2 Unix SMB/CIFS implementation.
4 trivial database library
6 Copyright (C) Andrew Tridgell 2005
8 ** NOTE! The following LGPL license applies to the tdb
9 ** library. This does NOT imply that all of Samba is released
12 This library is free software; you can redistribute it and/or
13 modify it under the terms of the GNU Lesser General Public
14 License as published by the Free Software Foundation; either
15 version 3 of the License, or (at your option) any later version.
17 This library is distributed in the hope that it will be useful,
18 but WITHOUT ANY WARRANTY; without even the implied warranty of
19 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
20 Lesser General Public License for more details.
22 You should have received a copy of the GNU Lesser General Public
23 License along with this library; if not, see <http://www.gnu.org/licenses/>.
26 #include "tdb_private.h"
31 - only allow a single transaction at a time per database. This makes
32 using the transaction API simpler, as otherwise the caller would
33 have to cope with temporary failures in transactions that conflict
34 with other current transactions
36 - keep the transaction recovery information in the same file as the
37 database, using a special 'transaction recovery' record pointed at
38 by the header. This removes the need for extra journal files as
39 used by some other databases
41 - dynamically allocated the transaction recover record, re-using it
42 for subsequent transactions. If a larger record is needed then
43 tdb_free() the old record to place it on the normal tdb freelist
44 before allocating the new record
46 - during transactions, keep a linked list of writes all that have
47 been performed by intercepting all tdb_write() calls. The hooked
48 transaction versions of tdb_read() and tdb_write() check this
49 linked list and try to use the elements of the list in preference
52 - don't allow any locks to be held when a transaction starts,
53 otherwise we can end up with deadlock (plus lack of lock nesting
54 in posix locks would mean the lock is lost)
56 - if the caller gains a lock during the transaction but doesn't
57 release it then fail the commit
59 - allow for nested calls to tdb_transaction_start(), re-using the
60 existing transaction record. If the inner transaction is cancelled
61 then a subsequent commit will fail
63 - keep a mirrored copy of the tdb hash chain heads to allow for the
64 fast hash heads scan on traverse, updating the mirrored copy in
65 the transaction version of tdb_write
67 - allow callers to mix transaction and non-transaction use of tdb,
68 although once a transaction is started then an exclusive lock is
69 gained until the transaction is committed or cancelled
71 - the commit stategy involves first saving away all modified data
72 into a linearised buffer in the transaction recovery area, then
73 marking the transaction recovery area with a magic value to
74 indicate a valid recovery record. In total 4 fsync/msync calls are
75 needed per commit to prevent race conditions. It might be possible
76 to reduce this to 3 or even 2 with some more work.
78 - check for a valid recovery record on open of the tdb, while the
79 global lock is held. Automatically recover from the transaction
80 recovery area if needed, then continue with the open as
81 usual. This allows for smooth crash recovery with no administrator
84 - if TDB_NOSYNC is passed to flags in tdb_open then transactions are
85 still available, but no transaction recovery area is used and no
86 fsync/msync calls are made.
88 - if TDB_NO_NESTING is passed to flags in tdb open then transaction
89 nesting is disabled. tdb_transaction_start() will then implicitely
90 cancel any pending transactions and always start a new transaction
91 context instead of nesting.
97 hold the context of any current transaction
99 struct tdb_transaction {
100 /* we keep a mirrored copy of the tdb hash heads here so
101 tdb_next_hash_chain() can operate efficiently */
102 uint32_t *hash_heads;
104 /* the original io methods - used to do IOs to the real db */
105 const struct tdb_methods *io_methods;
107 /* the list of transaction blocks. When a block is first
108 written to, it gets created in this list */
111 uint32_t block_size; /* bytes in each block */
112 uint32_t last_block_size; /* number of valid bytes in the last block */
114 /* non-zero when an internal transaction error has
115 occurred. All write operations will then fail until the
116 transaction is ended */
117 int transaction_error;
119 /* when inside a transaction we need to keep track of any
120 nested tdb_transaction_start() calls, as these are allowed,
121 but don't create a new transaction */
124 /* set when a prepare has already occurred */
126 tdb_off_t magic_offset;
128 /* old file size before transaction */
129 tdb_len_t old_map_size;
134 read while in a transaction. We need to check first if the data is in our list
135 of transaction elements, then if not do a real read
137 static int transaction_read(struct tdb_context *tdb, tdb_off_t off, void *buf,
138 tdb_len_t len, int cv)
142 /* Only a commit is allowed on a prepared transaction */
143 if (tdb->transaction->prepared) {
144 tdb->ecode = TDB_ERR_EINVAL;
145 TDB_LOG((tdb, TDB_DEBUG_FATAL, "transaction_read: transaction already prepared, read not allowed\n"));
146 tdb->transaction->transaction_error = 1;
150 /* break it down into block sized ops */
151 while (len + (off % tdb->transaction->block_size) > tdb->transaction->block_size) {
152 tdb_len_t len2 = tdb->transaction->block_size - (off % tdb->transaction->block_size);
153 if (transaction_read(tdb, off, buf, len2, cv) != 0) {
158 buf = (void *)(len2 + (char *)buf);
165 blk = off / tdb->transaction->block_size;
167 /* see if we have it in the block list */
168 if (tdb->transaction->num_blocks <= blk ||
169 tdb->transaction->blocks[blk] == NULL) {
170 /* nope, do a real read */
171 if (tdb->transaction->io_methods->tdb_read(tdb, off, buf, len, cv) != 0) {
177 /* it is in the block list. Now check for the last block */
178 if (blk == tdb->transaction->num_blocks-1) {
179 if (len > tdb->transaction->last_block_size) {
184 /* now copy it out of this block */
185 memcpy(buf, tdb->transaction->blocks[blk] + (off % tdb->transaction->block_size), len);
187 tdb_convert(buf, len);
192 TDB_LOG((tdb, TDB_DEBUG_FATAL, "transaction_read: failed at off=%d len=%d\n", off, len));
193 tdb->ecode = TDB_ERR_IO;
194 tdb->transaction->transaction_error = 1;
200 write while in a transaction
202 static int transaction_write(struct tdb_context *tdb, tdb_off_t off,
203 const void *buf, tdb_len_t len)
207 /* Only a commit is allowed on a prepared transaction */
208 if (tdb->transaction->prepared) {
209 tdb->ecode = TDB_ERR_EINVAL;
210 TDB_LOG((tdb, TDB_DEBUG_FATAL, "transaction_write: transaction already prepared, write not allowed\n"));
211 tdb->transaction->transaction_error = 1;
215 /* if the write is to a hash head, then update the transaction
217 if (len == sizeof(tdb_off_t) && off >= FREELIST_TOP &&
218 off < FREELIST_TOP+TDB_HASHTABLE_SIZE(tdb)) {
219 uint32_t chain = (off-FREELIST_TOP) / sizeof(tdb_off_t);
220 memcpy(&tdb->transaction->hash_heads[chain], buf, len);
223 /* break it up into block sized chunks */
224 while (len + (off % tdb->transaction->block_size) > tdb->transaction->block_size) {
225 tdb_len_t len2 = tdb->transaction->block_size - (off % tdb->transaction->block_size);
226 if (transaction_write(tdb, off, buf, len2) != 0) {
232 buf = (const void *)(len2 + (const char *)buf);
240 blk = off / tdb->transaction->block_size;
241 off = off % tdb->transaction->block_size;
243 if (tdb->transaction->num_blocks <= blk) {
244 uint8_t **new_blocks;
245 /* expand the blocks array */
246 if (tdb->transaction->blocks == NULL) {
247 new_blocks = (uint8_t **)malloc(
248 (blk+1)*sizeof(uint8_t *));
250 new_blocks = (uint8_t **)realloc(
251 tdb->transaction->blocks,
252 (blk+1)*sizeof(uint8_t *));
254 if (new_blocks == NULL) {
255 tdb->ecode = TDB_ERR_OOM;
258 memset(&new_blocks[tdb->transaction->num_blocks], 0,
259 (1+(blk - tdb->transaction->num_blocks))*sizeof(uint8_t *));
260 tdb->transaction->blocks = new_blocks;
261 tdb->transaction->num_blocks = blk+1;
262 tdb->transaction->last_block_size = 0;
265 /* allocate and fill a block? */
266 if (tdb->transaction->blocks[blk] == NULL) {
267 tdb->transaction->blocks[blk] = (uint8_t *)calloc(tdb->transaction->block_size, 1);
268 if (tdb->transaction->blocks[blk] == NULL) {
269 tdb->ecode = TDB_ERR_OOM;
270 tdb->transaction->transaction_error = 1;
273 if (tdb->transaction->old_map_size > blk * tdb->transaction->block_size) {
274 tdb_len_t len2 = tdb->transaction->block_size;
275 if (len2 + (blk * tdb->transaction->block_size) > tdb->transaction->old_map_size) {
276 len2 = tdb->transaction->old_map_size - (blk * tdb->transaction->block_size);
278 if (tdb->transaction->io_methods->tdb_read(tdb, blk * tdb->transaction->block_size,
279 tdb->transaction->blocks[blk],
281 SAFE_FREE(tdb->transaction->blocks[blk]);
282 tdb->ecode = TDB_ERR_IO;
285 if (blk == tdb->transaction->num_blocks-1) {
286 tdb->transaction->last_block_size = len2;
291 /* overwrite part of an existing block */
293 memset(tdb->transaction->blocks[blk] + off, 0, len);
295 memcpy(tdb->transaction->blocks[blk] + off, buf, len);
297 if (blk == tdb->transaction->num_blocks-1) {
298 if (len + off > tdb->transaction->last_block_size) {
299 tdb->transaction->last_block_size = len + off;
306 TDB_LOG((tdb, TDB_DEBUG_FATAL, "transaction_write: failed at off=%d len=%d\n",
307 (blk*tdb->transaction->block_size) + off, len));
308 tdb->transaction->transaction_error = 1;
314 write while in a transaction - this varient never expands the transaction blocks, it only
315 updates existing blocks. This means it cannot change the recovery size
317 static int transaction_write_existing(struct tdb_context *tdb, tdb_off_t off,
318 const void *buf, tdb_len_t len)
322 /* break it up into block sized chunks */
323 while (len + (off % tdb->transaction->block_size) > tdb->transaction->block_size) {
324 tdb_len_t len2 = tdb->transaction->block_size - (off % tdb->transaction->block_size);
325 if (transaction_write_existing(tdb, off, buf, len2) != 0) {
331 buf = (const void *)(len2 + (const char *)buf);
339 blk = off / tdb->transaction->block_size;
340 off = off % tdb->transaction->block_size;
342 if (tdb->transaction->num_blocks <= blk ||
343 tdb->transaction->blocks[blk] == NULL) {
347 if (blk == tdb->transaction->num_blocks-1 &&
348 off + len > tdb->transaction->last_block_size) {
349 if (off >= tdb->transaction->last_block_size) {
352 len = tdb->transaction->last_block_size - off;
355 /* overwrite part of an existing block */
356 memcpy(tdb->transaction->blocks[blk] + off, buf, len);
363 accelerated hash chain head search, using the cached hash heads
365 static void transaction_next_hash_chain(struct tdb_context *tdb, uint32_t *chain)
368 for (;h < tdb->header.hash_size;h++) {
369 /* the +1 takes account of the freelist */
370 if (0 != tdb->transaction->hash_heads[h+1]) {
378 out of bounds check during a transaction
380 static int transaction_oob(struct tdb_context *tdb, tdb_off_t len, int probe)
382 if (len <= tdb->map_size) {
385 return TDB_ERRCODE(TDB_ERR_IO, -1);
389 transaction version of tdb_expand().
391 static int transaction_expand_file(struct tdb_context *tdb, tdb_off_t size,
394 /* add a write to the transaction elements, so subsequent
395 reads see the zero data */
396 if (transaction_write(tdb, size, NULL, addition) != 0) {
404 brlock during a transaction - ignore them
406 static int transaction_brlock(struct tdb_context *tdb, tdb_off_t offset,
407 int rw_type, int lck_type, int probe, size_t len)
412 static const struct tdb_methods transaction_methods = {
415 transaction_next_hash_chain,
417 transaction_expand_file,
424 static int transaction_sync(struct tdb_context *tdb, tdb_off_t offset, tdb_len_t length)
426 if (fsync(tdb->fd) != 0) {
427 tdb->ecode = TDB_ERR_IO;
428 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction: fsync failed\n"));
433 tdb_off_t moffset = offset & ~(tdb->page_size-1);
434 if (msync(moffset + (char *)tdb->map_ptr,
435 length + (offset - moffset), MS_SYNC) != 0) {
436 tdb->ecode = TDB_ERR_IO;
437 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction: msync failed - %s\n",
446 int tdb_transaction_cancel_internal(struct tdb_context *tdb)
450 if (tdb->transaction == NULL) {
451 TDB_LOG((tdb, TDB_DEBUG_ERROR, "tdb_transaction_cancel: no transaction\n"));
455 if (tdb->transaction->nesting != 0) {
456 tdb->transaction->transaction_error = 1;
457 tdb->transaction->nesting--;
461 tdb->map_size = tdb->transaction->old_map_size;
463 /* free all the transaction blocks */
464 for (i=0;i<tdb->transaction->num_blocks;i++) {
465 if (tdb->transaction->blocks[i] != NULL) {
466 free(tdb->transaction->blocks[i]);
469 SAFE_FREE(tdb->transaction->blocks);
471 if (tdb->transaction->magic_offset) {
472 const struct tdb_methods *methods = tdb->transaction->io_methods;
475 /* remove the recovery marker */
476 if (methods->tdb_write(tdb, tdb->transaction->magic_offset, &zero, 4) == -1 ||
477 transaction_sync(tdb, tdb->transaction->magic_offset, 4) == -1) {
478 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_cancel: failed to remove recovery magic\n"));
483 /* remove any global lock created during the transaction */
484 if (tdb->global_lock.count != 0) {
485 tdb_brlock(tdb, FREELIST_TOP, F_UNLCK, F_SETLKW, 0, 4*tdb->header.hash_size);
486 tdb->global_lock.count = 0;
489 /* remove any locks created during the transaction */
490 if (tdb->num_locks != 0) {
491 for (i=0;i<tdb->num_lockrecs;i++) {
492 tdb_brlock(tdb,FREELIST_TOP+4*tdb->lockrecs[i].list,
493 F_UNLCK,F_SETLKW, 0, 1);
496 tdb->num_lockrecs = 0;
497 SAFE_FREE(tdb->lockrecs);
500 /* restore the normal io methods */
501 tdb->methods = tdb->transaction->io_methods;
503 tdb_brlock(tdb, FREELIST_TOP, F_UNLCK, F_SETLKW, 0, 0);
504 tdb_transaction_unlock(tdb);
505 SAFE_FREE(tdb->transaction->hash_heads);
506 SAFE_FREE(tdb->transaction);
512 start a tdb transaction. No token is returned, as only a single
513 transaction is allowed to be pending per tdb_context
515 int tdb_transaction_start(struct tdb_context *tdb)
517 /* some sanity checks */
518 if (tdb->read_only || (tdb->flags & TDB_INTERNAL) || tdb->traverse_read) {
519 TDB_LOG((tdb, TDB_DEBUG_ERROR, "tdb_transaction_start: cannot start a transaction on a read-only or internal db\n"));
520 tdb->ecode = TDB_ERR_EINVAL;
524 /* cope with nested tdb_transaction_start() calls */
525 if (tdb->transaction != NULL) {
526 tdb_trace(tdb, "tdb_transaction_start");
527 if (!tdb->flags & TDB_NO_NESTING) {
528 tdb->transaction->nesting++;
529 TDB_LOG((tdb, TDB_DEBUG_TRACE, "tdb_transaction_start: nesting %d\n",
530 tdb->transaction->nesting));
533 tdb_transaction_cancel_internal(tdb);
534 TDB_LOG((tdb, TDB_DEBUG_TRACE, "tdb_transaction_start: cancelling previous transaction\n"));
538 if (tdb->num_locks != 0 || tdb->global_lock.count) {
539 /* the caller must not have any locks when starting a
540 transaction as otherwise we'll be screwed by lack
541 of nested locks in posix */
542 TDB_LOG((tdb, TDB_DEBUG_ERROR, "tdb_transaction_start: cannot start a transaction with locks held\n"));
543 tdb->ecode = TDB_ERR_LOCK;
547 if (tdb->travlocks.next != NULL) {
548 /* you cannot use transactions inside a traverse (although you can use
549 traverse inside a transaction) as otherwise you can end up with
551 TDB_LOG((tdb, TDB_DEBUG_ERROR, "tdb_transaction_start: cannot start a transaction within a traverse\n"));
552 tdb->ecode = TDB_ERR_LOCK;
556 tdb->transaction = (struct tdb_transaction *)
557 calloc(sizeof(struct tdb_transaction), 1);
558 if (tdb->transaction == NULL) {
559 tdb->ecode = TDB_ERR_OOM;
563 /* a page at a time seems like a reasonable compromise between compactness and efficiency */
564 tdb->transaction->block_size = tdb->page_size;
566 /* get the transaction write lock. This is a blocking lock. As
567 discussed with Volker, there are a number of ways we could
568 make this async, which we will probably do in the future */
569 if (tdb_transaction_lock(tdb, F_WRLCK) == -1) {
570 SAFE_FREE(tdb->transaction->blocks);
571 SAFE_FREE(tdb->transaction);
575 /* get a read lock from the freelist to the end of file. This
576 is upgraded to a write lock during the commit */
577 if (tdb_brlock(tdb, FREELIST_TOP, F_RDLCK, F_SETLKW, 0, 0) == -1) {
578 TDB_LOG((tdb, TDB_DEBUG_ERROR, "tdb_transaction_start: failed to get hash locks\n"));
579 tdb->ecode = TDB_ERR_LOCK;
583 /* setup a copy of the hash table heads so the hash scan in
584 traverse can be fast */
585 tdb->transaction->hash_heads = (uint32_t *)
586 calloc(tdb->header.hash_size+1, sizeof(uint32_t));
587 if (tdb->transaction->hash_heads == NULL) {
588 tdb->ecode = TDB_ERR_OOM;
591 if (tdb->methods->tdb_read(tdb, FREELIST_TOP, tdb->transaction->hash_heads,
592 TDB_HASHTABLE_SIZE(tdb), 0) != 0) {
593 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_start: failed to read hash heads\n"));
594 tdb->ecode = TDB_ERR_IO;
598 /* make sure we know about any file expansions already done by
600 tdb->methods->tdb_oob(tdb, tdb->map_size + 1, 1);
601 tdb->transaction->old_map_size = tdb->map_size;
603 /* finally hook the io methods, replacing them with
604 transaction specific methods */
605 tdb->transaction->io_methods = tdb->methods;
606 tdb->methods = &transaction_methods;
608 /* Trace at the end, so we get sequence number correct. */
609 tdb_trace(tdb, "tdb_transaction_start");
613 tdb_brlock(tdb, FREELIST_TOP, F_UNLCK, F_SETLKW, 0, 0);
614 tdb_transaction_unlock(tdb);
615 SAFE_FREE(tdb->transaction->blocks);
616 SAFE_FREE(tdb->transaction->hash_heads);
617 SAFE_FREE(tdb->transaction);
623 cancel the current transaction
625 int tdb_transaction_cancel(struct tdb_context *tdb)
627 tdb_trace(tdb, "tdb_transaction_cancel");
628 return tdb_transaction_cancel_internal(tdb);
632 work out how much space the linearised recovery data will consume
634 static tdb_len_t tdb_recovery_size(struct tdb_context *tdb)
636 tdb_len_t recovery_size = 0;
639 recovery_size = sizeof(uint32_t);
640 for (i=0;i<tdb->transaction->num_blocks;i++) {
641 if (i * tdb->transaction->block_size >= tdb->transaction->old_map_size) {
644 if (tdb->transaction->blocks[i] == NULL) {
647 recovery_size += 2*sizeof(tdb_off_t);
648 if (i == tdb->transaction->num_blocks-1) {
649 recovery_size += tdb->transaction->last_block_size;
651 recovery_size += tdb->transaction->block_size;
655 return recovery_size;
659 allocate the recovery area, or use an existing recovery area if it is
662 static int tdb_recovery_allocate(struct tdb_context *tdb,
663 tdb_len_t *recovery_size,
664 tdb_off_t *recovery_offset,
665 tdb_len_t *recovery_max_size)
667 struct list_struct rec;
668 const struct tdb_methods *methods = tdb->transaction->io_methods;
669 tdb_off_t recovery_head;
671 if (tdb_ofs_read(tdb, TDB_RECOVERY_HEAD, &recovery_head) == -1) {
672 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_recovery_allocate: failed to read recovery head\n"));
678 if (recovery_head != 0 &&
679 methods->tdb_read(tdb, recovery_head, &rec, sizeof(rec), DOCONV()) == -1) {
680 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_recovery_allocate: failed to read recovery record\n"));
684 *recovery_size = tdb_recovery_size(tdb);
686 if (recovery_head != 0 && *recovery_size <= rec.rec_len) {
687 /* it fits in the existing area */
688 *recovery_max_size = rec.rec_len;
689 *recovery_offset = recovery_head;
693 /* we need to free up the old recovery area, then allocate a
694 new one at the end of the file. Note that we cannot use
695 tdb_allocate() to allocate the new one as that might return
696 us an area that is being currently used (as of the start of
698 if (recovery_head != 0) {
699 if (tdb_free(tdb, recovery_head, &rec) == -1) {
700 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_recovery_allocate: failed to free previous recovery area\n"));
705 /* the tdb_free() call might have increased the recovery size */
706 *recovery_size = tdb_recovery_size(tdb);
708 /* round up to a multiple of page size */
709 *recovery_max_size = TDB_ALIGN(sizeof(rec) + *recovery_size, tdb->page_size) - sizeof(rec);
710 *recovery_offset = tdb->map_size;
711 recovery_head = *recovery_offset;
713 if (methods->tdb_expand_file(tdb, tdb->transaction->old_map_size,
714 (tdb->map_size - tdb->transaction->old_map_size) +
715 sizeof(rec) + *recovery_max_size) == -1) {
716 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_recovery_allocate: failed to create recovery area\n"));
720 /* remap the file (if using mmap) */
721 methods->tdb_oob(tdb, tdb->map_size + 1, 1);
723 /* we have to reset the old map size so that we don't try to expand the file
724 again in the transaction commit, which would destroy the recovery area */
725 tdb->transaction->old_map_size = tdb->map_size;
727 /* write the recovery header offset and sync - we can sync without a race here
728 as the magic ptr in the recovery record has not been set */
729 CONVERT(recovery_head);
730 if (methods->tdb_write(tdb, TDB_RECOVERY_HEAD,
731 &recovery_head, sizeof(tdb_off_t)) == -1) {
732 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_recovery_allocate: failed to write recovery head\n"));
735 if (transaction_write_existing(tdb, TDB_RECOVERY_HEAD, &recovery_head, sizeof(tdb_off_t)) == -1) {
736 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_recovery_allocate: failed to write recovery head\n"));
745 setup the recovery data that will be used on a crash during commit
747 static int transaction_setup_recovery(struct tdb_context *tdb,
748 tdb_off_t *magic_offset)
750 tdb_len_t recovery_size;
751 unsigned char *data, *p;
752 const struct tdb_methods *methods = tdb->transaction->io_methods;
753 struct list_struct *rec;
754 tdb_off_t recovery_offset, recovery_max_size;
755 tdb_off_t old_map_size = tdb->transaction->old_map_size;
756 uint32_t magic, tailer;
760 check that the recovery area has enough space
762 if (tdb_recovery_allocate(tdb, &recovery_size,
763 &recovery_offset, &recovery_max_size) == -1) {
767 data = (unsigned char *)malloc(recovery_size + sizeof(*rec));
769 tdb->ecode = TDB_ERR_OOM;
773 rec = (struct list_struct *)data;
774 memset(rec, 0, sizeof(*rec));
777 rec->data_len = recovery_size;
778 rec->rec_len = recovery_max_size;
779 rec->key_len = old_map_size;
782 /* build the recovery data into a single blob to allow us to do a single
783 large write, which should be more efficient */
784 p = data + sizeof(*rec);
785 for (i=0;i<tdb->transaction->num_blocks;i++) {
789 if (tdb->transaction->blocks[i] == NULL) {
793 offset = i * tdb->transaction->block_size;
794 length = tdb->transaction->block_size;
795 if (i == tdb->transaction->num_blocks-1) {
796 length = tdb->transaction->last_block_size;
799 if (offset >= old_map_size) {
802 if (offset + length > tdb->transaction->old_map_size) {
803 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_setup_recovery: transaction data over new region boundary\n"));
805 tdb->ecode = TDB_ERR_CORRUPT;
808 memcpy(p, &offset, 4);
809 memcpy(p+4, &length, 4);
813 /* the recovery area contains the old data, not the
814 new data, so we have to call the original tdb_read
816 if (methods->tdb_read(tdb, offset, p + 8, length, 0) != 0) {
818 tdb->ecode = TDB_ERR_IO;
825 tailer = sizeof(*rec) + recovery_max_size;
826 memcpy(p, &tailer, 4);
829 /* write the recovery data to the recovery area */
830 if (methods->tdb_write(tdb, recovery_offset, data, sizeof(*rec) + recovery_size) == -1) {
831 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_setup_recovery: failed to write recovery data\n"));
833 tdb->ecode = TDB_ERR_IO;
836 if (transaction_write_existing(tdb, recovery_offset, data, sizeof(*rec) + recovery_size) == -1) {
837 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_setup_recovery: failed to write secondary recovery data\n"));
839 tdb->ecode = TDB_ERR_IO;
843 /* as we don't have ordered writes, we have to sync the recovery
844 data before we update the magic to indicate that the recovery
846 if (transaction_sync(tdb, recovery_offset, sizeof(*rec) + recovery_size) == -1) {
853 magic = TDB_RECOVERY_MAGIC;
856 *magic_offset = recovery_offset + offsetof(struct list_struct, magic);
858 if (methods->tdb_write(tdb, *magic_offset, &magic, sizeof(magic)) == -1) {
859 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_setup_recovery: failed to write recovery magic\n"));
860 tdb->ecode = TDB_ERR_IO;
863 if (transaction_write_existing(tdb, *magic_offset, &magic, sizeof(magic)) == -1) {
864 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_setup_recovery: failed to write secondary recovery magic\n"));
865 tdb->ecode = TDB_ERR_IO;
869 /* ensure the recovery magic marker is on disk */
870 if (transaction_sync(tdb, *magic_offset, sizeof(magic)) == -1) {
877 static int tdb_transaction_prepare_commit_internal(struct tdb_context *tdb)
879 const struct tdb_methods *methods;
881 if (tdb->transaction == NULL) {
882 TDB_LOG((tdb, TDB_DEBUG_ERROR, "tdb_transaction_prepare_commit: no transaction\n"));
886 if (tdb->transaction->prepared) {
887 tdb->ecode = TDB_ERR_EINVAL;
888 tdb_transaction_cancel(tdb);
889 TDB_LOG((tdb, TDB_DEBUG_ERROR, "tdb_transaction_prepare_commit: transaction already prepared\n"));
893 if (tdb->transaction->transaction_error) {
894 tdb->ecode = TDB_ERR_IO;
895 tdb_transaction_cancel_internal(tdb);
896 TDB_LOG((tdb, TDB_DEBUG_ERROR, "tdb_transaction_prepare_commit: transaction error pending\n"));
901 if (tdb->transaction->nesting != 0) {
902 tdb->transaction->nesting--;
907 /* store seqnum now, before reading becomes illegal. */
908 tdb_ofs_read(tdb, TDB_SEQNUM_OFS, &tdb->transaction_prepare_seqnum);
911 /* check for a null transaction */
912 if (tdb->transaction->blocks == NULL) {
916 methods = tdb->transaction->io_methods;
918 /* if there are any locks pending then the caller has not
919 nested their locks properly, so fail the transaction */
920 if (tdb->num_locks || tdb->global_lock.count) {
921 tdb->ecode = TDB_ERR_LOCK;
922 TDB_LOG((tdb, TDB_DEBUG_ERROR, "tdb_transaction_prepare_commit: locks pending on commit\n"));
923 tdb_transaction_cancel_internal(tdb);
927 /* upgrade the main transaction lock region to a write lock */
928 if (tdb_brlock_upgrade(tdb, FREELIST_TOP, 0) == -1) {
929 TDB_LOG((tdb, TDB_DEBUG_ERROR, "tdb_transaction_prepare_commit: failed to upgrade hash locks\n"));
930 tdb->ecode = TDB_ERR_LOCK;
931 tdb_transaction_cancel_internal(tdb);
935 /* get the global lock - this prevents new users attaching to the database
937 if (tdb_brlock(tdb, GLOBAL_LOCK, F_WRLCK, F_SETLKW, 0, 1) == -1) {
938 TDB_LOG((tdb, TDB_DEBUG_ERROR, "tdb_transaction_prepare_commit: failed to get global lock\n"));
939 tdb->ecode = TDB_ERR_LOCK;
940 tdb_transaction_cancel_internal(tdb);
944 if (!(tdb->flags & TDB_NOSYNC)) {
945 /* write the recovery data to the end of the file */
946 if (transaction_setup_recovery(tdb, &tdb->transaction->magic_offset) == -1) {
947 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_prepare_commit: failed to setup recovery data\n"));
948 tdb_brlock(tdb, GLOBAL_LOCK, F_UNLCK, F_SETLKW, 0, 1);
949 tdb_transaction_cancel_internal(tdb);
954 tdb->transaction->prepared = true;
956 /* expand the file to the new size if needed */
957 if (tdb->map_size != tdb->transaction->old_map_size) {
958 if (methods->tdb_expand_file(tdb, tdb->transaction->old_map_size,
960 tdb->transaction->old_map_size) == -1) {
961 tdb->ecode = TDB_ERR_IO;
962 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_prepare_commit: expansion failed\n"));
963 tdb_brlock(tdb, GLOBAL_LOCK, F_UNLCK, F_SETLKW, 0, 1);
964 tdb_transaction_cancel_internal(tdb);
967 tdb->map_size = tdb->transaction->old_map_size;
968 methods->tdb_oob(tdb, tdb->map_size + 1, 1);
971 /* Keep the global lock until the actual commit */
977 prepare to commit the current transaction
979 int tdb_transaction_prepare_commit(struct tdb_context *tdb)
981 tdb_trace(tdb, "tdb_transaction_prepare_commit");
982 return tdb_transaction_prepare_commit_internal(tdb);
986 commit the current transaction
988 int tdb_transaction_commit(struct tdb_context *tdb)
990 const struct tdb_methods *methods;
993 if (tdb->transaction == NULL) {
994 TDB_LOG((tdb, TDB_DEBUG_ERROR, "tdb_transaction_commit: no transaction\n"));
998 /* If we've prepared, can't read seqnum. */
999 if (tdb->transaction->prepared) {
1000 tdb_trace_seqnum(tdb, tdb->transaction_prepare_seqnum,
1001 "tdb_transaction_commit");
1003 tdb_trace(tdb, "tdb_transaction_commit");
1006 if (tdb->transaction->transaction_error) {
1007 tdb->ecode = TDB_ERR_IO;
1008 tdb_transaction_cancel(tdb);
1009 TDB_LOG((tdb, TDB_DEBUG_ERROR, "tdb_transaction_commit: transaction error pending\n"));
1014 if (tdb->transaction->nesting != 0) {
1015 tdb->transaction->nesting--;
1019 /* check for a null transaction */
1020 if (tdb->transaction->blocks == NULL) {
1021 tdb_transaction_cancel_internal(tdb);
1025 if (!tdb->transaction->prepared) {
1026 int ret = tdb_transaction_prepare_commit_internal(tdb);
1031 methods = tdb->transaction->io_methods;
1033 /* perform all the writes */
1034 for (i=0;i<tdb->transaction->num_blocks;i++) {
1038 if (tdb->transaction->blocks[i] == NULL) {
1042 offset = i * tdb->transaction->block_size;
1043 length = tdb->transaction->block_size;
1044 if (i == tdb->transaction->num_blocks-1) {
1045 length = tdb->transaction->last_block_size;
1048 if (methods->tdb_write(tdb, offset, tdb->transaction->blocks[i], length) == -1) {
1049 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_commit: write failed during commit\n"));
1051 /* we've overwritten part of the data and
1052 possibly expanded the file, so we need to
1053 run the crash recovery code */
1054 tdb->methods = methods;
1055 tdb_transaction_recover(tdb);
1057 tdb_transaction_cancel_internal(tdb);
1058 tdb_brlock(tdb, GLOBAL_LOCK, F_UNLCK, F_SETLKW, 0, 1);
1060 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_commit: write failed\n"));
1063 SAFE_FREE(tdb->transaction->blocks[i]);
1066 SAFE_FREE(tdb->transaction->blocks);
1067 tdb->transaction->num_blocks = 0;
1069 if (!(tdb->flags & TDB_NOSYNC)) {
1070 /* ensure the new data is on disk */
1071 if (transaction_sync(tdb, 0, tdb->map_size) == -1) {
1076 tdb_brlock(tdb, GLOBAL_LOCK, F_UNLCK, F_SETLKW, 0, 1);
1079 TODO: maybe write to some dummy hdr field, or write to magic
1080 offset without mmap, before the last sync, instead of the
1084 /* on some systems (like Linux 2.6.x) changes via mmap/msync
1085 don't change the mtime of the file, this means the file may
1086 not be backed up (as tdb rounding to block sizes means that
1087 file size changes are quite rare too). The following forces
1088 mtime changes when a transaction completes */
1090 utime(tdb->name, NULL);
1093 /* use a transaction cancel to free memory and remove the
1094 transaction locks */
1095 tdb_transaction_cancel_internal(tdb);
1102 recover from an aborted transaction. Must be called with exclusive
1103 database write access already established (including the global
1104 lock to prevent new processes attaching)
1106 int tdb_transaction_recover(struct tdb_context *tdb)
1108 tdb_off_t recovery_head, recovery_eof;
1109 unsigned char *data, *p;
1111 struct list_struct rec;
1113 /* find the recovery area */
1114 if (tdb_ofs_read(tdb, TDB_RECOVERY_HEAD, &recovery_head) == -1) {
1115 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_recover: failed to read recovery head\n"));
1116 tdb->ecode = TDB_ERR_IO;
1120 if (recovery_head == 0) {
1121 /* we have never allocated a recovery record */
1125 /* read the recovery record */
1126 if (tdb->methods->tdb_read(tdb, recovery_head, &rec,
1127 sizeof(rec), DOCONV()) == -1) {
1128 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_recover: failed to read recovery record\n"));
1129 tdb->ecode = TDB_ERR_IO;
1133 if (rec.magic != TDB_RECOVERY_MAGIC) {
1134 /* there is no valid recovery data */
1138 if (tdb->read_only) {
1139 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_recover: attempt to recover read only database\n"));
1140 tdb->ecode = TDB_ERR_CORRUPT;
1144 recovery_eof = rec.key_len;
1146 data = (unsigned char *)malloc(rec.data_len);
1148 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_recover: failed to allocate recovery data\n"));
1149 tdb->ecode = TDB_ERR_OOM;
1153 /* read the full recovery data */
1154 if (tdb->methods->tdb_read(tdb, recovery_head + sizeof(rec), data,
1155 rec.data_len, 0) == -1) {
1156 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_recover: failed to read recovery data\n"));
1157 tdb->ecode = TDB_ERR_IO;
1161 /* recover the file data */
1163 while (p+8 < data + rec.data_len) {
1169 memcpy(&len, p+4, 4);
1171 if (tdb->methods->tdb_write(tdb, ofs, p+8, len) == -1) {
1173 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_recover: failed to recover %d bytes at offset %d\n", len, ofs));
1174 tdb->ecode = TDB_ERR_IO;
1182 if (transaction_sync(tdb, 0, tdb->map_size) == -1) {
1183 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_recover: failed to sync recovery\n"));
1184 tdb->ecode = TDB_ERR_IO;
1188 /* if the recovery area is after the recovered eof then remove it */
1189 if (recovery_eof <= recovery_head) {
1190 if (tdb_ofs_write(tdb, TDB_RECOVERY_HEAD, &zero) == -1) {
1191 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_recover: failed to remove recovery head\n"));
1192 tdb->ecode = TDB_ERR_IO;
1197 /* remove the recovery magic */
1198 if (tdb_ofs_write(tdb, recovery_head + offsetof(struct list_struct, magic),
1200 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_recover: failed to remove recovery magic\n"));
1201 tdb->ecode = TDB_ERR_IO;
1205 /* reduce the file size to the old size */
1207 if (ftruncate(tdb->fd, recovery_eof) != 0) {
1208 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_recover: failed to reduce to recovery size\n"));
1209 tdb->ecode = TDB_ERR_IO;
1212 tdb->map_size = recovery_eof;
1215 if (transaction_sync(tdb, 0, recovery_eof) == -1) {
1216 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_recover: failed to sync2 recovery\n"));
1217 tdb->ecode = TDB_ERR_IO;
1221 TDB_LOG((tdb, TDB_DEBUG_TRACE, "tdb_transaction_recover: recovered %d byte database\n",