| /* OK, we found the spot, now let's advance until to an entry |
| * that is actually different from what we were previously |
| * looking at. This is necessary to handle entries which exist |
| * in two (or more) journal files, and which shall all be |
| * suppressed but one. */ |
| |
| for (;;) { |
| bool found; |
| |
| if (j->current_location.type == LOCATION_DISCRETE) { |
| int k; |
| |
| k = compare_with_location(f, c, &j->current_location); |
| if (direction == DIRECTION_DOWN) |
| found = k > 0; |
| else |
| found = k < 0; |
| } else |
| found = true; |
| |
| if (found) { |
| if (ret) |
| *ret = c; |
| if (offset) |
| *offset = cp; |
| return 1; |
| } |
| |
| r = next_with_matches(j, f, direction, &c, &cp); |
| if (r <= 0) |
| return r; |
| } |
| } |
| |
| static int real_journal_next(sd_journal *j, direction_t direction) { |
| JournalFile *f, *new_file = NULL; |
| uint64_t new_offset = 0; |
| Object *o; |
| uint64_t p; |
| Iterator i; |
| int r; |
| |
| if (!j) |
| return -EINVAL; |
| |
| HASHMAP_FOREACH(f, j->files, i) { |
| bool found; |
| |
| r = next_beyond_location(j, f, direction, &o, &p); |
| if (r < 0) { |
| log_debug("Can't iterate through %s, ignoring: %s", f->path, strerror(-r)); |
| continue; |
| } else if (r == 0) |
| continue; |
| |
| if (!new_file) |
| found = true; |
| else { |
| int k; |
| |
| k = compare_entry_order(f, o, new_file, new_offset); |
| |
| if (direction == DIRECTION_DOWN) |
| found = k < 0; |
| else |
| found = k > 0; |
| } |
| |
| if (found) { |
| new_file = f; |
| new_offset = p; |
| } |
| } |
| |
| if (!new_file) |
| return 0; |
| |
| r = journal_file_move_to_object(new_file, OBJECT_ENTRY, new_offset, &o); |
| if (r < 0) |
| return r; |
| |
| set_location(j, LOCATION_DISCRETE, new_file, o, new_offset); |
| |
| return 1; |
| } |
| |
| _public_ int sd_journal_next(sd_journal *j) { |
| return real_journal_next(j, DIRECTION_DOWN); |
| } |
| |
| _public_ int sd_journal_previous(sd_journal *j) { |
| return real_journal_next(j, DIRECTION_UP); |
| } |
| |
| static int real_journal_next_skip(sd_journal *j, direction_t direction, uint64_t skip) { |
| int c = 0, r; |
| |
| if (!j) |
| return -EINVAL; |
| |
| if (skip == 0) { |
| /* If this is not a discrete skip, then at least |
| * resolve the current location */ |
| if (j->current_location.type != LOCATION_DISCRETE) |
| return real_journal_next(j, direction); |
| |
| return 0; |
| } |
| |
| do { |
| r = real_journal_next(j, direction); |
| if (r < 0) |
| return r; |
| |
| if (r == 0) |
| return c; |
| |
| skip--; |
| c++; |
| } while (skip > 0); |
| |
| return c; |
| } |
| |
| _public_ int sd_journal_next_skip(sd_journal *j, uint64_t skip) { |
| return real_journal_next_skip(j, DIRECTION_DOWN, skip); |
| } |
| |
| _public_ int sd_journal_previous_skip(sd_journal *j, uint64_t skip) { |
| return real_journal_next_skip(j, DIRECTION_UP, skip); |
| } |
| |
| _public_ int sd_journal_get_cursor(sd_journal *j, char **cursor) { |
| Object *o; |
| int r; |
| char bid[33], sid[33]; |
| |
| if (!j) |
| return -EINVAL; |
| if (!cursor) |
| return -EINVAL; |
| |
| if (!j->current_file || j->current_file->current_offset <= 0) |
| return -EADDRNOTAVAIL; |
| |
| r = journal_file_move_to_object(j->current_file, OBJECT_ENTRY, j->current_file->current_offset, &o); |
| if (r < 0) |
| return r; |
| |
| sd_id128_to_string(j->current_file->header->seqnum_id, sid); |
| sd_id128_to_string(o->entry.boot_id, bid); |
| |
| if (asprintf(cursor, |
| "s=%s;i=%llx;b=%s;m=%llx;t=%llx;x=%llx", |
| sid, (unsigned long long) le64toh(o->entry.seqnum), |
| bid, (unsigned long long) le64toh(o->entry.monotonic), |
| (unsigned long long) le64toh(o->entry.realtime), |
| (unsigned long long) le64toh(o->entry.xor_hash)) < 0) |
| return -ENOMEM; |
| |
| return 0; |
| } |
| |
| _public_ int sd_journal_seek_cursor(sd_journal *j, const char *cursor) { |
| char *w, *state; |
| size_t l; |
| unsigned long long seqnum, monotonic, realtime, xor_hash; |
| bool |
| seqnum_id_set = false, |
| seqnum_set = false, |
| boot_id_set = false, |
| monotonic_set = false, |
| realtime_set = false, |
| xor_hash_set = false; |
| sd_id128_t seqnum_id, boot_id; |
| |
| if (!j) |
| return -EINVAL; |
| if (isempty(cursor)) |
| return -EINVAL; |
| |
| FOREACH_WORD_SEPARATOR(w, l, cursor, ";", state) { |
| char *item; |
| int k = 0; |
| |
| if (l < 2 || w[1] != '=') |
| return -EINVAL; |
| |
| item = strndup(w, l); |
| if (!item) |
| return -ENOMEM; |
| |
| switch (w[0]) { |
| |
| case 's': |
| seqnum_id_set = true; |
| k = sd_id128_from_string(item+2, &seqnum_id); |
| break; |
| |
| case 'i': |
| seqnum_set = true; |
| if (sscanf(item+2, "%llx", &seqnum) != 1) |
| k = -EINVAL; |
| break; |
| |
| case 'b': |
| boot_id_set = true; |
| k = sd_id128_from_string(item+2, &boot_id); |
| break; |
| |
| case 'm': |
| monotonic_set = true; |
| if (sscanf(item+2, "%llx", &monotonic) != 1) |
| k = -EINVAL; |
| break; |
| |
| case 't': |
| realtime_set = true; |
| if (sscanf(item+2, "%llx", &realtime) != 1) |
| k = -EINVAL; |
| break; |
| |
| case 'x': |
| xor_hash_set = true; |
| if (sscanf(item+2, "%llx", &xor_hash) != 1) |
| k = -EINVAL; |
| break; |
| } |
| |
| free(item); |
| |
| if (k < 0) |
| return k; |
| } |
| |
| if ((!seqnum_set || !seqnum_id_set) && |
| (!monotonic_set || !boot_id_set) && |
| !realtime_set) |
| return -EINVAL; |
| |
| reset_location(j); |
| |
| j->current_location.type = LOCATION_SEEK; |
| |
| if (realtime_set) { |
| j->current_location.realtime = (uint64_t) realtime; |
| j->current_location.realtime_set = true; |
| } |
| |
| if (seqnum_set && seqnum_id_set) { |
| j->current_location.seqnum = (uint64_t) seqnum; |
| j->current_location.seqnum_id = seqnum_id; |
| j->current_location.seqnum_set = true; |
| } |
| |
| if (monotonic_set && boot_id_set) { |
| j->current_location.monotonic = (uint64_t) monotonic; |
| j->current_location.boot_id = boot_id; |
| j->current_location.monotonic_set = true; |
| } |
| |
| if (xor_hash_set) { |
| j->current_location.xor_hash = (uint64_t) xor_hash; |
| j->current_location.xor_hash_set = true; |
| } |
| |
| return 0; |
| } |
| |
| _public_ int sd_journal_test_cursor(sd_journal *j, const char *cursor) { |
| int r; |
| char *w, *state; |
| size_t l; |
| Object *o; |
| |
| if (!j) |
| return -EINVAL; |
| if (isempty(cursor)) |
| return -EINVAL; |
| |
| if (!j->current_file || j->current_file->current_offset <= 0) |
| return -EADDRNOTAVAIL; |
| |
| r = journal_file_move_to_object(j->current_file, OBJECT_ENTRY, j->current_file->current_offset, &o); |
| if (r < 0) |
| return r; |
| |
| FOREACH_WORD_SEPARATOR(w, l, cursor, ";", state) { |
| _cleanup_free_ char *item = NULL; |
| sd_id128_t id; |
| unsigned long long ll; |
| int k = 0; |
| |
| if (l < 2 || w[1] != '=') |
| return -EINVAL; |
| |
| item = strndup(w, l); |
| if (!item) |
| return -ENOMEM; |
| |
| switch (w[0]) { |
| |
| case 's': |
| k = sd_id128_from_string(item+2, &id); |
| if (k < 0) |
| return k; |
| if (!sd_id128_equal(id, j->current_file->header->seqnum_id)) |
| return 0; |
| break; |
| |
| case 'i': |
| if (sscanf(item+2, "%llx", &ll) != 1) |
| return -EINVAL; |
| if (ll != le64toh(o->entry.seqnum)) |
| return 0; |
| break; |
| |
| case 'b': |
| k = sd_id128_from_string(item+2, &id); |
| if (k < 0) |
| return k; |
| if (!sd_id128_equal(id, o->entry.boot_id)) |
| return 0; |
| break; |
| |
| case 'm': |
| if (sscanf(item+2, "%llx", &ll) != 1) |
| return -EINVAL; |
| if (ll != le64toh(o->entry.monotonic)) |
| return 0; |
| break; |
| |
| case 't': |
| if (sscanf(item+2, "%llx", &ll) != 1) |
| return -EINVAL; |
| if (ll != le64toh(o->entry.realtime)) |
| return 0; |
| break; |
| |
| case 'x': |
| if (sscanf(item+2, "%llx", &ll) != 1) |
| return -EINVAL; |
| if (ll != le64toh(o->entry.xor_hash)) |
| return 0; |
| break; |
| } |
| } |
| |
| return 1; |
| } |
| |
| |
| _public_ int sd_journal_seek_monotonic_usec(sd_journal *j, sd_id128_t boot_id, uint64_t usec) { |
| if (!j) |
| return -EINVAL; |
| |
| reset_location(j); |
| j->current_location.type = LOCATION_SEEK; |
| j->current_location.boot_id = boot_id; |
| j->current_location.monotonic = usec; |
| j->current_location.monotonic_set = true; |
| |
| return 0; |
| } |
| |
| _public_ int sd_journal_seek_realtime_usec(sd_journal *j, uint64_t usec) { |
| if (!j) |
| return -EINVAL; |
| |
| reset_location(j); |
| j->current_location.type = LOCATION_SEEK; |
| j->current_location.realtime = usec; |
| j->current_location.realtime_set = true; |
| |
| return 0; |
| } |
| |
| _public_ int sd_journal_seek_head(sd_journal *j) { |
| if (!j) |
| return -EINVAL; |
| |
| reset_location(j); |
| j->current_location.type = LOCATION_HEAD; |
| |
| return 0; |
| } |
| |
| _public_ int sd_journal_seek_tail(sd_journal *j) { |
| if (!j) |
| return -EINVAL; |
| |
| reset_location(j); |
| j->current_location.type = LOCATION_TAIL; |
| |
| return 0; |
| } |
| |
| static void check_network(sd_journal *j, int fd) { |
| struct statfs sfs; |
| |
| assert(j); |
| |
| if (j->on_network) |
| return; |
| |
| if (fstatfs(fd, &sfs) < 0) |
| return; |
| |
| j->on_network = |
| (long)sfs.f_type == (long)CIFS_MAGIC_NUMBER || |
| sfs.f_type == CODA_SUPER_MAGIC || |
| sfs.f_type == NCP_SUPER_MAGIC || |
| sfs.f_type == NFS_SUPER_MAGIC || |
| sfs.f_type == SMB_SUPER_MAGIC; |
| } |
| |
| static int add_file(sd_journal *j, const char *prefix, const char *filename) { |
| char _cleanup_free_ *path = NULL; |
| int r; |
| JournalFile *f; |
| |
| assert(j); |
| assert(prefix); |
| assert(filename); |
| |
| if ((j->flags & SD_JOURNAL_SYSTEM_ONLY) && |
| !(streq(filename, "system.journal") || |
| streq(filename, "system.journal~") || |
| (startswith(filename, "system@") && |
| (endswith(filename, ".journal") || endswith(filename, ".journal~"))))) |
| return 0; |
| |
| path = strjoin(prefix, "/", filename, NULL); |
| if (!path) |
| return -ENOMEM; |
| |
| if (hashmap_get(j->files, path)) |
| return 0; |
| |
| if (hashmap_size(j->files) >= JOURNAL_FILES_MAX) { |
| log_debug("Too many open journal files, not adding %s, ignoring.", path); |
| return set_put_error(j, -ETOOMANYREFS); |
| } |
| |
| r = journal_file_open(path, O_RDONLY, 0, false, false, NULL, j->mmap, NULL, &f); |
| if (r < 0) { |
| if (errno == ENOENT) |
| return 0; |
| |
| return r; |
| } |
| |
| /* journal_file_dump(f); */ |
| |
| r = hashmap_put(j->files, f->path, f); |
| if (r < 0) { |
| journal_file_close(f); |
| return r; |
| } |
| |
| log_debug("File %s got added.", f->path); |
| |
| check_network(j, f->fd); |
| |
| j->current_invalidate_counter ++; |
| |
| return 0; |
| } |
| |
| static int remove_file(sd_journal *j, const char *prefix, const char *filename) { |
| char *path; |
| JournalFile *f; |
| |
| assert(j); |
| assert(prefix); |
| assert(filename); |
| |
| path = strjoin(prefix, "/", filename, NULL); |
| if (!path) |
| return -ENOMEM; |
| |
| f = hashmap_get(j->files, path); |
| free(path); |
| if (!f) |
| return 0; |
| |
| hashmap_remove(j->files, f->path); |
| |
| log_debug("File %s got removed.", f->path); |
| |
| if (j->current_file == f) { |
| j->current_file = NULL; |
| j->current_field = 0; |
| } |
| |
| if (j->unique_file == f) { |
| j->unique_file = NULL; |
| j->unique_offset = 0; |
| } |
| |
| journal_file_close(f); |
| |
| j->current_invalidate_counter ++; |
| |
| return 0; |
| } |
| |
| static int add_directory(sd_journal *j, const char *prefix, const char *dirname) { |
| char _cleanup_free_ *path = NULL; |
| int r; |
| DIR _cleanup_closedir_ *d = NULL; |
| sd_id128_t id, mid; |
| Directory *m; |
| |
| assert(j); |
| assert(prefix); |
| assert(dirname); |
| |
| log_debug("Considering %s/%s.", prefix, dirname); |
| |
| if ((j->flags & SD_JOURNAL_LOCAL_ONLY) && |
| (sd_id128_from_string(dirname, &id) < 0 || |
| sd_id128_get_machine(&mid) < 0 || |
| !(sd_id128_equal(id, mid) || path_startswith(prefix, "/run")))) |
| return 0; |
| |
| path = strjoin(prefix, "/", dirname, NULL); |
| if (!path) |
| return -ENOMEM; |
| |
| d = opendir(path); |
| if (!d) { |
| log_debug("Failed to open %s: %m", path); |
| if (errno == ENOENT) |
| return 0; |
| return -errno; |
| } |
| |
| m = hashmap_get(j->directories_by_path, path); |
| if (!m) { |
| m = new0(Directory, 1); |
| if (!m) |
| return -ENOMEM; |
| |
| m->is_root = false; |
| m->path = path; |
| |
| if (hashmap_put(j->directories_by_path, m->path, m) < 0) { |
| free(m); |
| return -ENOMEM; |
| } |
| |
| path = NULL; /* avoid freeing in cleanup */ |
| j->current_invalidate_counter ++; |
| |
| log_debug("Directory %s got added.", m->path); |
| |
| } else if (m->is_root) |
| return 0; |
| |
| if (m->wd <= 0 && j->inotify_fd >= 0) { |
| |
| m->wd = inotify_add_watch(j->inotify_fd, m->path, |
| IN_CREATE|IN_MOVED_TO|IN_MODIFY|IN_ATTRIB|IN_DELETE| |
| IN_DELETE_SELF|IN_MOVE_SELF|IN_UNMOUNT|IN_MOVED_FROM| |
| IN_ONLYDIR); |
| |
| if (m->wd > 0 && hashmap_put(j->directories_by_wd, INT_TO_PTR(m->wd), m) < 0) |
| inotify_rm_watch(j->inotify_fd, m->wd); |
| } |
| |
| for (;;) { |
| struct dirent *de; |
| union dirent_storage buf; |
| |
| r = readdir_r(d, &buf.de, &de); |
| if (r != 0 || !de) |
| break; |
| |
| if (dirent_is_file_with_suffix(de, ".journal") || |
| dirent_is_file_with_suffix(de, ".journal~")) { |
| r = add_file(j, m->path, de->d_name); |
| if (r < 0) { |
| log_debug("Failed to add file %s/%s: %s", |
| m->path, de->d_name, strerror(-r)); |
| r = set_put_error(j, r); |
| if (r < 0) |
| return r; |
| } |
| } |
| } |
| |
| check_network(j, dirfd(d)); |
| |
| return 0; |
| } |
| |
| static int add_root_directory(sd_journal *j, const char *p) { |
| DIR _cleanup_closedir_ *d = NULL; |
| Directory *m; |
| int r; |
| |
| assert(j); |
| assert(p); |
| |
| if ((j->flags & SD_JOURNAL_RUNTIME_ONLY) && |
| !path_startswith(p, "/run")) |
| return -EINVAL; |
| |
| d = opendir(p); |
| if (!d) |
| return -errno; |
| |
| m = hashmap_get(j->directories_by_path, p); |
| if (!m) { |
| m = new0(Directory, 1); |
| if (!m) |
| return -ENOMEM; |
| |
| m->is_root = true; |
| m->path = strdup(p); |
| if (!m->path) { |
| free(m); |
| return -ENOMEM; |
| } |
| |
| if (hashmap_put(j->directories_by_path, m->path, m) < 0) { |
| free(m->path); |
| free(m); |
| return -ENOMEM; |
| } |
| |
| j->current_invalidate_counter ++; |
| |
| log_debug("Root directory %s got added.", m->path); |
| |
| } else if (!m->is_root) |
| return 0; |
| |
| if (m->wd <= 0 && j->inotify_fd >= 0) { |
| |
| m->wd = inotify_add_watch(j->inotify_fd, m->path, |
| IN_CREATE|IN_MOVED_TO|IN_MODIFY|IN_ATTRIB|IN_DELETE| |
| IN_ONLYDIR); |
| |
| if (m->wd > 0 && hashmap_put(j->directories_by_wd, INT_TO_PTR(m->wd), m) < 0) |
| inotify_rm_watch(j->inotify_fd, m->wd); |
| } |
| |
| for (;;) { |
| struct dirent *de; |
| union dirent_storage buf; |
| sd_id128_t id; |
| |
| r = readdir_r(d, &buf.de, &de); |
| if (r != 0 || !de) |
| break; |
| |
| if (dirent_is_file_with_suffix(de, ".journal") || |
| dirent_is_file_with_suffix(de, ".journal~")) { |
| r = add_file(j, m->path, de->d_name); |
| if (r < 0) { |
| log_debug("Failed to add file %s/%s: %s", |
| m->path, de->d_name, strerror(-r)); |
| r = set_put_error(j, r); |
| if (r < 0) |
| return r; |
| } |
| } else if ((de->d_type == DT_DIR || de->d_type == DT_LNK || de->d_type == DT_UNKNOWN) && |
| sd_id128_from_string(de->d_name, &id) >= 0) { |
| |
| r = add_directory(j, m->path, de->d_name); |
| if (r < 0) |
| log_debug("Failed to add directory %s/%s: %s", m->path, de->d_name, strerror(-r)); |
| } |
| } |
| |
| check_network(j, dirfd(d)); |
| |
| return 0; |
| } |
| |
| static int remove_directory(sd_journal *j, Directory *d) { |
| assert(j); |
| |
| if (d->wd > 0) { |
| hashmap_remove(j->directories_by_wd, INT_TO_PTR(d->wd)); |
| |
| if (j->inotify_fd >= 0) |
| inotify_rm_watch(j->inotify_fd, d->wd); |
| } |
| |
| hashmap_remove(j->directories_by_path, d->path); |
| |
| if (d->is_root) |
| log_debug("Root directory %s got removed.", d->path); |
| else |
| log_debug("Directory %s got removed.", d->path); |
| |
| free(d->path); |
| free(d); |
| |
| return 0; |
| } |
| |
| static int add_search_paths(sd_journal *j) { |
| int r; |
| const char search_paths[] = |
| "/run/log/journal\0" |
| "/var/log/journal\0"; |
| const char *p; |
| |
| assert(j); |
| |
| /* We ignore most errors here, since the idea is to only open |
| * what's actually accessible, and ignore the rest. */ |
| |
| NULSTR_FOREACH(p, search_paths) { |
| r = add_root_directory(j, p); |
| if (r < 0 && r != -ENOENT) { |
| r = set_put_error(j, r); |
| if (r < 0) |
| return r; |
| } |
| } |
| |
| return 0; |
| } |
| |
| static int allocate_inotify(sd_journal *j) { |
| assert(j); |
| |
| if (j->inotify_fd < 0) { |
| j->inotify_fd = inotify_init1(IN_NONBLOCK|IN_CLOEXEC); |
| if (j->inotify_fd < 0) |
| return -errno; |
| } |
| |
| if (!j->directories_by_wd) { |
| j->directories_by_wd = hashmap_new(trivial_hash_func, trivial_compare_func); |
| if (!j->directories_by_wd) |
| return -ENOMEM; |
| } |
| |
| return 0; |
| } |
| |
| static sd_journal *journal_new(int flags, const char *path) { |
| sd_journal *j; |
| |
| j = new0(sd_journal, 1); |
| if (!j) |
| return NULL; |
| |
| j->inotify_fd = -1; |
| j->flags = flags; |
| j->data_threshold = DEFAULT_DATA_THRESHOLD; |
| |
| if (path) { |
| j->path = strdup(path); |
| if (!j->path) |
| goto fail; |
| } |
| |
| j->files = hashmap_new(string_hash_func, string_compare_func); |
| j->directories_by_path = hashmap_new(string_hash_func, string_compare_func); |
| j->mmap = mmap_cache_new(); |
| if (!j->files || !j->directories_by_path || !j->mmap) |
| goto fail; |
| |
| return j; |
| |
| fail: |
| sd_journal_close(j); |
| return NULL; |
| } |
| |
| _public_ int sd_journal_open(sd_journal **ret, int flags) { |
| sd_journal *j; |
| int r; |
| |
| if (!ret) |
| return -EINVAL; |
| |
| if (flags & ~(SD_JOURNAL_LOCAL_ONLY| |
| SD_JOURNAL_RUNTIME_ONLY| |
| SD_JOURNAL_SYSTEM_ONLY)) |
| return -EINVAL; |
| |
| j = journal_new(flags, NULL); |
| if (!j) |
| return -ENOMEM; |
| |
| r = add_search_paths(j); |
| if (r < 0) |
| goto fail; |
| |
| *ret = j; |
| return 0; |
| |
| fail: |
| sd_journal_close(j); |
| |
| return r; |
| } |
| |
| _public_ int sd_journal_open_directory(sd_journal **ret, const char *path, int flags) { |
| sd_journal *j; |
| int r; |
| |
| if (!ret) |
| return -EINVAL; |
| |
| if (!path) |
| return -EINVAL; |
| |
| if (flags != 0) |
| return -EINVAL; |
| |
| j = journal_new(flags, path); |
| if (!j) |
| return -ENOMEM; |
| |
| r = add_root_directory(j, path); |
| if (r < 0) { |
| set_put_error(j, r); |
| goto fail; |
| } |
| |
| *ret = j; |
| return 0; |
| |
| fail: |
| sd_journal_close(j); |
| |
| return r; |
| } |
| |
| _public_ void sd_journal_close(sd_journal *j) { |
| Directory *d; |
| JournalFile *f; |
| |
| if (!j) |
| return; |
| |
| sd_journal_flush_matches(j); |
| |
| while ((f = hashmap_steal_first(j->files))) |
| journal_file_close(f); |
| |
| hashmap_free(j->files); |
| |
| while ((d = hashmap_first(j->directories_by_path))) |
| remove_directory(j, d); |
| |
| while ((d = hashmap_first(j->directories_by_wd))) |
| remove_directory(j, d); |
| |
| hashmap_free(j->directories_by_path); |
| hashmap_free(j->directories_by_wd); |
| |
| if (j->inotify_fd >= 0) |
| close_nointr_nofail(j->inotify_fd); |
| |
| if (j->mmap) |
| mmap_cache_unref(j->mmap); |
| |
| free(j->path); |
| free(j->unique_field); |
| set_free(j->errors); |
| free(j); |
| } |
| |
| _public_ int sd_journal_get_realtime_usec(sd_journal *j, uint64_t *ret) { |
| Object *o; |
| JournalFile *f; |
| int r; |
| |
| if (!j) |
| return -EINVAL; |
| if (!ret) |
| return -EINVAL; |
| |
| f = j->current_file; |
| if (!f) |
| return -EADDRNOTAVAIL; |
| |
| if (f->current_offset <= 0) |
| return -EADDRNOTAVAIL; |
| |
| r = journal_file_move_to_object(f, OBJECT_ENTRY, f->current_offset, &o); |
| if (r < 0) |
| return r; |
| |
| *ret = le64toh(o->entry.realtime); |
| return 0; |
| } |
| |
| _public_ int sd_journal_get_monotonic_usec(sd_journal *j, uint64_t *ret, sd_id128_t *ret_boot_id) { |
| Object *o; |
| JournalFile *f; |
| int r; |
| sd_id128_t id; |
| |
| if (!j) |
| return -EINVAL; |
| |
| f = j->current_file; |
| if (!f) |
| return -EADDRNOTAVAIL; |
| |
| if (f->current_offset <= 0) |
| return -EADDRNOTAVAIL; |
| |
| r = journal_file_move_to_object(f, OBJECT_ENTRY, f->current_offset, &o); |
| if (r < 0) |
| return r; |
| |
| if (ret_boot_id) |
| *ret_boot_id = o->entry.boot_id; |
| else { |
| r = sd_id128_get_boot(&id); |
| if (r < 0) |
| return r; |
| |
| if (!sd_id128_equal(id, o->entry.boot_id)) |
| return -ESTALE; |
| } |
| |
| if (ret) |
| *ret = le64toh(o->entry.monotonic); |
| |
| return 0; |
| } |
| |
| static bool field_is_valid(const char *field) { |
| const char *p; |
| |
| assert(field); |
| |
| if (isempty(field)) |
| return false; |
| |
| if (startswith(field, "__")) |
| return false; |
| |
| for (p = field; *p; p++) { |
| |
| if (*p == '_') |
| continue; |
| |
| if (*p >= 'A' && *p <= 'Z') |
| continue; |
| |
| if (*p >= '0' && *p <= '9') |
| continue; |
| |
| return false; |
| } |
| |
| return true; |
| } |
| |
| _public_ int sd_journal_get_data(sd_journal *j, const char *field, const void **data, size_t *size) { |
| JournalFile *f; |
| uint64_t i, n; |
| size_t field_length; |
| int r; |
| Object *o; |
| |
| if (!j) |
| return -EINVAL; |
| if (!field) |
| return -EINVAL; |
| if (!data) |
| return -EINVAL; |
| if (!size) |
| return -EINVAL; |
| |
| if (!field_is_valid(field)) |
| return -EINVAL; |
| |
| f = j->current_file; |
| if (!f) |
| return -EADDRNOTAVAIL; |
| |
| if (f->current_offset <= 0) |
| return -EADDRNOTAVAIL; |
| |
| r = journal_file_move_to_object(f, OBJECT_ENTRY, f->current_offset, &o); |
| if (r < 0) |
| return r; |
| |
| field_length = strlen(field); |
| |
| n = journal_file_entry_n_items(o); |
| for (i = 0; i < n; i++) { |
| uint64_t p, l; |
| le64_t le_hash; |
| size_t t; |
| |
| p = le64toh(o->entry.items[i].object_offset); |
| le_hash = o->entry.items[i].hash; |
| r = journal_file_move_to_object(f, OBJECT_DATA, p, &o); |
| if (r < 0) |
| return r; |
| |
| if (le_hash != o->data.hash) |
| return -EBADMSG; |
| |
| l = le64toh(o->object.size) - offsetof(Object, data.payload); |
| |
| if (o->object.flags & OBJECT_COMPRESSED) { |
| |
| #ifdef HAVE_XZ |
| if (uncompress_startswith(o->data.payload, l, |
| &f->compress_buffer, &f->compress_buffer_size, |
| field, field_length, '=')) { |
| |
| uint64_t rsize; |
| |
| if (!uncompress_blob(o->data.payload, l, |
| &f->compress_buffer, &f->compress_buffer_size, &rsize, |
| j->data_threshold)) |
| return -EBADMSG; |
| |
| *data = f->compress_buffer; |
| *size = (size_t) rsize; |
| |
| return 0; |
| } |
| #else |
| return -EPROTONOSUPPORT; |
| #endif |
| |
| } else if (l >= field_length+1 && |
| memcmp(o->data.payload, field, field_length) == 0 && |
| o->data.payload[field_length] == '=') { |
| |
| t = (size_t) l; |
| |
| if ((uint64_t) t != l) |
| return -E2BIG; |
| |
| *data = o->data.payload; |
| *size = t; |
| |
| return 0; |
| } |
| |
| r = journal_file_move_to_object(f, OBJECT_ENTRY, f->current_offset, &o); |
| if (r < 0) |
| return r; |
| } |
| |
| return -ENOENT; |
| } |
| |
| static int return_data(sd_journal *j, JournalFile *f, Object *o, const void **data, size_t *size) { |
| size_t t; |
| uint64_t l; |
| |
| l = le64toh(o->object.size) - offsetof(Object, data.payload); |
| t = (size_t) l; |
| |
| /* We can't read objects larger than 4G on a 32bit machine */ |
| if ((uint64_t) t != l) |
| return -E2BIG; |
| |
| if (o->object.flags & OBJECT_COMPRESSED) { |
| #ifdef HAVE_XZ |
| uint64_t rsize; |
| |
| if (!uncompress_blob(o->data.payload, l, &f->compress_buffer, &f->compress_buffer_size, &rsize, j->data_threshold)) |
| return -EBADMSG; |
| |
| *data = f->compress_buffer; |
| *size = (size_t) rsize; |
| #else |
| return -EPROTONOSUPPORT; |
| #endif |
| } else { |
| *data = o->data.payload; |
| *size = t; |
| } |
| |
| return 0; |
| } |
| |
| _public_ int sd_journal_enumerate_data(sd_journal *j, const void **data, size_t *size) { |
| JournalFile *f; |
| uint64_t p, n; |
| le64_t le_hash; |
| int r; |
| Object *o; |
| |
| if (!j) |
| return -EINVAL; |
| if (!data) |
| return -EINVAL; |
| if (!size) |
| return -EINVAL; |
| |
| f = j->current_file; |
| if (!f) |
| return -EADDRNOTAVAIL; |
| |
| if (f->current_offset <= 0) |
| return -EADDRNOTAVAIL; |
| |
| r = journal_file_move_to_object(f, OBJECT_ENTRY, f->current_offset, &o); |
| if (r < 0) |
| return r; |
| |
| n = journal_file_entry_n_items(o); |
| if (j->current_field >= n) |
| return 0; |
| |
| p = le64toh(o->entry.items[j->current_field].object_offset); |
| le_hash = o->entry.items[j->current_field].hash; |
| r = journal_file_move_to_object(f, OBJECT_DATA, p, &o); |
| if (r < 0) |
| return r; |
| |
| if (le_hash != o->data.hash) |
| return -EBADMSG; |
| |
| r = return_data(j, f, o, data, size); |
| if (r < 0) |
| return r; |
| |
| j->current_field ++; |
| |
| return 1; |
| } |
| |
| _public_ void sd_journal_restart_data(sd_journal *j) { |
| if (!j) |
| return; |
| |
| j->current_field = 0; |
| } |
| |
| _public_ int sd_journal_get_fd(sd_journal *j) { |
| int r; |
| |
| if (!j) |
| return -EINVAL; |
| |
| if (j->inotify_fd >= 0) |
| return j->inotify_fd; |
| |
| r = allocate_inotify(j); |
| if (r < 0) |
| return r; |
| |
| /* Iterate through all dirs again, to add them to the |
| * inotify */ |
| if (j->path) |
| r = add_root_directory(j, j->path); |
| else |
| r = add_search_paths(j); |
| if (r < 0) |
| return r; |
| |
| return j->inotify_fd; |
| } |
| |
| _public_ int sd_journal_get_events(sd_journal *j) { |
| int fd; |
| |
| if (!j) |
| return -EINVAL; |
| |
| fd = sd_journal_get_fd(j); |
| if (fd < 0) |
| return fd; |
| |
| return POLLIN; |
| } |
| |
| _public_ int sd_journal_get_timeout(sd_journal *j, uint64_t *timeout_usec) { |
| int fd; |
| |
| if (!j) |
| return -EINVAL; |
| if (!timeout_usec) |
| return -EINVAL; |
| |
| fd = sd_journal_get_fd(j); |
| if (fd < 0) |
| return fd; |
| |
| if (!j->on_network) { |
| *timeout_usec = (uint64_t) -1; |
| return 0; |
| } |
| |
| /* If we are on the network we need to regularly check for |
| * changes manually */ |
| |
| *timeout_usec = j->last_process_usec + JOURNAL_FILES_RECHECK_USEC; |
| return 1; |
| } |
| |
| static void process_inotify_event(sd_journal *j, struct inotify_event *e) { |
| Directory *d; |
| int r; |
| |
| assert(j); |
| assert(e); |
| |
| /* Is this a subdirectory we watch? */ |
| d = hashmap_get(j->directories_by_wd, INT_TO_PTR(e->wd)); |
| if (d) { |
| sd_id128_t id; |
| |
| if (!(e->mask & IN_ISDIR) && e->len > 0 && |
| (endswith(e->name, ".journal") || |
| endswith(e->name, ".journal~"))) { |
| |
| /* Event for a journal file */ |
| |
| if (e->mask & (IN_CREATE|IN_MOVED_TO|IN_MODIFY|IN_ATTRIB)) { |
| r = add_file(j, d->path, e->name); |
| if (r < 0) { |
| log_debug("Failed to add file %s/%s: %s", |
| d->path, e->name, strerror(-r)); |
| set_put_error(j, r); |
| } |
| |
| } else if (e->mask & (IN_DELETE|IN_MOVED_FROM|IN_UNMOUNT)) { |
| |
| r = remove_file(j, d->path, e->name); |
| if (r < 0) |
| log_debug("Failed to remove file %s/%s: %s", d->path, e->name, strerror(-r)); |
| } |
| |
| } else if (!d->is_root && e->len == 0) { |
| |
| /* Event for a subdirectory */ |
| |
| if (e->mask & (IN_DELETE_SELF|IN_MOVE_SELF|IN_UNMOUNT)) { |
| r = remove_directory(j, d); |
| if (r < 0) |
| log_debug("Failed to remove directory %s: %s", d->path, strerror(-r)); |
| } |
| |
| |
| } else if (d->is_root && (e->mask & IN_ISDIR) && e->len > 0 && sd_id128_from_string(e->name, &id) >= 0) { |
| |
| /* Event for root directory */ |
| |
| if (e->mask & (IN_CREATE|IN_MOVED_TO|IN_MODIFY|IN_ATTRIB)) { |
| r = add_directory(j, d->path, e->name); |
| if (r < 0) |
| log_debug("Failed to add directory %s/%s: %s", d->path, e->name, strerror(-r)); |
| } |
| } |
| |
| return; |
| } |
| |
| if (e->mask & IN_IGNORED) |
| return; |
| |
| log_warning("Unknown inotify event."); |
| } |
| |
| static int determine_change(sd_journal *j) { |
| bool b; |
| |
| assert(j); |
| |
| b = j->current_invalidate_counter != j->last_invalidate_counter; |
| j->last_invalidate_counter = j->current_invalidate_counter; |
| |
| return b ? SD_JOURNAL_INVALIDATE : SD_JOURNAL_APPEND; |
| } |
| |
| _public_ int sd_journal_process(sd_journal *j) { |
| uint8_t buffer[sizeof(struct inotify_event) + FILENAME_MAX] _alignas_(struct inotify_event); |
| bool got_something = false; |
| |
| if (!j) |
| return -EINVAL; |
| |
| j->last_process_usec = now(CLOCK_MONOTONIC); |
| |
| for (;;) { |
| struct inotify_event *e; |
| ssize_t l; |
| |
| l = read(j->inotify_fd, buffer, sizeof(buffer)); |
| if (l < 0) { |
| if (errno == EAGAIN || errno == EINTR) |
| return got_something ? determine_change(j) : SD_JOURNAL_NOP; |
| |
| return -errno; |
| } |
| |
| got_something = true; |
| |
| e = (struct inotify_event*) buffer; |
| while (l > 0) { |
| size_t step; |
| |
| process_inotify_event(j, e); |
| |
| step = sizeof(struct inotify_event) + e->len; |
| assert(step <= (size_t) l); |
| |
| e = (struct inotify_event*) ((uint8_t*) e + step); |
| l -= step; |
| } |
| } |
| |
| return determine_change(j); |
| } |
| |
| _public_ int sd_journal_wait(sd_journal *j, uint64_t timeout_usec) { |
| int r; |
| uint64_t t; |
| |
| assert(j); |
| |
| if (j->inotify_fd < 0) { |
| |
| /* This is the first invocation, hence create the |
| * inotify watch */ |
| r = sd_journal_get_fd(j); |
| if (r < 0) |
| return r; |
| |
| /* The journal might have changed since the context |
| * object was created and we weren't watching before, |
| * hence don't wait for anything, and return |
| * immediately. */ |
| return determine_change(j); |
| } |
| |
| r = sd_journal_get_timeout(j, &t); |
| if (r < 0) |
| return r; |
| |
| if (t != (uint64_t) -1) { |
| usec_t n; |
| |
| n = now(CLOCK_MONOTONIC); |
| t = t > n ? t - n : 0; |
| |
| if (timeout_usec == (uint64_t) -1 || timeout_usec > t) |
| timeout_usec = t; |
| } |
| |
| do { |
| r = fd_wait_for_event(j->inotify_fd, POLLIN, timeout_usec); |
| } while (r == -EINTR); |
| |
| if (r < 0) |
| return r; |
| |
| return sd_journal_process(j); |
| } |
| |
| _public_ int sd_journal_get_cutoff_realtime_usec(sd_journal *j, uint64_t *from, uint64_t *to) { |
| Iterator i; |
| JournalFile *f; |
| bool first = true; |
| int r; |
| |
| if (!j) |
| return -EINVAL; |
| if (!from && !to) |
| return -EINVAL; |
| |
| HASHMAP_FOREACH(f, j->files, i) { |
| usec_t fr, t; |
| |
| r = journal_file_get_cutoff_realtime_usec(f, &fr, &t); |
| if (r == -ENOENT) |
| continue; |
| if (r < 0) |
| return r; |
| if (r == 0) |
| continue; |
| |
| if (first) { |
| if (from) |
| *from = fr; |
| if (to) |
| *to = t; |
| first = false; |
| } else { |
| if (from) |
| *from = MIN(fr, *from); |
| if (to) |
| *to = MAX(t, *to); |
| } |
| } |
| |
| return first ? 0 : 1; |
| } |
| |
| _public_ int sd_journal_get_cutoff_monotonic_usec(sd_journal *j, sd_id128_t boot_id, uint64_t *from, uint64_t *to) { |
| Iterator i; |
| JournalFile *f; |
| bool first = true; |
| int r; |
| |
| if (!j) |
| return -EINVAL; |
| if (!from && !to) |
| return -EINVAL; |
| |
| HASHMAP_FOREACH(f, j->files, i) { |
| usec_t fr, t; |
| |
| r = journal_file_get_cutoff_monotonic_usec(f, boot_id, &fr, &t); |
| if (r == -ENOENT) |
| continue; |
| if (r < 0) |
| return r; |
| if (r == 0) |
| continue; |
| |
| if (first) { |
| if (from) |
| *from = fr; |
| if (to) |
| *to = t; |
| first = false; |
| } else { |
| if (from) |
| *from = MIN(fr, *from); |
| if (to) |
| *to = MAX(t, *to); |
| } |
| } |
| |
| return first ? 0 : 1; |
| } |
| |
| void journal_print_header(sd_journal *j) { |
| Iterator i; |
| JournalFile *f; |
| bool newline = false; |
| |
| assert(j); |
| |
| HASHMAP_FOREACH(f, j->files, i) { |
| if (newline) |
| putchar('\n'); |
| else |
| newline = true; |
| |
| journal_file_print_header(f); |
| } |
| } |
| |
| _public_ int sd_journal_get_usage(sd_journal *j, uint64_t *bytes) { |
| Iterator i; |
| JournalFile *f; |
| uint64_t sum = 0; |
| |
| if (!j) |
| return -EINVAL; |
| if (!bytes) |
| return -EINVAL; |
| |
| HASHMAP_FOREACH(f, j->files, i) { |
| struct stat st; |
| |
| if (fstat(f->fd, &st) < 0) |
| return -errno; |
| |
| sum += (uint64_t) st.st_blocks * 512ULL; |
| } |
| |
| *bytes = sum; |
| return 0; |
| } |
| |
| _public_ int sd_journal_query_unique(sd_journal *j, const char *field) { |
| char *f; |
| |
| if (!j) |
| return -EINVAL; |
| if (isempty(field)) |
| return -EINVAL; |
| if (!field_is_valid(field)) |
| return -EINVAL; |
| |
| f = strdup(field); |
| if (!f) |
| return -ENOMEM; |
| |
| free(j->unique_field); |
| j->unique_field = f; |
| j->unique_file = NULL; |
| j->unique_offset = 0; |
| |
| return 0; |
| } |
| |
| _public_ int sd_journal_enumerate_unique(sd_journal *j, const void **data, size_t *l) { |
| Object *o; |
| size_t k; |
| int r; |
| |
| if (!j) |
| return -EINVAL; |
| if (!data) |
| return -EINVAL; |
| if (!l) |
| return -EINVAL; |
| if (!j->unique_field) |
| return -EINVAL; |
| |
| k = strlen(j->unique_field); |
| |
| if (!j->unique_file) { |
| j->unique_file = hashmap_first(j->files); |
| if (!j->unique_file) |
| return 0; |
| j->unique_offset = 0; |
| } |
| |
| for (;;) { |
| JournalFile *of; |
| Iterator i; |
| const void *odata; |
| size_t ol; |
| bool found; |
| |
| /* Proceed to next data object in the field's linked list */ |
| if (j->unique_offset == 0) { |
| r = journal_file_find_field_object(j->unique_file, j->unique_field, k, &o, NULL); |
| if (r < 0) |
| return r; |
| |
| j->unique_offset = r > 0 ? le64toh(o->field.head_data_offset) : 0; |
| } else { |
| r = journal_file_move_to_object(j->unique_file, OBJECT_DATA, j->unique_offset, &o); |
| if (r < 0) |
| return r; |
| |
| j->unique_offset = le64toh(o->data.next_field_offset); |
| } |
| |
| /* We reached the end of the list? Then start again, with the next file */ |
| if (j->unique_offset == 0) { |
| JournalFile *n; |
| |
| n = hashmap_next(j->files, j->unique_file->path); |
| if (!n) |
| return 0; |
| |
| j->unique_file = n; |
| continue; |
| } |
| |
| /* We do not use the type context here, but 0 instead, |
| * so that we can look at this data object at the same |
| * time as one on another file */ |
| r = journal_file_move_to_object(j->unique_file, 0, j->unique_offset, &o); |
| if (r < 0) |
| return r; |
| |
| /* Let's do the type check by hand, since we used 0 context above. */ |
| if (o->object.type != OBJECT_DATA) |
| return -EBADMSG; |
| |
| r = return_data(j, j->unique_file, o, &odata, &ol); |
| if (r < 0) |
| return r; |
| |
| /* OK, now let's see if we already returned this data |
| * object by checking if it exists in the earlier |
| * traversed files. */ |
| found = false; |
| HASHMAP_FOREACH(of, j->files, i) { |
| Object *oo; |
| uint64_t op; |
| |
| if (of == j->unique_file) |
| break; |
| |
| /* Skip this file it didn't have any fields |
| * indexed */ |
| if (JOURNAL_HEADER_CONTAINS(of->header, n_fields) && |
| le64toh(of->header->n_fields) <= 0) |
| continue; |
| |
| r = journal_file_find_data_object_with_hash(of, odata, ol, le64toh(o->data.hash), &oo, &op); |
| if (r < 0) |
| return r; |
| |
| if (r > 0) |
| found = true; |
| } |
| |
| if (found) |
| continue; |
| |
| r = return_data(j, j->unique_file, o, data, l); |
| if (r < 0) |
| return r; |
| |
| return 1; |
| } |
| } |
| |
| _public_ void sd_journal_restart_unique(sd_journal *j) { |
| if (!j) |
| return; |
| |
| j->unique_file = NULL; |
| j->unique_offset = 0; |
| } |
| |
| _public_ int sd_journal_reliable_fd(sd_journal *j) { |
| if (!j) |
| return -EINVAL; |
| |
| return !j->on_network; |
| } |
| |
| static char *lookup_field(const char *field, void *userdata) { |
| sd_journal *j = userdata; |
| const void *data; |
| size_t size, d; |
| int r; |
| |
| assert(field); |
| assert(j); |
| |
| r = sd_journal_get_data(j, field, &data, &size); |
| if (r < 0 || |
| size > REPLACE_VAR_MAX) |
| return strdup(field); |
| |
| d = strlen(field) + 1; |
| |
| return strndup((const char*) data + d, size - d); |
| } |
| |
| _public_ int sd_journal_get_catalog(sd_journal *j, char **ret) { |
| const void *data; |
| size_t size; |
| sd_id128_t id; |
| _cleanup_free_ char *text = NULL, *cid = NULL; |
| char *t; |
| int r; |
| |
| if (!j) |
| return -EINVAL; |
| if (!ret) |
| return -EINVAL; |
| |
| r = sd_journal_get_data(j, "MESSAGE_ID", &data, &size); |
| if (r < 0) |
| return r; |
| |
| cid = strndup((const char*) data + 11, size - 11); |
| if (!cid) |
| return -ENOMEM; |
| |
| r = sd_id128_from_string(cid, &id); |
| if (r < 0) |
| return r; |
| |
| r = catalog_get(CATALOG_DATABASE, id, &text); |
| if (r < 0) |
| return r; |
| |
| t = replace_var(text, lookup_field, j); |
| if (!t) |
| return -ENOMEM; |
| |
| *ret = t; |
| return 0; |
| } |
| |
| _public_ int sd_journal_get_catalog_for_message_id(sd_id128_t id, char **ret) { |
| if (!ret) |
| return -EINVAL; |
| |
| return catalog_get(CATALOG_DATABASE, id, ret); |
| } |
| |
| _public_ int sd_journal_set_data_threshold(sd_journal *j, size_t sz) { |
| if (!j) |
| return -EINVAL; |
| |
| j->data_threshold = sz; |
| return 0; |
| } |
| |
| _public_ int sd_journal_get_data_threshold(sd_journal *j, size_t *sz) { |
| if (!j) |
| return -EINVAL; |
| if (!sz) |
| return -EINVAL; |
| |
| *sz = j->data_threshold; |
| return 0; |
| } |