lib/tdb/common/transaction.c

   1  /*
   2    Unix SMB/CIFS implementation.
   3
   4    trivial database library
   5
   6    Copyright (C) Andrew Tridgell              2005
   7
   8      ** NOTE! The following LGPL license applies to the tdb
   9      ** library. This does NOT imply that all of Samba is released
  10      ** under the LGPL
  11
  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.
  16
  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.
  21
  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/>.
  24 */
  25
  26 #include "tdb_private.h"
  27
  28 /*
  29   transaction design:
  30
  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
  35
  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
  40
  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
  45
  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
  50     to the real database.
  51
  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)
  55
  56   - if the caller gains a lock during the transaction but doesn't
  57     release it then fail the commit
  58
  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
  62
  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
  66
  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
  70
  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.
  77
  78   - check for a valid recovery record on open of the tdb, while the
  79     open 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
  82     intervention.
  83
  84   - if TDB_NOSYNC is passed to flags in tdb_open then transactions are
  85     still available, but no fsync/msync calls are made.  This means we
  86     are still proof against a process dying during transaction commit,
  87     but not against machine reboot.
  88
  89   - if TDB_ALLOW_NESTING is passed to flags in tdb open, or added using
  90     tdb_add_flags() transaction nesting is enabled.
  91     It resets the TDB_DISALLOW_NESTING flag, as both cannot be used together.
  92     The default is that transaction nesting is allowed.
  93     Note: this default may change in future versions of tdb.
  94
  95     Beware. when transactions are nested a transaction successfully
  96     completed with tdb_transaction_commit() can be silently unrolled later.
  97
  98   - if TDB_DISALLOW_NESTING is passed to flags in tdb open, or added using
  99     tdb_add_flags() transaction nesting is disabled.
 100     It resets the TDB_ALLOW_NESTING flag, as both cannot be used together.
 101     An attempt create a nested transaction will fail with TDB_ERR_NESTING.
 102     The default is that transaction nesting is allowed.
 103     Note: this default may change in future versions of tdb.
 104 */
 105
 106
 107 /*
 108   hold the context of any current transaction
 109 */
 110 struct tdb_transaction {
 111         /* we keep a mirrored copy of the tdb hash heads here so
 112            tdb_next_hash_chain() can operate efficiently */
 113         uint32_t *hash_heads;
 114
 115         /* the original io methods - used to do IOs to the real db */
 116         const struct tdb_methods *io_methods;
 117
 118         /* the list of transaction blocks. When a block is first
 119            written to, it gets created in this list */
 120         uint8_t **blocks;
 121         uint32_t num_blocks;
 122         uint32_t block_size;      /* bytes in each block */
 123         uint32_t last_block_size; /* number of valid bytes in the last block */
 124
 125         /* non-zero when an internal transaction error has
 126            occurred. All write operations will then fail until the
 127            transaction is ended */
 128         int transaction_error;
 129
 130         /* when inside a transaction we need to keep track of any
 131            nested tdb_transaction_start() calls, as these are allowed,
 132            but don't create a new transaction */
 133         int nesting;
 134
 135         /* set when a prepare has already occurred */
 136         bool prepared;
 137         tdb_off_t magic_offset;
 138
 139         /* old file size before transaction */
 140         tdb_len_t old_map_size;
 141
 142         /* did we expand in this transaction */
 143         bool expanded;
 144 };
 145
 146
 147 /*
 148   read while in a transaction. We need to check first if the data is in our list
 149   of transaction elements, then if not do a real read
 150 */
 151 static int transaction_read(struct tdb_context *tdb, tdb_off_t off, void *buf,
 152                             tdb_len_t len, int cv)
 153 {
 154         uint32_t blk;
 155
 156         /* break it down into block sized ops */
 157         while (len + (off % tdb->transaction->block_size) > tdb->transaction->block_size) {
 158                 tdb_len_t len2 = tdb->transaction->block_size - (off % tdb->transaction->block_size);
 159                 if (transaction_read(tdb, off, buf, len2, cv) != 0) {
 160                         return -1;
 161                 }
 162                 len -= len2;
 163                 off += len2;
 164                 buf = (void *)(len2 + (char *)buf);
 165         }
 166
 167         if (len == 0) {
 168                 return 0;
 169         }
 170
 171         blk = off / tdb->transaction->block_size;
 172
 173         /* see if we have it in the block list */
 174         if (tdb->transaction->num_blocks <= blk ||
 175             tdb->transaction->blocks[blk] == NULL) {
 176                 /* nope, do a real read */
 177                 if (tdb->transaction->io_methods->tdb_read(tdb, off, buf, len, cv) != 0) {
 178                         goto fail;
 179                 }
 180                 return 0;
 181         }
 182
 183         /* it is in the block list. Now check for the last block */
 184         if (blk == tdb->transaction->num_blocks-1) {
 185                 if (len > tdb->transaction->last_block_size) {
 186                         goto fail;
 187                 }
 188         }
 189
 190         /* now copy it out of this block */
 191         memcpy(buf, tdb->transaction->blocks[blk] + (off % tdb->transaction->block_size), len);
 192         if (cv) {
 193                 tdb_convert(buf, len);
 194         }
 195         return 0;
 196
 197 fail:
 198         TDB_LOG((tdb, TDB_DEBUG_FATAL, "transaction_read: failed at off=%u len=%u\n", off, len));
 199         tdb->ecode = TDB_ERR_IO;
 200         tdb->transaction->transaction_error = 1;
 201         return -1;
 202 }
 203
 204
 205 /*
 206   write while in a transaction
 207 */
 208 static int transaction_write(struct tdb_context *tdb, tdb_off_t off,
 209                              const void *buf, tdb_len_t len)
 210 {
 211         uint32_t blk;
 212
 213         /* Only a commit is allowed on a prepared transaction */
 214         if (tdb->transaction->prepared) {
 215                 tdb->ecode = TDB_ERR_EINVAL;
 216                 TDB_LOG((tdb, TDB_DEBUG_FATAL, "transaction_write: transaction already prepared, write not allowed\n"));
 217                 tdb->transaction->transaction_error = 1;
 218                 return -1;
 219         }
 220
 221         /* if the write is to a hash head, then update the transaction
 222            hash heads */
 223         if (len == sizeof(tdb_off_t) && off >= FREELIST_TOP &&
 224             off < FREELIST_TOP+TDB_HASHTABLE_SIZE(tdb)) {
 225                 uint32_t chain = (off-FREELIST_TOP) / sizeof(tdb_off_t);
 226                 memcpy(&tdb->transaction->hash_heads[chain], buf, len);
 227         }
 228
 229         /* break it up into block sized chunks */
 230         while (len + (off % tdb->transaction->block_size) > tdb->transaction->block_size) {
 231                 tdb_len_t len2 = tdb->transaction->block_size - (off % tdb->transaction->block_size);
 232                 if (transaction_write(tdb, off, buf, len2) != 0) {
 233                         return -1;
 234                 }
 235                 len -= len2;
 236                 off += len2;
 237                 if (buf != NULL) {
 238                         buf = (const void *)(len2 + (const char *)buf);
 239                 }
 240         }
 241
 242         if (len == 0) {
 243                 return 0;
 244         }
 245
 246         blk = off / tdb->transaction->block_size;
 247         off = off % tdb->transaction->block_size;
 248
 249         if (tdb->transaction->num_blocks <= blk) {
 250                 uint8_t **new_blocks;
 251                 /* expand the blocks array */
 252                 new_blocks = (uint8_t **)realloc(tdb->transaction->blocks,
 253                                                  (blk+1)*sizeof(uint8_t *));
 254                 if (new_blocks == NULL) {
 255                         tdb->ecode = TDB_ERR_OOM;
 256                         goto fail;
 257                 }
 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;
 263         }
 264
 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;
 271                         return -1;
 272                 }
 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);
 277                         }
 278                         if (tdb->transaction->io_methods->tdb_read(tdb, blk * tdb->transaction->block_size,
 279                                                                    tdb->transaction->blocks[blk],
 280                                                                    len2, 0) != 0) {
 281                                 SAFE_FREE(tdb->transaction->blocks[blk]);
 282                                 tdb->ecode = TDB_ERR_IO;
 283                                 goto fail;
 284                         }
 285                         if (blk == tdb->transaction->num_blocks-1) {
 286                                 tdb->transaction->last_block_size = len2;
 287                         }
 288                 }
 289         }
 290
 291         /* overwrite part of an existing block */
 292         if (buf == NULL) {
 293                 memset(tdb->transaction->blocks[blk] + off, 0, len);
 294         } else {
 295                 memcpy(tdb->transaction->blocks[blk] + off, buf, len);
 296         }
 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;
 300                 }
 301         }
 302
 303         return 0;
 304
 305 fail:
 306         TDB_LOG((tdb, TDB_DEBUG_FATAL, "transaction_write: failed at off=%u len=%u\n",
 307                  (blk*tdb->transaction->block_size) + off, len));
 308         tdb->transaction->transaction_error = 1;
 309         return -1;
 310 }
 311
 312
 313 /*
 314   write while in a transaction - this variant never expands the transaction blocks, it only
 315   updates existing blocks. This means it cannot change the recovery size
 316 */
 317 static int transaction_write_existing(struct tdb_context *tdb, tdb_off_t off,
 318                                       const void *buf, tdb_len_t len)
 319 {
 320         uint32_t blk;
 321
 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) {
 326                         return -1;
 327                 }
 328                 len -= len2;
 329                 off += len2;
 330                 if (buf != NULL) {
 331                         buf = (const void *)(len2 + (const char *)buf);
 332                 }
 333         }
 334
 335         if (len == 0) {
 336                 return 0;
 337         }
 338
 339         blk = off / tdb->transaction->block_size;
 340         off = off % tdb->transaction->block_size;
 341
 342         if (tdb->transaction->num_blocks <= blk ||
 343             tdb->transaction->blocks[blk] == NULL) {
 344                 return 0;
 345         }
 346
 347         if (blk == tdb->transaction->num_blocks-1 &&
 348             off + len > tdb->transaction->last_block_size) {
 349                 if (off >= tdb->transaction->last_block_size) {
 350                         return 0;
 351                 }
 352                 len = tdb->transaction->last_block_size - off;
 353         }
 354
 355         /* overwrite part of an existing block */
 356         memcpy(tdb->transaction->blocks[blk] + off, buf, len);
 357
 358         return 0;
 359 }
 360
 361
 362 /*
 363   accelerated hash chain head search, using the cached hash heads
 364 */
 365 static void transaction_next_hash_chain(struct tdb_context *tdb, uint32_t *chain)
 366 {
 367         uint32_t h = *chain;
 368         for (;h < tdb->hash_size;h++) {
 369                 /* the +1 takes account of the freelist */
 370                 if (0 != tdb->transaction->hash_heads[h+1]) {
 371                         break;
 372                 }
 373         }
 374         (*chain) = h;
 375 }
 376
 377 /*
 378   out of bounds check during a transaction
 379 */
 380 static int transaction_oob(struct tdb_context *tdb, tdb_off_t off,
 381                            tdb_len_t len, int probe)
 382 {
 383         if (off + len >= off && off + len <= tdb->map_size) {
 384                 return 0;
 385         }
 386         tdb->ecode = TDB_ERR_IO;
 387         return -1;
 388 }
 389
 390 /*
 391   transaction version of tdb_expand().
 392 */
 393 static int transaction_expand_file(struct tdb_context *tdb, tdb_off_t size,
 394                                    tdb_off_t addition)
 395 {
 396         /* add a write to the transaction elements, so subsequent
 397            reads see the zero data */
 398         if (transaction_write(tdb, size, NULL, addition) != 0) {
 399                 return -1;
 400         }
 401
 402         tdb->transaction->expanded = true;
 403
 404         return 0;
 405 }
 406
 407 static const struct tdb_methods transaction_methods = {
 408         transaction_read,
 409         transaction_write,
 410         transaction_next_hash_chain,
 411         transaction_oob,
 412         transaction_expand_file,
 413 };
 414
 415
 416 /*
 417   start a tdb transaction. No token is returned, as only a single
 418   transaction is allowed to be pending per tdb_context
 419 */
 420 static int _tdb_transaction_start(struct tdb_context *tdb,
 421                                   enum tdb_lock_flags lockflags)
 422 {
 423         /* some sanity checks */
 424         if (tdb->read_only || (tdb->flags & TDB_INTERNAL) || tdb->traverse_read) {
 425                 TDB_LOG((tdb, TDB_DEBUG_ERROR, "tdb_transaction_start: cannot start a transaction on a read-only or internal db\n"));
 426                 tdb->ecode = TDB_ERR_EINVAL;
 427                 return -1;
 428         }
 429
 430         /* cope with nested tdb_transaction_start() calls */
 431         if (tdb->transaction != NULL) {
 432                 if (!(tdb->flags & TDB_ALLOW_NESTING)) {
 433                         tdb->ecode = TDB_ERR_NESTING;
 434                         return -1;
 435                 }
 436                 tdb->transaction->nesting++;
 437                 TDB_LOG((tdb, TDB_DEBUG_TRACE, "tdb_transaction_start: nesting %d\n",
 438                          tdb->transaction->nesting));
 439                 return 0;
 440         }
 441
 442         if (tdb_have_extra_locks(tdb)) {
 443                 /* the caller must not have any locks when starting a
 444                    transaction as otherwise we'll be screwed by lack
 445                    of nested locks in posix */
 446                 TDB_LOG((tdb, TDB_DEBUG_ERROR, "tdb_transaction_start: cannot start a transaction with locks held\n"));
 447                 tdb->ecode = TDB_ERR_LOCK;
 448                 return -1;
 449         }
 450
 451         if (tdb->travlocks.next != NULL) {
 452                 /* you cannot use transactions inside a traverse (although you can use
 453                    traverse inside a transaction) as otherwise you can end up with
 454                    deadlock */
 455                 TDB_LOG((tdb, TDB_DEBUG_ERROR, "tdb_transaction_start: cannot start a transaction within a traverse\n"));
 456                 tdb->ecode = TDB_ERR_LOCK;
 457                 return -1;
 458         }
 459
 460         tdb->transaction = (struct tdb_transaction *)
 461                 calloc(sizeof(struct tdb_transaction), 1);
 462         if (tdb->transaction == NULL) {
 463                 tdb->ecode = TDB_ERR_OOM;
 464                 return -1;
 465         }
 466
 467         /* a page at a time seems like a reasonable compromise between compactness and efficiency */
 468         tdb->transaction->block_size = tdb->page_size;
 469
 470         /* get the transaction write lock. This is a blocking lock. As
 471            discussed with Volker, there are a number of ways we could
 472            make this async, which we will probably do in the future */
 473         if (tdb_transaction_lock(tdb, F_WRLCK, lockflags) == -1) {
 474                 SAFE_FREE(tdb->transaction->blocks);
 475                 SAFE_FREE(tdb->transaction);
 476                 if ((lockflags & TDB_LOCK_WAIT) == 0) {
 477                         tdb->ecode = TDB_ERR_NOLOCK;
 478                 }
 479                 return -1;
 480         }
 481
 482         /* get a read lock from the freelist to the end of file. This
 483            is upgraded to a write lock during the commit */
 484         if (tdb_allrecord_lock(tdb, F_RDLCK, TDB_LOCK_WAIT, true) == -1) {
 485                 TDB_LOG((tdb, TDB_DEBUG_ERROR, "tdb_transaction_start: failed to get hash locks\n"));
 486                 goto fail_allrecord_lock;
 487         }
 488
 489         /* setup a copy of the hash table heads so the hash scan in
 490            traverse can be fast */
 491         tdb->transaction->hash_heads = (uint32_t *)
 492                 calloc(tdb->hash_size+1, sizeof(uint32_t));
 493         if (tdb->transaction->hash_heads == NULL) {
 494                 tdb->ecode = TDB_ERR_OOM;
 495                 goto fail;
 496         }
 497         if (tdb->methods->tdb_read(tdb, FREELIST_TOP, tdb->transaction->hash_heads,
 498                                    TDB_HASHTABLE_SIZE(tdb), 0) != 0) {
 499                 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_start: failed to read hash heads\n"));
 500                 tdb->ecode = TDB_ERR_IO;
 501                 goto fail;
 502         }
 503
 504         /* make sure we know about any file expansions already done by
 505            anyone else */
 506         tdb->methods->tdb_oob(tdb, tdb->map_size, 1, 1);
 507         tdb->transaction->old_map_size = tdb->map_size;
 508
 509         /* finally hook the io methods, replacing them with
 510            transaction specific methods */
 511         tdb->transaction->io_methods = tdb->methods;
 512         tdb->methods = &transaction_methods;
 513
 514         /* Trace at the end, so we get sequence number correct. */
 515         tdb_trace(tdb, "tdb_transaction_start");
 516         return 0;
 517
 518 fail:
 519         tdb_allrecord_unlock(tdb, F_RDLCK, false);
 520 fail_allrecord_lock:
 521         tdb_transaction_unlock(tdb, F_WRLCK);
 522         SAFE_FREE(tdb->transaction->blocks);
 523         SAFE_FREE(tdb->transaction->hash_heads);
 524         SAFE_FREE(tdb->transaction);
 525         return -1;
 526 }
 527
 528 _PUBLIC_ int tdb_transaction_start(struct tdb_context *tdb)
 529 {
 530         return _tdb_transaction_start(tdb, TDB_LOCK_WAIT);
 531 }
 532
 533 _PUBLIC_ int tdb_transaction_start_nonblock(struct tdb_context *tdb)
 534 {
 535         return _tdb_transaction_start(tdb, TDB_LOCK_NOWAIT|TDB_LOCK_PROBE);
 536 }
 537
 538 /*
 539   sync to disk
 540 */
 541 static int transaction_sync(struct tdb_context *tdb, tdb_off_t offset, tdb_len_t length)
 542 {
 543         if (tdb->flags & TDB_NOSYNC) {
 544                 return 0;
 545         }
 546
 547 #ifdef HAVE_FDATASYNC
 548         if (fdatasync(tdb->fd) != 0) {
 549 #else
 550         if (fsync(tdb->fd) != 0) {
 551 #endif
 552                 tdb->ecode = TDB_ERR_IO;
 553                 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction: fsync failed\n"));
 554                 return -1;
 555         }
 556 #ifdef HAVE_MMAP
 557         if (tdb->map_ptr) {
 558                 tdb_off_t moffset = offset & ~(tdb->page_size-1);
 559                 if (msync(moffset + (char *)tdb->map_ptr,
 560                           length + (offset - moffset), MS_SYNC) != 0) {
 561                         tdb->ecode = TDB_ERR_IO;
 562                         TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction: msync failed - %s\n",
 563                                  strerror(errno)));
 564                         return -1;
 565                 }
 566         }
 567 #endif
 568         return 0;
 569 }
 570
 571
 572 static int _tdb_transaction_cancel(struct tdb_context *tdb)
 573 {
 574         int i, ret = 0;
 575
 576         if (tdb->transaction == NULL) {
 577                 TDB_LOG((tdb, TDB_DEBUG_ERROR, "tdb_transaction_cancel: no transaction\n"));
 578                 return -1;
 579         }
 580
 581         if (tdb->transaction->nesting != 0) {
 582                 tdb->transaction->transaction_error = 1;
 583                 tdb->transaction->nesting--;
 584                 return 0;
 585         }
 586
 587         tdb->map_size = tdb->transaction->old_map_size;
 588
 589         /* free all the transaction blocks */
 590         for (i=0;i<tdb->transaction->num_blocks;i++) {
 591                 if (tdb->transaction->blocks[i] != NULL) {
 592                         free(tdb->transaction->blocks[i]);
 593                 }
 594         }
 595         SAFE_FREE(tdb->transaction->blocks);
 596
 597         if (tdb->transaction->magic_offset) {
 598                 const struct tdb_methods *methods = tdb->transaction->io_methods;
 599                 const uint32_t invalid = TDB_RECOVERY_INVALID_MAGIC;
 600
 601                 /* remove the recovery marker */
 602                 if (methods->tdb_write(tdb, tdb->transaction->magic_offset, &invalid, 4) == -1 ||
 603                 transaction_sync(tdb, tdb->transaction->magic_offset, 4) == -1) {
 604                         TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_cancel: failed to remove recovery magic\n"));
 605                         ret = -1;
 606                 }
 607         }
 608
 609         /* This also removes the OPEN_LOCK, if we have it. */
 610         tdb_release_transaction_locks(tdb);
 611
 612         /* restore the normal io methods */
 613         tdb->methods = tdb->transaction->io_methods;
 614
 615         SAFE_FREE(tdb->transaction->hash_heads);
 616         SAFE_FREE(tdb->transaction);
 617
 618         return ret;
 619 }
 620
 621 /*
 622   cancel the current transaction
 623 */
 624 _PUBLIC_ int tdb_transaction_cancel(struct tdb_context *tdb)
 625 {
 626         tdb_trace(tdb, "tdb_transaction_cancel");
 627         return _tdb_transaction_cancel(tdb);
 628 }
 629
 630 /*
 631   work out how much space the linearised recovery data will consume
 632 */
 633 static bool tdb_recovery_size(struct tdb_context *tdb, tdb_len_t *result)
 634 {
 635         tdb_len_t recovery_size = 0;
 636         int i;
 637
 638         recovery_size = sizeof(uint32_t);
 639         for (i=0;i<tdb->transaction->num_blocks;i++) {
 640                 tdb_len_t block_size;
 641                 if (i * tdb->transaction->block_size >= tdb->transaction->old_map_size) {
 642                         break;
 643                 }
 644                 if (tdb->transaction->blocks[i] == NULL) {
 645                         continue;
 646                 }
 647                 if (!tdb_add_len_t(recovery_size, 2*sizeof(tdb_off_t),
 648                                    &recovery_size)) {
 649                         return false;
 650                 }
 651                 if (i == tdb->transaction->num_blocks-1) {
 652                         block_size = tdb->transaction->last_block_size;
 653                 } else {
 654                         block_size =  tdb->transaction->block_size;
 655                 }
 656                 if (!tdb_add_len_t(recovery_size, block_size,
 657                                    &recovery_size)) {
 658                         return false;
 659                 }
 660         }
 661
 662         *result = recovery_size;
 663         return true;
 664 }
 665
 666 int tdb_recovery_area(struct tdb_context *tdb,
 667                       const struct tdb_methods *methods,
 668                       tdb_off_t *recovery_offset,
 669                       struct tdb_record *rec)
 670 {
 671         if (tdb_ofs_read(tdb, TDB_RECOVERY_HEAD, recovery_offset) == -1) {
 672                 return -1;
 673         }
 674
 675         if (*recovery_offset == 0) {
 676                 rec->rec_len = 0;
 677                 return 0;
 678         }
 679
 680         if (methods->tdb_read(tdb, *recovery_offset, rec, sizeof(*rec),
 681                               DOCONV()) == -1) {
 682                 return -1;
 683         }
 684
 685         /* ignore invalid recovery regions: can happen in crash */
 686         if (rec->magic != TDB_RECOVERY_MAGIC &&
 687             rec->magic != TDB_RECOVERY_INVALID_MAGIC) {
 688                 *recovery_offset = 0;
 689                 rec->rec_len = 0;
 690         }
 691         return 0;
 692 }
 693
 694 /*
 695   allocate the recovery area, or use an existing recovery area if it is
 696   large enough
 697 */
 698 static int tdb_recovery_allocate(struct tdb_context *tdb,
 699                                  tdb_len_t *recovery_size,
 700                                  tdb_off_t *recovery_offset,
 701                                  tdb_len_t *recovery_max_size)
 702 {
 703         struct tdb_record rec;
 704         const struct tdb_methods *methods = tdb->transaction->io_methods;
 705         tdb_off_t recovery_head, new_end;
 706
 707         if (tdb_recovery_area(tdb, methods, &recovery_head, &rec) == -1) {
 708                 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_recovery_allocate: failed to read recovery head\n"));
 709                 return -1;
 710         }
 711
 712         if (!tdb_recovery_size(tdb, recovery_size)) {
 713                 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_recovery_allocate: "
 714                          "overflow recovery size\n"));
 715                 return -1;
 716         }
 717
 718         /* Existing recovery area? */
 719         if (recovery_head != 0 && *recovery_size <= rec.rec_len) {
 720                 /* it fits in the existing area */
 721                 *recovery_max_size = rec.rec_len;
 722                 *recovery_offset = recovery_head;
 723                 return 0;
 724         }
 725
 726         /* If recovery area in middle of file, we need a new one. */
 727         if (recovery_head == 0
 728             || recovery_head + sizeof(rec) + rec.rec_len != tdb->map_size) {
 729                 /* we need to free up the old recovery area, then allocate a
 730                    new one at the end of the file. Note that we cannot use
 731                    tdb_allocate() to allocate the new one as that might return
 732                    us an area that is being currently used (as of the start of
 733                    the transaction) */
 734                 if (recovery_head) {
 735                         if (tdb_free(tdb, recovery_head, &rec) == -1) {
 736                                 TDB_LOG((tdb, TDB_DEBUG_FATAL,
 737                                          "tdb_recovery_allocate: failed to"
 738                                          " free previous recovery area\n"));
 739                                 return -1;
 740                         }
 741
 742                         /* the tdb_free() call might have increased
 743                          * the recovery size */
 744                         if (!tdb_recovery_size(tdb, recovery_size)) {
 745                                 TDB_LOG((tdb, TDB_DEBUG_FATAL,
 746                                          "tdb_recovery_allocate: "
 747                                          "overflow recovery size\n"));
 748                                 return -1;
 749                         }
 750                 }
 751
 752                 /* New head will be at end of file. */
 753                 recovery_head = tdb->map_size;
 754         }
 755
 756         /* Now we know where it will be. */
 757         *recovery_offset = recovery_head;
 758
 759         /* Expand by more than we need, so we don't do it often. */
 760         *recovery_max_size = tdb_expand_adjust(tdb->map_size,
 761                                                *recovery_size,
 762                                                tdb->page_size)
 763                 - sizeof(rec);
 764
 765         if (!tdb_add_off_t(recovery_head, sizeof(rec), &new_end) ||
 766             !tdb_add_off_t(new_end, *recovery_max_size, &new_end)) {
 767                 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_recovery_allocate: "
 768                          "overflow recovery area\n"));
 769                 return -1;
 770         }
 771
 772         if (methods->tdb_expand_file(tdb, tdb->transaction->old_map_size,
 773                                      new_end - tdb->transaction->old_map_size)
 774             == -1) {
 775                 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_recovery_allocate: failed to create recovery area\n"));
 776                 return -1;
 777         }
 778
 779         /* remap the file (if using mmap) */
 780         methods->tdb_oob(tdb, tdb->map_size, 1, 1);
 781
 782         /* we have to reset the old map size so that we don't try to expand the file
 783            again in the transaction commit, which would destroy the recovery area */
 784         tdb->transaction->old_map_size = tdb->map_size;
 785
 786         /* write the recovery header offset and sync - we can sync without a race here
 787            as the magic ptr in the recovery record has not been set */
 788         CONVERT(recovery_head);
 789         if (methods->tdb_write(tdb, TDB_RECOVERY_HEAD,
 790                                &recovery_head, sizeof(tdb_off_t)) == -1) {
 791                 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_recovery_allocate: failed to write recovery head\n"));
 792                 return -1;
 793         }
 794         if (transaction_write_existing(tdb, TDB_RECOVERY_HEAD, &recovery_head, sizeof(tdb_off_t)) == -1) {
 795                 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_recovery_allocate: failed to write recovery head\n"));
 796                 return -1;
 797         }
 798
 799         return 0;
 800 }
 801
 802
 803 /*
 804   setup the recovery data that will be used on a crash during commit
 805 */
 806 static int transaction_setup_recovery(struct tdb_context *tdb,
 807                                       tdb_off_t *magic_offset)
 808 {
 809         tdb_len_t recovery_size;
 810         unsigned char *data, *p;
 811         const struct tdb_methods *methods = tdb->transaction->io_methods;
 812         struct tdb_record *rec;
 813         tdb_off_t recovery_offset, recovery_max_size;
 814         tdb_off_t old_map_size = tdb->transaction->old_map_size;
 815         uint32_t magic, tailer;
 816         int i;
 817
 818         /*
 819           check that the recovery area has enough space
 820         */
 821         if (tdb_recovery_allocate(tdb, &recovery_size,
 822                                   &recovery_offset, &recovery_max_size) == -1) {
 823                 return -1;
 824         }
 825
 826         data = (unsigned char *)malloc(recovery_size + sizeof(*rec));
 827         if (data == NULL) {
 828                 tdb->ecode = TDB_ERR_OOM;
 829                 return -1;
 830         }
 831
 832         rec = (struct tdb_record *)data;
 833         memset(rec, 0, sizeof(*rec));
 834
 835         rec->magic    = TDB_RECOVERY_INVALID_MAGIC;
 836         rec->data_len = recovery_size;
 837         rec->rec_len  = recovery_max_size;
 838         rec->key_len  = old_map_size;
 839         CONVERT(*rec);
 840
 841         /* build the recovery data into a single blob to allow us to do a single
 842            large write, which should be more efficient */
 843         p = data + sizeof(*rec);
 844         for (i=0;i<tdb->transaction->num_blocks;i++) {
 845                 tdb_off_t offset;
 846                 tdb_len_t length;
 847
 848                 if (tdb->transaction->blocks[i] == NULL) {
 849                         continue;
 850                 }
 851
 852                 offset = i * tdb->transaction->block_size;
 853                 length = tdb->transaction->block_size;
 854                 if (i == tdb->transaction->num_blocks-1) {
 855                         length = tdb->transaction->last_block_size;
 856                 }
 857
 858                 if (offset >= old_map_size) {
 859                         continue;
 860                 }
 861                 if (offset + length > tdb->transaction->old_map_size) {
 862                         TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_setup_recovery: transaction data over new region boundary\n"));
 863                         free(data);
 864                         tdb->ecode = TDB_ERR_CORRUPT;
 865                         return -1;
 866                 }
 867                 memcpy(p, &offset, 4);
 868                 memcpy(p+4, &length, 4);
 869                 if (DOCONV()) {
 870                         tdb_convert(p, 8);
 871                 }
 872                 /* the recovery area contains the old data, not the
 873                    new data, so we have to call the original tdb_read
 874                    method to get it */
 875                 if (methods->tdb_read(tdb, offset, p + 8, length, 0) != 0) {
 876                         free(data);
 877                         tdb->ecode = TDB_ERR_IO;
 878                         return -1;
 879                 }
 880                 p += 8 + length;
 881         }
 882
 883         /* and the tailer */
 884         tailer = sizeof(*rec) + recovery_max_size;
 885         memcpy(p, &tailer, 4);
 886         if (DOCONV()) {
 887                 tdb_convert(p, 4);
 888         }
 889
 890         /* write the recovery data to the recovery area */
 891         if (methods->tdb_write(tdb, recovery_offset, data, sizeof(*rec) + recovery_size) == -1) {
 892                 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_setup_recovery: failed to write recovery data\n"));
 893                 free(data);
 894                 tdb->ecode = TDB_ERR_IO;
 895                 return -1;
 896         }
 897         if (transaction_write_existing(tdb, recovery_offset, data, sizeof(*rec) + recovery_size) == -1) {
 898                 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_setup_recovery: failed to write secondary recovery data\n"));
 899                 free(data);
 900                 tdb->ecode = TDB_ERR_IO;
 901                 return -1;
 902         }
 903
 904         /* as we don't have ordered writes, we have to sync the recovery
 905            data before we update the magic to indicate that the recovery
 906            data is present */
 907         if (transaction_sync(tdb, recovery_offset, sizeof(*rec) + recovery_size) == -1) {
 908                 free(data);
 909                 return -1;
 910         }
 911
 912         free(data);
 913
 914         magic = TDB_RECOVERY_MAGIC;
 915         CONVERT(magic);
 916
 917         *magic_offset = recovery_offset + offsetof(struct tdb_record, magic);
 918
 919         if (methods->tdb_write(tdb, *magic_offset, &magic, sizeof(magic)) == -1) {
 920                 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_setup_recovery: failed to write recovery magic\n"));
 921                 tdb->ecode = TDB_ERR_IO;
 922                 return -1;
 923         }
 924         if (transaction_write_existing(tdb, *magic_offset, &magic, sizeof(magic)) == -1) {
 925                 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_setup_recovery: failed to write secondary recovery magic\n"));
 926                 tdb->ecode = TDB_ERR_IO;
 927                 return -1;
 928         }
 929
 930         /* ensure the recovery magic marker is on disk */
 931         if (transaction_sync(tdb, *magic_offset, sizeof(magic)) == -1) {
 932                 return -1;
 933         }
 934
 935         return 0;
 936 }
 937
 938 static int _tdb_transaction_prepare_commit(struct tdb_context *tdb)
 939 {
 940         const struct tdb_methods *methods;
 941
 942         if (tdb->transaction == NULL) {
 943                 TDB_LOG((tdb, TDB_DEBUG_ERROR, "tdb_transaction_prepare_commit: no transaction\n"));
 944                 return -1;
 945         }
 946
 947         if (tdb->transaction->prepared) {
 948                 tdb->ecode = TDB_ERR_EINVAL;
 949                 _tdb_transaction_cancel(tdb);
 950                 TDB_LOG((tdb, TDB_DEBUG_ERROR, "tdb_transaction_prepare_commit: transaction already prepared\n"));
 951                 return -1;
 952         }
 953
 954         if (tdb->transaction->transaction_error) {
 955                 tdb->ecode = TDB_ERR_IO;
 956                 _tdb_transaction_cancel(tdb);
 957                 TDB_LOG((tdb, TDB_DEBUG_ERROR, "tdb_transaction_prepare_commit: transaction error pending\n"));
 958                 return -1;
 959         }
 960
 961
 962         if (tdb->transaction->nesting != 0) {
 963                 return 0;
 964         }
 965
 966         /* check for a null transaction */
 967         if (tdb->transaction->blocks == NULL) {
 968                 return 0;
 969         }
 970
 971         methods = tdb->transaction->io_methods;
 972
 973         /* if there are any locks pending then the caller has not
 974            nested their locks properly, so fail the transaction */
 975         if (tdb_have_extra_locks(tdb)) {
 976                 tdb->ecode = TDB_ERR_LOCK;
 977                 TDB_LOG((tdb, TDB_DEBUG_ERROR, "tdb_transaction_prepare_commit: locks pending on commit\n"));
 978                 _tdb_transaction_cancel(tdb);
 979                 return -1;
 980         }
 981
 982         /* upgrade the main transaction lock region to a write lock */
 983         if (tdb_allrecord_upgrade(tdb) == -1) {
 984                 TDB_LOG((tdb, TDB_DEBUG_ERROR, "tdb_transaction_prepare_commit: failed to upgrade hash locks\n"));
 985                 _tdb_transaction_cancel(tdb);
 986                 return -1;
 987         }
 988
 989         /* get the open lock - this prevents new users attaching to the database
 990            during the commit */
 991         if (tdb_nest_lock(tdb, OPEN_LOCK, F_WRLCK, TDB_LOCK_WAIT) == -1) {
 992                 TDB_LOG((tdb, TDB_DEBUG_ERROR, "tdb_transaction_prepare_commit: failed to get open lock\n"));
 993                 _tdb_transaction_cancel(tdb);
 994                 return -1;
 995         }
 996
 997         /* write the recovery data to the end of the file */
 998         if (transaction_setup_recovery(tdb, &tdb->transaction->magic_offset) == -1) {
 999                 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_prepare_commit: failed to setup recovery data\n"));
1000                 _tdb_transaction_cancel(tdb);
1001                 return -1;
1002         }
1003
1004         tdb->transaction->prepared = true;
1005
1006         /* expand the file to the new size if needed */
1007         if (tdb->map_size != tdb->transaction->old_map_size) {
1008                 if (methods->tdb_expand_file(tdb, tdb->transaction->old_map_size,
1009                                              tdb->map_size -
1010                                              tdb->transaction->old_map_size) == -1) {
1011                         tdb->ecode = TDB_ERR_IO;
1012                         TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_prepare_commit: expansion failed\n"));
1013                         _tdb_transaction_cancel(tdb);
1014                         return -1;
1015                 }
1016                 tdb->map_size = tdb->transaction->old_map_size;
1017                 methods->tdb_oob(tdb, tdb->map_size, 1, 1);
1018         }
1019
1020         /* Keep the open lock until the actual commit */
1021
1022         return 0;
1023 }
1024
1025 /*
1026    prepare to commit the current transaction
1027 */
1028 _PUBLIC_ int tdb_transaction_prepare_commit(struct tdb_context *tdb)
1029 {
1030         tdb_trace(tdb, "tdb_transaction_prepare_commit");
1031         return _tdb_transaction_prepare_commit(tdb);
1032 }
1033
1034 /* A repack is worthwhile if the largest is less than half total free. */
1035 static bool repack_worthwhile(struct tdb_context *tdb)
1036 {
1037         tdb_off_t ptr;
1038         struct tdb_record rec;
1039         tdb_len_t total = 0, largest = 0;
1040
1041         if (tdb_ofs_read(tdb, FREELIST_TOP, &ptr) == -1) {
1042                 return false;
1043         }
1044
1045         while (ptr != 0 && tdb_rec_free_read(tdb, ptr, &rec) == 0) {
1046                 total += rec.rec_len;
1047                 if (rec.rec_len > largest) {
1048                         largest = rec.rec_len;
1049                 }
1050                 ptr = rec.next;
1051         }
1052
1053         return total > largest * 2;
1054 }
1055
1056 /*
1057   commit the current transaction
1058 */
1059 _PUBLIC_ int tdb_transaction_commit(struct tdb_context *tdb)
1060 {
1061         const struct tdb_methods *methods;
1062         int i;
1063         bool need_repack = false;
1064
1065         if (tdb->transaction == NULL) {
1066                 TDB_LOG((tdb, TDB_DEBUG_ERROR, "tdb_transaction_commit: no transaction\n"));
1067                 return -1;
1068         }
1069
1070         tdb_trace(tdb, "tdb_transaction_commit");
1071
1072         if (tdb->transaction->transaction_error) {
1073                 tdb->ecode = TDB_ERR_IO;
1074                 _tdb_transaction_cancel(tdb);
1075                 TDB_LOG((tdb, TDB_DEBUG_ERROR, "tdb_transaction_commit: transaction error pending\n"));
1076                 return -1;
1077         }
1078
1079
1080         if (tdb->transaction->nesting != 0) {
1081                 tdb->transaction->nesting--;
1082                 return 0;
1083         }
1084
1085         /* check for a null transaction */
1086         if (tdb->transaction->blocks == NULL) {
1087                 _tdb_transaction_cancel(tdb);
1088                 return 0;
1089         }
1090
1091         if (!tdb->transaction->prepared) {
1092                 int ret = _tdb_transaction_prepare_commit(tdb);
1093                 if (ret)
1094                         return ret;
1095         }
1096
1097         methods = tdb->transaction->io_methods;
1098
1099         /* perform all the writes */
1100         for (i=0;i<tdb->transaction->num_blocks;i++) {
1101                 tdb_off_t offset;
1102                 tdb_len_t length;
1103
1104                 if (tdb->transaction->blocks[i] == NULL) {
1105                         continue;
1106                 }
1107
1108                 offset = i * tdb->transaction->block_size;
1109                 length = tdb->transaction->block_size;
1110                 if (i == tdb->transaction->num_blocks-1) {
1111                         length = tdb->transaction->last_block_size;
1112                 }
1113
1114                 if (methods->tdb_write(tdb, offset, tdb->transaction->blocks[i], length) == -1) {
1115                         TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_commit: write failed during commit\n"));
1116
1117                         /* we've overwritten part of the data and
1118                            possibly expanded the file, so we need to
1119                            run the crash recovery code */
1120                         tdb->methods = methods;
1121                         tdb_transaction_recover(tdb);
1122
1123                         _tdb_transaction_cancel(tdb);
1124
1125                         TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_commit: write failed\n"));
1126                         return -1;
1127                 }
1128                 SAFE_FREE(tdb->transaction->blocks[i]);
1129         }
1130
1131         /* Do this before we drop lock or blocks. */
1132         if (tdb->transaction->expanded) {
1133                 need_repack = repack_worthwhile(tdb);
1134         }
1135
1136         SAFE_FREE(tdb->transaction->blocks);
1137         tdb->transaction->num_blocks = 0;
1138
1139         /* ensure the new data is on disk */
1140         if (transaction_sync(tdb, 0, tdb->map_size) == -1) {
1141                 return -1;
1142         }
1143
1144         /*
1145           TODO: maybe write to some dummy hdr field, or write to magic
1146           offset without mmap, before the last sync, instead of the
1147           utime() call
1148         */
1149
1150         /* on some systems (like Linux 2.6.x) changes via mmap/msync
1151            don't change the mtime of the file, this means the file may
1152            not be backed up (as tdb rounding to block sizes means that
1153            file size changes are quite rare too). The following forces
1154            mtime changes when a transaction completes */
1155 #ifdef HAVE_UTIME
1156         utime(tdb->name, NULL);
1157 #endif
1158
1159         /* use a transaction cancel to free memory and remove the
1160            transaction locks */
1161         _tdb_transaction_cancel(tdb);
1162
1163         if (need_repack) {
1164                 return tdb_repack(tdb);
1165         }
1166
1167         return 0;
1168 }
1169
1170
1171 /*
1172   recover from an aborted transaction. Must be called with exclusive
1173   database write access already established (including the open
1174   lock to prevent new processes attaching)
1175 */
1176 int tdb_transaction_recover(struct tdb_context *tdb)
1177 {
1178         tdb_off_t recovery_head, recovery_eof;
1179         unsigned char *data, *p;
1180         uint32_t zero = 0;
1181         struct tdb_record rec;
1182
1183         /* find the recovery area */
1184         if (tdb_ofs_read(tdb, TDB_RECOVERY_HEAD, &recovery_head) == -1) {
1185                 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_recover: failed to read recovery head\n"));
1186                 tdb->ecode = TDB_ERR_IO;
1187                 return -1;
1188         }
1189
1190         if (recovery_head == 0) {
1191                 /* we have never allocated a recovery record */
1192                 return 0;
1193         }
1194
1195         /* read the recovery record */
1196         if (tdb->methods->tdb_read(tdb, recovery_head, &rec,
1197                                    sizeof(rec), DOCONV()) == -1) {
1198                 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_recover: failed to read recovery record\n"));
1199                 tdb->ecode = TDB_ERR_IO;
1200                 return -1;
1201         }
1202
1203         if (rec.magic != TDB_RECOVERY_MAGIC) {
1204                 /* there is no valid recovery data */
1205                 return 0;
1206         }
1207
1208         if (tdb->read_only) {
1209                 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_recover: attempt to recover read only database\n"));
1210                 tdb->ecode = TDB_ERR_CORRUPT;
1211                 return -1;
1212         }
1213
1214         recovery_eof = rec.key_len;
1215
1216         data = (unsigned char *)malloc(rec.data_len);
1217         if (data == NULL) {
1218                 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_recover: failed to allocate recovery data\n"));
1219                 tdb->ecode = TDB_ERR_OOM;
1220                 return -1;
1221         }
1222
1223         /* read the full recovery data */
1224         if (tdb->methods->tdb_read(tdb, recovery_head + sizeof(rec), data,
1225                                    rec.data_len, 0) == -1) {
1226                 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_recover: failed to read recovery data\n"));
1227                 tdb->ecode = TDB_ERR_IO;
1228                 return -1;
1229         }
1230
1231         /* recover the file data */
1232         p = data;
1233         while (p+8 < data + rec.data_len) {
1234                 uint32_t ofs, len;
1235                 if (DOCONV()) {
1236                         tdb_convert(p, 8);
1237                 }
1238                 memcpy(&ofs, p, 4);
1239                 memcpy(&len, p+4, 4);
1240
1241                 if (tdb->methods->tdb_write(tdb, ofs, p+8, len) == -1) {
1242                         free(data);
1243                         TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_recover: failed to recover %u bytes at offset %u\n", len, ofs));
1244                         tdb->ecode = TDB_ERR_IO;
1245                         return -1;
1246                 }
1247                 p += 8 + len;
1248         }
1249
1250         free(data);
1251
1252         if (transaction_sync(tdb, 0, tdb->map_size) == -1) {
1253                 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_recover: failed to sync recovery\n"));
1254                 tdb->ecode = TDB_ERR_IO;
1255                 return -1;
1256         }
1257
1258         /* if the recovery area is after the recovered eof then remove it */
1259         if (recovery_eof <= recovery_head) {
1260                 if (tdb_ofs_write(tdb, TDB_RECOVERY_HEAD, &zero) == -1) {
1261                         TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_recover: failed to remove recovery head\n"));
1262                         tdb->ecode = TDB_ERR_IO;
1263                         return -1;
1264                 }
1265         }
1266
1267         /* remove the recovery magic */
1268         if (tdb_ofs_write(tdb, recovery_head + offsetof(struct tdb_record, magic),
1269                           &zero) == -1) {
1270                 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_recover: failed to remove recovery magic\n"));
1271                 tdb->ecode = TDB_ERR_IO;
1272                 return -1;
1273         }
1274
1275         if (transaction_sync(tdb, 0, recovery_eof) == -1) {
1276                 TDB_LOG((tdb, TDB_DEBUG_FATAL, "tdb_transaction_recover: failed to sync2 recovery\n"));
1277                 tdb->ecode = TDB_ERR_IO;
1278                 return -1;
1279         }
1280
1281         TDB_LOG((tdb, TDB_DEBUG_TRACE, "tdb_transaction_recover: recovered %u byte database\n",
1282                  recovery_eof));
1283
1284         /* all done */
1285         return 0;
1286 }
1287
1288 /* Any I/O failures we say "needs recovery". */
1289 bool tdb_needs_recovery(struct tdb_context *tdb)
1290 {
1291         tdb_off_t recovery_head;
1292         struct tdb_record rec;
1293
1294         /* find the recovery area */
1295         if (tdb_ofs_read(tdb, TDB_RECOVERY_HEAD, &recovery_head) == -1) {
1296                 return true;
1297         }
1298
1299         if (recovery_head == 0) {
1300                 /* we have never allocated a recovery record */
1301                 return false;
1302         }
1303
1304         /* read the recovery record */
1305         if (tdb->methods->tdb_read(tdb, recovery_head, &rec,
1306                                    sizeof(rec), DOCONV()) == -1) {
1307                 return true;
1308         }
1309
1310         return (rec.magic == TDB_RECOVERY_MAGIC);
1311 }