log0recv.c 87.1 KB
Newer Older
Vadim Tkachenko's avatar
Vadim Tkachenko committed
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
/*****************************************************************************

Copyright (c) 1997, 2009, Innobase Oy. All Rights Reserved.

This program is free software; you can redistribute it and/or modify it under
the terms of the GNU General Public License as published by the Free Software
Foundation; version 2 of the License.

This program 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 General Public License for more details.

You should have received a copy of the GNU General Public License along with
this program; if not, write to the Free Software Foundation, Inc., 59 Temple
Place, Suite 330, Boston, MA 02111-1307 USA

*****************************************************************************/

19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112
/******************************************************
Recovery

Created 9/20/1997 Heikki Tuuri
*******************************************************/

#include "log0recv.h"

#ifdef UNIV_NONINL
#include "log0recv.ic"
#endif

#include "mem0mem.h"
#include "buf0buf.h"
#include "buf0flu.h"
#include "buf0rea.h"
#include "srv0srv.h"
#include "srv0start.h"
#include "mtr0log.h"
#include "page0cur.h"
#include "page0zip.h"
#include "btr0cur.h"
#include "ibuf0ibuf.h"
#include "trx0undo.h"
#include "trx0rec.h"
#include "trx0roll.h"
#include "row0merge.h"

#ifdef UNIV_HOTBACKUP
/* This is set to FALSE if the backup was originally taken with the
ibbackup --include regexp option: then we do not want to create tables in
directories which were not included */
UNIV_INTERN ibool	recv_replay_file_ops	= TRUE;
#endif /* UNIV_HOTBACKUP */

/* Log records are stored in the hash table in chunks at most of this size;
this must be less than UNIV_PAGE_SIZE as it is stored in the buffer pool */
#define RECV_DATA_BLOCK_SIZE	(MEM_MAX_ALLOC_IN_BUF - sizeof(recv_data_t))

/* Read-ahead area in applying log records to file pages */
#define RECV_READ_AHEAD_AREA	32

UNIV_INTERN recv_sys_t*	recv_sys = NULL;
UNIV_INTERN ibool	recv_recovery_on = FALSE;
UNIV_INTERN ibool	recv_recovery_from_backup_on = FALSE;

UNIV_INTERN ibool	recv_needed_recovery = FALSE;

UNIV_INTERN ibool	recv_lsn_checks_on = FALSE;

/* There are two conditions under which we scan the logs, the first
is normal startup and the second is when we do a recovery from an
archive.
This flag is set if we are doing a scan from the last checkpoint during
startup. If we find log entries that were written after the last checkpoint
we know that the server was not cleanly shutdown. We must then initialize
the crash recovery environment before attempting to store these entries in
the log hash table. */
UNIV_INTERN ibool	recv_log_scan_is_startup_type = FALSE;

/* If the following is TRUE, the buffer pool file pages must be invalidated
after recovery and no ibuf operations are allowed; this becomes TRUE if
the log record hash table becomes too full, and log records must be merged
to file pages already before the recovery is finished: in this case no
ibuf operations are allowed, as they could modify the pages read in the
buffer pool before the pages have been recovered to the up-to-date state */

/* Recovery is running and no operations on the log files are allowed
yet: the variable name is misleading */

UNIV_INTERN ibool	recv_no_ibuf_operations = FALSE;

/* The following counter is used to decide when to print info on
log scan */
UNIV_INTERN ulint	recv_scan_print_counter	= 0;

UNIV_INTERN ibool	recv_is_from_backup	= FALSE;
#ifdef UNIV_HOTBACKUP
UNIV_INTERN ibool	recv_is_making_a_backup = FALSE;
#else
# define recv_is_making_a_backup FALSE
#endif /* UNIV_HOTBACKUP */

UNIV_INTERN ulint	recv_previous_parsed_rec_type	= 999999;
UNIV_INTERN ulint	recv_previous_parsed_rec_offset	= 0;
UNIV_INTERN ulint	recv_previous_parsed_rec_is_multi = 0;

UNIV_INTERN ulint	recv_max_parsed_page_no		= 0;

/* This many frames must be left free in the buffer pool when we scan
the log and store the scanned log records in the buffer pool: we will
use these free frames to read in pages when we start applying the
log records to the database. */

Vadim Tkachenko's avatar
Vadim Tkachenko committed
113
UNIV_INTERN ulint	recv_n_pool_free_frames		= 1024;
114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592

/* The maximum lsn we see for a page during the recovery process. If this
is bigger than the lsn we are able to scan up to, that is an indication that
the recovery failed and the database may be corrupt. */

UNIV_INTERN ib_uint64_t	recv_max_page_lsn;

/* prototypes */

/***********************************************************
Initialize crash recovery environment. Can be called iff
recv_needed_recovery == FALSE. */
static
void
recv_init_crash_recovery(void);
/*===========================*/

/************************************************************
Creates the recovery system. */
UNIV_INTERN
void
recv_sys_create(void)
/*=================*/
{
	if (recv_sys != NULL) {

		return;
	}

	recv_sys = mem_alloc(sizeof(recv_sys_t));

	mutex_create(&recv_sys->mutex, SYNC_RECV);

	recv_sys->heap = NULL;
	recv_sys->addr_hash = NULL;
}

/************************************************************
Inits the recovery system for a recovery operation. */
UNIV_INTERN
void
recv_sys_init(
/*==========*/
	ibool	recover_from_backup,	/* in: TRUE if this is called
					to recover from a hot backup */
	ulint	available_memory)	/* in: available memory in bytes */
{
	if (recv_sys->heap != NULL) {

		return;
	}

	mutex_enter(&(recv_sys->mutex));

	if (!recover_from_backup) {
		recv_sys->heap = mem_heap_create_in_buffer(256);
	} else {
		recv_sys->heap = mem_heap_create(256);
		recv_is_from_backup = TRUE;
	}

	recv_sys->buf = ut_malloc(RECV_PARSING_BUF_SIZE);
	recv_sys->len = 0;
	recv_sys->recovered_offset = 0;

	recv_sys->addr_hash = hash_create(available_memory / 64);
	recv_sys->n_addrs = 0;

	recv_sys->apply_log_recs = FALSE;
	recv_sys->apply_batch_on = FALSE;

	recv_sys->last_block_buf_start = mem_alloc(2 * OS_FILE_LOG_BLOCK_SIZE);

	recv_sys->last_block = ut_align(recv_sys->last_block_buf_start,
					OS_FILE_LOG_BLOCK_SIZE);
	recv_sys->found_corrupt_log = FALSE;

	recv_max_page_lsn = 0;

	mutex_exit(&(recv_sys->mutex));
}

/************************************************************
Empties the hash table when it has been fully processed. */
static
void
recv_sys_empty_hash(void)
/*=====================*/
{
	ut_ad(mutex_own(&(recv_sys->mutex)));

	if (recv_sys->n_addrs != 0) {
		fprintf(stderr,
			"InnoDB: Error: %lu pages with log records"
			" were left unprocessed!\n"
			"InnoDB: Maximum page number with"
			" log records on it %lu\n",
			(ulong) recv_sys->n_addrs,
			(ulong) recv_max_parsed_page_no);
		ut_error;
	}

	hash_table_free(recv_sys->addr_hash);
	mem_heap_empty(recv_sys->heap);

	recv_sys->addr_hash = hash_create(buf_pool_get_curr_size() / 256);
}

#ifndef UNIV_LOG_DEBUG
/************************************************************
Frees the recovery system. */
static
void
recv_sys_free(void)
/*===============*/
{
	mutex_enter(&(recv_sys->mutex));

	hash_table_free(recv_sys->addr_hash);
	mem_heap_free(recv_sys->heap);
	ut_free(recv_sys->buf);
	mem_free(recv_sys->last_block_buf_start);

	recv_sys->addr_hash = NULL;
	recv_sys->heap = NULL;

	mutex_exit(&(recv_sys->mutex));
}
#endif /* UNIV_LOG_DEBUG */

/************************************************************
Truncates possible corrupted or extra records from a log group. */
static
void
recv_truncate_group(
/*================*/
	log_group_t*	group,		/* in: log group */
	ib_uint64_t	recovered_lsn,	/* in: recovery succeeded up to this
					lsn */
	ib_uint64_t	limit_lsn,	/* in: this was the limit for
					recovery */
	ib_uint64_t	checkpoint_lsn,	/* in: recovery was started from this
					checkpoint */
	ib_uint64_t	archived_lsn)	/* in: the log has been archived up to
					this lsn */
{
	ib_uint64_t	start_lsn;
	ib_uint64_t	end_lsn;
	ib_uint64_t	finish_lsn1;
	ib_uint64_t	finish_lsn2;
	ib_uint64_t	finish_lsn;
	ulint		len;
	ulint		i;

	if (archived_lsn == IB_ULONGLONG_MAX) {
		/* Checkpoint was taken in the NOARCHIVELOG mode */
		archived_lsn = checkpoint_lsn;
	}

	finish_lsn1 = ut_uint64_align_down(archived_lsn,
					   OS_FILE_LOG_BLOCK_SIZE)
		+ log_group_get_capacity(group);

	finish_lsn2 = ut_uint64_align_up(recovered_lsn,
					 OS_FILE_LOG_BLOCK_SIZE)
		+ recv_sys->last_log_buf_size;

	if (limit_lsn != IB_ULONGLONG_MAX) {
		/* We do not know how far we should erase log records: erase
		as much as possible */

		finish_lsn = finish_lsn1;
	} else {
		/* It is enough to erase the length of the log buffer */
		finish_lsn = finish_lsn1 < finish_lsn2
			? finish_lsn1 : finish_lsn2;
	}

	ut_a(RECV_SCAN_SIZE <= log_sys->buf_size);

	/* Write the log buffer full of zeros */
	for (i = 0; i < RECV_SCAN_SIZE; i++) {

		*(log_sys->buf + i) = '\0';
	}

	start_lsn = ut_uint64_align_down(recovered_lsn,
					 OS_FILE_LOG_BLOCK_SIZE);

	if (start_lsn != recovered_lsn) {
		/* Copy the last incomplete log block to the log buffer and
		edit its data length: */

		ut_memcpy(log_sys->buf, recv_sys->last_block,
			  OS_FILE_LOG_BLOCK_SIZE);
		log_block_set_data_len(log_sys->buf,
				       (ulint) (recovered_lsn - start_lsn));
	}

	if (start_lsn >= finish_lsn) {

		return;
	}

	for (;;) {
		end_lsn = start_lsn + RECV_SCAN_SIZE;

		if (end_lsn > finish_lsn) {

			end_lsn = finish_lsn;
		}

		len = (ulint) (end_lsn - start_lsn);

		log_group_write_buf(group, log_sys->buf, len, start_lsn, 0);
		if (end_lsn >= finish_lsn) {

			return;
		}

		/* Write the log buffer full of zeros */
		for (i = 0; i < RECV_SCAN_SIZE; i++) {

			*(log_sys->buf + i) = '\0';
		}

		start_lsn = end_lsn;
	}
}

/************************************************************
Copies the log segment between group->recovered_lsn and recovered_lsn from the
most up-to-date log group to group, so that it contains the latest log data. */
static
void
recv_copy_group(
/*============*/
	log_group_t*	up_to_date_group,	/* in: the most up-to-date log
						group */
	log_group_t*	group,			/* in: copy to this log
						group */
	ib_uint64_t	recovered_lsn)		/* in: recovery succeeded up
						to this lsn */
{
	ib_uint64_t	start_lsn;
	ib_uint64_t	end_lsn;
	ulint		len;

	if (group->scanned_lsn >= recovered_lsn) {

		return;
	}

	ut_a(RECV_SCAN_SIZE <= log_sys->buf_size);

	start_lsn = ut_uint64_align_down(group->scanned_lsn,
					 OS_FILE_LOG_BLOCK_SIZE);
	for (;;) {
		end_lsn = start_lsn + RECV_SCAN_SIZE;

		if (end_lsn > recovered_lsn) {
			end_lsn = ut_uint64_align_up(recovered_lsn,
						     OS_FILE_LOG_BLOCK_SIZE);
		}

		log_group_read_log_seg(LOG_RECOVER, log_sys->buf,
				       up_to_date_group, start_lsn, end_lsn);

		len = (ulint) (end_lsn - start_lsn);

		log_group_write_buf(group, log_sys->buf, len, start_lsn, 0);

		if (end_lsn >= recovered_lsn) {

			return;
		}

		start_lsn = end_lsn;
	}
}

/************************************************************
Copies a log segment from the most up-to-date log group to the other log
groups, so that they all contain the latest log data. Also writes the info
about the latest checkpoint to the groups, and inits the fields in the group
memory structs to up-to-date values. */
static
void
recv_synchronize_groups(
/*====================*/
	log_group_t*	up_to_date_group)	/* in: the most up-to-date
						log group */
{
	log_group_t*	group;
	ib_uint64_t	start_lsn;
	ib_uint64_t	end_lsn;
	ib_uint64_t	recovered_lsn;
	ib_uint64_t	limit_lsn;

	recovered_lsn = recv_sys->recovered_lsn;
	limit_lsn = recv_sys->limit_lsn;

	/* Read the last recovered log block to the recovery system buffer:
	the block is always incomplete */

	start_lsn = ut_uint64_align_down(recovered_lsn,
					 OS_FILE_LOG_BLOCK_SIZE);
	end_lsn = ut_uint64_align_up(recovered_lsn, OS_FILE_LOG_BLOCK_SIZE);

	ut_a(start_lsn != end_lsn);

	log_group_read_log_seg(LOG_RECOVER, recv_sys->last_block,
			       up_to_date_group, start_lsn, end_lsn);

	group = UT_LIST_GET_FIRST(log_sys->log_groups);

	while (group) {
		if (group != up_to_date_group) {

			/* Copy log data if needed */

			recv_copy_group(group, up_to_date_group,
					recovered_lsn);
		}

		/* Update the fields in the group struct to correspond to
		recovered_lsn */

		log_group_set_fields(group, recovered_lsn);

		group = UT_LIST_GET_NEXT(log_groups, group);
	}

	/* Copy the checkpoint info to the groups; remember that we have
	incremented checkpoint_no by one, and the info will not be written
	over the max checkpoint info, thus making the preservation of max
	checkpoint info on disk certain */

	log_groups_write_checkpoint_info();

	mutex_exit(&(log_sys->mutex));

	/* Wait for the checkpoint write to complete */
	rw_lock_s_lock(&(log_sys->checkpoint_lock));
	rw_lock_s_unlock(&(log_sys->checkpoint_lock));

	mutex_enter(&(log_sys->mutex));
}

/***************************************************************************
Checks the consistency of the checkpoint info */
static
ibool
recv_check_cp_is_consistent(
/*========================*/
			/* out: TRUE if ok */
	byte*	buf)	/* in: buffer containing checkpoint info */
{
	ulint	fold;

	fold = ut_fold_binary(buf, LOG_CHECKPOINT_CHECKSUM_1);

	if ((fold & 0xFFFFFFFFUL) != mach_read_from_4(
		    buf + LOG_CHECKPOINT_CHECKSUM_1)) {
		return(FALSE);
	}

	fold = ut_fold_binary(buf + LOG_CHECKPOINT_LSN,
			      LOG_CHECKPOINT_CHECKSUM_2 - LOG_CHECKPOINT_LSN);

	if ((fold & 0xFFFFFFFFUL) != mach_read_from_4(
		    buf + LOG_CHECKPOINT_CHECKSUM_2)) {
		return(FALSE);
	}

	return(TRUE);
}

/************************************************************
Looks for the maximum consistent checkpoint from the log groups. */
static
ulint
recv_find_max_checkpoint(
/*=====================*/
					/* out: error code or DB_SUCCESS */
	log_group_t**	max_group,	/* out: max group */
	ulint*		max_field)	/* out: LOG_CHECKPOINT_1 or
					LOG_CHECKPOINT_2 */
{
	log_group_t*	group;
	ib_uint64_t	max_no;
	ib_uint64_t	checkpoint_no;
	ulint		field;
	byte*		buf;

	group = UT_LIST_GET_FIRST(log_sys->log_groups);

	max_no = 0;
	*max_group = NULL;
	*max_field = 0;

	buf = log_sys->checkpoint_buf;

	while (group) {
		group->state = LOG_GROUP_CORRUPTED;

		for (field = LOG_CHECKPOINT_1; field <= LOG_CHECKPOINT_2;
		     field += LOG_CHECKPOINT_2 - LOG_CHECKPOINT_1) {

			log_group_read_checkpoint_info(group, field);

			if (!recv_check_cp_is_consistent(buf)) {
#ifdef UNIV_DEBUG
				if (log_debug_writes) {
					fprintf(stderr,
						"InnoDB: Checkpoint in group"
						" %lu at %lu invalid, %lu\n",
						(ulong) group->id,
						(ulong) field,
						(ulong) mach_read_from_4(
							buf
							+ LOG_CHECKPOINT_CHECKSUM_1));

				}
#endif /* UNIV_DEBUG */
				goto not_consistent;
			}

			group->state = LOG_GROUP_OK;

			group->lsn = mach_read_ull(
				buf + LOG_CHECKPOINT_LSN);
			group->lsn_offset = mach_read_from_4(
				buf + LOG_CHECKPOINT_OFFSET);
			checkpoint_no = mach_read_ull(
				buf + LOG_CHECKPOINT_NO);

#ifdef UNIV_DEBUG
			if (log_debug_writes) {
				fprintf(stderr,
					"InnoDB: Checkpoint number %lu"
					" found in group %lu\n",
					(ulong) checkpoint_no,
					(ulong) group->id);
			}
#endif /* UNIV_DEBUG */

			if (checkpoint_no >= max_no) {
				*max_group = group;
				*max_field = field;
				max_no = checkpoint_no;
			}

not_consistent:
			;
		}

		group = UT_LIST_GET_NEXT(log_groups, group);
	}

	if (*max_group == NULL) {

		fprintf(stderr,
			"InnoDB: No valid checkpoint found.\n"
			"InnoDB: If this error appears when you are"
			" creating an InnoDB database,\n"
			"InnoDB: the problem may be that during"
			" an earlier attempt you managed\n"
			"InnoDB: to create the InnoDB data files,"
			" but log file creation failed.\n"
			"InnoDB: If that is the case, please refer to\n"
			"InnoDB: http://dev.mysql.com/doc/refman/5.1/en/"
			"error-creating-innodb.html\n");
		return(DB_ERROR);
	}

	return(DB_SUCCESS);
}

Vadim Tkachenko's avatar
Vadim Tkachenko committed
593
#ifdef UNIV_HOTBACKUP
594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663
/***********************************************************************
Reads the checkpoint info needed in hot backup. */
UNIV_INTERN
ibool
recv_read_cp_info_for_backup(
/*=========================*/
				/* out: TRUE if success */
	byte*		hdr,	/* in: buffer containing the log group
				header */
	ib_uint64_t*	lsn,	/* out: checkpoint lsn */
	ulint*		offset,	/* out: checkpoint offset in the log group */
	ulint*		fsp_limit,/* out: fsp limit of space 0,
				1000000000 if the database is running
				with < version 3.23.50 of InnoDB */
	ib_uint64_t*	cp_no,	/* out: checkpoint number */
	ib_uint64_t*	first_header_lsn)
				/* out: lsn of of the start of the
				first log file */
{
	ulint		max_cp		= 0;
	ib_uint64_t	max_cp_no	= 0;
	byte*		cp_buf;

	cp_buf = hdr + LOG_CHECKPOINT_1;

	if (recv_check_cp_is_consistent(cp_buf)) {
		max_cp_no = mach_read_ull(cp_buf + LOG_CHECKPOINT_NO);
		max_cp = LOG_CHECKPOINT_1;
	}

	cp_buf = hdr + LOG_CHECKPOINT_2;

	if (recv_check_cp_is_consistent(cp_buf)) {
		if (mach_read_ull(cp_buf + LOG_CHECKPOINT_NO) > max_cp_no) {
			max_cp = LOG_CHECKPOINT_2;
		}
	}

	if (max_cp == 0) {
		return(FALSE);
	}

	cp_buf = hdr + max_cp;

	*lsn = mach_read_ull(cp_buf + LOG_CHECKPOINT_LSN);
	*offset = mach_read_from_4(cp_buf + LOG_CHECKPOINT_OFFSET);

	/* If the user is running a pre-3.23.50 version of InnoDB, its
	checkpoint data does not contain the fsp limit info */
	if (mach_read_from_4(cp_buf + LOG_CHECKPOINT_FSP_MAGIC_N)
	    == LOG_CHECKPOINT_FSP_MAGIC_N_VAL) {

		*fsp_limit = mach_read_from_4(
			cp_buf + LOG_CHECKPOINT_FSP_FREE_LIMIT);

		if (*fsp_limit == 0) {
			*fsp_limit = 1000000000;
		}
	} else {
		*fsp_limit = 1000000000;
	}

	/*	fprintf(stderr, "fsp limit %lu MB\n", *fsp_limit); */

	*cp_no = mach_read_ull(cp_buf + LOG_CHECKPOINT_NO);

	*first_header_lsn = mach_read_ull(hdr + LOG_FILE_START_LSN);

	return(TRUE);
}
Vadim Tkachenko's avatar
Vadim Tkachenko committed
664
#endif /* UNIV_HOTBACKUP */
665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701

/**********************************************************
Checks the 4-byte checksum to the trailer checksum field of a log block.
We also accept a log block in the old format < InnoDB-3.23.52 where the
checksum field contains the log block number. */
static
ibool
log_block_checksum_is_ok_or_old_format(
/*===================================*/
			/* out: TRUE if ok, or if the log block may be in the
			format of InnoDB version < 3.23.52 */
	byte*	block)	/* in: pointer to a log block */
{
#ifdef UNIV_LOG_DEBUG
	return(TRUE);
#endif /* UNIV_LOG_DEBUG */
	if (log_block_calc_checksum(block) == log_block_get_checksum(block)) {

		return(TRUE);
	}

	if (log_block_get_hdr_no(block) == log_block_get_checksum(block)) {

		/* We assume the log block is in the format of
		InnoDB version < 3.23.52 and the block is ok */
#if 0
		fprintf(stderr,
			"InnoDB: Scanned old format < InnoDB-3.23.52"
			" log block number %lu\n",
			log_block_get_hdr_no(block));
#endif
		return(TRUE);
	}

	return(FALSE);
}

Vadim Tkachenko's avatar
Vadim Tkachenko committed
702
#ifdef UNIV_HOTBACKUP
703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791
/***********************************************************************
Scans the log segment and n_bytes_scanned is set to the length of valid
log scanned. */
UNIV_INTERN
void
recv_scan_log_seg_for_backup(
/*=========================*/
	byte*		buf,		/* in: buffer containing log data */
	ulint		buf_len,	/* in: data length in that buffer */
	ib_uint64_t*	scanned_lsn,	/* in/out: lsn of buffer start,
					we return scanned lsn */
	ulint*		scanned_checkpoint_no,
					/* in/out: 4 lowest bytes of the
					highest scanned checkpoint number so
					far */
	ulint*		n_bytes_scanned)/* out: how much we were able to
					scan, smaller than buf_len if log
					data ended here */
{
	ulint	data_len;
	byte*	log_block;
	ulint	no;

	*n_bytes_scanned = 0;

	for (log_block = buf; log_block < buf + buf_len;
	     log_block += OS_FILE_LOG_BLOCK_SIZE) {

		no = log_block_get_hdr_no(log_block);

#if 0
		fprintf(stderr, "Log block header no %lu\n", no);
#endif

		if (no != log_block_convert_lsn_to_no(*scanned_lsn)
		    || !log_block_checksum_is_ok_or_old_format(log_block)) {
#if 0
			fprintf(stderr,
				"Log block n:o %lu, scanned lsn n:o %lu\n",
				no, log_block_convert_lsn_to_no(*scanned_lsn));
#endif
			/* Garbage or an incompletely written log block */

			log_block += OS_FILE_LOG_BLOCK_SIZE;
#if 0
			fprintf(stderr,
				"Next log block n:o %lu\n",
				log_block_get_hdr_no(log_block));
#endif
			break;
		}

		if (*scanned_checkpoint_no > 0
		    && log_block_get_checkpoint_no(log_block)
		    < *scanned_checkpoint_no
		    && *scanned_checkpoint_no
		    - log_block_get_checkpoint_no(log_block)
		    > 0x80000000UL) {

			/* Garbage from a log buffer flush which was made
			before the most recent database recovery */
#if 0
			fprintf(stderr,
				"Scanned cp n:o %lu, block cp n:o %lu\n",
				*scanned_checkpoint_no,
				log_block_get_checkpoint_no(log_block));
#endif
			break;
		}

		data_len = log_block_get_data_len(log_block);

		*scanned_checkpoint_no
			= log_block_get_checkpoint_no(log_block);
		*scanned_lsn += data_len;

		*n_bytes_scanned += data_len;

		if (data_len < OS_FILE_LOG_BLOCK_SIZE) {
			/* Log data ends here */

#if 0
			fprintf(stderr, "Log block data len %lu\n",
				data_len);
#endif
			break;
		}
	}
}
Vadim Tkachenko's avatar
Vadim Tkachenko committed
792
#endif /* UNIV_HOTBACKUP */
793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227

/***********************************************************************
Tries to parse a single log record body and also applies it to a page if
specified. File ops are parsed, but not applied in this function. */
static
byte*
recv_parse_or_apply_log_rec_body(
/*=============================*/
				/* out: log record end, NULL if not a
				complete record */
	byte		type,	/* in: type */
	byte*		ptr,	/* in: pointer to a buffer */
	byte*		end_ptr,/* in: pointer to the buffer end */
	buf_block_t*	block,	/* in/out: buffer block or NULL; if
				not NULL, then the log record is
				applied to the page, and the log
				record should be complete then */
	mtr_t*		mtr)	/* in: mtr or NULL; should be non-NULL
				if and only if block is non-NULL */
{
	dict_index_t*	index	= NULL;
	page_t*		page;
	page_zip_des_t*	page_zip;

	ut_ad(!block == !mtr);

	if (block) {
		page = block->frame;
		page_zip = buf_block_get_page_zip(block);
	} else {
		page = NULL;
		page_zip = NULL;
	}

	switch (type) {
	case MLOG_1BYTE: case MLOG_2BYTES: case MLOG_4BYTES: case MLOG_8BYTES:
		ptr = mlog_parse_nbytes(type, ptr, end_ptr, page, page_zip);
		break;
	case MLOG_REC_INSERT: case MLOG_COMP_REC_INSERT:
		if (NULL != (ptr = mlog_parse_index(
				     ptr, end_ptr,
				     type == MLOG_COMP_REC_INSERT,
				     &index))) {
			ut_a(!page
			     || (ibool)!!page_is_comp(page)
			     == dict_table_is_comp(index->table));
			ptr = page_cur_parse_insert_rec(FALSE, ptr, end_ptr,
							block, index, mtr);
		}
		break;
	case MLOG_REC_CLUST_DELETE_MARK: case MLOG_COMP_REC_CLUST_DELETE_MARK:
		if (NULL != (ptr = mlog_parse_index(
				     ptr, end_ptr,
				     type == MLOG_COMP_REC_CLUST_DELETE_MARK,
				     &index))) {
			ut_a(!page
			     || (ibool)!!page_is_comp(page)
			     == dict_table_is_comp(index->table));
			ptr = btr_cur_parse_del_mark_set_clust_rec(
				ptr, end_ptr, page, page_zip, index);
		}
		break;
	case MLOG_COMP_REC_SEC_DELETE_MARK:
		/* This log record type is obsolete, but we process it for
		backward compatibility with MySQL 5.0.3 and 5.0.4. */
		ut_a(!page || page_is_comp(page));
		ut_a(!page_zip);
		ptr = mlog_parse_index(ptr, end_ptr, TRUE, &index);
		if (!ptr) {
			break;
		}
		/* Fall through */
	case MLOG_REC_SEC_DELETE_MARK:
		ptr = btr_cur_parse_del_mark_set_sec_rec(ptr, end_ptr,
							 page, page_zip);
		break;
	case MLOG_REC_UPDATE_IN_PLACE: case MLOG_COMP_REC_UPDATE_IN_PLACE:
		if (NULL != (ptr = mlog_parse_index(
				     ptr, end_ptr,
				     type == MLOG_COMP_REC_UPDATE_IN_PLACE,
				     &index))) {
			ut_a(!page
			     || (ibool)!!page_is_comp(page)
			     == dict_table_is_comp(index->table));
			ptr = btr_cur_parse_update_in_place(ptr, end_ptr, page,
							    page_zip, index);
		}
		break;
	case MLOG_LIST_END_DELETE: case MLOG_COMP_LIST_END_DELETE:
	case MLOG_LIST_START_DELETE: case MLOG_COMP_LIST_START_DELETE:
		if (NULL != (ptr = mlog_parse_index(
				     ptr, end_ptr,
				     type == MLOG_COMP_LIST_END_DELETE
				     || type == MLOG_COMP_LIST_START_DELETE,
				     &index))) {
			ut_a(!page
			     || (ibool)!!page_is_comp(page)
			     == dict_table_is_comp(index->table));
			ptr = page_parse_delete_rec_list(type, ptr, end_ptr,
							 block, index, mtr);
		}
		break;
	case MLOG_LIST_END_COPY_CREATED: case MLOG_COMP_LIST_END_COPY_CREATED:
		if (NULL != (ptr = mlog_parse_index(
				     ptr, end_ptr,
				     type == MLOG_COMP_LIST_END_COPY_CREATED,
				     &index))) {
			ut_a(!page
			     || (ibool)!!page_is_comp(page)
			     == dict_table_is_comp(index->table));
			ptr = page_parse_copy_rec_list_to_created_page(
				ptr, end_ptr, block, index, mtr);
		}
		break;
	case MLOG_PAGE_REORGANIZE: case MLOG_COMP_PAGE_REORGANIZE:
		if (NULL != (ptr = mlog_parse_index(
				     ptr, end_ptr,
				     type == MLOG_COMP_PAGE_REORGANIZE,
				     &index))) {
			ut_a(!page
			     || (ibool)!!page_is_comp(page)
			     == dict_table_is_comp(index->table));
			ptr = btr_parse_page_reorganize(ptr, end_ptr, index,
							block, mtr);
		}
		break;
	case MLOG_PAGE_CREATE: case MLOG_COMP_PAGE_CREATE:
		ut_a(!page_zip);
		ptr = page_parse_create(ptr, end_ptr,
					type == MLOG_COMP_PAGE_CREATE,
					block, mtr);
		break;
	case MLOG_UNDO_INSERT:
		ptr = trx_undo_parse_add_undo_rec(ptr, end_ptr, page);
		break;
	case MLOG_UNDO_ERASE_END:
		ptr = trx_undo_parse_erase_page_end(ptr, end_ptr, page, mtr);
		break;
	case MLOG_UNDO_INIT:
		ptr = trx_undo_parse_page_init(ptr, end_ptr, page, mtr);
		break;
	case MLOG_UNDO_HDR_DISCARD:
		ptr = trx_undo_parse_discard_latest(ptr, end_ptr, page, mtr);
		break;
	case MLOG_UNDO_HDR_CREATE:
	case MLOG_UNDO_HDR_REUSE:
		ptr = trx_undo_parse_page_header(type, ptr, end_ptr,
						 page, mtr);
		break;
	case MLOG_REC_MIN_MARK: case MLOG_COMP_REC_MIN_MARK:
		/* On a compressed page, MLOG_COMP_REC_MIN_MARK
		will be followed by MLOG_COMP_REC_DELETE
		or MLOG_ZIP_WRITE_HEADER(FIL_PAGE_PREV, FIL_NULL)
		in the same mini-transaction. */
		ut_a(type == MLOG_COMP_REC_MIN_MARK || !page_zip);
		ptr = btr_parse_set_min_rec_mark(
			ptr, end_ptr, type == MLOG_COMP_REC_MIN_MARK,
			page, mtr);
		break;
	case MLOG_REC_DELETE: case MLOG_COMP_REC_DELETE:
		if (NULL != (ptr = mlog_parse_index(
				     ptr, end_ptr,
				     type == MLOG_COMP_REC_DELETE,
				     &index))) {
			ut_a(!page
			     || (ibool)!!page_is_comp(page)
			     == dict_table_is_comp(index->table));
			ptr = page_cur_parse_delete_rec(ptr, end_ptr,
							block, index, mtr);
		}
		break;
	case MLOG_IBUF_BITMAP_INIT:
		ptr = ibuf_parse_bitmap_init(ptr, end_ptr, block, mtr);
		break;
	case MLOG_INIT_FILE_PAGE:
		ptr = fsp_parse_init_file_page(ptr, end_ptr, block);
		break;
	case MLOG_WRITE_STRING:
		ptr = mlog_parse_string(ptr, end_ptr, page, page_zip);
		break;
	case MLOG_FILE_CREATE:
	case MLOG_FILE_RENAME:
	case MLOG_FILE_DELETE:
	case MLOG_FILE_CREATE2:
		ptr = fil_op_log_parse_or_replay(ptr, end_ptr, type, 0);
		break;
	case MLOG_ZIP_WRITE_NODE_PTR:
		ptr = page_zip_parse_write_node_ptr(ptr, end_ptr,
						    page, page_zip);
		break;
	case MLOG_ZIP_WRITE_BLOB_PTR:
		ptr = page_zip_parse_write_blob_ptr(ptr, end_ptr,
						    page, page_zip);
		break;
	case MLOG_ZIP_WRITE_HEADER:
		ptr = page_zip_parse_write_header(ptr, end_ptr,
						  page, page_zip);
		break;
	case MLOG_ZIP_PAGE_COMPRESS:
		ptr = page_zip_parse_compress(ptr, end_ptr,
					      page, page_zip);
		break;
	default:
		ptr = NULL;
		recv_sys->found_corrupt_log = TRUE;
	}

	if (index) {
		dict_table_t*	table = index->table;

		dict_mem_index_free(index);
		dict_mem_table_free(table);
	}

	return(ptr);
}

/*************************************************************************
Calculates the fold value of a page file address: used in inserting or
searching for a log record in the hash table. */
UNIV_INLINE
ulint
recv_fold(
/*======*/
			/* out: folded value */
	ulint	space,	/* in: space */
	ulint	page_no)/* in: page number */
{
	return(ut_fold_ulint_pair(space, page_no));
}

/*************************************************************************
Calculates the hash value of a page file address: used in inserting or
searching for a log record in the hash table. */
UNIV_INLINE
ulint
recv_hash(
/*======*/
			/* out: folded value */
	ulint	space,	/* in: space */
	ulint	page_no)/* in: page number */
{
	return(hash_calc_hash(recv_fold(space, page_no), recv_sys->addr_hash));
}

/*************************************************************************
Gets the hashed file address struct for a page. */
static
recv_addr_t*
recv_get_fil_addr_struct(
/*=====================*/
			/* out: file address struct, NULL if not found from
			the hash table */
	ulint	space,	/* in: space id */
	ulint	page_no)/* in: page number */
{
	recv_addr_t*	recv_addr;

	recv_addr = HASH_GET_FIRST(recv_sys->addr_hash,
				   recv_hash(space, page_no));
	while (recv_addr) {
		if ((recv_addr->space == space)
		    && (recv_addr->page_no == page_no)) {

			break;
		}

		recv_addr = HASH_GET_NEXT(addr_hash, recv_addr);
	}

	return(recv_addr);
}

/***********************************************************************
Adds a new log record to the hash table of log records. */
static
void
recv_add_to_hash_table(
/*===================*/
	byte		type,		/* in: log record type */
	ulint		space,		/* in: space id */
	ulint		page_no,	/* in: page number */
	byte*		body,		/* in: log record body */
	byte*		rec_end,	/* in: log record end */
	ib_uint64_t	start_lsn,	/* in: start lsn of the mtr */
	ib_uint64_t	end_lsn)	/* in: end lsn of the mtr */
{
	recv_t*		recv;
	ulint		len;
	recv_data_t*	recv_data;
	recv_data_t**	prev_field;
	recv_addr_t*	recv_addr;

	if (fil_tablespace_deleted_or_being_deleted_in_mem(space, -1)) {
		/* The tablespace does not exist any more: do not store the
		log record */

		return;
	}

	len = rec_end - body;

	recv = mem_heap_alloc(recv_sys->heap, sizeof(recv_t));
	recv->type = type;
	recv->len = rec_end - body;
	recv->start_lsn = start_lsn;
	recv->end_lsn = end_lsn;

	recv_addr = recv_get_fil_addr_struct(space, page_no);

	if (recv_addr == NULL) {
		recv_addr = mem_heap_alloc(recv_sys->heap,
					   sizeof(recv_addr_t));
		recv_addr->space = space;
		recv_addr->page_no = page_no;
		recv_addr->state = RECV_NOT_PROCESSED;

		UT_LIST_INIT(recv_addr->rec_list);

		HASH_INSERT(recv_addr_t, addr_hash, recv_sys->addr_hash,
			    recv_fold(space, page_no), recv_addr);
		recv_sys->n_addrs++;
#if 0
		fprintf(stderr, "Inserting log rec for space %lu, page %lu\n",
			space, page_no);
#endif
	}

	UT_LIST_ADD_LAST(rec_list, recv_addr->rec_list, recv);

	prev_field = &(recv->data);

	/* Store the log record body in chunks of less than UNIV_PAGE_SIZE:
	recv_sys->heap grows into the buffer pool, and bigger chunks could not
	be allocated */

	while (rec_end > body) {

		len = rec_end - body;

		if (len > RECV_DATA_BLOCK_SIZE) {
			len = RECV_DATA_BLOCK_SIZE;
		}

		recv_data = mem_heap_alloc(recv_sys->heap,
					   sizeof(recv_data_t) + len);
		*prev_field = recv_data;

		ut_memcpy(((byte*)recv_data) + sizeof(recv_data_t), body, len);

		prev_field = &(recv_data->next);

		body += len;
	}

	*prev_field = NULL;
}

/*************************************************************************
Copies the log record body from recv to buf. */
static
void
recv_data_copy_to_buf(
/*==================*/
	byte*	buf,	/* in: buffer of length at least recv->len */
	recv_t*	recv)	/* in: log record */
{
	recv_data_t*	recv_data;
	ulint		part_len;
	ulint		len;

	len = recv->len;
	recv_data = recv->data;

	while (len > 0) {
		if (len > RECV_DATA_BLOCK_SIZE) {
			part_len = RECV_DATA_BLOCK_SIZE;
		} else {
			part_len = len;
		}

		ut_memcpy(buf, ((byte*)recv_data) + sizeof(recv_data_t),
			  part_len);
		buf += part_len;
		len -= part_len;

		recv_data = recv_data->next;
	}
}

/****************************************************************************
Applies the hashed log records to the page, if the page lsn is less than the
lsn of a log record. This can be called when a buffer page has just been
read in, or also for a page already in the buffer pool. */
UNIV_INTERN
void
recv_recover_page(
/*==============*/
	ibool		recover_backup,
				/* in: TRUE if we are recovering a backup
				page: then we do not acquire any latches
				since the page was read in outside the
				buffer pool */
	ibool		just_read_in,
				/* in: TRUE if the i/o-handler calls this for
				a freshly read page */
	buf_block_t*	block)	/* in: buffer block */
{
	page_t*		page;
	recv_addr_t*	recv_addr;
	recv_t*		recv;
	byte*		buf;
	ib_uint64_t	start_lsn;
	ib_uint64_t	end_lsn;
	ib_uint64_t	page_lsn;
	ib_uint64_t	page_newest_lsn;
	ibool		modification_to_page;
	ibool		success;
	mtr_t		mtr;

	mutex_enter(&(recv_sys->mutex));

	if (recv_sys->apply_log_recs == FALSE) {

		/* Log records should not be applied now */

		mutex_exit(&(recv_sys->mutex));

		return;
	}

	recv_addr = recv_get_fil_addr_struct(buf_block_get_space(block),
					     buf_block_get_page_no(block));

	if ((recv_addr == NULL)
Vadim Tkachenko's avatar
Vadim Tkachenko committed
1228 1229
		/* bugfix: http://bugs.mysql.com/bug.php?id=44140 */
	    || (recv_addr->state == RECV_BEING_READ && !just_read_in)
1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356
	    || (recv_addr->state == RECV_BEING_PROCESSED)
	    || (recv_addr->state == RECV_PROCESSED)) {

		mutex_exit(&(recv_sys->mutex));

		return;
	}

#if 0
	fprintf(stderr, "Recovering space %lu, page %lu\n",
		buf_block_get_space(block), buf_block_get_page_no(block));
#endif

	recv_addr->state = RECV_BEING_PROCESSED;

	mutex_exit(&(recv_sys->mutex));

	mtr_start(&mtr);
	mtr_set_log_mode(&mtr, MTR_LOG_NONE);

	page = block->frame;

	if (!recover_backup) {
		if (just_read_in) {
			/* Move the ownership of the x-latch on the
			page to this OS thread, so that we can acquire
			a second x-latch on it. This is needed for the
			operations to the page to pass the debug
			checks. */

			rw_lock_x_lock_move_ownership(&(block->lock));
		}

		success = buf_page_get_known_nowait(RW_X_LATCH, block,
						    BUF_KEEP_OLD,
						    __FILE__, __LINE__,
						    &mtr);
		ut_a(success);

		buf_block_dbg_add_level(block, SYNC_NO_ORDER_CHECK);
	}

	/* Read the newest modification lsn from the page */
	page_lsn = mach_read_ull(page + FIL_PAGE_LSN);

	if (!recover_backup) {
		/* It may be that the page has been modified in the buffer
		pool: read the newest modification lsn there */

		page_newest_lsn
			= buf_page_get_newest_modification(&block->page);

		if (page_newest_lsn) {

			page_lsn = page_newest_lsn;
		}
	} else {
		/* In recovery from a backup we do not really use the buffer
		pool */

		page_newest_lsn = 0;
	}

	modification_to_page = FALSE;
	start_lsn = end_lsn = 0;

	recv = UT_LIST_GET_FIRST(recv_addr->rec_list);

	while (recv) {
		end_lsn = recv->end_lsn;

		if (recv->len > RECV_DATA_BLOCK_SIZE) {
			/* We have to copy the record body to a separate
			buffer */

			buf = mem_alloc(recv->len);

			recv_data_copy_to_buf(buf, recv);
		} else {
			buf = ((byte*)(recv->data)) + sizeof(recv_data_t);
		}

		if (recv->type == MLOG_INIT_FILE_PAGE) {
			page_lsn = page_newest_lsn;

			mach_write_ull(page + UNIV_PAGE_SIZE
				       - FIL_PAGE_END_LSN_OLD_CHKSUM, 0);
			mach_write_ull(page + FIL_PAGE_LSN, 0);
		}

		if (recv->start_lsn >= page_lsn) {

			if (!modification_to_page) {

				modification_to_page = TRUE;
				start_lsn = recv->start_lsn;
			}

#ifdef UNIV_DEBUG
			if (log_debug_writes) {
				fprintf(stderr,
					"InnoDB: Applying log rec"
					" type %lu len %lu"
					" to space %lu page no %lu\n",
					(ulong) recv->type, (ulong) recv->len,
					(ulong) recv_addr->space,
					(ulong) recv_addr->page_no);
			}
#endif /* UNIV_DEBUG */

			recv_parse_or_apply_log_rec_body(recv->type, buf,
							 buf + recv->len,
							 block, &mtr);
			mach_write_ull(page + UNIV_PAGE_SIZE
				       - FIL_PAGE_END_LSN_OLD_CHKSUM,
				       recv->start_lsn + recv->len);
			mach_write_ull(page + FIL_PAGE_LSN,
				       recv->start_lsn + recv->len);
		}

		if (recv->len > RECV_DATA_BLOCK_SIZE) {
			mem_free(buf);
		}

		recv = UT_LIST_GET_NEXT(rec_list, recv);
	}

Vadim Tkachenko's avatar
Vadim Tkachenko committed
1357 1358 1359 1360 1361 1362 1363 1364 1365 1366
#ifdef UNIV_ZIP_DEBUG
	if (fil_page_get_type(page) == FIL_PAGE_INDEX) {
		page_zip_des_t*	page_zip = buf_block_get_page_zip(block);

		if (page_zip) {
			ut_a(page_zip_validate_low(page_zip, page, FALSE));
		}
	}
#endif /* UNIV_ZIP_DEBUG */

1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516
	mutex_enter(&(recv_sys->mutex));

	if (recv_max_page_lsn < page_lsn) {
		recv_max_page_lsn = page_lsn;
	}

	recv_addr->state = RECV_PROCESSED;

	ut_a(recv_sys->n_addrs);
	recv_sys->n_addrs--;

	mutex_exit(&(recv_sys->mutex));

	if (!recover_backup && modification_to_page) {
		ut_a(block);

		buf_flush_recv_note_modification(block, start_lsn, end_lsn);
	}

	/* Make sure that committing mtr does not change the modification
	lsn values of page */

	mtr.modifications = FALSE;

	mtr_commit(&mtr);
}

/***********************************************************************
Reads in pages which have hashed log records, from an area around a given
page number. */
static
ulint
recv_read_in_area(
/*==============*/
			/* out: number of pages found */
	ulint	space,	/* in: space */
	ulint	zip_size,/* in: compressed page size in bytes, or 0 */
	ulint	page_no)/* in: page number */
{
	recv_addr_t* recv_addr;
	ulint	page_nos[RECV_READ_AHEAD_AREA];
	ulint	low_limit;
	ulint	n;

	low_limit = page_no - (page_no % RECV_READ_AHEAD_AREA);

	n = 0;

	for (page_no = low_limit; page_no < low_limit + RECV_READ_AHEAD_AREA;
	     page_no++) {
		recv_addr = recv_get_fil_addr_struct(space, page_no);

		if (recv_addr && !buf_page_peek(space, page_no)) {

			mutex_enter(&(recv_sys->mutex));

			if (recv_addr->state == RECV_NOT_PROCESSED) {
				recv_addr->state = RECV_BEING_READ;

				page_nos[n] = page_no;

				n++;
			}

			mutex_exit(&(recv_sys->mutex));
		}
	}

	buf_read_recv_pages(FALSE, space, zip_size, page_nos, n);
	/*
	fprintf(stderr, "Recv pages at %lu n %lu\n", page_nos[0], n);
	*/
	return(n);
}

/***********************************************************************
Empties the hash table of stored log records, applying them to appropriate
pages. */
UNIV_INTERN
void
recv_apply_hashed_log_recs(
/*=======================*/
	ibool	allow_ibuf)	/* in: if TRUE, also ibuf operations are
				allowed during the application; if FALSE,
				no ibuf operations are allowed, and after
				the application all file pages are flushed to
				disk and invalidated in buffer pool: this
				alternative means that no new log records
				can be generated during the application;
				the caller must in this case own the log
				mutex */
{
	recv_addr_t* recv_addr;
	ulint	i;
	ulint	n_pages;
	ibool	has_printed	= FALSE;
	mtr_t	mtr;
loop:
	mutex_enter(&(recv_sys->mutex));

	if (recv_sys->apply_batch_on) {

		mutex_exit(&(recv_sys->mutex));

		os_thread_sleep(500000);

		goto loop;
	}

	ut_ad(!allow_ibuf == mutex_own(&log_sys->mutex));

	if (!allow_ibuf) {
		recv_no_ibuf_operations = TRUE;
	}

	recv_sys->apply_log_recs = TRUE;
	recv_sys->apply_batch_on = TRUE;

	for (i = 0; i < hash_get_n_cells(recv_sys->addr_hash); i++) {

		recv_addr = HASH_GET_FIRST(recv_sys->addr_hash, i);

		while (recv_addr) {
			ulint	space = recv_addr->space;
			ulint	zip_size = fil_space_get_zip_size(space);
			ulint	page_no = recv_addr->page_no;

			if (recv_addr->state == RECV_NOT_PROCESSED) {
				if (!has_printed) {
					ut_print_timestamp(stderr);
					fputs("  InnoDB: Starting an"
					      " apply batch of log records"
					      " to the database...\n"
					      "InnoDB: Progress in percents: ",
					      stderr);
					has_printed = TRUE;
				}

				mutex_exit(&(recv_sys->mutex));

				if (buf_page_peek(space, page_no)) {
					buf_block_t*	block;

					mtr_start(&mtr);

					block = buf_page_get(
						space, zip_size, page_no,
						RW_X_LATCH, &mtr);
					buf_block_dbg_add_level(
						block, SYNC_NO_ORDER_CHECK);
Vadim Tkachenko's avatar
Vadim Tkachenko committed
1517

1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 2452 2453 2454 2455 2456 2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 2509 2510 2511 2512 2513 2514 2515 2516 2517 2518 2519 2520 2521 2522 2523 2524 2525 2526 2527 2528 2529 2530 2531 2532 2533 2534 2535 2536 2537 2538 2539 2540 2541 2542 2543 2544 2545 2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557 2558 2559 2560 2561 2562 2563 2564 2565 2566
					recv_recover_page(FALSE, FALSE, block);
					mtr_commit(&mtr);
				} else {
					recv_read_in_area(space, zip_size,
							  page_no);
				}

				mutex_enter(&(recv_sys->mutex));
			}

			recv_addr = HASH_GET_NEXT(addr_hash, recv_addr);
		}

		if (has_printed
		    && (i * 100) / hash_get_n_cells(recv_sys->addr_hash)
		    != ((i + 1) * 100)
		    / hash_get_n_cells(recv_sys->addr_hash)) {

			fprintf(stderr, "%lu ", (ulong)
				((i * 100)
				 / hash_get_n_cells(recv_sys->addr_hash)));
		}
	}

	/* Wait until all the pages have been processed */

	while (recv_sys->n_addrs != 0) {

		mutex_exit(&(recv_sys->mutex));

		os_thread_sleep(500000);

		mutex_enter(&(recv_sys->mutex));
	}

	if (has_printed) {

		fprintf(stderr, "\n");
	}

	if (!allow_ibuf) {
		/* Flush all the file pages to disk and invalidate them in
		the buffer pool */

		mutex_exit(&(recv_sys->mutex));
		mutex_exit(&(log_sys->mutex));

		n_pages = buf_flush_batch(BUF_FLUSH_LIST, ULINT_MAX,
					  IB_ULONGLONG_MAX);
		ut_a(n_pages != ULINT_UNDEFINED);

		buf_flush_wait_batch_end(BUF_FLUSH_LIST);

		buf_pool_invalidate();

		mutex_enter(&(log_sys->mutex));
		mutex_enter(&(recv_sys->mutex));

		recv_no_ibuf_operations = FALSE;
	}

	recv_sys->apply_log_recs = FALSE;
	recv_sys->apply_batch_on = FALSE;

	recv_sys_empty_hash();

	if (has_printed) {
		fprintf(stderr, "InnoDB: Apply batch completed\n");
	}

	mutex_exit(&(recv_sys->mutex));
}

#ifdef UNIV_HOTBACKUP
/***********************************************************************
Applies log records in the hash table to a backup. */
UNIV_INTERN
void
recv_apply_log_recs_for_backup(void)
/*================================*/
{
	recv_addr_t*	recv_addr;
	ulint		n_hash_cells;
	buf_block_t*	block;
	ulint		actual_size;
	ibool		success;
	ulint		error;
	ulint		i;

	recv_sys->apply_log_recs = TRUE;
	recv_sys->apply_batch_on = TRUE;

	block = buf_LRU_get_free_block(UNIV_PAGE_SIZE);

	fputs("InnoDB: Starting an apply batch of log records"
	      " to the database...\n"
	      "InnoDB: Progress in percents: ", stderr);

	n_hash_cells = hash_get_n_cells(recv_sys->addr_hash);

	for (i = 0; i < n_hash_cells; i++) {
		/* The address hash table is externally chained */
		recv_addr = hash_get_nth_cell(recv_sys->addr_hash, i)->node;

		while (recv_addr != NULL) {

			ulint	zip_size
				= fil_space_get_zip_size(recv_addr->space);

			if (zip_size == ULINT_UNDEFINED) {
#if 0
				fprintf(stderr,
					"InnoDB: Warning: cannot apply"
					" log record to"
					" tablespace %lu page %lu,\n"
					"InnoDB: because tablespace with"
					" that id does not exist.\n",
					recv_addr->space, recv_addr->page_no);
#endif
				recv_addr->state = RECV_PROCESSED;

				ut_a(recv_sys->n_addrs);
				recv_sys->n_addrs--;

				goto skip_this_recv_addr;
			}

			/* We simulate a page read made by the buffer pool, to
			make sure the recovery apparatus works ok. We must init
			the block. */

			buf_page_init_for_backup_restore(
				recv_addr->space, recv_addr->page_no,
				zip_size, block);

			/* Extend the tablespace's last file if the page_no
			does not fall inside its bounds; we assume the last
			file is auto-extending, and ibbackup copied the file
			when it still was smaller */

			success = fil_extend_space_to_desired_size(
				&actual_size,
				recv_addr->space, recv_addr->page_no + 1);
			if (!success) {
				fprintf(stderr,
					"InnoDB: Fatal error: cannot extend"
					" tablespace %lu to hold %lu pages\n",
					recv_addr->space, recv_addr->page_no);

				exit(1);
			}

			/* Read the page from the tablespace file using the
			fil0fil.c routines */

			if (zip_size) {
				error = fil_io(OS_FILE_READ, TRUE,
					       recv_addr->space, zip_size,
					       recv_addr->page_no, 0, zip_size,
					       block->page.zip.data, NULL);
			} else {
				error = fil_io(OS_FILE_READ, TRUE,
					       recv_addr->space, 0,
					       recv_addr->page_no, 0,
					       UNIV_PAGE_SIZE,
					       block->frame, NULL);
			}

			if (error != DB_SUCCESS) {
				fprintf(stderr,
					"InnoDB: Fatal error: cannot read"
					" from tablespace"
					" %lu page number %lu\n",
					(ulong) recv_addr->space,
					(ulong) recv_addr->page_no);

				exit(1);
			}

			/* Apply the log records to this page */
			recv_recover_page(TRUE, FALSE, block);

			/* Write the page back to the tablespace file using the
			fil0fil.c routines */

			buf_flush_init_for_writing(
				block->frame, buf_block_get_page_zip(block),
				mach_read_ull(block->frame + FIL_PAGE_LSN));

			if (zip_size) {
				error = fil_io(OS_FILE_WRITE, TRUE,
					       recv_addr->space, zip_size,
					       recv_addr->page_no, 0,
					       zip_size,
					       block->page.zip.data, NULL);
			} else {
				error = fil_io(OS_FILE_WRITE, TRUE,
					       recv_addr->space, 0,
					       recv_addr->page_no, 0,
					       UNIV_PAGE_SIZE,
					       block->frame, NULL);
			}
skip_this_recv_addr:
			recv_addr = HASH_GET_NEXT(addr_hash, recv_addr);
		}

		if ((100 * i) / n_hash_cells
		    != (100 * (i + 1)) / n_hash_cells) {
			fprintf(stderr, "%lu ",
				(ulong) ((100 * i) / n_hash_cells));
			fflush(stderr);
		}
	}

	buf_block_free(block);
	recv_sys_empty_hash();
}
#endif /* UNIV_HOTBACKUP */

/***********************************************************************
Tries to parse a single log record and returns its length. */
static
ulint
recv_parse_log_rec(
/*===============*/
			/* out: length of the record, or 0 if the record was
			not complete */
	byte*	ptr,	/* in: pointer to a buffer */
	byte*	end_ptr,/* in: pointer to the buffer end */
	byte*	type,	/* out: type */
	ulint*	space,	/* out: space id */
	ulint*	page_no,/* out: page number */
	byte**	body)	/* out: log record body start */
{
	byte*	new_ptr;

	*body = NULL;

	if (ptr == end_ptr) {

		return(0);
	}

	if (*ptr == MLOG_MULTI_REC_END) {

		*type = *ptr;

		return(1);
	}

	if (*ptr == MLOG_DUMMY_RECORD) {
		*type = *ptr;

		*space = ULINT_UNDEFINED - 1; /* For debugging */

		return(1);
	}

	new_ptr = mlog_parse_initial_log_record(ptr, end_ptr, type, space,
						page_no);
	*body = new_ptr;

	if (UNIV_UNLIKELY(!new_ptr)) {

		return(0);
	}

	/* Check that page_no is sensible */

	if (UNIV_UNLIKELY(*page_no > 0x8FFFFFFFUL)) {

		recv_sys->found_corrupt_log = TRUE;

		return(0);
	}

	new_ptr = recv_parse_or_apply_log_rec_body(*type, new_ptr, end_ptr,
						   NULL, NULL);
	if (UNIV_UNLIKELY(new_ptr == NULL)) {

		return(0);
	}

	if (*page_no > recv_max_parsed_page_no) {
		recv_max_parsed_page_no = *page_no;
	}

	return(new_ptr - ptr);
}

/***********************************************************
Calculates the new value for lsn when more data is added to the log. */
static
ib_uint64_t
recv_calc_lsn_on_data_add(
/*======================*/
	ib_uint64_t	lsn,	/* in: old lsn */
	ib_uint64_t	len)	/* in: this many bytes of data is
				added, log block headers not included */
{
	ulint	frag_len;
	ulint	lsn_len;

	frag_len = (((ulint) lsn) % OS_FILE_LOG_BLOCK_SIZE)
		- LOG_BLOCK_HDR_SIZE;
	ut_ad(frag_len < OS_FILE_LOG_BLOCK_SIZE - LOG_BLOCK_HDR_SIZE
	      - LOG_BLOCK_TRL_SIZE);
	lsn_len = (ulint) len;
	lsn_len += (lsn_len + frag_len)
		/ (OS_FILE_LOG_BLOCK_SIZE - LOG_BLOCK_HDR_SIZE
		   - LOG_BLOCK_TRL_SIZE)
		* (LOG_BLOCK_HDR_SIZE + LOG_BLOCK_TRL_SIZE);

	return(lsn + lsn_len);
}

#ifdef UNIV_LOG_DEBUG
/***********************************************************
Checks that the parser recognizes incomplete initial segments of a log
record as incomplete. */
static
void
recv_check_incomplete_log_recs(
/*===========================*/
	byte*	ptr,	/* in: pointer to a complete log record */
	ulint	len)	/* in: length of the log record */
{
	ulint	i;
	byte	type;
	ulint	space;
	ulint	page_no;
	byte*	body;

	for (i = 0; i < len; i++) {
		ut_a(0 == recv_parse_log_rec(ptr, ptr + i, &type, &space,
					     &page_no, &body));
	}
}
#endif /* UNIV_LOG_DEBUG */

/***********************************************************
Prints diagnostic info of corrupt log. */
static
void
recv_report_corrupt_log(
/*====================*/
	byte*	ptr,	/* in: pointer to corrupt log record */
	byte	type,	/* in: type of the record */
	ulint	space,	/* in: space id, this may also be garbage */
	ulint	page_no)/* in: page number, this may also be garbage */
{
	fprintf(stderr,
		"InnoDB: ############### CORRUPT LOG RECORD FOUND\n"
		"InnoDB: Log record type %lu, space id %lu, page number %lu\n"
		"InnoDB: Log parsing proceeded successfully up to %llu\n"
		"InnoDB: Previous log record type %lu, is multi %lu\n"
		"InnoDB: Recv offset %lu, prev %lu\n",
		(ulong) type, (ulong) space, (ulong) page_no,
		recv_sys->recovered_lsn,
		(ulong) recv_previous_parsed_rec_type,
		(ulong) recv_previous_parsed_rec_is_multi,
		(ulong) (ptr - recv_sys->buf),
		(ulong) recv_previous_parsed_rec_offset);

	if ((ulint)(ptr - recv_sys->buf + 100)
	    > recv_previous_parsed_rec_offset
	    && (ulint)(ptr - recv_sys->buf + 100
		       - recv_previous_parsed_rec_offset)
	    < 200000) {
		fputs("InnoDB: Hex dump of corrupt log starting"
		      " 100 bytes before the start\n"
		      "InnoDB: of the previous log rec,\n"
		      "InnoDB: and ending 100 bytes after the start"
		      " of the corrupt rec:\n",
		      stderr);

		ut_print_buf(stderr,
			     recv_sys->buf
			     + recv_previous_parsed_rec_offset - 100,
			     ptr - recv_sys->buf + 200
			     - recv_previous_parsed_rec_offset);
		putc('\n', stderr);
	}

	fputs("InnoDB: WARNING: the log file may have been corrupt and it\n"
	      "InnoDB: is possible that the log scan did not proceed\n"
	      "InnoDB: far enough in recovery! Please run CHECK TABLE\n"
	      "InnoDB: on your InnoDB tables to check that they are ok!\n"
	      "InnoDB: If mysqld crashes after this recovery, look at\n"
	      "InnoDB: http://dev.mysql.com/doc/refman/5.1/en/"
	      "forcing-recovery.html\n"
	      "InnoDB: about forcing recovery.\n", stderr);

	fflush(stderr);
}

/***********************************************************
Parses log records from a buffer and stores them to a hash table to wait
merging to file pages. */
static
ibool
recv_parse_log_recs(
/*================*/
				/* out: currently always returns FALSE */
	ibool	store_to_hash)	/* in: TRUE if the records should be stored
				to the hash table; this is set to FALSE if just
				debug checking is needed */
{
	byte*		ptr;
	byte*		end_ptr;
	ulint		single_rec;
	ulint		len;
	ulint		total_len;
	ib_uint64_t	new_recovered_lsn;
	ib_uint64_t	old_lsn;
	byte		type;
	ulint		space;
	ulint		page_no;
	byte*		body;
	ulint		n_recs;

	ut_ad(mutex_own(&(log_sys->mutex)));
	ut_ad(recv_sys->parse_start_lsn != 0);
loop:
	ptr = recv_sys->buf + recv_sys->recovered_offset;

	end_ptr = recv_sys->buf + recv_sys->len;

	if (ptr == end_ptr) {

		return(FALSE);
	}

	single_rec = (ulint)*ptr & MLOG_SINGLE_REC_FLAG;

	if (single_rec || *ptr == MLOG_DUMMY_RECORD) {
		/* The mtr only modified a single page, or this is a file op */

		old_lsn = recv_sys->recovered_lsn;

		/* Try to parse a log record, fetching its type, space id,
		page no, and a pointer to the body of the log record */

		len = recv_parse_log_rec(ptr, end_ptr, &type, &space,
					 &page_no, &body);

		if (len == 0 || recv_sys->found_corrupt_log) {
			if (recv_sys->found_corrupt_log) {

				recv_report_corrupt_log(ptr,
							type, space, page_no);
			}

			return(FALSE);
		}

		new_recovered_lsn = recv_calc_lsn_on_data_add(old_lsn, len);

		if (new_recovered_lsn > recv_sys->scanned_lsn) {
			/* The log record filled a log block, and we require
			that also the next log block should have been scanned
			in */

			return(FALSE);
		}

		recv_previous_parsed_rec_type = (ulint)type;
		recv_previous_parsed_rec_offset = recv_sys->recovered_offset;
		recv_previous_parsed_rec_is_multi = 0;

		recv_sys->recovered_offset += len;
		recv_sys->recovered_lsn = new_recovered_lsn;

#ifdef UNIV_DEBUG
		if (log_debug_writes) {
			fprintf(stderr,
				"InnoDB: Parsed a single log rec"
				" type %lu len %lu space %lu page no %lu\n",
				(ulong) type, (ulong) len, (ulong) space,
				(ulong) page_no);
		}
#endif /* UNIV_DEBUG */

		if (type == MLOG_DUMMY_RECORD) {
			/* Do nothing */

		} else if (!store_to_hash) {
			/* In debug checking, update a replicate page
			according to the log record, and check that it
			becomes identical with the original page */
#ifdef UNIV_LOG_DEBUG
			recv_check_incomplete_log_recs(ptr, len);
#endif/* UNIV_LOG_DEBUG */

		} else if (type == MLOG_FILE_CREATE
			   || type == MLOG_FILE_CREATE2
			   || type == MLOG_FILE_RENAME
			   || type == MLOG_FILE_DELETE) {
			ut_a(space);
#ifdef UNIV_HOTBACKUP
			if (recv_replay_file_ops) {

				/* In ibbackup --apply-log, replay an .ibd file
				operation, if possible; note that
				fil_path_to_mysql_datadir is set in ibbackup to
				point to the datadir we should use there */

				if (NULL == fil_op_log_parse_or_replay(
					    body, end_ptr, type, space)) {
					fprintf(stderr,
						"InnoDB: Error: file op"
						" log record of type %lu"
						" space %lu not complete in\n"
						"InnoDB: the replay phase."
						" Path %s\n",
						(ulint)type, space,
						(char*)(body + 2));

					ut_error;
				}
			}
#endif
			/* In normal mysqld crash recovery we do not try to
			replay file operations */
		} else {
			recv_add_to_hash_table(type, space, page_no, body,
					       ptr + len, old_lsn,
					       recv_sys->recovered_lsn);
		}
	} else {
		/* Check that all the records associated with the single mtr
		are included within the buffer */

		total_len = 0;
		n_recs = 0;

		for (;;) {
			len = recv_parse_log_rec(ptr, end_ptr, &type, &space,
						 &page_no, &body);
			if (len == 0 || recv_sys->found_corrupt_log) {

				if (recv_sys->found_corrupt_log) {

					recv_report_corrupt_log(
						ptr, type, space, page_no);
				}

				return(FALSE);
			}

			recv_previous_parsed_rec_type = (ulint)type;
			recv_previous_parsed_rec_offset
				= recv_sys->recovered_offset + total_len;
			recv_previous_parsed_rec_is_multi = 1;

			if ((!store_to_hash) && (type != MLOG_MULTI_REC_END)) {
#ifdef UNIV_LOG_DEBUG
				recv_check_incomplete_log_recs(ptr, len);
#endif /* UNIV_LOG_DEBUG */
			}

#ifdef UNIV_DEBUG
			if (log_debug_writes) {
				fprintf(stderr,
					"InnoDB: Parsed a multi log rec"
					" type %lu len %lu"
					" space %lu page no %lu\n",
					(ulong) type, (ulong) len,
					(ulong) space, (ulong) page_no);
			}
#endif /* UNIV_DEBUG */

			total_len += len;
			n_recs++;

			ptr += len;

			if (type == MLOG_MULTI_REC_END) {

				/* Found the end mark for the records */

				break;
			}
		}

		new_recovered_lsn = recv_calc_lsn_on_data_add(
			recv_sys->recovered_lsn, total_len);

		if (new_recovered_lsn > recv_sys->scanned_lsn) {
			/* The log record filled a log block, and we require
			that also the next log block should have been scanned
			in */

			return(FALSE);
		}

		/* Add all the records to the hash table */

		ptr = recv_sys->buf + recv_sys->recovered_offset;

		for (;;) {
			old_lsn = recv_sys->recovered_lsn;
			len = recv_parse_log_rec(ptr, end_ptr, &type, &space,
						 &page_no, &body);
			if (recv_sys->found_corrupt_log) {

				recv_report_corrupt_log(ptr,
							type, space, page_no);
			}

			ut_a(len != 0);
			ut_a(0 == ((ulint)*ptr & MLOG_SINGLE_REC_FLAG));

			recv_sys->recovered_offset += len;
			recv_sys->recovered_lsn
				= recv_calc_lsn_on_data_add(old_lsn, len);
			if (type == MLOG_MULTI_REC_END) {

				/* Found the end mark for the records */

				break;
			}

			if (store_to_hash) {
				recv_add_to_hash_table(type, space, page_no,
						       body, ptr + len,
						       old_lsn,
						       new_recovered_lsn);
			}

			ptr += len;
		}
	}

	goto loop;
}

/***********************************************************
Adds data from a new log block to the parsing buffer of recv_sys if
recv_sys->parse_start_lsn is non-zero. */
static
ibool
recv_sys_add_to_parsing_buf(
/*========================*/
					/* out: TRUE if more data added */
	byte*		log_block,	/* in: log block */
	ib_uint64_t	scanned_lsn)	/* in: lsn of how far we were able
					to find data in this log block */
{
	ulint	more_len;
	ulint	data_len;
	ulint	start_offset;
	ulint	end_offset;

	ut_ad(scanned_lsn >= recv_sys->scanned_lsn);

	if (!recv_sys->parse_start_lsn) {
		/* Cannot start parsing yet because no start point for
		it found */

		return(FALSE);
	}

	data_len = log_block_get_data_len(log_block);

	if (recv_sys->parse_start_lsn >= scanned_lsn) {

		return(FALSE);

	} else if (recv_sys->scanned_lsn >= scanned_lsn) {

		return(FALSE);

	} else if (recv_sys->parse_start_lsn > recv_sys->scanned_lsn) {
		more_len = (ulint) (scanned_lsn - recv_sys->parse_start_lsn);
	} else {
		more_len = (ulint) (scanned_lsn - recv_sys->scanned_lsn);
	}

	if (more_len == 0) {

		return(FALSE);
	}

	ut_ad(data_len >= more_len);

	start_offset = data_len - more_len;

	if (start_offset < LOG_BLOCK_HDR_SIZE) {
		start_offset = LOG_BLOCK_HDR_SIZE;
	}

	end_offset = data_len;

	if (end_offset > OS_FILE_LOG_BLOCK_SIZE - LOG_BLOCK_TRL_SIZE) {
		end_offset = OS_FILE_LOG_BLOCK_SIZE - LOG_BLOCK_TRL_SIZE;
	}

	ut_ad(start_offset <= end_offset);

	if (start_offset < end_offset) {
		ut_memcpy(recv_sys->buf + recv_sys->len,
			  log_block + start_offset, end_offset - start_offset);

		recv_sys->len += end_offset - start_offset;

		ut_a(recv_sys->len <= RECV_PARSING_BUF_SIZE);
	}

	return(TRUE);
}

/***********************************************************
Moves the parsing buffer data left to the buffer start. */
static
void
recv_sys_justify_left_parsing_buf(void)
/*===================================*/
{
	ut_memmove(recv_sys->buf, recv_sys->buf + recv_sys->recovered_offset,
		   recv_sys->len - recv_sys->recovered_offset);

	recv_sys->len -= recv_sys->recovered_offset;

	recv_sys->recovered_offset = 0;
}

/***********************************************************
Scans log from a buffer and stores new log data to the parsing buffer. Parses
and hashes the log records if new data found. */
UNIV_INTERN
ibool
recv_scan_log_recs(
/*===============*/
					/* out: TRUE if limit_lsn has been
					reached, or not able to scan any more
					in this log group */
	ibool		apply_automatically,/* in: TRUE if we want this
					function to apply log records
					automatically when the hash table
					becomes full; in the hot backup tool
					the tool does the applying, not this
					function */
	ulint		available_memory,/* in: we let the hash table of recs
					to grow to this size, at the maximum */
	ibool		store_to_hash,	/* in: TRUE if the records should be
					stored to the hash table; this is set
					to FALSE if just debug checking is
					needed */
	byte*		buf,		/* in: buffer containing a log segment
					or garbage */
	ulint		len,		/* in: buffer length */
	ib_uint64_t	start_lsn,	/* in: buffer start lsn */
	ib_uint64_t*	contiguous_lsn,	/* in/out: it is known that all log
					groups contain contiguous log data up
					to this lsn */
	ib_uint64_t*	group_scanned_lsn)/* out: scanning succeeded up to
					this lsn */
{
	byte*		log_block;
	ulint		no;
	ib_uint64_t	scanned_lsn;
	ibool		finished;
	ulint		data_len;
	ibool		more_data;

	ut_ad(start_lsn % OS_FILE_LOG_BLOCK_SIZE == 0);
	ut_ad(len % OS_FILE_LOG_BLOCK_SIZE == 0);
	ut_ad(len > 0);
	ut_a(apply_automatically <= TRUE);
	ut_a(store_to_hash <= TRUE);

	finished = FALSE;

	log_block = buf;
	scanned_lsn = start_lsn;
	more_data = FALSE;

	while (log_block < buf + len && !finished) {

		no = log_block_get_hdr_no(log_block);
		/*
		fprintf(stderr, "Log block header no %lu\n", no);

		fprintf(stderr, "Scanned lsn no %lu\n",
		log_block_convert_lsn_to_no(scanned_lsn));
		*/
		if (no != log_block_convert_lsn_to_no(scanned_lsn)
		    || !log_block_checksum_is_ok_or_old_format(log_block)) {

			if (no == log_block_convert_lsn_to_no(scanned_lsn)
			    && !log_block_checksum_is_ok_or_old_format(
				    log_block)) {
				fprintf(stderr,
					"InnoDB: Log block no %lu at"
					" lsn %llu has\n"
					"InnoDB: ok header, but checksum field"
					" contains %lu, should be %lu\n",
					(ulong) no,
					scanned_lsn,
					(ulong) log_block_get_checksum(
						log_block),
					(ulong) log_block_calc_checksum(
						log_block));
			}

			/* Garbage or an incompletely written log block */

			finished = TRUE;

			break;
		}

		if (log_block_get_flush_bit(log_block)) {
			/* This block was a start of a log flush operation:
			we know that the previous flush operation must have
			been completed for all log groups before this block
			can have been flushed to any of the groups. Therefore,
			we know that log data is contiguous up to scanned_lsn
			in all non-corrupt log groups. */

			if (scanned_lsn > *contiguous_lsn) {
				*contiguous_lsn = scanned_lsn;
			}
		}

		data_len = log_block_get_data_len(log_block);

		if ((store_to_hash || (data_len == OS_FILE_LOG_BLOCK_SIZE))
		    && scanned_lsn + data_len > recv_sys->scanned_lsn
		    && (recv_sys->scanned_checkpoint_no > 0)
		    && (log_block_get_checkpoint_no(log_block)
			< recv_sys->scanned_checkpoint_no)
		    && (recv_sys->scanned_checkpoint_no
			- log_block_get_checkpoint_no(log_block)
			> 0x80000000UL)) {

			/* Garbage from a log buffer flush which was made
			before the most recent database recovery */

			finished = TRUE;
#ifdef UNIV_LOG_DEBUG
			/* This is not really an error, but currently
			we stop here in the debug version: */

			ut_error;
#endif
			break;
		}

		if (!recv_sys->parse_start_lsn
		    && (log_block_get_first_rec_group(log_block) > 0)) {

			/* We found a point from which to start the parsing
			of log records */

			recv_sys->parse_start_lsn = scanned_lsn
				+ log_block_get_first_rec_group(log_block);
			recv_sys->scanned_lsn = recv_sys->parse_start_lsn;
			recv_sys->recovered_lsn = recv_sys->parse_start_lsn;
		}

		scanned_lsn += data_len;

		if (scanned_lsn > recv_sys->scanned_lsn) {

			/* We have found more entries. If this scan is
 			of startup type, we must initiate crash recovery
			environment before parsing these log records. */

			if (recv_log_scan_is_startup_type
			    && !recv_needed_recovery) {

				fprintf(stderr,
					"InnoDB: Log scan progressed"
					" past the checkpoint lsn %llu\n",
					recv_sys->scanned_lsn);
				recv_init_crash_recovery();
			}

			/* We were able to find more log data: add it to the
			parsing buffer if parse_start_lsn is already
			non-zero */

			if (recv_sys->len + 4 * OS_FILE_LOG_BLOCK_SIZE
			    >= RECV_PARSING_BUF_SIZE) {
				fprintf(stderr,
					"InnoDB: Error: log parsing"
					" buffer overflow."
					" Recovery may have failed!\n");

				recv_sys->found_corrupt_log = TRUE;

			} else if (!recv_sys->found_corrupt_log) {
				more_data = recv_sys_add_to_parsing_buf(
					log_block, scanned_lsn);
			}

			recv_sys->scanned_lsn = scanned_lsn;
			recv_sys->scanned_checkpoint_no
				= log_block_get_checkpoint_no(log_block);
		}

		if (data_len < OS_FILE_LOG_BLOCK_SIZE) {
			/* Log data for this group ends here */

			finished = TRUE;
		} else {
			log_block += OS_FILE_LOG_BLOCK_SIZE;
		}
	}

	*group_scanned_lsn = scanned_lsn;

	if (recv_needed_recovery
	    || (recv_is_from_backup && !recv_is_making_a_backup)) {
		recv_scan_print_counter++;

		if (finished || (recv_scan_print_counter % 80 == 0)) {

			fprintf(stderr,
				"InnoDB: Doing recovery: scanned up to"
				" log sequence number %llu\n",
				*group_scanned_lsn);
		}
	}

	if (more_data && !recv_sys->found_corrupt_log) {
		/* Try to parse more log records */

		recv_parse_log_recs(store_to_hash);

		if (store_to_hash && mem_heap_get_size(recv_sys->heap)
		    > available_memory
		    && apply_automatically) {

			/* Hash table of log records has grown too big:
			empty it; FALSE means no ibuf operations
			allowed, as we cannot add new records to the
			log yet: they would be produced by ibuf
			operations */

			recv_apply_hashed_log_recs(FALSE);
		}

		if (recv_sys->recovered_offset > RECV_PARSING_BUF_SIZE / 4) {
			/* Move parsing buffer data to the buffer start */

			recv_sys_justify_left_parsing_buf();
		}
	}

	return(finished);
}

/***********************************************************
Scans log from a buffer and stores new log data to the parsing buffer. Parses
and hashes the log records if new data found. */
static
void
recv_group_scan_log_recs(
/*=====================*/
	log_group_t*	group,		/* in: log group */
	ib_uint64_t*	contiguous_lsn,	/* in/out: it is known that all log
					groups contain contiguous log data up
					to this lsn */
	ib_uint64_t*	group_scanned_lsn)/* out: scanning succeeded up to
					this lsn */
{
	ibool		finished;
	ib_uint64_t	start_lsn;
	ib_uint64_t	end_lsn;

	finished = FALSE;

	start_lsn = *contiguous_lsn;

	while (!finished) {
		end_lsn = start_lsn + RECV_SCAN_SIZE;

		log_group_read_log_seg(LOG_RECOVER, log_sys->buf,
				       group, start_lsn, end_lsn);

		finished = recv_scan_log_recs(
			TRUE, (buf_pool->curr_size - recv_n_pool_free_frames)
			* UNIV_PAGE_SIZE, TRUE, log_sys->buf, RECV_SCAN_SIZE,
			start_lsn, contiguous_lsn, group_scanned_lsn);
		start_lsn = end_lsn;
	}

#ifdef UNIV_DEBUG
	if (log_debug_writes) {
		fprintf(stderr,
			"InnoDB: Scanned group %lu up to"
			" log sequence number %llu\n",
			(ulong) group->id,
			*group_scanned_lsn);
	}
#endif /* UNIV_DEBUG */
}

/***********************************************************
Initialize crash recovery environment. Can be called iff
recv_needed_recovery == FALSE. */
static
void
recv_init_crash_recovery(void)
/*==========================*/
{
	ut_a(!recv_needed_recovery);

	recv_needed_recovery = TRUE;

	ut_print_timestamp(stderr);

	fprintf(stderr,
		"  InnoDB: Database was not"
		" shut down normally!\n"
		"InnoDB: Starting crash recovery.\n");

	fprintf(stderr,
		"InnoDB: Reading tablespace information"
		" from the .ibd files...\n");

	fil_load_single_table_tablespaces();

	/* If we are using the doublewrite method, we will
	check if there are half-written pages in data files,
	and restore them from the doublewrite buffer if
	possible */

	if (srv_force_recovery < SRV_FORCE_NO_LOG_REDO) {

		fprintf(stderr,
			"InnoDB: Restoring possible"
			" half-written data pages from"
			" the doublewrite\n"
			"InnoDB: buffer...\n");
		trx_sys_doublewrite_init_or_restore_pages(TRUE);
	}
}

/************************************************************
Recovers from a checkpoint. When this function returns, the database is able
to start processing of new user transactions, but the function
recv_recovery_from_checkpoint_finish should be called later to complete
the recovery and free the resources used in it. */
UNIV_INTERN
ulint
Vadim Tkachenko's avatar
Vadim Tkachenko committed
2567 2568
recv_recovery_from_checkpoint_start_func(
/*=====================================*/
2569
					/* out: error code or DB_SUCCESS */
Vadim Tkachenko's avatar
Vadim Tkachenko committed
2570
#ifdef UNIV_LOG_ARCHIVE
2571 2572 2573
	ulint		type,		/* in: LOG_CHECKPOINT or LOG_ARCHIVE */
	ib_uint64_t	limit_lsn,	/* in: recover up to this lsn
					if possible */
Vadim Tkachenko's avatar
Vadim Tkachenko committed
2574
#endif /* UNIV_LOG_ARCHIVE */
2575 2576 2577 2578 2579 2580 2581 2582 2583 2584 2585 2586 2587 2588 2589 2590 2591 2592 2593
	ib_uint64_t	min_flushed_lsn,/* in: min flushed lsn from
					data files */
	ib_uint64_t	max_flushed_lsn)/* in: max flushed lsn from
					data files */
{
	log_group_t*	group;
	log_group_t*	max_cp_group;
	log_group_t*	up_to_date_group;
	ulint		max_cp_field;
	ib_uint64_t	checkpoint_lsn;
	ib_uint64_t	checkpoint_no;
	ib_uint64_t	old_scanned_lsn;
	ib_uint64_t	group_scanned_lsn;
	ib_uint64_t	contiguous_lsn;
	ib_uint64_t	archived_lsn;
	byte*		buf;
	byte		log_hdr_buf[LOG_FILE_HDR_SIZE];
	ulint		err;

Vadim Tkachenko's avatar
Vadim Tkachenko committed
2594
#ifdef UNIV_LOG_ARCHIVE
2595
	ut_ad(type != LOG_CHECKPOINT || limit_lsn == IB_ULONGLONG_MAX);
Vadim Tkachenko's avatar
Vadim Tkachenko committed
2596 2597 2598 2599 2600 2601
# define TYPE_CHECKPOINT	(type == LOG_CHECKPOINT)
# define LIMIT_LSN		limit_lsn
#else /* UNIV_LOG_ARCHIVE */
# define TYPE_CHECKPOINT	1
# define LIMIT_LSN		IB_ULONGLONG_MAX
#endif /* UNIV_LOG_ARCHIVE */
2602

Vadim Tkachenko's avatar
Vadim Tkachenko committed
2603
	if (TYPE_CHECKPOINT) {
2604 2605 2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617 2618
		recv_sys_create();
		recv_sys_init(FALSE, buf_pool_get_curr_size());
	}

	if (srv_force_recovery >= SRV_FORCE_NO_LOG_REDO) {
		fprintf(stderr,
			"InnoDB: The user has set SRV_FORCE_NO_LOG_REDO on\n");
		fprintf(stderr,
			"InnoDB: Skipping log redo\n");

		return(DB_SUCCESS);
	}

	recv_recovery_on = TRUE;

Vadim Tkachenko's avatar
Vadim Tkachenko committed
2619
	recv_sys->limit_lsn = LIMIT_LSN;
2620 2621 2622 2623 2624 2625 2626 2627 2628 2629 2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640 2641 2642 2643 2644 2645 2646 2647 2648 2649 2650 2651 2652 2653 2654 2655 2656 2657 2658 2659 2660 2661 2662 2663 2664 2665 2666 2667 2668 2669 2670 2671 2672 2673 2674 2675 2676 2677 2678 2679 2680 2681 2682 2683 2684 2685

	mutex_enter(&(log_sys->mutex));

	/* Look for the latest checkpoint from any of the log groups */

	err = recv_find_max_checkpoint(&max_cp_group, &max_cp_field);

	if (err != DB_SUCCESS) {

		mutex_exit(&(log_sys->mutex));

		return(err);
	}

	log_group_read_checkpoint_info(max_cp_group, max_cp_field);

	buf = log_sys->checkpoint_buf;

	checkpoint_lsn = mach_read_ull(buf + LOG_CHECKPOINT_LSN);
	checkpoint_no = mach_read_ull(buf + LOG_CHECKPOINT_NO);
	archived_lsn = mach_read_ull(buf + LOG_CHECKPOINT_ARCHIVED_LSN);

	/* Read the first log file header to print a note if this is
	a recovery from a restored InnoDB Hot Backup */

	fil_io(OS_FILE_READ | OS_FILE_LOG, TRUE, max_cp_group->space_id, 0,
	       0, 0, LOG_FILE_HDR_SIZE,
	       log_hdr_buf, max_cp_group);

	if (0 == ut_memcmp(log_hdr_buf + LOG_FILE_WAS_CREATED_BY_HOT_BACKUP,
			   (byte*)"ibbackup", (sizeof "ibbackup") - 1)) {
		/* This log file was created by ibbackup --restore: print
		a note to the user about it */

		fprintf(stderr,
			"InnoDB: The log file was created by"
			" ibbackup --apply-log at\n"
			"InnoDB: %s\n",
			log_hdr_buf + LOG_FILE_WAS_CREATED_BY_HOT_BACKUP);
		fprintf(stderr,
			"InnoDB: NOTE: the following crash recovery"
			" is part of a normal restore.\n");

		/* Wipe over the label now */

		memset(log_hdr_buf + LOG_FILE_WAS_CREATED_BY_HOT_BACKUP,
		       ' ', 4);
		/* Write to the log file to wipe over the label */
		fil_io(OS_FILE_WRITE | OS_FILE_LOG, TRUE,
		       max_cp_group->space_id, 0,
		       0, 0, OS_FILE_LOG_BLOCK_SIZE,
		       log_hdr_buf, max_cp_group);
	}

#ifdef UNIV_LOG_ARCHIVE
	group = UT_LIST_GET_FIRST(log_sys->log_groups);

	while (group) {
		log_checkpoint_get_nth_group_info(buf, group->id,
						  &(group->archived_file_no),
						  &(group->archived_offset));

		group = UT_LIST_GET_NEXT(log_groups, group);
	}
#endif /* UNIV_LOG_ARCHIVE */

Vadim Tkachenko's avatar
Vadim Tkachenko committed
2686
	if (TYPE_CHECKPOINT) {
2687 2688 2689 2690 2691 2692 2693 2694 2695 2696 2697 2698 2699 2700
		/* Start reading the log groups from the checkpoint lsn up. The
		variable contiguous_lsn contains an lsn up to which the log is
		known to be contiguously written to all log groups. */

		recv_sys->parse_start_lsn = checkpoint_lsn;
		recv_sys->scanned_lsn = checkpoint_lsn;
		recv_sys->scanned_checkpoint_no = 0;
		recv_sys->recovered_lsn = checkpoint_lsn;

		srv_start_lsn = checkpoint_lsn;
	}

	contiguous_lsn = ut_uint64_align_down(recv_sys->scanned_lsn,
					      OS_FILE_LOG_BLOCK_SIZE);
Vadim Tkachenko's avatar
Vadim Tkachenko committed
2701 2702 2703 2704 2705 2706
	if (TYPE_CHECKPOINT) {
		up_to_date_group = max_cp_group;
#ifdef UNIV_LOG_ARCHIVE
	} else {
		ulint	capacity;

2707 2708 2709 2710 2711 2712 2713 2714 2715 2716 2717 2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733 2734 2735 2736 2737 2738
		/* Try to recover the remaining part from logs: first from
		the logs of the archived group */

		group = recv_sys->archive_group;
		capacity = log_group_get_capacity(group);

		if (recv_sys->scanned_lsn > checkpoint_lsn + capacity
		    || checkpoint_lsn > recv_sys->scanned_lsn + capacity) {

			mutex_exit(&(log_sys->mutex));

			/* The group does not contain enough log: probably
			an archived log file was missing or corrupt */

			return(DB_ERROR);
		}

		recv_group_scan_log_recs(group, &contiguous_lsn,
					 &group_scanned_lsn);
		if (recv_sys->scanned_lsn < checkpoint_lsn) {

			mutex_exit(&(log_sys->mutex));

			/* The group did not contain enough log: an archived
			log file was missing or invalid, or the log group
			was corrupt */

			return(DB_ERROR);
		}

		group->scanned_lsn = group_scanned_lsn;
		up_to_date_group = group;
Vadim Tkachenko's avatar
Vadim Tkachenko committed
2739
#endif /* UNIV_LOG_ARCHIVE */
2740 2741 2742 2743 2744 2745
	}

	ut_ad(RECV_SCAN_SIZE <= log_sys->buf_size);

	group = UT_LIST_GET_FIRST(log_sys->log_groups);

Vadim Tkachenko's avatar
Vadim Tkachenko committed
2746
#ifdef UNIV_LOG_ARCHIVE
2747 2748 2749
	if ((type == LOG_ARCHIVE) && (group == recv_sys->archive_group)) {
		group = UT_LIST_GET_NEXT(log_groups, group);
	}
Vadim Tkachenko's avatar
Vadim Tkachenko committed
2750
#endif /* UNIV_LOG_ARCHIVE */
2751 2752

	/* Set the flag to publish that we are doing startup scan. */
Vadim Tkachenko's avatar
Vadim Tkachenko committed
2753
	recv_log_scan_is_startup_type = TYPE_CHECKPOINT;
2754 2755 2756 2757 2758 2759 2760 2761 2762 2763 2764 2765 2766
	while (group) {
		old_scanned_lsn = recv_sys->scanned_lsn;

		recv_group_scan_log_recs(group, &contiguous_lsn,
					 &group_scanned_lsn);
		group->scanned_lsn = group_scanned_lsn;

		if (old_scanned_lsn < group_scanned_lsn) {
			/* We found a more up-to-date group */

			up_to_date_group = group;
		}

Vadim Tkachenko's avatar
Vadim Tkachenko committed
2767
#ifdef UNIV_LOG_ARCHIVE
2768 2769 2770 2771
		if ((type == LOG_ARCHIVE)
		    && (group == recv_sys->archive_group)) {
			group = UT_LIST_GET_NEXT(log_groups, group);
		}
Vadim Tkachenko's avatar
Vadim Tkachenko committed
2772
#endif /* UNIV_LOG_ARCHIVE */
2773 2774 2775 2776 2777 2778

		group = UT_LIST_GET_NEXT(log_groups, group);
	}

	/* Done with startup scan. Clear the flag. */
	recv_log_scan_is_startup_type = FALSE;
Vadim Tkachenko's avatar
Vadim Tkachenko committed
2779
	if (TYPE_CHECKPOINT) {
2780 2781 2782 2783 2784 2785 2786 2787 2788 2789 2790 2791 2792 2793 2794 2795 2796 2797 2798 2799 2800 2801 2802 2803 2804 2805 2806 2807 2808 2809 2810 2811 2812 2813 2814 2815 2816 2817 2818 2819 2820 2821 2822 2823 2824 2825 2826 2827 2828 2829 2830 2831 2832 2833 2834 2835 2836 2837 2838 2839 2840 2841 2842 2843 2844 2845 2846 2847 2848 2849 2850 2851 2852 2853 2854 2855 2856
		/* NOTE: we always do a 'recovery' at startup, but only if
		there is something wrong we will print a message to the
		user about recovery: */

		if (checkpoint_lsn != max_flushed_lsn
		    || checkpoint_lsn != min_flushed_lsn) {

			if (checkpoint_lsn < max_flushed_lsn) {
				fprintf(stderr,
					"InnoDB: #########################"
					"#################################\n"
					"InnoDB:                          "
					"WARNING!\n"
					"InnoDB: The log sequence number"
					" in ibdata files is higher\n"
					"InnoDB: than the log sequence number"
					" in the ib_logfiles! Are you sure\n"
					"InnoDB: you are using the right"
					" ib_logfiles to start up"
					" the database?\n"
					"InnoDB: Log sequence number in"
					" ib_logfiles is %llu, log\n"
					"InnoDB: sequence numbers stamped"
					" to ibdata file headers are between\n"
					"InnoDB: %llu and %llu.\n"
					"InnoDB: #########################"
					"#################################\n",
					checkpoint_lsn,
					min_flushed_lsn,
					max_flushed_lsn);
			}

			if (!recv_needed_recovery) {
				fprintf(stderr,
					"InnoDB: The log sequence number"
					" in ibdata files does not match\n"
					"InnoDB: the log sequence number"
					" in the ib_logfiles!\n");
				recv_init_crash_recovery();
			}
		}

		if (!recv_needed_recovery) {
			/* Init the doublewrite buffer memory structure */
			trx_sys_doublewrite_init_or_restore_pages(FALSE);
		}
	}

	/* We currently have only one log group */
	if (group_scanned_lsn < checkpoint_lsn) {
		ut_print_timestamp(stderr);
		fprintf(stderr,
			"  InnoDB: ERROR: We were only able to scan the log"
			" up to\n"
			"InnoDB: %llu, but a checkpoint was at %llu.\n"
			"InnoDB: It is possible that"
			" the database is now corrupt!\n",
			group_scanned_lsn,
			checkpoint_lsn);
	}

	if (group_scanned_lsn < recv_max_page_lsn) {
		ut_print_timestamp(stderr);
		fprintf(stderr,
			"  InnoDB: ERROR: We were only able to scan the log"
			" up to %llu\n"
			"InnoDB: but a database page a had an lsn %llu."
			" It is possible that the\n"
			"InnoDB: database is now corrupt!\n",
			group_scanned_lsn,
			recv_max_page_lsn);
	}

	if (recv_sys->recovered_lsn < checkpoint_lsn) {

		mutex_exit(&(log_sys->mutex));

Vadim Tkachenko's avatar
Vadim Tkachenko committed
2857
		if (recv_sys->recovered_lsn >= LIMIT_LSN) {
2858 2859 2860 2861 2862 2863 2864 2865 2866 2867 2868 2869 2870 2871 2872 2873 2874 2875 2876 2877 2878 2879 2880 2881 2882 2883 2884 2885 2886 2887 2888 2889 2890 2891 2892 2893 2894 2895 2896 2897 2898 2899 2900 2901 2902 2903 2904 2905 2906 2907 2908 2909 2910 2911 2912 2913 2914 2915 2916 2917 2918 2919

			return(DB_SUCCESS);
		}

		ut_error;

		return(DB_ERROR);
	}

	/* Synchronize the uncorrupted log groups to the most up-to-date log
	group; we also copy checkpoint info to groups */

	log_sys->next_checkpoint_lsn = checkpoint_lsn;
	log_sys->next_checkpoint_no = checkpoint_no + 1;

#ifdef UNIV_LOG_ARCHIVE
	log_sys->archived_lsn = archived_lsn;
#endif /* UNIV_LOG_ARCHIVE */

	recv_synchronize_groups(up_to_date_group);

	if (!recv_needed_recovery) {
		ut_a(checkpoint_lsn == recv_sys->recovered_lsn);
	} else {
		srv_start_lsn = recv_sys->recovered_lsn;
	}

	log_sys->lsn = recv_sys->recovered_lsn;

	ut_memcpy(log_sys->buf, recv_sys->last_block, OS_FILE_LOG_BLOCK_SIZE);

	log_sys->buf_free = (ulint) log_sys->lsn % OS_FILE_LOG_BLOCK_SIZE;
	log_sys->buf_next_to_write = log_sys->buf_free;
	log_sys->written_to_some_lsn = log_sys->lsn;
	log_sys->written_to_all_lsn = log_sys->lsn;

	log_sys->last_checkpoint_lsn = checkpoint_lsn;

	log_sys->next_checkpoint_no = checkpoint_no + 1;

#ifdef UNIV_LOG_ARCHIVE
	if (archived_lsn == IB_ULONGLONG_MAX) {

		log_sys->archiving_state = LOG_ARCH_OFF;
	}
#endif /* UNIV_LOG_ARCHIVE */

	mutex_enter(&(recv_sys->mutex));

	recv_sys->apply_log_recs = TRUE;

	mutex_exit(&(recv_sys->mutex));

	mutex_exit(&(log_sys->mutex));

	recv_lsn_checks_on = TRUE;

	/* The database is now ready to start almost normal processing of user
	transactions: transaction rollbacks and the application of the log
	records in the hash table can be run in background. */

	return(DB_SUCCESS);
Vadim Tkachenko's avatar
Vadim Tkachenko committed
2920 2921 2922

#undef TYPE_CHECKPOINT
#undef LIMIT_LSN
2923 2924 2925 2926 2927 2928 2929 2930 2931 2932 2933 2934 2935 2936 2937 2938 2939 2940 2941 2942 2943 2944 2945 2946 2947 2948 2949 2950 2951 2952 2953 2954 2955 2956 2957 2958 2959 2960 2961 2962 2963 2964 2965 2966 2967 2968 2969 2970 2971 2972 2973 2974 2975 2976 2977 2978 2979 2980 2981 2982 2983 2984 2985 2986 2987 2988 2989 2990 2991 2992 2993 2994 2995 2996 2997 2998 2999 3000 3001 3002 3003 3004 3005 3006 3007 3008 3009 3010 3011 3012 3013 3014 3015 3016 3017 3018 3019 3020 3021 3022 3023 3024 3025 3026 3027 3028 3029 3030 3031 3032 3033 3034 3035 3036 3037 3038 3039 3040 3041 3042 3043 3044 3045 3046 3047 3048 3049 3050 3051 3052 3053 3054 3055 3056 3057 3058 3059 3060 3061 3062 3063 3064 3065 3066 3067 3068 3069 3070 3071 3072 3073 3074 3075 3076 3077 3078 3079 3080 3081 3082 3083 3084 3085 3086 3087 3088 3089 3090 3091 3092 3093 3094 3095 3096 3097 3098 3099 3100 3101 3102 3103 3104 3105 3106 3107 3108 3109 3110 3111 3112 3113 3114 3115 3116 3117 3118 3119 3120 3121 3122 3123 3124 3125 3126 3127 3128 3129 3130 3131 3132 3133 3134 3135 3136 3137 3138 3139 3140 3141 3142 3143 3144 3145 3146 3147 3148 3149 3150 3151 3152 3153 3154 3155 3156 3157 3158 3159 3160 3161 3162 3163 3164 3165 3166 3167 3168 3169 3170 3171 3172 3173 3174 3175 3176 3177 3178 3179 3180 3181 3182 3183 3184 3185 3186 3187 3188 3189 3190 3191 3192 3193 3194 3195 3196 3197 3198 3199 3200 3201 3202 3203 3204 3205 3206 3207 3208 3209 3210 3211 3212 3213 3214 3215 3216 3217 3218 3219 3220 3221 3222 3223 3224 3225 3226 3227 3228 3229 3230 3231 3232 3233 3234 3235 3236 3237 3238 3239 3240 3241 3242 3243 3244 3245 3246 3247 3248 3249 3250 3251 3252 3253 3254 3255 3256 3257 3258 3259 3260 3261 3262 3263 3264 3265 3266 3267 3268 3269 3270 3271 3272 3273 3274 3275 3276 3277 3278 3279 3280 3281 3282 3283 3284 3285 3286 3287 3288 3289 3290 3291 3292 3293 3294 3295 3296 3297 3298 3299 3300 3301 3302 3303 3304 3305 3306 3307 3308 3309 3310 3311 3312 3313 3314 3315 3316 3317 3318 3319 3320 3321 3322 3323 3324 3325 3326 3327 3328 3329 3330 3331 3332 3333 3334 3335 3336 3337 3338 3339 3340 3341 3342 3343 3344 3345 3346 3347 3348 3349 3350 3351 3352 3353 3354 3355 3356 3357 3358 3359 3360 3361 3362 3363 3364 3365 3366 3367 3368 3369 3370 3371 3372 3373 3374 3375 3376 3377 3378 3379 3380 3381 3382 3383 3384 3385 3386 3387 3388 3389 3390 3391 3392 3393 3394 3395 3396 3397 3398 3399 3400 3401 3402 3403 3404 3405 3406 3407 3408 3409 3410 3411 3412 3413 3414 3415 3416 3417 3418 3419 3420 3421 3422 3423 3424 3425 3426 3427 3428 3429 3430 3431 3432 3433 3434 3435 3436 3437 3438 3439 3440 3441 3442 3443 3444 3445 3446 3447 3448 3449 3450 3451 3452 3453 3454 3455 3456 3457 3458 3459 3460 3461 3462 3463 3464 3465 3466 3467 3468 3469 3470 3471 3472 3473 3474 3475 3476 3477 3478
}

/************************************************************
Completes recovery from a checkpoint. */
UNIV_INTERN
void
recv_recovery_from_checkpoint_finish(void)
/*======================================*/
{
	int		i;

	/* Apply the hashed log records to the respective file pages */

	if (srv_force_recovery < SRV_FORCE_NO_LOG_REDO) {

		recv_apply_hashed_log_recs(TRUE);
	}

#ifdef UNIV_DEBUG
	if (log_debug_writes) {
		fprintf(stderr,
			"InnoDB: Log records applied to the database\n");
	}
#endif /* UNIV_DEBUG */

	if (recv_needed_recovery) {
		trx_sys_print_mysql_master_log_pos();
		trx_sys_print_mysql_binlog_offset();
	}

	if (recv_sys->found_corrupt_log) {

		fprintf(stderr,
			"InnoDB: WARNING: the log file may have been"
			" corrupt and it\n"
			"InnoDB: is possible that the log scan or parsing"
			" did not proceed\n"
			"InnoDB: far enough in recovery. Please run"
			" CHECK TABLE\n"
			"InnoDB: on your InnoDB tables to check that"
			" they are ok!\n"
			"InnoDB: It may be safest to recover your"
			" InnoDB database from\n"
			"InnoDB: a backup!\n");
	}

	/* Free the resources of the recovery system */

	recv_recovery_on = FALSE;

#ifndef UNIV_LOG_DEBUG
	recv_sys_free();
#endif

	/* Drop partially created indexes. */
	row_merge_drop_temp_indexes();

#ifdef UNIV_SYNC_DEBUG
	/* Wait for a while so that created threads have time to suspend
	themselves before we switch the latching order checks on */
	os_thread_sleep(1000000);

	/* Switch latching order checks on in sync0sync.c */
	sync_order_checks_on = TRUE;
#endif
	if (srv_force_recovery < SRV_FORCE_NO_TRX_UNDO) {
		/* Rollback the uncommitted transactions which have no user
		session */

		os_thread_create(trx_rollback_or_clean_all_recovered,
				 (void *)&i, NULL);
	}
}

/**********************************************************
Resets the logs. The contents of log files will be lost! */
UNIV_INTERN
void
recv_reset_logs(
/*============*/
	ib_uint64_t	lsn,		/* in: reset to this lsn
					rounded up to be divisible by
					OS_FILE_LOG_BLOCK_SIZE, after
					which we add
					LOG_BLOCK_HDR_SIZE */
#ifdef UNIV_LOG_ARCHIVE
	ulint		arch_log_no,	/* in: next archived log file number */
#endif /* UNIV_LOG_ARCHIVE */
	ibool		new_logs_created)/* in: TRUE if resetting logs
					is done at the log creation;
					FALSE if it is done after
					archive recovery */
{
	log_group_t*	group;

	ut_ad(mutex_own(&(log_sys->mutex)));

	log_sys->lsn = ut_uint64_align_up(lsn, OS_FILE_LOG_BLOCK_SIZE);

	group = UT_LIST_GET_FIRST(log_sys->log_groups);

	while (group) {
		group->lsn = log_sys->lsn;
		group->lsn_offset = LOG_FILE_HDR_SIZE;
#ifdef UNIV_LOG_ARCHIVE
		group->archived_file_no = arch_log_no;
		group->archived_offset = 0;
#endif /* UNIV_LOG_ARCHIVE */

		if (!new_logs_created) {
			recv_truncate_group(group, group->lsn, group->lsn,
					    group->lsn, group->lsn);
		}

		group = UT_LIST_GET_NEXT(log_groups, group);
	}

	log_sys->buf_next_to_write = 0;
	log_sys->written_to_some_lsn = log_sys->lsn;
	log_sys->written_to_all_lsn = log_sys->lsn;

	log_sys->next_checkpoint_no = 0;
	log_sys->last_checkpoint_lsn = 0;

#ifdef UNIV_LOG_ARCHIVE
	log_sys->archived_lsn = log_sys->lsn;
#endif /* UNIV_LOG_ARCHIVE */

	log_block_init(log_sys->buf, log_sys->lsn);
	log_block_set_first_rec_group(log_sys->buf, LOG_BLOCK_HDR_SIZE);

	log_sys->buf_free = LOG_BLOCK_HDR_SIZE;
	log_sys->lsn += LOG_BLOCK_HDR_SIZE;

	mutex_exit(&(log_sys->mutex));

	/* Reset the checkpoint fields in logs */

	log_make_checkpoint_at(IB_ULONGLONG_MAX, TRUE);
	log_make_checkpoint_at(IB_ULONGLONG_MAX, TRUE);

	mutex_enter(&(log_sys->mutex));
}

#ifdef UNIV_HOTBACKUP
/**********************************************************
Creates new log files after a backup has been restored. */
UNIV_INTERN
void
recv_reset_log_files_for_backup(
/*============================*/
	const char*	log_dir,	/* in: log file directory path */
	ulint		n_log_files,	/* in: number of log files */
	ulint		log_file_size,	/* in: log file size */
	ib_uint64_t	lsn)		/* in: new start lsn, must be
					divisible by OS_FILE_LOG_BLOCK_SIZE */
{
	os_file_t	log_file;
	ibool		success;
	byte*		buf;
	ulint		i;
	ulint		log_dir_len;
	char		name[5000];
	static const char ib_logfile_basename[] = "ib_logfile";

	log_dir_len = strlen(log_dir);
	/* full path name of ib_logfile consists of log dir path + basename
	+ number. This must fit in the name buffer.
	*/
	ut_a(log_dir_len + strlen(ib_logfile_basename) + 11  < sizeof(name));

	buf = ut_malloc(LOG_FILE_HDR_SIZE + OS_FILE_LOG_BLOCK_SIZE);
	memset(buf, '\0', LOG_FILE_HDR_SIZE + OS_FILE_LOG_BLOCK_SIZE);

	for (i = 0; i < n_log_files; i++) {

		sprintf(name, "%s%s%lu", log_dir,
			ib_logfile_basename, (ulong)i);

		log_file = os_file_create_simple(name, OS_FILE_CREATE,
						 OS_FILE_READ_WRITE, &success);
		if (!success) {
			fprintf(stderr,
				"InnoDB: Cannot create %s. Check that"
				" the file does not exist yet.\n", name);

			exit(1);
		}

		fprintf(stderr,
			"Setting log file size to %lu %lu\n",
			(ulong) ut_get_high32(log_file_size),
			(ulong) log_file_size & 0xFFFFFFFFUL);

		success = os_file_set_size(name, log_file,
					   log_file_size & 0xFFFFFFFFUL,
					   ut_get_high32(log_file_size));

		if (!success) {
			fprintf(stderr,
				"InnoDB: Cannot set %s size to %lu %lu\n",
				name, (ulong) ut_get_high32(log_file_size),
				(ulong) (log_file_size & 0xFFFFFFFFUL));
			exit(1);
		}

		os_file_flush(log_file);
		os_file_close(log_file);
	}

	/* We pretend there is a checkpoint at lsn + LOG_BLOCK_HDR_SIZE */

	log_reset_first_header_and_checkpoint(buf, lsn);

	log_block_init_in_old_format(buf + LOG_FILE_HDR_SIZE, lsn);
	log_block_set_first_rec_group(buf + LOG_FILE_HDR_SIZE,
				      LOG_BLOCK_HDR_SIZE);
	sprintf(name, "%s%s%lu", log_dir, ib_logfile_basename, (ulong)0);

	log_file = os_file_create_simple(name, OS_FILE_OPEN,
					 OS_FILE_READ_WRITE, &success);
	if (!success) {
		fprintf(stderr, "InnoDB: Cannot open %s.\n", name);

		exit(1);
	}

	os_file_write(name, log_file, buf, 0, 0,
		      LOG_FILE_HDR_SIZE + OS_FILE_LOG_BLOCK_SIZE);
	os_file_flush(log_file);
	os_file_close(log_file);

	ut_free(buf);
}
#endif /* UNIV_HOTBACKUP */

#ifdef UNIV_LOG_ARCHIVE
/**********************************************************
Reads from the archive of a log group and performs recovery. */
static
ibool
log_group_recover_from_archive_file(
/*================================*/
					/* out: TRUE if no more complete
					consistent archive files */
	log_group_t*	group)		/* in: log group */
{
	os_file_t	file_handle;
	ib_uint64_t	start_lsn;
	ib_uint64_t	file_end_lsn;
	ib_uint64_t	dummy_lsn;
	ib_uint64_t	scanned_lsn;
	ulint		len;
	ibool		ret;
	byte*		buf;
	ulint		read_offset;
	ulint		file_size;
	ulint		file_size_high;
	int		input_char;
	char		name[10000];

	ut_a(0);

try_open_again:
	buf = log_sys->buf;

	/* Add the file to the archive file space; open the file */

	log_archived_file_name_gen(name, group->id, group->archived_file_no);

	file_handle = os_file_create(name, OS_FILE_OPEN,
				     OS_FILE_LOG, OS_FILE_AIO, &ret);

	if (ret == FALSE) {
ask_again:
		fprintf(stderr,
			"InnoDB: Do you want to copy additional"
			" archived log files\n"
			"InnoDB: to the directory\n");
		fprintf(stderr,
			"InnoDB: or were these all the files needed"
			" in recovery?\n");
		fprintf(stderr,
			"InnoDB: (Y == copy more files; N == this is all)?");

		input_char = getchar();

		if (input_char == (int) 'N') {

			return(TRUE);
		} else if (input_char == (int) 'Y') {

			goto try_open_again;
		} else {
			goto ask_again;
		}
	}

	ret = os_file_get_size(file_handle, &file_size, &file_size_high);
	ut_a(ret);

	ut_a(file_size_high == 0);

	fprintf(stderr, "InnoDB: Opened archived log file %s\n", name);

	ret = os_file_close(file_handle);

	if (file_size < LOG_FILE_HDR_SIZE) {
		fprintf(stderr,
			"InnoDB: Archive file header incomplete %s\n", name);

		return(TRUE);
	}

	ut_a(ret);

	/* Add the archive file as a node to the space */

	fil_node_create(name, 1 + file_size / UNIV_PAGE_SIZE,
			group->archive_space_id, FALSE);
#if RECV_SCAN_SIZE < LOG_FILE_HDR_SIZE
# error "RECV_SCAN_SIZE < LOG_FILE_HDR_SIZE"
#endif

	/* Read the archive file header */
	fil_io(OS_FILE_READ | OS_FILE_LOG, TRUE, group->archive_space_id, 0, 0,
	       LOG_FILE_HDR_SIZE, buf, NULL);

	/* Check if the archive file header is consistent */

	if (mach_read_from_4(buf + LOG_GROUP_ID) != group->id
	    || mach_read_from_4(buf + LOG_FILE_NO)
	    != group->archived_file_no) {
		fprintf(stderr,
			"InnoDB: Archive file header inconsistent %s\n", name);

		return(TRUE);
	}

	if (!mach_read_from_4(buf + LOG_FILE_ARCH_COMPLETED)) {
		fprintf(stderr,
			"InnoDB: Archive file not completely written %s\n",
			name);

		return(TRUE);
	}

	start_lsn = mach_read_ull(buf + LOG_FILE_START_LSN);
	file_end_lsn = mach_read_ull(buf + LOG_FILE_END_LSN);

	if (!recv_sys->scanned_lsn) {

		if (recv_sys->parse_start_lsn < start_lsn) {
			fprintf(stderr,
				"InnoDB: Archive log file %s"
				" starts from too big a lsn\n",
				name);
			return(TRUE);
		}

		recv_sys->scanned_lsn = start_lsn;
	}

	if (recv_sys->scanned_lsn != start_lsn) {

		fprintf(stderr,
			"InnoDB: Archive log file %s starts from"
			" a wrong lsn\n",
			name);
		return(TRUE);
	}

	read_offset = LOG_FILE_HDR_SIZE;

	for (;;) {
		len = RECV_SCAN_SIZE;

		if (read_offset + len > file_size) {
			len = ut_calc_align_down(file_size - read_offset,
						 OS_FILE_LOG_BLOCK_SIZE);
		}

		if (len == 0) {

			break;
		}

#ifdef UNIV_DEBUG
		if (log_debug_writes) {
			fprintf(stderr,
				"InnoDB: Archive read starting at"
				" lsn %llu, len %lu from file %s\n",
				start_lsn,
				(ulong) len, name);
		}
#endif /* UNIV_DEBUG */

		fil_io(OS_FILE_READ | OS_FILE_LOG, TRUE,
		       group->archive_space_id, read_offset / UNIV_PAGE_SIZE,
		       read_offset % UNIV_PAGE_SIZE, len, buf, NULL);

		ret = recv_scan_log_recs(
			TRUE, (buf_pool->n_frames - recv_n_pool_free_frames)
			* UNIV_PAGE_SIZE, TRUE, buf, len, start_lsn,
			&dummy_lsn, &scanned_lsn);

		if (scanned_lsn == file_end_lsn) {

			return(FALSE);
		}

		if (ret) {
			fprintf(stderr,
				"InnoDB: Archive log file %s"
				" does not scan right\n",
				name);
			return(TRUE);
		}

		read_offset += len;
		start_lsn += len;

		ut_ad(start_lsn == scanned_lsn);
	}

	return(FALSE);
}

/************************************************************
Recovers from archived log files, and also from log files, if they exist. */
UNIV_INTERN
ulint
recv_recovery_from_archive_start(
/*=============================*/
					/* out: error code or DB_SUCCESS */
	ib_uint64_t	min_flushed_lsn,/* in: min flushed lsn field from the
					data files */
	ib_uint64_t	limit_lsn,	/* in: recover up to this lsn if
					possible */
	ulint		first_log_no)	/* in: number of the first archived
					log file to use in the recovery; the
					file will be searched from
					INNOBASE_LOG_ARCH_DIR specified in
					server config file */
{
	log_group_t*	group;
	ulint		group_id;
	ulint		trunc_len;
	ibool		ret;
	ulint		err;

	ut_a(0);

	recv_sys_create();
	recv_sys_init(FALSE, buf_pool_get_curr_size());

	recv_recovery_on = TRUE;
	recv_recovery_from_backup_on = TRUE;

	recv_sys->limit_lsn = limit_lsn;

	group_id = 0;

	group = UT_LIST_GET_FIRST(log_sys->log_groups);

	while (group) {
		if (group->id == group_id) {

			break;
		}

		group = UT_LIST_GET_NEXT(log_groups, group);
	}

	if (!group) {
		fprintf(stderr,
			"InnoDB: There is no log group defined with id %lu!\n",
			(ulong) group_id);
		return(DB_ERROR);
	}

	group->archived_file_no = first_log_no;

	recv_sys->parse_start_lsn = min_flushed_lsn;

	recv_sys->scanned_lsn = 0;
	recv_sys->scanned_checkpoint_no = 0;
	recv_sys->recovered_lsn = recv_sys->parse_start_lsn;

	recv_sys->archive_group = group;

	ret = FALSE;

	mutex_enter(&(log_sys->mutex));

	while (!ret) {
		ret = log_group_recover_from_archive_file(group);

		/* Close and truncate a possible processed archive file
		from the file space */

		trunc_len = UNIV_PAGE_SIZE
			* fil_space_get_size(group->archive_space_id);
		if (trunc_len > 0) {
			fil_space_truncate_start(group->archive_space_id,
						 trunc_len);
		}

		group->archived_file_no++;
	}

	if (recv_sys->recovered_lsn < limit_lsn) {

		if (!recv_sys->scanned_lsn) {

			recv_sys->scanned_lsn = recv_sys->parse_start_lsn;
		}

		mutex_exit(&(log_sys->mutex));

		err = recv_recovery_from_checkpoint_start(LOG_ARCHIVE,
							  limit_lsn,
							  IB_ULONGLONG_MAX,
							  IB_ULONGLONG_MAX);
		if (err != DB_SUCCESS) {

			return(err);
		}

		mutex_enter(&(log_sys->mutex));
	}

	if (limit_lsn != IB_ULONGLONG_MAX) {

		recv_apply_hashed_log_recs(FALSE);

		recv_reset_logs(recv_sys->recovered_lsn, 0, FALSE);
	}

	mutex_exit(&(log_sys->mutex));

	return(DB_SUCCESS);
}

/************************************************************
Completes recovery from archive. */
UNIV_INTERN
void
recv_recovery_from_archive_finish(void)
/*===================================*/
{
	recv_recovery_from_checkpoint_finish();

	recv_recovery_from_backup_on = FALSE;
}
#endif /* UNIV_LOG_ARCHIVE */