Apply snapshot innodb-51-ss1644

Fixes:
- Bug #23710: crash_commit_before fails if innodb_file_per_table=1
- Bug #28254: innodb crash if shutdown during innodb_table_monitor is running
- Bug #28604: innodb_force_recovery restricts data dump
- Bug #29097: fsp_get_available_space_in_free_extents() is capped at 4TB
- Bug #29155: Innodb "Parallel recovery" is not prevented

storage/innobase/Makefile.am

@@ -55,7 +55,9 @@ noinst_HEADERS = include/btr0btr.h include/btr0btr.ic			\
 			include/ha0ha.ic include/hash0hash.h		\
 			include/hash0hash.ic include/ibuf0ibuf.h	\
 			include/ibuf0ibuf.ic include/ibuf0types.h	\
+			include/lock0iter.h				\
 			include/lock0lock.h include/lock0lock.ic	\
+			include/lock0priv.h include/lock0priv.ic	\
 			include/lock0types.h include/log0log.h		\
 			include/log0log.ic include/log0recv.h		\
 			include/log0recv.ic include/mach0data.h		\
@@ -129,7 +131,8 @@ libinnobase_a_SOURCES =	btr/btr0btr.c btr/btr0cur.c btr/btr0pcur.c	\
 			eval/eval0eval.c eval/eval0proc.c		\
 			fil/fil0fil.c fsp/fsp0fsp.c fut/fut0fut.c	\
 			fut/fut0lst.c ha/ha0ha.c ha/hash0hash.c		\
-			ibuf/ibuf0ibuf.c lock/lock0lock.c		\
+			ibuf/ibuf0ibuf.c lock/lock0iter.c		\
+			lock/lock0lock.c				\
 			log/log0log.c log/log0recv.c mach/mach0data.c	\
 			mem/mem0mem.c mem/mem0pool.c mtr/mtr0log.c	\
 			mtr/mtr0mtr.c os/os0file.c os/os0proc.c		\

storage/innobase/buf/buf0buf.c

@@ -903,8 +903,7 @@ buf_block_make_young(
 	/* Note that we read freed_page_clock's without holding any mutex:
 	this is allowed since the result is used only in heuristics */
 
-	if (buf_pool->freed_page_clock >= block->freed_page_clock
-				+ 1 + (buf_pool->curr_size / 4)) {
+	if (buf_block_peek_if_too_old(block)) {
 
 		mutex_enter(&buf_pool->mutex);
 		/* There has been freeing activity in the LRU list:
@@ -1648,6 +1647,15 @@ buf_page_init(
 
 	block->lock_hash_val	= lock_rec_hash(space, offset);
 
+#ifdef UNIV_DEBUG_VALGRIND
+	if (!space) {
+		/* Silence valid Valgrind warnings about uninitialized
+		data being written to data files.  There are some unused
+		bytes on some pages that InnoDB does not initialize. */
+		UNIV_MEM_VALID(block->frame, UNIV_PAGE_SIZE);
+	}
+#endif /* UNIV_DEBUG_VALGRIND */
+
 	/* Insert into the hash table of file pages */
 
 	if (buf_page_hash_get(space, offset)) {

storage/innobase/buf/buf0lru.c

@@ -244,7 +244,15 @@ buf_LRU_search_and_free_block(
 			frame at all */
 
 			if (block->frame) {
+				/* The page was declared uninitialized
+				by buf_LRU_block_remove_hashed_page().
+				We need to flag the contents of the
+				page valid (which it still is) in
+				order to avoid bogus Valgrind
+				warnings. */
+				UNIV_MEM_VALID(block->frame, UNIV_PAGE_SIZE);
 				btr_search_drop_page_hash_index(block->frame);
+				UNIV_MEM_INVALID(block->frame, UNIV_PAGE_SIZE);
 			}
 
 			ut_a(block->buf_fix_count == 0);
@@ -449,6 +457,7 @@ buf_LRU_get_free_block(void)
 		mutex_enter(&block->mutex);
 
 		block->state = BUF_BLOCK_READY_FOR_USE;
+		UNIV_MEM_ALLOC(block->frame, UNIV_PAGE_SIZE);
 
 		mutex_exit(&block->mutex);
 
@@ -864,6 +873,7 @@ buf_LRU_block_free_non_file_page(
 
 	block->state = BUF_BLOCK_NOT_USED;
 
+	UNIV_MEM_ALLOC(block->frame, UNIV_PAGE_SIZE);
 #ifdef UNIV_DEBUG
 	/* Wipe contents of page to reveal possible stale pointers to it */
 	memset(block->frame, '\0', UNIV_PAGE_SIZE);
@@ -871,6 +881,8 @@ buf_LRU_block_free_non_file_page(
 	UT_LIST_ADD_FIRST(free, buf_pool->free, block);
 	block->in_free_list = TRUE;
 
+	UNIV_MEM_FREE(block->frame, UNIV_PAGE_SIZE);
+
 	if (srv_use_awe && block->frame) {
 		/* Add to the list of mapped pages */
 
@@ -939,6 +951,7 @@ buf_LRU_block_remove_hashed_page(
 		    buf_page_address_fold(block->space, block->offset),
 		    block);
 
+	UNIV_MEM_INVALID(block->frame, UNIV_PAGE_SIZE);
 	block->state = BUF_BLOCK_REMOVE_HASH;
 }
 

storage/innobase/fsp/fsp0fsp.c

@@ -2829,7 +2829,7 @@ will be able to insert new data to the database without running out the
 tablespace. Only free extents are taken into account and we also subtract
 the safety margin required by the above function fsp_reserve_free_extents. */
 
-ulint
+ullint
 fsp_get_available_space_in_free_extents(
 /*====================================*/
 			/* out: available space in kB */
@@ -2895,7 +2895,8 @@ fsp_get_available_space_in_free_extents(
 		return(0);
 	}
 
-	return(((n_free - reserve) * FSP_EXTENT_SIZE)
+	return((ullint)(n_free - reserve)
+	       * FSP_EXTENT_SIZE
 	       * (UNIV_PAGE_SIZE / 1024));
 }
 

storage/innobase/handler/ha_innodb.cc

@@ -4533,17 +4533,16 @@ ha_innobase::position(
 /*********************************************************************
 If it's a DB_TOO_BIG_RECORD error then set a suitable message to
 return to the client.*/
-static
+inline
 void
 innodb_check_for_record_too_big_error(
 /*==================================*/
-	dict_table_t*	table,		/* in: table to check */
-	int		error)		/* in: error code to check */
+	ulint	comp,	/* in: ROW_FORMAT: nonzero=COMPACT, 0=REDUNDANT */
+	int	error)	/* in: error code to check */
 {
 	if (error == (int)DB_TOO_BIG_RECORD) {
-		ulint		max_row_size;
-
-		max_row_size = page_get_free_space_of_empty_noninline(table);
+		ulint	max_row_size
+			= page_get_free_space_of_empty_noninline(comp) / 2;
 
 		my_error(ER_TOO_BIG_ROWSIZE, MYF(0), max_row_size);
 	}
@@ -4657,9 +4656,7 @@ create_table_def(
 
 	error = row_create_table_for_mysql(table, trx);
 
-	/* We need access to the table and so we do the error checking
-	and set the error message here, before the error translation.*/
-	innodb_check_for_record_too_big_error(table, error);
+	innodb_check_for_record_too_big_error(flags & DICT_TF_COMPACT, error);
 
 	error = convert_error_code_to_mysql(error, NULL);
 
@@ -4783,9 +4780,8 @@ create_index(
 	sure we don't create too long indexes. */
 	error = row_create_index_for_mysql(index, trx, field_lengths);
 
-	/* We need access to the table and so we do the error checking
-	and set the error message here, before the error translation.*/
-	innodb_check_for_record_too_big_error(index->table, error);
+	innodb_check_for_record_too_big_error(form->s->row_type
+					      != ROW_TYPE_REDUNDANT, error);
 
 	error = convert_error_code_to_mysql(error, NULL);
 
@@ -4802,6 +4798,8 @@ int
 create_clustered_index_when_no_primary(
 /*===================================*/
 	trx_t*		trx,		/* in: InnoDB transaction handle */
+	ulint		comp,		/* in: ROW_FORMAT:
+					nonzero=COMPACT, 0=REDUNDANT */
 	const char*	table_name)	/* in: table name */
 {
 	dict_index_t*	index;
@@ -4810,13 +4808,11 @@ create_clustered_index_when_no_primary(
 	/* We pass 0 as the space id, and determine at a lower level the space
 	id where to store the table */
 
-	index = dict_mem_index_create((char*) table_name,
-		(char*) "GEN_CLUST_INDEX", 0, DICT_CLUSTERED, 0);
+	index = dict_mem_index_create(table_name, "GEN_CLUST_INDEX",
+				      0, DICT_CLUSTERED, 0);
 	error = row_create_index_for_mysql(index, trx, NULL);
 
-	/* We need access to the table and so we do the error checking
-	and set the error message here, before the error translation.*/
-	innodb_check_for_record_too_big_error(index->table, error);
+	innodb_check_for_record_too_big_error(comp, error);
 
 	error = convert_error_code_to_mysql(error, NULL);
 
@@ -4947,8 +4943,9 @@ ha_innobase::create(
 		order the rows by their row id which is internally generated
 		by InnoDB */
 
-		error = create_clustered_index_when_no_primary(trx,
-							norm_name);
+		error = create_clustered_index_when_no_primary(
+			trx, form->s->row_type != ROW_TYPE_REDUNDANT,
+			norm_name);
 		if (error) {
 			goto cleanup;
 		}
@@ -5873,9 +5870,9 @@ ha_innobase::update_table_comment(
 	mutex_enter_noninline(&srv_dict_tmpfile_mutex);
 	rewind(srv_dict_tmpfile);
 
-	fprintf(srv_dict_tmpfile, "InnoDB free: %lu kB",
-		   (ulong) fsp_get_available_space_in_free_extents(
-					prebuilt->table->space));
+	fprintf(srv_dict_tmpfile, "InnoDB free: %llu kB",
+		fsp_get_available_space_in_free_extents(
+			prebuilt->table->space));
 
 	dict_print_info_on_foreign_keys(FALSE, srv_dict_tmpfile,
 				prebuilt->trx, prebuilt->table);

storage/innobase/include/buf0buf.ic

@@ -28,7 +28,7 @@ buf_block_peek_if_too_old(
 	buf_block_t*	block)	/* in: block to make younger */
 {
 	return(buf_pool->freed_page_clock >= block->freed_page_clock
-	       + 1 + (buf_pool->curr_size / 1024));
+	       + 1 + (buf_pool->curr_size / 4));
 }
 
 /*************************************************************************

storage/innobase/include/fsp0fsp.h

@@ -245,7 +245,7 @@ will be able to insert new data to the database without running out the
 tablespace. Only free extents are taken into account and we also subtract
 the safety margin required by the above function fsp_reserve_free_extents. */
 
-ulint
+ullint
 fsp_get_available_space_in_free_extents(
 /*====================================*/
 			/* out: available space in kB */

storage/innobase/include/lock0lock.h

@@ -519,6 +519,18 @@ lock_is_table_exclusive(
 	dict_table_t*	table,	/* in: table */
 	trx_t*		trx);	/* in: transaction */
 /*************************************************************************
+Checks if a lock request lock1 has to wait for request lock2. */
+
+ibool
+lock_has_to_wait(
+/*=============*/
+			/* out: TRUE if lock1 has to wait for lock2 to be
+			removed */
+	lock_t*	lock1,	/* in: waiting lock */
+	lock_t*	lock2);	/* in: another lock; NOTE that it is assumed that this
+			has a lock bit set on the same record as in lock1 if
+			the locks are record locks */
+/*************************************************************************
 Checks that a transaction id is sensible, i.e., not in the future. */
 
 ibool
@@ -597,7 +609,7 @@ lock_validate(void);
 			/* out: TRUE if ok */
 /*************************************************************************
 Return approximate number or record locks (bits set in the bitmap) for
-this transaction. Since delete-marked records ma ybe removed, the
+this transaction. Since delete-marked records maybe removed, the
 record count will not be precise. */
 
 ulint

storage/innobase/include/mem0mem.ic

@@ -167,6 +167,8 @@ mem_heap_alloc(
 	mem_block_set_free(block, free + MEM_SPACE_NEEDED(n));
 
 #ifdef UNIV_MEM_DEBUG
+	UNIV_MEM_ALLOC(buf,
+		       n + MEM_FIELD_HEADER_SIZE + MEM_FIELD_TRAILER_SIZE);
 
 	/* In the debug version write debugging info to the field */
 	mem_field_init((byte*)buf, n);
@@ -177,8 +179,10 @@ mem_heap_alloc(
 
 #endif
 #ifdef UNIV_SET_MEM_TO_ZERO
+	UNIV_MEM_ALLOC(buf, n);
 	memset(buf, '\0', n);
 #endif
+	UNIV_MEM_ALLOC(buf, n);
 	return(buf);
 }
 
@@ -369,6 +373,8 @@ mem_heap_free_top(
 	if ((heap != block) && (mem_block_get_free(block)
 				== mem_block_get_start(block))) {
 		mem_heap_block_free(heap, block);
+	} else {
+		UNIV_MEM_FREE((byte*) block + mem_block_get_free(block), n);
 	}
 }
 

storage/innobase/include/page0page.h

@@ -531,6 +531,15 @@ page_get_free_space_of_empty(
 			/* out: free space */
 	ulint	comp)	/* in: nonzero=compact page format */
 		__attribute__((const));
+/*****************************************************************
+Calculates free space if a page is emptied. */
+
+ulint
+page_get_free_space_of_empty_noninline(
+/*===================================*/
+			/* out: free space */
+	ulint	comp)	/* in: nonzero=compact page format */
+		__attribute__((const));
 /****************************************************************
 Returns the sum of the sizes of the records in the record list
 excluding the infimum and supremum records. */

storage/innobase/include/row0mysql.h

@@ -460,19 +460,6 @@ row_check_table_for_mysql(
 					/* out: DB_ERROR or DB_SUCCESS */
 	row_prebuilt_t*	prebuilt);	/* in: prebuilt struct in MySQL
 					handle */
-/*************************************************************************
-Get the min of the maximum possible row sizes. */
-
-ulint
-page_get_free_space_of_empty_noninline(
-/*===================================*/
-					/* out: The (approx) maximum size
-					of a row, this is a conservative
-					estimate, since the size can be
-					slightly larger depending upon
-					the ROW_FORMAT setting.*/
-	dict_table_t*	table);		/* in: table for which max record
-					size required.*/
 
 /* A struct describing a place for an individual column in the MySQL
 row format which is presented to the table handler in ha_innobase.

storage/innobase/include/univ.i

@@ -83,6 +83,8 @@ memory is read outside the allocated blocks. */
 /* Make a non-inline debug version */
 
 #if 0
+#define UNIV_DEBUG_VALGRIND			/* Enable extra
+						Valgrind instrumentation */
 #define UNIV_DEBUG				/* Enable ut_ad() assertions */
 #define UNIV_LIST_DEBUG				/* debug UT_LIST_ macros */
 #define UNIV_MEM_DEBUG				/* detect memory leaks etc */
@@ -214,6 +216,8 @@ typedef __int64			ib_longlong;
 typedef longlong		ib_longlong;
 #endif
 
+typedef unsigned long long int	ullint;
+
 #ifndef __WIN__
 #if SIZEOF_LONG != SIZEOF_VOIDP
 #error "Error: InnoDB's ulint must be of the same size as void*"
@@ -298,5 +302,17 @@ typedef void* os_thread_ret_t;
 #include "ut0dbg.h"
 #include "ut0ut.h"
 #include "db0err.h"
+#ifdef UNIV_DEBUG_VALGRIND
+# include <valgrind/memcheck.h>
+# define UNIV_MEM_VALID(addr, size) VALGRIND_MAKE_MEM_DEFINED(addr, size)
+# define UNIV_MEM_INVALID(addr, size) VALGRIND_MAKE_MEM_UNDEFINED(addr, size)
+# define UNIV_MEM_FREE(addr, size) VALGRIND_MAKE_MEM_NOACCESS(addr, size)
+# define UNIV_MEM_ALLOC(addr, size) VALGRIND_MAKE_MEM_UNDEFINED(addr, size)
+#else
+# define UNIV_MEM_VALID(addr, size) do {} while(0)
+# define UNIV_MEM_INVALID(addr, size) do {} while(0)
+# define UNIV_MEM_FREE(addr, size) do {} while(0)
+# define UNIV_MEM_ALLOC(addr, size) do {} while(0)
+#endif
 
 #endif

storage/innobase/lock/lock0lock.c

@@ -6,10 +6,14 @@ The transaction lock system
 Created 5/7/1996 Heikki Tuuri
 *******************************************************/
 
+#define LOCK_MODULE_IMPLEMENTATION
+
 #include "lock0lock.h"
+#include "lock0priv.h"
 
 #ifdef UNIV_NONINL
 #include "lock0lock.ic"
+#include "lock0priv.ic"
 #endif
 
 #include "usr0sess.h"
@@ -319,42 +323,6 @@ ibool	lock_print_waits	= FALSE;
 /* The lock system */
 lock_sys_t*	lock_sys	= NULL;
 
-/* A table lock */
-typedef struct lock_table_struct	lock_table_t;
-struct lock_table_struct{
-	dict_table_t*	table;	/* database table in dictionary cache */
-	UT_LIST_NODE_T(lock_t)
-			locks;	/* list of locks on the same table */
-};
-
-/* Record lock for a page */
-typedef struct lock_rec_struct		lock_rec_t;
-struct lock_rec_struct{
-	ulint	space;		/* space id */
-	ulint	page_no;	/* page number */
-	ulint	n_bits;		/* number of bits in the lock bitmap */
-				/* NOTE: the lock bitmap is placed immediately
-				after the lock struct */
-};
-
-/* Lock struct */
-struct lock_struct{
-	trx_t*		trx;		/* transaction owning the lock */
-	UT_LIST_NODE_T(lock_t)
-			trx_locks;	/* list of the locks of the
-					transaction */
-	ulint		type_mode;	/* lock type, mode, LOCK_GAP or
-					LOCK_REC_NOT_GAP,
-					LOCK_INSERT_INTENTION,
-					wait flag, ORed */
-	hash_node_t	hash;		/* hash chain node for a record lock */
-	dict_index_t*	index;		/* index for a record lock */
-	union {
-		lock_table_t	tab_lock;/* table lock */
-		lock_rec_t	rec_lock;/* record lock */
-	} un_member;
-};
-
 /* We store info on the latest deadlock error to this buffer. InnoDB
 Monitor will then fetch it and print */
 ibool	lock_deadlock_found = FALSE;
@@ -400,20 +368,6 @@ lock_deadlock_recursive(
 				LOCK_MAX_DEPTH_IN_DEADLOCK_CHECK, we
 				return LOCK_VICTIM_IS_START */
 
-/*************************************************************************
-Gets the type of a lock. */
-UNIV_INLINE
-ulint
-lock_get_type(
-/*==========*/
-			/* out: LOCK_TABLE or LOCK_REC */
-	lock_t*	lock)	/* in: lock */
-{
-	ut_ad(lock);
-
-	return(lock->type_mode & LOCK_TYPE_MASK);
-}
-
 /*************************************************************************
 Gets the nth bit of a record lock. */
 UNIV_INLINE
@@ -611,8 +565,8 @@ UNIV_INLINE
 ulint
 lock_get_mode(
 /*==========*/
-			/* out: mode */
-	lock_t*	lock)	/* in: lock */
+				/* out: mode */
+	const lock_t*	lock)	/* in: lock */
 {
 	ut_ad(lock);
 
@@ -1017,7 +971,7 @@ lock_rec_has_to_wait(
 
 /*************************************************************************
 Checks if a lock request lock1 has to wait for request lock2. */
-static
+
 ibool
 lock_has_to_wait(
 /*=============*/
@@ -1098,7 +1052,7 @@ lock_rec_set_nth_bit(
 /**************************************************************************
 Looks for a set bit in a record lock bitmap. Returns ULINT_UNDEFINED,
 if none found. */
-static
+
 ulint
 lock_rec_find_set_bit(
 /*==================*/
@@ -1390,7 +1344,7 @@ lock_rec_copy(
 
 /*************************************************************************
 Gets the previous record lock set on a record. */
-static
+
 lock_t*
 lock_rec_get_prev(
 /*==============*/

storage/innobase/log/log0recv.c

@@ -57,6 +57,16 @@ ibool		recv_needed_recovery = FALSE;
 
 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. */
+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
@@ -99,6 +109,16 @@ the recovery failed and the database may be corrupt. */
 
 dulint	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. */
 
@@ -2284,6 +2304,23 @@ recv_scan_log_recs(
 
 		if (ut_dulint_cmp(scanned_lsn, recv_sys->scanned_lsn) > 0) {
 
+			/* 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 %lu %lu\n",
+					(ulong) ut_dulint_get_high(
+						recv_sys->scanned_lsn),
+					(ulong) ut_dulint_get_low(
+						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 */
@@ -2405,6 +2442,48 @@ recv_group_scan_log_recs(
 #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
@@ -2532,92 +2611,6 @@ recv_recovery_from_checkpoint_start(
 		recv_sys->recovered_lsn = checkpoint_lsn;
 
 		srv_start_lsn = checkpoint_lsn;
-
-		/* 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 (ut_dulint_cmp(checkpoint_lsn, max_flushed_lsn) != 0
-		    || ut_dulint_cmp(checkpoint_lsn, min_flushed_lsn) != 0) {
-
-			if (ut_dulint_cmp(checkpoint_lsn, max_flushed_lsn)
-			    < 0) {
-				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 %lu %lu, log\n"
-					"InnoDB: sequence numbers stamped"
-					" to ibdata file headers are between\n"
-					"InnoDB: %lu %lu and %lu %lu.\n"
-					"InnoDB: #########################"
-					"#################################\n",
-					(ulong) ut_dulint_get_high(
-						checkpoint_lsn),
-					(ulong) ut_dulint_get_low(
-						checkpoint_lsn),
-					(ulong) ut_dulint_get_high(
-						min_flushed_lsn),
-					(ulong) ut_dulint_get_low(
-						min_flushed_lsn),
-					(ulong) ut_dulint_get_high(
-						max_flushed_lsn),
-					(ulong) ut_dulint_get_low(
-						max_flushed_lsn));
-			}
-
-			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);
-			}
-
-			ut_print_timestamp(stderr);
-
-			fprintf(stderr,
-				"  InnoDB: Starting log scan"
-				" based on checkpoint at\n"
-				"InnoDB: log sequence number %lu %lu.\n",
-				(ulong) ut_dulint_get_high(checkpoint_lsn),
-				(ulong) ut_dulint_get_low(checkpoint_lsn));
-		} else {
-			/* Init the doublewrite buffer memory structure */
-			trx_sys_doublewrite_init_or_restore_pages(FALSE);
-		}
 	}
 
 	contiguous_lsn = ut_dulint_align_down(recv_sys->scanned_lsn,
@@ -2670,6 +2663,8 @@ recv_recovery_from_checkpoint_start(
 		group = UT_LIST_GET_NEXT(log_groups, group);
 	}
 
+	/* Set the flag to publish that we are doing startup scan. */
+	recv_log_scan_is_startup_type = (type == LOG_CHECKPOINT);
 	while (group) {
 		old_scanned_lsn = recv_sys->scanned_lsn;
 
@@ -2691,6 +2686,69 @@ recv_recovery_from_checkpoint_start(
 		group = UT_LIST_GET_NEXT(log_groups, group);
 	}
 
+	/* Done with startup scan. Clear the flag. */
+	recv_log_scan_is_startup_type = FALSE;
+	if (type == LOG_CHECKPOINT) {
+		/* 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 (ut_dulint_cmp(checkpoint_lsn, max_flushed_lsn) != 0
+		    || ut_dulint_cmp(checkpoint_lsn, min_flushed_lsn) != 0) {
+
+			if (ut_dulint_cmp(checkpoint_lsn, max_flushed_lsn)
+			    < 0) {
+				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 %lu %lu, log\n"
+					"InnoDB: sequence numbers stamped"
+					" to ibdata file headers are between\n"
+					"InnoDB: %lu %lu and %lu %lu.\n"
+					"InnoDB: #########################"
+					"#################################\n",
+					(ulong) ut_dulint_get_high(
+						checkpoint_lsn),
+					(ulong) ut_dulint_get_low(
+						checkpoint_lsn),
+					(ulong) ut_dulint_get_high(
+						min_flushed_lsn),
+					(ulong) ut_dulint_get_low(
+						min_flushed_lsn),
+					(ulong) ut_dulint_get_high(
+						max_flushed_lsn),
+					(ulong) ut_dulint_get_low(
+						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 (ut_dulint_cmp(group_scanned_lsn, checkpoint_lsn) < 0) {
 		ut_print_timestamp(stderr);
@@ -2747,20 +2805,9 @@ recv_recovery_from_checkpoint_start(
 	recv_synchronize_groups(up_to_date_group);
 
 	if (!recv_needed_recovery) {
-		if (ut_dulint_cmp(checkpoint_lsn, recv_sys->recovered_lsn)
-		    != 0) {
-			fprintf(stderr,
-				"InnoDB: Warning: we did not need to do"
-				" crash recovery, but log scan\n"
-				"InnoDB: progressed past the checkpoint"
-				" lsn %lu %lu up to lsn %lu %lu\n",
-				(ulong) ut_dulint_get_high(checkpoint_lsn),
-				(ulong) ut_dulint_get_low(checkpoint_lsn),
-				(ulong) ut_dulint_get_high(
-					recv_sys->recovered_lsn),
-				(ulong) ut_dulint_get_low(
-					recv_sys->recovered_lsn));
-		}
+		ut_a(ut_dulint_cmp(checkpoint_lsn, 
+				   recv_sys->recovered_lsn) == 0);
+
 	} else {
 		srv_start_lsn = recv_sys->recovered_lsn;
 	}

storage/innobase/mem/mem0mem.c

@@ -514,6 +514,7 @@ mem_heap_block_free(
 	mem_erase_buf((byte*)block, len);
 
 #endif
+	UNIV_MEM_FREE(block, len);
 
 	if (init_block) {
 		/* Do not have to free: do nothing */

storage/innobase/mem/mem0pool.c

@@ -229,6 +229,8 @@ mem_pool_create(
 
 		mem_area_set_size(area, ut_2_exp(i));
 		mem_area_set_free(area, TRUE);
+		UNIV_MEM_FREE(MEM_AREA_EXTRA_SIZE + (byte*) area,
+			      ut_2_exp(i) - MEM_AREA_EXTRA_SIZE);
 
 		UT_LIST_ADD_FIRST(free_list, pool->free_list[i], area);
 
@@ -300,6 +302,7 @@ mem_pool_fill_free_list(
 	UT_LIST_REMOVE(free_list, pool->free_list[i + 1], area);
 
 	area2 = (mem_area_t*)(((byte*)area) + ut_2_exp(i));
+	UNIV_MEM_ALLOC(area2, MEM_AREA_EXTRA_SIZE);
 
 	mem_area_set_size(area2, ut_2_exp(i));
 	mem_area_set_free(area2, TRUE);
@@ -400,6 +403,8 @@ mem_area_alloc(
 	mutex_exit(&(pool->mutex));
 
 	ut_ad(mem_pool_validate(pool));
+	UNIV_MEM_ALLOC(MEM_AREA_EXTRA_SIZE + (byte*)area,
+		       ut_2_exp(n) - MEM_AREA_EXTRA_SIZE);
 
 	return((void*)(MEM_AREA_EXTRA_SIZE + ((byte*)area)));
 }
@@ -482,6 +487,7 @@ mem_area_free(
 	}
 
 	size = mem_area_get_size(area);
+	UNIV_MEM_FREE(ptr, size - MEM_AREA_EXTRA_SIZE);
 
 	if (size == 0) {
 		fprintf(stderr,

storage/innobase/os/os0file.c

@@ -456,10 +456,9 @@ os_file_handle_error_no_exit(
 
 #undef USE_FILE_LOCK
 #define USE_FILE_LOCK
-#if defined(UNIV_HOTBACKUP) || defined(__WIN__) || defined(__FreeBSD__) || defined(__NETWARE__)
+#if defined(UNIV_HOTBACKUP) || defined(__WIN__) || defined(__NETWARE__)
 /* InnoDB Hot Backup does not lock the data files.
  * On Windows, mandatory locking is used.
- * On FreeBSD with LinuxThreads, advisory locking does not work properly.
  */
 # undef USE_FILE_LOCK
 #endif

storage/innobase/page/page0page.c

@@ -209,6 +209,18 @@ page_set_max_trx_id(
 	}
 }
 
+/*****************************************************************
+Calculates free space if a page is emptied. */
+
+ulint
+page_get_free_space_of_empty_noninline(
+/*===================================*/
+			/* out: free space */
+	ulint	comp)	/* in: nonzero=compact page format */
+{
+	return(page_get_free_space_of_empty(comp));
+}
+
 /****************************************************************
 Allocates a block of memory from an index page. */
 

storage/innobase/rem/rem0rec.c

@@ -753,7 +753,11 @@ rec_convert_dtuple_to_rec_old(
 	/* Calculate the offset of the origin in the physical record */
 
 	rec = buf + rec_get_converted_extra_size(data_size, n_fields);
-
+#ifdef UNIV_DEBUG
+	/* Suppress Valgrind warnings of ut_ad()
+	in mach_write_to_1(), mach_write_to_2() et al. */
+	memset(buf, 0xff, rec - buf + data_size);
+#endif /* UNIV_DEBUG */
 	/* Store the number of fields */
 	rec_set_n_fields_old(rec, n_fields);
 

storage/innobase/row/row0mysql.c

@@ -4059,25 +4059,3 @@ row_check_table_for_mysql(
 
 	return(ret);
 }
-
-/*************************************************************************
-Get the maximum row size. */
-
-ulint
-page_get_free_space_of_empty_noninline(
-/*===================================*/
-					/* out: The (approx) maximum size
-					of a row, this is a conservative
-					estimate, since the size can be
-					slightly larger depending upon
-					the ROW_FORMAT setting.*/
-	dict_table_t*	table)		/* in: table for which max record
-					size is required.*/
-{
-	ibool		compact;
-
-	compact = dict_table_is_comp(table);
-
-	return(page_get_free_space_of_empty(compact) / 2);
-}
-

storage/innobase/sync/sync0rw.c

@@ -15,16 +15,34 @@ Created 9/11/1995 Heikki Tuuri
 #include "mem0mem.h"
 #include "srv0srv.h"
 
+/* number of system calls made during shared latching */
 ulint	rw_s_system_call_count	= 0;
+
+/* number of spin waits on rw-latches,
+resulted during shared (read) locks */
 ulint	rw_s_spin_wait_count	= 0;
+
+/* number of OS waits on rw-latches,
+resulted during shared (read) locks */
 ulint	rw_s_os_wait_count	= 0;
 
+/* number of unlocks (that unlock shared locks),
+set only when UNIV_SYNC_PERF_STAT is defined */
 ulint	rw_s_exit_count		= 0;
 
+/* number of system calls made during exclusive latching */
 ulint	rw_x_system_call_count	= 0;
+
+/* number of spin waits on rw-latches,
+resulted during exclusive (write) locks */
 ulint	rw_x_spin_wait_count	= 0;
+
+/* number of OS waits on rw-latches,
+resulted during exclusive (write) locks */
 ulint	rw_x_os_wait_count	= 0;
 
+/* number of unlocks (that unlock exclusive locks),
+set only when UNIV_SYNC_PERF_STAT is defined */
 ulint	rw_x_exit_count		= 0;
 
 /* The global list of rw-locks */

storage/innobase/sync/sync0sync.c

@@ -115,6 +115,7 @@ ulint	mutex_system_call_count		= 0;
 
 /* Number of spin waits on mutexes: for performance monitoring */
 
+/* round=one iteration of a spin loop */
 ulint	mutex_spin_round_count		= 0;
 ulint	mutex_spin_wait_count		= 0;
 ulint	mutex_os_wait_count		= 0;

storage/innobase/trx/trx0sys.c

@@ -868,7 +868,16 @@ trx_sysf_create(
 		trx_sysf_rseg_set_page_no(sys_header, i, FIL_NULL, mtr);
 	}
 
-	/* The remaining area (up to the page trailer) is uninitialized. */
+	/* The remaining area (up to the page trailer) is uninitialized.
+	Silence Valgrind warnings about it. */
+	UNIV_MEM_VALID(sys_header + (TRX_SYS_RSEGS
+				     + TRX_SYS_N_RSEGS * TRX_SYS_RSEG_SLOT_SIZE
+				     + TRX_SYS_RSEG_SPACE),
+		       (UNIV_PAGE_SIZE - FIL_PAGE_DATA_END
+			- (TRX_SYS_RSEGS
+			   + TRX_SYS_N_RSEGS * TRX_SYS_RSEG_SLOT_SIZE
+			   + TRX_SYS_RSEG_SPACE))
+		       + page - sys_header);
 
 	/* Create the first rollback segment in the SYSTEM tablespace */
 	page_no = trx_rseg_header_create(TRX_SYS_SPACE, ULINT_MAX, &slot_no,

storage/innobase/trx/trx0trx.c

@@ -1570,19 +1570,21 @@ trx_commit_for_mysql(
 	the transaction object does not have an InnoDB session object, and we
 	set the dummy session that we use for all MySQL transactions. */
 
-	mutex_enter(&kernel_mutex);
-
 	if (trx->sess == NULL) {
 		/* Open a dummy session */
 
 		if (!trx_dummy_sess) {
-			trx_dummy_sess = sess_open();
+			mutex_enter(&kernel_mutex);
+
+			if (!trx_dummy_sess) {
+				trx_dummy_sess = sess_open();
+			}
+
+			mutex_exit(&kernel_mutex);
 		}
 
 		trx->sess = trx_dummy_sess;
 	}
-	
-	mutex_exit(&kernel_mutex);
 
 	trx_start_if_not_started(trx);
 

storage/innobase/ut/ut0mem.c

@@ -162,6 +162,8 @@ ut_malloc_low(
 #endif
 	}
 
+	UNIV_MEM_ALLOC(ret, n + sizeof(ut_mem_block_t));
+
 	((ut_mem_block_t*)ret)->size = n + sizeof(ut_mem_block_t);
 	((ut_mem_block_t*)ret)->magic_n = UT_MEM_MAGIC_N;